Bond strength of restorative materials to hydroxyapatite inserts and dimensional changes of insert-containing restorations during polymerization

Highlights

  • HAP inserts reduced polymerization shrinkage of the insert-containing restorations.

  • ‘Total-etch’ approach is recommended for adhesive bonding of composites to HAP.

  • Micromechanical interlocking is the major adhesion mechanism.

  • A universal composite is recommended to bond HAP inserts to dentin.

Abstract

Objective

To determine the shear bond strength (SBS) between synthetic controlled porous hydroxyapatite (HAP) inserts and restorative materials and dimensional changes of insert-containing restorations during curing.

Methods

Cylinder-shaped HAP inserts (4 mm in diameter, 1.6 mm thick) were cemented in dentin discs (5 mm × 1.6 mm), cut mid-coronally from human third molars, using one of the following materials: universal microhybrid composite Filtek Z250, flowable composite Filtek Ultimate or glass-ionomer Vitrebond (all 3M ESPE). SBS of the same materials to HAP inserts was tested in a universal testing machine. Three-dimensional digital image correlation system Aramis (GOM) was used to measure strains and displacements. Data were statistically analyzed using one-way ANOVA with Tukey’s post-test ( α = 0.05).

Results

SBS of restorative materials to HAP inserts ranged between 12.2 ± 2.1 MPa (Filtek Z250) and 0.7 ± 0.4 MPa (Filtek Z250 without an adhesive). The ‘total-etch’ approach of adhesive application significantly increased SBS of both Filtek Z250 (12.2 ± 2.1 MPa) and Filtek Ultimate flowable (9.5 ± 2.5 MPa) compared to the ‘self-etch’ approach (8.2 ± 1.6 MPa and 4.4 ± 0.9 MPa, respectively) ( p < 0.05). HAP inserts reduced polymerization shrinkage to below 0.5% as well as displacements in the central region of the restorations. Peripheral shrinkage of restorative materials was similar with and without HAP inserts as were displacements of Filtek Z250 and Vitrebond.

Significance

Replacing major part of dentin clinically, especially in large cavities, HAP inserts may shorten clinical working time, improve dimensional stability of the restoration by reducing central shrinkage and displacements and provide adhesive bonding to universal composites following a ‘total-etch’ approach.

Introduction

Attempts to overcome the adverse properties of resin-based dental materials, polymerization shrinkage and the associated stress can either be material-related (e.g. low shrinkage monomers or dental inserts ), technique-related (incremental layering versus bulk technique ) and light source-related (‘soft-start’ or ‘pulse’ light-curing regimes ).

The idea behind the use of dental inserts, prefabricated parts of a restoration similar to inlays, is to reduce the amount of unset material in the cavity. Early inserts were megafillers and later were ceramic, manufactured in different sizes for different cavities . Inserts were produced mostly from IPS Empress ceramics (Cerena), β-quartz ceramics (β-quartz glass ceramic inserts) and leucite-reinforced ceramic (SonicSys and Cerafil inserts). To the best of our knowledge, hydroxyapatite (HAP)-based inserts have not been tested in restorative dentistry.

HAP may be more beneficial than ceramics due to similar mechanical properties to dentin, e.g. HAP fracture toughness of 1.30 ± 0.015 MPa m 1/2 falls in the range of dentin fracture toughness values (1.13–2.02 MPa m 1/2 ) whereas the trend is to increase fracture toughness of ceramics . Another advantage of synthetic HAP over ceramics is the ability to chemically interact with functional monomer groups in adhesives .

Dimensional changes of insert-containing restorations were tested using dye penetration tests , microscopic evaluation of microgaps , strain gauges and mercury bath . The use of β-quartz ceramic inserts instead of 31–38 vol% of composite reduced polymerization shrinkage . Clinical investigation of β-quartz inserts has shown contradicting results. Kiremitci et al. have shown satisfactory and improved clinical performance of restorations according to the USHPS criteria . An 8-year clinical study confirmed the beneficial effect of these inserts on composite restoration properties .

