WorksPhysics and Astronomy
Current-induced deterministic switching of van der Waals ferromagnet at room temperature
Abstract
Recent discovery of emergent magnetism in van der Waals magnetic materials (vdWMM) has broadened the material space for developing spintronic devices for energy-efficient computation. While there has been appreciable progress in vdWMM discovery, with strong perpendicular magnetic anisotropy (PMA) and Curie temperatures exceeding room temperature, a solution for non-volatile, deterministic switching of vdWMMs at room temperature has been missing, limiting the prospects of their adoption into commercial spintronic devices. Here, we report the first demonstration of current-controlled non-volatile, deterministic magnetization switching in a vdW magnetic material at room temperature. We have achieved spin-orbit torque (SOT) switching of the PMA vdW magnet Fe3GaTe2 using a Pt spin-Hall layer up to 320 K, with a threshold switching current density as low as at room temperature. We have also quantitatively estimated the anti-damping-like SOT efficiency of our Fe3GaTe2/Pt bilayer system to be = 0.093, using second harmonic Hall voltage measurement technique. These results mark a crucial step in making vdW magnetic materials a viable choice for the development of scalable, future spintronic devices.
Introduction
Magnetic materials-based spintronic devices1–3 hold great promise as energy-efficient, non-volatile memories and building blocks of neuromorphic4 and probabilistic5,6 computing hardware. However, just a few optimal material systems, like CoFeB/MgO7–9 in particular, have been identified for scalable device applications over the last two decades. The discovery of emergent magnetism in van der Waals (vdW) magnetic materials10–13 has opened a new arena for material exploration for spintronic technologies. vdW magnetic materials provide scalable, perpendicular magnetic anisotropy (PMA) alternatives to CoFeB/MgO while also providing atomically smooth interfaces down to monolayer thicknesses to help maintain device performance. An important step in the development of spintronic devices with vdW materials is the deterministic switching of PMA magnetism using current or voltage drives. While there have been several reports of spin-orbit torque (SOT) induced control of magnetism in bilayer systems of vdW ferromagnets with heavy metals14–17, topological insulators18, and topological semimetals19–22 over the last few years, such control at room temperature has remained elusive.
Here, we report the non-volatile, deterministic magnetization switching at room temperature in a vdW ferromagnet employing FeGaTe/Pt bilayer devices. Bulk crystals of FeGaTe (FGaT) were grown using a self-flux method as reported earlier23, and its magnetic properties were studied in the bulk and in exfoliated sheets. Bilayer devices of multi-layer FGaT and 6 nm Pt were fabricated to demonstrate current-induced magnetization switching up to 320 K in the presence of a 100 Oe in-plane magnetic field with a threshold current density of 1.69 A/cm. Furthermore, second harmonic Hall voltage measurements were used to estimate the anti-damping-like field component which is responsible for the current-induced switching, and the SOT efficiency of the FGaT/Pt system.
Growth and characterization of FeGaTe
We first present our characterization of bulk crystalline vdW FGaT synthesized using a Te self-flux method (see Methods). The FGaT unit cell possesses a hexagonal symmetry with space group P6/mmc (no. 194), like that of isostructural FeGeTe, wherein two adjacent quintuple layer substructures with two inequivalent Fe crystallographic sites form a vdW gap between the tellurium layers (Figure 1A). Millimeter-sized hexagonal planar crystals were measured with powder X-ray diffraction and showcase prominent (00L) Bragg peaks which were analyzed using a Rietveld refinement (Figure 1B). To verify the composition of our crystals, we performed energy-dispersive X-ray spectroscopy elemental mapping on a diverse set of bulk crystals and exfoliated flakes which all exhibited the correct atomic ratio (Figure 1C). Hysteresis loops of the direct current magnetization on bulk FGaT were performed from 3 K to 400 K with a magnetic field applied out-of-plane (OOP) as shown in Figure 1D. Both the coercive field and the saturation magnetization gradually decrease with increasing temperature up to the transition temperature near 340 K. These are complemented by measurements of the Hall resistivity with magnetic field applied out-of-plane which showcase a prominent anomalous Hall effect (AHE) accompanying the magnetization up to the same temperature (Figure 1E) with a room temperature Hall angle of 2.6°. The temperature dependence of magnetization with an applied out-of-plane magnetic field of 1000 Oe under field-cooling shows a departure near the Curie temperature from the zero-field-cooling situation (Figure 1F). The heat capacity manifests a prominent peak (Figure 1G) and the temperature-dependent longitudinal resistivity displays a noticeable change of slope (Figure 1H) both near 340 K. Altogether, these complementary measurements consistently indicate room-temperature magnetic properties and the high ferromagnetic transition temperature of the bulk FGaT crystals.

