Available online at www.sciencedirect.com

www.elsevier.com/locate/wasman

Waste Management 28 (2008) 2456–2464

Modification of mechanical properties of recycled polypropylenefrom post-consumer containers

P. Brachet a,*, L.T. Høydal a, E.L. Hinrichsen a, F. Melum b

a SINTEF, Materials and Chemistry, P.O. Box 124, Blindern, Oslo, NO 01314, Norwayb NTNU, Programme for Industrial Ecology, Trondheim, NO 7491, Norway

Accepted 31 October 2007Available online 14 January 2008

Abstract

This study is conducted to look at the modification of mechanical properties of recycled polypropylene (PP) from post-consumer con-tainers with the addition of stabilizers, elastomer (ethylene–octene rubber, EOR) and calcium carbonate (CaCO3). The mechanical andthermal properties of the blends were evaluated. The results showed limited changes with the addition of elastomer and calcium carbon-ate on the mechanical properties of the recycled polypropylene. Some differences were observed, but the trends were not reproducibleover the different compositions. DSC analysis confirmed the presence of polyethylene (PE) in the recycled polypropylene. The polyeth-ylene impurity and the presence of many different qualities of polypropylene in the recycled material may have prevented any possibleimprovement in the mechanical properties by the addition of EOR and CaCO3, improvements seen in previous studies on virgin poly-propylene. The compatibility of the different homopolymers and copolymers of PP used in consumer packaging is not known, while poly-ethylene and polypropylene are known not to be miscible with each other. The mixture of qualities and materials may explain such apoor blending. Reusing and upgrading of recycled PP from post-consumer containers would therefore first require a better sorting ofthe post-consumer waste. The use of an adequate compatibilizer that would allow a uniform and homogeneous blending of the raw mate-rial mixture might enhance the mechanical properties.� 2007 Elsevier Ltd. All rights reserved.

1. Introduction

The use and demand of polypropylene is increasing at avery fast pace. In 2000, polypropylene represented 23% ofthe thermoplastic consumed in Western Europe. Its sales intonnage is the third most important amongst plastics in theworld (Zebarjad et al., 2006). Indeed polypropylene can beproduced from low-cost petrochemical raw material, mak-ing it an inexpensive thermoplastic, relative to others, whileshowing good processability and reasonably high mechan-ical properties. However, due to its chemically stabilizedstate for long service life and high volume-to-weight ratio,it is one of the most visible forms of waste in landfills. Thepressure to recycle has increased significantly for economic,political and environmental reasons in today’s society.

0956-053X/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.wasman.2007.10.021

* Corresponding author. Tel.: +47 98 24 39 47; fax: +47 22 06 73 50.E-mail address: Philippe.brachet@sintef.no (P. Brachet).

Waste is perceived as a major problem, especially for highconsumption plastics such as polypropylene (Papaspyridesand Poulakis, 1996). The recycling process of post-con-sumer waste has appeared to be complicated because ofthe necessity of materials separation, as well as lower mate-rial properties (strength, stiffness, stability. . .) of recycledmaterials. It is hard to conduct and maintain a constantquality of recycled material coming from household wastebecause of the many types and grades of polymers used. Onthe contrary, materials used in car bumpers, drink bottles,and drink caps for instance follow a closed recycling loopwhich assure their quality. Some studies show that up to10% of foreign materials can be found in recycled polypro-pylene despite being sorted centrally by trained personnel(Seiler, 1995).

Techniques, for instance automatic sorting facilitiesusing infra red technology, can be used to reduce theamount of impurities, but their use increases the cost signif-

P. Brachet et al. / Waste Management 28 (2008) 2456–2464 2457

icantly. Furthermore recycled polypropylene undergoesthermo-mechanical degradation events due to high temper-ature and shearing during processing (Goldberg and Zai-kov, 1987; Gonzalez et al., 1998; Tiganis et al., 1996).Similarly, recycled polypropylene has lower fracture prop-erties (Aurrekoetxea et al., 2001a,b) and heat ageing (Gah-leitner, 2002) compared to virgin material. Virginpolypropylene homopolymer has a rather low modulus, alimited impact resistance and high notch sensitivity (Yanget al., 2006). To improve the mechanical and thermal prop-erties, additives, such as mineral fillers and rubbers or elas-tomers, are blended with polypropylene (Yang et al., 2006;Chan and Wu, 2002; Wang et al., 2002; Levita et al., 1989;Ma et al., 2005; Chen et al., 2004). Elastomer such as eth-ylene–propylene rubber (EPR) is used to enhance tough-ness and impact resistance, but has the side effect of alsoreducing the yield strength and Young’s modulus. To com-plement as well as amplify the effect of the elastomer con-tent, inorganic fillers are added. Calcium carbonate,CaCO3, provides increased impact energy, improved hard-ness, and higher modulus and tensile strength, while lower-ing the cost (Jang, 1992; Nam et al., 2001; Premphet andHoranont, 2001). Consequently, the addition of both elas-tomer and filler in the polypropylene affects the mechanicalproperties of the hybrid material depending on their com-position, but also on their phase morphology (Zebarjadet al., 2006). Literature refers to two separate possibilitiesfor where the inorganic fillers end up, the elastomer parti-cles and rigid particles are scattered in the polypropylenematrix independently, or the elastomer particles attractand absorb the rigid particles (Jancar et al., 1993; Premphetand Horanont, 2000; Barbosa et al., 2000). While usingadditives on virgin polypropylene improves the mechanicaland thermal properties of the material, the phase morphol-ogies would appear to be critical for recycled polypropyl-ene, where impurities such as other polymers and paperare likely to be an important factor to take into accountduring the blending with additives. The objective of thisstudy is to look at the mechanical properties obtained withthe blending of a stabilized recycled polypropylene frompost-consumer waste with different compositions of ethyl-ene–octene rubber, EOR, and calcium carbonate, CaCO3.The idea is to enhance the properties of the recycled poly-propylene sufficiently for an intended technical application,without losing the ecological and economic advantages ofthe recycled material. The application in mind during thisstudy is the buckets used in the Tomra Recycling Centre(TRC). Tomra produces reverse vending machines forempty drink bottles and containers. The Tomra RecyclingCentre is developed to help consumers recycle and to facil-itate the return of several types of used, rigid packaging,including PET, PP and PE. Tomra wants to close materialloops by using recycled packaging materials in applicationsin the Recycling Centre. The buckets in the TRC requiremost of all high impact resistance due to the high numberof products (plastic bottles, glass bottles, tins, cans, andrigid packaging) falling into the buckets. The buckets also

