SELECTIVE FLOTATION OF DOLOMITE FROM APATITE
USING SODIUM OLEATE AS THE COLLECTOR
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN
PARTIAL FULFILLMENT OF THE REQUIREMENTS
FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
To my family
The author wishes to express his sincerest gratitude to
Dr. B. M. Moudgil for his invaluable guidance in this study.
The author is also very grateful to Dr. E. Dow Whitney
and Dr. D. O. Shah for their encouragement and constructive
Appreciation is due also to Dr. G. Y. Onoda, Dr. F. N.
Blanchard,and Dr. C. D. Batich for the helpful discussions;
to Mr. Rob Pelick for his assistance in measuring the pore
size distribution of the minerals; to Mr. Richard Crockett
for his assistance at various stages of the experiments.
The author is very grateful to the University of
Florida and the Florida Institute of Phosphate Research
(through grant no. 82-02-023) for providing the financial
support of this study. Any opinions, findings, conclusions
or recommendations expressed in this publication are those
of the author and do not necessarily reflect the views of
the Florida Institute of Phosphate Research.
The author also wishes to acknowledge International
Minerals & Chemicals Corp., Bartow, Florida for providing
the mineral samples and Carpco, Inc., Jacksonville, Florida
for their help in preparing the apatite sample.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS.... ..................... .......... iii
ABSTRACT .......... .... .... ............ ... ... ... vi
I. INTRODUCTION................................... 1
1.1. Statement of the Problem................ 1
1.2. Objective of this Study.................. 9
II. LITERATURE REVIEW .......... ................... 11
2.1. Characteristics of the Minerals.'......... 11
2.2. Past Investigations of Apatite/Dolomite
2.3. Basic Principles......................... 22
III. EXPERIMENTAL PLAN............................. 45
3.1. Selection of the Separation Technique.... 45
3.2. Selection of Minerals and
Preparation Methods...................... 46
3.3. Outline of the Investigation............. 46
3.4. Selection of the Experimental Techniques. 48
3.5. Selection of Experimental Variables...... 50
IV. MATERIALS AND METHODS.......................... 52
4.1. Materials................................ 52
4.2. Methods.................................. 73
V. RESULTS........................................ 98
5.1. Flotation Studies..................... 98
5.2. Oleate Adsorption Studies................ 135
5.3. Zeta Potential Measurements............. 149
5.4. Infrared Spectroscopy Studies........... 159
5.5. Bubble Size Measurements................ 162
5.6. Solubility of the Minerals............... 169
5.7. Depletion of Oleate Ions in the presence
Calcium and Magnesium Cations............ 174
VI. DISCUSSION...................................... 177
6.1. Flotation after Conventional Conditioning 178
6.2. Flotation after Two-Stage Conditioning... 217
VII. SUGGESTIONS FOR FUTURE WORK.................... 239
VIII. CONCLUSIONS...... .............................. 243
8.1. Flotation after Conventional Conditioning 243
8.2. Flotation after Two-Stage Conditioning... 244
LIST OF REFERENCES............ ....................... 246
APPENDIX 1 DETERMINATION OF OLEATE CONTENT IN
AQUEOUS SOLUTION BY GREGORY'S METHOD..... 257
APPENDIX 2 DETERMINATION OF FLOTATION RECOVERY OF
SINGLE MINERALS.......................... 259
APPENDIX 3 DETERMINATION OF FLOTATION RECOVERY OF
MIXED MINERALS......................... 260
BIOGRAPHICAL SKETCH................... ............ 262
Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
SELECTIVE FLOTATION OF DOLOMITE FROM APATITE
USING SODIUM OLEATE AS THE COLLECTOR
Chairman: Dr. E. Dow Whitney
Major Department: Materials Science and Engineering
The major problem with the future utilization of
phosphate deposits in South Florida is the dolomite
impurity, which interferes with the subsequent processing of
the phosphate mineral, apatite, into fertilizers. The
purpose of this study is to achieve and understand the
separation of dolomite from apatite by selective flotation.
Flotation behavior of single and mt tests in the pH
range of 7 to 11, was not observed in mixed mineral systems.
To understand the loss of selectivity in the mixed minerals,
electrokinetic, oleate adsorption and
solubility studies were conducted. The loss of selectivity
in the mixed mineral system at pH 7 to 10 is attributed to
the depletion of oleate by precipitation with cations,
dissolved mainly from dolomite. At pH 11, the selectivity
in mineral mixtures is lost due to possible modification of
the apatite surface when in contact with dolomite.
A new technique of two-stage conditioning prior to
flotation has been developed. This method involves
conditioning the feed at pH 10 with sodium oleate followed
by reconditioning at a lower pH. Selective flotation of
dolomite from apatite was observed both for single and mixed
minerals by reconditioning at pH lower than 4.5. To
understand the mechanisms of observed selective flotation,
further studies involving oleate adsorption, infra-red
spectroscopy, and solubility of the minerals were conducted.
Selective flotation of dolomite after two-stage conditioning
is attributed to the combined effect of higher oleate
adsorption on dolomite and the hydrolysis of the adsorbed
oleate species to oleic acid at lower pH values. This
technique has a potential for commercial application in the
flotation scheme currently practiced by the industry without
any additional reagent cost.
1.1. Statement of the Problem
Improved agricultural techniques in industrial nations
as well as in third world countries have led to a steady
growth in the production of fertilizers. To keep up with
this demand, world production of phosphate rock, 95 percent
of which is converted to fertilizers, also has steadily
increased. In 1952, the world production of phosphate rock
was 22.4 mm tons, as compared with 119.6 mm tons in 1982. In
the year 2000 the demand for the rock is expected to be 296.9
In the past, about 30 to 50 percent of the world supply
of phosphate rock came from the United States and 80 percent
of the U.S. production was contributed by the Florida
phosphate industry. The demand and supply for U.S. phosphate
rock from 1979 and the trend for next two decades are shown
in Table 1. The foreseeable deficit of phosphate minerals
after about a decade has led to the development of known
resources by new technology.
Until the present, the production of phosphate rock in
Florida has been from the Bone Valley formation. The
technology for processing the rock from this deposit has been
Supply and demand for
Year Supply Demand Deficit Surplus
1979 51.3 53.7 2.4
1980 54.7 55.0 0.3
1981 58.1 56.4 1.7
1982 60.9 57.8 3.1
1983 70.0 59.3 10.7
1984 72.1 60.8 11.3
1985 75.9 63.7 12.2
1986 82.8 63.9 18.9
1987 83.1 65.5 17.6
1988 85.0 67.3 17.7
1989 83.5 69.0 14.5
1990 84.1 70.7 13.4
1991 80.5 72.4 8.1
1992 77.7 74.3 3.4
1993 70.6 76.2 5.6
1994 71.5 78.2 6.7
1995 72.0 80.3 8.9
1996 68.8 82.4 13.6
1997 64.5 84.5 20.0
1998 66.3 86.8 20.5
1999 65.6 89.0 23.4
2000 63.3 91.3 28.0
Source: U. S. Bureau of Mine
A Supply-Demand Anal
May 19, 1981
ysis of the U. S. Phosphate
phosphate rock in United
adequately developed to obtain a suitable product required
for the manufacture of the fertilizers. The pebble product,
-20,+1 mm, is produced by simply desliming and sizing and the
concentrate is obtained by a double stage froth flotation
circuit(2-4) When these deposits are depleted, the
replacement production will come from the southern extension
of the central district shown in Figure 1(5-12).
The future deposits in the southern extension consist of
two distinct geological ore types often referred to as upper
and lower zone. As compared to the currently processed ores
from the Bone Valley, the yield and grade of the phosphate
rock from south Florida are significantly lower. If these
ores were processed by conventional techniques, products
having the specifications shown in Table 2 would be obtained.
The products from the lower zone would be inferior in grade
to that of the upper zone.
The major problem in the production of fertilizers from
phosphate rock from the southern extension is the presence of
dolomitic (Ca Mg carbonate) impurities, usually reported as
weight percent MgO. The MgO content of the currently
produced phosphate rock is less than 0.5 percent. By
comparison, the deposits from the South Florida district
would yield phosphate products ranging from 0.2 to 25 percent
in MgO if processed by the conventional techniques(13-19).
Greater dolomite impurity in the rock is undesirable because
it interferes with the chemical processing as discussed
Phosphate deposits in the current mining
district in Central Florida and its Southern
extension [after Bernardi and Hall(5)].
Table 2. Comparison of the chemical analyses of the
phosphate rock from Central Florida and its
Southern Extension Central Florida
Upper Zone Lower Zone Bone Valley
reference (18) (18) (5)
Pebble (+16 mesh)
BPL 20 77 5 70 67
mesh) BPL 10 48 10 44 27
+150 mesh) BPL 55 80 54 73 73
% MgO-Pebble 0.2 10 0.5 25 0.4
% MgO-Concentrate 0.2 4 0.2 13 0.3
% Iron & Aluminum
Oxide Pebble 1.4 8 1 8 2.5
% Iron & Aluminum
Oxide-Concentrate 1.3 3.5 1.3 3.5 2.3
% Insol Pebble 2 50 3 70 7.6
A. Phosphoric acid is an intermediate compound during
the chemical processing of the phosphate rock into
fertilizers. The viscosity of this acid
exponentially increases(22) with percent MgO as
illustrated for three grades of phosphoric acid in
Figure 2. The higher viscosity results in higher
pumping costs and in certain cases it could make
B. Magnesium ions also precipitate complex salts that
would plug the process filters used to separate the
calcium sulfate by-product.
C. The maximum possible grade of fertilizer is also
downgraded in relation to the amount of dolomite.
D. During the production of phosphoric acid, the
carbonates in the rock would react with sulfuric
acid without yielding any useful product, thus
increasing the consumption of sulfuric acid.
It is widely accepted that the tolerance limit of MgO in
phosphate rock for fertilizer production is 1 percent. If
economical methods are not developed to eliminate dolomitic
impurities from the phosphate resources in Florida, it is
estimated that approximately 2 billion tons(23) of phosphate
rock cannot be mined. As a result, the phosphate rock would
not only be costly, but to maintain the same level of
production larger acreage have to be disturbed.
