Ultra‐High‐Speed Sintering of 3.5Y‐TZP Zirconia: Time Reduction Without Compromising Translucency

ABSTRACT

Objectives

To assess, firstly, the impact of three sintering speeds on translucency and grain size of 3.5Y‐TZP zirconia (Katana HT) across clinically relevant thicknesses, and secondly, the effect of thermocycling on translucency.

Materials and Methods

Disc specimens were fabricated at six thicknesses (0.7–2.0 mm) and sintered using SLOW (7 h), FAST (90 min), or UHSS (18 min) protocols. Translucency (TP₀₀, TP_ab) was calculated after color measurements with a VITA Easyshade V on white and black background before and after thermocycling (n = 15). Grain size and crystalline composition were assessed was analyzed using SEM images (linear intercept). The effects of thickness, sintering, and thermocycling were quantified using a three‐way ANOVA with η2a linear mixed‐effects model with specimen as a random factor and partial η².

Results

Thickness was the primary factor determining translucency (pη ² = 0.97, p < 0.001). In addition, the sintering protocol had a statistically significant, but limited effect (pη ² = 0.360.29). FAST produced higher translucency than UHSS (p = 0.01)Sintering protocols resulted in statistically significative translucency differences, yet within the acceptability thresholds (ΔTP₀₀ < 2.627). The effects of thermocycling were negligible (pη ²  < 0.001 = 0.003). Grain size decreased with faster sintering (p < 0.05), with no SEM‐detected degradation after thermocycling. Crystalline composition was affected by sintering protocol. The effects of thermocycling were negligible for all outcomes.

Conclusions

Considering the 3.5Y‐TZP zirconia composition and the furnace specificity, translucency was predominantly influenced by thickness, while sintering speed exhibited minimal clinical impact. UHSS enables major time reduction without compromising optical performance.

Clinical Significance

Within the limitations of this in vitro study and within the tested furnace systems, ultra‐high‐speed sintering protocolUHSS allowed a substantial reduction in 3.5Y‐TZP processing time while maintaining translucency within clinically acceptable thresholds. Ultra‐high‐speed sintering protocols for 3.5Y–TZP restorations, appear to be a promising solution, reducing processing time while maintaining translucency within clinically acceptable thresholds. Restoration thickness remains the primary design parameter for achieving desired translucency.

1. Introduction

Zirconia is a cornerstone in contemporary prosthetic dentistry due to its excellent mechanical properties and biocompatibility [1, 2, 3, 4, 5, 6, 7, 8]. Historically, its high opacity was an esthetic limitation, necessitating its use primarily as a framework for porcelain veneering [9, 10, 11]. To improve translucency, manufacturers have undertaken modifications such as reducing alumina and increasing yttria (Y₂O₃) concentration [11, 12, 13, 14, 15]. These changes have made possible the use of micro‐layered or even monolithic zirconia, even in the anterior zone [4, 5, 6, 7, 8, 16, 17].

Furthermore, the translucency of zirconia is not only influenced by its composition but also by its porosities and grain size [12, 13]. Light transmission is generally enhanced by larger grain sizes, which decrease the number of grain boundaries that cause light scattering, and by low porosity [1, 18, 19]. However, it is the fabrication process that will determine both porosity and grain size.

The manufacturing of blocs or discs begins with the purification of ZrO₂ sand, followed by the incorporation of stabilizing agents and pigments, all with an impact on translucency [5, 6]. The powder is then compacted by isostatic pressing to obtain homogeneous blocks or discs that are then industrially pre‐sintered, facilitating machining. After milling, either in the laboratory or chairside, the final sintering completely densifies the material, conferring its definitive mechanical and optical properties. For the last decades, this final sintering was performed following a validated slow process, of 6 to 12 h [5, 7]. Contemporary rapid sintering protocols have reduced the final sintering time, streamlining zirconia production [20]. However, sintering parameters significantly influence translucency [12, 21, 22, 23, 24, 25], through modifications in microstructure, porosity, and phase composition [26, 27, 28, 29, 30, 31, 32].

Usually, three sintering strategies are reported:

• – Conventional sintering employs standard resistive heating via molybdenum disilicide or silicon carbide elements, with heat transfer occurring through radiation and convection. Operating with heating rates of 5°C/mi–10°C/min, this method requires 6–12 h cycles and 2–6 h dwell times at peak temperature [33].
• – Speed sintering utilizes similar resistive heating mechanisms but incorporates enhanced temperature control enabling faster ramp rates (20°C/mi–40°C/min) while maintaining thermal uniformity [23, 34]. Cycle times range around 60–120 min with 10–25 min dwell times [33].
• – Ultra high‐speed sintering (UHSS) sometimes employs electromagnetic induction to generate heat directly in the susceptor material surrounding the restoration [33], eliminating heat transfer delays and producing rapid, uniform heating [22, 35]. Maximum heating rates reach 300°C/min, with total cycle times as short as 14 min, requiring no pre‐heating [20, 36].

Drastically shortened dwell times (5–20 min) provide insufficient time for atomic diffusion to eliminate interparticle pores [12, 18, 37]. These pores scatter light due to refractive index differences between zirconia grains and air‐filled voids. Additionally, rapid heating inhibits grain fusion and growth, producing smaller grains with higher grain boundary density that scatter light [13, 38, 39, 40], consequently reducing translucency [12, 21, 22, 23, 24, 25]. Conversely, conventional sintering (6–12 h) promotes larger grain sizes, reducing grain boundary volume per unit area and enhancing light transmission. Furthermore, rapid cooling rates (up to 120°C/min) in high‐speed sintering may induce metastable yttria‐rich t′‐phase formation, which exhibits lower tetragonality than standard tetragonal phase, reducing birefringence and potentially offsetting translucency losses in some material [20, 41, 42, 43].

