- Email: [email protected]
The influence of the modification of etched bovine dentin on bond strengths
dental materials Dental Materials 16 (2000) 255–265 www.elsevier.com/locate/dental
The influence of the modification of etched bovine dentin on bond strengths S. Phrukkanon a,b, M.F. Burrow a,*, P.G. Hartley c, M.J. Tyas a a
School of Dental Science, The University of Melbourne, 711 Elizabeth Street, Melbourne, Vic. 3000, Australia b Faculty of Dentistry, Srinakharinwirot University, Bangkok, Thailand c School of Chemistry, The University of Melbourne, 711 Elizabeth Street, Melbourne, Vic. 3000, Australia Received 27 July 1999; received in revised form 2 December 1999; accepted 6 January 2000
Abstract Objective: The aim of this study was to modify demineralized bovine dentin surfaces by application of either 12.5% sodium hypochlorite (NaOCl), or 0.1% (w/w) Type I collagenase, after conditioning with phosphoric acid, to observe the demineralized surface and to investigate the effect on tensile bond strength. Methods: The NaOCl was applied to etched dentin for 30 s, 1 or 2 min and the collagenase for 1, 3 or 6 h. A control group was used without NaOCl or collagenase treatment. Prior to bonding, treated surfaces were examined using an Atomic Force Microscope (AFM). A 2.3 mm diameter area of dentin was conditioned, treated and bonded with either One Coat Bond or Single Bond following each manufacturer’s instructions, and a resin composite rod attached. Bonds were stressed in tension at a rate of 1 mm/min until failure. Mean bond strengths were calculated (MPa) and mode of failure was determined by observation at 20 × magnification. Results were analyzed using multiple regression analysis and LSD test at the 95% level of confidence (n 12: Results: AFM results showed progressive changes of the surface collagen as the treatment time of NaOCl or collagenase increased. For both bonding systems, the bond strengths of 1 min NaOCl and 3 h collagenase treatments were significantly higher than the control or other treatment groups (p ⬍ 0:05: Bond failure consisted of mostly adhesive failure between dentin and resin combined with small regions exhibiting cohesive failure of resin. Significance: Bond strengths were not dependent on the thickness of the hybrid layer, but rather quality of the hybrid layer. 䉷 2000 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved. Keywords: Dentin; Bonding; Hybrid layer; Collagenase; Sodium hypochlorite
1. Introduction The success of the bond between adhesive resin and dentin depends on the penetration of the primer and adhesive resin into the conditioned dentin surface. The application of an acidic conditioner results in removal of the smear layer, opening the dentinal tubules, demineralizing the dentin, and increasing the dentin permeability [1]. The increase in dentin permeability results in an increased wetness of the dentin surface [2]. Because of the hydrophobic nature of dentin bonding agents, the increased wetness can adversely affect the penetration of these agents into the demineralized dentin. Thus, a primer, which often contains a volatile solvent, is applied to the conditioned
* Corresponding author. Tel.: ⫹ 61-3-9341-0384; fax: ⫹ 61-3-93410437. E-mail address: [email protected] (M.F. Burrow).
dentin prior to the application of adhesive resin [3]. The primer is generally amphiphilic, containing bifunctional compounds. The hydrophilic moiety has a high affinity for the moist conditioned dentin, while the hydrophobic moiety has a high affinity for the adhesive resin. When the adhesive resin is applied and cured, the collagen fibers of the demineralized dentin become enveloped by the resinous material, and a mechanical bond between the adhesive resin and dentin is obtained. This layer, comprised of demineralized dentin and resin, is known as the ‘hybrid layer’ [4] or the ‘resin–dentin interdiffusion zone’ [5]. However, there is evidence that primer and adhesive resin cannot completely penetrate to the base of the demineralized dentin collagen layer [6–11]. Sano et al. [8,9] reported the presence of sub-micrometer leakage in the hybrid layer of dentin-bonded systems, which they called ‘nanoleakage’. Nanoleakage may allow the penetration of bacterial products, various acids and bases and water along the interface between adhesive restorations and tooth structure. The
0109-5641/00/$20.00 + 0.00 䉷 2000 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved. PII: S0109-564 1(00)00015-4
256
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
possible reason for nanoleakage may be the collapse of the superficial collagen network, the so-called ‘collagen smear layer’ [12] which might reduce the ability of resin to diffuse completely into the demineralized dentin. This problem may possibly be decreased by removing the collagen smear layer and reducing the thickness of the demineralized dentin. Recent studies investigating bonding mechanisms have shown that the thickness of the collagen matrix or the thickness of the hybrid layer may not be important to obtain good bond strengths [13,14]. Research has shown that the bond strengths obtained after the application of sodium hypochlorite (NaOCl) or collagenase following acid conditioning and prior to the priming step, could be equal to or higher than those obtained from specimens without application of NaOCl or collagenase [13–18]. However, the effect of treatment time for NaOCl and collagenase on the bond strengths and on the micromorphology of the resin–dentin interface has not been thoroughly studied. Kinney et al. [19,20] have used the atomic force microscope (AFM) to measure the hardness and elasticity of peritubular and intertubular dentin. The AFM has enabled threedimensional high resolution images of the specimen surfaces under water or liquid solvents to be observed. In addition, compared with the scanning electron microscope (SEM), AFM is a more conservative technique since there is no need for specimen fixation, coating, or desiccation. However, no research has investigated the effect of treatment time for NaOCl and collagenase on the micromorphology of the treated dentin surface using the technique. The purpose of this study was to modify phosphoric acid demineralized bovine dentin surfaces by the application of either 12.5% sodium hypochlorite or 0.1% (w/w) Type I collagenase for varying times in order to observe the influence on (a) the morphology of the treated dentin surface by atomic force microscopy (AFM), (b) tensile bond strengths, and (c) the structure of the bonded interfaces, with two dentin bonding systems. 2. Materials and methods Freshly extracted bovine lower incisor teeth were stored frozen until use. Each tooth was left at room temperature for about 3 h before the labial surface was wet ground using a model trimmer to expose a flat, superficial dentin surface. 2.1. Investigation of NaOCl or collagenase treated dentin surfaces Eight superficial dentin surfaces were polished with silicon carbide papers up to 2000-grit under running water. In order to obtain an approximately 1.0 mm thick dentin section, the teeth were cut parallel to the superficial dentin surface using a low-speed diamond saw under copious water spray. Each dentin disc was sectioned to obtain two 1 mm thick blocks of dentin approximately 2 mm × 2 mm. Two dentin blocks were conditioned according to one of six
treatments: 15% phosphoric acid (Colte`ne/Whaledent Inc., Altsta¨tten, Switzerland) or 35% phosphoric acid (3M Dental Products, St. Paul, MN) followed by 1, 3 or 6 h application of 0.1% (w/w) Type I collagenase (Sigma Aldrich Pvt. Ltd., NSW, Australia) at room temperature; or 35% phosphoric acid followed by application of 12.5% NaOCl for 30 s, 1 or 2 min at room temperature. The treated dentin blocks were thoroughly washed with water spray for 20 s and kept in distilled water prior to use. A multimode AFM (Nanoscope III Digital Instruments, Santa Barbara, CA) was used to obtain topographic images of the dentin surfaces. Imaging was performed under distilled water using the ‘tapping mode’. In this mode, a silicon nitride AFM tip was oscillated at its resonant frequency, and the damping of this oscillation, due to the presence of surface features, was monitored as the tip was raster scanned across the surface. A feedback loop adjusted the vertical position of the surface such that the oscillation amplitude remained constant. By monitoring the displacement required to keep the oscillation amplitude constant, the z-height of surface features was obtained for each x–y position of the surface beneath the tip. This information was then used to assemble a three-dimensional topographical representation of the surface. In some cases, phase imaging was also employed. This entailed monitoring the phase shift of the tapping mode oscillatory signal due to interactions with surface features, and was of particular benefit in highlighting differences in surface crystallinity. The sectioned dentin specimens were kept immersed in distilled water immediately prior to use, in order to minimize artifacts due to specimen drying. In preparation for imaging, a specimen was removed from the water and mounted on a clean glass microscope cover slip using a small quantity of 5 min curing epoxy resin (Araldite; Selleys Chemical Co., NSW, Australia). The glass slide was attached to a magnetic AFM sample puck using doublesided sticky tape, and the puck was placed on top of the piezoelectric scanner of the AFM. A drop of distilled water (⬃20 ml) was immediately placed on the exposed upper surface of the dentin section. The AFM head, including fluid cell and AFM tip, was manoeuvered into position such that the drop of distilled water formed a bridge between the cell and sample, with the AFM tip within it. The ‘O-ring’, which is normally used to seal the fluid cell onto the sample, was not employed due to the thickness of the dentin specimens. The tip was then located above the specimen using the in-built translation stage, and finally lowered towards the surface prior to the commencement of imaging. The roughness (Rq) values of the selective areas (7 mm × 7 (mm) for images were obtained using the AFM proprietary software, located in a region between the dental tubule orifices. 2.2. Bond strength measurement Each flat superficial dentin surface was finished with wet
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
257
Table 1 Materials, manufacturers, batch numbers, and system compositions (BIS-GMA bisphenol glycidyl methacrylate, HEMA 2-hydroxyethyl methacrylate, TEGDMA triethylene glycol dimethacrylate, UDMA urethane dimethacrylate) Material
Manufacturer
Batch No.
