Stability of interfaces in hybrid EBC/TBC coatings for Si-based ceramics in corrosive environments

Highlights

• Dense, uniform and phase pure mullite-based EBCs were deposited on SiC/SiC CMCs by CVD and overlaid by APS YSZ TBCs.

• The EBCs were graded from stoichiometric mullite at the EBC/CMC interface to alumina-rich mullite at the EBC/TBC interface.

• The coating system performed better than plasma sprayed coatings systems based on similar design with segmentation cracks.

• The segmentation cracks in the TBC propagated into the EBC for the plasma sprayed samples.

• The plasma sprayed mullite/CMC interface was weak while the CVD mullite/CMC interface was strong.

Abstract

SiC/SiC ceramic matrix composites (CMCs) are being used increasingly in the hot-sections of gas turbines, especially for aerospace applications. These CMCs are subject to recession of their surface if exposed to a flow of high-velocity water vapor, and to hot-corrosion when exposed to molten alkali salts. This research involves developing a hybrid system containing an environmental barrier coating (EBC) for protection of the CMC from chemical attack and a thermal barrier coating (TBC) that allows a steep temperature gradient across it to lower the temperature of the CMC for increased lifetimes. The EBC coating is a functionally graded mullite (3Al₂O₃•2SiO₂) deposited by chemical vapor deposition (CVD), and the TBC layer is yttria-stabilized zirconia (YSZ) deposited by air plasma spray (APS). The stability of this system is investigated, which includes the adhesion between the two coating layers and the substrate, the physical and chemical stability of each layer at high temperature, and the performance under severe thermal shock and during exposure to hot corrosion. The effect of vertical cracks in the TBC on the EBC layer below it is also examined.

Introduction

Gas turbines used in aerospace applications have been operating under increasing severe conditions during the course of its evolution. In order for gas turbines to operate under harsh environments, they often require protective coatings over the hot-zone components. The most common protective coatings are environmental barrier coatings (EBCs) and thermal barrier coatings (TBCs).

Advanced ceramics are being used more and more in high-temperature applications due to their high melting points, high hardness, low density, and independence from transition and rare earth elements compare to superalloys. One of the signature trends in the aerospace industry is the increasing operation temperature of the gas turbines used to impart thrust. The fundamental driving force behind this is that higher temperatures lead to increased specific core power, better efficiency, and a reduction in environmentally harmful byproducts [1]. The increase of operation temperature automatically brings higher requirements on the materials and current extreme turbine operation temperatures have exceeded the operation limit of superalloys. In the recent decades, the use of thermal barrier coatings has been a standard practice for alloy-based engine parts and has proven to be an effective strategy to enhance the durability of the engine components. The adoption of TBCs offers a quantum leap in temperature capability equivalent to three decades of advancement in alloy design [2]. They are designed to maintain a steep temperature gradient across them. The TBCs keep the temperature of the substrate materials significantly cooler than the surrounding hot gases. Thus, the role of a TBC is to act as a barrier to heat flow. TBCs have been a standard practice for alloy-based engine parts and have proven to be an effective strategy to enhance the durability of engine components [3].

Replacing superalloys with ceramic materials (typically SiC and Si₃N₄) significantly increases the melting temperature, allowing the turbines to operate at much higher temperatures. While mechanical behavior (cracking) was a major issue with monolithic ceramic materials, this has been largely solved by the introduction of ceramic matrix composites (CMCs). The most popular candidate among them is SiC/SiC composites consisting of SiC fibers embedded in a SiC matrix (referred as “SiC/SiC CMC” hereon). However, mechanical failure is not the only degradation mechanism for SiC in gas turbine combustion atmospheres. In general, two additional chemical degradation mechanisms are common. The first degradation mechanism is recession, which is the loss of the silicon based ceramics due to repeated cycle of oxidation to form silica followed by volatilization due to formation of vapor phase species in the presence of high velocity water vapor moving through the turbine [3]. The second degradation mechanism is hot-corrosion, which is the formation of pits in the silicon-based ceramic due to liquid silicates, formed in the presence of alkali salts which are commonly present in the combustion atmospheres [4]. The presence of an EBC is critical to protect the CMCs from these unwanted degradation modes.

