Dental Materials (2005) 21, 773–779

www.intl.elsevierhealth.com/journals/dema

Effects of silicon coating on bond strength of two

different titanium ceramic to titanium
Isil Özcan, Hakan Uysal*

Department of Prosthodontics, Faculty of Dentistry, Çukurova University, Balcali, 01330 Adana, Turkey

Received 6 July 2004; received in revised form 12 January 2005; accepted 13 January 2005

KEYWORDS Summary Objectives. This study investigated the effect of silicon coating (SiO2) by
Ceramic; magnetron sputtering on bond strength of two different titanium ceramics to
Titanium; titanium.
Bond strength; Methods. Sixty cast titanium specimens were prepared following the protocol ISO
Silicon coating; 9693. Titanium specimens were divided into two test and control groups with 15
Metal–ceramic specimens in each. Test groups were silicon coated by the magnetron sputtering
bonding technique. Two titanium ceramics (Triceram and Duceratin) were applied on both
test (coated) and control (uncoated) metal specimens. The titanium–ceramic
specimens were subjected to a three point flexural test. The groups were compared
for their bond strength. SEM and SEM/EDS analyses were performed on the
delaminated titanium surfaces to ascertain bond failure.
Results. The mean bond strength of Ti–Duceratin, Ti–Triceram, Si-coated Ti–
Duceratin and Si-coated Ti–Triceram were 17.22G2.43, 23.31G3.18, 23.21G3.81
and 24.91G3.70 MPa, respectively. While the improvement in bond strength was 30%
for Duceratin, it was statistically insignificant for Triceram. An adhesive mode of
failure was observed in the Duceratin control group. In the silicoated Duceratin
specimen, the bonded ceramic boundaries were wider but less than in the silicoated
Triceram specimen. In the coated Triceram specimen, the ceramic retained areas
were frequent and the failure mode was generally cohesive.
Significance. Silicon coating was significantly effective in both preventing titanium
oxide layer formation and in improving bond strength for Duceratin. However, it was
of less value for Triceram.
Q 2005 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.

