Field-Driven Hybrid Filament Formation Governs Switching in Ta–HfO 2 –Pt Memristors
Ashutosh Krishna Amaram, Aditya Koneru *, Subramanian KRS Sankaranarayanan ** · 2026
Read Field-Driven Hybrid Filament Formation Governs Switching in Ta–HfO 2 –Pt Memristors by Ashutosh Krishna Amaram, Aditya Koneru *, Subramanian KRS Sankaranarayanan ** (2026) 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-2605-29127.txt.
Introduction
Ashutosh Krishna Amaram, Aditya Koneru, and Subramanian KRS Sankaranarayanan simulate switching in Ta-HfO2-Pt memristors, where a conductive bridge was long blamed on oxygen vacancies. Molecular dynamics runs with charge transfer show metal cations also migrate, forming a hybrid filament rich in tantalum cations and poor in oxygen. Varying starting vacancy counts and layouts in the oxide changes the filament's shape and size. Switching follows field-driven formation and rupture of that filament, and the simulations match experiment, guiding designs that cut cycle and device variability.
Catalog record
| CREATOR | Ashutosh Krishna Amaram, Aditya Koneru *, Subramanian KRS Sankaranarayanan ** |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 53365 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.29127
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Field-Driven Hybrid Filament Formation Governs Switching in Ta–HfO 2 –Pt Memristors Ashutosh Krishna Amaram, Aditya Koneru *, Subramanian KRS Sankaranarayanan ** Abstract Memristive devices have gained significant attention for their potential in next-generation non-volatile memory and neuromorphic computing architectures. Among emerging candidates, transition metal oxides have proven particularly promising. While the switching mechanism in Ta/HfO 2 /Pt devices was long attributed solely to oxygen vacancy based filaments, recent experimental evidence suggests a more complex dual-regime: the diffusion of metal cations also contributes to the formation of a conductive bridge. However, the precise atomistic mechanisms governing this metal cation migration remain poorly understood. Additionally, the role of defects such as oxygen vacancies present in the transition metal oxide in determining the final filament size and shape is also not well understood. Here, we employ molecular dynamics (MD) simulations with dynamic charge transfer to provide a detailed analysis of the atomistic mechanisms governing the co-formation of Ta-cation and oxygen-deficient filaments. We clearly show how varying the initial oxygen vacancy concentrations and spatial configurations within the HfO 2 matrix influences the final morphology and dimensions of the conductive filament. The switching is governed by field-driven formation and rupture of a hybrid Ta-cation-rich, oxygen-deficient filament in HfO 2 . Our simulations closely match experiment, validating the model as a robust framework for understanding switching in oxide memristors and guiding designs that reduce cycle-to-cycle and device-to-device variability—key barriers to high-performance devices. 1 Introduction Memristors have gained a lot of attention owing to their applications in non-volatile memories [ 21 , 10 , 15 ] and human-brain inspired neuromorphic computing architectures [ 31 , 1 ] . The device architecture of a memristor is a simple metal-insulator-metal (MIM) stack where a variety of semiconducting materials, both transition metal oxides [ 41 ] and 2-dimensional materials [ 30 ] , have been tested and have shown promising results in the form of low voltage consumption [ 9 ] , fast-switching speeds [ 39 , 4 ] , and high endurance and retention characteristics [ 49 ] . These characteristics make memristors highly attractive for realizing high-density integration of electronics [ 17 ] . The physical mechanism associated with switching of different memristive devices is a characteristic of the semiconducting material. Transition metal oxides like VO 2 switch because of an insulator-metal transition that is triggered when the device reaches the set voltage and the phase transition temperature [ 35 ] . In complex perovskites like PCMO, switching is mainly caused by charge trapping at the metal-semiconductor interface [ 23 ] . However, the majority of semiconductors which have been explored for memristive applications form a filament that acts like a conductive bridge between the two electrodes. Depending upon the composition of the filament, this class of filamentary based memristors is divided into two sub-categories. One is valence-charge-memory (VCM) [ 20 , 5 , 12 , 7 , 16 , 51 ] where the metal electrodes are mostly inert in nature like Pt and the