Bonding Orthodontic Attachments To Composite Restorations Biology Essay

Published: Last Edited:

This essay has been submitted by a student. This is not an example of the work written by our professional essay writers.

Bonding orthodontic attachments to composite restorations with a composite-based orthodontic adhesive is only known to have been reported a few times previously (Kao et al., 1995; Chunhachteevachaloke and Tyas, 1997; Lai et al., 1999). None of these studies investigated the effects of acid etching or air abrasion on the shear bond strength. Previous literature from the field of composite repair indicated that acid etching of the composite surface did not significantly increase the bond strength (Boyer et al., 1978), suggesting possible elimination of this clinical procedure. In its place, mechanical surface treatment, such as air-abrasion, appears to be more effective in increasing composite to composite bond strengths (Swift et al., 1992; Turner and Meiers, 1993; Kupiec and Barkrneier, 1996; Bouschlicher et al., 1997; Brosh et aL, 1997).

A new cyanoacrylate adhesive (smartbondB, Gestenco International, Goteborg, Sweden) has been developed for orthodontic bonding. The manufacturer claims the product provides sufncient bond strength to composite restorative materials. A previous cyanoacrylate adhesive was tested by Howells and Jones (1989). They found the material to have acceptable handling qualities (easily mixed, satisfactory viscosity, adequate working time), and comparable initial bond strength to a composite adhesive (124N vs. 132N). However, after storage in normal saline for 7 or 98 days, the hydrolysis of the polymerized material weakened the bond (to 6N), rendering it unsuitable for clinical use. At present, no definitive protocol is known for bonding orthodontic attachments to restorative composite resin.


1. To determine the effect of six different adhesive/surface treatment combinations on the shead/peel bond strength of orthodontic attachments to restorative resin after 24 hours and after thennocycling.

2. To determine the effect of thermocycling on the shead/peel bond strength of orthodontic attachments to restorative resin of the six experimental groups.

3. To evaluate the pattern of failure after debonding of orthodontic attachments to restorative resin.

Null hypothesis

The null hypothesis states that there are no statistically significant differences in the shead/peel bond strength between the six adhesive/surface treatment groups. The null hypothesis also states that thermocycling has no effect on the shead/peel bond strengths.



Mechanical bonding or mechanical retention are the simplest methods of achieving a strong attachment between two substances. The first attempts at retention in dentistry involved the placement of undercuts in materials to facilitate the "locking in" of subsequently placed materials.

The ultimate goal in dental bonding is to achieve true adhesion. When two substances are brought into intimate contact with each other, the molecules of one substance may be attracted to molecules of the other. Such an attractive force is known as adhesion when different molecules are involved and cohesion when like molecules are attracted. In dentistry, bonding is achieved by the application of a liquid material (the adhesive) to promote adhesion to a solid substrate (the adherent). For adhesion to occur, the adhesive and adherent surfaces must be attracted to one another at their interface. The surface of a material has greater energy than its interior. At the surface, the outermost atoms are not equally attracted in ail directions by other atoms. While within the material, atoms held in the solid lattice are equally attracted to each other. The increase in energy per unit area of surface is known as the surface energy or surface tension of a material.

The greater the surface energy, the greater the potential for adhesion (Anusavice, 1996).

In addition to a high surface energy, the adhesive must be able to adequately wet the surface of the adherent. Wetting is the manifestation of the attractive forces between molecules of adhesive and adherent and may be defined as the process of obtaining molecular nearness or establishing interfacial contact The extent to which an adhesive is attracted to the surface of an adherent cm be determined by measuring the contact angle. The contact angle is the angle between the adhesive and the adherent at their interface. The smaller the contact angle, the more effective the adhesive is able to wet the surface of the adherent (Anusavice, 1996).


