| dc.description.abstract | Quartz possesses desirable properties such as piezoelectricity, insulation, transparency, high hardness, and thermal stability. These characteristics make it suitable for frequency control components, and quartz crystals exhibit various oscillation modes depending on the cut angle. Among them, AT-cut quartz demonstrates minimal frequency variation with temperature changes, providing precise frequency control. Consequently, AT-cut quartz has become the primary choice for frequency control devices.
In the era of the Internet of Things (IoT) and the rapid development of the 5G industry, electronic products are designed to be compact, lightweight, and feature-rich. Quartz frequency control components play a crucial role in communication devices, automotive applications, and various electronic products by providing stable frequency and clock control.
However, the traditional mechanical manufacturing processes are no longer sufficient to meet the design requirements for miniaturization. To address this challenge, Seiko Epson in Japan pioneered the concept of Quartz with Micro Electro Mechanical Systems (QMEMS) by integrating semiconductor fabrication techniques into the quartz industry. This approach enables the creation of special structures in quartz crystals that meet the market demands for high frequency and miniaturization.
Despite leveraging the experience gained from semiconductor wafer processing, applying semiconductor wafer-level processes to quartz frequency control component fabrication still requires substantial research efforts to overcome existing challenges.
ii
This study aims to apply the Inductively Coupled Plasma-Reactive Ion Etching (ICP-RIE) system to perform dry etching processes on AT-cut quartz wafers. The process can be divided into four stages. Firstly, under fixed process parameters, the composition ratios of two etching gases will be varied to identify the gas combination that yields the highest etching rate. Subsequently, within the stable process range of the equipment, an orthogonal array experiment using the Taguchi method will be conducted to analyze the impact of each parameter on the etching response. The effects of each factor will then be validated through a full-factorial experiment. Finally, machine learning techniques will be utilized to optimize the process parameters, aiming to achieve high etching rates and vertical sidewall structures. | en_US |