||With the advent of technology evolution, the needs of memory become more important. Resistance random access memory (RRAM) is one of the mainstream research topics in academia and industry to develop the next-generation non-volatile memory to replace flash memory. However, the resistance switching mechanism and conduction model of RRAM are still under debating and keep discussing so far. Meanwhile, the stability and uniformity issues also hinder RRAM real applications. Therefore, this thesis aims to improve stability of RRAM by use of doping and stacking method.|
With development of Moore’s Law, it is unavoidable to face physical limitation. However, the high storage density memory can miniature dimension. Based on our previous research, a device of Pt/LiSiO2/TiN structure has potential of high storage density memory. However, it also has drawbacks of instability and uniformity of resistance distribution. To solve this instability problem, we doped graphene in the Al2O3 switching layer to limit the drift of Li+ ions.
Experimental results show that the conducting current declines with applying a voltage smaller than RESET voltage. Therefore, the number of DC scanning also decides the resistance of RRAM device. In addition, this phenomenon has been further confirmed by AC operations. Data retention experiment with high and room temperature shows excellent capability to store data. The experimental result shows that the ethyne would react with silicon dioxide and form benzene at high temperature, which is beneficial to produce better endurance in RRAM. We proposed a conducting mechanism based on the carrier fitting results.