Basic Research—Technology

The Efficacy of Different Sealer Removal Protocols onBonding of Self-etching Adhesives to AH Plus–contaminatedDentinSteven Roberts, DDS,* Jong Ryul Kim, DMD, PhD,

†Li-sha Gu, DDS, MS,

‡

Young Kyung Kim, DDS, PhD,§

Quinn M. Mitchell, BDS,jj

David H. Pashley, DDS, PhD,{

and Franklin R. Tay, BDSc (Hons), PhD*{

AbstractIntroduction: Smearing of unset root canal sealers overthe pulp chamber dentin may adversely affect bonding ofself-etching adhesives and jeopardize their coronal sealingpotential. This study examined the influence of differentsealer removal protocols on the microtensile bondstrengths of two self-etching adhesives to AH Plus-contaminated dentin. Methods: Coronal dentin surfaceswere prepared from extracted human third molars. In thepositive control groups, these surfaces were not contami-nated with sealer and were bonded with Clearfil SE Bondor Clearfil Tri-S Bond. For the experimental groups, dentinsurfaces were contaminated with AH Plus and wiped witheither dry cotton pellets, cotton pellets saturated withethanol, or cotton pellets saturated with Endosolv R fol-lowed by rinsing the dissolved sealer with water prior tobonding with the two adhesives. Bonded specimenswere sectioned into resin-dentin beams for microtensilebond strength evaluation. Additional specimens wereprepared for transmission electron microscopy to examinethe ultrastructure and nanoleakage within the hybridlayers. Results: For both adhesives, microtensile bondstrengths significantly declined when the sealer wasremoved with dry cotton pellets or cotton pellets saturatedwith ethanol. Only the Endosolv R/water sealer removalprotocol restored tensile bond strengths to those of theuncontaminated positive controls without adverselyaffecting hybrid layer formation in intact dentin orincreasing nanoleakage within the resin-dentin interfaces.Conclusion: The Endosolv R sealer removal protocolappears to be effective in preventing the deterioration ofbond strengths of the two self-etching adhesives to AHPlus–contaminated dentin and warrants further clinicalinvestigation. (J Endod 2009;35:563–567)

JOE — Volume 35, Number 4, April 2009 Se

Key WordsCoronal seal, Endosolv R, formamide, root canal sealer, self-etch adhesives

The significance of the coronal seal (1) in optimizing the success of root canal treat-ment has increasingly been recognized (2–9). Because coronal seals generated by

temporary restorative materials are less predictable (10, 11), dentin adhesives havebeen advocated for creating additional orifice barriers and sealing of pulp chambers(12–16) to protect endodontically treated teeth against bacterial invasion, endotoxin,and saliva infiltration from the coronal direction. This provides an economic yet prac-tical means to achieve bondable coronal barriers without resorting to proprietary adhe-sive root filling–sealer systems because the clinical supremacy of those systems has notbeen substantiated (17). Moreover, there is no additional bonding performed beneathor above the adhesive layer that forms the coronal seal. Thus, large C-factors and poly-merization shrinkage stresses associated with bonding to a long narrow tube-like spaceand direct bulk bonding over the cured adhesive are simultaneously reduced (18). Thismaximizes the likelihood of success of the coronal seal.

Immediately after root canal obturation, self-etching adhesives, which are rela-tively easy to apply and have reduced technique sensitivity, are frequently used forcreating a coronal seal (13, 19). Bonding of self-etch adhesives/resin cements iscompromised when they are applied to oil-contaminated or temporary cement-contam-inated dentin (20, 21). From a practical perspective, excess unset sealers that arepresent around filled canal orifices are smeared over the pulp chamber dentin duringthe application of self-etching adhesives. Although eugenol-containing cements havea negligible adverse effect on the strength of self-etching adhesives to dentin (22),the effect of dentin contamination by AH Plus (Dentsply Caulk, Milford, DE), an epoxyresin-based sealer, on bonding of self-etching adhesives is unknown. Recently, a sealerremoval procedure was proposed to facilitate coronal sealing of AH Plus–filled rootcanals. This procedure involves cleaning the AH Plus–contaminated pulp chamberdentin with Endosolv R (Septodont USA, New Castle, DE) before the application ofself-etching adhesives. Endosolv R is a softening solvent for phenolic resin-formalde-hyde sealers (23, 24). It contains formamide and 2-phenylethanol as cosolvents,with the former being the major ingredient (25). Although Endosolv R appears to beclinically effective in removing unset AH Plus sealer from pulp chamber dentin, the effectof this solvent blend on the bond strengths of self-etching adhesives is unknown. Thus,

