Effect of Fe 2 O 3 Addition and Sintering Temperature on mechanical Properties and Translucence of Zirconia Dental Ceramics with Different Y 2 O 3 Content

2021 The influence of Fe 2 O 3 additions on the microstructure, mechanical properties and translucence of zirconia with 3 mol-% or 5 mol-% Y 2 O 3 was evaluated. Samples were pressed with different thickness and sintered at 1475, 1500, 1550 and 1600°C for 2 h. Density, phases, microstructure, roughness, strength and translucency were analyzed. All samples showed densification greater than 99%. After sintering at 1600 ° C, the phase analysis revealed only tetragonal ZrO 2 in samples with 3% Y 2 O 3 , while in samples with 5% Y 2 O 3 the cubic ZrO 2 phase was identified. Significant increase in the grain size was noted, when increasing the Y 2 O 3 content from 3% to 5% from 0.8 µm to 2.3 µm for sintered samples at 1600°C. The strength was affected with Y 2 O 3 content and no relevant influence of the Fe 2 O 3 on density, microstructure or mechanical properties. Visible reflectance spectroscopy analyses, related to the CIELab scale, indicate variation in the contrast ratio as function of thickness of samples and coloring agent Fe 2 O 3


Introduction
Over the past few decades, the demand for dental prostheses an aesthetic appearance of natural teeth has stimulated an intense technological rush to replace the metallic infrastructure in prostheses 1 . In this scenario, structural ceramics emerged as potential substitutes and so-called "metal free" prosthesis systems have been developed entirely composed of ceramic materials 2 .
Among ceramics applied in dentistry, zirconia, ZrO 2 , stands out due to its excellent mechanical properties, such as high fracture toughness and bending strength, as well as being biologically inert [1][2][3] . In particular, ZrO 2 stabilized with 3 mol% Y 2 O 3 , technically known as 3Y-TZP, exhibits outstanding mechanical properties due to the stabilization of the tetragonal phase. This phase, t-ZrO 2 , is metastable at room temperature and may undergo transformation into the stable, monoclinic phase, in the wake of a tensile stress field of a crack. This transformation is accompanied by a volume change in the order of 4% to 5%, creating itself compressive stresses in the matrix, which hinders crack propagation and resulting in the highest fracture toughness of all ceramic materials 4-6 .
One of the best-performing manufacturing processes for manufacturing dental prostheses, in terms of speed and precision, is CAD-CAM milling, computer-aided design (CAD) computer-aided manufacturing (CAM). In this process, the geometry of the prostheses is detailed by a software and pre-sintered, still porous and low strength zirconia blocks are automatically milled 7,8 . After milling, the prostheses are subjected to a second sintering cycle, aiming at eliminating porosity and maximizing mechanical and aesthetic properties.
The adjustment of color shades of zirconia to natural teeth occurs by adding metal oxides as coloring agents such as Fe 2 O 3 , CeO 2 , Er 2 O 3 , Bi 2 O 3 and MnO 2 . Two processing routes are commonly employed 9 : the oxides are added to the starting powder, homogenized, compacted and presintered and then directed to milling, or pre-sintered zirconia blocks are infiltrated with a solution of the coloring agents. However, the infiltration method is less effective because it generates zones with different concentrations of the solution and consequently different shades of color throughout the same block 9,10 .
Although 3Y-TZP has exceptional mechanical properties, the low translucency of the anisotropic tetragonal phase, prevents an imitation of natural teeth by monolithic zirconia ceramics, requiring the application of a covering porcelain layer. However, the mechanical properties of porcelain are inferior than that of the 3Y-TZP core and therefore more prone to failure 11 . Furthermore, the porcelain cover is applied manually, representing a time-consuming process with little replicability 4,7 .
Therefore, the production of dental prostheses of monolithic zirconia with a high degree of translucency and a coloration similar to natural teeth would drastically reduce the need for coverings, representing a significant technological advance in relation to the current manufacturing process, both in mechanical performance and in economic viability of the final product 4,12 .
The literature converges on the parameters that impact the optical properties of monolithic ceramic materials [13][14][15] . The thickness of the material is undoubtedly the most important parameter for determining the real fraction of light, since the sum of events referring to all other parameters, such as porosity, birefringence and absorption, depends on the thickness 16,17 . An equation to simulate real online transmission was proposed by Beer -Lambert -Budworth 16 Where "T" = transmittance; "R" = reflectance; "α" = coefficient of absorption intrinsic to the material; "S p " = coefficient of scattering, referring to residual porosity; "S b " = coefficient of scattering referring to grain boundary (by birefringence or a second phase) and "t" = thickness. The initial efforts to increase the translucency in zirconia were aimed at reducing the particle size of the starting powders, thus improving the sintering kinetics. In this way, it was made possible to reduce sintering temperature and firing time, as well as to reduce the porosity of the sintered bodies 18 . In parallel, the purity of the starting powder has also been improved minimizing contamination at grain boundaries and increasing the translucency of the blocks 19 .
In recent years, some publications have focused on the induction of the cubic phase, through stabilization with 5 mol% of Y 2 O 3 , as a means of maximizing translucency due to the isomorph characteristic of this phase, as it has a cubic lattice no birefringence is expected 12,20 . On the other hand, a higher amount of cubic phase diminishes the amount of tetragonal phase, which, in consequence, lowers the mechanical properties 12 . However, it is still unclear of how much a higher cubic phase content may improve the translucency of pigmented blocks in detriment of the final mechanical properties.
In this work, the impact of stabilizing ZrO 2 with 3 mol% or 5 mol% of yttrium oxide on the optical, crystallographic and mechanical properties of polycrystalline zirconia, without or with 0.136 wt% of iron oxide as pigment has been evaluated.

