Effect of grain boundary segregation on aging resistance and mechanical properties of tetravalent element-doped 3Y-TZP ceramics for dental restoration

ABSTRACT

In the humid oral environment, 3Y-TZP ceramics always suffer from low-temperature degradation (LTD) for a long time, which results in the degradation of mechanical properties and catastrophic failure. The low-temperature degradation (LTD) and mechanical properties of low-cost tetravalent (Ge⁴⁺, Ti⁴⁺) element-doped 3Y-TZP were investigated by analysing grain boundary segregation in samples with deferent contents. The results show that GeO₂ is superior to TiO₂ in limiting LTD but results in lower flexural strength and fracture toughness when the content is ≥1.5 mol%. This dilemma can be improved by adding only 0.1%-0.5 wt% Al₂O₃, and the flexural strength and fracture toughness of 0.25 wt% Al₂O₃ zirconia are then increased to 898 MPa and 4.68 MPa·m^1/2 compared with 1Ge-3Y, respectively. This work is expected to provide an effective reference for the development and application of budget dental materials.      

GRAPHICAL ABSTRACT

KEYWORDS:

• 3Y-TZP
• low-temperature degradation
• grain boundary segregation
• mechanical properties
• GeO₂

1. Introduction

3Y-TZP is a 3% yttria-stabilized tetragonal polycrystalline material with a high fracture toughness and strength and is widely used for dental crowns, fixed bridges, and dental implants [1]. However, their surface will spontaneously transform into the martensitic phase when subjected to a humid and complex oral environment for an extended period [2]. The transformation of the tetragonal phase (t) to the monoclinic phase (m) will lead to surface microcracking, and water molecules will pass through these cracks and penetrate the material, increasing the roughness and decreasing the mechanical properties, which greatly shorted the clinical life of zirconia restorations. This phenomenon is also known as low-temperature degradation (LTD) [3].

To solve this problem, researchers have introduced rare earth cations (including Ce⁴⁺, La³⁺, and Y³⁺) as dopants to ensure the aging resistance of 3Y-TZP ceramics. However, the price of rare earth cations is high, and there needs to be a systematic method for choosing a suitable dopant that simultaneously exhibits both aging resistance and excellent strength. According to the mechanism of electroneutrality, oxygen vacancies can be introduced by trivalent dopants [4], which can occupy these oxygen vacancies and destabilize the tetragonal phase. Tetravalent dopants (Ce⁴⁺, Si⁴⁺, Ge⁴⁺, and Ti⁴⁺) do not create anion vacancies [5] and can stabilize tetragonal zirconia against monoclinic distortion. Among these tetravalent dopants, Ce⁴⁺ (with an ionic radius of 0.097 nm) is a large cation that decreases oxygen overcrowding around Zr⁴⁺ (with an ionic radius of 0.084 nm) and stabilizes the tetragonal phase. CeO₂-doped TZP (Ce-TZP) shows improved LTD resistance compared to 3Y-TZP without introducing oxygen vacancies [6]. However, the flexural strength of Ce-TZP is much lower than that of standard 3Y-TZP due to its larger grain size. Ti⁴⁺ (with an ionic radius of 0.074 nm) and Ge⁴⁺ (0.053 nm) are smaller than Zr⁴⁺ and occupy a smaller space in the crystal lattice [7,8]. The smaller tetravalent cation radius will introduce tensile stresses, which can keep the tetragonal phase at room temperature [9]. At the same time, these cations can also segregate to grain boundaries and block the diffusion pathways of water molecules; however, the detailed mechanism of their stability is not yet understood. Moreover, previous studies [10] have estimated a decrease in flexural strength and fracture toughness with increasing tetravalent cation content. It is well known that [11] doping 3Y-TZP with small amounts of Al₂O₃ can enhance its mechanical properties and suppress the t-m transformation. Previous studies [12] have proposed that Al₂O₃ additions lead to Ce-TZP materials with a fine microstructure and significant transformability, as well as enhanced mechanical properties (including strength and fracture toughness). Recently, researchers [13] have introduced small amounts of Al₂O₃ and 1 mol% CaO to 10Ce-TZP to retard LTD without compromising the mechanical properties. However, the solubility of Al₂O₃ in zirconia is extremely limited [14,15], so adding a suitable amount of Al₂O₃ is essential.

