| Sign In to gain access to subscriptions and/or personal tools. |
Veneer vs. Core Failure in Adhesively Bonded All-ceramic Crown Layers
1 Materials Science and Engineering Laboratory, National Institute of Standards and Technology, 100 Bureau Drive, Mail Stop 8520, Gaithersburg, MD 20899-8520, USA; Correspondence: * corresponding author, brian.lawn{at}nist.gov
Joining a brittle veneer to a strong ceramic core with an adhesive offers potential benefits over current fabrication methods for all-ceramic crowns. We tested the hypothesis that such joining can withstand subsurface radial cracking in the veneer, from enhanced flexure in occlusal loading, as well as in the core. Critical conditions to initiate fractures were investigated in model crown-like layer structures consisting of glass veneers epoxy-joined onto alumina or zirconia cores, all bonded to a dentin-like polymer base. The results showed a competition between critical loads for radial crack initiation in the veneers and cores. Core radial cracking was relatively independent of adhesive thickness. Zirconia cores were much less susceptible to fracture than alumina, attributable to a relatively high strength and low modulus. Veneer cracking did depend on adhesive thickness. However, no significant differences in critical loads for veneer cracking were observed for specimens containing alumina or zirconia cores.
Key Words: adhesive joining • glass • occlusal loading • veneer failure • core failure
In a recent study, we considered the possibility of using resin-based interlayer adhesives to join a porcelain-like veneer onto a stiff and strong ceramic core in the fabrication of all-ceramic crowns (Lee et al., 2007a). Such adhesives offer a simple means for bonding 2 independently fabricated layers, providing a barrier to crack propagation from one layer to the next (Clegg et al., 1991; Lee et al., 2007a). In that earlier study, a model system consisting of a 1 mm glass veneer layer was bonded onto a much thicker glass substrate layer, and we tested the support capacity of the adhesive by measuring the critical load to initiate a radial crack in the veneer undersurface from loading at the top surface with a spherical indenter. Introduction of a compliant adhesive caused the veneer to flex, with resultant tensile stress at the veneer undersurface. The critical loads could be maintained above those experienced in normal occlusal biting function by suitably increasing the modulus or diminishing the thickness of the adhesive. However, that test procedure ignored one important feature of all-ceramic crowns. In reality, any such veneer/adhesive/core ’crown’ must itself be cemented onto a compliant dentin substrate, in which case the core itself becomes subject to flexure in occlusal loading and therefore susceptible to the same kind of radial cracking as the veneer (Deng et al., 2003; Lawn et al., 2007; Rekow and Thompson, 2007). Thus, proper design of adhesively joined veneer/core bilayers must account for the effects of interlayer thickness and modulus on the relative loads to produce fracture in both the veneer and the core.
In this study, we investigated this issue by testing a model flat-layer system consisting of a glass veneer of thickness dv = 1.0 mm bonded to an alumina or zirconia core of thickness dc = 0.5 mm with an epoxy adhesive interlayer of variable thickness h (Fig. 1
Materials from a previous study were used to fabricate the layer systems depicted in Fig. 1  (Hermann et al., 2006; Bhowmick et al., 2007; Lee et al., 2007a). Veneers were glass plates of 1-mm thickness. Glass plates were used because they are transparent, with mechanical properties similar to those of porcelain. The side walls were polished to allow for side viewing of veneer cracking during testing. Cores were plates of alumina (AD995, CoorsTek, Golden, CO, USA) or yttria-tetragonal-stabilized zirconia (Lava Frame, 3M ESPE, Morrow, GA, USA) with 0.5-mm thickness. Some glass plates of the same thickness were used as control cores for baseline comparison. Polycarbonate substrates (Makrolon®, Bayer Material Science AG, Leverkusen, Germany) of 12.5-mm thickness were used as a dentin-like base. In one set of specimens, the bottom surfaces of the veneer and core plates were abraded with 600-grit SiC, to introduce controlled flaws and thus to promote crack initiation (Chai et al., 1999; Zhang et al., 2006). In another set of specimens, the glass and ceramic plates’ undersurfaces were, respectively, etched and carefully polished, to inhibit crack initiation for comparative testing. Epoxy resin (Harcos Chemicals, Bellesville, NJ, USA) was used as a joining adhesive. This material lies at the low end of compliance for dental adhesives, and so represents a worst case as far as veneer vulnerability is concerned. Opposing veneer and core surfaces were first cleaned thoroughly and then bonded under light mechanical pressure, with spacers between the layers to attain prescribed interlayer thickness over a range h = 5 µm to 500 µm. The finished layer structures were indented with a tungsten carbide sphere (radius, 3.18 mm) to represent occlusal contact (Bhowmick et al., 2007; Lee et al., 2007a). Video cameras were placed so that we could view the onset of radial cracking within the veneer (side view) and the core (subsurface view through the polycarbonate base). No delamination or other mode of fracture was observed in any of these experiments.
