Durability of Resin Bonding to Zirconia Using Products

Containing 10-Methacryloyloxydecyl Dihydrogen

Phosphate
Lu Yanga / Bingzhuo Chena / Haifeng Xieb / Ying Chenc / Yue Chenc / Chen Chend

Purpose: To examine the durability of composite bonding to zirconia after artificial aging using different products
containing 10-methacryloyloxydecyl dihydrogen phosphate (MDP).
Materials and Methods: Conditioning methods were: none (control [Ctr]; Clearfil SA Luting [CSL]; Panavia SA Luting
Plus [PSLP]), MDP-containing zirconia primers (Z-Prime Plus [ZP]; Clearfil Ceramic Primer [CCP]), and MDP-contain-
ing universal adhesives (Single Bond Universal [SBU]; Clearfil Universal Bond [CUB]). For the Ctr, ZP, CCP, SBU,
and CUB groups, the Y-TZP plates were bonded with MDP-free composite cement. For the remaining two groups,
each Y-TZP plate was bonded with one of MDP-containing self-adhesive composite cements (CSL; PSLP). The speci-
mens in each group (n = 30) were divided into two subgroups, with one subgroup of samples subjected to 24-h
water storage and the other to aging through 30,000 thermocycles plus 180-day water storage. Shear bond
strength (SBS) was measured. For each conditioning method, inductively coupled plasma-mass spectrometry (ICP-
MS) was used to quantify the release of phosphorus, and microleakage was evaluated through a methylene-blue
dyeing technique.
Results: The Ctr group exhibited the lowest SBS regardless of aging. Thermocycling and water storage significantly
decreased the SBS in all groups, with the exception of the groups with the MDP-containing zirconia primers. Groups
with MDP-containing universal adhesives and self-adhesive composite cements presented brand-dependent higher
SBS, even after aging. ICP-MS detected phosphorus release from all of the MDP-containing products. Microleakage
was identified at the interfaces of all bonded specimens, with the highest rate detected in the Ctr group.
Conclusion: MDP-containing primers, universal adhesives, and composite cements create bonds to zirconia with
acceptable strength after long-term aging.
Keywords: yttrium stabilized tetragonal zirconia polycrystals, phosphate ester monomer, MDP, bonding, primer,
microleakage, ICP-MS, shear bond strength.

J Adhes Dent 2018; 20: 279–287. Submitted for publication: 14.03.18; accepted for publication: 11.06.18
doi: 10.3290/j.jad.a40989

a Postgraduate Student, Jiangsu Key Laboratory of Oral Diseases, Department

of Prosthodontics, Affiliated Hospital of Stomatology, Nanjing Medical Univer-
sity, Nanjing, China. Wrote the manuscript, performed the experiments, per-
formed statistical evaluation; first and second authors contributed equally to
this work.
b Associate Professor, Jiangsu Key Laboratory of Oral Diseases. Department of
Prosthodontics, Affiliated Hospital of Stomatology, Nanjing Medical University,
Nanjing, China. Idea, experimental design, wrote the manuscript.
c Postgraduate Student, Jiangsu Key Laboratory of Oral Diseases, Department
of Prosthodontics, Affiliated Hospital of Stomatology, Nanjing Medical Univer-
sity, Nanjing, China. Performed the experiments and statistical evaluation.
d Associate Professor, Jiangsu Key Laboratory of Oral Diseases, Department of
Endodontics, Affiliated Hospital of Stomatology, Nanjing Medical University,
Nanjing, China. Idea, hypothesis, experimental design, proofread the manu-
script.
Correspondence: Dr. Chen Chen, Han-Zhong Road 136th, Stomatological Hos-
pital of Jiangsu Province, Nanjing 210029, China, Tel: +86-25-8503-1822;
e-mail: ccchicy@njmu.edu.cn
A definitive approach for optimal adhesive bonding to yt-
trium stabilized tetragonal zirconia polycrystals (Y-TZP)
has yet to be found, and it remains an area which gener-
ates great research interest. It has been widely reported
that the application of products containing the phosphate
ester monomer 10-methacryloyloxydecyl dihydrogen phos-
phate (MDP) after surface roughening results in better
Y-TZP bonding.14,23 MDP conditioning is nondestructive
and serves to activate or functionalize the inert Y-TZP sur-
face, rendering it available for adhesive chemical bond-
ing.13 Traditionally, MDP has been incorporated into prim-
ers to pre-treat metal oxides such as zirconia before using
composite cements. Following the ongoing trend among
manufacturers to further simplify bonding technology to
satisfy clinicians’ demand for adhesive procedures that
are faster, less technique-sensitive and more user-

