Revolutionary High-entropy Zr-Y-Yb-Ta-Nb-O Oxides for Next Generation Thermal Barrier Coating Applications

China Abstract We report a revolutionary ceramic material with exceptional high temperature stability and superior thermo-mechanical properties for next generation thermal barrier coatings (TBCs) for aeroengines. The multicomponent oxides (Zr 1-4x Y x Yb x Ta x Nb x O 2 ) designed via a high entropy concept could exhibit a double tetragonal phase. The optimized composition breaks the limitation of intrinsic brittleness in previously reported TBC candidate materials and shows a superior toughness up to ~4.59 MPa m 1/2 due to ferroelastic and phase transformation toughening mechanisms. It also shows a remarkable high temperature stability at 1600 ºC, which is almost 400 ºC higher than the state-of-the-art yttria stabilized zirconia TBC material. In addition, it also exhibits a significantly lower thermal conductivity (~1.37 W∙m -1 ∙K -1 at 900 ºC) and a higher coefficient of thermal expansion (~11.3 × 10 -6 K -1 at 1000 ºC), as well as excellent corrosion resistance to molten silicate (~2.9 μm/h at 1300 ºC). This work provides a new approach to design ceramics by extending the high-entropy concept to both medium-entropy and high-entropy compositions searching for multifunctional


Introduction
Thermal barrier coatings (TBCs) are widely employed to protect the metallic components of gas turbine engines from direct exposure to high-temperature gas stream [1][2][3]. In the pursuit of enhanced engine efficiency, there has been an ever-increasing demand for increasing the operation temperature of engines. The state-of-the-art TBCs material, 7-8 wt.% yttria stabilized zirconia (7-8YSZ), is approaching its temperature limit, due to the destructive tetragonal to monoclinic (t-m) phase transformation when temperature exceeds 1200 ºC [4], and the devastating corrosion from the low-meltingpoint oxides CaO-MgO-Al2O3-SiO2 (CMAS) from sand, volcano ash and dust at elevated temperatures [5]. Driven by the urgent demand of next-generation TBCs materials working at > 1200 ºC, numerous efforts have been devoted to developing novel ceramics to replace YSZ. To work at higher temperatures, such materials have to meet a number of strict criteria including low thermal conductivity, outstanding hightemperature phase stability, high coefficient of thermal expansion (CTE), good chemical compatibility with thermally grown oxide (TGO), high resistance to CMAS corrosion and excellent mechanical properties. Although many new materials such as rare earth zirconates (RE2Zr2O7) [6], LaMgAl11O19 [7] and perovskites [8] have been proposed, none of them can satisfy all the above property requirements. Developing next generation TBCs materials remains a big challenge.
Recently, multicomponent high-entropy engineering offers a new approach to design high performance materials. High entropy materials usually have high lattice strain and thereby exhibit a low thermal conductivity, a high hardness, and other promising properties [9][10][11][12]. Tailoring the composition, defects, and degree of disorder of materials, high-entropy engineering provides a possible approach to achieve better and more tunable properties for TBCs applications. For instance, Lu et al. [11] designed a type of Y/Hf-doped AlCoCrFeNi high-entropy alloys (HEAs) with superior oxidation and spallation resistance, which are promising to replace the conventional NiCoCrAlY bond coat for TBCs. For high-entropy oxides (HEOs) as top coat for TBCs, Ren et al. [12] fabricated a multicomponent high-entropy rare earth zirconate (5RE2Zr2O7), which 3 shows a larger coefficient of thermal expansion (~11×10 -6 K -1 , 1273 K) and a lower thermal conductivity (0.86 W m -1 K -1 , 1273 K) than the monolithic RE2Zr2O5. However, the fracture toughness (~2.24 MPa m 1/2 ) of 5RE2Zr2O7 is too low to meet the requirement for TBCs. Nevertheless, HEOs show great potential as TBCs top coat materials, which motivates the search and development of novel HEOs with desired properties, in particular, high fracture toughness in addition to high-temperature stability, low thermal conductivity and high CTE.
Among the known candidate materials, the ZrO2-YO1.5-TaO2.5 (ZYTO) system shows promising properties for TBC applications. ZYTO oxides with equimolar YO1.5 and TaO2.5 demonstrate low and temperature-independent thermal conductivities (1.4~2.0 W•m -1 •K -1 ) owing to a minimum phonon mean free path [13], and meanwhile exhibit adequate toughness due to the ferroelastic transformation similar to 7-8YSZ [14]. However, the stable tetragonal region in the ZrO2-YO1.5-TaO2.5 phase diagram is extremely narrow [15], as a consequence, the composition and phase may easily deviate from its original state after high temperature operation, which significantly restricts their applicability in TBCs. This fatal problem may be solved by high entropy engineering as a large number of equivalent sites in a crystal with an intermediate sublattice will increase the configurational entropy and expand the elemental diversity containable in a single solid solution [16][17][18].
