J. Anal. Appl.

Pyrolysis 79 (2007) 409414

www.elsevier.com/locate/jaap

Petrochemical feedstock by thermal cracking of plastic waste

Andras Angyal *, Norbert Miskolczi, Laszlo Bartha
Department of Chemical Engineering, Faculty of Engineering, University of Pannonia, 10 Egyetem St.,
H-8201 Veszprem, P.O. Box 158 Veszprem, Hungary
Received 23 June 2006; accepted 18 December 2006
Available online 8 January 2007

Abstract
The thermal cracking is one of the possible ways to obtain petrochemical feedstock from polymer wastes. The properties of the cracking
products depend mainly on the type of the raw material. In this study the degradation of different mixtures of polymer wastes (polypropylene and
polystyrene) were investigated in a horizontal tube reactor. Temperature of 510520 8C and residence time of 1530 min were used. The effects of
the concentration of polystyrene on the yields and structure of products were also studied. Moreover, the interaction between the polymers during
the pyrolysis was investigated. The products of the thermal cracking were analyzed by gas-chromatography, infrared spectroscopy and different
standardized and non-standardized methods.
It was found, that the concentration of polystyrene significantly affected both the quantity and quality of degradation products. It was also
observed that the presence of the polystyrene in the raw material increased the reaction rate coefficient, and enhanced the degradation rate of
polypropylene. By higher concentration of polystyrene in raw materials, the yield of light products considerably increased, furthermore, in the
naphtha-like fraction the concentration of aromatics also increased. The characteristics of the products were changed with the polystyrene content
in the raw materials. The liquid products had advantageous properties for further utilization as fuel like or petrochemical feedstock applications.
# 2007 Elsevier B.V. All rights reserved.
Keywords: Mixed plastic waste; Tube reactor; Thermal degradation; Reaction rate; Utilization

1. Introduction
Recently, more than 150 million tonnes of plastics were
produced in the world. On the one hand, the consumption and
production of polymers are increasing, on the other hand, the
increasing amount of polymer wastes from them generates
further mainly environmental problems whereas the utilization
of wastes is important also from energetic and political aspects.
The thermal degradation of plastic wastes is one of the
prospective ways to solve this problem. This way the plastic
wastes are converted into fuels or other valuable feedstocks for
the petrochemical industry. Numerous researchers have
contributed to the theory and practice of thermal degradation
(with or without catalyst) of polymers. Mainly polyolefins
(polyethylene (PE) and polypropylene (PP)) and polystyrene
(PS) were the target polymers, because their cracking resulted
in products with favourable properties for further application

* Corresponding author. Tel.: +36 88 624 178; fax: +36 88 624 520.
E-mail address: angyala@almos.vein.hu (A. Angyal).
0165-2370/$ see front matter # 2007 Elsevier B.V. All rights reserved.
doi:10.1016/j.jaap.2006.12.031

[14]. According to results the products of polyethylene,

polypropylene and even polystyrene cracking have the most
favourable properties for further energetic applications. PEderived fuel has very high cetane and octane number, because it
is rich in linear paraffins and a-olefins. Linear paraffin content
in the gas oil fraction is advantageous, because this
hydrocarbon structure has the highest cetane rate, but also
has the lowest octane number in the naphtha fraction. However,
PE-derived naphtha-like fraction has high octane number,
which comes from the high content of olefins. The liquid
products of cracking of PP contain primarily olefins and
isoparaffins that resemble the molecular skeleton of PP. Both
hydrocarbon structures are advantageous for further utilization.
Neat PS feedstocks will depolymerise in cracking processes to
give predominately styrene monomer. Therefore, the fuel made
from polystyrene feedstock will be high by aromatic character
and have good energy content. The aromatic hydrocarbons have
the highest octane number in the naphtha fraction [5].
The degradation of these polymers in their individual form is
well described. The literature shows the results of different
degradation techniques, the yield and the characteristics of the

