Effect of primer solvent and curing mode on dentinshear bond strength and interface morphologyElisa Velazquez, DDSVJayalakshmi Vaidyanathan, PhD^/Tritala K, Vatdyanathan,Milton Houpt, DMD, PhDVZia Shey, DMD, MSVStanley Von Hagen, PhD='

There is a need to study the rnain and interactive bonding effects of differences in solvent and curingmode used for adhesive monomers in dentin bonding systems. Objective: Two solvents (acetone andethanoi] and curing methods (iight cure, dual cure) were evaluated on their effects on bond strength andinteriaciai morphoiogy. Method atid materials: The adhesjves studied were based on two mohomers, py-romeiiitafe ol glyceryl dimethacrylate (PMGDM) and 2-hydroxy ethyl methacryiale (HEMA), Four groups ofeight teeth each were cut to expose planar dentin sections and treated with (a) iight-cure system with ace-tone as soivent (LCA group]; (b) iight-cure system with ethanoi as solvent (LCE group); (c) duai-cure sys-tem with acetone as solvent (DCA group]; and (dj dual-cure system with ethanoi as solvent (DCE group].The treated sections were tested tor shear bond strength to composite discs and interfacial morphoiogy.Results: The mean (standard deviation) of shear bond strength values (MPa) for the different groupswere; LCA; 11.8 (2.3), LCE: 12,7 (2.7); DCA: 24,9 (9,3); and DCE: 21,6 (9,6], Aii bonded sections werecharacterized by a similar hybrid iayer, resin tags, and overall interfacial morphology. Conciusion: Therewas a significant difference in shear bond strength as a function of cure mode, but not of solvent. Themean bond strength was higher for dual-cure systems studied. Oxygen inhibition effects may account forthe difference between light-cure and duai-cure types. (Ouintessence Int 2003,34:548-555)

Key words: cure effect, hybridization, shear bond strength, solvent effect

CLINICAL RELEVANCE: Bonding is a critical step in ad-hesive dentistry. Bonding with popular iight-cure dentin ad-hesive systems may yieid iess than the highest bondStrength that can be attained wi!h dual-cure systems. Theiight-cure bonding methods need standardization to re-duce oxygen inhibition or other adverse effeots causing re-duced bond strength even under wef bonding conditions.

'Graduate Student, Depaitment of Oral Biology, New Jersey Dentai Sciiool,University oí fifledicine and Dentistry of New Jersey, Newark, New Jersey.

^Associate Professor, Department ol Restorative Denlistry, New JerseyDertai School, University of f^edicine and Dentistry of New Jersey,Newark, New Jersey

'Protessor, Deparlment ot Restorative Denlislry, New Jersey Oertai Scfiool,University of Medicine and Dentistry ot New Jersey, Newark, New Jersey

Professor, Department of Pédiatrie Dentistry. New Jersey Déniai Sotiool,Liniversity of Medicine ard Denlistry of New Jeisey Newark, New Jersey.

^Assistant Professor, Deparlrrient of Piiarmacoiogy New Jersey MedicaiSchool, University of Medicine and Dentistry of New Jersey, Newark, NewJersey.

Reprint requests : Dr J. Vaidyanathan, Department of RestorativeDentistry-, New Jersey Dental SGhooi, Liniversity ot Medicine and Dentistryof New Jersey 110 Bergen Street, Newark, NJ 07103. E-mail: jvaidyan®umdnj edu

Bonding is a critical step in dentistry to restore dis-eased or damaged teeth to their anatomy and

function, and in itnproving esthetics of detitition forpatients. There have heen rapid advances in adhesivedentistry in recent years through a process known ashybridization of tissues with restorative systems inwhich adhesives are used to create a structural conti-nuity between the restorative and tooth tissues acrossa hybridized transitional zone,'-'' This transitional re-gion is generated by adhesive infiltration and in situcuring within a spatial network at the subsurface layerof the tissue created by etching and removal of thetninerali:¡ed tissue phase,' This process of demineral-ization of the tissue substrate and creation of tissue-adhesive hybrid layer connecting the intact tissue tothe restorative material is an important adhesive step(known as bonding) in a variety of restorative proce-dures in operative and prostbodontic treatment of pa-tients. This is especially important when restorativeprocedure involves dentin as a substrate for bonding-Dentin is a complex structure chemically and morpbo-logically, Cbemicaliy, its main constituents are a min-eralized bydroxyapatite structure, an organic fibrous

