You are on page 1of 39

International Journal of Mineral Processing, 8 (1981) 289--327 289

Elsevier Scientific Publishing Company, Amsterdam -- Printed in The Netherlands

A RATIONAL INTERPRETATION OF THE ROLE OF PARTICLE SIZE


IN F L O T A T I O N

W.J. T R A H A R
C.S.I.R.O., Division of Mineral Engineering, P.O. Box 312, Clayton, Vic. 3168 (Australia)
(Received August 6, 1980; revised and accepted February 13, 1981)

ABSTRACT

Trahar, W.J., 1981. A rational interpretation of the role of particle size in flotation.
Int. J. Miner. Process., 8: 289--327.

From the examination of data from detailed plant surveys and associated laboratory
batch testing, the principal effects of particle size in flotation have been identified. The
current state of knowledge concerning the role of this variable is discussed in terms of the
evidence presented. It is concluded that the minimum degree of hydrophobicity neces-
sary for the flotation of a particle depends upon its size and as a result, recovery-size
curves are a valuable diagnostic aid to the assessment of flotation performance. Entrain-
ment is shown to be an important contributory mechanism to the recovery of fine par-
ticles which, when coupled with a low rate of genuine flotation, can account for much of
the observed behaviour of such fines. The significance of particle size and its consequences
in flotation research, in plant operations and in control schemes has been under-rated.
The separate conditioning or flotation or both of separate size fractions seems inevitable as
ores become increasingly difficult to concentrate.

INTRODUCTION

Although many of the effects of particle size on flotation have been known
since the pioneering paper by Gaudin, Groh and Henderson in 1931, there
has been little acknowledgement of the diagnostic value of size by size assess-
ment of flotation performance either in research or in practice, nor has there
been any significant effort to utilize this knowledge in the design, optimiza-
tion or control of flotation plants.
In this paper, the influence of particle size is discussed in terms of experi-
mental evidence from detailed plant surveys of concentrator operations in
several Australian mills (see, for example, Kelsall et al., 1974) and from re-
lated laboratory batch flotation tests. All of the plant data and much of the
batch data apply to flotation systems for which the conditions have been op-
timized by the best methods available. The information is presented mostly
as recovery-size curves in which the recovery from a particular size fraction is
plotted against the average size of particles in that fraction.
These curves have a characteristic shape, illustrated in Fig. 1, which may be

0301-7516/81/0000--0000/$02.50 © 1981 Elsevier Scientific Publishing Company


290

g
IO0

9O
-
cA '
7-15

CELLS
4-6
]

{ ,0 CELLS
1-3

60~ LEAD FLOTATION


IN ROUGHERS
O
D- BROKEN HILL SOUTH L T D
501 -
Z
W
40~

>- 30

FINE INTERMEDIATE COARSE


2O
O REGION REGION REGION
I

1C

i
2 10 20 50 100 200 500 t000
AVERAGE PARTICLE SIZE (Urn)

Fig. 1. Cumulative lead recoveries in lead roughers at Broken Hill South Ltd.
divided conveniently but arbitrarily into three regions. The region below 5--10
pm comprises the fines which are difficult to float and more difficult to sepa-
rate; the region from 10 pm to, say, 70 pm comprises the intermediate particles
which are usually the most floatable and that above 70 pm and below some
more or less undefined upper limit is the coarse particle region where flota-
tion may be easy or difficult according to the mineral and the conditions. Be-
cause the boundaries between the three regions are frequently ill-defined and
their locations variable, it is often more convenient to refer to fine, intermedi-
ate or coarse particle behaviour than to nominate specific sizes.
Recovery-size curves are usually fairly stable in shape for a given set of
flotation conditions. The data in Fig. 2 illustrate the reproducibility of the
size by size recoveries of cassiterite in the Renison tin flotation circuit over
a number of years, while it is clear from Fig. 3 that in laboratory batch tests,
the shape of the curve is reasonably independent of moderate variations in
particle size distribution, valuable mineral content and pulp density. Such be-
haviour is to be expected from any flotation system the kinetics of which can
be described by an exponential decay; the particles in any given flotation com-
ponent should not then interfere with the particles in any other flotation com-
ponent. However, interference between sizes may occur in pulps with high
contents of solids or of floating minerals or in which mutual interactions be-
tween particles, such as slime coatings, are possible; the shape of the recovery-
size curves may not then be independent of the flotation feed sizing or com-
position. In addition, it is necessary that a sufficient number of particles be
present in any size range for their recoveries to be statistically meaningful.
For the most part, only liberated particles will be considered. The principal
recovery mechanisms are presumed to be genuine flotation (i.e. bubble attach-
291

~oo[------T- i -- ] ~ F---]- l --T-

(~ 80 CASSITERITE -
: RENISON PLANT DATA =
,o~- o~ ~
6o- <~

11

"~ 20F-

L !
1 2 5 I0 20 50 100 200 500 1000
AVERAGE PARTICLE SIZE (pm)

Fig. 2. Average recovery of cassiterite in Renison tin flotation circuit taken from several
plant surveys.

1001 ] I I ~ I~ ~ ~ ] ---7--

e °i-
,<
Clt"
//~
}"'i
U,. /
70t /'// CHALCOPYRITE-QUAIqTZpH
11

O~ 6O // Eh 250 mV
rr ~/ KEX 0
h / PPG 400
~o~
W
O
~ 40[ %SOUDS %CHALCOPYRITE %Cp PASSING
200 MESH
3O O 22 6.7 61
[~ 10 67 61
~, ~ ~ 16 lOO 61
20, ~ 22 10.0 57
LLI
Dr | ~ 16 10.0 41
10I- ~ 16 10.0 71

I I L~ I I ___ I
I 2 5 10 20 50 100 200 500 1000

AVERAGE PARTICLE SIZE (pm)


Fig. 3. Reproducibility of batch flotation of chalcopyrite at pH 11 from chalcopyrite-
quartz mixtures for varying pulp density, valuable mineral content and particle size distribu
tion. (KEX is potassium ethyl xanthate; PPG400 is polypropylene glycol of average
molecular weight 400; Eh is the potential referred to S.H.E. of a platinum electrode in
the pulp. )
292

m e n t and levitation) and ent r a i nm ent (i.e. carry-over with water which enters
the c o n c e n t r a t e via the froth). The scarcity o f i nform at i on about size effects
on o t h e r possible recovery mechanisms, including e n t r a p m e n t in the frot h
(Gaudin, 1957, pp. 362--363) and carrier flotation (Greene and Duke, 1962)
precludes any useful assessment of their significance. The influence of slime
coatings (Hemmings, 1978), of froth modification by fines (Lovell, 1976) or
o f possible size effects associated with the return o f particles from froth to
pulp are n o t considered.
It should be n o t e d t ha t although e nt r ai nm e nt can in some respects be studied
separately f r o m flotation, the converse is n o t true, for the recovery of mineral
in a conventional flotation cell, batch or plant, is always a com bi nat i on of flo-
tation and en tr ain m e nt and separation of the individual contributions is pos-
sible only to a limited extent.
The aims of the present paper, therefore, are to identify the essential effects
o f particle size for which reliable experimental evidence exists, to resolve some
o f the uncertainties associated with the influence of this variable and to draw
a t t en tio n to conclusions concerning its application in realistic flotation systems.
T h e scheme used as a basis for presentation is outlined in Table I.

TABLE I

Outline of particle size effects in flotation

Recovery of fine particles


Recovery by flotation
Recovery by entrainment
Combined effects of flotation and entrainment in sulphide and non-sulphide systems
Methods for improving flotation in the fines region
Particle size, surface energy and adsorption density

Recovery of intermediate size particles

Recovery of coarse particles


Basic flotation behaviour of coarse particles
Interpretation of coarse particle behaviour

Discussion
Theoretical background
Surface coverage and floatability
Particle size and selectivity
Composite particles
Comparison of recovery-size curves from laboratory and plant data
293

RECOVERY OF FINE PARTICLES

Recovery by flotation

The efficient flotation of fine particles has probably received more atten-
tion than any other aspect of particle-size effects, as recent reviews (Trahar
and Warren, 1976; Jameson et al., 1977) indicate. The significant conclusion
is that the flotation rate of fines is low relative to that of other sizes, pri-
marily as a result of the decreased probability of collision between particles
and air bubbles as the particle size is reduced (Sutherland, 1948; Flint and
Howarth, 1971; Reay and Ratcliff, 1973). However, the form of the relation-
ship between flotation rate and particle size is not yet clear. The most recent
theoretical analyses suggest that:
k cx d n

where k is a suitable measure of the flotation rate, d is the particle diameter


and n is a number between 1.5 and 2. There is some support from microflota-
tion tests and single particle-bubble encounter experiments (Jameson et al.,
1977) for this relation but the information available from flotation tests
(Gaudin et al., 1942; Morris, 1952) is indicative of a linear relation between k
and d. The results shown in Fig. 4, in which the rate constants of several
floating minerals in controlled batch tests are plotted against particle size,
support a roughly linear relationship. The rate constants were calculated from
the average flotation rate over the first minute of flotation, which is long
enough to reduce timing errors but short enough to limit the contribution

10~ -q
!

i,

,z ,

Il
0 ~
ALERITE pHB
0.5~-
ENA pH8 j

r pH8
pH5 I
E prig
O2 ~ lOOmg KEX
I
r
0 1l - - J
1 2 5 10 20 50 100
AVERAGE PARTICLE SIZE ( prn )

Fig. 4. Variation in f l o t a t i o n rate c o n s t a n t w i t h particle size for the batch f l o t a t i o n o f


several sulphide minerals with e x c e s s collector.
294

from slow floating components and entrainment. The data are typical of m a n y
results and even allowing for a contribution from entrainment to each curve,
it is unlikely that the slopes would become steep enough to support a value
of n much greater than unity. There is no evidence in Fig. 4 of a levelling in
flotation rate in the fines region of the type predicted by Derjaguin and Dukhin
(1960--1961) or observed by Gaudin et al. (1942) in the flotation of galena.
On occasion it has been observed that the highest flotation rates are dis-
played by fine particles in circumstances where the effect can not be entirely
accounted for by entrainment. An example is presented in Fig. 5 which is a
100F T-- F E I I I l I

90 CHALCOCITE
KEX 0 4
~__~ 80~ ~ pPG 5rag/rain
[
( rain ) CHALCOCITE WATER [
,o o~ 13~ 167 I

i L~ . . . . 3°3

Z ~ 8 28 4 648

3°L

oo ~0

10 4

0 1 i •
1 12 5 110 210 510 100 200 500 1000
A V E R A G E PARTICLE SIZE (IJm)

Fig. 5. B a t c h f l o t a t i o n o f c h a l c o c i t e at p H 11 in t h e a b s e n c e of c o n v e n t i o n a l collectors.

