Proximity-tuned Magnetic and Transport Anomalies in All-epitaxial Fe5-xGeTe2/WSe2 Van der Waals Heterostructures

TL;DR

This study demonstrates scalable all-epitaxial Fe5−xGeTe2/WSe2 van der Waals heterostructures with perpendicular magnetic anisotropy and room-temperature ferromagnetism. Structural probes (GID/STEM) confirm atomically sharp, registry-aligned interfaces and minimal intermixing, while magnetic and transport measurements reveal proximity-driven phenomena, including a thickness- and temperature-dependent sign reversal of exchange bias, a non-monotonic PMA with higher-order contributions, and a topological Hall effect linked to skyrmions. The results highlight strong interfacial SOC from WSe2 as a key driver of PMA and related anomalies, and they establish a scalable platform for exploring emergent 2D magnetic physics and next-generation spintronic concepts. Collectively, the work advances 2D magnetism research by providing a manufacturable route to integrated, proximity-engineered vdW devices capable of hosting chiral spin textures and robust spintronic functionality.

Abstract

Van der Waals (vdW) heterostructures combining two-dimensional (2D) ferromagnets and semiconducting transition-metal dichalcogenides (TMDCs) offer highly promising opportunities for tailoring 2D magnetism through interfacial proximity effects, enabling unique physical phenomena inaccessible in 3D systems and achieving functionalities beyond conventional spintronics. However, current fabrication of vdW heterostructures still relies heavily on the manual stacking of exfoliated 2D flakes, leading to critical challenges in scalability, interfacial quality, thickness control and device integration. This work reports on the realization of all-epitaxial, high-quality Fe5-xGeTe2(FGT)/WSe2 heterostructures exhibiting perpendicular magnetic anisotropy (PMA) and room-temperature ferromagnetism. The FGT/WSe2 system demonstrates temperature-driven magnetic transitions, higher-order PMA contributions and large anisotropic magnetoresistance, highlighting sublattice-specific contributions to magnetic and transport properties. Notably, the FGT/WSe2 heterostructures display unconventional physical phenomena, including thickness- and temperature-dependent sign reversal of exchange bias, a reversed thickness trend in the unconventional Hall effect, and a non-monotonic PMA-thickness dependence. These anomalies indicate pronounced interfacial contributions arising from proximity effects enhanced by epitaxial interface quality. Collectively, this study provides deep insights into the magnetic and transport properties of FGT/WSe2 vdW heterostructures, establishing a scalable platform for exploring emergent 2D physics and advancing next-generation 2D spintronic technologies.

Figures (7)

• Figure 1: a) In-plane reciprocal space maps: (left) bare WSe$_2$/Al$_2$O$_3$(0001) and (right) 17 nm FGT/WSe$_2$/ Al$_2$O$_3$(0001). Due to trigonal symmetry, the H and K axes form an angle of 60$^{\circ}$. b) Angular scans intersecting the WSe$_2$ and FGT(30.0) reflections in the space map shown in the right panel of a). c,d) Schematics of the structural configuration of the synthesized FGT/WSe$_2$ stack according to the epitaxial alignment determined by GID: c) side-view, d) top-view. For simplicity, d) shows only the last, topmost Se layer in WSe$_2$ and the first Te layer in FGT. The blue triangle indicates the coincidence-site lattice matching between WSe$_2$ and FGT.
• Figure 2: a) STEM cross-section image of a 12 nm thick FGT film grown on WSe$_2$/Al$_2$O$_3$(0001). b) STEM image (left) and the EELS compositional maps for W (orange), S (purple), Fe (blue), Ge (yellow) and Te (red) obtained from the same region.
• Figure 3: Magnetic characterization of a 17 nm FGT/WSe$_2$ vdW heterostructure using SQUID magnetometry. a) Normalized magnetization $M/M_\mathrm{S}$ as a function of out-of-plane (OP, black) and in-plane (IP, blue) magnetic fields at 20 K. Red curves are fits based on Equation (\ref{['eq:hys']}), comprising a square-shaped PMA component (pink) and a zero-coercivity contribution (green). Inset: normalized hysteresis loops at 300 K. b) Temperature-dependent magnetization $M$ during zero-field warming. The green curve shows the first derivative $dM/dT$ in the OP configuration. Inset: $M_\mathrm{R}/M_\mathrm{S}$ extracted from the hysteresis loops at each temperature.
• Figure 4: AHE and exchange bias effect in FGT/WSe$_2$ vdW heterostructures. a,b) Hall resistivity ($\rho_{xy}$) measured during upward and downward sweeps of the OP magnetic field at different temperatures, with sweep directions indicated by black arrows. The red lines in (b) are guides to the eye for the linear dependence in the saturation region. c,d) Temperature-dependent $\rho_{xy0}$/$\rho_\mathrm{A}$ and $\rho_\mathrm{A}$/$\rho_\mathrm{A}^{max}$ for samples with different FGT thicknesses, where $\rho_{xy0}$ is the zero-field $\rho_{xy}$ and $\rho_\mathrm{A}^{max}$ is the maximum $\rho_\mathrm{A}$. e) Top: Schematic of the sign reversal of the exchange bias controlled by temperature. Bottom: Temperature-dependent $H_{ex}$, with pink and green lines guiding the interfacial and bulk exchange coupling contributions, respectively.
• Figure 5: In-plane magnetotransport properties of a 6-nm-thick FGT/WSe$_2$ vdW heterostructure. a) Schematic of magnetization reorientation (pink arrows with spheres) in a PMA material under an applied in-plane magnetic field $H_{\parallel}$, determined by the minima of the free-energy profile (colored curves). Inset: measurement configuration. b) Phase diagram of possible magnetic states as functions of $H_\mathrm{K1}$ and $H_\mathrm{K2}$, symbols mark values extracted from the fits in c). c) $\rho_{xy}$/$\rho_\mathrm{A}$ versus $H_{\parallel}$, with $\rho_\mathrm{A}$ obtained from AHE measurements (Figure \ref{['fig:AHE']}). The red curves are fits based on Equation (\ref{['eq:PMAenergy']}). d) Temperature dependence of $H_\mathrm{K1}$ and $H_\mathrm{K2}$ extracted from fits in c). e) Temperature dependence of $H_\mathrm{S}$ obtained from AHE loops with OP (Figure \ref{['fig:AHE']}a,b) and IP (see estimation in c) field orientations. f) In-plane longitudinal magnetoresistance $MR_{\parallel}$ versus $H_{\parallel}$ at different temperatures; the red line indicates the method used to estimate the AMR value.