Ml Research

Demineralizing effect of dental cementson human dentin

Yasushi Shimada, PhD'/Yasuyo Kondo, DDSVShigehisa lnokoshi, PhD*/JunjiTagami, PhDVJoseph M. Antonucci, PhD"

Objective: This study was undertaken to verify the hypothesis that dentin surfaces are demineralized dur-ing placement ot lour kinds ot chomically setting cements (une phosphate cement, luting glass-ionomer ce-ment, restorative glass-ionomer cement, and zinc polycarboxytate cement). Method and materials: Sixtycemented dentin disks were cbserved under scanning electron microsccpy and with confocai laser scan-ning microscopy after use of an argon-ion etching technique. To determine the surface effects of the ce-ments. 30 dentin surfaces were treated with 1 of 6 freshly mixed cements (5 per group) for 60 seconds. Thedisks were subjected to rinsing with a water spray and ultrasonic washing prior to scanning eiectron micro-scopic observation. Results: Observation of cemented dentin specimens revealed that the dentin was notcompletely demineraiized at the interface formed by the cement and dentin and that the extent and depth ofdemineraiization along the interface tended to be nonuniform. Zinc phosphate cement caused the greatestdemineraiization of dentin, followed by luting glass-ionomer cement. The extent of demineralization withrestorative glass-ionomer cement or zinc polycarboxylate cement was less discernible. Confocai laser scan-ning microscopy generally confirmed scanning electron microscopic observations and revealed that most ofthe specimens showed close adaptation of the cements to the dentin surfaces. Conclusion: Acid-contain-ing cements have self-etching properties that are effective, to various degrees, in removing the smear layerand promoting close adaptation to dentin suriaces. ¡Quintessence int 1999;30:267-273)

Key words: confocai iaser scanning microscopy, demineralization, dentin. glass-ionomer cement.scanning electron microscopy, zinc phosphate cement, zinc poiycarboxylate cement

CLINICAL RELEVANCE: Acidic cements promote inter-facial bonding. Their relatively mild self-etching propertiesshould result in iess irritation and long-term pulpal sensitivity.

GeneraUy, dental cements are made from liquid andpowder components tbat barden tbrougb acid-

base interaction. Tbe pH of the initiai mix is low andrises to a level approaching neutrality during tbe courseof the setting reaction.'- Because of tbe initial bigbacidity of tbe cement, botb irritation of tbe pulp and thedemineralization of tbe tootb surfaces in contact witbthe cement migbt be expected. Abe et aP reported tbatzinc pbospbate cement pastes have a demineralizing ef-fect on bovine dentin that is similar in severity to tbatof an aqueous solution of 65 wt% pbospboric acid.

Since tbe development of glass-ionomer cement/

'Department ot Operative Dentistry. Faculty ot Dentistry. Tokyo f^edicaland Dental University. Bunkyo-ku, Tokyo. Japan

"Dental and Medical Materials Group. Polymers Division, National Instituteol Standards and Technology, Gaithersburg. Maryland.

Reprint requests : Dr Yasushi Shimada. Department ot OperativeDentistry, Faculty of Dentistry, Tokyo Medical and Denial University, 1-5-45Viishima, Bunkyo-ku, Tokyo I t 3 , Japan. Fax; 81-3-5803-0195.

tbis adbesive dental material bas undergone variousmodifications and is considered indispensable forcertain types of dental treatments. Tbe setting ofconventional glass-ionomer cements occurs by a seriesof neutralization reactions involving the polyacid andcalcium and aluminum ions leached from a reactiveglass, resulting in tbe formation of a polysalt matrix.'Tbe biologic effects of glass-ionomer cements bavebeen reported by many investigators.""'- Despite tbegeneral agreement that restorative glass-ionomercements are biocompatible,'"'- luting glass-ionomercements are considered to be the cause of mild pulpalirritation wben used for crown cementation.'^ Tbeacidic nature of the cement liquid is a likely cause ofmost of the unfavorable postoperative complications.'-'-^

