Ultrathin scale tailoring of anisotropic magnetic coupling and

of great importance to explore the new fundamentalism in spintronics4-6. It has the potential to pave the way for a new paradigm shift in modern spint...
0 downloads 0 Views 3MB Size
Download PDF
Research Article www.acsami.org

Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Ultrathin Scale Tailoring of Anisotropic Magnetic Coupling and Anomalous Magnetoresistance in SrRuO3−PrMnO3 Superlattices Antarjami Sahoo,† Wilfrid Prellier,‡ and Prahallad Padhan*,† †

Department of Physics, Indian Institute of Technology Madras, Chennai 600036, Tamil Nadu, India Laboratoire CRISMAT, CNRS UMR 6508, ENSICAEN, 6 Bd du Marechal Juin, F-14050 Caen Cedex, France

‡

Downloaded via UNIV OF WINNIPEG on December 9, 2018 at 00:58:30 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

S Supporting Information *

ABSTRACT: A strong perpendicular magnetocrystalline anisotropy (PMA) in antiferromagnetically coupled SrRuO3(17 uc (unit cell))/PrMnO3(n uc) superlattices effectively reconstructs the interfacial spin ordering. The occurrence of significant anisotropic interfacial antiferromagnetic coupling between the Ru and Mn ions is systematically tuned by varying the PrMnO3 layer thickness in ultrathin scale from 3 to 12 uc, which is associated with a rise in PMA energy from 0.28 × 106 to 1.60 × 106 erg/cm3. The analysis using the Stoner−Wohlfarth model and density functional theory confirm that the exchange anisotropy is the major contribution to the PMA. The superlattices with PrMnO3 layer thickness ≥7 uc exhibit the tunneling-like transport of Ru 4d electrons, which is rather expected in the stronger antiferromagnetically coupled superlattices with thinner PrMnO3 layer. Tunneling-like transport at thicker spacer layer in the SrRuO3−PrMnO3 superlattice system is an unique feature of two ferromagnet-based superlattices. Our investigations show that the technologically important interfacial magnetic coupling, PMA, and tunneling magnetoresistance could be achieved in a periodically stacked bilayer and can be precisely manipulated by the size effect in ultrathin scale. KEYWORDS: interface engineering, perpendicular magnetic anisotropy, tunneling magnetoresistance, anisotropic exchange coupling, superlattice

1. INTRODUCTION Data storage via spin assembly is one of the key features of the information technology sector. The innovative technology to manipulate the spin degrees of freedom in ultrathin heterostructures can be held accountable for the proliferation of interest in spintronics. The realization of anisotropically oriented spins of the same material distanced by a spacer layer has found application in tunneling magnetoresistance (TMR)and giant magnetoresistance (GMR)-based hard disk drives.1 Performance of data storage and the energy efficiency with ultralow-power consumption of electronic devices have been increasing every year.1−3 The surgical control of intriguing spin physics in highly stabilized oxide heterostructures has been of great importance to explore the new fundamentalism in spintronics.4−6 It has the potential to pave the way for a new paradigm shift in modern spintronics-based devices. 4−6 Although the interfacial magnetic couplings for certain combinations of oxides have been widely explored,7−9 the origin and systematic experimental studies of the reconstructed interfacial spin structure are yet to be established. The preferred orientation of magnetization along some crystallographic axis and/or with the geometrical shape of the body is known as magnetic anisotropy, which accounts for a small amount of correction to the total magnetic energy after introducing the relativistic correction to the Hamiltonian of the system. This correction evolves due to the dipole−dipole © XXXX American Chemical Society

interaction and spin−orbit coupling and breaks the rotational invariance with respect to the spin quantization axis. The magnetic random access memory (MRAM), a subclass of nonvolatile memories, has been a breakthrough discovery in the field of spin electronics.3 More specifically, the spin transfer torque MRAM (STT−MRAM) has the potential to be the part of industrial application because of its high data retention capability and low write current.10−12 The STT−MRAM works on the basic principle of perpendicular magnetic anisotropy (PMA), originated due to strong spin−orbit coupling, which was initially reported in Co/Pt and Co/Pd multilayers.13 The large PMA observed due to the hybridization of dz2 orbitals with sp-oxygen orbitals at the interfaces of transition metals and oxygen is reported to be a potential candidate for the fabrication of out-of-plane magnetic tunnel junction for STT−MRAM.14 The suppression of oxygen octahedral rotation by increasing the Sr content in La1−xSrxMnO3/SrIrO3(LSMO/SIO) superlattice induces a PMA as large as 4 × 106 erg/cm3 because of the change in the strength of interfacial exchange interaction.15 Further, a uniaxial in-plane magnetic anisotropy is observed in amorphous CoFeB thin film grown on GaAs substrate, where the interfaceinduced anisotropy carries itself into the volume with the Received: October 9, 2018 Accepted: November 26, 2018 Published: November 26, 2018 A

