The influence of the veneering porcelain and different surface treatments

on the biaxial flexural strength of a heat-pressed ceramic

Giuseppe Isgrò, BMedSci,a Prem Pallav, DDS, PhD,b Jef M. van der Zel, PhD,c and
Albert J. Feilzer, DDS, PhDd
Department of Dental Material Science, ACTA Netherlands, Amsterdam, The Netherlands;
Elephant Dental, Hoorn, The Netherlands

Statement of problem. The strength of all-ceramic restorations can be adversely affected by surface defects,
leading to restoration failures. Additionally, when a 2-layer all-ceramic restoration is required for esthetic pur-
poses, part of the stronger ceramic core material is replaced by veneering porcelain.
Purpose. This study evaluated the effect of different surface treatments on the strength of a ceramic core
material and veneering porcelain, as well as the influence of veneering porcelain on the strength of a 2-layer
ceramic structure.
Material and methods. Fifty heat-pressed ceramic cores and 30 veneering porcelain discs (17 mm diameter
⫻ 2 mm) were made. From the ceramic core group, 20 discs were selected and reduced to a thickness of 1 mm
and veneered with 1 mm of porcelain. These specimens were divided into 2 groups of 10 each. The remaining 30
ceramic core and the 30 veneering porcelain discs were divided into 2 sets of 3 equal sized groups (n⫽10).
Ceramic core groups were prepared for testing having the following surfaces: airborne-particle abrasion, ground,
and overglazed. Veneering porcelain groups were tested: as fired (no additional treatment), ground, and over-
glazed. Biaxial flexural strength was measured using the ball-on-ring test method. All specimens were loaded to
fracture. One and 2-way analysis of variance were used to analyze the data (␣⫽.05).
Results. The ceramic core discs were significantly (P⫽.001) stronger than the veneering porcelain discs for the
airborne-particle abrasion, as-fired, and ground surface treatments (82 ⫾ 11 MPa vs 51 ⫾ 8 MPa and 93 ⫾ 14
MPa vs 60 ⫾ 6 MPa, respectively). For the overglazed treatment, there was not a significant difference between
the core (115 ⫾ 1 4 MPa) and the veneer materials (107 ⫾ 14 MPa). The ground 1-layer core was significantly
(P⫽.015) stronger (93 ⫾ 14 MPa) than the 2-layer with the core tested in tension (72 ⫾ 19 MPa). There was no
significant difference between 1-layer veneer overglazed (107 ⫾ 14 MPa) and 2-layer discs when tested with the
veneer in tension (105 ⫾ 16 MPa).
Conclusion. The overglazed surface treatment significantly improved the strength of the materials tested, as
well as the strength of 2-layer discs with the veneer in tension. The veneering porcelain influenced the strength
of 2-layer specimens only when tested with the ground ceramic core surface in tension. (J Prosthet Dent 2003;
90:465-73.)

CLINICAL IMPLICATIONS
In this in vitro study, the overglazed surface treatment increased the strength of ceramic core and
veneering porcelain materials. The final surface preparation is of clinical importance because it
may have a positive effect on the longevity of all-ceramic heat-pressed restorations, single- or
double-layered. When a portion of the ceramic core material is replaced by a more esthetic,
veneering porcelain, the strength of a 2-layer ceramic restoration may be affected and the
restoration may be prone to failure at lower loads.

D ental ceramic materials should be capable of

meeting the increasing demand for esthetic restorations.
Their characteristic translucency enhances the ability to
produce natural-appearing restorations. However,
rather low mechanical strength is a problem. Although
a strong under compressive force, ceramic materials are
PhD Student, Department of Dental Material Science, ACTA Neth-
erlands. brittle and unable to withstand the complex tensile
b
Senior Researcher, Department of Dental Material Science, ACTA stresses that occur during functional loading of restora-
c
Netherlands. tions.1 Dental ceramics have been strengthened by
Commissioned Scientific Officer, Elephant Dental, Hoorn, The
metal bonding and by adding a microcrystalline phase to
Netherlands.
d
Professor, and Department Head, Dental Material Science, ACTA the glass matrix.2 With respect to the latter, higher
Netherlands. strength ceramics compared with the aluminous or feld-

NOVEMBER 2003 THE JOURNAL OF PROSTHETIC DENTISTRY 465

THE JOURNAL OF PROSTHETIC DENTISTRY ISGRÒ ET AL

Table I. Study materials

Trade Name Manufacturer Material CTE 20–500°C E GPa Shade Lot number

Carrara Press Core Elephant BV Ceramic Core 13.8 ppm* 73* S4 005256
Hoorn, NL
Carrara Vincent Veneering Porcelain 15.1 ppm* 54* A3 Powder 100243
Porcelain Liquid 690
Carrara Press Overglaze 10.5 — — Powder 002670
Glaze Liquid 687
E, Modulus of Elasticity; ppm, part per million.
*Values provided by manufacturer.

