Mineral Processing and Extractive Metallurgy Review

An International Journal

ISSN: 0882-7508 (Print) 1547-7401 (Online) Journal homepage: http://www.tandfonline.com/loi/gmpr20

Effect of Seawater on Sulfide Ore Flotation: A

Review

Ricardo I. Jeldres, Liza Forbes & Luis A. Cisternas

To cite this article: Ricardo I. Jeldres, Liza Forbes & Luis A. Cisternas (2016) Effect of Seawater
on Sulfide Ore Flotation: A Review, Mineral Processing and Extractive Metallurgy Review, 37:6,
369-384, DOI: 10.1080/08827508.2016.1218871

To link to this article: http://dx.doi.org/10.1080/08827508.2016.1218871

Accepted author version posted online: 02

Aug 2016.
Published online: 02 Aug 2016.

Submit your article to this journal

Article views: 63

View related articles

View Crossmark data

Full Terms & Conditions of access and use can be found at

http://www.tandfonline.com/action/journalInformation?journalCode=gmpr20

Download by: [Universidad De Concepcion] Date: 28 November 2016, At: 11:36

MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW
2016, VOL. 37, NO. 6, 369–384
http://dx.doi.org/10.1080/08827508.2016.1218871

Effect of Seawater on Sulfide Ore Flotation: A Review

Ricardo I. Jeldresa,b, Liza Forbesc, and Luis A. Cisternasa,b
a
Department of Chemical Engineering & Mineral Process, Universidad de Antofagasta, Antofagasta, Chile; bCSIRO, Chile International Center of
Excellence, Las Condes, Santiago, Chile; cCSIRO, Process Science and Engineering, Clayton, Australia

ABSTRACT KEYWORDS
The flotation process in seawater is highly complex and multifaceted, wherein the chemistry is very Froth flotation; review;
different to that of pure water. The high saline environment compresses the electrical double layers saline water; seawater;
resulting in (i) enhancement of the floatability for surfaces that are already hydrophobic; (ii) mitigation of sulfide ores
the slime coatings; (iii) increase of the entrainment; (iv) reduction of the bubbles size; and (v) better
froths stability. In parallel, the secondary ions present in seawater cause a colloidal precipitation and a
strong buffering effect, which difficult the operation at high alkaline condition. The objective of this
review is to present a summary of the current knowledge on the subject of seawater flotation processes,
highlighting the copper sulphide ores flotation. This review includes a description of the underlying
flotation mechanisms affected by the presence of saline water and seawater, as well as a more practical
description of industrial flotation operations.

1. Introduction strength. In Las Luces,1 seawater is transported over a distance

of 7 km, mixed with tailing dam water and then used in
The Chilean Copper Commission (Cochilco) estimated that
grinding and flotation circuits. During the last 15 years, the
the total water used in Chilean copper mining operations is
increase in the total dissolved solid content of the process
likely to rise as high as 775 million m3 per annum by 2025, a
water in Las Luces was from 36.9 to 46.3 g/L (Moreno et al.
66% increase compared to the 2014 usage levels. Chile, as well
2011). Calculations suggest that this increase is largely due to
as many other countries whose economy relies heavily on the
solar evaporation where the evaporation rate reaches 50 m3/
mining industry (such as Australia, South Africa, China, and
ha/day. Another large copper plant using seawater in Chile is
India), is located in a highly arid region. Furthermore, the vast
Centinela Mine,2 This plant is processing ore at 95,000 tpd
majority of Chilean mining operations are located in the
using seawater without any pretreatment. Seawater is pumped
north of the country in the Atacama Desert, one of the driest
145 km from the Pacific Ocean to a 60,000 m3 pool at the
nonpolar places on earth. In such locations, high levels of
mine site located at an altitude of 2,300 m above sea level
water consumption carry a significant economic and environ-
(Castro 2012a).
mental impact.
The flotation process is highly complex, and the exact
In regions with low annual rainfall, the major sources of fresh
reasons for the differences in flotation performance in a saline
water are underground aquifers. Prolonged and extensive with-
environment are multifaceted. The objective of this review is
drawal of water from these reservoirs carries with it a significant
to present the summary of the current knowledge on the
risk of structural aquifer collapse. Such a collapse has the poten-
subject of the effect of the saline environment on the flotation
tial to have a catastrophic impact on the local flora and fauna.
processes, highlighting flotation of copper sulphide ores. This
The use of seawater in copper mining operations provides
review includes the description of the underlying flotation
a viable solution to the problem (Cisternas and Moreno 2014),
mechanisms affected by the presence of saline water, as well
but the history dates back to the 1930s, where small opera-
as a more practical description of the effect of salt water on
tions such as the Tocopilla (Chile) mine attempted to use
industrial flotation operations.
seawater to float chalcopyrite. Today, a great number of base
metal sulphide flotation plants use different types of saline
water (water with presence of electrolytes) including seawater. 2. Fundamentals
Some of them are summarized in Table 1.
2.1. Interparticle forces
The three nickel flotation plants (Mt Keith, Leinster mine,
and Kambalda Nickel Concentrator) in Western Australia Usually, the size of the particles is small enough to be unaf-
operated by BHP Billiton use bore water with high ionic fected by gravity and the behavior of suspensions is controlled

CONTACT Ricardo I. Jeldres jeldresresearch@gmail.com Department of Chemical Engineering & Mineral Process, Universidad de Antofagasta, Antofagasta,
Chile.
1
“Las Luces” is a copper–molybdenum plant located in Taltal, owned by the Las Cenizas Mining Group (Grupo Minero las Cenizas) of Chile.
2
“Centinela Mine” is located in Sierra Gorda, owned by Antofagasta Minerals S.A AMSA.
© 2016 Taylor & Francis
370 R. I. JELDRES ET AL.

Table 1. Base metal sulphide flotation plants using saline water (Drelich and significantly smaller than the particle radius (typically 1–10
Miller 2012).
nm; Israelachvili and Pashley 1983).
Project Company Country Water source
Batu Hijau Newmont Indonesia Sea and fresh water aAH
Las Luces Minera las Cenizas Chile Seawater
FVDW ¼ (2)
6H 2
Michilla Antofagasta Minerals Chile Seawater
KCGM Barrick/Newmont Australia Saline From equation 2 it can be seen that the van der Waals
Mt Keith BHP Billiton Australia Saline force between two identical bodies is always attractive (AH >
Raglan Xstrata Canada Saline
Texada Closed Canada Seawater 0). Furthermore, the van der Waals interaction between two
Tocopilla Closed Chile Seawater dissimilar bodies may be either attractive or repulsive depend-
Centinela Antofagasta Minerals Chile Seawater
ing upon the properties of the suspending medium.

2.1.3. Electrical double-layer forces

by the range and magnitude of the interparticles forces When a mineral particle is brought into contact with an
between solid surfaces. For example, a mineral pulp can be aqueous environment, it often acquires a substantial surface
dispersed by promoting a net repulsive interparticle force and charge. The mechanisms for oxide minerals involve the pro-
the particles form a suspension which is resistant to sedimen- tonation/deprotonation of surface hydroxyl sites (M-OH) via
tation. Then, the suspension will readily flow and even might the process:
be compressed to higher solids concentrations than if a net
OH OH
attractive interparticle force prevails. Otherwise, when the H2 O þ M  O $þ M  OH $þ M  OHþ
2
H H
attractive forces dominate, the particles begin to adhere and
+ –
form larger aggregates. This leads to higher settling rates and where H and OH are referred to as potential determining
lower fluidity of the pulps. ions. The presence of the charged surface promotes a redis-
In the simplest cases, the net interparticle force is governed tribution of ions in the surrounding solution, with counter-
by the sum of an attractive and a repulsive interaction force, ions preferentially attracted toward and co-ions repelled away
as defined by the DLVO theory. from the surface. The result is an ionic neutralization of the
surface charge over a short distance into the neighboring
2.1.1. DLVO theory solution in which the distribution of counter-ions and co-
The DLVO theory was independently developed by Derjaguin ions is uneven and this is referred to as the electrical double
and Landau (1941) and Verwey and Overbeek (1948). layer (Dukhin and Derjaguin 1976).
Although Derjaguin’s paper was published first, both sets of In this theory, the average distribution of charge and the
authors are given equal credit for the theory, as the interven- corresponding function for electrostatic potential are based on
ing years cover the duration of the Second World War. The the Poisson–Boltzmann equation (PBE; Israelachvili 1991).
theory states that the total force between two particles in an zeρ
z eψ
i
aqueous suspension is determined by the balance of the Ñ2 ψ ¼  0 exp  (3)
ε0 ε kT
attractive and repulsive forces, as outlined in equation 1.
In the equation 3, Ψ is the electrical potential, ε0 is the
FTot ¼ Fattractive  Frepulsive (1) permittivity of free space, ε is the dielectric constant of the
In the classic DLVO theory, the force responsible for the bulk solution, ρ0 in the number of density of ions of valency z
attraction is the van der Waals force, whereas the repulsive at the midplane, and k is the Boltzmann constant. The PBE
force is attributed to a combination of electrostatic and dou- was deduced by using a number of simplifications such as: (i)
ble-layer repulsion forces. However, the theory was later electrolytes are into an ideal solution with uniform dielectric
revised to include other forces such as the hydrophobic, properties; (ii) ions are point charges; and (iii) the potential of
hydration, or steric forces. The nature of these surface inter- mean force and the average electrostatic potential are identi-
actions has been described in detail by some authors (see, for cal. Furthermore, the PBE is applicable only to systems with
instance, Israelachvili 1991), and only a brief overview is next symmetrical electrolyte or a mixture of electrolytes with the
presented. same valence.
The double-layer interaction between surfaces or particles
2.1.2. van der Waals force of different geometries always decays exponentially with the
The van der Waals forces exist between atoms or molecules distance with a characteristic decay length equal to the Debye
and it can be divided into three groups: dipole–dipole force, length. This is quite different from van der Waals interaction
dipole–dipole induced force, and dispersion forces. The where the decay is a power law having very different expo-
attractive van de Waals force between the atoms is propor- nents for different geometries. Different expressions for the
tional to 1/r7, where r is the distance between them. van der Waals and electrostatic forces can be found in
However, the interactions between macroscopic bodies are Israelachvili (1991).
different, for example for two spherical particles, the force
can be expressed in terms of equation 2, where a is the 2.1.4. Hydrophobic attraction
particle radius, H is the distance between particles, and AH The strong inclination of water molecules to form H-bonds
is the Hamaker constant. The van der Waals attraction has a influences their interactions with immersed nonpolar mole-
short range and comes into play if the distance H is cules or surfaces. The effect of bringing the molecules together
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 371

