Effect of light-curing modes on degree of conversion and color stability of flowable bulk-fill composites: an in vitro experimental study
Article information
Abstract
Objectives
This study aims to evaluate the impact of light-curing modes on the degree of conversion (DC) and color change (CIEDE2000 color difference, ΔE00) of different flowable composite resins.
Methods
A total of 120 specimens (n = 10 per group) were prepared from four materials: G-aenial Universal Injectable (GUI; GC), Estelite Bulk Fill Flow (EBF; Tokuyama Dental), Tetric N-Flow Bulk Fill (TNF; Ivoclar Vivadent), and Filtek Bulk Fill Flowable (FBF; 3M ESPE). The specimens were polymerized using an LED light-curing unit under three light-curing modes: standard (1,000 mW/cm², 20 seconds), high-power (1,400 mW/cm², 3 × 4 seconds), and extra-power (3,200 mW/cm², 2 × 3 seconds). The DC was quantified using Fourier transform infrared spectroscopy. Color measurements were performed at baseline, and on the 7th and 14th days after immersion in coffee, using a spectrophotometer. The ΔE₀₀ was calculated using the CIEDE2000 formula. The data were analyzed using IBM SPSS software (α = 0.05).
Results
Material type and light-curing mode significantly affected both the DC and ΔE00 (p < 0.05). All composites exhibited lower DC in the extra-power mode compared to the high-power mode. On day 14, the extra-power mode resulted in significantly greater discoloration in TNF, FBF, and EBF compared to the high-power mode (p < 0.05).
Conclusions
Although curing with the extra-power mode saves time, it may result in a lower DC and an increase in ΔE00 in flowable composite resin.
INTRODUCTION
In recent years, advancements in polymer chemistry and technology have led to significant developments in restorative materials. Initially, the term “bulk-fill” referred to a resin composite application technique that allowed layering in 4–5 mm thicknesses; however, over time, it has been adopted as a new class of material [1]. The use of bulk-fill composites, which allow for the placement of thicker layers resulting in a homogeneous restoration, is advantageous because of their ease and speed of application. The use of additional or more efficient photoinitiators, combined with a reduction in filler content, enhances light transmission and deepens polymerization in bulk-fill composites [2,3].
The degree of conversion (DC) is a fundamental parameter of composite resins, significantly influencing their mechanical performance, polymerization shrinkage, and biocompatibility [4,5]. The DC is affected by various parameters, including the chemical composition of the dimethacrylate monomer, intrinsic elements like the type and quantity of photoinitiators, as well as extrinsic conditions such as the temperature during polymerization. Additionally, the increased viscosity of the polymerizing network and the restricted mobility of reactants due to the presence of filler particles impact the DC [4]. More translucent bulk-fill composite resins facilitate the conversion of monomers to polymers, thereby increasing the DC. Flowable bulk-fill composites, characterized by a low inorganic filler content and a high resin matrix ratio, exhibit higher monomer-to-polymer conversion rates [6]. Nanoparticle fillers have been integrated into new-generation flowable bulk-fill resin-based composites to improve their mechanical properties while preserving low viscosity [7]. However, the polymerization efficiency of these materials must be tested using direct or indirect methods to ensure clinical applicability [8].
Adequate polymerization requires the light source to operate within an appropriate wavelength range, provide sufficient light output, and be used for an effective light-curing time [9]. Recent experimental studies [9–12] have shown that curing kinetics and depth of cure are strongly dependent on light-curing unit (LCU) type, spectral match, and exposure protocol. Although Ajaj et al. [8] and Parra Gatica et al. [11] reported valuable insights into bulk-fill polymerization efficiency and flowable bulk-fill characteristics, they emphasized substantial heterogeneity in light-curing protocols, material compositions, and evaluation methods. In many studies, higher irradiance has not generally been associated with higher DC because different irradiance intensities combined with varying light-curing times can lead to similar irradiance exposures [11,12]. Although in vitro studies argue that sensitivity to changes in irradiance depends on the material and that calculations based solely on total energy are invalid, LCUs continue to increase irradiance intensity. Moreover, manufacturers persistently claim that sufficient polymerization can be achieved with short exposure times (5 seconds or less) at high irradiance levels [6].
