Micro-tensile bond strength of adhesives bonded to class-I cavity ...
Micro-tensile bond strength of adhesives bonded to class-I cavity ...
Micro-tensile bond strength of adhesives bonded to class-I cavity ...
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
Dental Materials (2005) 21, 999–1007<br />
<strong>Micro</strong>-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong> <strong>of</strong> <strong>adhesives</strong> <strong>bond</strong>ed <strong>to</strong><br />
<strong>class</strong>-I <strong>cavity</strong>-bot<strong>to</strong>m dentin after thermo-cycling<br />
Jan De Munck, Kirsten Van Landuyt, Eduardo Coutinho, André Poitevin,<br />
Marleen Peumans, Paul Lambrechts, Bart Van Meerbeek*<br />
Leuven BIOMAT Research Cluster, Department <strong>of</strong> Conservative Dentistry, School <strong>of</strong> Dentistry, Oral<br />
Pathology and Maxillo-Facial Surgery, Catholic University <strong>of</strong> Leuven, Kapucijnenvoer 7, 3000 Leuven,<br />
Belgium<br />
Received 8 June 2004; received in revised form 6 Oc<strong>to</strong>ber 2004; accepted 19 November 2004<br />
KEYWORDS<br />
Thermo-cycling;<br />
Adhesion;<br />
Dental adhesive;<br />
Enamel;<br />
Dentin;<br />
Bond <strong>strength</strong><br />
www.intl.elsevierhealth.com/journals/dema<br />
Summary A widely used artificial aging methodology is thermo-cycling. The ISO<br />
TR 11450 standard (1994) recommends 500 cycles in water between 5 and 55 8C.<br />
Recent literature revealed that more cycles are needed <strong>to</strong> mimic long-term <strong>bond</strong>ing<br />
effectiveness. Furthermore, the artificial aging effect induced by thermo-cycling is<br />
not clearly established. Two underlying mechanisms can be advanced: (1) hot water<br />
may accelerate hydrolysis and elution <strong>of</strong> interface components and (2) repetitive<br />
contraction/expansion stress can be generated.<br />
Objectives: The purpose <strong>of</strong> this study was <strong>to</strong> evaluate the relative contribution <strong>of</strong> both<br />
chemical (hydrolysis and elution <strong>of</strong> interface components) and mechanical (repetitive<br />
contraction/expansion stress) degradation pathways on the thermo-cycling-induced<br />
artificial aging <strong>of</strong> dentin–adhesive interfaces at the bot<strong>to</strong>m <strong>of</strong> <strong>class</strong>-I cavities.<br />
Methods: The micro-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong> (mTBS) <strong>of</strong> contemporary <strong>adhesives</strong><br />
(a three-step etch and rinse, a two-step and a one-step self-etch adhesive) <strong>bond</strong>ed<br />
<strong>to</strong> <strong>class</strong>-I <strong>cavity</strong>-bot<strong>to</strong>m dentin was determined after 20,000 cycles as well as after 20<br />
days <strong>of</strong> water s<strong>to</strong>rage (control). Res<strong>to</strong>red <strong>class</strong>-I cavities (repetitive contraction/expansion<br />
stress) as well as prepared micro-specimens (diffusion-dependent<br />
hydrolysis and elution) were subjected <strong>to</strong> the thermo-cycling regimen.<br />
Results: Thermo-cycling did not enhance chemical or mechanical degradation <strong>of</strong> the<br />
<strong>bond</strong>s produced by a two-step self-etch and a three-step etch and rinse adhesive <strong>to</strong><br />
dentin. The one-step self-etch adhesive tested was, however, not able <strong>to</strong> withstand<br />
polymerization shrinkage stress, nor thermo-cycling, when applied in <strong>class</strong>-I cavities.<br />
Significance: Thermo-cycling results in combined contraction/expansion stress and<br />
accelerated chemical degradation. However, the relative contribution <strong>of</strong> each is<br />
strongly dependent on the specific test set-up and the adhesive used.<br />
Q 2005 Academy <strong>of</strong> Dental Materials. Published by Elsevier Ltd. All rights reserved.<br />
* Corresponding author. Tel.: C32 16 33 75 87; fax: C32 16 33 27 52.<br />
E-mail address: bart.vanmeerbeek@med.kuleuven.ac.be (B. Van Meerbeek).<br />
0109-5641/$ - see front matter Q 2005 Academy <strong>of</strong> Dental Materials. Published by Elsevier Ltd. All rights reserved.<br />
doi:10.1016/j.dental.2004.11.005
1000<br />
Introduction<br />
Thermo-cycling is a widely used artificial aging<br />
methodology. The ISO TR 11450 standard [1]<br />
indicates that a thermo-cycling regimen comprising<br />
500 cycles in water between 5 and 55 8C is an<br />
appropriate artificial aging test. A recent literature<br />
review [2] concluded that 10,000 cycles corresponds<br />
approximately <strong>to</strong> 1 year <strong>of</strong> in vivo functioning,<br />
rendering 500 cycles, as proposed by the ISO<br />
standard, very minimal <strong>to</strong> mimic long-term <strong>bond</strong>ing<br />
effectiveness.<br />
The artificial aging effect induced by thermocycling<br />
can be two-fold: (1) hot water may accelerate<br />
hydrolysis <strong>of</strong> non-protected collagen and<br />
extract poorly polymerized resin oligomers [3–5];<br />
(2) due <strong>to</strong> the higher thermal contraction/expansion<br />
coefficient <strong>of</strong> the res<strong>to</strong>rative material (as<br />
compared <strong>to</strong> that <strong>of</strong> <strong>to</strong>oth tissue) repetitive<br />
contraction/expansion stresses are generated at<br />
the <strong>to</strong>oth–biomaterial interface. This may result in<br />
cracks that propagate along <strong>bond</strong>ed interfaces, and<br />
once a gap is created, changing gap dimensions can<br />
cause in- and outflow <strong>of</strong> pathogenic fluids, a process<br />
known as ‘percolation’ [2,6]. In the light <strong>of</strong> the first<br />
aging effect (diffusion-dependent hydrolysis and<br />