Here we aim to design a high-performance top coat material for TBCs used in next generation aeroengines via the high entropy strategy. We propose multicomponent HEOs based on ZYTO, in which equal-molar Yb and Nb are introduced to partially substitute Y and Ta, respectively. The multicomponent high entropy Zr1-4xYxYbxTa xNbxO2 tetragonal oxides exhibit significantly improved properties than the previously reported TBC candidate materials. The optimized composition shows a superior high temperature stability up to 1600 °C, a high fracture toughness of ~4.59 MPa m 1/2 , a low thermal conductivity of ~1.37 W•m -1 •K -1 at 900 ºC, a high CTE of ~11.3 × 10 -6 K -1 at 1000 ºC, as well as an excellent corrosion resistance to CMAS (~2.9 μm h -1 at 1300 ºC). Such revolutionary top coat materials are expected to facilitate the 4 development of next-generation aeroengines working at higher temperatures with significantly enhanced efficiency.

Results
Microstructure and phase composition. Zr1-4xYxYbxTaxNbxO2 (x = 0. which is supported by the homogeneous elemental distribution revealed by EDS ( Fig.S1b). It is worth mentioning that the composition Zr0.312Y0.344Ta0.344O2 with same ZrO2 content in the ZYTO system is a mixed monoclinic and tetragonal phase [19].
Therefore, the single T-ABO4(ZrO2) phase in t-2 suggests an enhanced phase stability 5 induced by high-entropy effects [20]. On the other hand, the enhanced Zr 4+ content from 20 mol % in t-1 to 31.2 mol % in t-2 decreases the monoclinic (M) → tetragonal (T) phase transformation temperature in ABO4 [21], which may also contribute to retaining the single T-ABO4(ZrO2) in t-2. Further decrease in x to 0.142 (t-3) and 0.114 (t-4) results in a double tetragonal structure (T-ABO4(ZrO2) + t-ZrO2(ABO4)), which can be clearly identified by XRD ( Fig.1a and 1b), EDS mapping and BSE observation ( Fig.S1c and 1d). Furthermore, t-4 shows a higher peak intensity of t-ZrO2(ABO4) than t-3 ( Fig.1b), which agrees with the increased amount of the Zr-rich phase observed by EDS mapping and BSE images in Fig.S1c and d. Formation of the two-phase structure results from the limited solubility of ZrO2 in the T-ABO4 solid solution. Furthermore, a high mixing configuration entropy, associated with mass difference, lattice distortion, and chemical bonding deviation, may cooperativity promote the formation of simple compounds with multiphase elements [22]. For x = 0.083 (t-5) corresponding to a Zr concentration of 66.8 mol %, the sample displays a single tetragonal phase (t-ZrO2(ABO4)), as revealed by XRD ( Fig.1a and 1b) and EDS mapping (  ZrO2(ABO4). The crystal structure of T-ABO4(ZrO2) (space group 141, I41/amd) is derived from the high-temperature tetragonal YTaO4 phase [23]. Zr 4+ ions are most likely to equally replace the neighboring A 3+ and B 4+ sites to preserve charge neutrality and to adopt a stable four-fold coordination for Zr 4+ in the T-ABO4 structure. For the t-ZrO2(ABO4) solid solution (space group 137, P42/nmc), the tetragonal ABO4 phase is rotated by 45° about the common c-axis and the volume of the tetragonal ZrO2 phase unit cell is approximately four times the volume of a tetragonal ABO4 phase unit cell [19]. The tetragonal structure of T-ABO4(ZrO2) can be retained at room temperature as revealed by the reflections in the [1 _ 11]ZAP (Fig.1f). Meanwhile, the {112} reflections in Notably, the complex co-doping at A and B sites of ABO4 structure results in shortrange disordered atomic arrangements and thus increases the configuration entropy, along with mass difference and local lattice distortion. Thermal properties and CMAS resistance. Fig.2a shows the thermal conductivities of Zr1-4xYxYbxTaxNbxO2 oxides between 200 and 900 ºC together with those for 8YSZ [24], ZYTO oxides [13] and the c-HEOs for comparison. Zr1-4xYxYbxTaxNbxO2 oxides show almost temperature-independent thermal conductivities despite of a noticeable 8 increase of thermal conductivity at T > 600 ºC for t-1 and t-5, which may be related to a thermal radiation effect. Why radiation effect is significant only in t-1 and t-5 remains unclear at the moment and needs further investigation. Nevertheless, Zr1-4xYxYbxTa xNbxO2 oxides exhibit much lower thermal conductivities (1.37~1.84 W•m -1 •K -1 at 900 ºC) than 8YSZ (> 2.5 W•m -1 •K -1 ). Among all the compositions, t-4 shows the lowest thermal conductivity (~1.37 W•m -1 •K -1 ) at 900 ºC, which is also significantly lower than ZYTO at the same temperature [13]. In Fig.2a phase contents from t-2 to t-4. It should be noted that t-4 presents a higher CTE than t-5 with a single t-ZrO2(ABO4) phase. This may be attributed to the higher Zr 4+ content in the t-ZrO2(ABO4) phase in t-4 (~73.7 mol.%) than that in t-5 (~ 66.8 mol.%) as later revealed by EDS mapping. The enhanced tetragonality [21] and anisotropic expansion of t-ZrO2(ABO4) contribute to the higher CTE in t-4.
