Bonding of ceramic brackets to enamel: Morphologic and stm ctural considerations Theodore Eliades, DDS,* Anthony D. Viazis, DDS, MS,** and George Eliades, DDS, Dr Odont*** Athens, Greece, and Minneapolis, Minn. The purpose of this study was to evalUate the form and microstructure of three types of ceramic bracket and to assess their interracial surface shapes and bond strengths with visible light-cured and chemically cured adhesives after thermocycling. One monocrystalline and two polycrystalline structures were identified. The form of the bonding bases implied three types of bonding with the adhesive: a chemical bonding, a combination of mechanical retention and adhesion, and a combination of micromechanical retention and adhesion. All the ceramic bases were covered with a layer of h,-methacryloxypropyltrimethoxysilane coupling agent. The thickness of the adhesive layer was affected by the design of the bracket bases. The highest bond strength was obtained from the brackets by a combination of micromechanical retention and adhesion, with the site of failure located at the resin/bracket interface. The other types of ceramic brackets had a greater amount of resin left on the enamel and some cases of cohesive bracket fractures. (AM J ORTHOD DENTOFACORTHOP 1991 ;99:369-75.)

The esthetic requirements of orthodontic treatment prompted the development of tooth-colored attachments as alternatives to metal brackets. ~'2 Various types of esthetic bracket, based on polycarbonates and metal-reinforced plastics, had been introduced for direct bonding, but they were found clinically inefficient because of their high debonding rates. ~Recently, the new generation of ceramic brackets became available. 2"aAlthough these appliances are widely used, limited information exists regarding their structure and performance. 5"7In addition, limited research data are available regarding the contribution of visible-light curing to the adhesive strength of ceramic brackets. 7 Visible-light curing has been proposed for orthodontic bonding because it provides the unique advantages of accurate bracket placement, controlled working time, and enhanced peripheral sealing that result from its low oxygen inhibition, s The purpose of this in vitro investigation was to evaluate the form and microstructure of the bonding bases of some ceramic brackets and to assess their interfacial topographies and bond strengths with etched

*Research Fellow, Research Center for Biomaterials, Athens, Greece. **Assistant Professor, Deparlment of Orthodontics, Baylor School of Dentistry, Dallas, Texas. Former Assistant Professor, Division of Orthodontics, School of Dentistry, University of Minnesota. ***Director of Quality Control, Research Center for Biomaterials, Athens, Greece. 8/1/19060

enamel in combination with self-cured and visiblelight-cured orthodontic adhesives.

MATERIALS AND METHODS Three adhesive systems (Table I) and four types of bracket (Table II) were included in the study. The form and microstructure of the brackets' bonding bases was investigated with an electron probe microanalyzer (JXA 733 Superprobe Electron Probe Microanalyzer, Jeol Ltd, Tokyo, Japan) by secondary imaging (SEI), and by topographic and compositional backscattered electron imaging at 20 kv accelerating voltage. The chemical composition of the uppermost layer of the bracket bases was investigated with radiographic photoelectron spectroscopy provided by an ESCA spectrometer (Perkin-Elmer PHI 5400 ESCA system, Physical Electronics Division, Eden Prairie, Minn.), which was equipped with a variable-aperture lens capable of analyzing small areas. Three areas on the bracket bases, 1.0 mm in diameter, were chosen randomly and analyzed by survey and high-resolution spectra with a Mg Kc~ anode at 250 W and 45 ° take-off angle. One hundred forty-four extracted incisors, kept in normal saline solution, were used for the evaluation of the topography and strength of the bracket-enamel interfaces. The teeth were embedded in fast-setting acrylic resin, with their labial surfaces left free, and divided into three groups. Each group was then divided into four subgroups of 12 specimens apiece. The enamel surfaces were ground fiat with silicon carbide papers to