From the *Department of Endodontics, School of Dentistry, Medical College of Georgia, Augusta, GA, USA; †Department of Conservative Dentistry, School ofDentistry, KyungHee University, Seoul, Korea; ‡Department of Operative Dentistry and Endodontics, Guanghua School of Stomatology, Sun Yat-sen University, Guang-zhou, China; §Department of Conservative Dentistry, School of Dentistry, Kyungpook National University, Daegu, Korea; jjDepartment of Oral Rehabilitation, School ofDentistry, Medical College of Georgia, Augusta, GA, USA; and {Department of Oral Biology, School of Dentistry, Medical College of Georgia, Augusta, GA.

Supported by funds provided by Dental Research Center, School of Dentistry, Medical College of GeorgiaAddress requests for reprints to Dr Franklin R. Tay, Department of Endodontics, School of Dentistry, Medical College of Georgia, Augusta, GA 30912-1129. E-mail

address: ftay@mail.mcg.edu.0099-2399/$0 - see front matter

Published by Elsevier Inc. on behalf of American Association of Endodontists.doi:10.1016/j.joen.2009.01.001

aler Removal Protocols and Bonding of Self-etching Adhesives to AH Plus–contaminated Dentin 563

Basic Research—Technology

the objective of the present study was to test the hypothesis that cleaningAH Plus–contaminated dentin with Endosolv R does not result in dete-rioration of the strength of resin-dentin bonds created by two self-etching adhesives.

Materials and MethodsForty recently extracted noncarious human third molars were

used in the study. A flat dentin surface was prepared perpendicularto the longitudinal axis of each tooth with a slow-speed Isomet saw(Buehler Ltd, Lake Bluff, IL) under water cooling. The cut surfacewas wet polished with 180-grit silicon carbide paper to create a bondingsubstrate in coronal dentin.

Self-etching Adhesives and Sealer RemovalA two-step self-etching adhesive (Clearfil SE Bond; Kuraray

Medical Inc, Tokyo, Japan) and a one-step self-etching adhesive (Clear-fil Tri-S Bond, Kuraray) were examined. For each adhesive group, thedentin specimens to be bonded were randomly divided into foursubgroups (n = 5). The teeth in the ‘‘uncontaminated’’ controlsubgroup were not contaminated with sealer and were bonded withthe respective self-etching adhesive according to the manufacturer’sinstructions.

The teeth in the three other subgroups were first contaminatedwith a layer of mixed AH Plus sealer that was spread evenly over thedentin surface and left undisturbed for 5 minutes. The unset sealer inthe first ‘‘cotton’’ subgroup was removed by using a size 2 cotton pellet.A visibly discernible smear of the sealer remained on the dentin surface.Unset sealer in the second ‘‘ethanol’’ subgroup was wiped with a size 2cotton pellet that was saturated with 95% ethanol. No additional rinsingwith water was performed; the latter appeared to be immiscible with thesealer-ethanol mixture. In the third ‘‘Endosolv R/water’’ subgroup, theAH Plus–contaminated dentin was first wiped with an Endosolv R–satu-rated size 2 cotton pellet. The sealer–EndoSolv R mixture was thenrinsed with deionized water for 3 seconds until the surface appearedvisibly clean.

