As semiconductors’ device fabrication is highly integrated, the number of the deposition processes is continuously increasing, and the chamber cleaning process becomes essential for deposition equipment to maintain a normal chamber condition. Although the use of NF₃ gas for the chamber cleaning is common, it causes several environmental and safety issues. However, not much research has been performed on NF₃ plasma at high pressures, such as in cleaning processes. To understand fluorine in NF₃, herein, oxygen was added to N₂ and NF₃ plasma and then compared. Plasma emission spectra were compared using an OES data, and their analyses were performed via a line-ratio method employing the collisional-radiative model. As a result confirmed that the changes in electron temperature, electron density, and chemical species in the plasma could be explained. Additionally, the characteristics of NF₃ plasmas with respect to fluorine were confirmed by comparing the oxygenated N₂ plasma and the NF₃ plasma.

Semiconductor industry is experiencing a rising demand for environmentally friendly processes with the emphasis on green policies and worldwide environmental sustainability. Nitrogen trifluoride (NF3), the most common plasma chamber cleaning agent gas, poses a significant concern as a potent greenhouse gas since it has global warming potential (GWP), 740 times and 6 times higher than that CO2 and N2O. This study investigated the exhaust gas using quadrupole mass spectroscopy (QMS) and analyzed the change in cleaning speed and the type of exhaust gas through plasma monitoring using optical mass spectroscopy (OES). The objective is to lower the use of the amount of NF3 gas in chamber cleaning process to partially contribute the environmental sustainability in the point of semiconductor manufacturing. When a small amount of N2 was added to NF3 whose ratio of 7:23, the cleaning efficiency reached to 90 % compared to NF3 gas alone. Addition of N2 positively affected electron density and temperature to increase the F-radical in remote plasma system. In conclusion, 18 % of NF3 usage amount was reduced during the SiO2 deposition chamber cleaning process.

Plasma plays an important role in semiconductor processes. With the recent miniaturization and integration, the control of plasma became essential for success in the critical dimension of a few nanometers and etch narrow and deep holes with their high aspect ratios. Recently, the etching process has reached physical limitations due to a significant increase in wafer surface temperature under the elevated amount of RF power, affecting not only the warpage phenomenon, but also etching uniformity and etching profiles. Therefore, the plasma characteristics are identified using an invasive single Langmuir probe (SLP) for wafer temperature diagnosis. Optical data is obtained through a non-invasive optical emission spectroscopy (OES) and the plasma parameters are derived to compare and verify with the SLP. Two variables, electron temperature and electron density, are substituted for the heat flux formula to derive the heat flux according to the location. Using a wafer-type temperature sensor, the trends of the derived heat flux values towards wafer chuck were investigated. This study presents a method to calculate heat flux values in real time, anticipate wafer temperatures, and potentially illuminate existing ion heating problems.

The fault detection and classification (FDC) modeling proposed in this study is a research approach that is intended to improve the performance of plasma process models by leveraging optical emission spectroscopy (OES) data containing plasma information (PI) and enhancing model interpretability using explainable artificial intelligence (XAI) algorithms. Status variable identification data that included normal and abnormal states of bias power, pressure, SF₆ gas flow, and O₂ gas flow were collected during a silicon etching process with SF₆, O₂ gas plasma. Additional variables were derived from the OES data and included additional PI, such as O and F radicals, which were computed using actinometry, and electron temperature and electron density computed using the line ratio method. By building a high-performance FDC model and interpreting its results using XAI algorithms, we propose solutions to the limitations of the FDC model in semiconductor plasma processes.