Metal–Insulator Transition Driven by Traps in 2D WSe₂ Field-Effect Transistor

Localized trap density (D_t) at the 2D channel–gate dielectric interface and its relative strength to carrier–carrier interactions depending on the thickness of the 2D channel can determine the nature of a metal–insulator transition (MIT) in 2D materials. Here, the MIT occurring in WSe₂ devices is systematically analyzed by varying the WSe₂ thickness from ≈20 nm to monolayer to explore the effects of D_t on MIT. The corresponding critical carrier density increases from ≈8.30 × 10¹¹ to 9.45 × 10¹² cm^–2 and D_t from ≈6.02 × 10¹¹ to 1.13 × 10¹³ cm^–2 eV^–1 as WSe₂ thickness decreases from ≈20 nm to monolayer. These large increments in D_t with decreasing thickness of WSe₂ induce a strong potential fluctuation in the band of WSe₂, causing charge density inhomogeneity in the system, which attributed to tuning the MIT. The critical percolation exponent is strongly dependent on WSe₂ thickness with an excellent agreement between the transport data and percolation theory achieved from thinner WSe₂ devices, while the transport data measured from multilayer WSe₂ devices does not obey the percolation theory. These results suggest that the nature of MIT strongly depends on the WSe₂ channel thickness and corresponding unscreened charge impurity and strength of D_t at the interface.