548 Volume 34, Number 7, 2003

• Veiazquez et ai

Structure of collagen, and water. Morpbologically, itcontains a matrix structure of intertubular and per-itubular dentin tbrougb wbich tubules leading to pulpare dispersed. Tbe tubular patbways to pulp dottingthe spatial structure of dentin contain dentinai fluid,which has traditionally posed eritical probietns forbonding of restorations to exposed dentin. Most hy-dropbobic resins fail to bond to sucb a substrate be-cause of limitation to wetting of sueh resins to the ex-posed tissue surfaee, wbich not only has low surfaceenergy due to a biofilm normally present on tbe tissuebut also because of the dentinal fluid. During me-chanical preparation of the cavity to remove carioustissue, a smear iayer is also produced. Attempts to usetbe smear layer as a substrate for bonding were notgenerally successful because of its poor cobesivestrengtb. In recent years, the smear iayer has been re-moved by etebing with an acid. The acid not onlyhelps remove tbe smear layer, but aiso demineralizesthe subsurface regions of tbe intact tissue.Demineralization creates an open spatial networkaround collagen fibrils in tbe intact tissue. In addition,it opens up tbe tubules. A bydropbilic monomer in awater-ebasing medium, such as acetone or ethyl alco-bol. is now used to prime the demineralized tissue sur-face, which is typically maintained moist lo avoid thecollapse of collagen fibrils that may occur under a"dry" environment.""" Tbe prepared tissue surface isreceptive to wetting and spreading under tbese eondi-tions, and the adhesive is able to infiltrate botb thespatial network around the coiiagen fibriis of tbe ma-trix structure, as well as tbe open tubules. Tbemonomer is now cured in situ. The cured structurethus contains a resin-tissue bybrid structure in the re-gion of adbesive infiltration into the tissue subsurfacesaiong with a pure resin layer, wbich is on tbe surfaceitself and often protrudes into the tubules.'-^-* Tbus astructural continuity is generated from tbe tissue tothe adbesive througb a transitional bybrid layer. Tbctransitional layer includes ¡a) the hybrid layer formedin the matrix surrounding the tubules, and (b) theresin tags wbicb protrude into tbe tubules. Interes-tingly, both the resin plugs and the hybrid layer haveimportant functions in improving the retention of therestorations (such as composite restorations, iniays,onlays, crowns, etc) bonded to such a structure. Theresin tags belp plug tbe openings into the tubules, andbelp seal patbways to the puip from toxins, bacteria,and otber contaminants from tbe oral environment.The hybrid layer helps integrate the tissue to therestoration through the adhesive and helps in furtherreducing dentin permeability and improving bondstrengtb by micromechanie and/or adhesive interac-tions. Tbe morpbologic features of tbe tissue adhesiveinterfacial region and the tissue adhesive bond

strength need critical examination under potentialvariability of bonding conditions used in restorativeprocedures. Severai authors have examined commer-cial and experimental adhesive systems to characterizetheir effect on morphologic and ultrastructural fea-tures across tissue-adhesive interface,^*"-'' permeabil-ity of dentin,'' interfacial fracture structural details,^solvent-induced dimensional cbanges,' deptb" andcompositional gradient- of infiltrated region, andmany other important properties.' There are currentlyno reports on tbe effect of cboice of primer solventand tbe curing mode on bonditig. In this study, tbe ob-jective is to examine the effect of typical variables,sucb as differences in solvent selection for a popularadbesive monomer system and differences in curingadditives used in tbe preparation of the formulations.The null hypotheses tested in the investigation were:(1) The shear bond strengtb (SBS) of the restorationto tooth and the hybrid layer morphoiogy are not sig-nificantly influenced by the soivent used in the adhe-sive system; and (2) the SBS and the hybrid layer mor-phology are not significantly influenced by tbe methodof curing.