Properties of the materials used in the present study are shown in the Table
Results were obtained for critical loads to initiate radial cracks in terms of properties of the veneer (v), adhesive interlayer (i), core (c), and substrate (s), over a range of adhesive thickness h. A plot of critical load Pv for veneer failure is shown as a function of h, for specimens with abraded veneer undersurfaces (Fig. 2a  ). Datapoints are results of individual tests, with each symbol representing a different core material. The point with error bar is the mean and standard deviation for 20 tests on specimens with zirconia cores at one clinically representative adhesive thickness, h = 100 µm. A similar mean and standard deviation were measured for 13 specimens with alumina cores, but that dataset overlaps the zirconia-core data and hence is omitted from the plot. An unpaired Student’s t test showed no significance difference (p >> 0.05) between the sets of alumina-and zirconia-core data at thickness h = 100 µm. The individual datapoints for different cores appear to overlap within scatter over the entire range of h, further confirming little effect of underlying core or substrate material on veneer failure. Observe that critical load Pv diminishes steadily with increasing h, by a factor of more than five over the thickness range 5 µm to 500 µm. The solid curve in the Fig. is a universal fit using material parameters from the Table along with the following equation for critical load (Chai and Lawn, 2000, 2002):
where d = layer thickness, E modulus, S strength, and B' is an interlayer thickness-dependent term,
with constants B = 1.35, β= 0.18, and
An analogous plot of critical load Pc for core failure as a function of adhesive thickness h is shown, for veneers with etched undersurfaces (Fig. 2b
with D = dv + dc = veneer + core thickness, and effective modulus
where
The analysis in Eqs. 1–4, validated by the above data fits (Figs. 2a, 2b
We used a simple contact loading test to evaluate the effectiveness of polymer-based adhesives as a means of bonding brittle veneers (glass) to ceramic cores (alumina and zirconia). Such adhesives offer the potential for simplified fabrication of all-ceramic dental crowns, in which veneer and core are prepared independently prior to joining, without the time-consuming hand-layering, drying, and fusing processes currently in use. Future technologies, such as direct-write fabrication and robocasting, are exciting prospects in this context (Lewis et al., 2006; Strub et al., 2006). The use of high-modulus adhesives to join such layers may minimize any potential swelling of an interlayer during subsequent oral function (Wang et al., 2007). Adhesive joining also avoids the need for matching the coefficients of thermal expansion so essential in traditional porcelain fusion procedures, and provides an internal barrier to crack propagation from one layer to the next (Lee et al., 2007b). [A previous study on systems close to those used here indicated that any residual stresses from shrinkage of the adhesive during curing, provided the interlayers are sufficiently thin, should be negligibly small compared with typical thermal mismatch stresses (Hermann et al., 2006).] Consequently, the system used here affords a model system for an understanding of the role of interlayer properties well beyond current dental technologies. The contact test compares critical loads Pv and Pc to produce radial fractures at the bottom surfaces of veneer and core layers, respectively. Analysis of experimental data on model glass-veneer/ceramic-core test specimens (thickness ratio, 1 mm/0.5 mm) bonded onto dentin-like substrates validates the theoretical relations in Eqs. 1 to 4, and thereby provides a sound physical basis for predicting fracture responses of real crown material systems. In the context of adhesive joining, the main requirement is to ensure that veneer failure never occurs. Practically, this may be achieved by choosing the join conditions so as to maintain Pv higher than Pc, and the core properties so that Pc always exceeds occlusal biting forces (>> 100 N) (McLean, 1979; Kelly, 1997, 1999). Design maps indicating critical load conditions for given material systems provide a useful graphic methodology for quantifying susceptibility to veneer and core failure. Of principal interest in this study is the role of the adhesive interlayer properties in determining such failure susceptibility. Changes in adhesive thickness h or modulus Ei can lead to substantial shifts in the critical load Pv for veneer fracture. Clearly, keeping the adhesive layer sufficiently thin is crucial for protection of the veneer, the more so for higher biting forces. By way of example, a porcelain veneer of thickness 1 mm seated on an adhesive interlayer of thickness h = 100 µm and modulus Ei = 20 GPa is predicted, from Eq. 1, to withstand biting forces up to 600 N. Further increasing the adhesive modulus would allow the veneer to sustain even higher forces. These considerations would appear to provide a strong motivation for the development of stiffer adhesives for joining (Wang et al., 2007).
The equations also enable us to predict the role of core material properties, namely, modulus Ec and strength Sc. Of these properties, strength would appear to be the more important, because it is subject to more variation (Table In this paper, we have examined the competition between veneer and core radial crack modes. Other fracture modes may occur under extreme conditions (Lawn et al., 2007): cone cracks may initiate at the veneer top surface, particularly in contacts with small radii; and delamination may occur at the bonding interface if the adhesive bonding is not strong enough. Both these additional modes may be exacerbated in fatigue loading. Nevertheless, intercomparison between the two radial fracture modes remains a simple means of assessing relative susceptibilities of veneer and core failure, and offers a convenient route for the evaluation of prospective joining materials for processing of all-ceramic crowns.
This work was supported by a grant from the US National Institute of Dental and Craniofacial Research (PO1 DE10976). The use of equipment, instruments, or materials in this study does not imply recommendation by the National Institute of Standards and Technology. Received for publication August 31, 2007. Revision received November 20, 2007. Accepted for publication December 14, 2007.
Journal of Dental Research, Vol. 87, No. 4,
363-366 (2008)
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
= 0.84 (
= Ev/Ec and 
= dv/dc. Note that there is no appearance of Ei or h in Eqs. 3 and 4, indicating an insensitivity to the adhesive material. (Strictly speaking, interlayer thickness h should be included in D in Eq. 3, but this will become important only in the region of large h, where veneer failure dominates anyway.) The predicted horizontal lines from these equations pass through the data for each core material, independent of h within the experimental scatter. Note also that Pc is greater for alumina than for glass cores, but smaller than predicted for zirconia, as may be expected. 