- Email: [email protected]
Resin–dentin bond strength as related to different surface preparation methods
journal of dentistry 35 (2007) 467–475
available at www.sciencedirect.com
journal homepage: www.intl.elsevierhealth.com/journals/jden
Resin–dentin bond strength as related to different surface preparation methods Vanthana Sattabanasuk a,*, Viracha Vachiramon a,b, Fang Qian c, Steven R. Armstrong b a
Department of Conservative Dentistry and Prosthodontics, Srinakharinwirot University, Bangkok, Thailand Department of Operative Dentistry, The University of Iowa, Iowa City, IA, USA c Department of Preventive and Community Dentistry, The University of Iowa, Iowa City, IA, USA b
article info
abstract
Article history:
Objective: To determine the microtensile bond strength and micromorphological structures
Received 29 September 2006
on bonding of two adhesives (OptiBond FL and Clearfil SE Bond) to dentin surfaces ground
Received in revised form
with different preparation methods.
4 January 2007
Methods: Extracted human molars were ground flat to expose mid-coronal occlusal dentin
Accepted 4 January 2007
surface with one of six preparation methods—P120 grit SiC paper, P400 grit SiC paper, P1200 grit SiC paper, medium grit diamond bur, fine grit diamond bur, and carbide bur. Each of the adhesives was used to bond resin-based composite to the dentin surface. Dumbbell-shaped
Keywords:
specimens were fabricated and microtensile bond strengths were determined. The sub-
Carbide bur
sequent debond pathway and micromorphological structures of representative dentin
Dentin surface preparation
surfaces were examined under scanning electron microscopy. ANOVA and survival ana-
Diamond bur
lyses were performed both assuming independence from and accommodating for within-
Microtensile bond test
tooth correlation between specimens.
Silicon carbide abrasive paper
Results: By ignoring the correlations between specimens, statistical analyses revealed no
Smear layer
surface preparation effect on microtensile bond strength for each adhesive system. However, effects of surface preparation method on dentin adhesion of both adhesives were detected when accommodating for any within-tooth specimen correlations. Overall, carbide bur group showed the lowest bond strength for both OptiBond FL and Clearfil SE Bond. Dentin surfaces ground with diamond burs tended to present more compact smear layer than those ground with SiC papers and, subsequently, produced an effect on resin–dentin bond strengths. Conclusions: The dentin surface preparation method affects smear layer characteristics and dentin surface topography and, therefore affects resin–dentin bond strength. Smear layer denseness, more so than thickness, may compromise bonding efficacy of adhesives, especially of the self-etch systems. # 2007 Elsevier Ltd. All rights reserved.
1.
Introduction
Tooth preparation removes the carious defect, forms the cavity shape and establishes the substrate for the restorative material
placement. Whenever the tooth structure is altered by manual or rotary instrumentations, the cut surface is covered by a zone of debris called smear layer.1 This layer is considered as an impediment in adhesive dentistry. Therefore, to obtain an
* Corresponding author at: Department of Conservative Dentistry and Prosthodontics, Faculty of Dentistry, Srinakharinwirot University, Sukhumvit 23, Wattana District, Bangkok 10110, Thailand. Tel.: +66 2 649 5212; fax: +66 2 649 5212. E-mail address: [email protected] (V. Sattabanasuk). 0300-5712/$ – see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jdent.2007.01.002
468
journal of dentistry 35 (2007) 467–475
adequate bond to the underlying dental substrate, the smear layer must be treated prior to the application of adhesive resin.2,3 Based on the current adhesion strategy, there are two major approaches to produce an effective bond between resin and dentin. The etch-and-rinse systems employ phosphoric acid to remove the smear layer, followed by primer/adhesive applications. On the other hand, non-rinsed self-etch systems utilize acidic monomers to modify the smear layer. The subsequent bonding process incorporates this modified smear layer within the resin–dentin bond.4 Most laboratory studies have prepared the dentin surface using various grit sizes of silicon carbide (SiC) abrasive paper. The thickness and roughness of smear layers created by different SiC papers may vary and thus influence the resin adhesion, especially when the self-etch adhesives are used.5– 10 While several studies have reported low resin–dentin bond strengths over thick smear layer,5,6 there has been much published work showing no influence of smear layer thickness on the bond strength.7–10 Regardless of the coarseness of SiC paper, this surface preparation technique may not be clinically relevant. The use of SiC paper produces a looser smear layer and tends to have more open dentinal tubules than that produced by dental burs.8 Also, dental burs are available in a variety of types and coarseness which could produce qualitative and quantitative different smear layers.8,11–14 Continued studies of these issues as related to the adhesive dentistry, therefore, are needed. Besides the differences in smear layer characteristics, the surface topography of dentin created by different preparation methods may influence measured bond strength.7 Relative to SiC paper, it is difficult to obtain flat dentin surfaces when rotary instruments are used manually. To eliminate this drawback of using a clinically relevant cutting instrument, computerized numerical controlled (CNC) dental handpiece machining is recommended.15 The motivation of this work was to aid laboratory investigators in selecting an appropriate, that is, most clinically relevant surface preparation methodology for resin–dentin adhesion studies. The purpose of this study, therefore, was to investigate the microtensile bond strengths and micromorphological structures of dentin surfaces ground with different preparation methods using two adhesive systems, a three-step etch-and-rinse system and a two-step self-etch system. The hypothesis tested was that, for each adhesive approach, there would be no difference in resin– dentin bond strengths among different surface preparation methods.
2.
Materials and methods
Sixty intact, non-carious, non-restored human molars that were extracted solely for clinical purposes were selected as the substrate. The teeth were stored in 0.5% chloramine T at 4 8C and used within 3 months following extraction. The teeth were cleansed of soft tissue, roots notched, and mounted in gypsum blocks to facilitate the manipulation. A wet model trimmer (RPL 051659; Whip Mix, Louisville, KY, USA) was used to partially flatten the occlusal surface perpendicular to the long axis of the tooth without removing all of the occlusal
enamel. The teeth were randomly allocated to six groups before removing the last remnants of occlusal enamel with six different surface preparation methods: P120 grit SiC paper, P400 grit SiC paper, P1200 grit SiC paper (Carbimet; Buehler, Lake Bluff, IL, USA), a cylindrical medium grit diamond bur, a cylindrical fine grit diamond bur (ISO 806 314 008; Brasseler, Savannah, GA, USA), and a cylindrical fissure carbide bur (US No. 55; Brasseler). In detail, the dentin surfaces prepared with SiC paper were abraded with one of the different grit sizes on a table-top grinding/polishing machine (Ecomet V; Buehler) at a speed of 150 rpm under water lubrication. Each block of tooth was grasped with a specially designed specimen holder while moving and rotating the tooth under a constant load commanded by an automatic controller (RotoForce-I/RotoPol-V; Struers, Cleveland, OH, USA). Grinding periods varied by coarseness of SiC paper: 20 s for P120 grit, 40 s for P400 grit, and 90 s for P1200 grit SiC paper. Our preliminary investigations found that these periods of grinding were sufficient to remove all remnants of enamel, exposing the mid-coronal dentin surface. For the groups prepared with the dental burs, grinding was performed using the CNC Specimen Former (The University of Iowa, Iowa City, IA, USA). Each bur was used at a speed of 200,000 rpm under constant water spray and operated at equivalent feed rates. The tooth surface was deepened in 0.3 mm increments until all remnants of enamel were removed. Because the tooth block had to be steadily held in the machine throughout the preparation, complete removal of enamel was confirmed by a brief application of 37.5% phosphoric acid for 3–5 s, rinsing, and drying. Subsequently, an additional 100 mm of dentin was then removed to present a fresh substrate for bonding. Immediately after the surface preparation, the exposed dentin surfaces were bonded with one of the adhesives (Table 1) and its respective hybrid resin-based composite. OptiBond FL adhesive was used with Point 4 resin-based composite (Kerr, Orange, CA, USA) and Clearfil SE Bond with Clearfil AP-X resin-based composite (Kuraray Medical, Okayama, Japan). Each of the adhesives was used according to the manufacturers’ instructions. Initially, a thin layer of resin-based composite, approximately 0.3–0.5 mm in thickness, was bonded to the cured bonding resin with margins located entirely in dentin area. A block of resin-based composite was then incrementally built up and light-cured to a height of 4 mm over the bonded surface. All photopolymerizing steps were performed using a quartz-tungsten halogen curing unit (Optilux 500; Kerr) with the light guide held perpendicularly and within 1 mm of the surface. The light output from the curing unit was verified at not less than 600 mW/cm2 throughout the study. The bonded teeth were stored for 24 h in artificial saliva at ambient temperature.
2.1.