Historical Background of Composite Resin

Synthetic resins have evolved as restorative materials principally because of their esthetic characteristics. The most widely used esthetic materials in dentistry today are composite resins (Gleich, 1999). The term composite means that it is a mechanical mixture of at least two different classes of materials that have limited mutual solubility. The combination of materials provides a product with properties superior or intermediate to those of the individual constituents (Anusavice, 1996). Dental composites are composed of synthetic polymers, inorganic fillers, molecules which promote the polymerization reaction, silane coupling agents to bond the inorganic filler particles to the polymer matrix, pigments, and small amounts of other additives to improve color stability (UV absorbers) and prevent premature polymerization (eg. hydroquinone) (Bayne et al., 1994; Ferracane, 1995; Anusavice, 1996).

Development of dental composites began in the late 1950's and early 1960's when Dr. R.L. Bowen began experiments on reinforcing epoxy resins with filler particles (Bowen, 1963). For the past forty years, 80-90% of composites utilized the Bis-GMA (2.2-bis [4 - (2-hydroxy-3-methacryloyloxy-propoxy) phenyl] propane) monomer developed by Dr. Bowen as the matrix-forming resin (Ruyer and Oysæd, 1987). Bis-GMA is extremely viscous at room temperature because of the hydrogen bonding interactions that occur between the hydroxyl groups on the monomer molecules (Ferracane, 1995). Diluents of a more fluid resin such as TEGDMA (triethylene glycol dimethacrylate), which is also a crosslinking agent, are added to produce pastes of clinically usable consistencies (Ferracane, 1995). Stannard (1993) reported that optimal properties are produced with a, 1: 1 ratio of Bis-GMA and TEGDMA.

Inorganic filler particles are used to improve the strength (Farracane et al., 1987; Eversoll and Moore, 1988), increase stiffness (Braem et al, 1989), change the coefficient of thermal expansion (Sodërholm, 1984; Brown, 1988), reduce polymerization shrinkage (Iga et al., 1991), and increase the radiopacity (van Dijken et al., 1989) of composite resins. Filler particles are most commonly produced by grinding or milling silicate particles containing oxides of barium, strontium, zinc, aluminum, or zirconium (Khan et al., 1992). Composite resins have been classified according to their filler particulate size and percentage (Anusavice, 1996). Microfilled materials have filler sizes of 0.04um and filler loading of 35-60 wt%. Small particle-filled composites have large filler sizes of 1 to 5pm and increased loading of fillers up to 80-90 wt%. Hybrids are a third classification which combines -04 and 1 pm particles in weight percentages in the 75-80 wt% range (Anusavice, 1996). Clinical Research Associates recommends hybrid resin use in Class 5 restorations where there is a high esthetic need (C.R.A., 1999). Their recommendation was based on an ordinal evaluation of 12 clinical characteristics such as color match and surface smoothness. These are the surfaces to which orthodontic bonding will be generally required. Thus, it is likely that the type of composite encountered will be of the hybrid classification. To achieve the optimal properties of the composite, coupling agents are used to chemically bond the filler particles to the resin matrix. Coupling agents displace adsorbed water and provide a strong chemical bond between the oxide groups on the glass filler surface and the polymer molecules of the resin (van Noort, 1994). Silane coupling agents have the general formula:


where R represents an organofunctional group and the X groups are hydrolysable groups bonded to the silane. The X units are hydrolysed and a tri-hydroxy-silanol is produced.

R-Si-X3 + 3H2O ---> R-Si(OH)3 + 3HX

These silanols form hydrogen bonds with other hydroxyl groups on the glass surface. Water is removed upon drying in a condensation reaction to form a covalent bond between the coupling agent and the glass. Finally, the organofiinctional group, R, reacts with the polymer to link the two materials (van Noort, 1994). The most common coupling agents are organosilanes such as 3 -methacryloxypropyl-trimethoxy-silane (MPS) (Ferracane, 1995).