The release of residual monomers into the oral cavity and pulp chamber, caused by incomplete polymerization, may compromise the physical integrity and long-term stability of dental materials [12]. Czasch and Ilie [13] reported that this condition leads to problems such as marginal leakage, discoloration, water absorption, and reduced mechanical strength in restorations. Color stability in resin composites is influenced by multiple compositional aspects, encompassing the resin matrix’s chemical characteristics, the filler type and load, as well as photoinitiator content. Although restorations are influenced by external factors according to their inorganic filler and surface characteristics, intrinsic discoloration may occur due to oxidation-induced changes in the resin matrix and at the interface between the matrix and fillers [14]. The presence of unreacted carbon double bonds increases material susceptibility to degradation, thereby reducing color stability [15]. Significant discoloration in restorative materials is a primary reason for the replacement of anterior composite resin restorations. The long-term color stability of bulk-fill resin composites remains a critical concern among clinicians [16]. Furthermore, despite the introduction of new photoinitiators and increased irradiance outputs, there is limited evidence regarding (1) the compatibility between recently developed flowable bulk-fill composites and third-generation polywave LCUs, (2) the influence of high-irradiance and short-exposure modes on both DC and color stability, and (3) the relationship between initial polymerization efficiency and post-staining color changes.
The compatibility of the third-generation polywave light-emitting diode (LED) LCU introduced to the market with the addition of new photoinitiators to the materials has not yet been verified with flowable bulk-fills, and the effectiveness of the modes has not been determined [17]. Therefore, this study aims to contribute novel data by evaluating how different polywave light-curing modes affect the DC and color stability of contemporary flowable bulk-fill resin composites, directly addressing the unresolved issues noted in recent reviews.
The null hypotheses of the study were: (1) the DC would not be affected by different light-curing modes and material types and (2) color stability would not be influenced by different light-curing modes, material types, and time.
METHODS
In this in vitro study, three flowable bulk-fill composite resins (Estelite Bulk Fill Flow [EBF; Tokuyama Dental, Tokyo, Japan], Tetric N-Flow Bulk Fill [TNF; Ivoclar Vivadent, Schaan, Liechtenstein], and Filtek Bulk Fill Flowable Restorative [FBF; 3M ESPE, St. Paul, MN, USA]) were evaluated, along with one injectable flowable composite resin (G-aenial Universal Injectable [GUI; GC Corp., Tokyo, Japan]), which served as the control group. The characteristics and specifications of the materials evaluated in this study are summarized in Table 1. The experimental workflow of this study is summarized in Figure 1.
Flowable resin composite materials used in the study
The sample size was determined using G*Power 3.1.9.7 software (Heinrich Heine University, Düsseldorf, Germany), based on data from a previous study [18]. Using one-way analysis of variance (ANOVA) with an effect size of 0.40, corresponding to an expected difference of 2.69 units with a standard deviation of 0.33, a significance level of α = 0.05, and a statistical power of 1 − β = 0.80, the required sample size was calculated. A total of 120 specimens—10 for each of the three subgroups of the four composite resins—were deemed sufficient.
Specimen preparation
Specimens (5 × 4 mm) were prepared using transparent silicone molds placed on glass plates with Mylar strips on both sides. The composites were injected, covered with another Mylar strip and glass plate, and light-cured using a third-generation LED LCU (VALO Cordless; Ultradent, South Jordan, UT, USA) positioned perpendicularly with the tip in contact with the glass surface. Polymerization was performed in three modes: standard (1,000 mW/cm², 20 seconds), high-power (1,400 mW/cm², 3 × 4 seconds), and extra-power (3,200 mW/cm², 2 × 3 seconds). Measurements taken at the tip of the device indicated corresponding radiant exposures of approximately 20.0, 16.8, and 19.2 J/cm², respectively. To determine the irradiance reaching the specimen, measurements were performed at the specimen plane through the interposed glass plate and Mylar strips using a calibrated radiometer (SDI RADII Plus; SDI Limited, Bayswater, VIC, Australia). The corresponding radiant exposures at the specimen plane were calculated as approximately 19.0, 15.8, and 18.1 J/cm². The corresponding radiant exposures at the specimen plane were calculated as approximately 19.0, 15.8, and 18.1 J/cm². The output intensity of the LCU was verified before each curing session using a calibrated radiometer to ensure consistency across all specimens.
