Effect of Long-Chain Silane on Mechanical Properties of Experimental Resin Composites

To investigate mechanical properties of experimental resin composites containing 8-methacryloxyoctyltrimethoxysilane (MOS) and a more commonly used silane monomer, 3-methacryloxypropyltrimethoxysilane (MPS). MOS and MPS were hydrolysed in 95vol% ethanol solution at pH 4 for 1 h before silanisation of SiO2-based bioactive glass powder fillers (Ø1.0 μm, Schott, Germany) at 3, 6, 9 and 12wt% of silane. After drying in a vacuum oven (80 °C, 5 kPa, 2 h) they were mixed with a monomer mixture of urethane dimethacrylate and triethylene glycol dimethacrylate in a 70:30 weight ratio, camphoroquinone 0.7wt% and 2-(dimethylamino)ethyl methacrylate 0.7wt% to produce eight experimental groups of resin composites with a filler load of 72.7wt%. After photo-polymerisation for 25 min and post-curing at room temperature (24 h), uniform rectangular beams were cut out for the three-point bending test (2 mm×2 mm×25 mm, n = 8). The surface Vickers microhardness (VH) test (press-load 2.94 N for 10s, n = 6) was performed before observation under 3-D optical microscopy. Statistical analysis was performed using ANOVA followed by the Tukey post hoc test (p < 0.05). ANOVA revealed that both the silane type and concentration significantly affected flexural modulus (p < 0.001), flexural strength (p = 0.001) and microhardness (p < 0.001). Resin composites containing MOS exhibited lower flexural modulus but higher strength at a lower concentration than MPS. MOS resin composites exhibited lower surface microhardness than MPS with the corresponding silane content. 8-Methacryloxyoctyltrimethoxysilane may confer different mechanical properties in resin composites compared to 3-methacryloxypropyltrimethoxysilane at the same concentration. Surface microhardness is reduced with 8-methacryloxyoctyltrimethoxysilane.


