Fe permanent magnet produced by ball-milling and warm compaction

Fe permanent magnet produced by ball-milling and warm compaction

Journal of Magnetism and Magnetic Materials 323 (2011) 2855–2858 Contents lists available at ScienceDirect Journal of Magnetism and Magnetic Materia...

968KB Sizes 0 Downloads 36 Views

Journal of Magnetism and Magnetic Materials 323 (2011) 2855–2858

Contents lists available at ScienceDirect

Journal of Magnetism and Magnetic Materials journal homepage: www.elsevier.com/locate/jmmm

Atom probe study on the bulk nanocomposite SmCo/Fe permanent magnet produced by ball-milling and warm compaction X.Y. Xiong a,b,n, C.B. Rong c, S. Rubanov d, Y. Zhang e, J.P. Liu c a

Monash Centre for Electron Microscopy, Monash University, Vic. 3800, Australia Department of Materials Engineering, Monash University, Vic. 3800, Australia c Department of Physics, University of Texas at Arlington, Arlington, TX 76019, USA d Electron Microscopy Unit, Bio21 Institute, University of Melbourne, Vic. 3052, Australia e Division of Materials Science and Engineering, Ames Laboratory, Iowa State University, Ames, IA 50011, USA b

a r t i c l e i n f o

abstract

Article history: Received 12 May 2011 Received in revised form 11 June 2011 Available online 29 June 2011

The microstructure and compositions of the bulk nanocomposite SmCo/Fe permanent magnet were studied using transmission electron microscopy and 3-dimensional atom probe techniques. The excellent magnetic properties were related to the uniform nanocomposite structure with nanometer a-Fe particles uniformly distributed in the SmCo phase matrix. The a-Fe phase contained  26 at% Co, and the SmCo phase contained  19 at% Fe, confirming that the interdiffusion of Fe and Co atoms between the two phases occurred. The formation of the a-Fe(Co) phase explained why the saturation magnetization of the nanocomposite permanent magnet was higher than that expected from the original pure a-Fe and SmCo5 powders, which enhanced further the maximum energy product of the nanocomposite permanent magnet. & 2011 Elsevier B.V. All rights reserved.

Keywords: SmCo permanent magnet Nanocomposite Microstructure Atom probe tomography TEM Ball-mill

1. Introduction Nanocomposite permanent magnets have been the focus of research for two decades with a view to producing more powerful and more economic permanent magnets [1,2]. The extensive research work has produced some successful results [3–7], although their maximum energy products were not as high as predicted by the modeling work [8,9]. As SmCo-based permanent magnets have the highest coercivity and highest Curie temperature among all permanent magnets [10], the nanocomposite SmCo/Fe permanent magnets are expected to have better magnetic properties than the nanocomposite NdFeB/Fe permanent magnets, especially for the high temperature applications. However, for bulk material, how to make an appropriate nanocomposite structure with the a-Fe phase present in nanometer scale and uniformly distributed in the SmCo phase, as suggested by the modeling work [8,9], remains very challenging. Some research groups tried to use powder metallurgy and sintering to produce bulk nanocomposite SmCo/Fe permanent magnets, but failed to achieve good magnetic properties [11,12]. Recently we reported a successful work on fabricating bulk nanocomposite SmCo/Fe

n Corresponding author at: Monash Centre for Electron Microscopy, Monash University, Vic. 3800, Australia. Tel.: þ61 3 99051727; fax: þ 61 3 99053600. E-mail address: [email protected] (X.Y. Xiong).

0304-8853/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.jmmm.2011.06.035

permanent magnets, which used ball-milling and warm compaction of powder mixtures [5]. The maximum energy product was increased more than double compared with that of the singlephase counterparts. The results demonstrated a promising way to produce high temperature and high performance permanent magnets at low cost. The purpose of this paper is to report in detail the microstructure and composition distributions in the high performance nanocomposite SmCo/Fe permanent magnet, together with some issues of the characterization technique. The composition of the nanometer scale particles has been determined quantitatively using a 3-dimensional atom probe (3DAP) technique. In combining with the transmission electron microscopy (TEM) results, the excellent magnetic properties are well understood by relating to the unique uniform nanocomposite structure achieved in this alloy.

