Interfacial Interactions in Calcium

Carbonate
Polypropylene Composites.
2: Effect of Compounding on the Dispersion and the
Impact Properties of Surface-Modified Composites
YEH WANG* and WEI-C. LEE

Department of Chemical Engineering

Tunghai University
Taichung, Taiwan, 407
Republic of China

This study was carried out to investigate the influences of compounding process
and surface treatment on calcium carbonate (CaCO3) filled polypropylene. The com-
pounding process is discussed with reference to a twin-screw extruder and an in-
ternal mixer. The calcium carbonate filler was surface-treated with a liquid titanate
coupling agent (LICA 12) and stearic acid. Composites of different weight fractions
were prepared by both compounding processes, and their impact properties were
evaluated. The notched Izod impact strength increased with CaCO3 content up to a
maximum at about 10 vol%, and then decreased. Surface treatment of CaCO3 filler
generally yielded composites of higher impact strength than untreated system.
Though LICA 12 was more effective than stearic acid in modifying the filler, the low-
cost stearic acid proved to be more effective when dealing with the impact properties
of composites. Moreover, the composites from a Brabender Plasti-corder exhibited
better gross uniformity than that from the twin-screw extruder. However, good
filler dispersion and uniform microscopic morphology, as revealed by SEM micros-
copy, was observed in the samples from the twin-screw extruder. Polym. Compos.
25:451– 460, 2004. © 2004 Society of Plastics Engineers.

INTRODUCTION and titanates have been widely used as effective coat-

ings for fillers in many particulate-filled thermoplastic
I t has been a widespread practice to incorporate
mineral fillers into plastics and elastomers to ex-
tend the matrix or to enhance certain properties. The
polymers (16, 17). The physical properties of the filled
compounds are enhanced through improved filler dis-
persion.
degree of improvement depends on the judicious
On the other hand, the particle dispersion and re-
choice of filler origin, particle size and shape, fraction
agglomeration can happen simultaneously during
of filler (1
7), and surface treatment to modify inter-
compounding of the polymer composites. The breakup
action between polymer the matrix and filler (8
12).
of agglomerates occurs when hydrodynamic forces in-
To optimize the properties of filled compounds, how-
duced by the melt flow overcomes the cohesion forces
ever, effective mixing is also of great importance in
of the agglomerates; and particle re-agglomeration oc-
order to achieve adequate spatial distribution and dis-
curs when aggregates are brought together by the
persion of filler particles (13
15).
non-homogeneous flow field. Therefore, effective mix-
The surface treatment is often needed particularly
ing requires not only dispersive mixing to break up
when a polar filler is incorporated into a nonpolar poly-
agglomerates, but also distributive mixing to promote
mer matrix. Improved filler dispersion is generally
wetting and minimize agglomeration after filler break-
achieved as a result of using surface coatings on partic-
up.
ulate mineral fillers. In this respect, both stearic acid
The mixing efficiency can be achieved in mixing de-
vices through the distributive and the dispersive mixing
* To whom correspondence should be addressed. mechanisms. However, so far, very few studies have
© 2004 Society of Plastics Engineers
Published online in Wiley InterScience (www.interscience.wiley.com). examined the influence of the surface treatment on the
DOI: 10.1002/pc.20038 final state of dispersion (18, 19). In order to maintain

