Effect of Resin Cement Type on the Microtensile Bond

Strength to Lithium Disilicate Ceramic and Dentin

Using Different Test Assemblies
Susana María Salazar Marochoa / Mutlu Özcanb / Regina Amaralc / Marco Antonio Bottinod /
Luiz Felipe Valandroe

Purpose: This study evaluated the microtensile bond strength (μTBS) of 3 different resin cements to lithium-disil-
icate ceramic using two assemblies: ceramic-cement-ceramic (CCC) and ceramic-cement-dentin (CCD).
Materials and Methods: The bonding surfaces of lithium disilicate ceramic blocks (5 × 5 × 4 mm) (Nblock = 90)
were etched with 4% hydrofluoric acid for 20 s and silanized. Flat dentin surfaces of human third molars were
conditioned according to the respective manufacturer’s specifications for three types of resin cements (ML:
Multilink, Ivoclar-Vivadent; PF: Panavia F, Kuraray; SB: Super Bond C&B, Sun Medical). While one set of ceramic
blocks (n = 30) was cemented to another equal set (CCC assembly), another set of ceramic blocks (n = 30) was
cemented on flat dentin (CCD assembly). The bonded specimens were stored in distilled water at 37ºC for 24 h,
and then sectioned along the x- and y-axes to obtain nontrimmed beam specimens. The beam specimens were
randomly divided into two conditions: dry condition (DC – immediate testing); and aging condition (AC – thermo-
cycling 12,000 times + water storage for 150 days). The μTBS bond strength test was performed using a univer-
sal testing machine (1 mm/min). After debonding, the substrate and adherent surfaces were analyzed using a
scanning electron microscope to categorize the failure types. The data were statistically evaluated using 2-way
ANOVA and Tukey’s test (5%).
Results: While the mean μTBS of CCC assemblies were significantly influenced by the cement type (p < 0.05)
and aging (p < 0.05), CCD assemblies showed a significant effect of the cement (p < 0.05) but not the aging
(p > 0.05). Without aging (DC), the mean μTBS (MPa) of SB (26.9) and PF (26.9) were significantly higher than
ML (18.5) (p < 0.05). For CCC after aging (AC), SB (26.6) showed higher mean μTBS than those of PF (16.4)
and ML (18.5) (p < 0.05). However, in CCD after AC, no significant difference was found between the groups
(p > 0.05). In both CCC and CCD assemblies, pre-test failures were the least with SB cement. Regardless of the
resin cement type employed and storage conditions, adhesive failures ranged between 35.3% and 88.9%, cohe-
sive failures in cement between 2.3% and 35.3%, and cohesive failures in ceramic between 3.3% and 6.8%.
Conclusion: SB resin cement demonstrated the highest bond strength to a lithium disilicate ceramic in both
tests assemblies with and without aging conditions.
Keywords: adhesion, aging, bond strength, ceramic, dentin, lithium disilicate, microtensile bond strength.

J Adhes Dent 2012; 14: 8 pages. Submitted for publication: 15.08.11; accepted for publication: 15.06.12
doi: 10.3290/j.jad.a28624 

a Research Assistant, São Paulo State University, São Jose dos Campos Den-
tal School, Department of Prosthodontics and Dental Materials, São José
dos Campos, Brazil. Performed experiments, analyzed data, prepared draft
of manuscript, wrote and edited manuscript, discussed results and com-
mented on manuscript at all stages.
b Professor, University of Zürich, Dental Materials Unit, Center for Dental and
Oral Medicine, Clinic for Fixed and Removable Prosthodontics and Dental
Materials Science, Zürich, Switzerland. Study design, discussed results and
commented on manuscript at all stages, co-wrote and edited manuscript.
c Faculty of Odontology, MSD/PhD Graduate Program in Oral Science, Pros-
Research Assistant, São Paulo State University, São Jose dos Campos
Dental School, Department of Prosthodontics and Dental Materials, São
José dos Campos, Brazil. Performed experiments, analyzed data, prepared
draft of manuscript, discussed results and commented on manuscript at
all stages.
d Professor, São Paulo State University, São Jose dos Campos Dental School,
Department of Prosthodontics and Dental Materials, São José dos Campos,
Brazil. Discussed results and commented on manuscript at all stages.
e
Associate Professor, Prosthodontic Unit, Faculty of Odontology, Federal Uni-
versity of Santa Maria, Santa Maria, Brazil. Study design, co-wrote and edited
manuscript, discussed results and commented on manuscript at all stages.
Correspondence: Dr. Luiz Felipe Valandro, Federal University of Santa Maria,
thodontic Unit. R. Marechal Floriano, 1184, 97015-372, Santa Maria, Brazil.
Tel: +55-55-3220-9276; Fax: +55-55-3220-9272. e-mail: lfvalandro@hotmail.com
Part of this study was presented at the 87th General Session and Exhibition
of the International Association for Dental Research (IADR), 1-4 April, 2009,
Miami, FL, USA.

