Dear all,
I have adhesion problem with LOR3A resist on glass substrates. I have three
types of glass substrate (boro-silicate glass BK7, soda lime glass Gil49 and
erbium doped glass). I use bilayer Lift off method (LOR 3A and Shipley S1805
photorezist) for making narrow stripes (2 - 7 um).
For cleaning of the substrate I use this procedure:
1.chromium trioxide 10 min, 2. boiling in isopropanol 10 min, 3. vapor of
isopropanol 10 min, 4. cleaning with dichlor-methan, UV cleaning + DI water
rinsing.
Then I dry glass wafer 2h/180 deg in oven and then I coat wafer by LOR 3A
4000rpm, bake at 150deg/30min. in oven, then I coat S1805 resist 4000rpm and
bake at 110deg/30min. Then I exposure and develop by MF24-A developer (50-65 s).
If I use BK7 glass substrate, the result is very nice, adhesion is good and
system works well. Nevertheless, when I use Gil 49 or erbium doped boro-silicate
glass, the adhesion of LOR3A resist is very poor, because the stripes of
LOR3A+S1805 float on the substrate surface (in developer) and in the end they
fall away. I thought that it should be better if I use promoter adhesion HMDS
(99.9% /6000rpm and 105deg bake/3min) as the first layer. I have tested also
20-50% concentration of HMDS and Ether, but without success.  I know, that
hydrogen is essential to start the reaction = to break the molecule of HMDS into
two particles of "-Si(CH3)3", but where do I got that hydrogen needed for
breaking HMDS? Do you expect that it is present on the wafer's surface (glass)?
Can I impress amount of  O-H bond on these substrate by e.g. cleaning, UV
cleaning,  plasma treatment,....? Now I use only these tree types of glass
substrates and only one of them (Bk7) is OK.

Could you give me some advice what can I do?

Thanks Ferda