recovery-size-time plot of the response of chalcocite in the absence of con-


ventional collectors. Entrainment was unlikely to be the sole recovery mecha-
nism, for the maximum recovery was not in the finest fraction and, further-
more, exceeded the recovery of water. Floatability of this type might be at-
tributed to the presence on the mineral of a small a m o u n t of sulphur produced
by oxidation, to adsorbed frother molecules which is possible but unlikely,
or to a contribution from a minor but more floatable mineral such as bornite
intergrown with the chalcocite (Heyes and Trahar, 1979). The increase in re-
covery with time was significant only below about 20 #m which is indicative
of weak hydrophobicity.
The data in Fig. 6 demonstrate the response of chalcocite after conditioning
with potassium ethyl xanthate (KEX). When compared with those in Fig. 5, it
can be seen that the rate of recovery of fines was only slightly increased. How-
ever, the initial water recoveries for the test in Fig. 6 were low and the com-
parison is more realistic if the recoveries of fines are plotted against that o f
295

9oF 6 ,N

70

2MIN I
o~ 60
LL ! CHALCOCITE
I'-- 50~ pH 11
~ 401~_, 1 MIN Eh 175 mV
o KEX 10mg

~ 30! O,,*/ RECOVERY%


LU ~ 0 5 MIN CHALCOCITE WATER
2oL oE] 82.6
90.9
46
10.2
/~ 93.9 18.3
it" 10 ~ 95.7 36.2
/ 96.6 63.3
0[~ ~ • ~ [ I L _J _ _ __1___
1 2 5 10 20 50 100 200 500 1000
AVERAGE PARTICLE SIZE (#m)

Fig. 6. Batch f l o t a t i o n o f c h a l c o c i t e in presence o f KEX.

water, which offsets any entrainment contribution. Figure 7 shows such a


plot; it includes the data for the fines from Figs. 5 and 6 and from two ad-
ditional tests included for verification. Although the influence o f a collector
on the flotation rate of the fines is shown to be greater than a comparison of

<[
'01i ESTIMATED
~O'jr"
iT ? 0//,~/ " / //ENTRAINMENT

8
>- ! 7" ,/ / ]
~" oI / /~ //"
~'3
LU 30~-- ] _ ~ /" CHA!COCITE
rr

ul
.-
0
U,J
(3.
!/ / ':"/
10~-~ / " /
r/ / / / /
• n-HEXANOL KEX
• 4 1PPG
G • 4DO KEX
i
1

0 10 20 30 40 50 60 70 80 90 100
PER CENT RECOVERY OF WATER

Fig. 7. R e l a t i o n b e t w e e n recovery o f fine c h a l c o c i t e in batch f l o t a t i o n in presence and


absence o f x a n t h a t e and that o f water. (The s y m b o l s o and v refer to similar but separate
tests).
296

Figs. 5 and 6 indicates, the change in response is still small compared with the
very large changes displayed by intermediate and coarse particles. Such be-
haviour leads to the conclusion that the addition even of a collector which
results in the formation of a highly hydrophobic reaction product -- for few
minerals float as fast as chalcocite when treated with xanthate -- produces a
relatively small change in the response of fines.

Recovery by entrainment

The recovery of fine particles by entrainment in a flotation froth has at-


tracted sporadic attention over a number of years (Gaudin et al., 1931, 1942;
Sutherland, 1948; Jowett, 1966; Johnson et al., 1974}, but the prominent role
of this mechanism has been under-rated. A proportion of supposedly typical
fine particle behaviour can be attributed to entrainment, the contribution of
which is often of such a magnitude that the true flotation rate of fines is lower
than it appears to be.
Recovery-size curves depicting the entrainment of quartz in batch flotation
tests, in the absence and presence of strongly floating chalcopyrite, are pre-
sented in Fig. 8 and demonstrate that there is a steady increase in quartz re-

I I IQUART z I I
! BATCH FLOTATION ,
Z
40t~
~..W,T, E.OT,ERONLY~
O CHALCOPYRITE
° i
<
rr
u_
20;

lo,
©
rr

QUARTZ
BATCH FLOTATION
(.) 4@
n- ' WiTH FROTHER ONLY
<:ALCO ..... E POESENT
'.AN

\
O 2o~
O I

ol
i 2 5 10 20 50 100
AVERAGE PARTICLE SIZE (pm)

Fig. 8. R e c o v e r y o f q u a r t z in b a t c h f l o t a t i o n b y e n t r a i n m e n t in t h e p r e s e n c e a n d a b s e n c e o f
s t r o n g l y f l o a t i n g c h a l c o p y r i t e . F l o t a t i o n t i m e s in m i n u t e s were o 1/2; a 1;/~ 2, v 4; 0 8.

covery below about 50 pm. The relation between fine (< 5 pm) quartz recovery
and water recovery is given in Fig. 9 using data from a number of batch tests.
The slope of the line is 0.72, i.e. the recovery of the finest quartz size range
was 72% of that of the water; any variable which changes the recovery of
water will therefore change the recovery of fine quartz. Data for other sizes are
given in Table II.
297

FROTHER PULP STRONGLY 4


ADDITION DENSITY FLOATING
SLOPE 0.72 i
i ( mg/min) (%SOLIDS) CHALCOPYRITE \ . ~ I
~= , o 2.~ ~6 NO \ oi,-" 4
5 :: :o S ,, !
U ~_ ~ 2,5 16 YES ~
<= / • 5 2, NO ~, !
/ • 5 ~6 NO ~ !
3or • ,o 2, NO ~ ]

J
20 FROTHER PPG 400

10 I

oL// [ L ] l • _ i 1 • L ~ ] • •
0 10 20 30 40 50 60 70
WATER RECOVERY PERCENT
Fig. 9. Recovery of fine (< 5 ~m) quartz by entrainment compared with recovery of
water in several batch flotation tests with and without floating chalcopyrite.

TABLE II

Influence of particle size of quartz on its entrainment

Average particle size Recovery of quartz relative


(/~m) to that of water
(%)

40 4
29 18
20 35
14 45
9 59
3.5 72

A d i s t i n c t i o n s h o u l d be m a d e b e t w e e n e n t r a i n m e n t and w h a t m i g h t be
t e r m e d " e n t r a p m e n t " (Gaudin, 1957, pp. 3 6 2 - - 3 6 3 ) in w h i c h n o n - f l o a t a b l e
particles are t h o u g h t to be held in the f r o t h b y bridging across floatable par-
ticles held b y adjoining bubbles. The absence o f differences b e t w e e n the t w o
series o f curves in Fig. 8 suggests t h a t e n t r a p m e n t was n o t significant in this
example.
298

Combined effects of flotation and entrainment in sulphide and non-sulphide


systems

The low rate of true flotation of fine floatable mineral and the entrainment
of both floatable and non-floatable minerals are factors which are common to
nearly all flotation operations. In combination, they can account for much
of the observed behaviour of fine particles.

Sulphide systems
Figure 10 presents detailed information on the separation of chalcopyrite
from quartz in one of the tests from Fig. 3. In the first minute the recovery

1OC - - - - 50
/ , O " " - - ~ ' " - ~~ I 1E MIN,
9(3 RECOVERYOF / t" ~-~
CHALCOPYRITE'"~//"
~ ) 1 MIN.
8(3-- (~/ 4O
/
/ ('b
0
Z
// / C)
m
--30 Z

5o~- j -- m
6~
"~o
l I m

~o
1 MIN. I
16
J

2 s lO 20 so Vloo " 200 500


Io
1000
AVERAGE PARTICLE SIZE (pm)

Fig. 10. E f f e c t o f e n t r a i n m e n t o f quartz on grade o f c h a l c o p y r i t e c o n c e n t r a t e in a typical


batch f l o t a t i o n test at pH 11 w i t h o u t collector. ( O p e n s y m b o l s are for 1 rain and closed
s y m b o l s are for 16 rain f l o t a t i o n time).

of all chalcopyrite coarser than 10 pm was more than 85% compared with 45%
for the finest fraction. The recovery of quartz was less than 10% throughout.
After a further 15 min the recovery of the finest chalcopyrite had risen to
97% but at the same time the recovery of the finest quartz by entrainment
had risen to 58%. Thus a high recovery of chalcopyrite in all sizes was at-
tained only at the expense of concentrate grade in the fine sizes. It should
be noted that no collector was used in this test, the only reagent additions
being pH modifier and frother; it is not possible, therefore, to attribute the
low grade in the fine sizes of the concentrate to the non-selective adsorption
of collector. The possibility that the frother PPG 400 (polypropylene glycol
299

of an average molecular weight of 400) was responsible for inducing genuine


flotation of fine quartz is remote. Compounds of this type have been shown
not to adsorb on quartz at pH 11 (Doren et al., 1975) and further evidence
against a collector function is given in Figs. 9 and 11 where the fine quartz
recovery is shown to be related to the flow of water to the concentrate rather
than to the quantity of frother (Fig. 9) or to the type of frother (Fig. 11).

50 F - T "--~ ~ -- T T i

QUARTZIN PRESENCE OF
STRONGLY FLOATINGPyRRHOTITE

MIBC 10 / rain
/ PPG 400 5 mg/chin ///~
30~ DAA ,Omg/m'n/ / ~ \ ~ 0 ~ 0°5

'° J

0 15 20 30 40 50 60 70

PER CENT RECOVERY OF WATER

Fig. 11. R e c o v e r y o f fine ( < 5 #m) q u a r t z b y e n t r a i n m e n t c o m p a r e d w i t h t h a t o f w a t e r


in t h e p r e s e n c e o f s t r o n g l y floating p y r r h o t i t e a n d d i f f e r e n t f r o t h e r s (MIBC is m e t h y l
i s o b u t y l c a r b i n o l ; P P G 4 0 0 is p o l y p r o p y l e n e glycol o f average m.w. 4 0 0 ; D A A is d i a c e t o n e
alcohol; o = MIBC; u = PPG; n = DAA).

The same data may be used as a basis for examining the influence of feed-
size distribution on the efficiency of the separation. From the detailed kinetic
data for chalcopyrite given in Table III and from the entrainment figures for
quartz given in Table II, it is possible to calculate curves of recovery versus

T A B L E III

Size b y size rate d a t a for t h e b a t c h f l o t a t i o n o f c h a l c o p y r i t e

Size F l o t a t i o n rate c o n s t a n t Fraction with rate constant


range (min-')
(~m)
Fast Intermed. Slow Fast Intermed. Slow
(kf) (ki) (ks) (~f) (~i) (~s)
> 208 2.56 0.57 0.011 0.6t~ 0.23 0.11
104--208 3.06 0.44 0.018 0.85 0.09 0.06
53--104 4.27 1.50 0.025 0.76 0.21 0.03
25--53 6.58 1.36 0.045 0.95 0.03 0.02
12--25 4.13 0.054 0.97 0.03
5--12 1.84 0.090 0.92 0.08
< 5 0.85 0.145 0.72 0.28
300

enrichment ratio (ratio of concentrate grade to feed grade) for increasingly


finer feeds. For the calculations, water flow rates to the concentrates were
based on the measured flows in the original tests from Fig. 3 (280--320 ml
per min) and feed-size distributions were derived by arbitrarily shifting the
original distributions for chalcopyrite and quartz along the size axis. It was
assumed that neither the water flow rates nor the flotation rate parameters
were influenced by the feed size distribution, i.e. no account was taken of
possible interference to chalcopyrite flotation of the increasing proportion
of quartz fines (Hemmings, 1978) or to a change in quartz behaviour from
the increasing proportion of chalcopyrite fines. The results of the simulations
given in Fig. 12 demonstrate that, even for a near-ideal system comprising a
totally liberated, strongly floating mineral and a non-floating mineral, the ef-
ficiency of separation is dependent upon the degree of gangue entrainment
and becomes progressively lower with decrease in feed size for otherwise
identical conditions.
2 0 r ~ ~ I ~ v T ~ ~ ~ - - I
i

18~ CHALCOPYRITE - QUARTZ FLOTATION

0 60% < 74 pm
~_ [ ] 75% < 74 pm
16 ~ 95% <: 74 prn
~ ) 95% < 53 pm
14~

_o ! o

p- i
<~ o.__~._

4r

2~

oL ~ i ~__ I L_ I i I _ I
0 10 20 30 40 50 60 70 8Q 90 100

RECOVERY PER CENT

Fig. 12. Influence of feed size distribution on grade-recovery curves for batch flotation of
chalcopyrite from quartz at pH 11 without collector.

