Shrinkage stress and elastic modulus assessment of bulk-fill composites

Abstract Bulk-fill composites were introduced in dentistry to accelerate clinical procedures while providing adequate outcomes. Concerns regarding the use of bigger composite increments rely on the polymerization shrinkage and shrinkage stress, which may generate gaps on the adhesive interface and result in a reduced success rate. Objective: To evaluate the polymerization shrinkage stress of different bulk-fill resin composites and their elastic modulus. Materials and Methods: Fourteen specimens were made for each of the nine different resin composites (seven with 12 mm3 and seven with 24 mm3): Surefill SDR flow (SDR), X-tra Base (XB), Filtek Bulk Fill Flowable (FBF), Filtek Z350XT Flow (Z3F); Tetric Evo Ceram Bulk Fill (TBF), X-tra Fil (XF), Filtek Bulk Fill (FBP), Admira Xtra Fusion (ADM) and Filtek Z350 XT (Z3XT). Linear shrinkage stress was evaluated for 300 s with the aid of a linear shrinkage device adapted to a Universal Testing Machine. For each composite group, seven additional specimens (2x2x25 mm) were made and Young's modulus was evaluated with a 3-point bending device adapted in a Universal Testing Machine with 0.5 mm/min crosshead speed and 50 KgF loading cell. Results: For 12 mm3 specimens, three-way ANOVA showed that only SDR and TBF generated lower stress after 20 s. Considering 300 s, TBF, SDR, and XF generated the lowest stress, followed by ADM, FBP, XB, and FBF, which were similar to Z3XT. Z3F generated the highest stress values for all time points. Considering 24 mm3 specimens after 20 s, all bulk fill composites generated lower stress than Z3XT, except XB. After 300 s, SDR, FBP, and ADM generated the lowest stress, followed by TBF and XF. For elastic modulus, one-way ANOVA showed that FBF, SDR, Z3F, and ADM presented the lowest values, followed by XB and TBF. FBP, Z3XT, and XF presented the highest elastic modulus among the evaluated composites. Conclusions: Bulk-fill resin composites presented equal to lower shrinkage stress generation when compared to conventional composites, especially when bigger increments were evaluated. Bulk-fill composites showed a wide range of elastic modulus values, but usually similar to “regular” composites.


Introduction
Despite advances in adhesive dentistry, resin composites still tend to fail in extensive posterior restorations due to wear, medium to long term adhesive interface deterioration, technical sensitivity, polymerization shrinkage and inadequate polymerization, particularly in class II restorations with cervical margins located in dentin or cementum. [1][2][3][4][5] Defects on the adhesive interface are generated by the characteristics of resinous materials during the polymerization process. Composites generate shrinkage (polymerization shrinkage) that depends on the material composition and volume. [5][6][7][8][9][10] Shrinkage can generate stresses that may lead to the formation of micro gaps and, thus resulting in microleakage of saliva and bacteria, adhesive interface degradation, secondary caries, pulpal changes, and consequently, clinical failure of the restoration. 4,5,11 T h e i n c r e m e n t a l i n s e r t i o n t e c h n i q u e i s recommended to ensure a better marginal integrity because it reduces the development of polymerization shrinkage stress. [12][13][14] However, despite the advantages of the incremental technique in ensuring a better polymerization and stress distribution, this technique is more laborious, technically sensible and timeconsuming. 8,10,15 Bulk-fill resin composites are advised to be used in larger increments without compromising the degree of conversion (up to 4 mm according to some manufacturers). Concerns with the polymerization of large increments relies on the polymerization shrinkage and on the stresses generated in the tooth/ restoration interface. 10,[16][17][18] Promising results have been reported with these materials, mainly due to lower polymerization shrinkage, 5,18,19 which also depends on the composite organic/inorganic matrix composition and properties such as viscosity and elastic modulus.
Although several materials with different viscosities and handling characteristics are commonly classified as bulk-fill resin composites, their properties can change considerably, especially due to modifications in the organic matrix, with the incorporation of monomers with higher molecular weight, as well as changes in filler content and incorporation of stress relievers. 5,10,16,18,[20][21][22][23] Composites can be subdivided according to their consistency in low-and high-viscosity. Higher shrinkage stress for flowable composites are expected since they generally have a higher organic content when compared to microhybrid and nanoparticulate composites, which can result in greater polymerization shrinkage and lower mechanical properties. 22

