Modeling and dispatching refinement for implantation to reduce the probability of tuning beam

Implantation equipment is expected to possess the highest possible productivity due to the short process time in a semiconductor manufacturing factory. The setup time of implantation plays a more important role in improving tool productivity as compared with other processes. Dedicated tools for individual recipe groups avoid setup occurrence but lead to low tool utilization and high equipment cost. Restraining the occurrence of setup is necessary for improving productivity in implant area. Aiming at how to improve same recipe rate, a novel dispatching algorithm is proposed in this paper, which politically selects timings of changing recipes for lots with Q-time constraints according to queue lengths of individual recipes. The proposed algorithm let the interruptions due to Q-time issue can be coupled with unavoidable recipe changes which are caused by that tools exhaust some kind of lots. Compared to the conventional dispatching in which the lots with low arrival rates would interrupt continuous production runs of tools when they almost suffer Qtime over, the rate of tuning beams can be reduced by 12%.