Dental Research

Scanning electron microscopic evaluation of resin-dentin and calciumhydroxide-dentin interface with resin composite restorationsGuido Goracci*/Giovanni Mori*

Abstract Calcium hyclro.xide has been used as a liner in resin composite restorations toprotect the pulp. Recent research has demonstrated thai pulpal inflammation iscaused by microleakage of restorations and by the .subsequent passage of bacteria.The present study involved scanning electron microscopic observation of cross-sections of resin composite-dentin interfaces after the interposition of a layer ofcalcium hydroxide. A tiew-generation adhesive system that involves etching of thedentin was used. Ultrastructural analysis indicated that polymerization shrinkageofihe resin composite caused he separation of the calcium hydroxide from thedentinal surface, forming 8- to 15-iim-wide interfacial gaps in 100% of theareas studied. Where the adhesive was applied directly to dentin, it adheredclosely, forming a gap-free attachment with evidence of an acid-resistant hybridlayer (4 to 6 ^ m in thickness) and resin tags of various lengths thai hermeticallysealed the dentinal tubules. (Ouintessence lnt 996:27:29-35.)

Clinical relevance

Calcium hydroxide-based linings do not adhere tothe dentinal surface whereas those placed underresin composite with dentin bonding agents tend topull away from the cavity surfaces, leaving a gapbetween the lining and dentin. In turn, a gap-freeattachment, considered to be the best protection forthe pulp, is produced where the adhesive is putdirectly on the dentinal surface.

Introduction

For a long time, the chemical action of restorativematerials was considered to be the main cause of pulpalirritation. Lefkowitz et al' concluded that particularchemical substances contained in numerous resins havea toxic action on the odontoblasts and the underlyinglayers of the pulp. They even considered the residualmonomers after polymerization of the resins as pos-sible causes of puipal irritation.'

In 1967, Stanley et al' used resin composites and

' Departmem of Operative Deiilistr>. Universit> La Sapienza, Rums.Italy.

Reprinl requesis: Ptof Guido Goracci. Via Tagliamenlo. 50. 00198Rome, Italy.

observed a pulpal reaction that is similar to thatobtained when silicate cements are used, although thechemical properties of both materials are different. Itwas therefore considered necessary to protect thepulpodentinal organ with liner materials based oncalcium hydroxide, thus isolating the dentin from theirritant action of resins. Brnnstrm and Nyborg,^ in astudy on the pulpal response following the placementof resin composite restorations, observed that theinflammatory process is always associated with thepresence of bacteria under the restorations.' In sub-sequent studies, zinc oxide-eugenol cement was ap-plied over those restorative materials considered to betoxic, such as silicate cement and resin composite, toeliminate bacterial leakage. After this precaution, nopuipal irritation was reported; thus it has been shownthat microinfiltration between the cavity walls and therestoration represents the main cause of puipal irrita-tion.^

Further in vivo studies-^ "'^ have demonstrated thatdental materials such as silicate cement, zinc phos-phate cement, resin composite, and amalgam arebiologically compatible, even if applied directly to theexposed dental pulp, provided that the latter ishermetically sealed and protected from any subsequentbacterial leakage. Moreover, the dental pulp has beenshown to have its own reparative capacity, capable notonly of healing but also of producing a dentinal bridge

Quintessence International Volume 27, Number 2/1996 129

oracci/Mori

in the absence of calcium hydroxide.^ However, theproblem of marginai infiltration is still unresolvedbecause resin composites contract during polymeriza-tion, thus detaching from the walls of the cavity. It hastherefore been necessary to isolate the dentin withintermediate materials, so thai, after the passage offiuids in the contraction gap. there is no possibiiity ofbacterial penetration inside tbe tubules, causing post-operatory sensitivity, puipai complications, and sec-ondary caries.

New adhesive systems have been recently intro-duced to improve the adhesion of resin composites tothe dentinal surface and thereby to counteract theformation of marginai defects. These adhesive systemsprovide etching of the dentin and the combined use ofmonomers with bifunctional groups (primers) as wellas hydrophilic resins able to penetrate the dentinaltubules and to chemically and mechanically bind to theperituhular and intertubular dentin." Further, acidetching of the dentin eliminates the smear iayer'*' anddemineralizes the dentinal surface, thus allowing thepenetration of resin tags in the tubules and theformation ofa demineralized resin-dentin interdifhi-sion zonethe hybrid layer.-'---

In spite of these considerations, many dentistscontinue to have doubts about the toxicity of the acidetching of vitai dentin and about the sealing providedand prefer to protect the deepest parts of the cavitywith intermediate materials based on calcium hydrox-ide, beiieving that caicium ions confer protectionagainst dentinai etching and stimulate the depositionof reparative and sclerotic dentins.