Materials and methods
This study evaluated nine different resin composites The specimens were removed from the matrix and stored in distilled water at 37°C for 24 hours, in absence of light. Following, any excesses were removed

Statistical analyses
For all statistical analyses, 5% was adopted as the significance level (p<0.05). All data were evaluated for homogeneity through the Shapiro-Wilk test. For polymerization shrinkage stress, three-way ANOVA was used (time, composites and volume). For Young's modulus assessment, one-way ANOVA was used. All parametric tests mentioned above were followed by Tukey's test.
In addition, a linear regression analysis was performed considering Young's modulus and filler content, as well as considering Young's modulus and shrinkage stress.    (Figures 4 and 5), and no correlation was observed for any of the composite groups.

Discussion
Polymerization stress generated by the inherent shrinkage of composites during light curing has been the subject of several researches for a long time, 5,28 since stress values that exceed adhesive resistance can lead to the formation of gaps in the interface. 5,29,30 Therefore, the ideal composite should generate the lowest shrinkage stress possible while ensuring a better seal. 31 To allow the insertion of larger increments, the molecular basis of bulk-fill composites was modified by the incorporation of stress relievers and monomers with higher molecular weight (low molecular weight monomers promote a higher number of double bonds per unit of weight, allowing a higher degree of conversion, but also leading to higher shrinkage and shrinkage stress). 5,16,18,[20][21][22][23] One may question the organic and inorganic matrixes of these composites since "conventional" and bulk-fill composites sometimes share similar composition. Nevertheless, manufacturers usually do not report the proportion between the different monomers, neither the filler content or their proprietary formulations. 16,20,22,28 Similarly, differences in filler content (e.g., when comparing high-and low-viscosity composites) may be critical in volumetric shrinkage (higher stress due to a higher amount of organic content and lower filler content is expected in low-viscosity composites).
Nevertheless, a lower Young's modulus may allow stress dissipation during the polymerization process, thus reducing the stress in bigger increments. 10,25,26,32,33 Considering high-viscosity composites with 12 mm 3 of material after 300 s, TBF and XF generated lower stress values when compared to the control group (Z3XT). The other bulk-fill composites presented values similar to Z3XT, but also similar to TBF and XF (Table 1). For low-viscosity/flowable composites, SDR generated the lowest stress values, followed by FBF and XB. The low-viscosity control group (Z3F) generated the highest shrinkage stress.
In general, high-viscosity bulk-fill composites generated lower shrinkage than low-viscosity bulk-fill composites as stated by other authors. 5,8,32 The only exception was SDR, which generated similar stress when compared to high-viscosity bulk-fill composites despite being flowable. Such results can be explained by the presence of a modified UDMA (monomer with high molecular weight -849 g/mol) which was stated to reduce shrinkage and, consequently, shrinkage stress. 34 It is interesting to note that all bulk-fill composites (high-and low-viscosity) generated similar or lower stress values when compared to the high-viscosity control (Z3XT).
Given that stress depends on the composite volume, 29 testing how the volume changes the impacts caused on shrinkage stress is important. 35 Increased volume (24 mm 3 ) resulted in increased stress for the evaluated composites. All bulk-fill composites with 24 mm 3 generated lower or similar (XB) shrinkage stress when compared to Z3XT after 300 s (Table 1). SDR, FBP and ADM generated the lowest stress while Z3F generated the highest stress among all tested composites.