DOI: 10.1021/acsami.8b17385 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 1. (a) Reciprocal space mapping around (001)pc of (001)STO/[17 uc SrRuO3/5 uc PrMnO3]15. (b) Reciprocal space mapping around (103)pc of (001)STO/[17 uc SrRuO3/n uc PrMnO3]15 superlattices with n = 4, 5, 6 7, 8, and 10. (c) In-plane lattice parameter of different superlattices. Atomic force microscopy image of the superlattice with n = 4 (d) and n = 12 (e).

increase in film thickness.16 Strain relaxation due to the presence of a buffer layer underneath the Y3Fe5O12 film has also been reported to be responsible for high PMA.17 Moreover, it is essential to investigate the diverse sources of origin of PMA in different heterostructures because of its everincreasing technological importance for the upgradation of modern storage devices. The presence of robust interfacial exchange coupling, which is very sensitive to the oxygen octahedral ordering and rotation in manganite/ruthenite superlattices,15,18 makes them ideal candidates to explore the magnetic anisotropy and the related derived properties. Here, in this article, we have reported the presence of a very high PMA in SrRuO3(SRO)/PrMnO3(PMO) superlattices, which can be tuned by decreasing the PMO layer thickness. The anisotropy in magnetoresistance and magnetization have been explained by the size-dependent strength of the interfacial antiferromagnetic (AFM) coupling.

The density functional theory (DFT)-based calculations are carried out by using the Slater exchange and Perdew−Zunger parametrization of the local density approximation (LDA) with the projected augmented wave (PAW) method as implemented in Quantum Espresso.19 The interface model to a supercell consists of 40 atoms with (001)-oriented tetragonal (I4/mcm) SRO−PMO bilayer using the crystal parameters obtained from the X-ray diffraction. The calculations are performed with strain structure using a plane-wave energy cutoff of 130 Ry and a 8 × 8 × 8 Monkhorst−Pack k-point mesh.

3. RESULTS AND DISCUSSION The θ−2θ XRD patterns of all of the superlattices confirm the high crystallinity, epitaxial, and (00l)-oriented unit cells.18 The superlattice period (Λ) of this series of superlattices calculated from the satellite peak positions obtained from the XRD patterns confirms the desired stacking fabrication of the superlattices.18 The reciprocal space mappings (RSM) constructed from the Bragg’s reflection measurements consisting of 2θ−ω coupling scans in (001)pc for n = 5 and (103)pc of various superlattices for different ω values are shown in Figure 1a,b, respectively. The RSM pattern of the superlattice with n = 5 shows the presence of (001) Bragg’s reflection from STO along with the signature satellite fringes of the superlattice up to second order, which are symmetrically present on either side of the zeroth-order reflection, in both position and intensity. The vertical alignment of all of the reflections, appearance of the cocentered contours, and gradual reduction of intensities of the fringes indicate the coherently grown highly epitaxial nature of the superlattices. The (103) Bragg’s peak positions of STO are same for all of the superlattices, whereas those of the zeroth-order Bragg’s peak of the superlattices are getting gradually shifted toward the peak position of STO with the increase in the PMO layer thickness, indicating the size effect-induced systematic strain relaxation. The RSM validates the stabilization of tetragonal crystal symmetry in these superlattices, as reported earlier.18 The

2. EXPERIMENTAL AND COMPUTATIONAL METHODS The growth of highly oriented epitaxial [17 − uc (unit cell)SrRuO3/n(= 3, 4, ..., 12) − uc PrMnO3]15/(001) SrTiO3(STO) has been reported elsewhere.18 The out-of-plane X-ray diffraction (XRD) pattern and reciprocal space mapping were recorded by a four-cycle X-ray diffractometer. Atomic force microscopy (AFM) was employed to analyze the surface morphology and roughness. The temperature and field-dependent magnetization measurements for both in-plane and out-of-plane orientations of magnetic field have been performed by superconducting quantum interference device-based vibrating sample magnetometer, whereas the anisotropic magnetoresistance has been studied by the physical property measurement system. The in-plane and out-of-plane resistivities of these superlattices were measured using four probe technique. An in-plane constant DC current was passed through the current probes, and the voltage drop across the voltage probes were measured in the presence of the in-plane- and out-of-plane-oriented magnetic field. B

