Journal of Magnetism and Magnetic Materials 262 (2003) 452–457 A Mossbauer study of the spin reorientation transition in DyFe11Mo J.M Le Bretona,*, N.H Ducb, V.T Hienb, N.P Thuyb,c, J Teilleta a Groupe de Physique des Mat!eriaux, UMR CNRS 6634, Universit!e de Rouen, Site Universitaire du Madrillet, avenue de l’Universite!-B.P 12, 76801 Saint Etienne du Rouvray Cedex, France b Cryogenic Laboratory, Faculty of Physics, National University of Hanoi, 334 Nguyen Trai, Thanh Xuan, Hanoi, Viet Nam c International Training Institute for Materials Science (ITIMS), Dai Co Viet, DHBK Hanoi, Hanoi, Viet Nam Abstract The spin reorientation transition in DyFe11Mo around the spin reorientation temperature (220 K) is investigated by Mossbauer spectrometry The temperature dependence of the hyperfine parameters for each Fe site reveals an obvious discontinuity of the hyperfine field The magnitude of the discontinuity is more important for the 8f site than for the 8i and 8j sites, indicating that the most prominent contribution to the overall anisotropy in the Fe sublattice should be from the Fe ion at the 8f site This is attributed to the 3d(Fe(8f))–3d(Mo(8i)) hybridization, which may play a quite important role in R(Fe,Mo)12 compounds r 2003 Elsevier Science B.V All rights reserved PACS: 75.30.Àm; 75.50.Bb; 76.80.+y Keywords: DyFe11Mo; Spin reorientation; Mossbauer spectrometry Introduction Because they exhibit interesting magnetic properties, the R(Fe,M)12 compounds (R=rare earth; M=Ti, V, Cr, Mn, Mo, W, Al or Si) have been extensively studied [1,2] These compounds crystallize in the ThMn12 type tetragonal structure In the compounds containing R elements with a negative second–order Stevens factor aJ ; the rare-earth (4f) sublattice shows a planar anisotropy whereas the Fe (3d) sublattice has a uniaxial easy axis *Corresponding author Tel.: +2-32-95-50-39; fax: +2-3295-50-32 E-mail address: jean-marie.lebreton@univ-rouen.fr (J.M Le Breton) anisotropy The competitive anisotropy contributions from the two sublattices can induce rotations of the resultant magnetic moments with respect to the crystallographic directions, leading to spin– reorientation phase transitions as the temperature changes This is related to a change of the sign of the overall magnetic anisotropy constant at a temperature Tsr (spin reorientation temperature) This is the case, for example, for the Nd(Fe,Mo)12 [3,4], Tb(Fe,Mo)12 [4,5] and Dy(Fe,Mo)12 [5–7] compounds This phenomenon is generally evidenced by means of magnetic measurements, e.g magnetization and susceptibility measurements either on magnetically non-oriented [3,6,7] or oriented [4–6] powders However, Mossbauer spectrometry 0304-8853/03/$ - see front matter r 2003 Elsevier Science B.V All rights reserved doi:10.1016/S0304-8853(03)00077-5 J.M Le Breton et al / Journal of Magnetism and Magnetic Materials 262 (2003) 452–457 analysis can provide useful information on local properties, as the discontinuity in the hyperfine fields at the 3d sites at Tsr reflects the orbital contribution to the 3d magnetic moment at a given site [1] This information, therefore, enables one to consider the local anisotropy at each 3d site in this type of compound [8] In this paper, we present a detailed Mossbauer study of the spin reorientation transition in DyFe11Mo The magnetic properties of this compound have been reported in a previous paper [7] We focus here on the changes of the hyperfine parameters around the spin reorientation temperature Tsr ¼ 220 K [7] 453 errors s given by the fitting program [9] The error bars indicate 3sU Results and discussions The Mossbauer spectra of the DyFe11Mo powder have to be fitted with the contributions of both the ThMn12 phase and the a-Fe phase, in agreement with the results of X-ray diffraction analysis The contribution of the ThMn12 phase is fitted according to a model that accounts for both the content of Mo and the crystal structure 3.1 Fitting model Experimental A DyFe11Mo sample was prepared by arcmelting the constituents in the nominal stoichiometric composition in a protective atmosphere of pure argon (99.99%) Pure metals (Dy of 99.9%, Fe and Mo both of 99.99% purity) were used In order to ensure its homogeneity, the as–melted sample was several times turned over and melted again We have added about wt% excess of Dy to compensate the rare-earth loss caused by evaporation during the repeated melting