Thermomechanical Properties of Virgin and RecycledPolypropylene Impact Copolymer/CaCO3

Nanocomposites

A. Elloumi,1 S. Pimbert,2,3 A. Bourmaud,2,3 C. Bradai11 Laboratoire des Systemes ElectroMecaniques (LASEM), Ecole Nationale d’Ingenieurs de Sfax (ENIS),B.P. W3038, Sfax, Tunisia

2 Laboratoire d’Ingenierie des Materiaux de Bretagne (LIMATB), Universite de Bretagne Sud, Lorient 56321,France

3 Universite Europeenne de Bretagne, France

The effect of successive injection moldings on thethermal, rheological, and mechanical properties of apolypropylene impact copolymer (PP) was investigated.The crystal content decreased as the molecular weightdecreased due to chain scission with repeated injec-tion molding. The Young modulus and the yield stressremained constant, despite a drop in the strain tobreak. Virgin and recycled PP matrix were filled withnanosized calcium carbonate (CaCO3) particles. Theeffect of morphology on the thermal and mechanicalproperties of nanocomposites of virgin and recycledPP filled with nanosized CaCO3 particles was alsostudied. The mechanical properties of the nanocompo-sites were strongly influenced by the intrinsic tough-ness of the matrix and the concentration and disper-sion of the filler. The yield strength and strain of virginPP decreased gradually, while its Young’s modulusincreased slightly with increasing CaCO3 loading.These phenomena were less pronounced for therecycled matrix. Incorporation of nanoparticles to vir-gin matrix produced an increase in tensile stiffnessand ductility, when good dispersion of the filler wasachieved. However, the impact strength dropped dra-matically for high filler contents. A significant increasein impact strength was observed for the recycledPP. POLYM. ENG. SCI., 50:1904–1913, 2010. ª 2010 Society ofPlastics Engineers

INTRODUCTION

Recycling has been practiced for many years by indus-

tries without any great accuracy. The new environmental,

economic, and petroleum crises have induced the scien-

tific community to increasingly deal with polymer reproc-

essing and sustainability. Numerous academic research

works have become more and more interested in the recy-

cling of the widespread polyolefins as polyethylene andpolypropylene (PP). Given that it is difficult to mimic the

real degradation undertaken during life service, manystudies focus on the recycling effects on the properties ofneat thermoplastics [1, 2]. Few articles have studied the

recycling of PP impact copolymer [3], postconsumedmaterials [4, 5], and filled thermoplastics [6, 7]. The

belief is that recycled polyolefins exhibit poorer perform-ances than that of neat ones as they undergo several types

of degradation, for instance, thermal degradation, mechan-ical degradation, and thermo-oxidative degradation, but infact final properties depend on the heterogeneity of the

structure, the compatibility, and the stability of polymercomponents, as well as the dispersion of extenders and

impurities.

The main objectives of recycling studies are selecting

the best stabilizer, amount, and types of modifiers that

could restore loss properties. Toughness and impact

strength are required properties for many industrial appli-

cations. Several theoretical and experimental studies have

investigated various types of thermoplastic toughness, and

have identified them as the combination of two phenom-

ena: particle cavitations and plastic extensibility of the

matrix material in the interparticle ligaments [8]. In fact,

the oriented polymers are generally ductile and they

undergo a ductile-brittle transition for low temperatures

and high strain speeds. To increase the properties of such

materials, it is interesting to minimize their plastic resist-

ance [9], the most effective method is to toughen poly-

mers with cavitated rigid particles to achieve a condition

of interparticle ligament dimension. Nanosized particles

confer significant property improvement with very low

loading levels, traditional microparticle additives requiring

Correspondence to: S. Pimbert; e-mail: sylvie.pimbert@univ-ubs.fr

DOI 10.1002/pen.21716

View this article online at wileyonlinelibrary.com.

VVC 2010 Society of Plastics Engineers

POLYMER ENGINEERING AND SCIENCE—-2010

much higher loading levels to achieve similar performan-

ces. These loads can generate cavitations or debonding,

transforming material into a cellular solid, and thus,

increasing ductility and tenacity [10]. Recent research

focusing on the toughening of PP [11, 12] indicate that

intrinsic properties of particles and matrix, including com-

patibility between PP and fillers, PP viscosity, surface

energy, molecular weight, and particle shape, are efficient

on thermoplastic toughness.

