Variable conditions of the oral environment (pH changes, diet, saliva, temperature) negatively affect the clinical life of restorative materials and cause failures such as secondary caries, marginal leakage, deformation of restoration edges, fracture, and discoloration [14, 15]. This is caused by the water absorption of the resin composite during the process in which it functions in the mouth and thus the degradation of the bonding surface between the organic resin matrix and inorganic filler particles [16]. Repair of failure resin composite restorations with resin composite is a favorite restorative technique [17]. This is because it is easier to apply compared to indirect restorative materials, the treatment process is short and it can be performed at a relatively lower cost [18, 19]. To test the repair bond strength in the study, all sample surfaces prepared with Filtek Z250 were first roughened using 320-grit silicon carbide-coated sandpaper. Thus, it was aimed to simulate the standardized procedure of restoration correction using a diamond bur in the clinic [20]. The samples were then tested for 5,000 thermal cycles ranging from 5–55 0C to simulate the process by which the restoration functions in the mouth [21, 22]. It has been demonstrated previously that the number of 5,000 cycles corresponds to the six-month aging process of the samples [21]. The thermal cycling test applied before the repair procedure was also applied to the samples after the repair procedure to expose the samples to the conditions of the oral environment during the six-month post-repair period.
As a result of the study, it was determined that the lowest shear bond strength values were obtained in the control group where no treatment was applied to the sample surfaces, followed by the bond values of the group where Er,Cr:YSGG laser was applied, and the highest bond strength values were obtained from the samples in the group where the sample surfaces were sandblasted with CoJet. This finding is not surprising since no surface preparation process was applied to the sample surfaces in the control group. This finding supports the findings of Covalcanti et al. [23] who reported that roughening the resin composite surface only with a diamond bur caused lower bond strength values to be obtained. In this study, the smear layer formed on the sample surface may affect the higher bond strength values obtained in Er,Cr:YSGG laser-treated samples compared to the control group. While a smear layer was formed on the resin composite surface after the diamond bur application, it was reported that no smear layer was formed on the surface after Er,Cr:YSGG laser treatment. The findings of the studies reporting that higher bond strength values were obtained on Er,Cr:YSGG laser-treated sample surfaces compared to diamond burs support the findings obtained from our study [23–26]. When the relationship between the surface preparation methods was analyzed, it was found that there was a statistically significant difference between the control group with no surface application and the other surface preparation groups. It was determined that there was a significant difference between Group 9 and Group 10, where only universal adhesive was used in self-etch mode and Filtek Z250 was used as a repair material, between the samples sandblasted with CoJet and Er,Cr:YSGG laser. The difference in the shear bond strength values obtained from the CoJet sandblasting and Er,Cr:YSGG laser-treated groups may be due to the different effects of surface preparation methods on the resin composite surface. This finding is similar to the findings of Dursun et al. [27] and Rashidi et al. [28] who reported that higher bond strength values were obtained in CoJet blasted samples compared to Er,Cr:YSGG laser-treated samples. Based on these findings, the first null hypothesis of our study, "Different surface treatments have no effect on repair bond strength." was partially rejected.
In the study, it was observed that the highest bond strength values in all surface preparation groups were obtained in the groups where the universal adhesive was used in total-etch mode, and higher shear bond strength values were obtained in the groups where the universal adhesive was used in self-etch mode. This finding is similar to the findings of studies reporting that higher bonding to dental tissues and resin composite is obtained with the use of adhesive systems in total-etch mode [29, 30]. Influence of different etching modes on bond strength and fatigue strength to dentin using universal adhesive systems. Based on these findings, the second null hypothesis of our study, "The application of the universal adhesive system in different modes (total-etch or self-etch) has no effect on the repair bond strength." was generally rejected. The universal adhesive Scotchbond Universal Plus used in the study differs from other universal adhesive systems in that it contains a double silane monomer. The manufacturer says that this makes it unnecessary to apply silane as an additional step [31]. Based on this information, no additional silane was applied to the sample surfaces prepared in the study. However, since this adhesive system used in this study has not been compared with other adhesives, further studies on this subject are needed.