The speed sintering protocols affect differently the translucency depending on the zirconia disc compositions. Rapid sintering can result in inhomogeneous distribution of the sintering aids, causing light scattering due to their different refractive indices compared to the zirconia matrix [13, 39]. Furthermore, the high temperatures and rapid fluctuations during speed sintering may affect the chemical stability of the prosthesis, by uneven oxidation or volatilization of oxide‐based colorants (such as iron) used to mimic natural tooth shades [13, 39]. However, some studies have reported an increase in translucency after speed sintering [15, 44, 45, 46]. Thus, the response of zirconia to accelerated sintering is highly dependent on its generation and the specific sintering protocol employed [12, 13].

In addition, regardless of the sintering protocol, the long‐term reliability of zirconia can be affected by LTD (low‐temperature degradation), or hydrothermal aging [20, 39, 44]. LTD is the progressive surface deterioration of yttria‐stabilized zirconia caused by exposure to water at low temperatures, characterized by a tetragonal‐to‐monoclinic phase transformation of surface grains, a small volume expansion inducing residual stresses, and therefore microcracking and grain pull‐out. In fact, low yttria containing zirconia appears to be most affected by LTD, with data showing that LTD decreases zirconia's translucency [20, 39, 44].

3.5Y‐TZP is a High Translucency zirconia, a second‐generation 3.5 mol% yttria‐stabilized (3.5Y) material with reduced alumina content [12, 47]. This material maintains a high flexural strength while offering improved aesthetics [1, 13]. While different sintering protocols have been studied on various zirconia (especially 3Y‐TZP) [1, 10, 12, 13, 14, 15, 18, 19, 21, 22, 23, 24, 25, 37, 39, 41, 42, 43, 44, 45, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65], there is limited research that addresses the influence of high‐speed sintering on 3.5Y‐TZP [12, 51]. Yet, achieving optimal translucency for 3.5Y‐TZP zirconia across sintering protocols is essential for dentists and prosthodontic laboratories.

The objective of this study is to evaluate 3.5Y‐TZP translucency and grain size after accelerated sintering and aging in a new high‐speed zirconia sintering furnace (Shenpaz CS). The first research hypothesis is that thickness affects translucency. The second research hypothesis is that sintering speed affects zirconia's translucency. The third research hypothesis is that thermocycling influences translucency. The fourth research hypothesis is that the sintering protocol and thermocycling affect grain size. The fifth research hypothesis is that the sintering protocol affects phase composition.

2. Materials and Methods

2.1. Samples Preparation

The study design is represented in the graphical abstract, Figure 1. 3.5Y‐TZP zirconia specimens (HT12, Katana, Kuraray) were prepared. The sample size (n = 15 measurements per group, obtained from 3 specimens × 5 readings per specimen) was set on the basis of comparable in vitro translucency studies on dental zirconia, which have used similar or smaller numbers of specimens per condition [18, 34, 62]. This is acknowledged as a limitation. Discs were milled to obtain final samples after sintering with a diameter of 10 mm and variable thicknesses of 0.70, 0.80, 0.90, 1.00, 1.50, and 2.00 mm. The thickness range was selected to span clinically relevant scenarios for 3.5Y‐TZP restorations. The lower thicknesses correspond to minimum thickness in monolithic restorations, where the optical sensitivity of the material is greatest. Intermediate and superior values were included to address bulk and occlusal thicknesses and to enable direct comparison with the existing translucency literature [30, 33, 42, 50]. The specimen dimensions were measured using a digital caliper with 0.01 mm precision; flatness was accepted within 0.02 mm thickness range. The composition and main physicochemical properties, according to the manufacturer, are presented in Table 1.

FIGURE 1.

Graphical abstract and study design.

TABLE 1.

Main physicochemical properties of HT12 Zirconia, Katana, Kuraray (manufacturer data).

Property	Unit	Value
Composition	ZrO2 + HfO2	90%–95%
	Y2O3	5%–8%
	Other	0%–2%
Type		Monoshade
Light transmission (%: D65 φ30mm T: 1 mm)	%	45
Flexural strength (30 mm span)	MPa	1150
Fracture toughness	MPa・√m	5.1
Elastic modulus	GPa	214
Vickers hardness	Hv	1280
Density	g/cm3	6.06

Specimens were randomly distributed into three groups according to the sintering protocol:

• SLOW group: Conventional sintering, 7 h (control)

• FAST group: Speed sintering, 1 h 30 min

• UHSS group: Ultra high‐speed sintering, 18 min

A different furnace was used for the SLOW group as the manufacturer specifications for the Shenpaz furnace do not include conventional slow‐speed sintering parameters. Peak sintering temperatures were adjusted for accelerated protocols (FAST: 1530°C; UHSS: 1515°C) compared to the conventional protocol (SLOW: 1500°C) to compensate for reduced dwell times and ensure complete densification. These temperatures were proposed by the furnace manufacturer. The sintering furnace reference, sintering program, milling machine, and disc batch numbers are presented in the Table 2 and Figure 2.

TABLE 2.

Description of materials, equipment, and sintering protocols.

Sintering group	Furnace	Sintering protocol		Materials
SLOW	VITA Zyrcomat 6000MS Serial number 2120160808	Step	Description	Disc D98.5—T14mm ZR KATANA HT color HT12
		Starting temperature	200°C
		Heating rate	10°C/min
		Maximum temperature	1500°C
		Holding temperature	1500°C
		Holding time	120 min
		Cooling rate	10°C/min (down to 400°C)
		Total cycle duration	~7 h
UHSS	SHENPAZ Sintra CS software 3.13 Serial number 200–12–0823‐241	Step	Description	Disc D98.5—T26mm ZR KATANA HT color HT12
		Starting temperature	450°C
		Heating rate to 1300°C	999°C/min
		Heating rate to 1450°C	180°C/min
		Heating rate to 1515°C	26°C/min
		Holding temperature	1515°C
		Holding time	7 min
		Total cycle duration	18 min
FAST		Step	Description	Disc D98.5—T14mm ZR KATANA HT color HT12
		Starting temperature	450°C
		Heating rate to 1260°C	999°C/min
		Heating rate to 1452°C	72°C/min
		Holding temperature	1452°C
		Holding time	10 min
		Heating rate to 1530°C	99°C/min
		Final holding temperature	1530°C
		Final holding time	52 min
		Cooling rate to 350°C	−99°C/min
		Total cycle duration	~1 h 20 min

FIGURE 2.

