Effect of spin-reorientation transition of cell boundary phases on the temperature dependence of magnetization and coercivity in Sm2Co17 magnets

Si-Si Tu(涂思思), Lei Liu(劉雷), Bo Zhou(周波), Chuang-Hui Dong(董創(chuàng)輝),Li-Ming Ye(葉力銘), Ying-Li Sun(孫穎莉), Yong Ding(丁勇),?,A-Ru Yan(閆阿儒), and Xin-Biao Mao(毛信表)

1College of Chemical Engineering,Zhejiang University of Technology,Hangzhou 310014,China

2CISRI and NIMTE Joint Innovation Center for Rare Earth Permanent Magnets,CAS Key Laboratory of Magnetic Materials and Devices,Ningbo Institute of Materials Technology and Engineering,Chinese Academy of Sciences,Ningbo 315201,China

Keywords: spin-reorientation transformation,low-temperature coefficient,Sm2Co17 magnets

1.Introduction

More accurate and light-weight low-cost magnetic and inertial measurement units are needed for the technological development of gyroscopes, accelerometers, sensors and traveling-wave tubes.[1,2]At the same time,more stringent requirements have been placed on the temperature stability of magnets.Sm2Co17magnets are the best candidates due to their high Curie temperature, good magnetic properties and excellent thermal stability.In general, the higher the coercivity of the magnet, the better the demagnetization resistance and the magnet’s stability.However, recent studies[3,4]have shown that the irreversible and reversible flux losses of magnets are affected by their remanence temperature coefficient (α), and the magnetic drift caused by the change in the anti-demagnetization capacity of magnets is affected by the temperature coefficient of the coercivity (β).As a result, to meet high-precision measurement requirements, theαandβvalues of magnets should be low.Researchers[5-8]have attempted to achieve a lowαvalue by replacing Sm with heavy rare earth (RE) elements (such as Gd, Er and Dy).Subsequently,researchers[9-12]discovered that a positive coercivity coefficient of temperature could be achieved by increasing the Sm content, reducing the content of Cu or performing an incomplete aging treatment(later referred to as the Cu-rich CB method).Since the Cu-rich CB method results in a positive coercive temperature coefficient by reducing the Cu content in the cell boundary phase (CBP), and the decrease in Cu content leads to a decrease in the coercivity of the magnet, this method contradicts existing knowledge regarding the fabrication of high-coercivity magnets.[13]

The Sm1-xDyx(Co0.695Fe0.2Cu0.08Zr0.025)7.2magnets with spin-reorientation transition (SRT) of the CBP (later referred to as the SRT-CB method)have reported a positiveβvalue.[4]This result revealed that SRT of CBPs was responsible for abnormal coercivity.In contrast to the Cu-rich CB method, the SRT-CB method could result in high coercivity,and the positive temperature range could be varied by adding a RE(Nd,Pr,Dy,Tb).[4,14]

Paradoxically, becauseαandβare both related to the RE, it is unclear what effect SRT of CBPs has on the temperature dependence of the magnetization and coercivity of Sm2Co17magnets.In this study,Sm0.6Gd0.1Dy0.3(Co0.695Fe0.2Cu0.08Zr0.025)7.8alloy with SRT of CBPs and Dy88Cu12alloy powder(Dy-Cu)are used as the base material and additive, respectively, to regulateT1:5SR(the temperature of SRT of CBPs).Four Sm2Co17magnets with SRT of CBPs are prepared by doping 0 wt.%,3 wt.%,6 wt.%and 9 wt.%of Dy-Cu with liquid-phase sintering.The effects of SRT of CBPs on the temperature dependence of magnetization and coercivity are systematically studied.

2.Experimental details

A series of magnets with SRT of CBPs are prepared by liquid-phase sintering.The alloys Sm0.6Gd0.1Dy0.3(Co0.695Fe0.2Cu0.08Zr0.025)7.8and Dy88Cu12were prepared by induction melting in an argon atmosphere.The alloys were then milled into powders with a particle size of approximately 3-5 μm.Additions of 0 wt.%, 3 wt.%,6 wt.% and 9 wt.% of Dy88Cu12as a base material were mixed and pressed under a magnetic field(3 T).Further compacting was followed by cold isostatic pressing to obtain a green compaction.After the green compaction was sintered at 1200-1220?C for 2 h the solution was treated at 1160-1190?C for 2 h and quenched in a furnace after being aged at 830?C for 12 h.

The phase structure was examined by x-ray diffraction(XRD)with CuKαradiation.The cellular structure and elemental distribution were observed with a transmission electron microscope(TEM)equipped for energy-dispersive x-ray spectroscopy.The SRTs were measured by a physical properties measurement system.The temperature stability was measured by a Quantum Design superconducting quantum interference device vibrating sample magnetometer.The magnetic properties were examined at the National Institute of Metrology(NIM-500C).

