analiza polimerilor in procesul de piroliza

Resources, Conservation and Recycling 51 (2007) 754–769

Analysis of products from the pyrolysis andliquefaction of single plastics and

waste plastic mixtures

Paul T. Williams ∗, Edward SlaneyEnergy & Resources Research Institute, University of Leeds,

Leeds LS2 9JT, UK

Received 28 September 2006; received in revised form 5 December 2006; accepted 27 December 2006Available online 30 January 2007

Abstract

Waste plastics in the form of two examples of real world municipal solid waste plastics anda simulated mixture of municipal waste plastics were pyrolysed and liquefied under moder-ate temperature and pressure in a batch autoclave reactor. In addition, the five main polymerswhich constitute the majority of plastics occurring in European municipal solid waste compris-ing, polyethylene, polypropylene, polystyrene, polyethylene terephthalate and polyvinyl chloridewere also reacted. The plastics were reacted under both a nitrogen (pyrolysis) and hydrogenpressure (liquefaction) and the yield and composition of products are reported. The hydrocarbongases produced were mainly methane, ethane, propane and lower concentrations of alkene gases.A mainly oil product was produced with the mixed plastic waste with significant concentrationsof aromatic compounds, including single ring aromatic compounds. The composition of the oilsand gases suggested that there was significant interaction of the plastics when they were pyrol-ysed and liquefied as a mixture compared to the results expected from reactions of the singleplastics.© 2007 Elsevier B.V. All rights reserved.

Keywords: Pyrolysis; Plastics; Liquefaction; Waste

∗ Corresponding author. Tel.: +44 1133432504; fax: +44 1132467310.E-mail address: [email protected] (P.T. Williams).

0921-3449/$ – see front matter © 2007 Elsevier B.V. All rights reserved.doi:10.1016/j.resconrec.2006.12.002

mailto:[email protected]

dx.doi.org/10.1016/j.resconrec.2006.12.002

P.T. Williams, E. Slaney / Resources, Conservation and Recycling 51 (2007) 754–769 755

1. Introduction

The average plastic content of municipal solid waste in Western Europe is 9.1 wt%,representing more than 13 million tones of plastic waste generated in Western Europeeach year (APME, 2004). However, the recycling rate for the plastic fraction of munic-ipal solid waste remains below 10%, representing a waste of a resource (APME, 2004).There are five main plastics which arise in European municipal solid waste which arepolyethylene (as low and high density polyethylene, (HDPE and LDPE), polypropy-lene (PP), polystyrene (PS) polyethylene terephthalate (PET) and polyvinyl chloride(PVC).

Plastic material separated from municipal solid waste has been used in a varietyof low grade applications. The separated plastic material is usually processed by theend user by being granulated or pelletised, melted or partially melted and extrudedto form the end product. The recycled plastic may be added to virgin plastic duringthe process. Applications for plastic mixtures have included plastic fencing, indus-trial plastic pallets, traffic cones, playground equipment and garden furniture. Otheruses for recycled plastic products include their use in the construction industry forpipes, damp-proof membranes, plastic lumber and plastic/wood composites (Williams,2005).

The low grade uses for mixed plastic recycled materials has led to research into alter-native processes for plastics recycling. For example, there has been enormous researchinterest into pyrolysis or feedstock recycling of waste plastic where the plastic waste mate-rials are processed back to produce basic petrochemicals that can be used as feedstockto make virgin plastic (Williams, 2005; Kaminsky et al., 2004; de Marco et al., 2002;Bhaskar et al., 2003; Williams and Williams, 1997; Williams and Williams, 1999a,b).The pyrolysis process produces a gas, an oil/wax and in some cases a char product,the proportions of which are related directly to the type of plastic, the reactor type andthe process conditions, particularly pyrolysis temperature and heating rate. The influenceof catalysts on the yield and composition of products has been investigated extensively(Williams and Bagri, 2004; Bagri and Williams, 2002; Manos et al., 2000). Liquefactionof waste plastics has been undertaken by several investigators, involving reactions in auto-clave high temperature and pressure reactors in the presence of hydrogen, or a hydrogendonor such as tetralin or oil and/or the presence of catalysts (Ramdoss and Tarrer, 1998;Pinto et al., 1999; Feng et al., 1996; Murty et al., 1996; Ding et al., 1997; Shabtai et al.,1997). Much of the research has been carried out on single plastics or in simple plasticmixtures, whilst there are few data on plastics collected from real world municipal solidwaste.

In this paper we report on the pyrolysis in nitrogen and liquefaction in hydrogen of fiveindividual plastics and two types of waste plastic mixtures collected from municipal solidwaste and a simulated plastic mixture representative of those common plastics found inmunicipal solid waste. The reactions were carried out at in a batch autoclave reactor underpressure of either nitrogen or hydrogen at 500 ◦C. The yield and composition of the gas andoil products are reported.