Visual representation of the sintering protocol.

2.2. Thermocycling

Half of the samples from each group were further randomly allocated to subgroups: TC (thermocycled) and N (without TC). Samples were submitted to thermocycling using a custom‐made machine. 10,000 thermocycles were performed during two weeks at temperatures between 5°C and 55°C, with a dwell time of 30 s at each bath temperature, as reported in other literature [9]. This thermocycling regimen was intended to simulate cyclic thermal stress and does not represent a hydrothermal aging protocol for low‐temperature degradation (LTD) of yttria‐stabilized zirconia.

2.3. Translucency Parameter (TP)

The translucency of each specimen was evaluated using the translucency parameter (TP), calculated according to both the CIELab (TP_ab) and CIEDE2000 (TP₀₀) formulas. TP represents the color difference of a material of a given thickness when measured against standardized black (b) and white (w) backgrounds.

(1)

$$\begin{array}{cc}{\mathit{TP}}_{00}=\hfill & \surd \left({\left(\mathit{kL}\mathit{SL}\mathrm{\Delta }{\mathit{L}}^{\prime }\right)}^2+{\left(\mathit{kC}\mathit{SC}\mathrm{\Delta }{\mathit{C}}^{\prime }\right)}^2+{\left(\mathit{kH}\mathit{SH}\mathrm{\Delta }{\mathit{H}}^{\prime }\right)}^2\right.\hfill \\ \hfill & \left.+\mathit{RT}\left(\mathit{kC}\mathit{SC}\mathrm{\Delta }{\mathit{C}}^{\prime }\right)\left(\mathit{kH}\mathit{SH}\mathrm{\Delta }{\mathit{H}}^{\prime }\right)\right)\hfill \end{array}$$ (2)

Equations (1) and (2) TP_ab and TP₀₀ formulas.

Using the CIELab approach, TP was calculated as the color difference between measurements performed on white and black backgrounds, where ΔL* represents the difference in lightness, Δa* the difference in chromaticity along the red–green axis, and Δb* the difference in chromaticity along the blue–yellow axis. TP_ab was retained to allow direct comparison with the substantial body of dental zirconia literature in which translucency is reported as the CIELAB‐based difference between black and white backgrounds [30, 37, 59, 62]. However, the CIEDE2000 (TP₀₀), a more perceptually accurate formula, was also used [37, 59].

In the TP₀₀ formula, ΔL′, ΔC′, and ΔH′ are the differences in lightness, chroma, and hue, respectively. The parametric factors kL, kC, and kH are correction terms for experimental conditions and were set to 1. SL, SC, and SH are weighting functions that adjust the total color difference considering the location variation of the color difference pair in L′, a′, and b′ coordinates. Finally, the parameter RT is a rotation function accounting for the interaction between hue and chroma in the blue region [66, 67, 68].

Color coordinates were recorded using a contact digital spectrophotometer (VITA EasyShade V; VITA Zahnfabrik, Bad Säckingen, Germany) by a single operator under constant lighting conditions, with a repositioning jig to ensure a perpendicular position of the spectrometer upon the sample. Five measurements were taken across five different areas of each disc, resulting in n = 15 measures per group.

The VITA EasyShade V exhibits high repeatability [69, 70] and operates with a 5 mm measurement area, a spectral range of 400–700 nm, standard illuminant D65, d0/0° illumination geometry, and a 2° standard colorimetric observer. Prior to measurements, the device was charged for 12 h and allowed to warm up for 10 min, in accordance with the manufacturer's recommendations. Calibration on the VITA EasyShade incorporated deck was performed before each measurement.

To ensure optical contact between the ceramic specimen and the substrate, a neutral try‐in paste (Variolink Esthetic Try‐In, neutral shade; Ivoclar Vivadent, Schaan, Liechtenstein) was applied over the entire substrate surface, as previously reported in the literature [71]. The neutral try‐in paste was selected instead of a saccharose or glycerol solution to better reproduce clinical conditions.

During measurements, the substrate–ceramic assembly was positioned inside a white acrylonitrile butadiene styrene (ABS) well (Zortrax Z‐ABS White; Zortrax, Olsztyn, Poland) with 2 mm thick walls (CIE L = 89.7, CIE C = 10.1, CIE h = 90.3).

The CIE L*, C*, h°, a*, and b* color coordinates of all substrate–specimen combinations were retrieved using the VITA Assist App (VITA Zahnfabrik, Bad Säckingen, Germany) and recorded for each measurement. Calculations were performed using MATLAB software (R2023b; MathWorks, USA), ensuring correct parameter implementation as recommended [66].

2.4. SEM Analysis

Three samples for each material were randomly selected. They were mounted on SEM stubs and sputter‐coated with gold–palladium (20/80), then analyzed using an SEM (Quanta, FEI Company, Eindhoven, The Netherlands, 10 kV) with a working distance of 10 mm as performed in previous studies [72].

Grain size was determined using the linear intercept method applied to five scanning electron microscopy (SEM) micrographs. High‐magnification SEM images were acquired from representative areas of each specimen, avoiding edge regions. Image analysis was performed using ImageJ software (National Institutes of Health, Bethesda, MD, USA). For each micrograph, a grid of straight, randomly oriented test lines was superimposed, and the number of grain boundary intercepts along each line was counted. The mean linear intercept length was calculated by dividing the total length of the test lines by the total number of grain boundary intersections [73], n = 1000 intercepts were counted.

