Resin-modi®ed glass-ionomers: effect of dentinprimer application on thedevelopment of bondstrengthMiyazaki M, Iwasaki K, Onose H, Moore BK. Resin-modi®ed glass-ionomers:effect of dentin primer application on the development of bond strength. Eur JOral Sci 1999; 107: 393±399. # Eur J Oral Sci, 1999

The purpose of this study was to investigate the rate of development ofdentin bond strengths of resin-modi®ed glass-ionomer cements (RMGIC) withuse of dentin primers. The prepared dentin surface was treated according toeach manufacturer's instruction or the dentin primer. Cements werecondensed into a vinyl mold and light activated. The shear bond strengthswere measured at a crosshead speed of 1.0 mm/min after 1, 5, 10, 30, 60 min,2, 5, and 24 h storage in water at 37³C. Presence of a signi®cant differencebetween the mean bond strength at 1 min and each of the other test timeswas analyzed. The ®rst time at which there was a signi®cant increase in bondstrength was de®ned as the ``initial increasing time''. As compared with themanufacturer's suggested dentin treatments, the bond strengths increasedsigni®cantly when the dentin primers were used. The initial increasing timeswhen the specimens were made following each manufacturer's instructionswere 10y60 min. When dentin primer was used, the initial increasing timeshortened to 5y10 min. It was concluded that the use of dentin primers forRMGIC restorations resulted in reduction of the initial increasing time.

Masashi Miyazaki1, KeisukeIwasaki1, Hideo Onose1, B. KeithMoore2

1Department of Operative Dentistry, Nihon

University School of Dentistry, Tokyo, Japan,2Department of Restorative Dentistry, Indiana

University School of Dentistry, Indianapolis,

Indiana, USA

Dr. Masahi Miyazaki, Nihon University School

of Dentistry, Department of Operative Dentistry,

1-8-13, Kanda-Surugadai, Chiyoda-ku, Tokyo

101-8310, Japan

Telefax: z81-3-3219-8347

E-mail: miyazaki-m@dent.nihon-u.ac.jp

Key words: resin-modi®ed glass ionomers;

dentin bond strength; dentin primer

Accepted for publication June 1999

Glass-ionomer cements (GIC) have many attrac-tive features such as adhesion to wet toothstructure for good marginal sealing, shades similarto the tooth, the release of ¯uoride ions over aprolonged period of time (1), and good biocompat-ibility (2). Compared with resin composites, thedisadvantages of conventional glass-ionomercements are slow rate of setting, susceptibility tomoisture contamination or dehydration during theearly stages of the setting reaction (3, 4). In orderto reduce this weaknesses, resin-modi®ed glass-ionomer cements (RMGIC) have been developed(5±7). The RMGIC generally have longer workingtimes, so that operators have control over thesetting reaction by initiating the light exposure.The light activated setting reaction results in higher

bond strengths and moisture resistance comparedto conventional GIC (8±11).

In a clinical situation, debonding might occur soonafter the restoration was placed, if it is subjected tostress. These stresses may be due to the contractionstress of the material during setting, procedures suchas contouring and polishing of the material, andnormal oral function including mastication. Ade-quate bond strength to tooth structure is one of thedeterminant factors contributing to the clinicalsuccess of dental restorations, and 24-h data havebeen used widely to measure the bond strengths. Theimportance of rapid development of dentin bondstrengths has been reported, and the materials usedin restorative dentistry must be strong enough towithstand both long-term and short-term forces

Eur J Oral Sci 1999; 107: 393±399

Printed in UK. All rights reserved

Copyright # Eur J Oral Sci 1999

EUROPEAN JOURNAL OFORAL SCIENCES

ISSN 0909-8836

(12). PRICE & HALL (13) studied 10-min and 24-hshear bond strengths and emphasized the impor-tance of early bond strengths when bonding systemsare evaluated. Our previous report (14) on RMGICsuggested the importance of knowledge of thedevelopment of dentin adhesion with time, toallow the materials suf®cient maturation timeprior to functional loading or other external stressapplications.

Close adaptation of a restorative material toprepared tooth substrate and its long-term retentionto the entire surface of the cavity are required inorder to protect the remaining tooth structure.Composite restoratives rely on dentin primers andbonding agents for wetting in order to create goodadaptability to the restored tooth (15). The role of adentin primer is to improve the wettability of thedentin surface by the adhesive and to enhancemonomer penetration into the hydrophilic dentinsubstrate (16). Though the bond strength values ofRMGIC to tooth structure have been improvedcompared to conventional GIC, strengths compar-able to resin matrix composites have not beenachieved. It has been found that dentin bondstrength was enhanced by the application of dentinprimers before the placement of RMGIC liners (17).

