Structural and magnetic properties of hydrides R3Fe29-xVxHy (R = Y, Ce, Nd, Sm, Gd, Tb, and Dy)
Xiu-Feng Han,Ren-Geng Xu,Xin-Hua Wang,Hongge Pan,Elisa M Baggio Baggio-Saitovitch,Terunobu Miyazaki,Fuming Yang,Changpin Chen
DOI: https://doi.org/10.1016/S0304-8853(98)00287-X
IF: 3.097
1998-01-01
Journal of Magnetism and Magnetic Materials
Abstract:A systematic investigation of crystallographic and intrinsic magnetic properties of the hydrides R 3 Fe 29− x V x H y (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) has been performed in this work. The lattice constants a, b , and c and the unit cell volume of R 3 Fe 29− x V x H y decrease with increasing rare-earth atomic number from Nd to Dy, except for Ce, reflecting the lanthanide contraction. Hydrogenation results in regular anisotropic expansions along the a -, b -, and c -axes in this series of hydrides. Abnormal crystallographic and magnetic properties of Ce 3 Fe 27.5 V 1.5 H 6.5 , like Ce 3 Fe 27.5 V 1.5 , suggest that the Ce ion is non-triply ionized. Hydrogenation leads to the increase in both Curie temperature for all the compounds and in the saturation magnetization at 4.2 K and RT for R 3 Fe 29− x V x with R=Y, Ce, Nd, Sm, Gd, and Dy, except for Tb. Hydrogenation also leads to a decrease in the anisotropy field at 4.2 K and RT for R 3 Fe 29− x V x with R=Y, Ce, Nd, Gd, Tb, and Dy, except for Sm. The Ce 3 Fe 27.5 V 1.5 and Gd 3 Fe 28.4 V 0.6 show the larger storage of hydrogen with y =6.5 and 6.9 in these hydrides. PACS 75.50 75.60.E 75.10 75.20.H 75.30.G Keywords Magnetic properties Hydride Magnetization Anisotropy field Curie temperature 1 Introduction High performance permanent magnetic materials have attracted considerable attention. Intensive investigations of the R–Fe–T phase diagrams (R=rare earth, T=transition metal as a stabilizing element) located at the iron rich corner, led to the discovery of the novel R 2 (Fe,T) 19 phase [1,2] . Its real stoichiometric composition and crystallographic structure were determined as a Nd 3 (Fe,Ti) 29 -type structure with the monoclinic symmetry [3] and the A 2 /m space group [4] by X-ray powder diffraction (XRD) and then confirmed by neutron powder diffraction [5,6] . Subsequent works on its interstitial nitrides and carbides indicated that the Sm 3 (Fe,Ti) 29 N y [7–9] compounds can be regarded as new candidates for hard magnet application, which further aroused a great upsurge in extensive studies of the R 3 (Fe,T) 29 compounds. Many members in this new families of R 3 (Fe,T) 29 compounds, such as R 3 (Fe,Ti) 29 (R=Ce, Pr, Nd, Sm [10–12] , Gd [13,14] , Tb [15] and Y [16,17] ), R 3 (Fe,V) 29 (R=Sm [18–20] , Y, Nd, Gd [18,20] , Ce, Tb, and Dy [20–22] ), R 3 (Fe,Cr) 29 (R=Ce, Nd [10,20,23] , Sm [10,19,20,24] , Gd, Tb, and Dy [20,25] ), Nd 3 (Fe,Mn) 29 [26] , R 3 (Fe,Mo) 29 (R= Ce, Nd, Sm, Gd, Tb, Dy, and Y) [27–30] , have been formed one after another with a good single-phased quality. Recently, the Gd 3 (Fe,Nb) 29 compound with Nb [31] as a stabilizing element was also obtained. The crystallographic and magnetic properties of these compounds have been studied in detail by the authors and by some researchers in other papers. At the same time, much attention has been paid to the studies of their interstitial nitrides and carbides in order to develop new hard magnets. The nitrides Sm 3 (Fe,T) 29 N y (T= Ti [7,8,32] , V [19,33] , Cr [19,24,34,35] , and Mo [30] ) and the carbides Sm 3 (Fe,T) 29 C y (T=Ti [9,32] and Cr [36] ) were considered as new potential candidates for permanent magnet application. Up to now, a little