POLYMER COMPOSITE, OCTOBER 2004, Vol. 25, No. 5 451

 Yeh Wang and Wei-C. Lee

the quality of finished products, it is necessary to define Compounding was carried out on a co-rotating self-
and quantify the state of dispersion generated from wiping twin-screw extruder. The PSM30 machine was
compounding operations. For thermoplastic systems, manufactured by Sino-Alloy Machinery Inc. with a
SEM of fractured surfaces and image analysis meth- screw diameter D of 31.2 mm, the distance between
ods are commonly used in order to define various dis- screw axes was 26.2 mm, screw tip clearance was
persion indexes and mean particle diameters (18, 0.25 mm, and the length to the diameter ratio L/D 
20
22). 45. The screw consisted of ten segmented barrels with
To control and to improve the mechanical properties three kneading zones. The first kneading zone started
of particulate-filled polymeric composites, it is desir- at the 2nd barrel consisting of high shear disk blocks,
able to understand the interactions between surface and ended with neutral blocks. The second zone
treatment, processing, and properties. Therefore, this started at the 4th barrel also consisting of high shear
study was initiated to investigate the difference be- elements with reverse elements at the end. The third
tween the continuous process, the twin-screw ex- zone started at the 7th barrel with only neutral ele-
truder, and the batch process, the internal mixer. We ments. In the first and the second kneading zones,
characterized the structure of composites in terms of more severe shearing action is assumed due to the
the state of dispersion of the filler through the use of high shear disk blocks and due to the presence of re-
scanning electron microscopy in conjunction with a verse flighted elements. The reverse elements increase
commercial image analysis system. Special attention the resistance to flow resulting in an increase of the
has been paid to the investigation of large aggregates fill degree and the residence time in the mixing section.
and agglomerates existing in the composites. Notched The neutral elements only induce gentle shearing and
Izod impact properties were analyzed, and their rela- homogenization of the polymer melt (24, 25). There-
tions to the state of dispersion were investigated. The fore, it is expected that the filler particles experience
effects of filler concentration and processing condi- high intensity of dispersive mixing in the first and the
tions are discussed as well. second kneading zones; and the distributive mixing
action dominates in the third kneading zone. The de-
EXPERIMENTAL tails of the screw configuration and element geome-
tries are shown in Fig. 1. A standard pelletizing die
Materials
plate was installed at the screw end. The strand was
The general-purpose isotactic polypropylene homo- solidified in a water bath and pelletized.
polymer (Yungsox 1040) provided by the Yung-Chia In addition to the screw configuration, the principal
Chemical Co., Taiwan, was used in this study. The processing variables were throughput rate, screw
density of polypropylene was 904 kg/m3 measured speed, and barrel temperature profiles. After several
with an electronic densimeter. The calcium carbonate trials, the screw speed was set at 100 rpm, and the
with the trade name Hydrocarb 90 was supplied by barrel temperatures were set from 180°C at the first
Omya, Switzerland. Its average particle size was 0.8 barrel to 200°C at the last barrel. The barrel tempera-
m and the specific gravity was 2.7, which were pro- ture profiles are also shown in Fig. 1. The screw speed
vided by the manufacturer. The surface treatment and the barrel temperatures were fixed for all experi-
was carried out by the dissolution method, whose de- mental runs. Finally, the throughput rate was in the
tails are given in part 1 (23). The calcium carbonate range from 5 kg/hr to 8 kg/hr for compounds of dif-
filler were treated with 0.3 phf LICA 12 (Kenrich Petro- ferent formulations.
chemicals Inc., USA), or treated with 7.4 phf stearic We also employed an internal mixer besides the
acid (Lancaster, UK). Here ‘phf ’ denotes parts per twin-screw extruder for the comparison purpose. The
hundred filler. LICA 12 is a liquid titanate coupling internal mixer was a Brabender PL 2000 Plasti-
agent that has been well known for its effectiveness in corder. A mixing head of W 50 type with a mixing vol-
modifying inorganic fillers, and stearic acid has been ume of 60 cm3 was installed. After loading polypropyl-
a widely used surfactant for inorganic fillers. To all ene, the calcium carbonate filler was fed into the
the compounds, 0.2 wt% heat stabilizer (Evernox-10) mixing chamber sequentially. The mixing temperature
from Everspring Chemical Co., Taiwan, was added to was set at 180°C, rotation speed was 100 rpm, and
prevent degradation of the polymer during compound- the total mixing time was 30 min, during which the
ing. Polypropylene and the stablizer were dry-blended imposed torque would reach an equilibrium value.
at room temperature after dehumidifying.
Notched Izod Impact Test
Compounding Procedures Test pieces for impact testing were compression
The polymer pellets and mineral additive were me- molded from the pellets under pressure at 200°C for
tered independently in the required proportions using 30 minutes before slow cooling. The notched Izod test
volumetric dosing units. The pellets were fed into the specimens followed the ASTM D256 norms. All test
hopper of the extruder through a forced feeder, and pieces had a sharp notch using a diamond cutter. The
the mineral powder was introduced separately through initial crack depth was 1 mm. The notched samples
the secondary feeding port where the polymer was were used for studying the impact properties with an
partially melted. Izod impact tester at room temperature around 25°C.