METHOD AND MATERIALS

Bond strength measurements

Power analysis. The number of teeth studied in thebond strength analysis was estimated based on apower analysis using a three-factor design, which in-cluded solvent type, cure mode, and solvent type'xuremode. The power analysis used assumed values of a(level of significance desired), a (maximum expectedSBS variation), and S (minimum deteetable differencein SBS), and a minimum power of 70 /0. The foiiowingvalues were assumed (based on previously reportedSBS data in MPa): o = 0.05, o- = 5 (in tbe range 15 <20 < 25. corresponding to a mean SBS of 20 MPawith a standard deviation of ± 5), and S = 3. Tbe totalnumber of teetb required for bond strengtb measure-ments at 70% power was estimated to be 24. In tbe ae-tual investigation, a total of 32 teeth were used.

Sample preparation for bond strength measure-ments. The teeth were prepared with the Isometslow-cutting saw (Buehler) by removing tbe entire oc-ciusai portion perpendicular to its long axis to exposemiddle dentin, creating a flat surface. The exposed sec-tion was examined to ensure no pulpal horn exposure.The exposed surfaces were abraded with 600-grit sili-con carbide paper under a stream of water to producea uniform smear layer, and then rinsed witb water. Allteeth were acid-etched with 37% phosphoric acid for20 seconds and then rinsed for 20 seconds. The excess

Quintessence intemationai 549

•Veiazquezet ai

water was removed for 2 seconds witb a ligbt streamof air, leaving a moist surface, Tbe treated teeth wererandomly assigned to one of four groups as follows:group I (LCA) = received treatment witb tbe acetonesolvent primer in a ligbt-cure adbesive system (Bond1); group II (LCE) = received treatment witb tbeethanol solvent primer in a light-cure adbesive system(Bond 1 C&B); group III (DCA) = received treatmentwith the acetone solvent primer in a dual-cure bond-ing system (Bond 1 witb activator); and group IV(DCE) = received treatment witb tbe etbanol solventprimer and tbe dual-cure bonding system (Bond 1C&B with activator).

Two consecutive primer coats, selected for eachgroup as indicated above, were applied on the condi-tioned surfaces and dried with a light stream of air(leaving the resin layer intact), giving a shiny surface.The bonding agent was ligbt eured for 10 seconds.

Composite discs (3.5-mm nominal diameter and1.5-mm approximate thiekness) were light eured ontothe bonded surfaces using a Bencor Multi-T testingsystem (Danville Engineering), and the sbear bondstrengtb (SBS) values (MPa) were determined for allgroups, using a mechanical testing system model 810(MTS). Tbe sample size for eacb group was n = 8.

The fracture surfaces after tbe sbear tests were ex-amined visually with a magnifying glass to make aqualitative assessment of tbe type of failure, ie,whether failure occurred adhesively or cohesively inthe bonded region. To simplify analysis, the retentionof the adbesive layer on botb fracture surfaces of aspecimen was deemed as eobesivc failure within tbebonding layer.

Interface characterization. Morphology at thedentin-adhesive interface was evaluated using dentindiscs with a thickness of approximately 1 to 1.5 mm.The discs were prepared from eight teeth, witb one discper tooth. The occlusal one third of tbe tooth was slicedusing an Isomet slow-speed saw with a diamond bladeoperating in water. The occlusal portion was discarded.The remaining tooth was then cut witb a second sec-tioning parallel to the first cut. Dentin discs were exam-ined and those with enamel remnants or exposure ofthe pulp chamber were eliminated. Wet sanding of thedentin surface witb 600-grit silicon carbide sandpaperfor 1 minute was used to produce standardized smearlayers. Tbe dentin discs were washed in water, plungedbriefly into ethanol, and immediately dried witb mois-ture-free air, leaving the bonding surface sligbtly moist,similar to tbe case of bond-strengtb specimens, Theeight discs were randomly assigned to one of fourgroups with two discs per group and the adhesive sys-tems were applied on the dises as previously described.

