In this study, the internal void formation of two single-shade bulk-fill flowable composite resin and a high-strength restorative composite at different thickness levels was investigated by 3D micro-CT analysis. Thickness did not have a significant effect on total porosity percentages (p = 0.994) so there were no significant differences within the same composite resin groups of different thicknesses (Table 3). Thus, the first null hypothesis was accepted. Material was not found as statistically effective on total porosity percentages (p = 0.509) so there were no significant differences within the same thickness groups of the tested composite resins (Table 3). As a result, the second hypothesis was also accepted.

Micro-CT allows a 3D assessment of the material structure without an invasive procedure on the material sample. Thus, it is often used for porosity analysis of composite resin restorations in the literature [14, 15, 22]. The manual modelling necessity of composite resins [16] and also air bubbles due to the application process may cause voids within the materials. Closed voids may disrupt the mechanical integrity of the materials, leading to crack formation, followed by bulk failure [14]. Clinical indications recommended by the manufacturers are similar for the tested materials in this study. Unlike the other tested composite resins, class IV cavities are not listed among the clinical indications of CH. A study evaluated the porosity of different composite resins in their compule and restoration forms, and it was stated that the amount of porosity (vol%) values decreased in restoration forms of the composite resins [17]. Another previous study evaluated the void formation of the sonic delivery method with various composite resins. It was reported that sonic delivery might increase porosity depending on the composite resin type [14]. These results may suggest that the proper application and handling can reduce the risk of porosity in the final form of the materials.

Although material, thickness, and material*thickness interaction were not significantly effective on total porosity (Table 3), all of these main effects significantly affected the number of total voids (Table 4). Regardless of thickness, total porosity, and the number of total voids from the lowest to the highest were as follows: INJ < OM < CH (Tables 3 and 4). The lower results of INJ may be due to its 2 mm-increment placement, and it is also not a bulk-fill composite resin, unlike the other tested materials. In addition, it was claimed that INJ has a highly thixotropic viscosity property. Thus, this thixotropic structure allows the material not to slump during placement, but to flow when moved around with an instrument [10]. These qualities may be advantageous for decreasing the risk of air bubble occurrence as the material is extruded from the syringe’s tip. A micro-CT analysis study investigated 2.5 mm-thickness fifteen restorative materials, including INJ [12]. It was reported that INJ demonstrated a lower porosity percentage than the flowable bulk-fill composite resins (SDR flow + and Filtek Bulk Fill Flowable), but there were no statistical differences between these materials (p > 0.05). Total porosity values of INJ (0.076 ± 0.018) were similar to our results of total porosity of INJ (0.07 (0.026–0.682)) (Table 3).

In clinical practice, commonly treated class II restorations require restorative material of 4–5 mm thickness [18]. Even though the composite resins have acceptable mechanical properties, internal voids of the restorations may jeopardise the materials’ performance under loading [13]. The number of total voids, regardless of composite resins, in the total thickness groups was observed from the highest to the lowest as follows: 6 mm > 4 mm > 2 mm (Table 4). This result may be due to the total material volume increase. Incremental layering of thin composite resins reduces air entrapment. While flowable bulk-fill composite resins still carry an air entrapment risk during injection, they eliminate the need for multiple layers. This prevents the voids that typically form when adapting and manipulating traditional composite resins with hand instruments.

Although internal voids are considered defects, an acceptable amount of porosity or number of voids for the clinical performance of the materials has not yet been reported. The total porosity of the tested materials in the current study were observed as 0.026–0.115% (Table 3) and the number of large void percentages ranged in 3.764–10.111% (Table 6). These results highlight that the overwhelming majority of the voids were small, which may reduce the risk of critical failures. Voids may act as initiation points or stress concentrators for crack and fracture propagation. Higher amounts of voids can also increase water sorption and subsequent discoloration [22].