Inserts are clearly advantageous in large cavities due to a reduction in the number of clinical steps. In general, it has been shown that ceramic inserts do have some positive effects on marginal adaptation but further improvements are required. A possible negative effect of increasing the number of interfaces between materials with different elastic moduli needs to be investigated. Dentin exhibits lower elastic modulus (2 to 15 GPa) than most contemporary composites (10 to 24 GPa) . Although the elastic modulus of synthetic HAP may vary significantly , modulating this property to be comparable to dentin may in fact have a positive rather than a negative effect on the restoration despite the increased number of interfaces.

Digital image correlation is a non-contact method for measuring volumetric shrinkage, based on tracking the position of surface markers by specialist software before and after polymerization and calculating strains and displacements. Earlier studies employed one-camera systems which provide information only for in-plane movements. Micro-computed tomography was also used for 3D measurements . However, radiopaque markers required in micro-computed tomography often exceed filler size in dental composites and may disturb polymerization. Two-camera systems for obtaining actual 3D data were seldom used . In these previous studies, no dental bonding systems were used in order to allow unrestricted shrinkage.

The aim of this study was to measure the shear bond strength (SBS) between HAP inserts and various restorative materials as well as the effect of HAP inserts on dimensional changes of restorations during material setting/polymerization. The null hypotheses were: (1) there is no difference in SBS to HAP inserts between restorative materials and (2) there is no difference in shrinkage strains and displacements between groups with and without HAP inserts.

Materials and methods

Insert preparation and properties

Starting HAP powder, composed of spherically agglomerated nanosized rod-like particles, was obtained using the modified hydrothermal synthesis described earlier , starting from a solution of CaCl 2 ·2H 2 O, Na 2 H 2 EDTA·2H 2 O, NaH 2 PO 4 ·2H 2 O and urea. The synthesis parameters and concentration of precursors were the same as described previously . The powder was further isostatically pressed in a stainless steel mold at 400 MPa into disc compacts, 4 mm in diameter and 1.6 mm thick, and additionally sintered at 1200 °C over 2 h at a heating rate of 20 °C/min. Processed controlled porous insert material (with calculated mean pore diameter of 0.74 μm, observed at the fractured surface of inserts), consisted of HAP as the dominant phase, with the presence of lower amounts of α- and β-TCP as a second crystalline phase . Insert material was characterized with the density value of 2.64 ± 0.02 g/cm 3 , fracture toughness of 1.30 ± 0.01 MPa m 1/2 and Vickers hardness of 3.05 ± 0.05 GPa.

Sample preparation and shear bond strength measurements

Thirty-five HAP inserts ( N = 5 per group) were prepared for SBS testing with materials listed in Table 1 . HAP inserts were used as fabricated without any further surface treatment providing standardized surface characteristics prior to the adhesive procedure. The following test groups were prepared:

  • Group 1—Z250_SBU_TE (universal composite Z250 with Single Bond Universal adhesive following the ‘total-etch’ protocol);

  • Group 2—Flow_SBU_TE (Filtek Ultimate Flowable composite with Single Bond Universal adhesive following the ‘total-etch’ protocol);

  • Group 3—Z250_SBU_SE (universal composite Z250 with Single Bond Universal adhesive following the ‘self-etch’ protocol);

  • Group 4—Flow_SBU_SE (Filtek Ultimate Flowable composite with Single Bond Universal adhesive following the ‘self-etch’ protocol);

  • Group 5—Vitrebond (resin-modified glass-ionomer cement without any surface treatment);

  • Group 6—Z250 (universal composite Z250 without the adhesive) and

  • Group 7—Flow (Filtek Ultimate Flowable composite without the adhesive).

