The Study of Hysteretic Characteristics of FeFETs for Neuromorphic Computing Using a Closed-Form Analytical Model

The emerging ferroelectric field effect transistor (FeFET) holds tremendous potential for neuromorphic computing beyond Moore’s law, offering advantages such as high-energy efficiency, easy scalability, high-density integration, and compatibility with mainstream complementary-metal-oxide-semiconductor (CMOS) technology. In this article, we propose an analytical current–voltage (
${I}-{V}$
) model, which involves solving coupled Poisson, Pah-Sah, and Landau–Khalatnikov equations for long-channel double-gate FeFETs with a metal–ferroelectric–metal–insulator–semiconductor (MFMIS) structure. We rigorously validate the accuracy of our developed models through technology computer aided design (TCAD) numerical simulations and experimental data. Subsequently, we derive analytical expressions for the two critical voltages that characterize the hysteretic behavior of FeFET-based switches. Furthermore, we explore the intricate influence of voltage bias and geometrical parameters on these critical voltages, utilizing a combination of analytical expressions and SPICE simulations. The findings from this study hold significant promise in optimizing the design of FeFET-based neurons.