doi: 10.3290/j.jad.a28624 1
Marocho et al
T he demands for esthetic restorations have increased
substantially in recent years due to the popularity of
all-ceramic materials as an alternative to metal-ceramic
restorations. One of the major damaging mechanisms
for ceramics is the extension and growth of microscopic
surface flaws, but the resistance to crack propagation
of ceramics has been improved by leucite, fluormica, or
alumina reinforcement of the glass matrix.32 At the end
of the 1990s, a high-strength, fine-grained lithium dis-
ilicate glass ceramic (IPS Empress II, Ivoclar Vivadent;
Schaan, Liechtenstein) was introduced for the fabrica-
tion of metal-free restorations.3
An overview of the clinical performance of adhesively
luted IPS-Empress II crowns showed that of 1487 resto-
rations identified in seven clinical studies 57 had frac-
tured (3.8%).19 The majority of fractures occurred between
the third and the sixth year after placement, in a mean
observation time of 4.5 ± 1.7 years, with a higher fail-
ure rate for molars and canines. In one study, 82 IPS-
Empress II all-ceramic crowns placed by general dental
practitioners showed survival function of 93.9% between
a minimum observation time of 15.2 to a maximum of
57.2 months.30 Fracture was observed in three crowns,
two of which showed a crack upon transillumination. Five
crowns were rated unsatisfactory for color match, one for
marginal adaptation, and none for discoloration, second-
ary caries, or sensitivity. On the other hand, in another
study, 79 IPS Empress 2 crowns presented a 95.2% sur-
vival rate, where only 1 fracture was observed after an
observation period of 12 to 60 months.50 This type of
clinical failure is very often associated with the brittle-
ness and low fracture toughness of ceramic materials.
Small flaws can initiate cracking and cause fracture of
the ceramic restoration. Scanning electron micrographs
of the lithium disilicate glass ceramic showed a closely
packed, multidirectionally interlocking microstructure of
numerous elongated lithium disilicate crystals protruding
from the glass matrix.32 The main advantage of this cer-
amic is that through the injection-molding process, which
uses heat and pressure, randomly oriented crystals incor-
porated in the material create barriers that predispose to
crack formation under tension.12,27,37 In other words, they
act like walls in the prevention of the crack propagation
and catastrophic failures of the all-ceramic restoration
during its function, provoking a tortuous path of crack
propagation. Also, this path could be both transgranular
and intergranular.18 Thus, the added crystals improve the
flexural strength and the fracture toughness; on average,
IPS Empress II can withstand 340 to 400 MPa and 2 to
3.3 MPa·m½.21
Resin cements have been widely used to establish a
durable and reliable bond between the IPS Empress II and
the substrate. The use of resin cements for adhesive ce-
mentation of this ceramic is justified by their low solubility
in the oral environment, optical properties, and bonding
quality.11,12 A significant increase of tensile bond strength
has been attributed to those cements when compared
to the phosphate cement, polycarboxylate cements, and
glass-ionomer cements.25 The disadvantages of adhesive
cementation are related to technique sensitivity, the resin
2 The Journal of Adhesive Dentistry
luting cement’s high index of wear, and the small number
of longitudinal studies with consistent results.7,15,27-29
The selection of a resin luting material often depends
on the type of material and on the clinical requirements
expected by the clinician, as well as the fit of the restor-
ation, thickness of the film, and bonding to the tooth.1
Even in clinical situations when little or no retention is
available in the design of a preparation, the bond to the
dental structure must be effective.5,13,14 The adhesion of
a resin cement in the luting of ceramic restorations has
two aspects: adhesion mediated by the utilization of an
adhesive system on the enamel and dentin, and adhesion
of the resin cement on the etched and silanized ceramic
surface.13,42 Thus, the durability of the bond to the cer-
amic will basically depend on the bond strength of the
luting cement both to the tooth and the ceramic.
Adhesion of resin cements to ceramic surfaces using
the microtensile bond test (μTBS) is typically tested using
the cement between two ceramic blocks or, more clini-
cally, by adhering the ceramic block to a dentin surface
using the resin cement at the interface.2,22,36,48 Little
information is available in the literature with regard to
the bond strength of the tooth/cement/ceramic restor-
ation complex using different resin cements.16,36 In both
test designs, especially after aging conditions, different
results could be expected.
The objectives of the present study therefore were
to a) evaluate the μTBS of 3 different resin cements to
lithium disilicate ceramic using two assemblies: ceramic-
cement-ceramic and ceramic-cement-dentin with and with-
out aging conditions and b) classify the failure types after
aging. The hypothesis tested was that all cements would
show similar bond strengths and aging would decrease
the bond strength for all cements in both test assemblies.
MATERIALS AND METHODS
Brand names, manufacturers, abbreviations, chemical
compositions, batch numbers, and the respective ap-
plication procedures of the resin cements used in this
study are listed in Tables 1a and 1b.
Ceramic-Cement-Dentin Assemblies
Thirty noncarious human molars, extracted for peri-
odontal reasons, were collected from patients through
informed consent. They were cleaned with periodontal
scalers and stored for less than 2 months in distilled
water at 37°C until starting the experiments. The bond-
ing surfaces (midcoronal dentin) were obtained by sec-
tioning the crowns transversally at low speed (LabCut
1010, Extec; Enfield, CT, USA) under intense water
cooling. The roots were removed 3 mm below the ce-
mentoenamel junction in order to facilitate easy sample
manipulation during the adhesive procedures. A smear
layer was created by wet grinding the dentin surface
with 600-grit silicon carbide paper for 60  s.4 Then, the
specimens were washed and randomly divided into
three groups according to the resin cement type listed
in Table 1b.
 Marocho et al

Table 1a Brand names, manufacturers, abbreviations, and the respective application procedures of the resin ce-
ments used in this study

Brand names, Abbre- Dentin pre-treatment procedure Manufacturers’ recommended protocol for mix-
manufacturer viation ing cement
Panavia F PF Apply the ED Primer (A + B) on the Dispense equal amounts of paste A and paste B
(Kuraray; Osaka, Japan) dentin surface for 60 s; gently air dry to (1:1); mix paste A and paste B on the mixing plate
remove the excess of primer. for 20 s; apply on the ceramic surface; remove ex-
cess cement and light cure for 20 s, apply oxyguard.