Since EBCs and TBCs improve the performance and lifetimes of the gas turbines by addressing different needs, it is logical to combine them together into a hybrid coating system. NASA's Ultra-Efficient Engine Technology (UEET) Program targeted SiC/SiC CMC materials to be able to perform at 1315 °C for 1000 hours. It was clear that the successful outcome would strictly rely on the performance of EBC-TBC protection layer(s) [5]. A recent review article has pointed out that in the next 10 years, a significant investment will be made on the development of effective EBC/TBC coating systems for SiC/SiC CMCs [6]. Reports of recent studies on EBC/TBC hybrid systems in the open literature have mainly been in the form of patents [7], [8], [9]. Ceramic thermal and environmental barrier coating (TEBC or T/EBC) systems are applied over ceramic composites to achieve improved performance [10]. Some researchers have suggested single-layered coating materials (e.g., Y₂SiO₅) that serve both as environmental and thermal barrier coating (ETBC) [11]. It is clear that having a coating or coating system with effective environmental barrier and thermal barrier properties over ceramic materials for advance turbine applications, is required.

For silicon-based ceramic components, mullite (3Al₂O₃•2SiO₂) is widely used as the EBC material mainly due to its close CTE match and chemical compatibility with SiC [12]. Lee et al. pioneered plasma-sprayed mullite-based EBC systems on silicon-based ceramics [13]. In order to combat various issues associated with this system, such as silica (from the mullite itself) volatilization, durability issues due to presence of alkali metals, adhesion between different layers, and crack formation [13], [14], [15], [16], [17], current state-of-the-art EBC coating systems consist of three layers, all deposited by air-plasma spraying (APS). These are as follows: 1) a layer of silicon bond coat on top of the silicon-based substrate that aids coating adhesion, 2) a mullite or mullite/Ba_1−xSr_xAl₂Si₂O₈ (BSAS) intermediate layer, and 3) a BSAS top-coat. BSAS is a low Si-activity material that is resistant to recession and hot-corrosion. With this 3-layer configuration, two recent independent field tests with thousands of hours of service showed that the silicon layers had undergone significant oxidation [18], [19]. The formation of this thick silica layer was usually the cause for cracking and spallation of the coating. These results show that there are durability issues with even the state-of-the-art EBC systems, and improvements are needed.

Mullite is a promising material of choice for the EBC coating for the following important reasons. 1) It has a good CTE match with SiC over a wide range of temperatures (Fig. 1). 2) It has high melting point (1840 °C) thus providing high temperature physical stability. 3) Mullite can be deposited by chemical vapor deposition (CVD) [20], and its composition can be manipulated over a range of alumina to silica ratios by tuning the input gas ratio into the CVD reactor, which can lead to functional gradation of coating properties [21], [22], [23], [24].

Yttria stabilized zirconia (YSZ) has been extensively used as a commercial TBC material [25], [26], [27] due to its low thermal conductivity, high durability, relatively low cost, and the ability to tune its properties by tailoring the stabilizer composition [28], [29]. The performance and high temperature mechanical and chemical stability of YSZ is very well studied and it has been proven to be a very reliable candidate for TBC applications and will be used as the TBC material of choice.

Both CVD and APS are mature technologies for coating deposition. Table 1 summarizes the main differences between CVD and APS deposited coatings that are significant to EBC/TBC applications. The high porosity and fast deposition rate make APS a good technique to deposit TBCs. The role of the TBC is to act as a heat transfer barrier, and the high porosity decreases the effective thermal conductivity and the higher thickness allows for a larger temperature difference across the coating. The properties of CVD coatings such as high density and controllable chemistry make CVD a better technique to produce EBCs. Dense coatings are crucial for an effective mass transfer barrier, and composition control is very useful to functionally grade the EBC. The slow growth rate also allows for fine control over coating properties as a function of thickness. Moreover, since CVD is not a line-of-sight process, it is ideally suited for batch processing and depositing uniform coatings over complex curvatures. The CVD process needs a significantly longer development time for a new coating. However, once the process is developed, coatings can be produced reliably at a mass scale using batch processing.

By combining the strengths and avoiding weaknesses of CVD and APS techniques, an EBC/TBC hybrid coating system is proposed (Fig. 2). It consists of a mullite-based EBC layer, deposited on the SiC/SiC CMC by CVD, and a YSZ TBC layer deposited on the EBC by APS. As demonstrated previously [22], [23], the mullite-based EBC will be functionally graded, such that it is stoichiometric mullite at the EBC/CMC interface, with the alumina content in mullite increasing towards the TBC/EBC interface. The advantages of this system are as follows: i) good CTE match at the EBC/CMC interface and at the EBC/TBC interface since the functionally graded mullite acts as a CTE bridge between the SiC and YSZ (Fig. 1), ii) improved hot-corrosion and recession resistance in the EBC at the EBC/TBC interface due to lower silica activity [24], iii) weight saving due to thinner EBC layer, and iv) lack of silicon bond coat, which is the cause of oxidation and spallation on long-term exposure [18], [19].