- Email: [email protected]
Pt multilayer structure
Accepted Manuscript Tuning Magnetic Properties of Non-Collinear Magnetization Configuration in Pt/[Pt/Co]6/Pt/Co/Pt Multilayer Structure Taner Kalaycı, Caner Deger, Salih Akbulut, Fikret Yildiz PII: DOI: Reference:
S0304-8853(17)30067-7 http://dx.doi.org/10.1016/j.jmmm.2017.04.008 MAGMA 62610
To appear in:
Journal of Magnetism and Magnetic Materials
Received Date: Revised Date: Accepted Date:
10 January 2017 2 April 2017 5 April 2017
Please cite this article as: T. Kalaycı, C. Deger, S. Akbulut, F. Yildiz, Tuning Magnetic Properties of Non-Collinear Magnetization Configuration in Pt/[Pt/Co]6/Pt/Co/Pt Multilayer Structure, Journal of Magnetism and Magnetic Materials (2017), doi: http://dx.doi.org/10.1016/j.jmmm.2017.04.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Tuning Magnetic Properties of Non-Collinear Magnetization Configuration in Pt/[Pt/Co]6/Pt/Co/Pt Multilayer Structure Taner Kalaycı a, Caner Deger a, Salih Akbulut b, and Fikret Yildiz b,* a
b
Department of Physics, Marmara University, 34722, Kadıköy, Istanbul, Turkey Department of Physics, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey
Abstract In this study, effects of Pt spacer and Co reference layers thickness in [Co/Pt]6/Pt/Co multilayer have been revealed to tailor magnetization directions in non-collinear configuration. Magneto-optic Kerr effect and ferromagnetic resonance techniques were employed to investigate magnetic properties. Bilinear coupling between [Co/Pt]6 and Co layers and anisotropy constants were determined by a micromagnetic simulation based on metropolis algorithm. 3 nm spacer causes ferromagnetic coupling while the samples have 4 and 5 nm spacer are coupled anti-ferromagneticaly. Also, tuning magnetic anisotropy of [Co/Pt]6 layer was accomplished by Co reference layer. It is revealed that controlling of noncollinear states in such systems is possible by variation of thickness of spacer and reference layers and [Co/Pt]6/tPt/tCo trilayer system can be used in multilayered magnetic systems. Keywords: Non-collinear magnetic configuration, interlayer coupling, magnetic anisotropy, spin transfer torque *
Corresponding Authors:
E-mail addresses: [email protected], [email protected]
1. Introduction Ferromagnetic (FM) layers separated by a nonmagnetic metallic spacer layer are attracted lots of attention since they are widely employed in spintronic applications such as magnetic recording and sensor technologies [1]. The electrical resistivity of this kind of structures can be manipulated under magnetic field since it depends on the spin direction. Correlatively, Spin Transfer Torque (STT) effect is a mechanism that can control and manipulate the magnetism of a ferromagnet by current. So, STT is an alternative way uses current instead of Oersted field. Also, spin torque nano-oscillator (STNO) is a promising candidate for future microwave generators since it can be tuned by current and field. The spin direction of FM layers can be parallel or antiparallel to each other due to nonmagnetic spacer layer thickness in magnetic multilayer systems. This parallel or antiparallel orientation of FM layers can be called as collinear magnetization configuration. On the other hand, it is also possible to obtain non-collinear magnetization configuration due to the competition between interlayer coupling energy and magnetic anisotropies of coupled FM layers for some structures [2-6]. Reducing magnetization switching currents while maintaining the thermal stability and increasing susceptibility of magnetic field sensors are the advantages of the non-collinear magnetization configuration [7]. In order to obtain this kind of configuration it is proposed to use FM layers with different magnetic easy axes such as one is aligned in the film plane, and another one is in the film normal [2-6]. Magnetic thin films with high perpendicular magnetic anisotropy (PMA) are required to construct these types of structures. PMA in magnetic thin films can be originated from interface anisotropy like Co/X (X=Pt, Pd, Ni) multilayered systems where Co and X layers have only a few atomic layer thickness [8, 9] and bulk anisotropy like Co1-XFex [10, 11]. In order to obtain significant magnetic properties in magnetic thin films, determining the growth rates [12] and thickness of layers [13], type of buffer layers [14] and annealing [15,16] is crucial for effectiveness of magnetic properties of multilayered films. Deposition technique is also important for such films [17, 18]. All these works demonstrated that different multilayered systems grown on different surface of Si (100) substrate have rich physical meaning and should be investigated in more detail. [Co/Pt]6/Pt/Co multilayer system have not been so far investigated for the effect of spacer and FM2 Co layer thickness on Si (100) surfaces. Here, we investigated the effect of spacer (Pt) and reference (Co) layer thickness in [Co/Pt]6/Pt/Co multilayer where [Co/Pt]6 and Co will be mentioned as FM1 and FM2, 2
respectively. Magnetic properties were probed by Magneto Optic Kerr effect (MOKE) and ferromagnetic resonance (FMR) techniques. Micromagnetic simulation based on metropolis algorithm was also used to determine the coupling between FM1-FM2 layers and anisotropy constants. The effect of spacer and reference layer thickness on [Co/Pt]6/tPt/tCo multilayered structure were revealed both experimentally and theoretically. 2. Sample preparation To investigate effect of interlayer coupling and effect of FM2 layer on non-collinear ordering of the multilayered structure, the thickness of Pt spacer and FM2 layers were changed and fabricated as sample set-1 and set-2 respectively. The sample sets are shown in Fig. 1a and 1b. The substrate was annealed at 750 K for 90 minutes. To obtain a perpendicularly magnetized FM layer, (Pt/Co)n multilayer with 6 repetition (n) were grown on Si (100) substrate. The thickness of nonmagnetic spacer Pt and reference Co layers were varied to obtain weakly coupled non-collinear magnetization configuration. All layers were deposited by using magnetron sputtering. At the bottom of the multilayer system Pt buffer layer was grown with a thickness of 40 Å in order to induce (111) textures. Finally, the multilayer system was covered with 40 Å Pt cap layer against to the oxidation. Pt cap layer was also proven to be inevitable to observe PMA in Co/Pt multilayer. The base pressure of the sputter chamber is 2x10-9 mbar and during the deposition Argon (Ar) pressure was 5x10-3mbar. Pt layers were grown by DC generators while Co layers were grown by RF generators. The deposition rates for Pt and Co layers were 8.75 and 0.44 Å/min, respectively [19]. Hysteresis loops of the samples were measured by MOKE technique. The MOKE measurements were done at room temperature for both longitudinal and polar MOKE geometry with a Nanosan Instruments MOKE magnetometer (Smart MOKE Magnetometer System). Also magnetic anisotropies of the samples were investigated by FMR measurements. The FMR measurements were carried out at room temperature with a microwave frequency of 9.8 GHz with a JEOL ESR spectrometer (JESFA300). We have performed FMR measurements for out of plane geometry (OPG) where the films were rotated from sample plane to sample normal with respect to the applied DC magnetic field. Magnetic field was swept from 0.0 T to 2.0 T.
