Morphological characteristics of the interface between resin composite andglass-ionomer cement to thin-walled roots: A microscopic investigation

ABSTRACT: Purpose: To identify how different treatments of the root dentin surface affect the microscopic appearance ofthe resin composite/glass-ionomer cement-to-dentin interface. Methods: The root canals of 70 extracted human single-rooted teeth were enlarged to reduce dentin wall thicknesses to 0.5 mm. The roots were randomly divided into seven testgroups (n= 10) according to the canal irrigant used: no irrigant (control), 5% hydrogen peroxide, 5% sodium hypochlorite,a combination of 5% hydrogen peroxide and sodium hypochlorite, 15% ethylenediaminetetraacetic acid (EDTA), 10%lactic acid, or 20% lactic acid. To simulate thin-walled roots, within each group, crowns were sectioned and the entiresurface of each root canal space was enlarged with ProfIle instrument. Half of treated root canals (n= 5) were fIlled withresin composite (PermaFlo) and the other half were filled with glass-ionomer cement (Fuji IT Le). A light-transmittingplastic post (Luminex) was used to create space for a fiber-reinforced post and to ensure polymerization of the restorativematerial. Specimens were critical-point dried and freeze fractured for scanning electron microscope analysis. Three sitesalong the root were evaluated (cervical, middle, and apical). Results: Scanning electron micrographs showed nodifferences in the morphology of the resin tags at the cervical, middle or apical levels with any of the irrigants or therestorative materials used. Also, no difference in surface topography was found within individual groups. A resin-dentininterdiffusion zone and resin tags developed after application of resin composite with lactic acid solutions and EDT A butnot with the glass-ionomer cement. (Am J Dent 2010;23: 103-107).

CLINICAL SIGNIFICANCE: The choice of irrigant acid seemed to promote formation of characteristic resin bonding featuresalong the resin/dentin interface. Lactic acid or EDTA irrigant might best prepare intraradicular dentin for resin compositebonding systems.

Introduction

The strength of an endodontically-treated tooth is directlyrelated to the amount of remaining sound tooth structure. 1 Inmany clinical situations the root has little remaining wall thick-ness due to immature development, caries of the canal wallarover-instrumentation. A thin residual root wall can seriouslycompromise the prognosis for long-term restorative success.

Decisions regarding the selection of materials and restora-tive techniques for restoring root canals with compromised wallstructure are made difficult by the number of options that havebeen proposed?-6 Moreover, doubts remain about bonding toroot dentin.7 Laboratory studies8-11evaluating the fracture resis-tance of root filled teeth have shown that resin composites andglass-ionomer cements reinforce remaining tooth structure bybonding to dentin. Several factors may have contributed to thediscrepancies in bond strength values, such as morphologicaldifferences between coronal and apical root canal dentin,12morphological variations,13 and polymerization contraction ofthe resin cement.14,15 Nevertheless, the efficacy of variousagents for cleansing root canals during and after endodonticinstrumentation has not been well studied.16,17

For a restorative material to reinforce the tooth, it mustbond to dentin.18,19An essential attribute of a good bond is theability of the restorative material to wet and infiltrate thedentin?0,21 Conditioning the tooth surface with an acid prior tobonding removes the smear layer, alters surface energy, anddemineralizes the dentin, exposing a fine network of collagenfibrils.22 Infiltration of this network with resin permits forma-tion of a resin-dentin interdiffusion zone with resin tags andadhesive lateral branches, thus creating micromechanical reten-

tion of the resin to the demineralized substrate.23,24In addition"acid conditioning removes surface contaminants before mate-rial placement, possibly permitting greater ion exchange andimproved bonding between the adhesive cement and the toothstructure.25,26 These solutions include proteolytic enzymes,27chlorine-releasing agents,28 chlorhexidine,29 citric acid,30 sodi-um hypochlorite,3l sulfuric acid,32 tannic acid,33 lactic acid/5

and ethylenediamine-tetraacetic acid (EDTA).34Resin composites may offer one solution, but an alternative

class of material, the glass-ionomer cements, may have somepotential for fulfilling this role. Glass-ionomer cements havebeen described as possessing the unique properties of self-adherence to enamel and dentin, release of anti cariogenicfluoride into adjacent tooth structure and a low coefficient ofthermal expansion similar to dentin?5.36 Using resin bonded todentin prior to the placement of glass-ionomer cement increasesits bond strength?7