Introduction
Silicon-based compounds play an important role in dentistry and its materials. Silanes (silane coupling agents) are silicon-containing synthetic, hybrid compounds which have essentially the synthetic ≡ C-Si ≡ bonds that can contribute to chemical adhesion between dissimilar materials, viz. organic and inorganic [1]. Silanes are ubiquitously used for adhesion promotion between dissimilar materials, e.g., glass fibres to the resinous matrix in fibre-glass technology and in demanding environments. In dentistry, silanes are used to bond inorganic silica-based glass fillers to a resin matrix in resin composites used as dental restorative materials, because the fillers must be immobilised and permanently fixed to the resinous matrix of the composite structure [2]. Silanes are also indicated for resin bonding, when the metallic prosthodontic structures are first silica-coated, silanised, and then luted with adhesive resin composite cement. For example, adhesion can be promoted between silanised, silica-coated metal alloys to luting resin cements [3,4], as well as bonding orthodontic brackets [5]. Silica-coating followed by silanisation may also be advantageous in minor repair of silver dental amalgam fillings [6]. Even coating titanium dental implants with silanes has been attempted at the laboratory level because of their antifungal properties [7].
However, the current gold standard silane monomer in dental applications, 3-methacryloxypropyltrimethoxysilane (MPS), is not without limitations. It, as all silane coupling agents, needs to be hydrolysed (activated) before use. Yet, pre-hydrolysed silanes have a relatively short shelf life [8] and in commercially available dental silanes this issue is often managed by adding suitable inhibitors. It has been shown that dental silane products are not exactly identical but vary what comes to their exact contents and adhesion potency [9]. Despite strong adhesion between silica and silanes via siloxane linkages (≡Si-O-Si≡) created by a condensation reaction [10], the hydrolytic stability remains one of the main drawbacks of MPS which occurs at the alkoxyl bonds in this adhesive siloxane interface [4,11]. As a result, silanisation using MPS is highly technique sensitive. Several other factors affect adhesive strength, including the silane type [12,13], solvent system [12] and drying, i.e., the condensation time after silanisation, or, how long an applied silane film should set and cure before luting [14].
Since the hydrolytic stability of MPS is a concern, some other reactive silane monomer alternatives, such as 3-isocyanatopropyltriethoxysilane, 3-acryloyloxypropyltrimethoxysilane, and fluoroalkyltrimethoxysilanes have been attempted [13,15,16]. These silanes have been regarded as potential coupling agents due to their contribution to improved shear bond strength (SBS) and enhanced resistance to hydrolysis. These features might exceed those for pre-hydrolysed MPS products currently commercially available. Functional silanes blended with cross-linker silanes (also called bis-silanes) may improve the hydrolytic stability of the siloxane film [16] and bond strength [6] due to the decreased thickness by molecular entanglement of the silane polymer network of the hydrophobic, interfacial silane layer [17]. The reactive silane monomer concentrations in solvents to apply for silanisation are always relatively low, at the range of ca. 0.5-2%. With a suitable optimised concentration, a significantly higher adhesion strength may result [11].
Recently, a novel long chain length silane, 8-methacryloxyoctyltrimethoxysilane (MOS), has been synthesised and introduced with some molecular variations compared with the current gold standard, 3-methacryloxypropyltrimethoxysilane, which is also known as 3-(trimethoxysilyl)propyl methacrylate. More specifically, 8-methacryloxyoctyltrimethoxysilane is special because of its longer hydrocarbon linker (spacer) with a length of 8 carbon atoms between the functional groups. To date, MOS has not been widely reported in dental literature, in particular its mechanical properties when used in experimental resin composites. It is noteworthy that MOS has some structural similarities to MPS, viz. it consists of a trimethoxysilane entity on one molecular end and a methacrylate group on the other end. That said, these two silanes are chemically expected to bond similarly in terms of reacting with unpolymerised resin from the organofunctional group side and glass powder with hydrolysable alkoxy groups, on the other end. The principal difference between MPS and MOS is that MOS has a longer hydrocarbon linker part with 8 carbon atoms instead of only three in MPS (Fig. 1).
The chemical function of a hydrocarbon chain linker part, -(CH 2 ) n -, is to act as a spacer between the hydrolysable Si-OR groups and the organofunctional moiety with, e.g., carbon-carbon double bonds, >C = C<. Given this, a longer hydrophobic hydrocarbon chain also may reduce the rate of hydrolysis of the bond and extend shelf life, which is a known limitation of current silane systems [18]. One previous laboratory study has investigated MOS and it was reported that for bonding between lithium disilicate and resin cement, higher initial bond strength was observed compared to MPS [19,20]. It was explained that the difference might be related to a reduced availability of hydrolysable groups due to a longer hydrophobic hydrocarbon chain in MOS, which is the only structural difference with MPS. This might increase the difficulty by which water molecules pass through such a hydrophobic layer created by interfacial MOS and, hence, reduce the rate of hydrolysis of siloxane bonds. The finding reiterates that of an earlier study performed on silanes with a 10-carbon hydrocarbon chain which was able to better resist hydrolytic degradation than MPS [18]. Another study found that by virtue of a long hydrocarbon chain length, experimental resin composites synthesised with MOS exhibit lower viscosity while having a higher flexural modulus and strength and being less brittle than MPS [21]. This may be explained by the 'restrained layer theory', whereby silanes between the inorganic substrate and resin polymer matrix are capable of deformation and act as a stress breaker [22]. However, in that study, such differences were not statistically significant [21].
The objective of this current laboratory study was to further investigate MOS and its effect on an experimental resin composite when compared and contrasted with a traditional silane coupling agent, MPS. The hypothesis tested was that a longer chain silane such as MOS would lead to lower stiffness and higher strength with a difference in hardness at the same concentration as MPS. The null hypothesis was that MOS performs not differently than MPS in terms of mechanical testing, including the Young's modulus of bending (flexural modulus), maximum bending stress at maximum load (flexural strength), extension from preload at break, and microhardness.

Materials and Methods
Materials used in this laboratory investigation are presented in Table 1.

