ORIGINAL ARTICLE

Enamel surface evaluation after removal of orthodontic composite remnants by intraoral sandblasting: A 3-dimensional surface prolometry study
Seong-Sik Kim,a Woo-Kyung Park,b Woo-Sung Son,c Hyung-Soo Ahn,b Jung-Hoon Ro,d and Yong-Deok Kime Pusan, Korea Introduction: The purpose of this study was to determine the utility of sandblasting to remove composite remnants after orthodontic bracket debonding. Methods: The sample consisted of 20 human premolars extracted for orthodontic purposes. The buccal surface of each premolar was divided into 3 parts: the upper half (control surface group, CS), the lower half left (LS group), and the lower half right (SS group). A composite resin paste (volume, 5 3 1 mm3) was bonded onto the LS and SS surfaces. Then it was removed by using 1 of 2 methods: low-speed handpiece with tungsten carbide bur in the LS group and sandblasting in the SS group. Temperature change and removal time were recorded, and surface proles were examined with 3-dimensional prolometry. Results: An independent t test showed a statistically signicant difference in temperature change between the LS and SS groups (P .01). ANOVA showed no signicant difference in surface prole between the LS and SS groups (P .5). Conclusions: The results suggest that intraoral sandblasting might be an alternative to rotatory instruments for resin remnant removal after orthodontic bracket debonding. (Am J Orthod Dentofacial Orthop 2007;132:71-6)

ith improvements of the physical and mechanical properties of resin adhesive systems, cleanup of resin remnants after orthodontic bracket debonding has become a clinical problem. The removal of adhesive remnants from tooth surfaces after orthodontic bracket debonding is a nal procedure to restore the surface as closely as possible to its pretreatment condition without inducing iatrogenic damage.1 If remnants are not completely removed, tooth surfaces are likely to become unesthetically discolored and entrap plaque with time.2 Therefore, many investigators have introduced various resin remnant-removal techniques. The most common removal technique uses ultraFrom Pusan National University, Pusan, Korea. a Assistant professor, Department of Orthodontics, College of Dentistry. b Postgraduate student, Department of Dental Materials, College of Dentistry. c Professor, Department of Orthodontics, College of Dentistry. d Assistant professor, Department of Biomedical Engineering, College of Medicine. e Assistant professor, Department of Oral and Maxillofacial Surgery, College of Dentistry. Supported by the Pusan Orthodontic Society Research and Education. Reprint requests to: Yong-Deok Kim, 602-739, Department of Oral and Maxillofacial Surgery, Pusan National University Hospital, 1-10, Ami-Dong, Seo-Gu, Pusan, Korea; e-mail, ydkimdds@pusan.ac.kr. Submitted, April 2005; revised and accepted, July 2005. 0889-5406/$32.00 Copyright 2007 by the American Association of Orthodontists. doi:10.1016/j.ajodo.2005.07.027

sonic scalers, a low-speed handpiece with a tungsten carbide bur, and a high-speed handpiece with a diamond bur.3-5 Recently, a new method of ND:YAG laser radiation to selectively remove residues of bonding resin was introduced.6,7 The most preferred method is to use a low-speed handpiece with a round tungsten carbide bur.8 This technique, however, is time-consuming and inefcient, and can damage tooth enamel.9 Buyukyilmaz and Zachrisson10 used a sandblasting technique to improve the bond strength of amalgam, porcelain, and gold crowns. Other investigators also reported that sandblasting bracket bases greatly increases their retentive surfaces.11 Concerning the tooth enamel surface, however, Reisner et al12 reported that sandblasting does not appear to damage the enamel surface and can be used as a substitute for polishing with pumice. The aim of this in-vitro study was to determine the applicability of sandblasting to resin remnant removal after orthodontic bracket debonding.
MATERIAL AND METHODS

Twenty human premolars extracted for orthodontic purposes were collected and stored in normal saline solution at room temperature (25.7C). The criteria for tooth selection included intact buccal enamel not sub71

72 Kim et al

American Journal of Orthodontics and Dentofacial Orthopedics July 2007

Fig 1. Experimental enamel surface embedded in dental white stone: CS, control group enamel surface; LS, resin remnant simulation with light pink color for lowspeed handpiece treated surface; SS, sandblasting treated surface.

