This is a no-revenue generating web site.
The content of this page is a thought experiment. The idea is the stability of pressure within the chamber as the processes within the recipe progress from one process-step to the next. (Some OEM's design their tools with the capability of running several process-steps within the same recipe, and each process-step may have its characteristic power, pressure, gas mixtures, etc.) To simplify the experiment, the entire recipe is run in the viscous flow region of pressures. First, I consider the dual-chamber single-pump case, then I consider single-chamber single-pump case. The reason for the former case is economizing on the cost of pumps.
Figures 1 and 2 are schematics of a dual-chambers single-pump system. The chambers are isolated from one another except where they meet at the pump inlet. The figures depict the movement of the valves: in Fig.1 the movement is linear, and in Fig. 2 it is rotary. The linear motion has no serious impact on the simultaneous pressures in both chambers but the rotary does. The geometric setup is shown in Fig.3. The difficulty of maintaining exactly similar pressures in both chambers when the action of the valve is rotary is visualized by the geometrical setup in Fig. 4. It is difficult to orient the armature of the valve in such a way as to have half the area of the pump inlet available to both cambers at all time. By the nature of the geometry of circles, plots of the fraction of the area of the pump inlet versus the central angle confined by the segments of the circles at the pump center show nonlinear relations for a variety of cases considered.
The process engineer, to increase the utility of the tool - that is, expand the portfolio of recipes that can be run for a variety of IC designs, would prefer a linear relation as shown in Fig. 6. In reality, because of the geometrical presentation, the behavior is more non-linear further limiting the process windows on the tool. Moreover, it is impossible for such a design to have both chambers delivering the identical process results because the slight differences in pressure; thus, chamber matching would not be feasible.
In conclusion, the designer may find a butterfly valve as a viable solution for the single-pump and dual-chamber configuration. The designer may be forced to do difficult things with the design to avoid patent infringements. This is another demonstration of the limitation the hardware may impose on a process and chamber matching.
This site was last updated 12/17/09
© 2009 Stephen Hyatt
Web Site Administrator Novella Web Design