NPG-GMA (N-phenylglycine glycidyl methacrylate), a surface-active comonomer, was the basis of Cervident (S.S. White, Lakewood, NJ, USA), the first commercially available dentin-bonding agent [21]. This comonomer could chelate with calcium on the tooth surface and was able to generate water-resistant chemical bonds to dentinal calcium [22, 23]. The in vitro dentin bond strengths of the first-generation dentin adhesives were in the range of 2–3 Mpa [24].

Several phosphate ester dentin-bonding systems were introduced in the early 1980s. These products were based on phosphorous esters of methacrylate derivates. The mechanism of action was based on enhanced surface wetting and ionic interaction between negatively charged phosphate groups in the resin and positively charged calcium in the smear layer [24]. Clearfil Bond System F^c (Kuraray, Osaka, Japan) was recognised as the first product of the second generation of dentin adhesives (phenyl-P and hydroxyl methacrylate [HEMA] in ethanol). Bond strengths ranged between 1 and 6 Mpa [25–27] as a result of a lack of dentin wetting (large contact angles) [28]. Moreover, the entire depth of the smear layer was not penetrated up to the dentin to establish ionic bonding or create resin extensions into the dentinal tubules [26]. Inadequate hydrolytic stability in the oral environment was also described [29, 30].

In 1984, Clearfil New Bond (Kuraray, Osaka, Japan) was introduced and with this the concept of phosphoric acid etching of dentin before the application of a phosphate ester-type bonding agent. This phosphate-based material contained HEMA and a 10-carbon molecule known as 10-MDP, which includes long hydrophobic and short hydrophilic components [31].

Most other third-generation materials, however, were designed not to remove the entire smear layer.

In 1982, Nakabayashi described the hybrid layer as an interconnection zone between adhesive and dentin based on a micromechanical bonding mechanism as still used by the present-day adhesive systems [32]. Extensive research in Japan had shown that strong, long-lived bonds between resin and living dentin could be formed when a monomer such as 4-META, which contains both hydrophilic and hydrophobic chemical groups, penetrated the dentin and polymerised in situ. This resin impregnation creates the so-called transitional ‘hybrid’ layer, that is neither resin nor tooth, but a hybrid of the two. The thin layer of resin-reinforced dentin locks the two dissimilar substances together on a molecular level, sealing the surface against leakage and imparting a high degree of acid resistance. Dentin was etched with an aqueous solution of 10 % citric acid and 3 % ferric chloride. After rinsing and drying, an aqueous solution of 35 % HEMA and a self-curing adhesive resin containing 4-META, MMA (methyl methacrylate) and TBB (tri-n-butyl borane) was applied.

The total-etch approach originated in Japan, where the application of a phosphate ester type of bonding followed an etch procedure with phosphoric acid [20]. Research during the begin period of the 3E&Ra did not demonstrate improvement in bond strengths, most probably as a result of the hydrophobic nature of the phosphonated resin [33]. Today, however, in vitro tensile bond strengths between 20–50 Mpa for enamel and 13–80 for dentin are demonstrated [34].

Acid etching alters the mineral content of dentin and changes its surface-free energy [35, 36]. The etched dentin (the resultant collagen network depleted of high surface energy hydroxyapatite) ends up as a low surface energy substrate. The primer in a 3E&Ra is needed to increase the critical surface tension of the dentin [34]. Both primer and bonding resins impregnate the intertubular dentin and a resin-dentin interdiffusion zone of hybrid layer is formed. Both resins also penetrate the dentinal tubules and form polymerised resin tags.

The 2E&Ras (fifth-generation adhesives) were a simplification of the adhesive system. Most of these systems have fallen somewhat short of their objective, but still, comparable bond strengths to those of the 3E&Ra (fourth-generation adhesives) are achieved [16].

A further demand for simplification resulted in the development of systems without a separate conditioning phase. This sixth generation consists of systems with a self-etching primer and a separate adhesive resin and those that combine the conditioner, primer and adhesive resin but require mixing and hence, the confusion when using the classification based on generations.

This group of adhesives use the smear layer on the enamel and dentin as bonding substrate as was seen with the second-generation adhesives. The acidity of the primer, however, is the main difference between the two generations. This generation adhesives contain acidic monomers such as 4-MET and 10-MDP [37, 38] which renders these adhesives much more hydrophilic compared to previous hydrophobic adhesive systems [39, 40].

The self-etching primers (SEPs) have been classified in three groups: mild, moderate and aggressive. Mild SEPs appear to provide excellent dentin bond strengths and poorer enamel bonds. Aggressive SEPs provide the reverse.

This generation of adhesives combine conditioning, priming and application of adhesive resin. Adhesives belonging to this generation are mixes of hydrophilic and hydrophobic components [41]. Furthermore, these all-in-one adhesives contain uncured ionic monomers that bind to composite restorative materials directly [42, 43].

