1

Introduction

Porcelain is a glass-based ceramic with low tensile strength and excellent esthetics . Thus, metal infrastructures were often used to produce large and/or posterior restorations, such as fixed partial dentures. The development of high crystalline content ceramics and the introduction of new fabrication methods allowed for the replacement of metal infrastructures . These all-ceramic restorations are also built in layers due to the opacity of the high crystalline content ceramic infrastructure .

Alumina and zirconia show several toughening mechanisms that could improve the mechanical behavior of ceramic materials. For alumina, toughening is primarily due to grain bridging in the wake of the propagating crack, which shields the crack tip from the applied load . Zirconia crystals are associated to a well-known phase transformation toughening mechanism responsible for the enhancement of the material’s fracture toughness . Thus, zirconia-based ceramics, such as In-Ceram Zirconia (IZ) and yttria partially stabilized tetragonal zirconia (Y-TZP), can be used as infrastructure materials for crowns and fixed partial dentures. Zirconia-based structures are considered the best candidates to replace metal infrastructures . In-Ceram Zirconia is an alumina-based zirconia-reinforced glass-infiltrated ceramic and Y-TZP is a polycrystalline ceramic composed by tetragonal zirconia stabilized by yttria .

In addition to the differences in composition and crystalline content, there are differences in the fabrication methods used to produce ceramic restorations, which are related to microstructural features, such as pore and crystal distribution, presence of inclusions and flaws, which could influence the mechanical behavior of materials . The majority of dental ceramics is available as blanks for processing using CAD-CAM technology (computer aided design–computer aided manufacture). The use of CAD-CAM blanks and standardized scanning and milling procedures could minimize the influence of the dental laboratory technician in the fabrication process resulting in restorations with a more uniform microstructure, less processing flaws and better adaptation . However, literature findings are controversial .

Clinical studies report high survival rates for all-ceramic restorations. Infrastructure failures are unusual when the ceramic material is correctly indicated, varying from 2 to 6% . Yet, the failure mode most frequently observed is chipping of the veneering material. This type of failure has been observed mainly when zirconia-based infrastructures are used, showing chipping rates that vary from 0 to 50%, depending on the follow-up period . Yet, failure has been described as crack propagation from a contact damage zone throughout the veneering porcelain, reaching or not the interface, which can lead to delamination and exposure of the infrastructure material in the oral environment . Among the factors related to such failure are: lack of porcelain support provided by the infrastructure , thermal incompatibility , low bond strength between materials and porcelain cooling protocol . In addition, studies suggested that the chipping rate for metal-ceramic restorations is significantly lower, around 2% . Therefore, the investigation of the failure behavior of prosthetic crowns produced with different infrastructure materials and fabrication methods should offer relevant clinical and scientific findings.

In vitro tests are frequently used to characterize the mechanical behavior of dental materials. However, these tests usually fail to induce the same stress state in which restorations are subjected in the oral cavity . In vitro testing protocols that provide clinically valid information should be developed. To achieve this goal, the fractographic analysis is an important tool. Fractography is the analysis of the fracture surface that contains characteristic features produced by the interaction of the crack tip in propagation with the material’s microstructure and stress fields. Through this analysis it is possible to identify the crack size and initiation site . The failure behavior observed in vitro can be correlated with the behavior reported for restorations that failed in vivo .

Thus, considering the clinical problems of high porcelain chipping rates, this study investigated, in vitro, the influence of the infrastructure material on the failure behavior of prosthetic crowns. Two types of ceramic (In-Ceram Zirconia and Y-TZP) and a metal infrastructure (IS) were evaluated. The hypothesis tested is that different IS materials result in different fracture load values and failure modes. In addition, the influence of the fabrication method (CAD-CAM or slip-cast) on the failure behavior of In-Ceram Zirconia prosthetic crowns was also studied.

2

Materials and methods

The materials and fabrication methods used in this study are shown in Table 1 . Ten restorations were produced for each experimental group.

The prosthetic crowns were designed based on a type 304 stainless steel die simulating a prepared first lower premolar, with 4.5 mm height, 6° axial taper and a 120° chamfer . Impression (polyvinyl siloxane, Aquasil™ Soft Putty, high viscosity paste, and Aquasil™ Low Viscosity, Dentsply, Petropolis, RJ, Brazil) of the master metal die was taken and adequate model was produced according to the IS fabrication method.

