Sustainable composite materials based on ethylene-vinylacetate copolymer and organo-modified silica

2.2.1 Preparation and modification of nanosilica:

SNPs were synthesized from TEOS and ecofriendly solvents by the following process: 50 ml TEOS was dissolved in 300 ml absolute ethanol under continuous stirring for 30 min. Deionized water (10 ml) was then gradually added into this solution and stirring was continued for another 30 min. The solution was adjusted to alkaline pH of 8.5 by dropwise addition of 25% NH₄OH solution. After 30 min of magnetic stirring, clear silica sol was formed. SNP powder was obtained by evaporation of this silica sol at room temperature for 48 h. The silica powder was dried in a circulating hot air oven at 150°C for 5 h and in vacuum at 120°C for another 5 h until constant weight was obtained [11]. Several batches were made using this process for enough SNP volume for sample preparation later. The dried SNP was milled, sprayed and kept in a desiccator for later use.

The SNP was modified by APTES with amounts of 5 wt.%, 10 wt.%, 15 wt.%, 20 wt.% and 50 wt.% (compared to original SNP weight) by the following process: APTES was dissolved in absolute ethanol at 70°C for 60 min, then a stoichiometric distilled water and certain weights of SNP were added into this solution under continuous stirring for 2 h. After filtering and washing with absolute ethanol several times, the obtained powder was dried in a vacuum oven at 80°C for 4 h [10]. The final products were denoted as m5-SNP, m10-SNP, m15-SNP, m20-SNP, and m50-SNP for the above APTES contents, respectively. SNPs and m-SNPs mean unmodified and modified SNPs, respectively.

2.2.2 Preparation of EVA/SNP and EVA/m-SNP nanocomposites:

EVA/1.5 wt.% unmodified SNP and EVA/1.5 wt.% modified SNPs nanocomposites (compared to EVA weight) were prepared by melt mixing in a Thermo-Haake intermixer (Germany) at 160°C, with a rotor speed of 50 rpm for 6 min. After melt mixing, the molten products were hot pressed at 160°C under pressure of 5–6 MPa for 5 min to obtain sheets of the nanocomposites with suitable thickness. The EVA/SNP and EVA/m-SNP nanocomposites using m5-SNP, m10-SNP, m15-SNP, m20-SNP, and m50-SNP were coded as EVA/SNP, EVA/m5-SNP, EVA/m10-SNP, EVA/m15-SNP, EVA/m20-SNP, and EVA/m50-SNP, respectively.

3.4.1 FTIR analysis of EVA, EVA/SNP and EVA/m15-SNP nanocomposites before accelerated weather testing

As discussed in previous studies [7], [19], the hydrogen bonds in EVA/SNP nanocomposites can be characterized by FTIR spectra analysis. As is well known, it is hard to detect the shifts of all peaks of the nanocomposites in comparison to the reference component materials. In fact, only some specific bands with distinguishable changes can be taken into account.

Figure 3 shows the FTIR spectra of EVA and EVA/SNP, EVA/m15-SNP nanocomposites and Table 3 lists some specific bands in spectra of initial EVA, nanosilica, and the nanocomposites. It can be clearly seen that the stretching vibration absorption bands (ν) for the C=O groups at around 1737 cm⁻¹ in EVA and the nanocomposites are not significantly changed. This is due to the numerous C=O groups existing in the nanocomposites and only their small part is available to make interaction with SNPs. Unfortunately, there is no observable peak attributed to the absorption of amino groups in spectra of the nanocomposites. A specific band attributed to ester stretching vibration of C-O-C (ν) at 1135 cm⁻¹ does not appear in the spectra of EVA/SNP and EVA/m-SNP nanocomposites because of the overlap with stretching vibration of Si-O groups at 1100 cm⁻¹ of silica. The wavenumber of the Si-O band is moved to a higher wavenumber at about 1122–1124 cm⁻¹ in spectra of the nanocomposites, corresponding to the shifts of 22–24 cm⁻¹. There are also small changes of bending vibration (δ) of Si-O groups in spectra of the nanocomposites compared to that of pure silica, corresponding to the shifts of 7–10 cm⁻¹. From the above wavenumber shifts, it can be suggested that there are hydrogen bonds between SNP or m-SNP and the EVA matrix. Theoretically, a proposal scheme of hydrogen bonds between EVA and SNP-modified by APTES can be illustrated in Figure 4.