jected to any bracket bonding procedures, no damage caused by the extraction forceps, and no caries. The teeth were cleaned with prophylactic brushes to remove debris, rinsed in water spray for 5 seconds, and air dried. The teeth were embedded in dental white stone blocks, keeping the buccal surfaces exposed. The buccal surface of each tooth was oriented upward from the bottom of the block so that its buccal surface would be perpendicular to the cantilever tip of the surface prolometry (Form Talysulf Series 2 PGI plus, Rank Taylor Hobson, Leicester, England) during enamel surface scanning. The buccal surface of each tooth was divided into 3 parts for comparison of the surface characteristics with the selected remnant-removal procedures: the upper half was the control surface (CS group), the lower left quadrant was the low-speed handpiece treated surface (LS group), and the lower right quadrant was the sandblasting treated surface (SS group) (Fig 1). Celluloid tape with 5 3 mm2 holes covered the buccal enamel surface in each group except for the CS group; then resin adhesive was bonded to the exposed part of the enamel. A conventional light-curing bonding method was used for 30 seconds of acid-etching with 37% phosphoric acid (etching agent, Reliance, Itasca, Ill), 10 seconds of rinsing with water spray and air drying, 10 seconds of light curing (curing light XL3000, 3M Unitek, Monrovia, Calif) of the primer (Transbond XT Light Cure Adhesive Primer, 3M Unitek) mixture with rhodamine B dye material (Fluka Chemica-Biochemica, Buchs, Switzerland) for visual

inspection, and then 20 seconds of light curing of the resin adhesive (Transbond XT Light Cure Adhesive Paste, 3M Unitek) of 5 3 1 mm3 volumes onto the exposed part of the enamel to simulate resin remnants. The resin adhesive in each group was removed according to the following regimen. In the LS group, the resin adhesive was removed with a low-speed handpiece with a tungsten carbide bur (FG4, Komet, Lemgo, Germany), with a new bur used for each surface. In the SS group, removal of the resin adhesive was carried out with 50-m aluminum-oxide particle sandblasting (Microetcher II, Danville Engineering, San Ramon, Calif). The nozzle of the sandblaster was xed in a position perpendicular to the enamel surface at a distance of 10 mm. The cleanup was performed by 1 operator (S.S.K.). Complete removal of the resin adhesive was veried by visual inspection (disappearance of pink color) under a dental operating light. The temperature changes and the time taken to remove the remnant from the enamel surface were simultaneously recorded by using a digital thermometer (DTM 317 digital thermometer, TECPEL, Taipei, Taiwan). The laboratory room temperature was 25.7C. The surface prole of each group was measured with the Form Talysulf Series 2 PGI plus. The blocks were mounted in the devices cross-table and adjusted by tilting the cross-table until each surface was perpendicular to the devices stylus tip. The measuring surfaces were 50 50 m, and the measuring units were 10 nm. The measurements were recorded by computer. The measurement parameters were as follows: Ra, the arithmetic average value of the prole departure from the mean line in a sampling length; Rz, the maximum height of the prole; Rsk, the measurement of the symmetry of the prole about the mean line; and Rku, the measurement of the sharpness of the prole. Scanning electron microscopic (SEM) (S3500N, Hitachi, Tokyo, Japan) images were made of all 3 groups.
Statistical analysis

Descriptive statistics were computed. An independent t test was used to determine whether statistically signicant differences (P .01) existed between the LS and the SS groups in time and temperature changes during the resin remnant-removal procedures. A 1-way analysis of variance (ANOVA) was performed to determine whether statistically signicant differences (P .05) existed among the 3 groups in the surface prole measurements. The determinative factor was the resin removal technique (sandblasting or low-speed handpiece). The Duncan multiple range test was used to compare the groups for statistically signicant differences.

American Journal of Orthodontics and Dentofacial Orthopedics Volume 132, Number 1

Kim et al 73

Table I.

Temperature change and removal time during resin remnant removal

Group N 20 20 20 20 Mean SD 2.79 0.47C 0.15 0.08C 3.13 0.95 sec 5.05 2.31 sec P value .000 .001*

Temperature changes Removal times *P .01; P .001. Table II.

LS SS LS SS

Surface prole measurement in each group

Mean SD 0.45 0.522 m 1.24 0.631 m 1.46 0.378 m 2.96 2.134 m 6.93 2.849 m 9.99 2.693 m 0.26 0.438 0.07 0.315 0.08 0.358 3.21 0.954 2.67 0.534 2.86 0.391 F value 21.011 P value .000 Duncan test A B B A AB B

Group Ra CS LS SS CS LS SS CS LS SS CS LS SS

Rz

4.146

.021*

Rsk

1.662

.199

Rku

3.381

.041*

A B AB

*P .5; P .01. Groups with same letter are not different (P .05). Measurement units are relative ratio about mean line.

