Study of Non-Transformable t’-YSZ by Addition of Niobia for TBC Application

Abstract

The high toughness of zirconia is paving the way for the development of new materials for application in TBC for gas turbine blades. The main aim of this work was the obtainment of tetragonal zirconia polycrystalline (TZP) with high density, from mixtures of high-purity powder of zirconia, yttria, and niobia with different compositions (14.5 to 21 mol%), through the processes of cold pressing by uniaxial pressing and by isostatic pressing, followed by air sintering processes at 1550 °C for 1 h. The samples were characterized for phase composition by X-ray diffraction, Rietveld analysis, and morphology by transmission electron microscopy and energy dispersive spectroscopy analyses. Mechanical and tribological resistance was evaluated by fracture toughness and nanoindentation tests as well as Weibull statistics. The incorporation of yttria and niobia resulted in relatively denser ceramics with stabilization of the tetragonal phase which was confirmed by detailed X-ray diffraction analysis. Modified ceramics for TBC with 17.5 mol% of yttria and niobia showed higher hardness and fracture toughness, 16.16 GPa and 173.38 GPa, respectively. Through nano hardness measurements, it was possible to verify the effect of the samples’ ferroelasticity. Thus, the addition of niobia and yttria to zirconia represents an opportunity for the development of new materials with increasing mechanical and tribological resistance for TBC application.

1. Introduction

Since the introduction of the coatings as thermal barrier coatings (TBCs) in gas turbines, the insulating layer of oxide has been ZrO₂ stabilized with 7.6 ± 1 mol% YO_1.5 (7YSZ). This choice is explained by the material’s higher thermal cycle durability and ease of production using plasma spray (APS) or electron beam physical vapor deposition (EB-PVD). Several materials with superior thermal resistance were suggested as alternatives for TBC; however, its implementation was not possible because of its low durability in service [1].

Analyzing the potential of new materials for application in TBC, it makes sense to explore new refractory materials [2,3,4,5,6]. However, with a large number of known crystal structures, both in mineralogy and chemistry, each one can be formed by several elements; there would be thousands of compounds to be searched. Because of this difficulty, initial research focused on the exploration of structures related to zirconia [7].

Research [8] revealed that, within the family of ZrO₂/HfO₂ (including 7YSZ and most new materials), the metastable tetragonal structure (t’-7YSZ) is preferred as it offers resistance substantially higher than that of cubic zirconia. The current hypothesis is that the zirconia t’ has greater durability for suffering the ferroelasticity hardening mechanism, a mechanism that can operate at high temperatures, effectively dependent on the tetragonality (c/a) of the structure.

When ZrO₂ is used for applications at high temperatures the polymorph phase, cubic (c) and tetragonal (t), should be stabilized at room temperature through the formation of solid solutions, which prevents the deleterious phase transformation, tetragonal to monoclinic (m).

The oxides that lead to stabilization are composed of earth elements, rare earths, and actinides. It has been suggested that the factors that may influence stabilization are dimensions, valence, and concentration of cations in the crystal structure of the solute and solute oxides, where the valence and concentration determine the number of oxygen vacancies created by the formation of substitutional solid solutions [9,10,11,12,13,14].

The introduction of a stabilizer required to avoid the deleterious effect of phase transformation from tetragonal to monoclinic is accompanied by the incorporation of a substantial number of defects [15]. When a trivalent oxide, Y₂O₃, is added to ZrO₂, a certain number of defects such as oxygen vacancies and negatively charged solutes are incorporated into the crystal lattice of zirconia [13].

The effect of doping with pentavalent oxides such as tantala and niobia (the cathode ray in the oxidation state is +5~0.68 Å for both) indicates that both ions enter as substitutional defects in the crystal lattice of zirconia (the ionic radius of Zr⁴⁺ is 0.79 Å), annulling the oxygen vacancies generated by yttria doping.

Materials containing tetragonal zirconia are promising for structural applications because of their stress-assisted tetragonal-to-monoclinic transformation near-room temperature, which enables the design of zirconia-toughened ceramics, and also because of their potential for ferroelastic domain switching at high temperatures [16].

The 20 mol% YTaO₄-stabilized zirconia is known to have Young’s modulus of 150 GPa compared with the 200 GPa modulus of 3–4YSZ. The lower modulus and the possibility of a higher coefficient of thermal expansion (CTE) were expected to reduce stresses owing to thermal expansion mismatches and therefore to be beneficial to TBC lifetimes as well [17,18,19,20,21].

