Abstract
Introduction
Resin bonding can be compared to making a sandwich with the tooth on one side and the restoration on the other, a layer of bonding resin is applied to either side and a filled resin (composite) placed in between. This presentation considers factors that influence the restoration side of the sandwich and various ways that the assembled testpieces may be “aged” prior to testing. The materials to be bonded may be either ceramic, metal or composite formed by methods such as casting, pressing, sintering or machining. The fabrication method determines the susceptibility of the bonding surface to physical or chemical modification. The treatment of the surface prior to bonding can be physical (e.g. sandblasting) or chemical (e.g. metal primer); but is more likely to be a combination (e.g. silica deposition + silane).
Preparation of the bonding surface
Successful bonding depends on establishing a surface with a high population of unreacted vinyl groups (–C C) that can then be cross-polymerized to the resin in the bonding composite. The physical approach has involved etching or sandblasting the surfaces; but the ability to form a microretentive surface in this way depends on a heterogeneous surface. Noble metals and modern high strength ceramics have a more homogeneous surface and are not easily etched. To overcome this problem a number of ways to deposit a silica layer on the bonding surface have been developed: the Silicoater that involves baking on a silica layer, the Rocatec technique (CoJet) that involves air blasting silica onto the surface in conjunction with an abrasive; and two more modern approaches: sol–gel evaporation and molecular vapor deposition (MVD). All these techniques require the subsequent application of a silane layer to provide the –C C moieties. The use of primers without an intervening silica layer has been tested and found to be successful with some specialized bonding systems that contain agents such as methacryloyloxydecyldihydrogen-phosphate (MDP) (PanaviaEX).
Aging of testpieces prior to bonding
The most common type of aging is storage in water at temperatures from ambient to 100 °C. This generally decreases bond strengths; but not to catastrophic values. A more exacting pre-test regime is thermal cycling. In some studies this caused a slightly greater reduction in bond strength than storage in water; but in other tests it resulted in total failure. As some testpieces have spontaneously debonded during thermal cycling, it seems sensible to include TC in any screening test of new materials. Mechanical cycling (fatigue) prior to testing has a very significant effect and the bond strength that can withstand 1,000,000 cycles can be one sixth of the bond strength in a simple monotonic test (tensile, shear or compression). Whereas simple monotonic tests provide a blunt instrument for eliminating poorly performing techniques their use for discriminating between established techniques is open to discussion.
1
Introduction
Resin bonding can be compared to making a sandwich with the tooth on one side and the restoration on the other, a layer of bonding resin is applied to either side and a filled resin (composite) placed in between.
The materials to be bonded may be either ceramic, metal or composite. These may be formed by various methods such as casting, pressing, sintering or machining.
In the case of metals and ceramics there is no requirement to age the surfaces prior to bonding whereas with composites, a number of studies have attempted to age the composite substrate in order to mimic the situation of bonding new composite to old composite in the mouth.
It is normal to prepare the bonding surfaces of the teeth and restoration prior to the application of the bonding resin. This normally involves a combined physical and chemical approach (e.g. sandblasting + silanation).
Having prepared the surfaces the testpiece is assembled. The assembly protocol may attempt to mimic the clinical situation, which is relatively uncontrolled, or attempt to control variables such as the precise location of the restoration and the “seating force”.
Having completed the assembly the testpiece may be aged prior to testing. In vitro aging is a function of time and environment plus additional features such as thermal or mechanical cycling.
This paper considers these factors; but is limited to preparation of the restoration side of the sandwich.