3. Micromagnetic simulations
3
The multilayered structures on silica-like substrate is constructed meticulously by taking whole layers have the same smoothness geometry. The energy Hamiltonian of the system is described by
N −∑t M H cos(θ )cos(θ ) + sin(θ )sin(θ )cos(ϕ − ϕ ) H i H i H i i =1 i i N H = + ∑t i K effi cos2(θ i ) i =1 N N −∑∑[ Ai , j cos(θi − θ j ) i =1 j =1
(1)
where (θi ,θH ,θj) and (φi , φH) are respectively the polar and azimuth angles for magnetization vector M and external dc field vector H with respect to the film normal. ti indicates thickness of the layers obtained from X-Ray reflectivity (XRR) analysis of the samples. The first term of the Hamiltonian is Zeeman energy of the samples in external dc field. The second term represents the magneto static energy due to demagnetizing field. The last term appears from interlayer coupling interaction of ferromagnetic layers. The interlayer coupling energy densities between layers are determined by Ai,j bilinear coupling constant. Ai,j can be either positive or negative depending on type of the interaction. Multilayer structure with different spacer and FM2 layer thickness in nano-level is simulated based on Metropolis algorithm and Markov Chain Monte Carlo (MC-MC) method as realistic as possible. It is decided that 1000 MC steps/layer are appropriate to reach equilibrium state of the system after some smaller trials of step number. Herewith observables are obtained by using expectation values. Once reaching the equilibrium state thermodynamically, dynamic response of magnetization is tracked under a pulse field. Therefore external magnetic field is swept from 0 to 2000 mT to determine the field corresponding to the maximum value of χ2 which called as resonance field (Hres). Dynamic susceptibility spectrum χ2 is calculated by using the following equation: [20-22]
4
1 ∂ 2E 2ω 4π M 0 Dk 2 + M 0 ∂θ 2 γ 2T2 χ2 = 2 ω 2 ω 2 2 0 − + 44ω 2 γ γ γ T2
(2)
where ω0 is precession frequency of the system, sometimes called as Larmor frequency, and calculated by the following equation: [23]
2
ω0 2 1 ∂ 2ETotal = Dk + M 0 ∂θ 2 γ ∂ 2ETotal 1 − M sin2 θ ∂θ ∂ϕ 0
2 ∂ 2ETotal 1 Dk + M 0 sin2 θ ∂ϕ 2
2
1 + 2 2 γ T2
(3)
While external magnetic field is increasing, Larmor frequency is getting closer to resonance frequency of microwave radiation ω and dynamic susceptibility χ2 starts to increase because of decreasing difference between ω and ω0. At ω = ω0, the corresponding external magnetic field is marked as resonance field. All calculations are performed for room temperature to fit outputs of simulation with experimental data. Obtained data from the theoretical results are summarized in following section.
4. Results and Discussion Interaction between magnetic layers and effect of Co layer were studied. The thickness of Pt spacer layer has been changed for investigation of the effect of interaction between magnetic layers whose easy axis’ are orthogonal. In the investigation of the second effect, the thickness of Pt spacer layer is fixed to 5 nm and FM2 (Co) layer’s thickness has been varied. Firstly, the perpendicular magnetic anisotropy of [Pt/Co]6 multilayer on Pt buffer layer has been confirmed by measuring MOKE and FMR. The reasons for choosing 6 repetitions are having comparable thickness with FM2 (Co) layer and getting significant signal intensity in FMR measurements. We have varied the Pt spacer layer thickness from 3 to 5 nm. MOKE hysteresis loops of samples are given in Fig. 2 below is measured in longitudinal and polar geometry, respectively. 5
MOKE measurements have been performed in order to determine the effect of interlayer coupling, i.e. either ferromagnetic or anti-ferromagnetic, and strength of coupling. The loops show considerable changes when the interlayer coupling changes with increasing of the Pt spacer. There is single loop in hysteresis curves for 3 nm FM1 layer thickness while there are double loops for 4 and 5 nm thicknesses in longitudinal geometry. Fig 2a and 2d are typical easy and hard axis MOKE curves which imply that two layers are coupled and easy axis is in the film plane. In the case of Pt spacer thickness of 3 nm, FM1 ([Co/Pt]6) and FM2 (Co) layers switch simultaneously for L-MOKE (Fig. 2a) and show typical hard axis hysteresis behavior for PMOKE (Fig. 2d). From the hysteresis curves, one can conclude the magnetization parallel to the sample plane due to the strong demagnetization field of Co layer for 3 nm Pt spacer thicknesses and coupling is relatively strong. When the coupling is reduced by increasing spacer thickness (4 and 5 nm), the FM1 and FM2 layers do not switch in synchrony any longer. One layer which has easy axis in the direction of applied field switches sharply, and second one gives contribution like hard axis behavior. This behavior implies that the layers are not coupled strongly. In polar MOKE loops, there are two switches; one is “normal” and the other one is “reversed”. The reversed loop corresponds to the Co and [Pt/Co]6 layers being AFM coupled [3, 24]. Applied field forces to rotate the FM2 layer to the field direction. However, owing to the AFM coupling, both cannot be in the same direction. Consequently, FM2 layer switches to opposite alignment in a certain magnetic field resulting in sharp reduction of magnetization.
FMR spectra of sample set-1 are given for the applied field parallel to the sample normal (Θ=90º) and parallel to the sample plane (Θ=0º) as shown in Fig. 3, respectively. Between 200-350 mT cavity and holder have signal where is backgrounded with grey color. There are two FMR peaks for the films with the thickness of 4 and 5 nm which represents weakly coupling of two magnetic layers or uncoupled layers. [Co/Pt]6 gives peak around 150 mT when applied field is parallel to film normal and peak of Co layer appears higher field around 750-900 mT for weakly coupled case (Fig 3a). When applied field is in the direction of the sample plane, the order of peaks are reversed. On the other hand there is only one FMR peak for the film with 3 nm Pt spacer layer. Even though there are two magnetic layers with different magnetic anisotropies, one can see only one peak for the sample with 3 nm thick of 6
spacer layer. The magnetic layers are following the applied field simultaneously due to strong coupling and results in single peak. A full picture of angular dependency of the resonance fields are given for [Co/Pt]6/tPt/Co (t=3, 4, 5 nm) thin films in Fig. 4. Solid and dashed lines indicate resonance fields of FM1 and FM2 layers respectively where round and square marks are experimental data in Fig 4. For 3 nm, easy axis of the magnetic layers is parallel to the film plane. In Fig 4a, there is a single peak which means that strong coupling between two magnetic layers and Co magnetic layer dominates the magnetic behavior of the system. On the other hand, two resonance peaks are observed for 4 and 5 nm (Fig 4b and 4c). In these cases, the interlayer coupling is relatively weaker and two magnetic layers are able to reflect their own belongings. In contrast to previous studies, the interaction between layers is weakly present, even if the spacer layer is relatively thick (4 and 5 nm). This is potentially interesting for new spintronic applications.