Super Bond C&B SB Apply the green activator (batch LE1) Brush-dip technique.
(Sun Medical; Moriyama, on the dentin surface for 5 to 10 s;
Japan) rinse and gently air dry.

Multilink ML Mix Multilink Primer A and B for 15 s; The single pastes are dispensed from the Multilink
(Ivoclar Vivadent; apply the primer on the dentin surface double-push syringe and mixed in a 1:1 ratio.
Schaan, Liechtenstein) for 15 s and air dry.

Table 1b Chemical compositions and batch numbers of the resin cements used in this study

Abbreviation Chemical Composition Batch number

PF Primer A: 10-MDP, HEMA, ethylene glycol methacrylate, N, N-di-(2-hydroxyethyl)-p-toluidine, 5-NMSA, 00198C
water

Primer B: benzene sulfinic acid, sodium salt, N, N-di-(2-hydroxyethyl)-p-toluidine, 5-NMSA, water 00078C

Paste A: 10-MDP, 5-NMSA, sílica, dimethacrylate monomer, photo-initiator, accelerator 00243A

Paste B: barium glass, sodium fluoride, dimethacrylate monomer 00138A

SB Green activator: 10% citric acid, 3% ferric chloride LG2

ML Multilink Primer A: an aqueous solution of initiators H14896

Multilink Primer B: HEMA, phosphoric acid and acrylic acid monomers

Paste A and B: 22%–26% dimethacrylates, 6%–7%, HEMA, 1% benzoyl peroxide, 40%

barium glass, ytterbium trifluoride, spheroid mixed oxide

Empress II ceramic blocks (5 mm × 5 mm × 4 mm) Ceramic-Cement-Ceramic Assemblies

(N = 30) were fabricated according to the manufacturer’s
specifications, polished under water cooling, and ultra-
sonically cleaned (Vitasonic, Vita Zanhfabrik; Bad Säck-
ingen, Germany) in distilled water for 5 min. The ceramic
surfaces were etched with 5% hydrofluoric acid (HF) (IPS
Ceramic Etching, Ivoclar Vivadent, batch #H33247) for
20 s, and then washed with running water for 20 s. Next,
the neutralizing agent (IPS Ceramic Neutralizing powder,
Ivoclar Vivadent, batch #H33006) was applied on the cer-
amic surface for 5 min and removed by washing the cer-
amic surfaces with air-water spray for 20 s, followed by air
drying. Following etching and neutralizing, MPS-based si-
lane coupling agent (Monobond S, Ivoclar Vivadent, batch
#G25687) was applied with a brush.
The dentin surfaces were conditioned according to the
manufacturer’s instructions of the corresponding resin
cements and the ceramics were bonded on the dentin
surfaces (ceramic-cement-dentin assemblies).
Another set of Empress II ceramic blocks (5 mm ×
5 mm × 4 mm) (N = 60) was fabricated according to
the manufacturer’s specifications and conditioned as
described above. After that, each conditioned ceramic
was cemented to another conditioned ceramic block
(ceramic-cement-ceramic assemblies) using the three
resin cements.
The cements were manipulated according to each man-
ufacturer’s recommendation and bonding was achieved
under a vertical load (750 g) for 10 min. During this
period, excess material was removed, and each surface
was photoactivated for 40 s (XL3000, 3M ESPE; St Paul,
MN, USA, operating at 600 mW/cm2). An oxygen-blocking
agent (Oxyguard, Kuraray; Tokyo, Japan) was applied to
the margins of the ceramic/dentin interface. The assem-
blies were washed and rinsed with water, then stored in
distilled water at 37°C for 24 h prior to sectioning into
test beams.