note was taken to the investigation of hydride R 3 Fe 29− x T x H y . Only the studies of Nd 3 (Fe,Ti) 29 H y [11] and Sm 3 (Fe,T) 29 H y (T=Ti, V, and Cr) [19] were reported. There is every reason to believe that the hydrides of R 3 Fe 29− x T x also have fascinating metal and magnetic properties, because the metal hydrides are interesting materials for applications in the areas of heat pumps, cooling systems, and thermal storage medium. For this reason, a systematic study on the phase formation and intrinsic magnetic properties of the hydrides R 3 Fe 29− x V x H y (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) has been performed in this work. 2 Sample preparation and experimental methods In order to prepare hydrides the alloys with the composition R 3 Fe 29− x V x (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) were prepared by arc melting the constituent elements with a purity of at least 99.9%, in an argon atmosphere. A 5–10% excess of rare-earth elements relative to the ideal stoichiometric composition was added to compensate the loss of the R during melting and annealing. The alloys were subsequently annealed in sealed quartz tubes under a protective argon atmosphere at 1183–1323 K for 1–5 days, in order to maximize the amount of R 3 Fe 29− x V x phase, and then by water quenched. The 3 : 29 single-phased character of alloys was examined by both X-ray diffraction (XRD) with Cu K α radiation and thermomagnetic analysis (TMA). Before preparing the hydride phase, the single-phased 3 : 29 alloys were pulverized into fine powders with a size of 100–1000 μm. Hydrogenation was performed by heating the fine powders in hydrogen at a pressure of 5 atm and at 573 K for R=Ce, Nd, Sm, Gd, and Tb or at 673 K for R=Y and Dy for a period of 2–4 h in a high pressure container. The XRD patterns and TMA show that each nitride of R 3 Fe 29− x V x had a single phase with the same structure as that of the parent compound, except for a little α-Fe inside. The values of the vanadium content, given in this works, in the 3 : 29 compounds are the nominal ones, but they have good reliability because all the samples are almost single-phased except from little α-Fe. For example, the nominal compositions of Nd, Fe, and V in the Nd 3 Fe 27.0 V 2.0 sample are in agreement with the results of chemical analysis. The hydrogen concentration in each hydride R 3 Fe 29− x V x H y was deduced from the atmosphere pressure difference before and after hydrogenation, according to the gas state equation PV = nRT at a constant temperature. The 3 : 29 structure is an intermediate structure between the rhombohedral Th 2 Zn 17 (2 : 17R) and tetragonal ThMn 12 (1 : 12) structures consisting of a combination of the 2 : 17R and 1 : 12 units in a ratio of 1 : 1 [3] . It can be expressed as R 2 Fe 17 + RFe 12−x T x ⇒ R 3 Fe 29−x T x . Therefore, by estimating and deducting the excess hydrogen in the form of little rare-earth hydrides under a partial decomposition of the parent compound and a little α-Fe precipitation, we can estimate the hydrogen to be absorbed at about 6 atoms per formula unit with the upper limit according to the following relations: R 2 Fe 17 H 5 + RFe 12−x T x H 1 ⇒ R 3 Fe 29−x T x H 6 . TMA was performed in a low field of 0.04 T with a vibrating sample magnetometer (VSM) from room temperature (RT) to above the Curie temperature. XRD with Cu K α radiation was used to determine the lattice parameters. The similar preparation and measuring procedure of magnetically aligned samples, in order to obtain the saturation magnetization and magnetic anisotropy field at 4.2 K and RT, was described in our previous works [20] . 