2.5. X‐Ray Diffraction

Crystalline phase composition of the sintered specimens was characterized by X‐ray diffraction (XRD). Diffraction patterns were recorded on a Bruker D8 Discover diffractometer in Bragg–Brentano geometry, equipped with a sealed copper X‐ray tube (Cu Kα1, λ = 1.5406 Å), an incident‐beam quartz monochromator and an energy‐resolving LynxEye XE‐T linear detector. Each specimen was scanned from 15° to 100° 2θ with a step size of 0.02°. Patterns were refined by the Rietveld method using the FullProf program [74]. Three zirconia polymorphs were identified against the ICDD PDF database: tetragonal ZrO2 (P42/nmc; ICDD 01‐079‐1764), cubic ZrO2 (Fm‐3 m; ICDD 00‐049‐1642) and monoclinic ZrO2 (P21/c; ICDD 04‐013‐6875). The refined variables were the background (seven interpolated points), the specimen‐displacement offset and, for each phase, the unit‐cell parameters, scale factor and crystallite size; phase abundances were expressed as weight fractions (wt%). Refinement quality was assessed from the reduced chi‐square (χ ²), the profile and weighted‐profile factors (Rp, Rwp) and the per‐phase Bragg factor (Rbragg). X‐ray diffraction was performed on two specimens per sintering × thermocycling condition (n = 2).

2.6. Statistical Analysis

Data were analyzed with Matlab Software (Matlab R2024b, MathWorks, Portola Valley, USA). A linear mixed‐effects model—fitted to the log‐transformed translucency parameter, with sintering, thermocycling and thickness as fixed factors and specimen as a random factor to account for the repeated readings—was used as the primary analysis; descriptive statistics and figures are reported in the original (untransformed) units. Type III tests were obtained, followed by Tukey's HSD post hoc tests and simple‐effects testing with multiplicity control; alpha value was set at 0.05.

Mean differences were calculated to quantify the direction and magnitude of color variation between experimental groups. For TP₀₀, pairwise comparisons among sintering protocols were performed within each thickness level for non‐thermocycled specimens, including SLOW versus FAST, SLOW versus UHSS, and FAST versus UHSS. In addition, the effect of thermocycling was assessed by computing the contrast TC—N_TC within each sintering protocol and thickness level. These comparisons were based on Welch independent‐samples tests, and the effect size was expressed as the difference in means. Welch's method was preferred for these follow‐up comparisons because it is robust to heterogeneity of variance between groups. For each subgroup, 95% confidence intervals and two‐sided p values were calculated using the Welch–Satterthwaite approximation for degrees of freedom. The sign of the mean difference was retained to indicate directionality of the effect. These data were summarized numerically and displayed graphically using forest plots.

An evidence‐based clinical interpretation of translucency differences among zirconia sintered under different conditions was applied using established 50:50 perceptibility (TPT) and 50:50 acceptability thresholds (TAT) [75]. For the CIEDE2000 formula, TPT and TAT values were set at 0.62 and 2.62, respectively [75]. Mean differences below the perceptibility threshold were interpreted as not visually perceptible, those between perceptibility and acceptability thresholds as perceptible but clinically acceptable, and those beyond the acceptability threshold as potentially clinically unacceptable.

3. Results

3.1. Translucency Parameter

TP₀₀ values are presented as means and standard deviations (Table S1) and in box‐and‐whisker plot format (Figure 3).

FIGURE 3.

TP₀₀ for the different thickness groups based on sintering protocol and thermocycling group.

The factorial effects were tested with a linear mixed‐effects model on the log‐transformed translucency parameter, with specimen included as a random factor to account for the repeated readings; there was a significant main effect of Sintering, χ ²(2) = 162.1, p < 0.001, partial η ² = 0.29. There was also a very strong main effect of Thickness, χ ²(5) = 11,810, p < 0.001, partial η ² = 0.97. In contrast, the main effect of Thermocycling was not significant, χ ²(1) = 1.62, p = 0.20, partial η ² = 0.003.

Several interaction effects were also statistically significant: the Sintering × Thickness interaction, χ ²(10) = 50.7, p < 0.001, partial η ² = 0.12; the Thickness × Thermocycling interaction, χ ²(5) = 22.1, p < 0.001, partial η ² = 0.04, indicating that the thermocycling effect varied with thickness. The three‐way interaction among Sintering, Thickness, and Thermocycling was significant, χ ²(10) = 57.0, p < 0.001, partial η ² = 0.10. The Sintering × Thermocycling interaction was not significant, χ ²(2) = 3.26, p = 0.20, partial η ² = 0.006.

Post hoc Tukey for marginal contrasts comparisons for Sintering showed that all sintering groups presented statistically significant differences. For Thermocycling, the comparison between N_TC and TC was not significant (p = 0.91). For Thickness, the pairwise contrasts showed widespread significant differences, with lower thickness values generally associated with higher response values, reflecting the strong monotonic decrease in the translucency as thickness increased.

Overall, the results indicate that Thickness was the dominant factor influencing the response, with an additional contribution from Sintering, while Thermocycling did not show a meaningful overall main effect.

TP_ab values are presented as means and standard deviations (Table S2) and in box‐and‐whisker plot format (Figure S1).

As shown in Figure 4 and Table S4, the mean difference sintering protocols for the N_TC group revealed that SLOW versus FAST produced the smallest mean differences overall, with values consistently close to zero. In contrast, comparisons involving UHSS showed greater divergence, particularly FAST versus UHSS, which yielded the highest positive mean differences at several thicknesses. SLOW versus UHSS showed intermediate variability, with both near‐zero and positive differences depending on thickness. Overall, the effect of sintering protocol on TP₀₀ was thickness‐dependent and more pronounced in thinner specimens, although most differences remained within the perceptibility threshold.