is the reorientation or restructuring of the water molecules so hydrophobic force, which was represented by Yoon and
that they can participate in H-bonds formation more or less as Aksoy (1999) as a power law with the same form as for the
in bulk water. Closely related to this hydrophobic effect is the van der Waals force.
hydrophobic interaction, which describes the unusually long-
range attraction between hydrophobic molecules and surfaces
2.1.5. Hydration forces
in water (Kurihara et al., 1990). In order to minimize the
When two mineral particles interacting across an aqueous
water perturbation, the nonpolar additives tend to aggregate
medium are brought into close proximity, a short-range
in aqueous solution more than is expected from a considera-
repulsive force is often observed (Chan et al. 1978).
tion of the van der Waals forces alone. The result is a decrease
Traditionally, the appearance of this force has been associated
in the interaction area between the nonpolar moiety and its
with the presence of a structured layer of water molecules at
local aqueous environment. One of the most studied examples
the silica surface. The overlap of these hydrated layers of
of this behavior is the aggregation of simple surfactant mono-
different mineral surfaces give rise to short-range repulsions,
mers to form micelles, in which the mechanism involves the
commonly known as hydration forces. This assumption is
aggregation of hydrophobic portions of the monomers into a
supported by the very hydrophilic nature of some minerals
hydrocarbon core. However, some researchers proposes that
(Israelachvili and Wennerström 1996). Otherwise, Leckband
vapour–gas nanobubbles are, at least, partially responsible for
and Israelachvili (2001) suggested that the repulsive forces are
the macroscopic hydrophobic forces, either by forming adhe-
due to the entropic repulsion of molecules groups thermally
sion necks between adjacent hydrophobic surfaces or by
exited, which stick out from the surfaces.
assisting the propagation of the water perturbation (Ishida
The origin of this interaction is a very controversial issue,
et al. 2000; Meyer et al. 2006). In this sense, Craig et al.
but it is known that is a short range force, which decays
(1993a, 2003b) analyzed the impact of several electrolytes on
exponentially with the distance (Pashley 1982).
the coalescence rate of gas bubbles in water. They found that

x
bubble coalescence decreases with the increase of electrolyte
FH ðxÞ ¼ CH exp  (4)
concentration, suggesting that the hydrophobic force growths λ
with declining the electrolyte concentration. In the equation 4, FH is the hydration force, x is the
Van der Waals alone is not enough to counterbalance the separation distance between the surfaces, λ is the decay length,
hydrodynamic repulsion between air and bubbles. Therefore, and CH is a hydration constant. Some authors have used two
it is necessary to assume the presence of a stronger attractive decay lengths to get a better fit of the experimental results
force. Some Nanobubbles would only be expected when there (Grabbe and Horn 1993).
is a “supply” of air to the surface such as when the surface is
inserted into the liquid from air or when solvent conditions
are changed so that bubbles nucleate on the surface (Zhang
et al. 2008). 2.2. Effect of electrolyte concentration on the
Despite advances, no theory has yet been able to describe interparticle forces
all experimental data, and so the origin of hydrophobic attrac- The thickness of the double layer is termed the Debye length
tion remains as a matter of much conjecture. However, (κ–1) and it is a strong function of the ionic strength of the
Ravishankar and Yoon (1997) suggested that flotation recov- aqueous medium (I), as shown in equation 5 and Figure 1.
ery of minerals is closely related to the long-range The range of the electrical double layer and associated forces

Figure 1. Effect of solution ionic concentration on double layer thickness.

372 R. I. JELDRES ET AL.

can vary significantly depending on the electrolyte concentra-

tion in the bulk solution.

1
κ1 / pffiffi (5)
I
The surface charge of the particle is commonly approxi-
mated by the zeta potential (ζ), which is the charge on the
outer plane of the adsorbed layer, also known as the Stern
plane. As the double layer is compressed, the magnitude of
the zeta potential decreases, in a manner illustrated in
Figure 1.
Figure 2 shows the ζ of quartz as a function of both pH and
electrolyte content which demonstrates that as solution ionic
strength approaches 0.1 M, the zeta potential of quartz
approaches zero (seawater has the ionic strength of 0.6 M).
Overall, when mineral particles are present in a high saline
environment (e.g., seawater) these particles begin to behave as
if they do not have an electrical charge (Hogg et al. 1966).
Salts are not an exemption to the hydration phenomenon Figure 3. Interfacial water structure at KI, KCl, and NaCl surfaces. The stronger
because these strongly modify the structure of water structure making character of the Na+ cation leads to a more stable hydration
state at the NaCl surface (contact angle: θ = 0°) compared to that at the KCl
(Figure 3). The structure making/structure breaking attributes surface (θ = 8°). The instability of surface water is further promoted when both
of ions affect the mobility of the water molecules in solution. cation and anion serve as structure breakers as is the case for KI surface (θ = 25°;
Water molecules are in a constant state of Brownian motion Hencer et al. 2001).
and the time that a water molecule spends in equilibrium
position with respect to another molecule is defined as the
bonded due to the presence of structure making ions, it will
average retention time (τ). If ions are present in solution, their
be hard for collector molecules to reach the surface and be
interaction with water molecules is different from that
adsorbed. However, if those ions have a tendency to destroy
between the water molecules themselves, and this causes a
the structure of water and create a condition for the adsorp-
shift in the equilibrium, resulting in a different retention time
tion of collector, the flotation of soluble salt minerals will
(τi). Ions with small ionic radii (such as Li+ and Na+) as well
become feasible. The interaction between the additives and
as polyvalent ions, exhibit structure making properties, where
the solid surface in a high saline environment is difficult to
τi > τ. These ions are referred to as structure making or
understand, and Jeldres et al. (2014) showed that the impact
kosmotropes. Large monovalent ions with low levels of hydra-
of the type of salt strongly depends on the charge type of the
tion, (such as Cs+ and K+) exhibit structure breaking proper-
surface and also the electrical charge of the reagents.
ties, where τi < τ. These ions are known as chaotropes.
Hencer et al. (2001) studied the flotation of soluble salts
and they found that if a collector adsorbs at the salt interface, 3. Saline water and flotation subprocesses
it would displace interfacial water or penetrate through the
3.1. Particle–bubble interactions
water structure. If the structure of water is strongly hydrogen
Air bubbles in aqueous medium behave similarly to solid
particles. It has been widely demonstrated that air bubbles
carry a strong negative charge where the exact magnitudes of
the charge estimates is variable but there is an overall agree-
ment between researchers that the charge is negative in the
range of pH above pH 2 (Oliveira and Rubio 2011).
Consequently, when a bubble is present in solution an elec-
trical double layer forms around its surface. If one bubble and
one particle come within close proximity to one another, their
respective electrical double layers begin to interact, giving rise
to a long range double layer repulsion. This repulsion then
acts as an energy barrier that needs to be overcome with a
balance of attractive forces and kinetic energy. This principle
is illustrated in Figure 4 (Fuerstenau et al. 1983). The figure
represents the relationship between the disjoining pressures
(π) as a function of separation distance between a bubble and
a particle surfaces ðxÞ. Referring to Figure 4, there are three
type of behaviors in disjoining pressure: (i) if ð@πD =@xÞT is
Figure 2. Zeta potential of quarts as a function of electrolyte concentration (Li negative (curve A), the film is stable. As the film thins, the
and De Bruyn 1966). disjoining pressure increases maintaining the film in
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 373

Figure 5. Flotation response of methylated quartz (θ = 53°) at pH = 6.1–6.5, as a

function of solution salinity (Laskowski et al. 1991).

