International Journal of Mineral Processing, 28 (1990) 139-149 139

Elsevier Science Publishers B.V., Amsterdam - - Printed in The Netherlands

Technical N o t e

M e a s u r e m e n t of the E f f e c t s of P h y s i c a l and
C h e m i c a l V a r i a b l e s on B u b b l e S i z e

C.T. O'CONNOR, E.W. RANDALL and C.M. GOODALL

Department of Chemical Engineering, University of Cape Town, Rondebosch 7700
(South Africa)
(Received October 11, 1988; accepted after revision May 1, 1989)

ABSTRACT

O'Connor, C.T., Randall, E.W and Goodall, C.M., 1989. Measurement of the effects of physical
and chemical variables on bubble size. Int. J. Miner. Process., 28: 139-149.

A method for measuring bubble sizes in a laboratory column cell for two- and three-phase sys-
tems is described. An increase in viscosity, solid-particle size, pulp density and air flow rate and a
decrease in temperature all resulted in larger bubbles. Bubble size decreased with an increase in
the frother concentration and ionic strength of the solution. An increase in pH caused an increase
in bubble size but the presence of xanthate collectors had no effect on bubble size. Reasons for
these observations are proposed.

INTRODUCTION

The importance of bubble size in the efficiency of the flotation process is

well known. Ahmed and Jameson (1985) studied bubble sizes in the range 75-
655 ttm and found that the smallest bubbles ( - 100 ~m) led to the most effi-
cient flotation. They also showed that small bubbles are needed to float fine
particles.
Dobby and Finch (1986) showed that bubble size played a critical role in
column flotation. Smaller bubbles lead to an increase in overall bubble surface
area, more stable froth phases and a greater column carrying capacity. They
found that bubbles of the order of 1 mm diameter were optimum.
Pryor ( 1965 ) states bubbles of less than 0.5 mm diameter are indiscriminate
in their adhering power and tend to promote too stable a froth. If the bubbles
are extremely small, they remain in the pulp and cannot lift themselves. An
upper limit to the size of the bubble for efficient flotation is about 3 mm.
In general the size of a bubble in the pulp phase is influenced mainly by the
following factors: (1) size of the aperture from which it emerges; (2) hydro-
static head against which it is compressed; ( 3 ) surface tension of the interface

0301-7516/90/$03.50 © 1990 Elsevier Science Publishers B.V.

11t~ • ' ! ' i}'('i)NN()t~ ET AI,

formed with the pulp as it emerges; (4) speed of emergence and the volume
and pressure of gas behind it; i 5 ) turbulence of the surrounding pulp.
When a frother is added to the pulp, the surface tension is lowered and the
power requirement reduced. The diameter of bubbles generated in mechanical
cells is generally between 0.05 and 1.5 mm diameter with most having a di-
ameter of about 1 mm. Pneumatic spargers generate bubbles averaging 2.5 to
3 mm diameter.
A number of methods have been described for determining the sizes of"bub-
bles in two- and three-phase systems. These include, inter alia, photographic
techniques (Miller, 1985), electroresistivity measurements (Yasumishi et al.,
1986), gas holdup and pressure measurements (G.S. Dobby and J.A. Finch
private commun., 1987) and calculations using empirical or semi-empirical
correlations etc. (e.g. Sada et al., 1978; Tsuge et al., 1981 ).
Dunne et al. (1976) have developed a method which is described in the ex-
perimental section of this paper. Bubbles are sucked via a capillary tube in-
serted into the pulp through a block containing optical sensors. With the aid
of a computer, bubble-size distributions may be determined in two- and three-
phase systems.
This paper presents results obtained in using this method to determine the
influence of various physical and chemical parameters on the sizes of bubbles
in a small laboratory column cell.

