Comparison of the Physical and Mechanical Properties of Resin Matrix with Two Photoinitiator Systems in Dental Adhesives

This study evaluated the physical and mechanical properties of resin matrices in dental adhesives with two photoinitiator systems. Resin matrix specimens were made with five different kinds of photoinitiators. Neat resin consisted of 60% 2,2-bis[4-2(2-hydroxy-3-methacryloxypropoxy)phenyl]propane (Bis-GMA) and 40% hydroxyethyl methacrylate (HEMA) by weight, along with camphorquinone (CQ, 1 mol %) and additional components (1 mol % each) as follows: Group 1, 2-(dimethylamino)ethyl methacrylate (DMAEMA); Group 2, ethyl-4-(dimethylamino) benzoate (EDMAB); Group 3, diphenyliodonium hexafluorphosphate (DPIHFP); Group 4, DMAEMA+DPIHFP; Group 5, EDMAB+DPIHFP. The degree of conversion (DC), flexural strength, flexural modulus, microhardness, and ultimate tensile strength were tested. The contribution of each photoinitiator to the DC in a selected group was analyzed with contour plots. One-way ANOVA and Tukey tests (p < 0.05) were used for statistical analyses. The DC of Groups 2, 4, and 5 was similar. The flexural strength was similar in all groups, but flexural modulus was significantly different. Group 3 had the lowest values for all physical and mechanical properties. Among all methods, the microhardness test revealed the greatest degree of difference among the five specimens. CQ, EDMAB, and DPIHFP were the most effective photoinitiators and CQ was the most influential factor for the DC rate.


Introduction
Since dental resin technology was first introduced over 50 years ago, its use has recently increased due to patients' demands for esthetic restorations and clinicians' needs for easy and direct application. It enables clinicians to follow a predictable, conservative, and reliable chairside protocol for enhancing patient smiles and restoring worn and decayed tooth structure. The ability to be minimally invasive and conserve tooth structure is another benefit in the use of dental resin. Therefore, products with resin components can be used on a daily basis to restore caries, close spaces, lengthen teeth, cover dark or discolored teeth, and fix fractured teeth.
The longevity of dental restorative composites is a critical concern for the long-lasting retention of restorations and depends on the stability of the hybrid layer formed between the composite resin and the dentin/enamel structure. Therefore, adhesive resins with an effective polymerization system are needed for the best retention of the hybrid layer. The adhesive or bonding resin promotes a microtensile tester. Each measurement was performed among experimental groups with different photoinitiators. The contribution of each component to the degree of conversion within a group composed of the most effective photoinitiators was analyzed by contour plots.

Resin Components
Neat resin consisted of 60% bis[4-2(2-hydroxy-3-methacryloxypropoxy)phenyl]propane (Bis-GMA) and 40% hydroxyethyl methacrylate (HEMA) by weight. All compounds were purchased from Esstech (Essington, PA, USA). Five kinds of adhesive resins were fabricated by mixing the neat resin with different photoinitiators. The commercial raw materials and mixing ratios are provided in Table 1. Groups 1 and 2 were binary systems with different tertiary amine groups. Groups 4 and 5 were ternary systems using diphenyliodonium hexafluorophosphate (DPIHFP) as the iodonium salt. Group 3 used DPIHFP and lacked a tertiary amine. The resin matrix was stirred with a magnetic stir bar for 1 h in a glass bottle in the dark. Each group was polymerized with a light-curing unit from 3M ESPE (St Paul, MN, USA, 540 mW/cm 2 ) according to the instructions.

Degree of Conversion (DC)
Real-time polymerization was performed in an FT-IR spectrometer (Nicolet 6700, Thermo scientific, West Palm Beach, FL, USA). Spectra were obtained over the 4000-600 cm´1 region and acquired with a resolution of 4 cm´1 for a total of 32 scans per spectrum. An adhesive resin in each group was placed separately on the attenuated total reflectance (ATR) module. The FT-IR spectrum was recorded after exposure to light for 20 s (total recording time: 4 min) utilizing a dental light-curing unit. For each spectrum, the height of the aliphatic C=C peak absorption at 1636 cm´1 and the aromatic C-C peak absorption at 1581 cm´1 was determined using a baseline method. The aromatic C-C vibration was used as an internal standard. The ratio of absorbance intensities was calculated for each group and compared. The DC at each irradiation time was calculated by using the following equation: All tests were performed independently three times.