A total of 120 specimens (n = 10 per subgroup) were fabricated and stored dry in a 37°C incubator for 24 hours to complete polymerization. Finishing and polishing procedures were conducted by a single operator to minimize operator-related variability. Surface standardization was achieved using sequential aluminum oxide discs of coarse, medium, fine, and superfine grits (Sof-Lex Discs; 3M ESPE) applied for 5 seconds each, followed by medium-fine diamond-grit wheels (Clearfil Twist Dia; Kuraray Europe GmbH, Hattersheim am Main, Germany) for 30 seconds each [14,19]. All procedures were performed dry at 10,000 revolutions per minute with overlapping unidirectional strokes to prevent heat and grooves. To ensure consistency, approximately constant manual pressure was applied using only the inherent weight of the handpiece, without extra force; the applied pressure was standardized on a force sensor (Axis, Gdańsk, Poland) at around 1–2 N, consistent with previous reports [20]. Discs were replaced after every three uses, and specimens were rinsed with water for 20 seconds and air-dried for 5 seconds between steps.
Specimen thickness was measured and verified at 4 ± 0.1 mm using a digital caliper (Mitutoyo Corp., Kawasaki, Japan).
Degree of conversion analysis
The DC was assessed via Fourier-transform infrared (FTIR) spectroscopy (Spectrum One; PerkinElmer, Waltham, MA, USA) equipped with a universal attenuated total reflectance (uATR) accessory featuring a single-reflection diamond crystal (refractive index n = 2.40) at a 45° incidence angle. A constant pressure of approximately 70 N was applied using the integrated clamp to ensure consistent specimen–crystal contact. The effective ATR penetration depth was calculated at 1,637 cm–1—the wavenumber used for DC analysis—using the standard equation for a 45° diamond ATR crystal. Assuming refractive indices of 1.52 for the composite resin and 2.40 for the diamond crystal, the penetration depth at 1,637 cm–1 was estimated to be approximately 1.2–1.4 µm.
Specimens were stored dry for 24 hours prior to evaluation to avoid potential interference of residual surface or absorbed water with ATR-FTIR spectra, and in light-proof containers to prevent polymerization changes. All FTIR spectra were recorded on the same day for consistency. Prior to measurements, the ATR crystal was cleaned with alcohol and background scans were performed for calibration, including void check and linearity confirmation. Spectra were collected in absorbance mode (4,000–650 cm–1, 4 cm–1 resolution, 32 scans) and processed using Spectrum ver. 5.0.1 software (PerkinElmer) with baseline correction and automatic peak detection. Uncured composites were placed directly on the ATR crystal, while cured specimens (5 × 5 × 4 mm) were positioned with the bottom (non-irradiated) surface against the crystal and slightly compressed to ensure proper contact. For reproducibility, three measurements were taken for each specimen, and the averaged data were used to calculate a single DC value. In addition, one representative specimen from each group was analyzed for both top and bottom surfaces to verify polymerization uniformity.
Spectra of each resin were analyzed. Baseline correction in the 1,450–1,700 cm–1 region was performed using a straight-line baseline defined between these limits, and peak integration was based on peak height relative to the corrected baseline [21]. The peak height around 1,637 cm–1, indicating the absorbance intensities of aliphatic C=C, was calibrated according to Rueggeberg et al. [21]. Absorbance at 1,637 and 1,510 cm–1, corresponding to the aliphatic and aromatic C=C vibrations, respectively, was recorded (Figure 2). The aromatic band at 1,510 cm–1 served as the internal reference [22]. The DC was calculated from the absorbance values of uncured and cured specimens using the following formula [4].