Recent studies undertaken by Kun Yang et al. [13] onthree kinds of PP (homopolymer PP, copolymer PP, and a

mixture of both) reinforced by nano inorganic carbonate,

highlight different toughening characteristics. Thio [14]reinforced PP with calcium carbonate (CaCO3) particles

of quite different sizes (0.04 lm, 0.7 lm, and 3.5 lm)and noticed that only 0.7 lm diameter particles gave good

reinforcement properties, resulting from a combined

mechanism of crack deflection toughening and local plas-tic deformation of interparticle ligaments following

matrix-particle debonding. Their observations where repli-cated by Zuiderduin et al. [15], who found that particle

sizes lower than 0.7 lm lowered the toughening effi-

ciency. Indeed particle aggregation enhanced by the greatamount of surface energy of smaller particle and/or large

particle sizes exhibited brittle fracture due to stress con-centration around particles.

The toughness and strength of recycled polyolefin

through the addition of polymers or fillers are largely

studied. Microfibrillar composites were recently developed

on the basis of polymer blends (virgin or recycled) and

showed higher impact strength [16]. Marisa [17] used

organoclays to toughen recycled polyethylene terephtalate

polymer. The modification of mechanical properties of

recycled PP from postconsumer containers was studied by

adding elastomer (ethylene-octene rubber) and microsized

CaCO3 [18]. Tzankova et al. [5] studied the effect of

glass fibres, CaCO3, and wood fibres on properties of

‘‘light fraction’’ from post consumer plastics.

The potential practical importance of a PP/nanoCaCO3

system and the possibility of significant enhancement of

stiffness and impact energy compared with neat matrix

made the choice of this system very attractive. The ques-

tion remains whether the toughening mechanism of neat

matrix is transferable to recycled matrix and to which

parameters is it related.

The aim of this article as follows:

• First, to study the effects of recycling on the thermomechan-

ical and rheological properties of a PP impact copolymer.

• Second, to explore if rigid nanoparticles could improve

the thermomechanical characteristics of a recycled PP

impact copolymer. Particularly, the impact strength of

nanocomposites will be investigated.

MATERIALS AND METHODS

PP PPC 7712 impact copolymer (grade 13 g/10 min)

generally used in the automotive industry was supplied as

pellets by Total Petrochemicals. Its main properties are

summarized in Table 1. It contains a small amount of talc

(3500 ppm) and glycerol mono stearate.

Surface-treated CaCO3 nanoparticles were produced by

Solvay Advanced Functional Minerals under the trade

name SocalTM 322. Particles are coated with 29 g/kg ste-

aric acid. They have a mean particle diameter of 50 nm

and a specific surface (BET) of 26 m2/g. CaCO3 nanopar-

ticles are noted NCC.

Sample Preparation

Recycled Samples. PP was injection molded five times

to simulate the recycling steps with a 90 ton DEMAG

injection-molding machine (screw diameter 45 mm, length

to diameter ratio L/D ¼ 20). The tensile test specimens

obtained were granulated in a knife mill for the next recy-

cling step. The temperature profile was 2008C–2108C–2208C, and 2208C for the nozzle. The injection pressure

was kept constant at 200 bars; the mold temperature was

set at 358C and a constant injection speed of 60 cm/s was

applied. Injection parameters were the same for all recy-

cling steps.

Samples are differentiated by following the symbol PP

with the recycling number. For instance, PPI5 is a five-

time recycled PP.

Nanocomposites. CaCO3 particles were dried in an

oven at 608C for 12 h. Virgin PP (PPV) and recycled PP

(PPI5) were melt compounded with, respectively, wt 3%,

wt 10%, and wt 20% fillers (noted 3%, 10%, and 20%

NCC) in a corotating twin-screw extruder Brabender1

(Brabender1, Duisburg, Germany) with a temperature

profile (200, 210, and 2208C) and a screw rotation speed

of 60 rpm to obtain a better dispersion of fillers. The PPV

and five-time recycled PP (PPI5) without fillers were also

extruded with the same processing conditions and then

pelletized and injected to ensure that they submit the

same cycle as composites.

CaCO3/PPV and CaCO3/PPI5 compounds were then

injection moulded with an Engel ES 80-35 machine

(screw diameter 30, L/D ¼ 22, temperature profile (200,

210, and 2208C), mould temperature 358C, injection pres-

TABLE 1. Physical characteristics of PPC 7712 polypropylene.

Properties Values

Density (g/cm3) (ISO 1183) 0.905

Ip 5.4

Mw (g/mol) 229,500

Tensile strength at yield (MPa) (ISO 527-2) 23

Melt Flow Index (g/10 min) (ISO 1133,2308C/2.16 kg) 13

Tensile modulus (MPa) (ISO 527-2) 1200

Charpy Impact Strength (notched) (kJ/m2) (ISO179) at 238C [40

Melting point (8C) 165

Ethylene monomer content 11%

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—-2010 1905

sure 200 bars). Dumb-bell tensile bars correspond to the

ISO 527-2 standard (cross section 4 3 10; type according

to Fig. 1).