Material composition
Single Bond
3M Dental Products, St Paul, MN
Etch: 7EL Bonding Agent: 7AD
One Coat Bond
Colte`ne, Altsta¨tten, Switzerland
Etch: FTB1 Bonding Agent: FTB1
Z-100
3M Dental Products, St Paul, MN
19960829
Glaze
Bisco, Inc., Itasca, IL
129050
35% Phosphoric acid aqueous HEMA, polyalkenoic acid copolymer, BIS-GMA, photoinitiator 15% Phosphoric acid, HEMA, hydroxypropylmethacrylate, polyalkenoate methacrylized, UDMA, amorphic silicic acid Resin composite containing: 66% (v/v) colloidal zirconia/silica, BIS-GMA, TEGDMA Flowable microfill restorative composite containing: 55% (v/v) amorphous silica, BIS-GMA, UDMA
failure was recorded and the tensile bond strength (MPa) was calculated.
600-grit silicon carbide paper. Self-adhesive vinyl masking tape with a 2.3 mm diameter hole was placed on the surface in order to limit the bonding area. The materials used are listed in Table 1. Table 2 describes the 14 different treatment and bonding procedures, seven for each of Single Bond (3M Dental Products, St. Paul, MN) and One Coat Bond (Colte`ne/Whaledent Inc., Altsta¨tten, Switzerland). Twelve teeth were treated for each bonding procedure. After the surfaces were bonded, a thin layer of Z100 resin composite (3M Dental Products, St. Paul, MN) was placed on the cured bonding resin. A resin composite rod (3 mm diameter and 6 mm long), made with Z100 and reinforced with a wire loop, was attached to the resin composite and cured for a period of 20 s from three directions around the bonded area. The specimens were stored in tap water at 37⬚C for 24 h prior to testing. The tensile bond strength was obtained using a universal testing machine (Instron Model 5544; Instron Corp, Canton, MA) at a cross-head speed of 1 mm/min. The force (N) at
2.3. Fracture mode The fractured surfaces were examined using a 20 × magnification light microscope to determine the mode of fracture, and each fractured surface was allocated to one of four types: adhesive failure between bonding resin and dentin (Type 1); partial adhesive failure between the bonding resin and dentin, and partial cohesive failure in bonding resin (Type 2); partial cohesive failure in dentin (Type 3); and cohesive failure in the bonding resin (Type 4). Several fractured specimens exhibiting the different failure patterns were then selected for further investigation using a scanning electron microscope (SEM 515; Philips, Eindhoven, Netherlands) to confirm the light microscope observations. The specimens were sputter-coated with gold prior to the observations.
Table 2 Bonding procedures Material
Group
Etching
Collagenase or NaOCl treatment
Bonding
Single Bond
Control group (35% phosphoric acid)
–
Apply and reapply with brush; gently dry 5 s; light cure 20 s
Single Bond
1, 3, 6 h collagenase
Etch, 15 s; rinse, 15 s; gently dry, 5s As above
As above
Single Bond
30 s, 1 min, 2 min NaOCl
As above
One Coat Bond
Control group (15% phosphoric acid)
One Coat Bond
1, 3, 6 h collagenase
Etch, 30 s; rinse, 20 s; gently dry, 5s As above
Apply collagenase for 1 or 3 or 6 h; rinse, 30 s; gently dry, 5 s Apply NaOCl for 30 s or 1 min or 2 min; rinse, 30 s; gently dry, 5s –
One Coat Bond
30 s, 1 min, 2 min NaOCl
As above
Apply collagenase for 1 or 3 or 6 h; rinse, 30 s; gently dry, 5 s Apply NaOCl for 30 s or 1 min or 2 min; rinse, 30 s; gently dry, 5s
As above
Scrubbing motion 20 s with applicator; gently dry 5 s; light cure 30 s As above As above
258
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 1. AFM images of dentin surface treated with 35% phosphoric acid at (A) low magnification, the dentinal tubules are opened and the surface appears slightly rough (artifact on the surface, asterisk); (B) high magnification, the surface is completely covered by organic material; (C) 15% phosphoric treated dentin surface low magnification, the dentinal tubules are opened and the surface appears slightly rough; (D) high magnification, the surface is mainly covered by organic material (black arrows) but some inorganic material, believed to be hydroxyapatite, remained (white arrows).
2.4. Statistical analysis The mean and standard deviation of the tensile bond strengths were calculated for each bonding procedure. In each bonding system, the bond strength values for different procedures were compared using the least-significant difference (LSD) multiple comparison test and multiple regression analysis. The Student’s t-test was used to compare bond strengths obtained from Single Bond and One Coat Bond for each bonding procedure. The same method was employed to analyze the fracture mode differences, and the Kruskal– Wallis and Mann–Whitney U tests were used to analyze the non-parametric fracture mode data. 2.5. Investigation of the bonded interface Fourteen flat superficial dentin surfaces were prepared as for the bond test. The teeth were cut parallel to the superficial dentin surface using a low-speed diamond saw under copious water spray to obtain an approximately 1.0 mm
thick dentin section. One disk was assigned to each of the bonding procedures listed in Table 2. After curing of the bonding resin, a thin layer of flowable resin composite (Glaze; Bisco Inc., Itasca, IL) was applied and light-cured for 40 s. The specimens were stored in tap water at 37⬚C for 24 h. Shallow grooves were placed in the surface of the resin composite and on the specimen underside in dentin using a diamond disk (ISO#335-220-74; Dentsply, Milford, DE) in a slow speed handpiece under copious water spray. The grooves were checked under a light microscope to ensure that they had not cut into the bonded interface and that no crack lines occurred in the grooves. The specimens were fixed in 10% phosphate buffered formalin for 24 h and kept in phosphate buffered saline for a further 24 h. The specimens were washed with running water for 15 min, dehydrated in ethanol (at least 15 min per step) using 10, 20, 30, 50, 70% ethanol in water and 100% ethanol dried on a molecular sieve three times. The specimens were placed in a critical point drier (Samdri PVT-3; Tousimis Research Corp., Rockville, MD) until all residual moisture was
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
259
Fig. 2. AFM images of dentin surface treated with 35% phosphoric acid and NaOCl for (A) 30 s, some thick collagen fibers (black arrow) can be observed and some are covered with inorganic material (white arrow); (B) 1 min, some small fibers (black arrows) can be observed but most of the surface is inorganic material (white arrow); C) NaOCl for 2 min showing little intertubular dentin remaining; (D) collagenase for 6 h, collagen fibers in the walls of the tubules (arrow). The fibers also stretch across the tubule orifice, this may be an effect of the collagenase treatment unravelling the fibers.
removed. Specimens were fractured along the prepared grooves. Fractured surfaces were gold sputter-coated and observed using a field emission scanning electron microscope (XL30 FEG; Philips, Eindhoven, Netherlands).
3. Results 3.1. Investigation of the NaOCl and collagenase treated dentin surfaces For the control group, the dentin surface treated with 35% phosphoric acid for 15 s showed a surface covered completely with organic material believed to be a collagen smear layer (Fig. 1A and B). The surface treated with 15% phosphoric acid for 30 s was mainly covered with organic material but some hydroxyapatite remained (Fig. 1C and D). The fiber structure of the collagen could be observed more clearly on the surface treated with 15% phosphoric acid than that treated with 35% phosphoric acid. All dentin surfaces treated with NaOCl or collagenase
showed greater opening of the dentinal tubule orifices compared with the control group. For both the NaOCl and collagenase treatments, the longer the treatment time, the wider the dentinal tubules were opened. Individual collagen fibers could be observed for all dentin surfaces treated with NaOCl or collagenase. For the 30 s NaOCl specimen, some thick collagen fibers were observed and some were covered with inorganic material (Fig. 2A), while the 1 min NaOCl specimen showed what were believed to be short, small fiber-like projections with most of the surface being covered with inorganic material (Fig. 2B). The 1 h collagenase treated specimen exhibited distinct collagen fibers, while the 3 h collagenase treated specimen showed some fibers at the surface and regions of inorganic material deposits similar to the NaOCl treated specimen. Both the 2 min NaOCl and 6 h collagenase treatments severely damaged the dentin surface, with the 2 min NaOCl treatment being the most aggressive (Fig. 2C and D). However, some short collagen fibers could still be observed. The roughness values of the surfaces are shown in Fig. 3. The dentin surface treated with 15% phosphoric acid showed greater roughness
260
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 3. Graphs demonstrating the surface roughness (Rq, nm) and the bond strengths (MPa; n 12:
than that treated with 35% phosphoric acid. However, the roughness of the surface was significantly increased for all treatment times with NaOCl or collagenase. The roughness created by collagenase treatments was greater than that by NaOCl. For each treatment, there was a steady increase in roughness, which peaked at 1 min for NaOCl and 3 h for collagenase. A marked decrease in roughness for the 2 min NaOCl and 6 h collagenase treatments was then observed. 3.2. Bond strength measurement The mean tensile bond strengths and standard deviations for each group are given in Table 3. For Single Bond, the bond strengths obtained from the 30 s NaOCl, 1 min NaOCl, 1 h collagenase, and 3 h collagenase treatments were significantly higher than those obtained for the control group (p ⬍ 0:05: For One Coat Bond, the bond strengths obtained from the 1 min NaOCl and 3 h collagenase treatments were significantly greater than those obtained for the control group (p ⬍ 0:05: There was no significant difference between the two bonding systems when the mean bond strengths with the same bonding procedure were compared (p ⬎ 0:05: The multiple regression analysis determined that the Table 3 Comparison of bond strengths (SD) (MPa; n 12), affected by the surface treatment (Mean values with the superscript letters are statistically different (p ⬍ 0:05) from other treatments, when the bond strength values were compared within the same bonding system. No statistically different groups were determined (p ⬎ 0:05) when the bond strength values were compared within the same treatment groups.) Treatment
