| Sign In to gain access to subscriptions and/or personal tools. |
Quantifying the Strength of a Resin-coated Dental Ceramic
1 Biomaterials Unit, University of Birmingham School of Dentistry, St. Chad’s Queensway, Birmingham B4 6NN, UK; and Correspondence: * corresponding author, addisono{at}adf.bham.ac.uk
Resin luting all-ceramic restorations increases clinical performance; however, the strengthening mechanisms are not fully understood. The authors have previously proposed the existence of a resin-ceramic hybrid layer, and the hypothesis tested was that ceramic strength enhancement was conferred by the characteristics of the resin-ceramic hybrid layer. Dentin porcelain discs were polished with a P4000-grade abrasive paper, and half were centrally indented at 9.8 N. Further discs were alumina-air-abraded. Groups of 30 specimens were coated with resin cement thicknesses varying from 0 to 250 ± 20 µm before bi-axial flexure testing. Following investigation of residual stresses by annealing, regression analysis enabled us to calculate the magnitude of ’actual’ strengthening for a theoretical ’zero’ thickness of resin cement on each surface texture. Accounting for resin bulk strengthening, resin cement coating significantly increased the mean strength that was attributed to a resin-ceramic hybrid layer sensitive to surface texture.
Key Words: feldspathic porcelain • bi-axial flexure strength • resin cement thickness
Luting all-ceramic restorations with resin-based composite cements results in improved clinical performance (Malament and Socransky, 1999a,b). Similarly, in vitro load-to-failure mechanical testing demonstrated that resin cement coating increases the flexural strength of dental ceramics (Rosenstiel et al., 1993; Pagniano et al., 2005). Despite these observations, the strengthening mechanisms responsible remain poorly understood, thereby preventing optimization of the operative steps involved in cementation (Fleming et al., 2006). Axisymmetric finite element analysis (Anusavice and Hojjatie, 1992) and quantitative fractography (Kelly et al., 1989, 1996; Kelly, 1999; Quinn et al., 2005) have demonstrated that all-ceramic crown restorations fail from the extension of pre-existing defects resident on the internal ’fit’ surfaces of a restoration under tensile loading. Therefore, clinically relevant in vitro testing must replicate the failure modes involved (Kelly, 1999). Bi-axial flexure replicates the clinical failure mode for all-ceramic restorations, and failure stresses can be derived from Timoshenko’s thin-plate analyses for small deflections of isotropic bodies (Timoshenko and Woinowsky-Krieger, 1959). Timoshenko’s equations have been applied to resin-coated ceramics; however, stresses generated due to the mismatch in both Poisson’s ratio and elastic modulus in cemented (bilayered) structures are ignored (Young and Budynas, 2002). Recently, advancement of the analytical solutions by Hsueh et al.(2005) facilitated the calculation of stress distributions at axial positions throughout a bi-axially loaded bilayered specimen. The solutions have been compared with finite element models and enable stress calculations to be made at the point of failure in resin-coated ceramic specimens, namely, the resin-ceramic interface (Hsueh et al., 2005). However, the magnitude of the failure stresses is sensitive to the relative thicknesses of the materials comprising the bilayer. Variations in the relative thickness of the materials will modify the neutral axis of bending position and, therefore, the magnitude of shear stresses generated at the interface (Young and Budynas, 2002). Recent in vitro investigations on resin-strengthening mechanisms of ceramics have demonstrated that the traditionally accepted models of crack healing (Marquis, 1992) or crack-closure stresses secondary to resin polymerization shrinkage (Nathanson, 1994) do not adequately describe the strengthening patterns and mechanisms (Fleming et al., 2006). A mechanism insensitive to individual defect severity, but sensitive to surface texture, has been suggested by the authors, whereby strengthening is dependent upon the creation and subsequent characteristics of a resin-ceramic hybrid layer (Addison et al., 2007a,b). To validate the existence of a resin-ceramic hybrid layer, we tested the hypothesis that strengthening was conferred by the resin-infiltrated ceramic surface rather than the resin bulk, and that the magnitude of strengthening was dependent on the ceramic surface texture. We used controlled surface textures and advanced analytical solutions, in combination with a regression technique, to quantify the ’actual’ degree of resin strengthening at a predicted theoretical ’zero’ resin thickness.
Specimen Condensation We manipulated 0.6 g of Vitadur Alpha dentin porcelain powder and 0.22 mL of Vita Modeling Fluid (LOT7789 and LOT14209R, respectively; Vita, Bad Säckingen, Germany) to form an optimum slurry consistency (Fleming et al., 2000). The slurry was condensed into a perspex mold (15 mm diameter and 0.9 mm thickness) secured to burnished aluminum (Fleming et al., 1999). Disc-shaped specimens were pre-dried at 600°C for 360 sec in a Vita Vacumat 40 furnace (Vita Zahnfabrik, Bad Säckingen, Germany) and fired under vacuum from 600°C to 960°C at a temperature increase of 60°C/min for 360 sec. At 960°C, the specimens were air-fired for a further 60 sec, then cooled at 20°C/min prior to storage in a desiccator to prevent hydrolysis (Anusavice and Lee, 1989).
