3B2v7:51c ED:chanakshi bg GML4:3:1 MAGMA : 8454 Prod:Type:com pp:123ðcol:fig::NILÞ PAGN: bmprakash SCAN: mangala ARTICLE IN PRESS 1 3 Journal of Magnetism and Magnetic Materials ] (]]]]) ]]]­]]] 5 7 Observation of the bulk spin-flop in an Fe/Cr superlattice 9 L. Botty!ana,*, L. De!aka, J. Dekosterb, E. Kunnenc, G. Langoucheb, 11 J. Meersschautb, M. Majora,b, D.L. Nagya, H.D. R.uterd, E. Szil!agyia, K. Temstc 13 a KFKI Research Institute for Particle and Nuclear Physics, P.O. Box 49, 1525 Budapest, Hungary b Instituut voor Kern-en Stralingsfysica, K.U. Leuven, Celestijnenlaan 200D B-3001, Leuven, Belgium c 15 Laboratorium voor Vaste-Stoysica en Magnetisme, K.U. Leuven, Celestijnenlaan 200C B-3001, Leuven, Belgium d II. Institut f.ur Experimentalphysik Universit.at Luruper Chaussee 149 D-22761, Hamburg, Germany 17 19 Abstract 21 The layer magnetisation reorientation transition (spin-flop, SF) was studied in an artificial layer antiferromagnet (AF), namely in MgO(0 0 1)/[57Fe(2.6 nm)/Cr(1.3 nm)]20 epitaxial superlattice (SL) by synchrotron M.ossbauer 23 reflectometry and Kerr effect (SMOKE). The SF occurs simultaneously in the entire SLstack (bulk SF) in an increasing field of HSF ¼ 13 mT along the easy direction parallel to the layer magnetisations. It is recognised by the kink 25 in the SMOKE loop and by the sharp up-rise of the AF Bragg peak in the delayed M.ossbauer reflectivity. The moderate value of observed HSF is compared with estimations from a spin-chain model and interpreted as due to intraplane 27 domain-wall motion during SF. r 2001 Published by Elsevier Science B.V. 29 Keywords: Artificial superlattices; Interlayer coupling; Kerr measurements; M.ossbauer spectroscopy; Synchrotron radiation 31 33 An interesting model system of an `artificial layer function of the angle of grazing incidence Y: Structural 57 35 antiferromagnet' is a periodic Fe/Cr antiferromagnetic Bragg peaks due to the electronic SLperiodicity are (AF) superlattice (SL) with even number of Fe layers. observed in the prompt and in the delayed signal, but the 59 37 When the external magnetic field is aligned along the magnetic (hyperfine) super-cell doubling in an AF SL easy axis of the Fe layers parallel/antiparallel to the appears only in the delayed TISMR. The AF Bragg- 61 39 magnetizations Mkðk ¼ 1; 2; 2nÞ; the anisotropy-stabi- peak intensity in TISMR is at maximum for the lised configuration becomes energetically unfavourable photon's wave vector k; parallel/antiparallel to Mk; 63 41 at a certain critical in-plane field strength and a sudden and zero for k>Mk [3]. Therefore, SMR is especially magnetisation reorientation is expected in a finite suitable for studying the spin-flop (SF) phenomena. 