Introduction

Titanium and its alloys have been alternative sub-

structure materials for metal–ceramic restorations,
* Corresponding author. Tel.: C90 322 3387330; fax: C90 322
3387331.
with their excellent biocompatibility, corrosion
E-mail address: hakanu@cu.edu.tr (H. Uysal). resistance, mechanical properties close to dental
0109-5641/$ - see front matter Q 2005 Academy of Dental Materials. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.dental.2005.01.014
774
gold alloys and low cost [1–3]. However, their
chemical properties provide difficulties in ceramic
bonding [4–8]. The formation of poorly adhering oxide
to titanium at dental ceramic firing temperature
causes bonding problems between titanium and
ceramic [4–11]. The formation of a metal–ceramic
bond at firing temperatures should be strong enough
to resist both transient and residual thermal stresses
and mechanical forces in function. The requirements
to provide such a bond are discussed as chemical,
thermal, and mechanical compatibility, and have
been reviewed in detail [8–12].
Various surface modifications and attempts to
overcome oxidation problems associated with tita-
nium ceramic systems have been considered in the
literature [10,11,13–21]. One of the approaches
involves coating the titanium substrate with another
layer to prevent the formation of excessive and non-
adherent Ti oxide layer during porcelain firing, and
to improve the bond between ceramic and titanium
by controlled oxidation [10]. However, the use of
coatings has given contradictory results [10,11,13,
14,21]. It was advocated by Oshida et al. [10] that
nitridation of the titanium surface could control the
film thickness that might ensure satisfactory bonding
with porcelain. But the mechanical test results, in the
second part of that study, failed to support their
hypothesis. No statistically significant difference
could be demonstrated between bond strengths of
fired porcelain on nitrided pure titanium and non-
coated titanium [11]. On the other hand, the
presented results for silicon nitride (Si3N4) coating
were encouraging [14]. It was reported by Wang [14]
that Si3N4 coatings were effective in preventing the
oxidation of titanium and also provided a significant
improvement in ceramic bonding.
It was postulated in this study that a silicon (SiO2)
coating, which could be expected to serve as an
oxygen diffusion barrier on titanium during the
ceramic firings, could be effective in forming an
improved bond between titanium and ceramic. The
authors aimed to evaluate the effects of silicon
deposition using the magnetron sputtering technique
(silicon coating) on the bond strength of two different
titanium ceramics to titanium. The mode of failure at
the interface and effect of silicon coating on reducing
titanium oxidation were investigated.
Materials and methods
Preparation of metal–ceramic test specimens
The specimens were prepared for a three point
bending test as specified in ISO 9693 [22].
I. Özcan, H. Uysal
Seventy-five autopolymerizing acrylic resin (GC
Pattern resin, GC Corporation Tokyo, Japan) pat-
terns for cast NiCr alloy and titanium were
fabricated using a mold with recess dimensions of
25 mm!3 mm!0.7 mm.
Fifteen patterns were sprued and invested with a
phosphate-bonded investment and NiCr alloy (Her-
aenium NA, Heraeus Kulzer, Berkshire, Germany)
were cast with a centrifugal casting machine. Sixty
ASTM Grade II CP titanium (J. Morita, Europe GmbH,
Frankfurt, Germany) strips were cast using a dental
casting machine with MgO (magnesium)-based
investment in an argon atmosphere.
The cast strips were gently wet-ground with SiC
paper (120, 240, 400, 600 grit) to final dimensions of
25G1 mm!3G0.1 mm!0.5G0.05 mm. All dimen-
sions were controlled with a digital micrometer
(measurement to the nearest 0.01 mm). Each metal
strip was air-abraded with 125 mm aluminum oxide
powder at a pressure of 4 bar. All metal specimens
were then rinsed in ethyl acetate each for 5 min
before final cleaning. Final cleaning was performed
with hot distilled water in an ultrasonic cleaner for
10 min prior to ceramic application.
Silicon deposition was conducted by magnetron
sputtering on 30 cast titanium strips with a
thickness of 500 nm.
For cast titanium specimens, a thin bonding
ceramic was applied to the central portion (8 mm!
3 mm) of the specimen surface. After the bonding
ceramic was fired, opaque ceramic was applied in a
masking layer, as it is common practice with the
ceramic fused to metal technique, and fired. Dentin
ceramic was applied onto opaque ceramic using a
custom-made jig that controlled the position and
thickness of the dentin. Low fusing ceramics for
titanium: Triceram (Esprident GmbH, Inspringen,
Germany) and Duceratin (Ducera Dental, Rosbach,
Germany), and conventional ceramic: Ceramco II
(Ceramco Inc., Burlington, NJ, USA) for NiCr alloy
were fired according to each manufacturer’s
