CAD/CAM-produced zirconia cube specimens (48 in total, Aidite Technology, Qinhuangdao, China) with a length of 2 cm were acquired for the study. These zirconia samples were randomly divided into six experimental groups based on the use of specific bonding agents:

• Group 1 (XT): Transbond™ XT Light Cure adhesive paste (XTL) + Transbond™ XT Light Cure Orthodontic Adhesive Primer (XTP)

Please refer to Table 1 for detailed information on the chemical composition and application procedures of the bonding agents and cleaning paste used in this study.

The zirconia cube specimens in each group were divided into two subgroups, each accommodating two types of brackets: metal brackets (Victory Series, 3M Unitek, United States) and ceramic brackets (Maia Series, Protect, Zhejiang, China). A total of 216 metal brackets and 216 ceramic brackets were utilized. The mean base surface areas of the metal and ceramic brackets were 11.94 mm² and 14.82 mm².

To prepare the bonding surfaces on the zirconia cube, sandblasting with 125 μm aluminum oxide particles was conducted for 60 s at a distance of 1 cm and 2.8 bar pressure. Subsequently, all specimens underwent ultrasonic cleaning in distilled water for 5 min and gentle air drying for 15 s. The application of Ivoclean (IVO) followed the manufacturer’s instructions.

The same pressure (5 N) was applied to each bracket using a consistent clamp across all groups. The application procedures for light curing (Kerr Demi Plus, Orange, CA, United States) and cement removal were standardized and detailed in Table 1.

Each group comprised a total of 72 bracket specimens, divided into two subgroups (metal and ceramic brackets) with 36 specimens each.

Half of the brackets in each subgroup were tested after storing in distilled water at 37°C for 24 h, while the rest underwent 5,000 cycles of thermocycling between 5°C and 55°C in a thermocycling device (SD Mechatronik, Feldkirchen-Westerham, Germany). The SBS test was conducted using a universal testing machine (WD-200 Weidu, Wenzhou, China) with a crosshead speed of 1 mm/min (Figure 1). The SBS formula used was P (MPa) = F (N)/S (mm²). Out of 18 results in each group, the highest 4 and lowest 4 data points were discarded, and the remaining 10 were considered for analysis (n = 10).

The ARI scoring was performed based on the amount of remaining cement on the zirconia surface with the help of a dental digital camera (EyeSpecial C-IV, shofu, Koyoto, Japan). The score is represented by a scale with 5 levels (Score A to Score E) as follows:

Score A: almost all the cement remained on the zirconia surface;

Score B: more than 90% of the cement remained on the zirconia surface;

Score C: more than 10% but less than 90% of the cement remained on the zirconia surface;

Score D: less than 10% of the cement remained on the zirconia surface;

Score E: no cement remained on the zirconia surface.

Images were scored by three calibrated examiners, and a majority opinion was adopted in cases of disagreement. Figure 2 demonstrates the ARI score on the samples.

Statistical analysis involved three-way ANOVA (bonding agents, storage conditions, and brackets) and the Games-Howell test using SPSS version 26.0, with a significance level of α = 0.05.

The debonding bracket base surface was analyzed using field emission scanning electron microscopy (FE-SEM) and energy dispersive X-ray spectrometry (EDS; Phenom Pharos G2, Netherlands) to determine elemental composition and distribution.

The SBS mean values and standard deviations for each group are summarized in Table 2. Notably, RSBU, SBPM, and GMP exhibited higher SBS for metal bracket groups after 24 h of water storage, with no significant difference among these groups (p > 0.05). However, RSBU’s SBS decreased significantly after thermocycling, while SBPM and GMP remained unchanged.

For ceramic brackets, SBPM displayed higher SBS after 24 h of water storage but significantly declined post-thermocycling, akin to other bonding agents (p < 0.05). Comparatively, ceramic brackets demonstrated a more significant decrease in SBS after thermocycling in contrast to metal brackets. Generally, the SBS of ceramic brackets was lower than that of metal brackets after both 24 h of water storage (p > 0.05) and thermocycling (p < 0.05), except for XTS group results (p > 0.05) post-24 h of water storage and XT group outcomes under both storage conditions (p < 0.05).

SBPM showcased higher SBS values after thermocycling, irrespective of bracket type, while XT exhibited the lowest SBS in the metal bracket group. XTS displayed a significant increase in SBS for metal brackets post-thermocycling (p < 0.05).

The SPSS analysis (Table 3) revealed significant influences on SBS values by brackets, bonding agents, and storage conditions (p < 0.001). Interactions were observed among brackets, bonding agents, and storage conditions (p < 0.001).

Regarding the SBS results in Tables 4, 5, metal brackets displayed significantly higher SBS than ceramic brackets after thermocycling (p < 0.05). Also, RSBU, RGBU, SBPM, and GMP exhibited significantly higher SBS after 24 h of water storage compared to XT and XTS (p < 0.05). XT demonstrated notably lower SBS than other groups post-thermocycling (p < 0.05). SBPM consistently demonstrated higher SBS irrespective of storage conditions.

The distribution of failure modes for metal and ceramic bracket groups are visualized in Figures 3, 4, respectively. Under 24 h water storage, ARI scores for various groups were concentrated within different categories. After thermocycling, shifts in ARI scores were observed across groups.

For ceramic brackets under both conditions, XT showed a tendency toward E in ARI scores, while GMP and SBPM exhibited a notable increase in A and a decrease to 0 in C after thermocycling.

Typical bracket fractures were observed in ceramic bracket groups after 24 h water storage, except for XT, predominantly in bracket wings with fewer base fractures in RGBU and GMP. After thermocycling, fractures were observed only in SBPM and GMP, with SBPM showing bracket base fractures. Figure 5 displays the two modes of bracket fractures.