elution), thermo-cycling should be applied <strong>to</strong><br />
micro-specimens, <strong>of</strong> which the interface is directly<br />
exposed <strong>to</strong> the changing temperature environment.<br />
Then, degradation <strong>of</strong> the adhesive–<strong>to</strong>oth interface<br />
is most severe, clinically corresponding <strong>to</strong> the most<br />
vulnerable res<strong>to</strong>ration margins. Aging <strong>of</strong> mTBSspecimens<br />
might also be more appropriate than<br />
aging <strong>of</strong> the larger ‘shear-<strong>bond</strong>’-<strong>strength</strong> specimens,<br />
because the surrounding composite and<br />
<strong>to</strong>oth tissue may thermally protect the adhesive–<br />
<strong>to</strong>oth interface. If the second aging effect (repetitive<br />
contraction/expansion stress) is focused upon,<br />
thermo-cycling should be applied <strong>to</strong> specimens in<br />
which stress similar <strong>to</strong> that occurring clinically can<br />
be generated. In vivo stress will be generated if the<br />
ratio <strong>of</strong> <strong>bond</strong>ed <strong>to</strong> un<strong>bond</strong>ed surfaces or the<br />
‘C-fac<strong>to</strong>r’ is high [7]. Clinically, about the highest<br />
C-fac<strong>to</strong>r is generated in narrow occlusal <strong>class</strong>-I<br />
cavities. In this case, thermo-cycling <strong>of</strong> the whole<br />
<strong>to</strong>oth including the high C-fac<strong>to</strong>r res<strong>to</strong>ration will<br />
result in the highest possible stress imposed at the<br />
interface.<br />
It is, however, not clear whether or not thermocycling<br />
affects the <strong>bond</strong> <strong>strength</strong> <strong>to</strong> dentin. A<br />
recent meta-analysis [8], concerning data published<br />
between 1992 and 1996, concluded that<br />
thermo-cycling has no significant effect on <strong>bond</strong><br />
<strong>strength</strong>. Most studies included in the meta-analysis<br />
were carried out following the ISO standard <strong>of</strong> 500<br />
cycles (mean number <strong>of</strong> cycles in the studies<br />
analyzed was 630). This number <strong>of</strong> cycles was<br />
probably <strong>to</strong>o low <strong>to</strong> obtain an aging effect [2,8,9].<br />
Also specimen geometry has <strong>of</strong>ten not been taken<br />
in<strong>to</strong> account. In most studies <strong>of</strong> this review,<br />
relatively large composite cylinders <strong>bond</strong>ed <strong>to</strong> flat<br />
surfaces were thermo-cycled, prior <strong>to</strong> being pulled<br />
apart following a shear or <strong>tensile</strong> <strong>bond</strong> <strong>strength</strong><br />
test pro<strong>to</strong>col [8]. As a result, a large part <strong>of</strong> the<br />
interface must have been thermally protected<br />
by surrounding dentin [10] and composite [11]<br />
(which are known <strong>to</strong> be good thermal insula<strong>to</strong>rs).<br />
Because <strong>of</strong> the low C-fac<strong>to</strong>r <strong>of</strong> a flat res<strong>to</strong>red<br />
surface (about 1/6), little repetitive expansion/<br />
contraction stress must have been generated at the<br />
interface. Both reasons might explain why thermocycling<br />
did not affect <strong>bond</strong>ing effectiveness in those<br />
studies.<br />
Therefore, the purpose <strong>of</strong> this study was <strong>to</strong><br />
evaluate the relative contribution <strong>of</strong> diffusiondependent<br />
chemical degradation and repetitive<br />
contraction/expansion stress on the thermocycling-induced<br />
degradation <strong>of</strong> dentin–adhesive<br />
interfaces at the bot<strong>to</strong>m <strong>of</strong> occlusal <strong>class</strong>-I cavities.<br />
The hypothesis tested was that thermo-cycling <strong>of</strong><br />
res<strong>to</strong>red <strong>class</strong>-I cavities (repetitive contraction/expansion<br />
stress), as well as <strong>of</strong> mTBS-sticks (diffusiondependent<br />
hydrolysis and elution) does not<br />
decrease <strong>bond</strong>ing effectiveness.<br />
Materials and methods<br />
Specimen preparation<br />
J. De Munck et al.<br />
For this study, non-carious human third molars<br />
(gathered following informed consent approved by<br />
the Commission for Medical Ethics <strong>of</strong> the Catholic<br />
University <strong>of</strong> Leuven), s<strong>to</strong>red in 0.5% chloramine<br />
solution at 4 8C were used within 1 month after<br />
extraction. First, all teeth were mounted in gypsum<br />
blocks in order <strong>to</strong> ease manipulation. A standard<br />
box-type <strong>class</strong>-I <strong>cavity</strong> (4.5!4.5 mm) was then<br />
prepared at the occlusal crown center with the<br />
pulpal floor ending at mid-coronal dentin, using a<br />
high-speed hand piece with a cylindrical mediumgrit<br />
(100 mm) diamond bur (842; Komet, Lemgo,<br />
Germany) mounted in a <strong>Micro</strong>Specimen Former<br />
(University <strong>of</strong> Iowa, Iowa City, IA, USA). Next, the<br />
cavities were subjected <strong>to</strong> a <strong>bond</strong>ing treatment<br />
(Table 1) using either a three-step etch and rinse<br />
adhesive (OptiBond FL), a two-step self-etch<br />
adhesive (Clearfil Protect Bond) or a one-step<br />
self-etch adhesive (iBOND). Subsequently, the<br />
cavities were filled in three horizontal layers with<br />
a resin-composite (Z100, 3M ESPE, St Paul, MN,
<strong>Micro</strong>-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong> <strong>of</strong> <strong>adhesives</strong> <strong>bond</strong>ed <strong>to</strong> <strong>class</strong>-I <strong>cavity</strong>-bot<strong>to</strong>m dentin after thermo-cycling 1001<br />
Table 1 Adhesives used.<br />
Adhesive Composition Application<br />
