1822 XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3
of the synthetic magnetite to hematite, which is an antifer-
romagnetic mineral, and cooling affect the Fe recovery in
the subsequent magnetic separation, which must be pre-
vented. The typical conventional cooling process is water
quenching, in which the high-temperature roasted product
is cooled using cool water, and oxidation of synthetic mag-
netite is avoided in the absence of contact with air (Yu et
al., 2020). Although water quenching is an effective cooling
method for the magnetization roasting of iron ore, there
are many practical problems for its industrial application.
For example, it consumes large volumes of water resources,
which largely limits the industrial application of magne-
tization roasting, particularly in regions with water short-
ages. Moreover, there is heat waste during the heat transfer
with the cool water, which makes the temperature of the
ore pulp high in water quenching, which is not conducive
to the subsequent magnetic separation. Thus, developing
a new dry cooling method for magnetization roasting is
essential.
Air cooling is a typical dry cooling method in indus-
trial production, and it is characterized by a low produc-
tion cost and a simple process. However, its application
in magnetization roasting is very difficult. This is because
high-temperature Fe3O4 is readily oxidized to α-Fe2O3 in
an air atmosphere, whereas α-Fe2O3 is a typical weakly
magnetic mineral whose presence affects the Fe recov-
ery in the subsequent low-intensity magnetic separation.
Nevertheless, some studies indicate that Fe3O4 can be oxi-
dized to maghemite (γ-Fe2O3) under controlled oxidation
conditions (Uwadiale, 1992, Rzepa et al., 2016). Because
maghemite is ferrimagnetic, its presence does not affect the
Fe recovery in the magnetic separation step (Thompson,
2012). Wu et al. (2012) investigated the magnetization
roasting of goethite ore and found that Fe3O4 was mainly
oxidized to γ-Fe2O3 in the air-cooling process. However,
γ-Fe2O3 is metastable and readily oxidized to α-Fe2O3 in
the temperature range of 370–600 °C (Cornell and
Schwertmann, 2003). Therefore, the key to the air-cooling
process is to control the oxidation of Fe3O4 to γ-Fe2O3 and
thereby avoid the formation of α-Fe2O3.
Although air cooling is theoretically feasible for the
magnetising roasting of limonite ore, it has not been dem-
onstrated in practice. In this study, the comparison of
cooling processes such as air cooling, water quenching,
and nitrogen cooling was first investigated. The effects of
the different cooling processes on the magnetic separation
index were studied. Additionally, the mineral phase trans-
formation and magnetism variation in different cooling
processes were studied.
EXPERIMENTAL
Materials
The limonite ore used in the study was obtained from
Kunming, China. The sample had a particle size of 60 wt.%
particles less than 74 μm. Chemical composition analysis
of the sample are shown in Table 1. As shown in Table 1,
the iron content was 34.50%, and the FeO content was less
than 0.10%. In addition, the contents of SiO2, CaO, MgO,
and Al2O3 were 34.12%, 0.27%, 0.37%, and 2.40%,
respectively. Meanwhile, the content of harmful element P
was 0.70%. The loss on ignition (LOI) was 6.99%.
Experimental Method
The magnetising roasting and cooling experiments were
conducted using a fluidization roasting system, and the
experimental diagram is shown in Figure 1. A horizontal
furnace (MTI OTF-1200X-S-VT, China) was used to pro-
vide the reaction temperature range from 20 °C to 1200
°C. The reaction gas rate was controlled using a multichan-
nel mass flowmeter (MTI GSL-LCD, China). Magnetising
roasting was conducted in a quartz tube (Φ 50 mm).
In this experiment, first, synthetic magnetite was pre-
pared from limonite ore via magnetization roasting in a
mixed CO, H2 and N2 atmosphere. The roasting condi-
tions were a roasting temperature of 480 °C, CO and H2
concentration of 20%, and reduction time of 20 min. Once
the reduction roasting of limonite ore was completed, the
reduction gas of CO and H2 was stopped, and N2 was intro-
duced to protect the synthetic magnetite. The next stage
was the cooling experiments of high-temperature roasted
ore. In nitrogen cooling process, the roasting furnace was
turned off and the furnace tube was cooled by injecting N2
about 20 min. After cooling to 20 °C, the cooled samples
were obtained. In water quenching, the roasted ore was
directly poured into cooled water of 20 °C, and the sample
was obtained after drying in an oven with a temperature of
100 °C. In air cooling process, the roasted ore was directly
poured into air atmosphere, and the sample was obtained
when its temperature decreased to 20 °C. The roasted ore
was ground using a vertical stirring mill, and the particle
size was 75 wt.% passing 0.043 mm. The ground product
Table 1. Chemical composition analysis of limonite ore (mass, %)
Element Fe FeO SiO2 CaO MgO Al2O3 S P LOI
Content 34.50 0.10 34.12 0.27 0.37 2.40 0.004 0.70 6.99
of the synthetic magnetite to hematite, which is an antifer-
romagnetic mineral, and cooling affect the Fe recovery in
the subsequent magnetic separation, which must be pre-
vented. The typical conventional cooling process is water
quenching, in which the high-temperature roasted product
is cooled using cool water, and oxidation of synthetic mag-
netite is avoided in the absence of contact with air (Yu et
al., 2020). Although water quenching is an effective cooling
method for the magnetization roasting of iron ore, there
are many practical problems for its industrial application.