P. 104.4% H-304
75% H PO
Conc. of Mg,wt. %.
Viscosity of simulated wet process phosphoric
acids [after Cate and Deming(22)].
Phosphate rock, which is essentially carbonate-
fluorapatite(24,25), could contain magnesium in one of the
following three possible ways.
i) Ionic substitution in the apatite lattice
ii) Second phase dolomite in the apatite matrix
iii) Discrete dolomite particles.
Magnesium present as the ionic substitute in the
lattice cannot be removed by any of the physical methods of
separation. The dolomite phase in the apatite matrix can be
physically removed provided it is liberated by further
grinding. Discrete dolomite particles can be separated by
the physical methods.
No detailed study of the distribution of magnesium ions
in the South Florida phosphate mineral has been reported.
One such investigation(25) is under way and the results are
yet to be reported. However it has been suggested(24) that
the amount of magnesium present as the substitute for Ca+2 in
the apatite lattice is well below the tolerance limit.
Furthermore, a few studies investigating the separation of
dolomite from apatite in the deposits of South Florida have
illustrated that significant reduction in MgO content can be
achieved by physical separation(13-19). This could be
possible only if a major fraction of the magnesium is present
as second phase dolomite either in apatite matrix or as
In the past, several investigators have concentrated on
the separation of calcite from apatite which is the major
problem in other phosphate deposits of the world(26-51). The
efforts to eliminate dolomite impurities from south Florida
apatite have been limited to a few engineering studies which
were undertaken by the phosphate industry(6,15,16,19,52,53),
the U. S. Bureau of Mines(17,54) and the Tennessee Valley
Authority(55). These investigations were conducted only
under specific conditions and their results indicate that the
methods developed are either inefficient or uneconomical for
commercial application. However, these studies provide a
direction for the future research to eliminate dolomitic
impurities from phosphate rock.
It should be noted that no complete and systematic basic
studies involving mineral-water-surfactant systems using
phosphate minerals from South Florida have been conducted.
Such a study would increase the fundamental understanding of
the current processes so that they could be further improved.
1.2. Objective of This Study
The overall objective of this study is to develop a
suitable technique for achieving effective separation of
dolomite from south Florida apatite and to elucidate the
mechanisms of separation. Specifically, the objective was to
investigate in two phases the selective flotation of dolomite
from apatite using sodium oleate as the collector. The aim
of the first phase was to test the selective floatability of
apatite and dolomite under a wide range of pH and collector
concentration by conventional conditioning. In the second
phase, the objective was to investigate selective flotation
of these minerals after two-stage conditioning. The two-
stage process involved first conditioning the solids at
alkaline pH and then reconditioning at a lower pH.
2.1. Characteristics of the Minerals
2.1.1. Characteristics of Apatites in Phosphate Rock
The principal phosphate mineral in all commercial
phosphate rocks is apatite, which varies in physical,
chemical and crystallographic properties in different
deposits. The compositions of apatites in crystalline
sedimentary deposits, which are encountered in south Florida
are essentially fluorapatite [Ca5(PO4)3F] with extensive
substitutions of carbonate for phosphate and fluorine and of
other metals for calcium(56-58). These carbonate-
fluorapatites (also called francolite) have been assigned the
where x is one fourth of the total number of carbon atoms.
The lattice structure of apatite is hexagonal and the lengths
of crystallographic axes are dependent on the extent of
substitution of other ions.
The carbonate apatite in most sedimentary phosphate
rock, including the south Florida deposits, is composed of
complex aggregates of microcrystals of size ranging from 0.02
to 0.2 microns(56,59). This morphology of the minerals
results in a very high specific surface area.
2.1.2. Characteristics of Dolomite
The chemical formula of dolomite is Ca Mg (CO3)2(60).
The crystal structure of dolomite is a rhombohedron with
cations at the corners and centers of the faces of the unit
cell and the CO3 ions at the center of the edges as well as
in the center of the cell. This mineral is known to have a
large surface area due to its porous nature.
2.2. Past Investigations of Apatite/Dolomite
Some of the physical properties of the apatite and the
dolomite minerals are shown in Table 3. Since both the
minerals are nonconducting and nonmagnetic, the separation
techniques based on these characteristics would not be
feasible. The specific gravities of both minerals are not
sufficiently different for the separation by conventional
gravity separation techniques. However,the 'apparent'
densities of the two minerals have been found sufficiently
different for the possible application of heavy media
separation(15). Another difference in the two minerals is
their hardness. Since dolomite is the softer mineral,
separation of dolomite after selective attritioning has been
attempted(16). Other methods of dolomite separation are
based on surface chemistry, namely selective froth flotation
and flocculation. The flotation studies to remove dolomite
from south Florida apatite have been limited to a few
laboratory scale tests under specific
Table 3. Comparison of physical properties of apatite and
Specific Gravity 3.1 3.35 2.86
Hardness (Mohs scale) 5 3.5 4
Magnetic nonmagnetic nonmagnetic
Electrical nonconducting nonconducting
conditions(6,15,16,19,52-55). However, several past
investigators(26-51) have studied the selective flotation of
apatite from calcite, which is the major impurity in other
phosphate deposits of the world. Selective flocculation has
not yet been attempted to remove dolomite from apatite. Some
of the relevant past studies are discussed below.
2.2.1. Heavy Media Separation
It was determined by Lawyer et al.(15) of the
International Minerals and Chemical Corporation that +1 mm
size fraction dolomite pebbles had more internal pores than
apatite pebbles of the same size. As a result, the dolomite
yielded a lower 'apparent' specific gravity than phosphate
pebbles because of the trapped air or moisture in these
pores. Measurements of numerous dolomite pebbles have
confirmed the true specific gravity in the range of 2.67-
2.84, and 'apparent' specific gravity was determined to vary
from 1.38 to 2.32. Porosities of dolomite has been estimated
to vary from 14.1 to 51.4 percent. The porosity differences
in the two minerals have permitted the heavy media separation
using an aqueous suspension of ferrosilicon or magnetite.
Laboratory scale heavy media separation using the Erickson
Cone showed a considerable promise as a processing tool for
separating dolomite from pebble phosphate in +1 mm size
fraction. However, the effective separation in the size
fraction smaller than 20 mesh was not achieved by this
technique. In the pilot plant tests, using a Dyna Whirlpool
(DWP) heavy media circuit, dolomite separation from apatite
was successful, but at phosphate recoveries of 52 to 74
2.2.2. Selective Attritioning
As indicated in Table 3, hardness of the apatite in Mohs
scale is higher than that of dolomite. It was also confirmed
by Iwasaki and Soto(61) that a constant agitation of mineral
slurries containing 33 percent solids resulted in a
relatively higher attritioning of dolomite. Predali et
al.(16) have also investigated the removal of dolomite by
selective attritioning. In this study three different size
fractions, -32,+250 mesh, -16,+32 mesh and +1 mm pebbles were
tested separately. Attritioning was done in a cell which was
fitted with a turbine comprising two propellers rotating at
940 rpm. It was also found that a cleaner phosphate product
could be obtained by introducing a weak grinding step in the
circuit. Forty to 60 percent of the dolomite was eliminated
by this method with phosphate losses of below 10 percent for
-1 mm fraction and 27 percent for +1mm fraction. To obtain
further reduction in MgO content, these investigators
recommended subsequent selective flotation of dolomite.
2.2.3. Selective Flotation of Minerals
As mentioned earlier, only a few studies have been
reported on selective flotation of dolomite from south
Florida apatite. However, several investigators have studied
the separation of calcite from foreign phosphate deposits.
Since the chemical composition of calcite (CaCO3) is
analogous to dolomite [Ca Mg (CO3)2], the results of some of
these investigations are discussed below.
Both cationic flotation using amines and anionic
flotation using mostly fatty acids or their salts have been
studied in the past.
18.104.22.168. Cationic flotation
An extensive investigation to separate francolite from
dolomite by cationic flotation was undertaken by the
International Minerals and Chemical Corporation in the late
seventies(14,46,52,53). The process utilizes the cationic
collectors, usually primary amines in combination with
kerosene, as the extender. For adequate removal of dolomite,
a rougher float followed by several cleaner flotation stages
was suggested. The requirement of several flotation stages
to reduce the MgO content to below 1 percent also decreases
the process efficiency. Another disadvantage of this process
is that the major component, francolite, which often
constitutes over 90 percent weight of the feed, must be
To improve the efficiency and to explore the possibility
of reversing the flotation procedure Iwasaki and Soto(61)
continued the above study on dolomite and apatite from south
Florida. They concluded that there is a chemical
interaction between the amine collector and the phosphate
ions present on the francolite surface. The selectivity was
attributed to the lower solubility of the amine-phosphate
complex formed on the apatite as compared to the amine-
carbonate complex that could form on dolomite. Thus, they
concluded that selective flotation of dolomite, instead of
apatite, can not be achieved by using the cationic collector.
Another study on Egyptian ores by Hanna(27), should also
be noted. These attempts to selectively float francolite
from calcite using cationic collectors were not successful.
Nonselectivity of flotation in this case was attributed to
the lack of significant differences in the adsorption power
of amine salts on the two minerals.
In order to improve the selective flotation of calcite
and apatite using cationic collectors, several investigators
studied the mechanisms of adsorption of amines on these
minerals. Some of the noteworthy studies include those of
Taggart and Arbiter(28), Klassen and Mokrousov(29), Cases et
al.(30), Solnyshkin and Cheng(31), Kuz'kin and Cheng(32) and
Levinskii(33). These studies found that amines chemically
interact with both apatites and carbonates to form complexes
of mineral anion and collector cation.
22.214.171.124. Anionic flotation
Four major studies pertaining to the separation of
dolomite from south Florida phosphate rock were conducted at
International Minerals and Chemicals Corp.(15), Minemet
Recherche, France(16), U.S. Bureau of Mines(17,54) and
Tennessee Valley Authority(55). The results of these studies
as well as some of the investigations of similar minerals
from other domestic and foreign deposits are discussed below.