Preparation of Silanised Glass Powder Fillers
Two silanes, 3-methacryloxypropyltrimethoxysilane monomers, MPS (Sigma-Aldrich), and a long chain silane 8-methacryloxyoctyltrimethoxysilane, MOS (Shin-Etsu, Japan), were first activated. They were hydrolysed for 1 h at room temperature in a mixture of 95 vol% ethanol in deionised water at pH 4.0 using acetic acid [3].
An amount of 10.00 g of high fluoride-containing SiO 2 -based bioactive calcium glass powder fillers (Ø 1.0 μm, Schott, Germany) was immersed in 0.3, 0.6, 0.9, and 1.2 g of hydrolysed silanes of both types to produce eight groups of silanised glass fillers and then mixed in an ultrasonic cleansing apparatus. The solvent was allowed to evaporate and silanised glass fillers were further dried in a vacuum oven at 80 °C at 50 kPa pressure for 2 h. Next, they were kept in a desiccator before the next step.

Preparation of Resin Composite
An amount of 8.0 g of a monomer mixture consisting of urethane dimethacrylate and triethylene glycol dimethacrylate at an experimental 70:30 weight ratio (Esschem Europe, Seaham, UK) with camphoroquinone at 0.7 wt% (Aldrich, USA) and 2-(dimethylamino)ethyl methacrylate at 0.7 wt% (Aldrich, Germany) was mixed with 3.0 g of each of the silanised fillers to produce eight groups of experimental resin composites with a filler load of 72.7 wt% (SpeedMixer, FlackTek, USA). One resin ratio only was used in the current study. The resin composites were transferred to a lab curing unit (Targis Power, Ivoclar Vivadent), for photo-polymerization for 25 min, for the top and bottom directions. Then, they were stored in darkness for 24 h and thereafter carefully

Three-Point Bending Test
The three-point bending test (n = 8) was carried out in a universal testing machine (LR30K Plus, Lloyd, Ametek, USA). The distance between the two roller supports was set 20.0 mm and the cross-head speed was set 1.0 mm/min [23,24]. The Young's modulus of bending, maximum bending stress at maximum load, and extension from preload to break at the moment of breakage were recorded with software (NEXYGEN Plus, Lloyd, Ametek, USA). The flexural strength of a block with a rectangular cross-section is given by the formula [25]: where σ fs = flexural strength, F f = load at fracture, L = distance between support points, b = width of cross-section perpendicular to load, and d = width of cross-section parallel to load. The flexural modulus is given by the formula [26]: where E B = modulus of elasticity, L = distance between support points, m = force applied over deflection, b = width of cross-section perpendicular to load, and d = width of crosssection parallel to load.

Vickers Hardness Test
Next, the specimens were mechanically polished with 4000grit silicon carbide polishing paper under running water for 2 min before the surface microhardness test (Vickers hardness testing, VH; Duramin 5, Struers) with the following parameters: press load 2.94 N, dwell time 10 s, and 5 random indentations. The tested specimens were subsequently examined and observed with a 3D optical microscope (Contour GT-X, Bruker).

Statistical analysis
The data obtained were analysed statistically with SPSS 28.0 (IBM SPSS Statistics for Windows, Version 28.0. Armonk, NY; IBM Corp). One-way ANOVA, followed by the Tukey post hoc test was used to compare results of mechanical testing according to two independent factors: silane type and concentration. The significance level was set at α = 0.05.