Bond strength testing
After storage, the teeth were sectioned perpendicular to the bonded surface using a slow-speed diamond saw (Isomet 1000; Buehler) under water cooling. Four bar-shaped specimens with square cross-sections of approximately 2 mm 2 mm wide were obtained from each tooth. The specimen was
469
journal of dentistry 35 (2007) 467–475
Table 1 – Materials, manufacturers, batch numbers, system compositions, and bonding procedures Material OptiBond FL (Kerr, Orange, CA, USA)
Batch number
Procedures
Etchant: 301194
37.5% phosphoric acid, water, silica thickener, dye colorant
Etch 15 s; rinse 15 s; blot with lint-free absorbent paper, leaving the surface visibly moist
Primer: 25881
HEMA, GPDM, PAMM, ethanol, water, photoinitiators TEGDMA, UDMA, GPDM, HEMA, bis-GMA, ytterbium trifluoride, fillers, photoinitiators, stabilizers
Apply 15 s with light scrubbing action; air dry 5 s
MDP, HEMA, hydrophilic dimethacrylate, photoinitiators, water MDP, bis-GMA, HEMA, hydrophobic dimethacrylate, photoinitiators, filler
Apply 20 s; air dry
Bond liquid: 25882
Clearfil SE Bond (Kuraray Medical, Okayama, Japan)
Compositions
Primer: 00539A
Bond liquid: 00760A
Apply and thin with new applicator; light-cure 30 s
Apply and thin with new applicator; light-cure 10 s
bis-GMA, bisphenol A diglycidylmethacrylate; GPDM, glycerol phosphate dimethacrylate; HEMA, 2-hydroxyethyl methacrylate; MDP, 10methacryloyloxydecyl dihydrogen phosphate; PAMM, phthalic acid monoethyl methacrylate; TEGDMA, triethylene glycol dimethacrylate; UDMA, urethane dimethacrylate.
placed into the pin-chuck of a lathe and the region at the bonded interface was gently milled to form a cylindrical constriction using a cylindrical ultrafine grit diamond bur (ISO 806 314 012; Brasseler) in the CNC Specimen Former. No spontaneous interfacial debonding occurred while the specimens were being fabricated. The resulted dumbbell-shaped specimens had a mean cross-sectional surface area of 0.52 0.02 mm2 at the bonded interface, with a 1 mm guage length and a 0.6 mm radius of curvature. Each specimen was examined under an optical microscope (Stemi 2000; Carl Zeiss, Thornwood, NY, USA) for any interfacial defects that would exclude the specimen from testing, e.g., gross voids, or less than 1 mm of remaining dentin thickness from the pulp. At no time throughout sectioning, trimming, and testing were the specimens permitted to dehydrate. The specimens were then
placed in a self-aligning, glue-less passive gripping device (Dirck’s device; The University of Iowa) and stressed to failure under tension with a calibrated universal testing machine (Zwick Z2.5; Zwick GmbH, Ulm, Germany) at a cross-head speed of 1 mm/min (Fig. 1).
2.2.
Scanning electron microscopic (SEM) evaluation
After debonding, all fractured specimens were dried, mounted on aluminum stubs, and sputter-coated with gold. Failure modes were observed using a scanning electron microscope (JSM-5410LV; JEOL, Tokyo, Japan). The fractured surfaces were recorded as one of the six failure categories either in dentin, resin, joint, or mixed failure modes, in which mixed failures were carefully classified as the surfaces
Fig. 1 – Schematic illustration of experimental design of the study.
470
journal of dentistry 35 (2007) 467–475
Table 2 – Mean microtensile bond strengths and standard deviations of all specimen groups (MPa) Surface preparation method (abrasive particle grit sizea)
Adhesive OptiBond FL (number of specimens tested)
P120 grit SiC paper (127 mm) P400 grit SiC paper (35 mm) P1200 grit SiC paper (15.3 mm) Medium grit diamond bur (100 mm) Fine grit diamond bur (30 mm) Carbide bur (not applicable) a
61.7 13.4 63.2 15.8 61.9 12.6 56.7 11.5 66.7 13.0 53.3 16.7
(19) (19) (18) (19) (19) (18)
Clearfil SE Bond (number of specimens tested) 54.9 12.0 51.1 19.2 43.3 16.1 42.0 18.4 43.4 14.9 41.8 19.0
(19) (18) (18) (17) (18) (18)
Information obtained from respective manufacturers.
comprising the predominance of failure of each substrate or joint. An additional 18 teeth were used to describe surface preparation effects under SEM. Dentin disks from mid-coronal dentin were prepared in the same manner as for the bond strength test. For the observation of etching characteristics on respective dentin surfaces, the phosphoric acid of OptiBond FL or the acidic primer of Clearfil SE Bond was applied to the surface as recommended by the manufacturer. After each application time, the phosphoric acid gel was flushed with water for 15 s, whereas the monomer components of the acidic primer were removed by rinsing the specimens with 100% ethanol for 5 min and placed in distilled water for further 5 min.8 The specimens were then fixed in 3% glutaraldehyde/ 3% formaldehyde in 0.1 M sodium cacodylate buffer at pH 7.35 (Tousimis Research Co., Rockville, MD, USA) overnight at 4 8C and washed in phosphate buffer solution at pH 7.2 (Spectrum, New Brunswick, NJ, USA) for 10 min with two changes, then dehydrated in ascending concentrations of ethanol and water up to 90% ethanol and placed in 100% ethanol three times, for 20 min each. The dehydrated specimens were immersed in hexamethyldisilazane solution for 30 min, placed on a filter paper inside a covered glass vial, and dried at room temperature.16 The specimens were gold sputter-coated and observed using SEM.
2.3.
using Fisher’s Exact test. All statistical analyses were performed using statistical software system (SAS/STAT version 9.1; SAS Institute Inc., Cary, NC, USA). The level of significance was set at p < 0.05 and marginally significance level was set at p 0.10.
3.
Results
The average bond strengths and standard deviations are summarized in Table 2. None of the specimens debonded before actual testing, referred to as ‘pre-testing failures’. The
Statistical analysis
The mean and standard deviation of the microtensile bond strength were calculated for each group. Within each adhesive system, comparisons between six surface preparation methods for differences in microtensile bond strength were conducted using One-way ANOVA followed by Tukey–Kramer’s post hoc-test. The analyses were done under an assumption of independence between dumbbell-shaped specimens obtained from each tooth. Also, a simple random effect in Mixed Model ANOVA was conducted to allow correlation between the specimens.17 Furthermore, parametric Weibull regression models were performed to evaluate whether there was a significant association between microtensile bond strength and the surface preparation methods within each adhesive system by Wald’s x2-test, assuming no correlation between specimens from the same tooth. A random effect Weibull regression model, known as frailty model, was also run to account for any within-tooth specimen correlations.17 The failure mode frequencies were analyzed
Fig. 2 – Weibull plot of probability of failure (%) against stress to failure (MPa) for each adhesive. s0, normalizing parameter, characteristic strength or scaling factor; m, Weibull modulus, reliability indicator or shape factor.
471
journal of dentistry 35 (2007) 467–475
Table 3 – Failure mode results Surface preparation method
P120 grit SiC paper P400 grit SiC paper P1200 grit SiC paper Medium grit diamond bur Fine grit diamond bur Carbide bur
OptiBond FL
Clearfil SE Bond
1
2
3
4
5
6
1
2
3
4
5
10 17 16 17 14 18
4 1 1 2 5 0
0 1 0 0 0 0
0 0 1 0 0 0
1 0 0 0 0 0
4 0 0 0 0 0
8 7 9 7 6 8
0 0 0 1 0 0
3 7 7 6 7 7
1 0 1 0 0 0
0 0 0 0 0 0
6 7 4 1 3 5 3
Type 1, cohesive failure in dentin; Type 2, cohesive failure in resin; Type 3, joint failure; Type 4, mixed failure involving more than 50% of dentin fracture; Type 5, mixed failure involving more than 50% of resin fracture; Type 6, mixed failure involving more than 50% of joint fracture.
