Abstract: At the core of the theoretical framework of the ferroelectric field-effect transistor (FeFET) is the thermodynamic principle that one can determine the equilibrium behavior of ferroelectric (FERRO) systems using the appropriate thermodynamic potential. In literature, it is often implicitly assumed, without formal justification, that the Gibbs free energy is the appropriate potential and that the impact of free charge accumulation can be neglected. In this Article, we first formally demonstrate that the Grand Potential is the appropriate thermodynamic potential to analyze the equilibrium behavior of perfectly coherent and uniform FERRO-systems. We demonstrate that the Grand Potential only reduces to the Gibbs free energy for perfectly non-conductive FERRO-systems. Consequently, the Grand Potential is always required for free charge-conducting FERRO-systems. We demonstrate that free charge accumulation at the FERRO interface increases the hysteretic device characteristics. Lastly, a theoretical best-case upper limit for the interface defect density D-FI is identified. The ferroelectric field-effect transistor, which has attracted much attention for application as both a highly energy-efficient logic device and a non-volatile memory device, has often been studied within the framework of equilibrium thermodynamics. Here, the authors theoretically demonstrate the importance of utilizing the correct thermodynamic potential and investigate the impact of free charge accumulation on the equilibrium performance of ferroelectric-based systems.

Abstract: The observation of a significant temperature-dependent variation in the ${I}$ – ${V}$ characteristics of tunneling devices is often interpreted as a signature of a trap-assisted-tunneling dominated current. In this letter, we use a ballistic 2D quantum-mechanical simulator, calibrated using the measured temperature-dependent ${I}$ – ${V}$ characteristics of Esaki diodes, to demonstrate that the temperature dependence of band-to-band tunneling (BTBT) current can vary significantly in both Esaki diodes and tunnel FETs. The variation of BTBT current with temperature is impacted by doping concentration, gate voltage, possible presence of a highly-doped pocket at the tunnel junction, and material.

Abstract: Discrepancies exist between the theoretically predicted and experimentally measured performance of band-to-band tunneling devices, such as Esaki diodes and tunnel field-effect transistors (TFETs). We resolve this discrepancy for highly-doped, direct-bandgap Esaki diodes by successfully calibrating a semi-classical model for high-doping-induced ballistic band-tails tunneling currents at multiple temperatures with two In0.53Ga0.47As Esaki diodes using their SIMS doping profiles, C-V characteristics and their forward-bias current density in the negative differential resistance (NDR) regime. The current swing in the NDR regime is shown not to be linked to the band-tails Urbach energy. We further demonstrate theoretically that the calibrated band-tails contribution is also the dominant band-tails contribution to the subthreshold swing of the corresponding TFETs. Lastly, we verify that the presented procedure is applicable to all direct-bandgap semiconductors by successfully applying it to InAs Esaki diodes in literature.