Plasma Treated AR-Glass Fibres in Experimental Reinforced Composites with Three Silanes: A Study on Mechanical Properties

To compare and contrast mechanical properties of experimental alkali-resistant (AR) glass fibre-reinforced composites (FRCs) silanised with one of three functional trialkoxysilanes: 3-methacryloxypropyltrimethoxysilane (3-MPS), 8-methacryloxypropyltrimethoxysilane (8-MOS) and 3-acryloxypropyltrimethoxysilane (3-APS). The experimental AR-FRCs were silanised with or without plasma pretreatment. Continuous unidirectional AR-glass fibres (ARcoteX® 5326 2400tex, Owen Corning) were cleansed ultrasonically, treated either with or without cold plasma (Piezobrush® PZ2, Relyon Plasma, Germany), then silanised by immersion for 10 min in activated, hydrolysed silane solution containing either 3-MPS, 8-MOS or 3-APS. The fibre bundles were next air-dried (24 h), and dried in vacuum oven (80 °C, 5 kPa, 2 h), producing 6 types of silanised AR-glass fibres. Next, silanised AR-glass fibres were laminated with a resin matrix, bis-GMA:TEGDMA at 60:40 wt%, CQ at 0.7 wt%, DMAEMA at 0.7 wt% for 10 min and paired randomly in rovings of two in stainless steel moulds (2 mm × 2 mm × 25 mm) to prepare an AR-FRC beam specimen (n = 16). All specimens were light-cured from the top and bottom directions for 60 s (Elipar S10, 3 M ESPE). Half of the samples (n = 8) underwent accelerated artificial ageing by immersion in boiling water (100 °C, for 16 h). Finally, 12 experimental AR-FRC groups were produced. All specimens were subjected to the three-point bending test with the upper surface (facing curing light) towards the cross-head. After testing, fibres, silanised and non-silanised, were inspected under a scanning electron microscope (SU1510, Hitachi, Tokyo, Japan). Statistical analysis was performed with three-way ANOVA and the Tukey post hoc test at α = 0.05. The flexural modulus, flexural strength, and fracture work of the materials were significantly affected by silane type (p < 0.05) and artificial ageing (p < 0.001). Plasma treatment on AR-glass fibres significantly reduces flexural modulus of FRCs (p < 0.001) and slows the plasticising effect of artificial ageing on FRCs. Long chain silanes impart higher flexural strength and lower flexural modulus to AR-FRCs. While artificial ageing halves flexural strength and plasticises them, plasma surface pre-treatment of AR-glass fibres before silanisation reduces the plasticising effect.