recommendations. The ceramic was then adjusted
to a final thickness of 1 mm after firing; excess
ceramic therefore was removed with a disk.
Mechanical testing of bond strength
Testing the bond strength of ceramic to cast metal
was performed with a three-point flexural device on
a universal testing machine (Testometric M500
25 kN, Testometric Co., Rochdale, UK) according
to the ISO 9693 protocol. Each fired specimen was
positioned on the supports with a span distance of
20 mm in the testing machine with the ceramic
positioned symmetrically on the side opposite to
Effect of silicon coating on bond strength
the applied load. A compressive load was applied at
the midline of the metal strip at a cross-head speed
of 1.5 mm/min. The radius of the supports and the
bending piston was 1 mm. The load was applied
until the disruption in the load–deflection curve,
which indicated bond failure.
The bond strength was calculated numerically
based on the flow chart in ISO 9693. The strength
data obtained from titanium subgroups were
statistically analyzed using one-way ANOVA fol-
lowed by Tukey’s B post hoc test (aZ0.05).
Differences between titanium subgroups and NiCr
group were evaluated by Duncan’s post hoc test
(aZ0.05).
SEM evaluation of bond interface
One specimen of each combination was prepared to
examine the titanium–ceramic interface for the
presence of an oxide layer and evaluate metal–
ceramic bonding with scanning electron microscopy
(JEOL-JMS-6400 Scanning Microscope, JEOL Ltd,
Tokyo, Japan). The specimens were embedded in
clear acrylic resin (Orthocryl EQ, Dentaurum,
Pforzheim, Germany). After the complete polym-
erization of the acrylic resin, the specimen/resin
assembly was sectioned along the longitudinal axis
of the specimens with a cutting and grinding
machine (Exact 310GP, Axact Apparatebau GmbH
& Co., Nordenstent, Germany), with a 0.2 mm
thickness band saw. The cutting rate was
150 m/min and the applied force was 50 g. The
sectioned specimens were ultrasonically cleaned
for 10 min and a flat surface was prepared by
mechanical polishing using SiC papers through grits
of 320, 400–600. Final polishing was accomplished
on a rotary polisher with aluminum oxide sizes 0.5
and 0.3 mm, respectively.
SEM/EDS evaluation of delaminated surfaces
Delaminated specimens of each titanium group
were obtained by continuing to load titanium–
ceramic specimens. The delaminated interfacial
surfaces of the randomly selected specimens were
examined with a scanning electron microscope. To
ascertain the mode of failure, elemental analysis
(X-Ray Energy-Dispersive Spectroscopy EDS, Noron
Instrument Inc., Series II X-Ray Microanalysis
System, Tokyo, Japan) was performed concurrently
with the SEM examination to identify the material
at different locations on delaminated interfacial
surfaces. Elemental analysis was also performed on
the bonding ceramics fired on titanium specimens
to determine their chemical compositions.
775
Results
Bond strengths
The mean bond strength data and standard devi-
ations are presented in Table 1. The mean bond
strengths of Ti–Duceratin, Ti–Triceram, Si-coated
Ti–Duceratin, Si-coated Ti–Triceram were 17.22,
23.31, 23.21, and 24.91 MPa, respectively. The Ti–
Duceratin group was significantly (p%0.000) lower
than the rest of the groups. While the improvement
on bond strength was 30% for Duceratin when
titanium was silicon-coated, it was statistically
insignificant for coated and uncoated Ti–Triceram
groups. The bond strength of NiCr–Ceramco was
38.54 MPa, which was significantly higher than all
titanium subgroups.
SEM examinations of cross-sections
The cross-section at the titanium–Duceratin inter-
faces showed a thick oxide layer along the inter-
faces between ceramic and titanium (Fig. 1a). No
definite oxide layer was observed at the interfaces
of the non-coated Triceram specimen (Fig. 1b) or
either of the silicon-coated titanium–ceramic speci-
mens (Fig. 1c and d).
SEM, SEM/EDS analyses of delaminated
surfaces
The compositions of the bonding ceramics in wt%
are presented in Table 2. The elements mapped on
delaminated surfaces were compared with the
composition of the bonding ceramic used in that
specimen. SEM/EDS analysis of the metal surface
indicated that the dark areas were bonding ceramic
while the light regions were titanium.
The Titanium–Duceratin SEM photomicrograph
of the titanium surface after debonding showed
no residual ceramic retained on the metal surface
Table 1 Bond strength of metal–ceramic systems as
measured by three point bending test (MPa).
Metal–ceramic system n Mean (SD)
Ti–Duceratin 15 17.22 (2.43)*
Ti–Triceram 15 23.31 (3.18)
Silicon-coated Ti–Duceratin 15 23.21 (3.81)
Silicon-coated Ti–Triceram 15 24.91 (3.70)
NiCr–Ceramco 15 38.54 (3.19)**
SD, standard deviation; *significant at p!0.000 among the
titanium subgroups; **significant at p!0.000 from the
titanium subgroups.
776 I. Özcan, H. Uysal

