Japanese Dental Science Review 56 (2020) 190–207

Contents lists available at ScienceDirect

Japanese Dental Science Review

journal homepage: www.elsevier.com/locate/jdsr

Review Article

Current perspectives on dental adhesion: (1) Dentin adhesion – not

there yet
Jorge Perdigão
Professor, University of Minnesota, Department of Restorative Sciences, 515 SE Delaware St, 8-450 Moos Tower, Minneapolis, MN 55455, USA

a r t i c l e i n f o a b s t r a c t

Article history: The essential goal of any adhesive restoration is to achieve a tight and long-lasting adaptation of the
Received 5 June 2020 restorative material to enamel and dentin. The key challenge for new dental adhesives is to be simul-
Accepted 24 August 2020 taneously effective on two dental substrates of conflicting nature. Some barriers must be overcome to
accomplish this objective. While bonding to enamel by micromechanical interlocking of resin tags within
the array of microporosities in acid-etched enamel can be reliably achieved and can effectively seal the
restoration margins against leakage, bonding effectively and durably to organic and humid dentin is the
most puzzling task in adhesive dentistry.
Much of the research and development of dental adhesives has focused on making the clinical procedure
more user-friendly by reducing the number of bottles and/or steps. Although clinicians certainly prefer
less complicated and more versatile adhesive materials, there is a trade-off between simplification of
dental adhesives and clinical outcomes. Likewise, new materials are launched with claims of being novel
and having special properties without much supporting evidence.
This review article discusses dental adhesion acknowledging pioneer work in the field, highlights the
substrate as a major challenge to obtain durable adhesive restorations, as well as analyzes the three
adhesion strategies and their shortcomings. It also reviews the potential of chemical/ionic dental adhe-
sion, discusses the issue of extensively published laboratory research that does not translate to clinical
relevance, and leaves a few thoughts in regard to recent research that may have implications for future
adhesive materials.
© 2020 The Author. Published by Elsevier Ltd on behalf of The Japanese Association for Dental
Science. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction When clinical studies are completed, often a new version of the
same material has already been made available on the market. In
Dental adhesion was responsible for a paradigm shift in den- fact, dental adhesives can be launched without proof of clinical
tistry (Table 1). Dental adhesives have become one of the most efficacy, as the FDA usually reviews “the Section 510(k) premar-
intriguing biomaterials in Health Sciences. Research efforts in the ket notification of intent to market the device and determines the
last 20 years have shifted from clinically-proven multi-step den- device to be substantially equivalent to legally marketed predicate
tal adhesives to simplified versions that do not perform adequately devices” used for the same indications [3–5].
in laboratory and clinical studies [1,2]. The ideal goals for clinical It is extremely difficult for practicing dentists to keep updated
effectiveness and durability of the restorations have been fre- as so many dental adhesives are constantly launched on the market
quently neglected in favor of fewer number of bottles and quicker and updated or relaunched within short periods of time. In addition,
application of newer dental adhesives. dentists do not have access to the latest evidence-based informa-
Several obstacles must be overcome to accomplish the objective tion. As a result, dentists rely on the information provided by the
of developing a dental adhesive that bonds effectively to enamel industry representatives or information disseminated in contin-
and dentin, and achieves durable restorations that seal the margins uing education courses and dental meetings, often without solid
and provide some form of resistance to recurrent caries lesions. evidence to support the claims [6].
The continuous development and frequent introduction of den- The objective of this review article is to summarize the current
tal adhesives render existing materials outdated within a few years. evidence on dental adhesion, from the challenging substrate to the
latest trends, many of which do not extrapolate to sound evidence
pertinent to clinical practice.

E-mail address: [email protected]

https://doi.org/10.1016/j.jdsr.2020.08.004
1882-7616/© 2020 The Author. Published by Elsevier Ltd on behalf of The Japanese Association for Dental Science. This is an open access article under the CC BY-NC-ND
license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207 191

Table 1
Changes that resulted from the introduction of adhesives in Dentistry.