A thin layer of bonding agent was applied to thebonded dentin surface, to wbich a 1-mm layer of com-

posite (Sculpt-it, Pentron) was placed and ligbt curedfor 20 seconds, Tbe resin-bonded dentin discs werecross-sectioned perpendicular to the resin-dentin in-terface witb tbe Isomet slow-speed saw and embeddedin a self-curing epoxy resin for SEM examination. Alitbe mounted discs were subjected to a standard metal-lographic polishing through 600-grit emery papers anda final polish with 0.05-)jm alumina suspension. Thepolished cross sections were then demineralized anddeproteinated using 6 mol/L hydrochloric acid (HCI)for 10 seconds for dissolution of the mineral compo-nent of dentin, followed by immersion in 1% sodiumhypocblorite (NaOCI) for 5 minutes for removal of tbeeoUagenous material. Tbe specimens were then sputtercoated with gold and examined under a scanning elec-tron microscope (Hitachi S-2500) using an accelera-tion voltage of 15 kV.

il^aterials

The primer formulations studied in this investigationwere commercial formulations and their modificationsin a factorial layout to incorporate two solvents [ace-tone and etbanol) and two curing modes (light cureand dual cure). The two commercial formulationsstudied were Bond 1 and Bond 1 C&B (Pentron).Both these systems are light-cured formulations (cam-phoroquinone initiator) based on pyromellitate ofglyceryi dimetbacrylate (PMCDM) and 2-bydroxyethyl methacryiate (HRMA) as monomers, but tbeirmain difference was the use of different solvents forthe adhesive: acetone for Bond 1 and etbanol forBond 1 CScB. Two experimental dual-cure formula-tions were prepared from tbese commerciai systems byincorporating benzoyi peroxide (BPO) and other addi-tives, sucb as a tertiary amine activator and an in-hibitor. Two drops of the commercial formulationswere mixed witb one drop of tbe activator to modifytbe light-cure formulations to dual-cure systems capa-ble of polymerization by both light and cbemical cure.

RESULTS

Figure la presents tbe means and standard deviations(SD) of SBS values for the four different groups. Themeans are similar for the same curing mode, with noapparent differences hetween different solvents, butsbow differences between tbe different curing modesused. However, the variances of SBS data were foundnot to be homogenous. Various transforms were exam-ined using the Box-Cox approach, and the reciprocalsquare root transform appeared to be the best mode oftransformation. Tberefore, al! analyses were eon-ducted using tbe transformation 1/VSBS of the data.

550 Volume 34, Number 7, 2003

• Velazquez el al

_ 30

1

s ng

th 1

• D

g 10SiIB

S

"^ 0

a

24,9 (9.3¡H H 21.6(9.6)

^ ^ H ^ ^ H -11,8(2,3) ^'^(^"^

DCA DCE LCA LCE

Fig 1 Comparison o( bond strength means' (a) Bar graph of SBS means of ditferent groups. Standard devia-tions are shown in parentheses, (b) Statistical comparisons ol 1/VSBS using mean diamonds and Tui<ey-Ktamercomparison ciroies.

Fiq 2 Scanning eiectron micrographs iliustrating interfaciai morphoiogy. (a) spiit image at 50x and 500Xoriqinai magnilication The marked square in fhe iower hail identifies Ihe region magnified to 500X in the upperhalf. The composite (CM) and dentin (DE) regions are marked in the iower haif. (b) Magnified image ot the in-terface region liiustrating the hybrid iayer (HL) and resin tags (RT],