doi: 10.3290/j.jad.a28624 3
Marocho et al
Table 2a Microtensile bond strength (SD) results
in MPa for ceramic-cement-ceramic assemblies, in dry
and aged conditions
Cement × storage Dry conditions Aged conditions
SB 26.9 (9)A,a 26.6 (12.2)A,a
ML 18.5 (5.2)B,a 18.5 (13.5)B,a
PF 26.9 (7.3)A,a 16.4 (5.7)B,b
*Same capital superscript letters in each column and lower-case letters
in each row indicates no significant difference (Tukey’s test, _ = 0.05).
SD: standard deviation.
Specimen Preparation for Microtensile Test
The stored sets of bonded assemblies were cut along
two axes (x and y) with a diamond disk under water
cooling in a cutting machine (LabCut 1010, Extec). This
process resulted in non-trimmed beam specimens with
a bonded area of approximately 0.8 mm2, as measured
with a digital caliper.
The beam specimens obtained from each group were
randomly divided into two testing conditions. While half of
the specimens in each group were submitted to the micro-
tensile test immediately after sectioning (DC, immediate
testing), the other half were subjected to aging conditions
(AC) where they were first aged in a thermocycling ma-
chine (Nova Etica; São Paulo, Brazil) for 12,000 cycles at
5°C–55°C with a dwell time of 30 s, and a transfer time
of 2 s, followed by storage in distilled water at 37°C for
150 days before testing.
Microtensile Test and Failure Analysis
Each beam specimen was attached to a custom-made
testing device with cyanoacrylate gel. The device con-
sisted of an adapted caliper meant to ensure that pure
tensile force was applied to the adhesive interface
without any shear or torque forces. The microtensile
test (crosshead speed: 0.5 mm/min, 100 N load cell)
was performed in a universal testing machine (Emic DL-
1000, Emic; São José dos Pinhais, PR, Brazil).
After debonding, specimen surfaces were analyzed
using optical microscopy analysis and SEM (SSX-550, Shi-
matzu; Kyoto, Japan) to characterize the failure modes.
After evaluating all SEM images, the failure types were
defined as a) adhesive failure between the cement and
the ceramic (ADHES); b) cohesive failure of cement and
ceramic (MIX); c) cohesive failure of the cement (COHES-
cem); d) cohesive failure of ceramic (COHES-cer).
Statistical Analysis
Statistical analysis was performed using SPSS 11.0
software for Windows (SPSS; Chicago, IL, USA). Micro-
tensile bond strength data (MPa) were submitted to
three-way ANOVA with the bond strength as the depen-
dent variable and cement type (3 levels; Multilink, Pana-
via F, Super Bond C&B), aging condition (2 levels; dry vs
artificial aging), and testing assembly (2 levels; ceramic-
4 The Journal of Adhesive Dentistry
Table 2b Microtensile bond strength (SD) results
in MPa for ceramic-cement-dentin assemblies, in dry
and aged conditions
Cement × Storage Dry conditions Aging conditions
SB 24.7 (18.2)A,a 20.8 (9.2)A,a
ML 7.8 (2.7)B,a 18.5 (13.6)A,b
PF 17.1 (15.3)A,a 19.9 (7.9)A,a
*Same capital superscript letters in each column and lower-case letters
in each row indicates no significant difference (Tukey’s test, _ = 0.05).
SD: standard deviation.
cement-ceramic, ceramic-cement-dentin) as indepen-
dent variables. Pre-test failures were not considered in
the statistical calculations. Multiple comparisons were
made using Tukey’s test. p-values less than 0.05 were
considered to be statistically significant in all tests.
RESULTS
While mean μTBS of CCC assemblies were significantly
influenced by the cement type (p < 0.05) and aging
(p < 0.05), CCD assemblies showed a significant effect
of the cement (p < 0.05), but not the aging (p > 0.05).
Without aging (DC), the mean μTBS (MPa) of SB (26.9)
and PF (26.9) were significantly higher than ML (18.5)
(p < 0.05). For CCC after aging (AC), SB (26.6) showed
a higher mean μTBS than did PF (16.4) and ML (18.5)
(p < 0.05). In contrast, no significant difference was
found between the groups for CCD specimens after AC
(p > 0.05).
In both CCC and CCD assemblies, the fewest pre-test
failures were found with SB cement (Tables 2a and 2b).
In both DC and AC conditions, with both test assemblies,
Super Bond resulted in fewer pre-test failures than did the
other cements (Table 3).
Regardless of the resin cement type employed and
storage conditions, ADHES failures ranged between
35.3% and 88.9%, COHES-cem failures between 2.3% and
35.3%, and COHES-cer between 3.3% and 6.8% (Table 4).
DISCUSSION
One of the most commonly used ceramic materials for
esthetic, fracture-resistant dental restorations is lithium
disilicate glass ceramic, commercially marketed by
Ivoclar Vivadent in 1998 as IPS Empress II, now called
IPS e.max Press. Typically, unsupported ceramics are
prone to fracture under chewing loads.6 Hence, in order
to gain support from the underlying tooth to dissipate
occlusal forces, ceramics should be adhesively bonded
to form an integrated structure. Although IPS Empress
II may be conventionally or adhesively cemented,39 it is
recommended to choose an adhesive luting agent when
 Marocho et al

Table 3 Distribution of the pre-test failures at different stages of the experiment and the final number (percentage)
of the tested beams in both test assemblies

Assemblies Cement No. of beams No. and % of pre- No. and % of spon- No. and % of tested
planned test failures during taneous pre-test beams in μTBS
cutting failures during ther-
mocycling
DC AC DC AC DC AC DC AC
Ceramic-Cement- PF 90 (100) 90 (100) 86 (95.6) 85 (94.4) 0 (0) 0 (0) 4 (4.4) 5 (5.6)
Dentin
SB 90 (100) 90 (100) 68 (75.6) 65 (72.2) 0 (0) 0 (0) 22 (24.4) 25 (27.8)

ML 90 (100) 90 (100) 83 (92.2) 80 (88.9) 0 (0) 0 (0) 7 (7.8) 10 (11.1)

Ceramic-Cement- PF 90 (100) 90 (100) 65 (72.2) 65 (72.2) 0 (0) 0 (0) 25 (27.8) 25 (27.8)

Ceramic
SB 90 (100) 90 (100) 60 (66.7) 60 (66.7) 0 (0) 0 (0) 30 (33.3) 30 (33.3)

ML 90 (100) 90 (100) 68 (75.6) 68 (75.6) 0 (0) 0 (0) 22 (24.4) 22 (24.4)

DC: no aging; AC: aging.