FIGURE 4.

Pairwise comparisons of mean TP₀₀ differences among the SLOW FAST AND UHSS sintering protocols at each specimen thickness for non‐thermocycled samples. Points represent the estimated mean differences for each protocol comparison, with horizontal error bars showing the associated 95% confidence interval. The shaded background indicates translucency threshold regions, green—below the perceptibility threshold, yellow—within the acceptability threshold, and red—above the acceptability threshold.

As shown in Figure 5 and Table S3, the mean difference TP₀₀ between N_TC and TC groups was generally small across all sintering protocols and thicknesses, with most values clustered around zero and remaining within the translucency perceptibility threshold. The largest deviations were observed in the thinnest specimens, whereas thicker specimens showed smaller differences. Among the protocols, UHSS exhibited the widest fluctuation across thicknesses, while FAST remained the most stable and SLOW showed intermediate behavior.

FIGURE 5.

Mean difference in color change TP₀₀ between non‐thermocycled (N_TC) and thermocycled (TC) measurements for the SLOW FAST AND UHSS sintering protocols at each specimen thickness. Symbols represent the mean differences, and horizontal error bars indicate the corresponding 95% confidence interval. The shaded background indicates translucency threshold regions, green—below the perceptibility threshold, yellow—within the acceptability threshold, and red—above the acceptability threshold.

3.2. SEM Analysis

SEM surface micrographs (Figure 6) revealed marked differences in microstructure as a function of sintering protocol, whereas thermocycling induced only minor changes. SLOW‐sintered specimens (Column 1) exhibited a relatively coarse and heterogeneous grain structure, with well‐defined grain boundaries and a broader grain size distribution observed both in non‐thermocycled (Row 1) and thermocycled conditions (Row 2). FAST‐sintered specimens (Column 2) displayed a more homogeneous and refined microstructure, characterized by smaller, more uniformly distributed grains and reduced intergranular contrast compared with SLOW specimens. In contrast, UHSS specimens (Column 3) showed the finest and most densely packed microstructure, with markedly reduced grain size and a smoother, more compact surface appearance. Yet, several voids were visible on the surface. Comparison between non‐thermocycled and thermocycled rows revealed no evident microstructural degradation, grain coarsening, or surface damage attributable to thermocycling for any sintering protocol.

FIGURE 6.

Representative SEM images of zirconia surfaces after different sintering protocols (SLOW, first column; FAST, second column; UHSS, third column) before (upper row) and after (lower row) thermocycling. When the sintering time is reduced, smaller grains are observed.

3.3. Grain Size

As reported in detail in Table 3, sintering protocol significantly affected the grain size (p < 0.05), with SLOW specimens showing the highest values, followed by FAST and UHSS specimens exhibiting the lowest values, while thermocycling did not induce significant differences within any sintering condition (p > 0.05).

TABLE 3.

Mean and standard deviation of grain size according to sintering protocols and thermocycling groups.

Group	Mean (nm)	Std. deviation
FAST_N	475	14	Aa
FAST_TC	510	24	Aa
SLOW_N	904	14	Bb
SLOW_TC	918	256	Bb
UHSS_N	319	8	Cc
UHSS_TC	339	15	Cc

Note: Different superscripts indicate significantly different groups (p < 0.05): upper cases for TC and lower cases for sintering conditions.

3.4. X‐Ray Diffraction

Two specimens per group were characterized by X‐ray diffraction and analyzed by Rietveld refinement against a three‐phase model: cubic ZrO2 (ICDD 00–049‐1642), tetragonal ZrO2 (ICDD 01–079‐1764), and monoclinic ZrO2 (ICDD 04–013‐6875). Refinements converged for every specimen with χ2 between 1.58 and 1.98. Full per‐specimen goodness‐of‐fit indices, a representative XDR pattern, refined lattice parameters and crystallite sizes are reported in Figure S2 and Tables S5 and S6.

Phase weight fractions varied monotonically with sintering schedule. The cubic ZrO2 fraction increased from 34.0% (SLOW) to 38.5% (FAST) and 46.0% (UHSS), with the tetragonal fraction decreasing correspondingly from 61.0% to 57.0% and 52.5%. Monoclinic content showed the same direction of change with progressively shorter cycles, declining from 5.0% (SLOW) to 4.1% (FAST) and 1.7% (UHSS).

Crystallite sizes derived from X‐ray peak broadening decreased with shorter total sintering cycle for both the tetragonal and monoclinic phases. Mean tetragonal‐phase crystallite size decreased from 1106 Å (SLOW) to 802 Å (FAST) and 740 Å (UHSS), and monoclinic‐phase crystallite size from 285 Å to 222 Å and 191 Å, respectively. Cubic‐phase crystallite size showed substantial within‐group dispersion and no monotonic trend at this sample size (group means 662, 645 and 669 Å).

No consistent difference in phase composition was observed between thermocycled and non‐thermocycled specimens.

Given the two specimens analyzed per condition, these findings are reported as observed trends and were not subjected to statistical analysis.

4. Discussion

This study found that thickness was the dominant determinant of translucency in 3.5Y‐TZP (Katana HT, Kuraray Noritake) zirconia, with sintering speed (Shenpaz Sintra CS) exerting a statistically significant but clinically limited secondary effect.

The first research hypothesis was that thickness affects translucency. In the current study, the translucency parameter decreased monotonically with increasing thickness, in accordance with the Beer–Lambert law, as thicker specimens provide more opportunities for light scattering at grain boundaries, phase interfaces, and residual pores [30, 34, 39]. Thus, the first hypothesis was accepted. This finding has critical clinical implications: restoration design is the primary parameter for achieving desired translucency in 3.5Y‐TZP zirconia, while processing parameters (sintering protocol, thermocycling) exert only secondary influences.

