2250 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
water. The suggested collector dosage was 260 g/t corre-
sponding to 16.4 mL for 1.4 kg of ore. Out of this quantity,
80% (29.12 mL) was added at the rougher and the rest
(7.12 mL) at the scavenger flotation stage. Frother dosage
of ca. 30 g/t was followed and accordingly 14 drops were
added at the rougher and 3 at the scavenger stage, with each
drop weighting 4.4 mg.
Conditioning of the pulp prior to addition of the
reagents was carried out as follows. Hot water of about
60 °C was added into the flotation cell together with the
ground pulp from the mill. Warm air was blown from out-
side to maintain pulp temperature at the desired levels. For
the tests at low temperatures, ice sockets stored in advance
inside a freezer were placed into the cell, as illustrated in
Figure 2. Before addition of the reagents, the white ice
socket (on the left) was removed, and cold water with tem-
perature of 4–6 °C was used to compensate the volume, as
well as wash water to skim the froth fraction.
To reach statistical representativeness, the tests were
repeated for each temperature (15, 20, 25, 30 and 35 °C),
with the rest parameters being kept constant. Therefore the
reported data present a mean from the duplicated tests.
From each flotation test, five products were recovered in
total: four concentrate fraction and bulk tailings, and their
chemistry was accordingly assayed.
Characterization
Chemical composition. ICP-AES was used for copper and
iron analysis, following an indispensable acid digestion of
the slags with aqua regia. XRF assay (BRUKER S8 TIGER)
was done for complete characterization of the slags. The
matrix with the higher confidence level was that of Fe3O4.
Another purpose of using XRF spectrometry was to bench-
mark XRF and ICP-AES protocols. Sulfur content was
determined by Leco analyzer. Table 2 represents main com-
position of the slags as determined by the above-described
methods.
Mineralogy. Mineralogical quantification (X-ray
diffraction) has been done using a Bruker D8-ECO dif-
fractometer, with CuKα radiation of (λ=1,9373 Å). The
respective sample was pulverized before being scanned
between 5 and 75 °2θ at a speed of 0.02 °2θ per second.
The identification of all minerals from the XRD patterns
was done with a Panalytical Xpert suite and WebPDF4 +
ICDD relational database. Figure 3 depicts the XRD data
for the untreated feed slag. It could be seen, that apart from
fayalite and magnetite being largely present, pyroxene is
detected, specifically identified as “diopside.” The refined
mesh parameters obtained from the analysis, namely a =
9.79 Å, b =8.94 Å, c =5.29 Å, and β =105.77°, sug-
gest values larger than those observed for pure diopside,
yet smaller than those for hedenbergite. Consequently, the
composition is likely to be closer to augite. It is important
to note, that the detected “diopside” is, in fact, a different
pyroxene variant with a higher iron content, and that these
mixed-phase compositions (as detected by the automated
mineralogy) align well with augite.
A SEM-based automated mineralogy inspection was
accomplished using a Zeiss Sigma300 system controlled
by a Mineralogic Mining software. The SEM-EDS system
was equipped with two Bruker xFlash 6|30 X-ray detectors.
System magnification was set to 6000× and voltage tension
to 20kV.
Sub-sampling of product streams for thin sections
preparation was performed with a rotary splitter to ensure
statistically reliable samples. The polished sections prepara-
tion for the SEM analysis was based on four-step procedure: Figure 2. Flotation cell setup for the tests with cooled pulp
Table 2. Mean chemical composition of the feed slags
Content, %Fe Cu S Ca Al Zn K Pb
ICP-AES 42.4 2.47
XRF 42.9 2.41 2.17 2.05 1.09 0.74 0.34
Leco 0.86
water. The suggested collector dosage was 260 g/t corre-
sponding to 16.4 mL for 1.4 kg of ore. Out of this quantity,
80% (29.12 mL) was added at the rougher and the rest
(7.12 mL) at the scavenger flotation stage. Frother dosage
of ca. 30 g/t was followed and accordingly 14 drops were
added at the rougher and 3 at the scavenger stage, with each
drop weighting 4.4 mg.
Conditioning of the pulp prior to addition of the
reagents was carried out as follows. Hot water of about
60 °C was added into the flotation cell together with the
ground pulp from the mill. Warm air was blown from out-
side to maintain pulp temperature at the desired levels. For
the tests at low temperatures, ice sockets stored in advance
inside a freezer were placed into the cell, as illustrated in
Figure 2. Before addition of the reagents, the white ice
socket (on the left) was removed, and cold water with tem-
perature of 4–6 °C was used to compensate the volume, as
well as wash water to skim the froth fraction.
To reach statistical representativeness, the tests were
repeated for each temperature (15, 20, 25, 30 and 35 °C),
with the rest parameters being kept constant. Therefore the
reported data present a mean from the duplicated tests.
From each flotation test, five products were recovered in
total: four concentrate fraction and bulk tailings, and their
chemistry was accordingly assayed.
Characterization
Chemical composition. ICP-AES was used for copper and
iron analysis, following an indispensable acid digestion of
the slags with aqua regia. XRF assay (BRUKER S8 TIGER)
was done for complete characterization of the slags. The
matrix with the higher confidence level was that of Fe3O4.
Another purpose of using XRF spectrometry was to bench-
mark XRF and ICP-AES protocols. Sulfur content was
determined by Leco analyzer. Table 2 represents main com-
position of the slags as determined by the above-described
methods.
Mineralogy. Mineralogical quantification (X-ray
diffraction) has been done using a Bruker D8-ECO dif-
fractometer, with CuKα radiation of (λ=1,9373 Å). The
respective sample was pulverized before being scanned
between 5 and 75 °2θ at a speed of 0.02 °2θ per second.
The identification of all minerals from the XRD patterns
was done with a Panalytical Xpert suite and WebPDF4 +
ICDD relational database. Figure 3 depicts the XRD data
for the untreated feed slag. It could be seen, that apart from
fayalite and magnetite being largely present, pyroxene is
detected, specifically identified as “diopside.” The refined
mesh parameters obtained from the analysis, namely a =
9.79 Å, b =8.94 Å, c =5.29 Å, and β =105.77°, sug-
gest values larger than those observed for pure diopside,
yet smaller than those for hedenbergite. Consequently, the
composition is likely to be closer to augite. It is important
to note, that the detected “diopside” is, in fact, a different
pyroxene variant with a higher iron content, and that these
mixed-phase compositions (as detected by the automated
mineralogy) align well with augite.
A SEM-based automated mineralogy inspection was
accomplished using a Zeiss Sigma300 system controlled
by a Mineralogic Mining software. The SEM-EDS system
was equipped with two Bruker xFlash 6|30 X-ray detectors.
System magnification was set to 6000× and voltage tension
to 20kV.
Sub-sampling of product streams for thin sections
preparation was performed with a rotary splitter to ensure
statistically reliable samples. The polished sections prepara-
tion for the SEM analysis was based on four-step procedure: Figure 2. Flotation cell setup for the tests with cooled pulp
Table 2. Mean chemical composition of the feed slags
Content, %Fe Cu S Ca Al Zn K Pb
ICP-AES 42.4 2.47
XRF 42.9 2.41 2.17 2.05 1.09 0.74 0.34
Leco 0.86