Vol 20, No 4, 2018 279

Yang et al
Table 1 Descriptive details of primers, adhesives, and resin cements used in each group
Groups Primers Universal adhesives
Ctr \ \
ZP MDP-containing/Z-Prime Plus \
CCP MDP-containing/Clearfil Ceramic Primer \
SBU \ MDP-containing/Single Bond
Universal*
CUB \ MDP-containing/Clearfil Universal Bond
CSL \ \
PSLP \ \
* Also known as Scotchbond Universal. **Also known as Panavia SA Cement or Panavia SA Luting.
friendly,7 multipurpose products such as universal adhe-
sives and self-adhesive composite cements have been
developed that are comprised of multiple functional com-
ponents such as MDP and silane. These products facili-
tate the bonding of Y-TZP to either dentin or enamel with-
out the need for pre-priming.7,29
According to several previous studies,3,13,27 composite
cements and primers that contain MPD produced much
stronger and more durable bonds to air-abraded zirconia
than when tribochemical silica coating and silanization
was used, and provided long-term durable bonding of non-
retentive zirconia restorations to enamel.15 Nevertheless,
disagreement remains about the bond durability of MDP-
containing products. Some researchers have reported that
application of MDP-containing primers, universal adhe-
sives, or composite cements on sintered Y-TZP failed to
sustain stable bond strength after artificial aging.2,6,10,28
As universal adhesives and self-adhesive cements are in-
spired by a multipurpose design approach, which em-
braces the versatility of functional components, it remains
unknown whether or not the chemical affinity of MDP to
Y-TZP is compromised by the other functional components
due to competing physicochemical characteristics or by
the varied chemical environments that enable their deliv-
ery. Moreover, a recent study revealed that hydrolysis of
the coordinate bond between MDP and zirconia weakens
the bond integrity between composite and MDP-condi-
tioned Y-TZP.8 Therefore, the question remains as to
whether MPD-containing composite cements and universal
adhesives can provide sufficient, uncompromised bond
durability to zirconia. Further research on the bonding of
MDP-containing adhesives to zirconia is needed to evalu-
ate their performance.
Accordingly, this study evaluated three types of MDP-
containing products (primer, universal adhesive, self-adhe-
sive composite cement) by assessing their bond strength
before and after long-term water storage and thermocycling.
Furthermore, microleakage and MDP to Y-TZP bond hydroly-
280
Cements
MDP-free composite cement/RelyX
Veneer
MDP-containing/Clearfil SA composite
cement**
MDP-containing cement/Panavia SA
Luting Plus
sis were determined. The null hypotheses tested were:
there is no difference in the bond durability, microleakage,
and bonding hydrolysis between MDP and Y-TZP, irrespec-
tive of which MDP-containing product was used.
MATERIALS AND METHODS
Shear Bond Strength
210 industrially manufactured Y-TZP plates (94% ZrO2, 6%
Y2O3, Shenzhen Santo Industry Technology Development;
ShenZhen, China), with dimensions of 10 × 10 × 1.5 mm3,
were subjected to airborne-particle abrasion with 50-μm alu-
mina particles from a distance of 10 mm for 20 s at
0.25 MPa (JNBP-2, Jianian Futong Medical Equipment; Tian-
jin, China). The 210 Y-TZP plates were randomly assigned
to seven groups (n = 30 per group), according to the condi-
tioning methods applied (Table 1):
Ninety air-abraded Y-TZP plates received no further pre-
treatment and were classified into CSL, PSLP (defined
below), and control (Ctr) groups.
MDP-containing primer, either Z-Prime Plus (ZP, Bisco;
Schaumburg, IL, USA) or Clearfil Ceramic Primer (CCP, Kura-
ray Noritake; Okayama, Japan), was applied to 60 air-
abraded Y-TZP plates according to the manufacturer’s in-
structions.
A coat of MDP-containing universal adhesive, Single
Bond Universal (SBU, 3M Oral Care; St Paul, MN, USA) or
Clearfil Universal Bond (CUB; Kuraray Noritake), was ap-
plied on another 60 air-abraded Y-TZP plates, followed by
15–20 s volatilization, air drying, and 10 s light polymeriza-
tion using an LED lamp (1000 mW/cm2, Elipar FreeLight 2,
3M Oral Care).
Pre-polymerized composite cylinders (n = 210; 6 mm
inner diameter; 3 mm height) were made from a light-curing
composite (Valux Plus, 3M Oral Care).
For the Ctr, ZP, CCP, SBU, and CUB groups, pre-polymer-
ized composite cylinders were cemented on pre-treated
The Journal of Adhesive Dentistry
 Yang et al

Table 2 Materials used in the shear bond strength tests

Material/Trade name Main composition Lot Manufacture

Y-TZP 94% ZrO2, 6% Y2O3 616087 ShenZhen;
GuangDong, China

Light-polymerized cement/ TEG-DMA, bis-GMA, 66 wt% (47 vol%) zirconia/silica N707003N711885 3M Oral Care;
Relyx Veneer St Paul, MN, USA

Light-polymerized composite/ Barium aluminum fluoride glass + highly dispersive N794469

Valux Plus silica (80–90%), bis-GMA (5–10%) N775044
N757874

Zirconia adhesive/ Vitrebond copolymer, MDP, silane 595819

Single Bond Universal*

Zirconia primer/ Ethanol (<90%), BPDM (<10%); HEMA (<20%), MDP 1600002233 Bisco; Schaumburg,
Z-Primer plus IL, USA