Scanning electron microscopic investigations38,39have beenused to evaluate factors such as the bonding mechanism ofadhesive cements and post systems which may affect postretention. The efficacy of the bonding system to developmicromechanical retention can be evaluated by observinguniformity and quality of the resin-dentin interdiffusion zone,resin tag adhesive lateral branch formation within the lutingmaterial or at the interface between it and the cavity walls andthe post.40,41However, there have been few reports2,l1,42onstrengthening the remaining root dentin with restorativematerials before placing post-retained foundation restorations.Therefore, this study evaluated resin composite and glass-ionomer cement for the replacement of lost dentin tissues inflared root canals. Scanning electron microscopy was used to

Fig. 1. SEM photomicrograph of control (no irrigant) specimen showing noevidence of dentin tubule penetration or tag formation. A. Resin composite. B.Glass-ionomer cement. Orig. mag. xIOOO.

Fig. 3. SEM photomicrograph of specimen irrigated with 5% sodiumhypochlorite showing no evidence of tag formation by dentin tubule penetration.A. Resin composite. B. Glass-ionomer cement. Orig. mag. xl,OOO.

assess variations in surface topography and to assess thephysical evidence of bonding. The null hypothesis was thatsurface modification of root dentin with different regimenswould have no influence on tag formation of flow able resincomposite and glass-ionomer cement.

The study protocol was approved by the InstitutionalResearch Board of King Abdulaziz University, Jeddah, SaudiArabia. Seventy intact recently extracted human single-rootedteeth were debrided to remove remnants of periodontal liga-ments. The teeth were stored in distilled water with 0.1% thymoldisinfectant at room temperature and equally divided into seventest groups (n= 10) according to the irrigant used. The inigantsused were: no irrigant (control), 5% hydrogen peroxide(pharmaplanea

), 5% sodium hypocWorite (Sainsbury's bleach~, a50-50% combination of 5% sodium hypocWorite and 5%hydrogen peroxide, 15% ethylenediaminetetraacetic acid (EDTAenlargement'), 10% lactic acid, and 20% lactic acid.d

Crowns of the selected teeth were sectioned perpendicularto the long axis, 2 ± I mm coronal to the cemento-enameljunction with a 0.15 mm diamond wafering bladee in an Isomet1000 slow speed saw,d to provide root lengths of 13 ± 1 mm.Access to the root canals was gained with diamond rotarycutting instruments.f Canals were endodontically instrumented;no fixing solution was used to avoid affecting dentin wallssurrounding the root canal spaces. All roots were held by handduring instrumentation, and the plane of greatest curvature wasaligned parallel to the plane of fJle oscillation. An ISO size 15file (K-flex£) was inserted in the root canal until the tip of thefJle was just visible at the root apex. The working length wasdetermined by subtracting 1 mm from the total length of the fileinside the root canal. A step-back technique was used to enlargethe canals enlarged to an ISO size 50 file (K-flexg

). Again, eachcanal was irrigated with 3 mL of the assigned inigating solutionwhen there was a file size change and after filing was complete.This was accomplished using a syringe fitted with a 27-gauge

Fig. 2. SEM photomicrograph of specimen irrigated with 5% hydrogen peroxideshowing no evidence of dentin tubule penetration. A. Resin composite. B.Glass-ionomer cement. Orig. mag. xl ,000.

Fig. 4. SEM photomicrograph of specimen irrigated with combination of 5%hydrogen peroxide and 5% sodium hypochlorite showing no evidence ofdentin tubule penetration. A. Resin composite. B. Glass-ionomer cement.Orig. mag. xl,oOo.

