Influence of different drying protocols on bond strength of epoxy resin- and calcium silicate- based root canal sealers in nonsurgical retreatment Authors: Derin Bugu Yuzer, Birgul Ozasir Published: 2026 ABSTRACT AbstractBackgroundThe condition of root canal dentin and the drying protocol applied prior to obturation may influence the bonding performance of root canal sealers, particularly in retreatment cases where dentin characteristics are altered. This study aimed to evaluate the effect of different drying protocols on the push-out bond strength of epoxy resin– and calcium silicate–based sealers in primary root canal treatment and nonsurgical retreatment.MethodsA total of 192 extracted human mandibular premolars were prepared and randomly allocated into groups according to drying protocol (paper point, EndoVac, ethanol, irrigation needle), sealer type (AH Plus, AH Plus Bioceramic), and treatment type (primary vs. retreatment). In retreatment groups, canals were re-prepared to a larger apical size (X4) without the use of solvents. After obturation, specimens were sectioned into coronal, middle, and apical slices, and push-out bond strength (MPa) was measured. Data were analysed using a four-factor linear mixed-effects model with tooth as a random effect. Post hoc comparisons were performed using Tukey adjustment (α = 0.05).ResultsBond strength values generally increased from the coronal to the apical region across all groups. Retreatment specimens exhibited higher bond strength values than primary treatment, particularly in the apical region based on estimated marginal means (6.29 ± 0.35 vs. 2.78 ± 0.35 MPa). Ethanol drying yielded higher bond strength values in several apical conditions compared with other drying methods. No significant differences were observed between AH Plus and AH Plus Bioceramic within the coronal, middle, or apical regions (P > .05). Linear mixed-effects analysis demonstrated significant effects of treatment type (P < .001), root section (P < .001), and drying protocol (P = .001), whereas sealer type showed no significant main effect (P = .669). Significant interactions were identified between treatment type and root section, drying protocol and root section, and among treatment type, sealer type, drying protocol, and root section (P = .003). Cohesive and mixed failures predominated across experimental groups.ConclusionsBond strength of root canal sealers is primarily influenced by root section and treatment condition, with drying protocol playing a secondary but context-dependent role. Retreatment procedures may enhance bond strength, particularly in the apical region, while sealer type alone does not significantly affect bonding performance.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12903-026-08813-6.Keywords: Bond strength, Calcium silicate sealer, Drying protocol, Epoxy resin sealer, Push-out test, Root canal retreatment Background Root canal treatment is a procedure performed to prevent or heal pulp infection and/or periapical pathology, involving stages such as preparation, irrigation, and obturation [1]. Although there are different techniques available depending on the case, obturation procedures are generally performed with gutta-percha and root canal sealers [2]. Since gutta-percha does not bond to dentin tissue, it is used in conjunction with sealers to fill the complex anatomical elements of the root canal [3, 4]. Epoxy resin-based sealers are considered the gold standard due to their ability to fill irregularities in the root canal and their high bonding strength to dentin, but cytotoxicity during the initial setting is a concern [1, 5, 6]. On the other hand, calcium silicate-based sealers with bioactive properties and low cytotoxicity are gaining popularity [1, 4, 7]. Due to differences in composition, the relationship between the filling materials and the root dentin, and therefore the bond strength, may vary [2, 7]. In cases of inadequate bonding, microbial leakage into the periapical region can occur through gaps and voids, leading to root canal failure, where retreatment may be necessary [1, 2, 4]. Non-surgical retreatment usually involves removing the initial treatment filling material from the root canal, reshaping and disinfecting the canal, and refilling it [8]. Although sealers have specific properties, the condition of the root dentin during filling plays a critical role in bonding. Thorough irrigation is applied to the root dentin to eliminate vital or non-vital tissues, to reduce the number of microorganisms and to remove the smear layer formed during preparation in both initial root canal therapy and retreatment [8–10]. Furthermore, dentin characteristics after retreatment procedures differ from those of primary treatment due to residual filling material, altered smear layer composition, and potential dentinal tubule obstruction, which may influence sealer penetration and bonding [11, 12]. Drying protocols are applied after the final irrigation procedures. Although there is no established protocol for the ideal moisture level that should remain in the root canal, it has been reported that dentin dryness affects the bond strength depending on the polarity of the sealers [3, 13–15]. In practice, drying processes are usually carried out using paper point cones. Drying can also be achieved using an irrigation needle through the principle of negative pressure. Isopropyl alcohol can also provide rapid drying due to its water-repellent effect. The EndoVac (Kerr Lab Sybron Dental, Orange, CA, USA) irrigation device can also be used for root canal drying; however, research on its effect on bond strength is limited [3, 7, 14]. Therefore, this study aimed to compare how different drying methods affect the bond strength of epoxy resin and calcium silicate-based sealers in primary root canal treatment and retreatment procedures. The null hypotheses were that (i) different drying protocols would not affect the bond strength of either sealer, (ii) there would be no difference in bond strength between epoxy resin- and calcium silicate-based sealers under the same drying conditions, and (iii) the treatment protocol (primary vs. retreatment) would not influence bond strength. Methods Selection and preparation of teeth A prior power analysis was performed using G*Power software (version 3.1.9.2; Heinrich Heine University, Düsseldorf, Germany) to determine the minimum required sample size. The calculation was based on a three-way fixed-effects ANOVA model (F tests – ANOVA: Fixed effects, special, main effects and interactions), considering drying protocol (4 levels), treatment procedure (2 levels), and sealer type (2 levels) as independent factors. Assuming a medium effect size (f = 0.25), a significance level of α = 0.05, and a statistical power (1 − β) of 0.80, with 16 experimental groups (4 × 2 × 2), the required total sample size was calculated as 179 specimens (numerator df = 3). The analysis yielded a noncentrality parameter (λ) of 11.19 and a critical F value of 2.66. To compensate for potential specimen loss and to maintain balanced group allocation, 192 teeth (12 specimens per subgroup) were included in the study, resulting in an actual power of 0.80. This study included 192 anonymized, single-rooted human mandibular second premolar teeth extracted for dental treatment procedures, unrelated to the present study. For disinfection, the teeth were immersed in 0.1% thymol solution for 48 h and kept in 4 °C distilled water until use. Periapical radiographs were taken of the teeth from the bucco-lingual and mesio-distal directions, and teeth with multiple root canals were excluded from the study. Additionally, teeth that had previously undergone root canal treatment, showed internal or external resorption, had calcified root canals, exhibited open apexes with incomplete root development, had caries in the crown or root, showed a crack line or fracture were excluded from the study. Teeth with root curvature greater than 10° were excluded to ensure comparable canal morphology. Additionally, to standardize initial canal diameter, teeth in which a size 15 K-file could passively extend beyond the apical foramen without resistance were excluded from the study. Before starting root canal preparation, teeth were standardized to a root length of 16 mm by cutting the crown parts with a diamond fissure bur under water cooling. After the canal orifices were detected using a diamond-tipped round bur (801 –016 C), the working length was determined using a #10 K-file (Technical & General Ltd, London, UK) to be 0.5 mm shorter than the length seen at the apex. The root tips were sealed with adhesive wax to create a closed-end system to prevent the irrigation solution from flowing through the apical foramen. The root canals of the specimens were enlarged to X3 file according to the operating principles of the ProTaper Next (PTN; Dentsply Maillefer, Ballaigues, Switzerland) rotary system. Each file was used for 4 canal preparations, then was discarded. For the removal of debris after each file, 2 ml of 2.5% sodium hypochlorite (NaOCl; Werax, İzmir, Turkey) was used. The irrigation procedure was performed with 30-gauge (ga) side-perforated irrigation needles (Ultradent Products Inc, South Jordan, UT, USA). Following root canal preparation, the root canals were rinsed with 5 ml of 2.5% NaOCl, then 5 ml of 17% ethylenediaminetetraacetic acid (EDTA; Werax, Izmir, Turkey) for 1 min to remove the smear layer, followed by 5 ml of 2.5% NaOCl solution. Finally, the irrigation process was completed with 5 ml of distilled water to remove the remaining solutions. Following chemo-mechanical preparation, the specimens were randomly allocated to the experimental groups using a computer-generated random number sequence (simple randomization). Allocation was performed to ensure equal group sizes (n = 48 per drying protocol group). Subsequent subgroup allocation according to sealer type was carried out using the same randomization method. In group 1, the drying procedure was performed using #30 size paper points until the last paper point appeared visibly dry. In group 2, following final irrigation, the macro-cannula tip of EndoVac system was inserted to the coronal third and moved with gentle in-and-out motions to perform suction. Subsequently, the microcannula tip was positioned at working length and maintained for 10 s to evacuate the remaining solution by negative pressure. The procedure was repeated until no visible fluid was observed at the cannula tip. In the group 3, the canal space was dried out with paper points as in group 1. Then, 2.5 ml 70% ethanol was delivered to the canal space to working length using a 30 ga side-perforated needle. After waiting for 5 s, the solution was aspirated with the needle, and a single paper point was used to remove the residual fluid. In group 4, the canals were dried by aspiration with a 30 ga side-vented irrigation needle. The needle was inserted 1 mm short of