EQUIPMENT AND EXPERIMENTAl, METHODS

A diagram of the system used is shown in Fig. 1. The laboratory column cell
used in this study was a perspex cylinder (25 c m × 6 cm). The bubbles were
generated via a glass frit (porosity 2 ) located at the base of the cell.
The bubble-sizing system consists of a capillary tube with a belled end pass-
ing between two pairs of photo-transistor-LED detectors mounted exactly 5
mm apart. Bubbles are drawn up the capillary under vacuum and are thus
transformed from near-spheres into cylinders. As the bubbles pass the detec-
tors they are monitored as a change in light intensity. This change arises from
the different refractive indices of air and liquid. In order to obtain an accurate
bubble-size distribution about 3000 bubbles are sized. The bubbles are then
collected in a gas burette so that the total volume of bubbles is known.
Pulses are generated by the passage of the bubble past the detectors. Due to
the leading and trailing edges of the bubbles, these pulses are not sharp, but
have rounded edges. To correct this, the pulses are fed into a circuit which
generates sharp up-going and down-going spikes. These spikes are then used
to generate square waves equivalent to the bubble velocity and the bubble length.
These square waves are sent to the data capture system.
The data capture system is based on the Motorola 6809 processor which
controls the operation of a M6845 programmable timer. This timer has three
EFFECTS OF PHYSICAL AND CHEMICAL VARIABLES ON BUBBLE SIZE 141

T-Tap
~-" - - 4
Micro r':~h,'::>
Comput.er ?.---- ,'
Capillary i
Tube Alignment Master [
i D~_~]__ ~ S c r e w " •Water I
I
Reservoir Slave 7 --~,
LED ~ O ~ P hCell
oto-
Timer &l ° ° :
Diode T-Tap ~ ~" V a c u u m Buffer [ . . . . . . ,
Housing
Gas
}urette I

Pulse o .

Belled
Y Generator :B :,,
I
Mercury End
I
Manometer I
J
Fig. 1. Apparatus to determine bubble size.

independent channels, which are used to time the run, the length pulses and
velocity pulses. The length and velocity pulses are stored in memory along with
the time at which they occurred. The capture system can have up to 56 kilob-
ytes of memory (RAM) allowing about 7000 bubbles to be processed. It has
been found that the system is capable of processing more than 50 bubbles per
second. This speed is far in excess of typical operating conditions (about 20
bubbles per second). Detail of the entire system are given by Randall et al.
(1988).
At the end of a run the data is downloaded from the buffers of the data
capture system to a micro-computer. The programme which is used to process
the data calculates the bubble volume from the velocity and length data for
each bubble. The bubble volumes are then normalized with respect to the total
volume collected. This normalization is necessary because a thin film of water
coats the inside of the capillary making the effective capillary diameter diffi-
cult to calculate. From these bubble volumes, a distribution of bubble diame-
ters can be calculated.
Dunne et al. (1976) found that the applied vacuum and the geometry of the
capillary bell can be critical factors in obtaining an accurate bubble-size dis-
tribution. If either of these parameters is incorrect bubble coalescence or
breakage can occur in the capillary.
[,|2 <,T ()'(:ONN{}I~ IqT AI.

RESULTS AND DISCUSSION

Unless otherwise stated the values of the various parameters used in this
study were as follows:
Axial position (capillary tip): 5 cm from fit
Radial position (capillary tip): centre
Air rate: 1.6 1/min Frother type: none
Temperature: 20 ° C Frother dosage: none
Liquid phase: tap water Collector type: none
pH +7 Collector dosage: none
Particle size: 70% < 75/lm Solids: a pyritic ore
Gas hold-up: 0.08 Pulp density: 10% solids