Table 1. Summary of magnetic parameters of the samples with respect to spacer thickness, effective magnetization MEFF, effective anisotropy constant KEFF, Gyro magnetic ratio Gamma and bilinear coupling constant obtained from the theoretical results. Spacer
Bilinear
(nm)
C.C.
KEFF (106J/m3) FM1
Gamma (1010T-1 s-1)
FM2
FM1
FM2
MEFF (MA/m) FM1
FM2
(mJ/m2 ) 3
1.46
- 0.517
2.90
1.087
4
-0.79
0.288
- 0.260
3.08
2.81
0.917
1.256
5
-0.61
0.220
- 0.136
2.91
2.81
1.143
1.269
From the simulation results, while the thickness of Pt spacer layer is increased, the magnitude of bilinear coupling constant decreases. This illustrates weakly coupled state between ferromagnetic layers for spacer thickness of 4 and 5 nm. Also sign of the coupling constant is switched to negative implying anti-ferromagnetic coupling between magnetic layers for these thicknesses. Similarly, effective anisotropy constants of each magnetic layer KEFF have positive sign for FM1 and negative sign for FM2 in weakly coupled case which means that
7
FM1 layer has perpendicular magnetization and FM2 layer is in plane magnetization. Summary of magnetic parameters are given in Table 1 for sample set-1. To investigate effect of thickness of FM2 layer on non-collinear ordering, thickness of the Co (FM2) layer has been varied from 2 to 4 nm. The other layers kept constant; Pt 5 nm and FM1 2.4nm. In this case ferromagnetic layers are weakly coupled. MOKE hysteresis curves are given in Fig 5. There are double loops for 2 and 3 nm of FM2 layer thicknesses while there is single loop for 4 nm thicknesses in longitudinal geometry. The discussion above for sample set1 is valid for the Polar MOKE loops. Nevertheless, coercive field of perpendicular FM1 layer is getting smaller by increasing thickness of reference layer. Finally for 4 nm thick Co, it is almost zero. Ferromagnetic Resonance fields of FM1 and FM2 layers as a function of angle for OPG is given in Fig 6. Theoretical resonance fields of the layers are represented by solid and dashed lines. Round and square marks are experimental data of the layers. Resonance field of the FM2 layer has changed from smaller to larger values when applied field direction was rotated from film plane to film normal. On the other hand, resonance fields of the FM1 layer were opposite behavior, meaning that these two magnetic layers have orthogonal easy axes respect to each other. The difference between resonance field of zero and ninety degrees corresponds to the value of effective magnetic anisotropy. Thus, it is seen that anisotropy of FM1 layer is decreasing as the thickness of FM2 layer is increased and anisotropy of FM2 is increasing. The anisotropy of FM2 is increased with thickness owing to demagnetization field of the layer. On the other hand, the effective anisotropy of FM1 layer is decreased due to indirect coupling with FM2 via spacer layer. Analysis of FMR resonance fields showed that two layers are anti-ferromagnetic weakly coupled regardless of FM2 layer thickness. It is consistent with P-MOKE curves as well. Magnetic parameters of the sample set-2 from analysis of the FMR resonance fields are given in Table 2.
8
Table 2. Summary of magnetic properties of all samples with respect to FM2 layer thickness saturation magnetization Ms, effective anisotropy constant Keff, Gyro magnetic ratio Gamma and bilinear coupling constant obtained from the theoretical results. FM2
Bilinear
(nm)
C.C.
KEFF (106 J/m3)
Gamma (10 10 T-1 s-1)
MS (MA/m)
FM1
FM2
FM1
FM2
FM1
FM2
2
(mJ/m ) 2
-0.61
0.220
- 0.136
2.91
2.81
1.143
1.269
3
-0.61
0.151
- 0.206
2.75
3.18
1.143
1.269
4
-0.61
0.098
- 0.260
2.67
3.62
1.143
1.269
Co (FM2) layer thickness is increased from 2 nm to 4 nm while thickness of Pt spacer is fixed to 5 nm. Therefore, bilinear coupling constant values were not changed. However, effective anisotropy constant of FM2 layer increased while magnitude of FM1 layer’s anisotropy constant gradually decreased. Thus, these results are in good agreement both FMR and MOKE measurements. 5. Conclusion The dependence of the magnetic anisotropy and magnetization reversal processes in [Co/Pt]6/tPt/tCo ultra-thin ferromagnetic films have been studied, as a function of the reference Co and Pt spacer layer thickness by magneto-optic Kerr effect technique and ferromagnetic resonance spectroscopy. The anisotropy and interlayer exchange constants have been determined from a micromagnetic simulation based on metropolis algorithm. The interaction is weakly presence even if the spacer layer is thick (4 and 5 nm). Decreasing the thickness of spacer causes an increase in strength of interaction. Also, increasing the thickness of reference layer (FM2) results an increase in number of polarized electrons in the system which is origin of interlayer coupling. The polarized electrons due to FM2 layer dominates the system and causes a diminish of effective magnetic anisotropy of the FM1 layer. In conclusion, controlling of non-collinear states in such systems is possible by variation thickness of spacer and reference layers. It is obviously seen that the [Co/Pt]6/tPt/tCo trilayer system can be used in multilayered magnetic systems especially spin torque applications.