The literature shows that only compositions—16 mol% YTaO₄ and 10 mol% YNbO₄ can be used for TBCs, as those containing lower amounts of the dopants destabilize into a monoclinic phase above room temperature [1,20].

To develop a study of ceramic compositions for application in TBC, collaboration was carried out between researchers from different countries in the Americas, targeting zirconia, yttria, and niobia ceramics with high density and ferroelasticity. Equimolar codoped zirconia ceramics with yttria and niobia were created aiming at ceramics with tetragonal and ferroelastic phases. Ferroelasticity was characterized by TEM and SEM/EDS images to characterize polarization arrangements in the domains. The resistance that the TBC application requires was evaluated by nanoindentation, which allows the determination of toughness in small volumes of materials, and together with measurements of load, hardness indentation, and modulus of elasticity, it was possible to calculate the critical intensity of mechanical stress (K₁C) related to fracture toughness. This mechanical resistance suggested a material with better surface characteristics and tribological resistance for TBC application.

2. Materials and Methods

Commercial powders of zirconium, yttrium, and niobium oxides were used in the preparation of sintered samples; they have high purity (99.9% purity) and fine particle size (D50 < 5 microns).

Table 1. Chemical composition of the sintered samples.

	Chemical Composition (mol%)			Density (Archimedes)	Tetragonality c/a
Sample	ZrO2	YO1.5	NbO2.5
14.5Y14.5Nb	71.0	14.5	14.5	5.27	1.0237
16Y16Nb	68.0	16.0	16.0	5.39	1.0230
17.5Y17.5Nb	65.0	17.5	17.5	5.38	1.0240
17Y17Nb	66.0	17.0	17.0	5.77	-
21Y21Nb	58.0	21.0	21.0	5.67	-

shows the range of chemical compositions to be studied in ZrO₂ samples doped with YO_1.5 and NbO_2.5, the specific density by the method based on Archimedes’ principle, and the lattice parameter c/a. The range of chemical composition studied was established considering the literature [17,19,20]. The powders were homogenized in a high-energy ball mill for 10 min. The samples were made with a matrix with a diameter of 10.0 mm and a thickness of 3.5 mm by uniaxial pressing (50 MPa) and isostatic pressing (300 MPa), then sintered in air at 1550 °C for 1 h.

The samples were characterized by X-ray diffraction, XRD (Rigaku, model Dmax2100/X’Pert-MRD), with Rietveld analysis. The microstructure was obtained by transmission electron microscopy, TEM (TECNAI G2—T20, FEI Company, Hillsboro, OR, USA, HT 200 KV, spot size 9, SA 8.700X TEM Bright Field), and energy dispersive spectroscopy, EDS, in the lamella prepared by FIB (Focus Ion Beam, model FEI 235 DB, FEI Company, accelerating voltage 5 kV and spot size 3, magnification 12,000×).

Nanoindentation tests (nanoindenter Hysitron, model TriboIndenter, Berkovich type sapphire tip) were conducted on samples 14.5Y14.5Nb, 16Y16Nb and 17Y17Nb.

The Weibull statistic is the best statistical treatment to characterize the mechanical strength to explain the behavior of a wide range of ceramic materials.

Employing the Weibull statistic, the following expression was obtained for the cumulative probability of fracture [22]:

$$P=1-\mathit{exp}\left[-\frac{V}{V_0} {\left(\frac{\sigma -\sigma u}{{\sigma }_0}\right)}^m\right]$$

(1)

where P = Failure probability; V = Tested volume of material; V₀ = Standard volume; m = The Weibull module; σ₀ = The tuning parameter; σ = The stress fracture of the material; σ_u = Stress below which the probability of fracture is zero.

This equation is known as Weibull distribution. For a sample with constant volume, Equation (1) can be simplified to Equation (2):

$$P=1-\mathit{exp}-{\left(\frac{\sigma -\sigma u}{{\sigma }_0}\right)}^m.$$

(2)

Constants m, σ₀, and σ_u are known as Weibull parameters; the stress–strain distribution is determined.