Representative FTIR (Fourier-transform infrared) spectra of materials under different curing modes, showing the average absorbance curves. Two reference peaks are indicated: 1,510 cm–1 (internal standard aromatic carbon double bond, C=C) and 1637 cm–1 (methacrylate C=C). Material abbreviations and manufacturer information are provided in Table 1.
Evaluation of color change
Baseline color measurements (T0) of the post-polymerized specimens were obtained using a VITA Easyshade V digital spectrophotometer (VITA Zahnfabrik, Bad Säckingen, Germany). In accordance with the Commission Internationale de l'Éclairage (CIE) system, the primary color parameters (L*, a*, b*) of each specimen were measured three times against a white background within a light box illuminated by D65 lighting, and the mean values were subsequently calculated. The L* value represents the black/white color range (lightness), a* represents the red (+)/green (−) color range, and b* represents the yellow (+)/blue (−) color range [14].
Specimens that underwent baseline (T0) color measurements were marked on their bottom surfaces and individually immersed in light-proof amber bottles containing coffee solution for 14 days. To prepare the coffee solution, 2 g of coffee (Nescafé Classic; Nestlé, Barcelona, Spain) was dissolved in 200 mL of boiling water, which was subsequently cooled to ambient temperature prior to application. The solution was renewed daily. Prior to each solution change, samples underwent a 60-second rinse using distilled water to eliminate surface residues. Ertaş et al. [23] reported that a 7-day coffee immersion simulated approximately 7 months of coffee consumption. Furthermore, given that the actual contact time of coffee with oral tissues is shorter during drinking, the 14-day coffee immersion used in this study represents a longer period than 14 months of actual coffee consumption. Color measurements were repeated on day 7 (T1) and day 14 (T2). Color variation (CIEDE2000, ΔE₀₀) was determined using the following formula [24].
In this formula, ΔL represents the change in lightness, ΔC the change in chroma, and ΔH the change in hue. SL, SC, and SH are weighting functions used to calibrate for visual non-uniformity. The parametric factors for environmental correction, KL, KC, and KH, were all set to 1. RT is a rotational function incorporated to compensate for the interaction between chroma and hue differences, particularly in the blue region of the color space.
Color change evaluation was based on a 50:50 perceptibility threshold (ΔEPT) set at 0.8 and a 50:50 acceptability threshold (ΔEAT) defined as 1.8 [24].
Statistical analysis
Statistical analyses were performed using IBM SPSS ver. 21.0 (IBM Corp., Armonk, NY, USA) at a 95% confidence level (p < 0.05). The assumption of normality was tested using the Shapiro-Wilk test, and homogeneity of variances was assessed with the Levene test. The effects of material type and light-curing mode on DC were evaluated using two-way ANOVA. A three-way repeated-measures ANOVA was conducted to evaluate the effects of material and light-curing mode as between-subjects factors, and time (two levels: day 7 and day 14) as the within-subjects factor on color change (ΔE₀₀). Because the within-subjects factor consisted of only two levels, the sphericity assumption was inherently satisfied. One-way ANOVA was performed to compare differences among materials within each light-curing mode, as well as differences among light-curing modes within each material. For pairwise comparisons, Bonferroni or Dunnett’s T3 post hoc tests were applied depending on the homogeneity of variances. The relationship between DC and ΔE₀₀ was examined through the Pearson correlation and linear regression analyses.
RESULTS
Degree of conversion results
The DC results of the tested materials across different light-curing modes are presented in Table 2. A statistically significant difference in DC was found between light-curing modes for all materials except TNF (p < 0.001). The highest DC values were observed in the high-power curing mode for all tested materials. FBF exhibited the highest DC in the high-power mode (82.04 ± 9.04), while EBF showed the lowest DC in the standard power mode (43.34 ± 15.10). Polymerization of EBF in the standard mode resulted in significantly lower DC compared to other light-curing modes (p < 0.05). For GUI and FBF, the high-power mode produced significantly higher DC values than other curing modes (p < 0.05).
