XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 2451
minerals. The only reference to the authors knowledge is a
case study presented by Ruonal et al. (1997) on the separa-
tion of gersdorffite (NiAsS) from pentlandite (Fe4.5Ni4.5S8)
in a nickel ore where Eh control was accomplished by vary-
ing the air/nitrogen ratio during flotation. Several patents
have been filed concerning the use of Eh control in flota-
tion of nickel and nickel-iron sulfides though these are
concerned with separation of the sulfides from non-sulfidic
gangue (Clark and Newell, 1998 Senior and Ahveninen,
1998 Senior et al., 2009 Pyke et al., 2003) or the differ-
ential flotation of chalcopyrite, pentlandite and pyrrhotite
from their ores (Wells et al., 1981) whilst Chimonyo et al.
(2017a, b) studied the effect of pulp potential during flota-
tion of base metal sulfides (chalcopyrite, pentlandite, pyr-
rhotite) from non-sulfidic gangue. Industrially, Eh control
appears to have been implemented in the flotation of cop-
per and nickel sulfides from low-grade, disseminated and
serpentinized ores from the Hitura and Vammala mines
(Finland) with measurement and control of Eh via synthetic
metal sulfide electrodes rather than noble metal electrodes
(Ruonal et al., 1997 Woods, 2003). Woods (2003) pro-
vides a summary of other industrial applications of pulp
potential control during froth flotation.
The case study presented here is a follow up to earlier
published work on the flotability of millerite (Smith et al.,
2011) and pentlandite (Senior et al., 1994) where it was evi-
dent from comparison of pulp potential vs mineral recov-
ery data using xanthate as a collector, that a window exists
where one mineral could be separated from the other by
control of the Eh. The purpose of this study was three-fold
firstly, to confirm if it is indeed possible, based on previous
single mineral flotation data and theoretical considerations,
to selectively separate millerite from pentlandite by Eh con-
trol, secondly, to determine the effect of collector type on
mineral flotation as a function of Eh, and finally, to build
upon the work of Senior et al. (1994) by obtaining flotation
data for pentlandite above the air-set potential and with
different collectors. An understanding of the flotation char-
acteristics of millerite and pentlandite is of interest given
that millerite is found in disseminated nickel sulfide depos-
its (Barnes et al., 2011) and is an important constituent of
some nickel sulfide deposits in Australia such as Mt Keith,
Black Swan, and Otter-Juan (Dowling et al., 2004 Grguric
et al., 2006 Keele and Nickel, 1974). Additionally, the
ability to potentially separate the two nickel sulfides by flo-
tation and produce separate concentrates may be advanta-
geous given the high Ni-content of millerite (~65%). Such
an approach could enable production of a separate high
Ni, low Fe-concentrate and low Ni, high Fe-concentrate,
and as suggested by Grguric et al. (2006), this allows for
the possibility to directly refine millerite-rich concentrates
without the need for smelting.
EXPERIMENTAL
Nickel Sulphide Sample
The nickel sulphide sample used in this study was concen-
trated from a high-grade millerite-pentlandite specimen
from Clarabelle, Canada. The sample was stage-crushed in
a laboratory jaw crusher to pass 1650 µm and screened at
212 µm to remove low-grade fines. A bulk –1650 +212
µm millerite composite was thus prepared. The chemical
analysis of the nickel sulphide sample after preparation was
55.2% Ni, 11.2% Fe, 30.3% S, 1.58% Si, 0.07% Cu and
0.15% Co. Quantitative x-ray diffraction (QXRD) analy-
sis (not shown here) confirmed that the sample contained
51.9% millerite, 47.4% pentlandite and 0.7% quartz.
Five size fractions were prepared ranging from –1700
µm to +425 µm and were mounted and examined by
Scanning Electron Microscopy (SEM) to determine the
extent of mineral liberation. There was no way of distin-
guishing between millerite and pentlandite based on tex-
ture or contrast, so elemental mapping of Ni, Fe and S was
used. Where only Ni and S was present, this was identi-
fied as millerite while the presence of all three indicated
pentlandite. Pentlandite and millerite were present as both
liberated and composite particles as seen by the SEM maps
in Figure 1. Qualitatively, better liberation was seen in the
finer sizes, with the pentlandite and millerite well liberated
in the +0.60 mm and +0.425 mm fractions. There were
some locked particles and intergrowths comprising nickel
sulphides and silicate gangue mostly associated with pent-
landite. Energy Dispersive X-Ray (EDX) analysis of some
individual pentlandite and millerite grains was also carried
out and data are presented in Table 1. Pentlandite in the
nickel sulfide sample contained on average 45.8% Ni and
23.5% Fe, the high Ni content being characteristic of pent-
landite in pentlandite-millerite assemblages (Harris and
Nickel, 1972).
Quartz
High quality quartz, obtained and prepared locally, was
used as the diluent in all the single mineral flotation tests.
The amount used in each test was 450 g of –1650 +212 µm
quartz.