Methods

Tensile properties were determined using a MTS Syn-

ergie RT1000 (MTS, Eden Prairie, MN) testing apparatus

equipped with a HTE extensometer with a nominal length

of 49.7 mm. Two crosshead speeds were used: 50 mm/

min for the elongation at break and 2 mm/min to calcu-

late the Young modulus. Tensile specimens were injection

molded according to ISO 529. At least eight samples were

conducted for each tensile test.

Two kinds of impact tests were used in this study:

Charpy impact tests for recycled samples and Izod impact

tests for nanocomposite samples.

Notched Charpy impact and Izod tests were performed

using a Tinuis Olsen machine at ambient temperature.

Specimens for Charpy test were cut from tensile speci-

mens using ISO 179 (80 mm 3 10 mm 3 4 mm); a

single-edge U-shaped notch (width ¼ 2 mm, depth ¼1.3 mm) was milled in the sample. A single-edge

V-shaped notch (width ¼ 4 mm, depth ¼ 2 mm, 458, tipradius ¼ 0.25 mm) was milled in the sample for the Izod

impact test. Notched Charpy impact was used for the

recycling study. This test was not adapted for testing

composites because samples were not broken with this

impact mode.

The rheological properties of PP were determined

using a Gemini 2000 from Bohlin Instruments (Cranbury,

NJ). The useful temperature range extends from 50 to

2508C with a precision of 60.18CAll experiments were carried out with parallel plate ge-

ometry of 25 mm diameter and temperature of 2208C.Tests in dynamic mode were performed with 2 mm gap,

5% strain, and a frequency range from 0.1 to 100 Hz.

Strain was fixed at 5% after determination of the linear

viscoelasticity range.

Differential scanning calorimetry (DSC) was carried

out with a Mettler-Toledo DSC 822 thermal analyzer

under nitrogen atmosphere. Microtomed sections of ten-

sile specimens were used as samples. The samples typi-

cally weighed 10 mg. Test conditions were as follows:

the samples were first heated at a rate of 108C/min from

25 to 2008C, held at this temperature for about 2 min and

then cooled from 200 to 258C at the same rate; a second

heating run was performed from 25 to 2008C under the

same conditions.

First heating was used to remove the thermal history

of the samples.

The crystal content percent was calculated according

to the following formula:

Crystal content ¼ DHfusion

DH1� 100 (1)

DHfusion: is the melting enthalpy of the sample as

obtained from DSC thermograms; for nanocomposite sam-

ples, this value is corrected by the weight fraction of PP

in the sample.

DH1: is the value of the enthalpy of fusion for a

100% crystalline PP. The value was taken as 209 J/g

[19].

The melting enthalpy is taken after the second heating

scan. At least two experiments were conducted for each

sample.

Scanning electron microscopy (SEM) observations

were conducted with a Jeol JSM 6460 LV electron scan-

ning microscope (voltage 15 kV). The fractured surface

of broken tensile specimens and the cryofractured surface

in nitrogen were sputter-coated with a thin layer of gold

in an Edwards Sputter Coater and then analyzed by SEM.

RESULTS AND DISCUSSION

Rheological Properties

Viscosity measurements are displayed in Fig. 2 in

which a Newtonian plateau at lower frequency can be

noticed. PP behaves as a Newtonian fluid. Comparison of

rheological curves corresponding to multiple injections

shows a decrease in the viscosity as the number of injec-

tion cycles increases; this can be correlated to a decrease

in the polymer average molecular weight corresponding

to the occurrence of b-scissions reactions [15]. In fact, the

FIG. 1. Dumb-bell tensile bar used for tensile tests (ISO 527-2

standard).

FIG. 2. Complex viscosity versus frequency for recycled PP samples.

1906 POLYMER ENGINEERING AND SCIENCE—-2010 DOI 10.1002/pen

highest molecular weight chains are preferentially broken

during multiple injections.

It should be emphasized that a MW decrease simulta-

neously causes a drop in tie molecule density [20] and an

increase in the crystallization because of the higher mo-

bility of the chains and its ability to fold.

Measurements in dynamic mode could explain the

viscoelastic behavior of PP. An observation of the com-

plex viscosity curves displayed in Fig. 2 shows that the

complex viscosity gradually decreases with repeated

injection because of the greater mobility of recycled

material produced by chain scissions. Similar results—that

is, a decrease in complex viscosity—have also been

reported from studies of the repeated extrusion of com-

mercial PP grades [4]. Modulus-frequency curves dis-

played in Fig. 3 indicate that for low frequencies, the loss

modulus G00 is greater than the storage modulus G0; the

material behaves as a viscous liquid until the crossover.