Positive Negative

The use of dental adhesives has expanded across different dental disciplines Clinicians tend to rely solely on adhesion as the source of primary retention in
(Operative Dentistry, Orthodontics, Pediatric Dentistry, Periodontology, clinical situations without enamel margins or without enough residual tooth
Prosthodontics, Endodontics) structure, as in core build-up composite resin restorations
Dental adhesives are used to retain a wide range of restorative materials – Potential for marginal bacterial leakage when the cavo-surface margin is in
glass-matrix ceramics, oxide ceramics, pre-polymerized composite resins, dentin/cementum
direct composite resins, metal-based restorations, fiber posts, fiber splinting
materials
More conservative tooth preparations (lesion-specific preparations) Post-operative sensitivity in posterior adhesive restorations related to
polymerization shrinkage stress
Reliable micromechanical retention to etched enamel without Enamel cracks in posterior adhesive restorations related to polymerization
macro-retention features shrinkage stress
Reinforcement of residual tooth structure; undermined enamel does not need Moisture contamination of the operatory field may be more detrimental for
to be always removed adhesive than for non-adhesive restorations
Increased resistance to recurrent caries lesions when dentin is impregnated Small monomers, such as HEMA, may easily seep into the pulp space and cause
with dental adhesive pulp inflammation
Increased resistance to caries lesions in sealed fissure systems of posterior Frequent open contacts in posterior composite resin restorations (compared to
teeth amalgam restorations)
High efficacy to treat root sensitivity Adhesives may cause pulp necrosis if applied in preparations close to the pulp
Some adhesives have antibacterial properties, which may prevent recurrent Some of the monomers in dental adhesives may cause contact dermatitis
caries lesions
Stronger retention of glass-matrix ceramics restorations; adhesives increase
the resistance to fracture of glass-matrix ceramics
Stable chemical adhesion to hydroxyapatite for some adhesive materials when
dentin is not etched with phosphoric acid

2. Milestones in dental adhesion Etching enamel with phosphoric acid (Fig. 1) is still considered,
sixty-five years later, as the gold standard for bonding resin-based
In 1952, a manuscript published in the British Dental Jour- materials to tooth structure. The interlocking of resin tags (Fig. 1)
nal by Kramer and McLean described an in-situ study that was within the microsized porosities left by enamel chemical etch-
carried out in 124 preparations of 118 teeth scheduled to be ing can effectively seal the restoration margins in the long-term
extracted for orthodontic reasons [7]. The authors used 15 com- [13]. There is clinical evidence that dental adhesives result in more
binations of restorative materials. Tooth sections were stained reliable clinical behavior when enamel is etched prior to the appli-
with hematoxylin/eosin upon extraction, and observed blindly cation of the adhesive [14]. Nonetheless, new adhesives are still
under the optical microscope. The authors then matched the being launched without recommendations for etching enamel with
data with the experimental groups. Some of the sections dis- phosphoric acid.
played “an altered reaction observed as a narrow zone of material Another milestone in adhesive dentistry occurred in 1960.
staining deeply with hæmatoxylin immediately bordering the cav- Rafael Bowen (who retired from the ADA Foundation Research