Analysis of variance (ANOVA) using transformed datareveaied significant differences of mcajis^ Figure lbshows the stafistical contrast of the 1/VSBS data usingmean diamond and Tukey-comparison cireies, andshows ciear separation of 95% confidence range ofmeans. These comparisons revealed significant differ-ences of means between curing modes (P < .0001) butno significant differences between different soivents

used.Figure 2a illustrates a split secondary electron mi-

crograph (50X [bottom] and 500x [top] original mag-nification) of the interfacial morphology of restored

sections with Bond 1 adhesive. The microstructure il-lustrates a typical section showing the dentin, the ad-hesive layer, and the composite iayer. Note the pres-ence of the hybrid iayer including resin tags into thetubules. Figure 2b is a magnified view of a hybrid layerregion in Fig 2a, Note the width of the hybrid iayer,which is approximately 6 to 8 pm. Figures 3a to 3c aretypicai illustrations of micrographs of the hybrid layerregion for groups 2, 3, and 4, All groups revealed simi-iar morphology of the hybrid layer and resin tags.Thus, the interfacial morphology was very simiiar inall groups, indicafing tiiat demineraiization conditions

Quintessence Internafional 551

• Velazquez et al

Fig 3 SEM illustrations of interface structures in dif-lerent groups' (a) LCA; (b¡ DCE: and (c) DCA.

were far more important in detertnining the morphol-ogy of the hybrid layer and resin tag formation in thecurrent study.

Visual examination of the sheared surfaces afterbond strength tests revealed that failure oceurredacross the dentin-composite interface, through thebonding layer. This would indicate that the range ofbond strength values observed result from differencesin the strength of the hybrid layer, resulting from po-tential differences in polymerization due to inhibitionof conversion of double bonds in the adhesive during

DISCUSSION

Two independent variables were evaluated in thisstudy: the curing mode (light eure and dual cure); andthe solvent [acetone and ethanol) used as the carrierof the tnonomer for infiltration. The statistical analysisof the SBS data used a three-factor linear tnodel givenby the equation

where b„ is a constant independent of the experimen-tal variables C (cure mode used) and S (solvent typeused); b, and b are the coefficients associated withcontributions of main effects C and S; bj is the coeffi-cient associated with the C"S interaction contribution;and e, is the error term. Table 1 lists a summary of theresults of the ANOVA indicating the significance ofeach of the different variables used in the model.Analysis of the sums of squares indicates that 530/0 ofthe variation in the experimental results was from thethree faetors [cure type, solvent type, artd their inter-action) used in the model (with a total of only 3 de-grees of freedom) as compared to 47% variation dueto error term with 28 degrees of freedom. Thus, themodel used was found to be valid at a very high levelof significance [P < .0001). Ninety-four percent of thevariation due to the three factors of the model resultedfrom cure-type difference, and otily 6% from the sol-vent and interaction effects combined. The results sup-port rejection of the null hypothesis for the cure modeand acceptance of null hypotheses for the solvent typeand cure type*solvent interaction effects.

552 Volume 34, Number 7, 2003

• Veiazquez et ai

TABLE 1 Summary of analysis of variance (ANOVA) results

ModeiCure typeSolvent typeCure'soiventTotai model

ErrorTotai (modei + error)

Degrees ottreedoni

1

1

1

3

28

31

Sum otsquares

0.03940.00030.00210.04180 03780.0796

Meansquare

0,03940.00030.00210.01390.0013

F ralio

29.210.211.59

10.32

P

< ,00016343.2266

< .0001

Jim

SB

S

0,35

0.30

0.25

0 20

0.15

0,35

0,30

0,25

0,20

0.15,2

RP

z5

P

•

'•

51

= 0.53< ,0001 .

- • • • • ' - " /

,20

Predicted

= .63

,253

a

25 .30 .35

TSBS"

•

c

,255 ,257

\

Solvent type leverage

0,30m

L^ 0,25

0.20

p<

. . •

\

.21

0,30

m 0.25

0.20

p =

,20

.0001

23 ,25Cure type

22

• • • • - • •

22 ,24

b

.27 ,29eve rage

• . , , - • • • " ^

•

d

,26 28 30Soivent type'cure type leverage

Fig 4 Anaiysis ot transtormed SBS data using a linear model ot cure, solvent and cure'soivent interactioneffects (a) Wiicie model plot, (b) cure lype ieverage etfect, (c) soivent effect: and (d) cure'soivent ettect.

