d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.intl.elsevierhealth.com/journals/dema
Effect on in vitro fracture resistance of the technique used to attach lithium disilicate ceramic veneer to zirconia frameworks M. Schmitter a,∗ , M. Schweiger b , D. Mueller a , S. Rues a a b
Department of Prosthodontics, Section of Material Sciences, University of Heidelberg, Germany Ivoclar Vivadent AG, Schaan, Principality of Liechtenstein, Liechtenstein
a r t i c l e
i n f o
a b s t r a c t
Article history:
Objectives. This in vitro study should assess the fracture resistance of veneered zirconia-
Received 4 March 2013
based crowns with either luted or fused veneer.
Received in revised form 7 June 2013
Methods. Thirty-two identical zirconia frameworks (IPS e.max ZirCAD; Ivoclar/Vivadent),
Accepted 18 October 2013
were constructed (inLab 3.80; Sirona Dental Systems). All frameworks were veneered with CAD/CAM-fabricated lithium disilicate ceramic (IPS e.max CAD; Ivoclar/Vivadent). For half the crowns (n = 16) the veneer was luted to the framework (Multilink Implant;
Keywords:
Ivoclar/Vivadent); for the other it was fused (IPS e.max Crystall./Connect; Ivoclar/Vivadent).
Fracture load
Half of the specimens were then loaded until failure without artificial aging; the other
CAD/CAM
half underwent artificial aging before assessment of the ultimate load. To compare the
Zirconia
two techniques further, finite element analysis (FEA) and fractographic assessment using
Attachment technique
SEM and EDX analysis were conducted. Statistical assessment was performed by use of non-parametric tests. Results. Initial fracture forces were higher in the fusion group (mean: 1388 ± 190 N versus 1211 ± 158 N). All specimens were insensitive to artificial aging. FEA showed that tensile stresses in the veneer at the frame–veneer interface were much higher for crowns with luted veneer; this may be the reason for their lower fracture resistance. Fractographic analysis revealed that both fused and luted specimens had cohesive and adhesive fracture patterns which resulted in partial delamination of the veneer. Significance. Fused crowns are superior to luted crowns. Comparison of fracture resistance with the maximum loads which may occur clinically (Fmax = 600 N on one tooth) suggests both techniques might be used clinically, however. © 2013 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Because chipping has been identified as a major technical complication for veneered zirconia restorations [1,2], several means of improving the performance of these restorations
have been reported in recent years. In addition to optimization of firing procedure [3] and framework design [4], etc., it has been suggested that CAD/CAM production improves the mechanical properties of the veneer [5,6]. CAD/CAMproduced veneers have many fewer flaws than hand-layered veneers. There are two main reasons for this: first, the blanks
∗ Corresponding author at: Section of Material Sciences, University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. Tel.: +49 06221 566036; fax: +49 06221 565371. E-mail address: Marc [email protected] (M. Schmitter). 0109-5641/$ – see front matter © 2013 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dental.2013.10.008
123
2.2.
Manufacture of the crowns
One trained dental technician (D.M.) produced all the specimens to avoid any operator effect.
2.2.1.
Manufacture of the zirconia framework
An artificial molar was prepared and an impression was taken which was poured using gypsum in order to produce a die. This die was duplicated 32 times in CoCrMo alloy. On these dies, 32 identical zirconia frameworks (IPS e.max ZirCAD), 0.6 mm thick with an anatomically shaped occlusal
600 820/840 550/820 7:00 2:00 840 820 30 30
Vacuum V21 /V22 [◦ C] Vacuum V11 /V12 [◦ C] Exposure time H2 [min:s] Exposure time H1 [min:s] Temp. T2 [◦ C] Temp. T1 [◦ C]
2
1) CAD-on with luted veneer (Multilink Implant) 2) CAD-on with fused veneer (IPS e.max Crystall./Connect)
403
With regard to chipping and/or delamination of the veneer, the performance of all-ceramic molar crowns fabricated with the new CAD/CAM technique was of interest. The particular focus of the study, on the effect of attachment technique on fracture load, resulted in two test series:
Increase of temp. t2 [◦ /min]
Test series
Increase of temp. t1 [◦ /min]
2.1.
Closing time [min]
Material and methods
Working temp. [◦ C]
2.
Table 1 – Firing procedure for attachment of the veneer to the framework by use of fusion ceramic.
themselves are industrially produced and contain fewer material defects; second, during manual production of the veneer, imperfections will inevitably be created, even when production is performed accurately. To improve the fracture resistance of all-ceramic crowns further, changeover from feldspathic ceramic to lithium disilicate ceramic for production of the veneer has been suggested, because of the superior mechanical properties of lithium disilicate ceramic [7]. Furthermore, frame and veneer were milled from ceramic blanks by use of the so-called “CAD-on” technique. After sintering of the zirconia framework, the lithium disilicate veneer can be attached to the framework by use of resin cement or by fusion of the ceramic. Although fusion is recommended by the manufacturer, it might be relevant to determine the effect of attachment technique on fracture resistance, because, e.g., in the dental laboratory or even chair-side luting is often preferred because of rapid and easy handling. However, the effect of these two approaches on the fracture resistance of the crowns is unknown. As recommended by Anusavice et al. [8], several approaches were included in this study to enable comprehensive analysis of the failures. FE computations enabled assessment of the effect of interface layer stiffness on stresses occurring within the specimens during simulated loading [9]. Fractographic analysis enabled identification of the site of initiation of the fracture [10] and EDX enabled adhesive and cohesive failure to be distinguished [11] and aided identification of the location (on core or veneer) of attachment material after failure. The results of all these methods were combined to enable better understanding of the effect of the material used to attach the veneer to the core. The hypothesis of this study was that, with regard to load to failure, crowns with fused veneer would outperform crowns with luted veneer.
Cooling [◦ C]
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
124
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
area, were constructed (Sirona inLab 3.80). The frames were then milled (Sirona inLab MC XL) and sintered (Vita ZYrcomat). After sintering, all frameworks were checked for flaws, etc., by light microscopy (Stemi SR; Carl Zeiss, Oberkochen). Also, for each frame, the thickness at five previously defined points was measured by use of calipers.
2.2.2.
Construction of the veneer
After construction of the veneer by use of inLab 3.80 (Sirona), the mesio-lingual cusp was modified by use of Rapidform Dental software (INUS Technology, Seoul, Korea) as follows: 1) constant 30◦ angle between the occlusal surface of the cusp and the tooth axis, and 2) construction of a mold (radius: 3.0 mm, depth: 0.1 mm) defining the loading site for mechanical testing. The thickness of the veneer at the mesio-lingual cusp was approximately 2.0 mm.
2.2.3.
Milling of the veneer
The modified veneer was milled (inLab MC XL; Sirona) from a lithium disilicate ceramic blank (IPS e.max CAD). After this process, wall thickness readings were checked and recorded to ensure correct geometric fabrication.
2.2.4.
