INTRODUCTION
Composites have become the material of choice in restorative dentistry because of their adequate mechanical behavior, aesthetic properties and above all, conservation of healthy tissue. According to Ferracane et al. 1 , one of the deficiencies of conventional composites with regard to the polymerization reaction is that the volume of the material contracts by about 3% 2 . This contraction is transmitted to the interface between dental tissue and restorative material and may cause marginal filtration, secondary caries, loss of the restoration, cuspal deflection 3 , and enamel micro-cracks 4 , leading to postoperative sensitivity, usually during chewing. Various techniques and composites have been developed to minimize polymerization contraction and its clinical effects. The incremental technique is widely recommended for minimizing these problems. Placing layers 2 millimeters thick enables light to reach deeper zones, thereby achieving an adequate degree of conversion. However, this technique has some disadvantages such as inclusion of air bubbles, increased risk of contamination between layers, and longer working time 5 , 6 . Most improvements have focused on modifying the filling, which has improved mechanical properties, mainly resistance to wear. Despite this progress, average lifespan of these composite restorations was found to be only 10 years 7 . Initially, composite matrix was exclusively based on the chemistry of methacrylate, more specifically BisGMA, TEGDMA, BisEMA and UDMA, and in the past 20 years, approximately, alternative monomers have been developed with the aim of reducing polymerization contraction and stress, emphasizing the association between the development of stress and the formation of gaps between the restoration and the tooth structure 8 . One alternative attempted was to apply ring opening polymerization or monomers of very high molecular weight. In the former, the only material developed was Filtek LS (3M), based on the silorane chemistry. Both strategies were successful in reducing the contraction coefficient to 1% 9 and above all, reducing polymerization stress, according to in vitro assessments 10 . Although there are a few available clinical trials, results reported are contradictory. Popoff et al. found similar clinical behavior after one year with silorane-based and dimethacry late-based resins in restorations 11 . However, they note that studies should have longer follow-up periods. Gonqalves et al. conducted an 18-month double blind, randomized study, finding that marginal integrity was worse in restorations with silorane-based composites than in those with dimethylacrylate, finding no benefit in using this kind of composites in the restoration of Class II
lesions 12 .
More recently, greater importance has been assigned to improving resistance to breakdown in the oral medium -including hydrolysis of ester groups present in the methacrylates- caused by saliva and bacterial enzymes, and to preventing biofilm formation on the surface and interface of a composite restoration 13 . The most recent strategy has been to develop materials requiring fewer steps in their protocol for use, such as bulk-fill and self-adhesive composites. Bulk-fill composites have become increasingly popular for general practices. Although they can be classified in several different ways, they are best identified according to in-depth polymerization capacity, since they may have high or low viscosity, higher and lower ceramic load, and a great variety of mechanical properties 14 . The term bulk-fill has thus been used by manufacturers to refer to composites that can be inserted and polymerized in a single block of 4-5 mm. Some examples of the modifications made to conventional composite are use of monomers with high molecular weight such as AUDMA (Aromatic urethane dimethacrylate) and monomers known as AFM (addition-fragmentation monomers) 15 . Regarding ceramic filling, the percentage in volume was reduced in both bulk-fill flowable composites and bulk-fill composites with regular viscosity, with the percentage of filling in volume being even lower in the former. This reduction in percentage of filling reduces the difference in the refraction index between the matrix and the ceramic filling, which increases translucency for the material, enabling in-depth curing 16 . Moreover, the elastic modulus of these materials is lower than in conventional composites, and the rigidity of bulk-fill flowable composites is even lower than that of regular viscosity bulk-fill composites. This is why bulk-fill flowable composite manufacturers recommend placing an additional 2 mm layer of a conventional composite in zones exposed to greater stress 17 . The main properties that have been studied are degree of conversión and depth of cure 18 . However, no papers were found studying bond strength between composite Z350 3M ESPE -Filtek Bulk Fill 3M ESPE and Z350 3M ESPE-Surefil SDR flowable or the mechanical properties of these combinations. Thus, the aim of this study was to determine shear bond strength and compressive strength of different combinations of bulk-fill composites and conventional composites.