It might be argued that entrainment does not present a serious problem


in practical sulphide flotation because entrained gangue can be rejected by
cleaning at reduced pulp densities and this is usually true provided enough
valuable mineral is in the intermediate or coarse size regions. However, when
most minerals are present as fine particles, separations even of sulphides may
become difficult. An extreme example is provided by the McArthur River
lead-zinc prospect in northern Australia in which entrainment probably plays
301

a major role in limiting the separation efficiencies attainable by conventional


flotation.

Non-sulphide systems
In non-sulphide systems, entrainment can sometimes assume a d o m i n a n t
role. When cassiterite is floated in batch tests from flotation feed from the
Renison concentrator in Tasmania, using p-tolyl arsonic acid (pTA) as col-
lector, it has been found that for collector additions up to 0.5 kg per tonne,
the gangue minerals (siderite, tourmaline and other silicates) are recovered
almost entirely by entrainment. A typical result is shown in Fig. 13. When

iooF

oo-

Z i
~ 80F
(J
\
O CASSITERITE
SIDERITE
~ 6o~ TOURMALINE
GANGUE
0
~t.~_5o~
N
>" 40~
rr
N

20~

,o!
0! 5 10 2O 5O t00
AVERAGE PARTICLE SIZE (urn)
2~)0 50T 100o

Fig. 13. B a t c h f l o t a t i o n o f sample o f f l o t a t i o n feed f r o m R e n i s o n tin c o n c e n t r a t o r using


p-tolyl arsonic acid ( p T A ) as a c o l l e c t o r at p H 5.5.

the recovery of fine gangue was plotted against that of water (Goodman and
Trahar, 1977) the curve was similar to that shown in Fig. 9 but with a higher
slope (0.85) probably reflecting the different character of the froth. When
the collector addition was increased, or when stronger collectors were used,
there was an increasing contribution to gangue recovery from genuine flota-
tion which can be seen as the development of a hump in the recovery-size
curves. Such behaviour, which is shown for tourmaline in Ffg. 14, produces
unusual curves in which the recovery is high only below 30 pm and falls
rapidly above 40 pm. It is only when there is a positive contribution to gangue
recovery from genuine flotation that depressants are of value in cassiterite
flotation; if used in circumstances such as those in Fig. 13, they serve no use-
ful purpose because depression of cassiterite is the only result (Goodman and
Trahar, 1977).
302

ioo r , ' ' ' '


TOURMALINE ' 1.
O pTA 225g/t pHS,5
90 [:] CA540 200g/t pH2.5|
/k CA540 100g/t pH6.0|
< 80 ~:~ CA540 200g/t DH6.0~
! ~ pTA" 45Oglt pH6.0l!

• Fe activated
'°i -
60~
I
i

s°I
~ 30~ ~

~) 20-- -
O

~: ,0~
1 2 5 10 20 50 100 200 500 10c0
AVERAGE PARTICLE SIZE (lJm)

Fig. 14. Change in shape o f recovery-size curves for batch f l o t a t i o n o f tourmaline as the
mineral b e c o m e s more floatable. ( C A 5 4 0 is a c o m m e r c i a l l y available s u l p h o - s u c c i n a m a t e . )

A qualitative assessment of the relative contributions of entrainment and


flotation is possible for curves of this type as shown in Fig. 15 for siderite. An
entrainment contribution calculated from the measured water recoveries,
using data of the type given for quartz in Table II, was subtracted from the
100 p IC ~ I F • I I } I [
Z 90 ,/s~'~/' SIDERITE WITH pTA COLLECTOR
0 0 16 MIN. ORIGINAL
~_ 1"7 4 MIN. FLOTATION
f'~
< 80 ~ 1 MIN. CURVES
tr
IJ- ENTRAINMENT
• 16 MIN. } DATA FROM
70 • 4 MIN. AUXILIARY
• 1 MIN.
:~ 60 ~
/11/' \~"~
~I1 TESTS
-- ~ ~ DIFFERENCE CURVES
l'- 50
Z
LLI

O 40
rr
U.I • "~ I
= '\ "/t
",, ',\ t
~ 2o

~ 10

I [ _ L I I
2 5 10 20 50 100 200 500 1000
AVERAGE PARTICLE SIZE (tJm)
Fig. 15. Qualitative separation o f recovery-size curve into e n t r a i n m e n t and f l o t a t i o n com-
p o n e n t s b y subtracting a calculated e n t r a i n m e n t curve f r o m the e x p e r i m e n t a l recovery
curve.
303

measured recoveries in each size range. The resultant curve, shown dotted,
is taken to be an estimate of the true flotation c o m p o n e n t which, though
positive, is indicative of a weaker response than is generally supposed to be
typical for this mineral. It is not strictly possible to separate recovery-size
curves by subtraction b u t the illustrated distinction between entrainment and
flotation provides a useful approximation of the contribution from each
mechanism.
The batch results were supported by data from surveys of the Renison flo-
tation plant (Fig. 16) which showed that the recovery curves for siderite were

100 [ I I I I r [ I

PLANT FLOTATION
AT RENISON
u. zq- 0 o
O CASSITERITE
SIDERITE

g i
~ 50--

~ 4O--
m

]
1 2 5 10 20 200 500 1000
AVERAGE PARTICLE SIZE (~m)

Fig. 16. Recovery of cassiterite and siderite in the Renison final concentrate with relation
to particle size.

similar to those from batch tests. This is interpreted as indicating that siderite
was recovered in the final concentrate largely by entrainment. From a detailed
analysis of a typical survey coupled with extensive batch testing, it has been
concluded (Goodman and Trahar, 1977) that pTA is a weak collector for cas-
siterite such that an acceptable rate of flotation can be achieved in the plant
only by the use of high collector and frother additions, high pulp levels and
high aeration rates to create a high froth flow rate. Such conditions inevitably
result in a high recovery of water in the concentrate (in the rougher-scavenger
circuit 55% of the incoming water went to rougher concentrate and a further
28% to scavenger concentrate) which leads to a high level of entrainment of
fine gangue that, despite initial desliming, comprised 28% by weight of the
flotation feed. Although some rejection of entrained gangue is possible during
cleaning, the weak floatability of cassiterite in all sizes results in low cleaning
efficiencies because it is easily lost from the froth; in the later cleaning stages,
entrainment accounts for much of the recovery of cassiterite itself judging by
304

the shape of the recovery-size curves (Goodman and Trahar, 1977). Recovery
in the finest fractions is therefore maintained at the cost of a substantial drop
in concentrate grade (see Fig. 17).
Results of this nature suggest that one of the principal benefits to be
gained from desliming prior to tin flotation may be the minimization of en-

100 F I ~ ] r 50

(.~ 80 RENISON (RB600) 40


~/~ TIN FLOTATION
U- CONCENTRATE (~)
I.,tJ 70
N / / \ RECOVERY i Z

!50~
:oi m

3° ~
> 20~
o
CONCENTRAT~- \ ~ 0
o • I

rr lcl

0L ~ ~ ~ -- ~ 0
1 2 5 10 20 50 100
AVERAGE PARTICLE SIZE (~lrn)

Fig. 17. E f f e c t o f particle size o n r e c o v e r y o f cassiterite a n d c o n c e n t r a t e grade in t h e


R e n i s o n final c o n c e n t r a t e .

trainment of fine gangue. This is illustrated by data given by Pol'kin et al.


(1973) who investigated the influence on tin flotation of progressively in-
creasing additions of tin ore fines to a deslimed feed. As the a m o u n t of < 13
pm material was increased from 0 to 20%, the weight of concentrate increased
from 5 to 17%, the grade decreased from 5 to 1.8% Sn and the recovery in-
creased from 52 to 72%. Although these results have been attributed to the
deleterious effects of slimes on flotation, it is possible to account for them by
assuming that the added fines were of head grade in the range 0.43 to 0.48%
Sn and that a realistic 60% of the a m o u n t added was entrained with the
minerals which floated. The new concentrate would then have comprised 17%
by weight with an assay of 1.8% Sn and a recovery of between 65 and 72%
according to which calculated head assay was used, this quantity being variable
for the examples quoted. In terms of this interpretation, the inclusion of the
fines had no effect on the flotation of the tin mineral as such.

Methods for improving flotation in the fines region

It is clear that entrainment sets a lower limit to the size at which conventional
flotation will be efficient and that the size and the efficiency will depend upon
305

the degree of difficulty of the separation. The limit is likely to be finer for the
separation of a sulphide from a non-sulphide than for a sulphide from a sulphide
or, especially, for an oxide from an oxide. Recovery by entrainment is close-
ly tied to the recovery of water and in principle this could be reduced by in-
creasing the flotation rate of the valuable mineral, by diluting the pulp which
flows to the concentrate by some procedure analogous to, say, the use of water
injection in cyclones or by washing the concentrate in such a manner that
valuable mineral is not lost. Most of the proposed modifications to convention-
al flotation for the enhancement of fine ~)article separation (Somasundaran,
1975, 1979) fall into the first of these categories, i.e. they seek to raise the
rate of fines flotation.
In general three procedures have been considered. The first is to increase
the probability of particle-bubble collision by increasing the effective size of
the particle by aggregation. Such methods include selective flocculation fol-
lowed by flotation of the floccules, carrier flotation in which coarse or inter-
mediate particles function as carriers for the fines and agglomerate or emul-
sion flotation in which the carriers are oil droplets. The most favoured appears
to be selective flocculation and flotation, but the progress reported to the
present time (see, for example, Koh and Warren, 1979) is not impressive,
especially if the feed is low grade. Little information exists about the general
applicability of the other methods. It should be recognized that each is sus-
ceptible to interference by entrainment and the magnitude of the increase
in flotation rate of fine mineral required to keep flotation times short enough
to minimize entrainment might well be b e y o n d their capabilities. Presumably
some improvement would follow if the gangue could be selectively flocculated
but this does not seem to have been investigated.
The second procedure relates to the proposal that selective chemisorbing
or chelating collectors, for the most part yet to be devised, might increase
the flotation rate of fines (Fuerstenau, 1975, 1980; Somasundaran, 1975).
The results of Fuerstenau et al. (1970) who floated a hematite ore, containing
45% Fe, ground to 70% < 15 pm to produce concentrates assaying from 64
to 68% Fe (i.e. an enrichment ratio of about 1.5) with 86% recovery using a
"chelating" collector (octyl hydroxamate) are frequently quoted as support-
ive evidence. It is difficult to assess proposals of this nature, for insufficient
size by size data have been presented to enable any conclusions to be drawn.
Furthermore, a distinction should be made between separations which are
successful in the presence of fines and those in which the fines themselves
have to be separated. In some circumstances xanthates fail to separate fine
sulphides as was found for the McArthur River example. Frequently it is the
rate of flotation of the coarse particles which is increased by the use of stronger
collectors, any benefits in the fine sizes deriving from the reduced entrainment
of fine gangue in the shorter flotation times made possible. This can only be
determined from studying size by size data. When considering the fines them-
selves, two factors have to be taken into account. On the one hand, there is
the evidence (Figs. 5--7) that the addition of a powerful and selective collec-
306