barrier opposing the attachment. The experiments were car-

ried out over the pH range (6.1–6.5) where ζ of the methy-
Figure 4. Bubble disjoining pressure (π) as a function of separation distance ðxÞ lated quartz particles was in the range of –35 and –40 mV.
between bubble and particle surfaces (Fuerstenau et al. 1983). The rate of flotation process carried out under such condi-
tions clearly depends on the electrolyte concentration and the
correlation of the flotation rate and energy barrier is quite
equilibrium. This would be the case of an aqueous wetting good. Furthermore, the increase in solution salinity decreases
film on a hydrophilic solid as mineral gangue including silica the degree of solubility of gas in that solution. This can
particles or clays; (ii) for hydrophobic surfaces, ð@πD =@xÞT is potentially cause air precipitation to occur on mineral sur-
positive (curve B), and the film is unstable. The net result is a faces, giving rise to nanobubbles (Zhang et al. 2008). The
spontaneous rupturing of the film; and (iii) most mineral presence of nanobubbles is said to enhance the hydrophobic
systems, however, show a curve located between curve A interaction between bubbles and particles, further improving
and curve B. This is because a competing forces appear. For particle–bubble attachment (Calgaroto et al. 2014).
example, long-range electrical double-layer repulsive forces
may succeed in making the total disjoining pressure positive
up to a point where the film thickness reaches a critical value, 3.2. Particle–Particle Interactions
and the van der Waals forces start to dominate. This is the
situation illustrated in curve C. The rupture occurs sponta- A further consequence of the presence of high concentrations
neously once the maximum on curve C is reached, for then of electrolyte in the flotation medium is its effect on particle–
ð@πD =@xÞT become positive. The mechanism of particle-bub- particle interaction within the flotation pulp. The two phe-
nomena that are the most affected by this are the entrainment
ble attachment can be divided into three stages: (1) particle-
and slimes coatings.
bubble approach, (2) thinning and rupture of the disjoining
film, and (3) formation of a stable particle-bubble aggregate
capable of withstanding the considerable disruptive turbules- 3.2.1. Entrainment
cence in a flotation cell. Entrainment has been described as a nonselective process,
As described in previous sections, the high electrolyte whereby hydrophilic gangue particles report to the froth
content of seawater has the effect of completely compressing phase by being upswept in the wake of rising bubbles, while
the electrical double layers. This causes that the attractive van not being attached to them (Gaudin 1957). The effect of
der Waals and non-DLVO forces to become dominant in the particle size in the entrainment has been widely studied and
interaction between bubbles and particles, making the attach- it has been shown that the entrain ability of mineral particles
ment more likely. This means that some particles with low increases dramatically at particle sizes below approximately 20
hydrophobicity which would previously float poorly due to m (Neethling and Cilliers 2002).
the presence of the energy barrier would become highly floa- In the case of hydrophilic slime particles, the balance of
table (Fuertenau et al. 1983; Lucay et al. 2015). Laskowski interparticle forces comprises of attractive van der Waals
et al. (1991) have demonstrated this using quartz made mildly force and repulsive double-layer force. Because the parti-
hydrophobic by methylation as a function of increasing sali- cles are hydrophilic, the attractive hydrophobic force is
nity of the flotation medium (Figure 5). Flotation rate does not present. The compression of the electrical double
not only depend on the hydrophobicity of the particles but layer in the flotation pulp results in largely eliminating
also—because these particles carry electrical charge—the par- double-layer repulsion between the mineral particles, mak-
ticle-to bubble attachment, which depends on the energy ing them behave as if they were uncharged. This means
374 R. I. JELDRES ET AL.

that in a saline environment, the dominant interparticle

force is the van der Waals attraction, resulting in coagula-
tion of fine mineral particles. The degree of coagulation
can be reliably measured by settling tests, as large particle
aggregates have more rapid settling speeds than small
individual particles in accordance to Stokes’s Law. In this
sense, Peng and Bradshaw (2012) have recently demon-
strated that the presence of saline water in the flotation
pulp has significantly reduced the settling rate of lizardite
suspensions. They also observed that the recovery of lizar-
dite particles from lizardite/pentlandinte mixtures was sig-
nificantly reduced in the presence of saline water. This
improvement was partially attributed to the reduction in
the entrainment of fine lizardite particles through its
Figure 6. Effect of saline bore water on pentlandite flotation and lizardite
coagulation. rejection: 10% pentlandite mixed with 90% lizardite (Peng and Bradshaw 2012).

3.2.2. Slimes coatings

The detrimental effect of “slimes coatings” on flotation is a respective liquid films come into contact with one another
phenomenon wherein relatively coarse valuable particles and become destabilized. This destabilization causes the thin-
become either completely or partially coated by a layer of ning of the adjoining films, ultimately followed by film rup-
hydrophilic slimes, inhibiting the collector adsorption and ture and subsequent bubble coalescence. However, the
rendering the valuables hydrophilic (Del Guidice 1934). This flotation operations are greatly benefit from a reduction in
mechanism has been widely studied and the consensus among bubble size, through the increase in the probability of parti-
the vast majority of researchers is that slime coatings arise due cle–bubble collisions. For this reason, surfactant reagents
to electrical double layer forces between oppositely charged of (frothers) are commonly added to flotation systems.
mineral particles, a phenomenon known as “heterocoagula-
tion” (Fuerstenau et al. 1958; Xu et al. 2003). A good example 3.3.1. Effect of flotation frothers
of this is the attachment of fine lizardite to pentlandite sur- One of the actions of a frothing agent is to adsorb at the air–
faces (Peng and Seaman 2011; Peng and Bradshaw 2012). water interface and stabilize the liquid film via a process called
As it was seen in previous sections, the major effect of the the Marangoni effect (Klassen and Makrousov 1963). The
presence of saline water in flotation pulp is the compression presence of frothers reduces the coalescence of bubbles and
of the electrical double layers around particles, which serves to prevent it completely when they exceed a particular concen-
retard not only the repulsive forces but also attractive ones, tration, referred to as the critical coalescence concentration
making the particles behave as if they are charge neutral. This (CCC; Cho and Laskowski 2002a, 2002b; Laskowski et al.
helps to mitigate the effect of slimes coatings by removing the 2003; Grau and Laskowski 2006).
additional attractive force between oppositely charged sul- The concept of using the CCC to characterize frothers was
phide and gangue particles. It was demonstrated by Peng proposed by Cho and Laskowski (2002a, 2002b), which is
and Bradshaw (2012) who showed that the recovery of pen- becoming the accepted standard. The CCC is calculated by
tlandite is vastly improved in the presence of highly saline evaluating the Sauter mean diameter (D32) from the measured
bore water, whereas the recovery of fine lizardite is signifi- bubble size distribution and plotting it as a function of the
cantly reduced, as shown in Figure 6. Furthermore, Zhang frother concentration (C). Then, the D32–C curve is charac-
et al. (2015) found that bentonite has a deleterious effect on terized by an initial sharp decrease in D32, which levels off to
copper and gold flotation in tap water, but this detrimental reach a minimum that defines the CCC, the concentration at
consequence was mitigated in seawater. which the coalescence is considered to be fully retarded.
In resume, it has been found that seawater might have a Although Cho and Laskowski (2002a) used a graphical
beneficial effect on one of the main challenges of the mineral method to estimate the CCC, Nesset et al. (2007) suggested
flotation since it may decrease the effect of slime coatings. to fit the D32–C data to the following three-parameter model
(equation 6), wherein Dl is the limiting D32 as concentration
3.3. Bubble size tends toward infinity, A is the difference between Dl and D32
in water (no solute present), B is the decay constant, and C
Flotation kinetics involves a number of mass transfer pro- solute concentration:
cesses with some taking place in the pulp phase (particle–
bubble collision and attachment, transport of particle–bubble D32 ¼ Dl þ A exp½BC (6)
aggregate to the froth phase) and some in the froth phase The model allows for calculation of CCCX, the concentra-
(recovery of particle from the froth phase to concentrate tion at which D32 is reduced by X% from that in water to the
launder). All of these subprocesses depend strongly on bubble Dl , as shown in equation 7:
size.
Usually, an interfacial liquid film coats the air bubbles in lnð1  xÞ
CCCX ¼  (7)
an aqueous medium. When two bubbles come together, their B
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 375

Figure 7. Sauter bubble diameter as function of frother concentration and Figure 9. Sauter mean diameter D32 as function of salt concentration. CCC95
graphical determination of the critical coalescence concentration (Castro et al. values indicated (Quinn et al. 2014).
2013).

Different mechanisms have been proposed to explain the

The CCC has been determined for many commercial stabilizing effect of salts on bubble coalescence, however as
frothers (Cho and Laskowski 2002a, 2002b; Grau et al. yet there is no consensus. Lessard and Zieminski (1971)
2005), with four examples shown in Figure 7 (Castro et al. investigated the effects of inorganic electrolytes, AlCl3,
2013). MgSO4, Na2SO4, CaCl2, MgCl2, NaCl, LiCl, and NaBr, on
Recently, Zhang et al. (2014) examined the effect of frother the bubble coalescence and interfacial gas transfer in aqu-
on bubble size using the concept of frother partitioning, eous solution. The coalescence experiments consisted of con-
which is expressed as the ratio of frother concentration tacting a number of pairs of bubbles and evaluating the
between the froth/water interface and the bulk liquid coalescence percentage as a function of the solute concentra-
(Cinterface/Cbulk). They found that a strong relationship exists tion. The concentration resulting in 50% coalescence was
between bubble size (D32) and the Cinterface/Cbulk ratio, sharply reduced when the size of ions decreased. These
demonstrating that the largest bubble size reduction occurred concentrations correlated well with ionic entropy of solution
when the concentration ratio was at its maximum. and with the self-diffusion ability of water in solution, where
both parameters increase with breaker ions.
3.3.2. Effect of salts Craig et al. (1993a, 1993b) found that bubble coalescence is
The interfacial air–liquid films can be stabilized in presence of inhibited by some salts at moderate concentrations (0.1 M)
certain organic and inorganic salts, which inhibit the bubble whereas others have no effect. It was found that the coales-
coalescence and increase the gas holdup in flotation systems cence behavior is not determined by either the anion or the
(Quinn et al. 2007; Kracht and Finch 2010). The CCC concept cation but rather the pair of ions present. In other words both
used for alcohol-based frothers can be transferred to the salts ions are important. The frothing properties of salts were
as the same general shape of the D32–C curve as is shown in therefore classified by empirically assigning a property termed
Figure 8 for NaCl solutions. Also, Quinn et al. (2014) deter- either α or β, in such a way that an αα or ββ combination
mined the CCC for a series of coalescence inhibiting inor- resulted in an inhibition of bubble coalescence compared to
ganic salts (KCl, NaCl, Na2SO4, CaCl2, and MgSO4; Figure 9). that of water, whereas either an αβ or βα combination left the
bubble coalescence unchanged. The resultant classification of
a number of ion combinations is summarized in Figure 10. At
the time of their proposal, the parameters α and β were
assigned empirically and nothing was known of their physical
meaning. For example, they do not correlate with either ion
size, polarizability, geometry, solution surface tension, or
viscosity. However, Henry and Craig (2010) reported that
the α and β designations correspond to the ion distribution
between bulk and interfacial regions at the air–water surface
(Figure 11), a property of which the Cinterface/Cbulk ratio
(introduced by Zhang et al (2014) for frothers) is reminiscent.
It was shown that

