betaFeedback

WorksPhysics and Astronomy

Rashba splitting in polar-nonpolar sandwich heterostructure : A DFT Study

Abstract

In this study, we employ density functional theory (DFT) based first-principles calculations to investigate the spin-orbit effects in the electronic structure of a polar-nonpolar sandwich heterostructure namely LAO2.5_{2.5}/STO5.5_{5.5}/LAO2.5_{2.5}. Our focus on the Ti-3d bands reveals an inverted ordering of the STO-t2g\rm t_{2g} orbital near the n-type interface, consistent with earlier experimental work. In contrast, toward the p-type interface, the orbital ordering aligns with the natural ordering of STO orbitals, influenced by crystal field splitting. Interestingly, we have found a strong inter-orbital coupling between t2gt_{2g} and ege_g orbital, which has not been reported earlier in SrTiO3\rm SrTiO_3 based 2D system. Additionally, our observations highlight that the cubic Rashba splitting in this system surpasses the linear Rashba splitting, contrary to experimental findings. This comprehensive analysis contributes to a refined understanding of the role of orbital mixing in Rashba splitting in the sandwich oxide heterostructures.

Introduction

The pursuit of effective methods for generating and regulating spin-polarized electric currents is a paramount objective within the realm of spintronics, a field devoted to creating resilient, energy-efficient electronic devices that demonstrate non-volatile behaviour [1]. Fundamental breakthroughs in understanding spin-dependent phenomena in bulk semiconductors have been achieved through the exploration of Rashba spin-orbital coupling (RSOC) [2, 3, 4, 5, 6]. In a solid, the Rashba effect emerges from the interplay of intrinsic spin-orbit coupling (SOC) and the disruption of structural inversion symmetry. This interplay locks the spin and momentum degrees of freedom, leading to the elimination of spin degeneracy in the bands [7, 8, 9, 10]. Moreover, the mixing of orbitals also plays a crucial role in determining the nature and the magnitude of Rashba spin-splitting (RSS) along with the strength of SOC in the system [11]. This aspect opens up the realm of spin-orbitronics. [12, 13].

The nature of RSOC makes it possible to control and tune the electron spin with only an external electric field [14]. Furthermore, it has the advantage of an improved interconversion between spin and charge currents [15], exhibiting phenomena like spin Hall effect [16, 17], Edelstein and inverse Edelstein effects [18]. These properties have a significant impact on the advancement of quantum computers based on semiconductor quantum dots [19] and have left an indelible mark on various domains within condensed matter physics [20]. The field of spin-orbitronics started with semiconductors [21] and metals [22]. Later on it expanded its domain in the non-centrosymmetric bulk oxides, such as BiAlO3_3 [23], KTaO3_3 [24], KIO3_3 [25], PbTiO3_3 [26]. Although the evidence of RSOC has been found in these bulk systems, the inherent requirement of ferroelectricity severely restricts the number of such materials. Moreover, a bulk system is not very suitable from the point of view of the fabrication of modern devices. Oxide-heterostructures, on the other hand, provide a much more flexible platform to explore the RSOC along with a plethora of unique features [27, 28, 29]. The long spin lifetimes and modest RSOC seen on oxide surfaces and interfaces render them appealing candidates for spin-orbitronic devices [30]. In particular, SrTiO3_3 has become one of the most widely studied materials after the detection of two-dimensional electron gas (2DEG) in its surface [31] and interface with LaAlO3_3 (LAO) [32]. 2DEG could also be found in a slab [33], thin film [34], or at the surface [35] of SrTiO3_3. This 2DEG could be metallic [36] or semiconducting [32] in nature, such as forming n-type and p-type interfaces [32, 37]. The bulk strontium titanate (STO) exhibits octahedral symmetry [38] with a band gap of ≈3.75\approx3.75 eV [39]. The presence of time-reversal and inversion symmetries leads to a two-fold spin degenerate band structure in the bulk STO. This doubly degenerate nature of the STO band structure remains unaltered in the presence of intrinsic spin-orbit coupling. However, the introduction of a surface [40] or interface with other transition metal oxides [41, 42, 43, 44] disrupts inversion symmetry, leading to the lifting of this spin degeneracy. The internal electric field in this set of structures can be tuned by adjusting the slab thickness or by engineering the surface and the interface. The tuning of the internal electric field can also be achieved by another alternative method, in which an STO slab could be sandwiched between a polar material on either side that is arranged in a manner to generate the internal electric field in the nonpolar STO.

A recent experimental study suggested that the manipulation of the internal electric field can be achieved by creating two oppositely charged interfaces in LAO/STO/LAO oxide heterostructures [45]. This internal electric field effectively breaks the structural inversion symmetry in STO, leading to the emergence of RSOC. Furthermore, it is claimed that by reducing the STO (0 0 1) slab thickness from 20 u.c to 8 u.c., the RSOC transitioned from a cubic to a linear type. This experimental work also concluded that the RSOC is of linear type in thin heterostructures in contrast to the previously reported cubic RSS in oxide interfaces. Their conclusion is based on the study of systems down to 8 u.c thin STO slab. However, there is no detailed comparison of experimental results with the first-principle study of the nature RSS in such a thin STO slab.

In this article, we conduct a thorough density functional theory (DFT) based first-principle analysis of the nature of Rashba coupling in the LAO2.5_{2.5}/STO5.5_{5.5} (0 0 1)/LAO2.5_{2.5} structure. In particular, we focus on the electronic structure, with an emphasis on the role of orbitals and the spin texture associated with the spin-split bands. Furthermore, we use an effective k⃗⋅p⃗\vec{k}\cdot\vec{p} Hamiltonian to determine the RSOC coefficient and ascertain the nature of the splitting. We find that the experimental conclusion regarding the orbital ordering is in tune with the DFT results, while the nature of Rashba coupling stands in contrast with our analysis.

