High coercivity Pr₂Fe₁₄B magnetic nanoparticles by a mechanochemical method

Abstract

Nd₂Fe₁₄B nanoparticles are widely used because of their outstanding hard magnetic properties. In fact, Pr₂Fe₁₄B has higher magneto-crystalline anisotropy than Nd₂Fe₁₄B, which makes Pr-Fe-B a promising magnetic material. However, the chemical synthesis route to Pr₂Fe₁₄B nanoparticles is challenging because of the higher reduction potential of Pr³⁺, as well as the complex annealing conditions. In this work, Pr₂Fe₁₄B nanoparticles were successfully synthesized via an efficient and green mechanochemical method consisting of high energy ball milling, annealing, and a washing process. Microstructural investigations revealed that the oxide precursors were uniformly wrapped by CaO and CaH₂, which formed an embedded structure after ball milling. Then, Pr₂Fe₁₄B powder was synthesized via a time-saving annealing process. The impact of the Pr₂O₃ content and the preparation conditions was investigated. The coercivity of the as-annealed powder with 100 wt% Pr₂O₃ excess is 18.9 kOe. After magnetic alignment, the coercivity, remanence, and maximum energy product were: 9.8 kOe, 78.4 emu g⁻¹, and 9.8 MGOe, respectively. The present work provides a promising strategy for preparing anisotropic Pr-Fe-B permanent magnetic materials.

---

Introduction

Magnetic materials have been broadly applied in many industrial fields, including information storage, electronic devices, aerospace, energy generation, hybrid vehicles, wind power generators and so on.^1–3 Rare earth transition metal (RE-TM) permanent magnets, especially Nd₂Fe₁₄B, have gained considerable attention and are commonly applied because they are a high energy product.^4–6 The Nd₂Fe₁₄B is one of the rare earth magnetic materials, found by Sagawa et al.⁷ Because of the higher magneto-crystalline anisotropy of Dy₂Fe₁₄B (H_A = 158 kOe) and Tb₂Fe₁₄B (H_A = 220 kOe) compared to that of Nd₂Fe₁₄B (H_A = 67 kOe),^3,8 partial substitution of Dy and Tb for Nd in Nd–Fe–B magnets has been proved to be a convenient method to raise coercivity (H_c).^7–11 However, the addition of Dy and Tb will decrease the magnetization and magnetic energy product due to the ferrimagnetic coupling between the rare earth atom and Fe and increases the cost due to the scarcity of Dy/Tb resources. Therefore, developing a high-coercivity Nd₂Fe₁₄B magnetic material without Dy/Tb is becoming a hot topic in the permanent magnetic materials field. Compared to Nd₂Fe₁₄B, Pr₂Fe₁₄B with a higher magneto-crystalline anisotropy (H_A = 87 kOe) is a promising hard magnetic material with high coercivity.¹²

The majority of the research on the fabrication of the Nd–Fe–B magnets has centered on physical methods, including powder metallurgy sintering,¹³ HDDR,¹⁴ melt-spinning,^15,16 high-energy ball-milling,^17,18 and so on. Chemical methods, for example, microwave combustion,^3,21 sol–gel^19,20 and thermal decomposition have been applied in the preparation of Nd₂Fe₁₄B.¹⁹ Compared with physical methods, a chemical method has distinct advantages in regulating size and morphology, which have a significant effect on the magnetic properties. The chemical synthesis used oxides as the precursors, which were reduction-diffused by CaH₂ or Ca to form Nd₂Fe₁₄B.^20–22 It is reported that the holding time during the annealing of the chemical synthesis method is generally greater than 90 min, which increased the energy loss and prolonged the production cycle so that limited the application in industrial production.^23–25 Recently, an efficient and green synthesis method, a mechanochemical method, has been used to synthesize Nd₂Fe₁₄B magnetic nanoparticles.^26–30 The mechanistic method can promote the reaction between solids quickly and quantitatively, without adding solvent or adding only a nominal amount of solvent.