Figure 1 SEM photographs (!2000 magnification) of cross-sections of titanium–ceramic systems (a) titanium–
Duceratin interfaces. Arrows show oxide layer along the interface between ceramic and titanium. No definite oxide
layer was observed at the interfaces in non-coated Triceram specimen (b) and either of the silicon-coated titanium–
ceramic specimens, coated titanium–Duceratin (c), coated titanium–Triceram (d) (M, metal; C, ceramic).

(Fig. 2a). EDS analysis of the metal surface revealed Discussion

a substantial amount of titanium. An adhesive
titanium–ceramic bond failure mode was observed.
Titanium–Triceram SEM examination of the
debonded titanium surfaces demonstrated sporadic
ceramic remnants on the surface of the titanium.
This indicated a mixed adhesive and cohesive
failure mode, but mainly adhesive for the tita-
nium–Triceram bonding. (Fig. 2b).
Silicon-coated groups. Ceramic remnants could be
identified on the debonded interfaces of Si-coated
specimens on SEM photomicrographs (Fig. 2c and d).
Both of the coated groups exhibited a mixed adhesive
and cohesive titanium–ceramic bonding failure. It
was observed that the bonded ceramic boundaries in
silicon-coated titanium–Duceratin were wider but
less than in the silicon-coated titanium–Triceram
specimen. In the silicoated Triceram group, the
ceramic retained areas were frequent and the failure
was mixed, but generally cohesive in ceramic.
Cross-sectional SEM examinations of uncoated
groups showed different characterizations of the
reaction zone at the interface. The Ti–Duceratin
interface was the only group presented as an oxide
layer. The adhesive mode of failure was observed in
the Ti–Duceratin specimen and the Duceratin group
had the lowest bond strength. Low bond values and
similar reaction zone characteristics have also been
shown for Duceratin in a recent study [23].
Compared with the uncoated group, silicon-coated
titanium–Duceratin specimens showed a bond
strength improvement of 30%, as no oxide layer
was observed. When the mechanical test results of
uncoated and coated Duceratin groups were
assessed with the SEM interface findings, it could
be suggested that the presence of the oxide layer
determines the bond strength of titanium–ceramic
systems. The results of this study support others,

Table 2 Chemical composition of bonding ceramics in atomic weight% at 20 kV.

Na Al Si K Sn Ba Ca
Duceratin Haft bond 21.16 9.53 53.58 5.36 8.80 1.57 –
Triceram bonder 17.62 11.78 48.46 8.18 8.25 2.15 3.57
Effect of silicon coating on bond strength 777

Figure 2 SEM photomicrograph (!150 magnification) of uncoated (a and b) and coated (c and d) titanium surface
after debonding. Ceramic remnants could be identified on metal as dark areas. Light regions represented titanium. EDS
analyses of elements of these remnants were consistent with the bonding ceramic composition. (a) Titanium–Duceratin;
no residual ceramic retained on the metal surface. (b) Titanium–Triceram; the sporadic ceramic remnants on the
surface. The bonding ceramic boundaries in coated ti–Duceratin (c) were larger but less than coated ti–Triceram
specimen (d).