ity. This zone averaged about 3 m in thickness and was seen to Center in 2018 after 62 years of continuous service) and Mario
be composed of dentine having an intense affinity for hæmatoxylin. Rodriguez presented a paper [15] at the IADR meeting in Chicago
This change was present in all of the 28 teeth filled with Sevriton- that reported the tensile strength of several materials, including
adhesive. Similar changes were absent from all of the 96 teeth filled a new silica-resin material that contained “about 70 per cent vinyl
with other materials.” The specific chemically-cured adhesive used silane-treated clear fused silica combined with about 30 per cent of an
in this study had been developed in 1949 by Oskar Hagger, a adduct of glycidyl methacrylate and bisphenol A”. It is worth under-
chemist who worked for DeTrey/Amalgamated Dental Company lining here the inclusion of a silane to bond the inorganic filler to the
[8]. The adhesive contained a phosphate monomer, later identi- new Bis-GMA resin. By 1963, the full composition of the new mate-
fied by Dr. Buonocore’s research group as glycerol phosphoric acid rial had been finalized [16]. Based on Bowen’s research with the
dimethacrylate [9], which is still used in a few dental adhesives Bis-GMA molecule, the first commercial composite resin (Addent,
as GPDM [10]. Remarkably, the findings of the 1952 manuscript 3 M) was introduced in 1964 as a chemically-cured paste-paste
[7] were the first reference to the concept currently known as material. Interestingly, since that time most changes in composite
the hybrid layer, elegantly illustrated with an image of dentin resin technology have been in the filler particle size and distribu-
altered by the adhesive [7]. In addition, the use of the phosphate tion rather than in the resin matrix, which is still based on Bis-GMA,
monomer GPDM as a dentin adhesive may now be part of history also known as Bowen’s resin.
as the first research report of a self-etch adhesive in the litera- Alan Wilson and Brian Kent, working at the Laboratory of the
ture. Government Chemist in the UK, invented in 1968 one the most
In 1955, a major advance for the history of dental adhesion was groundbreaking materials in dentistry, for which the patent was
published in the Journal of Dental Research [11]. Michael Buono- applied for in 1969 [17]. This self-adhesive material was widely
core used 85% phosphoric acid to change enamel surfaces and make known as glass-ionomer cement (GIC), although the correct ter-
them more suitable for mechanical adhesion, using an industrial minology is glass polyalkenoate cement [18]. The first report of
technique that improved the adhesion of paints to metal surfaces. their findings in the literature appeared in 1971 [19]. The commer-
Buonocore later expanded his acid-etch technique to clinical den- cial version was subsequently launched in Europe in 1975 under
tistry to seal pits and fissures, as reported in 1967 [12]. The authors the commercial name ASPA by Amalgamated Dental International,
used 50% phosphoric to etch pits and fissures, followed by the appli- DeTrey Division.
cation of a silica-filled methacrylate adhesive. This novel technique, Takao Fusayama, defying the general belief that etching dentin
which was not standard of care in 1967, resulted in a reduction of caused irreversible pulp damage, in 1979 reported that etching
caries incidence in pits and fissures by as much as 86.3% at 1 year dentin and enamel with 40% phosphoric acid for 60 sec substan-
[12]. tially improved the adhesion of Clearfil Bond System-F (Kuraray)
192 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207

