Diffusional Analysis for Growth of the Monosulfide Scale on Cobalt‐Iron Alloys

Abstract

A ternary diffusion model is invoked to interpret growth of the (CoFe)S scale on Co-Fe alloys containing greater than 40 atomic percent (a/o) Fe in H2S-H2 atmospheres, 3.4 • 10 -7 ~ Ps2 -~ 1 • 10 -5 atm, at temperatures 533 ~ 980~ Concentration profiles in the solid phases were determined from electron probe microanalyses; the iron concentration across a scale continuously decreased from its highest value at the scale/alloy interface. Computations demonstrated that the relative self-diffusivity of cobalt to iron in (CoFe)S was dependent on concentration increasing asymptotically from a value of 1.3 in essentially pure FeS to a value of 2.4 at 60 a/o FeS. A numerical computing method capable of determining sulfidation rates of a binary alloy from diffusional and thermodynamic data was used to demonstrate that calculated metal composition profiles in the (CoFe)S scales and values of the parabolic sulfidation rate constants were in good agreement with the experimental determinations.

Appendix: Corrosion issues in radioactive waste interim storage facilities * *Part of this appendix is reproduced from Hélie et al. [4], which is copyright (2006) by the American Nuclear Society, La Grange Park, Illinois.

“…in fact, the steady state hypothesis of Wagner's model is no longer valid if the microstructure or the equilibrium conditions at the interfaces evolve with time. In fact, parabolic oxidation models for a monophased binary alloy that forms a solid solution of oxide [12,13] mainly use diffusion and concentration data measured in monocrystals. A.11) already revealed a gap compared with the classical parabolic model after only 300 hours of test.…”

Computer Simulation of the Diffusional Growth of Ternary Oxide Scales on Binary Alloys

“…The four empirical constants Fco, Gco, FNi, and GNi are --2.8, --1.3, --2.9, and --1.4, respectively, independent of tempera= ture. The term, n, in the above equations is n = ncoo + (6 --ncoo)~ [17] where ncoo is the value for pure CoO. This latter term is obtained by differentiating Eq.…”

“…The input data, therefore, are the self-diffusivity of A and p : DB/DA for the oxide scale, the interdiffusion coefficient ~AB for the alloy, and the free energy change hG for the displacement reaction between the pure binary oxides and the alloy components. We have presented the flow chart for this computer simulation in an earlier paper (17) based upon the modified predictor-eorrector Hamming method containing automatic changes of step width (26). The appropriate equations for the computations to describe (CoFe)O and (CoNi)O scale growth on the Co-Fe and Co-Ni alloys, respectively, are presented in the Appendix.…”

“…Several investigators (4)(5)(6)(7)(8)(9)(10)(11)(12)(13)(14)(15)(16)(17) have studied the suitability of Wagner's equations for describing scale growth on binary alloys by measuring cation distributions in the scales and comparing these results with theoretical profiles. In order to minimize computing time for evaluation of the nonlinear differential equations, alternate approximate expressions (14,15) have been developed for calculating the reaction rate and cation profile in a scale.…”

“…(16) to clarify the effect of each kinetic and thermodynamic parameter on the diffusion-controlled growth of these types of scales. More recently, Young et al (17) have shown that a complete computer simulation of Wagner's equations capable of determining oxidation rates of the binary alloy using independently obtained diffusional and thermodynamic data provides a more rigorous description of binary alloy oxidation than suggested by the analyses in the earlier investigations.…”