In this study, anisotropic Pr₂Fe₁₄B magnetic powder with a high coercivity is synthesized by the simple mechanochemical method for the first time.^31–33 The reducing agent, CaH₂, was selected to reduce the Pr₂O₃, Fe₂O₃, and B₂O₃. The precursor samples were mixed uniformly and wrapped with CaH₂ after the high energy ball milling, which contributed to the completed reduction of the as-milled oxide in the process of annealing. In this work, the optimal ball milling time, Pr₂O₃ content as well as annealing conditions were determined by comparing the magnetic properties. Meanwhile, the transition of the phases and microstructure were systematically investigated to reveal the reaction process.

Experimental

Materials

The precursors were commercially available and used without further purification: Pr₂O₃ (99.9%, Adamas-beta), and B₂O₃ (99.9%, Aladdin) powders, Fe₂O₃ (99.9%, Aladdin), and CaH₂ granules (99.9%, Aladdin), NH₄Cl (99.9%, Sigma-Aldrich), and methanol (99.99%, Sigma-Aldrich).

Synthesis

The fabrication process of the Pr₂Fe₁₄B magnetic materials is demonstrated in Scheme 1. After milling for 2–5 h, the as-milled samples were annealed at 850 °C for 5 min in a glove box under an argon (Ar) atmosphere. The relative contents of Pr₂O₃, Fe₂O₃, B₂O₃ and were determined from the equation:

Pr₂O₃ + 7Fe₂O₃ + 0.5B₂O₃ + 25.5CaH₂ → Pr₂Fe₁₄B + 25.5CaO + 25.5H₂↑

Scheme 1 A schematic diagram of the fabrication process of Pr₂Fe₁₄B-based nanoparticles and their microstructure.

The CaH₂ acts as the reducing agent. The weight ratio of the preliminary reduction of Pr₂O₃ powders to CaH₂ was 1 : 1.5 to ensure full reduction. A portion of the mixture (2 g) was added to a stable steel vial under an Ar atmosphere. Because of the evaporation losses, an excess of 50 wt%, 100 wt% and 150 wt% Pr₂O₃ were added (denoted as W_50%, W_100% and W_150%). The ball milling was carried out using an 8000 M ball mill (SPEX Sample Prep) with a ball-to-powder ratio of 10 : 1, from 2 h to 5 h. The milling time was a key parameter and its influence on the phase transitions was investigated carefully. Subsequently the as-milled powders of 200 mg were annealed from room temperature to the target temperature in the RTP-1200 vacuum heating furnace. All the manipulations of the samples were performed in a LS800S glove box (Dellix, China) filled with a high purity Ar atmosphere. After agitating ultrasonically three times in a NH₄Cl/methanol solution, the annealed samples were cleaned for 10 min. The NH₄Cl/methanol solution was provided by dissolving 2 g of NH₄Cl in 100 ml of methanol. The NH₄Cl/methanol washing system was used, because CaO and NH₄Cl are able to react to generate NH₃ and CaCl₂ which dissolve clearly in methanol solution with very little heat generation.²⁶

The sample components were determined by X-ray diffraction (XRD) on a D/max 2550 instrument (Rigaku) with Cu-Kα radiation. The elemental content and microstructure were determined using field emission scanning electron microscopy on a Supra 40/Gemini column instrument (Zeiss) with energy dispersive spectroscopy (EDS) detection and transmission electron microscopy (TEM) on a Talos F200S (FEI). The aligned sample was prepared as follows: the final powder was agitated ultrasonically in a methanol solvent and then placed in a 1 T magnetic field. After the methanol solvent had evaporated, the powder was solidified by addition of epoxy. Prior to magnetic measurement, a 9 T pulsed magnetic field was used to magnetize the sample to saturation. The final magnetic properties were measured by a vibrating sample magnetometer, VSM-7410 (VSM, Lake Shore Cryotronics) and the largest magnetic field was 2.8 T at 300 K.