Zirconia primer/ Ethanol (>80%), TMSPMA, MDP C30008 Kuraray Noritake

Clearfil Ceramic Primer Medical; Okayama,
Japan
Zirconia adhesive/ MDP, HEMA, bis-GMA, hydrophilic aliphatic 1L0003
Clearfil Universal Bond dimethacrylate, silane coupling agent

Zirconia cement/ Bis-GMA, TEG-DMA, MDP, dimethacrylate, silanated 320089

Clearfil SA Luting** barium glass filler, silanated colloidal silica

Zirconia cement/ MDP, bis-GMA, TEG-DMA, hydrophobic aromatic 7R0067

Panavia SA Luting Plus dimethacrylate, HEMA, silanated barium glass filler,
silanated colloidal silica

* Also known as Scotchbond Universal. **Also known as Panavia SA Cement or Panavia SA Luting. TEG-DMA: triethyleneglycol dimethacrylate; bis-GMA: bi-
sphenol A glycidyl dimethacrylate; MDP: 10-methacryloyloxydecyl dihydrogen phosphate; BPDM: biphenyl dimethacrylate; HEMA: 2-hydroxyethyl methacrylate;
TMSPMA: 3-trimethoxysilylpropylmethacrylate.

Y-TZP plates by a layer of MDP-free composite cement
(RelyX Veneer, 3M Oral Care) under a constant load of
20 N. After removal of excess cement with a probe, the
cylinders were light cured for 40 s at four different locations
on both sides, to ensure complete polymerization of the
composite cement.
For the CSL and PSLP groups, the MDP-containing com-
posite cements Clearfil SA Luting (CSL, Kuraray Noritake)
and Panavia SA Luting Plus (PSLP, Kuraray Noritake), re-
spectively, were used to cement the pre-polymerized com-
posite cylinders.
All bonded specimens were assigned to two aging condi-
tions: distilled water storage at 37ºC for 24 h and artificial
aging by 30,000X thermocycling (TC-501F, Suzhou Weier
Labware; Suzhou City, China) between two water baths set
to 5°C and 55°C, with a dwell time of 15 s at each tem-
perature, followed by 180 days of distilled water storage at
37°C.
All bonded specimens were then subjected to shear
bond strength (SBS) tests at a crosshead speed of
1.0 mm/min using a universal testing machine (Instron
Model 3365, ElectroPuls; Norwood, MA, USA). The number
of prematurely debonded specimens was recorded and their
SBS values were recorded as 0 MPa.19,25
After satisfying the normality and homogeneity of vari-
ance assumptions of the datasets, the means of each
group were submitted to two-way ANOVA and post-hoc
Tukey’s HSD test to examine the influences of the indepen-
dent surface treatments and aging. These analyses were
performed using the statistical software SPSS 22.0 (SPSS;
Chicago, IL, USA). Statistical significance was set at
α = 0.05.
All debonded specimens were examined at 8X magnifica-
tion under a stereomicroscope (C-DSS230, Nikon; Tokyo,
Japan). When the residual composites were difficult to dis-
tinguish from one another, a higher magnification was used
or SEM (Carl Zeiss EVO18; Jena, Germany) was employed.
Fracture modes were classified as adhesive (fracture sites
entirely located between the composite cement and zirconia
surface), cohesive (fractures occurring exclusively within the
composite or cement), or mixed (partial composite cement
fractures and partial Y-TZP surface exposure).
Details of the materials used in the SBS tests are de-
tailed in Table 2.
Preparation of Bilayered Y-TZP Bonded Specimens
Another batch of air-abraded Y-TZP plates
(10 × 10 × 1.5 mm3) was prepared and divided into seven
groups (n = 11), as for similar SBS testing, to receive sur-
face treatments. Two identically pre-treated Y-TZP plates in
each group were bonded to each other with a layer of com-
posite cement, either RelyX Veneer, Clearfil SA Luting, or
Panavia SA Luting Plus. Before the final light curing, four
sides of each specimen were protected with an oxygen-
blocking gel. One of the four sides was then successively
wet-polished with silicon carbide papers of increasing fine-