Single Bond
One Coat Bond
Control 30 s NaOCl 1 min NaOCl 2 min NaOCl 1 h Collagenase 3 h Collagenase 6 h Collagenase
15.9 (3.7) 19.3 (3.6) a 20.4 (5.6) a,b 15.3 (3.6) 19.0 (3.8) a 23.0 (4.5) a,b 15.7 (2.9)
16.3 (5.2) 17.7 (4.5) 21.2 (6.1) c 15.5 (4.0) 17.8 (4.4) 20.0 (3.2) c 16.5 (4.8)
overall effect of the surface treatments was minimal, but the duration of application of either the NaOCl or collagenase was significant (r2 0:033: There was little difference between the effects of the NaOCl or the collagenase (r 2 0:063: If any effect was apparent it was masked by the wide range of the bond strengths recorded. 3.3. Fracture mode The modes of failure for each group are shown in Table 4. The statistical analysis indicated no statistically significant difference in the fracture modes for the different bonding procedures (p ⬎ 0:05: For both bonding systems, the fractured surfaces were predominantly Type 2 (partial adhesive failure between the bonding resin and dentin, and partial cohesive failure in the bonding resin). For both bonding systems, about 20% of specimens also exhibited each of Type 1 and 4 failures. Two specimens treated with 1 min NaOCl for Single Bond and only one specimen treated with 1 h collagenase for One Coat Bond showed cohesive failure in the dentin. The scanning electron microscopy of representative specimens confirmed the fracture mode recorded by the 20 × magnification light microscope. 3.4. The investigation of the bonding interface For the control groups, both bonding systems used acid etching, which removed all of the smear layer and demineralized the dentin. A hybrid layer and resin tags could be clearly observed for both materials. For Single Bond, the hybrid layer was about 2–3 mm thick. There was a porous zone located at the base of the hybrid layer just above the mineralized dentin (Fig. 4A). The structure of the collagen fibers could be observed in this porous zone. For One Coat Bond, the hybrid layer was about 1–2 mm thick. A porous zone between the base of the hybrid layer and the mineralized dentin was also observed. However, the thickness of the hybrid layer was not uniform. For both systems, the collagen fibers in the superficial part of the hybrid layer showed what was believed to be collapse of the collagen
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
261
Table 4 Fracture mode results (n 12) (no statistically significant difference (p ⬎ 0:05) when the fracture modes were compared within the same bonding system or within the same treatment). Fracture mode: Type 1: adhesive failure between bonding resin and dentin: Type 2: partial adhesive failure between the bonding resin and dentin and partial cohesive failure in bonding resin: Type 3: partial cohesive failure in dentin: and Type 4: cohesive failure in the bonding resin. Treatment
Control 30 s NaOCl 1 min NaOCl 2 min NaOCl 1 h Collagenase 3 h Collagenase 6 h Collagenase
Single Bond
One Coat Bond
Type 1
Type 2
Type 3
Type 4
Type 1
Type 2
Type 3
Type 4
2 2 1 2 2 2 3
8 8 8 10 8 8 6
– – 2 – – – –
2 2 1 – 2 2 3
1 2 1 2 2 – –
11 9 9 10 7 8 9
– – – – 1 – –
– 1 2 – 2 4 3
network, since many collagen fibers were observed to be packed within that part of the hybrid layer (Fig. 4B). For the NaOCl treated groups, the hybrid layer was very thin and no porous zone could be observed. The longer treatment times produced wider opening of the dentinal tubules. It was difficult to distinguish any variation in structure of the hybrid layer with the NaOCl treatment times for either bonding system (Fig. 5A and C). A hybrid layer with collagen fibers enveloped by bonding resin was still observed but it was much thinner; about 0.5 mm thick and sometimes almost indistinguishable (Fig. 5B and D). The bonded interfaces between the resin and demineralized dentin treated with NaOCl were more irregular and rougher than those of the control groups. For the collagenase treated groups, the hybrid layer was about 3–5 mm thick for the 1 h collagenase treatment in both bonding systems (Fig. 6A and B). These hybrid layers were thicker than those of the control groups. For Single Bond, the porous zone at the base of the hybrid layer and the mineralized dentin was reduced. However, for One Coat Bond, the porous zone was still present. The thickness of the hybrid layer was significantly reduced to about 0.5 mm thick by 3 and 6 h collagenase treatment groups. The structure of the hybrid layer treated with the 3 and 6 h collagenase was similar to that treated with NaOCl, where the hybrid layer was very thin, almost indistinguishable, and there was no porous zone. At the junction between the resin and dentin, collagen fibers enveloped by bonding resin were observed.
4. Discussion
Fig. 4. SEM photomicrographs of fractured surface specimens of Single Bond at (A) low magnification, the hybrid layer is 2–3 mm thick, and a porous zone (arrows) is observed at the base of the hybrid layer, Bar 10 mm; (B) high magnification, the structure of the collagen fibers (white arrow) can be observed in the porous zone. The collagen fibers (black arrows) are packed in the superficial part of the hybrid layer. Bar 2 mm.
Bovine dentin has been shown to be a good substitute for human dentin in bond strength testing, although slightly lower values are frequently obtained [21–23]. The collagenous matrix of bovine dentin is mainly Type I collagen, the same as human dentin [24–26]. In the present study, several treatment times for NaOCl and Type I collagenase treatments were used in order to modify the collagen network. Both of the bonding systems used combined the primer
262
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 5. SEM photomicrographs of fractured surface specimens treated with NaOCl for 1 min of Single Bond and One Coat Bond. (A) Single Bond low magnification and (B) high magnification. Fractured specimens of One Coat Bond treated with NaOCl for 1 min, (C) low magnification and (D) high magnification. All specimens showed intimate adaptation of resin at the interface (arrow) without any porous zone; A and C. The area of what is believed to be the hybrid layer is very rough and thin, about 0.5–1 mm (arrows); B and D. Low magnification bar 10 mm; high magnification bar 1 mm.
and adhesive resin into one solution. For the control groups, the bonding resin appeared unable to fully penetrate to the base of the demineralized dentin collagen layer, creating a porous zone at the base of the hybrid layer and confirming previous findings [6–11]. The presence of a porous zone is believed to be due to residual water which prevents complete resin penetration, and/or poorly polymerized primer and adhesive resin at the base of the hybrid layer, similar to an air-inhibited layer, and which was removed by the ethanol used during critical point drying. In addition, penetration of bonding resin into the collagen fiber matrix may have been inhibited by damage to the surface of the dentinal collagen, referred to as the ‘collagen smear layer’ [12]. Tapping mode AFM was used to observe the dentin surfaces, as it is a conservative technique which allows direct observation of surfaces in the fully hydrated state. The advantage of tapping mode over conventional ‘contact mode’ was that the intermittent contact between tip and sample reduced both normal and frictional forces, thereby minimizing sample damage and artifacts which may arise.
This was of particular importance in the imaging of surfaces consisting of soft or weakly attached material. In addition, the surface roughness can be directly calculated during the imaging process. Surfaces observed using an AFM must be smooth at the micrometer scale since the cantilever tip of an AFM has an amplitude limit of about 5 mm. Therefore, the dentin surfaces prepared for this part of the experiment were finished with 2000-grit silicon carbide paper. For this reason, a direct comparison of surface roughness from the AFM and bond strength measurement is difficult, due to the differences in finishing of the surfaces. The roughness measurements included the funnel-shaped dentinal tubule orifices. The NaOCl- and collagenase-treated specimens exhibited considerably wider dentinal tubule opening and funnel-shaping at the tubule orifices. This may have produced an overestimation of the average roughness of the surface in these cases, so the results must be regarded cautiously. The greatest opening was observed at the longest treatment times for both NaOCl and collagenase, and as a result the surface structure of the longest treated dentin had very little area of intertubular dentin remaining. However,
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 6. SEM photomicrographs of fractured surface specimens treated with collagenase for 1 h of (A) Single Bond and (B) One Coat Bond. Both specimens showed the thick layer, about 5 mm, which is believed to be the hybrid layer (arrows). A porous zone (asterisks) can be observed in both systems. The hybrid layer of Single Bond appeared to be more uniform than that of One Coat Bond. Bar 10 mm.