the working length without binding and moved coronally for 5 s while continuously pulling the needle. For each group, the samples were randomly divided into two subgroups (n = 24) described previously according to the sealer for obturation. AH Plus (AHP; DeTrey GmbH, Konstanz, Germany) was prepared according to the manufacturer’s recommendations and applied to the canals using a #25 size lentulo spiral. AH Plus Bioceramic (AHPB; Dentsply Sirona, Charlotte, NC, USA) was delivered directly into the canal using its own syringe system. Both sealers were applied to fill the canals up to the canal orifices. The root canals were filled with ProTaper X3 gutta-percha (Dentsply Maillefer, Ballaigues, Switzerland), which provides tug-back feel, using single-cone technique. Excess gutta-percha was cut away using a heated plugger and vertical condensation was applied. The canal orifices were sealed with light-curing composite (Filltek Z250 XT; 3 M ESPE, St. Paul, MN, USA) approximately 1 mm in depth, and the samples were left for 14 days at 37 °C and 100% humidity, to let the sealers set. Twelve samples from each group were randomly assigned for bond strength testing (n = 12). The remaining samples were subjected to a retreatment procedure. In the retreatment groups, the composite restoration was removed using a diamond round bur under water cooling. The existing root canal filling was removed using the D1, D2, and D3 files of the ProTaper Universal Retreatment rotary file system (Dentsply Sirona, Ballaigues, Switzerland), respectively. No gutta-percha solvent was applied to the canals. Removal of the existing root canal filling was confirmed radiographically. The preparation of the canals was completed with the X4 file of the PTN system. After each file, the canals were irrigated with 2 ml of 2.5% NaOCl. The final irrigation and drying protocols were applied in the same way as in the initial treatment. The root canals were obturated using ProTaper X4 gutta-percha (Dentsply Maillefer, Ballaigues, Switzerland) and the sealers (AHP, AHPB), sealed with composite and kept in the same way described previously. All teeth were embedded in acrylic resin (Meliodent, Heraues Kulzer, Hanau, Germany) vertically up to the composite restoration margin. Subsequently, 3 horizontal sections of 1 ± 0.1 mm thickness were taken from the blocks placed in the Isomet 1000 device (Buehler, Lake Bluff, IL, USA) under water cooling and at low speed, at distances of 3, 8 and 12 mm from the apex. The coronal surfaces of the sections were marked with a marker to serve as a reference for the bond strength test. In the bond strength test, the force required to dislodge the root canal filling material from the canal was measured using a Universal testing machine (Lloyd LRX; Lloyd Instruments Ltd, Hants, UK). Samples positioned in the centre of a metal base with three different diameter cavities were subjected to force from apical to coronal at a speed of 1 mm/min via pistons for apical, mid, and coronal segments with diameters of 0.5, 0.6, and 0.7 mm, respectively. The force applied at the moment of displacement of the filling material was recorded in Newtons, and the push-bond strength was calculated in megapascals (MPa) for each sample according to the following equation: Bond strength (MPa) = Fmax / adhesion surface area (mm²). The adhesion surface area (A) was calculated using the formula for the lateral surface area of a truncated cone: A = π (R + r) √ [h² + (R2 – r2)], where R represents the coronal radius, r the apical radius, and h the thickness of the slice. Adhesion surface area was calculated separately according to the respective canal dimensions, taking into account the differences in apical preparation sizes between the primary and retreatment groups. After the bond strength test, the sections were examined under a stereo microscope (Leica MZ6, Leica, Germany) at ⋅10 magnification, and the failure modes occurring along the canal wall were evaluated as follows: Adhesive: Complete separation of the filling material from the dentin (dentin surface without sealer). Cohesive: Failure occurred within the filling material (dentin superficially covered with sealer). Mixed: Combination of both situations (Dentin surface partially covered with sealer) (Fig. 1). Representative stereomicroscopic images illustrating adhesive (a), cohesive (b), and mixed (c) failure patterns. Arrows indicate representative sealer remnants. Scale bars = 0.5 mm Push-out testing and failure mode evaluation were performed by a blinded investigator, with group codes revealed only after data analysis. Statistical analysis Data analysis was performed using the IBM SPSS Statistics 25.0 software package (IBM Corporation, Armonk, NY, US). Bond strength values were analysed using a four-factor factorial linear mixed-effects model. The fixed effects included treatment type (primary vs. retreatment), sealer type, drying procedure (paper point, EndoVac, ethanol, irrigation needle), and root section (coronal, middle, apical), as well as all possible interaction terms among these factors. Tooth was included as a random intercept to account for clustering of multiple sections within the same specimen. Slices were treated as the unit of analysis, with clustering within teeth accounted for by including tooth as a random effect. Type III sums of squares were used to evaluate the significance of main effects and interaction effects. Degrees of freedom for fixed effects were estimated using