INTRODUCTION
Resin-based materials are widely used for anterior and posterior tooth restoration. Nevertheless, although their properties make them suitable for clinical use, these materials shrink during polymerization1, which is a disadvantage associated with clinical performance. The polymerization shrinkage of resin composites occurs due to the conversion of monomers into a polymer structure2 accompanied by shrinkage stress, considered a multifactorial phenomenon determined by different factors (e.g., volumetric shrinkage, viscoelastic behavior, kinetics of reaction)3. Moreover, chemical and mechanical stresses in the oral environment can have consequences due to material characteristics such as defects in restoration/tooth interface, debonding, enamel micro-cracking, postoperative sensitivity, and cusp deflection4-7.
The type of resin composite and clinical technique of application can affect the restoration properties4, 5. The insertion of several small increments involves more clinical variables and increases the mean time of the procedure, beyond the difficulty of filling small cavities8, besides being dependent on the skill and expertise of clinicians. Thus, bulk-fill composites have emerged on the market to enable restorations in layers of up to 4 or 5 mm, according to manufacturers. These materials are commercially available as low- or high-viscosity composites9, 10. Low-viscosity bulk-fill composites (flow/flowable) are usually inserted in cavities/tooth preparations with tips and may require a layer of conventional resin composite on top of the restorations11. The high-viscosity bulk-fill composites (conventional/ paste/sculptable) have photoinitiators with adequate activation in response to the light-curing units12, which increase due to the resin’s translucency and enable the passage of light more easily. The advantages of these materials include the simplification of the restorative procedure, time saving, low shrinkage stress depending on the technology used by the manufacturer, and adequate radiopacity10, 12. However, the bulk application method of tooth restoration can be associated with debonding and greater shrinkage vectors13.
The properties and behavior of these materials need to be fully investigated. A previous study14 has suggested that a gradual decrease in the microhardness values from the top to the bottom is composite-dependent, and that an increase in thickness could have a negative effect on the microhardness of conventional resin composites and does not interfere with bulk-fill resin composites. Microhardness could even indirectly indicate the degree of conversion of the polymer network15 or depth of cure, and studies with different variables and purposes are necessary. A gradual reduction in microhardness values can indicate impairment in the degree conversion and consequently affect the longevity of restorations, which must have a suitable conversion to the base of the increment16. Considering that high-viscosity bulk-fill composites need longer curing times than low-viscosity bulk-fill composites for optimal properties17, it is necessary to evaluate the influence of viscosity of bulk-fill composites on the microhardness and mechanical behavior of restored teeth.
In addition to physical properties, it is also relevant to consider performance during the cyclic efforts of mastication, extensive cavities, or other physicochemical challenges that can occur in the oral environment. The influence of viscosity on bulk-fill resins used for restoration under mechanical cycling and fracture strength of posterior teeth should be investigated, especially considering Class II mesioocclusal- distal cavities with loss of marginal ridges, which are important strengthening structures for tooth resistance18. Furthermore, the bond strength to dentin needs to be evaluated in Class I occlusal restoration, whereas deep cavities can present high stress levels according to the technique or material applied19.
Thus, the objective was to evaluate the influence of high- or low-viscosity bulk-fill composites on microhardness, bond strength to dentin in deep occlusal restorations, and fracture strength in molars with mesio-occlusal-distal restorations submitted or not to thermomechanical cycling. The null hypotheses tested were: 1) Increment thickness or viscosity of bulk-fill resin composites does not interfere with microhardness values; 2) Bond strength to dentin of Class I restorations is not affected by the resin composite used; and 3) Restoration with bulk-fill composites of different viscosities, submitted or not to thermomechanical cycling, does not affect the fracture strength of the restored tooth.
MATERIAL AND METHODS
Experiment 1
Resin composite samples and Knoop microhardness analysis.
Disk-shaped samples (ø 6 mm) were made from a lowviscosity bulk-fill composite (SDR Flow, Dentsply, Milford, DE, USA) and high-viscosity bulk-fill composites (Filtek Bulk Fill, 3M ESPE, Saint Paul, MN, USA; and Tetric-N Ceram Bulk Fill, Ivoclar Vivadent, Schaan, Liechtenstein). Table 1 provides was performed using a Knoop indenter with a 50-g load for 15 s in a digital microhardness tester (Pantec HVS-1000, Digimess, São Paulo, SP, Brazil). Three indentations were made on the top and bottom of disk-shaped samples, keeping a 100 μm distance between indentations.
