Correlation of Topography to BondStrength of Etched Ceramic

Vincent Jardel, DCD, DU^Michel Degrange, DCD, DSO, DUO''Bernard Picard, DCD, DSO^Gérard Derrien, DCD, DSO, DUO''

Purpose: Bonding techniques are commonly used today to ensure the durability ofporcelain laminate veneers and ceramic iniay-oniays, which are estheticaiiy pleasingsolutions to the problem of dental restoration. Acid etching and silanization of porcelainhave been widely used to enhance the retention between bonding resins and ceramicrestorations. The purpose of this study was to evaluate the effect of surface modificationswith 10% hydrofluoric acid gel and a coupling agent la dual component silane) on thebond strengih of 2 dentai feldspathic ceramics (CC and I VSl bonded with an unfiiledresin (Super-Bond), Materials and Methods: Eighty samples of GC and eighty samples ofPVS were built for a mechanical study. Mechanical tests were conducted using a Lloyd T6000 R tensiie machine, which determined the bond strengths of the 2 ceramics afterdifferent surface treatments (etching or silanizationl. in the next part of the study, 10surfaces of CC and 10 surtaces of PVS were studied with a scanning mechanicaimicroscope to evaiuate the action of hydrofluoric acid gel on the roughness of the 2ceramics. Results: Etching ceramic with hydrofluoric acid gel increased the developedsurface of feldspathic ceramics, especiaily for PVS, but this treatment was not sufficient toobtain the highest shear bond strength. The highly positive influence of silanization wasshown, particulariy tor CC ceramic bonded with an unfilled resin. Conclusion: Silanecombined with the action of hydrofluoric acid gel is the most effective surface treatmentfor ceramic. Int j Prosthodor>t ! 999:13:59-64.

The bonding of resins and dental ceramics hasintroducecl new restorative techniques and

aroused considerable interest. The surface treat-ment of ceramic is directly linked to ifs structure, its

'Aisod3te Profesor, Department of Proslhodonlics, l-sculty ofDentistry, Breit Univeraty, Brest, France.

''Profeííor and Director. Laboratory of Biomaleriak, Faculty ofDerHlstry, Paris V Univenily, Parh, France.

'Professor and Oireclor. Laboratory of Biomaleriali, Fatuity ofDentistry, Paris Vit University. Paris, Prance.

''Professor and Director, Department of Pro^thodontics, facultyof Dentistry, Brest University, Brest, France.

Reprint requests: Dr Vincent Jardel, 1 rue Amirai Lacaze,29200 Brest, France. E-mail: Vincent.Jardel^univ-brest.fr

possibilities for selective etching and coupling, andits mechanical properties.' In 1981 Pereimuter andMontagnon- proposed etching ceramic with hydro-fluoric acid to introduce mechanically retentiveholes in the surface of the porceiain, which wouldstrengthen the bond. This surface treatment was ac-ceptable, but the concentration of the acid used inthe study varied from 7,5% to 10%, and the etchingtimes aiso ranged widely, from 2 minutes to 10minutes.^"^" it was then proposed that silanizationshould follow etching to increase the bond strengthof ceramic. A silane coupling agent, 7-methacryl-oxypropyi trimethoxysilane, is widely used in den-tistry to modify a giass surface. Once coated, thesud^ace has an affinity to polymers. Because of the

Volume 12, Number I, 1999 59 The International Journal of Prostliodnntics

Correioition of Topography to Bond Strength of Etched Ceramic

Table 1 Specifications of Products Used

Product Specitications Batch number

GC ceramic Feidspathic ceramic tor inlays andporceiain laminate veneers

PVS ceramic Feidspathic ceramic tor iniays andporceiain laminate veneers

Hydroiiuoric add Concentration: tO%Siiicoup A: ethyiacetate

B: methacryloxypropyltrimethoKysilane

Super-Bond Cathaiysl: tri-N-butyl borane(TBB)