At the International Minerals and Chemical Corporation,
Lawyer et al.(15) examined selective flotation of dolomite
using several anionic collectors, including fatty acids and
their soaps, sodium alkyl sulfates, and petroleum sulfonates.
These collectors were used at pH 5.6-6.2 with inorganic
depressants such as phosphates and fluorides. It was
observed that sodium tripolyphosphate and hexametaphosphate
with an anionic collector (unidentified because of pending
patent application) yielded the best results. This technique
was tested in the laboratory and pilot plant. The
application of this process to size fraction finer than 48
mesh and containing less than 2 percent MgO resulted in less
than 1 percent MgO product at BPL recoveries in the range of
50 to 90 percent.
Predali et al., at Minemet Recherche, France(16),
developed an anionic flotation scheme to eliminate residual
carbonate in the phosphate rock in which dolomite was
partially separated by selective attritioning. After the
flotation feed had been washed, the phosphate concentrate was
subjected to conditioning with anionic collector and a
depressant for phosphate particles at pH 5.5. The reagents
used have not been identified in the literature. The final
product, containing less than 1 percent MgO, was obtained
with BPL recoveries of 48 to 85 percent. Higher recoveries
were obtained for particles finer than 32 mesh.
Llewellyn, Davis, and Sullivan at the U.S. Bureau of
Mines(17) conducted studies to recover phosphate from four
dolomitic southern Florida deposits. These samples contained
5.8 to 10.2 weight percent P205 and 1.7 to 4.8 percent MgO.
In the conventional one-step fatty acid-fuel oil batch
flotation tests at pH 9.2 to 9.6, phosphate was floated while
silica and carbonates were depressed. It was observed that
the addition of sodium silicate aided in further depression
of the dolomite particles. The rougher concentrate was
further cleaned three times. The process yielded phosphate
rock containing 29 to 31.3 percent P205, 0.47 to 1.36 percent
MgO and 4.0 to 6.4 percent insol on a laboratory batch scale.
The flotation recovery of P205 ranged from 72.4 to 96.1
percent. Since the final grade acceptable for the
manufacture of fertilizers could not be achieved by
flotation, the investigators recommended additional removal
of dolomite by SO2 leaching.
In another study by the U.S. Bureau of Mines, Rule et
al.(62-64) developed a technique to selectively float
carbonates from western phosphates. The feed material was
conditioned with fluosilicic acid to depress phosphate
minerals and with an aqueous fatty acid emulsion at slightly
acidic pH. The fluosilicic acid renders the surface of the
phosphate minerals hydrophillic and prevents adsorption of
the fatty acid. The fatty acid then becomes a selective
collector for carbonate minerals. The laboratory flotation
tests yielded concentrates containing about 0.75 percent MgO
with flotation recoveries of 70 to 80 percent P205.
A recent patent(55) issued to the Tennessee Valley
Authority describes the use of alkyl phosphonic acids to
depress Florida phosphate while floating dolomite with oleic
acid as the collector at pH of 6 to 7.
The anionic flotation of calcium and magnesium carbonate
impurities from finely ground (60-80 percent -200 mesh)
western U.S. phosphorites, using various phosphate and
fluoride compounds as the phosphate mineral depressant at
slightly acidic conditions have been developed by Cominco
Ltd. (65-68). Phosphate concentrates analyzing less than 1.0
percent MgO were obtained from dolomite flotation feeds
containing 1.6 percent MgO with P205 recoveries of 85 percent
Johnston and Leja(43) determined that calcite and
dolomite floated at pH 6,using oleic acid as the collector,
whereas fluorapatite is fully depressed in the presence of
phosphate ions at that pH. The difference in flotation
behavior may be explained by the characteristic hydrogen
bonded phosphate ions that adsorb strongly on apatite. On
the other hand, it is suggested that the weaker hydrogen
bonding is occurring on the dolomite due to CO2 gas evolution
from carbonates in the acidic solutions.
Smani et al.(47) have reported separation of calcite
from a phosphate ore obtained from a Moroccan deposit using
0.9 Kg/ton (2 lb./ton) of fatty acid and a few hundred grams
of a 2:1 mixture of aluminum sulfate and tartaric acid as the
phosphate depressant. The best result indicated an upgrading
of a 29 percent P205 feed to 34 percent P205 concentrate with
95 percent P205 recovery at pH 5 to 6.
Mitzmager and coworkers(48) have reported that the
addition of monosodium phosphate significantly improved the
flotation of calcite from phosphate using fatty acid as the
collector. The calcite used in this investigation was from
Israel's Negev region.
Calcite in the -100 mesh phosphate rock from eastern
Turkey was separated by Onal(49) using phosphoric acid and
sodium oleate. However, the minimum MgO content of the
concentrate achieved was 1.4 percent, which is higher than
what is considered to be acceptable by the fertilizer
producers in the United States.
The other noteworthy studies include that of Belash(35),
Hanna et al.(36), and Ratobylskaya(37). These studies like
the ones discussed above have shown that selective flotation
of dolomite can be achieved at slightly acidic conditions
provided a suitable phosphate depressant is used with the
fatty acid collector.
Investigation by Predali(69) on the flotation of
dolomite and magnesite from Kosice deposits, Czekoslovakia
should also be noted. He studied flotation as a function of
pH and fatty acid chain length. The flotation of carbonates
in the acidic pH range increased as chain length of the
collector decreased from 18 to 12. On the other hand in the
alkaline pH range, the recovery of carbonates was higher with
the sodium oleate (18 carbon chain) than with sodium laurate
(12 carbon chain).
The low separation efficiency and additional cost of the
apatite surface modifiers are some of the major limitations
in applying the anionic flotation scheme on a commercial
scale. It should be noted that none of the studies involving
anionic flotation of minerals from South Florida deposits was
reported for the complete pH range (pH 3 to pH 11) or under a
wide range of collector concentrations. Furthermore, none of
these studies attempted to explain the mechanisms of
The basic principles of aqueous suspensions and
flotation of minerals are discussed next.
2.3. Basic Principles
The froth flotation(70-74) is a process for separating
different solids based on their physico-chemical nature.
This process has permitted the mining of low grade and
complex ores which would have been worthless if more
conventional methods, like gravity concentration, were relied
upon. Flotation of mineral particles results from the
attachment of gas bubbles to the solids while they are
suspended in aqueous solution. The float fraction, which is
the fraction of the minerals attached to the gas bubbles, is
collected in the froth that is later skimmed off. Although
this process has been in use for seventy years, the basic
principles are still not completely understood.
The fundamental requirement for particles to float is
that they should be preferentially wetted by gas and not
water, i.e. the mineral surface should be hydrophobic(75-77).
Only a small fraction of minerals like sulfur are naturally
hydrophobic. In the case of most of the other minerals
hydrophobicity has to be imparted onto the surface with a
surfactant(78) that will selectively adsorb on the mineral to
be floated. This surfactant, also called the collector, is
added to the mineral slurry to achieve its adsorption at the
solid/liquid interface. The success of the flotation process
lies in obtaining the selective hydrophobic character of the
minerals to be separated. The particle bubble attachment and
froth stability are also important factors that would
influence the efficiency of the flotation process.
Apatite and dolomite minerals are both salt type
minerals, whose solubility is lower than that of simple salt
minerals, but higher than that of oxide and silicate
minerals. Separation of these minerals is extremely
complex(26) because of their dissolution.
To understand the flotation properties, it is necessary
to examine the solubility, the aging behavior and
electrokinetic behavior of these minerals and the adsorption
of various organic and inorganic species present in the
system that could influence the various interfaces, namely
solid/liquid, and liquid/gas.
2.3.1. Solubility of the Minerals
Solubility of the minerals is of importance because of
the the role of dissolved ions in determining both the
chemical composition of the aqueous phase and the charge
characteristic of the mineral interface. When a salt is in
contact with water, lattice ions will be transferred to the
solution until the chemical potentials in both phases are
identical. Several studies(79-84) have been attempted in an
effort to understand the solubility behavior of apatite, but
none of them is adequate. The major discrepancies in the
data are attributed to the impurity ions in the lattice or
the added electrolytes. However Saleeb and deBruyn(85) have
shown that a constant solubility product can be obtained if
precautions are taken to obtain a stoichiometric compound.
In the case of dolomite the published pK values are in the
range of 16.5 to 19.5(79).
In most salt-type minerals such as calcite, dolomite, and
apatite dissolved ions in solution undergo various hydrolyses
and complex reactions.
In aqueous slurry, fluorapatite undergoes dissolution to
produce Ca+, HPO4 -2, and F-1 ions, which would undergo
further hydrolysis and complex formations according to the
reactions given below(86).
Ca2 + OH-1 CaOH+1 (1)
CaOH+1 + OH-1 # Ca(OH)2 (2)
Ca(OH)(aq) Ca(OH)(s) (3)
PO4 3 + H20 # HPO4-2 + OH-1 (4)
HPO4-2 + H204 H2PO41 -+ OH-1 (5)
H2PO4-1 + H20 H3PO4 + OH-1 (6)
Ca+2 + HPO4-2 t CaHPO4 (7)
+2 -1 +1
Ca+ + H2PO4- CaH2PO4 (8)
F-1 + H20 + HF + OH1 (9)
Ca+2 + 2F-1 i CaF2 (10)
Calcite, when dissolved in water, will undergo the
CaCC Ir CaCo
CaCO3(s) Ca3(aq) (11)
CaCo q Ca2 + CO32 (12)
CO3-2 + H20 HCO3-1 + OH-1 (13)
HCO3-1 + H20 0 H2CO3 + OH-1 (14)
H2CO3 = CO2(g) + H0 (15)
Ca+2 + HCO3-1 CaHCO3+1 (16)
CaHCO3 +1 # H+1 + CaCO (17)
Ca+2 + OH-1 V CaOH+1 (18)
CaOH+1 + OH-1 Ca(OH)2(aq) (19)
Ca(OH)2(aq) & Ca(OH)2(s) (20)
For the dolomite-water-CO2 system the following
reactions should be added to the reactions 13-20(88).