After sealer removal, dentin surfaces were bonded with the respec-tive adhesive abiding by the manufacturer’s instructions. Light curing wasperformed by using an Optilux 500 halogen light-curing unit (Demetron/Kerr, Danbury, CT) at 600 mW/cm2. Incremental composite buildupswere constructed with EPIC-TMPT (Parkell Inc, Farmington, NY).

Microtensile TestingAfter water storage at 37�C for 24 hours, each tooth was sectioned

occlusogingivally into 0.9-mm thick serial slabs using the Isomet sawwith water cooling. The central slab was used for ultrastructural exam-ination. The two adjacent slabs were sectioned into 0.9 � 0.9-mmbeams, using the ‘‘nontrimming’’ technique of the microtensile test(26). The three longest beams from each slab were selected for tensiletesting, yielding 30 beams per subgroup. Each beam was stressed tofailure under tension by using a Vitrodyne V1000 universal tester(Liveco Inc, Burlington, VT) at a cross-head speed of 1 mm/min.Premature failures were assigned null strength values and included inthe statistical analysis. Beam dimensions were measured to the nearest0.01 mm to provide the bonding surface area for calculating tensilebond strength. Failure modes were examined by using a stereomicro-scope at 30� magnification and classified as adhesive, cohesive, ormixed failure.

Statistical AnalysisStatistical analysis was performed separately for each adhesive

because the intention of the study was not to compare bond strengths

564 Roberts et al.

of the two adhesives. A preliminary data regression analysis showedthat bond strength was not significantly affected by the origin of the teethfrom which the prepared beams were derived. Thus, each beam wasconsidered as a statistical unit. Because the normality (Kolmogorow-Smirnoff test) and homoscedasticity assumptions (Levene test) of thetensile strength data appeared to be violated, the four subgroups ineach adhesive group were analyzed by using Kruskal-Wallis analysisof variance and a Dunn multiple comparison test, with statistical signif-icance set at a = 0.05.

Transmission Electron MicroscopyThe central slab of each tooth (n = 5) was used for examining the

ultrastructure as well as potential regions of incomplete resin infiltra-tion (ie, nanoleakage) within dentin hybrid layers, according to themethods previously reported by Tay et al (27). Briefly, specimens forexamining ultrastructure only were fixed in Karnovsky’s fixative andosmium tetroxide, dehydrated, and embedded in epoxy resin. Speci-mens for nanoleakage examination were first immersed in 50 wt%ammoniacal silver nitrate tracer solution for 48 hours followed by expo-sure to fluorescent light for 8 hours to reduce diamine silver ioncomplexes into metallic silver grains before epoxy resin embedding.The extent of nanoleakage within dentin hybrid layers created in the‘‘uncontaminated’’ control and those created after the use of differentsealer removal protocols was examined qualitatively. For both ultra-structural and nanoleakage evaluation, undemineralized 90- to 100-nmthick sections were examined unstained by using a JEM-1230 transmis-sion electron microscope (JEOL, Tokyo, Japan) operated at 110 kV.

ResultsTensile bond strength results (TBS) are summarized in Figure 1.

For Clearfil SE Bond, tensile bond strengths for the four subgroups,uncontaminated control, cotton, ethanol, and Endosolv R/water, are(in MPa) 40.9 � 9.2, 28.1 � 12.6, 31.7 � 14.5, and 43.1 � 9.7,respectively. There was no difference between the uncontaminatedcontrol and Endosolv R/water subgroups (p > 0.05) and between

Figure 1. The efficacy of different sealer removal protocols on the tensilebond strengths of Clearfil SE Bond (a two-step self-etching adhesive) andClearfil Tri-S Bond (a one-step self-etching adhesive) to AH Plus–contami-nated dentin. For Clearfil SE Bond, subgroups with different upper case letters(ie, A, B) on top of the data bars are significantly different (p < 0.05). ForClearfil Tri-S Bond, subgroups with different numerals (ie, 1, 2, 3) on topof the data bars are significantly different (p < 0.05).