tor for chalcocite had a relatively small influence on the recovery rate of the
fines. On the other hand, the rejection of entrained gangue by reflotation is
demonstrably more successful from sulphide concentrates than from oxide
concentrates, where t o o much valuable mineral can be lost. While it is logical
to assume that this is the result of the greater hydrophobicity of collector
coated sulphide minerals, it might also be the result of greater persistence of
hydrophobicity in changing conditions.
Finally, it is pertinent to point o u t that in the Renison tin flotation circuit,
concentrates containing 25% Sn with recoveries ranging from 85 to 90% are
consistently maintained for a feed containing 2% Sn (i.e. an enrichment ratio
of 12,5) in which 96% of the cassiterite is less than 20 pm and some 30% of
the weight is under 5 pm (Goodman and Trahar, 1977). Nevertheless the col-
lector, pTA, appears to be only weakly adsorbed by cassiterite and the sepa-
ration efficiency for sizes below 5 ~m is low (Fig. 17).
The third procedure is to promote a particle-bubble collision mechanism
different from the direct encounter mechanism of conventional flotation,
such as by vacuum or by pressure release flotation in which gas is precipitated
on the particle to be floated. Whereas very fine particles can be floated by
such methods, for they are used to remove fine solids in the clarification of
water, there is no evidence that selective separations can yet be achieved.
By contrast with the above methods, little thought seems to have been
devoted to the reduction of gangue entrainment. In principle, entrainment
could be reduced by decreasing the water flow to the concentrate using a
suitable choice of frothers, froth modifiers or collector-frother combinations
or by reducing pulp densities but in practice such methods rarely seem to
provide an effective solution, especially in non-sulphide systems. Reference
has already been made to the possibility of diminishing entrainment by floccu-
lation of the gangue and there is no d o u b t that this is possible. For example,
in batch flotation tests on a naturally coagulated Mr. Isa ore sample, the re-
covery of ~ 2 pm material was as low as 10% for a water recovery of about
50% whereas for dispersed samples, the recovery of this size was typically 35
to 40%. However, for flotation to be selective, both the non-floating gangue,
comprising several minerals, and the floating mineral would need to be separate
ly flocculated and this does not seem to be feasible at present.
Of the many methods suggested for improving separations in this size
region, only column flotation combined with some form of refluxing and
froth washing (Somasundaran, 1975) appears to offer much prospect for re-
ducing entrainment. It should be emphasized that unless entrainment can be
reduced it is difficult to be optimistic about the ultimate applicability of
conventional flotation as the sole concentrating method to systems in which
fine particles predominate.

Particle size, surface energy and adsorption density

Several factors thought to be relevant to the flotation of fines originate


307

from the supposedly high surface energies of fine particles; they include the
non-selective adsorption of reagents, rapid surface reactions, especially oxi-
dation, and increased solubility (Klassen and Mokrousov, 1963, p. 403). Al-
though there is little experimental support for these postulates, some assess-
ment of their possible significance is necessary in view of the prominence
they have been accorded (Collins and Read, 1971; Fuerstenau et al., 1973;
Chander, 1978; Sastry, 1978; Somasundaran, 1979). Most emphasis will be
given to the consideration of surface energy and non-selective adsorption, in
particular to the size range relevant to flotation (say 1--300 pm). Of the other
factors, effects of particle size on solubility are unlikely to be significant
above about 0.1 um (Trahar and Warren, 1976) and the supposed detrimental
effects of rapid surface oxidation, which must relate primarily to sulphides,
seem to ignore the experimental finding that sulphide flotation has been the
least troubled by problems associated with fines; any useful assessment of the
possible role of enhanced surface hydration of fines is precluded by an absence
of relevant data.
Interpretation of the role of the surface energy in adsorption or flotation
is clouded by the fact that fines have a high specific surface, i.e. surface area
per unit volume or weight, which is inversely proportional to particle size and
which results in high total surface energies. It is not this quantity but the sur-
face energy per unit of surface area which is considered here. A distinction
is necessary because confusion between the two quantities probably accounts
for some current misconceptions about the behaviour of fine particles.
The postulated increase in surface energy per unit surface area with de-
crease in size has been attributed to an increased proportion of edges and
corners (Welch, 1952) and of cracks, dislocations and other lattice imperfec-
tions (Leja, 1952) arising from size reduction. Experimental verification of
these hypotheses does not appear to have been attempted for systems of flo-
tation significance. However, measurements of the heat of immersion per
unit area of crystalline solids in water indicate that this quantity in fact de-
creases with particle size; this result has been attributed to an increase in the
amorphous character of the surface (Wade et al., 1961).
The non-selective adsorption of reagents (Klassen and Mokrousov, 1963,
p. 403) and in particular of collector (Collins and Read, 1971) is thought to
be a consequence of the high surface energy of fine particles. It is implicit
in this argument that the adsorptive properties of minerals, especially gangue
minerals, are dependent upon particle size. It is also necessary to presume
that the adsorptive properties of the valuable mineral do not change in a
manner that increases its floatability otherwise the selectivity should not
have been reduced to the extent shown, for example, in Fig. 17. Such dis-
criminatory behaviour based solely on particle size is inherently unlikely.
The experimental evidence does not support any significant size effect on
adsorption density. Data on the uptake of copper by sphalerite (Anthony et
al., 1975) which are given in Tables IV and V show no evidence of size effect
from 9 to 124 pm in a competitive environment (i.e., with more than one
308

TABLE IV

Uptake of copper by sphalerite in a batch flotation concentrate

Average Calcd. Measured Calcd. Flotation


size surface uptake "adsorption" recovery %
(.m) area of Cu density in 2 min
(cm2/g) (.g/g) (.g/cm 2)

124 121 46 0.38 86


89 169 36 0.21 92
63 238 47 0.20 95
41 366 77 0.21 97
28 536 104 0.19 98
21 732 167 0.23 98
14 1079 215 0.20 98
9 1579 352 0.22 98

size fraction present, Table IV) or a slightincrease in adsorption density with


increased particle size over the range from 5 to 60 p m in a non-competitive
erivironment (singlesize fraction present, Table V). In terms of the argument
given by Welch (1952) the length of edge and the number of comers per unit
of surface would have increased by factors of 14 and 150 for the data in
Table IV and by 11.5 and 131 for those in Table V. Similarly, data given for

TABLE V

Uptake of copper by single sizes of sphalerite

Average Calcd. Measured Calcd.


size surface uptake "adsorption"
(,m) area of Cu density
(cm21g) (.g/g) (.g/cm2)
63 238 174 0.73
28 536 245 0.46
14 1079 502 0.47
5.3 2830 1010 0.36

the u p t a k e o f sulphide ion b y sized m a l a c h i t e (Kelsall and A s q u i t h , 1 9 5 8 )


d o e s n o t indicate a n y particle-size e f f e c t either o n a d s o r p t i o n d e n s i t y (Table
VI) or o n t h e rate o f u p t a k e (Table V I I ) over t h e size range f r o m 15 t o
1 5 0 p m o t h e r t h a n a slight decrease o f a d s o r p t i o n d e n s i t y w i t h decrease in
size w h i c h is t o be e x p e c t e d w h e n t h e t o t a l surface area available f o r a fixed
q u a n t i t y o f a d s o r b a t e increases. T h e surface areas in Tables IV to V I I were
c a l c u l a t e d o n t h e a s s u m p t i o n o f spherical particles so t h a t a l t h o u g h t h e areas
will differ f r o m the true values, t h o s e o f d i f f e r e n t sizes s h o u l d be in p r o p o r t i o n
309

TABLE VI

Uptake of sulphur by sized malachite (from Kelsall and Asquith, 1958)

Size range Naris Total S S incr. Calc. Total Coverage


(urn) concn. (g) on specific surface by S
(g/l) malachite surface (era 2) (~ g/cm 2 )
(g) (cm~/g)
147--124 1.4 0.0515 0.0065 111 194 33.5
124--104 1.4 0.0590 0.0070 132 231 30.3
88--74 1.4 0.0590 0.0110 185 324 34.0
53--44 1.4 0.0568 0.0138 312 546 25.3
44--15 1.4 0.0590 0.0209 577 1010 20.7

TABLE VII

Rate of uptake of sulphur by sized malachite (from Kelsall and Asquith, 1958)

Time Size Uptake of sulphur for Naris concn, of


(min) range
(~m) 0.35 0.7 1.4
(g/l) (g/l) (g/l)

1 124--104 6.06 8.66 20.3


53--44 n.d.* 11.7 26.9
44--15 7.33 13.3 22.6

2 124--104 10.4 14.7 29.4


53--44 n.d. 15.4 31.1
44--15 7.4 13.8 24.4

4 124--104 14.3 22.5 41.6


53--44 n.d. 17.8 33.0
44--15 7.7 13.9 24.8

6 124--104 15.6 26.4 47.2


53--44 n.d. 18.5 33.0
44--15 n.d. n.d. n.d.

*n.d. = not determined.

Perhaps the most compelling argument against a relation between particle


size a n d a d s o r p t i o n d e n s i t y is p r o v i d e d b y t h e w i d e s p r e a d a c c e p t a n c e o f t h e
measurement of the surface area of powders, independently of their particle-
size d i s t r i b u t i o n s , b y t e c h n i q u e s w h i c h d e p e n d o n a d s o r p t i o n f r o m t h e gas o r
l i q u i d p h a s e . I f a d s o r p t i o n d e n s i t y w e r e p a r t i c l e size d e p e n d e n t , so p r e s u m -
a b l y w o u l d b e t h e m e a s u r e d a r e a s . H o w e v e r , i t is t h e n n e c e s s a r y t o e n s u r e
310

that estimates of surface area are not based solely on adsorptive techniques
when testing the independence of adsorption density and particle size. It was
for this reason that calculated areas were used in Tables IV and V.

