A Mechanistic Analysis of Oxygen Vacancy Driven Conductive Filament Formation in Resistive Random Access Memory Metal/NiO/Metal Structures
Study could lead to more efficient electrically switchable resistive random access memory devices.
Resistive Random Access Memory (RRAM) devices have drawn much interest in the last decade, particularly the concept of a memristor. In this case, the so-called memristance, which provides the relationship between the change in charge (time integral of the current) and flux (time integral of the voltage), is not a constant as in linear elements, but a function of the charge, resulting in a nonlinear circuit element. Applications of such two-terminal electrical devices that provide high densities and low-power operation include, for instance, neuromorphic-type computing elements.
This area of research led to a study on the effects of ionizing radiation on such devices. Significant focus on filamentary-type resistive switching (RS) mechanisms emerged, where formation/rupture of a conductive filament (CF) ensures successive switching in the non-volatile metal-insulator-metal (MIM) memristors, dependent on the switching material. In such a RRAM device, binary oxide MIM structures are constructed using an insulating layer stacked between two electrodes, which can be built either symmetrically or asymmetrically using the same or different top or bottom electrodes, respectively.
In the filamentary RS mechanism, following the CF forming stage, where a compliance current is used for controlling its size, operation depends on the migration of ions across the metal oxide in the SET (RESET) stages upon application of positive (negative) voltage in a bipolar RRAM, or of the same polarity voltage in a unipolar system. The rupture of the CF causes a High Resistance State (HRS), and its re-formation results in a Low Resistance State (LRS). Proposed conductance mechanisms by metal cations or positively charged oxygen vacancies, namely electrochemical metallization memory or conductive-bridge memory, and valence change memory mechanisms, respectively, are most common.
A thermochemical mechanism was also proposed. However, despite much promise and progress in this field, many issues remain, requiring further understanding of the memristive mechanism for specific MIMs. Indeed, a consensus on materials selection has still not been reached because properties such as reliability, switching speed, or the range of resistance states, depend on the materials used.
Among a multitude of metal oxide selections for RRAM devices, the behavior of p-type NiO, one of the earliest studied, still raises questions on the mechanism of operation. Following early work, RS characteristics of NiO were determined in a number of examples demonstrating high stability and reliable memory characteristics, high speed, low voltage, fast programming, and compatibility with the CMOS process. Moreover, inclusion of an oxygen exchange layer (Ta) in a Pt/NiO/Cu device enabled reaching a very high resistance ON/OFF ratio of 106.
Characterization of NiO filaments by magnetoresistance demonstrated their structural evolution, and also that multi-filaments are involved in the LRS, rupturing separately during RESET. Unipolar memristive behavior for NiO-based MIMs was demonstrated, indicating that reliable RS depends on the oxygen partial pressure during growth, so that the initial oxygen vacancy defect configuration affected the reliability of RS. Experiments for NiO/Pt films using time-of-flight secondary ion mass spectroscopy and conductive atomic force microscopy (C-AFM) measurements showed that oxygen atoms move to the anode, changing the surface composition, and therefore the resistance.
Oxygen vacancy migration, such as in a valence change memory mechanism-type system, has also been postulated. Although Ni vacancies in p-type NiO can be present in films, when the growth conditions are modified, appreciable concentration of oxygen vacancies is achieved. Indeed, investigation of diffusion of oxygen vacancies in epitaxial NiO by local multimodal scanning probe microscopy was reported, consistent with earlier work. Nevertheless, the NiO-based MIM system for RRAM application, which demonstrates encouraging characteristics, still poses questions on the role of oxygen vacancies in RS for this MIM system.
This work was done by Handan Yildirim and Ruth Pachter for the Air Force Research Laboratory. AFRL-0260
This Brief includes a Technical Support Package (TSP).

A Mechanistic Analysis of Oxygen Vacancy Driven Conductive Filament Formation in Resistive Random Access Memory Metal/NiO/Metal Structures
(reference AFRL-0260) is currently available for download from the TSP library.
Don't have an account?
Overview
The document titled "A Mechanistic Analysis of Oxygen Vacancy Driven Conductive Filament Formation in Resistive Random Access Memory Metal/NIO/Metal Structures" is an interim report authored by Handan Yildirim and Ruth Pachter from the Air Force Research Laboratory, published on December 14, 2017. It focuses on the mechanisms underlying conductive filament (CF) formation in resistive random access memory (RRAM) devices, particularly those utilizing nickel oxide (NiO) as the oxide layer in metal-insulator-metal (MIM) structures.
The report highlights the growing interest in RRAM devices as nonvolatile memory candidates, driven by advancements in memristor technology. Despite this interest, the atomic-level mechanisms that facilitate CF formation in MIM structures, especially with NiO, remain inadequately understood. The authors employ density functional theory (DFT) with a Hubbard-type on-site Coulomb correction to analyze the propensity for CF formation in NiO when oxygen vacancies are introduced, considering the effects of different electrode materials such as silver (Ag) and platinum (Pt).
Key findings of the study include the identification of contributions from metal-induced gap states to the electronic density of states, which accommodate oxygen vacancy states. The research demonstrates that the interface region of the MIM structure is more susceptible to reduction than the bulk region, particularly in Ag/NiO/Ag configurations. Additionally, the study reveals a tendency for oxygen vacancy clustering, which is crucial for CF formation. By introducing ordered oxygen vacancies into the oxide layer, the authors observe the emergence of extended defect states within the forbidden gap, leading to defect-assisted transport mechanisms.
The report emphasizes the significance of understanding these mechanisms for the development of efficient RRAM devices. The findings contribute to the broader knowledge of how oxygen vacancies influence electronic properties and transport phenomena in oxide materials, which is essential for optimizing memory device performance.
Overall, this document serves as a valuable resource for researchers and engineers in the field of materials science and memory technology, providing insights into the fundamental processes that govern the behavior of RRAM devices and paving the way for future advancements in nonvolatile memory solutions.
Top Stories
INSIDERRF & Microwave Electronics
FAA to Replace Aging Network of Ground-Based Radars
PodcastsDefense
A New Additive Manufacturing Accelerator for the U.S. Navy in Guam
NewsSoftware
Rewriting the Engineer’s Playbook: What OEMs Must Do to Spin the AI Flywheel
Road ReadyPower
2026 Toyota RAV4 Review: All Hybrid, All the Time
INSIDERDefense
F-22 Pilot Controls Drone With Tablet
INSIDERRF & Microwave Electronics
L3Harris Starts Low Rate Production Of New F-16 Viper Shield
Webcasts
Aerospace
Powering America’s EV Future: Connect, Collaborate, Innovate
Energy
Hydrogen Engines Are Heating Up for Heavy Duty
Internet of Things
A FREE Two-Day Event Dedicated to Connected Mobility
Power
SAE Automotive Podcast: Solid-State Batteries
Power
SAE Automotive Engineering Podcast: Additive Manufacturing
Defense
A New Approach to Manufacturing Machine Connectivity for the Air Force