Figure 3:

Fourier transform infrared (FTIR) spectra of initial ethylene vinyl acetate copolymer (EVA), EVA/silica nanoparticle (SNP) and EVA/m15-SNP nanocomposites.

Figure 4:

Proposal scheme of hydrogen bonds between ethylene vinyl acetate copolymer (EVA) and silica nanoparticle (SNP) modified by 3-aminopropyltriethoxysilane (APTES).

3.4.2 FTIR analysis of EVA, EVA/SNP and EVA/m15-SNP nanocomposites after accelerated weather testing

Thermo- and photo-oxidation of EVA in accelerated weather testing leads to formation of several photo-oxidation products that can be characterized by the FTIR technique [22]. The FTIR absorbance spectra of EVA, EVA/SNP and EVA/m15-SNP nanocomposites are illustrated in Figure 5, where the bottom and upper curves are spectra of the samples before and after accelerated weather testing, respectively. It is clear that the C=O stretching peak areas of the samples after accelerated weather testing are broadened. An increase of intensity of carbonyl absorption band at around 1739 cm⁻¹ is also observed (the upper curves). This is related to the formation of new carbonyl groups under the effect of UV irradiation, heating and atmosphere oxygen [23].

Figure 5:

Fourier transform infrared (FTIR) spectra of: (A) ethylene vinyl acetate copolymer (EVA), (B) EVA/silica nanoparticle (SNP) nanocomposites and (C) EVA/m15-SNP nanocomposite for 0 h, 72 h, and 168 h of accelerated weather testing.

For quantitative comparison, Table 4 represents the results of CI evaluated from the analysis of FTIR spectra of EVA and the nanocomposites. The CI value reflects the relative amount of C=O which existed in the samples. The results show that CI values of the samples before accelerated weather testing are approximately equal to each other. After accelerated weather testing, CI values of investigated samples are increased with increasing testing time. The CI value of the EVA/SNP sample after 72 h and 168 h of testing is lower than that of EVA. In our previous study, it had been demonstrated that nanosilica could limit the photo-degradation of EVA under UV exposure [8]. In this study, a special attention is paid to the effect of SNP modified by APTES on the CI value of the EVA/m-SNP nanocomposites. As expected, the CI value of EVA/m15-SNP nanocomposite after 72 h and 168 h of accelerated weather testing is slightly lower than that of EVA/SNP nanocomposite. This reflects the lower concentration of C=O groups formed by thermo- and photo-oxidation for the EVA/m-SNP nanocomposites, because APTES can enhance the adhesion between SNP and EVA matrix through hydrogen bonds, leading to limitation of the penetration of oxygen into the nanocomposites.

3.4.3 Morphology of EVA/SNPs nanocomposites after weather testing

Figure 6 shows the micro cracks that developed on the surface of EVA/SNP and EVA/m-SNP nanocomposites after accelerated weather testing. It is easy to see that the micro cracks of the exposed EVA/SNP nanocomposite are deep, while the micro crack of exposed EVA/m-SNP nanocomposite is shallow. These mean that the APTES modification of nanosilica can reduce the development of the crack tip under the weathering test. It is worth mentioning that in our previous work [8], the long and deep micro cracks had been formed on the surface of EVA after 168 h of accelerated weathering [8]. In the case of polymers with a semicrystalline structure (like EVA), cracking of these polymers is due to internal stress distribution in crystalline and amorphous areas. Since degradation occurs mainly in the amorphous region, the neighboring crystalline areas contribute to stress development around the crack tip, causing orientation of EVA macromolecules. For completely amorphous specimens, a mechanism of the above orientation does not exist; therefore, the samples degrade uniformly [13]. In our study, m-SNP increases the extent of the amorphous phase and this may be a reason for a shallow crack tip developing in EVA/m-SNP nanocomposite. Another reason is that m-SNP can adhere with EVA matrix more strongly than unmodified SNP.

Figure 6:

Field emission scanning electron microscopy (FESEM) images of: (A) ethylene vinyl acetate copolymer (EVA)/silica nanoparticle (SNP) and (B) EVA/m15-SNP nanocomposites after 168 h of accelerated weather testing.