2. Experimental Fully dense bulk nanocomposite permanent magnet samples were prepared by ball-milling and warm compaction of hard magnetic and soft magnetic powders. The raw powder materials, commercial SmCo5 and a-Fe powders with particle sizes of 1–20 mm, were mixed as per 80 wt% SmCo5 and 20 wt% a-Fe in a high-energy ball-milling machine and milled for 4 h, followed by

2856

X.Y. Xiong et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 2855–2858

warm compaction at 400 1C under a pressure of  2.5 GPa to form bulk nanocomposite samples with disk-shape in 8 mm diameter. The microstructure of the samples was examined using TEM (Philips 200 kV). Specimens for atom probe analysis were prepared by cutting square rods of dimensions 0.3  0.3  8 mm3 from disk samples using a wire saw, and electropolishing the square rods to sharp needles (  10 mm in diameter) in a solution of 2% HClO4 þ98% 2-butoxyethanol at 16–18 V. The final sharpening was completed on FEI Nova dual beam, focused ion beam (FIB) system. The needle was first line-cut to square cross-section with 3 mm wide and 20 mm deep using 5 nA at 30 kV, followed by annular cone cutting with an internal diameter of 600 nm using 1 nA beam current. 3DAP analysis was conducted in ultrahigh vacuum ( o10–8 Pa) on an Oxford nanoScience energy-compensated 3DAP [13]. The specimen temperature was 60 K. The evaporation rate was kept at  0.01 ion per pulse with the pulse repetition rate of 20 kHz and pulse fraction of 20%.

3. Results and discussion A magnetic hysteresis loop of the sample is given in Fig. 1, showing that the coercivity, Hc ¼6.95 kOe, and the maximum energy product, (BH)max ¼18MGOe (144 kJ/m3). A significant enhancement of the magnetic properties was achieved in the nanocomposite magnets, as the (BH)max of the single phase magnet produced from the SmCo5 powder under the same condition was only 9MGOe (72 kJ/m3). A TEM micrograph of the sample is given in Fig. 2, showing a uniform distribution of small particles of 10 nm in size embedded in the matrix. The exact composition of the small particles could not be obtained from TEM due to the small size of the particles. According to the starting materials, the matrix could be the SmCo phase and the small particles could be the a-Fe particles. The excellent magnetic properties were attributed to the nanocomposite structure with uniformly distributed nanometer a-Fe particles in the matrix, leading to the exchange-coupling between the hard and soft magnetic phases. To determine the exact compositions of the particles and the matrix, the 3DAP analysis were conducted. Due to the nature of the material studied in this work, it took quite an effort to make good tip specimens for the atom probe analysis and achieve successful probing. It proved to be the defects that caused the difficulty for the atom probe analysis.

Fig. 1. Magnetic hysteresis loop of the nanocomposite SmCo/Fe permanent magnet.

Fig. 2. TEM micrograph of the nanocomposite SmCo/Fe permanent magnet.