369

370

Am. J. Orthod.Dentofac.Orthop. April 1991

Eliades, Viazis, and Eliades

! ROD

BONDED BRACKET

ACRYLIC BLOCK

LOAD CELL Fig. 2. Bonding base from a Allure III ceramic bracket. (Secondary electron imaging. Magnification x 40.) Fig. 1. Assembly for shear testing. Table II. Brackets used in study Table I. Adhesives used in the study

Adhesive

l

Type

l Group I Manufacturer

HeliositOrthodontic

Light cured

I

VP-862

Light cured

It

Concise Orthodontic

Chemically cured

lit

Vivadent, Liechtenstein, West Germany Vivadent, Liechtenstein, West Germany 3M Co, St. Paul, Minn.

600 grit, subjected to prophylaxis with a fluoride-free pumice, and then acid-etched with 37% orthophosphoric acid gel for 60 seconds. The first subgroup was bonded with Allure brackets (GAC International, Central Islip, N.Y.) (subgroup A), the second with Starfire brackets ("A"-Company, San Diego, Calif.) (subgroup B), and the third with Transcend brackets (Unitek 3M, Monrovia, Calif.) (subgroup C). The fourth subgroup was treated with metal mesh-back brackets and considered as the control group. All brackets of the first group were bonded to the etched enamel with HeliositOrthodontic, while the brackets of the second group were bonded with an experimental fluoride-releasing visible-light-cured adhesive (VP 862). Both these adhesives were exposed sequentially to an Elipar light (Espe, Seefeld, West Germany) from cervical and incisal directions for 2 × 20 seconds. In the third group a chemically cured two-paste system was used. All samples were immersed in normal saline solution at 37 ± 1° C for 1 week and allowed to reach equilibrium, after which they were thermocycled 200

Bracket

l

Type

l Subgroup I Manufacturer

Allure Ill

Ceramic

A

Starfire

Ceramic

B

Transcend

Ceramic

C

Dynabond

Metal

D

GAC International, Central Islip, N.Y. "A"-Company, San Diego, Calif. Unitek/3M, Monrovia, Calif. Unitek/3M, Monrovia, Calif.

times at between 5 ° and 60 ° C and at a frequency of 2 cycles per minute, to simulate accelerated aging by thermally induced stress. Two samples from each group and subgroup were cross-sectioned, polished to 600 grit, and bracket-enamel interface was examined in the microanalyzer. The remaining samples were sheared off in a universal testing machine (Tensometcr 10, Monsanto, England) at a crosshead speed of 1 mm per minute (Fig. 1). The fractured surfaces were examined under a stereomicroscope to reveal the types of failure. The failures were characterized as follows: Type I, for adhesive resin fractures at the bracket-resin interface (IA) and the resin-enamel interface (IB); Type II, for cohesive-enamel fractures; type III, for cohesive-resin fractures; and type IV, for cohesive-bracket fracture. The statistical analysis of the tested parameters was performed by one-way analysis of variance (ANOVA) and Scheffe's F test at a significance level o f p < 0.05.

Volume 99 Number 4

Bonding of ceramic brackets to etzamel 371

Fig. 3. A detail of the polycrystalline structure of a base from a Allure III ceramic bracket. (Secondary electron imaging. Magnification x 1000.)

Fig. 5. Topographical backscattered electron imaging image of the base of a Transcend (Unitek/3M) ceramic bracket. (Magnification x 480.)

o

'D

"- -/21- Q

-

-

•

ru

Fig. 4. Secondary electron image of the monocrystalline structure of the base of a Starfire ceramic bracket. (Magnification x 1000.)

Fig. 6. Compositional backscattered electron imaging image of the base of a Transcend bracket at the same region of Fig. 5. (Magnification x 480.)

