Hot-corrosion behavior of Ti₃SiC₂ in a eutectic mixture of LiCl–KCl salts in air

Abstract

High-corrosion-resistant materials are required for the pyrochemical reprocessing of spent fuels. Towards this end, the hot-corrosion behavior of polycrystalline Ti₃SiC₂ in a eutectic mixture of LiCl–KCl melts was investigated for 100 h at 550 °C, 650 °C, and 750 °C in air. The results indicate that Ti₃SiC₂ exhibits excellent corrosion resistance in LiCl–KCl at 550 °C. However, it undergoes hot-corrosion at 650 °C and 750 °C, with the mass gains per unit area being approximately 0.4 and 1.6 mg cm⁻², respectively. The main corrosion products at 650 °C and 750 °C were found to be TiO₂ and Li₂TiO₃, respectively. It was surmised that Ti₃SiC₂ exhibits good corrosion resistance at 550 °C and is partly corrosion resistant because the oxide products form a stable protective film on the specimen surface at 650 °C and 750 °C. Finally, the microstructures and phase compositions of the corroded samples were investigated using scanning electron microscopy/energy-dispersive X-ray spectroscopy and X-ray diffraction analysis.

---

1. Introduction

Pyrochemical reprocessing based on electrolysis in molten salts as the reaction media is being used as a method of dealing with high-burn-up, short-cooled, high-plutonium-concentration spent fuels.^1–5 With this in mind, eutectic mixtures of molten LiCl–KCl salts are being considering for use as the process medium for the electrowinning of spent fuels.^6,7 For instance, they are used extensively for the recovering and separating of thorium, uranium, plutonium and other actinides from fission products.^8–10 However, because the use of molten chlorides involves high temperatures, pyrochemical reprocessing requires that high-corrosion-resistant materials be employed for the process. A number of studies have investigated the corrosion of structural materials in molten chlorides, focusing on corrosion under inert conditions or in the presence of oxygen and on the underlying mechanisms.^11–13 The two types of materials that have been studied most widely are alloys and ceramics.^14–16

Ceramics exhibit many properties that make them suitable as high-temperature replacements for alloys. These include high melting points, good high-temperature strengths, low densities, and increased resistance to aggressive environments.^17,18 Ti₃SiC₂, which has a layered crystal structure, possesses properties of both metals and ceramics.¹⁹ Thus, it is an ideal structural material for high-temperature applications. For instance, Ti₃SiC₂ has been studied as a candidate material for lead cooled fast reactor, the results exhibited that the extent of corrosion was minimal.^20,21 So the hot corrosion behavior was studied in the molten salts. Liu et al. showed that Ti₃SiC₂ exhibited better corrosion resistance at 650 °C and undergoes severe hot corrosion in K₂CO₃–Li₂CO₃ melts at 700–850 °C.²² In Na₂SO₄–NaCl melts, Ti₃SiC₂ undergoes significant corrosion at 850 °C when the concentration of Na₂SO₄ is higher than 35 wt%. However, when the concentration of Na₂SO₄ is lower than 25 wt%, the corrosion rate of Ti₃SiC₂ is quite low, because a protective layer of TiO₂ and SiO₂ is formed on its surface.²³ Previous studies have shown that the corrosion of Ti₃SiC₂ is greatly influenced by the temperature, exposure time, the types of molten salts used, and the oxygen concentration of the melt.

In the pyrochemical reprocessing of spent fuels, molten chlorides are commonly used as the process media. However, the corrosion behavior of Ti₃SiC₂ in the mixtures of molten chlorides has not been reported previously. The aim of this study was to elucidate the hot-corrosion behaviors of Ti₃SiC₂ in eutectic mixtures of molten chloride salts. The results of this study should aid in the selection of the appropriate conditions for using this technologically important material.

2. Experimental

The Ti₃SiC₂ samples used in this study were fabricated by an in situ hot pressing/solid–liquid reaction process, which has been described elsewhere.²⁴ The density of Ti₃SiC₂ is 4.47 g cm⁻³, and the porosity of the samples used was less than 2%. Rectangular specimens with dimensions of 10 mm × 5 mm × 5 mm were cut by the electrical discharge method. The surfaces of the specimens were ground and polished with SiC paper of up to 1200# grit. The specimens were cleaned in acetone, ethanol, and distilled water in an ultrasonic bath before the tests.

