271 - PDFsam - Drzymala - Mineral Processing Foundations of Theory and Practice

12.
Flotation
12.1. Theoretical basis

Flotation is one of many methods of separation and can be used for separation of
phases, for instance to remove solid particles or oil drops from water. More frequently
flotation is used for separation of particles having different hydrophobicities. Hydro-
phobicity is a feature of material characterizing its ability to be wetted with a liquid in
the presence of a gas phase. In mineral processing, solids which can be easily wetted
with water are called hydrophilic, while solids with limited affinity for wetting are
called hydrophobic. As a result of hydrophobicity, particles adhere to a gas bubble
forming a particle-air aggregate which is lighter than water, and travels upwards to the
surface of water (Fig. 12.1). Hydrophilic particles do not adhere to the bubbles and fall
down to the bottom of a flotation tank.
water
gas bubble gasgas

bubble
bubble
intergroth
water
wate θ
hydrophobic particle particle
particl
hydrophilic particle
Fig. 12.1. Flotation Fig. 12.2. Contact angle in flotation
Substances can be hydrophobic to a different degree and the measure of their hy-
drophobicity is contact angle. The contact angle (Fig. 12.2) is determined by straight
lines drawn from the point of contact between a particle, gas and water which are tan-
gent to the solid–gas, gas–water, and solid–water interfaces. Commonly, contact angle
is expressed and measured as an angle between gas and solid phases, through the wa-
ter phase. The contact angle can be also defined as the angle between solid and water
phases, through the gas phase. Both ways of expressing contact angle are equally
12. Flotation 271
valid, since the sum of contact angle measured through the water phase, as well as the
angle expressed through a gas phase is 1800. To avoid misunderstanding, information
which phase contact angle is expressed through should be clearly stated.
The contact angle for hydrophilic materials is zero, while for hydrophobic sub-
stances is more than zero. Maximum value of contact angle for materials in contact
with water and air is about 1100 (paraffin, Teflon). Table 12.1 presents contact angles
for selected materials measured by flotometry (Drzymała and Lekki 1989a, 1989b).
Table 12.1. Hydrophobicity of materials. Contact angle is based on flotometric measurements expressed
in degrees
hydrophilic**
Strongly hydrophobic* hydrophobic weakly hydrophobic
θ=0
Material θ Material θ Material θ Material
Paraffin gypsum
90+ sulfides 44–0 fluorite, CaF2 10–13
CnH2n+2 CaSO4·2H2O
Teflon, C2F4 90+ silicon carbide SiC 27.6 arsenic, As2O3 9.3 ferrosilicon
dolomite
Sulfur, S 63.2 coal 26–0 perovskite, CaTiO3 9
CaMg(CO3)2
magnetite
Mercury, Hg 45.6 indium, In 25 scheelite, CaWO4 9
Fe3O4
Germanium, Ge 39.7 iodargyrite, AgI 23.5 diamond, C 7.9 halite, NaCl
Silicon, Si 35.4 cassiterite, SnO2 22– tin, Sn 7.5 brawn coal
Talc 35.2 silver, Ag 14 boric acid, H3BO3 64 kaolinite
ilmenite, Fe 14 graphite, C 6.2+ hematite, Fe2O3
molybdenite,MoS2 5.9+ PbJ2 6 quartz, SiO2
gold, Au 5 calcite, CaCO3
barite, BaSO4 5 anhydrite, CaSO4
corundum, Al2O3 4 bones
HgO 3,3 tourmaline
HgJ2 3 vegetables
copper, Cu 3 iron, Fe
amber
ice, D2O
* Flotometric method is able to measure contact angles smaller than 90o.
** Other hydrophilic materials: chromite, malachite, smithsonite, azurite, rutile, zircon, mica.
Contact angles measured on polished plates made of different naturally hydropho-

bic minerals are usually higher than the flotometric ones. Contact angles of different
materials on smooth plates are presented in Table 12.2.
Numerous measurements of contact angle revealed that most minerals are hydro-
philic. Hydrophilic materials can be rendered hydrophobic with appropriate reagents.
Artificially made hydrophobic materials acquire surface properties similar to those of
naturally hydrophobic substances.
272 Part III. Separation processes
Table 12.2. Advancing contact angle (in degrees) measured on polished plates
(after Adamson, 1967; data from other sources are denoted as *)
Substance Advancing contact angle Substance Advancing contact angle
Teflon 112 sulfur 86*
Paraffin 110 graphite 86
Polystyrene 103 stibnite (Sb2S3) 84
Human skin 90 iodargyrite (AgI) 17
Naphthalene 88 calcite (CaCO3) ~0*
Stearic acid 80 glass ~0
The simplest method of contact angle determination relies on a direct measurement

of an angle of bubble attachment to a mineral surface immersed in water. It is called
the captive bubble method (Fig. 12.3a). When using this method it is important for an
air bubble not to stick to particle edges because both different hydrophobicity of the
edge and additional forces cause distortion of the results. Additionally, to avoid hys-
teresis of contact angle resulting from surface roughness, the surface should be well
polished. Occurrence of hysteresis of contact angle can be detected when contact an-
gle, measured through a water phase, after slight increase in bubble size, shows lower
values than the angle after releasing the gas from a bubble. Minimum angle obtained
during such measurements is called receding or retreating angle, while maximum one
the advancing contact angle. The hysteresis of contact angle can also take place due to
other reasons, for instance surface energy heterogeneity. From a theoretical point of
view the hydrophobicity is best characterized by equilibrium contact angle which can
be rarely determined experimentally. Approximately equilibrium angles can be calcu-
lated by taking arithmetic mean of the advancing and receding angles.
a
b
soild
air
water drop
θ
bubble contact angle θ
water
solid
Fig. 12.3. Contact angle θ can be measured by the captive bubble (a) and sessile drop (b) methods
The contact angle can be also measured using the sessile drop method (Fig. 12.3b).
The sessile drop contact angle usually disagrees with the results obtained by the cap-
tive bubble procedure.
The contact angle depends not only on the method of measurement, physical or
chemical heterogeneity of the surface, but also on such parameters as drop and bubble
12. Flotation 273
size. The value of a contact angle can be also effected by the so-called line tension,
which is an analogue of a surface tension, but characterizing the energy of contact
lines of phases. Therefore, each measuring system of contact angle is specific in some
respect, which, if taken into account, can provide the equilibrium contact angle. Un-
fortunately, exact procedures for recalculating the measured contact angles into equi-
librium contact angles is not known. A list of methods used for measurement of con-
tact angle on smooth surfaces and particles is presented in Table 12.3.
Table 12.3. Methods of contact angle measurements

Smoothly polished
Source* Particles Source
surfaces
Captured bubble Adamson, 1967 heat of immersion Neumann and Good, 1979
Drzymała and Lekki, 1989a,b
Sessile drop Adamson, 1967 flotometry
Drzymała, 1995; 1999a,b,c
shape of border between
Tilted plate in liquid Hiemenz, 1986 Aveyard and Clint, 1995
phases
Force of detachment Hiemenz, 1986 levitation Li and współ., 1993
Heertjes and Kossen, 1967,
Wetting of plate Adamson, 1967 pressed disc
He and Laskowski, 1992
rate of penetration of a
Drop on a tilte plate Hiemenz, 1986 van Oss and współ., 1992
thin layer of particles
Ralson rate of penetration of a Washburn, 1921, Crowl
Drop shape
i Newcombe, 1992 column of particles and Wooldridge, 1967
Ralson captured bubble for parti-
Drop size Hanning and Rutter, 1989
i Newcombe, 1992 cles
Clint and Tylor, 1992;
Langmuir through
Aveyard and współ., 1944
centrifuge Huethorst and Leenaars, 1990
capillary rise without
White, 1982
probe liquid
capillary rise in column of
Bartell and Whitney, 1933
particles with probe liquid
*Additional source: Neumann and Good (1979)
There are many theories binding hydrophobicity (contact angle) with the properties
of multi-phase systems. The principal is the Young formula (1805) which correlates
the contact angle with the interfacial energies (γ) of three-phase system including solid
(s), gas (g), and liquid (c).
γsg = γsc + γcg cos θ, (12.1)
where:
θ – equilibrium contact angle,
γsg, γsc, γcg – interfacial energy for solid-gas, solid liquid, and liquid-gas phase
boundaries, respectively.
Graphical expression of the Young equation is presented in Fig. 12.4.
γsg = γsc+ γcgcosθ

γs - π = γsc+ γcgcosθ
g
γcg vapor
adsorption
liquid drop
c
cieczy γsc γsg
x x x x x x x x x x θ x x x x x
g
ss
Fig. 12.4. Graphical representation of the Young equation
Interfacial energy of a solid–gas system should be expressed for gas saturated with
vapor of the liquid forming the measurement system. It means it should take into ac-
count liquid adsorption on the solid surface. A relation between surface energy of the
material γ and its energy in the presence of liquid vapor γsg is determined by the equa-
tion:
γsg = γs – π, (12.2)
in which π is the so-called film pressure. It is commonly assumed that π is equal zero
for solids having surface energy lower than surface energy of a liquid. In the case of
water it should be lower than 72 mN/m. The so-called low-energy surfaces have sur-
face energy lower than about 70mN/m. For liquids surface energy is measured by sur-
face tension determination. For solids surface energy, i.e. energy necessary to form a
unit surface, is not an equivalent of surface tension (Adamson, 1967). The unit of sur-
face energy is mJ/m2, while equivalent unit is mN/m (1mN/m=1mJ/m2).
The Young equation, although very simple, is not often applied, due to lack of full
data for particular systems. Table 12.4 presents materials for which existing data pro-
vide contact angle using the Young equation. As one can see, this list is very short,
therefore there is a need of carrying out further investigation aiming at determining
surface energy of solids in vacuum and in water, as well as the data regarding water
adsorption on these bodies, which is crucial for determination of the film pressure π.
The π value can be, for instance, evaluated following the relation (Adamson, 1967):
P0
π = γ s − γ sg = − ∫ dγ = RT ∫ Γ d ln P , (12.3)
P
12. Flotation 275
where:
Γ – surface excess
P – vapor pressure
Po – saturation vapor pressure.
According to Adamson (1967) typical π values for solids are up to 100 mN/m.
Table 12.4. Contact angle determined directly from the Young equation
Contact angle in degrees (Drzymała, 1994)
Substance γs, mN/m πe, mN/m γsw, mN/m γcg, mN/m θcalculated θmeasured*
Ice 90–120 ~0 22–33 72.8 0 0
Quartz 120–135 ~small 46 72.8 ~0 0
Paraffin 50– 68 0 51 72.8 77– 91 110
Mercury 484 ~75 415 72.8 95 43–110
* Some works indicate that contact angle of ice (van Oss i współ., 1992) and quartz (Jańczuk i współ.,
1986) can be more than zero
Surface energy of solids is difficult to determine and, therefore the data available in
the literature can include a considerable error. Surface energy of solids can be esti-
mated on the basis of surface tension of the substance melted at higher temperature
and linear extrapolation to a desired temperature, without taking into account possible
non-linear alterations of surface energy during crystallization. Usually, the coefficient
of surface tension changes with temperature dγ/dT is about –0.1 mN/mK (Hutner,
1987). The values of surface energy for some solids are presented in Table 12.5.
Table 12.5. Surface energy of selected crystalline materials.

(Crystallographic planes are given in parenthesis)
γcg, mJ/m2 γcg, mJ/m2
Solid Solid
(Dereń et al., 1977) (Kitchener, 1992)
CaCO3 (100) 230 paraffin 45
LiF (100) 340 polyethylene 55
CaF2 (111) 450 mica 300
MgO (100) 1200 halite 400
Si (111) 1240 fluorite 500
diamond 4000–9000
The literature provides numerous data dealing with surface tension of liquids. The
values of surface tension for selected substances at the temperature of their liquid state
are shown in Table 12.6, while surface tension for selected liquids at room tempera-
ture are presented in Table 12.7.
Table 12.6. Surface tension (γ ) of selected liquids at different than ambient temperature s (after various
sources including Gaudin, 1963; CRC, 1986/87)
Liquid Temperature, °C γ, mN/m
Elements
Nitrogen –193 8.27
Bromine 0 45.5
Chlorine –61.5 31.61
Fluorine –20316 17.9
Argon –188 13.2
Helium –270 0.24
Iodine 130 53.1
Neon –248 5.5
Platinum 2000 1819
Sulfur 112.8 (melting point) 60.9
Sodium 100 206.4
Silver 970 480.3
Oxygen –193 15.7
Hydrogen –255 2.31
Organic compounds
Ethyl alcohol 0 24.05
Benzene 30 27.56
Carbon tetrachloride 200 6.53
Glycerol 150 51.9
Ethylene oxide –20 30.8
Inorganic compounds
Sodium bromide 747 (melting point) 102.8
Sodium chloride 803 113.8
Sodium fluoride 1010 199.5
Cesium iodide 654 73.1
Rubidium iodide 673 79.4
Sodium iodide 661 (melting point) 93.9
Water 20 72.75
Table 12.7. Surface tension of various liquids in contact with their vapors at room temperature
(after CRC, 1986/87)
Liquid γcg, mN/m Liquid γcg, mN/m
Mercury 484 (vacuum) Styrene 32.14
Hydrazine 91.5 Benzene 28.89
Water 72.8 Acetic acid 27.80
Sulfuric acid (98,5%) 55.1 Carbon tetrachloride 26.95
Glycerol 47.7 Acetone 23.70
Aniline 42.9 Ethyl alcohol 22.75
Pyridine 38.0 Oleic acid 32.50
12. Flotation 277
Table 12.8. Interfacial solid-water energy at room temperature. Data collected by Söhnel and Garside,
1992. Data after Stumm and Morgan (1970) are marked as *
Interfacial en-
Interfacial energy
Compound Compound ergy
mJ/m2
mJ/m2
AgBr 65, 110, 140 KNO3 29, 30, 54
AgBrO3 55 K2SO4 23, 29, 41
AgCH3COO 39, 113 MgC2O4⋅2H2O 100
AgCl 57, 88, 92, 103 Mg(OH)2 123
Ag2CrO4 107 MnCO3 104
Ag2SO4 96 MnC2O4⋅2H2O 95
Al(OH)3 25 (pseudo-boehmit); 67(gibbsite) NaCl 38
BaCO3 96, 109, 115 Na2SiF6 52
BaC2O4 . 2H2O 110 Na2S2O3 16
BaCl2 22 NH4Br 14
BaCrO4 120 NH4Cl 27, 30, 32, 60
BaF2 86 (NH4)2Cr2O7 53
BaMoO4 103 NH4H2PO4 4, 8, 12
BaSeO4 88 NH4I 13
BaSO4 117, 135 NH4NO3 27
BaWO4 116 NH4SCN 34
CaC2O4 123 Ni(NH4)2(SO4)2·12H2O 8, 12
CaC2O4 ·3H2O 135 Ni(OH)2 110
CaCO3 58, 68, 97 PbCO3 125
CaF2 167 PbC2O4 139, 145
Ca(OH)2 66, 79, 86 PbCrO4 170, 194
CaSO4·2H2O 76, 92, 95, 117 Pb(NO3)2 3
CaMoO4 118 PbSeO4 71
CaWO4 151 PbSO4 100, 104
CaHPO4·2H2O 46, 68 SrCO3 116
Ca4H(PO4)3·2H2O 59 SrC2O4 76
Ca5OH(PO4)3 45 SrMoO4 100
CdI2 19 SrSO4 82, 92
CuCl2 12 SrWO4 62
FeC2O4· 2H2O 90 TlBr 92
KAl(SO4)2· 2H2O 2, 5 TlCl 40, 92
KBr 15, 23, 24 Tl2CrO4 108
KBrO3 14, 44 TlO3 87
KCl 22, 27, 30 TlSCN 65
KClO3 49 SiO2 (amorphic) 46*
K2Cr2O7 33, 47, 66 glycine 29*
KH2PO4 12, 14, 16 ZnO 0,2MNaClO4 770*
KI 17 CuO 0,2MNaClO4 690*
KIO3 14, 44 Cu(OH)2 0,2MNaClO4 410*
There are also available data on interfacial energy for solid–water systems, espe-
cially for salts (Table 12.8). The interfacial energy for such systems is usually deter-
mined utilizing the solubility measurements for different particle sizes, speed of nu-
cleation, critical supersaturation indispensable for homogenous nucleation, or kinetics
of crystal growth (Söhnel and Graside, 1992).
Another way of determining interfacial energy was proposed by Fowkes (1964):
γ12 = γ1 + γ2 – D – N, (12.4)
where:
γ1, γ2, γ12 – surface energy of phase 1, phase 2 (both in vacuum) and interfacial energy
of phase 1 and 2.
D – dispersion interactions (London-van der Waals),
N – other interactions.
The dispersion interactions D are determined by the relationship:
(
D = 2 γ 1d γ 2d ) 1/ 2
, (12.5)
in which γd1 and γd2 denote dispersion component of surface energy of phase 1 and 2,
respectively. The non-dispersion contribution N can be divided into metallic, hydro-
gen, and other interactions.
After combining the Fowkes and the Young equations, one can obtain the relation:
2 (γ sd γ cd ) − N − π
cosθ = −1. (12.6)
γ cg
Although a number of formulas were proposed for determining N, all of them were
disqualified by Fowkes et al. (1990). Therefore Eq. (12.6) is useful for calculation of
contact angle for substances interacting with water only due to their dispersion forces,
since then N=0.
As it has already been stated the dispersion component for water is 21.8 mN/m,
while for the non-dispersion component is 51mN/m. For other substances they can be
determined from contact angle measurement using certain standard liquids or esti-
mated according to the so-called Hamaker constants and relation (Drzymała, 1994c):
A11 ⋅ 10 13
γ = d
0.20527 . (12.7)
The values of the Hamaker constant are known for many substances and some se-
lected ones are presented in the chapter dealing with flocculation. Table 12.9 presents
hydrophobicity of materials for which contact angle can be calculated from the
Fowkes-Young equation.
12. Flotation 279
Table 12.9. Contact angle for nonpolar materials calculated by means of the Fowkes–Young equation
(12.6). γ d for water is 21.8 mN/m, N = 0 (Drzymała, 1994c)
Solid γ d, mN/m π, mN/m θcalculated θmeasured
Paraffin 25.5 ~0 110 110
Teflon 21.7 ~0 113 112
Mercury 200 ~75 112 43–110
Graphite 115–132 19 81 86 (plane 0001)
Silver 194.9 0 (?) 38 57–62
There are other possible ways of contact angle calculation. Among them there is
the so-called equation of state of Neumann (1974), which, after several modifications,
has the form (Spelt and Li, 1996):
γ sg − β (γ cg −γ sg ) 2
cos θ = −1 + 2 e , (12.8)
γ cg
where β is a constant equal to 0.000115 while cg and sg denote liquid-gas and solid-
gas interfaces, respectively.
This equation has been frequently criticized (Fowkes et al. 1990; Johson and Det-
tre, 1989) and therefore it is not often used, except by Neumann and his team. Another
equation has been proposed by van Oss et al. (1988; van Oss and Good 1996).
(1 + cos θ )γc = 2[( γ cLWγ sLW )1/2 + ( γ s γ c )1/2 + ( γ s γ c )1/2],
+ − − +
(12.9)
where: c – liquid , s – solid, γ LW, γ + and γ – stand for Lifshitz–van der Waals, electron-
accepting, and electron-donor components, respectively.
It has been assumed that for water γcLW=21.8 mJ/m2, γc+=25.5 mJ/m2, γc-=25.5
mJ/m2. However, some works suggest that these constants should have other values
(Lee, 1996). The values of γLW, γ+, and γ- can be determined with the use of standard
liquids and contact angle measurements. The van Oss et al. equation has been the sub-
ject to experimental verification and criticized (Lyklema, 1999; Kwok, 1999). In Table
12.10, the values of γLW, γ+ and γ- were presented for several liquids. They are pres-
ently considered as standard liquids.
Table 12.10. γ LW, γ + and γ – values for selected liquids (Chibowski et al., 1996)
Liquid γ γ LW γ+ γ–
Water 72.8 21.8 25.5 25.5
Diiodomethane 50.8 50.8 0 0
α-bromonaphthalene 44.4 44.4 0 0
Formamide 58 39 2.28 39.6
Glycerol 64 34 3.92 57.4
Ethylene Glycol 48 29 1.92 47.0
Decane 23.8 23.8 0 0
According to Gaudin et al. (1957), solids build of molecular (iodine, sulfur, realgar,
paraffin), chain (Ag, selenium), and layered (molybdenite, graphite, talc, boric acid)
structures are hydrophobic. In these structures molecules, chains, and layers are bound
together with the van der Waals-London forces. The rule is not entirely correct, since
there are solids, for example iodargyrite (AgI), which has an ionic-covalent structure
of zinc blende and its contact angle is not zero but about 170. On the other hand SnCl2
has a chain-like structure but is, according to flotometric tests, hydrophilic. Gaudin’s
rule does not say anything about substances having metallic bond. The metals can be
either hydrophilic (Fe, Al, Au) or hydrophobic (Ag, Cu, In). On the bases of existing
data it can be said that hydrophilic are those substances on whose surfaces absorbed
water becomes chemically bound. Non-ionic surface structures (layer, chain or mo-
lecular nets) usually prevent water form being chemically bound, which provides hy-
drophobicity of the substance. It should be noticed that hydrophilic substances are ca-
pable to form hydrates (Drzymała, 2001). Therefore, oxides are mostly hydrophilic as
they form hydroxides. Hydrophilic are many salts since they can form hydrates. For
instance, NaCl or CaCO3 are hydrophilic because they form hydrates NaCl.2H2O (hy-
drohalite) and CaCO3.H2O (monohydrocalcite) (Bolewski and Manecki, 1993; Wells,
1993), respectively. Thus, the salts which do not form hydrates can be hydrophobic,
for instance, barite (BaSO4) or iodargyrite (AgI) (Drzymała 1994c). Following this
hypothesis, hydrophilic can also be the substances of molecular, chain or layer struc-
ture if, when reacting with water, they form hydrophilic hydrates. This rule, although
varying in its form, also is valid for metals, since metals, on whose surfaces reaction
with water takes place leading to the formation of surface -OH and –H groups, are hy-
drophilic (Drzymała (1994c).
12.2. Hydrophobicity modification

Hydrophobicity of the materials is effected by reagents (acids, bases, salts, organic
compounds) added to the solid-water system. Their action can be explained with the
Young equation, since they modify all three interfaces. As a rule, however, one of the
interfaces is mostly effected. The compounds modifying hydrophobicity and floata-
bility of naturally and rendered hydrophobic materials can be classified into four
groups: collectors, hydrophilization reagents, electrolytes (potential determining sub-
stances and salts), and modifiers (activators, depressors, frothers). Each of these
groups differently effects hydrophobicity.
Collectors increase surface hydrophobicity and contact angle mainly through the
decrease in surface energy of the treated material, i.e. γsg. This fact was proved ex-
perimentally by Smolders (1961) and theoretically by de Bruyn, Overbeek and Schu-
mann in 1954 (de Bruyn and Agar, 1962). Hydrophilization compounds decrease hy-
drophobicity mostly by decreasing surface tension of the liquids, whereas electrolytes
can either increase or decrease contact angle by effecting mainly the solid-water inter-
12. Flotation 281
face. The effect of modifiers (activators, depressors, frothers) on properties of inter-

faces is not well established. It is schematically presented in Fig. 12.5.
collector
γsg, γ sc , γ cg
θ contact angle ,
γ −γ
γsc, γ
sg sc
initial cosθ = --------------
contact angle γ cg cg , γ sg electrolytes
(pH regulators and
salts)
0 γcg, γ sc , γ sg
modifying
reagents
hydrophilization
reagents
Fig. 12.5. Flotation is influenced by reagents modifying the interfaces. Collectors strongly change the
solid-gas, surfactants water–gas, and electrolytes (pH reagents and salts) solid–water interfaces. The
extent of modification is expressed by the height of symbol
When hydrophobicity is regulated with the concentration of the collector, the inter-
facial energies of the solid-gas, solid–water, and water-gas interfaces decrease. How-
ever, the most significant is the drop in the interfacial energy of the solid-gas interface
(Fig.12.6). According to the Young equation it leads to the increase in hydrophobicity
and improves flotation. This is well visible in Fig. 12.6 for mercury.
Hydrophobizing substances, i.e. collectors, can increase the contact angle up to
1100, which is the value characteristic for paraffin. Generally high values of contact
angles can be obtained with increased concentrations of collectors and collectors with
larger hydrocarbon chain are more powerful.
450 90
θ
Fig. 12.6. Dependence of contact angle
425 and the state of mercury at interfaces on
surface or interfacial
70 concentration of collector Data by

tension, mN/m
contact angle, degree

surface tenion, mN/m
H2O Smolders (1961) taken from various

400
sources: contact angle in dodecyl sulfate
Hg/air solution and γ H2O (dodecyl sulfate) (Leja,
50
375 1982), γ Hg /H2O(decyl sulfonate) and γ Hg (decyl
Hg/H2O sulfonate) (de Bruyn i Agar, 1962))
350 30
-7 -6 -5 -4 -3 -2 -1
log (collector concentration, mol/dm3)
Investigation relating hydrophobicity and surface tension of the liquid was initiated
by Zisman (1964). The relation between contact angle of a solid and surface tension of
the liquids is called the Zisman plot. It is used in film flotation process, which relies
on placing fine particles on liquid surface and collecting floating particles. The surface
tension of the flotation medium is regulated by adding the liquid of lower-than-water
surface tension. The so-called gamma flotation is similar to the film flotation. In
gamma flotation particles are floated with gas bubbles passing through the aqueous
solution whose surface tension is regulated with low surface tension liquid. The film
flotation method was widely investigated by Fuerstenau et al. (1988) and Sablik,
(1998). It is applied for determination of the so-called surface tension of wetting or
critical surface tension (γc). This parameter allows to compare flotation properties of
different substances. Gamma flotation in aqueous solutions of alcohols (usually ethyl)
was initiated by Yarar and his co-workers (Yarar and Kaoma, 1984).
It should be stressed that γc is not equivalent to the surface energy. It is easy to
prove by determining γc for a high surface energy hydrophobic material such as mer-
cury. Its γc is below 72.8 mN/m while it surface tension is 484 mN/m. Fig. 12.7 shows
an agreement of contact angle (Zisman’s diagram) with surface flotation as a function
of surface tension of the solution.
1 100
ε
0,75 75 flotation recovery, ε , %
cosθ
0,5 cosθ 50
0,25 25
surface tension
of wetting
0 0
0 20 40 60 80
surface tension, γ wg, mN/m

Fig. 12.7. Typical shape of the cos θ = f (surface tension of liquid) relationship also called the Zisman
plot, and flotation of naturally hydrophobic materials
Electrolytes effect mostly the solid–water interface. Their action is complex and
depends on concentration and properties of the electrical double layer. As a result of
adding electrolytes the flotation can increase or decrease. If acid, or base, is used as
electrolyte and the pH regulation medium, change of hydrophobicity depends on the
electric properties of the solid–water interface. It appears that any deviation from pH
at which the interface is electrically neutral, causes a decrease in the hydrophobicity
and flotation. The relation between contact angle (or flotation) and pH can be called
12. Flotation 283
the electrocapillary curve of a second kind. For example, Fig. 12.8 shows a decrease in
germanium flotation, which occurs above and below pH 2.8. At that pH the interface
is electrically neutral.
100
80 Ge
flotation recovery, %
pHiep = 2.8
60
40
20
0
0 2 4 6 8 10 12
pH
Fig. 12.8. Influence of pH on germanium flotation

(after Drzymala et al., 1987)
If salt is applied as the electrolyte, hydrophobicity and flotation decrease or in-

crease, depending on the flotation system environment. This question will be dis-
cussed in details in a next chapter. Figure 12.9 shows changes of coal flotation as a
function of NaCl concentration.
100
80
coal
60
40
20
≈
0
∞
-6 -5 -4 -3 -2 -1 0
0.1 1,00
0.3 2
0.5
log (NaCl concentration, mol/dm 3) NaCl concentration
mol/dm 3
Fig. 12.9. Coal flotation containing 80.65% C as a function of sodium chloride concentration in aqueous
solution(based on data of Li and Somasundaran, 1993)
To make the reader more familiar with the terms used in different fields Fig. 12.10
presents schematically the influence of the factors discussed on hydrophobicity and
flotation.
hydrophobization
reagents
θ hydrophobicity,
electrolytes
initial hydropho- (acids, bases, salts, potential
hydrophobicity bicity determining species)
0
low surface tension reagents
hydrophobization
contact angle, θ
initial contact angle electrocapillary curve

contact angle
Fig. 12.10. Influence of various
factors on hydrophobicity, con-
0
tact angle, and flotation along
Zisman effect with terms used in different
fields:
b
a – colloidal physical chemistry
collector
(delineation of hydrophobicity),
b – colloidal physical chemistry
(phenomena),
c – flotation (applied reagents),
flotation
d – in practice (type of flotation)

initial flotation electrolyte
flotation (pH reg., salts)
0
reagent with lower than
water surface tension
collector flotation
flotation
initial salt flotation and

flotation
flotation flotation with pH
adjustment
0
film flotation
d
12. Flotation 285
12.3. Electrical phenomena at interfaces

Interactions between two separated or being in contact objects can be described as
a sum of different, including electrical, interactions. The theory and practice indicate
that electrical interactions are significant in the processes taking place in different sys-
tems including flotation. Therefore, this question will be discussed in more details.
Introduction of any liquid, solid, or gas as a phase into water causes creation of an
interface and formation of electrical double layer. The term electrical double layer re-
flects the situation that the electrical charge present within or close to one phase is
neutralized in the second phase. Together they form electrical double layer (Fig.
12.11) which is often abbreviated to edl.
There are many models of edl and they differently express electrical properties of
the interface. The main purpose of the models is to relate parameters of edl, especially
surface electrical charge and surface potential. Figure 12.2 presents the mostly used
edl models.
Description of electrical properties of interfaces can be based on the Helmholz (flat
capacitor) or Gouy-Chapman (one plate of the capacitor is diffused) models, or their
combinations. A simple combination of the Helmholz and Gouy-Chapman models re-
sults in the Stern model while two rigid and one diffuse layer is the essence of the
Graham model. Unfortunately, the simplest Helmholz model is capable to describe
only a few systems with a very low surface electrical charge. Otherwise the Gouy-
Chapman model is used as in the case of the AgI/H2O interface. Metal oxides in water
usually require more complex edl models such as the Graham and more complex theo-
ries. Among them is the site bonding model (SBM) and its modifications such as sur-
face complexation model (SCM) which incorporate the Helmholz and Gouy-Chapman
models. More details regarding the edl models can be founds in a number of works
including a review work by Janusz (1999).
The Helmholz model considers edl as a flat capacitor, for which the equation relat-
ing the electrical charge (σ0) and potential (ψ0) has the following form:
σo = εεoκψo, (12.10)
where:
ε – dimensionless dielectric constant
εo – permittivity of free space (8.854⋅10–12 C2 N–1 m–2)
1/κ – electrical double layer thickness.
+ - + -
- + - +
charged surface +
layer + + - + -
-
+ - + - +
-
particle - - + - + -
- + - + - +
- +
- - + - + -
+
+ - + - +
Galvani potential, Φ 0 - + - + potential Ψ
diffusive layer
surface potential, Ψ 0 cation concentration
zeta potential, ζ eg. [H+] = [H+] r exp (−Ψ /RT)
anion concentration
eg. [OH-] = [OH-] r exp (Ψ /RT)
Fig. 12.11. Electrical double layer (edl) model consisting of a charged surface layer within the solid and
neutralizing it diffused layer in the solution. The Galvani potential between two phased is also shown as
well as location of the slipping plane determining the zeta potential. Hr and OHr symbols denote concen-
tration of H+ and OH– ions in the solution H and OH without subscript r stand for concentration of the
ions at the surface
triple
Helmholz Stern layer
(flat capacitor) (rigid and diffuse layer )
Gouy-Chapman quadruple
Grahame layer
(diffuse layer)
(binding sites)
Ψo Ψo Ψo Ψo Ψo Ψo
Ψo
Ψi Ψi
Ψi
Ψd Ψd
σo -σo σo σd -----→ σo σi σd-----→ σo σi σd----→ σo σi σj σd→

H+ H+ H+ H+ K+ H+ K+ K+H2O
OH- OH- OH- OH- A- OH- A- A-H2O
Fig. 12.12. Models of electrical double layer (edl). The layers i, j, ..., form flat capacitors and can be
treated as connected consequently or parallel. σ0 = –σi – σj – σd. K+ – cation, A– – anion
12. Flotation 287
The Gouy–Chapman model assumes that the electrical charge σd is gathered in the
solution to neutralized the electrical surface charge within or near the surface, does not
form a uniform layer but is diffused. It is given by (Hunter, 1987):
− 2κ kT ε 0 ε ⎛ ze ψ ⎞
σd = sinh ⎜ 0 0 ⎟ , (12.11)
ze ⎝ 2 kT ⎠
where:
k – Boltzmann constant
T – absolute temperature
eo – elemental charge
z – valence of potential determining ions (positive number)
ψd – electrical potential of diffuse layer
σd – charge in the solution (σs = –σd).
For a symmetric electrolyte in water at 250C (298K) the electrical charge in the
Gouy-Chapman diffuse layer can be calculated from the relation:
σd = –11.74 c0.5sinh (19.46 zψo), (12.12)
where c is the concentration of electrolyte in solution, expressed in mol/dm and ψo is
3
the surface potential (in volts).

