d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 508–513

available at www.sciencedirect.com

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

Influence of saliva contamination on zirconia

ceramic bonding

B. Yang a , H.C. Lange-Jansen b , M. Scharnberg b , S. Wolfart a , K. Ludwig a ,

R. Adelung b , M. Kern a,∗
a Department of Prosthodontics, Propaedeutics and Dental Materials, Christian-Albrechts University at Kiel,
Arnold-Heller-Str. 16, 24105 Kiel, Germany
b Multicomponent Materials, Faculty of Engineering, Christian-Albrechts University at Kiel, Kaiser Str. 2, 24143 Kiel, Germany

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

Article history: Objectives. The purpose of this study was to investigate the influence of saliva contamination
Received 6 September 2006 and cleaning methods on adhesive bonding to dental zirconia ceramic with a phosphate-
Accepted 5 April 2007 monomer-containing luting resin.
Methods. After saliva immersion, airborne-particle abraded ceramic specimens were cleaned
with water rinsing, with isopropanol, with phosphoric acid gel, or with additional airborne-
Keywords: particle abrasion. Airborne-particle abraded specimens without contamination were used
Zirconia ceramic bonding as the control group. Chemical analysis of the ceramic surfaces of all groups was done
Saliva contamination using X-ray photoelectron spectroscopy (XPS). The influence of contamination and cleaning
Cleaning methods on ceramic bond durability was examined by tensile bond strength (TBS) testing
after 3 days or 150 days water storage with 37,500 thermal cycles.
Results. After saliva contamination XPS revealed an organic coating which was not removed
completely with water rinsing, with isopropanol, or with phosphoric acid. Using TBS testing
a strong influence of contamination and cleaning methods on resin bond strength and its
durability was found.
Significance. Saliva contamination significantly affected resin bonds to zirconia ceramic and
its durability. Airborne-particle abrasion was the most effective cleaning method.
© 2007 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

1. Introduction zirconia ceramic restorations in most cases [2]. Phosphate-

Recently, various types of zirconia ceramic restorations
including fixed partial dentures (FPDs), resin-bonded FPDs,
full-coverage crowns, implant abutments, and endodontic
posts, have been widely used clinically due to zirconia’s high
fracture strength and esthetic properties [1]. A strong, durable
resin–ceramic bond provides all-ceramic restorations with
high retention, improved marginal adaptation and increased
fracture resistance of the restored teeth and the restorations,
although conventional cements can also be used for luting
monomer-containing composite resins in particular showed
a long-term durable bond to zirconia ceramic after airborne-
particle abrasion [3,4].
However, a good resin–ceramic bond obtained in a strictly
controlled clean situation in vitro might be compromised in
clinical situations, leading to a significantly reduced bond.
During the try-in procedure of the restoration in the oral situ-
ation, the contamination of the inner surface by saliva, blood
or silicone indicators is very difficult to avoid [5]. Saliva con-
tamination is frequently one main reason for decreased resin

∗
Corresponding author. Tel.: +49 431 5972874; fax: +49 431 5972860.
E-mail address: mkern@proth.uni-kiel.de (M. Kern).
0109-5641/$ – see front matter © 2007 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.dental.2007.04.013

Descargado para Anonymous User (n/a) en Universidad El Bosque de ClinicalKey.es por Elsevier en agosto 26, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 508–513 509