RESULTS

Results of time and temperature changes during resin remnant removal for each group are summarized in Table I. The temperature change in the SS group was less than that in the LS group. The temperature of the LS group was elevated by 2.79C ( 0.47) during remnant removal, whereas the temperature elevation of the SS group was 0.15C ( 0.08). A statistically signicant difference existed between the LS and the SS groups (P .001). In elapsed time for remnant removal, however, the SS group required much more than did the LS group, with a statistically signicant difference in removal time between those groups (P .01). Results of the surface prole measurement for each group are summarized in Table II. The ANOVA for the arithmetic average value of the prole departure from the mean line in the sampling length (Ra) indicated signicant differences among the 3 groups (P .001). The Duncan test showed that the LS and SS groups had signicantly greater irregularities compared with the CS group. The ANOVA for the maximum height of prole (Rz) indicated a statistically signicant difference (P .05) among the 3 groups. The Duncan test

Fig 2. Computerized graphic image of enamel surfaces: A, normal enamel surface (CS group) shows short and repetitive spiky peaks; B, enamel surface treated with low-speed handpiece (LS group) shows sloped valley with large puddle; C, sandblasted enamel surface (SS group) shows small irregular peaks.

showed a signicantly greater enamel peak for the SS group than for the CS group. The Rsk value, the measurement of the symmetry of the prole about the mean line, showed no differences among the 3 groups. The ANOVA for measurement of the sharpness of the prole (Rku) indicated a signicant difference (P .05) among the 3 groups. The Duncan test showed signicantly greater sharpness for the SS group than for the CS group. Computerized images of the surface prolometry for each group are shown in Figure 2. SEM images of the tooth surfaces in each group are shown in Figure 3.
DISCUSSION

We compared the resulting characteristics of enamel surfaces after remnant removal between the conventional method with the low-speed handpiece and an 8-uted tungsten carbide bur and the new method of intraoral sandblasting. SEM, a subjective qualitative analysis, and

74 Kim et al

American Journal of Orthodontics and Dentofacial Orthopedics July 2007

Fig 3. SEM images (600): A, normal enamel surface (CS group) shows perikyma ridges that run around tooth over its entire surface; open enamel prism ends are small holes. B, Enamel surface treated with low-speed handpiece (LS group) shows slight faceting with scratch lines intermingled with perikyma ridges. C, Sandblasted enamel surface (SS group) does not show perikyma ridges.

contact optical prolometry, a precise quantitative measurement of surface roughness, were used. To determine the effectiveness of the sandblasting remnant-removal techniques in clinical orthodontics, temperature changes and remnant-removal time were compared with those of a low-speed handpiece. There were signicantly greater temperature changes in the LS group than in the SS group. The maximum temperature elevations of the pulp were 4.2C in the LS group and 0.3C in the CS group. Zach and Cohen13 demonstrated that pulpal necrosis occurred in 15% of teeth with a 5.5C increase in Rhesus macaca pulp tissues. Several authors recommended using water spray to remove the bulk resin.14 However, this method has a disadvantage, because water coolant results in poor contrast between resin remnant and enamel.9 In contrast to using the low-speed handpiece, sandblasting for removal of resin remnants did not cause pulpal damage. In removal time of resin remnants, the SS group (5.05 2.31 seconds) was signicantly slower than the LS group (3.13 0.95 seconds). A clinical implication of removal time is that the low-speed handpiece is more efcient than sandblasting. However, the vibration of the low-speed handpiece during resin remnant removal can be uncomfortable for the patient.15 Sandblasting did not vibrate the tooth during remnant removal. Other studies used various methods to compare enamel surfaces after remnant removal: SEM, visual inspection by photography, and adhesive remnant index.16-18 However, most of them did not compare enamel surface textures. To compensate for this limitation, we used prolometric analysis with a prolometer in this in-vitro study. The Ra values, the arithmetic average values of the prole departure from the mean line in the sampling length, were larger for the LS and the SS groups than for the CS group, and there was no signicant differ-