Tootb-restoration interfaces bave been observed byvarious techniques, including scanning electron mi-croscopy (SEM),""-" transmission electron micro-scopy,'''* and confocai laser microscopy.̂ -̂̂ ^ Clear ob-servation of tbe often vague structural differences ofvarious cement-dentin interfacial structures bas inber-ent difficulties, such as poor fracture resistance of mostcements and tbeir relatively weak bond strengths. Anargon-ion etching technique bas been reported to beuseful in disclosing tbe morpbologic variations ofdentin-restoration interface under SEM-'̂ "̂ " In addi-

Quintessence International 267

•Shimadaetal •

TABLE 1 Dental CÊments used in the study

Material

Super Cement

Eilte Cement

Fuji 1

Ketac-Cem

Fuji ionomer Type

Hy-Bond Carbo

Type' Mf

ZPC

ZPC

LGIC

LGIC

il BGIC

ZPCC

"ZPC = zinc phosphate cement; LGIC =RGIC = resloraliue glass-ionomer ce mecemem.

nutaoturer

Shofu

GC

GC

ESPE

GC

Shofu

uting glass-iol l ; ZPCC ^ zin

Batch No.

P:079330L:039413

P:130121L:030631

P,250532L,240533

P:0282L:0040

P300771L:t 50601

P:129110L, 12907

nom er cemerii,c pol yea rboxy late

TABLE 2

Material

SuperCement

EliteCement

Fuj i l

Ketac-Cem

FujiIonomerType if

H y-BondCarbo

Depth of demineraiized dentin (pm)

SEM obseri'ationMean * SD(min, max)

2.00 ±1,94 (0,0, 5.0)^

2 ,00±2. t t (0.0, 6.0Y

0,50 ±0.47 (0.0, 1,5)'"

0,75 ±0.89 (0,0,3.0)=»

0,10 ±0,21 (0,0,0.5]"

0.04 + 0,13(0,0,0.4]'"

CLSM observationMean± SD(min, max]

2.00 ±1.94 {0.0, 5.0]"

2,00 ±1,94 (0,0, 5,0]"

0,40 ±0.32 {0.0, 1,0)"

0,70 ± 0,75 {0.0, 2,5)».»

0.10 ± 0 21 (0.0,0,5)"

0.04 ±0,13 (0.0, 0,4)»

SEM = scahning electron microscopy, CLSM = confocal laser scanningmicroscopy, (min, man) = minimum and n-Groups idenfified by llie same superscripferent (P > 3.05),

aximum values, respectively.letter are not significantiy dif-

tion to electron microscopy, confocal laser scanningmicroscopy (CLSM) has been widely used in biologyfor noninvasive and nondestructive imaging in vivo ofmany organ tissues and has found numerous applica-tions in dental research.-'-^'' Specimens examined byCLSM do not require any special preparation and arenot subject to the distortions caused by the dehydra-tion that results from procedures such as SEM.

Tliis study was designed to verify the hypothesisthat dental cements can iiave a demineralizing effecton human dentin and that different types of cementsprovide different degrees of demineralization. Tbe ce-ment-dentin interfaces were observed under SEM bymeans of an argon-ion etcbing technique and underCLSM. In addition, dentin surfaces were observedunder SEM after application of tbe dental cements for60 seconds foliowed by tbcir removal with water.

METHOD AND MATERIALS

Six commercial brands of dental cements were ern-ployed in this study {2 zinc phosphate cements [ZPC] ; 2luting glass-ionomer cements [LGIC]; 1 restorativeglass-ionomer cement [RGIC]; and 1 zinc polycarboxy-late cement fZPCC]) (Table 1). Each material was tnixedat 2 3 ^ ± rC and 50% ± 5% relative bumidity and ban-died according to the instructions of its manufacturer.

