Corrosion of 9Cr Steel in CO2 at Intermediate Temperature II: Mechanism of Carburization

Abstract

In parallel to the formation of a duplex oxide scale, 9Cr-1Mo steel carburizes strongly under CO 2 at 550°C and this carburization accelerates with time. It is observed that an increase of the total CO 2 pressure in the environment from 1 to 250 bars induces a higher carbon deposition in the inner Fe-Cr rich spinel oxide layer. In order to explain this phenomenon, modelling of the carburization process was carried out. A mechanism involving gas diffusion of CO 2 and CO through the oxide layer, the Boudouard reaction and carbon diffusion through the metallic substrate is proposed.

Corrosion of 9Cr Steel in CO2 at Intermediate Temperature I: Mechanism of Void-Induced Duplex Oxide Formation

“…With the advance of the oxidation front into the metallic substrate, the proportion of annihilated vacancies decreased and their condensation into voids was more and more privileged. Since the only microstructure modification observed in the sub-scale with time was the increase of the carbides density (intragranular and intergranular) [15], it could mean also that carbides, more exactly the interfaces between carbides and the matrix, acted as high-diffusion paths and preferential condensation sites for vacancies. This observation was in good agreement with other observations made by Caplan et al [37] on the favorable influence of carbon in pore formation in pure nickel.…”

“…This observation was in good agreement with other observations made by Caplan et al [37] on the favorable influence of carbon in pore formation in pure nickel. The fact that, simultaneously, the depth of the carburized zone [15] and the internal oxidation zone thickness over the inner Fe-Cr rich spinel oxide layer thickness ratio increased with time ( Table 2) suggest strongly that carbides influences the depth of vacancy injection. A scenario of the formation of the inner Fe-Cr rich spinel oxide layer and the internal oxidation zone could be as following: at any time, iron diffused first from the Fe rich metallic substrate near the inner Fe-Cr rich spinel oxide/metal interface and secondly from a deeper zone near the carbide/metallic substrate interface.…”

“…In the first case, the flux of iron vacancies coming from the metal/oxide interface caused the formation of ''available space'' at the oxide/ metal interface by their condensation and thus participated to the growth of the inner Fe-Cr rich spinel oxide layer. In the second case, the formation of chromium rich carbides in the metallic substrate (originally present in T91 or mainly formed because of the carburization process [15]) induced an increase of iron activity around them and created a metal/carbide interfaces which could offer high diffusion paths for outwards iron diffusion. The flux of iron vacancies arriving around chromium rich carbides, then, condensed.…”

“…28, one expects as well that Mo did not oxidize in the inner spinel oxide layer. Nevertheless, because it is bigger and probably less mobile than Ni it staid mainly embedded in the inner oxide or precipitated within carbides [15]. On the other hand, Si was likely present in its oxidized form Si 2?…”

“…From these observations, a model of duplex oxide layer formation inspired by models already proposed in the past for 9 wt% Cr steels in CO 2 [8][9][10][11] and from a more recent model developed by Martinelli et al [12][13][14] for corrosion in oxygen saturated liquid Pb-Bi is proposed. The second article (part II) [15] focuses on the carburization phenomenon of T91. From a detailed study of the carbon content and its evolution over time, a mechanism of carburization, in good agreement with the proposed void-induced duplex oxide layer formation, is suggested.…”