Experiment 2
Microtensile bond strength test in occlusal restorations.
After approval by the Local Ethics Committee in Human Research (# 87973218.3.000.5374), 30 recently extracted human third molars were used. Interproximal radiographs were performed to verify the distance between the central sulcus and the pulp chamber of each tooth, selecting teeth with a minimum distance of 5 mm. Occlusal cavities were made using a cavity preparation machine (Elquip, São Carlos, SP, Brazil) and cylindrical diamond tips positioned parallel to the long axis of the tooth. Cavity dimensions were 4 x 5 x 4 mm, checked after each cavity preparation using a digital caliper. Then the teeth were randomly divided into the following restorative treatments (n = 10):
Restoration using a conventional nanocomposite (Filtek Z350 XT; 3M ESPE, Saint Paul, MN, USA) and incremental filling technique as a control. The insertion was performed using small increments (2 mm) and light-curing each portion for 10 s.
Restoration using a low-viscosity bulk-fill composite (Filtek Bulk-fill Flow, 3M ESPE, Saint Paul, MN, USA). The insertion was performed applying a 4-mm layer and lightcuring for 20 s.
Restoration using a high-viscosity bulk-fill composite (Filtek Bulk Fill, 3M ESPE). The insertion was performed using a 4-mm layer and light-curing for 20 s.
A LED curing light (BluePhase, Ivoclar Vivadent AG, Schaan, Liechtenstein) was used for the photoactivation of the resin composite and adhesive system. All teeth were restored using a universal adhesive (Adper Single Bond Universal, 3M ESPE, Saint Paul, MN, USA) applied as a self-etching adhesive system.
After the restorative procedures, the teeth were stored for 7 days at 37 ºC. After this time, the teeth were individually fixed on an acrylic plate that was attached to a precision cutting machine (Isomet 1000, Buehler, Lake Bluff, IL, USA) and high-concentration diamond disc (Buehler) was used to serially section the samples, providing stick-shaped specimens composed of resin composite bonded to dentin. Each tooth resulted in approximately 4 sticks of 1 mm2.
Tensile testing was performed in a universal testing machine (EMIC, São José dos Pinhais, PR, Brazil). The sticks were individually attached to the grips of a microtensile device. The test was conducted at a crosshead speed of 0.5 mm/min until debonding or fracture of the stick, and the corresponding force values were obtained in newtons (N). The stick debonding tensions were calculated in megapascals (MPa) after measuring the bonding area with a digital caliper. The interface of the fractured sticks was examined under a stereoscopic microscope (30x magnification) to classify the fracture pattern. The fractures were classified such as: (a) adhesive; (b) cohesive in resin; (c) cohesive in dentin; or (d) mixed.
Experiment 3
Fracture strength in molars with mesio-occlusaldistal restorations.
Human third molars with MOD cavities were evaluated according to the following treatments/ techniques: intact tooth (control, no treatment or preparation); restoration with conventional microhybrid composite (Z250, 3M ESPE, Saint Paul, MN, USA); restoration with low-viscosity bulk-fill composite (SureFil SDR Flow, Dentsply); or restoration with high-viscosity bulk-fill composite (Filtek Bulk Fill, 3M ESPE). These teeth were submitted or not to thermomechanical cycling and were assessed, with n = 10, for fracture strength by axial compressive loading (ACL) and qualitative evaluation of fracture pattern.