Reproducibility

Two- and three-phase reproducibility studies were carried out to establish

the precision of the results. In an air-water system the mean bubble diameter
was found to be 2.82 mm with an average standard deviation of 0.80 mm while
in an air-water-quartz system the mean bubble size was 3.09 mm with an av-
erage standard deviation of 0.92 mm. The standard deviation gives an indica-
tion of the distribution of bubble sizes. It was found that the reproducibility
was excellent. In the six repeat experiments for the two-phase system a mean
bubble size of 2.822 mm ( a = 0 . 0 1 5 m m ) was recorded. The accuracy of this
data is indicated by its proximity to the values expected for pneumatic spargers
(Pryor, 1965).
Tests were also carried out to establish the effect of the axial and radial
position of the capillary in the cell on bubble size. It was found that, for the
column cell being used, these effects were negligible. There was a slight in-
crease in bubble size near the liquid surface and this is ascribed to the effect of
the head of water and not to coalescence.

Effect of physical parameters on bubble size

Table I (A) shows the influence of air flow rate (superficial gas velocity vg)
on bubble size. As the flow rate increases the bubble size increases. The rela-
tionship between air flow rate and bubble size was found to fit a simple power
law relationship, db = kv~, where x = 0.40 for the two-phase system and x = 0.47
for the three-phase system. Dobby and Finch (1986) found a similar
relationship.
Table I ( B ) shows the effect of using quartz of different particle sizes. In
general the presence of the solid phase produced slightly larger bubbles. Fine
particle s ( - 38 ttm) produced bubbles smaller than those observed in the pres-
EFFECTS OF PHYSICALAND CHEMICALVARIABLESON BUBBLESIZE 143

TABLE I

Effect of physical parameters on bubble size

Parameter Value 2 -phase 3-phase

mean std. mean std.

size dev. size dev.
(mm) (mm) (mm) (mm)

(A) Air flow rate 1.0 1/min 2.36 0.61 2.34 0.73
1.2 1/min 2.54 0.66 2.54 0.80
1.4 1/min 2.70 0.71 2.73 0.90
1.6 1/min 2.84 0.80 2.91 0.96
1.8 1/min 2.97 0.89 3.07 1.00
2.0 l / m i n 3.10 0.96 3.22 1.07

(B) Particle size unsized 3.08 0.89

+ 106/lm 4.00 1.22
75-106 ttm 3.63 1.10
53-75/lm 3.15 0.97
38-53 ttm 2.95 0.86
-38/1m 2.71 0.79

(C) Ore type quartz 3.08 0.91

pyrite 2.98 0.87
pyritic gold ore 2.91 0.85

(D) Liquid viscosity Liquid Viscosity

methanol 0.60 cP 2.38 0.62

benzene 0.65 cP 2.50 0.65
water 1.03 cP 2.83 0.81
ethanol 1.20 cP 3.15 0.98
n-propanol 2.20 cP 3.53 1.12
n-butanol 2.95 cP 3.69 1.20
analine 4.40cP 4.05 1.28

(E) Pulp density* l Density Viscosity

10% 1.13 cP 2.97 0.83

15% 1.20 cP 3.37 0.98
20% 1.29 cP 3.75 1.13
30% 1.54 cP 4.05 1.40

(F) Pulp temperature Temp. Viscosity

and viscosity
2-phase 3-phase

16:C 1.11 cP 1.28 cP 2.90 0.86 2.93 0.98

20 ~C 1.03 cP 1.15 cP 2.83 0.81 3.00 0.91
30 ~C 0.80 cP 0.90 cP 2.74 0.74 3.81 0.86
40~C 0.65 cP 0.73 cP 2.61 0.72 3.86 0.84