452 POLYMER COMPOSITES, OCTOBER 2004, Vol. 25, No. 5

 Interfacial Interactions in CaCO3
PP Composites. 2.

CCD camera). The digitized gray level images repre-

senting a resolution of 480  640 pixels. The images
were quantified using a 586-based personal computer
equipped with an Optimas 5.1 image processing sys-
tem from Bioscan, USA. A gray level was set on a 0
(black) to 255 (white) scale and was held constant
throughout. Average diameter was computed based
on the illuminated pixel regions. Diameter data were
then inventoried and classified.

RESULTS AND DISCUSSION

Macroscopic Analysis
We first compare the mixing quality of compounds
from the two compounding devices based on the mac-
roscopic analysis. The compounds were prepared at
the weight concentrations of 13.6%, 25%, 34.5%, and
42.8%, which correspond to the volume concentra-
tions of 5%, 10%, 15%, and 20%. The results from
thermogravimetric (TGA) measurements for untreated
and treated composites are similar, and only the re-
sults for untreated system are presented as an exam-
ple. Figure 2 is the plot of measured weight concentra-
tion against preset concentration for untreated samples
from twin-screw extruder and internal mixer. The
measured weight concentration is the mean of six
samples. Ideally if the measured concentration is ex-
actly the same as the preset concentration, then the
plot will be a diagonal line starts from the origin. As
expected, the measurements from internal mixer fit
the diagonal line better. The small deviation from di-
agonal line for twin-screw extruder is mainly due to
the use of volumetric feeder, which is not a high preci-
sion device.
Next the results from density measurements for un-
treated system are shown in Fig. 3. The composite
densities were measured by an electronic densimeter.
The measured compound density is also the mean of
six samples. The measured density was plotted against
the preset volume concentration for untreated sam-
ples from twin-screw extruder and internal mixer. It
can be seen that the compound density almost varies
linearly with composite concentration. As expected,
Fig. 1. Schematics of twin-screw configuration and barrel the densities of the composites from the Plasti-corder
temperature profiles. exhibit smaller deviation from the predicted values
than composites from the twin-screw extruder.

Notched Izod Impact Strength

At least six specimens were tested for each com-
The results of the notched Izod impact strength ver-
pound. Fracture surfaces were also studied with the
sus filler concentration of i-PP/CaCO3 composites with
aid of SEM to understand the mechanism of failure.
and without surface treatment from the twin-screw
extruder are shown in Fig. 4. The impact strengths of
Electron Microscopy
the composites appear as functions of filler concentra-
Fracture surfaces for the examination of the mor- tion. The strength increases with CaCO3 content up to
phology were obtained from the room temperature im- a maximum at about 10 vol%, and then decreases.
pact test samples. The microstructures of specimens Similar results have been reported elsewhere (18, 26,
were studied using an SEM of TOPCON. All speci- 27). The increase in toughness at low concentrations
mens were gold-coated in a sputtering chamber prior of the filler may be due to the local microplastic defor-
to SEM studies. In addition the SEM micrographs mation arising from the microscopic cavities around
were digitized by an electronic scanner (UMAX 610S the filler particles. The reduction in toughness at high

POLYMER COMPOSITE, OCTOBER 2004, Vol. 25, No. 5 453

 Yeh Wang and Wei-C. Lee

Fig. 2. Measured weight concentration from TGA versus preset concentration for untreated samples.