(i) α cations and α anions are both excluded from the

interface;
(ii) β cations and β anions are both accumulated at the
Figure 8. Sauter bubble diameter as a function of the concentration of NaCl
solutions (Castro et al. 2013).
interface;
376 R. I. JELDRES ET AL.

Figure 10. Effect of a range of electrolytes on bubble coalescence. A tick indicates that the salt inhibits bubble coalescence and cross that no bubble coalescence
inhibition has been observed. Properties α and β have been assigned empirically and can be used with the combining rules to predict the bubble coalescence
behavior of an anion–cation combination (Henry and Craig 2010).

Regardless of the mechanism, it has been demonstrated

that in highly concentrated salt solutions (such as seawater),
the size of the air bubbles remains small in the absence of the
frothing agent, compared to deionized water. Hence, flotation
in saline water meets all the flotation process requirements:

(i) In the environment of high ionic strength, the energy

barrier opposing attachment of the hydrophobic par-
ticles to bubbles is reduced making attachment
possible.
(ii) At the same time, fine bubbles are generated under
such conditions.

Thus, in flotation systems with sufficiently high salt concen-

Figure 11. Schematic of ion partitioning and ion assignments from bubble
trations (ionic strength > 0.3), frother addition is not required
coalescence (Henry and Craig 2010). to produce small bubbles. At least one plant, Xstrata’s Raglan
concentrator, operates without frother but there are other
situations of high salt content where frother is added. Quinn
et al. (2007) found that the bubble size for a salt solution with
(iii) αα and ββ electrolytes inhibit coalescence; and ionic strength of 0.4 (0.4 M NaCl), as at the Raglan concen-
(iv) αβ and βα electrolytes are less inhibiting. trator, is similar to that of ca. 10 ppm MIBC, which is a typical
dosage for this common frother. The result was similar in two-
Ion surface affinities were presented in terms of an ion phase (solution-air) and three-phase (slurry-air) tests. The
surface partition coefficient, Kp,i, which can be thought as a result assists with the interpretation of why the Raglan plat
ratio of surface to bulk concentration. Kp,i was obtained by can operate efficiently without frother addition.
using the solute partitioning model (SPM) proposed by Castro et al. (2010, 2012b) have started to examine the
Pregram and Record (2007). It is interesting to note that the interaction between inorganic salts and frother which pro-
ordering of ions given by the partition coefficients of Pegram mises to be a useful area of study to understand why in
and Record approximately follow the Hofmeister series, but some cases frother is still added. Furthermore, the presence
with some clear exceptions. This may lead one to believe that of the frothing agent does not reduce the bubble size further
ion hydration properties may be instrumental in determining as depicted in Figure 12 (Castro et al. 2010). The concept of
the frothing properties of salts. Otherwise, Craig (2011), and the surface tension switch point (s.t.s.p.) was introduced to
more recently Firouzi and Nguyen (2014), postulated that characterize the effect of electrolyte concentration on frother
hydration forces play no role in the specific ion effects properties. The s.t.s.p corresponds to a single concentration of
observed in bubble coalescence as coalescence is governed surfactant, on the surface tension isotherm at witch a surfac-
by dynamic processes. However, the specific ion effects tant becomes more surface active while in the presence of an
observed are a manifestation of the particular arrangement electrolyte. The value for MIBC was determined to be around
of ions in the interface which ultimately is strongly influenced 120 ppm in NaCl electrolyte solutions (Figure 13).
by ion hydration. Thus ion hydration (but not the hydration
force) is critical to specific ion effects in bubble coalescence, as
3.4. Froth stability
it strongly influences the positioning of ions in the interfacial
region which ultimately determines whether the coalescence is The stability of flotation froths is an important property, which
inhibited. Importantly, seawater is dominated by maker salts has a strong effect on the overall flotation performance through
as sodium, calcium, and magnesium. different phenomena such as froth carrying capacity, entrainment,
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 377

phase systems (air–liquid–gas). Such froths not only depen-

dent on the presence and properties of the frothing agents,
but also on the presence and properties of mineral particles.
In particular, size, shape, and hydrophobicity/hydrophilicity
play a key role in determining the degree to which the parti-
cles stabilize froths (Klassen and Makrousov 1963; Johansson
and Pugh 1992). Despite this, there are still many disagree-
ments within the literature concerning the exact magnitudes
of these particle properties and their respective effects. It is
also not clear whether the effect of particles is more or less
dominant than the effect of frothing agents in determining
froth stability.

4. Seawater chemistry

Figure 12. Effect of MIBC frother on bubble size in seawater (Castro et al. 2010).
Seawater contains many ionic species (Table 2) that have a
strong effect on flotation mainly due to two phenomena: (a)
the buffering effect and (b) secondary ion effects.

4.1. Buffering effects

Lime is usually used in the copper mining industry to adjust
the flotation pulp pH to approximately 11. This is done to
depress pyrite and improve the recovery of molybdenum.
However, the immediate effect of performing flotation in
seawater is the dramatic increase in lime required to obtain
the conventional flotation pH regime. This is because the
ionic content of seawater gives it strong buffering properties,
which arises from the presence of carbonate/bicarbonate ion
pairs (HCO3–/CO3–2), as well as boric acid/borate ion pairs (B
(OH)3/B(OH)4–; Table 2). These ionic pairs engage in equili-
brium reactions with OH– and H+ and results in a strong pH
buffering effect (Pytkowicz and Atlas 1975). Jeldres et al.
(2015a, 2015b) used CaO and NaOH to modify the pH of a
Figure 13. Effect of MIBC concentration on surface tension of aqueous solution
in the presence of NaCl (Castro et al. 2012a). mineral suspension in seawater. They showed that the reagent
consumption strongly increases from pH 10.2 and to reach
pH 11 it is required around 25% more of CaO compared to
and drainage. Over-stable froths are generally characterized by NaOH (Figure 14). However, Castro (2012a) indicated that
reduced mobility with detrimental effects on flotation recovery. uptake of lime to achieve a change in pH from neutral to pH
High froth stability also tends to retard drainage of entrained 11 in seawater in ten times higher than in fresh water.
material, thus decreasing the grade of the froth product.
However, low froth stability dramatically reduces the froth recov-
ery (Subrahmanyam and Forssberg 1988). 4.2. Secondary Ion effects
Pure liquids do not foam because there is nothing to stop The Table 2 shows that seawater is rich in calcium and
the liquid from draining out of the bubble lamellae. When magnesium ionic species, mainly carbonates, sulphides, and
surface-active molecules are present, their adsorption at the hydroxides (Castro 2012a) wherein some of these ions can
gas/liquid interface serves to retard the loss of liquid from the form colloidal precipitates at pH > 10. The chemical reactions
lamellae and produce a more mechanically stable foam that occur in seawater include the speciation of CO2, Mg, and
(Klassen and Makrousov 1963). Ca, and the salts precipitation as follow:
The previous section on bubble size described in detail how
different types of ions tend to distribute differently between
the bulk solution and the froth–water interface. The retarda- Table 2. Composition of seawater (Millero et al. 2008).
tion of the bubble coalescence that drives the formation of Component Concentration (g/kg) Component Concentration (g/kg)
increasingly small bubbles is also a process that affects the H2O 964.83 HCO3– 0.104
Na+ 10.78 Br– 0.0672
stability of foams and froths. It is, therefore, likely that the Mg2+ 1.28 CO32– 0.0143
stabilising effects of inorganic ions on flotation froths could Ca2+ 0.41 B(OH)4– 0.00795
K+ 0.39 F– 0.0013
be described in the same terms, such as the α and β assign- Sr2+ 0.0079 OH– 0.00014
ments proposed by Henry and Craig (2010). In addition, the Cl– 19.35 B(OH)3 0.01944
types of froths encountered in flotation practice are three SO42– 2.71 CO2 0.00042
378 R. I. JELDRES ET AL.

