Simulation of the resistance switching performance and synaptic behavior of TiO2-based RRAM devices with CoFe2O4 insertion layers.

The electrothermal coupling model of Pt/CoFe2O4/TiO2/TiN devices was established to study their resistive switching characteristics and basic biological synaptic properties in our research. The processes of set and reset are simulated, and the distribution of the temperature, the electric field and the concentration of oxygen vacancies in the dielectric layer are obtained. The switching performance of the TiO2-based device is significantly improved after the CoFe2O4 layer is inserted, with the switching voltage, working current and power consumption being reduced, while the switching ratio is increased. By changing the thermal conductivity of the top electrode, the rupture position of the conductive filament can be controlled. The I-V characteristics of the Pt/CoFe2O4/TiO2/TiN device during the reset and set processes are fitted linearly in logarithmic coordinates, and the ohmic conduction mechanism or the space-limited charge conduction mechanism is mainly satisfied in the high and low resistance states. Finally, the application of dual-layer devices on biological synapses is studied, and the basic biological characteristics of enhancement, inhibition and paired pulse promotion are simulated successfully. In addition, the redox reaction induced by oxygen vacancy migration also promotes the formation and rupture of the conductive filament. Results of the study show this ferrite material as an insertion layer in a resistive random-access memory structure that offers potential for future information storage and bioneuromorphic computation devices.

Solving the issue of increasing forming voltage during device miniaturization in hafnium oxide-based resistive access memory using high-ksidewall material.

An electrothermal coupling model of resistive random access memory (RRAM) was established based on the oxygen vacancy conduction mechanism. By resolving the partial differential equation for the coefficients, the variation process of the device resistance was simulated. In our studies, a device model was proposed which can accurately simulate the whole process of RRAM forming, reset, and set. Based on the established model, a new high dielectric constant (high-k) material (La2O3) is introduced as the sidewall material. The La2O3sidewall material can concentrate the electric field and helps to speed up the formation of conductive filaments. The La2O3sidewall can effectively reduce the forming voltage increase during the miniaturization process. Then, the influence of sidewall thermal conductivity on forming voltage is studied, and it is discovered that low thermal conductivity helps to reduce the model's forming voltage and increase the temperature concentration. These findings serve as a foundation for more studies on the choice of sidewall materials.