Mechanical Test
The results of the three-point bending test are presented in Table 2 and Fig. 2. The highest flexural modulus was obtained in MPS with 3 wt% silane content: 11.6 ± 0.5 GPa. The lowest value obtained for MOS with 12 wt% silane content was: 8.4 ± 0.4 GPa.
Flexural strength was highest in specimens with 9 wt% MPS at 161 ± 14 MPa, while in MOS groups the highest value was obtained with 6 wt% silane was 160 ± 20 MPa.
A similar trend was observed for extension from preload at break where MPS with 9 wt% silane and MOS with 6 wt% silane exhibited greatest values, 0.53 ± 0.09 mm, and 0.59 ± 0.09 mm, respectively. The lowest extension was obtained in specimens with 3 wt% MPS: 1.36 ± 0.28%.
The highest hardness value was obtained with 6 wt% MPS, numerically closely followed by the same silane at 3 wt%, at 73.4 ± 4.6, and 72.6 ± 3.8, respectively. The MOS group with 12 wt% silane exhibited the lowest microhardness, 51.9 ± 2.4 VH.
Statistical analysis using two-way ANOVA revealed that both the silane type and concentration significantly affect flexural modulus (p < 0.001), flexural strength (p = 0.001), and microhardness (p < 0.001). Whereas, the extension from preload at break was only significantly affected by the silane concentration (p < 0.001).

3-D Optical Microscopy
3-D optical microscopy showed a uniform pattern of indentations created during VH measurements (Fig. 3). All indentations showed a uniform rhomboidal shape with equal diagonal lengths. MOS resin composites displayed a rougher surface around the indentation than MPS.