Degree of conversion (%) of tested materials in different light-curing modes
The differences in DC among the materials were statistically significant in the standard and high-power curing modes (p < 0.001), but not in the extra-power mode (p = 0.224). In the standard mode, FBF exhibited significantly higher DC values than GUI and EBF, and in the high-power mode, a statistically significant increase in DC values was observed in this material relative to the others (p < 0.05).
According to the two-way ANOVA results, material type, light-curing mode, and the interaction between material type and light-curing mode had a significant effect on the DC (p < 0.001) (Table 3).
Two-way analysis of variance results for degree of conversion (%)
Color change results
The color change results of the tested materials across different light-curing modes on days 7 and 14 are presented in Table 4. On day 7, differences in ΔE₀₀ between light-curing modes were not statistically significant for GUI (p = 0.215) and TNF (p = 0.081), whereas statistically significant differences were identified for EBF and FBF (p < 0.001). All materials showed lower ΔE₀₀ values when cured with the high-power mode compared to other modes. The lowest ΔE₀₀ was recorded for GUI in the high-power mode (1.76 ± 0.54), whereas the highest was observed for FBF in the extra-power mode (8.35 ± 1.06).
Color change (ΔE00) of tested materials in different light-curing modes
On day 14, differences in ΔE₀₀ between light-curing modes were statistically significant for all materials except GUI (p < 0.005). Although ΔE₀₀ increased over time (p < 0.05), the lowest value on day 14 was again observed for GUI in the high-power mode (2.83 ± 0.84). On both days 7 and 14, EBF exhibited significantly higher ΔE₀₀ values in the extra-power mode in comparison with the other modes (p < 0.05), whereas FBF showed significantly lower ΔE₀₀ values in the high-power mode in comparison with the others (p < 0.05).
At both time points, statistically significant differences in ΔE₀₀ were found among the materials across all light-curing modes (p < 0.001).
The three-way repeated-measures ANOVA revealed significant effects of material type, light-curing mode, time, and their interactions on the ΔE₀₀ of the tested materials (p < 0.001), except for the three-way interaction (time × material × light-curing mode), which was not significant (p = 0.136) (Table 5).
Three-way repeated-measures analysis of variance results for color change (ΔE00)
A significant negative association was observed between DC and ΔE00 values on day 7 (β = –0.208, p = 0.023) and day 14 (β = –0.238, p = 0.009). Furthermore, a strong positive correlation was found between the color changes measured on days 7 and 14 (r = 0.974, p < 0.001) (Tables 6 and 7).
Pearson correlation analysis between DC and color change
Linear regression analysis evaluating the association between DC and color change
DISCUSSION
The extent of polymerization, expressed as the DC, critically affects the performance and biocompatibility of resin composites. It is influenced by total energy input, which depends on irradiance and curing time [4]. However, no consensus exists on the optimal radiant exposure, as DC varies by material, making total energy-based protocols unreliable [11]. Manufacturers suggest that short, high-intensity exposures may be as effective as standard protocols. In this study, the DC and color changes of flowable bulk-fill composites polymerized using different light-curing modes—each involving varying energy densities and irradiation times—were investigated and compared with an injectable flowable composite resin. Results showed that both light-curing mode and material type significantly affected DC; therefore, the first null hypothesis of the study was refuted.