9
Acknowledgement This study was supported by Research Projects with Foundation Number FEN-C-DRP110316-0103, Marmara University Scientific Research Commission, Turkey. Simulations are carried out at the Simulation and Modeling Research Laboratory, Department of Physics, Marmara University. References [1] A. Friedrich, G. Creuzet, Magnetoresistive sensor, in, Google Patents, 1992. [2] F. Yildiz, M. Przybylski, J. Kirschner, Perpendicular anisotropy and oscillatory interlayer coupling in Fe0.5Co0.5/Rh/Fe0.5Co0.5 bilayers on Rh(001),J. Appl. Phys., 105 (2009)07C312-07C312-3. [3] F. Yildiz, M. Przybylski, J. Kirschner, Direct Evidence of a Nonorthogonal Magnetization Configuration in Single Crystalline Fe1−xCox/Rh/Fe/Rh(001) System, Phys. Rev. Lett., 103 (2009)147203. [4] J.M. Tonnerre, M. Przybylski, M. Ragheb, F. Yildiz, H. C. N. Tolentino, L. Ortega, J. Kirschner, Direct in-depth determination of a complex magnetic configuration in an exchange-coupled bilayer with perpendicular and in-plane anisotropy, Phys. Rev. B, 84 (2011)100407. [5] W. Kuch, XingyuGao, and J. Kirschner, Competition between in-plane and out-ofplanemagnetization in exchange-coupled magnetic films, Phys. Rev. B, 65 (2002) 064406. [6] A. Taga, L. Nordstrom, P. James, B. Johansson, and O. Eriksson, Non-collinear states in magnetic sensors, Nature 406 (2000)280-2. [7]S. M. Thompson, The discovery, development and future of GMR: The Nobel Prize2007, J. Phys. D 41 (2008) 093001. [8]M. Sakamaki, K. Amemiya, M. O. Liedke, J. Fassbender, P. Mazalski, I. Sveklo, and A. Maziewski, Perpendicular magnetic anisotropy in a Pt/Co/Pt ultra-thin film arising from a lattice distortion induced by ion irradiation, Phys. Rev. B 86 (2012)024418.
10
[9] P. F. Carcia, A. D. Meinhaldt, and A. Suna, Perpendicular magnetic anisotropy in Pd/Co thin film layered structures, Appl. Phys. Lett.47 (1985) 178–180. [10]F. Yildiz, M. Przybylski, X.-D. Ma, and J. Kirschner, Strong perpendicular anisotropy in Fe1−xCo x alloy films epitaxially grown on mismatching Pd(001), Ir(001), and Rh(001) substrates, Phys. Rev. B 80 (2009)064415 [11]F. Yildiz, F. Luo, C. Tieg, R. M. Abrudan, X. L. Fu, A. Winkelmann, M. Przybylski, and J. Kirschner, Strongly Enhanced Orbital Moment by Reduced Lattice Symmetry and Varying Composition of Fe1−xCox Alloy Films, Phys. Rev. Lett. 100 (2008) 037205 [12] D.H. Hernando, Conductance modulation by spin precession in noncollinear ferromagnet normal-metal ferromagnet systems, Phys. Rev. B 62 (2000) 5700 [13]L. You, R. C. Sousa, S. Bandiera, B. Rodmacq, and B. Dieny, Co/Ni multilayers with perpendicular anisotropy for spintronic device applications, Appl. Phys. Lett. 100 (2012)172411. [14]M.P.R. Sabino, S.T. Lim, M. Tran, Composition and Annealing Temperature Dependence of Magnetic Properties in MgO/Co–Fe–B/Ta/Co–Fe–B/MgO Films, J.Appl.Phys. 50 (2014) 115. [15]S. Vorobiov, I. Lytvynenko, T. Hauet, M. Hehn, D. Derecha, A. Chornous, The effect of annealing on magnetic properties of Co/Gd multilayers, Vacuum120 (2015) 9-12 [16]T. Das, P.D. Kulkarni, S.C. Purandare, H.C. Barshilia, S.Bhattacharyya and P. Chowdhury, Anomalous enhancement in interfacial perpendicular magnetic anisotropy through uphill diffusion, Sci. Rep. 4 (2014) 5328 [17]M.T. Johnson,R. Jungblut, P.J. Kelly, F.J.A. den Broeder, Perpendicular magnetic anisotropy of multilayers: recent insights, J. Magn. Magn.Mater.148 (1995) 118–124 [18]Y.B. Zhang, J.A. Woollam, Z.S. Shan, J.X. Shen, D.J. Sellmyer, Anisotropy and Magneto-Optical Properties Of Sputtered Co/Ni Multilayer Thin Films, IEEE Trans.Magn.30 (1994) 4440–4442 [19]S. Akbulut, A. Akbulut, M. Özdemir, F. Yildiz, Effect of deposition technique of Ni on the perpendicular magnetic anisotropy in Co/Ni multilayers, J. Magn. Magn.Mater.390 (2015) 137-141 [20]R.F. Soohoo, Ferromagnetic and spin-wave resonance in multilayer films, J. Appl. Phys., 63 (1988) 3829 [21]R.F. Soohoo, Excitation and boundary effects in spin-wave resonance, J. Appl. Phys., 32 (1961) 148-150 [22]R.F. Soohoo, General spin-wave dispersion relations, Phys. Rev. 120 (1978) 1960 11
[23]M. Özdemir, Electron spin resonance (ESR) and resistivity measurements on NiMn, NiMnPt and CrFe alloys thin films, Ph.D. dissertation, Dept. Phys., Istanbul Technical Univ., Inst. Pure Appl. Sci., Istanbul, Turkey, 1998 [24]T.Devolder, Ferromagnetic resonance study of composite Co/Ni-FeCoB free layers with perpendicular anisotropy, J.Appl. Phys. 119 (2016) 153905
12
Fig 1. Schematic representation of (a) [Co/Pt]6/Pt(3-5nm)/Co(2nm), (b) [Co/Pt]6/Pt(5 nm)/Co(2-4 nm) multilayer. Arrows show easy direction of magnetization for uncoupled cases of magnetic layers.
13
Fig 2. Hysteresis loops of samples where (a) d=3 nm, (b) e=4 nm and (c) f=5 nm Pt spacer layer.
14
Fig 3. FMR spectra of the sample set-1 at (a) the applied field perpendicular to the film plane (angle of 90°) and (b) the applied field parallel to the film plane (angle of 90°).
15
Fig 4. Ferromagnetic resonance fields of samples (a) for 3, (b) for 4, (c) for 5 nm Pt spacer layer is measured in OPG, respectively (Lines: Simulation outputs, Symbols: Experimental results)
16
Fig 5. Hysteresis loops of samples where (a) d=2 nm, (b) e=3 nm and (c) f=4 nm Co reference layer.
17
Fig 6. Ferromagnetic resonance fields of samples (a) for 2 nm, (b) for 3 nm, (c) for 4 nm FM2 (Co) layer is measured in out of plane geometry, respectively (Lines: Simulation outputs, Symbols: Experimental results)
18
Highlights
-
Effects of Pt spacer and reference layers thickness are investigated by magneto optic Kerr effect and Ferromagnetic resonance techniques.
-
Controlling of non-collinear states in multilayered thin film systems is studied.
-
Micromagnetic approach is used for modeling of magnetic multilayered structure.