Extracted noncarious hurnan molars, wbicb hadbeen stored in a preservative solution of 2 g of tbymolin 1 L of saline (Hikari, Hikari Pharmaceutical) at4''C, were used. Ninety-five dentin disks weresectioned perpendicularly to tbe long axis of tbe teetbwitb a diamond saw microtome (Leitz 1600 saw mi-crotome, Lica Instruments) under a stream of water.

Observation of cement-dentin interface

Sixty of 95 disks were randomiy selected and sub-jected to observation of their eement-dentin interfacialstructures by SEM and CLSM. Tbc disks were ran-domly divided into 6 groups of 10 disks, each grouprepresenting tbe different cements. Specimen prepara-tion involved applying the freshly mixed cement be-tween a pair of disks to trtake a dentin-cement-deniinsandwich witbout breaking the cement bonding area.Eacb group was composed of 5 specimens and pro-vided 10 interfaces to observe {n = 10) beeause eachspecimen bad 2 cement-dentin interfaces.

Twenty-four bours after cementation, the specimenwas sectioned perpendicular to tbe cemented surfacethrougb tbe center of the disk with a diamond bur(No. 201, Sbofu) tbat was attacbed to an air turbinebandpiece, with water spray. Both halves of each cutspecimen were embedded in epoxy resin separatelyand then ground and polished witb wet silicon carbidepapers and diamond pastes of decreasing abrasivenessdown to 0.25 pm.

One of the polisbed surfaces was subject to argoti-ion beam etcbing (EIS-IE, EHonix), for 10 minutes,under tbe foilowing operating conditions: an acceler-ating voltage of 1 kV and an ion current density of 0.2mA/cm^ with the ion beam directed perpendicular totbe poiished surface.'̂ "* The polished surface was thensputter coated with gold and observed under an SEM(JXA840, |EOL). The other half of the polished pairsurface was employed for direct viewing under CLSM(1LM21H/W, Lasertec).

Tbe depth of tbe demineraiized layer of each speci-men was determined by measuring tbe deepest ob-served altered zone within the specimen from the

268 Voiume 30, Number 4, 1999

• Shimada et al •

Fig 1 Scanning eiectfon microscopic image ol SuperCemenl-dentin interface. The depttt ot the demineralizeddantin is about 5 ijm (distance between arrowheads). Mostof the zinc ptiosptiate cement has been fractured awayIrom the overlying dentin specimen.

Fig 2 Scanning elecircn microscopic image of EiiteCement-denrin interface. The 2one ct demineralized dentinis much iess ciear than that observed tor Super Cement(see Fig 1),

Fig 3 Scanning eleclrcn microscopic image ot Ketac-Cem-dentin interface. The depth of the demineraüzeddentin IS abouL 1 |jm (distance between arrowfieads).

Fig 4 Scanning electron microscopic image ot iHy-BondCarbo-dentin interface. The zone of demineraiized dentin ismuch iess ciear than that of Ketac-Cem (see Fig 3¡.

SEM and CLSM photographs,'^" Statistical analysisof the results was performed with a parametric one-way analysis of variance and Duncan's multiple rangetest for differences among the cements.

Observation of the dentin surfaces

The remaining 35 dentin disks were assigned to 7groups of 5 specimens each (n = 5); 1 group served asa control. For the experimental groups, freshly mixedcetnents (see Table 1) were immediately applied to thedentin surface for 60 seconds. After that, the dentindisks were rinsed with water spray and then subjectedto ultrasonic washing for 60 seconds to remove the ce-ment pastes. For the specimens in the control group,isotonic saline was applied for 60 seconds withoutwater spray rinsing or ultrasonic washing.

