- Email: [email protected]
Etching conditions for resin-modified glass ionomer cement for orthodontic brackets
ORIGINAL ARTICLE
Etching conditions for resin-modified glass ionomer cement for orthodontic brackets Rudolfo M. Valente, DDS,a Waldemar G. de Rijk, PhD, DDS,b James L. Drummond, DDS, PhD, MBA,c and Carla A. Evans, DDS, DMScd Chicago, Ill This study reports the tensile bond strength of orthodontic eyelets (RMO, Inc, Denver, Colo) bonded to human extracted teeth with a resin-modified glass ionomer cement (RMGIC) (Fuji Ortho LC, GC America, Alsip, Ill) and various acid etchants (Etch-37 and All-Etch, Bisco, Schaumburg, Ill; Ultra Etch, 3M Unitek, St Paul, Minn) for enamel preparation before bonding. The enamel etch conditions were as follows: 37% phosphoric acid with silica; 37% phosphoric acid, silica-free; 10% phosphoric acid, silica-free; 10% polyacrylic acid; and unetched enamel. Bond strength was measured by pulling in tension on the eyelet with a 0.018-in steel wire perpendicular to the enamel surface with a testing machine (Instron model 1125, Canton, Mass) at a speed of 2 mm/min. A light-cured resin cement (Transbond XT, 3M Unitek, Monrovia, Calif) applied to enamel etched with 37% phosphoric acid containing silica served as a control. Each group included 30 specimens. The Weibull distribution (m) was used for statistical analysis with a 90% CI. The different etchants used with RMGIC did not affect tensile bond strength. The resin cement group had the highest tensile strength. Significantly lower bond strengths were observed when glass ionomer cement was used to bond orthodontic attachments to nonetched teeth. However, unlike resin cement, RMGIC can bond effectively to etched teeth in a moist environment without an additional bonding agent. (Am J Orthod Dentofacial Orthop 2002;121:516 –20)
R
esin-modified or resin-reinforced glass ionomer cements have gained popularity as alternatives to resin cements for directly bonding orthodontic attachments to teeth. Some investigators have shown that glass ionomer cements bond to teeth effectively in a moist environment without a need to etch or condition the teeth before bonding procedures to achieve clinically acceptable bond strengths.1-4 However, other investigators have shown that, although glass ionomer cements can bond to enamel in a moist environment, the bond strength of this material is significantly lower than that of resin cements when the teeth are not etched before bonding procedures.5,6 Ultimately, it would be desirable to find a material that achieves bond strengths comparable to those of the resin cements while still tolerating moisture during bond procedures. From the University of Illinois at Chicago. a Former orthodontic resident; private practice, Chicago. b Clinical Associate Professor of Biomaterials in Orthodontics. c Professor of Restorative Dentistry. d Professor of Orthodontics and Department Head. Reprint requests to: Carla A. Evans, Department of Orthodontics, College of Dentistry, University of Illinois at Chicago, 801 S Paulina St (M/C 841), Chicago, IL 60612–7211; e-mail, [email protected]. Submitted, May 2001; revised and accepted, October 2001. Copyright © 2002 by the American Association of Orthodontists. 0889-5406/2002/$35.00 ⫹ 0 8/1/122165 doi:10.1067/mod.2002.122165
516
In a 1993 in vitro study,7 the effects of 3 types of acid etchants on the bond strength of 3 dentin bonding systems were investigated. The preparation without silica and oxalate gave the highest bond strengths of the 3 bonding systems, and the acid etchant containing oxalate gave the lowest bond strength. These findings suggested that the bond strengths varied because of minerals or salts in the acid etchants deposited on the dentin surface. This surface debris could reduce contact between bonding material and dentin surface, thereby compromising bond strength. These results led the investigators to suggest that if dentin is to be treated with acid, the acid etchant should be free of silica and oxalate. The purpose of this in vitro study was to investigate how different acid etch preparations and concentrations affect the tensile bond strength of a visible light-cured resin-modified glass ionomer cement (RMGIC) (Fuji Ortho LC, GC America Inc, Alsip, Ill) for bonding orthodontic attachments. We also examined the amount of adhesive remaining on the tooth surface and its location after the attachments were debonded. These values and observations were compared with those obtained when a visible light-cured resin cement (Transbond XT, 3M Unitek, Monrovia, Calif) was used for bonding the attachments.