OptiBond FL Etchant: 37.5% phosphoric acid,<br />
Apply the etchant for 15 s; rinse for 15 s; gently<br />
(Kerr, Orange, silica thickener [301194]<br />
air dry for 5 s; scrub the surface for 15 s with<br />
CA, USA) Primer: HEMA, GPDM, PAMM, ethanol, water, primer; apply a thin coat <strong>of</strong> <strong>bond</strong>ing agent and<br />
pho<strong>to</strong>initia<strong>to</strong>r [212652]<br />
Bond: TEGDMA, UDMA, GPDM, HEMA, bis-GMA,<br />
filler, pho<strong>to</strong>initia<strong>to</strong>r [301335]<br />
light cure for 30 s<br />
Protect Bond Primer: MDP, MDPB, HEMA, initia<strong>to</strong>r, water Apply the primer for 20 s using a rubbing<br />
(Kuraray, [ABB-002]<br />
motion; gently air dry; apply the <strong>bond</strong>ing<br />
Osaka, Japan) Bond: MDP, HEMA, dimethacrylates, colloidal<br />
SiO2, surface treated NaF, initia<strong>to</strong>r [ABP-001]<br />
agent; light cure for 10 s<br />
iBOND (Heraeus Adhesive: UDMA, 4-MET, gluteraldehyde, Apply in three consecutive times and rub for<br />
Kulzer, Hanau, ace<strong>to</strong>ne, water, stabilizer, pho<strong>to</strong>initia<strong>to</strong>r 30 s; gentle air dry until adhesive moves no<br />
Germany) [010028]<br />
more; thoroughly air dry for 5 s; light cure<br />
for 20 s<br />
Bis-GMA, bisphenol-glycidyl methacrylate; GPDM, glycerol phosphate dimethacrylate; HEMA, 2-hydroxyethylmethacrylate; MDP,<br />
10-methacryloyloxydecyl dihydrogen phosphate; MDPB, 12-methacryloyloxydodecylpyridinium bromide; PAMM, phthalic acid<br />
monoethyl methacrylate; TEGDMA, triethylene glycol dimethacrylate; UDMA, urethane dimethacrylate; 4-MET, 4-methacryloyloxyethyl<br />
trimellitic acid.<br />
USA). Light-curing was performed using a highpower<br />
LED curing device (L.E.Demetron 1, Demetron/Kerr,<br />
Danbury, CT, USA).<br />
Then, the res<strong>to</strong>red cavities were sectioned<br />
perpendicular <strong>to</strong> the adhesive–<strong>to</strong>oth interface<br />
using an Isomet diamond saw (Isomet 1000,<br />
Buehler Ltd, Lake Bluff, IL, USA) <strong>to</strong> obtain<br />
rectangular sticks (1.8!1.8 mm wide; 8–9 mm<br />
long). Out <strong>of</strong> each <strong>to</strong>oth, four sticks were<br />
sectioned from the central <strong>cavity</strong> floor (Fig. 1).<br />
They were mounted in the pin-chuck <strong>of</strong> the<br />
<strong>Micro</strong>Specimen Former and trimmed at the biomaterial–<strong>to</strong>oth<br />
interface <strong>to</strong> a cylindrical hour-glass<br />
Figure 1 Schematic study design.<br />
shape with a <strong>bond</strong>ing surface <strong>of</strong> about 1 mm 2 using<br />
a fine cylindrical diamond bur (835KREF, Komet,<br />
Lemgo, Germany) in a high-speed handpiece under<br />
air/water spray coolant. Specimens were then fixed<br />
<strong>to</strong> Ciucchi’s jig with cyanoacrylate glue (Model<br />
Repair II Blue, Sankin Kogyo, Tochigi, Japan)<br />
and stressed at a crosshead speed <strong>of</strong> 1 mm/min<br />
until failure in a LRX testing device (LRX, Lloyd,<br />
Hampshire, UK) using a load cell <strong>of</strong> 100 N. The mTBS<br />
was expressed in MPa, as derived from dividing<br />
the imposed force (N) at the time <strong>of</strong> fracture by the<br />
<strong>bond</strong> area (mm 2 ). When specimens failed before<br />
actual testing, the mTBS was determined from
1002<br />
the specimens that survived specimen processing<br />
with an explicit note <strong>of</strong> the number <strong>of</strong> pre-testing<br />
failures.<br />
Study design<br />
All specimens were randomly divided in<strong>to</strong> nine groups<br />
(3 <strong>adhesives</strong>!3 experimental groups, Fig. 1) and<br />
subjected <strong>to</strong> a <strong>bond</strong>ing treatment strictly according<br />
<strong>to</strong> the respective manufacturer’s instructions<br />
(Table 1). After adhesive procedures, all teeth were<br />
s<strong>to</strong>red in water for 24 h at 37 8C. For each adhesive,<br />
three teeth were subjected <strong>to</strong> 20,000 thermal cycles<br />
(group 1: thermo-cycling/<strong>cavity</strong>), i.e. the res<strong>to</strong>red<br />
<strong>cavity</strong> was changed between two water baths <strong>of</strong> 5 and<br />
55 8C with a dwell time <strong>of</strong> 30 s at each temperature<br />
extreme (Thermocycler, Willytec, Munich,<br />
Germany). From six other teeth per adhesive, four<br />
mTBS specimens (per <strong>to</strong>oth) were prepared. Two <strong>of</strong><br />
these specimens were also subjected <strong>to</strong> the same<br />
thermo-cycling regimen (group 2: thermo-cycling/<br />
stick). The other half <strong>of</strong> these specimens were s<strong>to</strong>red<br />
for 20 days, the time needed for the thermo-cycling<br />
procedure, in 100% humidity <strong>to</strong> serve as control<br />
(group 3: control).<br />
Statistical analysis<br />
The results were analyzed at a significance level <strong>of</strong><br />
0.05 using a two-way ANOVA and post hoc Tukey–<br />
Kramer multiple comparisons. All statistics were<br />
performed using the statistical s<strong>of</strong>tware package<br />
(StatS<strong>of</strong>t, Tulsa, OK, USA).<br />
Failure analysis<br />
The mode <strong>of</strong> failure was determined light-microscopically<br />
at a magnification <strong>of</strong> 50! using a<br />
stereomicroscope, and recorded as either ‘failure<br />
within dentin’, ‘interfacial failure’ or ‘failure<br />
within resin’.<br />
From each group, representative mTBS-specimens<br />
were processed for field-emission gun scanning<br />
electron microscopy (Feg-SEM, Philips XL30,<br />
Eindhoven, The Netherlands) using common electron<br />
microscopic specimen processing techniques<br />
including fixation, dehydration, chemical drying,<br />
and gold-sputter coating [12].<br />
Results<br />
