Microscopic fracture mechanisms of octavinyl polyhedral oligomeric silsesquioxane-containing hybrid nanocomposite materials

2.1 Materials

The styrene monomer was purified by removing the inhibitor with the aid of an inhibitor-removal column. 2,2′- Azobis (isobutyronitrile) (AIBN) was recrystallized from methanol and dried in vacuum. The OV-POSS was purchased from Hybrid Plastics, Hattiesburg, MS, USA. All other reagents used for experiments were purchased from Sinopharm Chemical Reagent Co. (Shanghai, China).

2.2 Preparation of PS-OV-POSS composites

PS-OV-POSS composites with various mass fractions of OV-POSS were synthesized via bulk polymerization. Styrene, 50 ml, the desired amount of OV-POSS, and 0.1 g AIBN were fed into a three-necked flask. After stirring for 20 min, the mixture was heated rapidly to 60°C. The reaction was carried out at 60°C and under continuous stirring for 7 h. Subsequently, the prepolymer was cast into watch glasses and kept at 70°C under vacuum for 8 h until obtaining the desired PS-OV-POSS composites. For comparison, PS was also synthesized via bulk polymerization. Styrene, 50 ml, and 0.1 g AIBN were charged into a 100-ml three-neck flask. The mixture was heated quickly to 60°C after homogeneous mixing. The reaction mixture was stirred for 8 h at 60°C and then placed in a vacuum oven for 8 h at 70°C.

2.3 Characterization

The infrared spectra on OV-POSS, PS, and PS-OV-POSS composites were obtained using a Nexus Fourier transform infrared spectroscope (FTIR) (Mattson Instruments, Inc., Madison, WI, USA). Optical grade potassium bromide (KBr, International Crystal Laboratories, Garfield, NJ, USA) was used as the supporting medium. The resolution was ±2/cm.

The X-ray measurements were carried out on a D/Max-IIIA X-ray generator (Rigaku Co., Japan) with Cu-Kα radiation at a wavelength of 1.54 Å. The specimens were scanned from 5° to 40° with a scan speed of 0.5°/min.

Scanning electron microscopy imaging was obtained by means of a Toshiba S-4800 field emission scanning electron microscopy (FESEM, Tokyo, Japan). The fracture surfaces of the samples were contrasted with platinum.

The dynamic mechanical behavior of the samples was studied using a Perkin-Elmer DMA-7e dynamic mechanical analyzer (Wellesley, MA, USA). The samples of 20 mm×5 mm×2 mm were used for the dynamic mechanical analysis (DMA) testing, and the experiments were performed under ambient atmosphere in step mode every 5°C from 25°C to 200°C, and a frequency of 1 Hz was used, after the optimization of the static and dynamic loads.

The impact strengths were determined according to the Chinese standard GB/T 16420-1996 using an impact tester (XC-22Z, Chengde, China). The specimen sizes employed were 56 mm (length)×6 mm (width)×4 mm (thickness). The sample gap was set as 40 mm. The tensile stress-strain behavior was measured according to the Chinese standard GB/T 1043-92 on dumbbell-shaped specimens using a SANS CMT 5202 testing machine (Shenzhen, China). The cross-head speed was set at 5 mm/min, and the test continued until sample failure. The flexural stress-strain behaviors of the specimens (50 mm×10.5 mm×4.0 mm) were evaluated according to the Chinese standard GB/T 9341-2000 using a SANS CMT 5202 testing machine (Shenzhen, China) with a speed of 5 mm/min, and the sample gap was set as 30 mm. All mechanical tests were performed under ambient conditions. A minimum of five tests were analyzed for each sample, and the average values are reported.

3.1 Mechanical properties of PS-OV-POSS hybrid materials

The impact, tensile, and flexural data (see Figure 1) are utilized to assess the properties of PS-OV-POSS hybrid materials. The composites are observed to undergo brittle fracture in the mechanical behavior test. The incorporation of OV-POSS resulted in markedly more frangible behavior than the neat PS, presenting a sudden drop (from ∼7 to 1 KJ/m²) of the impact strength at an exceedingly low OV-POSS level. Correspondingly, both the tensile and flexural strengths of the composites calculated from the stress-strain data were maintained almost the same as the neat PS at OV-POSS contents below 3 wt%. The OV-POSS makes the composite a frangible material while retaining the inherent tensile and flexural properties, so the material is hopefully employed in the low impact resistant structure by simple fabrication, that is, current additional structure design to increase the frangibility of the structures could be not required any more, which is of significance for the low-cost manufacturing. The steady plateau of impact property between 0.6% and 3% (by weight) is expected to ensure performance stability and help promote the mass production of the materials. Moreover, the material is easy to dispose by means of a mechanical process like crushing due to its inherent frangible character and light weight, which would provide greater potential environmental benefits than current low impact resistant structure materials.

Figure 1

Impact strengths (A) of PS-OV-POSS composites with various OV-POSS contents and the tensile (B) and flexural (C) stress-strain curves.