Introduction
Fibre-reinforced composites, FRCs, can be alternatives to composites containing particulate fillers in dentistry.FRCs are being used in a number of dental applications, including (but not limited to) direct restorative materials such as flowable composites [1], fixed and removable dental prostheses [2,3] and root canal posts [4].A wide range of mechanical properties is possible by virtue of different fibre size, fibre composition, length and design.In particular, silanes play a vital role in load transfer from the weak resin matrix to the stronger glass fibre.In glass fibre-reinforced composites, the layer of the interfacial siloxane film is vital because it distributes stress peaks during loading owing to its elasticity as explained in the restrained layer theory [4,5].
In essence, FRCs are composed of three main composite components: glass fibres, resin matrix, and a silane coupling agent.Silane chemically binds the fibre and matrix Extended author information available on the last page of the article together.Regarding glass fibres alone, numerous studies have been conducted in relation to the fibre length, size or aspect ratio [1,6], fibre orientation [7], and fibre volume content [8], just to name a few.FRCs have also been investigated against prolonged water treatment which was shown to have a plasticising effect for FRCs [6,9], and could also lead to water sorption and over time, hydrolytic degradation [9][10][11].Humidity and water are most detrimental at the polysiloxane network in the resin-glass fibre interface, and, hence, the polysiloxane network that is related to the silanes that make up this interface.
Silane coupling agents are technically important synthetic bi-functional hybrid compounds consisting of unique ≡C-Si≡ bonds that enable the union of hydrolysable alkoxy groups and organofunctional entities.Silanes create chemical adhesion between dissimilar materials, such as organic and inorganic matrices [12].Silanes are ubiquitously used in modern dental resin composites, and for etched ceramics and silica-coated titanium and some other metals [4], and even as a sizing agent of glass fibres [13] in fibre-reinforced applications.Different types, formulations, and handling procedures of silanes have been compared and assessed in various laboratory studies.Silanes blended with a nonfunctional cross-linking silane, the so-called bis-silane [14], having a different hydrocarbon linker chain length [15][16][17], prepared as formulations combining (blending) various silanes [18,19] and at different concentrations [20] have been reported.Information on a range of advanced silane applications, such as amalgam repair by bonding to resin composite [18], and resin-titanium adhesion [20][21][22] have been researched.
In general, addition of glass fibres to a resin matrix significantly improves the mechanical properties of resin composites [23].A number of glass fibres with different compositions have been used in FRCs, particularly E-glass and S-glass [4,24].Reports on dental FRCs containing alkaliresistant (AR) glass fibres, however, are relatively scarce.One laboratory study has demonstrated that such glass fibres in a bis-GMA-TEGDMA resin monomer system could increase flexural strength [23].
In the current laboratory study, alkali-resistant AR-glass fibres containing a minimum of 16% ZrO 2 [25] were used.These fibres are advocated for their durability, high mechanical strength and good impregnation with polyester and epoxy vinyl ester resins [26].
Plasma surface treatment is widely used in industry, including varnishing, printing, and coating [27] to increase adhesion durability.It is a relatively new technique introduced to dentistry with limited applications to date [28][29][30].Plasma is fully or partially ionised gas containing electrons, ions and radicals produced by passing a current through a dielectric gas [31].Plasma is generated by, e.g., corona discharge, dielectric barrier discharge, and cold plasma torch [32].Piezoelectric direct discharge (PDD) utilises the ability of piezoelectric materials, such as PZT, to transform low supply voltage to high output voltage, thereby creating plasma [31].
While plasma treatment has already been reported on dental zirconia [29,30] and glass fibres in FRCs [28], resin composites containing AR-glass fibres treated with cold atmospheric pressure plasma by PDD has not yet been investigated.
The objective of this laboratory study was to research mechanical properties and characterize the surface of AR-FRCs containing one of the three silanes: 3-MPS, 8-MOS or 3-APS, as well as to examine the effect of fibre surface plasma treatment and accelerated artificial ageing.The hypothesis was that AR-FRCs would behave mechanically differently when AR-glass fibres are silanised with different silanes due to a longer hydrocarbon linker chain of 8-MOS and different functional groups of 3-APS when compared with the gold standard 3-MPS.The null hypothesis was that AR-FRCs perform similarly regardless of silane type, glass fibre surface treatment, and accelerated artificial ageing.

Materials and Methods
Materials used in this laboratory investigation are presented in Table 1.
Continuous unidirectional AR-glass fibres (ARcoteX® 5326 2400tex, Owen Corning) were first cleansed ultrasonically.Immediately before silanisation, half of the fibres were treated with cold atmospheric pressure plasma at 1 mm distance for 30 s according to the manufacturer's instruction (Piezobrush® PZ2, Relyon Plasma, Germany).Glass fibre silanisation was achieved by immersion into the three hydrolysed silane mixtures for 10 min, air drying for 24 h and further dried in a vacuum oven (80 °C, 50 kPa, 2 h) to produce six different group of silanised glass fibres.

Preparation of Experimental AR-Glass Fibre-Reinforced Resin Composite
A monomer mixture (i.e., the resin matrix) was prepared by mixing from the following constituents: bis-GMA and triethylene glycol dimethacrylate at a 60:40 weight ratio, and camphoroquinone at 0.7wt% (Aldrich, USA) and 2-(dimethylamino)ethyl methacrylate at 0.7wt% (Aldrich, Germany).Silanised AR-glass fibres in single-end rovings were resin-impregnated for 10 min by immersion in the resin mixture.A custom made stainless steel mould (2 mm × 2 mm × 25 mm) was used to build up rectangular beams of FRCs with two 25 mm rovings of silanised AR-glass fibres in the monomer mixture (n = 16) and photopolymerised for 60 s from both sides (Elipar S10, 3 M ESPE) [34].Half of the samples of each group were randomly selected and they underwent accelerated artificial ageing by immersion in boiling water at 100 °C for 16 h continually [35].Subsequently, 12 groups of samples (a total of 96 samples with 8 per group) of ARglass fibre-reinforced composites with a fibre load of 44.8 vol% / 65.7 wt% were created.A flowchart illustrates the preparation of treatment groups (Fig. 2).