Table II. Firing schedule for ceramic core and veneering porcelain

Firing name Start temp °C Pre-drying min Rate °C/min End temp Hold time/min

Press 650 — 55 in vacuum 1150 17

Conditioner 450 4 55 in air 810 1
1st Dentin Firing 450 5 60 in vacuum 845 1.3
2nd Dentin Firing 450 4 60 in vacuum 840 1.3
Overglazed 450 3.3 60 in air 835 1.3
Dry specimen 150 — 60 in air 200 60

spathic porcelains, are produced for fabricating metal-
free, all-ceramic restorations. A strengthened ceramic
can be used as the sole material for making an all-ceramic
crown, inlay, or veneer restoration or as a core substrate
for a ceramic crown.3 However, compressive occlusal
forces produce tensile stresses along the inner surface of
ceramic crowns,4 which can initiate crack propagation
from surface flaws and porosities. Flaws are usually in-
troduced during the fabrication process5 and may be
responsible for early restoration failures. Crack initiation
may be influenced by the quality of the ceramic surface.
The propagation of cracks can be influenced by the in-
ternal structure of ceramic materials, which can provide
resistance to crack growth.6 Currently, all-ceramic sys-
tems strengthened with microcrystalline phases are suit-
able for anterior and premolar single crowns. Greater
strength would be necessary for fixed partial dentures.
Two developmental directions are recognized in ef-
forts to improve the strength characteristics of all-ce-
ramic systems. The first is directed toward enhancing the
strength of core materials, while the second focuses on
improving processing techniques designed to produce
more homogeneous ceramic materials. Often both di-
rections are used in the development of new all-ceramic
restorative systems with improved strength.
Ceramic core materials with new chemical composi-
tions have been developed for use with processing meth-
ods combining pressure and high temperature.5,7 These
new materials, known as heat-pressed ceramics, are leu-
cite-reinforced feldspathic porcelains strengthened by
incorporating leucite (K2O Al2O3 4SiO2) crystals, ap-
proximately 45 volume percent, in the glass matrix.8
Flexural strengths for these materials have been reported
to be 120 MPa.9 Conventional feldspathic porcelains
designed for metal-ceramic restorations contain 12 to
25 volume percent leucite9 and have strengths in the
range of 60 MPa.8 The increase in strength has been
achieved through a heat treatment, which enhances the
formation of a highly crystallized microstructure that
resists crack propagation under stress.2 As a crack grows,
it must pass through or around these crystals. This dis-
sipates energy from the crack and may even arrest crack
growth. As a result, a restoration with an arrested crack
may continue to function rather than fracture and fail.10
Avoiding large pore formation is the major advantage
of the hot-pressing technique. This method also allows
better distribution of the crystalline phase within the
glass matrix.7,11 Another advantage of the heat-press
method, in combination with the lost wax technique, is
that 1-layer all-ceramic crowns, veneers, and inlays can
be made with final contours using only the leucite-rein-
forced core material. These restorations are then stained
and overglazed with a thin layer of glass12 to enhance
esthetics and improve the surface quality of the ceramic
material.13,14
Leucite-reinforced feldspathic porcelains have been
designed to be compatible with low fusing porcelains,
which provide optical properties necessary for making
the most esthetic restorations.15 Thus a 2-layer all-ce-
ramic crown can be fabricated with a heat-pressed ce-
ramic core. The core substrate is then veneered with low
fusing porcelain applied with a brush to achieve the de-
sired shade and structure.
The development of stronger veneering porcelains is
not an area of current research. Therefore veneering
material strengths have remained largely unchanged. It
is doubtful whether further advances in ceramic core
strength would improve the performance of all-ceramic

466 VOLUME 90 NUMBER 5

ISGRÒ ET AL THE JOURNAL OF PROSTHETIC DENTISTRY

Table III. Dimensions, surface finishing, and BFS results for each group

Group No. Diameter (mm) Height (mm) Surface finishing BFS (MPa)
(n ⴝ 10) Material (Mean ⴞ SD) (Mean ⴞ SD) tested in tension (Mean ⴞ SD)

1 17.1 ⫾ 0.1 2.1 ⫾ 0.06 Airborne particle abrasion 82 ⫾ 11

2 Ceramic Core 17.1 ⫾ 0.2 2.01 ⫾ 0.05 Ground 93 ⫾ 4
3 17.1 ⫾ 0.1 2 ⫾ 0.04 Overglazed 115 ⫾ 14
4 17.0 ⫾ 0.06 2.1 ⫾ 0.03 As-fired 51 ⫾ 8
5 Veneering Porcelain 16.9 ⫾ 0.3 2.03 ⫾ 0.03 Ground 60 ⫾ 6
6 16.9 ⫾ 0.2 2.03 ⫾ 0.02 Overglazed 107 ⫾ 14
7 Core/Veneer 16.7 ⫾ 0.1 2 ⫾ 0.03 Veneer Overglazed 105 ⫾ 16
8 Core/Veneer 16.8 ⫾ 0.1 2.03 ⫾ 0.03 Core Ground 72 ⫾ 19

Fig. 1. Preparation of ceramic core specimens using lost wax

technique.

systems, thus leading to more durable restorations.