As mentioned in Section. 2, the needle sample was electropolished and then sharpened by FIB. During the electropolishing, the thin section at the tip end often fell off before the normal sharp end was formed, due to some pores in the material. The needles prepared by electropolishing were about 10 mm in diameter. These needles had to be sharpened further on FIB. The pores in the sample were visible (in white contrast) on FIB and could be avoided for making sharp tips. This paper presents one of the successful 3DAP results. The 3-D atom maps of Sm, Fe, and Co are shown in Fig. 3(a). The size of the whole volume is 10  9  46 nm3. It can be seen that in the Fe atom map there are some Fe-enriched regions, corresponding to the Sm- and Co-depleted regions in the Sm and Co atom maps. Apart from these three elements, oxygen was also detected, mainly in the form of Sm-oxide. As seen in Fig. 3(a), O-enriched regions correspond to the Sm-enriched regions. The a-Fe particle shape was evident by analyzing the iso-concentration surface of Fe atoms, which showed spherical particles. The result is consistent with the TEM observation shown in Fig. 2. To determine the compositions of the individual phases, a small volume of each phase was chosen, as shown by the small box (2  2  19 nm3) passing through different phases in the Sm atom map in Fig. 3(a). The composition profiles along the length direction of the small box are shown in Fig. 3(b). The SmCo phase region (marked in SmCo) and a-Fe phase region (marked in FeCo) can be identified. A Sm-oxide region (marked in SmO) can be seen between these two regions. The measured compositions of the SmCo phase, a-Fe phase and the Sm-oxide are shown in Table 1. From Table 1, it can be seen that the SmCo phase contains  19 at% Fe and the a-Fe phase contains  26 at% Co. These results suggest that the original two phases, SmCo5 and a-Fe powders, have been changed after ball-milling and warm compaction. Not only the grain size was reduced from micron to nanometer scale, but also the compositions of the phases were changed due to the mechanical-alloying effect. The interdiffusion between the two phases must have occurred in the system, i.e., the Fe atoms in the a-Fe phase diffused into the SmCo5 phase to replace Co atoms, and the Co atoms in the SmCo5 phase diffused into the a-Fe phase. This is the first quantitative data showing the extent of interdiffusion of Fe and Co between the two phases in this nanocomposite permanent

X.Y. Xiong et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 2855–2858

Fig. 3. 3DAP analyses of the nanocomposite SmCo/Fe permanent magnet. (a) Atom maps of Sm, Co, Fe, and O (the box size is 10  9  46 nm3), and (b) composition profiles along the length direction of the selected box in the Sm atom map (the selected box size is 2  2  19 nm3).

Table 1 Measured compositions of the SmCo, a-Fe and Sm-oxide phases in the nanocomposite.

SmCo

Sm

Co

Fe

O

13 7 2 at%

68 7 3 at% 26 7 2 at% 31 7 2 at%

19 7 2 at% 74 7 3 at% 97 1 at%

16 7 2 at%

a-Fe Sm–O

44 7 2 at%

magnet. The driving force for this interdiffusion could be due to the negative enthalpy of mixing for Fe and Co, which was estimated to be 2 kJ/mol [14].