Fig. 5. SEM photomicrograph of specimen irrigated with 15% EDTA showingcomplete penetration of resin composite along the widely opened dentin tubuleswith no evidence of dentin tubule penetration with glass-ionomer cement. A.Resin composite. B. glass-ionomer cement. Orig. mag. xl ,000.

needle placed passively in the coronal canal opening. Themaximum depth of preparation of the needle tip/file was 1-2mm from the apical foramen. During irrigations, roots wereheld vertically, apices down, to ensure apical penetration ofinigant solutions. After the last inigation, canals werecompletely dried with paper points.h Ketac-Endo Aplicapi rootcanal sealer was mixed according to manufacturer's directionand the canals were filled using a lentulo spiral} The apicalthird of a size 35 master gutta percha conek was coated with thesealer, and then fully seated to the working length. The gutta-percha was then removed from each canal to a point 5 mm fromthe apex using a Gates-Glidden drill (Lexiconh). To simulateextensive clinical structure damage, the entire surface of eachroot canal space was further enlarged to reduce dentin wallthicknesses to 0.5 mm using the profJle nickel titanium files to asize #40 .06 taper (ProFileh). 0.5 mm thickness of dentin waschosen to represent the worse-case clinical situation.2,42 Postspace lengths of 8 mm were created with residual dentin wallthickness of 0.5 ± 0.2 mm at the cemento-enamel junction. Thebuccal aspect of each root at points 2.5 and 5.0 mm apical tothe coronal sectioned surface was measured for unifonnity inthickness among the specimens. Again, each post space wasrinsed with 10 mL of the corresponding inigant for 30 secondsto remove any remaining sealer. Inigating solutions wereremoved from the canal with sufficient paper points tocompletely dry the canal surface.

Fig. 6. SEM photomicrograph of specimen irrigated with 20% lactic acid showing penetration of resin composite along the opened dentin tubules with no evidence ofdentin tubule penetration with glass-ionomer cement. A. Resin composite. B. Glass-ionomer cement. Orig. mag. x 1,000. C. Orig. mag. x3,500. Note loose collagenfiber network represents demineralized dentin matrix.

To standardize the bond to be solely through microme-chanical interaction when dentin was etched prior to theapplication of resin composite or glass-ionomer cement,43,44theroot canal spaces were prepared by etching the surface with32% phosphoric acid for 15 seconds applied with a plasticneedle-nose application tip, until excess was seen extrudingfrom canal space. This was followed by rinsing with water for30 seconds and air drying. OptiBond Solo Plusg bonding agentwas placed in the root spaces according to manufacturer'sdirections. Within each group, half of the enlarged root canalspaces (n= 5) were. filled with resin composite (PermaFlog

)

and the other half were filled with glass-ionomer cement (FujiII LC\ A light-transmitting 1.4 mm diameter plastic post(Luminexm) was used to create space for a fiber-reinforcedpost and to allow the use of light polymerizing restorative ma-terials. Flowable light-polymerizing resin composite (PermaFIo) or light cured reinforced glass-ionomer cement (Fuji IILC) was injected into the canal spaces using a 21 rnm needletip (NaviTipO). Then smooth light transilluminating posts(Luminex) were inserted and centered in the root spaces, andthe resin was compacted around the posts. The curing light(UltraLume LED 5°) was placed to the end of the smooth lighttransilluminating post to polymerize the resin composite bytransmitting light down the length of the post for I minute.45,46Light intensity output was monitored with a curing radio-meterO to be at least 750 mW/cm2

• Next, the smooth lighttrans-illuminating post was removed and light was applied foranother 20 seconds.

The prepared specimens were sectioned longitudinally forexamination. A notch was prepared on each external rootsection using a long, cylindrical diamond rotary cuttinginstrumentP in a high speed handpiece, with water spray coolantto facilitate freeze fracture. Specimens were fixed in 2.5%glutraldehyde solution,q for 24 hours, and then fixed in 0.1mollL phosphate buffered in 2.5% glutaraldehyde (pH 7.4) foran additional 24 hours.16 This process was essential for anaccurate examination of dentin morphology without water lossor dimensional changes during preparation for scanningelectron microscopy.!? The moist state of the dentin wasmaintained with the use of liquid carbon dioxide as atransitional fluid under 1,300 psi pressure (CPD-2r). Thespecimens were then freeze-fractured using a hammer andchisel placed at the previously prepared notches and mountedon aluminum stubs to evaluate the formation and uniformity ofresin-dentin interdiffusion zone and resin tags. Specimens weresputter-coated with gold-palladium alloy (Cressingtons sputter

coater), and observed by a single investigator with a scanningelectron microscope (Philips Electron Optics BVt

) at three sitesalong the root canal walls (cervical, middle, and apical) for theformation and uniformity of resin-dentin interdiffusion zoneand resin tags. The investigator was blinded to the treatment ofthe specimens.

