Extending the Limits of the Sm2Co17 System

By Steve Constantinides  | Arnold Magnetic Technologies

The 2010-2011 rare earth crisis appears over. Prices are low and adequate raw material is available. Many of the efforts initiated to find a better non-rare earth permanent magnet have ended in disappointment. Programs such as Fe-nitride, core-shell exchange coupled magnets and enhanced alnico are continuing but with limited success to-date and significant hurdles remaining. It is in this framework that revisiting one of the more significant magnet materials seems a reasonable proposition. What material, you ask? Well, the strongest permanent magnet material at and above 150°C, the one with excellent corrosion resistance and superior temperature stability – Sm2Co17.

The theoretical limit of energy product for the Sm2Co17 system has been previously calculated at 34 MGOe (270 kJ/m3). Material near this limit has been produced commercially for several years and with the recent introduction of Recoma 35E is right at that limit. In order for further improvements to take place, both composition and process challenges must be overcome to deliver remanence in excess of 1.2T while still delivering intrinsic coercivity high enough to be of commercial interest. Developing alloys with greater magnetic remanence, while feasible, requires stringent control throughout manufacturing.

Arnold_Fig1Justification for working on improved SmCo can be explained by Figure 1 which shows that NdFeB experiences large reversible reduction in magnetization as a function of temperature and loses magnetization irreversibly above a critical temperature that varies by grade (dependent upon heavy rare earth and cobalt content). On the other hand, Sm2Co17 standard grades can be used to 350°C; high temperature grades can be used to over 500°C; and high energy product grades can be used above 250°C.

Arnold Magnetic Technologies has manufactured SmCo magnets since 1972. It is only natural that we should exert efforts to improve its performance, especially now with a lack of a significant new commercial magnet material and recognition of the limitations of competing magnetic products. These recent efforts have resulted in the introduction of Recoma 33E in 2013 and 35E in 2016.

Making Better SmCo
Several issues are important to improving the performance of magnets in general and SmCo specifically: contamination, particle alignment during compaction, elemental substitution, starting alloy condition and microstructure development, and thermal processing.

It is difficult to totally eliminate contaminants and they preferentially bind with Sm or Zr reducing the magnetization. For example, one weight percent oxygen binds with 6.2 weight percent samarium. To compensate for the samarium bound with oxygen requires very careful control of samarium and oxygen in the raw materials and minimization of oxygen pick-up during processing.

Small amounts of excess carbon reduce sintering temperature and improve demagnetization hysteresis loop shape[2]. However, controlling the exact amount present in the alloy system is difficult. Further, carbon binds preferentially with zirconium to form ZrC. The binding ratio is 7.6 weight percent zirconium per percent carbon. Carbon is present in small percentages in the raw materials, but additional amounts can be added via organic processing lubricants during milling or pressing and from organic vapors in vacuum sintering.

Maximum BR and energy product require that all grains are aligned in one direction. This is achieved during manufacture by reduction of the alloy to a particle size wherein each particle contains only one crystal. In Sm2Co17, this particle size is approximately 7 to 10 microns (laser analysis).[13] Prior to and during compaction of the alloy particles, domain orientation is impressed via an external magnetic field. Orienting torque is applied due to the reaction of the magnetized particle’s field with the applied orienting field. Compaction creates misalignment forces on the particles. The three methods of compaction are: Axial (parallel), Transverse (perpendicular), and Isostatic. Isostatic compression of the powder results in minimum misalignment of the grains. Transverse pressing is almost as effective. Axial/parallel pressing results in the most misalignment during compaction. Methods of measuring “perfection” of alignment include examination using magneto-optical Kerr effect and Schulz XRD crystallographic determination.[6,10,18,19]

Elemental Substitution
Optimal magnetic properties of Sm2Co17 have been achieved using a quinary alloy represented by the formula Sm(Coa Feb Cuc Zrd)z where “z” is somewhat less than 8.5 (=17/2). For example, a typical formula as reported by de Campos et al is Sm(Co0.66Fe0.2Cu0.1Zr0.04)8.[5] Iron has been substituted for cobalt at 5 to 20 weight percent. A 5 weight percent grade produces excellent high temperature capability while a 20 weight percent material maximizes magnetic strength. If 20 weight percent is good, would 25 weight percent produce even higher magnetic strength? What is the practical limit of iron substitution? According to Strnat, a 50:50 blend of Co:Fe could result in an energy product of 60 MGOe (475 kJ/m3).[4]

Thermal processing of the SmCo alloy becomes more difficult as the iron content increases. More importantly, intrinsic coercivity decreases rapidly as iron content rises above 20 weight percent although Liu and Ray show HcJ rising again above b=0.30 (mole ratio).[1] Complete substitution of iron for cobalt in Sm2Fe17 does not work as it generates planar rather than uniaxial anisotropy.