DOI: 10.1021/acsami.8b17385 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces same ω position of the substrate and sample peaks infer the pseudomorphic growth. Hence, the in-plane lattice of the superlattices is clamped (coherently strained) by the substrate and [001]t orientation is favored. The in-plane pseudocubic lattice parameter (aPC) extracted from the RSM pattern decreases gradually with the increase in n (see Figure 1c). The SRO (aPC = 3.93 Å), stacked on PMO (aPC = 3.85 Å), experiences in-plane biaxial tensile stress, which gets more strained as the PMO thickness increases, and hence, aPC decreases. This also infers the layer-by-layer coherent growth and desired artificial fabrication of the superlattices. The presence of Keissg’s fringes around the sample peak positions can also be seen from the two-dimensional RSM patterns, further elucidating the high crystallinity of the superlattices. The topographic images of the surface morphologies studied by atomic force microscopy for the superlattices with n = 4 and 12 are shown in Figure 1d,e, respectively. The root-mean-square (RMS) roughness for the superlattices with n = 4 and 12 are found to be 5.27 and 17.43 Å, respectively, indicating the layerby-layer growth of the unit cells with atomically smooth superlattice surface. The increase in RMS roughness for the superlattice with n = 12 is consistent with the thicknessdependent roughness variation20 due to the lattice mismatch of SRO and PMO. The 0.05 T field-cooled (FC) temperature-dependent magnetization M(T) of the superlattices with n = 3, 4, 5, and 12 for both in-plane ([100]PC) and out-of-plane ([001]PC) orientations of 0.05 T magnetic field are plotted in Figure 2a−d.

interfacial Ru spins, dragged by the AFM coupling energy, align at a canted angle with Mn spins at T < TN. As the PMO layer thickness increases, net ferromagnetic moment is resulted in the PMO layer by the canted spins of Mn3+ ions.21 The cumulative FM ordering in SRO and PMO dominates the AFM ordering at the interfaces, and it leads to the absence of TN for n > 6. The interfacial antiferromagnetic ordering is also influenced by the Zeeman energy, which is confirmed from the FC M(T) measured using different cooling fields. As the cooling field increases, the TN of the superlattice with n = 3 gradually decreases (see Figure 2e) because the Zeeman interaction becomes larger than the antiferromagnetic exchange coupling at the interfaces. Similar competition of the AFM coupling and Zeeman energy is also observed for the superlattice with n = 4. The presence of TN for higher cooling field (∼6 T) reflects the strongly antiferromagnetically coupled interface in the superlattices with n = 3 and 4, which is suppressed in other superlattices. For [001]PC-oriented field, the Ru spins of the superlattices with n > 3 align along the field direction and the Mn spins are dragged by AFM coupling energy for T < TN. However, with the increase in cooling field, the AFM coupling energy at the interfaces is suppressed by the Zeeman energy (not shown in figure). The absence of TN for the superlattice with n = 3 could be attributed to the larger spin−orbit coupling in strained ultrathin PMO,22 which can suppress the exchange couplinginduced spin canting of Mn ions. The spins of Mn ions of the superlattices with thinner PMO layer largely contribute to the interfaces and are strongly coupled with spins of Ru ions via Ru− O−Mn bond in tetragonally stabilized crystal symmetry for inplane orientation of magnetic field. But, as the thickness of PMO increases, the strong exchange interaction between the spins of 3d electrons of Mn ions weakens the interfacial coupling. The 0.05 T field applied along the [001]PC direction may not be sufficient to orient the spins of Mn ions along that direction as it is the hard axis of magnetization for PMO and results in a very weak FM moment and it is dominated by the interfacial AFM coupling in the low-temperature region. The zero-field-cooled (ZFC) field-dependent magnetization (M(H)) is measured at 20 K for both [100]PC and [001]PC orientations of magnetic field for all of the superlattices. The ZFC M(H) for n = 3 and 12 are depicted in Figure 3a,b,

Figure 2. Temperature-dependent 0.05 T field-cooled in-plane and out-of-plane magnetization of the superlattices with (a) n = 3, (b) n = 4, (c) n = 5, and (d) n = 12. (e) In-plane cooling field-dependent Néel temperature of the superlattice with n = 3.