Vol 20, No 4, 2018 281

Yang et al
Bi-layered specimen
Fig 1 Each bilayered specimen was separated by shear force using pliers.
ness (800-, 1000-, 1200-, 1500-, 2000-, 2500-, 3000-, and
4000-grit) using a metallographic polisher (PG-1, BiaoYu
Instrument; ShangHai, China), to completely expose the
bonding interface. The other three sides were painted with
nail polish to seal the interfaces.
Microleakage Evaluation
Bilayered specimens from all groups were assigned to two
aging-condition groups: 24-h water storage at 37°C (n = 5)
or aging by 30,000X thermocycling and 180-day water stor-
age at 37ºC (n = 5), similar to SBS testing.
The specimens were then immersed in 2% methylene
blue for 48 h, followed by rinsing with running water for 2 h.
Each bilayered specimen was separated by shear force
using pliers (Fig 1). Each debonded surface was imaged at
8X magnification under a stereomicroscope. Stained areas
along the interface were quantified by Photoshop CS6 v
13.0 (Adobe Systems; San Jose, CA, USA).
For each group, two subgroups that were subjected to
the two aging conditions were included. Comparison of the
leakage among the seven groups was performed by two-way
parametric ANOVA (independent variables were surface
treatments and aging) and post-hoc Tukey’s HSD test after
ascertaining normality and homoscedasticity of variance of
the datasets. Statistical analysis was set to α = 0.05.
Inductively Coupled Plasma-Mass Spectrometry
(ICP-MS)
One randomly selected bilayered specimen from each group
was stored in distilled water at 37ºC for 24 h. After washing
with running deionized water for 2 h, the specimens were then
dropped into a separate container filled with deionized water.
All of the three kinds of materials contained in the bilay-
ered specimen were fabricated into plate shapes, including
an air-abraded Y-TZP plate, a pre-polymerized composite
plate (Valux Plus), and three different brands of pre-poly-
merized cement plates (Relyx Veneer, Clearfil SA Luting,
282
and Panavia SA Luting Plus), and then placed in separate
deionized water containers to serve as the controls.
ICP-MS (ICP-MS 7500ce, Agilent; Santa Clara, CA, USA)
was used to quantify the phosphorus content in the soaking
liquids from each group, after soaking for 1, 7, 14, 30, 60,
and 90 days. Each measurement was repeated three times.
RESULTS
Shear Bond Strength
Two-way ANOVA and post-hoc tests of the SBS data re-
vealed that SBS was statistically significantly affected by
the surface treatments and aging (p < 0.01), while there
were no significant interactions between surface treatment
and aging (p = 0.153). Means and SD of SBS are detailed
in Fig 2 and Table 3.
The Ctr group, where no MDP-containing product was
used, presented the lowest mean SBS, regardless of
aging conditions. The PSLP group exhibited the highest
24-h SBS, followed by the SBU, CSL and CUB groups.
However, groups that contained MDP-containing primers
exhibited comparatively lower SBS. After thermocycling
and water storage, SBS decreased significantly
(p < 0.05) in groups using MDP-containing composite ce-
ments and universal adhesives (SBU, CUB, CSL, and
PSLP groups). The control group, in which MDP-containing
products were not used, showed a dramatic decrease in
bond strength, and presented 7 prematurely debonded
specimens. However, there were no significant SBS de-
creases (p > 0.05) observed for groups that used MDP-
containing primers (ZP and CCP groups). The PSLP group
still yielded the highest SBS after aging, followed by
groups using MDP-containing universal adhesives and
MDP-containing primers.
Failure modes are shown in Fig 3 and Table 3. Typical
SEM images of bonding interfaces are shown in Fig 4. The
The Journal of Adhesive Dentistry
 Yang et al

24-h water storage 24-h water storage 30,000 thermocycles plus

30,000 thermocycles plus 180 d water storage
Mpa 180 d water storage Adhesive Adhesive
16 e Mixed Mixed
* * c D 100%
14 cd
Shear Bond Strength (SBS)

bd
b C
b B
12 B B
80%

Failure percentage
B
10

8 60%

6 a
40%
4

2 A 20%

0
0
Ctr ZP CCP SBU CUB CSL PSLP Ctr ZP CCP SBU CUB CSL PSLP
Groups Groups

Fig 2 Group means and standard deviations of SBS. Different let- Fig 3 Failure modes observed in all groups after SBS testing. Ad-
ters indicate statistically significantly different group means hesive: failure at the ceramic surface; mixed: combination of adhe-
p < 0.05); lowercase letters refer to 24-h water storage groups, sive failure at ceramic surface and cohesive failure in luting
while the uppercase letters refer to groups aged with 30,000 ther- composite or composite cylinder.
mocycles and 180-day water storage. Asterisks mark SBS that was
not significantly different between 24-h water storage and after
aging (p < 0.05).

control group presented the most adhesive fractures Table 3 Group means and standard deviations of the
(73.4%) after 24-h water storage; the remaining 26.6% were SBS and failure mode results
mixed failures. After aging, all Ctr specimens exhibited ad-
hesive fractures (100%). In contrast, mixed failure was the Group Aging Failure
SBS (Mpa) percentage (%)
only failure mode observed in the experimental groups after
24-h water storage. After aging however, the ZP, CCP, and Means SDs Adhesive mixed
CUB groups presented adhesive failures in several speci-
Ctr Before 4.987 1.100 73.33 26.67
mens.
After 0.877 1.007 100 0
Microleakage
ZP Before 11.387 0.955 0 100
Microleakage was identified at the interfaces of all bonded
specimens, as shown by methylene blue dye penetration. After 10.575 1.206 6.67 93.33
Two-way parametric ANOVA showed that the extent of staining CCP Before 10.734 1.486 0 100
was significantly affected by the specific surface treatments
used (p < 0.001), but not by aging (p = 0.234). There were no After 10.245 1.306 6.67 93.33
significant interactions between surface treatment and aging SBU Before 13.719 1.215 0 100
(p = 0.914). The post-hoc test results showed the greatest
extent of microleakage in the control group (p < 0.001). How- After 11.822 0.960 0 100
ever, no significant differences were observed among the CUB Before 12.165 1.074 0 100
groups in which different MDP-containing products were ap-
After 9.887 0.977 13.33 86.67
plied (p > 0.05). The stained areas are shown in Fig 5.
CSL Before 12.946 0.826 0 100
ICP-MS After 10.887 1.106 0 100
According to the ICP-MS results (Fig 6), no phosphorus was
detected in the soaking solutions of the air-abraded Y-TZP PSLP Before 14.504 1.288 0 100
plate, the pre-polymerized composite plate, or the MDP-free After 13.338 1.127 0 100
composite cement plate, regardless of the timepoint of mea-
surement. Conversely, phosphorus was detected in all MDP-