bond strength [5–9]. In dental textbooks, organic solutions are

often recommended for the removal of saliva contamination
2. Materials and methods
on the internal surfaces of restorations before cementation
[10]. In the instructions of modern adhesive composite resins,
2.1. Specimen preparation
phosphoric acid gel is recommended to remove contaminants
Densely sintered partially stabilized zirconia ceramic disk-like
from the inner surface of restorations (Kuraray Medical, Inc.,
specimens (Cercon, DeguDent, Hanau, Germany) with a thick-
Osaka, Japan).
ness of 3.4 mm and a diameter of 8.5 mm were used in this
Composite resins containing an adhesive phosphate-
study. The bonding surfaces of the disks were polished with
monomer (10-methacryloyloxy-decyl dihydrogenphosphate,
600 grit silicon carbide papers under water rinsing and then
MDP) showed a long-term durable bond to airborne-particle
airborne-particle abraded with 50 ␮m Al2 O3 at 2.5 bar pres-
abraded zirconia ceramic after water storage for 150 days
sure for 15 s at a distance of 10 mm. Afterwards the surfaces
[11,12] and 2 years [4] with repeated thermal cycling. How-
were air-cleaned for 20 s with the Rocatector delta device (3M
ever, the effect of saliva contamination and different cleaning
Espe, Seefeld, Germany). For XPS examination specimens were
methods on the zirconia ceramic surface chemistry and
cleaned ultrasonically in a distilled water bath for 10 min.
how it might influence the long-term durability of the resin
Then, the cleaned specimens were immersed in saliva for
bond of MDP containing resins, has not been evaluated yet
1 min. Saliva was collected from one healthy female donor
[13].
who had refrained from eating and drinking 1.5 h prior to
Therefore in the current study, X-ray photoelectron spec-
the collection procedure. All experiments were performed
troscopy (XPS), a highly surface sensitive technique for
using fresh saliva collected on the same occasion. After saliva
determining the chemical composition of multiphase com-
immersion, the specimens were cleaned according to tested
pounds and for detecting surface contaminants [14–16], was
cleaning methods generally available in dental offices (Fig. 1):
used to identify the existence of saliva contamination on
zirconia ceramic surface after try-in simulation with saliva
immersion. In addition, the influence of long-term water stor- • (WA) rinsed with only tap water for 15 s and air-dried for
age and thermal cycling on the bonding durability of MDP 15 s;
monomer-containing composite resin to zirconia ceramic • (AL) immersed in 70% isopropanol for 2 min and then rinsed
after contamination and the use of different cleaning methods with tap water for 15 s;
was investigated in this study. • (PA) etched with phosphoric acid gel (K Etchant agent)
The null hypotheses tested were that there is no saliva (Kuraray Medical, Inc., Osaka, Japan) for 30 s twice and then
contamination left on the ceramic surface after try-in simula- rinsed for 30 s;
tion, and there is no influence of contamination and cleaning • (AA) airborne-particle abraded with 50 ␮m Al2 O3 at 2.5 bars
methods on zirconia ceramic bonding with a MDP monomer- pressure for 15 s at a distance of 10 mm and then cleaned
containing composite resin. with compressed air;

Fig. 1 – Study design for chemical analysis and bond strength testing.

Descargado para Anonymous User (n/a) en Universidad El Bosque de ClinicalKey.es por Elsevier en agosto 26, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
510 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 508–513