After reaching the crossover frequency, the material

behaves as a solid material. Curve inspections indicate

that both G0 and G00 decrease during repeated injections;

this can be explained by the higher mobility of the poly-

mer chains induced by a continuous mechanical degrada-

tion process. Despite a modulus decrease, the crossover

frequency remains constant at wc ¼ 340 rad s21, indicat-

ing that chain entanglements are not clearly affected by

repeated injections. We can conclude that although

repeated injections induce chain scissions resulting in

shorter molecule chains, PP still has several entanglement

points.

Structural changes in nanocomposites in the molten

state can be evaluated from frequency dependence of the

storage and loss moduli. Evaluation of dynamic moduli

gives information about the structure network and the

state of distribution of fillers in the matrix. The evolution

of storage modulus versus frequency is displayed in

Fig. 4, for PPV, recycled PP (PPI5), and the two related

nanocomposites with 3% of nano CaCO3 fillers. Recycled

and virgin matrices show a power-law dependence of

dynamic moduli at low frequency. The addition of CaCO3

to polymer melt shows an increase in the storage modulus

G0 because of the interaction between filler and matrix.

For the same filler content, the comparison of G0 curvesshows that in the case of recycled PP nanocomposite, G0

becomes nearly frequency-independent and shows a pla-

teau at low frequency that corresponds to a solid-like net-

work behavior. We could assume that a better distribution

and dispersion of nanofillers is obtained in the recycled

matrix at low contents with a significative influence on

the mechanical properties.

DSC Results

DSC results for the recycled PP samples are sum-

marized in Table 2; we can observe that both melting

enthalpy and crystallization increase, corresponding to

an increase in the degree of crystal content with the

number of injection cycles. Lower molecular weight

allows chains to fold and build more and more crystal

structure. The growth of crystal content degree can be

attributed to the recrystallisation of the molecule seg-

ments released by the scission of macromolecules with

repeated injections.

However, no significant evolution of the crystalliza-

tion temperature of PP is observed that is not in

accordance with the work of Aurrekoetxa [2]. We can

suppose that the PP degradation induced by only five

recycling steps compared with the 10 steps in [2] is

not sufficient to introduce enough impurities and conse-

FIG. 3. Complex modulus versus frequency for recycled PP samples.

FIG. 4. Storage modulus versus frequency for nanocomposites with 3%

wt CaCO3.

TABLE 2. DSC measurements on recycled PP samples.

Recycling

number DHC (J/g) DHm2 (J/g) Tm2 (8C) Tc (8C) v (%)

0 67.1 67.3 166.3 123.0 32.2

1 69.0 69.3 166.5 121.8 33.3

2 70.6 71.7 168.6 122.0 34.4

3 71.0 72.4 167.1 122.4 34.6

4 75.7 78.4 167.4 122.0 37.5

5 77.5 79.6 165.5 123.1 38.0

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—-2010 1907

quently heterogeneous nuclei leading to higher crystalli-

zation temperatures.

Thermal properties of the recycled and virgin nano-

composites are listed in Table 3. Tests are performed at

least twice for reproducibility.

No differences of PP melting and crystallization tem-

peratures can be observed between nanocomposites pre-

pared with respectively virgin and recycled matrix. Con-

sidering the previously reported nucleant effect of talc

present in PP, CaCO3 does not show any additional effect

on PP crystallization [15].

However, the crystal content percent of PPV blended

with nanofillers increases to reach 40% compared with

33% for pure PPV; it appears that the presence of nano-

sized CaCO3 in the PP matrix has a significant effect on

increasing the crystalline domain.

A small decrease in the crystal content of recycled

blends, compared with that of virgin blends, is observed

up to 10% fillers. It may due to an effect of the nanopar-

ticles that restrict molecular movements and hinder the

packing of molecule segments. For the highest filler con-

tent (20%), the recycled PP crystal content reaches 40%,

as for the PPV nanocomposite with 20% nanofillers. This

different crystallization behavior of virgin and recycled

PP matrix could be explained by the filler-matrix interac-

tions and the blend morphologies, affecting the interfacial

adhesion.

Mechanicals Tests and MEB Observations

Tensile tests conducted on virgin and recycled PP

show a ductile behavior. At low tensile strain, semicrys-

talline polymer exhibits an elastic deformation until yield

stress. Then plastic flow is generally followed by whiten-

ing in the necking zone. After reaching the plastic zone,

lamellae formed by folded chains are broken into small

blocks, the alignment along the drawing direction forming

a microfibrillar structure. Continued deformation of

microfibrillar structure is extremely difficult because of

high strength of individual microfibrils.

