- Email: [email protected]
Shear strength of core-veneer interface in bi-layered ceramics
Shear strength of core-veneer interface in bi-layered ceramics Hana M. Al-Dohan, BDS, MS,a Peter Yaman, DDS, MS,b Joseph B. Dennison, DDS, MS,c Michael E. Razzoog, DDS, MS, MPH,d and Brien R. Lang, DDS, MSe School of Dentistry, University of Michigan, Ann Arbor, Mich Statement of problem. Delamination of veneering porcelain from underlying ceramic substrates has been reported for all-ceramic restorations. Whether this phenomenon is an inherent weakness of the veneering porcelain due to a weak interface between the veneering and the core porcelains, or merely a fracture through the veneering porcelain itself, has not been explored. Purpose. The purpose of this study was to investigate the strength of the substructure and veneering porcelain interface in all-ceramic systems. Methods. The all-ceramic systems tested with their respective veneering porcelains were IPS-Empress2 with Eris (IE), Procera AllCeram with AllCeram (PA), Procera AllZircon with CZR (PZ), and DC-Zircon with Vita D (DC). The veneering porcelain recommended by the manufacturer for each material was fired to the ceramic core. A metal ceramic (MC) combination was tested as a control group. Sixty specimens, 12 for each system and control, were made from 1 master die. A cylinder of veneering porcelain 2.4 mm in diameter was applied using a specially designed aluminum split mold. After firing, the specimens were placed in a mounting jig and subjected to shear force in a universal testing machine. Load was applied at a crosshead speed of 0.50 mm/min until failure. Average shear strengths (MPa) were analyzed with a 1-way analysis of variance and the Tukey test (a=.05). The failed specimens were examined microscopically at original magnification 320 to classify the mode of failure as cohesive in the core, cohesive in the veneer, or adhesive at the interface. Results. The mean shear strengths (6SD) in MPa were MC control 30.16 6 5.88; IE bonded to Eris 30.86 6 6.47; PZ bonded to CZR 28.03 6 5.03; DC bonded to Vita D 27.90 6 4.79; and PA bonded to AllCeram 22.40 6 2.40. IE, PZ, and DC were not significantly different from the MC control. Microscopic examination showed that adhesive failure, or complete delamination, did not occur between the compatible ceramic core and veneering materials. Failure primarily occurred near the interface with residual veneering porcelain remaining on the core. IE with Eris exhibited cohesive failure in both the core and the veneer. Conclusion. The bond strengths of 3 of the tested all-ceramic materials (IE, PZ, and DC) were not significantly different from the control (MC) group. (J Prosthet Dent 2004;91:349-55.)
CLINICAL IMPLICATIONS This study demonstrated that the bond of veneering porcelain to a ceramic core for the materials tested is similar to that of the metal ceramic control. It may be expected that the clinical behavior would be similar.
E
sthetic restorative treatment in dentistry has made dental ceramics an often-used alternative for both anterior and posterior restorations. The dental profession seeks an ideal all-ceramic restoration with excellent physical properties, strength, marginal fit, Presented at the IADR, Goteborg, Sweden, June, 2003. a Graduate Student, Department of Cariology, Restorative Sciences and Endodontics. b Clinical Professor, Department of Cariology, Restorative Sciences and Endodontics. c Professor, Department of Cariology, Restorative Sciences and Endodontics. d Professor, Department of Biologic and Materials Sciences Division of Prosthodontics. e Professor, Department of Biologic and Materials Sciences Division of Prosthodontics.
APRIL 2004
and esthetics necessary for anterior, as well as posterior, restorations.1,2 In many situations all-ceramic crowns have a potential to be more esthetic than metal ceramic (MC) restorations. For the MC crown, the alloy structure produces an opaque appearance, and the metal margins are often visible. High-gold–content alloys are relatively expensive, and the alternatives may have disadvantages, such as risk of metal allergy, bond failure, or porcelain discoloration, and noble metal content alone is not indicative of clinical performance.3 Reinforced all-ceramic crowns consist of a highstrength porcelain core material, laminated with dentin and incisal porcelain.4 Successful performance and reliability of veneered ceramic prostheses may be limited THE JOURNAL OF PROSTHETIC DENTISTRY 349
THE JOURNAL OF PROSTHETIC DENTISTRY
AL-DOHAN ET AL
and (8) environmental effects.11 In the present study, a crosshead speed of 0.5 mm/min was chosen. It has been hypothesized that relatively high crosshead speeds may develop abnormal stress distributions during the shear test, inducing cohesive failure in either of the bonded materials, which would influence the bond strength values achieved. According to Hara et al,12 crosshead speeds of 0.50 and 0.75 mm/min result in more adhesive failures and are therefore preferable in shear bond strength tests. The purpose of this study was to develop a protocol to investigate the average shear strength of the core-veneer interface in bilayered all-ceramic systems and to use that test to compare metal ceramic to all-ceramic systems.
MATERIAL AND METHODS Fig. 1. Master die and core.