Serial SEM photomicrographs at xl,OOO and x3,500 ori-ginal magnifications were made of the canal walls at three sites.No difference in the microscopic appearance of resin-dentininterdiffusion zone was noticed at different locations withinindividual groups. SEM analysis of control specimens at x1,OOOrevealed no tag formation for resin composite (Fig.· lA) orglass-ionomer cement (Fig. IE) groups. Smear layer obscuredthe orifices of the dentin tubules in these groups. The dentinsurfaces of specimens irrigated with 5% hydrogen peroxidesolution did not reveal any evidence of tag formation with resincomposite (Fig. 2A) or glass-ionomer cement (Fig. 2B). SEMsof dentin surfaces treated with 5% sodium hypochlorite werealso obscured with debris with no evidence of resin composite(Fig. 3A) or glass-ionomer tag formation along the dentintubules (Fig. 3B). Root canal dentin treated with a combinationof 5% sodium hypochlorite and 5% hydrogen peroxide (Fig. 4)appeared similar to those treated with the 5% sodiumhypocWorite solution.

With the use of EDT A, complete tag formation for resincomposite along the widely opened dentin tubules was evident(Fig. 5A). However, no evidence was observed for glass-ionomer cement (Fig. 5B). Appearance was similar for dentinsurfaces treated with 10% or 20% lactic acid (Fig. 6). At highermagnification, the loose collagen fiber network represented thedemineralized dentin matrix (Fig. 6C).

Discussion

The data supported the null hypothesis that the surfacemodification of root canal dentin with the different regimensevaluated appeared most conducive to the development of tagformation with flowable resin composite but not for glass-ionomer cement. The increased demand for clinically con-venient treatment to restore a severely weakened endodon-tically treated tooth with a flared root canal has providedclinicians with a plethora of post and core based restorativeoptions. 1 However, abundant choices can present an under-standably difficult situation for clinicians trying to select thebest materials and techniques for an optimal result.

Research21,22 supports the concept that acid treatment ofdentin removes the smear layer and hydroxyapatite from thetreated dentin, demineralizing the dentin to a certain depth,leaving behind a collagen rich network for interaction withadhesive resins. This process results in the formation of ahybrid, or resin-dentin interdiffusion zone.23,24The diameter oftubule orifices was increased after lactic acid and EDT Atreatments due to loss of peritubular dentin, so the area ofintertubular dentin decreased. In the SEM observations from aprevious study/6 the etched and deproteinized dentin exhibitedthis morphology. In resin composite restorations, a dentin bondis produced when the resin monomers infiltrate the dentintubules and collagen in demineralized dentin, producing ahybrid layer.21,22,24In the SEM observations of the currentstudy, dentin treatment with an adhesive and resin compositeexhibited this same morphology. Resin tags were intimatelyadapted to the dentin tubules but tags were not observed inglass-ionomer cement groups, although the physical aspect of aglass-ionomer bond to dentin, that one would expect to see in aSEM image certifying a possible cement/dentin bond. This canbe explained by the different handling characteristics, compo-sitions and properties between resin composite and glass-ionomer. These differences may have an effect on their adhe-sion to root canal dentin. Another explanation is the placementof dentin adhesive prior to the application of glass-ionomercements that result in the loss of direct contact between theglass-ionomer cement and the cavity wall. This barrier mayaffect the ion exchange that normally OCClli'S between a glass-ionomer cement and tooth structure when setting.