Table 1
Materials used in the study.
Material Code Type Composition
Filtek™ Ultimate Flowable Flow Flowable composite Substituted dimethacrylate, BisGMA, TEGDMA, functionalised dimethacrylate polymer, EDMAB, silane treated ceramic (50–60 wt%), silane treated silica (5–10 wt%), ytterbium trifluoride (<5 wt%), TiO 2 , benzotriazol, diphenyliodonium hexafluorophosphate
Vitrebond™ Vitrebond Glass ionomer cement, light cure liner/base Powder : fluoro alumino silicate glass: SiO 2 , AlF 3 , ZnO, SrO, cryolite, NH 4 F, MgO, P 2 O 5
Liquid : modified polyacrilic acid with pendant methacrylate groups, HEMA, water, photoinitiator. P / L = 4/1wt
Filtek™ Z250 Universal Restorative Z250 Conventional composite, microhybrid BisEMA6, UDMA, BisGMA, TEGDMA, silane-treated ceramic (75–85 wt%), benzotriazol, EDMAB
Single Bond Universal Adhesive SBU Adhesive, ‘total-etch’ or ‘self-etch’ protocol BisGMA, HEMA, DMDMA, ethanol, water, silica filler, 2-propenoic acid, 2-methyl-, reaction products with 1,10-decanediol and phosphorous oxide (P 2 O 5 ), copolymer of acrylic itaconic acid, camphorquinone, EDMAB, toluene
All materials from 3M ESPE, St. Paul, MN, USA.
Abbreviations: BisGMA – Bisphenol A Diglycidyl Ether Dimethacrylate; TEGDMA – Triethylene Glycol Dimethacrylate; EDMAB – Ethyl 4-dimethyl Aminobenzoate; HEMA – Hydroxyethyl methacrylate; BisEMA6 – Bisphenol A Polyethylene GlycolDiether Dimethacrylate; UDMA – Diurethane Dimethacrylate; DMDMA – Decamethylene dimethacrylate.

In Groups 1 and 2, HAP inserts were first etched with 37% phosphoric acid for 15 s, copiously rinsed with water and mildly air-dried. Each sample was prepared by placing an insert into a custom-made mold, 4 mm in diameter and 3.6 mm deep. The insert occupied 1.6 mm at the bottom of the mold leaving the upper 2 mm free for the restorative material. The diameter of the inserts fitted the mold. In Groups 1–4, the adhesive was rubbed into the insert surface for 20 s, air-dried and light-cured for 20 s using an LED light-curing unit (LEDition, Ivoclar Vivadent, Schaan, Liechtenstein) at an intensity of 800 mW/cm 2 . The respective composite material was applied in one 2-mm increment by filling the entire mold and light-cured for 40 s using the same light source. In Group 5, Vitrebond was mixed according to the manufacturer’s instructions, applied directly to the untreated surface of HAP inserts and light-cured for 20 s.

SBS was measured in a universal testing machine (Force Gauge PCE-FM200, Southampton, United Kingdom), using a knife-edge shearing blade, within 30 min of sample preparation. Vertical force was applied to the dental material 1 mm from the bonded surface at 1 mm/min speed until fracture. SBS ( τ ) in MPa was calculated using the maximum force reached ( F in Newton) and bonded surface area ( A in mm 2 ) according to the following equation:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='τ=FA.’>τ=FA.τ=FA.
τ = F A .

The type of fracture was analyzed under a stereomicroscope and classified as: (1) ‘adhesive’ (fracture at the bonded surface between the insert and material), (2) ‘cohesive’ (fracture in the insert or material) or (3) ‘mixed’ (fracture at the bonded surface extending into the insert and/or restorative material).

Sample preparation and strain and displacement measurements

The study was approved by the Ethics Committee of the School of Dental Medicine, University of Belgrade. Thirty intact human third molars extracted for orthodontic reasons were embedded in gypsum up to the enamel–cementum junction. The occlusal third was cut off perpendicular to the long axis using a diamond saw (Buehler, Lake Bluff, IL, USA). In the center of the exposed flat dentin, a 5-mm wide circular cavity was prepared using a high-speed hand-piece. The teeth were sectioned 1.6 mm below the flat dentin surface to obtain dentin discs with a central hole, 1 mm wider than HAP inserts.

Six groups, 5 samples each, were prepared using the materials listed in Table 1 . The inner surface of the central hole was treated with Single Bond Universal adhesive according to the SE protocol prior to the application of Flow or Z250. The adhesive was cured for 20 s using the LEdition light-curing unit. No dentin treatment was performed for Vitrebond as per manufacturer’s instructions. Dentin discs were placed on Mylar strips and completely filled with material in groups without HAP inserts. In groups with HAP inserts, dentin discs were filled with material and then an insert was placed in the center of the disc, extruding excess material which was removed with a ‘plastic’ filling instrument. The outer surface of HAP inserts was not treated and was not adhesively bonded to the restorative material.