JOE — Volume 35, Number 4, April 2009

Basic Research—Technology

the cotton and ethanol subgroups (p > 0.05), with the rest of thecomparisons being statistically significant. For Clearfil Tri-S Bond,tensile bond strengths for uncontaminated control, cotton, ethanol,and Endosolv R/water subgroups are (in MPa) 36.5 � 11.1, 11.2 �11.7, 27.87� 10.7, and 35.0� 17.5, respectively. There was no differ-ence between the uncontaminated control and Endosolv R/watersubgroups; their TBS were significantly higher than the ethanolsubgroup (p < 0.05), which, in turn, was higher than the cottonsubgroup (p < 0.05). Most failures were mixed failures (>70%) exceptfor the Tri-S Cotton subgroup in which adhesive failure was the predom-inant failure mode (>90%).

The application of Endosolv R followed by rinsing with waterremoved most of the AH Plus sealer from the dentin surface. Thissealer removal protocol did not affect the ability of Clearfil SE Bond(Fig. 2A and B) or Clearfil Tri-S Bond to create hybrid layers in intactdentin. The extent of incomplete resin infiltration was qualitativelymanifested by the silver-impregnated regions with the hybrid layers.Nanoleakage was similar for all subgroups within one adhesive group

JOE — Volume 35, Number 4, April 2009 Sealer Removal Prot

(Fig. 3). In particular, the application of Endosolv R did not increasethe extent of nanoleakage along the bonded resin-dentin interface(Fig. 3A) when compared with that in the uncontaminated subgroup(Fig. 3B). Conversely, electron-dense fillers derived from the sealerwere always identified within the adhesive or composite in the ethanolor cotton subgroups of Clearfil SE Bond (Fig. 2C and D) and ClearfilTri-S Bond (Fig. 3C and D). Large voids representing dissolvedresinous sealer component were always identified adjacent to theAH Plus fillers in the two cotton subgroups (Fig. 2C and 3C). Thesevoids were less readily observed in the ethanol subgroups (Fig. 2Dand 3D).

DiscussionThe rationale for using coronal dentin instead of pulp chamber

dentin as a bonding substrate in the present study was to avoid intro-ducing unpredictable parameters such as curved surface areas, stressvectors, and altered dentin substrates (sclerotic and/or reparative

Figure 2. Representative transmission electron microscopy images taken from unstained, undemineralized sections of specimens bonded with Clearfil SE Bond.Similar ultrastructural features seen in the uncontaminated control and Endosolv R/water subgroups are collectively represented in A and B. Likewise, results fromthe cotton and ethanol subgroups were similar and are collectively represented in C and D. Generic descriptors: C, resin composite; A, filled adhesive; D, inter-tubular dentin. (A) Low magnification of the resin-dentin interface (Endosolv R/water) showing absence of AH Plus sealer remnants within the adhesive orcomposite (ie, original oxygen inhibition layer of the adhesive). The application of Endosolv R did not affect the ability of the self-etching adhesive to etch throughthe smear layer and create a 1-mm thick hybrid layer (open arrow). (B) High magnification (uncontaminated control) of the partially demineralized hybrid layer(between open arrows). P, smear plug; pointer, peritubular dentin; open arrowhead, adhesive nanofillers. (C) Low magnification view (cotton) showing AH Plusfiller remnants (arrows) and empty spaces (pointers) within the adhesive and composite. Open arrow: hybrid layer. (D) A high-magnification view (ethanol)showing entrapment of AH Plus fillers (arrows) within the filled adhesive. Open arrowhead: smear plug.