The aim of the present study was therefore toobserve, by scanning electron microscopy, the inter-face between the adhesive resin and the etcheddentinal surface and the interface between calciumhydroxide and untreated dentin following apphcationand polymerization of the resin composite.

Method and materials

Ten noncarious extracted human third molarsbelonging to subjects ranging in age between 18 and 22years were used in the present study. Teeth were fixedby immersion in 10% neutralized formalin immediatelyafter the avulsion. After 48 hours, 2-mm-deep heelswere made on the buccal and lingual surfaces at thelevel of the ccmcntocnamel junction of each tooth. Thespecimens were then ground fiat with a water-cooledorthodontic cast trimmer (Whip-Mix) until the heelsdisappeared. The dentin was then ground with wet

silicon carbide paper, in decreasing grits (240, 400,and 600), in an Ecomet grinder (Buehler).

A light layer of self hardening calcium hydroxide(Dycal, LD Caulk) was applied over the treateddentinal surfaces, adjacent to the pulp. Tbe dentinalsurfaces were then treated with the Scotchbond MP(3M Dental) dentinal adhesive system in accordancewith the manufacturer's instructions. Thereafter, alayer of hybrid resin composite (ZIOO, 3 M Dental) ofnot more than 2 mm was applied to the dentinalsurfaces and then light cured for 40 seconds (XL3000,3M Dental}.

The teeth were embedded in a self-curing epoxy resinand sectioned with a microtome (Leitz 1600 Micro-tome) along the longitudinal axis, thus passing throughthe center of the restorations (Fig 1 ). Section surfaceswere cleaned with 10% orthophosphoric acid for 3 to 5seconds and quickly rinsed with air-water spray for 15seconds to remove the smear layer. AH the specimenswere dehydrated through increasirtg concentrations ofethyl alcohol (30%, 50%, 70%, 90%, and 100%) andcritical point dried (substitution of alcohol by carbondioxide). Specimens were mounted with silver pasteon metallic stubs, and coated with about 20 A ofplatinum (Edwards Sputtering 150 S). They wereobserved with a Cambridge 150 A scanning electronmicroscope (Cambridge Instruments) used at anacceleration voltage of 7 to 10 kV Scanning electronmicrographs were made along the resin composite-dentin margin and subsequently colored with a pre-viously described method.^-'

Results

The scanning electron microscopic analysis was initial-iy carried out to evaluate the relationships among theresin composite, calcium hydroxide, and dentinalsurface (Fig 2). The layer of calcium hydroxide had athickness of about 200 laminall the specimens (Fig 3).At a higher magnification, it was possible to observethe good adhesion of the resin composite to thecalcium hydroxide ( interfacial gaps of 2 to 3 |j.m wereobserved in approximately 10% of the areas studied)and to note the presence ofa fissure between calciumhydroxide and dentin in 100% of the areas studied(Fig 4).

In all the specimens, polymerization shrinkage ofthe resin composite caused the detachment of thecalcium hydroxide from the dentinal surface. Thus, itwas possible to observe the opened dentinal tubulesinside the microfissure (Fig 5). The detachment of thecalcium hydroxide occurred all along the dentinal

130 Quintessence International Volume 27, Number 2/1996

Goracci/Mori

Fig 2 Composite resin (R, calcium hydroxide (C], anddentin (D) of about 1 mm in thickness, and the pulpalchamber |P].

Fig 1 Schematic representation of a specimen afteriongitudinai sectioning, Caicium hydroxide is placed in theproximity of the pulp.

Fig 3 Caicium hydroxide is about 200 \im thick in all thespecimens.

Fig 4 Cise adhesion of resin composite (R) to calciumhydroxide (C) (interfaciai gaps of 2 to 3 |im were observedin oniy approximateiy lO^ c of the areas studied). Formationof an interfacial gap between caicium hydroxide and dentin(D).

Fig 5 Partiaiiy cohesive detachment of caicium hydroxidefrom the dentinai surlace.

Quintessence International Volume 27, Number 2/1996 131

Goracci/Wori

Fig 6 in some zones, the gap between dentin and calciumhydroxide is 20 xm

Fig 7 Calcium hydroxide (C) interposed between resincomposite (R) and dentin (D) on the lell; resin compositeapplied directly to tfie denlin on the right.

