Testing Algorithm Parameter and Device Area Effect in HfO₂OX-RRAM

We demonstrated an optimization procedure of SET and RESET parameters in HfO2-based resistive random-access memory (OX-RRAM). The vertical layout for the OX-RRAM is TiN/AlOxTiOx/TiN (top electrode/top interfacial layer/insulator/bottom interfacial layer/bottom electrode). There are four parameters that could decide the four performances of OX-RRAM, which are compliance current, interval operation voltage, maximum interval voltage and device area. The OX-RRAM device performances are compared with resistance ratio (RRESET/RSET) and endurance. From the power consumption, resistance ratio and resistance endurance, we suggested a methodology of interval voltage optimization instead of compliance current in our OX-RRAM. After optimization, interval voltage 0.125 volts, maximum interval 4.0 volts with 10×10μm2 providing 2.22.4 resistance ratio (RRESET/RSET) with over 200 cycles was applied. The 10×10μm2 has higher resistance ratio (RRESET/RsET) 2.2,_,2.4 with 200 cycle. However, 35×35 μm2 could also sustain 255 cycles with 0.81.2 resistance ratio (RRESET/RSET). The flexibility of device area could provide different resistance ratio and endurance combination, where no further voltage or current optimization is needed. The smaller area OX-RRAM could apply as 'buffer' memory, which requires small device area with placement flexibility. The larger area OX-RRAM with longer endurance could apply as 'storage' memory, which requires robust endurance under higher operating frequency.