Figure 4B shows SEM images of each group at 50× and 100× magnifications. Visual inspections and SEM observations revealed thinner margin regions of the crown with printing interruptions. Adding particles of different sizes and concentrations resulted on differing effects on defects on the material surface. As the concentration increased, the extent of damage became progressively more severe. Figure 4 shows that the surface structure remained relatively intact in the control group, with clear layered structures. The surface structure exhibited an increase in the number and size of holes in the 1%S group, which became more pronounced in the 2%S group, with the layered structure being extensively disrupted. By comparison, groups with larger particles showed major damage to the layered structure, with defects larger than those observed in the 1%S and 2%S groups. In the 2%L group, the walls surrounding the defect had fractured and disappeared.
Silica is also known for its high biocompatibility, supporting its use in biomedical and dental application. Evaluating the effects of particle diameter and weight% on silica-incorporated resins is essential to achieving superior performance and long-term use in dental restorations. The aim of this study was to systematically determine how the size of silica particles affects resin properties in order to meet diverse application requirements. The findings indicated partial rejection of the null hypothesis. Particle size produced significant differences in most properties - larger silica yielded higher flexural strength and hardness but rougher surfaces, while nano-sized silica improved degree of conversion and smoother surfaces with a pronounced whitening effect - yet printing trueness showed no significant size-dependent difference (accuracy was similarly reduced for both nano and micro fillers at 2% wt). This indicates that although silica size clearly influences mechanical and surface outcomes, it does not appreciably affect printing accuracy, warranting a partial rather than full rejection of the null hypothesis.
This study compared the effects of adding small particles (5-20 nm) and large particles (0.5-10 μm) at concentrations of 1 wt% and 2 wt%. It has been reported that ground quartz micropowder (10-50 μm) and ground glass micropowder (0.6-10 μm) from crystallized silica significantly improved the strength of resins. Lankoff et al. found that the size of silica particles directly affected the filler loading capacity and test performance by comparing 12-nm and 40-nm particles. Such small spheres, which differ from those with micron dimensions, can act as a lubricant to allow the material to flow freely and exert a small effect on viscosity. Two types of particles were selected with sizes differed by at least 25-fold to cover both nanoscale and microscale ranges, ensuring that significant results with broad applicability. Silica concentration is another key determinant of resin performance. Previous studies have shown that silica particles can be incorporated with a uniform distribution in a resin matrix at concentrations of up to 4 wt%, whereas aggregation tends to occur above 6 wt%. Notably, at concentrations of ≤4 wt%, such particles are able to remain dispersed and exhibit stable properties. Gangil et al. reported a 15-30% reduction in tensile strength when silica content was increased from 2 wt% to ≥ 3 wt% in hemp-sisal epoxy composites. Importantly, our results and earlier studies show that even at this ~ 2 wt% loading, silica addition already had either a strengthening or deteriorating effect on the resin system (Table 1). Considering the need for a uniform distribution and that silica is predominantly used in filler systems with two or more composite formulations, this study set the maximum concentration at 2 wt%. Although industrial composite systems can achieve silica loadings of ~ 30 wt % or even higher by using engineer under high shear mixing to suppress aggregation and control rheology, our study at 1-2 wt % establishes a clear and indispensable baseline for isolating intrinsic size effects and guiding the design of future hybrid or high silica loading formulations. To overcome the narrow range limitation and fully characterize the sensitive results, future work will explore higher or intermediate levels with broader concentration gradients (0.5, 1.5, 3.0, and even 4.0 wt% for one size) even hybrid silica to identify the optimal balance between flexural strength, degree of conversion, surface roughness, and trueness.
This study found that when considering silica as a coloring agent, there was a slight difference in the white colors of the two powders, and that the whitening effect increased progressively in increments of 1.0% (P < 0.001). The silica particles appeared as a loose, dry, pure-white powder, with the larger microparticles showing a visible decrease in brightness to the naked eye as well as having a slightly gritty texture (Fig. 1A). These observations are consistent with a previous study finding that adding 10% glass silica to a resin resulted in a noticeable whitening effect. Additionally, NextDent C&B is a commercial resin whose color is similar to that of teeth, which would surely be changed by the addition of white-powder fillers (Fig. 1F). Alshamrani et al. provided a representative overview of the findings for their test bars, describing that the bar with 20% white-powder glass-silica filler appeared brighter in color, although no quantitative color measurements were made. Furthermore, Lim et al. emphasized that the quantitative lightness was strongly correlated with that of resin composites containing 10-70 wt% glass filler. The diameters of the two glass fillers in the study of Lim et al. were similar (0.77 μm and 0.50 μm), which may explain why they found no correlation between color lightness and particle diameter. This contrasts with our study investigating particles with a 25-fold difference in the diameter ranges (5-20 nm vs. 0.5-10 μm), which may change the color due to its effects on light scattering, facilitating the ability to characterize any correlation between color lightness and particle diameter. Clinically, ΔE*ab value above approximately 3.3 is considered the 50:50% acceptability threshold for dental restoration. The 2%S group (ΔE = 27.15 ± 0.60) exceeded acceptable limits and thus raised concern about a perceptible whitening effect. The color differences between specimens may also vary with the heat-treatment methods, oxygen inhibition, and the colorants. Moreover, higher filler content may cause surface deterioration during the postprocessing and particles precipitation during storage. The pronounced whitening effect may be clinically beneficial in cases such as anterior temporization or long-span provisional where increased brightness is preferred. Considering of these factors above, careful adjustment of filler size and concentration along with the use of pigment or opaquer when necessary is recommended to mitigate excessive whitening and ensure aesthetic harmony in dental restorations.
