REPOSITORY > RESULTS

Abstract

The rare-earth-based permanent magnets (REPM) are vital for modern electrical appliances and green energy technologies which has created a massive supply-demand frontier for the industries. Usually high thermal stability and resistance to demagnetization comes with the addition of heavy rare earth elements (HREEs) like Dy and Tb in Nd2Fe14B-type permanent magnets, which has the highest energy products (BHmax) per unit volume. The addition of HREEs to Nd-Fe-B permanent magnets has strong implications on natural abundance, costs as well as the reduction in magnetization (and BHmax) due to the antiferromagnetic coupling of the magnetic moments of the HREE atoms with that of iron. Since the 2011 rare earth (RE) crisis, which caused the prices of raw material (especially Nd, Dy, Tb) to surge by 700%; the global REPM producers are looking for supply chain and economic sustainability to fulfill this commercial demand. On the contrary, the scientific front has been looking to explore a new class of permanent magnets without the costly rare earth elements or try to develop the methods to recycle the end-of-life (EOL) REPMs. The former has not been very successful since the past decade or so, whereby the later route has shown high potential in commercial up-scaling if the processing parameters are controlled to preserve and control the microstructure. Direct reusage or direct/indirect recycling of EOL magnets is an area of interest for the researchers as well as the magnet producers. Slight modification of chemical composition towards rare earth rich compositions has helped developed REPMs with better magnetic properties than the starting EOLs, but the higher costs due to the addition of rare earth rich compositions to the existing systems may be a problem.
The hydrogen-based recycling route of REPM is an economically feasible and clean method due to its very low environmental footprint compared to pyro- and hydrometallurgical recycling methods. Normally the rejected waste in REPM production goes up to 25%, which is huge considering the aggregate global magnet production. This rejected scrap or EOL waste magnets can be processed by hydrogen-based recycling methods like: hydrogen decrepitation (HD) used for Nd-Fe-B and Sm-Co or hydrogenation–disproportionation–desorption–recombination (HDDR) used for Nd-Fe-B to develop magnetic powder can either be sintered by conventional powder metallurgical routes or readymade to assorted shapes in the form of plastic bonded magnets. The HD treatment facilitates the intergranular cracking in EOL Nd-Fe-B magnets due to formation of NdHX (causing volume expansion) and this friable non-magnetic powder can be further jet-milled to 3 – 5 µm size, magnetically aligned and sintered to final shape. The chemical composition can be easily modified by blending fine RE-rich alloy powder/dopant (DyF3) with the HD-treated recycled powder to achieve the desirable level of magnetic properties.
Identically the HDDR reprocessing has the additional benefit since it can develop anisotropic nanocrystalline grains in the size range of ~400 nm for achieving higher coercivities than the microcrystalline counterparts. The HDDR-treated recycled powders can be formed to shape by mixing them with the polymeric binders or sintered by rapid consolidation techniques like hot deformation or spark plasma sintering (SPS) to bulk magnets. Normally these nanocrystalline powders cannot be conventionally sintered, as higher temperatures for prolonged periods would result in exaggerated grain coarsening and the coercivity will decline severely. Therefore, rapid compaction by SPS is an alternate method in magnet recycling to achieve the magnetic properties at par with the commercial grade or fresh HD/HDDR powder. The SPS process is very rapid and heating rates can be controlled to prevent the deterioration to the microstructure, which in turn preserves the magnetic properties of the original powder and fully dense magnets can be formed in a matter of minutes at lower sintering temperatures than those in conventional sintering. Simultaneous application of pressure in the SPS is beneficial for achieving full densification and ramping it rapidly (> 100 MPa) and is beneficial in the texture in the Nd-Fe-B nanocrystalline material by hot forging (deformation).