The mechanical characteristics of PP during recycling

steps are summarized in Table 4. All PP samples exhibit

regions of elastic yielding, and plastic deformation

accompanied by cold drawing. One may note that PPV

and recycled PP samples have similar elastic and yield

domains but a reduction of the drawing behavior that may

be because of the higher crystal content is observed for

the recycled samples, as well as a decrease in the stretch-

ing of fibrils and microfibrils. All samples show stress-

whitening as a consequence of craze nucleation.

The Young modulus remains constant no matter the

number of recycling steps, although the increase in crystal

content observed is expected to also affect the stiffness of

recycled material. Yield stress does not show any signifi-

cant changes. The observed degradation process slightly

affects the small strain properties (yield stress and modu-

lus) [21].

Elongation at break, as seen in Fig. 5, decreases gradu-

ally with the recycling steps. This drop is correlated both

to the increase in crystal content, the molecular weight,

and the chain length decrease.

According to Aurrekoetxa [2], the shortest chains

involve more ends in the structure and consequently there

are less chains to sustain stress during tensile loading,

TABLE 3. DSC results for virgin PP and recycled PP nanocomposites.

Tm2 (8C) Tc (8C) DHC (J/g PP) DHm2 (J/g PP) v (%)

PPV (extruded and injected) 166.4 121.8 69.1 69.3 33

PPV 3% NCC 167.4 121.9 73.8 75.1 36

PPV 10% NCC 167.4 122.2 78.4 82.1 39

PPV 20% NCC 167.7 122.4 83.5 84.2 40

PPR (PPI5 extruded and injected) 166.2 122.6 77.5 78.1 37

PPR 3% NCC 166.2 122.7 77.2 66.5 32

PPR 10% NCC 167.1 121.3 69.2 72.2 34

PPR 20% NCC 165.8 122.7 87.5 86.5 41

DHm2, second melting enthalpy (J/g PP); DHC, crystallization enthalpy (J/g PP); Tm2, second melting temperature (8C); Tc, crystallization temperature

(8C); v, crystal percent (%).

TABLE 4. Mechanical characteristics of recycled PP samples.

Sample Young Modulus (MPa) Yield stress (MPa) Stress at break (MPa) Strain at break (%) Impact energy (kJ/m)

PPI1 958 (SD: 19) 23.0 (SD: 0.6) 18.5 (SD: 0.3) 125 (SD: 20) 38 (SD: 3)

PPI2 965 (SD: 7) 23.0 (SD: 0.7) 18.4 (SD: 0.7) 106 (SD: 15) 36 (SD: 3)

PPI3 970 (SD: 56) 23.5 (SD: 0.4) 18.5 (SD: 0.5) 91 (SD: 18) 35 (SD: 3)

PPI4 970 (SD: 23) 23.5 (SD : 0.4) 18.5 (SD: 0.5) 74 (SD: 17) 33 (SD: 2)

PPI5 980 (SD: 36) 23.5 (SD : 0.2) 18.3 (SD: 0.3) 67 (SD: 23) 32 (SD: 2)

SD, standard deviation.

1908 POLYMER ENGINEERING AND SCIENCE—-2010 DOI 10.1002/pen

causing failure at smaller elongation. Furthermore, long

chains have more C��C linkages to stretch. The decrease

in tie molecule density, which constitutes bridges between

lamellar crystals and acts as transmitters of stress among

lamellar could also explain the reduction of elongation at

break. The consolidation zone is reduced involving a

decrease in ductility behavior.

Table 5 presents the mechanical characteristics of the

different PP nanocomposites. We can notice an increase

in the PP Young modulus and a decrease in yield stress

versus filler content, both for virgin and recycled PP, but

in a less significant way than is observed in some studies

[22]; these differences could be justified by the processing

parameters, such as screw configuration and speed, extru-

sion temperature, and also blending fractions. For the

blends with PPV, the higher modulus values could be jus-

tified by a reinforcement effect of nanofillers and also by

the increase in the PP crystal content, as observed by Van

der Wal [23]. Elasticity improvement could also be

related to the reduction in spherolite size as demonstrated

by Zhang et al. [22]

Recycled PP blends also show a slight enhancement

of the rigidity, less marked than for PPV blends, until

a critical weight fraction of 20% for which a modulus

decrease is observed; this may be explained by a phe-

nomenon of nanoparticle agglomeration. Agglomeration

of surface-modified fillers could form a weak structure;

the interactions between fillers and matrix decrease,

facilitating debonding. This limited increase in rigidity

for recycled PP composites is also confirmed by Bra-

chet et al. [18].

