INTRODUCTION

Effective bonding to tooth tissue using tooth-colored restorative materials is an absolute necessity for clinical success. Although enamel adhesion is a predictable and established entity in contemporary restorative dentistry, an adequate bond to dentin is more difficult to achieve. This is partly due to the biologic characteristics of dentin; namely, a high organic content, tubular structure with the presence of odontoblastic processes, continuous moist conditions due to the presence of dentinal fluid, intratubular pressure, permeability and the presence of the smear layer formed immediately after cavity preparation (Lopes & others, 2002). There are many variables that can influence the bond strength of resin composite to dentin (Pashley & others, 1995). These include substrate variables, etching variables, priming variables, bonding variables and testing variables.

Many in vitro studies regularly report high bond strength of newly developed dentin bonding systems (Ogata & others, 2001; Bouillaguet & others, 2001; Rosa & Perdigão, 2000). One of the drawbacks of these bond strength studies was that they had been conducted without maintaining intrapulpal pressure during bonding. Compared to laboratory conditions, dentin is an intrinsically hydrated tissue in vivo, penetrated by a network of 1.0–2.5-μm diameter fluid-filled dentin tubules. The flow of fluid from the pulp to the dentin-enamel junction is the result of a slight but constant pulpal pressure (Brännström, Linden & Johnson, 1968). It is also difficult to simultaneously achieve uniform wetness on the axial, pulpal and gingival walls and between the superficial and deep dentin in Class II cavities because of regional differences in hydraulic conductance (Ozok, Wu & Wesserlink, 2002). Although the flow along individual tubules is very small, the average velocity of outward fluid flow per tubule is 1–2 μm/seconds. This is sufficient to oppose and greatly reduce the inward diffusion of substances from the oral environment (Stead, Orchardson & Warren, 1996). Lower values of bond strength have been reported in previous studies when intrapulpal fluid pressure was maintained (Mitchem & Gronas, 1991; Pioch & others, 2001; Mitchem, Terkla & Gronas, 1988).

Moreover, these laboratory studies are performed on flat tooth surfaces prepared by using silicon carbide abrasive papers; whereas, in a clinical situation, complex cavity designs are prepared by different cutting instruments such as diamond or steel burs (Bouillaguet & others, 2001). This geometrically determined contraction stress has been described as the C-factor (configuration factor) by Davidson, de Gee and Feilzer (1984). C-factor is the ratio of bonded to unbonded walls of the preparation. The C-factor for dental restorations typically ranges from 0.1 to 5.0, with higher values (>1.5) indicating a greater likelihood of high interfacial stresses (Feilzer, de Gee & Davidson, 1987). Box-like Class I cavities in which the walls have equal dimensions would have a C-factor of 5; whereas a flat surface, as in veneering, would have a C-factor of 1 (Bouillaguet & others, 2001). Laboratory studies conducted at a C-factor of 1 tend to overestimate bonding performance compared with complex cavity preparations with high C-factors (Yoshikawa & others, 1999; Bouillaguet & others, 2001).

A key element for adhesion is the intimate association of the adhesive and substrate. The wetting ability and extent to which the adhesive penetrates the tooth surface play a major role in determining the quality of bonding (Erickson, 1992). A major factor in the surface wetting of bonding resins is the surface topography of the prepared tooth surface, since a roughened surface creates a greater surface area for the adhesive bond (Eick & others, 1972; Vaysman, Rajan & Thompson, 2003). Therefore, information on the effect of cutting teeth with different burs on resin-dentin bond strength is essential for the appropriate clinical use of dentin bonding systems (Ogata & others, 2001).

After mechanical preparation of the cavity with a dental instrument such as a bur, an amorphous layer of organic and inorganic debris, such as the smear layer, is formed on the surface (Pashley, 1984). This dentin smear layer occludes dentinal tubules and reduces dentin permeability by 86% (Pashley, Livingstone & Greenhill, 1978). It is an established fact that the quantity and quality of the smear layer vary depending upon the manner in which they are created (Eick & others, 1970; Gilboe & others, 1980). Differences in smear layers prepared with bur cutting or abrasive paper have been reported to affect the bond strength of resin composite to dentin using self-etching primers (Ogata & others, 2001). Watanabe, Nakabayashi and Pashley (1994) reported that dentin bond strengths were affected by different smear layers created by different grits of abrasive papers using self-etching primers. On the other hand, Ogata and others (2002) reported that bond strength is affected by different types of smear layer in the case of self-etching primers only and not in the case of phosphoric acid etching.

