Regulations of Thermal Expansion Coefficients of Yb_1−xAl_xTaO₄ for Environmental Barrier Coatings Applications

Abstract

Environmental barrier coatings (EBCs) are widely used to protect ceramic matrix composites (CMCs, SiC_f/SiC, and Al₂O_3f/Al₂O₃), and they should have low thermal expansion coefficients (TECs) matching the CMCs and excellent mechanical properties to prolong their lifetime. Current EBC materials have disadvantages of phase transitions and insufficient mechanical properties, which affect their working temperatures and lifetime. It is necessary to develop new oxide EBCs. Ytterbium tantalate (YbTaO₄) is a stable and novel EBC material, and we have improved the mechanical properties and TECs of Yb_1−xAl_xTaO₄ (x = 0.05, 0.1, 0.2, 0.3, 0.4, 0.5) ceramics by replacing Yb with Al. XRD, SEM, and EDS are used to verify the crystal and microstructures, and nano-indentation is used to measure the modulus and hardness when changes in TECs are measured within a thermal expansion device. The results show that the phase structure of Yb_1−xAl_xTaO₄ (x = 0.05, 0.1, 0.2, 0.3, 0.4, 0.5) is stable at 25–1400 °C within air atmosphere, and their high-temperature TECs (6.4–8.9 × 10⁻⁶ K⁻¹, 1400 °C) are effectively regulated by introductions of different contents of Al, which enlarge their engineering applications for SiC_f/SiC and Al₂O_3f/Al₂O₃ CMCs. The evolutions of TECs are analyzed from structural characteristics and phase compositions, and the increased TECs make Yb_1−xAl_xTaO₄ potential EBCs for Al₂O₃ matrixes. Due to the high bonding strength of Al–O bonds, hardness, as well as Young’s modulus, are enhanced with the increasing Al content, with Yb_1−xAl_xTaO₄ (x = 0.05, 0.1, 0.2, 0.3, 0.4, 0.5) having a nano-hardness of 3.7–12.8 GPa and a Young’s modulus of 100.9–236.6 GPa. The TECs of YbTaO₄ are successfully regulated to expand their applications, and they match those of Al₂O₃ and SiC matrixes, as well as displaying improved mechanical properties. This work promotes applications of YbTaO₄ as potential EBCs and provides a new way to regulate the TECs of tantalates.

1. Introduction

Currently, high-temperature nickel-based alloys are the primary materials used as hot-end components of aviation engines. With the rapid advancement of aircraft components and the increases in the power of engines, the maximum working temperature (1150 °C) of nickel-based alloys cannot meet their requirements. Thermal barrier coatings (TBCs) are applied to the surface of nickel-based components to provide thermal insulation accompanied with cooling technologies, which increase working temperatures [1,2,3]. However, the effects of TBCs are limited, and the traditional zirconate-based TBCs have working temperatures of less than 1200 °C [4]. Additionally, alloys have a high density, which increases the overall weight of engines to reduce fuel efficiency.

Accordingly, lightweight high-temperature ceramic matrix composites (CMCs) with operating temperatures of 1000–1650 °C are noticed for their outstanding performances [5,6,7,8,9]. Using CMCs can reduce the weight of these components and significantly enhance the thrust-to-weight ratio [10]. Common CMCs include Al₂O₃, C, and SiC fiber-reinforced Al₂O₃, as well as C fiber-toughened SiC (C_f/SiC) and SiC fiber-reinforced SiC (SiC_f/SiC) [11,12,13]. CMCs are susceptible to corrosion by molten salts, oxygen, and steam at high temperatures, which significantly reduce their lifespan [14,15,16,17,18]. Based on the above situations, environmental barrier coatings (EBCs) are mostly applied to shield CMCs from damages caused by high-temperature steam and oxygen, thereby extending their service life [19,20,21]. Oxide EBCs should have low thermal expansion coefficients (TECs, 3~9 × 10⁻⁶ K⁻¹) to match those of CMCs, and the interface thermal stress should be reduced as much as possible. The main functions of EBCs are isolating high-temperature steam and oxides to avoid the premature failure of CMCs; thus, EBCs must be stable at high temperatures. It can be seen that excellent high-temperature stability, adjustable TECs, and outstanding mechanical properties are essential for novel EBC materials.