*1Solid fraction used was quartz ( 70% < 75/lm )

t |1 ; i ~)'t'()NN~ t~ E'I AI

ence of coarse particles ( + 106 zm ). These results are consistent with the find-
ings of Dobby and Finch (1987).
Table I(C) shows the effect of using different ores. The pyrite and pyritic
ore were finer than the quartz (pyrite: 75% < 15 tim; pyritic ore: 75% < 20 tim;
and quartz: 75% < 53 tim). The differences between the quartz and the other
two ores can be ascribed to a particle-size effect, but the pyritic ore and the
pyrite are almost the same size. Since the pyritic ore consists mostly of quartz
(2% sulphur), it is likely that the differences in bubble size between the pyrite
and the pyritic ore are due to surface charge effects. The point of zero charge
(PZC) of quartz occurs at a pH of about 2 while the PZC of pyrite is around a
pH of 7.
The effect of viscosity is seen in Table I ( D ) and Fig. 2. All these tests were
carried out at 20 ° C. It is clear that as viscosity increased, larger bubbles were
produced. This may be a result of liquid films forming faster at the point of
bubble formation as viscosity decreases, trapping less air in each bubble and
thus forming smaller bubbles. Dobby and Finch ( 19860 showed that the bubble
size increases with the viscosity of the liquid according to a power law
(d~, 3: ti~ ). It was found in this study that his correlation between bubble size
and viscosity for liquids fitted well for the two -phase system (this study: x = 0.26;
Dobby and Finch: x--0.25) but not for the three-phase system. Clearly there

14 "~-" . . . . . . . . . . . . . . . . . . . . . . . . . . T 40

\ T
E
M
y 1.2

•
i " t "
/7 30
P
E
R
A
o T
°o@oo WATER VISCOSITY • U
T1 1.0 W A T E R TEMPERATURE R
PULP VISCOSITY E
Y PULP TEMPERATURE

20 d
C
e
P 0.8 g

0.6 10
2.6 2.7 2.8 2.9 3.0
MEAN BUBBLE SIZE (mm)
Fig. 2. E f f e c t o f v i s c o s i t y a n d t e m p e r a t u r e o n b u b b l e s i z e .
EFFECTS OF PHYSICAL AND CHEMICAL VARIABLES ON BUBBLE SIZE 145

"'o'" LIQUIDVISCOSITY " ~^

PULP VISCOSITY ' / I ~U
V 4 ~ PULP DENSITY ~ ' ~ I uP

Co "

p;i o...e,'"'"'"'
l°Y
5 -

, 4 , , , I 0
2.0 2.5 3.0 3.5 4.0 4.5
MEAN BUBBLE SIZE ( r a m )

Fig. 3. Effectof viscosityand densityon bubble size.

are factors which affect bubble size in three-phase systems other than pulp
density and viscosity.
Table I (E) shows the influence of pulp density on bubble size. The viscosity
of the pulp as calculated from Mooney's correlation (1951) is also shown. These
results are plotted in Fig. 3. Table I (F) shows that the bubble size decreases
as temperature increased for both the two- and three-phase systems. Since
surface tension, like viscosity, decreases as temperature increases the trends
seen in Fig. 2 are expected.

Effect of chemical parameters on bubble size

Table II shows the effect of chemical parameters on the sizes of bubbles. Fig.
4 shows that the presence of frother (Dowfroth 250) reduces the size of the
bubbles. This is due to the lowering of the surface tension by the frother. Table
II (A and B ) indicate that neither collector type nor dosage have an effect on
the bubble size. This was expected because the xanthate collectors used in this
study have no frothing properties. Table II (C) shows that the presence of
frother was the dominant factor in determining bubble size when ore and col-
lector were mixed with frother.
The effects of pH on bubble size are shown in Fig. 5. It can be seen that an
increase in pH causes bubble size to increase. This result is consistent with the
observation that, in the case of pyrite, recovery decreased at higher pH's (Fuer-
stenau, 1982; Dimou, 1986). Apart from other effects associated with surface
adsorption on the collector at these pH's it is also possible that the lower re-
146 ~' I O'(7ONNOR ET AL

T A B L E lI

Effect of c h e m i c a l p a r a m e t e r s on I)ubble size

Parameter Value 2-Phase 3-Phase

mean std. mean std.

size dev. size dev.
(mm~ (mm) (mm) (mm)