by mechanical integrity and adhesion of the veneering porcelain to the ceramic substrates. The mechanical properties of core materials and veneering porcelains should match to a certain extent to achieve a durable bond.5 According to the cohesive plateau theory, a proper bond is one that is stronger than the materials joined together.6 Previous studies testing the porcelainto-metal bond reported shear strength equal to the shear strength of the veneering porcelain as being an adequate bond.7 However, if the metallic understructure is eliminated, the bond to the underlying porcelain core becomes an important issue. Without documented evidence of the strength of the bond between the core and veneering porcelain, the profession must rely on manufacturers’ claims to judge which material is best for patients. The bond between a metal core and veneering porcelain has been investigated using bending tests.8,9 While this type of test may be appropriate for ceramic metal systems, it cannot be applied to a ceramic core due to its brittle nature. The fracture mechanism of ceramics and metals are different due to their different structure and mechanism of bonding. Covalent and ionic bonds in ceramics are associated with large interatomic forces and hence produce a stronger resistance to plastic deformation than metals.10 External loads will therefore cause stress concentration at a crack tip instead of relaxation by plastic flow.10 There are several factors that are associated with the stress state created in dental ceramic restorations, including: (1) thickness of ceramic layers, (2) mechanical properties of each ceramic, (3) elastic modulus of the supporting substrate material, (4) direction, magnitude, and frequency of applied load, (5) size and location of occlusal contact areas, (6) residual stresses induced by processing, (7) restoration-cement interfacial defects, 350
Specimen preparation A master die was fabricated from dental stone (Jade Stone; Whip Mix Corp, Louisville, Ky). The stone die was a negative replica of a previously fabricated ceramic core that was 8 mm high with a flat occlusal surface, and 10 mm in diameter with rounded occlusal line angles and a 1-mm–deep circumferential chamfer (Fig. 1). The all-ceramic systems tested and the manufacturerrecommended veneering porcelains are listed in Table I. Sixty specimens, 12 for each group, were made from the master die. A 2.4-mm–diameter cylinder of the veneering porcelain was applied using a specially designed aluminum split mold (Fig. 2). Due to porcelain shrinkage, a total of 3 separate firings were required to establish the correct diameter.
Mounting The completed ceramic specimen was inverted on a glass slab with a 4-mm–diameter opening (Fig. 3). The coping was fixed in position using sticky wax (Kerr Corp, Orange, Calif), and a brass ring, 30 mm in diameter and 30 mm in height, was placed over the coping on the glass slab and held in position with glue (Quicktite; Manco, Inc, Avon, Ohio). A die-stone (Jade Stone; Whip Mix Corp) was mixed and poured into the ring as illustrated in Figure 3.
Testing Each specimen was mounted in a metal holder on the universal testing machine (Instron no. 5565; Instron Corp, Canton, Mass) and the load was applied with the jig (Custom-made; Ultradent Inc, South Jordan, Utah) that had a diameter of 2.4 mm corresponding to the diameter of the veneering porcelain. Each specimen was tightened and stabilized to ensure that the edge of the shearing jig was touching the core surface and was positioned as close to the veneer-core interface as VOLUME 91 NUMBER 4
AL-DOHAN ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Table I. All-ceramic systems and control group used in study Group
IE
Group name
Empress2
Core Veneer
PZ
Procera allZircon
Core Veneer
DC
PA
DC Zircon
Procera allCeram
MC
Fabrication
IPS-Empress (ingot) IPS-Empress (IPS Eris)
Fabricate core from waxings on the master die Cylinder: 2.37mm in diameter, 3mm in height using specially designed split mold Fabricate cores to digitized specifications of master die Same as above
Procera allZircon Core Cerabien CZR
Core
DC-Zircon core
Veneer
Vita D opaque porcelain
Core Veneer
Control group
Material
Core Veneer
Procera allCeram Alumina Core Degussa-Ney All ceram dentine Lodestar Noritake
possible (Fig. 4). A shear load was applied at a crosshead speed of 0.50 mm/min until failure. The ultimate load to failure was recorded by the system’s software (Merlin; Instron Corp) in Newtons (N). Average shear strengths (MPa) were calculated by dividing the load (N) at which failure occurred by the bonding area (mm2): Shear stress ðMPaÞ ¼ Load ðNÞ 4 Area ðmm2 Þ The mean failure load and the SD for each group were calculated from these data.
Fabricate cores to digitized specifications of master die Same as above
Fabricate cores to digitized specifications of master die Same as above Fabricate basic core from waxups on master die Same as above
Manufacturer
Ivoclar Vivadent, Amherst, NY Ivoclar Vivadent
Procera Sandvik, Inc, Gothenberg, Sweden Noritakekizai Co Ltd, Nagoya, Japan TKT Metoxit AG, Thayngen, Switzerland Vita Zahnfabrik H. Rauter GmbH & Co KG, Sackinien, Germany Procera Sandvik, Inc Dentsply Int., York, Pa Ivoclar Vivadent Noritakekizai Co, Ltd
Statistical analysis Statistical analysis was carried out using statistical software (SAS software; SYSTAT Software Inc, Evanston, Ill). The data were analyzed using a 1-way analysis of variance test (ANOVA) to determine whether significant differences existed between the shear strengths of the 5 groups (a=.05). Also, a Tukey multiple comparisons test (a=.05) was used to assess the differences among the specified materials.
Microscopic examination The fractured surfaces were visually analyzed with a microscope (SMZ-U Stereoscopic Zoom Microscope; Nikon Corp, Tokyo, Japan) at original magnification 320. Sketches of the fractured surfaces were plotted and scanned to a computer. The fractured surfaces were then analyzed using a computer software program (Scion Image; Scion Corp, Frederick, Md) to determine the percentage of cohesive fracture within the veneer, cohesive fracture within the core, or adhesive failure at the core/veneer interface. This was done by tracing the borders of the cohesive veneer/core fracture that remained within the bonded interface and calculating the area using the Scion program. The percentage was calculated by dividing this area by the total bonded area (Fig. 5). The failed surfaces were classified as cohesive fracture within the veneer (V), adhesive fracture between the core and the veneer (C/V), and cohesive fracture within the core (C). APRIL 2004
RESULTS Table II shows the mean shear strengths of the coreveneer interface of the 4 all-ceramic groups and control group (MC). The highest mean shear strength was recorded for IPS-Empress2 bonded to Eris (30.86 6 6.47 MPa). The mean shear strength of Procera AllZircon bonded to Cerabien CZR was 28.03 6 5.03 MPa; DCZircon bonded to Vita D was 27.90 6 4.79 MPa; Procera alumina bonded to AllCeram veneering porcelain was 22.40 6 2.40 MPa. The 1-way ANOVA showed a significant difference for the shear strengths among the materials tested at P\.001 (Table III). The Tukey multiple comparisons test was computed to make all pairwise comparisons among the 4 groups in this study plus the control group. These comparisons are listed in the last column of Table II. The P values of the different comparisons show that IE, PZ, and DC were not significantly 351
THE JOURNAL OF PROSTHETIC DENTISTRY
AL-DOHAN ET AL
Fig. 3. Mounting technique.