At low filler content, virgin matrix exhibits a duc-

tile deformation resulting in a continuous deformation

of microfibrils into the loading direction. The low-

magnification SEM micrographs of the fractured sur-

face of PPV 10% NCC nanocomposites are presented

in Fig. 6 along with the corresponding high-magnifica-

tion micrograph of fractured zones. The microstructure

of the sample appears porous (Fig. 5B) and consists in

highly deformed fibrils (Fig. 6A) with large and shal-

low voids.

Because of the high amount of hydrostatic stress pres-

ent in the centre of the specimen fibrillation, which is the

drawing of a network of fibrils is concentrated in the

centre of the fractured surface (Fig. 6A). Long fibrils,

located between debonded particles, are extremely elon-

gated in the drawing direction, illustrating an extensive

plastic deformation (Fig. 6D and E). Severely deformed

fibrils are expected to scatter the light more and exhibit

higher stress. Some aggregates seen in Fig. 6D could stop

the stretching of fibrils.

SEM micrographs of fractured surfaces of PPR/CaCO3

nanocomposites are presented in Fig. 7 in which fibrilla-

tion (Fig. 7A) and a brittle mode of deformation (Fig. 7B)

can be observed. Smooth surface characterizes a brittle

mode of strain, probably because of the intrinsic proper-

ties of the recycled matrix and to a lower density of tie

molecules. Although the region appears to be brittle, there

are a large number of unbroken fibrils that continue to

plastically deform.

The percentage of fibrillation is lower in this case

compared with the nanocomposites prepared with a virgin

matrix. It is suggested that the higher percentage of fibril-

lation is related to the higher crystal content and elonga-

tion to break of PP in the nanocomposite with a virgin

matrix compared with a recycled one.

The high-magnification SEM micrographs presented in

Fig. 7D and E show voids in both fibrillated and brittle

regions of fractured surface and detachment of particles

from the matrix, which results in stress whitening of the

FIG. 5. Elongation at break and Charpy impact strength vs. cycle num-

ber.

TABLE 5. Mechanical characteristics of virgin and recycled PP nanocomposites.

Young Modulus (MPa) Stress at yield (MPa) Strain at yield (%) Strain at break (%)

Virgin PP 969 (32) 21.7 (0.3) 7.0 (0.2) 158 (34)

PP 3% NCC 1081 (40) 21.1 (0.5) 6.1 (0.1) 160 (47)

PP 10% NCC 1136 (24) 20.8 (0.2) 5.8 (0.1) 82 (34)

PP 20% NCC 1253 (54) 19.7 (0.1) 4.1 (0.2) 38.6 (14.9)

Recycled PP 1025 (19) 19.9 (0.4) 6.3 (0.2) 61 (7.8)

PPR 3% NCC 1145 (16) 19.8 (0.4) 5.7 (0.4) 47.3 (9.5)

PPR 10% NCC 1122 (41) 20.0 (0.3) 5.0 (0.3) 36 (3.7)

PPR 20% NCC 992 (38) 17.6 (0.4) 5.4 (0.1) 43.4 (8.7)

Standard deviation between brackets.

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—-2010 1909

nanocomposite. It may be noted that the high rigidity of

the polymeric matrix could probably facilitate the debond-

ing of mineral particles with formation of a high density

of voids in the brittle area of the fractured surface.

For some samples, a lot of deviation of the strain at

break values can be noticed, reaching up to 400%, and

corresponding to an increase in macroscopic ductility; for

some other samples, failure takes place at lower strain. It

FIG. 6. SEM micrographs of fracture surface of PPV nanocomposites with 10% CaCO3.

FIG. 7. SEM micrographs of fracture surface of PPR nanocomposites with 10% CaCO3.

1910 POLYMER ENGINEERING AND SCIENCE—-2010 DOI 10.1002/pen

may be due to a nonuniform dispersion of particles in the

PP matrix although nano CaCO3 is surface modified. The

decrease in ductility despite the addition of CaCO3 indi-

cates a reduction in the matrix deformation because of the

introduction of mechanical stresses by the nondeformable

filler particles.

Impact Strength

The notched Charpy impact test results are presented

in Fig. 5, which shows that the absorbed energy by the

recycled PP decreases gradually with repeated injections

corresponding to a decrease in fracture toughness and

ductility. Toughness is defined as the ability of a material

to delocalize strain [24] and is evaluated by impact

energy.

These changes can be related to the decrease in PP

molecular weight as observed in Fig. 2 and expected by

[25]. According to Meijer [24], higher molecular weights

are beneficial for properties, such as strain-to-break and

impact strength. Aurrekoetxa et al. [26] show that the

high impact energy of PPV is because of its low crystal

content and high molecular weight that lead to a larger

number of crazes. Considering that the PP impact copoly-

mer is a heterogeneous structure consisting of a PP homo-

polymer, an ethylene-propylene rubber and a relatively

small portion of PE homopolymer phases, the decrease in

impact strength may be the consequence of changes in

the rubbery domains responsible for toughness in a solid

polymer [10].

