We have used x-ray photoelectron spectroscopy in a normal emission configuration to investigate the chemistry of formation of the buried polyimide on Cu interface. We find that oxidation of the Cu surface begins within an hour of the curing process, and that Cu atoms in contact with the polyimide overlayer are completely oxidized to CuO after being stored in air for several days. This conclusion is in stark contrast with that drawn for the inverse system, Cu on polyimide, which has been shown to be essentially unreactive in vacuum over a time scale of several hours. We suggest that the key difference between the two systems is a very slow corrosion reaction perhaps involving trace amounts of chloride ion, water trapped in the polyimide overlayer, and the Cu surface. In contrast, the same experiment involving a Au substrate shows no evidence of reaction over a time frame of weeks.