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 [34,36].The Young's modulus of bending, maximum bending stress at maximum load, work from preload to   [25] 27 μm diameter [26] ARcoteX® 5326 Tex number = 2400 Linear density 2400 g km −1 Density 2.68 g cm −1 maximum load, and work from preload to the moment of breakage, were recorded with software (NEXYGEN Plus, Lloyd, Ametek, USA).The difference in work of fracture energy was obtained by subtraction of work from preload to break from work from the preload to the maximum load [37]: Dif ference in work of fracture energy = (work from preload to maximum extension) − (work from preload to maximum load)

Surface Characterisation with SEM
Tested, fractured samples of AR-FRCs were examined and observed with a scanning electron microscope after preparation with Pd/Pt coating (SU1510, Hitachi, High-Technologies Corporation, Tokyo, Japan).The voltage was 15.0 kV, working distance 18-22 mm, and magnification 100X to 2000X.

Statistical Analysis
The data obtained were analysed statistically with SPSS 28.0 (IBM SPSS Statistics for Windows, Version 28.0.Armonk, NY).Normality of data was tested with Shapiro-Wilk test.Three-way ANOVA, followed by the Tukey post hoc test were used to compare and contrast results of mechanical testing, according to three independent factors: silane type, fibre surface treatment and artificial ageing.The significance level was set at α = 0.05.

Three-Point Bending Test
The results of the three-point bending test are presented in Table 2 and Figs. 3 and 4.  Flexural strength was the highest in specimens without accelerated artificial ageing in the 3-APS plasma-treated and 3-MPS not plasma-treated groups at 591 ± 32 MPa and 589 ± 41 MPa respectively.Aged specimens displayed only approximately half of the flexural strength than their nonaged counterparts, regardless of silane type and plasma treatment (p < 0.001).Comparison between silanes reveal that AR-glass fibres silanised with 8-MOS produce FRCs with significantly higher flexural strength than 3-MPS and 3-APS (p = 0.039 and p = 0.001 respectively).
The highest flexural modulus was obtained in the 3-MPS group without plasma treatment before artificial ageing at 21.4 ± 2.4 GPa, while the lowest value of 14.3 ± 2.8 GPa was obtained with 8-MOS with plasma treatment which underwent artificial ageing.Additionally, plasma treatment on AR-glass fibres significantly lowers flexural modulus (p < 0.001).Excluding the 3-MPS FRC group without plasma treatment and artificial ageing, all remaining groups showed no significant differences between each other (p > 0.05).In general, 8-MOS produces FRCs with significantly lower flexural modulus than those containing 3-MPS (p = 0.041).
The difference in the work of fracture energy represents fracture behaviour of the tested material where a higher value describes increasingly plastic behaviour [37].All groups of artificially aged samples showed higher values than their nonaged counterparts.Specimens that were not plasma treated and underwent accelerated artificial ageing exhibited the greatest difference in work of fracture energy, first the 3-APS group (14.6 ± 5.4 N cm), followed by the 3-MPS specimens (13.8 ± 2.9 N cm).When examined in combination, 3-APS was found to have a significantly greater difference in work of fracture energy than the 8-MOS group (p = 0.023).In aged specimens, AR-FRCs containing plasma-treated glass fibres showed the greatest work of fracture energy when 3-APS was used (13.3 ± 5.2 N cm), comparable with non-plasma treated blocks after artificial ageing.Three-way ANOVA showed that plasma treatment has a significant interaction effect with artificial ageing (p < 0.001).This is further illustrated by post hoc tests that, once AR-glass fibres have undergone plasma treatment, accelerated artificial ageing no longer had a significant effect on differences in the work of fracture energy of all silane groups, 3-MPS (p = 0.024), 8-MOS (p = 0.006) and 3-APS (p < 0.001), compared with samples containing non-plasma treated AR-glass fibres (p > 0.05).
Normality of data was confirmed with the Shapiro-Wilk test (p > 0.05).Statistical analysis using three-way ANOVA with silane type, plasma treatment and artificial ageing as covariates revealed that flexural modulus, flexural strength and difference in the work of fracture energy were significantly affected by silane type (p < 0.05) and accelerated artificial ageing (p < 0.001) (Table 3).
Stress-extension of each experimental group is plotted in Fig. 4. All groups produced the characteristic steady increase in stress with extension which then drops abruptly as extension further increases.

