Hi Jan,

I do the same thing on glass substrates. 100nm of Cr seems a little much. On
glass at least the Cr-layer's purpose is only to improve the gold's adhesion,
not to protect against HF (because Cr is crappy at withstanding HF :)

Possibly your problems result from the HF leeking through the gold through small
pinholes (unavoidable) and then undercutting the gold layer in the Cr-layer.
Consequently the gold rips off in big chunks.

Perhaps you reduce the thickness of the Chromium to something like 10nm (the
thickness I use) and try again. 300nm of gold seems enough to withstand more
than 10min in HF.

P.S.: Cleanliness of the substrate surface and a good, low pressure in the
sputtering machine are THE key to resistant surfaces. Let it pump out overnight
if possible.

Greetings,
Robert R.

> I have a pad metallization consisting of 100nm Chromium and 300nm Gold.
> Both are sputtered right after each other in vacuum. The problem is,
> that I need to etch with 49%HF for a few minutes and the Cr-Au layer is
> practically lifted off the siliconsubstrate. The Wafers were cleaned
> before sputtering but have a native oxide layer. Does anyone know a
> possibility to work with Cr-Au or another metallization which withstands
> the HF? I need gold on top.