Abstract
Objective
CAD/CAM ceramic restorative material is routinely bonded to tooth substrates using adhesive cement. This study investigates micro-crack growth and damage in the ceramic/dentin adhesive interface under fatigue shear testing monitored using the acoustic emission (AE) technique with optical coherence tomography (OCT).
Methods
Ceramic/dentin adhesive samples were prepared to measure the shear bond strength (SBS) under static load. Fatigue shear testing was performed using a modified ISO14801 method. Loads in the fatigue tests were applied at 80%, 70%, and 60% of the SBS to monitor interface debonding. The AE technique was used to detect micro-crack signals in static and fatigue shear bond tests.
Results
The results showed that the average SBS value in the static tests was 10.61 ± 2.23 MPa (mean ± standard deviation). The average number of fatigue cycles in which ceramic/dentin interface damage was detected in 80%, 70% and 60% of the SBS were 152, 1962 and 9646, respectively. The acoustic behavior varied according to the applied load level. Events were emitted during 60% and 70% fatigue tests. A good correlation was observed between crack location in OCT images and the number of AE signal hits.
Significance
The AE technique and OCT images employed in this study could potentially be used as a pre-clinical assessment tool to determine the integrity of cemented load bearing restored ceramic material. Sustainable cyclic load stresses in ceramic/dentin-bonded specimens were substantially lower than the measured SBS. Predicted S–N curve showed that the maximum endured load was 4.18 MPa passing 10 6 fatigue cyclic.
1
Introduction
With advances in adhesive methods and ceramic materials, adhesive restorations are advantageous because the macro-retentive design is no longer a prerequisite if an adequate amount of tooth surface is available for bonding. Clinicians are particularly interested in restoring minimal or absent macro-retentive preparations with extensive dentin- or enamel-bonded ceramic coverage . Increased demand in recent years for esthetic and metal-free restorations has led to the development of a computer design/manufacturing (CAD/CAM) system for fabricating ceramic inlays, onlays and veneers. CAD/CAM system generated ceramics are currently available that provide a novel means of restoring large cavities in posterior teeth and achieving chair-side design and automated production of all-ceramic monolithic single-unit restorations .
However, challenges remain when working with ceramics. One of the main reasons for extensive or total failure in ceramic restorations are luting defects or resin cement wear between the ceramic restoration and the tooth substance, especially to the dentin where obtaining a reliable bond is more challenging than to the enamel because of the higher water content of dentin .
Generally, in vitro shear or tensile testing is used to evaluate the effectiveness of an adhesive system involving bond strength measurement to the enamel or dentin. However, this is not a likely mode of failure for bonds in the mouth, where failure is considered to result from repeated loads over many months or years and at stresses well below the ultimate bond strength . This suggests that fatigue studies where cyclic loads in bonded specimens are evaluated may provide better insight into the in vivo performance and give more realistic sustainable stress values . Nevertheless, fatigue micro-crack growth in a bonded adhesive layer is difficult to monitor from in vitro studies requiring sectioning or dissolving sample dental tissues to confirm failure paths.
The acoustic emission (AE) technique is a non-destructive technique that has been used in engineering to detect the onset and failure progression in structures . The AE technique can offer the advantages of being a non-stop technique that can monitor the condition of materials under investigation throughout the test. Combined with conventional static fracture testing machines, AE can identify failure initiation, the initial damage site, damage propagation and catastrophic failure of the material and help elucidate the complex failure mechanism. It has been applied in the analysis of fracture behavior in different types of dental materials such as ceramics , composites and porcelain . However, the AE technique was rarely employed in dental ceramic fracture testing under cyclic loads.
This study applied the AE technique to monitor the failure process in CAD/CAM ceramic block/dentin interfaces using the total-etch adhesive system under different cyclic load stages. The adhesive layer image was obtained using optical coherence tomography (OCT) scanning to analyze the fatigue micro-crack growth and validate the AE signals results at different cyclic load stages.
2
Materials and methods
2.1
Specimen preparation
Thirteen caries-free, extracted human molars were embedded into epoxy resin blocks and sectioned mesio-distally using a low-speed saw (IsoMet 1000, Buehler Ltd., Lake Bluff, IL, USA). The facial coronal portion of the molar was milled from the outer surface of the block with a milling machine (VMM, King-Long Co., Taipei, Taiwan), polished with a grinding machine (P20FR, Holy Instrument Co., Taipei, Taiwan), and vinyl tape drilled with a hole 2 mm in diameter and approximately 0.05 mm in thickness (about the adhesive layer thickness) placed on the sample to expose a similar dentin surface. A ceramic piece (2 mm × 2 mm × 2 mm) was prepared from a ceramic block (Pro-CAD, Ivoclar Vivadent Inc., Schaan, Liechtenstein) using a saw machine. An etch-and-rinse Variolink II adhesive system was applied to bond the dentin and ceramic. The exposed dentin was acid etched with 35% phosphoric acid gel and air-dried. Heliobond was uniformly applied to the dentin surface. The ceramic piece was etched for 90 s with 6% hydrofluoric acid and cleaned with water spray, then treated with silane porcelain primer coupling etching. Light cured cement was adopted to bond the dentin and ceramic together ( Fig. 1 ).
