Empirical Tunneling Model Describing the Retention of 2.5 Mb HfO2 based ReRAM

For the analysis of the high resistive state (HRS) retention characteristics of resistive switching random access memory (ReRAM), we propose to use its intrinsic log-normal statistics (LNS) to evaluate the data regarding shift and tilt of the distribution. We present 2.5 Mb retention data, extract the impact of the bake temperature on the shift and tilt of the distributions and introduce an empirical model which describes the full retention characteristics and allows the extrapolation to different temperatures and times.