Studies have shown that a thin layer of veneering porce-
lain fired onto ceramic core material diminishes the
strength of the 2-layer test specimens.16,17 The veneer-
ing porcelain is still likely to be the weakest link.
The purpose of this study was to determine the influ-
ence of the veneering porcelain layer and 3 surface fin-
ishes, airborne-particle abrasion, ground, and over-
glazed, on the biaxial flexural strength of specimens Fig. 2. Porcelain slurry condensed in metallic mold using air
made with a heat-pressed leucite-reinforced ceramic re- pressure.
storative system. There were 2 hypotheses for this study.
First, the strength of a 2-layer ceramic specimen, a leu-
cite-reinforced ceramic core, veneered with low-fusing
dental porcelain, depends on the strength of the veneer- core can be pressed and veneered with ultra-low-fusing
ing porcelain and the surface finish provided. Second, a porcelain (Table II). The veneering porcelain is classi-
2-layer specimen is weaker than a specimen made with fied as ultra-low-fusing porcelain because its firing tem-
only 1 material. perature is below 870°C.18 In this system, both ceramic
core and veneering components are feldspathic porce-
MATERIAL AND METHODS lains containing different amounts of leucite. The chem-
An all-ceramic system (Carrara Press System; Ele- ical composition of the ceramic substrate material ac-
phant Dental BV, Hoorn, The Netherlands) was se- cording to the manufacturer is 59.0 wt% SiO2, 21.8 wt%
lected for this study (Table I). This system uses a heat- Al2O3, 10.4 wt% K2O, and 6.6 wt% Na2O; and the
pressed ceramic to fabricate 1-layer all-ceramic crowns, veneering porcelain is 64.0 wt% SiO2, 13.0 wt% Al2O3,
veneers, and inlays. Additionally, a ceramic substrate or 12.5 wt% K2O, and 7.0 wt% Na2O.

NOVEMBER 2003 467

THE JOURNAL OF PROSTHETIC DENTISTRY ISGRÒ ET AL

Specimen preparation
Fifty discs (17 mm diameter ⫻ 2 mm) of ceramic core
material and 30 discs of veneering porcelain were made
for test specimens. From the ceramic core group, 20
discs were selected and reduced to an even thickness of
1 mm and veneered with 1 mm of veneering porcelain
(Table III), restoring these discs to a 2-mm thickness.

Ceramic core discs

Initially 50 wax discs (17 mm diameter ⫻ 2 mm) were
made. Each disc, 1 at a time, were sprued using a 3-mm
round wax rod and invested with a graphite-free phos-
phate-bonded investment (Carrara Press Investment;
Elephant Dental BV) according to the manufacturer’s
recommendations (Fig. 1). The refractory material was
mechanically mixed (Model F; Whip Mix, Louisville,
Ky) under vacuum for 60 seconds. The casting ring was
filled with the investment material and left to bench set
for 15 minutes. The sprue base was carefully removed
and the casting ring placed in a burn-out furnace (Com-
bilar CL-V; Heraeus, Hanau, Germany) preheated to
800°C. The ring was heat soaked for 10 minutes at this
temperature to eliminate the wax. The temperature was
increased 10°C per minute to 950°C and held for 45
minutes. Next, the casting ring was removed from the
furnace, and the pressing channel was filled with Carrara
Press pellets together with the plunger. The mold was
then transferred into a special porcelain furnace (Stratos;
Elephant Dental BV) where the pressing process was
carried out according to the manufacturer’s instructions
(Table II). The ceramic core pellets were pressed into
the mold for 90 seconds at a pressure of 4 ⫻ 102 kPa.
The casting ring was allowed to bench-cool to room
temperature. Each ceramic disc was carefully divested
using airborne-particle abrasion (50-␮m glass beads),
first at 5 ⫻ 102 kPa and then at 3 ⫻ 102 kPa of pressure
to prevent damaging disc surfaces. Sprues were removed
with a diamond cutting disc (942 104 354524 200;
Komet, Lemgo, Germany), and attachment sites were
reshaped with a conical diamond bur (858 104 165524
014; Komet). The ceramic core discs were then ran-
domly divided into 5 groups of 10, numbered 1, 2, 3, 7,
and 8. Discs in Groups 1 to 3 received different surface
treatments, whereas those in Groups 7 and 8 were used
to make veneered 2-layer specimens.
Ceramic core discs in Groups 1 to 3 were prepared as
described below. Discs in Group 1 received no further
treatment after they were airborne-particle abraded.
Those in Group 2 were airborne-particle abraded then
mechanically wet ground on both sides, using a cast-iron
grinding wheel (PT 310; VEM Metallurgie BV, The
Netherlands) with a 30-␮m grit diamond abrasive slurry
at a speed of 40 rpm to produce discs with flat, smooth
surfaces. During the grinding process, discs were held
on the cast-iron wheel by means of a special holder de-
Fig. 3. Schematic representation of specimen under BFS
test.1, Stainless steel block; 2, circular slot filled with silicone
rubber; 3, steel disc; 4, two-layer specimen; 5, pins; 6, stain-
less steel ball.
signed by the manufacturer. Group 3 discs were treated
the same as in Group 2, and additionally, 1 side of these
discs were roughly ground with a hand piece (Kavo;
EWL, Leutkirch, Germany) using a conical diamond
bur (858 104 165524 014; Komet). Discs were cleaned
ultrasonically, first in acetone, and then in distilled wa-
ter. Group 3 discs were overglazed according to the
manufacturer’s instructions (Table II). Table III dis-
plays the surface finishes and the dimensions for each
group of specimens.
Veneering porcelain discs
Thirty discs of veneering porcelain (17 mm diameter ⫻
2 mm) were produced by use of a metal mold (18 mm
inner diameter and 2.5 mm high). A press with a plunger
diameter equal to the inner diameter of the mold, loaded
to a pressure of 6 ⫻ 102 kPa, was used to compact the
porcelain slurry (Fig. 2). Mold dimensions were larger than
the final size of the porcelain discs to compensate for ap-
proximately 26% volumetric porcelain shrinkage (5.8% di-
ameter and 20% in height) during the sintering process.12
The veneering porcelain powder was mixed with its
special carving liquid (Super Wet; Elephant Dental BV)
to obtain the correct consistency. To ensure a homoge-
nous sintering process during firing, the porcelain slurry
was poured into the metallic mold and condensed under

