d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419

available at www.sciencedirect.com

journal homepage: www.intl.elsevierhealth.com/journals/dema

The role of HEMA in one-step self-etch adhesives

K.L. Van Landuyt a , J. Snauwaert b , M. Peumans a , J. De Munck a ,

P. Lambrechts a , B. Van Meerbeek a,∗
a Leuven BIOMAT Research Cluster, Department of Conservative Dentistry, School of Dentistry, Oral Pathology
and Maxillo-Facial Surgery, Catholic University of Leuven, Kapucijnenvoer 7, B-3000 Leuven, Belgium
b Molecular and Nanomaterials, Department of Chemistry, Catholic University of Leuven,

Celestijnenlaan 200d, 3001 Heverlee, Belgium

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

Article history: In spite of its high allergenic potential, 2-hydroxyethyl methacrylate (HEMA), a low-
Received 13 August 2007 molecular-weight monomer, is frequently used in adhesives for its positive influence on the
Accepted 11 February 2008 bond strength. In addition, the presence of HEMA in one-component one-step adhesives
can prevent phase separation.
Objectives. In search of improved bonding effectiveness, the 24-h bond strength of four
Keywords: experimental one-step self-etch adhesives with different concentrations of HEMA to bur-cut
Adhesion enamel and dentin was determined using a micro-tensile bond strength protocol.
HEMA Methods. The tested experimental adhesives (Exp-0, Exp-10, Exp-19 and Exp-36) only differed
One-step self-etch adhesive in their concentration of HEMA, which was 0, 10, 19 and 36%, respectively. With an increasing
Phase separation concentration of HEMA, the concentration of acetone was decreased. Besides bond strength,
Droplets the adhesives were also examined by light-microscopy for phase separation. The interface
Osmosis was investigated by SEM and TEM.
Hydrophilic Results. Regarding bond strength, Exp-10 performed best. Even though Exp-36 was the only
adhesive formulation that did not exhibit phase separation on a glass plate, it yielded
the lowest bond strength. Accordingly, droplets could be observed by SEM and TEM in the
adhesive layers of all adhesives, except for Exp-36 on enamel.
Conclusion. A small amount of HEMA (10%) improved the bond strength of a one-step self-
etch adhesive. When added in higher concentrations, this beneficial effect of HEMA on the
bond strength is lost due to increased osmosis, which resulted in many droplets; due to
reduced polymerization conversion; and sub-optimal physico-mechanical properties of the
resultant poly-HEMA containing adhesive interface.
© 2008 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

1. Introduction [4]. HEMA has been reported to positively influence bond

2-Hydroxyethyl methacrylate (HEMA) is a water-soluble
methacrylate monomer frequently present in dental adhe-
sives [1]. Its polar properties and small dimensions enhance
the wetting properties of the adhesive solution [2,3] and the
penetration efficacy of the adhesive into demineralized dentin
strength to dentin [5]. Because of its hydrophilic character, it
is also frequently added to improve miscibility of hydrophobic
and hydrophilic components in an adhesive solution [6–8].
The major drawback of the use of HEMA is its high aller-
genic potential. Fast penetration of this monomer through
gloves and through skin due to its low molecular weight can

Corresponding author. Tel.: +32 16 33 75 87; fax: +32 16 33 27 52.

∗

E-mail address: bart.vanmeerbeek@med.kuleuven.be (B. Van Meerbeek).