The surface of each sample facing the two-camera system (Aramis 2M, GOM, Braunschweig, Germany) was sprayed with a fine layer of black and white paint (Kenda Color Acrilico, Kenda Farben) to produce irregularly shaped speckles for detection and analysis by the specialist software (Aramis). The unsprayed bottom surface, opposite the one facing the cameras, was light-cured for 40 s. Standardized conditions were created by mounting each sample and the light-curing unit in fixed holders, maintaining 1 mm distance between the irradiated surface and the light tip.

Images were taken before and after polymerization and von Mises strain and displacements in the Z -axis (directed toward center of the restoration) were determined according to the previously mentioned methodology . Briefly, von Mises strain is defined as an index gained from the combination of principal stresses at any given point to determine at which points stress occurring on the X -, Y -, and Z -axis will cause failure.

For the purpose of statistical analysis, values for von Mises strains and Z -axis displacements were plotted as a function of distance along a horizontal and vertical line through the center of each sample (Section 0 and 1) and along the circumferential, peripheral segment (Section 2). A representative digital image of restorations with and without HAP inserts is shown in Fig. 1 . Central parts of each sample (Section 0 and 1) were compared separately from the circumferential, peripheral zone (Section 2). In all groups, i.e. with and without HAP inserts, Section 2 corresponded only to the restorative material.

Fig. 1
Representative digital images of speckled samples with the superimposed analysis of von Mises strain (a and b) and displacements (c and d) in a conventional restoration (a and c) and a HAP-containing restoration (b and d). Section 0—vertical, Section 1—horizontal and Section 2—circumferential.

Scanning electron microscopy (SEM)

SEM was used to analyze the surface of HAP inserts, representative fractured fragments following SBS testing as well as dentin discs used for strain and displacement measurements. All samples were mounted on aluminium stubs, coated with Au–Pd alloy using a sputter coater (POLARON SC502, Fisions Instruments, Ipswich, UK) and analyzed using TESCAN FE-SEM (Mira 3 XMU, TESCAN a.s., Brno–Czech Republic), operating at 10 kV.

Statistical analysis

Statistical analysis was performed in Minitab 16 (Minitab Inc, State College, PA, USA). The data were analyzed using one-way ANOVA with Tukey’s post-hoc test for multiple comparisons at the level of significance α = 0.05. Where necessary, data transformation was done using the log or sqrt functions to stabilize the variance, which is a necessary requirement for parametric testing.

Materials and methods

Insert preparation and properties

Starting HAP powder, composed of spherically agglomerated nanosized rod-like particles, was obtained using the modified hydrothermal synthesis described earlier , starting from a solution of CaCl 2 ·2H 2 O, Na 2 H 2 EDTA·2H 2 O, NaH 2 PO 4 ·2H 2 O and urea. The synthesis parameters and concentration of precursors were the same as described previously . The powder was further isostatically pressed in a stainless steel mold at 400 MPa into disc compacts, 4 mm in diameter and 1.6 mm thick, and additionally sintered at 1200 °C over 2 h at a heating rate of 20 °C/min. Processed controlled porous insert material (with calculated mean pore diameter of 0.74 μm, observed at the fractured surface of inserts), consisted of HAP as the dominant phase, with the presence of lower amounts of α- and β-TCP as a second crystalline phase . Insert material was characterized with the density value of 2.64 ± 0.02 g/cm 3 , fracture toughness of 1.30 ± 0.01 MPa m 1/2 and Vickers hardness of 3.05 ± 0.05 GPa.

Sample preparation and shear bond strength measurements

Thirty-five HAP inserts ( N = 5 per group) were prepared for SBS testing with materials listed in Table 1 . HAP inserts were used as fabricated without any further surface treatment providing standardized surface characteristics prior to the adhesive procedure. The following test groups were prepared:

  • Group 1—Z250_SBU_TE (universal composite Z250 with Single Bond Universal adhesive following the ‘total-etch’ protocol);

  • Group 2—Flow_SBU_TE (Filtek Ultimate Flowable composite with Single Bond Universal adhesive following the ‘total-etch’ protocol);

  • Group 3—Z250_SBU_SE (universal composite Z250 with Single Bond Universal adhesive following the ‘self-etch’ protocol);

  • Group 4—Flow_SBU_SE (Filtek Ultimate Flowable composite with Single Bond Universal adhesive following the ‘self-etch’ protocol);

  • Group 5—Vitrebond (resin-modified glass-ionomer cement without any surface treatment);

  • Group 6—Z250 (universal composite Z250 without the adhesive) and

  • Group 7—Flow (Filtek Ultimate Flowable composite without the adhesive).

