Polymer Degradation and Stability 89 (2005) 101e108

www.elsevier.com/locate/polydegstab

Fluidised bed pyrolysis of polypropylene over cracking

catalysts for producing hydrocarbons
Y.-H. Lin*, H.-Y. Yen
Department of Biochemical Engineering & Graduate Institute of Environmental Polymeric Materials,
Kao Yuan Institute of Technology, 821 Kaohsiung, Taiwan, ROC
Received 20 October 2004; accepted 7 January 2005
Available online 19 February 2005

Abstract
Polypropylene (PP) was pyrolysed over various catalysts using a laboratory uidised-bed reactor operating isothermally at
ambient pressure. The yield of volatile hydrocarbons for zeolite catalysts (HZSM-5 O HUSY z HMOR) was higher than that for
non-zeolite catalysts (SAHA z MCM-41). Product distributions with HZSM-5 contained more olenic materials with about
60 wt% in the range of C3eC5. However, both HMOR and HUSY produced more paran streams with large amounts of isobutane
(i-C4) and both catalysts were deactivated during the course of the degradation. SAHA and MCM-41 showed the lowest conversion
and generated an olen-rich product with a rise to the broadest carbon range of C3eC7. Experiments carried out with various
catalysts gave good yields of volatile hydrocarbons with diering selectivity in the nal products dependent on reaction conditions.
The selectivity could be further inuenced by changes in reactor conditions; in particular, olens and iso-olens were produced by
low temperatures and short contact times. It is demonstrated that under appropriate reaction conditions suitable catalysts can have
the ability to control both the product yield and product distribution from polymer degradation, potentially leading to a cheaper
process with more valuable products.
2005 Elsevier Ltd. All rights reserved.
Keywords: Pyrolysis; Polypropylene; Fluidised-bed reactor; Catalysts; Degradation; Selectivity

1. Introduction
Polymer waste is a cause of increasing environmental
concern in the developed world [1]. In view of their nonbiodegradability, most polymers are felt unsuitable for
landll disposal. The destruction of wastes by incineration is prevalent, but this practice is expensive and often
generates problems with unacceptable emissions [2,3].
Several methods for the recycling of waste polymers have
been proposed, of which pyrolysis is a promising approach with potential applications. Pyrolysis, i.e. thermal
cracking of the polymers to give low molecular weight
* Corresponding author. Tel.: C886 7 6077777; fax: C886 7
6077788.
E-mail address: lin@cc.kyit.edu.tw (Y.-H. Lin).
0141-3910/$ - see front matter 2005 Elsevier Ltd. All rights reserved.
doi:10.1016/j.polymdegradstab.2005.01.006

materials, is attractive, but unfortunately gives a very

broad range of products. It also requires high temperatures typically more than 500
C and even up to 900
C
[4e6]. Catalytic pyrolysis is being investigated as a means
to address these problems [7e9].
The addition of catalyst is expected to reduce decomposition temperature, to promote decomposition
speed, and to modify the products. Catalytic pyrolysis
has been carried out by considering a variety of catalysts
with little emphasis on the reactor design, with only
simple adiabatic batch and xed bed reactors being used.
Studies of the eects of catalysts on the catalytic
degradation of polymer have been performed by contacting melted polymer with catalyst in xed bed reactors
[10e12], heating mixtures of polymer and catalyst
powders in reaction vessels [13e15], and passing the

102

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108

products of polymer pyrolysis through xed bed reactors

containing cracking catalysts [16e18]. The use of xed
beds or adiabatic batch where polymer and catalyst are
contacted directly leads to problems of blockage and
diculty in obtaining intimate contact over the whole
reactor. Without good contact the formation of large
amounts of residue are likely, and scale-up to industrial
scale is not feasible [19e21].
Much less is known about performance of catalysts in
polymer degradation using a uidised-bed reactor. A
uidised-bed reactor has been used for obtaining hydrocarbon products from the catalytic degradation of high
density polyethylene (HDPE) in our previous study [22].
It also seems that in the process of polymer pyrolysis,
a particular kind of catalyst is eective for a particular
kind of polymer. Of course, some polymeric materials,
e.g., polystyrene, can be decomposed thermally in high
yields to the monomers. However, this is not true for
polyethylene (PE) or polypropylene (PP), which are
among the most abundant polymeric waste materials
typically making up 60e70% of municipal solid waste
[23,24]. It would be desirable to convert these waste
polyolens into products of value other than the
monomers, because the products could be of sucient
value to oset the collection and pyrolysis costs. The
objective of this current work was to expand the
capabilities of a laboratory catalytic uidised-bed reaction system using various catalysts (i) for the study of
product distributions in the pyrolysis of PP and (ii) for
identication of suitable reaction conditions for achieving
polymer recycling.