The Gouy-Chapman edl model is shown in Fig. 12.11 with marked potential differ-
ences between the interior of the solid phase and the liquid phase, called the Galvani
potential (Φo), as well as surface potential (Ψo). The Galvani potential is equal to the
sum of surface potential and the potential drop transition between surface and bulk
solution, called the χ potential. Figure 12.11 also shows the zeta potential which plays
an important role not only in mineral processing. The zeta potential (ζ) is an electrical
potential at the place called the slipping plane, where a particle moving in water leaves
a part of its edl behind. A moving particle renews its edl very soon and the time
needed to achieve it is called the relaxation time. If edl renewal is impossible, as e.g.
during the fall of melting hail balls, then electrification takes place, in this case of the
Earth and clouds, leading to lightning (Drzymala, 2001).
The structure of the electrical double layer in water depends on the mechanism of
formation. In the case of metals whose surface does not react with water (precious
metals and mercury), their edl is formed due to electrons gathered at the interface on
the side of metal. This phenomenon results from either transition of some amount of
metal to the solution in the form of cations, or application of external potential to the
metal. Then, the metal surface becomes negatively charged. A positive charge on
metal surface occurs when electrons are removed from metal. The formation of elec-
trical charge in edl on metal surface is shown Fig. 12.13.
surface charge
negative positive
metals
⏐ ⏐ ⏐ ⏐ ⏐ ⏐ ⏐ ⏐⊕
⎯ Me ⎯ Me ⎯ Me ⎯ Me ⎯ H2O ⎯ Me ⎯Me ⎯ Me ⎯ Me
⏐ ⏐ ⏐ ⏐ ⏐ + ⏐ ⏐
⎯ Me ⎯ Me ⎯ Me ⎯ Me ⎯ - + n Me or ⎯ Me ⎯ Me + electrons
⎯ Me
⏐ ⏐ ⏐ ⏐ ⏐ ⏐
⏐
⎯ Me ⎯ Me ⎯ Me ⎯ Me ⎯ ⎯ Me ⎯ Me
⎯ Me ⎯Me
⏐ ⏐ ⏐ ⏐ ⏐ ⏐⊕
⏐ ⏐
oxides
⏐ ⏐ ⏐ ⏐
⏐ ⏐
⎯ Me ⎯ O ⎯ Me ⎯ O ⎯ H2O - +
⏐ ⏐ ⏐ ⏐ ⎯ O ⎯ Me⎯O ⎯ O ⎯ Me⎯OH2
⎯ O ⎯ Me ⎯ O ⎯ Me ⎯ ⏐ + ⏐ -
⏐ ⏐ ⏐ ⏐ ⎯ Me ⎯ OH +nH or ⎯ Me ⎯ OH + n OH
⎯ Me ⎯ O ⎯ Me ⎯ O ⎯ ⏐ ⏐
- +
⏐ ⏐ ⏐ ⏐ ⎯ O ⎯ Me⎯O ⎯ O ⎯ Me⎯OH2
⏐ ⏐
salts
⏐ ⏐ ⏐ ⏐ ⏐ ⏐
⎯ Me ⎯ X ⎯ Me ⎯ X ⎯ H2O ⎯ Me ⎯ X ⎯ X ⎯ Me
⏐ ⏐ ⏐ ⏐ ⏐ ⏐
- + + -
⎯ X ⎯ Me ⎯ X ⎯ Me ⎯ ⎯X + n Me or ⎯ Me +nX
⏐ ⏐ ⏐ ⏐ ⏐ ⏐
⎯ Me ⎯ X ⎯ Me ⎯ X ⎯ ⎯ Me ⎯ X ⎯ X ⎯ Me
⏐ ⏐ ⏐ ⏐ ⏐ ⏐
place of particle breakage particle /water interface

miejsce przełomu ziarna
Fig. 12.13. Formation of surface charge on metal, oxide and salt
A different electrical double layer is formed on particles which do not conduct

electrical current (Fig. 12.13). The cause of electrical charge on the surface of salts are
ions passing from the salt into the solution. For instance, for NaCl they will be Na+
and Cl- ions. The more ions of one kind pass into the solution, the higher the surface
electrical charge. The electrical charge can be further increased by adding to the solu-
tion another soluble salt having at least one identical ion as the solid particle. The ions
12. Flotation 289
which form the salt and regulate the electrical charge are called the potential deter-
mining ions.
⏐ ⏐ ⏐ ⏐ ⏐
⎯ Me ⎯ S ⎯ Me ⎯ S ⎯ ⎯ S ⎯ Me⎯OH
⏐ ⏐ ⏐ ⏐ H2O ⏐
⎯ S ⎯ Me ⎯ S ⎯ Me ⎯ ⎯ Me ⎯ SH
⏐ ⏐ ⏐ ⏐ ⏐
⎯ Me ⎯ S ⎯ Me ⎯ S ⎯ ⎯ S ⎯ Me⎯OH
⏐ ⏐ ⏐ ⏐ ⏐
other
⏐ ⏐
+ ⎯ S ⎯ Me⎯OH
⎯ S ⎯ Me⎯OH2
⏐ ⏐
- +
⎯ Me ⎯ SH + n OH ⎯ Me ⎯ S- +H
⏐ ⏐
+ ⎯ S ⎯ Me⎯OH
⎯ S ⎯ Me⎯OH2
⏐
⏐
Fig. 12.14. Formation of surface charge on sulfide
In the case of oxides the edl forming is more complicated than on salt type miner-
als. This is so because the original (primary) potential determining ions are Me (Me+,
Me2+, Me3+) cations and O2- anions. However, the O2- ions are not present in the solu-
tion. Instead, OH- ions are present as a result of reactions between O2- surface ions and
water. Therefore, OH- ions, as well as related to them (due to water molecules disso-
ciation) H+ cations, are the secondary potential determining ions for metal oxides. The
OH- and H+ ions are the secondary potential determining ions for all substances whose
surfaces undergo reaction with water. This refers also to sulfides (Fig. 12.14) for
which the original potential determining ions are S2- and Men+, while the secondary
potential determining ions are OH- and H+. Figure 12.15 shows a diagram for zinc sul-
fide indicating the areas of dominance of different surface species as function of con-
centration of H+ and HS- ions. For some solid–water systems the potential determining
ions can be still other species, for example ions introduced to the solution to regulate
pH and ionic strength of the system, that is sodium, potassium, halide, and other ions.
The potential determining ions can be also ions present in water as contaminants.
-2
-
-SH -SH -S
-
-SZnSH -SZnS -SZnS -
log (HS)tot -3
-4
-SH -SH -SH -
-S
-SH + Zn ++ +
-SZnOH 2 -SZnOH -SZnOH
-5
-6
4 6 8 10 12
pH
Fig. 12.15. Diagram of dominance of surface species on zinc sulfide (sphalerite)

depending on total amount of HS– and pH. Assumptions: (–SZn)tot = 1.0 mmol/dm3
(after Rönngren et al., 1994)
Potential determining ions regulate the electrical surface charge of materials in wa-
ter. The net charge can be either positive or negative, or sometimes zero. Fig. 12.16.
presents a typical shape of electrical charge plotted as a function of pH of the solution.
60
surface electrical charge, µC/cm 2
40
20 pzc
-20
0.001M NaCl
-40 0.010M NaCl
0.100M NaCl
-60
2 4 6 8 10
pH
Fig. 12.16. Surface charge vs. pH of solution and concentration of electrolyte in solution.
The pzc is also shown
12. Flotation 291
The point at which the electrical surface charge equals zero is a very important
property of the system and is called the point of zero charge or shortly pzc. Table
12.11 presents pzc values for different substances. Fig. 12.16 shows that the charge
depends on the concentration of salts (ions) present in the solution. It should be no-
ticed that for a particular pH the higher electrolyte concentration in the solution, the
higher values of electrical charge.
Table 12.11. Point of zero charge (pzc) and isoelectric point (iep)
for selected substances in aqueous solution. The values of pzc and iep
for a great number of solids were collected by Parks (1965) and Kosmulski (2001)
Substance pHpzc pHiep
Quartz, SiO2 <5 1.54
Oleic acid, C17H33COOH 2.0
Cassiterite, SnO2 <5.5 2.0–55
Sulfur, S – 2.1
Sulfides, MeS – 2.1–7.0
Ice, D2O 7.0 ±0.5 3.0–3.5
Hydrocarbons, CnH2n+2 6.3 3.3
Air, O2+N2+CO2 – 3.5
Diamond, C – 3.5
Bacteria (Nocardia) – 3.5
Rutile, TiO2 4.8–5.3
Ilmenite, FeTiO3 5,6
Hematite, Fe2O3 6.5–8.5 4.8–8.7
Barite, BaSO4 – 6.0–8.1
Tenorite, CuO 6.5–8.5 6.0–7.6
Dolomite, (Ca,Mg)CO3 – 7.5
Magnesite, MgCO3 7.5
Corundum, Al2O3 9.1
Periclase, MgO 12.0
The electrical state of the surface can also be determined by measuring either sur-
face potential or zeta potential. Fig. 12.17 shows a typical pattern of the relation be-
tween zeta potential and pH at different concentrations of an electrolyte in the solu-
tion. The point at which the surface assumes zero potential is called the isoelectric
point or iep. Typical pzc and iep values for different substances are presented in Table
12.11. It can be concluded from the table, the iep not always is equal to the pzc. It re-
sults from difference in the adsorption of cations and anions in the edl coming from
the electrolyte used, especially close to the pzc.
60
40
zeta potential, mV
20
iep
-20
0.100M NaCl
-40 0.010M NaCl
0.001M NaCl
-60
2 4 6 8 10
pH
Fig. 12.17. Zeta potential as a function of pH and concentration of electrolyte in the solution. The
isoelectric point (iep) is also shown
For a particular concentration of the potential determining ions the zeta potential
decreases as ionic strength of the solution, determined by the electrolyte concentration,
increases. This results from the compression of the electrical double layer leading to a
greater decrease in the potential with the distance from the surface. For metals the
electrical charge can also be regulated by applying external potential. At a certain
magnitude of the applied potential the metal reaches zero charge and this point is
called the potential of zero charge or Epzc. The value of Epzc is connected with the so-
called work function (w) that is the work needed to remove electron from metal. The
relationship is (Jakuszewski 1962):
Epzc = w – 4.78, (12.13)
where Epzc is expressed in relation to so-called hydrogen electrode. Both Epzc and w are
expressed in volts. It should be added that the curve describing the relation between
metal surface tension (surface energy) and applied external potential E is called the
electrocapillary curve, and is given by the Lippmann equation:
∂γ /∂E = –σ. (12.14)
where σ stands for surface charge and γ for surface tension.
The electrical charge has a significantly effect on contact angle and flotation. Ex-
aminations of flotation data indicate that highest contact angles and best flotation oc-
curs at Epzc, pzc or iep. Figures 12.8 - 12.19 show flotation results of different materi-
als, whose electric state of interface was regulated either by external potential (for
conducting substances) and by primary or secondary potential determining ions.
12. Flotation 293
120 30
pAg iep = 5.3

contact angle, θ , degree
100
contact angle θ , degree

Epzc
80 20
60
40 Hg 10
AgI
20
0 0
0 -0.5 -1 -1.5 -2 -2.5 2 4 6 8 10 12 14
applied potential, V pAg
a b
100
pHiep <3
80
60
40 talc
20
0
2 4 6 8 10 12 14
pH
c
Fig. 12.18. Influence of potential determining factors on contact angle and flotation:
a – mercury, depending on applied potential (activity of electrons)
(after Nakmura et al., 1973), b –iodargyrite (silver iodide) depending on concentration of potential deter-
mining ions (Ag+) (Billett et al., 1976),
c – talc (Mg3[(OH)2Si4O10]) depending on concentration of H+ ions (Chander et al., 1975)
100 100
80 pHiep = 2.8 Ge 80
pHiep = 3
60 60
40 40
20 20 molibdenit e
0 0
0 2 4 6 8 10 12 2 4 6 8 10
pH pH
100
80
94%
pHiep, silane-treated SiO2 = 2÷3
60
pHiep, SiO2 = 2÷3
40
silane-treated
20 quartz
27%
0
2 4 6 8 10 12
pH
Fig. 12.19. Influence of pH on collectorless flotation of various hydrophobic and rendered hydrophobic
materials: a – germanium (Drzymała et al., 1987),
b – molybdenite (Chander et al., 1975), c – quartz covered with ether 27 or 94% of silane hydrophobic
groups (Laskowski and Iskra, 1970)
A precise description of the effect of electrical charge on wetting angle and flota-
tion is not simple, since any alteration of the solution composition causes changes at
all three interfaces influencing the contact angle and flotation. For naturally hydro-
phobic substances it can be assumed, without a considerable error, that the solid–
water interface is the most sensitive to the changes of the electrical charge. Therefore
the surface energy γsc of this interface can be split into a chemical component γ0sc and
the part dealing with the effects involving electrical double layer Fedlsc (Ralston and
Newcombe, 1992). This can be written as:
γsc = γ sc0 + Fscedl . (12.15)

After inserting this relation into the Young equation we get an expression for a
wetting angle, measured through the water phase, depending on potential determining
12. Flotation 295
factor X, which causes the increase of the electrical charge and electrical potential at
the interface:
γ sg − (γ sc0 + Fscedl ( pX )) F edl ( pX )
cos θ ( pX ) = = cos θ pzc − sc
γ cg γ c cg , (12.16)
where pX denotes negative logarithm of X ions concentration (pX=-log[X]).

Basing on the electrocapillary data and other measurements enabling determination
of interfacial energy it can be concluded that since Fedlsc can be only negative, there-
fore, according to equation (12.16) the lowest value of cos θ, i.e. the largest contact
angle occurs at pzc, where Fedlsc = 0. Any decrease of the solid–liquid interfacial en-
ergy results in the decrease of the contact angle values, i.e. decreased hydrophobicity.
The value of Fedlsc depends on the system subjected to flotation, that is on the kind
of material and the method of regulation of the electrical charge. If electrical charge is
regulated using external source of electrons, as in the case of mercury, then it is differ-
ent from substances whose surface potential is determined by the Nernst equation, and
still different when the surface is described as a non-nerstian one. Table 12.12 presents
Fedlsc expressions useful in description of different flotation systems.
Table 12.12. Expression for calculating Fscedl , useful for determination of influence of different potential
determining species on contact angle in flotation systems by means of Eq. 12.16
(after Ralston and Newcombe, 1992)
Potential determin- edl
System Example Fsc
ing species
E − E0
Ideally polarized body electrons mercury–water F edl

sc
=−
∫ σ dE ***
E0
0
ψ0
Nernstian surface*
primary potential AgI -Ag+ aqueous
Fscedl
∫
= − σ 0 dψ
0
determining ions solution
8n0 kT ⎛ zeψ 0 ⎞
Fscedl = − ⎜ cosh − 1⎟⎟ **
κ ⎜⎝ 2kT ⎠
ψ0
solids–aqueous
∫
Fscedl = − σ 0 dψ +Fsw
edl
Non-nernstian surface secondary ions solution with regu- ,konf ( pX ) ****
lated pH 0
* For nernstian surface ψ 0 = f (pX ) fulfills the Nernst equation.
** For edl fulfilling the Gouy–Chapmann equation.
*** Modified equation of the Lippmann electrocapillary curve relating surface potential change and
applied potential change.
**** Fscedl (pX ) is a correction for entropy resulting from the structure of edl.
As it has already been stated, the occurrence of surface charge for naturally hydro-
phobic and rendered hydrophobic (outside the flotation environment, e.g. silane-
treated quartz) worsens flotation (Fig. 12.19c). However, if hydrophobization is not
permanent but resulting from the reaction between collector ions on the surface di-
rectly in the solution, the effect of concentration of potential determining ions be-
comes more complicated, since apart from diminishing hydrophobicity and flotation,
which results from electrical effects, there also is competition between potential de-
termining and collector ions for the surface.
The above discussion does not take into account the properties of solid–gas (air)
and water–gas interfaces. However, both interfaces can contribute to the hydrophobic-
ity of flotation systems. The solid–air interface can, as the pH and ionic strength
changes, alter their surface energy due to adsorption of water vapor, which leads to π
change. The water–air interface, in turn, has its iep near pHiep=3.5 (Tab. 12.11). Thus,
the electrical charge of this interface also depends on pH. Similarly, the surface ten-
sion of aqueous solutions depends on the electrolyte concentration, i.e. pH and salts
regulators. At low electrolyte concentrations surface tension decreases, which is con-
nected with the increase of electrical charge of the interface and it is predicted by the
Lippmann equation (12.14), as well as by properly solved Gibbs equation for solutions
of different pH (de Bruyn and Agar, 1962), or for salts (Ratajczak, 2001). The effect
of electrolytes on the decrease of surface tension finds its proof in the measurements
and it can be recorded for all kinds of interfaces. For the water–air interface the pres-
ence of small amount of salts results in the so-called Jones-Ray effect (Fig. 12.20a)
(Adamson, 1967). At higher electrolyte concentrations the surface tension of aqueous
solutions of acids (HCl, HNO3, HClO4) decreases, while for bases and most salts in-
creases (Fig. 12.20b). There are several theories trying to explain this phenomenon,
which are mainly based on the so-called negative adsorption (Pugh et al., 1997). Quite
probably there is simpler explanation, proposed by Gaudin (1963), that the increase in
water surface tension results from additivity of surface tensions of water and electro-
lyte with certain involvement of the positive and negative adsorption and electrical
charge. Indeed, surface tension of pure (undiluted) acids is lower than that of water,
while salt surface energy is higher than γcg = 72.8mN/m.
The electrical phenomena at the interfaces can be used for estimation of adsorption.
Especially useful can be measurements of zeta potential of particles in the solutions
before and after adsorption, as it will be shown in other chapters for the quartz-amine
system. It should be also noticed that the zeta potential for both hydrophobic and hy-
drophilic materials without acidic or basic functional groups are similar and their iep is
near pH = 3.5 (Fig. 12.21).
12. Flotation 297
relative surface tension, γ /γ H20 1.00020
1.00010
KCl
1.00000
0.99990
0.99980
0 0.002 0.004 0.006 0.008 0.01
electrolyte concentration, mol/dm3
a
realative surface tension, γ / γ H20
NaCl
1.60
KOH
0.80
NH 4NO 3
0.00
HCl
-0.80 HNO 3
HClO 4
-1.60
-2.40
0 0.2 0.4 0.6 0.8 1
electrolyte concentration, mol/dm 3
b
Fig. 12.20. Influence of electrolyte on surface tension of water:
a) the Jones–Ray effect at low concentrations of salts (Adamson, 1967),
b) at greater salt concentrations (Weissenborn and Pugh, 1996)
20
0
zeta potentia, mV
-20
D2O-ice ( ) (0.0001M)
diamond (•)
-40
air ( ), Nocardia sp.( )

-60
hexadekane
D2O-ice (0.001M)
-80
2 4 6 8 10 12
pH
Fig. 12.21. Comparison of zeta potential in aqueous solutions containing 10–3 M NaCl
for different materials without functional groups having either acidic or alkaline character
(Drzymała et al., 1999; with permission of Academic Press and the authors)
12.4. Delineation of flotation

Flotation, as any other process of separation, depends on properties of material
used as the feed, flotation device and the way flotation is conducted. The main mate-
rial parameter of the flotation is hydrophobicity. The fact that the hydrophobicity is
necessary for flotation (though not always sufficient) can be proved by simple experi-
ments. Gypsum subjected to flotation in water does not float and a drop of water put
on a larger piece of gypsum wets completely the surface pointing to the contact angle
equals zero. Such a simple experiments can fail in the case of other minerals, e.g.
quartz or calcite, which are also hydrophilic and do not float, but a certain contact an-
gle is sometimes detected on these minerals. This can result from adsorption of some
contaminants or perhaps the receding contact angle, not the equilibrium one, is impor-
tant in flotation, which for mentioned above SiO2 and CaCO3 is zero. In the literature
there is disagreement whether or not gold is hydrophobic. Presently, the prevailing
opinion is that gold is hydrophilic, yet sorption of air components, including water va-
pors, can cause its hydrophobicity (Drzymała 1994c). This phenomenon is connected
with high surface energy of gold, which easily adsorbs other substances to decrease it.
Hydrophilicity of gold can be theoretically predicted on the basis of the Fowkes equa-
12. Flotation 299
tions (12.4) and (12.7), using the Hamaker constant, which is exceptionally high for
gold (~5.10-19J).
The necessity of non-zero contact angle for flotation was also proved by Yarar and
Kaoma (1984). They floated naturally hydrophobic substances, including graphite and
sulfur, in the presence of increasing amount of ethyl alcohol which lowers the surface
tension of the solution. They recorded a cease of flotation as soon as the value of con-
tact angle reached zero. Similar results were obtained by other authors (Yarar and Al-
varez 1988, Hornsby and Leja, 1980, Kelebek, 1987).
The condition of non–zero contact angle to facilitate flotation can be also proved
by thermodynamic analysis of the process. Flotation can be treated as a process ana-
logical to a chemical reaction. It takes place between particle and bubble both im-
mersed in water:
particle + bubble = particle-bubble aggregate (12.17)
2
and change of the free enthalpy (thermodynamic potential) (mJ/m ) of the process can
be calculated taking into account the fact that attachment a unit area of particle with a
unit area of bubble creates a unit area of the particle–bubble interface:
ΔGflotation = ΔGproducts – ΔGsubstrates = γsg – (γsc + γcg). (12.18)
Formula 12.18 is know as the Dupre equation (Dupre, 1869). After introducing the
Young relation (Eq. 12.1) to Eq. (12.18) (Pockels, 1914), we get the Young-Dupre
formula:
ΔGflotation = γcg (cos θ – 1). (12.19)
It results form equation (12.19) that the contact angle will more than zero when
ΔGflotation is negative. This is in agreement with the second principle of thermodynam-
ics which says that reaction (in this case flotation) may take place when ΔG of the
process is negative.
The contact angle in flotation is effected not only by parameters included in the
Young equation, i.e. energy of interfaces involved in forming the particle–bubble ag-
gregate but also the size of bubble and particle as well as their densities. Therefore, in
some cases contact angle will not be formed, since a particle will be too heavy and
will detach from the bubble. Flotation can also fail when the hydrophobic particle is
too light (lack of successful collisions). For these reasons each method of flotation fea-
tures its own upper and lower limits of floating particle size (Ralston, 1992).
The magnitude of the interfacial tension depend on such factors as kind of bonding
(covalent, metallic, ionic) between the phases and electrical charge at the interface. On
the basis of these and further pieces of information one can build a pyramid of pa-
rameters, effecting hydrophobicity (Fig. 12.22a). A complete description of flotation
may require more than 100 parameters (Fig. 12.22b).
material
γ sc,γ sg,γ cg,π

γ d, γ p, γ , other
or γ LW, γ + , γ −
Rb hydrodynamics
Rp
device mode of work
Pa
θ, τ
R b, R p, R F ,
h crit , η , c b , γ cg , ρ p
other parameters
R b, R p
R b, R p θ , ρ p, E k
Re hydrodynamics
Pc Pstab
Fig. 12.22. Most important parameters in delineation of flotation:
a – static, b – dynamic delineation (symbols are explained in Table 12.13)
However, flotation is a dynamic, time dependent, process. Very often we want to

know how much time is needed for flotation. Complete description of flotation, in-
cluding the time as a parameter is difficult. There are some general models of flotation
which combine many parameters of flotation. On of them is the Schulze model
(Schulze, 1984; 1992; 1993; Nguyen et al., 1997) which is most advanced. Other
models were proposed by Mao and Yoon (1997), Varbanov Forssberg and Hallina
(1993), and Rubinstein and Samygin (1998). The common feature of all models is
that they divide flotation into three subprocesses, i.e.:
• collision of bubble with mineral particle (collision)
• forming thin film between bubble and particle, and breaking the film after
reaching a critical thickeness (hkryt), followed by formation of contact angle
• forming stable particle–bubble aggregate (stability).
The overall probability P of flotation is:
12. Flotation 301
P = Pc Pa Pstab, (12.20)
where probability of flotation process P is a product of probability of particle colli-
sions with bubble (Pc), particle adhesion to the bubble (Pa), and resistance to detach-
ment of particle from bubble during movement of the bubble-particle aggregate in the
flotation cell (Pstab).
The second common feature of general models of flotation is relating probability
with kinetics of the process. Although equations describing flotation kinetics can as-
sume different forms, it is assumed in most models that kinetic is of the first order:
dNp /dt = –k Np, (12.21)
where:
Np – concentration of floating particles in a unit volume
k – flotation rate constant
t – flotation time.
The third common feature of the general models of flotation is the relation between
the three subprocesses with different flotation parameters. Table 12.13 compares three
general models of flotation. More details regarding the models can be found in the
original publications.
Table 12.13. Models of flotation

General models of flotation (dNp/dt = –k Np)
Varbanov–Forssberg–
Mao–Yoon (1997) Schulze (1993)
–Hallin (1993)
k = PcPaPstab0.25 Sb k = Pc Pa Pstab Ptpc ZNb k = δEUCb
2 2
⎡ 3 4 Re 0.72 ⎤ ⎛ Rp ⎞ ⎛ R ⎞
Pc = ⎢ 2 + 15 ⎥ ⎜⎜ ⎟
⎟ Pc = ⎜⎜ b ⎟ δ = f (Pc) = RpRb
⎣ ⎦ ⎝ Rb ⎠ ⎟
⎝ Rb ⎠
⎛ E ⎞ ⎡ ⎛ τ − τ min ⎞⎤
Pa = exp⎜⎜ − 1 ⎟⎟ Pa = 1 − exp ⎢− ⎜⎜ c ⎟⎥
τi ⎟
⎝ Ek ⎠ ⎣⎢ ⎝ ⎠⎦⎥
τc = [π2 R 3p (ρp + 1.5ρ)/3γ]0,5(1.39 – E = f (Pa)
E = 1 – cos θ
– 0,46 ln Rp)
τi = 3η RF2 Rp/8γcbh2kryt,
h kryt = 23.3[γ (1 – cos θA)]0,16
Pstab= 1 – exp(1 – 1/Bo´),
Pstab = 1 – Pd ω* ≅ 180° – θ /2,
Pstab taken into account in
Bo´ = 4 R 2p (Δρg + ρpa) +
⎡ γ πR 2 (1 − cosθ )2 + E ⎤ other parts of expression for
Pd = exp ⎢− p 1
⎥ +3Rp(sin2 ω*) f (Rb)/C k
⎢ ´ ⎥ C = |6γ sin ω*sin (ω* + θ)|,
⎣ E k ⎦
f (RB) = (2γ /RB) – 2Rbρg
Ptpc ≅ 1 – exp (–τv /τtpc) U = g ( Rb2 ρ +2 R 2p Δρ)/(9η)
τv = 0.6(Rp + Rb)2/ν laminar flow
τv = 13(Rp + Rb)2/3/ε*1/3 turbulent flow Cb = 3Q/(4π Rb3 SV)
dN/dt – flotation rate,

a – centrifugal acceleration acting on particle-bubble aggregate in a vortex of liquid,
Bo´ – Bond’s number,
cb – rigidity of bubble surface (cb = 1 for rigid uneven surface, cb < 1 for movable smooth surface),
Cb – bubble concentration in pulp (number of bubbles in 1cm3 of pulp),
E – efficiency of attachment of particles to bubble surface (number of attached particles divided by
the number of particles colliding with considered bubble),
E1 – energy barrier for adhesion of bubble and particle,
Ek – kinetic energy of collision of bubble with particle,
Ek´ – kinetic energy of detachment of particle and bubble (calculated from the French–Wilson Eq.),
g – acceleration due to gravity,
hkryt – critical film thickness on surface of particle,
k – rate constant of flotation,
Nb – number of bubbles in flotation cell at a considered time,
Np – number of particles subjected to flotation at a considered time,
Pa – probability of adhesion of particles to a bubble,
Pd – probability of detachment of particles from a bubble (Pd = 1 – Pstab),
Pstab – probability of stability of particles-bubble aggregate,
Ptpc – probability of formation of particle-bubble-water contact,
Q – flow of air in flotation machine, m3/s,
Rp – particle radius,
Rb – bubble radius,
Rc – radius of the stream enabling collision of particle with bubble,
RF – size of thin film between particle and bubble during collision,
Re – Reynolds number,
S – cross section area of flotation machine, m2,
Sb – area of bubbles leaving flotation cell per time unit and per cross section area ot the cell,
V – rate of ascending bubble, m/s,
Vb –surface velocity of aeration defined as volume rate of aeration normalized per cross section of
the flotation column,
U – velocity of particles in relation to velocity of bubble in the pulp,
Z – number of particle collisions per unit time,
γ – surface tension of aqueous solution,
δ – quantity characterizing efficiency of collision between particles and bubbles,
ε* – dissipation energy in flotation cell,
Δρ – effective density of particle in water,
η – dynamic viscosity,
ν – kinematical viscosity,
ρ – pulp density,
ρp – particle density,
θ – contact angle,
θA – advancing contact angle,
τv – life time of liquid vortex in flotation cell which destroys the particle–bubble aggregate,
τtpc – time needed to form permanent three-phase solid-gas-liquid contact,
τc – collision time of bubble and particle,
τi – induction time (time of removing liquid film from particle and forming attachment),
τmin – minimal time of contact,
π = 3.14,
ω* = 180 – θ /2.
12. Flotation 303
The value of the periodic flotation rate constant can be determined from the change
of particle yield as a function of time, according to the relation:
ε (t) = 1 – exp (–kt). (12.22)
It should be added, that the particle - bubble adhesion takes place not only as a re-
sult of collision, but also sliding on the bubble. Therefore, the probability of adhesion
will depend on kind of collision and it can have two forms: probability of adhesion
through collision (Pac) and probability of adhesion resulting from sliding (Pasl). Ac-
cording to Schulze (1992) prevalence of any of these forms of collision depends on
many factors, not all of them being known. Generally, it can be stated that collision
involving deformation takes place for large and heavy mineral particles which move
perpendicularly to the bubble surface. The contact through sliding occurs for small and
light particles moving at a low speed. Probability of adhesion, resulting from sliding,
can be described with the use of the equation determining particle movement in the
area of flowing air bubble:
Pasl = sin2 ϕ crit, (12.23)
where ϕ crit is a critical value of the collision angle which, when exceeded, makes ad-
hesion due to sliding impossible.
The trajectory of moving particles causing either collision or sliding is shown in
Fig. 12.23.
rc
particle
droga cząstki
trajectory
zderzenie
collision
sliding
poślizg
ϕ rp
rb stream lines
linia prądu
Fig. 12.23. Flotation can take place as a result of either bubble-particle collision or catching particle by
bubble
The probability of adhesion resulting from sliding is strictly connected with the
character of flow (laminar or turbulent), property of the bubble surface, critical thick-
ness of the film at which its breaking takes place, as well as on particle size, because
smaller particles have higher Pasl value.
To calculate the conditions for good flotation using general theories of flotation,
not only contact angle but also such parameters as size of particles and bubble, induc-
tion time, as well as the value of energy barrier between particle and bubble have to be
taken into account. It makes delineation of flotation quite complex.
The contact time is an important parameter because is decides about forming a con-
tact angle in dynamic conditions of flotation. For a successful flotation the contact
time has to be greater than the induction time. The latter one is measured in the so-
called contact time apparatus. The formula presented in Table 12.13, as well as the
results of examinations indicate that contact time depends on such parameters as hy-
drophobicity (Fig. 12.24) and particle size (Fig. 12.25). Figure 12.26 shows, after
Laskowski and Iskra (1970), that at the same hydrophobicity, shortening the contact
time improves flotation. It can be accomplished for instance by introducing salt to the
solution.
60 2.00
1.50 log (induction time, τ i, ms)

40 θ
1.00
20 τi
0.50
0 0.00
0 LIG0.5SUB 1HVB 1.5
MVB 2
LVB 2.5
ANT
coalification degree
Fig. 12.24. Induction time depends on hydrophobicity
(after Ye and Miller (1988) and Ye et al., 1989).
Various coals from lignite (LIG) to anthracite (ANT).
Symbols LIG, SUB denote type of coal according to American standards
12. Flotation 305
The stability of the particle–bubble aggregates is described by the DLVO theory

discussed in details in chapter 13 which deals with oil agglomeration. Difficulties with
using the DLVO theory for the description of flotation can be caused by lack of data
on zeta potential of bubbles, because such measurements are quite complex. The zeta
potential of air bubbles was described by Schulze, 1984; Li and Somaundaran, 1993,
Laskowski et al., 1989; Usui and Sasaki, 1978.
1000
quartz
induction time, τ i, ms
100
10
0,1
10 100 1000
average particle size, μ m
Fig. 12.25. Induction time depends on particle size. Data after Yordan and Yoon, 1986
60 0
flotation recovery, ε , %
θ
log (induction time τ i, s)
40 -1
20 ε τ i -2
0 -3
0 0.5 1 1.5 2 2.5
KCl concentration, mol/dm3
Fig. 12.26. At certain hydrophobicity flotation increases with decreasing induction time
(after Laskowski and Iskra, 1970)
The diagram presenting the dependence of interaction energy E between bubble

and mineral particle on distance H between these objects is shown in Fig. 12.27. En-
ergy barriers in the flotation system are indicated.
flotation energy barrier

+
( Eb atHb)
Interaction energy , E
Ee + repulsion
Ed – attraction
0
H
Eh secondary minimum
( E2 at H2)
– barrier for a stable

particle-bubble attachment
primary minimum
( E1 at H1)
stable particle-bubble
attachment
Fig. 12.27. Example of interaction energy E between bubble and particle in water resulting from the
DLVO theory: Ee – electrostatic, Ed – dispersion, Eh – hydrophobic interactions.
Solid line is the sum of these interactions
To describe the rate of flotation (dN/dt) not only the first but also higher orders ki-
netic equations can be used. According to Arbiter and Harris (1962), Garcia–Zunga
(1935, Soc. Nac. Miner. Santiago, Chile, 47, p. 83) proposed a general equation:
dN
= − kN n , (12.24)
dt
where:
N – number of hydrophobic particles in the cell which are able to float at flotation
time t,
n – constant indicating the so-call order of kinetic equation,
k – flotation rate constant.
There are other approaches to kinetics of flotation which are summarized in Ta-
ble 12.14.
12. Flotation 307
Table 12.14. Models applied for kinetics of flotation

(after Dowling et al., 1985 and Lynch et al. 1981)
Model Relation
Classic first order ε = ε∞ [1 – exp (–k1t)]
Modified first order ε = ε∞ {1 – 1/(k2t)[1 – exp (–k2t)]}
For reactor with ideal mixing ε = ε∞ [1 – 1/(1 + t/k3)]*
Modified for gas–solid adsorption ε = ε∞ k4t/(1 + k4t)*
Kinetics of second order ε = (ε∞)2 k5t/(1 + ε∞ k5t)
Modified second order ε = ε∞ {1 – [ln (1 + k6t)]/(k6t)}
Two rate constants ε = ε∞ [1– {ϕ exp (–k7t) + (1 – ϕ ) exp(–k8t)}
∞
ε = ε∞ [1 – exp(–kt) f (k, 0) dk]
Distributed rate constants ∫
0
* Equivalent models because k3 = 1/k4.

