Bonding of an auto-adhesive luting material to enamel and dentin

Dental Materials (2004) xx, xxx–xxx
Bonding of an auto-adhesive luting material
to enamel and dentin
Jan De Munck a , Marcos Vargas b , Kirsten Van Landuyt a , Kazuhiro Hikita a,c ,
Paul Lambrechts a , Bart 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, Leuven 3000, Belgium
b Department of Operative Dentistry, University of Iowa, Iowa city, IA, USA
c Institute of Medical Science, Health Sciences University of Hokkaido, Sapporo, Japan
Received 14 October 2003; received in revised form 2 February 2004; accepted 1 March 2004
KEYWORDS
Adhesion; Cement; Selfadhesive;
Micro-tensile
bond strength; Fe-SEM;
TEM
*Corresponding author. Tel.: þ32-16-33-75-87; fax: þ32-16-
33-27-52.
E-mail address: bart.vanmeerbeek@med.kuleuven.ac.be
ARTICLE IN PRESS
www.intl.elsevierhealth.com/journals/dema
Summary Objectives: The objectives of this study were (1) to assess the bonding
performance of a new auto-adhesive cement (RelyX Unicem, 3M ESPE) to enamel and
dentin, using a standard micro-tensile bond strength (mTBS) test set-up, and (2) to
evaluate the interaction of this material with dentin by means of high-resolution
electron microscopy.
Methods: The mTBS of RelyX Unicem with and without prior acid etching was
determined to enamel and dentin after 24 h of water storage and compared to the
bonding effectiveness of the control cement (Panavia-F, Kuraray). In addition,
diamond-knife cut interfaces of RelyX Unicem and Panavia-F bonded to dentin were
examined using field-emission scanning (Fe-SEM) and transmission electron microscopy
(TEM).
Results: The mTBS of RelyX Unicem to enamel was significantly lower than that of
the control cement, whereas no significant difference was found when both cements
were bonded to dentin. Acid etching prior to the application of RelyX Unicem raised
the enamel mTBS to the same level as that of the control, but was detrimental for the
dentin bonding effectiveness. The latter must be attributed to inadequate infiltration
of the collagen mesh as revealed by Fe-SEM and TEM. Morphological evaluation
additionally revealed that RelyX Unicem only superficially interacted with enamel and
dentin, and that application using some pressure is required to ensure close adaptation
of the cement to the cavity wall.
Significance: (1) RelyX Unicem should always be applied with some pressure, to
ensure that the relatively high viscous cement intimately adapts to the cavity wall;
(2) The cement only interacted superficially with dentin and enamel; (3) The best
bonding effectiveness with this new auto-adhesive cement was obtained by selectively
acid-etching enamel prior to luting.
Q 2004 Published by Elsevier Ltd on behalf of Academy of Dental Materials. All rights
reserved.
0109-5641/$ - see front matter Q 2004 Published by Elsevier Ltd on behalf of Academy of Dental Materials. All rights reserved.
doi:10.1016/j.dental.2004.03.002

2
Introduction
Resin-based adhesive luting materials are widely
used for the fixation of inlays and onlays, crowns,
posts and veneers. Currently, all resin cements are
based upon the use of an etch-and-rinse or self-etch
adhesive along with a low-viscosity resin composite.
This multi-step application technique is complex
and rather technique sensitive, and consequently
may compromise bonding effectiveness. 1 Recently,
a new resin-based cement has been marketed that
combines the use of adhesive and cement in one
single application, eliminating the need for pretreatment
of both tooth and restoration. The
adhesive properties are claimed to be based upon
acidic monomers that demineralize and infiltrate
the tooth substrate, resulting in micro-mechanical
retention. Secondary reactions have been suggested
to provide chemical adhesion to hydroxyapatite (3M
ESPE RelyX Unicem product profile) a feature
currently only proven for glass-ionomers. 2,3
The purpose of this study was (1) to assess the
bonding performance of a new auto-adhesive resin
cement (RelyX Unicem, 3M ESPE) to enamel and
dentin, using a standard micro-tensile bond strength
(mTBS) test set-up, and (2) to evaluate the interaction
of this material with dentin by means of highresolution
electron microscopy. The hypotheses
tested were: (1) The use of a simplified application
procedure does not decrease bonding effectiveness
Table 1 List of materials used in this study.