The mean mTBS, SDs, the number <strong>of</strong> pre-testing<br />
failures (ptf) and the <strong>to</strong>tal number <strong>of</strong> specimens (n)<br />
are summarized per adhesive and experimental<br />
condition in Table 2, and graphically presented in<br />
box-whisker plots in Fig. 2. Thermo-cycling <strong>of</strong><br />
neither the mTBS specimens, nor the res<strong>to</strong>red<br />
cavities decreased the <strong>bond</strong> <strong>strength</strong> <strong>of</strong> the<br />
<strong>adhesives</strong> tested.<br />
Pre-testing failures were only recorded for the<br />
one-step self-etch adhesive (iBOND). All pre-testing<br />
failures occurred during specimen processing<br />
(mostly during preparation <strong>of</strong> the sticks with the<br />
diamond saw). No additional pre-testing failures<br />
were produced by thermo-cycling. Because <strong>of</strong> the<br />
high number <strong>of</strong> pre-testing failures, the data <strong>of</strong><br />
iBOND were excluded from the statistical analysis,<br />
as <strong>to</strong>o few valid data were available <strong>to</strong> perform an<br />
adequate analysis and thus <strong>to</strong> draw a valid<br />
conclusion regarding degradation <strong>of</strong> this adhesive.<br />
The two-way ANOVA analysis disclosed no significant<br />
difference in mTBS between OptiBond FL<br />
and Clearfil Protect Bond (pZ0.321), nor between<br />
the different experimental conditions (control,<br />
thermo-cycling/<strong>cavity</strong> and thermo-cycling/stick;<br />
pZ0.111). The <strong>bond</strong>ing effectiveness <strong>of</strong> the onestep<br />
self-etch adhesive tested, iBOND, was however<br />
already compromised at baseline, given the high<br />
number <strong>of</strong> pre-testing failures (Table 2).<br />
For none <strong>of</strong> the <strong>adhesives</strong>, were morphological<br />
changes induced by thermo-cycling observed using<br />
light-microscopy (Table 3) or Feg-SEM (Fig. 3–5) <strong>of</strong><br />
the fracture surfaces. For the one-step self-etch<br />
adhesive, most specimens failed within the resin<br />
(Table 3; Fig. 5). Especially in the areas that<br />
fractured very close (a few mm) <strong>to</strong> the interface,<br />
the resin appeared very porous, and at higher<br />
magnifications many porosities could be noticed.<br />
The porosity amount and density was clearly higher<br />
in the area near <strong>to</strong> the interface with dentin, but<br />
also in the adhesive resin itself, some larger<br />
porosities could be observed (Fig. 5).<br />
Table 2 mTBS <strong>to</strong> dentin.<br />
mTBS (SD)<br />
ptf/n<br />
Control no<br />
thermocycling<br />
Thermo-cycling (20,000<br />
cycles)<br />
Cavity Stick<br />
OptiBond FL 20.0 (3.6) 27.0 (11.5) 18.3 (9.8)<br />
0/11 0/11 0/13<br />
Protect<br />
Bond<br />
J. De Munck et al.<br />
23.8 (8.3) 24.7 (9.9) 23.1 (7.5)<br />
0/11 0/14 0/12<br />
iBOND 14.7 (11.9) 12.1 (4.9) 12.6 (3.8)<br />
6/9 8/12 11/20<br />
mTBS, micro-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong>, value in MPa; ptf, pretesting<br />
failure; n, <strong>to</strong>tal number <strong>of</strong> specimens; SD, standard<br />
deviation.
<strong>Micro</strong>-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong> <strong>of</strong> <strong>adhesives</strong> <strong>bond</strong>ed <strong>to</strong> <strong>class</strong>-I <strong>cavity</strong>-bot<strong>to</strong>m dentin after thermo-cycling 1003<br />
Figure 2 mTBS after thermo-cycling. The box represents the spreading <strong>of</strong> the data between the first and third quartile.<br />
The central vertical line represents the median. The whiskers denote the range <strong>of</strong> variance and outliers are represented<br />
by a dot.<br />
Discussion<br />
The hypothesis that thermo-cycling <strong>of</strong> res<strong>to</strong>red<br />
occlusal <strong>class</strong>-I cavities (repetitive contraction/<br />
expansion stress) as well as <strong>of</strong> mTBS sticks (diffusion-dependent<br />
hydrolysis and elution) does not<br />
decrease mTBS was confirmed, as for none <strong>of</strong> the<br />
<strong>adhesives</strong> a decreased mTBS was recorded after<br />
thermo-cycling (20,000 cycles). Because <strong>of</strong> the<br />
specific study design, in which thermo-cycled as<br />
well as non-thermo-cycled sticks originated from<br />
the same teeth (Fig. 1), an additional paired<br />
analysis was carried out <strong>to</strong> compare the control<br />
and thermo-cycling/stick group. This analysis is<br />
more powerful than the standard ANOVA analysis,<br />
because the variable ‘<strong>to</strong>oth’ was statistically<br />
excluded. Also using this more powerful analysis,<br />
no significant effect <strong>of</strong> thermo-cycling was<br />
recorded for OptiBond FL (pZ0.68), nor for Clearfil<br />
Protect Bond (pZ0.8624).<br />
S<strong>to</strong>rage <strong>of</strong> small mTBS specimens in water for<br />
relatively short periods (3 months and longer) can<br />
significantly reduce the mTBS [13]. Given the long<br />
time needed <strong>to</strong> implement the thermo-cycling<br />
Table 3 Failure analysis under the light microscope.<br />
Experimental group Interfacial Mixed failure<br />
failure<br />
a<br />
Failure in Total (n)<br />
resin<br />
OptiBond FL Control 1 5 5 11<br />
Thermo-cycling Cavity 2 5 4 11<br />
Stick 4 7 2 13<br />
Protect<br />
Bond<br />
Control 0 6 5 11<br />
Thermo-cycling Stick 1 5 8 14<br />
Cavity 0 8 4 12<br />
1 3<br />
Thermo-cycling Stick 0 3 b<br />
1 4<br />
Cavity 1 b<br />
5 b<br />
3 9<br />
iBOND Control 0 2 b<br />
a Mixed failure, interfacial failure and failure within resin.<br />
b Feg-SEM evaluation revealed that some interfacially failed areas, actually failed within resin (Fig. 5).