RECOVERY OF INTERMEDIATE SIZE PARTICLES

The principal mechanism for the recovery of intermediate particles is by


genuine flotation with a contribution from entrainment which becomes sub-
stantial towards the lower boundary of this size region. Intermediate par-
ticles are frequently liberated and float at a rate such that, for typical resi-
dence times, recovery is close to 100% over a considerable size range; these
particles pose no problems in flotation. The range of the intermediate region
varies with the mineral-collector system as indicated in Fig. 18 where it is

I - " ~ I
,°°L
90

8O

~ j
60--

50~

~
rr
4o~
/~ 30k
QUARTZ(DODECYLAMINEI
0 1
[J CASSITERITE(OLEICACID)
O0 20
f10 '&
7 CUPRITE (KEX)
TOURMALINE (SULPHO !
SUCClNAMATE)
~ CHALCOCITE{KEX) i
] J
2 5 10 51o ido ~;0 500 1000
A V E R A G E P A R T I C L E SIZE (#m)

Fig. 18. Typical recovery-size curves for a number of mineral-collector combinations


(data for quartz-dodecylamine adapted from Robinson, 1959--1960).

clear that intermediate behaviour extends to much coarser sizes for sulphides
than for oxides or silicates. There is a scarcity of information about the par-
ticle size response of non-sulphide systems other than oxides although the
data for fluorite of Lay and Bell (1962) indicate that it can be floated at
very coarse sizes in favourable circumstances.
The general properties of intermediate particles are evident in all of the
data presented and require no further comment. It should be noted however,
that the sizes defining the limits of intermediate behaviour and the propor-
tion of intermediate to other particles for each significant mineral constituent
are important parameters to establish when assessing the response of an ore
311

sample to conventional selective flotation, for they are indicative of the de-
gree of difficulty to be expected in effecting a separation for the whole ore.

RECOVERY OF COARSE PARTICLES

In general the recovery of coarse particles is by flotation with a negligible


contribution from entrainment. Referring back to Fig. 1, coarse particle re-
covery is usually lower than that of the intermediate sizes but this is not al-
ways true as will be evident in the detailed data presented below.

Basic flotation behaviour o f coarse particles

In Fig. 19 the recovery of chalcocite after one minute of flotation in four


separate tests with increasing single additions of collector is presented. For

--~I°°19o~- T. . . . . T .]j

80--

b,. 70F-
oo! /

'~ 4o
~z 50- '~ ~ ~

20~ 400
3 , :

~: ,O'Ioi :~i II~';," 2o ~2~o ~ 5OO 1OO0


AVERAGE PARTICLE SIZE (prn)
Fig. 19. Influence of the magnitude of single additions o f x a n t h a t e on the recovery o f
chalcocite.

small collector additions, the region of m a x i m u m floatability is narrow and is


located at small sizes; as the a m o u n t of collector increases, the recovery
plateau becomes progressively wider as the boundary between intermediate
and coarse particles shifts to coarser sizes. The response of the finest material
(average size ~ 2 pm) to increased collector, however, is slight which agrees
with the data presented earlier in Figs. 5 and 6. Size by size recoveries of
cassiterite as a function of collector (pTA) addition are given in Fig. 20 which
suggest that although the trend is similar to that shown by chalcocite, the
m a x i m u m floatability region is not as wide because the recovery drops fairly
sharply above 50 pm even when the collector addition is high (450 g/tonne).
312

Z
o_
8c

7c

6C i CASISITERITE i ~ i
0

w
0 40
O 113 g/tonne DTA
II
450 g/tonne oTA \
~" 3o
ul
\
~ 2o
ul

__ j j J L° i _ ~ i J
2 5 10 2 50 100 200 500 1000
AVERAGE PAR'rlCLE SIZE (pro)
Fig. 20. Influence of p-tolyl arsonic acid (pTA) on the recovery of cassiterite in batch
flotation at pH 5.5.

O t h e r reagents w h i c h p r o m o t e floatability have associated size e f f e c t s simi


lar to collectors. T h u s differences b e t w e e n frothers are clearly observable
o n l y at t h e coarser sizes, w h i l e the i n f l u e n c e o f activators is s h o w n to be
similar to that o f c o l l e c t o r s by t h e data in Fig. 21 w h i c h illustrates t h e e f f e c t
o n t h e floatability o f m a l a c h i t e o f increasing a d d i t i o n s o f h y d r o s u l p h i d e ion

70~ J

~Ok [',~HS ADDITIONS(Ib,/hr')


N /
~[ • ' 14+07+035

°
o ~,0 . . . . . . . . . . . . . ,\

I
oL I t 1 1 i l L ±
l 2 5 10 20 50 100 200 500 tO00
AVERAGE PARTICLE SIZE (urn.)
Fig. 21. Influence of increasing additions of sodium hydrosulphide on the recovery of
malachite in batch flotation ( f r o m Kelsa]l and Asquith, 1 9 5 8 ) .
313

(Kelsall and Asquith, 1958). Similar data have been given elsewhere for the
effect of copper sulphate on sphalerite (Kelsall et al., 1974).
The influence of depressants or suppressants is illustrated in Fig. 22 where
it is clear that the principal effect of increasing pH on the floatability of

loo[ [ I-. . . . . . . ~ - -

go

i 80
< I
F
W 70
N
60

~_) 40 / ~ PYRITE
¢r
W 0 pH5

>.. 30k OP Hll


n"

~• TIME 4 rain
0 2oF
(.,) KEX lOOmg
I
~L" PPG 400 5rag/min

1°I -I
0 1 I L 1 1_ _ • ___L__ ___L~
1 2 5 10 20 50 100 200 500 1000
AVERAGE PARTICLE SIZE (pro)

Fig. 22, T h e effect o f pH o n t h e b a t c h f l o t a t i o n o f p y r i t e w i t h excess collector.

pyrite in the presence of excess collector is the depression of the coarse par-
ticles in agreement with Blake's (1972) plant observations. Similarly flotation
is suppressed in the coarser sizes if the environment becomes too reducing or
too oxidizing (Heyes and Trahar, 1979). Even for the mineral chalcopyrite,
which requires no collector for floatability, the coarse particles are the most
susceptible to environmental changes as shown by the detailed data of Heyes
and Trahar (1977).
The feature c o m m o n to these data is the variability in the response of the
coarse particles and their sensitivity to the chemical environment when com-
pared with intermediate particles for which the response to change is moderate
and with fines where it is slight. Thus the first signs of a deficiency of collec-
tor or activator, of an excess of depressant, of an unfavourable pH or redox
condition, indeed of any variable which could reduce the hydrophobicity of
the mineral, are shown by a decrease in the rate of flotation of the coarse par-
ticles.

Interpretation of coarse particle behaviour

Excluding effects due to froth conditions or those relating to slime coatings,


two mechanisms have been proposed to account for the behaviour of coarse
particles, namely:
314

(1) that the degree of hydrophobicity required to promote a high level of


floatability increases with increase in particle size;
(2) that the presumed rapid and disproportionate consumption of collector
by fine particles leads to less complete surface coverage of the coarse particles
which accordingly are less floatable.
The first mechanism was discussed by Gaudin in 1927 who noted that:
" I n t h e f l o t a t i o n o f s p h a i e r i t e w i t h oleic acid, as in o t h e r cases, a d i s t i n c t relation-
ship has b e e n o b s e r v e d b e t w e e n t h e c o m p l e t e n e s s o f t h e a d s o r b e d c o l l e c t o r film
a n d t h e size: as t h e particle size is increased, a m o r e c o m p l e t e coating is necessary.
This result a p p e a r s logical e n o u g h w h e n t h e a d s o r b e d film is c o n s i d e r e d as a part-
ly c o m p l e t e m o n o m o l e c u l a r film. As a n e x t r a p o l a t i o n o f this o b s e r v a t i o n it w o u l d
s e e m t h a t a c o m p l e t e m o n o m o l e c u l a r film is r e q u i r e d t o f l o a t t h e coarsest m i n e r a l
grains t h a t c a n be f l o a t e d , a n d t h a t i n f i n i t e l y small particles w o u l d f l o a t w i t h an
i n f i n i t e l y i n c o m p l e t e m o n o m o l e c u l a r film.".
There is much experimental support for such a contention including data
for the systems sphalerite-oleic acid (Gaudin, 1927), galena-oleic acid and
galena-xanthate (Gaudin et al., 1928), sphalerite-copper sulphate (Ravitz and
Wall, 1934; A n t h o n y et al., 1975), galena-xanthate and pyrite-xanthate
(Mitrofanov et al., 1962; Glembotsky, 1968) and quartz-dodecylamine
(Robinson, 1959--1960). These results showed convincingly that the reagent
additions necessary to float the coarsest particles were much higher than for
other sizes; a typical example is the finding that five times as much copper
sulphate was required to float 125 pm sphalerite at the same rate as 15 pm
sphalerite (Anthony et al., 1975). In a slightly different approach, Kakovsky
et al. {1961) showed that the principal benefit of thiol collectors with long
hydrocarbon chains was in the recovery of coarse particles; such collectors
produced no change in the flotation rate of fine or intermediate particles. In
his careful analysis of particle size effects, Robinson (1959--1960) concluded
that there is a critical degree of hydrophobicity for the attainment of maxi-
m u m floatability, which can be distinguished from some threshold level at
which flotation just becomes perceptible. The evidence presented here pro-
vides strong confirmation for the contention that the difference between
threshold and critical levels increases rapidly with particle size. A form of
relationship between floatability and hydrophobicity based on such a sup-
position is presented in Fig. 23 which, though qualitative, is derived from de-
tailed data from the sphalerite-copper sulphate system (Anthony et al., 1975).
There is no good reason for attempting to distinguish between minerals
which become floatable by the addition of collector and those which become
floatable by partial oxidation of their surfaces, such as chalcopyrite, pyrrho-
tire and galena, for which the presence of sulphur is presumed to be respon-
sible (Richardson and Maust, 1976; Gardner and Woods, 1979). The domi-
nating requirement is the total degree of hydrophobicity attained from all
possible sources: mineral, collector, mineral-collector reaction product, min-
eral oxidation product, frother, etc.
The second mechanism, which agtempts to account for the reduced floata-
bility of coarse particles in terms of competitive adsorption between different
315

rn
<
I.-
o
._1
iJ_

HYDROPHOBICITY - - - - ~ - ~ -

Fig. 23. Qualitative representation o f suggested f o r m o f the influence of particle size


on the relationship b e t w e e n floatability and h y d r o p h o b i c i t y .

particle sizes, has wide support (Klassen and Mokrousov, 1963, p. 409; Rey,
1968; Arbiter, 1978) probably arising from the plausible presumption that
equal collector coverages should correspond to equal floatabilities indepen-
dently of particle size, coupled with the well known observation from flota-
tion practice that coarse particle recovery is facilitated b y stagewise collector
addition. Any theory on the operation of this mechanism needs careful defini-
tion for it is sometimes not obvious exactly what is intended. For example,
in discussing the sulphidation of malachite with sodium hydrosulphide, Set
et al. (1970) stated:
" I t has been shown that the reaction which forms the sulphide surface on o x i d e
copper minerals proceeds at a m u c h faster rate for particles in the finer size frac-
tions. Hence in a pulp to which sulphidizer is added, the fine copper particles tend
to adsorb a substantial p r o p o r t i o n of the sulphide ions present, with the result that
there are insufficient of these to adequately sulphidize the coarser fractions.".
It is not clear from this statement whether these authors believe that the up-
take of sulphide per unit of surface by fine malachite is greater than by coarse
malachite or merely that the fines consumed more reagent than the coarse
by virtue of their greater specific surface but the reference to the faster rate
of adsorption, which recurs elsewhere (Rey, 1968; Fisher and Deuchar, 1977)
suggests that they do. As noted earlier, it is difficult to find experimental
evidence in favour of an influence of particle size on adsorption density or
on the rate of adsorption, even though a change in rate when the particle size
approaches the thickness of the diffusion layer (for those systems where ad-
sorption is under diffusion control) might be expected on theoretical grounds
316