The determination of Weibull parameters is performed by modifying Expression 2 in order to transform it into the equation of a line under a system of axes ln [1/(1 − P)] versus ln σ [22]:

$$\mathit{ln}\frac{1}{1-P}=m\mathit{ln}\sigma -m\mathit{ln}{\sigma }_0$$

(3)

Using linear regression, it is possible to determine parameter values m and σ₀. The Weibull modulus “m” is a distribution parameter, characteristic of the material and related to the homogeneity of the aforementioned parameters, which may vary for different samples and/or in different manufacturing processes.

3. Results and Discussion

3.1. X-ray Diffraction

The XRD results are shown in Figure 1. It is observed that for the sample of just zirconia, the monoclinic phase (m) is predominant, while the samples with contents of yttria and niobia above 14.5% present tetragonal non-transformable (t’) and/or cubic (c) phases of zirconia due to the presence of a peak (111) at 30.2°. Samples with concentrations of yttria and niobia over 17.5% probably have a small amount of the m phase due to the presence of low-intensity peaks in monoclinic regions.

The XRD results of 14.5Y14.5Nb, 16Y16Nb, and 17.5Y17.5Nb samples are shown in Figure 2.

The determination of lattice parameters by Rietveld analysis was performed and the c/a ratio was obtained as a function of chemical composition. Regardless of the content of yttria and niobia, the c/a ratio remained constant at around 1.023; results are presented in Table 1. The c/a ratio for the (NbO_2.5-YO_1.5) mol% equal to zero was 1.022. The profile of the fabricated TBC ceramic showed no evident change in the lattice parameter.

The c/a ratio can be used as a criterion to decide between stable and unstable compositions. It can be noted that a room temperature c/a ratio of 1.023 (12–14 mol% YTaO₄) separates the stable and unstable compositions in the 1400 °C sintered samples. This is similar to the c/a ratio of pure tetragonal, 1.0234, at its transformation temperature of 1150 °C [17,18,19,20].

3.2. Electronic Microscopy

Figure 3 indicates by arrows the specific analyzes by EDS on the 17Y17Nb sample, which revealed different chemical compositions for the different regions. This analysis made it possible to study the distribution of chemical compositions and estimate the phases involved. The EDS analysis revealed different chemical compositions for the different regions for all samples under study, as can be seen in Figure 4 for composition 21Y21N.

Figure 5 and Figure 6 show images by TEM of Sample 17Y17Nb. It is possible to observe biphasic regions in Figure 7. In Figure 8, EDS point analysis is also carried out and the regions are identified as rich in tetragonal zirconia (14.5Y13.5Nb) and some small grains of YNbO₄ (43Y42Nb).

The micrographs obtained by TEM of the 21Y21Nb sample are presented in Figure 7. Figure 8 shows the biphasic regions. In the EDS analysis, regions rich in monoclinic zirconia YNbO₄ (43Y42Nb) were identified, as well as some grains of tetragonal zirconia (12Y11Nb).

Figure 3, Figure 4, Figure 5, Figure 6, Figure 7 and Figure 8 show that there are two kinds of grains in the samples, one with approximately 15Y15Nb and another with approximately 43Y43Nb. As for the appearance of these grains, the 15Y15Nb grains appear clear, and it is assumed that they are tetragonal. It is possible to find the [011] zone axis, but it is the [111] zone axis that shows the “forbidden” tetragonal spots (forbidden if the crystal is cubic; their presence is proof that the crystal is tetragonal). In the TEM images, there is a [011] zone axis diffraction image from a “tetragonal” grain in the samples. There is also diffraction of the YNbO₄ grain, and very strong “streaking” of the diffraction spots is shown, indicating twinning (Figure 7), which is characteristic of the monoclinic phase.

The results of the elemental analysis are consistent with the expected phase equilibrium of the ZrO₂-YO_1.5-NbO_2.5 system. All samples present two phases. The EDS analysis indicates a YNbO₄ phase (with 14–16 mol% of ZrO₂) and another phase with around 43–45 mol% of YO_1.5 and NbO_2.5, indicating tetragonal zirconia.

The electron diffraction patterns in the Y-Nb-Zr ceramic samples are not conclusive regarding the phases. Still, they indicate that the YNbO₄ phase is highly twinned (as expected for a monoclinic phase), and the other phase is a monoclinic phase. The crystal structure is related to FCC as cubic or tetragonal ZrO₂ ceramics.

3.3. Nanoindentation

The nanoindentation test results for Sample 14.5Y4.5Nb are shown in Figure 9.