the Satterthwaite approximation. When significant effects were detected, estimated marginal means (EMMs) were calculated, and pairwise comparisons were performed using Tukey-adjusted post hoc tests to control for multiple comparisons. For post hoc analyses involving interaction effects, degrees of freedom were estimated using the Kenward–Roger method. Model assumptions were assessed using residual diagnostics. Because several subgroup values approached the lower measurement boundary (0 MPa), with standard deviations exceeding the mean in selected strata, suggesting skewness and floor effects, a sensitivity analysis using square-root transformed bond-strength values was additionally performed within the same mixed-effects framework. Results The raw push-out bond strength values (MPa) for all experimental groups are presented in Table 1. Overall, bond strength values increased from the coronal to the apical region across both treatment conditions and sealer types. Retreatment groups generally exhibited higher bond strength values than primary treatment, particularly in the apical Sect. (6.29 ± 0.35 vs. 2.78 ± 0.35 MPa). Table 1Push-out bond strength values (MPa; mean ± SD) for all experimental groupsSealer typeRegionDrying method n Primary treatmentRetreatmentAH PlusCoronalPaper point120.35 ± 0.252.60 ± 1.70Endovac121.24 ± 1.713.29 ± 1.96Ethanol120.29 ± 0.252.10 ± 2.02Needle121.22 ± 1.052.73 ± 4.20MiddlePaper point121.18 ± 0.643.02 ± 1.97Endovac121.32 ± 1.061.60 ± 1.95Ethanol120.94 ± 0.631.76 ± 1.86Needle121.35 ± 0.920.04 ± 0.05ApicalPaper point123.93 ± 2.814.36 ± 3.97Endovac123.27 ± 2.350.74 ± 0.90Ethanol123.00 ± 1.5310.05 ± 6.42Needle123.00 ± 1.756.30 ± 4.46AH Plus BioceramicCoronalPaper point120.13 ± 0.183.22 ± 3.75Endovac121.01 ± 1.602.41 ± 2.32Ethanol120.53 ± 0.520.61 ± 0.73Needle120.72 ± 0.640.16 ± 0.49MiddlePaper point120.51 ± 0.731.75 ± 2.89Endovac120.62 ± 0.661.75 ± 3.83Ethanol120.53 ± 0.714.05 ± 6.11Needle120.98 ± 0.710.13 ± 0.31ApicalPaper point122.06 ± 2.964.93 ± 6.28Endovac121.68 ± 1.179.64 ± 8.25Ethanol123.26 ± 3.3211.58 ± 9.77Needle122.01 ± 1.142.74 ± 2.76SD Standard deviation. For post hoc pairwise comparisons among drying procedures within each root section, see Table 5. Push-out bond strength values (MPa; mean ± SD) for all experimental groups SD Standard deviation. For post hoc pairwise comparisons among drying procedures within each root section, see Table 5. Differences among drying protocols were most pronounced in the apical region, where ethanol drying yielded higher bond strength values compared with other methods in several conditions. In contrast, variations among drying protocols were less consistent in the coronal and middle regions. Pairwise comparisons between sealer types within each root section revealed no statistically significant differences (Table 2). In the coronal section, AH Plus showed slightly higher bond strength values than AH Plus Bioceramic (mean difference = 0.63 MPa, P = .161). No significant differences were observed in the middle (P = .807) or apical sections (P = .364). Table 2Comparison of sealer types within each root sectionRegionMean Difference (MPa)SEdftP valueCoronal0.630.455171.410.161Middle0.110.455170.240.807Apical-0.410.45517-0.910.364SE Standard error Comparison of sealer types within each root section SE Standard error Bond strength analysis Bond strength values were analysed using a four-factor factorial linear mixed-effects model including treatment type (primary vs. retreatment), sealer type (AH Plus vs. AH Plus Bioceramic), drying procedure (paper point, EndoVac, ethanol, irrigation needle), and root section (coronal, middle, apical) as fixed effects, along with all possible interaction terms (Fig. 2). Tooth was included as a random intercept to account for clustering of multiple sections within the same specimen. Type III sums of squares were used, and degrees of freedom were estimated using the Satterthwaite method (Table 3). Effect sizes were expressed as partial eta squared (η²p), calculated from the reported F and degrees of freedom values. Push-out bond strength (MPa) across experimental groups. Each dot represents an individual specimen (n = 12 per subgroup). Black diamonds indicate mean values, and error bars represent standard deviations Table 3Results of the four-factor linear mixed-effects model for push-out bond strength (Type III tests of fixed effects)Effectdf (num, den)F P Treatment type1, 52855.98< 0.001Sealer type1, 5280.180.669Drying procedure3, 5285.300.001Root section2, 52866.22< 0.001Treatment × Sealer1, 5283.400.066Treatment × Drying3, 5286.19< 0.001Sealer × Drying3, 5283.680.012Treatment × Root section2, 5289.83< 0.001Sealer × Root section2, 5281.340.263Drying × Root section6, 5285.23< 0.001Treatment × Sealer × Drying3, 5283.870.009Treatment × Sealer × Root section2, 5284.540.011Treatment × Drying × Root section6, 5283.650.001Sealer × Drying × Root section6, 5282.820.010Treatment × Sealer × Drying × Root section6, 5283.330.003Degrees of freedom were estimated using the Satterthwaite method. Statistical significance was set at P < .05 Results of the four-factor linear mixed-effects model for push-out bond strength (Type III tests of fixed effects) Degrees of freedom were estimated using the Satterthwaite method. Statistical significance was set at P < .05 Main effects A significant main effect of treatment type was observed (F(1, 528) = 55.98, P < .001, η²p = .10), indicating a moderate-to-large effect of retreatment on bond