8 Nanocomposites Based on Thermosetting Polyurethane Matrix. . .
139
motional cooperativity due to increasing free volume and loosening segmental
packing with elevating temperature. Coincidence of Q(T) plots below 50 ◦ C for neat
PU and for the nanocomposite (Fig. 8.11) confirms the above-mentioned suggestion
that nanotubes are concentrated mainly near the hard nanodomains of PU matrix,
and therefore their presence does not change the activation barriers of motion in the
PPG nanoregions undisturbed by MWCNTs.
However, at higher temperatures the Q(T) plots do not coincide (Fig. 8.11). In
this region, the Q maxima observed are associated with unfreezing of the networkjunction motion and the Q value increases to 160 kJ mol −1 at about 80 ◦ C for neat
matrix but up to 260 kJ mol −1 at 130 ◦ C for the nanocomposite. Moreover, the
latter peak covers the range from 100 to 150 ◦ C, which suggests the occurrence of
complicated dynamics in this region as well. Since the co-reaction of isocyanate
groups in the formation of the PU network with functional groups of MWCNTs
occurs, we assume that the high-temperature dynamics in the nanocomposite glass
transition may be considered as network-junction motions that are hindered to
different extents by their covalent attaching to MWCNTs. The notion about the
complicated glass transition dynamics in the nanocomposites studied over the
temperature range from −60 to 140 ◦ C was directly confirmed also from the creep
rate spectroscopy data (see next section).
It had been observed through another studies that varying the surface chemistry
and in turn the interface, albeit in those cases high filler levels of MWCNTs
were used and acid-oxidized-derived fulvic acids were not removed, could lead to
substantial changes in the final composite properties. Therefore, it was anticipated
that even at low filler levels, with a good degree of dispersion and after cleaning
nanotubes from fulvic acids, such changes would reflect more correctly the actual
interface state through the surface chemistry. Ultimately, the peculiar segmental
dynamics in the nanocomposites studied manifested itself indeed in their enhanced
mechanical properties.
Figure 8.12 shows the impact of MWCNTs introduced into PU on the polymer’s
dynamic modulus E’ values when measured at 1 Hz over the broad temperature
range. Figure 8.12a shows a considerable rise of E’ with increasing MWCNTox content, up to three times at 20–100 ◦ C at 0.25% nanofiller. Figure 8.12b
compares the E’(T) plots for the nanocomposites with 0.25% nanofiller with
different interfacial chemistry. Again, at van der Waals interfacial interactions (using
MWCNT-hemin or MWCNT-hemin-red), effect is practically absent in comparison
with unfilled matrix (Fig. 8.12b, curves 2 and 3). Whereas covalent interfacial
bonding (using MWCNT-red and especially MWCNT-ox) provides the maximal
enhancing of dynamic modulus of nanocomposites over the whole temperature
range (Fig. 8.12b, curves 4 and 5).
139
motional cooperativity due to increasing free volume and loosening segmental
packing with elevating temperature. Coincidence of Q(T) plots below 50 ◦ C for neat
PU and for the nanocomposite (Fig. 8.11) confirms the above-mentioned suggestion
that nanotubes are concentrated mainly near the hard nanodomains of PU matrix,
and therefore their presence does not change the activation barriers of motion in the
PPG nanoregions undisturbed by MWCNTs.
However, at higher temperatures the Q(T) plots do not coincide (Fig. 8.11). In
this region, the Q maxima observed are associated with unfreezing of the networkjunction motion and the Q value increases to 160 kJ mol −1 at about 80 ◦ C for neat
matrix but up to 260 kJ mol −1 at 130 ◦ C for the nanocomposite. Moreover, the
latter peak covers the range from 100 to 150 ◦ C, which suggests the occurrence of
complicated dynamics in this region as well. Since the co-reaction of isocyanate
groups in the formation of the PU network with functional groups of MWCNTs
occurs, we assume that the high-temperature dynamics in the nanocomposite glass
transition may be considered as network-junction motions that are hindered to
different extents by their covalent attaching to MWCNTs. The notion about the
complicated glass transition dynamics in the nanocomposites studied over the
temperature range from −60 to 140 ◦ C was directly confirmed also from the creep
rate spectroscopy data (see next section).
It had been observed through another studies that varying the surface chemistry
and in turn the interface, albeit in those cases high filler levels of MWCNTs
were used and acid-oxidized-derived fulvic acids were not removed, could lead to
substantial changes in the final composite properties. Therefore, it was anticipated
that even at low filler levels, with a good degree of dispersion and after cleaning
nanotubes from fulvic acids, such changes would reflect more correctly the actual
interface state through the surface chemistry. Ultimately, the peculiar segmental
dynamics in the nanocomposites studied manifested itself indeed in their enhanced
mechanical properties.
Figure 8.12 shows the impact of MWCNTs introduced into PU on the polymer’s
dynamic modulus E’ values when measured at 1 Hz over the broad temperature
range. Figure 8.12a shows a considerable rise of E’ with increasing MWCNTox content, up to three times at 20–100 ◦ C at 0.25% nanofiller. Figure 8.12b
compares the E’(T) plots for the nanocomposites with 0.25% nanofiller with
different interfacial chemistry. Again, at van der Waals interfacial interactions (using
MWCNT-hemin or MWCNT-hemin-red), effect is practically absent in comparison
with unfilled matrix (Fig. 8.12b, curves 2 and 3). Whereas covalent interfacial
bonding (using MWCNT-red and especially MWCNT-ox) provides the maximal
enhancing of dynamic modulus of nanocomposites over the whole temperature
range (Fig. 8.12b, curves 4 and 5).