The magnetization M and maximum energy product (BH)_max of the powders were calculated using the density of a 100%-dense block material, and it was assumed that the magnets synthesized from the magnetic powders behaved similarly in the magnetic field. The sample density measured by the Archimedes principle was about 7.5 mg cm⁻¹.^32,34

Results and discussion

The XRD patterns of the as-milled samples with different Pr₂O₃ contents are shown in Fig. 1(a)–(c). The Fe₂O₃ vanished and CaO, Ca(OH)₂ and α-Fe were observed, meaning that the Fe₂O₃ has been reduced to α-Fe by CaH₂. The Pr₂O₃ exists, especially in a sample with short milling time, suggesting that it is hard to reduce Pr₂O₃. The reduction potential of Fe₂O₃ and Pr₂O₃ were Fe³⁺/Fe (−0.44 eV) and Pr³⁺/Pr (−2.47 eV).^26,35 Therefore, Fe₂O₃ was reduced to α-Fe during high energy ball milling whereas Pr₂O₃ was seldom reduced. The Ca(OH)₂ is observed in Fig. 1(a)–(c). The formation of Ca(OH)₂ is due to the decomposition of CaH₂ being initially excessive, and it can easily convert into Ca(OH)₂ upon exposure to air.³⁶

Fig. 1 The XRD diffraction patterns of samples with different amounts of Pr₂O₃ before annealing ((a) W_50%, (b) W_100%, (c) W_150%) and after annealing ((d) W_50%, (e) W_100%, (f) W_150%), all ball milled for 2–5 h.

After annealing, the α-Fe and Pr₂O₃ phases gradually decreased and Pr₂Fe₁₄B was observed (Fig. 1(d)–(f)), suggesting that the Pr₂O₃ was reduced during the annealing and then reacted with Fe and B to produce Pr₂Fe₁₄B. The B₂O₃ was reduced during the high energy ball milling, but it was difficult to observe in the XRD patterns due to the small amount. The reduction of Pr₂O₃ and the formation of Pr₂Fe₁₄B were promoted by the ball milling. For W_50%, the relative content of α-Fe increased abnormally when milled for 5 h, which should be attributed to the insufficiency of Pr₂O₃ and hence, the Pr content, caused by the loss of Pr during the long time of ball milling. No α-Fe appeared in the samples with more Pr₂O₃ (W_100% and W_150%) and the long milling time, suggesting that the Pr₂O₃ content was sufficient.

The magnetization hysteresis loops of the as-milled powder (Fig. 2(a)) show the existence of abundant soft magnetic α-Fe phases, which is in agreement with the XRD results in Fig. 1(a). The VSM results in Fig. 2(a) show that the coercivity was about 500 Oe, which may be the reason for the existence of trace amounts of Pr₂Fe₁₄B during the ball milling process.

Fig. 2 Magnetization hysteresis loops of (a) an as-milled Pr₂O₃ sample (W_50%). The inset shows the corresponding magnified second quadrant of these M–H curves.

The demagnetization curves of the as-annealed samples are presented in Fig. 3(a)–(c). With an increase of ball milling time, the permanent magnetic properties were first enhanced and then weakened, the magnetism peaked at 4 h. The corresponding coercivity and remanence of the annealed samples ball milled for 4 h were 14.2 kOe and 24.1 emu g⁻¹, 18.9 kOe and 23.0 emu g⁻¹, 20.4 kOe and 19.2 emu g⁻¹, for W_50%, W_100%, and W_150%, respectively, see Table 1.³⁷ The poor permanent magnetic properties of the samples with short ball milling time was because of the existence of magnetically soft α-Fe, as proved by the XRD patterns. Whereas excessive ball milling might result in a grain size which was too small, which decreased the coercivity by the superparamagnetic effect from the magnetic properties shown in Fig. 3(d) and Table 1, an insufficient Pr₂O₃ content caused the appearance of α-Fe which decreased the coercivity, and an excessive Pr₂O₃ content decreased magnetization because of the incomplete reduction of no magnetic Pr₂O₃. Therefore, the optimum Pr₂O₃ content is W_100%, and the optimal ball milling time is 4 h.