2. Experimental
2.1. Materials and reaction preparation
The polymer used in this study was pure polypropylene
(PP; isotactic, r Z 853.6 kg m3, MW z 332,000, Aldrich). The catalysts employed are described in Table 1.
Prior to use, all the catalysts were pelleted, crushed and
sieved to give particle sizes ranging from 75 to 180 mm.
The catalyst (0.2e0.3 g) was then activated by heating in
the reactor in owing nitrogen (50 mL min1) to 120
C
at 60
C h1. After 2 h the temperature was increased to
520
C at a rate of 120
C h1. After 5 h at 520
C the
reactor was cooled to the desired reaction temperature.
Several uidisation runs were performed at ambient
temperature and pressure to select suitable particle sizes
(both catalyst and polymer) and to optimise the uidising gas ow rates to be used during the reaction. The
particle sizes of both catalyst (75e180 mm) and polymer
(75e250 mm) were chosen as being large enough to
avoid entrainment but not too large to be inadequately
uidised.

Table 1
Catalysts used in uidised-bed reactor for PP degradation
Catalyst

Micropore
size (nm)

BET area
(cm3/g)

HUSY

0.74

603

6.0

HZSM-5
HMOR
SAHA

0.55 ! 0.51
0.65 ! 0.70
3.15d

391
561
274

17
6.3
2.6

MCM-41

4.2e5.2d

845

17.5

a
b
c
d
e

Si/Al
ratio

Commercial
name
H-Ultrastabilised
Y zeolitea
H-ZSM-5 zeoliteb
H-Mordenitec
Synclyst 25a
(Silicaealumina)
ee

Croseld Chemicals, Warrington, UK.

BP Chemicals, Sunbury-on-Thames, UK.
Laporte, Warrington, UK.
Single point BET determination.
Synthesized by procedure outlined by Beck et al. [25].

2.2. Experimental procedures and product analysis

A process ow diagram of the experimental system is
given elsewhere [22] and shown schematically in Fig. 1.
The reactor consists of a pyrex glass tube
(170 mm ! 20 mm i.d., upper section; 30 mm ! 10 mm
i.d., middle section; 200 mm ! 10 mm i.d., lower section)
with a sintered distributor (10 mm i.d., in the middle
section) and a three-zone heating furnace with digital
controllers. High-purity nitrogen was used as the uidising gas and the ow was controlled by a needle valve.
The polymer feed system was designed to allow PP
particles, purged under nitrogen, to enter the top of the
reactor and to drop freely into the uidised bed at
t Z 0 min. The added polymer melts, wets the catalyst
surface and is pulled into the catalyst macropores by
capillary action [26]. At suciently low polymer/catalyst
ratios the outside of the catalyst particles are not wet with
polymer, so the catalyst particles move freely.
Volatile products leaving the reactor were passed
through a glass-bre lter to capture catalyst nes,
followed by an iceewater condenser to collect any
condensable liquid product. A three-way valve was used
after the condenser to route product either into a sample
gas bag or to an automated sample valve system with 16
loops. The Tedlar bags, 15 L capacity, were used to
collect time-averaged gaseous samples. The bags were
replaced at intervals of 10 min throughout the course of
reaction. The multiport sampling valve allowed frequent, rapid sampling of the product stream when
required. Spot samples were collected and analysed at
various reaction times (t Z 0.5, 1, 2, 3, 5, 8, 11, 15 min).
The rate of hydrocarbon production (Rgp, wt% min1)
was dened by the relationship:
Rgp Z

Hydrocarbon production rateg=min!100

Total hydrocarbon product over the whole rung

Gaseous products were analysed using a gas chromatograph equipped with (i) a thermal conductivity detector

103

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108
N2 V1
1

TC1
V3

6
4

TC2

V2

10

V4

11

TC3

Fig. 1. Schematic diagram of a catalytic uidised-bed reactor system: 1. Feeder; 2. Furnace; 3. Sintered distributor; 4. Fluidised catalyst; 5. Reactor;
6. Condenser; 7. Flow meter; 8. 16-loop automated sample system; 9. Gas bag; 10. GC; 11. Digital controller for three-zone furnace.