Additionally, other factors which are under the control of the operator, such as cutting speed, have also been found to be a contributing factor for bond strength. It is still unclear whether the bur or speed is the crucial factor in bond strength determination. Bond strength determination and mode of failure estimates the adhesion quality of tooth-resin interface.

This study evaluated the effect of cutting teeth with burs of different types operated at various speeds on shear bond strength using Prime and Bond NT bonding agent. The null hypothesis was that different surface preparation methods would have no effect on bond strength at the tooth-resin interface.

METHODS AND MATERIALS

Sixty extracted, caries-free human molars were cleaned and stored in distilled water until used for shear bond strength testing. The teeth were divided into the following six groups, with 10 teeth in each group, according to the type of cutting instrument and speed used (Table 1).

1. Fine grit straight fissure diamond bur in air rotor (DA).

2. Fine grit straight fissure diamond bur in micromotor (DM).

3. Crosscut fissure carbide bur in air rotor (CCA).

4. Crosscut fissure carbide bur in micromotor (CCM).

5. Plain fissure carbide bur in micromotor (CM).

6. #600-grit silicon carbide abrasive paper (SiC).

Table 1 Identification of Groups by Dentin Surface Preparation

View Fullscreen

All the teeth were kept under positive hydrostatic intrapulpal fluid pressure via pulp chambers filled with distilled water during tooth preparation and bonding procedures to simulate clinical conditions. To create a mechanism into the pulp chamber, a hole, 4-mm in diameter, was made in the furcation area of the teeth, between the roots. The roots were covered externally by a rubber sheath; the junction between the rubber sheath and tooth surface was covered by cyanoacrylate and clay to maintain the air-tight seal. The lower portion of the same rubber sheath covered a plastic tube filled with distilled water, which connected the pulp chamber to a water-filled plastic syringe (Nikaido & others, 1995). The column height of water was adjusted to 34 cm to provide approximately 25-mmHg of pressure, which is the average tissue pressure in healthy pulp (Mitchem & others, 1988; Van Hassel, 1971). During tooth preparation and restoration, this intra-pulpal fluid pressure was maintained for seven days (Figures 1 and 2).

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

This study evaluated the bond strength of composites to both enamel and dentin. Class II proximal box-only cavities of standard dimensions 4-mm buccolingual, 4-mm occlusogingival and 2-mm mesiodistal, with facial and lingual walls straight and parallel to each other, were prepared. The standard dimensions included both the enamel and dentin surfaces (Figures 3 and 4). In all groups, an initial preparation was done by fine-grit straight-fissure diamond bur in air rotor, then the surface was finished with different types of burs. Teeth in the air rotor groups, A and C, were prepared with burs rotating in a dental turbine at a high speed of 150,000 rpm (Contraangle PANA AIR T air rotor handpiece, NSK, Nakanishi Inc, Tochigi-ken, Japan). Teeth in Groups B, D and E were prepared with their respective burs mounted in a contra angle micromotor handpiece at 40,000 rpm (NSK, Nakanishi Inc). Handpieces were hand-held to simulate clinical conditions. In Group F, #600-grit silicon carbide abrasive paper was used. A small piece of this paper was glued to a metallic blank bur with cyanoacrylate. This blank was mounted in the micromotor at slow speed (40,000 rpm), and 30 passes were made across the tooth surface under copious air-water spray. The surface was then prepared by 10 strokes with the same mounted silicon carbide abrasive paper on a blank bur when the micromotor was not rotating to create uniform scratches. The abrasive paper was changed as soon as it got distorted. In all groups, the tooth surface was prepared by the bur under copious air-water spray until uniform scratches by each bur covered the entire tooth surface. Each bur was changed after preparing three cavities.