Fig. 3 Demagnetization curves of annealed Pr₂O₃ ((a) W_50%, (b) W_100%, (c) W_150%), and the demagnetization curves of samples annealed at the optimum ball milling time of 4 h (d).

Table 1 The coercivity (H_c) and remanence (M_r) values of Pr₂Fe₁₄B nanoparticles with different Pr₂O₃ levels

Sample name		Ball milling time
		2 h	3 h	4 h	5 h
W50%	Hc (kOe)	5.6	11.2	14.2	9.6
	Mr (emu g^−1)	18.7	21.5	24.1	20.1
W100%	Hc (kOe)	4.5	9.9	18.9	15.4
	Mr (emu g^−1)	18.3	18.9	23.0	20.4
W150%	Hc (kOe)	11.2	14.3	20.4	18.6
	Mr (emu g^−1)	16.3	18.3	19.2	17.9

---

The formation mechanism of Pr₂Fe₁₄B in this mechanochemical method during annealing was further explored. The influence of the ball milling temperature was investigated based on Pr₂O₃ (W_100%) ball milled for 4 h, as shown in Fig. 4. It was found that the diffraction peaks of Pr₂Fe₁₄B become more obvious with the increase of the annealing temperature. Whereas magnetically soft α-Fe phase was found when the temperature rose to 900 °C, which may be due to the loss of Pr. The demagnetization curves (Fig. 4(b)) show that the optimum temperature was 850 °C, at which point the optimum permanent magnetic properties were gained. According to Fig. 5(a) and (b), the optimum holding time was 5 min. Extending the holding time has the same effect as raising the temperature.

Fig. 4 The XRD diffraction patterns (a) and demagnetization curves (b) of Pr₂O₃ (W_100%) ball milled for 4 h and annealed at 700–900 °C.

Fig. 5 The XRD diffraction patterns (a) and demagnetization curves (b) of Pr₂O₃ (W_100%) milled for 4 h and then annealed to 850 °C and held for 2 to 10 min.

The XRD patterns of samples, of different Pr₂O₃ contents, washed with NH₄Cl/methanol, with a ball milling time of 4 h are shown in Fig. 6(a). It can be observed that the Pr₂Fe₁₄B phase is the major phase with a small amount of an unidentified phase. The magnetically soft α-Fe phase is observed in washed Pr₂O₃ (W_50%). The standard data of PDF#50-0678 (Pr₂Fe₁₄B) is also presented. Comparing the washed Pr₂O₃ with PDF#50-0678, the peaks shifted a little to the lower 2θ positions, which arose from the formation of a small amount of Pr₂Fe₁₄BH_x during the washing process. From the magnetic hysteresis curves (Fig. 6(b)), after removing CaO, the H_c of W_50%, W_100% and W_150% with a ball milling time of 4 h decreased to 6.8 kOe, 9.4 kOe and 11.7 kOe, respectively. The obvious decrease of coercivity may be attributed to an occurrence of an exothermic oxidation reaction, which lead to a partial Pr₂Fe₁₄B magnetic phase decomposition to soft magnetic phases. Accordingly, the magnetic properties of the final products were generally decreased, when the annealed samples were washed with an NH₄Cl/methanol solution to remove the CaO.^{23,24,35,38–40} The M_r of Pr₂O₃ (W_50%, W_100% and W_150%) at a ball milling time of 4 h increased to 53.3 emu g⁻¹, 56.3 emu g⁻¹ and 44.4 emu g⁻¹ after the washing process, respectively, because of the elimination of nonmagnetic phases such as CaO, as well as Ca(OH)₂. The maximum energy product of Pr₂O₃ (W_50%, W_100% and W_150%) prepared at optimum conditions were 3.7 MGOe, 5.0 MGOe and 3.2 MGOe.

