Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing

ABSTRACT Objective: This study aims to determine shear debonding strength of metal and ceramic brackets, and the degree of enamel crack healing. Material and Methods: Extracted human maxillary premolars were flattened on the buccal surface, and randomly separated into five groups (n = 15). In control groups (groups 1 and 2), metal and ceramic brackets were bonded on flat polished enamel, while in experimental groups (groups 3 and 4), metal and ceramic brackets were bonded on the surface with boundary where corner cracks were created. Additionally, fifteen specimens (group 5) were also prepared for an indentation procedure with no bracket installation. The degree of crack healing was measured. All brackets were detached with a universal testing machine, and the adhesive remnant index (ARI) was also identified. Healing degree and apparent fracture toughness were then calculated. Results: Between groups with similar bracket types, there was no statistically significant difference in debonding strength. Regarding bracket types, ceramic brackets provided significantly higher debonding strength than metal brackets. There was a significant difference in ARI scores between metal and ceramic brackets. The corner cracks showed signs of healing in both horizontal and vertical directions. No statistically significant difference in the healing rates among the groups was found and the apparent fracture toughness increased from the initial to the final measurement. Conclusions: Within the limitations of this study, even though ceramic brackets required significantly higher debonding force compared to metal brackets, debonding stress was limited to the bonding site and did not affect the surrounding cracks on enamel.


INTRODUCTION
Enamel cracks may be a consequence of several factors, including abnormalities in the maturation process, occlusal overloading, temperature variations, therapeutic procedures, and surface injuries from bracket removal - especially with the use of ceramic brackets. 1 Several studies have determined that bonding of ceramic brackets to enamel provided higher bond strength when compared to conventional metal brackets. [2][3][4] Such firm adhesion may cause some degree of micro surface damage in the form of crazing, crack or fracture on the enamel surface when brackets are removed. 5 The enamel micro-defects after bracket removal are of great interest for orthodontists who use fixed orthodontic appliances. 6,7 Presence of cracks may cause stain and plaque accumulation on the enamel and increase the risk for dental caries.
Additionally, propagation of cracks may lead to more surface disintegration and structural loss. 8 However, there is some evidence of enamel microcrack healing as a natural defense to prevent crack propagation to the dentin and to dental pulp. 9 Few studies have evaluated enamel defects before bonding, 1,7 and analyzed the presence of alteration of the control enamel microcracks before and after bonding brackets. Regarding the bracket types, there is a lack of knowledge on the relative microcrack characterization on debonded enamel after brackets removal.
Nimplod P, Tansalarak R, Sornsuwan T -Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing 5 From the fractographical and mechanical aspects, the objectives of this study were to compare debonding strength and degree of crack repair on the debonded enamel after removal of metal and ceramic brackets.

