Laser-aided debonding of orthodontic ceramic brackets Karlheinz Strobl, PhD,*.** Ted L. Bahns, BS,** Lee Willham, MS,*** Samir E. Bishara, DDS, BDS, D. Orth., MS,*** and W. C. Stwalley, PhD** Los A/amos, N.M., and Iowa City, Iowa
The removal of ceramic brackets from the enamel surface by means of laser heating was investigated with the use of CO2 and YAG lasers. The two bracket types investigated were polycrystalline alumina and monocryslalline alumina. The average torque force necessary to break the adhesive between the polycrystalline ceramic brackets and the tooth was lowered by a factor of 25 when the brackets were illuminated with a CO2 laser beam of 14 watts for 2 seconds. All polycrystalline brackets debonded with the CO2 laser resulted in a complete bracket detachment without bracket failure. The average torque force needed to debond monocrystalline brackets was lowered by a factor of 5.2 when illuminated with a laser setting of 7 watts. Monocrystalline brackets cracked along the bracket slot in 2 of 10 cases. Debracketing without laser heating resulted in a slightly higher incidence of bracket failure (12 of 50). Nevertheless, no visible damage to the enamel surface was observed. Advantages of the laser-aided bracket-removal techniques include the following: The heat produced is localized and controlled; the debracketing tool is essentially "cold"; and the method can be used for removal of various types of ceramic brackets, regardless of their design. (AMJ ORTHOD DENTOFACORTHOP 1992;101:152-8.)
T r a n s l u c e n t ceramic brackets became availabIe to the orthodontists in the last 5 years. Unlike metal brackets, which are ductile and can be peeled from the enamel tooth surface, ceramic brackets are brittle and cannot be readily removed by the conventional debonding methods. Manufacturers recommended that ceramic brackets be removed with pliers or wrenches, which apply shear or torsional forces, respectively. Several complications have been encountered with these techniques, including enamel fracture and bracket failure or breakage, t5 Various alternative techniques have been investigated, including the use of ultrasonic and electrothermal debonding. These techniques have their own advantages and limitations? '5 With the electrothermal approach, the composite adhesive softens above a critical temperature (approximately 150 ° to 200 ° C) and allows bracket debonding at a significantly reduced force magnitude. Since the detachment occurs within a short time interval (1 to 5 seconds), significant heat conduction
This project was supported, in part, by U.S. Public Health Service Grant 2 SO7-RR5313-28 awarded by the Bio-medical Support Grant Program, Division of Research Resources, National Institute of Health, and the Department of Orthodontics, College of Dentistry, University of Iowa. *Los Alamos National Laboratory. **Center for Laser Science and Engineering, University of Iowa. ***'Orthodontic Department, College of Dentistry, University of lov,,a. 811126589
that might result in patient discomfort or irreversible pulpal changes is avoided. The electrothermal device, still under development, uses a total energy of 30 joules for successful removal of the bracket. There are two main drawbacks to electrothermal debonding: (1) the whole assembly head must cool down after the removal of a few brackets and (2) the instrument is designed to fit a specific bracket design. Ideally, the orthodontist would like to have a universal tool that allows atraumatic debonding of ceramic brackets, regardless of the bracket design. Therefore the development of a compact, "cold" universal debracketing instrument based on a laser-aided technique should be of assistance to the orthodontist. The purpose of this study is to determine the efficiency of using COz and Nd:YAG (neodymium: yttrium-aluminum-garnet) lasers in debonding ceramic brackets from the enamel surface.