410

A. Angyal et al. / J. Anal. Appl. Pyrolysis 79 (2007) 409414

products and the degradation mechanism and kinetic data were

presented [613].
Description of decomposition reactions during degradation
of plastics is quite difficult, because they are very complex.
The determination method of the reaction kinetic depends on
the degradation technology of the plastic wastes. For example,
there are considerable differences in the thermal degradation
of plastic wastes between the absence and presence of
catalysts. Cracking experiments are done mostly in batch
reactors connected to thermogravimetric analysis by using
different techniques. Only a few researches have investigated
the cracking of polymer wastes under continuous or semicontinuous conditions. Some kinetic models for thermal
degradation have been proposed. The commonly used
approach is first-order kinetics to investigate the characteristics of degradation. In this approach first the weight loss
curve of polymers during the decomposition was determined,
and overall reaction rate constants were calculated. The
objectives of experiments were the determination of the
apparent activation energies and other reaction kinetic
parameters (reaction rate, pre-exponential constant, etc.) of
degradation resulting volatile products (gases and liquids).
Besides the development of kinetic models the prediction of
the yields or main properties of products of the raw material
are also important. The principal problem in this case is the
derivation of reaction rate constant because some factors
significantly affect the conversion of degradation. Usually, a
constant value of reaction rate was postulated and first-order
kinetic equation together with the Arrhenius equation was
used [14].
In the literature there are considerably less experimental
results on the degradation of mixtures of plastic wastes, e.g.
mixture of HPDE and PS, PP and PS, etc. The effect of the
composition of feeding polymers on the properties of products
was studied in the experimental work of Miskolczi et al. [3].
The thermal cracking of different mixtures of commercial
plastic wastes from the packaging- and other industry was
investigated. The chemical structure of polymer waste had a
considerable effect on the yield and properties of the products.

Table 1
The main properties of waste polymers
Properties

PP

PS

Polymer
Origin
Density (g/cm3)
Particle size (mm)
MFIa (g/10 min)
S content (ppm)
N content (ppm)

Polypropylene
Packaging
0.909
46
0.366
14.8
6.1

Polystyrene
Electrical
1.040
34
0.628
16.2
7.3

At 180 8C, using 2160 g load.

The yields of volatile products were higher by the degradation

of the polypropylene contained raw material. Furthermore, the
liquid products had low heteroatom content (sulphur, nitrogen).
In case of the cracking of polyurethane rubber and polyamide
contained polymer waste the sulphur and nitrogen concentration increased in the products.
Other studies in the literature [1520] have investigated the
degradation of polyethylene or polypropylene and polystyrene
mixtures, and the results of the interactions between these
polymers. Williams and Williams [17] have analyzed the
binary degradation of polystyrene and polypropylene, and it
was found that the overall degradation rate of the mixture was
greater than the combined degradation rates of the individual
components.
In our experimental work the degradation of different
mixtures of polymer wastes (polypropylene and polystyrene)
was investigated in a horizontal tube reactor. The goal of our
work was to study the effect of the concentration of polystyrene
on the yields and structure of products.
2. Experimental
2.1. Raw material
Commercial plastic wastes from the packaging- and
electrical industry were used as raw materials. The main

Fig. 1. Cracking apparatus: (1, 2) raw material tanks, (3) extruder, (4) electric motor, (5) reactor, (6) separator, (7) cooler, (8) separator, (9) distillation column, (10)
vacuum distillation column and (1114) product storage tanks.