Three Point Bending Test
Specimens [(25˘2)ˆ(2˘0.1)ˆ(2˘0.1) mm 3 ] were made from each material according to ISO 4049 and cured in five separate 20 s steps for each side (200 s in total). Tests were performed after 24 h of water storage at 37˝C using a universal testing machine (QTest, Instron, Miami, FL, USA, crosshead speed of 0.75 mm/min) according to ANSI/ADA specification No. . Flexural strength was calculated using σ = (3FL)/(2bh 2 ) and flexural modulus by E = (L 3 /4bh 3 )ˆ(F/Y) (both expressed in MPa), where F is the maximum strength, L the distance between rests (20 mm), b the width of the specimen, h the height of the specimen, and F/Y the slope of the linear part of the stress-strain curve.
Ten specimens were made and tested in each group.

Microhardness
The specimens in each group were fabricated by placing the adhesive resin in a stainless steel mold (15 mm in diameter and 1 mm thick) which was inserted between two sheets of clear matrix and photoactivated for eight separate 20 s steps using a light-curing unit. Surface hardness was tested using a dynamic ultra-microhardness tester (DUH-W201S, Shimadzu, Kyoto, Japan) followed by ASTM E384. Electromagnetic force was used to press an indenter (standard type: 115˝triangular pyramid indenter) against a specimen. The pressing force was increased at a constant rate from 0 to the preset test force (100 mN). The holding time of the maximum load was 2 s and the loading speed was 13.2 mN/s. The indentation depth was measured automatically as the indenter pressed against the specimen, allowing a dynamic measurement of the changes occurring in the specimen's resistance to deformation during the indentation process. Microhardness was calculated as 3.8584ˆF/h 2 , where F is the maximum load (mN) and h is the maximum indentation depth (µm). Three specimens were made for each group and each test repeated three times for each specimen (total: 9 times).

Ultimate Tensile Strength
A microtensile strength tester (Bisco, Inc., Schaumburg, IL, USA) tested I-shaped specimens (1 mm width and depth in the center) by ASTM D1708-13. A cyanoacrylate material was used to bond the ends of each specimen to the two free-sliding parts of a specially designed holding device. The jig was able to transmit purely tensile forces to the specimen without any torqueing or bending component. The tensile load was applied at a crosshead speed of 0.5 mm/min until the specimen fractured. The weight (in kilograms) of the loaded force was recorded and the ultimate tensile strength (in MPa) calculated based on the measured width and depth (in mm) of the specimens. After the tests, the specimens were inspected by a microscope (100ˆ) to exclude those with internal voids. All tests were performed 10 times in every group.

Adhesive Resins with Selected Photoinitiators
CQ, EDMAB, and DPIHFP (i.e., Group 5) were selected as the most effective photoinitiators for the adhesives after performing the DC, flexural strength, flexural modulus, microhardness, and ultimate tensile strength tests in five groups. To investigate the contribution of each material to the DC, the molar ratios of the three components were changed as shown in Table 2. After curing for 20 s using FT-IR, the conversion rate to polymer was calculated as described above. Three independent experiments were performed.

Statistical Analysis
One-way analysis of variance was used (α = 5%) among groups to determine significant differences. Pair-wise multiple comparisons were carried out using the Tukey test when the one-way analysis of variance test detected significant differences.  Figure 1 shows the DC for the five kinds of adhesive resins. Groups 2, 4, and 5 had nearly the same conversion rates with maximum average values in real-time FT-IR of 64.55%, 64.56%, and 65.32%, respectively. Group 3 lacked a tertiary amine and had the lowest DC (55.58%). Group 1 had an intermediate value.

Statistical Analysis
One-way analysis of variance was used (α = 5%) among groups to determine significant differences. Pair-wise multiple comparisons were carried out using the Tukey test when the one-way analysis of variance test detected significant differences. Figure 1 shows the DC for the five kinds of adhesive resins. Groups 2, 4, and 5 had nearly the same conversion rates with maximum average values in real-time FT-IR of 64.55%, 64.56%, and 65.32%, respectively. Group 3 lacked a tertiary amine and had the lowest DC (55.58%). Group 1 had an intermediate value.

Three Point Bending Test
The flexural strengths of the five adhesive resin groups were not significantly different (p > 0.05, Figure 2a). However, the flexural modulus was similar and highest among Groups 2, 4, and 5, followed by Group 1 and Group 3 (p < 0.05, Figure 2b). The flexural strengths of the five adhesive resin groups were not significantly different (p > 0.05, Figure 2a). However, the flexural modulus was similar and highest among Groups 2, 4, and 5, followed by Group 1 and Group 3 (p < 0.05, Figure 2b).