Fig. 4. Testing using semicircular blade attached to universal testing machine.
Fig. 2. Veneer fabrication. a, Specially designed aluminum split mold holds core. b, Core secured via screws in both sides of split mold. c, Veneering porcelain placed inside mold (2.4 mm in diameter). d, As screws are loosened, 2 parts of mold are separated. e, Specimen removed and veneer is fired.
352
different and were comparable to the control group. Procera alumina had significantly lower values than the control, yet it was not significantly lower than PZ and DC. The highest values were observed with IE, and the lowest were observed with PA. The mean values for the fracture analysis were calculated and are provided in Table IV. All test groups demonstrated cohesive failure within the veneer, as well as adhesive failure between the core and the veneer. Only Empress2/Eris demonstrated cohesive failure within the core. Procera alumina showed more surface failure at the core-veneer interface. VOLUME 91 NUMBER 4
AL-DOHAN ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Table II. Mean shear strength (MPa) of 4 groups of all-ceramic specimens and control group All ceramic system Group
Core
Sample size (n)
Mean (MPa)
SD
Sign*
Eris CZR
12 12
30.86 28.03
6 6.47 6 5.03
a ab
Vita D All Ceram
12 12
27.90 22.40
6 4.79 6 2.40
ab bc
Noritake
12
30.16
5.89
Veneer
IE PZ
Empress 2 Procera AllZircon DC DC-Zircon PA Procera alumina Control Metal group (Lodestar)
a
*Values with the same letter are not statistically different using Tukey test at P\.05.
Fig. 5. Sketch of fractured surface classified: cohesive within veneer (V), adhesive (C/V), and cohesive within core (C).
Table III. One-way ANOVA data of shear length of different all-ceramic specimens and control group Source
Sum of squares
df
Mean square
F-Ratio
P value
530.20 1434.65
4 55
132.55 26.09
5.08
.001*
Material Error
*Significant at 95% CI.
Table IV. Failed bonded surfaces divided by percentage: (V) cohesive within veneer, (C/V) combined surface failure, (C) cohesive within core Material group
MC (Control) Empress 2/Eris Procera AllZircon/CZR DC-Zircon/Vita D Procera alumina/AllCeram DC-Zircon/Eris
Fig. 6. Test scheme. A, Movement will be in one direction parallel to bonded surface. B, Failure due to bending moment.
DISCUSSION The currently accepted method of measuring shear bond strength for enamel and dentin adhesives was used in the present study. The shear bond strength (SBS) test is defined as a test in which 2 materials are connected via an adhesive agent and loaded in shear until separation occurs.13 The bond strength is calculated by dividing the maximum applied force by the bonded crosssectional area.14 This test is relatively simple and easy to perform, producing rapid results. However, some critical aspects must be considered in using an in vitro method to predict the clinical performance of adhesive materials. First, in vitro information cannot be extrapAPRIL 2004
V
V/C
C
56 42.8 41.6 58.6 30.3 27.9
43 8 57.4 40.4 68.7 71.1
1 49.2 1 1 1 1
olated directly to clinical situations, and second, the great variations in SBS test results have made the test questionable.15 An effort must therefore be made to standardize SBS test methods to improve the clinical usefulness of this in vitro test. Important aspects that should be considered include storage conditions, type of substrate, specimen preparation, rate of load application, cross-sectional surface area, and experience of the researcher. Establishing parameters for some of these issues was proposed in 1994 by an ISO standard.16 Studies of lap joints using mathematical stress analysis and finite element stress analysis demonstrate that the stresses at the interface between the adhesive and the substrate are not uniform and are highly dependent on the test geometry and loading configuration adopted.16 Van Noort et al17 demonstrated that the highest concentration of stress is seen at the loading point of the interface between the composite and dentin surface. 353
THE JOURNAL OF PROSTHETIC DENTISTRY
Van Noort et al17 compared a jig with a semicircular edge versus the straight knife edge. Since the initial contact of the knife is at one point on the edge of the bonded specimen, the stress is concentrated in a smaller area, resulting in premature failure. In theory, the curved knife wraps around the cylinder to be loaded, thereby contacting a larger area. It is thought that the curved knife distributes the stress over a larger surface area and minimizes load concentration. The data (Table II) illustrate that IE, IPS Eris veneering porcelain applied to IPS-Empress2 core, produced the highest values. Comparable results were found with both zirconia groups (PZ and DC) bonded to the veneering porcelain recommended by the manufacturer. AllCeram veneering porcelain applied to Procera alumina core (PA) demonstrated a weaker bond. During fabrication of the AllCeram/Procera alumina core specimens, a thin layer of opaque was applied in the first firing. This is the first step in the veneering sequence to improve the esthetic properties and mask the more yellowish color of the alumina core. The lower color value of this layer neutralizes the relatively bright aluminum oxide coping, which also provides the initial color to the restoration as well as bonding to the coping. It is very thin in consistency and when fired shrinks considerably. This might explain the weaker shear stress found in this group, since the opaque shrinks considerably after firing, leaving a very thin layer. The effect of this shrinkage was observed under high magnification where only remnants of the opaque were seen. Thus, when the load creates interfacial stresses, the