The first-generation EBCs are mullite, which has low TECs and outstanding chemical compatibility with SiC. However, mullite will produce cracks during thermal cycling [22]. By combining mullite with yttria-stabilized zirconia (YSZ), researchers have improved the properties of mullite. YSZ has high TECs, which lead to thermal stress, and the resultant cracks allow water vapor to penetrate through EBCs to corrode CMCs [23,24]. Barium strontium aluminosilicate (BSAS) shows a substantial improvement in performance, but its high-temperature volatility and poor chemical compatibility with SiO₂ limit its applications [25]. Rare earth silicates (Rare earth = RE, RE₂SiO₅, and RE₂Si₂O₇) with low TECs and excellent SiO₂ chemical compatibility have replaced previous EBCs [26,27]. RE₂SiO₅ can withstand high temperatures for extended periods; however, they are unstable at temperatures beyond 1200 °C, and the phase transformations between polymorphs can produce strain and cracks which cause failure. It can be seen that excellent high-temperature stability is essential for EBCs. Additionally, the TECs of CMCs span from about 3 × 10⁻⁶ K⁻¹ to 9 × 10⁻⁶ K⁻¹ [17,28,29,30,31,32]_, indicating that low TECs are necessary for EBCs.

Chen et al. [33,34] have reported that YbTaO₄ and AlTaO₄ are stable without phase changes at 25~1500 °C, and they show excellent chemical compatibility with SiO₂. At 1200 °C, the TECs of YbTaO₄ and AlTaO₄ are 7.6 × 10⁻⁶ K⁻¹ and 5.6 × 10⁻⁶ K⁻¹, respectively. The above characteristics suggest that YbTaO₄ and AlTaO₄ are novel oxide EBCs. YbTaO₄ has weaknesses of a relatively low modulus and low hardness, which are not beneficial for its EBC applications. It is necessary to optimize the properties of YbTaO₄ EBCs, including TECs and mechanical properties. To expand the engineering applications of YbTaO₄, we adjust its TECs by Al-substituted Yb to form Yb_1−xAl_xTaO₄ (x = 0.05, 0.1, 0.2, 0.3, 0.4, 0.5) oxides. AlTaO₄ has a higher hardness and modulus than YbTaO₄, and Al-substituting Yb can enhance the mechanical properties of YbTaO₄ [35,36]. The results indicate that Yb_1-xAl_xTaO₄ ceramics have TECs matching SiC_f/SiC and Al₂O_3f/Al₂O₃ CMCs, and their TECs are successfully regulated by Al substitution. Furthermore, the hardness and modulus of YbTaO₄ are simultaneously enhanced to improve their lifetime. In a future study, it is necessary to research the CaO-MgO-AlO_1.5-SiO₂ corrosion and oxidation resistance properties of YbTaO₄. Overall, Yb_1−xAl_xTaO₄ ceramics have significant potential as EBCs according to their adjustable TECs and improved mechanical properties.

2. Experiments

2.1. Sample Synthesis

The synthesis of Yb_1−xAl_xTaO₄ samples (x = 0.05, 0.1, 0.2, 0.3, 0.4, 0.5) used Yb₂O₃ (purity ≥ 99.99%), Al₂O₃ (purity ≥ 99.99%), and Ta₂O₅ (purity ≥ 99.99%) powders from Aladdin, China, and their particle sizes were less than 10 μm. The raw powders were maintained at 1000 °C for 3 h to remove any moisture and then were mixed based on their chemical formula. Anhydrous ethanol was mixed with the weighted powders, and the mixtures were ball-milled for 12 h at 300 rpm. The mixed powders were dried at 100 °C for 12 h. The powders were subsequently pressed under a pressure of 6 MPa for 2 min to obtain ceramic disks. Finally, the disks were sintered at 1500~1700 °C in atmospheric conditions for 3–6 h to produce dense Yb_1−xAl_xTaO₄ ceramics.