The purpose of this study was to evaluate therate of development of dentin bond strengths ofRMGIC from 1 min to 24 h after curing. Thehypothesis tested in this study was that thedevelopment of dentin bond strengths would beaffected by the application of dentin primer.

Material and methods

Dentin bond strength test

The RMGIC employed in this study were Fuji IILC (LC, GC Corp., Tokyo, Japan) and Vitremer(VT; 3M Dental Products, St. Paul, MN, USA) asshown in Table 1. Besides each manufacturer'srecommended conditioner/primer, two commer-cially available dentin primers were used (Table 2).The input voltage for the curing unit (New LightVL-2; GC Corp.) was adjusted by using a variabletransformer in order to control the light intensity(18) to 600 mW/cm2 as measured with a dental

radiometer (Model 100; Demetron/Kerr, Dunbury,CT, USA).

A total of 540 bovine, mandibular incisors,stored frozen (±20³C) up to 2 wk after extraction,were used as a substitute for human teeth. Afterremoving the roots with a low-speed saw (Buehler,Lake Bluff, IL, USA), the pulps were removed, andthe pulp chamber of each tooth was ®lled withcotton to avoid penetration of the embeddingmedia. The labial surfaces of the bovine incisorswere ground on wet 240-grit SiC paper to a ¯atsurface. Each tooth was then mounted in cold-curing acrylic resin (Tray Resin II; Shofu, Kyoto,Japan) to expose the ¯attened area and placedunder tap water to reduce the temperature risefrom the exothermic polymerization of the acrylic.Final ®nish was accomplished by grinding on wet600-grit SiC paper until a suf®cient area of dentinwas exposed. After ultrasonic cleaning in distilledwater for 3 min to remove debris, the surfaces werewashed and dried with oil-free compressed air.

Double-sided adhesive tape with a 4 mm dia-meter hole was ®rmly attached to the ¯attenedsurface to restrict the adhesive area, and the teethwere randomly divided into different treatmentgroups; 1) the dentin surface was treated with theDentin Conditioner for 20 s, rinsed with water andgently air dried; 2) theVitremer Primer was appliedto the dentin surface for 30 s, gently air-thinnedand light irradiated for 20 s; 3) OptiBond Primewas applied to the dentin surface with scrubbingmotion for 30 s, gently air-dried and light irra-diated for 20 s; and 4) one each drop of FluoroBond Primer A and B were mixed, applied tothe dentin surface for 10 s and gently air-dried(Table 2).

Vinyl molds (2 mm height, 4 mm internaldiameter) were used to form and hold therestoratives to the dentin surface. The restorativewas condensed into the mould, and then com-pressed with a 0.5 N load followed by lightexposure. After light exposure of the restorationsaccording to each manufacturer's instruction, the®nished specimens were transferred to 37³Cdistilled water and stored for 1, 5, 10, 30, 60 min,2, 5, and 24 h. Ten samples per test group weretested in a shear mode with a universal testing

Table 1

Materials tested

Filling material Lot. No. Conditioner/Primer Manufacturer

Fuji Ionomer P: 041034 Dentin conditioner GC Corp.Type II LC L: 240931Vitremer P: 474 Vitremer Primer 3M Dental Products

L: 433

394 M. Miyazaki et al.

machine (Instron Type 4204; Instron Corp.,Canton, MA, USA) at a crosshead speed of1.0 mm/min. Shear bond strength values (MPa)were calculated from the peak load at failuredivided by the specimen surface area. All the testswere conducted at a temperature of 23¡1³C andrelative humidity 50¡5%.

The mean and standard deviation for each groupwere tested for homogeneity of variance usingBartlett's test, and then subjected to one-wayANOVA followed by the Dunnet test to test forpresence of a signi®cant difference between themean bond strength at 1 min and each of the othertest times. The earliest test time when there was asigni®cant increase in bond strength was de®ned asthe ``initial increasing time''. Then the data weresubjected to the Duncan multiple range test todetermine if signi®cant differences existed amongthe dentin surface treatments employed. Allstatistical tests were performed at the 95% levelof signi®cance.