ARTICLE IN PRESS
to both enamel and dentin. (2) The use of phosphoric
acid prior to luting improves the bonding effectiveness.
(3) The bonding mechanism of RelyX Unicem to
dentin is similar to that of a self-etch adhesive.
Material and methods
Eighteen non-carious human third molars (gathered
following informed consent, approved by the
Commission for Medical Ethics of the Catholic
University of Leuven) were stored in 0.5% chloramine
in water at 4 8C and used within 1 month after
extraction. The teeth were randomly divided into
six experimental groups.
Enamel specimen preparation for mTBS
Group Composition Application
UNICEM (3M ESPE, Seefeld, Germany) Powder: glass powder, silica, calcium
hydroxide, pigment, substituted pyrimidine,
peroxy compound, initiator. Liquid:
methacrylated phosphoric ester,
dimethacrylate, acetate, stabilizer,
initiator [001/0001]
H3PO4 þ UNICEM Etchant: 35% phosphoric acid, silica
thickener (3M ESPE). RelyX Unicem see
above
Panavia-F (Kuraray, Osaka, Japan) Primer A: HEMA, 10-MDP, 5-NMSA, water,
accelerator [00108B]. Primer B: 5-NMSA,
accelerator, water, sodium benzene
sulfinate [0015B]. Paste A: 10-MDP,
hydrophobic aromatic dimethacrylate,
hydrophobic aliphatic dimethacrylate,
hydrophilic dimethacrylate, silanated silica,
photoinitiator, dibenzoyl peroxide
[00124A]. Paste B: hydrophobic aromatic
dimethacrylate, hydrophobic aliphatic
dimethacrylate, hydrophilic
dimethacrylate, sodium aromatic sulfinate,
accelerator, sodium fluoride, silanated
barium glass [00046]
J. De Munck et al.
Lingual and/or buccal enamel was flattened using a
high-speed diamond bur (842, Komet, Lemgo,
Germany), mounted in the MicroSpecimen Former
(University of Iowa, Iowa City, IA, USA). Prior to
cementing procedures, pre-cured resin composite
blocks (Paradigm MZ1OO, 3M ESPE, St Paul, MN,
USA) were cut using a low-speed diamond saw
(Isomet 1000, Buehler, Lake Bluff, IL, USA) to obtain
a standard surface roughness. The bonding surface
was prepared according to the manufacturers’
instructions of the respective cement (Table 1).
Next, the resin cement was applied on the enamel
Composite block: no pre-treatment. Tooth:
no pre-treatment. Cement: mix capsule for
15 00 (Rotomix, 3M ESPE); apply on surface;
lute resin block using light pressure; light
cure for 20 00 from each side
Composite block: no pre-treatment. Tooth:
apply etchant for 15 s; rinse; air-dry without
desiccation. Cement: as described above
Composite block: apply K-etchant for 5 00 ;
rinse with water; air dry; apply mixture of
Clearfil SE primer [00125C] with Porcelain
bond activator [00107B] for 5 00 ; gently air
blow; light cure for 10 00 . Tooth: mix ED
primer (A&B); apply undisturbed for 60 00 ;
gently air blow. Cement: mix cement,
(A&B); lute resin block using light pressure;
light cure for 20 00 from each side; apply
oxyguard [00334A] for at least 3 0
HEMA: 2-hydroxyethyl methacrylate; 10-MDP: 10-methacryloyloxydecyl dihydrogen phosphate; 5-NMSA: N-methacryloxyl-
5-aminosalicyclic acid.

