Effect of surface treatments of porcelain on the shear bond

strength of applied dual-cured cement

A. Nehir &den, PhD,a Funda Akaltan, PhD,a and Gtilsen Can, DDSb
Faculty of Dentistry, University of Ankara, Ankara, Turkey

The effect of different surface treatments of porcelain on the shear bond strength of
the applied dual-(cured cement was examined. Acid etching with hydrofluoric acid,
roughening the porcelain surface by means of grinding with a diamond bur, and the
use of silane coupling agent alone and with combinations of these surface treat-
ments were used. Silane treatment resulted in a shear bond strength on roughened
porcelain surfaces that was statistically significantly greater than all other
treatment groups Cp < 0.05). However the use of two mechanical bond systems
together with a chemical bond did not improve the shear bond strength. (J PROS-
THET DENT 1994;72:85-8.)

C omposite luting cements are used for cementation
of indirect porcelain restorations such as inlays, onlays, and
laminate veeners. Recently, dual-cured cements have been
introduced to eliminate some disadvantages of conven-
tional composite luting cements such as short working
time, the thickness of the material between dentin and
porcelain and the risk of fracture arising during the
cementation of indirect restorations.l, 2
Dual-cured cements that have been developed in the
combination of light-cured and conventional composite
luting cements contain the peroxide/amine components
present in chemical-cure systems in addition to the photo-
sensitizer camphoroquinone used in light-cured
A slow-acting peroxide/amine system has been used to
achieve an extended working time and to adequately
harden the surface of resin insufficiently exposed to light.
Dual-cured initiator systems have been found to improve
curing, solvent resistance, and the tensile strength of com-
posites2
The purpose of this study was to evaluate the shear bond
strength between porcelain and dual-cured cement with
different surface treatments of porcelain.
MATERIAL AND METHODS
Wiron 88 metal (Bego, Bremen, Germany) was used to
cast 77 metal buttons as the test samples. The buttons were
10 mm in diameter and 1 mm in thickness. With a brass in-
dex, feldspathic porcelain (Ivoclar, Liechtenstein) of 1 mm
thickness was applied on the flat surfaces of the metal but-
tons as recommended by the manufacturer.
aResearch Assistant, Department of Prosthodontics.
bProfessor, Department of Prosthodontics.
Copyright @ 1994 by The Editorial Council of THE JOURNALOF
PROSTHETICDENTISTRY.
0022-3913/94/$3.00 + 0. 10/l/54026
Then 77 samples were randomly divided into seven
treatment groups and the porcelain surface of the samples
received the following treatments:
(1) Silane coupling agent (S) applied
(2) Etching (E)
(3) Roughening (R)
(4) Silane coupling agent applied after roughening and
etching (RES)
(5) Silane coupling agent applied after roughening (RS)
(6) Silane coupling agent applied after etching (ES)
(7) Etching after roughening (RE)
In the first group 11 samples were coated with silane
coupling agent (Monobond S, Vivadent, Liechtenstein)
materials. according to manufacturer’s directions.
Eleven samples in the second group were etched with
36% (w/w) hydrofluoric acid gel (Ceramco, Johnson &
Johnson, New Brunswich, N.J.) for 10 minutes, rinsed for
30 seconds with tap water, and dried with oil-free air spray.
The porcelain surface of 11 test samples in the third
group were roughened with the use of a coarse diamond bur
(No:837-016, Northbel, Italy) moved in two different di-
rections at go-degree angles. Samples were then rinsed with
tap water and dried with oil-free air spray.
The samples of the other four groups were subjected to
different combinations of surface treatments by use of the
methods described.
A Teflon (Du Pont) index 8 mm in diameter was then
placed on top of the treated porcelain surfaces of the sam-
ples and dual-cured cement (Opaque dual-cured cement,
Vivadent, Liechtenstein) was bonded in accordance with
the manufacturer’s recommendations.
After all of the samples were polymerized with a curing-
light, they were placed in distilled water and stored at 37”
C for 24 hours, then thermocycled at temperatures between
6” C and 60” C with a total thermocycling of 100 cycles.
Samples were tested for shear bond strength by use of a
Hounsfield tensiometer (Tensometer Ltd., Croydon, En-
gland) with a crosshead speed of 2 mm/minute until fail-

JULY 1994 85
 QZDEN, AKALTAN, AlriD CAN

Fig. 1. Diagrammatic representation OT shear testing. u. Met& hutton: b, porcelain; c,

dual-cured cement; p ~ceo.