The second research hypothesis was that sintering speed affects zirconia's translucency and it was partially accepted, as there was a statistically significant difference for the translucency between the sintering groups. However, the clinical relevance of these differences must be evaluated against established perceptibility and acceptability thresholds. Thus, although the differences in translucency were statistically significant, their clinical effect was clinically limited, as most translucency differences were within perceptibility thresholds, and all differences within clinically acceptable thresholds. Statistically significant interactions between thickness and sintering protocol indicated that sintering effects are thickness‐dependent, with more variation for thinner samples. However, the small effect size of this interaction indicated that the main effects of thickness and not sintering provide adequate predictive models for most clinical applications. Nonetheless, because there was not a clear trend of protocol‐related differences in thinner specimens, high‐speed sintering should be avoided in very thin restorations. In these cases, minor TP differences could approach acceptability thresholds and may necessitate adjustments in staining, glazing, or cement shade to maintain esthetic consistency. Studies using prosthesis‐shaped specimens at different clinically relevant thicknesses could further clarify these observations.

These findings are in agreement with previous research [30, 34, 39, 57]. Several studies have found that sintering speed changes translucency, but within acceptability thresholds [23, 34, 57]. However, another study observed that speed and high‐speed sintering altered translucency, yielding considerably lower TP_ab values for 1 mm thickness 4Y‐PSZ specimens [12]. Other studies using speed sintering protocols across 4Y‐PSZ zirconia brands showed limited TP_ab changes, under the perceptibility thresholds, for samples with thickness from 0.5 to 1.2 mm [34, 39, 57]. Another study on 4Y‐PSZ zirconia found changes below the acceptability threshold comparing TP₀₀ for 1.5 mm samples with conventional or speed sintering protocols [15]. Similarly, for 3Y‐TZP specimens, high speed or speed sintering were found to considerably alter translucency [12, 44]. These contrasting findings can be explained by the zirconia type and brand and furnace type used in the present study.

The third research hypothesis was that thermocycling influences translucency and it was rejected. In the present study, hydrothermal aging susceptibility was assessed through thermocycling. The effect of aging was not statistically significant and pairwise comparisons between thermocycled and non‐thermocycled groups showed no significant differences. Similar protocols, with a reduced number of thermo‐cycles have been previously used [10, 44, 51, 60]. However, literature reports show that TC reduces translucency [44, 54, 59], within acceptability thresholds, contrasting with the present findings. Thus, in this study, the aging results should be interpreted conservatively, as the applied thermocycling regimen may not have been sufficiently aggressive to reproduce clinically relevant LTD‐related transformation phenomena for this material within the tested timeframe. More explicitly, the 10,000‐cycle thermocycling regimen used here cannot be interpreted as a hydrothermal aging or LTD protocol. Validated LTD studies for yttria‐stabilized zirconia rely on autoclaving in steam at 134°C and 2 bar for several hours, conditions that drive measurable tetragonal‐to‐monoclinic transformation. Future investigations should therefore include accelerated hydrothermal aging.

The fourth research hypothesis was that sintering protocols and thermocycling alter zirconia grain size. In this study, SEM observations and grain size measurement have shown a significant grain size difference, with smaller grains when the sintering time was reduced, yet no alterations related to thermocycling. Thus, the fourth research hypothesis was partially accepted. These findings align with several studies in the literature reporting significant grain size reduction after accelerated sintering [13, 28, 47]. This grain refinement likely relates to the limited time available for grain growth at peak temperature in accelerated protocols, highlighting the material‐specific and protocol‐dependent nature of this phenomenon. In contrast, contradictory findings exist in the literature. No significant impact of speed sintering on grain size has been reported in another study [1], whereas studies on 5Y‐PSZ found grain sizes coarsening after speed and high‐speed sintering protocols [18, 20]. These disparities likely reflect differences in furnace technology, peak temperatures, heating rates, and intrinsic material characteristics such as yttria content and starting powder morphology.

However, while grain refinement with accelerated sintering was observed, grain size alone cannot fully explain translucency changes in zirconia, which are also influenced by residual nanoscale porosity/densification and phase constitution (t, t′, c and potential m fraction after aging) [12]. The observation that UHSS exhibited smaller grains yet only modest TP differences suggests that multiple competing mechanisms are involved. Consequently, complementary analyzes such as polished cross‐sectional porosity quantification could strengthen the interpretation. In addition, the SEM analysis itself was primarily qualitative, based on a limited number of specimens and micrographs per condition. This is acknowledged as a methodological limitation, and a larger SEM dataset would be needed to fully support mechanistic claims.

The fifth research hypothesis was that sintering affects phase composition, and it was accepted, as cubic and tetragonal phase composition shifted with sintering speed. Phase composition indicated that the cubic phase accounted for approximately 34–46 wt%, intermediate between the values typically reported for 3Y‐TZP (8–20 wt%) and 5Y‐PSZ (50 wt%) [12, 20]. In this study, the refinement used a three‐phase cubic + tetragonal + monoclinic model. However, recent work indicates that the component indexed as cubic in 4–6 mol% yttria zirconia is more accurately described as a metastable, yttria‐rich tetragonal‐prime (t′) phase that is difficult to resolve from true cubic by X‐ray diffraction [18, 29, 41, 42, 60]. A proportion of the cubic content quantified here, and its apparent increase after UHSS, may therefore correspond to t′ formed by the rapid cooling of the accelerated cycle.

In this study, UHSS was associated with a higher cubic and a lower monoclinic content than SLOW or FAST sintering. This direction of change is in agreement with reports in which accelerated or higher‐temperature schedules raise the cubic or t′ fraction [14, 27, 43, 76], but it contrasts with studies in which speed sintering reduces the cubic content or promotes its conversion to t′ [34, 42], and with others reporting no measurable phase change [20, 77]. This heterogeneity most plausibly reflects differences in furnace technology, peak temperature and cooling rate between studies, together with the modeling decision to resolve, or not to resolve, the tetragonal‐prime phase.

