Glasnik Rudarsko-geološko-građevinskog fakulteta

Journal of Faculty of Mining, Geology and Civil Engineering

ISSN: Vol. 2013/1, pp. 25-32
Paper published at XIV Balkan Mineral Processing Congress, Tuzla, June 2011

C.T. O’Connor1
N. Chapman2
J. Mishura3
T. Khonthu4
N. Shackleton5
J. Wise6
K.C. Corin7

THE EFFECT OF VARIOUS COMMINUTION PROCEDURES ON THE FLOTATION OF SULFIDE ORES

Abstract: In the flotation process inadequate attention is given to the effect which different comminution procedures may have on the
floatability of the ore. Modern concentrators make use of a wide range of comminution devices and sometimes dry milling is used
instead of wet milling. All of these factors can produce particles with different flotation characteristics. The DO and pH levels in the
mill also play a role in affecting the floatability of the ore. This paper presents results of a set of investigations into the effect of
various comminution procedures and conditions on the flotation of base metal sulphide ores.

Key Words: Comminution, Flotation, Sulphides

EFEKTIVNOST RAZLIČITIH PROCEDURA USITNJAVANJA NA FLOTABILNOST SULFIDNIH RUDA

Abstract: U procesu flotacije adekvatna pažnja se ne pridaje efektu koji različite procedure usitnjavanja mogu imati na flotabilnost
rude. Moderni koncentratori koriste veliki broj mašina različitog koncepta usitnjavanja, a u nekim slučajevima bolja koncentracija
se dobiva suhim u odnosu na mokro mljevenje. Koncept mljevenja utiče na proizvodnju čestica sa različitim flotabilnim svojstvima.
DO i PH nivoi mlina takođe utiču na flotabilnost rude. Ovaj rad prezentuje rezultate niza proučavanja efekata različitih procedura
usitnjavanja i uvjeta na flotabilnost metala iz sulfidnih ruda.

Key Words: usitnjavanje, flotacija, sulfidi

1-7
Centre for Minerals Research, Department of Chemical Engineering, University of Cape Town, Private Bag X3, Rondebosch
7701, South Africa, Cyril.oconnor@uct.ac.za
1. INTRODUCTION

In the preparation of ore for flotation, the various grinding procedures used can influence the flotation
performance of the valuable minerals by changing the degree of liberation, the particle size distribution or other particle
characteristics such as morphology. With respect to liberation it has been proposed in a study of a copper/lead/zinc ore,
using MLA to analyze the mineralogy of the samples, that the composition distribution of a given size fraction from a
comminution circuit is constant for that mineral in that ore regardless of whether the particle has been crushed, ground
or subjected to impact breakage or compressive breakage and also regardless of the overall size distribution of the
comminution product [1]. However such constant composition distribution of particles does not necessarily infer that
they would have similar flotation characteristics. The comminution process should ultimately be designed to produce
particles which are conducive to downstream processes of which generally the most important is flotation. Hence the
various studies carried out on the differences between various comminution devices should include an evaluation of the
floatability of the mill product.
Modern concentrators make use of a variety of crushing devices such as high pressure grinding rolls (HPGR),
cone crushing, rod and ball mills, Isamills, etc. Various claims have been made about the ability of these different
devices to enhance liberation. For example, it has been proposed that HPGR as a comminution device not only
improves energy efficiency but may also lead to preferential liberation of valuables due to micro-cracking along grain
boundaries and finer grinding [2-6]. The Isamill results in the production of ultrafine material which, although better
liberated, may also result in the poor flotation recovery often associated with fine particles. The grinding media (e.g.
stainless vs. mild steel) or the grinding environment (wet vs. dry) can also affect the surface properties of the ore
particles, due to different galvanic interactions associated with such slurry conditions, and hence may result in different
flotation behavior.
In many concentrators the major focus of attention has traditionally been to produce a particle size distribution
which is most conducive to flotation. Much attention has also been given in recent years to developing devices which
will reduce the energy consumption in the comminution process. However the need to ensure that the particles of any
particular size are conducive to flotation has not enjoyed the attention it deserves. Simply put, the integration between
comminution and flotation is a key to ensuring the optimization of the overall concentration process. This can be done
by using flotation as a standard diagnostic procedure to evaluate the efficiency of the comminution process. Apart from
the range of new comminution devices being introduced into concentrators the chemical conditions existing during
milling is also a critical aspect that needs to be evaluated in terms of the ultimate flotation behavior.
The present paper presents results obtained as part of an overall programme of research aimed at a greater
understanding of the interactions between the milling and flotation processes. In Part 1 of this work the relationship
between processing ore using HPGR as opposed to cone crushing followed by wet or dry rod milling using either
stainless or mild steel media was investigated using a base metal sulfide (BMS) ore. In Part 2 the influence of
controlling the dissolved oxygen (DO) concentration and pH in the mill (while observing the redox potential (Eh) of the
pulp) on the flotation behavior of a BMS ore containing chalcopyrite, pentlandite and pyrrhotite was investigated. In
this study a highly instrumented laboratory scale Magotteaux mill ® was used. In Part 3 the relationship between
particles of similar size obtained from an Isamill and a ball mill respectively and their flotation behavior was studied
using the same ore as in Part 2.