Both the experimental and control specimens were

dehydrated hy exposure to ascending concentrationsof ethyl alcohol and then crifically point dried (JCPD,JEOL) with liquid CO,, used according to the manu-facturer's instructions. Finally, the surfaces were sput-ter coated with gold and ohserved under SEM,

RESULTS

Cement-dentin interface

The SEM observafion of the cement-dentin interfaceshowed that the dentin was partially demineralizedbut in a nonuniform pattern along tbe cement line.That is, the depth of demineralization was not uniformand ranged fiom 0,0 to 6-0 ym for ZPC, 0,0 to 3,0 umfor LGIC, 0,0 to 0,5 pm for RGIC, and 0,0 to 0,4 pmfor ZPCC (Table 2 and Figs 1 to 4), There was a signif-

Quintessence international 269

• Shimada et al •

Fig 5 Cuniùcal laser scanning microscopic image of theother naif of the Super Cement-dentin intertace shown inFig 1. The depth of the demineralized dentin is abcut 5 |jm(distance between arrowheads) Adaptation beiween thecement and dentin is close

Fig 6 Confccal laser scanning microscopic image of FujiIonomer Type ll-dentin interface. There is no ciear evi-dence of a zone ot demineraiized dentin. Adaptation of ttiecement to the dentin surface is exceiient (arrowhead).

icant difference among the rnean depth values of thegroups (one-way ANOVA; P < 0.01), and the meanvalues for the ZPCs were statistically different fromthose for the RGIC and ZPCC (Duncan's multiplerange test; P<0.05).

The ZPCs had a comparatively severe demineraliz-ing effect on the dentin; the second highest valueswere obtained with the LGlCs, but no statisticallysignificant difference existed between the LGICs andthe other eements (Duncan's multiple range test; P >0.05). The depth and extent of demineralized dentinat the interface were unelear in the RGIC and ZPCCspecimens, except for 2 of 5 RGIC specimens and 1of 5 ZPCC specimens, which showed the same de-gree of demineralization. Some SEM specimensshowed separation at the ceinent-dentin interface orfracture within the cement. In particular, more thanhalf of the ZPC specimens observed by SEM werefractured (see Fig 1).

The CLSM observation of demineralized dentinunder the eement revealed depths of demineralizationsimilar to those ohserved with SEM (see Table 2).There were also statistically significant differencesatnong the mean values of demineralization depthamong the groups (one-way ANOVA; P = 0.001); theZPCs were different from the RGIC and the ZPCC(Duncan's multiple range test; P < 0.05). Close adap-tation of the cement-dentin interface and minimalspecimen damage, even for ZPC specimens, were ob-served under CLSM (Figs 5 and 6).

Dentin surfaces

The use of water spray and ultrasonic washing com-pletely removed the cement pastes from the dentin

and revealed an altered surface under SEM. Figure 7shows the dentin surface after applieafion of ZPC for60 seconds followed by the aqueous washing proce-dure. Both the smear layer and the top of the smearplugs were removed, clearly revealing the fibrillar tex-tures in the decalcified dentin tnatrix. Likewise, whenthe surface was treated with the LGICs, the smearlayer and the top of the plugs were removed, and thefibriilar texture of demineralized dentin was evident(Fig 8). When the RGIC and ZPCC were used, smearlayers were removed, but most of the smear plugs re-mained within the tubule orifices (Fig 9). The fibrillartexture of the partially detTiineralized surfaee dentinwas only oecasionally observed.

DISCUSSION

The use of an argon-ion etchitig technique prior toSEM observation has been reported to be useful indisclosing the tnorphologic variation of the resin-dentin hybrid layer in several dentin bonding sys-tems.'̂ " '̂' Application of this technique in this study re-vealed the presence of demineralized dentin at thecement-dentin interface. When CLSM was used, thepresenee of demineralized dentin under the cemerttand the close adaptafion of the cement to the dentinsurface also could he observed. However, when theSEM technique was used for interfacial analysis, morethan half of the specimens of ZPC-dentin separated atthe interfaee or fractured within the cement.