the roughness of these dentin surfaces was decreased possibly due to artifacts in the roughness analysis, which treats void areas inside the wide, open dentin tubules as flat surfaces. The surface damage for the specimens treated with NaOCl or collagenase following acid-etching was much greater than for the control groups treated only with the acid etching, regardless of the concentration of acid used. For this reason, with respect to surface roughness and opening of tubule orifices, only surfaces etched with 35% phosphoric acid and treated with NaOCl or collagenase were observed under the AFM for all application times. AFM showed that a surface etched with 35% phosphoric acid was completely covered by a layer of organic material believed to be the ‘collagen smear layer’ (Fig. 1B). The phase imaging mode of the AFM confirmed that the structure consisted of only soft material, presumably organic. Hydroxyapatite crystals were observed attached to the
263
collagen fibers on the dentin surface treated with 15% phosphoric acid (Fig. 1D). It is possible that these crystals were responsible for the higher roughness of the surface for this group. For the collagenase treated specimens, the AFM showed many collagen fibers exposed in the 1 h specimens. This is consistent with the FE-SEM observations, where the hybrid layer was thick for both bonding systems. However, why the hybrid layers for both bonding systems were thicker than for the control groups is unclear. It is possible that the collagen fibers may not have been digested by 1 h at room temperature (about 20⬚C), but that the protein in the fibers had become swollen during the denaturing process. When the bonding resin was applied to the swollen fibers, the hybrid layer created was slightly thicker than the control. Where the collagenase had more time to digest the collagen fibers, in the 3 and 6 h collagenase specimens, the AFM showed fewer exposed collagen fibers on the surface. This is consistent with FE-SEM observations which showed the hybrid layers for both the 3 and 6 h collagenase-treated specimens were very thin, and almost undetectable at times. The higher roughness for the 3 h collgenase treatment, and subsequent lower roughness for the 6 h collagenase treatment, can be explained in a similar way to the differences observed for the 1 and 2 min NaOCl treatments. It was also noted that the tubule openings for the 6-h collagenase treated specimens seemed to have been covered with collagen fibers. This may have been caused by the effects of the collagenase unravelling and loosening the fibers. This needs to be further investigated. In the current study, as surface roughness increased, a similar increase in the bond strength was observed (Fig. 3). However, no correlation between surface roughness and bond strength was made, since the surfaces for AFM and bond strength specimens were finished differently. For the control group, the Single Bond specimens showed a more uniform hybrid layer than the One Coat Bond specimens. Both systems use HEMA as a hydrophilic monomer to penetrate the moist collagen, however, the other components and bonding procedures are different. These differences may affect the formation of the hybrid layer. The silica filler in One Coat Bond gave the bond a gel consistency, which has been claimed to improve the adhesive strength and reduce polymerization shrinkage. However, it may have an adverse effect on the penetration of the resin into the demineralized dentin. The scrubbing motion during the application of the One Coat Bond adhesive resin may damage parts of the demineralized dentin surface, which probably leads to the non-uniform thickness of the hybrid layer. Several specimens showed that the resin failed to penetrate to the base of demineralization leaving a porous region within the hybrid layer. Perdiga˜o et al. [27] have referred to this region as a ‘submicron hiatus’. The fracture modes showed most specimens exhibited partial adhesive failures between the bonding resin and dentin associated with partial cohesive failure in the
264
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
bonding resin. The fracture modes observed were similar in nature to other tests using larger surface areas [28,29]. Results from this study suggest that collagenase and NaOCl did not completely remove the collagen bundles on the surface dentin. The other published studies used the SEM for investigating the surface change after NaOCl or collagenase treatment [13–18]. The surfaces must be desiccated and coated with gold or gold–platinum in order to conduct an electron beam, which may have affected the detection of the remaining collagen fibers on the surface. The present study demonstrated that the bonding mechanism between resin and dentin was not dependent upon the thickness of the hybrid layer as stated by previous works [13,14,30]. However, we have shown that bond strengths can be increased by these treatments. The removal of the ‘collagen smear layer’ may be significant in improving the bond strengths. Optimal bond strengths can be obtained even where a very thin layer of collagen fibers exists at the dentin surface, associated with an increase in surface roughness. The bonding mechanism is believed to be dependent on not only the collagen fibers, but also on surface roughness, penetration of the bonding resin into the treated dentin, hydroxyapatite crystals projecting into the collagen matrix, and possible chemical interaction at the resin– dentin interface. In addition, penetration of the resin tags into the dentin tubules and inter-tubular anastomoses are important [14]. Further research on the bonding mechanism to dentin with regard to degree of surface demineralization to establish a good bond is now required.
Acknowledgements The research was supported by the Australian Dental Research Foundation Inc., St Leonards, NSW 2065, Australia. Dr P.G. Hartley acknowledges support from the Advanced Mineral Products Research Centre, an Australian Research Council Special Research Centre. The assistance of Jocelyn L. Carpenter, School of Botany, The University of Melbourne, with the FE-SEM imaging is also greatly appreciated.
References [1] Swift Jr. EJ, Perdiga˜o J, Heymann HO. Bonding to enamel and dentin: a brief history and state of the art. Quintessence Int 1995;26:95–110. [2] Pashley DH, Horner JA, Brewer PD. Interactions of conditioners on the dentin surface. Oper Dent 1992;17(Suppl 5):137–50. [3] Attal JP, Asmussen E, Degrange M. Effects of surface treatment on the free surface energy of dentin. Dent Mater 1994;10:259–64. [4] Nakabayashi N, Kojima K, Masuhara E. The promotion of adhesion by the infiltration of monomers into tooth substrates. J Biomed Mater Res 1982;16:265–73. [5] Van Meerbeek B, Inokoshi S, Braem M, Lambrechts P, Vanherle G. Morphological aspects of the resin–dentin interdiffusion zone with different dentin adhesive systems. J Dent Res 1992;71: 1530–40.
[6] Nakajima M, Sano H, Inai N, Tagami J, Burrow MF, Takatsu T. Resin–dentin interface—SEM observation using a freeze drying technique. In: Shimono M, Maeda T, Suda H, Takahashi K, editors. Proceedings of the International Conference on Dentin/Pulp Complex 1995 and the International Meeting on Clinical Topics of Dentin/Pulp Complex, Tokyo: Quintessence, 1995. p. 352–4. [7] Prati C, Ferrieri P, Galloni C, Mongiorgi R, Davidson CL. Dentine permeability and bond quality as affected by new bonding systems. J Dent 1995;23:217–26. [8] Sano H, Takatsu T, Ciucchi B, Horner JA, Matthews WG, Pashley DH. Nanoleakage: leakage within the hybrid layer. Oper Dent 1995;20:18–25. [9] Sano H, Yoshiyama M, Ebisu S, Burrow MF, Takatsu T, Ciucchi B, et al. Comparative SEM and TEM observations of nanoleakage within the hybrid layer. Oper Dent 1995;20:160–70. [10] Wieliczka DM, Spencer P, Kruger MB. Raman mapping of the dentin/ adhesive interface. Appl Spect 1996;50:1500–4. [11] Wieliczka DM, Kruger MB, Spencer P. Raman imaging of dental adhesive diffusion. Appl Spect 1997;51:1593–6. [12] Pashley DH, Ciucchi B, Sano H, Horner JA. Permeability of dentin to adhesive agents, Vol 24. Tokyo: Quintessence Int., 1993;24:618–31. [13] Gwinnett AJ, Tay FR, Pang KM, Wei SHY. Quantitative contribution of the collagen network in dentin hybridization. Am J Dent 1996;9:140–4. [14] Vargas MA, Cobb DS, Armstrong SR. Resin–dentin shear bond strength and interfacial ultrastructure with and without a hybrid layer. Oper Dent 1997;22:159–66. [15] Wakabayashi Y, Kondou Y, Suzuki K, Yatani H, Yamashita A. Effect of dissolution of collagen on adhesion to dentin. Int J Prosthodont 1994;7:302–6. [16] Hosoda H, Sugizaki H, Nakajima M, Shono T, Tagami J. A study on the mechanism of bonding between resin and dentin. Part 1. Bonding to dentin treated with sodium hypochlorite. Japan J Conserv Dent 1993;36:1054–8. [17] Inai N, Kanemura N, Tagami J, Watanabe LG, Marshall SJ, Marshall GW. Adhesion between collagen depleted dentin and dentin adhesives. Am J Dent 1998;11:123–7. [18] Prati C, Chersoni S, Pashley DH. Effect of removal of surface collagen fibrils on resin-dentin bonding. Dent Mater 1999;15:323–31. [19] Kinney JH, Balooch M, Marshall SJ, Marshall Jr. GW, Weihs TP. Hardness and Young’s modulus of human peritubular and intertubular dentine. Arch Oral Biol 1996;41:9–13. [20] Kinney JH, Balooch M, Marshall SJ, Marshall Jr. GW, Weihs TP. Atomic force microscope measurements of the hardness and elasticity of peritubular and intertubular human dentin. J Biomech Eng 1996;118:133–5. [21] Nakamichi I, Iwaku M, Fusayama T. Bovine teeth as possible substitutes in the adhesion test. J Dent Res 1983;62:1076–81. [22] Fowler CS, Swartz ML, Moore BK, Rhodes BF. Influence of selected variables on adhesion testing. Dent Mater 1992;8:265–9. [23] Barkmeier WW, Erickson RL. Shear bond strength of composite to enamel and dentin using Scotchbond Multi-Purpose. Am J Dent 1994;7:175–9. [24] Kuboki Y, Mechanic GL. Comparative molecular distribution of cross-link in bone and dentin collagen. Structure–function relationships. Calcif Tissue Int 1982;34:306–8. [25] Stetler-Stevenson WG, Veis A. Type I collagen shows a specific binding affinity for bovine dentin phosphophoryn. Calcif Tissue Int 1986;38:135–41. [26] Garbarsch C, Matthiessen ME, Olsen BE, Moe D, Kirkeby S. Immunohistochemistry of the intercellular matrix components and the epithelio-mesenchymal junction of the human tooth germ. Histochem J 1994;26:110–8. [27] Perdiga˜o J, Lambrechts P, Van Meerbeek B, Tome´ AR, Vanherle G, Lopes A. Morphological field emission-SEM study of the effect of six phosphoric acid etching agents. Dent Mater 1996;12:262–71.
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265 [28] Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, Carvalho R, et al. Relationship between surface area for adhesion and tensile bond strength—evaluation of a micro-tensile bond test. Dent Mater 1994;10:236–40. [29] Phrukkanon S, Burrow MF, Tyas MJ. Effect of cross-sectional surface
265
area on bond strengths between resin and dentin. Dent Mater 1998;14:120–8. [30] Burrow MF, Takakura H, Nakajima M, Inai N, Tagami J, Takatsu T. The influence of age and depth of dentin on bonding. Dent Mater 1994;10:241–6.