Fig. 1. (a) SEM micrograph of human enamel etched with 35% phosphoric acid for 15 s. Original magnification = 5,000X. (b) SEM micrograph of a replica of an interface of
etched enamel with a dental adhesive. After curing the adhesive, the specimen was left in 6N HCl for 12 h to dissolve the enamel. The enamel prisms at the interface were
not dissolved because the etched enamel was impregnated with polymerized adhesive, creating a hybrid layer. H – hybrid layer; Ad – adhesive; E – residual enamel. Original
magnification = 5,000X.

[20]. For the first time, the concept of “total-etch” was associated interlocking of resin monomers into the array of microporosities
with improved dentin adhesion. left by the acid chemical dissolution of enamel (Fig. 1). Bonding to
Nobuo Nakabayashi’s research team was responsible for another enamel after etching with phosphoric acid is certainly the founda-
breakthrough in 1982, when they reported for the first time tion for the durability of adhesive restorative procedures.
a “demineralization-resistant zone” [21] in dentin after etching Dentin is a complex biocomposite structure, defined by some
dentin with 10% citric acid-3% ferric chloride (10:3 solution) for authors as a puzzle of different types of dentin and by other authors
30 sec and applying 4-META (methacryloxyethyltrimellitate anhy- as a bone-like nanocomposite built of carbonated hydroxyapatite
dride) cured with tri-n-butyl borane. This concept of hybrid layer mineral particles, protein, and water [24,25]. As opposed to enamel,
in etched dentin was identified using the scanning electron micro- dentin is a humid and more organic substrate. Dentin adhesion
scope (SEM). The same authors also highlighted the importance of has been one of the most challenging and less predictable tasks
monomers with both hydrophobic and hydrophilic groups to pro- in adhesive dentistry due to the dynamic compositional differences
mote adhesion with tooth substrates by penetration and infiltration and complex histology of dentin. The ability of adhering restorative
dentin as a new concept in biocompatible materials for dental use materials intimately to dentin is affected by many factors, includ-
[21]. ing biological and clinical factors. These factors include the patient’s
The most recent milestone was associated with the adhesion- age, location of the tooth in the mouth, dentin depth and perme-
decalcification concept (AD-concept) for adhesion to dentin ability, pulpal fluid flow, presence of sclerotic and/or carious dentin,
[22,23]. This concept was originally described for carboxylic acids. radicular versus coronal dentin, type of restorative material and
When these acids are applied on hydroxyapatite they first form procedure, isolation, parafunctional habits, dentist’s experience,
ionic bonds to calcium, which may be dependent on the pKa of among others [26–31].
each acid. While some of the carboxylic acids, such as oxalic acid, The mineral phase (hydroxyapatite) of dentin is on average
stay attached to calcium on the hydroxyapatite surface resulting 45 vol%, while the organic matrix is 33 vol%, the remainder being
in insignificant decalcification, other carboxylic acids result in a water [32]. Type I collagen is the most abundant protein in the
significant decalcification of hydroxyapatite with minimal or no organic phase. Dentin encloses a maze of inverted-cone shaped
chemical attachment. This adhesion-decalcification (AD) concept is tubules that traverse dentin, radially oriented with the larger diam-
still relevant. Etch-and-rinse (ER) adhesives follow the decalcifica- eter facing the pulp [26]. Garberoglio and Brännström in 1976 [33]
tion pathway derived from phosphoric acid etching, whereas mild measured the area occupied by the tubules and the tubular diam-
self-etch (SE) adhesives (pH ≈ 2), such as those that contain 10- eter in 30 extracted teeth. The number of tubules near the pulp
methacryloyloxydecyl dihydrogen phosphate or 10-MDP (MDP), was 45,000 per square millimeter and their diameter 2.5 ␮m. In
tend to follow the adhesion pathway. Nevertheless, mild SE adhe- middle dentin, the number of tubules was 29,500/mm2 and the
sives still cause minimal decalcification, which is still required for average diameter was 1.2 ␮m. In superficial dentin, the area occu-
calcium release and subsequent formation of stable MDP-Ca salts pied by tubules was 20,000/mm2 and the average tubule diameter
and respective nanolayering, as discussed later in this article. was 0.9 ␮m [33]. The contents of water increase 20-fold from super-
ficial to deep dentin. The mean tubule volume in coronal dentin is
10% of the entire dentin volume, while near the DEJ it is 4% and
3. The substrate increases to 28% near the pulp [33] (Fig. 2).
Dentin tubules are permeated with fluid under constant out-
Enamel and dentin are the dental substrates to which we bond ward pulpal pressure estimated to be 25 to 30 mm Hg [34]. In
our restorative materials. Cementum may also be involved when addition, there is fluid present within the intertubular dentin
the cavo-surface margin is located apically to the cementum- area, making dentin an intrinsically moist hard tissue throughout
enamel junction. its internal structure. Dentin contains extensions of the odonto-
Enamel is a dry substrate without vital structures containing blast (odontoblastic processes) and intra-tubular collagen fibers
92 vol% of mineral phase (hydroxyapatite), which makes enamel in deeper areas (Fig. 3), less frequently in middle and superficial
almost the ideal substrate to form a tight adhesive joint. The dentin. These characteristics, which we sometimes overlook as clin-
acid-etch technique [11] is still the gold standard for bonding icians, attest the greater challenge when an adhesive restoration is
resin-based materials to tooth structure. The micromechanical inserted in deep dentin compared to restorations placed in more
interaction of adhesives with enamel is a result of the diffusion and superficial dentin.
 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207 193

Fig. 2. (a) SEM micrograph of fractured superficial dentin. Int – intertubular dentin; P – peritubular dentin; T – dentin tubule; Arrows – bacteria in the tubule lumen. Original
magnification = 10,000X. (b) SEM micrograph of fractured deep dentin 75 ␮m from the pulp of the same tooth in Fig. 2a. Original magnification = 10,000X.