65 43 multilayer stack [1­4] with surface spin-flop [5,6] or bulk Here, we report on TISMR of the (bulk) SF in a Fe/Cr spin-flop (BSF) [7] scenarios, in the cases of uniaxial and AF SLwith a four-fold in-plane anisotropy. The 67 45 four-fold in-plane anisotropy, respectively. observed HSF is compared with a spin-chain calculation Synchrotron M.ossbauer Reflectometry (SMR, [8­11]) with the aim of elucidating the magnetisation reorienta- 69 47 is sensitive to the alignment of local hyperfine fields in tion mechanism. the film. Consequently, in an 57Fe-containing magnetic The [57Fe(2.6 nm)/Cr(1.3 nm)]n (n ¼ 20) periodic mul- 71 49 SL, the 73 51 75 53 *Correspo UNCORRECTED PROOF Fe-layer magnetisation directions can be tilayer was grown on a MgO(0 0 1) substrate at 450 K by determined relative to the photon's propagation and MBE using an electron beam gun (Cr) and a Knudsen polarisation vectors [8,9]. Time integral (TI) SMR cell (57Fe) at a base pressure of 1 10 9 mbar following records the total number of delayed photons as a a degassing of the substrate at 873 K for 30 min. RHEED patterns and high-angle X-ray diffractograms confirmed the epitaxial quality and excellent layering nding author. Tel.: +36-1-392-2761; fax: +36-1- 77 55 of the SLfilm. L ow-angle X-ray diffraction at 395-9151. E-mail address: battyan@rmki.kfki.hu (L. Botty!an). l ¼ 0:086 nm (Fig. 1a) showed extended Kiessig-fringes 0304-8853/01/$ - see front matter r 2001 Published by Elsevier Science B.V. PII: S 0 3 0 4 - 8 8 5 3 ( 0 1 ) 0 0 8 9 4 - 0 MAGMA : 8454 ARTICLE IN PRESS 2 L. Botty!an et al. / Journal of Magnetism and Magnetic Materials ] (]]]]) ]]]­]]] 1 In the SL, Mk points parallel or anti-parallel to either 57 of the Fe[0 1 0] or Fe[1 0 0] easy axes in the film plane, 3 with AF domains oriented at random, parallel/antipar- 59 allel with either of those (inset (1) in Fig. 1). The initial 5 magnetic state was carefully prepared by aligning Mk 61 along a single easy axis: the SLfilm was magnetised to 7 46 mT in order to induce an SF (inset (2)). Then, the 63 field was decreased to 3 mT and the sample was rotated 9 through p=4: At this point, TISMR scan 1=b was 65 recorded. Since in this state (inset (3)), k>Mk; no AF 11 superreflections were observed (Fig. 1b, inset (4)). 67 Having increased the field to 35.3 mT, the 12-order AF 13 Bragg peak appeared (Fig. 1c) as a direct evidence of the 69 BSF. Due to the four-fold anisotropy, this state was 15 preserved when the magnetic field was decreased again to 71 3 mT (inset (5)). Accordingly, the 12-order AF Bragg- 17 peak intensity did not change (Fig. 1d). 73 The BSF transition was also confirmed by SMOKE 19 (Fig. 2). High-field loops (see inset in Fig. 2) were 75 indicative of AF coupling and a saturation field of 21 HSE0:9 T. First, the Mk were prepared in an easy 77 direction of Fe (by exerting and releasing a saturating 23 field), afterwards, the sample was rotated through p=4: 79 A kink was observed in the first loop around 25 HSF ¼ 13 mT, which did not re-occur until the sample 81 was turned to the perpendicular direction. This is in full 27 agreement with the TISMR scans. A difference between 83 SMOKE and TISMR is that the latter probes the entire 29 Fig. 1. Prompt (a) and TISMR (b­d) scans of MgO(0 0 1)/ multilayer stack at the AF Bragg angle, while the former 85 [57Fe(2.6 nm)/Cr(1.3 nm)]20 superlattice taken in (b): 3 mT, (c): remains more sensitive to the upper layers. The 31 35.3 mT and (d): repeated 3 mT magnetic fields, respectively. agreement indicates that the SF reorientation occurs 87 The appearance of the 12-order AF Bragg peak shows the simultaneously in the entire SLstack (bulk SF). 