Eighty human third molars without caries, stains, or cracks (Local Ethics Committee in Human Research approval - #60999616.4.0000.5374) were used in this assessment. The teeth were scraped with periodontal curettes (Duflex - SS White / Dental Articles Ltd., Rio de Janeiro, RJ, Brazil) and stored in 0.1% thymol solution. All teeth were submitted to dental X-ray to check the distance between the central sulcus and the pulp chamber, and teeth in which they were less than 5 mm apart were excluded. Other variables measured were total tooth, crown, and root length, in order to ensure that tooth sizes were distributed evenly among the groups. After the to size as small, medium and large, and randomly assigned to groups so that all groups consisted of similar amounts of each size. To simulate the periodontal ligament20, the dental roots were immersed in wax (# 7 Lysanda Produtos Odontológicos Ltd., São Paulo, SP, Brazil), obtaining a layer 0.2 mm thick. A polyvinyl chloride ring (PVC, Tigre S.A., Joinville, SC, Brazil) 25 mm in diameter and 25 mm high was placed around the root that was embedded with a polystyrene resin. After this, a soft polyether impression material (Impregum, 3M ESPE, Seefeld, Germany) was manipulated according to the manufacturer’s instructions and applied to the root. MOD cavities were prepared using a cavity preparation machine (Elquip, São Carlos, SP, Brazil) associated with copious air–water spray. The preparations were performed with 1/3 of the intercuspal width, within a 5 mm-deep occlusal box, and without a proximal box using a diamond bur (#3145 KG Sorensen Ind. e Com. Ltd, Cotia, SP, Brazil). All restored groups received an application of 35% phosphoric acid (3M ESPE) for 15 s for dentin and 30 s for enamel; flush with water for 15 s; removal of the excess water with a light air jet for 2 s; application of the adhesive system (Adper Single Bond 2, 3M ESPE, Saint Paul, MN, USA) according to the manufacturer’s recommendation; and photoactivation for 20 s by the LED curing light (BluePhase, Ivoclar Vivadent AG, Schaan, Liechtenstein). The groups were divided according to the following description:
Control: Intact tooth without preparation or restoration.
Conventional microhybrid composite (Z250): The composite was incrementally inserted in three oblique layers that were photoactivated individually.
Low-viscosity bulk-fill composite (SDR): The material was inserted in a single layer of 4 mm and photoactivated, followed by the layer insertion (1 mm) of microhybrid resin (Filtek Z250, 3M ESPE) and photoactivated.
High-viscosity bulk-fill composite (Filtek Bulk Fill): The resin was inserted in a single layer (5 mm) and photoactivated for 60 s: 20 s from the occlusal surface, 20 s from the buccal surface, and 20 s from the lingual surface.
The resin-based materials of all restored groups were photoactivated by the LED curing light unit mentioned above. Twenty teeth were subjected to each treatment, and half of each group was submitted to a thermomechanical cycling test, establishing n=10 per group.
Thermomechanical cycles were simulated to induce material fatigue (Elquip, São Carlos, SP, Brazil). The teeth received loading in the axial direction and were cycled 100,000 times with 50 N load and 2 Hz frequency. During the test, the teeth were stored at a relative humidity and submerged cyclically between 5 ºC and 55 ºC (1 min). The compressive loading test was performed in a Universal Testing Machine (EMIC DL 2000, São José dos Pinhais, PR, Brazil) with axial loading of compression, at 0.5 mm/ min (crosshead speed). The values obtained were expressed in newtons (N). After the fracture strength test, the teeth were evaluated for fracture pattern and classified as: (a) coronary fracture up to the middle third; (b) coronary fracture up to the cervical; (c) root fracture up to the cervical; and (d) severe root fracture in the middle and apical third.
Statistical analysis
The statistical models used followed the experimental design of each experiment. All analyses were performed at the SAS (SAS Institute Inc., Cary, NC, USA, Release 9.2, 2010) considering the significance level of 5%. After the exploratory analysis, the KHN data were submitted to split-plot analysis of variance (ANOVA) and Tukey’s test for multiple comparisons. The split-plot ANOVA was used because the experiment was performed considering two factors (bulk-fill composite and thickness), and the KHN values of the top and bottom were considered as a subplot. This analysis considered main factors, double and triple interactions.
The results obtained for occlusal restorations were evaluated by one-way ANOVA to determine whether the bond strength values were influenced by the resin composite used. A G-test was performed to assess the fracture pattern for microtensile bond strength test. For MOD restorations, the values of fracture strength after logarithmic transformation were analyzed by two-way ANOVA. The twoway ANOVA was used in order to consider the two factors (tooth restoration x thermomechanical cycling) and interactions. The fracture pattern was assessed by Fisher’s exact test.
The calculation of sample size was performed using GPower software. The sample size (n=10) was provided considering the power setting of 0.80, significance level of 0.05, and following parameters for the detectable minimum effect sizes: 0.51 (large) for KHN; 0.52 (large) for fracture pattern; and 0.38 (medium to large) for fracture strength.