-
Bilinear coupling between ferromagnetic layers and anisotropy constants are determined by a micromagnetic simulation based on metropolis algorithm.
S0304-8853(17)30067-7 http://dx.doi.org/10.1016/j.jmmm.2017.04.008 MAGMA 62610
To appear in:
Journal of Magnetism and Magnetic Materials
Received Date: Revised Date: Accepted Date:
10 January 2017 2 April 2017 5 April 2017
Please cite this article as: T. Kalaycı, C. Deger, S. Akbulut, F. Yildiz, Tuning Magnetic Properties of Non-Collinear Magnetization Configuration in Pt/[Pt/Co]6/Pt/Co/Pt Multilayer Structure, Journal of Magnetism and Magnetic Materials (2017), doi: http://dx.doi.org/10.1016/j.jmmm.2017.04.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Tuning Magnetic Properties of Non-Collinear Magnetization Configuration in Pt/[Pt/Co]6/Pt/Co/Pt Multilayer Structure Taner Kalaycı a, Caner Deger a, Salih Akbulut b, and Fikret Yildiz b,* a
b
Department of Physics, Marmara University, 34722, Kadıköy, Istanbul, Turkey Department of Physics, Gebze Technical University, 41400, Gebze, Kocaeli, Turkey
Abstract In this study, effects of Pt spacer and Co reference layers thickness in [Co/Pt]6/Pt/Co multilayer have been revealed to tailor magnetization directions in non-collinear configuration. Magneto-optic Kerr effect and ferromagnetic resonance techniques were employed to investigate magnetic properties. Bilinear coupling between [Co/Pt]6 and Co layers and anisotropy constants were determined by a micromagnetic simulation based on metropolis algorithm. 3 nm spacer causes ferromagnetic coupling while the samples have 4 and 5 nm spacer are coupled anti-ferromagneticaly. Also, tuning magnetic anisotropy of [Co/Pt]6 layer was accomplished by Co reference layer. It is revealed that controlling of noncollinear states in such systems is possible by variation of thickness of spacer and reference layers and [Co/Pt]6/tPt/tCo trilayer system can be used in multilayered magnetic systems. Keywords: Non-collinear magnetic configuration, interlayer coupling, magnetic anisotropy, spin transfer torque *
Corresponding Authors:
E-mail addresses: [email protected], [email protected]
1. Introduction Ferromagnetic (FM) layers separated by a nonmagnetic metallic spacer layer are attracted lots of attention since they are widely employed in spintronic applications such as magnetic recording and sensor technologies [1]. The electrical resistivity of this kind of structures can be manipulated under magnetic field since it depends on the spin direction. Correlatively, Spin Transfer Torque (STT) effect is a mechanism that can control and manipulate the magnetism of a ferromagnet by current. So, STT is an alternative way uses current instead of Oersted field. Also, spin torque nano-oscillator (STNO) is a promising candidate for future microwave generators since it can be tuned by current and field. The spin direction of FM layers can be parallel or antiparallel to each other due to nonmagnetic spacer layer thickness in magnetic multilayer systems. This parallel or antiparallel orientation of FM layers can be called as collinear magnetization configuration. On the other hand, it is also possible to obtain non-collinear magnetization configuration due to the competition between interlayer coupling energy and magnetic anisotropies of coupled FM layers for some structures [2-6]. Reducing magnetization switching currents while maintaining the thermal stability and increasing susceptibility of magnetic field sensors are the advantages of the non-collinear magnetization configuration [7]. In order to obtain this kind of configuration it is proposed to use FM layers with different magnetic easy axes such as one is aligned in the film plane, and another one is in the film normal [2-6]. Magnetic thin films with high perpendicular magnetic anisotropy (PMA) are required to construct these types of structures. PMA in magnetic thin films can be originated from interface anisotropy like Co/X (X=Pt, Pd, Ni) multilayered systems where Co and X layers have only a few atomic layer thickness [8, 9] and bulk anisotropy like Co1-XFex [10, 11]. In order to obtain significant magnetic properties in magnetic thin films, determining the growth rates [12] and thickness of layers [13], type of buffer layers [14] and annealing [15,16] is crucial for effectiveness of magnetic properties of multilayered films. Deposition technique is also important for such films [17, 18]. All these works demonstrated that different multilayered systems grown on different surface of Si (100) substrate have rich physical meaning and should be investigated in more detail. [Co/Pt]6/Pt/Co multilayer system have not been so far investigated for the effect of spacer and FM2 Co layer thickness on Si (100) surfaces. Here, we investigated the effect of spacer (Pt) and reference (Co) layer thickness in [Co/Pt]6/Pt/Co multilayer where [Co/Pt]6 and Co will be mentioned as FM1 and FM2, 2
respectively. Magnetic properties were probed by Magneto Optic Kerr effect (MOKE) and ferromagnetic resonance (FMR) techniques. Micromagnetic simulation based on metropolis algorithm was also used to determine the coupling between FM1-FM2 layers and anisotropy constants. The effect of spacer and reference layer thickness on [Co/Pt]6/tPt/tCo multilayered structure were revealed both experimentally and theoretically. 2. Sample preparation To investigate effect of interlayer coupling and effect of FM2 layer on non-collinear ordering of the multilayered structure, the thickness of Pt spacer and FM2 layers were changed and fabricated as sample set-1 and set-2 respectively. The sample sets are shown in Fig. 1a and 1b. The substrate was annealed at 750 K for 90 minutes. To obtain a perpendicularly magnetized FM layer, (Pt/Co)n multilayer with 6 repetition (n) were grown on Si (100) substrate. The thickness of nonmagnetic spacer Pt and reference Co layers were varied to obtain weakly coupled non-collinear magnetization configuration. All layers were deposited by using magnetron sputtering. At the bottom of the multilayer system Pt buffer layer was grown with a thickness of 40 Å in order to induce (111) textures. Finally, the multilayer system was covered with 40 Å Pt cap layer against to the oxidation. Pt cap layer was also proven to be inevitable to observe PMA in Co/Pt multilayer. The base pressure of the sputter chamber is 2x10-9 mbar and during the deposition Argon (Ar) pressure was 5x10-3mbar. Pt layers were grown by DC generators while Co layers were grown by RF generators. The deposition rates for Pt and Co layers were 8.75 and 0.44 Å/min, respectively [19]. Hysteresis loops of the samples were measured by MOKE technique. The MOKE measurements were done at room temperature for both longitudinal and polar MOKE geometry with a Nanosan Instruments MOKE magnetometer (Smart MOKE Magnetometer System). Also magnetic anisotropies of the samples were investigated by FMR measurements. The FMR measurements were carried out at room temperature with a microwave frequency of 9.8 GHz with a JEOL ESR spectrometer (JESFA300). We have performed FMR measurements for out of plane geometry (OPG) where the films were rotated from sample plane to sample normal with respect to the applied DC magnetic field. Magnetic field was swept from 0.0 T to 2.0 T.
