Influence of Hot Compaction on Microstructure and Magnetic Properties of Mechanically Alloyed Fe(Co)-based Amorphous Compositions

Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Influence of hot compaction on microstructure and magnetic properties of mechanically alloyed Fe(Co)-based amorphous compositions J.J. Ipus1, J.M. Borrego1, J.S. Blázquez1, M. Stoica2,3, V. Franco1, A. Conde1 1 Departamento de Física de la Materia Condensada, ICMSE-CSIC, Universidad de Sevilla, P.O. Box 1065, 41080, Sevilla, Spain 2 IFW Dresden, Institute for Complex Materials, D-01069, Dresde, Germany 3 Politehnica University Timisoara, 300006, Timisoara, Romania Abstract Amorphous Fe75Nb10B15 and (Fe70Co30)75Nb10B15 alloys were prepared by mechanical alloying from the elemental constituents and hot compacted at different temperatures within the supercooled liquid region. After compaction, the microstructure studied by X- ray diffraction shows an increase in the crystalline fraction for both compositions. Magnetic properties and magnetic entropy change of compacted samples of the Co-free alloy were enhanced with respect to the powder sample, meanwhile, a deterioration in these properties was observed for the Co containing alloy after compaction. Both changes could be ascribed to an enrichment in boron in the amorphous phase. Keywords: Amorphous and nanocrystalline alloys, Mechanical alloying, Hot compaction, Magnetocaloric effect Corresponding author: Javier S. Blázquez e-mail: [email protected] Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Introduction Nanocrystalline and/or amorphous alloys have received large attention from the scientific community in the last years due to their attractive physical properties (mechanical, magnetic, etc), Particularly, Hitperm type alloys [ 1 ] were developed in 1998 to extend the good soft magnetic properties presented by Nanoperm type alloys [ 2 ] developed in the 80’s to higher temperatures, as the Co addition increases the magnetic moment and the Curie temperature of both the amorphous and the crystalline phases [ 3 ]. These systems have been commonly prepared by rapid quenching techniques. However, mechanical alloying has become an important tool in materials science as an alternative route to directly prepare these amorphous and nanocrystalline microstructures in a wider compositional range [ 4 ]. Through this technique, starting from a mixture of pure elements or prealloyed materials, it is possible to process easily and economically a large amount of powder, which is an advantage from the commercial applications point of view. For example, although Co-based bulk amorphous samples have been produced by rapid solidification techniques such as copper mold casting [ 5 , 6 , 7 ], bulk samples with shapes and dimensions that cannot be obtained by direct casting have been successfully produced by powder compaction [ 8 , 9 ]. In fact, to obtain bulk amorphous and/or nanocrystalline samples in order to cover practical needs, powder should be compacted after milling. Compaction of amorphous powders into a bulk sample preserving the initial microstructure is a primary requirement but difficult to perform. Pressing at room temperature usually leads to mechanical degradation of the compacts and high temperature compaction is needed to obtain good interparticle binding. The use of a binding material enhances the mechanical stability of the compact but generally deteriorates its magnetic properties, as the amount of magnetic material is reduced. It is worth noting that applying high temperatures during compaction could lead to an undesirable coarsening of the crystals and even crystallization of the amorphous phase. In this work, two amorphous Fe(Co)NbB alloys are prepared by mechanical alloying. Bulk samples were obtained by hot pressing within the supercooled liquid region without the use of any binding material. Microstructure and magnetic properties as a function of the compaction temperature are studied and compared to those of the asmilled powders. Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Experimental Amorphous (Fe1-XCoX)75Nb10B15 (X=0, 0.30) alloys were prepared via mechanical alloying starting from pure crystalline Fe, Nb and Co and intermetallic FeB powders in order to enhance the incorporation of hard boron into the matrix [ 10 ]. Milling was performed in a planetary mill Fritsch Pulverisette 4 Vario, with steel balls (10 mm diameter) and hardened steel vials (250 cm3), initial powder mass of 30 g and a ball to powder ratio of 10:1. The rotational speed of the disc which supports the vials was 350 rpm and that of the vials was 700 rpm in opposite direction. Powder samples were taken out to study their microstructure, thermal stability and magnetic properties after 20 and 40 h milling. Disk-shaped samples 10 mm diameter and 4 mm thick were prepared by compaction of powders using an uniaxial hot press under a vacuum of 10-3 Pa. The powder was heated at 0.67 K s-1 up to the consolidation temperature, compacted under a pressure of 700 MPa for 1.5 min and then cooled down to room temperature using Ar flow. Density of the compacted samples, obtained from Archimedes method, shows, for (Fe0.7Co0.3)75Nb10B15 alloy, an increase from 5.30.4 g/cm3 for the sample compacted at 753 K to 6.90.5 g/cm3 for sample compacted at higher temperatures. In the case of the Co-free alloy, density increases from 6.10.4 g/cm3 for the sample compacted at 773 K to 6.90.5 g/cm3 for sample compacted at higher temperatures. This indicates about 90 % of the expected density for a bulk amorphous sample of these compositions. Microstructural characterization was studied by X-ray diffraction using Cu-Kα radiation in a Bruker D8I diffractometer and scanning electron microscopy using a Zeiss Auriga Crossbeam Focused Ion Beam (FIB)-Scanning Electron Microscopy (SEM). SEM images were taken on in situ polished surfaces using Ga ion beam in FIB mode. Thermal characterization of the samples was studied by differential scanning calorimetry (DSC) using a Perkin Elmer DSC7 at a heating rate of 20 K/min in a continuous Ar flow. Magnetic properties were studied using a Lakeshore 7407 vibrating sample magnetometer with a maximum applied magnetic field of 1.5 T. Isothermal magnetization curves were measured on parallelepipeds of about 1x1x3 mm in size in a temperature range from 100 to 753 K with increments of 10 K. Spontaneous specific magnetization, σ0, was obtained from the extrapolation of high field magnetization curves to zero field. Curie temperature, Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 TC, was calculated as the inflexion point of σ0(T) curves. Hysteresis loops as a function of temperature were recorded in a temperature range of 100-753 K with steps of 50 K. Isothermal magnetic entropy change, ΔSM, due to the application of a magnetic field, was calculated by processing the temperature and field dependent magnetization curves using a numerical approach to the expression: 2 1 0 H M H H M S dH T         , (1) where M is the magnetization and T the temperature and H1 and H2 are the minimum and maximum applied fields. The partial derivatives are replaced by finites differences and the integration is performed numerically. This procedure was performed with the help of the Magnetocaloric Effect Analysis Program [ 11 ]. The bulk samples were cut into small parallelepipeds and its longest axis was aligned with the applied field to minimize the demagnetizing field. In the case of powder samples the unavoidable demagnetizing factor, ND, is considered as 1/3 [ 12 ] for loosely packed spherical samples, in order to correct the isothermal magnetization curves as: app D H H N M , (2) where Happ is the applied magnetic field and H is the internal field. The refrigerant capacity represents a measure of the energy transferred between the hot and cold reservoirs. In this study the refrigerant capacity, RCFWHM, was calculated as the product of the peak magnetic entropy change times the full width at half maximum of the peak. Results and Discussion The microstructural evolution of powder samples was followed by XRD. Figure 1 shows the XRD patterns of samples after 20 and 40 h milling for both studied compositions, along with the deconvolution (a Gausssian profile to describe the amorphous halo and a Lorentzian profile for the nanocrystalline α-Fe type phase) used to extract the fraction of phases from the patterns when possible. It is shown that, after the formation of the bcc Fe-Co(NbB) supersaturated solid solution, the development of an amorphous phase takes place and its fraction increases as the crystalline one decreases. Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 An almost completely amorphous material is obtained for Fe75Nb10B15 alloy after 20 h milling and for the (Fe0.7Co0.3)75Nb10B15 alloy after 40 h. DSC records presented in figure 2a show the thermal stability of the powder samples after 40 h for both compositions. Different processes can be observed in the scans. At low temperatures, a gas desorption process starting at ~390 K followed by a broad structural relaxation process at ~500 K. Both features are typically observed in powder samples [ 13 ]. Although it is not fully accepted the existence of glass transition for amorphous systems not obtained by rapid quenching, an exothermic feature is observed that could be named as Tg (inset of Fig. 2a) and estimated at Tg ~760 and 785 K for (Fe0.7Co0.3)75Nb10B15 and Fe75Nb10B15 alloys, respectively, and the onset temperature of the main crystallization process was measured at Ton ~815 and 850 K for (Fe0.7Co0.3)75Nb10B15 and Fe75Nb10B15samples, respectively. In order to perform the hot compaction of the powder samples, one of the targets was to retain the previously observed microstructure of the initial powder. Therefore, temperatures close but below the crystallization temperatures were selected for compaction: 753, 773 and 793 K for (Fe0.7Co0.3)75Nb10B15 alloy and 773, 793, 803 and 813 K for Fe75Nb10B15 alloy. A compromise has been taken between the heating rate of the hot-press (limited to 0.67 K/min) and the DSC heating rate needed to get a significant signal to noise ratio. A range of temperatures has been explored to minimize the influence of the different rates on the conclusions derived. After compaction, the microstructure of bulk samples was examined by XRD and the same deconvolution procedure used for the case of as-milled powder was applied to the compacted samples, as shown in figure 1. Lattice parameter does not present significant changes with compaction (a=2.8780.005 Å for both compositions). All values are higher than that of pure bcc-Fe. This fact could be explained by the influence of Nb atoms into the lattice. [ 14 , 15 ]. After hot compaction, whereas for the Fe75Nb10B15 sample XC increases up to ~0.10 regardless compaction temperature, for the (Fe0.7Co0.3)75Nb10B15 sample XC continuously increases with temperature, reaching ~0.16 for the sample compacted at 793 K. This increment in XC is in agreement with the closer proximity of the compaction temperature to the crystallization temperature in (Fe0.7Co0.3)75Nb10B15 (T =22 K) than in Fe75Nb10B15 alloy (T =37 K). Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 DSC scans showing the thermal stability of compacted samples are presented in figures 2b-c. From these scans we can see that the low temperature processes, T <750 K, observed in powder samples and ascribed to microstuctural relaxation, have disappeared for all compacted samples. Moreover, no important changes in the peak temperature are observed after compaction for Fe75Nb10B15 alloy. For the (Fe0.7Co0.3)75Nb10B15 alloy, whereas for samples compacted at 753 and 773 K no significant changes are observed in the peak temperatures, an increase of ~10 K is detected for the sample compacted at 793 K. This change in the crystallization temperature for the (Fe0.7Co0.3)75Nb10B15 alloy could be ascribed to a partial consumption of the transformation due to crystallization of the bcc-(FeCo) type phase. Moreover, small compositional changes in the amorphous phase (e.g. an enrichment in boron content) during compaction at the highest temperature could also lead to an increase in the crystallization temperature of the FeCo-based amorphous alloys [ 16 ]. Coercivity, HC, at room temperature shows a reduction after compaction from ~2 and ~4 kA/m for Fe75Nb10B15 and (Fe0.7Co0.3)75Nb10B15 powders, respectively, down to ~1 kA/m for both compositions. This drop in HC could be ascribed to relaxation phenomena in the samples due to annealing during compaction. Spontaneous specific magnetization values of compacted samples present different behaviors for each alloy. For the Fe75Nb10B15 alloy at room