For example, it consumes large volumes of water resources,
which largely limits the industrial application of magne-
tization roasting, particularly in regions with water short-
ages. Moreover, there is heat waste during the heat transfer
with the cool water, which makes the temperature of the
ore pulp high in water quenching, which is not conducive
to the subsequent magnetic separation. Thus, developing
a new dry cooling method for magnetization roasting is
essential.
Air cooling is a typical dry cooling method in indus-
trial production, and it is characterized by a low produc-
tion cost and a simple process. However, its application
in magnetization roasting is very difficult. This is because
high-temperature Fe3O4 is readily oxidized to α-Fe2O3 in
an air atmosphere, whereas α-Fe2O3 is a typical weakly
magnetic mineral whose presence affects the Fe recov-
ery in the subsequent low-intensity magnetic separation.
Nevertheless, some studies indicate that Fe3O4 can be oxi-
dized to maghemite (γ-Fe2O3) under controlled oxidation
conditions (Uwadiale, 1992, Rzepa et al., 2016). Because
maghemite is ferrimagnetic, its presence does not affect the
Fe recovery in the magnetic separation step (Thompson,
2012). Wu et al. (2012) investigated the magnetization
roasting of goethite ore and found that Fe3O4 was mainly
oxidized to γ-Fe2O3 in the air-cooling process. However,
γ-Fe2O3 is metastable and readily oxidized to α-Fe2O3 in
the temperature range of 370–600 °C (Cornell and
Schwertmann, 2003). Therefore, the key to the air-cooling
process is to control the oxidation of Fe3O4 to γ-Fe2O3 and
thereby avoid the formation of α-Fe2O3.
Although air cooling is theoretically feasible for the
magnetising roasting of limonite ore, it has not been dem-
onstrated in practice. In this study, the comparison of
cooling processes such as air cooling, water quenching,
and nitrogen cooling was first investigated. The effects of
the different cooling processes on the magnetic separation
index were studied. Additionally, the mineral phase trans-
formation and magnetism variation in different cooling
processes were studied.
EXPERIMENTAL
Materials
The limonite ore used in the study was obtained from
Kunming, China. The sample had a particle size of 60 wt.%
particles less than 74 μm. Chemical composition analysis
of the sample are shown in Table 1. As shown in Table 1,
the iron content was 34.50%, and the FeO content was less
than 0.10%. In addition, the contents of SiO2, CaO, MgO,
and Al2O3 were 34.12%, 0.27%, 0.37%, and 2.40%,
respectively. Meanwhile, the content of harmful element P
was 0.70%. The loss on ignition (LOI) was 6.99%.
Experimental Method
The magnetising roasting and cooling experiments were
conducted using a fluidization roasting system, and the
experimental diagram is shown in Figure 1. A horizontal
furnace (MTI OTF-1200X-S-VT, China) was used to pro-
vide the reaction temperature range from 20 °C to 1200
°C. The reaction gas rate was controlled using a multichan-
nel mass flowmeter (MTI GSL-LCD, China). Magnetising
roasting was conducted in a quartz tube (Φ 50 mm).
In this experiment, first, synthetic magnetite was pre-
pared from limonite ore via magnetization roasting in a
mixed CO, H2 and N2 atmosphere. The roasting condi-
tions were a roasting temperature of 480 °C, CO and H2
concentration of 20%, and reduction time of 20 min. Once
the reduction roasting of limonite ore was completed, the
reduction gas of CO and H2 was stopped, and N2 was intro-
duced to protect the synthetic magnetite. The next stage
was the cooling experiments of high-temperature roasted
ore. In nitrogen cooling process, the roasting furnace was
turned off and the furnace tube was cooled by injecting N2
about 20 min. After cooling to 20 °C, the cooled samples
were obtained. In water quenching, the roasted ore was
directly poured into cooled water of 20 °C, and the sample
was obtained after drying in an oven with a temperature of
100 °C. In air cooling process, the roasted ore was directly
poured into air atmosphere, and the sample was obtained
when its temperature decreased to 20 °C. The roasted ore
was ground using a vertical stirring mill, and the particle
size was 75 wt.% passing 0.043 mm. The ground product
Table 1. Chemical composition analysis of limonite ore (mass, %)
Element Fe FeO SiO2 CaO MgO Al2O3 S P LOI
Content 34.50 0.10 34.12 0.27 0.37 2.40 0.004 0.70 6.99