Serving under long-term extreme high-temperature conditions, TBC materials should have preeminent high-temperature phase stability to avoid the catastrophic damage from phase transformation. However, most HEOs undergo a segregation of second phase after long-term high-temperature heat treatment [28]. To evaluate the long-term thermal phase stability of Zr1-4xYxYbxTaxNbxO2 oxides, they are subjected to heat treatment at 1600 ºC for 100 h. XRD patterns (Fig.S3) show absence of any second phase or phase transformation of Zr1-4xYxYbxTaxNbxO2 oxides (except for t-5) after exposure to 1600 ºC for 100 h, indicating the excellent high-temperature phase stability of these oxides. In particular, the t-4 sample with coherent double tetragonal phases is prone to form a single tetragonal phase after long-term heat treatment (Fig.2d), which suggests the development of an enhanced high-temperature phase stability.
As t-4 presents the lowest thermal conductivity, highest CTE and exceptional hightemperature stability, its resistance against CMAS corrosion is evaluated and compared with those of the state-of-the-art YSZ and the excellent CMAS-resistant material Gd2Zr2O7 [29], as shown in Fig.2e. At 1300 ºC, the total infiltration depth in t-4 shows a rapid increase (~26 μm) after 1 h CMAS corrosion, and then increases to ~144 μm after 50 h with a slow infiltration rate of ~ 2.9 μm h -1 , which is comparable to the superior CMAS resistance performance of Gd2Zr2O7 (~2.7 μm h -1 ). Cross-sectional SEM images of t-4 after CMAS corrosion at 1300 ºC for 1-50 h show the formation of a uniform CMAS infiltration layer (Fig.S4). The soluble product slightly precipitates  Gd2Zr2O7 [31], 5Re2Zr2O7 [12] and ReNbO4 [32] is presented in Fig.3a. The state-ofthe-art YSZ has a high fracture toughness of ~3.5 MPa m 1/2 due to a ferroelastic toughening mechanism [1]. Other materials, such as c-HEOs, Gd2Zr2O7, 5Re2Zr2O7 and ReNbO4, without any intrinsic toughening mechanism demonstrate a low fracture toughness. Among the Zr1-4xYxYbxTaxNbxO2 oxides, the medium-entropy composition t-4 with double tetragonal phases shows the highest fracture toughness (~4.59 MPa m 1/2 ), which is even higher than YSZ. SEM observations on the indents (Fig.S5) show distinctly different fracture behaviors between the single-phase samples (t-2 and t-5) and two-phase samples (t-3 and t-4). To understand the improved fracture toughness of t-4, morphology of the grains adjacent to an indent is observed by SEM (Fig.3b) Fig.3f) reveals that the area "g" (Fig.3e) has undergone ferroelastic reorientation to form three orthogonal twin variants in response to the varying stress fields, whereas only a single variant (the upper SAD pattern in Fig.3f) is found in the matrix away from the crack.
Bright field image of the ferroelastic domain process zone (Fig.3g) reveals patterns of nanoscale twin domains corresponding to the three tetragonal orientations. Two tetragonal domains are observed in the T-ABO4(ZrO2) phase (Fig.S6), demonstrating that the ferroelastic domain transformation of tetragonal ABO4(ZrO2) could also be stress-induced.
The coercive strain (γ T ) derived from ferroelasticity is directly related to the tetragonality (c/a) by [33]. The tetragonality of the single t phase (c/a ≈ 1.026) is higher than that of the T phase (c/a ≈ 1.022), which indicates that the t phase is more prone to the stress-induced domain switch. No monoclinic phase is found in this grain away from the indent, indicating that the stress-induced ferroelastic deformation precedes the martensitic transformation at low stresses due to the sufficiently low effective critical shear stress for domain (twin) wall motion [34]. All the above observations demonstrate that ferroelastic domain transformation can be retained both for t and T phase to provide sufficient fracture toughness even without phase toughening at high temperature for TBCs application. Besides the above discussed ferroelastic toughening, phase transformation observed in t-4 further increases the fracture toughness. Transformation of tetragonal to monoclinic phase is evident from the Raman spectrum collected from a crack tip (Fig.S7). The Raman peak at ~182 cm -1 of monoclinic ZrO2 phase is detected without presence of monoclinic ABO4, indicating that the t-m phase transformation only occurs in the t-ZrO2(ABO4) solid solution.
Large numbers of deformation twins induced by the phase transformation are found along an intragranular crack (Fig.4a). Twining deformations significantly consume the propagation energy of intergranular cracks and deflected them to the grain boundary with lower propagation energy. For example, as shown in Fig.4a, deformation of Twin 1 nucleates branch crack due to the lattice rotation at the twin boundary and large twining shear at the twin tip. This consumes the propagation energy of the main crack.
The multiple twins observed between Twin 2 and Twin 3 reveal discontinuous crack propagation where nucleation of twins consumes the stored energies at the crack front and impedes its growth. The crack is eventually directed towards the grain boundary along the shear direction of Twin 3.
All the above-mentioned twining deformations take place in the Zr-enriched t-ZrO2(ABO4) phase with a cation composition of Zr0.737Y0.048Yb0.048Ta0.066Nb0.101 according to the EDS mapping results (Fig.4b). The <111> SAD patterns obtained from the Twin 2 area and the matrix material (inset figure in Fig.4d) confirm the presence of the monoclinic ZrO2(ABO4) phase, suggesting the martensitic t→m transformation could be stress-induced, even though the t-ZrO2(ABO4) phase is retained at room temperature by increasing the configurational entropy using chemical doping. Notably, the t→m martensitic transformation of t-ZrO2(ABO4) results in a much lower dilatation strain than that of pure t-ZrO2 with about 4-6% dilatation due to its smaller monoclinic angle ( o 95.5   for m-YTaO4).
The role of deformation twinning is twofold. It alleviates the stress concentrations at the crack front and toughens the material while it also impedes dislocation motion and improves the strength of the material. Two types of twin boundary are observed: 15 coherent twin boundary (CTB) of the Twin 1 (Fig.4c) and incoherent twin boundary (ICTB) of the Twin 2 (Fig.4e). The lattice rotation associated with deformation Twin 1 is ~20° and the interplanar spacing becomes much smaller (~0.251 nm) when the (101) lattice plane of tetragonal phase transforms to (002) of monoclinic phase, leading to the nucleation of branch crack (Fig.4a) by the grouping of dislocations pile-up against the CTB. The HAADF STEM image (Fig.4d) shows large amount of precipitates T-ABO4 within a t-ZrO2(ABO4) matrix near Twin 2. The nanocrystalline structure of tetragonal precipitates within matrix could bring in short diffusion distances and short circuit paths for solute redistribution resembling to the 8YSZ [35] and ZYTO [36].  zone axis in the figure a (the insert is <111> SAD patterns); (e) Enlarged atomicresolution HAADF STEM images of Twin 2 area in the figure d. 16

Discussion
In summary, high-entropy principle was applied to stabilize the high-temperature tetragonal phase and successfully fabricated multicomponent medium-and highentropy Zr1-4xYxYbxTaxNbxO2 tetragonal oxides. Stress-induced ferroelastic domain variants and twining transformation arising from the martensitic phase transformations should be responsible to the enhanced fracture toughness (~4.59 MPa m 1/2 ) of mediumentropy composition with double tetragonal phase. Meanwhile, ultrahigh CTE (~11.3 × 10 -6 K -1 at 1000 ºC) was also achieved in this medium entropy composition.
Competitive temperature-independent low thermal conductivities (~1.37 W•m -1 •K -1 at 900 ºC) were obtained in the multicomponent Zr1-4xYxYbxTaxNbxO2 as result of the enhanced phonon scattering rate arise from higher mass disorder and shorter phonon mean free path. Besides, excellent high-temperature thermal stability at 1600 ºC and superior CMAS resistance (~2.9 μm h -1 at 1300 ºC) manifest that the multicomponent medium-entropy oxides reach a better balance in the overall properties than the highentropy compositions for TBC application. The present work provides a novel route to design TBC materials by extending the high-entropy concept to both medium-entropy and high-entropy compositions for improved properties, especially when mediumentropy ceramics precede their high-entropy counterparts.