The objective of the present work was to study the curvature of very thinly, veneered Y-TZP discs of different framework thicknesses submitted to different firing times.
Fifteen 20-mm-wide Y-TZP discs were produced in three different thicknesses: 0.75, 1, 1.5 mm. One disc from each group was left unveneered while the others were layered with a 0.1 mm veneering ceramic layer. All discs underwent five firing cycles for a total cumulative firing time of 30 min, 1, 2, 5 and 10 h at 900 °C. The curvature profile was measured using a profilometer after the veneering process and after each firing cycle respectively. A fitted curve was then used to estimate the, curvature radius. The coefficient of thermal expansion (CTE) measurements were taken on veneering, ceramic and Y-TZP beam samples that underwent the same firing schedule. Those data were used to calculate the curvature generated by CTE variations over firing time.
All bilayered samples exhibited a curvature that increased over firing time inversely to framework thickness. However non-veneered samples did not exhibit any curvature modification.
The results of the present study reveal that even a very thin veneer layer (0.1 mm) can induce a significant curvature of Y-TZP discs. The dilatometric results showed that Tg and CTE, variations are not sufficient to explain this curvature. A chemical-induced zirconia volume, augmentation located at the framework sub-surface near the interface could explain the sample, curvature and its increase with firing time.
About 10 years ago, Yttria-tetragonal-zirconia-polycrystal (Y-TZP) was introduced to replace metal as a framework material, giving better aesthetics and biocompatibility for dental crowns and fixed partial dentures (FPDs).
The Y-TZP phase transformation potential gives it a high strength and fracture toughness. Those mechanical properties compare well with those obtained for other ceramic framework materials, such as glass ceramics containing leucite or lithium disilicate crystals, or glass-infiltrated ceramics containing spinel, alumina or alumina/zirconia.
However, the clinical reports for Y-TZP based crowns and FPDs reveal a higher short-term failure rate when compared to the classic porcelain-fused-to-metal (PFM) restorations, and cohesive fracture of the veneering layer (“chipping”) is reported as the first complication . Reasons underlying these fractures are complex and remain unidentified.
Originally veneering ceramics for zirconia-based restorations were engineered with the PFM concept in mind. The veneering ceramics were adapted to zirconia frameworks performing Coefficient of Linear Thermal Expansion (CTE) measurements and thermal shock testing . The CTE mismatch between the bulk materials was designed to mimic that of the ceramo-metal CTE. The CTE of veneering ceramics was modified with the goal of achieving a lower value than that of the zirconia. Based on the principle that compressive stress improves the mechanical behavior of the veneering ceramic, this approach was intended to develop residual compressive stress within the veneer during the cooling process . Certain manufacturers proposed a slow cooling procedure to reduce fracturing in the veneer. Residual stress profiles were measured in veneering ceramic layered either from metal- or zirconia-disk frameworks . On the one hand, this experiment demonstrates the positive effect of tempering and the benefits of CTE mismatch on stress in the veneering ceramic layered on metal frameworks. On the other hand, this procedure has proven that zirconia-based samples behave differently than metal-based samples, which may be problematic. As expected metal-based samples exhibited exclusively compressive stress, either in the surface or in the depth of the veneer layer. However zirconia-based samples often exhibited tensile stress in the interior of the veneering ceramic layer in contact with the framework. The presence of interior tensile stress was related to the slow cooling rate and to high veneer/framework thickness ratio. It was hypothesized that the t–m phase transformation following the veneer firing process caused a volume increase of zirconia grains at the interface, which may explain these findings . The possible occurrence of a t–m transformation of zirconia during the veneering process was detected by SEM and X-Ray Diffraction (XRD) . Meanwhile a confocal Raman microscopic analysis of the interface did not reveal the presence of monoclinic phase but rather highlighted a 2 μm-deep silicon–zirconium inter-diffusion zone. Recently, a FIB nano-tomography study showed sub-surface structural changes in the zirconia layer. This undefined alteration located in the first micrometer of the zirconia surface could be the consequence of a diffusion phenomenon, as this behavior was observed in a sample that was cooled very slowly from 900 °C to room temperature with a 2 °C/min cooling rate . In fact, very slow cooling may intensify those changes, as a high temperature and time increase the diffusion process .
