controlled pyrolysis of polyethylene/polypropylene/polystyrene mixed plastics with high impact...

Polymer Degradation and Stability 92 (2007) 211e221www.elsevier.com/locate/polydegstab

Controlled pyrolysis of polyethylene/polypropylene/polystyrenemixed plastics with high impact polystyrene containing flame retardant:

Effect of decabromo diphenylethane (DDE)

Thallada Bhaskar a,*, William J. Hall b, Nona Merry M. Mitan a, Akinori Muto a,Paul T. Williams b,**, Yusaku Sakata a,***

a Department of Applied Chemistry, Graduate School of Natural Science and Technology, Okayama University,3-1-1 Tsushima Naka, 700-8530 Okayama, Japan

b Energy and Resources Research Institute, University of Leeds, Leeds LS2 9JT, United Kingdom

Received 23 August 2006; received in revised form 19 October 2006; accepted 20 November 2006

Available online 17 January 2007

Abstract

The pyrolysis of polyethylene(PE)/polypropylene(PP)/polystyrene(PS) mixed with high impact polystyrene (HIPS-Br) containing decabromodiphenylethane (DDE) as a brominated flame retardant with antimony trioxide as a synergist was performed under controlled temperatureprogrammed pyrolysis (two steps) conditions to understand the decomposition behaviour and evolution of brominated hydrocarbons fromflame-retardant additives. The liquid products were extensively analyzed by gas chromatographs equipped with FID, ECD, MSD, TCD,AED and FT-IR. The solid residue samples were analyzed by powder X-ray diffraction and combustion followed by ion-chromatography.The controlled pyrolysis of PE/PP/PS/HIPS-Br significantly affected the decomposition behaviour of HIPS-Br and subsequently the formationof decomposition products. GC/ECD analysis confirmed that the brominated hydrocarbons were concentrated in step 1 liquid products leavingless brominated hydrocarbons in the step 2 liquid products, similar to the decabromo diphenyl ether flame retardant containing mixed plastics.The yield of liquid products in step 1 from 3P/DDE-Sb(5) was 5 wt% and from 3P/DDE-Sb(0) was 2.4 wt%. The presence of antimony in theDDE containing plastics affected the yield of liquid, gas and residue products. ECD analysis showed that the presence of antimony increased theBr containing hydrocarbons and step 1 has 3e4 times higher brominated compounds than step 2 hydrocarbons in both the samples.� 2006 Elsevier Ltd. All rights reserved.

Keywords: HIPS-Br; Flame retardant; Pyrolysis; Decabromo diphenylethane; Thermal decomposition; Waste plastics

1. Introduction

The production of electric and electronic equipment (EEE)is increasing worldwide. Due to the presence of hazardous ma-terial contents, WEEE may cause environmental problemsduring the waste management process at the end of life, if it

* Corresponding author. Tel./fax: þ81 86 251 8082.

** Corresponding author. Tel.: þ44 113 3432504; fax: þ44 113 2467310.

***Corresponding author. Tel./fax: þ81 86 251 8082.

E-mail addresses: [email protected] (T. Bhaskar), p.t.

[email protected] (P.T. Williams), [email protected] (Y.

Sakata).

0141-3910/$ - see front matter � 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.polymdegradstab.2006.11.011

is not properly pre-treated. Recycling of WEEE is an impor-tant subject not only from the point of waste treatment butalso from the aspect of recovering valuable materials such asthe plastic portion of WEEE. A significant and increasing pro-portion of plastic waste is being used in consumer productssuch as television casing materials, computer equipment, etc.The conversion of waste plastics into chemical feedstock orfuel represents a sustainable way for the recovery of the or-ganic content of the polymeric waste. Approximately 50%of waste WEEE plastics are high impact polystyrenes(HIPS), with the next largest fraction being acrylonitrileebutadieneestyrene (ABS) [1]. High impact polystyrene makesup a significant proportion of WEEE plastics and is often fire

mailto:[email protected]




http://www.elsevier.com/locate/polydegstab

212 T. Bhaskar et al. / Polymer Degradation and Stability 92 (2007) 211e221

retarded with decabromo diphenyl ether and diantimony triox-ide to decrease the flammability of the plastics. Deca-BDE isadded to HIPS because the bromine radicals quench combus-tion reactions while diantimony trioxide is added as a synergistas it increases the rate of bromine release by forming antimonybromides and antimony oxybromides [2]. The flame retardantcontent of WEEE, particularly the presence of bromine andantimony, represents a major hindrance to the developmentof recycling technologies. Decabromo diphenyl ether (DDO)and decabromo diphenylethane (DDE) are the two most com-mon flame retardants in the WEEE plastics.