0109-5641/$ – see front matter © 2008 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.dental.2008.02.018
 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419
cause contact dermatitis (allergic reaction type IV), which is a
well-known occupational disease among dentists [9–11]. As
HEMA is not exclusively used in dentistry (varnish, paints,
lenses, fake nails, etc.), cross-allergic reactions have also
been described [11]. Beside that, HEMA, being a mono-
methacrylate, may also deteriorate the mechanical properties
of the polymerized adhesive, which can result in inferior
mechanical strength of the cured adhesive, enhanced water-
uptake, swelling and staining [8,12].
Recently, HEMA was shown to play an important role
in preventing phase-separation reactions in one-component
one-step self-etch adhesives [6]. In these ‘one-bottle’ solu-
tions, polar and apolar ingredients are blended together with a
solvent such as water, acetone and ethanol. The solvent keeps
the ingredients in solution, but once dispensed subsequent
evaporation of the solvent can trigger a phase-separation reac-
tion with the formation of multiple droplets. Macroscopically,
this can be observed as an uncured drop of adhesive solution
that becomes opaque. Microscopically, many droplets in the
uncured adhesive can be seen. When the adhesive is cured
before the end of the separation reaction and thus before com-
plete removal of the droplets, the droplets get frozen in the
adhesive layer. HEMA can prevent such a phase-separation
reaction by replacing the solvent and keeping the ingredients
in solution [1].
The objective of this study was first to investigate the effect
of HEMA on the adhesive performance of one-component
one-step self-etch adhesives and second to determine the con-
centration of HEMA needed to prevent phase separation. The
null hypothesis advanced in this study was that HEMA did not
influence the bond strength of one-step self-etch adhesives.
2. Materials and methods
Four experimental mild (pH 2) one-component (one-bottle)
one-step self-etch adhesives containing a carboxylate and
phosphate-based functional monomer were prepared. Their
ingredients are listed in Table 1. They had an identical compo-
sition but for the ingredient HEMA. Exp-0, Exp-10, Exp-19 and
Exp-36 contained respectively 0, 10, 19 and 36% of HEMA. With
an increasing concentration of HEMA, the concentration of the
solvent acetone was reduced. In Exp-36 the concentration of
acetone was substantially reduced.
Table 1 – Adhesives investigated, their composition and application procedure
Adhesive Manufacturer Compositiona,b
Exp-0 GC, Tokyo, Japan 4-MET, phA-m, DMA, acetone, water, filler,
photoinitiator, stabilizer
Exp-10 HEMA (10%), 4-MET, phA-m, DMA, acetone, water,
filler, photoinitiator, stabilizer
Exp-19 HEMA (19%), 4-MET, phA-m, DMA, acetone, water,
filler, photoinitiator, stabilizer
Exp-36 HEMA (36%), 4-MET, phA-m, DMA, acetone, water,
filler, photoinitiator, stabilizer
a
Abbreviations of monomers in alphabetical order: DMA, dimathacrylates;
methacryloxyethyltrimellitic acid; phA-m, phosphoric acid ester monomer.
b
Underlined words refer to differences in composition.
1413
3. Light-microscopy
All adhesive solutions were examined (uncured) for a poten-
tial phase-separation reaction and the presence of droplets
by light-microscopy (LM) (Olympus BH2, Hamburg, Germany).
A drop of each self-etching solution was dispensed onto a
glass plate, and imaged real-time at different magnifications
(140–280×) using a digital camera (JVC TK-870E, Yokohama,
Japan).
4. !TBS-testing
The bond strength to enamel and dentin was determined
using a standardized micro-tensile bond strength proto-
col. Non-carious human third molars (gathered following
informed consent approved by the Commission for Medical
Ethics of KULeuven) were stored in 0.5% chloramine/water
at 4 ◦ C and used within 1 month after extraction. To prepare
dentin samples, the occlusal crown third was removed with
a diamond saw (Isomet 1000, Buehler, Lake Bluff, IL, USA),
thereby exposing a flat mid-coronal dentin surface. A stan-
dardized bur-cut smear layer was produced by removing a thin
layer of the surface using a Micro-Specimen Former (Univer-
sity of Iowa, Iowa City, IA, USA), equipped with a high-speed
regular-grit (100 !m) diamond bur (842, Komet, Lemgo, Ger-
many). For enamel, a flat surface was ground using the same
bur at the buccal and lingual surface of a tooth. All four
adhesives were applied on air-dried, but not overly desiccated
enamel and dentin. After a dwell time of 10 s, the adhesive was
strongly air-blown with maximum pressure and subsequently
light-cured for 10 s (Table 1). Composite build-ups were made
with Gradia Direct Anterior (GC, color A2) in three or four lay-
ers to a height of 5–6 mm. After 24 h storage in distilled water
(37 ◦ C), rectangular sticks (2 mm × 2 mm wide; 8–9 mm long)
were sectioned perpendicular to the adhesive-tooth interface
using the Isomet saw. Only the four central dentin sticks
were used to eliminate substrate regional variability [13,14].
The sticks were trimmed at the interface into an hourglass
shape (diameter of ±1.1 mm) using the Micro-Specimen For-
mer, equipped with a fine-grit (30 !m) diamond bur (5835KREF,
Komet) in a high-speed handpiece under air/water coolant.
The specimens were fixed to Ciucchi’s jig with cyanoacrylate
Application
(1) Apply adhesive to the entire surface with a
disposable applicator
(2) Keep dentin wet (shiny surface) with adhesive for
at least 10 s; strongly air-dry
(3) Light-cure for at least 10 s
HEMA, 2-hydroxyethyl methacrylate; 4-MET, 4-
1414 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419

