Composition Effects on TBC/Silicate Melt (CMAS) Interaction Dynamics

Thermal barrier coatings (TBC) employed in the hot section of modern gas turbine engines are susceptible to degradation by molten deposits formed when siliceous debris is ingested into the engine [1]. These deposits, generically known as CMAS, may contain SiO₂, CaO, MgO, AlO_1.5, FeO_x, and minor fractions of other oxides; the compositions can vary widely based on the geological source and engine service history [2]. Regardless of the specific deposit composition, the melt wicks into the open features in the TBC microstructure and solidifies under the imposed thermal gradient, leading to stiffening of the coating material and, upon cycling, elevated thermal stresses in the infiltrated region may produce cracking and/or spallation.  Thermochemical attack may compound the damage, especially if the CMAS interacts with the thermally grown aluminum oxide that protects the alloy from oxidation.  

These degradation mechanisms can be mitigated, in part, with coatings designed to minimize the depth of melt penetration for a given exposure condition.  In this context, the most effective coatings react readily with the deposit, converting the penetrating melt to a combination of stable crystalline phases that block the open porosity near the coating surface. The desired behavior is most clearly demonstrated by rare earth (RE) zirconate TBCs, notably Gd₂Zr₂O₇, which react with CMAS to form a combination of fluorite and apatite (~Ca₂Gd₈Si₆O₂₆)[3]. Work with one model CMAS composition (33CaO-9MgO-13AlO_1.5-45SiO₂ in mol%, or Ca₃₃Mg₉Al₁₃Si₄₅) has shown that related RE zirconate and hafnate phases also form fluorite and apatite, among other reaction products. The efficacy of the crystallization reactions, however, varies with the RE cation radius, with the larger RE’s (i.e. La) acting as more potent apatite formers [4]. Efforts to design new TBC compositions would benefit from improved understanding of the role of the silicate deposit composition on the reaction processes. Elucidating these effects is the focus of this study.

This work examined the effect of melt composition on the phase equilibria between the melts and candidate TBC materials such yttria stabilized zirconia (YSZ) and yttrium zirconate (Y₄Zr₃O₁₂, YZO). The silicate melts were selected to span a relevant range of compositions while leveraging existing knowledge about the intrinsic melting and crystallization behaviors [5]. Initial experiments were performed in the simple, well-characterized CaO-AlO_1.5-SiO₂ (CAS) system with additional work in more complex silicate melts. Pre-reacted crystalline silicate powders were mixed with thermal barrier oxide (TBO) particles and annealed to identify the equilibrium crystalline phases and determine how the solubility limit of the TBO constituents in the melt changes with the melt composition.

The results reveal that the melt composition plays a significant role in determining the crystallization product identity and volume fraction. Figure 1 highlights the effect of systematic changes in melt composition on the phase equilibria with YZO at 1300ºC. Figure 1(a,b) show melt compositions in the CAS system.  Both systems form apatite while the primary ZrO₂-bearing phase changes from zircon (ZrSiO₄) to fluorite with a moderate shift in the SiO₂ concentration. Meanwhile Ca₃₃Mg₉Al₁₃Si₄₅, the CMAS composition used in many past laboratory studies and shown in Figure 1(c), forms only fluorite. The latter behavior is related to the relatively high solubility of YO_1.5 in both the melt and precipitated fluorite so the melt is not sufficiently saturated in YO_1.5 to promote apatite crystallization.

The findings shed new light on observations from ex-service TBCs that exhibit a range of deposit compositions and corresponding crystallization products and offer guidance for improved laboratory tests. The conclusions underscore the importance of considering a variety of factors related to melt chemistry when assessing the CMAS degradation of existing TBC materials and designing novel TBCs for improved CMAS resistance.

Figure 1: Reaction between Y₄Zr₃O₁₂ TBC material and selected silicate melt compositions (initially 25mol% YZO in the melt) produced a variety of crystallization products after 50h/1300ºC equilibration. Compositions (a) and (b) represent changes in the Ca:Si ratio but similar AlO_1.5 concentration in the CAS system; (a) has sufficient SiO₂ activity to form zircon (ZS) while (b) forms fluorite (F) as the only ZrO₂-containing reaction product. Melt composition (c) has lower SiO₂ activity and higher YO_1.5 solubility; as such, no apatite (Ap) crystallization is observed even with the higher CaO concentration than (a) or (b).  

[1] M.P. Borom, C.A. Johnson, L.A. Peluso, Surf. Coat. Technol, 86 (1996) 116-126.

[2] C.G. Levi, J.W. Hutchinson, M.-H. Vidal-Sétif, C.A. Johnson, MRS Bulletin, 37 (2012) 932-941.

[3] S. Krämer, J. Yang, C.G. Levi, J.Am. Ceram. Soc., 91 (2008) 576-583.

[4] D.L. Poerschke, C.G. Levi, J.Eur. Ceram. Soc., 35 (2015) 681-691.

[5] E.M. Zaleski, C. Ensslen, C.G. Levi, Submitted (2014)