There has been a plethora of research work on the recyclingof polymers containing brominated flame retardants and thedebromination of brominated hydrocarbons by using catalystsand/or sorbents [3e12]. Hall and Williams reported on the fastpyrolysis of halogenated plastics from waste computers andshowed that the pyrolysis products contained a chemically het-erogeneous mixture of hydrocarbons which would need to betreated before they were used in the hydrocarbon industry [3].Analytical pyrolysis of flame retarded polymers of electronicscrap and debromination was studied by Blazso et al. [4]. Itwas reported that brominated epoxy resins pyrolysed with so-dium hydroxide enhanced bromomethane evolution, whilea depressed brominated phenol formation was observed. Dueto the debromination of dibromo- and tribromophenyl groupsof brominated polystyrene, co-pyrolysis with sodium hydrox-ide and with basic zeolites resulted in a considerably reducedyield of dibromo- and tribromo-styrenes [4].

Hornung et al. reported that the optimal pyrolysis condi-tions were 350 �C with a residence time of 10e30 min forthe detoxification of brominated pyrolysis oil model com-pounds such as 2,6-dibromophenol and tetrabromobisphenol-A with polypropylene [5] and found that the polypropyleneacts as a reductive agent for the dehalogenation of brominatedorganic compounds [7]. Bhaskar et al. reported the pyrolysisof HIPS-Br mixed with PP/PE/PS/PVC and debrominationby carbon composite of calcium carbonate (CaeC) [10], aswell as the thermal decomposition of HIPS-Br containing var-ious brominated flame retardants and debromination by carboncomposite of iron oxide catalyst (FeeC) [11]. Antos and Sed-lar investigated the influence of flame retardant on thermal de-composition and found that brominated flame retardants mayundergo thermal decomposition at temperatures lower thanthose of polymer combustion [6]. The optimization of cata-lysts/sorbents for the effective removal of halogens especiallybrominated hydrocarbons is an intense area of research. In ourrecent report, we found that the controlled pyrolysis of PVC orPVDC containing PP/PE/PS mixed plastics produced the chlo-rine free hydrocarbons in the absence of catalyst/sorbent [12]and also found that the decabromo diphenyl ether (DDO) con-taining HIPS-Br plastics pyrolysis step 1 liquid products with2 times higher brominated hydrocarbons than step 2 liquidproducts [13]. In the present investigation, we report on the con-trolled (two step) pyrolysis of PE/PP/PS mixed with high impactpolystyrene containing decabromo diphenylethane (DDE)flame retardant and an extensive analysis of the pyrolysis prod-ucts. The liquid products were collected independently for the

two pyrolysis steps to better understand the pyrolysis behaviourof the flame retardant and subsequently the formation of bromi-nated compounds in the pyrolysis oils.

2. Experimental

2.1. Materials

High density polyethylene (PE; 2200 J) was obtained fromMitsui Chemical Co. Ltd., Japan; polypropylene (PP; J 105G)from Ube Chemical Industries Co. Ltd., Japan; polystyrene(PS; 666) from Asahi Kasei Industries Co., Ltd., Japan; high

Table 1

Bromine and Sb2O3 content in brominated flame retardant containing high

impact polystyrene (HIPS-Br) samples and their codes

Code Description Br

content

(wt%)

Sb2O3

content

(wt%)

3P/DDE-Sb(5) HIPS with decabromo

diphenylethane (DDE)

10.7 5

3P/DDE-Sb(0) HIPS with decabromo

diphenylethane (DDE)

10.7 0

3P: PE (3 g)þ PP (3 g)þ PS (3 g).

Fig. 1. (a) Schematic experimental setup for pyrolysis of 3P mixed with 3P/

DDE-Sb(5) and 3P/DDE-Sb(0) samples at 430 �C; (b) temperature profile

for the pyrolysis experiment.

213T. Bhaskar et al. / Polymer Degradation and Stability 92 (2007) 211e221

0

20

40

60

80

100

150 200 250 300 350 400 450 500 550 600

Temperature,°C

Wei

ght l

oss,

%

-2.E-02

-1.E-02

-8.E-03

-4.E-03

0.E+00

4.E-03

DTG

, mg/secDDE-Sb(5)

DDE-Sb(0)PSDDE-Sb(5)DDE-Sb(0)PS

Fig. 2. TG/DTA profiles of DDE-Sb(5), DDE-Sb(0) and PS.

impact polystyrene (HIPS) containing decabromo diphenyl-ethane (DDE) flame retardant with Sb2O3 (5 wt%) and withoutSb2O3 was commercially available. The composition of theflame retarded HIPS samples is shown in Table 1, as is the sym-bols used throughout this paper.