insulator materials are transition metal oxides like HfO 2 [ 50 ] , TiO 2 [ 37 ] and Ta 2 O 5 [ 19 ] . In this class of memristors, the primary switching mechanism for High Resistance State (HRS) to Low Resistance State (LRS) switching is the formation of a positively charged oxygen vacancy filament between the two electrodes. The other category of filamentary based memristive devices is electrochemical metallization memories (ECM) [ 27 , 11 , 26 , 2 ] . In this class, the memristor stack is composed of highly diffusible metal electrodes like Cu, Ag, Al, Ni etc. and the insulator is composed of transition metal oxides like HfO 2 , TiO 2 and Ta 2 O 5 or 2-dimensional materials like MoS 2 [ 32 , 8 , 48 , 25 , 28 , 3 , 24 ] or hBN [ 42 , 14 , 38 , 29 , 47 ] . In these devices, the primary switching mechanism is governed by the formation of a metallic filament between the two electrodes. Recent advancements in memristive devices have revealed a complex switching behavior governed by the simultaneous migration of metal cations and oxygen-deficient regions. Experimental studies using Scanning Tunneling Microscopy (STM) have confirmed that metal cation motion complements oxygen anion movement in AO x devices (where A = Ta, Ti, or Hf) [ 43 ] . Specifically, investigations into Ta/HfO 2 /Pt hetero-structures using Scanning Transmission Electron Microscopy (STEM) and Electron Energy Loss Spectroscopy (EELS) have directly visualized conductive bridges characterized by both Ta-cation enrichment and oxygen vacancy concentration [ 18 ] . These findings reveal that earlier models with a simplified theory of only oxygen vacancies participating in the switching mechanism are not sufficient and a hybrid model is required to understand the switching characteristics of the memristive device which would integrate the migration of both oxygen vacancies and metal cations. Despite previous experimental observations, a step-by-step physical mechanism governing the formation and dissolution of these hybrid filaments remains elusive. In particular, the temporal sequence of the participating species—that is, whether the oxygen vacancies are first nucleated to create a favorable pathway for metal cation diffusion or these cations diffuse first across the semiconducting layer causing the oxygen anions to migrate towards the electrode region leaving behind oxygen vacancies—remains unclear. Furthermore, there is a critical lack of research correlating the initial stochastic distribution of oxygen vacancies with the resulting filament geometry. These vacancies might arise while depositing the material or during the fabrication process. The distribution of these defects is not deterministic in nature and is rather randomly arranged; thus, the role of such defects in governing the geometric and electrical parameters of the conductive filament must be understood to fine-tune the processing parameters of device fabrication and help the user gain a better control over device performance. Addressing this gap is essential, as cycle-to-cycle and device-to-device variability remain the primary bottlenecks preventing the large-scale commercialization of memristive technology, particularly in the use of these memristive devices in non-volatile memories and neuromorphic computers. To overcome this challenge, a robust modeling framework is required where one can simulate different device configurations and study the entire kinetic process governing the switching mechanism of the device. Moreover, the modeling framework should be able to quantify the main cause of variability, thus allowing one to fabricate more reliable memristive systems. By bridging the gap between atomistic insights and continuum scale parameters, we will have a powerful tool capable of predicting the performance of such devices with different configurations and operating conditions. Here, we present a comprehensive molecular dynamics based framework with dynamic charge transfer that reveals the governing mechanism of Ta rich filament formations in Ta-HfO 2 -Pt memristive devices, which is further validated with charge maps and IV characteristics. Unlike previous approaches that had a pure continuum based modeling approach, our framework integrates atomistic scale trajectories produced by running MD simulations on devices exposed to different initial defects and electrical stimuli with IV characteristics. The IV characteristics are obtained by an analytical model that accounts for the number of Ta cations participating in the filament formation and rupture process. Thus, the model helps establish a coherent structure-property correlation for