require high stiffness due to the high temperature within theCentres in the summer, high wear resistance and resistanceto fatigue. This study investigates the possible upgrading ofmaterial properties of the recycled PP in order to fulfil therequirements for the buckets. This study investigates alsothe morphology, thermal properties, and fracture surfaceproperties of recycled polypropylene, by means of calorim-etry and microscopy analysis, in order to collect valuableinformation about the characterization of recycled poly-propylene blends.

2. Experimental procedures

2.1. Materials

Different blends were prepared using recycled polypro-pylene from post-consumer waste. The recycled polypro-pylene was delivered from Expladan in Denmark. Acombination of two kinds of stabilizers was used: IrganoxB215 FF, a processing stabilizer and Chimassorb 944FD, a hindered amine stabilizer (HAS), used for long termthermal properties. The rubber used during these tests isEngage 8401 from Dow Chemical Company, a polyolefinelastomer produced by metallocene technology which is acopolymer of the a-olefin octene and ethylene. Engage8401 has a melt flow rate, MFR, of 30 g/min (190 �C,2.16 kg) and a density of 885 kg/m3. The calcium carbonateused is Danchalk PC CaCO3. The particles have a meandiameter of 3 lm and are surface coated with 2% stearicacid, a low molecular weight organic compound. The stea-ric acid on the filler particles reduces the particle:particleinteraction, leading to a better dispersion of the particlesin the PP matrix (Suetsugu and White, 1987; Jancar andKucera, 1990).

2.2. Blending preparation

Compounding of the materials was performed on aClextral co rotating twin screw lab extruder with a set ofstandard polyolefin screws. Each screw has two reversedelements (back flow) and two kneading blocks for a betterblending of the materials. The length of the screws was800 mm with a diameter of 25 mm. In the extrusion set,the barrel temperatures varied from 180 to 215 �C, andthe production speed was 100 g/min. A masterbatch ofpolypropylene and stabilizers was produced prior to theblends in order to facilitate and compose an even disper-sion of stabilizers in the polypropylene matrix. Eight differ-ent blends with different compositions were made (Table1). Consequently, the recycled polypropylene was blendedtwice for the blends containing EOR and CaCO3.

2.3. Specimen preparation

Test specimens were injection moulded on a BattenfeldBSKM 45/20 HY (2 mm thick specimens) or a DSM miniextruder/injection moulder (4 mm thick specimens) at

Table 1Composition of blends showing weight percentage of stabilizers, elastomerand calcium carbonate used relative to the weight of the recycled PP inthese blends

Sample composition Chimassorb(%)

Irganox(%)

Elastomer(%)

CaCO3

(%)

PP + S 0.15 0.15 – –PP + S + 5%E 0.15 0.15 5 –PP + S + 5%E +10%C 0.15 0.15 5 10PP + S + 5%E + 20%C 0.15 0.15 5 20PP + S + 10%E + 10%C 0.15 0.15 10 10PP + S + 10%E + 20%C 0.15 0.15 10 20PP + S + 10%C 0.15 0.15 – 10PP + S + 20%C 0.15 0.15 – 20

Table 2Density and melt flow index for all blends

Sample composition Density (kg/m3) Melt flow rate(g/10 min)

PP + S 908 14.6PP + S + 5%E 906 19.7PP + S + 5%E +10%C 968 14.4PP + S + 5%E + 20%C 1033 14.6PP + S + 10%E + 10%C 973 16.1PP + S + 10%E + 20%C 1029 15.8PP + S + 10%C 974 14.3PP + S + 20%C 1038 14.5

2458 P. Brachet et al. / Waste Management 28 (2008) 2456–2464

230 �C. All test samples were conditioned for 48 h at 23 �Cand 50% humidity prior to testing.