MATERIALS AND METHODS Brackets. The following ceramic bracket types ,,,,'ereused throughout the experiment: 63 brackets made of polycrystalline alumina (A12OD (Transcend, Unitek/3M, Monrovia, Calif.) and 30 monocrystalline (sapphire) alumina brackets (Starfire, A Company/Johnson & Johnson, San Diego, Calif.). The former material is also known under the name Lucole.? and scatters the light mostly in a forward direction, which results in a good (diffuse) energy transmission between 200 nm and 6 p.m. The heat conductance of this material is
Volume 101 Number 2
roughly 0.1 to 0.05 cal/cmKs for the temperature range of 0 ° to 200 ° C and has a linear expansion coefficient of approximately 6.7 10 -6 K - ' at 200 ° C. All the brackets were made available by the manufacturers. The polycrystalline brackets used were for the maxillary central incisors, whereas the monocrystalline brackets used were for the mandibular canines. The two bracket types had a surface area of 11.7 mm 2 and 12.8 mm-', respectively. Preparation attd bonding of teeth. ;I'he brackets were bonded to human molars and incisors. After extraction, 48 molars and 45 incisors were stored in a thymol solution. Previous reports found no clinically significant differences in the debonding forces of brackets attached to incisors and molars. J Before bonding, the teeth were washed, scaled of calculus and soft tissue debris, rinsed, and dried. The teeth were mounted in 13A-inch copper rings with the use of improved stone. The surface of enamel to be bonded was polished with a slurry of a nonfluoridated pumice, rinsed, and dried. Impressions of the enamel surface were obtained with a polyvinyl siloxane impression material. This recorded the original condition of the enamel surface. The teeth were etched with a 37% orthophosphoric acid solution for I minute, rinsed with water for 20 seconds, and completely dried with an air-water syringe. The same chemical-cure two-paste orthodontic composite resins were used to bond the brackets to the enamel surface on all teeth. The excess composite was removed before polymerization with a hand scaler. The teeth were stored in 100% humidity at 37 ° C for more than 48 hours to ensure complete polymerization. All brackets were bonded in one setting to allow for better control of the experimental conditions. Composite. The composite used to bond the brackets to the enamel surface was Concise (3M, Minneapolis, Minn.). It is a two-paste mix essentially composed of a quartz filler (77%) and methyl methacrylate (23%). The adhesive layer between the enamel surface and the bonding surface of the bracket ranged in thickness between 0.1 and 0.5 mm. At this thickness, the adhesive absorbs more than 95% of ultraviolet (308 nm) and near-infrared (10 p.m) light. The composite is more translucent in the visible near-infrared spectral region. The thermal expansion of the resin is approximately 1.6 10 -~ K-t at 200 ° C, and its thermal conductance is approximately 1.6 10 -3 cal/cmKs. The tensile strength of the resin is temperature dependent and is significantly reduced above 150 ° to 200 ° C. The thermal expansion coefficient of the composite is more than twice as large as that of the ceramic brackets. Variables recorded. The molars and incisors, bonded with polycrystalline and monocrystalline brackets, were divided into subgroups. The following four variables were evaluated: the residual torque/surface area needed to debond the bracket, the incidence of bracket failure, the evaluation of residual adhesive on the surface of enamel, and the presence of enamel damage. The application of a torque force perpendicular to the bracket-enamel interface was chosen as the standard dcbonding technique throughout this study. Bracket failure was defined as any breakage of the bracket, partial or complete, in either the bracket base or the tie wings. The amount of residual
Laser-aided debonding of ceramic brackets 153
Ceramic Orthodontic Bracket
Fig. 1. Experimental setup for measuring residual debonding torque after illuminating ceramic brackets with set laser power. The shutter and translation stages are interfaced to a computer for proper synchronization. adhesive was classified in compliance with a previously defined adhesive remnant index (ARI), which uses a scale of 1 to 5 to assess the amount of resin material adhering to the e n a m e l SUlface. 7
When a bond failure was detected at the enamel-adhesive interface, a magnifying ( × 10) loop was used to assess the damage to the enamel surface. Any observed damage was compared with the pretreatment scanning electron micrographs of the polyvinyl siloxane impressions of the enamel surface. Residual torque meast~rement. A schematic diagram of the experimental setup used is presented in Fig. i. The laser was set to the desired intensity, and its beam was blocked by a shutter so that the laser could stabilize. A beam-splitter allowed simultaneous accurate measurement of the laser power. The beam was focused, with a lens, to a spot slightly smaller than l mm at the bracket surface. The tooth with the bonded bracket was mounted in a bolder. A thermally insulated fork (Teflon tape wrapped around the fork fingers) was clamped to the bracket. The thermal insulation was necessary to prevent fast cooling of the bracket through its contact with the fingers of the rectal fork. Without thermal insulation of the mechanical contacts, a significant increase in the debonding torque force was observed. Also, some of the higher residual threshold values recorded were probably caused by imperfect thermal insulation. The thermal insulation had to be done manually for each bracket tested.
