Revealing New Leakage Paths in GAAFETs using Predictive First-Principles Simulations

Abstract

Gate-All-Around Field-Effect Transistors (GAAFETs) are emerging as successors to Fin Field-Effect Transistors (FinFETs) and are central to continued scaling at advanced semiconductor nodes. Despite their improved electrostatic control, recent measurements show that the drain current in GAAFETs deviates strongly from thermionic behavior at negative gate voltages, increasing as the gate voltage becomes more negative. This leakage has been attributed to direct gate-drain tunneling through the high-$k$ dielectric, enhanced by drain-induced dielectric barrier lowering (DIDBL). Here, we use predictive, charge-self-consistent Non-Equilibrium Green Function (NEGF) simulations based on the Contact Block Reduction (CBR) method to spatially resolve this leakage pathway in GAAFETs. Unlike approaches based primarily on phenomenological leakage models, our first-principles, multi-terminal quantum-transport framework treats the source, drain, and gate as current-carrying quantum-mechanical contacts and resolves the local current density throughout the device. The spatially resolved current density shows that gate-injected carriers tunnel through the high-$k$ dielectric preferentially near the drain side of the gate, where the dielectric barrier is most strongly lowered by the drain bias, and are collected by the drain. This provides a direct real-space picture of the DIDBL-driven gate-drain leakage mechanism.