Table 1
Materials used in the study.
Material Code Type Composition
Filtek™ Ultimate Flowable Flow Flowable composite Substituted dimethacrylate, BisGMA, TEGDMA, functionalised dimethacrylate polymer, EDMAB, silane treated ceramic (50–60 wt%), silane treated silica (5–10 wt%), ytterbium trifluoride (<5 wt%), TiO 2 , benzotriazol, diphenyliodonium hexafluorophosphate
Vitrebond™ Vitrebond Glass ionomer cement, light cure liner/base Powder : fluoro alumino silicate glass: SiO 2 , AlF 3 , ZnO, SrO, cryolite, NH 4 F, MgO, P 2 O 5
Liquid : modified polyacrilic acid with pendant methacrylate groups, HEMA, water, photoinitiator. P / L = 4/1wt
Filtek™ Z250 Universal Restorative Z250 Conventional composite, microhybrid BisEMA6, UDMA, BisGMA, TEGDMA, silane-treated ceramic (75–85 wt%), benzotriazol, EDMAB
Single Bond Universal Adhesive SBU Adhesive, ‘total-etch’ or ‘self-etch’ protocol BisGMA, HEMA, DMDMA, ethanol, water, silica filler, 2-propenoic acid, 2-methyl-, reaction products with 1,10-decanediol and phosphorous oxide (P 2 O 5 ), copolymer of acrylic itaconic acid, camphorquinone, EDMAB, toluene
All materials from 3M ESPE, St. Paul, MN, USA.
Abbreviations: BisGMA – Bisphenol A Diglycidyl Ether Dimethacrylate; TEGDMA – Triethylene Glycol Dimethacrylate; EDMAB – Ethyl 4-dimethyl Aminobenzoate; HEMA – Hydroxyethyl methacrylate; BisEMA6 – Bisphenol A Polyethylene GlycolDiether Dimethacrylate; UDMA – Diurethane Dimethacrylate; DMDMA – Decamethylene dimethacrylate.

In Groups 1 and 2, HAP inserts were first etched with 37% phosphoric acid for 15 s, copiously rinsed with water and mildly air-dried. Each sample was prepared by placing an insert into a custom-made mold, 4 mm in diameter and 3.6 mm deep. The insert occupied 1.6 mm at the bottom of the mold leaving the upper 2 mm free for the restorative material. The diameter of the inserts fitted the mold. In Groups 1–4, the adhesive was rubbed into the insert surface for 20 s, air-dried and light-cured for 20 s using an LED light-curing unit (LEDition, Ivoclar Vivadent, Schaan, Liechtenstein) at an intensity of 800 mW/cm 2 . The respective composite material was applied in one 2-mm increment by filling the entire mold and light-cured for 40 s using the same light source. In Group 5, Vitrebond was mixed according to the manufacturer’s instructions, applied directly to the untreated surface of HAP inserts and light-cured for 20 s.

SBS was measured in a universal testing machine (Force Gauge PCE-FM200, Southampton, United Kingdom), using a knife-edge shearing blade, within 30 min of sample preparation. Vertical force was applied to the dental material 1 mm from the bonded surface at 1 mm/min speed until fracture. SBS ( τ ) in MPa was calculated using the maximum force reached ( F in Newton) and bonded surface area ( A in mm 2 ) according to the following equation:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-2-Frame class=MathJax style="POSITION: relative" data-mathml='τ=FA.’>τ=FA.τ=FA.
τ = F A .

The type of fracture was analyzed under a stereomicroscope and classified as: (1) ‘adhesive’ (fracture at the bonded surface between the insert and material), (2) ‘cohesive’ (fracture in the insert or material) or (3) ‘mixed’ (fracture at the bonded surface extending into the insert and/or restorative material).

Sample preparation and strain and displacement measurements

The study was approved by the Ethics Committee of the School of Dental Medicine, University of Belgrade. Thirty intact human third molars extracted for orthodontic reasons were embedded in gypsum up to the enamel–cementum junction. The occlusal third was cut off perpendicular to the long axis using a diamond saw (Buehler, Lake Bluff, IL, USA). In the center of the exposed flat dentin, a 5-mm wide circular cavity was prepared using a high-speed hand-piece. The teeth were sectioned 1.6 mm below the flat dentin surface to obtain dentin discs with a central hole, 1 mm wider than HAP inserts.