data were normally distributed for both OptiBond FL ( p = 0.174) and Clearfil SE Bond ( p = 0.203) based on the Shapiro–Wilk-test. When assuming independent samples, One-way ANOVA showed overall non-significant but a marginally significant effect for the surface preparations on each adhesive ( p = 0.060 for OptiBond FL and p = 0.086 for Clearfil SE Bond). The lowest bond strength for each adhesive system was presented in the carbide bur group, which, for OptiBond FL, was significantly weaker than the fine grit diamond bur group ( p = 0.047), but not different from the remaining groups. However, for Clearfil SE Bond, Tukey– Kramer’s-test failed to show significant differences in mean microtensile bond strengths between any surface preparation methods ( p 0.181 for each instance). Including the simple random effect in Mixed Model ANOVA for within-tooth specimen correlations maintained a statistically significant effect of the surface preparation methods for OptiBond FL ( p = 0.011) but not for Clearfil SE Bond ( p = 0.527). Regarding the survival analysis, when assuming independence between specimens, Weibull regression model revealed no significant association between tensile bond strengths and the surface preparation methods for OptiBond FL ( p = 0.221) and Clearfil SE Bond adhesive system ( p = 0.285). However,
when using a random effect Weibull regression model, for accommodating correlations between specimens from the same tooth, a significant difference was shown for Clearfil SE Bond ( p = 0.040) but not for OptiBond FL ( p = 0.156). Weibull plots of probability of failure against stress to failure for test group combinations of each adhesive are shown in Fig. 2. Table 3 presents the failure mode patterns of each adhesive system. Statistically significant difference in failure distributions was found for the specimens bonded using OptiBond FL ( p < 0.0001), but not for those bonded using Clearfil SE Bond ( p = 0.755). The OptiBond FL specimens with P120 grit SiC paper for the surface preparation were less likely to have failure pathway in dentin position than other surface preparations. Fig. 3 displays the characteristics of dentin surfaces ground with different preparation methods. The apparently uneven and undulating surfaces were discerned for the groups prepared with SiC papers (Fig. 3A–C). The dentin surfaces revealed many scratches left by the abrasive papers with the entire surface covered with a distinct smear layer and the dentinal tubules were apparently occluded by smear plugs. The bur prepared groups presented smoother dentin surfaces, especially for the groups with fine grit diamond bur and
Fig. 3 – Representative SEM photographs of dentin surfaces after preparation with different methods. (A) P120 SiC paper, (B) P400 SiC paper, (C) P1200 SiC paper, (D) medium grit diamond bur, (E) fine grit diamond bur and (F) carbide bur.
472
journal of dentistry 35 (2007) 467–475
Fig. 4 – Representative SEM photographs of specimens after phosphoric acid treatment. (A) The specimen with P120 SiC paper for the surface preparation; (B) the specimen with fine grit diamond bur for the surface preparation. Patent dentinal tubules with fibrous network of intertubular dentin and peritubular dentin are observed for both groups; (C) the specimen with carbide bur for the surface preparation. Patent dentinal tubules can also be seen. The collagen fibers appear to be less distinguishable than in other groups.
carbide bur for the surface preparation (Fig. 3D–F). Fine and evenly distributed scratches were observed on the relatively flat surfaces. The smear layer was found to be more compact and denser than that produced by abrasive papers, with an evidence of dentinal tubules occluded by smear plugs.
Fig. 5 – Representative SEM photographs of specimens after acidic primer treatment. (A) The specimen with P120 SiC paper for the surface preparation. The smear layer on the dentin surface is removed, exposing the intertubular microporosities. Peritubular dentin seems to be slightly etched. There are still remnants of smear plugs within the dentinal tubules; (B) the specimen with medium grit diamond bur for the surface preparation. The intertubular dentin is conditioned while the peritubular dentin seems still intact. Remnants of smear plugs still occlude in the dentinal tubules; (C) the specimen with carbide bur for the surface preparation. While the peritubular dentin at the tubule orifices is largely conditioned, the exposed intertubular collagen network appears to be less obvious.
journal of dentistry 35 (2007) 467–475
Phosphoric acid of OptiBond FL completely removed the smear layer and smear plugs. More loose hydroxyapatitedepleted collagen fibrils were shown for the dentin surface prepared by P120 grit SiC paper and fine grit diamond bur (Fig. 4A and B, respectively), compared with that prepared by carbide bur (Fig. 4C). This feature was more evident in crosssectional views (data not shown). On the other hand, the acidic primer of Clearfil SE Bond dissolved the smear layer on the dentin surface, partially exposing the collagen fibril network (Fig. 5). For the group with P120 grit SiC paper for the surface preparation, the intertubular dentin and the peritubular dentin matrices were both etched. There are still remnants of smear plugs within the dentinal tubules (Fig. 5A). The medium grit diamond bur group presented the fibrous network of intertubular dentin. Peritubular dentin was found to be slightly etched with residual smear plugs were retained within the dentinal tubules (Fig. 5B). The dentin surface prepared with carbide bur had a smooth texture but showed only a small extent of intertubular microporosities, even though much of the peritubular dentin of the tubule orifices was removed (Fig. 5C).
4.
Discussion
The presence of the smear layer on ground dentin has been regarded as a barrier for resin infiltration during bonding. This zone of debris is a mixture of partly denatured collagen fibrils, other organic materials, and several minerals, according to the underlying dentin surface.1 Moreover, differences in surface preparation methods during laboratory testing can produce a variety of smear layer characteristics that have been reported to affect the bond strengths of resin to dentin.5,6,8,11–14,18 It would appear that the dentin surfaces prepared with SiC papers presented rough profile and thick smear layers, relating to the coarseness of the abrasive particles. This feature was also true for the dentin prepared with diamond burs in which medium grit diamond bur created a more uneven surface than did a fine grit diamond bur. This finding is in agreement with the results reported by other studies.8,13 However, comparisons between different abrasive methods with similar coarseness, i.e., P120 grit SiC paper (approximately 127 mm) and medium grit diamond bur (100 mm), and P400 grit SiC paper (approximately 35 mm) and fine grit diamond bur (30 mm), showed that SiC papers produced more irregular surfaces and thicker smear layers. This observation is in contrast with previous studies, reporting that dentin prepared with diamond burs created a rug surface without any evidence of underlying dentinal tubules.11,12 Oliveira et al.8 demonstrated that dental burs have produced a thinner but more compact smear layer than the SiC papers. An accumulation of smear layer, therefore, might reflect underlying dentin surface topography.14 In this study, flat surfaces created by burs were prepared using the CNC Specimen Former. The CNC Specimen Former greatly reduces the technique sensitivity of surface preparation since any manual manipulation is avoided.15 The employment of this, or similar,8 device that firmly holds a dental handpiece during preparation will produce flat dentin surfaces that evaluate clinically relevant smear layers by the cutting instrument of interest.