For YZ-C and IZ-C groups, IS were produced using CEREC InLab CAD-CAM system (Sirona Dental Company, Charlote, NC, EUA). Thus, type IV plaster (CAM-base, Dentona AG, Dortmund, Germany) was used to produce the model, which was laser scanned in a CAD-CAM unit (Cerec inLab, Sirona Dental Company, Germany). The design of the first premolar IS was selected in the CAD-CAM software (InLab 3D version 2.90, Sirona Dental Company, Germany) and the ceramic blocks were milled. YZ-C IS was sintered in a special furnace (Zyrcomac, Vita Zahnfabrik, Germany). IZ-C IS was glass infiltrated (Glass Zirconia Powder Zahnfabrik Vita, Germany) using appropriate firing cycle (Inceramat 3 furnace, Vita Zahnfabrik, Germany). The excess glass was removed from the infiltrated IS with finishing burs (#1112F and 1112FF, KG Sorensen, Brazil). All procedures were performed according to manufacturer’s recommendations.

For the slip-casting IS (IZ-S), two layers of a spacer (Vita In-Ceram Interspace Varnish, Vita Zahnfabrik, Germany) were applied on the metal die, resulting in a 40 μm-thick spacer layer. Then, a silicon mold (Components 1 and 2, CEHA ECS White Duplicating Silicone, C HAFNER, Pforzheim, Germany) was used to duplicate the die into ten plaster models (Vita In-Ceram Special Plaster, Vita Zahnfabrik, Germany). The ceramic slurry was prepared by mixing 38 g of powder (Vita In-Ceram Zirconia Powder, Vita Zahnfabrik, Germany), 5 ml of liquid (Vita In-Ceram Zirconia Mixing Liquid, Vita Zahnfabrik, Germany) and one drop of additive (Vita In-Ceram Additive, Vita Zahnfabrik, Germany). The slurry was applied on the models and, after moisture absorption by the plaster, the green body was carefully removed from the model and sintered (Inceramat 3 furnace, Vita Zahnfabrik, Germany), according to manufacturer’s instructions. Same glass infiltration and finishing procedures described for IZ-C IS were performed for IZ-S IS.

Plaster models (type IV plaster, Bego Stone plus, Bego, Germany) were produced and used to fabricate NiCr alloy IS (Wironia® light, BEGO Bremer Goldschlägerei Wilh. Herbst GmbH & Co. KG, Bremen, Germany) using the lost wax technique, according to the manufacturer’s recommendation. The wax IS pattern was embedded in the refractory investment and the wax was eliminated at 840 °C. The alloy was heated at 1350 °C with an oxygen and acetylene gas burner and injected into the spinning ring. The IS was removed from the investment ring, cleaned and grinded with diamond burs. The metal IS were subjected to a thermal cycle, according to the manufacturer instructions, to produce an oxide layer. Subsequently a thin layer of Vita VM13 WASH OPAQUE (Vita Zahnfabrik, Germany) was applied over the IS and sintered as recommended by the manufacturer.

IS were veneered with the manufacturer recommended porcelain ( Table 1 ). The porcelain shade 2M2 was used for all experimental groups. Veneering was performed by an experienced dental laboratory technician. A mixture of porcelain powder and distilled water was applied on the IS, vibrated and excess humidity was removed with absorbing paper. The porcelains were sintered (Vita Vacumat 40 furnace, Vita Zahnfabrik, Germany) according to the manufacturer recommendations.

The external shape of the crowns corresponded to a simplified shape of a lower first premolar. All IS have a 0.5 mm thickness. The final thickness (IS + porcelain) was measured using a digital caliper in pre-determined locations of the crowns and were as follows: 1.2 ± 0.05 mm in the lateral walls, 2.0 ± 0.05 mm at the tip of the buccal and lingual cusps, and 1.0 ± 0.05 mm at the center of the occlusal fossa. A uniform porcelain thickness was obtained using abrasive burs (Supermax, Edenta, Switzerland). Finally, all crowns were subjected to a glaze firing cycle as recommended by the manufacturer (Vita Vacumat 40, Vita Zahnfabrik, Germany).