(Berner, 1971, p. 32). Ser et al. (1970) used the data of Kelsall and Asquith
(1958) as a basis for their interpretation but when these data are assessed in
terms of reasonable estimates of surface area no particular size effect can be
distinguished, as noted previously for Tables VI and VII.
The increased rate at which adsorbate is removed from solution when the
particle size of the adsorbent is reduced should not be taken to be evidence
of an increased specific rate of adsorption; it merely reflects the increased
surface area available. A flotation analogy would be to conclude that because
an increased feed tonnage led to an increased concentrate flow rate, the flo-
tation rate constant had increased.
The influence of depressant on coarse particle response is not consistent
with the concept that reagent concentrations are disproportionately depleted
by fine particles. If the depression of pyrite observed at high pH values in
Fig. 22 is caused by competition between collector and h y d r o x y l ions, which
is the accepted theory (Gaudin, 1957, pp. 282--288), it is logical to conclude
that the h y d r o x y l ions must have reacted to a disproportionate extent with
the coarse particles, for they are the first to be depressed. It must then be con-
cluded that, since fines presumably adsorb collector and activator preferentially
(Figs. 19 and 21), different sizes of the mineral can distinguish between the
different reagent types in the system, which is absurd.
It is concluded therefore that the evidence is in favour of the first postu-
late, i.e., that the degree of hydrophobicity required for maximum floatability
increases with particle size. In a range of sizes no single collector addition will
be appropriate for all sizes and the most efficient compromise will depend on
the size distribution of the important minerals. Insufficient evidence exists to
support the view that fine particles consume reagents disproportionately to
surface area and this point of view is unlikely to be true except in the sense
that when the proportion of fines is large, the adsorption density of reagent
on all particles will be reduced. In this case the high coverage requirements
of the coarsest particles might not be met and their recovery would be then
lower than that of particles whose coverage was adequate.

DISCUSSION

The effects of particle size in flotation are so diverse that it is not feasible
to discuss or even to recognize every possible consequence. Accordingly the
discussion which follows is limited to a consideration of some salient aspects.

Theoretical background

Although flotation is known to include a number of subprocesses as listed


by Trahar and Warren (1976), the three most important relate to particle-
bubble collision, adherence and detachment. The probability of flotation can
be related to the product of the probabilities of each of these (Schuhmann,
1942; Sutherland, 1948; Tomlinson and Fleming, 1963). Thus:
317

Pf ~ Pc" Pa" ( 1 - P d ) (1)


where pf, Pc, Pa and Pd refer to the probabilities of flotation, collision, ad-
herence and detachment. For present purposes it is simpler to replace ( 1 - p d )
by Ps where Ps is the probability that the particle-bubble combination, once
formed, will be stable enough to enter the concentrate, i.e.:
pf cc Pc " Pa " Ps (2)
The probability of particle-bubble collision has attracted much attention
(Sutherland, 1948; Flint and Howarth, 1971; Reay and Ratcliff, 1973) which
has recently been summarized by Jameson et al. (1977). Although the exact
form of the relation between collision frequency and other variables has not
been established beyond dispute, there is no real doubt that Pc is related to
the physical variables, i.e. particle and bubble diameters, particle and liquid
densities, liquid viscosity, relative particle-bubble velocity, etc. and in particu-
lar, Pc is directly related to the particle diameter d. No term relating to the
contact angle appears in any of the expressions f o r p c (Jameson et al., 1977)
which therefore should not be influenced significantly by the chemical nature
of the environment except indirectly through the number of bubbles generated
per unit volume of air and the relative electrical charges of particles and
bubbles (Collins and Jameson, 1976).
The variables which influence the probability of adherence, Pa, are less well
understood, for the problems are complex and the variables inter-related
(Derjaguin and Dukhin, 1960--1961). Furthermore, although it is simple in
concept to distinguish between collision and adherence, there is a real difficulty
in practice in knowing where the one ends and the other begins. In qualita-
tive terms, Pa is inversely related to the induction time, which is the time
taken for thinning and rupture of the water film between particle and bubble.
If the induction time is less than the particle-bubble contact time, adherence
occurs. The experimental results of Sven-Nilsson (1935), Eigeles (1939),
Glembotsky (1953, 1968) and Eigeles and Volova (1960) provide clear evi-
dence that induction time is inversely related to collector addition and tem-
perature and directly related to particle size such that relatively small in-
creases in particle diameter lead to large increases in induction time. Since
Pa is inversely related to induction time it must be directly related to collector
addition hence to contact angle and inversely related to particle size.
The third term Ps is dependent on the force of adhesion between particle
and bubble which in turn is directly related to contact angle and inversely
related to particle size (Wark, 1933; Morris, 1952; Gaudin; 1957, pp. 157--
162). The quantitative form of the relationship is probably complex (Morris,
1952).
In summary, of the terms in relation 2, Pc is directly related to a function
of d while Pa and Ps are directly related to a function of the hydrophobicity
and inversely related to a function of d i.e.:
Pc o: f[d] (where f[d] increases with d) (3)
318

Pa "Ps ~ f ' [ ¢ ] / f " [d] where f ' [~] increases with ~ and f"[d] increases {4)
with d)
The term ¢, though not necessarily the contact angle of standard flotation
terminology, is a measure of the degree of hydrophobicity of the mineral par-
ticle, being the net result of all the effects produced by the chemical environ-
ment, including surface coverage by the mineral-collector reaction product,
by neutral oil, by frother, by mineral oxidation products, by hydrophilic
groups and by water. No attempt will be made here to evaluate the separate
effects of Pa and Ps, for these have been discussed in detail elsewhere (Jowett,
1980). Presumably both may limit the flotation rate of coarse particles in ap-
propriate circumstances, either singly or conjointly.
According to these relations, the overall probability of flotation has a com-
plex relationship with particle size. Recovery curves will be a compromise be-
tween Pc on the one hand and Pa "Ps on the other such that there is a size region
where the recovery reaches a maximum. In addition there is a clear relation-
ship between the chemical environment and the location and width of the
maximum floatability region. It is presumed that for any particle there is a
critical value of the term which includes ~ and d in relation 4 which must be
exceeded before the particle will float. When ¢ is small (i.e., conditions un-
favourable}, the critical value can only be exceeded for small d; the flotation
maximum will then span a small size range as shown, for example, by the data
in Fig. 5. As ~ is increased, for example by adding more collector, the region
of maximum floatability becomes wider by extending to coarser particle sizes
as shown in Figs. 6, 19 and 21. When ¢ is very high the flotation rate is highest
for the coarsest sizes present. This follows from the fact that if ~ is very large
{just as when d is very small)Pa " Ps will approach unity (Tomlinson and
Fleming, 1963) and then the rate of flotation will be determined primarily
by Pc which increases with particle size. Such an interpretation is also in
accord with the observations made earlier that the cumulative recovery of
fines is determined more by the available flotation time than by the chemical
environment because the critical level of hydrophobicity is easily exceeded
for such particles.
When the effects of entrainment are added to those discussed above, the
resulting model is capable of qualitatively interpreting all of the data pre-
sented. The changes in shape of recovery-size curves and their evolution with
time as a result of increasing hydrophobicity are summarized in Fig. 24 which
depicts in detail the influence of increasing additions of copper sulphate on
the flotation behaviour of sphalerite. The transition from entrainment to flo-
tation is particularly clear. The slight hump in the curves in Fig. 24A, indica-
tive of incipient flotation, develops in Fig. 24B into a clearly defined flota-
tion peak which increases in height in Fig. 24C and in width in Figs. 24D and
24E. Finally in Fig. 24F the boundary separating intermediate and coarse
particles has practically disappeared; overall flotation is then very rapid.
319

,ooF . . . . . . ~ - - 10(3 r
B

50

i 1/2 i
OL _____
i lO lOO lOOO °l tO 100 100o
'0or ' ~ c~ :oo[-- ~ -~ DI

50~ 5O

1/

1~o loo 1 10OO

i
oL ~
10OO
I
AVERAGE PARTICLE SEE (~m}
Fig. 24. Influence of increasing additions of copper sulphate on batch flotation of sphal-
erite. The rates of copper sulphate additions in mg CuSO 4. 5H~O per g of sphalerite were:
(A) 0; (B) 0.17; (C) 0.33; (D) 0.67; (E) 1.67; (F) 3.33. (Flotation times in minutes were
o ]/2;D 1 ; ~ 2; v 4 ; 0 8.)

Surface coverage and floatability

There is a need for quantitative information on the collector requirements


of individual size fractions of different minerals if the efficiency of selective
flotation of minerals from increasingly refractory ores is to be improved. Ex-
perimental attempts to derive such relationships have produced results which
have been so variable that little recent research in this topic has been under-
taken. An extreme view is the assertion that there is an almost total lack of
correlation between collector adsorption and flotation response (Pope and
Sutton, 1971--1972); it is probable that this conclusion arose from a failure
to recognize the prominent role of particle size in any such relation. In those
320

studies where the average size of particles was very small, coverages of less
than 5% were sufficient to produce recognizable flotation (Edwards and
Ewers, 1951). In the presence of a range of sizes, and in particular of coarse
particles, higher coverages were necessary, as shown by the data of Gaudin
and Preller (1946). Although their results have been widely used as p r o o f that
collector coverages of 15--20% are sufficient to produce flotation, what they
really proved was that complete flotation requires coverages approaching
a nominal monolayer. In the interpretation of such data a complicating factor
is introduced by sulphides which exhibit floatability in the absence of added
collector, for it is not then possible to establish an unambiguous relation be-
tween added collector and flotation response. Detailed attempts by the author
to relate surface coverage by copper to the floatability of sized sphalerite and
surface coverage by xanthate to the floatability of sized chalcopyrite have
failed to produce quantitative relationships for this reason.