It was expected that these acidic cements wouldhave self-etehing properties, and as a consequence,adapt to dentin by micromechanical attachment. Inthe case of the glass-ionomer and zinc polycarhoxylate

270 Voiume 30, Number 4. 1999

• Stiimada et al •

Fig 7 Scanning electron microsoopio image of Ihe dentinsurtace after Elite Cement was applied tor 60 seconds an6then rinsed away. Tiie smear iayer and ttie top of the smearplugs are removed, and many coiiagen fibrils are visible.

Fig 8 Scanning electron microscopic image of tlie dentinsurtace after application of Ketac-Cem. The smear layerand tiie top at the smear piugs are removed, reveaiing ttiedentin surface to have a librlilar texture.

cements, bonding meehanistns also can involve spé-cifie chemical reactions with the dentin surface.^-'TheSEM images of the dentin surface after the cementpastes were applied for 60 seconds and then rinsedaway with water also showed that tbese dental ce-ments are capable of etching dentin surfaces.

Zinc phosphate cements removed the smear layerand had relatively severe déminerai i zing effects on theunderlying dentin surfaces. In contrast to poiycarboxy-late cements, it is widely recognized that ZPCs do notbond to htjman teeth sufficiently, although they pos-sess self-etching properties. The weak adhesive bondsof the ZPC might be due in part to the poor fractureresistance of the cement itself and in part to the chem-ical nature of this cement, which lacks specific chemi-cal reactivity to promote adhesion with the dentin sur-faces.=-̂ In the SEM part of this study, more than halfof the ZPC specimens were damaged by distortionscaused hy dehydration resuhing from the specimenpreparation process. In contrast, CLSM specimensshowed less damage because they were never dried.

The more severe demineralizing effect of LGICcompared witb the other polyacid-based cements.RGIC and ZPCC, may be due in part to the longer set-iing time of LGIC,'- the different content of powderand liquid components,-^ and different powder-liquidratios.'-̂ Tartaric acid, an important additive that con-trols the setting of LGIC systems, inhibits the initialacid-hase reaction and allows tbe cement to retain flu-idity for an extended period.-*' In addition, tartaric acidis a stronger acid than the polyacid that initially reactswith the glass filler.^' It is conceivable that the moreacidic tartaric acid reacts not only with the ion-ieach-able glass particles but also with the mineral phase ofthe dentin surface to cause demineralization. The basic

Fig 9 Scanning eiectron microscopio image of tiie ûentmsurfaoe atter application ot Fu|i Ionomer Type ii. Some olttie smear iayer and the top of the smear piugs are re-moved, aithojgh the tubuies remain ocoiuded witii residualsmear piug material. Oniy a lew coiiagen fibrils are visibie.

zinc oxide powder of ZPCC reacts more rapidly andcompletely with fhe polyacid than does the glass fillerof the glass-ionomer cement, thereby causing the pH ofthe ZPCC to reach neutrality faster.''̂

Similarly, Fuji Ionomer Type II is intended for useas a restorative material and is used in a thicker mix-ture [higher powder-liquid ratio) tban is the case forthe more fluid luting material. Hence, both ZPCCsand RGICs would he expected to have mild self-etch-ing effects on dentin surfaces. Nitta^* and Shono'^ re-ported that treatment of the smear layer with anacidic conditioner promotes bonding of RGIC tohuman enamel and dentin. These studies indicatedthat the self-etching effect of the RGIC is weak and

Ouintessence international 271

• Shimada et ai

that an acidic conditioner is necessary to acbieve ade-quate bonding strength. Tbese studies support thefindings of the present study, tbat is, tbat FujiIonotner Type 11 exhibits a relatively mild etcbing ef-fect on tbe dentin surface.