There needs to be more research on the LTD of tetravalent element (GeO₂ or TiO₂)-doped 3Y-TZP, the mechanism of the aging resistance still needs to be determined, and a suitable content remains to be studied. Therefore, this study investigated the effect of the ionic radius of the tetravalent dopant on the LTD behavior and mechanical properties in 3Y-TZP to further understand the diffusion stabilization process of GeO₂ during the nonprecursor transformation of tetragonally stabilized zirconia. Further, we also focused on improving the mechanical properties and aging resistance of 3Y-TZP by codoping with Al₂O₃ and GeO₂.

2. Materials and methods

2.1 Sample preparation

The 3 mol% yttria-stabilized zirconia powder (TZ-3Y, purity >99.99%, Tosoh, Japan) had an impurity concentration of approximately 0.1 wt%, with SiO₂, Fe₂O₃, NaO₂, and Al₂O₃ being the chief impurities. The specific surface area of 3Y-TZP was 16 ± 3 m²/g, and the mean particle diameter was 0.6 μm. The other starting powders were GeO₂ (purity >99.999%, Aladdin), TiO₂ (purity >99.8%, Aladdin), and Al₂O₃ (purity >99.99%, Aladdin) powders. The 3Y-TZP powders were mixed with 0–2 mol% GeO₂, TiO_2, or codoped GeO₂ and Al₂O₃ powders by wet milling in a Teflon jar for 24 h using 5 mm and 1 mm zirconia balls as the milling media and ethanol as the mixing medium. TZ-3Y refers to 3Y-TZP, and the mol% is indicated in the nomenclature (e.g. 0.5Ge-3Y refers to 0.5 mol% GeO₂ and 3 mol% Y₂O₃). Samples were dried in an oven at 80°C for 8 h to obtain xGe/Ti-3Y and 1Ge-xAl-3Y zirconia powders.

The powders were isostatically cold pressed at 250 MPa for 2 min and pressureless sintered in air at 1450°C for 2 h, with constant heating and cooling rates of 3°C/min and 10°C/min, respectively. The samples were then cut into 36 × 4 × 3 mm rods. The indicated sintered samples were ground in turn by 120, 240, 600, 800, 1200, 2000, 3000, 4000, and 5000 grade SiC papers and subsequently polished with 1 and 0.05 μm diamond pastes to obtain an optically reflective surface.

2.2 Microstructural analysis

The densities of the samples were tested in deionized water according to Archimedes’ principle. The grain size distributions were obtained on polished surfaces after thermal etching at 1350°C for 20 min. The grain size was measured according to the linear intercept method, performed on pictures recorded by field-emission scanning electron microscopy (SEM, InLens, Germany). At least 1000 grains were counted by Nano Measurer 1.2.5 software, and the outcomes were reported as average results (± standard deviation). The distribution of dopant cations (Y³⁺, Ge⁴⁺, Ti⁴⁺, and Al³⁺) and Zr⁴⁺ near the grain boundaries in thin foils was analyzed by scanning transmission electron microscopy (STEM, FEI Talos F200×, U.S.A), energy-dispersive X-ray spectroscopy (EDS) and a high-angle annular dark-field (HAADF) detector.

2.3 Assessment of aging kinetics

Each group’s LTD of polished samples (n = 3) was tested in deionized water for 10, 15, 20, and 30 h at 134°C and a pressure of 2 bar using an autoclave. The amount of all specimens that transformed was measured by X-ray diffraction (XRD, Japan Smartlab) using Cu-Kα radiation at 40 kV and 40 mA, a scan range of 20° to 80° (2θ), and a step size of 0.02 s⁻¹. Fractions of m-ZrO₂ and t-ZrO₂ were measured by the diffraction patterns in accordance with the method used by Garvie [16] and modified by Toraya [17]:

(1) $X_m=\frac{I_m(-111)+I_m(111)}{I_m(-111)+I_m(111)+I_t(101)}$(1)

where X_m is the relative amount of the monoclinic phase, I_m and I_t are the intensities of the monoclinic and tetragonal peaks, respectively.

