| Sign In to gain access to subscriptions and/or personal tools. |
Apparent Interfacial Fracture Toughness of Resin/Ceramic Systems
1 School of Dentistry, The University of Passo Fundo, PO Box 611, Passo Fundo, RS, 99001-970, Brazil; Correspondence: * corresponding author, dbona{at}upf.br
We suggest that the apparent interfacial fracture toughness (KA) may be estimated by fracture mechanics and fractography. This study tested the hypothesis that the KA of the adhesion zone of resin/ceramic systems is affected by the ceramic microstructure. Lithia disilicate-based (Empress2-E2) and leucite-based (Empress-E1) ceramics were surface-treated with hydrofluoric acid (HF) and/or silane (S), followed by an adhesive resin. Microtensile test specimens (n = 30; area of 1 ± 0.01 mm2) were indented (9.8 N) at the interface and loaded to failure in tension. We used tensile strength (  ) and the critical crack size (c) to calculate KA (KA = Yc1/2) (Y = 1.65). ANOVA and Weibull analyses were used for statistical analyses. Mean KA (MPa·m1/2) values were: (E1HF) 0.26 ± 0.06; (E1S) 0.23 ± 0.06; (E1HFS) 0.30 ± 0.06; (E2HF) 0.31 ± 0.06; (E2S) 0.13 ± 0.05; and (E2HFS) 0.41 ± 0.07. All fractures originated from indentation sites. Estimation of interfacial toughness was feasible by fracture mechanics and fractography. The KA for the systems tested was affected by the ceramic microstructure and surface treatment.
Key Words: structural reliability • fracture • fractography • fracture toughness • Weibull modulus
Bond strength tests have been used to predict the clinical performance of repaired fractured ceramic restorations and resin-bonded ceramic restorations, even though most of these tests exhibit a wide variability in fracture patterns and bond strength. The commonly used shear bond test often produces fracture at a distance from the resin-ceramic adhesion zone, and may lead to erroneous conclusions about bond quality. Such failures of the substrate prevent the measurement of interfacial bond strength and limit further improvements in bonding systems (Della Bona and van Noort, 1995; Versluis et al., 1997; Chadwick et al., 1998; Della Bona et al., 2003b). The microtensile test was developed to eliminate non-uniform stress distribution at the adhesive interface and to minimize the influence of interfacial defects (Sano et al., 1994). This test has been used to measure the bond strength of composite to dental tissues (Sano et al., 1994; Pashley et al., 1995; Schreiner et al., 1998; Shono et al., 1999; Carvalho et al., 2003) and to ceramics (Della Bona et al., 2000; El Zohairy et al., 2003). The strength values determined from the microtensile test are considered a reliable indicator of the composite-ceramic bond quality, since fractures occur within the adhesion zone. In addition, the microtensile test produces variable fracture-surface morphology and fracture origins for the same adhesive interfaces within the adhesion zone. Therefore, a careful interpretation of the failure mode is required to prevent inappropriate conclusions about the utility of the microtensile test and the adhesion zone phenomena. Fracture-surface characterization combined with analyses of fracture mechanics parameters is of great importance to our ability to understand and predict bonded interface reliability (Chen and Mecholsky, 1993; Della Bona et al., 2003b). One method for assessing bond resistance to fracture is an estimation of the apparent interfacial fracture toughness (KA) of the adhesion zone by the promotion of crack initiation within the bonded interface. The initial critical flaw can be identified by fractography, and KA can be assessed according to fracture mechanics principles. KA is expected to be independent of crack size, whereas bond strength is dependent on crack size (Chen and Mecholsky, 1993). Thus, KA reflects the ability of a material to resist unstable crack propagation at the interface (Mecholsky and Barker, 1984; Tam and Pillar, 1993). The objective of this study was to use fracture mechanics and fractography to determine the KA of the adhesion zone of resin/ceramic systems, testing the hypothesis that KA is affected by ceramic microstructure and ceramic surface treatments.
