Design space of Hafnia-FeCap-based 6T2C-nvSRAM
NvSRAM is promising for IoT applications where devices operate in low activity frequency and standby leakage current essentially dominates the energy consumption. Among various proposed nvSRAMs, the 6T2C structure uses FeCaps as the non-volatile elements, and becomes one of the most encouraging options in terms of density and cost, especially after the discovery of ferroelectricity in CMOS-compatible HfO2. The proof-of-concept of the 6T2C nvSRAM with HfO2 FeCaps has been experimentally demonstrated by M. Kobayashi et al. However, a deep analysis of the 6T2C nvSRAM based on Hafnia FeCaps including its working principle and performance optimization is still lacking, which is critical for the further development of such technology.
This work comprehensively evaluates the performance of Hafnia-FECap-based 6T2C nvSRAM through simulations. The interplay between properties of transistors and FECaps is presented, and optimization strategies are proposed.
Future of ferroelectric tunnel junction - nonvolatile or volatile?
We developed a computationally efficient and comprehensive model for metal-ferroelectric-insulator-metal (MFIM) Ferroelectric Tunnel Junctions (FTJs).
-
Computationally efficient: inhomogeneous E-field in multi-domain HZO film is parameterized into a single partial differential equation; ×10 faster than KMC model when FE domains > 103;
-
Comprehensive: image force (IF) and screening charge (SC) effects are involved in potential profile calculations, responsible for ON-state current density JON and tunneling electroresistance ratio (TER);
Based on the model, we investigated design spaces of material properties to optimize the non-volatile/volatile performance, towards the guidance of MFIM-based FTJs applications.
Transition barriers between topological spin textures can be tailored
Fatigue mechanism of NbOx neuron device is unveiled
This study explores the fatigue mechanisms of NbOx-based neuron devices and proposes a physical model to elucidate the threshold switching (TS) behavior in these devices. It highlights the importance of understanding the shift in transition voltages and voltage window shrinkage that affect device reliability. By developing a model that considers the role of oxygen vacancies (Vos), we simulate the continuous drift of operation voltages and propose a method to mitigate the Vo migration, improving device endurance. The key findings include the successful simulation of performance degradation and the identification of Vos redistribution as a key mechanism. The paper contributes to optimizing the performance of NbOx-based devices for artificial neuron applications.
Skyrmionic abacus in magnetic multilayer
Exchange stiffness determined by FMR perpendicular standing spin waves technique
An interconnect design based on skyrmionic textures
Mechanism of resistive switching in HfOx-based memristors
We conducted an in-depth study of the resistance-switching mechanism in classical HfOx-based devices. Through atomic-level characterisation of the dynamic evolution of conductive filaments, we confirmed that the filament system in this type of device has a quasi-core-shell structure, consisting of a metallic Hf6O core and an insulating HfO2 crystalline shell. constitute. The study pinpointed multiple …