Minerals Engineering xxx (2016) xxxxxx

Contents lists available at ScienceDirect

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

Behaviour of swelling clays versus non-swelling clays in flotation

Saeed Farrokhpay a,b,, Bulelwa Ndlovu a, Dee Bradshaw a
a
SMI/JKMRC, The University of Queensland, Brisbane, QLD 4068, Australia
b
Universit de Lorraine, GeoRessources Laboratory, CNRS, UMR 7359, 2 rue du Doyen Marcel Roubault, 54518 Vanduvre-ls-Nancy, France

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

Article history: The deleterious effects of clays on flotation performance are widely acknowledged but the mechanisms
Received 12 January 2016 involved are not clearly established. Moreover, the concentrations beyond which clay minerals become
Revised 18 March 2016 problematic are not clearly defined. One major parameter is the difference between swelling and non-
Accepted 16 April 2016
swelling clays which is evaluated in this study. The ore slurry pulp rheology and froth stability were mon-
Available online xxxx
itored in the absence and presence of different clay minerals. It was found that swelling clays can
adversely affect the flotation performance mainly via adsorbing water which changes the rheology and
Keywords:
froth stability, reducing both flotation grade and recovery. Non-swelling clays had a lower effect on
Flotation
Clay
the rheology. Kaolinite increases the froth stability and reduces the flotation grade but illite showed
Phyllosilicate the least effect on the flotation performance in this study. The potential mechanisms and critical concen-
Froth stability trations are discussed.
Swelling clays 2016 Elsevier Ltd. All rights reserved.
Rheology

1. Introduction made up of aluminium-oxygen octahedrons with shared apical

The mineral processing industries are facing finely disseminated
and low grade ores. The processing of these ores is hindered by the
presence of clay minerals. A comprehensive review on the effect of
clays on mineral processing has been published recently (Ndlovu
et al., 2013). It revealed that clays have the potential to significantly
affect the planning, operation and economics of mineral processing
projects. The various deleterious roles of clays in froth flotation have
been investigated by a number of authors (Bulatovic et al., 1999;
Farrokhpay et al., 2013; Forbes et al., 2014; Zhang and Peng,
2015). These effects include coating of the mineral surfaces, increas-
ing reagent consumption due to the high surface area, transferring
large quantities of clay minerals into the concentrate during the
flotation process by entrainment, increasing pulp viscosity, and
changing froth stability (Arnold and Aplan, 1986; Bulatovic, 2007;
Tao et al., 2010; Cruz et al., 2013; Farrokhpay and Ndlovu, 2013).
It has been reported that bentonite and kaolin cause different issues
in flotation (Wang et al., 2015).
Clay minerals are phyllosilicates which are made up of various
combinations of sheets stacked on each other and bonded together.
A tetrahedral sheet is made up of silicon-oxygen tetrahedrons with
shared basal oxygen molecules (T layer). An octahedral sheet is
Corresponding author at: Universit de Lorraine, GeoRessources Laboratory,
CNRS, UMR 7359, 2 rue du Doyen Marcel Roubault, 54518 Vanduvre-ls-Nancy,
France.
E-mail address: saeed.farrokhpay@univ-lorraine.fr (S. Farrokhpay).
and basal oxygen (O layer). These tetrahedral or octahedral sheets
bond together through hydrogen bonding. The 1:1 clay minerals
are formed by bonding of one tetrahedral and one octahedral sheet.
They are electrically neutral therefore they will not attract posi-
tively or negatively charged species, and they do not swell in
water. On the other hand, 2:1 clay minerals are formed by bonding
of two tetrahedral and one octahedral sheet. Due to ion substitu-
tion in the layers, some of these clay minerals have a negative
charge on the surface and will attract water or hydrated cations.
The negatively charged surfaces of 2:1 clays attract positively
charged water molecules, allowing the water molecules to enter
between the layers.
It should be noted that swelling is related to the hydration of
clays but not all clays swell when hydrated. For example, while
kaolinite shows little or no swelling on hydration, montmorillonite
considerably swells when presents in water (up to 1500%, Taylor
and Smith, 1986). Kaolinite and illite are both non-swelling, albeit
with different structures. Kaolinite has 1:1 structure but illite has
2:1 structure (Fig. 1), but they can be easily dispersed and trans-
ported to the concentrates (i.e. entrainment). These differences in
the swelling characteristics of different clays are related to their
chemical composition and structure. There can also be changes in
the degree of isomorphous replacements in their structures as well
as in the amount and nature of their associated exchangeable
cations. It has been shown that different crystalline forms of
kaolinite exist and can affect flotation differently (Ndlovu et al.,
2015).

http://dx.doi.org/10.1016/j.mineng.2016.04.011
0892-6875/ 2016 Elsevier Ltd. All rights reserved.

Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
2 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx
Kaolinite
Fig. 1. Schematic of clay structures showing the T and O layers and potential adsorption of water (adapted from Ndlovu et al., 2014).
Froth stability is known to play an important role in determin-
ing mineral flotation recovery and selectivity (Tang et al., 1989;
Barbian et al., 2005). Over-stable froth is often difficult to handle
(Farrokhpay, 2011). Therefore, optimum froth stability is of utmost
importance in flotation. The presence of clay minerals in ores can
dramatically affect the froth stability. When clay minerals are pre-
sent as individual particles, they may adsorb a large amount of
frother due to their large surface area (for example, adsorption of
MIBC by coal particles as reported by Miller et al., 1983). Therefore,
they may decrease the froth stability. Clay aggregates can also
interact with the hydrophobic valuable minerals resulting in high
froth stability.
Rheology has been recognized as an important factor in differ-
ent aspects of mineral processing, including flotation (e.g. (Boger,
2009; Farrokhpay, 2012a). Rheology data can be used to under-
stand the interactions occurring in a flotation process. For example,
a higher recovery of coarse composite copper bearing particles has
been reported by increasing the viscosity of the flotation medium
(Farrokhpay et al., 2011). Sweet et al. (2012) have also reported
that by diluting the feed slurry with water and adjusting viscosity
in a Pt plant in South Africa, the flotation recovery has increased
more than 10%.
The aim of this study is to understand the differing effects of
swelling and non-swelling clays on the flotation of a copper ore
through examining the grade and recovery, pulp rheology, and
froth characteristics. The findings could be beneficial towards
ongoing studies to manage high clay containing ores in flotation
which has been referred as a mineral processing nightmare
(Connelly, 2011).
2. Materials and experimental methods
2.1. Materials
A copper ore from Australia was used as the baseline in this
investigation. The chemical assays of the ore sample contains
0.5% Cu, 3.8% Fe and 0.3% S. The mineralogy of the ore using XRD
analysis has showed that it has 2.3% sulfides (chalcopyrite 0.7%,
bornite 1.3%, and pyrite 0.3%), 39.4% quartz, 44.4% feldspar, 4.2%
amphibole, 3.2% oxides, and 7.7% carbonates.
The clay minerals used in this study were illite, kaolinite and
montmorillonite. Kaolinite Q38 was provided by Unimin Australia
Limited. All other clay samples were obtained in a pre-ground form
from Wards Minerals.
The particle size distribution of clay samples were measured by
wet size light scattering technology using a Malvern Mastersizer
(Malvern, UK). In each case, 0.5 g sample was dispersed in water.
Sodium hexametaphosphate (Calgon) was also used as a dispersant
Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
Illite Montmorillonite
at a concentration of 1 wt%. The sample was sonicated for 60 s at
50% ultrasonic power for 60 s before the measurement. Pump
and stirring speeds were 330 and 350 rpm, respectively. For each
sample, tests were carried out in triplicate for reproducibility.
The size distribution shown for each mineral is representative of
the average size distribution of three individual measurements.
Fig. 2 shows that the clays samples had more or less same particle
size distribution. This range is the typical size of clay minerals in
real ore systems.
2.2. Flotation tests
The ore was ground by wet milling at 60% solid in a laboratory
rod mill to achieve the P80 of 90 lm. The ground sample was trans-
ferred to a 5 L bottom driven batch flotation cell (JKTech, Brisbane),
and the required amount of water was added to obtain about 25%
solid ratio. Brisbane tap water was used in the flotation experi-
ments. The air flowrate and impeller speed were 15 L/min and
800 rpm, respectively. Tests were conducted on slurries compris-
ing the copper ore and varying concentrations of each clay mineral
(030 wt%). In each case, the pH was maintained at pH 8 using KOH
and/or HCl (This is the pH condition at which most industrial flota-
tion runs are conducted). Potassium amyl xanthate (PAX) and
methyl isobutyl carbinol (MIBC) were used as collector and frother,
at the dosages of 100 g/t and 40 ppm, respectively. The P80 and the
type and amount of flotation reagents were chosen based on the
100
80
Cumulative passing (%)
60
40
20
0
0.1 1 10 100
Particle size (m)
Fig. 2. Particle size distribution of the clay samples (illite j, kaolinite d and
montmorillonite ) by wet size light scattering using a Malvern Mastersizer
(Malvern, UK) (data are the average of 3 tests).
 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx
ore supplier recommendations. For each test, three concentrates
were collected at 10 s intervals over 1, 2 and 5 min (8 min in total).
During flotation, froth images were taken and analysed using
VisioFroth software developed by Metso Minerals Cisa for bubble
size distribution. The camera was positioned directly above the
flotation cell. The image capture area for bubble size measurement
was 11.4
11.4 cm. The average bubble size was calculated by
using the bubble size percentage distribution information. A full
description of the VisioFroth system is given by Runge et al.
(2007) and it has been used in different studies (e.g. Wei et al.,
2014).
The copper flotation rate constants in the presence of different
clays, at two levels of low and high concentrations were calculated
on the basis of the first order flotation rate constant models
(Eq. (1)).
R Rmax 1  ekt
The flotation rate constants were calculated by plotting ln
(1  R/Rmax) against the flotation time (t). The slope of the resulted
lines present the k constant values.
2.3. Frother adsorption tests
A 0.1% MIBC solution by weight was prepared with distilled
water in a conical flask using a magnetic stirrer. For each sample,
5 g of mineral was added to a round bottomed flask together with
a predetermined volume of 0.1% frother solution. The bottle was
filled up to 500 mL with distilled water and was shaken vigorously