All-in-one adhesive systems must be acidic enough to demineralise the enamel and the dentin smear layer. As a consequence, the hydrophilicity of their resin monomers is high and might result for some insusceptibility for water degradation [44]. Due to the complex nature of the mixed solutions, these adhesives are prone to phase separation and formation of water droplets within their adhesive layers [41]. The adhesive layers can also act as semipermeable membranes, permitting bidirectional water currents [45]. Lower bond strengths were registered with the all-in-one systems as compared to the 3E&Ra and the 2E&Ra [45, 46]. More recent investigations demonstrate higher bond strengths to dentin for 2SEA (Clearfil SE Bond) as compared to 1SEA. The performance of one-step self-etch systems to dentin appears to be material dependent [47, 48].

De Moor and Delmé [14] concluded in their review of 2009 that acid etching with phosphoric acid remained mandatory on laser-irradiated enamel with respect to bond strength and marginal seal.

There was no value-added benefit of laser etching of the enamel after laser preparation. It has to be mentioned that laser etching is an insufficient term; laser conditioning is a more accurate term. Laser conditioning is a term used to describe a reduced ablative effect obtained with energy just above the threshold of ablation (15 J/cm² for Er:YAG and 20 J/cm² for Er;Cr:YSGG) (Figs. 5.1, 5.2, 5.3 and 5.4). In general, these investigations on laser etching were performed with etch-and-rinse adhesives. At that time, there was insufficient information on the effect of self-etching systems on the lased enamel.

The hybrid layer created with etch-and-rinse adhesives after erbium laser irradiation of dentin and after subsequent acid etching is thinner as compared to acid-etched bur-cut dentin (Figs. 5.5, 5.6, 5.7 and 5.8).

Different explanations have been put forward for the less profound hybridization:

It was concluded that there is a reduced thickness of the hybrid layer after erbium laser irradiation even though phosphoric acid was used. Subsurface damage is thought to be responsible for the decrease in bond strength and cohesive failure in the subbonding layer in dentin [82–84].

Also with self-etching adhesives, both subsurface damage and the thinner or absent hybrid layer resulted in low tensile bond strengths. It was recommended in 2,009 not to use ‘non-rinse’ or ‘self-etch’ adhesives. Higher microleakage at the enamel and dentin margins was also observed [14].

In recent years, however, more attention has been paid to the laser parameters (see Sect. 5.5.2) [85–89]. The latter explains the changed approach at present for the sixth-generation or two-step self-etch adhesives.

Pulse duration of the erbium lasers is related to the laser ablation ability and surface morphology, which are a very important factor for bond strength of adhesives to both the enamel and dentin [90]. Furthermore, one study demonstrated higher microtensile bond strength (μTBS) for Er:YAG than for Er,Cr:YSGG in combination with two-step etch-and-rinse adhesives (2E&Ra) (fifth generation) [91].

In general, acid etching of enamel margins resulted in better tensile bond strengths than the sole use of two-step self-etch adhesives (2SEA) (sixth generation). Only laser-pretreated enamel surfaces appeared not to achieve clinically acceptable values due to the more acid-resistant-lased surface. These findings also coincide with the present-day recommendations still to rely on the use of phosphoric acid to etch the lased enamel margins.

Investigations with two-step self-etch adhesives (2SEA) demonstrated that treatment of lased dentin with these 2SEA in general still resulted in lower μTBS values as compared to conventionally bur-prepared dentin surfaces [92–94]. The reduction in bond strength for Clearfil SE Bond self-etching adhesive (SEA), however, was lower than for a 2E&Ra [93]. The latter accounts for both Er:YAG and Er,Cr:YSGG [92]. These studies have in common that there is a superior adhesion for the adhesive system with 10-MDP, which has the potential to chemically interact with interfacial hydroxyapatite [93, 94].

Increasing etching time was also not able to increase the bonding strength of a 2E&Ra to irradiated dentin [92]. This is in contrast with 2E&Ra where etching during the 90 seconds resulted in increased bond strength as compared to 30 and 15 s [95]. In this respect, previous studies had already reported an increase in acid resistance of the enamel and dentin after laser irradiation [96–98]. An additional cleaning after with NaOCL after the etch procedure was also proposed [93] in order to improve the penetration of adhesive monomer into the irradiated dentin, as little or no hybrid layer formation was reported in previous studies [99, 100].

Research in the 2010s considered laser pretreatment again. Finishing the dentin at low fluency after laser cavity preparation resulted in higher bond strength then without an additional finishing for 2SEAs [101]. Depending on the formulation, a number of 1SEAs (seventh generation) demonstrate identical bond strengths on lased dentin as compared to bur-cut dentin [102]. In general, these adhesives belong to the group of ‘mild’ self-etch adhesives. More research however is needed, to explain these apparently unexpected different effects.