Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
SEPARATION SCIENCE AND TECHNOLOGY, 27(13), pp. 1743-1758, 1992
Dissolved-Air Flotation of Metal Ions
N. K. LAZARIDIS, K. A. MATIS, G. A. STALIDIS,
and P. MAVROS*
LABORATORY OF GENERAL & INORGANIC CHEMICAL TECHNOLOGY
DEPARTMENT OF CHEMISTRY (BOX 114)
ARISTOTLE UNIVERSITY
THESSALONIKI 54006, GREECE
Abstract
Metal ions (copper, nickel, zinc, and ferric ions) were separated from dilute
aqueous solutions by dissolved-air flotation. The ions were either precipitated as
sulfides or floated (as ions) by xanthates. Copper and nickel were selectively separated; promising results were obtained with single, binary, and ternary mixtures.
The effect of several parameters (solution pH, addition of chemical reagents at
varying concentrations, and the presence of other ions) on the removal of ions was
studied. The collectorless flotation of copper ions was also investigated.
INTRODUCTION
Metals are often contained in industrial effluent, and since several tons
of them are lost each year, there is considerable interest in recovering
them. This has the additional benefit of complying with environmental
conservation policies which are already being implemented.
Similar problems are faced in the hydrometallurgical recovery of metals
(e.g., copper) from ores, where great volumes of aqueous solutions are
often produced, and which have to be treated to recover the valuable ion.
Flotation, a rather unique case of a three phase (gas/liquid/solid) process, is one of the most significant unit operations in mineral processing,
but it may also be applied to water and wastewater treatment ( I ) and to
hydrometallurgy. Flotation is envisaged at present to replace both the
concentration and separation processes, if feasible. In a recent review (2)
it was suggested that one of the main advantages of flotation may be the
selective recovery of ions from complex solutions with several coexisting
metals.
*To whom correspondence should be addressed.
1743
Copyright 0 1992 by Marcel Dekker, Ine.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1744
LAZARIDIS ET AL.
Previous work-in the broad field-involves, among others, the flotation
as hydroxide precipitates of metal-bearing wastes mainly by amines (3) or
by fatty acids ( 4 ) and as sulfide precipitates ( 5 , 6), a comparative study
for a column technique and a flotation machine on copper industry wastewaters ( 7 ) , and the application of LIX reagents (known from liquid-liquid
extraction) in flotation (8). Adsorbing colloid flotation of metals, simultaneously for ferric or manganese hydroxides, has also been applied to
deep-sea nodules ( 9 ) . A short-chain xanthate was used as the collector for
copper ions during a study of the selective separation of copper, zinc, and
arsenic ions from solution by flotation (10). Copper and zinc ions were
separated by precipitate flotation as sulfides in batch ( 2 2 ) and continuous
flow (12) conditions.
In this work the removal of single or multiple metal ions from aqueous
solutions is investigated, and conditions are determined for the removal
of the ions, with or without collectors, in order to achieve total removal
or selective separation.
EXPERIMENTAL
A dissolved-air flotation unit, previously described ( l o ) , was used for
the experiments. Chemical analysis of the solution samples was carried out
by AAS, and the results are expressed as the percentage of initial ion
content removed (or recovered) by flotation.
The reagents used were copper nitrate, nickel sulfate, zinc nitrate, and
ferric sulfate ( p r o analysi grade). The pH of the solution (deionized water)
was adjusted by nitric acid or sodium hydroxide solutions.
Sodium sulfide was applied for the precipitation at a stoichiometric
amount (unless otherwise stated) at a pH of approximately 3.
In the first part of this work-precipitate flotation-laurylamine (0.1 %
ethanolic solution), cetyl pyridinium chloride, sodium lauryl sulfate, and
cetyl trimethyl ammonium bromide were used as collectors. In the second
part-ion flotation-potassium 0-ethyl-dithiocarbonate (xanthate) was
used.