The superlattices undergo a paramagnetic to ferromagnetic (FM) phase transition around the Curie temperature (TC), which corresponds to the transition temperature of SRO. The in-plane magnetization of the superlattice with n = 3 increases with the decrease in temperature below TC up to ∼100 K, followed by a gradual decrease down to the lowest temperature. The drop in magnetization at the temperature which we have marked TN is a signature of ferromagnetic-to-antiferromagnetic transition. Qualitatively, similar M(T) with a lower TN is observed for the superlattices with n = 4−6. The appearance of TN in M(T) is attributed to the antiferromagnetic exchange coupling between the spins of Mn and Ru at the interfaces. The antiferromagnetic exchange coupling at the interfaces is not detected in the M(T) of the superlattices with n > 6 for [100]PC orientation of the magnetic field. As the orientation of the external magnetic field changes to [001]PC, a drop in magnetization is observed for all of the superlattices below a significantly lower TN ∼ 30 K, except for the superlattice with n = 3. Both the SRO and PMO exhibit ferromagnetic behavior18 with their easy axes of magnetization along the [001]PC and [100] PC directions, respectively. Thus, for the [100] PC orientation of magnetic field, in the superlattices with n = 3− 6, the Mn spins align along the field direction, whereas the

Figure 3. Field-dependent zero-field-cooled (ZFC) in-plane and outof-plane magnetization of superlattices with (a) n = 3 and (b) n = 12. (c) In-plane and out-of-plane loop opening of the ZFC magnetic hysteresis loops.

respectively. All of the hysteresis cycles reveal the presence of two-step magnetization reversal mechanism with soft PMO reversing first followed by the hard SRO. For [100] PC orientation of magnetic field, the Mn spins switch at a lower field because of the low coercive field of PMO and favorable AFM coupling nature. As the field increases, the Zeeman energy overcomes the AFM coupling energy and Ru spins switch at a very high field (HC ∼ 3 T) compared to the coercive field of SRO thin film, which demonstrates the presence of very strong AFM C

DOI: 10.1021/acsami.8b17385 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

× 106 erg/cm3 for n = 3, and it increases with n up to n = 10 and saturates beyond that (see Figure 4b). The observed PMA is larger than that observed in some metal/metal oxide heterostructures3 and comparable to the Keff of LSMO/SIO superlattice.15 This SRO−PMO superlattice system shows an increase of Keff, ΔKeff ≈ 1.32 × 106 erg/cm3 by increasing the PMO layer thickness ∼2.7 nm. The orthorhombic SRO has a strong magnetocrystalline anisotropy with its easy axis tilted at 45° in between the [100]PC and [001]PC directions. But, it is relatively easier to magnetize SRO along the [001]PC direction compared to the [100]PC direction for both the orthorhombic and tetragonal crystal symmetries.23 In contrast, in the SRO− PMO superlattices, the [100]PC direction is magnetically more easy than the [001]PC direction (see Figure 3a,b). In addition, the thickness of the SRO in the superlattice series is constant and the magnetocrystalline contribution of the SRO to the PMA of each superlattice is expected to be same. On the other hand, the manganites are known to be soft magnets with relatively very weak magnetic anisotropy24 and thus the contribution of PMO to the total large magnetic anisotropy of this strongly coupled superlattice system can be negligible. The Stoner−Wohlfarth model for the thin films has been adopted to interpret the variation of Keff with the PMO layer thickness, where the magnetization reversal in a single domain is considered.25 The effective anisotropy can be expressed as17