1004<br />
J. De Munck et al.<br />
Figure 3 Feg-SEM <strong>of</strong> OptiBond FL. (a) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface <strong>of</strong> a control specimen (s<strong>to</strong>red as mTBS<br />
stick in water for 20 days) at the dentin side. The specimen failed at the interface (I) and within the adhesive resin (Ar).<br />
(b) Composite counterpart <strong>of</strong> (a). (c) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface <strong>of</strong> a thermo-cycling/<strong>cavity</strong> specimen at<br />
the dentin side. The specimen failed at the interface (I) and within the adhesive resin (Ar). (d) Higher magnification <strong>of</strong><br />
the composite counterpart <strong>of</strong> (c). The specimen failed at the <strong>to</strong>p <strong>of</strong> the hybrid layer (Hy), with the heads <strong>of</strong> the resin tags<br />
still attached <strong>to</strong> the adhesive resin. (e) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface (dentin side) <strong>of</strong> a thermo-cycling/stick<br />
specimen. The specimen failed at the interface (I) and within the adhesive resin (Ar). (f) Higher magnification <strong>of</strong> (e) at<br />
an area that failed near the interface. The specimen actually failed at the <strong>to</strong>p <strong>of</strong> the hybrid layer (Hy).<br />
Figure 4 Feg-SEM <strong>of</strong> Clearfil Protect Bond. (a) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface <strong>of</strong> a control specimen (s<strong>to</strong>red<br />
as mTBS stick in water for 20 days) at the dentin side. The specimen failed mainly near the interface (I), apart from a<br />
small part that failed within the adhesive resin (Ar). (b) Higher magnification <strong>of</strong> the composite counterpart <strong>of</strong> (a). The<br />
specimen failed within the hybrid layer (Hy). Some resin flashes (Ar) remained attached <strong>to</strong> the dentin side <strong>of</strong> the beam.<br />
(c) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface (dentin side) <strong>of</strong> a thermo-cycling/<strong>cavity</strong> specimen. The specimen failed at<br />
the interface (I) and within the adhesive resin (Ar). (d) Higher magnification <strong>of</strong> the composite counterpart <strong>of</strong> (c). The<br />
specimen failed within the hybrid layer (Hy). (e) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface (dentin side) <strong>of</strong> a thermocycling/stick<br />
specimen. The specimen failed at the interface (I) and within the adhesive resin (Ar). (f) Higher<br />
magnification <strong>of</strong> (e) at an area that failed near the interface. Again the specimen failed within the hybrid layer (Hy). No<br />
morphological differences with the control group could be observed.
<strong>Micro</strong>-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong> <strong>of</strong> <strong>adhesives</strong> <strong>bond</strong>ed <strong>to</strong> <strong>class</strong>-I <strong>cavity</strong>-bot<strong>to</strong>m dentin after thermo-cycling 1005<br />
Figure 5 Feg-SEM <strong>of</strong> iBOND. (a) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface <strong>of</strong> a control specimen (s<strong>to</strong>red mTBS<br />
specimen in water for 20 days) at the dentin side. The specimen failed mainly within the adhesive resin (Ar). A small part<br />
failed near the interface (I). (b) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface (dentin side) <strong>of</strong> a thermo-cycling/<strong>cavity</strong><br />
specimen. The specimen failed entirely within the adhesive resin (Ar). A large part failed, however, very near <strong>to</strong> the<br />
interface (Ar–I) and appeared very porous. (c) Higher magnification <strong>of</strong> the area marked by the hand-pointer in (b). Many<br />
small porosities can be observed in the resin part close <strong>to</strong> the dentin interface (Ar–I). Also in the adhesive resin itself,<br />
some porosities can be observed. (d) Pho<strong>to</strong>micrograph <strong>of</strong> the fractured surface (dentin side) <strong>of</strong> a thermo-cycling/mTBS<br />
stick specimen. The specimen failed entirely within the adhesive resin (Ar). Again a large part failed near the interface<br />
(Ar–I) and appeared porous. (e) Composite counterpart <strong>of</strong> (d). Part <strong>of</strong> the adhesive resin chipped <strong>of</strong>f during processing<br />
(arrow) and disclosed numerous large voids within the adhesive resin. (f) Higher magnification <strong>of</strong> (e) at an area that<br />
failed near the interface. Many small porosities (0.5–7.5 mm) can be observed.<br />
regimen (20 days), one can speculate that the<br />
degradation <strong>of</strong> the interface <strong>of</strong> the thermo-cycling/<br />
stick group is caused by water exposure rather than<br />
by the thermo-cycling itself. To rule out this option,<br />
the mTBS sticks <strong>of</strong> the control group were also<br />
s<strong>to</strong>red in water for 20 days at 37 8C, this in contrast<br />
<strong>to</strong> previous studies that had 24-h controls [14,15].<br />
However, degradation caused by 20 days <strong>of</strong> water<br />
s<strong>to</strong>rage should have been minimal, as the <strong>bond</strong>s<br />
produced by three-step etch and rinse <strong>adhesives</strong><br />