The influence of the modification of etched bovine dentin on bond strengths S. Phrukkanon a,b, M.F. Burrow a,*, P.G. Hartley c, M.J. Tyas a a
School of Dental Science, The University of Melbourne, 711 Elizabeth Street, Melbourne, Vic. 3000, Australia b Faculty of Dentistry, Srinakharinwirot University, Bangkok, Thailand c School of Chemistry, The University of Melbourne, 711 Elizabeth Street, Melbourne, Vic. 3000, Australia Received 27 July 1999; received in revised form 2 December 1999; accepted 6 January 2000
Abstract Objective: The aim of this study was to modify demineralized bovine dentin surfaces by application of either 12.5% sodium hypochlorite (NaOCl), or 0.1% (w/w) Type I collagenase, after conditioning with phosphoric acid, to observe the demineralized surface and to investigate the effect on tensile bond strength. Methods: The NaOCl was applied to etched dentin for 30 s, 1 or 2 min and the collagenase for 1, 3 or 6 h. A control group was used without NaOCl or collagenase treatment. Prior to bonding, treated surfaces were examined using an Atomic Force Microscope (AFM). A 2.3 mm diameter area of dentin was conditioned, treated and bonded with either One Coat Bond or Single Bond following each manufacturer’s instructions, and a resin composite rod attached. Bonds were stressed in tension at a rate of 1 mm/min until failure. Mean bond strengths were calculated (MPa) and mode of failure was determined by observation at 20 × magnification. Results were analyzed using multiple regression analysis and LSD test at the 95% level of confidence (n 12: Results: AFM results showed progressive changes of the surface collagen as the treatment time of NaOCl or collagenase increased. For both bonding systems, the bond strengths of 1 min NaOCl and 3 h collagenase treatments were significantly higher than the control or other treatment groups (p ⬍ 0:05: Bond failure consisted of mostly adhesive failure between dentin and resin combined with small regions exhibiting cohesive failure of resin. Significance: Bond strengths were not dependent on the thickness of the hybrid layer, but rather quality of the hybrid layer. 䉷 2000 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved. Keywords: Dentin; Bonding; Hybrid layer; Collagenase; Sodium hypochlorite
1. Introduction The success of the bond between adhesive resin and dentin depends on the penetration of the primer and adhesive resin into the conditioned dentin surface. The application of an acidic conditioner results in removal of the smear layer, opening the dentinal tubules, demineralizing the dentin, and increasing the dentin permeability [1]. The increase in dentin permeability results in an increased wetness of the dentin surface [2]. Because of the hydrophobic nature of dentin bonding agents, the increased wetness can adversely affect the penetration of these agents into the demineralized dentin. Thus, a primer, which often contains a volatile solvent, is applied to the conditioned
* Corresponding author. Tel.: ⫹ 61-3-9341-0384; fax: ⫹ 61-3-93410437. E-mail address: [email protected] (M.F. Burrow).
dentin prior to the application of adhesive resin [3]. The primer is generally amphiphilic, containing bifunctional compounds. The hydrophilic moiety has a high affinity for the moist conditioned dentin, while the hydrophobic moiety has a high affinity for the adhesive resin. When the adhesive resin is applied and cured, the collagen fibers of the demineralized dentin become enveloped by the resinous material, and a mechanical bond between the adhesive resin and dentin is obtained. This layer, comprised of demineralized dentin and resin, is known as the ‘hybrid layer’ [4] or the ‘resin–dentin interdiffusion zone’ [5]. However, there is evidence that primer and adhesive resin cannot completely penetrate to the base of the demineralized dentin collagen layer [6–11]. Sano et al. [8,9] reported the presence of sub-micrometer leakage in the hybrid layer of dentin-bonded systems, which they called ‘nanoleakage’. Nanoleakage may allow the penetration of bacterial products, various acids and bases and water along the interface between adhesive restorations and tooth structure. The
0109-5641/00/$20.00 + 0.00 䉷 2000 Academy of Dental Materials. Published by Elsevier Science Ltd. All rights reserved. PII: S0109-564 1(00)00015-4
256
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
possible reason for nanoleakage may be the collapse of the superficial collagen network, the so-called ‘collagen smear layer’ [12] which might reduce the ability of resin to diffuse completely into the demineralized dentin. This problem may possibly be decreased by removing the collagen smear layer and reducing the thickness of the demineralized dentin. Recent studies investigating bonding mechanisms have shown that the thickness of the collagen matrix or the thickness of the hybrid layer may not be important to obtain good bond strengths [13,14]. Research has shown that the bond strengths obtained after the application of sodium hypochlorite (NaOCl) or collagenase following acid conditioning and prior to the priming step, could be equal to or higher than those obtained from specimens without application of NaOCl or collagenase [13–18]. However, the effect of treatment time for NaOCl and collagenase on the bond strengths and on the micromorphology of the resin–dentin interface has not been thoroughly studied. Kinney et al. [19,20] have used the atomic force microscope (AFM) to measure the hardness and elasticity of peritubular and intertubular dentin. The AFM has enabled threedimensional high resolution images of the specimen surfaces under water or liquid solvents to be observed. In addition, compared with the scanning electron microscope (SEM), AFM is a more conservative technique since there is no need for specimen fixation, coating, or desiccation. However, no research has investigated the effect of treatment time for NaOCl and collagenase on the micromorphology of the treated dentin surface using the technique. The purpose of this study was to modify phosphoric acid demineralized bovine dentin surfaces by the application of either 12.5% sodium hypochlorite or 0.1% (w/w) Type I collagenase for varying times in order to observe the influence on (a) the morphology of the treated dentin surface by atomic force microscopy (AFM), (b) tensile bond strengths, and (c) the structure of the bonded interfaces, with two dentin bonding systems. 2. Materials and methods Freshly extracted bovine lower incisor teeth were stored frozen until use. Each tooth was left at room temperature for about 3 h before the labial surface was wet ground using a model trimmer to expose a flat, superficial dentin surface. 2.1. Investigation of NaOCl or collagenase treated dentin surfaces Eight superficial dentin surfaces were polished with silicon carbide papers up to 2000-grit under running water. In order to obtain an approximately 1.0 mm thick dentin section, the teeth were cut parallel to the superficial dentin surface using a low-speed diamond saw under copious water spray. Each dentin disc was sectioned to obtain two 1 mm thick blocks of dentin approximately 2 mm × 2 mm. Two dentin blocks were conditioned according to one of six
treatments: 15% phosphoric acid (Colte`ne/Whaledent Inc., Altsta¨tten, Switzerland) or 35% phosphoric acid (3M Dental Products, St. Paul, MN) followed by 1, 3 or 6 h application of 0.1% (w/w) Type I collagenase (Sigma Aldrich Pvt. Ltd., NSW, Australia) at room temperature; or 35% phosphoric acid followed by application of 12.5% NaOCl for 30 s, 1 or 2 min at room temperature. The treated dentin blocks were thoroughly washed with water spray for 20 s and kept in distilled water prior to use. A multimode AFM (Nanoscope III Digital Instruments, Santa Barbara, CA) was used to obtain topographic images of the dentin surfaces. Imaging was performed under distilled water using the ‘tapping mode’. In this mode, a silicon nitride AFM tip was oscillated at its resonant frequency, and the damping of this oscillation, due to the presence of surface features, was monitored as the tip was raster scanned across the surface. A feedback loop adjusted the vertical position of the surface such that the oscillation amplitude remained constant. By monitoring the displacement required to keep the oscillation amplitude constant, the z-height of surface features was obtained for each x–y position of the surface beneath the tip. This information was then used to assemble a three-dimensional topographical representation of the surface. In some cases, phase imaging was also employed. This entailed monitoring the phase shift of the tapping mode oscillatory signal due to interactions with surface features, and was of particular benefit in highlighting differences in surface crystallinity. The sectioned dentin specimens were kept immersed in distilled water immediately prior to use, in order to minimize artifacts due to specimen drying. In preparation for imaging, a specimen was removed from the water and mounted on a clean glass microscope cover slip using a small quantity of 5 min curing epoxy resin (Araldite; Selleys Chemical Co., NSW, Australia). The glass slide was attached to a magnetic AFM sample puck using doublesided sticky tape, and the puck was placed on top of the piezoelectric scanner of the AFM. A drop of distilled water (⬃20 ml) was immediately placed on the exposed upper surface of the dentin section. The AFM head, including fluid cell and AFM tip, was manoeuvered into position such that the drop of distilled water formed a bridge between the cell and sample, with the AFM tip within it. The ‘O-ring’, which is normally used to seal the fluid cell onto the sample, was not employed due to the thickness of the dentin specimens. The tip was then located above the specimen using the in-built translation stage, and finally lowered towards the surface prior to the commencement of imaging. The roughness (Rq) values of the selective areas (7 mm × 7 (mm) for images were obtained using the AFM proprietary software, located in a region between the dental tubule orifices. 2.2. Bond strength measurement Each flat superficial dentin surface was finished with wet
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
257
Table 1 Materials, manufacturers, batch numbers, and system compositions (BIS-GMA bisphenol glycidyl methacrylate, HEMA 2-hydroxyethyl methacrylate, TEGDMA triethylene glycol dimethacrylate, UDMA urethane dimethacrylate) Material
Manufacturer
Batch No.