The surface roughness (which was quantified as the Ra in this study) of resin-based composites significantly effects their aesthetic and biological properties, contributing to discoloration, increased wear, biofilm accumulation, and cytotoxicity. Since manually polished composites exhibit greater instability than unpolished surfaces, the original surfaces were retained for the Ra measurements. The Ra values in the experimental groups demonstrated a distinct smoothing effect after the addition of nanoparticles and a roughening effect after adding microparticles, both of which were strongly correlated with the particle concentration (Fig. 1G). Similarly, Rodríguez et al. compared the roughness between two types of silica particles (silica-zirconion-dioxide nanoclusters and silica nanoclusters), and concluded that the surface was smoother for the smaller particles. The surface area and pore structure of mesoporous silica allow it to be uniformly dispersed in a resin matrix, filling in microscopic defects and thus reducing the surface roughness. This beneficial effect was even more obvious in our 2%S group, in which the Ra reduced to 55.60 ± 7.91 nm (Table 1). Ra values exceeding 0.2 μm has been reported to increase biological concerns regarding plaque deposition. In the present study, the highest Ra observed for 2 wt% silica microparticles remained below 0.2 μm (Fig. 1G; Table 1), thereby minimizing these risks. Thus, maintaining the microparticles at a concentration at or below 2% is recommended for controlling roughness and for reducing oral hygiene risks when using resin-based restorations.
Figure 2 demonstrates that no distinct fluorescent spots indicating the agglomeration of silica nanoparticles were found in the SEM images or EDS mapping. The silicon signals observed in the control group's EDS mapping can be attributed to instrumental artifacts, including scattered X-rays from silicon-based components within the SEM chamber (such as detector windows and sample stages) and inherent background noise from the X-ray spectrometer system itself, as silicon was confirmed to be absent from both the resin formulation and post-processing procedures. The control group appeared relatively clean, whereas the surfaces of the groups with silica nanoparticles had a granular appearance. In contrast, the microparticle groups exhibited pit-like voids or even a wavy appearance. There is usually fairly good consistency between Ra values and SEM findings, as found by Aytac et al.. The Ra value of the discs in our 2%L group was 114.73 ± 8.19 nm, which is consistent with the gritty, wavy surface evident in the SEM images. The finishing and postprocessing steps would only remove residual monomer and unintegrated silica, resulting in other particles either remaining embedded on the surface or detaching from it. The smaller nanoparticles -- with their larger surface area -- are more likely to remain embedded, while the larger microparticles tend to detach more easily from the surface. Similar exposed filler particles and voids resulting from filler loss were observed by Aytac et al. and Aljabo et al.. SEM images obtained in other studies revealed silica-filler-induced voids at fracture sites deep inside the resin. Although our study did not observe the degradation of mechanical properties due to internal voids, these observations suggest that further evaluations of factors such as abrasion and fractures are necessary to evaluate the distribution of internal fillers.
The presence of nanoparticles demonstrated a distinct enhancement effect as well as a concentration-dependent improvement in the DC (Fig. 3A). It could be inferred that smaller nanoparticles cause increased light scattering, which leads to a higher DC. The ultimate DC in all groups in the present study typically fell within the range from 50 to 80%, which is consistent with reported values for most commercial dental resins. Therefore, future studies should investigate the effects of higher concentrations of particles with clinically relevant standard deviations in order to further elucidate the effects on DC.
The mechanical properties of a resin crucially affect the long-term performance of a dental prosthesis. Adequate flexural strength is essential for ensuring that temporary crowns and bridges can withstand chewing forces before the final restoration is fabricated. According to ISO 4049:2019, Type 2 polymer-based restorative materials must exhibit a minimum flexural strength of 50 MPa. All tested groups in the present study comfortably exceed this requirement, with flexural strength values not falling below 116.73 MPa (Table 1). This study found that the flexural strength was significantly lower in the 2%S group than in the control group, widening the gaps with the group with larger silica particles at the same concentration and with the baseline resin (Fig. 3B). Conversely, Khan et al. observed improved flexural strength in nanofillers with a particle diameter of 20 nm. The larger surface area of nanoparticles will increase their interactions with polymer chains. This allows more polymer segments to be adsorbed on the particle surface but also has the potential to cause aggregation at higher concentrations that will create areas of stress concentration that can induce crack initiation.