Particle size and selectivity

Although the influence of size on selectivity has long been recognized


(Gaudin et al., 1931; Carr, 1952; Fleming, 1959) it has seldom been utilized
in flotation testing or until recently in the assessment of plant performance.
The limit to the selectivity attainable in practice is set by the floatability
of the coarse particles of the more floatable mineral relative to that of the
intermediate particles of the less floatable mineral as a detailed examination
of the galena-sphalerite system has shown (Trahar, 1976). If these overlap,
selectivity is difficult to maintain for high recoveries. Typical data are given
in Fig. 25 in the form of selectivity curves. Assuming first order kinetics,

0.£ /
0~ O 1 150~m Pb vs 1501Jrn Zn
0.7 [] 2 150#m Pb vs 30pro Zn
3 150pm Pb vs 30~m Zn WITH SEPARATE r-I t

f/
0.6 CONDITIONING

~: 05

0.t

Z
_o
I-- 0.2

0,1
0.1
' I
0.2
I
0.3 0.4
/
IE~ J
0.5
I
0.6
i
I I I
0.7 0.8 1.0
FRACTION OF ZINC UNFLOATED

Fig. 25. Selectivity curves (ratio of average rate constants) for batch flotation of galena
from sphalerite.
321

the slope of the lines is the ratio of average rate constants for the flotation
of galena and sphalerite. While there is excellent selectivity at 150 pm, that
between 150 pm galena and 30 pm sphalerite is greatly reduced. This curve
represents the limiting condition in this example. In such circumstances the
selection of a suitable environment will be a compromise between the dif-
fering requirements of all particles present and cannot be o p t i m u m for all
of the individual components. Such a conclusion reduces the value of test-
work in which reagent regimes for selectivity have been measured on single
size ranges of minerals using microscale flotation tests and provides a reason
why such regimes are seldom found to be effective when tested on a full
range of sizes.
These limitations on selectivity can be alleviated in practice by the use of
separate circuits for fine, intermediate and coarse particles. However, the
technique has seldom been utilized, suggesting that the improvement in ef-
ficiency has not been considered commensurate with the increased complex-
ity; but in the long term, separate treatment of selected size ranges will be-
come necessary for the more difficult ore types. A modified version can be
useful in those systems in which the mineral-collector interaction is irreversible
within normal flotation times, whereby different size fractions may be con-
ditioned separately in environments which are more nearly optimum and
then recombined prior to flotation {Glembotsky, 1968). Such a procedure
has been shown to have several potential advantages including increased rates
of flotation of coarse mineral, increased selectivity or reduced reagent con-
sumption in the activation of malachite by sodium hydrosulphide (Kelsall and
Asquith, 1958), in the activation of sphalerite by copper sulphate (Anthony
et al., 1975} and in the separation of galena from sphalerite (Trahar, 1976).
A measure of the increased selectivity in the last system is included with the
data in Fig. 25. In practice, the flotation feed is divided into two or three size
ranges. Reagents which p r o m o t e flotation (collectors, activators) are added
principally to the coarse fraction at high pulp densities while those which aid
selectivity by depression of gangue are best added to the intermediate frac-
tion. After suitable conditioning periods, the fractions are recombined for
flotation. The optimum sizes for the initial separation and the most beneficial
apportioning of reagents need to be established experimentally.

Composite particles

The behaviour of composites in relation to particle size effects merits


c o m m e n t even though composite behaviour has not been included in this
practical study.
It is c o m m o n l y supposed that composite particles float less readily than
free particles and that a relation should exist between surface composition
and floatability. Although this is a plausible view, for which some evidence
exists (Steiner, 1974), composite behaviour is not well documented or
understood; there is a scarcity of data which enables an unambiguous distinc-
322

tion to be made between the effects of particle composition and the effects
of particle size, since composites tend to be in the slower floating coarse
sizes. If the interpretation of particle size behaviour suggested here is true,
i.e. that relatively low levels of hydrophobicity are necessary to float fine
and intermediate particles, then, provided the percentage of valuable mineral
exceeds some threshold level, the flotation properties of composites in these
size regions will not differ significantly from those of free particles. How-
ever, the hydrophobicity requirements of coarse composites would be diffi-
cult to achieve and in this region the floatability would be expected to de-
crease as valuable mineral content decreased, i.e. composites probably ex-
hibit coarse particle behaviour at smaller absolute sizes, depending to some
extent on the contribution to the hydrophobic or hydrophilic character of
the particle from the other minerals present.

Comparison of recovery-size curves from laboratory and plant data

Because many of the data which have been presented were from labora-
tory batch tests under idealized conditions, an obvious question which arises
concerns the extent to which these observations are relevant to plant con-
ditions and can be supported by plant data. Available information shows that
the general pattern of behaviour appears to be typical of that observed in full
scale operations. For example, Fig. 26 gives data relating to the rougher cir-
cuit and to the whole flotation circuit in the tin flotation section at Renison

100: I I ~--b-~ I I I
O ,,~" -9.-.4'~,,"t\

8o- \ ,

0-
(~ 30: ~ PLANT RUN 1 13112172 "
>"
n- A PLANT RUN 2 18/1/73 /

2o . . . . , 3 FLOTATION RUN
.PLANT ~ n I \\
O BATCH TEST
10~ • ROW 1 ROUGHER 17/1173
/
oL ± ~ ~ I I I
1 2 5 10 20 0 200 500 1000
A V E R A G E PARTICLE SIZE (pm)
Fig. 26. Comparison of plant data and batch flotation results on Renison flotation feed
samples. (The plant results shown by open points refer to the whole circuit; those with
the closed points refer to a typical rougher result. )
323

and data from a typical batch test on Renison flotation feed. The curves are
similar in shape with some differences at the coarse end as might be expected.
When it is considered that the plant results include substantial contributions
from recycled products, however, the differences are not great. Plant and
laboratory results on a Mt. Isa lead-zinc ore are compared in Fig. 27 where
the agreement is excellent over the whole size range of the feed. Other com-
parisons have been given elsewhere (Trahar and Warren, 1976).

sot

(3_ ) PRIMARYPb ROUGHER


>- 30~ ~ SECONDARyPb ROUGHER
~-~
[]~ 20~1 ! C TOV~ER
!A
! LILO ~ i iPb
T ~CIRCUIT
Hz~

oi "
1 2 5 ~O 20 5O 100 200 500 1000
AVERAGE PARTICLE SIZE (prn)

Fig. 27. Comparison of plant data and batch flotation results on Mt. Isa lead-zinc flotation
feed.

PRINCIPAL OBSERVATIONS AND CONCLUSIONS

(1) Particle size is a flotation variable of great significance. Although recog-


nition of its importance has been widely professed, the intelligent implemen-
tation of its consequences, which may lead to higher efficiencies in operation
and to greater insight in research, has seldom been responsibly practiced.
(2) Entrainment is a major contributor to the recovery of fine minerals
in flotation which, in combination with a slow rate of genuine flotation, can
account for much of the observed behaviour of fine particles. Unless some
means for diminishing the entrainment contribution can be'devised, this
mechanism will always limit the lower size to which efficient separations
can be extended using conventional flotation.
(3) There is a qualitative b u t a clear and fundamental relationship between
the degree of hydrophobicity necessary to effect flotation and the particle
size. Accordingly, the shape of recovery-size curves and its evolution with
time is a valuable diagnostic aid to assessing flotation data. The sensitivity
324

of the coarse particle response in particular is invaluable as an indicator of the


suitability of the chemical environment for any envisaged separation.
(4) Floatable minerals generally display high flotation rates only within a
limited size range, the location of the coarser boundary being determined by
the system itself, i.e. the mineral and its chemical environment.
(5) Even when the minerals in an ore are wholly liberated, the maximum
size for which recovery is significant is determined by the requirements of
selectivity. If the selectivity is high, the region for high recoveries may ex-
tend to very coarse sizes (300 pm or more), but if the selectivity is low, par-
ticles above 40--50 pm may become difficult to recover. Any attempt to in-
crease coarse particle recovery in conventional circuits (i.e. without some
form of size separation) under these circumstances inevitably leads to a decrease
in selectivity.
(6) Feed preparation for flotation, i.e. grinding and classification, in ad-
dition to its role in the liberation of valuable mineral, should be designed
where possible to shape the size distribution of valuable mineral such that
as large a proportion as possible falls within the range associated with high
flotation rates. It is in this sense in particular that attempts should be made
to integrate the control of grinding and flotation circuits.
(7) There is a need for quantitative information on the relationship be-
tween hydrophobicity and particle size and particularly on the distribution
of added reagents between the different size fractions of different minerals
when present in the same environment. Such data will be indispensible for the
successful beneficiation of increasingly refractory ores. Coupled with appro-
priate on-line sizing and assaying equipment which would permit the separate
assaying of at least three size fractions, such data can provide a basis for the
application of realistic control strategies to other than simple separations.
The introduction of circuits with separate flotation of separate size regions,
more advanced in concept than simple sand-slime floats, seems inevitable in the
long term.

ACKNOWLEDGEMENTS

The author is indebted to many of his colleagues at the Division of


Mineral Engineering, especially Alan Jowett, for useful discussion of much
of the content of this paper.

REFERENCES

Anthony, R.M., Kelsall, D.F. and Trahar, W.J., 1975. The effect of particle size on the
activation and flotation of sphalerite. Proc. Australas. Inst. Min. Metall., 254: 47--58.
Arbiter, N., 1978. Problems in sulphide ore processing. N.S.F. Workshop on Fine Particle
Technology, Sterling Forest, N.Y.
Blake, R.A., 1972. Effect of partial removal of xanthate collector on size of mineral
floated. Trans. Am. Inst. Min. Metall. Eng., 252: 39--42.
325