strength. Root section demonstrated a strong main effect (F(2, 528) = 66.22, P < .001, η²p = .20), with bond strength values increasing progressively from the coronal to the apical region. Drying procedure had a statistically significant but small effect on bond strength (F(3, 528) = 5.30, P = .001, η²p = .03), suggesting a limited but measurable influence of drying method. In contrast, sealer type did not show a significant independent effect (F(1, 528) = 0.18, P = .669, η²p < .001). Given the presence of significant higher-order interactions, these main effects should be interpreted with caution. Interaction effects A significant interaction between treatment type and root section was identified (F(2, 528) = 9.83, P < .001, η²p = .04), indicating that the effect of retreatment varied depending on the anatomical level of the root canal. The interaction between drying procedure and root section was also significant (F(6, 528) = 5.23, P < .001, η²p = .06), demonstrating that the influence of drying protocols differed across coronal, middle, and apical regions. Significant two-way interactions were observed for treatment × drying (F(3, 528) = 6.19, P < .001, η²p = .03) and sealer × drying (F(3, 528) = 3.68, P = .012, η²p = .02), indicating that the effectiveness of drying procedures depended on both treatment condition and sealer type. Higher-order interactions were also statistically significant, including treatment × sealer × drying (F(3, 528) = 3.87, P = .009, η²p = .02), treatment × sealer × root section (F(2, 528) = 4.54, P = .011, η²p = .02), treatment × drying × root section (F(6, 528) = 3.65, P = .001, η²p = .04), sealer × drying × root section (F(6, 528) = 2.82, P = .010, η²p = .03), and the four-way interaction among all variables (F(6, 528) = 3.33, P = .003, η²p = .04). Section-specific analysis of treatment effect To further investigate the significant treatment × root section interaction, section-specific estimated marginal means (EMMs) and pairwise comparisons were analysed (Table 4). Table 4Estimated marginal means (± SE) of push-out bond strength according to treatment type within each root sectionRegionTreatment typeMean ± SE95% CICoronalPrimary treatment0.69 ± 0.350.00–1.37CoronalRetreatment2.14 ± 0.351.46–2.82MiddlePrimary treatment0.93 ± 0.350.25–1.61MiddleRetreatment1.76 ± 0.351.08–2.44ApicalPrimary treatment2.78 ± 0.352.10–3.46ApicalRetreatment6.29 ± 0.355.61–6.97SE Standard error, CI Confidence interval Estimated marginal means (± SE) of push-out bond strength according to treatment type within each root section SE Standard error, CI Confidence interval In the coronal section, retreatment resulted in significantly higher bond strength values compared with primary treatment (2.14 ± 0.35 vs. 0.69 ± 0.35 MPa; mean difference = 1.46, P = .003). In the middle section, although retreatment showed higher bond strength values than primary treatment (1.76 ± 0.35 vs. 0.93 ± 0.35 MPa), this difference was not statistically significant (mean difference = 0.83, P = .089). In the apical section, retreatment demonstrated markedly higher bond strength values compared with primary treatment (6.29 ± 0.35 vs. 2.78 ± 0.35 MPa), and this difference was highly significant (mean difference = 3.52, P < .001). Post hoc comparisons for drying procedures Post hoc comparisons based on estimated marginal means demonstrated that differences among drying procedures were most pronounced in the apical section. In this region, ethanol drying exhibited distinct bond strength values compared with other drying protocols, indicating a differential effect under apical conditions. In contrast, differences among drying protocols in the coronal and middle sections were less pronounced and did not follow a consistent pattern across treatment or sealer types (Table 5). Table 5Pairwise comparisons of drying procedures within each root section based on estimated marginal means (Tukey-adjusted)RegionComparisonMean difference (MPa)*SEdft P CoronalPaper Point – EndoVac-0.4100.633517-0.6470.917Paper Point – Ethanol0.6940.6335171.0950.693Paper Point – Needle0.3690.6335170.5820.938EndoVac – Ethanol1.1030.6335171.7420.303EndoVac – Needle0.7780.6335171.2290.609Ethanol – Needle-0.3250.633517-0.5130.956MiddlePaper Point – EndoVac0.2930.6335170.4630.967Paper Point – Ethanol-0.2040.633517-0.3220.988Paper Point – Needle0.9900.6335171.5630.401EndoVac – Ethanol-0.4970.633517-0.7850.861EndoVac – Needle0.6970.6335171.1000.690Ethanol – Needle1.1940.6335171.8850.236ApicalPaper Point – EndoVac-0.0150.633517-0.0241.000Paper Point – Ethanol-3.1540.633517-4.979< 0.001Paper Point – Needle0.3060.6335170.4820.963EndoVac – Ethanol-3.1390.633517-4.955< 0.001EndoVac – Needle0.3210.6335170.5070.958Ethanol – Needle3.4590.6335175.462< 0.001* Identical standard errors reflect equal cell sizes and the homogeneous variance structure assumed by the linear mixed-effects model. Degrees of freedom were estimated using the Kenward–Roger approximation Pairwise comparisons of drying procedures within each root section based on estimated marginal means (Tukey-adjusted) * Identical standard errors reflect equal cell sizes and the homogeneous variance structure assumed by the linear mixed-effects model. Degrees of freedom were estimated using the Kenward–Roger approximation Sensitivity analysis Because several subgroup means were near zero and standard deviations exceeded the mean in selected conditions, a sensitivity analysis using square-root transformed bond-strength values