Fig. 1 – Light-microscopical analysis to detect phase separation. In the upper row, the phase-separation reaction of Exp-0 is
demonstrated. As soon as the drop of the uncured adhesive’s solution was exposed to air and the solvent has started to
evaporate, a phase-separation reaction was triggered. A multitude of droplets was formed. This could also be observed with
the bare eye, as the drop turned opaque. These droplets disappeared with time. After 1 min, the majority of droplets were
still present. Notice the droplet-free halo at the periphery of the drop. It took more than 6 min before all droplets had
vanished. In the lower row, the HEMA-containing experimental adhesives are shown. Only Exp-36 did not exhibit phase
separation, not even after several minutes.

glue (Model Repair II Blue, Dentsply-Sankin, Ohtawara, Japan)
and stressed in tension at a crosshead speed of 1 mm/min
using a universal testing device (LRX, Lloyd, Hampshire, UK).
The !TBS was derived by dividing the imposed force at the
time of fracture by the bond area (mm2 ). Statistical differences
were examined using Kruskal–Wallis non-parametric statis-
tical analysis (˛ = 0.05). The mode of failure was determined
with a stereomicroscope at 50× magnification.
Representative dentin and composite !TBS-fracture
planes, exhibiting the most frequently observed failure mode
and a !TBS close to the mean, were processed for field-
emission gun scanning electron microscopy (Feg-SEM, Philips
XL30, Eindhoven, The Netherlands) using common specimen
processing described previously [15].
5. TEM and SEM interface characterization
In order to investigate the effect of HEMA on the adhe-
sive interface, dentin specimens of all adhesives tested
were prepared for TEM. Further specimen processing fol-
lowed the procedure described in detail by Van Meerbeek
et al. [16]. Non-demineralized ultra-thin sections were cut
(Ultracut UCT, Leica) and stained with 50 wt.% ammo-
niacal silver-nitrate solution [17], after which they were
examined by TEM (Philips CM10). As voids observed by
TEM, especially when situated at the adhesive–composite
interface, cannot be distinguished from artifacts, the epoxy-
resin-embedded blocks that remained after diamond-knife
ultra-microtomy were also examined using SEM to bet-
ter determine the presence and the position of interfacial
droplets.
6. Results
LM analysis of uncured adhesive solutions revealed a phase-
separation reaction in all experimental adhesives but Exp-36
(Fig. 1, Table 2).
Addition of HEMA to the HEMA-free Exp-0 did improve the
bond strength to enamel, however only significantly for small
concentrations (10 and 19% HEMA). When the concentration
of HEMA was increased to 36%, no significant effect on the
bond strength compared with Exp-0 was evident (Fig. 2).
A similar beneficial effect of HEMA on the bond strength
to dentin could be observed (Fig. 3). 10% HEMA increased the
bond strength significantly. Higher concentrations of HEMA
(19 and 36%) did not improve the bond strength. Moreover,
36% of HEMA even reduced the bond strength compared to
Exp-0, however not significantly.
Fracture analysis by light microscopy revealed a predom-
inantly mixed adhesive failure pattern on both enamel and
dentin. In particular Exp-36 on dentin mainly failed at the
adhesive–composite interface.
Electron microscopy revealed the presence of droplets in all
adhesives tested on both enamel and dentin, except for Exp-36
that did not exhibit droplets on enamel (Figs. 4 and 5, Table 2).
In the adhesive layers of Exp-0 and Exp-10, the few droplets
observed were located mainly throughout the adhesive layer.
The size of these droplets varied between 1 and 10 !m. On
 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419
Table 2 – Observations of droplets in the adhesive layer of the experimental adhesives on a glass plate by light