filler concentrations is due to limited plastic flow of depends on the structure and properties of the com-
the PP matrix, i.e., the ductile matrix of PP is replaced posites. Structural inhomogeneities, such as aggrega-
by the rigid dispersed particulates. tion, voids, etc., in a poorly dispersed system would
Surface treatment shows a positive effect on impact favor initiation of fracture.
strength. As can be seen from Fig. 4, impact strengths Figure 5 shows the results of the notched Izod
of the treated composites are higher than those of impact strength versus filler concentration of i-PP/
untreated ones. It can also be seen that stearic acid CaCO3 composites with and without surface treat-
treated composites generally give higher strengths ment from the Plasti-corder. Note that we did not add
than LICA 12. The difference in impact strength be- stabilizer into untreated and LICA 12 treated compos-
comes more significant at higher volume concentra- ites. However, since stearic acid is less thermally sta-
tions. While the impact strength of LICA 12 treated ble than LICA 12, the stabilizer was added to stearic
compounds at 20 vol% is lower than the neat PP, acid treated composites in order to prevent the dra-
stearic acid treated compounds show higher strengths matic drop of the equilibrium torque due to excessive
than the neat PP. We noticed that stearic acid treated degradation. As expected, stearic acid treated compos-
CaCO3 showed greater drop in surface energy than ites again exhibit higher strength than that of LICA12
LICA 12 treated filler (23), though it took much higher treated composites, and untreated composites have
concentration of stearic acid than LICA 12 in order to the lowest strength. Note that even at 20 vol% the im-
achieve similar effect of surface treatment. Therefore pact strength of stearic acid treated composites is still
the particle-particle interaction was reduced more ef- higher than the neat PP. Moreover, when comparing
fectively in stearic acid treated CaCO3 fillers, particu- Fig. 5 with Fig. 4, it can be clearly seen that the impact
larly at high loadings of filler. strengths of composites, either treated or untreated,
It has been well known that the extent of stress from the Plasti-corder are lower than those from the
concentration around the inclusion in a matrix is pro- twin-screw extruder. Though the mixing time and ro-
portional to the inclusion size. Thus a low degree of tation speed of the Plasti-corder may not be properly
filler dispersion, which will be discussed through mi- adjusted to achieve the same mixing quality as in the
croscopic observation in the following section, may give twin-screw extruder, too long mixing time or too high
rise to large agglomerates of CaCO3 particles, which rotation speed would induce excessive degradation of
would initiate sample failure and lower the impact the PP matrix. Note that the macroscopic analysis
energy. Furthermore, failure initiation and propagation through TGA and density measurement indicates better

454 POLYMER COMPOSITES, OCTOBER 2004, Vol. 25, No. 5

 Interfacial Interactions in CaCO3
PP Composites. 2.

Fig. 3. Compound density versus concentration for untreated samples.