by hand to wet the mineral. Using a magnetic stirrer the solution
was mixed for 30 min. The pH and temperature was measured
and the flask was sealed before being placed in the fridge to settle
overnight. The supernatant was removed from the flask and
poured into a 250 mL plastic bottle. Sodium hydroxide and
hydrochloric acid was added to adjust the pH. All samples, with
a sample of pure water (to represent no frother), were sent to
Levay & Co. Environmental Services in Adelaide for frother analysis
using liquid-liquid extraction (with the detection limit of
0.004 mg/L). The amount of adsorbed frother was then calculated
from the difference of the amount of frother before and after the
adsorption tests.
2.4. Froth structure and stability evaluation
The effects of the different clay minerals on froth stability were
investigated using a specially designed column. Fig. 3 gives a sche-
matic of the froth column and the associated parameters. A similar
column has previously been used to assess the effects of different
frothers on the froth stability (Zanin and Grano, 2006;
Farrokhpay and Zanin, 2012). The column is a 100 cm long, with
10
10 cm cross section, in which pulp samples are introduced
for testing foaming properties. The column is equipped with an
impeller to ensure sufficient pulp agitation. A porous plate at the
bottom of the column ensures uniform air diffusion. Air flow is
regulated by means of a flowmeter. The same Jg as in flotation
(1 cm/s) was maintained during the tests. In each case, a 2 L slurry
comprising the copper ore and clays (at different loadings) was
made up. The suspension was conditioned with collector and
frother and placed into the apparatus. Once the pulp slurry was
introduced into the column (time t = 0), the stirrer was turned
on, in order to ensure sufficient agitation and avoid settling of par-
ticles. The froth height was then measured as a function of time
until reaching an equilibrium height. At this point, air was turned
off, and the froth decay was measured. The half-life of the froth
(t1/2), i.e. the time needed for the froth to collapse to half its equi-
librium height, was taken as an indicator for froth stability.
Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
3
2.5. Rheology measurements
Rheological measurements were performed using an AR1500EX
(Thermal Analysis, Germany), with a standard 4-panel vane rotor
geometry (with 18 mm width and 25 mm height). This geometry
is more suitable for analysis of suspensions in minerals processing
industry. It is less susceptible to artefacts arising from large parti-
cle sizes and results in less sample disturbance and thixotropic
breakdown during the measurements (Nguyen and Boger, 1985).
It also eliminates variable gap effects which commonly arise when
dealing with non-Newtonian fluids containing yield stress
(Keentok et al., 1985; Yan and James, 1997; Fisher et al., 2007).
The rheological analyses were conducted on slurries obtained
from the feed material to the flotation tests. Slurries comprising
the ore doped with increasing concentrations of clay minerals (0
30 wt%) were prepared using 0.01 mol dm3 NaCl as a background
1 electrolyte. The stress-strain tests were conducted in a shear rate
controlling regime, with an upward and downward ramp within
the range 0 s1 to 300 s1 over a 60 s time period (No turbulence
was observed at the maximum shear rates). No significant hysteresis
was observed between the upward and downward ramps. The shear
stress values was obtained automatically from the rheometer using
the software provided by the rheometer manufacturer.
3. Results and discussion
3.1. Flotation
Fig. 4 shows the effect of concentration of the selected clay min-
erals on the final copper flotation grade and recovery. It should be
noted that while it was possible to increase the level of illite and
kaolinite to up to 30 wt% it was only possible to test montmoril-
lonite up to 15 wt%. With no clay mineral addition, a maximum
copper grade of about 18% was obtained (rougher concentrate).
However, a significant decrease in the copper grade was observed
in the presence of all clay minerals.
The copper grade decreased to about 1% in the presence of 30%
or 15% of kaolinite or montmorillonite, respectively and to about
5% in the presence of 30% illite. This may be due to the transporta-
tion or entrainment of clay fine particles to the froth phase and/or
the rheological effects which affect the hydrodynamics within the
cell. These very small particles can move upwards to the froth layer
reducing the copper grade. Recently, it has been reported that the
addition of 25% kaolinite has caused the flotation grade of a copper
ore to be reduced from 4% to 1.5% (about 60% reduction) (Wang
et al., 2015) (although strangely, the copper grade in the absence
of clays is extremely low). In the current study, the data presented
in Fig. 4 indicate a slight decrease in the copper recovery in the
presence of kaolinite or illite while the copper grade considerably
decreases to 14% when 30% of these clays present. The reason of
decreasing the grade to such low values while the recovery stays
almost constant can be the entrainment. For example, the total
weight of the flotation concentrates in the presence of 10% and
30% illite were about 20 and 60 g, respectively. It is expected that
the extra mass is due to the entrainment of the clay particles.
Wang et al. (2015) have reported that an increase in the amount
of bentonite did not significantly affect the copper grade. This is
contrary to the findings of the current investigation with a reduc-
tion of about 40% copper grade upon the addition of bentonite (The
addition of 20% bentonite reduced the copper grade from 4% to
2.5%). The discrepancy could be attributed to differences in the
bentonite samples used in these two studies. The bentonite sample
used by Wang et al. (2015) comprised about 63% montmorillonite,
while the current investigation used a montmorillonite sample of
higher purity (+90%) therefore, one might not see the full effect
of this clay due to the high impurity (25% albite and 12% quartz).
4 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx

Feed

Equilibrium froth
height (Leq)

Equilibrium froth
height (Leq)
Initial froth
1/2 Equilibrium froth height (Li)
height (1/2 Leq)

Initial froth
height (Li)

Air

(a) (b)
Fig. 3. (a) A schematic of the froth column and (b) a physical representation of the initial and equilibrium froth heights.

20 and kaolinite (i.e. 30%), the final copper recovery only slightly
18 changed to 88%. A similar trend has been reported recently for
16 the effect of clay minerals on the copper flotation using chalcopy-
rite (Farrokhpay and Ndlovu, 2013). The reduction in the copper
14
Cu grade (%)

recovery due to the presence of montmorillonite content in flota-

12
tion can be due to the reduced bubble-particle collisions and/or
10
the mobility of bubble-particle aggregates. High slurry viscosity
8 can cause the frequency of bubble-particle collision to be reduced
6 resulting in the poor flotation performance. The effect of slime
4 coating and/or entrainment should not be ignored, however, more
2 details analysis using Cryo SEM (for example as conducted by
0 Wang et al., 2013) is needed to evaluate it.
0 5 10 15 20 25 30 From a practical view point, the overall flotation process is gen-
100 erally represented by the first order system with lumped parame-
ters, assuming the collection zone is perfectly mixed. Table 1
96 shows the flotation rate constants in the presence of different clays
Cu recovery (%)

at two level of low and high concentration, for the total flotation
92 results (three concentrates). It can be seen that the flotation rate
constant decreases due to the presence of higher amount (i.e.
88
30%) of kaolinite suggesting the possible occurrence of slime coat-
ing, or physical adsorption of kaolinite particles onto the copper
84
containing mineral surface. It can be argued that the flotation rate
80 constant is dependent on hydrodynamics of a flotation system.
Under a fixed hydrodynamic condition as in the current case, the
76 observed variation in flotation rate constant cannot be attributed
0 5 10 15 20 25 30 to hydrodynamic, but rather related to the changes in the system
Clay content (%) chemistry (Kasongo et al., 2000). The resulted flotation rate

Fig. 4. Effect of different clay minerals (illite j, kaolinite d and montmorillonite )

on the flotation of the copper ore (the error bars represent 95% confidence Table 1
intervals). Flotation tests were conducted at 25% solid and pH 8, in a 5 L bottom The copper flotation rate constant in the presence of different clays, at two levels of
driven batch flotation cell. (air flowrate: 15 L/min; impeller speed: 800 rpm; low and high (clay) concentrations.
collector: PAX, 100g/t; frother: MIBC, 40 ppm).
Kaolinite Illite Montmorillonite
Low concentration (i.e. 10%) 0.68 0.68 0.60
Montmorillonite was also found to have the most significant High concentration (i.e. 30% for 0.51 0.49 0.33
effect on the copper recovery. The final recovery decreased from kaolinite and illite, and 15% for
90% to about 80% in the presence of 15% montmorillonite. montmorillonite)
Zero concentration (no clay) 0.70 0.70 0.70
However, with the addition of the maximum tested amount of illite

Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx 5

constants suggest that kaolinite particles triggered the interactions 20

between the copper bearing minerals and collectors. On the other 18
hand, no change in the flotation rate constant was observed in the
16
presence of montmorillonite and illite.
To better understand the behaviour of clay minerals on the 14