ε – flotation recovery after time t,
ε∞ – maximum recovery,
ϕ – fraction of particles having lower flotation rate constant, k7,
k – flotation rate constant.
12.5. Flotation reagents
12.5.1. Collectors
Flotation requires hydrophobicity of the mineral particle but only a few mineral
substances are naturally hydrophobic. Therefore, there is a need to use various re-
agents, called collectors, to render hydrophilic and slightly hydrophobic particles hy-
drophobic. The hydrophobizing power of a collectors results from its chemical and
physical interactions with the surface and is usually described from physical chemistry
point of view. The role of chemistry in the description, analysis and assessment of flo-
tation process is shown in Fig. 12.28.
Flotation process can be used to separate particles from water but it is most often
used to separate particles of different materials. In such a case the applied collector has
to be selective. Selectivity of a collector relies on favored adsorption of one type of
particles of the flotation suspension. Flotation collectors effect not only particle hy-
drophobicity but also other parameters of flotation, including the time of particle-
bubble contact required to form a stable particle-bubble aggregate, as well as the sta-
bility of froth.
FLOTATION
classification evaluation
upgrading
splitting
γ
sorting, etc.
analysis
ε−−−−− λ
delineation c chemistry of flotation

phy s cs
x1 x2
an mi
a
sics
x4
me dyn
r x3
i
he
ot
o
xn
ch
m
Cp
er
th
Cs P2
P1
Fig. 12.28. Chemistry plays an important role in flotation
There are many various reagents used as flotation collectors (Gaudin, 1963; Las-
kowski, 1969), and there are many possible ways of their classification. One of them
is presented in Fig. 12.29. Collectors are usually divided into polar and a non-polar.
The non-polar collectors can be further divided into hydrocarbons and their derivates,
sulfur organic compounds, and alcohols with their derivates. Generally, the non-polar
collectors are insoluble or only slightly soluble in water. The adsorption of alcohol–
type collectors takes place mainly due to hydrogen bonding, while in the case of re-
maining non-polar collectors the adsorption is due to the van der Waals interactions.
The family of polar collectors can be divided into cationic, anionic, amphoteric and
chelating collectors. The cationic collectors have positively charged functional groups,
anionic collectors are negatively charged, while amphoteric ionic collectors contain
both cationic and anionic functional groups. A characteristic feature of simple ionic
collectors is forming single bonds between collector and mineral surface. Far more
complex are reaction occurs with the chelating collectors. They form strong chemical
rings involving the chelating compound and bulk and surface ions. The adsorption of
non-polar, ionic, and chelating collectors is shown in Fig. 12.30.
12. Flotation 309
collectors
ionic non-polar
sulfur hydrocarbons
chelating simple compounds and derivatives
alcohols
S-S xanthates
cationic amines
O-O fatty acids
N-N diamines anionic merkaptans
S-N carbamines
amphoteric amino-
O-S monothiocarbonates alkylacids
O-N oximes
Fig. 12.29. One of several ways of collectors classification
Formation of a surface bond between the ionic collector and a metal ion on the par-
ticle surface (–MeOH) can be written as a chemical reaction:
R–SH + HO–Me– = R–S–Me– + H2O (12.25)
while the formation of a chelating bond, using hydroxamic acid as an example, is
given by the reaction:
R C O R C O
+ HO Me = Me + H2O (12.26)
HN OH HN O
where sign refers to additional coordinate bond with the use of a free electron pair
coming from the oxygen atom while Me- denotes metal ion forming the particle sur-
face.
Chelating bonds can be formed by collectors capable of interacting with metal ion
on the surface. They form four, five or six-segment ring. One of the bonds is ionic–
covalent–metallic in character, while the second one is a coordination bond. The col-
lectors with functional groups containing two atoms out of the group of N, S, P, and O
are capable of forming the chelating bond. Therefore, the chelating collectors can be
divided into subgroups such as S-S, O-O, S-O, S-N, and N-P collectors.
oil
-O-H -O
Me S S
Me S
Fig. 12.30. Possible modes of adsorption of collectors at particle-water interface:

a – adsorption of oil on hydrophobic particle with van der Waals forces,
b – adsorption of apolar molecule of collector by means of hydrogen bonding,
c – adsorption of polar collector by means of simple chemical bond
or electrostatic attraction, d – adsorption with formation of chelating bond.
Hydrophobic part of the collector is shown in white while hydrophilic as black
Not to scale.
Some researchers believe that certain ionic collectors, although capable of forming
chelating bonds, do not form them on mineral surfaces (Sprokin, 1976, Parfitt and
Russell, 1977; Cornell and Schindler 1980), especially when four-segment rings are
involved.
Tables 12.15–12.20 list, after Nagaraj (1988), selected non-polar, cationic, anionic,
amphoteric and chelating collectors. In the tables symbols R, R’ and R’’ stand for hy-
drocarbon radicals.
12. Flotation 311
Table 12.15. Nonionic collectors

Collector Example
fuel oil, naphtha, heptane, benzene, halogen derivatives of
Hydrocarbons and derivatives
hydrocarbons
dixanthogen (R–O–C(=S)–S–)2
Sulfur compounds formic xanthate R–O–C(=S)–S–C(=O)–O–R´, alkyl
disulfides R–S–S–R
alkylphenyl(polyethoxy) alcohols (Triton, Tergitol, Brij),
Alcohol and derivatives
alkylphenols, higher alcohols
Table 12.16. Ionic collectors

Collector Example
Cationic
n-dodecylamine C12H25–NH2 or C12H25–NH2·HCl
Primary amines* R–NH2
(C12H25–N+H3Cl–)
Secondary amines* R–(R´)N–H di-n-amylamine, (C5H11)2–NH
Tertiary amines* R–(R´)N–R´´ tri-n-amylamine, (C5H11)3–N
R' R''
-
Quarternary ammonium salts N+ Cl
R R''
Diamines and triamines diamine R–NH–(CH2)x–NH2

Cl-
N+
Pyridinium salts R
CH2-CH2
+ -
Morpholinium salts R NH O Cl
CH2-CH2
R' R''
Sulfonium salts S+
-
R Cl
Anionic
Due to unpleasant smell mercaptans are not used
Mercaptans, R–SH
by industry
amphoteric
N-dodecyl-2-aminopropionic acid C12H25–N+H2–(CH2)2–COO–
C12H25–NH–CH2–COONa,
other C12H25–N(CH2–COONa)2, R–(CH3)2N+–CH2–
COO– (alkyl betaines)
* In practice hydrochloric forms of amine are used (for example R–NH2·HCl).
Table 12.17. Selected chelating collectors. Type O–O. Based on Nagaraj, 1988
Collector Formula Example
Carbonic acid derivatives
R–COOH oleic, linoleic, linolenic, stearic acids
(fatty acids)
Sulfuric acid derivatives
R–O–SO3H dodecyl sulfate
(sulfates)
Sulfuric acid derivatives
R–SO3H dodecyl sulfonates
(sulfonates)
Phosphoric acid derivatives
(RO)2–P(O)–OH dialkyl phosphoric acid
(phosphates)
dialkyl phosphonic acid
Phosphonic acid derivatives (RO)–(R)P(O)–OH
Flotol-7,9 (1-hydroxyalkylidene-1,1-
(phosphonates, diphosphonates) R–(PO3H2)2
diphosphonic acid)
Phosphinic acid derivatives
(R)2–P(O)–OH dialkyl phosphinic acid
(phosphinate)
Nitrosophenylhydroxylamine
(Ar–N(O–)–N=O)NH4 Cupferron
(ammonium salt)
Salicylaldehyde OH–Ar–CHO Salicylaldehyde
α-nitroso-β-naphthol, β-nitroso-α-
Nitrosonaphthols ON–nA–OH
naphthol
Nitrosophenols R–(OH)Ar(OH)–NO nitroso alkyl resorcinol
Organic dyes alizarin and derivatives
Benzohydroxamic acid, potassium
Hydroxamic acids R–CO–NH–OH
octylhydroxamate, IM-50 (C7-C9)
Table 12.18. Selected chelating collectors. Type S–S. Based on Nagaraj, 1988
Collector Formula Example
S
Potassium ethyl xanthate
Dithiocarbonates (xanthate) - C - O-
(R–OCSSK)
S
S
Trithiocarbonates (tioxanthate)
- C - S-
S
S
Dithiophosphates - P(OR)2 Aerofloat ((RO)2P(=S)–SK)
S
S
Dithiophosphinates - PR2 Aerofins
S
S
Dithiocarbamates - C - NR2 Sodium diethyldithiocarbamate
S
12. Flotation 313
Table 12.19. Selected chelating collectors. Type O–N. Based on Nagaraj, 1988
collector formula Example
α-benzoin oxime
CH C
CH C
Oximes OH
N OH
OH
N OH
LIX65N
C 9 H 19
Hydroxyoximes (LIX series)

C
OH N
OH
8-hydroxyquinoline and derivatives N

8-hydroxyquinoline (oxine)
OH
Table 12.20. Selected chelating collectors. Type S–N. Based on Nagaraj, 1988
Collector Formula
C S
C SH
Mercaptobenzothiazols C N
(flotagen)
R C N
Mercaptothiodiazoles C SH
N S
C NH
Thiotertrahydroglyoxaline C SH
C N
S
Mono- and dithiocarbamates
C4 H9 O C NH C2 H5
N
S C H
Phenylthiourea H
N
Ionic and some nonionic collectors have very characteristic structure. They consist
of a hydrophilic part which is adsorbed on the particle surface and a hydrophobic part
which remains in aqueous phase (Fig. 12.30 and 12.31).
CH3CH2CH2CH2 CH2CH2CH2CH2 COO–
tail head
(hydrophobic) (hydrophilic)
Fig. 12.31. A sketch of structure of ionic collectors using carboxylic acid as an example
Dissolved in water collectors having large hydrophobic part can mutually interact
and form, depending on concentration, pre-micelles, micelles, big micelles and other
forms (Fig. 12.32). The micelles occur at a certain surfactant concentration called the
critical micelle concentration or shortly CMC.
a b c
Fig. 12.32. Collector ions can be present in aqueous solution as free ions (a), premicellar species (b)
spherical micelles (c). The structures appear with increasing concentration of collector in aqueous solu-
tion.
Symbol o denotes ion appositively charged to the collector ion
The CMC values for selected anionic, cationic, and amphoteric collectors are pre-
sented in Tables 12.21-12.23.
12. Flotation 315
Table 12.21. Critical micelle concentration (CMC) of selected anionic collectors at room temperature
(data after Mishra, 1988)
Anionic surfactant CMC (kmol/m3)
Sodium caprylate, C8H17COONa 3.5·10–1
Sodium capriate, C10H21COONa 9.4·10–2
Sodium laurate, C12H25COONa 2.6·10–2
Sodium myristate, C14H29COONa 6.9·10–3
Sodium palmitate, C16H33COONa 2.1·10–3
Sodium stearate, C18H37COONa 1.8·10–3
Sodium elaidate, trans-C17H33COONa 1.4·10–3
Sodium oleate, cis-C17H33COONa 2.1·10–3
Sodium dodecyl sulfate, C12H25SO4Na 8.2·10–3
Sodium dodecylo sulfonate, C12H25SO3Na 9.8·10–3
Sodium dodecylo benzene sulfonate,
1.2·10–3
C12H25–C6H4–SO3Na
Table 12.22. Critical micelle concentration (CMC) of selected cationic collectors at room temperature
(data after Smith, 1988), m = number of CH2 groups in w hydrocarbon chain R)
Cationic surfactant Temperature °C (K) CMC, kmol/m3
n-alkylammonium chlorides (R–NH3Cl) 30 (303) log CMC = 1.31÷0.270m
N,N-dimethyl-n-dialkylammonium chlorides, log CMC = 2.78÷0.544m
30 (303)
R2(CH3)–N–CH3Cl (m = 8, 10, 12)
N,N,N-trimetylalkylammonium bromides, log CMC = 1.98÷0.311m
25(60) (298)(333)
R(CH3)2–N–CH3Br (m = 8 ÷16)
N-(n-alkyl)-propyl-1,3-diammonium chloride, log CMC = 1.72÷0.295m
25 (298)
R(H)2–(CH2)3NH3Cl (m = 10÷16)
log CMC = 1.70÷0m
n-alkylopyridinium bromides 30 (303)
(m = 10÷16)
Table 12.23. Critical micelle concentration (CMC) of selected amphoteric collectors at room temperature
(data after Smith, 1988)
Amphoteric surfactant Temperature °C (K) CMC, kmol/m3
Diheptanoyl phosphatidylcholine 25 (298) 1.4·10–3
Dioctanoyl phosphatidylcholine 25 (298) 2.7·10–4
Didecanoyl phosphatidylcholine 25 (298) 5.0·10–6
Dipalmitoylophosphatidylocholine 25 (298) 4.7·0–10
N-dodecyl-2-aminopropionic acid 40 (313) 1.65·10–3
N, N-dimethyl-N-(3 sulfopropyl dodecylamine 30 (303) 3.6·10–3
+ –
C12H25(CH3)2–N –CH2–COO 30 (303) 1.7·10–3
Micelles of ionic collectors are not only a collection of ions. Examinations and cal-
culations indicate that on average 45% of ionic micelles ions are bonded with the
counter ions (Drzymała, 1990; Dutkiewicz, 1998). Therefore, adsorption layers of
ionic collectors contain both surfactant ions and surfactant molecules. Increasing the
concentration of inorganic salts in the solution results in higher number of non-

dissociated surfactant molecules in the micelles and adsorbed layers. The micelles eas-
ily adsorb at the water-air interface (Fig. 12.33). Micelles are formed by such collec-
tors as oleates, sulfates, sulfonates and other anionic, cationic, amphoteric, and non-
ionic surfactants. Certain thiocompounds including xanthates do not form micelles,
and therefore, do not adsorb at the water–air interface, and thus, are not called surfac-
tant. The micellization of xanthates requires a higher number of CH2 groups in the hy-
drocarbon chain, usually 10 or more groups, while typical xanthates possess from 1 to
about 9 groups.
inorganic salts
surface tension, mN/m
72
thiocompounds
R1
R2
R1 , R 2 - alcohols of decreasing solubility
concentration, kmol/m 3
b
72
CMC
surface tension, mN/m
R1
R 1 + contaminants
45 R2
R1 , R 2 - surfactants of increasing length
R 2 + salt of hydrocabon chain
CMC
concentration, kmol/m 3
Fig. 12.33. Change of surface tension for aqueous solutions in the presence of collectors:
micelle-forming (a) and not forming micelles (b). Surface tension of water in the presence of inorganic
salts and alcohols are also presented (after Leja, 1982)
12. Flotation 317
The decrease in the surface tension caused by surface active collectors does not re-
sult from a simple mixing of two substances that is water (surface tension 72,8 mN/m)
and organic collector (surface tension from 25 to 35 mN/m), but from a considerable
adsorption of the hydrophobic surfactant ions or molecules at the water–air interface.
Significant adsorption of collector ions at the interface makes the surface tension, in
the vicinity of CMC, nearly equal to surface tension of pure (anhydrous) collector. It
usually corresponds to a monolayer of the collector at the water-air interface.
The behavior of collectors at the air-water interface is different from that of inor-
ganic salts, since the surface tension of aqueous solutions in the presence of salts, after
a slight decrease (Fig. 12.20a), increases as salt concentration increases (Fig. 12.33a).
80
60
40
galena
20
0
0 2 4 6 8 10
concentration of potassium ethyl xanthate, g/m 3
a
100
80 pH = 7
pH = 9
60
pH = 10
40
pH = 10.5
20
hematite - NaOl
0 -7 -6 -5 -4 -3
10 10 10 10 10
concentration of sodium oleate, kmol/m 3
b
Fig. 12.34. Change of hydrophobicity (contact angle) of mineral particles as a function collector concen-
tration: a – collector not forming micelles (Wark and Cox, 1934, Gaudin, 1963), b – micelles-forming
collector (after data of Yap et al., 1981)
The association of collectors influences the hydrophobicity of mineral surface. For

non-associating collectors the contact angle increases as the collector concentration
increases and reaches plateau a certain constant value (Fig. 12.34a) which is character-
istic for a particular hydrocarbon chain (Table 12.24). Table 12.24 shows maximum
values of contact angle for different non-associating collectors.
Table 12.24. Maximum contact angle for different collectors

with ethyl and butyl chain.
After Gaudin, 1963
Collector Chain length*
ethyl butyl
Dithiocarbaminate 60 77
Mercaptane 60 74
Xanthate 60 74
Dithiophosphate 59 76
Trithiocarbonate 61 74
Monothiocarbonate 61 73
* For methyl chain (C1) θ ≅ 50°, propyl (C3) 68°, C5 78°, C6 81°, for greater about 90°, and for C16 98°
(Aplan and Chander, 1988).
The long-chained collectors associate and their contact angle, after reaching maxi-
mum, begins to decrease (Fig. 12.34) due to formation of a second layer on the exist-
ing first layer (Fig. 12.35). In the second layer the collector ion is adsorbed with a hy-
drophobic part on the hydrophobic part of the collector in the first layer.
a b c
Fig. 12.35. Adsorption of ionic collector on the surface of particle with the formation of hemimicelle (a),
monolayer (b) and a second layer leading to hydrophilicity (c)
It should be noticed, however, that no correlation between CMC and the point of
diminishing flotation is observed (Fig. 12.36).
12. Flotation 319
100
recovery, %
80 CMC SOS CMC SDS
60
40
20
alumina
0 -7 -6 -5 -4 -3 -2 -1
10 10 10 10 10 10 10
collector concentration, kmol/m 3
Fig. 12.36. An example of lack of correlation between CMC and flotation (after Freund and Dobias,
1995). SOS – sodium octyl sulfate, SDS – sodium dodecyl sulfate
On the other hand, there is a quite satisfactory correlation between flotation, ad-
sorption, hydrophobicity, and other properties of flotation systems (Fig. 12.37).
50 100 100
adsorption density, mol/m2 x1011
quartz-dodecylamine
40 0.80 80 60
zeta potential, mV
flotation recovery
cos θ
30 60 20
20 0.90 40 -20
10 20 -60
0 -6 -5 -4 -3
0
-2
-100
10 10 10 10 10
collector concentration, kmol/m 3
Fig. 12.37. Flotation of particles increases with increasing concentration of collector in the system
and is proportional to collector adsorption and hydrophobicity caused by the adsorption. Collector adsorp-
tion is manifested by the increase of zeta potential of particles (after Fuerstenau et al., 1964 and Fuer-
stenau and Urbina, 1988), pH = 6–7
Adsorption of collectors and other compounds occurs, to a different extent, at any

interface including those important in flotation, i.e. solid–water, solid–gas and water–
gas interface. The property of the interfacial region can be characterized by the Gibbs
equation. It is based on thermodynamics and assumption that for two-phase systems
there are two pure bulk phases α and β, and a transitory interfacial region σ, situated
between the bulk phases. The total internal energy U of each phase of the system will
be a sum of different forms of energy:
Uα = TSα – pVα + Σμnα, (12.27)
Uβ = TSβ – pVβ + Σμnβ, (12.28)
Uσ = TSσ – pVσ + γA + Σμnσ, (12.29)

that is thermal TS, surface γA, and chemical Σμn energies as well as the volumetric
work pV
where:
α, β, σ – phases
T – temperature in kelvins
S – entropy
p – pressure
V – volume
γ – interfacial energy
A – interfacial area
μ –chemical potential
n – number of molecules or ions (or moles) of a given substance.
Symbol Σ indicates that the chemical energy is a sum of energies of all chemicals
species.
The interfacial region σ contains chemical individuals present in both α and β
phases To calculate the excess internal energy Us of the interfacial area we subtract
energies of the α and β phases from the total energy of interfacial region.
Us = TSs + γA + Σμns. (12.30)
Total derivative of Eq. (12.30) assumes the form:
dUs = TdSs + SsdT + γdA + Adγ + Σμdns + Σnsdμ. (12.31)
It results from thermodynamics that when a small and reversible change in surface
phase σ takes place then:
dUs = TdSs + γdA + Σμdns, (12.32)
which means that the remaining segments of Eq. (12.30) have to fulfill the relation:
12. Flotation 321
SsdT + Adγ + Σ(nidμi)s = 0, (12.33)

and for the processes taking place at a constant temperature T and knowing that
Γ=(n1)s/A, we obtain the relation:
(dγ )T = (–ΣΓidμi)T. (12.34)
Relation (12.34) is general and complete and describes the state of an interface and
it is called the Gibbs equation. Taking into account that the chemical potential μ is de-
termined by the dependence:
dμi = RT d ln ai, (12.35)
where α is the activity of a given chemical species, the Gibbs equation can be trans-
formed into:
(dγ )T = –RT ΣΓid (ln ai)T = (–RT ΣΓidai /ai)T. (12.36)
The Gibbs equation combines three parameters: adsorption (strictly speaking sur-
face excess Γ), activity (or concentration c, since a = fc where f is the activity coeffi-
cient) of a chemical species in the solution (a) and the interfacial energy γ. Further
transformation of the Gibbs equation is possible taking into account additional data
regarding the interfacial region, e.g. dissociative properties of molecules, location of
ions or molecules in interfacial region, as well as their mutual interactions. For in-
stance, a change of the interfacial energy resulting from the adsorption of potential
determining ions H+ and OH- on oxides at constant concentration of other ions origi-
nating from dissolved salts (in order to maintain constant ionic strength of the solu-
tion) and at a constant temperature, is described by the equation (Bruyn and Agar,
1962):
dγ = 2.303 RT (ΓH+ – ΓOH–) d(pH) = 2.303RTσ0 d(pH). (12.37)
where σ0 is the surface charge.
After relating pH and electric potential, Eq. 12.37 can be transformed into the
Lippmann electrocapillary curve. The Lippmann equation can be directly derived from
the Gibbs equation (Davis and Rideal, 1963).
The interfacial energy change due to variation of the salt concentration (Csoli) in the
solution at a constant temperature and pH can be described by the following equation
(Ratajczak, 2001):
(dγ /dlog csalt)pH = –2.303RTF–1(σ+/z+ – σ–/z–), (12.38)
where:
σ+, σ– – ionic components of surface charge (Lyklema, 1972) σo =-(σ+ +σ–),
z – valence of ions of the salt.
The pH change due to variation of the salt concentration at a constant surface

charge and constant temperature is given by the Esin–Markov equation (Lyklema,
1972):
(d pH/d log a)σ 0 = β, (12.39)
where β is a constant.
The Szyszkowski formula relating a change of the surface tension π (being the dif-
ference between the surface tension of pure solvent γ0 and that of surfactant solu-
tion γ) due to increasing mole fraction x (concentration) of the surfactant in the solu-
tion fulfills the Gibbs equation (Koopal, 1992) and has the form:
π = γ0 – γ = RT Γm ln (1 + xi K iP1 ). (12.40)
Equation 12.40 is one of several forms of the equation proposed by Szyszkowski as
an empirical relation, which was later found to be a special case of the Gibbs formula
in which:
γ0 – surface tension of solvent
γ – surface tension of solution at a given concentration of surfactant expressed in
mole fraction form (x)
Γm – constant for a complete monolayer adsorption
xi – mole fraction
K iP1 – equilibrium constant for the surface reaction:
(1)s + i = 1 + (i)s, (12.41)
where:
1 – molecule of solvent
i – molecule of adsorbed component
s – surface.
For reaction 12.41 the equilibrium constant, in which concentration is expressed as
a mole fraction x, assumes the following form:
⎛ x x s ⎞⎛ f f s ⎞
K iP1 = ⎜⎜ 1 is ⎟⎟⎜⎜ 1 i s ⎟⎟ , (12.42)
⎝ xi x1 ⎠⎝ fi f1 ⎠
where f stands for activity coefficient.
Thermodynamically the equilibrium constant is expressed by the relation:
⎡ A0 (γ 0 − γ i0 ) ⎤
K iP1 = exp ⎢ i 1 ⎥, (12.43)
⎣ RT ⎦
where:
Ai0 – area occupied by one mole at monolayer adsorption
12. Flotation 323
γ 0 – surface tension of pure substance.

It can be said that the formula for the reaction equilibrium constant for a monolayer
adsorption is a bridge between the Szyszkowski and Gibbs equations.
The equations derived from the Gibbs formula as a rule combine two out of three
parameters (c, γ, Γ) for finding the third one. Therefore, the Gibbs equations can be
divided, as shown in Fig. 12.38, into three families: 1) the relations relating adsorption
with concentration Γ = f(c), i.e. adsorption isotherm (Γ-c curves), 2) surface tension
change with concentration (γ-c curves), and 3) changes of interfacial tension as a func-
tion of adsorption (γ-Γ curves), or more often π = γ-γ0 = f(Γ)).
– for two components Esin-Markov coefficient
– for more componets C1 = f (C2) (T, σ = const)
C
concentration
– for organic electrolytes – Langmuir

– for inorganic electrolytes (salts) – Stern-Graham
– for nonelectrolytes – Temkin
ad
– Szyszkowski equation – Frumkin

so
es
rp
– other – other
ur v
tio
ni
cc
so
γ-
th e
general Gibbs eq.

rm
dγ = -Σ Γ i dμ i (T = const)
γ, interfacial energy (or Π)

two-dimensional
equation of state Γ , surface excess
– Schofield π [(1/Γ ) – Ao )] = bkT

– Volmer (b = 1)
Fig. 12.38. Central role played by the Gibbs equation

in description of relationships between concentration of adsorbing reagent c,
its adsorption Γ and interfacial tension γ
and the place of other equations based on the Gibbs formula
The γ-c curves are applied in mineral processing for monitoring the influence of
different reagents (collector, frother, salt, etc.) on thermodynamical properties of inter-
faces for possible further calculation of hydrophobicity (contact angle) and its
changes. The π = γ - γ0 = f (Γ) curves are also useful in mineral processing and they
are called the equations of state. The equivalent of either Langmuir’s or Szysz-
kowski’s equations for a monomolecular adsorption of substances adsorbing itself
without any additional interactions, based on the Gibbs equation for adsorption at a
constant temperature, has the following form:
π = γ0 – γ = –RT Γm ln (1 – xis ) lub = –kT Γm ln (1 – Γi /Γm), (12.44)

where Γmax is the density of adsorption when monolayer is formed.
Equation 12.44 describes a rigid adsorption of chemical species which do not
change the site of adsorption. For mobile adsorbants, there are other equations and
they can be empirical (Shofield and Rideal or Vomer) (Jaycock and Parfitt, 1981), as
well as combined empirical and Gibbs equations, which usually contain added infor-
mation on adsorption conditions and interactions (Davis and Rideal, 1963).
The simplest adsorption isotherm is that of Langmuir. It describes a uniform ad-
sorption leading to a monolayer in which there no interactions between the adsorbing
species. The Langmuir adsorption isotherm can also be derived from chemical reac-
tions constants, which at the same time have to fulfill the Gibbs equation. The Lang-
muir adsorption isotherm has a form:
Γ max xi K iP1
Γi = , (12.45)
1 + xi K iP1
where K is the adsorption reaction constant.
There are many adsorption isotherms relating concentration of adsorbed species as
a function of their concentration in solution at a constant temperature. They represent
complex isotherms which take into account various interactions between adsorbed
species as well as simplified isotherms. A simplified Stern–Graham isotherm can be
considered as a Langmuir’s simplified isotherm. It is valid for low concentration of the
substance undergoing adsorption in the solution (l > xKp) and it has the form (Hunter,
1987):
Γi = 2rc exp(–Δ Gads
0
/RT), (12.46)
where:
Γi – adsorption, mol/cm2
Δ Gads
0
– standard free enthalpy of adsorption (Δ Gads
0
= –RT ln Kij)
r – radius of adsorbed counter ion
c – number of pairs of ions per 1 cm3 of solution.
When transforming the Langmuir into Stern-Graham equation, Γmax becomes de-
pendent on the size r of ions adsorbed.
There are many other adsorption isotherms, including the Freundlich, (Adamson
1967), Temkin (Davis and Rideal, 1963), Frumkin (Leja, 1982), and other isotherms
(Jaycock and Parfitt, 1981). Details on adsorption isotherms can be found in other
sources.
Interfacial collector adsorption takes place as a result of the action of different
forces. To describe the forces, researchers employ certain chemical rules based on
chemical bonds and interactions. In vacuum, ionic, covalent, coordinative, metallic
12. Flotation 325
bonds, as well as van der Waals interactions can take place. In aqueous solutions, ad-
ditional interactions such as hydrogen bonds, electric interactions between electrically
charged surface and entirely or partly hydrated ions in the solution, hydrophobic inter-
actions originating from different structure of water at hydrophobic surface and the
structure of ions or molecules undergoing interactions with the surface and showing a
tendency to associate can be found. Therefore, a general complete formula for the free
enthalpy energy of adsorption of ions, molecules, associated forms, and colloidal ob-
jects at interfaces, can be written as follows:
ΔGad
0
= Δ G 0chemical adsorption + ΔG physical
0
adsorption + Δ G hydrogen bonding
0
(12.47a)
or more precisely:
ΔGad
0
= {( ΔGchem
0
+ ΔGchel
0
)} +{ ΔGWaals
0
+ ΔGelec
0
+ ΔGCH
0
+
2
(12.47b)
+ ΔGhydrofob
0
}+{ ΔGhydrogen
0
},
where:
ΔG0chem + ΔG0chel – free enthalpy describing the effect of simple chemical and coordi-
native bonds ΔG0chem leading to chelating structures
{ΔG0Waals + ΔG0elec + ΔG0CH2 + ΔG0 hydrophob }- free enthalpy of physical (ΔG0Waals),
electrical (ΔG0elec), interaction of associating CH2 groups from hydrocarbon
chains (ΔG0CH2), and hydrophobic interactions (ΔG0hydrophob)
ΔG chem – term resulting from the effect of hydrogen bonds.
0
It should be added that the van der Waals interactions consist of three other interac-
tions, namely, dispersive (London), induction (Debye inducted dipoles) and orienta-
tion (Keesom permanent dipoles) (Laskowski, 1969).
This diversity of forces makes the adsorption and consequently flotation a complex
process (Fig. 12.39). In the case of ionic collectors, there are three dominating mecha-
nisms: ionic (electrostatic, physical sorption), chemical, and van der Waals (physical
adsorption). The adsorption via chemical bonding usually leads to the best flotation in
the vicinity of the pzc (Fig. 12.39a). The electrostatic mechanism is typical for sub-
stances which are not capable of forming chemical bonds with collectors, e.g. in flota-
tion with anionic collectors which render minerals hydrophobic only up to the isoelec-
tric point (iep), while with cationic collectors above the iep (Fig. 12.39b). Purely elec-
trostatic mechanism takes place at low collector concentrations. At higher concentra-
tions, apart from electrostatic adsorption, sorption is due to the van der Waals or
chemical interactions. Then, the effect of surface electric charge on flotation becomes
complicated. The nonionic collectors undergo adsorption on mineral surfaces and then
induce flotation due to, first of all, van der Waals forces, especially the London disper-
sive interaction (Fig. 12.39c). The most efficient adsorption takes place on hydropho-
bic surfaces leading to an additional increase of hydrophobicity which considerably
improves flotation. Other nonionic collectors render minerals floatable due to the hy-
drogen bonds (Fig. 12.39d).
100 100
goethite
80 80
kyanit e
recovery, %
collector
recovery, %
60 (collector: oleic acid) collector
60 RSO 4Na RNH3Cl
iep 6.7
40
40
iep 6.9
20
20
0
0 2 4 6 8 10 12 14 0
pH 2 4 6 8 10 12
pH
100 100
80 coal 80 quartz
(collector: tridecane) (collector: octylphenylpolyethoksyethanol)
recovery, %
recovery, %
60 60
40 40
iep 2÷ 3
20
iep 2÷ 3
20
0
0 2 4 6 8 10 12
2 4 6 8 10 12
pH
pH
Fig. 12.39. Examples of domination of different mechanisms of adsorption leading to flotation as a func-
tion of pH and iep of material: a – chemical adsorption of oleate (10–4 M) on an aluminosilicate (kyanite)
(maximum flotation is at iep) (Choi i Oh, 1965), b – physical adsorption (electrostatic) of dodecyl sulfate
(10–3 M) on goethyte (flotation is up to iep, while for dedecyl amine above iep) (after Iwasaki et al.,
1960), c – adsorption due to van der Waals interactions of apolar oil with coal (flotation is in vicinity of
iep and decreases with increasing surface electrical charge) (Stachurski and Michałek, 1986), d – adsorp-
tion of nonionic polymeric collector (1.6·10–4 M, 10 etoxy units) with hydrogen bonding on quartz (flota-
tion is proportional to the number of undissociated surface –OH groups which maximum number is at
iep) (Doren et al., 1975)
Knowing the mechanism of adsorption as well as the extent of adsorption depend-

ing on the collector concentration in water, one can calculate the free enthalpy ΔGads
of adsorption on the basis of an appropriate adsorption isotherm and then predict flota-
tion results and selectivity. However, the calculations are far from being easy. Table
12.25 presents ΔG0ads values for different minerals and collectors based on the Gra-
ham-Stern isotherm, which is a simplified form of the Langmuir isotherm (Hunter,
1989).
12. Flotation 327
Table 12.25. Free enthalpy of adsorption ΔGads calculated from the Graham–Stern isotherm
and collector concentration (CT), at which adsorption is equal to 1 μmol/m2, r = 2.5·10–10 m
(after Dobias, 1995)
Collector concentration
ΔGa
Mineral Collector CT
3 kJ/mol
kmol/m
Fluorite Sodium tetradecyl sulfate 4·10–5 –26.3
Fluorite Sodium laurate 6·10–5 –25.4
Calcite Sodium dodecyl sulfonate 2·10–4 –22.4
–4
Fluorite Sodium dodecyl sulfonate 3·10 –21.4
Scheelite Sodium dodecyl sulfonate 3.5·10–4 –21.0
Calcite Sodium dodecyl sulfate 7.5·10–4 –19.2
Fluorite Sodium dodecyl sulfate 1·10–3 –18.5
Fluorapatite Sodium dodecyl sulfonate 2·10–4 –16.8
It results from Table 12.25 that the strongest adsorption involves sodium tetradecyl
sulfate on fluorite, while the weakest regards sodium dodecyl sulfonate on fluorapa-
tite.
In flotation, mixtures of two or more collectors are often applied. Then one of them
is usually called promoter. Some flotation promoters, when used separately, show a
very weak flotation effect. They feature synergism only in the presence of an appro-
priate collector. Numerous investigations on the flotation promoters involved coal
(Sablik, 1998).
12.5.2. Frothers
The main role of frother is to disperse gas, form stable froth and accelerate flota-
tion. These tasks can be achieved because frothers are preferentially adsorbed at the
water–gas interface. The sorption is usually accompanied by a decrease in surface ten-
sion (Fig. 12.40a). Introduction of a frother into solution results in reduction of air
bubbles size (Fig. 12.40b). The bubbles size reduction is connected with a decrease in
the surface tension of the solution. It becomes evident when the bubbles are formed in
a capillary, since the following relation takes place:
6aγ wg
dp = 3 , (12.48)
g ( ρc − ρ g )
where:
γwg – surface tension of frother aqueous solution
a – capillary diameter
ρc – aqueous solution density
ρg – gas density
g – gravity.
2
80
Aston et al., 1983
1.6 Tucker et al., 1994