(TCD) tted with a 1.5 m ! 0.2 mm i.d., molecular sieve

13! packed column and (ii) a ame ionisation detector
(FID) tted with a 50 m ! 0.32 mm i.d., PLOT Al2O3/KCl
capillary column. A calibration cylinder containing 1%
C1eC5 hydrocarbons was used to help identify and
quantify the gaseous products. The remaining solid
deposited on the catalyst after the catalytic degradation
of the polymer was deemed residues and contained
involatile products and coke. The amount and nature of the
residues were determined by thermogravimetric analysis
(TA Instruments, SDT 2960 Simultaneous DTA-TGA) as
described elsewhere [27]. Data analysis was carried out
using a Microsoft Excel spreadsheet. The reactor and
various units of the collection system were weighted before
and after the runs to determine the mass balance.

3. Results and discussion

Catalytic pyrolysis products (P) are grouped together
as hydrocarbon gases (!C5), gasoline up to C9 (C5eC9),
liquids (condensate in condenser and lter) and residues
(coke and products, involatile at reaction temperature,
deposited on catalyst) to enable the overall pyrolysis
processes to be described more easily. The term yield
as used in this paper is dened by the relationship:
Yieldwt%Z

Pg!100
Polymer Fedg

A number of runs were repeated in order to check their

reproducibility. It was found that the experimental error
was within G5%. Due to the high nitrogen ow rates
used in this study, it is dicult to completely recover all
the lower molecular weight material, and this results in

some loss in the mass balance. Mass balances of

90 G 5% were obtained for all experiments.
3.1. Degradation of PP over various catalysts
The reaction yield (based on the feed) of PP
degradation for each catalyst is summarised in Table 2.
The yield of volatile hydrocarbons for zeolite catalysts
(HZSM-5 O HUSY z HMOR) was higher than for nonzeolitic catalysts (SAHA z MCM-41) and the highest
yield (nearly 95 wt%) was obtained for HZSM-5. Overall,
the bulk of the products observed was in the gas phase
Table 2
Summary of the main products of PP degradation at reaction
temperature of 360
C over various catalysts (uidising N2 rate Z
570 mL min1, catalyst particle size Z 75e180 mm, and polymer to
catalyst ratio Z 40 wt%)
Degradation results

Catalyst type
HUSY HZSM-5 HMOR SAHA MCM-41

Yield (wt% feed)

Gaseous
89.49
Liquida
3.75
Residueb
6.76
Involatile residue 3.32
Coke
3.44
Mass balance (%)

93.71

94.77
2.31
3.92
2.27
1.25

88.29
4.54
7.17
4.96
3.21

86.44
4.73
9.83
7.49
2.34

86.19
5.07
8.74
6.83
1.91

95.32

93.24

89.68

88.46

22.54

25.47

63.65
0.25

60.56
0.16

Distribution of gaseous products (wt% feed)

59.86
Hydrocarbon gases 36.73 67.41
(SC1eC4)
27.95
Gasoline (SC5eC9) 51.83 25.54
BTXc
0.93
1.82
0.48
a
b
c

Condensate in condenser and captured in lter, unidentied.

Coke and involatile products.
Benzene, toluene and xylene.

104

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108

3.2. Inuence of operating conditions

on PP degradation
The inuence of operation conditions including
temperature (290e430
C), ow rates of uidising gas
(270e900 mL min1), and ratios of PP to catalyst feed
(0.1:1 to 1:1) has been investigated in this work. Some
similar trends in product yields were observed with
HUSY as the reaction temperature was increased. Gas
and coke yields increased and involatile residues
(unreacted or partially reacted PP) and liquids decreased

HUSY

HZSM-5

HUSY

HZSM-5

HMOR

SAHA

Yields (wt%)

20
15
10

MCM-41

12
10
8
6
4
2
0

C2

C3

i-C4

i-C5

i-C6

i-C7

C8

Carbon number
Fig. 3. Comparison of alkane products for the catalytic degradation of
PP at 360
C over various cracking catalysts (polymer to catalyst
ratio Z 40% wt/wt, rate of uidisation gas Z 570 mL min1).