33 reorientation of the layer magnetisations at a bulk SF transition In order to relate the layer parameters to the 89 field HSF: The system of co-ordinates in the insets is fixed to the measured HSF; for simplicity, an infinite `two-sublattice' 35 substrate. spin-chain scheme is invoked. The energy E per 91 unit area of a SLwith quadratic anisotropy (experi- 37 mentally found for Fe/Cr on MgO(0 0 1)) in an external 93 and structural SLreflections up to the third order (not 39 shown) with a bilayer period of 3.9 nm and root-mean- 95 square interface roughness of 0.43 nm. The Fe/Cr 41 thickness ratio was determined by Rutherford back- 3 97 scattering. Conversion electron M.ossbauer spectroscopy 10 2 5 43 revealed an in-plane orientation of the Fe moments, an 0 99 expected consequence of the shape anisotropy. 1 -5 57 45 Fe SMR experiments were performed on the BW4 -10 0 -1.0 -0.5 0.0 0.5 1.0 101 nuclear resonance beamline in HASYLAB, Hamburg, at H (T) Hsf 47 room temperature in vertical scattering geometry. -1 103 Motorized permanent magnets provided horizontal up Kerr angle (mdeg) down 49 fields between -2 up again 105 -3 51 -30 -20 -10 0 10 20 30 107 H (mT) 53 UNCORRECTED PROOF 3 and 46 mT perpendicular to k: TISMR scans were recorded at grazing angles between 0 and 20 mrad. The SMR results are shown in Fig. 1. The solid lines in (a)­(d) are simulations [9,12]. Peaks labelled `0', `12' and `1' are the total reflection peak [13,14], the AF Fig. 2. Surface magneto-optical Kerr loops of a superlattice 109 Bragg peak and the structural Bragg peak, respectively. MgO(0 0 1)/[57Fe(2.6 nm)/Cr(1.3 nm)]20. SF occurs around 55 The presence or absence of the 12 peak reveals if k8Mk or HSF ¼ 13 mT only once following a 901 rotation of the 111 k>Mk: substrate relative to the field direction. MAGMA : 8454 ARTICLE IN PRESS L. Botty!an et al. / Journal of Magnetism and Magnetic Materials ] (]]]]) ]]]­]]] 3 1 field H is HPRI-CT-1999-00040), the Flemish-Hungarian bilateral 43 EðH; W Project No. BIL98/20 and Project No. T 029409 of the 1; W2Þ ¼ J1 cosðW1 W2Þ J2 cos2ðW1 W2Þ 3 Hungarian Scientific Research Fund (OTKA) is grate- 45 þ Aðsin2 2W1 þ sin2 2W2Þ fully acknowledged. J.M. and K.T. are Post-Doctoral 5 HMðcos W1 þ cos W2Þ; ð1Þ Fellows of the Flemish FWO. 47 where J1; J2 and A are the bilinear and biquadratic 7 coupling coefficients between the two sublattices and the 49 magneto-crystalline energy (J1; J2o0; A > 0). The bulk 9 References anisotropy energy K 51 1 ¼ 4A=ttot Fe ; with ttot Fe and M being the Fe sublattice layer thickness (i.e. 26 nm in the present [1] F.C. N.otermann, R.L. Stamps, A.S. Carri-co, R.E. 11 case), and moment per unit area M ¼ njM 53 kj=2; k ¼ 1; 2: Camley, Phys. Rev. B 46 (1992) 10847. M1 and M2 decline by W1 and W2; respectively, from the [2] M. Major, L. Botty!an, L. De!ak, D.L. Nagy, in: E.A. 13 field (the latter pointing along an easy direction of Fe). G.orlich, A. Pedziwiatr (Eds.), Proceedings of the XXXIV, 55 The SF and saturation occur in increasing and decreas- Zakopane School of Physics, Jagellonian University, 15 ing fields, at which the energy given by Eq. (1) is no Cracow, 1999, p. 165. 