FJg 8 Wherever there is no caicium hydroxide (C). theresin composite (R) adheres perfectly to the dentinaisurface (Dl by means of the formation of a hybrid iayer (H].

Fig 9 Characteristic bond following the use of Scotch-bond MP in direct contacf with the dentin Structuraicontinuity between resin composite (R) and dentin can beobserved. Note the hybrid iayer (H) and ihe resin tagspenetrating the tubules (T).

surface, and the gap thus formed reached 20 [xm(Fig 6),

The point of passage between calcium hydroxide-treated zones and those in which the adhesive wasapplied directly to the dentinal surface (Fig 7) revealedthat, when the calcium hydroxide layer finished, theresin composite was able to adhere perfectly to thedentinal surface, forming a hybrid layer (Fig 8}.

The relationship between the resin composite andthe dentinal surface in the outermost areas wherecalcium hydroxide was not applied was also examined.Excellent adhesion of the composite resin-adhesivecomplex to the dentin, mediated by a hybrid layer, was

evident; the dentin appeared to be crossed by resin tagsthat penetrated the tubules for about 20 \im (Figs 9and 10). In fact, the treatment of the sectioned surfaceswith phosphoric acid involved the superficial demine-ralization of the untreated dentin for a depth of 2 \im,to expose the resin tags that had penetrated the tubulesand to uncover a transitional acid-resistant zonebetween resin and dentinthe hybrid layer. At a highermagnification, the structural continuity between resincomposite and dentin, as well as the hybrid layer,which was about 5 |im thick, could be clearly observed(Fig 11).

132 Quintessence international Volume 27, Number 2/1996

Goracci/Mori

Fig 10 lieft to nghH The reconstructive material, anacid-resjstant hybrid layer, the dentjn, and the tesin tagspenetrating the tubules lor about 20 ^m

Fii 11 Higher iTiagnificatton showing Ihe junction betweenresin composite (R) and dentin treated with Scotchbond MPadhesive system. The hybrid layer (H) presents a thicknessol about 5 |im and continues wilh the resin tags penetratingthe tubules (T).

Discussion

Good marginal adaptation of restorative rnaterialsreduces inicroleakage, staining, pulpal irritation, andrectirrent caries,-''Recently, newadhcsivesysterns withhigher bond strengths have been introduced to resistcontraction stresses developed by the resin compositeduring polymerization, thus preventing the detach-ment of the resin from the dentinal surface.

The new enamel-dentinal adhesive systems includethe acid conditioning of not only the enamel but alsoIhe dentin, with the complete removal of smear layerand smear plug considered obstacles to adhesion.-""'^The application of etchant to the enarnel removes theinterprismatic substance: at the level of the dentin, theetchant opens the dentinal tubules and detnineralizesthe dentinal surface, thus exposing the collagen fibersfor a depth of 3 to 10 |xm, depending on Pka, concen-tration, and time of application of the selected acid,-*"

The subsequent use of an intermediate agent (pri-mer), to be applied to dentin before the piacement ofan adhesive resin, promotes the wettabMity of the resin,which must undergo intimate adhesive relationshipswith the dentina! tissue and establish the conditions fora real resin-dentin chemical bond. In fact, the com-monly used primers contain biflinctional molecules{hydroxyethyl methacrylate, hydrophilic monomers),able to bind to both the dentinal surface and themethacr>'late ( hydrophobic) groups of the resins, thusmodifying the dentinal surface frorn a hydropiiiiic to a

hydrophobic one,- '^ Successively, the adhesive fluidresin (bonding agent) is applied and light cured.

Scanning electron microscopic examination of crosssections ofthe resin-dentin interface revealed, in 100%of the areas studied, a gap-free attachment betweenresin composite adhesive and detitin and a 4 to 6-[imhybrid layer. Thus, a chemical and micromechanicalbond was obtained, based on both the penetration ofresin tags inside the tubules and the impregnation ofthe collagen fibers by the adhesive monomer. Thisdemineralized dentin-resin interdifHtsion zone is acidresistant and has the double function of binding therestoration to the dentin and rendering it cariesresistant,-' This resin-dentin interdiusion zone notonly allows adequate mechanical retention of therestoration but also is likely to be provided with acoefficient of elasticity that contributes to the mainte-nance of the bond. Such an elastic zone should be ableto compensate for and uniformly distribute the stressformed by the contraction of the resin composite inrelation to the rigid dentinal substrate, thus enhancingthe marginal adaptation of the restoration,^*