Cryofacture surfaces of the PPV nanocomposites at

3, 10, and 20% of fillers are, respectively, exhibited in

figures (Fig. 8E, C, and A); they show many heteroge-

neities and aggregates with diameters of more than 300

nm for the nanocomposite with 10% of particles. Inter-

particle distances decrease as a function of weight frac-

tions.

The curve of Fig. 9 clearly indicates that the incorpora-

tion of nanosized CaCO3 to PPV decreases the Izod-

notched impact strength. This decrease of the absorbed

energy may be because of different factors, such as the

kind of the PP matrix (ethylene-propylene copolymer),

the size, distribution, and dispersion of fillers, and also

the interfacial adhesion between the filler and the polymer

[27]. Rothon [28] estimates that in the case of a PP copol-

ymer and an impact-modified PP, impact strength is low-

ered by hard particulate fillers. This is because the fillers

interact with the soft or rubbery component and nullifies

its ability to absorb energy during impact.

Recycled PP nanocomposites behave differently with

an increase in impact strength from 8 kJ/m2 for PPR with-

out fillers to 18 kJ/m2 for PPR with 20% CaCO3, as seen

in Fig. 10. The impact strength values remain quite con-

stant up to 10% fillers and then sharply increase for 20%

fillers. This behavior may be due to a weak adhesion

between the recycled matrix and nano CaCO3 particles

since debonding matrix-particle facilitates the plastic

stretching of the polymer ligaments between the filler

particles.

Large clustering is a basic imperfection because of the

flow channeling in the usual forms of industrial polymer

processing practices and to the great energy surface of

FIG. 8. Cryofactured surface of virgin and recycled PP nanocompo-

sites. PPV: (A) 20% CaCO3, (C) 10% CaCO3, (E) 3% CaCO3; PPR: (B)

20% CaCO3, (D) 10% CaCO3, (F) 3% CaCO3.

FIG. 9. Izod impact strength and Young’s modulus vs. CaCO3 wt% of

PPV nanocomposites.

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—-2010 1911

nanoparticles that furthers the agglomeration as seen in

(see Fig. 8).

Cryofacture surfaces of recycled PP nanocomposites,

presented in figures (Fig. 8F, D, and B), show narrower

interparticle distances for lower filler contents.

The polymer with the highest viscosity requires more

energy and shear forces to disperse nanoparticles than the

lowest one. Matrix with the lowest viscosity (PPR) seems

to have a better dispersion and distribution of the particles

than the matrix (PPV) with the highest melt viscosity.

There are more aggregates in the PPV nanocomposites

than in the PPR nanocomposites. At higher filler content

20% (Fig. 8A), particles are so huge that they build great

agglomerates that influence ultimate properties.

We should emphasize that crystal content and mor-

phology of nanocomposites are not the only factor influ-

encing the impact strength; the intrinsic strength of the

matrix plays an important role on the composite strength

and toughness. The dispersion of rigid fillers in a rubber

phase could improve the compatibility between the fillers

and the matrix [29, 30]. A brittle matrix could be tough-

ened by the addition of a rubbery component [30].

Forthcoming work must be conducted on the character-

ization of ternary blends by adding a rubber to recycled

PP-CaCO3 nano- or microcomposites to obtain composites

with higher impact strength and rigidity.

CONCLUSIONS

Recycling steps consisting in successive injection

moldings have a crucial effect on the thermomechanical

properties of a PP copolymer with some important effects

on its toughness and impact strength. Modulus and yield

stress remain constant. The rheological and DSC results

show both a decrease in PP molecular weight density and

an increase in its crystal content induced by chain scis-

sion.

The thermal and mechanical properties of virgin and

recycled PP filled with rigid CaCO3 nanoparticles are

compared. The fillers do not appear to have any effect on

PP crystallization.

An increase in Young modulus with increasing CaCO3

content is observed for both matrices whereas yield

strength decreases slightly for the highest CaCO3 content.

Nanocomposites prepared with PPV matrix show a

decrease in impact resistance; this can be explained by a

bad dispersion of fillers despite a pretreatement by stearic

acid; an increase in impact strength, even with the highest

filler content, is observed for the nanocomposites prepared

with recycled PP; this may be explained by the lowest

viscosity of PP allowing for a better dispersion of the

nanoparticles. At the highest filler content (20%) the

behavior of nanocomposites is mainly influenced by the

presence of aggregates. The effect of PP matrix (virgin or

recycled) is minor.