1.1
Materials to be bonded
1.1.1
Metals
Resin bonding of indirect restorations began with the metal “Rochette” splint in 1973 . This relied on macromechanical retention via composite “rivets” through large perforations in the metal. The metals were predominantly alloys of chromium, nickel, beryllium and cobalt formed as a casting via the lost wax process. This retention technology was superseded by electrolytic etching , or sandblasting of similar alloys. The early papers on sandblasting were mainly in German . A disadvantage of etching and sandblasting was that it precluded the use of noble metals that had a more homogeneous surface. This led to development of devices that could deposit a layer of silica on the fitting surface (see Silicoater and Rocatec below). These techniques could be used with noble metals. Lost wax casting has remained the main method for preparing metal core restorations although more recently, driven by the introduction of CAD–CAM technology, a number of laboratories have pioneered laser sintered cobalt chrome. There are a limited number of studies comparing cast and sintered non-precious metals, but, as this is a new technique, they have concentrated on the ceramic to metal bond rather than the bond to tooth . In terms of non-precious metals the inclusion of nickel and beryllium is being phased out because of concerns over antigenicity and the material of choice is cobalt–chrome.
1.1.2
Ceramics
Resin bonded ceramics were introduced in the early 1980s following the use of indirect composite veneers . The main method of manufacture was sintering of a feldspathic porcelain frit. This left a relatively heterogeneous surface that could be etched with hydrofluoric acid . Sorenson et al. observed that etching feldspathic porcelain with 20% hydrofluoric acid for 3 min significantly increased its bond strength to composite resin. An alternative technique for fabricating ceramic restorations was casting or pressing of a glass blank. For these materials (Dicor) the etching gel was a solution of 10 percent (w/v) ammonium bifluoride (NH 4 HF 2 ) .
The fact that sintered and cast (pressed) ceramic could be etched is indicative of a heterogeneous surface that will therefore have a relatively large inherent flaw size or even overt surface deficiencies. These make the material susceptible to fracture. The use of glass infiltration to improve fracture toughness, in techniques such as In-ceram, has meant that surfaces are more homogeneous and consequently less susceptible to acid etching. A similar problem occurs with fully crystalline alumina and Y-TPZ (zirconia) that are machined from blanks. These materials are far more homogeneous and not susceptible to acid etching or other mechanical forms of roughening .
1.1.3
Composites
Indirect composite restorations such as inlays and onlays have always been made by hand moulding by the dental technician. For laboratory testpieces this can be controlled by using a PTFE mould to standardize the dimensions of the testpiece.
1.1.4
Aging of composite substrate
A number of investigators have studied the bond strength of new composite to composite that has been aged in water or artificial saliva . Tezvergil et al. aged the first layer of composite by boiling it for 8 h and then storing in water for 3 weeks . Most authors agree that the inclusion of an intermediary resin layer improves the bond strength .
1.1.5
Choice of materials substrates for bonding
Most papers do not give a reason for the choice of individual materials. It is likely that this was strongly influenced by the material available in the dental school.
2
Preparation of the bonding surface
Successful bonding depends on establishing a surface with a high population of unreacted vinyl groups (–C C) that can then be cross-polymerized to the resin in the bonding composite. Three approaches have been made: physical, chemical and combined.
2.1
Physical approach
This involves creating a rough surface into which a resin is added. The resin is then set leaving it micromechanically locked into the material with a population of –C C– moieties in the resin on the fitting surface. Initially the roughness was created by preferentially removing some of the heterogeneous surface by techniques such as acid etching or sandblasting . Etching results in a surface with definite microtags similar in size to those formed in etched enamel. However, whereas enamel can be etched with orthophosphoric acid, metals depended on electrochemical etching and ceramics on the use of hydrofluoric acid. Noble metals were not susceptible to electrolytic etching because of their homogeneous structure ( Fig. 1 ).
The inability to etch precious metals was part of the impetus to create a silica deposition layer in devices such as the Silicoater and Rocatec. In fact there were two Silicoaters: one applied a flame sprayed silica surface (Silicoater Classic; Kulzer) and the other applied a silicate layer coated in chrome oxide sintered into the alloy surface (Silicoater MD; Kulzer) . The alternative Rocatec technique is a so-called tribo-deposition process whereby silica modified aluminum oxide particles are blasted onto the surface. The force of impact causes the Al 2 O 3 to bounce back off the surface leaving a layer of silica behind . Although this technique is said to be “cold” compared to the Silicoaters there is a considerable amount of energy exchange at the points of impact . All silica layer techniques require a silane coupling agent to facilitate bonding of the resin composite to the treated surface. As both techniques involve sandblasting there is always some degree of surface loss .
