XXXI International Mineral Processing Congress 2024 Proceedings/Washington, DC/Sep 29–Oct 3 3233
magnets were coarser than the slow cooled magnets after
grinding with a d80 of 99 μm compared to 79 μm. This
difference is thought to be due to how the treated magnets
were loaded into the ring and puck and not a result of the
different thermal treatments. It is however evidence that
the quenching did not cause substantial disintegration of
the magnets.
Magnetic separation was not successful for separating
Co and SmFeO3 in the oxidized magnets. Passing a magnet
over the pulverized, oxidized magnets recovered all of the
powder. Examining the microstructure of a slow cooled,
oxidized magnet in Figure 7, it is evident why the separa-
tion was unsuccessful. The oxide particles are at most a few
micrometers across while most of the pulverized magnet
particles were tens of micrometers in diameter. Liberation
was not achieved, so magnetic separation was not successful.
Researchers at the United States Bureau of Mines encoun-
tered a similar issue when doing a comparable selective oxi-
dation process with NdFeB magnets (Lyman and Palmer
1993). If the oxide and metal particles can be liberated by
Figure 5. Cross section of a quenched, oxidized magnet
showing the oxidized layer intact
Figure 6. Particle sized distribution of oxidized, pulverized magnets
Figure 7. Cross section of a slow cooled, oxidized magnet
showing the microstructure of the oxides (dark grey), Co
(light grey, upper right), and unoxidized magnet (light grey,
lower right)
magnets were coarser than the slow cooled magnets after
grinding with a d80 of 99 μm compared to 79 μm. This
difference is thought to be due to how the treated magnets
were loaded into the ring and puck and not a result of the
different thermal treatments. It is however evidence that
the quenching did not cause substantial disintegration of
the magnets.
Magnetic separation was not successful for separating
Co and SmFeO3 in the oxidized magnets. Passing a magnet
over the pulverized, oxidized magnets recovered all of the
powder. Examining the microstructure of a slow cooled,
oxidized magnet in Figure 7, it is evident why the separa-
tion was unsuccessful. The oxide particles are at most a few
micrometers across while most of the pulverized magnet
particles were tens of micrometers in diameter. Liberation
was not achieved, so magnetic separation was not successful.
Researchers at the United States Bureau of Mines encoun-
tered a similar issue when doing a comparable selective oxi-
dation process with NdFeB magnets (Lyman and Palmer
1993). If the oxide and metal particles can be liberated by
Figure 5. Cross section of a quenched, oxidized magnet
showing the oxidized layer intact
Figure 6. Particle sized distribution of oxidized, pulverized magnets
Figure 7. Cross section of a slow cooled, oxidized magnet
showing the microstructure of the oxides (dark grey), Co
(light grey, upper right), and unoxidized magnet (light grey,
lower right)