Track: Automation and Control

Abstract

In CMOS technology semiconductor fabrication with typical 30,000 wafers per month capability, there are 300 to 1000 process steps, 200 to 300 product variations and 500 equipment. The processes are Lithography, Etching, Cleaning, Film Deposition, Ion Implantation and Chemical Mechanical Planarization (CMP). The focus is on CMP, which represents about 10% of the total processes in wafer fabrication. It is defined as a planarization process by polishing the wafers to certain thickness and eventually producing smooth surface for next process requirement. The challenges to this process are the product variations, the incoming lot and the correct thickness. As a result, lot queue is high due to waiting time on measurement results and adjusting recipe manually to get the right polishing time. The focus in this paper is to improve lot queue time at CMP, from ex-situ manual process adjustment into automated in-situ data integration of measurement and process adjustment. The study proposes an object oriented algorithm that integrates Manufacturing Execution Systems (MES) with process and measurement equipment by using Visual Basic language. The main goal is to control CMP process and to provide real time feedback data from measurement equipment to the CMP process. Based on data collected, the proposal has successfully reduced CMP waiting time by 13%.