microscopy (not cured) and by SEM during the failure analysis (cured)
Phase separation in SEM interfacial analysis: presence and location of droplets
uncured adhesive at the interface (SEM)
drop (LM)
Enamel
Exp-0 Yes Some droplets (throughout adhesive layer)
Exp-10 Yes Some droplets (throughout adhesive layer)
Exp-19 Yes Very few droplets (throughout adhesive layer)
Exp-36 No No droplets
Fig. 2 – Box-whisker plot (min-[lower
quartile–median-upper quartile]-max) of the !TBS to
enamel. On the right, mean (standard deviation); n, total
number of tested specimens; ptf, pre-testing failure. The
diamonds represent the mean !TBS and the vertical line in
the box is the median. Means with unlike superscript
letters are statistically significantly different.
Fig. 3 – Box-whisker plot (min-[lower
quartile–median-upper quartile]-max) of the !TBS to
dentin. On the right, mean (standard deviation); n, total
number of tested specimens; ptf, pre-testing failure. The
diamonds represent the mean !TBS and the vertical line in
the box is the median. Means with unlike superscript
letters are statistically significantly different.
1415
Origin of
droplets [26]
Dentin
Some droplets (throughout adhesive layer) Mainly phase
separation
Some droplets (throughout adhesive layer) Mainly phase
separation
Many droplets (very few in the adhesive Phase
layer, mainly at the interface between separation
adhesive layer and composite) and osmosis
Many droplets (at the interface between Osmosis
adhesive layer and composite)
enamel, Exp-19 exhibited only very few droplets which were
similar to those in Exp-0 and Exp-10. On dentin, however,
numerous smaller (0.3–1.5 !m) droplets could be found located
at the transition between the adhesive layer and the lining
composite. In Exp-36, no droplets at all could be found on
enamel, in contrast to the numerous droplets on dentin, again
located between the adhesive layer and the composite.
All experimental adhesives produced a similar partially
demineralized hybrid layer with a thickness between 1 and
1.5 !m. In spite of the variable thickness of the adhesive layer,
depending on the smear layer and on the scratches made by
the diamond bur, the adhesive layer in Exp-0 and Exp-10 was
overall wider than in Exp-19 and Exp-36. The average width of
the adhesive layer in Exp-0 and Exp-10 was 10 !m, compared
to 5 !m in Exp-19 and Exp-36. Moreover, a distinct oxygen-
inhibition layer could only clearly be discerned in Exp-0 and
Exp-10.
7. Discussion
The composition of the four experimental adhesives in this
study was identical, except for the concentration of HEMA
and correspondingly the amount of acetone. As a result, the
concentrations of the other ingredients, such as cross-linkers
and functional monomers, initiators, inhibitors and fillers,
was equal for all adhesives tested, hence allowing for an
indisputable evaluation of the effect of HEMA on adhesive
performance [1].
It is a long-known and generally accepted fact that HEMA
pretreatment of demineralized dentin or HEMA in the primer
of etch and rinse adhesives improves the bond strength to
demineralized dentin [4,18,5,19–21]. Regarding bond strength,
the optimal concentration of HEMA in the primer was deter-
mined to be 30–40% [18,20]. Beyond this concentration, the
bond strength decreases [18]. Besides lowering the viscos-
ity of primers (or adhesive resins in one-bottle systems) and
improving the wetting capacity, HEMA also promotes resin
penetration into the demineralized dentin [22].
In contrast to etch and rinse adhesives, self-etch adhesives
do not require a separate etching phase as they contain acidic
monomers that demineralize and infiltrate enamel and dentin
simultaneously and do not need to be rinsed off. Together
1416 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419

