Tailoring Emergent Magnetic Moment in La$_{0.7}$Sr$_{0.3}$MnO$_3$-Bi$_2$Te$_3$ Heterostructures via Interfacial Reconstructions

Abstract

We report emergent magnetic behavior in heterostructures composed of (111)-oriented La$_{0.7}$Sr$_{0.3}$MnO$_3$ (LSMO) and (00$l$)-oriented Bi$_2$Te$_3$ (BT), controlled by interfacial reconstructions. When BT is deposited directly onto LSMO, an intermediate interfacial layer forms between the two materials. Polarized Neutron Reflectometry modeling reveals that this reconstructed region stabilizes a secondary magnetically ordered phase that is coupled to the underlying ferromagnetic LSMO layer. As a consequence, the heterostructures exhibit unconventional self-crossing magnetic hysteresis loops at room temperature, characterized by a reversal of the net magnetization at low applied fields. In contrast, the introduction of a tellurium seed layer results in a sharper LSMO-BT interface, while preserving the anomalous hysteresis behavior and enhancing the saturation magnetization. Element-specific X-ray absorption spectroscopy suggests that the emergent magnetic phase originates from the chemical reconstruction of manganese species. These results demonstrate that interface engineering in magnetic oxide-topological insulator heterostructures provides a pathway to control emergent magnetic coupling and emergent magnetic states in oxide-topological insulator heterostructures.

Figures (6)

• Figure 1: a) Schematic illustration of the LSMO-BT heterostructures discussed. One stack is grown with a direct contact between LSMO and BT layers (left), whereas the other has an additional tellurium seed layer deposited prior to the growth of the BT layer (right). b) RSM of (111)-LSMO around asymmetrical (123)$^+$ peak of STO. The RSM confirms the epitaxial relation between the film and the substrate, showing that LSMO thin film grows under tensile strain on STO. c) HRXRD of (111)-LSMO film deposited onto (111)-STO. The thickness is estimated using the InteractiveXRDFit package. d) HRXRD of (006)-Bi2Te3 with Te seed layer and estimated 20 nm thick layer.
• Figure 2: PNR data (reflectivity $R$ and calculated spin asymmetry $SA$) and $\rho_N$/$\rho_M$ profiles of LSMO-BT heterostructures grown using the direct growth (a) and the Te seed layer (b) approach. The data was collected at $T$ = 300 K and 110 K. The 300 K data is shifted along the $y$-axis for clarity. The shaded areas on the SLD profiles mark the 95% Bayesian confidence intervals. The dashed gray lines indicate the boundary between the pristine and the oxidized BT layers. The achieved figure of merit $\chi^2$ values are written above the plots. For the direct growth sample, the Refl1D fit of $\chi^2=1.93$ predicts formation of an intermediate interfacial phase with induced magnetic moment. For the seed layer sample, the fit of $\chi^2=1.56$ is achieved without the interfacial layer and with induced magnetic moment within the BT layer.
• Figure 3: X-ray absorption and dichroism spectra of the Mn L$_{3,2}$ edge. a) XAS and XMCD of the heterostructure with the direct growth. The dichroism intensity is expressed as a fraction of the XAS signal, normalized to the edge's maximum. b) Comparative XAS of the direct-growth heterostructure and a single LSMO thin film. The plot inset shows the Mn L$_{2}$ edge.
• Figure 4: a) X-ray absorption spectra of the Te M$_{5,4}$. b) The corresponding dichroism spectra (green line) and the dichroism spectra collected upon the magnetic field reversal (purple line).
• Figure 5: Zoomed-in hysteresis loops around $H$ = $\pm$100 Oe. a) LSMO-BT heterostructures with the direct growth (left) and seed layer growth (right). Scans collected at two temperatures are plotted to compare the magnetic response. b) Comparative heterostructure with the thinner LSMO layer. c) Single LSMO layer.