Replicas of the original metal dies were produced using a dentin analog material (epoxy filled with woven glass fibers, NEMA grade G10, Accurate Plastics, Inc., USA). Crowns were sonically cleaned in isopropyl alcohol bath for 5 min before cementation onto the epoxy fiber-reinforced dies using resin cement (Panavia F, Kuraray, Japan). Surface treatment of the bonding area of the dies followed reported recommendations , i.e., the surface was etched with 10% hydrofluoric acid for 1 min (Condac Porcelana, FGM, Brazil), washed in water for 30 s, dried using oil-free air, followed by the application of a silane bonding agent (Silano Agente de União, Angelus, Brazil) and an adhesive system (ED Primer A + B, Kuraray, Japan). Cement pastes were mixed and applied to the internal surface of the crowns, which were placed onto the dies. No treatment was performed in the internal surface of the crowns. A static load of 20 N was applied to the occlusal surface of the crown using a cementation device and excess cement was removed from the finishing line. Each crown surface was light cured for 20 s (Radii-cal LED curing light, SDI, Victoria, Australia) from each side (buccal, lingual, mesial and distal). The cemented crowns were stored in 37 °C distilled water for 24 h before loading to failure.

All restorations were loaded in compression in a 30° angle with a spherical stainless steel piston (6 mm diameter). The test was performed in 37 °C distilled water bath using a universal testing machine (EMIC DL 2000, São Jose dos Pinhais, PR, Brazil) at a cross-head speed of 0.5 mm/min.

After testing, the fracture surfaces of all crowns were examined under a stereomicroscope (Serie ZTX ZOOM, Instrumento óptico Co. de Nigbo Wason Ltd., Nigbo/Zhejiang, China) to identify the failure modes and fractographic features. The failure mode was classified as chipping (porcelain fracture with or without IS exposure) or catastrophic failure (fracture of the porcelain and the IS). Scanning electron microscopy (SEM- Superscan SSX-550, Shimadzu Corporation, Kyoto, Japan) analysis was also performed to map the fracture surface and to identify the flaw origin, following fractographic principles .

Fracture load data were statistically analyzed using one-way analysis of variance (ANOVA) and Student–Newman–Keuls tests ( α = 0.05). Chi-square test was used to verify the relation between failure modes and IS materials ( α = 0.05).

2

Materials and methods

The materials and fabrication methods used in this study are shown in Table 1 . Ten restorations were produced for each experimental group.

The prosthetic crowns were designed based on a type 304 stainless steel die simulating a prepared first lower premolar, with 4.5 mm height, 6° axial taper and a 120° chamfer . Impression (polyvinyl siloxane, Aquasil™ Soft Putty, high viscosity paste, and Aquasil™ Low Viscosity, Dentsply, Petropolis, RJ, Brazil) of the master metal die was taken and adequate model was produced according to the IS fabrication method.

For YZ-C and IZ-C groups, IS were produced using CEREC InLab CAD-CAM system (Sirona Dental Company, Charlote, NC, EUA). Thus, type IV plaster (CAM-base, Dentona AG, Dortmund, Germany) was used to produce the model, which was laser scanned in a CAD-CAM unit (Cerec inLab, Sirona Dental Company, Germany). The design of the first premolar IS was selected in the CAD-CAM software (InLab 3D version 2.90, Sirona Dental Company, Germany) and the ceramic blocks were milled. YZ-C IS was sintered in a special furnace (Zyrcomac, Vita Zahnfabrik, Germany). IZ-C IS was glass infiltrated (Glass Zirconia Powder Zahnfabrik Vita, Germany) using appropriate firing cycle (Inceramat 3 furnace, Vita Zahnfabrik, Germany). The excess glass was removed from the infiltrated IS with finishing burs (#1112F and 1112FF, KG Sorensen, Brazil). All procedures were performed according to manufacturer’s recommendations.

For the slip-casting IS (IZ-S), two layers of a spacer (Vita In-Ceram Interspace Varnish, Vita Zahnfabrik, Germany) were applied on the metal die, resulting in a 40 μm-thick spacer layer. Then, a silicon mold (Components 1 and 2, CEHA ECS White Duplicating Silicone, C HAFNER, Pforzheim, Germany) was used to duplicate the die into ten plaster models (Vita In-Ceram Special Plaster, Vita Zahnfabrik, Germany). The ceramic slurry was prepared by mixing 38 g of powder (Vita In-Ceram Zirconia Powder, Vita Zahnfabrik, Germany), 5 ml of liquid (Vita In-Ceram Zirconia Mixing Liquid, Vita Zahnfabrik, Germany) and one drop of additive (Vita In-Ceram Additive, Vita Zahnfabrik, Germany). The slurry was applied on the models and, after moisture absorption by the plaster, the green body was carefully removed from the model and sintered (Inceramat 3 furnace, Vita Zahnfabrik, Germany), according to manufacturer’s instructions. Same glass infiltration and finishing procedures described for IZ-C IS were performed for IZ-S IS.