Strobl et el.
Am. J. Orthod. Dentofac. Orthop. Februa~" 1992
Table I. Measured residual torque force (Nmm/mm'-) needed to debond the ceramic brackets without laser application Brackets Polycrysta/line Teeth
20.9 23. I
N = sample size; x = mean; and SD = standard deviation.
Table II. Frequency distribution of various Adhesive Remnant Index scores and bracket failure rate (breakage o f tie wings or base) during debonding without laser application Brackets Polycrystalline ARI scores*
1 2 3 4 5
7/10 2/10 1/10
*1, 100% of composite remains on enamel surface along with impressions of bracket base. 2, More than 90% of composite remains. 3, More than 10% but less than 90% of composite remains. 4, Less than 10% of composite remains. 5, All composite removed from enamel surface during debonding.
The pin of a force meter touched the fork at a distance of 8.76 cm from the bracket center. The force meter was mounted on a motorized translation stage. The laser beam shutter was set to allow an exposure of 2 seconds after receiving a trigger signal. Both the translation stage and the shutter were computer interfaced for proper synchronization. As soon as the shutter was closed again, a motorized translation stage pushed the force meter forward (! cm/sec). After a certain travel distance, the bracket popped off, and the maximum torque force applied to the bracket was read from the force meter. The time delay between the closing of the shutter and the bracket detachment was approximately 2 seconds. Computerizing the data acquisition made the measurement reproducible. The torque force was applied only after the shutter was closed and the bracket was already cooling off. Therefore the measurements presented could be interpreted as upper limit values for the real residual debonding torque force.
RESULTS Brackets debonded withottt laser. To allow for a reliable assessment of the laser-aided debonding techniques, 50 brackets (30 bonded to molars and 20 bonded
to incisors) were debonded without the use o f a laser. For each bracket the torque force necessary to break the bond between the ceramic bracket and the tooth surface was measured with the setup shown in Fig. 1, but without the laser beam being activated. The measured residual torque was then normalized to the surface area of the bracket to allow for standardization and appropriate comparisons between the different bracket types. The debonding torque forces for monocrystalline and polycrystalline brackets bonded to molars and incisors are presented in Table I. No statistically significant differences in the debonding forces for the four subgroups were observed. The A R I scores and the incidence of bracket failure are presented in Table II. No enamel damage was observed either in the form o f cracks in the enamel surface or as missing enamel fragments.
Brackets debonded with the aid of COe laser. At the COz laser wavelength (10.6 I-tm), both bracket materials are highly absorptive. On illumination of the bracket, the energy is absorbed and converted into heat
Volume 101 Number 2
Laser-aided debonding of ceramic brackets 155 30
• Polycrystalllne Brackets • Monocrystalline Brackets
E tT~ Ct.-¢
Total CO2 Energy (J) Fig. 2. Residual debonding torque forces for polycrystalline and monocrystalline brackets in dependence of applied CO2 laser energy. Error bars show one standard deviation.