Vol 20, No 4, 2018 283

Yang et al

Fig 4 Representative SEM images of frac-

tured bonding interfaces. A: adhesive fail-
A B ure (26X); B: mixed failure (26X).

24-h water storage

(x10-) 30,000 thermocycles plus
3.50 a 180 d water storage
3.25
3.00
2.75 a
2.50
Dying area (cm2)

2.25
2.00
1.75 b
b b
1.50 b
b b
1.25 b b b b
b b
1.00
0.75
0.50
0.25 Fig 5 Mean and standard deviation of
0.00 areas stained by methylene blue dye. Dif-
Ctr ZP CCP SBU CUB CSL PSLP ferent superscript letters represent statisti-
Groups cally significantly different group means
(p < 0.05).

0.30 1.9
0.28 1.8
1.7
0.26
1.6
Amount of P released (ppm)

0.24 1.5
Amount of P released (ppm)

0.22 1.4
0.20 1.3
1.2
0.18
1.1
0.16 1.0
0.14 0.9
0.12 0.8
0.7
0.10
0.6
0.08 0.5
0.06 0.4
0.04 0.3
0.2
0.02
0.1
0.00 0.0
-0.02 -0.1
24 h 1w 2w 4w 2m 3m 24 h 1w 2w 4w 2m 3m

Time Time

ZP CCP SBU Y-TZP Relyx Veneer Valux Plus

a CUB CSL PSLP b Panavia SA Luting Plus Clearfil SA Luting

Fig 6 Phosphorus release as detected by ICP-MS. a: bilayered specimens; b: the materials contained in the bilayered specimens (Y-TZP,
composite, and composite cements).