Surface Characterisation with SEM
All specimens from the 3-point bend test underwent incomplete fracture and remained as one piece at the end of the test.In order to visualise the fracture surfaces on the tension side of AR-FRC beams, specimens were separated and oriented for SEM observation.As a result, fracture surfaces were only observed on the tension half of the fracture plane along the direction of the beam while the compression side was observed perpendicular to the beam direction.Representative SEM images are seen in Fig. 5.
Examination under SEM revealed similar appearances amongst experimental AR-FRC groups.Fibres were evenly distributed throughout the resin matrix (Fig. 5a).Most fibres fractured close to the plane of fracture with few fibres pulled out (Fig. 5b) and resin tags remain on the surface of fibres after fracture (Fig. 5e, f), indicating good adhesion between silanised fibres and the resin matrix.The mode of failure of AR-FRCs under tensile stresses of the 3-point bend test appeared to be a mixture of cohesive and adhesive failures occurring simultaneously in the resin and/or AR-glass fibres (Fig. 5a).The quantity of resin tags observed on glass fibre surfaces was reduced with artificial ageing (Fig. 5e vs. h).

Mechanical Properties
Flexural strength is a measure of the stress at the point of breakage ('strength') determined in a bending test, in this case, a three-point bending test [38].In this research, flexural strength was halved in all groups after hydrothermal accelerated ageing.This observation resonates with other studies involving resin composites [35] and could be explained by the induction of hydrolysis of interfacial silane between the resin and the fillers [35,39] which, in our case, were AR-glass fibres.The loss of interfacial bond strength could be regarded as the major factor in a reducing of mechanical properties in artificially aged FRC specimens [39], and this could be confirmed microscopically with the reduced quantity of resin tags remaining on AR-glass fibre surfaces under SEM.
As for comparison amongst silanes, FRCs prepared with a longer chain length silane, 8-MOS, performed better in terms of exhibiting higher flexural strength than both 3-MPS and 3-APS, which had a hydrocarbon chain length of only three units, in agreement with a recent finding using particulate fillers [17].Differences in the FRCs' mechanical properties could arise from the different trialkoxysilanes used in the composite preparation process.3-MPS is a commonly used silane found in many commercially available products and has been regarded as a 'gold standard', and evaluated in numerous studies [17,[20][21][22]33].Interestingly, 8-MOS consists of the same functional groups, a trimethoxysilane at one end and methacrylate on the other, except for having a longer hydrophilic hydrocarbon linker chain with 8 -CH 2 -units [17], while 3-MPS only has three -CH 2 -hydrocarbon chain units [15,16].However, 3-APS is another type of silane with a different end group, a very reactive acrylate, without a protective methyl group, -CH 3 .3-APS has been shown to exhibit high shear bond strength in resin-Ti bonding, and in amalgam repair because of a more reactive acrylate group than the methacrylate group as the organofunctional entity [14,18,40].
These research results reflected that extra energy was required to cause breakage of resin composites silanised if a long linker chain silane was selected, regardless of the organofunctional group.This also confirms the pivotal role of silane coupling agents on adhesion between the resin and filler component: without silanisation the glass fibres would act as voids and weaken the FRC [39].
A higher flexural modulus represents lower elasticity of a tested material [38].In this study, FRCs silanised with 8-MOS produced significantly lower flexural modulus than when 3-MPS was used, which is in agreement with studies investigating long chain silanes [15,17].The experiment also revealed that AR-glass fibres treated with plasma significantly lowered the flexural modulus of FRC.On the other hand, it has been shown that atmospheric pressure low-temperature plasma is capable of improving adhesion between zirconia and resin cement with similar efficacy as treatment with alumina powder grit-blasting [30].Similarly, it is possible that silane adhesion enhanced by plasma treatment on zirconia-containing AR-glass fibres would promote stress transfer between the resin matrix and AR-glass fibres via the interfacial silane layer.This would permit the silane to impart greater elasticity to FRC.The difference in the work of fracture is calculated between work from preload to the maximum extension and the work from the preload to the maximum load [37].The rationale behind computing this difference in work is that a higher value would indicate additional energy to result in complete fracture of the material, i.e., the material would be safer to use and immediate fractures would be avoided.A higher value also reflects increased plasticity of the material.We know that fracture energy is the sum of surface energy and plastic work done during the fracture process: where Γ = fracture energy, 2γ = energy of newly created surfaces and w = work done in plastic deformation.Whereas, the newly created surfaces might be similar in similarly prepared AR-FRCs, varying work done in plastic deformation might occur because of silane type, plasma and accelerated ageing treatment.
In the current study, all hydrothermally aged samples exhibited a greater difference in work than non-aged FRCs, confirming the plasticising effect on FRC upon exposure to water, and in this very case we applied accelerated artificial aging by using boiling water [9,39].The graphs in Fig. 4  as freshly prepared FRC samples started to break abruptly after 500 MPa of stress is applied, artificially aged FRCs withstood no more than 350 MPa of stress, yet they were subject to a greater extension as measured stress decreases concomitantly.
When considering differently silanised FRCs, the 3-APS group exhibited a greater difference in work than the other two silane groups.This might be contributed by the acrylate functional group which distinguishes it from the methylcontaining silanes, viz.3-MPS and 8-MOS.Most interestingly, plasma treatment and accelerated hydrothermal ageing appeared to interact in the plasma-treated samples, which were less affected by artificial ageing treatment by reducing the gap of the difference in work of fracture energy between aged and the corresponding non-aged groups.Indeed, the plasticising effect of hydrothermal treatment is due to hydrolysis of the adhesive siloxane interface over time [35].Plasma treatment could have possibly strengthened the chemical connection between silane and AR-glass fibres, thus, delaying the hydrolysis and slowing the ageing process.The bond strengthening process might be contributed by the following means: first of all, plasma produces a cleaning effect primarily by removal of organic contaminants, more specifically, achieved by removal of carbon in such organic contaminants [29,41].Plasma interacts with the top 10 nm thin layer of a material surface [41] producing several effects, such as reducing metal oxides, increasing wettability by deposition of hydroxyl groups [42], reducing the contact angle, and increasing surface free energy [43] and even increasing surface roughness, compared with grit-blasting [29].Moreover, advantages of PDD plasma generation used in this study include high feasibility in clinical or chairside environments, because cold atmospheric pressure plasma is generated at < 50 °C [27].It is also considerably less messy than other chemical or physical surface treatment methods, such as acid-etching [1] and grit-blasting [30], and in the case of resin bonding, does not require additional silica-coating or special primers [44,45].In the case of zirconia substrates, cold plasma at atmospheric pressure does not affect the crystal structure, unlike grit-blasting [30].The current results suggest that plasma treatment might not only improve adhesion, but can lessen the detrimental effect of ageing on the mechanical properties of FRCs.More investigation with other blends of silanes and glass fibre varieties on a microscopic level would be warranted to explain this phenomenon.
This study comprised a single resin formulation and one fibre material.The effect of other silane blends and varieties, such as cross-linking silanes, on plasma pre-treated glass fibres is yet unknown.Furthermore, this is the first study to investigate the effect of atmospheric pressure-low temperature plasma treatment on FRC and a single application of plasma on AR-glass fibres.Moreover, it has shown to modify the mechanical properties and limit plasticity of FRCs caused by accelerated hydrothermal ageing.Other effects of plasma treatments would merit further research as well.
Integration of these findings would be beneficial in developing FRCs with greater flexural strength that resist hydrolytic degradation better.