MIBC
surface tenion, mN/m
60 MIBC
bubble size, mm
1.2
40
0.8
20 Laskowski, 2004 0.4

CCC
Laskowski, 2004 CCC
CMC
0 0
0.1 1 10 100 1000 0 4 8 12 16 20
frother concentration, mmol/dm 3 frother concentration, ppm
a b
Fig. 12.40. Influence of concentration of frother on properties of the gas–water system for methyl isobu-
tyl carbinol (MIBC) as an example: a – surface tension (Aston et al., 1983), CCC (Laskowski, 2004), and
CMC, b – size of bubbles (Tucker et al., 1994) CCC and CMC. For MIBC 1 ppm ≈10-2 mol/dm3
When many bubbles are created in flotation cells the relation between bubble size
and surface tension is less obvious because the size of bubbles reach constant size at
very small concentrations when the surface tension drop is not significant in relation
to water. Figures 12.40a,b show that methyl isobutyl carbinol (4-methyl-2-pentanol or
MIBC) lowers surface tension of aqueous solution down to about 200 mmole/dcm3
(CMC), while the size of bubbles stabilizes at 0.1 mmol/dm3 or 10 ppm called by Las-
kowski (2004) the critical coalescence concentration or CCC.
According to Laskowski (1998) reagents with branched molecule are better froth-
ers because they form less compact films at the gas–water interface.
Air bubbles dispersed in water travel up to the liquid surface and either burst or
remain on the surface in a form of froth. And this is another role of frothers in flota-
tion – to form stable froth. The presence of a stable froth is necessary for keeping
floated out material on the water surface. This provides enough time to remove float-
ing particles. The froth, however should not be too stable. A stable froth creates prob-
lems with its handling. The froth exists when a thin water film is formed around the
bubbles and it has the structure preventing a rapid leaking of water from inter-bubble
space. This phenomenon prevents bubble disappearance, that is coalescence. The
frothers, sorbing themselves on the bubble surface, cause the formation of “rough”
layers. This is confirmed by the fact that pure liquids do not form froth, due to the fact
that water easily leaks out of thin layers, and therefore the bubble coalescence pro-
gresses rapidly.
Froth structure depends on the place in the flotation cell and the extent of mixing
(Fig. 12.41). Froth layer undergoes destruction due to bubble breaking as a result of
12. Flotation 329
slow dewatering, while new bubbles are continuously transported from the bottom of
the cell. Upper part of the froth is usually of hexagonal shape. In the middle part of the
froth the bubbles are round and in the bottom part the froth contains bubbles sus-
pended in water. Hydrophobic particles, present in froth, stick to bubbles. In the
spaces between bubble bulk water contains hydrophilic particles mechanically trans-
ported to the froth and their amount is usually directly proportional to the amount of
water in the froth (Ross, 1991; Waksmundzki et al., 1972). The decrease in content of
gangue particles in froth voids can be regulated by froth hydration, i.e. quantity and
kind of frother, as well as by appropriate procedures, like spraying the froth with wa-
ter. The relation between frothing and flotation is not a simple, since both processes
are effected by the type of collector used for flotation as well as by interaction be-
tween frother and collector. Many problems involving froths, their stability and inter-
actions between frothers and collectors have not been solved. More details about froth
in flotation can be found in the “Frothing in flotation II” (Laskowski and Woodburn,
1998).
froth height in flotation cell
pulp
Fig. 12.41. Froth and change of it structure with the position in the flotation cell. Not to scale.
Another role played by the frother is acceleration of the flotation process. This
takes place due to diminishing the so called induction time, i.e. the time needed by a
bubble to attach itself to a particle and to remove water film between bubble and parti-
cle. The effect of diminishing induction time results from the fact that good frothers
rapidly change their orientation at interface water-gas at the moment of the collision
between bubble and particle on which the collector molecule is adsorbed. To do this
the frother molecule has to behave at interface as a gas film, i.e. it should not associ-
ate but to evenly spread at the interface in all stages of flotation, starting from a bubble
collision with a particle to the formation a particle–bubble aggregate (Laskowski,
1998).
Table 12.26. Flotation frothers (Laskowski 1998, with some modifications)

Group Frother
1. aliphatic alcohols
a) linear from amyl to decanol
iso-amyl
b) branched
2-ethyl hexanol, methyl isobutyl carbinol (MIBC)
c) with additional group diacetone
2. Cyclic
a) linear cyclohexanol
b) branched α-terpineol
cresols
3. Aromatic
xylenols
4. Alkoxy-hydrocarbons 1,1,3-triethoxybutane
R(X)nOH
5. Polyglycols and hydrocarbon poly- R=H or CnH2n+1 or any hydrocarbon
glycol ethers X=EO (ethylene oxide), PO (propylene oxide)
BO (butylene oxide) from 3 to 7
Many chemical compounds can be used as frothers and their classification takes
into account either their properties or chemical composition. Classification regarding
the properties divides the frothers into acidic, basic, and neutral. Another classification
follows the division into normal (standard) frothers and frothers–collectors. The froth-
ers can be divided into aliphatic, cyclic and aromatic alcohols, hydrocarbons with eth-
oxy groups and polymeric frothers based on polyglycols (Table 12.26) frothers. Ali-
phatic alcohols contain from four to about 10 carbon atoms and one or more –OH
group. The simplest frothers are aliphatic normal alcohols, such as pentanol, hexanol,
heptanol. Their solubility in water decreases as molecular weight increases. The solu-
bility of alcohols increases when they contain additional oxygen atom, like in diace-
tone alcohol. In the industry, branched frothers are commonly used. Methyl isobutyl
carbinol (MIBC) belongs to this group. Long-chained amines, sulfonates, sulfates,
and fatty acid (Klimpel and Hansen, 1988) are frothers which play also the role of col-
lectors. They are used in flotation of non-sulfide ores. The simple and complex alco-
hols, especially having high molecular weight and weakly soluble in water belong the
nonionic frothers-collectors family. When accompanied by other collectors, they are
called promoters.
The properties of froth also depend on salts present in the solution. Investigations
by Iskra and Laskowski (unpublished but discussed in details by Laskowski, 1969, pp.
91-93) showed that salts make the foam less voluminous and containing less water. It
is caused by so-called salting out of the frother leading to a decreasing surface tension
12. Flotation 331
of frother aqueous solution. Therefore, flotation in the presence of salts generally pro-
vides better results (Ratajczak, Drzymala, 2003)
12.5.3. Activators
Activators initiate or improve flotation in the presence of collectors. The question
whether or not a reagent is an activator depends not only on reagent properties but also
on its interaction with the collector. It happens that activator in the presence of another
collector functions as a depressor. It also depends on reagent concentration. Most of-
ten cations of hydrolyzing multivalent metals ions are activators. Their list includes
Fe2+, Fe3+, Al2+, Pb2+, Mn2+, Mg2+, Ca2+, and other ions. Some anions are also activa-
tors, especially fluoride and sulfide ions. There are other activators with a complex
structure, known as promoters, which are applied in selected flotation systems, e.g. in
coal flotation (Sablik, 1998).
The metal cations become activators due to sorption at the solid-water interface.
The adsorption of a particular metal cation takes place within precisely determined pH
ranges in the Stern layer of the electric double layer. This results from the difference
in the sign of the electric charge of the ion and the surface. The experimental data in-
dicate that the best sorption occurs when the metal ion is in the form of a monohy-
droxycomplex (Me(OH)n+). Figure 12.42 shows flotation of quartz in the presence of
different metal cations with the use of sulfonate as a collector (Fuerstenau and Palmer,
1976). Similar relations have been recorded for many other anionic collectors.
100
80
recovery, %
60
Fe 3+ Al3+ Pb 2+ Mn 2+ Mg 2+ Ca 2+
40
20
0
0 2 4 6 8 10 12
pH
Fig. 12.42. Activating action of hydrolyzing cations of multivalent metals. Flotation of quartz in the pres-
ence of 10–4 M sulfonate. There in lack of flotation in the absence of activating cations. In presence of
10–4 M of cations there is flotation above characteristic pH, at which the dominating species in solution is
the monohydroxy complex MeOHn+ (Fuerstenau and Palmer, 1976, with permission of SME)
Another mechanism of surface activation with metal cations takes place with sul-
fide minerals. There is a displacement of the original metal cations from the surface by
the added cations as it was described by the reaction in Table 12.27 for sphalerite. The
activation reactions are of electrochemical character. On the basis of free enthalpy of
reaction ∆Gr0, a series of metal cations abilities to activate sulfides can be arranged. In
the activation series the greater ability of activation the more negative is the value of
∆Gr0.
Table 12.27 shows the values of ∆Gr0 for the activation reactions of selected sul-
fides with different metal ions. The values represent ∆Gr0 (kJ/mol) for the reactions of
sulfides activation with metal ions calculated for the reactions MeIS + MeII2+ = MeIIS
+ MeI,2+, taking into account the free energy of formation (∆Gf0) of each species i and
the relation: ΔGr0 = Σνi ΔG 0f in which v means reaction stoichiometric coefficient
(see 12.73).
The ∆G0f values are taken from Woods and Garrels (1987), but there are many
other sources of such thermodynamic data, e.g. CRS (1986).
Table 12.27. Activation reaction of sphalerite with selected metal cations

and calculated free enthalpy of the reactions
Activation reaction free enthalpy, ΔGr0 (kJ/mol)
2+ 2+
ZnS +Fe =FeS+Zn 35.2
ZnS +Pb2+=PbS+Zn2+ –17.3
ZnS +Cu2+=CuS+Zn2+ –62.9
ZnS +2Ag+=Ag2S+Zn2+ –142.3
Table 12.28. Free enthalpy of the activation reactions for sulfides reacting with metal ions
Fe2+ Zn2+ Pb2+ Cu2+ Ag+
FeS –35.2 –52.5 –98.1 –177,5
ZnS 35.2 –17.3 –62.9 –142,3
PbS 52.5 17.3 –45.6 –125,0
CuS 98.1 62.9 45.6 –79,4
Cu2S 170.7 136.1 118.2 –6,8
Ag2S 177.5 142.3 125.0 79.4
It can be concluded from Table 12.28 that pyrrhotite (FeS) can be activated with all
considered cations (∆Gr0 is negative), sphalerite with all cation except Fe3+, while ga-
lena (PbS) only with Cu2+, and Ag+ ions. Both cooper sulfides can be activated only
with Ag+, while argentite (Ag2S) cannot be activated at all (∆G0j is positive). It can be
noticed that activation series for sulfides is, with some exceptions, identical with the
so-called galvanic series of metals.
The activation process induces sphalerite flotation in the presence of xanthates af-
ter its activation with copper ions (Laskowski, 1969). The activation process can also
be harmful as it can lead to a decreased selectivity of sulfides flotation. This phe-
nomenon can be observed during galena separation from copper minerals (Bigosinski,
1998), because galena surface becomes similar to copper minerals.
12. Flotation 333
Fluoride ions can also be applied as activators, especially for silicates, e.g. beryl
(Fig. 12.43), while quartz itself is not a subject to activation.
1000 1000
dodecylamine, mg/dm 3
beryl beryl
flotation
flotation
oleate, mg/dm 3
100 100 flot

atio
n wi t
h N
aF
no flotation
10 10
with NaF
no flotation no flotation
1 1
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH
Fig. 12.43. Fluoride ions as activating and depressing media in flotation of silicates:
a – regions of lack of flotation in the presence of oleate as collector,
b – regions of activation in the presence of dodecyl amine as collector (Manser, 1975)
Some authors consider silicate ions as activators. It was proved that at low concen-
trations the SiO32- ion facilitates flotation of some minerals (Kłasen and Mokrousov,
1963) (Fig. 12.44) and coal (Sablik, 1998). The mechanism of this process, however,
is not fully known. It is assumed that silicate ions remove foreign ions and particles
from the surface, which are present there due to crushing, grinding and other form of
contacts with small particles present in the aqueous solution, by peptization (decoagu-
lation).
100
80
cerussite
recovery, %
60
40 malachite
20 apatite
0
0 1 2 3 4 5
dosage of sodium silicate, mol/Mg
Fig. 12.44. Silicate ions improve flotation results for many minerals,
provided that their concentration is not too high (Klassen and Mokrousov, 1963)
The sulfide ion can also play a role of activator. They are applied as activators in
flotation of oxidized minerals. They are added to the system before flotation with xan-
thates. However, too high sulfide concentration depresses flotation, since the sulfide
ions compete with xanthate ions for adsorption sites on the surface. Therefore, it is
very often necessary to remove sulfide ions before flotation, that is immediately after
the activation process. The mechanism of activation with sulfides relies on forming
metal sulfides on the surface of an activated mineral (Fuerstenau and Han, 1988).
According to Sablik (1998), difficult-to-float coals can be activated with hydrocar-
bon polyglycol ethers, esthers and amines (promoters). The mechanism of their activa-
tion is not known.
12.5.4. Depressors
Depressors are used to elevate flotation selectivity. The selectivity is achieved by
depressing one or more components of the flotation suspension while the valuable
mineral should float without changes. Sometimes the selectivity can be obtained in a
reverse process, i.e. through depression of a useful component, while gangue under-
goes flotation. Depressor can be organic, inorganic, acid, salt, base, redox, and com-
plexing reagent, etc. The list of often used depressors is presented in Table 12.29.
Table 12.29. Depressors used in flotation

Inorganic depressor Formula Organic depressor
Alum KAl(SO4)2 starch
Ammonia NH3 quebracho
Calgon polyphosphates tannin
Sodium cyanide NaCN lignin derivatives
Sodium dichromate Na2Cr2O7 synthetic polymers
Phosphates various acetic acid
Sodium silicate Na2SiO3 dextrin
Fuoric acid HF humic acid
Sulfuric acid(IV) H2SO3, SO2 cellulose derivatives
Sulfuric acid(VI) H2SO4 alginate
Hydrochloric acid HCl chitin derivatives
Sodium sulfate(IV) Na2SO3 citric acid
Zinc sulfate(VI) ZnSO4 hydrazine
Caustic soda NaOH thioglycolic acid
Oxygen O2 chelating compounds
Lime Ca(OH)2 electrons (redox substances)
Sodium carbonate Na2CO3
Sodium hydrosulfide NaHS
Depressors work through different mechanisms. In this monograph they were clas-
sified into 3 groups: depressors desorbing collector from the surface, redox depressors
12. Flotation 335
changing the chemical composition of surface, and depressors which decompose col-
lector.
12.5.4.1. Depressors acting through adsorption

Depressors, when adsorbed on mineral surface, block its hydrophobization, that is
make the sorption of a collector impossible. When a collector is already present on the
surface, the depressor causes its desorption. Activity of depressors depends both on
their properties and reactivity of the floating material. Generally, depression results
from a competition between depressor and collector ions for the place on the surface
of floated particles. The mechanism of action of adsorbing depressors can be described
with the use of exchange reactions. Such a description, however, is not simple, since
the chemistry of surface reactions is not well known and there is a lack of a complete
theory of electric double layer and exact relation between surface charge and surface
potential.
The simplest depressor competing with the collector for the places on particle sur-
face is hydrogen ions (H+). The hydrogen ions simultaneously regulate acidity of the
solutions. The list of acids include H2SO4 and HCl, while lime, sodium carbonate and
caustic soda (NaOH) are the most often used reagents for alkalinity regulation in flota-
tion.
The results of flotation involving pH regulators are usually presented as recovery
versus pH curves. Figure 12.45 shows the influence of pH on chalcocite flotation. It
can be concluded from the figure that cease of flotation takes place both at low and
high pH values.
100
80
recovery, %
60
chalcocite
KButX 0.0001M
40
20
0
2 4 6 8 10 12
pH
Fig. 12.45. Influence of pH on flotation of chalcocite in the presence of butyl xanthate.

Too high and low pH depresses flotation of minerals
The concentration of hydrogen and hydroxyl ions is presented on one pH scale,

since both parameters are related. This results from dissociation of water molecules
into ions, according to the reaction:
H2O = H+ + OH– (12.49)

Reaction equilibrium constants are expressed as a quotient of chemical products
activities to chemical substrates activities taking part in the reaction. It should be no-
ticed that species on the right side of the reaction are products while on the left side
are substrates. For reaction (12.49) the equilibrium constant is given by:
(H ) (OH )
+ −
c0 c0
K Wa = , (12.50)
(H 2 O )
c0
where the round parenthesis denote activity. Activities are expressed in kmol/dm3. The
term co is constant and always is equal to 1 kmol/m3, and is introduced to the equation
for equilibrium constant to guarantee that the constant is a dimensionless quantity. As
co always equals 1 kmol/m3, this quantity is customarily omitted when writing or cal-
culating equilibrium constants. It will be also practiced in this book. Activity, denoted
( ) or a, is, in turn, the product of concentration ([ ] or c) and activity coefficient f,
then:
a = ( ) = c f = [ ] f. (12.51)
For low concentrations the activity coefficient is close to 1. Since in mineral proc-
essing low concentrations or reagents are used, it is usually assumed that the activity
coefficient is equal to 1. The expression for the equilibrium constant is as follows:
KW =
( )(
H + OH −
≅
) [ ][
H + OH −
.
] (12.52)
(H 2O ) [H 2O]
At room temperature (21oC ≈ 298K) quantity KW amounts 1.80.10-16 and since at
this temperature concentration of water in water [H2O] is 55.56 koml/m3, the product
of OH- and H+ ions concentration, i.e. ionic water product equals:
(H+) (OH–) = KW [H2O] = 1.00·10–14. (12.53)
After finding the logarithm of both sides of this equation:
pH = 14 – pOH, (12.54)
where pH=–log(H+) and pOH=–log(OH-).
This equation connects concentration of H+ ions with concentration of OH- ions.
Symbol p stands for mathematical function –log and it can used not only for H+ or OH-
ions, but also any other ion, e.g. pAg = –log(Ag+), pBa = –log(Ba2+) etc.
The mechanism of action of the pH regulating depressors, similarly to the many
other depressors, relies on competition with collector ions or molecules. Generally, the
competition can be expressed in the form of an exchange reaction:
–MC + D = –MD + C (12.55)
12. Flotation 337
where:
D – depressor ion
C – collector ion
–MC – collector adsorbed or forming surface compound with mineral
–MD – depressor adsorbed or forming surface compound with mineral.
When reaction strictly follows Eq. (12.55), it is possible, on the basis of the equi-
librium constants of this reaction,
[C] [− MD]
KM = 12.56),
[D] [− MC]
to obtain the so-called Barski relation
[C] = [− MD] K = const , (12.57)
[D] [− MC] M
which says that for a particular mineral-collector-depressor system a decrease in flota-
tion takes place at a constant ratio between the concentration of collector and depres-
sor. It means that if collector concentration increases in the solution, to obtain flotation
depression we should add more depressor and that the stoichiometric ratio of C to D
equals 1. Qualitatively, such a relation holds for many depressor-collector systems.
For example, Fig. 12.46 shows lines of flotation – no flotation transition for sulfides
floated in the presence of ditiophosphate at different pH values.
3
collector concentration, mg/dm
0.6 pyrite
0.5
galena
0.4
0.3
chalcopyrite
0.2
0.1
0
4 6 8 10 12
pH
Fig. 12.46. Competition between collector and depressor expressed as edges of transition between flota-
tion and lack of flotation for selected sulfides in the presence of dithiophoshate as collector. The region of
flotation occurs on the left side of the curve. The curves were drawn basing on attachment of particles to
bubbles measurements (data after Wark and Cox, 1934)
However, it was claimed by Chander (1988) (Table 12.30) that the stoichiometric
coefficient for different collector–OH- systems is not equal to 1, but it follows a gen-
eral relation:
[C] =
[collector ] = const
[D] 12.58
[OH ]
y y
−
Table 12.30. Values of parameter y (Eq. 12.58) for selected mineral–collector systems
(here pH regulator determines concentration of OH- ions) (after Chander, 1988)
Mineral–collector y pH range
Galena/KEX 0.65 9–12
Galena/KDEDTP 0.53 6–9
Sphalerite/DEDTC 0.56 6–8
Sphalerite/DBDTC 0.72 7.5–10
Sphalerite/DADTC 0.75 8–11
Chalcopyrite/NaDEDTP 0.71 8.5–11
Chalcopyrite/KEX - 11–13
Pyrite/NaDEDTP 0.62 4–6
Pyrite/KEX - 10–12
E – ethyl, D – di, T– tio, A – amyl, B – butyl, X- xanthate, P – phosphate, C – carbonate , K – potassium,
Na – sodium. In highly alkaline environment Eq.12.58 is not working
A discrepancy between the Barski formula and Eq. 12.58 was explained by
Chander (1988) by the differences between the speeds of collector desorption and de-
pressor adsorption, i.e. lack of equilibrium of the exchange process. A different expla-
nation is possible if we take into account that surface reactions can considerably differ
from hypothetical reaction (12.55). Investigation on sulfides proved that on their sur-
− SH
faces there dominate − SMeX groups in the presence of xanthates in neutral medium,
−S−
while in alkaline medium − SMeX groups are the prevailing ones (Rönngren et al., 1995).
Displacing xanthate groups with OH- groups leads to formation of the –SH–SMeOH
and –S-–SMeOH groups. Depressing activity of the hydroxyls can be presented as the
reaction:
− SH − S−
(12.59)
− SMeX + 2OH − = −SMeOH + X − + H 2 O
In order to write down equilibrium constant of surface reaction (12.59), one should
take into account the fact that concentration of ions taking part in the reaction close to
the surface (Ip) is different from the one in the bulk solution (I) and is determined by
the relations (Drzymala et al. 1979).
⎛ Fψ 0 ⎞ ⎛ Fψ 0 ⎞
I −p = I − exp⎜ ⎟; I p+ = I + exp⎜ − ⎟ (12.60)
⎝ RT ⎠ ⎝ RT ⎠
12. Flotation 339
where:
Ip – concentration of ion at interface
I – concentration of ion in the bulk solution.
There a negative sign is used for positive ions and a positive sign for negative ions
while Ψo stands for the surface potential. For negatively charged surface Ψo is nega-
tive, while for the surface charged positively is positive. Therefore, the surface equi-
librium constant can be presented as:
[X ] exp ⎛⎜ FRTψ ⎞⎟ [− S
− 0 −
− SMeOH ]
K OH = ⎝ ⎠ =
⎛ Fψ 0 ⎞ 2
[OH ]
− 2
exp ⎜ ⎟ [− SH − SMeX − MC][H 2 O]
⎝ RT ⎠ (12.61)
=
[X − ][− S− − SMeOH ] .
[OH − ]2 exp ⎛⎜ FRTψ 0 ⎞⎟ [− SH − SMeX − MC][H 2O]
⎝ ⎠
Assuming that flotation depression takes place at a constant–S–SMeOH to –SH-

SMeX ratio, we obtain the relation:
[X ] −
= K * exp
Fψ 0
[OH ] − 2 RT
.
(12.62a)
Thus, Eq. 12.62a is a slightly different equation from that of Barski. It can be con-
cluded from the above equation that depression depends not only on concentration of
collector ions and depressor ions, but also on electric surface potential. The latter one
can change as pH changes (from the point of pHpzc) in nernstian mode, i.e. 59mV/pH.
This value, however, is usually lower, which can be expressed by introducing an addi-
tional factor m. Then Eq. (12.62a) assumes the form:
[X ] −
= K * exp
Fm (pH pzc − pH)
[OH ] − 2 RT
.
(12.62b)
The equation obtained is quite complicated for it contains three variables. On the
basis of measurements for many oxides (Drzymala, 1979), it has been shown that for
surface reactions involving ions empirical dependence between surface potential and
OH- ions concentration can be used:
Fψ 0
[OH ] − − ( from 0.5 to 0.7 )
= K x exp
RT
(12.63)
where Kx is a constant.
The above relation has been derived by combining the theoretical equation for the
surface equilibrium constant with an empirical expression of the following type:
pH = pK – n log [(1/α) – 1)] (12.64)

which represents a modified Temkin isotherm. In Eq. 12.64 α denotes the degree of
surface site dissociation, while n and pK are empirical constants.
Introduction of the above relation into general equation (12.63) causes that the ef-
fect of OH- ions on cease of flotation is not so strong as a part OH- ions is involved in
establishing the surface potential. Due to that fact, the general form of the Barski
equation becomes:
[X ]−
≈ k,
[C] ≈ k.
[OH ]
(12.65)
[D] y
− y
There is a large group of depressors which especially strongly bind with ions pre-
sent in the solution as free or surface ions. These depressors are capable of forming
chemical compounds of high stability. If a complex is especially stable, it is not possi-
ble to desorbed it from the surface by using excessive amount of a collector, since in
that case enormously high and impossible-to-achieve collector concentration, higher
one than its solubility, should be used. Then, the complex constant can be accepted as
a parameter characterizing the depressor’s strength. It results from the assumption that
the more stable complex the better depressor. To exemplify this statement, Table
12.31 shows concentrations of cyanide needed for sulfides depression in the presence
of 1.56.10-4 mole/dm3 of potassium ethyl xanthate along with equilibrium constants of
forming cyanide complexes with iron and copper ions. It can be concluded from the
mentioned table that stability constants for iron ions are much higher than those of
copper, therefore, cyanide concentration needed for copper sulfides depression is con-
siderably higher than that of iron sulfides. For mixed iron-copper sulfides, for example
for bornite, the critical depression concentration assumes intermediate values.
Table 12.31. Comparison of concentration of cyanide necessary for depression of sulfides subjected to
flotation with 1.56·10–4 mol/dm3 of potassium ethyl xanthate and stability constants of formation of cya-
nide complexes with iron and copper ions. Data after Rogers, 1962 and Chander, 1988
Concentration of
pH range of
Mineral Chemical formula CN– pβ = –log β
flotation
(mol /dm3)
1 2 3 4 5
Pyrite FeS2 7–9 4.0·10–6 36 (Fe2+(CN)64–)
Chalcopyrite CuFeS2 7–11 1.6·10–5
Marcasite FeS 8–10 1.0·10–4 36 (Fe2+(CN)64–)
Bornite Cu5FeS4 8.5–13 2.5·10–4
Tetrahedrite Cu3SbS3,25 8.5–13 8.0·10–4
Covellite CuS 9–12 1.4·10–3 26 (Cu2+(CN)42–)
20 (Cu1+(CN)2–)
Chalcocite Cu2S 9–13 7.0·10–3
28 (Cu1+(CN)43–)
12. Flotation 341
A high constant of iron ions complexation is, evidently, responsible for a very sta-
ble, irreversible depression of iron sulfides, while copper sulfides depression takes
place at much higher CN- concentrations. And it can be reversed, which means that
after removing CN- ions from the solution, they will float again. These questions are
very important when it comes to selective flotation of sulfide polymetallic ores.
The properties of other complexing depressors can be estimated in the same way.
Equilibrium constants can be found in literature or calculated on the basis of thermo-
dynamic data, especially using free enthalpy of forming substances ΔG0f (Sillen and
Martell, 1964; Garrels and Christ, 1965; Naumov and Others, 1971; CRC, 1986). For
this purpose a general reaction of complex forming can be written:
Mem+ + nX– = MeX nn − m . (12.66)
Since the central Mem+ ion can, depending on reaction conditions, binds different
number of complexing ions (ligands), therefore there do exist different constants of the
complexing equilibrium. There are constants expressing subsequent stages of a ligand
bonding, assigned with symbol K, as well as cumulated constants expressing transition
of the central ion to the state of making complexes with certain number of ligands,
assigned with symbol β, as presented in Table 12.31. A general equation determining
stability constants for the reaction of complexing metal ions with a ligand (depressor)
is:
βn =
[MeX ] = Π K ,
n−m
n
n
(12.67)
[Me ][X ]
m+ − n i =1
i
where Π is a mathematical symbol for multiplying.