3. Micromagnetic simulations
3
The multilayered structures on silica-like substrate is constructed meticulously by taking whole layers have the same smoothness geometry. The energy Hamiltonian of the system is described by
N −∑t M H cos(θ )cos(θ ) + sin(θ )sin(θ )cos(ϕ − ϕ ) H i H i H i i =1 i i N H = + ∑t i K effi cos2(θ i ) i =1 N N −∑∑[ Ai , j cos(θi − θ j ) i =1 j =1
(1)
where (θi ,θH ,θj) and (φi , φH) are respectively the polar and azimuth angles for magnetization vector M and external dc field vector H with respect to the film normal. ti indicates thickness of the layers obtained from X-Ray reflectivity (XRR) analysis of the samples. The first term of the Hamiltonian is Zeeman energy of the samples in external dc field. The second term represents the magneto static energy due to demagnetizing field. The last term appears from interlayer coupling interaction of ferromagnetic layers. The interlayer coupling energy densities between layers are determined by Ai,j bilinear coupling constant. Ai,j can be either positive or negative depending on type of the interaction. Multilayer structure with different spacer and FM2 layer thickness in nano-level is simulated based on Metropolis algorithm and Markov Chain Monte Carlo (MC-MC) method as realistic as possible. It is decided that 1000 MC steps/layer are appropriate to reach equilibrium state of the system after some smaller trials of step number. Herewith observables are obtained by using expectation values. Once reaching the equilibrium state thermodynamically, dynamic response of magnetization is tracked under a pulse field. Therefore external magnetic field is swept from 0 to 2000 mT to determine the field corresponding to the maximum value of χ2 which called as resonance field (Hres). Dynamic susceptibility spectrum χ2 is calculated by using the following equation: [20-22]
4
1 ∂ 2E 2ω 4π M 0 Dk 2 + M 0 ∂θ 2 γ 2T2 χ2 = 2 ω 2 ω 2 2 0 − + 44ω 2 γ γ γ T2
(2)
where ω0 is precession frequency of the system, sometimes called as Larmor frequency, and calculated by the following equation: [23]
2
ω0 2 1 ∂ 2ETotal = Dk + M 0 ∂θ 2 γ ∂ 2ETotal 1 − M sin2 θ ∂θ ∂ϕ 0
2 ∂ 2ETotal 1 Dk + M 0 sin2 θ ∂ϕ 2
2
1 + 2 2 γ T2
(3)
While external magnetic field is increasing, Larmor frequency is getting closer to resonance frequency of microwave radiation ω and dynamic susceptibility χ2 starts to increase because of decreasing difference between ω and ω0. At ω = ω0, the corresponding external magnetic field is marked as resonance field. All calculations are performed for room temperature to fit outputs of simulation with experimental data. Obtained data from the theoretical results are summarized in following section.
4. Results and Discussion Interaction between magnetic layers and effect of Co layer were studied. The thickness of Pt spacer layer has been changed for investigation of the effect of interaction between magnetic layers whose easy axis’ are orthogonal. In the investigation of the second effect, the thickness of Pt spacer layer is fixed to 5 nm and FM2 (Co) layer’s thickness has been varied. Firstly, the perpendicular magnetic anisotropy of [Pt/Co]6 multilayer on Pt buffer layer has been confirmed by measuring MOKE and FMR. The reasons for choosing 6 repetitions are having comparable thickness with FM2 (Co) layer and getting significant signal intensity in FMR measurements. We have varied the Pt spacer layer thickness from 3 to 5 nm. MOKE hysteresis loops of samples are given in Fig. 2 below is measured in longitudinal and polar geometry, respectively. 5
MOKE measurements have been performed in order to determine the effect of interlayer coupling, i.e. either ferromagnetic or anti-ferromagnetic, and strength of coupling. The loops show considerable changes when the interlayer coupling changes with increasing of the Pt spacer. There is single loop in hysteresis curves for 3 nm FM1 layer thickness while there are double loops for 4 and 5 nm thicknesses in longitudinal geometry. Fig 2a and 2d are typical easy and hard axis MOKE curves which imply that two layers are coupled and easy axis is in the film plane. In the case of Pt spacer thickness of 3 nm, FM1 ([Co/Pt]6) and FM2 (Co) layers switch simultaneously for L-MOKE (Fig. 2a) and show typical hard axis hysteresis behavior for PMOKE (Fig. 2d). From the hysteresis curves, one can conclude the magnetization parallel to the sample plane due to the strong demagnetization field of Co layer for 3 nm Pt spacer thicknesses and coupling is relatively strong. When the coupling is reduced by increasing spacer thickness (4 and 5 nm), the FM1 and FM2 layers do not switch in synchrony any longer. One layer which has easy axis in the direction of applied field switches sharply, and second one gives contribution like hard axis behavior. This behavior implies that the layers are not coupled strongly. In polar MOKE loops, there are two switches; one is “normal” and the other one is “reversed”. The reversed loop corresponds to the Co and [Pt/Co]6 layers being AFM coupled [3, 24]. Applied field forces to rotate the FM2 layer to the field direction. However, owing to the AFM coupling, both cannot be in the same direction. Consequently, FM2 layer switches to opposite alignment in a certain magnetic field resulting in sharp reduction of magnetization.
FMR spectra of sample set-1 are given for the applied field parallel to the sample normal (Θ=90º) and parallel to the sample plane (Θ=0º) as shown in Fig. 3, respectively. Between 200-350 mT cavity and holder have signal where is backgrounded with grey color. There are two FMR peaks for the films with the thickness of 4 and 5 nm which represents weakly coupling of two magnetic layers or uncoupled layers. [Co/Pt]6 gives peak around 150 mT when applied field is parallel to film normal and peak of Co layer appears higher field around 750-900 mT for weakly coupled case (Fig 3a). When applied field is in the direction of the sample plane, the order of peaks are reversed. On the other hand there is only one FMR peak for the film with 3 nm Pt spacer layer. Even though there are two magnetic layers with different magnetic anisotropies, one can see only one peak for the sample with 3 nm thick of 6
spacer layer. The magnetic layers are following the applied field simultaneously due to strong coupling and results in single peak. A full picture of angular dependency of the resonance fields are given for [Co/Pt]6/tPt/Co (t=3, 4, 5 nm) thin films in Fig. 4. Solid and dashed lines indicate resonance fields of FM1 and FM2 layers respectively where round and square marks are experimental data in Fig 4. For 3 nm, easy axis of the magnetic layers is parallel to the film plane. In Fig 4a, there is a single peak which means that strong coupling between two magnetic layers and Co magnetic layer dominates the magnetic behavior of the system. On the other hand, two resonance peaks are observed for 4 and 5 nm (Fig 4b and 4c). In these cases, the interlayer coupling is relatively weaker and two magnetic layers are able to reflect their own belongings. In contrast to previous studies, the interaction between layers is weakly present, even if the spacer layer is relatively thick (4 and 5 nm). This is potentially interesting for new spintronic applications.