473
Carbide bur, on the other hand, generated the smoothest surface with apparently thin smear layers and smear plugs.8,12,14 This might be due to the fact that carbide bur uses bladed cutting rather than the abrasive cutting of diamond burs or SiC papers. The blades scrap the dentin and produce new surface, where as the abrasive particles wear the surface down while debris is displaced laterally by passage of the abrasive particles.19 However, even though the carbide bur created the thinnest smear layer of all burs, it tended to produce the weakest bond strength. The results are contrary to the studies that have been previously reported8,12; thus is yet to be clarified. Under TEM and micro-RAMAN spectroscopy observations, Spencer et al.20 reported that carbide bur created a fibrous smear layer, composed of well arranged and undisrupted collagen fibrils. This smear layer might not be as easily dissolved by phosphoric acid or acidic monomer conditioning steps and, subsequently, might interfere with the permeation of bonding resin.21 As resin infiltration of the demineralized intertubular dentin has been reported to account for a substantial proportion of the adhesion to dentin,2 relatively less resin infiltration in this region with carbide bur preparation might explain the lower bond strengths observed in the current investigation. Further micromorphological and chemical studies of the resin–dentin interfaces as related to different surface preparation methods, however, are required to clarify this issue. Overall, abrasive preparation methods produced comparable bond strengths. Contrary to carbide bur, SiC papers and diamond burs created loosely organized smear layers; therefore, even though thicker, they may be more easily removed prior to the application of bonding resin.20 When morphologically comparing diamond cutting instruments to SiC papers of similar abrasive particle grit size, the dentin surfaces treated with phosphoric acid showed patent dentinal tubules and exposed peritubular collagen network. Phosphoric acid could totally remove the smear layer and smear plugs, exposing the porous interfibrillar spaces, regardless of the grinding method or the coarseness of abrasive particles. The bond strength results of phosphoric acid etch-and-rinse adhesive system, OptiBond FL, seemed not to be affected by both concerned matters. Acidic primer supplied with Clearfil SE Bond, in contrast, was less effective in complete removal of the smear layer. The ability of acidic primer to etch through the smear layer might depend directly upon the smear layer denseness. This characteristic of smear layer has been reported as it varied by the method of grinding, but not by the coarseness of abrasive particles.8 With the use of SiC papers, the thicker, but less compact smear layers created by this preparation method might not hinder an efficacy of acidic primer of Clearfil SE Bond to penetrate through and hybridize the underlying dentin. The adhesive performances of Clearfil SE Bond do not seem to be dependent on the dentin smear layer thickness.5,8–10,15 On the other hand, dentin surface prepared with diamond burs, as previously mentioned, created more compact smear layers and more closed dentinal tubules than that prepared with SiC papers.8 The smear layer modification and the infiltration of bonding resin into demineralized dentin might be compromised; thus, impeding the bond strength results. With similar coarseness, dentin with SiC papers for the surface preparation tended to present
474
journal of dentistry 35 (2007) 467–475
higher bond strengths than that with diamond burs for the surface preparation using Clearfil SE Bond. This study found significant differences in dentin bond strengths between surface preparation methods when accounting for within-tooth specimen correlation. Eckert and Platt have recently reported that the analyses assuming independence between specimens would affect statistical significance levels and might lead to incorrect interpretations.17 Correlations between the specimens should be taken into account and incorporated in the statistical model.17,22 Furthermore, in some cases, statistical analyses of bond strength are also performed using Weibull statistics.17,23–25 One advantage of this test is to provide information related to the overall performance and properties of the adhesive material.23,25 It has been mentioned that the mean bond strength value represents only one point on the failure distribution curve of the resin–dentin bond, analogous to the characteristic strength (s0) in Weibull analysis.23,24 Weibull modulus (m), a dependability parameter, characterizes the spread of failure data with respect to the applied stress.23 With higher m value, a more predictable and, possibly, clinically reliable preparation method might be implied since it indicates an increased homogeneity of the flaw distribution in a material.23–25 The group with carbide bur for the dentin surface preparation method presented the lowest Weibull modulus for both adhesives with surface preparation method having a significant effect on probability of failure for Clearfil SE Bond while accounting for specimen correlation. However, the results of the current study need to be interpreted with caution since the number of test specimens used for calculating the Weibull statistics was minimally adequate for obtaining reliable conclusions.26 Within the limitations of this study, the null hypothesis, that dentin surface preparation method exerts no influence on resin bond strength, was rejected. While accounting for any within-tooth specimen correlation effects, the carbide bur preparation method revealed significant weaker bond strengths for the OptiBond FL adhesive system and, using survival analysis, the specimen preparation method also affected tensile bond strength for Clearfil SE Bond adhesive system. Surface preparation methods produce different smear layers that affect the resin–dentin bond strengths. From the clinical point of view, however, most preparations include the regions of normal and caries-affected dentin. The composition of smear layers may also change due to the underlying dental substrates from which they are formed.1 Furthermore, other additional factors, e.g., speed of dental bur, amount of pressure applied, and coolant used during preparation, could also affect the smear layer formation which, in turn, influence the resin adhesion to dentin.11,15,18,19 Dental manufacturers, therefore, are encouraged to understand the effect of their surface preparation method during the adhesive development, especially if the dentin surface is not prepared in a clinically relevant manner, i.e., abrasive paper grinding. The method of applying the cutting instrument, the cutting instrument, and the substrate being cut affect surface topography and smear layer characteristics and need to be considered in experimental methods to more fully understand the effect of dentin smear layer on resin bonding.
Acknowledgment This investigation was supported in part by Research Grant 173/2549 from Faculty of Dentistry, Srinakharinwirot University, Bangkok, Thailand.
references
[1] Eick JD, Wilko RA, Anderson CH, Sorensen SE. Scanning electron microscopy of cut tooth surfaces and identification of debris by use of the electron microprobe. Journal of Dental Research 1970;49:1359–68. [2] Swift Jr EJ, Perdiga˜o J, Heymann HO. Bonding to enamel and dentin: a brief history and state of the art, 1995. Quintessence International 1995;26:95–110. [3] Eick JD, Gwinnett AJ, Pashley DH, Robinson SJ. Current concepts on adhesion to dentin. Critical Review in Oral Biology and Medicine 1997;8:306–35. [4] Van Meerbeek B, De Munck J, Yoshida Y, Inoue S, Vargas M, Vijay P, et al. Buonocore memorial lecture. Adhesion to enamel and dentin: current status and future challenges. Operative Dentistry 2003;28:215–35. [5] Tay FR, Carvalho RM, Sano H, Pashley DH. Effect of smear layers on the bonding of a self-etching primer to dentin. Journal of Adhesive Dentistry 2000;2:99–116. [6] Koibuchi H, Yasuda N, Nakabayashi N. Bonding to dentin with a self-etching primer: the effect of smear layers. Dental Materials 2001;17:122–6. [7] Tani C, Finger WJ. Effect of smear layer thickness on bond strength mediated by three all-in-one self-etching priming adhesives. Journal of Adhesive Dentistry 2002;4:283–9. [8] Oliveira SS, Pugach MK, Hilton JF, Watanabe LG, Marshall SJ, Marshall Jr GW. The influence of the dentin smear layer on adhesion: a self-etching primer vs. a total-etch system. Dental Materials 2003;19:758–67. [9] Kenshima S, Reis A, Uceda-Gomez N, de L Tancredo L, Filho LE, Nogueira FN, et al. Effect of smear layer thickness and pH of self-etching adhesive systems on the bond strength and gap formation to dentin. Journal of Adhesive Dentistry 2005;7:117–26. [10] Reis A, Grandi V, Carlotto L, Bortoli G, Patzlaff R, de L Accorinte M, et al. Effect of smear layer thickness and acidity of self-etching solutions on early and long-term bond strength to dentin. Journal of Dentistry 2005;33: 549–59. [11] Ogata M, Harada N, Yamaguchi S, Nakajima M, Pereira PNR, Tagami J. Effects of different burs on dentin bond strengths of self-etching primer bonding systems. Operative Dentistry 2001;26:375–82. [12] Dias WR, Pereira PNR, Swift Jr EJ. Effect of bur type on microtensile bond strengths of self-etching systems to human dentin. Journal of Adhesive Dentistry 2004;6:195–203. [13] Hosoya Y, Shinkawa H, Suefiji C, Nozaka K, Garcı´a-Godoy F. Effects of diamond bur particle size on dentin bond strength. American Journal of Dentistry 2004;17:359–64. [14] Barros JA, Myaki SI, No¨r JE, Peters MC. Effect of bur type and conditioning on the surface and interface of dentine. Journal of Oral Rehabilitation 2005;32:849–56. [15] Van Meerbeek B, De Munck J, Mattar D, Van Landuyt K, Lambrechts P. Microtensile bond strengths of an etch & rinse and self-etch adhesive to enamel and dentin as a function of surface treatment. Operative Dentistry 2003;28:647–60. [16] Perdiga˜o J, Lambrechts P, Van Meerbeek B, Vanherle G, Lopes ALB. Field emission SEM comparison of four
journal of dentistry 35 (2007) 467–475
[17]
[18]
[19]
[20]
[21] [22]
postfixation drying techniques for human dentin. Journal of Biomedical Materials Research 1995;29:1111–20. Eckert GJ, Platt JA. A statistical evaluation of microtensile bond strength methodology for dental adhesives. Dental Materials 2007;23:385–91. Gupta R, Tewari S. Effect of rotary instrumentation on composite bond strength with simulated pulpal pressure. Operative Dentistry 2006;31:188–96. Thompson JY, Bayne SC. Instruments and equipment for tooth preparation. In: Roberson TM, Heymann HO, Swift Jr EJ, editors. Sturdevant’s art and science of operative dentistry. 5th ed. St. Louis: Mosby; 2006. p. 325–64. Spencer P, Wang Y, Walker MP, Swafford JR. Molecular structure of acid-etched dentin smear layers—in situ study. Journal of Dental Research 2001;80:1802–7. Pashley DH, Carvalho RM. Dentine permeability and dentine adhesion. Journal of Dentistry 1997;25:355–72. Loguercio AD, Barroso LP, Grande RH, Reis A. Comparison of intra- and intertooth resin–dentin bond strength
[23]
[24]
[25]
[26]
475
variability. Journal of Adhesive Dentistry 2005;7: 151–8. Robin C, Scherrer SS, Wiskott HW, de Rijk WG, Belser UC. Weibull parameters of composite resin bond strengths to porcelain and noble alloy using the Rocatec system. Dental Materials 2002;18:389–95. Dickens SH, Cho BH. Interpretation of bond failure through conversion and residual solvent measurements and Weibull analyses of flexural and microtensile bond strengths of bonding agents. Dental Materials 2005;21: 354–64. Burrow MF, Thomas D, Swain MV, Tyas MJ. Analysis of tensile bond strengths using Weibull statistics. Biomaterials 2004;25:5031–5. American Society for Testing and Materials. C1239-06a standard practice for reporting uniaxial strength data and estimating Weibull distribution parameters for advanced ceramics. Annual book of standards 15.01. Philadelphia: ASTM International; 2006.
available at www.sciencedirect.com
journal homepage: www.intl.elsevierhealth.com/journals/jden
Resin–dentin bond strength as related to different surface preparation methods Vanthana Sattabanasuk a,*, Viracha Vachiramon a,b, Fang Qian c, Steven R. Armstrong b a
Department of Conservative Dentistry and Prosthodontics, Srinakharinwirot University, Bangkok, Thailand Department of Operative Dentistry, The University of Iowa, Iowa City, IA, USA c Department of Preventive and Community Dentistry, The University of Iowa, Iowa City, IA, USA b
article info
abstract
Article history:
Objective: To determine the microtensile bond strength and micromorphological structures
Received 29 September 2006
on bonding of two adhesives (OptiBond FL and Clearfil SE Bond) to dentin surfaces ground
Received in revised form
with different preparation methods.