Dentin conditioning with lactic acid and the use of resincomposite appears to be an efficient method to reconstruct lostdentin tissues, This may have been achieved by the low pH(104) of the lactic acid.16 After etching, the adhesive infiltratesthe exposed collagen with hydrophilic monomers, which thencopolymerize with the subsequently placed adhesive resin.21,25

Resin tag formation involves penetration of adhesive resininto the dentin tubules.21,24-26The formation of resin tags and aresin-dentin interdiffusion zone is achievable with availableadhesives and resin composites,7 However, the bond strength isunpredictable and may be influenced by morphological featuresof the dentin? Ferrari et aZ12 found increased resin tags in thecervical compared to the middle and apical thirds. However,other SEM observations13 reported increased resin tags in theapical third of root dentin compared to middle and cervicalregions. Results of the current study showed no differencesbetween cervical, middle, and apical levels where a flared rootcanal prevents bonding agent accumulation in the apical third.

The results obtained from this study are introductory andcomparative. A limitation was the use of microscopic investi-gations that did not permit collection of numeric data orstatistical analysis. Moreover, tensile or shear strengths of thebonded interfaces were not evaluated. Future investigations areneeded to evaluate the fracMe resistance of internallyreinforced roots with flared root canals. Also, study is requiredto investigate the physical properties of the interface betweenroot dentin and restorative materials and to determine theoptimal dentin preparation protocol for intraradicular bonding.

a. Fresenius, Bad Homberg, Germany.b. Sainbury, London, UK.

c. Produits Dentaires, Vevey, Switzerland.d. Fisher Chemicals, Fair Lawn, NJ, USA.e. Buehler, Lake Bluff, IL, USAf. Brasseler USA, Savannah, GA, USA.g. Kerr, Romulus, MI, USAh. Dentsply Maillefer, Tulsa, OK, USA.1. 3M ESPE, 81. Paul, MN, USA.j. Henry J. Schein, Port Washington, NY, USA.k. Hygienic Corp, Akron, OH, USAI. GC America, Alsip, IL, USAm. Dentatus USA Ltd, New York, NY, USA.n. Ultradent Products, Inc, South Jordan, UT, USA.o. DemetronlKerr, Danbury, CT, USA.p. Monoesteril Diamonds, Cormano, Italy.q. Sigma-Aldrich Corp, St Louis, MO, USA.r. TED Pella Inc, Redding, CA, USA.s. Cressington Scientific Instruments Ltd., Watford, United Kingdom.1. Philips Electron Optics, BV. Achseweg Noords, The Netherlands.

Acknowledgement: Research supported by King Abdulaziz City for Science andTechnology, Riyadh, Saudi Arabia, Project number AT-25-93.

Dr. Ayad is Professor, Section of Restorative Dentistry, Prosthodontics andEndodontics, College of Dentistry, University of Tanta, Egypt and KingAbdulaziz University, Jeddah, Saudi Arabia. Dr. Bahannan is AssociateProfessor, Department of Oral and Maxillofacial Rehabilitation, College ofDentistry, King Abdulaziz University, Jeddah, Saudi Arabia. Dr. Rosenstiel isProfessor, Division of Restorative and Prosthetic Dentistry, College ofDentistry, Ohio State University, Columbus, Ohio, USA.

1. Rosenstiel SF, Land MF, Fujimoto J. Contemporary fixed prosthodontics.4th ed, St. Louis: ElsevierlMosby, 2006; 336-374.

2. Saupe W A, Gluskin AH, Radke Jr RA. A comparative study of fractureresistance between morphologic dowel and cores and a resin-reinforceddowel system in the intraradicular restoration of structurally compromisedroots. Quintessence Int 1996; 27: 483-491.

3. Fernandes AS, Sharat Shetty S, Coutinho 1. Factors determining postselection: A literature review. J Prosthet Dent 2003; 90: 556-562.

4. Scurria MS, Shugars DA, Hayden WJ, Felton DA. General dentist'spatterns of restoring endodontically treated teeth. J Am Dent Assoc 1995;126: 775-779.

5. Smith CT, Schuman NJ. Restoration of endodontically treated teeth: Aguide for restorative dentistry. Quintessence Int 1997; 28: 457-462.