The clinical relevance of these phase differences lies in their predicted optical consequences. Cubic zirconia is optically isotropic, so boundaries between cubic grains do not generate birefringence‐driven light scattering, whereas the monoclinic phase introduces additional scattering through refractive‐index mismatch and the surface disruption that accompanies its formation [6, 78]. On the basis of phase composition alone, the UHSS group, with the highest cubic and the lowest monoclinic content, would be expected to be the most translucent. This was not borne out by the translucency measurements, in which UHSS did not produce the highest TP₀₀. Therefore, phase composition does not account for the observed translucency ranking. The most coherent interpretation is that translucency in this material is governed by several competing scattering mechanisms, and that the favorable phase composition of the UHSS specimens is offset by a competing factor. This factor could be residual porosity introduced by the short sintering cycle, since increased residual porosity has been identified as a principal cause of reduced translucency after high‐speed sintering [18]. Thus, the limited effect of sintering speed on translucency therefore cannot be ascribed to crystalline phase composition.

The absence of a consistent phase change between thermocycled and non‐thermocycled specimens is consistent with thermocycling being a milder hydrothermal stress than autoclave low‐temperature degradation [79, 80], and parallels the negligible effect of thermocycling on translucency observed in the present study.

Several limitations qualify this phase analysis. It was performed on two specimens per condition and is therefore descriptive rather than statistically powered. Furthermore, the three‐phase model does not resolve the tetragonal/tetragonal‐prime distinction, so the cubic fraction reported here is best regarded as an upper estimate of the true cubic content. Lastly, the elevated monoclinic Bragg factor indicates that the small monoclinic fractions are semi‐quantitative, with a possible contribution of machining‐induced surface transformation [52].

The novelty of this research resides in the furnace used for the speed sintering protocols and in the zirconia used. Firstly, to the best current knowledge, the Shenpaz Sintra CS has not previously been studied for its performance on zirconia speed sintering. According to the manufacturer, this is a conventional resistance‐heat furnace capable of achieving high‐speed heating ramps. Another high‐speed furnace, scarcely described in the literature [18, 20, 36, 37], is the Dentsply SpeedFire furnace. This furnace operates with an electromagnetic induction‐based heating system and has been described to enable single‐unit cycles of ~15–30 min and chairside glazing/crystallization, supporting one‐visit zirconia workflows. Mechanical, geometrical and optical changes have been observed after sintering in the Speed Fire furnace [18, 20, 36], with lower mechanical properties, a reduced thickness‐dependent translucency, yet better marginal and occlusal fit [18, 20, 36]. Interestingly, 5Y‐PSZ zirconia has proved to be least reliable after high‐speed sintering, encouraging final users to apply high‐speed sintering protocols only on validated zirconia [18, 20, 36].

Secondly, the KATANA HT12 zirconia, used in the present study, is innovative by its composition. While categorized as 4Y‐PSZ (partially stabilized zirconia) [48, 57], one source refers to it as the monolayer equivalent to certain 3Y‐TZP materials [60]. This confusion comes from the KATANA HT yttrium composition. Indeed, the manufacturer specifies 5%–8% wt% of yttrium oxides, and claims this zirconia to be designed as a high‐translucency material for monolithic restorations, offering higher translucency than first‐generation zirconia. It is this specific weight composition that makes KATANA HT difficult to classify. Usually 3Y‐TZP zirconia contains 4%–6% wt% of Yttrium, while 4Y‐PSZ contains 6–8 wt%¹¹. Thus, KATANA HT cannot be classified in either 3Y or 4Y, but more as a 3.5Y. Research comparing different zirconia brands concluded that the translucency of Katana zirconia remained unaffected by accelerated sintering protocols [29, 34] and presented better values. These observations were linked to Katana purity, homogeneous yttria distribution and suppressed yttria segregation during rapid thermal cycles [29, 34], leading to consistent phase composition, even after speed sintering. In a recent 2025 study, a 3.5Y zirconia, manufactured from Tosoh zirconia powder, has been proposed as a better suit for high‐speed sintering protocols to provide an optimal balance of high strength and high translucency [37]. The study subjected the 3.5‐YSZ zirconia to an 18 min high‐speed sintering protocol (1560°C with a 5 min dwell) and found it was 22% more translucent than 4YSZ and 73% more translucent than 3YSZ with similar high speed sintering protocols [37]. However, this study did not evaluate the conventional sintering protocol. Therefore, this study aimed to address this gap, for the Katana zirconia.

For these findings, a methodological limitation concerns the selection of measurement instrumentation. While the VITA EasyShade V is designed primarily as an accurate and precise [69, 70] dental office spectrophotometer, its application in this study under controlled lighting conditions within a standardized experimental setting represents an appropriate methodological choice [12, 37, 54, 62]. Importantly, the device was employed for calculating differential parameters derived from two distinct measurements rather than for direct reporting of LCH values.

Other methodological limitations warrant acknowledgment in this research. First, different furnaces were employed for conventional and accelerated sintering protocols. This introduces potential confounding variables related to furnace‐specific heating characteristics and temperature control precision [65]. This is a major limitation of the present design: because the SLOW (conventional) protocol was carried out in a VITA Zyrcomat 6000MS while the FAST and UHSS protocols were carried out in a Shenpaz Sintra CS, the effect of sintering speed cannot be isolated from furnace design. Differences in heating element technology, thermal mass, atmosphere control, and effective temperature uniformity directly influence densification, grain growth, residual porosity, and phase distribution—all of which determine optical behavior. The findings of this study should therefore be interpreted as describing the combined effect of sintering schedule and furnace, within the tested furnace systems, rather than as an independent effect of sintering speed.