Valente et al 517
American Journal of Orthodontics and Dentofacial Orthopedics Volume 121, Number 5
MATERIAL AND METHODS
The crowns of 180 molars, without caries and decalcification, were embedded in acrylic blocks with the buccal surfaces exposed; the specimens were stored in 10% formalin until testing. A total of 6 test groups were studied, each consisting of 30 teeth. Just before placing the orthodontic attachments, the teeth were rinsed with tap water, the exposed buccal surface was cleaned for 30 seconds with a nonfluoride oil-free pumice paste with a prophy cup attached to a slow-speed hand piece, and the teeth were rinsed again and dried for 2 seconds with an oil-free compressed air nozzle. The procedures specified for each experimental group were then followed. The 6 conditions evaluated in this study were as follows. For group 1, the enamel was etched for 30 seconds with 37% phosphoric acid containing silica and rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. For group 2, the enamel was etched for 30 seconds with 37% phosphoric acid that was silica-free and rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. The enamel of group 3 was etched for 30 seconds with 10% phosphoric acid in a silica-free form and rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. For group 4, the enamel was etched with 10% polyacrylic acid for 30 seconds and rinsed with tap water, allowing the surface to remain moist for bonding with the glass ionomer cement. The enamel of group 5 was not etched but was rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. For group 6, the enamel was bonded with a 1-step light-cured resin cement (Transbond XT, 3M Unitek) following the manufacturer’s instructions (tooth surface etched for 30 seconds with 37% phosphoric acid with silica; tooth rinsed and then dried for 2 seconds with an oil-free air stream; primer applied to tooth surface; attachment coated with bonding material, bonded to the tooth surface, and light-cured for a total of 20 seconds). The teeth in groups 1 through 5 were light-cured for 10 seconds each on the gingival, interproximal, and occlusal aspects for a total of 40 seconds with an Ortholux-XT visible light-curing unit (3M Dental Products, St Paul, Minn). The specimens in group 6 served as controls. All teeth were allowed to bench cure for 10 minutes before being placed in a 37°C water bath for 24 hours before testing. The attachments that were used were mesh-backed orthodontic eyelets (RMO, Inc, Denver, Colo). Each sample was placed in a universal testing jig to allow complete freedom of rotation and movement in 3
dimensions, to test tensile strength. The jig was then secured in the vise of a universal testing machine (Instron model 1125, Canton, Mass). A 14-in long 0.018 –in round stainless steel wire was threaded through the eyelet, and the ends were secured in an opposing vise, creating a loop. The crosshead speed was set at 2 mm/min. The tensile bond strength was determined by dividing the fracture load by the surface area of the bonded attachment. After debonding, the location of adhesive was scored with the following scale: A, all cement remained on the tooth surface; B, less than 50% of the cement remained on the tooth surface; C, more than 50% of the cement remained on the tooth surface; D, all cement was debonded with the eyelet. RESULTS
The load at fracture measured in kilograms was determined for every specimen in each group. The raw data were converted to megapascals (MPa) by dividing fracture load in newtons by the eyelet base area (3.0 ⫻ 3.5 mm). The data were used to derive the Weibull modulus and the characteristic strength (SO). Under a given load, Pf ⫽ 1 – exp(S/SO)m, where Pf is the probability of failure, S is the load applied, SO is a constant known as the characteristic value or scale parameter, and m is a constant known as the Weibull modulus or shape parameter. The Weibull modulus represents the underlying flaw distribution that is a measure of the closeness of the group data points. High values for m indicate a small range within a specific group and less variation. Therefore, the Weibull modulus gives a good indication of the reliability of a cement or a bonding agent. A low value indicates a wide distribution of results and a higher probability that a given specimen will have a low strength. The Weibull distribution also allows for censored data; this is helpful because there were cases of enamel or attachment fracture during test procedures. In this study, 1 instance of enamel tear-out occurred, in which the enamel under the eyelet pad fractured from the rest of the tooth before the eyelet attachment could be debonded (this occurred in group 2, in which enamel was etched with 37% phosphoric acid without silica before bonding with glass ionomer cement). In another test group (group 4, in which enamel was etched with 10% polyacrylic acid before bonding with glass ionomer cement), an eyelet fractured from the attachment base pad before debonding from the tooth surface. The Weibull distribution is an extreme value (weakest link) distribution that is ideally suited for this type of fracture data.5,8 With the use of Newton-Raphson iteration, the Weibull modulus, and
518 Valente et al
Table I.
American Journal of Orthodontics and Dentofacial Orthopedics May 2002
Bond strength after enamel surface exposure to various acid etch preparations Bonding material
Bond strength (MPa)
CI (90%)
Shape parameter (m)
CI (m) (90%)
Glass ionomer cement Glass ionomer cement Glass ionomer cement Glass ionomer cement Glass ionomer cement Composite resin
10.5b 10.7b 11.3b 11.1b 6.9a 12.5c
10.0-11.0 10.2-11.1 11.0-11.6 10.7-11.5 5.7-8.2 11.8-13.2
7.0 7.4 10.7 9.5 1.8 6.4
5.2-8.5 5.2-9.0 8.1-13.1 7.1-11.7 1.4-2.3 4.7-7.7
⬎ 50% cement on tooth surface
All cement on eyelet surface
Fracture before debonding of eyelet
1
Preparation 37% Phosphoric acid with silica 37% Phosphoric acid without silica 10% Phosphoric acid without silica 10% Polyacrylic acid Nonetched enamel 37% Phosphoric acid with silica
Letters indicate statistically different values at 90% CI. m, Weibull modulus. Table II.