Figure 1. (A) Crystal structure of FeGaTe under different orientations: general view, -plane and -plane. (B) Powder x-ray diffraction (PXRD) pattern of the bulk crystal with indexed prominent Bragg peaks. (C) Scanning electron microscope image of an exfoliated flake on SiO/Si substrate (i) with corresponding elemental Fe (ii), Ga (iii) and Te (iv) maps and energy-dispersive spectra (v). The scale bar on the SEM image is 20 μm. (D) Spontaneous magnetization dependence with magnetic field applied out-of-plane at different temperatures. (E) Hall resistivity of the bulk sample measured at different temperatures with the magnetic field applied in the out-of-plane direction. Inset shows the measured anomalous Hall angle as a function of temperature. (F) Spontaneous magnetization under field-cooled (FC) and zero-field-cooled (ZFC) conditions with respect to temperature with the magnetic field (1000 Oe) applied out-of-plane. (G) Specific heat capacity measurement with zero and a 10 kOe magnetic field applied in the out-of-plane direction. Inset shows a fine scan at zero field near the magnetic transition. (H) Temperature dependence of the bulk longitudinal resistivity. Inset shows the derivative of the longitudinal resistivity with respect to temperature. A broad peak indicates the presence of a magnetic transition.
Next, we have characterized magnetism in exfoliated nanosheets of FGaT using a combination of AHE and polar magneto-optic Kerr effect (MOKE) measurements. Figure 2A shows the temperature dependent AHE hysteresis loops for OOP field sweep for a 29 nm thick FGaT flake. As depicted in the schematic (Figure 2A inset), current is applied along the – axis and the transverse voltage () is measured along – axis for the AHE measurements. The hysteresis loops exhibit a rectangular nature right up to room temperature, indicating a strong perpendicular magnetic anisotropy (PMA). Figure 2B shows the temperature dependence of remanent anomalous Hall resistance. The plotted here is the difference in transverse resistance of the device, recorded while warming up the device at zero magnetic field, after cooling under +3 T and -3 T out-of-plane field, respectively. The sharp drop in above 320 K (inset of Figure 2B) marks the transition from ferromagnetic to paramagnetic state with a Curie temperature K, slightly lower than the bulk value. The material exhibits an anisotropy field of kOe at 300 K as evident from the plot for near in-plane magnetic field sweeps () in Figure 2C. The inset provides vs with finer field steps indicating a room temperature coercivity of 220 Oe for the OOP field sweep. To further confirm room temperature magnetic properties of the FGaT flake, we report polar MOKE measurement of a 48 nm-thick flake in Figure 2D. The hysteresis plot shows a near 100% remanence confirming the presence of strong PMA, with a coercivity of 235 Oe. The two-step nature of the hysteresis loop can be attributed to presence of multi-domains in the relatively thick flake, as previously reported24. We would like to note that the MOKE plots correspond to a flake exposed directly to air for more than a week, indicative of the lack of material degradation (barring surface oxidation which cannot directly be concluded from MOKE) and hence, the air-stability of FGaT nanosheets.

Figure 2. (A) Temperature dependent hysteresis plots of the device for out-of-plane field sweeps (). Data offset along -axis for clarity. Inset: Schematic of measurement geometry. (B) Variation of anomalous Hall resistance of a (29 nm) FGaT flake against temperature, indicating a Curie temperature of K. Insets – optical image of the FGaT Hall bar on bottom left, zoomed in view of close to the Curie temperature. (C) Comparison of room temperature Hall resistance curves of the FGaT device for field swept OOP () and in-plane (), with anisotropy field of kOe denoted by vertical dashed lines. Inset: Low field zoom-in of the plot. (D) Room temperature polar MOKE curve of a (48 nm) FGaT flake, indicative of strong PMA and a coercivity Oe. (E) Optical image of the FGaT (57.9 nm)/Pt (6 nm) device (left) and its AFM micrograph (right). (F) Current sequence applied to the FGaT/Pt device for magnetization switching experiments, with the inset clarifying the nature of write pulses () and read pulses (). (G) Cyclic magnetization switching curves observed for the FGaT/Pt devices over four consecutives current pulsing loops, at 300 K under in-plane bias field of 100 Oe. Scale bars: 5 m.