Table 1. Summary of magnetic parameters of the samples with respect to spacer thickness, effective magnetization MEFF, effective anisotropy constant KEFF, Gyro magnetic ratio Gamma and bilinear coupling constant obtained from the theoretical results. Spacer
Bilinear
(nm)
C.C.
KEFF (106J/m3) FM1
Gamma (1010T-1 s-1)
FM2
FM1
FM2
MEFF (MA/m) FM1
FM2
(mJ/m2 ) 3
1.46
- 0.517
2.90
1.087
4
-0.79
0.288
- 0.260
3.08
2.81
0.917
1.256
5
-0.61
0.220
- 0.136
2.91
2.81
1.143
1.269
From the simulation results, while the thickness of Pt spacer layer is increased, the magnitude of bilinear coupling constant decreases. This illustrates weakly coupled state between ferromagnetic layers for spacer thickness of 4 and 5 nm. Also sign of the coupling constant is switched to negative implying anti-ferromagnetic coupling between magnetic layers for these thicknesses. Similarly, effective anisotropy constants of each magnetic layer KEFF have positive sign for FM1 and negative sign for FM2 in weakly coupled case which means that
7
FM1 layer has perpendicular magnetization and FM2 layer is in plane magnetization. Summary of magnetic parameters are given in Table 1 for sample set-1. To investigate effect of thickness of FM2 layer on non-collinear ordering, thickness of the Co (FM2) layer has been varied from 2 to 4 nm. The other layers kept constant; Pt 5 nm and FM1 2.4nm. In this case ferromagnetic layers are weakly coupled. MOKE hysteresis curves are given in Fig 5. There are double loops for 2 and 3 nm of FM2 layer thicknesses while there is single loop for 4 nm thicknesses in longitudinal geometry. The discussion above for sample set1 is valid for the Polar MOKE loops. Nevertheless, coercive field of perpendicular FM1 layer is getting smaller by increasing thickness of reference layer. Finally for 4 nm thick Co, it is almost zero. Ferromagnetic Resonance fields of FM1 and FM2 layers as a function of angle for OPG is given in Fig 6. Theoretical resonance fields of the layers are represented by solid and dashed lines. Round and square marks are experimental data of the layers. Resonance field of the FM2 layer has changed from smaller to larger values when applied field direction was rotated from film plane to film normal. On the other hand, resonance fields of the FM1 layer were opposite behavior, meaning that these two magnetic layers have orthogonal easy axes respect to each other. The difference between resonance field of zero and ninety degrees corresponds to the value of effective magnetic anisotropy. Thus, it is seen that anisotropy of FM1 layer is decreasing as the thickness of FM2 layer is increased and anisotropy of FM2 is increasing. The anisotropy of FM2 is increased with thickness owing to demagnetization field of the layer. On the other hand, the effective anisotropy of FM1 layer is decreased due to indirect coupling with FM2 via spacer layer. Analysis of FMR resonance fields showed that two layers are anti-ferromagnetic weakly coupled regardless of FM2 layer thickness. It is consistent with P-MOKE curves as well. Magnetic parameters of the sample set-2 from analysis of the FMR resonance fields are given in Table 2.
8
Table 2. Summary of magnetic properties of all samples with respect to FM2 layer thickness saturation magnetization Ms, effective anisotropy constant Keff, Gyro magnetic ratio Gamma and bilinear coupling constant obtained from the theoretical results. FM2
Bilinear
(nm)
C.C.
KEFF (106 J/m3)
Gamma (10 10 T-1 s-1)
MS (MA/m)
FM1
FM2
FM1
FM2
FM1
FM2
2
(mJ/m ) 2
-0.61
0.220
- 0.136
2.91
2.81
1.143
1.269
3
-0.61
0.151
- 0.206
2.75
3.18
1.143
1.269
4
-0.61
0.098
- 0.260
2.67
3.62
1.143
1.269
Co (FM2) layer thickness is increased from 2 nm to 4 nm while thickness of Pt spacer is fixed to 5 nm. Therefore, bilinear coupling constant values were not changed. However, effective anisotropy constant of FM2 layer increased while magnitude of FM1 layer’s anisotropy constant gradually decreased. Thus, these results are in good agreement both FMR and MOKE measurements. 5. Conclusion The dependence of the magnetic anisotropy and magnetization reversal processes in [Co/Pt]6/tPt/tCo ultra-thin ferromagnetic films have been studied, as a function of the reference Co and Pt spacer layer thickness by magneto-optic Kerr effect technique and ferromagnetic resonance spectroscopy. The anisotropy and interlayer exchange constants have been determined from a micromagnetic simulation based on metropolis algorithm. The interaction is weakly presence even if the spacer layer is thick (4 and 5 nm). Decreasing the thickness of spacer causes an increase in strength of interaction. Also, increasing the thickness of reference layer (FM2) results an increase in number of polarized electrons in the system which is origin of interlayer coupling. The polarized electrons due to FM2 layer dominates the system and causes a diminish of effective magnetic anisotropy of the FM1 layer. In conclusion, controlling of non-collinear states in such systems is possible by variation thickness of spacer and reference layers. It is obviously seen that the [Co/Pt]6/tPt/tCo trilayer system can be used in multilayered magnetic systems especially spin torque applications.