For instance, for the reaction of divalent metal ion with four cyanide ions:
Me2+ + 4CN– = Me(CN) 24 − . (12.68)
The cumulated constant has the form:
β4 =
[Me(CN) ] 2−
4
, (12.69)
[Me ][CN ]
m+ − 4
while K4 will describe the reaction:

Me(CN)3− + CN– = Me(CN) 24 − , (12.70)
and the expression for the constant is:
K4 =
[Me(CN) ] . 2−
4
(12.71)
[Me(CN) ] [CN ]
−
3
−
Equilibrium constants of any arbitrary chemical reaction can be calculated on the

basis of the free enthalpy of the reaction, since reaction equilibrium constant K, and
also β, is related according to:
ΔGr0 = –RT ln K, (12.72)
and the ΔG0r value can be calculated from free enthalpy of formation ΔG 0f
ΔGr0 = Σνi ( ΔG 0f )i , (12.73)
where Σ means the sum of products of standard free enthalpy of formation of particu-
lar species i taking part in the reaction and reaction stoichiometric factor v being posi-
tive for the products and negative for the substrates. As it has already been mentioned,
ΔG0f values are available in literature.
12.5.4.2. Redox depressors

The course of some minerals flotation can be modified by reagents which are ca-
pable of changing the character of the molecules, collector ions, surface species in an
electrochemical way. They are redox substances and they can be both oxidizing and
reducing reagents. The redox reagents are able to exchange electrons and oxidizing
species accepts electrons undergoing reduction, while reducing species give up elec-
trons and undergo oxidations. The reactions involving electrons can cause changes in
chemical composition and structure of substance or conditions under which the sorp-
tion of flotation reagents occurs through the change of redox potential. The simplest
way of effecting flotation with redox potential would be regulation of electrons con-
centration in the flotation system. Unlike protons (H+), there are no free electrons in
water solution, since they can exist only in a bonded form with the ions or molecules
present in the solution. Therefore, the description, performance, as well as control of
redox reaction are complicated.
Basic relation describing redox reaction is called the Nernst equation. Its traditional
derivation is based on difficult notions of solution pressure or osmotic pressure (Pa-
jdowski, 1993). A better derivation was used by Chander (1988). It is based on reac-
tion equilibrium constant. To derivate the expression for redox reaction involving
electrons, first the reaction should be written in a general form:
oxidized form of a substance (u) + electron (e) =
(12.74)
= reduced form of a substance (z),
where e denotes electron.
A practical example of such a reaction can be iron(III) ions reduction to iron(II)
ions:
Fe3+ + e = Fe2+ (12.75)
12. Flotation 343
which can be written in a general form as:

u + e = z. (12.76)
The equilibrium constant for this reaction is given by:
K=
[z ] , (12.77)
[u ] [e]
where the square brackets mean concentration of a particular form which are ex-
pressed in kmol/m3.
Now both sides of the equation undergo logarithm operation and then are multi-
plied by RT:
RT ln K = RT ln
[z ] . (12.78)
[u ] [e]
The expression RT ln K equals the standard potential of chemical reaction -ΔG0,
therefore Eq. (12.78) can be written as:
− ΔG 0 = RT ln
[z ] . (12.79)
[u ] [e]
If this equation is divided by nF, i.e. the product of the Faraday constant and the
number stating how many electrons are exchanged in the process, i.e. n, then:
− ΔG 0 RT
= ln
[z ] . (12.80)
nF nF [u ] [e]
Because –ΔG0/nF is known as the standard redox potential E0, introducing this in-
formation results in:
E0 =
RT
ln
[z ] = − RT ln [u ] − RT ln e = − RT ln [u ] + E , (12.81)
nF [u ] [e] nF [z ] nF nF [z ]
which means that expression E0, featuring equilibrium constant of the reaction with
participation of electrons, was divided into two segments. The first one depends on the
concentration of the oxidized and reduced forms of the considered substance while the
second one characterizes electrons activity.
Equation (12.81) is the well known Nernst relation. It can be seen after its rear-
rangement:
RT [u ] RT
E = E0 + ln = E0 − ln K * , (12.82)
nF [z ] nF
where K* stands for the equilibrium constant of redox half reaction without taking into
account electrons in an explicit form, but taking into account all species taking part in
the reaction, including H+ and OH- ions. It should be remembered that the concentra-
tion of the oxidized species is displayed in the denominator and the reduced form in
the numerator.
As it has already been mentioned, depressing action of redox substances relies on
transporting electrons which participate in forming new chemical compounds and de-
termine flotation. In the case of some metal sulfides (MS) in the presence of xanthate
(X), the following electrochemical oxidation reactions can take place:
X– → Xads + e, (12.83)
2X– → X2 + 2e, (12.84)
2X– + MS → MX2 + S + 2e, (12.85)
2X– + MS + H2O → MX2 + SO 24 − + 8H+ + 8e, (12.86)
MS + 6H2O → M(OH)2 + SO 24 − + 10H+ + 8e, (12.87)

where:
MeX2 – metal xanthate
X– – xanthate ion in solution
Xads – xanthate adsorbed on the surface
X2 – dixanthogen (on surface as oil or solid).
The species X2, MeX2 and Xads are generally hydrophobic, M(OH)2 is hydrophilic,
while MeS is hydrophilic or slightly hydrophobic (Chmielewski, 1990). Therefore,
flotation or its lack depends on the degree of surface occupation by these species. Yet,
the condition under which the reactions presented in Eqs. (12.83)-(12.87) can take
place is the presence of a reducing itself substance, i.e. the one accepting electrons.
The simplest of such substances can be oxygen, which, after accepting electrons, un-
dergoes reduction to O2-.
O2 + 2H2O + 4e → 4OH– (12.88)
There exist many substances capable of redox reactions. Some of them are pre-
sented in Table 12.32.
It often happens that hydrogen or hydroxy ions take part in electrochemical reac-
tions (e.g. (12.86)-(12.87). It is then convenient to present the relation between redox
potential of these reactions and pH in E-pH diagrams also called the Pourbaix dia-
grams (1963) (Letowski, 1975). The diagram for the Cu-H2O system, after Letowski
(1975), is shown in Fig. 12.47. E-pH diagrams can be plotted for any system in which
12. Flotation 345
in reaction between the examined substance and the collector, depressor and/or activa-
tor takes place with an exchange of electrons (Lekki, 1979).
1.4
1.2
-6
-2
-6
-4
0
10
10
10
10
10
1
2-
Cu 2+
2
0.8
HCuO 2 , CuO
CuO
Eh potential, V
0.6
-
0
0.4 10
0.2
-2 -4 -6 Cu 2O
0 10 10 10
-0.2
-0.4
-0.6 Cu
-0.8
-1
0 2 4 6 8 10 12 14
pH
Fig. 12.47. Eh–pH diagram for Cu–H2O system at 25 °C (298 K). Diagram is based on reactions: Cu2O +
H2O = 2CuO + 2H+ + 2e (E = 0,747 – 0,0591 pH); 2Cu + H2O = Cu2O + 2H+ + 2e (E = 0,471 – 0,0591
pH); Cu = Cu2+ + 2e (E = 0,337 + 0,0295 lg [Cu2+]); Cu2O + 2H+ = 2Cu2+ + H2O + 2e (E = 0,203 + 0,0591
pH + 0,0591 lg [Cu2+]); Cu2+ + H2O = CuO + 2H+ (pH = 3,44 – 0,5 lg [Cu2+]) (Łętowski, 1975)
On the basis of the Pourbaix diagrams it is possible to predict what E and pH

should be used in order to produce proper species on the surface of particles. E value
can be either calculated or taken from the tables. It can also be measured with the use
of properly prepared electrode, usually made of platinum. The E values are a measure
of electrons activity in the system. However, the E values are not absolute, but they
are related to a reference system, called standard system, which usually is the potential
of the so called hydrogen electrode having E values related to the to normal hydrogen
electrode denoted as Eh. In practice, potentials are measured in reference to other elec-
trodes, usually saturated calomel electrode. The standard potential of such electrode
must be taken into account. The normal potentials for selected redox reactions related
to the hydrogen electrode scale are presented in Table 12.32. To directly determine
redox properties of a particular substance the best way is to compare E0 values for the
reactions written in such a way that on the left side there are electrons. If OH- ions
take part in the reaction they should be rewritten in such a way that H+ ions should
occur without OH- ions. As it can be concluded from Tab. 12.32, the substance of the
highest redox potential, i.e. oxidizers, are S2O8 2- ions, then hypochlorite (ClO-) and
permanganate (MnO4-). Strong reducers are also hydrazine, zinc and sulfide ions S2-,
since their redox potential is highly negative.
Table 12.32. Standard potentials for selected materials based on standard hydrogen electrode potential
(data after Pajdowski, 1993; Barycka and Skudlarski, 1993; Schmidt, 1983 (hydrazine), and other sources
Normal potential
Electrode reaction Short notation
Eh0 (V)
S2 O82 − + 2e = 2 SO 24 − S2 O82 − / SO 24 − 2.050
– + – – –
ClO + 2H + 2e = Cl + H2O ClO /Cl 1.640
MnO −4 +
+ 8H +5e = Mn + 12H2O 2+
MnO −4 /Mn 2+ 1.510
–
Cl2 +2e = 2Cl Cl2/2Cl– 1.360
O2 + 4H+ + 4e = 2H2O O2/O2– 1228
Fe3+ + e = Fe2+ Fe3+/Fe2+ 0.771
O2 + 2e + 2H+ = H2O2 O2/H2O2 0.680
(CN)2 + 2H+ + 2e = 2HCN (CN)2/HCN 0.370
Fe(CN) 36− + e = Fe(CN) 64 − Fe(CN) 36− / Fe(CN) 64 − 0.363
2+ +
Cu + e = Cu Cu2+/Cu+ 0.167
2H+ + 2e = H2 H+/H2 0.000
SO 24 − + 2H+ + 2e = SO 32 − + H2O SO 24 − / SO 32 − –0.103
+
N2 + 4H + 4e = N2H4 (hydrazine) N2 /N2– –0.333
S + 2e = S2– S/S2– –0510
Zn2+ + 2e = Zn Zn2+/Zn –0.763
100
80
Cu 2S
recovery, %
60 Cu 5FeS 4
40 CuFeS 2
FeS 2
20
0
-0.3 -0.2 -0.1 0 0.1 0.2 0.3 0.4
potential, V
Fig. 12.48. Relationship between flotation recovery of sulfides and potential applied to a platinum wire
immersed in aqueous suspension of a sulfide containing ethyl xanthate at 1.44·10–5 kmol/m3. After
Richardson and Walker, 1985
Sulfides flotation and their depression can be regulated not only with the use of re-
dox substances, but also by applying external potential through the wires conducting
electrons or electrodes immersed in the suspension. It is necessary to provide a contact
between each particle and the electrode delivering or removing electrons and the parti-
cles have to be conductive. Although technically difficult, such flotation can be per-
formed. Figure 12.49 shows the effect of applied potential on flotation of selected
12. Flotation 347
metal sulfides. It can be concluded from the mentioned figure that by the means of
applied potential one can not only depress, but also activate sulfides for flotation.
12.5.4.3. Depressors decomposing the absorbed collector

Some depressors are capable of destruction of collector molecule or ion absorbed
on a mineral surface. Sulfate(IV) ion can serve as an example, which, as it is sup-
posed, decomposes xanthate present on the surface. It works according to the reaction
(Yamamoto, 1980):
2C2H5OC(S)SK + 2K2SO3 + 2KHSO3 + 2O2 → 2C2H5OC(S)SOK →
(12.89)
2C2H5OH + 4K2S2O3 + 2CO2
Reaction (12.89) takes place mostly in neutral pH regions.
Removing collectors from particles surface is important when there is a necessity
of flotation with one collector and next with another one. In such cases desorption of
the first collector using appropriate depressing ions, for instance pH regulators, is car-
ried out. In literature, however, there are not many examples of such depressors.
12.6. Flotation of mineral matter

Many ores and raw materials can be upgraded by flotation. Since ores contain min-
erals of different properties, a proper collector is needed for flotation of one or a group
of minerals. Eigeles (1964) classified materials into six groups according to their flota-
tion response to certain collectors. It is presented in Table 12.33.
Table 12.33. Classification of materials according to their flotation properties (after Eigeles, 1964)
Class Example Applied collectors
Non-metals and solids with signifi- sulfur, graphite, coal, talc hydrocarbons, nonionic liquids
cant natural hydrophobicity insoluble in aqueous
Native metals and sulfides gold, chalcocite chalcopy- xanthates, aerofloats
rite, galena sphalerite
Oxidized minerals of non-ferrous cerusite, smithsonite mala- xanthates (after sulfidization),
metals chite, tenorite, cuprite anionic and cationic
Oxides, hydroxides and silicates hematite, ilmenite anionic and cationic (with and
corundum, cassiterite without activation using metal
chromite, feldspars ions)
kaolinite
Sparingly soluble salts fluorite, barite, calcite, anionic and cationic
apatite, dolomite
Soluble salts halite, silvinite, carnalite, cationic, seldom anionic
kieserite
Minerals of the same class feature similar flotation properties. Thus, for their sepa-
ration the same or similar collectors are used, while minerals of different classes un-
dergo flotation by applying other collecting reagents. These issues are discussed in
details in following chapters 12.6.1-12.6.6.
12.6.1. Naturally hydrophobic substances

Graphite, talc, sulfur and some sulfides have a very specific crystalline structure
which makes them hydrophobic. They form sheets mutually connected with the van
der Waals forces. Crushing and grinding split the particles mostly along the sheets
whose surfaces are naturally hydrophobic. Hydrophobic planes easily adsorb different
reagents, especially non-ionic, hence their ability to float with non-ionic collectors.
Other planes are usually hydrophilic because they contain broken chemical bonds,
which easily adsorb hydrophilic hydroxyl groups. This is the reason why particles of
many naturally hydrophobic substances upon grinding contain also hydrophilic planes.
The latest investigation show that that talc (Michot et al., 1994), graphite Schrader,
1975) and coal (Drzymała, 1999c), known as naturally hydrophobic substances, can be
only slightly hydrophobic or completely hydrophilic when, after grinding, the particles
are prevented from adsorption of air components such as N2, CO2, and pollutants (e.g.
hydrocarbons), on their surface. These minerals become hydrophilic when the pollut-
ing reagents are removed from their surface, for instance by vacuum degassing. Hy-
drophilization of naturally hydrophobic substances can also take place as a result of
oxygen adsorption. For example, sulfur exposed to air becomes less and less hydro-
phobic due to oxygen binding.
The presence of hydrophilic planes, oxidation and hydroxylation make flotation of
naturally hydrophobic materials inefficient. Therefore, either frothers having collect-
ing properties or collectors and frothers are applied to float naturally hydrophobic ma-
terials. Among collectors, usually apolar compounds such as diesel oil and kerosene,
are used. Flotation of naturally hydrophobic materials sometimes requires the use of
special flotation promoters, especially in the case of oxidized coals. Coal is an organic
substance characterized by a complex structure. It is not possible to assign it a deter-
mined chemical formula. It is known that coal contains many organic groups: mostly
aromatic, then aliphatic, as well as certain amount of oxygen compounds and smaller
amount of other elements (S, N). There are very many simplified chemical formulas
for coal and one of them is shown in Fig. 12.49.
H H
H
H OH Fig. 12.49. Simplified model of a coal unit. After

C H2 Krevelen (Czapliński, 1994)
H
H2C CH
CH3
12. Flotation 349
Oxygen present in coal can be found in the form of different functional groups.
Their amount changes with the degree of coalification. Coalification results in removal
of oxygen and increase in coal aromatization. The last stage of coalification is a com-
plete carbonation of coal, i.e. it contains only elemental carbon (graphite) (Fig. 12.50).
32
% of oxygen in oxygen-containg species
28 coal
-OCH3
24 -COOH
20
16 -C=O
12 -OH
4 -C-O-C
and other groups
0
65 75 85 95
carbon content, %
Fig. 12.50. Distribution of oxygen containing groups in coals as a function of their coalification (after
Zyla, 1994; with permission of Ucz. Wyd. Nauk.-Dyd. AGH and the author)
80
70
sessile drop
60
50
40
bubble attachment
30
captive bubble
20 θr
10 flotometry
0
0 2 4 6 8 10 12 14
moisture content , %
Fig. 12.51. Influence of method of measurement on values of contact angles of coals measured by: ses-
sile drop and captive bubble (Gutierrez–Rodriguez et al., 1984),
bubble attachment (*) – Hanning and Rutter (1989), sessile drop aqueous receding contact angle θr) –
Good and Keller (1989), flotometric contact angle (Drzymala, 1999c) (permission of © AA Balkema
Publishers)
The degree of coalification effects a number of coal properties including its hydro-
phobicity (Fig. 12.51). The coal hydrophobicity, measured as contact angle, consid-
erably depends on the method of measurement, which was shown in Fig. 12.52. This
results from coal heterogeneity and its susceptibility to changes in the course of pre-
paratory operations for measurements.
80
60
40 coal
20
0
80 82 84 86 88 90 92 94
carbon content in coal , %
Fig. 12.52. Change of coal hydrophobicity as a function of its coalification

(after Brown, 1962)
+
subb
cok
itum
ant
zeta potential, mV
ing
hra
inou
cite
10
8
6
s
4
2
0
0 2 4 6 8 pH 10
oxid
ized
-
Fig. 12.53. Shape of zeta potential for coal as a function of pH

(after Laskowski and Miller, 1984)
Coal is very easily oxidized by air components. It causes an increase in coal oxy-
gen content, especially in the form of carboxyl groups, which, in turn, causes the shift
12. Flotation 351
of the iep to low pH values. This can be demonstrated, for example, by zeta potential
measurements (Fig. 12.53).
As the degree of coalification increases, the content of the volatile matter and other
components in coal decreases. This could be a base for coal classification. The Polish
classification of coal is presented in Table 12.34. Many countries uses their own clas-
sification. There also exists the so-called international classification (Czapliński,
1994).
Table 12.34. Polish classification of coal (PN-82/6-97002)
Volatile mat-
Coal type Symbol Roga #* Application
ter, %
1 2 3 4 5
Flame 31.1 > 28 < or = 5 Energetic coal for all firebeds
32.1 > 5 to 20
Gas-flame > 28 Energetic coal for: all firebeds; gas production
32.2 21 to 40
Energetic coal for: grate firing; dust firing; indus-
Gas 33 > 28 > 40 to 55
trial furnaces; gas plant; coke
Gas- 34.1
> 28 > 55 For: coke; gas plant; gas-coke plant
-coke 34.2
35.1 > 26 do 31
Orthocoke > 45 For coke
35.2 > 20 do 26
Methacoke 36 > 14 do 20 > 45 For coke
37.1 > 20 do 28 For: special coke mixtures; energy by plants with
Semicoke > or = 5
37.2 > 14 do 20 special firebeds; smokeless fuel
Lean 38 > 14 do 28 <5 As coal 37
Anthracitic 41 > 10 do 14 - As coal 37
Anthracite 42 ≥ 3 do 10 - For energy by plant with special firebeds
Metaanthracite 43 <3 -
Graphite
*ability of sintering
Typical values of moisture, ash, chemical composition and density of different coal
types are presented in Table 12.35.
Table 12.35. Properties of coals (according to Polish classification) on dry and ash-free basis (after Czap-
linski, 1994)
Properties–coal type 31 32 33 34 35 37 38 41 42
Moisture, % 9.10 7.45 1.88 2.48 1.15 0.94 0.93 0.98 1.4
Ash, % 4.97 1.24 5.74 5.11 7.89 3.95 5.91 3.69 4.51
Volatile, % 35.29 29.81 26.99 32.36 23.68 16.84 12.67 6.72 5.39
C content, % 66.34 73.58 79.62 79.10 80.19 86.63 85.16 88.59 87.48
H content, % 4.30 4.16 4.66 5.17 4.90 4.51 4.07 3.85 3.87
O+N content, % 13.4 13.19 7.57 7.62 5.31 3.56 2.95 2.17 2.37
Organic S content, % 1.82 0.38 0.53 0.52 0.56 0.77 0.98 0.72 0.73
Real density, g/cm3 1.48 1.53 1.38 1.35 1.33 1.33 1.41 1.45 1.44
Apparent density, g/cm3 1.21 1.29 1.34 1.27 1.30 1.20 1.33 1.38 1.38
Porosity, % 18.06 14.19 2.69 4.54 1.69 11.96 5.72 4.82 4.54
The location of hard coals in relation to other carbon-containing substances taking

into account carbon, hydrogen, and oxygen content is shown in Fig. 12.54.
0 100%
20 80
en
og
40 60
dr
hy
ox
yg
60 40
en
cellulose
80 brown 20
hard coal
lignin
antrac ite
100% graphite
0 20 40 60 80 100%
carbon
Fig. 12.54. Place of coal among other carbonaceous matters (after Wert and Weller, 1982, with permis-
sion of the American Institute of Physics)
Flotation aims at separation of combustible carbonaceous matter from minerals re-

sponsible for the formation of ash. Coal contains from a few to tens of percent of ash.
Flotation usually involves coal fines, which are difficult to upgrade by other methods.
Since coal is naturally hydrophobic, non-polar liquids are used for its flotation. They
are mainly products of coal and crude oil distillation. Most often crude oil is used,
which is a mixture of hydrocarbons, mostly aliphatic. According to Sablik (1998)
post-cumene tar also called the FK reagent can be also used for flotation. It consists of
di- and tri-isopropylbenzene, isobuthylbenzene, isopropylindene, diphenylmethane
and diphenylpropane. Another collector capable of coal flotation is Pyrolysat (BF), a
mixture of high-boiling aliphatic and aromatic hydrocarbons. A good flotation can
also be achieved using dibutylmaleate (MDB) or diocthylmaleate (MDO). In coal flo-
tation numerous different substances are used as frothers. They are often waste prod-
ucts coming from local producers, being either pure substances or waste mixtures.
Therefore, their chemical composition and commercial names vary considerably and
they often change. Generally, pure higher aliphatic alcohols (AC) contaminated with
aldehydes, ketones, and esters are used as frothers. They are available on the market in
the form of such reagents as Alifal N, Flotanol, Mekol, Reagent HO, ect. Earlier pine
oil and products containing phenol were used. Industrially used methyl isobutyl carbi-
nol is a good frother and it is considered as standard in laboratory investigation. Ter-
penes and polyethylene glycol derivatives can also be used as frothers. In coal flota-
tion various modifying substances are also applied. They are similar, or the same, as
12. Flotation 353
those used in mineral ores flotation. pH regulators (NaOH, H2SO4), stability regulators
(aqueous-glass, CaCl2), organic colloids (starch, dextrin, tannin) belong to this group.
Some coals do not float well and require special reagents called promoters (Sablik,
1998). The well-known cationic collectors (amines, pyridine salts), anionic collectors
(sulfonates, sulfates), non-ionic (polyglicolic ethers) can be used as promoters but
should be applied in small quantities. Usually they are combined with diesel oil. Ac-
cording to Sablik (1998) activity of coal flotation promoters depends on their adsorp-
tion on oxidized coal sites, which provides the sites available for apolar collector.
Promoters also cause higher degree of emulsification of diesel oil drops.
Coal flotation takes place due to adsorption of apolar oil at non-polar coal sites.
The sorption is physical in character and is caused by the van der Waals forces. There-
fore, the best flotation can be observed when there are many non-polar sites, e.g. near
iep (Fig. 12.39). It results form the research of Yarar and Leja (1982) that non-
oxidized coals float well, while their zeta potential does not considerably effect flota-
tion, although it is possible to see the maximum flotation in the vicinity of neutral pH.
Oxidized coals, however, float the worse the lower their zeta potential is. Adsorption
of non-polar collector, and thus flotation of coals, is the better the more hydrophobic
coal is. As it was shown in Fig. 12.52, the hydrophobicity and coal flotation increases
with carbon content up to about 88%, and then it slightly decreases as result of higher
number of aromatic groups in coal.
In the presence of simple inorganic salts (NaCl, KCl, etc.), one can notice an in-
crease in coal flotation (Klassen and Mokrousov, 1963; Laskowski, 1966). To observe
improved flotation, the salt concentration has to be higher than about 0.1 mol/dm3.
Otherwise flotation is reduced (Li and Somasundaran, 1993). The mechanism of salt
flotation, in spite of the attempts by many authors (Klassen and Mokrousov, 1963;
Laskowski, 1966; Paulson and Pugh, 1996; Pugh et al., 1997), has not been well ex-
plained. Salts does not seem to change the hydrophobicity of the floated material
(Laskowki, 1986) but mainly make flotation faster (Yoon, 1982) and shorten the so-
called contact time. It can be supposed that the increase in density of aqueous solution
also plays a certain role. Salt flotation is not commonly applied because of corrosivity
of saline aqueous solutions (Ozbayoglu, 1987).
Sulfur, talc, and graphite also belong to the group of naturally hydrophobic sub-
stances. For flotation of sulfur, collectors are used along with frothers. Any apolar
compound can serve as a collector. In Poland, at the beginning of sulfur industry de-
velopment in the sixties of the previous century, kerosene was used as the collector
and turpentine as a frother while water-glass was applied as gangue minerals depres-
sor. Since 1970’s diesel oil was used and turpentine replaced by mixture of pure alco-
hols (AC) or byproducts containing AC (Flotanal) (Bara and Kisielewski, 1984). Be-
cause of the increased sulfur production from other sources (natural gas, crude oil, ex-
haust gases desulfurization, coking plants, bituminous shale, underground melting)
sulfur ore upgrading has no significance nowadays (Bilans, 1997).
1.8
pH=5.0-6.5
adsorption density, mg/m 2 1.5
coal
molibdenite
1.2 talc
0.9
coal pyrite
0.6
0.3 pyrite
quartz
0
0 20 40 60 80 100 120 140 160 180
equilibrium concentration of dextrin, mg/dm 3
Fig. 12.55. Strong adsorption of dextrin on hydrophobic minerals and coal and weak adsorption on hy-
drophilic quartz and pyrite (Miller et al., 1984) (Overseas Publishers Association. With permission of
Tylor & Francis Ltd.)
Other naturally hydrophobic minerals are talc and graphite. Talc is a silicate (Mg3
[(OH)2Si4O10]) and graphite is one of many allotropic varieties of carbon. Presently six
allotropic groups of this element are known: graphite (α and β), diamond (diamond
and lonsdaleite), carbine (chaoite and cumulene), fullerene (fullerite and nanotubes),
carbon (linear and cyclocarbon) (Pierzak and Drzymała, 1998).
Graphite, talc and bituminous matter accompany some ores and raw materials and
their easy flotation is an undesirable phenomenon. Additionally, they can alter froth
property causing its stabilization. Dextrin is commonly used to depress naturally hy-
drophobic materials (Laskowski, 1983). Figure 12.55. presents that a considerable ad-
sorption of dextrin on different naturally hydrophobic minerals and negligible adsorp-
tion on hydrophilic minerals such as quartz and pyrite occurs. This phenomenon is
used for depression of naturally hydrophobic materials.
12.6.2. Native metals and sulfides

Such metals as iron, mercury, copper, gold, and platinum group can be found as na-
tive materials (Bolewski and Manecki, 1993). Their flotation is conducted using col-
lector with sulfydryl groups, especially xanthates. To float precious and non-ferrous
metals it is necessary to use higher concentrations of sulphydryl collector or collectors
containing higher number of CH2 groups in the hydrocarbon radicals, i.e. five or more.
Dithiophosphates as well as xanthate + mercaptobenzothiazole, and dithiophosphate+
mercaptobenzothiazole mixtures (Aplan and Chander, 1988) can be used for flotation.
12. Flotation 355
The hydrophobization mechanism of metal surfaces with sulphydryl collectors is simi-

lar to that of sulfides and is of an electrochemical nature. Good flotation of metallic
silver, gold, and copper is connected with a high stability constant of metal ions with
sulphydryl flotation regents (Fig. 12.56). Metals, not only the native ones, float well
with fatty acids and amines (Wark, 1955). Some of them, however, especially Al and
Pb, provide small contact angles with amines. Metal sulfides can be floated with dif-
ferent collectors, yet the key role is played by sulphydryl compounds and their alka-
line salts (–SNa or –SK). The list of collectors containing sulfur in their structure is
presented in Table 12.36 (Aplan and Chander, 1988).
Me +
-5 ++
Me
log (solubility product)
-15 Cd
Ag
Zn
Cu
Ni
-25
Pb
Au
-35 Cu
Hg
-45
0 2 4 6 8 10 12 14
number of carbon atoms in xanthate
Fig. 12.56. Solubility products of metal xanthatates

(after Aplan and Chander, 1988)
Among sulphydryl collectors the most important are dithiocarbonates (xanthates)

and dithiophosphates (aerofloates). Their strong collective activity result from their
good reactivity with metal ions, which is expressed by a low solubility product. Figure
12.56 shows the solubility products of mono– and divalent metals with xanthates hav-
ing different number of carbon atoms in the hydrocarbon radical. Only metal ions
which form sulfides of industrial significance were considered.
It results from Fig. 12.56 that xanthates form the most stable compounds with gold,
mercury and copper, while the most soluble is the one with Zn2+. Although the con-
stants represent the solubility product for bulk phases not for xanthate-ion compounds
on the sulfide surface, there can be observed considerable similarity between metal
xanthates solubility and sulfide flotation.
Table 12.36. Collectors containing sulfur applied for flotation of sulfides (after Aplan i Chander, 1988)
Collector type Formula Chemical name Manufacturer and des-
ignation
Mercaptan R–SH Pennwalt, Philips
Dithiocarbonate AmCy Dow
(xanthate) R–O–(C=S)–SK potassium ethyl 303 Z–3
R–O–(C=S)–SNa sodium ethyl 325 Z–4
potassium isopropyl 322 Z–9
sodium isopropyl 343 Z–11
potassium butyl – Z–7
sodium isobutyl 317 Z–14
potassium sec-butyl – Z–8
sodium sec-butyl 301 Z–12
potassium amyl 355 –
sodium amyl 350 Z–6
potassium sec-amyl – Z–5
potassium hexyl – Z–10
Trithiocarbonate R–S–(C=S)–SNa Philips (Orform C0800)
Xanthogen formate R–O–(C=S)–S–(C=O)–OR´ Dow Minerec
R=ethyl, R´=ethyl Z–1 A
R=izopropyl, R´=ethyl – 2048
R=butyl, R´=ethyl – B
Xanthic ester R–O–(C=S)–S–R’ AmCy Minerec
R=amyl, R´=allyl 3302 1750
R=heksyl, R´=allyl 3461 2023
Monothiophosphate (R–O–)2(P=S)–ONa Amcy 194, 3394
Dithiophosphate (R–O–)2(P=S)–SNa AmCy (Aerofloat)
sodium diethyl Na Aerofloat
sodium di-isopropyl Aerofloat 211, 243
sodium di-izobutyl Aerofloat 3477
sodium di-isoamyl Aerofloat 3501
sodium di-iso-sec-butyl Aerofloat 238
sodium di-methylamyl Aerofloat 249
(R–O–)2(P=S)–SH cresylic acid+P2S5 Aerofloat 15
Dithiophosphinate (R–)2(P=S)–S–Na AmCy3418
Thiocarbamate R–(NH)–(C=S)–OR´ Dow Minerec
N-methyl-O-isopropyl – 1703
N-methyl-O-butyl – 1331
N-methyl-O-isobutyl – 1846
N-ethyl-O-isopropyl Z–200 1661
N-ethyl-O-isobutyl – 1669
Thiourea derivatives (C6H5NH2)C=S AmCy Aero. 130
(thiocarbanilide)
Mercaptobenzo- AmCy 400 series
thiazole
Flotation of sulfides takes place as a result of collectors adsorption. In the case of

xanthate flotation of sulfides, hydrophobization mechanism is complex and is not well
12. Flotation 357
understood. This is so because there are many reactions which can create hydrophobi-
zation of a sulfide in the presence of xanthates. According to Woods (1988) hydro-
phobization of sulfides with sulphydryl compounds, especially with xanthates, results
from electrochemical reactions in which electrons are transmitted from a collector to a
sulfide mineral (anodic process), and then the electrons return to aqueous solution due
to catodic reduction of oxygen:
O2 + 2H2O + 4e = 4OH- (12.90)
During anodic oxidation, the following products can be formed:
a) chemisorbed xanthate Xad created from X- ion coming from the aqueous solution
and a metal ion sitting in the crystalline structure of sulfide:
X– → Xad + e (12.91)
-
b) dixanthogene X2, as a result of X ion oxidation
2X– → X2 + 2e (12.92)
c) metal xanthate MeX2, due to the reaction of X- ion with metal sulfide MS
2X– + MS → MX2 + S + 2e, (12.93)
and the product of this reaction is element sulfur S, though other products can also be
formed, e.g. thiosulfate, sulfate(IV) or sulfate(VI). The formation of sulfate(VI) can be
also written as:
2X– + MS + 4H2O → MX2 + SO 24 − + 8H+ + 8e (12.94)
The chemisorbed xanthate Xad can be transformed into either a dithiol:
2Xad → X2 or Xad + X– → X2 + e (12.95)
or metal xanthate:
2Xad + M2+ + 2e → MX2 (12.96)
Metal xanthates MX2 can also be formed on partly oxidized sulfide surface. Sur-
face oxidation can be written as the reaction:
MS + 6H2O → M(OH)2 + SO 24 − + 10H+ + 8e (12.97)
while the formation of metal xanthate as:
M(OH)2 + 2X– → MX2 + 2OH– (12.98)
According to Leja (1982), after introduction of xanthate into flotation pulp many
other compounds can be formed, for instance xanthogenic acid HX, hydroxyxanthates,
perxanthates, disulfide carbonates, etc.
The redox reactions taking place in flotation systems can be shown in a Eh –pH
diagram. However, their complexity for metal sulfides in the presence of xanthates
and oxygen can be significant due to numerous possible reactions in the system.
Therefore, the diagram is usually based on certain assumptions, e.g. the number of
considered components and their oxidation degree are limited. Well-constructed dia-
grams should reflect real conditions in the flotation system and allow to predict quan-
titatively the presence of different species in the system, especially those which are
responsible for flotation. Usually the diagrams provide quantitative information be-
cause they are based on the assumption that all predicted reactions would take place
without any difficulties within a short period of time, which may not be true.
Figure 12.57 shows a Eh–pH diagram for PbS in the presence of potassium ethyl
xanthate (X-) plotted on the basis of redox equations for this system applying the pro-
cedure already used for the Cu-H2O system (Fig. 12.57). When drawing the diagram
an assumption was made that the highest oxidation form of sulfur is elementary sulfur
(So). A comparison of real data on flotation for PbS xanthate flotation at varying Eh
values and pH=8 (Fig. 12.58) with the areas of predomination of different species pre-
sent in the flotation system indicates that PbX2 is responsible for a rapid PbS flotation,
while the role of dixanthate (X2) is considerably less important. An interpretation of
PbS flotation when the sulfide was kept under oxidizing conditions before flotation, is
much more difficult since it requires drawing new Eh – pH diagrams at different as-
sumptions.
0.5
Pb(OH)2 + S + X2
0.4 -
HPbO2 + S + X2
0.3
0.2 PbX 2 + S
E h, mV
-
HPbO2- + S + X
0.1
0 -
PbS + X
-0.1
-0.2
galena
-0.3
6 8 10 12 14
pH
Fig. 12.57. Eh–pH diagram for galena in the presence of ethyl xanthate assuming
that total amount of xanthate species was 10–4 M and that elemental sulfur is created during oxidation
(after Woods, 1988)
12. Flotation 359
100
recovery during first minut of flotation, %

a
80
60
40
b
20
galena
0
-0.5 -0.3 -0.1 0.1 0.3 0.5
E Pt, mV
Fig. 12.58. Galena flotation with ethyl xanthate at pH = 8 as a function of applied potential to a platinum
electrode in solution: a – galena kept in oxidizing environment before flotation, b – kept in reducing envi-
ronment (after Richardson, 1995; based on Guy and Trahar, 1985)
Xanthates easily render sulfides hydrophobic. Even a low concentration of xanthate

in the solution causes flotation of pyrite (Fig. 12.59). Flotation usually takes place
within a wide range of pH. Usually there is a maximum pH value, also called the criti-
cal pH (Lekki, 1979) at which flotation disappears. The critical pH depends on the
concentration and structure of the collector. Only flotation time, according to Las-
kowski (1986), who based his claim on the data by Lekki (1980), does not effect the
critical pH of flotation.
100
-4
2x10 M KEtX
80
recovery, %
60
-5
10 M KEtX
40
20
pyrite
0
0 2 4 6 8 10 12
pH
Fig. 12.59. Pyrite flotation in the presence of different concentration of potassium ethyl xanthate
(after Fuerstenau et al., 1968)
Sulfide flotation usually takes place at relatively small adsorption of sulphydryl