The resins themselves polymerize by an addition mechanism initiated by free radicals (Anusavice, 1996). No byproducts are formed in addition polyrnerization reactions; the structure of the monomer is simply repeated numerous times. An external stimulus is used to activate the addition reaction (Anusavice, 1996). In self-cured resins the induction system is chemically activated by mixing two separate pastes. Typically, one paste contains benzoyl peroxide (the initiator) and the other paste contains an amine, dihydroxyethyl-p-toluidine (DHEPT) which works as an activator (Anusavice, 1996). The most popular composite resins are now light-activated (C.R.A., 1999). The initiator in these products is camphoroquinone (CQ) which generates free radicals when exposed to blue light in the 470nm range of the electromagnetic spectrum (Ferracane, 1995). These products are popular because they are one-part systems, and working time is somewhat controlled by the operator.

Aging and Water Sorption

Composite resins are not inert materials. They have been shown to be soluble in water and in organic solvents such as ethanol (Craig and Payton, 1975; Pearson, 1979; Ferracane and Condon, 1990; Soderholm, 1990; Ferracane, 1994). The leaching of components has a potential impact on both the structural stability and biocompatibility of the material (Ferracane, 1994). Of course, the former is of concern to this study, as the structural stability of the composite may be a factor in the shear bond strength of orthodontic attachments.

It has been estimated by infrared spectroscopy that only 50-75% of the resin monomer double bonds are actually polymerized after light curing (Eliades et al., 1987; Ferracane, 1990, 1994; Ferracane and Condon, 1992). As mentioned previously, the Bis-GMA and TEGDMA monomers polymerize by a free radical addition reaction involving substantial cross-linking. The cross-linking reaction produces a gel structure which severely reduces molecular mobility and greatly slows the rate of polymerization (Ferracane, 1994). Most of the unreacted carbon-carbon double bonds are on molecules which have reacted at one end and are thus bound to the polymer chain and are not free to elute (Ferracane, 1994). However, the polymer ma& does contain a portion of totally free monomer molecules. Studies have verified that virtually all of the components in composites may be leached into solution- Bis-GMA and TEGDMA have been identified in solution by many different sources (Inoue and Hayashi, 1982; Thompson et al., 1982; Ferracane and Condon, 1990; Rathbum et al-, 1991; Tanaka et al., 1991).

Approximately 5-10% of the unbound monomer elutes into an aqueous solution. This equals about 2% of the weight of the resin component for most composites (Ferracane, 1994). Filler particles are also known to leach ions such as silicon, barium, strontium and sodium when stored in water (Soderholm, 1983, 1990; Soderholm et al., 1984; 0ysæd and Ruyter, 1986).

Fluoride has also been shown to be released from composite resins (Teminnd Csuriz, 1988; Swift, 1989; Wiltshire and van Rensburg, 1995). The efficacy of such fluoride release as an anticariogenic agent is still under investigation (Swift, 1989).

The loss of components and subsequent water sorption is rapid during the initial soaking period and slows substantially within hours. Approximately 75% of the elutable species are extracted within the first several hours (Pham and Ferracane, 1989; Ferracane and Condon, 1990; Wiltshire and van Rensburg, 1995).

Water uptake has been reported to be 2-3% of the weight of the composite (Fan et al., 1985; Ferracane and Condon, 1990). Such water sorption may affect the, mechanical properties of the composite and has been explained by filler-matrix debonding and hydrolytic degradation of the fillers (Soderholm et al., 1984). Soderholm et al. (1984) hypothesized that the breakdown of the filler articles raised internal osmotic pressure within the composite structure and microcracks formed dong the matrix-filler interfaces. These cracks propagated until they reached the surface of the composite, as evidenced in S.E.M. studies (Solderholm, 1984).