2857

This alloying effect is beneficial for the magnetic properties of the nanocomposite magnet. The resulting a-Fe(Co) phase contained 26 at% Co, which is the composition that possesses a higher saturation magnetization than the pure a-Fe. This explained why the measured saturation magnetization (120 emu/g) for the nanocomposite magnet was higher than the expected value (107 emu/g) calculated using the saturation magnetizations of the original a-Fe powder and the SmCo5 powder and their weight percentages in the alloy. The maximum energy product was enhanced accordingly. This result helped successfully to modify the alloy composition to achieve a higher maximum energy product (19.2MGOe) by changing the soft magnetic component from 20 wt% Fe to 25 wt% Fe65Co35 powder. It should be pointed out that the resulting SmCo phase is different from the original SmCo5 phase. The measured concentration of Sm (13 at%) in the SmCo phase is lower than the stoichiometric value of 16.7 at% in the SmCo5 phase. Compared with our previous work [15,16], the measured concentration of Fe (19 at%) in the SmCo phase is much higher than those in the SmCo5 phase for the sintered SmCo-based permanent magnets (11–13 at% Fe [15]) and the melt-spun SmCo-based permanent magnets (about 5 at% Fe [16]). This high Fe concentration may suggest that the alloy system was not in the equilibrium state after ball-milling. It was reported that in the equilibrium state the Fe solubility in the SmCo5 phase was limited,o10 at% at 1200 1C [17]. At a lower temperature, a lower solubility of Fe in the SmCo5 phase is expected. In the contrast, this high Fe concentration seems to approach the value in the Sm2Co17 phase for the sintered and melt-spun SmCo-based permanent magnets, which was 19–22 at% Fe [15,16]. It is known that in the Sm2Co17 phase the Fe atoms can replace Co atoms completely. Therefore, it is reasonable that with a further high temperature treatment, the SmCo phase is likely to transform into Sm2(Co, Fe)17 phase, which was confirmed by our X-ray diffraction analysis. This phase transformation would decrease the anisotropy field [18], hence reducing the coercivity. This may be the reason that the good magnetic properties could not be obtained after a higher temperature treatment. It is worth mentioning that some minor impurities, O, C, N, Al, Cr, and Mn, were detected in the specimen, as shown in the mass spectrum of the sample (Fig. 4). Their atom maps showed that they were randomly distributed in the matrix, except oxygen forming Sm-oxide. The measured impurities in the whole volume were 1.8270.03 at% O, 0.9570.02 at% C, 0.3870.01 at% N, 0.2270.01 at% Al, 0.1570.01 at% Cr, and 0.1270.01 at% Mn. These impurities could be introduced during the ball-milling. Oxygen could be introduced with the raw powders as well, as samarium has a high chemical affinity for oxygen [18]. These impurities are all detrimental to the magnetic properties of the nanocomposite permanent magnets. The result that are obtained suggests that there is still room for improving the magnetic properties of these nanocomposite permanent magnets. If these impurities could be reduced, even higher saturation magnetization and higher maximum energy product could be achieved. Another element, gallium, was detected in the analyzed volume of the tip specimen. It is believed that gallium was introduced by the FIB during sharpening of the tip specimen, as the Ga atoms were present at the surface of the tip specimen with a thickness of  5 nm according to the Ga atom map. One thing that should be noted here, which is relevant to the atom probe analysis technique, is that the charge states of Sm ions during probing for this alloy (as shown in Fig. 4) are the same as observed in the melt-spun SmCo-based permanent magnets [19]. This supports that the probing conditions used in this work were appropriate. As observed in the previous work [19], during the probing process, the major Sm ions were doubly charged from

2858

X.Y. Xiong et al. / Journal of Magnetism and Magnetic Materials 323 (2011) 2855–2858

Fig. 4. 3DAP mass spectrum of the nanocomposite SmCo/Fe permanent magnet.

the alloy. No singly charged Sm ions were observed, but a minor portion of triply charged Sm ions was observed in the mass spectrum (Sm3 þ in Fig. 4). These observations are not consistent with the prediction calculated from the evaporation fields of pure Sm metal, which gave the appearance probability sequence as doubly charged, singly charged, and triply charged ions [19]. This discrepancy suggests that the calculated evaporation fields of the pure Sm metal may not apply to the Sm alloys completely.

4. Conclusions The microstructure and compositions of the bulk nanocomposite SmCo/Fe permanent magnet produced by ball-milling and warm compaction have been characterized using TEM and 3DAP techniques. The excellent magnetic properties of the nanocomposite permanent magnet can be related to the nanocomposite structure with nanometer a-Fe(Co) particles uniformly distributed in the hard magnetic SmCo phase matrix, leading to the exchange-coupling between the hard and soft magnetic phases. The original a-Fe and SmCo5 powders were changed after ballmilling and warm compaction not only in size (from micron to nanometer scale), but also in composition. The Co concentration in the a-Fe phase was  26 at%, and the Fe concentration in the SmCo phase was  19 at%, confirming that the interdiffusion of Fe and Co atoms between the a-Fe and SmCo5 phases occurred. The formation of the a-Fe(Co) phase increased the saturation magnetization of the original a-Fe phase. This explained why the saturation magnetization of the nanocomposite permanent magnet was higher than that calculated using the original a-Fe powder, and enhanced further the maximum energy product. A small amount of Sm-oxide was observed in the nanocomposite. The magnetic properties of the nanocomposite permanent magnet could be improved further if the impurities and defect were reduced.