(Table 3). The rate of hydrocarbon production as

a function of time for PP degradation over HUSY at
dierent reaction temperatures is compared in Fig. 4
and, as expected, faster rates were observed at higher
temperatures. At 430
C, the maximum rate of hydrocarbon production was 56 wt% min1 (calculated as
described in Section 2.2) after only 0.5 min with all the
polymer degraded after approximately 3 min. As the
temperature of reaction was decreased, the initial rate of
hydrocarbon production dropped and the time for PP to
be degraded lengthened. At 290
C the rate of hydrocarbon production was signicantly lower with the
polymer being degraded more slowly over 15 min. The
change in the hydrocarbon yield with temperature was
similar, for all catalysts used, with faster rates observed
at higher temperatures.
Table 3
Product distributions shown from HUSY catalysed pyrolysis of PP at
dierent reaction temperatures (uidising N2 rate Z 570 mL min1,
catalyst particle size Z 75e125 mm, and polymer to catalyst
ratio Z 40 wt%)

MCM-41

25

SAHA

14

Degradation results
30

HMOR

16

Yields (wt%)

with less than 5 wt% liquid collected. The highest level of

unconverted polymer was observed with SAHA and
MCM-41, while the highest coke yields were observed
with HUSY and HMOR. Some similarities were observed
between HZSM-5 and HMOR with C1eC4 and C5eC9
yields, which were approximately 60e67 wt% and
26e28 wt%, respectively. However, with HUSY
the C1eC4 and C5eC9 yields were approximately
30e37 wt% and 52e55 wt%, respectively.
The individual volatile hydrocarbon products of PP
degradation over various catalysts are demonstrated in
Figs. 2 and 3. Product distributions with HZSM-5
contained more olenic materials with about 60 wt% in
the range of C3eC5. The results of the products of PP
degradation reect the diering cracking eect of the
zeolite compared with the non-zeolite materials. Zeolites
(HUSY, HZSM-5 and HMOR) produced more paran
streams with large amounts of C4. Both SAHA and
MCM-41 resulted in a highly olenic product and gave
rise to the broadest carbon range of C3eC7. The
dierences in the product distributions between the
zeolites can be seen with HUSY producing a wider
molecular weight range than HZSM-5 and HMOR.
HMOR was particularly selective to C4 products
generating over 40 wt% with the highest yield of i-C4
(15 wt%).

Temperature (
C)
290

330

360

390

430

Yield (wt% feed)

Collection time (min)
Gaseous
Liquida
Residueb
Involatile residue
Coke

30
20
15
15
15
85.52 87.61 89.49 91.78 93.71
5.92 4.26 3.75 2.64 1.12
9.56 8.13 6.76 5.58 5.27
7.37 5.53 3.32 2.04 1.63
2.19 2.60 3.44 3.54 3.64

Mass balance (%)

89.69 89.31 93.71 92.62 93.83

5
0

C2=

C3=

C4=

C5=

C6=

C7=

C8=

Carbon number
Fig. 2. Comparison of olenic products for the catalytic degradation of
PP at 360
C over dierent catalysts (polymer to catalyst ratio Z
40% wt/wt, rate of uidisation gas Z 570 mL min1).

Distribution of gaseous products (wt% feed)

Hydrocarbon gases (SC1eC4) 34.31 35.98 36.73 37.24 37.88
51.08 51.22 51.83 53.56 54.71
Gasoline (SC5eC9)
BTXc
0.13 0.41 0.93 0.98 1.12
a
b
c

Condensate in condenser and captured in lter, unidentied.

Coke and involatile products.
Benzene, toluene and xylene.

105

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108

390C

Table 4
Product distributions shown from HUSY catalysed pyrolysis of PP at
dierent uidising N2 rates (reaction temperature Z 360
C, catalyst
particle size Z 75e125 mm, polymer to catalyst ratio Z 40 wt%, and
total time of collection Z 15 min)

430C

Degradation results

60
290C

Rgp (wt%/min)

50

330C
360C

40
30

900

20
10
0

10

12

14

16

Time (min)
Fig. 4. Comparison of hydrocarbon yields as a function of time at
dierent reaction temperatures for the degradation of PP over HUSY
(rate of uidisation gas Z 570 mL min1, catalyst particle size Z
75e125 mm, polymer to catalyst ratio Z 40 wt%).