2.2
Shear bond strength and cyclic load testing
Eight specimens were produced for the shear bond strength (SBS) tests using the Instron E3000 (Instron, Canton, MA, USA) material testing machine. A custom-made stainless steel knife-edged chisel, aligned 0.5 mm away from the dentin substrate, was used as the loading instrument ( Fig. 2 ). The load was applied with a crosshead speed of 0.5 mm/min until complete debonding occurred. The maximum shear bond loads/strengths were recorded.
The fatigue tests were carried out according to the maximum shear load of the static test modified with the standard ISO 14801(2007) test method to evaluate the mechanical resistance of the ceramic/dentin adhesive system. The cyclic loads were set at 80%, 70%, and 60% of the static maximum shear load. The R value (Fmax/Fmin) was set at 10. The test frequency was set at 1 Hz because the human mastication frequency was found to be 0.94 Hz–2.17 Hz from the literature . Three specimens were tested at each cyclic load. The number of cycles at each load was recorded until debonding occurred on ceramic/dentin adhesive interface.
2.3
AE analysis
During the SBS and cyclic load tests an AE signal wide band transducer (Broadband sensor S9225, Physical Acoustic Corporation (PAC) Princeton Junction, NJ, USA) was glued with resin (Triad Gel, Dentsply, York, PA, USA) to the resin block in which the sample was embedded ( Fig. 2 ). Signals detected by the transducer were passed through 40 dB gain preamplifiers with a band pass of 100 k–2 MHz (Model 2/4/6, PAC) . AE signals were then recorded during the load period, load versus AE hit in SBS testing and number of cycles versus AE hit in cyclic load test for each specimen were recorded to evaluate how the adhesive system failure process differed between static and fatigue tests.
2.4
OCT scanning
In order to monitor the micro-crack propagation in the adhesive interface another nine specimens were prepared and divided into three groups to perform the fatigue tests under 80%, 70%, and 60% of the static maximum shear load. However, the corresponding number of cycles at each load was set to half the average total number of cycles at which debonding occurred in the previous cyclic tests, i.e. 76 (152/2), 983 (1965/2) and 4823 (9646/2) numbers for 80%, 70%, and 60% of load, respectively ( Table 1 ).
| Load percentage | Cyclic numbers of debonding | Average total cycle |
|---|---|---|
| 80 | 194 (sample 1) | 152 |
| 130 (sample 2) | ||
| 132 (sample 3) | ||
| 70 | 2427 (sample 1) | 1965 |
| 2206 (sample 2) | ||
| 1262 (sample 3) | ||
| 60 | 12,527 (sample 1) | 9646 |
| 9858 (sample 2) | ||
| 6562 (Sample 3) |
Ceramic/dentin interface debonding was not found in all specimens and exposed ceramic surfaces were scanned perpendicularly using a swept-source OCT system (OCM1300SS, Thorlabs Inc., Newton, NJ) after testing. Several two-dimensional (2D) OCT images were acquired and recorded to obtain 3D datasets at each time point. The swept-source OCT system had a median wavelength of 1310 nm, an axial resolution of around 10 μm in tissue, total power of 10 mW and an A-scan rate of 16 kHz (Huber et al., 2005). For each measurement, a 3D dataset covering a volume of 3 mm × 3 mm × 3 mm corresponding to 512 × 512 × 512 pixels in the X – Y – Z directions was obtained within 1 min.
2
Materials and methods
2.1
Specimen preparation
Thirteen caries-free, extracted human molars were embedded into epoxy resin blocks and sectioned mesio-distally using a low-speed saw (IsoMet 1000, Buehler Ltd., Lake Bluff, IL, USA). The facial coronal portion of the molar was milled from the outer surface of the block with a milling machine (VMM, King-Long Co., Taipei, Taiwan), polished with a grinding machine (P20FR, Holy Instrument Co., Taipei, Taiwan), and vinyl tape drilled with a hole 2 mm in diameter and approximately 0.05 mm in thickness (about the adhesive layer thickness) placed on the sample to expose a similar dentin surface. A ceramic piece (2 mm × 2 mm × 2 mm) was prepared from a ceramic block (Pro-CAD, Ivoclar Vivadent Inc., Schaan, Liechtenstein) using a saw machine. An etch-and-rinse Variolink II adhesive system was applied to bond the dentin and ceramic. The exposed dentin was acid etched with 35% phosphoric acid gel and air-dried. Heliobond was uniformly applied to the dentin surface. The ceramic piece was etched for 90 s with 6% hydrofluoric acid and cleaned with water spray, then treated with silane porcelain primer coupling etching. Light cured cement was adopted to bond the dentin and ceramic together ( Fig. 1 ).
2.2
Shear bond strength and cyclic load testing
Eight specimens were produced for the shear bond strength (SBS) tests using the Instron E3000 (Instron, Canton, MA, USA) material testing machine. A custom-made stainless steel knife-edged chisel, aligned 0.5 mm away from the dentin substrate, was used as the loading instrument ( Fig. 2 ). The load was applied with a crosshead speed of 0.5 mm/min until complete debonding occurred. The maximum shear bond loads/strengths were recorded.