CaMg(CO3)2 C Ca+2 + Mg+2 + 2Co3-2 (21)
+2 -1 +1
Mg2 + HCO3-1 MgHCO3 (22)
MgHCO3+1 0 MgCO3(aq) + H+1 (23)
Mg+2 + OH-1 0 MgOH+1 (24)
MgOH+1 + OH-1 v Mg(OH)2(s) (25)
MgOH+1 + OH-1 # Mg(OH)2(aq) (26)
2.3.2. Aging Behavior of Mineral Slurries
When the dry minerals are suspended in water, the
dissolution of the minerals followed by the hydrolysis of the
dissolved ions and complex formation reactions, as presented
in the previous section, causes the pH of the slurry to
shift with time to an equilibrium value. This is often
referred to as the "aging phenomenon". It is of importance
in the present investigation since pH is the most critical
flotation variable. The aging behavior of apatites and
carbonates has been studied by the previous investigators.
Somasundaran(86) observed that the pH of the apatite
slurries decreased with time if the initial value was greater
than 7, whereas slurry pH increased if the initial value was
lower than 7. No significant changes were observed if the
starting pH was 7. He also noted- that it was necessary to
condition the minerals for two weeks to obtain constancy in
pH and zeta potential as a function of aging time. It should
be further noted that not only the slurry pH but also the
zeta potential (obtained by streaming potential) changed in a
similar manner. The isoelectric point, defined as the pH of
zero zeta potential value, shifted from pH 4 to 6 as
equilibrium was reached. Iwasaki and Soto(61) observed an
equilibrium pH value of about 6.7 for apatite from Florida
Agar and Somasundaran(87) studied the calcite-water
system. They observed a trend similar to that of the apatite-
water system. The equilibrium pH was found to be 8.2, which
was in agreement with the value calculated from thermodynamic
data. However, the isoelectic point of the mineral had an
initial value of 10.8, which shifted to 9.5 upon
Predali and Cases(88) studied the aging phenomena of
magnesite and dolomite from Kosice, Czekoslovakia. They
obtained equilibrium pH values very close to the calculated
isoelectric point of the solution, namely 8.46 and 7.96 for
magnesite and dolomite, respectively. Iwasaki and Soto(61)
determined an equilibrium pH value of about 8.4 for dolomite
2.3.3. Electrochemical Aspects of Mineral-Water Interface
126.96.36.199. Electrical double layer
When a mineral surface is brought into contact with
water, it acquires a surface charge(79,89-92). The possible
charging mechanisms are ionisation, ion adsorption and ion
dissolution. Besides this charging mechanism, the charge
characteristics of the mineral particles can be altered by
the adsorption of the dipolar molecules from the solution.
The surface charge on the solid surfaces influences the
distribution of the dissolved ions in the aqueous media.
Ions of opposite charge (counter ions) are attracted towards
the surface and the ions of like charge (co-ions) are
repelled from the surface. This, together with the thermal
effects, leads to the formation of an electric double layer.
This layer is made up of a charged surface and a neutralising
excess of counter ions over co-ions distributed in a diffuse
manner. The theory of double layer deals with the magnitude
of the electrical potentials which occur on the charged
surfaces. This is an important step toward understanding
several experimental observations such as electrokinetic
behavior of minerals and electrostatic forces governing the
adsorption of organic or inorganic ions. Details of the
electrical double layer phenomenon can be found in several
188.8.131.52. Quantitative determination of the potential at the
It is widely accepted that the measurement of the
electrokinetic effect is representative of the potential at
solid/liquid interface. The electrokinetic phenomenon is the
mobility of the charged particles caused by applying an
electric field and conversely an electric field induced by
the relative motion of such a surface. In each case the slip
plane between the double layer and the medium is involved.
Thus the electrokinetically measured value, called zeta
potential, is the potential at the plane of shear in the
double layer and strictly speaking not the surface potential.
In the absence of any direct means of measuring the actual
surface potential, zeta potential has provided valuable
information about the nature of solid/liquid interfaces.
There are four electrokinetic methods of measuring zeta
potential, namely electrophoretic mobility, streaming
potential, electro-osmosis, and sedimentation potential.
Electrophoretic mobility is the most developed method of
measuring zeta potential because reproducible results could
be obtained with relative ease.
Electrophoresis is the movement of charged colloidal
particles in suspension under a potential gradient. The
velocity of the particles can be the measured by a
microscope. This motion of a particle under applied field
directly proportional to its zeta potential:
S = ( )(4n )
1 = v/H
where 5 = zeta potential
y = electrophoretic mobility
n = viscosity of the fluid
e = dielectric constant
v = velocity of the particles in the
H = Applied potential drop per unit length
184.108.40.206. Isoelectric point of the minerals
In the past two decades, several attempts(84-87,93-101)
have been reported to characterize the electrokinetic
behavior and to understand the surface charge generation
mechanism for apatite and calcite. A few attempts have also
been reported(88,102-104) for dolomite. All investigators
have found H and OH to be the potential determining ions
for these minerals. In other words the value and sign of the
zeta potential is dependent on the pH of the slurry. A
particular mineral-aqueous system has a characteristic
isoelectric point pH at which the zeta potential has a zero
value. Below or above this pH, the mineral has either
positive or negative potential. Thus, the isoelectric point
is the parameter that is often used to compare the
electrokinetic behavior of different mineral slurries. The
reported values of the isoelectric point for apatites
(hydroxyapatite and fluorapatite), calcite, and dolomite are
summarized in Table 4, Table 5, and Table 6, respectively.
As seen from these data, there is a wide range of variations
observed in the results obtained by different investigators.
This could be attributed to the differences in the
experimental techniques, sample preparation, and the
impurities present on the mineral surfaces.
220.127.116.11. Mechanisms for surface charge generation
The mechanisms of surface charge generation for salt-
type minerals are basically different from those of simple
oxide minerals(26,75,77,86-88,105-108), owing to the
relatively high solubility of the former. In addition to any
preferential dissolution of the constituent species, the
surface charge development is also .influenced by the
formation of various complex species due to the hydrolysis
reactions discussed in section 2.3.1. These ionic species
may be produced at the solid/liquid interface or in the bulk
Isoelectric point of apatite.
Sample pH Method IEP electrolyte Ref.
Streaming pH 5.6
Electro- pH 3.5
Electro- pH 5.5
pH 6.5 KNO
pH 7.15 KNO3
pH 6.5 KNO3
Table 5. Isoelectric point of calcite.
Sample pH Method IEP electrolyte Ref.
1. Natural Electro- pH 10.8 _(100)
Aged Hours phoresis
2. Natural HNOI/ Streaming pH 10.8- KNO3 (87)
Iceland Spar NaOH potential 9.5
3. Natural HC1/ Streaming pH 5-6 (45)
Morrocco NaOH potential
4. Natural NaOH Electro- pH 9.5 (101)
5. Natural HC104/ Electro- pH 8.2 NaClO4 (95)
Broken Hill NaOH
6. Natural Electro- pH 10.1 (93)
Table 6. Isoelectric point of dolomite
Sample pH Method IEP electrolyte Ref.
1. Natural HC1/ Streaming pH<7 KC1 (88)
Kosice, NaOH Potential
2. Natural Streaming pH<6 (103)
Bela Stena Potential
solution and subsequently adsorb on the mineral in amounts
proportional to their concentration in the solution.
Besides the H+, OH and constituent ions of the
minerals, other 'foreign' ions, both inorganic and organic,
could also affect the zeta potential of the minerals by their
adsorption(75,109-111). The adsorption of polyvalent metal
cations which results in significant alteration of the
minerals surface charge characteristics is known(112,113) to
occur when the cations involved are hydrolyzable. In the
case of quartz, a charge reversal, from negative to positive,
was observed in the presence of Pb+2(114), Fe 2(112) and
Mg+2(114-118) at the pH values corresponding to the
hydrolysis of these cations. In flotation or flocculation
processes, the surfactants or polymers are added so that the
particles will acquire the desired surface properties. The
charge characteristics of the adsorbed organic ions are also
known(111,119) to have an influence on the electrokinetic
behavior of the minerals.
2.3.4. Adsorption of Surfactants
In flotation, the adsorption of surface active agents on
solid/liquid and liquid/ gas interfaces is of significance.
In the literature, the adsorption of the flotation
collectors, as well as their interactions with coadsorbents,
like extenders, brothers. and polymers(74,91,120-126), has
been reported. Since this study does not include the use of
any other surface active agents, only the collector
adsorption on mineral/water and water/air interfaces will be
18.104.22.168. Adsorption on mineral-water interface
Except for some naturally occurring hydrophobic
minerals, hydrophobicity on most minerals to be floated is
achieved by adsorption of the collectors at the solid/liquid
interface. A suitable collector and adsorption conditions
are established so that the selective hydrophobicity is
obtained for the particular minerals to be separated.
Adsorption mechanisms. Adsorption of the surfactants or
their complexes in solution occurs on the minerals due to the
interaction between the solute and surface species on the
solid. Density of adsorption of the surfactant is commonly
interpreted using the Stern-Graham equation(127-130):
6 = 2r C exp (- AGoads /RT)
where rF = adsorption density in the plane 6 ,which is
at the distance of closest approach of counter ions to the
r = the effective radius of the adsorbed ion.
C = the bulk concentration of the adsorbate.
R = gas constant.
T = absolute temperature.
AG ads= the driving force for adsorption. This is the
sum of a number of contributing interactions between the
solute and the mineral surface as shown by the equation
AG0 = AG0 + AG0 + AG0 + AG0 + AG0
ads elec chem c-c c-s H
AG0 = electrostatic interaction between the electrical
field on the mineral/water interface and the charge on the
surfactant or its complex.
AGo = chemical interaction due to covalent bonding of
surfactant to the mineral surface specie.
AG = lateral interaction term between the long carbon
chain of the adsorbed surfactant.
AG = interaction terms between the hydrocarbon chains
of the adsorbed surfactant and the hydrophobic sites on the
AGH = interaction due to the adsorption of surfactant
by hydrogen bonding.