Composite To Composite Bonding

The majority of the literature on resin to resin bonding deals with the repair of previously placed composite resin restorations. The interfacial bond strength of composite to fresh (i.e., only a few minutes old) composite has been reported to be the same as the cohesive strength of the material (Lloyd et al., 1 980; Boyer et al., 1984). Boyer et al. (1984) reported transverse bond strength of between 53.2 and 109.4 MPa, depending on the type of composite resin. However, the bond strength to polished, untreated composite surfaces is in the range of 20% to 75% of the cohesive strength of the respective substrate (Lloyd et al., 1980; Vankerckhoven et al., 1982; Chan and Boyer, 1983; Boyer et al., 1984; Azarbal et al., 1986; Pounder et al., 1987; Kao et al., 1988b; Puckett et al., 1991). Absolute values for the cohesive and adhesive shear bond strengths reported in the above mentioned studies ranged from as low as 5.74 MPa to as high as 11 1.6 MPa depending on numerous factors such as substrate, surface treatments, adhesives, storage conditions, and bond strength testing methods.

There are three possible mechanisms of a 'new" composite resin bonding to an "old" composite resin. Chemical bonds may form with the resin matrix (Brosh et al., 1997), chemical bonds may form with the exposed filler particles (Brosh et al., 1997), and micromechanical retention may be gained by penetration of resin monomer into undercuts on the restoration surface and perhaps even into microcracks in the matrix (Brosh et al., 1997). This micromechanical bonding may be further enhanced by solvent bonding. Solvent bonding occurs as the methylmethacrylate monomer diffuses into the "old" resin, resulting in swelling of the set resin, and allowing penetration of the "new" resin (Powers et al., 1997). Upon polymerization, the "new" resin is mechanically locked onto the surface of the "old" resin.

Chemical bonding between resin matrices is dependent upon the concentration and availability of unreacted methacrylate groups in the substrate resin (Vankerckhoven et al., 1982; Puckett et al., 1991; Tumer and Meiers, 1993; Li, 1997). The concentration of such unreacted methacrylate groups decreases from 100% to approximately 50% as the resin polymerizes, so the potential for chemical bonding diminishes as the resin ages (Vankerckhoven et al., 1982; Swifi et al., 1992). Also, when the restoration surface is polished, inorganic filler particles are exposed and the degree of unsaturated methacrylate groups is, decreased to 25%. This limits the chemical bonding between resin matrices (Vankerckhoven et al., 1982).

Chemical bonding to the filler particles would require the use of a silane coupling agent, much like the organosilane incorporated into individual composite materials as discussed in section 2.2.1. Silane coupling agents have been shown to improve the bond of composite to etched, sandblasted and roughened porcelain by up to 3 MPa (Stangel et al., 1987; Andreasen and Stieg, 1988; Kao et al., 1988a; Major et al., 1995; Roulet et al., 1995). However, silanes have failed to predictably increase the bond strength of new and old composites compared with dentine/enamel bonding agents (Azarbal et al., 1986; Saunders, 1990; Sodërholm and Roberts, 1991; Swift et al., 1994; Bouschlicher et al., 1997; Brosh et al., 1997). This may suggest that mechanical interlocking is the most significant factor contributing to repair strength, and thus, silanes are not recommended for clinical application of composite to composite bonding (Sodërholm and Roberts, 199 1; Swift et al., 1994; Bouschlicher et al., 1997; Brosh et al., 1997).

Numerous surface treatments and bonding agents have been advocated to improve the repair strength of composites. Most treatments attempt to increase the micromechanical bonding between the substrates (Brosh et al., 1997). Wetting of the restoration surface by the repair material is a major factor controlling the repair bond strength as discussed in section 2.1. Unfilled resins improve the bonding of fresh material to polished composite surfaces by up to 6 to 25 MPa after 24 hours (Boyer et al., 1984; Azarbal et al., 1986; Puckett et al., 1991; Swift et al., 1992).