ocols and Bonding of Self-etching Adhesives to AH Plus–contaminated Dentin 565

Basic Research—Technology

Figure 3. Representative transmission electron microscopy images taken from silver-impregnated, undemineralized sections of Clearfil Tri-S Bond specimens.Similar nanoleakage seen in the uncontaminated control and Endosolv R/water subgroups are collectively represented in A and B. Generic descriptors: C, resincomposite; A, filled adhesive; D, intertubular dentin. (A) The Endosolv R/water subgroup showing nanoleakage (pointer) in the form of silver deposits within thehybrid layer. AH Plus sealer remnants were not seen within the adhesive or composite. The application of Endosolv R did not increase the extent of nanoleakagealong the bonded resin-dentin interface. (B) High magnification of the control subgroup showing the extensive nanoleakage (pointer) present within the partiallydemineralized hybrid layer created by this one-step self-etch adhesive. P, smear plug; open arrowhead, nanofillers. (C) The cotton subgroup showing electrondense AH Plus fillers (arrow) and resinous sealer components (open arrowheads) within the adhesive and composite. For this particular specimen, the sealerremnants spanned the entire thickness of the adhesive layer. Pointer: nanoleakage within hybrid layer. (D) The ethanol subgroup showing entrapment of AHPlus filler remnants (arrows) within the adhesive when the contaminated dentin was wiped with ethanol before adhesive application.

dentin), which may adversely affect the interpretation of bond strengthresults (28).

The silver penetrant used for nanoleakage examination facilitateda similar demonstration of voids in the manner that a contrast mediumis used in radiology. Unlike microleakage that is an indirect manifesta-tion of interfacial gaps, nanoleakage occurs in the absence of interfacialgaps (29). Incomplete infiltration of adhesive resins into the full depthof acidic resin monomer-demineralized dentin may leave exposedcollagen fibrils and cause seepage of water into these regions through20- to 100-nm-sized interfibrillar spaces. This may result in subsequenthydrolytic degradation of these collagen fibrils and the hybrid layer.

Cleaning AH Plus–contaminated dentin with dry or ethanol-satu-rated cotton pellets appears to be ineffective for creating coronal sealwith self-etching adhesives because the resin and/or the filler compo-nents of the sealer are retained within the resin-dentin interfaces. Theepoxy resin–based sealer is chemically distinct from the adhesive meth-acrylate resins, and the two classes of resin monomers do not copoly-merize. Unlike silanized adhesive nanofillers that are designed to

566 Roberts et al.

bond to the adhesive resins, nonsilanized fillers derived from the sealerare not bondable to methacrylate resins. Thus, remnant resin and fillercomponents from the AH Plus sealer may act as stress raisers thatconcentrate instead of dissipate stresses (30). This was reflected bythe significant decline in TBS after the adoption of these two sealerremoval protocols. Although the microtensile testing is quasi-static innature, the results suggest the possibility of a lower resistance to fatiguecrack growth during cyclic stresses or functional loading (31). This maylead to an accentuated breach of the coronal seal that allows ingress offluid, endotoxins, and bacteria through the root filling into the periradic-ular regions. The appearance of large voids adjacent to the remnantfillers further suggests that the epoxy resin components of the sealerwere insoluble in the self-etching primer/adhesive and were trappedwithin the adhesives after polymerization. They were probably dissolvedby the propylene oxide used for specimen processing (27), leavingbehind empty spaces on transmission electron microscopy examination.

EndoSolv R contains formamide and 2-phenylethanol as cosol-vents (29). It was originally designed for softening NaOH-catalyzed,

JOE — Volume 35, Number 4, April 2009

Basic Research—Technology

hardened Resorcinol-formaldehyde resin (Russian Red) (32). In thepresent study, Endosolv R was used as a solvent for incompletely poly-merized epoxy resin. The efficacy of a solvent to soften a polymer ordissolve a solute may be explained by the concept ‘‘like dissolveslike’’ (ie, polar solvents are better at dissolving polar compounds).Nonpolar solvents typically will not dissolve polar substances and viceversa. It may be assumed that water or ethanol alone is either immiscibleor incompletely miscible with the epoxy resin component of the AH Plussealer. Conversely, the epoxy resin is probably miscible with the form-amide/2-phenylethanol cosolvents, enabling it to be dissolved by the En-dosolv R. Because formamide and 2-phenylethanol are miscible withwater, this permits the dissolved epoxy resin to be rinsed away by water.