14

12
Reagent consumption (g/kg)

NaOH
10 CaO

7 8 9 10 11 12

pH Figure 15. Cu recovery as a function of pH adjusted by lime for a copper ore

floated in fresh and seawater in a chalcopyrite/bornite rich ore (Castro 2012a).
Figure 14. CaO and NaOH consumption in seawater (Jeldres et al. 2015a).

Speciation of CO2
 
CO2ðgÞ þ H2 OðlÞ $ H2 CO3 K1 ¼ H2 CO3 =PCO2
   
H2 CO3 $ HCO 3 þH
þ
K2 ¼ HCO þ
3 ½H = HCO3
  
HCO 2
3 $ CO3 þ H
þ
K3 ¼ CO2 þ
3 ½H = HCO3


Speciation of Mg and Ca
    
CaCO03 $ Caþ2 þ CO2
3 Ke1 ¼ Caþ2 CO2 3 = CaCO3
0

MgHCOþ 3 $ Mg
þ2
þ HCO 3
    
Ke2 ¼ Mgþ2 HCO2 = MgHCOþ
  3 3
MgOHþ $ Mgþ2 þ OH Ke3 ¼ Mgþ2 ½OH =½MgOH 
Salt precipitation
  
CaCO3 $ Caþ2 þ CO2 Kp1 ¼ Caþ2 CO2
3  þ2  32 
MgCO3 $ Mgþ2 þ CO2 Kp2 ¼ Mg CO3 Figure 16. Cu recovery from chalcopyrite at pH 8, in the absence of collector, as
3  þ2  a function of water quality (Smith and Heyes 2012).
MgðOHÞ2 $ Mgþ2 þ 2OH Kp3 ¼ Mg ½OH 2
The presence of hydroxy metal complexes has long been
known to affect mineral surfaces through adsorption (James and they found that the presence of seawater did not improve
and Healy 1972). The adsorption of such ions onto hydro- the overall chalcopyrite recovery, but it has had an effect of
phobic mineral surfaces can often lead to the loss of their mildly improving the flotation kinetics, as shown in Figure 16.
hydrophobicity and subsequent depression from flotation Chalcocite appears to be negatively affected by the presence
(Fuerstenau et al. 1988; Laskowski and Castro 2012). of seawater (Alvarez and Castro 1976). Copper recoveries from
pure chalcocite and a chalcocite/covelite rich ore were found to
display a clear peak around pH 10, with a sharp drop in
5. Seawater in base metal sulphide flotation recovery on either side of the pH range (Figure 17). Similarly,
seawater was found to have a detrimental effect on the float-
5.1. Copper sulphide minerals
ability of bornite, as shown in Figure 18 (Smith and Heyes
Examples of copper minerals include chalcopyrite (CuFeS2), 2012). In a salt-water environment at pH 8, the recovery of
bornite (CuFeS4), chalcocite (CuS2), and covellite (CuS). Head bornite was reduced so substantially that it was likely taking
grades are typically less than 2% Cu, due to the presence of place by means of entrainment alone, rather than true flotation.
other metal sulphides, and non-sulphide gangue minerals The reasons why chalcocite and bornite are so strongly affected
such as pyrite (FeS2), quartz, clays, between others. by seawater whereas chalcopyrite is not are unclear. One possibi-
Chalcopyrite, the major mineral for copper production, lity is that these three minerals display different degrees of oxida-
often occurs in association with iron sulphide minerals, parti- tion, which may or may not be further influenced by seawater.
cularly pyrite. Alvarez and Castro (1976) found that the use of Fullston et al. (1999) used ζ measurements to rank the degree to
seawater has little or no effect on the flotation of pure chal- which copper bearing sulphide minerals oxidize. They found that
copyrite and later Castro (2012a) confirmed it (Figure 15). the oxidation was most prevalent in the following decreasing
Similarly, Smith and Heyes (2012) examined the effect of order: chalcocite > bornite > covellite > chalcopyrite. In other
water quality on the flotation of chalcopyrite/bornite ore words, chalcocite and bornite are much more prone to oxidation
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 379

and edges. The planes have different surface properties,

whereby the faces tend to be hydrophobic, whereas the
edges are hydrophilic (Chander and Fuerstenau 1972). This
makes the flotation behavior more complex than of those
minerals with a simple and homogenous surface structure.
However, as is the case of many other plate-like minerals,
the basal planes are not smoothly crystalline but are covered
in imperfections such as steps and craters. Such “steps” intro-
duce an edge component to the basal planes, adding hydro-
philic sites to an otherwise hydrophobic surface. The
Figure 19a shows a SEM image of a molybdenite particle,
depicting its layered nature (Triffett et al. 2008), whereas
Figure 19b depicts the imperfect nature of a molybdenite
basal plane surface (Komiyama et al. 2004).
Figure 17. Chalcocite recovery in fresh and seawater (Alvarez and Castro 1976).

5.2.2 Effect of high salt concentration

Lucay et al. (2015) found that the floatability of molybdenite is
enhanced in the presence of sodium chloride (Figure 20). This
effect could be explained by the compression of the electrical
double layer with increasing the solution ionic strength
(Troncoso et al. 2014), which leads to a decrease in the double-
layer repulsion force between particles and bubbles. However,
the opposite effect has been observed with seawater at pH levels
higher than 9.0–9.5 (Figure 21). Such strongly alkaline condi-
tions are normally used to depress pyrite but, as was discussed in
Section 4.2, the addition of lime into seawater produces large
quantities of precipitated secondary ions, specifically calcium
and magnesium complexes (hydroxides, carbonates, and sul-
phates). These ions have been shown to be detrimental for
Figure 18. Cu recovery from bornite at pH 8, in the absence of collector, as a molybdenite flotation and possibly for the flotation of chalcocite
function of water quality (Smith and Heyes 2012). and bornite (Alvarez and Castro 1976; Smith and Heyes 2012).

than chalcopyrite. The alteration of the mineral species in distilled 5.2.3 Secondary ion effects
water as a function of pH was attributed to the dissolution of the The effect of calcium and magnesium ion complexes on molyb-
surface copper hydroxide layer (pH < 8) and precipitation of denite flotation was investigated by Castro (2012b), who showed
copper hydroxide on the mineral surface (pH > 6). that the Ca+2 and Mg+2 cations strongly depressed molybdenite
in 0.6M NaCl solutions at pH 10. It was also shown that the
detrimental effect the Mg+2 ions is significantly stronger than
that of Ca+2 ions (Castro et al. 2016).
5.2. Molybdenite
The effect of both ions (Ca+2 and Mg+2) may be attributed
5.2.1 Molybdenite anisotripic structure to the hydrolysis and adsorption of MOH+ hydroxyl-com-
Molybdenite poses a special case, as this mineral is anisotropic plexes. This is followed by the subsequent precipitation of
with a layered platy structure and distinct basal planes (faces) hydrophilic colloidal hydroxides at the molybdenite/solution

Figure 19. (a) SEM image of a molybdenum particle, showing the layered structure, adapted from Triffett et al. (2008). (b) TEM image of a molybdenum basal plane,
showing surface imperfections indicated by arrows, adapted from Komiyama et al. (2004).
380 R. I. JELDRES ET AL.

Figure 20. Natural floatability of molybdenite particles as a function of the

concentration of NaCl in the solution (Lucay et al. 2015). Figure 22. Froth layer thickness under rougher flotation conditions at laboratory
scale as a function of pH, for fresh water and seawater (frother: Mathfroth-355,
10 g/ton; Castro et al. 2012a).

Figure 21. Chalcopyritic Cu and Mo rougher flotation recovery at laboratory

scale in fresh water and seawater as a function of pH (35% solids content; Ramos
et al. 2013).

Figure 23. Natural floatability of molybdenite with two different sizes as func-
tion of SO2
4 concentration in presence of 0.05 mol/L Ca
+2
(Lucay et al. 2015).
interface and in bulk solution. Notably, magnesium hydroxide
is much less water soluble than calcium hydroxide. In this
sense, Castro (2010) proposed that the pretreatment of sea- 2
and HCO3– had no detrimental effect on molybdenite flota-
water with partial removal of calcium and magnesium is tion performance. In addition, Lucay et al. (2015) showed that
enough to achieve good recoveries of molybdenum. an increased concentration of SO4–2 in the presence of Ca+2
Ramos et al. (2013) also compared profiles of Cu and Mo might be detrimental, due to the formation of gypsum
recovery (Figure 21) with the froth layer thickness (Figure 22) (CaSO4·2H2O) precipitates which are deposited on molybde-
as a function of pH but no relationship was observed. nite basal planes, rendering them hydrophilic (Figure 23).
Laskowski et al. (2013) suggested that in the flotation of
Cu–Mo sulphide ores, the effect of seawater ions on molyb-
denite depression is very different from the role that these
ions play in stabilizing the froth. 5.3. Pyrite
Another important factor to be considered is the charge of Pyrite is the most prominent sulphide gangue in base metal
bubbles, particularly when cations hydroxyl-complexes are sulphide flotation processes. Given the right conditions, it is
formed into the system. Li and Somasundaran (1991) showed highly floatable, readily entering flotation concentrates, redu-
that magnesium hydroxyl-complexes and magnesium hydro- cing their grades. Pyrite can be depressed if the pH of flota-
xide are characterized by a very high affinity toward the tion pulps is brought to alkakine levels (above pH 10), which
liquid/gas interface and their adsorption makes the bubbles is commonly realized by lime addition. The presence of lime
positively charged. Later, Han et al. (2004) confirmed that results in the deposition of Ca(OH)2 and CaSO4 precipitates
bubbles become cationic in MgCl2 above 10–2 M and espe- (Fe(OH)3 precipitates are also expected) on the pyrite sur-
cially above pH 9. faces, rendering them completely hydrophilic. However, as
The effect of anions on molybdenite flotation is less well described in Section 4.1, the buffering effect of seawater
studied. Castro (2012b) indicated that the anions such as SO4– greatly increases the consumption of lime in the pH
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 381