756 P.T. Williams, E. Slaney / Resources, Conservation and Recycling 51 (2007) 754–769

2. Materials and methods

2.1. Waste plastic samples

Individual single plastic types were reacted under pressure and at 500 ◦C in either anitrogen atmosphere, pyrolysis conditions or hydrogen atmosphere, liquefaction condi-tions. Five plastics which comprise the majority of European municipal solid waste wereused comprising, polyethylene (as high density polyethylene), polypropylene, polystyrene,polyethylene terephthalate and polyvinyl chloride. The polymers were obtained from BPChemicals Ltd., Grangemouth, Scotland and were virgin polymers of 2–5 mm diametersize.

Two samples of real world post-consumer, mixed waste plastic obtained from municipalsolid waste were used for the experiments. One sample was obtained from Germany viaDuales System Deutschland (DSD). The plastic sample was collected in Germany as partof the green dot recycling scheme. The plastic sample was shredded and washed to removenon plastic impurities. The samples were sized at 2–5 mm granules.

The second waste plastic sample was collected post-consumer municipal waste mixedplastic. This sample was collected in Belgium by Fost Plus, the plastic was flaked and wasseparated into a low density fraction through air separation. The low density fraction wouldbe expected to have a reduced fraction of the high density polymers, polyvinyl chloride andpolyethylene terephthalate, with an increased fraction of lower density polymers, includingpolyethylene, polypropylene and polystyrene. The sample size was approximately 5–10 mmsized flakes.

The simulated mixed waste plastic sample was made up from the main plastics foundin typical European municipal solid waste and comprised, polyethylene (as high densitypolyethylene), polypropylene, polystyrene, polyethylene terephthalate and polyvinyl chlo-ride (APME, 2004). The actual proportions of each plastic used are shown in Table 1. Onlythe main polymer fractions were used to make up the simulated municipal solid waste plas-tic mixture, with no account for the numerous minor components in the remainder of theplastics waste stream.

The sample mass used in the reactor for all samples was between 30 g and 40 g. Thisrelatively large sample mass was important for the heterogeneous samples of real worldwaste samples, DSD and Fost Plus, and the simulated mixture of plastics. Such a large massof sample ensured that the samples were a representative mixture of the plastic waste.

Table 1Proportion of each plastic in the simulated municipal solid waste plastic mixture

Plastic Proportion (wt%)

Polyethylene (HDPE) 44.4Polypropylene (PP) 21.2Polystyrene (PS) 13.3Polyvinyl chloride (PVC) 12.2Polyethylene terephthalate (PET) 8.9


Fig. 1. Schematic diagram of the pyrolysis/liquefaction reactor.

2.2. Pyrolysis/liquefaction reactor

The same reactor was used for pyrolysis and liquefaction of the plastics and plasticmixtures. The reactor used was an a Parr Mini Bench Top Reactor, Type 4561 m stirredpressure reactor and was obtained from the Parr Instrument Co., Moline, IL, US. Thereactor had a volume capacity of 300 ml and a maximum operating pressure of 19.2 MPaat 500 ◦C and was constructed of type T316 stainless steel, with a flexible graphite gas-ket. Fig. 1 shows a schematic diagram of the reactor system. The reaction vessel wasstirred using a Parr Instrument Co. magnetic drive. A stirrer shaft with rotor blades at thebase was attached to an inner magnetic drive rotor powered through an outer magneticdrive rotor. The reactor was heated using an external mantle type furnace, which con-tacted the sides and bottom of the reaction vessel. The reactor was fitted with a safetyburst disc. Gas pressures were recorded from the pressure gauge on the reactor, measuringthe reactor internal pressure. The temperature of the waste plastic sample in the reactorwas monitored via a J type thermocouple, which was sheathed. A programmable tem-perature controller controlled power input to the heater. Heating power was controlled tofollow a pre-set temperature profile based on the reactor internal thermocouple reading. Amaximum heating rate of 5 ◦C min−1 was established under experimental conditions to afinal temperature of 500 ◦C and the reactor was held at that condition for 1 h. The reac-tor components and sample weight were recorded prior to assembling the reactor for eachexperiment. Gas samples were taken from the system after it had cooled to room temper-ature after the final pressure at room temperature had been recorded. The final weight ofnon-gaseous products was determined at the end of each experiment after de-pressurizingand disassembling the reactor. The initial pressure of nitrogen used was 0.2 MPa, generat-ing a maximum pressure of approximately 10 MPa at 500 ◦C. The initial hydrogen pressureused was 1 MPa, which was higher than for the initial nitrogen atmosphere to allow forincreased hydrogen availability, representing 0.61 mol, or 1.2 g of hydrogen gas available


as a reagent for the reaction. The final pressure at the 500 ◦C final reaction temperature was18 MPa.

