Designing spin and orbital sources of Berry curvature at oxide interfaces

Pattern progress

The nine-unit-cell-thick LaAlO₃ crystalline layer is grown on the TiO-rich floor of a (111)-oriented SrTiO₃ substrate, from the ablation of a high-purity (>99.9%) LaAlO₃ sintered goal by pulsed laser deposition utilizing a KrF excimer laser (wavelength, 248 nm). We carry out the real-time monitoring of progress by following depth oscillations, in a layer-by-layer progress mode, of the primary diffraction spot utilizing reflection high-energy electron diffraction (Prolonged Knowledge Fig. 7a). This permits us to cease the expansion at exactly the essential thickness of 9 unit cells of LaAlO₃ (ref. ⁴⁶) vital for the (111)-oriented LaAlO₃/SrTiO₃ 2DES to type. The SrTiO₃(111) substrate was first heated to 700 °C in an oxygen partial strain of 6 × 10⁻⁵ mbar. The LaAlO₃ layer was grown in these situations at a laser fluence of 1.2 J cm⁻² and laser repetition charge of 1 Hz. Following the expansion of the LaAlO₃ layer, the temperature is ramped right down to 500 °C earlier than performing one-hour-long in situ annealing in a static background strain of 300 mbar of pure oxygen, to recuperate the oxygen stoichiometry of the lowered heterostructure. Lastly, the pattern is cooled down at –20 °C min⁻¹, and saved in the identical oxygen atmosphere at zero heating energy for at the least 45 min.

System fabrication

The (111)-oriented LaAlO₃/SrTiO₃ blanket movies had been lithographically patterned into two Corridor bars (width W = 40 μm; size L = 180 μm), oriented alongside the 2 orthogonal crystal-axis instructions of ([bar{1}10]) and ([bar{1}bar{1}2]). The Corridor bars are outlined by electron-beam lithography right into a poly(methyl methacrylate) resist, which is used as a tough masks for argon-ion milling (Prolonged Knowledge Fig. 7c). The dry-etching period is calibrated and timed to be exactly stopped when the LaAlO₃ layer is totally eliminated to keep away from the creation of an oxygen-deficient conducting SrTiO_3−δ floor. This leaves an insulating SrTiO₃ matrix surrounding the protected LaAlO₃/SrTiO₃ areas, which host a geometrically confined 2DES.

Electrical transport measurements

The Corridor bars are linked to a chip service by an ultrasonic wedge-bonding method wherein the aluminium wires type ohmic contacts with the 2DES by way of the LaAlO₃ overlayer. The pattern is anchored to the chip service by homogeneously coating the bottom of the SrTiO₃ substrate with silver paint. A d.c. voltage V_g is sourced between the silver back-electrode and the specified Corridor-bar gadget to allow electrostatic-field-effect gating of the 2DES, leveraging the massive dielectric permittivity of strontium titanate at low T (~2 × 10⁴ under 10 Ok)^47,48. Non-hysteretic dependence of σ_xx (σ_yy) on V_g is achieved following an preliminary gate-forming process⁴⁹.

Normal four-terminal electrical (magneto-)transport measurements had been carried out at 1.5 Ok in a liquid helium-4 circulation cryostat, outfitted with a superconducting magnet (most magnetic discipline, B = ±12 T). An a.c. excitation present I^ω ∝ |I^ω|sin(ωt), of frequency ω/(2π) = 17.77 Hz, is sourced alongside the specified crystallographic course. The sheet resistance, ({R}_{{{{rm{s}}}}}={sigma }_{{{{{xx}}}}}^{-1}), of a Corridor-bar gadget is expounded to the first-harmonic longitudinal voltage drop V_xx based on R_s = (V_xx/I_x)(W/L). When the a.c. present is sourced alongside (hat{{{{bf{x}}}}}parallel [bar{1}10]) ((hat{{{{bf{y}}}}}parallel [bar{1}bar{1}2])), we make use of a typical lock-in detection method to concomitantly measure the first-harmonic longitudinal response V_xx (V_yy), and both the in-phase first-harmonic ({V}_{{{{{xy}}}}}^{omega }) (({V}_{{{{{yx}}}}}^{omega })) or out-of-phase second-harmonic ({V}_{{{{{yxx}}}}}^{2omega }) (({V}_{{{{{xyy}}}}}^{2omega })) transverse voltages (Fig. 1a). We outline the first- and second-harmonic transverse resistances as ({R}_{{{{{xy}}}}}^{omega }={V}_{{{{{xy}}}}}^{omega }/| {I}_{{{{{x}}}}}^{omega }|) and ({R}_{{{{{y}}}}}^{2omega }={V}_{{{{{yxx}}}}}^{2omega }/| {I}_{{{{{x}}}}}^{omega } ^{2}), respectively. First- and second-harmonic measurements are carried out at 10 and 50 μA, respectively.