3. Methodology

3.1. Density and melt flow index

The density and melt flow index were measured for alleight blends. The measurement of the melt flow rate wasperformed according to standard ISO 1133 at 230 �C, witha load of 2.16 kg. The density was measured according tostandard ISO 1183, using the materials extruded duringthe melt flow rate test.

3.2. Tensile tests

Tensile tests were performed at three different tempera-tures, at room temperature, at �20 �C and at 60 �C. Thetensile tests were measured in accordance with ISO 527.At room temperature the tests were performed on aZWICK tensile tester model Z250 with a load cell of50 kN and an Universal extensometer ‘‘Multisens” (accu-racy grade 1 to EN ISO 9513). At �20 �C, and at 60 �C,the tests were carried out on a SCHENCK TREBELRM 100 tensile tester with a load cell of 100 kN; a clip-on extensometer MTS 634.11F-55 was attached to thesamples during the tests (maximum extension 20%). Theinitial speed during testing was set at 1 mm/min and froma stress of 5 MPa the speed was increased to 50 mm/min.The data sampling frequency used during the tests was setat 50 Hz.

3.3. Charpy impact tests

Notched impact tests were carried out using a RosandInstrumented Falling Weight Impact Tester, type 4. Thetests were performed for all specimens at �20 �C. The sam-ples were kept inside the temperature chamber for 30 minprior to testing in order to get the desired temperature.The rectangular shaped samples had a 2 mm notch withan angle of 45�. The tests were carried out according tothe standard ISO 179.

3.4. Ageing test

The ageing tests were conducted according to a methodbased upon ASTM D3045 ‘‘Standard practice for heataging of plastics without load”. Five blends of recycledpolypropylene and one virgin polypropylene were tested.Standard ‘‘dog bone” samples were used (cross section of12 mm2). Through a hole at one end of each sample, ametallic cable was used to hold to samples. The rigid cablewas attached to the top of the autoclave; therefore, thesamples were not in contact with each other and did nottouch the bottom of the autoclave. A sample of each mate-rial was put in autoclaves at temperatures of 120 �C,130 �C, 140 �C and 150 �C. (The test was repeated twice,the temperature was controlled with Eurotherm controllersand an electronic thermocouple, and was accurate within1 �C.) The autoclaves were covered with glass wool andaluminium sheets, and therefore were not sealed. Over timethe samples were inspected manually to detect failure,which was done while holding the sample in both endswhile bending slightly the sample 20–30� at the most.

3.5. Differential scanning calorimeter testing

Differential scanning calorimeter (Perkin–Elmer Pyris 1,DSC) analysis was conducted to determine the meltingpoint of selected blends. The instrument was calibratedby measuring the melting point of indium. All measure-ments were performed under nitrogen flow. The tests wereconducted as follows: the samples are heated from 30 �C to250 �C, then cooled from 250 �C to 30 �C and reheated asbefore to 250 �C. The heating and cooling rate was set at20 �C per min. To erase differences in initial morphology,the melting point is measured on the second heating.

4. Test results

4.1. Density and melt flow index

Table 2 shows the density and melt flow index for thedifferent blends of recycled polypropylene. Eight tests wereperformed for each material for better accuracy. The den-sity increases significantly with the addition of calcium car-bonate, with a density of 908 kg/m3 for the recycled

P. Brachet et al. / Waste Management 28 (2008) 2456–2464 2459

polypropylene and stabilizers only, to 1029 kg/m3 for thepolypropylene, stabilizers and 20% calcium carbonate.The addition of elastomer does not affect the density; how-ever, it increases the melt flow index, 14.6 g/10 min for thepolypropylene and stabilizers only to 19.6 g/10 min for thepolypropylene, stabilizers and 5% EOR. The addition ofcalcium carbonate does not affect the melt flow rate buttends to reduce it in combination with the elastomer.

4.2. Tensile tests at different temperatures

Table 3 represent the tensile properties of polypropyleneat �20 �C, 23 �C and 60 �C with different composition ofcalcium carbonate and elastomer blended in. Nine speci-mens of each composition were tested. The Young’s mod-ulus was measured between 1 and 3 MPa. The results at�20 �C and 23 �C show limited influence of the calciumcarbonate on the Young’s modulus, while there is a limitedreduction with the addition of elastomer. The data shows agreat degree of variability for all blends, especially forthose with a large amount of calcium carbonate. Similarlyas the concentration of calcium carbonate increases, theyield stress decreases slightly, while the addition of elasto-mer shows limited reduction. The maximum elongationdoes not reveal significant changes but shows, however, agreat degree of variability, which does not allow a properanalysis. Necking is observed for all samples. The changesobserved during these tests are limited, which is contradic-

Table 3Tensile tests performed at �20 �C (a), +23 �C (b) and +60 �C (c)

Sample composition Young’s modulus (Mpa) Yield st

(a)PP + S 3400 ± 200 43.8 ± 0PP + S + 5%E 2800 ± 300 43.5 ± 0PP + S + 5%E + 10%C 3300 ± 600 41.4 ± 0PP + S + 5%E + 20%C 3100 ± 500 39.0 ± 0PP + S + 10%E + 10%C 2800 ± 300 39.8 ± 0PP + S + 10%E + 20%C 3000 ± 700 37.9 ± 0PP + S + 10%C 3300 ± 300 42.4 ± 0PP + S + 20%C 3700 ± 700 40.4 ± 0