View Figure

• Download as Powerpoint
• Download Figure

View Figure

• Download as Powerpoint
• Download Figure

The prepared enamel surface was etched with 38% ortho phosphoric acid (DPI tooth conditioning gel, Dental Products of India Ltd, Mumbai, India) for 30 seconds, with the help of a brush. Then, the etchant was applied to the prepared dentin surface for 15 seconds. This resulted in enamel etching for 45 seconds and dentin etching for 15 seconds. The etchant was rinsed with distilled water for 10 seconds with a high force of combined air-water spray and blot dried with a cotton pellet to keep the surface moist. For this purpose, excess water from a cotton pellet saturated with water was removed by blotting it on a gauze pad before using the pellet to blot the tooth. Then, one coat of the bonding agent Prime and Bond NT (Dentsply/DeTrey, Konstanz, Germany) was applied with a bristle brush. The surface was kept wet for 20 seconds and gently air-dried for five seconds to a glossy surface, then photo-polymerized using a light intensity of 600 mW/cm² for 10 seconds. A mylar strip was now applied to cover the proximal box. Spectrum TPH, the hybrid resin composite, (Dentsply/DeTrey) was packed in 2-mm thick horizontal increments. The last layer was made flush with the enamel cavosurface margins. Each layer was exposed to the curing light for at least 40 seconds from the occlusal side. Details for all restoration materials are provided in Table 2. Finishing and polishing of the composite restoration was not done in order to eliminate the influence of these variables on the surface properties of the composite restoration. All teeth were stored in distilled water at 37°C for seven days, maintaining intrapulpal pressure.

Table 2 Materials Used for Bonding

View Fullscreen

The specimens were removed from the assembly apparatus and placed in a fixture that positioned them for loading at an angle of 45° (Summitt, Della Bona & Burgess, 1994; Della Bona & Summit, 1998). For this purpose, metallic rings with a one- and two-inch high diameter were used. The upper half of the ring was cut at a 45° angle to the long axis of the ring (Figure 5). The rings were filled with auto polymerizing acrylic resin up to their uppermost margin. All the specimens were placed in the rings, with the roots covered with acrylic resin. The specimens were placed perpendicular to the acrylic resin surface. The mounted specimens in rings were stored in distilled water until testing was performed.

View Figure

• Download as Powerpoint
• Download Figure

Bond strength between the restorative material and tooth surface was measured in the shear mode with the Universal Testing Machine Z 010 (Zwick, Ulm, Germany). The specimens were mounted in a jig, while a straight knife-edge rod 2-mm-wide was applied at the tooth-restoration interface at a crosshead speed of 2-mm/minute. This resulted in a shear force at a 45° angle to the tooth surface. Load was applied until restoration failure occurred. Bond strength was recorded in Newtons. The total bonded surface area of the proximal box cavity preparation was 40 mm², and it was calculated as the sum of the surface area of the gingival wall (8 mm²), facial wall (8 mm²), lingual wall (8 mm²) and axial wall (16 mm²). Loads were converted to MPa by dividing the loads in Newton by the total bonded surface area.

The mode of failure of the bond at tooth-restoration interface was determined with a Stereomicroscope at 10x magnification, then classified into three categories: adhesive, cohesive and mixed types of failure (Royer & Meiers, 1995).

1. Adhesive mode of failure was recorded if the restorative material was completely detached from the tooth surface.

2. Cohesive mode of failure was recorded if the bond failure occurred entirely within the restorative material.

3. Mixed mode of failure was recorded if the bond failure was a combination of the adhesive and cohesive modes of failure.

Statistical analysis of the shear bond strengths was performed using one-way, two-way ANOVA and t-test at the 95% level of confidence.

RESULTS

Figure 6 and Table 3 show the shear bond strength results of each group. For air rotor, there was no statistically significant difference in mean bond strength among fine grit straight fissure diamond burs and crosscut fissure carbide burs: DA: 258.04 ± 150.18 N (6.45 MPa); CCA: 288.501 ± 196.22 N (7.21 MPa). For micromotor, the silicon carbide paper group produced the lowest mean bond strength, and a statistically significant difference was observed among all groups: DM: 64.81 ± 42.71 N (1.62 MPa); CCM: 243.86 ± 110.94 N (6.09 MPa); CM: 51.03 ± 47.95 N (1.27 MPa); SiC: 26.93 ± 14.71 (0.67 MPa). One-way and two-way ANOVA analysis revealed a statistically significant interaction between type of bur and speed used for tooth surface preparation.