Figure 4a is a wbole model plot of tbe transformedSBS data against tbe predicted transformed data asper tbe equation above, along witb tbe gSit/o confi-dence range indicated by tbe curved lines near the lineof fit, Tbe dasbed horizontal line is tbe grand mean ofall data. The grand mean line is not fully enclosed bythe 95% confidence range, indicating the sigtiificantcumulative effect of model factors on tbe mean of1/VSBS,

Figure 4b sbows a cure mode leverage plot, wbichpools tbe data for different solvent types, and tbe re-sults clearly indicate tbat shear bond strength meansare significantly higher with dual-cure mode whencompared to light-cure mode. According to Reinhardt, ^the amount of unconverted double bonds in a 20-pm

surface layer of cured composite is 28% to Wla (ascompared to only 18% to 32% at 60 ym away from tbesurface). This would suggest tbat oxygen ¡nbibition ef-fects may adversely influence polymerization in tbinlayers, Tberefore, tbere is a strong likelibood tbat tbemonomer adbesive in tbin layers is more sensitive tooxygen inhibition effects especially when curing is ef-fected only through light activation, fti dual-cure sys-tems, chemical cure initiates conversion effects as soonas the activator is mixed witb the monomer, and poly-merization is further supplemented by light activation,Tbese effects may enhance overaii conversion in dual-cure systems, and additional research is therefore nec-essaty to determine the adverse effects of oxygen inhibi-tion and limitations of double bond conversion on SBS.

Quintessence Internationai 553

• Veiazquez et al

Figure 4c is an effect plot of solvent leverage on1/VSBS. Tbe closeness of tbe ettect plot to tbe horizon-tal grand mean line (fully enclosed by the 95% confi-dence range) and the significance level (P = .63) bothconfirm tbat tbere is very little effect of solvent selec-tion on bond strengtb. The optimal primer solvent isone that evaporates easily and removes water withoutthe need for excessive drying. ' Solvents typically usedin dentin adhesives arc acetone and ethanol. Previousreports indicate that acetone-based systems havegreater solvent volatility, which may have significant ef-fect during long-term storage.'* The results in this studywere based on short-term experiments without signifi-cant storage time, and consequently, long-term storagevolatility effects are not expected to have significantbearing on the current results. Tbe role of the soivent isto act primariiy as a carrier for the monotner to helpdiffuse into tbe open space around exposed collagenfibrils on the surface of demineralized dentin. Onec dif-fusion is achieved, the solvent is typically dried witb agentle stream of air so as not to adversely affect thepolymerization of tbe monotner during cure. Tbus,there is no direct role for the solvent in tbe bondstrength under optitnized conditions of bonding, andthe results tend to confirm this in tbe study. In additionto tbe main eñccts discussed above, tbe only interactioneffect due to crossing of solvent and cure type effectswas also found to be nonsignificant (P — .22), as seen inFig 4d. Tbus, tbe results eonfirm tbat the bond strengthdifferences can be primarily attributed to differences ineuring associated witb oxygen inbibition effects, as dis-cussed previously

The effect of polymerization on bonding of poly-mers to solid surfaees has been an important topic ofinterest in adhesion phenomena. For exatnple, molec-ular models have been proposed in the past to de-scribe how polymer molecules could promote adhe-sion between two different polymers or between apolymer and a solid surface. Sucb molecules calledconnector molecules have been suggested as a sourceof enbancement of adbesive bond strength. Leger etaV recently reviewed the concept of connector mole-cules and tbeir possible effect on increase in tbe adbe-sion energy over tbe tbermodynamic work of adbe-sion, even if tbe surfaee-ancbored cbains are notchemically attached to the substrate surfaee. Adhesionenergy increment associated witb the interaction atthe interface was deduced as

wwbere G^-W is the adhesion energy increment due toconnector molecules, G , is the adhesion energj' at thevanishing velocity of fracture, W is tbe normal work of