2.2. Structural Analysis

An X-ray diffraction (XRD) device (MiniFlex600, Rigaku, Tokyo, Japan) was used to characterize phases of Yb_1−xAl_xTaO₄. The microstructure, as well as elemental distribution and content, were analyzed using energy-dispersive X-ray energy spectroscopy (EDS, Quantax 200, Bruker, Karlsruhe, Germany) and scanning electron microscopy (SEM, SIGMA-300, ZEISS, Oberkochen, Germany). The actual density (ρ) was obtained by the Archimedes principle:

$$\rho ={\displaystyle \frac{m_1}{m_3 - m_2}}$$

(1)

where m₁ was the initial weight, m₂ was the weight placed in water, and m₃ was the saturated wet weight of the sample.

2.3. Properties Measurements

The TECs (α_p) were measured within a thermal expansion device (DIL 402, NETZSCH, Bavaria, Germany) at 30–1400 °C based on the thermal expansion rate (dL/L₀):

$${\alpha }_p={\displaystyle \frac{\mathit{dL}\mathit{/}{L}_0}{\mathsf{\Delta }T}}$$

(2)

where dL, L₀, and ΔT were elongations in length, original length, and change in temperature, respectively.

The Young’s modulus and hardness of Yb_1−xAl_xTaO₄ were measured using a nano-indentation device (Nanomechanisc, Inc. iMicro, Milpitas, CA, USA) according to the nano-indentation depth (h ≤ 0.2 μm) specified in the national standard GB/T21838.1-2008 [37]. The load was 50 mN, and the holding time was 10 s because the nano Young’s modulus and hardness were stable when the load was higher than 20 mN in our previous study. A holding time of 10 s was used to stabilize the contact area between the sample and the nano-indenter.

3. Results and Discussion

3.1. Crystal and Microstructure

Figure 1a shows experimental XRD patterns of Yb_1−xAl_xTaO₄ ceramics compared with the standard ICDD cards. When x is 0.05–0.1, Yb_1−xAl_xTaO₄ is crystallized into a monoclinic-prime (m′) phase, and its corresponding standard card is PDF#24-1416. When x = 0.2–0.5, XRD patterns correspond to the standard card of PDF#24-1413 and 24-1416, indicating that Yb_1−xAl_xTaO₄ (x = 0.2–0.5) is a mixture of m′ and monoclinic (m) phases. Figure 1b shows that the XRD peaks of Yb_1−xAl_xTaO₄ are obviously shifted to a higher angle with an increment in Al content. The Al³⁺ ion has a shorter ionic radius than Yb³⁺, and Al³⁺ substitutes Yb³⁺ will contract lattice volumes [37]. The cell parameters and theoretical densities of Yb_1−xAl_xTaO₄ are obtained from XRD Rietveld refinement and are shown in Figure 2a–f and

Table 1. Lattice parameters (a, b, c/Å), β angle (β/°), phase content (wt%), unit cell volume (V/ų), density (ρ/g·cm⁻³), and relative density (ρ_R/%) of Yb_1−xAl_xTaO₄ ceramics.

Samples	Phase	Content	a	b	c	β	V	ρ	ρR
0.05	m′	100	5.07	5.26	5.44	96.17	144.192	9.496	97
0.1	m′	100	5.07	5.43	5.26	96.18	144.061	9.357	97
0.2	m′	91.5	5.07	5.43	5.25	96.16	144.005	8.972	98
	m	8.5	6.89	5.27	10.86	133.13	287.507
0.3	m′	85.5	5.07	5.43	5.25	96.16	143.999	8.650	97
	m	14.5	6.81	5.25	10.89	132.95	285.041
0.4	m′	80.9	5.07	5.25	5.43	96.18	143.643	8.309	97
	m	19.1	6.86	5.23	10.84	133.03	284.464
0.5	m′	76.0	5.07	5.25	5.43	96.16	143.518	8.017	96
	m	24.0	6.83	5.19	10.85	132.92	281.963

. For most samples, the weighting factor (R_wp) is less than 15% and the fitting factor (x²) is less than 3, which indicates a high credibility of lattice parameters [38].