Location of adhesive after bond failure
Preparation 37% Phosphoric acid with silica 37% Phosphoric acid without silica 10% Phosphoric acid without silica 10% Polyacrylic acid Nonetched enamel 37% Phosphoric acid with silica Total
⬍ 50% cement on tooth surface
Bonding material
All cement on tooth surface
Glass ionomer cement
27
2
Glass ionomer cement
23
6
Glass ionomer cement
18
1
11
Glass ionomer cement Glass ionomer cement Composite resin
23 2 20
10 2
6 9 8
9
113
13
42
10
the characteristic strength, we determined the value at which 63% of the bonds failed. The tensile bond strengths and the shape parameters, or Weibull moduli, with 90% CIs, are summarized in Table I. The control group (group 6), in which eyelets were bonded to the teeth with RMGIC following the manufacturer’s instructions, had the highest tensile bond strengths, with a mean value of 12.5 MPa. Group 5 (bonded with RMGIC to a nonetched enamel surface) had the lowest tensile bond strengths, with a mean value of 6.9 MPa. This group also had the lowest value for the Weibull modulus (1.8 MPa) and a wide range of values in the failure data. Groups 5 and 6 were significantly different from each other and from groups 1 through 4 with respect to tensile bond strength (90% CI). The differences in tensile bond strength among groups 1 through 4 were not statistically significant at an 81% CI. The only statistically different value for the Weibull modulus was obtained from group 5. The analysis of the site of fracture and the location of the remaining adhesive showed that the test group with the most bonding material left on the tooth surface after debonding was group 1 (teeth etched with 37% phosphoric acid with silica and bonded with RMGIC).
1 enamel tearout
1 eyelet failure
2
Group 5, in which the teeth were not etched but were bonded with RMGIC, showed less bonding material remaining on the tooth surface after debonding; this was also the group with the lowest tensile bond strength. However, a nonparametric Spearman correlation analysis showed that the amount of cement on the surface correlates to bond strength at a P value of .295, far from P ⫽ .05. Therefore, there is no strong correlation between the surface cover of the cement and the tensile bond strength. The location and amount of debonded adhesive are summarized in Table II. As stated previously, an instance of enamel tear-out occurred (in group 2), and an instance of eyelet failure occurred (in group 4) during test procedures. The corresponding tensile bond strengths were not used to calculate means and SDs; however, they were used as censored data, ie, a lower-limit estimate in the Weibull calculations. DISCUSSION
Glass ionomer cements have gained popularity for directly bonding orthodontic attachments to teeth because this material can be used effectively in a moist environment. Its chemical composition requires mois-
American Journal of Orthodontics and Dentofacial Orthopedics Volume 121, Number 5
ture. Conversely, without the use of a bonding agent, resin cements do not tolerate moisture well, and resin cement bond failures are often a result of moisture contamination during bonding procedures. Ultimately, the orthodontic practitioner desires a bonding adhesive that provides the highest bond strength possible without harming the tooth surface, especially during debonding. It would be undesirable for an adhesive to bond so well to the tooth surface that debonding procedures would be lengthened or made more difficult, causing discomfort to the patient and damage to the teeth during appliance removal. The practitioner seeks bond strengths that are high enough to allow treatment to progress predictably and without delays caused by appliance breakage. In a previous study conducted in part by Valente,5 it was found that RMGIC approaches the strength observed for conventional resin cements while tolerating moisture from water, saliva, and saliva substitute. In that study, the teeth were conditioned with 10% polyacrylic acid before bonding with glass ionomer cement. One hypothesis of the present study was that the tensile bond strength of RMGIC would be greater because of increased mechanical bonding if the teeth were first etched with a stronger acid (phosphoric acid). This acid is traditionally used when bonding orthodontic brackets with resin cement. Because RMGIC contains polyacrylic acid, it was believed that bond strength could be enhanced by etching the teeth with phosphoric acid and that the chemical bond would be supplemented by micromechanical adhesion provided by phosphoric acid. The results of this study showed no significant difference in tensile bond strength when phosphoric acid or 10% polyacrylic acid was used to etch the tooth surface before bonding. First, etching with 10% phosphoric acid was compared with etching with 10% polyacrylic acid. The bond strengths were 11.3 and 11.1 MPa, respectively; there was no significant difference. It was also hypothesized that a stronger version of the acid, 37% phosphoric acid, would significantly increase tensile bond strength of the glass ionomer cement, but it did not. Mean tensile bond strength was 10.5 MPa when enamel was prepared with 37% phosphoric acid containing silica and 10.7 MPa with 37% phosphoric acid without silica. Another hypothesis of this investigation was sparked by a 1993 study by Kanca,7 previously described. Although the structures of enamel and dentin are different, it was still thought that various minerals or salts in the etchant preparation, possibly deposited onto the enamel surface, would reduce contact between the bonding material and the enamel surface and thus reduce the tensile bond strength of the glass ionomer
Valente et al 519