Attachment of the veneer to the framework
In one group (luting group; n = 16) the veneer was attached to the framework by use of Multilink Implant after separate crystallization and glaze firing of the veneer shell. Multilink Implant was used in accordance with the manufacturer’s guidelines: etching of the veneer for 20 s with 5% hydrofluoric acid (IPS Ceramic Etching Gel; Ivoclar Vivadent) and application of Monobond Plus (Ivoclar Vivadent). The zirconia frameworks were sandblasted (P-G400; Harnisch & Rieth,
Winterbach, Germany, aluminum oxide, 50 m, 2.0 bar) and cleaned for 2 min with 95% isopropyl alcohol in an ultrasonic bath (Sonorex super RK 102H; Bandelin, Berlin, Germany). After cementation, the specimens were stored at 37 ± 1 ◦ C for 24 h. In the second group (fusion group; n = 16) the veneer was attached to the framework in a furnace (Programat 700; Ivoclar Vivadent) by use of IPS e.max Crystall./Connect. The fusion ceramic was used in accordance with the manufacturer’s guidelines; details of the firing procedure are given in Table 1. This method was no more time-consuming than the first, because crystallization of the lithium disilicate veneer and fusion to the zirconia framework were performed in one firing.
2.3.
In vitro tests
The purpose of this study was to assess the effect of different attachment techniques on forces leading to delamination (adhesive failure at the frame–veneer interface) and/or chipping (cohesive failure within the veneer ceramic) of CAD-on molar crowns. Mechanical tests were performed by applying a load to the mesio-lingual cusp at an angle of 30◦ to the direction of insertion (Fig. 1a). For chewing simulation and fracture testing the metal dies were embedded in steel molds at an angle of 30◦ to the direction of insertion with cold-curing resin (Technovit 4071; Heraeus Kulzer GmbH, Hanau, Germany). The placement of the metal dies could be reproduced exactly by use of a special metal key, i.e. a CAM-manufactured metal block containing the negative form of the prepared tooth surface with 30◦ inclination, orienting the metal dies in the desired direction during use of a parallelometer table. Consequently, the loaded cusp was oriented horizontally (Fig. 1a) after the crowns were luted to the metal dies with Panavia F 2.0 (Kuraray
Fig. 1 – (a) Schematic diagram of the mechanical tests. The load was applied to the mesio-lingual cusp at an angle of 30◦ to the direction of insertion. (b) In vitro test setup. (c) FE model of the tested crown, with force application by contact of a descending steel sphere.
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Medical Inc. Japan). Before in vitro testing the specimens were stored for 24 h at 37 ± 1 ◦ C. For each of the two attachment techniques, two series of tests were conducted (n = 8 for each series). In the first series, the initial fracture resistance (Zwick/Roell Z005, Ulm, Germany; crosshead speed of 0.5 mm/min., force shutdown threshold of 100 N) was determined directly after manufacture of the crowns whereas the second series underwent artificial aging (10,000 thermal cycles between 6.5 and 60 ◦ C, dwell time 45 s; Willytec THE-1100; SD Mechatronik, Feldkirchen, Germany) and 1.2 million chewing cycles with a force magnitude of Fmax = 108 N, water storage (Willytec CS3) before the final fracture tests were performed. Steel spheres 6 mm in diameter were used as antagonists for both chewing simulation and fracture tests. The test setup is illustrated in Fig. 1b. The crowns were inspected after each step (manufacture, thermal cycling, and chewing simulation) under a microscope (Stemi SR; Carl Zeiss, Oberkochen) to detect formation of potential flaws or cracks.
2.4.
Finite element computations
Because the stiffness of the interfacial layer between frame and veneer will affect stress distribution, finite element computations (ANSYS 13.0) were performed with different interface materials on the basis of the original geometry (Fig. 1c). The Young’s modulus of the interface layer was varied, i.e., E = 3.5 GPa for Multilink Implant, and E = 70 GPa for IPS e.max Crystall./Connect (material properties in accordance with manufacturers’ data). The properties of all the linear–elastic materials are listed in Table 2. The FE model consisted of approximately 65,000 20-node hexahedral elements and
Table 2 – Properties of the materials used in the FE analysis. Material
e.max CAD (Ivoclar Vivadent)a Fusion ceramic (Ivoclar Vivadent)a e.max ZirCAD (Ivoclar Vivadent)a Multilink CoCrMo alloya Dentin a
Fig. 2 – Whisker and box plots of fracture loads of specimens with and without artificial aging.
Statistics
Means, standard deviations, and medians were analyzed by use of descriptive statistics. tests (Kruskal–Wallis test and Non-parametric Mann–Whitney U-test with ˛ < 0.05 as significance level) were used to assess the effect of the different attachment techniques on fracture load. The results were depicted by use of whisker and box plots. A post hoc power analysis was performed using the program “PS power and sample size calculation V 2.1.30” [12].
2.5.
125
Young’s modulus [GPa]
According to information provided by the manufacturer.
2.6.
Fractographic analysis and EDX
Fractographic analysis was performed by use of an FE SEM (Field Emission Scanning Electron Microscope; Supra 40 VP; Zeiss, Germany) equipped with an EDX detector (Energy Dispersive X-ray Spectroscopy). Before analysis the fracture surfaces were cleaned with ethanol in an ultrasonic bath.
3.
Results
3.1.
Initial fracture resistance
The ultimate loads (mean value and standard deviation) were:
Poisson’s ratio [–]
95 70 210 18.6 200 18
240,000 nodes. The CoCrMo stump was assumed to be rigid, i.e., all displacements of nodes located on the inner surface (stump–cement interface) were restricted. As a consequence, the stump was not included in the computations. Loading was performed by contact (area-to-area contact formulation) with a steel sphere (6 mm diameter), by analogy with loading in the in vitro tests (Fig. 1b). For computations for each interface material, the corresponding mean ultimate loads Fu,m from the in vitro tests were chosen as final loads, i.e. at this load critical stresses should be seen within the restoration. Thermal residual stress which might occur to some extent within the fused crowns were not included in the FE analysis.
0.20 0.21 0.26 0.28 0.30 0.30
• crowns attached by use of resin cement Fu = 1211 ± 158 N (Fu,min = 904 N, Fu,max = 1412 N) • crowns attached by use of fusion ceramic Fu = 1388 ± 190 N (Fu,min = 1118 N, Fu,max = 1677 N)
(n = 8): (n = 8):
There was no significant difference between the groups (Mann–Whitney U-test, p = 0.105). The power of this test was 0.4. The results are given in Fig. 2.
126
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Fig. 3 – Representative examples of the fracture modes. (a) Fusion ceramic. (b) Resin cement.
3.2.
Artificial aging and tests after artificial aging
3.2.1.
Chewing simulation
During chewing simulation and thermal cycling, no specimen failed.
3.2.2.
Remaining fracture resistance
Force values recorded during fracture testing of crowns surviving thermal cycling and chewing simulation were: • crowns attached by use of resin cement Fu = 1226 ± 290 N (Fu,min = 765 N, Fu,max = 1545 N) • crowns attached by use of fusion ceramic Fu = 1492 ± 206 N (Fu,min = 1117 N, Fu,max = 1674 N)
(n = 8): (n = 8):
The difference between these values was significant (Mann–Whitney U-test; p = 0.028). The results are displayed in Fig. 2. In comparison with the initial fracture resistance, fracture forces for fused crowns were approximately the same after aging. In contrast, the lowest observed fracture force for luted crowns decreased by approximately 140 N.
3.2.3.
Macroscopic fracture appearance
In both luting and fusion groups, adhesive failures (delamination) of the veneer were predominant. Delamination areas were smaller for the fusion group than for the luting group. Occasionally, especially for high fracture forces, damage of the margin of the zirconia framework occurred in the fusion group; this was not observed in the luting group (Fig. 3a and b).
3.3.