Reagents
The collectors used were a high purity (99%) potassium
ethyl xanthate (KeX) from Sigma-Aldrich, a commercial
grade thionocarbamate, Cytec 3894, and a commercial
grade dithiophosphate, Cytec Aero 3501. A dilute solution
minerals. The only reference to the authors knowledge is a
case study presented by Ruonal et al. (1997) on the separa-
tion of gersdorffite (NiAsS) from pentlandite (Fe4.5Ni4.5S8)
in a nickel ore where Eh control was accomplished by vary-
ing the air/nitrogen ratio during flotation. Several patents
have been filed concerning the use of Eh control in flota-
tion of nickel and nickel-iron sulfides though these are
concerned with separation of the sulfides from non-sulfidic
gangue (Clark and Newell, 1998 Senior and Ahveninen,
1998 Senior et al., 2009 Pyke et al., 2003) or the differ-
ential flotation of chalcopyrite, pentlandite and pyrrhotite
from their ores (Wells et al., 1981) whilst Chimonyo et al.
(2017a, b) studied the effect of pulp potential during flota-
tion of base metal sulfides (chalcopyrite, pentlandite, pyr-
rhotite) from non-sulfidic gangue. Industrially, Eh control
appears to have been implemented in the flotation of cop-
per and nickel sulfides from low-grade, disseminated and
serpentinized ores from the Hitura and Vammala mines
(Finland) with measurement and control of Eh via synthetic
metal sulfide electrodes rather than noble metal electrodes
(Ruonal et al., 1997 Woods, 2003). Woods (2003) pro-
vides a summary of other industrial applications of pulp
potential control during froth flotation.
The case study presented here is a follow up to earlier
published work on the flotability of millerite (Smith et al.,
2011) and pentlandite (Senior et al., 1994) where it was evi-
dent from comparison of pulp potential vs mineral recov-
ery data using xanthate as a collector, that a window exists
where one mineral could be separated from the other by
control of the Eh. The purpose of this study was three-fold
firstly, to confirm if it is indeed possible, based on previous
single mineral flotation data and theoretical considerations,
to selectively separate millerite from pentlandite by Eh con-
trol, secondly, to determine the effect of collector type on
mineral flotation as a function of Eh, and finally, to build
upon the work of Senior et al. (1994) by obtaining flotation
data for pentlandite above the air-set potential and with
different collectors. An understanding of the flotation char-
acteristics of millerite and pentlandite is of interest given
that millerite is found in disseminated nickel sulfide depos-
its (Barnes et al., 2011) and is an important constituent of
some nickel sulfide deposits in Australia such as Mt Keith,
Black Swan, and Otter-Juan (Dowling et al., 2004 Grguric
et al., 2006 Keele and Nickel, 1974). Additionally, the
ability to potentially separate the two nickel sulfides by flo-
tation and produce separate concentrates may be advanta-
geous given the high Ni-content of millerite (~65%). Such
an approach could enable production of a separate high
Ni, low Fe-concentrate and low Ni, high Fe-concentrate,
and as suggested by Grguric et al. (2006), this allows for
the possibility to directly refine millerite-rich concentrates
without the need for smelting.
EXPERIMENTAL
Nickel Sulphide Sample
The nickel sulphide sample used in this study was concen-
trated from a high-grade millerite-pentlandite specimen
from Clarabelle, Canada. The sample was stage-crushed in
a laboratory jaw crusher to pass 1650 µm and screened at
212 µm to remove low-grade fines. A bulk –1650 +212
µm millerite composite was thus prepared. The chemical
analysis of the nickel sulphide sample after preparation was
55.2% Ni, 11.2% Fe, 30.3% S, 1.58% Si, 0.07% Cu and
0.15% Co. Quantitative x-ray diffraction (QXRD) analy-
sis (not shown here) confirmed that the sample contained
51.9% millerite, 47.4% pentlandite and 0.7% quartz.
Five size fractions were prepared ranging from –1700
µm to +425 µm and were mounted and examined by
Scanning Electron Microscopy (SEM) to determine the
extent of mineral liberation. There was no way of distin-
guishing between millerite and pentlandite based on tex-
ture or contrast, so elemental mapping of Ni, Fe and S was
used. Where only Ni and S was present, this was identi-
fied as millerite while the presence of all three indicated
pentlandite. Pentlandite and millerite were present as both
liberated and composite particles as seen by the SEM maps
in Figure 1. Qualitatively, better liberation was seen in the
finer sizes, with the pentlandite and millerite well liberated
in the +0.60 mm and +0.425 mm fractions. There were
some locked particles and intergrowths comprising nickel
sulphides and silicate gangue mostly associated with pent-
landite. Energy Dispersive X-Ray (EDX) analysis of some
individual pentlandite and millerite grains was also carried
out and data are presented in Table 1. Pentlandite in the
nickel sulfide sample contained on average 45.8% Ni and
23.5% Fe, the high Ni content being characteristic of pent-
landite in pentlandite-millerite assemblages (Harris and
Nickel, 1972).
Quartz
High quality quartz, obtained and prepared locally, was
used as the diluent in all the single mineral flotation tests.
The amount used in each test was 450 g of –1650 +212 µm
quartz.
Reagents
The collectors used were a high purity (99%) potassium
ethyl xanthate (KeX) from Sigma-Aldrich, a commercial
grade thionocarbamate, Cytec 3894, and a commercial
grade dithiophosphate, Cytec Aero 3501. A dilute solution