3 Results and discussion As several examples, Fig. 1 shows XRD patterns with Cu K α radiation for R 3 Fe 29− x V x H y (R=Y, Ce, Tb, and Dy). The patterns shown can be quite well indexed assuming the Nd 3 (Fe,Ti) 29 -type structure with monoclinic symmetry and A 2 /m space group. As an example, the indices ( h , k , l ) of Ce 3 Fe 27.5 V 1.5 H 6.5 are given in Fig. 1 b. The peak positions of the XRD for each hydride are shifted to lower angles compared to that for R 3 Fe 29− x V x (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) [20] which demonstrated the increase of the lattice parameters after hydrogenation. The crystallographic parameters of the hydrides were obtained by fitting the XRD patterns with monoclinic symmetry and A 2 /m space group. Table 1 gives the crystallographic parameters, the unit cell volume and the X-ray density ρ derived based on the analysis of XRD data for all hydrides investigated here, together with their parent compounds. It can be seen that the lattice constants a, b , and c and the unit cell volume V of R 3 Fe 29− x V x H y decrease with increasing rare-earth atomic number from Nd to Dy, except for Ce, reflecting the lanthanide contraction. A larger fraction of the interstitial hydrogen atoms in the Gd 3 Fe 28.4 V 0.6 H 6.9 resulted in a larger expansion of its lattice constants a, b , and c and the unit-cell volume upon hydrogenation compared with that of the Sm 3 Fe 26.7 V 2.3 H 4.8 . The decrease of unit cell volume leads to an increase in density with increasing rare-earth atomic number from Nd to Dy, except for Ce. The lattice constants of a, b , and c of Ce 3 Fe 27.5 V 1.5 and Ce 3 Fe 27.5 V 1.5 H 6.5 are unusually smaller than those of R 3 Fe 29− x V x and R 3 Fe 29− x V x H y (R=Nd and Sm). This abnormal property may originate from the Ce ions being as non-triply ionized in each series of R 3 Fe 29− x V x and R 3 Fe 29− x V x H y compounds. It is noticeable that Ce 3 Fe 27.5 V 1.5 and Gd 3 Fe 28.4 V 0.6 show the maximum content of hydrogen with y =6.5 and 6.9 in this series of compounds with V as a stabilizing element. So they can be interesting candidates for further investigation in order to develop new hydrogen storage medium. Table 2 gives the expansions of the lattice parameters and the unit cell volume, the variations of the angle β and the X-ray density ρ upon hydrogenation for all hydrides compared with their parent compounds. It can be seen that the lattice expansions are different along the a -, b -, and c -axes for each hydride in this series of compounds. The regular anisotropic expansions of R 3 Fe 29− x V x H y , such as δa / a > δb / b > δc / c for R=Y, Ce, Nd, Sm, and Gd, and δa / a ≃ δb / b > δc / c for R=Tb and Dy, indicate that the expansions occur mainly in the basal c -plane rather than along the c -axis for all the hydrides. These results are in agreement with the hypothesis [19] that hydrogen atoms mainly occupy the 4i and 4f sites with A 2 /m space group (corresponding to the 4e 1 and 4e 2 sites with P2 1 /c space group [19,5] ) which are located in the basal plane referred to the 1 : 5 cell. After hydrogenation the average relative volume expansion is about 2%, except for Y and Ce in which the smaller and larger amount of the absorbed interstitial H atoms in their compounds resulted in the smallest and largest expansions of their unit-cell volumes. Fig. 2 shows thermomagnetic curves of R 3 Fe 29− x V x H y (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) measured