coupling. The qualitative features of the minor loops and HC values of all of the superlattices for in-plane orientation of the magnetic field are similar. In addition, these superlattices exhibit same hysteresis loops for the [010]PC-oriented magnetic field, which is consistent with the observed tetragonal crystal symmetry18 (a0a0c− glazer notation), where the BO6 octahedral rotation around the in-plane crystallographic axes ([100]PC, [010]PC) is absent. The hysteretic nature significantly changes with the rotation of magnetic field from [100]PC to [001]PC direction. The HC at which the Ru spins switch for [001]PCoriented magnetic field is lower than that for the [100]PC direction, which infers the lowering of AFM coupling strength with the change in the orientation of the magnetic field. Superlattices with lower PMO layer thickness do not exhibit distinguishable minor loops because of the weak AFM coupling along [001]PC. As the thickness of PMO increases, the ferromagnetic ordering in PMO gets stabilized and the minor loop starts appearing for the out-of-plane M(H), due to the low coercivity of PMO. The loop opening, which is proportional to the biased SRO layer thickness, systematically decreases from 19 to 6% and from 50 to 13% for in-plane and out-of-plane orientations of magnetic field, respectively (see Figures 3c and S2), concluding that the interfacial magnetic coupling and the derived physical properties can be effectively tuned by controlling the spacer layer thickness in ultrathin scale. A perpendicular magnetoanisotropy (PMA) is observed for all of the superlattices, whereas the in-plane magnetic anisotropy is absent, consistent with the a0a0c− type of octahedral rotation pattern. The PMA of these superlattices is represented by the area enclosed by the in-plane and out-of-plane first quadrant M(H)s (see Figure 4a). The perpendicular magnetoanisotropic

Keff = Kcrystalline + K induced + K shape + Kexchange

where the crystalline anisotropy, strain-induced anisotropy, shape-induced anisotropy, and anisotropy evolved because of the exchange interaction between the spins of Ru and Mn ions are represented by Kcrystalline, Kinduced, Kshape, and Kexchange, respectively. These anisotropic constants are essentially determined by the nature of rotation of spins of Ru and Mn ions. To investigate the origin of the change in the Keff (ΔKeff), the effect of PMO layer thickness on different anisotropic contribution was studied. ΔKcrystalline (change in the Kcrystalline) and ΔKinduced (change in the Kinduced) depend on the change in the orbital overlapping, which in turn could change the spin− orbit coupling and hence, the anisotropy.26 Since the superlattices with n = 3−12 exhibit tetragonal crystal structure, the contribution of ΔKcrystalline is negligible. We have performed the first-principles calculations for the superlattices with n = 3 and 10 with the refined lattice parameters obtained from RSM and out-of-plane XRD patterns.27 The density of states (DOSs) for the interfacial Ru and Mn ions are plotted in Figure 4c,d, respectively. The presence of high DOSs for both the spin channels (majority and minority) at Fermi level (EF) for Ru ion, mainly reflecting the t2g character, confirms the metallic nature of SRO.28 Strong hybridization of its eg orbitals with oxygen is responsible for the presence of DOSs at the band edge (∼8 eV). However, a significantly lower DOSs at EF for PMO infers its weak metallic character, which is consistent with the previous results.29 Contrary to Ru, the electron distribution in the minority spin channel is significantly lowered compared to the majority spin channel for Mn ions. The similar DOSs have been observed for both the superlattices, indicating that the subtle structural deformation is not affecting the orbital overlapping in larger picture. Hence, ΔKinduced can be neglected. Further, the magnitude of the strain-induced anisotropy is found by using the standard expressions30

Figure 4. (a) First-quadrant field-dependent hysteresis loops of the superlattices with n = 3 and 12. The shaded regions show the perpendicular magnetic anisotropy. (b) Effective anisotropic constant of different superlattices. Density of states of Ru (c) and Mn (d) ions of the SRO−PMO bilayer simulated using the lattice parameters of the superlattice with n = 3 and 10. (e) Saturated magnetization of various superlattices extracted from the hysteresis loops and calculated from the spin-only moment of SRO and PMO. (f) First-quadrant fielddependent hysteresis loops with in-plane field (top curves for n = 3 and 12) and out-of-plane field (bottom curves for n = 3 and 12) of the superlattices.

constant (Keff) values of all of the superlattices were calculated using15

∫0

Keff = μ

HS

∫0

MIP dH − μ

HS

MOP dH

(2)

(1)

where MOP and MIP are the out-of-plane and in-plane magnetizations, respectively, and HS is the corresponding saturation field. The Keff represents the shaded area between the MOP and MIP, as shown in the Figure 4a.15 The PMA is 0.28

Han = 3λσ /MS , and K induced = 3λσ /211 D

(3)

