Chemistry Reference
In-Depth Information
geometry of the unit cell of the crystal lattice. Polymers crystallized from the
melt are polycrystalline. Their structures are a conglomerate of disordered mate-
rial and clusters of crystallites that developed more or less simultaneously from
the growth of many nuclei. Distinct crystal faces cannot be distinguished, and the
ordered regions in semicrystalline polymers are generally much smaller than those
in more perfectly crystallized micromolecular species. X-ray maxima are broad-
ened by small crystallite sizes and by defects in larger crystals. In either case
such data may be interpreted as indicating that the highly ordered regions in semi-
crystalline polymers have dimensions of the order of 10
2
5
10
2
6
cm. These
domains are held together by “tie molecules” which traverse more than one crys-
tallite. This is what gives a semicrystalline polymer its mechanical strength.
Aggregates of crystals of small molecules are held together only by secondary
forces and are easily split apart. Such fragility is not observed in a polymer sam-
ple unless the ordered regions are large enough to swallow most macromolecules
whole and leave few interregional molecular ties.
The term
crystallite
is used in polymer science to imply a component of an
interconnected microcrystalline structure. Metals also belong to the class of
microcrystalline solids, since they consist of tiny ordered grains connected by
strong boundaries.
4.3.2.1
Nucleation of Crystallization
Crystallization begins from a nucleus that may derive from surfaces of adventi-
tious impurities (heterogeneous nucleation) or from the aggregation of polymer
segments at temperatures below
T
m
(homogeneous nucleation). The latter process
is reversible up to the point where a critical size is reached, beyond which further
growth results in a net decrease of free energy of the system. Another source of
nuclei in polymer melts is ordered regions that are not fully destroyed during the
prior melting process. Such nuclei can occur if segments in ordered regions find
it difficult to diffuse away from each other, because the melt is very viscous or
because these segments are pinned between regions of entanglement. The domi-
nant effect in bulk crystallization appears to be the latter type of nucleation, as
evidenced by in nuclear magnetic resonance spectroscopy relaxation experiments
and other observations that indicate that polyolefins contain regions with different
segmental densities at
3]
.
Although segments of macromolecules in the most compact of these regions are
not crystalline, as measured by calorimetry or X-ray diffraction, they would
remain close together even when the bulk of the polymer is molten and can
reform crystallites very readily when the temperature is lowered. The number of
such nuclei that are available for crystal growth is a function of the degree of
supercooling of the polymer. Incidentally, this explains why polyethylene has
never been observed in the completely amorphous state; even when the melt is
quenched in liquid N
2
crystallites will form since they are produced simply by
the shrinkage of the polymer volume on cooling. An alternative mechanism
that is postulated involves heterogeneous nucleation on adventitious impurities.
temperatures above their melting temperatures
[1