3.4.4 Mechanical properties of EVA, EVA/SNP, EVA/m-SNP nanocomposites before and after accelerated weather testing

Table 5 shows initial values of tensile strength and elongation at break of EVA, EVA/SNP, and EVA/m-SNP nanocomposites (SNP and m-SNP were used with 1.5 wt.% compared with EVA) before accelerated weather testing. With introducing unmodified SNP into EVA matrix, the tensile strength of EVA increases slightly from 16.94 MPa to 17.23 MPa. When using m-SNPs, the tensile strength of EVA/m-SNPs nanocomposites is higher than that of EVA/SNP nanocomposite and increases with increasing the APTES modifier content up to 20 wt.%, then it tends to decrease for EVA/m50-SNP nanocomposite (50 wt.% APTES modifier). The enhancement in tensile strength of EVA/m-SNPs nanocomposites can be explained by contribution of APTES modifier grafted on the surface of nanosilica. In our previous studies [9], [10], the grafted APTES weights on the silica surface were increased as 2.66 wt.%, 5.54 wt.%, 6.89 wt.% and 7.52 wt.% when modifying with APTES contents of 5 wt.%, 10 wt.%, 15 wt.%, and 20 wt.%, respectively (all were compared to silica weight). The filler/matrix interaction contributed by APTES coupling agent can be attributed to the formations of hydrogen bond interaction, as illustrated in Figure 4, and molecular interdiffusion interaction between grafted APTES molecules and EVA macromolecules. Therefore, stress transfer at the particle/matrix interfaces is more efficient in the nanocomposites using m-SNP, and tensile strength of nanocomposites using m-SNP is higher than that of nanocomposites using bare SNP. Unfortunately, when modifying with APTES content larger than 15 wt.%, although the grafted weight increased, the higher concentration of silanol that was catalyzed by higher amino group concentration, caused the agglomeration of silica particles, which reduced the effect of APTES on tensile strength of nanocomposites [3]. In contrast, it is observed that elongation at break of EVA decreases with increasing the APTES coupling agent used. This is due to the greater the effective stress transfer, the less elongation at break is gained. As a result, break of nanocomposites occurred sooner. When APTES coupling agent content is larger than 15 wt.%, some defects can be formed in the nanocomposites because of agglomerated m-SNP, and a much higher decrease in elongation at break is observed.

Table 5 also displays retentions of tensile strength, and elongation at break of EVA, EVA/SNP, and EVA/m-SNP nanocomposites after accelerated weather testing for two periods (72 h and 186 h). It is clear that the retentions of tensile strength and elongation at break of the nanocomposites after testing periods are higher than those of EVA. The enhancement effect of SNP on the retentions of tensile properties of EVA can be explained by the limitation of degradation of SNP for EVA and heat dissipation under testing conditions. This result is in good agreement with our previous studies, in which commercial SNP (average particle diameter of 12 nm) had been used with SNP content which varied from 2 wt.% to 5 wt.% [8]. In this study, SNP was only used with small content of 1.5 wt.% for studying the effect of APTES modifier on tensile properties retentions of EVA. As expected, the retentions of tensile strength and elongation at break of nanocomposites using m-SNP are improved. The remarkable improvement in tensile properties retentions is observed for the nanocomposite using m15-SNP (1.5 wt.% SNP modified by 15wt.% of APTES). It can be suggested that there is an optimal dosage usage of APTES weight for modifying surface of SNP, because of the limited amount of surface silanol for a certain bare silica weight. Shortage of usage of APTES would cause low hydrophobicity for SNP. Exceeding the amount would cause unwanted effects, such as agglomeration of silica, and hydrogen bonds between amino groups on m-SNP surface. In general, better filler/matrix interaction may provide the better restriction of oxygen and moisture permeation into the nanocomposites, which reduces the thermo-, photo-oxidation and scission reaction of EVA macromolecules. For this reason, adding m-SNP to EVA can limit the reduction in tensile strength and elongation at break of EVA/m-SNP nanocomposites under accelerated weather testing. Remarkably, a nanocomposite using m15-SNP, that was modified by APTES with 15 wt.%, may be an optimal dosage of APTES coupling agent, which possesses the highest weather resistance.

When using m-SNP that was modified with APTES content of 20 wt.% or larger, the agglomeration of m-SNPs and unfavorable intermolecular hydrogen bonds between amino groups would cause irregular dispersion of m-SNP in EVA matrix and the appearance of defects in the nanocomposites [10]. This reduces the effect of coupling agent on prevention of thermo-, photo-oxidation for the nanocomposites. Therefore, a tendency of decreased retention in tensile strength and elongation at break of EVA/m-SNP nanocomposites after accelerated weather testing is observed.