Computational methods

First-principle density functional theory computations are done with the Quantum Espresso (QE) package (QE) [46, 47]. The Perdew-Burke-Ernzerhof (PBE) [48] functional is used to account for electron exchange-correlation effects. Projector augmented wave (PAW) [49] basis set is used to account for valence electron-core ion interaction with a 75 Ry cut-off energy and 750 Ry cut-off charge density. All ionic relaxation are performed for geometry optimization using the quasi-Newton Broyden-Fletcher-Goldfarb-Shann (BFGS) algorithm [50]. This study uses Γ\Gamma-centered 8×8×18 \times8 \times1 k-point grids, as suggested by Monkhorst and Pack [51], to calculate the first Brillouin zone. Using a minimum thickness of ≥30\geq30 Å vacuum layer in the z-direction (perpendicular to the plane) reduces interaction between adjacent slabs. The total energy convergence criteria for geometrical relaxation are set to be <10−4< 10^{-4} Ry. All the atomic positions are optimized until the interatomic forces are less than 10−310^{-3} Ry/Bohr. The correlated Ti − 3d state has been treated with ortho-atomic Hubbard correction using U = 5 eV and J = 0.64 eV as reported in the earlier studies [52, 53, 54, 55]. To examine collinear magnetism, spin-polarized computations were done on the slab system. Collinear magnetism was found to be missing from the system. Therefore, the slab system with a starting magnetism of 0 μB\mu_B on Ti was used for all subsequent calculations involving spin-orbit interaction.

Structural details

In this study, we investigate the non-stoichiometric LAO/STO/LAO sandwich structure, where two unequal interfaces have been created to achieve tunable electric polarization as suggested in Ref. [45]. STO and LAO are both cubic perovskites with a general configuration of ABO3_3. LAO is composed of a stacked arrangement of LaO+^+ and AlO2−_2^-, whereas, STO are composed with the alternative arrangements of TiO20_2^0 and SrO0^0 layers. Eventually, to keep the stoichiometry intact, two types of interfaces can be formed with these two materials, such as TiO20_2^0 and LaO+^+ forming n-type interface and AlO2−_2^- and SrO0^0, forming p-type interface, as mentioned in the Section I. In the present work, the theoretical simulation is done with a 5.5 u.c slab of STO that has been put on the AlO2−_2^-ended LAO substrate and this TiO2_2-ended STO slab is capped with 2.5 u.c. LAO. In this particular arrangement, the disparity in potential between the two surfaces generates an inherent electric field that disrupts the inversion symmetry of STO and generates an internal electric field within the STO slab, as presented in Figure 1. The in-plane lattice parameter of the asymmetric sandwich structure is fixed as the theoretical lattice parameter of LAO, i.e., 3.813 Å. The theoretical lattice parameter of STO is 3.939 Å, so making the sandwich heterostructure implies a compressive strain on the STO ≈ -3.2 %. In this structural geometry, the STO has a polarization along [0 0 1] with Sr and Ti ions below the oxygen atoms in each SrO and TiO2_2 layers.

Structural model of the LAO/STO/LAO sandwich heterostructure

Figure 1. Structural model for the LAO/STO/LAO sandwich hetero structure with n-type, i.e., LaO+^+/AlO2−_2^- interface and p-type, i.e., AlO2−_2^-/SrO0^0 at both sides of 5.5 u.c STO slab.

Results

Orbital ordering

We explore the stoichiometric LAO/STO/LAO sandwich heterostructure in Figure 1 in the presence of SOC. It is important to note that the electronic structure calculations reveal the system to be non-magnetic. In Figure 2, we present the atomic layer-resolved total density of states (LRIDOS) for each layer in the heterostructure. This figure indicates that due to the interface formation, the bands of the STO move towards the Fermi energy; however, there is no charge accumulation at the interface in the system. Moreover, no metallic state is found throughout the STO slab. In previously reported STO-based heterostructures, the RSS is notably observed in the Ti-3d orbitals [56]. Consequently, our study concentrates on the Ti-3d orbitals within this system. In the current structure, in the absence of SOC, Ti-3d orbitals are split into t2gt_{2g} and ege_g orbitals due to crystal field splitting. Due to slab formation, the t2gt_{2g} levels are further splitted into dxyd_{xy} and dxz/yzd_{xz/yz}. For direct visualization of the t2gt_{2g} and ege_g orbitals, we display the corresponding projected density of states for all TiO2_2 layers in Figure 3. This orbital-resolved representation offers insights into the orbital ordering of the Ti-3d bands originating from each layer. Close to the n-type interface, in the TiO2_2-I and TiO2_2-II, the observed inverted orbital ordering of the t2gt_{2g} orbital aligns with previously reported experimental findings [45]. Interestingly, this inverted orbital ordering of t2gt_{2g} orbital is specific in the vicinity of n-type interface. However, the orbital ordering changes as it moves towards the p-type interface (TiO2_2-III to TiO2_2-VI), which agrees with the reported orbital ordering of the SrTiO3_3 slab [34, 57].

Figure 2. Calculated layer-resolved density of states available for the LAO/STO/LAO asymmetric sandwich structure.

Orbital resolved projected density of states for every TiO₂ layer in the sandwich heterostructure

Figure 3. Orbital resolved projected density of states for every TiO₂ layer in the sandwich heterostructure. The shifting of orbital ordering is presented by the available states near the Fermi energy.

