The aim of this study was to investigate the influence of shrinkage strain, modulus, and instrument compliance on the polymerization shrinkage stress measurement of light-cured composites, and to determine whether the silorane-based low-shrinkage composite shows a low-polymerization shrinkage stress.
A universal hybrid; Z250 (Z2), a flowable; Z350 (Z3), and a silorane-based; P90 (P9) composite was examined. A modified “bonded disc method” was used to measure the axial shrinkage strain of the composite. For the measurement of the initial modulus development of composites during light curing, a dynamic oscillatory shear test was undertaken using a custom-made oscillation rheometer. A frequency of 6 Hz and strain amplitude of 0.0091 rad for 20 s was employed and the complex shear modulus ( G * ) was determined. A newly designed stress–strain analyzer was made to measure the shrinkage stress of the composites with two modes: (1) high compliance, or (2) low compliance. The shift between the two modes was controlled by an On–Off switch of a negative feedback circuit. Theoretical shrinkage stress was calculated from the shrinkage strain and modulus measured above, and compared with experimentally measured stress. Data were analyzed with one-way ANOVA and Tukey’s post hoc test ( α = 0.05), and correlation analysis was done to investigate the relationship between measured stress and shrinkage strain, modulus, and theoretical stress.
The shrinkage strain of Z3 (4.12%) at 10 min was the highest, followed by Z2 (2.31%) and P9 (0.77%). At 10 s after light curing, Z2 showed the highest modulus (466.2 MPa), Z3 (154.1 MPa), and P9 the lowest (130.7 MPa). The measured stresses with low compliance were much higher than those with high compliance. With high compliance, the contraction stress of Z3 was the highest (2.75 MPa), followed by Z2 (1.54 MPa) and P9 (0.48 MPa). In low-compliance mode, the stresses of Z3 (7.93 MPa) and Z2 (7.48 MPa) were similar ( p = 0.323) while the stress of P9 (3.23 MPa) was much lower. A strong correlation was observed between the theoretical stress and the measured stress with low compliance ( R = 0.996). In high-compliance mode, the shrinkage strain also showed a near-linear relationship with the stress measured ( R = 0.937), but the modulus showed a low correlation with the measured stress ( R = 0.398).
Depending on the instrument compliance, polymerization shrinkage stress showed significant differences for each material. In high-compliance shrinkage strain played a greater role, while in low-compliance shrinkage strain and elastic modulus contributed comparably in determining the shrinkage stress. The low-shrinkage silorane-based composite demonstrated considerable reduction in shrinkage strain and stress.
Dental resin composites are today’s most widely used direct restorative material. One of their main advantages is their adhesive capacity allowing for minimal cavity preparation and superior esthetics. However, as a major drawback, most composites undergo volumetric shrinkage during the polymerization process as the van der Waals forces loosely binding the monomers are exchanged for shorter covalent bonds. As the polymer chains grow in length and cross-linking occurs, polymerization shrinkage is accompanied by an increase in rigidity . Together, these events result in the development of polymerization shrinkage stress within the restored tooth. This stress can manifest as debonding, enamel microcracks, cusp flexure, and microleakge, which in turn can lead to post-operative sensitivity and secondary caries .
Numerous studies have been carried out on polymerization shrinkage stress of composites, and measured stresses exhibit a wide range of values from below 1 MPa to above 15 MPa . Different stress values are even reported for the same material, which primarily can be attributed to the difference in compliance of the measuring device. An experimental setup allowing a high degree of compliance will result in lower values due to relief of some of the shrinkage stress , while a setup restricting the compliance by negative feedback will result in higher stress values . Thus, an inverse relationship exists between instrument compliance and shrinkage stress .
Conflicting results concerning the effect of inorganic filler on shrinkage stress can also be ascribed to differences in compliance of the measuring device, and an increased filler content has reported to be related to both a decrease and increase in shrinkage stress. While the addition of inorganic filler to the resin matrix decreases the overall shrinkage of the composite, it also increases its elastic modulus. Which effect plays the larger role in determining the polymerization shrinkage stress is still a topic of debate.
