In generally, silver or gold are easy be reduced and formed on the graphene or graphene oxide surface from their salt compound, but it's not take place on SiO2 surface. What is difference in the bonding or surface properties between them?
(i) Different kind of functional groups on SiO2 surface and graphene oxide. GO contains various kind of surface functional groups such as hydroxy, carboxylic, carbonyl and epoxy etc. which all easily hold the NPs.
(ii) Moreover, zeta potential also plays important role for NPs adhesion with graphene layers. The negatively charged graphene layers (due to the various kind of oxygen functionalities, as described above) easily attracts the cations (Ag+ and Au3+) provides the electrostatic interaction for them.
For achieving good adhesion " some kind of reaction at the interface is perhaps the best possible way" , and is applicable to a great extent for films deposited on a substrate. This is precisely the reason we need thin buffer layers in case of gold and platinum when deposited on SiO2 surfaces.
How about the SiO2 surface? Actually, we can treat SiO2 surface to form the oxygen functional groups (O2 plasma, ...) but can not also deposit on it. I wonder whether that carbon affect on the reduction processing or not