Microhardness
The dynamic microhardness of the adhesive resins based on the differences in maximum indentation depth and elastic/plastic deformation is shown in Figure 3a-e. A summary of the microhardness values is displayed in Figure 3f. Group 5 had the highest microhardness (p < 0.05) and Group 3 had the lowest microhardness (p < 0.05). Groups 2 and 4 had similar values (p > 0.05). The differences in microhardness measurements were more significant among the groups than differences from other measurement methods.

Microhardness
The dynamic microhardness of the adhesive resins based on the differences in maximum indentation depth and elastic/plastic deformation is shown in Figure 3a-e. A summary of the microhardness values is displayed in Figure 3f. Group 5 had the highest microhardness (p < 0.05) and Group 3 had the lowest microhardness (p < 0.05). Groups 2 and 4 had similar values (p > 0.05). The differences in microhardness measurements were more significant among the groups than differences from other measurement methods. The flexural strengths of the five adhesive resin groups were not significantly different (p > 0.05, Figure 2a). However, the flexural modulus was similar and highest among Groups 2, 4, and 5, followed by Group 1 and Group 3 (p < 0.05, Figure 2b).

Microhardness
The dynamic microhardness of the adhesive resins based on the differences in maximum indentation depth and elastic/plastic deformation is shown in Figure 3a-e. A summary of the microhardness values is displayed in Figure 3f. Group 5 had the highest microhardness (p < 0.05) and Group 3 had the lowest microhardness (p < 0.05). Groups 2 and 4 had similar values (p > 0.05). The differences in microhardness measurements were more significant among the groups than differences from other measurement methods.

Ultimate Tensile Strength
None of the specimens had defects on the fractured surfaces after the tests as evaluated by a microscope. The microtensile strength of Groups 2, 4, and 5 were not significantly different (p > 0.05, Figure 4). Group 3 had the lowest strength (p < 0.05).

Contribution of Each Photoinitiator to the DC in the Selected Group
The contour plots in Figure 5 describe the relationship between two chosen components. The areas of the same color indicate identical conversion rates when the remaining photoinitiator is set to the lowest molar ratio level: EDMAB = 1 mol % (Figure 5a), CQ = 0.3 mol % (Figure 5b), and DPIHFP = 0 mol % (Figure 5c). As shown in Figure 5, CQ was the most important factor for DC followed by DPIHFP and EDMAB. When EDMAB was set to 1 mol % (Figure 5a), the DC did not change significantly, showing a wide width with the same colors. When CQ was set to 0.3 mol % (Figure 5b), the color width was much narrower than the others. Without DPIHFP, the DC was 59%-63%. When DPIHFP was present in the maximum amount, the DC increased to 67%.

Ultimate Tensile Strength
None of the specimens had defects on the fractured surfaces after the tests as evaluated by a microscope. The microtensile strength of Groups 2, 4, and 5 were not significantly different (p > 0.05, Figure 4). Group 3 had the lowest strength (p < 0.05).

Ultimate Tensile Strength
None of the specimens had defects on the fractured surfaces after the tests as evaluated by a microscope. The microtensile strength of Groups 2, 4, and 5 were not significantly different (p > 0.05, Figure 4). Group 3 had the lowest strength (p < 0.05).

Contribution of Each Photoinitiator to the DC in the Selected Group
The contour plots in Figure 5 describe the relationship between two chosen components. The areas of the same color indicate identical conversion rates when the remaining photoinitiator is set to the lowest molar ratio level: EDMAB = 1 mol % (Figure 5a), CQ = 0.3 mol % (Figure 5b), and DPIHFP = 0 mol % (Figure 5c). As shown in Figure 5, CQ was the most important factor for DC followed by DPIHFP and EDMAB. When EDMAB was set to 1 mol % (Figure 5a), the DC did not change significantly, showing a wide width with the same colors. When CQ was set to 0.3 mol % (Figure 5b), the color width was much narrower than the others. Without DPIHFP, the DC was 59%-63%. When DPIHFP was present in the maximum amount, the DC increased to 67%.

Contribution of Each Photoinitiator to the DC in the Selected Group
The contour plots in Figure 5 describe the relationship between two chosen components. The areas of the same color indicate identical conversion rates when the remaining photoinitiator is set to the lowest molar ratio level: EDMAB = 1 mol % (Figure 5a), CQ = 0.3 mol % (Figure 5b), and DPIHFP = 0 mol % (Figure 5c). As shown in Figure 5, CQ was the most important factor for DC followed by DPIHFP and EDMAB. When EDMAB was set to 1 mol % (Figure 5a), the DC did not change significantly, showing a wide width with the same colors. When CQ was set to 0.3 mol % (Figure 5b), the color width was much narrower than the others. Without DPIHFP, the DC was 59%-63%. When DPIHFP was present in the maximum amount, the DC increased to 67%.