A. Angyal et al. / J. Anal. Appl. Pyrolysis 79 (2007) 409414

411

properties of raw materials are shown in Table 1. In the

selection of the raw materials the main consideration was
the possibility of collection and the purity of the polymer
wastes.
2.2. Cracking apparatus
The decomposition process of the mixtures of the PP and
PS polymers was carried out in a tube reactor system
consisted of three main parts: an extruder, a reactor and a
separator (Fig. 1). The waste polymers were fed by an
extruder then the preheated polymer was driven from the
extruder into the reactor. Inside the reactor the polymers were
melted and their carbon chain cracked into lower molecular
fragments. The reactor wall temperature was 510520 8C and
residence time of 1530 min was used. In preliminary
separator the formed hydrocarbons were separated into
volatile products and heavy oil. The separator temperature
was fixed at 350 8C. The non-condensed fraction at this
temperature leaved the separator on the top and the bottom
products were the condensed hydrocarbons. The volatile
fraction was driven through the water cooler to another
separator. In this separator the gas and the liquid fraction
were separated at ambient temperature. Furthermore, the
liquid fraction was separated by laboratory atmospheric and
vacuum distillations to three other products: naphtha, middle
distillates and light oil.
2.3. Analysis of products
Each fraction of cracking was analyzed by using the
following standardized test methods: liquid density measurement (MSZ EN ISO 12185), determination of distillation curve
(ASTM-D 1078), of sulphur and nitrogen content of naphtha
and middle distillate fractions (ASTMD 6428 99 and ASTMD
6366 99), of flash point (ISO 2719:2002 and MSZ 15967:1979),
of cold filter plugging point (CFPP) (MSZ EN 116:1999), of
kinematic viscosity (MSZ ISO 3105:1998) and copper
corrosion (MSZ EN ISO2160:2000). Gases were analyzed
by a Carlo-Erba Vega Series GC 6000 gas chromatograph
provided with a 50 m 0.32 mm fused silica column with Al2O3/
KCl coating, at 40 8C. Liquid products were analyzed also by
gas chromatographs a TRACE GC gas chromatograph with a
flame ionization detector. It was provided with a
30 m  0.32 mm Rtx1 (Crossbond1 100% dimethyl-polysiloxane) column. The olefin content of liquids was determined
by infrared technique by a SHIMADZU IR-470 type spectrometer (resolution: 2.7 cm1, illumination: Globar light,
monocromator: Littrow prism, detector: KRS 5 type detector
with a pyrometer in crystal window vacuum cell) in the 400
4000 cm1 wavenumber range. The concentration of sulphur in
light oil and heavy oil fractions was determined by energy
dispersive X-ray fluorescence spectrometer (EDXRF), PHILLIPS PW 4025/02 (MiniPal). Research octane number (RON),
motor octane number (MON) and cetane number were
determined from their infrared spectra with a ZX-101c type
instrument.

Fig. 2. Yields of the volatile products by different PS concentrations and

residence time.

3. Results and discussion

3.1. Yields of products
The yields of volatile products mainly gases, naphtha,
middle distillates and light oil fractions are shown in Fig. 2. The
percentage of the volatile hydrocarbons means the conversion
of the degradation reactions. In this case the heavy oil fraction
is the residue of the cracking of polymer waste. The amount of
this residue belongs to the mass of the unconverted polymer in
the later kinetic calculations. According to data of Fig. 2 the
yields of valuable light products considerably increased with
the residence time and the concentration of polystyrene in the
raw material, which could be caused by the differences in the
thermal stability of polymer chains and the possible interactions between the cracked fragments. As it is well known that
the activation energy of destruction is approximately equal in
the case of PP and PS, therefore, the considerable increase of
conversion can be the consequence of the interactions between
radicals from PP and PS.
The interaction between the polymers in the degradation
process was investigated on the calculation of the degradation
reaction rates at different concentrations of the polystyrene.
Thus, the degradation reaction rates were determined.
Considering the complexity of the degradation reaction
Eqs. (1) and (2) were used for the calculation of constant
value of reaction rates supposed the constant reaction
temperature and the first-order kinetic equation.
dm
kmn
dt

(1)

k A0 eE=RT

(2)

m m0 ekt

(3)

m
kt
m0

(4)

ln

Where m0 is the weight of the sample, n the reaction order, k the

reaction rate coefficient, m the weight of the residue, t the time
of degradation, E the activation energy and A0 is the preexponential constant.

412

A. Angyal et al. / J. Anal. Appl. Pyrolysis 79 (2007) 409414

Fig. 3. The experimental reaction rates as ln(m/m0) by different PS concentrations and residence time.
Fig. 5. The yield of products at 23 min residence time.