468 VOLUME 90 NUMBER 5

ISGRÒ ET AL
Fig. 4. Biaxial flexural strength (BFS) of 1-layer specimens versus surface treatments.
a pressure of 6 ⫻ 102 kPa. Excess water was removed
from the slurry with absorbent paper after the plunger
was withdrawn. Next, the porcelain disc was extruded
from the mold and replaced upside down in the mold.
The disc was pressed again and use of absorbent paper
was repeated. Porcelain discs were fired in a program-
mable and calibrated vacuum porcelain furnace (Austro-
mat 3001; Dekema, Freilassing, Germany) according to
the firing cycle shown in Table II.
The fired porcelain discs were examined with a ste-
reomicroscope (Olympus; Tokyo, Japan) at original
magnification ⫻20 to evaluate the specimens for small
surface cracks and flaws. Specimens displaying visible
defects were replaced. The discs were divided into 3
groups of 10, identified as Groups 4, 5, and 6. The fired
surfaces of the discs in Group 4 received no further
treatment; those in Group 5 and 6 were prepared and
finished in the same way as those in Groups 2 and 3,
which received, respectively, ground and overglazed
surface treatments (Table III).
Two-layer discs
The ceramic core discs in Groups 7 and 8 (17 mm
diameter ⫻ 2 mm) were reduced in height from 2 mm to
an even thickness of 1 mm and veneered with 1 mm
veneering porcelain. The thickness reduction was
achieved manually using a hand piece (Kavo EWL; Leut-
kirch, Germany) and a diamond conically shaped bur
(858 104 165524 014; Komet). The discs were ground
flat on both sides as described for Group 2. The sides of
the discs to be veneered were further manually roughly
ground with the same conically-shaped diamond. After
ultrasonic cleaning, first in acetone then in distilled wa-
NOVEMBER 2003
THE JOURNAL OF PROSTHETIC DENTISTRY
ter, a conditioner (Carrara Press Conditioner; Elephant
Dental BV) supplied by the manufacturer was applied to
the roughened surface (Table II), and the discs were
fired. The conditioner was used to improve the bond
between the ceramic core and the veneering porcelain.
The veneering dentin porcelain slurry was applied to the
ceramic core disc with a brush and gentle vibration, then
condensed under pressure, 0.5 ⫻ 102 kPa. Moisture was
removed with absorbent paper. Discs were fired twice,
first a dentin firing cycle, followed by a dentin correction
firing (Table II). The second firing compensated for
shrinkage that occurred during the first dentin cycle. At
the end of these firing cycles, the veneering porcelain
surface was ground flat and parallel to the external core
surface, with the cast-iron grinding wheel used as de-
scribed previously. The veneering porcelain surface was
further treated in the same way as the discs in Group 3
and then overglazed (Table I) according to the firing
cycle presented in Table II.
Disc dimensions were measured and recorded before
strength testing. The thickness in mm was measured in 5
different areas, randomly selected, with a digital microme-
ter (Mitutoyo; Tokyo, Japan). Diameters were measured in
3 equidistant locations around the periphery of discs using
an electronic digital calliper (Time Technology Europe,
Maastricht, Netherlands). Mean values for these measure-
ments were used to calculate the biaxial flexural strength
(BFS) value for each disc. The mean thickness and diame-
ter for each group of discs are shown in Table III.
Biaxial flexural strength test
The BFS test is widely used and is considered the
most reliable method of assessing the strength of brittle
469
THE JOURNAL OF PROSTHETIC DENTISTRY
dental ceramic materials.19 The BFS value of the ceramic
core, veneering porcelain, and 2-layer discs was deter-
mined by the ball-on-ring test method, using a universal
testing machine (Instron 6022; Instron Limited, High
3F共1 ⫹ v兲 · 关1 ⫹ 2ln(Ds/B兲 ⫹ 兵1 ⫺ B2/共2 · Ds2兲其 · 共Ds2/D2兲 · 共1 ⫺ v兲/共1 ⫹ v兲]
BFS ⫽
4␲T2
Wycombe, United Kingdom) at a cross-head speed of
0.5mm/min at a room temperature of 25°C. The me-
chanical test was carried out according to the ISO 6872
for dental ceramic materials.20
Before the test, all the specimens, except those in
Groups 1 and 4, were dried in an oven (Austromat
3001; Dekema, Freilassing, Germany) according to the
firing cycle in Table II. This procedure removed any
moisture introduced during surface treatment (for ex-
ample wet-grinding or ultrasonic cleaning). This proce-
dure conformed to standard designation F394-78 of the
American Society for Testing Materials (ASTM) for bi-
axial flexural strength of ceramic substrates.21
The ball on ring experimental design used in this
study (Fig. 3) conforms to the ASTM designation F
