This article is one of eight papers to be presented exclusively on the web as part of the January 2000 JOM-e the electronic supplement to JOM.
	The following article appears as part of JOM-e, 52 (1) (2000), http://www.tms.org/pubs/journals/JOM/0001/Seal/Seal-0001.html JOM is a publication of The Minerals, Metals & Materials Society

Functional Coatings: Research Summary

Ceria-Based High-Temperature Coatings for Oxidation Prevention

S. Seal, S.K. Roy, S.K. Bose, and S.C. Kuiry

TABLE OF CONTENTS
INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS ACKNOWLEDGEMENTS References

The beneficial effects of a cerium-oxide coating on the isothermal cyclic oxidation behavior of 316, 321, 304, and 347 austenitic grade steels were studied. The best oxidation resistance was noted for 321, followed by 316, 347, and 304. The internal oxidation of silicon, which acted as a pegging action for better scale adherence, was observed. A fine-grained oxide structure was achieved in the presence of the coating for better scale plasticity. The oxidation mechanism was changed from an outward action migration to a predominately inward oxygen transport, leading to the condensation of vacancies and the prevention of voids, which suggests a compact and uniform scale.

INTRODUCTION

Austenitic-grade stainless steels find wide usage in applications, such as superheaters, reheater tubes, vanes and turbine blades, and various components in fuel-conversion units, that are subject to thermal fluctuations under normal operating conditions and, hence, need protection from high-temperature degradation. As a result, researchers have directed their attention toward the use of composites and ceramics for improving the lifetime and increasing the service temperature of materials used in high-temperature oxidizing environments. Numerous articles have been written concerning the beneficial effect of rare-earth elements (e.g., Y, Ce, Se, or La) or their oxides (either in the form of oxide-dispersion alloy additions or by superficial coating of these oxides to both chromia- or alumina-forming alloys), which have exhibited improved scale adherence to the metal/alloy substrates when exposed to high-temperature environments.^1-7 Several oxidation mechanisms have been put forth to explain these beneficial effects for high-temperature alloys.^8-16 Both chromia and alumina are known to have protective Cr₂O₃ and Al₂O₃ oxide scales to also prevent high-temperature degradation of the underneath layer.

Based on observations in studies on the effect of ceria on the high-temperature oxidation behavior of austenite-grade steels^6,17-22 and a detailed review by J. Jedlinski, we proposed the following factors to account for the effectiveness of reactive elements in improving scale-spallation resistance:

• An overall reduction in oxidation rate.
• Better cohesion between the scale and the substrate.
• The early formation of a Cr₂O₃/Al₂O₃ scale.
• Improved scale plasticity.

The application of cerium-oxide coatings is considered to be the most practical method for ensuring a high local concentration of the active element within the oxide scale, thus protecting the material at elevated temperatures. Such coatings offer potential advantages over alloy additions with respect to their low cost, relative ease of application, and ability to avoid problems related to alloy manufacturing during fabrication.

EXPERIMENTAL PROCEDURES

Typical bulk compositions of AISI 316, 304, 321, and 347 steels are presented in Table I. Test specimens 40 mm × 40 mm × 1 mm were cold rolled and annealed in an argon atmosphere at 1,223 K. Surface preparation was done using 600 SiC grit paper, followed by cleaning in acetone.

It has been reported in a number of articles that the high-temperature oxidation behavior of pure metals^2,23-25 as well as Fe-Cr binary and ternary Fe-Cr-Ni alloys^26-31 is strongly dependent on the surface-preparation procedures, such as polishing, electroplating, etching, short pining, etc. Each procedure has a definite impact on the oxidation rate of the alloys.^23,27 A specimen abraded mechanically facilitates early establishment of a protective Cr-rich spinel layer, essentially acting as a diffusion barrier and resulting in a reduced oxidation rate. Moreover, this type of surface is more conducive to ceria coating due to more available nucleation sites.^20-22

The steel specimens were coated in a CeO₂ slurry (average particle size of 50 mm) to a final coating thickness of 2-3 mm. The oxidation experiments were conducted in dry air in a vertically placed quartz tube at Po₂ = 21.27 kPa. A Sartorious electronic microbalance (MP8) was used to record the change in mass of each sample with an accuracy of +0.01 mg. After each experimental run, the oxidized sample was allowed to cool to room temperature inside the reactor itself.