Fig. 6 The XRD diffraction patterns (a) and the magnetic hysteresis curves (b) of washed samples with different Pr₂O₃ levels after annealing.

Fig. 7 shows the hysteresis loops, parallel and perpendicular to the alignment direction, of magnetically aligned Pr₂O₃ (W_100%). Two hysteresis loops show a distinct difference, meaning that the particle is a partial crystal and can be made into anisotropic magnets. With the original washed samples, the M_r value obviously increased, which demonstrated that most of the washed Pr₂Fe₁₄B powders were single domain particles.⁴¹ Fig. 7(b) shows the coercivities, remanence, saturation magnetization and maximum energy product of the magnetically aligned Pr₂Fe₁₄B particles of W_50%, W_100% and W_150%. The M_r increased from 53.3 emu g⁻¹, 56.3 emu g⁻¹, 44.4 emu g⁻¹ to 77.5 emu g⁻¹, 78.4 emu g⁻¹, 61.1 emu g⁻¹, after alignment, respectively. The best permanent magnetic property, a remanence of 78.4 emu g⁻¹, coercivity of 9.8 kOe, and a maximum energy product of 9.8 MGOe, was obtained in Pr₂O₃ (W_100%).

Fig. 7 The magnetization hysteresis loops of magnetically aligned Pr₂O₃ (W_100%) (a), and H_c, M_s, M_r, and (BH)_max data for aligned samples measured along the direction of alignment (b).

To explore the microstructural evolution, SEM and TEM measurements were performed. It can be seen in Fig. 8(a), that the particles were relatively large after a ball milling time of 4 h. The TEM images (Fig. 9(a) and (b)) display sheet-shaped structures of W_100% after milling. Meanwhile, the HRTEM and HAADF images demonstrate that the as-milled powders have an embedded morphology, as shown in Fig. 9(c) and (d). The excessive amounts of CaH₂ ensure that the Pr₂O₃, α-Fe and B₂O₃ were uniformly wrapped by CaO or CaH₂ after high energy ball milling. This special structure may contribute to the fact that the reduction reaction happened in a short time. After annealing, the size of the particles did not change significantly, as shown in Fig. 8(b), because of the covering of CaO on the Pr₂Fe₁₄B particles. The existence of the CaO shell effectively inhibited the growth of the Pr₂Fe₁₄B particles. After removal of the external CaO covering with a NH₄Cl/methanol solution, the irregular spherical or fine rod shaped particles with a smaller size were obtained, which is shown in Fig. 8(c).

Fig. 8 SEM images of powder with a Pr₂O₃ content of W_100% ball milled for 4 h after (a) milling, (b) annealing, and (c) washing.

Fig. 9 The TEM (a) and HRTEM (c) images of Pr₂O₃ (W_100%) powder after milling. (b) A magnified view of the red box in (a). (d) A HAADF image and elemental mapping of Pr, Fe, and Ca. (e) An SEM image of powder with W_100% after washing. The EDS element mapping of (f) Pr L, (g) Fe K, and (h) Ca K.

Elements were uniformly distributed, as shown in Fig. 9(f) and (g) in the element mapping images. Due to the limitation of SEM, it was not possible to see the boron element. The atomic ratio between Pr and Fe was near to 0.21, which is similar to the best reactant ratio of Pr as well as the Fe precursors. The trace atomic ratio of Ca was 0.04% which means that the nonmagnetic phases as CaO and Ca(OH)₂ were eliminated in the cleaning process.