4 January 2007
Methods: Extracted human molars were ground flat to expose mid-coronal occlusal dentin
Accepted 4 January 2007
surface with one of six preparation methods—P120 grit SiC paper, P400 grit SiC paper, P1200 grit SiC paper, medium grit diamond bur, fine grit diamond bur, and carbide bur. Each of the adhesives was used to bond resin-based composite to the dentin surface. Dumbbell-shaped
Keywords:
specimens were fabricated and microtensile bond strengths were determined. The sub-
Carbide bur
sequent debond pathway and micromorphological structures of representative dentin
Dentin surface preparation
surfaces were examined under scanning electron microscopy. ANOVA and survival ana-
Diamond bur
lyses were performed both assuming independence from and accommodating for within-
Microtensile bond test
tooth correlation between specimens.
Silicon carbide abrasive paper
Results: By ignoring the correlations between specimens, statistical analyses revealed no
Smear layer
surface preparation effect on microtensile bond strength for each adhesive system. However, effects of surface preparation method on dentin adhesion of both adhesives were detected when accommodating for any within-tooth specimen correlations. Overall, carbide bur group showed the lowest bond strength for both OptiBond FL and Clearfil SE Bond. Dentin surfaces ground with diamond burs tended to present more compact smear layer than those ground with SiC papers and, subsequently, produced an effect on resin–dentin bond strengths. Conclusions: The dentin surface preparation method affects smear layer characteristics and dentin surface topography and, therefore affects resin–dentin bond strength. Smear layer denseness, more so than thickness, may compromise bonding efficacy of adhesives, especially of the self-etch systems. # 2007 Elsevier Ltd. All rights reserved.
1.
Introduction
Tooth preparation removes the carious defect, forms the cavity shape and establishes the substrate for the restorative material
placement. Whenever the tooth structure is altered by manual or rotary instrumentations, the cut surface is covered by a zone of debris called smear layer.1 This layer is considered as an impediment in adhesive dentistry. Therefore, to obtain an
* Corresponding author at: Department of Conservative Dentistry and Prosthodontics, Faculty of Dentistry, Srinakharinwirot University, Sukhumvit 23, Wattana District, Bangkok 10110, Thailand. Tel.: +66 2 649 5212; fax: +66 2 649 5212. E-mail address: [email protected] (V. Sattabanasuk). 0300-5712/$ – see front matter # 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.jdent.2007.01.002
468
journal of dentistry 35 (2007) 467–475
adequate bond to the underlying dental substrate, the smear layer must be treated prior to the application of adhesive resin.2,3 Based on the current adhesion strategy, there are two major approaches to produce an effective bond between resin and dentin. The etch-and-rinse systems employ phosphoric acid to remove the smear layer, followed by primer/adhesive applications. On the other hand, non-rinsed self-etch systems utilize acidic monomers to modify the smear layer. The subsequent bonding process incorporates this modified smear layer within the resin–dentin bond.4 Most laboratory studies have prepared the dentin surface using various grit sizes of silicon carbide (SiC) abrasive paper. The thickness and roughness of smear layers created by different SiC papers may vary and thus influence the resin adhesion, especially when the self-etch adhesives are used.5– 10 While several studies have reported low resin–dentin bond strengths over thick smear layer,5,6 there has been much published work showing no influence of smear layer thickness on the bond strength.7–10 Regardless of the coarseness of SiC paper, this surface preparation technique may not be clinically relevant. The use of SiC paper produces a looser smear layer and tends to have more open dentinal tubules than that produced by dental burs.8 Also, dental burs are available in a variety of types and coarseness which could produce qualitative and quantitative different smear layers.8,11–14 Continued studies of these issues as related to the adhesive dentistry, therefore, are needed. Besides the differences in smear layer characteristics, the surface topography of dentin created by different preparation methods may influence measured bond strength.7 Relative to SiC paper, it is difficult to obtain flat dentin surfaces when rotary instruments are used manually. To eliminate this drawback of using a clinically relevant cutting instrument, computerized numerical controlled (CNC) dental handpiece machining is recommended.15 The motivation of this work was to aid laboratory investigators in selecting an appropriate, that is, most clinically relevant surface preparation methodology for resin–dentin adhesion studies. The purpose of this study, therefore, was to investigate the microtensile bond strengths and micromorphological structures of dentin surfaces ground with different preparation methods using two adhesive systems, a three-step etch-and-rinse system and a two-step self-etch system. The hypothesis tested was that, for each adhesive approach, there would be no difference in resin– dentin bond strengths among different surface preparation methods.
2.
Materials and methods
Sixty intact, non-carious, non-restored human molars that were extracted solely for clinical purposes were selected as the substrate. The teeth were stored in 0.5% chloramine T at 4 8C and used within 3 months following extraction. The teeth were cleansed of soft tissue, roots notched, and mounted in gypsum blocks to facilitate the manipulation. A wet model trimmer (RPL 051659; Whip Mix, Louisville, KY, USA) was used to partially flatten the occlusal surface perpendicular to the long axis of the tooth without removing all of the occlusal
enamel. The teeth were randomly allocated to six groups before removing the last remnants of occlusal enamel with six different surface preparation methods: P120 grit SiC paper, P400 grit SiC paper, P1200 grit SiC paper (Carbimet; Buehler, Lake Bluff, IL, USA), a cylindrical medium grit diamond bur, a cylindrical fine grit diamond bur (ISO 806 314 008; Brasseler, Savannah, GA, USA), and a cylindrical fissure carbide bur (US No. 55; Brasseler). In detail, the dentin surfaces prepared with SiC paper were abraded with one of the different grit sizes on a table-top grinding/polishing machine (Ecomet V; Buehler) at a speed of 150 rpm under water lubrication. Each block of tooth was grasped with a specially designed specimen holder while moving and rotating the tooth under a constant load commanded by an automatic controller (RotoForce-I/RotoPol-V; Struers, Cleveland, OH, USA). Grinding periods varied by coarseness of SiC paper: 20 s for P120 grit, 40 s for P400 grit, and 90 s for P1200 grit SiC paper. Our preliminary investigations found that these periods of grinding were sufficient to remove all remnants of enamel, exposing the mid-coronal dentin surface. For the groups prepared with the dental burs, grinding was performed using the CNC Specimen Former (The University of Iowa, Iowa City, IA, USA). Each bur was used at a speed of 200,000 rpm under constant water spray and operated at equivalent feed rates. The tooth surface was deepened in 0.3 mm increments until all remnants of enamel were removed. Because the tooth block had to be steadily held in the machine throughout the preparation, complete removal of enamel was confirmed by a brief application of 37.5% phosphoric acid for 3–5 s, rinsing, and drying. Subsequently, an additional 100 mm of dentin was then removed to present a fresh substrate for bonding. Immediately after the surface preparation, the exposed dentin surfaces were bonded with one of the adhesives (Table 1) and its respective hybrid resin-based composite. OptiBond FL adhesive was used with Point 4 resin-based composite (Kerr, Orange, CA, USA) and Clearfil SE Bond with Clearfil AP-X resin-based composite (Kuraray Medical, Okayama, Japan). Each of the adhesives was used according to the manufacturers’ instructions. Initially, a thin layer of resin-based composite, approximately 0.3–0.5 mm in thickness, was bonded to the cured bonding resin with margins located entirely in dentin area. A block of resin-based composite was then incrementally built up and light-cured to a height of 4 mm over the bonded surface. All photopolymerizing steps were performed using a quartz-tungsten halogen curing unit (Optilux 500; Kerr) with the light guide held perpendicularly and within 1 mm of the surface. The light output from the curing unit was verified at not less than 600 mW/cm2 throughout the study. The bonded teeth were stored for 24 h in artificial saliva at ambient temperature.
2.1.