slab [34, 57]. Interestingly, in contrast to the earlier report [45], when SOC is applied, the inverted orbital ordering is only found in the TiO2_2-I layer, and the remaining layers show the usual orbital ordering present in other STO-based interfaces and heterostructures. Moreover, Ti-3d orbitals experience a further splitting in the presence of SOC. In our DFT calculations, we have established that the wave vectors (k⃗\vec{k}) consistently lie in the kxk_x–kyk_y plane, aligning along the (1 0 0) direction between Γ\Gamma and X, and the (1 1 0) direction along the Γ\Gamma–M path. The spin component is zero in the z [0 0 1] direction, corresponding to the polar axis normal to the surface, which is a characteristic feature of the Rashba effect. A comprehensive band structure depicting different spin polarizations is presented in Appendix § 1.3.1. Around the Γ\Gamma, Ti−3dyz/zxd_{yz/zx} bands originating from different layers coincide, making them indistinguishable layer by layer. In the presence of SOC, the true eigenstates of the system are combinations of the original dd−orbitals [58], and hence it is imperative to identify the bands in terms of its total angular momentum quantum number jj and the corresponding mjm_j values. In bulk SrTiO3_3, around the Γ\Gamma point, the original ege_g and t2gt_{2g} bands remain well separated even in the presence of SOC. SOC split the t2gt_{2g} band to a twofold degenerate Γ7\Gamma_7 and a fourfold degenerate Γ8\Gamma_8 level [58]. This fourfold Γ8\Gamma_8 level is identified with ∣3/2,±3/2⟩|3/2,\pm3/2\rangle and ∣3/2,±1/2⟩|3/2,\pm1/2\rangle [57, 59]. The Γ8\Gamma_8 level further splits to Γ6\Gamma_6 and Γ7\Gamma_7 under tetragonal distortion [58]. Surprisingly, however, in our system, we have observed from our data that there is mixing of the original ege_g and t2gt_{2g} bands in the presence of SOC and tetragonal distortion. In Figure 4, we present fat bands illustrating different angular momentum components. Our analysis reveals a significant mixing of orbitals within each band. For instance, in the Γ7\Gamma_7 band, we observe the presence of ∣5/2,±5/2⟩|5/2,\pm5/2\rangle, ∣5/2,±3/2⟩|5/2,\pm3/2\rangle, and ∣3/2,±3/2⟩|3/2,\pm3/2\rangle. Similarly, in the Γ61\Gamma_6{}_1 band is found to be a combination of ∣5/2,±1/2⟩|5/2,\pm1/2\rangle and ∣3/2,±1/2⟩|3/2,\pm1/2\rangle. The combination of ∣j,mj⟩|j,m_j\rangle for all the bands designated in Figure 4 are presented in Table I. Interestingly, this kind of interband coupling has not been reported earlier in SrTiO3_3. However, such interband mixing is not unusual. For example, in reference [60], a significant ege_g and t2gt_{2g} interband coupling has been found in a tantalate monolayer on BaHfO3_3, which arises due to strong confinement that breaks the inversion symmetry maximally. More recently it has been pointed out that such an interband coupling is crucial to explain the RSS in KTaO3_3 heterostructure [1].

Calculated fat band structure with the orbital angular momentum associated with the bands in the SrTiO₃ layers

Figure 4. Calculated fat band structure with the orbital angular momentum associated with the bands in the SrTiO₃ layers. The bands are designated with their level as Γ₇ᵢ and Γ₆ᵢ, as they splitted due to inversion symmetry breaking.

Band level∣j,mj⟩\lvert j,m_j\rangle
Γ₇₁, Γ₇₂, Γ₇₃, Γ₇₅, Γ₇₆, Γ₇₇∣52,±52⟩\lvert\frac{5}{2},\pm\frac{5}{2}\rangle, ∣52,±32⟩\lvert\frac{5}{2},\pm\frac{3}{2}\rangle, ∣32,±32⟩\lvert\frac{3}{2},\pm\frac{3}{2}\rangle
Γ₇₄, Γ₇₈, Γ₇₉, Γ₇₁₀, Γ₇₁₁, Γ₇₁₂∣52,±32⟩\lvert\frac{5}{2},\pm\frac{3}{2}\rangle, ∣32,±32⟩\lvert\frac{3}{2},\pm\frac{3}{2}\rangle
Γ₆₁, Γ₆₂, Γ₆₃, Γ₆₄, Γ₆₅, Γ₆₆∣52,±12⟩\lvert\frac{5}{2},\pm\frac{1}{2}\rangle, ∣32,±12⟩\lvert\frac{3}{2},\pm\frac{1}{2}\rangle

Table I. This table represents the role of total angular momentum quantum number jj and the corresponding mjm_j values present in each band designated in Figure 4.

RSS and spin texture

Next, we proceed to analyse the splitting of band pairs in this structure. To quantify the band splitting and comprehend the impact of symmetry on the spin texture, we employ the k⃗⋅p⃗\vec{k}\cdot\vec{p} Hamiltonian associated with the structure. The introduction of a sandwich structure alters the point group symmetry of SrTiO3_3 to c4vc_{4v}. The high symmetry points of the 2D slab structure are Γ (0, 0, 0), X (0.5, 0, 0), and M (0.5, 0.5, 0). In Figure 4, the levels denoted as Γ₇₁ and Γ₇₅ originate from TiO₂-V and TiO₂-I layers, respectively, while Γ₆₁ and Γ₇₄ are present in all TiO₂ layers. A zoomed-in view of the Γ₆₁, Γ₇₄, and Γ₇₅ bands is presented in Figure 5(a) and Figure 5(b). To study the RSS in these bands, an effective two-band Hamiltonian derived from the k⃗⋅p⃗\vec{k}\cdot\vec{p} model is presented here to characterize the splitted bands at Γ point. The Hamiltonian for the c4vc_{4v} point group [57, 61, 62] is written up to the third order:

Spin-splitting band structure of Ti-t2g orbital originating from TiO2 − I layer

Figure 5. Spin-splitting band structure of Ti-t2g orbital originating from TiO2 − I layer. (a) present the spin-splitted Γ61\Gamma_{61} and Γ74\Gamma_{74} bands along −X–Γ\Gamma–X direction with the projection of electron spins towards y^\hat{y}. (b) represents Γ75\Gamma_{75} band along M–Γ\Gamma–X–M. Red and blue coloured lines represent positive and negative spin polarization along y^\hat{y} direction.

Hc4v=γ(kx2+ky2)+βkz2+αR1(kxσy−kyσx)+αR2kxky(kyσy−kxσx)+αR3(kx3σy−ky3σx)(1)H_{c_{4v}}=\gamma(k_x^2+k_y^2)+\beta k_z^2+\alpha_{R_1}(k_x\sigma_y-k_y\sigma_x)+\alpha_{R_2}k_xk_y(k_y\sigma_y-k_x\sigma_x)+\alpha_{R_3}(k_x^3\sigma_y-k_y^3\sigma_x) \tag*{(1)}

The parameter γ\gamma is related to the effective mass m∗m^{*} by the ∣γ∣=ℏ22m∗|\gamma|=\frac{\hbar^{2}}{2m^{*}} of the associated bands, αR1(kxσy−kyσx)\alpha_{R1}\left(k_{x}\sigma_{y}-k_{y}\sigma_{x}\right) is the linear RSO interaction term, whereas αR2\alpha_{R2} and αR3\alpha_{R3} are the coefficients of cubic RSO interaction, with the cubic RSO interaction term αR2(kxky2σy−kx2kyσx)\alpha_{R2}\left(k_{x}k_{y}^{2}\sigma_{y}-k_{x}^{2}k_{y}\sigma_{x}\right) and αR3(kx3σy−ky3σx)\alpha_{R3}\left(k_{x}^{3}\sigma_{y}-k_{y}^{3}\sigma_{x}\right). kik_{i} (i=x,y,zi=x,y,z) are the components of the wave vector k⃗\vec{k} close to the HS points, and σi\sigma_{i} are the Pauli spin matrices.