cement. However, results for test groups etched with silica-free phosphoric acid and silica-containing acid were not significantly different. As the specimens were being prepared, it was observed that the acid etchants thickened with silica remained where applied on the tooth surface without spreading. However, some silicafree acid spread onto the surrounding acrylic, raising concerns of inadvertent soft-tissue etching and increased risk of ingestion if used intraorally. It was hypothesized that more bonding material would remain on the tooth surface if it were etched with phosphoric acid rather than polyacrylic acid before bonding. It was thought that the roughened enamel surface created by the stronger phosphoric acid would cause greater adhesion with its micromechanical lock and would result in more adhesive tags after the attachments were debonded. This hypothesis was also refuted. Although the test group etched with 37% phosphoric acid with silica had the most teeth in which all RMGIC remained on the surface after debonding (27 of 30 samples), the results were similar to those of other test groups. For example, in group 2 (teeth etched with 37% phosphoric acid without silica), 23 of 30 specimens retained all of the glass ionomer cement on the tooth surface. The same result was achieved when teeth were etched with 10% polyacrylic acid before bonding. Statistical analysis showed that any correlation of the amount of cement on the surface correlates to bond strength at P ⫽ .295. Therefore, there was no strong correlation between the surface cover of the cement and the tensile bond strength of the materials tested. In this study, there was an instance of enamel tear-out and an instance of the attachment fracturing before debonding. The tear-out was probably due to the tooth’s structural weakness and not to dangerously high bond strengths. Possible explanations for enamel weakness include advanced age of the tooth, prolonged storage time, and tooth damage during extraction. The fracture of the attachment before debonding was probably caused by a weak weld joint on the attachment. Etching teeth with phosphoric acid rather than polyacrylic acid before bonding with RMGIC did not produce significantly higher tensile bond strength, possibly because polyacrylic acid contains functional groups potentially capable of hydrogen bonding to the tooth surface,9 thus facilitating cleaning and wetting of the enamel surface before bonding. This could enhance bond strength. The polyacrylic acid in RMGIC itself might not be at a sufficiently high concentration to facilitate cleaning and wetting the tooth surface. Therefore, even though the phosphoric acid provided increased micromechanical retention, its effectiveness
520 Valente et al
was offset by its inability to increase the readiness of the tooth surface to bond with the glass ionomer cement. Tensile tests were chosen as the mode for measuring bond strength in this study. Shear tests were not conducted because finite element analyses of bond strength protocols10,11 showed that when shear forces are applied to specimens at a distance from the adhesive interface, tensile and compressive forces can occur, so that a true shear strength value cannot be obtained.5 This study did not attempt to reproduce conditions in a patient’s mouth. Instead, it was designed to quantify the inherent strength of the bond of a material to a tooth surface under various scenarios and to compare it with the strength of a control material that traditionally has been used for bonding orthodontic attachments. Although superior tensile strengths were achieved with resin cement versus RMGIC, RMGIC might still be strong enough to be an acceptable alternative for bonding orthodontic appliances in patients under certain circumstances. Because RMGIC bonds well in a moist environment, it might be useful in bonding orthodontic attachments to teeth in areas of the mouth where dry isolation is difficult (eg, second molars, surgically exposed teeth, or the lingual surfaces of mandibular teeth). CONCLUSIONS
The following conclusions can be made from this in vitro study: (1) the etchants used in this study did not affect the tensile bond strength of RMGIC; (2) it is still necessary to prepare the tooth surface with an etchant before bonding procedures to achieve acceptable bond strengths; (3) significantly lower tensile bond strengths occur when RMGIC is used to bond orthodontic attachments to nonetched teeth; (4) resin cement’s tensile bond strength was superior to that of RMGIC; and (5) unlike resin cements, RMGIC can bond effectively to
American Journal of Orthodontics and Dentofacial Orthopedics May 2002
etched teeth under moist conditions without the need for an additional bonding agent to promote bond strength. We thank GC America Inc, 3M Unitek, Bisco, and RMO, Inc, for their generous donations of materials to the Department of Orthodontics, University of Illinois at Chicago, for this project. REFERENCES 1. Silverman E, Cohen M, Demke RS, Silverman M. A new light-cured glass ionomer cement that bonds brackets to teeth without etching in the presence of saliva. Am J Orthod Dentofacial Orthop 1995;108:231-6. 2. Cacciafesta V, Bosch C, Melson B. Clinical comparison between a resin-reinforced self-cured glass ionomer cement and a composite resin for direct bonding of orthodontic brackets. Part 1: Wetting with water. Clin Orthod Res 1998;1:29-36. 3. Cacciafesta V, Bosch C, Melson B. Clinical comparison between a resin reinforced self-cured glass ionomer cement and a composite resin for direct bonding of orthodontic brackets. Part 2: Bonding on dry enamel soaked with saliva. Clin Orthod Res 1999;2:186-93. 4. Cacciafesta V, Jost-Brinkman P, Subenberger U, Miethke R. Effects of saliva and water contamination on the enamel shear bond strength of a light-cured glass ionomer cement. Am J Orthod Dentofacial Orthop 1998;113:402-7. 5. Jobalia SB, Valente RM, de Rijk WG, Be Gole EA, Evans CA. Bond strength of visible light-cured glass ionomer cement. Am J Orthod Dentofacial Orthop 1997;112:205-8. 6. Tringas AJ. Clinical trial of orthodontic bonding agents [thesis]. Chicago: University of Illinois; 1998. 7. Kanca J. Etchant composition and bond strength to dentin. Am J Dent 1993;6:287-90. 8. Lawless JF. Statistical models and methods for lifetime data. New York: John Wiley; 1982. 9. Powis DR, Folleras T, Merson SA, Wilson AD. Improved adhesion of a glass ionomer cement to dentin and enamel. J Dent Res 1982;61:1416-22. 10. Versluis A, Douglas WH. Why do shear tests pull out dentin? J Dent Res 1996;73:177. 11. Thomas R, de Rijk WG, Evans CA. Tensile and shear stresses in the orthodontic attachment layer using 3D finite element analysis. Am J Orthod Dentofacial Orthop 1999;116:530-2.