Material composition
Single Bond
3M Dental Products, St Paul, MN
Etch: 7EL Bonding Agent: 7AD
One Coat Bond
Colte`ne, Altsta¨tten, Switzerland
Etch: FTB1 Bonding Agent: FTB1
Z-100
3M Dental Products, St Paul, MN
19960829
Glaze
Bisco, Inc., Itasca, IL
129050
35% Phosphoric acid aqueous HEMA, polyalkenoic acid copolymer, BIS-GMA, photoinitiator 15% Phosphoric acid, HEMA, hydroxypropylmethacrylate, polyalkenoate methacrylized, UDMA, amorphic silicic acid Resin composite containing: 66% (v/v) colloidal zirconia/silica, BIS-GMA, TEGDMA Flowable microfill restorative composite containing: 55% (v/v) amorphous silica, BIS-GMA, UDMA
failure was recorded and the tensile bond strength (MPa) was calculated.
600-grit silicon carbide paper. Self-adhesive vinyl masking tape with a 2.3 mm diameter hole was placed on the surface in order to limit the bonding area. The materials used are listed in Table 1. Table 2 describes the 14 different treatment and bonding procedures, seven for each of Single Bond (3M Dental Products, St. Paul, MN) and One Coat Bond (Colte`ne/Whaledent Inc., Altsta¨tten, Switzerland). Twelve teeth were treated for each bonding procedure. After the surfaces were bonded, a thin layer of Z100 resin composite (3M Dental Products, St. Paul, MN) was placed on the cured bonding resin. A resin composite rod (3 mm diameter and 6 mm long), made with Z100 and reinforced with a wire loop, was attached to the resin composite and cured for a period of 20 s from three directions around the bonded area. The specimens were stored in tap water at 37⬚C for 24 h prior to testing. The tensile bond strength was obtained using a universal testing machine (Instron Model 5544; Instron Corp, Canton, MA) at a cross-head speed of 1 mm/min. The force (N) at
2.3. Fracture mode The fractured surfaces were examined using a 20 × magnification light microscope to determine the mode of fracture, and each fractured surface was allocated to one of four types: adhesive failure between bonding resin and dentin (Type 1); partial adhesive failure between the bonding resin and dentin, and partial cohesive failure in bonding resin (Type 2); partial cohesive failure in dentin (Type 3); and cohesive failure in the bonding resin (Type 4). Several fractured specimens exhibiting the different failure patterns were then selected for further investigation using a scanning electron microscope (SEM 515; Philips, Eindhoven, Netherlands) to confirm the light microscope observations. The specimens were sputter-coated with gold prior to the observations.
Table 2 Bonding procedures Material
Group
Etching
Collagenase or NaOCl treatment
Bonding
Single Bond
Control group (35% phosphoric acid)
–
Apply and reapply with brush; gently dry 5 s; light cure 20 s
Single Bond
1, 3, 6 h collagenase
Etch, 15 s; rinse, 15 s; gently dry, 5s As above
As above
Single Bond
30 s, 1 min, 2 min NaOCl
As above
One Coat Bond
Control group (15% phosphoric acid)
One Coat Bond
1, 3, 6 h collagenase
Etch, 30 s; rinse, 20 s; gently dry, 5s As above
Apply collagenase for 1 or 3 or 6 h; rinse, 30 s; gently dry, 5 s Apply NaOCl for 30 s or 1 min or 2 min; rinse, 30 s; gently dry, 5s –
One Coat Bond
30 s, 1 min, 2 min NaOCl
As above
Apply collagenase for 1 or 3 or 6 h; rinse, 30 s; gently dry, 5 s Apply NaOCl for 30 s or 1 min or 2 min; rinse, 30 s; gently dry, 5s
As above
Scrubbing motion 20 s with applicator; gently dry 5 s; light cure 30 s As above As above
258
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 1. AFM images of dentin surface treated with 35% phosphoric acid at (A) low magnification, the dentinal tubules are opened and the surface appears slightly rough (artifact on the surface, asterisk); (B) high magnification, the surface is completely covered by organic material; (C) 15% phosphoric treated dentin surface low magnification, the dentinal tubules are opened and the surface appears slightly rough; (D) high magnification, the surface is mainly covered by organic material (black arrows) but some inorganic material, believed to be hydroxyapatite, remained (white arrows).
2.4. Statistical analysis The mean and standard deviation of the tensile bond strengths were calculated for each bonding procedure. In each bonding system, the bond strength values for different procedures were compared using the least-significant difference (LSD) multiple comparison test and multiple regression analysis. The Student’s t-test was used to compare bond strengths obtained from Single Bond and One Coat Bond for each bonding procedure. The same method was employed to analyze the fracture mode differences, and the Kruskal– Wallis and Mann–Whitney U tests were used to analyze the non-parametric fracture mode data. 2.5. Investigation of the bonded interface Fourteen flat superficial dentin surfaces were prepared as for the bond test. The teeth were cut parallel to the superficial dentin surface using a low-speed diamond saw under copious water spray to obtain an approximately 1.0 mm
thick dentin section. One disk was assigned to each of the bonding procedures listed in Table 2. After curing of the bonding resin, a thin layer of flowable resin composite (Glaze; Bisco Inc., Itasca, IL) was applied and light-cured for 40 s. The specimens were stored in tap water at 37⬚C for 24 h. Shallow grooves were placed in the surface of the resin composite and on the specimen underside in dentin using a diamond disk (ISO#335-220-74; Dentsply, Milford, DE) in a slow speed handpiece under copious water spray. The grooves were checked under a light microscope to ensure that they had not cut into the bonded interface and that no crack lines occurred in the grooves. The specimens were fixed in 10% phosphate buffered formalin for 24 h and kept in phosphate buffered saline for a further 24 h. The specimens were washed with running water for 15 min, dehydrated in ethanol (at least 15 min per step) using 10, 20, 30, 50, 70% ethanol in water and 100% ethanol dried on a molecular sieve three times. The specimens were placed in a critical point drier (Samdri PVT-3; Tousimis Research Corp., Rockville, MD) until all residual moisture was
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
259
Fig. 2. AFM images of dentin surface treated with 35% phosphoric acid and NaOCl for (A) 30 s, some thick collagen fibers (black arrow) can be observed and some are covered with inorganic material (white arrow); (B) 1 min, some small fibers (black arrows) can be observed but most of the surface is inorganic material (white arrow); C) NaOCl for 2 min showing little intertubular dentin remaining; (D) collagenase for 6 h, collagen fibers in the walls of the tubules (arrow). The fibers also stretch across the tubule orifice, this may be an effect of the collagenase treatment unravelling the fibers.
removed. Specimens were fractured along the prepared grooves. Fractured surfaces were gold sputter-coated and observed using a field emission scanning electron microscope (XL30 FEG; Philips, Eindhoven, Netherlands).
3. Results 3.1. Investigation of the NaOCl and collagenase treated dentin surfaces For the control group, the dentin surface treated with 35% phosphoric acid for 15 s showed a surface covered completely with organic material believed to be a collagen smear layer (Fig. 1A and B). The surface treated with 15% phosphoric acid for 30 s was mainly covered with organic material but some hydroxyapatite remained (Fig. 1C and D). The fiber structure of the collagen could be observed more clearly on the surface treated with 15% phosphoric acid than that treated with 35% phosphoric acid. All dentin surfaces treated with NaOCl or collagenase
showed greater opening of the dentinal tubule orifices compared with the control group. For both the NaOCl and collagenase treatments, the longer the treatment time, the wider the dentinal tubules were opened. Individual collagen fibers could be observed for all dentin surfaces treated with NaOCl or collagenase. For the 30 s NaOCl specimen, some thick collagen fibers were observed and some were covered with inorganic material (Fig. 2A), while the 1 min NaOCl specimen showed what were believed to be short, small fiber-like projections with most of the surface being covered with inorganic material (Fig. 2B). The 1 h collagenase treated specimen exhibited distinct collagen fibers, while the 3 h collagenase treated specimen showed some fibers at the surface and regions of inorganic material deposits similar to the NaOCl treated specimen. Both the 2 min NaOCl and 6 h collagenase treatments severely damaged the dentin surface, with the 2 min NaOCl treatment being the most aggressive (Fig. 2C and D). However, some short collagen fibers could still be observed. The roughness values of the surfaces are shown in Fig. 3. The dentin surface treated with 15% phosphoric acid showed greater roughness
260
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 3. Graphs demonstrating the surface roughness (Rq, nm) and the bond strengths (MPa; n 12:
than that treated with 35% phosphoric acid. However, the roughness of the surface was significantly increased for all treatment times with NaOCl or collagenase. The roughness created by collagenase treatments was greater than that by NaOCl. For each treatment, there was a steady increase in roughness, which peaked at 1 min for NaOCl and 3 h for collagenase. A marked decrease in roughness for the 2 min NaOCl and 6 h collagenase treatments was then observed. 3.2. Bond strength measurement The mean tensile bond strengths and standard deviations for each group are given in Table 3. For Single Bond, the bond strengths obtained from the 30 s NaOCl, 1 min NaOCl, 1 h collagenase, and 3 h collagenase treatments were significantly higher than those obtained for the control group (p ⬍ 0:05: For One Coat Bond, the bond strengths obtained from the 1 min NaOCl and 3 h collagenase treatments were significantly greater than those obtained for the control group (p ⬍ 0:05: There was no significant difference between the two bonding systems when the mean bond strengths with the same bonding procedure were compared (p ⬎ 0:05: The multiple regression analysis determined that the Table 3 Comparison of bond strengths (SD) (MPa; n 12), affected by the surface treatment (Mean values with the superscript letters are statistically different (p ⬍ 0:05) from other treatments, when the bond strength values were compared within the same bonding system. No statistically different groups were determined (p ⬎ 0:05) when the bond strength values were compared within the same treatment groups.) Treatment
