Modeling And Simulation Of ECM-type Resistive Switching Memory

With the vigorous development of portable mobile multimedia electronic products, non-volatile memory Flash quickly occupys the memory chip market. However, according to the Moore’s Law, the feature size is shrinking as usual, so that after entering the 20 nm technology node, Flash memory is facing severe challenges due to its physical limits. The industry generally believes that 16nm-3D technology will be the final stage of Flash memory. In order to deal with the crisis, many types of new memory, such as ferroelectric random access memory(Fe RAM), phase change random access memory(Pc RAM), magnetoresistive random access memory(MRAM), and resistive random access memory(RRAM) emerged. RRAM is extremely attractive to high integration and low-power technology requirements, because of its simple structure, fast operating speed, ability of scaling down to nanometer, compatibility with CMOS technology and many other advantages. Especially the RRAM based on binary metal oxides has become the focus of the study, and was regarded as the most powerful competitor of the new generation of non-volatile memory. But the current study is not yet ripe for RRAM, for example, there is no unified theory of the resistance switching mechanisms; the durability of the device and consistency is poor in performance, so that a lot of research is required.This paper makes research on the resistance switching mechanisms in two ways based on the current status of research. One is that we carry out a complete mathematical model for describing the resistance switching behavior of a ECM unit on the foundation of the electrochemical type conductive filament mechanism(ECM) that is generally accepted by the industry and some of the results of previous studies. The variation of current with voltage and the growth or dissolution process of a conductive filament are described in detail in a ECM unit during the SET and RESET process. The second part deals with the simulation of the model established in different parameters, such as the variation of current, voltage and the tunneling gap during the SET and RESET process, the effect on I-V characteristics of different electrode materials and many other content using MATLAB tools. On the one hand we can have a better image understand of conduction mechanism in RRAM and verify the reasonableness of the model; on the other hand we can put forward reasonable proposals in the actual RRAM memory manufacturation and what should be noted in order to improve the characteristics of RRAM.