2. EXPERIMENTAL

Table 1 summarizes the different comminution devices and ore types used in these investigations.

Part Device Ore type Product psd

1 HPGR (60/90 bar) BMS - sphalerite <3.3 mm
1 Cone crusher BMS - sphalerite <3.3 mm
#
1 Rod mill (wet/dry; MS / SS ) BMS - sphalerite d80 = 90 µm
##
2 Magotteaux Mill BMS – Cp, Pn, Po d80 = 75 µm
##
3 Isamill – 4 passes BMS – Cp, Pn, Po d80 = 24 µm
3 Ball mill (SS) BMS – Cp, Pn, Po## d80 = 24 µm
# MS= mild steel rods; SS = stainless steel rods
##Cp = Chalcopyrite; Pn = pentlandite; Po = pyrrhotite
Table 1. Different comminution devices, ore types used and product psd for each process
 In Part 1 the ore used was a BMS in which the dominant sulfide was sphalerite. Ore from the HPGR and cone
crusher was screened at 90 µm and the +90 µm size fraction was fed to the mill. The milling time was determined to
ensure that the reconstituted feed to float of the milled product and the -90 µm size fraction had a P80 of 90 µm. Details
of the procedure used have been described elsewhere [7]. In Part 2 a highly instrumented Magotteaux Mill ® was used
[8] and the ore used was a base metal sulfide (BMS) containing chalcopyrite, pentlandite and pyrrhotite. DO
concentrations of 0 ppm in the mill were obtained by continually purging the system with a pure nitrogen stream. DO =
5.5 or 6.5 ppm (as required) was obtained by purging with an air stream under controlled conditions. The pH in the
slurry was held constant at 9-9.2 by controlled addition of lime. Samples were extracted at various times during the
milling process to determine the concentration of xanthate by measuring the UV absorption at its characteristic
wavelength of 301 nm. The ORP (oxidation – reduction potential) was measured using a platinum electrode. The
grinding media used were 21 % Cr balls. In Part 3 the same ore was used as in Part 2. The feed material was first
crushed in a ball mill using ceramic balls to <106 µm. This material was then fed respectively to an M4 IsaMillTM
containing ceramic beads (3.5 mm) and a ball mill containing stainless steel balls (11 mm). The product after 4 passes
in the IsaMillTM and after 64 minutes in the ball mill both had a d 80 of 24 µm. The mills used are shown in Figure 1.

a) b) c)
Figure 1. Milling devices used in Part 2 and 3 of this study. (a) Magotteaux Mill® (b) Conventional ball mill (c) M4 IsaMillTM
Batch flotation tests were carried out using synthetic plant water (I=2.0E-02; TDS=1030) [9]. Table 2 provides
details of the conditions used for the flotation tests. In Parts 1 and 2 a laboratory batch flotation cell was used and in
Part 3 microflotation tests were carried out. The system has been described elsewhere [10].