was conducted to evaluate the robustness of the mixed-effects findings. The transformed analysis demonstrated findings broadly consistent with the primary analysis (Table 6). Treatment type, drying procedure, root section, and higher-order interactions remained statistically significant. However, unlike the primary analysis, sealer type demonstrated a statistically significant main effect (P = .001), with AH Plus showing higher bond strength than AH Plus Bioceramic in the coronal and middle sections (Table 7). Table 6Sensitivity analysis using square-root–transformed push-out bond strength values (Type III tests of fixed effects from the linear mixed-effects model)Effectdf (num, den)F P Treatment type1, 52833.04< 0.001Sealer type1, 52810.490.001Drying procedure3, 5284.110.007Root section2, 52882.78< 0.001Treatment × Sealer1, 5281.180.278Treatment × Drying3, 5289.46< 0.001Sealer × Drying3, 5283.780.010Treatment × Root section2, 5285.420.005Sealer × Root section2, 5280.970.379Drying × Root section6, 5285.62< 0.001Treatment × Sealer × Drying3, 5283.620.013Treatment × Sealer × Root section2, 5285.120.006Treatment × Drying × Root section6, 5284.12< 0.001Sealer × Drying × Root section6, 5283.050.006Treatment × Sealer × Drying × Root section6, 5283.360.003Degrees of freedom were estimated using the Satterthwaite approximation. Statistical significance was set at P < .05 Sensitivity analysis using square-root–transformed push-out bond strength values (Type III tests of fixed effects from the linear mixed-effects model) Degrees of freedom were estimated using the Satterthwaite approximation. Statistical significance was set at P < .05 Table 7Pairwise comparisons between sealer types within root sections in the square-root sensitivity analysisRoot sectionComparisonMean differenceSEdft P CoronalAH Plus – AH Plus Bioceramic0.310.115172.760.006MiddleAH Plus – AH Plus Bioceramic0.230.115172.030.043ApicalAH Plus – AH Plus Bioceramic0.090.115170.810.416Degrees of freedom were estimated using the Kenward–Roger method Pairwise comparisons between sealer types within root sections in the square-root sensitivity analysis Degrees of freedom were estimated using the Kenward–Roger method Failure mode analysis Failure mode distributions were analysed using the Fisher–Freeman–Halton exact test under fixed drying protocol and root section conditions. Overall, cohesive and mixed failure modes predominated across most experimental groups, whereas adhesive failures were less frequent (Fig. 3). Distribution of failure modes according to treatment type, drying protocol, and sealer type Paper point No significant differences were observed between AH Plus and AH Plus Bioceramic within either treatment group at any root level (all P > .05). When treatment types were compared: Apical section (AH Plus): A significant difference was observed (P = .041), with mixed failures predominating in the primary group and cohesive failures (66.7%) in the retreatment group. Middle section (AH Plus Bioceramic): A significant difference was detected (P = .024), with adhesive failures present only in the primary treatment group. EndoVac No significant differences were observed between sealers at any root level (all P > .05). Between treatment types: Apical section (AH Plus): Significant differences were observed (P < .05), with cohesive failures predominating in retreatment specimens. Middle section (AH Plus Bioceramic): A significant difference was detected (P = .024), with cohesive failures predominating in the retreatment group. Ethanol No significant differences were found between sealers within the same treatment group (all P > .05). Between treatment types: Apical section (AH Plus): Cohesive failures were more frequent in retreatment specimens (P < .05). Middle section (AH Plus Bioceramic): Mixed failures predominated in the primary group (P < .05). Coronal section (retreatment): A significant difference between sealers was observed (P = .013). Irrigation needle No significant differences were observed between sealers at apical and middle levels (P > .05). Apical section (AH Plus): Significant differences between treatment types were observed (P < .05). Middle section (AH Plus Bioceramic): A significant difference between treatment types was detected (P < .05), with mixed failures predominating in the primary group. Coronal section (primary treatment): A significant difference between sealers was detected (P = .042), with fewer adhesive failures in the AH Plus group. Discussion The present study evaluated the influence of different drying protocols on the bond strength of epoxy resin– and calcium silicate–based sealers under both primary and retreatment conditions. Overall, the findings partially rejected the null hypotheses, as bond strength was significantly affected by treatment type, root section, and drying protocol, while sealer type alone did not exhibit a significant independent effect. One of the most prominent findings of this study was the strong influence of root section on bond strength, with values increasing progressively from the coronal to the apical region. This effect demonstrated the largest magnitude among all factors (η²p = .20), indicating that anatomical location plays a dominant role in bonding performance. This finding is consistent with previous studies reporting improved adaptation and mechanical interlocking of filling materials in the apical region [16–18], which may be attributed to increased frictional resistance [19], and