2.3. Gas analysis

Gas samples were taken from the reactor after reaction using sample syringes. The gaseswere immediately analysed for hydrocarbon gases. Hydrocarbons from C1 to C4 wereanalysed using a Varian CP-3380 gas chromatograph with a Flame Ionisation Detector(GC/FID). The column used was 2 m long with 2 mm diameter packed with 80–100 meshHysesp. Nitrogen was used as the carrier gas. The temperature was programmed to startat 60 ◦C for 3 min, heated at a heating rate of 10 ◦C min−1 to 100 ◦C, held for 3 min, andfinally heated to 120 ◦C at a heating rate of 20 ◦C min−1 and held for 9 min at 120 ◦C.The injector temperature was held at 150 ◦C while the detector temperature was 200 ◦C.Where the sample included PVC, HCl was expected as a product gas from the reaction, butwas not included in the analysis of the gases. The concentration of each component of theproduct gas was calculated by gas chromatography, and the weight of the gas componentcalculated from the concentration and final pressure of the reactor by using the Ideal GasLaw.

2.4. Oil analysis

The oil yield was represented by the non-gaseous products soluble in dichloromethane(DCM), with the char the insoluble fraction in DCM. The combined mass of oil and charwas noted after reaction. The yield of solid char produced was determined via dissolutionof the oil in DCM followed by filtration of the char. The filter was washed with clean DCMto remove the soluble sample, and to ensure that all insoluble char sample was collected onthe filter paper. The filter papers were first dried in a fume cupboard at room temperature,and then fully dried in an oven at 110 ◦C, and reweighed to find the weight of non solublechar, the weight of oil was calculated by difference.

The main hydrocarbons present in the oils were analysed by capillary column gas chro-matography. A Carlo Erba HRGC 5300 Mega Series GC, fitted with a flame ionizationdetector was used for the quantification of compounds found in the oil product. The columnwas an HP5 (5% Ph-Me-silicone) capillary column, 30 m length with 0.53 mm diameterand 5.0 �m film thickness. Helium was used as a carrier gas. The temperature program usedwas, initial temperature 50 ◦C for 5 min followed by a heating rate of 5 ◦C min−1 to 270 ◦Cand held at 270 ◦C for 20 min.

To confirm the identification of hydrocarbon species, the oils were also analysed bycoupled gas chromatography/mass spectrometry. The system used was a Hewlett Packard(HP) 5890 Series 2 GC, fitted with an HP 5971 MS mass spectrometer detector. The columnwas a 30 m long by 0.25 mm internal diameter Restek Rtx-5MS capillary column withfused silica 5% diphenyl/95% dimethyl polysiloxane coating of 0.25 �m film thickness.Helium was used as the carrier gas. The gas chromatographic temperature programmewas the same as that used for the gas chromatography/flame ionization detector system.The identification of peaks was conducted on the software using the NIST 98 chemicallibrary.


Table 2Product yield from the pyrolysis and liquefaction of different plastic types under nitrogen and hydrogenatmospheres

Plastic Oil (wt%) Gas (wt%) Residue (wt%)

Pyrolysis (nitrogen)Polyethylene 93 7 0Polypropylene 95 5 0Polystyrene 71 2 27Polyvinyl chloride - - -Polyethylene terephthalate 15 32 53

Liquefaction (hydrogen)Polyethylene 95 5 0Polypropylene 95 5 0Polystyrene 77 2 22Polyvinyl chloride 2 38 52Polyethylene terephthalate 27 32 41

3. Results and discussion

3.1. Product yield

Table 2 shows the yield of oil, gas and char from the pyrolysis and liquefaction of thevarious individual plastics under nitrogen pressure and hydrogen pressure where the sampleswere heated at 5 ◦C min−1 to the final temperature of 500 ◦C and held at 500 ◦C for 60 min.The reaction of polyethylene and polypropylene in both nitrogen pyrolysis and hydrogenliquefaction conditions produced a high conversion of plastic to an oil product with lowerconcentrations of gas and no solid residue product. Polystyrene liquefaction also produceda high conversion to an oil product at 71 wt% conversion under pyrolysis and 77 wt% underhydrogen liquefaction conditions, but also resulted in the production of a significant solidresidue. The production of oil under hydrogen liquefaction was increased compared topyrolysis under nitrogen, with a consequent decrease in the solid residue from pyrolysis of27 wt% compared to 22 wt% with hydrogen liquefaction. Polyvinyl chloride pyrolysis couldnot be completed due to problems of corrosion of the reactor burst disc, due to formation ofhigh concentrations of hydrogen chloride. However, with hydrogen liquefaction of polyvinylchloride the products produced were 2 wt% conversion to an oil, 38 wt% gas and 52 wt%solid residue. Polyethylene terephthalate also produced a high conversion to a solid residueunder nitrogen pyrolysis conditions with a yield of 53 wt% with a consequent production of15 wt% oil. But for hydrogen liquefaction conditions the yield of solid residue was 41 wt%and 27 wt% oil. Pinto et al. (1999) undertook pyrolysis experiments with polyethylene,polypropylene and polystyrene in an autoclave reactor at an initial nitrogen pressure of0.41 MPa and temperature of 430 ◦C, leading to a mean experimental pressure of 3.5 MPa.They reported over 90 wt% conversion of the plastics to oil. However, the pressure used intheir system was significantly lower than that used in this work. Whilst some authors havereported low solid residue yields for polystyrene, others have reported significant productionof non-volatile material. For example, McCaffrey et al. (1996) examined the pyrolysis ofpolystyrene at 390 ◦C in a continuous purge of nitrogen, they also reported a high yield of