Fig. 4 – Transmission-electron microscopy images of the adhesive–dentin interface produced by the four experimental
one-step adhesives. All sections were silver-stained, but the amount of silver-stain varied considerably even within one
section. No remarkable differences between the hybrid layers could be distinguished. All four experimental adhesives
yielded a partially demineralized hybrid layer of 0.5–1.5 !m, in which hydroxyapatite crystals could still be found. Even
though the thickness of the adhesive layer varied considerably in one specimen according to the grooves of the smear layer
(see also Fig. 5), the adhesive layers of Exp-0 and Exp-10 were generally wider than those of Exp-19 and Exp-36. The
average thickness of the adhesive layers of Exp-0 (a) and Exp-10 (b) was approximately 10 !m. Moreover, in these two
adhesives, a well-defined oxygen-inhibition layer of several micrometers could be discerned. The width of the adhesive
layer in Exp-19 (c) and Exp-36 (d) was around 5 !m, and most often, no oxygen-inhibition layer was visible. Abbreviations:
Ar, adhesive resin; C, composite; Dt, dentinal tubule; Hy, hybrid layer; O2 I, oxygen-inhibition layer; Ud, unaffected dentin.

Fig. 5 – Pairs of scanning-electron microscopy images, representing left (a, c, e, and g) failed enamel or dentin !TBS
specimens and right (b, d, f, and g) the respective interface of the experimental adhesive with dentin. These samples were
prepared as for transmission electron microscopy, including sectioning by the ultra-microtome (b). This procedure allowed
good evaluation of the exact location of the droplets with reference of the adhesive layer, which was not possible with the
failed specimens. (a) Shows the presence of phase-separation droplets in Exp-0 on enamel. (b) Shows a TEM specimen after
trimming and ultra-microtomy. Notice that the adhesive layer varies significantly regarding thickness. In this dentin
specimen, no droplets could be found. In the adhesive layer of Exp-10 some phase-separation droplets could be found (c).
Evaluation of the interface indeed showed that these droplets were located within the adhesive layer (d). In contrast to
Exp-0 and Exp-10, many very small droplets could be observed in dentin samples of Exp-19 (e) and Exp-36 (g). Moreover,
these droplets were located near the transition between the adhesive layer and the composite. These droplets were
therefore assigned to an osmosis process, in which water travels from the dentin through the cured adhesive layer.
Abbreviations: Ar, adhesive resin; C, composite; D, dentin; E, epon; Ud, unaffected dentin.
d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419 1417
1418 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419
with the fact that the acidic monomers in self-etch adhesives
may also contribute to the wetting properties of an adhe-
sive, the need for HEMA is reduced in self-etch adhesives. In
view of the allergenic disadvantages of HEMA, some manu-
facturers may want to omit HEMA from the composition of
adhesives.
Nevertheless, the results of this study indicate that a
small amount of HEMA is beneficial to the 24-h bond
strength of one-step self-etch adhesives to both enamel and
dentin. An optimal concentration, yielding maximum bond
strength, could be determined at 10% of HEMA. Of course,
this result should be confirmed by long-term studies, evalu-
ating bond strength not after 24 h, but after longer periods.
Concentrations beyond 10% did not have any supplemen-
tary advantageous effect on the bond strength to enamel
and dentin. Higher concentrations of HEMA even lowered the
bond strength, especially to dentin. The null hypothesis that
HEMA does not influence the adhesive performance of one-
step adhesives must thus be rejected.
Besides influencing the bond strength, HEMA also affected
the stability of the adhesive solutions in this study. Only very
high amounts of HEMA, replacing the majority of acetone,
were able to prevent phase separation. Only Exp-36 was free
of droplets from phase separation. Nevertheless, this adhesive
obtained the lowest bond strength results of all tested adhe-
sives. Moreover, in spite of droplet formation due to phase
separation, many small droplets could be found in this adhe-
sive when applied to dentin.
This study also demonstrated an effect of the amount
of HEMA on the thickness of the adhesive layer. It is
clear that the small molecular weight of HEMA decreases
the viscosity of the dental adhesive, resulting in thinner
adhesive layers. Overly thinned adhesive layers have been
associated with reduced physico-mechanical strength due to
oxygen inhibition of the polymerization, resulting in sub-
optimal curing rates [23,24]. The absence of a clearly distinct
oxygen-inhibition layer in Exp-19 and Exp-36 however does