Materials
Four different commercial zirconia powders of the TOSOH Corporation -Japan were used in this work, Zpex (G3) and Zpex Yellow (G3Y), stabilized with 3 mol% of Y 2 O 3 , designated as G3 and G3Y, and Smile (G5) and Smile Yellow (G5Y), stabilized with 5 mol% of Y 2 O 3 , designated G5 and G5Y. The "yellow" powders contain also 0.136 wt.% Fe 2 O 3 corresponding to the highest iron oxide concentration necessary to imitate the yellowest natural teeth, as described by color scales used in dentistry. The manufacturer data of the powder characteristics are presented in Table 1.

Processing
Discs (Ø12.5 x 1.3 mm discs) were compacted uniaxially under a pressure of 100 MPa-30s. They were then sintered at four different final sintering temperatures of 1475, 1500, 1550 or 1600 °C, maintaining a 2-hour threshold. The samples was sintered in air, with a heating rate adopted was 2°C/ min, up to 1100°C with an isothermal holding time of 120 min and further heating up to the pre-established final temperature, with a rate of 5°C/min. This rate was also used for cooling down to room temperature.

Characterizations
The apparent density of the sintered specimens was measured using the Archimedes principle, following the instructions described in the ASTM C373-06 standard 21 . The relative density (%) was calculated by the ratio between the apparent density and the actual density of each studied composition obtained by measuring powder samples of each composition, after a heat treatment at 1600°C-2hr, in a helium pycnometry, Ultrapyc model 1200e.
The identification of the crystal phases was conducted by X-ray diffraction analysis using a Panalytical -Empyriam diffractometer, Cu-Kα radiation (1.5405 Å) in the 2θ range of 10 o to 90°, angular step of 0.01° and a counting time of 100s. The diffractograms were compared with standard crystallographic information sheets (CIF), obtained from the database of inorganic crystallographic structures (ICSD) from the online repository administered by the FIZ Karlsruhe -Leibniz institute. To compare the diffraction patterns obtained with the catalog cards, the X'pert Highscore plus (PANalytical) program was used. For the quantitative phase analysis, Rietveld refinement was done using the FullProf program.
The microstructure of the sintered samples was observed by scanning electron microscopy, JEOL -JSM 7100FT, of polished surfaces. In order to reveal the grain boundaries, the samples were thermally etched at a temperature of 1400°C for 15 min and heating and cooling rates of 20°C/min and 10 °C/min, respectively. After the thermal attack, the sample surfaces were coated with a gold layer of approximately 20 nm, using a BAL-TEC metallizer, mod.MED 020. The determination of the average grain size was done using the ImageJ software, measuring a minimum of 700 grains per sintering condition, following the methodology proposed by Peregrina-Barreto et al. 22 .
The surface roughness of the samples used for determining the bending strength and of the spectrophotometry tests was determined by optical interferometry with a 3D optical profilometer, Zygo, model NewView 7300.