Table 4 Distribution of the frequencies of failure types (%) in both test assemblies

Assemblies Cement Type of failure

ADHES MIX COHES-cem COHES-cer
Ceramic-Cement- PF 8 (88.9) 1 (11.1) 0 (0) 0 (0)
Dentin
SB 40 (85.1) 7 (14.9) 0 (0) 0 (0)

ML 6 (35.3) 5 (29.4) 6 (35.3) 0 (0)

Ceramic-Cement- PF 35 (70) 12 (24) 3 (6) 0 (0)

Ceramic
SB 47 (78.3) 7 (11.7) 4 (6.7) 2 (3.3)

ML 33 (75) 7 (15.9) 1 (2.3) 3 (6.8)

ADHES: adhesive failure between the cement and the ceramic; MIX: cohesive failure of cement and ceramic; COHES-cem: cohesive failure of the cement;
COHES-cer: cohesive failure of the ceramic.

retentive preparation is not available. For durable adhe-
sion, the prerequisites are HF acid etching, followed by
silane application of the intaglio ceramic surface and
bonding to dentin using dentin bonding agents and ap-
propriate resin luting cements.18,32,33 Despite the fact
that HF acid etching exerted a weakening effect on the
flexural strength of this type of ceramic,24,56 adhesion
to dentin is required to achieve suitable bonding of the
resin cements to dentin structures.8
Bonding effectiveness may influence the prognosis of the
ceramic restoration and the clinical success.47 Thus, it is
important to identify the most reliable and effective method
of bonding both at the cement/tooth and the cement/cer-
amic restoration interface, which relies on formation of
micromechanical interlocking as well as chemical bonds.
Hence, this study aimed to evaluate the μTBS bond strength
of 3 different resin cements to a lithium disilicate ceramic in
ceramic-cement-ceramic (CCC) and ceramic-cement-dentin
(CCD) assemblies. Since the cement factor showed a sig-
nificant effect in both assemblies, the first hypothesis that
all cements would show similar bond strength could be re-
jected. However, aging had a significant effect on the results
with some exceptions; therefore, the second hypothesis
that aging would decrease the bond strength for all cements
in both test assemblies could be only partially accepted.
It should be noted that a high number of pre-test failures
occurred in this study. This could be partially attributed to
the brittleness of the ceramic in general; in addition, the
cutting blade possibly resulted in more stresses and torque
forces especially at the cement/dentin interface. Compared
to PF and ML, SB resulted in fewer failures in both the CCD
and CCC groups. Since SB is a chemically polymerized resin
cement, the interface was probably polymerized better than
the dual-polymerizing resin cements, considering that pre-
test failures were not completely eliminated in the SB group.
During cutting procedures, only one site of the assembly

doi: 10.3290/j.jad.a28624 5
Marocho et al
was glued to the metal holder, but securing the whole as-
sembly with a resin material could be tried in future studies.
In the CCC group, conditioning ceramic with hydrofluoric
acid and silanization seemed to deliver better adhesion
compared to the CCD group, where dentin was conditioned.
The ceramic-ceramic assembly was most likely more stable
with better conditioning and adhesion of the resin cement
to both ceramic substrates. When one side of the bond
of the adhesive material to the substrate suffers from in-
ferior adhesion, this may also trigger delamination of the
adhesive material from the other substrate. This competing
crack propagation eventually leads to pre-test failure. In this
study, the almost negligible frequency of cohesive failure in
the ceramic substrates indicates that the pre-test failure is
partially due to the weak cohesive strength of the cement.
A careful evaluation of the site (dentin-cement or cement-
ceramic) that triggers the first delamination needs to be
identified in μTBS studies, but this was not performed in the
present study. During cutting procedures, primarily ceramic
debonded as a block before all sticks were obtained. Yet the
high number of pre-test failures in the CCD groups vs CCC
with all cements tested indicates weaker adhesion of the
resin cements to dentin.
Several previous studies investigated the durability of
the bond strength either by storing the specimens in water
at 37°C, 60°C, and 100°C for different time periods, or
by thermocycling.16,20,22,23,36,40,42,44,46,48,54 Long-term
water storage at a constant temperature or thermocycling
are the most commonly practiced methods to simulate the
aging process that influences the resin bonds to ceramic
surfaces for cementation or repair procedures. Thermo-
cycling promotes differences in the coefficient of thermal
expansion of resin cements based on their components
(matrix and fillers); these differences increase the internal
stresses and subject the bonded interface,22 be it resin-
dentin or resin-ceramic, to hydrothermal degradation.16,36
