Modeling and Simulation of Hysteron Dynamics in HfO2-Based Memristive Devices

Published studies on memristive devices primarily focus on the conventional current-voltage major hysteresis loop, often overlooking the switching dynamics that emerge under controlled amplitude excitation. The resulting major and minor loops, often referred to as hysterons, provide a more stringent...

Descripción completa

Detalles Bibliográficos
Autores: Miranda, E.|||0000-0003-0470-5318, García, H.|||0000-0003-1329-8806, Vinuesa, G.|||0000-0003-0389-3409, Suñé, Jordi|||0000-0003-0108-4907, Campabadal, Francesca|||0000-0001-7758-4567, Bargallo Gonzalez, Mireia|||0000-0001-6792-4556, Castán, Helena|||0000-0002-3874-721X, Dueñas, S.|||0000-0002-2328-1752
Tipo de recurso: artículo
Fecha de publicación:2026
País:España
Institución:Universitat Autònoma de Barcelona
Repositorio:Dipòsit Digital de Documents de la UAB
Idioma:inglés
OAI Identifier:oai:dnet:uabarcelona_::e1a2a8817bd9a82b623572b279db263e
Acceso en línea:https://ddd.uab.cat/record/328201
https://dx.doi.org/urn:doi:10.1109/LED.2026.3679261
Access Level:acceso abierto
Palabra clave:Memristor
Resistive switching
SPICE
Descripción
Sumario:Published studies on memristive devices primarily focus on the conventional current-voltage major hysteresis loop, often overlooking the switching dynamics that emerge under controlled amplitude excitation. The resulting major and minor loops, often referred to as hysterons, provide a more stringent benchmark for model validation, while also offering deeper insight into the key physical mechanisms governing resistive switching. In this work, hysteron behavior in HfO₂-based memristive devices is analyzed within the framework of the Dynamic Memdiode Model (DMM). Hysterons are particularly relevant for neuromorphic computing, as they encode synaptic weights. Both experimental and modeling results indicate that the switching dynamics are predominantly governed by the applied voltage, whereas thermal effects associated with dissipation in the conductive filament appear to play a secondary role or remain implicit within the proposed description. Overall, these findings demonstrate that the DMM provides a robust and effective framework for the analysis and compact modeling of memristive devices. An updated LTspice implementation of the DMM is provided.