Corrosion is a very complex process that involves many chemical and physical factors. Energy of the open surface and grain boundaries is only one of them.

In corrosion of stainless steels play important role micro-galvanic elements, which arise from inhomogeneities of the composition. At grain boundaries optimal conditions for formation of such galvanic elements arise because the composition of the grain interior and grain boundary differ. In the case of grain boundaries, the second condition of corrosion is also better met than on the surface: electrolyte (water) remains there for a longer time. The deeper the pitting corrosion spreads, the better are the conditions for corrosion and the faster it becomes. By contrast, the smooth steel surface allows the water to be removed quickly.

For the electrochemical corrosion to start, the difference in electrochemical potentials has to be present between the constituents of a mateerial (usually a metal alloy or a cermet). Within a single grain, assuming it has a homogeneous microstructure, the electrochemical potential may be considered equal. At the same time, at a grain boundary the electrochemical potential starts varying from that of the grain due to a alteration of the microstructure (e. g., higher concentration of impurities, dislocations, etc).

If a material is immersed into a liquid or a gaseous substance, which conducts electricity (i.e., contains positively and negatively discharged ions), electrons start moving from the site with the higher potential to the site with a lower potential (in other words,. a galvanic couple forms). As the result, positively discharged ions start appearing at the place with the initially higher potential. The reaction of these ions with the negatively discharged ions from the above-mentioned liquid or gaseous substance is the process of the electrochemical corrosion.

Surface or grain boundary free energy alone doesn't mean very much for the mass transport through the surface and/or grain boundary drift diffusion. Even though without the grain boundary contribution at triple junction the surface diffusion driven by the capillary forces, which involve the product of sürface free energy by the local curvature can cause grooving at the triple junction by transferring mass out of that region towards the surface of the specimen. Especially the local strain energy density due to prestraining as well as the applied uniaxial tension may accelerate rate of grooving drastically, and compressive stress may hinder grooving but induce the formation of hillocks around the grain boundary edges. Corrosive environments may change the surface free energy at the groove lips, and/or accelerates local dissolutions of surface layer, which changes the local curvature at the groove tip. The secondly this dissolution of surface at the groove region is nothing to do with surface mass transfer by diffusion but it is a bulk /liquid reaction taking place locally, which is also treated in our theory dealing with surface and interface stabilities.

Thank you all for your answers.

It is clear that the energy level is just a small aspect of corrosion and there are many different aspects that play a role. One important aspect seems to be the potential differences between the material inside the grain and the grain boundary. Does that meen the grain boundary alsways acts as the more anodic part? Becaus dissolution of Fe is the anodic reation. And if so why?

I am not an expert in corrosion, but I would suggest that a grain boundary may play the role of as cathode, as anode, depending on the chemical composition of the alloy, its condition (plastically deformed, annealed, etc.). For example, if we take a Cr-steel with a sufficient C content for Cr3C2 to precipitate at the grain boundaries, I would suggest that grain boundary will be a cathode in a corrosion process. If we take a cold rolled low-carbon steel, assuming that dislocation density is higher at the grain boundary, I would offer that grain boundary will be an anode in the corrosion process.

What concerns Fe dissolution. In a steel or a cast iron pure Fe usually exhibits a higher electrochemical potential, than other chemical elements (e.g., Cr, Ni) or compounds (e.g., Fe3C), present in the alloy. Moreover, when during corrosion FeOOH an similar compounds form, they will also have lower electrochemical potential than Fe. That is the reason why during the corrosion process Fe all the time plays a role of anode.