Ab initio study of the intrinsic exchange bias at the SrRuO${}_{3}$/SrMnO${}_{3}$ interface

Ab initio study of the intrinsic exchange bias at the SrRuO $_{3}$ /SrMnO $_{3}$ interface

Shuai Dong, Qinfang Zhang, Seiji Yunoki, J.-M. Liu, and Elbio Dagotto

Phys. Rev. B 84, 224437 – Published 29 December 2011

Abstract

In a recent publication [S. Dong et al., Phys. Rev. Lett. 103, 127201 (2009)], two (related) mechanisms were proposed to understand the intrinsic exchange bias present in oxides heterostructures involving G-type antiferromagnetic perovskites. The first mechanism is driven by the Dzyaloshinskii-Moriya interaction, which is a spin-orbit coupling effect. The second is induced by the ferroelectric polarization, and it is only active in heterostructures involving multiferroics. Using the SrRuO $_{3}$ /SrMnO $_{3}$ superlattice as a model system, density-functional calculations are here performed to verify the two proposals. This proof-of-principle calculation provides convincing evidence that qualitatively supports both proposals.

Received 25 October 2011

DOI:https://doi.org/10.1103/PhysRevB.84.224437

Authors & Affiliations

Shuai Dong^1,2, Qinfang Zhang^3,4,5, Seiji Yunoki^4,5,6, J.-M. Liu^2,7, and Elbio Dagotto^8,9

¹Department of Physics, Southeast University, Nanjing 211189, China
²National Laboratory of Solid State Microstructures, Nanjing University, Nanjing 210093, China
³Key Laboratory for Advanced Technology in Environmental Protection of Jiangsu Province, Yancheng Institute of technology, Yancheng 224051, China
⁴Computational Condensed Matter Physics Laboratory, RIKEN, Wako, Saitama 351-0198, Japan
⁵CREST, Japan Science and Technology Agency (JST), Kawaguchi, Saitama 332-0012, Japan
⁶Computational Materials Science Research Team, RIKEN AICS, Kobe, Hyogo 650-0047, Japan
⁷International Center for Materials Physics, Chinese Academy of Sciences, Shenyang 110016, China
⁸Department of Physics and Astronomy, University of Tennessee, Knoxville, Tennessee 37996, USA
⁹Materials Science and Technology Division, Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831, USA

Article Text (Subscription Required)

Click to Expand

References (Subscription Required)

Click to Expand

Issue

Vol. 84, Iss. 22 — 1 December 2011

Reuse & Permissions

Access Options

Author publication services for translation and copyediting assistance advertisement

Images

Figure 1
Sketch of possible FM-AFM interfaces. The upper/lower (blue/red) layers denote FM/AFM components, respectively. Local magnetic moments are marked as arrows. (a) A fully uncompensated AFM interface, as it occurs for example in an (001) surface of A-type AFM state or (111) surface of G-type AFM state. (b) and (c) Fully compensated AFM interfaces, as they occur in a G-type AFM state. In (a) and (b), the FM and AFM magnetic moments are collinear. In (c), the FM and AFM magnetic moments are noncollinear, namely, the magnetic easy axes of the FM and AFM materials are different.Reuse & Permissions
Figure 2
(a) Optimized superlattice crystal structure (drawn by vesta).[31] The space group is $P m c 2_{1}$ (orthorhombic). The in-plane lattice constants $a$ and $b$ are fixed as $3.905 \sqrt{2}$ Å to match the SrTiO $_{3}$ substrate. The optimized $c$ -axis constant is $15.5$ Å. The pseudocubic $x$ - $y$ - $z$ axes are shown. The collective rotations/tiltings of oxygen octahedra are obvious to the eye, giving rise to bended Mn-O-Mn, Mn-O-Ru, and Ru-O-Ru bonds. (b) The optimized Mn-O-Ru bond angles $α$ are $170 . 4^{\circ}$ , while the bending directions between nearest-neighbor pairs are staggered. The $\vec{D}$ 's point “in” ( $⊙$ ) and “out” ( $\otimes$ ) for these two bonds respectively. (c) and (d) When an (in-plane) FE polarization $\vec{P}$ (the interfacial oxygens displaced along the [110] direction by $0.005 \sqrt{2}$ Å) is imposed to the original bending bonds, then the bond angles become modulated: $β = 168 . 5^{\circ}$ and $γ = 171 . 6^{\circ}$ . Here, the difference between $β$ and $γ$ is enlarged for a better visual comparison.Reuse & Permissions
Figure 3
Sketches of interfacial noncollinear spin configurations after the self-consistency calculation reaches convergence. Here, the initial spins ${\vec{S}}_{Mn}$ 's are along the (100) direction while ${\vec{S}}_{Ru}$ 's are along (001). All output spins show some canting, but small in value, from the initial axes. The canting components are shown in the insert boxes, where the lengths of arrows are in proportional to the values of canting momenta. (a) Results without SOC. The canting angles of ${\vec{S}}_{Mn}$ 's are about $4^{\circ}$ while those of ${\vec{S}}_{Ru}$ 's are about $5^{\circ}$ . After the flips of ${\vec{S}}_{Ru}$ 's (from left to right in the figure), all canting directions also flip symmetrically. (b) Results with SOC. The canting angles of ${\vec{S}}_{Mn}$ 's are about $3^{\circ}$ (left) or $4^{\circ}$ (right), while those of ${\vec{S}}_{Ru}$ 's are about $9^{\circ}$ (left) or $4^{\circ}$ (right). The canting angles (directions and values) are $n o$ longer symmetrical after the flips of ${\vec{S}}_{Ru}$ 's, which also suggests a bias due the SOC effect.Reuse & Permissions

Physical Review B

covering condensed matter and materials physics