Effect of the stiffness of the interface layer
Tensile stresses are of most interest when brittle materials fracture. As is apparent from Fig. 4, tensile stresses within the veneer, frame, and interface layer of the undamaged restoration loaded with a critical value of Fu,m were highly affected by the stiffness of the interface layer. In a radial section including the center of the loading area, lower tensile and compressive stresses were observed within the zirconia framework of luted crowns when compared to fused crowns, because a more resilient interface layer leads to more uniform
load transfer along the interface area. However, the tensile stresses in the zirconia framework never exceeded 300 MPa, which is approximately one third of the material’s strength. Hence, fracture of the restoration originating from the frame is highly improbable. For the fused crowns, tensile stresses within the fusion layer reached critical values I > 100 MPa whereas for the luted crowns critical tensile stresses ( I = 375 MPa, II = 260 MPa, strength of the veneering material: u,veneer = 360 MPa) occurred at the inner surface of the veneer. As a consequence, in both cases, fracture may originate from the interface layer. However, high tensile stresses were also present in the vicinity of the loading area, so a fracture starting from the loading site is also probable.
3.4.
Fractographic analysis by SEM and EDX
Fractographic analysis by SEM showed that for fused crowns a mixed failure mode (cohesive and adhesive failure of the fusion ceramic) occurred (Fig. 5a). The site of fracture initiation was identified in different specimens as either in the occlusal surface area (Fig. 5b) of the veneer or in the interface between veneer and core (Fig. 5c). There were no differences between aged and non-aged specimens in the fusion group. For the luted non-aged crowns predominately adhesive failures between the core and Multilink implant were observed (Fig. 6a). After aging, for the luted crowns adhesive failures between the veneer and Multilink implant were observed. The presence of Multilink Implant cement on the adhesively fractured surface was detected by EDX. The elements Ba and Yb are characteristic indicators of cement composition. The sites of fracture initiation for these crowns were comparable with those for the fused crowns (Figs. 5b and 6b).
4.
Discussion
In recent years it has become obvious that cohesive and adhesive failures of zirconia-based, veneered (hand-layered) restorations often occur [13,14]. Several approaches have therefore been assessed as possible means of reducing this high occurrence of failure. Although some of these approaches have been shown, in vitro, to reduce the incidence of veneer failure, additional effort is necessary to
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
127
Fig. 4 – Principal stresses I , II , and III for all-ceramic crowns with the veneer fused (left) or luted (right) to the zirconia framework. In both cases, the mean ultimate load Fu,m obtained from in vitro tests with aged and non-aged crowns was applied.
improve the performance of these restorations further. In this context CAD/CAM-based manufacture of the restoration might be a promising technique. Beside the use of monolithic restorations based on lithium disilicate [15] or zirconia [16], the CAD/CAM based fabrication of the veneer and attachment of the veneer to the framework by use of different techniques have been proposed [5,6]. One advantage of CAD/CAM is the use of industrially produced ceramic blanks with fewer faults than manually produced veneer [17]. Use of lithium disilicate veneer has also been suggested, because the mechanical properties of this material are superior to those of feldspathic ceramic [7], i.e., lithium disilicate is approximately three times as strong as feldspathic ceramics. However, larger restorations (e.g. long-span fixed dental prosthesis) cannot be produced by monolithic lithium disilicate because of the mechanical properties of the material. In these cases veneered or non-veneered (monolithic) zirconia restorations have to be used if the patient desires all-ceramic restorations. However, both over-pressed and hand-layered zirconia restorations show reduced fracture loads [18], although the over-pressed zirconia restorations show higher fracture loads than the hand-layered ones [19]. Furthermore, for monolithic zirconia restorations there is a lack of clinical studies assessing the wear of antagonistic teeth (although in vitro
studies showed promising results [16]). Additionally, esthetic considerations might limit the use of monolithic zirconia restorations. Thus, the technique presented in the present study might be an auspicious way to restore or replace decayed teeth. Previous in vitro testing [6] of manually layered crowns with the same design as used in this study revealed a substantial effect of artificial aging—7 of 8 crowns failed during chewing simulation with Fmax = 108 N. In the work reported here, all crowns survived chewing simulation, and the difference between fracture force for aged and non-aged crowns was rather small, especially in the fusion group. Therefore, crowns consisting of a zirconia framework and CAD/CAM-produced lithium disilicate ceramic veneer were almost insensitive to the artificial aging process used in this study. The fracture resistance, Fu , of aged and non-aged crowns with the veneer fused with IPS e.max Crystall./Connect was in the range 1100–1700 N. The aged crowns with luted veneer fractured at significantly lower forces than the fused crowns and a small decrease of minimum fracture resistance because of aging was also observed, although this minimum fracture resistance, Fu,min = 765 N, still exceeds the maximum load of approximately 500–700 N which can be expected to act on a single tooth in the molar region in vivo.
128
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Fig. 5 – Fractographic analysis of fusion ceramic after initial fracture. (a) Overview of fracture surface, with fusion ceramic adhering to veneer. (b) Loading area with possible fracture initiation. (c) Crack formation within the fusion.
FE analysis revealed that tensile stress equaling the material’s strength (360 MPa for e.max CAD) was reached at the inner surface of veneer with soft resin cement as interface material when the crown was loaded with Fu,m = 1218 N. In contrast, crowns with a hard interface layer made of fusion ceramic had critical tensile stress values within the interface layer which reached Fu,m = 1440 N. Stresses in the vicinity of the contact region were not included in the FE evaluation, because small imperfections will have a large effect on the
computed results. Next to the loading site, however, high tensile stresses acting along the outer surface of the crown could initiate cone cracking [20] within the veneer. The use of fractography to identify the origin of the fracture has been recommended by several authors [8,11]. In this study fractographic analysis revealed that the sites of fracture initiation in both groups are the loading point and the interface between the veneer and the core. The high stress concentration beneath the loading area lead to crack initiation
Fig. 6 – Fractographic analysis of resin-cemented restoration after initial fracture ceramic. (a) Overview of fracture surface, cement adhering to glass-ceramic veneer. (b) Loading area with possible fracture initiation.