Conclusions

In summary, Pr₂Fe₁₄B nanoparticles were prepared from Pr₂O₃, Fe₂O₃, B₂O₃, and CaH₂ powder via applying a mechanochemical method. The powders were partially reduced during high-energy ball milling, leading to the formation of a Pr₂Fe₁₄B nucleus, which helped the as-milled powders to transform to Pr₂Fe₁₄B in a time-saving annealing process. The phase compositions and magnetic properties can be enhanced via fine-tuning the Pr/Fe ratio and preparation conditions. Coercivity of 18.9 kOe was obtained in the sample with 100 wt% Pr₂O₃ excess, ball milled for 4 h and annealed to 850 °C for 5 min. After washing and magnetic alignment, the coercivity, remanence, and maximum energy product were improved to 9.8 kOe, 78.4 emu g⁻¹, and 9.8 MGOe, respectively. Via fine-tuning the particle size and microstructure, it is possible to further improve the magnetic performance of Pr₂Fe₁₄B materials.

Conflicts of interest

There are no conflicts to declare.

References

1. X. Zhou, Y. L. Tian, H. Y. Yu, H. Zhang, X. C. Zhong and Z. W. Liu, Waste Manag., 2019, 87, 645–651.
2. H. Tu, Y. Yan, D. Cai, J. Wang, F. Li, F. Su and X. Du, J. Magn. Magn. Mater., 2019, 490, 165497.
3. X. Tan, H. Parmar, Y. Zhong, V. Chaudhary and R. V. Ramanujan, J. Magn. Magn. Mater., 2019, 471, 278–285.
4. D. B. D. Mooij and K. H. J. Buschow, J. Less-Common Met., 1988, 142, 349–357.
5. D. Goll, M. Seeger and H. Kronmüller, J. Magn. Magn. Mater., 1998, 185, 49–60.
6. N. Cannesan, J. M. Lebreton, A. J. Williams and I. R. Harris, J. Magn. Magn Mater., 2002, 242, 1372–1374.
7. M. Sagawa, S. Fujimura, N. Togawa, H. Yamamoto and Y. Matsuura, J. Appl. Phys., 1984, 55, 2083–2087.
8. S.-q. Hu, K. Peng and H. Chen, J. Magn. Magn. Mater., 2017, 426, 340–346.
9. M. Yue, X. Zhang and J. P. Liu, Nanoscale, 2017, 9, 3674–3697.
10. D.-s. Wang, Z.-b. Li, F. Liu, Q. Ma, Y.-f. Li, Q. Zhao and X.-f. Zhang, Mater. Lett., 2018, 228, 509–512.
11. R. Goto, M. Matsuura, S. Sugimoto, N. Tezuka, Y. Une and M. Sagawa, J. Appl. Phys., 2012, 111, 2083.
12. W.-W. Yang, J. Luo, H.-J. Peng, Y.-L. Sui and X.-X. Gao, Rare Met., 2014, 33, 171–175.
13. J. Song, S. Guo, G. Ding, K. Chen, R. Chen, D. Lee and A. Yan, J. Magn. Magn. Mater., 2019, 469, 613–617.
14. J. J. Zhang, Y. Yan, H. M. Gao, H. M. Geng, X. Y. Feng, Z. P. Hou, H. D. Li, W. Q. Wang, F. Su and X. B. Du, J. Magn. Magn. Mater., 2015, 374, 317–320.
15. I. Shchetinin, P. Aggrey, I. Bordyuzhin, A. Savchenko, M. Gorshenkov, M. Zhelezniy, V. Menushenkov and P. Mogil’nikov, Metals, 2019, 9, 497.
16. X. Lin, Y. Luo, H.-j. Peng, Y.-f. Yang, Y.-k. Dou, Z.-l. Wang, K.-s. Xu, S.-l. Diao and D.-b. Yu, J. Magn. Magn. Mater., 2019, 490, 165454.