surface. The auto-adhesive cement was applied
with and without prior phosphoric acid etching. The
pre-cured composite block was pressed on the
cement using light pressure, after which excess
cement was removed. Light-curing was performed
from four directions parallel along the cement
interface using an Optilux 500 (Demetron/Kerr,
Danbury, CT, USA) device with a light output not
less than 550 mW/cm 2 . At each moment during
specimen processing, care was taken to prevent
dehydration of the specimens.
Dentin specimen preparation for mTBS
The occlusal third of the molars was removed using
the Isomet slow-speed diamond saw. Dentin surfaces
were controlled for absence of enamel and/or
pulp tissue using a stereomicroscope (Wild M5A,
Heerbrugg, Switzerland). The teeth were mounted
in a chuck and a standard smear layer was produced
by removing a thin layer of the surface using a highspeed
medium-grit (100 mm) diamond bur (842,
Komet) mounted in the MicroSpecimen Former. The
cements were subsequently applied following the
methodology described above.
mTBS testing
After 24 h of water storage, the teeth were
sectioned perpendicular to the adhesive–tooth
interface using the Isomet saw to obtain rectangular
sticks of about 1.8 £ 1.8 mm wide and 8–9 mm long.
Specimens were trimmed at the biomaterial–tooth
surface to a cylindrical hourglass shape with a
diameter of about 1.2 mm using the MicroSpecimen
Former and fine cylindrical diamond burs
(B ¼ 1.2 mm, 835KREF, Komet, Lemgo, Germany)
under continuous air/water spray. Next, the crosssectional
diameter was precisely (accuracy ¼ 0.001 mm)
measured using a precision measuring instrument
transformed from a x–y multi-purpose modular
microscope (Leitz, Wetzlar, Germany). Specimens
were then fixed to Ciucchi’s jig (fixation height 4–
6 mm) with cyanoacrylate glue (Model Repair II
Blue, Dentsply-Sankin, Ohtawara, Japan) and
stressed at a crosshead speed of 1 mm/min 4 until
failure in a LRX material testing device (LRX, Lloyd,
Hampshire, UK). The mTBS was expressed in MPa, as
derived from dividing the imposed force (N) at the
time of fracture by the bond area (mm 2 ). When
specimens failed before actual testing, the mTBS
was determined from the specimens that
survived specimen processing with an explicit note
of the number of pre-testing failures. The means of
the different groups were compared using
one-way ANOVA analysis and a Scheffe’s multiple
ARTICLE IN PRESS
Bonding of an auto-adhesive luting material to enamel and dentin 3
comparisons test at a significance level of 0.05. All
statistics were performed using the Statistica software
package (StatSoft, Tulsa, OK, USA). The mode
of failure was determined at a magnification of 50 £
using a stereomicroscope (Wild M5A, Heerbrugg,
Switzerland).
Electron microscopy (Fe-SEM and TEM)
Bonding effectiveness to dentin was also ultramorphologically
assessed by electron microscopy.
Two dentin surfaces were prepared for each group
in the same way as for mTBS testing. RelyX Unicem
was applied using proper pressure during application
by pressing the cement with a glass plate.
The same cement was also applied to two dentin
surfaces that were fractured with a forceps and
consequently were free of smear layer debris. Next,
a thin layer of a low-viscosity resin composite
(Clearfil Protect Liner F, Kuraray) was applied.
Then, the resin-bonded dentin specimens were
cross-sectioned perpendicular to the resin–dentin
interface to 1 mm wide sticks using the lsomet
diamond saw. Half of the specimens were demineralized
and fixed simultaneously in a 10% formaldehyde–formic
acid solution (Gooding and Stewart
Fluid, Prosan, Gent, Belgium) for at least 36 h.
Figure 1 Box plots of the mTBS results. The central box
represents the spreading of the data between the first
and third quartile. The central vertical line in the box
represents the median value. The range of the data is
represented by whiskers extending to the minimum and
maximum value measured. Asterisks indicate significant
differences (p , 0:05; Scheffé multiple comparisons)
within each group (enamel or dentin).

4
Table 2 mTBS results.