2.4 Characterization of mechanical properties

According to ISO 6872, the three-point bending strength was measured on 36 × 4 × 3 mm³ test bars (5 bars were tested for each sample).

(2) ${\sigma }_{\mathrm{t}}=\frac{3FL}{2b{d}^2}$(2)

where F is the maximum load, L is the space of fixtures, b is the sample width, and d is the sample thickness.

The Vickers hardness and fracture toughness [18,19] were measured by the microindentation method. Indentations were made on the surfaces of samples with different compositions at 98.1 N, with expected cracks originating at the corners of the indentation after loading. SEM imaged indentations and the apparent indentation fracture toughness were analyzed by using the equation proposed by Niihara:

(3) $K_{IC}=\frac{0.035}{3}(\frac{l}{a}{)}^{-0.5}{H}_V\sqrt{a}(\frac{H_V}{3E}{)}^{-0.4}$(3)

where ${\mathrm{H}}_{\mathrm{V}}$ is the Vickers hardness, E is the elastic modulus of ceramic materials (taken as 210 GPa [20,21]), a is half of the diagonal of the indentation, c is the separation between the indentation center and the tip of the crack, and l is the propagation crack length. Equation (3)(3) $K_{IC}=\frac{0.035}{3}(\frac{l}{a}{)}^{-0.5}{H}_V\sqrt{a}(\frac{H_V}{3E}{)}^{-0.4}$(3) can be applied when 0.25≤(c-a)/a≤2.5 [22] in the case of small indentation cracks.

2.5 Statistical analysis

One-way analysis of variance (ANOVA) with Tukey’s post hoc test and Tukey’s multiple comparisons test was employed in the statistical analysis of mechanical properties (n = 5). GraphPad Prism 8 software was employed in the analysis, where α was set to 0.05. Mean values plus standard deviations are presented in the results, and the different letters in the figures indicate statistical significance.

3. Results and discussion

3.1 Effect of the tetravalent GeO₂/TiO₂ content on the hydrothermal aging of 3Y-TZP

From the XRD analysis (Figure 1(a), 2(a), all of the samples before aging consisted of the tetragonal ZrO₂ phase regardless of the amount of added GeO₂/TiO₂. After aging for 30 h (Figures 1(b), 2(b), the I_t (101) main peak at 30.2° (2θ) and the monoclinic I_m (−111), I_m (111), and I_m (002) peaks positioned at 2θ values of 28.2°, 31.3°, and 34° start to emerge, with their intensities tending to decrease with increasing tetravalent dopant amount. The monoclinic phase content of 3Y-TZP containing different amounts of GeO₂/TiO₂ is plotted in Figure 3 against accelerated aging time. Note that compared with GeO₂/TiO₂-doped 3Y-TZP, undoped 3Y-TZP exhibited a higher monoclinic phase content. The lines indicate that increasing the stabilizer (Ge⁴⁺/Ti⁴⁺) will improve the resistance to hydrothermal degradation. Moreover, when the aging time exceeded 15 h, the 1.5Ge-3Y samples started to age slowly and stabilized at a monoclinic phase content of approximately 9%, less than the monoclinic phase content in 1.5Ti-3Y ceramics. This suggests that 1.5Ge-3Y displayed more excellent aging resistance than 1.5Ti-3Y after 30 h in steam, corresponding to 90 years in the human body [23]. Since a transformation can only be detected at a depth of up to 8 μm by XRD, we used SEM to further study the degradation area of the cross-section (in Figure 4). After aging for 20 h, the thickness of the degradation layer in 1.5Ge-3Y (~6 μm) was smaller than that in 3Y-TZP (~16 μm) and 1.5Ti-3Y (~8 μm). These observations demonstrate that GeO₂-doped 3Y-TZP shows better resistance to low-temperature aging than TiO₂-doped 3Y-TZP.