Bond strength testing
After storage, the teeth were sectioned perpendicular to the bonded surface using a slow-speed diamond saw (Isomet 1000; Buehler) under water cooling. Four bar-shaped specimens with square cross-sections of approximately 2 mm 2 mm wide were obtained from each tooth. The specimen was
469
journal of dentistry 35 (2007) 467–475
Table 1 – Materials, manufacturers, batch numbers, system compositions, and bonding procedures Material OptiBond FL (Kerr, Orange, CA, USA)
Batch number
Procedures
Etchant: 301194
37.5% phosphoric acid, water, silica thickener, dye colorant
Etch 15 s; rinse 15 s; blot with lint-free absorbent paper, leaving the surface visibly moist
Primer: 25881
HEMA, GPDM, PAMM, ethanol, water, photoinitiators TEGDMA, UDMA, GPDM, HEMA, bis-GMA, ytterbium trifluoride, fillers, photoinitiators, stabilizers
Apply 15 s with light scrubbing action; air dry 5 s
MDP, HEMA, hydrophilic dimethacrylate, photoinitiators, water MDP, bis-GMA, HEMA, hydrophobic dimethacrylate, photoinitiators, filler
Apply 20 s; air dry
Bond liquid: 25882
Clearfil SE Bond (Kuraray Medical, Okayama, Japan)
Compositions
Primer: 00539A
Bond liquid: 00760A
Apply and thin with new applicator; light-cure 30 s
Apply and thin with new applicator; light-cure 10 s
bis-GMA, bisphenol A diglycidylmethacrylate; GPDM, glycerol phosphate dimethacrylate; HEMA, 2-hydroxyethyl methacrylate; MDP, 10methacryloyloxydecyl dihydrogen phosphate; PAMM, phthalic acid monoethyl methacrylate; TEGDMA, triethylene glycol dimethacrylate; UDMA, urethane dimethacrylate.
placed into the pin-chuck of a lathe and the region at the bonded interface was gently milled to form a cylindrical constriction using a cylindrical ultrafine grit diamond bur (ISO 806 314 012; Brasseler) in the CNC Specimen Former. No spontaneous interfacial debonding occurred while the specimens were being fabricated. The resulted dumbbell-shaped specimens had a mean cross-sectional surface area of 0.52 0.02 mm2 at the bonded interface, with a 1 mm guage length and a 0.6 mm radius of curvature. Each specimen was examined under an optical microscope (Stemi 2000; Carl Zeiss, Thornwood, NY, USA) for any interfacial defects that would exclude the specimen from testing, e.g., gross voids, or less than 1 mm of remaining dentin thickness from the pulp. At no time throughout sectioning, trimming, and testing were the specimens permitted to dehydrate. The specimens were then
placed in a self-aligning, glue-less passive gripping device (Dirck’s device; The University of Iowa) and stressed to failure under tension with a calibrated universal testing machine (Zwick Z2.5; Zwick GmbH, Ulm, Germany) at a cross-head speed of 1 mm/min (Fig. 1).
2.2.
Scanning electron microscopic (SEM) evaluation
After debonding, all fractured specimens were dried, mounted on aluminum stubs, and sputter-coated with gold. Failure modes were observed using a scanning electron microscope (JSM-5410LV; JEOL, Tokyo, Japan). The fractured surfaces were recorded as one of the six failure categories either in dentin, resin, joint, or mixed failure modes, in which mixed failures were carefully classified as the surfaces
Fig. 1 – Schematic illustration of experimental design of the study.
470
journal of dentistry 35 (2007) 467–475
Table 2 – Mean microtensile bond strengths and standard deviations of all specimen groups (MPa) Surface preparation method (abrasive particle grit sizea)
Adhesive OptiBond FL (number of specimens tested)
P120 grit SiC paper (127 mm) P400 grit SiC paper (35 mm) P1200 grit SiC paper (15.3 mm) Medium grit diamond bur (100 mm) Fine grit diamond bur (30 mm) Carbide bur (not applicable) a
61.7 13.4 63.2 15.8 61.9 12.6 56.7 11.5 66.7 13.0 53.3 16.7
(19) (19) (18) (19) (19) (18)
Clearfil SE Bond (number of specimens tested) 54.9 12.0 51.1 19.2 43.3 16.1 42.0 18.4 43.4 14.9 41.8 19.0
(19) (18) (18) (17) (18) (18)
Information obtained from respective manufacturers.
comprising the predominance of failure of each substrate or joint. An additional 18 teeth were used to describe surface preparation effects under SEM. Dentin disks from mid-coronal dentin were prepared in the same manner as for the bond strength test. For the observation of etching characteristics on respective dentin surfaces, the phosphoric acid of OptiBond FL or the acidic primer of Clearfil SE Bond was applied to the surface as recommended by the manufacturer. After each application time, the phosphoric acid gel was flushed with water for 15 s, whereas the monomer components of the acidic primer were removed by rinsing the specimens with 100% ethanol for 5 min and placed in distilled water for further 5 min.8 The specimens were then fixed in 3% glutaraldehyde/ 3% formaldehyde in 0.1 M sodium cacodylate buffer at pH 7.35 (Tousimis Research Co., Rockville, MD, USA) overnight at 4 8C and washed in phosphate buffer solution at pH 7.2 (Spectrum, New Brunswick, NJ, USA) for 10 min with two changes, then dehydrated in ascending concentrations of ethanol and water up to 90% ethanol and placed in 100% ethanol three times, for 20 min each. The dehydrated specimens were immersed in hexamethyldisilazane solution for 30 min, placed on a filter paper inside a covered glass vial, and dried at room temperature.16 The specimens were gold sputter-coated and observed using SEM.
2.3.
using Fisher’s Exact test. All statistical analyses were performed using statistical software system (SAS/STAT version 9.1; SAS Institute Inc., Cary, NC, USA). The level of significance was set at p < 0.05 and marginally significance level was set at p 0.10.
3.
Results
The average bond strengths and standard deviations are summarized in Table 2. None of the specimens debonded before actual testing, referred to as ‘pre-testing failures’. The
Statistical analysis
The mean and standard deviation of the microtensile bond strength were calculated for each group. Within each adhesive system, comparisons between six surface preparation methods for differences in microtensile bond strength were conducted using One-way ANOVA followed by Tukey–Kramer’s post hoc-test. The analyses were done under an assumption of independence between dumbbell-shaped specimens obtained from each tooth. Also, a simple random effect in Mixed Model ANOVA was conducted to allow correlation between the specimens.17 Furthermore, parametric Weibull regression models were performed to evaluate whether there was a significant association between microtensile bond strength and the surface preparation methods within each adhesive system by Wald’s x2-test, assuming no correlation between specimens from the same tooth. A random effect Weibull regression model, known as frailty model, was also run to account for any within-tooth specimen correlations.17 The failure mode frequencies were analyzed
Fig. 2 – Weibull plot of probability of failure (%) against stress to failure (MPa) for each adhesive. s0, normalizing parameter, characteristic strength or scaling factor; m, Weibull modulus, reliability indicator or shape factor.
471
journal of dentistry 35 (2007) 467–475
Table 3 – Failure mode results Surface preparation method
P120 grit SiC paper P400 grit SiC paper P1200 grit SiC paper Medium grit diamond bur Fine grit diamond bur Carbide bur
OptiBond FL
Clearfil SE Bond
1
2
3
4
5
6
1
2
3
4
5
10 17 16 17 14 18
4 1 1 2 5 0
0 1 0 0 0 0
0 0 1 0 0 0
1 0 0 0 0 0
4 0 0 0 0 0
8 7 9 7 6 8
0 0 0 1 0 0
3 7 7 6 7 7
1 0 1 0 0 0
0 0 0 0 0 0
6 7 4 1 3 5 3
Type 1, cohesive failure in dentin; Type 2, cohesive failure in resin; Type 3, joint failure; Type 4, mixed failure involving more than 50% of dentin fracture; Type 5, mixed failure involving more than 50% of resin fracture; Type 6, mixed failure involving more than 50% of joint fracture.