(b)PP + S 1010 ± 60 24.6 ± 0PP + S + 5%E 910 ± 30 23.6 ± 0PP + S + 5%E + 10%C 970 ± 40 22.2 ± 0PP + S + 5%E + 20%C 1080 ± 120 20.4 ± 0PP + S + 10%E + 10%C 890 ± 50 20.4 ± 0PP + S + 10%E + 20%C 960 ± 80 19.5 ± 0PP + S + 10%C 1140 ± 40 22.9 ± 0PP + S + 20%C 1250 ± 60 21.6 ± 0

(c)PP + S 590 ± 140 14.5 ± 0PP + S + 5%E 390 ± 40 13.1 ± 0PP + S + 5%E + 10%C 410 ± 80 12.7 ± 0PP + S + 5%E + 20%C 380 ± 40 12.1 ± 0PP + S + 10%E + 10%C 320 ± 30 12.2 ± 0PP + S + 10%E + 20%C 340 ± 20 11.3 ± 0PP + S + 10%C 420 ± 70 13.9 ± 0PP + S + 20%C 450 ± 60 13.6 ± 0

The errors given are standard deviations.

tory to previous studies (Zebarjad et al., 2006; Premphetand Horanont, 2001; Goldman and Copsey, 2000; Zuider-duin et al., 2003; Bertin and Robin, 2002; Thio et al., 2002)performed at room temperature, where the addition of cal-cium carbonate and elastomer on virgin PP affect greatlythe tensile properties and stiffness. The tensile propertiesmeasured at 60 �C show different trends. Unlike the resultsfound at �20 �C and 23 �C, the addition of calcium car-bonate to the polypropylene does not increase the Young’smodulus, and the addition of elastomer reduces it signifi-cantly. The Young’s modulus of the recycled PP is higherthan all of the blends, even though it shows a great degreeof variability. Goldman and Copsey (2000), on the otherhand, reported a 10% increase of the Young’s modulus at42 �C with the addition of 20% of calcium carbonate to ahomopolymer PP. Furthermore, in our study, as the con-centration of calcium carbonate and elastomer increase,the yield stress decreases; however, this effect is only minor.

4.3. Charpy impact tests

Table 4 shows the impact energy and maximum forceafter impact measured at �20 �C on the Charpy test. Sixspecimens of each composition were tested. The resultsshow that the addition of elastomer and/or calcium car-bonate does not influence the results of the impact energyand maximum force to any large extent. Within errors,all of the results are more or less equal, even though a sta-

ress (MPa) Yield strain (%) Maximum elongation (%)

.7 3.4 ± 0.1 110 ± 50

.6 4.1 ± 0.6 95 ± 30

.3 3.6 ± 0.3 120 ± 30

.3 3.4 ± 0.3 120 ± 40

.6 4.3 ± 0.4 160 ± 80

.5 3.9 ± 0.4 200 ± 100

.3 2.8 ± 0.2 160 ± 70

.6 2.4 ± 0.1 120 ± 50

.1 6.7 ± 0.1 16 to 580

.4 8.9 ± 0.7 >150

.9 8.5 ± 0.1 >150

.1 8.3 ± 0.1 >150

.2 12.5 ± 0.1 >150

.1 9.5 ± 1.1 >150

.1 6.2 ± 0.2 >150

.1 5.6 ± 0.2 >150

.1 10 ± 2 >150

.1 13.5 ± 0.5 >150

.2 11 ± 2 >150

.1 12 ± 1 >150

.2 12 ± 4 >150

.1 13 ± 2 >150

.1 11.1 ± 0.6 >150

.1 9.9 ± 0.2 >150

Table 4Impact tests performed at �20 �C showing the energy to failure andmaximum force with their standard deviation

Sample composition Energy (kJ/m2) Max force (N)

PP + S 2.0 ± 0.3 270 ± 34PP + S + 5%E 1.7 ± 0.2 240 ± 21PP + S + 5%E + 10%C 2.2 ± 0.2 275 ± 35PP + S + 5%E + 20%C 1.8 ± 0.1 270 ± 11PP + S + 10%E + 10%C 2.5 ± 0.9 280 ± 26PP + S + 10%E + 20%C 2.0 ± 0.3 280 ± 32PP + S + 10%C 2.1 ± 0.3 330 ± 69PP + S + 20%C 2.0 ± 0.3 270 ± 33

2460 P. Brachet et al. / Waste Management 28 (2008) 2456–2464

tistical analysis (to be discussed later) shows that the addi-tion of elastomer only gives a small but statistically signif-icant decrease in impact properties. The literature on virginPP, however, shows significant increase on the impactproperties with the addition of elastomer (20-fold increasewith 20% elastomer) and calcium carbonate (from 50% to300% increase with 10% to 30% CaCO3) (Chan and Wu,2002; Wang et al., 2002; Premphet and Horanont, 2001;Goldman and Copsey, 2000; Yang et al., 2006). However,a study from Bertin and Robin (2002) showed partial frac-ture during a Charpy impact test at 0 �C for post-consumerPP with 5% elastomer EPDM (ethylene propylene dienemonomer). The authors put forward the idea that thepost-consumer PP can contain foreign materials that affectthe results, and that the EPDM did not work as a compat-ibilizer, therefore explaining these partial fractures. Theresults measured in this study can partially confirm Bertinand Robin’s theory, where the addition of both elastomerand calcium carbonate on recycled polypropylene doesnot show any significant difference on the maximum force,while a degree of variability is observed.