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
due to Hertzian cone crack formation. This is in accordance with the finding of the FE analysis which came to the same result. Additionally, fractographic analysis could show in some specimens a secondary flaw system within the fusing ceramic, which reduces the stiffness of the material. These flaws might be a result of thermal residual stress of the fusion ceramic which were not considered in the FEA analysis. The combination of FEA and fractographic analysis provided a deeper understanding of the failure of the specimens, as reported by Anusavice [8], and has therefore be recommended. Although the manufacturer recommends fusing the veneer with the core, luting of the veneer could be desirable for two reasons: first, it is much easier, so it could be performed chairside in the dental practice; second, the frame and the veneer can be sintered at the same time, which can save time, in contrast with fused crowns, for which the frame must be sintered first. Both aspects might justify use of luting instead of the fusion technique. The mechanical properties of resin cement are, however, inferior to those of fusion ceramic. This might, as shown by the finite element analysis, have a negative effect on the fracture load of these restorations, because the point of initiation of the fracture might be located at the ceramic–composite interface. In a previous study it was shown that zirconia crowns veneered with lithium disilicate ceramic outperform zirconia crowns manually veneered with feldspathic ceramic [6]. Although esthetic aspects of feldspathic veneer might be slightly superior to those of lithium disilicate veneer, both materials are suitable for production of appealing dental restorations [21]. Clinically, chipping is often observed after the restoration has been in situ for some time, i.e., aging has occurred [1]. This study showed that both attachment techniques are relatively insensitive to aging, because initial fracture loads were comparable with fracture loads after artificial aging, especially for fused crowns. For manually layered feldspathic ceramic, results were completely different: artificial aging drastically reduces the fracture resistance of veneered zirconia crowns [6]. Both initial fracture loads and fracture loads after artificial aging should, however, exceed expected chewing forces, which are lower than 700 N in the posterior region [22]. In this study this was achieved for both techniques—the lowest value (Fu,min = 765 N) was measured for one specimen in the luting group. For all other specimens fracture resistance, Fu , was >900 N. Only one cusp was loaded in the fracture tests because, eventually, one or a few unique critical force applications during the entire lifetime of a ceramic restoration will cause technical failure, i.e., the critical load case does not have to fit the regular loading scheme. For the aging process, however, regular loading conditions should be simulated. Here, the applied load simulates part of a vertical biting force which is transferred to several cusps. The regular vertical bite force corresponding to our test setup for periodic loading on n cusps with 30◦ inclination is, then, given by: n × 108 N × cos 30◦ which is either 281 N or 374 N for n = 3 or n = 4. In this study a metal die was used instead of a natural tooth and the stump was assumed to be rigid in FE computations. Rosentritt [23] found that the die material has a significant effect on the fracture resistance of all-ceramic crowns. In a
129
recent study, however, it was shown by finite element analysis that for our test setup the stump’s stiffness had almost no effect on the stresses in the critically loaded structures (veneer and interface layer) because the zirconia framework does not fracture [6] and “shields” the veneer from the die. In addition, Rosentritt used axial loading, which led to complete fracture of the crown, including the Empress 2 framework (lithium disilicate ceramic) for which the stresses were highly dependent on the properties of the underlying material. Consequently, use of metal dies seems to be adequate for our test setup, and fracture resistance results similar to the situation in vivo can be expected. Because in vitro experiments always include simplification of the real situation and reduction to a few selected load cases and other variables, however, transfer to a clinical setting is always a challenge. For subjects with bruxism and/or in the event of exceptional loading conditions, especially, the risk of failure might be underestimated in this study. In addition, flaws generated either by the dental technician (inaccurate processing, sharp edges) or the dentist (inappropriate preparation, grinding of the occlusal surface) may lead to lower fracture resistance. To summaries, the hypothesis of this study must be accepted. However, both techniques seem to be acceptable for clinical use because of the high fracture resistance.
5.
Conclusion
Although both attachment techniques seem acceptable for use in a clinical setting, fusing the veneer increased the load to failure.
Acknowledgments The authors would like to thank Ivoclar/Vivadent for the material supply and the fractographic analysis. Additionally, we would like to thank Ian Davies for the proof-reading of the manuscript.
references
[1] Al-Amleh B, Lyons K, Swain M. Clinical trials in zirconia: a systematic review. J Oral Rehabil 2010;73:641–52. [2] Triwatana P, Nagaviroj N, Tulapornchai C. Clinical performance and failures of zirconia-based fixed partial dentures: a review literature. J Adv Prosthodont 2012;4:76–83. [3] Rues S, Kroger E, Muller D, Schmitter M. Effect of firing protocols on cohesive failure of all-ceramic crowns. J Dent 2010;38:987–94. [4] Rosentritt M, Steiger D, Behr M, Handel G, Kolbeck C. Influence of substructure design and spacer settings on the in vitro performance of molar zirconia crowns. J Dent 2009;37:978–83. [5] Beuer F, Schweiger J, Eichberger M, Kappert HF, Gernet W, Edelhoff D. High-strength CAD/CAM-fabricated veneering material sintered to zirconia copings—a new fabrication mode for all-ceramic restorations. Dent Mater 2009;25:121–8. [6] Schmitter M, Mueller D, Rues S. Chipping behaviour of all-ceramic crowns with zirconia framework and CAD/CAM manufactured veneer. J Dent 2012;40:154–62.
130
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
[7] Baltzer A. All-ceramic single-tooth restorations: choosing the material to match the preparation—preparing the tooth to match the material. Int J Comput Dent 2008;11:241–56. [8] Anusavice KJ, Kakar K, Ferree N. Which mechanical and physical testing methods are relevant for predicting the clinical performance of ceramic-based dental prostheses? Clin Oral Implants Res 2007;18(Suppl. 3):218–31. [9] Yettram AL, Wright KW, Pickard HM. Finite element stress analysis of the crowns of normal and restored teeth. J Dent Res 1976;55:11–004. [10] Kelly JR, Campbell SD, Bowen HK. Fracture-surface analysis of dental ceramics. J Prosthet Dent 1989;62:536–41. [11] Derand T, Hero H. Bond strength of porcelain on cast vs. wrought titanium. Scand J Dent Res 1992;100:184–8. [12] Dupont WD, Plummer WD. PS power and sample size program available for free on the Internet. Controlled Clin Trials 1997;18:274. [13] Silva NR, Bonfante EA, Zavanelli RA, Thompson VP, Ferencz JL, Coelho PG. Reliability of metalloceramic and zirconia-based ceramic crowns. J Dent Res 2010;89:6–1051. [14] Sailer I, Pjetursson BE, Zwahlen M, Hammerle CH. A systematic review of the survival and complication rates of all-ceramic and metal-ceramic reconstructions after an observation period of at least 3 years. Part II: Fixed dental prostheses. Clin Oral Implants Res 2007;18(Suppl. 3):86–96. [15] Kern M, Sasse M, Wolfart S. Ten-year outcome of three-unit fixed dental prostheses made from monolithic lithium disilicate ceramic. J Am Dent Assoc 2012;143:234–40.
[16] Stawarczyk B, Ozcan M, Schmutz F, Trottmann A, Roos M, Hammerle CH. Two-body wear of monolithic, veneered and glazed zirconia and their corresponding enamel antagonists. Acta Odontol Scand 2012;71:102–12. [17] Christensen RP, Ploeger BJ. A clinical comparison of zirconia, metal and alumina fixed-prosthesis frameworks veneered with layered or pressed ceramic: a three-year report. J Am Dent Assoc 2010;141:29–1317. [18] Eisenburger M, Mache T, Borchers L, Stiesch M. Fracture stability of anterior zirconia crowns with different core designs and veneered using the layering or the press-over technique. Eur J Oral Sci 2011;119:253–7. [19] Stawarczyk B, Ozcan M, Roos M, Trottmann A, Sailer I, Hammerle CH. Load-bearing capacity and failure types of anterior zirconia crowns veneered with overpressing and layering techniques. Dent Mater 2011;27:53–1045. [20] Lawn BR, Deng Y, Lloyd IK, Janal MN, Rekow ED, Thompson VP. Materials design of ceramic-based layer structures for crowns. J Dent Res 2002;81:433–8. [21] Culp L, McLaren EA. Lithium disilicate: the restorative material of multiple options. Compend Contin Educ Dent 2010;31, 716–20, 22, 24–5. [22] Helkimo E, Carlsson GE, Helkimo M. Bite force and state of dentition. Acta Odontol Scand 1977;35:297–303. [23] Rosentritt M, Plein T, Kolbeck C, Behr M, Handel G. In vitro fracture force and marginal adaptation of ceramic crowns fixed on natural and artificial teeth. Int J Prosthodont 2000;13:387–91.