Fig. 3. (a) SEM micrograph of fractured middle dentin showing an odontoblastic process extending from the tubule Int – intertubular dentin; P – peritubular dentin; T –
dentine tubule. Original magnification = 10,000X. (b) SEM micrograph of fractured deep dentin showing intratubular collagen (asterisk) with the characteristic 64 nm collagen
banding pattern. Int – intertubular dentin; P – peritubular dentin; T – dentin tubule. Original magnification = 25,000X.

ical bonding to dentin is facilitated by the smaller size and the

crisscross orientation of the hydroxyapatite crystallites [35]. The
different crystallite size facilitates chemical bonding to dentin to
the detriment of chemical bonding to enamel [35].
Physiological changes resulting from dentin aging or in response
to carious lesions and other aggressive stimuli increase the degree
of mineralization of dentin, resulting in an increase in dentin thick-
ness and a reduction of dentin permeability [28,29]. Reduction of
permeability with age may have a direct effect on dentin bond
strengths [36] as dentin permeability affects the adhesion process.

4. Relevance of in vitro testing for dental adhesion

Several peer-reviewed publications have covered this topic

very well [37–39]. It is still important to emphasize that labora-
tory tests of dental adhesives usually employ extracted impacted
Fig. 4. SEM micrograph of fractured dentin-enamel junction (DEJ) area. Note the
third molars and premolars extracted for orthodontic reasons. This
morphological differences between enamel and dentin hydroxyapatite. Original ‘laboratory-type’ (or unaffected dentin substrate) lacks clinical rel-
magnification = 5,000X. Arrows – DEJ; E – enamel; Int – intertubular dentin; T – evance, as clinicians place adhesive restorations in teeth that have
dentin tubule. had carious lesions, failed restorations, or non-carious cervical
lesions with sclerotic dentin continuously exposed to the oral envi-
In addition to the compositional differences, there is a subtle ronment.
difference between the enamel and dentin hydroxyapatite crys- The clinically relevant substrates for dentin adhesion include
tallites. Enamel crystallites are larger and have a more regular affected dentin, which is located immediately underneath the car-
and parallel oriented arrangement, whereas dentin hydroxyapatite ious dentin area. Affected dentin is slightly decalcified but sill
crystals are smaller and arranged in a crisscross pattern within amenable to recalcification, with odontoblastic processes, sound
the organic matrix (Fig. 4), rendering it more difficult to establish collagen fibers, and apatite crystals bound to the fibers [40,41].
a micro-mechanical interlocking with dentin. Conversely, chem- Continuous deposition of mineral within the tubules underneath
194 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207

ent steps used in the adhesive procedure. This nomenclature will

be used in this manuscript (Table 2). Adhesives that include a
phosphoric acid-etching step are known as etch-and-rinse (ER)
adhesives. They dissolve and remove the smear layer and smear
plugs (Fig. 7b). Adhesives that do not use a separate etching step
are known as self-etch (SE) adhesives, as they do not remove the
smear layer, but incorporate into the adhesive interface (Fig. 7c).
Self-adhesive (SA) materials (dental adhesive and restorative com-
ponent all-in-one material), belong to another strategy for adhesion
to tooth structure, as either composite resins or GIC-based materi-
als. The latter use a separate polyalkenoic acid (often polyacrylic)
dentin conditioner that is rinsed off (Fig. 7d).
The basic components of a dental adhesive system are