differences in dentinal tubule density and orientation [20]. In contrast, the coronal region, with its wider and oval anatomy, and greater sealer thickness formed owing to the single cone obturation technique [21], may present less favourable conditions for stable adhesion to dentin. The finding of higher bond strength values in the apical region contrasts with several previous studies reporting reduced adhesion due to lower dentinal tubule density [22–24]. It should also be noted that dentinal tubule density and sealer penetration do not necessarily correlate directly with bond strength. Previous studies have demonstrated that deeper sealer penetration into dentinal tubules does not always result in improved adhesion, suggesting that mechanical factors and interfacial adaptation may play a more dominant role in bond strength outcomes [16, 19, 25, 26]. Therefore, the increased bond strength observed in the apical region in the present study likely reflects a combination of geometric, mechanical, and procedural factors rather than dentinal tubule density alone. The effect of treatment type was also substantial (η²p = .10), with retreatment generally resulting in higher bond strength values compared with primary treatment, particularly in the apical region. This finding may initially appear counterintuitive, as retreatment procedures are often associated with alterations in dentin structure, residual filling materials, and potential obstruction of dentinal tubules [27, 28]. However, the observed increase in bond strength may be explained by the creation of a more irregular dentin surface during retreatment, which could enhance micromechanical retention [29]. Additionally, remnants of previous filling materials may act as a coupling layer, potentially contributing to improved interfacial adaptation in certain regions [30]. Another important factor that may have contributed to the increased bond strength observed in the retreatment groups, particularly in the apical region, is the difference in apical preparation size between the treatment protocols. In the present study, root canals in the primary treatment group were prepared up to ProTaper Next X3, whereas retreatment specimens were further enlarged to X4. This additional enlargement may have enhanced the removal of filling materials and smear layer, and improved irrigant penetration, resulting in a cleaner dentin surface and greater accessibility of dentinal tubules [31, 32]. Moreover, increased apical enlargement may facilitate deeper sealer penetration and improved mechanical interlocking, especially in the apical region where anatomical constraints typically limit material adaptation. This may partly explain why the effect of retreatment on bond strength was most pronounced at the apical level. Similar observations have been reported in previous studies demonstrating that greater apical preparation sizes can improve debris removal and dentin cleanliness [33–35]. Another factor that may have influenced the bond strength outcomes in the present study is the absence of gutta-percha solvents during retreatment procedures. Chemical solvents such as chloroform, eucalyptol, or orange oil are commonly used to facilitate the removal of filling materials; however, previous studies have shown that these agents may alter the dentin surface, leave residual films, and negatively affect the interaction between sealers and dentin [36–38]. In the present study, no solvent was used during retreatment, and filling removal was performed mechanically. This approach may have contributed to preserving the integrity of the dentin surface and preventing the formation of residual layers that could interfere with sealer adhesion. Consequently, the higher bond strength values observed in the retreatment groups—particularly in the apical region—may partly be attributed to the absence of solvent-related adverse effects. Although drying protocol exhibited a statistically significant effect on bond strength, its magnitude was relatively small (η²p = .03) compared with other factors. Nevertheless, its interaction with root section and treatment type suggests that its role is context-dependent. In particular, drying methods appeared to exert a more pronounced influence in the apical region, where ethanol-based drying demonstrated distinct behaviour compared with other protocols. This may be attributed to the ability of ethanol to displace residual moisture [39], especially in the apical region where moisture control is more challenging [15, 40]. These findings are in agreement with previous reports suggesting that optimal moisture conditions are critical for maximizing sealer–dentin interaction, particularly for materials with different hydrophilic or hydrophobic properties [41, 42]. The use of ethanol as an irrigation agent may influence dentin structure by removing water from the collagen network, leading to dehydration. Excessive dehydration has been associated with collagen collapse, which may hinder the infiltration of filling materials into dentin [43]. Previous studies [40, 44] have described high-concentration ethanol protocols (e.g., 95%) as producing “dry” dentin conditions. Both overly wet and overly dry dentin substrates have been reported to impair adhesion [40]. In the present study, the use of 70% ethanol, which contains a higher proportion of water compared to absolute ethanol, may have resulted in a more balanced moisture condition rather than complete dehydration. This may