Hysteresis-Free High Mobility Graphene Encapsulated in Tungsten Disulfide

Technical Summary: Hysteresis-Free High Mobility Graphene Encapsulated in Tungsten Disulfide

Problem Statement
High carrier mobility is a fundamental requirement for advanced electronic and optoelectronic applications, including field-effect transistors, modulators, and sensors. While graphene possesses ultra-high room-temperature mobility, its performance is often limited by scattering from residual charge carriers and substrate-induced roughness. Currently, the state-of-the-art for high-quality graphene devices relies on encapsulation with hexagonal boron nitride (hBN). However, scaling hBN encapsulation presents significant challenges: precisely controlling the number of layers in large-area flakes remains elusive, and chemical vapor deposition (CVD) grown hBN often suffers from point defects, wrinkles, and grain boundaries that increase surface roughness and degrade mobility.

Transition metal dichalcogenides (TMDs), specifically tungsten disulfide (WS₂), offer a scalable alternative as they can be grown in large, multi-layer flakes via CVD. However, a critical barrier to their adoption as graphene encapsulants is severe hysteresis in electrical characteristics. When gating graphene through untreated WS₂, the charge neutrality point (CNP) exhibits large hysteretic shifts (Δn ∼ 2.6 × 10¹¹ cm⁻²). This hysteresis is attributed to charge trapping at sulfur vacancies within the WS₂ lattice and at the interface with the substrate, which hinders the reliable operation of graphene devices.

Methodology
The authors propose a chemical passivation strategy to eliminate hysteresis in WS₂-encapsulated graphene. The core methodology involves treating exfoliated WS₂ flakes with a super-acid, bis(trifluoromethane)sulfonimide (TFSI), dissolved in 1,2-dichloroethane (DCE).

1. Fabrication: WS₂ flakes (6–25 nm thick) and single-layer graphene were exfoliated onto SiO₂/Si substrates. The WS₂ flakes were immersed in the TFSI solution for approximately 30 minutes inside a nitrogen glovebox to remove contaminants and passivate vacancies, followed by nitrogen blow-drying.
2. Heterostructure Assembly: Using a dry-transfer stamping technique with a polycarbonate (PC) film on a polydimethylsiloxane (PDMS) support, the authors assembled a Tr-WS₂/Gr/Tr-WS₂ heterostructure (where "Tr" denotes TFSI-treated). Both the top and bottom encapsulating layers were treated.
3. Control and Characterization: For comparison, an untreated WS₂/Gr/WS₂ heterostructure was fabricated. Devices were patterned into Hall bar geometries. Electrical transport measurements were performed using a four-probe setup with a back-gate configuration. The study included systematic tests of hysteresis (Δn) as a function of sweep rate, sweep range, and device aging (up to 120 days).
4. Material Analysis: Atomic Force Microscopy (AFM) and Raman spectroscopy were employed to assess morphological changes and defect passivation efficacy post-treatment.

Key Results

• Hysteresis Suppression: The TFSI treatment drastically reduced hysteresis. In untreated devices, the charge neutrality point shifted by Δn ∼ 2.6 × 10¹¹ cm⁻². In contrast, the treated Tr-WS₂/Gr/Tr-WS₂ devices exhibited hysteresis below the measurement noise level, with Δn < 5 × 10¹⁰ cm⁻² (a reduction of at least a factor of five).
• Mobility Performance: The treated heterostructures demonstrated high room-temperature mobility. At a carrier density of n ∼ 1 × 10¹² cm⁻², the Drude mobility reached ∼6.2 × 10⁴ cm²V⁻¹s⁻¹.
• Stability and Robustness:
  • Sweep Rate/Range: The low hysteresis was maintained across varying sweep rates (0.005 to 0.1 Vs⁻¹) and voltage sweep ranges, confirming that charge trapping mechanisms were effectively suppressed.
  • Aging: The devices retained their low-hysteresis performance after 120 days of storage in ambient laboratory conditions, indicating the resilience of the chemical treatment.
  • Morphology: AFM analysis showed no significant change in surface roughness (standard deviation remained ~145–155 pm) after treatment.
  • Raman Spectroscopy: A slight blue-shift in the E¹₂g and A₁g modes was observed, consistent with the removal of substitutional oxygen atoms and the passivation of sulfur vacancies.
• High-Doping Behavior: Unlike hBN-encapsulated graphene, which typically shows a ~70% mobility drop at high carrier densities due to acoustic phonon coupling, the WS₂-encapsulated devices showed a more stable mobility profile, with only a ~30% drop at n ∼ 3.8 × 10¹² cm⁻².

Significance and Claims
The paper claims to report the first demonstration of hysteresis-free electrical behavior in high-quality graphene encapsulated by TMDs. By utilizing TFSI chemical treatment, the authors successfully passivate sulfur vacancies and remove surface contaminants in multilayer WS₂, transforming it from a hysteretic dielectric into a reliable encapsulant.

The authors position chemically treated WS₂ as a valid and scalable alternative to hBN for high-performance graphene devices. This is particularly significant for applications requiring large-area integration via CVD, where controlling hBN layer thickness is difficult. The results suggest that TMD/Gr/TMD heterostructures are promising candidates for integrated optoelectronic applications, such as photodetectors and modulators, where low hysteresis and high mobility at high carrier concentrations are critical. The study establishes that the chemical passivation of TMD defects is a viable pathway to overcoming the intrinsic limitations of TMD-based dielectrics without compromising the morphological integrity of the material.