4.4. Ageing test

Fig. 1 shows the ageing process of a virgin polypropyl-ene and five different blends of recycled polypropylene.The virgin polypropylene grade (Moplen EP240M) is a typ-

60

70

80

90

100

110

120

130

140

150

160

1 10 100 1000 10000 100000Time (hours)

Tem

pera

ture

(deg

)

PP+SPP+S+5%EPP+S+5%E+20%CPP+S+10%E+10%CPP+S+20%CMOPLEN EP240M

10 YEARS

Fig. 1. Degradation of selected blends of recycled polypropylene and onevirgin polypropylene material.

ical grade used in packaging. This was included in order tocompare the life expectancy time of a typical virgin poly-mer to that of recycled polypropylene, in addition to seethe effect of additives on the recycled grade. All blends wereadded stabilizers (0.15% of Chimassorb, and 0.15% of Irga-nox) as previously described (2.1). The objective of thisdegradation test was only to get an indication of life expec-tancy times. Therefore, only two specimens of each compo-sition were tested. With so few specimens, the test onlyprovides an indication of ageing and expected life time.Many more parallels are needed in order to obtain the fail-ure time distribution and reliable estimates of life expec-tancy times. Literature studies have shown thatdegradation of recycled polypropylene is heterogeneous(Richters, 1970; Goss et al., 2003; Ahlblad et al., 2001;Hamskog et al., 2003). Such heterogeneity has been attrib-uted amongst other things to the presence of polymeriza-tion catalyst residues initiating the degradation (Celinaet al., 1993; Goss et al., 2003), morphological effects (Bill-ingham et al., 1976) or to the enhanced oxidation rate ofpolymers by oxygenated structures. Recycled polypropyl-ene, as expressed previously, is composed of many gradesof polypropylene in addition to contamination from otherpolymers. Such contamination reduces significantly themechanical properties, as well as reducing the life time ofthe material. Previous degradation studies were carriedout on recycled polypropylene (Jansson et al., 2003; Ham-skog et al., 2006). Hamskog et al. (2006) reported that theaddition of virgin material to recycled polypropylene wasineffective in terms of improved ageing properties: Whether80% or 20% of virgin polypropylene was added to recycledpolypropylene, the life expectancy of the recycled polypro-pylene did not improve, indicating that even a smallamount of recycled polypropylene in virgin polypropylenecan significantly shorten its ageing properties. The additionof stabilizers, however, increased drastically the ageingproperties of recycled polypropylene. The results in Fig. 1show that it is possible to obtain the same life expectancyof recycled PP as in virgin material with the addition ofextra stabilizers. The addition of elastomer and moderateamounts of calcium carbonate does not seem to have anyeffect. However the addition of larger amounts of calciumcarbonate seems to have a negative effect on the life expec-tancy. The samples with 20% of calcium carbonate, bothwith and without elastomer added, appear to break fasterthan all of the other samples, including the original recy-cled material. Thus, despite the addition of extra stabilizersfor long life property, a larger amount of calcium carbon-ate appears to accelerate the degradation of the material.The extrapolation of the results suggest, however, that allblends, when properly stabilized, are usable for 10 yearsat 60 �C, which is one of the main criteria for the intendeduse of this specific recycled PP, but it may be observed thatthe difference between a stabilized recycled PP and thesame blend with 20% calcium carbonate is significant, at70 �C, as the difference in life expectancy more than dou-bles between the two blends.

Fig. 3. Recycled polypropylene with the addition of 10% EOR and 10%CaCO3. The magnifications in image (a) and (b) is 250 and 1500,respectively.

P. Brachet et al. / Waste Management 28 (2008) 2456–2464 2461

4.5. Differential scanning calorimeter testing

Tensile and Charpy mechanical tests do not reveal sig-nificant improvements or changes to the recycled PP withthe addition of calcium carbonate and elastomer. In orderto get a better understanding of the performance of theblends, a calorimetry analysis was conducted; the tests wereperformed three times for reproducibility.

The DSC analysis (Fig. 2) reveals similar thermographsfor all four selected blends. The graphs show two peaks,one occurring at 126 �C and one at 160 �C. The meltingpoint taking place at 160 �C corresponds to the meltingpoint of polypropylene (homopolymer), while the meltingpoint taking place at 126 �C is characteristic of polyethyl-ene. Comparing these peaks for the four samples showedonly small variations – the standard deviation of peak tem-peratures are 0.1 and the standard deviation of area (i.e.,the heat of fusion) on the PE and PP peaks are 0.1 and0.2, respectively. The area of polyethylene is significantlysmaller than the area of polypropylene; nonetheless it stillemphasizes the presence of PE in all blends of recycledpolypropylene. From the area of the PE peak we can deter-mine the crystallisable volume fraction of PE in the recy-cled PP material using that the heat of fusion ofcrystalline polyethylene (290 J/g (Mark, 1996) (the crystaldensity of HDPE is 1 g/cm3)). From all four DSC graphswe observe that the fraction of PE crystalline material isroughly 6%. In addition to the crystalline regions of PE,there is probably the same amount of amorphous PE pres-ent. Thus, we can estimate the total PE content in the recy-cled material to be double that of the crystalline PEfraction (Neway et al., 2001). The presence of this fractionis either through impurities or within the different PPcopolymers used for packaging materials. In an attemptto get a better understanding of the homogeneity of the

