3510 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
Anodic reaction: R(s) =Rn+(aq.) +ne– (1)
Cathodic reaction: Mm+(aq.)+ me – =M(s) (2)
Therefore, in this study, various experimental conditions,
including solution pH, temperature, molar ratio of Co
or Ni over Al, mixing ratio of FeS2 and Al, and particle
size of FeS2, were investigated to assess the applicability of
this newly developed method on FeS2. By using the two
magnetic elements, this study also explores the differences
in their deposition behavior on pyrite. In addition, mag-
netic separation was applied after all tests on the particles
obtained after the cementation reaction to determine the
magnetic fraction.
MATERIALS &METHODS
Co and Ni sulfate solutions were prepared by dissolving
cobalt sulfate heptahydrate (CoSO4·7H2O, purity 99.9%,
Chem Supply, Pty Ltd.) and nickel sulfate hexahydrate
(NiSO4·6H2O, purity 99.9%, Sigma Aldrich, Ltd.) in
deionized (DI) water. The concentration of Ni and Co was
calculated by the molar ratio of their reaction with Al, using
the below Eq. (3) and (4). The molar ratio varied from 0.50
(about 40,000 mg/L), 0.25 (about 20,000 mg/L), and 0.10
(about 10,000 mg/L). The initial pH was adjusted at pH 2
using 4 mol/L sulfuric acid (H2SO4, 98%, Ajax Finechem,
Pty Ltd.) and measured using a pH meter (TPS, WP-80,
Pty Ltd.).
3Ni2+
(aq.) +2Al(s) =3Co(s) +2Al3+
(aq.) (3)
3Co2+
(aq.) +2Al(s) =3Co(s) +2Al3+
(aq.) (4)
The particle size of the Al powder used was 45 μm (Purity
99.9%, Barnes, Pty Ltd.). FeS2 sample was ground using
a vibratory disc pulverizer (ring mill, C+PB, ROCKLAB,
Ltd.) and sieved to generate particles of less than 75 μm.
The D50 and D80 of the prepared sample were determined
to be 26.5 μm and 70.2 μm, respectively (Mastersizer 2000,
Malvern Penalytical, Ltd.). The FeS2 powder used was a
pure sample, but to make sure adhered impurities on the
surface are removed, the sample was ultrasonically washed,
as described in the study by McKibben and Barnes (1986).
The washing procedure followed included: (1) ultrasoni-
cally ethanol washing for 30 s (UCS-10, Lab companion,
Ltd.), (2) 2% HNO3 washing (70%, Ajax Finechem, Pty
Ltd.) for 1 min, and (3) filtration and DI water washing
followed by ethanol.
Once 10 mL of the prepared magnetic metal (Co or
Ni) solution was added into the 50 mL Erlenmeyer flask,
N2 gas was purged into the solution for 15 min to remove
the dissolved oxygen (O2). Afterward, a pre-determined
amount of 0.24 g of Al powder and 0.16 g of FeS2 sample
were added into the flask and tightly capped with the rub-
ber stopper. The flask was placed in a water bath shaker
(ZWY-110X30, Labwit Scientific, Pty Ltd.) and shaken at
pre-determined temperatures (25 °C, 40 °C, and 50 °C)
and 100 rpm. The flasks were collected at pre-determined
time intervals (1 h, 3 h, 5 h, and 6 h). The solution was
filtered using a 0.45 μm membrane filter and immediately
diluted using 2% HNO3.
The obtained particles were washed using DI water fol-
lowed by ethanol to suppress the surface oxidation of FeS2
and dried at room temperature (about 25 °C). After that,
the dried particles were separated from the mixture using
the neodymium (Nd) disc magnet (d=18 mm and h=24
mm). The mass of magnetic and non-magnetic fractions
were used to determine the magnetic separation fraction
(MS, %)according to the following equation Eq. (5) where
M.F. is the amount of magnetic fraction and N.F. is the
amount of nonmagnetic fraction.
Magnetic fraction M.F. M.F. N.F.
M.F. 100 #=+^h (5)
RESULTS &DISCUSSION
Figure 1. shows the influence of the feed type, FeS2, Al,
and the mixture of FeS2 and Al, on the magnetic fraction
with Ni and Co cementation. Due to the semiconductor
property of FeS2, the galvanic interaction between Al and
FeS2 was expected to facilitate electron transfer from Al to
FeS2, forming metal deposition on the surface of FeS2. The
Al powder can act as a reducing agent for both Co and Ni
cementation, resulting in magnetic deposition on its sur-
face. On the other hand, the use of only FeS2 cannot offer
electrons like Al powder for cementation reactions and
does not make any magnetic fraction, as shown in Figure 1.
However, when adding FeS2 and Al simultaneously, the
magnetic fraction increases to about 100 wt. %with Ni
cementation and 88.93 wt. %with Co cementation. From
these results, the interaction between FeS2 and Al enhanced
the reaction to make a magnetic deposition on the surface
of FeS2.