data were normally distributed for both OptiBond FL ( p = 0.174) and Clearfil SE Bond ( p = 0.203) based on the Shapiro–Wilk-test. When assuming independent samples, One-way ANOVA showed overall non-significant but a marginally significant effect for the surface preparations on each adhesive ( p = 0.060 for OptiBond FL and p = 0.086 for Clearfil SE Bond). The lowest bond strength for each adhesive system was presented in the carbide bur group, which, for OptiBond FL, was significantly weaker than the fine grit diamond bur group ( p = 0.047), but not different from the remaining groups. However, for Clearfil SE Bond, Tukey– Kramer’s-test failed to show significant differences in mean microtensile bond strengths between any surface preparation methods ( p 0.181 for each instance). Including the simple random effect in Mixed Model ANOVA for within-tooth specimen correlations maintained a statistically significant effect of the surface preparation methods for OptiBond FL ( p = 0.011) but not for Clearfil SE Bond ( p = 0.527). Regarding the survival analysis, when assuming independence between specimens, Weibull regression model revealed no significant association between tensile bond strengths and the surface preparation methods for OptiBond FL ( p = 0.221) and Clearfil SE Bond adhesive system ( p = 0.285). However,
when using a random effect Weibull regression model, for accommodating correlations between specimens from the same tooth, a significant difference was shown for Clearfil SE Bond ( p = 0.040) but not for OptiBond FL ( p = 0.156). Weibull plots of probability of failure against stress to failure for test group combinations of each adhesive are shown in Fig. 2. Table 3 presents the failure mode patterns of each adhesive system. Statistically significant difference in failure distributions was found for the specimens bonded using OptiBond FL ( p < 0.0001), but not for those bonded using Clearfil SE Bond ( p = 0.755). The OptiBond FL specimens with P120 grit SiC paper for the surface preparation were less likely to have failure pathway in dentin position than other surface preparations. Fig. 3 displays the characteristics of dentin surfaces ground with different preparation methods. The apparently uneven and undulating surfaces were discerned for the groups prepared with SiC papers (Fig. 3A–C). The dentin surfaces revealed many scratches left by the abrasive papers with the entire surface covered with a distinct smear layer and the dentinal tubules were apparently occluded by smear plugs. The bur prepared groups presented smoother dentin surfaces, especially for the groups with fine grit diamond bur and
Fig. 3 – Representative SEM photographs of dentin surfaces after preparation with different methods. (A) P120 SiC paper, (B) P400 SiC paper, (C) P1200 SiC paper, (D) medium grit diamond bur, (E) fine grit diamond bur and (F) carbide bur.
472
journal of dentistry 35 (2007) 467–475
Fig. 4 – Representative SEM photographs of specimens after phosphoric acid treatment. (A) The specimen with P120 SiC paper for the surface preparation; (B) the specimen with fine grit diamond bur for the surface preparation. Patent dentinal tubules with fibrous network of intertubular dentin and peritubular dentin are observed for both groups; (C) the specimen with carbide bur for the surface preparation. Patent dentinal tubules can also be seen. The collagen fibers appear to be less distinguishable than in other groups.
carbide bur for the surface preparation (Fig. 3D–F). Fine and evenly distributed scratches were observed on the relatively flat surfaces. The smear layer was found to be more compact and denser than that produced by abrasive papers, with an evidence of dentinal tubules occluded by smear plugs.
Fig. 5 – Representative SEM photographs of specimens after acidic primer treatment. (A) The specimen with P120 SiC paper for the surface preparation. The smear layer on the dentin surface is removed, exposing the intertubular microporosities. Peritubular dentin seems to be slightly etched. There are still remnants of smear plugs within the dentinal tubules; (B) the specimen with medium grit diamond bur for the surface preparation. The intertubular dentin is conditioned while the peritubular dentin seems still intact. Remnants of smear plugs still occlude in the dentinal tubules; (C) the specimen with carbide bur for the surface preparation. While the peritubular dentin at the tubule orifices is largely conditioned, the exposed intertubular collagen network appears to be less obvious.
journal of dentistry 35 (2007) 467–475
Phosphoric acid of OptiBond FL completely removed the smear layer and smear plugs. More loose hydroxyapatitedepleted collagen fibrils were shown for the dentin surface prepared by P120 grit SiC paper and fine grit diamond bur (Fig. 4A and B, respectively), compared with that prepared by carbide bur (Fig. 4C). This feature was more evident in crosssectional views (data not shown). On the other hand, the acidic primer of Clearfil SE Bond dissolved the smear layer on the dentin surface, partially exposing the collagen fibril network (Fig. 5). For the group with P120 grit SiC paper for the surface preparation, the intertubular dentin and the peritubular dentin matrices were both etched. There are still remnants of smear plugs within the dentinal tubules (Fig. 5A). The medium grit diamond bur group presented the fibrous network of intertubular dentin. Peritubular dentin was found to be slightly etched with residual smear plugs were retained within the dentinal tubules (Fig. 5B). The dentin surface prepared with carbide bur had a smooth texture but showed only a small extent of intertubular microporosities, even though much of the peritubular dentin of the tubule orifices was removed (Fig. 5C).
4.
Discussion
The presence of the smear layer on ground dentin has been regarded as a barrier for resin infiltration during bonding. This zone of debris is a mixture of partly denatured collagen fibrils, other organic materials, and several minerals, according to the underlying dentin surface.1 Moreover, differences in surface preparation methods during laboratory testing can produce a variety of smear layer characteristics that have been reported to affect the bond strengths of resin to dentin.5,6,8,11–14,18 It would appear that the dentin surfaces prepared with SiC papers presented rough profile and thick smear layers, relating to the coarseness of the abrasive particles. This feature was also true for the dentin prepared with diamond burs in which medium grit diamond bur created a more uneven surface than did a fine grit diamond bur. This finding is in agreement with the results reported by other studies.8,13 However, comparisons between different abrasive methods with similar coarseness, i.e., P120 grit SiC paper (approximately 127 mm) and medium grit diamond bur (100 mm), and P400 grit SiC paper (approximately 35 mm) and fine grit diamond bur (30 mm), showed that SiC papers produced more irregular surfaces and thicker smear layers. This observation is in contrast with previous studies, reporting that dentin prepared with diamond burs created a rug surface without any evidence of underlying dentinal tubules.11,12 Oliveira et al.8 demonstrated that dental burs have produced a thinner but more compact smear layer than the SiC papers. An accumulation of smear layer, therefore, might reflect underlying dentin surface topography.14 In this study, flat surfaces created by burs were prepared using the CNC Specimen Former. The CNC Specimen Former greatly reduces the technique sensitivity of surface preparation since any manual manipulation is avoided.15 The employment of this, or similar,8 device that firmly holds a dental handpiece during preparation will produce flat dentin surfaces that evaluate clinically relevant smear layers by the cutting instrument of interest.