gross uniformity for compounds from the Plasti-corder. proceeds along the PP-CaCO3 interface so that abun-
However, it does not guarantee better ultimate proper- dant CaCO3 particles can be observed on the fracture
ties like impact strength, which is intimately related surfaces, except for a few embedded fillers. It is rather
to the microscopic morphology and dispersion degree clear here that the untreated CaCO3 fillers do not
of compounds. Further study of mixing quality, the bond the matrix well, which reveals that the filler-ma-
dispersion degree in the composites, and the morphol- trix interface is weak and adhesive failures prevail.
ogy of the fracture surfaces through scanning electron The micrograph shown in Fig. 6b at high magnifica-
microscopy shall be undertaken next. tion allows us to take a close look at the interface be-
tween fillers and matrix. The popped-up particles are
Morphological Observation scattered on the fracture surface without PP residue
covering the surface of the particle. Many uneven holes
Scanning electron micrographs of the fractured sur- suggest the occurrence of filler pullout. Moreover, the
faces of the i-PP/CaCO3 composites from impact test- clean surface of CaCO3 particles with bordered pits
ing are presented in Figs. 6 through 8. Since the effect printed in the matrix clearly indicates they are merely
of surface treatment is most significant at high con- embedded in the matrix without interfacial bonding.
centrations, we show only the pictures of composites Figures 7a and 7b show the fracture surfaces of
at 20 vol%, such that the effectiveness of surface composites with LICA 12 treated CaCO3 from the
treatment could be judiciously evaluated. twin-screw extruder at magnifications of 1000 and
Figures 6a and 6b show the fracture surface of com- 5000, respectively. Both micrographs display nearly
posites with untreated CaCO3 from the twin-screw ex- the same morphology as shown in Figs. 6a and 6b for
truder at magnifications of 1000 and 5000, re- untreated CaCO 3. Figure 7a shows the randomly
spectively. We first examine the fracture surfaces at propagated fracture surface; and Fig. 7b shows popped-
1ow magnification in Fig. 6a. It can be seen that the up particles and uneven holes. In short, surface treat-
randomly propagated fracture surfaces in these mi- ment of CaCO3 filler by LICA 12 does not seem very
crographs clearly confirm the brittle behavior induced effective, and indeed the impact strength of LICA 12
by the PP matrix at room temperature (21, 28, 29). treated composites is only slightly greater than that of
The observed crack has spread mainly through the untreated systems at all loadings of filler, as seen
weakest track in the composite. Hence, the fracture from Figs. 4 and 5.

POLYMER COMPOSITE, OCTOBER 2004, Vol. 25, No. 5 455

 Yeh Wang and Wei-C. Lee

Fig. 4. Impact strength versus filler concentration of treated and untreated composites from twin-screw extruder.

Figures 8a and 8b show the fracture surfaces of twin-screw extruder are shown in Fig. 9. It can be
composites with stearic acid treated CaCO3 from the clearly seen that stearic acid treated CaCO3 shows
twin-screw extruder at magnifications of 1000 and higher fractions of particles with size less than 1 m,
5000, respectively. Both micrographs display quite while untreated CaCO3 shows higher fractions of par-
different morphologies from those shown in the previ- ticles bigger than 3 m. The calculated mean particle
ous figures. In Fig. 8a, though the propagation of the sizes are 1.69 m for untreated CaCO3, 1.21 m for
fracture is not very regular, there are many fewer LICA 12 treated CaCO3, and 1.15 m for stearic acid
structural inhomogeneities, such as uneven cracks treated CaCO3, which is the smallest. Smaller mean
and voids, compared to the composites filled with un- particle size would suggest better quality of particle
treated or LICA 12 treated CaCO3. In Fig. 8b, it can be dispersion.
seen that more particles are embedded in the polymer Figures 10a and 10b show the fracture surfaces of
matrix without gaps between fillers and matrix, and composites with stearic acid treated CaCO3 from the
the fracture surface looks more uniform than that Plasti-corder at magnifications of 1000 and 5000,
shown previously. These morphological observations respectively. It can be clearly seen that these micro-
indicate improved adhesion between fillers and matrix graphs display more voids and irregular cracks, and
due to surface treatment with stearic acid, and explain more popped-up particles without the PP matrix ad-
why the impact strengths of stearic acid treated com- hering to the surface of particle, when compared with
posites are better than untreated or LICA 12 treated Figs. 8a and 8b. However, the mean particle size is
systems, particularly at high loadings of filler. 1.19 m, which is only slightly larger than that from
Image analysis of the fracture surfaces was carried the twin-screw extruder. Apparently, the degree of
out in order to look into the particle size distribution, filler dispersion is comparable to that from the twin-
which is closely related to the quality of filler disper- screw extruder, but thermal degradation may induce
sion. We used only micrographs with magnifications the deterioration of interfacial adhesion due to im-
at 1000. We think, with lower magnification, more proper compounding conditions in the Plasti-corder,
particles would be available for analysis, leading to and this confirms the inferior impact strengths of
more unbiased results. The particle size distributions composites from the Plasti-corder relative to those
of untreated and treated CaCO3 at 20 vol% from the from the twin-screw extruder.