The mixture of LiCl–KCl powders was prepared using a LiCl–KCl mass ratio of 0.42 : 0.58. In each test, one sample and the mixed LiCl–KCl salts were placed in an Al₂O₃ crucible, which was then put in an electrical box furnace. To remove the residual water in the salts, the LiCl–KCl mixtures were firstly heated to 300 °C for 2 h. The tests were conducted in air at 550 °C, 650 °C, and 750 °C. The specimens were immersed completely in the molten salts during the tests. After the hot corrosion tests, we washed the samples with boiling distilled water to dissolve any remaining LiCl–KCl and other dissolvable salts. Then, the samples were dried in hot air. Each sample was weighed before and after the test; the sensitivity of the balance used was 10⁻⁵ g. The microstructures of the corroded samples were analyzed using a scanning electron microscopy (SEM) system equipped with an energy-dispersive X-ray spectroscopy (EDS) attachment. The phase compositions of the corrosion layers of the samples were determined by X-ray diffraction (XRD) analysis.

3. Results and discussion

3.1. Kinetics of hot corrosion

Fig. 1 shows the change in the mass per unit area of Ti₃SiC₂ as a function of the exposure time during hot corrosion in the mixture of LiCl–KCl melts at 550 °C, 650 °C, and 750 °C. When the test sample was corroded in a eutectic salt at 550 °C for 100 h, the change in the mass per unit area increased slightly (0.4 mg cm⁻²). This suggests that Ti₃SiC₂ is not readily corroded at this temperature. When the temperature was raised to 650 °C, the sample underwent a greater degree of corrosion as well as a greater change in mass. However, the final change in the mass per unit area was approximately 1.6 mg cm⁻² after 100 h at 750 °C. These results indicate that Ti₃SiC₂ undergoes hot corrosion at 750 °C. Thus, the corrosion process of Ti₃SiC₂ is affected mainly by the temperature. Fig. 1 also shows that, for all the runs, the change in mass versus time curves exhibit two distinct regions, which have different kinetics. The first region (t < 24 h) corresponds to rapid mass gain, while the second region corresponds to slow mass gain. Thus, we can conclude from the kinetic data that the generated corrosion products have an adverse effect on the corrosion process.

Fig. 1 The change in the mass per unit area as a function of the corrosion time corresponding to the hot corrosion of Ti₃SiC₂ in the mixture of LiCl–KCl melts at 550 °C, 650 °C, and 750 °C.

3.2. Phase compositions and microstructures of the corrosion products

Fig. 2 shows the XRD patterns of the untreated Ti₃SiC₂ sample and of the samples subjected to hot corrosion in the LiCl–KCl melts for 100 h at 550 °C, 650 °C, and 750 °C. The diffraction peaks in Fig. 2a could be assigned to Ti₃SiC₂ (JCPDS card no. 74-0310). In the case of the sample corroded by the LiCl–KCl melts at 550 °C, no new peak was observed in the XRD pattern (see Fig. 2b). This suggests that the sample was barely corroded by the LiCl–KCl melts, corresponding that there was almost no change in its mass. In Fig. 2c, which shows the XRD pattern of the sample corroded at 650 °C, peaks related to TiO₂ and Li₂TiO₃ can be seen. This means that this sample did undergo corrosion. And the main corrosion product in the case of the sample corroded at 650 °C was TiO₂, as can be seen from the XRD pattern in Fig. 2c. After corrosion at 750 °C, the corrosion product TiO₂ was replaced by Li₂TiO₃ (see Fig. 2d), which means that the sample continued to undergo corrosion and the main corrosion product changed.

Fig. 2 XRD patterns of the samples corroded in LiCl–KCl melts for 100 h at different temperatures: (a) uncorroded, (b) 550 °C, (c) 650 °C, and (d) 750 °C.