Part d80 (µm). Collector Frother Activator pH

1 90 SEX (25 g/t) DF200 (25 g/t) CuSO4 (600 g/t) 9
2 75 SIBX (20 g/t) DF200 (20 g/t) - 9.2
3 24 SIBX (50 g/t) DF200 (50 g/t) - 9

Table 2. Details of conditions used for flotation experiments

3. RESULTS AND DISCUSSION

In Part 1 of this set of investigations it was found that after milling there was very little difference in the particle
size distribution between cone crushing (CC) followed by rod milling, wet (WM) or dry (DM) compared to the HPGR
followed by rod milling (Figure 2). The fact that mild steel (MS) balls were used for dry milling compared to stainless
steel (SS) for wet milling is not relevant for the present purposes. The primary product from HPGR was however finer
than from cone crushing but no significant differences in liberation were observed between the two processes. Figure 3
shows that there were significant differences in the zinc distribution by size fraction between wet and dry milling. The
d50 of the sphalerite in the case of wet milling was 27 µm and 23 µm for cone crushing and the HPGR respectively
whereas the equivalent results in the case of dry milling were 12 µm and 14 µm respectively. Figure 4 shows the
grade-recovery results obtained in the flotation of these samples. As can be seen there was a significant improvement in
the flotation performance after dry milling (DM) compared to wet milling (WM) in the case of both HPGR and cone
crushing (CC) treatment. It should be noted that the dry milled samples were finer than the wet milled samples and thus
would usually tend to show a poorer flotation performance. Chapman et al. [11], in a similar study of a platinum group
mineral ore, showed surprisingly that for that ore dry milling led to a much poorer flotation performance. The major
difference between these two ores is the fact that the PGM ore contains very little sulphides and hence the improved
performance of the dry milled sample may be due to two factors, viz. an increase in the concentration of high energy
sites due to the very abrasive nature of dry milling and to subsequent significant galvanic effects after exposure to water
during flotation. It is also interesting to note how ore specific such experimental results can be. It is also worth noting
that in neither ore did the use of HPGR improve the degree of liberation.
 110

100

90

Cummulative % passing
80

70

60

50

40

30

20

10

0
1 10 100 1000 10000
Particle size (um)

CC-product CC-WM-SS CC-DM-MS

HPGR-product HPGR-WM-SS HPGR-DM-MS

Figure 2. Particle size distribution of the mill feed and mill product sphalerite ore after various pretreatments

70

60

50
Zinc Distribution (%)

40

30

20

10

0
-25 25-38 38-53 53-75 75-90 90-106 106-150 150-212 212-300
Size fraction (µm)

CC-WM-SS CC-DM-MS HPGR-WM-SS HPGR-DM-MS

Figure 3. Zinc distribution versus particle size based on a screen analysis (sphalerite ore).
 60

50

Zn Grade (%) 40

30

20

10

0
0 10 20 30 40 50 60 70 80 90 100
Zn Recovery (%)

CC-WM-SS CC-DM-MS HPGR-WM-SS HPGR-DM-MS

Figure 4. Grade-recovery results after different comminution procedures (sphalerite ore).

The focus in Part 2 of this set of experiments was to study the relationship between the chemical conditions
present during the milling process and the consequent flotation using the BMS ore described previously. Figures 5 and 6
as well as Table 3 show the relevant results obtained in these tests. Figure 5 shows that the recovery of chalcopyrite
(represented by Cu) was hardly affected by the presence or otherwise of air during milling even though the Eh value
increased considerably as air was introduced. The recovery of pentlandite (represented as Ni) decreased as the DO
concentration in the mill increased and this decrease was even more significant in the case of pyrrhotite. It is known that
these two minerals, especially pyrrhotite, are very reactive to oxygen and readily oxidize which would explain the poor
recoveries. The implications of these results for the flotation or depression of pyrrhotite are self-evident.
During milling samples were extracted and analyzed for xanthate concentration. At DO=0 ppm there was hardly
any consumption of xanthate. This is presumably due to the absence of oxygen which is needed for the reaction in
which xanthate is converted into dixanthogen, viz.
( ) ( ) ( ) (1)