adjustment process (Figure 14) and also large quantities of weight maintain a strong depressing effect on the flotation
Ca+2 ions are introduced into the flotation system (Section of molybdenite. Thus, there is not a direct relationship
5.2). For this reason, the depression of pyrite in seawater between the flocculant effectiveness and its depressing effect
needs to be achieved by other means while operating at a on molybdenite recovery. Even polyacrylamide degraded by
lower pH. shear tourned out to be effective molybdenite depressants. For
Traditionally, reagents such as cyanides, sulphites, and this reason, the recommendation is that the use of flocculant
ferro-cyanides have been used as pyrite depressants. in the Cu–Mo flotation process must be limited because they
However, their use has been significantly curbed because and their degradation product exhibit a strong depressing
they present severe environmental hazards. Consequently, effect on molybdenite flotation.
organic polymers have been considered as an alternative,
taking advantage of their natural, biodegradable, and nontoxic
5.4. Pentlandite
properties. In this sense, Lopez-Valdivieso et al. (2004)
showed that dextrin is as effective as cianide, at least in Some of the effects of saline water environment on pentlan-
fresh water. This could be explained by the adsorption of dite flotation have already been discussed in Section 3.2.2.
dextrin onto ferric oxyhidroxide, which is a product of the Peng and Seaman (2011) reported that bore water increased
pyrite surface oxidation. Also, Bicak et al. (2007) indicated pentlandite recovery while reducing lizardite entrainment.
that guar gum can adsorb and depress pyrite even at low Later, Peng and Bradshaw (2012) proposed two possible
dosages. However, these tests were performed with pure pyrite mechanisms responsible for the improved flotation separa-
and were not extended to complex ores. Chen et al. (2010) tion in bore water: (a) the reduction of the electrical double-
showed that organic compounds containing single hydroxyl layer forces between the particles in bore water might miti-
(–OH), carboxyl (–COOH), or amino (–NH2) functional gate the coating of pentlandite surfaces by lizardite slimes;
groups, are ineffective in depressing marmatite, jamesonite, and (b) the reduction of the electrical double-layer forces
and pyrite. They found that thioglycollic acid containing might also induce to the aggregation of fine lizardite parti-
reductive tyiol (–SH) functional groups, performed well as a cles, and therefore enhance the lizardite rejection. A similar
depressant for marmatite and pyrite. The depressing perfor- result was previously observed by Wellham et al. (1992),
mance was also enhanced by the presence of a benzene ring in who reported that the use of seawater made a significant
the molecule structure. improvement to the grade and recovery of nickel from
More recently, Sarquís et al. (2014) have used tannin nickel sulfide transition ores at Leister Mine in Western
(obtained from quebracho tress) as another organic depres- Australia. The improvement was attributed to the reduced
sant that notably promotes pyrite depression and increases the slime coatings and the improved hydrophobicity of pentlan-
Cu/Fe ratio in concentrates between 8% and 40% (Figure 24). dite surfaces. In that study, the use of seawater was found to
More detailed information on the use of organic compounds be significantly more effective than that of carboxymethyl
that promote pyrite depression is contained in Liu et al. cellulose, which is a reagent commonly used as a depres-
(2009), Koleini et al. (2012), and Mu et al. (2015). sant/dispersant of magnesium oxide–based gangue.
Unfortunately, polymeric depressants are also exceedingly Moreover, Peng et al. (2012) studied three type of
effective in depressing molybdenite as well as pyrite (Braga Pluronic triblock copolymers—varying in solubility and
et al. 2014; Castro et al. 2016). Castro and Laskowski (2015) hydrophile–lipophile balance (HLB) values—on the flotation
showed that polyacrilamide flocculants are strong flotation of nickel sulphide minerals. Pluronic triblock copolymers
depressants of fine particles of molybdenite and this effect are a type of nonionic dispersants. They are synthesized by
increases as the polymer concentration is increased and it was the simultaneous polymerization of two monomers, poly-
also found that short chain segments with low molecular ethylene oxide (PEO) and polypropylene oxide (PPO), in
three blocks denoted as (PEO)n1–(PPO)m–(PEO)n2. The
pluronic PEO–PPO–PEO block copolymers are available in
a range of molecular weights and PPO/PEO composition
ratios. The authors found the two copolymers with higher
solubility or HLB value produced more stable mineral sus-
pensions and more stable froth in flotation, resulting in a
better nickel grade and recovery.

6. Conclusions
6.1. Effect of seawater on mineral and solution chemistry
The effect of seawater on flotation operations can be attrib-
uted to two specific aspects: (i) the high concentration of
electrolytes, and (ii) presence of secondary and multivalent
ions with their precipitates. These characteristics have impor-
tant consequences on the flotation performance, as listed
Figure 24. Comparison between pyrite and chalcopytite, with and without below:
tannin (Sarquís et al. 2014).
382 R. I. JELDRES ET AL.

The compression of the electrical double layer around both 6.3. Research opportunities
particles and bubbles results in the following:
The use of seawater as an alternative to fresh water in flota-
tion operations is becoming increasingly important. However,
● Enhancement of floatability in surfaces that are already the topic is complex, and many issues require more detailed
hydrophobic. investigations, as mentioned below:
● Mitigation of the slime coatings.
● The increase of the entrainment. ● The oxidation of copper sulphide and the galvanic inter-
action between pyrite and copper mineral surfaces has
been well approached in fresh water. However, there is a
The high salinity of seawater also results in the following
lack of knowledge about the oxidation in seawater.
effects on the air–water interface:
● It is well known that a high saline environment changes
the interaction of particles surfaces with impact on the
● The reduction of the bubble size. rheological properties of the pulps. However, there is
● The increase of stability of the flotation froths. not studies that evaluate the rheological properties of
sulphide minerals pulps in seawater and their relation-
ship with the flotation performance.
The presence of secondary ions in seawater results in a
● Lack of knowledge about the behaviour of reagents
number of effects, including
beyond frothers and copper collector, for example
many organic pyrite depressants have only been tested
● a strong buffering effect, necessitating a large quantity of in fresh water. The same happens with the molybdenite
reagents to alter the solution pH; and collectors, which might be helpful to address the detri-
● the colloidal precipitation of secondary ions, specifically, mental effect of the hardness of seawater.
the precipitation of Magnesium hydroxide, caused by
lime addition.

Funding
The authors thank INNOVA CORFO Projects Csiro Chile 10CEII-9007,
and L.A.C. also thanks CONICYT and the Regional Government of
6.2. Effect of seawater on the flotation of base metal
Antofagasta for their funding through the PAI program, Project Anillo
sulphides ACT 1201.
The presence of seawater has a varied effect on the floatability
of different base metal sulphides, specifically:
References
● Seawater might have a beneficial effect on mineral flota- Alvarez, J., and Castro, S., 1976, Flotation of chalcocite and chalcopytire
tion since the effect of slime coatings is reduced. At in seawater and salty water. IV. Encontro Naciaonal de Tratamento de
present, the studies have only been focused on bento- Minerios. Sao Jose Dos Campos, Brazil, Anais. 1. pp. 39–44 (Spanish
Text).
nite, but further investigations must incorporate other
Bicak, O., Ekmekci, Z., Bradshaw, D. J., and Harris, P. J., 2007,
types of clays. “Adsorption of guar gum and CMC on pyrite.” Minerals Engineering
● Overall, the flotation of chalcopyrite is not sufficiently 20(10). pp. 996–1002.
sensitive to the effects of seawater. Braga, P. F., Chaves, A. P., Luz, A. B., and França, S. C., 2014, “The use of
● Seawater decreases the recovery of chalcocite and bor- dextrin in purification by flotation of molybdenite concentrates.”
International Journal of Mineral Processing, 127. pp. 23–27.
nite. Despite this behavior has not been explained, it was
Calgaroto, S., Wilber, K. Q., and Rubio, J., 2014, “On the nanobubbles
found that the floatability can be improved through the interfacial properties and future applications in flotation.” Minerals
use of collectors. Engineering, 60. pp. 33–40.
● Little or no effect on the floatability of pyrite. However, Castro, S., 2012a, “Challenges in flotation of Cu-Mo sulfide ores in sea
the buffering effect of seawater makes difficult to operate water.” In: Proc. 1st International Symposium. Water in Mineral
Processing., Drelich, J. (Ed.). SME, Englewood, pp. 29–40.
at high alkaline conditions. Thus new alternatives to
Castro, S., 2010, Use of seawater and other saline waters (salinity between
depress pyrite must be found. 1-7%) pre-treated with lime or other alkalinizing reagents, on miner-
● The decrease in the recovery of molybdenite at high als flotation processes, where the pre-treatment removes the ions of
alkaline conditions. This is due to the deposition of Ca seawater hardness without significantly altering its salinity. Chilean
and Mg complex on the molybdenite surfaces and sub- patent, 52248, 2010.
Castro, S., 2012b, Depression of molybdenite in sea water. In XXVI
sequent precipitation. In this sense, new alternatives
International Mineral Processing Congress, New Delhi, India. pp.
have been proposed as the use of seawater with partial 737–752.
precipitation of calcium and magnesium instead the Castro, S., and Laskowski, J. S., 2015, “Depressing effect of flocculants on
complete desalination by reverse osmosis. molybdenite flotation.” Minerals Engineering, 74. pp. 13–19.
● Improvement in the flotation of pentlandite. This can be Castro, S., Lopez-Valdivieso, A., and Laskowski, J.S., 2016, “Review of the
flotation of molybdenite. Part I: Surface properties and floatability.”
attributed to the mitigation of slime coating effects.
International Journal of Mineral Processing, 148. pp. 48–58.
 MINERAL PROCESSING AND EXTRACTIVE METALLURGY REVIEW 383