have helped preserve or allow partial re-expansion of the collagen network, potentially facilitating better sealer adaptation and contributing to the higher bond strength values observed under certain conditions [43]. However, this interpretation should be considered with caution, as the interaction between dentin moisture and sealer properties is complex and may vary depending on material characteristics. In contrast, sealer type did not demonstrate a significant main effect on bond strength, indicating that both epoxy resin–based and calcium silicate–based sealers performed comparably under the conditions tested. However, the presence of significant interaction effects involving sealer type suggests that its performance may depend on the specific combination of drying protocol, treatment condition, and root section. This finding aligns with the literature indicating that the bonding behaviour of epoxy resin– and calcium silicate–based sealers is highly dependent on environmental and procedural factors rather than material composition alone [3, 6, 7, 15, 16, 45, 46]. Failure mode analysis further supported the bond strength findings, with cohesive and mixed failures predominating across most groups, indicating adequate adhesion between the sealer and dentin. The relatively low frequency of adhesive failures suggests that the sealer–dentin interface was generally stronger than the internal strength of the filling material. Notably, retreatment groups exhibited a higher prevalence of cohesive failures in several conditions, which may reflect improved interfacial adaptation or altered stress distribution following retreatment procedures. The present study has several limitations that should be considered. First, as an in vitro investigation, it does not fully replicate the complex biological and mechanical conditions of the clinical environment. Factors such as intraoral temperature fluctuations, occlusal loading, and long-term aging were not simulated. Second, although a standardized methodology was employed, variations in dentin structure among specimens may have influenced the results. In addition, although the mixed-effects model accounted for clustering within teeth, the use of sectioned specimens may still introduce inherent dependencies that cannot be entirely eliminated. The markedly higher bond strength values observed in the apical region should also be interpreted with caution. Although push-out values were normalized according to the calculated adhesion surface area, narrower apical geometries may increase frictional resistance and wedging effects during dislodgement testing. Therefore, part of the measured resistance may reflect mechanical interlocking and geometric constraints rather than purely adhesive interactions. This limitation is inherent to push-out testing methodologies and may contribute to disproportionately higher bond strength values in small-diameter apical slices. An additional consideration relates to the distributional characteristics of push-out bond-strength measurements. Several subgroup means were close to the lower measurement boundary (0 MPa), with standard deviations exceeding the mean in selected strata, suggesting skewness and floor effects. Sensitivity analysis using square-root transformed values generally supported the primary findings; however, differences related to sealer type became statistically significant after transformation. Therefore, conclusions regarding sealer-specific performance should be interpreted cautiously, as they may depend on modelling assumptions. Future studies should focus on evaluating the long-term stability of the sealer–dentin interface under thermomechanical aging conditions and exploring the interaction between drying protocols and newer-generation sealers with enhanced bioactivity. Additionally, advanced imaging techniques such as micro-computed tomography or confocal microscopy could provide further insight into sealer penetration and interfacial adaptation at different root levels. Within the limitations of this study, it can be concluded that bond strength of root canal sealers is predominantly influenced by root section and treatment condition, with drying protocol playing a secondary but context-dependent role. Retreatment procedures were associated with higher bond-strength values, particularly in the apical region, while the choice of sealer appears less critical when considered independently. Supplementary Information Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Acknowledgements Not applicable. Abbreviations ProTaper Next Sodium hypochlorite Ethylenediaminetetraacetic acid AH Plus AH Plus Bioceramic Estimated marginal mean Authors' contributions DB.Y conceived the study and collected and analysed the data. B.O drafted the manuscript and contributed to the literature search and visualisation. All authors read and approved the final manuscript. Funding This study was funded by Baskent University Research Fund (Project no: D-KA24/05). Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate This study was approved by Baskent University Institutional Review Board and Ethics Committee (Project no: D-KA24/05), according to the principles of The World Medical Association Helsinki Declaration. Informed written consent forms were obtained from each patient whose extracted teeth were used in the study. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Supplementary Material 3. Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.