procedure The ingot obtained was then annealed at 1000 C for 70 h in a pure argon atmosphere At the end of the annealing procedure the sample was quenched in water down to room temperature The powder was characterized by X-ray diffraction, and the pattern shows typical reflections of the ThMn12 structure, and the presence of some aFe, as an impurity phase The lattice parameters of the ThMn12 structure indicate that the Mo content in the ThMn12 phase is very close to the nominal composition [7], i.e DyFe11Mo The Mossbauer spectra were recorded in the temperature range from 77 to 300 K in transmission geometry using a 57Co source in a rhodium matrix The Mossbauer sample contains about 10 mg cmÀ2 of natural iron The isomer shift (relative to metallic a-Fe at room temperature), quadrupolar shift and hyperfine field are denoted d, e and B; respectively Estimated errors for the hyperfine parameters originate from the statistical The magnetic properties of the R–Fe compounds being mainly governed by the Fe–Fe exchange interactions, the hyperfine field at one Fe site mainly depends on both the number of Fe nearest neighbours (NN) and the corresponding interatomic distances The 8i site having 13 Fe NN, the corresponding hyperfine field B(8i) is higher than at the 8j and 8f sites, both having 10 Fe NN The Fe–Fe interatomic distances around the 8f site being weaker than those around the 8j site [2], this should result in a lower magnetic moment (and consequently, a lower hyperfine field) at the 8f site than at the 8j site, in agreement with neutron diffraction experiments [2,10] Consequently: B(8i)>B(8j)>B(8f) This order corresponds to that of the average Fe–Fe distances for ð8iÞ ð8jÞ ð8fÞ each Fe ion site (dFeÀFe > dFeÀFe XdFeÀFe ) [2] Neutron diffraction experiments showed that in R(Fe,M)12 compounds, the M atoms substitute to Fe preferentially on the 8i site [11,12] This is in agreement with the positive enthalpy contribution associated with R and Mo, the R atoms having four nearest 8i neighbours compared with eight nearest 8j and 8f neighbours [12] Consequently, it is considered here that the M atoms occupy the 8i sites only The ramdom occupancies of the Fe atoms in the ThMn12 unit cell can be calculated according to a binomial distribution, which gives the probability of finding Mo atoms in the vicinity of a given Fe site According to the Mo content, the calculation shows that 13 sextets must be used J.M Le Breton et al / Journal of Magnetism and Magnetic Materials 262 (2003) 452–457 for the Mossbauer contribution of the DyFe11Mo phase: five sextets for 8i, four sextets for 8j and four sextets for 8f This fitting procedure results in a high number of contributions, with numerous hyperfine parameters, and was not used here With the aim to look for a simpler model, with a lesser number of contributions, the distribution of environments around each Fe site is simulated by two broad sextets, having the same relative intensity, and each hyperfine parameter d; 2e or B corresponding to each Fe site is the mean value of the corresponding distribution The contribution of the ThMn12 phase is thus fitted with six sextets As an example, the contributions of the different Fe sites for the spectrum recorded at 77 K are presented in Fig The relative intensities are constrained to the values calculated from the atomic distribution of Fe and Mo atoms in the crystal structure According to the Mo content, the Mossbauer relative intensities of the different Fe sites in the DyFe11Mo phase were thus constrained to the following values: 27.2% for 8i, 36.4% for 8j and 8f The Mossbauer relative intensity of the a-Fe contribution can be deduced from the fitting of the room temperature spectrum, as its contribution is clearly distinguishable from that of the pure DyFe11Mo phase: the obtained value is 5% Thus, the contribution of a-Fe is fitted in each spectrum with a relative intensity fixed to 5% At each temperature, the contribution of the DyFe11Mo phase thus represents 95% of the intensity of the spectrum The spectra were fitted consistently in the whole temperature range according to these considerations, and the spectra are reported in Fig 3.2 Analysis of the Mossbauer data From the fittings, the hyperfine parameters of each Fe site contribution in the DyFe11Mo compound and the mean hyperfine field /BS of the DyFe11Mo compound were obtained Their temperature