After the testing, the specimens were examinedin an optical microscope (SZH-131; Olympus,Tokyo, Japan) at a magni®cation of 610 to identifythe location of the bond failure. The test area onthe tooth was divided into eight segments and thepercentage that was free of material was estimated.The type of failure was determined based on thepercentage of substrate free material as adhesive,adhesive-cohesive and cohesive.

Scanning electron microscopy

For the ultrastructural observation of the RMGIC/dentin interface, bonded specimens stored in 37³Cdistilled water for 24 h were embedded in epoxyresin and then longitudinally sectioned with a

diamond saw. The sectioned surfaces were polishedto a high gloss abrasive discs and diamond pastesdown to 0.1 mm particle size successively. Theywere dehydrated in ascending grades of butanol(50% for 20 min, 75% for 20 min, 95% for 20 min,and 100% for 2 h), and then transferred from the®nal 100% butanol bath to a chamber of a critical-point dryer (Model ID-3; Elionix, Tokyo, Japan)for 30 min. The polished surfaces were thensubjected to argon-ion beam etching (EIS-200ER; Elionix) for 15 s with the ion beam(accelerating voltage 1.0 kV, ion current density0.4 mA/cm2) directed perpendicular to thepolished surface (19). The surfaces were coatedin a vacuum evaporator (Quick Coater Type SC-701; Sanyu Denshi, Tokyo, Japan) with a thin ®lmof Au. Observation was done under the scanningelectron microscope (JSM-5400; JEOL, Tokyo,Japan). The primer/conditioner treated dentinsurfaces also were observed by SEM.

Results

The data for mean shear bond strength at varioustime intervals with various dentin treatments areshown in Table 3. At 24 h, all the materials testedexhibited their highest bond strengths and thevalues obtained were 6.7¡1.2y11.8¡1.5 MPa forLC and 4.3¡1.3y11.4¡1.5 MPa for VT. Com-pared with the manufacturer's suggested dentintreatments, the bond strengths increased signi®-cantly when dentin primers were used.

The dentin bond strengths increased with time,and the increasing tendency was different betweenthe RMGIC and among the dentin treatmentsused. The initial increasing time, when a ®rstsigni®cant increase in bond strength was observed,

Table 2

Main components of conditioner/ primer used and application methods

CodeConditioner/Primer

(Lot. No.) Main component Procedure Manufacturer

DC Dentin Conditioner 10% Polyacrylic acid 20 s apply, GC Corp.(290501) rinse

VP Vitremer Primer HEMA, Ethanol, CQ 30 s apply, 3M Dental Products(420) Polyacrylic copolymer 20 s irradiation

OB OptiBond Prime GPDM, PAMM, HEMA, 30 s scrub, Kerr/Sybron(750685) CQ, Ethanol 20 s irrad.

FB Fluoro Bond 4-AET, 4-AETA, HEMA, 10 s apply Shofu(A: 328, B: 392) Ethanol

HEMA: 2-hydroxyethyl methacrylateCQ: CamphorquinoneGPDM: Glycerophosphate dimethacrylatePAMM: Phtalic acid amino ethyl methacrylate4-AET: 4-acryloxyethyltrimellitic acid4-AETA: 4-acryloxyetyltrimellitate anhydride

Dentin bond of resin-modi®ed GIC 395

was 10 min for LC and 60 min for VT. With use ofdentin primers, the initial increasing timedecreased to 5 min for LC and 5y10 min for VT.

A trend toward differences in failure modeamong the test time periods and the dentin primersemployed was seen. The fracture mode of RMGIC,when used according to each manufacturer'sinstruction, was found to be adhesive at thedentin interface and partially cohesive failure inthe cement regardless of the test time period. Whenthe dentin primers were employed, the failure modetended to be cohesive within the cement, andpartially in the dentin after the 2-h test time period.

Fig. 1A illustrates the smear layer produced bythe #600 SiC paper grinding. The entire surface ofthe dentin and the dentinal tubules were coveredwith smear layer. With the 10% polyacrylic acidtreatment, the smear layer was removed but thedentinal tubules were still occluded by smear plugsat their ori®ces (Fig. 1B). With the use of dentinprimer, the smear layer and smear plugs werepartially removed and the porous intertubulardentin revealed (Fig. 1C).

SEM micrographs of the interface betweenRMGIC and dentin is shown in Fig. 2. Thoughthe cement matrix/dentin interdiffusion zones ofLC (Fig. 2A) and VT (Fig. 2B) were not clearlyseen, a thin layer of cement matrix-dentin inter-diffusion zone (0.5 mm) was observed for LC.When the dentin surfaces were treated with dentinprimers, a layer with low resistance to argon ionetching was identi®ed as the dentin primer-cementmatrix-dentin interdiffusion zone (Fig. 3). Thethickness of this layer, which ranged from1.0y1.5 mm, was different between the dentinprimers used.

