Nucleation of Polypropylene Crystallization by Single-Walled Carbon Nanotubes
Brian P. Grady,* Francisco Pompeo, Robert L. Shambaugh, and Daniel E. Resasco
Department of Chemical Engineering and Materials Science, The UniVersity of Oklahoma,
Norman, Oklahoma 73019
ReceiVed: December 20, 2001; In Final Form: March 28, 2002
Nonisothermal and isothermal crystallization experiments were performed on polypropylene mixed with carbon
nanotubes produced by disproportionation of CO on Co-Mo catalysts. Functionalization of the nanotubes
with octadecylamine made the tubes hydrophobic and allowed the tubes to be solubilized in an organic solvent.
Mixing of the nanotubes with the polymer was accomplished by adding the nanotubes to a Decalin solution
that contained dissolved polypropylene, followed by evaporation of the solvent. Dynamic mechanical analysis
indicated very little difference in the small-strain mechanical properties between filled and unfilled polymers
at the very low solid levels that were tested. By contrast, the crystallization behavior of the filled and unfilled
polymer was quite different. Nanotubes promoted growth of the less-preferred beta form of crystalline
polypropylene at the expense of the alpha form. In nonisothermal crystallization, the total amount of crystalline
material in the sample was the same for the filled and unfilled materials. However, for isothermal crystallization
experiments, the percent crystallinity in the filled materials was slightly higher. Most importantly, the rate of
crystallization was substantially higher in the filled system. The results presented in this paper clearly show
that carbon nanotubes nucleate crystallinity in polypropylene.
1. Introduction
The unusually high Young’s modulus and tensile strength
coupled with the low density of single-walled nanotubes
(SWNT) have prompted investigations of these materials as
reinforcement in polymer composites.
1
Of course, the effective
utilization of SWNT as reinforcement materials is not only
related to their intrinsic properties but also to characteristics
such as dispersion, alignment, and interfacial properties.
2
In our
laboratory, we have been particularly interested in carbon
nanotubes as reinforcing fillers for thermoplastics. One unique
aspect of nanotubes is that they can retain a high aspect ratio
even when processed as melt-spun fibers, unlike typical micron
diameter fillers. The resultant morphologies in this situation,
especially in the case where the filler can nucleate crystallinity,
might be totally unique.
Carbon fibers, which are closely related to carbon nanotubes,
can nucleate crystallinity in isotactic polypropylene.
3-9
This
nucleating ability leads to transcrystallization, which is a region
of highly oriented crystalline material with the c-axis parallel
to the fiber surface at regions very close to the fiber surface.
10
This orientation is a result of very dense nucleation on the fiber,
which consequently restricts crystal growth in the direction
normal to the fiber. However, several studies have shown that
the crystalline growth rate, that is, the rate of the crystal front
movement independent of nucleation, is no different for a
transcrystalline layer versus a normal spherulite.
5,9
The fundamental morphological characteristics of the carbon
fiber that lead to the formation of this transcrystalline layer have
not been established. Thomason and Van Rooyen
9
showed that
high modulus (HM) carbon fibers tend to form transcrystalline
layers with isotactic polypropylene, while high-strength (HS)
fibers tend to not form transcrystalline layers. A similar
distinction was found for syndiotactic polystyrene transcrystal-
linity.
11
HM fibers can be produced from either polyacrylonitrile
(PAN) or pitch, while the HS fibers are typically only produced
from PAN. However, there is not enough evidence to conclude
that only pitch-based fibers can cause transcrystallinity. In a
second paper by Thomason and Van Rooyen,
12
the authors
suggested that transcrystallinity is due to stress between the fiber
and the matrix, and hence large differences in axial thermal
expansion coefficients between the fiber and the polymer favor
transcrystallinity. A similar explanation with some quantification
was proposed by another set of authors.
13
However, both sets
of authors note that fiber-melt interaction does seem to be
important. In the particular case of carbon fibers, the fact that
some carbon fibers form transcrystalline layers and others do
not would suggest that the interaction between the polymer and
the fiber surface is almost certainly important. Sizings (propri-
etary coatings often applied to fibers to improve processing
characteristics or fiber-matrix compatibility) may also play a
role. For glass fibers, different sizings have been shown to affect
the ability of the fiber to nucleate crystallinity.
14,15
Clearly, the characteristics required to induce transcrystallinity
have not been established. Even if a priori such a prediction
could be made for carbon fibers, the specific relevance to carbon
nanotubes would not be clear. Hence, the only way to determine
whether nanotubes will nucleate crystallinity is through experi-
mental measurement. The purpose of this study is to determine
whether carbon nanotubes nucleate crystallization in polypro-
pylene.
2. Experimental Section
2.1 Materials. The SWNT used in this work have been
obtained by a catalytic decomposition method based on the
disproportionation of CO over Co-Mo/SiO
2
catalysts at a total
pressure of 5 atm and 1123 K. In a previous study,
16
we fully
characterized the structure and chemical state of these catalysts
by EXAFS, XANES, UV/vis-DRS, H
2
TPR, XPS, and DRIFTS * To whom correspondence should be addressed.
5852 J. Phys. Chem. B 2002, 106, 5852-5858
10.1021/jp014622y CCC: $22.00 © 2002 American Chemical Society
Published on Web 05/21/2002