dependences are presented in Figs 3– In each curve, the spin reorientation temperature (Tsr ¼ 220 K [7]) is evidenced The order sequence of the magnitudes of the isomer shift is d(8i)>d(8j)Ed(8f) in the whole temperature range (Fig 3) This is in good -10 Velocity (mm/s) +10 1.00 8i site 0.97 1.00 8j site Absorption (%) 454 0.97 1.00 8f site 0.97 1.00 α-Fe 0.97 Fig Mossbauer spectrum at 77 K of the DyFe11Mo powder The contributions of the Fe sites of the ThMn12 phase and the a-Fe phase are displayed agreement with the literature [13,14] This order can be understood as the consequence of the order of the average Fe–Fe distances for each Fe ion site ð8iÞ ð8jÞ ð8fÞ (dFeÀFe > dFeÀFe XdFeÀFe ) [2] For each Fe site, a continuous decrease of the curve is observed, and no obvious discontinuity can be evidenced As the behaviour of 2e is connected with the change of angle between the easy axis of magnetisation and the electric field gradient [15], a discontinuity of the corresponding curves is expected in the region around the spin reorientation temperature From the temperature dependence of the quadrupolar shift reported in Fig 4, this discontinuity is only suggested The J.M Le Breton et al / Journal of Magnetism and Magnetic Materials 262 (2003) 452–457 -10 Velocity (mm/s) 455 0.2 +10 8i 0.15 8j 0.1 1.00 77 K δ (mm/s) 8f 0.05 -0.05 -0.1 Tsr -0.15 0.97 1.00 -0.2 50 100 150 200 250 300 350 Temperature (K) 120 K Fig Temperature dependence of the isomer shift d for each Fe site of the DyFe11Mo compound The full lines are guides for the eye 0.97 1.00 0.15 0.1 Absorption (%) 170 K 2ε (mm/s) 0.05 0.97 1.00 8i 8j -0.05 8f -0.1 -0.15 220 K -0.2 Tsr 50 100 150 200 250 300 350 Temperature (K) 0.97 1.00 Fig Temperature dependence of 2e (e is the quadrupolar shift) for each Fe site of the DyFe11Mo compound The lines (full for 8j and dotted for 8i and 8f) are guides for the eye 240 K 0.97 1.00 270 K 0.97 1.00 300 K 0.97 Fig Mossbauer spectra of the DyFe11Mo powder in the temperature range from 77 to 300 K discontinuity is not clearly evidenced probably because the value of 2e for each site is obtained from a distribution of Mossbauer sextets (related to Mo/Fe substitution effects on the 8i site) which simulates the contribution of the corresponding Fe atoms to the spectrum Each site contribution being not easily resolved from the distributions corresponding to the other sites, and 2e being treated by the fitting program as a perturbation of B; this does not allow to measure 2e with the highest possible accuracy The temperature dependence of the hyperfine field of each Fe site, and that of the mean hyperfine field of the DyFe11Mo compound are reported in Figs 5a and b, respectively The hyperfine field gradually decreases as the J.M Le Breton et al / Journal of Magnetism and Magnetic Materials 262 (2003) 452–457 456 40 35 30 B (T) 25 20 15 8i 10 8j 8f Tsr 0 50 100 (a) 150 200 250 300 350 Temperature (K) 29 27 (T) 25 23 21 Tsr 19 17 15 (b) 50 100 150 200 Temperature (K) 250 300 350 Fig Temperature dependence of: (a) the hyperfine field B at each Fe site and (b) the mean hyperfine field /BS of the DyFe11Mo compound The full lines are guides for the eye means that the most prominent contribution to the overall anisotropy in the Fe sublattice should be from the Fe ion at the 8f site (if one takes 1.7 T/ 3=0.57 T, so rather closely to 0.6 T) The fact that the Fe ion at the 8f site gives the largest contribution to the overall 3d anisotropy in the DyFe11Mo compound is contradictory to the Mossbauer results reported for RFe12ÀxTix compounds [1,2] This effect is associated with the preferential substitution of Mo for Fe at the 8i site and suggests that the 3d(Fe(8f))–3d(Mo(8i)) hybridization may be stronger than the 3d(Fe(8f))– 3d(Ti(8i)) hybridization It is worthwhile to mention that among the Fe–Fe distances around the (8i) site, the mean Fe(8f)–Fe(8i) distance is the shortest one Consequently, the 3d(Fe(8f))– 3d(Mo(8i)) hybridization would be the strongest and the density of the negative 3d(Mo) spin around Fe(8f) site would be the highest [18] As the change of hyperfine field is related to a change in the spin density, this leads to a strong reduction of B at the 8f-site Conclusion temperature increases and, on each curve, an obvious discontinuity is evidenced in the temperature region around the spin reorientation phase transition The discontinuity in the temperature