To provide a basic characterization of the energy spin splitting, irrespective of the point group, one can model the difference in energy between the two bands of opposite spin, denoted as ΔE±\Delta E_{\pm}, as a third-order polynomial in terms of the wave vector k⃗\vec{k}, specifically, ΔE±(k)=ak+bk3\Delta E_{\pm}(k)=ak+bk^{3}. This polynomial also follows the fitting criteria obtained from the eigenvalue Eqn.1. The detailed derivation of the eigenvalues for different kk-paths is given in Appendix B. For the energy spin splitting along the Γ\Gamma–X path, the coefficient bb is determined as 2αR32\alpha_{R3}, while along the Γ\Gamma–M direction, b=αR2+αR3b=\alpha_{R2}+\alpha_{R3}. In both directions along the kk-path, aa is consistently equal to 2αR12\alpha_{R1}. These findings collectively contribute to generating contour and orbital spin textures around the Γ\Gamma point in the kxk_{x}–kyk_{y} plane. Next, we will apply this method to analyze the Rashba spin splitting (RSS) and the spin texture in the Γ61\Gamma_{61}, Γ71\Gamma_{71}, Γ74\Gamma_{74}, and Γ75\Gamma_{75} bands. To determine the contour for the particular pair of spin-splitting bands, we fit our DFT data numerically with Hamiltonian presented in eqn. 1. The kk-limit is important in determining the spin texture and concentric iso-energy contour. We commence our analysis with the Γ61\Gamma_{61} band pairs of TiO2-V atomic layer. To understand the Rashba spin-splitting. Along Γ\Gamma–X, the effective mass of the Γ61\Gamma_{61} band is determined to be 1.46 m0m_{0}. The linear and cubic Rashba spin-orbit (RSO) coefficients are found to be αR1=0.025\alpha_{R1}=0.025 eVÅ and αR3=−2.16\alpha_{R3}=-2.16 eVÅ3^{3} within a kk-limit of 0.046 Å−1^{-1}. However, along the Γ\Gamma–M direction, spin-

splitting is observed, though it may not strictly adhere to the Rashba-type. This characteristic is further confirmed by the iso-energy spin texture with energy contour in the Γ\Gamma–X–M plane presented in Figure 6(b), where the two bands exhibit different spin-topologies, potentially signifying anisotropic RSOC in the system. The disparity in spin length along the inner band for the entire path suggests a potential helical spin texture. This phenomenon, although not uncommon, has been observed previously in the KTaO3/LaAlO3 heterostructure, as reported in earlier first-principle studies [63]. In the atomic layer TiO2-V, the next band pair under consideration is the Γ74\Gamma_{74} 5(a). The effective mass of the band along Γ\Gamma–X is 1.12 m0m_{0}, and along Γ\Gamma–M is 1.23 m0m_{0}. These bands also display an iso-energy spin texture, featuring two concentric circles with tangent spins rotating in inverse directions, indicative of a cubic Rashba spin-splitting. This is corroborated by our fits of the Γ74\Gamma_{74} bands around the Γ\Gamma point, yielding an isotropic cubic-Rashba parameter of 1.27 eVÅ3^{3} along Γ\Gamma–X and 0.13 eVÅ3^{3} along Γ\Gamma–M. The

Spin texture projected on the $k_x$–$k_y$ plane for three bands

Figure 6. The spin texture derived from DFT data are projected on kxk_x–kyk_y plane for, (a) Γ74\Gamma_{74} bands with energy contour at E−Ef=1.1666E-E_f=1.1666 eV, (b) Γ61\Gamma_{61} bands with energy contour at E−Ef=1.1416E-E_f=1.1416 eV, and (c) Γ75\Gamma_{75} bands with energy contour at E−Ef=1.1915E-E_f=1.1915 eV. The contours are derived by using the Hamiltonian presented in Eqn. (1) along with a reasonable klimitk_{\mathrm{limit}}.

Band Levelkk-pathE−EfE-E_{\mathrm{f}} (eV)m∗m^{*} (m0m_{0})αR1\alpha_{R1} (eVÅ)αR3\alpha_{R3} (eVÅ3^{3})αR3\alpha_{R3} (eVÅ3^{3})
Γ71\Gamma_{71}Γ→X\Gamma\to X0.90350.53.24×10−33.24\times10^{-3}0.43
Γ71\Gamma_{71}Γ→M\Gamma\to M0.90350.54.12×10−34.12\times10^{-3}0.33
Γ61\Gamma_{61}Γ→X\Gamma\to X1.14071.462.55×10−22.55\times10^{-2}-2.16
Γ74\Gamma_{74}Γ→X\Gamma\to X1.16581.122.66×10−22.66\times10^{-2}1.27
Γ74\Gamma_{74}Γ→M\Gamma\to M1.16581.232.67×10−22.67\times10^{-2}0.13
Γ75\Gamma_{75}Γ→X\Gamma\to X1.18520.51.11×10−21.11\times10^{-2}-3.33
Γ75\Gamma_{75}Γ→M\Gamma\to M1.18520.511.77×10−21.77\times10^{-2}-1.34

Table 2. The table represents the effective mass m∗=ℏ22∣γ∣m^{*}=\frac{\hbar^{2}}{2|\gamma|}, linear and cubic Rashba spin-splitting parameter αR1\alpha_{R1}, αR3\alpha_{R3} & αR3\alpha_{R3} for different band levels along Γ\Gamma-X and Γ\Gamma-M path obtained from the two-band fitting with polynomial ΔE±(k)=ak+bk3\Delta E_{\pm}(k)=ak+bk^{3}.