Fracture in PP nanocomposites occurs by mixed modes

consisting of fibrillation and brittle mode.

REFERENCES

1. H.M. Da Costa, V.D. Ramos, and M.C.G. Rocha. Polym.Test, 24, 86 (2005).

2. J. Aurrekoetxa, M.A. Sarrionandia, and I. Urrutibeascoa. J.Mat. Sci., 36, 2607 (2001).

3. J. Tochacek, J. Jancar, J. Kalfus, P. Zborilova, and Z.

Buran, Polym. Degrad. Stab., 9, 770 (2008).

4. A. Jansson, K. Moller, and T. Gevert, Polym. Degrad. Stab.,82, 37 (2003).

5. N.Tzankova Dintcheva, F.P. La Mantia, F. Trotta, M.P.

Luda, G. Camino, M. Paci, L. Di Maio, and D. Acierno,

Polym. Adv. Technol., 12, 552 (2001).

6. M.R. Thompson, and K.K. Yeung., Polym. Degrad. Stab.,91, 2396 (2006).

7. A. Bourmaud and C. Baley. Polym. Degrad. Stab., 92, 1034(2007).

8. Z. Bartczak, A.S. Argon, R.E. Cohen, and M. Weinberg,

Polymer, 40, 2347 (1999).

9. A.S. Argon and R.E. Cohen, Polymer, 44, 6013 (2003).

10. A. Van der Wal and R.J. Gaymans, Polymer, 40, 6067 (1999).

11. S.P. Bao and S.C. Tjong, Composites Part A, 38, 378 (2007).

12. Y. Liu and M. Kontopoulou, Polymer, 47, 7731 (2007).

13. K. Yang, Q.Y. Guangxian Li, Y. Zhang, and P. Zhang,

Polym. Eng. Sci., 47(2), 96 (2007).

14. Y.S. Thio, A.S. Argon, R.E. Cohen, and M. Weinburg,

Polymer, 43, 3661 (2002).

15. W.C.J. Zuiderduin, C. Westzaan, J. Huetink, and R.J. Gay-

mans, Polymer, 44, 261 (2003).

16. R.J. Shields, D. Bhattacharyya, and S. Fakirov, J. Mater.Sci., 43, 6758 (2008).

17. M.T.M. Bizarria, A.L.F.de.M. Giraldi, C.M. de Carvalho,

J.I. Velasco, M.A. d’Avila, and L.H.I. Mei, J. Appl. Polym.Sci, 104, 1839 (2007).

18. P. Brachet, L.T. Høydal, E.L. Hinrichsen, and F. Melum,

Waste Manag., 28, 2456 (2008).

19. R. Greco, C. Mancarella, E. Martuscelli, G. Ragosta, and Y.

Jinghua, Polymer, 28, 1929 (1987).

FIG. 10. Izod impact strength and Young’s modulus vs. CaCO3 wt% of

recycled PP nanocomposites.

1912 POLYMER ENGINEERING AND SCIENCE—-2010 DOI 10.1002/pen

20. J. Karger-Kocsis, Macromol. Symp., 143, 185 (1999).

21. H.M. da Costa, V.D. Ramosa, and M.G. de Oliveira, Polym.Test, 26, 676 (2007).

22. Q.X. Zhang, Z.Z. Yu, X.L. Xie, and Y.W. Mai, Polymer,45, 5985 (2004).

23. A. Van der Wal, J.J. Mulder, J. Oderkerk, and R.J. Gay-

mans, Polymer, 26, 6781 (1998).

24. H.E.H. Meijer and L E. Govaert, Prog. Polym. Sci., 30, 915 (2006).

25. T.R. Crompton, Polymer Reference Book, Smithers RapraTechnology, England (2006).

26. J. Aurrekoetxa, M.A. Sarrionandia, and I. Urrutibeascoa, J.Mater. Sci., 36, 5073 (2001).

27. K. Yang, Q.Y.G. Li, Y. Sun, and D. Feng, Mater. Lett., 60,805 (2006).

28. N. Roger Rothon, Ed., Particulate-Filled Polymer Compo-sites, 2nd ed., Smithers Rapra Technology, England (2003).

29. J. Yu, G. Wang, J. Chen, X. Zeng, and W. Wang, Polym.Eng. Sci., 201 (2007).

30. C.J.G. Wang, X. Zeng, H. Zhao, D. Cao, J. Yun, and C.K.

Tan, J. Appl. Polym. Sci., 94, 796 (2004).

DOI 10.1002/pen POLYMER ENGINEERING AND SCIENCE—-2010 1913