Figure 1. X-ray diffraction pattern obtained from samples of GeO₂-doped 3Y-TZP (a) before and (b) after 30 h aging at 134°C.

Figure 2. X-ray diffraction pattern obtained from samples of TiO₂-doped 3Y-TZP (a) before and (b) after 30 h aging at 134°C.

Figure 3. Monoclinic phase content from 3Y-TZP with different GeO₂/TiO₂ content against accelerated aging time.

Figure 4. SEM images of the degraded zone from the cross-sections of (a) 3Y-TZP (b) 1.5ge-3Y (c) 1.5ti-3Y (d) 1ge-0.25al-3Y after 20 h aging at 134°C, the transformation zone is indicated by the area between the dotted lines, the upper zones are resin which was used to imbed the zirconia.

Increasing the content of GeO₂/TiO₂ is expected to improve the susceptibility of 3Y-TZP to LTD, and the chemistry of the dopant and its concentration are closely related to grain boundary segregation, which can enhance the stability of local zirconia grain boundaries and thus decrease the susceptibility to LTD [5].

3.2 Grain boundary segregation in MO₂-3Y-TZP: M=Ge/Ti

Grain boundary segregation in tetravalent cation-doped 3Y-TZP is common when sintering at high temperatures. Tetravalent cations such as Ce⁴⁺, Mn⁴⁺, Si⁴⁺, Ge⁴⁺, and Ti⁴⁺ [24–28] can stabilize the tetragonal phase by inducing grain boundary segregation at room temperature, with the detailed mechanisms depending on the cation type and ionic radius. In the current study, GeO₂ and TiO₂ were specifically chosen, as their cations have a lower tetravalent ionic radius than the host cation. The cation radius evolution was as follows: Y³⁺ (0.104 nm) [7] >Zr⁴⁺ (0.084 nm) > Ti⁴⁺ (0.074 nm) > Ge⁴⁺ (0.053 nm) [5,29]. We aimed to research the influence of cation radius on low-temperature degradation.

The EDS elemental maps reveal the segregation of Ge⁴⁺, Ti⁴⁺, Y³⁺, and Zr⁴⁺ to the grain boundaries (Figure 5a-b). The distribution of Ge⁴⁺, Ti⁴⁺, Y³⁺, and Zr⁴⁺ across the grain boundary is plotted in line scans (Figure 5d-e). Ge⁴⁺ tends to segregate to a wide area within up to 10 nm from the grain boundary, while Y³⁺ is apt to segregate to a narrow area within ~5 nm from the grain boundary. At the grain boundaries, the content of GeO₂ reached 8%, while the content of Y₂O₃ was 2%, and the concentration of GeO₂ was greatly enhanced at the grain boundaries. In 1.5Ti-3Y, Y³⁺ ions and Ti⁴⁺ ions both tend to segregate to a wide area approximately 5 nm from the grain boundary, with concentrations of ~3–4%.

Figure 5. HAADF-STEM images, EDS elemental mappings and corresponding chemical compositions of Ge⁴⁺, Ti⁴⁺, Y³⁺, Zr⁴⁺ and Al³⁺ across the grain boundaries of (a,d) 1.5ge-3Y, (b,e) 1.5ti-3Y and (c,f) 1ge-0.25al-3Y. The dots represent the experimental data, and the black lines represent the fitting curves.

The discrepancies between the radii of dopant cations and the host cation were mainly responsible for the different grain boundary segregation behaviors of various cations. The differences among the cation sizes of Y³⁺, Ti⁴⁺, and Zr⁴⁺ are similar, resulting in similar grain boundary segregation behavior. The Ge-O bond is much smaller than the Ti-O and Zr-O bonds. The smaller substitutional Ge⁴⁺ cation could decrease the bond length and increase the binding strength, allowing for greater mobility of the Ge⁴⁺ atoms and therefore strengthening grain boundary segregation [7,29,30]. Grain boundary segregation is more substantial when the difference between the dopant cation radius and the host cation radius is more significant. Increased tetravalent element grain boundary segregation can strengthen grain boundaries and inhibit water molecules’ diffusion, leading to excellent aging resistance in GeO2-doped 3Y-TZP.