UNICEM H3PO4 þ UNICEM PANAVIA-F
Enamel Dentin Enamel Dentin Enamel Dentin
Mean 19.6 B
15.9 a
35.6 A
5.9 b
35.4 A
SD 6.0 3.9 13.2 3 9.5 5.9
n 15 11 16 12 14 10
ptf 1 1 0 0 0 0
SD, standard deviation; n; total number of specimens prepared; ptf, pre-testing failures; means with the same superscript are not
significantly different (p , 0:05; Scheffé multiple comparisons).
Table 3 Failure analysis from light microscopy.
ARTICLE IN PRESS
J. De Munck et al.
Figure 2 Fe-SEM photomicrographs of RelyX Unicem bonded to bur-cut dentin. (a) RelyX Unicem (Un) applied to a flat
surface. At dentin (D), tubules were not obturated by resin tags (hand indicator). Within the RelyX Unicem cement and
especially at the interface with dentin, many porosities can be detected (arrow). A ¼ Air bubble resulting from mixing.
(b) Higher magnification of (a), clearly showing porosities at the interface (arrow). Within the cement, glass particles
(Gp) as well as voids can be noted. (c) Fe-SEM overview of a resin composite inlay (Ci)/RelyX Unicem (Un)/dentin (D)
interface (inlay cemented under pressure). The cement is approximately 80 mm thick, packed with glass particles
ranging in size from 1 to 5 mm and larger. No voids can be observed at the interface with dentin. (d) Higher magnification
of (c), though RelyX Unicem adapted pore-free with dentin, no distinct morphological manifestation of interaction with
unaffected dentin can be observed.
17.5 a
Cohesive in tooth Tooth/adhesive Adhesive Adhesive/resin Cohesive in resin Total
UNICEM Enamel 0 0 12 3 0 15
Dentin 0 3 2 6 0 11
H3PO4 þ UNICEM Enamel 3 2 2 1 8 16
Dentin 0 0 12 0 0 12
PANAVIA-F Enamel 0 2 1 8 3 14
Dentin 0 0 8 0 2 10

Further specimen preparation of both the demineralized
and non-demineralized sections was performed
in accordance with common procedures
used for ultra-structural Transmission Electron
Microscopy (TEM) examination of biological tissues.
5 Then, 70–90 nm thick sections were cut
through the fractured plane using a diamond-knife
(Diatome, Bienne, Switzerland) in an ultramicrotome
(Ultracut UCT, Leica, Vienna, Austria). For
evaluation of collagen, TEM sections were positively
stained with 5% uranyl acetate (UA) for
20 min and saturated lead citrate (LC) for 3 min
prior to TEM examination (Philips CM10, Eindhoven,
The Netherlands). Unstained sections were evaluated
as well. The remaining diamond-knife cut
blocks, from which the TEM sections were prepared,
were gold-sputtered and evaluated by fieldemission
scanning electron microscopy (Fe-SEM,
Philips XL30, Eindhoven, The Netherlands). To
evaluate the importance of applying the cement
under pressure, also two Class I cavities were
prepared with the gingival floor ending into midcoronal
dentin, after which resin composite inlays
were cemented using RelyX Unicem. Consequently,
in this group the cement was applied using
ARTICLE IN PRESS
Bonding of an auto-adhesive luting material to enamel and dentin 5
a clinically relevant pressure, opposed to the first
groups that were applied to a flat surface using a
glass plate, which results in very light pressure.
Results
mTBS testing
The mean mTBS values are presented per experimental
group in Fig. 1 and Table 2. The values of
the control cement (Panavia-F) were always among
the highest values obtained (Table 2), this when
bonded to enamel as well as to dentin. The mTBS of
RelyX Unicem to enamel was significantly lower
than that of the control cement, whereas no
significant difference was found when both cements
were bonded to dentin (Table 2). The application of
RelyX Unicem to acid-etched enamel significantly
increased the mTBS to a level that was not
significantly different from that of Panavia-F
bonded to (non-etched) enamel. However, the
mTBS of RelyX Unicem to acid-etched dentin was
significantly lower than when dentin was not
acid etched, and significantly lower than when
Figure 3 TEM photomicrographs of RelyX Unicem bonded to bur-cut dentin. (a) Unstained non-demineralized section.
