Advanced mass flow controllers (MFC) with EtherCAT communication protocol and embedded self diagnostics

摘要

Mass Flow Controllers (MFCs) are the most critical components of a gas delivery system used for semiconductor manufacturing. As the technology transitions to advanced nodes below 10 nm, so have the requirements for the MFCs to have embedded self-diagnostics to support Fault Detection and Classification. A few examples of factors contributing to MFC performance degradation are sensor drift and control valve leak-by. Sensor drift can affect accuracy and repeatability of the device which in turn, can cause shifts in CD uniformity. Valve leak-by can cause `first wafer effect' essentially impacting yield. Therefore, developing an in-situ method to detect sensor drift and valve leak-by is beneficial to process tool owners to minimize tool downtime and improve process yield. An algorithm based approach to detect sensor drift and valve leak-by has been developed without interruption to process flow steps or requiring any hardware changes. Utilizing advances in EtherCAT communication protocol as a backbone, data from the devices on these two critical parameters can be logged and monitored to ensure fault-free and consistent device performance. The algorithm and implementation relies on standard EtherCAT indices for commissioning and diagnostic routines.