linear RSO coefficient for this band is determined to be 0.027 eVÅ. The prevailing cubic Rashba spin splitting is further exemplified by the DFT spin texture presented in Figure 6(a). The Rashba spin-splitting of the nearest bands to the Fermi energy is denoted as Γ71\Gamma_{71}. The effective mass of the band is 0.5 m00.5\,m_0 around Γ\Gamma-point. The coefficient of linear spin-splitting for Γ71\Gamma_{71} band pair is found to be 3.235×10−33.235\times10^{-3} eVÅ and 4.12×10−34.12\times10^{-3} eVÅ along Γ\Gamma–X and Γ\Gamma–M, respectively. In this Γ71\Gamma_{71} level, the cubic Rashba is more prominent than the linear, and the cubic RSO coefficients are found to be 0.43 eVÅ3^3 and 0.33 eVÅ3^3 along Γ\Gamma–X and Γ\Gamma–M, respectively. The DFT spin texture is presented in Figure 6(c). Finally, we discuss the band Γ75\Gamma_{75}, originating from TiO2_2-I layer, which is the closest to the n-type interface. The effective mass of the band is 0.5 m00.5\,m_0. Along Γ\Gamma–X, it exhibits predominantly cubic Rashba spin-splitting with -3.33 eVÅ3^3 up to 0.15 Å−1^{-1}, while the linear splitting is 1.11×10−21.11\times10^{-2}, retaining linearity in the closest vicinity of the Γ\Gamma point, i.e., up to 0.03 Å−1^{-1}. Along Γ\Gamma–

M, the linear coefficient of RSOC is 1.77×10−21.77\times10^{-2}, and the cubic Rashba coefficient is determined to be -1.34 eVÅ3^3. The summarized results are presented in Table II. For the remaining bands, we do not observe any significant Rashba spin splitting around Γ\Gamma. However, the Γ73\Gamma_{73} band pairs exhibit splitting primarily around the band crossing region, and the Rashba spin splitting is distinctly cubic with αR3=1.86\alpha_{R3}=1.86 eVÅ3^3 along Γ\Gamma–X and α~R3=1.2\tilde{\alpha}_{R3}=1.2 eVÅ3^3 along Γ\Gamma–M. The other Γ6\Gamma_6 levels, located at higher energy regions, do not display Rashba-like spin splitting. Furthermore, no linear Rashba spin splitting is evident in the spin-splitting of the Ti bands.

It is already known that the spin orientation changes as a function of the wave vector k⃗\vec{k} together with the modification of the orbital contributions of the interface states. We derive the projected spin components ⟨Sx⟩\langle S_x\rangle and ⟨Sy⟩\langle S_y\rangle from our density functional theory (DFT) results to explore the potential presence of spin arrangements for a single band of the spin-splitted pair. Precise determination of the RSS is a cumbersome task within specific klimitk_{\mathrm{limit}}, however, the orientation of the spin can be determined properly from the DFT spin texture. However, the length of the spin presented in the one band spin texture are not the real data. Subsequently, we depict the projected spin vectors for a single band as ⟨Sx⟩i^+⟨Sy⟩j^\langle S_x\rangle\hat{i}+\langle S_y\rangle\hat{j} in the kxk_x–kyk_y plane, within a suitable energy range. These figures are presented for one band spin texture in Figure 7(a), (Figure 7b), and (Figure 7c). Figure 7(a) represents the one band spin texture for Γ74\Gamma_{74} band which is pure cubic type RSS around the Γ\Gamma point in the X–Γ\Gamma–M plane. On the contrary, Figure 7(b) portrays the anisotropy of the Γ61\Gamma_{61} band pair. In Figure 7(c), the spins are oriented perpendicular to the Γ\Gamma–M line, which indicates the absence of αR2\alpha_{R_2} in the Hamiltonian of the Γ75\Gamma_{75} bands.

One band spin projection in the $k_x$–$k_y$ plane

Figure 7. One band spin projection in the kxk_x–kyk_y plane derived from the Hamiltonian (Eqn. (1)). (a) and (c) present the outer band projection. (b) presents the inner band projection. The orientation of the spins are taken same as that of the DFT spin orientation (Figure 6). The length of the arrows are not the real length of arrows presented in DFT spin texture.

Conclusion

In conclusion, our investigation into Rashba spin-orbit coupling in the LAO/STO/LAO sandwich heterostructure involves the construction of a model featuring a thin STO slab with 5.5 unit cells. Through DFT-based first-principle calculations, we delve into the intricate role of orbitals within different layers of the STO slab. We find a presence of significant interorbital mixing of t2gt_{2g} and ege_g orbital in the STO slab. We report this phenomenon for the first time in STO-based 2D structure. The identification of this mixing in sandwich structure is pivotal, providing valuable insights for future researchers in modelling and understanding the complexities of the system. Moreover, in this study, we encompass the spin texture for splitted bands to elucidate the nature of Rashba spin splitting in the system, and reveal a predominant cubic Rashba spin-orbit coupling (RSOC). This research not only enriches the understanding of the specific heterostructure but also establishes a foundation for prospective advancements in comprehending Rashba spin-orbit coupling in complex oxide heterostructures.

SOC included band structure

This section provides the band structure for the entire structure (Figure 1) along M–Γ–X–M. The band structure presented in Figure 8 highlights the unique characteristics of the Rashba effect within the system. Figure 8(a), (b) and (c) represent the band structure with spin-polarization along x^\hat{x}, y^\hat{y} and z^\hat{z}. Specifically, the spins exhibit polarization perpendicular to the wave vector. The red and blue solid circles represent the positive and negative spin, whereas the size of the circles represent the weight of the spin associated with the bands. Along z^\hat{z}, the spin weight are present only at the high symmetry points, which is a characteristics of the RSOC in the system. The complete system, including the capping LAO layers, exhibits metallic states. These metallic states may be attributed to either oxygen (O) or lanthanum (La) in the LAO layers, as depicted in Figure 2.

Band structures of the LAO/STO/LAO sandwich model along $M-\Gamma-X-M$

Figure 8. Band structures of the LAO/STO/LAO sandwich model along M−Γ−X−MM-\Gamma-X-M calculated with the spin-orbit interaction. Red and blue coloured circles represent up and down spin polarization. The circle size represents the weight of the spin. (a), (b) and (c) represent the sxs_x, sys_y and szs_z spin-polarization projected along M−Γ−X−MM-\Gamma-X-M.

Effective k.p Hamiltonian

The strength of the RSO interaction can be estimated by solving Eq. (1) separately for Γ–X and Γ–M high symmetry path, represented in the Table II. It is evident that the Hamiltonian of Eq. (1) only exhibits terms that rely on σx\sigma_x and σy\sigma_y, as per symmetry concerns. We confirm from our DFT findings that the spin polarization contribution is zero along the zz [0 0 1] axis for every k⃗\vec{k} vector.