Table 3Product yields from the pyrolysis and liquefaction of plastic wastes

Plastic waste Oil (wt%) HC Gas (wt%) Residue (wt%)

Pyrolysis (nitrogen)Simulated waste 48.7 3.7 34.6Calculated mixture 72.3a – –Waste DSD 32.5 0.5 50.2Waste Fost Plus 64.1 4.3 23.3

Liquefaction (hydrogen)Simulated waste 55.8 4.4 22.7Calculated mixture 75.12 11.1 12.9Waste DSD 48.2 2.2 35.1Waste Fost Plus 70.6 4.9 17.8

a Estimated.

solid residue of 14.3 wt%. Carniti et al. (1991) undertook pyrolysis of polystyrene undervacuum and in a closed vessel at up to 420 ◦C. They reported that volatile products from thepyrolysis reached a maximum at 60 wt% conversion of polystyrene. They reported that theproduct yield and composition was directly related to the conditions of the closed systemwhich prevented the volatile products and intermediates from leaving the reaction vessel.They reported that styrene, a high yield product from the thermal degradation of polystyrene,was not found in significant concentrations in their system due to the saturation of styreneto form ethylbenzene. They also suggested further reactions to produce heavy products toaccount for the low conversion of the polystyrene to volatile products.

Table 3 shows the product yield for the pyrolysis and liquefaction of the waste mixtures,simulated plastic mixture prepared in relation to the percentages shown in Table 1, and thereal municipal solid waste derived plastic mixtures, Waste DSD from Germany and WasteFost Plus from Belgium. Also shown in Table 3 are the calculated oil, gas and residuepercentages, based on the product yields shown in Table 2 for the individual plastics inrelation to their percentages representative of municipal solid waste plastic shown in Table 1.

The simulated waste plastic mixture gave an oil yield of 48.7 wt% for nitrogen pyrolysisand 55.8 wt% for hydrogen liquefaction, with a significant production of solid residueand low yield of gas. The calculated yield of oil, gas and solid residue suggested that amuch higher yield of oil should have been produced, of the order of more than 70 wt%for both pyrolysis and liquefaction conditions; thus suggesting that there was significantinteraction of the plastics to produce a higher yield of solid residue. Previous work hasshown that plastic mixtures interact during nitrogen pyrolysis conditions and produce adifferent product slate and product composition to that expected from reactions of theindividual pure plastic (Williams and Williams, 1999a). The real world waste derived plasticmixtures, DSD and Fost Plus produced very different results. The DSD sample produced alow oil yield of 32.5 wt% and solid residue of 50.2 wt% for pyrolysis, whilst for hydrogenliquefaction the oil yield was 48.2 wt% and the solid residue yield was 35.1 wt%. However,the Fost Plus sample from Belgium produced high oil yields and correspondingly lower solidresidue yields for both pyrolysis and liquefaction. The DSD sample contained significantamounts of paper and dirt and was not as pure a plastic mixture as that produced by the


Fost Plus process. The Fost Plus plastic was a low density fraction of municipal solid wasteplastic and as such would have an expected composition with increased concentrations ofpolyethylene, polypropylene and polystyrene polymer fractions and reduced fractions ofpolyvinyl chloride and polyethylene terephthalate. The real world plastic waste samplesmay also contain additives and fillers which would not be present in the pure single plastics.Shah et al. (1999) also examined the liquefaction of DSD waste plastic at 445 ◦C in a bombmicroreactor with hydrogen. They reported an oil yield of greater than 80 wt% for the DSDplastic waste and that the yield was markedly influenced by both temperature and reactiontime below 445 ◦C.

3.2. Gas composition

Figs. 2 and 3 show the hydrocarbon gas composition for polypropylene, polyethylene,polystyrene and polyethylene terephthalate and polyvinyl chloride for nitrogen pyrolysisand hydrogen liquefaction, respectively. Gas composition data for polyvinyl chloride withnitrogen could not be obtained, due to corrosion of the reactor burst disc due to the formationof hydrogen chloride gas. The main hydrocarbons for pyrolysis were the alkane gases,methane, ethane, propane and butane, with lower concentrations of the alkenes, ethane,propene and butane. The gas composition was quite similar under liquefaction conditionswith hydrogen compared to pyrolysis, with however, some small but significant differences.For example, for polystyrene and polyethylene terephthalate, the presence of hydrogenresulted in increased methane production compared to the pyrolysis in nitrogen. Pinto et al.(1999) also showed that the gas composition produced from the pyrolysis of polyethylene,polypropylene and polystyrene in nitrogen was dominated by alkane gases compared toalkene gases, under similar process conditions as used here. In addition to the hydrocarbongases, hydrogen chloride gas was generated in high concentration from polyvinyl chlorideand carbon dioxide and carbon monoxide from polyethylene terephthalate (Williams andWilliams, 1999a).