(thiol) collectors. When partly oxidized sulfides are subjected to flotation, xanthate
consumption is higher, since some part of it is used for removing the surface oxidized
layers by xanthate ions, while the remaining part makes the new surface hydrophobic.
The calculated apparent total adsorption of xanthate is equivalent to several or a dozen
of statistical monolayers. The flotation improves when longer chain xanthates are used
in experiments. Since hydrophobization with xanthates is of electrochemical character,
the iep of sulfide does not effects flotation. The isoelectric point on the pH scale, de-
termined from the electrokinetic potential for sulfides is about 5 for ZnS and about 9
for CuS (Lekki, 1979). For oxidized sulfides the iep falls to about 2, i.e. takes the
value of elementary sulfur (Ney, 1973).
Sulfides can be also floated with other collectors including oleates (Abramov et al.,
1982) and amines. As far as amines are concerned, flotation takes place at low pH val-
ues after sulfide ions addition (Aplan et al., 1980).
Application of xanthates requires the use of frothers. Different reagents including
simple aliphatic alcohols (AD), branched alcohols (MIBC), complex alcohols (pine
oil, α-terpineol), aromatic alcohols (crezoles), polyglycols (Arefroth 65) and Dow-
forth, as well as trietoxybuthane (TEB) can be used as frothers.
Many papers discuss natural flotation of sulfides. It is known that sulfides having
layered (MoS2, molybdenite; As2S3 (auripigment), molecular (As4S4, realgar) or chain
(Sb2S3) structures are naturally hydrophobic (Gaudin et al., 1957; Drzymala, 1994c).
Other sulfides which have ionic-covalent-metallic structure, depending on their geo-
logical history and the way of their preparation before the measurement, are either hy-
drophilic or hydrophobic. They are characterized by different contact angles, ranging
from very small to that of elementary sulfur. This phenomenon results from formation,
due to oxidation, surface layers containing excess of sulfur atom in relation to metal
ions, or even elemental sulfur. A further oxidation can lead to hydrophilization of sul-
fide surface. Metal sulfides from Polish copper deposits are hydrophilic or slightly hy-
drophobic and their flotometric contact angle is presented in Table 12.37.
Table 12.37. Natural hydrophobicity of sulfides and other minerals mostly from Polish copper deposits
in Lubin (LGOM). The hydrophobicity is given in term of contact angle calculated from flotometric data.
After Drzymala and Bigosinski (1995), Drzymala (1994) as well as Lekki and Drzymala (1990)
Sulfide Contact angle Sulfide Contact angle
Sulfur 63.2 *covelline 1.9
Pyrite 44.0–0 *djurleite 0
*Bornite 6.5–9.6 *chalcocite 0
*Djurleite – bornite 4.5 *bituminous shale 0
*Galena 4.0 calcite 0
*Chalcopyrite 3.6 quatz 0
* denotes mineral from LGOM.
12. Flotation 361
12.6.3. Oxidized non-ferrous metals minerals

Flotation of this group of minerals can be performed in two different ways. The
first approach is designed for sulfides and relies on using thiol collectors after treating
the oxidized sulfide surface with sulfide ions. The second path is for silicates, oxides
and hydroxides which are floated with either cationic or anion collectors. Lead: cerru-
site (PbCO3), vanadinite (Pb5[Cl(VO4)3]), anglesite (PbSO4), copper: malachite
(CuCO3Cu(OH)2), azurite (2CuCO2Cu(OH)2), chrysocola (hydrated cuprum silicate),
tenorite (CuO), cuprite (Cu2O), and zinc: smitsonite (ZnCO3) minerals are suitable for
flotation with thiol after sulfidization. Sulfidization of some minerals (cerusite, mala-
chite, azurite) is easily, while with others (anglesite, tenorite, cuprite) experiences dif-
ficulty (Leja, 1982).
Sulfidization of mineral relies on treating them with aqueous solutions containing
sulfide ions, which react with surface mineral ions, forming metal sulfides and ele-
mentary sulfur on the surface. The process, for instance for an oxide, can be written
as:
-MO + S2- + 2H+ = -MS + H2O (12.99)
where –MO is an oxide surface group which becomes metal sulfide –MS under the
influence of sulfide ion S2-.
The driving force of this reaction is lesser solubility of metal sulfides in relation to
metal oxides. Metal sulfides easily undergo decomposition in the presence of oxygen,
forming such oxidized sulfur compounds as thiosulfates, sulfates(IV) or sulfates(VI).
Yet, it should be noticed that sulfidization is conducted as a surface reaction. Sulfidi-
zation is rather a complicated process.
Sulfide ions regulate redox potential of flotation and they create so low Eh that
thiol collector adsorption (e.g. xanthate) does not take place. Therefore, flotation of
sulfidized minerals generally does not occur in the presence of sulfide ions
(Fig. 12.60, line 3). The best way to create proper Eh is removing sulfide ions from the
solution after sulfidization using clean aqueous instead of sulfide solution (Fig. 12.60,
line 1) or oxidation of sulfide ions by passing through the solution air (Fig. 12.60,
line 2), oxygen, or other oxidizers. However, oxidizing can cause oxidation and the
removal of surface sulfide layers which is responsible for flotation.
Some oxidized minerals float with thiols directly without sulfidization, but it re-
quires considerably higher xanthate concentration. According to Fleming (1952) and
Leja (1982) vanadinite floats with xanthates at 1 to 2 kg of collector per 1 megagram
of the ore, but it requires only 0.1 kg/Mg when sulfide minerals are sulfidized. In the
case of chrysocolla, roasting at 773-873 K significantly improves its flotation (Las-
kowski et al., 1985).
100
80
1
2
recovery, %
60
malachite
40
20
3
0
0 10 20 30 40
dosage of amyl xanthate, mg/dm 3
Fig. 12.60. Influence of conditions of flotation on recovery of malachite sulfidized with 960 mg/dm3 of
Na2S·9H2O in the presence of frother (amyl alcohol 60 mg/l): 1 – flotation when after sulfidization the
solution is replaced with pure aqueous, 2 – flotation after 25 minutes of air bubbling through the solution
containing sulfide ions, 3 – flotation directly after sulfidization in the presence of sulfide ions (after Soto
and Laskowski, 1973)
The use of anionic and cationic collectors for flotation of oxidized minerals of non-
ferrous metals was discussed in the section on oxide and hydroxide flotation.
12.6.4. Oxides and hydroxides

This is the most numerous and diversified group of minerals. It includes simple
oxides (Fe2O3, SnO2), hydroxioxides (AlOOH), as well as complex oxides and hy-
droxides (spinels, silicates, aluminosilicates). Their flotation can be conducted with
different collectors. The selection of a collector depends on mineral properties, includ-
ing isoelectric point and the ability of a cation which forms the mineral to chemically
react with collectors. Detailed discussion of this mineral group would require a sepa-
rate description of each system and it would be too lengthy. Therefore, only selected
flotation systems will be discussed here and application of fatty acids and amines as
collectors. Other aspects of flotation of minerals belonging to this group have already
been discussed in chapters on flotation reagents (activators, depressants, collectors).
Fatty acids and their aqueous-soluble potassium salts, called soaps, are eagerly
used as collectors for oxides, hydroxides, and salts. The most popular is oleic acid
which is a cheap by-product the wood industry and is called the tall oil. The tall oil
contains oleic acid and such acids as linoleic, linolenic, and abietic. It is known that
pH range of flotation of many minerals with fatty aids is similar (Polkin and Najfonov,
12. Flotation 363
1965); Mishra, 1988) (Fig. 12.61). Yet, their mechanisms of adsorption and flotation
can be quite different. It results from Fig. 12.61 that some minerals do not float well or
do not float at all as quartz (Fig.12.61) or silicates (Table 12.38). This diversity has
become the basis of separation of minerals with the use of fatty acids. It can also be
concluded from Fig. 12.61 that oleate flotation ceases at extreme pH values. This fact
results from a competition of OH- and H+ ions. This has already been discussed in
chapter 12.5.4.1.
100
80
varous minerals
recovery, %
60
40
albite
20
quartz
0
2 4 6 8 10 12
pH
Fig. 12.61. Flotation of many minerals in the presence of 1 kg/Mg NaOl depending on the pH of the solu-
tion. Quartz and some silicates do not float under such conditions. Investigated minerals: columbite, zir-
con, tantalite, ilmenite, rutile, granate, perovskite (based on data of Polkin and Najfonov (1965) supple-
ments with the line of quartz)
Table 12.38. Influence of structure of silicates on their flotation with anionic and cationic collectors (after
Manser, 1975)
Silicate group
Collector
orthosilicates pyroxene amphibole frame
Anionic good week None none
Cationic satisfactory* satisfactory * Good very good
* Flotation depends on pH.
Fatty acids form different species in aqueous solution shown in Fig. 12.62. There
are three groups of oleate species: ions (Ol-, HOl2- premicelle), micelles and such
phases as liquid oleic acid and acid soap (NaHOl2). All these forms play an important
role in flotation process. It should be also noticed that sodium oleate is present in most
oleate forms including HOl oil phase, ionic species (acid soap ion) and micelles. Since
the structure of minerals and oleate species are diverse, the mechanisms of fatty acids
adsorption on mineral surface is highly complex.
Fig. 12.62. Concentration - pH diagram for sodium oleate aqueous solutions. It shows predominance of
various oleate species (Drzymala, 1990): c – activity of oleate species, mol/dm3, B (or β) – degree of
binding oleate with sodium ions in associated species (number of sodium ions per one oleate ion in the
associate) (data for drawing the diagram are given in Table 12.39. It should be noticed that sodium ions
also take part in most oleate species)
There are many hypothesis regarding adsorption of oleate species on mineral sur-
faces. Two of them seem to be the most realistic. The firs hypothesis, proposed by
Lekki (1984) and Laskowski (1981) assumes that flotation is caused by oleate ions
which form with mineral-building cations non-soluble soaps. According to such an
assumption flotation pH range can be calculated on the basis of oleate soaps solubility
product, as flotation takes place when under particular conditions of pH and collector
concentrations in the system an oleate soap is formed. The argument of this hypothesis
is a good agreement of flotation data with the pH range of soap formation. This is
shown in Table 12.40.
According to the second hypothesis, proposed by Fuerstenau and Palmer (1976),
adsorption of oleate ions occurs when cations are able to form monohydroxy com-
plexes. It is not important whether the ions originate form mineral crystalline structure
or are purposefully introduced into the solution. It has been noticed that flotation of
quartz activated by metal ions introduced into the solution is similar to flotation of
minerals containing the same metal ion in their crystalline structure. Table 12.41 pre-
12. Flotation 365
sents calculated ranges of metal monohydroxycomplexes formation and observed flo-

tation pH ranges of different minerals. The fact that there is an agreement between flo-
tation and pH ranges of bulk soap formation suggest that both hypothesis can be com-
bined. They become compatible providing that flotation takes place due to forming
soaps with hydrolized metal ions present near the mineral surface (Fuerstenau, 1995).
These combined hypothesis can be called the activation hypothesis of flotation with
fatty acids.
Table 12.39. Reactions and corresponding equilibrium constants of reactions taking place in oleate solu-
tion. Activity coefficient of NaOl and HOl in the oily phase and activities of NaHOl2 as well as micellar
forms (Naβ Olβ–1)n were assumed to be 1 (Drzymala, 1990)
Nr Reaction reaction constant pK
1 HOlaq ⇔ H+aq + Ol–aq K1 = (H+aq)·(Ol–aq)/(HOlaq) 4.95
2 HOlaq ⇔ H+aq + Ol–aq K2 = (H+aq)·(Ol–aq)/XHOl oil 12.3
3 HOloil ⇔ HOlaq K3 = (HOlaq)/XHOl oil 7.35
4 HOlaq + Ol–aq ⇔ HOl2–aq K4 = (HOl2– aq)/[(Ol–aq)·(HOlaq)] –7.6
5 2Ol–aq ⇔ Ol22–aq K5 = (Ol22– aq)/(Ol–aq)2 –4.0
6 HOloil + Na+aq⇔ NaOlw HOl + H+aq K6 = (H+aq)·XNaOl w HOl/[(Na+ aq)·XHOl oil ] 6.38
7 NaOlIn HOl ⇔ Na aq + Ol aq
+ –
K7 = (Na+aq)·(Ol–aq)/XNaOl in HOl 5.92
8 NaHOl2 ⇔ Na+aq + H+aq + 2Ol–aq K8 = (Na+aq)·(H+aq)·(Ol–aq)2 19.01
9 2HOloil + Na+aq ⇔ NaHOl2 + H+aq K9 = (H+aq)/[(Na+aq)·(XHOl = 0,8)2 5.59
10 0,5 NaHOl2 + (β – 0,5)Na+aq ⇔ K10 = (H+aq)0,5/(Na+aq)β–0,5, dla 10–3M Na+ 4.85
NaβOlβ–1 + 0,5 H+aq for 10–2M Na+ 4.58
for 10–1M Na+ 4.44
11 βNa+aq + Ol–aq ⇔ (NaβOlβ–1)n Kβ = 1/[(Na+aq)β·(Ol–aq)] or
log [Oltotal] = A – β log (Na+aq) : CMC β = 0,5 A = –4.4
premicellar form β = 0,4 A = –5.0
cylindrical micelles β = 0,8 A = –1.8
12 –HOls ⇔ –Ol–s + H+aq Kinta2 = {–Ol–s}(H+aq) exp(–Fψ0/RT )/{–HOls} 4.40
13 –HOls + Na+aq ⇔ –Ol–Na+s + H+aq KintNa = {–Ol–Na+s}(H+aq) exp(–Fψ0 /RT)/ 3.3
/[{–HOls}(Na+aq) exp(–Fψb /RT)] ±0.3
14 iep pH=2.0
ψ0– surface potential,
ψb – surface potential at the Na+ ions binding plane
{} stands for activity of surface species,
parenthesis ( ) denote activity of other species.
Table 12.40. Comparison of calculated pH range of existence of soaps with the range of pH of different
minerals according to a hypothesis of Lekki (1984) (data collected by Drzymala, 1986)
pH range of soap exis- pH range of flotation
Soap Mineral
tencea with 10–5M NaOl
FeOl3 2.57–8.01 Magnetite 3.0–8.0b
FeOl3 2.57–8.01 Ilmenite 3.7–8.0c
FeOl3 2.57–8.01 Hercynite 3.9–8.0c
AlOl3 3.98–8.22 Pleonaste 3.6–7.9c
MgOl2 7.92–10.37 Pleonaste 8.7–10.4c
a) Drzymala (1986), experimental data: b) Luszczkiewicz et al. (1979), c) Lekki (1984).
Table 12.41. Comparison of pH ranges of oleate flotation of minerals as well as activated quartz and pH
of existence of metal monohydroxy complexes
Flotation (after a) pH of flotation b
Monohydroxy Range of pH pH of maxi- mineral pH of activated quartz
complex at concentration> mum concen- maximum
10–7 M tration flotation
++
FeOH 0–3.9 2.7 augite 2.9 2–8*
AlOH+ 2.1–5.9 4.3 2–8
PbOH+ 3.2–12.4 8.7
MnOH+ 7.6–11.6 9.5 pirolusite 9
MgOH+ 8.4–12.5 10.5 magnesite 10.4 7–13
CaOH+ > 8.5 13.1 Augite 11 7–13
CuOH+ 5.1–8.1 6.5
FeOH+ 4.5–12.1 8.7 chromite and 8.7, ∼8
other iron min-
erals
a – Fuerstenau and Palmer (1976), b – Daellenbach and Tiemann (1964).
* The participation of FeOH+ ions in widening the pH range of flotation of activated quartz activated
with FeOH++ ions cannot be ruled out.
A hypothetical mode of fatty acid adsorption at mineral surface via hydrolized

metal cation is shown in Fig. 12.63.
As it results from the oleate forms predominance diagram (Fig. 12.62), the compo-
sition, apart from ions, involves non-soluble forms of fatty acids, namely the oil or
solid phases. Theoretically, they can attach themselves to mineral surface by hetero-
coagulation, contact, and semi-contact (Drzymala, 1986).
Fig. 12.63. Schematic illustration of adsorption of

fatty acid ion on mineral surface through hydrol-
Me ized metal cation leading to surface hydrophobiza-
OH tion
Adhesion by heterocoagulation (Fig. 12.64a) takes place as a result of the van der
Waals-London attraction forces between oil or solid phase and solid surface, like in
the DLVO theory. The mechanism of this sorption is similar to heterocoagulation.
Since electrical double layers between two attracting objects are partly preserved, such
attachment can be called a non-contact adsorption, analogically to the nomenclature
used by Derjagin (1989) for adhesion of air bubbles to a solid surface. The heteroco-
agulation adhesion takes place when interacting objects have opposite electrical
charge. If the objects are of the same charge, adhesion is determined by the van der
12. Flotation 367
Waals-London attractive forces. Such adhesion can be described using the DLVO the-
ory. An example of flotation resulting from the heterocoagulation adhesion is oleate
flotation of rutile (Purcell and Sun 1963) up to pH about 4,5.
The contact adhesion (Fig. 12.64 b) relies on direct adhesion of oil phase drops to a
solid surface. Adhesion force is determined by the force balance determined by the
Young equation (12.1), which occur at the three phase boundaries, i.e. aqueous-oil,
aqueous-solid, and oil-solid. In turn, the force of interaction of a particular interface
depends on dispersion and the non-dispersion forces (polar, chemical, hydrogen
bonds and others), which is described by the Fowkes relation (12.4). Flotation at low
pH values of such hydrophobic forces as germanium (Drzymała el al., 1987) or mer-
cury (Volke et al., 1984) can serve as an example of contact adhesion.
The semi-contact adhesion consists of adsorption of simple collector ions on the
surface and formation of hemicelles or even monolayers which form a bridge for addi-
tional adsorption of oily drop or particles of collectors as a result of hydrophobic in-
teraction. Figure 12.64c. presents this kind of sorption. The adhesion is characterized
by adsorption isotherms. Flotation resulting from the semi-contact adsorption can take
place at low pH values when non-soluble collector phases predominate in the solution
and hydrophobization of particle surface is due to adsorption of single soluble collec-
tor species. Many minerals can float according to this mechanism.
particle
oil
x
a b c
Fig. 12.64. Schematic illustration of modes of adhesion of a colloidal collector (here as an oil drop) to
solid surface: a – contactless (heterocoagulation), b – contact, c – semicontact adhesion
Investigations showed that at higher concentrations of fatty acids in the solution the
co-adsorption of different species is very common. Coadsorption of oleate ions and
oleic acid drops can take place in pH ranges when drops of oleic acid are present in the
solution that is in acid or neutral solutions (Leja, 1982). This has already been dis-
cussed and shown in Fig. 12.64. Since oleate species are often accompanied by so-
dium oleate (NaOl), they are frequently present in adsorption layers on minerals. In
the case of coadsorption of ionic and micelle species, flotation may be reduced to zero.
This phenomenon can be observed at high oleate concentrations and is caused by ad-
sorption of hydrophilic micelles. This can be illustrated by the results of zircon flota-
tion shown in Fig. 12.65.
100 5
adsorption density, mol/cm 2 x 10 4

80 4
recovery, %
60 3
40 2
zircon
20 1
0 0
0 2 4 6 8 10 12
initial concentration of sodium oleate, mol/dm 3 (x10 4 )
Fig. 12.65. At high oleate species concentrations flotation decreases even though the oleate adsorption
increases. It is assumed that it results from adsorption of hydrophilic micelles (based on data of Dixit and
Biswas, 1973)
100
oleic
80
recovery, %
linoleic
60
linolenic
lauric
40
20 iep 6.9 kyanite
0
0 2 4 6 8 10 12 14
pH
Fig. 12.66. Kyanite flotation with 10–4 kmol/m3 of fatty acids (Choi and Oh, 1965). Applied acids: laurate
(C11H23COOH), linoleic (C5H11–CH=CH–CH2–CH=CH–(CH2)7COOH), linolenic CH3–[CH2–
CH=CH]3(CH2)7COOH and oleic (C17H33COOH)
Oxides and hydroxides flotation depends on the length of hydrocarbon chain of

fatty acid applied in flotation, as well as on number of double bonds in the molecule.
Fig. 12.66 presents kyanite flotation with different fatty acids. The figure shows that
the longer the chain the wider pH range of flotation, which is connected with greater
stability constant of created soap. On the other hand, the presence of double bonds at
the same number of carbon atoms in hydrocarbon radicals, worsens flotation. It can be
12. Flotation 369
seen in Fig. 12.66 that flotation with linolenic acid, which has three double bonds,
takes place in narrower pH range than with linoleic acid, which has two double bonds.
Still better is flotation with oleic acid having only one double bond. Oleic, linoleic and
linolenic acids have 17 carbon atoms in hydrocarbon radical.
Amines are also used in flotation of oxides and hydroxides. It is assumed that their
flotation takes place solely due to physical sorption of amines species on mineral sur-
faces. Therefore, such parameters as the state of mineral surface, mainly the sign of
the electrical charge and hydrophobicity, as well as the chemistry of amine collectors
solutions considerably govern flotation. On the basis of the equilibrium constant of
amine reaction with water, the CMC and iep of the amine solid phase, presented in
Table 12.42, it is possible to draw a diagram expressing predomination of different
dodecylamine species in aqueous solutions as a function of pH.
Table 12.42. Equilibrium constants of selected reactions, iep and CMC for dodecylamine in aqueous
(after Laskowski, 1988)
Reaction K
R–NH2 (aq)+H2O ⇔ R–NH3+ (aq)+OH– 4.3·10–4
R–NH2 (s) ⇔ R–NH2 (aq) 2.0·10–5
micellization CMC = 1.3·10–2 M
iep pH = 11
The dodecylamine domain diagram (Fig. 12.67) proves, that similarly to fatty ac-
ids, ions, micelle and amine phases are present in the collector-aqueous system. In
contrary to fatty acids, amine ions are positively charged, since in aqueous solution
they hydrolyze to –NH3+. Correctness of the phase diagram for amine drawn on the
basis of equilibrium constants was experimentally confirmed by Laskowski (1988)
and Laskowski et al. (1988). They found that the iep of non-soluble dodecylamine
equals 11 and that in certain pH ranges and concentrations unstable suspension with
coagulating colloidal amine particles was formed, while in other areas a stable amine
colloidal suspension was present. This is also shown in the diagram.
Typical course of adsorption and flotation of mineral oxides with amines has been
already shown in Fig. 12.37 for quartz. Quartz flotation as well as other minerals flota-
tion takes place as a result of electrostatic adsorption of positively charged amine ions
with negatively charged sites on particle surfaces. Flotation of minerals with amines
generally increases with pH starting from the iep. From a certain pH value flotation
decreases as a result of amine precipitation, which takes place along the transition line
for the reaction R-NH+3(aq) → R-NH2(s) presented in the predominance of amine spe-
cies diagram (Laskowski, 1988). This regularity has been known for a long time (Wat-
son and Manser, 1968; Leja, 1982) but without taking into account the concentration–
pH diagram, the decrease in flotation had been attributed to micellization while in fact
it is caused by amine precipitation.
-1 iep
log (amine concentration, kmol/m 3 )

micelle
+
(R-NH3 (aq))n colloidal suspension
-2
unstable
+ -
-3
aqueous R-NH 2 (s)
solution
-4
+ stable
R-NH 3 (aq)
-5
R-NH 2(aq)
-6
5 7 9 11 13
pH
Fig. 12.67. Diagram of predomination of various forms of dodecylamine
as a function of pH of solution (data after Laskowski, 1988)
100
quartz
80
R-NH2 precipitation
recovery, %
60
iep
40
5×10 -4 M C 12H25-NH2×HCl
20
competition of OH -
0
6 8 10 12 14
pH
Fig. 12.68. Relationship between quartz flotation with amine and pH. Following a good flotation in alka-
line solutions there is a drop in flotation as a result of precipitation of coagulating amine. At high pH an
increase of flotation is caused by stable of amine suspension (after Laskowski et al., 1988)
As it has already been mentioned amine ions are positively charged up to pH 11

and above this value they become negatively charged. This phenomenon not only af-
fects the properties of amine colloid suspensions but also flotation. This is so because
as pH increases, coagulating amine precipitates and flotation decreases. The decrease
in flotation continues until pH 11, which corresponds to the iep of amine and mini-
12. Flotation 371
mum of its stability. At higher pH values amine suspension becomes stable, flotation
slightly increases. However, at pH about 13-14 flotation disappears as a result of com-
petition with OH- ions (Fig. 12.68).
Different amines are applied for flotation. It was established that the longer the hy-
drocarbon chain of amine the lower amine concentration is needed flotation (Fig.
12.69). Amine with four carbon atoms requires very high amine concentrations, there-
fore, in practice amines having ten or more carbon atoms in the molecule are used.
100
80
QUARTZ
recovery, %
60
40
18 16 14 12 10 8 6 4
20
0 -8 -7 -6 -5 -4 -3 -2 -1 0
10 10 10 10 10 10 10 10 10
amine concentration, kmol/m 3
Fig. 12.69. Influence of length of the amine hydrocarbon chain on quartz flotation at pH = 6–7 and in-
creasing concentration of alkylamine acetate (Fuerstenau et al., 1964)
12.6.5. Sparingly soluble salts

Many useful minerals belong to this group. Among them the most important are:
fluorite (CaF2), barite (BaSO4), celestite (SrSO4), calcite (CaCO3), magnesite
(MgCO3), dolomite (CaMg(CO3)2), bastnäsite ((La,Ce)CO3F, scheelite (CaWO4), and
apatite (Ca5F(PO4)3) and its modifications. These minerals can float with different
ionic collectors. They float within wide pH and collector concentration ranges (Fig.
12.70).
When sparingly soluble salts type minerals are present in the same ore, there is a
need to separate them. Their flotation, at a particular pH, depends on collector concen-
tration. There are differences in flotation resulting from the way the separation is per-
formed as well as the origin of minerals (Fig. 12.71a and b).
Generally, the use of collector alone is not sufficient to selectively separate useful
mineral of salt type from the ore. Therefore, the use of depressants for selective flota-
tion of sparingly soluble salt type minerals is necessary. Numerous chemical com-
pounds are applied as depressants, including simple organic substances (water glass),
simple organic (citric) and complex (humic) acids, as well as carbohydrates (starch,
dextrin) and tree extracts (tannin, quebracho). Figure 12.72 presents separation of
fluorite from calcite by application of mixed Al2(SO4)3 + Na2SiO3 depressants. With-
out depressants the separation of fluorite from calcite is not be possible.
100 100
SDS
chrysocolla
80 80
DDA calcite
recovery, %
recovery, %
60 60
fluorit e
40 40 ..
NaOl bastnasite
20 20
barite
0 0
2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
pH pH
a b
Fig. 12.70. Flotation of sparingly soluble minerals: a) fluorite flotation with sodium oleate (cNaOl = 2·10–5
M, Pugh, 1986), dedecylamine (DDA) 9·10–5 M (Pugh, 1986), and sodium dodecyl sulfate (SDS) 1·10–4
M (Takamori and Tsunekawa, 1982), b) flotation of different minerals with potassium octyl hydroxamate:
chrysocolla 8·10–4 M, bastnäsite 5·10–4 M, calcite 5·10–4 M, barite 1·10–3 M (Fuerstenau and Pradip,
1984). pH range of flotation is usually greater at a higher collector concentration
100 100
80 80
ionic strength
calcite
0.002 M NaClO4
recovery, %
recovery, %
pH = 9.5
60 60
fluorite calcite chloro
apatite
40 40
fluorite
20 20 apatite
barite
0 0 -6 -5 -4 -3
-6 -5 -4 -3 10 10 10 10
10 10 10 10
sodium oleate concentration, mol/dm 3 concentration of sodium oleate, mol/dm3
a b
Fig. 12.71. Flotation of sparingly soluble minerals with oleic acid: a – after Finkelstein (1989), natural
pH, b – after Parsonage et al., (1982)
12. Flotation 373
Table 12.43 presents collectors and depressors used for separation of barite from
fluorite by flotation.
Table 12.43. Influence of different collectors and depressants on barite and fluorite flotation
(table after Pradel, 2000 based on Sobieraj, 1985)
Reagent Flotation of
barite fluorite
Collectors Alkyl sulfate
Pretopon floats well at pH 8–12 reduced flotation at pH 8
Siarczanol N-2 floats well at pH 4–12 flotation at pH 6–10
Sodium dodecyl sulfate
floats well at pH 4–12 cease of flotation at pH > 7
(SLS)
Alkyl sulfonate
gradual cease of flotation at
Oleic sulfosuccinate floats well at pH 5–12
pH < 8
Streminal ML floats well at pH 5–12 floats well at pH 5–12
Sodium kerylbenzosul-
floats well at pH 4–12 cease of flotation at pH > 7
fonate
Fatty acids
Sodium oleate floats well at pH 6.5–8.5 floats well at pH 4–10
Other collectors
Kamisol OC,
floats well at pH 3–12 flotation at pH 3–12
cationic collector
Rokanol T-16,
weak collecting power weak collecting power
nonionic collector
Depressant Tannins
Quebracho S no flotation in alkaline solu- total cease of flotation in alka-
(+ SLS) tions line solutions
Quebracho S
cease of flotation at pH > 6 cease of flotation at pH > 6
(+ Pretopon G)
cease of flotation in alkaline cease of flotation in alkaline
Tannin (+ SLS)
solutions solutions
Gallic acid cease of flotation in alkaline cease of flotation in alkaline
(+ SLS) solutions solutions
Tannin D (+ SLS)
solutions solutions
Tannin M (+ SLS)
solutions solutions
Other depressants
Dextrin
no flotation in alkaline solutions flotation at pH 6–9
(+ sodium oleate)
no flotation in acidic envi-
Glycerol full flotation depression at pH
ronment; no week flotation in
(+ sodium oleate) 5–11
alkaline solutions
100
80 fluorite
recovery, % calcite
60 +
depressor
+ depressor
40
20 fluorite calcite
0
2 4 6 8 10 12
pH
Fig. 12.72. Influence of depressant (70 mg/dm3 Al2(SO4)3 and 70 mg/dm3 Na2SiO3) on flotation of fluo-
rite and calcite mixture (dashed line) in the presence of sodium oleate (100 mg/dm3)
(after Abeidu, 1973). Solid line indicates flotation in the absence of depressant
Table 12.44. Solubility products (Ir) for selected compounds at 293 K (after Barycka and Skudlarski,
1993)
Compound Ir Compound Ir
1 2 3 4
Fluoride sulfate
CaF2 4.0·10–11 BaSO4 9.8·10–11
–9
SrF2 2.5·10 SrSO4 6.2·10–7
–9
MgF2 6.5·10 CaSO4 9.1·10–6
Chloride sulfide
AgCl 1.8·10–10 HgS 1.9·10–53
–5
PbCl2 1.7·10 Ag2S 6.3·10–50
Bromide Cu2S 7.2·10–49
AgBr 4.6·10–13 CuS 4.0·10–36
PbBr2 2.8·10–5 PbS 6.8·10–29
Iodide ZnS 1.2·10–28
–17
AgI 8.3·10 NiS 1.0·10–24
–9
PbI2 7.1·10 CoS 3.1·10–23
Carbonate FeS 5.1·10–18
–14
PbCO3 7.2·10 MnS 1.1·10–15
–11
ZnCO3 1.7·10 Cyanide
CaCO3 7.2·10–9 Hg2(CN)2 5.0·10–40
–8
MgCO3 3.5·10 CuCN 3.2·10–20
Hydroxide chromate
Fe(OH)3 4.5·10–37 PbCrO4 2.8·10–13
–17
Zn(OH)2 3.3·10 BaCrO4 1.2·10–10
–11
Mg(OH)2 1.2·10 CuCrO4 3.6·10–6
12. Flotation 375
The mechanism of reaction between polar collectors and sparingly soluble salts
minerals is quite complex and the degree of complication is even higher for minerals
from the oxide and hydroxide group. It results from the fact that additional parameters
become involved, namely concentration of anions and cations in flotation system
which are determined by the solubility products of minerals. The solubility products of
selected compounds at 293 K, after Barycka and Skudlarski (1993) are gathered in
Table 12.44.
The solubility products indicate what the concentration of salt ions in mineral
aqueous suspension is. For fluorite the solubility product at 200C (293 K) is 9.0.10-11.
It means that due to fluorite dissolution the concentration of calcium ions will be cCa++
= (Ir/4)1/3 = 2.2. 10-4 kmol/m3, while the concentration of fluoride ions will be twice
higher. These values are in agreement with direct measurements of Marinakis and
Shergold (1985). Solubility products of other compounds, including salts, can be
found in the literature or calculated on the basis of thermodynamic data (free enthalpy
of formation) of chemical species taking part in the reaction. The data can be taken
from handbooks (CRC, 1986/87) and other sources. When calculating ions concentra-
tion one should remember that solubility product Ir is equal to the equilibrium constant
of a dissolution reaction Kr, since the activity of the solid phase is 1. Fluorite dissolu-
tion can be written as:
CaF2 = Ca2+ + 2F– , (12.100)
and the equilibrium constant of the reaction and the solubility product will be given
by:
Ir = Kr = exp(– ΔGr0 /RT) = (Ca2+) (F–)2/(CaF2), (12.101)
where
R – gas constant
T – temperature, K.
( ) denotes activity which in technical calculations can be replaced with concentration
in kmol/m3, activity of the solid phase of CaF2 is 1.
In Eq. 12.101 ΔGr0 is the standard free enthalpy of reaction calculated from the re-
lation:
ΔGr0 = Σνi ΔGi0, f ,
where:
ΔGi0, f – standard of formation (formation of a given species)
νi – stoichiometric coefficient of reaction which is positive for products and nega-
tive for substrates.
According to Fuerstenau and Urbina (1988) adsorption of a collector on sparingly