No intermediary zone of interaction between RelyX Unicem and dentin (D) can be detected. Smear plugs (Sp) still
occlude the dentinal tubules. Pd: Peri-tubular dentin, Gp: glass particle. (b) Stained non-demineralized section. An
irregular interaction layer (I), ranging from nearly 0 to 1.5 mm is present in between cement and unaffected dentin.
(c) High magnification of an unstained non-demineralized section at a site of limited interaction. The interaction zone (I)
is only a few hundred nanometers thick. Hand pointer ¼ nanofiller. (d) Stained demineralized section. Under the thick
irregular interaction zone (I), a deeper zone (Z) showed some heavy-metal stained strings.

6
Panavia-F was bonded to dentin. Failure analysis of
the mTBS fracture surfaces in general corroborated
these results (Table 3), as nearly all non-etched
enamel specimens failed adhesively, in contrast
to mostly ‘mixed’ and ‘cohesive’ failures in the
acid-etched enamel group. Noteworthy for the
dentin specimens is that after phosphoric acid
etching, all specimens failed solely ‘adhesively’
between RelyX Unicem and dentin, whereas the
application of RelyX Unicem without prior acid
etching on the other hand resulted in mostly
‘mixed’ failures.
Morphology
Fe-SEM analysis revealed that the adaptation of
RelyX Unicem to the dentin surface substantially
improved when it was used to lute a composite inlay
in a Class-I cavity. This indicates that some pressure
during luting is necessary to prevent voids at the
interface (Fig. 2), which otherwise must be detrimental
to the bonding longevity. Though with some
pressure, an intimate adaptation to the cavity walls
could be achieved, RelyX Unicem only interacted
superficially with the underlying dentin (Fig. 2). No
ARTICLE IN PRESS
J. De Munck et al.
hybrid layer nor resin tags were observed using SEM.
The cement itself is heavily packed with glass
particles, ranging in diameter from less than 1 up to
5 mm and more.
TEM analysis revealed that no real hybrid layer
was formed. Nevertheless, an irregular interaction
zone was disclosed ranging from nearly 0 up to 2 mm
(Fig. 3). Below this interaction zone, cement
components appeared to have infiltrated deeper,
as disclosed by a deeper heavy-metal stained zone
(Fig. 3d). To exclude any smear layer effects and to
elucidate the actual interaction with unaffected
dentin, the cement was applied to fractured
dentin. Hardly any interaction was detected
(Fig. 4), apart from a very thin, more densely
stained layer at the dentin–cement interface. As
fractured dentin tubules are not occluded by smear
plugs, the cement also infiltrated into the tubules.
Within the tubules, RelyX Unicem interacted with
the tubule wall in a similar way as at the
intertubular dentin surface. The body of the resin
tag seemed, however, to have picked up more
staining.
When RelyX Unicem was applied to phosphoric
acid-etched dentin, the weak link in the bonding
Figure 4 TEM photomicrographs of RelyX Unicem bonded to fractured dentin. (a) Unstained non-demineralized
section. Despite the high magnification, no interaction zone was observed. D, Unaffected dentin; Gp, Glass particle;
Rt, Resin tag; Hand pointer, nanofiller. (b) Stained non-demineralized section. At the cement–dentin interface, a thin
electron dense line appeared, which had strongly reacted with the metal stain (arrows). (c) Stained non-demineralized
section. The same interaction line as in (c) can be detected (arrows). Also along the tubule wall, no smear plug
prevented the cement from forming a resin tag. (d) Higher magnification of (c). The resin matrix (Rm) seems to have
infiltrated dentin with an only limited demineralization effect.

