Archive: Jun 2016

Rare-earth reduction

Comments Off on Rare-earth reduction

New methods leading to the reduction of heavy rare-earth elements in NdFeB permanent magnets with enhanced coercive force are being implemented by magnet manufacturers

The advent of rare earth-transition metal alloys, or inter-metallic compounds, such as samarium cobalt and a ternary system composed of neodymium iron and boron (NdFeB) permanent magnets revolutionized the usage of permanent magnets. The latter alloys (Figure 1) with their extremely high coercive forces and substantially higher energy products (as compared with their predecessor systems, alnico and ceramic) not only simplified the design of magnetic circuits, but made possible the miniaturization of a vast array of magnetic devices. Nevertheless, the lower Curie temperature of this alloy has made it prone to loss of coercive force (Hci) at higher operating temperatures.
There has been a constant improvement in coercivity of NdFeB magnets since their inception. Figure 2 shows the latter improvement. Increase in demand for NdFeB magnets coupled with a relative shortage of heavy rare earths (HRE) such as dysprosium (Dy) and terbium (Tb) has persuaded magnet manufacturers to price the alloy, among other factors, as a function of the content of HRE, causing fear, uncertainty and skepticism among magnetic device designers and manufacturers. This was the genesis of the decision by magnet manufacturers to launch an orchestrated attempt to reduce the content of HREs or partially replace the latter rare earths,
with more abundant and less used HREs such as holmium and others.

Using heavy rare earths

Light rare earths couple ferromagnetically with transition metals, resulting in higher saturation magnetization. NdFeB, with its high saturation magnetization and extremely high-energy product, does have a drawback. That is, the coercive force of this alloy system suffers exponentially as the operating temperature of the magnet increases.

Figure 1: The coercive forces of rare earth-transition metal alloys and inter-metallic compounds are increasing faster than previously

To remedy this situation selectively, a portion of Nd (light rare earth or LRE) is replaced by HRE (Figure3). Although HREs coupling with transition metals ferromagnetically produces a somewhat lower saturation magnetization, it does increase the anisotropy field of the alloy and therefore increases the intrinsic coercive force (Hci) of the magnet. The degree of enhancement of Hci is a function of the content of HRE. The higher the content of HRE, the higher the coercive force will be. However, there is an upper limit beyond which the coercive force starts to show a gradual decline.

Improved melting technology

Strip casting technology was a much more powerful tool affecting the efficiency and quality of NdFeB magnets than any other changes since the inception of manufacturing of sintered Nd magnets. It is prudent to observe that the
transition from the classical book mold and cup cake (ingot melting) process to strip casting hasimproved the efficiency of usage of HREs (which are associated with higher vapor pressure and excessive loss of HREs in the melting process) to a great extent.

Figure 2: The steady improvement in coercive force of NdFeB magnets since their inception

Strip casting is another term for rapid quenching process or centrifugal casting. This process produces thin strips of alloy (as opposed to that produced by ingot casting) by dropping the molten alloy on a single or twin rollers rotating at high speed, causing the molten alloy to solidify to a thickness of 0.03mm to perhaps as thick as 4-5mm. The strips used for production have been in a narrow range of thicker than 0.1mm and thinner than 0.3mm. Strips of solidified alloy exhibit a columnar crystal that grows from the surface of the roller in the direction of thickness of the strip.

The remarkable improvement in the microstructure of the rapidly solidified NdFeB strips (columnar structure of hard magnetic grain) greatly enhances the homogeneity and the unit properties of the magnet produced from the strip cast material.

Strip cast of NdFeB type composition, containing niobium (Nb) or molybdenum (Mo) in the range of one atomic weight percent, has considerably improved the coercive force of this alloy without a substantial loss in the in (Br).

Without the advent of rapidly solidified strip casting of NdFe magnets, the current degree of improvement in the use of HREs would have been virtually impossible. Figures 4 and 5 show the effect of Nb on Hci and Br. Figure 6 shows the grinding resistance for the sintered NdFeB magnet as a function of the content of Nb – the higher level of Nb significantly increases the resistance to grinding. Therefore, there is a limit as to how much Nb can be added to the alloy.

Figure 3: A portion of Nd (light rare earth or LRE) replaced by HRE to counter the detrimental effects of increased operating temperatures

Figure 4: The effect of niobium on the intrinsic coercive force and residual induction of NdFeB magnets using strip casting
and ingot casting technology

Dysprosium grain boundary diffusion

One successful method of enhancing the coercive force of NdFeB magnets is to diffuse dysprosium into the grain boundary of an alloy made of Nd, Fe, B (no heavy rare-earth elements) either in the form of strip-cast or finished magnets.
The crystal anisotropy of the magnetic grains surface region is always lower than that of the inside of the grains, as this region suffers defects and is under stress.
The principle of grain boundary diffusion is based on the premise that when a reverse magnetic field is applied to a magnet, magnetization reversal happens to all particles making up the mass of the magnet. The latter reversal starts from those particles that nestle on the surface of the particles. If the concentration of dysprosium (high anisotropy heavy RE) is increased on the surface, the magnetization reversal becomes difficult in the particles, thus increasing the coercive force of the magnet. Since the concentration of dysprosium decreases in the central region of the particles, the loss of saturation magnetization, and thus the drop in the Br of the magnet, will not be significant. Although there are no available guidelines as to how much dysprosium can be saved by using grain boundary diffusion, generally this process saves as much as 50% dysprosium used in a typical melt for a given value of coercive force, with very little loss in the value of Br of the magnet. Figure 7 shows the effect of Dy grain boundary diffusion in the strip- cast stage of manufacturing.