DOI: 10.1021/acsami.8b17385 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces where σ is the biaxial strain experienced by the superlattice, calculated using the lattice parameters obtained from RSM, Han is the anisotropy field, which is assumed to be the saturation field for the magnetic field oriented along the hard axis i.e., [001]pc, and MS is the saturation magnetization. The Kinduced for the superlattices with n = 3 and 10 are found to be 2.24 × 103 and 4.47 × 103 erg/cm3, respectively, which are negligible compared to Keff. Hence, ΔKinduced ≈ 2.23 × 103 erg/cm3 does not play an important role in the rise of PMA. The shape anisotropy of the superlattices is evaluated by using the expression17 K shape =

2πMS2

Figure 5. Field-cooled (0.1 and 5 T) in-plane (a) and out-of-plane (b) minor hysteresis loops of the superlattice with n = 12.

magnetic field for all of the superlattices.18 For low Hcool, the interfacial AFM coupling energy opposes the switching of Mn spins, which are oriented along the in-plane field, results in the alignment of the interfacial Ru spins at an angle δ > 90° (δ is the angle between the spins of Ru and Mn ions), and leads to the negative exchange bias. However, for high in-plane Hcool, the interfacial AFM coupling energy is suppressed by the Zeeman energy, results in the alignment of the interfacial Ru spins at an angle δ < 90° with the spins of Mn ions, and leads to the positive exchange bias.18 The interfacial spin glass state has been held responsible for the exchange bias shift observed in LSMO/ SrMnO332 and LSMO/La2CuO433 heterostructures. The presence of a drop in magnetization in the field-cooled M(T) below TN, very large switching field for SRO in M(H) (∼3 T), and the strong AFM coupling lead crossover of negative exchange bias to positive exchange bias18 infer that the interfacial glassy state has been considerably suppressed by the strong interfacial AFM coupling in SRO−PMO superlattices. As the orientation of magnetic field switches to the perpendicular direction, the minor loop shifts along the magnetization axis (see Figure 5b). The vertical shift, i.e., incomplete reversal of Ru spins,34 can be ascribed to the pinning of Ru spins due to the spin−orbit coupling of SRO rather than the observed in-plane exchange biasing, mediated by the AFM coupling. The vertical shift of the minor loops could be attributed to the stronger spin− orbit coupling of SRO.35,36 An appreciable amount of vertical shift is absent for in-plane orientation of magnetic field inferring the shape-dependent anisotropic spin−orbit coupling in SRO. The ZFC field-dependent anisotropic magnetoresistance (AMR) of different superlattices are calculated by using the relation37

(4)

where MS is the saturation magnetization of the superlattices. Figure 4e shows the MS of these superlattices determined after the diamagnetic corrections for the substrate from M(H)s and also the calculated MS for the spin-only moments of SRO (1.55 μB/Ru)18 and PMO (3.54 μB/Mn).18 The MS of the superlattice with n = 3 is significantly lower than the theoretical value. However, as “n” increases, MS increases, and becomes closer to 1.6 μB/uc for n ≈ 6, which indicates the stabilization of weak ferromagnetic PMO around n = 6. Using the MS of the superlattices with n = 3 and 10, ΔKshape (Kshape(n = 10) − Kshape(n = 3)) is found to be ≈1.6 × 105 erg/cm3, which is 1 order less than ΔKeff. Thus, ΔKexchange = ΔKeff − (ΔKcrystalline + ΔKinduced + ΔKshape) ≈ 106 erg/cm3, which confirms that the dominant contribution to the rise in PMA is the variation in the exchange interaction between the SRO and PMO at the interfaces. Further, the rotation of spins of Ru and Mn ions in the first quadrant is almost similar for all of the superlattices for the outof-plane-oriented magnetic field (see Figure 4f). The AFM coupling strength decreases with the increase in n for the inplane orientation of magnetic field.18 The stronger AFM coupling for the superlattice with n = 3 triggers the rotation of spins of Mn ions at a relatively higher positive field compared to that of the superlattice with n = 12 (see Figure 4f) for in-plane orientation of field and thus the increase of ΔKeff. Hence, the variation in exchange interaction between Ru and Mn ions during the field sweep can be held responsible for the rise in PMA. The AFM coupling strength depends on the orbital occupancy of the dZ2 orbitals of the interfacial Ru and Mn ions as the superexchange interaction arises between two vacant dorbitals mediated by the oxygen.31 Fascinatingly, the dZ2 orbital occupancy of Ru and Mn ions calculated from the partial density of states increases only by