Available online at www.sciencedirect.com
ScienceDirect journal homepage: www.intl.elsevierhealth.com/journals/dema
Effect on in vitro fracture resistance of the technique used to attach lithium disilicate ceramic veneer to zirconia frameworks M. Schmitter a,∗ , M. Schweiger b , D. Mueller a , S. Rues a a b
Department of Prosthodontics, Section of Material Sciences, University of Heidelberg, Germany Ivoclar Vivadent AG, Schaan, Principality of Liechtenstein, Liechtenstein
a r t i c l e
i n f o
a b s t r a c t
Article history:
Objectives. This in vitro study should assess the fracture resistance of veneered zirconia-
Received 4 March 2013
based crowns with either luted or fused veneer.
Received in revised form 7 June 2013
Methods. Thirty-two identical zirconia frameworks (IPS e.max ZirCAD; Ivoclar/Vivadent),
Accepted 18 October 2013
were constructed (inLab 3.80; Sirona Dental Systems). All frameworks were veneered with CAD/CAM-fabricated lithium disilicate ceramic (IPS e.max CAD; Ivoclar/Vivadent). For half the crowns (n = 16) the veneer was luted to the framework (Multilink Implant;
Keywords:
Ivoclar/Vivadent); for the other it was fused (IPS e.max Crystall./Connect; Ivoclar/Vivadent).
Fracture load
Half of the specimens were then loaded until failure without artificial aging; the other
CAD/CAM
half underwent artificial aging before assessment of the ultimate load. To compare the
Zirconia
two techniques further, finite element analysis (FEA) and fractographic assessment using
Attachment technique
SEM and EDX analysis were conducted. Statistical assessment was performed by use of non-parametric tests. Results. Initial fracture forces were higher in the fusion group (mean: 1388 ± 190 N versus 1211 ± 158 N). All specimens were insensitive to artificial aging. FEA showed that tensile stresses in the veneer at the frame–veneer interface were much higher for crowns with luted veneer; this may be the reason for their lower fracture resistance. Fractographic analysis revealed that both fused and luted specimens had cohesive and adhesive fracture patterns which resulted in partial delamination of the veneer. Significance. Fused crowns are superior to luted crowns. Comparison of fracture resistance with the maximum loads which may occur clinically (Fmax = 600 N on one tooth) suggests both techniques might be used clinically, however. © 2013 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
1.
Introduction
Because chipping has been identified as a major technical complication for veneered zirconia restorations [1,2], several means of improving the performance of these restorations
have been reported in recent years. In addition to optimization of firing procedure [3] and framework design [4], etc., it has been suggested that CAD/CAM production improves the mechanical properties of the veneer [5,6]. CAD/CAMproduced veneers have many fewer flaws than hand-layered veneers. There are two main reasons for this: first, the blanks
∗ Corresponding author at: Section of Material Sciences, University of Heidelberg, Im Neuenheimer Feld 400, 69120 Heidelberg, Germany. Tel.: +49 06221 566036; fax: +49 06221 565371. E-mail address: Marc [email protected] (M. Schmitter). 0109-5641/$ – see front matter © 2013 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.dental.2013.10.008
123
2.2.
Manufacture of the crowns
One trained dental technician (D.M.) produced all the specimens to avoid any operator effect.
2.2.1.
Manufacture of the zirconia framework
An artificial molar was prepared and an impression was taken which was poured using gypsum in order to produce a die. This die was duplicated 32 times in CoCrMo alloy. On these dies, 32 identical zirconia frameworks (IPS e.max ZirCAD), 0.6 mm thick with an anatomically shaped occlusal
600 820/840 550/820 7:00 2:00 840 820 30 30
Vacuum V21 /V22 [◦ C] Vacuum V11 /V12 [◦ C] Exposure time H2 [min:s] Exposure time H1 [min:s] Temp. T2 [◦ C] Temp. T1 [◦ C]
2
1) CAD-on with luted veneer (Multilink Implant) 2) CAD-on with fused veneer (IPS e.max Crystall./Connect)
403
With regard to chipping and/or delamination of the veneer, the performance of all-ceramic molar crowns fabricated with the new CAD/CAM technique was of interest. The particular focus of the study, on the effect of attachment technique on fracture load, resulted in two test series:
Increase of temp. t2 [◦ /min]
Test series
Increase of temp. t1 [◦ /min]
2.1.
Closing time [min]
Material and methods
Working temp. [◦ C]
2.
Table 1 – Firing procedure for attachment of the veneer to the framework by use of fusion ceramic.
themselves are industrially produced and contain fewer material defects; second, during manual production of the veneer, imperfections will inevitably be created, even when production is performed accurately. To improve the fracture resistance of all-ceramic crowns further, changeover from feldspathic ceramic to lithium disilicate ceramic for production of the veneer has been suggested, because of the superior mechanical properties of lithium disilicate ceramic [7]. Furthermore, frame and veneer were milled from ceramic blanks by use of the so-called “CAD-on” technique. After sintering of the zirconia framework, the lithium disilicate veneer can be attached to the framework by use of resin cement or by fusion of the ceramic. Although fusion is recommended by the manufacturer, it might be relevant to determine the effect of attachment technique on fracture resistance, because, e.g., in the dental laboratory or even chair-side luting is often preferred because of rapid and easy handling. However, the effect of these two approaches on the fracture resistance of the crowns is unknown. As recommended by Anusavice et al. [8], several approaches were included in this study to enable comprehensive analysis of the failures. FE computations enabled assessment of the effect of interface layer stiffness on stresses occurring within the specimens during simulated loading [9]. Fractographic analysis enabled identification of the site of initiation of the fracture [10] and EDX enabled adhesive and cohesive failure to be distinguished [11] and aided identification of the location (on core or veneer) of attachment material after failure. The results of all these methods were combined to enable better understanding of the effect of the material used to attach the veneer to the core. The hypothesis of this study was that, with regard to load to failure, crowns with fused veneer would outperform crowns with luted veneer.
Cooling [◦ C]
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
124
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
area, were constructed (Sirona inLab 3.80). The frames were then milled (Sirona inLab MC XL) and sintered (Vita ZYrcomat). After sintering, all frameworks were checked for flaws, etc., by light microscopy (Stemi SR; Carl Zeiss, Oberkochen). Also, for each frame, the thickness at five previously defined points was measured by use of calipers.
2.2.2.
Construction of the veneer
After construction of the veneer by use of inLab 3.80 (Sirona), the mesio-lingual cusp was modified by use of Rapidform Dental software (INUS Technology, Seoul, Korea) as follows: 1) constant 30◦ angle between the occlusal surface of the cusp and the tooth axis, and 2) construction of a mold (radius: 3.0 mm, depth: 0.1 mm) defining the loading site for mechanical testing. The thickness of the veneer at the mesio-lingual cusp was approximately 2.0 mm.
2.2.3.
Milling of the veneer
The modified veneer was milled (inLab MC XL; Sirona) from a lithium disilicate ceramic blank (IPS e.max CAD). After this process, wall thickness readings were checked and recorded to ensure correct geometric fabrication.
2.2.4.
