polymers, under the same conditions, are subjected to diversified confinement
effects. Seemingly, under conditions of varying film thickness, correlations do
exist among fragility and the amplitude of the T g change [123]. Alongside the
experimental studies, simulations of polymers under confinement have shown that
a free interface, or the usage of repulsive interactions among the polymer and the
solid walls, induces acceleration in the dynamical response of the polymer [124–
126], whereas the use of attractive polymer-wall interactions has the opposite effect
[16, 127]; see Fig. 19 where a tremendous slowing down of the segmental relaxation
in polymer matrix is shown upon approaching the nanofiller surface. Essentially, it
has been confirmed that the glass transition temperature in films is significantly
different than that of the bulk and mostly depends on interfacial phenomena
[99, 128]. On the other hand, despite having closely attended to the effect of the
degree of confinement and the adhesion interactions on the glass transition temperature and the segmental dynamics in thin polymer films, the effect of cross-links on
the properties of confined polymers is much less investigated — especially with
molecular dynamics simulations [129].
Concerning the particulate model, the nanofiller particles have been simulated
explicitly. Each nanoparticle consisted of a specified number of beads which were
randomly packed inside a sphere of a given diameter σ. Due to the random packing
of the filler beads, though, each nanoparticle was only approximately spherical. The
polymer matrix was composed of homopolymer chains with the same length as those
in the film model. The varying parameters in the simulations of the particulate model
were the filler volume fraction and the radius of the nanoparticles. Our simulation
results of many-filler nanocomposites showed a linear dependence of the reinforcement on the inverse radius of the NPs, in agreement with recent experimental studies
[17]. Further, the simulated systems with attractive NP-NP interactions displayed a
sharply increasing reinforcement once the average distance between the surfaces of
the NPs became smaller than the NP interaction radius. This observation may serve
as an indication that, for a high enough volume fraction of fillers, the development of
a filler network could indeed be an important source of reinforcement. Further,
comparing the film- and many-particles simulation models, we saw that their
structural properties (e.g. density profiles), as well as the values of the reinforcement
(Fig. 20), are quite similar. This allows us to conclude that the confinement effects,
present in the film model, are replaced by another reinforcing factor in the particulate
model. This additional factor seems to be the direct, attractive interactions among the
nanofillers, which was absent in the film model. Further simulations with a larger
number of filler particles are needed so as to study the sensitivity of the reinforcement on the nanofiller interactions.
Rheology, Rupture, Reinforcement and Reversibility: Computational Approaches. . .
119
effects. Seemingly, under conditions of varying film thickness, correlations do
exist among fragility and the amplitude of the T g change [123]. Alongside the
experimental studies, simulations of polymers under confinement have shown that
a free interface, or the usage of repulsive interactions among the polymer and the
solid walls, induces acceleration in the dynamical response of the polymer [124–
126], whereas the use of attractive polymer-wall interactions has the opposite effect
[16, 127]; see Fig. 19 where a tremendous slowing down of the segmental relaxation
in polymer matrix is shown upon approaching the nanofiller surface. Essentially, it
has been confirmed that the glass transition temperature in films is significantly
different than that of the bulk and mostly depends on interfacial phenomena
[99, 128]. On the other hand, despite having closely attended to the effect of the
degree of confinement and the adhesion interactions on the glass transition temperature and the segmental dynamics in thin polymer films, the effect of cross-links on
the properties of confined polymers is much less investigated — especially with
molecular dynamics simulations [129].
Concerning the particulate model, the nanofiller particles have been simulated
explicitly. Each nanoparticle consisted of a specified number of beads which were
randomly packed inside a sphere of a given diameter σ. Due to the random packing
of the filler beads, though, each nanoparticle was only approximately spherical. The
polymer matrix was composed of homopolymer chains with the same length as those
in the film model. The varying parameters in the simulations of the particulate model
were the filler volume fraction and the radius of the nanoparticles. Our simulation
results of many-filler nanocomposites showed a linear dependence of the reinforcement on the inverse radius of the NPs, in agreement with recent experimental studies
[17]. Further, the simulated systems with attractive NP-NP interactions displayed a
sharply increasing reinforcement once the average distance between the surfaces of
the NPs became smaller than the NP interaction radius. This observation may serve
as an indication that, for a high enough volume fraction of fillers, the development of
a filler network could indeed be an important source of reinforcement. Further,
comparing the film- and many-particles simulation models, we saw that their
structural properties (e.g. density profiles), as well as the values of the reinforcement
(Fig. 20), are quite similar. This allows us to conclude that the confinement effects,
present in the film model, are replaced by another reinforcing factor in the particulate
model. This additional factor seems to be the direct, attractive interactions among the
nanofillers, which was absent in the film model. Further simulations with a larger
number of filler particles are needed so as to study the sensitivity of the reinforcement on the nanofiller interactions.
Rheology, Rupture, Reinforcement and Reversibility: Computational Approaches. . .
119