• (CL) clean airborne-particle abraded specimens not treated

Table 1 – Mean ratios of carbon (C), oxygen (O) and
with saliva as control group. nitrogen (N) elements in groups
Groups C/O C/Zr O/Zr N/Zr
2.2. XPS examination
CL 0.2 1.4 6 0
Specimens of the above test groups were examined with XPS WA 2.9 47.6 16.6 3
to identify the contamination chemically and to evaluate the AL 1.2 9.9 8.1 2.2
effectiveness of cleaning methods. All measurements were PA 0.4 3.2 7 0
AA 0.7 3.9 5.8 0
done using a XPS system (Vacuum Generators, England) with
an X-ray source providing Al K␣ X-rays and kinetic energy Group codes see Fig. 1
of 1486.6 eV. A hemispherical analyzer with five channeltrons
was used for detection of the photoelectrons. A pass energy
setting of 100 allowed adequate quantitative analysis. In order WA), the C/Zr, O/Zr and N/Zr ratios drastically increased indi-
to examine the surface composition of the test specimens high cating that the ceramic surface was covered with an organic
resolution scans of the carbon (C 1s), oxygen (O 1s), zirconium coating, mainly composed of carbon (C), oxygen (O) and nitro-
(Zr 3d) and nitrogen (N 1s) peaks were taken. gen (N) compared to group CL. After cleaning with alcohol
(group AL), N still remained on the surface, and the C/Zr ratio
2.3. Tensile bond strength (TBS) testing was lower than water cleaned specimens, but higher than
those specimens cleaned by phosphoric acid and airborne-
Plexiglas tubes with an inner diameter of 3.2 mm were filled particle abrasion. For groups PA and AA, the ratios for O/Zr
with auto-curing composite resin (Clearfil FII, Kuraray Med- and C/Zr were reduced, comparable to those of group CL. Also,
ical, Inc., Osaka, Japan). The filled tubes were bonded with N was no longer detected.
Panavia F 2.0 (PF 2.0) (Kuraray Medical, Inc., Osaka, Japan) to
the ceramic surface using an alignment apparatus [17] under 4.2. TBS results
a load of 750 g according to the manufacturer’s instruction.
The specimens bonded with PF 2.0 were light-cured for 20 s Medians, means and standard deviations of TBS in MPa of the
from two opposite sides with a dental curing light machine tested groups are shown in Table 2. Clean specimens (group
(Optilux 500, Kerr, Danbury, USA). Then the bonded specimens CL) showed relatively high initial TBS, which remained statis-
were placed for 10 min at room temperature, and then stored tically stable over 150 days water storage with thermal cycling.
in 37 ◦ C water for 3 days or 150 days with 37,500 thermal cycles After saliva immersion and water cleaning (group WA) or alco-
from 5 to 55 ◦ C (Fig. 1). After different storage conditions, ten- hol cleaning (group AL) initial TBS was significantly lower than
sile bond strength (TBS) was tested with a universal testing in group CL (p ≤ 0.01). During 150 days water storage with ther-
apparatus (Zwick Z010/TN2A, Germany) at a crosshead speed mal cycling all specimens of groups WA and AL debonded
of 2 mm/min using a chain loop alignment which provided a spontaneously. The initial TBS of group PA was statistically
moment-free axial application [17]. higher than those of groups WA and AL (p ≤ 0.05), but statisti-
cally lower than that of groups AA and CL (p ≤ 0.05). However,
3. SEM examination after 150 days storage and thermal cycling, long-term TBS of
group PA decreased significantly. Group AA showed a high and
A scanning electron microscope (SEM, XL 30 CP, Philips, Kas-
sel, Germany) operating at 10–25 kV was used to observe the
failure modes of debonded ceramic specimens after tensile
Table 2 – Tensile bond strength to zirconia ceramic after
testing. Failure modes were classified into one of the following
saliva contamination and different cleaning methods
modes: A, adhesive failure at the ceramic surface; C, cohesive
failure in the luting composite resin or in the tube filling com- Groups Panavia F 2.0
posite resin. Failure areas of each mode were calculated and 3 days 150 days/37,500 TC
expressed as a percentage of the total bonding surface area
Means (S.D.) Medians Means (S.D.) Medians
for each test group.

3.1. Statistical analysis CL 33.9 (11.8) 34.0 A, ␣ 38.3 (13.6) 41.9 A, ␣

WA 12.8 (3.1) 13.0 B, ␣ 0a 0a B, ␤
Statistical analyses were performed using the Kruskal–Wallis AL 10.0 (3.4) 9.8 B, ␣ 0a 0a B, ␤
test followed by multiple pair-wise comparisons PA 25.4 (10.7) 22.0 A, ␣ 7.8 (3.6) 7.7 C, ␤
(Mann–Whitney test) at ˛ = 5%. AA 35.3 (5.9) 35.3 A, ␣ 29.63 (4.8) 28.4 A, ␣

Means, standard deviations (S.D.) and medians in MPa (n = 8).

