*AV*--The die to die is a big deal as stated. 3 degress C, 3 sigma may not sound like much and it may not matter much, but as wafer diameters increase or as internal components deteriorate you can run into problems. I know of no heater that is uniform that does not degrade over time. I have seen this on hotplates for the coat and develop process equipment. We are talking about the heating elements, filaments, etal. all used to keep a wafer chuck uniformly heated. There are too many things that can occur on a daily basis that can affect this type of plate uniformity.
When you are dealing with this type of process and the needs of the ensuing technologies at 0.18u and below, the more precise and accurate you are, the better. If given a choice between the two, I would still opt for the UTEK version even if the throughputs were identical.
This entire discussion is almost a replay of the projection aligner vs stepper discussions of years ago. The projection aligner scanned the entire wafer through a mask. No matter how much you were able to control focus, intensity, wavelength, spectral output of the lamps, lamp uniformity, magnification and distortion of the lenses, there were enough variations to cause issues. This assumed you had the ultimate perfect mask. Couple this with the flatness of the wafer and the flatness of the wafer chuck and it was easy to see why using a perfect reticle amd site by site alignment and exposure on a stepper would yield superior results.
While not as complicated, we are looking at a scaled down version of this same situation. high throughput on a site by site basis is almost always preferable.
Juat my opinion. Besides, UTEK still can do it at smaller feature sizes thereby giving it an edge.
Andrew |