Figure 2. shows the magnetic fraction with the final
pH value of the solution using Ni or Co cementation over
time. With Ni cementation, the magnetic fraction remark-
ably improved to about 100 wt. %within 5 h. The final
recovery efficiency and grade of FeS2 achieved were 100
wt. %and 47.23 wt. %,respectively. Notably, the pH of
the solution was observed to be related to the formation
of magnetic deposition. The achievement of the magnetic
fraction can be estimated by the increase in pH over time
where the magnetic fraction is present, pH is increased by
Anodic reaction: R(s) =Rn+(aq.) +ne– (1)
Cathodic reaction: Mm+(aq.)+ me – =M(s) (2)
Therefore, in this study, various experimental conditions,
including solution pH, temperature, molar ratio of Co
or Ni over Al, mixing ratio of FeS2 and Al, and particle
size of FeS2, were investigated to assess the applicability of
this newly developed method on FeS2. By using the two
magnetic elements, this study also explores the differences
in their deposition behavior on pyrite. In addition, mag-
netic separation was applied after all tests on the particles
obtained after the cementation reaction to determine the
magnetic fraction.
MATERIALS &METHODS
Co and Ni sulfate solutions were prepared by dissolving
cobalt sulfate heptahydrate (CoSO4·7H2O, purity 99.9%,
Chem Supply, Pty Ltd.) and nickel sulfate hexahydrate
(NiSO4·6H2O, purity 99.9%, Sigma Aldrich, Ltd.) in
deionized (DI) water. The concentration of Ni and Co was
calculated by the molar ratio of their reaction with Al, using
the below Eq. (3) and (4). The molar ratio varied from 0.50
(about 40,000 mg/L), 0.25 (about 20,000 mg/L), and 0.10
(about 10,000 mg/L). The initial pH was adjusted at pH 2
using 4 mol/L sulfuric acid (H2SO4, 98%, Ajax Finechem,
Pty Ltd.) and measured using a pH meter (TPS, WP-80,
Pty Ltd.).
3Ni2+
(aq.) +2Al(s) =3Co(s) +2Al3+
(aq.) (3)
3Co2+
(aq.) +2Al(s) =3Co(s) +2Al3+
(aq.) (4)
The particle size of the Al powder used was 45 μm (Purity
99.9%, Barnes, Pty Ltd.). FeS2 sample was ground using
a vibratory disc pulverizer (ring mill, C+PB, ROCKLAB,
Ltd.) and sieved to generate particles of less than 75 μm.
The D50 and D80 of the prepared sample were determined
to be 26.5 μm and 70.2 μm, respectively (Mastersizer 2000,
Malvern Penalytical, Ltd.). The FeS2 powder used was a
pure sample, but to make sure adhered impurities on the
surface are removed, the sample was ultrasonically washed,
as described in the study by McKibben and Barnes (1986).
The washing procedure followed included: (1) ultrasoni-
cally ethanol washing for 30 s (UCS-10, Lab companion,
Ltd.), (2) 2% HNO3 washing (70%, Ajax Finechem, Pty
Ltd.) for 1 min, and (3) filtration and DI water washing
followed by ethanol.
Once 10 mL of the prepared magnetic metal (Co or
Ni) solution was added into the 50 mL Erlenmeyer flask,
N2 gas was purged into the solution for 15 min to remove
the dissolved oxygen (O2). Afterward, a pre-determined
amount of 0.24 g of Al powder and 0.16 g of FeS2 sample
were added into the flask and tightly capped with the rub-
ber stopper. The flask was placed in a water bath shaker
(ZWY-110X30, Labwit Scientific, Pty Ltd.) and shaken at
pre-determined temperatures (25 °C, 40 °C, and 50 °C)
and 100 rpm. The flasks were collected at pre-determined
time intervals (1 h, 3 h, 5 h, and 6 h). The solution was
filtered using a 0.45 μm membrane filter and immediately
diluted using 2% HNO3.
The obtained particles were washed using DI water fol-
lowed by ethanol to suppress the surface oxidation of FeS2
and dried at room temperature (about 25 °C). After that,
the dried particles were separated from the mixture using
the neodymium (Nd) disc magnet (d=18 mm and h=24
mm). The mass of magnetic and non-magnetic fractions
were used to determine the magnetic separation fraction
(MS, %)according to the following equation Eq. (5) where
M.F. is the amount of magnetic fraction and N.F. is the
amount of nonmagnetic fraction.
Magnetic fraction M.F. M.F. N.F.
M.F. 100 #=+^h (5)
RESULTS &DISCUSSION
Figure 1. shows the influence of the feed type, FeS2, Al,
and the mixture of FeS2 and Al, on the magnetic fraction
with Ni and Co cementation. Due to the semiconductor
property of FeS2, the galvanic interaction between Al and
FeS2 was expected to facilitate electron transfer from Al to
FeS2, forming metal deposition on the surface of FeS2. The
Al powder can act as a reducing agent for both Co and Ni
cementation, resulting in magnetic deposition on its sur-
face. On the other hand, the use of only FeS2 cannot offer
electrons like Al powder for cementation reactions and
does not make any magnetic fraction, as shown in Figure 1.
However, when adding FeS2 and Al simultaneously, the
magnetic fraction increases to about 100 wt. %with Ni
cementation and 88.93 wt. %with Co cementation. From
these results, the interaction between FeS2 and Al enhanced
the reaction to make a magnetic deposition on the surface
of FeS2.
Figure 2. shows the magnetic fraction with the final
pH value of the solution using Ni or Co cementation over
time. With Ni cementation, the magnetic fraction remark-
ably improved to about 100 wt. %within 5 h. The final
recovery efficiency and grade of FeS2 achieved were 100
wt. %and 47.23 wt. %,respectively. Notably, the pH of
the solution was observed to be related to the formation
of magnetic deposition. The achievement of the magnetic
fraction can be estimated by the increase in pH over time
where the magnetic fraction is present, pH is increased by