Deterministic switching of vdW magnet
To demonstrate current induced switching of magnetism in FGaT, we have fabricated bilayer devices of exfoliated FGaT flakes (bottom) and 6 nm sputtered Pt (top). The bilayer stack is patterned into a Hall bar device and anomalous Hall resistance measurements are used to track the magnetization state of FGaT in the switching experiments. Figure 2E shows the optical image of one such device with 57.9 nm FGaT/ 6 nm Pt patterned into a 5 m wide Hall bar. The device is subjected to a current waveform as shown in Figure 2F, with 1 ms write pulses up to a maximum of 6 mA, at 1 s intervals, with being recorded after each current pulse. Figure 2G shows the Hall resistance across this device for the cyclic current sweeps. The measurements are performed at 300 K in the presence of a 100 Oe in-plane magnetic field applied parallel to current direction.
The device transitions from a low resistance to a high resistance state and vice versa, for a current pulse of mA, indicating switching of the OOP magnetization, with appreciable repeatability across four consecutive measurement cycles. This corresponds to a threshold switching current density of A/cm, which compares well with previous reports of ferromagnet/Pt systems (Supplementary Section S3). The small asymmetry in the positive and negative current switching cycles can be attributed to an OOP component of external magnetic field acting on the device due to slight misalignment between the field direction and the true sample plane. We have observed similar current-induced switching behavior in two more of our FGaT/Pt devices as documented in section S2 of Supplementary Information.
As depicted in Figure 3A, a charge current injected in the plane of the device results in the creation of a transverse spin-current in the Pt layer due to spin-Hall effect. The vertical component of the spin current results in an in-plane oriented spin accumulation at the FGaT/Pt interface. The in-plane damping-like torque applied by the moment of these spins drives the magnetization in-plane while the current pulse is active. The torque from the accompanying in-plane field, parallel to current direction, favors relaxation of the magnetization to one of the two OOP directions (opposite for positive and negative current) after the current pulse is removed25, resulting in deterministic, non-volatile switching of magnetization in the PMA material. Figure 3B shows the robust control of magnetization state in the device at room temperature using a train of random current pulses, 1 ms wide and mA in magnitude. We further studied the dependence of magnetization switching curves for varying in-plane field and increasing temperature. In Figure 3C, we observe that increasing the magnitude of applied in-plane magnetic field results in vertical shrinking of the hysteresis curves. Increasing in-plane magnetic field drives the steady state magnetization of FGaT further away from the -axis, decreasing the OOP magnetization component. Thus, we can expect a reduction in the anomalous Hall resistance of the device () on increasing the in-plane magnetic field, in agreement with our observations (Figure 3E). Similarly, we have found that the AHE splitting of the hysteresis curves decreases on increasing the temperature above 300 K (Figure 3D). We continue to see a sharp transition in resistance up to 320 K, but at 330 K, there is no clear change in across the current cycle. Once again, this agrees well with expectations that the magnitude of magnetization decreases with increasing temperature and goes to zero beyond the Curie temperature ( K). In fact, the trend of observed for these current pulse sweeps (Figure 3F) matches well with the vs observed in our FGaT-only devices (Figure 2A).
Figure 3. (A) Schematic diagram showing generation of spin-current in Pt layer in response to a planar current injection () due to spin-Hall effect. Vertical component of the in-plane polarized spin-current is incident at the Pt/FGaT interface, resulting in spin accumulation and spin-orbit torque on FGaT magnetization . (B) Deterministic, non-volatile switching of FGaT magnetization by a train of current pulses, 1 ms wide and mA in magnitude. Modulation of current-induced hysteresis curves upon variation of (C) externally applied in-plane magnetic field magnitude and (D) temperature of the system. Data offset along y-axis for clarity. (E, F) Variation of anomalous Hall resistance against in-plane field (E) and system temperature (F).