AGo 0= dissolvation term due to displacement of any
species from the interface due to adsorption.
The adsorption of surfactants at the solid/liquid
interface can be classified into the two general categories,
physical adsorption or chemisorption. These categories have
been arbitrarily defined by the amount of adsorption energies
involved. If the free energy of adsorption is less than 10
kcal/mole(106), then adsorption is physical, whereas it is
chemisorption if the energy of adsorption is higher. Except
for the adsorption by covalent bonding chemisorptionn), all
other adsorption mechanisms, mentioned in the previous
paragraph, are physical. Physical adsorption is reversible,
whereas chemisorption is irreversible. For a particular
mineral-water-collector system, the adsorption could be
physical, chemical or a combination of both.
The major factor that determines the type and the amount
of adsorption beside the syrface charge of the minerals is
the nature of the surfactant. Even for the same surfactant,
adsorptive power may vary as a function of pH because of
changes in the oleate solution chemistry(129). Furthermore,
for a particular salt-type mineral-water-oleate system, the
adsorption by surface precipitation of oleate also has been
recently found to be the major contributing factor(131).
Influence of oleate solution chemistry. In addition to
micellization and precipitation, certain surfactants, like
fatty acids and amines, undergo hydrolysis and associative
interaction to form dimers, trimers and acid-soap
complexes(132,133). The concentration of the hydrolyzed
species and various complexes is influenced most by the
solution pH. Each of these species would not only have
different adsorption behavior but would also induce varying
degree of hydrophobicity on the mineral surface. The
solution chemistry of oleate has been studied in detail by
Somasundaran et al.(134-137). They constructed an
equilibrium diagram, based on estimated and experimentally
determined reaction constants, showing the activities of
various species of oleate at a concentration of 3 x 10-
kmol/m3. It is seen further in Figure 3 that the
precipitation of oleic acid occurs below pH 7.78. The
maximum acid-soap complex formation also occurs at this pH.
3 4 5 6 7 8 9 10 II
Oleate species distribution as a function of pH
[after Ananthpadmanabhan and Somasundaran(136)].
In spite of the low numerical value of the acid-soap complex
concentration, it is sufficient to affect the adsorption and
flotation. The surface tension studies on oleate solution as
a function of pH suggest the presence of highly surface
active species in the pH range at which these complexes are
expected to be present in maximum amount. Also, in the case
of hematite, the narrow alkaline pH range in which flotation
is observed was found to coincide with the pH at which these
complexes are present. Moreover, as shown in Figure 4(136),
a number of minerals have exhibited maximum flotation with
fatty acid collector around this pH range. These studies
suggest an active role of acid-soap complex in adsorption and
flotation of minerals. The adsorption of monomers and dimers
could also occur at alkaine pH. However, at higher values,
pH 11 or above, there would be an increased competition for
adsorption from OH ions(139). It should be noted that the
contribution of oleate dimers in making the mineral surface
hydrophobic would depend on the orientation of the adsorbed
Precipitation of oleate soaps. Oleate anions are known
to precipitate readily with multivalent cations present in
the solution(140). This is attributed to high pK value of
the precipitation reactions as shown for the calcium and
magnesium oleate precipitates in Table 7(141-143). The
precipitation could occur in the bulk solution or at the
mineral/water interface. The bulk precipitation would
Flotation recovery of minerals as a function
of pH using fatty acid collector, 1 Kg/ton.
1) columbite 2) zircon 3) tantalite 4)
ilmenite 5) rutile 6) garnet, 7) tourmaline,
8) albite, 9) perovskite [after
Ananthpadmanabhan and Somasundaran (136)].
Table 7. Solubility product of calcium oleate and
ref. 141 ref. 142 ref. 143
Calcium Oleate 3.98 x 10-13 2.51 x 10-16 2.04 x 10-20
Magnesium Oleate 1.58 x 10-11 6.31 x 10-16 4.68 x 10-19
deplete the free oleate anions available for adsorption. On
the other hand, the precipitation on the mineral surface
would increase the oleate adsorption at the interface, thus,
it would be an additional mechanism for adsorption. Since
the precipitation requires the presence of dissolved cations,
it would be significant for the salt-type minerals. On the
basis of adsorption isotherm for calcite at pH 10 as
illustrated in Figure 5(145), this mechanism was
recently(131,144) proposed. The slope of the isotherm
increased several fold beyond a certain minimum collector
concentration. It has been suggested, that due to the
soluble nature of the salt-type minerals, there is a higher
concentration of dissolved cations at the interface. As a
consequence in such systems, the oleate precipitation could
occur on the surface at lower collector concentration than in
the bulk. It should be noted that a similar adsorption
isotherm was also observed for magnesite by a Predali and
Cases(146). The mechanisms of surface precipitation
phenomenon are not yet completely understood.
22.214.171.124. Adsorption on solution-gas interface
The attachment of particles to gas bubbles is the most
important step in the flotation process. If more gas bubbles
are available for attachment, the flotation kinetics could be
further improved. In a pure liquid, the smaller bubbles
coalesce to form larger bubbles, whereas in surfactant
solutions, a higher number of finer and more stable bubbles
Eq. Conc. of Oleate, Kmol/m3
Adsorption isotherm of oleate on calcite at pH
9.6 [after Somasundaran(145)].
are present. The adsorption of the collector on the
liquid/gas interface, therefore, aids the flotation kinetics
by increasing the number of stable gas bubbles available for
particle attachment. It has also been suggested that the
adsorbed collector at the L/G interface could affect the
induction time for particle-bubble attachment. The
phenomenon of particle-bubble attachment is not sufficiently
well understood to isolate this role of the collector
adsorption at L/G interface in the overall flotation process.
Orientation of the collector molecules adsorbed at the
water/air interface is such that the polar group stays in the
solution and the nonpolar hydrocarbon chain points toward the
gas phase resulting in lower surface tension. Measurement of
surface tension, therefore, represents one of the common
methods of determining adsorption surfactants at the L/G
interface. The surface tension measurements are mostly
conducted under equilibrium conditions. In the flotation
process, a bubble is only a fraction of a second old and
adsorption on its surface can be different from that
determined under equilibrium conditions. The effect of the
dynamic nature of surface tension of oleate solution has been
illustrated in a previous investigation(147).
3.1. Selection of the Mineral Separation Technique
Past investigations of some of the separation methods
(refer to Section 2.2.) suggest that selective flotation is
the technique of greatest potential for separating these
minerals. For a better process efficiency in a large-scale
operation, the selective flotation of the minor mineral,
dolomite, is preferred(15). Cationic flotation using amines
as the collectors, as attempted in the past, yielded lower
process efficiency, as well as resulted in the flotation of
the major mineral, apatite. On the other hand, anionic
flotation resulted in the selective flotation of dolomite.
Although the past studies on the anionic flotation did not
result in an economically feasible technique for a commercial
application, they have demonstrated a potential for
eliminating dolomite impurities in the phosphate rock. Thus
anionic flotation, using sodium oleate as the collector, has
been selected for investigation in the current study. It
should be noted that sodium oleate, which solubilizes readily
in water, was selected because oleic acid is the major
constituent of the fatty acid collectors currently used by
the mineral industry.
3.2. Selection of Minerals and Preparation Methods
Selection of the natural minerals from South Florida
deposits for a systematic study was governed by the following
A. Since the objective of this study is to develop a
technique to eliminate dolomite impurity in south Florida
phosphate rock, the choice of natural minerals from the same
location was considered to be appropriate.
B. Synthetic dolomite and carbonate-fluorapatite are
not commercially available. In the laboratory, they could be
synthesized only in a powder (-200 mesh size) form, which is
not a suitable size for flotation tests.
The sample preparation method was selected so that these
minerals are not exposed to any surfactants prior to testing.
The dolomite pebbles, which did not contain any significant
apatite impurity, required only a washing step to remove the
clays. On the other hand, since apatite pebbles of
sufficient purity could not be obtained, the flotation feed
containing the mixture of apatite and silica was used.
Electrostatic separation, instead of the conventional
flotation using amines or fatty acids, was selected as a
method to remove silica from the feed.
3.3. Outline of The Investigation
The following steps were followed in the current study:
A. Characterization of the mineral samples by
determining their chemical compositions, surface
area, solubility and aging behavior.
B. Flotation tests by conventional conditioning under
a wide range of pH and collector concentration. It
should be noted that all earlier studies(15-
17,54,55) were conducted at slightly acidic pH
conditions using an additional surface active agent
to depress apatite flotation. This part of the
current study was planned to determine if selective
flotation of the minerals could be obtained at
other pH values under a wider range of collector
concentrations without using any other types of
C. Flotation tests by two-stage conditioning which
involve first conditioning at high alkaline pH and
then reconditioning at a lower pH. Reconditioning
was done both with and without reducing collector
concentration. In the preliminary tests on calcite
and alumina, it was determined(148) that after two-
stage conditioning, there was selective flotation
of carbonate. Therefore, it was also planned to
test the selective flotation of the dolomite from
apatite by similar technique. It should be noted
that there is a need to test new conditioning
methods of achieving surface hydrophobicity,
because conventional conditioning techniques, using
some of the less costly and easily available
collectors like fatty acids, are not effective.
D. Elucidation of mechanisms involved in the observed
selective flotation by oleate adsorption studies,
electrokinetic measurements, determination of the
nature of the adsorbed species, and froth
3.4. Selection of the Experimental Techniques
The following experimental techniques were selected.
126.96.36.199. Apparatus for flotation tests
A Hallimond cell was selected for flotation studies.
This device was preferred because it is easier to control the
parameters like flotation time, agitation, air flow rate and
collection of the float fraction in the Hallimond cell than
with other laboratory scale flotation apparatus. It is also
recognized, however, that a direct correlation of Hallimond
cell results to a laboratory cell in terms of the percent
recovery may not be achieved, but it is expected that the
trends in the flotation behavior will be reproduced in the
bench scale tests(149,150).