View Figure

• Download as Powerpoint
• Download Figure

Table 3 Results of Shear Bond Strengths for Each Group (mean ± SD)

View Fullscreen

When examined under a stereomicroscope (10x), the representative micromorphology of the failure pattern was classified as either adhesive or mixed. There was no remarkable difference in failure patterns among all groups.

DISCUSSION

The use of the adhesive systems in restorative dentistry allows for use of more conservative preparations, a reduction of microleakage in the tooth-restoration interface and the prevention of recurrent caries, marginal discoloration and the reduction of postoperative sensitivity. One of the primary objectives of researchers is to achieve a strong, durable, predictable union between restorative materials and tooth structure. Shear bond strength tests have been the method of choice for testing tooth bonding and are included in the International Organization for Standardization for testing dental materials and adhesives, ISO's TR 11405 (Blomlof & others, 2001).

In a tooth composed of brittle materials, the stresses responsible for failure, called Von Mises stresses, are a combination of the resultant normal stresses and shearing stresses (Yaman, Alacam & Yaman, 1998). Load applied at a 45° angle simulates the effect of high shearing stresses (Yaman & others, 2000). Microtensile tests were not used in this study, because they are not easy to perform, and they are labor intensive, technically demanding, require special equipment, rapid dehydration of small samples occur and it is difficult to measure bond strengths less than 5 MPa (Pashley & others, 1995). A push-out test is more time-consuming and cannot be applied for evaluating enamel bond strength. Plus, there are chances of fracture of adjacent tooth structure during testing (Van Meerbeek & others, 2003).

Shear bond strength values recorded in this study are lower—in the range of 0.67-7.21 MPa. These are the total bond strength values obtained by resin composites to both enamel and dentin surfaces. Bouillaguet and others (2001) reported a bond strength of 18.5 MPa for flat dentin surfaces and a bond strength of 13 MPa for mesioocclusodistal cavities, including both enamel and dentin surfaces. Rosa and Perdigão (2000) observed bond strengths of 18.2 MPa and 23.4 MPa for flat dentin and enamel surfaces, respectively. Perdigão, Baratieri and Lopes (1999) reported bond strengths of 20.5 MPa and 27.0 MPa for flat dentin and enamel surfaces, respectively. These lower values could be attributed to polymerization shrinkage forces with a high C-factor in Class II cavities, a large bonded surface area, more axial dentin permeability and interference by dentinal fluid due to intrapulpal pressure.

Bouillaguet and others (2001) reported a 20% reduction in bond strength in the cavity bonding group as compared to those measured in the flat bonding group. Yoshikawa and others (1999) reported similar results for bulk-filled Class I cavity surfaces. The lower shear bond strength recorded in this study can be explained by the fact that shrinkage forces in Class II cavities have a high C-factor that cannot be relieved by resin flow, resulting in debonding of one or more walls as compared to flat surfaces with a low C-factor (Bouillaguet & others, 2001).

The results of this study are supported by studies by Mitchem and Gronas (1991), who reported a low bond strength of 1 ± 1.3 MPa in tubules full of fluid under pressure as compared to 18.3 MPa in empty tubules. Pioch and others (2001) reported that bond strength using Prime & Bond NT dropped significantly from 14.8 to 8.7 MPa when the dentin of extracted teeth was perfused under intrapulpal pressure (34cm H₂O). The presence of fluid inside the dentinal tubules tends to dilute the dentin conditioner, decrease its potential for demineralization of the intertubular and peritubular dentin and eventually lower bond strength (Perdigão & others, 1996).

The purpose of this study was not to ratify how good the bonds were, but simply to compare the bond strength achieved by different burs at different speeds. Consequently, negative controls (the same burs using the same test methods with no pulpal pressure) were not included.

In this study, all burs in micromotor, except crosscut fissure carbide burs, resulted in significantly lower bond strength than in air rotor. This can be explained by the fact that cutting in micromotor requires a relatively heavy force application and produces vibrations of high amplitude and low frequency (Marzouk, Simonton & Gross, 2001). At low speeds, burs have a tendency to roll out of the tooth preparation, mar the proximal margin of tooth surface and leave micro-cracks. The presence of micro-cracks in the bonding interface may leave flaws that will diminish the bond strength of resin composite to tooth surface.