in a very thin surface layer of the bracket. From that point, the absorbed energy propagates by heat conduction to the opposite side of the bracket and softens the composite. To determine the appropriate level of laser heating to be used for debonding, polycrystalline and monocrystalline brackets were exposed for 2 seconds to a CO2 laser beam at different power levels. The residual debonding torque force needed for bracket removal was recorded. This setting was based on earlier experiments, which indicated that use of longer exposure times at constant total energy leads to cooling through the fork fingers and results in an increased residual debonding force. On the other hand, with short exposure times ( < 1 second), brackets tended to crack from the extreme temperature gradient generated by the higher local heat intensity needed to cause debonding. Since the main objective of the present approach is to debond the bracket from the tooth painlessly, a 2-second exposure time was chosen. Such an exposure time has the following advantages; minimization of enamel damage with minimal force for debonding and achievement of the required laser energy level to soften the composite in 2 seconds. This time factor is within a range of what could be considered clinically efficient debonding time. The observed decrease in the residual debonding torque force with the corresponding increase in laser energy level is presented in Fig. 2. The data points within a certain energy interval (typically three to five brackets) have been grouped, and the averages with standard deviations plotted. Values with larger y-standard errors
contain fewer data points. The x-standard errors represent the energy intervals used for the various averages. For monocrystalline brackets, a two to four times larger beam waist was used to minimize plasma and plume formation and occasional bracket cracking. Such a behavior has been significantly reduced through enlargement of the incident laser beam by moving the bracket closer to the lens until the laser beam approximately filled the bracket surface. On the other hand, no such behavior was observed when debonding polycrystalline brackets with a beam waist of I mm or less. On the basis of the results presented in Fig. 2, a fixed laser energy power level of 14.1 watts applied for 2 seconds was selected as most optimal for additional debonding experiments. These settings resulted in a significant reduction in the required debonding torque force. Also, this power level is in the range of energy produced by the currently available thermal debonding instrument that operates at approximately 15 Watts. Table III presents the residual debonding torque forces with the COz laser at the optimal setting. Only a few monocrystalline brackets were tested at this setting. From the overall results, the polycrystalline brackets and monocrystalline brackets debonded differently as shown in Fig. 2. All polycrystalline brackets debonded without any bracket failure. On the other hand, the monocrystalline brackets had a tendency to be more sensitive, both to the anaount of clamping force with which the fork gripped the bracket and to the laser beam waist. No enamel damage in the form of cracks in the enamel surface or missing enamel fragments were ob-
Strobl et el.
Am. J. Orthod. Dentofac. Orthop. February 1992
j * t l l f l t T | r , . , t l , t l l l l * t l t * t t | l s , . t , , ,
total YAG laser energy
Fig. 3. Residual debonding torque in dependence on applied Nd :YAG laser energy.
served during the debonding of brackets of either type. Brackets debonded with the YAG laser. At the YAG laser frequency of 1.06 g.m, the transmission level of the incident light was at 8 3 . 4 % - 2.5% and 92% _ 2.5% for polycrystalline and monocrystalline brackets, respectively. For a typical 350 to 400 p.m thickness of the adhesive, the absorption level was approximately 18% - 3%. This would indicate that approximately 69% to 75% of the incident light reaches the enamel surface and can potentially cause damage or pain. Depending on the reflection of the enamel surface, only 15% to 26% of the incident light in the case of polycrystalline bracket (17% to 29% for monocrystalline) is deposited in the adhesive layer and converted into heat. In fact, laser-aided bracket removal with a 5-second exposure with a total energy of 120 joules did not produce any significant additional reduction in the debonding force. Two methods suggested to improve this situation involved modification of either the adhesive or the bracket surface to make them more absorptive. Changing the composition of the adhesive, for example, by adding a dye would also require a reevaluation of the effective bond strength. This approach was not attempted in this study. To artificially enhance laser absorption by the bracket, the exposed surface of the translucent ceramic bracket was painted with a thin film of a dark (high absorptive) paste. The coating was made of a mixture of graphite powder (for high laser absorption) and a high-temperature silicone grease (for optimum heat transfer and mechanical adhesion). The weight ratio of
the two materials was approximately 5: 1. The silicone paste has a high thermal conductivity of k = 0.07 cal/cmKs. The paste used in the present study is not necessarily compatible with the one that will be eventually used intraorally. The results of this modification of the bracket surface are shown in Fig. 3. Although fewer teeth were tested, one can readily observe that the coated brackets, when exposed to a total energy of approximately 30 joule, resulted in a significantly reduced debonding force. DISCUSSION