The present study showed that intetiacial gaps of 8to 15 |irn wide were coincident with the presence ofcalcium hydroxide. In tact, the research dernonstratedthat, due to the poor adhesion of calcium hydroxide tothe dentinal surface, the polymerization shrinkage ofthe resin composite caused detachment of the calciumhydroxide, leading to the formation of an interfacial gapbetween calcium hydroxide and dentin. Therefore, it

Quintessence International Volume 27, Number 2/1996 133

Goracci/Mori

could be concluded that the application of calciumhydroxide underneath resin composite restorationsnot only is useless but also could be harmful if it leadsto the formation of a microfissure in the zones where,due to the number and orientation of the tubulesfavoring the passage of bacteria, it would be moreappropriate to obtain hermetic seal capable of pro-tecting the pulp. In fact, under the calcium hydroxidein all the specimens, unsealed dentinal tubules werepresent, as well as microfissures of about 10 \im. Incontrast, in the zones in which the adhesive had beenapplied in direct contact with the dentin, an excellentadhesion of the resin composite to the surface wasevident. This result is in agreement with that ofMcConnel et ai-' who reported that a material such asV.L.C. Dycal placed under resin composite with abonding agent tends to be pulled away from the cavitysurface, leaving a gap between the lining and thedentin.

Reinhardt and Chalkey-*" reported that calciumhydroxide does not adhere to the smear iayer or to thedentinal tubule complex. The present data showed thatthe bond strength of adhesive resin to calciumhydroxide was higher than the bond strength ofcalcium hydroxide to dentin. Interfacial gaps only 2 to3 jim wide were observed between calcium hydroxideand adhesive in approximately 10% of the areasstudied, whereas 8 to 15-nm interfacial gaps betweencalcium hydroxide and dentin were observed in 100%of the cases.

According to Brnnstrm et al.-" such a microfis-sure will be quickly colonized by pulpal fluids, and,because of inward and outward fluid movement, mostof the calcium hydroxide paste will disappear withtime. In addition, the dissolution will happen morequickly, the greater the width of the marginal fissure.'This observation agrees with clinical reports ondisappearing Dycal.-'' In fact, the fluid present in thedentinal tubules of vital teeth is in a continuousoutward now because of a gradient of pressure. Thepresence of Gram-positive bacteria inside these spaceshas been reported; therefore, calcium hydroxide doesnot constitute an adequate long-term protection in thepresence of a marginal defect because it does notpresent an obstacle to the passage of bacteria or theirmetabolic products to the pulp.^'

Therefore, according to the present results, thenegative consequences of the interposition of a layer ofcalcium hydroxide between dentin and adhesive areevident, because the layer will detach from the dentinal

surface and will notably reduce the dentina! surfaceavailable for adhesion of the restoration.

In the present in vitro study excellent and con-tinuous attachment was evident wherever adhesiveresin was directly applied to flat surfaces of con-ditioned dentin. Further in vivo studies using wide,disk-shaped cavities are necessary. In more concavecavities, contraction becomes hindered in the threedimensions, shrinkage stress will be less compensatedfor by flow, and not only calcium hydroxide-dentin butalso the resin composite-dentin bond may be disrupted.

References1. Letlowilz W. Seelig A. Zachinshy L. PJ IP response lo self-turing

acrylie niling material. NY Dent J 1949; 15:373-386.2. Stanley HR. Swerdlow H. Stanwich L. Snare; C A comparison of

the biologieal effeets of filbng malerials with recommendations forpulp protection. Am Acad Goid Foil Oper 19A7; 12:55-63.

3. Brannstrm M. Kyborg H. Pulpa] reaction to composite resinrestorations. J Prostliet Dem 1973:27:181-189.

4. Briinsitm M. Dentin and Pulp in Restorative Dentistr>'. London:Wolfe Medical. 1982.

5. Asmussen li. Uno S. Adhesion of restorative resins to denlin:Chemical and physiochemical aspects. Oper Dent 1992:(SLLppl5):6S-74.

6. Beri;enholtz G. Cox CF. LoescheWJ, Syed SA. Bacterial leakagearound dental restorations: Its effect on the dental pulp. J OriilPalhol 19S2:11.439-450.

7. Cox CF, Bergenhoiti G, Fitzgerald M. Heys DR, ileys RJ. BaktrJA. Pulp capping of dental pulp mechanically exposed to oralmicroflora: A 5 week observation of wound healing in the monkey.J Oral Pathol l982;ll:327-339.