2.2. Thermogravimetry

TG/DTG experiments were carried out on a ShimadzuTGA-51 thermobalance. Samples of 6e10 mg were heatedin an alumina pan at a heating rate of 5 �C min�1, while thefurnace was flushed with 200 mL/min of nitrogen.

2.3. Pyrolysis procedure

Pyrolysis of the 3P plastic samples mixed with differentHIPS-Br samples [4 samples] [(3P is PE (3 g)þ PP(3 g)þ PS (3 g)] was performed in a Pyrex glass reactor(length: 35 cm; i.d. 3 cm) under atmospheric pressure by batchoperation using the experimental setup and conditions shownin Fig. 1(a) and (b). Briefly, 10 g of mixed plastics was loaded

into the reactor and the temperature program was as follows:step 1: ambient temperature to 330 �C at 5 �C min�1 andhold for 2 h at 330 �C with an N2 carrier gas flow of 55mL/min; step 2: 330e430 �C at 15 �C min�1 and hold at430 �C till the end of the experiment with an N2 carrier gasflow of 30 mL/min. The plastic bed temperature was measuredand it was taken as the temperature of decomposition. The va-pour products were condensed to liquid using a cold watercondenser and trapped in a two different measuring cylinders,one for each of the two different pyrolysis steps. The hydrogenbromide evolved during pyrolysis was trapped in a flask con-taining ion-exchanged water. The hydrocarbon gaseous prod-ucts were trapped in a Tedlar bag and analyzed by GC/TCDonce at the end of the pyrolysis experiment.

2.4. Analysis of pyrolysis products

The pyrolysis oils were characterised using gas chromato-graph (GC) equipped with flame ionization detector (GC/FID). The GC was a Varian 3380 fitted with a ZB-5MS column(30 m� 0.32 mm; i.d. 0.25 mm film thickness). The GC was

0

2

4

6

8

10

0 100 200 300 400 500 600 700 800Time, min.

Cum

ulat

ive

Volu

me,

mL

0

100

200

300

400

500

Temperature,°C

3P/DDE-Sb(5)3P/DDE-Sb(0)

Fig. 3. Cumulative volume of liquid products obtained during pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0).


programmed to hold at 40 �C for 15 min and then heat to 280 �Cat 5 �C min�1 and then hold for a further 15 min. The injectorand FID were heated to 300 �C. The results from the GC/FIDwere compared with those from a coupled gas chromatogra-phemass spectrometer that used the same type of column andGC program. The GC/MSD was a Shimadzu QP2010 fittedwith a 30 m RTX-5 column. The oils were all diluted in meth-anol and analyzed. The injector temperature was 285 �C and theoven was held at 40 �C for 15 min and then ramped to 280 �C at5 �C min�1 and then held for 15 min. The mass spectrometerelectron energy was 70 eV and the ion source and couplingtemperatures were 220 �C and 300 �C, respectively. The ionmass spectra derived were automatically compared to spectrallibraries. The MSD was not switched on until 5 minof therun, meaning that benzene and toluene could not be analyzed.

GC/ECD analysis of PBDEs in the pyrolysis oil was carriedout using a Varian 3380 GC fitted with a Varian CP-SIL 5CBcolumn (15 m� 0.25 mm; i.d.� 0.25 mm film thickness). TheGC was calibrated using a Cambridge Isotope Laboratories’predominant congener mixture. The GC oven was pro-grammed to hold at 100 �C for 1 min, ramped to 140 �C at2 �C min�1, then ramped to 220 �C at 4 �C min�1, thenramped to 280 �C at 8 �C min�1, hold for 36.5 min and the in-jector temperature was 300 �C and the ECD temperature waskept at 310 �C.

Table 2

Material balance from the pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0) sam-

ples and properties of liquid products

Sample Temperature

program

Yield of degradation

products, wt%

Liquid

properties

Liquid

(L)

Gas

(G)

Residue

(R)

Cnp Density,

g/mL

3P/DDE-Sb(5) Step 1 5 17.1 14.8 14.3 0.77

Step 2 63.1 12 0.81

3P/DDE-Sb(0) Step 1 2.4 15.3 4 13.7 0.60

Step 2 78.3 13.3 0.88

The pyrolysis oils derived from the controlled pyrolysis ofthe plastic samples were analyzed by Fourier transform infra-red spectrometer to determine the functional group composi-tion of the oils. A Nicolet 560 FT-IR with data processingand spectral library facilities was used to analyze the func-tional groups present in the oil. A small amount of pyrolysisoil was coated on a KBr disk and the spectrometer scannedthe sample from 450 to 4000 cm�1.

