Logo image
A New Methodology to Precisely Induce Wake-Up for Reliability Assessment of Ferroelectric Devices
Conference proceeding

A New Methodology to Precisely Induce Wake-Up for Reliability Assessment of Ferroelectric Devices

Tiang Teck Tan, Yu-Yun Wang, Joel Tan, Tian-Li Wu, Nagarajan Raghavan and Kin Leong Pey
IEEE International Reliability Physics Symposium proceedings, Vol.2023-, pp.1-7
03/2023

Abstract

Constant Voltage Stress Current measurement Defect Distribution Degradation Electrodes Fatigue Ferroelectric Ferroelectric devices Performance evaluation Ramp Voltage Stress Reliability Trap Assisted Tunneling Voltage Wake-up
Studies on ferroelectric (FE) device degradation are performed on "woken up" devices. The process of waking up a device is typically done by applying a logarithmically increasing number of pulsed, alternating bipolar switching voltage cycles. However, this method has low resolution in precisely achieving the wake-up state, resulting in ambiguity in the current stage of the life cycle of the device. Furthermore, ferroelectric device performance depends heavily on the spatio-temporal distribution of defects in the device stack, which are very different in the wake-up and fatigue phases of the life cycle. The standard bipolar pulsed stressing scheme as well as asymmetric device structure further complicate the analysis of the effects of voltage stressing on defect drift and subsequent device degradation. Here, we propose a new stressing methodology leveraging on an alternating stress-sense scheme using CVS/RVS and positive-up-negative-down (PUND) waveforms to better control the extent of wake-up in the device. Wake-up and the associated changes to the spatio-temporal mapping of the charged defect concentrations can be more confidently ascertained using the proposed methodology, thereby enabling better understanding of the reliability physics governing wake-up and fatigue for FE devices in the future for lifetime prediction from accelerated life tests.

Metrics

1 Record Views

Details

Logo image