394-7821 with 1 exception. In the ASTM standard, the
supporting ring consists of a series of balls resting on a
circular groove in a steel block. Because this test requires
that the load be distributed uniformly over the support-
ing balls, the specimen discs must be absolutely flat.
Therefore discs with an untreated surface could not be
tested. The ring of balls of the ASTM standard was
replaced by a series of pins (Fig. 4). A circular slot was
machined in the top face of a stainless steel block. This
slot was filled with a soft, but also incompressible, type
of silicone rubber (PMC-121/30; Smooth-On, Easton,
Pa), which has a shore A hardness of 30 as specified by
the manufacturer.
The slot was covered by a 3-mm steel disc with a circle
of 16 holes for the pins. The silicone rubber in the slot
worked as a hydraulic fluid, distributing the load evenly
over the pins. The pins have rounded ends to guarantee
an accurate definition of the diameter of the support
circle.
With this method, the disc was supported by a ring 16
mm in diameter. It was centrally loaded, from the top
until failure, with a stainless steel ball 5 mm in diameter.
The experiment was designed so that the bottom surface
of the disc was loaded in tension with the maximum
stress occurring in a small circular area at the center,
eliminating edge failures. The diameter of this area
where the stress was uniform depended on the thickness
of the disc and the diameter of the loading ball. In this
470
ISGRÒ ET AL
study design, the diameter of this area was about two
thirds of the thickness of the disc.22 The load at fracture
was recorded and the BFS for each disc was calculated
using Shetty’s equation:23
(1)
In this equation, the BFS is calculated using the force
at fracture (F); Poisson’s ratio (␯); the diameter of the
support circle (Ds); the diameter of the disc (D); the
thickness of the disc (T), and the diameter (B) of the
circular area, with uniform stress on the bottom surface.
This equation assumes a uniform elastic modulus and
Poisson’s ratio throughout the entire disc, and for this
reason, it could not be used in this form to calculate the
BFS of the 2-layer discs. Therefore laminated plate the-
ory was applied.24 First, the position of the neutral plane
(hn) was calculated, and then, a value for elastic modulus
(E0) and Poisson’s ratio (␯0) was obtained, to represent
the entire 2-layer disc. Finally, the BFS of the specimens
was assessed by evaluating the following equation:
2E1共1 ⫺ v0兲BFS(T ⫺ hn)
BFS bi ⫽ (2)
E0共1 ⫺ v1兲T
Where BFSbi is the stress at the center of the bottom
surface, the BFS is the value obtained from Shetty’s
equation, T is the thickness of the disc, hn height of the
neutral line from the top, E1 and ␯1 are the elastic mod-
ulus, and Poisson’s ratio values of the bottom material.
The test was carried out as follows. The 1-layer speci-
mens in Groups 3 and 6, were tested by subjecting the
overglazed surface to tensile stresses; the 2-layer speci-
mens in Groups 7 and 8 (for example, ceramic core
veneered with porcelain) were tested by subjecting both
the overglazed veneering porcelain and the ground ce-
ramic core to tensile stress (Table III).
Statistical analysis
The BFS mean values for each group of 1-layer spec-
imens (for example, Groups 1 to 6) were subjected to
statistical analysis by means of 2-way analyses of variance
(ANOVA), with strength as the dependent variable. The
material (for example, ceramic core or veneering porce-
lain) was the first between-subjects factor, and the sur-
face treatment (airborne-particle abrasion, as-fired,
ground, or overglazed) was the second. The mean BFS
values of the specimens made of 2-layer materials (for
example, Groups 7 and 8) were statistically evaluated by
means of 1-way ANOVA. The layer of the specimens
VOLUME 90 NUMBER 5
ISGRÒ ET AL
Fig. 5. Biaxial flexural strength of 2-layer specimens compared with biaxial flexural of 1-layer discs.
tested under tensile stress (veneering overglazed, or core
ground) was the between subjects factor.
For all of the analyses, the results were considered
significant for P⬍.05. When the effects were statistically
significant, the data were further analyzed, using part of
the ANOVA approach for simple effects and pairwise
comparisons. The statistical analyses were carried out by
use of the statistical software (General Linear Mode, a
subprogram of the SPSS for Windows, version 9.01;
SPSS Inc, Chicago, Ill).
RESULTS
The mean BFS values and SD for each group of spec-
imens are shown in Table III. Two-way ANOVA
showed that the method-by-material interaction effect
was significant (P⫽.002). The strength of both materials