and mild two-step self-etch <strong>adhesives</strong> resisted up <strong>to</strong><br />
1-year direct water exposure [16]. This assumption<br />
is substantiated by the fact that for OptiBond FL,<br />
the highest mTBS was recorded in the thermocycling/<strong>cavity</strong><br />
group, the only group not directly<br />
exposed <strong>to</strong> water for 20 days.<br />
The composite used in this study is known for its<br />
high E-modulus (21 GPa) [17]. Applying this composite<br />
in a relatively small <strong>class</strong>-I <strong>cavity</strong>, must have<br />
resulted in high polymerization shrinkage stress [7].<br />
By using a high-efficiency, high-power LED curing<br />
device, the polymerization reaction must have<br />
been rather fast, so that also the plastic flow <strong>of</strong><br />
composite, which can reduce the shrinkage stress,<br />
must have been limited. As a result, the <strong>cavity</strong><br />
model used in this study represents a clinical ‘worst<br />
case scenario’. In this study, only OptiBond FL and<br />
Clearfil Protect Bond were able <strong>to</strong> withstand the<br />
shrinkage stress in this challenging situation, in<br />
contrast <strong>to</strong> iBOND that performed unreliably.<br />
From a previous review, it was concluded that<br />
10,000 thermal cycles corresponds <strong>to</strong> 1 year in vivo<br />
degradation [2]. Therefore, the <strong>bond</strong>s produced by<br />
OptiBond FL should be durable for at least 2 years <strong>of</strong><br />
clinical service. This postulation was supported by<br />
many in vitro studies, in which OptiBond FL<br />
successfully withs<strong>to</strong>od <strong>to</strong> up <strong>to</strong> 4 years <strong>of</strong> water<br />
s<strong>to</strong>rage, thermo-cycling and/or mechanical loading<br />
[14,–16,18–20]. Only when miniature (0.4–0.6 mm 2 )<br />
mTBS specimens were aged, was a significant<br />
decrease in mTBS observed for this three-step etch<br />
and rinse adhesive [21,22]. All other types <strong>of</strong><br />
adhesive did, however, decrease at least <strong>to</strong><br />
the same extent in a similar study [23]. Also in<br />
clinical <strong>class</strong>-V studies, this three-step etch and<br />
rinse adhesive performed very reliably for up <strong>to</strong> 5<br />
years <strong>of</strong> clinical service [24,25].<br />
The <strong>bond</strong> <strong>strength</strong>s obtained with the mild twostep<br />
self-etch adhesive Clearfil Protect Bond were<br />
not significantly different from the three-step
1006<br />
etch and rinse control (Table 2; Fig. 2). Given the nonchanged<br />
mTBS and fracture surface ultra-morphology<br />
(Fig. 4), this adhesive also resisted <strong>to</strong> the thermocycling<br />
regimen very well. This is consistent with in<br />
vitro research, in which its predecessor Clearfil SE<br />
(very similar in composition <strong>to</strong> Clearfil Protect Bond,<br />
apart from the antibacterial monomer added <strong>to</strong> the<br />
latter) performed very well. This two-step self-etch<br />
adhesive resisted <strong>to</strong> 1 year in vivo functioning [26,27],<br />
up <strong>to</strong> 30,000 thermo-cycles in a ‘shear-<strong>bond</strong>’ <strong>strength</strong><br />
test [5], and combined thermal and occlusal loading<br />
[9]. Long-term water s<strong>to</strong>rage <strong>of</strong> prepared mTBSbeams<br />
on the other hand, decreased the <strong>bond</strong><br />
<strong>strength</strong> <strong>to</strong> dentin [16,23]; other types <strong>of</strong> adhesive<br />
did, however, decrease at least <strong>to</strong> the same extent in<br />
a similar study [23]. Also in clinical <strong>class</strong>-V studies,<br />
this adhesive performed very well [28,29].<br />
iBOND was not able <strong>to</strong> produce a strong <strong>bond</strong> <strong>to</strong><br />
dentin at the bot<strong>to</strong>m <strong>of</strong> an occlusal <strong>class</strong>-I <strong>cavity</strong><br />
(Table 2). For the control and the thermo-cycling/<br />
stick group, all pre-testing failures occurred after<br />
24 h during further specimen preparation. Because<br />
<strong>of</strong> this low <strong>bond</strong>ing effectiveness at baseline and<br />
the low number <strong>of</strong> remaining specimens, no<br />
conclusion can be drawn regarding degradation <strong>of</strong><br />
the resultant adhesive–<strong>to</strong>oth <strong>bond</strong>. Nonetheless,<br />
the <strong>bond</strong> <strong>strength</strong> <strong>of</strong> this adhesive is <strong>to</strong>o low <strong>to</strong><br />
resist the polymerization shrinkage <strong>of</strong> the res<strong>to</strong>rative<br />
composite in a <strong>class</strong>-I <strong>cavity</strong>. Analysis <strong>of</strong> the<br />
fracture planes revealed that in all groups porosities<br />
were observed in the adhesive resin near the<br />
interface. This certainly must have weakened the<br />
<strong>bond</strong> and is <strong>to</strong> a large extent responsible for the low<br />
<strong>bond</strong>ing effectiveness recorded. Similar porosities<br />
were observed by Tay et al. [30,31]. These<br />
porosities may be due <strong>to</strong> residual solvent (H2O)<br />
that was not adequately removed because <strong>of</strong><br />