Etching conditions for resin-modified glass ionomer cement for orthodontic brackets Rudolfo M. Valente, DDS,a Waldemar G. de Rijk, PhD, DDS,b James L. Drummond, DDS, PhD, MBA,c and Carla A. Evans, DDS, DMScd Chicago, Ill This study reports the tensile bond strength of orthodontic eyelets (RMO, Inc, Denver, Colo) bonded to human extracted teeth with a resin-modified glass ionomer cement (RMGIC) (Fuji Ortho LC, GC America, Alsip, Ill) and various acid etchants (Etch-37 and All-Etch, Bisco, Schaumburg, Ill; Ultra Etch, 3M Unitek, St Paul, Minn) for enamel preparation before bonding. The enamel etch conditions were as follows: 37% phosphoric acid with silica; 37% phosphoric acid, silica-free; 10% phosphoric acid, silica-free; 10% polyacrylic acid; and unetched enamel. Bond strength was measured by pulling in tension on the eyelet with a 0.018-in steel wire perpendicular to the enamel surface with a testing machine (Instron model 1125, Canton, Mass) at a speed of 2 mm/min. A light-cured resin cement (Transbond XT, 3M Unitek, Monrovia, Calif) applied to enamel etched with 37% phosphoric acid containing silica served as a control. Each group included 30 specimens. The Weibull distribution (m) was used for statistical analysis with a 90% CI. The different etchants used with RMGIC did not affect tensile bond strength. The resin cement group had the highest tensile strength. Significantly lower bond strengths were observed when glass ionomer cement was used to bond orthodontic attachments to nonetched teeth. However, unlike resin cement, RMGIC can bond effectively to etched teeth in a moist environment without an additional bonding agent. (Am J Orthod Dentofacial Orthop 2002;121:516 –20)
R
esin-modified or resin-reinforced glass ionomer cements have gained popularity as alternatives to resin cements for directly bonding orthodontic attachments to teeth. Some investigators have shown that glass ionomer cements bond to teeth effectively in a moist environment without a need to etch or condition the teeth before bonding procedures to achieve clinically acceptable bond strengths.1-4 However, other investigators have shown that, although glass ionomer cements can bond to enamel in a moist environment, the bond strength of this material is significantly lower than that of resin cements when the teeth are not etched before bonding procedures.5,6 Ultimately, it would be desirable to find a material that achieves bond strengths comparable to those of the resin cements while still tolerating moisture during bond procedures. From the University of Illinois at Chicago. a Former orthodontic resident; private practice, Chicago. b Clinical Associate Professor of Biomaterials in Orthodontics. c Professor of Restorative Dentistry. d Professor of Orthodontics and Department Head. Reprint requests to: Carla A. Evans, Department of Orthodontics, College of Dentistry, University of Illinois at Chicago, 801 S Paulina St (M/C 841), Chicago, IL 60612–7211; e-mail, [email protected]. Submitted, May 2001; revised and accepted, October 2001. Copyright © 2002 by the American Association of Orthodontists. 0889-5406/2002/$35.00 ⫹ 0 8/1/122165 doi:10.1067/mod.2002.122165
516
In a 1993 in vitro study,7 the effects of 3 types of acid etchants on the bond strength of 3 dentin bonding systems were investigated. The preparation without silica and oxalate gave the highest bond strengths of the 3 bonding systems, and the acid etchant containing oxalate gave the lowest bond strength. These findings suggested that the bond strengths varied because of minerals or salts in the acid etchants deposited on the dentin surface. This surface debris could reduce contact between bonding material and dentin surface, thereby compromising bond strength. These results led the investigators to suggest that if dentin is to be treated with acid, the acid etchant should be free of silica and oxalate. The purpose of this in vitro study was to investigate how different acid etch preparations and concentrations affect the tensile bond strength of a visible light-cured resin-modified glass ionomer cement (RMGIC) (Fuji Ortho LC, GC America Inc, Alsip, Ill) for bonding orthodontic attachments. We also examined the amount of adhesive remaining on the tooth surface and its location after the attachments were debonded. These values and observations were compared with those obtained when a visible light-cured resin cement (Transbond XT, 3M Unitek, Monrovia, Calif) was used for bonding the attachments.
