https://www.researchgate.net/post/What_is_the_best_way_to_promote_adhesion_between_glass_and_chromium

This may help a bit.

Normally these kind of optical coating materials should stick well on optical glasses. There might be a few reasons why you have adhesion problems:

1. Your glass is not clean and contains carbon like contamination. This can be solved by proper pre-treatment of the glass (e.g. Oxygen plasma)

2. Your vacuum system is not clean enough or you have oil contamination from the vacuum pumps. You can find out if this is the case by putting an RGA on the system and do a gas analysis. Also make sure you have low enough base pressure before you start the deposition (e.g < 10-6 T).

3. In general e-beam evaporation generates lower kinetic energy species arriving at the glass surface compared to sputtered layers. This also depends a lot on the gas pressure you use during deposition. This should be as low as possible. If kinetic energy is low you have less atom intermixing at the surface and therefore adhesion will be less.

4. In sputtering very often the substrate gets a Ar-sputter etch clean to remove physisorbed and chemisorbed water and carbon contamination. This greatly improves adhesion of materials. If you can't do this, then (pre) heating the glass before you load the glass in the vacuum chamber to get rid of water may help adhesion.

5. Depending in the condition of deposition these oxide materials may have high intrinsic film stress which may lead to delamination. You can influence this by playing with the pressure and power levels. But if the glass surface cleanliness is not ok this will not help you since the interface quality is compromises.