Our results showed that after 24-h storage, the adhesive
cementation using SB and PF in both CCC and CCD com-
plexes resulted in higher bond strength values than did ML.
In ceramic-cement-dentin assemblies, SB resulted in bond
strengths that did not change as a function of long-term
water storage. In fact, hydrolytic or chemical degradation
is assumed to be diffusion and time dependent; it takes
time to penetrate the interface and cause chemical break-
down.49 However, variations in chemical composition, wet-
ting capacity, viscosity, and mechanical properties for each
resin cement could also be responsible for variations in
the bonding capacity to the ceramic substrate and for the
different failure rates. Since there are no previous stud-
ies evaluating the bond strength of SB cement to lithium
disilicate glass ceramics, it was not possible to compare
the present data. However, as the chemical formulation of
any material determines it performance in clinic, the high-
est bond strength values of SB to the ceramic surface can
be related and discussed according to its composition.
In comparison to the other bonding systems used in this
study, SB monomer is composed of 4-methacryloxyethyl
trimellitate anhydride (4-META) that polymerizes with MMA
to form a co-polymer. Despite the fact that 4-META bears
two carboxyl groups attached to the aromatic group that
6 The Journal of Adhesive Dentistry
provide acidic and thus demineralizing properties while
also enhancing wetting, its aromatic group is hydrophobic
and it decreases the acidity and the hydrophilicity of the
carboxyl groups.51 Hence, this monomer is well soluble in
acetone, moderately soluble in ethanol, and not well solu-
ble in water. The reason why SB cement also produced high
bond strength values to dentin, which did not change after
aging, may be attributed to the etching effect of the green
activator on decalcification of the dentin surface, removing
the smear layer and the hydroxyapatite, thus allowing the
penetration of the monomer into that zone. As a function
of the 4-META component, the diffusion of the monomer is
enhanced, creating a strong bond to this prepared dentin.
However, the ionic bond formed by 4-META with calcium
is less intense than 10-methacryloyloxydecamethylene
phosphoric acid (10-MDP).55 This monomer is capable of
forming strong ionic bonds with calcium due to the low
dissolution rate of the resulting calcium salt in its own
solution.55 In a literature review,53 10-MDP was rated as
the most promising monomer for chemical bonding to hy-
droxyapatite of enamel or dentin. This fact may explain the
high bond strength values obtained with Panavia F in the
CCD assemblies, which is due to favorable dentin bonding.
Only CCC assemblies bonded with PF presented a sig-
nificant decrease in bond strength values after aging.
These results may indicate that the resin cements studied
here are subject to some degree of dissolution after water
storage, but apparently PF underwent more water sorption
and solubility compared to the other cements, which may
have a number of undesirable consequences in its clinical
application.41 Water sorption has been shown to result in
the deterioration of mechanical/physical properties of the
material,35,43,44 as it may lead to degradation of the filler/
matrix interface and act as a plasticizer within the poly-
mer matrix.11 Such hydrolytic degradation weakens the
cement layer further and can adversely affect clinical per-
formance of the complex formed. Hydrolytic degradation
can facilitate debonding through the outer atomic layers
of the ceramic surface. Thus, as the ceramic-composite
bond ages, the ceramic restoration may debond.40
Despite the fact that ML does not contain adhesive
phosphate monomer, its dimethacrylate monomers and
HEMA seemed to improve its mechanical properties,
which may be responsible for high bond strength results.
Unfortunately, the results were not consistent in the CCD
assemblies, partially due to high number of pre-test fail-
ures. The reaction of ML dentin adhesives need to be
further investigated.
Comparing the bond strength results in CCC assem-
blies with CCD assemblies, early microtensile bond
strength values (DC) were higher in the former. This may
be associated with the use of silane in both cementation
surfaces, which could have improved the bond strength
between the ceramic and resin based-materials.10 The
silane coupling agent use in this study couples the silica
oxides present in the glassy matrix and lithium disilicate
crystal in the IPS Empress II48 to the organic matrix of
resin cements by means of siloxane bonds.31,45 Moreo-
ver, due to the bifunctional characteristics of silane, its
application on the etched ceramic surface promotes the