Specimen Surface Modification
Specimen Cementation A quantity of Rely-XTM Veneer Cement (shade A3, 3M ESPE, St. Paul, MN, USA) was applied to the center of the prepared surface of the remaining groups and covered with Mylar. A glass slide was pressed until the resin spread to the edges. The resin was light-activated (Optilux 501, SDS Kerr, Danbury, CT, USA) for 20 sec at an intensity of 740 mW cm–2 with a 13-mm-diameter curing tip. Groups B, G, and L, Groups C, H, and M, Groups D, I, and N, and Groups E, J, and O were coated with a 60 ± 20, 120 ± 20, 180 ± 20, and 250 ± 20 µm layer of cement, respectively, with the use of a custom-made jig.
Investigation of Residual Stresses
We determined the bi-axial flexure strength by centrally loading the glazed surface onto a 10-mm-diameter ring-support with a 4-mm-diameter spherical ball indenter at 1 mm/min, following 24 hrs of dry storage. The neutral plane (tn) was calculated as a function of the porcelain and resin thicknesses
and
The bi-axial flexure stresses were calculated at axial positions (z) at the center of the disc-shaped specimens, where the bonded interface was located at z = 0, the porcelain surface at z = t1, and the resin surface at z = –t2.
(0
(– t2
P was the load at fracture, v1 and v2 were the Poisson’s ratios of the ceramic (0.25, from Zeng et al., 1998) and resin (0.27, from De Jager et al., 2004). a, b, and R were the radii of the knife-edge support, loaded region, and specimen, respectively. We used a contact stylus profilometer (Talysurf CLI 2000, Taylor-Hobson Precision, Leicester, UK) to examine representative surface textures of the 3 prepared porcelain surfaces. Scanning electron microscopy (SEM) with a JEOL JSM 5300 LV (JEOL Ltd., Akishima Tokyo, Japan) was used for the examination of representative prepared ceramic surfaces and cross-sections of the resin-coated ceramic specimens, following deposition of a 2-µm gold layer.
Statistical Analysis
The bi-axial flexure strength of the uncoated polished disc-shaped specimens (Group A) was 75.4 ± 11.8 MPa, while the introduction of a controlled defect (Group F) or surface roughening with alumina particle air abrasion (Group K) resulted in a significant strength reduction (P < 0.05), to 60.4 ± 10.6 and 57.0 ± 4.0 MPa, respectively (Table 1  ). Surface profilometry characterized the polished surface as showing a high frequency of low-amplitude depressions < 3 3m (Fig. 1a ). The polished-indented surface was equivalent to the polished surface, with the exception of a centrally oriented deformation with a depth of 16 3m (Fig. 1b). Electron micrographs illustrated radial cracks associated with the Vickers indent of between 40 and 70 3m (Fig. 1b). The alumina-abraded surface texture had a high frequency of regular peaks and depressions, to 10 3m (Fig. 1c). Annealing the ’as-fired’ specimens resulted in no significant strengthening (Table 2). Annealing of the polished and polished-indented specimens resulted in a significant reinforcement of a similar magnitude, whereas increased strengthening was observed following annealing of the alumina-air-abraded specimens. Significant strengthening was observed on all surfaces following resin-cementation (P < 0.05). Electron microscopy highlighted increased interpenetration of the ceramic surface with resin cement on alumina-air-abraded surfaces compared with polished and polished-indented surfaces (Fig. 1). During bi-axial flexure testing, specimens from Groups C, D, and E delaminated prior to failure (Table 1), skewing the mean bi-axial flexure strength data. Following exclusion of the delaminated specimens, a univariate analysis revealed a significant interaction between resin thickness and mean bi-axial flexure strength on each of the 3 surfaces investigated (P < 0.05). A regression analysis demonstrated linear strengthening with increasing resin thickness for the polished (Fig. 2a, R2 = 0.97) and polished-indented specimens (Fig. 2b, R2 = 0.97). Less confidence in linear strengthening on the alumina-air-abraded surface was observed (Fig. 2c, R2 = 0.92), and a ’best fit’ cubic trend line was substituted (R2 = 0.99). The projected strengths of 96, 83, and 90 MPa for a theoretical ’zero’ resin thickness (Fig. 2) represented a 21-, 23-, and 33-MPa reinforcement of the uncoated polished (75 MPa), polished-indented (60 MPa), and alumina-air-abraded controls (57 MPa), respectively.