Estimation of Spin-Orbit Torque efficiency
To quantify the spin-orbit torque in our FGaT/Pt bilayer system, we have performed second harmonic Hall (SHH) voltage measurements26,27. As depicted in Figure 4A, the SHH voltage () is measured along the -axis for an ac current excitation along the -axis, in presence of an externally applied in-plane magnetic field, , of varying magnitude and azimuthal angle with respect to the -axis. For an arbitrary orientation of magnetization, with polar angle and azimuthal angle , the transverse resistance contains contributions from the anomalous Hall effect and the planar Hall effect,
Figure 4. (A) Schematic illustration of the second harmonic Hall (SHH) voltage measurement. External field is applied in the sample (-) plane at an angle from -axis. Current is applied along -axis and SHH voltage () is measured along the -axis. (B, C) measured for in-plane magnetic field rotation for (B) mA and (C) mA. Solid black lines are fits to equation (2). Data offset in y-axis for clarity. (D) Hollow squares represent the amplitude of components of in equation (2). Dotted lines are fits for the linear, thermal contribution to from ordinary Nernst effect and spin Seebeck effect. (E) Anti-damping-like field contribution to (solid squares) and their theoretical fit, with kOe. Inset: extracted for the two current level, and their fitting line, with near zero y-intercept. Measurements taken at 300 K.
where, and are the anomalous Hall resistance and planar Hall resistance of the device, respectively. Upon injecting an ac current into the device, the current induced field-like and anti-damping-like torques drive periodic oscillations of magnetization about its equilibrium position, resulting in harmonic oscillation of the transverse resistance . Thus, the recorded transverse voltage contains a component which can be detected through lock-in measurement. However, the SHH voltage also has contributions from thermal effects which appear in addition to the SOT-induced components and can lead to overestimation of SOT efficiency unless systematically eliminated from 28. Joule heating in the device, proportional to (and hence containing component), creates a vertical thermal gradient. The voltage measured along the –axis, is thus proportional to component of along –axis through the ordinary Nernst effect (ONE)29 and to through the anomalous Nernst effect (ANE) and spin-Seebeck effect (SSE). As a result, the SHH voltage takes the form29,
where, , and are the effective fields corresponding to current induced anti-damping-like (ADL) torque, field-like torque and Oersted field, respectively. kOe is the effective anisotropy field, is the SHH voltage per unit applied field and is the field-independent SSE and ANE contribution.
The SHH voltage corresponding to ac excitation of amplitude 1 mA and 1.5 mA is plotted in Figure 4B and Figure 4C, respectively. The solid black lines correspond to least squared error fit of the recorded voltage to Eq. (2). The components of the voltages, corresponding to combined ADL torque, ONE and SSE/ANE contribution, is plotted against in Figure 4D. As can be noted from Eq. (2), the ADL field dwindles upon increasing far over and we expect to see a linear scaling of at high fields due to the dominant ONE and SSE/ANE. Thus, a linear fit to at high fields (dotted lines in Figure 4D) can be used to eliminate thermal contributions from the SHH voltage. Figure 4E shows the SHH voltage corresponding solely to ADL torque, obtained upon subtraction of thermal contributions from in Figure 4D. We estimate an anti-damping-like field per unit current density of Oe/Am, where is the current density in Pt only, calculated based on the parallel resistor model (Supplementary Information, S1). The anti-damping-like spin torque efficiency can then be calculated as,
Using the bulk saturation magnetization of our FGaT crystals, emu/g (or A/m), we obtain for our FGaT/Pt device. This value is in good agreement with that expected for SOT system using Pt as the spin-Hall material with a metallic ferromagnet30. We provide a comparison of threshold switching current density, ADL torque per unit current and efficiency of various previous reports of deterministic magnetization switching in vdW magnetic materials in Supplementary Information, S3. Our FGaT/Pt device not only works beyond room temperature but also achieves switching at one of the lowest current densities (barring the CGT/Pt system16, where the FM is an insulator and hence not suitable for magnetic tunnel junctions) providing an energy efficient alternative for spintronic devices. During the preparation of this manuscript, we came across an archived report by Li et al. on a similar system of FGaT/Pt31. We would like to note that the threshold switching current density we report is almost an order of magnitude smaller than that of Li et al. Furthermore, while their reported is larger, it may be attributed to oversight of thermal contributions to the SHH voltage even upon operating at current densities larger than ours, as has been argued previously16 to result in overestimation of SOT effects15. By fully accounting for these contributions, we assume that our SOT efficiency value, although smaller, might be a more representative estimate for the FGaT/Pt bilayer system.