which attribute lower bond strength to the thick
oxide layer on the metal surface [5,7,10,11,24,25].
It appeared that the silicon coating on cast titanium
surfaces could effectively control the oxide layer
formation. Therefore, remarkably enhanced bond
strengths were obtained in Duceratin specimens.
Wang et al. [14] used Si3N4 coatings as a protective
coating on Ti substrate and fired Duceratin. Their
control specimens without coating had much lower
adhesion test values than the coated group and
produced multilayered Ti oxide at which the
fractures had always occurred. Their test results
in terms of failure loads are not comparable with
those in this study since the test performed in that
study was four point bending. However, our findings
related to the reduction of thick oxide layer are
consistent with the results of that study.
In contrast to the Duceratin control specimen,
the Ti–Triceram interface presented no definite
oxide layer. The higher mechanical test values in
the Triceram group may be due to the lack of an
oxide layer. The silicon coating did not improve the
bond strength in the Triceram group. It should be
noted that if Triceram had been the only ceramic
material used in this study, the effects of the
coating would not have been observed. It is
apparent that the Triceram bonding ceramic pre-
vented the formation of a thick oxide layer.
The bonding ceramic may provide improved bond
strength both by controlling the oxide layer
formation and reducing the thermal expansion
mismatch between titanium and ceramic [26]. The
chemical compositions of porcelain and ratios of
crystallized compositions have been demonstrated
to play an important role in the titanium–ceramic
bond [27]. SEM/EDS analyses revealed that the
bonding ceramics used in this study have different
chemical compositions. The lack of a specific oxide
layer on uncoated Triceram–titanium substrate may
be attributed to the effectiveness of the bonding
ceramic in preventing excess oxidation of titanium
during ceramic firing. The stresses at the metal–
ceramic interface at room temperature are
affected not only by mismatch in the thermal
contraction of metal and ceramic, but also by any
stress relaxation in the ceramic during cooling [28].
Crystalline oxide particles in the glass matrix
change the expansion coefficients of the ceramic
[28]. Thus, the different chemical compositions of
the ceramics used in this study may explain the
difference in their interfacial reaction behavior and
bond strengths.
778
In this study, a sophisticated technique for
measuring metal–ceramic adherence as developed
by Ringle and Mackert [29] was not used. But, SEM/
EDS examination of debonded specimens revealed
that the ceramic remnants on coated Ti–Triceram
specimen were significantly larger and more frequent
than in uncoated specimens. This finding might be an
indication of a superior adhesion of Triceram–ceramic
to coated titanium. However, mechanical test results
of the three point bending test showed no improve-
ment in the coated group since the bonds strength of
uncoated and coated Triceram groups was statisti-
cally insignificant. In a recent study, the adhesion of
porcelain to titanium in various bonding systems was
determined in terms of the strain energy release rate
using a four point bending test [23]. The authors
concluded that SEM photographs might not show any
major distinguishing characteristics that would sup-
port the ranking and adhesion measurement values
among the groups. Papazoglu and Brantley’s [30]
study to characterize the metal–ceramic bond
showed there was no correlation between the
percent of ceramic retained and the failure bond.
However, they pointed to the value and necessities
of SEM examination to investigate metal–ceramic
bonds [31].
The bond strength of coated titanium–ceramic
groups was 47–64% of the NiCr metal–ceramic system
in this study, similar to other studies [18,32,33]. The
mean strength values were significantly lower than
the clinically proven NiCr group. The results of this
study suggest that eliminating the oxide layer on the
titanium substrate alone may not necessarily provide
clinically proven bond strength values. Other factors
such as residual stress, which is dependent on thermal
contraction difference between the titanium and
ceramics, the glass transition temperature of cer-
amics, the geometry of the samples and the elastic
constants of the materials, may effectively contrib-
ute to limiting the bond strength. The findings of this
study support Könönen’s and Kvilahti’s [8] statement
that reactivity is necessary for chemical bonding
whereas in titanium–ceramic systems, brittle reac-
tion products may impair the mechanical compat-
ibility of the joint. Reaction products as well as the
interface material, which consists of the elements
that react with titanium, should have sufficient
strength to resist stresses due to thermo-mechanical
incompatibility [8].
Conclusions
The effect of a SiO2 coating, applied by magnetron
sputtering, on the adhesion of titanium and two
different low fusing ceramics was studied.
I. Özcan, H. Uysal
The bonding ceramic of Triceram was more
effective than that of Duceratin in reducing oxide
film during ceramic firings. The bond strength of
titanium–Triceram was significantly greater than