adhesion in the absence of connector molecules, N istbe index of polymerization of tbe connector chains,and 1 is the density of connector chains at the inter-face. Por a fracture advancing at a finite velocity, theadhesive strength may further be enbanced due to fric-tion between tbe connector molecules and the sub-strate molecules. It is interesting to note that botb co-besive and adhesive interactions may play a criticalrole in the bond strength, and in tbe case of dentinbonding, tbe index of polymerization may representtbe cohesive interactions between tbe adhesive poly-meric cbains in tbe bybrid layer. Thus, analysis of ad-hesion of polymeric materials to solid surfaces tendsto support the role of polymerization in determiningthe bond strengtb observed in the current studies.Adhesive penetration into the subsurfaces of a de-mineralized dentin may also enbance the bondstrength because of increased cobesive, adbesive, andfrictional interactions.

Tbc results furtber indicate tbat tbe bonding ofdemineralized dentin to dental restorations throughtbe adbesive primers used in the study leads to tbe for-mation of a hybrid layer and resin tags to the dentinaitubules. The hybrid layer formed at the interface hasbeen shown by many authors in the past to be an inti-mate mixture of collagen fibrils of demineralizeddentin and tbe adbesive.'-'' The tbickncss of tbe hybridlayer may be influenced by etebing conditions and for-muiation variables. Tbe present study was conductedunder similar etebing conditions but used different for-mulations. Under these conditions, the bybrid layerthickness ranged from 6 to S pm witb no significantdifferences between different groups. Tbus, the soiventand cure formulation differences used in the study hadno detectable effect on the bybrid layer. However,Finger and Fritz ' have sbown tbat bybrid layer thick-ness is not a valid, but an unspecific parameter toevaluate bonding efficiency, and the results tend tosupport such a conciusion in this study.

One possible limitation of most bond strengtb mea-surement metbods is the absence of positive fluid flowpresent under in vivo conditions. Tbis may displacetbe adbesive during its placement and thus limit the invivo bond strengtb to less tban tbe in vitro measure-ments reported in bond strength studies. Tbis limita-tion is of course considerably more significant withhydrophobic monomers used in tbe past than themore reeent hydrophilic monomers used in tbe cur-rent study, Tbis is because the hydrophilic monomersin water-chasing solvents (such as acetone andethanol) tend to be attracted to tbe dentinal fluidrather tban be repelled as in tbe ease of bydrophobicmonomers.

554 Voiume 34, Number 7, 2003

• Veiazquez et al

Clinical implications

Clinically, this study brings into focus the importanceof efficient polymerization of the resin phase of thehybrid layer, Altbougb light cure is preferred from aclinical point of view, the limitations of polymeriza-tion of unfilled resins under light-cure conditions areto be considered by both tbe manufacturer and tbeclinician. In particular, oxygen inbibition conditionsshould be carefully avoided to ensure efficient poly-merization, Tbe dentin adbesives with Httle or no fillercontent may be especially sensitive to oxygen inhibi-tion in tbin layers, Tbe incorporation of smallamounts of filler in tbe adhesive or the use of a flow-able composite layer after adbesive application priorto final placement of the restoration are potential ap-proaches to alleviate this problem, Furtber research isneeded to examine tbis and otber aspects of polymer-ization in tbe hybrid layer.

CONCLUSION

The results of this study indicate tbat optimized curingconditions are critical to cstabüsb efficient bondingthrougb bybridization. However, etbanol or acetonefunction effectively as a solvent carrier for monomerinfiltration.

ACKNOWLEDGMENT

This Sludy was part of a thesis submitted by EMsa Velazquez (o thegraduace school of Biotiiedit-al Sciences al (he University ofMedicitie and Dentistry of New Jetsey. Newark. New Jersey, towardthe requiretDents of a MS degree in oral biology.