The effects of the laser-aided debonding of ceramic brackets were investigated with the use of both the COz and the YAG lasers. The CO2 laser debonding was investigated more extensively, whereas for the Nd:YAG laser only a proof of principle was tested. This is because both lasers cause essentially the same physical effects, namely, softening of the composite with heat generated by the laser beam. Therefore it was concluded that little more could be learned by duplicating the more comprehensive CO2 laser experiments. Also, no attempt has been made in this study to measure the transmitted temperatures, within the tooth itself, resulting from the use of the laser-aided debonding technique. Reduction in debonding force. The initial measurement series for debonding without laser heating provides a baseline (Table I) that can be used to compare the effectiveness of the laser heating methods in reducing the debonding force. A significant percentage of brackets (12/50) fractured at the tie wing or at the
Volume 101 Number 2
Laser-aided debonding of ceramic brackets
Table III. Measured residual torque force (Nmm/mm 2) needed to debond ceramic brackets with COz laser at
the 14.1 watt for 2-second illumination Brackets Monocrystalline
N = Sample size; x = mean; and
= standard deviation.
bracket base during debonding, with the four subgroups having approximately the same bracket failure rates. CO2 laser-aided bracket-debonding techniques (Fig. 2 and Table III) resulted in significantly lower residual debonding torque force when compared with nonlaser debonding. The detachment behavior of polycrystalline and monocrystalline brackets was not identical (Fig. 2). The latter required a significantly lower laser energy to cause an equivalent debonding torque force reduction. As an example, at a total energy of 15 joules, a 5.2-fold reduction (in the residual torque debonding force) was obtained with monocrystalline brackets. On the other hand, polycrystalline brackets showed only a 1.3-fold decrease under the same conditions. It is possible that the different behaviors observed are, in part, the result of differences in the design (shape and dimensions) of the two brackets, as well as in their different microscopic structure. The polycrystalline material consists of an agglomerate of small microcrystals with random orientation, size distribution, and shape that result in a much higher energy diffusion (heat and light) than the basically homogeneous single-crystal material (Sapphire) used in the monocrystalline brackets. In other words, with the use of similar energy levels in debonding both bracket types, the heat transmits through the body of the monocrystalline bracket with much less lateral spreading, which results in a significantly hotter spot at the bracket-adhesive interface. Since the adhesive must be heated to approximately 150 ° to 200 ° C before significant softening of the composite occurs, a more localized hot spot is more effectively obtained with the monocrystalline brackets. This explanation is supported by the observation that at higher laser energy levels, the monocrystalline brackets have a tendency, during the debonding process, to show plasma or plume formation, occasionally associated with the cracking of the bracket along the wire slot. The polycrystalline brackets showed no such behavior at the same energy levels. At a power level of approximately 14 watts (similar
to the level used for thermal debonding techniques) applied for 2 seconds, the residual debonding torque force for polycrystalline brackets bonded to molars showed an average 25-fold reduction. The average reduction for incisors was half as much as that recorded for molars (Table IV). This still may not be significant, since the standard error is larger than the differences in debonding forces. Furthermore, in this study only a few brackets were tested. Additional work is necessary to determine whether the observed differences are real. Bracket faihtre and enamel damage. All laser-aided polycrystalline bracket debondings resulted in no bracket fracture, while 12 of 50 brackets debonded without laser were damaged during the bonding procedure. Since the residual debonding torque force was obviously significantly reduced by the laser heating, such an improvement in the bracket failure rate is not surprising. No enamel damage was found when the surface was inspected with a X10 magnifying glass. With monocrystalline brackets, the failure rate with laser-aided debonding was 25%, with most of the breakage observed along the bracket slot, particularly when laser energies of between 17 and 28 joules were used. As discussed earlier, it is suggested that the reduced lateral heat diffusion of this material results in a higher thermal gradient in the bracket body, causing the material to be more sensitive. In addition, the monocrystalline bracket design might have an additional effect since it has a thinner bridge between the bracket halves compared with the polycrystalline design. Such bracket damage can be overcome by a more appropriate beam size and a different clamping tool. Effects of laser debonding on the ARI scores. When the laser was not applied, 80% of the brackets debonded had an ARI score of 1 (100% of the composite remains on the enamel surface). With laser-aided debonding, the frequency of the ARI score of 1 increased to 88%. An even greater increase (95%) was noticed when the polycrystalline brackets were debonded with only 28 joule. Therefore, laser-aided debonding significantly reduces the probability of bond failure at the enamel-
