Prediction on MRAM Etching Endpoint by Response Surface Method

Abstract: STT-MRAM (Spin-Transfer-Torque Magnetic Random Access Memory) with high-density is considered as one of the most promising storage candidates potential applications. In process MRAM manufacturing, etching step should be stopped precisely at specific material layer. The dielectric layer protected certain coverage. Then subsequent steps continue. It crucial to detect endpoint during fabrication process.
 paper, factors influencing rate are analysed, including gas pressure, temperature, ion sheath thickness, self-biased DC voltage and RF power frequency, respectively. An approach based on Response Surface Method (RSM) adopted predict process. optimized interplay relationship set up among et al.. shows that RSM an effective statistical method for optimization stop technology, especially when complex condition options involved. simulation results demonstrate sidewall smoothness can improved under environment configuration.