Liquid: 4-META (monomer)Polymer: 4-META

No, 171061

No, 509815

GC Dental Industrial

SS White

SY-G-92-04372 SymphyseA: Ch-B 904026 Hearaeus KuizerB: Ch-B 28

Cataiyst: 009041 Sun MedicalLiquid: C 1892Polymer: 0OBO1

silica content of dental porcelain, use of a silanecoupling agent is an important factor in achievingbonding to porcelain.''' Wben the silane couplingagent is applied to a glass surface and subsequentlydried, the resulting condensation forms a strongchemical bond. The siiane-treated giass, with itsmethacrylate groups, can then form a bond with themethacryiate groups in the resin.^^•'^'^^•^^"^''Currently, many authors recommend etching porce-lains with hydrofiuoric acid before silane pretreat-ment.^''^^'^^-^"'^ Bonding techniques are routinelyused for fixed partial dentures, porcelain iaminateveneers, and ceramic iniays and oniays. The resultsof mechanicai studies of adhesion vary dependingon the materials investigated or the methods usedto evaluate them, and they are difficult to comparebecause of the unique designs used for in vitro re-search. Three-dimensional tactile profilometry canbe used as a nondestructive method to evaluate tbeaction of hydrofluoric acid gel on a ceramic sur-face.-'-^ Therefore, in this study, the adhesion of 2feidspathic dental ceramics, GC ceramic (CCDental Industrial) and PVS ceramic (SS White)(Table Î ), was evaluated using tensile strength tests.As these filled giasses contain a iow density of crys-taliine fillers, and a large vitreous phrase can be at-tacked using hydrofiuoric acid, a surface study wasaiso performed to evaluate the action of hydrofiuo-ric acid gel on the 2 ceramics.

Materials and Methods

Eighty sampies of GC ceramic and eighty samples ofPVS ceramic, randomly divided into eight subgroupsof twenty samples each, were selected for a study ofmechanical properties. Ten samples of CC ceramicand ten samples of PVS ceramic, randomly dividedinto four subgroups of five samples each, were se-lected for a study of surface topography. The ce-ramic test samples were built in two parts (Fig 1 ).

Metallic Portion

Base metal nickel-chromium (Ni-Cr) alloy rods 20mm iong and 5 mm in diameter were used (Rexil-lium 111, Jeneric Industries). These rods were formedfrom industrialiy calibrated 5-mm diameter Plexi-glass (Weber Métaux & Plastiques) tubes cut into20-mm sections, sealed with wax at eacb end, in-vested (Ceramigold, Whip-Mix), and then cast inRexillium ill alloy to create standardized bars. Themetallic portions were inserted into the jaws of thetensile machine.

Ceramic Portion

Modified feidspathic ceramic GC or PVS (Table 1)disks 4 mm high were used; they were fired to 1 ofthe tips of the metal rods described. After firing andremoving excess material, the thickness of the ce-ramic mass was adjusted with a Dap-U polisher(Struers). Each ceramic-tipped rod was positionedperpendicuiar to the surface of a 22G-grain abrasivedisk turning at 300 rpm and constantly irrigated withwater to prepare a flat, normal ceramic surface.

Two tests were performed in this study: mechan-ical tests to quantify tbe adhesive strength of twoceramics (GC and PVS) bonded to themselves, anda topographic study to characterize the effect ofetching on these ceramics.

Mechanical tests were conducted using a Lloyd T6000 R tensile machine Ü) Lloyd) fitted with a buiit-inextensometer (Lloyd 4500 WM, |J Lloyd). Ceramicsurfaces of the samples were either etched for 5 min-utes with hydrofluoric acid gel (Table 1 ) -'^-s.i'''ií'.i' orleft unetched and either silanized with Siiicoup (dualcomponent nonhydrolized silane, Hearaeus Kuizer)or ieft unsilanized (Table i).i .i5.ia,20-22,24.29 jhesamples were bonded with a 4-META bonding resin,Super-Bond (Sun Medical), a reference resin for un-filled bonding.'^-i'^^'^'^'^^ After bonding, the sample

The International lournal of Frostliodonticî 60 Volume12, Number 1,1999

Fig 1 Samples cf GC and PVSassembled for meohanicai tests.The metailic portions were in-serted into the jaws ot the tensilemachine, c = ceramic porijon; m =metallic portion; F = traction force.

lardei et ai Currelalion of Topography to Bond Strength of Etched Ceramic

Table 2 Results of Mechanical Stutjy of Bond Strength

Groups

ABCDEFGH

Ceramic

GCGCGCGCPVSPVSPVSPVS

Treatment

NoneEtchedSilanizedEtched and silanizedNoneEtchedSilanizedEtched and silanized

Mean (MPa)'

3,311 619.621.7

6,111,212,414,6

SD (MPa)*

6.2

2.6i-.O

'Mean and standard devialior (SD) ot 10 tests lor each group.

pairs were positioned with their ceramic parts on ahonding bench to ensure their coaxial alignment (Fig1). After assembly, the samples were stored in an in-cubator at 37°C and 100% humidity. Tensile strengthtests were performed 1 hour later. Because the ce-ramic-metal bond was stronger than the ceramic-ceramic bond tested, failure of the bond between the2 ceramic surfaces could be assessed, A totai of 80tests were performed (10 tests for each of 8 groups;see Tabie 2),