Acknowledgment This work has been supported in part by the US Office of Naval Research/MURI project under Grant N00014-05-1-0497 and by the University of Texas-Arlington. References [1] J.P. Liu, in: J.P. Liu, E. Fullerton, O. Gutfleisch, D.J. Sellmyer (Eds.), Nanoscale Magnetic Materials and Applications, Springer Science Business Media, LLC, New York, 2009, p. 309. ¨ [2] M. Marinescu, A. Gabay, G.C. Hadjipanayis, in: H. Kronmuller, S. Parkin (Eds.), Handbook of Magnetism and Advanced Magnetic Materials, vol.4: Novel Materials, John Wiley & Sons, Ltd, New York, 2007, pp. 1–22. [3] H. Zeng, J. Li, J.P. Liu, Z.L. Wang, S. Sun, Nature 420 (2002) 395. [4] J. Zhang, Y.K. Takahashi, R. Gopalan, K. Hono, Appl. Phys. Lett. 86 (2005) 122509. [5] C. Rong, Y. Zhang, N. Poudyal, X.Y. Xiong, M.J. Kramer, J.P. Liu, Appl. Phys. Lett. 96 (2010) 102513. ¨ [6] J. Bauer, M. Seeger, A. Zern, H. Kronmuller, J. Appl. Phys. 80 (1996) 1667. ¨ [7] D. Goll, M. Seeger, H. Kronmuller, J. Magn. Magn. Mater. 185 (1998) 49. ¨ [8] H. Kronmuller, R. Fischer, M. Bachmann, T. Leineweber, J. Magn. Magn. Mater. 203 (1999) 12. [9] C. Rong, H. Zhang, R. Chen, S. He, B. Shen, J. Magn. Magn. Mater. 302 (2006) 126. [10] K.J. Strnat, R.M.W. Strnat, J. Magn. Magn. Mater. 100 (1991) 38. [11] N.V. Rama Rao, R. Gopalan, M. Manivel Raja, V. Chandrasekaran, D. Chakravarty, R. Sundaresan, R. Ranganathan, K. Hono, J. Magn. Magn. Mater. 312 (2007) 252. [12] J.M. Le Breton, R. Larde, H. Chiron, V. Pop, D. Givord, O. Isnard, I. Chicinas, J. Phys. D: Appl. Phys. 43 (2010) 085001. [13] O. Jagutzki, A. Cerezo, A. Czasch, R. Dorner, M. Hattass, M. Huang, V. Mergel, U. Spillmann, K. Ullmann-Pfleger, T. Weber, H. Schmidt-Bocking, G.D.W. Smith, IEEE Trans. Nucl. Sci. 49 (2002) 2477. [14] H. Bakker, Enthalpies in Alloys—Miedema’s Semi-Empirical Model, Scitec Publications Ltd, Switzerland, 1998, p. 68. [15] X.Y. Xiong, T. Ohkubo, T. Koyama, K. Ohashi, Y. Tawara, K. Hono, Acta Mater. 52 (2004) 737. [16] X.Y. Xiong, T.R. Finlayson, J. Appl. Phys. 104 (2008) 103910. [17] G. Schneider, E-Th. Henig, H.L. Lukas, G. Petzow, J. Less-Common Met 110 (1985) 159. [18] K.J. Strnat, in: E.P. Wohlfarth, K.H.J. Buschow (Eds.), Ferromagnetic Materials, vol.4, Elsevier Science Publishers B V, Amsterdam, 1988, p. 158. [19] X.Y. Xiong, T.R. Finlayson, Ultramicroscopy 107 (2007) 781.