The oxidized scales (with or without coating) formed on the sample were subsequently analyzed using scanning electron microscopy (SEM), energy dispersive x-ray analysis (EDAX), x-ray diffraction (XRD), and x-ray photoelectron spectroscopy. These analytical techniques are useful in understanding the oxidation process in terms of scale morphology and chemistry in the presence or absence of ceria coating.

Figure 1. Typical kinetic data for the isothermal oxidation of 321 at 1,273 K in dry air under different surface preparations after nonisothermal exposure at a heating rate of 1,423 K.

RESULTS

The oxidation kinetics data for all AISI grade stainless steels, with or without ceria coating, under isothermal/cyclic conditions were plotted to monitor the oxidation rate. The mechanically polished samples showed improved oxidation resistance compared to the electropolished sample, and the improvement is more pronounced in the presence of CeO₂ coating. Such an improvement in the oxidation behavior imparted by the mechanical polishing has been reported previously.³² This improvement is due to the change in defect concentration within the material due to polishing conditions, leading to enhanced diffusion through the distorted lattice structure as well as through the grain boundaries to the alloy surface. This further facilitates an easy and early establishment of a protective external Fe-Ni-doped chromia-rich spinel layer. Table II summarizes the relative mass gain rate for all steels under similar oxidizing conditions.

Typical kinetic data for the isothermal oxidation of mechanically polished 321 AISI steel at 1,273 K in dry air after nonisothermal exposure at 1,423 K (heating rate 6 K/min.) are shown in Figure 1. These data include a parabolic fitting of the isothermal oxidation kinetics and clearly depict a marked improvement in the oxidation resistance of the steel specimen in its simple, mechanically polished condition. The parabolic rate constant, k_p, calculated from Equation 1 for coated samples are approximately 2-3 orders of magnitude lower than that of the corresponding uncoated ones.

The parabolic growth kinetics is expressed by

where Dm is the total mass gain per unit surface area for isothermal holding, k_p is the isothermal parabolic rate constant, and t is the time for isothermal holding.

From the data in Figure 1 and Table II, it is found that the coated steel registered a marked improvement in reducing the rate of scale growth. A comparison of all four grades of steels showed that 321 has superior oxidation resistance performance when coated with cerium oxide. On comparing the oxidation kinetics under identical experimental conditions, all grades of steels under CeO₂ coating follow a general trend in their performance of 321 > 316 > 347 > 304. Note that in the case of 321 steel (Figure 1), subsequent cycles show enhanced improvement in oxidation resistance. This can be attributed to the formation of TiO and FeTiO₃ at the end of the first cycle, resulting in titanium depletion in the alloy, which may subsequently help to increase the interdiffusion of chromium. This result is favorable in the formation of a continuous Cr₂O₃ rich layer. The poor performance exhibited by 304 might be due to the presence of microcracks in the scale; such features are also reported earlier in the literature.³

Figure 2. SEM micrographs of the top oxide-scale surface on (a-upper left) coated 316, (b-upper right) coated 321, (c-lower left) bare 347, and (d-lower right) coated 347 (i-represents substrate, ii-scale).
Figure 3. An SEM micrograph of the top scale surface on ceria-coated-oxidized 347 revealing white cerium oxide particles (from EDX) distributed along the oxide grain boundaries.