Several flocculants were also tested: Percol-173 (anionic), Magnafloc351 (nonionic), and Zetag-92 (cationic), kindly supplied by Allied Colloids
(UK). When these were used, the solution was agitated at 3.33 Hz for
600 s during the initial flocculation stage. The stirring was later slowed
down during flotation and stopped completely when pressurized water was
introduced into the cell.
Finally, microfiltration tests were conducted using membrane filters having pore sizes of 8.0, 0.45, and 0.05 pm.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
1745
RESULTS AND DISCUSSION
Precipitate flotation generally includes all processes in which an ionic
species is concentrated from an aqueous solution by forming a precipitate,
which is subsequently removed by flotation.
Dissolved air is often used in conjunction with precipitate flotation. Air
is dissolved into water under pressure and, when the pressure is released,
fine air bubbles with diameters of less than 120 pm are produced. The
rising bubbles constitute the process transport medium. These have been
found to float fine particles effectively (13).From the solubility isotherms
of air in water, it has been calculated that the quantity of air released as
fine bubbles from 100 cm3 of pressurized water is roughly 7 cm3, which we
introduced into a 1-L solution of metal ions.
Collectorless Flotation of Copper Ions
Experiments were initially conducted with only copper ions in solution;
sodium sulfide was used to precipitate a colloidal CuS (artificial covellite).
Figure 1 illustrates the results obtained without any additional reagent. Cu
removal drops to about 5% in the pH range from 3 to 6; above pH 6 it
increases back to almost 100% for doses of NazS less or equal to the
Copper
removal (%)
. .
100
80
60
"a2SMCul
40
0.80
0
20
-II16
0
1-00
10
12
1.80
PH
FIG. 1. Effect of sodium sulfide and pH on the removal of copper ions by dissolved-air
precipitate flotation. [Cu2+1: 50 rng/L. Sodium sulfide expressed as percentage equivalent to
copper. Dissolvcd-air flotation recycle 10%.
LAZARIDIS ET AL.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1746
Comer removal 1%1
00
40
20
10
12
PH
FIG.2 . Use of flocculants in the dissolved-air precipitate flotation of copper ions. ( A ) 5 mg/
L; (B) 10 mglL. [Cu+]: SO mg/L. Stoichiometric amount of sodium sulfide. (a) Percol 173;
(b) Magnafloc 351; (c) Zetag 92. [NazS]/[Cuz+]
= 1.0.
% of
CuS passing filter [%I
100 80 -
60 -
40
20 0
10
12
PH
FIG. 3. Microfiltration tests; percentage of copper sulfide passing through filter. [Cu]:
SO mglL. [Na2S]/[Cuz+]
= 1.0.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
1747
stoichiometric amount required. This seems to contradict previous results,
where a 30% excess of sodium sulfide was found to be adequate (If), but
a possible explanation can be traced to froth flotation ( 1 4 ) where sodium
sulfide constitutes a depressing agent for sulfide ores.
Three different polymers [a cationic (Zetag 92), an anionic (Percol 173),
and a nonionic (Magnafloc 351)] were then tested as flocculants; Fig. 2
presents the results obtained with 5 and 10 mg/L. The same flotation
behavior was noticed, more or less, with the anionic and the nonionic
flocculant, but with slightly higher removals. On the contrary, copper removal was always over 80% with the cationic polyelectrolyte (concentration
at 10 mg/L).
As a further test of the natural floatability of the copper sulfide precipitate, the dispersion (after precipitation) was filtered through several filters
having various pore sizes (from 8.00 down to 0.05 Fm). Figure 3 shows
that in the 3-6 pH range the precipitate particles are very fine, and this
may explain why collectorless flotation is so inefficient in this pH range.
Flotation of Copper Precipitates Using Collectors
Several anionic and cationic surfactants were tested as collectors for the
removal of copper precipitate: cetyl pyridinium chloride (CPCl), cetyl trimethyl-ammonium bromide (CTMABr), sodium laurylsulfate (SLS), and
laurylamine (LA). The results are illustrated in Figs. 4 and 5 .