inefficient drying in a narrow <strong>cavity</strong>. Alternatively,<br />
these porosities may also be caused by an osmotic<br />
driven water uptake from dentin and/or the<br />
environment, as these one-step self-etch <strong>adhesives</strong><br />
can act as semi-permeable membranes [30]. The<br />
large amount and density, as seen in this study, may<br />
be due <strong>to</strong> the s<strong>to</strong>rage in water for 20 days that<br />
allowed this water uptake <strong>to</strong> take place <strong>to</strong> its full<br />
extent. The <strong>bond</strong>ing effectiveness after 24 h would,<br />
however, not have been that different, as all pretesting<br />
failures occurred during specimen preparation<br />
1 day after adhesive procedures. The most<br />
plausible explanation that follows out <strong>of</strong> recent<br />
research [32,33] is that these porosities represent<br />
water droplets that separated from the monomers<br />
that no longer remained dissolved in water upon<br />
evaporation <strong>of</strong> ace<strong>to</strong>ne.<br />
In conclusion, thermo-cycling did not result in an<br />
enhanced chemical or mechanical degradation <strong>of</strong><br />
the <strong>bond</strong>s <strong>to</strong> dentin produced by a two-step selfetch<br />
and a three-step etch and rinse adhesive. The<br />
<strong>bond</strong>ing effectiveness <strong>of</strong> the one-step self-etch<br />
adhesive tested was, however, <strong>to</strong>o low <strong>to</strong> withstand<br />
polymerization shrinkage stress, as produced in an<br />
occlusal <strong>class</strong>-I <strong>cavity</strong>.<br />
References<br />
J. De Munck et al.<br />
[1] International Organization for Standardization. ISO TR<br />
11405. Dental materials—guidance on testing <strong>of</strong> adhesion<br />
<strong>to</strong> <strong>to</strong>oth structure 1994.<br />
[2] Gale MS, Darvell BW. Thermal cycling procedures for<br />
labora<strong>to</strong>ry testing <strong>of</strong> dental res<strong>to</strong>rations. J Dent 1999;27:<br />
89–99.<br />
[3] Hashimo<strong>to</strong> M, Ohno H, Kaga M, Endo K, Sano H, Oguchi H. In<br />
vivo degradation <strong>of</strong> resin–dentin <strong>bond</strong>s in humans over 1 <strong>to</strong> 3<br />
years. J Dent Res 2000;79:1385–91.<br />
[4] Hashimo<strong>to</strong> M, Tay FR, Ohno H, Sano H, Kaga M, Yiu C, et al.<br />
SEM and TEM analysis <strong>of</strong> water degradation <strong>of</strong> human<br />
dentinal collagen. J Biomed Mater Res 2003;66B:287–98.<br />
[5] Miyazaki M, Sa<strong>to</strong> M, Onose H, Moore BK. Influence <strong>of</strong><br />
thermal cycling on dentin <strong>bond</strong> <strong>strength</strong> <strong>of</strong> two-step<br />
<strong>bond</strong>ing systems. Am J Dent 1998;11:118–22.<br />
[6] Versluis A, Douglas WH, Cross M, Sakaguchi RL. Does an<br />
incremental filling technique reduce polymerization shrinkage<br />
stresses? J Dent Res 1996;75:871–8.<br />
[7] Feilzer AJ, De Gee AJ, Davidson CL. Setting stress in<br />
composite resin in relation <strong>to</strong> configuration <strong>of</strong> the restauration.<br />
J Dent Res 1987;66:1636–9.<br />
[8] Leloup G, D’Hoore W, Bouter D, Degrange M, Vreven J.<br />
Meta-analytical review <strong>of</strong> fac<strong>to</strong>rs involved in dentin<br />
adherence. J Dent Res 2001;80:1605–14.<br />
[9] Nikaido T, Kunzelman KH, Chen H, Ogata M, Harada N,<br />
Yamaguchi S, et al. Evaluation <strong>of</strong> thermal cycling and<br />
mechanical loading on <strong>bond</strong> <strong>strength</strong> <strong>of</strong> a self-etching<br />
primer system <strong>to</strong> dentin. Dent Mater 2002;18:269–75.<br />
[10] Harper RH, Schnell RJ, Swartz ML, Phillips RW. In vivo<br />
measurements <strong>of</strong> thermal diffusion through res<strong>to</strong>rations <strong>of</strong><br />
various materials. J Prosthet Dent 1980;43:180–5.<br />
[11] Watts DC, McAndrew R, Lloyd CH. Thermal diffusivity <strong>of</strong><br />
composite res<strong>to</strong>rative materials. J Dent Res 1987;66:<br />
1576–8.<br />
[12] Perdigão J, Lambrechts P, Van Meerbeek B, Vanherle G,<br />
Lopes ALB. Field emission SEM comparison <strong>of</strong> four postfixation<br />
drying techniques for human dentin. J Biomed Mater<br />
Res 1995;29:1111–20.<br />
[13] Shono Y, Terashita M, Shimada J, Kozono Y, Carvalho RM,<br />
Russell CM, et al. Durability <strong>of</strong> resin–dentin <strong>bond</strong>s. J Adhes<br />
Dent 1999;1:211–8.<br />
[14] De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Vargas M,<br />
Suzuki K, et al. Four-year water degradation <strong>of</strong> <strong>to</strong>tal-etch<br />
<strong>adhesives</strong> <strong>bond</strong>ed <strong>to</strong> dentin. J Dent Res 2003;82:136–40.<br />
[15] De Munck J, Van Meerbeek B, Yoshida Y, Inoue S, Suzuki K,<br />
Lambrechts P. Four-year water degradation <strong>of</strong> a resinmodified<br />
glass–ionomer adhesive <strong>bond</strong>ed <strong>to</strong> dentin. Eur<br />
J Oral Sci 2004;112:73–83.<br />
[16] Shirai K, De Munck J, Yoshida Y, Inoue S, Lambrechts P,<br />
Shintani H, et al. Effect <strong>of</strong> <strong>cavity</strong> configuration and aging on<br />
the <strong>bond</strong>ing effectiveness <strong>of</strong> six <strong>adhesives</strong> <strong>to</strong> dentin. Dent<br />
Mater 2005; 21:110–24.