The strength of dental ceramics is dependent on the presence of surface flaws, and the stress at failure is related to the size of the critical flaw, assuming the presence of equivalent residual stresses (Griffith, 1921). The electron micrographs of the polished-indented surfaces illustrated an increased critical surface flaw size, resulting from Vickers indentation, when compared with the polished surfaces. The polished and polished-indented groups were reinforced by a similar magnitude following annealing. Therefore, the electron microscopy observations were considered to correlate well with the fracture strength data manifested as a significant strength reduction of the polished specimens following Vickers indentation. Clinically, the internal ’fit’ surfaces of dental ceramic restorations are roughened to promote adhesion (Horn, 1983; Wolf et al., 1993). Alumina particle air abrasion resulted in the creation of a radically altered surface texture when compared with the polished surface condition. In contrast, the mean bi-axial flexure strengths of the alumina-air-abraded and polished-indented groups were comparable when considered in combination with the annealing data, which suggests similar critical flaw sizes. A significant interaction between resin cement thickness and strengthening was observed. For the polished specimens (groups C–E), visual examination of the fracture surfaces during testing revealed several delaminations that skewed the fracture strength data. Delamination of the resin-ceramic structures under bi-axial flexure results from shear stresses generated parallel to the resin-ceramic interface. When the critical shear stress at the interface was exceeded before a critical tensile stress was generated in the ceramic, the strengthening effect was negated, resulting in low tensile failure stresses. Exclusion of the delaminated specimens revealed strengthening associated with increasing resin thickness, for the polished and polished-indented specimens to be equivalent. This was reflected in the regression line gradients and the total reinforcement conferred. In contrast, the pattern of strengthening conferred by the resin cement on alumina-air-abraded specimens was altered for initial increases in resin cement thickness, since a linear trend could not be extrapolated. Additional electron microscopy and surface metrology demonstrated that resin coating resulted in an interpenetrating layer of a significantly greater thickness than observed on the polished specimens. At a theoretical ’zero’ resin thickness, the strengthening observed on each surface is of sufficient magnitude to affect clinical performance and disproves the hypothesis that the resin bulk solely provides strengthening. Strengthening of the polished and polished-indented specimens was of a similar magnitude, despite the contrasting critical flaw sizes. The findings are in agreement with those of Fleming et al.(2006), who concluded that the traditionally accepted resin-strengthening mechanisms of crack healing (Marquis, 1992) and crack-closure stresses (Nathanson, 1994) did not represent the observed strengthening, since a sensitivity to defect size was implied. Alternatively, the magnitude of strengthening following resin coating of the alumina-abraded-air surface was increased compared with that in the polished-indented specimens. De Jager et al.(2000) proposed that macroscopic surface roughness may act as a cumulative stress concentrator on critical flaws. Consequently, rather than differentially ’healing’ defects, the strengthening effect observed occurs due to the interaction of the resin with the entire surface defect population, and the magnitude of strengthening is dependent upon a ’resin-ceramic hybrid layer’. Under loading conditions, the stressing patterns of the resin in the hybrid layer are dependent on the complex interactions of various elements, and the system becomes sensitive to the characteristics of the layer. The hypothesis tested assumed the resin-ceramic hybrid layer to be the dominant source of strengthening. Electron microscopy demonstrated the existence of a resin-ceramic hybrid layer on each surface texture, and a significant reinforcement was observed following extrapolation to a ’zero’ resin thickness. Strengthening may be attributed to the exclusion of moisture, thereby preventing crack tip hydrolysis; however, the pattern of strengthening was expected to be sensitive to defect severity. If the resin-ceramic hybrid layer was dominant, sensitivity to surface texture would be implicit. When the hybrid-layer was radically modified, an enhancement of reinforcement was exhibited consistent with the hypothesis tested. For the alumina-air-abraded surface texture, a tolerance to initial increases in resin thicknesses was observed, since failure may potentially be governed by the presence of an extensive resin-ceramic hybrid layer. Extrapolation from baseline reinforcement, to the stress at the maximum resin thickness tested, established a trend similar to that observed on the polished surface textures, reflecting the consistent influence on the strengthening conferred by the bulk resin thickness on all surface textures. In conclusion, utilizing novel analytical solutions in combination with a regression technique, the authors have demonstrated that the magnitude of strengthening conferred following resin coating is sensitive to ceramic surface texture. The authors have previously demonstrated that clinical variables, including pre-cementation silane coating (Addison et al., 2008) and resin elastic moduli (Addison et al., 2007b), may influence the scale, but not the pattern, of the observations reported. The description of a ’resin-ceramic hybrid layer’ offers a convenient yet simplistic model of the complex strengthening mechanisms operative in all-ceramic crown restorations.
This work was performed in partial fulfillment of the requirements for a self-funded PhD by Owen Addison at The University of Birmingham. Received for publication February 26, 2007. Revision received October 16, 2007. Accepted for publication February 8, 2008.
Journal of Dental Research, Vol. 87, No. 6,
542-547 (2008)
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
z 