The oxidized scales were analyzed using SEM, EDAX, and XRD to investigate the nature of the top-surface morphology and alloy/scale cross section for scale adherence. XRD analysis of the scale showed the presence of simple, complex, and mixed oxides (spinels), including Fe₂O₃, Ni₂Cr₂O₄, FeCr₂O₄, MoO₂, NiFe₂O₄, Fe-Ni-Cr spinels, FeTiO₃, TiO₂, Cr₂O₃, and NiNb₂O₆ for uncoated samples and CeO₂, Ce₂O₃, Cr₂O₃, Fe₂O₃, MnFe_xCr_2-xO₄, NiCr₂O₄, FeCr₂O₄, and Fe₃O₄ for ceria-coated samples. Coated 316 samples showed the presence of Ce₂O₃ and CeO₂, whereas coated 304 and 321 only indicated Ce₂O₃, which is a thermodynamically more stable compound than CeO₂.^16,34,35 The niobium-containing 347 steels showed the formation of NiNb₂O₆-type compounds, but not that of complex carbides (e.g., Nb₃M₃C as predicted by others).^36,37 This suggests that at prevailing oxygen partial pressures (dry air), NbC particles have probably undergone a conversion to Nb₂O₅, which, in turn, has reacted with initially formed NiO, resulting in the formation of NiNb₂O₆-type oxides.

Figure 2 shows the typical top surface morphologies of CeO₂-coated 316 (Figure 2a) and 321 (Figure 2b), bare 347 (Figure 2c), and CeO₂-coated 347 (Figure 2d). Oxide scales are found to spall off in the uncoated samples; whereas the coated samples showed fine-grained, uniform, and well-adherent compact scale formation. Furthermore, Figure 3 represents a backscattered SEM image of CeO₂-coated 347, showing the preferential segregate of CeO₂ particles (as confirmed by EDAX) along the oxide grain boundaries. In all the coated steels, the presence of CeO₂ is confirmed by EDX at the top surface layer, primarily at the outer oxide/air interface. Numerous cracks and voids are found in the uncoated samples.

Alloy cross sections of all four steels were further studied by SEM with electron probe microanalysis to follow the distribution of various elements in the scales. Nevertheless, all ceria-coated samples showed remarkable improvement and a drastic reduction in scale growth as compared to the uncoated steels. A typical SEM cross-section image of bare (Figure 4a) and ceria-coated 316 (Figure 4b) is shown in Figure 4. It should be noted that in the presence of the coating, the scale growth is reduced by 10-20 times that of the uncoated alloy. The layered oxide structures as observed in the uncoated oxidized sample were modified to more refined and compact in the coated samples.

A detailed study using x-ray mapping of the coated sample shows cerium (Figure 5a) present in the outermost layer; whereas chromium is observed at the scale base, suggesting an early establishment of the Cr₂O₃ layer in the presence of ceria. The presence of silicon as an internal oxide at the grain boundary of the substrate alloy is quite evident. A silicon x-ray image of ceria-coated 316 shows silicon internal oxidation (Figure 5b). Similarly, Fe, Cr, and Ni x-ray maps of 304 show that the scale base is chromium-rich, dissolved in iron, but devoid of nickel, while in the case of coated steels, chromium is definitely enriched at the alloy scale interface (Figure 5c), with silicon internal oxidation between the scale and the substrate.^20-22

Figure 4. SEM micrographs of the alloy/scale cross section of (a-left) bare and (b-right) ceria-coated oxidized 316 isothermally heated at 1,273 K, 6 K min-1.
Figure 5. (a-upper left) A cerium x-ray image, (b-upper right) silicon x-ray image of ceria coated 316, and (c-bottom) chromium x-ray image of ceria-coated 304.

In general, all grades of austenitic stainless steels receive protection against high-temperature degradation due to the formation of a compact Cr₂O₃ healing layer, which thickens slowly. The improvement in oxidation resistance in the presence of a ceria coating is also seen from the kinetic curves. The suggested Cr₂O₃ growth mechanism is attributed to either cation diffusion along the high-angle grain boundaries as well as other short-circuit (dislocation) diffusion paths or to anion ingress through the initial oxide grain boundaries. The CeO₂ particles on the alloy surface appear to have acted as inert markers, and the post-oxidation analysis has identified this reactive-oxide particle at the scale/air interface.

