Dear Alex,
I do the same photolitography on glass (Bk7,soda-lime Gil49 adn newly
designed Er.dop.). I had similar problem with pattern of stripes one month ago.
I use LOR3A and S1805. The samples must be very dry and in some case you
 need to use HMDS (hexamethyldisilazane) as promotor adhesion.
The solution of the second problem is very easy, you need
 cca. 1.2-1.3 times thicker layer of the LOR resist, becouse
 you need make undercut. In your case you should use LOR5A
or LOR7B and S1805. For Lift-Off you don't use CD-26, it's
developer, you must use some REMOVER, like R1165 (works very well).
If you put the substrate in R1165 for 10 min, the metal comes out.

Ferda
> Dear All,
> When I do photolithography on glass substrate I found
> two problems:
> 1) The pattern cannot develop well on the photoresist
> (LOR + S1805), although I found that there is no such
> problem for using SiO2 as substrate.
> 2) It is extremely difficult to lift-off the metal
> layer (50nm Cr and 400nm Au). I put the substrate in
> CD-26 for 2 day but no metal comes out.
>
> Can someone give me ideas to solve the problems?
> Thanks a lot in advance,
> Alex