In addition, after 300 s, SDR, FBP and ADM with 24 mm 3 , showed values similar to Z3XT with 12 mm 3 , and FBP and SDR generated similar values for both 12 and 24 mm 3 (Table 1). Such results demonstrate a great capability of bulk-fill composites in dealing with the generation of shrinkage stress, even in big increments, as previously reported. 22,36 FBP relies on monomers with higher molecular weight (AUDMA, UDMA and 1, 12-dodecane-DMA), associated with a relatively higher filler content (76.5%) when compared to low-viscosity composites, to reduce polymerization shrinkage. The effect of monomers with higher molecular weight can be also observed in FBF, which substituted TEGDMA (286 g/ mol) for UDMA and, despite presenting the same filler content as Z3F (~65% in weight), generated lower  Considering shrinkage stress development, a rapid increase during the first 10 s of light curing (20 s total) can be observed, followed by a slower increase until the LED light is turned off. The fast subsequent cooling of the composite might be responsible for a second shrinkage peak, as reported by other authors. 22,34,39 Shrinkage stress development seems to be slower in bulk-fill composites when compared to conventional resins. This can be observed in Figure 3, in which stress development in bulk-fill composites took longer when compared to their regular counterparts. This is especially true when comparing bulk-fill composites within the same viscosity classification (i.e., high-viscosity bulk-fill composites with lower elastic modulus: ADM and TBF, showed slower stress development). This is important because a slower stress generation allows a better stress distribution and may contribute to the bonding integrity, since the material has more time to accommodate the shrinkage stress before the elastic modulus (composite stiffness) starts to increase. 40 In addition, the stress curve is flat for all composites after 200 s, showing that most of the shrinkage develops during the initial minutes. This explains the current option for assessing shrinkage stress up to 5 minutes instead of several hours, as also observed by other authors. 7,32,39 The authors of this study performed correlation tests between the elastic modulus and filler content.
No correlation was observed for high-viscosity composites, but a strong correlation was observed for low-viscosity resins as reported by other authors. 39,40 The low correlation between high-viscosity composites may have occurred because ADM and TBF present a relatively lower elastic modulus when compared to their filler content, as previously discussed.
In addition, no correlation was observed between shrinkage stress and elastic modulus for any of the composite groups (Figures 4 and 5). These results corroborate other authors. 34  the high volumetric shrinkage of Z3XT and the use of monomers with higher molecular weight in FBP. Such results demonstrate the fundamental role of volumetric shrinkage on the generation of shrinkage stress. 34,40 This statement supports the results of this study, since all flowable composites (with lower filler content) are expected to present higher shrinkage and generate higher shrinkage stress. 32 The SDR group is an outlier as already discussed and as previously reported. 34 Although having benefits that may reflect in easier and faster cavity restorations, bulk-fill composites still require further studies to assess the influence of their properties on the long-term maintenance of internal and marginal adaptation. Assessing the interaction between bulk-fill composites and tooth structure regarding adaptation, cusp deflection, among other factors, will also be important.
These results show, in general, a better behavior for bulk-fill composites regarding the generation of shrinkage stresses, mainly when larger increments are used. Nevertheless, it is important to note that despite being classified as bulk-fill resin composites, the different tested materials can show very different behavior, not only regarding the different classifications (low-and high-viscosity) as would be expected.
Further tests are advised to clarify the best indication for each composite to clinicians. In addition, bulk-fill composites and regular composites also showed very different properties as previously discussed and, thus, the initial hypothesis was accepted.

Conclusion
Considering the limitations of this study, it was possible to conclude that bulk-fill composites present very heterogeneous behavior, which is related to their composition (monomers and filler content).
In addition, it can be concluded that: Bulk-fill resin composites present equal to lower shrinkage stress generation when compared to conventional composites, mainly with bigger increments.
Bulk-fill composites show a wide range of elastic young's modulus values, but usually similar to "regular" composites.
Volumetric shrinkage seems to be more important than elastic modulus for polymerization stress development.