(1) To investigate the properties, and their correlations, of the acoustic emission (AE) from interfacial debonding of Class-I composite restorations during curing. (2) To establish the relationship between the theoretical shrinkage stress and the level of interfacial debonding in such restorations as determined by AE measurement.
An AE sensor was attached onto the surface of human molars with a Class-I composite restoration of 4 mm (length) × 3 mm (width) × 2 mm (depth) to monitor their debonding from the tooth tissues during curing. Background signals were analyzed before curing to determine the threshold amplitude for noise filtering. Three groups ( n = 3) of composites with different levels of shrinkage were tested: (1) Z100™, (2) Filtek™ Z250, and (3) Filtek™ LS. All restorations were cured with an LED blue light operated at 1200 mW/cm 2 for 40 s. AE signals were recorded continuously from the start of curing for 10 min, and their frequency, amplitude and duration were analyzed. Finally, the cumulative number of AE events was compared with the theoretical maximum shrinkage stress that could be generated by the composites.
The amplitude of the background signals was below 30 dB, which was chosen as the threshold for noise filtering. The amplitude of all debonding events ranged from 30 to 50 dB, and their duration was below 100 μs. The peak frequency had two main bands: 100–200 kHz and 700–800 kHz. The duration time increased with increasing amplitude, but no correlation was found between the peak frequency and the other two parameters. The cumulative number of AE events was 30.67 ± 2.31, 14.00 ± 7.81 and 5.67 ± 3.06 for Z100, Z250 and LS, respectively, which corresponded well with the theoretical maximum shrinkage stress they could produce, i.e. 42.5, 97.5 and 182.5 MPa. R 2 = 0.9955 for the linear regression. The theoretical shrinkage stress below which no AE events were detected was about 14.3 MPa.
For the materials considered, the amount of interfacial debonding produced in a Class-I restoration during curing increased linearly with the theoretical maximum shrinkage stress of the composite. The theoretical stress below which no AE events were detected was similar to composite–dentin bond strength reported in the literature.
Dental composite resins are now widely used for repairing decayed and damaged tooth tissues because of their esthetics, ease of manipulation, acceptable mechanical properties and environmentally friendliness . Unfortunately, debonding of the restoration from the tooth tissues caused by polymerization shrinkage remains a serious problem for these restorative materials. Leakage and bacterial invasion may follow interfacial debonding, resulting in marginal staining, pulpal inflammation and eventually secondary caries .
It is difficult, if not impossible, to measure shrinkage stress in real teeth due to their small size and complex geometry. While performed routinely, measurement of shrinkage stress generated in simple disc specimens using either uniaxial or cantilever beam configurations may not be representative of the clinical situation. Thus, model teeth with simple cavity shapes have been used to assess the shrinkage stress generated by dental composites through the measurement of “cuspal deflection” .
Alternatively, one may infer the level of shrinkage stress produced by a resin composite from the amount of debonding it causes to the restoration. Traditional methods for measuring debonding or the presence of interfacial gaps rely on the assessment of leakage using optical microscopy, scanning electron microscopy (SEM) or transmission electron microscopy (TEM) . However, they all require destructive sectioning of the specimens and they are, at best, semi-quantitative in their assessment. As a non-destructive method, X-ray micro-computed tomography (micro-CT) has recently been used to study interfacial debonding of composite restorations . However, it is not always possible to differentiate between interfacial gaps and adhesives, which are often radiolucent. A radiopaque dye, such as silver nitrate solution, may be used to provide the contrast , but one should bear in mind that the adhesive layer can be permeable, allowing the dye to penetrate into the tooth even in the absence of debonding. Another non-destructive method, acoustic emission (AE) measurement, has also been applied to the detection of interfacial debonding in composite restorations. The advantage of this method lies in its ability to measure debonding in real time during curing of the composite restoration. Choi et al. applied the AE method to detect marginal fracture in ring-shaped composite restorations using different substrates. They found that a larger number of AE hits were accompanied by larger marginal gaps. Also using AE, Liu et al. found that the higher the C-factor of a restoration, the more interfacial debonding there was. According to Li et al. , there appeared to be a positive correlation between the total number of AE events and the shrinkage stress created in a tensometer. Kim et al. further showed that the relationship between the two was linear.
In this paper, the properties, and their correlations, of the AE events are examined in more depth with the aim of gaining new insights into debonding of composite restorations. Comparisons are also made with predicted shrinkage stress based on a recent analytical treatment to verify its validity.
Materials and methods
Nine intact human third molars, with similar dimensions, that had been extracted and kept in saturated thymol solution at 4 °C for less than a month were used to prepare the specimens. The teeth were washed in running tap water and then kept in de-ionized water at room temperature for 24 h prior to preparation. All the teeth were prepared with a Class-I cavity of 4 mm (length) × 3 mm (width) × 2 mm (deep) ( Fig. 1 ) by the same operator following standard clinical procedures with a high-speed hand piece and conically shaped carbide burs. The dimensions of each cavity were measured with a vernier caliper (Mitutoyo, Japan) to ensure their uniformity. The prepared teeth were then divided into three groups ( n = 3) randomly.
The newly cut cavity surfaces were wiped using ethanol pads, washed under running water and dried with compressed air. Three different composite resins with different reported shrinkage stresses were used to restore the cavities: (1) Z100™, (2) Filtek™ Z250, and (3) Filtek™ LS (all 3 M ESPE, USA). In accordance with the manufacturer’s instructions, Adper™ Single Bond Plus was used as the bonding agent for Groups (1) and (2), and LS System Adhesive for Group (3), both from the same manufacturer. All the composites were bulk-cured with an Elipar™ S10 LED light-curing unit (3 M ESPE, USA) for 40 s. The intensity of the curing unit was set at 1200 mW/cm 2 . To control their dimensions, the top surface of each restoration was pressed flat before curing.
An AE test system (PCI-2, Physical Acoustic Corporation, USA) was used to monitor the interfacial debonding between the tooth structure and composite resin during curing. 40 dB was chosen for pre-amplification, and 100 kHz–2 MHz for the band pass. For each test, an AE sensor (S9225, Physical Acoustic Corporation, USA) was adhered to one of the proximal surfaces of the tooth ( Fig. 1 ) with cyanoacrylate adhesive (Super Bond, Staples Inc, USA). During testing, the tooth was wrapped with a piece of wet paper to keep it moist so that it would not crack from drying and generate spurious AE signals.
The collection of AE signals started prior to curing the composite. Background signals thus collected were analyzed to determine the threshold amplitude for noise filtering. Then, AE signals were recorded continuously for 10 min from the start of curing. Their cumulative number, peak frequency, amplitude and duration ( Fig. 2 ) were analyzed and compared between the different composites.
Shrinkage stress calculation
According to Fok , when a composite is highly constrained, the shrinkage stress ( σ ) it produces under uniaxial conditions could be approximated by the following equation:
where E is the elastic modulus and ɛ the linear shrinkage strain. The composite in a Class-I restoration is highly constrained, especially at the floor of the cavity, where it is reasonable to assume that material is being constrained in all three directions. In this case, Generalized Hooke’s Law gives an equi-triaxial stress that is 1/(1 − 2 ν ) times that of the uniaxial case, i.e.,
where ν is Poisson’s ratio. Thus, theoretical maximum shrinkage stress was calculated for the three composite materials using Eq. (2) above, together with elastic modulus and linear shrinkage strain values obtained from their technical data sheets . A Poisson’s ratio of 0.3 was assumed for all three materials. The calculated values were then plotted against the cumulative numbers of AE events to examine their relationship.