9
Acknowledgement This study was supported by Research Projects with Foundation Number FEN-C-DRP110316-0103, Marmara University Scientific Research Commission, Turkey. Simulations are carried out at the Simulation and Modeling Research Laboratory, Department of Physics, Marmara University. References [1] A. Friedrich, G. Creuzet, Magnetoresistive sensor, in, Google Patents, 1992. [2] F. Yildiz, M. Przybylski, J. Kirschner, Perpendicular anisotropy and oscillatory interlayer coupling in Fe0.5Co0.5/Rh/Fe0.5Co0.5 bilayers on Rh(001),J. Appl. Phys., 105 (2009)07C312-07C312-3. [3] F. Yildiz, M. Przybylski, J. Kirschner, Direct Evidence of a Nonorthogonal Magnetization Configuration in Single Crystalline Fe1−xCox/Rh/Fe/Rh(001) System, Phys. Rev. Lett., 103 (2009)147203. [4] J.M. Tonnerre, M. Przybylski, M. Ragheb, F. Yildiz, H. C. N. Tolentino, L. Ortega, J. Kirschner, Direct in-depth determination of a complex magnetic configuration in an exchange-coupled bilayer with perpendicular and in-plane anisotropy, Phys. Rev. B, 84 (2011)100407. [5] W. Kuch, XingyuGao, and J. Kirschner, Competition between in-plane and out-ofplanemagnetization in exchange-coupled magnetic films, Phys. Rev. B, 65 (2002) 064406. [6] A. Taga, L. Nordstrom, P. James, B. Johansson, and O. Eriksson, Non-collinear states in magnetic sensors, Nature 406 (2000)280-2. [7]S. M. Thompson, The discovery, development and future of GMR: The Nobel Prize2007, J. Phys. D 41 (2008) 093001. [8]M. Sakamaki, K. Amemiya, M. O. Liedke, J. Fassbender, P. Mazalski, I. Sveklo, and A. Maziewski, Perpendicular magnetic anisotropy in a Pt/Co/Pt ultra-thin film arising from a lattice distortion induced by ion irradiation, Phys. Rev. B 86 (2012)024418.
10
[9] P. F. Carcia, A. D. Meinhaldt, and A. Suna, Perpendicular magnetic anisotropy in Pd/Co thin film layered structures, Appl. Phys. Lett.47 (1985) 178–180. [10]F. Yildiz, M. Przybylski, X.-D. Ma, and J. Kirschner, Strong perpendicular anisotropy in Fe1−xCo x alloy films epitaxially grown on mismatching Pd(001), Ir(001), and Rh(001) substrates, Phys. Rev. B 80 (2009)064415 [11]F. Yildiz, F. Luo, C. Tieg, R. M. Abrudan, X. L. Fu, A. Winkelmann, M. Przybylski, and J. Kirschner, Strongly Enhanced Orbital Moment by Reduced Lattice Symmetry and Varying Composition of Fe1−xCox Alloy Films, Phys. Rev. Lett. 100 (2008) 037205 [12] D.H. Hernando, Conductance modulation by spin precession in noncollinear ferromagnet normal-metal ferromagnet systems, Phys. Rev. B 62 (2000) 5700 [13]L. You, R. C. Sousa, S. Bandiera, B. Rodmacq, and B. Dieny, Co/Ni multilayers with perpendicular anisotropy for spintronic device applications, Appl. Phys. Lett. 100 (2012)172411. [14]M.P.R. Sabino, S.T. Lim, M. Tran, Composition and Annealing Temperature Dependence of Magnetic Properties in MgO/Co–Fe–B/Ta/Co–Fe–B/MgO Films, J.Appl.Phys. 50 (2014) 115. [15]S. Vorobiov, I. Lytvynenko, T. Hauet, M. Hehn, D. Derecha, A. Chornous, The effect of annealing on magnetic properties of Co/Gd multilayers, Vacuum120 (2015) 9-12 [16]T. Das, P.D. Kulkarni, S.C. Purandare, H.C. Barshilia, S.Bhattacharyya and P. Chowdhury, Anomalous enhancement in interfacial perpendicular magnetic anisotropy through uphill diffusion, Sci. Rep. 4 (2014) 5328 [17]M.T. Johnson,R. Jungblut, P.J. Kelly, F.J.A. den Broeder, Perpendicular magnetic anisotropy of multilayers: recent insights, J. Magn. Magn.Mater.148 (1995) 118–124 [18]Y.B. Zhang, J.A. Woollam, Z.S. Shan, J.X. Shen, D.J. Sellmyer, Anisotropy and Magneto-Optical Properties Of Sputtered Co/Ni Multilayer Thin Films, IEEE Trans.Magn.30 (1994) 4440–4442 [19]S. Akbulut, A. Akbulut, M. Özdemir, F. Yildiz, Effect of deposition technique of Ni on the perpendicular magnetic anisotropy in Co/Ni multilayers, J. Magn. Magn.Mater.390 (2015) 137-141 [20]R.F. Soohoo, Ferromagnetic and spin-wave resonance in multilayer films, J. Appl. Phys., 63 (1988) 3829 [21]R.F. Soohoo, Excitation and boundary effects in spin-wave resonance, J. Appl. Phys., 32 (1961) 148-150 [22]R.F. Soohoo, General spin-wave dispersion relations, Phys. Rev. 120 (1978) 1960 11
[23]M. Özdemir, Electron spin resonance (ESR) and resistivity measurements on NiMn, NiMnPt and CrFe alloys thin films, Ph.D. dissertation, Dept. Phys., Istanbul Technical Univ., Inst. Pure Appl. Sci., Istanbul, Turkey, 1998 [24]T.Devolder, Ferromagnetic resonance study of composite Co/Ni-FeCoB free layers with perpendicular anisotropy, J.Appl. Phys. 119 (2016) 153905
12
Fig 1. Schematic representation of (a) [Co/Pt]6/Pt(3-5nm)/Co(2nm), (b) [Co/Pt]6/Pt(5 nm)/Co(2-4 nm) multilayer. Arrows show easy direction of magnetization for uncoupled cases of magnetic layers.
13
Fig 2. Hysteresis loops of samples where (a) d=3 nm, (b) e=4 nm and (c) f=5 nm Pt spacer layer.
14
Fig 3. FMR spectra of the sample set-1 at (a) the applied field perpendicular to the film plane (angle of 90°) and (b) the applied field parallel to the film plane (angle of 90°).
15
Fig 4. Ferromagnetic resonance fields of samples (a) for 3, (b) for 4, (c) for 5 nm Pt spacer layer is measured in OPG, respectively (Lines: Simulation outputs, Symbols: Experimental results)
16
Fig 5. Hysteresis loops of samples where (a) d=2 nm, (b) e=3 nm and (c) f=4 nm Co reference layer.
17
Fig 6. Ferromagnetic resonance fields of samples (a) for 2 nm, (b) for 3 nm, (c) for 4 nm FM2 (Co) layer is measured in out of plane geometry, respectively (Lines: Simulation outputs, Symbols: Experimental results)
18
Highlights
-
Effects of Pt spacer and reference layers thickness are investigated by magneto optic Kerr effect and Ferromagnetic resonance techniques.
-
Controlling of non-collinear states in multilayered thin film systems is studied.
-
Micromagnetic approach is used for modeling of magnetic multilayered structure.
-
Bilinear coupling between ferromagnetic layers and anisotropy constants are determined by a micromagnetic simulation based on metropolis algorithm.