increased through the different surface treatments (for
example, airborne-particle abrasion [ceramic core] as-
fired [veneering porcelain], ground, and overglazed);
(Fig. 4).
The difference in BFS values between ceramic core
and veneering porcelain was not significant for the over-
glazed surface treatment. For the airborne-particle abra-
sion, as-fired, and the ground surface treatment; the
difference between the 2 materials was statistically sig-
nificant (P⫽.001). The different surface treatments were
also compared separately for each material. The statisti-
cal analysis showed that for each material the difference
in BFS value between the airborne-particle abrasion, as-
fired surface treatments, and the ground surface treat-
ment was quite small. Although the difference was not
significant for the veneering porcelain, it was for the
ceramic core material, however, only marginally
(P⫽.049). For both materials, the BFS values for the
NOVEMBER 2003
THE JOURNAL OF PROSTHETIC DENTISTRY
overglazed surface treatment were significantly higher
than for the ground surface treatment (P⫽.001).
For the 2-layer specimens, 1-way ANOVA demon-
strated that the mean BFS value obtained for the 2-layer
specimens with the ceramic core ground and tested un-
der tension was significantly lower than the BFS mean
value recorded for the 2-layer specimens with the ve-
neering porcelain overglazed under tension (P⫽.001).
This result is shown in Figure 5.
The mean BFS value of 1-layer ceramic core, ground,
was compared with that of 2-layer specimen with core
ground, and tested under tension. The BFS mean value
of 1-layer discs was significantly higher (P⫽.015) than
that of 2-layer discs. Finally, the mean BFS value of
1-layer veneering porcelain, overglazed, was compared
with the mean BFS value of 2-layer discs with the ve-
neering porcelain overglazed, tested under tension. The
difference was not significant (Fig. 5).
DISCUSSION
It has been well established that crack propagation
brought about by tensile stresses, can cause a brittle
ceramic to fracture.4 However, if the strength of the
material, affected by internal microstructure and surface
quality3 is improved, then the performance of all-ce-
ramic dental restorations should improve with respect to
fracture resistance.
The BFS test method is suitable for assessing the
effect of different surface preparations (for example, air-
borne-particle abrasion, as-fired, ground, overglazed)
on the strength of ceramic materials.20 In addition, the
silicone rubber in this experiment (Fig. 3) distributed
the load during the test uniformly over the support cir-
cle. Therefore, it was possible to test specimens that
471
THE JOURNAL OF PROSTHETIC DENTISTRY
were not absolutely flat and to obtain results that were
unaffected by this condition.
When the effects of surface treatments were com-
pared, the BFS values for the overglazed ceramic core
and veneering porcelain specimens were significantly
higher than for the other 2 surface treatments. The ce-
ramic core was stronger than the veneering porcelain,
when the airborne-particle abraded, as-fired, and
ground surfaces were compared. This finding may be
related to the methods used to process the materials and
on their microstructure. The ceramic core is processed
by the heat-and-press technique, which may have a pos-
itive influence on its strength. Drummond et al10
showed that a heat-pressed ceramic core compared with
conventional porcelains is stronger, because of a more
homogeneous distribution of the leucite crystals in the
glass matrix. This is brought about by the technique
used to process the material.
However, there was no statistical difference in BFS
values when the ceramic core was compared with the
veneering porcelain, which had been overglazed. The
strength of the veneering porcelain increased by 78%,
from 60 MPa, the BFS value for the ground surface, to
107 MPa. For the ceramic core the increase was only
23%, from 93 MPa, the BFS value for the ground sur-
face, to 115 MPa. (Table III).
The overglazing surface treatment is a routine proce-
dure in a dental laboratory. It produces crowns with
smooth and shiny surfaces and has a positive effect on
the BFS of ceramic materials.13 The term “overglazing”
describes the firing of a low-fusing colorless glass on the
ceramic core or veneering porcelain.12 This thin layer of
about 4 ␮m of glass, produced after 60 seconds of hold
time at the final temperature13 reduces the size of flaws
present on the surface (probably introduced during the
fabrication of the restoration), thus increasing the
strength of the materials.14,15 Furthermore, it is used to