3.Results and discussion

The XRD patterns of the magnets doped with 0 wt.%,3 wt.%, 6 wt.%and 9 wt.%Dy-Cu are shown in Fig.1.According to the PDF card, all samples consist mainly of two phases: a Th2Zn17-type rhombohedral phase(2:17R,i.e.,cell phase) and a CaCu5-type hexagonal phase (1:5H, i.e., CBP).With increasing Dy-Cu doping,the diffraction peak intensity of the magnets at approximately 42.7?tends to increase; this could be the increase in the 1:5H phase content of the magnets.

The cellular microstructure and the Dy and Cu element distributions were observed by TEM.The results for the magnets doped with 0 wt.%,3 wt.%and 9 wt.%Dy-Cu are shown in Figs.2(a)-2(c).As shown in Fig.2(a), the cell structure of the magnet without Dy-Cu doping is nonuniform, and the CBP is discontinuous.The statistical results show that the average size of the cell structure is approximately 100 nm.The line scanning results show that the CBP is Cu-rich and Dypoor compared with the cell interiors.High-resolution electron microscope analysis shows that the thickness of the CBP is approximately 10 nm.As shown in Fig.2(b), the cellular structure of the 3 wt.% Dy-Cu doped magnet appears to be more uniform and complete than that of the magnet without Dy-Cu doping.The statistical calculation results show that the average size of the cell structure is approximately 80 nm.The distribution of Dy in the cellular structure seems more uniform,and the thickness of the CBP increases to approximately 15 nm.As shown in Fig.2(c),the size of the cellular structure of the magnet doped with 9 wt.%Dy-Cu is small.As Dy stabilizes the 2:17H phase and prolongs the aging time,[15]some areas are not as large as they would normally be.The statistical results show that the average size of the cell structure is approximately 50 nm,and the Dy content in the CBP is higher than that in the cell interiors.The thickness of the CBP is approximately 23 nm.The results of morphological observation,statistical calculation and high-resolution electron microscopy analysis show that the number,thickness and volume fraction of the 1:5H CBP increase with the ratio of Dy-Cu doping.This is consistent with the results of XRD analysis.Moreover,the addition of Dy-Cu is conducive to the segregation of Dy at the cell boundary,and(Sm,Dy)(Co,Cu)5compounds with spin reorientation are formed.[16-18]

Fig.1.XRD patterns of the magnets doped with 0 wt.%,3 wt.%,6 wt.%and 9 wt.%Dy-Cu.

Fig.2.TEM analysis of magnets doped with(a)0 wt.%,(b)3 wt.%and(c)9 wt.%Dy-Cu.

Fig.3.Temperature dependence of the real part of the AC susceptibility of the magnets doped with 0 wt.%,3 wt.%,6 wt.%and 9 wt.%Dy-Cu.

Fig.4.(a)Temperature dependence of the magnetization(M-T curves)for the magnets doped with 0 wt.%,3 wt.%,6 wt.%and 9 wt.%Dy-Cu.(b)Temperature dependence of dM/dT for the magnets doped with 0 wt.%,3 wt.%,6 wt.%and 9 wt.%Dy-Cu.

Fig.5.Temperature dependence of the coercivity of magnets doped with 0 wt.%,3 wt.%,6 wt.%and 9 wt.%Dy-Cu.

It is found that SRT of CBPs can optimize the coercivity temperature coefficient with little effect on the temperature dependence of magnetization.Moreover,the temperature range of low-temperature-dependent magnetization and coercivity tends to move towards higher temperatures, which is conducive to the preparation of magnets with a low temperature coefficient at higher temperatures.The magnet doped with 6 wt.% Dy-Cu exhibits ideal temperature stability near room temperature.As a result, the demagnetization curves of the magnets were precisely compared in the range from room temperature (RT) to 100?C with a NIM 500C magnetic measuring device every 20?C.As shown in Fig.6(a),a low point is observed for the magnet without Dy-Cu doping, for which the magnetic properties areBr= 8.71 kG,Hcj=16.37 kOe,(BH)max=17.61 MGOe,α=0.01%·?C-1andβ=-0.32 %·?C-1.Lowαandβvalues for the magnet with 6 wt.% Dy-Cu doping are shown in Fig.6(b), at which point the magnetic properties areBr=7.80 kG,Hcj=8.03 kOe, (BH)max=13.45 MGOe,α=0.01 %·?C-1andβ=-0.03%·?C-1.Both theirα(RT-100?C)values are similar,while theβ(RT-100?C)values decrease from-0.32%·?C-1to 0.03%·?C-1.

Fig.6.Demagnetization curves of the magnets with different amounts of Dy-Cu doping of(a)0 wt.%and(b)6 wt.%.in the range from RT to 100 ?C.

4.Conclusion

Acknowledgements

Project supported by the National Key R&D Program of China(Grant Nos.2021YFB3803003 and 2021YFB3503101),Youth Innovation Promotion Association of CAS (Grant No.2023311),Major Project of‘Science and Technology Innovation 2025’ in Ningbo (Grant No.2020Z044), and Zhejiang Provincial Key Research and Development Program(Grant No.2021C01172).