FTIR, a quantitative method used to measure DC, is considered the gold standard; however, the choice of reference peak may influence calculated conversion values. As an internal standard, the band at 1,608–1,610 cm−1 representing the aromatic C=C absorption is commonly used [25]. The availability of this vibrational band depends on the presence of methacrylate monomers containing an aromatic core, such as bisphenol A glycidyl methacrylate (Bis-GMA) and bisphenol A dimethacrylate (Bis-EMA). In composites based on aliphatic monomers, e.g., urethane dimethacrylate (UDMA), the vibrational band at 1,608–1,610 cm−1 is absent, and other alternative reference bands are used [25]. According to the manufacturer’s Safety Data Sheet information, the tested material GUI contains only aliphatic monomers such as UDMA and triethylene glycol dimethacrylate (TEGDMA), and no aromatic monomer is specified. However, the literature reports that GUI may also contain aromatic dimethacrylates such as Bis-EMA [26]. In the uncured FTIR spectra obtained in this study, it was determined that no distinct peak was present in the 1,608–1,610 cm–1 region for the tested materials, and that only very weak signals were observed in some samples. In contrast, the 1,510 cm–1 band exhibited a more stable and stronger peak in all four materials (Figure 2). Therefore, the 1,510 cm–1 region was preferred as the internal reference for the DC calculations.
Studies investigating the DC in bulk-fill composite resins show considerable methodological variability in terms of both measurement timing and the surface from which DC is assessed. In a network meta-analysis synthesizing the findings of in vitro studies, Hatipoğlu et al. [5] reported significant differences between top and bottom surface DC values when measured immediately after curing, whereas no significant difference was observed between the two surfaces when measurements were performed 24 hours post-curing. Since bottom-surface DC provides more meaningful and comparable information regarding curing efficiency among materials, DC measurements in the present study were performed on the bottom surfaces of the specimens and at 24 hours after curing. Additionally, to verify polymerization homogeneity, one representative specimen from each group was analyzed on both the top and bottom surfaces, and the top surface was found to exhibit approximately 1.5%–2% higher DC values.
The present research found that while radiant exposures did not significantly differ between the three curing modes, the high-power mode produced greater DC in every material assessed. Similarly, in a study by Yenidünya et al. [27], which examined the effect of different light-curing modes on the DC in various materials, significantly higher DC values were reported under the high-power mode. The higher DC observed in the high-power mode may be attributed to the greater initial irradiance (1,400 mW/cm²) compared to the standard mode. This high irradiance could lead to a greater increase in exothermic heat within the polymeric matrix, enhancing the mobility of polymer chains and consequently increasing the DC [28]. However, relatively lower DC values were recorded in the extra-power mode. Very high irradiance combined with very short curing times accelerates radical generation at the beginning; however, the rapid weight gain leads to early vitrification, which restricts chain mobility. This condition alters the rate of bimolecular termination and, consequently, the final hardness/conversion depending on the material’s depth profile [12]. The relatively low DC observed in the extra-power mode in this study may be attributed to the accelerated polymerization rate. On the other hand, in cases where the reaction proceeds at a normal rate, an increase in post-irradiation DC has been reported for most resin matrices [29]. After irradiation, polymerization continues depending on factors such as residual monomer content, cross-linking density, radical concentration, and radical lifetimes [30]. The high-power mode may have generated more free radicals by providing increased photon availability, enhancing post-irradiation polymerization [31].
The viscosity of materials is an important factor affecting the DC. Compared to high-viscosity materials, flowable resin-based bulk-fill composites contain a greater proportion of organic matrix, which may lead to a reduction in the refractive index [6]. The matrices of flowable resin-based composites consist of low molecular weight monomers characterized by increased flexibility and reactivity, which have been reported to augment the number of bonding sites during light curing, consequently improving the DC [6]. In the current study, FBF exhibited a higher DC compared to the other materials across all curing modes, with statistically significant differences noted specifically in the standard and high-power curing modes. FBF contains Bis-GMA, TEGDMA, Bis-EMA, and a procrylate monomer. The procrylate monomer is a high molecular weight monomer similar to Bis-GMA; however, it does not contain pendant hydroxyl groups that contribute to increased viscosity [32]. Dental polymers incorporating Bis-EMA and urethane-based monomers with lower viscosity have been demonstrated to achieve higher DC values relative to conventional Bis-GMA/TEGDMA resins [32]. This may account for the elevated DC observed in FBF relative to the other flowable materials.
