A facile way to synthesize rare-earth-free Mn–Bi@Bi magnetic nanoparticles

Abstract

Rare-earth-free manganese bismuth (MnBi) is a promising candidate as an earth abundant permanent magnet for its large magnetocrystalline anisotropy and high energy density. A facile aqueous solution method was designed to synthesize MnBi@Bi nanoparticles, which exhibits a saturation magnetization of 2.37 emu g⁻¹, the coercivity of 8 kOe at 5 K and superparamagnetism at 300 K. This processing approach eliminates severe synthesis equipment and makes it a versatile scalable strategy for preparing MnBi nanoparticles.

---

Introduction

Permanent magnets have emerged as a critical component of modern society.¹ High-performance rare earth magnets, such as Nd–Fe–B and Sm–Co systems, have been specially developed as a result of the growing demand for green energy technologies.^2–5 However, due to the high cost and limited supplies of rare earth elements, substitute magnets without rare earth elements draw intense attention from material science researchers. Inter-metallic compound manganese bismuth (MnBi) is a promising permanent magnetic material without rare earth element for its large magnetocrystalline anisotropy, which can lead to large coercivity and high energy density.

In the past, solid-state syntheses of MnBi have been well studied, such as arc melting, rapid solidification, and ball milling.^6–16 Nevertheless, few chemical synthesis of MnBi has been reported. Due to the large negative reduction potential of soluble Mn salts (Mn²⁺/Mn = −1.185 V, Bi³⁺/Bi = −0.31 V), it is quite difficult to co-reduce manganese and bismuth simultaneously in chemical synthesis process. Ren et al. reported a unique synthesis utilizing the metal-redox method to generate stable and nanosized MnBi particles.^17–19 This method eliminates the need for harsh reducing agents, while severe oxygen-free Schlenk-line is still essential, which demands complicated and elaborate operation. Rowe et al. reported a similar method substituting zero-valence Mn precursor with homemade MnLAERC and further thermal annealing treatment can improve microstructure and magnetic properties of the resultant MnBi particles.⁸

Here we report a new hydrothermal method to synthesize Mn–Bi@Bi nanoparticles. This method simplifies and eliminates the need for standard air-free Schlenk line technique, which is a promising candidate for industry conversion. As we know, chemical synthesis methods have been widely used to prepare nanostructure materials. In chemical methods, nanoparticles size can be controlled by adjusting reaction parameters such as time, temperature, and concentration of reagents. Particles growth can be minimized by controlling the reaction time and temperature. Alloying through parameters modification can also be accomplished by chemical synthesis techniques.^20–22 The synthesized Mn–Bi@Bi nanoparticles exhibit a saturation magnetization (M_s) of 2.37 emu g⁻¹ and the coercivity (H_c) of 8 kOe at 5 K. The magnetic properties of the Mn–Bi@Bi nanoalloys can be readily modified by changing reaction time, offering a versatile scalable strategy for manufacturing MnBi.

Experimental section

Materials

Manganese acetate tetrahydrate (99%), bismuth nitrate pentahydrate (98+%), hydrazine hydrate (80%) and polyvinylpyrrolidone (PVP, K = 15) were purchased from Alfa Aesar and used as received without further purification.

Synthesis of MnBi

For a typical MnBi synthesis, 0.2 mmol of manganese acetate tetrahydrate (Mn(Ac)₂·4H₂O), 0.2 mmol polyvinylpyrrolidone (PVP, K = 15) and 0.1 mmol bismuth nitrate pentahydrate (Bi(NO₃)₃·5H₂O) were dissolved into 10 mL of distilled water under rapid stirring for 15 min, and finally 3 mL of hydrazine hydrate (N₂H₄·H₂O) was dropwise added to the solution. The solution in a Teflon-lined reactor with a capacity of 25 mL was sealed in the autoclave and maintained at 240 °C for proper reaction time. After the reaction, the autoclave was cooled down to room temperature by itself. The black-solid products were washed with distilled water and ethanol, and then separated by centrifugation and purified by being re-dispersed. The product after purification was collected and dried under Ar flow to avoid overpressure during annealing. In order to obtain Mn–Bi@Bi alloy, the as-synthesized powders were annealed at 600 °C for 2 h at H₂/Ar atmosphere.