Plaster models (type IV plaster, Bego Stone plus, Bego, Germany) were produced and used to fabricate NiCr alloy IS (Wironia® light, BEGO Bremer Goldschlägerei Wilh. Herbst GmbH & Co. KG, Bremen, Germany) using the lost wax technique, according to the manufacturer’s recommendation. The wax IS pattern was embedded in the refractory investment and the wax was eliminated at 840 °C. The alloy was heated at 1350 °C with an oxygen and acetylene gas burner and injected into the spinning ring. The IS was removed from the investment ring, cleaned and grinded with diamond burs. The metal IS were subjected to a thermal cycle, according to the manufacturer instructions, to produce an oxide layer. Subsequently a thin layer of Vita VM13 WASH OPAQUE (Vita Zahnfabrik, Germany) was applied over the IS and sintered as recommended by the manufacturer.

IS were veneered with the manufacturer recommended porcelain ( Table 1 ). The porcelain shade 2M2 was used for all experimental groups. Veneering was performed by an experienced dental laboratory technician. A mixture of porcelain powder and distilled water was applied on the IS, vibrated and excess humidity was removed with absorbing paper. The porcelains were sintered (Vita Vacumat 40 furnace, Vita Zahnfabrik, Germany) according to the manufacturer recommendations.

The external shape of the crowns corresponded to a simplified shape of a lower first premolar. All IS have a 0.5 mm thickness. The final thickness (IS + porcelain) was measured using a digital caliper in pre-determined locations of the crowns and were as follows: 1.2 ± 0.05 mm in the lateral walls, 2.0 ± 0.05 mm at the tip of the buccal and lingual cusps, and 1.0 ± 0.05 mm at the center of the occlusal fossa. A uniform porcelain thickness was obtained using abrasive burs (Supermax, Edenta, Switzerland). Finally, all crowns were subjected to a glaze firing cycle as recommended by the manufacturer (Vita Vacumat 40, Vita Zahnfabrik, Germany).

Replicas of the original metal dies were produced using a dentin analog material (epoxy filled with woven glass fibers, NEMA grade G10, Accurate Plastics, Inc., USA). Crowns were sonically cleaned in isopropyl alcohol bath for 5 min before cementation onto the epoxy fiber-reinforced dies using resin cement (Panavia F, Kuraray, Japan). Surface treatment of the bonding area of the dies followed reported recommendations , i.e., the surface was etched with 10% hydrofluoric acid for 1 min (Condac Porcelana, FGM, Brazil), washed in water for 30 s, dried using oil-free air, followed by the application of a silane bonding agent (Silano Agente de União, Angelus, Brazil) and an adhesive system (ED Primer A + B, Kuraray, Japan). Cement pastes were mixed and applied to the internal surface of the crowns, which were placed onto the dies. No treatment was performed in the internal surface of the crowns. A static load of 20 N was applied to the occlusal surface of the crown using a cementation device and excess cement was removed from the finishing line. Each crown surface was light cured for 20 s (Radii-cal LED curing light, SDI, Victoria, Australia) from each side (buccal, lingual, mesial and distal). The cemented crowns were stored in 37 °C distilled water for 24 h before loading to failure.

All restorations were loaded in compression in a 30° angle with a spherical stainless steel piston (6 mm diameter). The test was performed in 37 °C distilled water bath using a universal testing machine (EMIC DL 2000, São Jose dos Pinhais, PR, Brazil) at a cross-head speed of 0.5 mm/min.

After testing, the fracture surfaces of all crowns were examined under a stereomicroscope (Serie ZTX ZOOM, Instrumento óptico Co. de Nigbo Wason Ltd., Nigbo/Zhejiang, China) to identify the failure modes and fractographic features. The failure mode was classified as chipping (porcelain fracture with or without IS exposure) or catastrophic failure (fracture of the porcelain and the IS). Scanning electron microscopy (SEM- Superscan SSX-550, Shimadzu Corporation, Kyoto, Japan) analysis was also performed to map the fracture surface and to identify the flaw origin, following fractographic principles .

Fracture load data were statistically analyzed using one-way analysis of variance (ANOVA) and Student–Newman–Keuls tests ( α = 0.05). Chi-square test was used to verify the relation between failure modes and IS materials ( α = 0.05).