The goal of our kinetic study was the calculation of the

reaction rates at constant temperature and different polystyrene
content, with Eq. (4). The decomposition of pure PS was not
investigated in this study, because during the cracking of pure
PS the formation of the coke was very intensive and it caused
technical problems by this reactor system. Thus, two types of
reaction rates were calculated one from the experimental data
and one from the hypothetical reaction. In case of the
hypothetical reaction it was stated that there were not any
interactions between radicals from PP and PS supposing that
the conversion of the volatile products of PP was equal with the
experimental conversion of volatile products of pure PP by the
calculation of the hypothetical reaction rates. Furthermore, the
PS conversion was 100% by the hypothetical reaction. The
experimental and hypothetical reaction rates are shown in
Figs. 3 and 4. The slopes of the curves (Fig. 3) were given by the
reaction rate coefficients of the thermal cracking of the
polymers in the continuous tube reactor in the 510520 8C
interval. In Fig. 4, the slope of the curves was given by the
hypothetical reaction rate coefficients.
In case of the degradation of the pure polypropylene the
reaction rate coefficient was 0.0162. The same coefficient was
0.0279 in the case of polystyrene concentration of 20% in the
raw material. Furthermore, the experimental reaction rates were
always higher than the hypothetical data. So it could be found
that, the radical fragments from preliminary degradation of PS
could react with other still not degraded molecules and initiate

Fig. 4. The hypothetical reaction rates as ln(m/m0) by different PS concentrations and residence time.

the further cracking reactions. It is in agreement with the results

of Kruse et al. [15]. It was proposed that low molecular weight
radicals were diffusing into polypropylene and abstracting
hydrogen from polypropylene, resulting in an enhanced
degradation rate of polypropylene.
The yields of products at 510 8C temperature and 23 min
residence time are shown in Fig. 5. According to the tendencies
the yields of valuable light products considerably increased
with the concentration of polystyrene. The cracking parameters
(residence time and temperature) did not significantly affect the
composition of the final product, as it was presented by the
following characteristics of these products from the thermal
cracking at 510 8C and 23 min.
3.2. Gases
In the gases formed in the cracking of PP waste the propene
was the main component. In case of the cracking of the
mixtures of PP and PS the cracking products of PP were
dominant in the gas fraction, because the yield and the
composition of the gases were not changed with the
concentration of PS in the raw material (Figs. 5 and 6). So it
could be assumed that the gases derived from the degradation of
PP. The calculated heating values of gases were more than

Fig. 6. The composition of gases by different PS concentrations.

A. Angyal et al. / J. Anal. Appl. Pyrolysis 79 (2007) 409414

413

Fig. 7. The hydrocarbon structural composition of naphtha-like fraction by

different PS concentrations.
Fig. 9. The composition of aliphatic hydrocarbons of the naphtha-like fraction.

45 MJ/kg, which were quite high for energy utilization, but this
high olefin (propene) content could be advantageous also for
this utilization as petrochemical feedstock.
3.3. Naphtha
The naphtha-like fraction contained C5C15 hydrocarbons.
Identifying the individual compounds in the liquid fractions is
rather difficult because of the great number of isomer compounds
formed in the cracking of polypropylene. Thus, the paraffin,
olefin and aromatic hydrocarbon structure of the products were
analyzed (Fig. 7). According to our previous experiences the
formation of aromatic hydrocarbons was not detectable in the
cracking process of polyolefins below 560 8C in this reactor
system [3]. In this study the aromatic content was observed only
in the cracking products of polystyrene containing raw materials.
The concentration of the aromatics was increased with the
polystyrene concentration of the raw materials. It is well known
that cracking of polystyrene results styrene monomer, ethylbenzene and toluene in the highest probability, beside formation
of benzene, xylenes and other aromatics which were observed at
a smaller extent (Fig. 8). In the composition of aliphatic
hydrocarbons the triplet sequence could be observed, which is
typical for decomposition of polypropylene. Due to the high
propylene content of samples the C6, C9 and C12 hydrocarbons
were present in the highest concentration (Fig. 9).
3.4. Middle distillates