As iron is increased, the familiar phase diagram for SmCo changes and copper and zirconium must also be adjusted to maintain the composition “in balance”. Morita shows us the change to the phase diagram resulting from changes of Fe and Cu wherein the solution treatment region with hexagonal SmCo stable at higher temperatures and rhombohedral stable at the lower temperatures narrows and stretches downward.[3,11,12]

Other rare earth elements (REEs) can be substituted for Sm. Strnat teaches us that Pr, Nd and Y in 2:17 compounds should produce higher magnetization than Sm.[4] In the 1:5 SmCo alloys Pr has been substituted for up to about half of the samarium with a concomitant increase in energy product.

Arnold_Fig2During the development of Sm2(Co,Fe,Cu,Zr)17 numerous refractory elements such as V, Ti, Hf, Mo, and Zr were evaluated. In trials, zirconium produced the greatest increase in coercivity with the least reduction in magnetization.

Starting Alloy Condition and Microstructure Development
At least three conditions of the alloy have been investigated relevant to improving magnetic performance: grain size, alloy homogeneity, and alloy crystal structure as the result of the casting process. Earliest efforts were targeted at solution treating the cast alloy to ensure an homogenous phase structure which would fracture and mill into single crystal particles. SmCo alloy cast and cooled under controlled rates, either bulk or strip cast, has been found to yield suitable microstructure for subsequent milling and alignment during compaction obviating the additional solutionizing step.

Milling of the cast alloy to fine particle size has been accomplished via ball (or attritor) milling and via fluid bed jet milling. Ball milling efficiency is improved in the presence of liquid carriers and surfactants. These organic materials introduce a residual amount of carbon which can, if not adequately removed or controlled, produce variable sintering effects plus the dilution effect of the ZrC non-magnetic phase. Further, particle size distribution in ball milling is broader than that obtained via jet milling since the process does not provide for elimination of the super fines or the less-milled larger particles. Super-fines oxidize more easily while larger particles may not be single crystal.

Fluid bed jet milling has a cut-off for large particles controlled by a variable speed rotating “squirrel-cage” at the gas/particle exit. Small particle content is eliminated using one or more stages of cyclone separation upon exhaust from the mill. The result is a narrower particle size distribution with minimal contamination from organics. Milled powders are sensitive to oxidation.

Arnold_Fig3The final structure of SmCo 2:17 (Figure 3) has been extensively reported. Ray provides a clear explanation for the phase development.[8] Although additional phases have been proposed/reported, the basic structure contains the 2:17 rhombohedral (main) phase, a hexagonal Sm(Co,Cu)5 cell boundary phase and the Zr-rich lamella structure. Coercivity and “loop squareness” is developed as the result of both the Sm(Co,Cu)5 grain boundary and the Zr-rich lamella phases.

Thermal Process
Standard powder metallurgical processes are used for the commercial manufacture of SmCo magnets. Once the powder has been aligned and compacted, it must be densified via liquid phase sintering. This is accomplished in either a vacuum or inert atmosphere furnace. Large scale production coupled with the operational problems of large tube furnaces and increasing expense of helium dictate the use of vacuum furnaces to accomplish densification and phase development.

Samarium has a relatively high vapor pressure such that at and near sintering temperature it volatizes from the magnet surface. Sintering is therefore a compromise of maximum densification in a vacuum (to avoid trapped pores) coupled with partial pressure inert gas (argon) to suppress volatization. Each manufacturer has, no doubt, optimized around a suitable combination of time/temperature/partial pressure.

