Decoupling Electric Field and Temperature-Driven Atomistic Forming Mechanisms in TaO x /HfO 2 -Based ReRAMs using Reactive Molecular Dynamics Simulations
Simanta Lahkar, Valeria Bragaglia, Behnaz Bagheri, Donato Francesco Falcone, Matteo Galetta, Marilyne Sousa, Aida Todri-Sanial · 2025
Read Decoupling Electric Field and Temperature-Driven Atomistic Forming Mechanisms in TaO x /HfO 2 -Based ReRAMs using Reactive Molecular Dynamics Simulations by Simanta Lahkar, Valeria Bragaglia, Behnaz Bagheri, Donato Francesco Falcone, Matteo Galetta, Marilyne Sousa, Aida Todri-Sanial (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-24468.txt.
Introduction
Resistive memory devices store data by changing how well a thin film conducts, and a first forming step sets that behavior. Simanta Lahkar and co-authors simulated that forming step at the atomic level in 2025 for a two-layer device. Their model let them separate the effects of the applied electric field from the effects of heat, which experiments struggle to pull apart.
Tantalum atoms moved most under voltage, leaving an oxygen-rich zone near the top electrode. Oxygen vacancies clustered near the bottom electrode, where the conducting filament begins to grow. A minimum voltage was needed to start that clustering. Growth then ran on heat-driven creation of new vacancies near the young filament.
Catalog record
| CREATOR | Simanta Lahkar, Valeria Bragaglia, Behnaz Bagheri, Donato Francesco Falcone, Matteo Galetta, Marilyne Sousa, Aida Todri-Sanial |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 65 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-24468
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Decoupling Electric Field and Temperature-Driven Atomistic Forming Mechanisms in TaO x /HfO 2 -Based ReRAMs using Reactive Molecular Dynamics Simulations Simanta Lahkar, Valeria Bragaglia, Behnaz Bagheri, Donato Francesco Falcone, Matteo Galetta, Marilyne Sousa, Aida Todri-Sanial Abstract Resistive random access memories (ReRAMs) with a bilayer TaO x /HfO 2 stack structure have shown unique multi-level resistive switching capabilities. However, the physical processes governing their behavior, and specifically the atomistic mechanisms of forming, remain poorly understood. In this work, we present a detailed analysis of the forming mechanism at the atomic level using molecular dynamics (MD) simulations. An extended charge equilibration scheme, based on a combination of the charge transfer ionic potential (CTIP) formalism and the electrochemical dynamics with implicit degrees of freedom (EChemDID) method, is employed to model the localized effects of applied voltage. Our simulations reveal that tantalum ions exhibit the highest displacement under applied voltage, followed by hafnium ions, while oxygen ions respond only minimally. This results in the formation of a tantalum-depleted, oxygen-rich zone near the positive top electrode (anode), and the clustering of oxygen vacancies near the negative bottom electrode (cathode), where the conductive filament nucleates. This ionic segregation partially shields the bulk dielectric from the applied electric field, hindering further migration of ions in the vertical direction. We find that a minimum threshold voltage is required to initiate vacancy clustering. Filament growth proceeds through a localized mechanism, driven by thermally activated generation of oxygen vacancy defects, which are stabilized near the edge of the nucleated filament at the cathode. † † affiliation: NanoComputing Research Lab, Integrated Circuits Group, Electrical Engineering Department, Eindhoven University of Technology, Eindhoven, 5612AZ, The Netherlands † † affiliation: IBM Research Europe-Zurich, Rueschlikon, Switzerland † † affiliation: NanoComputing Research Lab, Integrated Circuits Group, Electrical Engineering Department, Eindhoven University of Technology, Eindhoven, 5612AZ, The Netherlands † † affiliation: Department of Applied Physics, Eindhoven University of Technology, Eindhoven, 5612AZ, The Netherlands † † affiliation: IBM Research Europe-Zurich, Rueschlikon, Switzerland † † affiliation: IBM Research Europe-Zurich, Rueschlikon, Switzerland † † affiliation: IBM Research Europe-Zurich, Rueschlikon, Switzerland † † affiliation: NanoComputing Research Lab, Integrated Circuits Group, Electrical Engineering Department, Eindhoven University of Technology, Eindhoven, 5612AZ, The Netherlands † † email: [email protected] 1 Introduction Filamentary resistive random access memories (ReRAMs) are resistive switching devices that have gained significant attention for their potential in neuromorphic computing and non-volatile memory technologies owing to their promising characteristics like deep scalability, ease of programmability, and large on/off ratios Chen et al. (2012) ; Lee et al. (2011) ; Govoreanu et al. (2011) . These devices are typically built using a metal-insulator-metal stack design, where the resistance of the insulator can be dynamically altered based on the history of charge flow through it. Chen et al. (2012) ; Lee et al. (2011) ; Dittmann et al. (2021) ; Ma et al. (2019) However, typically, an initial process known as electroforming is required before acquiring this dynamic resistance change capability. During electroforming, a high voltage is applied across the electrodes, inducing