Table I. ANOVA summaq treatment groups (p <0.05). A statistically significant dif-

-~-. .
ference was also found between the groups with silane ap-
Source s,4. 1)F F P
--- _l_l-l___ - plied and the group with etching applied on the roughened
S.3rlace treatment 580.2 !: 10 4F” 0.000I surface (p <0.0.5). The silane coupling agent on intact por-
Errcr 663.5 6C ceiain surface showed higher mean shear bond strength
‘I’oLai 1244.6 66 compared wit.h the combination of roughening and etching.
s4lr,!l, = 0.05 No statistically significant difference was found among
the groups of silane, etching and silane combined, and
etching and the combinations of three kinds of surface
treatments (p <0.05). They showed comparable mean
ure (Fig. 1). The calculated shear bond strengths were de shear bond strengths. There was no statistical difference
termined by the force where bond failure nccurre& dividec among the last four treatment groups (Table III) (p cO.05).
5;; the bonding- area The use of silane on intact or etched porcelain surfaces
did not significantly effect the shear bond strength. 1Jsing
RESULTS the combinations of three kinds of surface treatments did
The groups of porcelain surface treatments were arm cot improT.re the mean shear bond strength.
lyzed by one-way analysis of variance (ANOVA) and Stu
dent-Newman-Keul’s multiple range test. The full-model DISCUSSION
AXOVA of the treatment groups indicat,ed a significant Dual-cured resin cements are used for cementing indi-
ditfyrence at A confidence levei of’&% (p ~10.5) among the rect restorations such as resin- or porcelain-bonded inlays,
shear bond strengths (Table 11. oniays, and veneers. These indirect restorations rely solel>
Mean shear bond strengths and standard deviations arc on the retention and support derived from mechanical
presenrrd in Table 11 and the results of Student-iGewmar.- and/or chemical bonding of the luting agent to the tooth
Keu!‘s multipie range test are seen in ‘Table III. Because and restoration. ‘Thus a principal factor in determining the
rougllrned test samples failed after tbermocycling and na long-term clinical success of veneers lies with tne resin lut -
data could he obtained. the ANOVA procedure was done ing cement and t,he strength and durability oi‘ its bond.”
oniv on the remaining six groups and this t:eatmenr grcu!. Ideally, luting agents should have good wetting capabil-
was not inciuded in Tables II and III. ities, be mechanically strong, resist v+ater sorption: and
The use of silane couplmg agent after mechanical rough wear, and be color stable and nontoxic. Physical prosartie;
ening yieided the highest mean shear bond stren@.h ant of the commercial materiais, including bond sicrength.s, iire
~I?(Ywc e statist.icall:v significant dil?nrenre Flour. al: other largei!- unknrwn ant‘ require independent :wessmw:
OZDEN, AKALTAN, AND CAN THE JOURNAL OF PROSTHETIC DENTISTRY

Table II. Surface treatment groups with mean shear bond strengths (MPa) and calculated standard deviations
Treatment MPa

Treatment symbol Roughening Etching Silane Mean SD

RS + - + 11.548 6.590
S - - + 6.500 3.560
ES - + + 5.586 2.022
E - + - 5.226 1.437
RES + + + 3.952 .I.476
RE + + - 1.873 1.424