There was no statistically significant difference in mean bond strength attained by crosscut fissure carbide burs in air rotor and micromotor and between fine grit straight fissure diamond burs and crosscut fissure carbide burs in air rotor. A possible explanation might be that etching may have produced a similar bonding substrate despite of the different types of smear layers created by different cutting instruments.

The lower values of bond strength with fine grit straight fissure diamond burs, compared to crosscut fissure carbide burs in micromotor, can be attributed to more frictional stresses created due to greater contact between the diamond fine grit and tooth surface as compared to a cutting instrument like a carbide fissure bur (Ogata & others, 2001). Watson, Flanagan and Stone (2000) reported no significant differences in sub-surface enamel cracking by diamond and carbide fissure burs; however, they reported a slight increase in temperature by frictional heat using diamond burs as compared to carbide fissure burs. Ogata and others (2001) reported a significantly lower bond strength with regular grit diamond burs at high speed as compared to both plain and crosscut fissure carbide burs at low speed using self-etching systems; while in this study, bond strength achieved with a fine grit straight fissure diamond bur was lower than a crosscut fissure carbide bur but significantly higher than a plain fissure carbide bur.

A higher bond strength with a crosscut fissure carbide bur as compared to a plain fissure carbide bur can be attributed to the increased surface area of tooth substrate. The true contact area between the materials involved may be much greater, because of a mechanically rough interface that allows for better infiltration of the bonding resins. The above findings are well supported by Vaysman and others (2003), who reported the highest reduction in etched and bonded dentin permeability by increased surface area produced by extensively serrated carbide burs.

Unlike the results of previous studies, lower bond strengths depicted with silicon carbide abrasive paper in this study might be due to inaccessibility of the instrument and variations in the amount of pressure application during each in Class II cavities.

The ranking of bond strengths from highest to lowest yielded the following results CCA> DA> CCM> DM> CM> SiC. Statistically significant differences existed (p<0.05). From these observations, it is clear that both the type of cutting instrument and its speed may affect the shear bond strength of composite-tooth interface, although the crosscut fissure carbide bur produced high bond strength at either speed used. These differences in mean bond strength values were recorded despite of etching with 38% orthophosphoric acid. These findings are in opposition to the study by Ogata and others (2002), who reported no significant difference in bond strength with different burs after phosphoric acid etching, as it completely removed the smear layer created by the different burs. The results of this study warrant further exploration to determine the degree of roughness with different burs after removal of the smear layer.

Lower bond strength exhibited in this study is further supported by the type of failures seen, which is predominantly adhesive and mixed in nature. These types of failure suggest poor bond strengths at the composite-tooth interface and an improvement in the wetting properties or chemical reactions, with the substrate probably being necessary to improve bond strength.

The variations in bond strength results obtained by the different researchers are significant. Standardization of test methods is needed in order to give comparable values that can be used for both guidance to users and further improvement in adhesives. Further research is required, as this study did not take into account the effect of interproximal contacts between teeth, the smooth carbide burs and silicon carbide paper in air rotor, a lack of standardized preparations by fixing the handpiece in a jig, the different types of bonding agents, the effects of temperature changes on restoration, the effects of longevity on the composite restoration and the different types of testing methods.

While significant progress has been made in the area of adhesion and esthetic restorative materials, numerous questions still remain unanswered in the area of adhesive materials. In particular, the confirmation of preliminary laboratory results with controlled clinical research seldom surfaces or is delayed beyond the commercial life of proprietary adhesive systems. Without such data, little confidence in adhesive behavior can be obtained. More importantly, without a thorough understanding of the performance of adhesives in vivo, knowledge will never be gained relative to adhesive mechanisms that are necessary to further technology in this area.

CONCLUSIONS

Within the limitations of this in vitro study, it can be concluded that shear bond strength values recorded by maintaining the intrapulpal pressure were lower. Groups with bond strengths from highest to lowest were CCA> DA> CCM> DM> CM> SiC. The bond strength achieved with a fine grit straight fissure diamond bur or a crosscut fissure carbide bur in air rotor and a crosscut fissure carbide bur in micromotor was significantly higher than a fine grit straight fissure diamond bur, a plain fissure carbide bur and #600 grit silicon carbide abrasive paper in micromotor. Therefore, both the bur and its speed may affect bonding, although the crosscut fissure carbide burs produced high bond strength at either speed used.