50 70 90 110 130 150 170 190Temperature (deg C)

160.0 °C

160.0 °C

159.6 °C

160.0 °C

126.4 °C

126.1 °C

126.1 °C

126.1 °C

PP + S + 10%EOR + 10%CaCO3

PP + S

PP + S + 5%EOR

PP + S + 10%CaCO3

Perkin Elmer Thermal Analysis

Fig. 2. DSC analysis of recycled polypropylene with different composi-tions of elastomer, calcium carbonate. Stabilisers are added to all blends.The curves are obtained from a second heating run at 20 �C/min after firstbeing heated to 250 �C. Two melting transitions, one happening at 126 �Cand one at 160 �C are visible.

blends of recycled polypropylene, elastomer and calciumcarbonate, we made scanning electron microscopy (SEM)images of the fracture surfaces of a few samples.

Fig. 3a and b shows the fracture surface of a samplecontaining 10% EOR and 10% calcium carbonate, ana-lyzed with the SEM with magnifications of 250 and 1500.The particles of calcium carbonate are clearly visible inboth pictures and confirmed a homogenous dispersion.However, the elastomer added to the recycled polypropyl-ene is not visible in these images, so we cannot say anythingabout how well this phase is dispersed or blended into thePP.

5. Discussion and conclusion

An elastomer and a calcium carbonate were added torecycled polypropylene to look at the changes of themechanical properties, especially the improvement of theimpact property. Studies on virgin polypropylene haveshown that such additives greatly improve the impactand tensile properties. The elastomer, EOR, was chosenin order to enhance toughness and impact properties, espe-

Table 5Trends in property changes of recycled PP with the addition of elastomer,calcium carbonate or both

Additive �20 �C 23 �C 60 �C

E-modulus (tensile)

CaCO3 0 + �EOR � � �CaCO3 + EOR � 0 �

Additive r Yield stress e Yield strain

Yield stress and strain

CaCO3 � �EOR � +CaCO3 + EOR � +

Table 6Significant differences in the impact properties for different temperaturesbetween the eight different blends, obtained from an ANOVA statisticalanalysis, is obtained between the pair of blends indicated

Statistical differences between pair ofblends

Energy at impact �20 �C 1–2 2–3 2–6 2–7 3–4 4–7Maximum force at impact �20 �C 2–4 2–5 2–6 2–7

(1) – PP + S;(2) – PP + S + 5%E;(3) – PP + S + 5%E + 10%C(4) – PP + S + 5%E + 20%C;(5) – PP + S + 10%E + 10%C;(6) – PP + S + 10%E + 20%C;(7) – PP + S + 10%C;(8) – PP + S + 20%C.

2462 P. Brachet et al. / Waste Management 28 (2008) 2456–2464

cially at low temperature, and the calcium carbonate wasused to improve hardness and stiffness while lowering thecost. However, our tests did not show the general enhance-ment of the properties reported in the literature. Onlyminor changes were observed. A summary of the tensiletest results in Section 4.2 is given in Table 5. This table indi-cates the trends in the results, although minor, as a result ofblending elastomer and calcium carbonate with recycledpolypropylene.

As discussed previously, the mechanical and impactproperties showed rather large scatter among parallels.In order to get a feeling of whether or not any of theseminor trends are statistically significant, we performed astatistical analysis of variance (ANOVA, one way modelwith Tukey’s 95% simultaneous confidence intervals). Weanalysed the significance of these small differences inYoung’s modulus, impact energy and maximum forceat impact between the eight different blends. This statis-tical analysis shows that at �20 �C, the decrease inYoung’s modulus by adding elastomer only is significant.However, the addition of calcium carbonate did not haveany significant statistical influence on the results, eventhough the data indicates that the addition of calciumcarbonate reduces the effect of added rubber. At 23 �C,the ANOVA statistical analysis showed that the decreasein Young’s modulus with the addition of elastomer andthe increase with the addition of CaCO3 are both signif-icant trends. At 60 �C the ANOVA analysis confirmedthat the Young’s modulus for the recycled PP with stabi-lizers was significantly different than all of the otherblends, by being stiffer, while the addition of 10% elasto-mer made the Young’s modulus significantly softer. Thecalcium carbonate seemed not to have any statisticallysignificant effect.

Generally speaking the literature reported much largerincreases than the minor changes obtained in our trials,especially on the Young’s modulus with the addition ofcalcium carbonate (Zebarjad et al., 2006; Premphet andHoranont, 2001; Goldman and Copsey, 2000; Zuiderduinet al., 2003; Bertin and Robin, 2002; Thio et al., 2002; Yanget al., 2006; Wu et al., 2005; Rong et al., 2001,; Lehmannet al., 2003). Generally, the results observed during the ten-

sile tests showed a great degree of variability when testingparallels of the same material, which prevented a morecomplete analysis but emphasized the idea that the blend-ing of the different materials was not as homogenous asit optimally should have been.