Conclusion
We have demonstrated current-induced, deterministic switching of out-of-plane magnetization in a /Pt bilayer, spin-orbit torque system up to 320 K, with a low switching current of A/cm at room temperature. Furthermore, using second harmonic Hall voltage measurements, we have quantitively estimated the effective anti-damping-like field and spin-torque efficiency of the FGaT/Pt system, and found it to be in good agreement with previous reports of Pt-based SOT systems. This work marks an important step in the adoption of vdW magnetic materials for building spintronic devices for non-volatile, energy-efficient memories and computing devices.
Methods
Bulk crystal growth and characterization
Single crystals of were synthesized through a Te self-flux method as described in Ref. 23. A mixture of Fe powder (Beantown Chemical, 99.9%), Ga ingot (Alfa Aesar, 99.99999%), and Te pieces (Sigma Aldrich, 99.999%) were weighed in a molar ratio of 1:2:2 in a nitrogen-filled glovebox (with and levels less than 0.1 ppm) and placed in an alumina Canfield crucible. The mixture-filled crucible was flame-sealed in an evacuated quartz tube with quartz wool and subsequently heated to 1000°C from room temperature within an hour. The mixture dwelled for 24 hours and subsequently cooled to 880°C within an hour followed by a slow cooling to 780°C at a rate of 1°C/h. Centrifugation was performed to remove the excess flux and afterwards the products were heat-treated to alleviate the concentration of tellurium defects. The resulting products contain a mixture of products with a silver luster among which contain crystals which are millimeter sized.
Powder X-ray diffraction (PXRD) data were measured on bulk samples using an X’Pert Pro diffractometer (PANalytical) in Bragg-Brentano geometry operating with a curved Ge(111) monochromator and Cu K radiation with a wavelength of 0.154 nm. Scanning electron microscope (SEM) images on both bulk crystals and exfoliated flakes on SiO substrates are measured using a Zeiss Merlin high-resolution SEM system with images acquired at an acceleration voltage of 20 kV, a current of 1000 pA, and a working distance of 8.5 mm. SEM-energy dispersive X-ray spectroscopy (EDS) elemental maps were taken using the EDAX APEX software.
Electrical transport measurements were performed on the bulk crystals in a Physical Property Measurement System (PPMS Dynacool, Quantum Design) in a five-probe geometry with contacts made of silver epoxy H20E and platinum wires. Direct current magnetization of the sample was acquired using the Vibrating Sample Magnetometer (VSM) option on crystals placed in the proper orientation using Kapton tape on a quartz holder (OOP) or using a Brass holder (IP). Temperature-dependent field-cooled and zero-field-cooled magnetization measurements were performed with an applied magnetic field of 1000 Oe. Specific heat capacity was measured on a 4.4 mg crystal using the Heat Capacity (HC) option with Apiezon H vacuum grease applied on the platform.
Device Fabrication
flakes were exfoliated on Si/SiO dies in a nitrogen-filled glovebox. For FGaT/Pt devices, the dies after exfoliation were sealed in the glovebox and only briefly exposed to air while transferring to the load-lock chamber of the sputterer. An RF plasma cleaning step was performed, under 4 mTorr Ar pressure and 40 W for 15 seconds, to remove any nascent oxide from FGaT surface. Subsequently, 6 nm Pt was deposited at a rate of 1.8 Å/s. 6-terminal Hall bars were patterned into the FGaT/Pt stack through e-beam lithography using negative resist ma-N2403, followed by Ar ion-milling (300 keV, 300s, inclination). Finally, lift-off through photolithography was used to pattern contact traces and pads for the etched Hall bars. Ti/Au (5 nm/70 nm) was e-beam deposited to create these metallic contacts. FGaT-only devices for magneto-transport characterization were fabricated using e-beam lithography with PMMA resist and Ti/Au (5 nm/35 nm) contacts, ensuring under a minute of direct exposure of FGaT flakes to ambient air. These devices were additionally encapsulated with thick hBN flakes in a glovebox using PDMS-based dry transfer.
