Giant spin-orbit torque (SOT) from topological insulators (TIs) provides an energy efficient writing method for magnetic memory, which, however, is still premature for practical applications due to the challenge of the integration with magnetic tunnel junctions (MTJs). Here, we demonstrate a functional TI-MTJ device that could become the core element of the future energy-efficient spintronic devices, such as SOT-based magnetic random-access memory (SOT-MRAM). The state-of-the-art tunneling magnetoresistance (TMR) ratio of 102% and the ultralow switching current density of 1.2 × 105 A cm−2 have been simultaneously achieved in the TI-MTJ device at room temperature, laying down the foundation for TI-driven SOT-MRAM. The charge-spin conversion efficiency θSH in TIs is quantified by both the SOT-induced shift of the magnetic switching field (θSH = 1.59) and the SOT-induced ferromagnetic resonance (ST-FMR) (θSH = 1.02), which is one order of magnitude larger than that in conventional heavy metals. These results inspire a revolution of SOT-MRAM from classical to quantum materials, with great potential to further reduce the energy consumption.

The large spin Hall effect in topological insulators (TIs) is very attractive for ultralow-power spintronic devices. However, evaluation of the spin Hall angle and spin-orbit torque (SOT) of TIs is usually performed on high-quality single-crystalline TI thin films grown on dedicated III-V semiconductor substrates. Here, we report on room-temperature ultralow power SOT magnetization switching of a ferrimagnetic layer by non-epitaxial BiSb TI thin films deposited on Si/SiO 2 substrates. We show that non-epitaxial BiSb thin films outperform heavy metals and other epitaxial TI thin films in terms of the effective spin Hall angle and switching current density by one to nearly two orders of magnitude. The critical SOT switching current density in BiSb is as low as 7 × 10 4 A/cm 2 at room temperature. The robustness of BiSb against crystal defects demonstrate its potential applications to SOT-based spintronic devices. Charge-to-spin conversion utilizing the strong spin-orbit coupling (SOC) in non-magnetic materials has become a very attractive concept with possible applications to various spintronic devices, such as spin-orbit torque (SOT) magnetoresistive random access memories (MRAM) 1 , racetrack memories 2 , and spin torque nano-oscillators 3,4. These SOT-based spintronic devices are superior to their spin-transfer torque (STT)-based counterparts in terms of driving current, speed, and long-term durability 5,6. In SOT-based devices, a perpendicular pure spin current density J s is generated by an in-plane charge current density J e in the non-magnetic layer through the spin Hall effect (SHE) 7-9 , whose charge-to-spin conversion efficiency is characterized by the spin Hall angle θ SH = (2e/ℏ) J s /J e. Thus, finding spin Hall materials with large θ SH and high electrical conductivity is crucial for SOT applications, and there have been huge efforts so far to achieve that goal. In the well-studied heavy metals (HMs) such as Pt 10-13 , Ta 14 , and W 15-17 , θ SH is of the order of ~ 0.1, and the typical critical switching current density J c in bilayers of heavy metals/ferromagnet with perpendicular magnetic anisotropy is typically of the order of 10 7 A/cm 2 for continuous direct currents (DC) and 10 8 A/cm 2 for nano-second (ns) pulse currents. The large switching current density requires large driving transistors, whose size limits the bit density of SOT-MRAM. Meanwhile, large θ SH (> 1) have been observed in topological insulators (TIs) 18,19 thanks to their strong SOC and Dirac-point-driven singularity of the Berry phase on their topologically protected surface states 20. Thus, significant reduction of the driving current density from 10 7-10 8 A/cm 2 to 10 5-10 6 A/cm 2 can be expected for SOT-based devices 21-25 , particularly in SOT-MRAM whose their large writing current density is the major obstacle for reducing the writing power consumption and increasing the bit density. However, evaluation of θ SH and SOT switching by TIs is usually performed on single-crystalline TI thin films deposited on dedicated III-V...

We report on the growth and characterization of BiSb thin films deposited on sapphire substrates by sputtering deposition with Ar and Kr plasma. By optimizing the growth conditions, we are able to obtain quasi-single-crystal BiSb(001) thin films with equivalent twin crystals. The conductivity of BiSb at the studied thicknesses exceeds 10 5 Ω −1 m −1 , reaching 1.8 × 10 5 Ω −1 m −1 at 10 nm. From the temperature dependence of the electrical resistivity, we confirm the existence of metallic surface states. Our results demonstrate that it is possible to obtain sputtered BiSb thin films with quality approaching that of epitaxial BiSb grown by molecular beam epitaxy.

The MoS2 film for chip-size area was synthesized by two step processes consisting of MoS2-compound sputtering and post sulfurization. We intentionally revealed that the crystallinity of sulfurized MoS2 film depends on that of just-after-sputtered film. Therefore, a crystallinity improvement just-after sputtering is mandatory to achieve an excellent quality MoS2 film after sulfur-vapor annealing for thin film transistor, sensor and human interface device applications.

We report edge-induced ferromagnetism in a sputtered molybdenum disulfide (MoS2) film having a long whole-edge length, with the effects of crystallinity improvement including edge reconstruction by sulfur vapor and argon annealing. Strong edge-induced ferromagnetism was observed by annealing, and its saturation magnetization of 13–26 emu/cc was larger than that of a chemical-vapor deposition sample with edge-induced ferromagnetism, as reported previously. Whereas both the annealing steps improved the crystallinity of the sputtered MoS2 film, argon annealing significantly enhanced the ferromagnetism. We conclude that the difference of the ferromagnetism enhancement between the sulfur and argon annealing steps is attributed to the edge reconstruction shape, which depends on the sulfur chemical potential.