Subsequent to densification, the crystal structure requires homogenization (solution treatment). Phase diagrams have been generated for several composition variants and these show the change in shape of stable-phase regions. [3,11,12]

Tempering is achieved in three stages: an extended hold at or near 850°C; a ramp from 850°C to 400°C over an extended period and can be accomplished as one continuous ramp or as a stepped process; hold time ~400°C to achieve optimum hysteresis loop shape.[16]

The spectrum of magnetic materials has seen few new industrially interesting entrants since the advent of NdFeB. Since its introduction followed so soon after the commercialization of SmCo, there remains more opportunity to find optimization and improvements in the SmCo system than in other, more thoroughly researched materials. The chemistry of SmCo suggests several paths to increasing the energy density. However, many of the changes made to increase energy product also make processing more difficult. Modern equipment and a thoughtful examination of every step of the process has resulted in a strategy that can reliably deliver commercial materials with magnetic performance superior to previous limits. Combined with the natural high-temperature capability of SmCo, this creates compelling new products for the market.

Note: This has been adapted from a more extended technical paper presented at WMM’16 in Rome, Italy, written by Steve Constantinides, Dave Maybury, Urs Wyss and Gerhard Martinek and presented by Dave Maybury.

For more information visit www.arnoldmagnetics.com.

The following references were down-selected from a universe of more than 440 documents relating to SmCo.

  1. 2-17 type RE-TM magnets with improved magnetic properties; Liu and Ray; REPM 10; 1989
  2. Behaviour of residual carbon in Sm(CoFeZr)z permanent magnets; Tian et al; Journal of Alloys and Compounds 440 (2007) 89-93
  3. Effect of the variation in Sm/Cu/Zr content on phase stability of an Sm(Co,Fe,Cu,Zr)7.4 permanent magnet alloy; Chin et al; IEEE Transactions on Magnetics, Vol. 25, No. 5, September 1989
  4. Hard Magnetic Properties of Rare Earth-Transition Metal Alloys; Strnat; IEEE Transactions on Magnetics, Vol. 8, No. 3, September 1972
  5. Impurity phases in Sm(CoFeCuZr)z magnets: The role of Zr; de Campos et al; Journal of Alloys and Compounds 403 (2005) 329-334
  6. Metallographic method for the determination of crystal alignment in Co-R permanent magnets; Martin; AIP Conf. Proc. 29, 614 (1976)
  7. Metallurgical and magnetic properties of iron-enriched Ce-Co-Fe-Cu-Zr magnets; Tawara et al; REPM 1987
  8. Metallurgical behavior of Sm(Co,Fe,Cu,Zr)z alloys; Ray; J. Appl. Phys. 55 (6), March 1984
  9. Microstructural characterization of Sm-Co magnets; Victoria et al; IMECE2014-37106; 2014
  10. Orientation and Magnetization of SmCo5 Magnets; Swift et al; AIP Conf. Proc. 29, 612 (1976)
  11. Phase diagrams for Sm2Co17 magnets; Morita et al; MRS Int’l. Mtg. on Adv. Mats. Vol. 11, 1989
  12. Phase Transformation at high Temperature and Coercivity of Sm(Co,Cu,Fe,Zr)7-9 magnet alloys; Morita et al; IEEE Transactions on Magnetics, Vol. MAG-23, No. 5, September 1987
  13. Rare Earth-Cobalt Permanent Magnets; Strnat; Ferromagnetic Materials, Vol. 4, Edited by E.P. Wohlfarth and K.H.J. Buschow, Elsevier Science Publishers; 1988
  14. Rare earth-cobalt permanent magnets near the 2-17 composition; Tawara and Strnat; IEEE Transactions on Magnetics, Vol. MAG-12, No. 6, November 1976
  15. Rare earth magnet alloy; Aichi Steel; US Patent 5017247; 1991
  16. Rare earth permanent magnets Sm2(Co,Fe,Cu,Zr)17 for high temperature applications; Peng et al; Journal of Rare Earths, Vol. 26, No. 3, June 2008
  17. Rare-earth Intermetallics for Permanent Magnet Applications; Fidler et al; Handbook of Magnetism and Advanced Magnetic Materials, Volume 4 Novel Materials, John Wiley & Sons, Ltd.; 2007; ISBN: 978-0-470-02217-7
  18. Statistical analysis of the orientation of sintered SmCo5 magnets; Trout and Graham; IEEE Transaction on Magnetic, Vol. MAG-12, No. 6, November 1976
  19. Study of internal particle alignment on rare earth cobalt magnets; Searle et al; REPM 1981