Comer removal [%I
100
80
50
40
0
20
(a) SLS
(b) CPCl
(c) CTMABr
PH
FIG. 4. Use of various collectors in copper flotation: sodium lauryl sulfate (SLS), cetyl
pyridinium chloride (CPCI), and cetyl trimethyl-ammonium bromide (CTMACI). [Cu" 1:
50 mg/L. Surfactant concentration: 10 mg/L. [Na2S]/ICu2+]= 1.0.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1748
LAZARIDIS ET AL.
Cower removal [%I
Cower removal [%I
100 -.
80 -
80
40 -
20 -
\
2
I
6 8
PH
40
J
2 0 - x stoich. Na2S
o 30% exc8ss Na2S
0.5
10
12
PH
Flti. 5. (A) Effect of laurylamine (LA) on copper ion flotation. (B) Excess (30%) of Na,S.
[Cu?']: 50 mg/L. [NazS]/[Cu?+]= 1.0.
Comer removal
[%I
40 -
20 -
10 mg/L LA
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
1749
The first three surfactants display a similar behavior; they were unable
to float the copper precipitate in the 3-6 p H range. Laurylamine, however,
proved more efficient; it successfully removed the copper precipitate when
it was added at a dosage of 10 mg/L (Fig. 5A). An excess of 30% of Na,S
further enhanced its efficiency to approximately 100% across almost the
whole p H range (Fig. 5B). Laurylamine also proved to be as efficient at
higher copper concentrations (250 mg/L, Fig. 6). It should be noted, however, that at this high Cu concentration, even the collectorless removal
was significantly higher.
Generally, it seems most probable that flotation will be required to solve
problems connected with dilute rather than highly concentrated solutions.
Therefore, the observed reduction in copper ion removal in the 3-6 p H
range may be corrected by adding some suitable cationic reagent
(surfactant).
Collectorless Floatation of Multi-Ion Solutions
The removal of the copper precipitate in the presence of other ions was
also investigated. The analysis of an actual scrubber wastewater from a
copper smelting plant (6) was used as a guide for the ionic species and the
respective concentrations that may be found in an industrial effluent.
Zinc ions often coexist with copper. Figure 7 illustrates the effect of
adding 50 mg/L of zinc ions to the copper ion solution (at pH 5 ) . It seems
Ion removal
0.0
[%I
1.0
2.0
3,O
4.0
[Na2Sl/[Cul ratio
FIG. 7. Effect of sodium sulfide and zinc ions (50 mg/L) on copper ion flotation (pH approximately 5 ) . [CU~']:50 mg/L.
LAZARIDIS ET AL.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1750
Copper removal [%I
#
loot
80 -
"
--
60 40
20 OL
0
0
8 1 0 1 2
PH
Sulfate ions
1 ',",",:;:
FIG.8. Effect ofSOi ions on copper ion flotation. [Cu"]: 50 mg/L. (Na,S]/[Cu''] = 1.0.
that the presence of Zn2+enhances the collectorless removal of the copper
ion.
It should be remembered (from Fig. 1) that, at pH 5 , there is very little
removal of Cu2+,even with an excess of 30% of Na2S. When Zn is added,
a 60% removal is achieved for a stoichiometric amount of Na2S.A selective
separation of the two species is achieved at this dosage. A similar selective
separation has been achieved at pH 2 by using laurylamine (11).
If, on the other hand, a total removal of the ionic species is required,
then an excess of over 150% of NazS is required to remove both Cu2+and
Zn?+.
The addition of sulfate ions (200 or 800 mg/L SO:-) had no effect on
the (collectorless) flotation of copper (Fig. 8).