complex was clearly located at the level of the
exposed collagen matrix, as shown by failure analysis
(Table 3) and Fe-SEM evaluation of the mTBS fracture
surfaces (Fig. 5a,b). This was substantiated by TEM
evaluation, which revealed the presence of a noninfiltrated
demineralized collagen mesh in between
cement and unaffected dentin (Fig. 5b,c).
Opposite to RelyX Unicem, the application of
Panavia-F resulted in the formation of a ‘true’
hybrid layer (Fig. 6). Demineralization of the
surface was however incomplete, so that much
hydroxyapatite remained within the hybrid layer
(Fig. 6b). On top of this thin (0.5–1 mm) hybrid
layer, irregular, amorphous, hybridized remnants
of the smear layer can be observed (Fig. 6c).
Discussion
In this study, the bonding effectiveness of a new
cement with a simplified application procedure was
compared to that of a conventional composite
cement. The first hypothesis that the use of
ARTICLE IN PRESS
Bonding of an auto-adhesive luting material to enamel and dentin 7
Figure 5 RelyX Unicem bonded to phosphoric acid-etched dentin. (a) Fe-SEM photomicrograph of the composite side
of a fractured mTBS specimen, viewed at an angle of 608. The specimen failed 100% adhesively (A) between RelyX
Unicem (Un) and dentin. C, Composite; Arrows, Air bubbles. (b) Higher magnification of (a). On top of RelyX Unicem,
remnants of the collapsed collagen mesh can be detected (hand pointer). (c) Non-demineralized stained TEM section.
Phosphoric acid etching removed the smear layer and completely demineralized dentin to a depth of about 3 mm,
exposing a collagen mesh (Cm). RelyX Unicem infiltrated into the tubules to create particle-filled tags (Rt), but hardly
infiltrated the collagen mesh (arrow). Consequently, most of the collagen fibrils remained unprotected. D, Dentin; Gp,
Glass particle. (d) Higher magnification of (c). Apart from some superficial infiltration (arrow), the collagen mesh (Cm)
was not hybridized and appears amorphous.
a simplified application procedure does not
decrease bond strengths must be rejected, as the
mTBS of the experimental cement to enamel was
significantly lower than that of the control cement
Panavia-F. Nevertheless, the mTBS of RelyX Unicem
to dentin was in the same range of that of Panavia-F
to dentin. The mTBS values obtained in this study
for the control cement were comparable to those
obtained by Mak et al. 1 In that study, however,
Panavia-F underscored the other cements investigated.
The authors ascribe this to the inhibition of
polymerization of the cement by acidic monomers.
This effect must have been negligible in this study
as the cement was light cured directly upon luting.
Although the pH of the mixed material is very low
(,2 during the first minute, data supplied by 3M
ESPE), nearly no demineralization of the dentin
surface was noticed (Figs. 2–4). This may be due to
the relatively high viscosity of the material and the
limited penetration/interaction time (the material
was light cured directly after application). These
factors along with the thixotropic properties of this
cement can also explain the improved adaptation of

8
Figure 6 (a) Panavia-F bonded to bur-cut dentin. TEM photomicrograph of an unstained, non-demineralized section.
The cement is packed with glass particles (Gp). In between the cement and unaffected dentin (D), a true hybrid layer (H)
is formed. The adhesive was, however, not able to dissolve the smear plug (Sp) in the dentinal tubule. (b) High
magnification of the hybrid layer in (a). Hydroxyapatite crystals can be observed throughout the full depth of the hybrid
layer, suggesting partial demineralization induced by the self-etch adhesive. (c) TEM photomicrograph of a stained,
demineralized section of Panavia-F bonded to bur-cut dentin. Dentin is completely deprived from its mineral
component. Individual collagen fibrils can be observed in the hybrid layer (H). On top of the hybrid layer, a more
amorphous, stained layer (hand pointer) appeared, probably including hybridized remnants of the irregular, diamond
bur prepared smear layer. (d) TEM photomicrograph of a stained, demineralized section of Panavia-F applied to
fractured dentin. The unaffected dentin was demineralized and consecutively infiltrated with resin for about 500 nm,
creating a hybrid layer (H). Because the surface is free of smear, the resin could freely infiltrate the tubules to create
resin tags (Rt). D, Unaffected dentin, Gp, Glass particle.
the cement when applied under pressure, as
evidenced by a reduced amount of porosities at
the interface (Fig. 2). This is clinically very
relevant, as most indirect restorations are luted
with pressure, as for example crowns, inlays and
onlays.