MATERIAL AND METHODS
Seventy five extracted human maxillary premolars were used for this research. These premolars were extracted due to orthodontic indications. These specimens, originated from both genders, between 16 and 40 years of age, were collected from patients at the surgical department in the School of Dentistry, Naresuan University, and private dental clinics, following an ethical approval protocol by the Institutional Review Board of Naresuan University.
All premolars were caries-free, with no existing restorations nor root canal fillings, and with no sign of prominent cracks, abrasion or erosion. After extraction, all specimens were washed in running water to remove all blood and adhered tissue, stored in 0.1% thymol solution and then tested within a month of extraction, to reduce the potential for organic and inorganic losses.
After root separation using a high-speed carborundum disc, the specimens were positioned in a 2-cm diameter plastic ring with the most convex buccal surface of the tooth 2-3 mm above the surface of a self-cured acrylic resin, and then kept in 25°C water for 24 hours, for complete resin polymerization. A series of abrasive papers, with grits P1000, P1200, and 3-µm and Nimplod P, Tansalarak R, Sornsuwan T -Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing 6 1-µm diamond pastes were consecutively used to standardize the curvature of the buccal surface of the teeth. The polishing protocol consisted of the use of a grinder polisher, driven with a 20-Newton force for 20 seconds, to achieve a flat area to bond the bracket base (9.28 ± 0.08 mm 2 for metal and 10.38 ± 0.08 mm 2 for ceramic brackets). The polishing was carried out horizontally relative to the cutting plane of the plastic ring.
All samples were randomly divided into five groups depending on the bracket type and with or without indentations: Before indention making in groups 3, 4, and 5, a four-millimeter-width rectangular barrier was attached to the middle of the polished area to separate the indented from the bonded areas ( Fig 1A). Six micro-indentations were performed close to the edges of the barrier using a microhardness tester with a Vickers Nimplod P, Tansalarak R, Sornsuwan T -Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing 7 diamond indenter (Zwick/Roell; Indentec) loaded with a 500-gram force for 10 seconds (Fig 1B). Three indentations were created at the upper boundary including A, B, and C points from left to right, and another three indentations were created at the lower boundary including D, E, and F points from right to left (Fig 1B). before bracket bonding and also after debonding. By using machine software (Zwick/Roell) to draw measuring lines, which were calibrated with the size and depth of the indentation diagonal, between the indentation's corner and the prominent crack tips, the crack length can be precisely measured.
The illustration of non-indented and indented specimens before bracket attachment is shown in Figure 2. The specimen's unbonded areas were then covered with a barrier tape to avoid adhesive contamination on the microcracks and to control the bonding area ( Fig 1C).The bonded surface of each  A) The first barrier used for separation before indentation, to locate indentation zone at the bracket boundary. B) Indentation making (nomenclature of the indented microcracks performed at the boundary according to their directions: A 1 , B 1 , C 1 , D 1 , E 1 , and F 1 (centripetal vertical cracks); A 2 , B 2 , C 2 , D 2 , E 2 , and F 2 (centrifugal vertical cracks); A 3 , B 3 , C 3 , D 3 , E 3 , and F 3 (clockwise horizontal cracks); A 4 , B 4 , C 4 , D 4 , E 4 , and F 4 (counterclockwise horizontal cracks). C) The second barrier used for protection of the indentations from resin infiltration before bracket attachment. specimen was prepared by etching with 37% phosphoric acid (3M Unitek) for 30 seconds, followed by 15-second water rinsing and 10-second drying with oil-free compressed air. The  In addition, the crack length was used to analytically calculate the apparent fracture toughness (K c(app) ) for each indentation according to the following equation: 11 (3) where HV, F, L, and c are the Vickers hardness, indentation load, average diagonal length, and crack length, respectively.
The elastic modulus (E) for enamel was obtained elsewhere. 12 Standard descriptive statistics means and standard deviations were calculated for all parameters. The difference in the debonding strength and degree of crack alteration was compared statistically using a Kruskal-Wallis test. One-way analysis of variance (ANOVA) was used to determine the difference of healing rate of the microcracks and the apparent fracture toughness between groups. Any differences were further investigated using the post-hoc test. A statistically significant level was predetermined at 0.05 for all tests.