4. Results a
All specimens debonded spontaneously during water storage
with thermal cycling. TC = thermal cycles. Within the same col-
4.1. XPS results umn, medians with the same letters (A–C) are not statistically
different (p > 0.05), within the same row, medians with the same
The peak intensity ratios of C/O, O/C, C/Zr and N/Zr in test Greek letters (␣ and ␤) are not statistically different (p > 0.05).
Group codes see Fig. 1.
groups are shown in Table 1. After saliva contamination (group

Descargado para Anonymous User (n/a) en Universidad El Bosque de ClinicalKey.es por Elsevier en agosto 26, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 508–513 511

stable TBS over time. No statistical difference was detected

between groups AA and CL (p > 0.05) for initial and long-term
TBS.

4.3. SEM observation and fractographic analysis

Mean percentages of areas assigned to the failure modes

observed in the bonding groups after tensile bond strength
testing are shown in Fig. 2. Figs. 3–5 show representative SEM
photographs of fractured interfaces after tensile bond testing
following 3 days or 150 days water storage with 37,500 thermal
cycles. For groups CL and AA with high TBS, most failures were
found to be cohesive in the luting composite resin (Panavia F
2.0) or the tube filling composite resin (Clearfil FII) (Fig. 3). In
contrast, for groups WA and AL the failure mode was com-
pletely adhesive at the zirconia ceramic surface, without any
composite resin remaining on the ceramic surface for both
storage conditions (Fig. 4). For groups PA, the failure mode was
mostly cohesive with a high initial TBS, however, after 150 days
storage, there was a significant increase in the adhesive failure Fig. 2 – Mean percentages of areas assigned to the failure
area with mainly cohesive failure in PF 2.0 luting resin (Fig. 5). modes observed in test groups after tensile bond strength
testing. A, adhesive failure at ceramic surface; C, cohesive
failure in luting composite resins (Panavia F 2.0) or tube
5. Discussion filling composite resin.

In the application of adhesive resin in ceramic bonding, factors

influencing ceramic bonding include contamination during

Fig. 3 – (a) Low magnification SEM micrograph showing a representative sample of groups CL and AA with a mixed but
predominant cohesive failure mode; (b) high magnification SEM micrograph of A in (a) indicating adhesive failure at the
zirconia ceramic surface; (c) area C1 in (a) indicating cohesive failure in the luting resin (Panavia F 2.0); (d) area C2 in (a)
indicating cohesive failure in tube filling composite resin (Clearfil FII).

Descargado para Anonymous User (n/a) en Universidad El Bosque de ClinicalKey.es por Elsevier en agosto 26, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
512 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 508–513

Fig. 4 – (a) Low magnification SEM micrograph of complete adhesive failure mode in test groups WA and AL; (b) high
magnification SEM micrograph indicating an adhesive failure at the ceramic surface without luting resin residues.