Figure 3 shows the crystal structures of Yb_1−xAl_xTaO₄ with different space groups. The unit cell of the m′ phase has twisted [YbO₈] dodecahedra and [TaO₆] octahedra. Yb and Ta are layered in different planes, and the structure relies on the connection between [YbO₈] dodecahedra when two kinds of polyhedral are periodically continued along the a-axis. The primary distinctions between the m and m′ phases are that the oxygen coordination number of Ta atoms in the m phase is 4 and 6 for the m′ phase. According to Hazen and Prewitt’s theory, the thermal expansion of metallic ions in polyhedron is affected by their oxygen coordination number [39,40,41,42]. The higher the coordination number, the lower the TECs. Thus, m-Yb_1−xAl_xO₄ may have higher TECs than those with the m′ phase, and it will be discussed in Section 3.3.

Figure 4a–f shows that the grain sizes of each sample are less than 15 µm. The elemental distribution is analyzed by EDS mapping, and the elemental distribution of Yb_1−xAl_xTaO₄ is uniform when x = 0.05 and 0.1, while Yb_1−xAl_xTaO₄ ceramics are two-phase composites when x is greater than 0.1. Figure 4 shows SEM in the backscatter mode, while the differences in color are not obvious because they consist of the same elements. The points marked as 1~8 in Figure 4c–f are measured using an EDS point scan to estimate their elemental content, as shown in

Table 2. The compositions of Yb_1−xAl_xTaO₄ (x = 0.2, 0.3, 0.4, 0.5) ceramics, as marked in Figure 4, obtained from an EDS point scan.

Samples	Position	Yb (at%)	Al (at%)	Ta (at%)	O (at%)
Yb0.8Al0.2TaO4	1	3.8	19.73	15.22	61.25
	2	10.95	20.84	19.53	48.68
Yb0.7Al0.3TaO4	3	5.62	20.66	17.98	55.47
	4	9.93	15.75	16.56	57.77
Yb0.6Al0.4TaO4	5	3.86	20.71	17.39	58.04
	6	11.44	17.86	18.66	52.04
Yb0.5Al0.5TaO4	7	3.61	17.30	15.67	63.41
	8	10.39	15.82	17.44	56.35

. The Yb content in point 1 is less than that in point 2, and a similar situation is detected in the other three samples. It is believed that segregations of Yb and Al exist in Yb_1−xAl_xTaO₄ solid solutions; subsequently, this leads to co-formations of m and m′ phases. Previous studies prove that lanthanide contraction leads to the m-m′ phase transition in RETaO₄ ceramics [28,34,39], and Al³⁺ substituting Yb³⁺ reduces the A-site ionic radius in ABO₄-type YbTaO₄. The formations of m + m′ Yb_1−xAl_xTaO₄ composites indicate that the solid solubility of Al in YbTaO₄ is less than 0.2. The EDS point scan is semi-quantitative, and it is unable to obtain the accurate chemical forms of m- and m′-phase Yb_1−xAl_xTaO₄ ceramics.

3.2. Mechanical Properties

Young’s modulus is directly related to the bonding strength of crystal [43]. In a practical working environment, turbine engines are subject to impact friction from objects entrained in high-speed airflow. Therefore, a high modulus and high hardness are needed to improve the service life of YbTaO₄. Young’s modulus (100.9–236.6 GPa) is shown in Figure 5a for Yb_1−xAl_xTaO₄, which is lower than that of YSZ (250 GPa) but similar to those of La₂Zr₂O₇ (237 GPa) and Yb₂SiO₅ (160 GPa) [44,45,46]. The highest modulus is 236.6 GPa when x = 0.5, and the lowest modulus is 100.9 GPa when x = 0.05. The substitution between Al and Yb increases the bonding strength, which enhances the modulus. Figure 5b shows the hardness (3.7–12.8 GPa) for Yb_1−xAl_xTaO₄. The highest hardness (12.8 GPa) of Yb_0.5Al_0.5TaO₄ is higher than other EBC materials, and this phenomenon can be explained by the enhanced bonding strength. When Al substitutes Yb, the Al-O bond is stronger than the Yb-O bond, which increases the bonding strength and hardness. Both RETaO₄ and AlTaO₄ are monoclinic phases, and AlTaO₄ has a far higher bonding strength and modulus than most RETaO₄, except ScTaO₄. Thus, using Al to substitute Yb can enhance the hardness and modulus of YbTaO₄.