6. Dietschi D, Duc 0,Krejci I, Sadan A. Biomechanical considerations for the.restoration of endodontically treated teeth: A systemic review of theliterature-Part 1. Composition and micro- and macrostructure alterations.QuintesselU:e Int 2007; 38: 733-743.

7. Vichi A, Grandini S, Davidson CL, Ferrari M. An SEM evaluation ofseveral adhesive systems used for bonding fiber posts under clinicalconditions. Dent Mater 2002; 18: 495-502.

8. Marchi GM, Paulillo LA, Pimenta LA, Lima FA Effect of different fillingmaterials in combination with intraradicular posts on the resistance tofracture of weakened roots. J Oral Rehabil2003; 30: 623-629.

9. Cormier C, Bums D, Moon P. In vitro comparison of the fracture resistanceand failure mode of fiber, ceramic, and conventional post system at variousstages of restoration. J Prosthodont 2001; 10: 26-36.

10. Wu X, Chan AT, Chen Ya-Ming, Yip K, Smales R. Effectiveness anddentin bond strengths of two materials for reinforcing thin-walled roots.Dent Mater 2007; 23: 479-485.

11. Katebzadeh N, Dalton BC, Trope M. Strengthening immature teeth duringand after apexification. J Endodont 1998; 24: 256-259.

12. Ferrari M, Mannocci F, Vichi A, Gqgidiaco MC, Mjor IA. Bonding toroot canal: Structural characteristics of the substrate. Am J Dent 2000;13: 255-260.

13. Gaston BA, West LA, Liewehr FR, Fernandes C, Pashley DH. Evaluationof regional bond strength of resin cement to endodontic surfaces. JEnnodont 2001; 27: 321-324.

14. Bouillaguet S, Troesch S, Wataha JC, Krejci I, Meyer JM, Pashley DH.Microtensile bond strength between adhesive cements and root canaldentin. Dent Mater 2003; 19: 199-205.

15. Sano H, Takatsu T, Ciucchi B, Homer JA, Matthews WG, Pashley DH.Nanoleakage: Leakage within the hybrid layer. Opel' Dent 1995; 20: 18-25.

16. Clarkson RM, Moule AJ. Sodium hypochlorite and its use as an endodonticirrigant. Aust Dent J 1998; 43: 250-256.

17. Ayad MF, Rosenstiel SF, Farag AM. A pilot study of lactic acid as a dentinconditioner for dentin-bonding agent development. J Prost!let Dem 1996;76: 254-259.

18. Weiger R, Heuchert T. Hahn R, Lost C. Adhesion of a glass-ionomer cementto human radicular dentin. Endod Dent Trallmatol1995; 11: 214-219.

19. Molla K, Park HS, Haller B. Bond strength of adhesive/compositecombinations to dentin involving total-and self-etch adhesives. J AdhesDent 2002; 4: 171-180.

20. Nakabayashi N, Nakamura M, Yasuda N. Hybrid layer as a dentin-bondingmechanism. J Esthet Dent 1991; 3: 133-138.

21. Gwinnett AJ. Altered tissue contribution to interfacial bond strength withacid conditioned dentin. Am J Dent 1994; 7: 243-246.

22. Drummond JL, Toepke RS, King TJ. Thermal and cycling loading ofendodontic posts. EliI' J Oral Sci 1999; 107: 220-224.

23. PasWey DH, Ciucchi B, Sano H. Permeability of dentin to adhesive agents.Quintessence lilt 1993; 24: 618-631.

24. Chappel RP, Cobb CM, Spencer P. Dentinal tubule anastomosis: Apotential factor in adhesive bonding? J ProsthetDellt 1884; 72: 183-188.

25 Ayad MF. Lactic acid root canal irrigation for dowel and core treatment: Apilot study. J Prosthet Dent 2004; 92: 540-545.

26. Carvalho RM, Yoshiyama M, Brewer PD, Pashley DH. Dimensionalchanges of demineralized human dentin during preparation for scanningelectron microscopy. Arch Oral Bioi 1996; 41: 379-386.

27. Tay FR, Pashley DH, Loushine RJ, Doyle MD, Gillespie WT, Weller RN,King NM. Ultrastructure of smear layer-covered intraradicular dentinfollowing irrigation with BioPure MTAD. J Endodont 2006; 32: 218-221.