8. Cox CF, Bergenholtz G, Heys DR, Syed SA, Fitzgerald M, Heys RJ.Pulp capping of dental pulp mechanically exposed to oral micro-flora: A 1-2 year observation of wound healing in the monkey. J OralPalhol I985;I4:156-I68.

9. Cox CF. Biocompatibility of dental materials in absence of bacteria!infection. Oper Dent 1987; 12:146-153.

lU. Cot CF. Feltoii D. Bergcnholti G. Histopathological response ofinfected cavities treated with Gluma and Scotchbond dentin bondingagents. AmJ Dent 1988;1;189-194.

11. Felton DA, Cox CF. Odom M, Kanoy BE. Pulpal response tochemically cured and experimental light-cured glass ionomer cavityliners. J Prosthet Dent 199h65:704-7l2.

12. Fusayama T. New Concepts in Operative Dentistry. Chicago;Quintessence, 1980.

13. Fusayama T. Factors and prevention of pulp irritation by adhesivecomposite resin reslorations. Quintesseiiee Inl 1987:18:633-641.

14. Inokoshi S, Iwaku M. Fusayania T. Pulpal response to a newadhesive restorative rebin ] Dent Res 19a2;61:1014-1019.

15. Kaiika J. Pulpal studies: Bioeompatibility or effectiveness ofmarginal seal? Quintessence lnt l99O;2l:775-779.

16. Pashley DH. The effects of acid-etching on the pulpodentineomplex. Oper Dent l992;!7;229-242.

17. Stanley HR. Pulpal consideration of adhesive materials Oper DentI992;(suppl 5); 151-164.

134 Quintessence International Volume 27, Number 2/1996

18. Torslenson B. Nordenvall KJ. Brnnstrm M. Puipai reaelion andmicroorganisms under t-learfil eomposile resin in deep e avilie s withacid elched dentin. Swed DeniJ 1982;6:167-176.

19. GoracciG. Mori G, Casa de'ManinisL.BazziiechiM.Olturazioniin eomposilo ed adesione alle strullure dentarie. Milan: Ed Masson.1994.

20. Pashley DH. Tlie smear layer: Physiological considerations. OperDent 19S4;(sijppl 31:13-19,

21. Nakabajiashi N. The hybrid layer: A resin-dintin eoinposile. ProeFtnDertSoc l992;SS(suppl l|;323-329.

22. Van Meetbeek B, Diiem A. Goret-Nieaise M, BraemM. LambrechtsP. Vanherie G. Comparative SEM and TEM examination of theultrastrtieture of the resin-denlin interdi Rus ion zone. J Dem Res1993:72:495-501.

23. Goracci G. Ba^zucchi M. Mori G. Casa de' Martinis L. In vivo andin viiro analysis ofa bonding agent. Qtiintessenee Inl 1994i25:627-635.

24. Nakabayashi N. Adhesive bondingI992;{siippl5):125-13.

.th 4-META. Oper Dent

25. KankaJ, A method for bonding to tuoth struetiire using phosphoricacid as a dentin enamel conditioner. Quintessence Int 199l;22:285-290.

26. FusayatTia T, Optimal cavity treatmenl for adhesive restorations.J Esthel Dem 1990:2:95-99,

27. Nakabayashi N. Nakamiira M. Yasuda N. Hybrid layer as adentin bonding meciianism. J Esthct Dent I99l;3:l33-i38,

28. Watanabe I. Nakabayashi N. Bonding durability of photocuredphcnyl-P in TEGDMA lo smear I ayer-retained bovine dentin.Otiintessenee Int l993;24:335-342.

29. McConnellRJ.BoksmanL. Hunter JK.Gratton DR. The effect ofrestorative materials on the adaptation of two bases and a dentinbonding agent to internal cavity walls. Quintessence Int 1986;17:703-710

30. Reinhardt JW, Chalkey Y, Softening efl'ects of bases on compositeresins. Chn Prev Dent 1983:5:5-12.

3 I. Brannstrm M. Mattson B. Torstenson B. Material techniques forlining composite resin restorations: A critical approach. J Dem1991:19:71-79.

32. Akester J. Disappearing Dycal, Br Dent J 1979:147:369-375. D

HARVARDsurvivesthe testof time

I Please ask for brochures

Richter & Hoffmann Harvard Dental-GesellschaftJohannisbcrger Strae 24, 14197 Berlin, Germany

Quintessence International Volume 27, Number 2/1996 135