crack propagation occurs in the weaker dentin layer near the core-veneer. An adequate bond for metal-ceramics has been determined to occur when the fracture stress is greater than 25 MPa.11 Adequate bond strength for all-ceramic materials has not been determined. Table II shows that the control group (MC) demonstrated higher (30.15 MPa) mean shear strength than recommended, and that IE, PZ, and DC were not significantly different from the control group. A study by Smith et al18 to gain insight into the fracture behavior of prostheses under incisal-directed, load-to-failure testing showed that 50% of the allceramic In-Ceram crowns failed by delamination of veneering glass alone, leaving a thin layer of residual glass on the core surface. The term ‘‘delamination’’ indicates complete debonding or adhesive failure, and microscopic examination of all the specimens in the present study substantiated this finding. Figure 6 explains the sequence of events the specimens may undergo during testing. In Figure 6, A, the applied stress exceeds the ultimate shear strength, and the veneer moves as a whole in the direction of the load, and complete debonding occurs. In this situation, the failure occurs due to shear stress. In Figure 6, B, the bond is stronger than the stress and resists the shear 354
AL-DOHAN ET AL
force. In this example, the applied tensile stress has exceeded the ultimate tensile stress; the veneer will rotate, and the failure will be a cohesive and adhesive combination. In Figure 2, Procera AllZircon/CZR group and DCZircon/Vita D group showed more veneer left on the core surface; failures were primarily within the veneer (55%-60%), and the remainder were at the interface with veneer residue left on the core surface. Also, the Empress2/ Eris group showed failure to be cohesive within both the veneer (45.83%) and core (52.5%). Empress2 core was the only material that was pulled out at the interface. A study by Wagner and Chu19 compared biaxial flexural strength and indentation fracture toughness of 3 ceramic core materials, Empress, In-Ceram, and Procera AllCeram. The results showed that the average flexural strength of Empress (134 MPa) and fracture toughness (1.74 MPam1/2) were significantly lower than the other two materials, Porcera AllCeram (687 MPa), and In-Ceram (352 MPa). Adhesive failure was not seen between the veneer and core due to the fact that the 2 materials fuse together and certain elements diffuse across the interface. This was explained by Smith et al18 in a study that investigated the possibility that elements unique to the core or the veneering glass diffused across the interface, or the possibility that a layer of excess infiltration glass developed on the core surface during processing. Either of these occurrences may cause a chemical alteration of the glass layer adjacent to the core, possibly changing physical properties at the interface, such as strength or coefficient of thermal expansion. Elemental analysis was performed for lanthanum, calcium, silicon, sodium, and potassium and showed that residual infiltration glass was not present on the core surface, and chemical alteration in the veneering glass was limited to 1 to 2 mm.18 Adhesive failure does not occur in the presence of a good bond between a compatible ceramic core and veneer material. Microscopic examination of Procera AllCeram showed failure to be primarily at the interface, with residue of the veneering porcelain remaining on the core. According to Oden et al,20 the strength of this veneering porcelain in combination with the aluminum oxide coping was demonstrated to be excellent. The veneering porcelains are chemically bonded to the densely sintered aluminum oxide by ionic and covalent bonds.20 Cohesive failure within the veneer was also seen, but to a somewhat lesser extent compared with the other groups.
CONCLUSIONS Within the limitations of this study, the following conclusions are made: 1. Eris veneering porcelain applied to IPS-Empress2 core showed the highest shear strength values and were VOLUME 91 NUMBER 4
AL-DOHAN ET AL
not significantly different from the metal ceramic (MC) control surface. 2. The bond of the veneering porcelains to the 2 zirconia cores was not significantly different from 1PSEmpress2/Eris or the MC control. 3. AllCeram applied to the Procera alumina core showed a significantly weaker bond compared to the other systems. Remnants of the opaque layer on the core were observed microscopically after failure. 4. Surface analysis of failure modes demonstrated that the bond between the core and the veneer was cohesive in the veneer and adhesive at the interface for most systems tested. REFERENCES 1. Tam LE, McComb D. Shear bond strengths of resin luting cements to laboratory-made composite resin veneers. J Prosthet Dent 1991;66: 314-21. 2. Malament KA, Grossman DG. The cast glass-ceramic restoration. J Prosthet Dent 1987;57:674-83. 3. O’Brien WJ. Dental materials and their selection. 3rd ed. Chicago: Quintessence; 2002. p. 200-9. 4. O’Brien WJ. Dental materials and their selection. 3rd ed. Chicago: Quintessence Int; 2002. p. 216-20. 5. Blatz MB. Long-term clinical success of all-ceramic posterior restorations. Quintessence Int 2002;33:415-26. 6. Gwinnett AJ. A new method to test the cohesive strength of dentin. Quintessence Int 1994;25:215-8. 7. Yamada NH. Dental porcelain: the state of the art. 1st ed. Los Angeles: University of Southern Calif School of Dentistry; 1977. p. 137-40. 8. Ozcan M. Fracture reasons in ceramic-fused-to-metal restorations. J Oral Rehabil 2003;30:265-9. 9. Baran G, Boberick K, McCool J. Fatigue of restorative materials. Crit Rev Oral Biol Med 2001;12:350-60.