The addition of Fe3+ had, however, a noticeable effect (Fig. 9). The
removal of Cu2+was enhanced in the 3-6 pH region (compared with the
single ion behavior-Fig. l), but then dropped dramatically. At pH over
6, heavy flocs appeared in the cell due to the precipitation of Fe(OH),
which could not be floated under these collectorless conditions. However,
at pH < 3 the selective separation of Cu2+from Fe'+ is apparent, with
recoveries of Cuz+over 80%.
Metal Ions Flotation as Xanthates
Thiol reagents, such as the xanthates, are common collectors in froth
flotation of sulfide minerals (15) and have been extensively studied ( 1 6 ) .
Here, potassium 0-ethyl-dithiocarbonate (KEX) was used for the dissolved-air flotation of several ions.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
Ion removal [%I
100-
80
60
1751
____
40 20 i
0
0
10
12
PH
FIG.9. Effect of iron ions (150 mg/L) and pH on removal of ions. [SO:-] = 800 mg/L.
[Cu?'1: SO m g / L . [Na,S]/[Cu'+] = 1.0.
Copper ions were previously floated with KEX as a yellow precipitate
in the pH range from 2 to 6, displaying satisfactory rates of removal (10).
When Cu2+reacts with xanthate anions, it yields an unstable complex which
decomposes, producing dixanthogen and cuprous xanthate, CuEX.
Trivalent iron was also floated with KEX; the conditioning time was
300 s at an agitator speed of 5 Hz. Stirring was then slowed down to 1.33
Hz for another 300 s to help flocculation. A stoichiometric amount of KEX
was used. Removals of almost 100% were observed for p H values over
4.5 (Fig. lOA), where one (or more) of the xanthate anions in the reaction
product, Fe(EX)3, was expected to bc replaced by OH-.
Iron, depending upon the conditions, precipitates out of solution at
around p H 3.5 and, due to hydrolysis, ferric hydroxo-complexes are
formed. When ferric ions come in contact with a xanthate solution, a dark
brown precipitate forms. This is possibly ferric xanthate, because ferrous
xanthates have considerable solubility in water ( 1 6 ) . It is interesting to
note that a stable ferric xanthate was formed in the case of iron, even
though (under the same conditions) cupric xanthate is reduced to the cuprous state.
Experiments with Ni2+gave similar results (Fig. 10B) for a 10% excess
of KEX over the stoichiometric amount. Removals of over 90% were
obtained for p H values over 4. Even better results, with removals of almost
loo%, occurred at pH over 8, where nickel precipitation (as hydroxide)
LAZAR\D\S ET AL.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1752
20
[KEX] / [Fe3+] = 1.0
I
NI removal [%I
I [KEX] / [Ni2+]
201
1.1
6
1
1
10
PH
FIG. 10. Flotation of single metal (Fe3+,Ni2+)solutions using potassium-0-ethyl dithiocarbonate (KEX).
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
1753
starts. Nickel xanthate, Ni(EX)*, was yellow and colloidal initially; an
activity product of 1.4 x
has been reported for nickel ethyl xanthate
(17 ) .
The above dissolved-air flotation experiments covered a wide pH range.
Thiol anions, which exhibit a high reactivity for heavy metal ions, may
hydrolyze in acidic conditions, forming free acids; the latter are normally
short-lived and decompose rapidly. For example, at p H 2.5 the half-life of
ethyl xanthate is only 120 s (18). In an alkaline medium, xanthate can be
considered as stable during any normal flotation operation.
Tests were performed with binary systems (Cu-Fe, Ni-Fe, Ni-Cu). Figure
11 illustrates the results with the Cu-Fe system; equal concentrations of
the ions were used and the amount of KEX was that stoichiometrically
required for the copper ion only. At pH values over approximately 6,
nearly complete flotation was obtained due to the precipitation of metal
hydroxides. In the more acidic region, copper removal was of the order
of 50%, since part of the xanthate was consumed by the ferric ions. Since
even at low pH values both ions exhibit some percentage of removal, no
appreciable separation may be achieved.