The diffraction analysis indicates that both GeO₂ and TiO₂ can stabilize the tetragonal phase Figure 6a-b; the GeO₂-doped grains have a tetragonal crystal structure, whereas the TiO₂-doped grains have a tetragonal or cubic crystal structure. According to the grain boundary segregation-induced phase transformation mechanism (GBSIPT), a large amount of Y³⁺ starts to segregate to grain boundaries when temperatures exceed 1300°C, decreasing its concentration in the tetragonal phase grain regions. Consequently, the regions in 1.5Ti-3Y with high Y³⁺ ion concentrations transform into cubic grains. A possible explanation for this behavior was proposed by Chevalier, who suggested that large cubic grains will act as new nucleation sites for tetragonal-to-monoclinic (t-m) martensite phase transformation, and monoclinic phase grains are more susceptible to degradation than tetragonal grains [30]. This mechanism may also explain why 1.5Ge-3Y is superior to 1.5Ti-3Y in resisting LTD.

Figure 6. Diffraction analysis of the area where the Roman numerals located form (a) 1.5ge-3Y, (b)1.5ti-3Y, and (c) 1ge-0.25al-3Y.

To further understand the grain boundary segregation behavior of GeO₂, we studied the diffusion mechanism of GeO₂ in 1.5Ge-3Y at lower temperatures. At 1250°C (Figure 7), the EDS elemental mappings also reveal the segregation of Ge⁴⁺, Ti⁴⁺, and Y³⁺ to the grain boundaries, whereas at 1450°C, Ge⁴⁺ tends to segregate to a wide area within up to 10 nm from the grain boundary, and the GeO₂ content at the grain boundaries reaches 6.7%; however, the grain boundary segregation behavior of Y³⁺ is slightly weaker than that of 1.5Ge-3Y sintered at 1450°C.

Figure 7. HAADF-STEM analysis of 1.5ge-3Y sintered at 1250°C: (a) EDS elemental mappings. (b) Diffraction analysis of the corresponding area in the images of TEM. (c) Corresponding chemical compositions of Ge⁴⁺, Y³⁺, and Zr⁴⁺ across the grain boundaries.

3.3 Mechanical properties of MO₂-3Y-TZP: M=Ge/Ti

Grain boundary segregation can affect not only the aging resistance of 3Y-TZP but also the mechanical properties of the material. First, increasing the doping amount of GeO₂ or TiO₂ slightly lowers the hardness. Second, as shown in Table 1, the flexural strength of 3Y-TZP decreases with GeO₂/TiO₂ doping and decreases after a significant increase in TiO₂ doping amount (and has almost no effect on that doped with 1.5 mol%). The flexural strength will drop greatly with increasing grain size for samples containing the same dopant. The grains of xTi-3Y are smaller than those of xGe-3Y, so the flexural strength of xTi-3Y is higher than that of xGe-3Y because doping with tetravalent elements will promote grain growth and coarsening and lead to a decrease in density and flexural strength. Another possible explanation for the difference between the strengths of xGe-3Y and xTi-3Y is that a large amount of GeO₂ grain boundary segregation leads to an increase in the number of impurities at the grain boundaries, which affects bridging between grains and decreases the density, thus influencing mechanical properties [30] (in Table 1). In Figure 8(c), the flexural surface was investigated by SEM, and the 1.5Ge-3Y materials exhibited predominantly intergranular fracture modes. The segregation of many impurities results in much lower cohesion between grains, leading to intergranular fracture and lower flexural strength [3].

Figure 8. SEM images of polished and thermally etched surfaces of (a) 1.5ge-3Y and (b) 1ge-0.25al-3Y zirconia ceramics sintered for 2 h at 1450°C and fractured surfaces of (c) 1.5ge-3Y and (d) 1ge-0.25al-3Y after a flexural strength test. TEM microstructure images of (e) 1.5ge-3Y and (f) 1ge-0.25al-3Y.