Composite Resin Surface Treatments

Many different surface treatment protocols have been recommended. Perhaps the most important procedure in composite to composite bonding is the roughening of the mature resin surface (Swift et al., 1992; Turner and Meiers, 1993; Kupiec and Barkmeier, 1996). Turner and Meiers (2993) compared the shear bond strengths achieved with various surface treatments and with different adhesives. In a two-way ANOVA, they found the surface treatment to be highly significant (p<0.0001) and the adhesive to be less significant (p=0.643). Etching with 37% phosphoric acid, while vital in bonding to enamel (Abdullah and Rock, 1993; Johnston et al., 1996; Olsen et al., 1996; Powers et al., 1997), appears to be relatively ineffective in composite bonding (Boyer et al., 1978). Boyer et al. (1997) reported no significant difference in tensile bond strength between resin layers with the use of 37% phosphoric acid. Hydrofluoric acid (HF) is used to etch porcelain surfaces for indirect restorations, inraoral repair, or orthodontic bonding (Zachrkson and BuyukyiImaz, 1993; Kem and Thompson, 1994; Major

et al., 1995). HF can etch the glass filler particles in hybrid and small particle, composites. HF has been shown to cause some slight surface changes such as minor roughening (Kula et al., 1983; Kula et al., 1986); however, research does not reveal significant increases of bond strength with the use of HF for composite repair (Cnimpler et al., 1989; Swift et al., 1992; Swift et al., 1994; Brosh et al-, 1997). Mitchem et al. (1991) even recommended against the use of HF with hybrid composites as etching softens the resin surface. This change is likely as a result of the removal of the hard filler particles, as it has been shown that there is no change in hardness in unfilled resins after exposure to 15% HF for 6 months (Al-Jezairy and Williams, 1996).

Mechanical treatment of the composite surface may be accomplished with rotary instruments such as diamond or green carborundum stones or through the use of air abrasion with aluminum oxide particles. Brosh et al. (1997) descnbed the surface treatment of composites with diamonds or green stones as providing "macro" retentive features which are controlled by the operator. Air abrasion resulted in "micro" retentive features that are under the control of the instrument. In the same study (Brosh et al., 1997) the authors concluded that the "micro" retentive features demonstrated superior shear bond strengths over the ''macro" retentive features in combination with a bonding agent. Smdblasting increased the shear bond strength by 4.19 MPa and roughening with a diamond stone increased the bond strength by 0.64 MPa (Brosh et aL, 1997). Microetching with air abrasion is less invasive of the composite restoration than roughening with handpieces (Brosh et al., 1997). Also, it is a relatively simple and quick procedure, and does not involve the use of strong acids intraorally. Recent studies have reported air abrasion as an effective means for surface preparation of aged composites (Swift et al., 1992; Turner and Meiers, 1993; Kupiec and Barkmeier, 1996; Bouschlicher et al., 1997).


Direct Bonding

Arguably, the most significant achievement in dentistry in the 20th century was the acid etch technique developed by Buonocore (1955). Utilizing various acidic mixtures, of which 30% to 40% phosphoric acid seems to be the most effective (Moin and Dogon, 1974; Retief, 1974; Legler et al., 1990; Wang et al., 1994; Olsen et al-, 1996), enamel surfaces are "etched," a process in which the acid preferentially dissolves the centers or peripheries of the enamel rods. The etching time has also been debated and studied numerous times. Britton et al. (1990) compared bond strengths between 15- second and 60-second etch times. Their results indicated increased bond strengths in the 15-second group. Gorelick (1977) evaluated the effects of 60- and 90-second etching times, Barbier et al. (1985) compared 15- and 60-seconds of etching, and Beech and Jalaly (1980) evaluated 5-, 15, 60-, and 120-second intervals. They all reported no decrease in bond strength as the result of shortened etching times. The most recent reviews on the subject of etching time (Olsen et al., 1996) also concluded there was no significant effect on bond strength between 10- or 30-second etching intervals. Etched enamel allows a bonding agent of low viscosity to penetrate into the microscopic undercuts as evidenced by "resin tags" when seen under S.E.M. (Pahlavan et al., 1976). Once polymerized, a micromechanical bond is established.