The results with the Ni-Fe system are shown in Fig. 12(A); similar behavior was noticed for pH values over 7, with total removal of both ionic
species. The experiments were repeated without any xanthate collector
00
-.........................................................
.............
80 -
60 40 -
20 -
Fe
[KEXI / [Cu2+1 = 1,O
I
LAZARIDIS ET AL.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1754
ion removal
,oo
80
60
40
20
[%I
.................................................................................................
..........
10
PH
ion removal [%I
....................
60
(B)
0
0
10
PH
FIG.12. Flotation of Ni-Fe solution with and without KEX. [Ni?]
[Fe] = 50 mg/L.
(Fig. 12B) in order to check whether there was any adsorption of nickel
on ferro-hydroxo complexes. The results showed that Ni+ floats only in
the pH region where its hydroxide forms.
When the binary solution Ni-Cu was examined at equal concentrations,
shown in Fig. 13(A), the xanthate was more easily tied to CU*+than to
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
1755
Ion removal [%I
100
............................................
..... ... ..... ..............................................
d
r . r ,
cu
eo 60
40
20
NI
n
[KEX] / [Cu2+] = 1.2
I-b
10
10
PH
Ion removal [%I
100
00
-0
PH
FIG.13. Flotation of Ni-Cu solution. [Ni?'] = [CuL+]= 50 m g L
Ni?+, probably as a result of its lower solubility (1.54 x 1 V S g/L) as
compared to the solubility of nickel xanthate (2.07 x lo-' g/L). Copper
was completely removed over the entire pH range. An increase in the
amount of KEX added resulted in higher nickel removals (Fig. 13B). From
these results it was inferred that an excess of collector would be detrimental
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1756
LAZARIDIS ET AL.
1ooi- - ---- - E3-P-
_.
"
.
PI
fl
I_
10
PH
FIG.14. Flotation of Ni-Cu solution ([Nil
50 rngiL). Effect of Cu concentration.
to the process because it lowered the selectivity of the separation process;
hence, the amount of xanthate used in all subsequent experiments was
stoichiometric toward copper alone.
Figure 14 shows the results with increased copper concentrations ranging
from 100 to 250 mg/L for the same binary Cu-Ni system. Nickel usually
occurs in lower concentrations than copper in the effluents from metalproducing plants. In all cases a similar behavior was observed. Copper
removals were steadily high; at pH over about 7, a 100% removal was
reached due to the simultaneous precipitation of copper hydroxide. On
the other hand, Ni remained in solution; it started floating only when
Ni(OH)2 was formed.
It should be noted that in hydrometallurgy the separation of Ni-Cu
mixtures is conventionally accomplished by liquid-ion exchange ( 1 9 ) .
Therefore, flotation may be regarded as an alternative process, especially
for dilute solutions.
Finally, a ternary Ni-Cu-Fe system was tested. A t low initial Cu concentrations (Fig. 15A), copper and iron floated more or less together,
except at very low p H (<3) where removals of copper reached 50% whereas
iron reached only 25%. Nickel, however, remained in solution until the
pH was high enough for the nickel hydroxide to precipitate. Thus, some
separation may be achieved, especially at pH around 6, with copper and
iron in the foam and nickel remaining mainly in solution. At higher copper
concentrations (Fig. 15B), however, the copper removal was greatly enhanced. Even at p H 2 its removal was over 80%, while the removals of
iron and nickel remained practically the same as before. Therefore, it
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
DISSOLVED-AIR FLOTATION OF METAL IONS
1757
........".........
(A)
J
10
PH
Ion removal [%I
.....
,00 _
"
...
"
..... ................
"
80 (8)
60 -
20 40
Fe
NI
O0
[Cu2+b260mg/L
10
PH
FIG. 15. Flotation of Ni-Cu-Fe solution. [Ni"] = (FeZ+]= 50 rng/L.
should be possible, in principle, to separate the ternary system into two
binary ones, which may be processed as before in order to separate all
three species.
