The aim of this study was to evaluate the fatigue limits of the dentin–composite interfaces established either with an etch-and-rinse or an one-step self-etch adhesive systems under tensile and bending configurations.
Flat specimens (1.2 mm × 5 mm × 35 mm) were prepared using a plexiglass mold where dentin sections from human third molars were bonded to a resin composite, exhibiting the interface centrally located. Syntac Classic and G-Bond were used as adhesives and applied according to the manufacturer’s instructions. The fluorochrome Rhodamine B was added to the adhesives to allow for fractographic evaluation. Tensile strength was measured in an universal testing machine and the bending strength ( n = 15) in a Flex machine (Flex, University of Antwerp, Belgium), respectively. Tensile (TFL) and bending fatigue limits (BFL) ( n = 25) were determined under wet conditions for 10 4 cycles following a staircase approach. Interface morphology and fracture mechanisms were observed using light, confocal laser scanning and scanning electron microscopy. Statistical analysis was performed using three-way ANOVA (mod LSD test, p < 0.05).
Tensile and bending characteristic strengths at 63.2% failure probability for Syntac were 23.8 MPa and 71.5 MPa, and 24.7 MPa and 72.3 MPa for G-Bond, respectively. Regarding the applied methods, no significant differences were detected between adhesives. However, fatigue limits for G-Bond (TFL = 5.9 MPa; BFL = 36.2 MPa) were significantly reduced when compared to Syntac (TFL = 12.6 MPa; BFL = 49.7 MPa). Fracture modes of Syntac were generally of adhesive nature, between the adhesive resin and dentin, while G-Bond showed fracture planes involving the adhesive–dentin interface and the adhesive resin.
Cyclic loading under tensile and bending configurations led to a significant strength degradation, with a more pronounced fatigue limit decrease for G-Bond. The greater decrease in fracture strength was observed in the tensile configuration.
Recently, a great discussion is being raised on the validity of bond strength tests and their true power of inference to the clinical scenario as a predictive tool for bond durability. Strength measurements are commonly made under static loads to create a linear stress build-up within a material or interface until failure. Such a loading scenario would only take place clinically during polymerization shrinkage, in which the bond created between the adhesive system and the substrate is stressed as the restorative material contracts . The adhesion can then withstand these shrinkage forces or not, leading to a sealed interface or an early bond failure.
Bond failure, however, may also happen later in the in the lifetime of a restoration as a result of stresses accumulated overtime. Such failure mode is termed fatigue failure and occurs via progressive localized damage under cyclic loads. Limited investigations have dealt with dentin bond fatigue resistance until now, whilst initial bond strength under static loads has been extensively researched. This contrast reflects the current insubstantial comprehension of fatigue behavior of dentin–composite interfaces and calls for further insights on the issue in order to bring in vitro bond durability tests closer to clinical relevance. Bond strength assessments would therefore gain increased applicability if tested under protocols capable of mimicking oral mastication effects to determine the stress fatigue limits of a bonding interface. An investigation on the viscoelastic behavior of dental adhesives as thin films has recently warned on the misleading effect of fracture mechanics under static loads for studying their time-dependent mechanical changes . Long-term exposure of viscoelastic polymer materials to loads under its actual yield strength leads to strain accumulation and sudden failure following a fracture mechanics that differs from the resembling brittle fracture patterns resulting from static-loading strength tests . Also, molecular weight, cross-linking density and degree of conversion drive plasticization effects in methacrylic polymers in wet environments , which in turn affect creep compliance and thus time-dependent responses to loading regimens.
The high water affinity of current adhesive systems have shown to bring negative consequences to interfacial sealing and to dentin bond strength, as water tends to accumulate on the surface of the hybrid layer , inhibit copolymerization inside the adhesive and expedite polymer softening and degradation . Droplets and water channels within the adhesive layer can be mechanically interpreted as internal structural flaws around which stresses concentrate under load, increasing sites for slow crack growth and, therefore, also increasing failure probability. Under static load, surface defects are the key mechanism for failure. However under cyclic fatigue conditions, internal defects (i.e., inside the adhesive or at the interfaces adhesive–dentin or adhesive–composite) gain a more relevant role as sites for slow crack growth coalescing to catastrophic failure .
In order to decrease polymer hydrophilicity and water droplet formation, HEMA-free formulations were introduced. The decrease in polarity within these adhesive polymer chains has led to osmotic process inhibition, but as a side effect, large droplet formation resulting from phase separation reactions has then been observed . Internal flaws have therefore not been eliminated, and sites for fatigue structural degradation persist. In contrast, total-etch systems that ensure the subsequent application of a hydrophobic resin over hydrophilic primers fail to show such droplet structures .
With regard to the complex forces to which the adhesive interfaces are subjected in function, the purposes of this study were to compare the Weibull modulus, characteristic strength and fatigue limit of two dental adhesives using two testing configurations (bending and tensile). The three-fold null hypothesis tested were that there are no differences between the two adhesives tested, the two testing configurations and between the characteristic strength and the fatigue limit.
Materials and methods
Specimens for initial strength and fatigue limit, both for tensile and flexural tests, were manufactured in the same way and randomly divided among test groups.