RESULTS

Fig. 2 shows detail from the secondary electron image of a bracket-bonding base from subgroup A. The base is perforated with six major symmetric recesses. The area between the recesses is smooth and flat, of a polycrystalline structure composed of irregular grain boundaries at a random distribution (Fig. 3). Detail from the bracket-bonding bases used in subgroup B is presented in Fig. 4. The surface is flat and smooth, but with a total absence of crystalline boundaries. Some waveform lines detected on the surface can probably be attributed to the finishing process. Figs. 5 and 6 show the topographic and compositional backscattercd electron images of a bracket base from subgroup C. The polycrystalline consistency of the base is apparent. However, the surface is rough,

Fig. 7. A detail from the compositional backscattered electron imaging image of a Transcend bracket at higher magnification. (Magnification x 1000.)

372

Am. J. Orthod. Dentofac. Orthop. April 1991

Eliades, Viazis, and Eliades

deft flCQ1]ME=3.67 .in

ESCA SURVEY ANGLE= FILE: Eliades_27 Saeple X SCALE FACTOR. OFFSE]= I0.7S0.

I

I

10 9

I

l

1.700 k cls P~SS ENERGY= 89.'150 eV. Mg 250R

I

I

~

l

I

I

l

l

I

I

I

I

I

t~ oJ

o.

I

¢_

8 7 o

G

U

d.

5 v

4 3 2 1 0

I

1100.0

990.0

I

I

880.0

I

I

770.0

I

I

GGO.0

I

I

I

550.0

IIINI)ING ENERGY,

I , I

"140.0

I

330.0

220.0

110.0

0.0

eV

Fig. 8. Radiographic photoelectron spectroscopy survey spectrum of a bonding base of Starfire bracket.

Fig. 9. Interfacial topography of a Starfire bracket bonded to enamel. (Magnification x300.)

and the orientation of the crystals is random, with many edges protruding outward. At higher magnification, the compositional image of the base reveals that part of the crystals is embedded in a radiographically transparent medium (Fig. 7).

A representative spectrum of a subgroup B bracket base is shown in Fig. 8. The uppermost layer of the base is composed of carbon, oxygen, and silicon at atomic concentration ratios 7:2: 1. These values correspond to the structure of -,/-methacryloxypropyltrimethoxysilane. The layer at the tested positions was continuous and of uniform thickness. Similar spectra were recorded for subgroup A. The silane layer in subgroup C had discontinuities, with peaks arising from the ceramic substrate. The secondary image of subgroup B, series 2 interface (Fig. 9), shows a continuous transitional zone free of pores and cracks. Identical images were obtained from all the tested groups. Measurements of the adhesive layer thickness at three different points on each print are exhibited in Table III. Subgroup B manifested the thinnest adhesive zone, while subgroup A showed the thickest. The results from the shear bond strength study are shown in Table IV. There is no statistically significant difference among the orthodontic adhesives for each bracket. However, there are significant differences among the various brackets for each adhesive. The highest values were obtained from subgroup C, folo

Volume 99 Number 4

Bonding of ceramic brackets to enamel

373

Table IIh Average thickness of the adhesive layer (i.tm, mean ___ 1 SD)

Adhesive group

Bracketsubgroup A

I

B

I

c t

I i

i

I

215.30 .4- 25.13 I I I

138.02 ± 2.19

147.21 ± 2.45 I

I i

II

216.51 -~ 28.27

1 1 9 . 2 0 _.I. 1.96

I I

IIl

t

"36.08 ± 26.45 I I

183.89 _~ 1.52

I

i

147.43 ~_l 2.10

193.32 _" 1.72

l I

148.89 ± 1.49 I

190.95 -- 2.09 I I

I

221.86 ± 2.30 ;

Horizontal bars indicate values of statistical significance at p = 0.05

lowed by A and B. The failure pattern was as follows: The fractures in subgroup C were located at the adhesive-enamel interface, combined with partial bracket-resin and resin-cohesive failures. Most of the resin was removed with the bracket base. In one case of subgroup C, enamel cracking was detected after debonding. Subgroups A and B had failures, in addition to Type IA, IB, and III, within the bracket (Type IV). In subgroup D the failures were of types IA and III, in which most of the resin remained attached to enamel. DISCUSSION

