| Sign In to gain access to subscriptions and/or personal tools. |
Polymerization Shrinkage Influences Microtensile Bond StrengthDepartment of Restorative Dentistry, Faculty of Odontology, Complutense University of Madrid, Plaza Ramón y Cajal s/n, Ciudad Universitaria, 28040 Madrid, Spain; Correspondence: * corresponding author, macorra{at}odon.ucm.es
Shrinkage results from a complex spatial strain network, producing movements within materials. The purpose of this study was to test whether microtensile bond strength (µTBS) of a light-curing resin composite to enamel depends on distance to the center of the curing mass. Labial surfaces of bovine incisors were ground flat, divided into 2 groups (n = 8), acid-etched, and coated with an unfilled resin bond. A resin-based composite was placed in one increment (group A) or separately at gingival, central, and incisal sites (group B), and light-cured. Teeth were sectioned, yielding stick-shaped specimens assigned to one of 9 groups according to distance to incisal edge of restoration (NDistanc). Microtensile bond strength was transformed to percentages of its maximum values within each tooth (PMPa). Comparisons within groups showed (group A) that mean PMPa decreased from central to gingival and from central to incisal (p < 0.01). Comparisons between groups showed that mean PMPa was significantly lower in group A compared with group B, only at gingival and incisal sites. Microtensile bond strength significantly decreased as the distance increased to the center of the curing mass.
Key Words: polymerization shrinkage • resin composite • enamel • regional bond strength • microtensile testing
The adverse effects of polymerization shrinkage were reported as early as 1975 (Jorgensen et al., 1975). Since then, many methods to measure polymerization shrinkage have been used: dilatometers (De Gee et al., 1981), tensilometers (Davidson et al., 1984), linometers (Feilzer et al., 1989), and strain gauges (Sakaguchi et al., 1991). Also, Finite Element Analysis (FEA) (Versluis et al., 1998), photo-elastic analysis (Kinomoto et al., 1999), and elastic micromechanics computer modeling (Sakaguchi et al., 2004) have been used to estimate the magnitude and effects of polymerization shrinkage. The competition between contraction stress and the maturing bond to walls is one of the main causes of clinical problems in composite resin dental restorations (Davidson et al., 1984). As long as the polymerizing material can flow before reaching the gel point (Bausch et al., 1982), shrinkage can be compensated for, and further contraction stress can be dissipated (Davidson and De Gee, 1984). Shrinkage can be described as a vector, with a magnitude and a direction (Watts and Cash, 1991). The vector that describes the material’s shrinking block behavior is the result of a spatially complex strain network, and will produce displacements within the curing material. Optimal adaptation is required for adhesion; consequently, these movements may be detrimental to the infiltration of the substrate (Feilzer et al., 1990). Light-cured materials were thought to shrink toward the light source, and self-cured materials toward the center of the curing mass (Asmussen and Peutzfeldt, 1999). FEA studies (Versluis et al., 1998) and photo-elastic analysis (Kinomoto et al., 1999) have shown that the direction of force vectors during polymerization shrinkage flow is determined by the conditions at the boundary (Loguercio et al., 2004a). This direction is also supposed to be dependent on the bonding features of the material to substrate (Asmussen and Peutzfeldt, 1999). Watts and Cash (1991) described a contraction pattern, not only perpendicular to the interface, but also shearing alongside it. It was suggested that this pattern may affect, specifically, the peripheral area of the material being bonded (Feilzer et al., 1990). Microtensile bond testing (Sano et al., 1994) made possible the measurement of bond strengths in several areas within the same tooth. Yoshiyama et al.(1996) measured bond strengths at different sites of root dentin at cervical, middle, and apical regions. This resulted in significant differences when a total-etch dentin bonding system was used. Shono et al.(1999) also used the microtensile test and observed inconsistencies in bond strength across occlusal dentin, but drew no conclusions as to their causes. For a properly bonded resin-based composite restoration, shrinkage vectors should theoretically be oriented toward the bonded interfaces (Cho et al., 2002). This would cause the resin composite to be pulled away from peripheral zones (Feilzer et al., 1990), toward the center, at the expense of optimal bonding (Loguercio et al., 2004b) during polymerization. This effect should be stronger as distances increase. The purpose of this study was to investigate whether resin-enamel bond strength over a relatively large area exhibits regional variations, within the same sample, due to shearing polymerization contraction stresses.
Specimen Preparation and Microtensile Testing Sixteen sound bovine incisors, stored in distilled water at room temperature for a month after extraction, were selected. All samples were collected and used following the guidelines of the Animal Research Committee of the Complutense University. Roots were removed ca. 1 mm apical to the cement-enamel junction (Fig., i  ), with the use of a diamond saw (3031 CP/N Exakt; Norderstedt, Germany) under constant water-cooling.