Fig. 2. Hydrocarbon product gas distribution from the pyrolysis of individual plastics at 500 ◦C for 60 min withnitrogen pressure, for polyethylene (PE), polypropylene (PP), polystyrene (PS) and polyethylene terephthalate(PET).


Fig. 3. Hydrocarbon product gas distribution from the liquefaction of individual plastics at 500 ◦C for 60 minwith hydrogen, for polyethylene (PE), polypropylene (PP), polystyrene (PS) polyethylene terephthalate (PET)and polyvinyl chloride (PVC).

Fig. 4 shows the hydrocarbon gases derived from pyrolysis and Fig. 5 shows the hydro-carbon gases derived from liquefaction of the of the simulated plastic mixture based onthe composition of plastic found in municipal solid waste shown in Fig. 1, together withthe real world waste samples, DSD and Fost Plus. Also shown in Figs. 4 and 5 are thecalculated hydrocarbon gas compositions, based on the gas yields shown in Figs. 2 and 3for the individual plastics in relation to their percentages representative of municipal solidwaste plastic shown in Table 1.

The calculated hydrocarbon gas composition and that obtained experimentally fromthe pyrolysis of the simulated waste plastic mixture were similar, with the alkane gases,methane, ethane, propane and butane dominating. However, there was a higher concen-

Fig. 4. Hydrocarbon gas distribution from the pyrolysis of mixed plastic, at 500 ◦C for 60 min with nitrogen.


Fig. 5. Hydrocarbon gas distribution from the liquefaction of mixed plastic, at 500 ◦C for 60 min with hydrogen.

tration of ethane and lower concentration of propene for the experimental pyrolysis of thesimulated mixture compared to the calculated results. Liquefaction under hydrogen alsoshowed that the calculated hydrocarbon gas concentration was similar to the experimentalgas concentrations. In addition, as was the case for pyrolysis, the concentration of ethanewas higher and propene were lower for the experimental liquefaction of the simulatedmixture compared to the calculated data.

The hydrocarbon gas composition for the liquefaction of the real world waste plasticsamples, DSD and Fost Plus were also dominated by the alkane gases with lower concen-trations of the alkene gases (Fig. 4). The hydrocarbon gas composition derived from thepyrolysis of the waste DSD and Waste Fost Plus plastic samples produced higher concen-trations of ethane, propene and butane and lower concentrations of methane and ethanecompared to pyrolysis of the simulated plastic mixture. Liquefaction of the waste DSDand Fost Plus plastics showed significantly different gas compositions compared to theliquefaction of the simulated mixture of plastics (Fig. 5). For the liquefaction of the DSDplastic sample, the derived gases were higher in concentration for methane and ethene andpropene and lower in concentration for ethane. Liquefaction of the Waste Fost Plus plasticsample also showed significant difference in hydrocarbon gas composition compared tothat derived from the simulated plastic mixture, with lower methane and higher ethane andpropene. Consequently, it is difficult to estimate the hydrocarbon gas composition for eitherpyrolysis under nitrogen pressure or liquefaction under hydrogen derived from individualplastics in a simulated mixture, either experimentally or through calculation of the propor-tion of gases derived from the individual plastics for real world plastic mixtures comparedto mixtures of single plastics. The real world plastic samples would contain contaminantsand other plastics at lower concentrations than the main ones found in municipal solid wasteplastic fractions and used in this study.

It should also be noted that significant yields of hydrogen chloride would be obtainedfrom any polyvinyl chloride present and carbon dioxide and carbon monoxide from anypolyethylene terephthalate for both pyrolysis and liquefaction of the waste plastic mixtures.


Fig. 6. Oil product n-alkane and alk-1-ene concentration from the pyrolysis of mixed waste plastics at 500 ◦C fora 60 min residence time with nitrogen.

3.3. Oil composition

Fig. 6 shows oil product n-alkane and alk-1-ene concentration from the pyrolysis ofmixed waste plastics and Fig. 7 shows the n-alkane and alk-1-ene in nitrogen and hydrogen,respectively. Pyrolysis and liquefaction of the simulated plastic mixture and real world DSDand Fost Plus plastic wastes resulted in significant production of n-alkanes and alkenes. Theproduction of n-alkanes would be expected from the pyrolytic conditions under nitrogen ofthe two main components of the polymer mix, polyethylene and polypropylene (Williams,2006). For the pyrolysis of the simulated plastic mixture, very low concentrations of alkenes(alk-1-enes) were produced (Fig. 6). The oil composition from liquefaction with hydro-gen also showed a significant concentration of n-alkanes, but additionally alkenes wereproduced, albeit at lower concentrations than calculated, with 53 wt% of the total concen-tration calculated, significantly increased compared to pyrolysis. The low concentration of