The reduction in the rate of oxidation is due to the segregation of Ce³⁺ and Ce⁴⁺ ions at the oxide grain boundaries, causing hindrance to cation migration. This can be due to fine-grained oxide layer formation that has taken place due to heterogeneous nucleation caused by the presence of a reactive-oxide particle. Detailed XPS studies³⁸ of the top surface showed only the presence of the Ce⁺⁴ oxidation state.

At the same time, oxygen availability at the alloy/oxide interface becomes a favorable process for scale growth as a result of a large grain boundary area. Johnson et al.²⁶ suggested that solute ion segregation at the grain boundary is quite likely in this case. This was further supported by the verification during the oxidation of Y-implanted Fe-20Cr-25Ni stainless steel, in which Y ions from particles of Y₂O₃ and YCrO₃ were found to be segregated along the grain boundaries of the scale. The transport mechanism for the scale growth in the presence of a coating completely changed from cation (Cr³⁺) migration to predominately anion (oxygen) ingress and the growth taking place at the alloy/oxide interface.

The other beneficial effects of ceria coating on scale adherence may involve a reduction in compressive stress within the scale and an enhancement in the scale/substrate interface adhesion. These are direct outcomes of the modification of the oxide grain structure. The fine-grained structure is expected to have better creep properties that allow the scale to deform plastically to accommodate generated compressive stresses. Furthermore, the inward transport of oxygen suppresses the possibility of void formation and coalescence and as a result, improves scale adherence to the alloy substrate. The difference in the topographic morphology of the oxide scale in the coated and uncoated steel suggests that Cr₂O₃ should be nucleated heterogeneously at the CeO₂ particles with limited lateral growth. Besides, an internal oxidation of silicon is favored by the presence of coating. These internal oxide stringers create a pegging action for better scale adhesion. In the case of uncoated alloy, the absence of internal oxidation is supported by the scale spallation at some location on the alloy surface. Improved scale adherence in the presence of the coating is also due to high plasticity achieved by the fine grain structure.^{1,2,5,35,40-42}

ACKNOWLEDGEMENTS

The authors thank continued support and useful advice from Indian Institute of Technology, KGP, India; Institute National Polytechnique de Grenoble, France; and Advanced Materials Processing and Analysis Center (AMPAC) and Mechanical, Materials, and Aerospace Engineering Department, University of Central Florida.