temperature a continuous increase from ~46 emu/g (powder) up to ~54 emu/g (compacted at 813 K) with compaction temperature is observed. However, for the (Fe0.7Co0.3)75Nb10B15 alloy no important changes in σ0 are observed at room temperature. The temperature dependence of σ0 is presented in figure 3 for the different samples. It is observed that TC for the (Fe0.7Co0.3)75Nb10B15 alloy does not change after compaction at 753 and 773 K with respect to the value observed for the powder sample, with TC =563 K. However, a reduction down to TC =533 K for the sample compacted at 793 K is observed. For the Fe75Nb10B15 samples, TC increases from 320 K, for the powder sample, up to 340 K for each compacted sample from 773 K to 803 K, but a smaller increase (TC =330 K) is observed for the sample compacted at 813 K. These features could be related to compositional changes (a better dissolution of boron) in the amorphous matrix during the compaction. In fact, an increase in boron content in amorphous Fe-B leads to an increase in TC of ~20 K/at.% B [ 17 ], as well as an increase in σ0 [ 18 ]. In the case of Co-B amorphous alloys, a decrease of ~30 K/at.% B in TC is reported [ 19 ]. Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Figure 4 shows the temperature dependence of coercivity for Fe75Nb10B15 sample compacted at 813 K. All samples exhibit a similar behavior. An initial increase of the coercivity can be seen as the measurement temperature reaches TC and then a decrease is observed for higher temperatures. The initial trend can be explained by a suppression of the exchange coupling in between the remaining nanocrystals as the amorphous matrix become paramagnetic. Above the Curie temperature of the amorphous phase, the gradual disappearance of the dipolar coupling between nanocrystals causes a reduction in coercivity as they approach the superparamagnetic regime as temperature increases [ 20 ]. The temperature dependence of the magnetic entropy change, ΔSM, for a magnetic field change of 1 T is shown in figure 5 for all studied samples. It can be seen that, after compaction, a ~15 K increase of the peak temperature takes place for the Fe75Nb10B15 alloy. For the (Fe0.7Co0.3)75Nb10B15 alloy the peak temperature shows a reduction of ~20 K only for the sample compacted at the highest temperature. The peak temperature, Tpk, as a function of the compacted one, Tcomp, for both alloys is presented in figure 6a. The shift of the peak temperature for both alloys agrees with the shift observed in TC and could be ascribed to the increase in B content in the amorphous phase during hot temperature compaction. In fact, an incorporation as small as ~0.7 at.% B in the amorphous matrix could explain the observed shifts of the peak temperature for both compositions as reported in [17,18]. The evolution of the peak value of ΔSM and refrigerant capacity as a function of compaction temperature are presented in figure 6 (b and c). A continuous reduction in the peak value of |ΔSM| for compacted (Fe0.7Co0.3)75Nb10B15 samples with respect to the asmilled one can be observed. However, for the case of Fe75Nb10B15 alloy an increase of ~13% in the peak value of ΔSM for compacted samples is observed. The continuous reduction for the (Fe0.7Co0.3)75Nb10B15 alloy after compaction could be ascribed to the progressive crystallization during hot compaction. In the other case, the increase observed for the Fe75Nb10B15 samples after compaction can be explained by the effect of boron incorporation in the amorphous matrix which enhances the magnetic moment in Fe-based amorphous alloys [10, 21 ], in agreement with the increase of TC. The refrigerant capacity shows two different behaviors for the two different compositions. In the case of Fe75Nb10B15 alloy this value decreases with respect to that of powder sample, which is related to the sharpening of the ΔSM curve due to a reduction of Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 the TC distribution with increasing compaction temperature, as normally observed for powder samples [ 22 , 23 , 24 ]. For the (Fe0.7Co0.3)75Nb10B15 alloy significant changes in RC are not initially observed but a large reduction is measured after compaction at 793 K, which can be explained by the reduction