Within this thesis two different kinds of recycled magnetic powders HD-treated SmCo5 and HDDR Nd-Fe-B, were investigated by consolidation with the SPS. In the former case the SPS temperature range was identified around 900 – 950 °C with holding time of 1 – 5 minutes and 100 MPa pressure in 0.2 mbar Ar pressure for achieving > 97 % densification. The resultant magnetic properties of isotropic SPS-ed SmCo5 magnets (HCi > 2000 kA/m & Br = 0.47 T) were slightly better than the vacuum-sintered magnets due to better microstructural control (optimal distribution of Sm-rich Sm2Co7 phase in the SmCo5 matrix) achieved with the SPS. As compared to the commercial grade material, the finer HD-treated recycled powder had a tighter grain size control (4 – 20 µm), which resulted in nearly full density and room temperature HCi > 2000 kA/m and better thermal stability at 180 °C.
Similarly, the recycled HDDR Nd-Fe-B powder was optimally SPS-ed at 750 – 800 °C (approx. 30 % lower sintering temperature) for 1 minute and 100 MPa to reach HCi ~ 1200 kA/m which was slightly better than the fresh HDDR powder MF-15P (1180 kA/m). The starting HDDR powder with Nd13.4Dy0.6Fe78.6B6.1Nb0.4Al0.7 composition had the HCi = 830 kA/m. The effect of post SPS thermal treatments was thoroughly investigated for the recycled HDDR powder and with HR-SEM and EDXS it was confirmed the redistribution of Nd-rich grain boundary phase is important for achieving higher coercivities after the SPS fabrication. These optimally SPS-ed magnets were hot deformed at 750 °C by varying the hot deformation pressure to 100 and 150 MPa in WC + Co cermet dies to induce texture in the isotropic bulk magnets. As compared to the starting HDDR powder with Br = 0.9 T, the resultant improvement after 150 MPa hot deformation in Br was ~ 1.01 T (BHmax = 180 kJ/m3) with 55% height reduction of the sample.
It is important to consider that hydrogen-based recycling is effective only if the protective coating on EOL magnets is non present or ruptured for decrepitation to occur. In secondary recycling, these coatings are removed from the fragmented powder with an average particle size less than 100 µm by mechanically crushing the magnets. On the contrary, the HDDR powder is prone to excessive oxidation if the reduction in average particle size is too significant. But this critical particle size range has not been identified previously, which also restrained the commercial scale up. The in-depth studies were performed to verify the relationship of HDDR powder particle size with the oxygen content, sinterability and the magnetic properties. The particle fractions from 700 µm to < 50 µm were individually tested for the oxygen content, the magnetic properties and the sinterability. It was evaluated that particles sized below 100 µm took up twice the oxygen (> 7000 ppm) and as a result, the overall Nd-rich phase declined due to excessive Nd2O3 formation. This resulted in poor sinterability and severe degradation of the magnetic properties (HCi < 200 kA/m). Therefore, it was suggested to limit sizing down the HDDR powder particles not less than the 100 µm range for achieving optimal magnetic properties.
Lastly the improvement in HCi of the recycled HDDR Nd-Fe-B powder was realized by blending it with DyF3 in varying weight fractions from 1 – 5 %. The blends were compacted at 750 °C for 1 minute and 100 MPa. These magnets were subjected to vacuum thermal treatments at 600 – 900 °C for 1 – 6 hours. The optimal thermal treatment condition: 750 °C for 1 hour for 2 wt. % DyF3 resulted in HCi ~ 1410 kA/m which is approx. 70% higher than the starting HDDR powder. This doping technique developed uniform a (Dy,Nd)2Fe14B core-shell structure throughout the sample due to localized Dy diffusion and solution reprecipitation upon solidification below the ternary eutectic temperature 665 °C. Excessive DyF3 content ≥ 5 wt. %, prolonged thermal treatment (≥ 3 hours) and higher temperatures (900 °C) deteriorated the magnetic properties due to the decomposition of core-shell structure and formation of complex Nd-Fe-F-type intermetallic phases besides the excess of unreacted, non-ferromagnetic DyF4 & NdF4 phases.