soluble minerals can take place physically or chemically, as well as it can be a surface
reaction. It is possible to distinguish:
• physical adsorption of a collector due to electrical double layer with preservation
of ions hydration (amine on barite),
• chemical adsorption of a collector on non-hydrolized surface ion located in a
crystalline structure (fatty acids on barite) or on a hydrolized ion (hydroxamate
on bastnesite),
• surface reaction of a collector with surface metal ion leading to its removal from
crystalline structure and forming at the interface multilayer collector-metal ions
compound having lower solubility than the mineral (oleate on calcite),
• surface reaction of a collector with metal ion situated near the surface (outside
crystalline structure) and forming a collector-metal ion compound at the mineral
surface (tridecane on hydroxapatite).
The above modes of adsorption after Fuerstenau and Pradip (1984) can schemati-
cally be delineate as:
a) physical sorption
Me++M + H2O–collector → Me++MH2O–collector
Me++M + H2O(OH–)H2O → Me++MH2O OH– H2O
b) chemical sorption
Me++M + –collector → Me+Mcollector
++
Me M + H–collector → Me+Mcollector + H+
Me++M + OH– → Me+MOH
+
Me MOH + collector -
→ MeM (OH)collector
c) adsorption, and then surface reaction
Me++M + –collector → M(Mecollector)+
Me++M + H–collector → M(Mecollector)+ + H+
++
Me M + OH –
→ M(MeOH)+
Me++M + HCO 3− → M(MeHCO3)+
d) surface reaction
M(MeOH)+ + –collector → M(Mecollector)+ +OH–
M(MeOH)+ + H-collector → M(Mecollector)+ + H2O
M(MeOH)+ + OH– → M(MeOH)2
M(MeOH) + 2Hcollector → MMe(collector)2 + H+ + H2O
+
where:
symbol M denotes metal ions surface,
symbol denotes site after removing metal ion from the crystalline structure.
The list does not cover all the reaction which can take place between collector ions
and mineral ions. A compound of collector and mineral ion can be formed in the bulk
12. Flotation 377
solution and it can undergo sorption on the surface leading to flotation. If the collec-
tor–ion compound formed as a result of surface reaction is removed from the surface,
flotation is not possible.
It is assumed that chemical reactions between collector and cation dominate in flo-
tation systems involving fatty acids, alkylosulfonates, alkylosulfates, as well as chelat-
ing reagents (hydroxamates) while physical adsorption prevails in flotation with
amines. Since pH, ionic strength of the solution, concentration of ions released from
mineral and the way the mineral is added to a flotation system govern the reactions,
the flotation results obtained by different researches can be considerably different both
for the same minerals floated by different investigators and for the same minerals of
different origin (Fig. 12.71).
A characteristic feature of the discussed flotation systems is possible occurrence of
collector multilayers, especially of oleate, on mineral surfaces. According to Rao and
Forssberg (1991) the multilayer adsorption of collectors and metal ions (coming from
the mineral) compounds is characteristic for the system floated in suspensions having
low content of a mineral. The dissolved metal ions form chemical compound with col-
lector ions in the solution and they are adsorbed on the surface of particles which total
area is small. Thus, the adsorption is multilayered. If flotation is performed in dense
suspensions, the amount of formed ion-collector compound is the same but the surface
area of particles is large. It is dictated by the solubility product which is not dependent
on the amount of mineral particles in the suspension. The collector-metal ions com-
pounds spread on a very large area, therefore adsorption becomes limited. At appro-
priate conditions, before multilayer precipitation, there is a monolayer of oleate on
calcite, while a double layer exists on such minerals as fluorite, apatite, scheelite and
barite (Rao and Forssberg, 1991).
Chemical composition of the mono- or multilayers depends on the number of
metal ions available in the solution. If there are few of them, the adsorption of the col-
lector ions on the surface takes place not only through reaction and bridging with cal-
cium ions, but also with participation of sodium ions, as it has been predicted by
Drzymala (1986). According to Rao and Forssberg (1991), depending on the sign of
surface potential and its value for calcium minerals, the following reactions, leading to
the formation of mono- and double layers of compounds, take place:
on electrically neutral sites:
–CaOH + –OOCR = –Ca+ –OOCR + OH–
–CaOH + Na+ –OOCR + OH– = –CaO Na OOCR– + H2O
–CaOH + Ca++ –OOCR + OH– = –CaO Ca OOCR– + H2O
on positively charged sites:
–CaOH2+ + –OOCR + OH– = –Ca+ –OOCR + H2O
on negatively charged sites:
–CaOδ– Na+ + –OOCR = –CaO Na OOCR, where δ < 1,
–CaOδ– Ca++ + –OOCR = –CaO Ca OOCR, where δ ≤ 1.
Somasundaran and Ananthapadmanabhan (1986) investigated the role of the rela-

tion between mineral ions and collector concentrations. They found out that very often
it is possible to record a simple dependence between flotation and onset of formation
of collector-metal ions bulk compounds. This is the case for apatite-oleate, dolomite-
oleate, magnesite-fatty acids, calcite-fatty acids. If there is in the solution some surplus
of cations released from the mineral, the precipitation can occur in the solution, result-
ing in significant worsening of flotation. The cause of the differences in flotation can
also be ions diffusion. If the speed of chemical dissolution is high, the system will
tend to form the metal ions-collector compound in the solution. It is better if diffusion
advances slowly – then the formation of the compound between a collector and metal
ions occurs close to the surface.
To float sparingly soluble salt type minerals one can use hydroxamates of a general
formula:
R1 N OH
(12.102)
R2 C O
.
These compounds can be called N-alkyl hydroxylamine derivatives or carboxylic
acids oximes. The characteristic feature of these chelating collector is that they react
best at pH about 9 and form multilayers. It suggests that surface layers contain both
anions and neutral molecules of hydroxamic acid, as their pK, indicating at which pH
concentration of both forms is identical, is about 9 (Fuerstenau and Pradip, 1984).
12.6.6. Soluble salts

Flotation of soluble salts such as halite (NaCl), silvinite (KCl), carnallite
(KCl.MgCl2.6H2O), and kieserite (MgSO4.H2O) is conducted with the use of cationic
and anionic collectors from their saturated solutions. Flotation of soluble salt type
minerals is performed in order to separate one salt form another, as well as to remove
from gangue minerals which can be fine particles of alumino-silicates (illite, chlorites)
and other minerals, e.g. carbonates (dolomite). The separation of soluble salts is usu-
ally based on their surface charge (Table 12.45).
According to Miller et al. (1992) a sign of surface charge of soluble salts is possi-
ble to predict using the theory of ions hydration (Table. 12.45). For instance, NaCl is
positively charged while KCl is negatively charged and therefore, these minerals can
be separated by flotation using amines (in that case KCl floats) or fatty acids (to pro-
voke NaCl flotation) (Miller et al., 1992). Other soluble salt type minerals also follow
the electrostatic mechanism and their flotation using amine is shown in Fig. 12.73.
The mechanism of collector - ions interaction in saline solutions is not well known.
Very likely there is no a single mechanism of flotation. It has been assumed that KCl
flotation takes place as a result of heterocoagulation of positively charged amine col-
12. Flotation 379
lector with negatively charged KCl particles at concentrations higher than amine solu-
bility i.e. when amine in an aqueous solution is present in colloidal form.
Table 12.45. Sign of surface charge for selected soluble salts (after Miller et al., 1992)
Salt Surface charge sign Salt Surface charge sign
measured predicted* measured predicted*
LiF + +– KBr – +
NaF + + RbBr – +
KF + + CsBr + +
RbF + + LiI – –
CsF + + NaI – –
LiCl – – KI +
NaCl + – RbI – –
KCl – + CsI + +–
RbCl + + NaI·2H2O +
CsCl + + K2SO4 –**
LiBr – – Na2SO4·10H2O –**
NaBr – – Na2SO4 –**
* Predicted from the ions hydration theory for ions in crystalline lattice (Miller et al., 1992).
** Hancer et al., 1997.
100
80
K2SO 4 KCl Na2SO 4×10H 2O
60
40
20 Na2SO 4 NaCl
0 -6 -5 -4 -3 -2
10 10 10 10 10
dodecylamine hydrochloride, kmol/m 3
Fig. 12.73. Flotation of different salts in the presence of dodecylamine

(after data of Hancer et al., 1997, data for NaCl after Miller et al., 1992)
This mechanism is valid for K2SO4 flotation, since flotation with amines takes place at
concentrations much lower than concentration needed for precipitation. Investigations
of Hancer et al. (1997) showed that successful flotation requires thermodynamic sta-
bility of a salt. Sodium sulfate can serve as an example since it exists in anhydrous
(Na2SO4) and hydrated (Na2SO4 .10 H2O) forms. Na2SO4 does not float with amine
despite its negative charge. Yet, this salt is thermodynamically unstable, since it hy-
drates and forms Na2SO4.10H2O which floats well because its surface charge is nega-
tive and it is thermodynamically stable at ambient temperature. Thus, the presence of

hydrated salt layers does not interfere with amine flotation if the salt is negatively
charged. The lack of flotation of halite with amine can be explained by the unmatched
size of amine cation and surface sodium ion in NaCl crystalline structure, which is not
the case for KCl (Laskowski, 1969).
A technical problem in flotation of soluble minerals is not only the separation of
different salts, but also removal of fine gangue particles. The unwanted fine particles
significantly increase the consumption of a collector, lower the quality of final prod-
ucts, and harmfully stabilize froth. Reagents like carboxymethylocellulose (CMC),
polyacryloamides (PAM), and natural polymers (starches) (Alonso and Laskowski,
1999; Laskowski, 1994) are used as depressors of fine gangue particles. The reagents
depressing fine gangue particles which are applied in flotation of soluble salts are also
called blinders. Other approaches of separation of gangue fines from salt minerals in-
clude selective flocculation of fines with high molecular polyacrylamide followed by
flotation of flocculated fines with amine (Perucca and Cormode, 1999). Figure 12.74
presents the effects of selected fined depressants on the results of KCl flotation with
amine. In addition to a collector, traditional frothers are used in salt flotation. Mixing
frother with a collector to achieve better amine dispersion is recommended.
100
KCl
80
recovery, %
60 CMC
PAM
guar
40
20 fines
0
0 50 100 150 200 250
dosage of depressor, g/Mg
Fig. 12.74. Application of depressants for removing fines of gangue minerals during amine flotation of
KCl (after Alonso and Laskowski, 1999). CMC denotes carboxymethylcellulose
PAM - polyacrylamide of low molecular weight, while guar is a natural polysaccharide
12.7. Flotation devices

Flotation can be performed on a small and large scale. The smallest scale involves
flotation of a single particle. This can be accomplished in a Hallimond microflotation
cell (tube). Tones of material per hour are processed in industrial flotation machines.
The microlaboratory scale flotation devices provide a quick and cheap way of deter-
12. Flotation 381
mination of flotation properties of ores and raw materials and their components. Gen-
erally, microlaboratory cells are designed to test flotation of one minerals, not mixture
of minerals, because in a lean ore a very small sample may have no a single particle
of the investigated mineral. It can happen in flotation of gold ore containing several
ppm (parts per million) of Au. A disadvantage of small scale flotation devices, except
that of Siwek et al., 1981 (Fig. 12.75e), is high mechanical flotation (entrainment) of
hydrophilic particles, which masks the true flotation results. The entrainment in the
Hallimond tube shown in Fig. 12.75b is governed by the following relation (Drzymala
and Lekki, 1989a; 1989b; Drzymala, 1984):
amax (ρp – ρw)/ρw) = LH = 0.023 ±0.002 (cm), (12.103)
for density of particles greater than 2.0g/cm3.
For particles of density lower than 2.09g/cm3 another expression is used (Drzy-
mala, 1994, 1999):
amax ((ρp – ρw)/ρw)0.75 = LL = 0.0225 ±0.0025 (cm), (12.104)
where:
ρp – particle density,
ρw – aqueous density,
amax – maximum particle size which can float mechanically; amax is equal to d50, read
from the partition curve representing recovery of mechanical flotation as a func-
tion of particle size,
LL, LH – constants.
Entrainment decreases as Hallimond cell height decreases. For the cell 1 m in
height mechanical flotation of quartz and magnetite decreases from about 100μm to
about 20μm (Lukaszewska, 1998). Therefore, relatively large particles (in order of
mm in size) should be used for Hallimond tube flotation tests. The appropriate size
can be calculated from Eqs (12.103) and (12.104). Another approach to flotation tests
in the Hallimond tube, aiming at avoiding mechanical flotation, is conducting flotation
measurements within a very short period of time, typically from 15 to 60 seconds.
Frequently used microflotation cells are presented in Fig. 12.75.
When working out technological basis of the enrichment of raw materials, labora-
tory flotation machines are used ranging in capacity from several grams to several
kilograms. Their list includes Denver, Mechanobr, Wameco, and IMN laboratory flo-
tation machines. Laboratory flotation columns are less frequently used. Figure 12.76
shows the Mechanobr laboratory flotation machine and its flotation cell, Fig. 12.77
presents the Denver laboratory flotation machine while Fig 12.78 other laboratory flo-
tation devices.
water level
air bubble + mineral particle
scale
concentrate
capillary
capillary
a b
mm
mm
25
25
20
0m 20
m 0m
m
30°
30°
100 mm
100 mm
stirrer
porous
capillary glass
60 μm
magnetic magnetic
stirrer stirrer
8 mm 8 mm
c d
stirrer
stirring tank
120-300 mm
A
100-140 mm
receiver C B1 B2
magnetic
stirrer capillary magnetic
manometer stirrer
Fig. 12.75. Flotation cells used for flotation on a microlaboratory scale: a – original Hallimond tube (Hal-
limond, 1944), b – contemporary Hallimond tube designed by Ewers (Sutherland and Work, 1955)
equipped with a scale by Lekki (1970), c – modified Hallimond tube (Fuerstenau et al., 1957),
d – porous plug version (Nagy et al., 1962), e – cell for fine particles (Siwek et al., 1981), f) universal cell
(Dobias, 1983)
12. Flotation 383
drive
air
a b
c
Fig. 12.76. Mechanobr laboratory flotation machine
(with subareation ) (Drzymala, 2001) (a), its flotation cell (b) (after Luszczkiewicz –unpublished, based
on Berger and Efremov, 1962, and its photograph (c)
a b
Fig. 12.77. Denver laboratory flotation machine with cell volume 5 dm3 (Konopacka, 2005 and Bulletin,
1994) (a) and its photograph (b)
water level
deflector
magnetic stirrer
porous glass
water level
x
gas
froth
product stirrer
porous glass
flotaton product
gas
a b
Fig. 12.78. Other laboratory flotation devices. a) Cylindrical cell equipped with magnetic stirrer (Fuer-
stenau, 1964), b) laboratory flotation device of Partridge and Smith, 1971
12. Flotation 385
FEED
AIR
WATER
CONCENTRATE
CELL
TAILING
a
FEED
water
(0.01 m/s) (0.0005-0.003 m/s)
concentrate FROTH LAYER
phase border
air bubbles
(0.5-3 mm)
FLOTATION
REGION
gas
(0.005-0.03 m/s)
TAILING
(~0.01 m/s)
AIR
FEED
c
Fig. 12.79. Industrial flotation machines: a – Jameson cell (Evans et al., 1995), b – column cell (Finch et
al., 1995), c – tank cell (Denver DR) (Laskowski and Luszczkiewicz, 1989)
Mechanical flotation of hydrophilic particles can be also observed in laboratory

and industrial flotation machines. The entrainment is frequently proportional to the
amount of water collected along with the concentrate (Ross, 1991). The size of the
entrained particles in the Mechanobr laboratory flotation machine is governed by the
equation (Hrycyna, 1999):
L = d50 (ρp – ρw )1.65 = 0.1, (12.105)
where:
d – diameter particle at 50% recovery (d50), mm
ρ = ρp – ρw – density of particle in water, g /cm3.
Equation (12.105) is valid for the flotation cell 200 cm3 in volume and using α-
terpineol as a frother at concentration of 12.5 mg/cm3. Equation (12.105) can be used
to describe entrainment when only hydrophilic particles are tested. The delineation of
entrainment of hydrophilic particles in the presence of floating hydrophobic grains
has not been systematically investigated.
Different flotation machines are used by industry. They can be divided into me-
chanical, pneumatic, mechanical-pneumatic, vacuum, column, and others. Their
names include PA, Wemco, Fagergreen, Mechanobr (Sablik, 1998). Mechanical flota-
tion machines are equipped with an impeller which disperses air and stirs particles and
bubbles. The pneumatic flotation machines do not have impellers. The air is supplied
from an outer source. During depressurizing the air is dispersed and stirs the suspen-
sion. Pneumatic Jameson flotation machine is shown in Fig. 12.79a (Evans et al.,
1995). The pneumatic-mechanical flotation machines represent a variation of the me-
chanical ones in which pressurized air is supplied (IZ flotation machine, Szczerba et
al., 1999, Denver (Fig. 12.79c), and Agitar). In vacuum flotation machines air bubbles
are formed directly on the particle surfaces due to decreased pressure in flotation sys-
tem.
In Poland flotation machines are used in coal and copper industries (e.g. IMN, IZ,
OK, HG, DR types and others).
The capacity of flotation machines can be as high as 100 m3 and more.
Flotation of some ores and raw materials is conducted in flotation columns. A
scheme of flotation column is presented in Fig. 12.79b. Theoretically, they are supe-
rior to other flotation machines as the main flotation stage and several cleaning flota-
tion steps take place in the same column in one flotation.
Literature
Abeidu A.M., 1973. Selective depression of calcite from fluorite, Trans. IMM, Section C., 82, C49-C50.
Abramov A.A., Leonov S.B., Sorokin M.M., 1982. Chemistry of flotation systems, Nedra, Moskva, in Rus-
sian.
Adamson A.W., 1967. Physical chemistry of surfaces, 3rd edition, Wiley, New York.
12. Flotation 387
Alonso E.A., Laskowski J.S., 1999. Selection of polymers used as slime modifiers in potash ore, in: Poly-
mers in mineral processing, J. Laskowski (ed.), Proc. 3rd UBC-McGill Bi-annual Inter. Symp. on Fun-
damentals of mineral processing, Met. Soc. CIM Montreal, 405–419.
Aplan F.F., Chander Ch., 1988. Collectors for sulfide mineral flotation, in: Reagents in mineral technology,
Surfactant science series, Vol. 27., P. Somasundaran, B.M. Moudgil (eds.), 335–369, Dekker, New
York.
Aplan F.F., Simkovich E.Y., Spearin E.Y., Thompson K.C., 1980. In: The physical chemistry of mineral –
reagent interactions in sulfide flotation, P.E. Richardosn et al. inni (eds.), US Bureau of Mines, IC 8818,
s. 215.
Arbiter N., Cooper H., Fuerstenau M.C., Harris C.C., Kuhn M.C., Miller J.D. and Yap R.F., 1985. Flotation,
in: N.L. Weiss (ed.), SME Mineral Process. Handbook Vol. 1, SME, N.Y., 39–109.
Arbiter N., Harris C.C., 1962. Flotation kinetics, in: Froth flotation – 50th anniversary volume, D.W. Fuer-
stenau (ed.), AIMM/SME, New York, 215–246.
Aston J.R., Drummond C.J. Scales F.J., Healy T.W., 1983. Frother chemistry in fine coal processing,
Proc.2nd Australian Coal Prep. Congress, R.I. Whitmore (ed.), Westminster Press, Brisbane, 148–160.
Aveyard R. and Clint J.H., 1995. Liquid droplets and solid particles at surfactant solution interface, J. Chem.
Soc., Faraday Trans., 91 (17), 2681–2697.
Aveyard R., Binks B.P., Fletcher P.D.I. and Rutherford C.E., 1994. Measurements of contact angles of
spherical monodisperse particles with surfactant solutions, Colloids and Surfaces, 83: 89–98.
Bara D., Kisielewski W., 1984. Analiza zwiększania uzysku przy produkcji siarki metodą flotacyjno-
rafineryjną, w: materiały seminarium „Procesy technologiczne surowców mineralnych z uwzględnie-
niem międzyfazowych zjawisk powierzchniowych”, Baranów Sandomierski, 1–14.
Bartell F.E., Whitney C.E., 1933. Adhesion tension, J. Phys. Chem., 36: 3115–3119.
Barycka I., Skudlarski K., 1993. Podstawy chemii, Wydawnictwo Politechniki Wrocławskiej, Wrocław.
Berger G.S., Efimov I.A., 1962. Methods of separation of mono-mineral fractions, 2nd edition, Gos-
geoltechnizdat, Moscow, in Russian.
Bigosiński J., 1998. Wpływ zawartości rozdzielanych składników na wzbogacanie metodą flotacji, Ph.D.
thesis, Instytute of Mioning Eng. Wroclaw, Poland
Bilans 1997. Siarka, w: Bilans gospodarki surowcami mineralnymi w Polsce na tle gospodarki światowej –
1995, PAN, Krakow, Poland
Billett D.F, Hough D.B., Ottewill R.H., 1976. Studies on the contact angle of the charged silver iodide-
solution-vapour interface, J. Electroanalytical Chem., 74, 107–120.
Bolewski A., Manecki A., 1993. Mineralogia szczególowa, Wyd. PAE, Warszawa.
Brown D.J., 1962. Coal flotation, in: Froths flotation, 50th Anniversary Volume, D.W. Fuers- tenau (ed.),
AIME, New York, 518–538.
Bulletin No. L-FLO-1. 1994. Joy Process Equipment Limited (Denver Equipment). England
Chander S., 1988. Inorganic depressants for sulfide minerals, in: Reagents in Mineral Technology, P. Soma-
sundaran i B.M. Moudgil (eds.), Surfactants Science Series, Vol. 27, Chap. 14, 429–469, Dekker, New
York.
Chander S., Wie J.M., Fuerstenau D.W., 1975. On the native floatability and surface properties of naturally
hydrophobic solids, in: Advances in interfacial phenomena of particulate/solution/gas systems: applica-
tions to flotation research, P. Somasundaran and R.B. Grieves (eds.), AICHE Syp. Series, No. 150, Vol.
71, AICHE, New York, 183–191.
Chibowski E., Hołysz L., Zdziennicka F., Gonzalez-Caballero F., 1996. Is the contact angle hysteresis due to
film present behind the drop?, in: Surfactants in solutions, A.K. Chattopadhyay and K.L. Mittal (eds.),
Dekker, New York, 64, 31–53.
Chmielewski T., 1990. Some aspects of pyrite hydrophobicity – investigation on cleaved electrodes, Journal
of Mining and Metallurgy (Glastnik Rudarstwa i Metaurgije), 26(2), 133–144.
Choi H.S., Oh J., 1965. Silicates flotation, J. Inst. Min. Metall, Japan, 81, 614–620.
Clint J.H. and Tylor S.E., 1992. Particle size and interparticle forces of overbased detergents– a Langmuir
trough study, Colloids Surfaces, 65(1): 61–67.
Cornell R.M., Schindler P.W., 1980. Infrared study of the adsorption of hydroxy carboxylic acids on α-
FeOOH and amorphous iron(III) hydroxide, Colloid and Polymer Sci., 258, 1171–1175.
CRC, 1986/87. Handbook of chemistry and physics, 67th ed. (R.C. Weast (ed.), CRC Press, Boca Raton,
Florida, USA.
Crowl V.T. and Wooldridge W.D.S., 1967. A method for the measurement of adhesion tension of liquids in
contact with powders, SCI Monograph, 25: 200–212.
Czaplinski A., 1994. Charakterystyka węgli kamiennych i ich właściwości technologiczne, in: Węgiel ka-
mienny, A. Czapliński (red.), Wyd. AGH., Kraków.
Daellenbach C.B., Tiemann T.D., 1964. Chelation of quartz activation ions in oleic acid flotation, Trans.
AIME 229, 59–64.
Davies J.T., Rideal E.K, 1963. Interfacial Phenomena, Academic Press, New York.
de Bruyn P.L., Agar G.E., 1962. Surface chemistry of flotation, in: Froth flotation – 50th anniversary vol-
ume, D.W. Fuerstenau (ed.), AIMM/SME, New York, 91–138.
Deren J., Haber J., Pampuch R., 1977. Chemia ciała stałego, PWN, Warszawa.
Derjaguin B.V., 1989. Theory of stability of colloids and thin films, Consultants Bureau, New York.
Dixit S.G., Biswas A.K., 1973. Studies on zircon-sodium oleate flotation system: 1-solid-liquid interfacial
parameters, Trans. IMM, Sec. C., 82, C200-206.
Dobias B., (ed.), 1993. Coagulation and flocculation, Surfactant Science Ser., 47, Marcel Dekker, N.Y.,
Chap. 7, 321–353.
Dobias B., 1983. New modified Hallimond tube for study of flotation of minerals from kinetic data, Trans.
IMM., Sec.C, 92, C164–166.
Dobias B., 1995. Salt-type minerals, w: Flotation Science and Technology, K.A. Matis (ed.), Dekker, New
York, 179–259.
Doren A., Vargas D., Goldfarb J., 1975. Non-ionic surfactants as flotation collectors, Trans. IMM, Sec. C.,
C34-C37.
Dowling E.C., Klimpel R.R., Aplan F.F., 1985. Model discrimination in the flotation of porphyry copper ore,
Minerals and Metallurgical Processing, 2, 87–97.
Drzymala J., 1979. Adsorption isotherm of potential determining ions at oxide/aqueous solution interface,
Polish Journal of Chemistry, 53, 1809–1820.
Drzymala J., 1986. O hydrofobizacji powierzchni ciał stałych w wodnych roztworach oleinianowych, Fizy-
kochemiczne Problemy Mineralurgii, 18, 63–84.
Drzymala J., 1989. Chemistry of oleic acid-H2O-NaCl system at 25°, in: Surfactants in Solutions, Vol. 7.,
K.L. Mittal (ed.), Plenum Press, New York.
Drzymala J., 1990. Własciwosci wodnych roztworów i emulsji oleinianowych, Prace Naukowe Instytutu
Chemii Nieorganicznej i Metalurgii Pierwiastków Rzadkich Politechniki Wrocławskiej 61, Monografie
29, Wydawnictwo Politechniki Wrocławskiej, Wroclaw.
12. Flotation 389
Drzymala J., 1994a. Characterization of materials by Hallimond tube flotation. Part 1. Maximum size of
entrained particles, Int. J. Miner. Process., 42: 139–152.
Drzymala J., 1994b. Characterization of materials by Hallimond tube flotation, Part 2. Maximum size of
floating particles and contact angle, Int. J. Miner. Process., 42, 153–167 oraz errata, Int. J. Miner. Pro-
cess. 43 (1995), 135.
Drzymala J., 1994c. Hydrophobicity and collectorless flotation of inorganic materials,
Advances in Colloid and Interface Sci., 50, 143–185.
Drzymala J., 1995. Określanie możliwości zmniejszenia zawartości ołowiu w koncentratach miedziowych
metodami flotacyjnymi przy wykorzystaniu najnowszych osiągnięć z fizykochemii powierzchni, Raport
Instytutu Górnictwa Politechniki Wrocławskiej I, 1995.
Drzymala J., 1999a. Characterization of materials by Hallimond tube flotation. Part 3. Maximum size of
floating and interacting particles, Int. J. Miner. Process, 55: 203–218.
Drzymala J., 1999b. Entrainment of particles with density between 1.01 and 1.10 g/cm3 in a monobubble
Hallimond flotation tube, Mienerals Eng. 12(3), 329–331.
Drzymala J., 1999c. Flotometric hydrophilicity of coal particles, in: Mining Science and Technology ’99, H.
Xie i T Golosinski (eds.), Balkema, Rotterdam, 533–537.
Drzymala J., 2001. Hydrofobowość naturalna soli nieorganicznych (in preparation).
Drzymala J., 2001. Ice/aqueous interface, in: Encyclopedia of surface & colloid science, A. Hubbard (ed.),
Dekker, New York.
Drzymala J, 2001. Podstawy mineralurgii, Oficyna Wydawnicza PWr.
Drzymala J., Bigosinski J., 1995. Collectorless flotation of sulfides occurring in the fore-Sudetic copper
minerals deposit of SW Poland, Mineralogia Polonica, 26, 63–73.
Drzymala J., Chmielewski T., 1992. Flotometry of very hydrophobic materials in the multibubble Hallimond
tube, Prace Naukowe Instytutu Górnictwa Politechniki Wroclawskiej, nr 65, Studia i Materialy nr 23,
1992.
Drzymala J., Laskowski J., 1981. Zastosowanie odczynników chelatujących we flotacji, Fizykochemiczne
Problemy Mineralurgii, 13, 39–64.
Drzymala J., Lekki J., 1989a. Mechanical, contactless, and collector flotation in the Hallimond tube, J. Col-
loid Interface Sci., 130: 197–204.
Drzymala J., Lekki J., 1989b. Flotometry – another way of characterizing flotation, J. Colloid Interface Sci.,
130, (1), 205–210.
Drzymala J., Lekki J., Kiełkowska M.M., 1987. A study of the germanium-sodium oleate flotation system,
Powder Technology, 52, 251–256.
Drzymala J., Lekki J., Laskowski J., 1979. Surface dissociation constants for solid
oxide/aqueous solution systems, Colloid and Polymer Sci., 257, 768–772.
Drzymala J., Sadowski Z., Hołysz L., Chibowski E., 1999. Ice/aqueous interface: zeta potential, point of zero
charge and hydrophobicity, J. Colloid Interface Sci., 220, 229–234.
Dupre A., Theorie Mecanique de la Chaleur Gauthier-Villars, Paris, 1869, pp. 369ff
Dutkiewicz E.T., 1998. Fizykochemia powierzchni, WNT, Warszawa.
Eigeles M.A., 1964. Principles of non-sulfide minerals flotation, Nedra, Moskva, in Russian.
Ek C., 1992. Flotation kinetics. Innovation in flotation technology, Vol. 208, Kluwer Academic Publishers,
London.
Evans G.M., Atkinson B.W., Jameson G.J., 1995. The Jameson cell, in: Flotation Science and Egineering,
K.A. Matis (ed.), Dekker, New York, 1995, Chap. 12, 331–362.
Finch J.A., Urbe-Salas A., Xu M., 1995. Column flotation, in: Flotation Science and Engineering, Dekker,
K.A. Matis (ed.), New York.
Finkelstein N.P., 1989. Review of interactions in flotation of sparingly soluble calcium minerals with anionic
collectors, Transaction IMM, Sec. C, 98, C157–177.
Fleming M.G., 1952. Effects of alkalinity in the flotation of lead minerals, Trans. AIME., 193, 1231–1236.
Fowkes F.M., Riddle F.L., Pastore W.E., Weber A., 1990. Interfacial interactions between self-associated
polar liquids and squalane used to test equations for solid-liquid interfacial interactions, Colloids and
Surfaces, 43, 367–387.
Fowkes M. F., Attractive forces at interfaces, Ind. and Eng. Chem, 56(12), 40–52.
Freund J., Dobias B., 1995. The role of surface tension, in: Flotation science and technology, K.A. Matis
(ed.), Dekker, New York.
Fuerstenau D.W., Healy T.W., Somasundaran P., 1964. The role of the hydrocarbon chain of alkyl collectors
in flotation, Trans. AIME, 229, 321–323.
Fuerstenau D.W., Metzeger P.H., Seele G.D., 1957. How to use this modified Hallimond tube for better
flotation testing, Eng. Mining Journal, 158 (3), 93–95.
Fuerstenau D.W., Pradip, 1984. Mineral flotation with hydroxamate collectors, in: Reagents in the mineral
industry, M.J. Jones and R. Oblatt (eds.), IMM, 161–168.
Fuerstenau D.W., Urbina H., 1988. Flotation fundamentals, in: Reagents in mineral technology, Surfactant
Science Series, Vol.. 27, P. Somasundaran and B.M. Moudgil (eds.), Dekker, New York, 1–38.
Fuerstenau D.W., Wiliams M.C., Narayanan K.S., Diao J.L., Urbina R.H., 1988. Assessing the wettability
and degree of oxidation of coal by film flotation, Energy and Fuels, Vol. 2(3).
Fuerstenau M.C. 1964. An improved micro-flotation technique. Eng. Min. J., 165(11): 108-109.
Fuerstenau M.C., 1995. Oxide and silicate flotation, in: Flotation science and engineering, K.A. Matis (ed.),
Dekker, New York, 89–126.
Fuerstenau M.C., Han K.N., 1988. Activators, in: Frothers, in: Reagents in Mineral Technology, P. Soma-
sundaran (ed.), Surfactants Science series, Vol. 27, Dekker, New York, Chp.13, 411–428.
Fuerstenau M.C., Kuhn M.C., Elgilliani D.A., 1968. The role of dixanthogen in xanthate flotation of pyrite,
Trans. AIME, 241, 148–156.
Fuerstenau M.C., Palmer B.R., 1976. Anionic flotation of oxides and silicates, in: Flotation
– AM Gaudin memorial volume, Vol. 1, Chap. 7, M.C. Fuerstenau (ed.), AIME, New York, 148.
Garrels R.M, Christ Ch.L., 1965. Solutions, Minerals and Equilibria, Harper and Row, New York.
Gaudin A.M., 1963. Flotacja, Slask, Katowice.
Gaudin A.M., Miaw H.L., Spedden H.R., 1957. Native floatability and crystal structure, in: Electrical phe-
nomena and solid/liquid interface, Proc. 2nd Int. Congr. Surface Activity, London. Butterworths, Lon-
don, 202–219.
Good R., Keller D.V., 1989. Fundamental research on surface science of coal in support of physical benefi-
ciation of coal, Quarterly Report No. 7, State University of New York at Buffalo, 25.
Gutierrez-Rodriguez J.A., Purcell R.J., & Aplan F.F., 1984. Estimating the hydrophobicity of coal, Colloids
and Surfaces 12, 1–25.
Guy P.J., Trahar W.J., 1985. The effects of oxidation and mineral interactions on sulfide flotation, in: Flota-
tion of sulfide minerals, K.S.E. Forssberg (ed.), Elsevier, Amsterdam, 61–79.
Hallimond A.F., 1944. Laboratory apparatus for flotation tests, Mining Magazine, 70, 87–91.
Hancer M., Hu Y., Fuerstenau M.C., Miller J.D., 1997. Amine flotation of soluble sulfate salts, Proc. XX
Inter. Mineral Processing Congress, Vol. 3, 715–721, GMDB, Clausthal-Zeller- feld, Germany.
12. Flotation 391
Hanning R.N., Rutter P.R., 1989. A simple method of determining contact angles on particles and their rele-
vance to flotation, Int. J. Miner. Process., 27: 133–146.
He Y.B., Laskowski J.S., 1992. Contact angle measurements on discs compressed from fine coal, Coal
Preparation, 10: 19–36.
Heertjes P.M., Kossen N.W.F., 1967. Measuring the contact angles of powder-liquid systems, Powder Tech-
nology, 1(1), 33–42.
Hiemenz P.C., 1986. Principles of Colloid and Surface Chemistry, Dekker, New York.
Hornsby D.T., Leja J., 1980. Critical surface tension and the selective separation of inherently hydrophobic
solids, Colloids and Surfaces, 1, 425–429.
Hornsby D.T., Leja J., 1983. Critical surface tension of floatability, Colloids and Surfaces, 7(4), 339–348.
Hrycyna E., 1999. Wyniesienie mechaniczne w laboratoryjnej maszynce flotacyjnej typu Mechanobr, M.Sc.
thesis, Wydział Górniczy Politechniki Wrocławskiej, Wrocław, Poland.
Huethorst J.A.M. and Leenaars A.F.M., 1990. A new method for determining the contact angle of liquid
against solid spherical particles, Colloids and Surfaces, 50: 101–111.
Hunter R.J., 1987. Foundation of colloid science, Vol. 1, Clarendon Press, Oxford.
Iwasaki I., Cooke S.R.B., Colombo A.F., 1960. Report of investigation no. 5593, U.S. Bureau of Mines,
USA.
Jakuszewski B., 1962. Współczesne zagadnienia elektrochemii teoretycznej, PWN, Warszawa.
Janczuk B., Chibowski E., Bialotiotrowicz T., 1986. Time dependence wettability of quartz with aqueous,
Chem. Papers, 40(3), 349–356.
Janusz W., 1999. Electrical Double layer at the metal oxide-electrolyte interface, in: Interfacial Forces and
Fiels: Theory and Applications, J.P. Hsu (ed.), Surfactants Science Series, 85, Dekker, New York,
Chap. 4, 135–206.
Jaycock M.J., Parfitt G.D., 1981. Chemistry of interfaces, Ellis Horwood (Wiley), New York.
Johson R.E. Jr., Dettre R.H., 1989. An evaluation of Neumann’s „Surface equation of state”, Langmuir, 5,
293–295.
Kelebek S., 1987. Wetting behavior, polar characteristics and flotation of inherently hydrophobic minerals,
Trans. IMM, Sec. C., 96, C103–C106.
Kitchener J.A., 1992. Minerals and surfaces, in: Colloid chemistry in mineral processing, J.S. Laskowski i J.
Ralston (eds.), D.W. Fuerstenau (ed.) Development in Mineral Processing, Vol. 12, 1–35.
Klassen V.I., Mokrousov V.A., 1963. An introduction to the theory of flotation, translation by J. Leja and
G.W. Poling, Butterworths, London.
Klimpel R.R., Hansen R.D., 1988. Frothers, in: Reagents in Mineral Technology, Somasundaran and B.M.
Moudgil (eds.), Surfactants Science Series, Vol. 27, Dekker, New York, 385–409.
Konopacka 2005, Flotacja solna, Oficyna Wydawnicza PWr.
Koopal L.K., 1992. Adsorption, in: Colloid chemistry in mineral processing, J. Laskowski, J. Ralston (eds.),
Elsevier, 37–96.
Kosmulski M., 2001. Chemical properties of material surfaces, Surfactant Science Series, v.102, Dekker,
New York
Kwok D.Y., 1999. The usefulness of the Lifshitz–van der Waals/acid–base approach for surface tension
components and interfacial tensions, Colloid and Surfaces. A: Physicochemical and Engineering
Aspects, 156, 191–200.
Laskowski J., 1966. Flotacja minerałów o naturalnej hydrofobowości w roztworach z podwyższonym stęże-
niem soli nieorganicznych, Zeszyty Naukowe Politechniki Śląskiej, Górnictwo, Gliwice 1966, no.16.
Laskowski J., 1969. Chemia fizyczna w procesach mechanicznej przeróbki kopalin, Śląsk, Katowice.
Laskowski J., 1981. Redox conditions in flotation: oxides, in: Interfacial phenomena in mineral processing,
B. Yarar i D.J. Spottiswood (eds.), Proc. Eng. Found. Conf., Rindge, New Hampshire.
Laskowski J., 1983. Dextrin adsorption by oxidized coal, Colloids and Surfaces, 8, 137–151.
Laskowski J., 1986. The relationship between floatability and hydrophobicity, in: Advances in mineral proc-
essing, P. Somasundaran (ed.), SME Littleton, Colorado, USA, 189–208.
Laskowski J., 1994. Flotation of potash ores, in: Reagents for better metallurgy, P.S. Mulukutla (ed.) SME.,
Littleton, 225–243.
Laskowski J., 1998. Frothers and frothing, in: Frothing in Flotation II, J.S. Laskowski, E.T. Woodburn
(eds.), Gordon and Breach, Australia, 1–49.
Laskowski J., 2004, Testing flotation frothers, Physicochemical Problems of Mineral Processing, 38, 13-22
Laskowski J., Fuerstenau D.W., Gonzales G., Urbina R.H., 1985. Studies on the flotation of chrysocolla,
Mineral processing and technology review, 1985, Vol. 2, 135–155.
Laskowski J., Iskra J., 1970. Role of capillary effect in bubble particle collision in flotation Trans. IMM,
Sec. C., Vol. 79. C6–C10.
Laskowski J., Luszczkiewicz A., 1989, Przeróbka kopalin. Wzbogacanie surowców mineralnych, Wydaw-
nictwo Politechniki Wrocławskiej, Wrocław.
Laskowski J., Miller J.D., 1984. New reagent in coal flotation, in: M.J. Jone and R. Oblatt (eds.), Reagent in
the minerals industry, IMM, London, 145–154.
Laskowski J.,1988. Weak electrolyte type collectors, in: Mineral processing and process control, COBRE 87
– International Conference, Vina del Mar, A.L. Mular, G. Gonzalez, C. Barahona (eds.), University of
Chile, Vol. 2, 137.
Laskowski J.S., Vuredela R.M., Liu Q., 1988. The colloid chemistry of weak-electrolyte collector flotation,
Proc. XVI Inter. Min. Proc. Congr., E. Forsberg (ed.), Elsevier, Amsterdam, 703–715.
Laskowski J.S., Woodburn E.T. (eds.), 1998. Frothing in flotation II, Gordon and Breach, Australia.
Laskowski J.S., Yordan J.L., Yoon R.H., 1989. Electrokinetic potential of microbubbles generated in aque-
ous solutions of weak electrolyte type surfactants, Langmuir, 5, 373–376.
Lee L.-H., 1996. Correlation between Lewis acid – base surface interaction components and linear solvation
energy relationship solvatochromic α and β parameters, Langmuir, 12, 1681–1687.
Leja J., 1982. Surface chemistry of froth flotation, Plenum Press, New York.
Lekki J., 1970, unpublished
Lekki J., 1979. Fizykochemiczne podstawy flotowalności minerałów siarczkowych, Prace Naukowe Instytu-
tu Chemii Nieorganicznej 41, Seria Monografie 16, Wydawnictwo Politechniki Wrocławskiej, 1–112.
Lekki J., 1984. Korelacja oleinianowej flotacji minerałów Fe(II), Al(III) oraz Mg(II) z fazami układu tlenek–
oleinian–woda, materiały seminarium, Baranów Sandomierski, 243–259.
Lekki J., Drzymala J., 1990. Flotometric analysis of the collectorless flotation of sulfide materials, Colloids
and Surfaces, 44, 179–190.
Letowski F., 1975. Podstawy hydrometalurgii, PWN, Warszawa.
Li C., Somasundaran P., 1993. Role of electrical double layer forces and hydrophobicity in coal flotation in
NaCl solutions, Energy and Fuels, 7, 244–248.
Li C., Somasundaran P., Harris C.C., 1993. A levitation technique for determining particle hydrophobicity,
Colloids and Surfaces Colloids and Surfaces, A 70: 229–232.
Lukaszewska I., 1998. Flotacja w celce Hallimonda, M.Sc. thesis, Wydział Górniczy Politechniki Wrocław-
skiej, Wroclaw, Poland
Luszczkiewicz A., Lekki J., Laskowski J., 1979. Floatability of ilmenite, Proc. XIIIth Inter. Min. Proc. Con-
gr., Warszawa, Round Table Seminar, 2, 163–182.
12. Flotation 393
Luszczkiewicz, 2004, unpublished.

