Switching Dynamics of Multidomain Ferroelectric Devices

Abstract

This work presents a comprehensive modeling framework for ferroelectric (FE) devices that can be integrated into hierarchical circuit simulators, enabling technology-device-circuit co-design. The framework employs an enhanced nucleation limited switching (NLS) model with updated Merz law and incubation time formulations to accurately capture the polarization dynamics under arbitrary voltage excitations and cumulative stimulation. The model is validated against measured data from fabricated HZO capacitors (FECaps), ferroelectric tunnel junctions (FTJs), and ferroelectric FETs (FeFETs). The switching time and the temperature sensitivity of FECap is explored. FTJ device readout is modeled through a self-consistent module that solves the Poisson equation, charge balance, tunneling transport, and drift-diffusion currents. FeFET device framework is modeled through enforcing electrostatic charge-continuity conditions at the gate stack interface with standard BSIM-CMG FinFET model. Utilizing its predictive capability, we explore the influence of technology parameters, including FE thickness, remanent polarization, effective mass, semiconductor doping, and program/erase voltages, on key FTJ and FeFET performance metrics such as on/off current, tunneling electroresistance (TER), and memory window behavior. The resulting Verilog-A–compatible framework provides a unified platform for analyzing and optimizing FE-based devices including FECAPs, FTJs, and FeFETs within emerging heterogeneous computing architectures.