concentration of aromatics was not more than 1%. It was

concluded that the aromatic hydrocarbons derived from the
decomposition of PS and their cracking products with lower
boiling point stayed in the naphtha-like fraction. The
aliphatic hydrocarbon content of the middle distillates
fraction had also a triplet sequence and the main compounds
were the C15, C18 and C21 hydrocarbons. The composition of
middle distillates is shown in Fig. 10. In all cases the
percentage of paraffin and olefin contents was almost similar,
about 50%.
3.5. Light and heavy oil
It was found that the light oil fraction had only olefin and
paraffin content without aromatics and naphthenes. This
fraction appeared in a solid phase at room temperature due
to their paraffinic hydrocarbon structure with boiling point
range of 250350 8C. Their hydrocarbon composition did not
change significantly as the function of concentration of
polystyrene in the raw material mixture.
The residue of cracking is the heavy oil fraction
(bp > 350 8C) that contained mainly paraffins, olefins and
aromatics. Aromatics were present in the shape of oligomers of
styrene with higher carbon number, which could not distill from
melted polymers. These fractions were wax-like materials with
melting points over 100 8C, and high molecular weight, about
2500 g/mol.

The middle distillates fraction contained C12C28 hydrocarbons. Independent of the raw materials composition the

Fig. 8. The composition of aromatic hydrocarbons of the naphtha-like fraction.

Fig. 10. The hydrocarbon structural composition of middle distillates.

414

A. Angyal et al. / J. Anal. Appl. Pyrolysis 79 (2007) 409414

Table 2
The main properties of naphtha-like fraction
PS concentration (%)

10

20

Density (g/cm )
RON
MON
Sulphur content (ppm)
Nitrogen content (ppm)
Corrosion test

0.764
78
70
17.1
16.5
Group 1

0.751
79
71
9.5
10.2
Group 1

0.764
90
76
18.4
14.8
Group 1

0.783
96
80
14.6
11.0
Group 1

Distillation data (8C)

IBP
30
50
70
FBP

38
109
137
146
206

40
99
127
137
204

38
81
129
141
201

37
71
131
141
205

Table 3
The main properties of middle distillates fraction
PS concentration, %

10

20

Density (g/cm3)
Viscosity 408C (mm2/s)
Sulphur content (ppm)
Nitrogen content (ppm)
CFPP (8C)
Flash point (8C)
Cetane number
BMCI

0.831
4.30
14.6
16.2
2
81
63
40

0.826
4.57
10.1
16.0
1
78
63
39

0.825
4.30
9.3
14.6
1
75
62
38

0.833
4.20
12.7
13.5
2
71
61
41

Distillation data (8C)

IBP
30
50
70
FBP

64
185
196
215
269

58
183
190
214
268

62
180
194
220
271

70
187
197
222
265

3.6. Properties of liquid products

The main properties of cracking products are shown in
Tables 2 and 3. According to data of the analysis, both fractions
contained low amounts of heteroatoms (sulphur and nitrogen).
The low heteroatom content seemed to be advantageous for the
further fuel-like applications, because of the European Union
criterions prescribed the maximum sulphur (50 ppm) content
both in the gasoline and gas oil fuels. In the naphtha-like
fraction the aromatic content was increased with the
polystyrene concentration in the feed plastic mixture. Due to
high olefins and aromatics content the octane numbers of the
naphtha-like fraction were high. The fractions of middle
distillates were rich in paraffin compound, however, their
cetane numbers were high, while the CFPPs were rather low.
The further utilization of middle distillates as petrochemical
feedstock can be also advantageous, because the aromatics
concentration of this fraction was under 1%, and the Bureau of
Mines Correlation Index (BMCI) was also low in all case, about
40. A high value (>80) of BMCI indicates a highly aromatic
hydrocarbon fraction, a low value (<30) a highly saturated
hydrocarbon fraction. The BMCI calculated empirically from
the hydrocarbon fractions density and from its molal average
boiling point. These properties of the middle distillates can be