Attachment of the veneer to the framework
In one group (luting group; n = 16) the veneer was attached to the framework by use of Multilink Implant after separate crystallization and glaze firing of the veneer shell. Multilink Implant was used in accordance with the manufacturer’s guidelines: etching of the veneer for 20 s with 5% hydrofluoric acid (IPS Ceramic Etching Gel; Ivoclar Vivadent) and application of Monobond Plus (Ivoclar Vivadent). The zirconia frameworks were sandblasted (P-G400; Harnisch & Rieth,
Winterbach, Germany, aluminum oxide, 50 m, 2.0 bar) and cleaned for 2 min with 95% isopropyl alcohol in an ultrasonic bath (Sonorex super RK 102H; Bandelin, Berlin, Germany). After cementation, the specimens were stored at 37 ± 1 ◦ C for 24 h. In the second group (fusion group; n = 16) the veneer was attached to the framework in a furnace (Programat 700; Ivoclar Vivadent) by use of IPS e.max Crystall./Connect. The fusion ceramic was used in accordance with the manufacturer’s guidelines; details of the firing procedure are given in Table 1. This method was no more time-consuming than the first, because crystallization of the lithium disilicate veneer and fusion to the zirconia framework were performed in one firing.
2.3.
In vitro tests
The purpose of this study was to assess the effect of different attachment techniques on forces leading to delamination (adhesive failure at the frame–veneer interface) and/or chipping (cohesive failure within the veneer ceramic) of CAD-on molar crowns. Mechanical tests were performed by applying a load to the mesio-lingual cusp at an angle of 30◦ to the direction of insertion (Fig. 1a). For chewing simulation and fracture testing the metal dies were embedded in steel molds at an angle of 30◦ to the direction of insertion with cold-curing resin (Technovit 4071; Heraeus Kulzer GmbH, Hanau, Germany). The placement of the metal dies could be reproduced exactly by use of a special metal key, i.e. a CAM-manufactured metal block containing the negative form of the prepared tooth surface with 30◦ inclination, orienting the metal dies in the desired direction during use of a parallelometer table. Consequently, the loaded cusp was oriented horizontally (Fig. 1a) after the crowns were luted to the metal dies with Panavia F 2.0 (Kuraray
Fig. 1 – (a) Schematic diagram of the mechanical tests. The load was applied to the mesio-lingual cusp at an angle of 30◦ to the direction of insertion. (b) In vitro test setup. (c) FE model of the tested crown, with force application by contact of a descending steel sphere.
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Medical Inc. Japan). Before in vitro testing the specimens were stored for 24 h at 37 ± 1 ◦ C. For each of the two attachment techniques, two series of tests were conducted (n = 8 for each series). In the first series, the initial fracture resistance (Zwick/Roell Z005, Ulm, Germany; crosshead speed of 0.5 mm/min., force shutdown threshold of 100 N) was determined directly after manufacture of the crowns whereas the second series underwent artificial aging (10,000 thermal cycles between 6.5 and 60 ◦ C, dwell time 45 s; Willytec THE-1100; SD Mechatronik, Feldkirchen, Germany) and 1.2 million chewing cycles with a force magnitude of Fmax = 108 N, water storage (Willytec CS3) before the final fracture tests were performed. Steel spheres 6 mm in diameter were used as antagonists for both chewing simulation and fracture tests. The test setup is illustrated in Fig. 1b. The crowns were inspected after each step (manufacture, thermal cycling, and chewing simulation) under a microscope (Stemi SR; Carl Zeiss, Oberkochen) to detect formation of potential flaws or cracks.
2.4.
Finite element computations
Because the stiffness of the interfacial layer between frame and veneer will affect stress distribution, finite element computations (ANSYS 13.0) were performed with different interface materials on the basis of the original geometry (Fig. 1c). The Young’s modulus of the interface layer was varied, i.e., E = 3.5 GPa for Multilink Implant, and E = 70 GPa for IPS e.max Crystall./Connect (material properties in accordance with manufacturers’ data). The properties of all the linear–elastic materials are listed in Table 2. The FE model consisted of approximately 65,000 20-node hexahedral elements and
Table 2 – Properties of the materials used in the FE analysis. Material
e.max CAD (Ivoclar Vivadent)a Fusion ceramic (Ivoclar Vivadent)a e.max ZirCAD (Ivoclar Vivadent)a Multilink CoCrMo alloya Dentin a
Fig. 2 – Whisker and box plots of fracture loads of specimens with and without artificial aging.
Statistics
Means, standard deviations, and medians were analyzed by use of descriptive statistics. tests (Kruskal–Wallis test and Non-parametric Mann–Whitney U-test with ˛ < 0.05 as significance level) were used to assess the effect of the different attachment techniques on fracture load. The results were depicted by use of whisker and box plots. A post hoc power analysis was performed using the program “PS power and sample size calculation V 2.1.30” [12].
2.5.
125
Young’s modulus [GPa]
According to information provided by the manufacturer.
2.6.
Fractographic analysis and EDX
Fractographic analysis was performed by use of an FE SEM (Field Emission Scanning Electron Microscope; Supra 40 VP; Zeiss, Germany) equipped with an EDX detector (Energy Dispersive X-ray Spectroscopy). Before analysis the fracture surfaces were cleaned with ethanol in an ultrasonic bath.
3.
Results
3.1.
Initial fracture resistance
The ultimate loads (mean value and standard deviation) were:
Poisson’s ratio [–]
95 70 210 18.6 200 18
240,000 nodes. The CoCrMo stump was assumed to be rigid, i.e., all displacements of nodes located on the inner surface (stump–cement interface) were restricted. As a consequence, the stump was not included in the computations. Loading was performed by contact (area-to-area contact formulation) with a steel sphere (6 mm diameter), by analogy with loading in the in vitro tests (Fig. 1b). For computations for each interface material, the corresponding mean ultimate loads Fu,m from the in vitro tests were chosen as final loads, i.e. at this load critical stresses should be seen within the restoration. Thermal residual stress which might occur to some extent within the fused crowns were not included in the FE analysis.
0.20 0.21 0.26 0.28 0.30 0.30
• crowns attached by use of resin cement Fu = 1211 ± 158 N (Fu,min = 904 N, Fu,max = 1412 N) • crowns attached by use of fusion ceramic Fu = 1388 ± 190 N (Fu,min = 1118 N, Fu,max = 1677 N)
(n = 8): (n = 8):
There was no significant difference between the groups (Mann–Whitney U-test, p = 0.105). The power of this test was 0.4. The results are given in Fig. 2.
126
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Fig. 3 – Representative examples of the fracture modes. (a) Fusion ceramic. (b) Resin cement.
3.2.
Artificial aging and tests after artificial aging
3.2.1.
Chewing simulation
During chewing simulation and thermal cycling, no specimen failed.
3.2.2.
Remaining fracture resistance
Force values recorded during fracture testing of crowns surviving thermal cycling and chewing simulation were: • crowns attached by use of resin cement Fu = 1226 ± 290 N (Fu,min = 765 N, Fu,max = 1545 N) • crowns attached by use of fusion ceramic Fu = 1492 ± 206 N (Fu,min = 1117 N, Fu,max = 1674 N)
(n = 8): (n = 8):
The difference between these values was significant (Mann–Whitney U-test; p = 0.028). The results are displayed in Fig. 2. In comparison with the initial fracture resistance, fracture forces for fused crowns were approximately the same after aging. In contrast, the lowest observed fracture force for luted crowns decreased by approximately 140 N.
3.2.3.
Macroscopic fracture appearance
In both luting and fusion groups, adhesive failures (delamination) of the veneer were predominant. Delamination areas were smaller for the fusion group than for the luting group. Occasionally, especially for high fracture forces, damage of the margin of the zirconia framework occurred in the fusion group; this was not observed in the luting group (Fig. 3a and b).
3.3.