A hot-pressed leucite-based ceramic (E1; IPS Empress, Ivoclar-Vivadent, Schaan, Liechtenstein) and a hot-pressed lithia disilicate-based ceramic (E2; IPS Empress2, Ivoclar-Vivadent, Schaan, Liechtenstein) were selected for the study (Della Bona et al., 2004b). Fifteen ceramic blocks each of the E1 and E2 ceramics were fabricated and polished through 1 µm alumina abrasive (Mark V Laboratory, East Granby, CT, USA). All ceramic blocks were ultrasonically cleaned in distilled water for 10 min after being polished. The methods for investigating the microstructure, composition, mechanical and physical properties, and the HF-treated ceramic topography of E1 and E2 ceramics have been fully described in previous studies (Della Bona et al., 2000, 2003a,Della Bona et al., b, 2004a,Della Bona et al., b). The ceramic blocks were randomly divided into 3 groups for each ceramic (n = 5) and treated as follows: E1HF: 9.5% hydrofluoric acid (HF) (Ultradent Products Inc., South Jordan, UT, USA) applied for 1 min on E1 ceramic, rinsed for 30 s, and dried with oil-free air. E1S: silane coating (S) (Ultradent Products Inc., South Jordan, UT, USA) applied to E1, allowed to evaporate for 5 min, and air-dried. E1HFS: HF applied for 1 min, rinsed for 30 s, air-dried, followed by application of S. E2HF: HF applied as for E1HF on E2 ceramic. E2S: S coating applied as for E1S on E2 ceramic. E2HFS: HF and S applied as for E1HFS on E2 ceramic. Adhesive (Scotchbond Multi-Purpose Plus, 3M Dental Products, St. Paul, MN, USA) and 2-mm-thick layers of a resin composite (Z100, 3M Dental Products, St. Paul, MN, USA) were applied to the ceramic-treated surfaces and light-cured for 10 s and 40 s, respectively, with the use of a Coltene Coltolux-4 unit (Coltene/Whaledent Inc., Mahwah, NJ, USA; light intensity = 430 mW/cm2). The ceramic-adhesive-composite sets were cut by means of a slow-speed diamond wheel saw (model 650, South Bay Technology Inc., San Clemente, CA, USA), which produced bar specimens with a mean bonding area of 1 ± 0.01 mm2 (Della Bona et al., 2000). The bar specimens were examined for flaws by optical microscopy (microscope model SCW30L, Fisher Scientific, Bangkok, Thailand). The exclusion criteria included the presence of any obvious flaw or specimen debonding before testing. Thirty specimens per group were selected at random and stored in air at 23°C and 50% humidity for 7 days before indentation. We performed a preliminary study to determine the indentation load (P). A microhardness tester (Model MO Tukon Microhardness Tester, Wilson Instruments Inc., Binghamton, NY, USA) with a Vickers diamond indenter was used to indent the center of the ceramic/resin interface of additional specimens. Indentation loads ranged from 4.9 N to 29.4 N (dwell time, 20 s). An indentation load of 9.8 N was selected based on a constant value of P/c3/2;o plotted against P, where co is the dimension of the radial/median crack (Anstis et al., 1981). In addition, some interfaces (E1S and E2S) consistently debonded during the indentation procedure, when loads greater than 9.8 N were used.
After indentation, the crack was allowed to grow and stabilize for 24 h in air at 23°C and 50% humidity before being tested. Each specimen was loaded to failure in tension (Fig. 1
Tensile bond strength ( ) and KA data were analyzed statistically by one-way ANOVA and Tukey’s multiple range test ( = 0.05). We also performed Weibull analysis to evaluate the structural integrity of the adhesion zone (Della Bona et al., 2003b).
We prepared fracture surfaces for SEM examination to determine the mode of failure, which was confirmed by x-ray elemental maps (EDAX). We used quantitative fracture surface analysis (fractography) to determine the critical crack size (c) and to calculate KA (KA = Y
The elastic modulus (E), hardness (H), and other relevant properties of the materials used in this study were reported previously (Table 1  ).
The mean bonded area of the specimens was 1.00 ± 0.01 mm2. Linear regression analysis showed that tensile bond strength was statistically independent of the size of the bonded area. The load and location of indentations were adequate to produce controlled critical defects (Fig. 1  ).
Ceramics treated with S only showed the lowest m value for both E1 and E2 ceramics. The E1HFS group revealed the greatest m value of all tested groups. For each surface treatment, the mean
The mean KA value of E2HFS was significantly greater than that for all the other groups (p  0.0001). There was no difference in the ranking of the mean and KA values, which varied with the ceramic microstructure and surface treatment. Both and KA mean values increased when HF and S were used (Table 2).