References
1. D.P. Whittle and J. Stringer, Phil. Trans. Roy. Soc. London, A295 (1980), pp. 309.
2. D.P. Moon and M.J. Bennett, Mater. Sci Forum, 43 (1989), pp. 269.
3. P. Papaiacovou et al., Oxid. Met., 33 (1990), pp. 19.
4. M.F. Stroosnijder et al., Oxid. Met., 33 (1990), pp. 371.
5. L.V. Ramanathan, Proc. 11th ICMI, Vol. 4 (1990), p. 177.
6. S.K. Mitra, S.K. Roy, and S.K. Bose, Oxid. Met., 34 (1990), pp. 101.
7. M.J. Bennett and D.P. Moon, The Role of Active Elements in the Oxidation Behavior of High Temperature Metals, ed. E. Lang (London: Elsevier, Commission of European Communities, 1989), p. 111.
8. C.G. Giggins et al., Met. Trans., 5 (1975), p. 1685.
9. J.K. Tien and F.S. Pettit, Met. Trans., 3 (1972), p. 1587.
10. W.W. Smeltzer, Oxid. Met., 23 (1985), p. 251.
11. J. Antill and K. Peakhall, J. Iron Steel Inst., 205 (1967), p. 1136.
12. H. Pfeiffer, Werks. Korrs., 8 (1957), p. 574.
13. F. Golightly, F. H. Stott and G.C. Wood, Oxid. Met.,10 (1976), pp. 163.
14. J. Stringer et al., Corros. Sci., 12 (1972), p. 625.
15. H.J. Grabke, Oxid. Met., 23 (1985), p.251.
16. T.N. Rhy-Jones, H.J. Grabke, and P. Schmidt, High Temperature Alloys, Their Exploitable Potential, ed. J.B. Marriot et al. (London: Elsevier, Commission of the European Communities, 1987), p. 97.
17. S.K. Roy, A. Auddya, and S.K. Bose, React. Solids., 6 (1989), p. 301.
18. S.K. Mitra, S.K. Roy, and S.K. Bose, Oxid. Met., 37 (1992), p. 95.
19. S.K. Mitra, S.K. Roy, and S.K. Bose, Proc. Symp. Kinetics of Metall. Processes, ed. A.K. Chakraborty and U.K. Chatterjee (Kharagpur, India: IIT, 1987), p. 199.
20. S.K. Roy et al., J. Mater. Sci. and Lett., 12 (1993), pp. 249.
21. S. Seal, S.K. Roy, and S.K. Bose, Oxid. Met., 41 (1994), pp. 139.
22. S.C. Kuiry et al., Iron and Steel Inst. Jpn., 34 (7) (1994), pp. 599.
23. J. Jedlinski , Corros. Sci., 35 (1993), pp. 863.
24. L.B. Pfeil, U.K. patent 459,848 (1937).
25. J. Stringer, Mater. Sci. Eng., A120 (1989), pp. 129.
26. D.P. Whittle and J. Stringer, Philos. Trans. R. Soc. (London), A205 (1980), p. 309.
27. F.H. Scott and G.C. Wood, Mater. Sci. Eng., 87 (1987), p. 267.
28. M.J. Bennett and G.C. Wood, The Role of Active Elements in the Oxidation Behavior of High Temperature Metals and Alloys, ed. E. Lang (London: Elsevier, 1989), p. 81.
29. D.P. Moon, Mater. Sci. Technol., 5 (1987), p. 754.
30. J. Jedlinski, Solid State Phenomena, 21/22 (1992), p. 335.
31. S.K. Mitra, S.K. Roy, and S.K. Bose, Oxid. Met., 39 (1993), p. 221.
32. D. Caplan, A. Harvey, and Cohen, J. Electrochem. Soc., 108 (1961), p. 134.
33. M. Landkof et al., Corrosion, 41 (1985), p. 344.
34. L. Coudurier, D.W. Hopkins, and I. Wilokomirsky, Fundamentals of Metallurgical Processes (New York: Pergamon, 1978), p. 79.
35. T.N. Rhy- Jones, H.J. Grabke, and H. Kudielka, Corros. Sci., 27 (1987), p. 49.
36. G.C. Allen et al., Oxid. Met., 21 (1984), p. 187.
37. R.C. Lobb and M.J. Bennett, Oxid. Met., 35 (1991), p. 35.
38. S. Seal et al., J. Vac. Sci. and Tech., A17 (4) (Jul/Aug, 1999), p. 1109.
39. W.C. Johnson, Met. Trans., 8A (1970), p. 1413.
40. M. Landkof et al., Corrosion, 41 (1985), p. 344.
41. Y. Saito and T. Maruyama, Proc. MRS Int. Meeting Advanced Materials, ed. M. Dayama, S. Samiya, and R.P.H. Chang (Pittsburgh, PA: MRS, 1989), p. 345.
42. S. Mrowee, J. Jedlinski, and A. Gil, Mater. Sci. Eng., A120 (1989), p.169.

S. Seal is an assistant professor in the Advanced Materials Processing Analysis Center and the Mechanical, Materials and Aerospace Engineering Department at the University of Central Florida. S.K. Roy and Prof. S.K. Bose are with the Indian Institute of Technology. Dr. S.C. Kuiry is with the Mukund Iron and Steel Company.

For more information, contact S. Seal, University of Central Florida, Advanced Materials Processing and Analysis Center and Mechanical, Materials, and Aerospace Engineering, 4000 University Boulevard, Orlando, Florida 32816; (407) 823-5277; fax (407) 823-0208; e-mail sseal@pegasus.cc.ucf.edu.