Analysis by nanoindentation shows the curves of load versus penetration; it is possible to see the level of loading (pop-in) with different colors in Figure 9. Figure 10 is an example of the behavior of ferroelasticity, where there is a detail of a load versus penetration curve for a load level, indicated by the arrow.

Figure 11 and Figure 12 show, respectively, the hardness result and the elastic modulus for the 14.5Y14.5Nb sample, and the results of the Weibull statistics.

Table 2. Elastic modulus (Er) and hardness (H) of samples.

Sample	Modulus Er (GPa)	Hardness H (GPa)	Weibull Regression
			Modulus	Hardness
14.5Y14.5Nb	169.32 ± 9.70	10.96 ± 1.33	−107.04x + 20.75	−24.11x + 9.70
16Y16Nb	154.99 ± 8.22	10.87 ± 1.36	−109.00x + 11.71	−23.63x + 9.70
17.5Y17.5Nb	173.38 ± 11.96	12.16 ± 1.01	−87.56x + 16.88	−36.29x + 14.32

summarizes the results of nanoindentation for samples 14.5Y14.5Nb, 16Y16Nb and17.5Y17.5Nb. As the samples consist mainly of the tetragonal phase and a small amount of the monoclinic phase (Figure 7), the present values of the elastic modulus and hardness are similar. The Weibull regression shows straight lines with great inclination, especially for the elastic modulus, indicating samples with very homogeneous properties.

Previous research has demonstrated that the stability of the non-transformable zirconia base phase is confirmed by detailed X-ray diffraction analysis, where the diffusion of Nb⁵⁺ into the zirconia and yttria matrix results in further stabilization of its tetragonal phase [21]. Notably, tetragonal phases are identified in all structures of each composition of zirconia samples.

The results confirm that samples sintered from powders of high-purity zirconia codoped with approximately 17 mol% yttria and niobia are monophasic and tetragonal, with negligible amounts of the monoclinic phase. The identification in the DRX of the t phases is evident in Figure 1 but with no evident shift in the lattice parameter. The microstructure of this coating (Figure 6) shows fully coherent phases with Y-rich domains.

The literature [23] describes the effect of Nb₂O₅ codoping on the structure of 11 mol % Sc₂O₃ stabilization. ZrO₂ confirmed that niobium exists in a 5+ valence state in the sintered stabilized ZrO₂ samples and revealed that the addition of Nb₂O₅ not only assists in the densification of Sc₂O₃-ZrO₂ but also leads to exaggerated grain growth.

Moreover, the addition of these oxides increases the distortion of the tetragonal lattice, which is consistent with the increase in fracture toughness [24]. Analysis of load versus penetration curves by nanoindentation was obtained with a level of loading (pop-in), showing a result consistent with the expected behavior for materials with ferroelasticity.

In studies describing zirconia doped with titania and yttria, it was observed that tetragonality increases with increasing TiO₂. As a result, ferroelastic strain increases, causing higher fracture energy and therefore better toughness if only lattice variation is considered [16].

Therefore, the results are consistent with the hypothesis that fracture toughness can be improved by controlling the tetragonality of the structure through a proper selection of chemical composition.

The addition of NbO_2.5 can be used to achieve simultaneous improvements in toughness and phase stability, which are desirable for applications in TBC, and thus contribute to improvements in the tribological effect of the ceramic surface.

4. Conclusions

This study shows the following significant and systematic benefits of the sintered ZrO₂ samples doped with YO_1.5 and NbO_2.5 (14.5 to 21 mol%) for TBC:

• It is possible to identify biphasic regions in all samples and, by punctual EDS analysis, regions rich in tetragonal zirconia and YNbO₄.
• In the nanoindentation test, the results of the loading curves are consistent with the ferroelastic mechanism, although direct evidence of the spin domain has not yet been obtained.
• The Weilbull statistics for the modulus of elasticity indicate homogenization of values, which results in better mechanical and tribological resistance. Reducing the effects of erosion, surface wear, and abrasion represents an opportunity to develop better and longer-lasting TBCs.
• Compositions with higher doping levels (17% NbO_2.5) are also resistant to monoclinic transformation. The concept of increasing tetragonality and toughness by adding cations (Nb⁵⁺) represents an opportunity to create new materials with low thermal conductivity by combining cations. Thus, the addition of niobia and yttria to zirconia represents an opportunity to develop better TBCs by increasing their mechanical- and tribological-resistant (friction and wear) characteristics for the prolonged application.