A) Etchant, currently phosphoric acid in a concentration between

Fig. 5. Non-carious cervical lesion (NCCL) with sclerotic dentin. 30% and 40%. Most phosphoric acid gels are thickened with sil-
ica microparticles, although there are a few that contain other
thickeners such as xanthan gum. A color dye is always included
a carious lesion process results in tubular obliteration and the for-
to improve the application accuracy and ensure that all gel is
mation of sclerosis, and potentially reducing bond strengths [28].
washed off. Glycol is often added to improve wettability and
Another type of clinically relevant dentin that may be found in
decrease viscosity. The etchant is always rinsed off from the
deep caries lesions is reactionary tertiary dentin, which is formed by
tooth surface.
odontoblasts in the pulp chamber wall near the area corresponding
B) Primer which is a hydrophilic solution of resin monomers,
to the carious lesion or to the occlusal trauma. In case of a pulp expo-
organic solvent (alcohol or acetone), water, and stabilizers. The
sure, newly differentiated odontoblast-like or odontoblastoid cells
hydrophilic groups boost the wettability to the dentin surface,
replace the irreversibly injured odontoblasts at the exposure site
which is a humid environment. The role of primers in dentin
and form a bridge of atubular dentin known as reparative tertiary
adhesives is comparable to that of primers in paints. The primer
dentin [42].
adheres to surfaces and forms a binding layer that is better pre-
Sclerotic dentin is common in areas where dentin has been
pared to receive the paint, in this case, the bonding resin. Primers
exposed to the oral environment, such as non-carious cervical
are not usually rinsed off nor cured once placed on the tooth
lesions (NCCLs) (Fig. 5). These lesions contain a complex dentin
surface; they are only air-dried.’
substrate with different ultrastructural layers [43], including a
C) The bonding resin is a solvent-free (hydrophobic) low-viscosity
hypermineralized layer on the surface with denatured collagen
resin that is applied over the primer and then light-cured. The
and bacteria (Fig. 6). The tubules appear obliterated by crystalline
hydrophobic groups interact and copolymerize with the restora-
deposits. Etching sclerotic dentin in NCCLs is difficult to achieve
tive material and make dentin bonding more stable and more
[43,44], as phosphoric acid does not etch underneath the hyper-
durable by sealing the bonded interface against nanoleakage
mineralized surface layer and is unable to dissolve the sclerotic
[47,48]. The hydrophobic resin improves both the polymeriza-
casts in the tubules. Due to the intricacies of the substrate, restora-
tion rate of the primer and the mechanical properties of the
tions of NCCLs have a higher tendency to fail than restorations in
adhesive and hybrid layer [48]. Adhesive systems that have this
other areas of the mouth [43,45]. Roughening the superficial area
separate bonding step result in better in vitro and clinical out-
of hypermineralized dentin (and enamel) in NCCLs improves the
comes [49].
survival rate of restorations [46].

Dental adhesives may involve three separate procedures or

5. Current dental adhesives
steps, two of them merged in one, or all of them merged in one
Dental adhesives are currently categorized using two different (Table 2).
classifications.
5.1. Etch-and-rinse (ER) adhesives
I By generation – from first to eighth generations.
The advantages and disadvantages of ER adhesives ae summa-
It is used mostly by the dental industry to highlight the latest rized in Table 3. The first ER adhesives were 3-step adhesives. They
trend. It is a confusing nomenclature, as the first dentin adhe- include an acid etchant applied to enamel and dentin simultane-
sives that used a phosphoric acid etchant on enamel and dentin ously to remove the smear layer and demineralize the substrate.
are known as the fourth generation. This classification is not very Phosphoric acid has been the standard etchant but other acids
informative either, especially when considering the missing com- have been unsuccessfully used, namely maleic, nitric and oxalic
ponents of the dental adhesive, i.e., etchant, primer and bonding acids [57–59]. The 3-step ER adhesives also include a primer and
resin. a hydrophobic bonding resin that infiltrates the demineralized
dentin (Dd in Fig. 7b) to form a hybrid layer of collagen and resin
II By adhesion strategy – with or without etching enamel and (Fig. 8). Regarding the phosphoric acid etching agent, the current
dentin simultaneously with phosphoric acid (Table 2). market tendency is for newer gels (Fig. 9a) being less viscous (run-
nier) than older phosphoric acid gels (Fig. 9b), which is likely to be
This classification is easier to understand for practicing den- as a result of extra glycol added to the gel. Although this detail is
tists than the classification by generation, as adhesives are grouped often overlooked, newer gels are much less aggressive on enamel
according to their interaction with the tooth structure, more pre- and dentin than their predecessors [56] (Table 4) without reducing
cisely according to the way they interact with the smear layer dentin bond strengths. However, their effect on enamel bonding
(Fig. 7a). In addition, it is informative in regards to the differ- durability needs to be further investigated.
 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207 195

