Effect of Network Chain Length on the Tearing Energy Master Curves of Poly(Dimethyldiphenyl Siloxane)

Abstract

Abstract Five different molecular weights of vinyl-terminated poly(dimethyldiphenyl siloxane) were endlinked with tetrakis(dimethylsiloxy)silane to produce networks of various crosslink densities. A hydrosilation endlinking reaction was chosen, since this method provides the advantage of easily controlling the network chain length and network morphology. Tear energies were determined using a modified trouser-tear test piece at several rates and temperatures. All of the tear energy data for each crosslink density could be shifted to a single master curve with the same WLF coefficients, C1=6 and C2=108 K. The threshold tear energy of PDMDPS is different from unsubstituted PDMS because the average molecular weight per backbone atom and density are increased by the presence of phenyl groups. Each of the tear-energy master curves of PDMDPS shift to a single curve with Mc1/2, except in regions near the glass transition. Deviations of the curves are observed to coincide with increases in the crack-tip diameter with increasing molecular weight between crosslinks. In the lower transition region of the tear-energy master curve, the loss function is determined by the hysteresis of the material. Near the glass transition, the effect of the strain energy and strain-energy distribution, as well as the hysteresis, must be considered. Deviations of the master curves with crosslink density can also be explained from the changes in the loss function for each curve due to the differences in strain and strain distribution for different crack-tip diameters which were visually evident. Andrews' representation of the loss function of the tear energy is known to be dependent on hysteresis, strain energy, and the strain-energy distribution near the crack tip. Modifying his relation, the tear energy was shown to increase with hysteresis as long as the strain energy and strain-energy distribution remained constant. The strain energy distribution was also shown to increase directly with the molecular weight between crosslinks of the network.