In tbis study, only chemically setting cements wereinvestigated. In tbe case of resin-modified glass-ionomer cetnent, tbe self-etching effect of cementpaste also tnay be tnoderated because of tbe effect ofthe resin component and because, after photocuring,gelation is almost itnmediatc. Researchers have re-ported tbat acidic conditioning is necessary beforeplacement of resin-modified glass-ionomer cement intbe cavity to obtain maximum bond strengtb.'-'^''

During the setting period, ion excliange occurs be-tween tbe cement and tbe mineral pbase of the dentinsurfaces. In contrast to resin-dentin bonding, wberegenerally tbe resin-dcntin hybrid layer tends to formuniformly in the etcbed dentin zone,'^"'"'^" tbe de-mineralized dentin zone caused by dental cementpastes is less uniform in extent and depth. One of tbereasons for this irregularity may be tbe beterogenousnature of the smear layer existing on the dentin sur-face, eg, variations in thickness, distribution, and de-gree of attacbment.''"'' If parts of the smear layer aretbick, tbe acidic cement may not be able to remove allof the mineral component of tbe smear layer becauseit is being buffered by the basic powder as it is appliedto tbe dentin surface.

The viscosity of tbe materials is also an importantfactor, because the viscosity of the cement paste canlimit its penetration of the smear layer and dentin sur-face. Tbe increasing viscosity of tbe cement paste, bylimiting diffusion of sufficient protons from tbe acid,makes removal of tbe smear layer more difficult tbanis tbe case witb tbe less viscous conditioners usedwitb resin cotnposite systems. Anotber reason tbatmay contribute to tbe poorly defined cetiient-dentininterface is the formation of an intermediary layer,consisting of precipitated components from botb ce-ment and dentin.-^"'-' For all of the cements, deminer-alized dentin under tbe cement migbt be remineral-ized by tbe migration and precipitation of mineralions eluted from tbe cement (unpublisbed data), orthe cement paste itself might impregnate tbe demincr-alized dentin. It has been shown that glass-ionomercements can impregnate tbe dentin after treatmentwitb an acid conditioner.'̂ -^^ Tbe interplay of fhesefactors may obscure the morpbologic state of dem-ineralized dentin.

Results of this study indicated tbat ZPC and LGICdernineralized dentin because of fbeir self-etcbingproperties. These findings are in accord witb short-term histopathologic investigations and clinical re-sponses to these cements.''^ However, their etcbing ef-

fect was not uniform, and tbe demineralization of sur-face dentin was incomplete. Because of tbis partialand moderate demineralization of dentin, tbe etchingeffect of tbese cements sbould be minimally irritatingand not likely to be associated witb an increase inlong-term pulpal sensitivity.''^'"

Tbe experiments described in this article were per-formed on extracted teetb, in wbicb no dynamic physi-ologic effect can take place. Outward tissue fluid flowresulting from physiologic intrapulpal pressure wetstbe dentin surface and can upset the setting processand reduce tbe etcbing effects of the dental cemetitsbecause of tbe mineral and protein content of tbe pul-pal fluid.'''̂ '' Tbese bypotbeses must be substantiatedby in vivo studies.

CONCLUSION

f. The results of this study indicated tbat acidic ce-ments, wben placed in contact witb dentin, haveself-etcbing effects tbat aid in promoting interfacialbonding and adaptation to cavity walls.

2. Observation of cement-dentin interfacial structuresby scanning electron microscopy and confocai laserscanning microscopy demonstrated that the de-mineralized dentin tbat formed after treatment ofdentin surfaces witb various acid-base cements wasincomplete and nonuniform. The deptb of the dem-ineralized dentin ranged from 0.0 to 6.0 ]jm for ziticphosphate cement, 0.0 to 3.0 pm for luting glass-ionomer cement, 0.0 to 0.5 pm for restorative glass-ionomcr cement, and 0.0 to 0.4 pm for zinc poly-carboxylate cement.

3. Scanning electron photomicrographs of the dentinsurfaces also revealed the self-etcbing effects of thedental cements.

4. Confocai laser scanning microscopic observationrevealed close adaptation of tbe cement-dentin in-terface for all cements. In SEM observation, inter-facial adaptation was not always well delineated;for example, tnore than balf of the SEM specimensof the zinc pbosphate cements showed interfaciaiseparation or fracture witbin tbe cetnent.