Valente et al 517
American Journal of Orthodontics and Dentofacial Orthopedics Volume 121, Number 5
MATERIAL AND METHODS
The crowns of 180 molars, without caries and decalcification, were embedded in acrylic blocks with the buccal surfaces exposed; the specimens were stored in 10% formalin until testing. A total of 6 test groups were studied, each consisting of 30 teeth. Just before placing the orthodontic attachments, the teeth were rinsed with tap water, the exposed buccal surface was cleaned for 30 seconds with a nonfluoride oil-free pumice paste with a prophy cup attached to a slow-speed hand piece, and the teeth were rinsed again and dried for 2 seconds with an oil-free compressed air nozzle. The procedures specified for each experimental group were then followed. The 6 conditions evaluated in this study were as follows. For group 1, the enamel was etched for 30 seconds with 37% phosphoric acid containing silica and rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. For group 2, the enamel was etched for 30 seconds with 37% phosphoric acid that was silica-free and rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. The enamel of group 3 was etched for 30 seconds with 10% phosphoric acid in a silica-free form and rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. For group 4, the enamel was etched with 10% polyacrylic acid for 30 seconds and rinsed with tap water, allowing the surface to remain moist for bonding with the glass ionomer cement. The enamel of group 5 was not etched but was rinsed with tap water, allowing the surface to remain moist for bonding with RMGIC. For group 6, the enamel was bonded with a 1-step light-cured resin cement (Transbond XT, 3M Unitek) following the manufacturer’s instructions (tooth surface etched for 30 seconds with 37% phosphoric acid with silica; tooth rinsed and then dried for 2 seconds with an oil-free air stream; primer applied to tooth surface; attachment coated with bonding material, bonded to the tooth surface, and light-cured for a total of 20 seconds). The teeth in groups 1 through 5 were light-cured for 10 seconds each on the gingival, interproximal, and occlusal aspects for a total of 40 seconds with an Ortholux-XT visible light-curing unit (3M Dental Products, St Paul, Minn). The specimens in group 6 served as controls. All teeth were allowed to bench cure for 10 minutes before being placed in a 37°C water bath for 24 hours before testing. The attachments that were used were mesh-backed orthodontic eyelets (RMO, Inc, Denver, Colo). Each sample was placed in a universal testing jig to allow complete freedom of rotation and movement in 3
dimensions, to test tensile strength. The jig was then secured in the vise of a universal testing machine (Instron model 1125, Canton, Mass). A 14-in long 0.018 –in round stainless steel wire was threaded through the eyelet, and the ends were secured in an opposing vise, creating a loop. The crosshead speed was set at 2 mm/min. The tensile bond strength was determined by dividing the fracture load by the surface area of the bonded attachment. After debonding, the location of adhesive was scored with the following scale: A, all cement remained on the tooth surface; B, less than 50% of the cement remained on the tooth surface; C, more than 50% of the cement remained on the tooth surface; D, all cement was debonded with the eyelet. RESULTS
The load at fracture measured in kilograms was determined for every specimen in each group. The raw data were converted to megapascals (MPa) by dividing fracture load in newtons by the eyelet base area (3.0 ⫻ 3.5 mm). The data were used to derive the Weibull modulus and the characteristic strength (SO). Under a given load, Pf ⫽ 1 – exp(S/SO)m, where Pf is the probability of failure, S is the load applied, SO is a constant known as the characteristic value or scale parameter, and m is a constant known as the Weibull modulus or shape parameter. The Weibull modulus represents the underlying flaw distribution that is a measure of the closeness of the group data points. High values for m indicate a small range within a specific group and less variation. Therefore, the Weibull modulus gives a good indication of the reliability of a cement or a bonding agent. A low value indicates a wide distribution of results and a higher probability that a given specimen will have a low strength. The Weibull distribution also allows for censored data; this is helpful because there were cases of enamel or attachment fracture during test procedures. In this study, 1 instance of enamel tear-out occurred, in which the enamel under the eyelet pad fractured from the rest of the tooth before the eyelet attachment could be debonded (this occurred in group 2, in which enamel was etched with 37% phosphoric acid without silica before bonding with glass ionomer cement). In another test group (group 4, in which enamel was etched with 10% polyacrylic acid before bonding with glass ionomer cement), an eyelet fractured from the attachment base pad before debonding from the tooth surface. The Weibull distribution is an extreme value (weakest link) distribution that is ideally suited for this type of fracture data.5,8 With the use of Newton-Raphson iteration, the Weibull modulus, and
518 Valente et al
Table I.
American Journal of Orthodontics and Dentofacial Orthopedics May 2002
Bond strength after enamel surface exposure to various acid etch preparations Bonding material
Bond strength (MPa)
CI (90%)
Shape parameter (m)
CI (m) (90%)
Glass ionomer cement Glass ionomer cement Glass ionomer cement Glass ionomer cement Glass ionomer cement Composite resin
10.5b 10.7b 11.3b 11.1b 6.9a 12.5c
10.0-11.0 10.2-11.1 11.0-11.6 10.7-11.5 5.7-8.2 11.8-13.2
7.0 7.4 10.7 9.5 1.8 6.4
5.2-8.5 5.2-9.0 8.1-13.1 7.1-11.7 1.4-2.3 4.7-7.7
⬎ 50% cement on tooth surface
All cement on eyelet surface
Fracture before debonding of eyelet
1
Preparation 37% Phosphoric acid with silica 37% Phosphoric acid without silica 10% Phosphoric acid without silica 10% Polyacrylic acid Nonetched enamel 37% Phosphoric acid with silica
Letters indicate statistically different values at 90% CI. m, Weibull modulus. Table II.