During slurry transfer to the flotation cell the DO concentration increased to about 3 ppm and the xanthate
concentration decreased significantly. However, when nitrogen was sparged into the flotation cell (Table 3) the xanthate
concentration increased significantly again. This indicates that the oxidation of xanthate to dixanthogen is reversible
and this is consistent with the findings of Pomianowski and Leja [12] who noted that this reaction was reversible at
pH>8. As shown in Table 3 the recovery of all the minerals decreased after the sparging with nitrogen which further
illustrates the critical role which dixanthogen plays in the recovery of the sulfides given the above discussion of the
reversibility of the xanthate-dixanthogen reaction. The grade of chalcopyrite and pentlandite is strongly influenced by
the recovery of pyrrhotite which is essentially the gangue mineral. However the recoveries are relatively so low after
nitrogen sparging that this effect is hardly noticeable.
Figure 6 shows the variation in recoveries of the various minerals when the pH in the mill was varied from 9.2 to
11 at a DO concentration of 5.5 ppm. As previously noted the recovery of pyrrhotite is adversely affected by the
presence of oxygen due to its being readily oxidized [13]. Chalcopyrite also floated readily at all the pHs tested.
Pentlandite was the most significantly affected by an increase in the pH with its recovery decreasing from 58 % to 41
%. It is also significant to observe the Eh values of the mill slurry in the two figures. At a constant pH of 9.2 the Eh
value increased from about -120 mV to almost 250 mV as the DO concentration increased from 0 to 6.5 ppm (Figure 5).
At a constant DO concentration of 5.5ppm the Eh value decreased from about 225 mV to 150 mV as the pH increased
from 9.2 to 11. The use of Eh to control a flotation system is complex and it seems that the best that one can achieve is
to monitor its value and that controlling the DO and pH values is far more important.
 100 350
250
80
150
Recovery (%) 50

Eh (mV)
60
-50
40 -150
-250
20
-350
0 -450
0 1 6,5
Dissolved oxygen level (in ppm)

Cu Ni Po Eh

Figure 5. Variation in the mineral recovery by flotation after milling at different DO concentrations (Eh values of mill slurry shown;
pH=9.2).

Mineral DO=0/air sparging* DO=0/air sparging* DO=0/N2 sparging* DO=0/N2 sparging*

Recovery (%) Grade (%) Recovery(%) Grade (%)
Chalcopyrite(Cu) 84.2 1.5 55.1 1.2
Pentlandite (Ni) 69.0 4.9 46.0 4.3
Pyrrhotite# 29.8 15.7 16.4 9.7
# Determined by residual S balance assuming pyrrhotite is FeS and negligible pyrite present.
* Sparging by air or nitrogen in the flotation cell after slurry transfer
Table 3. Recoveries and grades after milling at DO=0 and different flotation cell sparging procedures (pH = 9.2)
In Part 3 of this study a preliminary study was made of the effect of using an Isamill compared to a ball mill on
the floatability of the same ore as used in Part 2. Milling curves showed that the d 80 value of the product material from
the Isamill after four passes was the same as that from the ball mill after 64 minutes milling time, viz. 24µm. In both
cases the feed material had a d90 of 106 µm. The equivalent mean particle size obtained after the comminution step was
important in order to enable an investigation into the effect of other particle characteristics such as morphology and
degree of liberation on the flotation properties of the mill product. This was done using a microflotation cell [10].

100,0 250

80,0 200
Recovery (%)

Eh ( mV)

60,0 150

40,0 100

20,0 50

0,0 0
9,2 10 11
pH
Cu Ni Po Eh
Figure 6. Variation in the mineral recovery by flotation after milling at different pHs (DO=5.5ppm) (Eh values of mill slurry shown).
 100

Cumulative mass recovery (%)

90
80
70
60
50 IsaMill
40 Pass4
30
20
10
0
0 5 10 15 20 25
Time (min)

Figure 7. % Mass recovery by flotation time after milling with an Isamill vs. a ball mill at equivalent d 90 of 106 m
As can be seen in Figure 7 the % mass recoveries from the different samples were essentially the same. The
microflotation cell essentially determines the relative hydrophobicities of the particles since it measures recovery in a
non-turbulent environment arising from bubble-particle attachment. Thus these results showed that in these preliminary
experiments there seems to be little difference in particle hydrophobicities between particles produced using the two
different milling devices. Further work is currently under way to determine whether there are any differences between
the shapes of the particles emanating from these two devices. This involves determining the ferret diameter, particles
roundness and the aspect ratio of the particles by particle size using scanning electron micrographs and proprietary
software [14].