FE-SEM images (Figure 6) and EDS results (Figures 7, 8) indicated the presence of Si on bracket bases of XT, XTS, RSBU, RGBU, and GMP groups. Zr particles were detected in RSBU and RGBU metal bracket bases. SBPM showed Zr particles in both ceramic and metal bracket bases. Fe and Al were detected on metal bracket bases, whereas Fe was absent on ceramic bracket bases.

These observations provide insights into the elemental composition and characteristics of the bracket bases across different bonding agents and brackets used in the study.

The study aimed to evaluate the bond strength and durability of six bonding agents with metal or ceramic brackets on zirconia. Table 5 highlighted significant differences in shear bond strength (SBS) among bonding agents, rejecting the null hypothesis that no significant difference exists in the bond strength of different bonding agents. Typically, for optimal orthodontic efficacy, the bond strength of orthodontic brackets should ideally be within the range of 6–8 MPa at least (Uysal et al., 2004; Tecco et al., 2005). According to Table 5, except for the XT group, all metal bracket groups achieved long-term bond strengths deemed clinically acceptable. In contrast, among the ceramic bracket groups, while acceptable strength was achieved in the short term by all groups, only the SBPM group exhibited sustained durability.

Using XT as a control group for bonding brackets to enamel, it demonstrated higher SBS in ceramic brackets after 24 h of water storage compared to metal brackets. This outcome suggests the potential influence of ceramic translucence in enhancing XT resin polymerization, contributing to improved short-term bonding (Al-Hity et al., 2012; Reginato et al., 2013). However, XT’s SBS significantly decreased after thermocycling, potentially due to mismatched coefficients of shrinkage/expansion between XT resin and brackets, failing to meet the required clinical bond strength.

The XTS group, employing SBU, an universal adhesive, exhibited improved bonding performance attributed to 10-MDP’s presence, facilitating chemical bonding to zirconia via phosphate groups (Khan et al., 2017; Nagaoka et al., 2017). This bond was indicated by ARI scores concentrated on A, consistent with SBU’s ability to strengthen XTL resin-zirconia bonds.

The absence of HEMA in GBU aimed to evaluate HEMA’s impact on bracket-zirconia bonding. RSBU (containing HEMA) for metal brackets demonstrated significantly higher SBS than HEMA-free RGBU after 24 h water storage. HEMA’s role in enhancing component miscibility and forming a uniform adhesive layer could explain RSBU’s superiority (Toledano et al., 2001; Moszner et al., 2005; Van Landuyt et al., 2005), despite its decreased SBS post-thermocycling. However, RGBU’s hydrophobic nature prevented water absorption, resulting in more stable bonding durability, aligning with earlier findings (Hu et al., 2022).

Studies have shown that HEMA-containing adhesives, due to continuous water absorption, lead to decreased bond strength post-polymerization (Ito et al., 2005; Takahashi et al., 2011). RSBU demonstrated significant SBS reduction after 5,000 cycles of thermocycling, attributed to water absorption. Conversely, RGBU’s SBS remained stable in metal brackets group, indicating its resistance to water-induced degradation. Therefore, Three-way ANOVA demonstrated a statistically significant interaction among bonding agents, storage conditions and brackets (p < 0.001), which was not detected in previous experiment (Hu et al., 2022).

GMP, containing a self-adhesive resin cement, showed comparable SBS to SBPM in the metal bracket group after 5,000 thermocycles, suggesting GMP’s stability post-thermal stress.

SBPM exhibited significantly higher SBS for ceramic brackets, despite a notable 60% bracket breakage rate. This could be attributed to MMA resin’s water resistivity (Ikemura and Endo, 2010; Aoki et al., 2011) and the absence of silicon components in EDS analysis. Due to the absence of inorganic fillers, SBPM is typically polymerized in a linear form, resulting in better toughness compared to cross-linked polymers. On the other hand, SBPM can promote free radical polymerization of the resin using oxygen and water in the presence of TBB, a polymerization initiator, which can significantly improve the bond strength and long-term durability of ceramic brackets to zirconia porcelain (Tanaka et al., 2004; Meguro et al., 2006; Shinagawa et al., 2019). The presence of Zr in the EDS results also reveals the viewpoint. This finding is consistent with those of Shimoe et al., who also observed an increase in the [C] and [O] intensity peaks when the zirconia surface was treated with 4-META, indicating that 4-META chemically adheres to the zirconia surface effectively (Shimoe et al., 2018).

While most groups demonstrated decreased SBS after thermocycling, XTS displayed a significant SBS increase, possibly due to radical mobility enhancement in high-temperature conditions during resin solidification (Al Jabbari et al., 2014). This variance partially rejected the null hypothesis of storage conditions not affecting bonding agents’ strength.

Metal and ceramic brackets showcased comparable SBS after 24 h water storage, but ceramic brackets exhibited significantly lower SBS post-thermocycling, rejecting the null hypothesis of similar bonding durability between metal and ceramic brackets on zirconia.

Limitations include in vitro conditions not entirely mimicking oral temperatures, potentially impacting XTS’s clinical bond strength, and bracket fractures affecting measured SBS accuracy in ceramic brackets. Meanwhile, the study solely applied sandblasting treatment to zirconia surfaces, without comprehensive assessment of other factors capable of zirconia surface modification, such as laser treatment (Hou et al., 2020) and silica coating (Galvão Ribeiro et al., 2018), which have been demonstrated to have a positive impact on the bonding efficacy between zirconia and resin. The subsequent phase of research may encompass the effects of diverse surface modification treatments between orthodontic brackets and zirconia and investigate their clinical practicality.