Single Bond
One Coat Bond
Control 30 s NaOCl 1 min NaOCl 2 min NaOCl 1 h Collagenase 3 h Collagenase 6 h Collagenase
15.9 (3.7) 19.3 (3.6) a 20.4 (5.6) a,b 15.3 (3.6) 19.0 (3.8) a 23.0 (4.5) a,b 15.7 (2.9)
16.3 (5.2) 17.7 (4.5) 21.2 (6.1) c 15.5 (4.0) 17.8 (4.4) 20.0 (3.2) c 16.5 (4.8)
overall effect of the surface treatments was minimal, but the duration of application of either the NaOCl or collagenase was significant (r2 0:033: There was little difference between the effects of the NaOCl or the collagenase (r 2 0:063: If any effect was apparent it was masked by the wide range of the bond strengths recorded. 3.3. Fracture mode The modes of failure for each group are shown in Table 4. The statistical analysis indicated no statistically significant difference in the fracture modes for the different bonding procedures (p ⬎ 0:05: For both bonding systems, the fractured surfaces were predominantly Type 2 (partial adhesive failure between the bonding resin and dentin, and partial cohesive failure in the bonding resin). For both bonding systems, about 20% of specimens also exhibited each of Type 1 and 4 failures. Two specimens treated with 1 min NaOCl for Single Bond and only one specimen treated with 1 h collagenase for One Coat Bond showed cohesive failure in the dentin. The scanning electron microscopy of representative specimens confirmed the fracture mode recorded by the 20 × magnification light microscope. 3.4. The investigation of the bonding interface For the control groups, both bonding systems used acid etching, which removed all of the smear layer and demineralized the dentin. A hybrid layer and resin tags could be clearly observed for both materials. For Single Bond, the hybrid layer was about 2–3 mm thick. There was a porous zone located at the base of the hybrid layer just above the mineralized dentin (Fig. 4A). The structure of the collagen fibers could be observed in this porous zone. For One Coat Bond, the hybrid layer was about 1–2 mm thick. A porous zone between the base of the hybrid layer and the mineralized dentin was also observed. However, the thickness of the hybrid layer was not uniform. For both systems, the collagen fibers in the superficial part of the hybrid layer showed what was believed to be collapse of the collagen
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
261
Table 4 Fracture mode results (n 12) (no statistically significant difference (p ⬎ 0:05) when the fracture modes were compared within the same bonding system or within the same treatment). Fracture mode: Type 1: adhesive failure between bonding resin and dentin: Type 2: partial adhesive failure between the bonding resin and dentin and partial cohesive failure in bonding resin: Type 3: partial cohesive failure in dentin: and Type 4: cohesive failure in the bonding resin. Treatment
Control 30 s NaOCl 1 min NaOCl 2 min NaOCl 1 h Collagenase 3 h Collagenase 6 h Collagenase
Single Bond
One Coat Bond
Type 1
Type 2
Type 3
Type 4
Type 1
Type 2
Type 3
Type 4
2 2 1 2 2 2 3
8 8 8 10 8 8 6
– – 2 – – – –
2 2 1 – 2 2 3
1 2 1 2 2 – –
11 9 9 10 7 8 9
– – – – 1 – –
– 1 2 – 2 4 3
network, since many collagen fibers were observed to be packed within that part of the hybrid layer (Fig. 4B). For the NaOCl treated groups, the hybrid layer was very thin and no porous zone could be observed. The longer treatment times produced wider opening of the dentinal tubules. It was difficult to distinguish any variation in structure of the hybrid layer with the NaOCl treatment times for either bonding system (Fig. 5A and C). A hybrid layer with collagen fibers enveloped by bonding resin was still observed but it was much thinner; about 0.5 mm thick and sometimes almost indistinguishable (Fig. 5B and D). The bonded interfaces between the resin and demineralized dentin treated with NaOCl were more irregular and rougher than those of the control groups. For the collagenase treated groups, the hybrid layer was about 3–5 mm thick for the 1 h collagenase treatment in both bonding systems (Fig. 6A and B). These hybrid layers were thicker than those of the control groups. For Single Bond, the porous zone at the base of the hybrid layer and the mineralized dentin was reduced. However, for One Coat Bond, the porous zone was still present. The thickness of the hybrid layer was significantly reduced to about 0.5 mm thick by 3 and 6 h collagenase treatment groups. The structure of the hybrid layer treated with the 3 and 6 h collagenase was similar to that treated with NaOCl, where the hybrid layer was very thin, almost indistinguishable, and there was no porous zone. At the junction between the resin and dentin, collagen fibers enveloped by bonding resin were observed.
4. Discussion
Fig. 4. SEM photomicrographs of fractured surface specimens of Single Bond at (A) low magnification, the hybrid layer is 2–3 mm thick, and a porous zone (arrows) is observed at the base of the hybrid layer, Bar 10 mm; (B) high magnification, the structure of the collagen fibers (white arrow) can be observed in the porous zone. The collagen fibers (black arrows) are packed in the superficial part of the hybrid layer. Bar 2 mm.
Bovine dentin has been shown to be a good substitute for human dentin in bond strength testing, although slightly lower values are frequently obtained [21–23]. The collagenous matrix of bovine dentin is mainly Type I collagen, the same as human dentin [24–26]. In the present study, several treatment times for NaOCl and Type I collagenase treatments were used in order to modify the collagen network. Both of the bonding systems used combined the primer
262
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 5. SEM photomicrographs of fractured surface specimens treated with NaOCl for 1 min of Single Bond and One Coat Bond. (A) Single Bond low magnification and (B) high magnification. Fractured specimens of One Coat Bond treated with NaOCl for 1 min, (C) low magnification and (D) high magnification. All specimens showed intimate adaptation of resin at the interface (arrow) without any porous zone; A and C. The area of what is believed to be the hybrid layer is very rough and thin, about 0.5–1 mm (arrows); B and D. Low magnification bar 10 mm; high magnification bar 1 mm.
and adhesive resin into one solution. For the control groups, the bonding resin appeared unable to fully penetrate to the base of the demineralized dentin collagen layer, creating a porous zone at the base of the hybrid layer and confirming previous findings [6–11]. The presence of a porous zone is believed to be due to residual water which prevents complete resin penetration, and/or poorly polymerized primer and adhesive resin at the base of the hybrid layer, similar to an air-inhibited layer, and which was removed by the ethanol used during critical point drying. In addition, penetration of bonding resin into the collagen fiber matrix may have been inhibited by damage to the surface of the dentinal collagen, referred to as the ‘collagen smear layer’ [12]. Tapping mode AFM was used to observe the dentin surfaces, as it is a conservative technique which allows direct observation of surfaces in the fully hydrated state. The advantage of tapping mode over conventional ‘contact mode’ was that the intermittent contact between tip and sample reduced both normal and frictional forces, thereby minimizing sample damage and artifacts which may arise.
This was of particular importance in the imaging of surfaces consisting of soft or weakly attached material. In addition, the surface roughness can be directly calculated during the imaging process. Surfaces observed using an AFM must be smooth at the micrometer scale since the cantilever tip of an AFM has an amplitude limit of about 5 mm. Therefore, the dentin surfaces prepared for this part of the experiment were finished with 2000-grit silicon carbide paper. For this reason, a direct comparison of surface roughness from the AFM and bond strength measurement is difficult, due to the differences in finishing of the surfaces. The roughness measurements included the funnel-shaped dentinal tubule orifices. The NaOCl- and collagenase-treated specimens exhibited considerably wider dentinal tubule opening and funnel-shaping at the tubule orifices. This may have produced an overestimation of the average roughness of the surface in these cases, so the results must be regarded cautiously. The greatest opening was observed at the longest treatment times for both NaOCl and collagenase, and as a result the surface structure of the longest treated dentin had very little area of intertubular dentin remaining. However,
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
Fig. 6. SEM photomicrographs of fractured surface specimens treated with collagenase for 1 h of (A) Single Bond and (B) One Coat Bond. Both specimens showed the thick layer, about 5 mm, which is believed to be the hybrid layer (arrows). A porous zone (asterisks) can be observed in both systems. The hybrid layer of Single Bond appeared to be more uniform than that of One Coat Bond. Bar 10 mm.