ence between the LS and the SS groups. This means that the average depth of removed enamel surface is similar between the low-speed handpiece and sandblasting. However, the amounts of enamel loss supposed that the Ra values of this study (Table II) were less than those of previous studies in a clinical situation. Because the Ra value describes the surface prole above the mean line, the Rz value, the maximum height of prole, should be used for direct clinical observation. Our results showed that the amounts of removed enamel (the Rz value) were 6.93 ( 2.849) m for the LS group and 9.99 ( 2.693) m for the SS group. Pus and Way19 found that a low-speed handpiece with a tungsten carbide bur removes approximately 10 m of enamel, whereas Thompson and Way20 showed that an average of 14.2 m of enamel was lost when prophylaxis with zirconium silicate was performed with a rotating bristle brush. The differences between those 2 studies might not be in removal procedure but in cleanup time. The removal procedure in the former study was performed for 15 seconds, and that for the latter study was performed for 30 seconds. In this study, the cleanup times were 3.13 ( 0.95) and 5.05 ( 2.31) seconds for the LS and the SS groups, respectively. These results agree with previous studies that indicated that the amounts of enamel loss are proportional to cleanup time and suggest that most resin remnantremoval procedures have a similar effect on the enamel surface, including minor scratching, gouging, dimpling, and cracking with respect to the selected procedure.21 Although resin remnant-removal procedures cause deterioration of the enamel surface, no clinical problem was caused, because the enamel surface layer contained a uoride-rich layer of 50 m depth. Furthermore, OReilly and Featherstone22 demonstrated that teeth exposed to daily rinsing with 0.05% sodium uoride or weekly topical application of acidulated phosphate

American Journal of Orthodontics and Dentofacial Orthopedics Volume 132, Number 1

Kim et al 75

uoride (APF) gel during a 1-month period showed rehardening or inhibition of demineralization, and a hypermineralized (92%) outer layer to a depth of 25 m.22 If the sharpness of the prole curve has a balanced Gaussian shape, the Rku value measuring the sharpness of the prole is about 3. A blunt surface will yield a value of less than 3, and a peaky or spiky surface, a value of more than 3. In this study, the Rku value of the CS group (3.21 0.954) was greater than those of the LS group (2.67 0.534) and the SS group (2.86 0.391). This result might be due to the perikyma ridges of the natural enamel surface. Although the enamel surfaces after treatment with the low-speed handpiece and sandblasting were irregular, the natural enamel also was slightly repetitive and spiky over the whole enamel surface area (Fig 2, A). The Rsk value, another variable that indicates the nature of the surface, supports the results for the Rku value in this study. A negative Rsk value indicates that the surface has a uniform structure. The Rsk value of the CS group in this study was more negative than the values of the LS and the SS groups. However, no signicant statistical difference in Rsk value existed among the 3 groups (P .1). Microetcher II, used in this study, is an FDAapproved intraoral sandblasting machine.10 Micromechanical retention of metals can be increased by at least 300% by using intraoral sandblasting. Chung et al23 showed that sandblasting metal and porcelain surfaces obtained bracket bond strength comparable with that for an etched enamel surface. By contrast to restoration preparation, Reisner et al12 concluded that sandblasting does not appear to damage the enamel surface, and that sandblasting before etching can be a substitute for polishing with pumice. Although the usual amount of enamel loss caused by acid etching is similar to that caused by sandblastingfrom 3 to 10 mthe clinical implication for retention ability is dissimilar.24 Acid etching is a form of microetching, but sandblasting can be regarded as a form of macroetching. SEM evaluation of an acid-etched enamel surface demonstrates that the prism centers are removed preferentially, and the prism periphery demonstrates the repetitive head-and-tail arrangement of the prism. However, SEM evaluation of sandblasted enamel shows a large and irregular roughened surface without subtle histologic alterations, creating mechanical interlocks (Fig 3, C). The new method for resin remnant removal from enamel surfaces after bracket debonding with an intraoral sandblasting machine is considered a highly complex procedure. In clinical situations, a plastic device (Sand-Trap, Danville Engineering) designed to

Fig 4. Resin remnant removal after orthodontic bracket debonding with chair-side intraoral sandblasting. Standard high-volume suction tip is inserted close to side hole of Sand-Trap; tip of microetcher ts into small hole snugly.

conne and suction away aluminum oxide particles during intraoral sandblasting is used (Fig 4). Furthermore, various factors, including sandblasting pressure, duration of sandblasting, particle size, type of particle, and protection for operator and patient should be considered in an appropriate clinical application.
CONCLUSIONS