the bond strength of titanium–Duceratin.
Silicon coating effectively prevented excessive
oxidation of titanium and significantly increased
the bond between titanium and Duceratin. How-
ever, the silicon coating did not provide additional
improvement in bond strength for Triceram, which
had a bonding ceramic effective in reducing oxide
film.
Bond strength values of NiCr–ceramic were
47–64% higher than all titanium subgroups.
Acknowledgements
This study was supported by Cukurova University
Research Fund SBE2002D11 and conducted
at Biomechanics and Materials Test Laboratory,
Division of Dental Research, Faculty of Dentistry.
References
[1] Hruska AR. Intraoral welding of pure titanium. Quintessence
Int 1987;18:683–8.
[2] Lautenschlager EP, Moaghan P. Titanium and titanium
alloys as dental materials. Int Dent J 1993;43:245–53.
[3] Wang RR, Fenton A. Titanium for prosthodontic appli-
cations: a review of the literature. Quintessence Int 1996;
27:401–8.
[4] Taira Y, Matsumura H, Yoshida K, Tanaka T, Atsuta M.
Influence of surface oxidation of titanium on adhesion.
J Dent 1998;26:69–73.
[5] Adachi M, Macket Jr JR, Parry EE, Fairhurst CW. Oxide
adherence and porcelain bonding to titanium and Ti–6Al–4V
alloy. J Dent Res 1990;69:1230–5.
[6] Kimura H, Horng CJ, Okazaki M, Takahashi J. Oxidation
effects on porcelain–titanium interface reactions and bond
strength. Dent Mater J 1990;9:91–9.
[7] Könönen M, Kivilahti J. Bonding of low fusing dental
porcelain to commercially pure titanium. J Biomed Mater
Res 1994;28:1027–35.
[8] Könönen M, Kivilahti J. Fusing of dental ceramics to
titanium. J Dent Res 2001;80:848–54.
[9] Pang I, Gilbert JL, Chai J, Lautenschlager EP. Bonding
characteristics of low fusing porcelain bonded to pure
titanium and palladium–copper alloy. J Prosthet Dent 1995;
3:17–25.
[10] Oshida Y, Hashem A. Titanium–porcelain system. Part I.
Oxidation kinetics of nitrided pure titanium, simulated to
porcelain firing process. Biomed Mater Eng 1993;3:185–98.
[11] Oshida Y, Fung LW, Isikbay SC. Titanium–porcelain system.
Part II. Bond strength of fired porcelain on nitrided pure
titanium. Biomed Mater Eng 1997;7:13–34.
[12] Bagy M, Marshall SJ, Marshall GW. Metal ceramic compat-
ibility: a review of the literature. J Prosthet Dent 1990;63:
21–5.
Effect of silicon coating on bond strength
[13] Wang RR, Fung KK. Oxidation behavior of surface-modified
titanium for titanium–ceramic restorations. J Prosthet Dent
1997;77:423–34.
[14] Wang RR, Welch GE, Monteiro O. Silicon nitride coating on
titanium to enable titanium–ceramic bonding. J Biomed
Mater Res 1999;46:262–70.
[15] Reyes MJD, Oshida Y, Andres CJ, Barkoo T, Hovijitra S,
Brown D. Titanium–porcelain system. Part III. Effects of
surface modification on bond strengths. Biomed Mater Eng
2001;11:117–36.
[16] Oshida Y, Reyes MJD. Titanium–porcelain system. Part IV.
Some mechanistic considerations on porcelain bond
strengths. Biomed Mater Eng 2001;11:137–42.
[17] Fischer J. Ceramic bonding to a dental gold–titanium alloy.
Biomaterials 2002;23:1303–11.
[18] Atsu S, Berksun S. Bond strength of three porcelains to
forms of titanium using two firing atmospheres. J Prosthet
Dent 2000;84:467–74.
[19] Yoda M, Konno T, Takada Y, Iijima K, Griggs J, Okuno O,
et al. Bond strength of binary titanium alloys to porcelain.
Biomaterials 2002;22:1675–81.
[20] Cai Z, Bunce N, Nunn ME, Okabe T. Porcelain adherence to
dental cast CP titanium: effects of surface modifications.
Biomaterials 2002;2:979–86.
[21] Sadeq A, Cai Z, Miller AW. Effects of interfacial variable on
ceramic adherence to cast and machined commercially
pure titanium. J Prosthet Dent 2003;90:10–17.
[22] ISO 10477-9693. Metal–ceramic dental restorative systems.
2nd ed. Switzerland: International Organization for Stan-
dardization; 1999, p. 12–5.
[23] Saunsuwan N, Swain MV. Adhesion of porcelain to titanium
and a titanium alloy. J Dent 2003;31:509–18.
779
[24] Dent RJ, Preston JD, Moffa JP, Caputo A. Effect of oxidation
on ceramometal bond strength. J Prosthet Dent 1982;47:
59–62.
[25] Hautaniemi JA, Juhanoja JT, Suoninen EJ, Yli-Urpo AU.
Oxidation of for palladium-rich ceramic fusing alloys.
Biomaterials 1990;11:62–72.
[26] Gilbert JL, Covey DA, Lautenschlager EP. Bond character-
istics of porcelain fused to milled titanium. Dent Mater
1994;10:134–40.
[27] Mo A, Wang J, Liao Y, Cen Y, Shi X. The influences of
crystallized compositions in the porcelain on bonding
strength of titanium to porcelain. Hua Xi Kou Qiang Yi Xue
Za Zhi 2001;9:357–62.
[28] Kvam K, Hero H. Stress relaxation in titanium–
ceramic beams during veneering. Biomaterials 2002;
22:1379–84.
[29] Ringle RD, Mackert Jr JR, Fairhurst CW. An x-ray spectro-
metric technique for measuring porcelain–metal adher-
ence. J Dent Res 1983;62:933–6.
[30] Papazoglou E, Brantley WA. Porcelain adherence vs force to
failure for palladium–gallium alloys: a critique of metal–
ceramic bond testing. Dent Mater 1998;14:112–9.
[31] Papazoglou E, Brantley WA, Carr AB, Johnston WM. Porce-
lain adherence to high-palladium alloys. J Prosthet Dent
1993;70:386–94.
[32] Pröbster L, Maiwald U, Weber H. Three-point bending
strength of ceramics fused to cast titanium. Eur J Oral Sci
1996;104:313–9.
[33] Suansuwan N, Swain MV. New approach for evaluating
metal–porcelain interfacial bonding. Int J Prosthodont
1999;2:547–52.