17. Z. Wang, X. Liu, Z. Chen, R. Zhu, W. Yang, Z. Zhang, W. Wang, Q. Wu, K. M. U. Rehman and M. Shezad, J. Supercond. Novel Magn., 2018, 32, 715–720.
18. B. Z. Cui, W. F. Li and G. C. Hadjipanayis, Acta Mater., 2011, 59, 563–571.
19. L. Yu, C. Yang and Y. Hou, Nanoscale, 2014, 6, 10638–10642.
20. H. Rahimi, A. Ghasemi, R. Mozaffarinia and M. Tavoosi, J. Magn. Magn. Mater., 2017, 444, 111–118.
21. L. Wu, A. Mendoza-Garcia, Q. Li and S. Sun, Chem. Rev., 2016, 116, 10473–10512.
22. Z. Ma, T. Zhang and C. Jiang, Chem. Eng. J., 2015, 264, 610–616.
23. P. K. Deheri, V. Swaminathan, S. D. Bhame, Z. Liu and R. V. Ramanujan, Chem. Mater., 2010, 22, 6509–6517.
24. C.-Q. Chen, D. Kim and C. Choi, J. Magn. Magn. Mater., 2014, 355, 180–183.
25. H. Rahimi, A. Ghasemi, R. Mozaffarinia and M. Tavoosi, J. Magn. Magn. Mater., 2017, 424, 199–206.
26. Y. Zhong, V. Chaudhary, X. Tan, H. Parmar and R. V. Ramanujan, Nanoscale, 2017, 9, 18651–18660.
27. A. M. Gabay, X. C. Hu and G. C. Hadjipanayis, J. Alloys Compd., 2013, 574, 472–476.
28. A. Pal, A. Gabay and G. C. Hadjipanayis, J. Alloys Compd., 2012, 543, 31–33.
29. S. L. James, C. J. Adams, C. Bolm, D. Braga, P. Collier, T. Friscic, F. Grepioni, K. D. Harris, G. Hyett, W. Jones, A. Krebs, J. Mack, L. Maini, A. G. Orpen, I. P. Parkin, W. C. Shearouse, J. W. Steed and D. C. Waddell, Chem. Soc. Rev., 2012, 41, 413–447.
30. V. Chaudhary, Y. Zhong, H. Parmar, X. Tan and R. V. Ramanujan, Chemphyschem, 2018, 19, 2370–2379.
31. K. Filipecka, P. Pawlik, A. Kozdraś, J. Filipecki and K. Pawlik, Acta Phys. Pol., A, 2019, 135, 226–228.
32. P. Mingxiang, Z. Pengyue, G. Hongliang, Y. Hangfu and W. Qiong, J. Magn. Magn. Mater., 2012, 324, 2953–2957.
33. O. Koylu-Alkan, J. M. Barandiaran and D. Salazar, et al. Submicron R₂Fe₁₄B particles, AIP Adv., 2016, 6(5), 056027.
34. H. Tu, Y. Yan, D. Cai, J. Wang, X. Shang, H. Tian, F. Su and X. Du, Mater. Des., 2019, 183, 108140.
35. H. X. Ma, C. W. Kim, D. S. Kim, J. H. Jeong, I. H. Kim and Y. S. Kang, Nanoscale, 2015, 7, 8016–8022.
36. A. M. Gabay, X. C. Hu and G. C. Hadjipanayis, J. Magn. Magn. Mater., 2014, 368, 75–81.
37. W. F. Li, A. M. Gabay, X. C. Hu, C. Ni and G. C. Hadjipanayis, J. Phys. Chem. C, 2013, 117, 10291–10295.
38. V. Swaminathan, P. K. Deheri, S. D. Bhame and R. V. Ramanujan, Nanoscale, 2013, 5, 2718–2725.
39. J. H. Jeong, H. X. Ma, D. Kim, C. W. Kim, I. H. Kim, J. W. Ahn, D. S. Kim and Y. S. Kang, New J. Chem., 2016, 40, 10181–10186.
40. Y. Wang, J. Ahn, D. Kim, W. J. Ren, W. Liu, Z. D. Zhang and C. J. Choi, J. Magn. Magn. Mater., 2017, 439, 91–94.
41. C. Li, Q. Wu, Z. Ma, H. Xu, L. Cong and M. Yue, J. Mater. Chem. C, 2018, 6, 8522–8527.

---