Fig. 7. Oil product n-alkane and alkene concentration from the liquefaction of mixed waste plastics at 500 ◦C fora 60 min residence time with hydrogen.


alkenes present in the oils for the simulated waste mixture is significant. Previous studies onpyrolysis at atmospheric pressure and under nitrogen (Williams and Williams, 1999a) haveshown that alkenes are produced from the pyrolysis of polyethylene and polypropylene,which represent major components of the simulated plastic mixture. A wide spectrum ofhydrocarbon fragments which may contain any number of carbon atoms may be expectedfrom the thermal degradation of the polyalkene plastics which occurs through random scis-sion (Albertson and Karlsson, 1990; Wampler, 1995). The C–C bond is the weakest in thepolyethylene structure and during the degradation process the stabilisation of the resultantradical after chain scission leads to the formation of carbon double bonds (C C) in the struc-ture in addition to production of alkanes (Pacakova and Leclercq, 1991). Consequently, alarge number of compounds with carbon double bonds, that is, alkenes in the resultant oilwould be expected. The thermal degradation of polypropylene has also been assigned to arandom scission reaction which leads to the formation of a large number of hydrocarbonspecies. Polypropylene is similar in structure to polyethylene and thermal degradation viarandom scission would also result in the production of alkenes as well as alkanes (Turi,1996). Consequently, the influence of the nitrogen pressure is suppressing the formation ofalkenes in favour of alkanes and aromatic compounds. The oil produced from the pyrolysisof polystyrene and polyvinyl chloride would be expected to produce mainly aromatic com-pounds and the oil from polyethylene terephthalate should contain oxygenated compounds(Williams and Williams, 1999a). However, it has been suggested by Carniti et al. (1991)that reactions in closed vessels (polystyrene), where the reaction products cannot leave thereaction system leads to saturation of aromatic compounds at the ethylenic double bond,leading to a wide spectrum of aromatic compounds.

Analysis of the oil from the pyrolysis of the DSD waste plastic under nitrogen showedsignificant production of n-alkanes, with a maximum production for C9 (Fig. 6). Produc-tion of n-alkanes reduced with increased carbon number to less of an extent than withthe other wastes, with significantly reduced concentrations of ≤C11 and increased con-centrations of ≥C15 compared to the other waste plastics used. Alkenes were produced atincreased concentrations compared to the simulated waste plastic mixture. Fig. 7 showsthat the oil produced with hydrogen liquefaction of the DSD waste plastic showed that theconcentrations of alkanes for the higher carbon numbers were reduced for liquefaction inhydrogen compared to pyrolysis in nitrogen, but for the lower carbon numbers, there wereno significant differences in concentration compared to nitrogen.

Pyrolysis of the Fost Plus waste plastic produced an oil with high concentrations ofalkanes compared to the simulated plastics mixture and the DSD waste plastic sample(Fig. 6). This may be expected since the Fost Plus sample would be expected to contain anincreased fraction of polyethylene and polypropylene. Significant production of alkenes wasalso shown. The production of alkanes and alkenes was reduced for hydrogen liquefactioncompared to pyrolysis in nitrogen.

Table 4 shows the concentration of single ring and polycyclic aromatic hydrocarbons(PAH) present in the oils derived from the pyrolysis and liquefaction of the individual plas-tics. All of the product oils produced significant concentrations of aromatic compounds.Previous work with nitrogen pyrolysis at atmospheric pressure has shown that for polyethy-lene and polypropylene, the formation of aromatic compounds was minimal, the oil beingcomposed almost entirely of aliphatic compounds (Williams and Williams, 1999a). The


Table 4Oil product single ring aromatic and PAH compounds identified from the pyrolysis and liquefaction of individualplastics at 500 ◦C for a 60 min residence time

Plastic waste 1 Ring aromatic(wt%)

2 Ring PAH(wt%)

3 Ring PAH(wt%)

Total (wt%)

Pyrolysis (nitrogen)Polyethylene 7.8 8.1 0.5 16.4Polypropylene 10.1 7.4 1.3 18.8Polystyrene 74.5 12.5 13.1 100.1Polyvinyl chloride – – – –Polyethylene terephthalate 11.1 13.9 7.7 32.7

Liquefaction (hydrogen)Polyethylene 7.3 7.8 0.4 15.5Polypropylene 8.7 7.3 1.1 17.1Polystyrene 77.6 11.9 10.9 100.4Polyvinyl chloride 18.1 25.9 8.7 52.7Polyethylene terephthalate 8.2 11.4 2.4 22.0