57 longer positive-definite. The respective field values are [3] M. Major, Master's Thesis, E.otv.os Lor!and University, ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi p ffiffiffiffiffiffiffiffiffiffiffiffi 17 H0 Budapest, 1999 (in Hungarian). 59 SF ¼ 4 Að4A J1 þ 2J2Þ=M and HS ¼ 2ðJ1 þ 2J [4] A.L. Dantas, A.S. Carri-co, Phys. Rev. B 59 (1999) 1223. 2 þ 4AÞ=M: (As shown by Dantas and Carri-co [4], 19 H [5] R.W. Wang, D.L. Mills, E.E. Fullerton, J.E. Mattson, 61 S remains unaffected, while the SF field H0SF is lowered S.D. Bader, Phys. Rev. Lett. 72 (1994) 920. in a finite SLstack due to the dangling surface layers. [6] N.S. Almeida, D.L. Mills, Phys. Rev. B 52 (1995) 13504. 21 For a strongly AF-coupled finite SL( J 63 1 þ 2J p 2bA), ffiffiffi [7] K. Temst, E. Kunnen, V.V. Moshchalkov, H. Maletta, this lower value of the SF field is HSFEH0SF= 2: This H. Fritzsche, Y. Bruynseraede, Physica B 276­278 (2000) 23 latter H 65 SF value is considered in the following estima- 684. tions.) From the SMOKE loops, HSE0:9 T. Using [8] D.L. Nagy, L. Botty!an, L. De!ak, E. Szil!agyi, H. Spiering, 25 this and literature value of K 67 1 ¼ 4:5 kJ/m3 [15], HSF J. Dekoster, G. Langouche, Hyperfine Interactions 126 was calculated. Assuming a pure bilinear coupling (2000) 349. 27 (J [9] L. De!ak, L. Botty!an, D.L. Nagy, H. Spiering, Phys. Rev. 69 2=J1 ¼ 0), this gives Hcalc SF ¼ 260 mT: Allowing for a variation of 0oJ B 53 (1996) 6158. 2=J1o0:45; a range of Hcalc SF was 29 estimated. For K [10] R. R.ohlsberger, Hyperfine Interactions 123/124 (1999) 71 1 ¼ 4:7 kJ/m3, Hcalc SF >130 mT. Varying 455. K1 in a range as broad as 2.4 kJ/m3oK1o4.7 kJ/m3, [11] A.I. Chumakov, D.L. Nagy, L. Niesen, E.E. Alp, 31 Hcalc 73 SF remains by a factor of 5 above the measured value. Hyperfine Interactions 123/124 (1999) 427. These facts imply that, as expected, rather than by [12] H. Spiering, L. De!ak, L. Botty!an, Hyperfine Interactions 33 coherent rotation of the sublattice magnetisations, the 125 (2000) 197. 75 SF is likely to occur by intralayer domain wall motion in [13] A.Q.R. Baron, J. Arthur, S.L. Ruby, A.I. Chumakov, 35 this artificial layer antiferromagnet. The latter requires G.V. Smirnov, G.S. Brown, Phys. Rev. B 50 (1994) 10354. 77 much lower field to overcome the anisotropy barrier. [14] L. De!ak, L. Botty!an, D.L. Nagy, Hyperfine Interactions 92 37 The balance between the Zeeman energy and the (1994) 1083. 79 anisotropy energy at the SF field was found to be [15] H.-P. Klein, E. Keller, Phys. Rev. 144 (1966) 372. 39 essential in shaping the AF domain structure [16]. [16] D.L. Nagy, L. Botty!an, B. Croonenborghs, L. De!ak, 81 B. Degroote, J. Dekoster, H.J. Lauter, V. Lauter­Pasyuk, O. Leupold, M. Major, J. Meersschaut, O. Nikonov, 41 Support by the IHP Programme `Access to Research 83 A. Petrenko, R. R.uffer, H. Spiering, E. Szil!agyi, Phys. Rev. Infrastructures' of the European Commission (Contract Lett, submitted for publication. UNCORRECTED PROOF