Freshly extracted caries-free human third molars were cut longitudinally with a low-speed diamond saw (Isomet, Buehler, Lake Bluff, IL, USA) under water irrigation into 1.4 mm slabs. The slabs were further cut to remove the enamel and create dentin slabs of 1.4 mm in thickness, 9 mm in length and 5 mm in width. The dentin occlusal surfaces were grinded wet for 30 s using 600-grit SiC paper to produce relevant dentin smear layer. An one-step self-etch adhesive (G-Bond, GC, Japan) and one three-step etch-and-rinse adhesive (Syntac Classic, Ivoclar Vivadent, Liechtenstein) were applied to the occlusal dentin surface according to Table 1 . The fluorochrome Rhodamine B was compounded into the adhesives (primer + adhesives for Syntac) before application to allow posterior fractographic evaluation. A low concentration (0.1 ppm) was used in order not to affect polymer conversion and bond strength .
|G-Bond, GC, Tokyo, Japan||0612121||One-step self-etch adhesive||4-MET, phA-m, DMA, ethanol, water, filler, photo-initiator, stabilizer||1. Shake bottle
2. Apply on dentin
3. Leave undisturbed for 10 s
5. Light-cure for 10 s
|Syntac Classic, Ivoclar Vivadent, Schaan, Liechtenstein||Primer: K36299
|Three-step etch-and-rinse adhesive||Etchant: 36% phosphoric acid
Primer: Maleic acid, TEGDMA, water, acetone
Adhesive: PEGDMA, glutaraldehyde, water
Heliobond: Bis-GMA, TEGDMA, UDMA
|1. Apply etchanto on dentin for 15 s, rinse and blot-dry
2. Apply primer, leave undisturbed for 15 s, air-dry
3. Apply adhesive, leave undisturbed for 10 s, air-dry
4. Apply Heliobond, air-thin
5. Light-cure for 20 s
A plexiglass split-mold with internal space of 1.4 mm × 5 mm × 35 mm was used to prepare the specimens. With the adhesive interface positioned at the center of the mold, a nanohybrid composite (Grandio, VOCO, Germany) was inserted at both sides of the dentin slabs (occlusal and apical) until the mold was filled. The mold was closed and pressed before light-curing the composite for a total of 60 s on each side (Elipar Trilight, 3M ESPE, Germany) with an output intensity of 750 mW/cm 2 . After removal from the mold, the specimens were wet grinded with 600-grit SiC paper to the final dimensions (1.2 mm thickness, 5 mm in width and 35 mm in length). The specimens were stored for 24 h at 37 °C in distilled water before testing.
For the initial tensile strength (TS), 15 specimens were wet tested in tension in an universal testing machine (Z 2.5, Zwick, Germany) at a cross-head speed of 1 mm/min (see Fig. 1 ). To evaluate the initial bending strength, 15 specimens were wet tested until fracture in a three-point bending Flex machine (Flex, University of Antwerp, Belgium; Span length = 30 mm) at a comparable cross-head speed (see Fig. 2 ).
Tensile fatigue limits (TFL) and bending fatigue limits (BFL) were determined for 10 4 cycles using the same testing conditions as for the initial strength measurements at a frequency of 0.5 Hz. Details of the bending fatigue test have been described elsewhere . The ‘staircase’ approach was used to determine the adhesive fatigue limits .
Five specimens from Syntac and G-Bond were randomly selected for evaluation of the interfacial morphology using a confocal laser scanning microscope (CLSM) under fluorescence and reflectance modes (488 nm wavelength). Prior to application, the adhesives had been previously stained with a fluorescent dye (Rhodamine B), which allowed for fractographic examination due to the spectra excitation.
The fractured surfaces of TS and BS specimens as well as the failed specimens from TFL and BFL were evaluated under a light microscope (LM) (SV 6, Zeiss, Germany) to classify the failure mode in: adhesive (along the interface without composite or dentin involvement), cohesive (totally in dentin or in composite) and mixed (at the interface involving dentin or/and composite). Fractured specimens were further observed under a CLSM (TCS SL, Leica, Germany) under fluorescence and reflectance modes (488 nm wavelength) to highlight the adhesive zone. Further five specimens were randomly selected, sputter coated with gold and observed under a scanning electron microscope (SEM) (SR 50, Leitz ISI, Japan).
A uniform distribution of material defects results in a strength variability which can be statistically treated using the Weibull approach, describing the failure probability <SPAN role=presentation tabIndex=0 id=MathJax-Element-1-Frame class=MathJax style="POSITION: relative" data-mathml='PF(σc)’>PF(σc)PF(σc)
P F ( σ c )
P F ( σ c ) = 1 − exp − σ c σ 0 m
where σ 0 is the characteristic strength at a failure probability of 63.2% ( <SPAN role=presentation tabIndex=0 id=MathJax-Element-3-Frame class=MathJax style="POSITION: relative" data-mathml='PF(σc)=63.2%’>PF(σc)=63.2%PF(σc)=63.2%
P F ( σ c ) = 63.2 %
) and m is the Weibull modulus, respectively . The strength data was evaluated according to this two parameter cumulative Weibull distribution by plotting the failure probability <SPAN role=presentation tabIndex=0 id=MathJax-Element-4-Frame class=MathJax style="POSITION: relative" data-mathml='PF(σc)’>PF(σc)PF(σc)
P F ( σ c )
versus fracture strength σ c .
ln ln 1 1 − P F ( σ c ) = m ⋅ ln σ c − m ⋅ ln σ 0
The parameters m and σ 0 were determined using a maximum likelihood approach. Weibull modulus m , characteristic strength σ 0 and their respective 95% confidence intervals ( D l / D u and C l / C u ) were corrected by a factor corresponding to n = 15 ( b = 0.908), according to the European standard EN 843-5 . The 95% confidence limits for the groups were calculated and differences were considered to be significant when the 95% confidence intervals did not overlap.
FFL = X 0 + d ∑ i n i ∑ n i ± 0.5
SD = 1.62 d ∑ n i ∑ i 2 n i − ∑ i n i 2 ∑ n i 2 + 0.029