284 The Journal of Adhesive Dentistry


containing bilayered specimen solutions and the solutions of
the two MDP-containing composite cement plates. Moreover,
the variation of content suggests a time-dependent release
of phosphorus, as the concentration of phosphorus in-
creased with increased soaking time. This upward tendency
did not last when soaking lasted beyond 1 or 2 months for
the ZP, SBU, and CSL groups. Notably, continued release of
phosphorus was detected for up to 3 months for the two
MDP-containing composite cement plate solutions.
DISCUSSION
Although numerous studies have reported the effects of
MDP-containing products on bonding to zirconia when ap-
plied individually or in combination, comparisons of bond
strength, durability, and microleakage among different
types of MDP-containing products (primer/universal, adhe-
sive/self-adhesive, composite cement) are rare. In this
study, we performed dynamic monitoring of chemical bond
hydrolysis between MDP and zirconia for three different
types of MDP-containing products, which – to the best of
our knowledge – has not been presented previously. The
results described above demonstrate that there were dif-
ferences between MDP and Y-TZP bond durability, micro-
leakage, and bonding hydrolysis when different kinds of
MDP-containing products were used. Therefore, the null
hypothesis has to be rejected.
A commonly used SBS test was employed in the pres-
ent study to investigate the bonding performance of the
different kinds of MDP-containing products. Although a
combination of tribochemical silica coating and silaniza-
tion has proven to provide good bonding to Y-TZP,19
based on the purpose of this study, we chose air abra-
sion rather than tribochemical silica coating to obtain a
standardized pre-roughened Y-TZP surface prior to applica-
tion of MDP products. The reason for this was to avoid
potential interference due to chemical reactions between
MDP and silica, and chemical bonding between silica and
silane. Previous studies found the load for cementing the
composite/zirconia bonded specimen to vary from 4 N,
6 N, 7.35 N, 9.8 N, to 20 N.9,16,22,26,31 Because there is
still no standard for SBS testing of bonding to zirconia,
any of the above loads can be employed as long as the
same load is used for each bonded specimen. Pre-polym-
erized, 3-mm-high composite cylinders are somewhat too
thick for complete polymerization; however, 3 mm was
also a commonly used thickness in previous studies.1,21
Moreover, although the light-curing time has been proven
to affect composite cement polymerization, when suffi-
ciently long light-curing time was provided, adequate poly-
merization could be achieved even at very low light irradi-
ance. 17 The normal curing time of 20 s can attain a
clinically required degree of cure. In this study, prolonged
light-curing time was performed from different locations to
ensure complete polymerization of composite cement.
According to the SBS results, air abrasion alone was
not capable of producing reliable, strong bonds between
Vol 20, No 4, 2018
Yang et al
composite and Y-TZP. Indeed, the lowest SBS and bond
durability were observed for the control group. Notably,
premature debonding occurred after aging, with most of
the specimens in this group presenting pure adhesive fail-
ure. Conversely, the combination of air abrasion and any
of the MDP-containing primers/adhesives/composite ce-
ments improved the composite bond strength to Y-TZP,
suggesting MDP improved adhesion at the composite/Y-
TZP interface for all types of MDP-containing products this
chemical adhesion is necessary to obtain adequate bond
strength. Higher SBS were recorded for the multipurpose
products than for MDP-containing primers, which suggests
that the addition of other elements in addition to MDP
would at least not compromise the composite/Y-TZP cou-
pling ability of MDP. However, within the present study,
inconsistent bond durability was observed among different
brands or types of MDP-containing products. The groups
that used MDP-containing primers exhibited better aging
resistance than did groups employing other types of MDP-
containing products. Nevertheless, MDP-containing primer
is a single-purpose product focused on conditioning zirco-
nia and applied in a single step; its bond improvement
ability was not overwhelmingly superior to the multipur-
pose MDP-containing products such as universal adhesive
and self-adhesive composite cement without additional
prior priming. Therefore, the latter are more favorable for
clinical applications.
Associating MDP-containing primers or adhesives with
MDP-containing composite cements has been reported to
provide synergic effects and higher bond strength after
aging in some previous studies.7,21,29 However, based on
the aim of the present study, the contribution of MDP in
each type of MDP-containing product cannot be clearly
known when being used in combination. Therefore, the as-
sociation of these MDP-containing products was not inves-
tigated.
A stereomicroscope was used to examine the failure
modes. As the colors of the resin cement and composite
were the same, it was difficult to distinguish the residual
resins; in such cases, a higher magnification or SEM was
used. According to the failure mode analysis, all the MDP-
containing samples presented solely mixed failures after
24-h water storage, with the exception of Ctr group sam-
ples. However, 30,000x thermocycles and 180-day water
storage led to a low frequency of adhesive failures in the
two groups where primers were used and in the Clearfil
Universal Bond group, suggesting weakened bond quality.
The fact that solely adhesive failures were observed in the
Ctr group after aging indicates that using a conventional
MDP-free composite cement alone does not maintain sta-
ble bonding during aging. The failure modes and SBS re-
sults showed dependency on brands or types of MDP-con-
taining products.
In order to more accurately compare the bond durability
of the three MDP-containing products, leakage evaluation
and high-sensitivity ICP-MS were performed in the present
study. The purpose of the microleakage test is to deter-
mine whether water can penetrate the barrier and whether
285
Yang et al
there is a difference in the degree and extent of water