The structure and form of the bonding bases varied significantly among the tested brackets. The bracket bases of Allure III (subgroup A) were composed of polycr:ystalline alumina. The design included symmetrical recesses on each base, which were apparently for additional mechanical retention with the polymerized adhesive layer. The base structure between the recesses was fiat and smooth for maintaining close contact with the adhesive. The bracket bases of Starfire (subgroup B) had a smooth, fiat surface of monocrystalline alumina without any retentive grooves or slots for mechnical anchoring. While a macroscopic observation led to the same conclusions for the bracket bases of Transcend (subgroup C), the photomicrographs revealed a rough bonding base composed of aluminum oxide crystals that increased the bonding area dramatically. All the tested bases were found to be covered with a silane layer, which was continuous in the smooth bracket bases of subgroups A and B and had some discontinuities in subgroup C. It is not known whether the noncontinuous layer of the.radiographically trans-

parent medium detected at the secondary image" of Fig. 7 is associated with the findings from radiographic photoelectron spectroscopy. In all the cases, the identified silane was ",/-methacryloxypropyltrimethoxysilane. Silanol groups of activated silane probably adhere to the hydration layer of alumina crystals by means of hydrogen bonding, while methacrylate groups react in a second step with the adhesive resin, forming covalent bonds. 9 Thus a type of chemical bonding is established between the bracket and the adhesive. The choice of "/-methacryloxypropyltrimethoxysilane is based on the compatibility of this specific silane with the reactive methacrylate groups of the adhesives.I° As a result, the propensity for primary bonding between the silane molecule and the adhesives is subtantially increased. The brackets of subgroup B had the thinnest adhesive layer, compared with the other brackets, regardless of the type of adhesive used. A possible explanation is that the flat and smooth surface of the bracket pressed the adhesive resin to a uniform layer. The presence of the recesses and the retentive crystals in subgroups A and D increased the average film thickness. However, increased thickness beyond a critical point may result in a weak interface, caused by greater polymerization shrinkage and thermal expansion of the resin matrix." The bond strength results are similar to the findings of Gwinnett, 3 ~degaard and Segar, 4 and Viazis et al. 5 The combination of micromechanical retention and silane treatment of subgroup C provides the highest bond strength values, even after thermocycling (Table IV). Subgroup C also had the highest rate of failures at the resin-enamel interface, with most of the resin remaining attached to the debonded bracket. It seems that

374

Am. J. Orthod. Dentofac. Orthop. April 1 9 9 1

Eliades, Viazis, and Eliades

Table IV. Results from shear bond strength test (MPa mean +__ I SD) Bracket subgroup Adhesive group

A

c

1.35-',- 1.46 I I

°

h

i

I

I

i

15.91 - 6.53

5,30 ± 1.95

4.59 4- 2.01

I

I

1

II

11.30 4- 4.44

15.30 ± 7.42

6.32 '4- 1.91 I

II1

I

13.69 ± 5.30

5.90 4- 2.71

I

IL.

I

15.55 ---7.53 I

t

5.46 ± 1.68 l

i

7.93 --- 2.98 I

Horizontal bars indicate values of statistical significance at p = 0.05.

the protruding crystals o f the polycrystalline base reinforce the adhesive layer, which fractures at higher values than that o f the resin-etched enamel. In one case, c o h e s i v e - e n a m e l fracture was detected. Subgroups A and B failed at the r e s i n - b r a c k e t and r e s i n - e n a m e l interface. However, in two cases c o h e s i v e - b r a c k e t failures were detected. An explanation o f the bracket failures, especially o f the strong monocrystalline structure, might be based on their fracture toughness and inherent defects o f their structure such as cracks or pores, which contribute to crack growth and propagation.~2~4 The control group o f metal brackets had frequent failures at the a d h e s i v e - b r a c k e t interface, in which most of the adhesive was left on the etched enamel. Apart from the brackets in subgroup B, all other ceramic brackets had significantly greater ( p < 0.05) bond strength values than the metal brackets, regardless o f the adhesive used. The present study supports the work o f previous investigators 35 and agrees with the various explanations regarding bond strengths and sites o f bond failure. It also goes a step further in clarifying the nature o f the high chemical bond attributed to the Transcend ceramic brackets, which is essentially a dual bond o f (1) micromechanical retention, provided b y the protruding crystals o f the polycrystalline alumina, and (2) chemical adhesion, provided by the coupling effect o f the ",/methacryloxypropyltrimethoxysilane layer. CONCLUSIONS The following conclusions may be drawn from this study: 1. All ceramic bracket bases are covered with a