Characterizations

The X-ray diffraction (XRD) was done at room temperature using monochromated Cu-K_α radiation on a Bruker diffraction system operating at 30 kV and 20 mA. A drop of sample diluted in hexane was placed on a 100 mesh carbon-coated copper grid for TEM observation on a Tecnai G2-F20 at 200 kV. SEM images and energy dispersive spectroscopy (EDS) were operated on a Supra-50 at 100 kV. The magnetic hysteresis (M–H) loops were taken on a superconducting quantum interference device (SQUID) from 300 K down to 5 K.

Results and discussion

The Mn–Bi nanocrystal is synthesized using a hydrothermal approach, which is followed by a high temperature (600 °C) thermal annealing to induce the elemental diffusion between Mn and Bi for the formation of MnBi alloys. Based on our previous work, quick reduction of the bismuth salt occurs upon the temperature increasing above 130 °C, and Bi nanospheres are formed within several minutes. There is no manganese present in these initial particles due to their different reduction potential and rates (reduction potentials for Mn²⁺ and Bi³⁺ being −1.185 V and −0.31 V, respectively). It is important to note that the initial Bi nanospheres begin to merge together, likely due to their softening, as the reaction temperature is nearing bismuth's melting point of 271.5 °C. Once the temperature reaches 240 °C, the manganese element is released from the precursor, and manganese monomers nucleate on the surface of bismuth nanospheres. It's important to notice that smaller manganese particle is formed during the hydrothermal process, which does not participate in the Bi sphere. Further thermal annealing may induce the manganese to diffuse into the bismuth core nanospheres, where alloying happens.

Fig. 1a shows the Rietveld refined XRD patterns of the annealed Mn–Bi@Bi alloy nanoparticles synthesized for 72 h, which confirms the formation of crystalline MnBi alloy phase. Large amount of Bi appears as the co-product, which may serve as the template for Mn atoms. MnO phase in the XRD patterns of annealed samples indicates that Mn atoms locate on the surface of Bi nanospheres and get oxidized during the following process. Furthermore, the intensity of MnBi increases with reaction time and the intensity of MnO phase recedes with the increasing of reaction time (Fig. 1b), which is the evidence for more manganese atoms diffusing into the bismuth core. From the results of Rietveld refinement of XRD patterns, the fraction of MnBi in the sample is 4.2 ± 0.4% and further improvement is still needed to increase the content of MnBi. In this study, N₂H₄·H₂O was treated as the reducing agent for the formation of zero-valence Mn. It is difficult to confirm the formation of Mn nanoparticles by XRD directly, due to the Mn nanoparticles is nanocrystalline or amorphous and it is challenging to be measured because of its oxophilicity. This has been confirmed by the formation of the MnO from the XRD patterns (Fig. 1a).

Fig. 1 (a) The XRD patterns of the annealed Mn–Bi@Bi synthesized for 72 h. (b) The XRD patterns of the annealed Mn–Bi@Bi with reaction time from 12 h to 48 h.

Fig. 2a shows TEM images of Bi nanospheres exhibiting a smooth surface, and Fig. 2b shows TEM images of as-synthesized Mn–Bi particles which exhibit a rough surface. TEM image of annealed Mn–Bi@Bi was shown in Fig. 2c. It is important to notice that the Mn shell (light dark shell in Fig. 2b) cannot be seen in Fig. 2c, which indicates that Mn and Bi diffuse into each other and form MnBi alloy. The outer shell of Mn–Bi@Bi is MnO (Fig. 4d), which can be seen as the evidence of Mn monomers. The nano-SAED image of MnBi region is the inset of Fig. 4(d). Bismuth displaying such a low melting temperature (271.5 °C) affects the shape of the particles as seen.

Fig. 2 (a) Initial bismuth particles after rapid reduction. (b) Mn–Bi@Bi particles before annealing exhibiting manganese around bismuth particles. (c) Mn–Bi@Bi particles after annealing. (d) MnO shell on the annealed Mn–Bi@Bi particles (inset: nano-SAED image of MnBi region with the zone axis along [21̄1]).

