Scheduling of cluster tools with k-periodic chamber cleaning operations

Abstract

Cluster tools are widely being used in diverse wafer fabrication processes as the wafer quality issues become critical concerns in semiconductor manufacturing. As the process quality control in cluster tools has become increasingly stringent and important, fabs have introduced periodic chamber cleaning operations to prevent wafer quality failures caused by impurities with in process chambers. Recently, it has been reported that about 70–80% of the cluster tools in actual fabs are periodically cleaned to eliminate residual particles and heats inside process chambers. However, the chamber cleaning operations complicate cluster tool operations significantly. In particular, larger chamber cleaning periods significantly increase tool operation cycles and make scheduling analysis and optimization intractable. So, little is known on scheduling cluster tools with a general cleaning period. Therefore, in this research, we examine the scheduling of cluster tools with general k-periodic cleaning operations. First, We confirm that k-periodic chamber cleaning induces an additional decision problem of optimizing the cleaning plan, and introduce a simple algorithm to determine an efficient cleaning plan under a determined sequence. Next, we developed a new robot task sequence that minimizes the tool cycle time for most process recipes for single process step cluster tools and proposed a new scheduling strategy that shows powerful performance in series-parallel cluster tools. Finally, we devised an near-optimal scheduling strategy for linear