We systematically decompose each first- and second-harmonic magneto-responses into their field-symmetric ({R}_{{{{rm{sym}}}}}^{(2)omega }) and field-antisymmetric ({R}_{{{{rm{as}}}}}^{(2)omega }) contributions based on

Specifically, the first-harmonic transverse resistance is solely discipline antisymmetric, and therefore, we selected the simplified notation of ({R}_{{{{{xy}}}}}equiv {R}_{{{{{xy}}}},{{{rm{as}}}}}^{omega }).

Estimation of the Rashba spin–orbit power from magnetoconductance measurements within the weak antilocalization regime

In a 2DES, within the presence of a spin rest mechanism induced by an extra spin–orbit interplay, the conductance is topic to weak localization corrections at decrease temperatures. Prolonged Knowledge Fig. 4a exhibits the gate-modulated magnetoconductance curves of the 2DES, which exhibit a attribute low-field weak antilocalization behaviour. The magnetoconductance curves, normalized to the quantum of conductance G_Q = e²/(πħ), are fitted utilizing a Hikami–Larkin–Nagaoka mannequin that expresses the change in conductivity Δσ(B_⟂) = σ(B_⟂) – σ(0) of the 2DES below an exterior out-of-plane magnetic discipline B_⊥, within the diffusive regime (with negligible Zeeman splitting), as follows^50,51:

the place Ψ is the digamma perform; ħ = h/(2π) is the lowered Planck fixed; ({B}_{{{{rm{i}}}},{{{rm{so}}}}}=hslash /left(4eD{tau }_{{{{rm{i}}}},{{{rm{so}}}}}proper)) are the efficient fields associated to the inelastic and spin–orbit rest instances (τ_i and τ_so, respectively); and D = πħ²σ(0)/(e²m*) is the diffusion fixed. The final time period in equation (3), proportional to ({B}_{perp }^{2}), comprises A_Ok, the so-called Kohler coefficient, which accounts for orbital magnetoconductance.

Therefore, from the match to the weak antilocalization magnetoconductance curves, the efficient Rashba spin–orbit coupling α_R will be calculated as

based mostly on a D’yakonov–Perel’ spin rest mechanism⁵¹. A abstract of the dependence of the extracted parameters on the 2DES’ sheet conductance is plotted in Prolonged Knowledge Fig. 5b. The spin–orbit power Δ_so can then be estimated based on

the place, in two dimensions, the Fermi wavevector is given by ({ok}_{{{{rm{F}}}}}=sqrt{2uppi {n}_{2{{{rm{D}}}}}}), assuming a round Fermi floor. The sheet service density n_2D is experimentally obtained for every doping worth from the (strange) Corridor impact (Supplementary Be aware III), measured concomitantly with the magnetoconductance traces.

Spin-sourced and orbital-sourced BCD calculations

We first estimate the BCD as a consequence of spin sources in time-reversal symmetry situation as a perform of service density contemplating the low-energy Hamiltonian for a single Kramers’-related pair of bands (Supplementary Be aware I):

the place the momentum-dependent mass will be unfavourable near the Γ level (Supplementary Be aware I). Though this mannequin Hamiltonian is supplied with a finite BC, its dipole is pressured to fade by the three-fold rotation symmetry (Supplementary Be aware I). We seize the rotation symmetry breaking of the low-temperature construction on the main order by assuming inequivalent coefficients for the spin–orbit coupling phrases linear in momentum. In different phrases, we make the substitution α_R(σ_xok_y – σ_yok_x)→v_yok_yσ_x – v_xok_xσ_y. Because the dipole is a pseudo-vector, the residual mirror symmetry ({{{{mathcal{M}}}}}_{x}) forces it to be directed alongside the (hat{{{{bf{x}}}}}) course. Within the relaxation-time approximation, it’s given by