A flat enamel surface was produced on each incisor’s labial aspect by wet-grinding in a grinder-polisher (Struers A/S Dap-7; Rodøvre, Denmark) with 500-grit SiC paper (Fig., g  ). All labial enamel surfaces were bonded and restored in a similar and standardized sequence, with the same bonding system and resin composite restorative materials (Appendix Table), placed in two different ways (Group A and Group B, n = 8 each). All surfaces were etched (20 sec, 37% phosphoric acid etchant, Scotchbond; 3M ESPE, St. Paul, MN, USA), thoroughly water-rinsed, and gently air-dried. A bonding agent (Heliobond; Ivoclar-Vivadent, Amherst, NY, USA) was applied, gently air-extended, and light-cured for 20 sec (Optilux 501; Demetron/Kerr Co., Orange, CA, USA; 800 mW/cm2).
In Group A, we placed a 2-mm-thick resin composite build-up (Filtek Z250; 3M ESPE, St. Paul, MN, USA) (Fig., A1
Each tooth was cut, producing parallelepipedic sticks with their long axis perpendicular to the pulp chamber (Fig., A2, B2
In group A, the distance (in mm) to the gingival edge of each sample was registered. In group B, 6 specimens per tooth (2 gingival, 2 central, and 2 incisal) were obtained (Fig., B2
Each specimen was glued (Model Repair II Blue; Dentsply-Sankin KK, Tokyo, Japan) to a testing device and subjected to tension until failure (Hounsfield HTI, Croydon, UK), at a crosshead speed of 1 mm/min, and the force recorded from a 500-N cell load was converted to tensile stress units (MPa). Debonded areas were analyzed with a stereoscopic microscope analyzer (Leica Qwin Q500 IW; Leica Imaging Systems Ltd., Cambridge, UK) for determination of the type of fracture (adhesive or cohesive). Only adhesive fractures were analyzed statistically.
Data Management To test this, we related the independent variable to the distance to the gingival end, within samples. This was due to: (i) not all samples having equal total length; and (ii) in evaluation of progressive specimen’s material separation from the center (the distance to it), specimen separation was found at the far end of the mass. Its effects must also be considered. Thus, the NDistanc variable was calculated in group A, creating 9 categories of distance to the gingiva. In group B, specimens were assigned to gingival, central, or incisal categories, corresponding to categories 1, 5, or 9 of group A. This correspondence is most important, because relative microtensile bond strength results obtained from the same relative locations, in both groups, were to be compared, to exclude local differences in the enamel’s micro- or macro-morphological characteristics. Variability in tooth morphology and histology is expected to influence the mean results of microtensile bond strength. To overcome this, and because eventual differences in microtensile bond strength within each tooth can be studied, we reported the dependent variable (PMPa) as a percentage of maximum microtensile bond strength results within each tooth.
Statistical Analysis To compare groups A and B, thus determining if differences within groups could be explained by local enamel morphological variations, we made adequate (parametric or non-parametric) comparisons among corresponding distances (1, 5, and 9 categories).
In group A, distances were distributed in 9 quartiles (NDdistanc), from cervical to incisal, along the labial surface area (Panel B in the Fig .). Numbers of cases, means and standard deviations of PMPa, distributed in NDistanc categories, in groups A and B are given in Table 2. All categories showed a normal distribution, except for categories 8 and 9 in group A. For comments on results’ variability, see the APPENDIX.
Numbers of debonded or discarded specimens are shown in Table 1 , per group (see APPENDIX for detailed analysis). Kruskal-Wallis test results were, for Group A, p = 0.001, and for group B, p = 0.082. Only in group A were differences among one or more groups statistically significant. Jonckheere-Terpstra test results showed that the probability of results from categories 1 to 5, defined by NDistanc, being increasingly ordered by chance was p = 0.002. For categories 5 to 9, this probability was p = 0.01.