For Γ–X high symmetry path, ky=0k_y = 0. Imposing this condition, we solve Eq. (1) to obtain the following two eigenvalue equations are given by,

ϵ1=γkx2+αR1kx+αR3kx3\epsilon_1 = \gamma k_x^2 + \alpha_{R_1} k_x + \alpha_{R_3} k_x^3
ϵ2=γkx2−αR1kx−αR3kx3.\epsilon_2 = \gamma k_x^2 - \alpha_{R_1} k_x - \alpha_{R_3} k_x^3.

The amount of momentum dependent spin-splitting ΔRX\Delta_{RX} is given by,

ΔRX=ϵ1−ϵ2=2αR1kx+2αR3kx3(2)\Delta_{RX} = \epsilon_1 - \epsilon_2 = 2\alpha_{R_1} k_x + 2\alpha_{R_3} k_x^3 \tag*{(2)}

For Γ–M high symmetry path, kx=kyk_x = k_y, and k∥=kx2+ky2k_{\parallel} = \sqrt{k_x^2 + k_y^2}. By imposing these conditions, solutions for Eq. (1) give two eigenvalue equations as follows,

ϵ1′=γk∥2+αR1k∥+(αR2+αR32)k∥3\epsilon'_1 = \gamma k_{\parallel}^2 + \alpha_{R_1} k_{\parallel} + \left(\frac{\alpha_{R_2} + \alpha_{R_3}}{2}\right) k_{\parallel}^3
ϵ2′=γk∥2−αR1k∥−(αR2+αR32)k∥3.\epsilon'_2 = \gamma k_{\parallel}^2 - \alpha_{R_1} k_{\parallel} - \left(\frac{\alpha_{R_2} + \alpha_{R_3}}{2}\right) k_{\parallel}^3.

The amount of momentum dependent spin-splitting ΔRM\Delta_{RM} is given by,

ΔRM=ϵ1′−ϵ2′=2αR1k∥+(αR2+αR3)k∥3=2αR1k∥+α~R3k∥3(3)\begin{aligned} \Delta_{RM} &= \epsilon'_1 - \epsilon'_2 \\ &= 2\alpha_{R_1} k_{\parallel} + (\alpha_{R_2} + \alpha_{R_3}) k_{\parallel}^3 \\ &= 2\alpha_{R_1} k_{\parallel} + \tilde{\alpha}_{R_3} k_{\parallel}^3 \tag*{(3)} \end{aligned}

where, α~R3=(αR2+αR3)\tilde{\alpha}_{R_3} = (\alpha_{R_2} + \alpha_{R_3}) is the modified cubic RSO interaction strength. The linear and cubic coupling strengths are obtained by fitting Eq. (2) and Eq. (3) to the DFT data. All the RSO interaction strengths are presented in Table II. The eqns. B1 and B2 can be written in the form of polynomial ak+bk3ak + bk^3.