Fig. 3 shows the surface microstructures of the Ti₃SiC₂ samples subjected to hot corrosion in the mixed LiCl–KCl melts for 100 h at 550 °C, 650 °C, and 750 °C. The corresponding EDS spectra are shown in Fig. 4. For comparison, an SEM image of the untreated Ti₃SiC₂ sample is shown in Fig. 3a. The untreated sample exhibits a smooth surface and contains only three elements, Ti, Si, and C, as can be seen from Fig. 4a. Further, as shown in Fig. 3b, the Ti₃SiC₂ sample corroded at 550 °C still has a smooth surface similar to that of the untreated one, and contains elemental O in a very small amount, as indicated by its EDS spectrum. These results suggest that Ti₃SiC₂ possess good corrosion resistance at 550 °C. On the other hand, as shown in Fig. 3c, for the Ti₃SiC₂ sample treated at 650 °C, the smooth surface has been replaced by a layer of particles. The concentration of elemental O has also increased notably (see Fig. 4c). This suggests that the surface of the Ti₃SiC₂ sample has undergone significant corrosion. When combined with the results of the XRD analysis, this result suggests that the main corrosion product on the surface of this Ti₃SiC₂ sample is TiO₂. Finally, in the case of the sample hot corroded at 750 °C, a rough corrosion layer is noticed on the sample surface (see Fig. 3d). The results of the XRD and EDS analyses show that the main corrosion product was now Li₂TiO₃. Moreover, pores and cracks were barely found on the surfaces of the corrosion products. This indicated that Ti₃SiC₂ would possess high corrosion resistant in LiCl–KCl melts because pores and cracks should accelerate the corrosion process.

Fig. 3 SEM micrographs showing the surface morphologies of the Ti₃SiC₂ samples corroded in the LiCl–KCl melts for 100 h at different temperatures: (a) uncorroded, (b) 550 °C, (c) 650 °C, and (d) 750 °C.

Fig. 4 EDS spectra of the Ti₃SiC₂ samples corroded in the LiCl–KCl melts for 100 h at different temperatures: (a) uncorroded, (b) 550 °C, (c) 650 °C, and (d) 750 °C.

Fig. 5 shows the cross-sectional microstructures of the Ti₃SiC₂ samples corroded in the LiCl–KCl melts at 550 °C, 650 °C, and 750 °C. The SEM images revealed that the cross-sections of all the samples had pits. The mechanism of formation of these pits in the substrate might be related to the microstructural characteristics of ternary-layered Ti₃SiC₂. As is the case with most ceramics, a number of pores and large grain boundaries existed in the bulk Ti₃SiC₂ (the porosity of the Ti₃SiC₂ samples used in this study was approximately 2%). It is likely that these pores and large boundaries served as short paths for the inward diffusion of the molten chlorides. In these areas, the corrosion rate was high. No oxide layer is observed in Fig. 5a, which means that the specimen treated at 550 °C for 100 h was barely corroded. The SEM image in Fig. 5b shows a cross-section with a two-layered microstructure. The thickness of the scale-like oxide layer was approximately 2 μm after corrosion for 100 h at 650 °C. Fig. 5c and d show cross-sections of the specimens corroded at 750 °C for 48 h and 100 h, respectively. The thickness of the oxide layer reached up to 6 μm after 48 h at 750 °C, which suggests that the temperature is a critical factor influencing the corrosion process. When the corrosion time was increased to 100 h, the thickness of the oxide layer increased to approximately 11 μm. Thus, the corrosion time also has an effect on the corrosion process of Ti₃SiC₂. Moreover, pores and cracks were barely observed in the oxide layer in all the cross-sections. Thus, it is likely that the compact oxide layer retards the corrosion of Ti₃SiC₂.

Fig. 5 Cross-sections of Ti₃SiC₂ samples corroded in LiCl–KCl melts at various temperatures: (a) 550 °C for 100 h, (b) 650 °C for 100 h, (c) 750 °C for 48 h, and (d) 750 °C for 100 h.