3.3. Thermal Expansion Performance

Figure 6a shows the thermal expansion rates of Yb_1−xAl_xTaO₄, and their slopes are almost constant with the increasing temperature, which can prove their excellent phase stability. Figure 6b shows that the TECs of Yb_1−xAl_xTaO₄ increase steadily with the increasing temperature, and they are approximately 6.0~9.0 × 10⁻⁶ K⁻¹ at 1400 °C. The TECs of single-phase Yb_1−xAl_xTaO₄ are lower than that of RE₂SiO₅, while the TECs of bi-phase Yb_1−xAl_xTaO₄ are similar to or slightly greater than that of RE₂SiO₅ [46]. The TECs of Yb_1−xAl_xTaO₄ ceramics are similar to those of CMCs, including Al₂O_3f/Al₂O₃ (8–9 × 10⁻⁶ K⁻¹) and SiC_f/SiC (4–6 × 10⁻⁶ K⁻¹). Indeed, the TECs of Yb_0.5Al_0.5TaO₄ match that of Al₂O_3f/Al₂O₃, and the TECs of Yb_0.95Al_0.05TaO₄ match that of SiC_f/SiC. The mathematical expression for TECs is [47]

$$\mathrm{TECs}\left(T\right)={\mathrm{TECs}}_0\left(T\right)+{\mathrm{TECs}}_1\left(T\right)$$

(3)

$${\mathrm{TECs}}_0\left(T\right)={\displaystyle \frac{3k_B}{2\phi r_{0}U}}{\left({\displaystyle \frac{T}{T_D}}\right)}^3{g}_0\left(x_D\right)$$

(4)

$${\mathrm{TECs}}_1\left(T\right)={\displaystyle \frac{\mathit{3}{k}_B}{2\phi r_{0}U}}{\displaystyle \frac{k_{B}T}{2U}}{\left({\displaystyle \frac{T}{T_D}}\right)}^3{g}_1\left(x_D\right)$$

(5)

$$g_0\left(x_D\right)={\int }_0^{{\displaystyle \frac{T_D}{T}}}{\displaystyle \frac{x^4{e}^x}{{\left(e^x-\mathit{1}\right)}^2}}dx$$

(6)

$$g_1\left(x_D\right)={\int }_0^{{\displaystyle \frac{T_D}{T}}}{\displaystyle \frac{x^5{e}^x\left(1+e^x\right)}{{\left(e^x- 1\right)}^3}}dx$$

(7)

where φ is the reciprocal of width for the energy barrier, k_B is the Boltzmann’s constant, r₀ is the equilibrium inter-ion distance, T_D is the Debye temperature, and U is the crystal-binding energy. The ionic radius of Al³⁺ is 0.535 Å, and the ionic radius of Yb³⁺ is 0.863 Å. The equilibrium inter-ionic distance decreases with the increasing Al content, and the crystal lattice is more tightly packed. The r₀ is inversely proportional to the TECs; subsequently, the TECs are increased with the increasing Al content. The crystal-binding energy (U) is [48]

$$U={\displaystyle \frac{N_{A}Az+z-e^2}{r_0}}(1 - {\displaystyle \frac{1}{B}})$$

(8)

where N_A is the Avogadro’s constant, A is the Madelung’s constant, z is the ionic charge, and B is the Born’s index. A decrease in the equilibrium distance causes an increase in the crystal-binding energy which reduces the TECs, which is contrary to the experimental results. The TECs are not determined by a single index since Yb_1−xAl_xTaO₄ is a composite when x is higher than 0.2. Yang et al. [49] have found that the TECs of m-phase RETaO₄ are higher than that of m′-phase RETaO₄ because m′-phase RETaO₄ has a higher crystal-binding energy. For m-phase RETaO₄, the cations and anions are located at two different positions, whereas the cation and anion are located at three different positions in m′-phase RETaO₄. The structural differences affect their thermodynamic stability and force constants. The force constant is directly proportional to the bonding strength indicated by the Debye temperature (T_D) [50,51]:

$$T_D={\displaystyle \frac{h}{k_B}}{\left({\displaystyle \frac{3e}{4\pi V}}\right)}^{{\displaystyle \frac{1}{3}}}{V}_M$$