APRIL 2004
THE JOURNAL OF PROSTHETIC DENTISTRY
10. Kvam K, Hero H, Oilo G. Fracture toughness measurements of some dental core ceramics: a methodologic study. Scand J Dent Res 1991; 99:527-32. 11. Craig RG. Mechanical properties in restorative dental materials. 11th ed. New York: Mosby; 2002. p. 551-92. 12. Hara AT, Pimenta LA, Rodrigues AL Jr. Influence of cross-head speed on resin-dentin shear bond strength. Dent Mater 2001;17:165-9. 13. Craig RG, Powers JM. Restorative dental materials. 11th ed. Mosby; 2002. p. 85. 14. Versluis A, Tantbirojn D, Douglas WH. Why do shear bond test pull out dentin? J Dent Res 1997;76:1298-307. 15. Hadavi F, Hey JH, Ambrose ER, Louie PW, Shinkewski DJ. The effect of dentin primer on the shear bond strength between composite resin and enamel. Oper Dent 1993;18:61-5. 16. International Organization for Standardization. ISO TR 11405, Dental Materials-Guidance on testing of adhesion to tooth structure. 1994. 17. Van Noort R, Noroozi S, Howard IC, Cardew G. A critique of bond strength measurements. J Dent 1989;17:61-7. 18. Smith TB, Kelly JR, Tesk JA. In vitro fracture behavior of ceramic and metal-ceramic restorations. J Prosthodont 1994;3:138-44. 19. Wagner WC, Chu TM. Biaxial flexural strength and indentation fracture toughness of three dental core ceramics. J Prosthet Dent 1996;76: 140-4. 20. Oden A, Andersson M, Krystek-Ondracek I, Magnusson D. Five-year clinical evaluation of Procera AllCeram crowns. J Prosthet Dent 1998; 80:450-6. Reprint requests to: DR PETER YAMAN DEPARTMENT OF CARIOLOGY, RESTORATIVE SCIENCES, UNIVERSITY OF MICHIGAN 1101 N. UNIVERSITY AVE ANN ARBOR, MI 48109 FAX: 734-936-1597 E-MAIL: [email protected]
AND
ENDODONTICS
0022-3913/$30.00 Copyright ª 2004 by The Editorial Council of The Journal of Prosthetic Dentistry doi:10.1016/j.prosdent.2004.02.009
355
CLINICAL IMPLICATIONS This study demonstrated that the bond of veneering porcelain to a ceramic core for the materials tested is similar to that of the metal ceramic control. It may be expected that the clinical behavior would be similar.
E
sthetic restorative treatment in dentistry has made dental ceramics an often-used alternative for both anterior and posterior restorations. The dental profession seeks an ideal all-ceramic restoration with excellent physical properties, strength, marginal fit, Presented at the IADR, Goteborg, Sweden, June, 2003. a Graduate Student, Department of Cariology, Restorative Sciences and Endodontics. b Clinical Professor, Department of Cariology, Restorative Sciences and Endodontics. c Professor, Department of Cariology, Restorative Sciences and Endodontics. d Professor, Department of Biologic and Materials Sciences Division of Prosthodontics. e Professor, Department of Biologic and Materials Sciences Division of Prosthodontics.
APRIL 2004
and esthetics necessary for anterior, as well as posterior, restorations.1,2 In many situations all-ceramic crowns have a potential to be more esthetic than metal ceramic (MC) restorations. For the MC crown, the alloy structure produces an opaque appearance, and the metal margins are often visible. High-gold–content alloys are relatively expensive, and the alternatives may have disadvantages, such as risk of metal allergy, bond failure, or porcelain discoloration, and noble metal content alone is not indicative of clinical performance.3 Reinforced all-ceramic crowns consist of a highstrength porcelain core material, laminated with dentin and incisal porcelain.4 Successful performance and reliability of veneered ceramic prostheses may be limited THE JOURNAL OF PROSTHETIC DENTISTRY 349
THE JOURNAL OF PROSTHETIC DENTISTRY
AL-DOHAN ET AL
and (8) environmental effects.11 In the present study, a crosshead speed of 0.5 mm/min was chosen. It has been hypothesized that relatively high crosshead speeds may develop abnormal stress distributions during the shear test, inducing cohesive failure in either of the bonded materials, which would influence the bond strength values achieved. According to Hara et al,12 crosshead speeds of 0.50 and 0.75 mm/min result in more adhesive failures and are therefore preferable in shear bond strength tests. The purpose of this study was to develop a protocol to investigate the average shear strength of the core-veneer interface in bilayered all-ceramic systems and to use that test to compare metal ceramic to all-ceramic systems.
MATERIAL AND METHODS Fig. 1. Master die and core.
by mechanical integrity and adhesion of the veneering porcelain to the ceramic substrates. The mechanical properties of core materials and veneering porcelains should match to a certain extent to achieve a durable bond.5 According to the cohesive plateau theory, a proper bond is one that is stronger than the materials joined together.6 Previous studies testing the porcelainto-metal bond reported shear strength equal to the shear strength of the veneering porcelain as being an adequate bond.7 However, if the metallic understructure is eliminated, the bond to the underlying porcelain core becomes an important issue. Without documented evidence of the strength of the bond between the core and veneering porcelain, the profession must rely on manufacturers’ claims to judge which material is best for patients. The bond between a metal core and veneering porcelain has been investigated using bending tests.8,9 While this type of test may be appropriate for ceramic metal systems, it cannot be applied to a ceramic core due to its brittle nature. The fracture mechanism of ceramics and metals are different due to their different structure and mechanism of bonding. Covalent and ionic bonds in ceramics are associated with large interatomic forces and hence produce a stronger resistance to plastic deformation than metals.10 External loads will therefore cause stress concentration at a crack tip instead of relaxation by plastic flow.10 There are several factors that are associated with the stress state created in dental ceramic restorations, including: (1) thickness of ceramic layers, (2) mechanical properties of each ceramic, (3) elastic modulus of the supporting substrate material, (4) direction, magnitude, and frequency of applied load, (5) size and location of occlusal contact areas, (6) residual stresses induced by processing, (7) restoration-cement interfacial defects, 350
Specimen preparation A master die was fabricated from dental stone (Jade Stone; Whip Mix Corp, Louisville, Ky). The stone die was a negative replica of a previously fabricated ceramic core that was 8 mm high with a flat occlusal surface, and 10 mm in diameter with rounded occlusal line angles and a 1-mm–deep circumferential chamfer (Fig. 1). The all-ceramic systems tested and the manufacturerrecommended veneering porcelains are listed in Table I. Sixty specimens, 12 for each group, were made from the master die. A 2.4-mm–diameter cylinder of the veneering porcelain was applied using a specially designed aluminum split mold (Fig. 2). Due to porcelain shrinkage, a total of 3 separate firings were required to establish the correct diameter.