REFERENCES

1. Nakabayashi N, Pashiey DH. Hybridization of Dental HardTissues Chicago: Quintessence, 1998.

2. Nakabayashi N, Ashizawa M, Nakatnura M. Identificationof a resin-dentin hybrid layer in vital human dentin createditi vivo: Durable bonding to vital dentin. Quintessence Int1992;23:135-141.

3. Nakabayashi H, Nakatnura M, Yasuda N. Hybrid layer as adentin bonding ttiechanism. ] Esthel Dent 1992;3:133-138.

4. Pashiey DH, Ciucchi B, Satio H, Homer JA. Permeability ofdentin to adhesive agents. Quintessence Int 1993;24:618-631.

21

Pashiey DH, Agee IÍA, Nakajima M, et ai. Solvent induceddimensional changes in EDTA-deminerahzed dentin tnatrix.J Biotiied Mater Res 2000;56:273-281,Tay PR, Carvaiho RM, Yiu CKY, et al. Mechanical disrup-tion of dentin collagen fibrils during resin-dcntiti bonding, JAdhesive Dent 2000;2; 175-192.

Eick ]D, Robinson SI, Byerley TJ, Chapell RP, Spencer P,Chappelow CC, Scanning transmission eieclron tni-eroscopy/energy dispersive spectroscopy analysis of [hedentin adhesive interface using a labeled 2-hydroxyethyl-methacrylate analogue, J Dent Res 1995;74:1246-1252.Van Meerbeek B. Dentin Adhesion: Morphological,Physieo-Chemical and Clinical Aspects [thesis], Leuven,Belgium: Calholic University of Leuven, 1993.Spencer P, Wang Y, Walker MP, Weiliczka DM, SwaffordjM. Interfaciai chemistry of Ihe denlin/adhesive interface. JDent Res 2000,79:1456-1463.

Van Meerbeek B, Inokoshi S, Braem M, Lamhrechts P,Vanherle G. Morphological aspects of interdiffusion zonewith different dentin adhesive systems. J Dent Res1992:71:1530-1540.Van Meerbeek B, Dheri A, Goret-Nicaise M, Braem M,Lambrechts P, Vanherle G. Comparalive SEM and TEM ex-amination of the ultraslructure of the dentin-resin interdif-fusion zone, J Dent Res 1993:72:495-501.Eick JD, Gwinnell AJ, Pashiey DH, Robinson SJ. Currentconcepts on adbesion to dentin. Cril Rev Oral Biol Med1997;8:306-335.Perdigao J. An ultra-morphological study of human dentineexposed to adbesive systems [thesis]. Leuven, Belgium:Catholic University of Leuven, 1995.Kanca |. Improving bond strength through acid etching ofdentin and bonding to wet dentiti surfaces. | Am DentAssoc 1992:123:35-43.Gwinnett AJ. Moist versus dry dentin: Its effect on shearbond strength. Am J Dent 1992;5:127-129.Eliades G, Palaghias G, Vougiouldalds G. Effect of aeidic con-ditioners on dentin morphology, molecular composition andcollagen conformation in situ. Dent Mater 1997;13:24-33,liinney JH, Balooch M, Marshall GW, Marshall SJ, Atomicforce microscope study of ditnensional changes in dentinduring drying. Arch Oral Biol 1993 ;38:1003-1007,Reinhardt KJ. Unconverted double bonds and interface phe-nomena in composite material. Deutsehe ZahnarztlicheZeitschrift 1991:46 204-208.Abate PP, Rodriguez VÎ, Macchi L Evaporation of solventin one-bottle adhesives. J Dent 2000:28:437-440.Leger L, Hervet C, Tardivat C, Koutsos V, Portigliatti M.Molecular Mechanisms of Adhesion at Polymer-solidInterfaces. [Proceedings of the 25th Annual Meeting of theAdhesion Society and Second World Congress on Adhesionand Related Phenomena, 10-14 Feb 2002, Orlando,Blacksburg, VA: Adhesion Society, 2002. 2002:2-5,Finger W, Fritz U. Laboratory evaluation of one-componentenamel/dentin banding agents. Am J Dent 1996;9:206-210.

Quintessetice Intetnational 555