In-situ Process Monitoring Data of PECVD Plasma
The storage capacity of 3D-NAND flash memory has been enhanced by the multi-layer dielectrics. The deposition process has become more challenging due to the tight process margin and the demand for accurate process control. To reduce product costs and ensure successful processes, process diagnosis techniques incorporating artificial intelligence (AI) have been adopted in semiconductor manufacturing. Recently there is a growing interest in process diagnosis, and numerous studies have been conducted in this field. For higher model accuracy, various process data are required, such as plasma, gases, equipment control state, and process result. Also it is important to data must be have the reproducibility of raw data is difficult to acquire due to the data security policy of the chip manufactures.
This data presents various data on the multiple oxide and nitride (ON) dielectric stack deposition process. They include optical emission spectroscopy (OES), the quadrupole mass spectrometer (QMS), state variable identification (SVID) data, and the film thickness each of layer in the multi-layer dielectric. This data can provide valuable information on the equipment control and plasma during the ON stack process. It can be utilized for developing the diagnosis model and identifying the effect of view-port contamination along the ON stack.