Six groups, 5 samples each, were prepared using the materials listed in Table 1 . The inner surface of the central hole was treated with Single Bond Universal adhesive according to the SE protocol prior to the application of Flow or Z250. The adhesive was cured for 20 s using the LEdition light-curing unit. No dentin treatment was performed for Vitrebond as per manufacturer’s instructions. Dentin discs were placed on Mylar strips and completely filled with material in groups without HAP inserts. In groups with HAP inserts, dentin discs were filled with material and then an insert was placed in the center of the disc, extruding excess material which was removed with a ‘plastic’ filling instrument. The outer surface of HAP inserts was not treated and was not adhesively bonded to the restorative material.

The surface of each sample facing the two-camera system (Aramis 2M, GOM, Braunschweig, Germany) was sprayed with a fine layer of black and white paint (Kenda Color Acrilico, Kenda Farben) to produce irregularly shaped speckles for detection and analysis by the specialist software (Aramis). The unsprayed bottom surface, opposite the one facing the cameras, was light-cured for 40 s. Standardized conditions were created by mounting each sample and the light-curing unit in fixed holders, maintaining 1 mm distance between the irradiated surface and the light tip.

Images were taken before and after polymerization and von Mises strain and displacements in the Z -axis (directed toward center of the restoration) were determined according to the previously mentioned methodology . Briefly, von Mises strain is defined as an index gained from the combination of principal stresses at any given point to determine at which points stress occurring on the X -, Y -, and Z -axis will cause failure.

For the purpose of statistical analysis, values for von Mises strains and Z -axis displacements were plotted as a function of distance along a horizontal and vertical line through the center of each sample (Section 0 and 1) and along the circumferential, peripheral segment (Section 2). A representative digital image of restorations with and without HAP inserts is shown in Fig. 1 . Central parts of each sample (Section 0 and 1) were compared separately from the circumferential, peripheral zone (Section 2). In all groups, i.e. with and without HAP inserts, Section 2 corresponded only to the restorative material.

Fig. 1
Representative digital images of speckled samples with the superimposed analysis of von Mises strain (a and b) and displacements (c and d) in a conventional restoration (a and c) and a HAP-containing restoration (b and d). Section 0—vertical, Section 1—horizontal and Section 2—circumferential.

Scanning electron microscopy (SEM)

SEM was used to analyze the surface of HAP inserts, representative fractured fragments following SBS testing as well as dentin discs used for strain and displacement measurements. All samples were mounted on aluminium stubs, coated with Au–Pd alloy using a sputter coater (POLARON SC502, Fisions Instruments, Ipswich, UK) and analyzed using TESCAN FE-SEM (Mira 3 XMU, TESCAN a.s., Brno–Czech Republic), operating at 10 kV.

Statistical analysis

Statistical analysis was performed in Minitab 16 (Minitab Inc, State College, PA, USA). The data were analyzed using one-way ANOVA with Tukey’s post-hoc test for multiple comparisons at the level of significance α = 0.05. Where necessary, data transformation was done using the log or sqrt functions to stabilize the variance, which is a necessary requirement for parametric testing.

Results

Universal composite Z250 bonded to HAP inserts following the ‘total-etch’ protocol (Z250_SBU_TE) showed significantly higher SBS to HAP inserts than other groups ( p < 0.05) except the flowable composite group bonded following the ‘total-etch’ protocol (Flow_SBU_TE) ( p > 0.05). Both universal and flowable composite showed increased SBS following the ‘total-etch’ protocol compared to the ‘self-etch’ protocol ( p < 0.05) ( Fig. 2 ). Fig. 3 illustrates the differences between intact and acid-etched surfaces of HAP inserts.

Fig. 2
Shear bond strength of HAP inserts to dental materials. The same letters indicate no statistically significant difference between groups ( p > 0.05).

Fig. 3
The surface of HAP inserts, (a) without any treatment and (b) etched with 37% phosphoric acid for 15 s. Intact HAP inserts are characterized with flat surface and scarce pores up to 1 μm in diameter. Acid etching partially demineralizes micron-sized HAP grains leaving rough surface with numerous pores extending to about 3–4 μm in diameter. Insert in picture (b) shows a cross-section of an etched insert (white arrow shows etched surface). The effect of acid etching was only superficial.