Fig. 6. Sequence of SEM micrographs to illustrate the morphological characteristics of the bonding substrate in a NCCL of a recently extracted mandibular canine. (a) SEM
micrograph depicting a general view of the NCCL. The incisal aspect is on the right side (E-enamel), with the cervical aspect on the left side (R-root). The white arrows point
to the natural incisal cavo-surface margin. The dark arrows point to the natural cervical cavo-surface margin. Original magnification = 30X. (b) SEM micrograph of the area
included in the rectangle in Fig. 6a. The horizontal dotted line separates the unetched area (upper half) from the area that was etched with 35% phosphoric acid for 15 s
(lower half). Original magnification = 100X. (c) SEM micrograph of the area included in the rectangle of Fig. 6b (etched area). Note the sclerotic casts in the tubules (circles)
and, overall, hypermineralized dentin. Original magnification = 1,000X. (d) SEM micrograph of a sclerotic cast (asterisk) obliterating the tubule (T). Note how intertubular
dentin (Int) is densely mineralized. Original magnification = 15,000X. (e) SEM micrograph of bacteria (arrows) ‘fossilized’ into the mineralized area of intertubular dentin
(Int). Original magnification = 15,000X. (f) SEM micrograph of a longitudinal fracture of dentin in a NCCL. Note how the tubule is obliterated with rhombohedral mineral
crystals, which were elegantly described in 1989 [173]. Int – intertubular dentin; P – peritubular dentin; T – dentin tubule. Original magnification = 15,000X.

Since the early 1990s, the concept of wet or moist dentin for ER ter retention rate for ER adhesives when compared to dry dentin
adhesives [60,61] has been widely advocated and taught in den- [64,65].
tal schools. The collagen left in the area of demineralized dentin Leaving the dentin moist after etching and rinsing may not be so
collapses when dentin is air-dried after rinsing off the etchant. crucial with current simplified adhesives, as agitation of the adhe-
This collapse results in incomplete infiltration of the adhesive into sive during application improves infiltration of the monomers into
demineralized intertubular dentin [62] and lower bond strengths etched dentin. In fact, a clinical study in non-carious cervical lesions
[63]. For this reason, keeping the dentin substrate moist (glis- (NCCLs) found that passive application of the adhesive resulted in
tening) after rinsing off the etching gel has been recommended 82.5% retention rate after 2 years compared to 92.5% retention rate
based on in vitro testing. However, in vitro tests are carried out in of the restorations in which the adhesive was scrubbed vigorously
“laboratory-type” or unaffected dentin, as discussed above. When [66]. Furthermore, leaving dentin moist has been shown to cause
the degree of dentin moisture was tested in clinical trials with degradation of the resin dentin interfaces at 6 months [67]. Den-
the same or similar ER adhesives, including a recent study with tists are advised to gently dry dentin after rinsing off the gel without
a universal adhesive, moist dentin was not associated with bet- inducing desiccation.
196 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207

Table 2
Classification of dental adhesives by adhesion strategy.

Ac – Phosphoric acid; Pr – Hydrophilic primer; BR – Non-solvated bonding resin; GIC– glass ionomer cement; PAA -
polyacrylic acid.

Fig. 7. (a) SEM micrograph of a fractured human dentin specimen with smear layer and a smear plug created with diamond bur. Int – intertubular dentin; P – peritubular
dentin; T – dentin tubule; Sp – smear plug; Oc –occlusal surface; Dotted circle – another tubule plugged with smear layer. Original magnification = 10,000X. (b) SEM
micrograph of human dentin etched with liquid phosphoric acid for 15 sec. Int – intertubular dentin; P – peritubular dentin; T – dentin tubule; Oc – occlusal surface; Pc –
exposed peritubular collagen from dissolution of the peritubular dentin; Dd – dentin demineralized by the etching agent; arrows – other tubules. Original magnification =
10,000X. (c) SEM micrograph of human dentin treated with the Clearfil SE primer (Kuraray) from the 2-step SE adhesive Clearfil SE Bond. The asterisk marks the area of dentin
partially decalcified by the primer (pH = 1.8–2.0). Upon application of the respective hydrophobic bonding resin, this 0.5 ␮m deep area will become the hybrid layer. Int –
intertubular dentin; T – dentin tubule; Oc – occlusal surface; Sp – primer-infiltrated smear plug. Original magnification = 15,000X. (d) SEM micrograph of occlusal view of
human dentin treated with GC Cavity Conditioner (20% polyacrylic acid with 3% aluminum chloride hexahydrate) for 10 sec, and rinsed with water for 15 sec. Note residual
smear layer (ovals) and some patent tubules (T). The intertubular dentin does not have morphological characteristics of demineralization (no visible collagen fibers). Original
magnification = 5,000X.
 J. Perdigão / Japanese Dental Science Review 56 (2020) 190–207 197

Table 3
Etch-and-rinse adhesives.

