Experimental and Theoretical Investigation of Minimization of Forming-Induced Variability in Resistive Memory Devices
- PDF / 10,227,963 Bytes
- 6 Pages / 612 x 792 pts (letter) Page_size
- 27 Downloads / 184 Views
Experimental and Theoretical Investigation of Minimization of Forming-Induced Variability in Resistive Memory Devices Brian L. Geist1, Dmitri Strukov2 and Vladimir Kochergin1 1 MicroXact, Inc., Blacksburg, VA 24060-6376, U.S.A. 2 Electrical and Computer Engineering Department, UC Santa Barbara, Santa Barbara, CA 93106-9560, U.S.A. ABSTRACT Resistive memory materials and devices (often called memristors) are an area of intense research, with metal/metal oxide/metal resistive elements a prominent example of such devices. Electroforming (the formation of a conductive filament in the metal oxide layer) represents one of the often necessary steps of resistive memory device fabrication that results in large and poorly controlled variability in device performance. In this contribution we present a numerical investigation of the electroforming process. In our model, drift and Ficks and Soret diffusion processes are responsible for movement of vacancies in the oxide material. Simulations predict filament formation and qualitatively agreed with a reduction of the forming voltage in structures with a top electrode. The forming and switching results of the study are compared with numerical simulations and show a possible pathway toward more repeatable and controllable resistive memory devices. INTRODUCTION Resistive memory materials and devices (often called memristors) are an area of intense research, due to the enormous promise that such devices hold for digital data storage and processing applications. A memristor [1,2] is a 2-terminal electrical circuit element that changes its resistance depending on the position of vacancies inside the device which can form a filament through electroforming. [3] Typically, a memristor is made of a specific thin film material sandwiched between two metallic electrodes (wires). The resistance of a thin film can be switched either in continuous or binary fashion by applying a voltage to the electrodes. At present, despite of significant progress made to date, the understanding of resistive switching (forming, as well as the processes taking place during set and reset operations) is still not fully developed. Recently, the numerical reset model based on temperature/field-driven ion migration in metal-oxide-metal memristive devices was suggested. [4] In this contribution we significantly expand the model by including a thermophoresis contribution and by modeling not only the reset process but also the device forming process. THEORY Development of a solid model with cylindrical symmetry Titanium oxide was chosen as the modeling material in order to analyze the predicted behavior of memristor devices. The electrical and thermal properties of titanium oxide are already well established [5] which allow for the material to be modeled using COMSOL®. Pure titanium dioxide has a much higher resistivity relative to Magnèli phases of titanium oxide compared to what has been reported for hafnium dioxide [4] making it a more suitable candidate for studying the filament forming process.
Memristive behavi
Data Loading...