Mounting The completed ceramic specimen was inverted on a glass slab with a 4-mm–diameter opening (Fig. 3). The coping was fixed in position using sticky wax (Kerr Corp, Orange, Calif), and a brass ring, 30 mm in diameter and 30 mm in height, was placed over the coping on the glass slab and held in position with glue (Quicktite; Manco, Inc, Avon, Ohio). A die-stone (Jade Stone; Whip Mix Corp) was mixed and poured into the ring as illustrated in Figure 3.
Testing Each specimen was mounted in a metal holder on the universal testing machine (Instron no. 5565; Instron Corp, Canton, Mass) and the load was applied with the jig (Custom-made; Ultradent Inc, South Jordan, Utah) that had a diameter of 2.4 mm corresponding to the diameter of the veneering porcelain. Each specimen was tightened and stabilized to ensure that the edge of the shearing jig was touching the core surface and was positioned as close to the veneer-core interface as VOLUME 91 NUMBER 4
AL-DOHAN ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Table I. All-ceramic systems and control group used in study Group
IE
Group name
Empress2
Core Veneer
PZ
Procera allZircon
Core Veneer
DC
PA
DC Zircon
Procera allCeram
MC
Fabrication
IPS-Empress (ingot) IPS-Empress (IPS Eris)
Fabricate core from waxings on the master die Cylinder: 2.37mm in diameter, 3mm in height using specially designed split mold Fabricate cores to digitized specifications of master die Same as above
Procera allZircon Core Cerabien CZR
Core
DC-Zircon core
Veneer
Vita D opaque porcelain
Core Veneer
Control group
Material
Core Veneer
Procera allCeram Alumina Core Degussa-Ney All ceram dentine Lodestar Noritake
possible (Fig. 4). A shear load was applied at a crosshead speed of 0.50 mm/min until failure. The ultimate load to failure was recorded by the system’s software (Merlin; Instron Corp) in Newtons (N). Average shear strengths (MPa) were calculated by dividing the load (N) at which failure occurred by the bonding area (mm2): Shear stress ðMPaÞ ¼ Load ðNÞ 4 Area ðmm2 Þ The mean failure load and the SD for each group were calculated from these data.
Fabricate cores to digitized specifications of master die Same as above
Fabricate cores to digitized specifications of master die Same as above Fabricate basic core from waxups on master die Same as above
Manufacturer
Ivoclar Vivadent, Amherst, NY Ivoclar Vivadent
Procera Sandvik, Inc, Gothenberg, Sweden Noritakekizai Co Ltd, Nagoya, Japan TKT Metoxit AG, Thayngen, Switzerland Vita Zahnfabrik H. Rauter GmbH & Co KG, Sackinien, Germany Procera Sandvik, Inc Dentsply Int., York, Pa Ivoclar Vivadent Noritakekizai Co, Ltd
Statistical analysis Statistical analysis was carried out using statistical software (SAS software; SYSTAT Software Inc, Evanston, Ill). The data were analyzed using a 1-way analysis of variance test (ANOVA) to determine whether significant differences existed between the shear strengths of the 5 groups (a=.05). Also, a Tukey multiple comparisons test (a=.05) was used to assess the differences among the specified materials.
Microscopic examination The fractured surfaces were visually analyzed with a microscope (SMZ-U Stereoscopic Zoom Microscope; Nikon Corp, Tokyo, Japan) at original magnification 320. Sketches of the fractured surfaces were plotted and scanned to a computer. The fractured surfaces were then analyzed using a computer software program (Scion Image; Scion Corp, Frederick, Md) to determine the percentage of cohesive fracture within the veneer, cohesive fracture within the core, or adhesive failure at the core/veneer interface. This was done by tracing the borders of the cohesive veneer/core fracture that remained within the bonded interface and calculating the area using the Scion program. The percentage was calculated by dividing this area by the total bonded area (Fig. 5). The failed surfaces were classified as cohesive fracture within the veneer (V), adhesive fracture between the core and the veneer (C/V), and cohesive fracture within the core (C). APRIL 2004
RESULTS Table II shows the mean shear strengths of the coreveneer interface of the 4 all-ceramic groups and control group (MC). The highest mean shear strength was recorded for IPS-Empress2 bonded to Eris (30.86 6 6.47 MPa). The mean shear strength of Procera AllZircon bonded to Cerabien CZR was 28.03 6 5.03 MPa; DCZircon bonded to Vita D was 27.90 6 4.79 MPa; Procera alumina bonded to AllCeram veneering porcelain was 22.40 6 2.40 MPa. The 1-way ANOVA showed a significant difference for the shear strengths among the materials tested at P\.001 (Table III). The Tukey multiple comparisons test was computed to make all pairwise comparisons among the 4 groups in this study plus the control group. These comparisons are listed in the last column of Table II. The P values of the different comparisons show that IE, PZ, and DC were not significantly 351
THE JOURNAL OF PROSTHETIC DENTISTRY
AL-DOHAN ET AL
Fig. 3. Mounting technique.
Fig. 4. Testing using semicircular blade attached to universal testing machine.
Fig. 2. Veneer fabrication. a, Specially designed aluminum split mold holds core. b, Core secured via screws in both sides of split mold. c, Veneering porcelain placed inside mold (2.4 mm in diameter). d, As screws are loosened, 2 parts of mold are separated. e, Specimen removed and veneer is fired.