influence of pressure is therefore to encourage the formation of aromatic compounds.Polystyrene and polyvinyl chloride produced highly aromatic oils and polyethylene tereph-thalate produced an oil high in oxygenated compounds such as carboxylic acids, aldehydesand ketones. The thermal degradation mechanism of polyvinyl chloride is initiated by adechlorination reaction resulting in the formation of hydrogen chloride. The process of theelimination of a chloride atom from the structure results in the formation of a carbon dou-ble bond in addition to hydrogen chloride (Madorsky, 1964). Further carbon double bondsare formed as more hydrogen chloride is evolved from the resultant chain. Eventually thechain undergoes cyclisation to yield aromatic and alkylaromatic compounds (Madorsky,1964). Lattimer and Kroenke (1980) have suggested that hydrogen chloride may promotecross linking through a Diels-Alder type mechanism. The thermal degradation mechanismfor polystyrene in inert atmospheres has been shown to be via firstly chain scission andthen random scission (Raave, 1997). This results in the formation of the styrene monomerand also styrene dimer, trimer and tetramer and other aromatic hydrocarbons includingbenzene, toluene, ethylbenzene and methylbenzene, naphthalene and other polycyclic aro-matic hydrocarbons (Williams and Williams, 1999a; Madorsky, 1964). Styrene was notdetected in any significant concentrations in this work, a result also reported by Pinto etal. (1999) for the pyrolysis of polystyrene under nitrogen pressure. The influence of pres-sure and the long residence time resulting in the degradation of any styrene formed toproduce other aromatic compounds. Carniti et al. (1991) have also reported that styreneand methyl styrene degradation products from polystyrene are readily degraded to ethyl-benzne and cumene, respectively, in closed vessel reaction systems, as used in this work.The main single ring aromatic compounds formed in the pyrolysis and liquefaction productoils in this work were, benzene, toluene, xylenes and ethylbenzene. In general, liquefactionof the plastics with hydrogen produced a less aromatic oil compared to pyrolysis undernitrogen.

Table 5 shows the concentration of aromatic and polycyclic aromatic hydrocarbonsin the oils derived from the pyrolysis and liquefaction of the simulated plastic mixture


Table 5Oil product single ring aromatic and PAH compounds identified from the pyrolysis and liquefaction of mixedwaste plastic at 500 ◦C for a 60 min residence time

Plastic waste 1 Ring aromatic(wt%)

2 Ring PAH(wt%)

3 Ring PAH(wt%)

Total (wt%)

Pyrolysis (nitrogen)Simulated waste 21.1 4.4 3.1 28.6Waste DSD 23.2 5.2 1.2 29.6Waste Fost Plus 20.6 6.0 1.0 27.6

Liquefaction (hydrogen)Simulated waste 20.0 2.7 1.8 24.5Waste DSD 20.0 3.5 0.6 24.1Waste Fost Plus 19.0 4.8 1.1 24.9

and the real world waste DSD and Fost Plus plastics. Liquefaction in hydrogen producedlower levels of aromatic hydrocarbons compared to pyrolysis in nitrogen. The productionof single ring aromatic compounds from liquefaction was around 20 wt% of the productoil concentration. This represents between 6.5 wt% conversion of the total DSD plasticwaste to single ring aromatic compounds and 14 wt% for the total Fost Plus waste plastic.The production of aromatic hydrocarbons, particularly benzene, toluene, ethylbenzene andxylenes from waste plastics represents a potentially significant route for waste plasticsas a source of chemical feedstock. Benzene, toluene and the xylenes have a major useas chemical feedstocks. The major industrial products from benzene are derivatives suchas ethylbenzene, cyclohexane and cumene together with miscellaneous other derivatives.The derivatives are used as basic materials for the production of plastics, resins, fibres,surfactants, dyestuffs and pharmaceuticals, and long chain alkylbenzenes, which are usedas feedstocks in the manufacture of surfactants (Franck and Stadelhofer, 1988). Xylenes arealso regarded as major industrial chemicals and have applications in the plastics industry,for example o-xylene is used to produce phthalic anhydride which is used to produceplasticisers, dyes and pigments, m-xylene derivatives have applications in the polyesterresin and fibre industries and p-xylene derivatives are used in the production of polyesterfibres (Franck and Stadelhofer, 1988). Toluene has a wide range of applications as a chemicalfeedstock and is used for example, in the production of pesticides, dyestuffs, surfactantsand solvents.

4. Conclusions

The main plastics found in municipal solid waste and three samples of mixed plasticwaste have been pyrolysed in an autoclave reactor at 500 ◦C under nitrogen pressure andliquefied at 500 ◦C in a hydrogen atmosphere. The pyrolysis and liquefaction of polyethy-lene and polypropylene produced a mainly oil product and a gas composed largely ofmethane, ethane, propane and butane. Polyvinyl chloride produced significant concen-trations of hydrogen chloride gas and mostly a solid residue. Pyrolysis and liquefactionof polyethylene terephthalate produced mainly carbon dioxide and carbon monoxide and


mostly a solid residue. The hydrocarbon gas composition obtained experimentally from thesimulated plastic mixture was similar to that calculated from the gas compositions producedfrom the individual plastics. The oils produced from the mixed waste plastic samples con-tained high concentrations of alkanes but also high concentrations of single ring aromaticcompounds.

Acknowledgements

The research was supported by the Faraday Packaging Partnership via a research studentscholarship.