in the peak value of SM due to the partial crystallization of this sample. There are two processes which cause B enrichment of the amorphous phase: crystallization of the α-Fe type phase and dissolution of the remnant boron inclusions. In the former case, after a 10 % crystallization the composition of the residual amorphous phase, neglecting the presence of B inclusions, should be (FeCo)72Nb11B17, which could reasonably account for the boron enrichment. However, previous studies on Fe-Zr-B [ 25 ] report on the dissolution of boron inclusions after high temperature treatment. In order to consider this mechanism, SEM micrographs were obtained on cross sections of compacted samples after being polished using a beam of Ga ions. Figure 7 shows these SEM images at the lowest and highest compaction temperatures for Fe75Nb10B15 (Fig. 7a and 7b) and (Fe0.7Co0.3)75Nb10B15 alloy (7c-7d), respectively. B inclusions are still clearly observed as dark spots (regions of low interaction with electrons due to the low atomic number of B atoms) in the compacted samples, even at the highest compaction temperature, and not significant changes are evident with respect to the samples compacted at lower temperatures. However, slight boron incorporation from the inclusions, below our resolution, cannot be discarded. Conclusions Mechanically alloyed Fe75Nb10B15 and (Fe0.70Co0.30)75Nb10B15 powder compositions were hot compacted at different temperatures and the evolution of the microstructural and magnetic properties were studied. Around 10 % crystalline fraction was developed after compaction within the supercooled liquid region for Fe75Nb10B15 samples. For the (Fe0.70Co0.30)75Nb10B15 samples a continuous increase up to XC ~16% was observed. The magnetic properties show changes with the compaction temperature for both compositions. Small increases in the room temperature spontaneous specific magnetization, the Curie temperature and the peak value of the magnetic entropy change Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 for the Fe75Nb10B15 alloy were observed. For the case of (Fe0.7Co0.3)75Nb10B15 alloy a reduction in TC and the peak value of the magnetic entropy change was evidenced, while σ0 is almost constant. In both cases, changes in TC could be explained after a small boron enrichment of the amorphous phase during compaction. Crystallization of the α-Fe type phase can explain the enhancement of σ0 in Fe75Nb10B15 alloy as well as its constant value in (Fe0.7Co0.3)75Nb10B15 alloy while the peak value of |SM| decreases. Acknowledgements This work was supported by the Spanish Ministry of Science and Innovation (MICINN) and EU FEDER (project MAT2013-45165-P) and the PAI of the Regional Government of Andalucía. References [ 1 ] M.A. Willard, D.E. Laughlin, M.E. McHenry, D. Thoma, K. Sickafus, J.O. Cross, V.G. Harris, J. Appl. Phys. 84 (1998) 6773-6677. [ 2 ] K. Suzuki, N. Kataoka, A. Inoue, A. Makino, T. Matsumoto, Mater. Trans. JIM 31 (1990) 743-746. [ 3 ] M.E. McHenry, M.A. Willard, D.E Laughlin, Prog. Mater. Sci. 44 (1999) 291- 433. [ 4 ] C. Suryanarayana, Prog. Mater. Sci. 46 (2001) 1-184. [ 5 ] A.H. Taghvaei, M. Stoica, K.G. Prashanth, J. Eckert, Acta Mater. 61 (2013) 6609-6621. [ 6 ] K. Biswas, S. Ram, S. Roth, L. Schultz, J. Eckert, J. Mater. Sci. 41 (2006) 3445- 3450. [ 7 ] A. Inoue, Acta Mater. 48 (2000) 279-306. [ 8 ] P. Kollar, J. Bednarcik, S. Roth , H. Grahl, J. Eckert, J. Magn. Magn. Mater. 278 (2006) 373-378. [ 9 ] J. Fuezer, J. Bednarcik, P. Kollar, S.Roth, J. Magn. Magn. Mater. 316 (2007) E834-E887. [ 10 ] J.J. Ipus, J.S. Blázquez, C.F. Conde, J.M. Borrego, V. Franco, S. Lozano-Pérez, A. Conde, Intermetallics 49 (2014) 98-105. Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Figure 5. 100 200 300 400 -0.6 -0.4 -0.2 0.0 Fe75Nb10B15 powder 773 K 793 K 803 K 813 K SM (J kg-1K-1) T (K) 400 500 600 700 -0.8 -0.6 -0.4 -0.2 0.0 (Fe70Co30)75Nb10B15 powder 753 K 773 K 793 K SM (J kg-1K-1) T (K) Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Figure 6. 0.4 0.6 0.8 |SM pk| (J kg-1K-1) (b) 293 740 760 780 800 820 40 50 60 70 (c) TComp (K) |RCFWHM| (J kg-1) 320 340 540 570 (a) Tpk (K) Journal of Alloys and Compounds, 653 (2015) 546-551 http://dx.doi.org/10.1016/j.jallcom.2015.09.074 Figure 7.