CONCLUSIONS
It is known that flotation presents some advantages over other, more
conventional processes, such as sedimentation, because of its better effluent
quality and, mainly, its appreciably lower residence time, which implies a
lowering of plant costs, including the need for less floor space.
The present work demonstrated that dissolved-air flotation may achieve
two goals: concentration and separation of metal ions. Conditions for the
collectorless flotation of Cu2+precipitates have been determined, and they
LAZARIDIS ET AL.
Downloaded By: [HEAL- Link Consortium] At: 09:40 18 June 2007
1758
permit the selective separation and removal/recovery of this ionic species
from a complex metal-containing effluent.
Acknowledgments
The help of Ms. I. Kosti and Mr. E. Megalonidis in the experimental
part of this work is gratefully acknowledged.
REFERENCES
I. A. I. Zouboulis, K. A. Matis, and G. A. Stalidis. in Innovutions in Flotation Terhnology
(P. Mavros and K. A. Matis, cds.). Kluwer Academic Publishers, Dordrecht, 1992, p.
499.
2. K. A. Matis and P. Mavros, Sep. Purif Methods, 20. 1 (1991).
3 . D. Pearson and J . M. Shirley, J. Appl. Chem. Biotechnol., 23, 101 (1973).
4. L. D. Skrylev and V. A . Borisov. Zh. Prikl. Khim., 51, 434 (1978).
5 . M. C. Kuhn, M. J. Noakes, and A. D. Rovig. CIM Bufl., p. 103 (June 1975).
6. D. Bhattacharyya, A. B. Jumawan, and R. B. Grieves, Sep. Sci. Technol., 14, 441 (1979).
7. W. Charevicz and W. Walkowiak. Environ. Protect. Eng., 8, 67 (1982).
8. N. A. Mumallah and D. J . Wilson, Sep. Sci. Terhnol., 15, 1753 (1980).
9. E. H. De Carlo, B. D. Bleasdell, H. Zeitlin. and Q. Fernando, h i d . , 17, 1205 (1982).
10. G . A. Stalidis, K. A . Matis, and N . K. Lazaridis, /bid., 24, 97 (1989).
11. G. A. Stalidis, K. A. Matis, and N. K. Lazaridis, Int. J. Miner. Process., 24, 203 (1988).
12. G. A. Stalidis, N. K. Lazaridis, K. A. Matis, and I. N. Papadoyannis, Sep. Sci. Technol.,
24, 1033 (1989).
13. K. A. Matis and G. P. Gallios, in Mineral Processing at a Crossroads (B. A. Wills and
R. W. Barley, eds.), Nijhoff, Dordrecht, 1986, p. 37.
14. S. M. Bulatovic and D. M. Wyslouzil, in Complex Sulfides-Processing of Ores, Concentrates and By-products (A. D. Zunkel, R. s. Boorman, A. E. Morris, and R. J .
Wesely, eds.), TMSIAIME, Pennsylvania, 1985, p. 101.
15. I. Bayraktar, U. A. Ipekoglu, and R. Tolun, in Innovations in Flotation Technology (P.
Mavros and K. A. Matis, eds.), Kluwer Academic Publishers, Dordrecht, 1992, p. 307.
16. G. W. Poling, in Flotation-A. M. Gaudin Memorial Volume (M. C . Fuerstenau, ed.),
SME/AIME, New York, 1976, p. 334.
17. D. W. Fuerstenau and R. K. Mishra, in Complex Sulphide Ores, IMM Symposium,
London, 1980, p. 271.
18. I. Kakovsky, in Proceedings, 2nd International Conference on Surface Activity, Vol. 4,
Butterworths, London, 1957, p, 225.
19. J . C. Agarwal, N. Beecher, G. L. Hubred, D. L. Natwig, a n d R . R. Shrab, Eng. Min.
.I.
p.,
74 (December 1976).
Received by editor December 27, I991