Discussion
Dental adhesives are often used before applying polymer-based filling resins or other kinds of esthetic restorations in order to completely seal and protect the tooth from chemical/physical stimulation and marginal leakage. Adhesives are composed of methacrylate-based monomers that undergo free-radical polymerization and provide adhesion between a restoration and the natural tooth. Thus far, an evaluation of the physical and mechanical properties of the adhesives has been considered prior to the adaptation. One solution could be a change in the photoinitiators used in dentin adhesives, which in turn can affect the physical and mechanical aspects in the improvement of clinical performance [11,13,20,21].
Usually, a binary photoactivator system comprising CQ and an amine are used. Diaryliodonium salts with complex metal halides as weakly nucleophilic counter ions are efficient photoinitiators for UV-irradiated monomer systems which absorb light below 300 nm and are extensively used in paints and coatings [22]. However, curing based on visible light is preferred because commercial lightcuring units used in dentistry emit 400-500 nm light. CQ can compensate for this drawback by changing the highest absorbance peak of resin composites to the visible light region [23]. Thus, the three component system is a practical and realistic method that can be easily adapted in dentistry. In this study, photoinitiators with different polymerization kinetics were investigated through various evaluation methods. The relationship among the most effective photoinitiators are expressed as contour plots.
As a result, Group 3 (1.0 mol % CQ and DPIHFP) was significantly inferior in all tests. The sensitizer lacks a tertiary amine group and cannot perform electron transfer or polymerize effectively. In the appropriate excited state, the diketone combines with the reducing agent to form an excited state complex (exiplex), which then breaks down to form reactive free radicals [24]. Group 3 had insufficient radical formation, resulting in the poorest physical and mechanical properties. Accordingly, amine products such as EDMAB or DMAEMA are most effective for proper polymerization.
EDMAB is a more stable product compared with DMAEMA because it is in a powder rather than a volatile liquid form. Previously, EDMAB was reported to have a higher double bond conversion rate than DMAEMA [25]. Group 2 (1.0 mol % CQ and EDMAB) had superior DC, flexural modulus, and microhardness values compared to those of Group 1 which contained DMAEMA. Group 4 contained CQ, DMAEMA, and DPIHFP and had similar DC, flexural strength, flexural modulus, and ultimate tensile strength, but had a larger microhardness value when compared with Group 5 (CQ, EDMAB, and DPIHFP). In conclusion, a ternary system is less sensitive to the selection of DMAEMA or EDMAB as an electron donor compared with a binary system. However, a photoinitiator system with EDMAB as an electron donor is more likely to possess superior physical and mechanical properties. Therefore, a proper tertiary amine (i.e., EDMAB) has to be selected prior to the comparison of a two-or three-component system.