3.2 Bonding between OV-POSS and PS

Figure 2 shows the FTIR spectra for OV-POSS, neat PS, and PS-OV-POSS composites with various OV-POSS contents. The strong absorption bands in the pure PS and PS-OV-POSS composites are observed at 1600, 1492, and 1452/cm, assigned to the C=C stretching vibration of the phenyl ring, and at 696 and 756/cm, attributed to the monosubstituted benzene. The characteristic absorbance at 1117/cm corresponded to the Si-O-Si stretching vibrations that are observed in the OV-POSS and PS-OV-POSS composites spectrum, as well as the intensity of the band increases with the OV-POSS content in PS-OV-POSS composites, indicating the existence of OV-POSS in the PS matrix. The absorbance at 1117/cm is absent in the neat PS spectrum. The characteristic peak of CH=CH₂ in OV-POSS at 1604/cm is overlapped with the absorption band of the phenyl ring. The chemical bonding between the OV-POSS and the PS molecules would decrease the local chain mobility resulting in the enhanced frangibility of the composites.

Figure 2

FTIR spectra for OV-POSS, neat PS, and PS-OV-POSS composites with various OV-POSS contents.

3.3 Swelling behavior of PS-OV-POSS composites

In order to further elucidate the nature of interaction between OV-POSS and PS, swelling experiments have been performed with PS-OV-POSS systems at ambient temperature using dichloromethane (DCM) and toluene for 24 h. The results indicate that these composites swelled but do not dissolve in both DCM and toluene and kept their original shapes, making us believe that the synthesized composite is most likely in cross-linked form rather than in star-shaped structure, and the OV-POSS is deemed to serve as a nanocross-linker. The cross-linking serves to confine the mobility of the chain segments and, thus, decrease the energy absorption of the material under impact loading, that is, leading to the frangibility.

3.4 Microscopic analysis of the PS-OV-POSS composite structure

Figure 3 shows the X-ray diffraction patterns of the OV-POSS, PS, and PS-OV-POSS composites (0.6–3 wt%) with Bragg’s angle (2θ) varying from 5° to 40°. Each of the PS-OV-POSS composites shows only one broad peak at 18°, which is quite similar to the amorphous PS, and the peak intensity increases with the OV-POSS content. For the PS-OV-POSS composites at OV-POSS contents below 3 wt%, the sharp characteristic peaks of OV-POSS at 9.7°, 13°, 22.7°, and 23.6° disappeared, which suggests that OV-POSS is homogeneously dispersed in the PS matrix at the molecular level, which is favorable to the formation of cross-linked networks. The composites with 3 wt% OV-POSS shows a diffraction peak in the vicinity of 8°, which may be attributed to the increasing interaction between closely spaced OV-POSS molecules with the increase in OV-POSS content. This indicates the presence of non-cross-linked OV-POSS, which deteriorates the overall mechanical properties.

Figure 3

X-ray diffraction patterns of OV-POSS, neat PS, and PS-OV-POSS composites with various OV-POSS contents.

The SEM analysis provided further important information concerning the distribution of cross-linked networks. Figure 4 shows that the fracture surface of the neat PS appears much rougher and more disorderly than that of the composite. Spherical regions in the diameter range from ∼20 to 30 nm are homogeneously dispersed in the composite, as depicted in Figure 4B. Considering that the overall diameter of OV-POSS molecule ranges from 1 to 3 nm and a molecular-level dispersion of OV-POSS in the composite, one can infer that the spherical regions are not OV-POSS aggregates but densely cross-linked microregions with higher cross-linking density. The presence of the dispersed spherical regions could reflect the distribution of the cross-linking density in the system, that is, large-scale homogeneity and local fluctuation. The fluctuation of cross-linking density induced local phase separation and, thus, led to the formation of spherical densely cross-linked microregions with higher cross-linking density than other regions. The densely cross-linked microregions act as stress concentration sources when subjected to impact loading, resulting in drastic reduction in the impact strength. The large-scale homogeneity of cross-linked networks would contribute to maintain the tensile and flexural properties of virgin polymer [17]. More experimental evidence is needed to gain more insight into the effect of the incorporation of OV-POSS on the tensile and flexural properties.

Figure 4

SEM images of freeze-fracture surfaces of neat PS (A) and PS-OV-POSS composite containing 2 wt% OV-POSS (B).

3.5 Effect of OV-POSS on the chain dynamics

The loss tangents tan δ and tensile storage moduli E′ vs. temperature at 1 Hz with increasing OV-POSS content are presented in Figure 5 A and B, respectively. The glass transition temperature (T_g) shifted to a higher temperature for the PS-OV-POSS composite. The incorporation of a multifunctional OV-POSS into PS is responsible for the increase in the T_g of the composites. It indicates that OV-POSS moieties confined the motion of chain segments and affected primary (α) molecular mobility of the matrix polymer.

Figure 5

DMA traces that show loss tangents tan δ (A) and tensile storage moduli E’ (B) as a function of temperature.

Similar results were obtained from storage moduli vs. temperature data (Figure 5B). The E′ values of the composites increase in the glassy state due to the incorporation of OV-POSS into PS through chemical bonding. The OV-POSS moieties incorporated into the PS network retarded the polymer chain segmental motion. During curing, the OV-POSS moieties react with propagating radical chains and become part of a growing polymer chain and, then, participate in the cross-linking due to its multifunctionality. As styrene monomers are consumed, the OV-POSS moieties are immobilized as joints of growing net, retarding chain segmental motion of PS and increasing the E′ values of the composites.

Actually, more OV-POSS does not mean an increased T_g or a higher E′ value. The explanation would be, during polymerization, the probability of the encounter between the actively propagating chain ends of PS and OV-POSS moieties increases with the increasing OV-POSS loading, promoting chain termination and, further, resulting in decreased molecular weights of the PS chain segments or even the occurrence of non-cross-linked OV-POSS and, thus, lower T_g, which is due to the extra fractional free volume or larger configurational entropy [18].