All samples fractured adhesively in groups without an adhesive whereas only ‘mixed’ fractures occurred in the universal composite group adhesively bonded to HAP inserts following the ‘total-etch’ protocol ( Table 2 ). Other groups showed similar incidence of ‘adhesive’ and ‘mixed’ fractures. Fractured samples showing representative ‘adhesive’ and ‘mixed’ fractures are presented in Fig. 4 .

Table 2
Fracture types following shear bond strength testing.
Group Type of fracture Pre-test failures
Adhesive Cohesive Mixed
Z250_SBU_TE 0 0 5 0
Flow_SBU_TE 2 0 3 0
Z250_SBU_SE 2 0 3 0
Flow_SBU_SE 3 0 2 0
Vitrebond 5 0 0 0
Z250 5 0 0 0
Flow 5 0 0 0

Fig. 4
Representative SEM photomicrographs of the fracture modes, (a) ‘Adhesive’ fracture between a HAP insert and Z250 composite in the Z250_SBU_SE group. Arrows indicate resin remnants attached to the insert surface. (b) ‘Mixed’ fracture between a HAP insert and Vitrebond. The severed part of the insert (right) exposes the internal porous structure of HAP inserts.

HAP inserts significantly decreased von Misses strain (i.e. volumetric shrinkage) in the central region of the restorations in all groups compared to the corresponding control groups without HAP inserts ( p < 0.05). There were no strain differences centrally in groups with HAP inserts irrespective of the restorative material ( p > 0.05). Peripherally, von Mises strain was significantly higher for Flow_HAP (3.4 ± 2.3%) and Vitrebond_HAP (3.7 ± 1.5%) than Z250_HAP (1.3 ± 0.4%) ( p < 0.05). Mean values of von Mises strain were lower in the flowable composite group (2.1 ± 1.6%) and Vitrebond group (2.4 ± 1.3%) without HAP inserts compared to their counterparts with HAP inserts, respectively, but no statistically significant difference was detected ( Fig. 5 ).

Fig. 5
Von Mises strain in the central (section 0 + 1) and peripheral zones (section 2). The same upper case letters indicate no statistically significant differences for sections 0 and 1 and the same lower case letters indicate no differences for section 2 ( p > 0.05).

HAP inserts significantly decreased central displacements in all groups compared to the non-HAP-containing controls ( p < 0.05). Peripheral displacements were comparable in Vitrebond and Z250 groups with and without HAP inserts ( p < 0.05) ( Fig. 6 ). Greater displacement in Section 2 found for the flowable composite without HAP inserts compared to the situation with HAP inserts indicates a change of the displacement vector from the XY axis to the Z axis when HAP inserts were used. Fig. 7 shows the interfaces between materials and HAP inserts as well as materials and dentin in restorations with and without HAP inserts.

Fig. 6
Displacement in microns of central and peripheral zones of restorations. The same upper case letters indicate no statistically significant differences for sections 0 and 1 and the same lower case letters indicate no differences between groups for section 2 ( p > 0.05).

Fig. 7
Representative SEM images of the samples used for strain and displacement measurements. No microgaps or cracks were observed at the interface of all tested materials with either HAP inserts or dentin. The presence of HAP inserts did not affect the gap-free interfaces between the tested materials and dentin. The thickness of the adhesive layer varied between approximately 5 μm and 15 μm showing intimate contacts with both composites and dentin. HAP inserts presented irregular rough surfaces similar to Vitrebond. Co—universal or flowable composite; A—adhesive; HAP—HAP insert; Vb—Vitrebond, D-dentin.

Discussion

The tested null hypotheses were rejected as significant differences in SBS were found as well as the differences in volumetric shrinkage and displacements between restorations with and without HAP inserts.

In the present study a macro shear test set-up was chosen to allow bond strength testing to be performed without any cutting, rinsing, gluing or other sample pre-treatment which might affect HAP inserts. The test is less procedure sensitive than micro test set-ups albeit the results tend to show a relatively large coefficient of variation . The present results showed no outliers and comparable variances between the groups and, thus, may be considered an acceptable indicator of the quality of bond strength between HAP inserts and various materials. No data were found in the literature on the bond strength values between ceramic inserts and restorative materials.