penetration into the bonding interface after application of
different types of MDP products. Our previous studies
have demonstrated that the chemical bond between MDP
and zirconia can be hydrolyzed;20 it weakened the bond
strength between composite and MDP-conditioned Y-TZP.8
More severe microleakage suggests deeper water infiltra-
tion. Hydrolysis of the chemical bond leads to more pro-
nounced bond strength weakening. Therefore, we hypoth-
esize that bond durability is related to the extent of
leakage. Previous studies have evaluated the leakage of
zirconia crowns by preparing complete cross sections, in
which dye penetration depths were identified along the ce-
ment/dentin interface and/or the cement/Y-TZP inter-
face.18,24 In the present study, leakage was identified at
the interface between composite and Y-TZP in all of the
bilayered specimens, which suggests that the composite/
Y-TZP interface contains hydrophilic materials that attract
water to the bonded interface. Air abrasion alone resulted
in the highest microleakage among the groups in this
study, regardless of aging. This result is consistent with
the SBS results and confirms that the absence of chem-
ical bonding results in the highest microleakage, and that
better water resistance can be achieved through chemical
adhesion by MDP coupling. Although no significant differ-
ences in microleakage were found between the MDP-con-
taining products, regardless of aging, it is reasonable to
conclude that the MDP-containing products influenced the
hydrophilicity of the bonding interface by preventing micro-
leakage, when compared to bilayered specimens in which
no MDP-containing product was used.
Based on the similarity in the present results, it must be
acknowledged that microleakage results might not precisely
reflect the quality of the marginal seal of a bonding inter-
face. However, a potential release pathway of phosphorus
from the bonding interface can still be proposed by analyz-
ing the results of the microleakage test. In the present
study, all the materials in the bilayered specimens, as well
as the bilayered specimens in which different bonding strat-
egies were used, were soaked individually for phosphorous
detection according to the same method employed in a pre-
vious study.8 As elemental phosphorus was only present in
MDP molecules and was absent from other bonding mater-
ials, the release of phosphorus can be attributed to unre-
acted MDP and/or degradation of the MDP-zirconia co-ordi-
nate (dative covalent) bond.8 Phosphorus detection in a
soaking solution implies that MDP molecules or their mo-
lecular fragments containing phosphorus were released dur-
ing water immersion. As deionized water soaking solution
for each group was subjected to ICP-MS, the specimens
were washed under running deionized water for 2 h before
analysis, in order to completely remove any potential con-
tamination through water-soluble phosphorus. The present
ICP-MS results were reasonable, since these MDP-free ma-
terials release no phosphorous. Moreover, no phosphorus
was detected in the deionized water soaking solutions of
any of the phosphorus-free materials, which suggested that
there was no phosphorus in the deionized water. Therefore,
286
it was not necessary to set deionized water as a separate
negative control group. According to the presented ICP-MS
results, for the plates of two brands of MDP-containing com-
posite cement (Clearfil SA Luting and Pananvia SA Luting
Plus) and the bilayered specimens that employed these two
composite cements, phosphorus was released into their
soaking liquids within 24 h, which can be attributed to
water sorption and hydrolytic degradation of the composite
cements.5,12,30
The marginal gaps within zirconia ceramic crowns vary
from between 7 and 529 μm,11 with these areas of bond-
ing interface exposed to gingival crevicular fluid and/or
moisture in the oral environment. The composite cements
located at the margin of the bonding interface are inevita-
bly exposed to moisture, which leads to hydrolysis of com-
posite cements and, in turn, release of phosphorus. This
explains why ICP-MS detected phosphorus after 24 h in
both of the two MDP-containing composite cement groups.
Phosphorus was detected after one or more weeks in the
MDP-containing primer and adhesive groups. It is impos-
sible that all the MDP formed a co-ordinate bond with zir-
conia; therefore, unreacted MDP molecules still exist in
the primer or adhesive-conditioned Y-TZP surfaces. Never-
theless, since the application of either MDP-containing
primers or adhesives must be combined with the use of a
composite cement, MDP at the primer (adhesive)/compos-
ite cement/Y-TZP interface, whether involved in a co-ordi-
nate bond or unreacted, was covered by the MDP-free
composite cement. When the composite cement was de-
graded by water, the inner unreacted MDP would have
been exposed, which would likely have led to the release
of phosphorus into the solution. Moreover, thermodynamic
calculations have shown the bond integrity between MDP
and zirconia to be susceptible to water aging via hydrolysis
of the Zr-O-P bond, which leads to the release of MDP mol-
ecule fragments containing phosphorus into the solution.8
Interestingly, however, no phosphorus was detected after
24 h in the MDP-containing primer or universal adhesive
bonding groups. It may be assumed that the superficial
composite cement covering the unreacted MDP molecules
was not completely hydrolytically degraded, so that these
unreacted MDP molecules were not exposed in the short
term. In the present study, a positive correlation between
time and phosphorus concentration release was observed.
Whether the time-dependent release of phosphorus was
caused by hydrolysis of the composite cements to expose
the inner unreacted MDP or by hydrolysis of the coordinat-
ing bonds between MDP and zirconia, could not be distin-
guished, which is a limitation of this study.
CONCLUSIONS
MDP-containing universal adhesives and self-adhesive ce-
ments presented higher bond strength to Y-TZP than did
MDP-containing primers, although the latter presented bet-
ter bond durability. No MDP-containing product prevented
water from infiltrating into the composite/Y-TZP interface.
The Journal of Adhesive Dentistry

