Bond strengths for orthodontics are variable and depend on a multitude of factors, such as the type of adhesive used, the design of the bracket base, the type of substrate bonded, the system of forces applied by the appliance and the technique of the practitioner [2, 11, 13, 14].

The aim of this study was to evaluate the SBS and failure modes of metal orthodontic brackets to glazed lithium disilicate reinforced glass-ceramic and zirconia according to various surface treatment protocols, including sandblasting, the use of a universal primer alone or with hydrofluoric acid etching and four adhesive systems.

In our study, for LD, sandblasting significantly improved the SBS values for groups with adhesive systems tested alone or with adhesive systems combined with MP alone. These results agree with those of previous studies: sandblasting increases surface roughness, creates micropores and exposes silica oxides present in ceramics, enhancing the micromechanical interlocking between the exposed surface oxides and the bonding resin [1, 9, 10, 13,14,15,16,17,18].

For LD, except for the OFL adhesive, the groups that received HF + MP with and without sandblasting were significantly different from the same groups that did not receive HF etching. Our results agree with the existing consensus on the beneficial role of surface pretreatment by HF etching prior to bonding to lithium disilicate glass ceramics. HF etching removes the glaze and partially dissolves the glass phase of the ceramic to create surface micropores favorable to the micromechanical anchoring of the bonding material [1,2,3,4,5, 15,16,17]. However, our results did not reveal a significant difference in the SBS values between the groups that were subjected to both sandblasting with HF etching and MP and the same groups that were not subjected to sandblasting. These findings are consistent with the results of other scholars: HF etching and sandblasting are both micromechanical treatments with similar effects on increasing the surface roughness of lithium disilicate glass ceramics [1, 5, 16, 17].

For the Zr groups, sandblasting significantly increased the SBS values only for the groups in which adhesive systems were tested compared with those of the same groups without sandblasting. Sandblasting alumina or silica particles enhances the micromechanical anchorage of the bonding material on zirconia and its wettability by creating surface irregularities and increasing the roughness [1, 9, 10, 17, 18]. Combining HF etching with MP application does not seem to improve the SBS values for sandblasted zirconia samples compared with those of the same groups without sandblasting. This finding is consistent with other studies. Zr is a polycrystalline ceramic and does not contain a glass phase; therefore, HF etching does not have a significant effect on enhancing bonding to this ceramic [1, 9, 10, 17, 18].

Surprisingly, without sandblasting, the combination of HF etching and MP application resulted in significantly higher SBS values than those of the same groups with MP application alone; this is contrary to the results reported in the literature, which do not demonstrate the effectiveness of HF on zirconia surfaces [1, 9, 10, 17, 18]. However, unlike our study, polishing samples were used, which could explain the difference observed with our results [1, 9, 10, 17, 18]. In our study, the Zr samples were covered by glazing composed of a glassy phase [1,2,3]. Several authors, using glazed zirconia samples, reported a significant effect of HF, which is consistent with our results [19, 20]; this could suggest an attack of the glaze by HF, increasing the surface roughness suitable for mechanical and chemical anchorage of the bonding resin to zirconia.

For bonding to ceramic restorations, a primer is applied after sandblasting or HF etching to promote adhesion between the ceramic restoration and the bonding resin [2]. Recently, to simplify bonding procedures, universal primers have been developed to be used universally for both metals and ceramics, such as zirconia and glass ceramics, in contrast to specific primers [21,22,23,24,25]. In our study, adding a universal primer (MP) increased the SBS values for both LD and Zr compared with those of the different adhesive systems tested alone. Monobond Plus, the universal primer used in our study, contains a silane (3-(trimethoxysilyl) propyl methacrylate (3-MPTS)), a sulfide methacrylate and 10-MDP. The manufacturer’s instructions for the use of this product explain that these components form bonds between the silane and glass ceramics and between 10-MDP and zirconia, improving the bond strength of these two types of ceramic. Moreover, in our study, for LD, the application of the universal primer tested (MP) after sandblasting the surface significantly increased the SBS values obtained. For Zr, the addition of MP after sandblasting also increased the SBS values, although not significantly. These findings are consistent with previous studies showing that universal primers are more effective or as effective as specific ceramic primers when applied after sandblasting [1,2,3, 5,6,7,8,9,10, 21,22,23,24,25]. Sandblasting enhances the microroughness of the ceramic surface and the wettability of primers suitable for chemical bonding to LD or Zr [21,22,23,24,25,26,27].

In addition, except for the TXTP adhesive, HF etching + MP without sandblasting resulted in a significant difference compared with the groups with MP alone for both LD and Zr. Similar interactions to those observed between the silica atoms of LD and the silane in the primer could occur with the silica atoms contained in the glaze and the silane in the universal primer, which could explain the increase in SBS values obtained. This interaction is facilitated by HF etching, which can dissolve the glaze, creating micropores and increasing the wettability of the universal primer at the surface. This finding was also demonstrated by Kwak et al., who suggested applying a silane to glazed zirconia surfaces after HF etching [20].