The microstructure of the annealed Mn–Bi@Bi sample was further studied by SEM and elemental mapping (Fig. 3). Elemental mapping shows that both the Mn and Bi elements are distributed throughout the particle, where the surface exhibits high manganese counts and large amounts of bismuth reside inside the particle. Combined with XRD patterns, it is confirmed that dispersed MnBi alloys are decorated on the surface of Bi particles, which indicates that the thermal annealing induces the element diffusion and alloying. Furthermore, elemental mapping shows that Mn disperses around the Mn–Bi@Bi particles, which may serve as the evidence of manganese losses. EDS indicates that the manganese and bismuth are dominant of the annealed particles, and the oxygen is from the MnO phase.

Fig. 3 (a) SEM element mapping images of annealed Mn–Bi@Bi particles. (b) EDS of annealed Mn–Bi@Bi particles.

Fig. 4 shows the magnetization versus magnetic field (M(H)) curves for the annealed Mn–Bi@Bi particles with different reaction time measured at 300 K and 5 K. Since the Bi contents of each sample are different, the linear part of M(H) curves has been deducted to subtract the background signal. No detectable hysteresis is observed at 300 K as shown in Fig. 4(a). It can be seen that particles is not saturated even up to 50 kOe at 5 K (Fig. 4(b)). The non-hysteresis and non-saturation indicate that a superparamagnetism (SPM) phase dominates over the Mn–Bi@Bi particles at 300 K. The Langevin function was used to fit the SPM M(H) curves of annealed 72 h sample measured at 300 K.²⁰ The inset of Fig. 4(a) shows that the Langevin function fits well with the measured data, and the density of SPM clusters N and the magnetic moment of a cluster μ are 2.64 × 10¹⁴ and 7.055 × 10⁻¹⁸ emu, respectively. Based on the fitting results and MnBi configuration,²³ the average volume of a cluster is about 17.6 nm³, which verifies the XRD results (Fig. 1(a)). The zero-field-cooled (ZFC) and field-cooled (FC) magnetization versus temperature curves with the blocking temperature ∼ 37 K are shown in the inset of Fig. 4(b). From the blocking temperature, the magnetocrystalline anisotropy value can be estimated to be about ∼2.3 × 10⁶ erg cm⁻³, giving the coercivity around a few thousands Oersteds as seen in Fig. 4(b). As pointed out by Suzuki et al.,²⁴ the uniaxial magnetic anisotropy constant Ku of MnBi is found to monotonously increase with temperature from 3 × 10⁶ erg cm⁻³ at 150 K to 1.5 × 10⁷ erg cm⁻³ at 400 K. Furthermore, according to the Rietveld refinement results, the fraction of MnBi in the sample is about 4.2 ± 0.4%, and the measured saturation magnetizations of annealed 72 h sample are 1.84 × 10⁻² emu g⁻¹ and 2.37 emu g⁻¹ at 300 K and 5 K, respectively. If the whole particles are pure MnBi, the spontaneous magnetizations would be 0.44 emu g⁻¹ and 56.43 emu g⁻¹ at 300 K and 5 K respectively, which reckons without the enhanced particles interaction. Both the low temperature magnetocrystalline anisotropy and spontaneous magnetization values reported in our work are comparable with Sellmyer's and other groups.^24–28 However, the room temperature value of saturated magnetization is too small and closed to the value reported by Kavita et al.²⁹ The above results further mean that the nominal Mn–Bi@Bi particles can be regarded as a composite system where the MnBi clusters are embedded in the Bi nonmagnetic matrix.

Fig. 4 Magnetic hysteresis (M–H) loops of MnBi nanoparticles (a) at 300 K (inset: Langevin fitting of sample 72 h at 300 K) and (b) at 5 K (inset: ZFC/FC loops of sample 72 h). (c) Saturation magnetization increases from 12 h to 72 h reaction.

Initial stoichiometry of the precursors played a critical role in producing the proper MnBi phase. In general, losses of Mn atoms induce the low amount of MnBi alloys. Some Mn clusters will be washed out during the magnetic separation at the end of the reaction, which does not participate in alloying into the bismuth spheres. Therefore, the reaction required excess Mn rather than stoichiometry 1 : 1 ratio within the final Mn–Bi@Bi particles. As mentioned before, XRD patterns of thermally annealed particles shown in Fig. 1b clearly demonstrate that MnBi phase can be enhanced with reaction time extending from 12 h to 48 h. Meanwhile, magnetization increase monotonically with the reaction time from 12 h to 72 h for both 5 K and 300 K (Fig. 4c), which could be attributed to high density Mn deposition on the surface of particles with increasing reaction time. Low M_s can be attributed to the high Bi concentration within the particles. An increasing in M_s is observed, which is consistent well with lower Bi percentage within the particles because of better diffusion and alloying (Fig. 1b).