Lyklema J., 1972. The Esin and Markov coefficient for double layer with colloid chemical importance, Elec-
troanal. Chem. Interfacial Electrochem., 37, 53–60.
Lyklema J., 1999. The surface tension of pure liquids. Thermodynamic components and corresponding
states, Colloid and Surfaces. A: Physicochemical and Engineering Aspects, 156, 413–421.
Lynch A.J., Johnson N.W., Manlapig E.V., Thorne C.G., 1981. Mineral and coal flotation circuits, Chap. 3:
Mathematical models of flotation, in: D.W. Fuerstenau (ed.), Development in Mineral Processing, No.
3, Elsevier, Amsterdam.
Manser R.M., 1975. Handbook of silicate flotation, Warren Spring Laboratory, England.
Mao L., Yoon R.H., 1997. Predicting flotation rates using equation derived from first principles, Interna-
tional Journal of Mineral Processing, 51, 171 – 181.
Marinakis K.I., Shergold H.L., 1985. The mechanism of fatty acid adsorption in the presence of fluorite,
calcite and barite, Int. J. Mineral Process., 14, 161–176.
Michot L.J., Villieras F., Francois M., Yvon J., Le Dred R., Cases J.M., 1994. The structural microscopic
hydrophilicity of talc, Langmuir, 10, 3765–3773.
Miller J.D., Lin C.L., Chang S.S., 1984. Coadsorption phenomena in the flotation of pyrite from coal by
reverse flotation, Coal Preparation, 1, 21–32.
Miller J.D., Yalamanchili M.R., Kellar J.J., 1992. Surface charge of alkali halide particles as determined by
laser-Doppler electrophoresis, Langmuir, 8, 1464–1469.
Mishra S.K., 1988. Anionic collectors in nonsulfide mineral flotation, in: Reagents in Mineral Technology,
Surfactant Science Series 27, P. Somasundaran i B.M. Moudgil (eds.), Dekker, New York, 195–217.
Nagaraj D.R., 1988. The chemistry and application of chelating or complexing agents in mineral separations,
in: Reagents in Mineral Technology, Surfactant science series Vol. 27, P. Somasundaran i B.M.
Moudgil (eds.), Dekker, New York, 257–334.
Nagy E., Van Cleave A.B., 1962. A study of the characteristics of galena, Can. J. Chem. Eng., 40(2), 776–
781.
Nakamura Y., Kamada K., Katoh Y., Watanabe A., 1973. Studies on secondary electrocapillary effects. I –
The confirmation of Young-Dupre equation, J. Colloid Interface Sci., 44, 517–524.
Naumov G.B., Ryżenko B.N., Chodakowski J.L., 1971. Handbook, Atomizdat, Moscow, in Russian.
Neumann A.W., 1974. Contact angles and their temperature dependence. Thermodynamic status, measure-
ments, interpretation, and application, Advances in Colloid and Interface Sci., 4(2–3), 105–191.
Neumann A.W., Good R.J., 1979. Techniques of measuring contact angles, in: Surface and Colloid Science,
Vol. 11, R.J. Good i R.R. Stomberg (eds.), Plenum Press, New York, 31–91.
Ney P., 1973. Zeta-Poteniale und Flotierbarkeit von Mineralen, Springer-Verlag, Wien–New York.
Nguyen A.V., Schulze H.J., Ralston J., 1997. Elementary steps in particle-bubble attachment, Inter. J. Min-
eral. Process., 51, 183–195.
Ozbayoglu G., 1987. Coal flotation, in: Mineral processing design, B. Yarar and Z.M. Gogan (eds.), 77–105,
Martinus Nijhoff Publisher, Dordrecht.
Pajdowski L., 1993. Chemia ogólna, PWN, Warszawa.
Parfitt R.L., Russell J.D., 1977. Adsorption on hydrous oxides. V – Mechanisms of adsorption of various
ions on goethite, J. Soil Sci., 28(2), 297–305.
Parks, G.A., 1965. Isoelectric points of solid oxides, solid hydroxides, and aqueous complex systems,
Chemical Reviews, 65(2), 177–198.
Parsonage P., Watson D., Hickey T.J., 1982. Surface texture, slime coating and flotation of some industrial
minerals, XIV IMPC, CIM Toronto, V-5.1–V-5.22.
Partridge A. C., Smith G.W. 1971. Small-sample flotation testing a new cell. Trans. Inst. Min. Metall., Sec.
C:80: C199–C203.
Paulson O., Pugh R.J., 1996. Flotation of inherently hydrophobic particle in aqueous solutions of inorganic
electrolytes, Langmuir 12, 4808-4813.
Perucca C.F., Cormode D.A., 1999. The use of polymers in potash beneficiation at Agrium Potash, in:
Polymers in mineral processing, J. Laskowski (ed.), Proc. 3rd UBC-McGill Bi-annual Inter. Symp. on
fundamentals of mineral processing, Met. Soc. CIM Montreal, 393–404.
Pierzak J., Drzymala J., 1998. Alotropowe odmiany węgla pierwiastkowego, Prace Naukowe Instytutu Gór-
nictwa Politechniki Wrocławskiej, nr 85, Studia i Materiały nr 27, 63–68, Wrocław.
Pockels A., 1914, Phys. Z., 15, 39.
Polkin S.I., Najfonow T.V., 1965. Concerning the mechanism of collector and regulator interaction in the
flotation of silicate and oxide minerals, VII IMPC, Vol. 1., N. Arbiter (ed.) Gordon and Breach, New
York, 307–318.
Pourbaix M., 1963. Atlas d’equilibres electrochimiques a 25 °C, Gauthier-Villars, Paris.
Pradel K., 2000. Wykorzystanie efektów kinetycznych do wzbogacania surowców fluorytowo-barytowych,
Ph.D. thesis, Instytut Górnictwa Politechniki Wrocławskiej, Wrocław.
Pugh R.J., 1986. The role of the solution chemistry of dedecyloamine and oleic acid collectors in the flota-
tion of fluorite, Colloids and Surfaces, 18, 19–41.
Pugh R.J., Weissenborn P., Paulson O., 1997. Flotation in inorganic electrolytes, Int. J. Miner. Process., 51,
125–138.
Purcell G., Sun S.C., 1963. Significance of double bonds in fatty acids flotation – an electrocinetic study,
Trans. AIME, 226, 6–16.
Rönngren L., Sjöberg S., Sun Z.-X., 1995. Complexation at the sulfide aqueous interface, in: Flotation Sci-
ence and Engineering, K.A. Matis (ed.), Dekker Mew York.
Rönngren L., Sjöberg S., Sun. Z., Forsling W., 1994. Surface reaction in aqueous metal sulfide systems. 5.
Surface complexation of sulfide ions, J. Colloid Interface Sci., 162: 227–235.
Ralston J., 1992. The influence of particle size and contact angle in flotation, in: Colloid Chemistry in Min-
eral Processing, J.S. Laskowski i J. Ralston (eds.), Development in Mineral Processing, D.W. Fuer-
stenau (ed.), Vol. 12, Chap. 6, 203–224.
Ralston J., Newcombe G., 1992. Static and dynamic contact angles in flotation, in: Colloid Chemistry in
Mineral Processing, J.S. Laskowski i J. Ralston (eds.), Development in Mineral Processing, D.W. Fuer-
stenau (red. serii), Vol. 12, Chap. 6, 173–202.
Rao K.H., Forssberg E., 1991. Mechanism of oleate interactions on salt-type minerals, Part III. Adsorption,
zeta potential, and diffuse reflectance FT-IR studies of schelite in the presence of sodium oleate, Col-
loids and Surfaces, 54(1–2), 161–187.
Ratajczak T., 2001. Mechanizm flotowalności ziarn mineralnych w roztworach soli nieorganicznych, Poli-
technika Wrocławska, Wrocław, Ph.D. thesis, Wroclaw, Poland.
Richardson P.E., 1995. Surface chemistry of sulfide flotation, in: Mineral surfaces, D.J. Vaughan and R.A.D.
Pattrick (eds.), Mineralogical Society Series, No. 5, Chapman and Hall, 261–302.
Richardson P.E., Walker G.W., 1985. The flotation of chalcocite, bornite, chalcopyrite, and pyrite in an elec-
trochemical flotation cell, XVth Inter. Min. Proc. Congr., Vol. 2, 198–210, Soc. Ind. Min. – Bur. Res.
Geol. Min., Cannes.
Rogers J., 1962. Principles of sulfide mineral flotation, in: Froth flotation, 50th anniversary volume, D.W.
Fuerstenau (ed.), AIMM, New York, Chap. 6, 139–169.
12. Flotation 395
Ross V.E., 1991. Comparison of methods for evaluation of the flotation and entrainment, Trans. IMM, Sec-
tion C., 100, C121–126.
Rubinstein J.B., Melik-Gaikazyan V.I., 1998. Frothing in flotation II, Chap. 5: Charakterization of flotation
froth, Edited by J.S. Laskowski and E.T. Woodburn, Gordon and Breach Science Publishers, Australia.
Rubinstein J.B., Samygin V.D., 1998. Effect of particle and bubble size on flotation kinetics, in: Frothing in
flotation II, J.S. Laskowski i E.T. Woodburn (eds.), Gordon and Breach, Australia, Chap. 2, 51–80.
Söhnel O., Garside J., 1992. Precipitation and industrial applications, Butterworth, Oxford.
Sablik J., 1998. Flotacja węgli kamiennych, GIG, Katowice.
Schmidt E.W., 1983. Hydrazine and its derivatives, Wiley, New York.
Schrader M.E., 1975. Ultrahigh vacuum techniques in the measurements of contact angle, IV. Aqueous on
graphite (0001), J. Phys Chem., 79, 2508–2515.
Schulze H.J., 1984. Physico-chemical elementary processes in flotation, Developments in Mineral Process-
ing, 4, D.W. Fuerstenau (ed.), Elsevier, Amsterdam, 1984.
Schulze H.J., 1992. Elements of physically – based modelling of the flotation process, Innovation in Flota-
tion Technology, Vol. 208, Kluwer Academic Publishers, London.
Schulze H.J., 1993. Flotation as a heterocoagulation process: possibilities of calculating the probability of
flotation, in: Coagulation and flocculation. Theory and applications, Marcel Dekker Inc., New York.
Sillen L.G., Martell A.E., 1964. Stability constants of metal–ion complexes, Chem. Soc., Burlington House,
London.
Siwek B., Zembala M., Pomianowski A., 1981. A method for determining fine particle floatabilty, Int. J.
Miner. Proces., 8, 85–88.
Smith R.W., 1988. Cationic and amphoteric collectors, in: Reagents in Mineral Technology, Surfactant sci-
ence series, Vol. 27, P. Somasundaran (ed.), 219–256.
Smolders C.A., 1961. Contact angles, wetting and dewetting of mercury, praca doktorska, University of
Ultrecht, Holandia.
Sobieraj S., 1985. Badanie procesu flotacji krajowych rud barytowych, Zeszyty Naukowe Politechniki Ślą-
skiej, 135, 27–45.
Somasundaran P., Ananthapadmanabhan K.P., 1986. In: Advances in mineral processing. P. Somasundaran
(ed.), Society of Mining Eng. of AIME, Colorado, 1986, 8.
Sorokin M.M., 1976. Processing of mineral row materials, Izdat. Nauka, Moskva, in Russian.
Soto H., Laskowski J., 1973. Redox conditions in the flotation of malachite with a sulphidizing agent, Trans.
IMM., 82, C153–C158.
Spelt J.K., Li D., 1996. The equation of state approach to interfacial tensions, in: Applied surface thermody-
namics, A.W. Neumann i J.K. Spelt (eds.), Surfactant Science Series, 63, Dekker, New York, 239–292.
Stachurski J., Michałek M., 1986. Zjawiska elektrokinetyczne w emulsyjnej flotacji różnych typów węgli,
referat VIII seminarium naukowo-przemysłowego pt. „Flotacja węgla”, Katowice, GIG, 96–104.
Stumm W., Morgan J.J., 1970. Aquatic chemistry, Wiley, New York.
Sutherland K.L., 1948. Kinetics of the flotation process, J. Phys. Chem. 52, 394–425.
Sutherland K.L., Wark I.W., 1955. Principles of flotation, Australasian Institute of Mining and Metallurgy,
Melbourne.
Szczerba E., Komorowski J., Skorupska B., 1999. Nowa konstrukcja maszyn flotacyjnych do flotacji grubo-
ziarnistych zawiesin o dużym zagęszczeniu, V międzynarodowa konferencja „Przeróbka Rud Metali
Nieżelaznych” ICNOP 99, Instytut Metali Nieżelaznych, KGHM, CBPM Cuprum, 45–58.
Takamori T., Tsunekawa M., 1982. Separation of calcite from fluorite ore by the adsorption – washing –
flotation method, XIV IMPC, CIM Toronto, V-1.1–V-1.12.
Tucker J.P., Deglon D.A., Franzidis J.-P., Harris M.C., O’Connor C.T., 1994. An evaluation of direct
method of bubble size distribution measurements in a laboratory batch flotation cell, Minerals Eng., 7,
667.
Usui S., Sasaki H., 1978. Zeta potential measurements of bubbles in aqueous surfactant solutions, J. Colloid
Interface Sci., 65, 36–45.
van Oss C.J., Giese R.F., Li Z., Murphy K., Norris J., Chaudhury M.K. and Good R.J., 1992. Determination
of contact angle and pore sizes of porous media by column and thin layer wicking, J. Adhesion Sci.
Technol., 6, 413–445.
van Oss C.J., Giese R.F., Wenzek R., Norris J., Chuvilin J., 1992. Surface tension parameters of ice obtained
from contact angle data from positive and negative particles adhesion to advancing freezing fronts, Ad-
hesion Sci. Technol., 6(4), 503–516.
van Oss C.J., Good R.J., 1996. Hydrogen bonding, interfacial tension and the aqueous solubility of organic
compounds, J. Dispersion Sci. Tech., 17(4), 433–449.
van Oss C.J., Good R.J., Chaudhury M.K., 1988. Additive and nonadditive surface tension components and
the interpretation of contact angles, Langmuir, 4, 884–891.
Varbanov R., Forssberg E., Hallin M., 1993. On the modelling of the flotation process, International Journal
of Mineral Processing, 37, 27–43, Elsevier, Amsterdam.
Volke K., Para G., Pawlikowska-Czubak J., Neubert H., 1984. Flotation and adsorption investigation of the
system n-dodecanoic acid/mercury at different pH values, Coll. Polymer Sci., 262, 245–251.
Waksmundzki A., Neczaj-Hruziewicz J., Ptanik M., 1972. Mechanism of carryover of gangue slimes during
flotation of sulfur ore, Trans. IMM, Section C., 81, C249–251.
Wark I., Cox A.B., 1934. Principles of flotation, Trans. AIMME, 112, part I: 189–302, part III, 267–302.
Washburn, E., 1921. The dynamics of capillary flow, Phys. Rev., 17, 273–283.
Watson D., Manser R.M., 1968. Some factors affecting the limiting conditions in cationic flotation of sili-
cates, Trans. IMM, 77, C57–60.
Weissenborn P.K., Pugh R.J., 1996. Surface tension of aqueous solution of electrolytes: relationship with ion
hydration, oxygen solubility, and bubble coalescence, J. Colloid Interface Sci., 184, 550–563.
Wells A.F., 1993. Strukturalna chemia nieorganiczna, WNT, Warszawa.
Wert C.A., Weller M., 1982. Polymeric character of coal, J. Appl. Phys., 53 (10), 6505–6512.
White L.R., 1982. Capillary rise in powders, J. Colloid Interface Sci., 90: 536–538
Woods L.T., Garrels R.M., 1987. Thermodynamic values at low temperatures for natural inorganic materi-
als, Oxford University Press.
Woods R., 1988, Flotation of sulfide minerals, in: Reagents in Mineral Technology, Surfactant science series
Vol. 27, P. Somasundaran, B.M. Moudgil (eds.), 39–78.
Yamamoto T., 1980. Complex Sulfide Ores, M.J. Jones (ed.), IMM London, 71.
Yap S.N.Y., Mishra R.K., Raghavan S., Fuerstenau D.W., 1981. The adsorption of oleate from aqueous so-
lution onto hematite, in: Adsorption from aqueous solutions, P.H. Tewari (ed.), Plenum Press, New
York, 119–142.
Yarar B., Alvarez J., 1988. Separation of sulfur minerals by the „Gamma flotation” process, XVI Int. Min.
Proc. Congr., E. Forssberg (ed.), 547–562, Elsevier, Amsterdam.
Yarar B., Kaoma J., 1984. Estimation of the critical surface tension of wetting of hydrophobic solids by flo-
tation, Colloids and Surfaces, 11, 429–436.
Yarar B., Leja J., 1982. Flotation of weathered coal fines from Western Canada, IX. Int. Coal Prep. Con-
gress, New Delhi, paper C-5.
12. Flotation 397
Ye Y., Khandrika S.M., Miller J.D., 1989. Induction time measurements at a particle bed, Int. J. Miner. Proc.
25, 221–240.
Ye Y., Miller J.D., 1988. Bubble/particle contact time in the analysis of coal flotation, Coal Preparation, 5,
147–166.
Yoon R.H., 1982. Flotation of coal using microbubbles and inorganic salts, Mining Congress J., 68, 76–80.
Yoon R.H., Yordan J.L., 1986. Zeta potential measurements on microbubbles generated using various sur-
factants, 1986, J. Colloid Interface Sci., 113, 430–438.
Yordan J.L., Yoon R.H., 1986. Induction time measurements for the quartz–amine flotation systems, Annual
SME Meeting New Orleans 1986, preprint 86–105, 1–10.
Young T., 1805, Philos.Trans., R. Soc., London, 95, 65
Zisman W.A., 1964. Relation of the equilibrium contact angle to liquid and solid costitution. Contact angle,
wettability and adhesion, ACS Advances in Chem. Series, Washington D.C., 43.
Zyla M., 1994. Chemiczny charakter powierzchni węgla kamiennego, w: Węgiel kamienny, A. Czapliński
(red.), Wyd. AGH, Krakow.

271 - PDFsam - Drzymala - Mineral Processing Foundations of Theory and Practice

Uploaded by

Copyright:

Available Formats

271 - PDFsam - Drzymala - Mineral Processing Foundations of Theory and Practice

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

271 - PDFsam - Drzymala - Mineral Processing Foundations of Theory and Practice

Uploaded by

Copyright:

Available Formats

12.

12.1. Theoretical basis

gas bubble gasgas

Fig. 12.1. Flotation Fig. 12.2. Contact angle in flotation

Contact angles measured on polished plates made of different naturally hydropho-

The simplest method of contact angle determination relies on a direct measurement

Table 12.3. Methods of contact angle measurements

γsg = γsc+ γcgcosθ

Fig. 12.4. Graphical representation of the Young equation

Table 12.5. Surface energy of selected crystalline materials.

12.2. Hydrophobicity modification

face. The effect of modifiers (activators, depressors, frothers) on properties of inter-

70 concentration of collector Data by

contact angle, degree

H2O Smolders (1961) taken from various

surface tension, γ wg, mN/m

Fig. 12.8. Influence of pH on germanium flotation

If salt is applied as the electrolyte, hydrophobicity and flotation decrease or in-

initial contact angle electrocapillary curve

d – in practice (type of flotation)

initial salt flotation and

12.3. Electrical phenomena at interfaces

surface potential, Ψ 0 cation concentration

zeta potential, ζ eg. [H+] = [H+] r exp (−Ψ /RT)

eg. [OH-] = [OH-] r exp (Ψ /RT)

σo -σo σo σd -----→ σo σi σd-----→ σo σi σd----→ σo σi σj σd→

the surface potential (in volts).

place of particle breakage particle /water interface

A different electrical double layer is formed on particles which do not conduct

Fig. 12.15. Diagram of dominance of surface species on zinc sulfide (sphalerite)

-40 0.010M NaCl

pAg iep = 5.3

contact angle θ , degree

γsc = γ sc0 + Fscedl . (12.15)

where pX denotes negative logarithm of X ions concentration (pX=-log[X]).

Ideally polarized body electrons mercury–water F edl

relative surface tension, γ /γ H20 1.00020

air ( ), Nocardia sp.( )

12.4. Delineation of flotation

γ sc,γ sg,γ cg,π

device mode of work

However, flotation is a dynamic, time dependent, process. Very often we want to

Table 12.13. Models of flotation

dN/dt – flotation rate,

1.50 log (induction time, τ i, ms)

The stability of the particle–bubble aggregates is described by the DLVO theory

The diagram presenting the dependence of interaction energy E between bubble

flotation energy barrier

– barrier for a stable

Table 12.14. Models applied for kinetics of flotation

* Equivalent models because k3 = 1/k4.

12.5. Flotation reagents

delineation c chemistry of flotation

Fig. 12.28. Chemistry plays an important role in flotation

N-N diamines anionic merkaptans

Fig. 12.30. Possible modes of adsorption of collectors at particle-water interface:

Table 12.15. Nonionic collectors

Table 12.16. Ionic collectors

Diamines and triamines diamine R–NH–(CH2)x–NH2