28. Kuruvilla JR, Kamath MP. Antimicrobial activity of 2.5% sodiumhypochlorite and 0.2% cWorhexidine gluconate separately and combined,as endodontic irrigants. J Endodollt 1998; 24: 472-476.

29. White RR, Hays GL, Janer LR. Residual antimicrobial activity after canalirrigation with cWorhexidine. J Endodont 1997; 23: 229-231.

30. Yamaguchi M, Yoshida K, Suzuki R, Nakamura H. Root canal irrigationwith citric acid solution. J Endodont 1996; 22: 27-29.

31. Torabinejad M, Cho Y, Khademi AA, Bakland LK, Shabahang S. Theeffect of various concentrations of sodium hypocWorite on the ability ofMTAD to remove the smear layer. J Endodont 2003; 29: 233-239.

32. Von der Fehr FR, Nygaard 0B. Effect of EDT AC and sulfuric acid on rootcanal dentin. Oral Surg 1963; 16: 199-205.

33. Bitter NC. A 25% tannic acid solution as a root canal irrigant cleanser: A

scanning electron microscope study. Oral SlIrg Oral Med Oral Pathol OralRadial Endod 1989; 67: 333-337.

34. Kuah HG, Lni IN, Tseng PS, Chen NN. The effect of EDTA with andwithout ultrasonics on removal of the smear layer. J Endodollt 2009; 35:393-396.

35. Hotta M, Aono M. Adaptation to the cavity floor of the light-cured glass-ionomer cement base under a composite restoration. J Oral Rehabil 1994;21: 679-685.

36. Mitra SB, Kedrowski BL. Long-term mechanical properties of glass-ionomers. Dent Mater 1994; 10: 78-82.

37. Pereira PN, Yamada T, Inohoshi S, Burrow MF. Adhesion of resin-modifiedglass-ionomer cements bonding systems. JDent 1998; 26: 479-485.

38. Mannocci F, Innocenti M, Ferrari M, Watson T. Confocal and scanningelectron microscopic study of teeth restored with fiber posts, metal posts,and composite resins. J Endodont 1999; 25: 789-794.

39. Mitsui FH, Marchi GM, Pimenta LA, Ferraresi PM. In vitro study offracture resistance of bovine roots using different intraradicular postsystems. Quintessence lnt 2004; 35: 612-616.

40. Mannocci F, Ferrari M, Watson T. Stereomicroscopic and scanningelectron microscopic study of roots obturated with vertically condensedgutta-percha, epoxy resin cement and dentin bonding agent. J Endod 1998;24: 397-400.

41. Ferrari M, Mannocci F. A 'one-bottle' adhesive system for bonding·a fiberpost into a root canal: A SEM evaluation of the post-resin interface. liltEndodont J 2000; 33: 397-400.

42. Marchi GM, Mitsui FHO, Cavalcanti AN. Effect of remaining dentinestructure and thermal-mechanical aging on the fracture resistance of bovineroots with different post and core systems. lilt Endodollt J2008; 41: 969-976.

43. De Dunck J, van Meerbeek B, Yoshida Y, Inoue S, Suzuki K, LambrechtsP. Four-years water degradation of a resin-modified glass-ionomeradhesive bonded to dentin. EliI' J Oral Sci 2004; 112: 73-83.

44. Coutinho E, Van Landuyt K, De Dunck J, Poitevin A, Yoshida Y, InoueS, Peumans M, Susuki K, Lambrechts P, Van Meerbeek B. Vanherle G.Development of a self-etch adhesive for resin-modified glass-ionomers. JDent Res 2006; 85: 349-353.

45. Yoldas 0, Ala~am T. Microhardness of composites in simulated root canalscured with light transmitting post and glass-fiber reinforced compositeposts. J Endod2005; 31: 104-106.

46. Goncalves LA, Vansan LP, Paulino SM, Sousa Neto MD. Fractureresistance of weakened roots restored with a transilluminating post andadhesive restorstive materials. J Prosthet Dent 2006; 96: 339-44.