Bending strength
The biaxial flexural strength of sintered samples was determined by the piston on 3 ball (P-3B) method, using a piston and a base with three semi-spheres of high hardness tempered steel, positioned in a circular and equidistant manner with an angle of 120º to each other, supporting the specimens. A thin plastic film was used to reduce the friction between the spheres of the device and the surface of the specimens. The device was then coupled to an universal testing machine, model EMIC-DL10000, and subjected to a loading rate of 0.5 mm/min until rupture of the samples. The where F represents the failure probability, σ the bending strength in MPa; σ o the characteristic Weibull strength in MPa and m the Weibull modulus. Although the number of samples are relatively low for the Weibulls modulus analisys, n=15 instead of n=30, turning the coefficient of variation higher them the expectable, with due care and reinforced by other indications this data can be useful for further discussions.

Visible reflectance spectroscopy
The influence of thickness and composition on translucency and color of sintered samples was analyzed using reflectance colorimeter, X-Rite model CF57 / CA, geometry 45/0, following the guidelines of ISO 2471: 2008 25 , using samples with thicknesses of 0.35 mm, 0.8 mm, 1.0 mm, 1.2 mm and 1.35 mm.
The equipment was powered with a light source based on CIE standards for the CIELab measurement, D65 standard, and standard samples were used for the white (L = 97,53; a = -0,04; b = 1,62) and black (L = 18,67; a = 0,03; b = 0, 14) backgrounds. The results obtained were processed using equations that correlate the Lab quantities to obtain the contrast ratio (Equations 3 and 4) and the color difference (Equation 5).
Where: "CR" is contrast ratio; "Y" is spectral reflectance; "L" is luminance; "a" is coordinate "a*"; "b w " is coordinate "b*" and the b and w indices are related to the black and white backgrounds, respectively. Also, "Y n " is the spectral reflectance for light reflected by a diffuser with perfect reflection, illuminated by the same light source as the object = 100; "ΔE" is the color difference.

Relative density
The relative density of the samples sintered under different conditions is shown in Figure 1. All samples exhibited relative density greater than 99%, indicating that all sintering cycles were effective in densification of the specimens and that the residual porosity of samples is very low.
The effect of iron oxide (Fe 2 O 3 ) on the densification and grain growth of yttrium-stabilized zirconia (YSZ) has been debated in the literature. Guo and Xiao 26 did not detect any statistically relevant variation in the values measured for the relative density of zirconia doped with 3 or 8 mol% of Y 2 O 3 compared to doping with up to 2 mol% of Fe 2 O 3 . The data reported by these authors are in agreement with the results measured in the present study, where a maximum amount of 0.136 wt.% Fe 2 O 3 was added, affirming that within the limits of the analysis of this work, no effective variation of relative density was observed.  Qualitatively, all G3Y samples exhibited the tetragonal ZrO 2 phase (spatial group P42 / nmc) as major crystalline phase. Although the phase diagram presented by Scott 27 proposes the existence of a cubic ZrO 2 phase (space group Fm3 ̅ m), no trace of this phase could be found in the diffractograms. However, evidence of a secondary crystal phase was observed; see Figure 2b. The peaks of this phase were identified and properly refined as being of a tetragonal ZrO 2 phase (space group P42 / nmc) with network parameters significantly different from those detected for the main phase and a c/a√2 ratio of approximately 1.005, which points to a more stable unit cell, with parameters close to those described for the cubic ZrO 2 phase.