Discussion
The aim of this pilot study was to compare and contrast some important mechanical properties of experimental composites containing a novel long chain length silane, 8-methacryloxyoctyltrimethoxysilane (MOS), with 3-methacryloxypropyltrimethoxysilane (MPS) which is almost ubiquitously used in commercially available resin composites. The null hypothesis that MOS performs similarly to MPS with respect to the flexural modulus, flexural strength, extension from preload at break and microhardness, is rejected.
The three-point bending test is widely used for assessing denture base materials and resin composites [21,23,27,28]. Our current laboratory study samples have demonstrated mechanical properties with values at the same magnitude compared with an earlier study with a similar setup [21]. It was explained that the different properties obtained in the various silanes investigated depended on the structure of the silane molecule, whereby silanes containing a larger number of -CH 2 -linker units in a longer and unbranched alkyl chain had the potential to bend flexibly under stress between the rigid resin network and glass filler. This would allow more energy to be absorbed, resulting in higher fracture toughness [21]. Perhaps unsurprisingly, that study found that resin composites containing a long chain length silane exhibited a lower elastic modulus, while for higher flexural strength a lower concentration of long chain length silane was required for silanisation, compared to MPS. This is consistent with our current laboratory investigation using a different experimental resin formulation where a decreasing trend in the flexural modulus was observed with an increasing silane Another reason might be that greater hydrophobicity of a longer linker chain length of MOS, causes it to be less soluble in the silanisation solvent system. This could lead to a greater mass of silane being available for deposition on the glass powder surface during the silanisation of fillers, compared to if a more soluble silane with a shorter linker part was used [15]. One of the observations is that a further increase in the silane content beyond 9 wt% did not benefit the mechanical properties of the experimental resin composite. It is well reported that various silane concentrations form siloxane films of different thicknesses on a glass surface depending on the exposure time. One study has suggested that a monolayer is formed first [28] -while the innermost layer is strongly chemisorbed to the glass surface, and additional layers are weakly physisorbed and, thereby, more prone to removal [10,28]. Accordingly, an adequate concentration of silane according to the filler quantity and geometry should be used for effective silanisation and to subsequently produce a new resin composite with better mechanical properties. Proper silanisation coverage and coating of glass powder with silanes may also be ensured by using an ultrasonic bath. We also observed that resin composites behaved differently when containing MPS or MOS. Although the values were not significantly different, below 9 wt% of either silane, the peak values occurred at a lower concentration for MOS (6 wt%) compared to MPS (9 wt%), where a similar trend is observed for flexural strength. A greater ability to extend before fracture might be beneficial in resin composites whereby failure of the material could be subjected to more energy before catastrophic failure. By comparison, MOS could impart both higher flexural strength and extension from preload at break to a resin composite at a lower silane content (6 wt%) than if MPS was used (9 wt%).
Furthermore, long chain length silanes were associated with lower hardness on VH. In the current laboratory study, we found that experimental resin composites containing MPS exhibited greater hardness than MOS for the same silane content, by 14% on average. The VH values were greatest in the MPS group with 6 wt% silane (73.4 ± 4.6 VH). A further increase in the silane content reduced the surface hardness regardless of the silane type. This could probably be explained by an increase in flexibility offered by a longer molecular hydrocarbon body (of MOS) which contributed to lower resistance to permanent deformation. On the indentation images, however, differences between specimens with the different silane type or content were not as conspicuous, although there appears to be a rougher surface on specimens containing the longer chain silane monomer, and indentations seem to measure slightly wider with longer diagonals than those containing the conventional MPS. While indentation may provide some information regarding the mechanical property of the polymer matrix, other imaging methods, such as scanning electron microscopy and AFM, might give a deeper insight into the differences at the filler-matrix interface depending on the different silane type and concentration by observing the fracture surfaces.
In terms of water storage, Maruo et al. observed that MOS exhibited greater initial bond strength than MPS and explained that this might be due to longer hydrophobic hydrocarbon chains in MOS which increased the resistance to hydrolytic degradation [19]. Moreover, a longer hydrocarbon spacer might impart greater flexibility to resin composite, a similar finding to a long chain length silane being tested for bond strength between lithium disilicate glass ceramic and resin cement [19]. In addition to that, MOS could achieve similar, equal flexural strength at a lower concentration than if MPS was used. Clinically, resin composite manufactured with MOS as the silane coupling agent might exhibit different handling properties and a surface texture due to the increased flexibility and lower surface hardness. This might make it easier to finish but, however, it might become less resistant to wear.
It is noteworthy that the filler glass in our study is a high fluoride-content SiO 2 -based bioactive calcium glass with the particle size of ~ 1 μm. It is understood that these glass particles act as a dispersed phase in resin composite and when blended in resin composites they would be regarded as 'micro-filled', exhibiting high compressive strength, hardness, and elastic modulus when compared and contrasted with unfilled counterparts [29]. The mechanical test results obtained in this pilot study are based on a calculated silane content coupled with the same filler glass with a specific geometry and surface area. Given that, differences in the test results could be attributed to the silane type and concentration only. Moreover, modifying the glass filler characteristics, such as the use of finer glass powder, or even changing the geometry by incorporating glass fibres, affects the available surface area for bonding and would necessitate further research with regards to the silane content and resin composite preparation [28]. In the current study, high fluoride-containing SiO 2 -based glass fillers were used. This type of filler was expected to exhibit some form of bioactivity such as fluoride-releasing or even cariostatic activity (but this aspect was not studied this time) [30]. In the future, it might be possible to develop resin composite with this novel silane coupling agent which might modify the capability of fluoride leaching ability. Statistically, a power analysis was not carried out; however, statistically different mechanical behaviours could be shown according to the number of specimens (n = 8) in each experimental group.
As this pilot study is a cross-sectional investigation with a single timepoint only for evaluation, the long-term effects after artificial aging on the mechanical properties of the experimental resin composite merit further studies. Nevertheless, this time the storage time in this pilot study was limited to 24 h of dry storage only. The influence of the storage time might also be investigated as it relates to the long-term hydrolytic stability of siloxane bonds and, hence, the effect on the mechanical properties. It is known that silane-based bonding is susceptible to hydrolytic degradation which may cause the mechanical properties of resin composite to deteriorate. Laboratory experiments involving hydrophobicity tests, storage under water or in conditions simulating the human oral cavity, as well as artificial ageing might provide greater longitudinal insight to the stability of the bonds in a moist environment. In addition, the inclusion of some other trialkoxysilane monomers to create a blend, such as also using a cross-linking silane, and varying the resin and filler composition might reveal a better match with such long chain silanes as is MOS.

Conclusion
A trialkoxysilane with a longer hydrocarbon linker part confers lower stiffness, higher strength, higher extension from preload to break, and lower microhardness to an experimental resin composite compared to the trialkoxysilane with a shorter chain silane. By incorporating a long chain silane at different silane contents, the mechanical properties of resin composite might be optimised for applications requiring lower microhardness at similar stiffness and unchanged flexural strength.