RESULTS
The KHN results (Table 2) had the following p-values: p(composite) = 0.0004; p(thickness) = 0.04; p(composite vs. thickness) = 0.01; p(top/ bottom) = 0.51; p(composite vs. top/bottom) = 0.16; p(thickness vs. top/bottom) = 0.38; p(composite vs. thickness x top/bottom) = 0.11. There was no significant difference between the top and bottom for KHN values (p = 0.51). At the top, the highviscosity bulk-fill composite (Tetric-N) showed significantly higher KHN values for an increment thickness of 4 mm in comparison to 2 mm (p = 0.01). At the bottom and increment thickness of 4 mm, the other high-viscosity bulk-fill composite (Filtek BF) presented a significantly lower surface microhardness values than the other composites (p = 0.01).
Concerning the results of microtensile bond strength test for occlusal restorations (Table 3), there was no significant difference for MPa values among all groups, regardless of the restorative material used (p = 0.15). Nevertheless, there was a statistically significant difference among groups regarding fracture pattern (p = 0.04). Adhesive-type fracture patterns were more prevalent in high-viscosity bulkfill (46.7%) and nanocomposite (47.4%) than in low-viscosity bulk-fill composite (20%). The dentin cohesive-type fracture pattern was more frequent in the tooth restored with low-viscosity bulk-fill composite (50%).
The results of fracture load (Table 4) presented the following p-values: p(treatment) = 0.88; p(cycling) = 0.81; and p(treatment vs. cycling) = 0.34. There was no significant difference between treatments or thermomechanical cycling (with and without) for fracture strength. Fisher’s exact test showed that the distribution of the fracture pattern varied according to the treatment (p = 0.007), and these results are presented in Table 5. All fractures were coronary, with the majority being coronary until the middle third for the following groups: intact tooth (with and without cycling); conventional microhybrid composite (Z250) with cycling; and high-viscosity bulk-fill composite (Filtek BF) without cycling. For the conventional microhybrid composite (Z250) without cycling, 70% of the teeth presented a coronary fracture until the middle third, and the remainder had a root fracture up to the cervical. In the treatments with low-viscosity bulk-fill composite (SDR) or high-viscosity bulk-fill composite (Filtek BF) with cycling, the root fractures were observed in the middle and apical thirds.
DISCUSSION
The present study showed there was no significant difference between the top and bottom in the surface microhardness of bulk-fill composites with different viscosities, so the results fail to reject the first null hypothesis. Bulk-fill composites have higher translucency than conventional resin composites12, and the translucency of resins depends on the factors of increment thickness, dispersion/absorption coefficients of material, pigments, and opacifiers21-23. Moreover, increase in the cure depth of a bulk-fill resin can relate not only to higher translucency compared to a conventional resin but also to modified monomers, incorporation of stress relievers, or photoinitiator systems included in its composition21, especially because the decrease in polymerization shrinkage is manufacturer-dependent and may be associated with different attenuation mechanisms7.
The viscosity of material is influenced by monomer and filler content associating with the reaction kinetics and final polymerization. Modifications of the monomer and filler components make bulkfill resins more translucent/transparent by adding so-called polymerization modulators or initiation boosters14, 24. Considering the low-viscosity bulkfill composites studied, SDR (Dentsply) presented favorable and constant results regarding top or bottom microhardness and fracture strength of restored molars. SDR presents a uniform degree of conversion at a depth of 1-4 mm and a low proportion of internal gaps in dental restoration25. According to the manufacturer, an adequate degree of conversion and reduction of shrinkage stress is related to chemically modified polymer formation, which is flexible with the homogeneous network. Furthermore, SDR is composed of a modulator chemically incorporated to UDMA that could interact synergistically with camphorquinone, culminating in adequate polymerization21.
Considering the KHN results, Tetric-N was the high-viscosity bulk-fill composite that presented a significantly higher top KHN value at 4 mm than at the increment thickness of 2 mm. Tetric-N has a photoinitiator described as a polymerization booster (Ivocerin®) which, associated with the camphorquinone/amine initiator system, can polymerize the material in depth21. Ivocerin® is a dibenzoyl germanium derivative system with the highest absorption of wavelengths around 370 to 460 nm26. At the bottom, with increment thickness of 4 mm, the other high-viscosity bulk-fill composite studied (Filtek Bulk Fill) showed significantly lower surface microhardness compared to the other bulkfill composites. This can be explained by the absence of TEGDMA in the composition of this material, which has approximately half the molecular weight of the other monomers27. The microhardness variable has a high correlation with the filler content of material, and the lower microhardness values at the bottom surface of the Filtek Bulk Fill could occur due to light attenuation28. Nevertheless, no differences were found between increments of 2 mm or 4 mm, indicating an adequate depth of cure.
