Performance enhancement of NbOI₂-based field-effect transistor by piezotronic effect for object recognition

Two-dimensional field-effect transistors (2D FETs) have attracted significant interest in logic gate circuits, non-volatile memory, and sensors, owing to their superior gate control capability. However, high contact resistance (R_c) and limited effective carrier mobility (μ_eff) caused by the Schottky barrier impede the development of high-performance 2D FETs. Herein, the as-synthesized 2D NbOI₂ with a high piezoelectric coefficient was transferred onto flexible SiO₂/Si substrates to fabricate flexible NbOI₂ FET whose on/off ratio can reach 10⁴. Owing to piezotronic effect, the Schottky barriers at the FET metal-semiconductor interface are modulated by strain-induced piezoelectric polarization charges, which improve its electrical characteristics. Under a tensile strain of 3.32 ‰, the μ_eff and R_c of the NbOI₂ FET are increased by 56 % and decreased by 72.3 %, respectively, while the on-state current (I_on) has increased by 77.2 %. Notably, the I_on exhibits a good linear correlation with tensile strain, indicating its potential application in piezotronic sensors. Therefore, a strain sensing array consisting of four FETs responds to different objects by generating four-channel electrical output signals, achieving 99.9 % accuracy in object recognition with the aid of deep learning. This study proposes a novel strategy for designing high-performance 2D FET and highly integrated intelligent sensing systems.