cement wettability on the ceramic surface,2,36,38 enhanc-
ing the contact with resin cements in irregularities caused
on ceramic surface by etching.
The effect of humid conditions not only on sintered
ceramic structures but also on human dentin must also
be taken into account. The detrimental effects of silane
hydrolysis and degradation40 may eventually decrease the
dentin-ceramic bond, yielding delamination. However, de-
pending on the polymerization degree, cohesive strength,
water sorption and adhesion level of the cement to den-
tin, this delamination may be accompanied by dentin-ce-
ment debonding. Notwithstanding that dentin in extracted
human teeth lacks dentinal fluid,26,37 it takes up water
during water storage, diffusing into intertubular dentin.52
Moreover, it is important to emphasize the function of the
other factors in adhesion, since silane coupling agents
themselves are only capable of forming bonds with both
inorganic and organic surfaces but do not promote hydro-
lytic stability.31
Looking for a method that produces uniform stress dis-
tribution across the bonded interfaces, previous studies
have evaluated similar adhesive systems under different
bond testing configurations.9,25 These studies suggest
that a microtensile bond strength test may be more ap-
propriate for evaluating the bond strength of adhesive
interfaces because of more uniform interfacial stresses
at the adhesive interface.
The bond strength data should be considered in con-
junction with the failure types. Overall, it may be inferred
that in CCC, the resin cement created stronger bonds
than CCD assemblies, since in the latter, more adhesive
failures were observed. In addition to the chemical degra-
dation that confronts resin cements along cement/dentin
interfaces, water storage can lead to erosion caused by
the release of unreacted monomers, which may result in
massive loss of the resin cement.17,34
In short, the adhesive performance of resin cements
should be cautiously evaluated in the future, consider-
ing both the test assembly and the aging procedures
practiced.
CONCLUSIONS
From this study, the following could be concluded:
y Testing the adhesive strength of resin cements in
ceramic-cement-ceramic or ceramic-cement-dentin as-
semblies showed differences in non-aged conditions,
but after aging, the results were comparable.
y In both non-aged and aged testing conditions, with
both test assemblies, chemically cured Super Bond
C&B resulted in fewer pre-test failures than did other
cements, indicating more durability to both the ce-
ramic and the dentin.
y Regardless of the resin cement type and the storage
conditions, adhesive failures between the cement
and the ceramic and dentin surfaces were more com-
monly observed. While cohesive failures in dentin
were not observed with any of the cements, cohesive
failures of the ceramic were restricted.
doi: 10.3290/j.jad.a28624
Marocho et al
ACKNOWLEDGMENT
The authors would like to acknowledge the manufacturers for
generous provision of the resin cements tested. S.M.S.M. acknowl-
edges a fellowship from CAPES/CNPq - IEL Nacional - Brazil.
REFERENCES
1. ADA Council on Scientific affairs. Direct and indirect restorative materi-
als. J Am Dent Assoc 2003;134:463-472.
2. Aida M, Hayakawa T, Mizukawa K. Adhesion of composite to porcelain
with various surface conditions. J Prosthet Dent 1995;73:464-470.
3. Beham G. IPS Empress: eine neue Keramik-Technologie. Zahnärztl Welt
Reform 1991;100:404-408.
4. Belli R, Sartori N, Peruchi LD, Guimarães JC, Vieira LC,  Baratieri LN,
Monteiro S Jr. Effect of multiple coats of ultra-mild all-in-one adhesives
on bond strength to dentin covered with two different smear layer thick-
nesses. J Adhes Dent 2011;13:507-516.
5. Browning WD, Nelson SK, Cibirka R, Myers ML. Comparison of luting
cements for minimally retentive crown preparations. Quintessence Int
2002;33:95-100.
6. Burke FJ, Watts DC. Fracture resistance of teeth restored with dentin-
bonded crowns. Quintessence Int 1994;25:335-340.
7. Burke FJT, Qualthrough AJE, Wilson NHF. A retrospective evaluation of
a series of dentin-bonded crowns. Quintesense Int 1998;29:103-106.
8. Chen JH, Matsumura H, Atsuta M. Effect of different etching periods
on the bond strength of a composite resin to a machinable porcelain. J
Dent 1998;26:53-58.
9. Della Bona A, van Noort R. Shear vs. tensile bond strength of resin
composite bonded to ceramic. J Dent Res 1995;74:1591-1596.
10. Della Bona A, Shen C, Anusavice KJ. Work of adhesion of resin on
treated lithia disilicate-based ceramic. Dent Mater 2004;20:338-344.
11. Diaz-Arnold AM, Arnold MA, Williams VD. Measurement of water sorp-
tion by resin composite adhesives with Near-Infrared Spectroscopy. J
Dent Res 1992;71:438-442.
12. Dietschi D, Maeder M, Meyer JM, Holz J. In vitro resistance to frac-
ture of porcelain inlays bonded to tooth. Quintessence Int 1990;21:
823-831.
13. Dong JK, Lüthy H, Wohlwend A, Schärer P. Heat-pressed ceramics: tech-
nology andstrength. Int J Prosthodont 1992;5:9-16.
14. Duken van JWV. Multiple versus one-bottle bonding systems. Rèalités
Cliniques 1999;10:199-222.
15. Duken van JWV, Hasselrot L, Örmin A, Olofsson AL. Clinical evaluation
of extensive dentin/enamel-bonded all ceramic onlays and onlays-
crowns. Eur J Oral Sci 2001;109:222-229.
16. Foxton RM, Pereira PNR, Nakajima M, Tagami J, Miura H. Durability of
the dual-cure resin cement / ceramic bond with different curing strate-
gies. J Adhes Dent 2002;4:49-55.
17. Gopferich A. Mechanisms of polymer degradation and erosion. Biomate-
rials 1996;17:103-114.
18. Guazzato M, Albakry M, Ringer SP, Swain MV. Strength, fracture
toughness and microstructure of a selection of all-ceramic materials.
Part I. Pressable and alumina glass-infiltrated ceramics. Dent Mater
2004;20:441-448.
19. Heintze SD, Rousson V. Fracture rates of IPS Empress all-ceramic
crowns – a systematic review. Int J Prosthodont 2010;23:129-133.
20. Helvatjoglu-Antoniades M, Koliniotou-Kubia E, Dionyssopoulos P. The
effect of thermal cycling on the bovine dentine shear bond strength of
current adhesive systems. J Oral Rehabil 2004;31:911-917.
21. Höland W, Schweiger M, Frank M, Rheinberger V. A comparison of the
microstructure and properties of the IPS Empress 2 and the IPS Em-
press glass-ceramics. J Biomed Mater Res 2000;53:297-303.
22. Hooshmand T, van Noort R, Keshvad A. Bond durability of the resin
bonded and silane treated ceramic surface. Dent Mater 2002;18:
179-188.
23. Hooshmand T, van Noort R, Keshvad A. Storage effect of a pre-activated
silane on the resin to ceramic bond. Dent Mater 2004;20:635-642.
24. Hooshmand T, Parvizi S, Keshvad A. Effect of surface acid etching on
the biaxial flexural strength of two hot-pressed glass ceramics. J Prosth-
odont 2008;17:415-419.
25. Kitasako Y, Burrow MF, Nikaido T, Harada N, Inokoshi S, Yamada T,
Takatsu T. Shear and tensile bond testing for resin cement evaluation.
Dent Mater 1995;11:298-304.
7
Marocho et al