Highly translucent resin-based composites offer better curing efficiency due to greater light transmission through the material, allowing light to reach deeper layers [3]. The relatively low DC observed in TNF compared to the other materials may be attributed to its lower translucency. Although TNF contains a photoinitiator system (Ivocerin; Ivoclar Vivadent) that exhibits higher curing activity by absorbing visible light at a wider wavelength range than germanium-based camphorquinone (CQ), it does not fully compensate for the reduced translucency in deeper layers. This finding is consistent with the study by Ilie et al. [33].
Although GUI, an injectable composite, is not a bulk-fill composite, it is frequently used in clinical practice at thicknesses exceeding 2 mm. In the present study, considering these clinical conditions, GUI specimens were prepared at a thickness of 4 mm to evaluate critical thresholds for DC. Although there is no consensus on a minimum DC value, values between 55%–65% are generally considered critical for mechanical performance [11]. Notably, this range reflects unreacted C=C bonds, not unpolymerized monomers [15]. GUI showed acceptable DC (55%–65%) despite exceeding the recommended thickness. In contrast, EBF and TNF under standard mode exhibited DC below this range, suggesting suboptimal polymerization. Given that curing efficiency varies with monomer type [34], high-power mode may be more appropriate for EBF and TNF. It should also be noted that due to vitrification or gelation, DC rarely reaches 100% [35].
The present study demonstrated that color change was significantly affected by light-curing mode, material type, and time (Table 5), leading to the rejection of the study’s second null hypothesis. The best performance in terms of color stability was observed in specimens polymerized with the high-power mode, while groups cured for shorter durations using the extra-power mode exhibited greater color change. During prolonged polymerization, increased light penetration into deeper layers—similar to that at the beginning of irradiation—explains the limitations of the short exposures in the extra-power mode [12]. The unstable resin matrix and reduced DC caused by the short exposure may contribute to increased discoloration of the materials.
In dental composites, CQ and tertiary amines—used as co-initiators to enhance free radical generation—constitute the most common photoinitiator system [36]. To reduce the disadvantages of this photoinitiator, such as weak bleaching properties, yellowish staining, and toxicity, Rap Technology (radical amplified photopolymerization; Tokuyama Dental), which recycles CQ during initiator production, has been developed [36]. The lower color change observed in EBF compared to FBF can be explained by its reduced CQ content due to Rap Technology. Across all curing modes, GUI exhibited less color change than the other materials, which may be attributed to the full-coverage silane coating technology claimed by the manufacturer to enhance filler-matrix bonding [37].
Staining of composite resins is influenced by the polishing protocol and the resulting surface roughness [38]. Because the presence of water during polishing makes it difficult to accurately assess surface texture and gloss, several authors have recommended dry polishing [39]. In addition, immediate finishing and polishing may lead to plastic deformation, as the material has not yet reached its final hardness due to ongoing polymerization. For this reason, postponing finishing and polishing for 24 hours after light curing has been suggested [38], and the same approach was followed in the present study. However, the heat generated during dry finishing and polishing may cause degradation at the filler-matrix interface or microcrack formation, potentially contributing to increased discoloration. These factors should be considered when interpreting acceptability thresholds, as in vitro protocols may not fully replicate clinical conditions. Clinically, perceptibility (ΔE₀₀ = 0.8) and acceptability (ΔE₀₀ = 1.8) thresholds are critical benchmarks for esthetic performance [24]. In this study, all groups exceeded the perceptibility threshold after 7 and 14 days of coffee immersion. On day 7, only GUI polymerized in high-power mode remained within the acceptability limit, whereas all groups surpassed it by day 14. Although continuous coffee exposure without saliva or brushing likely intensified staining, these results highlight that both material formulation and curing parameters directly affect long-term esthetic stability.
Staining of composites is affected by surface roughness, filler characteristics, and the hydrophilicity of the resin matrix [40]. Nanofill composites, with smaller particles, offer smoother surfaces, reduced water absorption, and better filler-matrix integration [41,42]. In this study, GUI, which showed the least color change, is a nanofill composite, while FBF, which exhibited the highest discoloration, is a microhybrid composite. The findings of the present study are in agreement with those reported by Gürdal et al. [43].
