In vitro performance of an autocured universal adhesive system in bonding to dentin

A total of 25 caries-free human third molars were collected from Tokyo Medical and Dental University hospital according to the protocol approved by the Institutional Review Board of Tokyo Medical and Dental University, Tokyo, Japan (D2013-022–02). All included teeth were extracted for reasons unrelated to the study and are so-called excess material. Inclusion criteria for teeth selection were the absence of caries and complete coronal and root development and the absence of root fracture and resorption, endodontic treatment. The teeth were stored at 4 °C in a0.1% thymol solution.

Sample preparation was performed as shown in Fig. 1. The roots were cut at the cementoenamel junction, and two dentin disks that were 2-mm-thick were cut for each tooth using a diamond blade (IsoMet, Buehler, Lake Bluff, IL, USA). The dentin disks were polished with water-resistant abrasive paper with grit levels reaching #600 under running water to flatten the surface. Fifty dentin disks were randomly assigned to one of the following 5 groups of 10 dentin disks each. All the materials were used in this study according to the manufacturers’ instruction. The materials used in this study are listed in Table 1 and below.

We used a combination of a bonding system and a composite resin from the same manufacturer. For autocured bonding materials, BE and BO group were created. For light-cured bonding materials, PS, SF and GG group were created.

For sample preparation, Tygon tubes (R-3603, Norton Performance Plastics) of 0.8 mm in diameter and 1.0 mm in height were cut, two per disk were placed on the dentin disks with adhesives to fill the corresponding composite resin inside the tube space. As a result, two small resin cylinders of 0.8 mm in diameter and 1.0 mm in height for μSBS test were bonded to the dentin disks. The composite resin was light cured for 20 s (Valo LED Curing Light, high-power mode at 1,400 mW/cm², Ultradent, South Jordan, UT, USA), and the tube was carefully removed from the resin composite cylinders. The specimens were stored in distilled water at 37 °C for 24 h before testing. The μSBS test was measured using a universal testing machine (EZ-Test-500N, Shimadzu, Kyoto, Japan) at a crosshead speed of 1.0 mm/min. The data obtained were statistically processed using one-way analysis of variance (ANOVA) at a significance level of 5%. After the μSBS test, the fracture surface was observed using SEM (JSM-5310LV, JEOL, Tokyo, Japan) to examine the failure mode. The area of each failure type (%) was calculated, and 70% of the surface was used as the threshold for classification. Failure modes were categorized into one of the following three types: cohesive failure of dentin, adhesive failure at the interface between adhesive and dentin, and mixed failure (combination of adhesive and cohesive failure). The results were statistically processed using the chi-square test. All analyses were performed using the Statistical Package for Social Science (SPSS; Windows, Version 23.0, IBM, Armonk, NY, USA).

Cavity adaptation of each adhesive system was evaluated nondestructively using SS-OCT. The SS-OCT system (IVS-2000, Santec, Komaki, Japan) used in this study was a swept-source OCT. The system was incorporated with a low-coherence near-infrared light source and had the configuration of a Mach–Zehnder-type interferometer. The near-infrared light was produced by a fast sweeping laser that repetitively swept the wavelength from 1260 to 1360 nm (centred at 1310 nm) at a rate of 20 kHz. The axial resolution of this SS-OCT system in air was 11 μm, corresponding to 7 μm in dental tissue with a refractive index of approximately 1.5 [25]. The lateral resolution of the system was ∼17 μm, which was determined by the objective lens at the probe. The probe connected to the interferometer had a power of 5 mW, which was within the safety limit of the American Standard Institute. The laser source emitted from the probe was directed onto the sample at the desired location in the X and Z dimensions. The backscattered light carrying information from each single scan point of the sample was returned to the system, digitized on a time scale, and analysed in the Fourier domain to disclose the depth information (A-scan) of the sample. By combining a series of A-scans in a linear fashion across the sample, a cross-section (B-scan) was obtained. Finally, cross-sectional images could be created by converting the B-scan raw data into a greyscale image with 2001 × 1019 pixels.