Discussion

To evaluate the bond to human teeth in vivo is theultimate goal in bond strength tests. A large

number of teeth are required for conductingbond strength tests, and it is dif®cult to obtainintact extracted human teeth for laboratory studies.Therefore, bovine teeth were used as a substitutefor human teeth, as it has been indicated byprevious studies (20, 21) that there was little or nodifference in laboratory bond strengths to humanand bovine tooth structure. In this study, a shearbond strength test was employed to evaluate thebonding ability of RMGIC because of the testperiod of data acquisition. Much of the researchrelated to dentin bonding has been done in anattempt to assess the integrity and/ or strength ofthe interfacial bond. Experimental approaches forthe measurement of bond strengths of RMGIChave consisted of tensile and shear bond strengthdeterminations. The reliability and validity oftensile and shear bond strength determinations ofdentin-bonded interface have been questioned(22). Although the testing procedures employedare apparently similar, the results presented indifferent studies may differ tremendously. Largevariations in bond strength determinations and thelack of standardized laboratory test procedureshave contributed to ambiguities in data interpreta-tion (23).

Although the exact mechanism of bonding ofGIC to human tooth is not fully understood, itseems that the mechanism involves wetting of thetooth surface by the cement and the subsequentformation of ionic bonds with the conditionedtooth substrate (24). It is generally believed thatthe adhesion of GIC might be the result of thedevelopment of an ion-exchange mechanism,polyacrylate ions replacing phosphate ions in thesurface of hydroxyapatite (25). The fracturesurfaces after the bond strength test with theGIC showed adhesive failures as well as cohesivefailures in the material. The increase in bondstrengths over 24 h seen for all materials testedcould be explained by a maturation of the material.

Table 3

Shear bond strengths (MPa) of the restorative materials measured at various storage times

Storage Time

1 min 5 min 10 min 20 min 30 min 60 min 2 h 5 h 24 h

Fuji II LC/DC 3.4 (1.2) 3.5 (1.1) 4.9 (1.3)* 5.1 (1.2) 5.8 (1.3) 6.2 (1.2) 6.5 (1.3) 6.5 (1.4) 6.7 (1.2)Fuji II LC/OB 4.8 (1.0) 6.5 (0.9)* 8.2 (1.5) 8.4 (1.4) 9.0 (1.2) 9.8 (1.5) 9.9 (1.5) 10.9 (2.0) 11.8 (1.5)Fuji II LC/FB 5.1 (1.3) 7.5 (1.6)* 7.7 (1.9) 7.6 (1.4) 8.0 (1.4) 8.0 (1.5) 9.8 (1.4) 10.0 (1.8) 11.2 (1.6)Vitremer/VP 1.8 (0.3) 2.3 (0.4) 2.3 (0.5) 2.3 (0.7) 2.6 (0.8) 3.2 (1.0)* 3.6 (0.9) 3.9 (1.1) 4.3 (1.3)Vitremer/OB 3.7 (0.6) 4.7 (0.9)* 6.0 (1.1) 6.5 (1.5) 7.3 (1.3) 8.4 (1.1) 9.4 (1.7) 10.1 (2.1) 11.4 (1.5)Vitremer/FB 3.8 (0.7) 4.2 (0.9) 4.8 (0.8)* 5.7 (1.1) 6.3 (1.4) 7.2 (1.3) 8.5 (1.7) 10.3 (1.4) 11.1 (1.7)

Standard deviations are in parentheses, n~10.*: Indicate the test period when a signi®cant increase in bond strength compared with 1 min value (initial increasing time) was foundby Dunnet's test (pv0.05).

396 M. Miyazaki et al.

It has been proposed that the increase in thestrengths of the GIC was caused by the formationof a silica matrix network developed after initialstage of the setting reaction (26). For the resin-modi®ed glass-ionomer cements, the acid-basereaction continues after the cement has beenexposed to the curing light, and the poly-HEMAand polyacrylic metal salts form a homogeneousmatrix which surrounds the glass particles (5). Itwas suggested that the adhesion of the RMGIC to

dentin was primarily chemical in nature. Thesurface demineralization might improve surfacewetting and facilitate penetration HEMA compo-nent, thus contributing to later micromechanicalinterlocking (27). The change in ¯exural propertiesover testing interval might re¯ect the extent of thecontinuing acid-base setting reaction (28).

The HEMA concentration of RMGIC providesgood wetting ability on the hydrophilic dentinsubstrate. For the RMGIC used in this study, thecement matrix including the resin monomerpenetrates the dentin and creates a micromecha-nical reinforcement (29). Such a micromechanicalreinforcement layer was more clearly seen whenthe dentin primers were employed. Along with theimproved mechanical properties of RMGIC,

Fig. 1. Scanning electron micrographs of dentin surfaces, A)untreated (polished with #600 SiC paper); B) treated withDentin Conditioner; and C) treated with Fluoro Bond Primer.Original magni®cation63,500.

Fig. 2. A) Scanning electron micrograph of Fuji II LC (C)/dentin (D) interface after argon ion-beam etching. The dentinsurfaces were treated with the Dentin Conditioner followed bycement application. A resin-rich layer was slightly observedbetween the cement and dentin. B) Scanning electronmicrograph of Vitremer (C)/dentin (D) interface after argonion-beam etching. The dentin surfaces were treated with theVitremer Primer followed by cement application. A resin-richlayer cannot be clearly observed. Original magni®cation63,500.

Dentin bond of resin-modi®ed GIC 397

micromechanical retention through resin compo-nent impregnation as seen with the hybrid layermight have contributed to higher dentin bondstrengths when the dentin primer was employed.

An additional function of the dentin primer iswetting the dentin surface to improve contactbetween the RMGIC and the hydrophilic dentin.The primer penetrates into the dentin andincreases the surface energy for good wetting byRMGIC, resulting in the formation of a continuousinterface. The resin component of RMGIC mayreact with the dentin primer, then cement matrixpenetrates into the super®cial dentin layer creatinga cement matrix-dentin interdiffusion zone. ForVitremer used with Vitremer Primer, the pH valueof the primer may not be low enough to completelydissolve the smear layer and create such a zone.

It has been reported that the use of dentinbonding agents with RMGIC improved the bondstrength to bovine dentin (30). From a clinicalconsideration, use of a bonding agent creates aresin layer so that the diffusion of ¯uoride ionsfrom RMGIC may be somewhat reduced. This willdecrease the ¯uoride ion penetration into dentinand may diminish the cariostatic potential ofRMGIC. However, the use of dentin primersimproved the dentin bond strength of RMGICwithout creating a thick resin ®lm which mightaffect ¯uoride penetration.

The results of the present study suggest that theuse of the dentin primers may play an importantrole in increasing the rate of development of bondstrength as well as enhancing the ultimate dentinbond strength. And also, it is important to pay

attention to the rate of development of dentinadhesion (the initial increasing time) to allow thematerials suf®cient maturation time prior tofunctional loading or other stress application.The use of dentin primers prior to the RMGICplacement is recommended to enhance theirbonding ability.

References

1. SWARTZ ML, PHILLIPS RW, CLARK HE. Long-term F releasefrom glass ionomer cements. J Dent Res 1984; 63: 158±160.

2. WALLS AWG. Glass polyalkenoate (glass-ionomer)cements; a review. J Dent 1986; 14: 231±246.

3. CAUSTON BE. The physico-mechanical consequence ofexposing glass ionomer cements to water during setting.Biomaterials 1981; 21: 112±115.

4. CHO E, KOPEL H, WHITE SW. Moisture susceptibility ofresin-modi®ed glass-ionomer materials. Quintessence Int1995; 26: 351±358.

5. SMITH DC. Comparison and characteristics of glass ionomercements. J Am Dent Assoc 1990; 120: 20±22.

6. MOUNT GJ. Glass ionomer cements and future research.Am J Dent 1994; 7: 286±292.

7. SIDHU SK, WATSON TF. Resin-modi®ed glass ionomermaterials-a status report for the American Journal ofDentistry. Am J Dent 1995; 8: 59±67.

8. RUSZ JE, ANTONUCCI M, EICHMILLER F, ANDERSON MH.Adhesive properties of modi®ed glass-ionomer cements.Dent Mater 1992; 8: 31±36.

9. NICHOLSON JW, ANSTICE M, MCLEAN JW. A preliminaryreport on the effect of storage in water on the properties ofcommercial light-cured glass ionomer cements. Br DentJ 1992; 173: 98±101.

10. MOMOI Y, HIROSAKI K, KOHONO A, MCCABE JF. Flexuralproperties of resin-modi®ed ``hybrid'' glass ionomers incomparison with conventional acid-base glass-ionomers.Dent Mater J 1995; 14: 109±119.

11. MIYAZAKI M, MOORE BK, ONOSE H. Effect of surfacecoatings on ¯exural properties of glass ionomers. EurJ Oral Sci 1996; 104: 600±604.

12. BURROW MF, TAGAMI J, NEGISHI T, NIKAIDO T, HOSODA H.Early tensile bond strengths of several enamel and dentinbonding systems. J Dent Res 1994; 73: 522±528.

13. PRICE RBT, HALL GC. In vitro comparison of 10-minuteversus 24-hour shear bond strengths of six dentin bondingsystems. Quintessence Int 1999; 30: 122±134.

14. MIYAZAKI M, IWASAKI K, SOYAMURA T, ONOSE H, MOORE

BK. Resin-modi®ed glass ionomers: dentin bond strengthversus time. Oper Dent 1998; 23: 144±149.

15. VAN MEERBEEK B, PERDIGAÈ O J, LAMBRECHTS P, VANHERLE G.The clinical performance of adhesives. J Dent 1998; 26: 1±20.

16. ERICKSON RL. Surface interactions of dentin adhesivematerials. Oper Dent 1992; Suppl 5: 81±94.

17. HINOURA K, MIYAZAKI M, ONOSE H. Dentin bond strengthof light-cured glass-ionomer cements. J Dent Res 1991; 70:1542±1544.

18. MIYAZAKI M, HINOURA K, ONOSE H, MOORE BK. In¯uenceof light intensity on shear bond strength to dentin. AmJ Dent 1995; 8: 245±248.

19. INOKOSHI S, HOSODA H, HARNIRATTISAI C, SHIMADA Y.Interfacial structure between dentin and seven dentinbonding systems revealed using argon ion beam etching.Oper Dent 1993; 18: 8±16.

20. NAKAMICHI I, IWAKU M, FUSAYAMA T. Bovine teeth aspossible substitutes in the adhesion test. J Dent Res 1983;62: 1076±1081.

Fig. 3. Scanning electron micrograph of Fuji II LC (C)/dentin(D) interface after argon ion-beam etching. The dentin surfaceswere treated with the Fluoro Bond Primer followed by cementapplication. Close adaptation of the cement to dentin and alayer with low resistance to argon ion etching was clearlyobserved. Original magni®cation63,500.

398 M. Miyazaki et al.

21. FOWLER CS, SWARTZ ML, MOORE BK, RHODES BF. In¯uenceof selected variables on adhesion testing. Dent Mater 1992;8: 265±269.

22. VAN NOORT R, NOROOZI S, HAWARD IC, CARDEW G. Acritique of bond strength measurements. J Dent 1989; 17:61±67.

23. MIYAZAKI M, OSHIDA Y, XIROUCHAKI L. Dentin bondingsystem. Part I: literature review. Bio-Med Mater Eng 1996;6: 15±31.

24. LIN A, MCINTYRE NS, DAVIDSON RD. Studies on theadhesion of glass-ionomer cements to dentin. J Dent Res1992; 71: 1836±1841.

25. WILSON AD, PROSSER HF, POWIS DM. Mechanism ofadhesion of polyeletrolyte cements to hydroxyapatite.J Dent Res 1983; 62: 590±592.

26. WASSON EA, NICHOLSON JW. New aspects of the setting ofglass-ionomer cements. J Dent Res 1993; 72: 481±483.

27. TITLEY KC, SMITH DC, CHERNECKY R. SEM observations ofthe reactions of the components of a light-activated glasspolyalkenoate (ionomer) cement on bovine dentine. J Dent1996; 24: 411±416.

28. MITRA SB, KEDROWSKI BL. Long-term mechanical proper-ties of glass ionomers. Dent Mater 1994; 10: 78±82.

29. MIYAZAKI M, RIKUTA A, IWASAKI K, ANDO S, ONOSE H.In¯uence of environmental conditions on bond strength of aresin-modi®ed glass ionomer. Am J Dent 1997; 10: 287±290.

30. PEREIRA PNR, YAMADA T, INOKOSHI S, BURROW MF, SANO H,TAGAMI J. Adhesion of resin-modi®ed glass ionomercements using resin bonding systems. J Dent 1998; 26:479±485.

Dentin bond of resin-modi®ed GIC 399