Research

We focus on developing smart pixel architectures that integrate light sensing with in-pixel computing, inspired by the neuromorphic computing paradigm. Our research emphasizes the use of emerging back-end-of-line (BEOL) materials—such as 2D MoS₂ for the pixel array and indium tin oxide (ITO) for peripheral circuitry—to enable efficient data processing and facilitate monolithic three-dimensional (3D) integration. In parallel, we focus on developing scalable and efficient neuromorphic devices at the technology level and on demonstrating their utilization at the circuit, (micro)architecture, and system levels to support efficient, flexible AI fine-tuning and inference on the edge.

We work on wafer-scale synthesis and integration of two-dimensional (2D) materials, aiming to unlock their potential for next-generation electronic and optoelectronic platforms. Our goal is to translate the unique physics of 2D materials into functional systems, with applications spanning neuromorphic computing, retinomorphic devices, and advanced circuits. This research involves scalable materials growth, heterogeneous integration, and device-to-circuit-level exploration, complemented by comprehensive materials, electrical, and optical characterization.

We focus on enhancing key performance parameters such as ON current, threshold voltage stability, and overall reliability in indium tin oxide (ITO)-based amorphous oxide semiconductor (AOS) transistors. Our goal is to enable their eventual integration into back-end-of-line (BEOL) applications, including DRAM. This research involves device-level design and fabrication, along with comprehensive electrical characterization, including C–V, I–V, and accelerated stress/recovery testing.

We are developing non-destructive capacitance–voltage measurement-based techniques to investigate reliability-limiting defects in two-terminal GaN devices. Our work also focuses on electrical characterization methods to determine the spatial and energy distribution of defects in GaN. This effort combines physics-based device modeling with electrical measurements on unconventional device geometries.