5. Clinically, tbe relatively mild self-etcbing propertiesof acid-base cements sbould result in less irritationand long-term pulpal sensitivity.

ACKNOWLEDGMENTS

We gratefully thank Ms Kohori, Lasertet Co of Tokyo, for tier úHi^technical assistance and usage of a CLSM ( 1LM2 i HAV).

This investigation was supported by Tokyo Medical and DentalUniversity. Tokyo. Japan, and Nationat Institute of Standards andTechnology, Gaithersburg, Maryland.

272 Voiume 30, Number 4,

• Shimada et al

DISCLAIMER

Certain commercial materials and equipment are identified in thispaper to specify the experimental procedure. In no instance does suchidentification imply recommendation or endorsement by ihe N;itionulInstitutes of Health or ihe National Inslimie of Standards andTechnology or that the muleriol and equipment is necessarily the bestavailable for the purpose. Ttie opinions and assertions containedherein are the pri\ ate \ iews of the aiilhors and ore not to be construedas official or as reflecting the views of iin\ department or agency ofthe United States government.

REFERENCES

1. Smith DC, Ruse ND. Acidity of glass ionomer cements dur-ing setting and its relation to pulp sensitivity, J Am DentAssoc 1986:112:654-657.

2. Wasson EA, Nicholson JW. Cbange in pH during setting ofpolyelectrol^ie dental cements. J Dent 1993:21,122-126.

3. Abe Y, Mizunuma T. Nakamura M, Nakabayashi N.Affection of ZPC on dentin, J Jpn Soc Dent Mater Devices1984:3:79-84,

4. Wilson AD, Kent BE, A new treatment cement fcr dentistry,Br DentJ 1972:132:133-135,

5. Wilson AD, McLean JW, Glass-ionomer Cement. Chicago:Quintessence, 1988:43-55,

6. Tobias RS. Browne RM, Plant CG, Ingran DV. Pulpal responseto a glass ioncmer cement. BrDent J 1978:144:345-350.

7. Stanley HR, Pulpal responses to ionomer c e m e n t s -Biologicai cbaracteristics. | Am Dent Assoc 1990:120:25-29.

8. Mjór IA, Nordabl I, Tronstad L, Glass ionomer cement anddental pulp. Endod Dent Traumatol 1991;7:59-64.

9. Pameijer CH. Stanley HR, Ecker G. Biocompatibility of aglass ionomer luting agent. Part II. Crown cementation. AmJDent 1991:4:134-141.

10. Hosoda H, Inokoshi S. Shimada Y, Hamirattisai C, Otsuki M.Pulpal response to a new light-cured composite placed inetched glass-ionomer lined cavities. Oper Dent 1991:16:122-129.

11. Shimada Y. Ptilpal response to various dentin conditioningagents. Jpn J Conserv Dent 1992:35:69-100.

12. Sbono T. Pulpal responses to light-cured restorative glasspolyaßtenoate cements, and ultrastructure of cement-dentininterface. Jpn J Conserv Dent 1995:38:514-548,

U. Inokoshi S, Hosoda H, Hamirattisai C, Shimada Y, TatsumiT. A study on the resin impregnated layer of dentin. Part 1.A comparative study on tbe decalcified sections and tbe ap-plication of argon ion beam etching to disclose tbe resin im-pregnated layer of dentin. Jpn J Conserv Dent 1990;33:427-442.

M. Itiokoshi S, Hosoda H. Hamirattisai C, Sbimada Y, Inter-facial structure between dentin and seven dentin bondingsystems revealed using argon ion beam etching. Oper Dent1993:18:8-16.

15. Sugizaki J. Tbe effect of tbe various primers on tbe dentinadhesion of resin composites, SEM and TEM observationsof tbe resin impregnated layer and adhesion promoting ef-fect of tbe primers. Jpn J Conserv Dent 1991:34:228-265.