Previous studies have found that incorporating 3-µm microparticles increases the hardness of dental composites more effectively than when using 30-nm nanoparticles at 2.5% or 5.0%. Moreover, Rodríguez et al. concluded that the mechanical properties are influenced more by the filler concentration than by the particle size. By comprehensively investigating both of these factors, our study revealed a clear hardening effect with silica microparticles, which appeared greater when the concentration raised up to 2 wt% (Fig. 3C). This suggests that voids present in the resin are gradually filled by crystalline silica rather than the monomer, leading to an increase in hardness as the filler content increases. In the present study, mechanical properties were measured according to the size and concentration of silica particles, which were found to influence the agglomeration effect, crystalline substance resistance, and light scattering. The results support the importance of considering particle size to achieving the optimal trade-off between mechanical properties and filler characterization.
The surface geometry of a DLP provisional crown is limited by the layer-by-layer printing pattern, which appears step-like when the crown is placed horizontally (Figs. 4A and 6B). The lower trueness of the marginal areas across all groups as indicated in Fig. 4A non-green surrounding collar area, may attribute to inherent distortions from vertical printing pattern. Since crown margins are very thin, inaccuracies caused by the step-like morphology and layer discontinuities would be particularly significant, especially in geometry aligned with the projection axis. Additionally, the regions highlighted in red (Fig. 4A) indicate a localized clustering of elevated RMS values, which also correspond to concave and convex distortions induced by the layer-by-layer printing and orientation. However, the localized deviations have disproportionately influenced the overall values by group-level, contributing unevenly to RMS values. This study analyzed overall intaglio surface for better comparison. It is recommended to consider region-wise segmentation or direction-specific assessment in clinically contoured crown models to enable more precise evaluation of printing parameters. Besides, layer thickness is demonstrated to play a key role influencing trueness. Specifically, groups with a 30-µm layer thickness showed higher trueness than the 50-µm groups in several studies. Given that the design of the present study included the incorporation of silica particles, a regular layer thickness of 50 μm was selected to improve comparability between other 3D-printing studies. The obtained trueness values were relatively high (calibration cube, 33.52 ± 7.81 μm; control, 71.40 ± 8.97 μm); in contrast, Son et al. reported a trueness of only 29.5 ± 3.3 μm when using a 25-µm printing-layer thickness in their crown-investigation study. In clinical cases or other different scenarios where high precision of the crown margins is required, it is recommended that the printing layer thickness be reduced, or different parameters be used than for conventional crown manufacturing processes.
Previous research suggested that the trueness of the intaglio surface trueness should be between 50 and 100 μm for it to be used in fixed dental prostheses. This study measured RMS values of up to 108.94 μm in 2%S. The increase in RMS at 2% loading may compromise the marginal and internal fit of prosthetic restorations, potentially affecting clinical performance. But considering the comparison purpose, the methodology compromised via adoption of a 50 μm layer thickness and tailored changes in crown morphology, resulting in larger RMS values and greater distinction between the groups included. Prior studies showed that reducing layer thickness to 20-30 μm markedly improves trueness and margin quality, which would generally narrow the spread of RMS values between experimental groups. As shown in Fig. 4B, a thinner area was delineated to target and compare holes and walls at the same location when designing the crown STL file. In the control group, margins were intact, and the interlayer microstructure was clearly similar to the ideal side surface with a natural gradient. However, the same area exhibited scattered holes in the 1%S group, while these holes started to emerge in the 2%S group to form larger defects with pronounced disruption of the layered structure. In contrast, groups with larger added particles showed major damage to the printing continuity and thin wall retainability. Becker et al. suggested that the overall layer structure of nanocomposites comprises a blend of intercalated and exfoliated structures. It is generally believed that the phase boundary is the weakest point of layered materials, and where destruction in the form of delayering begins. In our study, interlayer disruption may have resulted from poor bonding of the silica particles to the resin matrix, which was also concentration dependent. Smaller particles intercalated into one or two 50‑µm layers producing minute air voids Fig. 4B. Larger particles are more prone to embedding and exfoliation within several layers (50 μm) to form rigid zones that created stress‑concentrated spots, triggering cracks and edge discontinuities. Therefore, silica particle size dictated the defect mechanisms while the 50 μm layer thickness emerged as a significant confounder, exaggerating interlayer damage. Using thinner layer below 50 μm would certainly improve RMS trueness and margin quality, but residual differences from particle size or concentration would likely shrink. Thus, it is recommended to carefully select layer thickness or other parameters like printing orientation, segmental accuracy analysis based on specific goals or clinical use, to balance trueness, marginal continuity and structural integrity.
This study had some Limitations. Firstly, only one commercial 3D-printing resin product was evaluated, which might be highly specific to the unique resin chemistry of NextDent C&B. In other formulation systems with Bis-GMA, UDMA-, TEGDMA or bis-acrylic matrices, silica incorporation can influence viscosity, polymerization shrinkage and mechanical behavior in ways that differ from those observed in NextDent C&B. Other resin systems need to be explored in future work. Secondly, more detailed gradients should be investigated, with appropriate increases in silica concentration, as well as attempts to create more complex formulations by varying the sizes of the silica particles. It has also been reported that a glass-silica-based system improved the resin properties. The long-term evaluations including of color stability, water sorption, residual monomer release and wear resistance are planned in future material studies. To fully characterize biosafety, sub chronic and chronic assessments beyond 72‑hour cytotoxicity tests are planned in follow-up research.