<strong>Micro</strong>-<strong>tensile</strong> <strong>bond</strong> <strong>strength</strong> <strong>of</strong> <strong>adhesives</strong> <strong>bond</strong>ed <strong>to</strong> <strong>class</strong>-I <strong>cavity</strong>-bot<strong>to</strong>m dentin after thermo-cycling 1007<br />
[17] Abe Y, Lambrechts P, Inoue S, Braem MJA, Takeuchi M,<br />
Vanherle G, et al. Dynamic elastic modulus <strong>of</strong> ‘packable’<br />
composites. Dent Mater 2001;17:520–5.<br />
[18] Blunck U, Roulet JF. Effect <strong>of</strong> one-year water s<strong>to</strong>rage on the<br />
effectiveness <strong>of</strong> dentin <strong>adhesives</strong> in <strong>class</strong> V composite resin<br />
res<strong>to</strong>rations. J Dent Res 2002;81 [Special Issue B, abstr.<br />
946].<br />
[19] Frankenberger R, Strobel WO, Krämer N, Lohbauer U,<br />
Winterscheidt J, Winterscheidt B. Evaluation <strong>of</strong> the fatigue<br />
behavior <strong>of</strong> the resin–dentin <strong>bond</strong> with the use <strong>of</strong> different<br />
methods. J Biomed Mater Res Part B: Appl Biomater 2003;<br />
67B:712–21.<br />
[20] Krejci I, Häusler T, Sägesser D, Lutz F. New <strong>adhesives</strong> in<br />
<strong>class</strong> V res<strong>to</strong>rations under combined load and simulated<br />
dentinal fluid. Dent Mater 1994;10:331–5.<br />
[21] Armstrong SR, Keller JC, Boyer DB. Mode <strong>of</strong> failure in the<br />
dentin–adhesive resin–resin composite <strong>bond</strong>ed joint as<br />
determined by <strong>strength</strong>-based (mTBS) and fracture-based<br />
(CNSB) mechanical testing. Dent Mater 2001;17:201–10.<br />
[22] Armstrong SR, Keller JC, Boyer DB. The influence <strong>of</strong> water<br />
s<strong>to</strong>rage and C-fac<strong>to</strong>r on the dentin–resin composite micro<strong>tensile</strong><br />
<strong>bond</strong> <strong>strength</strong> and de<strong>bond</strong> pathway utilizing a<br />
filled and unfilled adhesive resin. Dent Mater 2001;17:<br />
268–76.<br />
[23] Armstrong SR, Vargas MA, Fang Q, Laffoon JE. <strong>Micro</strong><strong>tensile</strong><br />
<strong>bond</strong> <strong>strength</strong> <strong>of</strong> a <strong>to</strong>tal-etch 3-step, <strong>to</strong>tal-etch 2-step, selfetch<br />
2-step, and a self-etch 1-step dentin <strong>bond</strong>ing system<br />
through 15-month water s<strong>to</strong>rage. J Adhes Dent 2003;5:<br />
47–56.<br />
[24] Van Meerbeek B, Kanumilli P, De Munck J, Van Landuyt K,<br />
Lambrechts P, Peumans M. A Randomized controlled trial<br />
evaluating the three-year clinical effectiveness <strong>of</strong> two etch &<br />
rinse <strong>adhesives</strong> in cervical lesions. Oper Dent 2004; 29:376–85.<br />
[25] Boghosian A. Clinical evaluation <strong>of</strong> a filled adhesive system<br />
in Class 5 res<strong>to</strong>rations. Compend Contin Educ Dent 1996;17:<br />
750–4.<br />
[26] Sano H, Yoshikawa T, Pereira PN, Kanemura N, Morigami M,<br />
Tagami J, et al. Long-term durability <strong>of</strong> dentin <strong>bond</strong>s made<br />
with a self-etching primer, in vivo. J Dent Res 1999;78:<br />
906–11.<br />
[27] Takahashi A, Inoue S, Kawamo<strong>to</strong> C, Omina<strong>to</strong> R, Tanaka T,<br />
Sa<strong>to</strong> Y, et al. In vivo long-term durability <strong>of</strong> the <strong>bond</strong> <strong>to</strong><br />
dentin using two adhesive systems. J Adhes Dent 2002;4:<br />
151–9.<br />
[28] Van Meerbeek B, De Munck J, Yoshida Y, Inoue S, Vargas M,<br />
Vijay P, et al. Buonocore memorial lecture: adhesion <strong>to</strong><br />
enamel and dentin: current status and future challenges.<br />
Oper Dent 2003;28:215–35.<br />
[29] Türkün SL. Clinical evaluation <strong>of</strong> a self-etching and a onebottle<br />
adhesive system at two years. J Dent 2003;31:<br />
527–34.<br />
[30] Tay FR, Pashley DH, Suh BI, Carvalho RM, Itthagaruna A.<br />
Single-step <strong>adhesives</strong> are permeable membranes. J Dent<br />
2002;30:371–82.<br />
[31] Tay FR, King NM, Chan KM, Pashley DH. How can<br />
nanoleakage occur in self-etching adhesive systems that<br />
demineralize and infiltrate simultaneously? J Adhes Dent<br />
2002;4:255–69.<br />
[32] Van Meerbeek B, Van Landuyt K, De Munck J, Yoshida Y,<br />
Inoue S, Lambrechts P. Bonding effectiveness <strong>of</strong> experimental<br />
one-step self-etch <strong>adhesives</strong> <strong>to</strong> enamel and dentin.<br />
J Dent Res 2004;83 [Special issue B Abstract 0028].<br />
[33] Van Landuyt K, De Munck J, Snauwaert J, Coutinho E,<br />
Poitevin A, Yoshida Y, Inove S, Peumans M, Suzuki K,<br />
Lambrechts P, Van Meerbeek B. Monomer-solvent phaseseparation<br />
in contemporary one-step self-etch <strong>adhesives</strong>.<br />
J Dent Res 2005;84:183–8.