Transport Measurements
All transport measurements were performed in a 9 T PPMS Dynacool system. Anomalous Hall effect measurements on the FGaT-only devices were performed using the Electrical Transport Option of the PPMS Dynacool, with a drive current of 100 A. Current-induced switching experiments were performed by interfacing a Keithley 6221 current source and 2182A nanovoltmeter with the PPMS Dynacool. Current input sequence consisted of a 1 ms write-pulse followed by 999 ms of read pulses ( A). For second harmonic Hall measurements, ac current was supplied by the Keithley 6221, at a frequency 1711.97 Hz. Two lock-in amplifiers, Stanford Research Systems (SRS) SR860 were used to simultaneously measure and transverse voltage components.
Current distribution across Pt and FGaT
Figure S1. Longitudinal resistivity of 6nm Pt on Si/SiO₂.
For a FGaT/Pt bilayer Hall bar, of width and length , equating voltage drop across both materials (parallel current channels), we have,
where, , , and are resistivities of Pt, FGaT and thicknesses of Pt and FGaT flake in the device, respectively.
Current induced switching - Additional FGaT/Pt devices
Figure S2. A. Current induced switching in device D1, for consecutive current pulsing loops. Inset: Optical image of the device; scale bar – 5 um. B. Deterministic, non-volatile switching in response to an arbitrary train of pulses. Measurements performed at 300 K under 2 kOe field parallel to current input.
Figure S3. Current induced switching curves for device D2, for in-plane field of 300 Oe. Chirality reverses upon changing field direction. Measurements at 300 K. Inset: Optical image of the device; scale bar – 5 um.
Current induced switching in vdW magnetic materials
| Ferro-magnet | Spin-Hall material | (K) | (Am) | (Oe/Am) | Reference | |
| FeGeTe | Pt | 180 | 0.14 | Alghamdi et al.1 | ||
| FeGeTe | Pt | 100 | 0.12 | Wang et al.2 | ||
| FeGeTe | (Bi,Sb)Te | 180 | Fujimura et al.3 | |||
| FeGeTe | WTe | 200 | Kao et al.4 | |||
| FeGeTe | WTe | 150 | 4.6 | Shin et al.5 | ||
| FeGeTe | WTe | 120 | Wang et al.6 | |||
| CrTe | ZrTe | 50 | Ou et al.7 | |||
| CrGeTe | Pt | * (insulator) | 0.25 | Gupta et al.8 | ||
| FeGaTe | Pt | 300 | 0.22† (likely overestimate) | Li et al.9‡ (unpublished) | ||
| FeGaTe | Pt | 300 | 0.093 | This work. |
Table 1.
References
H. Spin–Orbit Torque Switching in an All-Van der Waals Heterostructure. Adv. Mater. 34, 1–7 (2022).
S6. Wang, L., Xiong, J., Cheng, B., Dai, Y., Wang, F., Pan, C., Cao, T., Liu, X., Wang, P., Chen, M., Yan, S., Liu, Z., Xiao, J., Xu, X., Wang, Z., Shi, Y., Cheong, S. W., Zhang, H., Liang, S. J. & Miao, F. Cascadable in-memory computing based on symmetric writing and readout. Sci. Adv. 8, 1–11 (2022).
S7. Ou, Y., Yanez, W., Xiao, R., Stanley, M., Ghosh, S., Zheng, B., Jiang, W., Huang, Y. S., Pillsbury, T., Richardella, A., Liu, C., Low, T., Crespi, V. H., Mkhoyan, K. A. & Samarth, N. ZrTe2/CrTe2: an epitaxial van der Waals platform for spintronics. Nat. Commun. 13, 1–9 (2022).
S8. Gupta, V., Cham, T. M., Stiehl, G. M., Bose, A., Mittelstaedt, J. A., Kang, K., Jiang, S., Mak, K. F., Shan, J., Buhrman, R. A. & Ralph, D. C. Manipulation of the van der Waals Magnet Cr2Ge2Te6by Spin-Orbit Torques. Nano Lett. 20, 7482–7488 (2020).