Conclusion
Silanes with longer hydrocarbon chains might increase flexural strength and reduce the flexural modulus of FRCs.Accelerated artificial hydrothermal ageing predictably halves flexural strength.Atmospheric pressure-low temperature plasma pre-treatment of glass fibres potentially improves and modifies mechanical properties over ageing.

Fig. 2
Fig. 2 Sample preparation with AR-glass fibres, followed by artificial ageing or no ageing treatment

Fig. 3 Fig. 4
Fig. 3 Mechanical test results of experimental groups according to the silane type, plasma treatment and accelerated artificial ageing.Upper left: Flexural strength; upper right: flexural modulus; lower left: difference in work of fracture energy.Error bars: standard deviation

Fig. 5
Fig. 5 SEM micrographs of AR-FRCs after the 3-point bend test.a: A non-aged, plasmatreated specimen containing 3-APS and showing failure of AR-FRC block as a result of three events: cohesive failure within fibre (black arrows), cohesive failure within resin (white arrows), and adhesive failure between fibre and resin (hollow arrows); b: fracture surface of a similar block containing 8-MOS showing majority of fibre fractures occur at short distances from the resin matrix; c and d: view from compression side (towards cross head) showing majority of fibres fracturing at short distance from resin breakage; e and f: resin tags (white arrows) remaining on fibre surface of non-aged 3-APS samples; g and h: fewer resin tags (white arrows) remain attached to fibre surface in artificially aged samples (8-MOS and 3-MPS respectively)

Table 1
Materials used in this study

Table 3
Results of three-way ANOVA investigating the effects of silane type, plasma treatment and accelerated artificial ageing on mechanical properties of AR-FRC.df: degree of freedom; F: F statistic *Asterisks mean the factors are considered together, which is the usual way of presenting results of the ANOVA statistical analysis