Loss of indirect restorations, especially veneers, have Table III. Student-Newman-Keul’s multiple range test
been primarily attributed to an adhesive failure between for shear bond strengths (MPa) of six groups of surface
the resin-luting cement and the veneer. Although resin ve- treatments
neers rely on a chemical bond between the veneer and ce- Treatment Student-Newman-
ment, this study shows that a micromechanical retention symbol Mean N Keul’s groups
mechanism is valid for the porcelain veneers.
RS 11.548 11 A
Porcelain restorations may have some gaps, dependent
S 6.500 11 B
on particle size and condensation of the porcelain powder,
ES 5.586 11 B, C
and microcracks can occur during the firing procedure.4 E 5.226 11 %, c
These areas and gaps appear superficially after surface RES 3.952 11 B. c
treatments such as etching or grinding. Hydrofluoric acid RE 1.873 11 C
etching is a well-accepted pretreatment regimen that pro-
Key to Student-Newman-Keul’s groups: Groups with same letter do not
duces an irregular surface topography with large exposed differ significantly (alpha = 0.05).
quartz crystals on indirect porcelain restorations. However,
etching may not be sufficient for bonding; thus it is recom-
mended that, ideally, both a mechanical and a chemical
bond should be applied for all indirect porcelain restora- temperatures between 16’C and 60’ C with a total thermo-
tions.3 cycling of 100 cycles. No standard agingl, lo-l2 and thermo-
The effect of mechanical and chemical surface treat- cycling6 7s13,14procedure has been advocated in the liter-
ments on the bonding characteristics between porcelain ature.
and luting cements is not widely presented in the literature. Silane coupling agents with the general chemical formula
Because most clinical failures of indirect restorations result X-(CH2) Si-(OR) are capable of forming chemical bonds
from shear stresses, this investigation was planned to de- with both organic and inorganic surfaces.sThe use of silane
termine the effect of etching, roughening with a diamond as an organofunctional coupler between polymers and in-
bur, and applying silane coupling agent alone and in com- organic substances is well known. Stangel et al.l” investi-
binations of these three methods on the shear bond gated the effects of different bonding methods such as un-
strength between the porcelain and dual-cured cement. filled resin, silane, and silane with dentin adhesive on both
Previous studies evaluating bond strengths have differ- etched and nonetched porcelain. They concluded that
ent conclusions about the effect of aging and thermocycling composite resin bonded to etched porcelain both with
of the specimens. However, various terms of thermocycling silane and silane with dentin adhesive produced high shear
procedures are being recommended by many authors.5-s bond strengths. Boyer et a1.5measured the int’erfacial bond
Recently, the in vitro tensile and shear bond strengths of strengths of light-activated composites and they also
different porcelain repair systems were investigated and showed that use of a bonding agent on both uncut and
the results showed that thermocycling caused a statistically ground surfaces improved bond strengths. Shortah et a1.13
significant decrease in the mean bond strengths of some evaluated the marginal seal of an injection-molded ceramic
porcelain repair materialsss crown system used in combination with three adhesive lut-
Bailey9 also compared the flexural strengths of porcelain ing cements. Minimal microleakage was dem.onstrated with
bonded to composite resin specimens with four organosi- the composite luting cement used with specific porcelain
lane materials. They found that unhydrated specimens had and dentin bonding agents. Thus, the increa.se in dual-ce-
significantly higher bond strengths than hydrated speci- ment bond strength resulting from silane used on both in-
mens with ail products. Thus we stored our test samples in tact and roughened porcelain is not surprising.
37” C distilled water for 24 hours. To simulate clinical con- This finding is consistent with the study of Shortall et
ditions, it was decid.ed to thermocycle the test samples at a1.13and Stangel et al.r” Our results showed that the high-

JULY 1994 87
 OZDEN, AKAETAN, AND CAN

est bond st.rengths l,a.rrere

obtaineri wirk the use of’ siiane I\:: with diamond bur wou!d only consume time: it did not im-
roughened porcelain. prove the mean shea.r bond strengt.h significantly.
In view ofprevious investigations and our findings, it was 5. .4fter thermocycling, all of the test samples in the
c.nncluded that etching can contribute to bond strength group subjectec; to grinding faileci before shear testing.
between composit.e resin and porcelain, but siiane coupling
agent alone or with a?ly of the mechanica! bond s!rstenc her REFEHESCES
a superior effect on shear bond sLrcngth in vitro. 1.
It bvas expected that. :he use of !wo of the mechanica,
bonding procedures such as etching and roughening wit’. 2.
the diamond bur together with silane would increase t,hc:
3.
mean shear bond strength, but this procedure exhibiter.
lower shear bond st.rengths than that of silane on rougP-
eced porcelain. ‘Thus, combining two mechanical surface
Lreatments with siIal;e for composite !uting cement bond-
ing to porc&in k not recommendeb because it is tin?.
consuming.
The roughening procedure alone was not a reliable su;
face treatment of porcelain for honding with dual-curec 7.
cement. The cohesive failure that occurred in all of the-
loughened test sampies after thermocvclirlg indicated thaw
8.
qing and thermocycling may deoeasc the bond st.rength
of’ lud-cured cemeni md porcekk. 9.

10.

This study investigated the shear bond strength betwet> !1.

dual-cured cement and porcelain following difi’erent cu*
face treatmert.s applied on porcelain surfaces. 12.
1. The highest mean bond strength was achieved with the
use Gsilane togethc with a mechanicai -ouphening prccr
13.
dure.
:2. .4n kcreaaed d&-cured cement bond strength rc-
sul~rd wher: &lane was used on intact, porcelain compa:ec !I.

v:itI- -I.he boncj st-en& aft.er etchink rx: ~~ouphener! ~OICP-

!ain.
,3. Etching contributed bond strength hekveen compos
ire resin and po:ceiain, and silane coupiing agent togelhe!
wit!; et.ching resulteti :‘n a shea,. bon& strength similar tc
etcthing alone.
4. ?‘ile use of a siiane coupling agent t.ogether with twt
-rec+scicaI bond s:~r:tems SW+ as etching and rctigheninp