188.8.131.52. Particle size of the minerals
Generally, the size fraction -16,+150 mesh is used for
the flotation of phosphate rock. However, at the extremes of
this size range, the flotation recovery usually becomes
sensitive to the particle size. To eliminate the possible
effect of particle size on the recovery, a narrow size range
-65,+100 mesh was selected. As discussed in Chapter IV, this
size range was also found to be suitable for the oleate
3.4.2. Oleate Adsorption Studies
These tests were conducted to determine the amount of
oleate adsorbed on the surface by altering the conditioning
or reconditioning variables. Oleate adsorption measurement
using solution depletion method was selected.
3.4.3. Electrokinetic Measurements
The determination of zeta potential of minerals can lead
to a better understanding of oleate adsorption
characteristics and thus their flotation behavior. In this
study, the electrophoretic technique of measuring the zeta
potential was selected.
3.4.4. Determination of the Nature of the Adsorbed
To explain the differences in flotation behavior under
different pH conditions, it is necessary to identify the
nature of the oleate species adsorbed. Since the oleate
adsorption studies discussed earlier are not expected to
yield any information about the nature of the adsorbed
species, transmission IR spectroscopy was selected to
identify the adsorbed species.
3.4.5. Bubble and Froth Characteristics
Knowledge of changes in the size of the bubbles and the
characteristics of froth would aid in understanding their
role on the flotation behavior. Unlike past investigators,
who extrapolated the surface tension measurements to the
flotation conditions, the in-situ determination of bubble
size and froth characteristics in the Hallimond cell was
selected to study the role of liquid/gas interface on the
3.5. Selection of Experimental Variables
Based on the preliminary tests and the data reported in
the literature, the following variables were selected.
pH of the slurry is one of the important variables for
flotation because the surface charge characteristics of the
minerals and the solution chemistry of the collectors are
governed by this parameter. By correlating the flotation,
oleate adsorption, electrokinetic behavior of the minerals
and solution chemistry of the collectors as a function of pH,
the observed selectivity in mineral flotation can be
understood. The pH range, 3 to 11, which is feasible in
practice, was selected for testing.
3.5.2. Collector Concentration
Collector concentration was selected as the system
variable since this factor is known to influence the
adsorption and flotation processes. On the basis of the
preliminary flotation tests, the range of collector
concentration studied was determined to be in the range of 0
to 1.87 x 104 kmol/m However oleate adsorption tests were
conducted up to 4.7 x 10-4 kmol/m3
3.5.3. Presence of Foreign Ions
Since both apatite and dolomite are soluble, the
dissolved ions from one mineral could affect the surface
charge behavior of the other mineral. Wherever necessary the
apatite tests were conducted with the ions dissolved from
dolomite and vice versa.
3.5.4. Ionic Strength
Ionic strength effects due to modifications in pH and to
the presence of the dissolved ions were tested up to 3 x 10-2
3.5.5. Apatite:Dolomite Ratio in the Feed
Since the final objective of the study is to separate
the minerals from their mixture ranging from low (5 wt.
percent) to high (50 percent) dolomite content, the
individual minerals, as well as mineral mixtures, in these
ratios were selected as the flotation feed.
MATERIALS & METHODS
4.1.1. Mineral Preparation
184.108.40.206. Preparation of apatite
The flotation feed sample, containing 10.7 percent
P205, from South Florida phosphate deposits was supplied by
the International Minerals and Chemical Corp., Bartow,
Florida. This sample, consisting of -16,+150 mesh size
fraction of phosphate rock and silica grains, was obtained
after desliming and sizing to remove the fines (-150 mesh)
and the coarse pebbles (+16 mesh) associated with the
Silica, the gangue mineral in the feed sample, was
separated electrostatically at Carpco, Inc., Jacksonville,
Florida. The sample was first dried at 100 OC and then fed
into the electrostatic separator. The scheme of the
separation is shown in Figure 6. The final sample, observed
under an optical microscope, did not contain free silica
grains. The removal of silica upgraded the P205 content of
the feed from 10.7 percent to 36.78 percent yielding only 6
Ibs. of the final product out of the total feed of 37 lbs.
The particle size analysis of the cleaned sample is reported
in Table 8.
Figure 6 Electroseparation scheme to remove quartz
Table 8. Size analysis of apatite concentrate after
removal of quartz by electroseparator.
Size Range weight Percent
-16, +20 1.3
-20, +35 24.9
-35, +65 49.9
-65, +100 19.6
-100, +150 3.5
The cleaned sample was confirmed by x-ray diffraction
to be carbonate-fluorapatite (also called francolite). In
this study the carbonate-fluorapatite sample will be
referred to as apatite.
220.127.116.11. Preparation of dolomite
Dolomite, +1/2 inch hand-picked pebbles from South
Florida deposits, was supplied by the International Minerals
and Chemical Corp., Bartow, Florida. From this large batch
of sample, tan colored pebbles were further sorted out.
X-ray diffraction tests confirmed that the mineral was
Both apatite and dolomite samples were ground and
sieved to yield -65,+100 mesh and -325 mesh size fractions.
The coarser size fraction was used for flotation and oleate
adsorption tests and the finer sample was utilized in zeta
potential and IR spectroscopy studies. The procedures for
preparing these samples are outlined below.
The coarser size fractions, -65,+100 mesh, were
prepared in an alumina mortar and pestle. To improve the
yield of the particles in this narrow size range, care was
taken to avoid overgrinding by frequently screening the
samples. In spite of this measure, 58 percent of the
apatite and 71 percent of the dolomite samples were lost as
a -100 mesh fraction. The ground samples were washed 10
times with triple distilled water, dried at 50 C and stored
in a glass bottle.
The -325 mesh samples were prepared by grinding in a
steel rod mill for six hours. With a hand magnet the excess
iron impurities in the ground powders, that may have been
introduced from the steel mill, were removed. These samples
were also stored in a glass bottle.
4.1.2. Mineral Characterization
18.104.22.168. Chemical analysis
The chemical analysis of the apatite and the dolomite
conducted by Thornton Laboratories Inc., Tampa, Florida is
reported in Table 9. It is clear from the data presented
that the amount of MgO (0.12 percent) in the apatite and the
amount of P205 (1.62 percent) in the dolomite are not
significant. The major impurity in the cleaned samples is
silica which is reported as acid insolubles (3.34 percent in
apatite and 5.11 percent in dolomite). The other minor
constituents are Fe 03 and A1203, which are both less than 1
percent. The CO2 (1.62 percent) and fluorine (3.82 percent)
content in apatite confirm that it is a carbonate-
22.214.171.124. Surface area measurements
The surface area was measured by B.E.T. technique using
nitrogen as the adsorbate in the Quantasorb Sorption System,
supplied by Quantachrome Corp., New York. The measured
surface areas of the -65,+100 mesh size fraction of the
minerals are reported as follows.
Table 9. Chemical analysis of apatite and dolomite.
Mineral P 05 Fe2 0 A103 Acid MgO CaO CO F Organic
% % % Insoluble % % % % Carbon
Apatite 36.78 0.42 0.93 3.34 0.12 50.35 1.68 3.82 0.19
Dolomite 1.67 0.94 0.96 5.11 17.58 32.33 42.16 0.21 0.06
(loss of ign.)
Apatite: 18.83 m2/g
Dolomite: 4.19 m 2/g
The surface areas of the minerals are higher than
expected for nonporous particles in this size range. The
presence of fine open pores in both the minerals was
confirmed by pore size distribution which was determined by
a mercury porosimeter, supplied by Quantachrome Corp., New
York. The pore size distributions of apatite and dolomite
are reported in Figures 7 and 8, respectively. The pore
size distribution of the two minerals is similar, except for
10 percent higher number of pores of size lower than 200
angstrom in apatite as compared to dolomite. Due to high
surface area of the minerals, flotation and oleate
adsorption tests could be conducted on the same size
126.96.36.199. Scanning electron microscopy
Apatite and dolomite particles (-65,+100 mesh) were
aged for two hours at pH 10 in a similar manner as in
flotation and adsorption tests. After drying at 50 0C, the
samples were observed, as illustrated in Figure 9, under the
SEM at 100x and 10,000x magnification. High surface
porosity of both the minerals is further confirmed in the
micrographs at 10,000x magnification. It is to be noted
that apatite particles are composed of aggregates of
iO-3 16-2 ICo I o o I 1 2
Pore Size, microns
Pore size distribution on apatite.
I Po3 re2 Size microns 02
Pore Size, microns
Pore size distribution on dolomite.
SEM micrographs of apatite and dolomite (-65, +100
mesh size fraction).
submicron size crystals. A similar characteristic of
carbonate apatite has been observed in most sedimentary
188.8.131.52. Aging behavior
The soluble minerals, such as apatite and dolomite, are
known to exhibit the aging phenomenon, i.e. pH of an aqueous
slurry of these minerals will vary as a function of time.
Since pH is an important variable in the present study, it
was decided to examine the aging behavior of the minerals by
monitoring the pH of a 1 wt. percent suspension as a
function of time. These tests were conducted using -65,+100
and -325 mesh material.
Aging behavior of apatite. The aging data at different
levels of initial pH are presented in Figures 10 and 11 for
-65,+100 and -325 mesh, respectively. With -65,+100 mesh
size fraction, the equilibrium pH was reached after
approximately 600 minutes. The equilibrium value of pH 6
and 7 was observed with the slurry of initial pH 4 and 10,
respectively. No significant shift in pH was observed in
the case of slurry with initial value of pH 7. For -325
mesh size fraction, the system attains equilibrium
instantaneously in the acidic pH range. Aging behavior of
-325 mesh fraction at pH 10 was found to be similar to that
of the coarser size fraction. Similar pH shifts toward
equilibrium, also observed by previous investigators(86,61),
have been attributed to the dissolution of the mineral.
-65, 100 mesh
.... ..... 57
-- ..A l-. .-.-.-s- --7C.- ..- 7.-. 7. 7 .
--- "- ---.-------. . .- 0
Aging Time, min.
Effect of aging on pH of 1 wt. percent apatite
slurry (size fraction -65,+100 mesh).
