Stable Subloop Behavior in Ferroelectric Si-Doped HfO₂

Abstract

The recent demand for analogue devices for neuromorphic applications requires modulation of multiple nonvolatile states. Ferroelectricity with multiple polarization states enables neuromorphic applications with various architectures. However, deterministic control of ferroelectric polarization states with conventional ferroelectric materials has been met with accessibility issues. Here, we report unprecedented stable accessibility with robust stability of multiple polarization states in ferroelectric HfO 2 . Through the combination of conventional voltage measurements, hysteresis temperature dependence analysis, piezoelectric force microscopy, first-principles calculations, and Monte Carlo simulations, we suggest that the unprecedented stability of intermediate states in ferroelectric HfO 2 is due to the small critical volume size for nucleation and the large activation energy for ferroelectric dipole flipping. This work demonstrates the potential of ferroelectric HfO 2 for analogue device applications enabling neuromorphic computing.

Large Remnant Polarization in a Wake-Up Free Hf0.5Zr0.5O2 Ferroelectric Film through Bulk and Interface Engineering

“…This behavior was attributed to the effect of the ozone pulse on the suppression of the mechanisms that drive the wake-up effect [28]. That is, oxygen vacancies and the electric field-induced phase transition from the non-ferroelectric phases (m-and t-phase) to the ferroelectric o-phase [28][29][30][31][32][33][34][35][36].…”

Ferroelectric Properties and Polarization Fatigue of La:HfO2 Thin‐Film Capacitors

“…By fitting the log t 0 versus 1/ E plots with Merz's law (), [ 54 ] the activation field α for polarization switching can be extracted, which is 9.9 MV cm −1 for the fresh La:HfO 2 and increases to 11.9 MV cm −1 with pinned domains after the fatigue measurement. The activation field of La:HfO 2 is larger than that of Hf 1− x Zr x O 2 and Si:HfO 2 thin films deposited by the means of atomic layer deposition, [ 53,55 ] as listed in Table S1, Supporting Information.…”

Compact Modeling of Multidomain Ferroelectric FETs: Charge Trapping, Channel Percolation, and Nucleation-Growth Domain Dynamics

“…Because of the setup and devicerelated parasitic RC time constants (τRC) in the measurement (~ 10 -9 s -10 -6 s) [20], the extracted NLS parameters (τ0, Ea, etc.) can exhibit considerable discrepancies among literature [11,24,33], depending upon the exact experimental setups. Nevertheless, the possible imperfection in the NLS parameters in Table II is not expected to compromise their use in interpreting our eMSM (>10 -2 s, Fig.…”

“…14(a)). This would misleadingly equate the "incipient" critical nuclei (~10 0 nm 3 [33]) to fully reversed domains (>10 1 nm 3 [29,33,34]). Rather, we believe that even for the non-ratelimiting "fast growth" step in NLS [31], we need to model it as a finite-rate process instead of an instant one.…”