Two universal adhesives were used in our study: Scotchbond Universal, which contains silane and 10-MDP, and Adhese Universal, which contains 10-MDP without silane. With respect to SBS values, universal adhesive applied alone after sandblasting yielded satisfactory values above the orthodontic threshold defined by Reynolds et al. [11] for both LD and Zr, and no significant difference was noted between the two universal adhesives used. Sandblasting before universal adhesive application increases surface roughness for micromechanical anchorage and improves the wettability of the adhesive [1, 2, 8, 15, 28,29,30].

For LD, the groups with HF etching + MP and SU or AU presented the highest SBS values compared with those of the two other adhesives, with no difference with or without surface sandblasting. These findings confirm the results of previous studies, which demonstrated a better efficacy of universal adhesives after sandblasting and/or HF etching followed by the application of silane alone or a universal primer containing silane, such as Monobond Plus [1, 2, 8, 15, 29,30,31].

For the sandblasted samples with MP alone and with SU or AU on Zr, our results are in accordance with the literature. Some scholars have demonstrated the efficacy of universal adhesives after sandblasting the surface due to the chemical bonding of 10-MDP with the zirconia surface and an increase in the bond strength if the universal adhesive is combined with a zirconia primer or a universal primer [6, 16, 22, 28]. The authors explained the lower results obtained with the universal adhesive alone than with the addition of a primer because the silane in universal adhesives is less stable than that contained in primers is, which may affect the chemical bonding of 10-MDP to the ceramic surface, particularly in the long term [31, 32].

Adhesion values must be high enough to withstand the orthodontic forces applied during the treatment but must allow the debonding of the brackets at the end of the treatment without damaging the supporting material (enamel or, in this case, prosthetic restoration) [2, 13, 14]. Our results show fractures of ceramic restorations (ARI 4) only for groups combining the two universal adhesives tested with HF + MP for the LD samples. With this surface treatment, this implies a significant risk of ceramic restoration damage during the removal of the fixed appliance that is not suitable, especially when restorations are in the anterior zone or are recently made. Our results are consistent with those reported in the literature, wherein authors demonstrated higher incidences of ceramic fractures, especially for lithium disilicate with universal adhesives, than for conventional adhesives, suggesting better anchoring of the bonding material due to the chemical bonding provided by these adhesives [1, 9, 10, 13, 29]. Our results also demonstrated ARI scores of 0 and 1, indicating that no or less than 50% of the adhesive remained on the ceramic surface after debonding, for all the groups employing OFL and TXTP and for the groups employing the two universal adhesives alone, alone with sandblasting or combined with MP alone. It was demonstrated that ARI scores of 0 or 1 are suitable because they allow easier cleaning of the restoration surface during debonding, which limits scratching of the ceramic surface with the bur [2, 13, 14].

Regarding our results concerning SBS values and ARI scores, as well as the use of hydrofluoric acid—which poses a risk for both the patient and the practitioner (such as burns in the case of contact with skin and mucous membranes)—combining a universal primer (Monobond Plus in our study) with a universal adhesive seems to be the best bonding protocol for both Zr and LD restorations, which would avoid the use of this dangerous acid.

Sandblasting before the application of the universal primer followed by the use of a universal adhesive enhances the bond strength, mainly for LD, but several authors have demonstrated that it can damage the ceramic surface by creating surface microdefects [2,3,4]. Our findings suggest that this step could also be avoided. It could be interesting to simplify the bonding procedures and the inventory assortment by using the same primer for both restoration types.

The in vitro design of this study cannot simulate all the parameters of the oral environment and therefore does not provide a complete understanding of the clinical behavior of all the materials tested (e.g., presence of saliva, effects of occlusal forces and those delivered by orthodontic appliances, effects of mastication and hygiene procedures).

Shear bond strength tests were used in our study. Compared with other tests, such as tensile strength tests, SBS tests provide not only shear stresses but also tensile and compressive stresses during interfacial analysis, which increase SBS values [33,34,35]. Additionally, during SBS tests, shear force is applied not only to the adhesive/composite joint but also to all the adhesive/joint/resin composite assemblies, which are stressed [33,34,35]. Universal adhesives form a thinner adhesive layer than do conventional adhesives, such as TXTP or OFL. The thickness of the composite resin layer is increased, and consequently, increased mechanical properties of the adhesive/composite layer are achieved [33,34,35]; this would lead to a bias in the SBS values obtained for the two universal adhesives tested, which are higher than those of the other groups.

Furthermore, the actual timeline of orthodontic treatment (approximately two to three years) was not simulated in this study, as shear bond strength testing was performed one week after bonding. It could be interesting to conduct these tests over extended periods or to use thermocycling to induce mechanical fatigue and simulate the effect of intraoral aging of bonding materials to approach clinical conditions [16, 23, 36]. However, in the clinic, immediate bond strength is also important because the archwire is engaged in the brackets, exerting forces immediately after the bonding procedure.

Therefore, additional clinical studies