provide large surface compression, which strengthens
the ceramic body. The coefficient of thermal expansion
(CTE) value of the overglaze material used in this study
is lower than ceramic core and veneering porcelain (Ta-
ble I). This means that on cooling, the underlying ma-
terial shrinks more than overglaze, placing the latter
under compressive forces. These strengthening methods
inhibit the propagation of cracks from the surface of the
specimens.3 However, the overglazed surface treatment
had more effect on the veneering porcelain than ceramic
core (Fig. 4). This is probably due to a large number of
surface defects present on the veneering porcelain,
which are introduced during the traditional sintering
process6 and repaired by the overglaze material.
Previous studies conducted by Hopkins16 and Zeng
et al17 observed that the veneering porcelain had a neg-
ative effect on the strength of some ceramic core mate-
rials. However, Hopkins16 did not indicate which side of
the double-layered specimens was under tensile stress,
472
ISGRÒ ET AL
whereas Zeng et al17 did not test the 2-layer specimen
with the veneering porcelain under tension. In this
study, the decline in strength of the 2-layer specimens
was only recorded when the ground ceramic core was
tested in tension (Group 8 in Table III). Specimens
consisted of 1 mm of ceramic core and 1 mm of veneer-
ing porcelain. The ceramic core specimens, before being
layered with veneering porcelain, were subjected to air-
borne-particle abrasion, cleaned with acetone, and then
washed with distilled water. Subsequently, the condi-
tioner was fired as described previously. This may have
improved the bond between ceramic core and veneering
porcelain and effectively transferred stresses from 1 layer
to the other. Under these circumstances, part of the
veneering porcelain may also be subjected to tensile
stresses, during testing, leading to a failure at even low
forces. This was demonstrated by comparing the
strength of the ground single-layer ceramic core (h⫽2
mm) (Group 2 in Table III) with double-layer speci-
mens of equal thickness, tested with the ceramic core
under tension (Group 8 in Table III). The single-layer
group of specimens was stronger than the double-layer
group. Another possible explanation for this finding is
that the surface finish of the ceramic core was ground
and the mean BFS value obtained from single-layer ce-
ramic core ground (Group 2) was significantly weaker
than overglazed specimens (Group 3). This is likely due
to the presence of surface defects. Moreover, the CTE
for the veneering porcelain was higher than for the core
material (Table I). This may induce residual tensile stress
in the veneer porcelain and, also, result in reduced
strength. On the other hand, when tested under ten-
sion, the 2-layer specimens with the overglazed veneer-
ing porcelain (Group 7 in Table III) proved to be as
strong as the overglazed single-layer veneering porcelain
(h⫽2 mm) (Group 6 in Table III). It appears that in this
situation, the thickness of the ceramic core did not affect
the BFS of the 2-layer specimens. Moreover, none of the
double-layer specimens presented any delamination af-
ter testing. By process of elimination, this result may be
related to the beneficial effect of overglazing, which has
already been described,13 and to the acceptable mis-
match CTE values between glaze material and underly-
ing structure.
The first hypothesis of this study, that the strength of
a 2-layer ceramic specimen depends on the strength of
the veneering porcelain and the surface finish provided,
was accepted, because the different surface treatments
did influence the biaxial flexural strength of both mate-
rials. The second hypothesis, that a 2-layer specimen is
weaker than a specimen of only 1 material, must be
rejected, in part, because the strength decreases in dou-
ble-layer specimens were only recorded when the
ground core was tested under tension. A limitation of
this study is that only 1 ceramic core and 1 veneering
porcelain were tested. Further research should be con-
VOLUME 90 NUMBER 5
ISGRÒ ET AL
ducted to evaluate how these results apply to other all-
ceramic systems with comparable material composi-
tions.
The clinical implication of this study is that the final
overglazed surface preparation may have a positive effect
on the longevity of single and double layer all-ceramic
restorations. The veneering porcelain may weaken
2-layer restorations.
CONCLUSIONS
The influence of different surface treatments and the
effect of veneering porcelain on the BFS of a heat-
pressed ceramic were evaluated; and, within the limita-