References

  1. [1]S. Wolf, D. Awschalom, R. Buhrman, J. Daughton, v. S. von Molnár, M. Roukes, A. Y. Chtchelkanova, and D. Treger, Spintronics: a spin-based electronics vision for the future, science 294, 1488 (2001).pubmed.ncbi.nlm.nih.gov/11711666
  2. [2]E. Rashba and V. Sheka, Combinational resonance of zonal electrons in crystals having a zinc blende lattice, Soviet Physics-Solid State 3, 1257 (1961).
  3. [3]E. Rashba and V. Sheka, Landau level spectroscopy (1991).
  4. [4]E. Rashba, Properties of semiconductors with an extremum loop. I. Cyclotron and combinational resonance in a magnetic field perpendicular to the plane of the loop, Sov. Phys.-Solid State 2, 1109 (1960).
  5. [5]F. Vas’ ko, Spin splitting in the spectrum of two-dimensional electrons due to the surface potential, Soviet Journal of Experimental and Theoretical Physics Letters 30, 541 (1979).
  6. [6]E. Rashba and V. Sheka, Symmetry of energy bands in crystals of wurtzite type II. Symmetry of bands with spin-orbit interaction included, (2015).
  7. [7]W. Knap, C. Skierbiszewski, A. Zduniak, E. Litwin-Staszewska, D. Bertho, F. Kobbi, J. L. Robert, G. E. Pikus, F. G. Pikus, S. V. Iordanskii, V. Mosser, K. Zekentes, and Y. B. Lyanda-Geller, Weak antilocalization and spin precession in quantum wells, Phys. Rev. B 53, 3912 (1996).DOI
  8. [8]R. Winkler, S. Papadakis, E. De Poortere, and M. Shayegan, Spin-orbit coupling in two-dimensional electron and hole systems, Vol. 41 (Springer, 2003).
  9. [9]R. Winkler, Rashba spin splitting in two-dimensional electron and hole systems, Phys. Rev. B 62, 4245 (2000).DOI
  10. [10]S. Faniel, T. Matsuura, S. Mineshige, Y. Sekine, and T. Koga, Determination of spin-orbit coefficients in semiconductor quantum wells, Phys. Rev. B 83, 115309 (2011).DOI
  11. [11]S. Varotto, A. Johansson, B. Göbel, L. M. Vicente-Arche, S. Mallik, J. Bréhin, R. Salazar, F. Bertran, P. L. Févre, N. Bergeal, et al., Direct visualization of rashba-split bands and spin/orbital-charge interconversion at KTaO3 interfaces, Nature Communications 13, 6165 (2022).
  12. [12]G. Bihlmayer, S. Blügel, and E. V. Chulkov, Enhanced Rashba spin-orbit splitting in Bi/ Ag (111) and Pb/ Ag (111) surface alloys from first principles, Phys. Rev. B 75, 195414 (2007).DOI
  13. [13]P. Kim, K. T. Kang, G. Go, and J. H. Han, Nature of orbital and spin rashba coupling in the surface bands of SrTiO3 and KTaO3, Phys. Rev. B 90, 205423 (2014).DOI
  14. [14]E. I. Rashba and A. L. Efros, Orbital mechanisms of electron-spin manipulation by an electric field, Phys. Rev. Lett. 91, 126405 (2003).pubmed.ncbi.nlm.nih.gov/14525382
  15. [15]D. C. Vaz, P. Noël, A. Johansson, B. Göbel, F. Y. Bruno, G. Singh, S. Mckeown-Walker, F. Trier, L. M. Vicente-Arche, A. Sander, et al., Mapping spin-charge conversion to the band structure in a topological oxide two-dimensional electron gas, Nature materials 18, 1187 (2019).
  16. [16]Y. Kato, R. Myers, A. Gossard, and D. Awschalom, Coherent spin manipulation without magnetic fields in strained semiconductors, Nature 427, 50 (2004).
  17. [17]J. Sinova, S. O. Valenzuela, J. Wunderlich, C. H. Back, and T. Jungwirth, Spin hall effects, Rev. Mod. Phys. 87, 1213 (2015).DOI
  18. [18]V. M. Edelstein, Spin polarization of conduction electrons induced by electric current in two-dimensional asymmetric electron systems, Solid State Communications 73, 233 (1990).DOI
  19. [19]L. M. Vandersypen and M. A. Eriksson, Quantum computing with semiconductor spins, Physics Today 72, 38 (2019).DOI
  20. [20]A. Manchon, H. C. Koo, J. Nitta, S. Frolov, and R. Duine, New perspectives for rashba spin–orbit coupling, Nature materials 14, 871 (2015).DOI
  21. [21]M. I. Dyakonov and V. Perel, Current-induced spin orientation of electrons in semiconductors, Physics Letters A 35, 459 (1971).DOI
  22. [22]J. R. Sánchez, L. Vila, G. Desfonds, S. Gambarelli, J. Atané, J. De Teresa, C. M. Magén, and A. Fert, Spin-to-charge conversion using rashba coupling at the interface between non-magnetic materials, Nature communications 4, 2944 (2013).
  23. [23]L. G. D. da Silveira, P. Barone, and S. Picozzi, Rashba-dresselhaus spin-splitting in the bulk ferroelectric oxide BiAlO₃, Phys. Rev. B 93, 254159 (2016).
  24. [24]G. Venditti, M. E. Temperini, P. Barone, J. Lorenzana, and M. N. Gastiasoro, Anisotropic rashba coupling to polar modes in KTaO₃, Journal of Physics: Materials 6, 014007 (2023).DOI
  25. [25]S. Sheoran, M. Kumar, P. Bhumla, and S. Bhattacharya, Rashba spin splitting and anomalous spin textures in the bulk ferroelectric oxide perovskite KIO₃, Materials Advances 3, 4170 (2022).DOI
  26. [26]J. Gosteau, R. Arras, P. Chen, H. J. Zhao, C. Paillard, and L. Bellaiche, Spin-orbit effects in ferroelectric PbTiO₃ under tensile strain, Phys. Rev. B 103, 024416 (2021).DOI
  27. [27]A. Soumyanarayanan, N. Reyren, A. Fert, and C. Panagopoulos, Emergent phenomena induced by spin-orbit coupling at surfaces and interfaces, Nature 539, 509 (2016).DOI
  28. [28]M. Bibes, J. E. Villegas, and A. Barthélémy, Ultrathin oxide films and interfaces for electronics and spintronics, Advances in Physics 60, 5 (2011).DOI
  29. [29]S. Chen, C. Ren, and S. Liang, Spintronics phenomena of two-dimensional electron gas at oxide interfaces, Advanced Devices & Instrumentation 4, 0024 (2023).DOI
  30. [30]J. Varignon, L. Vila, A. Barthélémy, and M. Bibes, A new spin for oxide interfaces, Nature Physics 14, 322 (2018).DOI
  31. [31]M. Vivek, M. O. Goerbig, and M. Gabay, Topological states at the (001) surface of SrTiO₃, Phys. Rev. B 95, 165117 (2017).DOI
  32. [32]A. Ohtomo and H. Hwang, A high-mobility electron gas at the LaAlO₃/SrTiO₃ heterointerface, Nature 427, 423 (2004).pubmed.ncbi.nlm.nih.gov/14749825
  33. [33]N. Sivadas, H. Dixit, V. R. Cooper, and D. Xiao, Thickness-dependent carrier density at the surface of SrTiO₃ (111) slabs, Phys. Rev. B 89, 075303 (2014).
  34. [34]E. B. Guedes, S. Muff, M. Fanciulli, A. P. Weber, M. Caputo, Z. Wang, N. C. Plumb, M. Radović, and J. H. Dil, Single spin-polarized fermi surface in SrTiO₃ thin films, Phys. Rev. Res. 2, 033173 (2020).