Particle Size: 325 mesh
Pulp Density: I wt. %
..... ..... 5.7
-" -- "- 3.9
0 500 1000 1500 2000
Aging Time, min.
Effect of aging on pH of 1 wt. percent apatite
slurry (size fraction -325 mesh).
Since the mineral solubility is higher in acidic pH range
and for finer size fraction, the equilibration pH under
these conditions is therefore also attained faster. Effect
of ionic strength on the aging behavior of apatite was also
investigated. It is clear from Figure 12 that the aging
behavior in the presence of 3 x 10-2 kmol/m3 KNO3 remained
the same as that in the absence of the electrolyte.
Aging behavior of dolomite. The aging behavior of
dolomite is plotted in Figures 13 and 14 for -65,+100 mesh
and -325 mesh size fractions. Size fraction -65,+100 mesh
reached an equilibrium value of pH 8.2 after 800 minutes.
In the case of the finer size fraction (-325 mesh), the
equilibrium value of pH 8.5 was attained after 1800 minutes.
The equilibrium pH range 8.2 to 8.5 corresponds closely to
the equilibrium value reported in the literature(88,61). As
illustrated for dolomite in Figure 15, the presence of 3 x
10 kmol/m KNO3 did not affect its aging behavior.
From the aging data it is observed that equilibration
time of 8 hours or more would result in a narrow pH range of
6 to 8 thus eliminating pH as a variable. For -65,+100 mesh
size fraction of apatite and dolomite, if the mineral slurry
was aged for 2 hours, there would be a possibility of
studying the effect of wider pH range. After 2 hours of
aging, the pH drift during the next 25 minutes, which is the
maximum duration of flotation tests, would not be more than
+0.1 pH units. It was decided, therefore, that all the
Particle Size =-325 mesh
Pulp Density = Wt %
0 0 Kmol/m3 KNO3
a 3X10-2 Kmol/O KNO
Aging Time, min.
Effect of aging on pH of 1 wt. percent apatite
slurry with and without KNO3 (size fraction
Particle Size: 65, + 100 mesh
Pulp Density: I wt.%
Aging Time; min.
Effect of aging on pH of 1 wt. percent
dolomite slurry (size fraction -65,+100 mesh).
- 325 mesh
SI I I i I I I I
Aging Time, min.
Effect of aging
on pH of 1 wt. percent
(size fraction -325 mesh).
Particle Size=-325 mesh
Pulp Density=l Wt.%
0 0 Kmol / m KNO
y A 3X10 Kmol/m KNO
I I I I
Effect of aging on pH of 1 wt. percent
dolomite slurry with and without KNO3
fraction -325 mesh).
Aging Time, min.
samples be pre-aged for two hours at the desired pH prior to
flotation and adsorption tests.
184.108.40.206. Solubility of the minerals
The solubility of the two minerals was measured by
analysing the dissolved cations in the supernatant of the 1
wt. percent slurry of -65,+100 mesh size fraction after
aging it for two hours. The cations in the solution were
measured by atomic absorbtion (Perkin Elmer 6000
spectrophotometer). The dissolved cations from apatite and
dolomite as a function of pH are plotted in Figure 16. It
is to be noted from these results that dissolution of the
dolomite is significantly higher than that of apatite.
Apatite is slightly soluble in the pH range of 7 to 10 and
negligibly soluble at PH 11. On the other hand the
solubility of dolomite is significant in the pH range 7 to
11. At pH below 7, the solubility of both the minerals
increases exponentially as the pH decreases. The solubility
of individual and mixed minerals at various pH levels is
shown in Table 10.
Purified sodium oleate from Fisher Scientific Company
was used in this study. ACS certified grade potassium
hydroxide and nitric acid were used to modify the pH. Also,
ACS grade calcium and magnesium nitrate standards, potassium
carbonate, potassium phosphate, and 99.999 percent potassium
nitrate were used in this investigation.
Suspension aging time
= 2 hrs
Effect of pH on the amount of cations
dissolved from apatite and dolomite (-65,+100
mesh) 1 wt. percent slurry.
I I I
Amount of dissolved calcium and magnesium
ions in the supernatant of apatite and
dolomite slurries after aging for two hours.
pH mineral Ca+ Mg+
11 Apatite 0.1 0.0
Dolomite 1.4 0.0
Apatite 0.3 0.0
10+3 Dolomite 1.2 0.4
-- 1:1 mixture 0.9 0.3
95:5 mixture 0.3 0.0
Apatite 0.6 0.2
7+0.7 Dolomite 5.0 0.0
- 1:1 mixture 3.3 1.8
95:5 mixture 0.7 0.3
Apatite 8.0 0.7
Dolomite 70 32
4- 1:1 mixture 50 15
95:5 mixture 10 1.4
1 ppm Ca+
= 2.5 x 10-5 kmol/m3
1 ppm Mg+ = 4.1 x 10-5 kmol/m3
In this study, deionized, triple distilled water of
less than 1.2 micromhos specific conductivity was used.
4.2.1. Flotation Tests
220.127.116.11. The Hallimond cell apparatus
Flotation tests were conducted using a modified
Hallimond cell(149,150) as sketched in Figure 17. The
complete Hallimond cell flotation arrangement is shown in
Figure 18. Since 'prepurified' nitrogen containing less
than 1 ppm CO2 was used, no further attempt was made to
purify the gas. The gas was first passed through a column
of Drierite D, to remove moisture from the gas stream. The
sparger T2 is provided to remove the solids from the gas.
The liquid traps, T1 and T3, prevent water from entering the
rest of the circuit. Next the nitrogen gas enters a 50
litre glass reservoir, R. This large volume helps to
maintain a constant gas pressure, which is measured by the
manometer, M. A solenoid valve, V, is provided to turn the
gas supply on and off by a remote timer. The gas flow to
the flotation cell is regulated by a flowmeter, F. The
outgoing gas from the flowmeter is connected to the
Hallimond cell. The magnetic stirrer, preadjusted to a
fixed speed, can also be turned on and off by the remote
timer. The arrangement is designed so that the gas supply
to the cell and the magnetic stirrer can be turned on/off
simultaneously or individually by the timer.
'ork S topper
Hallimond Flotation Cell
Hallimond flotation cell.
Hallimond Cell Flotation Arrangement
Arrangement for nitrogen gas supply and Hallimond cell.
Prior to flotation, the mineral slurries were
conditioned in 100 ml volumetric flasks by tumbling around
an horizontal axis at 28 rpm.
18.104.22.168. Flotation test procedure
The flotation test procedure consists of the following
A. Solution preparation
B. Aging of the mineral slurry
Solution preparation. A stock solution of 1 x 10-2
kmol/m3 sodium oleate was prepared. After 15 days, a white
curd-like precipitate was observed at the air/liquid
interface. The effect of this precipitate was tested by
obtaining flotation recovery as a function of stock solution
age. It was observed as shown in Figure 19 that the
flotation of apatite and dolomite was not affected by using
up to three-week-old stock solution. However, as a
precaution, solutions more than 5 days old were not used in
this study. Approximately 30 minutes prior to the
conditioning of the minerals this stock solution was diluted
to 2 x 10-3 kmol/m3 sodium oleate concentration for further
Some of the experiments involved using the solutions
containing calcium nitrate, magnesium nitrate, potassium
carbonate and potassium nitrate. The solutions of desired
IXIO-2Kmol/n3 Sodium Oleate.
(1.45 X 10-4 Kmol/m3 Sodium Oleote)
(5.61 X 105 kmol/m3 Sodium Oleate)
S5 10 15 20 25
Stock Solution Aging Time,days
Effect of sodium oleate stock solution age on the
flotation of apatite and dolomite.
concentrations of these inorganic chemicals were prepared
just prior to testing.
Aging of the mineral slurry. It was noted in the
earlier section on aging characteristics of the mineral
slurries that pH of the apatite and dolomite slurries shift
with time. In order to perform the flotation experiments
over a wide pH range without significant shifts in pH, the
optimum aging time of 2 hours was determined to be
necessary. The detailed procedure for aging the mineral
slurry is described below.
One gram sample of apatite, dolomite or their mixtures
were suspended in 107 + 1 ml (the total volume of 100 ml
volumetric flask) of water. The pH of this water was
preadjusted so that the desired pH was obtained after
conditioning. The samples were aged without any agitation
for 120 + 5 minutes at room temperature 25 + 1 0C.
Conditioning. Flotation experiments were conducted
after reagentizing the minerals with sodium oleate by
conventional (one- stage) and two-stage conditioning. The
detailed conditioning procedures were as the following:
A. In conventional (one-stage) conditioning, the
minerals were treated with reagents directly at the pH at
which they were floated. First, the collector solution (2 x
10-3 kmol/m3) was added to the volumetric flask containing
the aged slurry. This was done by first pouring out 40 ml
of the supernatant in a clean glass beaker and adding a
desired volume of the sodium oleate solution. The
supernatant in the glass beaker was immediately transferred
to the flask containing the remaining mineral suspension.
Next the mineral slurry with the collector was tumbled in
the conditioner for a desired duration at room temperature.
It was determined that flotation recovery, as shown in
Figure 20, was not affected significantly by conditioning
time of up to 10 minutes. To allow adequate time of mixing
with minimum attritioning effects, the conditioning time of
five minutes was selected. After conditioning, the pH of
the supernatant was measured.
In one series of tests, the aging and conditioning at
pH 10 of apatite and dolomite samples were conducted in a
specially constructed glass container, as shown in Figure
21, instead of a standard 100 ml volumetric flask. This
container was built so that apatite and dolomite could be
conditioned together without the minerals coming into
physical contact with eachother.
B. In two-stage conditioning, the mineral slurry was
first conditioned with the desired amount of the collector
at pH 10 for 2.5 minutes followed by reconditioning for 2.5
minutes at a lower pH. The second stage conditioning was
done with and without reducing the collector concentration.
It should be noted that the total tumbling time of 5 minutes
was maintained both in the case of one-stage and two-stage
conditioning techniques to obtain similar attritioning
0 5 1(
Effect of conditioning time on the flotation
of apatite and dolomite at pH 10.