Froth half-life(s)
flotation, the flotation products were assayed for Al2O3 and SiO2. 12
It should be noted that the amount of Al2O3 and SiO2 in the concen- 10
trate is indications of the clay particles reporting to the concen-
8
trate. In the current study, the recovery of Al2O3 and SiO2
increased by increasing the clay content. The concentrations of 6
Al2O3 and SiO2 in the concentrate decreased in the order montmo- 4
rillonite > kaolinite > illite. This corresponds well with the reduc- 2
tion in the copper grade as discussed.
0
0 5 10 15 20 25 30

3.2. Frother adsorption tests 45

40

Mximum Froth height (cm)

The adsorption of frother by the ore and/or clay minerals were
tested. The remaining concentration of MIBC after adsorption tests
for the ore, and different clays is presented in Table 2. It can be
seen that the concentration of MIBC after absorption tests is close
to the original MIBC content of the solution (before absorption) for
all samples. So it can be concluded that, no frother adsorption (by
the ore or clay minerals) has occurred.
3.3. Froth stability and structure
Fig. 5 shows the changes in the froth stability upon the addition
of different clay minerals (using maximum froth height and froth
half-life as proxy). It must be noted that the maximum froth height
and froth half-life indicate different aspects of froth stability. While
the former represents the static stability the latter shows more or
less the dynamic behaviour of froth stability. On the other hand,
the froth height can be considered as an indicator of the amount
of material that may report to the concentrate while the half-life
indicates the structure of the froth itself. A very high half-life
may mean that the froth is too stable and cannot be transported.
In the presence of only 15% montmorillonite, the froth height
increased from 20 to 40 cm, and froth t1/2 from 8 to 14 s. Kaolinite
had more or less the same effect but at higher clay concentrations:
froth height increased from 20 to 30 cm, and froth t1/2 from 8 to
14 s when 30% kaolinite was present. On the other hand, illite
had a negligible effect on the froth stability. Montmorillonite
shows the highest effect on the flotation recovery and grade as well
as froth stability. On the other hand, illite has the least effect on
both flotation performance and froth stability. Kaolinite also affects
the copper flotation grade (but not the recovery) while they also
affect the froth stability.
In addition to froth stability measurement using the froth col-
umn, the surface structure of the froth and (surface) bubble size
during flotation was also analysed using froth images. Fig. 6 pre-
sents a summary of the observations for different clays, and at dif-
ferent levels. It can be seen that kaolinite (at higher concentration)
and montmorillonite affect the bubble size distribution. In fact, the
smallest bubble size is evident when montmorillonite is present.
Table 2
The concentration of MIBC in the supernatant after adsorption tests (the concentra-
tion of MIBC in the solution before the adsorption tests was 0.48 mg/L).
Sample MIBC (mg/L)
Ore 0.47
Kaolinite 0.48
Montmorillonite 0.47
Illite 0.48
Water <0.01
35
30
25
20
15
10
5
0
0 5 10 15 20 25 30
Clay content (%)
Fig. 5. The effects of the different clays (illite j, kaolinite d and montmorillonite
) on the froth stability in froth column (using maximum froth height and froth
half-life as proxy) (the error bars represent 95% confidence intervals).
On the other hand, illite does not seem to have a considerable
effect on the bubble size.
The bubble size distribution of the froth surface in different
conditions is also presented in Fig. 6. Kaolinite and montmoril-
lonite have a different bubble size distribution to illite. When
montmorillonite is present, the bubble size decreases while it
increases when kaolinite presents. In fact, the smallest bubble size
in Fig. 6 is observed in the presence of montmorillonite. On the
other hand, illite does not have a considerable effect on the bubble
size.
In summary, the effect of different clay minerals on the froth
stability is ranked as montmorillonite > kaolinite > illite. Illite
interlayer space is occupied by poorly hydrated K+ ions responsible
for the absence of swelling property. Kaolinite is a non-swelling
clay but can be easily dispersed to the froth zone. In contrast,
montmorillonite has a 2:1 layer structure, a large base exchange
capacity and will readily adsorb Na+ and other cations, all leading
to a high degree of swelling.
3.4. Rheology
Fig. 7 presents the rheograms of the ore slurry (25% solid in
water) in presence of maximum amount of tested clays (15% mont-
morillonite and 30% of kaolinite or illite). It can be seen that the
slurries have non-Newtonian behaviour (pseudo-plastic). The
rheograms exhibit linear shear stress, shear rate behaviour after
an initial shear stress threshold has been reached. This is indeed
the definition of a Bingham fluid. Therefore, the Bingham model
was used to estimate the suspension yield stress in each case.
The Bingham model is represented by Eq. (2) (Hunter, 2001),
where s is the shear stress (Pa), s0 is the Bingham yield stress

Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
6 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx

Ore

Illite
(left to right: 10,
20 and 30 wt%)

Kaolinite
(left to right: 10,
20 and 30 wt%)

Montmorillonite
(left to right: 10
and 15 wt%) Not Available

100

98
Cumulave %

96
Bubble size
distribution 94
of the froth 92
surface
90
10 100 10 100 10 100
size (mm) size (mm) size (mm)
Fig. 6. Froth images from the first flotation concentrate (top surface) in the absence and presence of different clays, with bubble size distribution of the froth surface (ore with
no clays +, ore with illite j, kaolinite d, and montmorillonite ).