The bond strength of composite to CoCr using six bonding regimes was compared in 1995 . The six techniques were sandblasting + Bis-GMA, silanation + Bis-GMA, Rocatec, Silicoater, methacryloyloxydecyldihydrogen-phosphate (MDP) modified composite (PanaviaEX), and a modified presentation of the latter. Each group was subdivided into 3 which were tested immediately, after 30 days with 7500 thermal cycles and after 150 days with 37,500 thermal cycles. The results showed that the two Bis-GMA techniques had far lower bond strengths than the other preparations. In some cases water storage and thermal cycling for 30 days increased the bond strength whereas it decreased after 150 days. An insight into the resin bond to ceramics can be obtained by considering the literature on porcelain repair. A review of the literature determined that the following treatments are all cited by the manufacturers of porcelain repair kits: etching with hydrofluoric acid (9.6%), etching with acidulated phosphate fluoride, micromechanical roughening (with a diamond bur) and air abrasion with Al 2 O 3 . Many of these physical treatments were followed by the application of a silane primer. The review indicated great variation in the results; but a consistent finding was that silane alone was not adequate for porcelain repair. Bailey noted significant differences between some systems depending on whether the porcelain substrate had been stored in water or at ambient humidity .
The Rocatec technique was marketed by 3M-ESPE as Rocatector and more recently as the CoJet system (3M-ESPE). These products are a chairside microsandblaster advocated for porcelain repairs including ceramic fractures that expose metal. In a similar way orthodontists have studied the ability to bond to ceramic as they occasionally need to bond brackets to ceramic crowns. In these cases some degree of physical roughening is followed by the application of a silane layer .
As ceramics such as glass infiltrated ceramic (In-ceram) and fully crystalline ceramic (Y-TZP) are not susceptible to acid etching with hydrofluoric acid both the Silicoater and Rocatec have been applied to these materials. More recent techniques for the deposition of a silica layer have been the sol–gel process and molecular vapor deposition (MVD) . In a recent study a thin layer (2.6 nm) of SixOy deposited by MVD gave higher bond strengths than CoJet (Rocatec) whereas with a thicker layer (23 nm), the bond strength was less .
An alternative approach to create a mechanical undercut has been to use preformed wax that was manufactured with inherent bars, webs or ribs. This is standard for orthodontic brackets. A simple “in house” technique was to push the wax onto crystals of table salt before making the wax pattern and then wash out the salt to leave a roughened surface. More recently spark erosion has been compared with the Rocatec system . After 24 h spark erosion gave significantly higher shear bond strengths on titanium than Rocatec. This was reduced after thermal cycling.
2.2
Chemical approach
Silane layers have traditionally been used to enhance bonding. Silanes are difunctional molecules with one part able to bond to the ceramic and the other having the –C C group to cross polymerize with the bonding resin . Activated silanes with silanols are adsorbed, deposited, polymerized, and finally covalently bonded to the substrate surface. This procedure is called silanization or silanation . There are various forms of silane the most common being 3-methacryloyloxypropyltrimethoxysilane . These are discussed in Section 2.3 as they are usually used in combination with a physical process. An alternative to the silane system was a primer developed by Kojima et al. in 1987 . The active molecule is 6-(4-vinylbenzyl- n -propyl)amino-1,3,5-triazine-2,4-dithione (VBATDT). It is used in conjunction with a methyl methacrylate-tri- n -butylborane (MMA-TBB) resin and marketed as V-Primer. An alternative metal primer is l0-methacryloyloxydecyldihydrogen thiophosphate (MlOPS) . It is marketed as Metal Primer. These two primers were used in conjunction with three bonding agents specifically advised for metal bonding: Imperva Dual (Shofu, Japan), Panavia 21 (Kuraray, Japan) and Super-Bond (Sun Medical, Japan). Generally they enhanced the bond strength to a silver–palladium copper–gold alloy and a Type IV gold .