The topographic characterization of the surfacesof the 20 sampies was performed using 3-dimensional roughness measurements tai<en with ascanning mechanicai microscope (Talysurf) con-nected to an IBM PC/AT-compatibie computer. Thissystem provided point-by-point, iine-by-line scan-ning of surfaces using a pyramid-shaped diamondstylus with a 2-|jm curve radius, A test-sample surface of N^ points was described by piot-ting N profiies of N points each, spaced in steps "p."Three-dimensional cartography from the topogra-phic surface studies was plotted in real time. Pro-cessing data concerning surfaces and 3-D cartogra-phy were determined,^'''^^ This was accomplishedusing the perspective of a receding angie set at ß =45 degrees. Profiles were represented in planes (X,Z) perpendicular to the receding line (Y), Thus, eachpoint M (x, y, z) of the true surface referenced in O

(x, y, z) ordinates had a corresponding point M' (X,Z) on the plot; for example: X = x -i- yCosO and Z = z+ ySinQ, '''-^ The following calculations were made:

1, Rt: Total roughness measurement, which wasthe difference between the highest peak andlowest valley encountered (Rt = Z^.^^.^- Z^¡^).This parameter provided a general measure-ment of roughness,

2, Ra: Arithmetic roughness measurement, whichwas the arithmetic mean of Z measurements (Zmeasurements represented z values in O [x, y,z] ordinates), Ra is preferable to Rt, which hasno statistical value. A low Ra generally charac-terizes low roughness and was expressed in \¡m.

3, Rq: Mean square deviation. The Rq was ex-pressed in |jm,

4, Sk: Skewness, The Sk described the distributionsymmetry of peaks and valleys. The Sk was ex-pressed without units; for Gaussian distribu-tion"'28 Sk = 0,

5, Ek: Kurtosis, which described the accuracy ofGaussian distribution. The Ek was expressed with-out units: for Gaussian distribution^^'^^ Ei< = 3,

6, DS: Developed surface. The DS represented therelation between the true surface of a recorded3-D measurement and the reference surface,DS was expressed in percentages.

Volume 12, Number 1, 1999 61 The Inlerrationai Journal of Prostliodonlics

Correlation oí Topograptiy to Bond SIrengtIi oí Etched Ceramic ¡ardel et ai

Table 3 Summary of Mean Values and StandardDeviations Obtained from Surface Study

Group

1JKL

Ceramic

GC polishedGC etchedPVS polishedPVS etctied

Ra (pm)

Mean

Q.7t.Ot.Ot.6

SD

0.20.20.40.1

DS (%)

Mean

100.9103.5101.61091

SD

0.31.21.02.9

Five tests were performed tor each group ot five samples. Ra = ariiti-metic roughness measurement; DS = dei/eloped SLirface.

A statistical software package (Stat View II BrainPower) on a workstation (Macintosh PowerBook1400cs/166, Apple Computer) was used for a 1-way analysis of variance (ANOVA) and a Scheffe'stest at 95% confidence level for all of the results.

Results

Mechanical Study

The results of the mechanical studies are presentedin Table 2. For GC ceramic, there were significantdifferences (P < 0.0001) between the groups; A(GC poiished) =i C (GC silanized), A {GC polished)* D (GC elched and silanized), B (GC etched) Í C(GC silanized), and B (GC etched) * D (GC etchedand silanized}. For GC ceramic, etching is less effi-cient than silanization. This mechanical study as-sessed the main qualitative contribution of silani-zation for GC ceramic. For PVS ceramic, all bondstrength values were lower than the bond strengthvalues for GC. There were significant differences(P< 0.0001) between the groups: F (PVS poiished)^ F (PVS etched), F (PVS poiished) ^ G (PVSsilanized), and E (PVS polished) * H (PVS etchedand silanized). For PVS, etching is as efficient as5Ílanization. Statistically for PVS, there was no dif-ference between silanization and etching, but thecombination of the 2 treatments increased thebond strength.

Topographic Characterization

The results recorded for each ceramic are listed inTable 3. They represent the mean Ra and DS mea-sured for the 20 samples. The topographic studyquantified the etching action (DS parameter) andprovided qualitative characterization of the effect ofthe acid (Ra parameter). A 1-way ANOVA andScheffe's test at 95% confidence level showed asignificant etching action on PVS ceramic. Therewas a significant difference (P < 0.0001) betweenthe groups: 1 (GC polished) =i ) (GC etched), I (GC

polished) ^ L (PVS etched), | (GC etched) ^ I (PVSetched), and K (PVS polished) * I (PVS etched).Etching ceramic surfaces with a hydrofiuoric acidgel increased the developed surface by only 2.4%for GC ceramic and 7.3% for PVS ceramic, com-pared with polishing. These results varied with re-spect to the ceramics studied. When changes in thedeveloped surface were considered, PVS ceramicshowed the greatest surface changes because ofetching; conversely, GC ceramic was only slightlyaffected by hydrofluoric acid. Comparison of theroughness values measured for GC and PVS in thepolished state and in the etched state (Table 3)shows that etching had a greater effect on PVS, in-creasing its Ra by 69%, than on GC, increasing itsRa by only 40%.