3.3. Mechanism of corrosion of Ti₃SiC₂

Sato et al. proved that the presence of oxygen accelerates the reaction between non-oxide ceramics and a eutectic melt mixture.²⁵ Liu et al. studied the corrosion behavior of Ti₃SiC₂ in Li₂CO₃ and K₂CO₃ melts; the results suggested that Ti₃SiC₂ does not react with Li₂CO₃ and K₂CO₃ directly.²² Similarly, the increase in the mass of Ti₃SiC₂ ceramics in the eutectic LiCl–KCl mixture is caused by the oxidation of Ti₃SiC₂ owing to the presence of dissolved oxygen in the melt. The corresponding reaction is as follows:²⁶

Ti₃SiC₂(s) + 5O₂(g) → 3TiO₂(s) + SiO₂(s) + 2CO(g) (1)

It is worth noting that the results of the EDS analysis (Fig. 4) revealed that four elements, Ti, O, Si, and Li, were present in all the samples. The reason a crystalline Si-containing phase was not detected by XRD analysis (Fig. 2) is that either the SiO₂ content in the corrosion films was too low or SiO₂ does not exist in the films in crystalline form. A similar phenomenon was observed in Ti₃SiC₂ oxidized at 1100–1300 °C, in that Si was detected by EDS but no reflections related to SiO₂ were noticed in the XRD spectrum.²⁶

In the initial stage of the corrosion process, Ti₃SiC₂ is oxidized by the oxygen dissolved in the LiCl–KCl melts according to Reaction (1). With an increase in the corrosion time and temperature, the oxygen diffuses through the molten salts to the test samples and reacts with the eutectic melt and the newly formed TiO₂ and SiO₂.²³ In the (Li, K) eutectic mixture, Li-containing corrosion products are formed readily. For example, the corrosion products are Li₂TiO₃ and Li₂SiO₃ after Ti₃SiC₂ is corroded in the eutectic mixture of (Li, K) carbonates. Therefore, the corrosion products are formed primarily according to the following reactions:^27,28

4LiCl(l) + O₂(g) + 2TiO₂(g) → 2Li₂TiO₃(g) + 2Cl₂(g) (ΔG = 281 kJ mol⁻¹ at 1023 K) (2)

4LiCl(l) + O₂(g) + 2SiO₂(g) → 2Li₂SiO₃(g) + 2Cl₂(g) (ΔG = 261 kJ mol⁻¹ at 1023 K) (3)

However, both Reactions (2) and (3) are not favored thermodynamically. Considering that most of the oxygen dissolved in the melt is consumed by Reaction (1), the oxygen partial pressure in the melt is very low, which is adverse for Reactions (2) and (3).²³ This is the reason Ti₃SiC₂ exhibits excellent corrosion resistance in the eutectic LiCl–KCl melts.

Two main rate-determining processes occur during hot corrosion.^22,29 These are the transport of the oxidant from the gas phase through the molten salts as well as the corrosion products to the active oxidation site and the dissolution/reprecipitation of the oxide within the corrosion product layer owing to the existence of a solubility gradient. In this study, the distribution of the salts within the pores of the Ti₃SiC₂ substrate was a key factor that determines the reaction rate in the early stages of the hot-corrosion process. Thus, it was concluded that fluxing is very likely a rate-controlling process. The SEM images showed that the corrosion layer has a compact structure. While the actual hot-corrosion rate increased considerably with an increase in the corrosion time, the oxidant transport was controlling. The two rate-controlling processes correspond with there are two different kinetics (Fig. 1). This mechanism also demonstrates that the corrosion layer retards the corrosion process.

4. Conclusion

In this study, the hot-corrosion behavior of polycrystalline Ti₃SiC₂ was studied in a eutectic mixture of LiCl–KCl melts for 100 h at 550 °C, 650 °C, and 750 °C. The microstructures and phase compositions of the corroded samples were investigated by SEM/EDS and XRD analysis. The results indicated that Ti₃SiC₂ exhibits excellent corrosion resistance in LiCl–KCl at 550 °C. However, it undergoes significant corrosion in the mixture of the LiCl–KCl melts when the temperature is raised up to 650 °C and 750 °C, while exhibiting mass gains per unit area of approximately 0.4 and 1.6 mg cm⁻², respectively. The main corrosion products at 650 °C and 750 °C were determined to be TiO₂ and Li₂TiO₃, respectively. It was concluded that Ti₃SiC₂ exhibits good corrosion resistant at 550 °C and partly corrosion resistant because the oxide products form a stable protective film on the specimen surface at 650 °C and 750 °C. Therefore, Ti₃SiC₂ could be a potential material for use during the pyrochemical reprocessing of spent fuels.