352
different and were comparable to the control group. Procera alumina had significantly lower values than the control, yet it was not significantly lower than PZ and DC. The highest values were observed with IE, and the lowest were observed with PA. The mean values for the fracture analysis were calculated and are provided in Table IV. All test groups demonstrated cohesive failure within the veneer, as well as adhesive failure between the core and the veneer. Only Empress2/Eris demonstrated cohesive failure within the core. Procera alumina showed more surface failure at the core-veneer interface. VOLUME 91 NUMBER 4
AL-DOHAN ET AL
THE JOURNAL OF PROSTHETIC DENTISTRY
Table II. Mean shear strength (MPa) of 4 groups of all-ceramic specimens and control group All ceramic system Group
Core
Sample size (n)
Mean (MPa)
SD
Sign*
Eris CZR
12 12
30.86 28.03
6 6.47 6 5.03
a ab
Vita D All Ceram
12 12
27.90 22.40
6 4.79 6 2.40
ab bc
Noritake
12
30.16
5.89
Veneer
IE PZ
Empress 2 Procera AllZircon DC DC-Zircon PA Procera alumina Control Metal group (Lodestar)
a
*Values with the same letter are not statistically different using Tukey test at P\.05.
Fig. 5. Sketch of fractured surface classified: cohesive within veneer (V), adhesive (C/V), and cohesive within core (C).
Table III. One-way ANOVA data of shear length of different all-ceramic specimens and control group Source
Sum of squares
df
Mean square
F-Ratio
P value
530.20 1434.65
4 55
132.55 26.09
5.08
.001*
Material Error
*Significant at 95% CI.
Table IV. Failed bonded surfaces divided by percentage: (V) cohesive within veneer, (C/V) combined surface failure, (C) cohesive within core Material group
MC (Control) Empress 2/Eris Procera AllZircon/CZR DC-Zircon/Vita D Procera alumina/AllCeram DC-Zircon/Eris
Fig. 6. Test scheme. A, Movement will be in one direction parallel to bonded surface. B, Failure due to bending moment.
DISCUSSION The currently accepted method of measuring shear bond strength for enamel and dentin adhesives was used in the present study. The shear bond strength (SBS) test is defined as a test in which 2 materials are connected via an adhesive agent and loaded in shear until separation occurs.13 The bond strength is calculated by dividing the maximum applied force by the bonded crosssectional area.14 This test is relatively simple and easy to perform, producing rapid results. However, some critical aspects must be considered in using an in vitro method to predict the clinical performance of adhesive materials. First, in vitro information cannot be extrapAPRIL 2004
V
V/C
C
56 42.8 41.6 58.6 30.3 27.9
43 8 57.4 40.4 68.7 71.1
1 49.2 1 1 1 1
olated directly to clinical situations, and second, the great variations in SBS test results have made the test questionable.15 An effort must therefore be made to standardize SBS test methods to improve the clinical usefulness of this in vitro test. Important aspects that should be considered include storage conditions, type of substrate, specimen preparation, rate of load application, cross-sectional surface area, and experience of the researcher. Establishing parameters for some of these issues was proposed in 1994 by an ISO standard.16 Studies of lap joints using mathematical stress analysis and finite element stress analysis demonstrate that the stresses at the interface between the adhesive and the substrate are not uniform and are highly dependent on the test geometry and loading configuration adopted.16 Van Noort et al17 demonstrated that the highest concentration of stress is seen at the loading point of the interface between the composite and dentin surface. 353
THE JOURNAL OF PROSTHETIC DENTISTRY
Van Noort et al17 compared a jig with a semicircular edge versus the straight knife edge. Since the initial contact of the knife is at one point on the edge of the bonded specimen, the stress is concentrated in a smaller area, resulting in premature failure. In theory, the curved knife wraps around the cylinder to be loaded, thereby contacting a larger area. It is thought that the curved knife distributes the stress over a larger surface area and minimizes load concentration. The data (Table II) illustrate that IE, IPS Eris veneering porcelain applied to IPS-Empress2 core, produced the highest values. Comparable results were found with both zirconia groups (PZ and DC) bonded to the veneering porcelain recommended by the manufacturer. AllCeram veneering porcelain applied to Procera alumina core (PA) demonstrated a weaker bond. During fabrication of the AllCeram/Procera alumina core specimens, a thin layer of opaque was applied in the first firing. This is the first step in the veneering sequence to improve the esthetic properties and mask the more yellowish color of the alumina core. The lower color value of this layer neutralizes the relatively bright aluminum oxide coping, which also provides the initial color to the restoration as well as bonding to the coping. It is very thin in consistency and when fired shrinks considerably. This might explain the weaker shear stress found in this group, since the opaque shrinks considerably after firing, leaving a very thin layer. The effect of this shrinkage was observed under high magnification where only remnants of the opaque were seen. Thus, when the load creates interfacial stresses, the crack propagation occurs in the weaker dentin layer near the core-veneer. An adequate bond for metal-ceramics has been determined to occur when the fracture stress is greater than 25 MPa.11 Adequate bond strength for all-ceramic materials has not been determined. Table II shows that the control group (MC) demonstrated higher (30.15 MPa) mean shear strength than recommended, and that IE, PZ, and DC were not significantly different from the control group. A study by Smith et al18 to gain insight into the fracture behavior of prostheses under incisal-directed, load-to-failure testing showed that 50% of the allceramic In-Ceram crowns failed by delamination of veneering glass alone, leaving a thin layer of residual glass on the core surface. The term ‘‘delamination’’ indicates complete debonding or adhesive failure, and microscopic examination of all the specimens in the present study substantiated this finding. Figure 6 explains the sequence of events the specimens may undergo during testing. In Figure 6, A, the applied stress exceeds the ultimate shear strength, and the veneer moves as a whole in the direction of the load, and complete debonding occurs. In this situation, the failure occurs due to shear stress. In Figure 6, B, the bond is stronger than the stress and resists the shear 354
AL-DOHAN ET AL
force. In this example, the applied tensile stress has exceeded the ultimate tensile stress; the veneer will rotate, and the failure will be a cohesive and adhesive combination. In Figure 2, Procera AllZircon/CZR group and DCZircon/Vita D group showed more veneer left on the core surface; failures were primarily within the veneer (55%-60%), and the remainder were at the interface with veneer residue left on the core surface. Also, the Empress2/ Eris group showed failure to be cohesive within both the veneer (45.83%) and core (52.5%). Empress2 core was the only material that was pulled out at the interface. A study by Wagner and Chu19 compared biaxial flexural strength and indentation fracture toughness of 3 ceramic core materials, Empress, In-Ceram, and Procera AllCeram. The results showed that the average flexural strength of Empress (134 MPa) and fracture toughness (1.74 MPam1/2) were significantly lower than the other two materials, Porcera AllCeram (687 MPa), and In-Ceram (352 MPa). Adhesive failure was not seen between the veneer and core due to the fact that the 2 materials fuse together and certain elements diffuse across the interface. This was explained by Smith et al18 in a study that investigated the possibility that elements unique to the core or the veneering glass diffused across the interface, or the possibility that a layer of excess infiltration glass developed on the core surface during processing. Either of these occurrences may cause a chemical alteration of the glass layer adjacent to the core, possibly changing physical properties at the interface, such as strength or coefficient of thermal expansion. Elemental analysis was performed for lanthanum, calcium, silicon, sodium, and potassium and showed that residual infiltration glass was not present on the core surface, and chemical alteration in the veneering glass was limited to 1 to 2 mm.18 Adhesive failure does not occur in the presence of a good bond between a compatible ceramic core and veneer material. Microscopic examination of Procera AllCeram showed failure to be primarily at the interface, with residue of the veneering porcelain remaining on the core. According to Oden et al,20 the strength of this veneering porcelain in combination with the aluminum oxide coping was demonstrated to be excellent. The veneering porcelains are chemically bonded to the densely sintered aluminum oxide by ionic and covalent bonds.20 Cohesive failure within the veneer was also seen, but to a somewhat lesser extent compared with the other groups.
