Embedded phase change memory (PCM) show optimized performance and reliability thanks to Ge enrichment of the active GeSbTe material. However, excess Ge tends to segregate in the virgin state, which requires an initial forming process for initializing the PCM device. This work presents the detailed energy landscape model for embedded PCMs before forming and after forming. The model predicts the distribution of PCM resistance as a function of forming conditions and read temperature. The model is validated by physical and electrical data, providing a physical interpretation of the forming process in terms of nanoscale element migration within the PCM.
12 Dec 2020
Pages: 13.3. 1-13.3. 4
2020 IEEE International Electron Devices Meeting (IEDM)