Discussion

This study compared the resuits of conventionalmechanical testing and topographic characteriza-tion. Under identical bonding and surface process-ing conditions, GC ceramic held up better thanPVS ceramic in tensile tests (Table 2). This investi-gation showed that when GC ceramic and PVS ce-ramic were compared, GC ceramic provided thebest bond for porcelain laminate veneers and ce-ramic iniay-oniay restorations. Etching the ceramicwith hydrofluoric acid gel increased the deveiopedsurface and the Ra (Table 3), especially for PVS ce-ramic (see results, group L, Table 3), therefore sug-gesting that PVS ceramic was more sensitive toetching and its structure was better suited to thistype of surface treatment (Figs 2 and 3). The PVSceramic contained more vitreous phase than GCceramic; this is why the hydrofluoric acid treatmentwas more important. For PVS there was only asmall difference between bond strength after etch-ing (see results, group F, Table 2) or after silaniza-tion (see results, group G, Tabie 2); the associationof etching and silanization (see results, group H,Table 2) provided the highest bond strength.However, to obtain the best bond strength in themechanical study, etching was not sufficient (seeresults, groups B and F, Table 2). Hydrofluoric acidcreated retentive holes in the surface of the ceram-ics, but these cavities were not large enough me-chanically to ensure the penetration of the unfilledresin. Silanization with a dual-component, nonhy-droiized silane (Silicoup) increased the bondstrength and provided a desirable surface treatment,particularly for GC ceramic (see results, group C,Tabie 2). Because of its high wettability and chem-ical contribution to adhesion, the silane agent sub-stantially increased the reliability of the bonded

Journal of Prosthodontics 62 Volume12, Number!, 1999

Jardcl et al Correlation of Topography to Bond Strength of Etched Ceiamic

0.0 |j 765.0 pm

Fig 2 Three-dimensionai cartography tor test sampie made of polished PVS ceramic.

Fig 3 Three-dimensional cartography tor test sampie made o( etohed PVS ceramic.

Volume 12, Number 1,1999 6 3 The International Journal of Proithodontics

Correlation oí Topography lo Bond Strengih of Elched Ceramic lardel el al

ceramic joints, Silane provided a bond between tbeluting resin and ceramics. This coupling agent is alarge molecule; one end consists of a methacrylategroup, which is capable of bonding to the resin,and the other end is a silanol, which is capable ofbonding to the ceramic sud'ace. This investigationdemonstrated that etching the ceramic with hydro-fluoric acid gel is an insufficient inner surface treat-ment, Silanization is an easily implemented surfacetreatment in clinical practice. It is as effective oreven more effective than etching, A combination ofetching and silanization is probably the best meansof obtaining optimal bond strength for GC and PVSceramics (see results, groups D and H, Table 2),

It can be concluded that these investigationsprovided an innovative, comprehensive approachto evaluation of adhesive phenomena, because thedata generated enabled the bonding potential ofboth ceramics to be assessed without long, meticu-lous, and destructive mechanical tests. Mechanicalstudies were informative and useful for the evalua-tion of etching and siianization of ceramics. Thetopographic investigations conducted on CC andPVS ceramics supported results from the initialstudy of their mechanical properties.

Clinical Implications

Following try-in of porcelain laminate veneers andceramic inlays and onlays, the practitioner must treatthe inner surface of proslhetic restorations. However,the treatment chosen must be the most appropriateand ciinically easy to perform and depends on thetype of ceramic employed. The inner suKace of therestoration to be bonded should be etched with a hy-drofluoric acid gel, rinsed in ciear water whiie heldwith a pair of 3- or 4-prong piiers, and placed in dis-tilled water in an ultrasonic bath for 3 minutes. Afterthe restoration has dried, a dual-component silane(nonhydrolized) should be applied to the piece in 2coats using a brush. The solvent should be allowedto evaporate completely between the iayers and atthe end of the treatment. Both treatments carried outin succession are needed to obtain optimum results.This preparation should be made up by tbe practi-tioner just before bonding, after taking the necessaryprecautions for the use of hydrofluoric acid.