not indicate that there was no oxygen inhibition in the
HEMA-rich experimental adhesives. This rather indicates
the absence of a good polymerized part of the adhesive
layer.
Recently, it was demonstrated that one-step adhesives
may exhibit droplets from two different origins: those aris-
ing from phase separation [6] and those from water attraction
and osmosis through the cured adhesive layer [25,26]. These
two kinds of droplets can be differentiated by their loca-
tion with reference to the adhesive layer, their size, and
their behavior. Droplets originating from a phase-separation
reaction are located throughout the adhesive layer and are
generally larger than osmosis droplets. The latter are not
only smaller, but can also only be found at the transition
between adhesive layer and lining composite, which corre-
sponds to the oxygen-inhibition layer, a zone with uncured
monomers [23,27]. As the oxygen-inhibition layer represents
a hypertonic area, water from dentin will be attracted by diffu-
sion. The osmosis droplets exhibit a time-dependent behavior
as their number will increase when the lining composite is
delay-cured [25,26]. Nonetheless, it was shown that osmosis
droplets can already be found even when the lining com-
posite is immediately cured. It was also already shown that
HEMA, due to its low molecular weight and its polar prop-
erties, enhances the diffusion process of water through the
adhesive layer. When the adhesive contains a high concentra-
tion of HEMA, the oxygen-inhibition layer will also contain a
proportionally high amount of HEMA monomers, which will
increase the oncotic pressure, thus enhancing the osmosis
process. Even though an osmosis process must certainly also
take place in Exp-0 and Exp-10, it was concluded that the
droplets in these adhesives mainly originated from phase sep-
aration (Table 2). Low concentrations of HEMA (up to 10%)
did not seem to induce a significant osmosis process, at least
not when the composite was immediately cured. However, in
Exp-19, the higher concentration of HEMA (19%) did induce
an osmosis process that resulted in a significant number of
droplets even when the composite was immediately cured
(Table 2).
Hence, the negative effect of high concentrations of HEMA
on the bond strength can be explained by several factors.
First, HEMA exacerbates the osmosis process, which results
in an abundance of droplets on dentin [26]. It was already
shown that these droplets may reduce the bond strength,
especially after delay-curing the composite. The weakening
effect of the osmosis droplets was also confirmed by failure
analysis of the Exp-36 samples on dentin, which were shown
to have failed predominantly between the composite and the
adhesive layer. SEM confirmed that the fracture plane most
frequently involved the osmosis droplets (Fig. 5). Secondly,
increased amounts of HEMA may have led to reduced polymer-
ization due to its hydrophilicity and water attraction, leading
to dilution of the monomers [28,29]. In addition, as higher con-
centrations of HEMA result in a thinner adhesive layer, oxygen
inhibition may inhibit polymerization even more [30]. Low
conversion rates have been associated with declined mechan-
ical strength of adhesives [31]. A third plausible explanation
may be the reduced cross-link density of the resulting poly-
mer in the adhesive layer, as HEMA only has one polymerizable
group [1]. Cross-linked polymers have been shown to exhibit
better physico-mechanical strength than linear polymers
[12,32,33].
The beneficial effect of 10% HEMA must be explained by
the good wetting properties of this monomer [2,4].
Some authors have warned against one-step self-etch
adhesives, which are too hydrophilic in comparison to multi-
step adhesives [34]. This study showed that not all one-step
adhesives are hydrophilic and that depending on their chem-
ical composition (for example depending on the amount
of HEMA), one-step adhesives can be classified as rather
hydrophobic or hydrophilic. Whereas hydrophobic one-step
adhesives are more prone to phase-separation reactions,
hydrophilic one-step self-etch adhesives tend to be more sus-
ceptible to osmosis.
By and large, this study showed that small amounts (10%)
of HEMA improve the bonding efficacy of a one-step self-etch
adhesive by enhancing the wetting behavior of the adhesive.
Yet, this concentration of HEMA was not capable of prevent-
ing phase separation. Paradoxally, high amounts of HEMA that
indeed block a phase-separation reaction and subsequent for-
mation of droplets, lead to even more droplets by osmosis. The
adverse effect of high concentrations of HEMA on the bond
strength must be attributed to a combined effect of the pres-
 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 1412–1419 1419