Effect of the stiffness of the interface layer
Tensile stresses are of most interest when brittle materials fracture. As is apparent from Fig. 4, tensile stresses within the veneer, frame, and interface layer of the undamaged restoration loaded with a critical value of Fu,m were highly affected by the stiffness of the interface layer. In a radial section including the center of the loading area, lower tensile and compressive stresses were observed within the zirconia framework of luted crowns when compared to fused crowns, because a more resilient interface layer leads to more uniform
load transfer along the interface area. However, the tensile stresses in the zirconia framework never exceeded 300 MPa, which is approximately one third of the material’s strength. Hence, fracture of the restoration originating from the frame is highly improbable. For the fused crowns, tensile stresses within the fusion layer reached critical values I > 100 MPa whereas for the luted crowns critical tensile stresses ( I = 375 MPa, II = 260 MPa, strength of the veneering material: u,veneer = 360 MPa) occurred at the inner surface of the veneer. As a consequence, in both cases, fracture may originate from the interface layer. However, high tensile stresses were also present in the vicinity of the loading area, so a fracture starting from the loading site is also probable.
3.4.
Fractographic analysis by SEM and EDX
Fractographic analysis by SEM showed that for fused crowns a mixed failure mode (cohesive and adhesive failure of the fusion ceramic) occurred (Fig. 5a). The site of fracture initiation was identified in different specimens as either in the occlusal surface area (Fig. 5b) of the veneer or in the interface between veneer and core (Fig. 5c). There were no differences between aged and non-aged specimens in the fusion group. For the luted non-aged crowns predominately adhesive failures between the core and Multilink implant were observed (Fig. 6a). After aging, for the luted crowns adhesive failures between the veneer and Multilink implant were observed. The presence of Multilink Implant cement on the adhesively fractured surface was detected by EDX. The elements Ba and Yb are characteristic indicators of cement composition. The sites of fracture initiation for these crowns were comparable with those for the fused crowns (Figs. 5b and 6b).
4.
Discussion
In recent years it has become obvious that cohesive and adhesive failures of zirconia-based, veneered (hand-layered) restorations often occur [13,14]. Several approaches have therefore been assessed as possible means of reducing this high occurrence of failure. Although some of these approaches have been shown, in vitro, to reduce the incidence of veneer failure, additional effort is necessary to
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
127
Fig. 4 – Principal stresses I , II , and III for all-ceramic crowns with the veneer fused (left) or luted (right) to the zirconia framework. In both cases, the mean ultimate load Fu,m obtained from in vitro tests with aged and non-aged crowns was applied.
improve the performance of these restorations further. In this context CAD/CAM-based manufacture of the restoration might be a promising technique. Beside the use of monolithic restorations based on lithium disilicate [15] or zirconia [16], the CAD/CAM based fabrication of the veneer and attachment of the veneer to the framework by use of different techniques have been proposed [5,6]. One advantage of CAD/CAM is the use of industrially produced ceramic blanks with fewer faults than manually produced veneer [17]. Use of lithium disilicate veneer has also been suggested, because the mechanical properties of this material are superior to those of feldspathic ceramic [7], i.e., lithium disilicate is approximately three times as strong as feldspathic ceramics. However, larger restorations (e.g. long-span fixed dental prosthesis) cannot be produced by monolithic lithium disilicate because of the mechanical properties of the material. In these cases veneered or non-veneered (monolithic) zirconia restorations have to be used if the patient desires all-ceramic restorations. However, both over-pressed and hand-layered zirconia restorations show reduced fracture loads [18], although the over-pressed zirconia restorations show higher fracture loads than the hand-layered ones [19]. Furthermore, for monolithic zirconia restorations there is a lack of clinical studies assessing the wear of antagonistic teeth (although in vitro
studies showed promising results [16]). Additionally, esthetic considerations might limit the use of monolithic zirconia restorations. Thus, the technique presented in the present study might be an auspicious way to restore or replace decayed teeth. Previous in vitro testing [6] of manually layered crowns with the same design as used in this study revealed a substantial effect of artificial aging—7 of 8 crowns failed during chewing simulation with Fmax = 108 N. In the work reported here, all crowns survived chewing simulation, and the difference between fracture force for aged and non-aged crowns was rather small, especially in the fusion group. Therefore, crowns consisting of a zirconia framework and CAD/CAM-produced lithium disilicate ceramic veneer were almost insensitive to the artificial aging process used in this study. The fracture resistance, Fu , of aged and non-aged crowns with the veneer fused with IPS e.max Crystall./Connect was in the range 1100–1700 N. The aged crowns with luted veneer fractured at significantly lower forces than the fused crowns and a small decrease of minimum fracture resistance because of aging was also observed, although this minimum fracture resistance, Fu,min = 765 N, still exceeds the maximum load of approximately 500–700 N which can be expected to act on a single tooth in the molar region in vivo.
128
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
Fig. 5 – Fractographic analysis of fusion ceramic after initial fracture. (a) Overview of fracture surface, with fusion ceramic adhering to veneer. (b) Loading area with possible fracture initiation. (c) Crack formation within the fusion.
FE analysis revealed that tensile stress equaling the material’s strength (360 MPa for e.max CAD) was reached at the inner surface of veneer with soft resin cement as interface material when the crown was loaded with Fu,m = 1218 N. In contrast, crowns with a hard interface layer made of fusion ceramic had critical tensile stress values within the interface layer which reached Fu,m = 1440 N. Stresses in the vicinity of the contact region were not included in the FE evaluation, because small imperfections will have a large effect on the
computed results. Next to the loading site, however, high tensile stresses acting along the outer surface of the crown could initiate cone cracking [20] within the veneer. The use of fractography to identify the origin of the fracture has been recommended by several authors [8,11]. In this study fractographic analysis revealed that the sites of fracture initiation in both groups are the loading point and the interface between the veneer and the core. The high stress concentration beneath the loading area lead to crack initiation
Fig. 6 – Fractographic analysis of resin-cemented restoration after initial fracture ceramic. (a) Overview of fracture surface, cement adhering to glass-ceramic veneer. (b) Loading area with possible fracture initiation.