Within the limits of this ex vivo study, it may be concluded that theEndosolv R sealer removal protocol appears to be effective in preventingthe deterioration of resin-dentin bond strengths of the two self-etchingadhesives when they were bonded to AH Plus–contaminated dentin.However, it must be stressed that bond strength measurement is onlyan indirect surrogate for assessing the sealing properties of a coronalseal. Moreover, a statistically significant finding in this type of ex vivoresearch does not necessarily imply clinical significance. Thus, furtherclinical trials have to be performed to evaluate the clinical effectivenessof the Endosolv R sealer removal protocol on the integrity and longevityof coronal seal placed in endodontically treated teeth that are obturatedwith the AH Plus sealer. Although Endosolv R has a clinical track recordas an endodontic softening solvent, it is prudent to point out that form-amide is a potential tetratogen. Thus, female workers/patients of child-bearing age must be made aware that a known teratogen is being usedand appropriate ‘‘right-to-know’’ compliance policies be adhered to byproviding them with data from appropriate animal studies (33, 34).

AcknowledgmentsWe thank Drs Jason Ames and Brian Babb, endodontic resi-

dents from the Medical College of Georgia, who offered their clinicaltechnique for ex vivo evaluation. The self-etch adhesives andcomposites used in this study were generous gifts from KurarayMedical Inc, 3M ESPE, and Parkell, Inc.

References1. Swanson K, Madison S. An evaluation of coronal microleakage in endodontically

treated teeth. Part I. Time periods. J Endod 1987;13:56–9.2. Ray HA, Trope M. Periapical status of endodontically treated teeth in relation to the tech-

nical quality of the root filling and the coronal restoration. Int Endod J 1995;28:12–8.3. Tronstad L, Asbjørnsen K, Døving L, et al. Influence of coronal restorations on the

periapical health of endodontically treated teeth. Endod Dent Traumatol 2000;16:218–21.

4. Hommez GM, Coppens CR, De Moor RJ. Periapical health related to the quality ofcoronal restorations and root fillings. Int Endod J 2002;35:680–9.

5. Heling I, Gorfil C, Slutzky H, et al. Endodontic failure caused by inadequate restor-ative procedures: review and treatment recommendations. J Prosthet Dent 2002;87:674–8.

6. Iqbal MK, Johansson AA, Akeel RF, et al. A retrospective analysis of factors associ-ated with the periapical status of restored, endodontically treated teeth. Int J Pros-thodont 2003;16:31–8.

7. Siqueira JF Jr, Rocas IN, Alves FR, et al. Periradicular status related to the quality ofcoronal restorations and root canal fillings in a Brazilian population. Oral Surg OralMed Oral Pathol Oral Radiol Endod 2005;100:369–74.

JOE — Volume 35, Number 4, April 2009 Sealer Removal Pro

8. Sabeti MA, Nekofar M, Motahhary P, et al. Healing of apical periodontitis afterendodontic treatment with and without obturation in dogs. J Endod 2006;32:628–33.

9. Ng YL, Mann V, Rahbaran S, et al. Outcome of primary root canal treatment: system-atic review of the literature—Part 2. Influence of clinical factors. Int Endod J 2008;41:6–31.

10. Koagel SO, Mines P, Apicella M, et al. In vitro study to compare the coronal micro-leakage of Tempit UltraF, Tempit, IRM, and Cavit by using the fluid transport model.J Endod 2008;34:442–4.

11. Madarati A, Rekab MS, Watts DC, et al. Time-dependence of coronal seal of tempo-rary materials used in endodontics. Aust Endod J 2008;34:89–93.

12. Leonard JE, Gutmann JL, Guo IY. Apical and coronal seal of roots obturated witha dentine bonding agent and resin. Int Endod J 1996;29:76–83.

13. Wolanek GA, Loushine RJ, Weller RN, et al. In vitro bacterial penetration ofendodontically treated teeth coronally sealed with a dentin bonding agent. J Endod2001;27:354–7.