The results shown in Fig. 5 illustrate that for ecient

polymer degradation good mixing is required, with
a dramatic drop-o in the rate of degradation observed
only at the lowest uidising ow used (270 mL min1).
Furthermore, changing the uidising ow rate inuences
the product distribution (Table 4). At low ow rates
(longer contact times), secondary products are observed
with increased amounts of coke precursors (BTX)
although the overall degradation rate is slower as shown
by increasing amounts of partially depolymerised
products. The amount of catalyst used in the degradation
of PP remained constant and, therefore, as more
polymers were added to the reactor then fewer catalytic
sites per unit weight of catalyst were available for
cracking. The overall eect of increasing the polymer to
catalyst ratio from 1:10 to 1:1 on the rate of hydrocarbon
generation was small but predictable (Fig. 6). As the
polymer to catalyst ratio increases, the possibility of PP
adhesion to the reactor wall increases as the amount of
unreacted polymer in the reactor rises. However, for the
work carried out in this paper no such problems were
observed. Also, the maximum rate observed dropped
slightly and the time taken to generate the maximum rate

570

420

270

89.77 87.58 89.49 88.12 86.53

4.19 4.53 3.75 4.49 5.25
6.04 7.89 6.76 7.39 8.22
3.25 4.75 3.32 4.23 4.91
2.79 3.14 3.44 3.16 3.31

Mass balance (%)

87.41 91.53 93.71 92.36 89.67

Distribution of gaseous products (wt% feed)

Hydrocarbon gases (SC1eC4) 37.21 36.43 36.73 34.64 34.86
Gasoline (SC5eC9)
51.93 50.57 51.83 51.83 50.71
0.63 0.58 0.93 1.65 1.96
BTXc
a
b
c

Condensate in condenser and captured in lter, unidentied.

Coke and involatile products.
Benzene, toluene and xylene.

extended from 1 min to 2 min. The total product yield

after 15 min (Table 5) showed only a slight downward
trend even after a 10-fold increase in added polymer.
This can be attributed to the high activity of HZSM-5
and excellent contact between PP and catalyst particles.
Consequently, as more polymers were added, lower
C1eC4 hydrocarbon gases and coke yields but higher
liquid yields and involatile products were observed. In
addition, more BTX (coke precursor) was produced but
increasing the polymer to catalyst ratio had only
virtually no eect on gasoline production.
3.3. Product stream variation with catalyst
deactivation and reaction conditions
The deactivation of the catalyst was examined by the
transient change in the amount of gaseous compounds
35

900 ml/min

35

25

420 ml/min

20

270 ml/min

Rgp (wt%/min)

570 ml/min

20% wt/wt

30

720 ml/min

30

15
10

40% wt/wt

25

60% wt/wt
80% wt/wt

20

100% wt/wt

15
10
5

5
0

720

Yield (wt% feed)

Gaseous
Liquida
Residueb
Involatile residue
Coke

40

Rgp (wt%/min)

Fluidising N2 rates (mL min1)

10

12

14

16

Time (min)
Fig. 5. Comparison of hydrocarbon yields as a function of time at
dierent uidising gas ows for the degradation of PP over HUSY
(reaction temperature Z 360
C, catalyst particle size Z 75e125 mm,
polymer to catalyst ratio Z 40 wt%).

10

12

14

16

Time (min)
Fig. 6. Comparison of hydrocarbon yields as a function of time at
dierent ratios of polymer to catalyst for the degradation of PP over
HUSY (reaction temperature Z 360
C, catalyst particle size Z
75e125 mm, rate of uidisation gas Z 570 mL min1).

106

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108

Table 5
Product distributions shown from HUSY catalysed pyrolysis of PP
at dierent ratios of polymer to catalyst (reaction temperature Z 360
C, catalyst particle size Z 75e125 mm, uidising N2
rate Z 570 mL min1, and total time of collection Z 15 min)
Degradation results

Ratio of polymer to catalyst (wt%)

10

20

60

80

100

Yield (wt% feed)

Gaseous
Liquida
Residueb
Involatile residue
Coke

91.51 91.27 89.22 88.41 86.39

1.97 2.54 3.63 3.88 4.98
6.52 6.29 7.15 7.71 8.63
2.31 2.45 3.72 5.15 6.68
4.21 3.84 3.43 2.56 1.95

Mass balance (%)

88.76 91.82 93.73 88.51 96.83

Distribution of gaseous products (wt% feed)

Hydrocarbon gases (SC1eC4) 38.69 37.57 35.87 34.62 33.79
Gasoline (SC5eC9)
52.39 53.06 52.38 52.56 53.23
0.43 0.64 0.97 1.23 1.37
BTXc
a
b
c

Condensate in condenser and captured in lter, unidentied.