MATERIALS AND METHODS
The following materials were used: Filtek Bulk Fill (FBF) and Z350 (both 3M-ESPE), and Surefil SDR flow (SFF) (Dentsply). To determine shear bond strength, test specimens (n=3) 10 mm thick and 4 mm in diameter were prepared in cylindrical molds. This 10 mm thickness consisted of 6 mm of one material and 4 mm of another in the following combinations: G1: FBF- FBF; G2: Z350-Z350, G3: FBF-Z350, G4: Z350-SFF and G5: SFF-SFF. They were polymerized using a Coltolux LED light-curing unit (Coltene) with intensity 1100 mw/cm2 for 20 seconds per layer, following the thickness is recommended by the manufacturers in their respective instructions. An extra-fine marker (Edding 1880 Drawliner 0.1) was used to draw a line between the two parts. Specimens were embedded in self-curing acrylic resin cylinders (Subiton SL, Argentina) to make them easier to handle during mechanical assays. Specimens were stored for 24 hours at 37 °C in distilled water. The blocks prepared were placed on a support and load was applied at the level of the interface between the two materials with a metal needle until fracture occurred. Bond strength was determined using a universal testing machine for mechanical assays (1100. Instron Corporation) at a crosshead speed of 1 mm/minute. Fig. 1 shows the system used for determining shear bond strength. To determine compressive strength, cylindrical specimens (n=4) 4 mm in diameter and 6 mm thick were prepared. These 6 mm corresponded to 4 mm + 2 mm of the same combinations as described above. Specimens were stored in distilled water at 37 °C for 24 hours. Compressive strength was determined by placing an axial load using the same machine at the same crosshead speed as described for bond strength. Results were analyzed using ANOVA and Tukey’s test with significance level p<0.05.
RESULTS
Mean and standard deviation in MPa for each group were the following: Bond strength: G1: 435.87 (65.86), G2: 233.6 (108.15), G3: 279.2 (22.05), G4: 449.1 (109.35) and G5: 196.6 (51.16); Compressive strength: G1: 138.61 (18.92), G2: 156.06 (9.71), G3: 167.18 (35.89), G4: 199.3 (39. 28) and G5: 171.23 (28.71). Figs. 2 and 3 show mean and standard deviation for each. Evaluation with ANOVA showed no significant difference between combinations for compressive strength (p>0.05) but did show significant differences in bond strength (p<0.05). Tukey’s Test showed three homogeneous groups for comparison of shear bond strength, suggesting statistically significant difference between G4 and G5 (between combination G4: Z350-SFF and combination G5: SFF-SFF. This is shown in Fig. 2 with the same letters indicating homogeneous groups.
DISCUSSION
The introduction on the market ofbulk-fill composites raised two major questions: one regarding their ability to be photopolymerized adequately at the depths stated by the manufacturer, and the other regarding whether their mechanical properties might be deficient. As mentioned in the review by Camila Nuñez et al. regarding the advantage of shorter operative times, it would be interesting to evaluate whether the single block technique really saves clinical time 19 . Tiba et al. report that using bulk-fill flowable composites, which require an additional 2 mm occlusal layer of conventional composite, necessarily involves filling cavities with at least two increments, with different composites, which may not differ much from the operative times required for 4 mm cavities filled with conventional composites 20 . As there is some evidence of the lower mechanical properties in bulk-fill flowable and regular viscosity composites, dentists often choose to use them as a base and cover them with conventional composites. The in vitro studies evaluating marginal seal have found results which are comparable to those using conventional composites. However, the same is not true for the evaluation of their mechanical properties 21 , 22 . Esteves Lins et al. evaluated the mechanical properties of different composites, finding results in which they were unable to associate composite type -in terms of form of insertion- with compressive strength, having found similar values with no statistically significant difference compared to those inserted as a block 23 .
This is consistent with the behavior observed in the current study, where there was no significant difference between the different combinations. This may also be attributed to the fact that this type of material does not differ from conventional materials in ceramic load only, but in a combination offactors 24 . De Assis et al. reached similar conclusions 25 . According to Haughen et. al., the composite FBF presents a specific monomer with high molecular weight without free hydroxy groups in order not to increase its viscosity, in addition to lower ceramic content. Although this should provide better light penetration, some of the particles are silica or silanized zirconia, which have a high refractive index, leading to lower light transmission than in other bulk fill composites such as SFF, which could cause a lower degree of conversion 14 . In our study, group 4 (Z350- SFF) had the highest mean bond strength, even though different materials had been combined. This could be associated to the lower degree of conversion of monomers in the organic matrix of Z350 26 , which may leave a larger number of double bonds available for bonding to SFF. In turn, SFF has the lowest percentage of ceramic filling in volume (45 vol%), which may also contribute to providing a larger quantity of available monomers for bonding between the two materials.
Under the experimental conditions in this study, it can be concluded that the combinations of materials evaluated have similar mechanical behavior. However, bonding between them was affected in the combinations.