in a low field of 0.04 T on free powder samples and by using a VSM from RT to above the Curie temperature. The Curie temperature T C was deduced from M 2 – T plots, derived from the thermomagnetic curve M ( T ), by extrapolating M to zero. Hydrogenation leads to an increase in Curie temperature T C of R 3 Fe 29− x V x . The increase in Curie temperature of each hydride reaches about 60–110 K compared with its parent compound, except for R=Y and Sm which only increases about 10 and 20 K upon hydrogenation, respectively. Fig. 3 displays the magnetic isotherms at 4.2 K for the magnetically aligned samples of R 3 Fe 29− x V x H y (R=Y(a), Ce(b), Nd(c), Sm(d), Gd(e), Tb(f), and Dy(g)) measured with a VSM. Fig. 4 shows the magnetic isotherms at RT for the magnetically aligned samples of R 3 Fe 29− x V x H y (R=Y(a), Ce(b), Nd(c), Sm(d), Gd(e), Tb(f), and Dy(g)) measured by pulsed magnetic fields (PMF) up to 12 T, together with the singular-point-detection (SPD) signal plots (d M 2 /d t 2 versus H ). The SPD technique [37] was used to measure the anisotropy field H A of the magnetically aligned samples at RT. According to the SPD theory, a singularity in the second derivative of the magnetization M with respect to the magnetic field H can be observed at the anisotropy field H A . Similarly, this method can also be used to detect a first-order magnetization process (FOMP). Table 3 gives the saturation magnetization M S at 4.2 K and RT of the hydrides R 3 Fe 29− x V x H y (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) and their parent compounds investigated here. The values of the saturation magnetization at 4.2 K and at RT were derived from the high-field data of the magnetization curves measured in the external field applied parallel to the alignment direction of the powder samples, based on the law of approach to saturation. Table 4 gives the Curie temperature T C , the saturation magnetization M S , and anisotropy field H A at 4.2 K and RT of the hydrides R 3 Fe 29− x V x H y (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) and their parent compounds. In this table, the unit of the saturation magnetization at 4.2 K and at RT were converted from A m 2 /kg into μ B /f.u. for the physical discussions. The anisotropy field H A at 4.2 K was deduced from the interception of the two magnetization curves, by extrapolating the straight line parts, measured in magnetic field applied parallel and perpendicular to the alignment direction of the powder samples. The anisotropy field H A at RT was determined from the singularities in the SPD signal plots (d M 2 /d t 2 versus H ) or the interception of the two magnetization curves measured in a magnetic field applied parallel and perpendicular to the alignment direction of the powder samples. The experimental data of the saturation magnetization at RT and the anisotropy field at 4.2 K and RT, lacked in this table, for Dy 3 Fe 27.9 V 1.1 can refer to those of Dy 3 Fe 27.7 V 1.3 [20] in the following discussions. It can be seen that introducing H as interstitial atoms resulted in improvements for all hydrides R 3 Fe 29− x V x H y in the saturation magnetization at RT compared with their parent compounds owing to the increase in Curie temperature. Hydrogenation also led to an increase in the saturation magnetization at 4.2 K for R 3 Fe 29− x V x with R=Y, Ce, Nd, Sm, Gd, and Dy, except for Tb. Hydrogenation brings about an decrease in the anisotropy field at 4.2 K and RT for R 3 Fe 29− x V x with R=Y, Ce, Nd, Gd, Tb, and Dy, except for Sm. The saturation magnetization of Y 3 Fe 