the roughness of these dentin surfaces was decreased possibly due to artifacts in the roughness analysis, which treats void areas inside the wide, open dentin tubules as flat surfaces. The surface damage for the specimens treated with NaOCl or collagenase following acid-etching was much greater than for the control groups treated only with the acid etching, regardless of the concentration of acid used. For this reason, with respect to surface roughness and opening of tubule orifices, only surfaces etched with 35% phosphoric acid and treated with NaOCl or collagenase were observed under the AFM for all application times. AFM showed that a surface etched with 35% phosphoric acid was completely covered by a layer of organic material believed to be the ‘collagen smear layer’ (Fig. 1B). The phase imaging mode of the AFM confirmed that the structure consisted of only soft material, presumably organic. Hydroxyapatite crystals were observed attached to the
263
collagen fibers on the dentin surface treated with 15% phosphoric acid (Fig. 1D). It is possible that these crystals were responsible for the higher roughness of the surface for this group. For the collagenase treated specimens, the AFM showed many collagen fibers exposed in the 1 h specimens. This is consistent with the FE-SEM observations, where the hybrid layer was thick for both bonding systems. However, why the hybrid layers for both bonding systems were thicker than for the control groups is unclear. It is possible that the collagen fibers may not have been digested by 1 h at room temperature (about 20⬚C), but that the protein in the fibers had become swollen during the denaturing process. When the bonding resin was applied to the swollen fibers, the hybrid layer created was slightly thicker than the control. Where the collagenase had more time to digest the collagen fibers, in the 3 and 6 h collagenase specimens, the AFM showed fewer exposed collagen fibers on the surface. This is consistent with FE-SEM observations which showed the hybrid layers for both the 3 and 6 h collagenase-treated specimens were very thin, and almost undetectable at times. The higher roughness for the 3 h collgenase treatment, and subsequent lower roughness for the 6 h collagenase treatment, can be explained in a similar way to the differences observed for the 1 and 2 min NaOCl treatments. It was also noted that the tubule openings for the 6-h collagenase treated specimens seemed to have been covered with collagen fibers. This may have been caused by the effects of the collagenase unravelling and loosening the fibers. This needs to be further investigated. In the current study, as surface roughness increased, a similar increase in the bond strength was observed (Fig. 3). However, no correlation between surface roughness and bond strength was made, since the surfaces for AFM and bond strength specimens were finished differently. For the control group, the Single Bond specimens showed a more uniform hybrid layer than the One Coat Bond specimens. Both systems use HEMA as a hydrophilic monomer to penetrate the moist collagen, however, the other components and bonding procedures are different. These differences may affect the formation of the hybrid layer. The silica filler in One Coat Bond gave the bond a gel consistency, which has been claimed to improve the adhesive strength and reduce polymerization shrinkage. However, it may have an adverse effect on the penetration of the resin into the demineralized dentin. The scrubbing motion during the application of the One Coat Bond adhesive resin may damage parts of the demineralized dentin surface, which probably leads to the non-uniform thickness of the hybrid layer. Several specimens showed that the resin failed to penetrate to the base of demineralization leaving a porous region within the hybrid layer. Perdiga˜o et al. [27] have referred to this region as a ‘submicron hiatus’. The fracture modes showed most specimens exhibited partial adhesive failures between the bonding resin and dentin associated with partial cohesive failure in the
264
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265
bonding resin. The fracture modes observed were similar in nature to other tests using larger surface areas [28,29]. Results from this study suggest that collagenase and NaOCl did not completely remove the collagen bundles on the surface dentin. The other published studies used the SEM for investigating the surface change after NaOCl or collagenase treatment [13–18]. The surfaces must be desiccated and coated with gold or gold–platinum in order to conduct an electron beam, which may have affected the detection of the remaining collagen fibers on the surface. The present study demonstrated that the bonding mechanism between resin and dentin was not dependent upon the thickness of the hybrid layer as stated by previous works [13,14,30]. However, we have shown that bond strengths can be increased by these treatments. The removal of the ‘collagen smear layer’ may be significant in improving the bond strengths. Optimal bond strengths can be obtained even where a very thin layer of collagen fibers exists at the dentin surface, associated with an increase in surface roughness. The bonding mechanism is believed to be dependent on not only the collagen fibers, but also on surface roughness, penetration of the bonding resin into the treated dentin, hydroxyapatite crystals projecting into the collagen matrix, and possible chemical interaction at the resin– dentin interface. In addition, penetration of the resin tags into the dentin tubules and inter-tubular anastomoses are important [14]. Further research on the bonding mechanism to dentin with regard to degree of surface demineralization to establish a good bond is now required.
Acknowledgements The research was supported by the Australian Dental Research Foundation Inc., St Leonards, NSW 2065, Australia. Dr P.G. Hartley acknowledges support from the Advanced Mineral Products Research Centre, an Australian Research Council Special Research Centre. The assistance of Jocelyn L. Carpenter, School of Botany, The University of Melbourne, with the FE-SEM imaging is also greatly appreciated.
References [1] Swift Jr. EJ, Perdiga˜o J, Heymann HO. Bonding to enamel and dentin: a brief history and state of the art. Quintessence Int 1995;26:95–110. [2] Pashley DH, Horner JA, Brewer PD. Interactions of conditioners on the dentin surface. Oper Dent 1992;17(Suppl 5):137–50. [3] Attal JP, Asmussen E, Degrange M. Effects of surface treatment on the free surface energy of dentin. Dent Mater 1994;10:259–64. [4] Nakabayashi N, Kojima K, Masuhara E. The promotion of adhesion by the infiltration of monomers into tooth substrates. J Biomed Mater Res 1982;16:265–73. [5] Van Meerbeek B, Inokoshi S, Braem M, Lambrechts P, Vanherle G. Morphological aspects of the resin–dentin interdiffusion zone with different dentin adhesive systems. J Dent Res 1992;71: 1530–40.
[6] Nakajima M, Sano H, Inai N, Tagami J, Burrow MF, Takatsu T. Resin–dentin interface—SEM observation using a freeze drying technique. In: Shimono M, Maeda T, Suda H, Takahashi K, editors. Proceedings of the International Conference on Dentin/Pulp Complex 1995 and the International Meeting on Clinical Topics of Dentin/Pulp Complex, Tokyo: Quintessence, 1995. p. 352–4. [7] Prati C, Ferrieri P, Galloni C, Mongiorgi R, Davidson CL. Dentine permeability and bond quality as affected by new bonding systems. J Dent 1995;23:217–26. [8] Sano H, Takatsu T, Ciucchi B, Horner JA, Matthews WG, Pashley DH. Nanoleakage: leakage within the hybrid layer. Oper Dent 1995;20:18–25. [9] Sano H, Yoshiyama M, Ebisu S, Burrow MF, Takatsu T, Ciucchi B, et al. Comparative SEM and TEM observations of nanoleakage within the hybrid layer. Oper Dent 1995;20:160–70. [10] Wieliczka DM, Spencer P, Kruger MB. Raman mapping of the dentin/ adhesive interface. Appl Spect 1996;50:1500–4. [11] Wieliczka DM, Kruger MB, Spencer P. Raman imaging of dental adhesive diffusion. Appl Spect 1997;51:1593–6. [12] Pashley DH, Ciucchi B, Sano H, Horner JA. Permeability of dentin to adhesive agents, Vol 24. Tokyo: Quintessence Int., 1993;24:618–31. [13] Gwinnett AJ, Tay FR, Pang KM, Wei SHY. Quantitative contribution of the collagen network in dentin hybridization. Am J Dent 1996;9:140–4. [14] Vargas MA, Cobb DS, Armstrong SR. Resin–dentin shear bond strength and interfacial ultrastructure with and without a hybrid layer. Oper Dent 1997;22:159–66. [15] Wakabayashi Y, Kondou Y, Suzuki K, Yatani H, Yamashita A. Effect of dissolution of collagen on adhesion to dentin. Int J Prosthodont 1994;7:302–6. [16] Hosoda H, Sugizaki H, Nakajima M, Shono T, Tagami J. A study on the mechanism of bonding between resin and dentin. Part 1. Bonding to dentin treated with sodium hypochlorite. Japan J Conserv Dent 1993;36:1054–8. [17] Inai N, Kanemura N, Tagami J, Watanabe LG, Marshall SJ, Marshall GW. Adhesion between collagen depleted dentin and dentin adhesives. Am J Dent 1998;11:123–7. [18] Prati C, Chersoni S, Pashley DH. Effect of removal of surface collagen fibrils on resin-dentin bonding. Dent Mater 1999;15:323–31. [19] Kinney JH, Balooch M, Marshall SJ, Marshall Jr. GW, Weihs TP. Hardness and Young’s modulus of human peritubular and intertubular dentine. Arch Oral Biol 1996;41:9–13. [20] Kinney JH, Balooch M, Marshall SJ, Marshall Jr. GW, Weihs TP. Atomic force microscope measurements of the hardness and elasticity of peritubular and intertubular human dentin. J Biomech Eng 1996;118:133–5. [21] Nakamichi I, Iwaku M, Fusayama T. Bovine teeth as possible substitutes in the adhesion test. J Dent Res 1983;62:1076–81. [22] Fowler CS, Swartz ML, Moore BK, Rhodes BF. Influence of selected variables on adhesion testing. Dent Mater 1992;8:265–9. [23] Barkmeier WW, Erickson RL. Shear bond strength of composite to enamel and dentin using Scotchbond Multi-Purpose. Am J Dent 1994;7:175–9. [24] Kuboki Y, Mechanic GL. Comparative molecular distribution of cross-link in bone and dentin collagen. Structure–function relationships. Calcif Tissue Int 1982;34:306–8. [25] Stetler-Stevenson WG, Veis A. Type I collagen shows a specific binding affinity for bovine dentin phosphophoryn. Calcif Tissue Int 1986;38:135–41. [26] Garbarsch C, Matthiessen ME, Olsen BE, Moe D, Kirkeby S. Immunohistochemistry of the intercellular matrix components and the epithelio-mesenchymal junction of the human tooth germ. Histochem J 1994;26:110–8. [27] Perdiga˜o J, Lambrechts P, Van Meerbeek B, Tome´ AR, Vanherle G, Lopes A. Morphological field emission-SEM study of the effect of six phosphoric acid etching agents. Dent Mater 1996;12:262–71.
S. Phrukkanon et al. / Dental Materials 16 (2000) 255–265 [28] Sano H, Shono T, Sonoda H, Takatsu T, Ciucchi B, Carvalho R, et al. Relationship between surface area for adhesion and tensile bond strength—evaluation of a micro-tensile bond test. Dent Mater 1994;10:236–40. [29] Phrukkanon S, Burrow MF, Tyas MJ. Effect of cross-sectional surface
265
area on bond strengths between resin and dentin. Dent Mater 1998;14:120–8. [30] Burrow MF, Takakura H, Nakajima M, Inai N, Tagami J, Takatsu T. The influence of age and depth of dentin on bonding. Dent Mater 1994;10:241–6.