TEM analysis revealed a very irregular interaction
zone in between the cement and dentin. This zone
probably corresponds to the rough and irregular
smear layer, as produced by the regular-grit
diamond bur. 6 This smear layer was partially
demineralized and subsequently infiltrated by the
resin cement. Below this infiltrated smear layer, a
vague deeper infiltration was detected when the
specimens were stained (Fig. 3). This probably
corresponds to some phosphoric-ester monomer
infiltration, as this monomer was previously shown
to be highly stainable. 7,8 The low demineralization
effect of RelyX Unicem, despite its low initial pH,
was confirmed by applying it to fractured dentin.
Even though the interaction was not hampered due
ARTICLE IN PRESS
J. De Munck et al.
to the lack of an intermediary smear layer, dentin
did not appear to be demineralized at all in the
unstained sections (Fig. 4a). Staining with heavy
metals on the other hand disclosed some limited
interaction within a range of 100–200 nm (Fig. 4c,
d). This probably corresponds to the deeper
infiltration of phosphoric-acid monomers below
the smear layer, as seen in Fig. 3d. Consequently,
the third hypothesis that the bonding mechanism of
RelyX Unicem to dentin is similar to that obtained
with a self-etch adhesive, must be rejected, as no
distinct demineralization and hybridization was
observed, as commonly seen with self-etch
materials (Fig. 6).
Acid etching prior to the application of RelyX
Unicem raised the enamel mTBS to the same level as
that of the control (Table 2), thus corroborating the
second hypothesis that use of phosphoric acid prior
to luting improves the bonding effectiveness to
enamel. This hypothesis must, however, be
rejected for the dentin group, as the mTBS was

significantly lower than when applied without
phosphoric acid. The phosphoric acid thus
decreased the dentin bonding effectiveness, whilst
an increased bonding effectiveness was expected,
because the ‘weak’ smear layer 9 was removed by
phosphoric acid. 10 Apparently, the thick and compact
collagen mesh prevented the viscous cement
from reaching the deeper unaffected dentin
(Fig. 5c). Consequently, a weak layer of hydroxyapatite-depleted
collagen remained in between
RelyX Unicem and unaffected dentin. We concluded
that this poorly infiltrated collagen mesh is the
weak link in the whole adhesive complex because
(1) a 100% adhesive failure rate was observed for
RelyX Unicem bonded to acid-etched dentin specimens
(Table 3), (2) significant lower bond strengths
were recorded, despite the formation of particlereinforced
resin tags and complete smear layer
removal (Table 2), and because (3) remnants of the
poorly infiltrated collagen mesh remained attached
to the resin cement after mTBS testing (Fig. 5b).
Such a demineralized, non resin-reinforced, collagen
layer is known to be prone to degradation
processes 11,12 and will also enhance other degradation
processes as resin elution and nanoleakage.
13 – 15 Hence, a phosphoric acid pre-treatment
should be limited to enamel only. It might however
be worthwhile to investigate if other smear layer
removing procedures that do not demineralize
dentin, as for example air abrasion, may improve
the mTBS of this cement to dentin.
Conclusion
(1) RelyX Unicem should always be applied with
some pressure to ensure that the cement intimately
adapts to the cavity wall. (2) The cement only
interacted superficially with dentin and enamel.
(3) The best bonding effectiveness with this new
auto-adhesive cement was obtained by selectively
acid-etching enamel prior to luting.