Coke and involatile products.
Benzene, toluene and xylene.

produced. Deactivation of both HMOR and HUSY was

observed (Fig. 7) when the spot samples, taken during
the course of the reaction, were analysed. The deactivation is reected in the decrease of the amount of
isobutane (i-C4) and isopentane (i-C5) produced (product of bimolecular reaction) and the relative increase in
olens (product of monomolecular reaction), exempli-

]
ed by, C]
4 and C5 . HZSM-5 and non-zeolite catalysts
(SAHA and MCM-41) product streams remain virtually
unchanged throughout the degradation of PP. HZSM-5
is resistant to coking when coke builds up on outer
surface and the product stream remains essentially
unchanged, whereas the weakness and lower density of
the acid sites in SAHA and MCM-41 along with the
increased tolerance to coke in the larger mesopore
systems provide the most likely reasons for the lack of
variation in the product stream over these catalysts.
Equilibrium ratios of i-butane/n-butane and i-butene/
Sbutenes were predicted using Gibbs free energy
minimisation on the PRO/II ow sheets package for
the temperatures used experimentally and are presented
alongside the corresponding experimental results in
Tables 6 and 7. The i-butene/Sbutenes ratio is very
close to the predicted equilibrium values and thus the
reactions involved in the production and interconversion of butenes are very fast over zeolite catalysts, and
their ratio is primarily equilibrium controlled (Table 6).
The i-butane/n-butane ratio reects the involvement of
tertiary C4 carbenium ions in bimolecular hydrogen
transfer reactions and since tertiary carbenium ions are
more stable than primary ions, a higher yield of
isobutane would be expected. As can be seen in Table
6, the observed i-C4/n-C4 ratios at 360
C are well above
calculated equilibrium values and are consistent with the
steric constraints of their structures following the order:

HZSM5iC4 =nC4 Z4:72!HUSYiC4 =nC4 Z6:56

!HMORiC4 =n  C4 Z9:53
25

25

(a) isobutane

15
10
5
0

(b) isopentanes

20
Wt %

Wt %

20

15
10
5
0

0 1 2 3 4 5 6 7 8

0 1 2 3 4 5 6 7 8

35
30
25
20
15
10
5
0

Time (min)

(c) butenes
Wt %

Wt %

Time (min)

0 1 2 3 4 5 6 7 8

35
30
25
20
15
10
5
0

(d) pentenes

Table 6
Inuence of catalyst types on product selectivity for the degradation of
PP at reaction temperature of 360
C: experimental and predicted
equilibrium results
0 1 2 3 4 5 6 7 8

Time (min)
HZSM-5

HUSY

Both the carbon number distribution of the products of

PP cracking at 360
C over the various catalysts used in
this study and the nature of the product distribution
were found to vary with the catalyst used. The less acidic
amorphous SAHA and larger mesopores MCM-41
P
catalysts
resulted in a highly olenic product ( olen/
P
paran (o/p) Z 11.72 and 8.56) and the largest
amount of involatile residue (7.49 wt% and 6.83 wt%,
see Table 2). By comparison, the stronger acid sites of

Ratio

Time (min)
HMOR

SAHA

MCM-41

Fig. 7. Comparison of some of the main hydrocarbon products ((a)

isobutane; i-C4, (b) butenes; SC]
4 , (c) isopentane; i-C5 and (d)
pentenes; SC]
5 ) as a function of time for the degradation of PP over
dierent catalysts (reaction temperature Z 360
C, rate of uidisation
gas Z 570 mL min1, catalyst particle size Z 75e180 mm, polymer to
catalyst ratio Z 40 wt%).

Catalyst typea

VELb

HUSY HZSM-5 HMOR SAHA MCM-41

i-butene/Sbutenes 0.52
i-butane/n-butane 6.56
Solens/Sparansc 1.73
a

0.49
4.72
3.21

0.61
9.53
2.57

0.67 0.64
1.65 3.81
11.72 8.56

0.52
0.95
ed

Represents a series of runs where 40% wt/wt polymer to catalyst

feed and 570 mL min1 N2 uidising rate.
b
Predicted equilibrium data.
c
Denotes the ratio of the sum of all olenic to paranic products.
d
Data not available.

107

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108

Table 7
Inuence of operating conditions on product selectivity for the catalysed degradation of PP using HUSY: experimental and predicted equilibrium
results
Ratio

Reaction conditions
Reaction temperaturea (
C)

i-butane/n-butane
i-butane/n-butaned
i-butene/Sbutenes
i-butene/Sbutenesd
Solens/Sparanse
a
b
c
d
e

Fluidising N2 ratec (mL min1)

Polymer to catalyst
ratiob (wt%)