Discussion
Dental adhesives are often used before applying polymer-based filling resins or other kinds of esthetic restorations in order to completely seal and protect the tooth from chemical/physical stimulation and marginal leakage. Adhesives are composed of methacrylate-based monomers that undergo free-radical polymerization and provide adhesion between a restoration and the natural tooth. Thus far, an evaluation of the physical and mechanical properties of the adhesives has been considered prior to the adaptation. One solution could be a change in the photoinitiators used in dentin adhesives, which in turn can affect the physical and mechanical aspects in the improvement of clinical performance [11,13,20,21].
Usually, a binary photoactivator system comprising CQ and an amine are used. Diaryliodonium salts with complex metal halides as weakly nucleophilic counter ions are efficient photoinitiators for UV-irradiated monomer systems which absorb light below 300 nm and are extensively used in paints and coatings [22]. However, curing based on visible light is preferred because commercial light-curing units used in dentistry emit 400-500 nm light. CQ can compensate for this drawback by changing the highest absorbance peak of resin composites to the visible light region [23]. Thus, the three component system is a practical and realistic method that can be easily adapted in dentistry. In this study, photoinitiators with different polymerization kinetics were investigated through various evaluation methods. The relationship among the most effective photoinitiators are expressed as contour plots.
As a result, Group 3 (1.0 mol % CQ and DPIHFP) was significantly inferior in all tests. The sensitizer lacks a tertiary amine group and cannot perform electron transfer or polymerize effectively. In the appropriate excited state, the diketone combines with the reducing agent to form an excited state complex (exiplex), which then breaks down to form reactive free radicals [24]. Group 3 had insufficient radical formation, resulting in the poorest physical and mechanical properties. Accordingly, amine products such as EDMAB or DMAEMA are most effective for proper polymerization.
EDMAB is a more stable product compared with DMAEMA because it is in a powder rather than a volatile liquid form. Previously, EDMAB was reported to have a higher double bond conversion rate than DMAEMA [25]. Group 2 (1.0 mol % CQ and EDMAB) had superior DC, flexural modulus, and microhardness values compared to those of Group 1 which contained DMAEMA. Group 4 contained CQ, DMAEMA, and DPIHFP and had similar DC, flexural strength, flexural modulus, and ultimate tensile strength, but had a larger microhardness value when compared with Group 5 (CQ, EDMAB, and DPIHFP). In conclusion, a ternary system is less sensitive to the selection of DMAEMA or EDMAB as an electron donor compared with a binary system. However, a photoinitiator system with EDMAB as an electron donor is more likely to possess superior physical and mechanical properties. Therefore, a proper tertiary amine (i.e., EDMAB) has to be selected prior to the comparison of a two-or three-component system.
In contrast to flexural strength and ultimate tensile strength, the microhardness tests had clear differences among the five groups. These results were in accord with previous studies [26,27]. It was also proven that dynamic indentation as controlled by a computer program resulted in obvious differences and relatively little standard deviation among groups [28]. Thus, this protocol should be additionally performed with other testing methods like the three point bending test and DC to detect definitive physical and mechanical differences among resin combinations.
A three-component system containing iodonium salts as the third factor could result in a higher DC and enhanced mechanical properties compared with a two-component system. The results presented here were in accordance with previous observations [23]. In general, Group 5 had an excellent DC, flexural modulus, microhardness, and ultimate tensile strength. From these results, CQ, EDMAB, and DPIHFP were chosen as effective photoinitiators, and further studies with different molar ratios were performed. Each was a necessary component for effective polymerization, but CQ was determined to be the most influential factor by effectively changing the DC in the contour plots.
The silanized fillers (about 10 wt %) could be mixed with the resin matrix for the resin adhesives. Generally, it has been known that enough filler can reduce polymerization shrinkage, improve resistance to external forces, and increase the durability of filling resins (packable or flowable resins). As a result, it is said that fillers improve the biocompatibility of resins by reducing the release of monomers [29][30][31][32][33]. Kim et al., however, did not find significant differences in the DC of adhesives with 0.5%-3% filler [34]. As such, it was assumed that addition of filler to the adhesive resins would not affect the DC so they were not instituted in these studies. Otherwise, Halvorson et al. observed that the DC of composites vary inversely with the percentage of filler in the material [35]. The researchers determined that the fillers might scatter light which can hinder light penetration. This is especially problematic when particle size approaches the output wavelength of the light-curing unit [36,37]. Their results were opposed to our initial hypothesis. Future studies should address these effects with microor nanofillers combined with different photoinitiator systems directly.
The durability of a photoinitiator system with DPIHFP is an important factor for evaluating the clinical usages of the photoinitiation system. In the previous study, a ternary photoinitiator system with iodonium salts maintained increased bond strength with dentin in a model of self-etching adhesive system after 1 year of aging [17]. Improvement was also observed in the polymerization kinetics of a model of dental adhesive resin using a ternary photoinitiator system which made the material less sensitive to the residual presence of a solvent before photoactivation [38]. The yellow effect of CQ could be reduced by the white color of the powdered iodonium salts additionally. In the further studies, the dental adhesives with alternative photoinitiators in this study need to be evaluated with respect to color stability, durability, degradation rate after polymerization, and the deformation examination after hybrid formation with different solvents in an animal experiment.
Despite these limitations, this study showed the effectiveness of the EDMAB as a tertiary amine and ternary photoinitiator system in dental adhesives. Additionally, dynamic microhardness tests were capable of identifying clear differences in properties among test groups. Additionally, it was proven that the molar ratio of CQ was the most influential on DC rate.

Conclusions
This study found that dental adhesives with EDMAB had better mechanical and physical properties as compared with adhesives containing DMAEMA. Also, iodonium salts, DPIHFP, did not affect the mechanical and physical properties without amine initiators. Dynamic microhardness tests were a more effective test method than the flexural strength and ultimate tensile strength tests for investigating the differences among photoinitiator systems. It was demonstrated that the molar ratio of CQ was the most influential factor on the DC rate in the present study.