The acid etch technique provided the background for direct orthodontic bonding. Direct bonding eliminates the need for bands. This has several advantages such as enhanced ability for plaque removal by the patient (Zachrisson, 1976) minimizing soft tissue irritation and hyperplastic gingiva (Zachrisson, 1976) minimizing the danger of decalcification with loose bands (Zachrisson, 1976) elimination of the need for separation, absence of post-treatment band spaces, and improved esthetics during treatment. There is some controversy over who should get credit for the fist direct orthodontic bonding. Most references, and the American Association of Orthodontists, seem to support the claim of George Newman (Newman, 1992). He reported a bonding technique using acrylic resins in 1965 (Newman, 1965). The early resins had a 15-minute setting time, which limited their acceptance in clinical practice (Retief and Sadowsky, 1975).

Orthodontic Adhesives

Currently, there are numerous materials available for direct orthodontic bonding. These include diacrylate composite resin-based products and glass-ionomer adhesives, which are available in either chemical or dual-cured systems (Powers, 1997), and cyanoacrylate systems (Ortendahl and Ortengren, 2000). The composite resin adhesives are mostly based on the Bis-GMA resin (Bowen, 1963) modified to suitable viscosity for clinical use. The new chemically-cured resins have a reduced setting time of three to eight minutes (Wang and Meng, 1992; Mitchell, 1994; Lloyd and Scrimgeour, 1995). In vitro bond strengths do not appear to be affected by whether the composite is chemically-cured, light-cured, or dual-cured (Bradburn and Pender, 1992; Wang and Meng, 1992; Smith and Shivapuja, 1993; Whitlock et al., 1994; Eliades et al., 1995; Kao et al., 1995; Chamda and Stein, 1996).

Glass ionomer materials are also now available as either chemically or light-cured systems for orthodontic bonding (Powers et al., 1997). They have not replaced composite resins, as they have lower in vitro bond strengths to either etched or non-etched enamel (Rezk-Lega and Ogaard, 1991; Oen et al., 199 1; Wiltshire, 1994; Powers et al., 1997). For example, Wiltshire (1994) recorded mean shear bond strength of an orthodontic attachment bonded to etched enamel with a chemically-cured glass ionomer cement to be 5.5 MPa, and the comparative bond strength with a composite cement was 26 MPa. A new group of resin-modified glass-ionomer (RMGI) cements do provide for increased bond strengths over conventional GI cements (Enckson and Glasspoole, 1994). Erickson and Glasspoole (1994) reported shear bond strength of 20.5 MPa for a RMGI compared to 7.2 MPa for a conventional GI. Powers et al. (1997) also reported tensile bond strengths of 8 MPa to 25 MPa for five different RMGI's to unetched enamel. However, a recent publication reported low initial bond strength (after 30 minutes) for an orthodontic RMGI cement of 0.4 MPa vs. 5.2 MPa for resin cement (Bishara et al., 1999). Recently, a cyanoacrylate adhesive system- has been developed for direct, orthodontic bonding. The manufacturer, Gestenco Int., claims adequate bonding to many surfaces including enamel, porcelain, and composite (Gestenco Int., 1999). Previous cyanoacrylate adhesives have failed to gain acceptance in the orthodontic community because of their poor durability in a wet environment. Crabb and Wilson (1971) studied three cyanoacrylate adhesives (Cyanodont, Eastman 9 10, and Permabond). After storage in 37" C saline for 24 hours, al1 of the bond strengths were reduced to near zero. More recently, Howells and Jones (1989) reported on another cyanoacrylate adhesive developed exclusively for orthodontic bonding. Again in their study, the material proved to be too susceptible to deterioration after storage in water. Bond strengths went from a mean of 124 N after one hour to 26 N after seven days and decreased to only 6 N after 98 days.