the bonding procedure that will interfere with the develop-
ment of good adhesion, the wettability of ceramic by adhesive
resin, the roughness of the ceramic surface, the composition
of adhesives, and the performance of adhesives. To detect
and remove contaminants is very important to achieve a
long-term durable ceramic bonding clinically. Therefore, in
this study, the influence of saliva contamination on ceramic
bonding by simulating a clinical try-in procedure was inves-
tigated. Long-term water storage combined with thermal
cycling at regular intervals was used as a standard aging mode
[11,17–19].
Saliva consists of organic materials such as salivary pro-
teins, bacteria and food debris in water solution. After saliva
immersion, salivary protein adsorption would occur not
only on the tooth surface [20], but also on the restorative
materials [21,22]. The hydrophilic ceramic surfaces show a
roughened “activated” surface after airborne-particle abra-
sion with 50 ␮m Al2 O3 particles, prior to saliva immersion
[17].
Non-covalent adsorption of salivary proteins occurred on
this surface after the immersion of ceramic into saliva for
60 s. This organic coating could not be removed by rinsing with
tap water for 15 s, as shown by XPS, revealing a considerable
increase in C, O, and N after saliva immersion and rinsing
(group WA). Therefore, the first hypothesis that there is no
saliva contamination after try-in with saliva exposure has to
be rejected.
In addition, it was not possible to achieve a durable
bond to the ceramic after saliva immersion with only
water rinsing (group WA). All specimens showed low ini-
tial bond strengths and debonded spontaneously after 150
days with 100% adhesive failure at the ceramic surface.
However, the control group (CL) confirmed previous results
showing that a long-term durable bond to clean zirconia
ceramic is achieved when using airborne-particle abrasion
prior to phosphate-monomer-containing composite resins
[3,4].
Chemical cleaning can be accomplished by using alkaline,
acidic, neutral, solvent, or emulsion. In this study, alcohol
cleaning seems not to be very effective in removing salivary
organic contaminants, since the reduction of C, O, and N was
limited as shown by XPS. In addition, initial TBS of group
AL was significantly lower than those of groups PA and AA
with 100% adhesive failure at the ceramic surface. All speci-
mens of group AC debonded spontaneously during 150 days
storage.
Acidic cleaning is mainly based on the removal of organic
residues which are readily dissolved in acid. In this study,
phosphoric acid cleaning (group PA) seems to be an effective
chemical cleaning method to remove salivary contamination
as indicated by XPS. The XPS result showed that organic con-
taminants are removed by phosphoric acid since the ratios
C/Zr, O/Zr, and C/O were approaching the ratios of the con-
trol group CL, and the N element was not detectable on the
specimen surface in this group. However, the relatively high
initial TBS of group PA was statistically significantly reduced
after 150 days storage with thermal cycling and the adhesive
failure mode increased greatly. In this study, non-filled phos-
phoric acid gel and water spraying for 30 s were used to avoid
gel residues, especially inorganic fillers, which might influence
bonding [23]. A possible explanation for the decrease in TBS
is that phosphoric acid might change the surface energy at
the ceramic surface and therefore reduce the bonding prop-
erties although the organic contamination as measured by
XPS was removed. However, this assumption has to be further
investigated.
Among the cleaning methods tested in this study, airborne-
particle abrasion (group AA) was the most effective cleaning
method to remove the saliva contaminants, as shown by XPS
and TBS results which were comparable to those of the control
group CL. In addition, after 150 days storage long-term TBS in
group AA remained stable compared to initial TBS.
Therefore the null hypothesis that there is no influence
of saliva contamination and cleaning methods on zirco-
nia ceramic bonding has to be rejected. The combination
of chemical surface characterization with XPS and long-
term TBS performance was useful to determine the effects
of the contamination and cleaning methods on ceramic
bonding. Airborne-particle abrasion was the most effective
method to remove saliva contaminants and to create a zir-
conia ceramic surface suitable for long-term durable resin
bonding.

