RF Vacuum Microelectronics
Quarterly technical rept. no. 21 Jul-30 Sep 1992
HUGHES AIRCRAFT CO TORRANCE CA ELECTRON DYNAMICS DIV
Pagination or Media Count:
Most of mask set 1 was delivered in the previous quarter. In the second quarter July we took delivery of the final mask layer and re-ordered one other layer due to problems that we discovered with the original design rules. All of the mask layers were checked for registration and found to be within acceptable limits. We found that all the tip sizes were easily resolved with our standard lithographic techniques. Even 1 micrometer diameter opening to etch the smallest molds were easily obtained. The key problem that we encountered in using the newly designed mask was poor layer to layer registration of patterns defined prior to plating on the silicon mold to patterns defined after plating on the silver substrates. It appears that the dimensional stability of the silver is not as good as a typical semiconductor. As a result, there was about 10 micrometers of runout per centimeter when we aligned the gate metal pattern to the tip arrays. This amount of runout is tolerable but unexpected and did result in the loss of some samples due to subsequent emitter to gate shorting. Note that the tips themselves are made with a self-aligned technique and do not require any layer to layer registration tip geometry was not affected by the problem in layer registration.
- Inorganic Chemistry
- Electrical and Electronic Equipment
- Electricity and Magnetism