PhD Seminar Notice - Hyunjae Lee

Tuesday, July 19, 2022 12:00 pm - 12:00 pm EDT (GMT -04:00)

Candidate: Hyunjae Lee
Title: A Novel Computational Framework for Negative Capacitance Devices: From Ferroelectric-based Capacitors to Negative Capacitance Field-Effect Transistors
Date: July 19, 2022
Time: 12:00
Place: online
Supervisor(s):  Yoon, Youngki

Abstract:

Negative capacitance (NC) devices generally manifest hysteresis originating from the ferroelectric materials (FE) with nonlinear hysteretic responses of spontaneous polarization (P) to the applied electric field, while hysteresis should be suppressed for logic devices. The significant nonlinearity of P behavior in the FE entangles the manipulation and prediction of not only performance but also hysteresis in the NC-based devices when multiple materials are considered simultaneously within the device.

Here, hysteretic jump points (HJPs) are modelled by exploring the energy landscape based on the Gibbs free energy with the Landau approach in a metal-FE-metal-oxide-semiconductor capacitor. With the aid of HJPs, the hysteresis response to the key design parameters in the NC device can be described for various ferroelectric materials, which can contribute to the hysteresis engineering.

While the Landau approach can provide insights into the polarization behavior of FE with a so-called “S-curve” of polarization versus electric field across the FE () characteristics, the Miller model (MM) is more suitable for a gradual transition of polarization switching for the stand-alone FE in steady states, since it can precisely capture the overall hysteresis loop including the polarization switching transition region. Firstly, the static NC characteristics in a ferroelectric-dielectric

(FE-DE) capacitor are presented with significant internal voltage amplification using MM. Notably, the effect of different transitions of polarization switching that the S-curve from Landau approach cannot handle is explored.

After that, the FE-DE capacitor can be further extended to more advanced device such as a NC field-effect transistors (NCFETs), established on the understanding of NC characteristics with MM. Especially, NCFETs composed of FE exhibiting gradual switching transitions are demonstrated in steady-states using a novel computational framework. By iteratively solving three modules: (i) non-equilibrium Green’s function (NEGF) for carrier transport, (ii) Poisson’s equation for electrostatics, and (iii) the Miller model for the polarization behavior in FE, steep switching with hysteresis-free characteristics of NCFET can be achieved.

Besides, various materials and device parameters are engineered to understand their effects on the device performance, which enables performance optimization as required for different target applications.

Although the demonstrated NCFET model provides in-depth understanding of NC in the device, an upgraded model including a more realistic behavior of FE that the  could be varied by the maximum electric field applied due to a property called a partially polarized FE (PPFE) can present more precise prediction of the NCFET. For the PPFE, the MM-based FE minor loop model has been used to describe its switching behavior. By elucidating the difference between the NCFET considering PPFE and ideal behavior of FE, the overestimated subthreshold swing for a NCFET under the ideal FE behavior is observed as compared to the PPFE.

Furthermore, with the understanding of PPFE, a drain-induced barrier lowering (DIBL) in the PPFE-NCFET is investigated in comparison to the FPFE-NCFET.

This dissertation can provide not only a novel computational framework for NCFET simulation from FE-based capacitor to transistor but also irreplaceable physical insight into the NCFET behavior.