ACKNOWLEDGMENTS
The authors thank Mr. Taoran Ma and Dr. Hao Wang (Kuraray) for
providing Clearfil Ceramic Primer and Panavia SA Luting Plus.
This study was supported by the National Natural Science Foun-
dation of China (grant 81400539); Natural Science Foundation of
Jiangsu Province (grants BK20150998 and BK20140913), and
Jiangsu Higher Education Institutions (grants 15KJB320003 and
2014-37).
REFERENCES
1. Al-Ansari A, Al-Harbi F, Baba NZ. In vitro evaluation of the bond strength
of composite resin foundation materials to dentin. J Prosthet Dent
2015;114:529-535.
2. Aboushelib MN, Mirmohamadi H, Matinlinna JP, Kukk E, Ounsi HF, Sal-
ameh Z. Innovations in bonding to zirconia based materials. Part II: Fo-
cussing on chemical interactions. Dent Mater 2009;25:989-993.
3. Attia A, Kern M. Long-term resin bonding to zirconia ceramic with a new
universal primer. J Prosthet Dent 2011;106:319-327.
4. Breschi L, Cammelli F, Visintini E, Mazzoni A, Vita F, Carrilho M, et al. In-
fluence of chlorhexidine concentration on the durability of etch-and-rinse
dentin bonds: a 12-month in vitro study. J Adhes Dent 2009;11:191-198.
5. Bagheri R, Mese A, Burrow MF. Comparison of the effect of storage
media on shear punch strength of resin luting cements. J Dent 2010;38:
820-827.
6. Bömicke W, Schürz A, Krisam J, Rammelsberg P, Rues S. Durability of
Resin-Zirconia Bonds Produced Using Methods Available in Dental Prac-
tice. J Adhes Dent 2016;18:17-27.
7. Chen C, Niu LN, Xie H, Zhang ZY, Zhou LQ, Jiao K, et al. Bonding of uni-
versal adhesives to dentine-Old wine in new bottles? J Dent 2015;43:
525-536.
8. Chen C, Chen Y, Lu Z, Qian M, Xie H, Tay FR. The effects of water on deg-
radation of the zirconia-resin bond. J Dent 2017;64:23-29.
9. Chen Y, Lu Z, Qian M, Zhang H, Chen C, Xie H. Chemical affinity of
10-methacryloyloxydecyl dihydrogen phosphate to dental zirconia: Effects
of molecular structure and solvents. Dent Mater 2017;33:e415-e427.
10. de Sá Barbosa WF, Aguiar TR, Francescantonio MD, Cavalcanti AN, de
Oliveira MT, Giannini M. Effect of water storage on bond strength of self-
adhesive resin cements to zirconium oxide ceramic. J Adhes Dent
2013;15:145-150.
11. Fleming GJ, Dobinson MM, Landini G, Harris JJ. An in-vitro investigation
of the accuracy of fit of Procera and Empress crowns. Eur J Prosthodont
Restor Dent 2005;13:109-114.
12. Ito S, Hashimoto M, Wadgaonkar B, Svizero N, Carvalho RM, Yiu C, et al.
Effects of resin hydrophilicity on water sorption and changes in modulus
of elasticity. Biomaterials 2005;26:6449-6459.
13. Kern M, Wegner SM. Bonding to zirconia ceramic: adhesion methods and
their durability. Dent Mater 1998;14:64-71.
14. Klink A, Hüttig F. Zirconia-based anterior resin-bonded single-retainer
cantilever fixed dental prostheses: a 15- to 61-month follow-up. Int J
Prosthodont 2016;29:284-286.
15. Kern M. Bonding to oxide ceramics—laboratory testing versus clinical
outcome. Dent Mater 2015;31:8-14.
Vol 20, No 4, 2018
Yang et al
16. Lüthy H, Loeffel O, Hämmerle C. Effect of thermocycling on bond strength
of luting cements to zirconia ceramic. Dent Mater 2006;22:195-200.
17. Musanje L, Darvell BW. Polymerization of resin composite restorative ma-
terials: exposure reciprocity. Dent Mater 2003;19:531-541.
18. Owittayakul D, Lerdrit W, Pittayachawan P. Microleakage of zirconia
frameworks cemented with two types of phosphate monomer-based resin
cements. Eur J Gen Dent 2015;4:92-97.
19. Passos SP, May LG, Barca DC, Özcan M, Bottino MA, Valandro LF. Con-
ventional adhesive resin cement to y-tzp ceramic before and after aging
conditions. Oper Dent 2010;35:689-696.
20. Qian M, Lu Z, Chen C, Zhang H, Xie H. Alkaline nanoparticle coatings im-
prove resin bonding of 10-methacryloyloxydecyldihydrogenphosphate-con-
ditioned zirconia. Int J Nanomed 2016;11:5057-5066.
21. Sanohkan S, Kukiattrakoon B, Larpboonphol N, Sae-Yib T, Jampa T,
Manoppan S. The effect of various primers on shear bond strength of zir-
conia ceramic and resin composite. J Conserv Dent 2013;16:499-502.
22. Souza GD, Hennig D, Aggarwal A, Tam LE. The use of MDP-based mater-
ials for bonding to zirconia. J Prosthet Dent 2014;112:895-902.
23. Sasse M, Kern M. Survival of anterior cantilevered all-ceramic resin-
bonded fixed dental prostheses made from zirconia ceramic. J Dent
2014;42:660-663.
24. Sener I, Turker B, Valandro LF, Özcan M. Marginal gap, cement thick-
ness, and microleakage of 2 zirconia crown systems luted with glass ion-
omer and MDP-based cements. Gen Dent 2014;62:67-70.
25. Souza LC, Apolonio FM, Souza LP, Carrilho MRO, Saboia VPA. Stability of
the Adhesive Interface after Different Artificial Aging Techniques. J Dent
App 2015;2:153-157.
26. Vechiato-Filho AJ, Matos AO, Landers R, Goiato MC. Surface analysis and
shear bond strength of zirconia on resin cements after non-thermal
plasma treatment and/or primer application for metallic alloys. Mat Sci
Eng C 2017;72:284-292.
27. Wegner SM, Kern M. Long-term resin bond strength to zirconia ceramic.
J Adhes Dent 2000;2:139-147.
28. Wille S, Lehmann F, Kern M. Durability of resin bonding to lithium disili-
cate and  zirconia  ceramic using a self-etching primer. J Adhes Dent
2017;19:491-496.
29. Xie H, Li Q, Zhang F, Lu Y, Tay FR, Qian M, Chen C. Comparison of resin
bonding improvements to zirconia between one-bottle universal adhe-
sives and tribochemical silica coating, which is better? Dental Mater
2016;32:403-411.
30. Yiu CK, King NM, Carrilho MR, Sauro S, Rueggeberg FA, Prati C, Carvalho
RM,Pashley DH, Tay FR. Effect of resin hydrophilicity and temperature on
water sorption of dental adhesive resins. Biomaterials 2006;27:
1695-1703.
31. Yang B, Lange-Jansen HC, Scharnberg M, Wolfart S. Influence of saliva
contamination on zirconia ceramic bonding. Dent Mater 2008;24:
508-513.
Clinical relevance: MDP-containing universal adhesives
and self-adhesive cements provide acceptable bond du-
rability to zirconia without pre-priming.
287