No water storage was done prior to bond strength testing. In this first experiment on HAP inserts it was important to avoid any changes in HAP composition and/or properties that might occur during storage. Furthermore, in a clinical situation HAP inserts would be completely enclosed by restorative material and would not be directly exposed to either saliva or tubular liquid. Nevertheless, the effect of storage on the composition and properties of HAP inserts as well as the bond strength to various materials will be tested in the future work.

Acid etching of HAP inserts prior to adhesive application increased SBS by 50% in universal composite Z250 and 100% in Filtek Ultimate Flowable compared to the ‘self-etch’ approach. Higher SBS following the ‘total-etch’ approach may be attributed to superficial HAP demineralization similar to acid etching of natural tooth tissue. Surface roughness and a number of open pores increased resembling dentinal tubules. Micromechanical interlocking plays a major role in adhesion of composites to HAP inserts. Additional chemical bonding to HAP, already proven for Single Bond Universal via the interaction of 10-MDP monomer and Ca ++ from HAP , may be considered a secondary adhesion mechanism. SEM analysis of the fractured surfaces revealed that Single Bond Universal applied to HAP inserts following the ‘self-etch’ approach also increased surface roughness and opened additional pores but to a smaller extent than phosphoric acid etching. Therefore, interaction of functional monomers and HAP inserts during the ‘self-etch’ adhesive application was not sufficient to reach the SBS values obtained with the ‘total-etch’ approach.

Vitrebond exhibited the weakest bond strength to HAP inserts corroborating previous studies that found more efficient bonding of composites than glass-ionomers to tooth substrates . Vitrebond group was included since HAP inserts are intended as dentin restoratives, especially in large cavities. Glass-ionomers are also indicated as base materials under composites in deep cavities so the intention was to test the glass-ionomer-HAP combination as well.

Two cameras allowed 3D measurements of dimensional changes during polymerization including out-of-plane strains and displacements. With one camera systems in-plane displacements are measured and out-of-plane are assumed based on material properties . In the current set-up, out-of-plane Z -axis displacements could be actually measured by the two-camera system and used to calculate volumetric shrinkage. Furthermore, the use of human dentin with the applied adhesive system allows more clinically relevant results than material testing in stainless steel, Teflon or similar molds .

HAP inserts significantly reduced the amount of polymerizable, thus shrinkable, restorative material. In the current model, the volume of the restorative material in non-insert-containing groups was 30.4 mm 3 , nearly three times as much as in the insert-containing groups (11.3 mm 3 ). The present results showed somewhat higher mean shrinkage of the tested flowable composite and glass-ionomer in the insert-containing groups although the differences were not statistically significant. However, even the same or similar percentage of volumetric shrinkage of substantially different volumes of material are actually different in terms of the amount of shrinking material.

Volumetric shrinkage and the imposed restrictions by the bonded surfaces have been shown to affect the associated shrinkage stress responsible for the loss of marginal integrity of adhesively bonded restorations . In the present study, cylindrically-shaped cavities had two unbonded surfaces required for photographing (top) and light-curing (bottom). A limitation of the study is that two exposed surfaces are required, one for the black-and-white stochastic pattern and the other for light curing. In this way, it is impossible to actually measure shrinkage in Class I cavity, i.e. the clinical situation with the most unfavorable C -factor value.

C ’ factors in the current experimental groups with and without HAP inserts as well as the corresponding clinical Class I cavities can be calculated using the following equations:

<SPAN role=presentation tabIndex=0 id=MathJax-Element-3-Frame class=MathJax style="POSITION: relative" data-mathml='‘C’factorwithoutinsert,experimentalcase=2r1πh2r12π=0.64,’>Cfactorwithoutinsert,experimentalcase=2r1πh2r21π=0.64,‘C’factorwithoutinsert,experimentalcase=2r1πh2r12π=0.64,
‘ C ’ factor without insert, experimental case = 2 r 1 π h 2 r 1 2 π = 0.64 ,
Only gold members can continue reading. Log In or Register to continue

Nov 23, 2017 | Posted by in Dental Materials | Comments Off on Bond strength of restorative materials to hydroxyapatite inserts and dimensional changes of insert-containing restorations during polymerization
Premium Wordpress Themes by UFO Themes