favourable for further gas oil steam cracking producing also

olefin monomers and realizing by this way the total recycling of
polymer wastes.
4. Conclusions
In our experimental work the thermal cracking of
polypropylene and polystyrene mixtures were investigated. It
was found, that the concentration of polystyrene significantly
affected both the quantity and quality of degradation products.
It was also observed that the presence of the polystyrene in the
raw material increased the reaction rate coefficient, and
enhanced the degradation of polypropylene. With increasing
concentration of polystyrene in raw materials, the concentration of aromatics considerably increased, furthermore, the
characteristic of products also changed notably (e.g. both the
Research Octane Number and Motor Octane Number of the
gasoline fractions were increased). However, the middle
distillates had in all case low BMCI. The liquid products
had low heteroatom content (sulphur, nitrogen) and moderate
olefin content.
Acknowledgements
The authors gratefully acknowledge the financial support for
this work of the Chemical Engineering Institutes Cooperative
Research Centre (especially Ministry of Education of Hungary
and Hungarian Oil and Gas (MOL) PLC).
References
[1] I.H. Metecana, A.R. Ozkana, R. Islera, J. Yanikb, M. Saglamc, M.
Yukselc, Fuel 84 (2005) 619.
[2] K. Gobin, G. Manos, Polym. Degrad. Stab. 83 (2004) 267.
[3] N. Miskolczi, L. Bartha, G. Deak, B. Jover, Polym. Degrad. Stab. 86
(2004) 357.
[4] A. Demirbas, J. Anal. Appl. Pyrolysis 72 (2004) 97.
[5] J. Scheirs, in: J. Scheirs, W. Kaminsky (Eds.), Feedstock Recycling and
Pyrolysis of Waste Plastics, Wiley, 2006, p. 383 (Chapter 15).
[6] K. Gobin, G. Manos, Polym. Degrad. Stab. 86 (2004) 225.
[7] Y.H. Lin, H.Y. Yen, Polym. Degrad. Stab. 89 (2005) 101.
[8] D.P. Serrano, J. Aguado, J.M. Escola, J.M. Rodrguez, J. Anal. Appl.
Pyrolysis 74 (2005) 353.
[9] T. Faravelli, M. Pinciroli, F. Pisano, G. Bozzano, M. Dente, E. Ranzi, J.
Anal. Appl. Pyrolysis 60 (2001) 103.
[10] S.C. Cardona, A. Cormab, Catal. Today 75 (2002) 239.
[11] R.A. Garca, D.P. Serrano, D. Otero, J. Anal. Appl. Pyrolysis 74 (2005)
379.
[12] K. Murata, K. Sato, Y. Sakata, J. Anal. Appl. Pyrolysis 71 (2004) 569.
[13] H. Bockhorn, A. Hornung, U. Hornung, D. Schawaller, J. Anal. Appl.
Pyrolysis 48 (1999) 93.
[14] N. Miskolczi, in: J. Scheirs, W. Kaminsky (Eds.), Feedstock Recycling and
Pyrolysis of Waste Plastics, Wiley, 2006, p. 225 (Chapter 9).
[15] T.M. Kruse, S.E. Levine, H.W. Wong, E. Duoss, A.H. Lebovitz, J.M.
Torkelson, L.J. Broadbelt, J. Anal. Appl. Pyrolysis 73 (2005) 342.
[16] C.G. Jung, A. Fontana, in: J. Scheirs, W. Kaminsky (Eds.), Feedstock
Recycling and Pyrolysis of Waste Plastics, Wiley, 2006, p. 252 (Chapter
10).
[17] P.T. Williams, E.A. Williams, Energy Fuels 13 (1999) 188.
[18] J.R. Kim, J.H. Yoon, D.W. Park, Polym. Degrad. Stab. 76 (2002) 61.
[19] K.H. Lee, D.H. Shin, Waste Manage. 27 (2007) 168.
[20] K.H. Lee, D.H. Shina, Y.H. Seob, Polym. Degrad. Stab. 84 (2004) 123.