(Pa), gp is the Bingham viscosity (Pas) and c is the shear rate
(s  1). Bingham model has been widely used to model rheology
data of mineral suspensions (Bhattacharya et al., 1998;
Farrokhpay et al., 2004; Farrokhpay, 2012b).
Bingham model : s s0 gp c
The yield stress data as obtained using Eq. (2) are presented in Fig. 8
(the ore suspensions, with and without clay minerals). The ore in its
pure form has a simple rheological behaviour, as shown by its low
yield stresses (<1 Pa) at 30 wt%. However, upon addition of the
different clay minerals, an increase in the suspension yield stress is
observed. Illite does not significantly alter the suspension rheology
within the tested concentration. In fact, illite shows similar rheolog-
ical behaviour to the ore suspension. However, the addition of
montmorillonite results in an extremely high suspension yield
2 stresses obtained at only 10 wt%. The pseudo-plastic behaviour
observed at this low concentration can be attributed to the high
swelling capacity of montmorillonite. The other minerals, all with
different variations in platy morphology, have higher critical con-
centrations. Bekkour et al. (2005) demonstrated that rheological
behaviour of montmorillonite is related to the changes occurring

Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx 7

30 30 30
25 25

Shear stress (Pa)

25
20 20 20
15 15 15
10 10 10
5 5 5
0 0 0
0 100 200 300 0 100 200 300 0 100 200 300
Shear rate (1/s) Shear rate (1/s) Shear rate (1/s)
Fig. 7. The rheograms of the ore slurry (25% solid in water) in presence of 15% montmorillonite (left), 30% of kaolinite (middle) and 30% illite (right).

14 (within the tested range). The degree of complexity decreases in

the order of montmorillonite > kaolinite > illite. This is similar to
12 findings observed by Ndlovu et al. (2014) which demonstrated cor-
responding complexity and ranking of clay minerals in their pure
10
Yield Stress (Pa)

form. It should also be noted that all tested concentrations fall

within the range below which complex suspension behaviour is
8
expected (about 25 vol%, (Ndlovu et al., 2014). The complex rheol-
6 ogy behaviour observed for montmorillonite was attributed to its
swelling behaviour as explained. However, the observed increase
4 in the yield when kaolinite was present cannot be simply
explained. This suggests that kaolinite behaviour may be more
2 complex than expected in multi-component systems. The com-
plexity of kaolinite arising from differences in crystallinity has
0
0 5 10 15 20 25 30 been reported (Ndlovu et al., 2015). It should be also noted that
Clay content (%) Dodd et al. (1954) have shown that non-swelling clays also interact
with water, but this interaction is much less than swelling clays. A
Fig. 8. The yield stress of the copper ore suspension at different solid% (+), and in summary of the effect of investigated clays on the copper flotation
the presence of different clays (illite j, kaolinite d and montmorillonite ) (the is presented in Table 3.
error bars represent 95% confidence intervals). The data were obtained by fitting the
rheograms with Bingham model.
4. Conclusions

in the inner structure of the fluid due to the particle interaction
forces like the Vander Waals forces which are responsible for the
formation of flocks and aggregates. According to Murray (1991)
the high charge on the montmorillonite layers is satisfied by water
layers containing cations such as sodium, calcium, magnesium or
iron. These characteristics give montmorillonite, when mixed with
water, a fluid with a high viscosity. Moreover, in contrast to kaolin-
ite and illite, in which the layers are strongly held together, the
montmorillonite layers are held so loosely that water can easily
penetrate between them. Montmorillonite is subject to inter crys-
talline swelling, a process in which the interlayer distance can
increase to 1 nm (Mller-Vonmoos and Loken, 1989). This swelling
behaviour causes the high viscosity of montmorillonite mixture
suspensions.
In summary, the rheological behaviour of the ore suspensions is
altered and becomes more complex with the addition of clays
In general, a common agreement was found between the flota-
tion performance and swelling property of clays. Montmorillonite
has considerable effect on the flotation grade and recovery due
to its high swelling behaviour. Although the effect of slime coating
and/or entrainment should not be ignored.
On the other hand, it is possible to manage the deleterious
effect of illite and kaolinite (within the tested loading levels).
Kaolinite and illite, which do not significantly reduce the copper
recovery, contaminate the copper concentrate by entrainment (at
higher loading of illite compared to kaolinite). The calculated
change in the flotation rate constant also suggests physical adsorp-
tion, or slime coating, of clays.
In conclusion, one should carefully consider the type and
amount of the clays in the ore through systematic mineralogy anal-
ysis, as well as the critical concentrations at which the clays may
cause problem. It should be highlighted that whether or not clays

Table 3
Effect of clays on the copper flotation.

