XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3231
the partial oxidation of Sm2Co17 to Sm2O3 and Co has a
pO2,eq of 1.3×10–48 atm, and the complete oxidation of
Sm2Co17 to Sm2O3 and CoO has a pO2,eq of 1.9×10–20
atm (Barin 1995 Yuan 2011). Hence, the partial pressure
of oxygen must be kept between 1.9×10–20 and 1.3×10–48
atm to selectively oxidize Sm during the roasting process.
After the roasting, if the Sm2O3 and Co can liberated,
it should be possible to separate them using magnetic sepa-
ration with Co being ferromagnetic and Sm2O3 being para-
magnetic (Sucksmith 1938 Williams 1918). Additionally,
it is thought that, by quenching the oxidized magnets from
a high temperature, it may be possible to crack the brittle
Sm2O3 and ductile Co along their interfaces thereby caus-
ing liberation.
EXPERIMENTAL
Commercial SmCo magnets in the shape of 1.5 mm x
1.5 mm x 2.0 mm rectangular prisms were selectively oxi-
dized at 800 °C. To create the required atmosphere, the
bottom of an alumina boat was filled with an equimolar
mixture of Co and CoO powder. An alumina plate was
placed on top of the powder and magnets were placed on
this plate. The boat was then covered with a loose-fitting
alumina lid as shown in Figure 1. When the Co/CoO mix-
ture is heated, it reacts to establish the equilibrium pO2
for Co/CoO inside the boat which, as described previously,
will cause the Sm in the magnets to oxidize but not the
Co (Rhines 1940 Kupcis 1969). This whole assembly was
placed in a tube furnace which was then evacuated and
backfilled with titanium gettered argon three times before
heating in a static atmosphere up to the target tempera-
ture at a ramp rate of 3 °C/min. Once at temperature, the
magnets were held at 800 °C for 48 hours to oxidize the
Sm within the magnets, then titanium-gettered argon
was flowed through the tube furnace at a rate of 100 mL/
min while cooling back to room temperature at a rate of
3 °C/min.
After oxidizing, half of the magnets were set aside,
and the remainder were reheated to approximately 800 °C
under titanium gettered argon before being quenched in
water using the apparatus depicted in Figure 2. When the
magnets reached 800 °C as measured by a pyrometer, the
rubber stopper at the bottom of the tube was first removed,
then the external magnet was removed dropping the hot
magnets into a beaker of water below.
Magnets from each treatment were cross sectioned to
look for evidence of microcracking. The remaining magnets
from each treatment were pulverized using a ring and puck
for 2 minutes each. A wet sieve analysis was performed to
measure particle size distributions of the pulverized mag-
nets. Acetone was used instead of water for sieving since
agglomeration was observed when the pulverized magnets
were exposed to water. After grinding, the powders were
spread thinly over a glass sheet, then a ferrite magnet was
passed above the powder to separate the ferromagnetic
cobalt, from the paramagnetic Sm2O₃.
RESULTS
The selective oxidation step was partially successful. After
48 hours at 800 °C, the oxidation front had consumed a
Figure 1. Assembly used to create the required atmosphere for selectively oxidizing the SmCo magnets
Figure 2. Apparatus used to heat and quench the oxidized
magnets
the partial oxidation of Sm2Co17 to Sm2O3 and Co has a
pO2,eq of 1.3×10–48 atm, and the complete oxidation of
Sm2Co17 to Sm2O3 and CoO has a pO2,eq of 1.9×10–20
atm (Barin 1995 Yuan 2011). Hence, the partial pressure
of oxygen must be kept between 1.9×10–20 and 1.3×10–48
atm to selectively oxidize Sm during the roasting process.
After the roasting, if the Sm2O3 and Co can liberated,
it should be possible to separate them using magnetic sepa-
ration with Co being ferromagnetic and Sm2O3 being para-
magnetic (Sucksmith 1938 Williams 1918). Additionally,
it is thought that, by quenching the oxidized magnets from
a high temperature, it may be possible to crack the brittle
Sm2O3 and ductile Co along their interfaces thereby caus-
ing liberation.
EXPERIMENTAL
Commercial SmCo magnets in the shape of 1.5 mm x
1.5 mm x 2.0 mm rectangular prisms were selectively oxi-
dized at 800 °C. To create the required atmosphere, the
bottom of an alumina boat was filled with an equimolar
mixture of Co and CoO powder. An alumina plate was
placed on top of the powder and magnets were placed on
this plate. The boat was then covered with a loose-fitting
alumina lid as shown in Figure 1. When the Co/CoO mix-
ture is heated, it reacts to establish the equilibrium pO2
for Co/CoO inside the boat which, as described previously,
will cause the Sm in the magnets to oxidize but not the
Co (Rhines 1940 Kupcis 1969). This whole assembly was
placed in a tube furnace which was then evacuated and
backfilled with titanium gettered argon three times before
heating in a static atmosphere up to the target tempera-
ture at a ramp rate of 3 °C/min. Once at temperature, the
magnets were held at 800 °C for 48 hours to oxidize the
Sm within the magnets, then titanium-gettered argon
was flowed through the tube furnace at a rate of 100 mL/
min while cooling back to room temperature at a rate of
3 °C/min.
After oxidizing, half of the magnets were set aside,
and the remainder were reheated to approximately 800 °C
under titanium gettered argon before being quenched in
water using the apparatus depicted in Figure 2. When the
magnets reached 800 °C as measured by a pyrometer, the
rubber stopper at the bottom of the tube was first removed,
then the external magnet was removed dropping the hot
magnets into a beaker of water below.
Magnets from each treatment were cross sectioned to
look for evidence of microcracking. The remaining magnets
from each treatment were pulverized using a ring and puck
for 2 minutes each. A wet sieve analysis was performed to
measure particle size distributions of the pulverized mag-
nets. Acetone was used instead of water for sieving since
agglomeration was observed when the pulverized magnets
were exposed to water. After grinding, the powders were
spread thinly over a glass sheet, then a ferrite magnet was
passed above the powder to separate the ferromagnetic
cobalt, from the paramagnetic Sm2O₃.
RESULTS
The selective oxidation step was partially successful. After
48 hours at 800 °C, the oxidation front had consumed a
Figure 1. Assembly used to create the required atmosphere for selectively oxidizing the SmCo magnets
Figure 2. Apparatus used to heat and quench the oxidized
magnets