Advantages Disadvantages

Three-step ER adhesives have been available since the 1990s, therefore they Acetone-based adhesives need more applications than those recommended by
have a long-track record. the respective manufacturers [51].
High immediate dentin and enamel bond strengths in laboratory studies Over-etching decreases bond strengths [52].
Excellent bonding to enamel in vitro and durable restorations in clinical More technique sensitive than SE adhesives, as the potential for incomplete
studies. However, retention rates for 2-step ER adhesives are lower than for infiltration of the adhesive into the etched dentin depends on several
3-step ER adhesives [49]. contributing factors occurring simultaneously in a very short time.
Clinical studies over 5 years with excellent results for specific 3-step ER Hydrolytic degradation of the bonds occurs when margins are located in
adhesives. Optibond FL (Kerr) resulted in excellent retention at 13 years in dentin.
NCCLs [50]. Optibond FL is still the reference against which all ER adhesives
are compared.
As these adhesives contain organic solvents such as ethanol or acetone, minor The clinical and in vitro performance of 2-step ER adhesives undergo
dentin contamination with saliva does not always decrease bond strengths degradation faster than that of 3-step ER adhesives.
in vitro.
As opposed to 2-step ER adhesives, 3-step ER adhesives contain a hydrophobic Bond strengths may vary with the degree of moisture, depending on the
bonding resin that prevents or delays the degradation of the resin-dentin specific adhesive.
interface by making the interface impermeable and increasing the film
thickness. The lack of solvent increases the degree of conversion.
ER adhesives may result in mechanical interlocking with etched dentin Although clinical evidence demonstrates that ER adhesives do not cause more
provided that the dentin in not overetched, the primer/adhesive for 2-step post-operative sensitivity than SE adhesives [53,54], some clinicians claim that
ER adhesives is applied in an active scrubbing mode, and that there is not ER adhesives cause higher incidence of post-operative sensitivity with
excessive water within the interfibrillar spaces. posterior composite restorations.
Ability to bond composite, porcelain, fiber posts, etched or sandblasted metals, Solvent air-drying time recommended by the manufacturers is insufficient
or amalgam. [55] and must be extended.

Table 4
Median intertubular dentin demineralization of current phosphoric acid gels [56].

Etching gel Intertubular dentin

demineralization (␮m)
Etching time =15 sec

Ultra-Etch 35% 1.155A

(Ultradent)
Scotchbond Universal Etchant 32% 1.675AB
(3 M Oral Care)
Tooth Conditioning Gel 34% 2.185B
(Dentsply Sirona)
Total Etch 37% 2.820C
(IvoclarVivadent)
Select HV Etch 35% 3.090C
(Bisco)
Scotchbond Etchant 35% 3.205C
3 M Oral Care)
Gel Etchant 37.5% 4.033D
(Kerr)
Liquid etchant 35% 4.636D
(prepared in laboratory) Fig. 8. TEM micrograph of the adhesive-dentin interface formed by the 3-step
ER adhesive OptiBond FL (Kerr). The particle-filled hydrophobic bonding adhesive
Three dentin disks per etching gel; 4 measurements in each half disk in identical
resulted in filled resin tags (Rt). Ad – adhesive; H – hybrid layer; D – dentin. Original
areas, 24 measurements per etching gel; measurements taken with image analysis
magnification = 6,000X.
software (ImageJ, NIH). Statistical analysis included Kruskal-Wallis and Median tests
at p < 0.05

Fig. 9. (a) SEM micrograph of human dentin etched with 32% phosphoric acid (Scotchbond Universal Etchant, 3M). Original magnification = 7,000X. (b) SEM micrograph
of human dentin etched with 35% phosphoric acid (Scotchbond Etchant, 3M). Original magnification = 7,000X. Int – intertubular dentin; P –peritubular dentin; T – dentin
tubule; Oc – Occlusal surface; Pc – exposed peritubular collagen from dissolution of the peritubular dentin; Dd – dentin demineralized by the etching agent; Circles – silica
thickening agent; Arrows – intertubular anastomoses.