the place Ω_z is the BC of our two-band mannequin that we write in a dimensionless type by measuring energies in models of ({ok}_{{{{rm{F}}}}}^{2}/2m({ok}_{{{{rm{F}}}}})), lengths in models of 1/ok_F and densities in models of ({n}_{0}={ok}_{{{{rm{F}}}}}^{2}/2uppi). Right here ok_F is a reference Fermi wavevector. For simplicity, we have now thought-about a constructive momentum-independent efficient mass. For the BCD proven in Fig. 4a, the remaining parameters have been chosen as v_x = 0.4, v_y = (1.2, 1.4, 1.6) × v_x and λ = 0.1. Furthermore, we account for orbital degeneracy by tripling the dipole of a single Kramers’ pair. This offers an higher sure for the spin-sourced BCD.

We’ve got additionally evaluated the BCD as a consequence of orbital sources contemplating the low-energy Hamiltonian for spin–orbit-free t_2g electrons derived from symmetry rules (Supplementary Be aware I) and studying

the place we launched the Gell–Mann matrices as

and Λ₀ is the id matrix. Within the Hamiltonian above, Δ is the splitting between the a_1g singlet and ({e}_{mathrm{g}}^{{prime} }) doublet ensuing from the t_2g orbitals in a trigonal crystal discipline. Right here Δ_m is the extra splitting between the doublet brought on by rotational symmetry breaking. Lastly, α_OR and α_m are the strengths of the orbital Rashba coupling. Be aware that within the presence of three-fold rotation symmetry, α_m ≡ 0, wherein case the BC is pressured to fade. For simplicity, we have now evaluated the BC for the ({{{{mathcal{C}}}}}_{mathrm{s}}) level group-symmetric case assuming α_m ≡ α_OR. In our continuum SU(3) mannequin, the BC will be computed utilizing the strategy outlined elsewhere⁵². We’ve got subsequently computed the corresponding dipole measuring, as earlier than, energies in models of ({ok}_{{{{rm{F}}}}}^{2}/2m), lengths in models of 1/ok_F and densities in models of ({n}_{0}={ok}_{{{{rm{F}}}}}^{2}/2uppi). The dimensionless orbital Rashba coupling has been different between α_OR = 1 and α_OR = 2, whereas we have now fastened Δ = –0.1 and Δ_m = 0.005. The worth of the crystal discipline splitting Δ is per the amplitude decided by X-ray absorption spectroscopy⁵³ of the order 8 meV, and subsequently, it’s virtually one order of magnitude smaller than our power unit of ~40 meV for a reference ({ok}_{mathrm{F}}^{-1}) ≃ 0.5 nm and efficient mass m ≃ 3m_e (Supplementary Be aware III). The calculated dipole (Fig. 4a) has been lastly multiplied by two to account for spin degeneracy. As proven in Supplementary Be aware I, we comment that the mannequin Hamiltonian for the spin sources of BC (equation (6)) and the mannequin Hamiltonian for the orbital sources (equation (8)) derive from a single six-band mannequin the place orbital and spin levels of freedom are handled on an equal footing.

Estimation of BCD magnitude from nonlinear Corridor measurements

The nonlinear present density is mathematically given by ({j}_{alpha }^{2omega }={chi }_{alpha beta gamma },{E}_{beta },{E}_{gamma }), the place χ_αβγ is the nonlinear transverse-conductivity tensor. When an a.c. present density ({I}_{{{{{x}}}}}^{omega }/W={sigma }_{{{{{xx}}}}}{E}_{{{{{x}}}}}^{omega }) is sourced alongside (hat{{{{bf{x}}}}}), the second-harmonic transverse present density creating alongside (hat{{{{bf{y}}}}}) is expounded to the BCD D based on⁹

the place τ is the momentum rest time and e is the elementary cost. Because of the mirror symmetry ({{{{mathcal{M}}}}}_{{{{{x}}}}}equiv {{{{mathcal{M}}}}}_{[bar{1}10]}), the dipole is discovered to level alongside (hat{{{{bf{x}}}}}); within the quasi-d.c. restrict, that’s, (ωτ) ≪ 1, the BCD expression reduces to

which is the express expression for equation (1), when it comes to experimentally measurable portions solely, and the place

are the measured nonlinear transverse-conductivity tensor parts proven in Fig. 4c,e.