S9. Li, W., Zhu, W., Zhang, G., Wu, H., Zhu, S., Li, R., Zhang, E., Zhang, X., Deng, Y., Zhang, J., Zhao, L., Chang, H. & Wang, K. Room-temperature van der Waals 2D ferromagnet switching by spin-orbit torques. arXiv 2304.10718, (2023).
References
- [1]Alghamdi, M., Lohmann, M., Li, J., Jothi, P. R., Shao, Q., Aldosary, M., Su, T., Fokwa, B. P. T. & Shi, J. Highly Efficient Spin-Orbit Torque and Switching of Layered Ferromagnet Fe3GeTe2. Nano Lett. 19, 4400–4405 (2019).DOI
- [2]Wang, X., Tang, J., Xia, X., He, C., Zhang, J., Liu, Y., Wan, C., Fang, C., Guo, C., Yang, W., Guang, Y., Zhang, X., Xu, H., Wei, J., Liao, M., Lu, X., Feng, J., Li, X., Peng, Y., Wei, H., Yang, R., Shi, D., Zhang, X., Han, Z., Zhang, Z., Zhang, G., Yu, G. & Han, X. Current-driven magnetization switching in a van der Waals ferromagnet Fe3GeTe2. Sci. Adv. 5, 2–8 (2019).
- [3]Fujimura, R., Yoshimi, R., Mogi, M., Tsukazaki, A., Kawamura, M., Takahashi, K. S., Kawasaki, M. & Tokura, Y. Current-induced magnetization switching at charge-transferred interface between topological insulator (Bi,Sb)2Te3 and van der Waals ferromagnet Fe3GeTe2. Appl. Phys. Lett. 119, 1–4 (2021).DOI
- [4]Kao, I. H., Muzzio, R., Zhang, H., Zhu, M., Gobbo, J., Yuan, S., Weber, D., Rao, R., Li, J., Edgar, J. H., Goldberger, J. E., Yan, J., Mandrus, D. G., Hwang, J., Cheng, R., Katoch, J. & Singh, S. Deterministic switching of a perpendicularly polarized magnet using unconventional spin–orbit torques in WTe2. Nat. Mater. 21, 1029–1034 (2022).DOI
- [5]Shin, I., Cho, W. J., An, E. S., Park, S., Jeong, H. W., Jang, S., Baek, W. J., Park, S. Y., Yang, D. H., Seo, J. H., Kim, G. Y., Ali, M. N., Choi, S. Y., Lee, H. W., Kim, J. S., Kim, S. D. & Lee, G. H. Spin–Orbit Torque Switching in an All–Van der Waals Heterostructure. Adv. Mater. 34, 1–7 (2022).
- [6]Wang, L., Xiong, J., Cheng, B., Dai, Y., Wang, F., Pan, C., Cao, T., Liu, X., Wang, P., Chen, M., Yan, S., Liu, Z., Xiao, J., Xu, X., Wang, Z., Shi, Y., Cheong, S. W., Zhang, H., Liang, S. J. & Miao, F. Cascadable in-memory computing based on symmetric writing and readout. Sci. Adv. 8, 1–11 (2022).DOI
- [7]Ou, Y., Yanez, W., Xiao, R., Stanley, M., Ghosh, S., Zheng, B., Jiang, W., Huang, Y. S., Pillsbury, T., Richardella, A., Liu, C., Low, T., Crespi, V. H., Mkhoyan, K. A. & Samarth, N. ZrTe2/CrTe2: an epitaxial van der Waals platform for spintronics. Nat. Commun. 13, 1–9 (2022).DOI
- [8]Gupta, V., Cham, T. M., Stiehl, G. M., Bose, A., Mittelstaedt, J. A., Kang, K., Jiang, S., Mak, K. F., Shan, J., Buhrman, R. A. & Ralph, D. C. Magnetomanipulation of the van der Waals Magnet Cr2Ge2Te6by Spin-Orbit Torques. Nano Lett. 20, 7482–7488 (2020).DOI
- [9]Li, W., Zhu, W., Zhang, G., Wu, H., Zhu, S., Li, R., Zhang, E., Zhang, X., Deng, Y., Zhang, J., Zhao, L., Chang, H. & Wang, K. Room-temperature van der Waals 2D ferromagnet switching by spin-orbit torques. arXiv 2304.10718, (2023).arxiv.org/abs/2304.10718