Our results suggest that the enamel surface structure after remnant removal with intraoral sandblasting is similar to that after removal with a low-speed handpiece. However, resin remnant removal with sandblasting has advantages for pulpal health and patient comfort. With a specially designed device for protecting operator and patient, sandblasting can be an acceptable alternative to rotatory handpieces for restoring the enamel surface to its near-original state. We thank Dr Sang-Don Lee from the Laboratory of Tribology for his advice and work that was instrumental in completing this project.
REFERENCES 1. Campbell PM. Enamel surfaces after orthodontic bracket debonding. Angle Orthod 1995;65:103-10. 2. Hong YH, Lew KKK. Quantitative and qualitative assessment of enamel surface following ve composite removal methods after bracket debonding. Eur J Orthod 1995;17:121-8. 3. Krell KV, Courey JM, Bishara SE. Orthodontic bracket removal using conventional and ultrasonic debonding techniques, enamel loss, and time requirements. Am J Orthod Dentofacial Orthop 1993;103:258-66. 4. Oliver RG, Grifths J. Different techniques of residual composite removal following debondingtime taken and surface enamel appearance. Br J Orthod 1992;19:131-7. 5. Retief DH, Denys FR. Finishing of enamel surfaces after debonding of orthodontic attachments. Am J Orthod 1979;49:1-10.

76 Kim et al

American Journal of Orthodontics and Dentofacial Orthopedics July 2007

6. Thomas BW, Hook CR, Draughn RA. Laser-aided degradation of composite resin. Angle Orthod 1996;66:281-6. 7. Alexander R, Xie J, Fried D. Selective removal of residual composite from dental enamel surfaces using the third harmonic of a Q-switched ND:YAG laser. Lasers Surg Med 2002;30:240-5. 8. Zachrisson BU, Arthun J. Enamel surface appearance after various debonding techniques. Am J Orthod 1979;75:121-37. 9. Zachrisson BU. Chap. 10. Bonding in orthodontics. In: Graber TM, Varnasdal RJ, editors. Orthodontics: current principles and techniques. 2nd ed. St Louis: Mosby; 1994. p. 570-83. 10. Buyukyilmaz T, Zachrisson BU. Improved orthodontic bonding to silver amalgam. Part 2. Lathe-cut, admixed, and spherical amalgams with different intermediate resins. Angle Orthod 1998;68:337-44. 11. Millett D, McCabe JF, Gordon PH. The role of sandblasting on the retention of metallic brackets applied with glass ionomer cement. Br J Orthod 1993;20:117-22. 12. Reisner KR, Levitt HL, Mante F. Enamel preparation for orthodontic bonding: a comparison between the use of a sandblaster and current techniques. Am J Orthod Dentofacial Orthop 1997;111:366-73. 13. Zach L, Cohen G. Pulp response to externally applied heat. Oral Surg Oral Pathol 1965;19:515-30. 14. Zarrinnia K, Eid NM, Kehoe MJ. The effect of different debonding techniques on the enamel surface: an in vitro qualitative study. Am J Orthod Dentofacial Orthop 1995;108:284-93. 15. Williams OL, Bishara SE. Patient discomfort levels at the time of debonding: a pilot study. Am J Orthod Dentofacial Orthop 1992;101:313-9.

16. Diedrich P. Enamel alterations from bracket bonding and debonding: a study with the scanning electron microscope. Am J Orthod 1981;79:500-22. 17. Benson PE, Pender N, Higham SM, Edgar WM. Morphometric assessment of enamel demineralization from photographs. J Dent 1998;26:669-77. 18. David VA, Staley RN, Bigelow HF, Jakobsen JR. Remnant amount and cleanup for 3 adhesives after debracketing. Am J Orthod Dentofacial Orthop 2002;121:291-6. 19. Pus MD, Way DC. Enamel loss due to orthodontic bonding with lled and unlled resins using various clean-up techniques. Am J Orthod 1980;77:269-83. 20. Thompson RE, Way DC. Enamel loss due to prophylaxis and multiple bonding/debonding of orthodontic attachments. Am J Orthod 1981;79:282-95. 21. Howell S, Weekes WT. An electron microscopic evaluation of the enamel surface subsequent to various debonding procedures. Aust Dent J 1990;35:245-52. 22. OReilly MM, Featherstone JDB. Demineralization and remineralization around orthodontic appliances: an in vivo study. Am J Orthod Dentofacial Orthop 1987;92:33-40. 23. Chung K, Hsu B, Berry T, Hsieh T. Effect of sandblasting on the bone strength of the bondable molar tube bracket. J Oral Rehabil 2001;28:418-24. 24. Levitt HL, Zachrisson BU. Orthodontic bonding. In: Marks MH, Corn H, editors. Atlas of adult orthodontics: functional and esthetic enhancement. Philadelphia: Lea and Febiger; 1989. p. 506-11.