Andrew J. Lohn

A physical model of switching dynamics in tantalum oxide memristive devices

We present resistive switching model for TaOx memristors, which demonstrates that the radius of a tantalum rich conducting filament is the state variable controlling resistance. The model tracks the flux of individual oxygen ions and permits the derivation and solving of dynamical and static state equations. Model predictions for ON/OFF switching were tested experimentally with TaOx devices and shown to be in close quantitative agreement, including the experimentally observed transition from linear to non-linear conduction between RON and ROFF. This work presents a quantitative model of state variable dynamics in TaOx memristors, with direct comparison to high-speed resistive switching data.

Optimizing TaOx memristor performance and consistency within the reactive sputtering “forbidden region”

Standard deposition processes for depositing ReRAM oxides utilize mass flow of reactive gas to control stoichiometry and have difficulty depositing a precisely defined sub-stoichiometry within a “forbidden region” where film properties are discontinuous with mass flow. We show that by maintaining partial pressure within this discontinuous “forbidden region,” instead of by maintaining mass flow, we can optimize tantalum oxide device properties and reduce or eliminate the electroforming step. We also show that defining the partial pressure set point as a fraction of the “forbidden region” instead of as an absolute value can be used to improve wafer-to-wafer consistency with minimal recalibration efforts.

Reactive sputtering of substoichiometric Ta2Ox for resistive memory applications

A major class of resistive memory devices is based on transition metal oxides, where mobile oxygen vacancies allow these devices to exhibit multiple resistance states. Ta2O5 based devices in particular have recently demonstrated impressive endurance and forming-free results. Deposition of substoichiometric Ta2Ox (x < 5) films is a critical process in order to produce the required oxygen vacancies in these devices. This paper describes a physical vapor deposition (PVD) reactive sputtering process to deposit substoichiometric Ta2Ox films. The desired film stoichiometry is achieved by feedback control of the oxygen partial pressure in the PVD chamber. A calibration procedure based on Rutherford backscattering spectroscopy is described for locating the optimum oxygen partial pressure.

Memristive switching: physical mechanisms and applications

Memristive switches are devices which modulate their resistance state depending on the history of the applied current and/or voltage, and have substantial technological potential and as well as scientific interest. In this paper, we review the current state of research on memristive systems. We review the many classes of memristive switching, their physical mechanisms, and their fabrication and operational modes. Next, we discuss the state of the art experimental and modeling/simulation techniques, and the physical insight afforded from these approaches. Finally, we discuss the applications and believed technological potential of these fascinating and exciting devices.

Evaluating tantalum oxide stoichiometry and oxidation states for optimal memristor performance

Tantalum oxide has shown promising electrical switching characteristics for memristor devices. Consequently, a number of reports have investigated the electrical behavior of TaOx thin films. Some effort has been made to characterize the composition of the TaOx films and it is known that there must be an optimal stoichiometry of TaOx where forming and switching behavior are optimized. However, many previous reports lack details on the methodology used for identifying the chemistry of the films. X-ray photoelectron spectroscopy has been the most commonly used technique; however, peak fitting routines vary widely among reports and a native surface oxide of Ta2O5 often confounds the analysis. In this report a series of large area TaOx films were deposited via sputtering with controlled O2 partial pressures in the sputtering gas, resulting in tunable oxide compositions. Spectra from numerous samples from each wafer spanning a range of oxide stoichiometries were used to develop a highly constrained peak fitting...

Dynamics of percolative breakdown mechanism in tantalum oxide resistive switching

Switching dynamics are studied for tantalum oxide resistive random access memory subjected to long-duration constant current pulses for both SET and RESET transitions. The processes draw parallels to the widely studied percolation model for dielectric breakdown. The RESET transition is shown to consist of changes to critical local conduction sites and their effect on performance parameters such as switching speed and energy are discussed. Additionally, the SET transition shows an unexpected minimum stable resistance state. When driven below that state the device is found to increase resistance, returning to the stable state.

Room-temperature Coulomb staircase in semiconducting InP nanowires modulated with light illumination

Detailed electron transport analysis is performed for an ensemble of conical indium phosphide nanowires bridging two hydrogenated n(+)-silicon electrodes. The current-voltage (I-V) characteristics exhibit a Coulomb staircase in the dark with a period of ∼ 1 V at room temperature. The staircase is found to disappear under light illumination. This observation can be explained by assuming the presence of a tiny Coulomb island, and its existence is possible due to the large surface depletion region created within contributing nanowires. Electrons tunnel in and out of the Coulomb island, resulting in the Coulomb staircase I-V. Applying light illumination raises the electron quasi-Fermi level and the tunneling barriers are buried, causing the Coulomb staircase to disappear.

Optical properties of indium phosphide nanowire ensembles at various temperatures.

Ensembles that contain two types (zincblende and wurtzite) of indium phosphide nanowires grown on non-single crystalline surfaces were studied by micro-photoluminescence and micro-Raman spectroscopy at various low temperatures. The obtained spectra are discussed with the emphasis on the effects of differing lattice types, geometries, and crystallographic orientations present within an ensemble of nanowires grown on non-single crystalline surfaces. In the photoluminescence spectra, a typical Varshni dependence of band gap energy on temperature was observed for emissions from zincblende nanowires and in the high temperature regime energy transfer from excitonic transitions and band-edge transitions was identified. In contrast, the photoluminescence emissions associated with wurtzite nanowires were rather insensitive to temperature. Raman spectra were collected simultaneously from zincblende and wurtzite nanowires coexisting in an ensemble. Raman peaks of the wurtzite nanowires are interpreted as those related to the zincblende nanowires by a folding of the phonon dispersion.

Analytical estimations for thermal crosstalk, retention, and scaling limits in filamentary resistive memory

We discuss the thermal effects on scaling, retention, and error rate in filamentary resistive memories from a theoretical perspective using an analytical approach. Starting from the heat equation, we derive the temperature profile surrounding a resistive memory device and calculate its effect on neighboring devices. We outline the engineering tradeoffs that are expected with continued scaling, such as retention and power use per device. Based on our calculations, we expect scaling to continue well below 10 nm, but that the effect of heating from neighboring devices needs to be considered for some applications even at current manufacturing capabilities. We discuss possible designs to alleviate some of these effects while further increasing device density.

Degenerate resistive switching and ultrahigh density storage in resistive memory

We show that in tantalum oxide resistive memories, activation power provides a multi-level variable for information storage that can be set and read separately from the resistance. These two state variables (resistance and activation power) can be precisely controlled in two steps: (1) the possible activation power states are selected by partially reducing resistance, then (2) a subsequent partial increase in resistance specifies the resistance state and the final activation power state. We show that these states can be precisely written and read electrically, making this approach potentially amenable for ultra-high density memories. We provide a theoretical explanation for information storage and retrieval from activation power and experimentally demonstrate information storage in a third dimension related to the change in activation power with resistance.