By combining different characterization techniques such as TEM, EDS and XRD, we have shown that annealed Mn–Bi@Bi nanoparticles are alloyed structures with an ordered arrangement with a Bi matrix and decorated MnBi nanoparticles on the shell. It is interesting to consider how the heterogeneous Mn–Bi@Bi alloy formed and how they evolved. Here, a conjecture of growth mechanism has been made to explain the formation of MnBi. In the synthesis of Mn–Bi@Bi nanoparticles, we believe that the standard reduction potential and surface free energy are the most important intrinsic parameters for determining the final heterogeneous nanostructures and subsequently lead to the distribution of MnBi and Bi within the particles. On one hand, the higher value of standard reduction potential of Bi in solution at room temperature (Bi³⁺/Bi = −0.31 V) suggests that the Bi ions are reduced quickly by the addition of N₂H₄, which forms first. However, since the reduction potential of Mn is −1.185 V, Mn would be reduced more difficultly during the nucleation and growth stage and would be found in the outer shell. On the other hand, surface free energy, defined by the change of the free enthalpy of the surface-creating process under an isothermal-isobaric condition, gives rise to the same trend for the Mn/Bi atomic distribution compared to the reduction potential. The surface free energy of Mn at room temperature is higher than that of Bi (Mn = 1.440 J m⁻², Bi = 0.382 J m⁻²).³⁰ This would then conclude that Mn atoms would cluster near the surface of the product, while the Bi atoms are relegated to the core region. These two parameters lead to the final distribution of final particles, which MnBi locates on the surface and excess Bi on the core region.

Conclusion

In conclusion, solution-processed Mn–Bi@Bi nanoparticles that exhibit a saturation magnetization (M_s) of 2.37 emu g⁻¹ and a coercivity (H_c) of 8 kOe at 5 K has been reported. This aqueous solution process eliminates the need for harsh reducing agents and severe oxygen-free equipment, which offers a versatile scalable strategy for manufacturing MnBi. Due to decorated MnBi clusters embedded in the Bi matrix, SPM phase of MnBi is observed at room temperature. Increasing the reaction time can strength the Mn–Bi@Bi magnetization, and further efforts will be paid on other reaction parameters.

Acknowledgements

The project is supported by the National Basic Research Program of China (Grant No. 2015CB921502), the National Natural Science Foundation of China (Grant No. 11474184 and 11627805), the 111 project under Grant No. B13029, and the Fundamental Research Funds of Shandong University.