Concentration Cu grade Cu recovery Froth stability Pulp yield stress Overall (within
the tested
concentrations)a
Illite (non-swelling) Up to 30% Considerably Slightly decreases No change No change Manageableb
decreases (185%) (9088%)
Kaolinite Up to 30% Considerably Slightly decreases Slightly increases at higher Slightly increases at higher May cause
(non-swelling) decreases (181%) (9088%) concentrations concentrations problem
Montmorillonite Up to 15% Considerably Decreases (90 Considerably increases Considerably increases Problematic
(swelling) decreases (181%) 80%)
a
These results are valid within the investigated concentrations (for each clays) which are all below their critical concentrations.
b
The behaviour of any clay mineral beyond its critical concentration is unmanageable. Also, the behaviour of these minerals may change once in suspension with other
minerals due to antagonistic and synergistic effects.

Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
8 S. Farrokhpay et al. / Minerals Engineering xxx (2016) xxxxxx
are mineral processing nightmare also depends on their type and
concentration. However, this statement is certainly more applica-
ble for swelling clays.
This paper showed the deleterious effects of swelling and non-
swelling clays in flotation, with discussing some possible mecha-
nisms. However, how to minimize these deleterious effects is
another question. Further investigations in this area are undoubt-
edly warranted.
Acknowledgements
Ms. Elin Tranvik from Lulea University is acknowledged for her
contribution to this project during her training at the JKMRC.
References
Arnold, B.J., Aplan, F.F., 1986. The effect of clay slimes on coal flotation, part I: the
nature of the clay. Int. J. Miner. Process. 17, 225242.
Barbian, N., Hadler, K., Ventura-Medina, E., Cilliers, J.J., 2005. The froth stability
column: linking froth stability and flotation performance. Miner. Eng. 18 (3),
317324.
Bekkour, K., Leyama, M., Benchabane, A., Scrivener, O., 2005. Time dependent
rheological behavior of bentonite suspensions: an experimental study. J. Rheol.
49, 13291345.
Bhattacharya, I.N., Panda, D., Bandopadhyay, P., 1998. Rheological behaviour of
nickel laterite suspensions. Int. J. Miner. Process. 53, 251263.
Boger, D.V., 2009. Rheology and the resource industries. Chem. Eng. Sci. 64 (22),
45254536.
Bulatovic, S.M., 2007. Handbook of Flotation Reagents: Chemistry, Theory and
Practice: Flotation of Sulphide Ores. Elsevier, Boston.
Bulatovic, S.M., Wyslouzil, D.M., Kant, C., 1999. Effect of Clay Slimes on Copper,
Molybdenum Flotation from Porphyry Ores. In: Copper 99-Cober 99
International Conference, Phoenix.
Connelly, D., 2011. High clay ores a mineral processing nightmare. Austral. J. Min.
July/August edition: Miner. Process.-Flot. Sep., 2829
Cruz, N., Peng, Y., Farrokhpay, S., Bradshaw, D., 2013. Interactions of clay minerals in
copper-gold flotation: Part 1 - rheological properties of clay mineral
suspensions in the presence of flotation reagents. Miner. Eng. 5051, 3037.
Dodd, C.G., Conley, F.R., Barnes, P.M., 1954. Clay minerals in petroleum reservoir
sands and water sensitivity effects. Clays Clay Miner. 3 (1), 221228.
Farrokhpay, S., 2011. The significance of froth stability in mineral flotation- a
review. Adv. Colloid Interface Sci. 166, 17.
Farrokhpay, S., 2012a. The importance of rheology in mineral flotation: a review.
Miner. Eng. 3638, 272278.
Farrokhpay, S., 2012b. Rheology of titania pigment slurry. Appl. Rheol. 22, 55285.
Farrokhpay, S., Ametov, I., Grano, S., 2011. Improving the recovery of low grade
coarse composite particles in porphyry copper ores. Adv. Powder Technol. 22,
464470.
Farrokhpay, S., Bradshaw, D., Dunne, R., 2013. Rheological investigation of flotation
performance of a high clay containing gold ore from Carlin Trend. In: World
Gold, Brisbane. The Australasian Institute of Mining and Metallurgy.
Farrokhpay, S., Morris, G.E., Fornasiero, D., Self, P., 2004. Role of polymeric
dispersant functional groups in the dispersion behaviour of titania pigment
particles. Prog. Colloid Polym. Sci. 128, 216220.