ence of osmosis droplets, reduced co-polymerization due to [14] Yoshiyama M, Matsuo T, Ebisu S, Pashley D. Regional bond
both water-attraction by HEMA and oxygen inhibition, and last strengths of self-etching/self-priming adhesive systems. J
to the inferior physico-mechanical strength of the resultant Dent 1998;26:609–16.
[15] Perdigao J, Lambrechts P, Van Meerbeek B, Vanherle G, Lopes
poly-HEMA co-polymer. Adding high amounts of HEMA to one-
AL. Field emission SEM comparison of four postfixation
step self-etch adhesive to prevent phase separation is thus not drying techniques for human dentin. J Biomed Mater Res
recommended, as this will not lead to better bond strengths, 1995;29:1111–20.
and will also result in droplets, albeit from a different [16] Van Meerbeek B, Yoshida Y, Lambrechts P, Vanherle G, Duke
origin. ES, Eick JD, et al. A TEM study of two water-based adhesive
systems bonded to dry and wet dentin. J Dent Res
1998;77:50–9.
Acknowledgements [17] Tay FR, Pashley DH, Yoshiyama M. Two modes of
nanoleakage expression in single-step adhesives. J Dent Res
K.L. Van Landuyt has been granted a Ph.D. fellowship of the 2002;81:472–6.
[18] Munksgaard EC, Asmussen E. Bond strength between dentin
Research Foundation—Flanders (FWO). We thank Dr. A. Arita
and restorative resins mediated by mixtures of HEMA and
and GC Japan for preparing the experimental adhesives. glutaraldehyde. J Dent Res 1984;63:1087–9.
[19] Doi J, Itota T, Torii Y, Nakabo S, Yoshiyama M. Effect of
references 2-hydroxyethyl methacrylate pre-treatment on
micro-tensile bond strength of resin composite to
demineralized dentin. J Oral Rehabil 2004;31:1061–7.
[20] Hussain LA, Dickens SH, Bowen RL. Shear bond strength of
[1] Van Landuyt KL, Snauwaert J, De Munck J, Peumans M,
experimental methacrylated beta-cyclodextrin-based
Yoshida Y, Poitevin A, et al. Systematic review of the
formulations. Biomaterials 2005;26:3973–9.
chemical composition of contemporary dental adhesives.
[21] Pilo R, Cardash HS, Oz-Ari B, Ben Amar A. Effect of
Biomaterials 2007;28:3757–85.
preliminary treatment of the dentin surface on the shear
[2] Toledano M, Osorio R, Moreira MA, Cabrerizo-Vilchez MA,
bond strength of resin composite to dentin. Oper Dent
Gea P, Tay FR, et al. Effect of the hydration status of the
2001;26:569–75.
smear layer on the wettability and bond strength of a
[22] Nakabayashi N, Watanabe A, Gendusa NJ. Dentin adhesion
self-etching primer to dentin. Am J Dent 2004;17:310–4.
of “modified” 4-META/MMA-TBB resin: function of HEMA.
[3] Rosales JI, Marshall GW, Marshall SJ, Watanabe LG, Toledano
Dent Mater 1992;8:259–64.
M, Cabrerizo MA, et al. Acid-etching and hydration influence
[23] Biernat M, Rokicki G. Oxygen inhibition of
on dentin roughness and wettability. J Dent Res
photopolymerization processes and methods of its
1999;78:1554–9.
suppression. Polimery 2005;50:633–45.
[4] Nakabayashi N, Takarada K. Effect of HEMA on bonding to