Phase composition
Similar results were reported in previous studies [28][29][30] , the secondary phase being described as a tetragonal ZrO 2 phase t´, which, although coherent with the space group P42 / nmc, exhibits a slightly deformed cationic sub-network of fluorite, with oxygen atoms displaced along the "c" axis and not in their preferred positions of the fluorite structure. Lipkin et al. 29 and Krogstad et al. 30 studied the existence of this tetragonal ZrO 2 t´ phase by high-temperature, high-resolution X-ray diffraction (synchrotron radiation) and argued that the phase is the result of regions with a high saturation of Y 2 O 3 , but which still does not have the stability to maintain the fluorite structure, characteristic of the cubic ZrO 2 phase. This phase presents a tetragonality close to that of the cubic ZrO 2 phase but at the same time has a lower Y 2 O 3 concentration as expected for the cubic ZrO 2 phase. The X-ray diffractograms of the ZrO 2 samples stabilized with 5 mol.% Y 2 O 3 and containing 0.136 wt.% Fe 2 O 3 , G5Y, sintered at different temperatures are shown in Figure 3.
Although the phase diagram points out that the composition of the G5Y sample, in the temperature range of the applied heat treatments, is located in a two-phase region, the experimental data obtained point to a three-phase system. Qualitatively, it is possible to infer that the cubic ZrO 2 phase is the major phase, coexisting with a tetragonal ZrO 2 t´´ phase, of low tetragonality, similar to that identified in the G3Y samples, and the expected tetragonal ZrO 2 phase. Figure 3b clarifies this observation, although the phases have planes at very close 2θ positions. However, in some regions, the peaks were not completely superposed, allowing for an accurate identification of the position of the peaks and the relationship between the network parameters of the phases. Tables 2 and 3 show the results of Rietveld's refinement, according to data published by Kogstad 28 .  The quantitative phase analysis of the G3Y samples points to a reduction of the content of Y 2 O 3 in solid solution of the tetragonal ZrO 2 phase from 2.43 mol.% to 2.23 mol.% with increasing sintering temperature from 1475 °C to 1600 °C . These findings are consistent with the phase transformation mechanism induced by grain contour segregation (GBSIPT) proposed by Matsui 31 . This phenomenon is also accompanied by an increase in tetragonality, expressed by an increase of the c/a√2 ratio, of the tetragonal ZrO 2 phase, due to the concomitant incorporation of anionic vacancies, and a reduction of tetragonality of the tetragonal ZrO 2 t´´ phase, as can be verified by diminishing c/a√2 ratios. On the other hand, the G5Y samples, presented a different behavior from the one identified for the G3Y samples: The Y 2 O 3 -rich, cubic ZrO 2 and t´´ tetragonal ZrO 2 phases, showed some loss of Y 2 O 3 with increasing sintering temperature, a tendency to stabilize within the analyzed limits was not identified, with a fraction in solid solution slightly higher than 5.8 mol% in samples sintered at 1600 °C. Consequently, the tetragonal ZrO 2 phase showed a slight increase in the Y 2 O 3 content, to values close to those proposed by Scott 27 for the stabilization of this phase.