dependence of both the average and the individual Fe site hyperfine field observed at the spin reorientation temperature Tsr is closely related to the second-order anisotropy constant which is determined by the residual orbital moment quenched by the crystal field [16,17] At Tsr ; the sign of the hyperfine field change is positive: DB ẳ ẵBMJc ị2BMconical ị > in case DK1 ẳ ẵK1 T > Tsr ị2K1 ToTsr ị > 0: This is in good agreement with what was found in many other reports concerning the discontinuity of the hyperfine field at Tsr [16,17] The magnitude of the discontinuity in the average (DB) and in the individual site (DBðiÞ ) hyperfine field is proportional to the overall Kl and the individual site K1ðiÞ anisotropy, respectively [16] A close inspection of the data presented in Fig shows that, DBE0:6 T whereas DBð8iÞ E0:2; DBð8jÞ E0:4; and DBð8fÞ E1:7 T: This The spin reorientation transition in DyFe11Mo around the spin reorientation temperature (220 K) was investigated by Mossbauer spectrometry, focusing on the temperature dependence of the hyperfine parameters for each Fe site No discontinuity was observed for the isomer shift A discontinuity is only suggested for the quadrupolar shift, in relation with the fitting procedure used to fit a complex Mossbauer spectrum However, the results show an obvious discontinuity of the hyperfine field, which is related to the temperature dependence of the second-order anisotropy constant The magnitude of the discontinuity, which is proportional to the individual site first-order anisotropy constant, is more important for the 8f site than for the 8i and 8j sites, indicating that the most prominent contribution to the overall anisotropy in the Fe sublattice should be from the Fe ion at the 8f site This is attributed to the 3d(Fe(8f))–3d(Mo(8i)) hybridization, which may be stronger than the 3d(Fe(8f))–3d(Ti(8i)) hybridization in R(Fe,Ti)12 compounds J.M Le Breton et al / Journal of Magnetism and Magnetic Materials 262 (2003) 452–457 Acknowledgements This work is partly supported by the Program of the Fundamental Research of Vietnam, nr 420 301 References [1] B.P Hu, H.S Li, J.P Gavigan, J.M.D Coey, J Phys.: Condens Matter (1989) 755 [2] H.S Li, J.M.D Coey, in: K.H.J Buschow (Ed.), Handbook of Magnetic Materials, Vol 6, Elsevier, Amsterdam, 1991, p [3] Y.Z Wang, B.P Hu, X.L Rao, G.C Liu, L Song, L Yin, W.Y Lai, J Appl Phys 75 (1994) 6226 [4] R Vert, D Fruchart, D Gignoux, J Magn Magn Mater 242–245 (2002) 820 [5] E Tomey, M Bacmann, D Fruchart, J.L Soubeyroux, D Gignoux, J Alloys Compounds 231 (1995) 195 [6] C.P Yang, Y.Z Wang, B.P Hu, J.L Wang, Z.X Wang, Z.L Jiang, C.L Ma, J Zhu, J Alloys Compounds 290 (1999) 144 [7] V.T Hien, J.M Le Breton, N.T Hien, L.T Tai, N.P Thuy, N.H Duc, N.P Duong, J Teillet, J Magn Magn Mater 237 (2001) 10 457 [8] N.P Thuy, J Zukrowski, H Figied, J Przewoznik, K Krop, Hyperfine Interactions 40 (1988) 441 [9] J Teillet, F Varret, 1983, unpublished MOSFIT program [10] H Fujii, H Sun, in: K.H.J Buschow (Ed.), Handbook of Magnetic Materials, Vol 9, Elsevier, Amsterdam, 1995, p 303 [11] W.B Yelon, G.C Hadjipanayis, IEEE Trans Magn 28 (1992) 2316 [12] D.B de Mooij, K.H.J Buschow, J Less-Common Met 136 (1988) 207 [13] Z.W Li, X.Z Zhou, A.H Morris, Y.C Yang, J Phys.: Condens Matter (1990) 4253 [14] Y.Z Wang, G.C Hadjipanayis, Z.X Tang, W.B Yelon, V Papaefthymiou, A Moukarika, D.J Sellmeyer, J Magn Magn Mater 119 (1993) 41 [15] P Gutlich, R Link, A Trautwein, Mossbauer Spectroscopy and Transition Metal Chemistry, Springer, Berlin, Heidelberg, New York, 1978 [16] N.P Thuy, J.J.M Franse, N.M Hong, T.D Hien, J Phys 49 (1988) 499 [17] P.C.M Gubbens, A.M van der Kraan, K.H.J Buschow, Hyperfine Interactions 40 (1989) 389 [18] N.H Duc, A Fnidiki, J Teillet, J Ben Youssef, H Le Gall, J Appl Phys 88 (2000) 1265 ... procedure The ingot obtained was then annealed at 1000 C for 70 h in a pure argon atmosphere At the end of the annealing procedure the sample was quenched in water down to room temperature The powder... considerations, and the spectra are reported in Fig 3.2 Analysis of the Mossbauer data From the fittings, the hyperfine parameters of each Fe site contribution in the DyFe11Mo compound and the mean... The ramdom occupancies of the Fe atoms in the ThMn12 unit cell can be calculated according to a binomial distribution, which gives the probability of finding Mo atoms in the vicinity of a given