silane layer identified as 3,-methacryloxypropyltfimethoxysilane. 2. The thickness of the adhesive layer is significantly affected by the design o f the bracket base. 3. The highest bond strength values are obtained from the ceramic brackets that combine micromechanical retention, provided by the protruding crystals of the polycrystalline material, along with chemical adhesion, provided by the coupling effect o f the silane layer. 4. There are no statistically significant differences among the mean shear-bond strengths o f the adhesive systems used. 5. Certain ceramic brackets have significantly higher shear-bond strengths (p < 0.05) than do metal brackets. 6. Bracket failure, as well as tooth damage, can occur in vitro when ceramic brackets are used. REFERENCES 1. Viazis AD. Direct bonding of orthodontic brackets--a review. J Pedod 1986;11:1-23. 2. Swartz ML. Ceramic brackets. J Clin Orthod 1988;22:82-8. 3. Gwinnett AJ. A comparison of shear strengths of metal and ceramic brackets. AM J OR'rHODDENTOFACORTHoP 1988;93: 346-8. 4. ~degaard J, Segnar D. Shear bond strength of metal brackets compared with a new ceramic bracket. AMJ OR'mOPDr~zrOFAC OR'nZor' 1988;94:201-6. 5. Viazis AD, DeLong R, Doublas WH, Bevis RR, Speidel TM. Enamel abrasion from ceramic orthodontic brackets: a special case report. Ar,t J OR'I'HOD DENTOFACORTtlOP 1990;98: 103-9.

6. Viazis AD, DeLong R, Bevis RR, Douglas WH, Pintado M. Enamel abrasion from ceramic orthodontic brackets using an artificial oral environment. AM J ORTIIODDENTOFACORTIIOP i 989;98:103-9.

Volume 99 Number 4 7. Viazis AD, Cavanaugh G, Bevis RR. Bond strength of ceramic brackets under shear stress: an in vitro report. AM J ORTHOD DEwrOFACOgnaoP 1990;98:214-2 I. 8. Lekka MP, Papagiannoulis L, Eliades G, Caputo AA. A comparative in vitro study of visible-light cured sealants. J Oral Rehabil 1989;13:287-99. 9. BascomWD. Structure ofsilane adhesion promoter film on glass and metal surfaces. J Macrocol Sci Macromol Chem 1972;5: 792-99. 10. Temin SL, Richards MC. Color-stable compositions. US Patent no. 3991009. 11. Evans LB, Powers JM. Factors affecting in vitro bond strength of no-mix orthodontic cements. AMJ ORTHODDENTOFACORTHOP 1985;87:508-12. 12. Kusy RP. Morphology of polycrystalline alumina brackets and

Bonding o f ceramic brackets to enamel

375

its relationship to fracture toughness and strength. Angle Orthod 1988;58:197-203. 13. Scott GE Jr. Fracture toughness and surface cracks--the key to understand ceramic brackets. Angle Orthod 1988;58:5-8. 14. Phillps HW. The advent of ceramics: the editor's comer. J Clin Orthod 1988;22:69-70. Reprint requests to: Dr. Anthony D. Viazis Assistant Professor Department of Orthodontics Baylor School of Dentistry 3302 Gaston Ave. Dallas, TX 75246