Microstructural analysis
The microstructures of samples G3Y and G5Y sintered at different temperatures are shown in Figures 4 and 5.
The microstructure of G3Y samples are composed of equiaxed grains with a homogeneous size distribution and essentially pore free. With increasing sintering temperature from 1475°C to 1600°C an increase of the average grain size from 0.55 µm to 0.79 µm can be observed. Some previous studies [32][33][34] analyzed the relationship between an increase of the average grain size and the stability of the t-ZrO 2 phase, and its effect on the mechanical properties of Y-TZP. In this context, Wang and Stevens 32 reported results similar to those presented by Zhang 12 and Shahmiri et al. 35 , regarding the existence of a critical grain size and concluded that samples with an average grain size larger than 1.37 µm showed a reduction in fracture toughness of about 11% and 23.5% in   flexural strength. On the other hand, Xue et al. 34 analyzed submicrometric microstructures, reported that samples with an average grain size of 0.1 µm to 0.4 µm showed no variation of the mechanical properties, specifically of fracture toughness. Basu et al. 33 proposed that the increase in grain size can be directly associated with the reduction of free energy of the grain boundary and consequently with a reduction of the metastability of the t-ZrO 2 phase. For the authors, this change in phase stability means that, when exposed to a stress field, even of low intensity, the grains transform almost immediately and subsequently induce the transformation of neighboring grains in an autocatalytic manner, where the transformation of a new grain would be induced by the stress field created by a neighboring grain and not necessarily by the stress field of a propagating crack, resulting in the formation of a larger and more effective transformation zone.
Even so, the effective determination of a possible critical size for the spontaneous transformation of t-ZrO 2 grains is a complex issue, since it depends on multiple factors, such as grain size, Y 2 O 3 concentration and residual thermal stresses, which can be affected by the degree of anisotropy, tetragonality, or the existence of secondary phases. Quantitatively, literature points to a critical grain size close to 1 -1.2 µm 12,32,35 . Thus, within the sintering conditions employed in this study, the increase in the average grain size of the G3Y samples will not be prejudicial for the mechanical properties. As we have demonstrated in our recent publication 36 , working with the same compositional range of Fe 2 O 3 , the coloring additive did not have a significant impact in the microstructure of the ZrO 2 based ceramics. Based on this, only the SEM´s for the colored samples are showed in this manuscript.
On the other hand, samples stabilized with 5 mol.% of Y 2 O 3 , G5Y showed a significant grain growth, see Figure 5, with increasing sintering temperature from 0.62 µm to 2.32 µm, when sintering at 1475 °C or 1600 °C , respectively, representing an increase of 275%. The grain growth observed for samples of this composition, G5Y, is therefore much higher than that observed for the G3Y samples, under otherwise identical sinter conditions, see Figure 6.
Several authors studied the grain growth kinetics of the cubic ZrO 2 phase as a function of the of the Y 2 O 3 content of the starting powder 37,38 . In comparison, the authors observed a grain growth for the c-ZrO 2 phase about thirty times greater as for t-ZrO 2 , under otherwise identical processing conditions. In the present work, the analyzed micrographs of G5Y samples sintered at 1475°C indicate a homogeneous microstructure of mostly equiaxed grains with a narrow size distribution. However, for sintering temperatures above 1500 °C, a duplex microstructure, composed of a major population of large grains and a second population, of significantly smaller average size, uniformly distributed among the large grains, can be observed.
Matsui et al. 39 proposed that the precipitation of c-ZrO 2 , rich in Y 2 O 3 , in a matrix of t-ZrO 2 , with a nominal composition of 3 mol% Y 2 O 3 , occurs through the mechanism described by them as GBSIPT. The solute trawling theory is considered as the driving force of this mechanism, since the diffusion of Y 2 O 3 to the grain boundary of two tetragonal grains, leads to the saturation of this region and the nucleation of c-ZrO 2 , rich in Y 2 O 3 , may start.
The measured 3D roughness of the polished surfaces of the G3Y and G5Y samples indicate that the surface finishing of all samples was similar, regardless of the sintering conditions or compositions used. The average Ra parameter, which represents the arithmetic mean of the peaks and valleys, was 26 nm for G3Y samples and 22 nm for G5Y samples, with maximum peak and valley variations of +298 nm / -698 nm for G3Y samples and +302 nm / -278 nm for G5Y samples.