Location of adhesive after bond failure
Preparation 37% Phosphoric acid with silica 37% Phosphoric acid without silica 10% Phosphoric acid without silica 10% Polyacrylic acid Nonetched enamel 37% Phosphoric acid with silica Total
⬍ 50% cement on tooth surface
Bonding material
All cement on tooth surface
Glass ionomer cement
27
2
Glass ionomer cement
23
6
Glass ionomer cement
18
1
11
Glass ionomer cement Glass ionomer cement Composite resin
23 2 20
10 2
6 9 8
9
113
13
42
10
the characteristic strength, we determined the value at which 63% of the bonds failed. The tensile bond strengths and the shape parameters, or Weibull moduli, with 90% CIs, are summarized in Table I. The control group (group 6), in which eyelets were bonded to the teeth with RMGIC following the manufacturer’s instructions, had the highest tensile bond strengths, with a mean value of 12.5 MPa. Group 5 (bonded with RMGIC to a nonetched enamel surface) had the lowest tensile bond strengths, with a mean value of 6.9 MPa. This group also had the lowest value for the Weibull modulus (1.8 MPa) and a wide range of values in the failure data. Groups 5 and 6 were significantly different from each other and from groups 1 through 4 with respect to tensile bond strength (90% CI). The differences in tensile bond strength among groups 1 through 4 were not statistically significant at an 81% CI. The only statistically different value for the Weibull modulus was obtained from group 5. The analysis of the site of fracture and the location of the remaining adhesive showed that the test group with the most bonding material left on the tooth surface after debonding was group 1 (teeth etched with 37% phosphoric acid with silica and bonded with RMGIC).
1 enamel tearout
1 eyelet failure
2
Group 5, in which the teeth were not etched but were bonded with RMGIC, showed less bonding material remaining on the tooth surface after debonding; this was also the group with the lowest tensile bond strength. However, a nonparametric Spearman correlation analysis showed that the amount of cement on the surface correlates to bond strength at a P value of .295, far from P ⫽ .05. Therefore, there is no strong correlation between the surface cover of the cement and the tensile bond strength. The location and amount of debonded adhesive are summarized in Table II. As stated previously, an instance of enamel tear-out occurred (in group 2), and an instance of eyelet failure occurred (in group 4) during test procedures. The corresponding tensile bond strengths were not used to calculate means and SDs; however, they were used as censored data, ie, a lower-limit estimate in the Weibull calculations. DISCUSSION
Glass ionomer cements have gained popularity for directly bonding orthodontic attachments to teeth because this material can be used effectively in a moist environment. Its chemical composition requires mois-
American Journal of Orthodontics and Dentofacial Orthopedics Volume 121, Number 5
ture. Conversely, without the use of a bonding agent, resin cements do not tolerate moisture well, and resin cement bond failures are often a result of moisture contamination during bonding procedures. Ultimately, the orthodontic practitioner desires a bonding adhesive that provides the highest bond strength possible without harming the tooth surface, especially during debonding. It would be undesirable for an adhesive to bond so well to the tooth surface that debonding procedures would be lengthened or made more difficult, causing discomfort to the patient and damage to the teeth during appliance removal. The practitioner seeks bond strengths that are high enough to allow treatment to progress predictably and without delays caused by appliance breakage. In a previous study conducted in part by Valente,5 it was found that RMGIC approaches the strength observed for conventional resin cements while tolerating moisture from water, saliva, and saliva substitute. In that study, the teeth were conditioned with 10% polyacrylic acid before bonding with glass ionomer cement. One hypothesis of the present study was that the tensile bond strength of RMGIC would be greater because of increased mechanical bonding if the teeth were first etched with a stronger acid (phosphoric acid). This acid is traditionally used when bonding orthodontic brackets with resin cement. Because RMGIC contains polyacrylic acid, it was believed that bond strength could be enhanced by etching the teeth with phosphoric acid and that the chemical bond would be supplemented by micromechanical adhesion provided by phosphoric acid. The results of this study showed no significant difference in tensile bond strength when phosphoric acid or 10% polyacrylic acid was used to etch the tooth surface before bonding. First, etching with 10% phosphoric acid was compared with etching with 10% polyacrylic acid. The bond strengths were 11.3 and 11.1 MPa, respectively; there was no significant difference. It was also hypothesized that a stronger version of the acid, 37% phosphoric acid, would significantly increase tensile bond strength of the glass ionomer cement, but it did not. Mean tensile bond strength was 10.5 MPa when enamel was prepared with 37% phosphoric acid containing silica and 10.7 MPa with 37% phosphoric acid without silica. Another hypothesis of this investigation was sparked by a 1993 study by Kanca,7 previously described. Although the structures of enamel and dentin are different, it was still thought that various minerals or salts in the etchant preparation, possibly deposited onto the enamel surface, would reduce contact between the bonding material and the enamel surface and thus reduce the tensile bond strength of the glass ionomer
Valente et al 519