Farrokhpay, S., Ndlovu, B., 2013. Effect of Phyllosilicate Minerals on the Rheology,
Colloidal and Flotation Behaviour of Chalcopyrite Mineral. In: Chemeca 2013.
Brisbane.
Please cite this article in press as: Farrokhpay, S., et al. Behaviour of swelling clays versus non-swelling clays in flotation. Miner. Eng. (2016), http://dx.doi.
org/10.1016/j.mineng.2016.04.011
Farrokhpay, S., Zanin, M., 2012. An investigation into the effect of water quality on
froth stability. Adv. Powder Technol. 23, 493497.
Fisher, D.T., Clayton, S.A., Boger, D.V., Scales, P.J., 2007. The bucket rheometer for
shear stress-shear rate measurement of industrial suspensions. J. Rheol. 51 (5),
821831.
Forbes, E., Davey, K.J., Smith, L., 2014. Decoupling rheology and slime coatings effect
on the natural floatability of chalcopyrite in a clay-rich flotation pulp. Miner.
Eng. 56, 136144.
Hunter, R.J., 2001. Foundations of Colloids Science. Oxford University Press, Oxford.
Kasongo, T., Zhou, Z., Xu, Z., Masliyah, J., 2000. Effect of clays and calcium ions on
bitumen traction from Athabasca oil sands using flotation. Can. J. Chem. Eng. 78,
674681.
Keentok, M., Milthrope, J.F., ODonovan, E., 1985. On the shearing zone around
rotating vanes in plastic liquids- theory and experiment. J. Flow Curv. Non-
Newton. Fluids 17, 2335.
Miller, J.D., Lin, C.L., Chang, S.S., 1983. MIBC adsorption at the coal/water interface.
Colloids Surf. 7 (4), 351355.
Mller-Vonmoos, M., Loken, T., 1989. The shearing behaviour of clays. Appl. Clay
Sci. 4, 125141.
Murray, H.H., 1991. Overview-clay mineral application. Appl. Clay Sci. 5, 379395.
Ndlovu, B., Farrokhpay, S., Bradshaw, D., 2013. The effect of phyllosilicate minerals
on mineral processing industry. Int. J. Miner. Process. 125, 149156.
Ndlovu, B., Farrokhpay, S., Forbes, E., Bradshaw, D., 2015. Characterisation of
kaolinite colloidal and flow behaviour via crystallinity measurements. Powder
Technol. 269, 505512.
Ndlovu, B.N., Forbes, E., Farrokhpay, S., Becker, M., Bradshaw, D., Deglon, D.A., 2014.
A preliminary rheological classification of phyllosilicate group minerals. Miner.
Eng. 55, 190200.
Nguyen, Q.D., Boger, D.V., 1985. Direct yield stress measurement with the vane
method. J. Rheol. 29 (3), 335347.
Runge, K., McMaster, J., Wortley, M., La Rosa, D., Guyot, O., 2007. A correlation
between VisiofrothTM measurements and the performance of a flotation cell. In:
9th Mill Operators Conference, Melbourne. Australian Institute of Mining and
Metallurgy.
Sweet, J.A., Harris, M.C., Franzidis, J-P., Plint, N., Tustin, J., 2012. The AGDP 2012-
Nine Years of Exceptional Graduate Training. In: International Mineral
Processing Congress (IMPC), New Delhi.
Tang, F., Xiao, Z., Tang, J., Jiang, L., 1989. The effect of SiO2 particles upon
stabilization of foam. J. Colloid Interface Sci. 131 (2), 498502.
Tao, D., Dopico, P.G., Hines, J., Kennedy, D., 2010. An experimental study of clay
binders in fine coal froth flotation. In: International Coal Preparation Congress,
Lexington, USA. Society for Mining, Metallurgy, and Exploration (SME).
Taylor, R.K., Smith, T.J., 1986. The engineering geology of clay minerals; swelling,
shrinking and mudrock breakdown. Clay Miner. 21, 235260.
Wang, B., Peng, Y., Vink, S., 2013. Diagnosis of the Surface Chemistry Effects on Fine
Coal Flotation Using Saline Water. Energy Fuels 27 (8), 48694874.
Wang, Y., Peng, Y., Nicholson, T., Lauten, R.A., 2015. The different effects of bentonite
and kaolin on copper flotation. Appl. Clay Sci. 114, 4852.
Wei, T., Peng, Y., Farrokhpay, S., 2014. Froth stability of coal flotation in saline water.
Mineral Processing and Extractive Metallurgy (Trans. Inst. Min Metall. C) 123
(4), 234240.
Yan, J., James, A.G., 1997. The yield surface of viscoelastic and plastic fluids in a vane
viscometer. J. Nonnewton. Fluid Mech. 70 (3), 237253.
Zanin, M., Grano, S., 2006. Selecting Frothers for the Flotation of Specific Ores by
Means of Batch Scale Foaming Tests. Metallurgical Plant Design and Operating
Strategies (MetPlant), Perth.
Zhang, M., Peng, Y., 2015. Effect of clay minerals on pulp rheology and the flotation
of copper and gold minerals. Miner. Eng. 70, 813.