  35. [35]S. N. Rebec, T. Jia, H. M. Sohail, M. Hashimoto, D. Lu, Z.-X. Shen, and R. G. Moore, Dichotomy of the photo-induced 2-dimensional electron gas on SrTiO₃ surface terminations, Proceedings of the National Academy of Sciences 116, 16687 (2019).DOI
  36. [36]P. Delugas, A. Filippetti, V. Fiorentini, D. I. Bilc, D. Fontaine, and P. Ghosez, Spontaneous 2-dimensional carrier confinement at the n-type SrTiO₃/LaAlO₃ interface, Phys. Rev. Lett. 106, 166807 (2011).pubmed.ncbi.nlm.nih.gov/21599400
  37. [37]R. Pentcheva and W. E. Pickett, Electronic phenomena at complex oxide interfaces: insights from first principles, Journal of Physics: Condensed Matter 22, 043001 (2010).DOI
  38. [38]S. Piskunov, E. Heifets, R. Eglitis, and G. Borstel, Bulk properties and electronic structure of SrTiO₃, BaTiO₃, PbTiO₃ perovskites: an ab initio HF/DFT study, Computational Materials Science 29, 165 (2004).DOI
  39. [39]K. Van Benthem, C. Elsässer, and R. French, Bulk electronic structure of SrTiO₃: Experiment and theory, Journal of applied physics 90, 6156 (2001).DOI
  40. [40]A. Santander-Syro, O. Copie, T. Kondo, F. Fortuna, S. Pailhès, R. Weht, X. Qiu, F. Bertran, A. Nicolaou, A. Taleb-Ibrahimi, et al., Two-dimensional electron gas with universal subbands at the surface of SrTiO₃, Nature 469, 189 (2011).
  41. [41]J. Mannhart and D. G. Schlom, Oxide interfaces–an opportunity for electronics, Science 327, 1607 (2010).DOI
  42. [42]Z. Chen, Z. Chen, Z. Q. Liu, M. E. Holtz, C. J. Li, X. R. Wang, W. M. Lü, M. Motapothula, L. S. Fan, J. A. Turcaud, L. R. Dedon, C. Frederick, L. R. Xu, R. Gao, A. T. N’Diaye, E. Arenholz, J. A. Mundy, T. Venkatesan, D. A. Muller, L.-W. Wang, J. Liu, and L. W. Martin, Electron accumulation and emergent magnetism in LaMnO₃/SrTiO₃ heterostructures, Phys. Rev. Lett. 119, 156801 (2017).DOI
  43. [43]J. Li, D. Yin, Q. Li, R. Sun, S. Huang, and F. Meng, Interfacial defects induced electronic property transformation at perovskite SrVO₃/SrTiO₃ and LaCrO₃/SrTiO₃ heterointerfaces, Physical Chemistry Chemical Physics 19, 6945 (2017).DOI
  44. [44]Z.-C. Wang, Y. Zhang, Z.-N. Li, S.-S. Li, J.-S. Ying, W. Yan, L. Chen, S.-J. Zhang, F.-S. Luo, W. Zhao, et al., High-mobility spin-polarized two-dimensional electron gas at the interface of EuTiO₃/SrTiO₃ heterostructures, Journal of Physics: Condensed Matter 36, 015001 (2023).DOI
  45. [45]W. Lin, L. Li, F. Doğan, C. Li, H. Rotella, X. Yu, B. Zhang, Y. Li, W. S. Lew, S. Wang, et al., Interface-based tuning of rashba spin-orbit interaction in asymmetric oxide heterostructures with 3 d electrons, Nature communications 10, 3052 (2019).DOI
  46. [46]P. Giannozzi, S. Baroni, N. Bonini, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, G. L. Chiarotti, M. Cococcioni, I. Dabo, et al., Quantum espresso: a modular and open-source software project for quantum simulations of materials, Journal of physics: Condensed matter 21, 395502 (2009).DOI
  47. [47]P. Giannozzi, O. Andreussi, T. Brumme, O. Bunau, M. B. Nardelli, M. Calandra, R. Car, C. Cavazzoni, D. Ceresoli, M. Cococcioni, et al., Advanced capabilities for materials modelling with quantum espresso, Journal of physics: Condensed matter 29, 465901 (2017).
  48. [48]J. P. Perdew, K. Burke, and M. Ernzerhof, Generalized gradient approximation made simple, Phys. Rev. Lett. 77, 3865 (1996).pubmed.ncbi.nlm.nih.gov/10062328
  49. [49]P. E. Blöchl, Projector augmented-wave method, Phys. Rev. B 50, 17953 (1994).pubmed.ncbi.nlm.nih.gov/9976227
  50. [50]B. G. Pfrommer, M. Côté, S. G. Louie, and M. L. Cohen, Relaxation of crystals with the quasi-newton method, Journal of Computational Physics 131, 233 (1997).
  51. [51]H. J. Monkhorst and J. D. Pack, Special points for Brillouin-zone integrations, Physical review B 13, 5188 (1976).DOI
  52. [52]T. Mizokawa and A. Fujimori, Unrestricted hartree-fock study of transition-metal oxides: Spin and orbital ordering in perovskite-type lattice, Phys. Rev. B 51, 12880 (1995).pubmed.ncbi.nlm.nih.gov/9978079
  53. [53]S. Okamoto, A. J. Millis, and N. A. Spaldin, Lattice relaxation in oxide heterostructures: LaAlO₃/SrTiO₃ superlattices, Phys. Rev. Lett. 97, 056802 (2006).pubmed.ncbi.nlm.nih.gov/17026128
  54. [54]M. Altmeyer, H. O. Jeschke, O. Hijano Cubelos, C. Martins, F. Lechermann, K. Koepernik, A. F. Santander-Syro, M. J. Rozenberg, R. Valentí, and M. Gabay, Magnetism, spin texture, and in-gap states: Atomic specialization at the surface of oxygen-deficient SrTiO₃, Phys. Rev. Lett. 116, 157203 (2016).DOI
  55. [55]W. Kong, T. Yang, J. Zhou, Y. Z. Luo, T. Zhu, J. Chen, L. Shen, Y. Jiang, Y. P. Feng, and M. Yang, Tunable rashba spin-orbit coupling and its interplay with multi-orbital effect and magnetic ordering at oxide interfaces, Phys. Rev. B 104, 155152 (2021).
  56. [56]H. Liang, L. Cheng, L. Wei, Z. Luo, G. Yu, C. Zeng, and Z. Zhang, Nonmonotonically tunable rashba spin-orbit coupling by multiple-band filling control in SrTiO₃-based interfacial d-electron gases, Phys. Rev. B 92, 075309 (2015).
  57. [57]K. V. Shanavas, Theoretical study of the cubic rashba effect at the SrTiO₃(001) surfaces, Phys. Rev. B 93, 045108 (2016).
  58. [58]R. Bistritzer, G. Khalsa, and A. H. MacDonald, Electronic structure of doped d⁰ perovskite semiconductors, Phys. Rev. B 83, 115114 (2011).DOI
  59. [59]L. W. van Heeringen, A. McCollam, G. A. de Wijs, and A. Fasolino, Theoretical models of rashba spin splitting in asymmetric SrTiO₃-based heterostructures, Phys. Rev. B 95, 155134 (2017).
  60. [60]M. Kim, J. H. Im, and S. B. Chung, Strongly enhanced rashba splittings in an oxide heterostructure: A tantalate monolayer on BaHfO₃, Phys. Rev. B 94, 115431 (2016).DOI
  61. [61]S. Vajna, E. Simon, A. Szilva, K. Palotas, B. Ujfalussy, and L. Szunyogh, Higher-order contributions to the Rashba-Bychkov effect with application to the Bi/Ag(111) surface alloy, Phys. Rev. B 85, 075404 (2012).
  62. [62]R. Arras, J. Gosteau, H. J. Zhao, C. Paillard, Y. Yang, and L. Bellaiche, Rashba-like spin-orbit and strain effects in tetragonal PbTiO₃, Phys. Rev. B 100, 174415 (2019).
  63. [63]S. Bhattacharya and S. Datta, Evidence of linear and cubic Rashba effect in non-magnetic heterostructure, Journal of Physics: Condensed Matter 35, 205501 (2023).DOI

Paper details

vlnc:W_x6b6jnmq
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents