Early marginal leakage and shear bond strength of adhesive restorative systems C. Prali C. Nucci C.L. Davidson~ G. Montanari Io Department of Operative Dentistry School of Dentistry University of Bologna Via San Vitale 59 40125 Bologna, Italy 1 Department of Clinical Materials Science ACTA, University of Amsterdam Louwesweg 1 1066 EA Amsterdam The Netherlands Received September 11, 1989 Accepted March 22, 1990 Dent Mater 6:195-200, July, 1990
Abstract-The aims of this study were to assess in vitro the early marginal seal effectiveness in Class V restorations at the cemento-enamel junction, and the early shear bond strength of several composite and dentin bonding system combinations. The seal effectiveness was studied by evaluating the dye penetrations: (a) along the superficial margin of restorations (width of surface microleakage); (b) along the dentin cavity walls (length of microleakage); and (c) toward the pulp (depth of microleakage). Good correlations were demonstrated among the seal evaluations, but no relationships were apparent between these and shear bond strength. Our study suggests that the marginal seal is influenced by composite properties more than by bonding agents. The shear bond analysis indicates only the adhesive potential of the materials without considering the composite shrinkage effects on marginal gap after the photocuring setting reactions.
arly m a r g i n a l gap formation around composite restorations is the r e s u l t of resin s h r i n k a g e during polymerization and adhesion of the resin to the wall of the cavity (Bausch et at., 1982; Davidson et al., 1984). The gap between restoration and tooth structure is responsible for bacteria and fluid penetration, marginal discoloration, and, consequently, clinical failure (Triadan, 1987). Marginal microleakage along the dentinal wall and shear bond strength tests have been frequently utilized to study the performance of dentin bonding systems (DBS) (Crim and Shay, 1987; Hammesfahr et al., 1987; Gordon et al., 1986; Phair and Fuller, 1985). However, the relationship between these two tests is not clear (Finger, 1988; Retief et al., 1988). It has been suggested that the bond strength value is not an adequate indication of the effectiveness of DBS, although it is used extensively to evaluate the bonding performances of DBS (Komatsu and Finger, 1986). The aim of this study was to evaluate the relationship between the initial adhesion to dentin and early marginal seal ability of several composites used in combination with different new DBS and glass-ionomer cement liners (GIC1). Seal ability was assessed by rheasurement of the superficial marginal circumferential leakage (Zidan et al., 1987), the length of dye penetration around the restoration walls, and the depth of dye penetration toward the pulp (Hammesfahr et al., 1987; Prati et al., 1989).
E
MATERIALS and METHODS
Materials used in this study are listed in Table 1. The combinations of DBS and composites and the sequences of application are shown in Tables 2 and 3, respectively. We used extracted human third molars of healthy young patients
(age, from 18-30 years) stored in saline solution at room temperature for no more than 72 h. The Visilux II (3M Co., St. Paul, MN) was used as a light source for polymerization initiation. M i c r o / e a k a g e Procedure. - Non-retentive Class V restorations were prepared at the cemento-enamel junction (CEJ) of buccal and lingual surfaces of each tooth. A new No. 255 Intensiv (Swiss) high-speed diamond bur with water cooling was used for the cavity preparation. The dimensions of the cavities were 3 ± 0.2 mm (mesio-distal), 3 _+ 0.2 mm (occluso-gingival), and 2.2 ± 0.2 mm (depth). All enamel margins were etched with a 37% w/w phosphoric acid gel (3M Co., St. Paul, MN) for 30 s, washed with water for 30 s, and then airdried for 30 s. Clearfil N e w Bond. Clearfil New Bond was applied with a small brush on dentin and enamel surfaces and then covered with composite. G C L i n i n g Cement. Powder and liquid were mixed according to manufacturer's directions and applied on all cavity dentin surfaces by use of the method of McLean et al. (1985). After five rain, the glass-ionomer surface and enamel were covered with a Scotchbond 2 layer that was photo-cured for 30 s. Gluma Bond. The dentin surface was treated with EDTA (pH 7.4) (Gluma cleanser, Bayer, Leverkusen, West Germany) for 60 s with a small brush, followed by water spray for 30 s and a blast of compressed air. A layer of Gluma primer was applied, dried with a gentle air spray to obtain a reflective dried dentin surface, and then coated with a layer of Gluma Resin L that was photocured for 60 s. Finally, the composite layers were applied as described. Scotchbond 2. Scotchprep primer was applied to the dentin and brushed continuously onto the surface for
Dental Materials/July 1990 195
TABLE 1 DENTINAL BONDINGSYSTEMS(DBS) AND GLASS-IONOMERCEMENTLINERS* (GlCl) USED IN THE STUDY
Material ClearfiI New Bond G C lining cement* Gluma Bond Scotchbond 2 Scotchbond DC Vitrabond lining cement* Tripton
Manufacturer Kuraray, Osaka, Japan G C International Corp., Tokyo, Japan Bayer AG, Leverkusen, W. Germany 3M, St. Paul, MN, USA 3M 3M ICI, Macclesfield, England
TABLE 2 COMPOSITESUSED IN THE STUDY
Material C[earfil Lustre Pekalux Lumifor Experimental D 653 Silux Valux P 30 P 50 Opalux Occlusin
3M Co., St. Paul, MN, USA
ICl, Macclesfield, England
Evaluation of length of dye penetration along dentinal wall (Hammesfahr et al., 1987; Gordon et al., 1986). After superficial evaluation,
TABLE 3 SEQUENCESOF APPLICATIONOF DENTIN PRE-TREATMENT,PRIMER, BONDING RESINS, AND COMPOSITE
Primer Gluma Primer
Bonding Resin Clearfil New Bond Resin L
-
Scotchprep
Scotchbond 2
-
-
Scotchbond DC
-
Tripton Dentin Primer G C lining cement Vitrabond
Tripton Universal Bond Scotchbond 2
Polyacrylic Acid -
60 s, then gently dried with compressed air for 10 s to obtain a reflective dried dentin surface. A layer of Scotehbond 2 was then applied to the dentin surface with a brush and photo-cured for 30 s. Scotchbond DC. Scotchbond DC was applied (with a small brush) onto dentin and photo-cured for 20 s. T~pton. Tripton primer was applied with a small brush to the dentin for 60 s and then gently dried for 20 s with compressed air. Tripton Resin was then applied, gently airdried, and photo-cured for 30 s.
Scotchbond 2
Composite LUSTRE PEKALUX LUMIFOR D 653 SILUX VALUX P50 SILUX P30 OCCLUSIN OPALUX SlLUX SILUX VALUX
Vitrabond. Vitrabond was prepared according to the manufacturer's directions, applied until the entire dentin surface of the cavity was coyered, and then photo-cured for 40 s (McLean eta/., 1985; Prati et a!., 1989). A layer of Scotchbond 2 was then applied and photo-cured for 30 s. The cavities were filled with the composites by use of the three-increment technique (Zidan et al., 1987). The gingival third was filled first, then the occlusal third, and finally the middle third. In the sandwich restorations, the GIC layer was
196 PRATIet al./MARGINALLEAKAGEAND DENTINADHESION
Evaluation of marginal circumferential width of surface dye penetration (Zidan e t a [ , 1987). Buccal and lingual surfaces of teeth were gently cleaned and inspected under an optical stereomicroscope (x 60) along the interface between composite and dentin or enamel. Dye penetration was used as an indication of defective sealing. Marginal microleakage was expressed as the dyed percentage of the total surface width of dentin (c) or enamel (e). Separate evaluations were made for dentin and enamel surfaces, as illustrated in Fig. 1.
Manufacturer Kuraray, Osaka, Japan Bayer AG, Leverkusen, W. Germany
Dentin Pre-treatment EDTA
considered the first gingival increment (Prati et al., 1989). After five min, the restorations were finished with abrasive discs (Sof-Lex, 3M Co., St. Paul, MN) and immediately immersed in 2% erythrosin B solution for one h, then washed in tap water and dried. Eight specimens were prepared for each combination of materials.
the teeth were longitudinally serialsectioned with a diamond disc. Dye penetration along the dentinal wall (d) was considered as the length of microleakage along the interface between dentin and composite. Two evaluations were made (mesial and distal sections) and pooled. The length of microleakage was expressed as the percentage of dye penetration with respect to the total length of the cervical dentin cavity wall. Infiltration up to and not beyond the apex of the cavity was assessed as 100%, while dye penetration affecting both walls, i.e, gingival and incisal, was calculated as 200%, as schematicized in Fig. 1.
Evaluation of depth of dye penetration toward the pulp (Hammesfahr et al., 1987). Dye penetration toward the pulp (p) was evaluated as above and expressed as the percentage of dye penetration with respect to the distance between dentinal wall and pulp.
Shear Bond Strength Procedure. Occlusal and buccal enamel were removed with a high-speed diamond bur with water cooling. The teeth were then mounted with self-curing acrylic resin in box molds. Buccal dentin was pol-
ished by use of a standard technique with a 320-grit abrasive paper (SofLex Pop-On, 3M Co., St. Paul, MN). Only dentin just below the dentinenamel junction was utilized. Each DBS was applied one or two rain after dentin preparation, as previously described. After application of material on the dentin, a cylindrical Teflon tube (4 mm internal diameter and 5 mm length) incrementally filled with composite was applied and photo-cured for 120 s. Each increment was photo-cured for 40 s. After five min, the samples were tested in a universal testing machine. The cylindrical tubes were clamped to the machine and loaded with a stainless steel wire loop to the vertical axis of the cylinder (Prati et al., 1989). The cross-head load speed was 0.5 cm/ rain. Six specimens were made for each combination of materials. The Teflon tubes were not removed during the bond strength test. No bonding occurred between the tube and the dentin-material interface. With regard to GICI, a similar technique was utilized. Cylindrical Teflon tubes (4 mm internal diameter and 3 mm length) were completely filled with the GIC1 and then applied to dentin surfaces. After five rain (Vitrabond) or 60 rain (G C lining cement), the samples were tested in a universal testing machine as previously described. In fact, complete setting of G C was obtained only after about 60 rain.
Statistical Analysis. - T h e data were analyzed by ANOVA with differences determined by Tukey's contrasts at the 95% level. Pearson's R correlation test was used to determine intra-evaluator or inter-evaluator microleakage agreement and the relationships between the microl e a k a g e values and s h e a r bond strength. No significant difference between intra-evaluator and interevaluator microleakage scores was observed. RESULTS Table 4 reports the marginal circumferential width of dye penetration, the length and the depth of microleakage, and the shear bond strength values of each combination of materials. When ANOVA and Tukey's con-
trast were used, there were statistically significant differences among the material combinations. Figs. 2, 3, and 4 show the mean microleakage value ranked by Tukey's post hoc contrast test at p < 0.05. As shown in Fig. 2, the anterior composites (Silux, Valux, Opalux, Pekalux, and Lustre) presented a lower width of microleakage with respect to the posterior composites. These results are also confirmed for the length and depth microleakage (Figs. 3 and 4). As shown in the Figs., SB 2/Silux combination presented, on the whole, the lowest microleakage values for DBS restorations. G C lining cement (FU/SX) showed the lowest microleakage values for GIC1 restorations. Shear bond strength of materials ranked by Tukey's post hoc contrast t e s t is shown in Fig. 5. W h e n Scotchbond 2 was utilized with different composites, it gave a similar bond strength (Fig. 5). However, the SB 2/P 50 combination showed significantly higher microleakage values compared with SB 2/SX and SB 2/VX. Gluma presented significantly different microleakage values when used with Lumifor, Pekalux, or D 653. Gluma/I) 653 showed a higher shear bond value than Pekalux but less sealability. Vitrabond and G C lining cement restorations showed a better seal although presenting a lower bond strength. Table 5 shows the relationship between early shear bond strength and
Fig. I. This Fig. shows: (right) the marginal circumferential width of surface dye penetration evaluated along the dentin-composite (c) and enamel-composite (e) junctions and (left) the length (d) and the depth (p) of dye penetration along dentinal wall and toward the pulp. Circumferential microleakage (right) was expressed as the dyed percentage of the total width of dentin-composite or enamel-composite interface. With regard to length of microleakage (left), dye penetration up to and not beyond the apex of the cavity was assessed as 100%, while microleakage affecting both gingival and incisal walls was calculated as 200%.
marginal microleakage evaluations. No correlations were found between microleakage evaluations and shear bond strength at p = 0.05. A strong correlation was present between length and depth of dye penetration (p
Abstract-The aims of this study were to assess in vitro the early marginal seal effectiveness in Class V restorations at the cemento-enamel junction, and the early shear bond strength of several composite and dentin bonding system combinations. The seal effectiveness was studied by evaluating the dye penetrations: (a) along the superficial margin of restorations (width of surface microleakage); (b) along the dentin cavity walls (length of microleakage); and (c) toward the pulp (depth of microleakage). Good correlations were demonstrated among the seal evaluations, but no relationships were apparent between these and shear bond strength. Our study suggests that the marginal seal is influenced by composite properties more than by bonding agents. The shear bond analysis indicates only the adhesive potential of the materials without considering the composite shrinkage effects on marginal gap after the photocuring setting reactions.
arly m a r g i n a l gap formation around composite restorations is the r e s u l t of resin s h r i n k a g e during polymerization and adhesion of the resin to the wall of the cavity (Bausch et at., 1982; Davidson et al., 1984). The gap between restoration and tooth structure is responsible for bacteria and fluid penetration, marginal discoloration, and, consequently, clinical failure (Triadan, 1987). Marginal microleakage along the dentinal wall and shear bond strength tests have been frequently utilized to study the performance of dentin bonding systems (DBS) (Crim and Shay, 1987; Hammesfahr et al., 1987; Gordon et al., 1986; Phair and Fuller, 1985). However, the relationship between these two tests is not clear (Finger, 1988; Retief et al., 1988). It has been suggested that the bond strength value is not an adequate indication of the effectiveness of DBS, although it is used extensively to evaluate the bonding performances of DBS (Komatsu and Finger, 1986). The aim of this study was to evaluate the relationship between the initial adhesion to dentin and early marginal seal ability of several composites used in combination with different new DBS and glass-ionomer cement liners (GIC1). Seal ability was assessed by rheasurement of the superficial marginal circumferential leakage (Zidan et al., 1987), the length of dye penetration around the restoration walls, and the depth of dye penetration toward the pulp (Hammesfahr et al., 1987; Prati et al., 1989).
E
MATERIALS and METHODS
Materials used in this study are listed in Table 1. The combinations of DBS and composites and the sequences of application are shown in Tables 2 and 3, respectively. We used extracted human third molars of healthy young patients
(age, from 18-30 years) stored in saline solution at room temperature for no more than 72 h. The Visilux II (3M Co., St. Paul, MN) was used as a light source for polymerization initiation. M i c r o / e a k a g e Procedure. - Non-retentive Class V restorations were prepared at the cemento-enamel junction (CEJ) of buccal and lingual surfaces of each tooth. A new No. 255 Intensiv (Swiss) high-speed diamond bur with water cooling was used for the cavity preparation. The dimensions of the cavities were 3 ± 0.2 mm (mesio-distal), 3 _+ 0.2 mm (occluso-gingival), and 2.2 ± 0.2 mm (depth). All enamel margins were etched with a 37% w/w phosphoric acid gel (3M Co., St. Paul, MN) for 30 s, washed with water for 30 s, and then airdried for 30 s. Clearfil N e w Bond. Clearfil New Bond was applied with a small brush on dentin and enamel surfaces and then covered with composite. G C L i n i n g Cement. Powder and liquid were mixed according to manufacturer's directions and applied on all cavity dentin surfaces by use of the method of McLean et al. (1985). After five rain, the glass-ionomer surface and enamel were covered with a Scotchbond 2 layer that was photo-cured for 30 s. Gluma Bond. The dentin surface was treated with EDTA (pH 7.4) (Gluma cleanser, Bayer, Leverkusen, West Germany) for 60 s with a small brush, followed by water spray for 30 s and a blast of compressed air. A layer of Gluma primer was applied, dried with a gentle air spray to obtain a reflective dried dentin surface, and then coated with a layer of Gluma Resin L that was photocured for 60 s. Finally, the composite layers were applied as described. Scotchbond 2. Scotchprep primer was applied to the dentin and brushed continuously onto the surface for
Dental Materials/July 1990 195
TABLE 1 DENTINAL BONDINGSYSTEMS(DBS) AND GLASS-IONOMERCEMENTLINERS* (GlCl) USED IN THE STUDY
Material ClearfiI New Bond G C lining cement* Gluma Bond Scotchbond 2 Scotchbond DC Vitrabond lining cement* Tripton
Manufacturer Kuraray, Osaka, Japan G C International Corp., Tokyo, Japan Bayer AG, Leverkusen, W. Germany 3M, St. Paul, MN, USA 3M 3M ICI, Macclesfield, England
TABLE 2 COMPOSITESUSED IN THE STUDY
Material C[earfil Lustre Pekalux Lumifor Experimental D 653 Silux Valux P 30 P 50 Opalux Occlusin
3M Co., St. Paul, MN, USA
ICl, Macclesfield, England
Evaluation of length of dye penetration along dentinal wall (Hammesfahr et al., 1987; Gordon et al., 1986). After superficial evaluation,
TABLE 3 SEQUENCESOF APPLICATIONOF DENTIN PRE-TREATMENT,PRIMER, BONDING RESINS, AND COMPOSITE
Primer Gluma Primer
Bonding Resin Clearfil New Bond Resin L
-
Scotchprep
Scotchbond 2
-
-
Scotchbond DC
-
Tripton Dentin Primer G C lining cement Vitrabond
Tripton Universal Bond Scotchbond 2
Polyacrylic Acid -
60 s, then gently dried with compressed air for 10 s to obtain a reflective dried dentin surface. A layer of Scotehbond 2 was then applied to the dentin surface with a brush and photo-cured for 30 s. Scotchbond DC. Scotchbond DC was applied (with a small brush) onto dentin and photo-cured for 20 s. T~pton. Tripton primer was applied with a small brush to the dentin for 60 s and then gently dried for 20 s with compressed air. Tripton Resin was then applied, gently airdried, and photo-cured for 30 s.
Scotchbond 2
Composite LUSTRE PEKALUX LUMIFOR D 653 SILUX VALUX P50 SILUX P30 OCCLUSIN OPALUX SlLUX SILUX VALUX
Vitrabond. Vitrabond was prepared according to the manufacturer's directions, applied until the entire dentin surface of the cavity was coyered, and then photo-cured for 40 s (McLean eta/., 1985; Prati et a!., 1989). A layer of Scotchbond 2 was then applied and photo-cured for 30 s. The cavities were filled with the composites by use of the three-increment technique (Zidan et al., 1987). The gingival third was filled first, then the occlusal third, and finally the middle third. In the sandwich restorations, the GIC layer was
196 PRATIet al./MARGINALLEAKAGEAND DENTINADHESION
Evaluation of marginal circumferential width of surface dye penetration (Zidan e t a [ , 1987). Buccal and lingual surfaces of teeth were gently cleaned and inspected under an optical stereomicroscope (x 60) along the interface between composite and dentin or enamel. Dye penetration was used as an indication of defective sealing. Marginal microleakage was expressed as the dyed percentage of the total surface width of dentin (c) or enamel (e). Separate evaluations were made for dentin and enamel surfaces, as illustrated in Fig. 1.
Manufacturer Kuraray, Osaka, Japan Bayer AG, Leverkusen, W. Germany
Dentin Pre-treatment EDTA
considered the first gingival increment (Prati et al., 1989). After five min, the restorations were finished with abrasive discs (Sof-Lex, 3M Co., St. Paul, MN) and immediately immersed in 2% erythrosin B solution for one h, then washed in tap water and dried. Eight specimens were prepared for each combination of materials.
the teeth were longitudinally serialsectioned with a diamond disc. Dye penetration along the dentinal wall (d) was considered as the length of microleakage along the interface between dentin and composite. Two evaluations were made (mesial and distal sections) and pooled. The length of microleakage was expressed as the percentage of dye penetration with respect to the total length of the cervical dentin cavity wall. Infiltration up to and not beyond the apex of the cavity was assessed as 100%, while dye penetration affecting both walls, i.e, gingival and incisal, was calculated as 200%, as schematicized in Fig. 1.
Evaluation of depth of dye penetration toward the pulp (Hammesfahr et al., 1987). Dye penetration toward the pulp (p) was evaluated as above and expressed as the percentage of dye penetration with respect to the distance between dentinal wall and pulp.
Shear Bond Strength Procedure. Occlusal and buccal enamel were removed with a high-speed diamond bur with water cooling. The teeth were then mounted with self-curing acrylic resin in box molds. Buccal dentin was pol-
ished by use of a standard technique with a 320-grit abrasive paper (SofLex Pop-On, 3M Co., St. Paul, MN). Only dentin just below the dentinenamel junction was utilized. Each DBS was applied one or two rain after dentin preparation, as previously described. After application of material on the dentin, a cylindrical Teflon tube (4 mm internal diameter and 5 mm length) incrementally filled with composite was applied and photo-cured for 120 s. Each increment was photo-cured for 40 s. After five min, the samples were tested in a universal testing machine. The cylindrical tubes were clamped to the machine and loaded with a stainless steel wire loop to the vertical axis of the cylinder (Prati et al., 1989). The cross-head load speed was 0.5 cm/ rain. Six specimens were made for each combination of materials. The Teflon tubes were not removed during the bond strength test. No bonding occurred between the tube and the dentin-material interface. With regard to GICI, a similar technique was utilized. Cylindrical Teflon tubes (4 mm internal diameter and 3 mm length) were completely filled with the GIC1 and then applied to dentin surfaces. After five rain (Vitrabond) or 60 rain (G C lining cement), the samples were tested in a universal testing machine as previously described. In fact, complete setting of G C was obtained only after about 60 rain.
Statistical Analysis. - T h e data were analyzed by ANOVA with differences determined by Tukey's contrasts at the 95% level. Pearson's R correlation test was used to determine intra-evaluator or inter-evaluator microleakage agreement and the relationships between the microl e a k a g e values and s h e a r bond strength. No significant difference between intra-evaluator and interevaluator microleakage scores was observed. RESULTS Table 4 reports the marginal circumferential width of dye penetration, the length and the depth of microleakage, and the shear bond strength values of each combination of materials. When ANOVA and Tukey's con-
trast were used, there were statistically significant differences among the material combinations. Figs. 2, 3, and 4 show the mean microleakage value ranked by Tukey's post hoc contrast test at p < 0.05. As shown in Fig. 2, the anterior composites (Silux, Valux, Opalux, Pekalux, and Lustre) presented a lower width of microleakage with respect to the posterior composites. These results are also confirmed for the length and depth microleakage (Figs. 3 and 4). As shown in the Figs., SB 2/Silux combination presented, on the whole, the lowest microleakage values for DBS restorations. G C lining cement (FU/SX) showed the lowest microleakage values for GIC1 restorations. Shear bond strength of materials ranked by Tukey's post hoc contrast t e s t is shown in Fig. 5. W h e n Scotchbond 2 was utilized with different composites, it gave a similar bond strength (Fig. 5). However, the SB 2/P 50 combination showed significantly higher microleakage values compared with SB 2/SX and SB 2/VX. Gluma presented significantly different microleakage values when used with Lumifor, Pekalux, or D 653. Gluma/I) 653 showed a higher shear bond value than Pekalux but less sealability. Vitrabond and G C lining cement restorations showed a better seal although presenting a lower bond strength. Table 5 shows the relationship between early shear bond strength and
Fig. I. This Fig. shows: (right) the marginal circumferential width of surface dye penetration evaluated along the dentin-composite (c) and enamel-composite (e) junctions and (left) the length (d) and the depth (p) of dye penetration along dentinal wall and toward the pulp. Circumferential microleakage (right) was expressed as the dyed percentage of the total width of dentin-composite or enamel-composite interface. With regard to length of microleakage (left), dye penetration up to and not beyond the apex of the cavity was assessed as 100%, while microleakage affecting both gingival and incisal walls was calculated as 200%.
marginal microleakage evaluations. No correlations were found between microleakage evaluations and shear bond strength at p = 0.05. A strong correlation was present between length and depth of dye penetration (p
