Ion Depletion Defects in RRAMs
Characterization, Modeling and Test
Hanzhi Xun (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Moritz Fieback (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Sicong Yuan (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Hassen Aziza (Aix Marseille Université)
Mottaqiallah Taouil (TU Delft - Electrical Engineering, Mathematics and Computer Science)
Said Hamdioui (TU Delft - Electrical Engineering, Mathematics and Computer Science)
More Info
expand_more
Other than for strictly personal use, it is not permitted to download, forward or distribute the text or part of it, without the consent of the author(s) and/or copyright holder(s), unless the work is under an open content license such as Creative Commons.
Abstract
The commercialization of resistive random access memories (RRAMs) has been invested in by many companies as an alternative to conventional memories. This necessitates a thorough comprehension of manufacturing defects to create high-quality, efficient test and diagnosis solutions that will facilitate high-volume production. This work presents the identification and characterization of a novel defect, called ion depletion (ID), utilizing silicon measurements. In the selected defective device, after excluding the first 100 unstable cycles, 14% of the remaining 836 cycles exhibit ID-defective high resistance state (HRS) degradation, while the low resistance state (LRS) remains unaffected. Furthermore, this unique defect is non-linear and inside the device itself; hence, the conventional defect and fault modeling, which relies on linear resistors beside the device, cannot be applied to model the ID defect. To tackle this issue, the device-aware (DA) defect modeling approach is used. A calibrated RRAM model representing the defective device is developed using measurement data to precisely characterize how the defect influences the electrical properties of the memory hardware. Subsequently, fault analysis is conducted based on the DA defect model, leading to the introduction of suitable fault models. These models demonstrate that the ID defect can trigger undefined state faults. Finally, dedicated test and diagnosis solutions specifically designed for the ID defect are proposed.