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
due to Hertzian cone crack formation. This is in accordance with the finding of the FE analysis which came to the same result. Additionally, fractographic analysis could show in some specimens a secondary flaw system within the fusing ceramic, which reduces the stiffness of the material. These flaws might be a result of thermal residual stress of the fusion ceramic which were not considered in the FEA analysis. The combination of FEA and fractographic analysis provided a deeper understanding of the failure of the specimens, as reported by Anusavice [8], and has therefore be recommended. Although the manufacturer recommends fusing the veneer with the core, luting of the veneer could be desirable for two reasons: first, it is much easier, so it could be performed chairside in the dental practice; second, the frame and the veneer can be sintered at the same time, which can save time, in contrast with fused crowns, for which the frame must be sintered first. Both aspects might justify use of luting instead of the fusion technique. The mechanical properties of resin cement are, however, inferior to those of fusion ceramic. This might, as shown by the finite element analysis, have a negative effect on the fracture load of these restorations, because the point of initiation of the fracture might be located at the ceramic–composite interface. In a previous study it was shown that zirconia crowns veneered with lithium disilicate ceramic outperform zirconia crowns manually veneered with feldspathic ceramic [6]. Although esthetic aspects of feldspathic veneer might be slightly superior to those of lithium disilicate veneer, both materials are suitable for production of appealing dental restorations [21]. Clinically, chipping is often observed after the restoration has been in situ for some time, i.e., aging has occurred [1]. This study showed that both attachment techniques are relatively insensitive to aging, because initial fracture loads were comparable with fracture loads after artificial aging, especially for fused crowns. For manually layered feldspathic ceramic, results were completely different: artificial aging drastically reduces the fracture resistance of veneered zirconia crowns [6]. Both initial fracture loads and fracture loads after artificial aging should, however, exceed expected chewing forces, which are lower than 700 N in the posterior region [22]. In this study this was achieved for both techniques—the lowest value (Fu,min = 765 N) was measured for one specimen in the luting group. For all other specimens fracture resistance, Fu , was >900 N. Only one cusp was loaded in the fracture tests because, eventually, one or a few unique critical force applications during the entire lifetime of a ceramic restoration will cause technical failure, i.e., the critical load case does not have to fit the regular loading scheme. For the aging process, however, regular loading conditions should be simulated. Here, the applied load simulates part of a vertical biting force which is transferred to several cusps. The regular vertical bite force corresponding to our test setup for periodic loading on n cusps with 30◦ inclination is, then, given by: n × 108 N × cos 30◦ which is either 281 N or 374 N for n = 3 or n = 4. In this study a metal die was used instead of a natural tooth and the stump was assumed to be rigid in FE computations. Rosentritt [23] found that the die material has a significant effect on the fracture resistance of all-ceramic crowns. In a
129
recent study, however, it was shown by finite element analysis that for our test setup the stump’s stiffness had almost no effect on the stresses in the critically loaded structures (veneer and interface layer) because the zirconia framework does not fracture [6] and “shields” the veneer from the die. In addition, Rosentritt used axial loading, which led to complete fracture of the crown, including the Empress 2 framework (lithium disilicate ceramic) for which the stresses were highly dependent on the properties of the underlying material. Consequently, use of metal dies seems to be adequate for our test setup, and fracture resistance results similar to the situation in vivo can be expected. Because in vitro experiments always include simplification of the real situation and reduction to a few selected load cases and other variables, however, transfer to a clinical setting is always a challenge. For subjects with bruxism and/or in the event of exceptional loading conditions, especially, the risk of failure might be underestimated in this study. In addition, flaws generated either by the dental technician (inaccurate processing, sharp edges) or the dentist (inappropriate preparation, grinding of the occlusal surface) may lead to lower fracture resistance. To summaries, the hypothesis of this study must be accepted. However, both techniques seem to be acceptable for clinical use because of the high fracture resistance.
5.
Conclusion
Although both attachment techniques seem acceptable for use in a clinical setting, fusing the veneer increased the load to failure.
Acknowledgments The authors would like to thank Ivoclar/Vivadent for the material supply and the fractographic analysis. Additionally, we would like to thank Ian Davies for the proof-reading of the manuscript.
references
[1] Al-Amleh B, Lyons K, Swain M. Clinical trials in zirconia: a systematic review. J Oral Rehabil 2010;73:641–52. [2] Triwatana P, Nagaviroj N, Tulapornchai C. Clinical performance and failures of zirconia-based fixed partial dentures: a review literature. J Adv Prosthodont 2012;4:76–83. [3] Rues S, Kroger E, Muller D, Schmitter M. Effect of firing protocols on cohesive failure of all-ceramic crowns. J Dent 2010;38:987–94. [4] Rosentritt M, Steiger D, Behr M, Handel G, Kolbeck C. Influence of substructure design and spacer settings on the in vitro performance of molar zirconia crowns. J Dent 2009;37:978–83. [5] Beuer F, Schweiger J, Eichberger M, Kappert HF, Gernet W, Edelhoff D. High-strength CAD/CAM-fabricated veneering material sintered to zirconia copings—a new fabrication mode for all-ceramic restorations. Dent Mater 2009;25:121–8. [6] Schmitter M, Mueller D, Rues S. Chipping behaviour of all-ceramic crowns with zirconia framework and CAD/CAM manufactured veneer. J Dent 2012;40:154–62.
130
d e n t a l m a t e r i a l s 3 0 ( 2 0 1 4 ) 122–130
[7] Baltzer A. All-ceramic single-tooth restorations: choosing the material to match the preparation—preparing the tooth to match the material. Int J Comput Dent 2008;11:241–56. [8] Anusavice KJ, Kakar K, Ferree N. Which mechanical and physical testing methods are relevant for predicting the clinical performance of ceramic-based dental prostheses? Clin Oral Implants Res 2007;18(Suppl. 3):218–31. [9] Yettram AL, Wright KW, Pickard HM. Finite element stress analysis of the crowns of normal and restored teeth. J Dent Res 1976;55:11–004. [10] Kelly JR, Campbell SD, Bowen HK. Fracture-surface analysis of dental ceramics. J Prosthet Dent 1989;62:536–41. [11] Derand T, Hero H. Bond strength of porcelain on cast vs. wrought titanium. Scand J Dent Res 1992;100:184–8. [12] Dupont WD, Plummer WD. PS power and sample size program available for free on the Internet. Controlled Clin Trials 1997;18:274. [13] Silva NR, Bonfante EA, Zavanelli RA, Thompson VP, Ferencz JL, Coelho PG. Reliability of metalloceramic and zirconia-based ceramic crowns. J Dent Res 2010;89:6–1051. [14] Sailer I, Pjetursson BE, Zwahlen M, Hammerle CH. A systematic review of the survival and complication rates of all-ceramic and metal-ceramic reconstructions after an observation period of at least 3 years. Part II: Fixed dental prostheses. Clin Oral Implants Res 2007;18(Suppl. 3):86–96. [15] Kern M, Sasse M, Wolfart S. Ten-year outcome of three-unit fixed dental prostheses made from monolithic lithium disilicate ceramic. J Am Dent Assoc 2012;143:234–40.
[16] Stawarczyk B, Ozcan M, Schmutz F, Trottmann A, Roos M, Hammerle CH. Two-body wear of monolithic, veneered and glazed zirconia and their corresponding enamel antagonists. Acta Odontol Scand 2012;71:102–12. [17] Christensen RP, Ploeger BJ. A clinical comparison of zirconia, metal and alumina fixed-prosthesis frameworks veneered with layered or pressed ceramic: a three-year report. J Am Dent Assoc 2010;141:29–1317. [18] Eisenburger M, Mache T, Borchers L, Stiesch M. Fracture stability of anterior zirconia crowns with different core designs and veneered using the layering or the press-over technique. Eur J Oral Sci 2011;119:253–7. [19] Stawarczyk B, Ozcan M, Roos M, Trottmann A, Sailer I, Hammerle CH. Load-bearing capacity and failure types of anterior zirconia crowns veneered with overpressing and layering techniques. Dent Mater 2011;27:53–1045. [20] Lawn BR, Deng Y, Lloyd IK, Janal MN, Rekow ED, Thompson VP. Materials design of ceramic-based layer structures for crowns. J Dent Res 2002;81:433–8. [21] Culp L, McLaren EA. Lithium disilicate: the restorative material of multiple options. Compend Contin Educ Dent 2010;31, 716–20, 22, 24–5. [22] Helkimo E, Carlsson GE, Helkimo M. Bite force and state of dentition. Acta Odontol Scand 1977;35:297–303. [23] Rosentritt M, Plein T, Kolbeck C, Behr M, Handel G. In vitro fracture force and marginal adaptation of ceramic crowns fixed on natural and artificial teeth. Int J Prosthodont 2000;13:387–91.