The statistical analysis performed on the impact testresults showed very little differences between most ofthe blends, as seen in Table 6. The only blend thathad a behaviour significantly different from all the otherswas the blend having elastomer only added. Elastomer isadded for improved impact properties, but this blendactually had the lowest energy to break and maximumforce. Our data seems to indicate that adding elastomerand 10% calcium carbonate gives a slight increase inthe impact properties. However, the ANOVA analysisdoes not confirm this trend to be statistically significant.Overall the impact properties of recycled polypropylene(impact energy, maximum force) measured during thesetests did not improve significantly with the additives,while previous studies reported a substantial increasewith the addition of elastomer (20-fold increase with20% addition of EOR) and calcium carbonate (anincrease ranging from 50% to 300% with the additionof 10% to 30% CaCO3) (Chan and Wu, 2002; Wanget al., 2002; Premphet and Horanont, 2001; Goldmanand Copsey, 2000; Thio et al., 2002; Yang et al., 2006;Pukansky, 1995; Baker et al., 1987; Liu et al., 2002;Zuiderduin et al., 2001a,b). Low impact properties ofthe recycled PP at low temperatures can also be attrib-uted to the presence of homopolymer PP, which is brit-tle. The DSC analysis provided valuable information tounderstand the blending of the different materials. Thethermographs revealed for all four samples two meltingpoints, one characterizing the polypropylene (160 �C)and one of lesser intensity characterizing polyethylene(126 �C). The crystalline fraction of PE present in therecycled polypropylene was calculated to be around6%, so one can assume the total content of PE (amor-phous and crystalline) to be double this value (Newayet al., 2001). The presence of this phase can be attributedto impurities present in the recycled polypropylene and

P. Brachet et al. / Waste Management 28 (2008) 2456–2464 2463

to PE segments present in different (block) copolymerspresent in the recycled polypropylene. This can affectthe composition of the recycled PP matrix, and limitthe enhancement of the mechanical properties throughblending. Sufficient compatibility with the additives hasnot been achieved during these tests. Ageing testsrevealed that samples with a high content of calcium car-bonate degraded significantly faster than the others, eventhough stabilizers were added. Blends with elastomeronly and with calcium carbonate showed similar ageingas virgin material as well as recycled PP with added sta-bilizers only. Calcium carbonate intensified the degrada-tion process. Samples with a high concentration ofcalcium carbonate would therefore not have a sufficientlife expectancy for the purpose of its use. It appears thatthe use of the elastomer EOR, did not act as a goodcompatibilizer between the PP and the PE presentthrough impurities. As better sorting would significantlyincrease the price of recycled material, the most appro-priate solution for enhancing the properties might bethe use of a different quality of elastomer, which wouldact as a compatibilizer, providing easier blending withinthe materials, in addition to optimising the processingparameters used during blending.

Acknowledgements

This project was sponsored by the Norwegian ResearchCouncil, TOMRA Systems ASA Norway and Emballasjer-etur AS.

References

Ahlblad, G., Gijsman, P., Terselius, B., Gijsman, P., Jansson, A., Moller,

K., 2001. Thermo-oxidative stability of PP waste films as studied by

imaging chemiluminescence. Polymer Degradation and Stability 73, 15.

Aurrekoetxea, J., Sarrionandia, M.A., Urrutibeascoa, I., Maspoch, M.L.,

2001a. Journal of Material Science 36, 1.

Aurrekoetxea, J., Sarrionandia, M.A., Urrutibeascoa, I., Maspoch, M.L.,

2001b. Journal of Material Science 36, 5073.

Baker, R.A., Koller, L.L., Kummer, P.E., 1987. Handbook of Fillers for

Plastics, second ed. Van Nostrand Reinhold, New York.

Barbosa, S.E., Kenny, J., Capiti, N.J., 2000. Polymer Composites 21, 377.

Bertin, S., Robin, J.J., 2002. Study and characterization of virgin and

recycled LDPE/PP blends. European Polymer Journal 38, 2255–2264.

Billingham, N.C., Prentice, P., Walker, T.J., 1976. Some effects of

morphology on oxidation and stabilization of polymers. Journal of

Polymer Science Symposium 57, 287.

Celina, M., George, G.A., Billingham, N.C., 1993. Physical spreading of

oxidation in solid polypropylene as studies by chemiluminescence.

Polymer Degradation and Stability 42, 335.

Chan, C.M., Wu, J.S., 2002. Polymer 43, 2981.

Chen, J.F., Wang, G.Q., Zeng, X.F., 2004. Journal of Applied Polymer

Science 94, 796.

Gahleitner, M., 2002. Journal of Macromolecular Science, Part B: Physics

41, 833.

Goldberg, V.M., Zaikov, G.E., 1987. Polymer Degradation and Stability

19, 221.

Goldman, A.Y., Copsey, C.J., 2000. Polyproplyene toughened with

calcium carbonate mineral filler. Material Research Innovation 3,

302–307.

Gonzalez, V.A., Neira-Velazquez, G., Angulo Sanchez, J.L., 1998.