Orthodontic Bonding to Various Materials


The primary surface to which orthodontic brackets are directly bonded is enamel. The acid-etch procedure allowed for direct orthodontic bonding to be possible, and is discussed in section 2.3.1. Application of 30% to 40% phosphoric acid for at least 10 seconds seems to be the most effective in preparing enamel surfaces. Shear bond strengths of metal brackets bonded to etched, dry enamel with composite resin adhesives may attain values near 26 MPa (Wiltshire, 1994). Although glass ionomer cements can be bonded in a wet environment, the bond strengths are only 5-8 MPa (Chung et al., 1999). Even when bonded dry, glass ionomers still produce lower bond strengths than composite resins (3-10 MPa vs. 26 MPa) (Rezk-Lega and Ogaard, 1991; Oen et al., 199 1; Wiltshire, 1994; Powers et al., 1997, Chung et al., 1999). The new RMGI cements are promising as they had similar shear bond strengths compared to a composite cement in a recent article (8.8 MPa vs. 10.4 MPa) (Bishara et al., 1999).


Bonding orthodontic attachments directly to porcelain or ceramic restorations has also been extensively studied. Major et al. (1995) compared adhesion promoters and recommended the use of a silanating agent. Use of a silane produced bond strengths of 6-14 MPa, whereas bond strengths without the primer ranged from 0.4 to 4 MPa (Whitlock et al., 1994; Major et al., 1995; Zachrisson et al., 1996). Porcelain prepared with acidulated phosphate fluoride solutions produced low bond strengths of less than 5 MPa (Barbosa et al., 1995; Zachrisson et al., 1996). Roughening of the porcelain surface seems to be generally contraindicated, as bond strengths become excessive and porcelain fractures occur upon debonding. Cochran et al. (1997) reported bond strengths of 28-39 MPa with sandblasting and silanization, and Barbosa et al. (1995) obtained bond strengths of 28-47 MPa with diamond bur surface preparation and silanization.


Buyukyilmaz et al. (1995) compared sandblasting the gold surface to roughening with a diamond bur and found that sandblasting produced better bond strengths (20 MPa) than the diamond bur treatment (10 MPa). Superbond C & BTM (Sun Medical, Kyoto, Japan), a 4-META metal-bonding adhesive resin also produced superior bond strengths compared to a conventional composite resin (BüyiiSilmaz et al., 1995; Nollie et al., 1997).


Bond strengths of both conventional and 4-META adhesives to amalgam are generally low (3-6 MPa) (Zachrisson et al., 1995). Sperber et al. (1999) recently sandblasted an amalgam surface and produced shear bond strengths similar to bonding to etched enamel with a resin cement (1 1.77 MPa vs. 10.76 MPa).

Composite Resin:

Bonding of orthodontic brackets to composite resin surfaces is only known to have been reported a few times. Newman et al. (1984) studied orthodontic bonding to a heat-cured composite resin (Isosit TM, Vivadent Corp., Buffalo, N.Y.). They compared the shear bond strengths between brackets bonded with Concise TM, (3M, St. Paul, MN) with or without silane application to brackets bonded to etched, natural teeth. No significant differences existed between the three groups, and their bond strengths were 1120-1300 lbs/in2 (7.7 - 8.9 MPa).

Schwartz et al. (1990) studied the tensile bond strengths of metal brackets to resin substrates using three composite resin adhesives. They bonded to either untreated surfaces or surfaces treated with 37% phosphoric acid, reduced with a diamond bur, coated with silane agent, or coated with a dentine bonding agent. Though most of their results were not reported, they did report bond strengths of 4.3 +/- 2.0 MPa using Contacto on untreated resin surfaces. It was also reported that the use of Mono-Lok 2 and Unite produced tensile bond strengths of 10.5 +/- 3.2 MPa and 10.3 +/- 2.6 MPa. Kao et al. (1995) compared the torsional bond strength of ceramic brackets bonded to composite resin veneer laminates and enamel. Silux Plus TM (3M, St. Paul, MN) veneers were fabricated and ceramic brackets were bonded with either a light-cured or chemically-cured composite adhesive. All samples were acid-etched before bonding and subsequently thermocycled. Torsional bond strengths of between 30 MPa and 60 MPa were recorded for both the chemically- and light-cured adhesives.