Another finding obtained in this study is that higher shear bond strength values were obtained in the groups where Filtek Ultimate was used as repair material compared to Filtek Z250 and a statistical difference was generally detected between them. Based on this finding, the third null hypothesis of our study, "The use of different resin composites as repair material has no effect on repair bond strength." was generally rejected. Although the organic matrix of both resin composites is similar, the filler percentages and types are different. The organic matrix of both resin composites used in the study contains Bis-GMA, Bis-EMA, UDMA, and TEGDMA monomers. The inorganic components of the micro-hybrid resin composite Filtek Z250 consist of 0.01–3.5 µm silica and zirconium fillers. The percentage of filler added is 75–85 by weight and 60 by volume. The inorganic components of the nanofiller resin composite Filtek Ultimate consist of 20 nm silica, 4–11 nm zirconium, and 0.6–10 µm silica. The proportion of fillers added is 72.5 by weight and 55.6 by volume [32]. Nassoohi et al. [33] reported that they obtained higher repair bond strength values with micro-hybrid resin composites than with nanofiller resin composites in the study in which they examined the repair bond strength of two nano-filled and one micro-hybrid resin composites. This finding is opposite to the findings obtained in our study. This may be due to the different surface preparation methods as well as the structural properties of the resin composites tested in both studies.
Based on all the findings obtained in the study, we can say that the use of the universal adhesive in total-etch or self-etch mode does not affect the bond strength values if the repair application will be performed using a nano-filled resin composite such as Filtek Ultimate after only bur removal on the resin composite surface, on the contrary, if a micro-hybrid resin composite such as Filtek Z250 is used, higher bond strength values can be obtained by using the universal adhesive in total-etch mode. This finding is partially opposite to the findings of Irmak et al. [34]. This may be due to the different adhesive systems used in both studies. When sandblasting was performed with CoJet on the resin composite surface, it was determined that the use of universal adhesive in total-etch mode in both resin composites used as repair material caused an increase in bond strength values. This finding is in agreement with the findings of Nassoohi et al. [33] who reported that they obtained higher bond strength values with 35% phosphoric acid application after sandblasting. However, their methodology was different from our study. The findings obtained from the Er,Cr:YSGG laser-treated groups were similar to the findings obtained from the CoJet sandblasting group. Therefore, we believe that if Er,Cr:YSGG laser is to be applied to the sample surfaces, the use of universal adhesive in total-etch mode, regardless of what is used as repair material, will increase the bond strength values. We think that the use of the adhesive system in total-etch mode contributes to the cleaning of the sample surfaces after Cojet blasting or Er,Cr:YSGG application. Accordingly, the adhesive system may have penetrated the irregularities formed on the surface better, and higher bonding values may have been obtained.
Studies are showing that adhesive-type fracture is more common in adhesive systems with low bond strength values and cohesive and mixed fracture types are more common in adhesive systems with high bond strength values and that there is a relationship between bond strength values and fracture type [35]. When the fracture types obtained in the study are examined, it is difficult to make such a conclusion. This may be due to the methodology of the studies and the different materials used.
The study findings of Ahmadizenouz et al. [36], who reported that the most cohesive type fracture was observed on the sample surfaces roughened with diamond burs, are opposite to the findings obtained from the control group. This may be because the materials used in the study were different from the materials used in our study. Similar to our study, Fornazari et al. [37] reported that more cohesive fractures were observed in the specimens sandblasted with CoJet. The study findings of Kiomarsi et al. [38], who reported that more adhesive-type fractures were observed on Er,Cr:YSGG laser-treated sample surfaces, support our study findings.
In the evaluation made by SEM, a smoother surface image was obtained in the control group since no treatment was applied to the sample surfaces. On the opposite, it was observed that the sample surfaces in the group sandblasted with CoJet were rougher and aluminum oxide particles were present on the surface. In the Er,Cr:YSGG laser-treated group, it was found that wave-shaped craters were formed on the surface, but no side effects related to laser use occurred. These findings support the findings obtained as a result of the shear bond strength test. It is also in agreement with the findings of Cho et al. [39]. EDS scanning after SEM evaluation revealed that only the filler particles of the resin composite were present on the surfaces of the control group or Er,Cr:YSGG laser-treated samples. This shows that the laser parameters we used were correct and no thermal damage occurred on the resin composite surface during laser treatment. The sample surfaces sandblasted with CoJet showed the formation of aluminum oxide particles in addition to the filler particles of the resin composite. This finding supports the findings of studies reporting that aluminum oxide particles are embedded in the surface after sandblasting with CoJet.