**1. Introduction**

To improve their efficiency and design, turbine engines use ceramic-coated components. These coatings, known as Thermal Barrier Coatings (TBC), are designed for use at high temperatures [1–6]. TBCs serve as a protection for the base metal and super-alloy components by preventing them from experiencing high-temperature degradation [2]. They also increase the efficiency and lifetime of the components, besides providing creep resistance, thermal shock resistance, strain tolerance, higher temperature stability with respect to the substrate material and protection against hot corrosion [1–4]. Reducing the temperature of the metal is one objective of the top ceramic TBC. A state-of-the-art TBC is yttria-stabilized zirconia (YSZ) composed of ZrO2 with 6%–8% Y2O3 [3–5]. The microscopic structure of TBC is always very inhomogeneous [6].

Atmospheric plasma spray (APS) is used for depositing ceramic TBC [6]. So called "splats" are created by the plasma spray as flattened particles. Due to its quick solidification, the sprayed powder produces a coating on a clean surface of a substrate. An incomplete bonding between splats caused by the relaxation of residual stresses when the splat is being cooled, or by trapped gases, and also lack of adhesion, are typical APS–TBC drawbacks [6,7]. The structure of the porous TBC, which contains

cracks, pores, crack/coating interfaces, pore/coating interfaces and layer interfaces, affects its effective thermal conductivity [1,2,6,7].

In the case of sprayed TBC, the microstructural defects consist of three types of pores: interlamellar pores—that result from the build-up of micro-splats; microcracks—that result from the stress relief after coating deposition; and globular pores—that are due to a lack of complete filling. These three types of pores fall into different ranges of size distribution. It has been shown that the thermal conductivity is strongly dependent on the pore morphology and porosity [8,9]. Optimum porosity level in TBC is considered to be about 15% total porosity as a compromise between positive function of porosity to minimize thermal conductivity and negative function of worsening the coating integrity and mechanical properties. Some 20% to 25% of the total porosity is formed by cracks [9].

Among the investigated ceramic TBC candidates, SrZrO3 with perovskite structure has a high melting point, low thermal conductivity as well as the possibility of extensive substitution of ions at the *A* or/and *B* site, making it a promising TBC candidate material [7,10]. Application of pure SrZrO3 for the TBC seems to be limited because of its high-temperature phase transformations [10]. SrZrO3 exhibits a pseudo-tetragonal structure at 750 ◦C and a tetragonal structure at temperatures higher than 840 ◦C. The pseudo-tetragonal structure creates a lattice mismatch between Sr-O atoms, so that coating delaminates easily [11]. Delamination at the boundary bond-coat/top-coat is the most frequent case [2,5]. Elsewhere [12], the transformation sequence at heating is described as follows: orthorhombic (*Pnma*)→730 ◦C→pseudo-tetragonal (*Imma*, c/a < 1)→860 ◦C→tetragonal phase (*I4*/*mcm*, c/a<sup>&</sup>gt;1)→1170 ◦C→cubic (*Pm*3*m*).

When the gas turbine operates in a severe environment, such as the desert or the vicinity of a volcanic eruption, siliceous mineral debris matters (dust, sand and ash) in air are ingested by the turbine and deposited on the hot TBC surfaces as molten calcium–magnesium–alumino-silicate (CMAS) [13]. Immediately when CMAS melting takes place, it infiltrates into the TBC material via open porosities in APS deposited TBC and undergoes a series of chemical reactions with the TBC-forming oxides [14]. Upon cooling, CMAS solidifies and develops a high stress level because the pores were blocked. In conventional TBC, YSZ partially dissolves in the CMAS, causing disruptive phase transformation (from tetragonal ZrO2 to monoclinic ZrO2) accompanied with a significant volume increase (up to 5%) [15,16]. The size of CMAS particles could be from a nanometer range up to approximately half a millimeter. Kakuda et al. [17] investigated the effect of amorphous CMAS infiltration on the thermal properties and heat transport of plasma-sprayed (APS) coatings and observed a rise in both volumetric heat capacity and thermal conductivity of the coating upon infiltration. Concerning the mechanical attack of the dust on TBC, cavitation is mentioned [18], but the chemical attack is considered as more serious.

A paper dealing with an interaction of CMAS and the La2Zr2O7-SrZrO3 composite TBC coating [19] expressed a challenge that the interaction behavior of SrZrO3 in contact with CMAS melt at high temperatures requires further investigations in the future. This challenge to elaborate such experiments is now accepted by a co-author of papers dealing with multifunctional SrZrO3 coatings [20–22]—and this is in a focus of the actual paper. Keeping in mind that pure SrZrO3 is not optimal for a thermal barrier application because of high-temperature phase transformations but to study it directly in a composite with other refractories, for example La2Zr2O7-SrZrO3, is difficult [19], the decision was to study it as a single-component coating material. This is the main novelty of the current paper.

## **2. Materials and Methods**

## *2.1. Sample Preparation*

Plasma spray grade strontium zirconate powder supplied by Cerac Incorporated (Milwaukee, WI, USA) was used as the feedstock. The powder size was 74–150 μm. Plasma spraying was done by the water-stabilized plasma system (WSP) torch [23] at 150 kW power (500 A, 300 V). The feeding distance (from the plasma exit nozzle to the point of the powder feeding in the plasma

stream) was 80 mm and spray distance (from the plasma exit nozzle to the substrate) was 350 mm. Compressed air was used as the feeding gas and the substrates were preheated to 460 ◦C. This high temperature was selected with the purpose to minimize the cooling rate for impacting splats. Preheating was done by plasma torch passes with the powder feeding switched o ff. After each pass of the torch, manipulated by a robotic arm, the temperature rose to 350 ◦C and was pushed down to 170 ◦C by a compressed air flux before the next pass started [20].

Reactivity of the SrZrO3 coating with calcium—magnesium—aluminum-silicate (CMAS) powder was tested. The CMAS powder was Ultrafine Test Dust "Arizona desert sand" produced by Powder Technology (Arden Hills, MN, USA). Its chemical composition provided by the producer is SiO2 (50%, semi-quantitative content), Na(AlSi3O8) albite (32), CaMg(Si2O6) diopside (16), CaCO3 (2). The size of CMAS particles was in single micrometers. The dust powder with a concentration of 30 mg·cm<sup>−</sup><sup>2</sup> [19] was mixed with ethanol and the resulting slurry was applied on the coating surface by a brush. Then, this sample was dried in air at room temperature for 3 h and subsequently annealed in air (laboratory furnace Clasic, Revnice, Czech Republic) using dwell times of 1, 4, 8 and ˇ 10 h, respectively, at 1250 ◦C. Heating ramp was 8 ◦C·min−<sup>1</sup> and cooling ramp 6 ◦C·min−1. At this temperature (and during this time) the CMAS melt interacts with the coating, i.e., it both infiltrates into the coating and reacts with it. To better understand the phase constituents of the reaction products, the CMAS powder was mixed with the SrZrO3 feedstock powder with a weight ratio of 1:1 [19] by mechanical blending, followed by heat treatment with 10 h dwell time under the same conditions as the TBC specimens.