Castro, S., Miranda, C., Toledo, P., and Laskowski, J. S., 2013, “Effect of Grau, R. A., Laskowski, J. S., and Heikamem, K., 2005, “Effect of frothers
frothers on bubble coalescence and foaming in electrolyte solutions and on bubble size.” Internationcal Journal of Mineral Processing, 76. pp.
seawater.” International Journal of Mineral Processing, 124. pp. 8–14. 225–233.
Castro, S., Ramos, O., Cancino, J. P., and Laskowski, J. S., 2012a, Han, M. Y., Ahn, H. J., Shin, M. S., and Kim, S. R., 2004, “The effect of
Frothing in the flotation of copper sulphide ores in sea water. In divalent metal ions on the zeta potential of bubbles.” Water Science
Water in Mineral Processing: Proceedings of the First International and Technology, 50. pp. 49–56.
Symposium. SME, Englewood, USA. pp. 211–223. Hencer, M., Celik, M. S., and Miller, J. D., 2001, “The significance of
Castro, S., Toledo, P. G., and Laskowski, J. S., 2012b. Foaming properties interfacial water structure in soluble salt flotation systems.” Journal of
of flotation frothers at high electrolyte concentration. In Water in Colloid and Interface Science, 235. pp. 150–161.
Mineral Processing: Proceedings of the First International Symposium. Henry, C. L., and Craig, V. S., 2010, “The link between ion specific
SME, Englewood, USA. pp. 51–60. bubble coalescence and Hofmeister effects is the partitioning of ions
Castro, S., Venegas, I., Landero, A., and Laskowski, J. S., 2010. Frothing within the interface.” Langmuir, 26. pp. 6478–6483.
in seawater flotation systems. In XXV International Mineral Processing Hogg, R., Healy, T. W., and Fuerstenau, D. W., 1966, “Mutual coagula-
Congress. Brisbane, Australia, AusIMM. pp. 4039–4047. tion of colloidal dispersions.” Transactions of the Faraday Society, 62
Chan, D. Y. C., Mitchell, D. J., Ninham, B. W., and Pailthorpe, A., 1978, (0). pp. 1638–1651.
“Short-range interactions mediated by a solvent with surface adhe- Ishida, N., Inoue, T., Miyahara, M., and Higashitani, K., 2000, “Nano
sion.” Molecular Physics, 35. pp. 1669–1679. bubbles in a hydrophobic surface in water observed by tapping-mode
Chander, S., and Fuerstenau, D. W., 1972, “On the natural floatability of atomic force microscopy.” Langmuir, 16. pp. 6377–6380.
molybdenite.” Transactions of the SME, 252. pp. 62–69. Israelachvili, J., 1991, Intermolecular & Surface Forces. 2nd ed. New York:
Chen, J., Li, Y., and Long, Q., 2010, “Molecular structures and activity of Hacourt Brace & Company.
organic depressants for marmatite, jamesonite and pyrite.” Israelachvili, J. N., and Pashley, R. M., 1983, “Measurement of the
Transactions of Nonferrous Metals Society of China, 20. pp. 1993–1999. hydrophobic interaction between two hydrophobic surfaces in aqu-
Cho, Y. S., and Laskowski, J. S., 2002a, “Effect of flotation frothers on eous electrolyte solution.” Journal of Colloid and Interface Science, 98
bubble size and foam stability.” International Journal of Mineral (2). pp. 500–514.
Processing, 64. pp. 69–80. Israelachvili, J., and Wennerström, H., 1996, “Role of hydration and
Cho, Y. S., and Laskowski, J. S., 2002b, “Bubble coalescence and its effect water structure in biological and colloidal interactions.” Nature, 379.
on dynamic foam stability.” The Canadian Journal of Chemical pp. 219–225.
Engineering, 80. pp. 299–305. James, R. O., and Healy, T. W. 1972, “Adsorption of hydrolyzable metal
Cisternas, L., and Moreno, L., 2014, Seawater in Mining: Fundamentals ions at the oxide-water interface. I, Co(II) adsorption on SiO2 and
and Applications (in Spanish), Santiago: Master Ril. TiO2 as model systems.” Journal of Colloid and Interface Science, 40.
Craig, V. S., 2011, “Do hydration forces play a role in thin film drainage pp. 42–52.
and rupture observed in electrolyte solutions?” Current Opinion in Jeldres, R. I., Toledo, P. G., Concha, F.; Stickland, A. D., Usher, S. P., and
Colloid and Interface Science, 16. pp. 597–600. Scales, P. J., 2014, “Impact of seawater salts on the viscoelastic beha-
Craig, V. S., Ninham, B. W., and Pashley, R. M. 1993a, “Effect of vior of flocculated mineral suspensions.” Colloids and Surfaces A, 261.
electrolytes on bubble coalescence.” Nature, 364. pp. 317–319. pp. 295–302.
Craig, V. S. J., Ninham, B. W., and Pashley, R. M. 1993b, “The effect of Jeldres, R. I., Calisaya, D., and Cisternas, L. A., 2015a, Effect of different
electrolytes on bubble coalescence in water.” The Journal of Physical saline water on floatability of copper-molybdenum ores. 11st
Chemistry, 97(39). pp. 10192–10197. International Mineral Processing Conference, Santiago, 2015.
Del Guidice, G. R. M. 1934, “A study of slime coatings in flotation.” Jeldres, R. I., Cortes, L., and Cisternas, L. A., 2015b, Impact of seawater
Transactions AIME, 112. pp. 398–409. with calcium and magnesium removal on floatability of copper-
Derjaguin, B. V., and Landau, L. 1941, “Theory of stability of strongly molybdenum ores. Minerals Engineering International Conferences,
charged lyophobic sols and the adhesion of strongly charged particles Flotation 2015, Cape Town, South Africa.
in solution of electrolytes.” Acta Physiochim, 14. pp. 633–662. Johansson, G., and Pugh, R. J., 1992, “The influence of particle size and
Drelich, J., and Miller, J. D., 2012, Induction time measurements for air hydrophobicity on the stability of mineralized froths.” International
bubbles on chalcopyrite, bornite, and gold in seawater. In The First Journal of Mineral Processing, 34. pp. 1–21.
International Symposium on Water in Mineral Processing. Society for Klassen, V. I., and Makrousov, V. A., 1963, Introduction to the Theory of
Mining, Metallurgy, and Exploration, Seattle, USA. Flotation, London: Butterworth.
Dukhin, S. S., and Derjaguin, B. V., 1976, Electrophoresis, Moscow: Komiyama, M., Kiyohara, K., Li, Y., Fujikawa, T., Ebihara, T., Kubota, T.,
Academy of Science, USSR, Russian text. and Okamoto, Y., 2004, “Crater structures on a molybdenite basal
Firouzi, M., and Nguyen, A. V., 2014, “Effects of monovalent anions and plane observed by ultra-high vacuum scanning tunneling microscopy
cations on drainage and lifetime of foam films at different interface and its implication to hydrotreating.” Journal of Molecular Catalysis A:
approach speeds.” Advanced Powder Technology, 25. pp. 1212–1219. Chemical, 215(1–2). pp. 143–147.
Fuerstenau, D. W., Gaudin, A. M., and Miaw, H. L., 1958, “Iron oxide Kracht, W., and Finch, J. A., 2010, “Effect of frother on initial bubble
slime coatings in flotation.” Transactions AIME, 211. pp. 792–793. shape and velocity.” International Journal of Mineral Processing, 94.
Fuerstenau, D. W., Rosenbaum, J. M., and Laskowski, J., 1983, “Effect of pp. 115–120.
surface functional groups on the flotation of coal.” Colloids and Koleini, S. M., Abdollahy, M., and Soltani, F., 2012. The comparation
Surfaces, 8(2). pp. 153–173. between sodium cyanide and white dextrin as pyrite depressants in the
Fuerstenau, M. C., Lopez-Valdivieso, A., and Fuerstenau, D. W., 1988, sequential flotation Cu-Zn massive sulphide ore-irantaknar.
“Role of hydrolyzed cations in the natural hydrophobicity of talc.” Proceedings 26th International Mineral Processing Congress 2012,
International Journal of Mineral Processing, 23(3–4). pp. 161–170. New Delhi, India. pp. 5120–5128.
Fullston, D., Fornasiero, D., and Ralston, J., 1999, “Zeta potential study Kurihara, K., Kato, S., and Kunitake, T., 1990, “Very strong long range
of the oxidation of copper sulfide minerals.” Colloids and Surfaces A: attractive forces between stable hydrophobic monolayers of a poly-
Physicochemical and Engineering Aspects, 146(1–3). pp. 113–121. merized ammonium surfactant.” Chemistry Letters, 19(9). pp. 1555.
Gaudin, A. M., 1957, Flotation, New York: McGraw Hill. Laskowski, J., and Castro, S., 2012, Hydrolyzing ions in flotation cir-
Grabbe, A., and Horn, R. G., 1993, “Double-layer and hydration forces cuits: sea water flotation. XIII International Minerals Processing
measured between silica sheets subjected to various surface treat- Symposium. H. ÖZDAĞ, V. BOZKURT, H. İPEK and K. BİLİR.
ments.” Journal of Colloid Interface Science, 157(2). pp. 375–383. Bodrum, Turkey.
Grau, R. A., and Laskowski, J. S., 2006, “Role of frothers in blubble Laskowski, J. S., Castro, S., and Ramos, O., 2013, “Effect of seawater main
generation and coalescence in a mechanical flotation cell.” The components on frothability in the flotation of Cu-Mo sulfide ore.”
Canadian Journal of Chemical Engineering, 84. pp. 170–182. Physicochemical Problems of Mineral Processing, 50(1). pp. 17–29.
384 R. I. JELDRES ET AL.