Berner, R.A., 1971. Principles of Chemical Sedimentology. McGraw Hill, New York, N.Y.,
240 pp.
Carr, J.S., 1952. Contribution to discussion. Recent Developments in Mineral Dressing,
Proc. Syrup., London 1952. Inst. Min. Metall., 1953, pp. 90--91.
Chander, S., 1978. Recent developments in floatability of fine particles -- a review. Trans.
Indian Inst. Met., 31(1): 12--19.
Collins, G.L. and Jameson, G.J., 1976. Experiments on the flotation of fine particles --
the influence of particle size and charge. Chem. Eng. Sci., 31: 985--991.
Collins, D.N. and Read, A.D., 1971. The treatment of slimes. Miner. Sci. Eng., 3: 19--31.
Derjaguin, B.V. and Dukhin, S.S., 1960--1961. Theory of flotation of small and medium
sized particles. Trans. Inst. Min. Metall., 70: 221--245.
Doren, A., Vargas, D. and Goldfarb, J., 1975. Non-ionic surfactants as flotation collec-
tors. Trans. Inst. Min. Metall., 84: C34--C37.
Edwards, G.R. and Ewers, W.E., 1951. The adsorption of sodium cetyl sulphate on cas-
siterite. Aust. J. Sci. Res., 4(4): 627--643.
Eigeles, M.A., 1939. Kinetics of adhesion of mineral particles to air bubbles in flotation
suspensions. Dokl. Akad. Nauk. S.S.S.R., 24(4): 340--344.
Eigeles, M.A., and Volova, M.L., 1960. Kinetic investigation of effect of contact time,
temperature and surface condition on the adhesion of bubbles to mineral surfaces.
5th Int. Miner. Process. Congr., London. Inst. Min. Metall., 1960, pp. 271--284.
Fisher, J.F.C. and Deuchar, A.D., 1977. Some problems in the metallurgical processing of
oxide copper ores in Zambia. l l t h Int. Miner. Process. Congr., Sao Paolo, Brazil.
Round Table on Copper, 2, Oxidized, pp. 73--92.
Fleming, M.G., 1959. Selectivity factors in flotation. Chem. Ind., Oct. 3, pp. 1230--1238.
Flint, L.R. and Howarth, W.J., 1971. The collision efficiency of small particles with
spherical air bubbles. Chem. Eng. Sci., 26: 1155--1168.
Fuerstenau, D.W., Chander, S. and Abouzeid, A.M., 1973. The concentration of fine par-
ticles. Congreso Latinamericano de Mineria y Metalurgia Extractiva, Santiago, Chile.
Fuerstenau, D.W., 1975. Flotation. N.S.F. Workshop on Research Needs in Mineral Pro-
cessing, New York, Report, P. Somasundaran and D.W. Fuerstenau (Editors), 1976,
pp. 63--68.
Fuerstenau, D.W., 1980. Fine particle flotation. In: P. Somasundaran (Editor), Fine
Particles Processing, Vol. 1, AIME, pp. 669--706.
Fuerstenau, M.C., Harper, R.W. and Miller, J.D., 1970. Hydroxamate vs. fatty acid flota-
tion of iron oxide. Trans. Am. Inst. Min. Metall. Eng., 247: 69--73.
Gardner, J.R. and Woods, R., 1979. An electrochemical investigation of the natural flota-
bility of chalcopyrite. Int. J. Miner. Process., 6" 1--16.
Gaudin, A.M., 1927. Flotation mechanism, a discussion of the function of flotation rea-
gents. Am. Inst. Min. Metall. Eng., Tech. Publ. No. 4.
Gaudin, A.M., 1957. Flotation. 2nd Edn. McGraw-Hill, New York, N,Y., 573 pp.
Gaudin, A.M. and Preller, G.W., 1946. Surface areas of flotation concentrates and thickness
of collector coatings. Trans. Am. Inst. Min. Metall. Eng., 169: 248--256.
Gaudin, A.M., Glover, H., Hansen, M.S. and Orr, C.W., 1928. Flotation fundamentals.
Univ. Utah Expt. Station, Tech. Paper No. 1, Part A, Galena, pp. 6--30.
Gaudin, A.M., Groh, J.O. and Henderson, H.B., 1931. Effect of particle size on flotation.
Am. Inst. Min. Metall. Eng., Tech. Publ. No. 414.
Gaudin, A.M., Schuhmann, R., Jr. and Schlecten, A.W., 1942. Flotation kinetics, II. The
effect of size on the behaviour of galena particles. J. Phys. Chem., 46: 902--910.
Glembotsky, V.A., 1953. Rate of adhesion of air bubbles to mineral particles during
flotation and methods for its measurement. Izv. Akad. Nauk. S.S.S.R., Otdel. Tekh.
Nauk., pp. 1524--1531. Chem. Abstr., 49: 6667d.
Glembotsky, V.A., 1968. Investigation of separate conditioning of sands and slimes rea-
gents prior to joint flotation as a method of intensifying flotation. 8th Int. Miner. Pro-
cess. Congr., Leningrad. Paper S-16.
326

Goodman, R.H. and Trahar, W.J., 1977. Flotation of cassiterite at the Renison tin mine,
Renison Bell, Tasmania. Intern. Tin Symp., La Paz, Bolivia.
Greene, E.W. and Duke, J.B., 1962. Selective froth flotation of ultrafine minerals or slimes.
Trans. Am. Inst. Min. Metall. Eng., 223: 389--395.
Hemmings, C.E., 1978. Repression of galena by slime gangue in an agitated pulp. Int. J.
Miner. Process., 5: 85--92.
Heyes, G.W. and Trahar, W.J., 1977. The natural flotability of chalcopyrite. Int. J. Miner.
Process., 4: 317--344.
Heyes, G.W. and Trahar, W.J., 1979. Oxidation-reduction effects in the flotation of
chalcocite and cuprite. Int. J. Miner. Process., 6: 229--252.
Jameson, G.J., Nam, S. and Moo Young, M., 1977. Physical factors affecting recovery
rates in flotation. Miner. Sci. Eng., 9(3): 103--118.
Johnson, N.W., McKee, D.J. and Lynch, A.J., 1974. Flotation rates of non-sulphide
minerals in chalcopyrite flotation processes. Trans. Am. Inst. Min. Metall. Eng., 256:
204--209.
Jowett, A., 1966. Gangue mineral contamination of froth. Brit. Chem. Eng., 2: 330--333.
Jowett, A., 1980. Formation and disruption of particle-bubble aggregates in flotation. In:
P. Somasundaran (Editor), Fine Particles Processing. Vol. 1, AIME, pp. 720--754.
Kakovsky, I.A., Grebnev, A.N. and Silina, E.I., 1961. The relationship between the floata-
bility of mineral particles of various sizes, their structures and the consumption of col-
lectors. Tsvetnye Metal, New York, 2(8): 7--16.
Kelsall, D.F. and Asquith, J., 1958. A preliminary study of the sulphidization of malachite.
T.I. No. 248, R.M.S.L., Kitwe, Zambia.
Kelsall, D.F., Stewart, P.S.B. and Trahar, W.J., 1974. Diagnostic metallurgy, a method of
plant optimisation. Symp. on Opt. Control Miner. Process. Plants, Brisbane, July, pp.
53--65.
Klassen, V.I. and Mokrousov, V.A., 1963. An Introduction to the Theory of Flotation.
Butterworths, London, 493 pp.
Koh, P.T.L. and Warren, L.J., 1979. Flotation of an ultrafine scheelite ore and the effect
of shear flocculation. 13th Int. Miner. Process. Congr., Warsaw. Preprints of Papers, J.
Laskowski (Editor), Vol. 1, pp. 229--253.
Lay, W.C. and Bell, G.M., 1962. Fluorspar beneficiation at Alcoa Raw Materials Division
Plant, Rosiclare, Illinois. In: D.W. Fuerstenau (Editor), Froth Flotation, 50th Anni-
versary Volume. AIME, New York, N.Y., pp. 482--493.
Leja, J., 1952. Contribution to discussion. Recent Developments in Mineral Dressing,
Proc. Syrup., London, 1952. Inst. Min. Metall., 1953, pp. 426--428.
Lovell, V.M., 1976. Froth characteristics in phosphate flotation. In: M.C. Fuerstenau
(Editor), A.M. Gaudin Memorial Flotation Symposium. AIME, New York, N.Y., Vol. 1,
pp. 597--621.
Mitrofanov, S.I., Frumkina, R.A. and Ratnikova, O.A., 1962. Dependence on the mean
density of the collector on the flotation rate of galena particles of varying size. Sb.
Nauchn. Tr., Gos. Nauchn.-Issled. Inst. Tsvetn. Metal. 19: 44--62. Chem. Abstr., 60:
3761.
Morris, T.M., 1950. Measurement of equilibrium forces between an air bubble and an
attached solid in water. Trans. Am. Inst. Min. Metall. Eng., 187: 91--96.
Morris, T.M., 1952. Measurement and evaluation of the rate of flotation as a function of
particle size. Trans. Am. Inst. Min. Metall. Eng., 193: 794--798.
Polkin, S.I., Laptev, S.F., Matsuev, L.P., Adamov, E.V., Krasnukhina, A.V. and Purvinskii,
O.F., 1973. Theory and practice in the flotation of cassiterite fines. In: M.J. Jones (Editor)~
Proc. 10th Int. Miner. Process. Congr., London 1973. Inst. Min. Metall., 1974, pp.
593--614.
Pope, M.I. and Sutton, D.I., 1971--1972. Collector adsorption during froth flotation.
Powder Technol., 5: 101--104.
327

Ravitz, S.F. and Wall, W.A., 1934. The adsorption of copper sulphate by sphalerite and
its relation to flotation. J. Phys. Chem., 38: 13--18.
Reay, D. and Ratcliff, G.A., 1973. Removal of fine particles from water by dispersed air
flotation: effects of bubble size and particle size on flotation efficiency. Can. J. Chem.
Eng., 51: 178--185.
Rey, M., 1968. Quelques probl~mes pratiques et th~oretiques de la flotation des minerais,
8th Int. Miner. Process. Congr., Leningrad, Paper D-1.
Richardson, P.E. and Maust, E.E., Jr., 1976. Surface stoichiometry of galena in aqueous
electrolytes and its effect on xanthate interactions. In: M.C. Fuerstenau (Editor), A.M.
Gaudin Memorial Flotation Symposium. AIME, New York, N.Y., Vol. 1, pp. 364--392.
392.
Robinson, A.J., 1959--1960. Relationship between particle size and collector concentra-
tion. Trans. Inst. Min. Metall., 69: 45--62.
Sastry, K.V.S., 1978. Flotation separation of mineral fines. N.S.F. Workshop on Benefici-
ation of Mineral Fines -- Problems and Research Needs, Tuxedo, N.Y.
Schuhmann, R., Jr., 1942. Flotation kinetics, 1. Methods for steady-state study of flota-
tion problems. J. Phys. Chem., 46: 891--902.
Ser, F., MacDonald, J.D., Whyte, R.M. and Hillary, J.E., 1970. Sulphydric flotation of
previously sulphidized oxide copper minerals of Nchanga Consolidated Copper Mines
Limited. Rudy (Prague), 18(5): 169--174.
Somasundaran, P., 1975. Fine particles treatment. N.S.F. Workshop on Research Needs
in Mineral Processing, New York, Report, P. Somasundaran and D.W. Fuerstenau
(Editors), 1976, pp. 125--133.
Somasundaran, P., 1979. Processing mineral fines. Eng. Min. J., 180 (12): 64--68.
Steiner, H.J., 1974. Kinetic aspects of the flotation behaviour of locked particles. In:
M.J. Jones (Editor), Proc. 10th Int. Miner. Process. Congr., London, 1973. Inst. Min.
Metall., 1974, pp. 653--666.
Sutherland, K.L., 1948. Kinetics of the flotation process. J. Phys. Chem., 52: 394--425.
Sven-Nilsson, I., 1935. Effect of contact time between mineral and air bubble on flotation.
Ing. Vetenskap° Akad. Handl., No. 153.
Tomlinson, H.S. and Fleming, M.G., 1963. Flotation rate studies. In: A. Roberts (Editor),
Proc. 6th Int. Miner. Process. Congr., Cannes, Pergamon, 1965, pp. 563--579.
Trahar, W.J., 1976. The selective flotation of galena from sphalerite with special reference
to the effects of particle size. Int. J. Miner. Process., 3: 151--166.
Trahar, W.J. and Warren, L.J., 1976. The flotability of very fine particles -- a review. Int.
J. Miner. Process., 3: 103--131.
Wade, W.H., Cole, E.D., Meyer, D.E. and Hackerman, N., 1961. Adsorptive behaviour of
fused quartz powders. In: L.E. Copeland (Editor), Advances in Chemistry Series: No.
33, Solid Surfaces and the Gas-Solid Interface. Am. Chem. Soc., pp. 35--41.
Wark, I.W., 1933. Physical chemistry of flotation, 1. The significance of contact angle
in flotation. J. Phys. Chem., 37: 623--644.
Welch, A.J.E., 1952. The relation of crystal lattice discontinuities to mineral dressing.
Recent Developments in Mineral Dressing, Proc. Syrup., London, 1952. Inst. Min.
Metall., 1953, pp. 387--392.

You might also like