Materials exhibiting a high DC demonstrate reduced deformation due to increased polymer chain entanglement and elevated cross-linking density, which collectively constrain molecular mobility within the polymeric network [5]. Composites with lower DC release greater amounts of unreacted monomers, such as Bis-GMA and TEGDMA, which may diffuse toward the pulp and induce cytotoxic or inflammatory responses. Moreover, residual monomers can undergo oxidative and hydrolytic degradation, accelerating color instability and surface deterioration [15]. As reported by Sham et al. [44], chemical discoloration is primarily attributed to the oxidation of the polymer matrix, the oxidation of residual monomer double bonds that remain unreacted, and the formation of degradation products induced by water diffusion. In the present study, the observed negative correlation between DC and ΔE₀₀ supports that insufficient polymerization may contribute to greater color change in resin composites.
A key limitation of this study is that the 1,510 cm–1 band used as the internal reference does not represent a single, chemically isolated vibrational mode. In composites containing Bis-GMA/Bis-EMA, this region primarily reflects the aromatic ring C=C stretching vibration, but it also receives measurable contributions from aromatic C–H in-plane bending. In UDMA-based systems, additional overlap from the amide II mode (a combination of N–H bending and C–N stretching) may further broaden this band [45]. Therefore, the 1,510 cm–1 signal should not be interpreted as a structurally invariant and “chemically clean” reference band in the same sense as the conventional 1,608–1,610 cm–1 aromatic stretching peak.
Al-Zain et al. [46] evaluated both dry and wet storage conditions in terms of DC and reported that, under wet storage, the intermolecular bonds within the polymer system were softened due to the hydrolytic effect of water, resulting in a tendency toward reduced DC. The dry storage condition adopted in the present study to avoid potential FTIR spectral alterations related to water absorption may not fully simulate the clinical scenario in which dental materials are placed and may yield comparatively higher DC values. The use of a single light-curing unit represents another limitation, as potential radiometric inaccuracies and directional variations in light output at a 4-mm depth could have influenced the results. Additionally, the lack of a distilled water control group and a non-polished comparison group limited the ability to distinguish water absorption and surface effects from those caused by staining solutions. Future studies including a distilled water control group, various staining solutions, a non-polished comparison group, thermal aging, simulated brushing, and DC variations after coffee staining in relation to optical properties would enhance clinical relevance and clarify the mechanisms underlying color changes.
CONCLUSIONS
Within the limitations of this study, the different light-curing modes significantly affected both the DC and the discoloration behavior of the flowable composites. Although the extra-power mode shortens curing time, it may lead to a reduction in DC and an increase in discoloration. The discoloration of the tested composites progressed over time, and color stability was influenced not only by DC but also by the material type and its chemical composition. Clinically, these findings suggest that using excessively high curing intensities for short durations may compromise the long-term color stability and polymerization quality of flowable composites. Therefore, clinicians should balance curing efficiency with adequate polymerization to optimize the esthetic longevity of restorations.
Notes
CONFLICT OF INTEREST
No potential conflict of interest relevant to this article was reported.
FUNDING/SUPPORT
The financial support of this study was provided by Ordu University Scientific Research Projects Coordination Unit with project number B-2340.
ACKNOWLEDGEMENTS
The authors would like to thank the Scientific and Technological Research Center of Inonu University for their support in the FTIR analysis.
AUTHOR CONTRIBUTIONS
Conceptualization, Data curation, Funding acquisition, Investigation, Project administration, Resources, Software, Supervision: Karaaslan E, Karademir SA; Formal analysis, Validation: Karademir SA; Methodology, Visualization: Karaaslan E; Writing - original draft: Karaaslan E, Karademir SA; Writing - review & editing: Karaaslan E, Karademir SA. All authors read and approved the final manuscript.
DATA SHARING STATEMENT
The entirety of the data analyzed in this study has been included in the published article. The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.
DISCLOSURE OF GENERATIVE AI IN SCIENTIFIC WRITING
Artificial intelligence was used solely to assist with the language editing of this article.