290

360

430

10

60

100

270

570

900

7.54
1.17
0.57
0.56
2.47

6.56
0.95
0.52
0.52
1.73

4.38
0.81
0.45
0.48
1.46

6.47

6.64

6.73

5.82

6.56

8.79

0.56

0.51

0.53

0.48

0.52

0.53

1.72

1.67

1.65

1.34

1.73

1.81

Represents a series of runs where 40% wt/wt polymer to catalyst feed and 570 mL min
Reaction temperature Z 360
C and uidising N2 Z 570 mL min1.
Polymer to catalyst ratio Z 40 wt% and reaction temperature Z 360
C.
Predicted equilibrium data.
Denotes the ratio of the sum of all olenic to paranic products.

the zeolite catalysts resulted in increased bimolecular

hydrogen transfer in the order:
HZSM5o=pZ3:21!HMORo=pZ2:57
!HUSYo=pZ1:73
Bimolecular reactions, such as hydrogen transfer, are
sterically hindered within the 10 ring channel system of
HZSM-5. However, the 12 ring channels of HMOR are
less restrictive; hence the increase in the paran
product. The combination of 12 ring pore openings
and large internal super-cages of HUSY allows significant bimolecular reactions and yielded a saturate-rich
product. As shown in Table 7, the selectivity could be
varied by changes in dierent operating conditions used
in this study. At fast ow rates (short contact times),
primary cracking products are favoured as evidenced by
the increasing ratios of i-butene/Sbutenes and olen/
paran. The yield of smaller cracked products increased
with temperature as did the yields of coke and BTX
(Table 3). Further evidence of the increase in secondary
reactions was seen in the lowering of the i-C4/n-C4 and
o/p ratios with temperature (Table 7).

4. Conclusions
A laboratory catalytic uidised-bed reactor has been
used to obtain a range of volatile hydrocarbons by
catalytic degradation of polypropylene in the temperature range 290e430
C. The uidised bed has a number
of special advantages for the pyrolysis of PP; it is
characterised by excellent heat and mass transfer, it has
a low tendency for clogging with molten polymer and
has the ability to maintain a nearly constant temperature
throughout the reactor. The catalytic degradation of PP
performed in uidised-bed reactor was demonstrated to

1

N2 uidising rate.

be a useful method for the production of potentially

valuable hydrocarbons.
Product streams varied markedly depending on
catalysts type and structure. The acidic zeolite catalysts,
HZSM-5, HMOR and HUSY, catalysed degradation
resulted in much more amounts of volatile hydrocarbons
compared with degradation over non-zeolite catalysts
(MCM-41 and SAHA). MCM-41 with large mesopores
and SAHA with weaker acid sites resulted in a highly
olenic product and gave a wide carbon number
distribution, whereas HUSY yielded a saturate-rich
product with a wide carbon number distribution and
substantial coke levels. Greater product selectivity was
observed with HZSM-5 and HMOR as catalysts with
about 60% of the product in the C3eC5 range and
HMOR generating the highest yield of i-C4 for all
catalysts studied. The larger pore zeolites (HUSY and
HMOR) showed deactivation in contrast to the more
restrictive HZSM-5. Observed dierences in product
yields and product distributions under identical reaction
conditions can be attributed to the microstructure of
catalysts.
The systematic experiments discussed in this paper
show that the use of various catalysts improve the yield of
volatile products and provide better selectivity in the
product distributions. The selectivity could be further
inuenced by changes in reaction temperature, ratios of PP
to catalyst feed and ow rates of uidising gas. Valuable
hydrocarbons of olens and iso-olens were produced by
low temperatures and short contact times used in this
study. It is concluded that under appropriate conditions
the resource potential of polymer waste can be recovered.

References
[1] Brandrup J, Bittner M, Michaeli W, Menges G. Recycling and
recovery of plastics. Munich, New York: Carl Hanser Verlag;
1996.