Sample preparation was performed, as shown in Fig. 2. A total of 40 sound bovine maxillary incisors were extracted for this study. The root was cut at the cementoenamel junction, and the labial enamel was polished with #1000-grit silicon carbide paper to remove the superficial layer and expose the dentin to form a flat surface. Class I dish-shaped cavities (1 mm in diameter × 1 mm in depth) were prepared using a diamond bar (#144 Diamond Point FG, Shofu, Kyoto, Japan) attached to a high-speed turbine handpiece under water cooling; thus, the margin was on the labial flat surface and the dentin cavity bottom. Two cavities per tooth were formed. For bonding operations, an adhesive system and composite of the aforementioned five groups were applied to the prepared cavities and cured according to the manufacturer's instructions. Then, each composite resin was filled and light-irradiated for 20 s (Valo LED Curing Light, high-power mode at 1,400 mW/cm², Ultradent, South Jordan, UT, USA). The samples were stored in distilled water at 37 °C for 24 h.

The samples were used for SS-OCT observation to capture the 2D images. Each sample was fixed to the microscale head stage of the SS-OCT, and the scanning laser beam was directed perpendicular to the repair surface. The sample was moved in a mid-range direction across the laser beam, and a cross-sectional B-scan image was captured at the maximum depth. When air was in a defect at the restoration–tooth interface, as the light passed between these two mediums with different refractive indices, the light partially reflected from the interface, which appeared as bright areas on the OCT image [13]. In this study, to analyse images, we adopted sealed interface percentage (SI%) parameters to calculate and evaluate the bonding performance levels of the adhesives. 2D SS-OCT raw tomograms were imported to image analysis software (ImageJ version 1.53t), and a median filter was applied to decrease background noise [26]. An experimental threshold determination algorithm developed as a plugin for ImageJ under JAVA was used for image analysis [22]. As shown in Fig. 3, the region of interest (ROI) was selected as a polygon around the whole length of the restoration interface, excluding the specimen surface. The width of the ROI was approximately 80 pixels, and the adhesive interface was placed in the centre of the ROI. The pixel values on each vertical line (A-scan corresponding to 2 pixels in width) in the ROI were ranked by the software plugin. The pixels in the top 10%, which were presented on the same line, were selected. Among them, the pixels bearing higher intensity values that were equal to or greater than the sum of the background noise and median values were designated as target pixels (white); all other pixels were designated as null (black). The total percentage of these white pixels (gap) over the ROI length was automatically measured by the plugin. To obtain the total SI% value, the obtained total gap percentage was subtracted from 100%. Finally, the mean of the SI% values for each specimen was calculated. The data were subjected to normality analysis, and a nonparametric test was selected. The mean SI% values for each adhesive were statistically analysed by multiple comparisons using the Kruskal‒Wallis test. All analyses were performed using SPSS.

After SS-OCT imaging, the cavities were cut perpendicularly to the flat surface using an IsoMet. The cut surface was polished under running water with silicon carbide paper (#600–2000 grit) to the location of the cross section captured by the imaging of SS-OCT, followed by diamond paste to 0.25 μm in grain size. Finally, the polished specimens were observed with CLSM (VK-X 150, Keyence, Tokyo, Japan) at 100 × and 400 × magnification levels.

The means and standard deviations of the μSBS values are shown in Table 2 and Fig. 4a. One-way ANOVA showed that the mean μSBS values were not significantly different among the five groups. The frequencies of failure modes and the results of the analysis are shown in Fig. 5. SEM images of representative failure modes are shown in Fig. 6. The results of the chi-square test showed no significant differences in the frequencies of failure modes among the five groups (p > 0.05). Most of the failure modes occurred at the bond interface between the bond and dentin.

A representative 2D SS-OCT image of the cavity is shown in Fig. 7. Representative confirmatory CLSM images are shown in Fig. 8. Figure 9 shows that the increase in signal intensity at the interface in the 2D OCT cross section corresponded to a gap in the confirmatory CLSM image. Table 3 and Fig. 4b show the results of SS-OCT analysis of SI% for each group. The mean SI% values and their standard deviations for all groups are plotted in Table 3. Conducting the Kruskal‒Wallis test on the OCT data revealed that groups with significantly different SI% values existed, rejecting the null hypothesis of this study (p < 0.05). The BE group of autocured bonding materials showed the highest mean SI% that significantly differed from those of the BO, SF, and GG groups. Although not significantly different, the mean SI% values of the BE and BO groups of the autocured bonding materials were higher than those of the light-cured bonding materials.