Acknowledgements

This work is supported by the National Natural Science Foundation of China (no. 21301163), the Major Research Plan of the National Natural Science Foundation of China (no. 91426302), the Institute of Nuclear Physics and Chemistry.

References

1. T. Suzuki, M. Tanaka and S. Koyama, Prog. Nucl. Energy, 2011, 53, 969.
2. S. Noh, J. Kang, D. Kwak, P. Fischer and B. Han, Energy, 2014, 68, 751.
3. Y. Sakamura and M. Akagi, Nucl. Technol., 2012, 179, 220.
4. A. Venkatesh, S. Ghosh, S. Vandarkuzhali, B. P. Reddy, K. Nagarajan and P. R. V. Rao, Nucl. Technol., 2013, 182, 98.
5. H. Choi, Ann. Nucl. Energy, 2011, 38, 2338.
6. S. Delpech, Pure Appl. Chem., 2013, 85, 71.
7. T. J. Kim, Y. Jung, J. B. Shim, S. H. Kim, S. Paek, K. R. Kim, D. H. Ahn and H. Lee, J. Radioanal. Nucl. Chem., 2011, 287, 347.
8. Y. L. Liu, Y. D. Yan, W. Han, M. L. Zhang, L. Y. Yuan, K. Liu, G. A. Ye, H. He, Z. F. Chai and W. Q. Shi, RSC Adv., 2013, 3, 23539.
9. T. Murakami, T. Kato, A. Rodrigues, M. Ougier, M. Iizuka, T. Koyama and J. P. Glatz, J. Nucl. Mater., 2014, 452, 517.
10. E. Mendes, R. Malmbeck, C. Nourry, P. Soucek and J. P. Glatz, J. Nucl. Mater., 2012, 420, 424.
11. A. R. Shankar, K. Thyagarajan and U. K. Mudali, Corrosion, 2013, 69, 655.
12. M. Takeuchi, Y. Arai, T. Kase and Y. Nakajima, J. Nucl. Mater., 2013, 432, 35.
13. J. Sure, A. R. Shankar, S. Ramya, C. Mallika and U. K. Mudali, Carbon, 2014, 67, 643.
14. T. J. Kim, G. Y. Kim, D. Yoon, D. H. Ahn and S. Paek, J. Radioanal. Nucl. Chem., 2013, 295, 1855.
15. A. R. Shankar, A. Kanagasundar and U. K. Mudali, Corrosion, 2013, 69, 48.
16. L. Wang, B. Li, M. Shen, S. Y. Li and J. G. Yu, Int. J. Miner., Metall. Mater., 2012, 19, 930.
17. H. M. Lee and S. Lai, J. Electrochem. Soc., 1976, 123, C244.
18. Y. Fu, X. J. Zhai, B. Bai, X. S. Zhang and Z. W. Wang, J. Rare Earths, 2005, 23, 89.
19. H. B. Zhang, Y. W. Bao and Y. C. Zhou, J. Mater. Sci. Technol., 2009, 25, 1.
20. M. Utili, M. Agostini, G. Coccoluto and E. Lorenzini, Nucl. Eng. Des., 2011, 241, 1295.
21. L. A. Barnes, N. L. D. Rago and L. Leibowitz, J. Nucl. Mater., 2008, 373, 424.
22. G. M. Liu, M. S. Li, Y. C. Zhou and Y. M. Zhang, J. Eur. Ceram. Soc., 2003, 23, 1957.
23. G. M. Liu, M. Li, Y. C. Zhou and Y. M. Zhang, J. Eur. Ceram. Soc., 2005, 25, 1033.
24. Y. C. Zhou, Z. M. Sun, S. Q. Chen and Y. Zhang, Mater. Res. Innovations, 1998, 2, 142.
25. T. Sato, Y. Kanno and M. Shimada, Int. J. High Technol. Ceram., 1986, 2, 279.
26. Z. M. Sun, Y. C. Zhou and M. S. Li, Corros. Sci., 2001, 43, 1095.
27. M. Hara and Y. Kitagawa, Oxid. Met., 1999, 52, 77.
28. N. S. Jacobson, Oxid. Met., 1989, 31, 91.
29. K. N. Lee and D. A. Shores, J. Electrochem. Soc., 1990, 137, 859.

---