108

Y.-H. Lin, H.-Y. Yen / Polymer Degradation and Stability 89 (2005) 101e108

[2] Lee M. Feedstock recycling: new plastic for old. Chem Br

1995;31:515.
[3] Billingham NC. Polymers and the environment. London: Gerald
Scott, Royal Society of Chemistry; 1999.
[4] Hardman S, Leng SA, Wilson DC. BP Chemicals Limited,
Polymer Cracking. Eur Patent Appl 567292; 1993.
[5] Conesa JA, Font R, Marcilla A, Garcia AN. Pyrolysis of
polyethylene in a uidized reactor. Energy Fuels 1994;8:1238.
[6] Sodero SF, Berruti F, Behie LA. Ultrapyrolytic cracking of
polyethylene e a high yield recycling method. Chem Eng Sci
1996;51:2805.
[7] Audisio G, Bertini F, Beltrame PL, Carniti P. Catalytic
degradation of polyolens. Makromol Chem Macromol Symp
1992;257:191.
[8] Garforth AA, Lin YH, Sharratt PN, Dwyer J. Catalytic polymer
degradation for producing hydrocarbons over zeolites. Stud Surf
Sci Catal 1999;21:202.
[9] Aguado J, Serrano DP, Escola JM, Garagorri E, Fernandez JA.
Catalytic conversion polyolens into fuels over zeolite beta.
Polym Degrad Stab 2000;70(1):97.
[10] Sakata Y, Uddin MA, Koizumi K, Murata K. Catalytic
degradation of polypropylene into liquid hydrocarbons using
silicaealumina catalyst. Chem Lett 1996:245.
[11] Garforth AA, Fiddy S, Lin YH, Ghanbari-Siakhalu A,
Sharratt PN, Dwyer J. Catalytic degradation of high density
polyethylene: an evaluation of mesoporous and microporous
catalysts using thermal analysis. Thermochim Acta 1997;294:65.
[12] Lee KH, Noh NS, Shin DH, Seo YH. Comparison of plastic
types for catalytic degradation of waste plastics into product
with spent FCC catalyst. Polym Degrad Stab 2002;78(3):
539.
[13] Ishihara Y, Nanbu H, Saido K, Ikemura T, Takesue T. Back
biting reaction during the catalytic decomposition of polyethylene. Bull Chem Soc Jpn 1991;64:3585.
[14] Aguado J, Serrano DP, Romero MD, Escola JM. Catalytic
conversion of polyethylene into fuels over mesoporous MCM-41.
Chem Commun 1996:725.
[15] Dawood A, Miura K. Catalytic pyrolysis of g-irradiated polypropylene over HY-zeolite for enhancing the reactivity and the
product selectivity. Polym Degrad Stab 2002;76(1):45.

[16] Ohkita H, Nishiyama R, Tochihara Y, Mizushima T, Kakuta N,

Morioka Y, et al. Acid properties of silicaealumina catalysts and
catalytic degradation of polyethylene. Ind Eng Chem Res
1993;32:3112.
[17] Cardona SC, Corma A. Tertiary recycling of polypropylene by
catalytic cracking in a semibatch stirred reactor; use of spent
equilibrium FCC commercial catalyst. Appl Catal B 2000;25:151.
[18] Uemichi Y, Nakamura J, Itoh T, Sugioka, Garforth AA,
Dwyer J. Conversion of polyethylene into gasoline-range fuels
by two-stage catalytic degradation using silicaealumina and
HZSM-5 zeolite. Ind Eng Chem Res 1999;38:385.
[19] Lin YH, Sharratt PN, Garforth AA, Dwyer J. Catalytic
conversion of polyolens to chemicals and fuels over various
cracking catalysts. Energy Fuels 1998;12:767.
[20] Guohua L, Tomohiko S, Satomi Y, Kunio K. Catalytic
degradation of high density polyethylene and polypropylene into
liquid fuel in a powder-particle uidized bed. Polym Degrad Stab
2000;70(1):97.
[21] Mastellone ML, Perugini F, Ponte M, Arena U. Fluidized bed
pyrolysis of a recycled polyethylene. Polym Degrad Stab
2002;76(3):479.
[22] Lin YH, Yang MH, Yeh TF, Ger MD. Catalytic degradation of
high density polyethylene over mesoporous and microporous
catalysts in a uidised-bed reactor. Polym Degrad Stab
2004;86:121.
[23] Fortelny I, Michalkova D, Krulis Z. An ecient method of
material recycling of municipal plastic waste. Polym Degrad Stab
2004;85:975.
[24] Kaminsky W, Predel M, Sadiki A. Feedstock recycling of
polymers by pyrolysis in a uidised bed. Polym Degrad Stab
2004;85:1045.
[25] Beck JS, Calabro DC, Mccullen SB, Pelrine BP, Schmitt KD,
Vartuli JC. US Patent 5145816; 1992.
[26] Lin YH, Hwu WH, Ger MD, Yeh TF, Dwyer J. A combined
kinetic and mechanistic modeling of the catalytic of polymers.
J Mol Catal A: Chem 2001;171:143.
[27] Lin YH, Sharratt PN, Garforth AA, Dwyer J. Investigation of the
deactivation of ultrastable Y zeolite by coke formation during the
catalytic degradation of high density polyethylene. Thermochim
Acta 1997;294:45.