Two techniques have been proposed to reduce thermal-expansion mismatches between (a) substrates made of silicon, silicon-based ceramics, and silicon-based-ceramic composite materials and (b) surface coats that protect the substrates against chemical attack in oxidizing and/or corrosive environments. Typical substrate materials include SiC/ Si composites. A typical coating material is mullite (Al6Si2O13), which can protect silicon-based substrates against water-free oxidizing and corrosive environments. Mullite can also be applied as intermediate coating layers to relax stresses and enhance the adhesion of overlying protective layers of zirconia (ZrO2) or nonstoichiometric anorthite (stoichiometric composition CaAl2Si2O8). The coefficients of thermal expansion (CTEs) of mullite and of some other typical oxide coating materials are greater than the CTEs of silicon-based substrates and, as a result, the coatings tend to crack through their thicknesses. The cracks become pathways for the entry of the chemical species from which one seeks to protect the substrates.
In one proposed technique, one or more lower-CTE phase(s) would be incorporated into a mullite coating to reduce the CTE of the coating for a better CTE match with the substrate. Suitable lower-CTE compounds include cordierite (2MgO.2Al2O3.5SiO2) and fused silica (see Figure 1). Mullite, cordierite, and fused silica would be chemically compatible with the substrate, with each other, and with typical other oxide coating materials. A composite coating of mullite with cordierite and/or fused silica could be applied by plasma spraying or by a wet chemical process.
The CTE of a polycrystalline material like a mullite/cordierite/fused silica composite can be approximated by a rule of mixtures :ac ≈ αi Vi here αi is the CTE of the composite, αi is the CTE of the ith constituent, and Vi is the volume fraction of the ith constituent. Initially, the proportions of cordierite and/or fused silica could be chosen to obtain a desired value of ac according to this rule. However, because of the complexity of the phase composition of the mullite/cordierite/fused silica system, a process of trial and error would likely be necessary to establish the optimum composition.
In the second proposed technique, zircon (ZrSiO4) would be applied as an intermediate layer between a substrate and an overlying protective coating. Optionally, if a dense, crack-free zircon coating could be produced, then it could be used, instead of mullite, as a protective coating, provided that there is no water vapor in the environment. In comparison with mullite, zircon has a CTE closer to the CTEs of the typical substrate constituents SiC and Si. If resistance to water is needed, then a protective coating of zirconia (ZrO2) or of various silicates could be applied over the zircon layer. Zircon would be chemically compatible with both the protective coating and the thin layer of SiO2 that typically forms on the surface of an Si-based substrate.
Like a mullite/cordierite/fused silica composite coating, a zircon coating could be applied by plasma spraying or by a wet chemical process. Plasma spraying could be complicated by the fact that zircon melts and freezes incongruently, forming cubic zirconia first upon cooling from the liquid phase. It might be necessary to add Y2O3 or CaO to the starting composition to stabilize the cubic phase and prevent volumetric changes while allowing the conversion to zircon to take place. Post-spray annealing might be necessary to help the zircon coating reach equilibrium and enhance its stability.
The CTE of zircon is slightly less than that of SiC, though greater than that of Si(see Figure 2). In the case of zircon plasma-sprayed on SiC, the slight difference between the CTEs results in a small compressive stress in the zircon. Inasmuch as the compressive strength of zircon exceeds its tensile strength, this small compressive stress could be advantageous in that it might offset small residual local tensile stresses and thereby help to prevent cracking. As in the first technique, one could incorporate lower-thermal-expansion phases like cordierite and/or fused silica to obtain a lower overall CTE; for example, to obtain a greater compressive stress in a coating on an SiC substrate or to obtain a closer CTE match with an Si substrate.
This work was done by Hongyu Wang of General Electric Co. for Lewis Research Center.
Inquiries concerning rights for the commercial use of this invention should be addressed to
NASA Lewis Research Center
Commercial Technology Office
Attn: Tech Brief Patent Status
Mail Stop 7 - 3
21000 Brookpark Road
Refer to LEW-16393.