References

Albertson AC, Karlsson S. Glass JE, Swift G, editors. Agricultural and synthetic polymers, acs symposium series433. Washington DC: American Chemical Society; 1990. p. 60–4.

APME. Good practices guide on waste plastics recycling a guide by and for local and regional authorities. Brussels,Belgium: Association of Plastics Manufacturers in Europe; 2004.

Bagri R, Williams PT. Catalytic pyrolysis of polyethylene. J Anal Appl Pyrolysis 2002;63:29–41.Bhaskar T, Uddin MA, Murai K, Kaneko J, Hamano K, Kusaba T, Muto A, Sakata Y. Comparison of thermal

degradation products from real municipal waste plastic and model mixed plastics. J Anal Appl Pyrolysis2003;70:579–87.

Carniti P, Beltrame PL, Armada M, Gervasini A, Audisio G. Polystyrene thermodegradation. 2. Kinetics offormation of volatile products. Ind Eng Chem Res 1991;30:1624–9.

de Marco I, Caballero B, Torres A, Laresgoiti MF, Chomon MJ, Cabrero MA. Recycling polymeric wastes bymeans of pyrolysis. J Chem Technol Biotechnol 2002;77:817–24.

Ding W, Liang J, Anderson LL. Thermal and catalytic degradation of high density polyethylene and commingledpost-consumer plastic waste. Fuel Process Technol 1997;51:47–62.

Feng Z, Zhao J, Rockwell J, Bailey D, Huffman G. Direct liquefaction of waste plastics and co-liquefaction ofcoal-plastic mixtures. Fuel Process Technol 1996;49:17–30.

Franck HG, Stadelhofer JW. Industrial aromatic chemistry. London: Springer-Verlag; 1988.Kaminsky W, Predel M, Sadiki A. Feedstock recycling of polymers by pyrolysis in a fluidised bed. Polym Degrad

Stab 2004;85:1045–50.Lattimer RP, Kroenke WJ. Formation of volatile pyrolyzates from polyvinyl chloride. J Anal Appl Pyrol

1980;25:101–10.Manos G, Garforth A, Dwyer J. Degradation of high density polyethylene on different zeolitic structures. Ind Eng

Chem Res 2000;39:1196–8.Madorsky SL. Thermal degradation of organic polymers. New York: Interscience Publishers; 1964.McCaffrey WC, Brues MJ, Cooper DG, Kamal MR. Thermolysis of polyethylene/polystyrene mixtures. J Aappl

Polym Sci 1996;60:2133–40.Murty MVS, Rangarajan P, Grulke EA, Bhattacharyya D. Thermal degradation/hydrogenation of commodity

plastics and characterisation of their liquefaction products. Fuel Process Technol 1996;49:75–90.Pacakova V, Leclercq PA. Gas-chromatography mass-spectrometry and high-performance liquid-chromatographic

analyses of thermal-degradation products of common plastics. J Chromatog 1991;555:229–37.Pinto F, Costa P, Gulyurtlu I, Cabrita I. Pyrolysis of plastic wastes; 1. Effect of plastic waste on product yield. J

Anal Appl Pyrolysis 1999;51:39–55.Raave A. Organic chemistry of macromolecules—an introductory textbook. London: Edward Arnold Pubs; 1997.Ramdoss PK, Tarrer AR. High temperature liquefaction of waste plastics. Fuel 1998;77:293–9.Shabtai J, Xiao X, Zmierczak. Deploymerisation—liquefaction of plastic and rubbers. 1. Polyethylene, polypropy-

lene and polybutadiene. Energy Fuels 1997;11:76–87.


Shah N, Rockwell J, Huffman GP. Conversion of waste plastic to oil: direct liquefaction versus pyrolysis andhydroprocessing. Energy Fuels 1999;13:832–8.

Turi E. Thermal analysis. Berlin: Brandrup and Immergut Pubs; 1996.Wampler TP. Wampler TP, editor. Applied polymer handbook. New York: Marcel Dekker Inc.; 1995.Williams PT. Waste treatment & disposal. 2nd ed. Chichester: John Wiley and Sons Ltd.; 2005.Williams PT. Yield and composition of gases and oils/waxes from the feedstock recycling of waste plastics. In:

Schiers J, Kaminsky W, editors. Feedstock recycling and pyrolysis of waste plastics. Chichester: John Wiley& Sons; 2006.

Williams PT, Bagri R. Hydrocarbon gases and oils from the recycling of polystyrene waste by catalytic pyrolysis.Int J Energy Res 2004;28:31–44.

Williams EA, Williams PT. Analysis of products derived from the fast pyrolysis of plastic waste. J Anal ApplPyrolysis 1997;40/41:347–63.

Williams PT, Williams EA. Interaction of plastics in mixed plastics pyrolysis. Energy Fuels 1999a;13:188–96.Williams PT, Williams EA. Fluidised bed pyrolysis of low density polyethylene to produce petrochemical feed-

stock. J Anal Appl Pyrolysis 1999b;51:107–26.

analiza polimerilor in procesul de piroliza

Documents