(9)

where e is the total atomic weight per cell and h is the Planck’s constant. The Debye temperature is positively correlated with the interatomic bonding force. The strength of the Al-O bond is stronger than that of the Yb-O bond, leading to the Debye temperature being increased. Thus, the TECs of single-phase m′-Yb_1−xAl_xTaO₄ are reduced with an increasing Al content when x is 0.05~0.2. For x = 0.3~0.5, the TECs of Yb_1−xAl_xTaO₄ are dramatically increased due to the production of m-Yb_1−xAl_xTaO₄. Furthermore, thermal expansion is the non-harmonic oscillations of lattices reflected by the Grüneisen parameter (γ):

$$\gamma ={\displaystyle \frac{{\alpha }_p·V{·K}_T}{C_V}}$$

(10)

where V is the molar volume, K_T is the isothermal bulk modulus, and C_V is the constant volume heat capacity. Without considering C_V and K_T, the TECs of the m phase are higher than those of the m′ phase. Therefore, the Grüneisen parameter of the m phase is larger than that of the m′ phase, and a similar situation is reported in Sm_1−xYb_xTaO₄ ceramics [39]. There is a theory suggesting that the TECs of metal–oxygen bonds (TECs_M) are related to their oxygen coordination number (P) and cation charge (Z) [52,53,54]:

$${\mathit{TECs}}_M=32.9·(0.75 -{\displaystyle \frac{Z}{P}})\times {10}^{-6}$$

(11)

Equation (11) expresses that increasing the oxygen coordination number can enhance the TECs when the cation charge is unchanged, indicating that the TECs of Ta-O bonds in [TaO₄] of m-Yb_1−xAl_xTaO₄ should be lower than that of Ta-O bonds in [TaO₆] of the m′ phase. However, a current study of Yb_1−xSm_xTaO₄ shows that the m phase has higher TECs than the m′ phase attributed to the higher an-harmonic vibrations of lattices. It is believed that crystal structures are the dominators of the TECs for tantalates.

4. Conclusions

In this study, Yb_1−xAl_xTaO₄ (x = 0.05, 0.1, 0.2, 0.3, 0.4, 0.5) ceramics are successfully prepared to regulate their properties. The substitution between Al and Yb improves the hardness and modulus because Al-O bonds are stronger than Yb-O bonds. The nano hardness of Yb_1−xAl_xTaO₄ (x = 0.2, 0.3, 0.4, 0.5) (9.6–12.8 GPa) and the Young’s modulus (146.8–236.6 GPa) are increased effectively compared with YbTaO₄ and other EBC materials. The TECs of Yb_1−xAl_xTaO₄ are 6.4–8.9 × 10⁻⁶ K⁻¹ at 1400 °C, which are similar to those of the Al₂O₃ and SiC matrixes (3–9 × 10⁻⁶ K⁻¹). For single-phase m′-Yb_1−xAl_xTaO₄ (x = 0.05, 0.1), the addition of Al can slightly reduce the TECs because Al-O bonds are stronger than Yb-O bonds, which leads to a decrement in the an-harmonic vibrations of lattices. For bi-phase Yb_1−xAl_xTaO₄ (x = 0.2–0.5), their TECs increase with the increasing Al content due to the formation of the m phase. Crystal structures are the dominators of the TECs for Yb_1−xAl_xTaO₄ tantalates, which govern their lattice an-harmonic vibrations. Thus, the TECs of YbTaO₄ can be regulated by Al substitution. Accompanied with the enhanced mechanical properties, the studied Yb_1−xAl_xTaO₄ ceramics are promising EBCs.