tions of this in vitro study, the following conclusions
were drawn:
1. Ceramic core material was significantly stronger
than veneering porcelain when equal surface treatments
were compared; airborne-particle abrasion (ceramic
core) versus as-fired (veneering porcelain) and ground,
respectively.
2. The overglazed surface treatment significantly in-
creased the BFS of the materials tested.
3. Under tensile force, the 2-layer specimens with the
overglazed, veneering porcelain, were significantly
stronger than the 2-layer specimens with the ground
ceramic core. In both specimens, strength depended on
the final surface preparation. In addition, the fact that
the double-layered specimens with ground ceramic core
showed a decrease in strength could also be attributed to
the reduction of its thickness (h⫽1 mm).
The authors thank Mr A. Werner for technical support and Mr
F. C. van Ginkel, for carrying out the statistical evaluation of the data.
The advice and support of Mr M. de Kler of Elephant Dental Industry
BV is also appreciated, as well as for supplying equipment and
material.
REFERENCES
1. Cattell MJ, Clarke RL, Lynch EJ. The biaxial flexural strength and reliability
of four dental ceramics—Part II. J Dent 1997;25:409-14.
2. O’Brien WJ. Strengthening mechanism of current dental porcelains com-
pendium 2000;8:625-30.
3. Anusavice KJ. Phillips’ science of dental materials. 10th ed. Baltimore: WB
Saunders; 1996. p. 583– 4.
4. Piddock V, Qualtrough AJ. Dental ceramics—an update. Review J Dent
1990;18:227-35.
5. Kingery WD, Bowen HK, Uhlmann DR. Introduction to ceramics. 2nd ed.
New York: John Wiley; 1976. p. 787.
NOVEMBER 2003
THE JOURNAL OF PROSTHETIC DENTISTRY
6. Piddock V, Marquis PM, Wilson HJ. The mechanical strength and micro-
structure of all-ceramic crowns. J Dent 1987;15:153-8.
7. Dong JK, Luthy H, Wohlwend A, Schärer P. Heat-pressed ceramics: tech-
nology and strength. Int J Prosthodont 1992;5:9-16.
8. Williamson RT, Kovarik RE, Mitchell RJ. Effects of grinding, polishing and
overglazing on the flexure strength of a high-leucite feldspathic porcelain.
Int J Prosthodont 1996;9:30-7.
9. Shareef MY, Van Noort R, Messer PF, Piddock V. The effect of microstruc-
tural features on the biaxial flexural strength of leucite reinforced glass-
ceramics. J Mater Sci Mater Med 1994;5:113-8.
10. Giordano R 2nd. A comparison of all-ceramic systems, Part 1. Gen Dent
1999;47:566-70.
11. Drummond JL, King TJ, Bapna MS, Koperski RD. Mechanical property
evaluation of pressable restorative ceramics. Dent Mater 2000;16:226-33.
12. Baharav H, Laufer BZ, Pilo R, Cardash HS. Effect of glaze thickness on the
fracture toughness and hardness of alumina-reinforced porcelain. J Pros-
thet Dent 1999;8:515-9.
13. McLean J, Hubbard JR, Kedge MI. The science and art of dental ceramics.
Chicago: Quintessence Publishing; 1979. p. 39 – 40.
14. Brackett SE, Leary JM, Turner KA, Jordan RD. An evaluation of porcelain
strength and the effect of surface treatment. J Prosthet Dent 1989;61:446-
51.
15. Leinfelder KF. Porcelain esthetics for the 21st century. J Am Dent Assoc
2000;131 Suppl:S47-51.
16. Hopkins K. An investigation into the role of porcelain thickness in deter-
mining the load-carrying capacity of porcelain laminates. Br Dent J 1989;
67:201-4.
17. Zeng K, Odén A, Rowcliffe D. Evaluation of mechanical properties of
dental ceramic core materials in combination with porcelains. Int J Prosth-
odont 1998;11:183-9.
18. Craig RG, Powers JM. Restorative dental materials. 11th ed. St Louis:
Elsevier; 2001. p. 553.
19. Ban S, Anusavice KJ. Influence of test method on failure stress of brittle
dental materials. J Dent Res 1990;69:1791-9.
20. International standard for dental ceramic. ISO 6872 2nd ed. 1995E. p. 6.
21. American Society for Testing and Materials. Standard test method for
biaxial flexural strength (modulus of rupture) of ceramic substrates. Des-
ignation: F 394-78 re-approved 1996. p. 3.
22. Westergaard HM. Stresses in concrete pavements computed by theoretical
analysis. Public roads. US Dept of Agriculture, Bureau of Public Roads.
1926.
23. Setty DK, Rosenfield AR, McGuire P, Bansal GK, Duckworth WH. Biaxial
flexure test for ceramics. Ceram Bull 1980;12:1193-7.
24. Timoshenko S, Woinowsky-Krieger S. Theory of plates and shells. 2nd ed.
New York: McGraw-Hill; 1959. p. 51–78.
Reprint requests to:
G. ISGRÒ
ACTA
DEPARTMENT OF DENTAL MATERIALS SCIENCE
LOUWESEWEG 1
1066 EA AMSTERDAM
THE NETHERLANDS
FAX: 31-0-206692726
E-MAIL: g.isgro@acta.nl
Copyright © 2003 by The Editorial Council of The Journal of Prosthetic
Dentistry.
0022-3913/2003/$30.00 ⫹ 0
doi:10.1016/j.prosdent.2003.08.003
473