158 Strobl eta/.
composite interface, further minimizing the risk of enamel damage. Excimer-aided debonding. Ultraviolet light (300 to 200 rim) was found to readily transmit through the ceramic material. 6 As a result, an excimer laser (308 nm) has been successfully used to debond monocrystalline ceramic brackets, s Tocchio et al. s achieved debonding with an XeC1 excimer laser operating at a setting of 21 millijoule per pulse, 400 Hz, and a pulse width of 10 ns. Under these conditions, the mean illumination time for bracket removal was found to be 3.7 _ 0.7 seconds, resulting in an average energy of 31 ___ 6 joule. 4 A 10 N shear load was continuously applied during laser exposure. This excimer-aided debonding technique required laser energies similar to those found in the present study with both the CO2 and the N d : Y A G lasers. Such observations indicate that the primary debonding mechanism with excimer laser is also a tranformation o f the ultraviolet energy into heat, which results in a softening of the adhesive. A photochemical breakdown of the methylmethacrylate could still happen at a different laser setting. CONCLUSIONS The proposed laser-aided debonding technique was found to significantly reduce the residual debonding force, t h e risk o f enamel damage and the incidence of failure as compared with the conventional debonding techniques. Therefore this method has the potential to be more atraumatic (less painful) and safer (less risk of enamel damage) for the patient. Further research is necessary to determine the effects of the heat generated
Am. J. Orthod. Dentofac. Orthop. February 1992
on the pulp tissues and also to improve on the laser debonding procedure. We thank K. H. Yang for his helpful suggestions. Equipment was provided by the Iowa Laser Facility and the Physics Department of the University of Iowa. REFERENCES I. Joseph VP, Rossouw E. The shear bond strengths of stainless steel and ceramic brackets used with chemically and light-activated composite resins. AM J ORTHOD DEN'rOFACORTHOP 1990;97: 121-9. 2. Viazis AD, DeLong R, Bevis RR, Rudney JD, Pintado MR. Enamel abrasion from ceramic orthodontic brackets under an artificial oral environment. AsI J ORTHOD DENIOFACORTHOP 1990;98:103-9. 3. Bishara SE, Trulove TS. Comparisons of different debonding techniques of ceramic brackets: an in vitro study. Part I. Background and methods. AM J ORIHODDENTOFACORTHOP 1990; 98:145-53. 4. Viazis AD, Cavanaugh G, Bevis RR. Bond strengths of ceramic brackets under shear stress: an in vitro report. AM J ORTItOD DE:',rFOFACORTHOP1990;98:214-21. 5. Bishara SE, Trulove "IS. Comparisons of different debonding techniques for ceramic brackets: an in vitro study. Part lI. Findings and clinical implications. AM J ORTHOD DENTOFACORTHOP 1990;98:263-73. 6. Lucolex Ceramic Tubing brochure, General Electric. Richmond Heights, Ohio: General Electric Co. 7. Oliver RG. The effect of different methods of bracket removal on the amount of residual adhesive. AM J ORTHODDENTOFAC ORTHOP 1988;93:196-200. 8. Tocchio R, Williams PT, Mayer F. Laser debonding of sapphire orthodontic brackets (Abstract 1007). J Dent Res 1989;68:993. Reprint requests to:
Dr. Samir Bishara College of Dentistry University of Iowa Iowa City, IA 52242