The objective of the present work was to study the influence of firing time and framework thickness on the curvature of Y-TZP discs layered with a very thin veneer layer (0.1 mm). This veneer thickness was intended to reduce the impact of residual stress development within the veneer layer on zirconia framework. Dilatometric analyses were also performed in order to compute the influence of Tg and CTE variations during firing processes. The research hypothesis was that if a curvature of the samples occurs, it can only be related to a volume increase in the zirconia framework sub-surface. Additionally, a variation of curvature with firing time could suggest the occurrence of diffusion phenomena. In the present study, SEM imaging was added to observe the veneer–zirconia interface.
Materials and methods
Discs samples composed of veneering ceramic sintered on Y-TZP (VITA In-Ceram YZ, VITA Zahnfabrik, Bad Säckingen, Germany) frameworks were manufactured following standard dental laboratory procedures and manufacturer’s recommendations. Pre-sintered Y-TZP discs suitable for CAD/CAM machines were cut into a 20 mm diameter cylinder. This cylinder was then sliced into discs with a diamond wheel mounted on an IsoMet saw (Buehler Ltd., Lake Bluff, IL, USA). These discs were sequentially ground with 500-grit and 800-grit silicon carbide discs either to a 0.9 mm ( n = 7), a 1.29 mm ( n = 5), or a 1.93 mm ( n = 5) ± 0.02 mm thickness (Struers LabPol polishing machine, Copenhagen, Denmark). The discs were sintered in the VITA ZYRCOMAT furnace at 1530 °C for 120 min. Three different disc thicknesses were obtained: 0.75 mm ( n = 7), 1 mm ( n = 5) and 1.5 mm ( n = 5). One sample per group was left un-veneered and served as a control.
A thick (approximately 1 mm) veneering ceramic layer (VITA VM9 Base Dentin Shade 3M2) mixed with VITA VM Modeling Liquid was fired at 910 °C on the framework of the other samples (VITA VACUMAT 4000 Premium Furnace). The thickness of this veneer layer was then reduced to 0.1 mm using a wet grinder with a 500 μm and then an 800 μm grinding paper.
To study the influence of the firing time on the curvature of the discs, the four test samples of each group went through five additional firing cycles at 900 °C so as to obtain a cumulative firing time for each sample, of respectively 30 min, 1, 2, 5 and 10 h (long firing time at 900 °C protocol, LF). The curvature of the different samples was measured after each firing cycle. The two remaining 0.75 mm veneered samples underwent five short firing cycles (6 min each) (short firing time at 900 °C protocol, SF). Those samples were used as a control to isolate the temperature variation effect from the firing time effect.
The curvature of the sample was measured on the zirconia inner surface along two perpendicular diameters using a profilometer composed of a motorized sample holder (for x -axis with a 10 μm step) and a column stand with a linear encoder (Solartron LE25S for the z -axis with a 0.05 μm resolution).
The vertical position ( z ) of the sensor was registered using labview (National Instrument, Austin, TX, USA) forwards and backwards along both axes, and the mean was calculated ( Fig. 1 ) using excel (Microsoft Corp., Redmond, WA, USA).
The curvature measurements performed before the five additional firing cycles served as the reference value. These references values were subtracted from each test measurement. Resulting curves were then fitted with a curve (Abscissa software, Rüdiger Brühl, Germany) using formula (1) .
From this fit, the radius of curvature is calculated using the formula (2) .
The calculated radius was positive if the curve concavity was facing down (i.e. on the veneering side) and negative if the curve concavity was facing up (i.e. on the zirconia side).
A VM9 beam sample was manufactured using a rectangular mold into which powders were condensed and sintered. For Vita In-Ceram YZ, a rectangular sample was manufactured, sintered and submitted to a regeneration firing procedure at 1000 °C.
The CTE at temperatures above and below Tg for VM9 (Tg ∼ 600 °C, Ts ∼ 670 °C, according to the manufacturer) and zirconia were measured during cooling using a single pushrod dilatometer (Netzsch DIL 402C, Selb, Germany) after a heating it to 2 °K min −1 up to 700 °C.
The same firing protocols than bilayered samples were applied to the VM9 beam and the Y-TZP beam, measurements being taken after each cycle respectively.
The CTE and Tg measurement data were used to estimate their impact on disc curvature using the formula (3)
1 R = 6 E z ′ E c ′ t z t c ( t z + t c ) ( α c − α z ) Δ T E z ′ 2 t z 4 + E c ′ 2 t c 4 + 2 E z ′ E c ′ t z t c ( 2 t z 2 + 2 t c 2 + 3 t z t c )