RESULTS
A comparison of the debonding strengths within groups of similar bracket type and between metal and ceramic bracket groups is presented in Table1. The median debonding strengths of metal and ceramic groups were 23.06 MPa and 37.37 MPa Nimplod P, Tansalarak R, Sornsuwan T -Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing 13 for the non-indented groups, and 20.30 MPa and 31.85 MPa for the indented groups, respectively. There was no statistically significant difference in the strength between non-indented and indented specimens within the similar bracket type. However, ceramic brackets had significantly higher debonding strength than metal brackets (p < 0.001).
Alteration of surface indented microcracks between metal and ceramic bracket groups is presented in Table 2. There was some degree of crack healing after removal of both metal and ceramic brackets, which are comparable to that for indentation on the surfaces without brackets. However, there were no statistically significant differences in the healing degree among the groups in both vertical (p=0.852) and horizontal (p=0.071) directions. Table 3 summarizes ARI scores of the debonded interfaces of all specimens. There were 13 specimens (43.3%) of metal brackets that failed at the enamel-adhesive interface (score 5), and twelve samples (40%) left adhesive on enamel surface more than 10% but less than 90% (score 3). However, there was no adhesive remnant on the enamel surface of the ceramic bracket group (score 5). Additionally, four samples bonded with the ceramic brackets (13.3%) presented enamel chipping.    Nimplod P, Tansalarak R, Sornsuwan T -Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing Table 4: Comparison of average microcrack healing rates between groups after bracket debonding (nm/s). Comparison of average microcrack healing rates among the groups after bracket debonding is presented in Table 4. No statistically significant difference was found among the groups (p = 0.792 for vertical and p = 0.215 for horizontal directions).  The cracks partially repaired soon after the removal of the brackets (Fig 4). Interestingly, even with the highest debonding strength observed in the ceramic groups, the same healing rates of the corner cracks as those without brackets could still be found at the bracket's boundary. The stress seems to be limited only to the bonded interface. This finding is consistent with the enamel chipping located on the proximity of incisal or gingival borders of the brackets in the ceramic group (Fig 5), and it is also consistent with a report in which finite element analysis of shear stress distribution in the enamel-adhesive interface was used. The researchers reported a pattern that was quite heterogeneous, and the stress concentration was limited to the upper and lower margins of the brackets. 22 After indentation with a Vickers microhardness tester on the enamel surface, the length of the diagonal microcracks reduced with time. It has been reported that indented microcracks repaired around 9% of their initial length in the first 24 hours and reached a plateau level (10 % of the initial length) in 48 hours. 9 The repair process may be the consequence of a viscoelastic recovery and extrinsic toughening mechanism of organic protein in enamel. 9 When indentation on the enamel surface and microcrack generated, the crack  Even though enamel microcracks healing is a normal process on a vital tooth preventing crack propagation to the dentin and dental pulp, this healing process can be found in extracted tooth. According to ISO/TS 11405:2015, the teeth that have been extracted for longer than six months may undergo degenerative changes in enamel and dentinal protein. 24 However, the teeth used in this study have been tested within a month after extraction. Therefore, the remaining organic protein in enamel still had an influence on crack closure stress, and healing process occurred.
Since in this study the corner cracks were remeasured at least 24 hours after indentation, the crack length was a resultant of crack healing combined with the stress at the bracket's boundary. The expected length was then calculated by using a healing degree of 9%, as suggested in the literature. 9 The original and expected 24-hour crack lengths, both in the vertical and the horizontal directions are presented in Figures 6 and 7, respectively. It was observed in this study that the cracks healed slightly more than expected in both directions.
Even for the specimens with high debonding strength, such as those with enamel chipping, the stress on the surrounding area during removal of a bracket was minimal and did not extend the microcracks. Additionally, the healing degree in this study might have been more efficient because the outer enamel was polished, since the fracture toughness, as well as the organic content, was found to increase from the surface enamel to the dentin-enamel junction. 9,25 Comparing the differences between the crack lengths alone  debonding strengths compared to metal brackets, there was no significant difference in the degree of healing between the groups. These similar healing rates confirm that bracket removing stress did not affect the healing process of boundary enamel microcracks.
The stress intensity factor (KI) is another mechanical parameter used to describe resistance of any material to critical crack growth. The critical value of KI or fracture toughness (Kc) of enamel can be evaluated by an indentation approach. 25 It has been reported that 10% reduction of the apparent fracture toughness is associated with the degree of microcrack healing in the enamel surface. That reduction in K c(app) is consistent with bridging by the organic matrix in enamel that can be defined as follows: 9 (4) where is the nominal bridging stress on the protein matrix (assumed to be equivalent to the yield strength of protein, f b is the area fraction of protein matrix bridging ligaments, and l b is the bridging zone length. For this study, due to the crack reduction, there was an increase in K c(app) from the initial indentation to the final observation, of approximately 15% (Table 4). This higher reduction is probably due to the investigated area on the enamel surface, which was located healing in this study is larger than that found by Rivera et al. 9 For the ARI scores, combination groups of the same bracket type (groups 1 and 3, as well as groups 2 and 4) were performed due to no statistically significant difference of debonding force between groups within similar bracket type. Bond failure for brackets was found to be more prevalent at the enamel-adhesive interface, especially in ceramic brackets (100%) ( Table 1).
The result might be due to a higher bond strength between the ceramic bracket bases and the adhesive. The predominant failure type of debonded ceramic brackets was found to be at the bracket-adhesive interface. 15,17 This kind of failure is beneficial to the enamel surface because it is left intact, although more time is required to remove the adhesive remnant. 26 On the contrary, there is a higher probability of enamel damage if the unit fails at the enamel-adhesive interface. 27 All brackets used in this study were subjected to the shear strength test with a universal testing machine to deliver shear force. The unilateral axial load applied to the bonding surface by this testing machine creates pure shear stress, which might differ Nimplod P, Tansalarak R, Sornsuwan T -Effect of the different debonding strength of metal and ceramic brackets on the degree of enamel microcrack healing 25 from removing pliers used clinically. 17 Consequently, the stress generated by a bracket removing plier is not directly comparable to the condition used in this study. 28 Debonding strength exerted by bracket removing plier has been reported to be 30% less than the shear strength delivered by the universal testing machine. 29 There are some limitations to this study. Firstly, a standardized laboratory setup may be extrapolated to a complex clinical situation, e.g., changes in temperature, humidity, acidity, mechanical and masticatory stress on brackets. Besides, moisture control in vitro is superior to in vivo. Secondly, the delayed measurement of microcrack length after bracket debonding could not be a real-time crack analysis. Thirdly, the location of the crack tip was difficult to identify by using the microscope, which resulted

CONCLUSIONS
Within the limitations of this study, the conclusions are as follows: Removal of ceramic brackets required a higher debonding strength and was more susceptible to enamel fracture than with metal brackets.
The surrounding cracks partially healed after bracket debonding.
The debonding stress from bracket removal was quite localized and did not affect the healing degree of surrounding microcracks.