Polymer Degradation and Stability 60, 33.

Goss, B.G.S., Nakatan, H., George, G.A., Terano, M., 2003. Catalyst

residue effects on the heterogeneous oxidation of polypropylene.

Polymer Degradation and Stability 82, 119–126.

Hamskog, M., Terselius, B., Gijsman, P., 2003. Mapping of heterogeneous

degradation of polypropylene using imaging chemiluminescence.

Polymer Degradation and Stability 82, 181–186.

Hamskog, M., Klugel, M., Forsstrom, D., Terselius, B., Gijsman, P.,

2006. The effect of adding virgin material or extra stabilizer on the

recyclability of polypropylene as studies by multi-cell imaging chemi-

luminescence and microcalorimetry. Polymer Degradation and Stabil-

ity 91, 429–436.

Jancar, J., Dibenedatto, A.T., Kucera, J., 1993. Journal of Polymer 34, 1684.

Jancar, J., Kucera, J., 1990. Journal of Polymer Engineering Science 30,

707.

Jang, B.Z., 1992. Composites Science and Technology 44, 333.

Jansson, A., Moller, K., Gevert, T., 2003. Degradation of post consumer

polypropylene materials exposed to simulated recycling – mechanical

properties. Polymer Degradation and Stability 82, 37–46.

Lehmann, B., Friedrich, K., Zhang, M.Q., 2003. Journal of Material

Science Letters 22, 1027.

Levita, G., Marchetti, A., Lazzeri, A., 1989. Polymer Composites 10, 39.

Liu, Z.H., Argon, A.S., Cohen, R.E., Weinberg, M., 2002. Polymer 43,

2501.

Ma, C.G., Rong, M.Z., Friedrich, K., 2005. Polymer Engineering Science

45, 529.

Mark, E.J., 1996. Physical Properties of Polymers Handbook. American

Institute of Physics (AIP) Press, Chapter 11, 123.

Nam, P.H., Maiti, P., Okamoto, M., Kotaka, T., Hasegawa, N., Usuki,

A., 2001. Polymer 42, 9633.

Neway, B., Hedenqvist, M.S., Mathot, V.B.F., Gedde, U.W., 2001. Free

volume and transport properties of heterogeneous poly(ethylene-co-

octene)s. Polymer 42, 5307–5319.

Papaspyrides, C.D., Poulakis, J.G., 1996. In: Salomone, J.C. (Ed.), The

Polymeric Materials Encyclopedia. CRC Press, Inc., p. 7403.

Premphet, K., Horanont, P., 2000. Journal of Applied Polymer Science 76,

1929.

Premphet, K., Horanont, P., 2001. Improving performance of polypro-

pylene through combined use of calcium carbonate and metallocene-

produced impact modifier. Polymers Plastics Technology and Engi-

neering 40 (3), 235–247.

Pukansky, B., 1995. In: Karger-Kocsis, J. (Ed.), Polypropylene: Structure

Blends and Composites. Chapman & Hall, London (Chapter 1).

Richters, P., 1970. Initiation process in the oxidation of polypropylene.

Macromolecules 3, 262.

Rong, M.Z., Zhang, M.Q., Friedrich, K., 2001. Polymer 42, 167.

Rong, M.Z., Zhang, M.Q., Friedrich, K., 2001. Polymer 42, 3301.

Seiler, E., 1995. Properties and Applications of Recycled Polypropylene.

Recycling and Recovery of Plastics. Carl Hanser Verlag.

Suetsugu, Y., White, J.L., 1987. Advanced Polymer Technology 7, 427.

Thio, Y.S., Argon, A.S., Cohen, R.E., Weinberg, M., 2002. Toughening of

isotactic polypropylene with calcium carbonate particles. Polymer 43,

3661–3674.

Tiganis, B.E., Shanks, R.A., Long, Y., 1996. Journal of Applied Polymer

Science 59, 663.

Wang, G., Chen, X.Y., Huang, R., 2002. Journal Materials Science

Letters 21, 985.

Wu, C.L., Zhang, M.Q., Friedrich, K., 2005. Composites Science and

Technology 65, 635.

Yang, K., Yang, Q., Li, G., Sun, Y., Feng, D., 2006. Morphology and

mechanical properties of polypropylene/calcium carbonate nanocom-

posites. Materials Letters 60, 805–809.

Zebarjad, S.M., Sajjadi, S.A., Tahani, M., 2006. Modification of fracture

toughness of isotactic polypropylene with a combination of EPR and

CaCO3 particles. Journals of Materials Processing Technology 175 (1–

3), 446–451.

2464 P. Brachet et al. / Waste Management 28 (2008) 2456–2464

Zuiderduin, W.C.J., Westzaan, C., Huetink, J., Gaymans, R.J., 2001a.

ANTEC, spe, ISBN 1-56676-804. pp. 2848–2852.

Zuiderduin, W.C.J., Westzaan, C., Huetink, J., Gaymans, R.J., 2001b.

ANTEC, spe, ISBN 1-56676-804. pp. 3493–3497.

Zuiderduin, W.C.J., Westzaan, C., Huetink, J., Gaymans, R.J., 2003.

Toughening of polypropylene with calcium carbonate particles.

Polymer 44, 261–275.