473
Carbide bur, on the other hand, generated the smoothest surface with apparently thin smear layers and smear plugs.8,12,14 This might be due to the fact that carbide bur uses bladed cutting rather than the abrasive cutting of diamond burs or SiC papers. The blades scrap the dentin and produce new surface, where as the abrasive particles wear the surface down while debris is displaced laterally by passage of the abrasive particles.19 However, even though the carbide bur created the thinnest smear layer of all burs, it tended to produce the weakest bond strength. The results are contrary to the studies that have been previously reported8,12; thus is yet to be clarified. Under TEM and micro-RAMAN spectroscopy observations, Spencer et al.20 reported that carbide bur created a fibrous smear layer, composed of well arranged and undisrupted collagen fibrils. This smear layer might not be as easily dissolved by phosphoric acid or acidic monomer conditioning steps and, subsequently, might interfere with the permeation of bonding resin.21 As resin infiltration of the demineralized intertubular dentin has been reported to account for a substantial proportion of the adhesion to dentin,2 relatively less resin infiltration in this region with carbide bur preparation might explain the lower bond strengths observed in the current investigation. Further micromorphological and chemical studies of the resin–dentin interfaces as related to different surface preparation methods, however, are required to clarify this issue. Overall, abrasive preparation methods produced comparable bond strengths. Contrary to carbide bur, SiC papers and diamond burs created loosely organized smear layers; therefore, even though thicker, they may be more easily removed prior to the application of bonding resin.20 When morphologically comparing diamond cutting instruments to SiC papers of similar abrasive particle grit size, the dentin surfaces treated with phosphoric acid showed patent dentinal tubules and exposed peritubular collagen network. Phosphoric acid could totally remove the smear layer and smear plugs, exposing the porous interfibrillar spaces, regardless of the grinding method or the coarseness of abrasive particles. The bond strength results of phosphoric acid etch-and-rinse adhesive system, OptiBond FL, seemed not to be affected by both concerned matters. Acidic primer supplied with Clearfil SE Bond, in contrast, was less effective in complete removal of the smear layer. The ability of acidic primer to etch through the smear layer might depend directly upon the smear layer denseness. This characteristic of smear layer has been reported as it varied by the method of grinding, but not by the coarseness of abrasive particles.8 With the use of SiC papers, the thicker, but less compact smear layers created by this preparation method might not hinder an efficacy of acidic primer of Clearfil SE Bond to penetrate through and hybridize the underlying dentin. The adhesive performances of Clearfil SE Bond do not seem to be dependent on the dentin smear layer thickness.5,8–10,15 On the other hand, dentin surface prepared with diamond burs, as previously mentioned, created more compact smear layers and more closed dentinal tubules than that prepared with SiC papers.8 The smear layer modification and the infiltration of bonding resin into demineralized dentin might be compromised; thus, impeding the bond strength results. With similar coarseness, dentin with SiC papers for the surface preparation tended to present
474
journal of dentistry 35 (2007) 467–475
higher bond strengths than that with diamond burs for the surface preparation using Clearfil SE Bond. This study found significant differences in dentin bond strengths between surface preparation methods when accounting for within-tooth specimen correlation. Eckert and Platt have recently reported that the analyses assuming independence between specimens would affect statistical significance levels and might lead to incorrect interpretations.17 Correlations between the specimens should be taken into account and incorporated in the statistical model.17,22 Furthermore, in some cases, statistical analyses of bond strength are also performed using Weibull statistics.17,23–25 One advantage of this test is to provide information related to the overall performance and properties of the adhesive material.23,25 It has been mentioned that the mean bond strength value represents only one point on the failure distribution curve of the resin–dentin bond, analogous to the characteristic strength (s0) in Weibull analysis.23,24 Weibull modulus (m), a dependability parameter, characterizes the spread of failure data with respect to the applied stress.23 With higher m value, a more predictable and, possibly, clinically reliable preparation method might be implied since it indicates an increased homogeneity of the flaw distribution in a material.23–25 The group with carbide bur for the dentin surface preparation method presented the lowest Weibull modulus for both adhesives with surface preparation method having a significant effect on probability of failure for Clearfil SE Bond while accounting for specimen correlation. However, the results of the current study need to be interpreted with caution since the number of test specimens used for calculating the Weibull statistics was minimally adequate for obtaining reliable conclusions.26 Within the limitations of this study, the null hypothesis, that dentin surface preparation method exerts no influence on resin bond strength, was rejected. While accounting for any within-tooth specimen correlation effects, the carbide bur preparation method revealed significant weaker bond strengths for the OptiBond FL adhesive system and, using survival analysis, the specimen preparation method also affected tensile bond strength for Clearfil SE Bond adhesive system. Surface preparation methods produce different smear layers that affect the resin–dentin bond strengths. From the clinical point of view, however, most preparations include the regions of normal and caries-affected dentin. The composition of smear layers may also change due to the underlying dental substrates from which they are formed.1 Furthermore, other additional factors, e.g., speed of dental bur, amount of pressure applied, and coolant used during preparation, could also affect the smear layer formation which, in turn, influence the resin adhesion to dentin.11,15,18,19 Dental manufacturers, therefore, are encouraged to understand the effect of their surface preparation method during the adhesive development, especially if the dentin surface is not prepared in a clinically relevant manner, i.e., abrasive paper grinding. The method of applying the cutting instrument, the cutting instrument, and the substrate being cut affect surface topography and smear layer characteristics and need to be considered in experimental methods to more fully understand the effect of dentin smear layer on resin bonding.
Acknowledgment This investigation was supported in part by Research Grant 173/2549 from Faculty of Dentistry, Srinakharinwirot University, Bangkok, Thailand.
references
[1] Eick JD, Wilko RA, Anderson CH, Sorensen SE. Scanning electron microscopy of cut tooth surfaces and identification of debris by use of the electron microprobe. Journal of Dental Research 1970;49:1359–68. [2] Swift Jr EJ, Perdiga˜o J, Heymann HO. Bonding to enamel and dentin: a brief history and state of the art, 1995. Quintessence International 1995;26:95–110. [3] Eick JD, Gwinnett AJ, Pashley DH, Robinson SJ. Current concepts on adhesion to dentin. Critical Review in Oral Biology and Medicine 1997;8:306–35. [4] Van Meerbeek B, De Munck J, Yoshida Y, Inoue S, Vargas M, Vijay P, et al. Buonocore memorial lecture. Adhesion to enamel and dentin: current status and future challenges. Operative Dentistry 2003;28:215–35. [5] Tay FR, Carvalho RM, Sano H, Pashley DH. Effect of smear layers on the bonding of a self-etching primer to dentin. Journal of Adhesive Dentistry 2000;2:99–116. [6] Koibuchi H, Yasuda N, Nakabayashi N. Bonding to dentin with a self-etching primer: the effect of smear layers. Dental Materials 2001;17:122–6. [7] Tani C, Finger WJ. Effect of smear layer thickness on bond strength mediated by three all-in-one self-etching priming adhesives. Journal of Adhesive Dentistry 2002;4:283–9. [8] Oliveira SS, Pugach MK, Hilton JF, Watanabe LG, Marshall SJ, Marshall Jr GW. The influence of the dentin smear layer on adhesion: a self-etching primer vs. a total-etch system. Dental Materials 2003;19:758–67. [9] Kenshima S, Reis A, Uceda-Gomez N, de L Tancredo L, Filho LE, Nogueira FN, et al. Effect of smear layer thickness and pH of self-etching adhesive systems on the bond strength and gap formation to dentin. Journal of Adhesive Dentistry 2005;7:117–26. [10] Reis A, Grandi V, Carlotto L, Bortoli G, Patzlaff R, de L Accorinte M, et al. Effect of smear layer thickness and acidity of self-etching solutions on early and long-term bond strength to dentin. Journal of Dentistry 2005;33: 549–59. [11] Ogata M, Harada N, Yamaguchi S, Nakajima M, Pereira PNR, Tagami J. Effects of different burs on dentin bond strengths of self-etching primer bonding systems. Operative Dentistry 2001;26:375–82. [12] Dias WR, Pereira PNR, Swift Jr EJ. Effect of bur type on microtensile bond strengths of self-etching systems to human dentin. Journal of Adhesive Dentistry 2004;6:195–203. [13] Hosoya Y, Shinkawa H, Suefiji C, Nozaka K, Garcı´a-Godoy F. Effects of diamond bur particle size on dentin bond strength. American Journal of Dentistry 2004;17:359–64. [14] Barros JA, Myaki SI, No¨r JE, Peters MC. Effect of bur type and conditioning on the surface and interface of dentine. Journal of Oral Rehabilitation 2005;32:849–56. [15] Van Meerbeek B, De Munck J, Mattar D, Van Landuyt K, Lambrechts P. Microtensile bond strengths of an etch & rinse and self-etch adhesive to enamel and dentin as a function of surface treatment. Operative Dentistry 2003;28:647–60. [16] Perdiga˜o J, Lambrechts P, Van Meerbeek B, Vanherle G, Lopes ALB. Field emission SEM comparison of four
journal of dentistry 35 (2007) 467–475
[17]
[18]
[19]
[20]
[21] [22]
postfixation drying techniques for human dentin. Journal of Biomedical Materials Research 1995;29:1111–20. Eckert GJ, Platt JA. A statistical evaluation of microtensile bond strength methodology for dental adhesives. Dental Materials 2007;23:385–91. Gupta R, Tewari S. Effect of rotary instrumentation on composite bond strength with simulated pulpal pressure. Operative Dentistry 2006;31:188–96. Thompson JY, Bayne SC. Instruments and equipment for tooth preparation. In: Roberson TM, Heymann HO, Swift Jr EJ, editors. Sturdevant’s art and science of operative dentistry. 5th ed. St. Louis: Mosby; 2006. p. 325–64. Spencer P, Wang Y, Walker MP, Swafford JR. Molecular structure of acid-etched dentin smear layers—in situ study. Journal of Dental Research 2001;80:1802–7. Pashley DH, Carvalho RM. Dentine permeability and dentine adhesion. Journal of Dentistry 1997;25:355–72. Loguercio AD, Barroso LP, Grande RH, Reis A. Comparison of intra- and intertooth resin–dentin bond strength
[23]
[24]
[25]
[26]
475
variability. Journal of Adhesive Dentistry 2005;7: 151–8. Robin C, Scherrer SS, Wiskott HW, de Rijk WG, Belser UC. Weibull parameters of composite resin bond strengths to porcelain and noble alloy using the Rocatec system. Dental Materials 2002;18:389–95. Dickens SH, Cho BH. Interpretation of bond failure through conversion and residual solvent measurements and Weibull analyses of flexural and microtensile bond strengths of bonding agents. Dental Materials 2005;21: 354–64. Burrow MF, Thomas D, Swain MV, Tyas MJ. Analysis of tensile bond strengths using Weibull statistics. Biomaterials 2004;25:5031–5. American Society for Testing and Materials. C1239-06a standard practice for reporting uniaxial strength data and estimating Weibull distribution parameters for advanced ceramics. Annual book of standards 15.01. Philadelphia: ASTM International; 2006.