References

1. N. Jones, Nature, 2011, 472, 22–23.
2. S. Tasoglu, C. H. Yu, H. I. Gungordu, S. Guven, T. Vural and U. Demirci, Nat. Commun., 2014, 5, 4702.
3. M. A. Wheeler, C. J. Smith, M. Ottolini, B. S. Barker, A. M. Purohit, R. M. Grippo, R. P. Gaykema, A. J. Spano, M. P. Beenhakker, S. Kucenas, M. K. Patel, C. D. Deppmann and A. D. Guler, Nat. Neurosci., 2016, 19, 756–761.
4. W. L. Zuo, X. Zhao, J. F. Xiong, M. Zhang, T. Y. Zhao, F. X. Hu, J. R. Sun and B. G. Shen, Sci. Rep., 2015, 5, 13117.
5. W. L. Zuo, X. Zhao, T. Y. Zhao, F. X. Hu, J. R. Sun and B. G. Shen, Sci. Rep., 2016, 6, 25805.
6. X. Guo, X. Chen, Z. Altounian and J. O. Ström-Olsen, Phys. Rev. B: Condens. Matter Mater. Phys., 1992, 46, 14578–14582.
7. M. Gjoka, C. Sarafidis, G. Giannopoulos and D. Niarchos, INTERMAG, 2015, EU09.
8. M. P. Rowe, E. Skoropata, Y. Wrocynskyj, R. D. Desautels and J. Van Lierop, INTERMAG, 2015, GD11.
9. S. Saha, R. T. Obermyer, B. J. Zande, V. K. Chandhok, S. Simizu, S. G. Sankar and J. A. Horton, J. Appl. Phys., 2002, 91, 8525.
10. J. B. Yang, Y. B. Yang, X. G. Chen, X. B. Ma, J. Z. Han, Y. C. Yang, S. Guo, A. R. Yan, Q. Z. Huang, M. M. Wu and D. F. Chen, Appl. Phys. Lett., 2011, 99, 082505.
11. J. B. Yang, W. B. Yelon, W. J. James, Q. Cai, S. Roy and N. Ali, J. Appl. Phys., 2002, 91, 7866.
12. W. Xie, E. Polikarpov, J.-P. Choi, M. E. Bowden, K. Sun and J. Cui, J. Alloys Compd., 2016, 680, 1–5.
13. Y. L. Ma, X. B. Liu, V. V. Nguyen, N. Poudyal, M. Yue and J. P. Liu, J. Magn. Magn. Mater., 2016, 411, 116–119.
14. M. Y. Sun, X. W. Xu, X. A. Liang, X. W. Sun and Y. J. Zheng, J. Alloys Compd., 2016, 672, 59–63.
15. T. J. Williams, A. E. Taylor, A. D. Christianson, S. E. Hahn, R. S. Fishman, D. S. Parker, M. A. McGuire, B. C. Sales and M. D. Lumsden, Appl. Phys. Lett., 2016, 108, 192403.
16. S. K. Sharma, H. R. Prakash and S. Ram, AIP Conf. Proc., 2016, 1728, 020657.
17. J. Shen, H. Cui, X. Huang, M. Gong, W. Qin, A. Kirkeminde, J. Cui and S. Ren, RSC Adv., 2015, 5, 5567–5570.
18. A. Kirkeminde, J. Shen, M. Gong, J. Cui and S. Ren, Chem. Mater., 2015, 27, 4677–4681.
19. A. Kirkeminde, S. Spurlin, L. Draxler-Sixta, J. Cooper and S. Ren, Angew. Chem., 2015, 54, 4203–4207.
20. S. Kang, G. X. Miao, S. Shi, Z. Jia, D. E. Nikles and J. W. Harrell, J. Am. Chem. Soc., 2006, 128, 1042–1043.
21. S. Kang, J. W. Harrell and D. E. Nikles, Nano Lett., 2002, 2, 1033–1036.
22. S. Sun, D. Weller, L. Folks and A. Moser, Science, 2000, 287, 1989–1991.
23. P. Toson, A. Asali, G. A. Zickler, and J. Fidler, Materials Science, 2015, arXiv:1508.03208v1.
24. T. Suzuki, T. Hozumi, J. Barker, S. Okatov, O. Mryasov and T. Suwa, IEEE Trans. Magn., 2015, 51, 2102804.
25. V. V. Nguyen, N. Poudyal, X. B. Liu, J. P. Liu, K. Sun, M. J. Kramer and J. Cui, IEEE Trans. Magn., 2014, 50, 2105506.
26. P. Kharel, V. R. Shah, R. Skomski, J. E. Shield and D. J. Sellmyer, IEEE Trans. Magn., 2013, 49, 3318–3321.
27. J. Cui, J. P. Choi, G. Li, E. Polikarpov, J. Darsell, M. J. Kramer, N. A. Zarkevich, L. L. Wang, D. D. Johnson, M. Marinescu, Q. Z. Huang, H. Wu, N. V. Vuong and J. P. Liu, J. Appl. Phys., 2014, 115, 17A743.
28. N. R. Christopher, N. Singh, S. K. Singh, B. Gahtori, S. K. Mishra, A. Dhar and V. P. S. Awana, J. Supercond. Novel Magn., 2013, 26, 3161–3165.
29. S. Kavita, U. M. R. Seelam, D. Prabhu and R. Gopalan, J. Magn. Magn. Mater., 2015, 377, 485–489.
30. L. Z. Mezey and J. Giber, Jpn. J. Appl. Phys., 1982, 21, 1569–1571.

---