26. Komabayashi T, Ahn C, Zhang S, Zhu Q, Spångberg LS. Chronologic 43. Scarret DC, Söderholm KJ, Ybatich CD. Water and abrasive effects on
comparison of root dentin moisture in extracted human teeth stored three-body wear of composites. J Dent Res 1991;70:1074-1081.
in formalin, sodium azide, and distilled water. Oral Surg Oral Med Oral 44. Söderholm KJM, Roberts MJ. Influence of water exposure on the tensile
Pathol Oral Radiol Endod 2009;108:e50-e54. strength of composites. J Dent Res 1990;69:1812-1816.
27. Krämer N, Frankenberger R. Clinical performance of bonded leucite- 45. Söderholm KJM, Shang SW. Molecular orientation of silane at the sur-
reinforced glass ceramic inlays and onlays after eight years. Dent Mater face of colloidal silica. J Dent Res 1993;72:1050-1054.
2005;21:262-271.
46. Sorensen JA, Engelman MJ, Torres TJ, Avera SP. Shear bond strength of
28. Krejci I, Krejci D, Lutz F. Clinical evaluation of a new pressed glass- composite resin to porcelain. Int J Prosthodont 1991;4:17-23.
ceramic inlay material over 1.5 years. Quintessence Int 1992;23:
47. Sorensen JA, Choi C, Fanuscu MI, Mito WT. IPS Empress Crown
181-186.
System: Three-year clinical trial results. J Calif Dent Assoc 1998;26:
29. Krejci I, Lutz F, Reimer M, Heinzmann J. Wear of ceramic inlays, their 130-136.
enamel antagonists and luting cements. J Prosthet Dent 1993;69:
48. Spohr AM, Sobrinho LC, Consani S, Sinhoreti MA, Knowles JC. Influence
425-430.
of surface conditions and silane agent on the bond of resin to IPS Em-
30. Mansour YF, Al-Omiri MK, Khader YS, Al-Wahadni A. Clinical perfor- press 2 ceramic. Int J Prothodont 2003;16:277-282.
mance of IPS-Empress 2 ceramic crowns inserted by general dental
49. Staninec M, Kim P, Marshall GW, Ritchie RO, Marshall SJ. Fatigue
practitioners. J Contemp Dent Pract 2008;1;9:9-16.
of dentin-composite interfaces with four-point bend. Dent Mater
31. Matinlinna JP, Lassila LVJ, Özcan M, Yli-Urpo A, Vallittu PK. An introduc- 2008;24:799-803.
tion to silanes and their clinical applications in dentistry. Int J Prostho-
50. Toksavul S, Toman M. A short-term clinical evaluation of IPS Empress 2
dont 2004;17:155-164.
crowns. Int J Prosthodont 2007;20:168-172.
32. Oh SC, Dong JK, Lüthy H, Schärer P. Strength and microstructure of IPS
51. Unemori M, Matsuya Y, Matsuya S, Akashi A, Akamine A. Water ab-
Empress 2 glass-ceramic after different treatments. Int J Prosthodont
sorption of poly(methyl methacrylate) containing 4-methacryloxyethyl
2000;13:468-472.
trimellitic anhydride. Biomaterials 2003;24:1381-1387.
33. Ohyama T, Yoshinari M, Oda Y. Effects of cyclic loading on the strength
52. van der Graaf ER, ten Bosch JJ. The uptake of water by freeze-dried
of all-ceramic materials. Int J Prosthodont 1999;12:28-37.
human dentine sections. Arch Oral Biol 1990;35:731-739.
34. Ortengren U, Wellendorf H, Karlsson S, Ruyter IE. Water sorption and
53. Van Landuyt KL, Snauwaert J, De Munck J, Peumans M, Yoshida Y,
solubility of dental composites and identification of monomers released
Poitevin A, Coutinho E, Suzuki K, Lambrechts P, Van Meerbeek B.
in an aqueous environment. J Oral Rehabil 2001;28:1106-1115.
Systematic review of the chemical composition of contemporary dental
35. Oysaed H, Ruyter IE. Composites for use in posterior teeth. Mechanical adhesives. Biomaterials 2007;28:3757-3785.
properties tested under dry and wet conditions. J Biomed Mater Res
54. Wegner SM, Gerdes W, Kern M. Effect of different artificial aging
1986;20:261-271.
conditions on ceramic-composite bond strength. Int J Prosthodont
36. Özcan M, Vallittu PK. Effect of surface conditioning methods on the 2002;15:267-272.
bond strength of luting cement to ceramics. Dent Mater 2003;19:
55. Yoshida Y, Nagakane K, Fukuda R, Nakayama Y, Okazaki M, Shintani H,
725-731.
Inoue S, Tagawa Y, Suzuki K, De Munck J, Van Meerbeek B. Compara-
37. Ozok AR, Wu MK, De Gee AJ, Wesselink PR. Effect of dentin perfusion tive study on adhesive performance of functional monomers. J Dent Res
on the sealing ability and microtensile bond strengths of a total-etch 2004;83:454-458.
versus an all-in-one adhesive. Dent Mater 2004;20:479-486.
56. Zogheib LV, Bona AD, Kimpara ET, McCabe JF. Effect of hydrofluoric
38. Phoenix S, Shen C. Characterization of treated porcelain surfaces via acid etching duration on the roughness and flexural strength of a lithium
dynamic contact angle analysis. Int J Prosthodont 1995;8:187-194. disilicate-based glass ceramic. Braz Dent J 2011;22:45-50.
39. Pröbster L, Geis-Gerstorfer J, Kirchner E, Kanjantra P. In vitro evalua-
tion of a glass-ceramic restorative material. J Oral Rehabil 1997;24:
636-645.
40. Roulet JF, Söderholm KJ, Longmate J. Effects of treatment and stor-
age conditions on ceramic/composite bond strength. J Dent Res
1995;74:381-387. Clinical relevance: Considering both the microtensile
41. Ruyter IE. Physical and chemical aspects related to substances re- bond strength data and the failure types, chemically
leased from polymer materials in an aqueous environment. Adv Dent
Res 1995;9:344-347.
polymerized Super Bond C&B and its corresponding
42. Salvio LA, Correr-Sobrinho L, Consani S, Sinhoreti MA, de Goes adhesive system provided the most favorable adhe-
MF, Knowles JC. Effect of water storage and surface treatments on sion to hydrofluoric acid-etched and silanized lithium
the tensile bond strength of IPS Empress 2 ceramic. J Prosthodont dilisilicate ceramic and the conditioned dentin.
2007;16:192-199.

8 The Journal of Adhesive Dentistry