4. CONCLUSIONS

These results have shown the importance of evaluating the efficiency of a comminution process not only by
determining the energy consumption or the mill product particle size distribution but most importantly by determining
the floatability of the product particles. In the present study it has been shown that the use of different comminution
devices and different chemical environments viz. presence of water, dissolved oxygen concentration and pH, can also
have a major effect of the activity of the collector molecules present in the mill. Hence different flotation methods such
as batch or microflotation can be used effectively to evaluate the performance of the comminution process.

ACKNOWLEDGEMENTS

The authors wish to thank AngloResearch for permission to publish parts of this paper and to Dr C Greet for providing
the Magotteaux Mill®.

5. LITERATURE

1. Bearman, R.: High-pressure grinding rolls: Characterizing and defining process performance for engineers. In S. Komar
Kawatra (Ed.), Advances in Comminution (pp. 3-14), Denver, Colorado: Society for Mining, Metallurgy and Exploration,
2006.
2. Bradshaw, D.J. and O’Connor, C.T. : Measurement of the sub-process of bubble loading in flotation. Minerals Engineering,
9, 4, pp.443-448, 1996.
3. Chapman, N.A., Shackleton,N.J., Malysiak, V., and O'Connor, C.T.: The effect of using different comminution procedures
on the flotation of Platinum-Group Minerals, Minerals Engineering , In Press, 2011.
4. Cotton, J., OneEighty Engineering Solutions
5. Daniel, M.J.: Energy efficient mineral liberation using HPGR technology. PhD Thesis, University of Queensland, Brisbane,
Australia, 2007.
6. Dunne, R., Goulsbra, A., & Dunlop, I. : High pressure grinding rolls and the effect on liberation: Comparative test results.
Proceedings of Randol Gold Forum (pp. 49-54), Golden, Colorado: Randol International, 1996.
7. Greet, C.J. et al.: The Magotteaux Mill®: investigating the effect of grinding media on pulp chemistry and flotation
performance. Minerals Engineering. 17(7-8),pp. 891-896, 2004.
8. Klymowsky, R., Patzelt, N., Knecht, J., & Burchardt, E. : An overview of HPGR technology. In M.J. Allan (Ed.)
International autogenous and semiautogenous grinding technology :Proceedings of the SAG conference (pp.11-26),
Vancouver, BC, Canada: Department of Mining Engineering, University of British Columbia, 2006.
9. Morley, C.: High pressure grinding rolls: A technology review. In S. Komar Kawatra (Ed.), Advances in Comminution (pp.
15-39), Denver, Colorado: Society for Mining, Metallurgy and Exploration, 2006.
10. N.A. Palm, N.J. Shackleton, V. Malysiak, C.T. O’Connor : The effect of using different comminution procedures on the
flotation of sphalerite, Minerals Engineering 23, 1053–1057, 2010.
11. O’Connor, C.T., and Allison,S.A., An investigation into the flotation behaviour of pyrrhotite, International Journal of
Mineral Processing, 98, pp. 202-207, 2011.
12. Pomianowski, A. and Leja,J. : Spectrophotometric study of xanthate and dixanthogen solutions, Can. J Chem, 41, pp.2219-
2230, 1963.
13. Shackleton,N.J., Malysiak, V., and O'Connor, C.T.: Surface Characteristics and flotation behaviour of platinum and
palladium tellurides, Mineral Engineering 20, pp.1237-1245, 2007.
14. Wightman, E., Evans, C.L., Vizcarra, T., & Sandoval, G.: Ninth International Congress for Applied Mineralogy (pp. 475-
482), Brisbane, Australia: The Australasian Institute of Mining and Metallurgy, 2008.