Laskowski, J. S., Tlhone, J., Williams, P., and Ding, K., 2003, copolymer. (4190, Ed.) XXVI International Mineral Processing
“Fundamental properties of the polyoxypropylene alkyl ether flotation Congress. New Delhi, India. pp. 4179.
frothers.” International Journal of Mineral Processing, 72. pp. 289–300. Peng, Y., and Seaman, D., 2011, “The flotation of slime–fine fractions of
Laskowski, J. S., Xu, Z., and Yoon, R. H., 1991, The energy barrier in Mt. Keith pentlandite ore in de-ionised and saline water.” Minerals
particle to bubble attachment and its effect on flotaiton kinetics. XVII Engineering, 24(5). pp. 479–481.
International Minerals Processing Congress. Dresden, Germany. pp. Pregram, L. M., and Record, M. T., 2007, “Hofmeister salt effects on
237–249. surface tension araise from partitioning of anions and cations between
Leckband, D., and Israelachvili, J., 2001, “Intermolecular forces in biol- bulk water and the air-water interface.” The Journal of Physical
ogy.” Quartely Reviews of Biophyisics, 34(2). pp. 105–267. Chemistry B, 111. pp. 5411–5417.
Lessard, R. R., and Zieminski, S. A., 1971, “Bubble coalescence and gas Pytkowicz, R. M., and Atlas, E., 1975, “Buffer intensity of seawater.”
transfer in aqueous electrolytic solutions.” Industrial and Engineering Limnology and Oceanography, 20(2). pp. 222–229.
Chemistry Fundamentals, 10. pp. 260–269. Quinn, J. J., Kracht, W., Gomez, C. O., Cagnon, C., and Finch, J. A.,
Li, C., and Somasundaran, P., 1991, “Reversal of bubble charge in multi- 2007, “Comparing the effect of salts and frother (MIBC) on gas
valent inorganic salt solutions - effect of magnesium.” Journal of dispersion and frtoh properties.” Minerals Engineering, 20. pp. 1296–
Colloid and Interface Science, 146. pp. 215–218. 1302.
Li, H. C., and De Bruyn, P. L., 1966, “Electrokinetic and adsorption Quinn, J. J., Sovechles, J. M., Finch, J. A., and Waters, K. E., 2014,
studies on quartz.” Surface Science, 5(2). pp. 203–220. “Critical coalescence concentration of inorganic salt solutions.”
Liu, R., Sun, W., Hu, Y., and Wang, D., 2009, “Effect of organic depres- Minerals Engineering, 58. pp. 1–6.
sant lignosulfonate calcium on separation of chalcopyrite from pyr- Ramos, O., Castro, S., and Laskowski, J. S., 2013, “Copper-molybdenum
ite.” Journal of Central South University, 16. p. 0753–0757. ores flotation in sea water: floatability and frothability.” Minerals
López Valdivieso, A., Celedón Cervantes, T., Song, S., Robledo Cabrera, Engineering, 53. pp. 108–112.
A., and Laskowski, J., 2004, “Dextrin as a non-toxic depressant for Ravishankar, S. A., and Yoon, R. H., 1997, “Long-range hydrophibic
pyrite in flotation with xanthates as collector.” Minerals Engineering, forces in the aimne flotation of quartz.” Minerals and Metallurgical
17. pp. 1001–1006. Processing, 14(2). pp. 10–17.
Lucay, F., Cisternas, L. A., Gálvez, E. D., and López-Valdivieso, A., 2015, Sarquís, P. E., Menéndez-Aguado, J. M., Mahamud, M. M., and Dsioba,
“Study of the natural floatability of molybdenite fines in saline solu- R., 2014, “Tannins: the organic depresants alternative in selective
tions. Effect of gypsum precipitation.” Minerals & Metallurgical flotation of sulfide.” Journal of Cleaner Production, 84. pp. 723–726.
Processing Journal, 32. pp. 203–208. Smith, L. K., and Heyes, G. W., 2012, The effect of water quality on the
Meyer, E. E., Rosenberg, K. J., and Israelachvili, J., 2006, “Recent progress collectorless flotation of chalcopyrite and bornite. In Third
in understanding hydrophobic interactions.” Proceedings of the International Congress on Water Management in the Mining
National Academy of Sciences of United States of America, 103(43). Industry, Santiago, Chile.
pp. 15739–15746. Subrahmanyam, T. V., and Forssberg, E., 1988, “Froth stability, particle
Millero, F. J., Feistel, R., Wright, D. G., and McDougall, T. J., 2008, “The entrainment and drainage in flotation—a review.” International
composition of standard seawater and the definition of the reference- Journal of Mineral Processing, 23(1–2). pp. 33–53.
composition salinity scale.” Deep Sea Research Part I: Oceanographic Triffett, B., Veloo, C., Adair, B. J. I., and Bradshaw, D. 2008, “An
Research Papers, 55(1). pp. 50–72. investigation of the factors affecting the recovery of molybdenite in
Moreno, P. A., Aral, H., Cuevas, J., Monardes, A., Adaro, M., Norgate, T., the Kennecott Utah Copper bulk flotation circuit.” Minerals
and Bruckard, W., 2011, “The use of seawater as process water at Las Engineering, 21(12–14). pp. 832–840.
Luces copper-molybdenum beneficiation plant in Taltal (Chile).” Troncoso, P., Saavedra, J. H., Acuña, S. M., Jeldres, R., Concha, F., and
Minerals Engineering, 24(8). pp. 852–858. Toledo, P. G. 2014, “Nanoscale adhesive forces between silica surfaces
Mu, Y., Peng, Y., and Lauten, R. A., 2015, “Electrochemistry aspects of in aqueous solutions.” Journal of Colloid and Interface Science, 424.
pyrite in the presence of potassium amyl xanthate and a lignosulfo- pp. 56–61.
nate-based biopolymer depressant.” Electrochimica Acta, 174. pp. Verwey, E. J. W., and Overbeek, J. T. G., 1948, Theory of Stability of
133–142. Lyophobic Solids, Amsterdam: Elsevier.
Neethling, S. J., and Cilliers, J. J., 2002, “The entrainment of gangue into Wellham, E. J., Elber, L., and Yan, D. S. 1992, “The role of carboxy
a flotation froth.” International Journal of Mineral Processing, 64(2–3). methyl cellulose in the flotation of a nickel sulphide transition ore.”
pp. 123–134. Minerals Engineering, 5(3–5). pp. 381–395.
Nesset, J. E., Finch, J. A., and Gomez, C. O., 2007, Operating variables Xu, Z., Liu, J., Choung, J. W., and Zhou, Z. 2003, “Electrokinetic study of
affecting the bubble size in forced-air mechanical flotation machines. clay interactions with coal in flotation.” International Journal of
Proceedings of the 9th Mill Operators Conference. Fremantle, Australia. Mineral Processing, 68(1–4). pp. 183–196.
pp. 55–65. Yoon, R. H., and Aksoy, B. S. 1999, “Hydrophobic forces in thin water
Oliveira, C., and Rubio, J., 2011, “Zeta potential of single and polymer- films stabilized by dodecyl ammonium chloride.” Journal of Colloid
coated microbubbles using an adapted microelectrophoresis technique.” and Interface Science, 211(1). pp. 1–10.
International Journal of Mineral Processing, 98(1–2). pp. 118–123. Zhang, M., Peng, Y., and Xu, N. 2015, “The effect of sea water on copper
Pashley, R. M., 1982, “Hydration forces between mica surfaces in elec- and gold flotation in the presence of bentonite.” Minerals Engineering,
trolyte solutions.” Advances in Colloid and Interface Science, 16(1). pp. 77. pp. 93–98.
57–62. Zhang, W., Nesset, J. E., and Finch, J. A. 2014, “Correspondence of
Peng, Y., and Bradshaw, D., 2012, “Mechanisms for the improved flota- bubble size and frother partitioning in flotation.” Journal of Central
tion of ultrafine pentlandite and its separation from lizardite in saline South University, 21. pp. 2383–2390.
water.” Minerals Engineering, 36–38(0). pp. 284–290. Zhang, X. H., Quinn, A., and Ducker, W. A. 2008, “Nanobubbles at the
Peng, Y., Liu, D., and Chen, X., 2012, Selective flotation of ultrafine interface between water and a hydrophobic solid.” Langmuir, 24. pp.
nickel sulphide from serpentine in saline water by pluronic triblock 4756–4764.