27.4 V 1.6 H 2.2 at 4.2 K is 57.3 μ B /f.u. corresponding to an average Fe magnetic moment of 2.09 μ B /Fe. Hydrogenation leads to an increase of 0.28μ B /Fe in the average Fe magnetic moment compared with that in the Y 3 Fe 27.4 V 1.6 at 4.2 K. The increase in the magnetic moment of Fe sublattice for Tb 3 Fe 27.2 V 1.8 H 5.2 might be cancelled by the magnetic moment of Tb sublattice owing to the antiferromagnetic coupling between the magnetic moments of Fe and Tb sublattices, which resulted in no net increase in the saturation magnetization for Tb 3 Fe 27.2 V 1.8 H 5.2 at 4.2 K. The decreases in the anisotropy field of both hydrides Y 3 Fe 27.4 V 1.6 H 2.2 and Gd 3 Fe 28.4 V 0.6 H 6.9 at 4.2 K and RT indicate that hydrogenation decreases the anisotropy field of the Fe sublattice in the R 3 Fe 29− x V x (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) compounds. The decrease of the Fe sublattice's anisotropy field results in the reduction of the anisotropy field for the R 3 Fe 29− x V x H y compounds. The increase of the anisotropy field for Sm 3 Fe 26.7 V 2.3 H 4.8 at 4.2 K and RT may originate from the larger increase of the Sm sublattice's anisotropy field (or crystalline electric field coefficient A 0 2 ) after hydrogenation, like that in Sm 3 (Fe,Ti) 29 N y [7,38] . 4 Conclusions A systematic study on the phase formation, crystallographic properties and intrinsic magnetic properties of the hydrides R 3 Fe 29− x V x H y (R=Y, Ce, Nd, Sm, Gd, Tb, and Dy) has been performed in this work. The lattice constants a, b , and c and the unit cell volume V of R 3 Fe 29− x V x and R 3 Fe 29− x V x H y decreases with increasing rare-earth atomic number from Nd to Dy, except for Ce, reflecting the lanthanide contraction. Hydrogenation leads to regular anisotropic expansions mainly along the a - and b -axes rather than along the c -axis for each hydride. This result indicates that the hydrogen atoms mainly occupy the 4i and 4f sites which are located in the basal plane referred to the 1 : 5 cell. The lattice constants of a, b , and c of Ce 3 Fe 27.5 V 1.5 H 6.5 , like those in Ce 3 Fe 27.5 V 1.5 , are unusually smaller than that of the R 3 Fe 29− x V x H y (R=Nd and Sm). Abnormal crystallographic and magnetic properties indicate that the Ce ion is non-triply ionized. Hydrogenation leads to an increase both in the Curie temperature for each compound and in the saturation magnetization at 4.2 K and RT for R 3 Fe 29− x V x with R=Y, Ce, Nd, Sm, Gd, and Dy, except for Tb. Hydrogenation results in a decrease in the anisotropy field at 4.2 K and RT for R 3 Fe 29− x V x with R=Y, Ce, Nd, Gd, Tb, and Dy, except for Sm. An increase of 0.28 μ B /Fe in the average Fe magnetic moment is achieved for Y 3 Fe 27.4 V 1.6 H 2.2 compared with that in its parent compound at 4.2 K. The decreases in the anisotropy field of both Y 3 Fe 27.4 V 1.6 H 2.2 and Gd 3 Fe 28.4 V 0.6 H 6.9 at 4.2 K and RT indicate that hydrogenation decreases the anisotropy field of the Fe sublattice in the R 3 Fe 29− x V x compounds. The Ce 3 Fe 27.5 V 1.5 and Gd 3 Fe 28.4 V 0.6 , showing the larger content of hydrogen with y =6.5 and 6.9 in these hydrides, can be considered as interesting candidates for further investigation in order to develop new hydrogen storage media. Acknowledgements Project supported by the State Committee of Science Technology and the National Natural Science Foundation of China, and in part by the CNPq-Brazil, Japan Society for the Promotion of Science, and National Key Laboratory of Theoretical Computational Chemistry, Jilin University. References [1] S.J. Collocott, R.K. Day, J.B. Dunlop, R.L. Davis, Proc. 7th Int. Symp. on Magnetic anisotropy and Coercivity in Rare-Earth-Transition Metal Alloys, Canberra, 1992, ed. Hi-Perm Laboratory, Research Center for Advanced Mineral and Material Procession, University of Western Australia, p. 437. [2] Ye V Shcherbakova G.V Ivanova A.S Yermolenko Ye V Belozerov V.S Gaviko J. Alloys Compounds 182 1992 199 [3] H.S Li J.M Cadogan R.L Davis A Margarian J.B Dunlop Solid State Commun. 