Figure 5: The effect of niobium on the intrinsic coercive force and residual induction of NdFeB magnets using strip casting

Figure 6: The grinding resistance for the sintered NdFeB magnet as a function of the content of Nb

Figure 7: The effect of the Dy grain boundary diffusion concept in the strip-cast stage of manufacturing

The concept of grain boundary diffusion started with finished uncoated magnets. In light of the fact that the process consists of a secondary operation, at temperatures exceeding 700ºC, the process was not deemed appropriate for magnets with less favorable aspect ratios (the ratio of the magnet thickness to the magnet surface area). Thinner magnets with a large surface area and small thickness-unfavorable aspect ratio did suffer enough deformation and warping to be classified as rejects. So the process would lend itself only to a select group of geometries. This process may be ideal for applications where the magnet is small and not prone to deformation on account of high temperature exposure.

The alternative method of grain diffusion with which the magnet manufacturing industry has been successful is the Dy grain boundary diffusion of cast strips, and the ultimate manufacturing of the magnets by using the conventional powder metal process.

Nanocomposite NdFeB magnets

Quantum size effects have spawned much interest among magnet scientists and technologists. As the size of particles in a nanocomposite structure decreases, there is an increase in the fraction of surface energy (surface atoms), which could lead to a considerably higher energy product in the permanent magnet.

The magnetic properties of magnets made from nanoparticles, among other factors, is a function of the chemical composition, the microstructure, the size and the shape of the particle, the crystal lattice and the morphology of particles. Controlling the latter parameters within a given range in the process of synthesis of nanomaterial will be, if not impossible, very difficult. Therefore, the properties of nanomaterial of the same type will be significantly different. The majority of the current high-energy magnetic alloys are of complex composition. As a result, this causes the range of parameters in the synthesis of nanomaterial to be extremely narrow.

The various processes of synthesis of magnetic nanomaterial also go completely against the principle of maintaining the proper stoichiometric ratio that is necessary in order to achieve sound magnetic properties.

For example, if the stoichiometric ratio of Nd2Fe14B or SmCo5 alloys changes during the synthesis, the resultant phases may be either slightly magnetic or non-magnetic; thus it significantly reduces the unit magnetic properties of the stated alloys.
The synthesis of such nanocomposites may require a manufacturing method that is entirely different from the existing powder metal method of manufacturing. The cost of the magnet is a function of the new manufacturing technology. If the latter cost is too high, the product may not be deemed viable.
The manufacturing of nanocomposite of NdFeB was orginally proposed by a Japanese firm as early as the late 1990s. The firm, in cooperation with a major university in the USA and one in Japan, actually managed to produced a very high- saturation and low-Hci magnet using a unique method of synthesis. Figure 8 shows the second quadrant demagnetization curve of a magnet using a non-powder metal process, a nanocomposite technology. The magnet exhibits sound magnetic properties. Nevertheless, that was the last time it was heard of. The company may have decided to abandon the project for reasons unknown, or it may have become part of a larger firm.

In summary, although the concept of nanocomposites may be a viable method of manufacturing NdFeB magnets, the cost and method of producing high-energy NdFeB magnets may be questionable.

Figure 8: The second quadrant demagnetization curve of a magnet using a non-powder metal process

Temperature control methods Magnetic circuit designers and final device engineers have used their imaginations to design mechanisms that control the working temperature of the device, making the magnet less prone to high-temperature exposure. This supplementary device may take the form of a cooling device external to the magnet or any other temperature- control mechanism that will essentially perform the latter task.
There are many existing and new methods
being developed where magnet manufacturers are reducing the use of HRE elements. There is also
an intensive effort being launched to eliminate the use of the latter family of elements altogether. n


1) Yukata Matsuura: Current Status of NdFeB Magnets In
Japan- China Magnetics Shanghai, September 2006, China
Magnetics Conference
2) Karl J Strnat, Herb Mildrum,T.K.Tan –Sintered NdFeB magnets Modified with Holmium and Cobalt- “The Tenth International Workshop pm Rare Earth Magnets May 1989
3) Private dialogs with Matsuura in reference to Grain Boundary
Diffusion of Dysprosium
4) Naoko Ono, Massato Sagawa, R.Kasada” Production of High Performance Magnets by Grain Boundary Diffusion, Journal of Magnetics and Magnetism, February 2011
5) Hongwang Zhang,Shen Peng, Shuan-bing Rong,J.Ping Liu, Ying
Zhang,M.J.Kramer,Journal ,of Material Chemistry June 6th
6) Magnetic Nanocomposite Materials for High Temperature Applications- Frank Johnson, Amy Hsio,ColinAshe, David Laughlin, David Lambeth, Michael McHenry- Carnegie Mellon University and Lajors K. Varga ,Hungarian Academy of Sciences, Budapest, Hungary
7) S P Gubin, Yu A Koksharov,G.B.Khomutov,G Yu Yurkov, Russian
Chemical Reviews 2005
8) Shuk Rashidi, Use and Reduction and Probable Remedies of Heavy Rare Earths in Proceedings of the International Workshop on Technology and Economics of Rare Earths and Metals – 2011
9) Shuk Rashidi, Global Magnet Market: NdFeB Price Escalation and Revised Manufacturing Paradigm- China Magnetics Conference 2007 Beijing China
10) Optimization of the Geometry of Strip Cast NdFeB Alloy – Beijing Zhone Ke San Huan High Tech Ltd- private communication