456 POLYMER COMPOSITES, OCTOBER 2004, Vol. 25, No. 5

 Interfacial Interactions in CaCO3
PP Composites. 2.

Fig. 5. Impact strength versus filler concentration of treated and untreated composites from Plasti-corder.

(a) (b)
Fig. 6. Scanning electron micrographs of fractured composites with untreated CaCO3 at 20 vol% from twin-screw extruder; (a) magni-
fication at 1000, (b) magnification at 5000.

POLYMER COMPOSITE, OCTOBER 2004, Vol. 25, No. 5 457

 Yeh Wang and Wei-C. Lee

(a) (b)
Fig. 7. Scanning electron micrographs of fractured composites with L ICA 12 treated CaCO3 at 20 vol% from twin-screw extruder;
(a) magnification at 1000, (b) magnification at 5000.

(a) (b)
Fig. 8. Scanning electron micrographs of fractured composites with stearic acid treated CaCO3 at 20 vol% from twin-screw extruder;
(a) magnification at 1000, (b) magnification at 5000.

CONCLUSION causes a significant drop in impact strength. Note that

This study clearly demonstrates that incorporation of effective surface treatment by stearic acid results in a
fine CaCO3 fillers greatly modifies the impact strength composite with impact strength higher than neat PP
of polypropylene, but the impact strength decreases even at 20 vol% of CaCO3 filler.
with the loading of filler after the maximum strength The effectiveness of surface treatment may be further
occurs at about 10 vol%. The submicron particles pro- justified through microscopic observation. SEM micro-
mote microplastic deformation in the brittle matrix graphs of fracture surfaces with image analysis can
without creating structural inhomogeneities. How- provide a quick and reliable method for quantitative
ever, at high loadings of filler, the replacement of the evaluation of filler dispersion. Particle agglomeration
thermoplastic matrix by the more rigid particulates and poor interfacial adhesion was found to be directly

458 POLYMER COMPOSITES, OCTOBER 2004, Vol. 25, No. 5

 Interfacial Interactions in CaCO3
PP Composites. 2.

Fig. 9. Particle size distribution of CaCO3 filled composites from twin-screw extruder at 20 vol%.

(a) (b)
Fig. 10. Scanning electron micrographs of fractured composites with stearic acid treated CaCO3 at 20 vol% from Plasti-corder;
(a) magnification at 1000, (b) magnification at 5000.