Second, consistent with previous investigations [26, 27, 65], the sintering temperature for accelerated protocols exceeded that of conventional sintering. For 3Y and 4Y zirconia, elevated sintering temperatures are typically employed to compensate for reduced dwell times, theoretically achieving comparable mechanical and optical properties. However, this methodological approach inherently confounds temperature and time variables, limiting the ability to isolate the independent effects of each parameter.

Third, the study examined only a single brand and type of zirconia material, constraining the interpretation of these findings across the commercially available zirconia formulations. Given documented variations in composition, grain size, and manufacturing processes among different brands [20, 26], extrapolation of these results is only possible to zirconia of similar composition, in a similar furnace. Further research examining this particular furnace system is warranted, as current literature remains limited.

All conclusions are specific to Katana HT 3.5Y‐TZP sintered in VITA Zyrcomat 6000MS (conventional) or Shenpaz Sintra CS (accelerated) furnaces. Generalizability to other materials and equipment requires validation. Ultimately, clinical adoption of ultra‐high‐speed sintering is contingent on confirmation that mechanical properties are not compromised.

5. Conclusions

Within the limitations of the current study, it was concluded that:

1. Thickness affected translucency. For 3.5 mol% Y₂O₃‐stabilized zirconia (Katana HT), specimen thickness was the dominant determinant of translucency.

2. Sintering speed of Katana HT zirconia in the Shenpaz CS furnace affected translucency, but the effects were clinically modest. All differences between ultra‐high‐speed (18 min) and conventional (7 h) sintering, between fast (90 min) and conventional, and between fast and ultra‐high‐speed were within clinically acceptable thresholds (ΔTP₀₀ < 2.62). Thus, within the limitations of this in vitro study and within the tested furnace systems, ultra‐high‐speed sintering appeared to enable a substantial reduction in processing time while maintaining translucency within clinically acceptable thresholds. Direct attribution of this outcome to sintering speed alone is not possible because the SLOW protocol was performed in a different furnace from the FAST and UHSS protocols.

3. Thermocycling did not affect translucency. Thermal cycling (10,000 cycles, 5°C–55°C) produced negligible translucency changes and no detectable microstructural degradation, confirming translucency stability under simulated thermal stress.

4. Accelerated sintering produced clear grain refinement relative to conventional sintering.

5. Shortening the sintering cycle changes the crystalline phase composition of this 3.5Y‐TZP.

Author Contributions

Tatiana Roman: conceptualization, methodology, investigation, formal analysis, data curation, writing – original draft, writing – review and editing, visualization, project administration. Olivier Etienne: conceptualization, methodology, investigation, formal analysis, resources, data curation, writing – original draft, writing – review and editing, visualization, supervision, project administration. Hamdi Jmal: methodology, writing – review and editing. Adrien Greffet: data curation, visualization, writing – review and editing. Teissir Ben Ammar: investigation, data curation, writing – review and editing. Marc Lenertz: investigation, data curation, writing – review and editing. All authors approved the final manuscript.

Funding

The authors have nothing to report.

Ethics Statement

This study was conducted entirely in vitro using commercially available dental materials. No human subjects, human tissues, or animal subjects were involved. Therefore, ethics committee approval was not required.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Table S1: Mean and standard deviation of TP₀₀ for all experimental groups, by sintering protocol (SLOW, FAST, UHSS), thermocycling condition (TC, thermocycled; N_TC, non‐thermocycled) and specimen thickness (0.70–2.00 mm). SD, standard deviation; TP₀₀, translucency parameter based on the CIEDE2000 color‐difference formula.

Figure S1: Box‐and‐whisker plots of TP_ab for each thickness group (0.70–2.00 mm), by sintering protocol (SLOW, FAST, UHSS) and thermocycling condition (TC, thermocycled; N_TC, non‐thermocycled). TP_ab, translucency parameter based on the CIE 1976 (ΔE*ab) color‐difference formula.

TABLE S2: Mean and standard deviation of TP_ab for all experimental groups, by sintering protocol (SLOW, FAST, UHSS), thermocycling condition (TC, thermocycled; N_TC, non‐thermocycled) and specimen thickness (0.70–2.00 mm). SD, standard deviation; TP_ab, translucency parameter based on the CIE 1976 (ΔE*ab) color‐difference formula.

Table S3: Effect of thermocycling on TP₀₀: mean difference (TC minus N_TC) with 95% confidence interval and p value, by specimen thickness and sintering protocol. CI, confidence interval; TC, thermocycled; N_TC, non‐thermocycled.

TABLE S4: Pairwise differences in TP₀₀ between sintering protocols (SLOW vs. FAST, SLOW vs. UHSS, FAST vs. UHSS) within the non‐thermocycled (N_TC) group, with 95% confidence interval and p value, by specimen thickness. CI, confidence interval.

Figure S2: Representative X‐ray diffraction (XRD) pattern of a SLOW, thermocycled (TC) specimen. The reflections are indexed to three phases of zirconium oxide (cubic, tetragonal and monoclinic); principal Bragg peaks are labeled and all Bragg positions are indicated by vertical tick marks below the pattern.

TABLE S5: Rietveld‐refinement goodness‐of‐fit indices and phase weight fractions for the non‐thermocycled specimens, expressed as the arithmetic mean of n = 2 specimens per sintering regime. Per‐specimen values and refinement uncertainties are available in the underlying dataset.

Table S6: Refined lattice parameters, unit‐cell volumes, crystallite sizes and sample‐height (alignment) parameter for the same regimes, expressed as the arithmetic mean of n = 2 specimens per sintering regime. Crystallite sizes are XRD coherent‐domain estimates from peak broadening and are not equivalent to SEM or EBSD grain sizes.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.