cement. However, results for test groups etched with silica-free phosphoric acid and silica-containing acid were not significantly different. As the specimens were being prepared, it was observed that the acid etchants thickened with silica remained where applied on the tooth surface without spreading. However, some silicafree acid spread onto the surrounding acrylic, raising concerns of inadvertent soft-tissue etching and increased risk of ingestion if used intraorally. It was hypothesized that more bonding material would remain on the tooth surface if it were etched with phosphoric acid rather than polyacrylic acid before bonding. It was thought that the roughened enamel surface created by the stronger phosphoric acid would cause greater adhesion with its micromechanical lock and would result in more adhesive tags after the attachments were debonded. This hypothesis was also refuted. Although the test group etched with 37% phosphoric acid with silica had the most teeth in which all RMGIC remained on the surface after debonding (27 of 30 samples), the results were similar to those of other test groups. For example, in group 2 (teeth etched with 37% phosphoric acid without silica), 23 of 30 specimens retained all of the glass ionomer cement on the tooth surface. The same result was achieved when teeth were etched with 10% polyacrylic acid before bonding. Statistical analysis showed that any correlation of the amount of cement on the surface correlates to bond strength at P ⫽ .295. Therefore, there was no strong correlation between the surface cover of the cement and the tensile bond strength of the materials tested. In this study, there was an instance of enamel tear-out and an instance of the attachment fracturing before debonding. The tear-out was probably due to the tooth’s structural weakness and not to dangerously high bond strengths. Possible explanations for enamel weakness include advanced age of the tooth, prolonged storage time, and tooth damage during extraction. The fracture of the attachment before debonding was probably caused by a weak weld joint on the attachment. Etching teeth with phosphoric acid rather than polyacrylic acid before bonding with RMGIC did not produce significantly higher tensile bond strength, possibly because polyacrylic acid contains functional groups potentially capable of hydrogen bonding to the tooth surface,9 thus facilitating cleaning and wetting of the enamel surface before bonding. This could enhance bond strength. The polyacrylic acid in RMGIC itself might not be at a sufficiently high concentration to facilitate cleaning and wetting the tooth surface. Therefore, even though the phosphoric acid provided increased micromechanical retention, its effectiveness
520 Valente et al
was offset by its inability to increase the readiness of the tooth surface to bond with the glass ionomer cement. Tensile tests were chosen as the mode for measuring bond strength in this study. Shear tests were not conducted because finite element analyses of bond strength protocols10,11 showed that when shear forces are applied to specimens at a distance from the adhesive interface, tensile and compressive forces can occur, so that a true shear strength value cannot be obtained.5 This study did not attempt to reproduce conditions in a patient’s mouth. Instead, it was designed to quantify the inherent strength of the bond of a material to a tooth surface under various scenarios and to compare it with the strength of a control material that traditionally has been used for bonding orthodontic attachments. Although superior tensile strengths were achieved with resin cement versus RMGIC, RMGIC might still be strong enough to be an acceptable alternative for bonding orthodontic appliances in patients under certain circumstances. Because RMGIC bonds well in a moist environment, it might be useful in bonding orthodontic attachments to teeth in areas of the mouth where dry isolation is difficult (eg, second molars, surgically exposed teeth, or the lingual surfaces of mandibular teeth). CONCLUSIONS
The following conclusions can be made from this in vitro study: (1) the etchants used in this study did not affect the tensile bond strength of RMGIC; (2) it is still necessary to prepare the tooth surface with an etchant before bonding procedures to achieve acceptable bond strengths; (3) significantly lower tensile bond strengths occur when RMGIC is used to bond orthodontic attachments to nonetched teeth; (4) resin cement’s tensile bond strength was superior to that of RMGIC; and (5) unlike resin cements, RMGIC can bond effectively to
American Journal of Orthodontics and Dentofacial Orthopedics May 2002
etched teeth under moist conditions without the need for an additional bonding agent to promote bond strength. We thank GC America Inc, 3M Unitek, Bisco, and RMO, Inc, for their generous donations of materials to the Department of Orthodontics, University of Illinois at Chicago, for this project. REFERENCES 1. Silverman E, Cohen M, Demke RS, Silverman M. A new light-cured glass ionomer cement that bonds brackets to teeth without etching in the presence of saliva. Am J Orthod Dentofacial Orthop 1995;108:231-6. 2. Cacciafesta V, Bosch C, Melson B. Clinical comparison between a resin-reinforced self-cured glass ionomer cement and a composite resin for direct bonding of orthodontic brackets. Part 1: Wetting with water. Clin Orthod Res 1998;1:29-36. 3. Cacciafesta V, Bosch C, Melson B. Clinical comparison between a resin reinforced self-cured glass ionomer cement and a composite resin for direct bonding of orthodontic brackets. Part 2: Bonding on dry enamel soaked with saliva. Clin Orthod Res 1999;2:186-93. 4. Cacciafesta V, Jost-Brinkman P, Subenberger U, Miethke R. Effects of saliva and water contamination on the enamel shear bond strength of a light-cured glass ionomer cement. Am J Orthod Dentofacial Orthop 1998;113:402-7. 5. Jobalia SB, Valente RM, de Rijk WG, Be Gole EA, Evans CA. Bond strength of visible light-cured glass ionomer cement. Am J Orthod Dentofacial Orthop 1997;112:205-8. 6. Tringas AJ. Clinical trial of orthodontic bonding agents [thesis]. Chicago: University of Illinois; 1998. 7. Kanca J. Etchant composition and bond strength to dentin. Am J Dent 1993;6:287-90. 8. Lawless JF. Statistical models and methods for lifetime data. New York: John Wiley; 1982. 9. Powis DR, Folleras T, Merson SA, Wilson AD. Improved adhesion of a glass ionomer cement to dentin and enamel. J Dent Res 1982;61:1416-22. 10. Versluis A, Douglas WH. Why do shear tests pull out dentin? J Dent Res 1996;73:177. 11. Thomas R, de Rijk WG, Evans CA. Tensile and shear stresses in the orthodontic attachment layer using 3D finite element analysis. Am J Orthod Dentofacial Orthop 1999;116:530-2.