First, you can calculate the adhesion between the SiO2 and gold/silver by their surface energies. Or, you  can directly measure the adhesion between them by AFM or other devices.

Hope that helps

Hao

Both gold and silver do not adhere very well to SiO2. For good adhesion you usually use an adhesion layer like Ti or Cr. These materials form a SiO2-MeOx intermediate layer which is a chemical bond. Au and to a lesser extent Ag are noble materials and don't form intermetallic oxide layers. Ag oxidizes a little and might from Ag2O intermediate layers. It is my experience that both Au and Ag sputtered layers stick to SiO2 but just for some hours due to van der Waals bonding. After a few hours water will diffuse into the interface and the films will start to buckle. Even when you do not see buckling, the film is easily removed with tape (tapetest) showing the intrinsic adhesion of these noble metals is bad. I would advise to use a sticking layer of Ti, Cr, Al. 

Thank you, Peter Knapen and Hao Zhang for your answers. It is really helpful. Dear Knapen sir, is there any other sticking materials been used, apart from  Ti, Cr, Al?

Hi Nabarun,

In thin film technology Cr and Ti are the most commonly used adhesion layers but any material which forms an oxide intermediate bond with the SiO2 will stick. Besides Cr and Ti, I have used in the past Al, Mo, Ta, Nb, V, Hf. They form stable oxide bonds. An adhesion  layer of 5-10 nm thickness should suffice to get good adhesion to Au and Ag.

thank you sir.

I  fully agree with Peter. A thin (5nm) Cr interlayer helps to improve the adhesion. However I want to highlight that the surface roughnes and sutface pre-treatment will also play some role. Neverless without interlayer you will never obtain good adhesion of Au or Ag.

The scheme of Peter Knapen is good. Normally, Cr is very common, such as  the coating of AFM probe.