Acknowledgements
This study was supported in part by a Research
Grant of the Fund for Scientific Research Flanders
(F.W.O.-grant ‘Krediet aan Navorsers’ 1.5.142.02),
in part by 3M ESPE (Seefeld, Germany) and in part by a
fund of the Toshio Nakao Chair for Adhesive Dentistry
inaugurated at the Catholic University of Leuven
ARTICLE IN PRESS
Bonding of an auto-adhesive luting material to enamel and dentin 9
with B. Van Meerbeek and P. Lambrechts awarded as
chairholders. We thank the respective manufacturers
for the generous donation of materials.
References
1. Mak YF, Lai SC, Cheung GS, Chan AW, Tay FR, Pashley DH.
Micro-tensile bond testing of resin cements to dentin and an
indirect resin composite. Dent Mater 2002;18:609—21.
2. Wilson AD, Prosser HJ, Powis DM. Mechanism of adhesion of
polyelectrolyte cements to hydroxyapatite. J Dent Res 1983;
62:590—2.
3. Yoshida Y, Van Meerbeek B, Nakayama Y, Snauwaert J,
Hellemans L, Lambrechts P, Vanherle G, Wakasa K. Evidence
of chemical bonding at biomaterial—hard tissue interfaces.
J Dent Res 2000;79:709—14.
4. Pashley DH, Carvalho RM, Sano H, Nakajima M, Shono Y,
Fernandes CA, Tay F. The microtensile bond test: a review.
J Adhes Dent 1999;1:299—309.
5. Van Meerbeek B, Conn Jr LJ, Duke ES, Eick JD, Robinson SJ,
Guerrero D. Correlative transmission electron microscopy
examination of nondemineralized and demineralized resin—
dentin interfaces formed by two dentin adhesive systems.
J Dent Res 1996;75:879—88.
6. Tay FR, Sano H, Carvalho R, Pashley EL, Pashley DH. An
ultrastructural study of the influence of acidity of selfetching
primers and smear layer thickness on bonding to
intact dentin. J Adhes Dent 2000;2:83—98.
7. Van Meerbeek B, Vargas M, Inoue S, Yoshida Y, Perdigão J,
Lambrechts P, Vanherle G. Microscopy investigations. Techniques,
results, limitations. Am J Dent 2000;13:3D—18D.
8. De Munck J, Van Meerbeek B, Vargas M, Iracki J, Van Landuyt
K, Poitevin A, Lambrechts P. One day bonding effectiveness
of new self-etch adhesives to bur-cut enamel and dentin.
Oper Dent 2004;29: in press.
9. Koibuchi H, Yasuda N, Nakabayashi N. Bonding to dentin with
a self-etching primer: the effect of smear layers. Dent Mater
2001;17:122—6.
10. Perdigao J, Lambrechts P, Van Meerbeek B, Tomé AR,
Vanherle G, Lopes AB. Morphological field emission-SEM
study of the effect of six phosphoric acid etching agents on
human dentin. Dent Mater 1996;12:262—71.
11. Santerre JP, Shajii L, Leung BW. Relation of dental
composite formulations to their degradation and the release
of hydrolyzed polymeric-resin-derived products. Crit Rev
Oral Biol Med 2001;12:136—51.
12. Hashimoto M, Tay FR, Ohno H, Sano H, Kaga M, Yiu C,
Kumagai H, Kudou Y, Kubota M, Oguchi H. SEM and TEM
analysis of water degradation of human dentinal collagen.
J Biomed Mater Res Part B 2003;66B:287—98.
13. Sano H, Shono T, Takatsu T, Hosoda H. Microporous dentin
zone beneath resin-impregnated layer. Oper Dent 1994;19:
59—64.
14. Hashimoto M, Ohno H, Kaga M, Endo K, Sano H, Oguchi H. In
vivo degradation of resin-dentin bonds over 1 to 3 years.
J Dent Res 2000;79:1385—91.
15. De Munck J, Van Meerbeek B, Lambrechts P, Vanherle G.
Four-year water degradation of total-etch adhesives bonded
to dentin. J Dent Res 2003;82:136—40.