(A) Collector t y p e sodium ethyl xanthate 2.83 0.78 2.91 0.90

(0.5 m m o l / l ) pot. a m y l x a n t h a t e 2.81 0.81 2.89 0.93
sod. n - o c t y l x a n t h a t e 2.84 0.82 2.88 0.90

(B) Collector 0.0 m m o l / l 2.85 0.78 2.94 0.91

concentration 0.5 m m o l / l 2.81 0.81 2,89 0.93
(PAX) 1.0 m m o l / l 2.82 0.82 2.90 0.92

(C) Pulp conditioning none 2.85 0.78 2.94 0.91

P A X (0.5 r e t o o l / l ) 2.81 0.81 2.89 0.93
10 p p m D o w f r o t h 250 1.76 0.30 1.63 0.33
P A X + D o w f r o t h 250 1.76 0.30 1.62 0.33

(D) Ionic s t r e n g t h Solution Ionic st

dist. w a t e r 0.00 3.34 1.03 3.28 1.21

tap water 0.04 2.83 0.81 2.91 0.85
0.1 M K C I 0.05 2.70 0.77 2.87 0.90
1.0 M KC1 0.50 2.40 0.71 2.46 0.82

F
R 10
o
T
HE ! ,! ....... WATER i
R 0 ! • PULP I
L J
C
0
N 6
C
E
N
T
R 4
A
T
I
O
N 2

P
P o , , , ~
m
- 1.5 1.8 2.1 2.4 2.7
MEAN BUBBLE SIZE ( m m )

Fig. 4. E f f e c t o f f r o t h e r c o n c e n t r a t i o n o n b u b b l e s i z e .
EFFECTS OF PHYSICAL A N D C H E M I C A L VARIABLES O N B U B B L E SIZE 147

I
14 0.6 0
....,.'"' N
I
12 C
0.5
S
•. le' • T
i0
0.4 R
E
N
8 G
P 0.3 T
H
H 6
0.2 m
- -

O
4 1
.... / "'- \ I ""6- pH (PULP) /
~ T H (PULP) 0.1 1
'"- ..
2 i
t
r
0 0.0 e
2.2 2.7 3.2 3.7
M E A N BUBBLE SIZE (mm)

Fig. 5. Effect of pH and ionic strength on bubble size.

T A B L E III

Effect of cation-collector interactions on bubble size

PBX Mean bubble diameter ( m m )

concentration
( mg/l ) Cu 2+ p b 2+ Zn 2+ Fe 2+
(50 p p m ) (50 p p m ) (50 p p m ) (50 p p m )

0 1.70 1.70 1.69 1.67

5 1.84 1.73 1.31 1.32
10 1.84 1.77 1.24 1.28
20 1.88 1.84 1.21 1.21
30 1.99 1.90 1.20 1.24
40 2.14 1.96 1.19 1.21
50 2.24 2.04 1.19 1.21

coveries are due to a smaller total surface area and lower collection efficiencies
(Anfruns and Kitchener, 1976).
Table II(D) shows that an increase in ionic strength leads to a decrease in
bubble size.Thus the presence of ions in solution leads to larger bubble surface
area and hence a better collectionefficiency.These observations are consistent
with the findings of Barker (1986), who found that in many instances increas-
ing the ionic strength increased the final recoveries in a pyrite/quartz system.
Table III shows the effect of adding various ions to a solution containing
l,t,~ i O'('()NN~d4.E'I AI,

potassium butyl xanthate. In these tests a smaller porosity sintered dis(: l No.
4 ) was used, so these results cannot be directly compared with the other results
in this paper. Copper and lead ions lead to an increase in bubble size. This is
thought to be due to the formation of ultrafine hydrophobic precipitates of
copper or lead xanthates which lead to bubble coalescence due to increased
contact angle. The presence of zinc and iron ions led to a decrease in bubble
size. This decrease can be ascribed to the presence of ions in solution. These
results compare well with those of Livshitz and Dudenkov (1965) who studied
the same variables using a nephelometric and photographic method.