According to Hooke’s law, and in the absence of permanent deformation, polymerization shrinkage stress can be expressed as the product of the shrinkage strain multiplied by the elastic modulus. Much data is available on the shrinkage strain and elastic modulus of the final polymerized material. However, substituting these values into Hooke’s equation does not result in the true shrinkage stress because during the polymerization process, both factors are changing; each increment of time corresponds to a different shrinkage strain and elastic modulus. Therefore, if one could obtain the changing values of strain and modulus according to time, multiply them at every time point to calculate the stress, and then integrate the results, a clearer understanding of the generation of polymerization shrinkage stress would be possible. Feilzer et al. estimated the theoretical shrinkage stress of a slow chemically cured composite by adding the product of the shrinkage strain and elastic modulus at certain distinct time points. However, no studies have been able to determine the theoretical stress of light-cured composites where the shrinkage strain and elastic modulus increase at a rapid rate during initial light curing.
In an effort to decrease the contraction stress of dental composites, novel low-shrinkage resins utilizing silorane monomers have emerged. The polymerization reaction of these composites is based on a cationic ring-opening polymerization, which results in shrinkage strain values below 1% . A relative decrease of shrinkage stress by 27% compared to methacrylate-based composites has been reported in the literature .
The aim of this study was to investigate the influence of shrinkage strain, elastic modulus, and instrument compliance on the shrinkage stress measurement of light-cured composites, and to compare the measured stress to the theoretical stress estimated from obtained strain and modulus values. In addition, we aimed to determine whether the silorane-based low-shrinkage composite shows a low-polymerization shrinkage stress.
Materials and methods
A universal hybrid, a flowable, and a low-shrinkage composite was used in this study ( Table 1 ). A VIP Junior (Bisco Inc., Schaumburg, IL, USA) was used as the light-curing unit at an irradiance of 600 mW/cm 2 . All measurements were made at 25 ± 0.5 °C.
|Composite (code)||Lot No.||Resin matrix||Filler (wt.%)||Manufacturer|
|1370A3||Bis-GMA, Bis-EMA, TEGDMA, UDMA||0.01–3.5 μm (average 0.6 μm) Zr/silica particles (82 wt.%)||3M ESPE, St. Paul, MN, USA|
|6031A3||Bis-GMA, Bis-EMA, TEGDMA||5–20 nm Zr/silica nanoparticles + 0.6–1.4 μm nano-clusters (65 wt.%)||3M ESPE, St. Paul, MN, USA|
|4762A2||Siloranes||0.01–3.5 μm (average 0.47 μm) quartz particles, yttrium fluoride (76 wt%)||3M ESPE, St. Paul, MN, USA|
Measurement of polymerization shrinkage strain
A modified “bonded disc method” was used to measure the axial shrinkage strain of the composite specimens ( Fig. 1 ). A fixed amount of composite was pressed between a slide glass (Marienfeld, Germany) and a flexible cover glass (Marienfeld, Germany). A spacer consisting of a 0.5 mm diameter metal wire was used to produce specimens 0.5 mm in thickness and 6.0 mm in diameter (C-factor = 6). The tip of a linear variable differential transformer (LVDT) probe was placed on the center of the cover glass and set to the zero point. Baseline data was recorded for 20 s, after which the curing light was automatically switched on. Light exposure time was 40 s and the output voltage from the LVDT was stored on a computer using a data acquisition device (USB-6009, National instrument, Mopac Expwy, Austin, TX, USA). Five specimens were tested for each composite at a rate of 10 data points/s for 10 min.
Measurement of the modulus change during initial light curing
A custom-made rheometer was used to measure the change in shear modulus of the polymerizing composite ( Fig. 2 ). The rheometer consisted of three parts: (1) two glass rods producing parallel plates on which the specimen was placed, (2) an oscillatory shear strain induction unit with a DC motor and crank mechanism, and (3) a stress-measuring unit using an electromagnetic torque sensor.
A sample is loaded between the parallel plates. A sinusoidal oscillating shear strain generated by the DC motor is then transmitted through a crank mechanism to the upper parallel plate. The amplitude of the shear strain is measured by a potentiometer attached to the upper glass rod. As the composite polymerizes, the modulus increases resulting in an increase in the shear force transmitted to the lower glass rod. This force tends to rotate the arm of the torque sensor from its null position, and this deviation of the arm is detected by a photo sensor composed of an infrared LED and a dual photo diode. The electrical signal from the sensor feeds into a servo-amplifier, which flows current into an actuator coil to restore the arm to its original position, instantaneously creating a feedback mechanism. The driving current of the servo-amplifier, which is proportional to the torque, is converted to a continuous voltage signal and stored onto a computer, which is then used to obtain a shear strain–torque curve.