Descargado para Anonymous User (n/a) en Universidad El Bosque de ClinicalKey.es por Elsevier en agosto 26, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
 d e n t a l m a t e r i a l s 2 4 ( 2 0 0 8 ) 508–513
Fig. 5 – (a) Low magnification SEM micrograph showing a
representative sample of group PA after 150 days water
storage with thermal cycling. There is a significant increase
of adhesive failure area (b) accompanied with low TBS, and
(c) cohesive failure was mostly in the luting resin (Panavia F
2.0).
references
[1] Blatz MB, Sadan A, Kern M. Resin–ceramic bonding: a review
of the literature. J Prosthet Dent 2003;89:268–74.
[2] Bishara SE, Olsen ME, Damon P, Jakobsen JR. Evaluation of a
new light-cured orthodontic bonding adhesive. Am J Orthod
Dentofacial Orthop 1998;114:80–7.
Descargado para Anonymous User (n/a) en Universidad El Bosque de ClinicalKey.es por Elsevier en agosto 26, 2018.
Para uso personal exclusivamente. No se permiten otros usos sin autorización. Copyright ©2018. Elsevier Inc. Todos los derechos reservados.
513
[3] Kern M. Bond strength of luting cements to zirconium oxide
ceramics. Int J Prosthodont 2000;13:350.
[4] Wegner SM, Kern M. Long-term resin bond strength to
zirconia ceramic. J Adhes Dent 2000;2:139–47.
[5] van Schalkwyk JH, Botha FS, van der Vyver PJ, de Wet FA,
Botha SJ. Effect of biological contamination on dentine bond
strength of adhesive resins. SADJ 2003;58:143–7.
[6] Aboush YE. Removing saliva contamination from porcelain
veneers before bonding. J Prosthet Dent 1998;80:649–
53.
[7] Bishara SE, Oonsombat C, Ajlouni R, Denehy G. The effect of
saliva contamination on shear bond strength of orthodontic
brackets when using a self-etch primer. Angle Orthod
2002;72:554–7.
[8] Cacciafesta V, Sfondrini MF, Baluga L, Scribante A, Klersy C.
Use of a self-etching primer in combination with a
resin-modified glass ionomer: effect of water and saliva
contamination on shear bond strength. Am J Orthod
Dentofacial Orthop 2003;124:420–6.
[9] Eiriksson SO, Pereira PN, Swift Jr EJ, Heymann HO,
Sigurdsson A. Effects of saliva contamination on resin–resin
bond strength. Dent Mater 2004;20:37–44.
[10] Rosenstiel SF, Land MF, Fujimoto J. Contemporary fixed
prosthodontics. 2nd ed. St. Louis: Mosby; 1995. p. 622.
[11] Kern M, Wegner SM. Bonding to zirconia ceramic: adhesion
methods and their durability. Dent Mater 1998;14:64–
71.
[12] Wegner SM, Gerdes W, Kern M. Effect of different artificial
aging conditions on ceramic/composite bond strength. Int J
Prosthodont 2002;15:267–72.
[13] Quaas AC, Yang B, Kern M. Panavia F 2.0 bonding to
contaminated zirconia ceramic after different cleaning
procedures. Dent Mater 2007;23:506–12.
[14] Hantsche H. Comparison of basic principles of the
surface-specific analytical methods: AES/SAM, ESCA (XPS),
SIMS, and ISS with X-ray microanalysis, and some
applications in research and industry. Scanning
1989;11:257–80.
[15] Kern M, Thompson VP. ESCA surface characterization of the
alumina ceramic in-ceram after various conditioning
methods for resin bonding. J Dent Res 1994;73(Special
issue):197 [abstr no. 763].
[16] Ratner BD, Castner DG. Electron spectroscopy for chemical
analysis. In: Vickerman JC, editor. Surface analysis. The
principle techniques. Chichester: John Wiley & Sons; 1997. p.
43–96.
[17] Hummel M, Kern M. Durability of the resin bond strength to
the alumina ceramic Procera. Dent Mater 2004;20:498–
508.
[18] Kern M, Strub JR. Bonding to alumina ceramic in restorative
dentistry: clinical results over up to 5 years. J Dent
1998;26:245–9.
[19] Lüthy H, Loeffel O, Hammerle CHF. Effect of thermocycling
on bond strength of luting cements to zirconia ceramic.
Dent Mater 2006;22:195–200.
[20] Baier RE, Glantz PO. Characterization of oral in vivo films
formed on different types of solid surfaces. Acta Odontol
Scand 1978;36:289–301.
[21] Carlen A, Nikdel K, Wennerberg A, Holmberg K, Olsson J.
Surface characteristics and in vitro biofilm formation on
glass ionomer and composite resin. Biomaterials
2001;22:481–7.
[22] Hannig M. Transmission electron microscopic study of in
vivo pellicle formation on dental restorative materials. Eur J
Oral Sci 1997;105:422–33.
[23] Eliades G, Palaghias G, Vougiouklakis G. Effect of acidic
conditioners on dentin morphology, molecular composition
and collagen conformation in situ. Dent Mater 1997;13:24–33.