Table 1. Group, density, grain size, and mechanical properties of GeO₂/TiO₂-doped 3Y-TZP.

Groups	Density (g/cm^3)	Grain size (nm)	Flexural strength (MPa)	Vickers hardness, Hv (GPa)	Indentation toughness, KIF (MPa·m^1/2)
TZ-3Y	6.02±0.10^[Aab]	320.62±113.41^[Aa]	815.48±48.72^[ABCa]	13.00±0.06^[Aa]	5.04±0.05^[Aa]
0.5Ge-3Y	6.13±0.01^[a]	345.28±112.57^[b]	628.21±22.14^[1b]	12.71±0.09^[b]	5.01±0.07^[1a]
1.0Ge-3Y	6.11±0.04^[a]	342.90±116.28^[b]	609.69±24.93^[1bc]	12.47±0.10^[c]	4.14±0.03^[1b]
1.5Ge-3Y	6.00±0.07^[ab]	342.55±124.22^[b]	611.16±32.52^[1bc]	12.24±0.08^[d]	4.07±0.13^[1b]
2.0Ge-3Y	5.94±0.00^[b]	396.04±137.64^[c]	509.07±71.53^[1c]	12.22±0.08^[d]	4.11±0.08^[1b]
0.5Ti-3Y	6.07±0.02^[A]	319.88±107.05^[A]	714.39±7.39^[2ABD]	12.78±0.08^[B]	4.93±0.19^[1A]
1.0Ti-3Y	6.05±0.03^[A]	318.12±112.46^[A]	786.41±61.11^[2BC]	12.50±0.07^[C]	5.01±0.14^[2A]
1.5Ti-3Y	6.04±0.02^[A]	330.79±118.89^[A]	864.56±52.12^[2C]	12.28±0.09^[D]	4.98±0.03^[2A]
2.0Ti-3Y	5.97±0.52^[A]	407.53±143.32^[B]	618.62±31.86^[1D]	12.24±0.12^[D]	5.21±0.26^[2A]

Note: Values are means ± standard deviations, the same superscript letters in square brackets means no difference, lowercase letters indicate comparison between groups with different amounts of Ge⁴⁺, capital letters denote the comparison between groups with different contents of Ti⁴⁺. Roman numerals indicate the comparison between two groups doped with the same content of Ge⁴⁺ and Ti⁴⁺. A different letter or number means a statistical difference.

Moreover, fracture toughness is likely related to the segregation behavior of many solute atoms at grain boundaries; however, the detailed mechanism underlying this behavior is unclear. Generally, the fracture toughness of xTi-3Y is typically higher than that of xGe-3Y. On the one hand, based on the line scan in Figure 5d-e, more Y₂O₃ is located at grain boundaries in 1.5Ti-3Y than in 1.5Ge-3Y, which leads to less Y³⁺ within grains. The composition of internal grains is responsible for the higher fracture toughness of 1.5Ti-3Y due to the t-m transformation toughening mechanism [31–34]. The monoclinic phase content on the fracture surface represents the stress-induced phase transformation rate, and the phase transformation rate of 1.5Ti-3Y is 12.7%, which is higher than the 3.1% phase transformation rate of 1.5Ge-3Y in Figure 9.

Figure 9. XRD patterns at the fractured surfaces of 1.5ge-3Y, 1.5ti-3Y ceramics.

On the other hand, we can observe stacking faults in 1.5Ti-3Y (areas 1 and 2 in Figure 6(b)). The scattered regions of a raised surface may result from the volume increase associated with martensitic transformation near grain boundaries [35]. Thus, local tensile stress will be present at the grain boundaries and lead to the formation of metastable tetragonal zirconia. Cubic grains will act as nucleation sites for the tetragonal-to-monoclinic transformation [36], which increases the fracture toughness of xTi-3Y; however, this phenomenon is not common [37].