Parallel plates were fabricated by cutting glass rods 3.0 mm in diameter at a 90° angle with diamond disks. The cut ends were sandblasted with 50 μm aluminum oxide (Al 2 O 3 ) particles (Basic Mobil, Renfert, Hilzingen, Germany), treated with silane (Porcelain Primer, Bisco Inc., Schaumburg, IL, USA), and coated with a thin layer of bonding resin supplied for each composite from the manufacturer and light cured. A fixed volume (14.2 mm 3 ) of composite was placed between the 2 mm gap between the glass rods. An oscillatory shear strain of frequency 6 Hz and oscillation amplitude of 0.0091 rad was applied to the sample through the upper rod, and the resultant shear stress was measured by the lower rod. With an oscillatory shear strain of frequency ω , and measured torque T , the complex shear modulus G * can be expressed as follows.
T : torque (N m), R : diameter of the glass rod (m), G * : complex shear modulus of the material (Pa), ω : oscillation amplitude of angular displacement (rad), and h : distance between the glass rods (m).
The device was set so that measurement started simultaneously with the switching on of the light-curing unit. Five measurements were made for each composite at a rate of 1000 data points/s for 20 s. From the resulting torque vs. time curve, the change of complex shear modulus G * for each 0.5 s increment of time was obtained for each material.
Measurement of polymerization shrinkage stress
A newly designed stress–strain analyzer was made to measure the polymerization shrinkage stress of the composites. The compliance of the instrument could be controlled by turning the negative feedback mechanism either on or off ( Fig. 3 ).
The stress–strain analyzer consisted of an LVDT probe as the displacement sensor, a cantilever load cell, and a servo-amplifier and servomotor for the feedback mechanism . A load cell and cantilever were horizontally attached to a crosshead capable of vertical movement. A metal rod of diameter 3 mm was hung onto the cantilever. Another metal rod of identical diameter was fixed onto the metal base so that the two metal rods lined up in the same plane. The space between the rods, where the sample was placed, was continuously measured by the LVDT probe, and could be kept constant by a feedback mechanism. A simple On/Off switch could either close or open the feedback circuit, therefore controlling the compliance of the instrument. In the high-compliance mode without feedback (Stress_Hi_Comp), the axial strain of the sample and force on the load cell caused by the polymerization contraction were measured. The instrument compliance in the high-compliance mode was 0.47 μm/N. In the low-compliance mode with feedback (Stress_Lo_Comp), the polymerization contraction detected by the LVDT activated the servo-amplifier which rotated the servomotor so that an instantaneous offset of the contraction occurred. In this way, the initial thickness of the composite was maintained, and the effect of instrument compliance was highly restricted. The measured strain and force on the load cell were stored on a computer.
The cut ends of the metal rods were sandblasted with 50 μm aluminum oxide (Al 2 O 3 ) particles (Basic Mobil, Renfert, Hilzingen, Germany) and treated with silane (Porcelain Primer, Bisco Inc., Schaumburg, IL, USA), and coated with a thin layer of bonding resin supplied for each composite from the manufacturer and light cured. Composite was placed between the 1.0 mm gap between the metal rods. Baseline data was recorded for 20 s, after which the curing light was automatically switched on for 40 s. Five specimens were tested for each composite at a rate of 10 data points/s for 10 min. The experiment was performed for both the high- and low-compliance modes.
Estimation of theoretical polymerization shrinkage stress during initial light curing
Theoretical shrinkage stress was calculated from the shrinkage strain and modulus measured above. Since the shrinkage strain and modulus of a composite change as polymerization proceeds, multiplying the change in thickness at a specific moment by the elastic modulus of the material at that moment would correspond to the shrinkage stress increment at that moment. Integrating these products from the beginning of light curing to a final set time gives the total cumulative stress in the specimen, expressed as:
σ = ∫ 0 t ( E Specific _ Time × d ε Specific _ Time _ Segment )