Bending strength
The results of the bending strength tests are shown in Figure 7. Specimens of G5 and G5Y, presented mean values of flexural strength in the order of 600 MPa, consistent, but close to the minimum requirement for the manufacture of dental prostheses of class 4 (σ ≥ 500 MPa), such as threeelement bridges involving molar restoration. However these ceramics are not suitable for the manufacture of class 5 dental prostheses (σ ≥ 800 MPa), such as prostheses of four or more elements. On the other hand, the specimens of G3 and G3Y, presented average values of flexural strength in the order of 1180 MPa, much higher than the requirements for class 5 prostheses.
Typically, ceramic materials exhibit fragile fractures, so failure is usually associated with the presence of a critical size defect, such as a pore or inlay at the sample surface or volume, from where crack propagation will start. Thus, it is expected that the distribution of the measured values for the rupture stress will present an asymmetric distribution, being necessary the use of a non-parametric statistical approach for the adequate understanding of the results. Figure 7b shows the Weibull distribution and failure probability regarding the dispersion of the measured values for each set of samples. Figure 7b shows that samples without Fe 2 O 3 have a higher Weibull modulus, m ≈ 13, than those containing the coloring agent, m ≈ 8. The Weibull modulus m describes the shape parameter of the distribution, i.e higher m values correspond to a sharper distribution and therefore also to a higher reliability of the material. Holz et al. 40  the authors did not consider Weibull's statistical analysis when processing the results, and there are no records in this regard. Sedda et al. 41 analyzed bleached, pre-colored and colored blocks when pre-sintered, with the use of coloring liquids. The authors also did not detect any variation in the average strength results, in the order of 1270 MPa. However, the Weibull modulus measured for the samples without color (m = 16.8) was significantly higher than that of the pre-colored samples (m = 12.5), indicating a wider spread of results for this group. This tendency is similar to the observations made in the present work.
The results of the characteristic strength σ 0 of the Weibull analysis, corresponding to the stress level at which failure probability is 63.21%, are also presented in Figure 7b.

Optical properties
The results of the optical properties are discussed according to the contrast ratio, which considers only the results computed for the luminance axis (L), and the color difference. Figure 8 shows a comparison between the contrast ratio (CR) of specimens G3 and G5, sintered at different temperatures, depending on the thickness of the analyzed specimens.

Contrast ratio
The results indicate a tendency decreasing translucency and an increase in the contrast ratio, with increasing sample thickness. This behavior is explained by the strong correlation between sample thickness and the number of possible scattering events. The formalism of this trend is explained in several models dedicated to the description of optical behaviors, as described by Equation 4 12,[42][43][44] .
where: "T" is transmittance; "R" is reflectance; " α " is absorption coefficient intrinsic to the material; "S p " is scattering coefficient due to secondary phases and porosity; "S b " is spreading coefficient due to birefringence and "t" is thickness.
Another trend observed is the relationship between the increase in sintering temperature and the reduction of the contrast ratio. This trend can be attributed to three main phenomena: reduction of residual porosity due to the increase in sintering temperature; the decrease of the interface regions, i.e. grain boundary density in the volume of the specimens, due to the increase in the average size of the grains. These results are consistent with data reported in previous studies 45,46 .
In a comparative analysis, samples G3 and G3Y showed less translucency than samples G5 and G5Y. This behavior is attributed to the predominance of the ZrO 2 -Cubic phase, isotropic, and of a phase with low tetragonality and consequently less anisotropic in the crystallographic composition of the material, tending to reduce losses due to birefringence. Figure 9 shows the results measured for the contrast ratio, in groups of samples G3Y and G5Y, sintered at different temperatures, depending on the sample thickness.
The variations of the contrast ratio, depending on the thickness and sintering temperature, of the G3 and G5 samples are similar. However, the analysis of the results also points to a loss of translucency in the samples doped with Fe 2 O 3 . Although the contrast ratio analysis considers only the variations in the luminance axis, the results indicate that, indirectly, the addition of Fe 2 O 3 had a negative impact on the contrast ratio of all samples. Another factor observed is that the addition of the coloring agent reduced the difference of the contrast ratio between samples G3Y and G5Y. While a contrast ratio difference of 9.5% can be observed between samples G3 and G5, sintered at 1600 °C, the difference between samples G3Y and G5Y, treated under the same conditions, was only 5, 5%, approximately.
Although there is no consensus as to the position occupied by Fe 2 O 3 , if dissolved in the ZrO 2 network or accumulated at grain boundaries, the scientific community agrees that the accumulation of impurities in the grain boundary causes loss of translucency, due to difference between the phase refraction indices 12,[47][48][49] .
The results of the contrast ratio also indicate that high sintering temperatures, above 1500 o C, reduces the contrast ratio. However, considering the results of porosity, in addition to the crystallographic parameters and the increase in the average grain size, it is plausible to assume that the contrast ratio is also impacted by the presence of residual porosity. Figure 10 illustrates some of the few cases in which residual porosity has been detected by SEM analysis.
For different starting materials, the pores detected were located at grain boundaries and were of an average size of less than 100 nm. According to the literature, the existence of residual porosity, even on a scale of less than 1%, significantly affects the optical properties of ceramic materials 44,50 . According to the authors, the pore size is also an important factor, with a theoretical limit of 50 nm being proposed as the threshold in which the porosity affects the beam transmittance in the visible light spectrum.

