This is a question that has many different definitions (depending on who you ask), and therefore “levels” of cleanliness from visibly dirty all the way down to free of impurity at the atomic level. How much cleanliness is required depends heavily on the application; knowing what the right level is for surface coatings demands some intimate knowledge of the system being evaluated.
It is helpful to know the nature of the substrate being coated in terms of surface chemistry, the basic constituents of the coating being applied (and how they react with each other and surfaces), whether it relies on mechanical interlock and/or chemical bonding, as well as what kinds of stresses the coated article will be subjected to under normal use conditions.
By far, proper selection of surface preparation (cleaning) methodology has been the most difficult challenge we’ve faced while scaling up phosphonic acid monolayer treatments on metals. Reaction of organophosphonic acids with metal oxides generally requires removal of as much surface contamination as possible in order to expose the underlying metal oxide(s) present on metal substrates. If there just a single molecular layer of fatty acids bound to the surface as metal carboxylates it will dramatically hinder the reaction of organophosphonic acids with metal oxides. This issue can be extended to many other coatings that rely on chemical bonding of the coating to the substrate. For the purpose of simplicity in this discussion, however, I will focus on illustrating the formation of organophosphonic acid monolayers on metal oxide surfaces.