CONCI,USIONS

A number of physical and chemical parameters have been found, by direct

measurement, to have a significant effect on bubble size in both a two- and
three-phase system. Among the physical parameters, increasing the viscosity,
the particle size of the ore, the pulp density and the air flow rate all resulted in
an increase in bubble size. An increase in temperature reduced bubble size.
The reduced viscosity which leads to a smaller bubble also increases collection
efficiency. Of the chemical parameters, the presence of frother and ions in
solution decreased bubble size and an increase in pH increased the bubble size.
In the presence of a collector in a two-phase system, copper and lead ions
increased bubble size, and zinc and iron ions decreased bubble size.

ACKNOWLEDGEMENTS

The authors wish to thank the University of Cape Town and the Council for
Mineral Technology (Mintek) for financial support. The assistance and ad-
vice of Dr P. Harris of Mintek and Mr R.C. Dunne of the Western Australian
School of Mines is gratefully acknowledged.

REFERENCES

Ahmed, N. and Jameson, G.S., 1985. The effect of bubble size on the rate of flotation of fine
particles. Int. J. Miner. Process., 14: 195-215.
Anfruns, J.P. and Kitchener, J.A., 1976. In: M.C. Fuerstenau (Editor), Flotation, A.M. Gaudin
Memorial Volume. AIME, New York, N.Y., Vol. 2, p. 632.
Barker, L.M., 1986. The Effect of Electrolytes on the Flotation of Pyrite. M.Sc. Thesis, University
of Cape Town.
Dimou, A., 1986. The Flotation of Pyrite Using Xanthate Collectors. M.Sc. Thesis, University of
Cape Town.
Dobby, G.S. and Finch, J.A., 1986. Particle collection in columns - - gas rate and bubble size
effects, Can. Metall. Q., 25 ( 1 ).
Dunne, R.C., Forbes, A.W., Harris, P.J., Hulse, N.D. and Moys, M., 1976. The Measurement of
EFFECTSOF PHYSICALANDCHEMICALVARIABLESON BUBBLESIZE 149

Bubble Size Distributions in Flotation Pulps. Council for Mineral Technology, Tech. Mem.
No. 14032, Randburg.
Fuerstenau, M.C., 1982. Sulphide mineral flotation. In: P.R. King (Editor), Principles of Flota-
tion. S. Aft. Inst. Min. Metali. Monogr. Ser., 3: 159-182.
Livshitz, A.K. and Dudenkov, S.V., 1965. Proc. 8th Int. Miner. Process. Conf., New York, N.Y.,
p. 367.
Miller, R.S., 1985. Photographic observations of bubble formation in flashing nozzle flow. Trans.
ASME, Vol. 107, pp. 750-755.
Mooney, J., 1951. The viscosity of a concentrated suspension of spherical particles. J. Colloid Sci.,
6: 162-170.
Pryor, E.J., 1965. Mineral Processing. Elsevier, Amsterdam, 3rd ed., pp. 501-503.
Randall, E.W., Goodall, C.M., Fairlamb, P.M., Dold, P.L. and O'Connor, C.T., 1988. A method of
determining the sizes of bubbles in two and three phase systems. J. Phys. E. Sci. Instrum., 22:
827-833.
Sada, E., Yasumishi, A., Katoh, S. and Nishioka, M., 1978. Bubble formation in a flowing liquid.
Can. J. Chem. Eng., 56: 669-672.
Tsuge, H., Tanaka, Y. and Hiberio, S., 1981. Effect of the physical properties of gas on the volume
of bubble formed from a submerged single orifice. Can. J. Chem. Eng., 59: 569-572.
Yasumishi, A., Fujkuma, M. and Muroyama, K., 1986. Measurement of behaviour of gas bubbles
and gas holdup in a slurry bubble column by a dual electroresistivity probe method. J. Chem.
Eng. J., 19(5): 444-449.