CONCLUSIONS Within the limitations of this study, the following conclusions are made: 1. Eris veneering porcelain applied to IPS-Empress2 core showed the highest shear strength values and were VOLUME 91 NUMBER 4
AL-DOHAN ET AL
not significantly different from the metal ceramic (MC) control surface. 2. The bond of the veneering porcelains to the 2 zirconia cores was not significantly different from 1PSEmpress2/Eris or the MC control. 3. AllCeram applied to the Procera alumina core showed a significantly weaker bond compared to the other systems. Remnants of the opaque layer on the core were observed microscopically after failure. 4. Surface analysis of failure modes demonstrated that the bond between the core and the veneer was cohesive in the veneer and adhesive at the interface for most systems tested. REFERENCES 1. Tam LE, McComb D. Shear bond strengths of resin luting cements to laboratory-made composite resin veneers. J Prosthet Dent 1991;66: 314-21. 2. Malament KA, Grossman DG. The cast glass-ceramic restoration. J Prosthet Dent 1987;57:674-83. 3. O’Brien WJ. Dental materials and their selection. 3rd ed. Chicago: Quintessence; 2002. p. 200-9. 4. O’Brien WJ. Dental materials and their selection. 3rd ed. Chicago: Quintessence Int; 2002. p. 216-20. 5. Blatz MB. Long-term clinical success of all-ceramic posterior restorations. Quintessence Int 2002;33:415-26. 6. Gwinnett AJ. A new method to test the cohesive strength of dentin. Quintessence Int 1994;25:215-8. 7. Yamada NH. Dental porcelain: the state of the art. 1st ed. Los Angeles: University of Southern Calif School of Dentistry; 1977. p. 137-40. 8. Ozcan M. Fracture reasons in ceramic-fused-to-metal restorations. J Oral Rehabil 2003;30:265-9. 9. Baran G, Boberick K, McCool J. Fatigue of restorative materials. Crit Rev Oral Biol Med 2001;12:350-60.
APRIL 2004
THE JOURNAL OF PROSTHETIC DENTISTRY
10. Kvam K, Hero H, Oilo G. Fracture toughness measurements of some dental core ceramics: a methodologic study. Scand J Dent Res 1991; 99:527-32. 11. Craig RG. Mechanical properties in restorative dental materials. 11th ed. New York: Mosby; 2002. p. 551-92. 12. Hara AT, Pimenta LA, Rodrigues AL Jr. Influence of cross-head speed on resin-dentin shear bond strength. Dent Mater 2001;17:165-9. 13. Craig RG, Powers JM. Restorative dental materials. 11th ed. Mosby; 2002. p. 85. 14. Versluis A, Tantbirojn D, Douglas WH. Why do shear bond test pull out dentin? J Dent Res 1997;76:1298-307. 15. Hadavi F, Hey JH, Ambrose ER, Louie PW, Shinkewski DJ. The effect of dentin primer on the shear bond strength between composite resin and enamel. Oper Dent 1993;18:61-5. 16. International Organization for Standardization. ISO TR 11405, Dental Materials-Guidance on testing of adhesion to tooth structure. 1994. 17. Van Noort R, Noroozi S, Howard IC, Cardew G. A critique of bond strength measurements. J Dent 1989;17:61-7. 18. Smith TB, Kelly JR, Tesk JA. In vitro fracture behavior of ceramic and metal-ceramic restorations. J Prosthodont 1994;3:138-44. 19. Wagner WC, Chu TM. Biaxial flexural strength and indentation fracture toughness of three dental core ceramics. J Prosthet Dent 1996;76: 140-4. 20. Oden A, Andersson M, Krystek-Ondracek I, Magnusson D. Five-year clinical evaluation of Procera AllCeram crowns. J Prosthet Dent 1998; 80:450-6. Reprint requests to: DR PETER YAMAN DEPARTMENT OF CARIOLOGY, RESTORATIVE SCIENCES, UNIVERSITY OF MICHIGAN 1101 N. UNIVERSITY AVE ANN ARBOR, MI 48109 FAX: 734-936-1597 E-MAIL: [email protected]
AND
ENDODONTICS
0022-3913/$30.00 Copyright ª 2004 by The Editorial Council of The Journal of Prosthetic Dentistry doi:10.1016/j.prosdent.2004.02.009
355