There are various monomers in the organic matrix of the resin composites in the market, such as Bis-GMA, UDMA, TEGDMA, and Bis-EMA. The organic matrix composition can influence the composite resins’ properties regarding polymerization and, as a result, their mechanical properties [2]. A previous study reported that UDMA groups demonstrated higher depth of cure than the Bis-EMA groups [1]. All the tested materials in this study contain UDMA. INJ had the least UDMA ratio (0.5– < 1%) in its matrix, followed by CH (≥ 10– < 20%) and OM (10–30%). INJ had the most varied monomer composition (Bis-EMA;10– < 25%, TCD-DMA; 5– < 10%, NPGDMA; 25– < 5%, and melamine formaldehyde resin; 1– < 25%), which may provide the material with diverse properties. Even though the resin matrices of CH and INJ present differences, Bis-EMA ratios of the materials (≥ 10–25% and 10– < 25%, respectively) are similar. Bis-EMA is the ethoxylated analog of Bis-GMA. Bis-EMA differs in that it lacks two hydroxyl groups [2]. TEGDMA was reported to present higher polymerization shrinkage [23]. There are different ratios of TEGDMA in OM and CH, which may be attributed to their higher number of total voids than INJ (Table 4).

Filler particle properties may influence the microstructure of resin composites. It was argued that resin composites with a hybrid filler composition may reduce the risk of air entrapment, thereby improving the packing density of the material's inorganic structure [3]. On the other hand, OM, a nanofill composite resin with homogeneous spherical fillers (Fig. 5), performed better than CH in terms of total porosity and the number of voids. However, OM and CH are both single-shade materials; OM uniquely does not contain any dyes or pigments and achieves its single-shade property through structural color technology [7], whereas CH employs adaptive light matching technology. Further, CH, a nanohybrid resin composite, has irregularly shaped fillers (Fig. 5) with a filler size range of 0.02 μm to 5 μm, which may have affected its internal void formation. The nanofill INJ exhibited lower total porosity and number of total voids than nanohybrid CH (Tables 3 and 4). This may be due to the Full-Coverage Silane Coating (FSC) technology of INJ. FSC technology provides an almost fully coated filler surface with the silane coupling agent. INJ was reported to exhibit the lowest water sorption and the highest flexural strength compared to another injectable restorative and traditional flowable composite resin [24, 25]. The material, when injected, keeps its shape without slumping, providing improved thixotropy and reduced extrusion pressure [10]. The smaller filler size (0.15 µm) (irregularly shaped, Fig. 5) of INJ may contribute to its better results in total porosity (Table 3) and number of total voids (Table 4) than OM (0.2–0.4 µm).

In this study, the polymerization protocols for sample preparation including light intensity and irradiation time adhered strictly to each manufacturers’ instructions. Although variations in polymerization protocols could be viewed as a limitation, following these guidelines ensures the study mimics clinical reality. Particularly, composite resins such as single-shade bulk-fill and high-strength restorative require unique irradiation parameters tailored to their specific layer thicknesses.

Micro-CT analyses for void formation also have some limitations, such as long scanning times, complex data processing, and limited device capacity for multiple-sample analysis compared to 2D digital radiography [26]. As a limitation of the study, the long-term effects of the oral environment, such as water sorption and thermal changes, were not simulated. These factors may impact the formation of voids and also the clinical consequences of materials’ porosity. The small sample size is another limitation of the study. Closed voids were investigated in the present study, it can be suggested that further studies may focus on the marginal adaptation of these materials in various cavity designs. In addition, employing mechanical analyses such as degree of conversion and microhardness would enrich the study and provide a more plausible interpretation of the clinical implications of porosity.

Within the limitations of the study, the following conclusions were drawn:

There were no significant differences between the test groups in terms of total porosity percentages.

The total number of voids increased as the thickness increased for the bulk-fill composite resins.

Overall number of total voids of the composite resins from the lowest to the highest were as follows: INJ < OM < CH.