the formation of a conductive filament within the insulating layer. Transition metal oxides (TMOs) such as \text{TaO}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{2}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{2}}\text{}} , \text{Ta}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{2}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{2}}\text{}}\text{O}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{5}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{5}}\text{}} , and HfO 2 are common candidates to form the insulating layer, which are often amorphous, in conventional ReRAMs. In such TMO-based ReRAM devices, the filament is formed as a connected region of oxygen-deficient conductive phases within the dielectric, commonly pictured as a continuous network of oxygen vacancies clustered together in the insulating layer Cartoixa et al. (2012) ; Rosário et al. (2018) ; Rosário et al. (2019) ; Ma et al. (2018) ; Urquiza et al. (2021) . The structural evolution of and around this filament allows the device to switch between low- and high-resistance states, enabling the resistive switching behavior that ReRAMs are known for Chen et al. (2012) ; Lee et al. (2011) ; Dittmann et al. (2021) ; Ma et al. (2019) ; Ma et al. (2020) . In recent years, bilayer ReRAM structures combining a layer of a conductive metal oxide (CMO) material, like TiO x , TaO x , and AlO x , and a dielectric HfO 2 layer stacked together between the electrodes have shown significant improvements over single-layer devices Yao et al. (2017) ; Woo et al. (2016) ; Cüppers et al. (2019) ; Stecconi et al. (2022) . These bilayer configurations offer enhanced analog bipolar switching, the ability to achieve multiple resistive states, and reduced noise and stochasticity in switching. Such properties make bilayer ReRAMs particularly attractive for energy-efficient neuromorphic computing applications, such as ReRAM-based crossbar arrays for accelerating deep neural network inference and training Stecconi et al. (2024) ; Wan et al. (2022) ; Gong et al. (2022) ; Choi et al. (2025) , as well as oscillatory neural networks (ONNs), where they act as tunable resistive coupling elements between oscillating neurons Delacour et al. (2021) ; Todri-Sanial et al. (2024) ; Falcone et al. (2025) . Experimental studies on TaO x /HfO 2 -based ReRAMs have shown how CMO and HfO 2 material properties impact their performance characteristics, like resistance range, switching behavior, and yield Stecconi et al. (2022) ; Stecconi et al. (2024) . Furthermore, a physics-based compact model Galetta et al. (2024) accounting for the electro-thermal distribution across the material stack Falcone et al. (2024) ; Falcone et al. (2023) has been recently proposed that can simulate their switching behavior with high accuracy. However, despite the promising characteristics of bilayer ReRAMs, the literature lacks studies addressing their electroforming process from an atomistic perspective, which is crucial to clarify and validate the physical mechanisms underlying the behavior of these devices. On the other hand, the electroforming in conventional single-layer ReRAMs has been extensively studied, providing insights into their behavior aiding in the optimization of the device characteristics Cartoixa et al. (2012) ; Rosário et al. (2018) ; Rosário et al. (2019) ; Ma et al. (2018) ; Urquiza et al. (2021) ; Ma et al. (2020) . Nevertheless, the primary electroforming mechanism in TaO 2 , \text{Ta}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{2}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{2}}\text{}}\text{O}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{5}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{5}}\text{}} , or HfO 2 -based ReRAMs remains a matter of contention in the literature, even for single-layer devices Kumar et al. (2016) ; Kumar et al. (2016) ; Ma et al. (2018) ; Ma et al. (2020) . The behavior of single-layer \text{TaO}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{2}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{2}}\text{}} , \text{Ta}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{2}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{2}}\text{}}\text{O}{\vphantom{\text{X}}}_{\vphantom{\text{2}}\smash[t]{\text{5}}}^{\vphantom{\smash[t]{\text{2}}}\hphantom{\text{5}}\text{}} , or HfO 2 -based ReRAMs has conventionally been attributed to the valence change mechanism (VCM) Waser et al. (2009) . In the VCM model, it is the electric field-driven migration of oxygen vacancies, which generate at the anode Padovani et al. (2012) ; Jeong et al. (2008) , that results in the formation of a conductive filament Urquiza et al. (2021) ; Dittmann et al. (2021) ; Waser et al. (2009) . This filament can break and reform, upon subsequent reversing of the voltage polarity applied to the electrode, enabling resistive switching. Joule heating can further accelerate this process by providing the thermal energy necessary to enhance the electric field-driven migration of oxygen vacancies Dittmann et al. (2021) . In contrast, the behavior of some ReRAMs, such as those based on NiO, is explained by the thermochemical mechanism (TCM), where thermally activated conductivity increase in the metal oxide layer creates a hot channel, leading to ionic segregation driven by