The grain boundary structure, and the presence of stacking faults near the grain boundary can affect the distribution of solute elements and influence the strength and toughness of the materials [35,38]. In short, the degree of segregation depends on the impurity concentration at the grain boundaries and can explain the differences between the mechanical properties of GeO₂- and TiO₂-doped 3Y-TZP.

3.4 Co-doping 3Y-TZP with Al₂O₃ and GeO₂ to improve aging resistance and mechanical properties

GeO₂ doping can significantly improve the aging performance of 3Y-TZP, but it has a detrimental influence on mechanical properties. This conclusion is consistent with the research of Mastui et al. [39]. Therefore, how to improve the mechanical properties while maintaining the excellent aging resistance of GeO₂-doped 3Y-TZP ceramics is the focus of current research.

As shown in Table 2, the flexural strength and fracture toughness of 1Ge-3Y were enhanced by adding 0.1 wt% to 0.5 wt% Al₂O₃. This enhancement was not observed for the hardness. From the micrographs provided in Figure 8a-b, all samples were densified without obvious porosity and adding Al₂O₃ slightly enhanced the grain growth of 1Ge-3Y. In addition to the role of Al₂O₃ on the flexural strength of the materials, 0.25 wt% Al₂O₃-doped 1Ge-3Y exhibited more transgranular fracture than 1.5Ge-3Y Figure 8d, which may be related to the segregated Al³⁺ increasing grain boundary cohesion [40–42]. In our study, addingAl₂O₃ can slightly enhance the fracture toughness of 1Ge-3Y; the material doped with 0.25 wt% Al₂O₃ had a 12.32% higher Vickers indentation hardness than 1Ge-3Y.

Table 2. Group, density, grain size, and mechanical properties of GeO₂ and Al₂O₃ co-doped 3Y-TZP.

Groups	Density (g/cm^3)	Grain size (nm)	Strength (MPa)	Hardness, Hv (GPa)	Toughness, KIF (MPa· m^1/2)
1.0Ge-3Y	6.11±0.04^[a]	342.90±116.28^[a]	609.69±24.93^[a]	12.47±0.10^[a]	4.14±0.03^[a]
1Ge-0.1Al-3Y	6.08±0.03^[a]	363.67±123.70^[b]	665.99±50.34^[a]	12.55±0.03^[ab]	4.58±0.05^[b]
1Ge-0.25Al-3Y	6.02±0.06^[a]	385.22±150.34^[c]	898.03±92.98^[b]	12.34±0.12^[ac]	4.68±0.03^[c]
1Ge-0.5Al-3Y	6.03±0.04^[a]	376.99±155.24^[bc]	782.23±138.56^[ab]	12.51±0.09^[ab]	4.59±0.05^[b]

Note: Values are means ± standard deviations, the same superscript letters in square brackets means no difference, a different letter means a statistical difference.

Three effects may explain the influence of alumina on toughness. One possible explanation may be that increased Y³⁺ segregation can slightly improve the indentation toughness [43]. A second explanation is that the various laws of grain size and fracture toughness are similar (Table 2); a larger grain size corresponds to a greater tendency for phase transformation, contributing the stress-induced phase transformation toughening mechanisms. To further determine the toughening mechanism of ceramic materials, Raman spectra were collected from the indentation center, crack root and crack tip in Figure 10a-b. For any component, the monoclinic phase content at the crack root is the largest, followed by that at the indentation center, with that at the crack tip being the smallest. In addition, 1Ge-0.25Al-3Y induces greater t-m transformability around the crack than 1Ge-3Y, thus resulting in higher toughness.

Figure 10. Raman spectra near the indentation zone of zirconia ceramics: (a)1.0ge-3Y, (b) 1ge-0.25al-3Y. (c) SEM images of indentation-induced cracking in 1ge-0.25al-3Y ceramics.

Moreover, other traditional toughening mechanisms, including fracture bridging, oxide wedging, and crack deflection, are also presented in Figure 10(c).

