Proceedings of the
9th International Conference of Asian Society for Precision Engineering and Nanotechnology (ASPEN2022)
15 – 18 November 2022, Singapore
doi:10.3850/978-981-18-6021-8_OR-17-0226
Simulation of Planarization Model Considered with Temperature Function in Patterned Cu CMP
1Department of Mechanical Engineering, Pusan National University, Busan, South Korea
ABSTRACT
In CMP, unlike blanket wafers, pattern wafers are affected by parameters such as line width and pattern density. In the previous research, a contact mechanism-based mathematical model considering the geometrical characteristics of oxide film pattern was derived, and it showed good consistency in predicting the step height reduction of oxide film pattern. However, the results of the copper pattern compared using the proposed model showed a relatively large error. Therefore, for copper CMP, not only geometrical properties, but also chemical factors are essential for deriving a model equation. Copper removal follows Preston's equation, which is proportional to the product of pressure and the relative velocity of the pad and wafer, and the chemical reactivity of the material follows Arrhenius' equation. In general, copper is more affected by temperature than oxide during CMP, and as the polishing temperature rises, the amount of material removal also increases. Due to these characteristics, in this study, a modified semi-empirical model considering the effect of temperature in the existing model equation was presented. The empirical formula obtained through the experiment can be expressed as a function of temperature, and is as follows. (1) Etching amount (2) total removal amount according to polishing temperature. First, the copper coupon was etched while controlling the slurry temperature to obtain copper etching rate according to the temperature. Through the experiment, it was confirmed that the etching rate of copper increased as the temperature increased, and it was found that the etching rate was proportional to the process time. Second, polishing was performed several times using polishing equipment to confirm the amount of copper removal according to the polishing temperature. As a result of the experiment, a polishing temperature profile was confirmed for each wafer, and the amount of copper removal also increased as the polishing average temperature increased. Finally, we present a modified MRR expression using the obtained semi-empirical models and contact mechanism theory.
Keywords: Chemical mechanical planarization, Copper, Temperature, Dynamic etch, Removal amount, Modeling
1Department of Mechanical Engineering, Pusan National University, Busan, South Korea
ABSTRACT
In CMP, unlike blanket wafers, pattern wafers are affected by parameters such as line width and pattern density. In the previous research, a contact mechanism-based mathematical model considering the geometrical characteristics of oxide film pattern was derived, and it showed good consistency in predicting the step height reduction of oxide film pattern. However, the results of the copper pattern compared using the proposed model showed a relatively large error. Therefore, for copper CMP, not only geometrical properties, but also chemical factors are essential for deriving a model equation. Copper removal follows Preston's equation, which is proportional to the product of pressure and the relative velocity of the pad and wafer, and the chemical reactivity of the material follows Arrhenius' equation. In general, copper is more affected by temperature than oxide during CMP, and as the polishing temperature rises, the amount of material removal also increases. Due to these characteristics, in this study, a modified semi-empirical model considering the effect of temperature in the existing model equation was presented. The empirical formula obtained through the experiment can be expressed as a function of temperature, and is as follows. (1) Etching amount (2) total removal amount according to polishing temperature. First, the copper coupon was etched while controlling the slurry temperature to obtain copper etching rate according to the temperature. Through the experiment, it was confirmed that the etching rate of copper increased as the temperature increased, and it was found that the etching rate was proportional to the process time. Second, polishing was performed several times using polishing equipment to confirm the amount of copper removal according to the polishing temperature. As a result of the experiment, a polishing temperature profile was confirmed for each wafer, and the amount of copper removal also increased as the polishing average temperature increased. Finally, we present a modified MRR expression using the obtained semi-empirical models and contact mechanism theory.
Keywords: Chemical mechanical planarization, Copper, Temperature, Dynamic etch, Removal amount, Modeling