POLYMER COMPOSITE, OCTOBER 2004, Vol. 25, No. 5 459

 Yeh Wang and Wei-C. Lee

related to inferior notched impact strength. Enhanced 3. M. Sumita, T. Ookuma, K. Miyasaka, and K. Ishikawa,
dispersions, revealed by SEM microscopy, were ob- J. Appl. Polym. Sci., 27, 3039 (1982).
4. M. Sumita, T. Shizuma, K. Miyasaka, and K. Ishikawa,
tained through surface treatment. Furthermore, the J. Macromol. Sci. Phys., B22, 601 (1983).
stearic acid modified composites yield better disper- 5. V. P. Chacko, R. J. Farris, and F. E. Karasz, J. Appl.
sion with decreased size of filler agglomerates than Polym. Sci., 28, 2701 (1983).
the LICA 12 treated system. In other words, the exten- 6. S.-R. Dai and M. R. Piggott, Polym. Compos., 7, 19
(1986).
sive acid-base interaction between stearic acid and
7. G. Levita, A. Marchetti, and A. Lazzeri, Polym. Compos.,
CaCO3 filler would, on the one hand, diminish the 10, 39 (1989).
particle-particle interaction and better facilitate the 8. I. Sasaki, K. Ito, T. Kodama, and F. Ide, Kobunshi Ron-
dispersion of the filler in the non-polar matrix, and, bunshui, 33, 162 (1976).
on the other hand, the long alkyl chain of stearic acid 9. T. Nakatuka, H. Kawasaki, and K. Itadani, J. Colloid In-
terface Sci., 82, 298 (1981).
promote interfacial adhesion between the filler and 10. Y. N. Sharma, R. D. Patel, I. H. Dhimmar, and I. S.
the matrix (23). Bhardwai, J. Appl. Polym. Sci., 27, 97 (1982).
Finally, the macroscopic analysis through thermo- 11. K. Mitsuishi, S. Kodama, and H. Kawasaki, Polym. Eng.
gravimetric analysis (TGA) and density measurement Sci., 25, 1069 (1985).
12. K. Mitsuishi, S. Kodama, and H. Kawasaki, Kobunshi
indicates better gross uniformity for compounds from Ronbunshu, 43, 43 (1986).
the Plasti-corder than from the twin-screw extruder, 13. J. W. Ess, P. R. Hornsby, S. Y. Yin, and M. J. Bevis,
because the volumetric dosing unit for twin-screw Plas. Rubber Proc. Appl., 4, 7 (1984).
compounding may not be very precise. However, mor- 14. A. M. Riley, C. D. Paynter, P. M. McGenity, and J. M.
Adams, Plas. Rubber Proc. Appl., 14, 85 (1990).
phological uniformity and high degree of filler disper- 15. S. F. Xavier, J. M. Schultz, and K. Friedrich, J. Mater.
sion, as revealed by SEM microscopy, was observed Sci., 25, 241 (1990).
in the compounds from the twin-screw extruder. The 16. A. Tabtiang and R. Venables, Composite Interfaces, 6,
microscopic uniformity confirms the high impact 65 (1999).
17. S. J. Monte, Paper 16, presented at the ACS Rubber Di-
strength of composites from the twin-screw extruder,
vision Meeting, Rhode Island (2001).
even at high loadings of CaCO3 filler. In short, the 18. Y. Wang and J.-S. Huang, J. Appl. Polym. Sci., 60, 1779
maximum improvement in the mechanical properties (1996).
of the composites could be achieved only through a 19. B. Haworth, C. L. Raymond, and I. Sutherland, Polym.
proper combination of mixing device and compound- Eng. Sci., 40, 1953 (2000).
20. J. W. Ess and P. R. Hornsby, Plast. Rubber Tech., 8,
ing conditions in conjunction with effective surface 147 (1987).
treatment of filler particles. 21. Y. Suetsugu, Inter. Polym. Proc., 5, 184 (1990).
22. M. Bories, M. A. Huneault, and P. G. Lafleur, SPE
ACKNOWLEDGMENT ANTEC Tech. Paper, 350 (1999).
23. Y. Wang and W.-C. Lee, Polym. Compos., 24, 119 (2003).
The financial support of this research by the Na- 24. J. L. White, Twin Screw Extrusion, Hanser Publishers,
tional Science Council of Taiwan, Republic of China, Munich (1990).
25. Y. Wang, Compounding in Co-Rotating Twin-Screw Ex-
under grant NSC-89-2216-E-029-001 is gratefully ac- truders, Rapra Review Report, vol. 10, no. 8 (2000).
knowledged. 26. S. N. Maiti and P. K. Mahapatro, J. Appl. Polym. Sci.,
42, 3101 (1991).
REFERENCES 27. V. Svehlova and E. Poloucek, Die Angewandte Makro-
molekulare Chemie, 153, 197 (1987).
1. L. E. Nielsen, Mechanical Properties of Polymers and 28. M. Bramuzzo, A. Savadori, and D. Bacci, Polym. Compos.,
Composites, Marcel Dekker, New York, 1974, p. 380. 8, 1 (1985).
2. L. Nicolais and R. A. Mashelker, J. Appl. Polym. Sci., 20, 29. E. Nezbedova, J. Ponesicky, and M. Sova, Acta Polymer-
561 (1976). ica, 41, 36 (1990).

460 POLYMER COMPOSITES, OCTOBER 2004, Vol. 25, No. 5