14. Galvan RR Jr, West LA, Liewehr FR, et al. Coronal microleakage of five materials usedto create an intracoronal seal in endodontically treated teeth. J Endod 2002;28:59–61.

15. Belli S, Zhang Y, Pereira PN, et al. Adhesive sealing of the pulp chamber. J Endod2001;27:521–6.

16. Schwartz RS, Fransman R. Adhesive dentistry and endodontics: materials, clinicalstrategies and procedures for restoration of access cavities: a review. J Endod2005;31:151–65.

17. Cotton TP, Schindler WG, Schwartz SA, et al. A retrospective study comparing clinicaloutcomes after obturation with Resilon/Epiphany or Gutta-Percha/Kerr sealer.J Endod 2008;34:789–97.

18. Tay FR, Pashley DH. Monoblocks in root canals: a hypothetical or a tangible goal.J Endod 2007;33:392–8.

19. Korasli D, Ziraman F, Ozyurt P, et al. Microleakage of self-etch primer/adhesives inendodontically treated teeth. J Am Dent Assoc 2007;138:634–40.

20. Chaiyabutr Y, Kois JC. The effects of tooth preparation cleansing protocols on thebond strength of self-adhesive resin luting cement to contaminated dentin. OperDent 2008;33:556–63.

21. Matos AB, Oliveira DC, Vieira SN, et al. Influence of oil contamination on in vitrobond strength of bonding agents to dental substrates. Am J Dent 2008;21:101–4.

22. Peutzfeldt A, Asmussen E. Influence of eugenol-containing temporary cement onbonding of self-etching adhesives to dentin. J Adhes Dent 2006;8:31–4.

23. Gambrel MG, Hartwell GR, Moon PC, et al. The effect of endodontic solutions onresorcinol-formalin paste in teeth. J Endod 2005;31:25–9.

24. Rosiers L, De Bruyne MAA, De Moor RJG. Solubility of AH 26 at different consisten-cies using Endosolv E and R root canal filling removers. Int Endod J 2005;38(suppl1):20–1 (abstr R62).

25. Endosolv R MSDS data. Septodont Corp. Available at: http://www.septodont.ca/Septodont/francais/pdf/ftss/endo/mendvr_f.pdf. Accessed December 21, 2008.

26. Pashley DH, Carvalho RM, Sano H, et al. The microtensile bond test: a review.J Adhes Dent 1999;1:299–309.

27. Tay FR, Pashley DH, Yoshiyama M. Two modes of nanoleakage expression in single-step adhesives. J Dent Res 2002;81:472–6.

28. Leloup G, D’Hoore W, Bouter D, et al. Meta-analytical review of factors involved indentin adherence. J Dent Res 2001;80:1605–14.

29. Pioch T, Staehle HJ, Duschner H, et al. Nanoleakage at the composite-dentin inter-face: a review. Am J Dent 2001;14:252–8.

30. Pilkey WD, Pilkey DF. Stress concentration analysis and design, Peterson’s stressconcentration factors. 3rd ed. New York, NY: John Wiley & Sons, Inc; 2007:457–512.

31. Soappman MJ, Nazari A, Porter JA, et al. A comparison of fatigue crack growth inresin composite, dentin and the interface. Dent Mater 2007;23:608–14.

32. Schwandt NW, Gound TG. Resorcinol-formaldehyde resin ‘‘Russian Red’’endodontic therapy. J Endod 2003;29:435–7.

33. George JD, Price CJ, Marr MC, et al. Evaluation of the developmental toxicity of form-amide in New Zealand white rabbits. Toxicol Sci 2001;69:165–74.

34. National Toxicology Program. NTP Toxicology and Carcinogenesis Studies of Form-amide (CAS No. 75-12-7) in F344/N Rats and B6C3F1 Mice (Gavage Studies). NatlToxicol Program Tech Rep Ser 2008;541:1–192.

tocols and Bonding of Self-etching Adhesives to AH Plus–contaminated Dentin 567