90 1994 487 [4] O Kalogiro V Psychairs L Saettas D.N Niarchos J. Magn. Magn. Mater. 146 1995 335 [5] W.B Yelon Z Hu J. Appl. Phys. 79 1996 1330 [6] Z Hu W.B Yelon O Kalogiro V Psychairs J. Appl. Phys. 80 1996 2955 [7] F.M Yang B Nasunjilegal J.L Wang J.J Zhu W.D Qin N Tang R.W Zhao B.P Hu Y.Z Wang H.S Li J. Phys. Condens. Matter 7 1995 1679 [8] B.P Hu G.C Liu Y.Z Wang B Nasunjilegal R.W Zhao F.M Yang H.S Li J.M Cadogan J. Phys. Condens. Matter 6 1994 L197 [9] B.P Hu G.C Liu Y.Z Wang B Nasunjilegal N Tang F.M Yang H.S Li J.M Cadogan J. Phys. Condens. Matter 6 1994 L595 [10] C.D Fuerst P.E Pinkerton J.F Herbst J. Appl. Phys. 76 1994 6144 [11] D.H Ryan J.M Cadogan A Margarian J.B Dunlop J. Appl. Phys. 76 1994 6150 [12] A Margarian J.B Dunlop R.K Day W Kalceff J. Appl. Phys. 76 1994 6153 [13] B Nasunjilegal F.M Yang J.J Zhu H.Y Pan J.L Wang W.D Qin N Tang B.P Hu Y.Z Wang H.S Li J.M Cadogan Acta Phys. Sin. 5 1996 544 [14] L Morellon P.A Algarabel M.R Ibarra J Kamarád Z Arnold L Pareti F Albertini A Paoluzi J. Magn. Magn. Mater. 150 1994 L285 [15] M.R Ibarra L Morellon J Blasco L Pareti P.A Algarabel J Garcı̀a F Albertini A Paoluzi J. Phys. Condens. Matter 6 1994 L717 [16] H.S Li D Courtois J.M Cadogan J.M Xu S.X Dou J. Phys. Condens. Matter 6 1994 L771 [17] F.M Yang X.F Han F.R de Boer H.S Li J. Phys. Condens. Matter 9 1997 1339 [18] Ye V Shcherbakova G.V Ivanova M.I Bartashevich V.I Khrabrov Ye V Belozerov J. Alloys Compounds 240 1996 101 [19] K Koyama H Fujii S Suzuki J. Magn. Magn. Mater. 161 1996 118 [20] X.F Han F.M Yang H.G Pan Y.G Wang J.L Wang H.L Liu N Tang R.W Zhao H.S Li J. Appl. Phys. 81 1997 7450 [21] X.F Han F.M Yang J.J Zhu H.G Pan Y.G Wang J.L Wang N Tang R.W Zhao J. Appl. Phys. 81 1997 3248 [22] D Courtois H.S Li J.M Cadogan IEEE Trans. Magn. 33 1997 3844 [23] O Kalogiro V Psychairs M Gjoka D Niarchos C.D Fuerst J. Appl. Phys. 79 1996 5539 [24] S Suzuki S Suzuki M.K Awasaki IEEE Trans. Magn. 31 1995 3695 [25] X.F Han H.G Pan H.L Liu F.M Yang Y.W Zheng Phys. Rev. B 56 1997 8867 [26] C.D Fuerst P.E Pinkerton J.F Herbst J. Magn. Magn. Mater. 129 1994 L115 [27] H.G Pan F.M Yang C.P Chen X.F Han N Tang J.F Hu J.L Wang R.W Zhao K Zhou Q.D Wang J. Magn. Magn. Mater. 159 1996 352 [28] H.G Pan F.M Yang C.P Chen X.F Han N Tang Q.D Wang J. Magn. Magn. Mater. 161 1996 177 [29] H.G Pan F.M Yang C.P Chen N Tang X.F Han Q.D Wang J. Magn. Magn. Mater. 163 1996 221 [30] H.G Pan F.M Yang Y Chen X.F Han N Tang C.P Chen Q.D Wang J. Phys. Condens. Matter 9 1997 2499 [31] Jing-Kui Liang et al., to be published. [32] K.H Müller J.B Dunlop A Handstein B Gebel P.A.P Wendhausen J. Magn. Magn. Mater. 157/158 1996 117 [33] X.F Han F.M Yang Q.S Li M.C Zhang S.Z Zhou J. Phys. Condens. Matter 10 1998 151 [34] X.F Han F.M Yang H.G Pan Y.G Wang Y.Z Wang J. Appl. Phys. 83 1998 4366 [35] Y.Z Wang B.P Hu G.C Liu H.S Li X.F Han C.P Yang J. Phys. Condens. Matter 9 1997 2787 [36] Y.Z Wang B.P Hu G.C Liu H.S Li X.F Han C.P Yang J. Phys. Condens. Matter 9 1997 2793 [37] G Asti F Bolzoni J. Appl. Phys. 58 1985 1924 [38] H.S Li D Courtois J.M Cadogan J. Appl. Phys. 79 1996 4622