Color difference (ΔE)
The differences in the yellow shades of the specimens G3Y and G5Y sintered at different temperatures, depending on the sample thickness, are shown in Figure 11.
The results indicate to an increase in the color difference as a function of the sintering temperature and the thickness of the specimens, with the specimens with the lowest Fe 2 O 3 content showing the greatest variation. Some authors have developed studies focused on understanding the limit of human perception of color variations, within the limits of interest in restorative dentistry 49,[51][52][53] . In general, all authors detected an increase in color variation in samples submitted to a sintering cycle at a higher temperature, or for a longer time. However, none of the authors proved the mechanism responsible to cause this phenomenon. On the other hand, the literature converges regarding the perception limit of the human observer of color variation, establishing an ΔE of the order of 4 as being the perception limit, that is, color variations greater than that can be distinguished by the human eye.  Regarding these optical parameters, the G5Y samples with a thickness less than 0.75 mm and 1 mm, sintered at 1550°C or 1600°C, respectively, are above the perception limit. On the other hand, basically all G5Y samples sintered at the same temperatures are above the perception limit. Although a tendency was observed, the data obtained are insufficient to propose a mechanism to clarify this behavior.

Conclusions
The increase of the yttria content of the ZrO 2 powders from 3 mol% to 5 mol% resulted in significant changes of the phase composition and microstructure of the sintered materials. The variations in phase composition and microstructure of samples containing 3 or 5 mol% Y 2 O 3 also reflect in the mechanical properties of the materials. Specifically, bending strength, as expressed by the characteristic strength σ 0 , decreased from about 1250 MPa to about 640 MPa when increasing the Y 2 O 3 content from 3 to 5 mol%. Therefore, the ZrO 2 ceramics containing 5 mol% attend the requirements for the manufacture of dental prostheses of class 4 (σ ≥ 500 MPa), such as three-element bridges involving molar restoration, but not the manufacture of class 5 dental prostheses (σ ≥ 800 MPa), such as prostheses of four or more elements. These strength requirements are met by the ZrO 2 ceramics containing 3mol%, much higher than the requirements for class 5 prostheses. The Fe 2 O 3 addition had no relevant effect on phase composition and microstructure. However, while the characteristic strength σ 0 has not been affected by the Fe 2 O 3 addition, strength showed a larger scattering, as represented by a decreasing Weibull modulus m from 13.85 to 8.7 for samples stabilized with 3 mol% Y 2 O 3 and from 12.6 to 8.7 for samples stabilized with 5 mol% Y 2 O 3 .
Regarding the optical properties it has been found that all samples showed decreasing translucency and an increase in the contrast ratio with increasing sample thickness, while higher sinter temperatures reduced the of the contrast ratio. This reduction has been attributed to the smaller residual porosity and the reduction of grain boundaries by a larger grain growth at higher temperatures. It has also been observed that samples containing 3 mol% Y 2 O 3 were less translucent compared to samples containing 5 mol% Y 2 O 3 , with or without Fe 2 O 3 additions, under otherwise identical sinter conditions. This behavior is attributed to the presence of the cubic, isotropic ZrO 2 phase besides a phase of low tetragonality in samples containing 5 mol% Y 2 O 3, reducing losses due to birefringence. Furthermore, an increase in the color difference as a function of the sintering temperature and the thickness of the specimens was observed, with the specimens with the lowest Fe 2 O 3 content showing the greatest variation.