[24] Nunes TG, Ceballos L, Osorio R, Toledano M. Spatially
dentin. Dent Mater 1992;8:125–30.
resolved photopolymerization kinetics and oxygen
[5] Nakaoki Y, Nikaido T, Pereira PN, Inokoshi S, Tagami J.
inhibition in dental adhesives. Biomaterials 2005;26:
Dimensional changes of demineralized dentin treated with
1809–17.
HEMA primers. Dent Mater 2000;16:441–6.
[25] Tay FR, Pashley DH, Suh BI, Carvalho RM, Itthagarun A.
[6] Van Landuyt K, De Munck J, Snauwaert J, Coutinho E,
Single-step adhesives are permeable membranes. J Dent
Poitevin A, Yoshida Y, et al. Monomer–solvent phase
2002;30:371–82.
separation in one-step self-etch adhesives. J Dent Res
[26] Van Landuyt KL, Snauwaert J, De Munck J, Coutinho E,
2005;84:183–8.
Poitevin A, Yoshida Y, et al. Origin of interfacial droplets
[7] Toledano M, Osorio R, de Leonardi G, Rosales-Leal JI,
with one-step adhesives. J Dent Res 2007;86:739–44.
Ceballos L, Cabrerizo-Vilchez MA. Influence of self-etching
[27] Gauthier MA, Stangel I, Ellis TH, Zhu XX. Oxygen inhibition
primer on the resin adhesion to enamel and dentin. Am J
in dental resins. J Dent Res 2005;84:725–9.
Dent 2001;14:205–10.
[28] Jacobsen T, Soderholm KJ. Some effects of water on dentin
[8] Moszner N, Salz U, Zimmermann J. Chemical aspects of
bonding. Dent Mater 1995;11:132–6.
self-etching enamel-dentin adhesives: a systematic review.
[29] Wang Y, Spencer P. Continuing etching of an all-in-one
Dent Mater 2005.
adhesive in wet dentin tubules. J Dent Res 2005;84:350–4.
[9] Sandberg E, Bergenholtz G, Eklund C, Dahlgren UI. HEMA
[30] Zheng L, Pereira PN, Nakajima M, Sano H, Tagami J.
bound to self-protein promotes auto-antibody production in
Relationship between adhesive thickness and microtensile
mice. J Dent Res 2002;81:633–6.
bond strength. Oper Dent 2001;26:97–104.
[10] Lonnroth EC, Wellendorf H, Ruyter E. Permeability of
[31] Yamamoto A, Tsubota K, Takamizawa T, Kurokawa H, Rikuta
different types of medical protective gloves to acrylic
A, Ando S, et al. Influence of light intensity on dentin bond
monomers. Eur J Oral Sci 2003;111:440–6.
strength of self-etch systems. J Oral Sci 2006;48:21–6.
[11] Goossens A. Contact allergic reactions on the eyes and
[32] Asmussen E, Peutzfeldt A. Influence of selected components
eyelids. Bull Soc Belge Ophtalmol 2004:11–7.
on crosslink density in polymer structures. Eur J Oral Sci
[12] Tay FR, King NM, Chan KM, Pashley DH. How can
2001;109:282–5.
nanoleakage occur in self-etching adhesive systems that
[33] Odian G. Principles of polymerization. New York: Willey
demineralize and infiltrate simultaneously? J Adhes Dent
Interscience; 2004.
2002;4:255–69.
[34] Tay FR, Pashley DH. Have dentin adhesives become too
[13] Phrukkanon S, Burrow MF, Tyas MJ. The effect of dentine
hydrophilic? J Can Dent Assoc 2003;69:726–31.
location and tubule orientation on the bond strengths
between resin and dentine. J Dent 1999;27:265–74.