16. Hamiraftisai C, Inokosbi S, Shimada Y, Hosoda H, ¡nterfacialmorphology of an adhesive composite resin and etched cariesaffected dendn. Oper Dent 1992; 17:222-228.

17. Hamirattisai C, Inokosbi S, Sbimada Y, Hosoda H, Adhesiveinterface between resin and etched dentin of cervical ero-sion/abrasion lesions. Oper Dent 1993; 18:138-143.

18. Van Meerbeek B, Inokoshi S, Braem M, Lambrecbts P,Vanberle G. Morphological aspects of tbe resin-dentin inter-diffusion zone witb different dentin adbesive systems J DentRes 1992:71:1530-1540.

19. Sbimada Y, Hamirattisai C, Inokoshi S, ButTOw MF, TakatsuT. In vivo adhesive interface between resin and dentin, OperDent 1995;20:204-210.

20. Scbumacber GE, Antonucci JM, Bennett PS, Code JE. N-pbenyliminodiacetic acid as an etcbant/primer for dentinbonding. J Dent Res 1997;76:602-609.

21. Watson TF. A confocal optical microscopy study of the mor-phology of the tootb/restoration interface using Scotchbond 2dentin adbesive. J Dent Res 1986:68:1124-1131

22. Lin A. Mcintyer NS. Davidson RD. Studies on tbe adhesion ofglass-ionomer cements to dentin. | Dent Res 1992;71:1836-1841.

23. Gonzalez-Cabezas C, Li Y, Noblitt TW, Gregory RL, KafrawyAH, Stookey GK. Detection of mutans streptococci in sec-ondary caries lesions using immunofluorescent tecbniques andconfocal laser scanning microscopy. Caries Res1995:29:198-203.

24. Pontana M, Li Y, Dunipace AJ, Noblift TW, Fischer G, KaUBP, Stookey GK. Measurenient of enamel demineralizationusing microradiograpby and confocal microscopy. Caries Res¡996:30:317-325.

25. Antonucci ]M, Stansbury JW, Molecularly designed dentalpolymers. In: Arshady R [ed] Desk Reference of FunctionalPolymers: Syntbeses and Apphcations, Wasbington, DC:American Cbemical Society, 1997:719-738.

26. Crisp S, Wilson AD. Reactions in glass ionomer cements. V.Effect of incorporating tartaric acid in tbe cement liquid. JDent Res 1976:55:1023-1031.

27. Prosser HJ, |erome SM, Wdson AD. The effect of additives onsetting properties of a glass-ionomer cement. J Dent Res1982:61:1195-1198.

28. Nitta Y. Effect of surface treafment on microleakage and ten-sile bond strengtb of glass polyalkenoate cement. Jpn JConserv Dent 1992:35 1346-1373.

29 Prati C, Montanari G, Eiagini G, Fava K Pasbiey DH, Effectof dentin surface treatments on tbe sbear bond strength ofVitrabond. Dent Mater 1992:8:21-26.

30. Nakabayashi N. Biocompatibiiity and promotion of adbesionto tooth substrates. Crit Rev Biocompat 1984:1:25-32

31. Eick JD, Wilco RA, Anderson CH, Sorensen SE, Scanningelectron microscopy of cut tooth surfaces and identification ofdebris by use of the electron microprobe, J Dent Res1970;49:1359-1368.

32. Pasbley DH. .Micbelich V, Kebl T. Dentin permeability:Effects of smear layer, J Prosthet Dent 1981:4653]-537

33. Pasbley DH, Kepler EE, Williams EC, Okabe A, The effectsof acid etching of in vivo permeability of dentine in tbe dog.Arch Oral Biol 1983:28:555-559.

34. Geiger SB, Weiner S, Fluoride carbonatoapatite in the inter-mediate layer between glass ionomer and dentin. DentMater 1993;9:33-36.

35. Meryon SD, Tobias RS, Jakeman KJ. Smear removal agents:A quantitative study in vivo and in vitro. J Prostbet Dent1987;57:174-179.

Quintessence International 273