Gold and Aluminium chip bondpads

The pads were always aluminum as most metallizations were aluminum. The bond wires were either gold or aluminum. Gold is better in encapsulated devices as it is more flexible than aluminum and less suspecptible to fatique. The problem with gold/aluminum bonds is "purple plague." Do a search on that term to find out more.

Al

Thin film recording heads use Au pads. One advantage is that the testers make good contact without having to break through any surface oxides.

Nils

Microunity used gold metallization in their attempt to jump ahead of the rest of the industry. The combination of gold's lower resistance, air bridges for insulators, sub-wavelength features to increase photolithographic resolution, etc. was supposed to provide enough of a leap to knock out the competition. They got as far as building the fab when they ran out of money.

It's true almost everybody uses aluminum pads. There are commercial services that will bump the pads with gold, if you have a need for that.

The formation of brittle gold-aluminum intermetallics is extremely well studied. See back issues of IEEE-CHMT. I believe there was a paper about 5 years ago by some Germans that had an excellent analysis of the metallurgy of gold-wire/aluminum-pad bonds.

Yes, it's been studied to death. But I have seen, every 5 years or so, when a new set of engineering graduates come into the production world, a repeat of the problems solved eons ago by thier predecessors.

Why should this happen? Seems like the experienced, but highly paid, professionals are being replaced by new graduates with new ideas. How often they are wrong and have no one to guide them is amazing.

And it seems to occur in far too many scientific/engineering fields. It's amazing how much talent in this country is idle for this reason.

Al