Chunhacheevachaloke and Tyas (1997) also studied bonding to resin composite. Again, they compared two types of ceramic brackets and the effect of roughening r the composite surface with a coarse Soflex TM disk (3M, St. Paul, MN). All samples were acid-etched and bonded with Transbond TM (3MNnntek, Monrovia, CA). No significant differences were found with their shear bond strengths which ranged from 17.1 MPa to 19.2 MPa. Nineteen of the forty composite samples had cohesive failures upon debonding. The most recent and thorough study of orthodontic bonding to composite resin was reported by Lai et al. (1999). They bonded metal, ceramic and polycarbonate brackets to Silux Plus TM (3M, St. Paul, MN) samples (roughened with Soflex TM discs) using either a light-cured resin modified glass ionomer cement, a chemical-cured composite, or a light-cured composite system. Half of the samples were tested after 24 hours and half were thermocycled. They concluded all groups to have clinically acceptable bond strengths (10.0 to 30.1 MPa) except the polycarbonate/chemically-cured group (3.58 MPa). 210 of 288 samples had damaged resin surfaces after debonding.

Bond Strength in Orthodontics

The bond strength of orthodontic attachments must be able to withstand both functional stresses (from occlusion and mastication) and operator stresses (from the orthodontic appliances) (Powers et al., 1997). Newman et al. (1994) stated that "maximum strength is needed to compensate for the unfavorable, moist environment in which the polymer adhesive system operates, as well as variations in pH, thermal changes, impact forces from sticky, chewy, or hard foods, and sports accidents." However, the direct bonding of orthodontic attachments is a temporary procedure; after treatment, the attachments must be removed with minimal or no damage to the substrate and this is best achieved with low bond strength (Powers et al., 1997). Although it has been attempted, it is difficult to evaluate orthodontic bond strengths in vivo (Voss et al., 1993). Laboratory testing allows for improved standardization of testing procedures and the use of more sensitive equipment.

Different types of bond strengths are reported in the literature including shear, shead/peel, tensile, and torque (Ostertag et al., 1991; Powers et al., 1997). In shear bond strength testing, the debonding force is applied directly and parallel to the junction of the bracket and adhesive (Fox et al., 1994). True shear bond strength is impossible to determine practically. Using a 3D finite element analysis, Thomas et al. (1999) found the tensile and compressive stresses exceeded the shear component. The term shear/peel is used in the literature to reflect this phenomenon (Katona, 1994). Most studies reporting shear bond strength are actually testing the shead/peel bond strength (Katona, 1997). In tensile strength testing, the debonding force is applied perpendicularly to the substrate surface. Up P to 15% of the stresses are in fact shear and compressive in nature, again directing the results to a tensile/peel force (Thomas et al., 1999). A final method of bond strength testing is torsion loading, in which the attachment is "twisted" off. This method is less favoured because most mechanical testing machines cannot perform it (Katona, 1997). Also, the results of a torsion test are reported in N/m and cannot be directly compared to tests of shear, tensile, or peel strength which are reported in MPa (Katona, 1997).

A bond strength is only relevant if it can be correlated clinically. The absolute value for clinically adequate shear bond strength most often quoted is from the work of Reynolds (Reynolds, 1975; Reynolds and von Fraunhofer, 1976). They recommend bond strength of 60-80 kg/cm2 (5.9-7.9 MPa). However, the original paper does not state any scientific method for calculating this value. Just as important as the minimum recommended bond strength is the maximum recommended bond strength. Retief s work (1974a) must also be considered when accessing bond strengths, as he demonstrated enamel fractures on specimens with bond strengths as low as 9.7 MPa.