Differences between groups in categories 1 (45%) and 9 (22.5%) were statistically significant, and in category 5 (6.9%), non-significant (Table 3
Results demonstrated that there are regional differences when resin composite is built up on enamel in one increment over a large area. These differences decrease from the center to the outer boundaries of the mass, and are not caused by local differences in the substrate. Regional differences in dentin bond strength have been shown before. Microtensile bond strength of resin composite bonded on cervical and middle thirds of root dentin was significantly lower than at the coronal or apical third, when a total-etching adhesive system was used (Yoshiyama et al., 1996). In the present study, there was higher microtensile bond strength in the central than in the cervical or incisal sites, which produced similar results. Two years later, regional differences in microtensile bond strength were reported when self-etching/self-priming systems were used (Yoshiyama et al., 1998), on enamel and apical root dentin, which were significantly lower than those in coronal, cervical, and mid-root dentin. These results are in accordance with the conclusions reported here. Results were related to the reduction of dentinal tubule density in apical root dentin, decreasing the contributions of resin tags to overall bond strength. However, this is not applicable to enamel, because it lacks tubules, which may participate in bonding. In contrast, differences in regional bond strength have not always been clearly demonstrated. Similar dentin bond strength results were obtained between resin-dentin specimens centrally located and those in the periphery in the same molar crown (Loguercio et al., 2005). This may be due to the use of small bonded areas that could not produce relevant differences in microtensile bond strength. Since dentin is a heterogeneous substrate, bond strength of adhesive systems largely depends on microstructure at the bonding site (Giannini et al., 2001). Tubule orientation and the possibility of resin tag formation are different in superficial, middle, or deep dentin (Pereira et al., 1999). Scanning electron microscopy (SEM) also demonstrated regional discrepancies in interfaces, due to differences in number and diameter of dentinal tubules, which facilitate both acid-etching and resin penetration (Yoshiyama et al., 1996). To avoid these confounding characteristics, we used enamel as the bonding substrate in this study. Human and bovine enamel show similar morphological characteristics. There were no significant differences between the two, when bond strength was measured on either surface (Reis et al., 2004). The labial aspect of bovine incisor enamel allows for large bonding areas. The suitability of microtensile testing for enamel may be questionable, due to the tissue’s brittleness and anisotropy (Carvalho et al., 2000). It is well-known that surface microcracks seriously weaken brittle materials. Enamel microcracks may occur during specimen preparation. A SEM investigation of the structural integrity of specimens prepared for microtensile testing, before being loaded, revealed the existence of a higher number of micro-fractures in enamel as compared with dentin specimens (Ferrari et al., 2002). This would contribute to the explanation of our relatively large standard deviations (see APPENDIX). This situation would require a relatively small specimen size (1.3 ± 0.2 mm2 in this study). Differences in substrate morphology can be relevant. Microtensile bond strength of a composite resin to enamel is dependent on the orientation of prisms, being considerably lower when the load is perpendicular to the general orientation of the prisms (Carvalho et al., 2000). Enamel is significantly stronger when tested parallel to its prismatic orientation (Giannini et al., 2001). In our report, as in most clinical situations, resin composite was bonded to enamel roughly parallel to the prisms. Influence of substrate was successfully minimized, since differences in data from group B were not statistically significant. This may explain the present results: Differences in group A were not expected to be caused by distinct prism orientation, but rather by the direction of polymerization contraction vectors. This confirms the suitability of enamel over dentin, making regional differences irrelevant. Interactions of force vectors created during polymerization shrinkage may produce effects detrimental to the material being bonded. In this report, curing resin composite was placed on a surface longitudinally. Before the gel point was reached, the material was theoretically intimately interrelated with enamel, filling the available spaces on its surface. Due to the experimental design, major force vectors tend to align parallel to the interface, following the longest length of the restoration. Resulting centripetal resin displacement could well affect the initial interrelationship between material and tissue, because the former is pulled away from its initial position. Direct experimental determination of the directions of polymerization shrinkage vectors is extremely difficult; this study is based on the interpretation of indirect observations, namely, final microtensile bond strength. Bond strengths of current adhesive systems may surpass polymerization shrinkage stresses, and might possibly affect the magnitude and direction of resin composite polymerization shrinkage vectors (Versluis et al., 1998). In general, it can be said that boundary conditions strongly affect these vectors (Loguercio et al., 2004a). A recent study reported that optimal configuration conditions were insufficient to obtain uniform microtensile bond strength results when adhesive materials were bonded to large areas. Polymerization shrinkage vectors still damage the quality of bonding in outer areas, being more favorable at the center of the mass (group A). In conclusion, the distance to the geometric center of a light-cured mass of composite bonded to enamel influences microtensile bond strength results. This effect is possibly due to centripetal dislodgment of curing resin along dental tissues, because of polymerization contraction forces. This negatively affects adhesion, mainly in peripheral areas. These differences in microtensile bond strength do not result from morphologic differences in enamel.
Dr. E. Asmussen made valuable suggestions, and Dr. C. Davidson helped extensively in the final version of the manuscript. Research was supported by the Dept. of Estomatología II, Complutense University. Part of a thesis to be submitted in partial fulfillment of the requirements for the PhD degree of Mrs. Cabrera. Partially presented as a poster to the 41st Annual Meeting of the Continental European (CED) and the Scandinavian (NOF) Divisions of the International Association for Dental Research (IADR). Amsterdam, Netherlands, September 15–17, 2005. Received for publication March 26, 2006. Revision received October 27, 2006. Accepted for publication November 7, 2006.
Journal of Dental Research, Vol. 86, No. 3,
227-231 (2007) This article has been cited by other articles:
|
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
|
|
 |
|
Sign In to gain access to subscriptions and/or personal tools.
TOP
ABSTRACT
INTRODUCTION