All fractures originated from the indentation semicircular crack (Fig. 2
Evaluation of KC or KA requires knowledge of the E/H ratio (Anstis et al. , 1981). KC is determined to an accuracy of 30 to 40% for any well-behaved material. In this context, uncertainties in the value of E/H are relatively unimportant; indeed, this ratio varies between 10 and 50 for most ceramics. Replacement of the ‘calibration’ constant  (E/H)1/8;by an averaged quantity = 0.59 [(E/H)1/8] = 0.88 would add no more than 10% to the error in KC evaluation for a material whose elastic/plastic parameters are unknown (Chantikul et al., 1981). In this study, the calibration constants for all materials were in agreement with the above values: E1 = 0.83; E2 = 0.83; AR = 0.86; and RC = 0.85. Previous investigators have discussed the generation and limitations of the indentation technique (Hagan and Swain, 1978; Kruzic and Ritchie, 2003), and a brief discussion is included in the APPENDIX. To test the integrity of bonded interfaces, one can subject a bonded assembly to a variety of loading conditions to control the crack path along the interface or within the interfacial region. Analyses of bond tests have revealed several problems associated with most common test arrangements, suggesting a lack of reliability of such measurements in assessing the adhesive behavior of bonded dental materials (Anusavice et al., 1980; van Noort et al., 1989; Della Bona and van Noort, 1995; DeHoff et al., 1995; Versluis et al., 1997; Sudsangiam and van Noort, 1999). The non-trimming method we used in this study to obtain specimens for the microtensile test produces less stress in the adhesion zone (Pashley et al., 1999). Since no specimen finishing is necessary, this method also avoids areas of stress concentration produced by polishing materials that differ in hardness and particle size (Della Bona et al., 2000, 2003b). Effective etching of the ceramic surface is considered an essential step for the clinical success of indirect ceramic-bonded restorations and direct-repaired ceramic prostheses. Structural and surface analyses of etched ceramics have showed that different etching patterns are created according to the ceramic microstructure and composition (Della Bona et al., 2003b), and the concentration, application time, and type of etchant (al Edris et al., 1990; Kupiec et al., 1996; Chen et al., 1998; Della Bona and van Noort, 1998; Jardel et al., 1999; Della Bona et al., 2000).
The results of this study show that there is a synergistic effect of HF and S on the The evaluation of the structural integrity of the adhesion zone by Weibull analysis revealed the highest Weibull modulus for E1HFS specimens, which is in agreement with the results of previous research (Della Bona et al., 2000). The strength obtained by the microtensile test can be a reliable indicator of the composite-ceramic bond quality, if all fractures occurred within the adhesion zone. In addition, the microtensile test produced variable fracture surface morphology for the same adhesive interfaces within the adhesion zone, which included failure modes 2, 4, and 5. Compared with optical microscopy observations, a thorough SEM examination of the fracture surfaces and confirmation of composition through the use of x-ray elemental map analysis will ensure a more consistent and complete description of the fracture process and the modes of failure. An appropriate way to assess the interfacial bond is to analyze the energy per unit crack surface area, GI, that is required for a crack to advance in the bond plane. Toughness is related to the critical strain energy release rate (GIC) and is a measure of the resistance of the bond to fracture, since GIC represents the relative energy required to create new surfaces.
A positive relationship was found between the mean values of The observation of microcracks at both adhesive interfaces for the bonded E2 ceramic specimens may be explained by the greater difference in E and H between the resin components of the adhesion zone and the E2 or E1 ceramic components. The E and H differences between the adhesive and both the composite and the ceramics used in this study result in different residual stresses at the interfaces. In both systems, the strongest interface bond occurred at the adhesive-composite interface. Thus, we expect all failures to occur at the weaker adhesive-ceramic interface, agreeing with previous results from non-indented fractured resin-ceramic systems (Della Bona et al., 2000, 2003b). Therefore, the results suggest that the interfacial crack size, and consequently the KA value, is associated with the E/H ratio of the substrates. Since all fractures occurring within the adhesion zone originated from the Vickers indentation, this study suggests that the microtensile test may be preferable to conventional shear or flexural tests as an indicator of composite-ceramic bond quality. A thorough SEM examination of the fracture surfaces following the principles of fractography and confirmation of surface composition through the use of x-ray elemental map analysis produce a more consistent and complete description of the fracture process and the modes of failure. These analyses avoid simplistic interpretations, such as the ‘mixed mode of failure’, that often follows ‘adhesive and/or cohesive’ observations. Thus, the quality of the bond should not be assessed based on bond strength data alone. Yet, estimation of interfacial toughness is possible with fracture mechanics and appropriate fractographic analyses that should reduce the risk for data misinterpretation and provide important information leading to predictions of clinical performance limits, which is the ultimate test of any adhesive system. Future studies should focus on GIC and apply different methods to test the interfacial fracture toughness, validating this procedure.
This study was supported by CNPq do Brazil grant 300659/2003-2 and by NIH/NIDCR Grant DE06672. The authors thank Ivoclar Vivadent (Liechtenstein) for providing the ceramic materials used in this study.
A supplemental appendix to this article is published electronically only at http://www.dentalresearch.org. Received for publication February 24, 2006. Revision received May 26, 2006. Accepted for publication July 6, 2006.
Journal of Dental Research, Vol. 85, No. 11,
1037-1041 (2006)
|
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
), Elastic Modulus (E), Poisson’s Ratio (
), Vickers Hardness (H), Four-point Flexural Strength Tested in 37°C Distilled Water, and Fracture Toughness by Fractography and Fracture Mechanics of Empress Ceramics (E1 and E2), the Adhesive and the Resin Composite (Z100)