2.3
Combined approach
Laboratories have normally adopted a combined approach whereby some degree of physical roughness has been created by sandblasting and then augmented with a silane layer. All the silica deposition techniques (Silicoater, Rocatec, sol–gel and molecular vapor deposition) require a subsequent layer of silane suggesting that this layer is more reactive than the underlying ceramic or metal.
Durability of bonding between composite resin and ceramic formed with chemical agents was markedly inferior to alteration of the ceramic surface with either aluminum oxide air abrasion, hydrofluoric acid or a combination of both . Many studies have continued to compare the combination of etch, sandblasting, priming, and silanation .
Matinlinna et al. compared the use of three silane molecules after a silica deposition by Rocatec . The silanes used were: (a) 3-methacryloyloxypropyltrimethoxysilane, (b) 3-acryloyloxypropyltrimethoxysilane, and (c) 3-isocyanatopropyltriethoxysilane. These were used in conjunction with an experimental Bis-GMA resin, or a commercially available adhesive based on Bis-GMA; but containing the adhesion promoter molecule: 10-methacryloyloxydecyldihydrogen-phosphate (RelyX ARC; 3M-ESPE). Half of the testpieces were thermal cycled. It was found that the testpieces formed with 3-isocyanatopropyltriethoxysilane spontaneously debonded during thermal cycling. The testpieces formed with RelyX ARC had greater bond strengths than the Bis-GMA material. Kern et al. compared the use of non-silane primers with air abraded and non-abraded zirconia (Cercon) disks . Half of the samples were stored in water for 150 days with thermal cycling. The samples that had not been air abraded debonded spontaneously during thermal cycling.
A similar finding occurred when bonding densely sintered zirconia (Cercon) disks . After 150 days storage and thermal cycling, only the air abraded specimens bonded with Panavia F showed high bond strengths of 39.2 MPa, whereas most other specimens debonded spontaneously or showed very low bond strengths.
Recently Yanagida et al. compared the use of two metal primers : Alloy Primer (Kuraray, Japan) containing VTD, [6-(4-vinylbenzyl- n -propyl)amino-1,3,5-triazine-2,4-dithiol, dithione tautomer] and MDP, [10-methacryloyloxydecyldihydrogen-phosphate] vs. ML Primer (Shofu, Japan) containing MDDT, [10-methacryloxydecyl 6,8-dithiooctanoate]; MHPA, [6-methacryloxyhexylphosphonoacetate]; and γ-MTPS, [γ-methacryloxypropyl trimethoxysilane]. Before priming the specimens were treated with the Rocatec/silane system. The substrates were Type IV gold and commercially pure titanium. They found that the initial bond strength to titanium was initially higher but decreased after thermal cycling.
2
Preparation of the bonding surface
Successful bonding depends on establishing a surface with a high population of unreacted vinyl groups (–C C) that can then be cross-polymerized to the resin in the bonding composite. Three approaches have been made: physical, chemical and combined.
2.1
Physical approach
This involves creating a rough surface into which a resin is added. The resin is then set leaving it micromechanically locked into the material with a population of –C C– moieties in the resin on the fitting surface. Initially the roughness was created by preferentially removing some of the heterogeneous surface by techniques such as acid etching or sandblasting . Etching results in a surface with definite microtags similar in size to those formed in etched enamel. However, whereas enamel can be etched with orthophosphoric acid, metals depended on electrochemical etching and ceramics on the use of hydrofluoric acid. Noble metals were not susceptible to electrolytic etching because of their homogeneous structure ( Fig. 1 ).