Sodium Oleate Conc. = 1.87XIO04 Kmol/m3
Conditioning PH =I0.1 0.2
100 ml Flasks (Glued).
Diagram of the cell designed to treat apatite and
dolomite together without contacting each other.
effects. The first stage of conditioning at pH 10 was done
as described earlier in the case of conventional
conditioning. Next, 75 ml of the supernatant (S) was
transferred to a clean glass beaker. In order to
recondition the mineral slurry at a lower pH only, the pH of
the supernatant (S) was adjusted to a lower value before it
was poured back into the original flask containing the
solids. Another series of samples was tested by
reconditioning at lower pH with reduced collector
concentration. This was achieved by replacing the
supernatant (S) with triple distilled water of preadjusted
pH (lower than pH 10).
Flotation. After the mineral slurry was conditioned
with the desired amount of sodium oleate, the pH of the
supernatant was measured prior to flotation. The mineral
slurry was next transferred to the Hallimond cell, which was
then connected to the rest of the cell arrangement (see
Figure 18). With the solenoid valve open for 25 seconds and
the flotation cell valve V1 closed, the system was
pressurized with nitrogen gas. This step was necessary to
avoid the significant time lag (10 to 15 seconds) between
turning on the solenoid valve and observing the gas bubbles
emerging from the bottom of the cell. Next the nitrogen
gas, preadjusted to flow at 48+ 3 ml per minute, and the
magnetic stirrer were turned on for exactly 60 seconds,
using the automatic timer. Simultaneously the valve, V1,
was opened manually. This would start the flotation
process, in which the floated particles were lifted up by
the gas bubbles and collected in the stem of the cell.
After the flotation test was completed, the cork at the
bottom of the stem was opened to drain out the floated
material into a clean glass beaker and the pH of this slurry
was measured. The float and sink fraction were filtered on
pre-weighed filter paper and dried overnight at 50 0C in
air. The percent flotation recovery was calculated based on
the dried weight (and P205 content in the case of mixed
minerals) as shown in Appendix 1.
All the experimental parameters were maintained
constant unless otherwise stated.
4.2.2. Oleate Adsorption Tests
Since the oleate adsorption tests were conducted on -
65,+100 mesh size fraction, the samples were conditioned by
the same techniques as used for the flotation tests. The
oleate adsorption was measured by the solution depletion
method. The oleate concentration in the supernatant was
determined by Gregory's method(151) as outlined in Appendix
2. In conducting these tests three main problems were
encountered and resolved as described below:
A. The maximum detectable limit of oleate
concentration in the aqueous solution was determined to be 1
x 10-4 kmol/m3. Gregory's method involves the determination
of oleate concentration by comparing the optical densities
of unknown samples to those of the standards. As shown in
Figure 22, the optical density of oleate solution varies
nonlinearly with the concentration above 1 x 10-4 kmol/m3
All the samples were appropriately diluted so that the final
oleate concentration was within the linear range of optical
B. As observed in Figure 23, at values below pH 10,
the oleate concentration in centrifuged samples was found to
be significantly lower than the actual amount, whereas for
uncentrifuged solutions the oleate content by analysis was
the same as the actual amount added. The discrepancy in
oleate analysis under aidicic pH conditions could be due to
phase separation of the oleic acid. This was confirmed by
observing an oily layer above water after centrifuging the
sodium oleate solution at pH 5 or lower. It should be noted
that since -65,100 mesh size fraction was used for
adsorption experiments, centrifugation of the suspension was
not necessary to obtain a supernatant sample for oleate
analysis. It was also shown by analyzing a known amount of
the added sodium oleate to the supernatant that any trace
amounts of mineral fines generated during the conditioning
step did not interfere with the oleate analysis.
Furthermore, oleate analyses of the uncentrifuged samples
was also observed after they were stored for a prolonged
time. The oleate content of the samples which were
transferred for analysis after 12 hours, as shown in Table
5 10 15
Conc. of Sodium
Oleate, X l-5Kmol/m3
Effect of sodium oleate concentration on the optical
density of its extract prepared by Gregory's
O Not Centrfuged
O Centrfuged at 15000 rpm
for 3 hrs
Total Sodium Oleate Cone. =9.4 X 10
Oleate analysis of sodium oleate standards as a function
of pH with and without centrifugation.
11, was reported to be lower than that of samples taken for
analysis immediately after conditioning. The lower oleate
analysis could also be attributed to the phase separation
upon prolonged storage. Based on the above findings, the
supernatant from the conditioned slurries was immediately
transferred to 75 ml glass tubes for the analysis.
C. Calcium and magnesium ions, dissolved from these
minerals (refer to Figure 16) can bring about the
precipitation of calcium or magnesium oleate soaps in the
alkaline pH range upon addition of sodium oleate. The
precipitation of Ca/Mg oleates would result in a decrease of
the free oleate ions. It should be noted that in Gregory's
method of analysis, the pH of the solution is raised to
11.5. Thus, if calcium or magnesium ions are present in the
oleate solution at any pH, the precipitation of the soaps
would result in lower surfactant analysis. To illustrate
this, 1.87 x 10-4 kmol/m3 sodium oleate solution containing
60 ppm Ca+2 and 35 ppm Mg+2 ions at pH 3.7 (Sol L) was
analyzed. It was determined to contain 1.36x10-4 kmol/m3
oleate ions, instead of 1.87 x 10-4 kmol/m3. It should be
noted that calcium and magnesium oleate precipitates are
very fine. Calcium oleate would settle to the bottom of the
tube only after centrifuging at 15000 rpm, for more than 2
hours. On the other hand the magnesium oleate precipitate
flocculated to a gel type mass, which did not settle upon
centrifuging at 15000 rpm for up to three hours. Therefore,
Table 11. Effect of prolonged storage of collector
solution on its oleate analysis.
Oleate Analysis, Kmol/m3
immediately 12 hours after
1. 1.57 x 10-4 1.40 x 10-4
2. 1.66 x 10-4 1.51 x 10-4
3. 1.26 x 10-4 1.04 x 10-4
to prevent the depletion of the oleate ions due to
precipitation in the bulk solution, the modified Gregory'
method using EDTA(151) was followed. EDTA is a sequestering
agent for the cations and thus its presence would prevent
the precipitation of oleate soaps. The effectiveness of
this modified method was confirmed by observing the correct
oleate analysis of the above solution (Sol L) containing
calcium and magnesium ions. It should be noted that
analyzing by this method would include the oleate ions as
well as any oleate soaps that may have formed in the bulk
solution. Knowing the analyzed oleate content in the
supernatant and the initial amount added, the true oleate
adsorption on the minerals would be obtained.
The above procedures were incorporated in analyzing the
known sodium oleate standards (1.87 x 10-4 kmol/m3) as set
forth in Table 12. (Since the oleate content in the above
standards exceeded the maximum limit of 1 x 10-4 kmol/m ,
the solutions were diluted by half.) The analysis within 95
percent confidence interval was 9.40 x 10-5 + o.41 x 10-5
kmol/m3. This is an accuracy of within 4 percent.
After the supernatant was analyzed, the amount of
oleate adsorbed on the minerals was calculated as
micromole/g as shown in Appendix 3.
4.2.3. Electrokinetic Measurements
Electrophoretic mobility was measured by a Rank II
electrophoresis apparatus. Two tenths gram of -325 mesh
Table 12. Reproducibility of the oleate analysis
pH Added Cations, ppm Oleate Analysis
95% Confidence Interval
Initial amount of sodium oleate in the solution=9.4 x 105
sample was suspended in 100 ml of 3 x 10-2 kmol/m3 KNO3
solution of a desired pH in a glass cylinder covered with
Teflon tape. The container was inverted 10 times and
contents allowed to settle for 10 minutes; 60 ml of the
supernatant containing fine solids was transferred to
another container. After aging the samples for two hours,
they were sonicated in Bransan sonicator (model B-52, 240
watts) for five minutes. After measuring pH of the sample,
the electrophoretic mobility was determined at a number of
electrode voltage settings before the particles settled in
the cell. The pH of the slurry was measured again. Zeta
potential was calculated from the average electrophoretic
mobility values using the following equation.
S= 4(1 ) n/c
where 5 = zeta potential
1i = electrophoretic mobility
n = viscosity
S= dielectric constant
Assuming 5 = 78.5 and n = 0.89 cps under the
experimental conditions, the above relation would reduce to
E= 12.8 (u ).
4.2.4. Transmission Infra-red Spectroscopy
Transmission IR spectroscopy was done on -325 mesh
mineral samples. The sample preparation technique was
similar to that used by Peck et al.(152) A brief
description of it is as follows. One gram of -325 mesh
sample was suspended in 100 ml of water at the desired pH in
a glass cylinder covered with Teflon tape. The suspension
was agitated by inverting the cylinder 10 times and then
allowing contents to settle for 5 minutes. Eighty ml of the
supernatant containing fine particles was reagentized with
50 ml of 10-2 kmol/m3 sodium oleate solution at pH 10 and
reconditioned at pH 4 to 5. In another test, the minerals
were conditioned at pH 4 to 5. The samples obtained at
different stages of the conditioning were centrifuged at
15000 rpm for 5 minutes. The settled solids in the
centrifuge tube were dried at 50 0C for 2 days in the air,
and 0.01 g of dried apatite and 0.005 g of dolomite were
mixed with 2 g of KBr with a diamonite mortar and pestle.
After the KBr mixture had been dried overnight in a vacuum
oven at 80 o C, 0.5-inch pellets were pressed. A MX-1 FT IR
NICOLET Spectrometer was used to obtain the transmission IR
spectra of these pellets.
4.2.5. Bubble Size Measurements
The size of the bubbles in the Hallimond cell was
measured photographically as follows. To facilitate the
viewing of the bubbles, the top part of the flotation cell
was modified as shown in Figure 24. The camera was arranged
as shown in Figure 25. The cell was filled with the
collector solution without any solids in it. After the gas
was turned on for 30 seconds, the photograph of the bubbles
was taken so that solid/liquid interface would appear at the
Diagram of modified Hallimond cell used in
measuring the size of the bubbles.