Third, an amorphous phase is located at the triple junctions, which may reduce residual strains at the grain boundaries of rounded grains compared with faceted and sharp grains in 1.5Ge-3Y Figure 8e-f. The amorphous phase mainly consists of Ge⁴⁺ and Al³⁺ (Table 3). According to a previous study [11], the formation of a grain boundary amorphous phase caused by the addition of alumina may contribute to the improved mechanical properties and resistance to t-m transformation; however, the detailed mechanism of LTD resistance has not been determined.

Table 3. The elements compositions of the amorphous phase.

Elements	Ge	Y	Al
Atomic %	13.54	2.04	11.05

According to ISO standard 13,356:2008, a monoclinic phase content of less than 25% should be expected for dental ceramics after heating for 5 h at 134°C in an autoclave [44]. In the present study, XRD analysis (Figure 11) indicates that the aging resistance of GeO₂- and Al₂O₃-codoped 3Y-TZP is superior to that of GeO₂ singly doped 3Y-TZP. Moreover, Figure 4 shows the average thickness of the degradation layer of 1Ge-0.25Al-3Y (varying from 1 to 3 μm, which was in accordance with the XRD results) after 20 h of hydrothermal treatment, which is much lower than that of 1Ge-3Y. We can conclude that GeO₂- and Al₂O₃-codoped 3Y-TZP materials have sufficient aging resistance compared with currently known bioceramics and meet clinical requirements. From the TEM and diffraction analyses of 1Ge-0.25Al-3Y in Figure 5c-f, Ge⁴⁺ is more segregated to the grain boundaries than Al³⁺ and Y³⁺ [45] and stabilizes the tetragonal phase at room temperature. Previous studies [3] have reported that segregated Al₂O₃ may also play a significant role in influencing the nucleation of aging and the concentration of free oxygen vacancies, thus reducing the sensitivity to low-temperature aging. In the present study, adding 0.25 wt% Al₂O₃ improved the aging resistance of 1Ge-3Y; however, the aging resistance was not further improved with the addition of 0.1 wt% or 0.5 wt% Al₂O₃, which may be related to the solubility of Al₂O₃ in 3Y-TZP. It is likely that Ge⁴⁺ and Al³⁺ segregated at grain boundaries and played major and minor roles, respectively, in retarding the LTD of 1Ge-xAl-3Y.

Figure 11. X-ray diffraction pattern obtained from samples of GeO₂- and Al₂O₃-codoped 3Y-TZP (a) before and (b) after 30 h aging at 134°C. (c) Monoclinic phase content of GeO₂- and Al₂O₃-codoped 3Y-TZP against accelerated aging time.

4. Conclusion

The present investigation has shown that tetravalent oxides (GeO₂ or TiO₂) with different cation radii resulted in different trends in the aging behavior and mechanical properties as a function of dopant concentration in the range of 0.5–2 mol%. On the one hand, the present work showed that adding more significant amounts of GeO₂ or TiO₂ resulted in substantially higher hydrothermal aging resistance than undoped 3Y-TZP. More importantly, GeO₂ is superior to TiO₂ in limiting low-temperature degradation, which could be directly related to the large discrepancy between the radii of the tetravalent cation and Zr⁴⁺; Ge⁴⁺ exhibited strong segregation at the grain boundaries with increasing temperature, which was preferred. On the other hand, TiO₂ is better than GeO₂ in terms of mechanical properties, which may also be relevant to the grain boundary structure. Various parameters must be considered, including the mechanical properties, in selecting the most suitable zirconia ceramic for biomedical applications. Therefore, we codoped GeO₂ and Al₂O₃ in 3Y-TZP and found that 1Ge-0.25Al-3Y exhibited excellent mechanical properties while maintaining the hydrothermal aging resistance. However, within the scope of this work, additional efforts will be made further to explore the aging resistance and mechanical properties of Al₂O₃- and GeO₂-codoped 3Y-TZP sintered at different temperatures. It is necessary to study the stable zirconia obtained from the transformation process of nonprecursors and the diffusion stabilization process of the subsequent addition of rare earth elements to further guide grain boundary engineering.

Disclosure statement

No potential conflict of interest was reported by the authors.