References

1. Slangel I, Nalbanson D, Hsu CS, Shear strength of the compos-ite bond lo eiched porcelain. J Dent Res 1987;66:1460-1465.

2. Pereimuter S, Morlagnon |. Reinforced jacl<et crowns. CahProthèse 19Bl;36:97-109.

3. Qlamia JR. Etched porcelain veneers: The current stale of theart. Quintessence Int l9a5;16:5- ]2.

4. Caiamia |R, Materials and technique for etched porcelain fa-cial veneers. Acta Odonlologia 1988;8] :48-51,

5. Calamia JR. Ciinicai evaiualiar of etched porcelain veneers.Am) r)ent1989;2:9-15.

5. Caiartiia IR. Anterior and post:erior higli resistance bondedporceiain. Clin Odontol 1990;11:53-63.

7. Faunce FR, Myers DR. Lamínale veneer restoration of perma-nent incisors. | Am Dent Assoc 1976;93:790-792.

8. Friedman M. Multiple potential of elched porceiain laminatevsneers. J Am Dent A Î M C 1907;(special issue) :B3E-87E.

9. iHeyde |B, Cammarato VT, A restorative system for the repairof defecls in anterior teeth. The laminate veneers, Deni ClinNorth Am 1981;25:337-345,

10. Howe DF, Denehy GE, Anterior fixed partial dentures utiiiz-ing the acid-etch technique and a cast metal framework. JProsl) ietDentl977;37:28-31.

11, Horn HR, A new lamination; Porcelain bonded to enamel.NY Stale Dent] 1983,49:401^03.

12 Horn Hf i . Porceiain iaminate veneers bonded to etc)iedenamel. Deni Clin North Am 1983:27:67]-685.

13. Malsumura H, Nakamura M, Nakabayashi N, Tanaka T,Atsuta M. Effect of silane coupiing agent and ferric chlorideon the bonding of porceiain, quartz and alumina with 4-META/MMA-TBB resin. Dent Maler I 1987;6:135-139.

14. MinkIR, TimmonslH. Laminate veneers Dent: Clin Norlh Ami 984;28:l 87-203.

15. Nakabayashi N. Effectiveness of 4-META/MMA-TBB resin andresin-bonded fixed partial dentures, in: DeGrange M, RouletI-F (eds). Min imal ly Invasive Restorations wi th Bonding.Chicago: Quintessence, 1997:177-183.

16. Nation W, ledrychowski |R, Caputo AA. Effects of surfacetreatments on the retention of restorative materials to dentin. |Proslhet Dent 19BQ;44:638-641,

17. Nowlin TP, Barghi N, Noriing BK, Evaiuation of the bondingof three porcelain repair syslems. | Prosthet Deni 1981,46:516-518,

IB. Pereimuter S, Launois C, Bonded laminate veneers. InformDent l9Ö7; l :13 -2 l .

19. Röchelte AL. Attachment o í a spiint to enamei of lower ante-rior teeth. I Prosthet Dent 1973;30:418-423.

20. Tjan AHL, Nemel3 H, A comparison of the shear bondstrength between two composite resins and two etched ce-ramic materials, Int I Prosthodont 1988,1:73-79.

2 1 . Degrange M. Coronal fiDing biomaterials: Criteria for choice.Real Clin 199O;l:9-25.

22. Launois C. Myloiys's treatment with bonded iaminates ve-neers. Inf Dent 1988;15:1203-1206,

23. Sub BL. All-bond-fourth generation dentin bonding system. ]FsthetDentl991;3:139-143.

24. Touat) B, Plissart-Vanackere A, Ceramic laminate veneers to-ward a reduced fixed partial demure. Real Clin 1990:1:51-65,

25. Stacey GD. A shear siress anaiysis of the bonding of porcelainveneers to enamei, J Prosthet Dent 1993:70:395-401.

26. Tirlet C, Bouter D, Degrange M. Fractuie mechanics appliedto bonded ceramics. J Biomater Dent 1994;9:15-23.

27. Degrange M, Wehbi D, Roques-Carmes C. Topographicalmodification of a casi sampie with reference to the patternsurface roughness,] Biomater Dent 1985:1:133-149,

28. Degrsnge M, Ramspacher V, Wehbi D, Roqucs-Carmes C,Topographic approach to the interfaces between dental stoneand class A alginates, | Biomater Deni 1985;1:263-275,

29. Tavernier B, Picard B, Ogoinik R. Mechanical fatigue ofbonded joints. I Biomater Dent 1991 ;6:165-169.

The Internation^i )ournai ol Proslhodontics 64 VDlume12, Numberi, 1999