Consistently with the KHN results, the bond strength results (MPa) showed that there were no differences between the conventional nanocomposite, high- and low-viscosity bulk-fill composites, so the results fail to reject the second null hypothesis. In posterior teeth cavities, bond strength is expected to be equal to or lower than the dental cohesive strength, because in case of failures after the adhesion process, it is preferable that these failures should occur in the material, protecting the remaining dental structure. Nevertheless, sufficient bond strength is necessary in order to resist the mechanical and chemical challenges in the oral environment. In the present study, the bulk-fill composites did not differ from conventional nanocomposite for MPa.
This result is relevant because the conventional nanocomposite has been extensively studied and its performance is considered satisfactory29. Adhesivetype failures were more usual in the high-viscosity composites (bulk-fill or conventional) that are nonflowable and sculptable, while cohesive-type failures occurred more frequently in low-viscosity bulk-fill composite. Flowable resin composites generally have lower filler loading and are more fluid30, promoting adequate adaptation in the pulpal floor and decreasing internal irregularities of the preparation. Other factors could contribute to the fracture pattern results, such as the elastic modulus of high-viscosity composites, which is higher than in the low-viscosity composites31 and consequently promotes lower capacity for flow and adaptation on the deeper walls. Furthermore, bond strength was measured in the pulpal floor in deep occlusal cavities. This area is challenging for adhesive procedures due to the humidity, permeability, and characteristics of the intertubular dentin32, which may accentuate premature loss of adhesion.
Considering the fracture strength in molars with MOD restorations, the results fail to reject the third null hypothesis because no difference was observed in the fracture strength values of cavities restored with a conventional resin composite or bulk-fill composites of different viscosities, including the comparison with the intact tooth. This result is compatible with a previous study33 in which teeth treated endodontically with conventional composite resins or bulk-fill composites were tested for fracture strength, and there was also no difference between these materials, even when compared to the intact teeth. A high elastic modulus can inhibit the ability to deform, generating greater stress in the dental structures34. The adequate elastic modulus of bulkfill resins to substitute dentin or enamel20 allows the material to deform and absorb the stress generated during the thermomechanical cycles, similarly to the microhybrid resin used as a control. However, further studies should evaluate the behavior of teeth fully restored with low-viscosity composites, since in the present study, a conventional composite surface layer was used, considering the high occlusal load to which MOD cavities are submitted.
In a clinical situation, the size, type, and location of the cavity should guide the choice of material; therefore, the present study investigated the behavior of these materials in cavities/restorations of different configurations. The mechanical properties of the different materials vary considerably. Thus, the low-viscosity composites seem appropriate for liner, deep cavities and restorations after endodontic treatments, since the low viscosity facilitates adaptation in less accessible spaces30. On the other hand, high-viscosity composites are materials with more filler content and could be used in cavities considering their resistance to fracture or wear11. The fracture strength and the bond strength presented by both viscosities of bulk-fill composites are similar. The fracture strength of molars restored with these composites is equivalent to that of the tooth structure and the bond strength is comparable to that of a conventional composite. However, these resin composites require care during the insertion step, especially in the deep walls, in order to reduce adhesive or adaptation failures. Under controlled situations, as employed in the present study, the behavior of the different conventional or bulk-fill resin composites, regardless of viscosity, was similar, in agreement with previous investigations35-37 that reported clinical performance of bulk-fill resin composites similar to that of conventional resin composites. Further in vitro studies and clinical trials of bulk-fill composites remain necessary to continue their validation.
In general, according to the results of experiments, relevant findings were: the viscosity of bulk-fill resin composites included in this study did not influence the microhardness of top and bottom, regardless of increment thickness (2 or 4 mm); the dentin bond strength of bulk-fill resin composites, regardless of viscosities inserted as a single increment, was similar to conventional nanocomposite incrementally inserted in deep occlusal cavities, although adhesive failures were less frequent in low-viscosity bulkfill composites compared to other materials.
Moreover, the fracture strength of molars with MOD cavities restored with bulk-fill composites, regardless of viscosity, was similar to intact tooth and conventional microhybrid resin restorations, even after thermomechanical cycling.