The quantitative analysis of the liquid products (for CeNPgrams) collected at the end of pyrolysis experiment was per-formed using a gas chromatograph equipped with flame ioni-zation detector (GC/FID; YANACO G6800). The liquidproducts were separated on a 100% methyl silicone column(50 m� 0.25 mm� 0.25 mm). The GC oven was programmedto hold at 40 �C for 15 min, then ramped to 280 �C at the rateof 5 �C min�1 and hold for 37 min. Thus, the quantity of hy-drocarbons and the carbon number distribution of the liquidproducts were determined by GC/FID.

The amount of gaseous inorganic Br trapped in the waterflask was determined by an ion chromatograph (DIONEX,DX-120). The quantitative estimation of Br in the residuewas carried out by burning a sample of the residue usinga combustion flask followed by ion chromatographic analysis.In brief, a small portion (8e15 mg) of the residue sample wascombusted with O2 in a Pyrex flask, containing a pair of elec-tric wires, a Pt sample pan held by the pan holder, and a Pt fil-ament for firing the sample. The combustion products wereabsorbed in about 40 mL water containing H2O2 and subse-quently analyzed by the ion chromatograph. Powder X-ray dif-fraction analysis of solid residue was performed using a X-raydiffractometer (RINT2500/RIGAKU).

3. Results and discussion

The controlled pyrolysis of PE/PP/PS (3P) mixed plasticswith flame retardant containing high impact polystyrene(HIPS) with DDO was performed using the experimental setup

0

10

20

30

40

5 7 9 11 13 15 17 19 21 23 25

3P/DDE-Sb(5) Step 13P/DDE-Sb(5) Step 23P/DDE-Sb(0) Step 13P/DDE-Sb(0) Step 2

g (C

n) x

100

/g (o

il), w

t %

Carbon Number

Fig. 4. CeNP gram of liquid products obtained during the pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0).


10

20

30

40

50

60

70

80

90

100

% T

rans

mitt

ance

1000200030004000Wavenumbers (cm-1)

3P/DDE-Sb(5) Step 1

10

20

30

40

50

60

70

80

90

100

% T

rans

mitt

ance

1000200030004000Wavenumbers (cm-1)

3P/DDE-Sb(5) step 2

10

20

30

40

50

60

70

80

90

100

% T

rans

mitt

ance

1000200030004000Wavenumbers (cm-1)

10

20

30

40

50

60

70

80

90

100

% T

rans

mitt

ance

1000200030004000Wavenumbers (cm-1)

3P/DDE-Sb(0) step 1

3P/DDE-Sb(0) step 2

(a) (c)

(b) (d)

Fig. 5. FT-IR analysis of liquid products from pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0). Typical FT-IR spectra of liquid product from: (a) step 1 of 3P/DDE-

Sb(5); (b) step 2 of 3P/DDE-Sb(5); (c) step 1 of 3P/DDE-Sb(0); (d) step 2 of 3P/DDE-Sb(0).

and temperature profile shown in Fig. 1(a) and (b). The tem-perature profiles were selected based on the thermal decompo-sition behaviour of flame-retardant additives in HIPS-Brsamples shown in Fig. 2. It is known from our earlier studies[2] that the thermal decomposition of HIPS and HIPS-DDOsamples affected by the presence of Sb2O3 synergist which al-ters the thermal decomposition of HIPS considerably by initi-ating the decomposition of brominated flame retardant and PS,as shown in Fig. 2. The presence of the Sb2O3 synergist lowersthe decomposition temperature by about 50e330 �C and theDTG curves split into two major peaks. It was established inour previous work [2] that the brominated additives andSb2O3 could be removed from HIPS by heating the plasticto 330 �C and then, once the brominated compounds andSb2O3 had been removed, the HIPS could be pyrolysed toform bromine-free gas, oil, and char. In this work, the bromi-nated HIPS was firstly heated to 330 �C for 2 h to remove thebrominated compounds and was then pyrolysed at 430 �C.

The pyrolysis products were classified as liquid, gas, andresidue, which were collected at the bottom of the reactor.The cumulative volume profiles of the liquid products ob-tained during the controlled pyrolysis of 3P/DDE samplesare presented in Fig. 3 and the mass balance of the pyrolysis

products and the properties of the pyrolysis oils are given inTable 2.

The formation of liquid products from the pyrolysis of 3P/DDE-Sb(5) was found earlier (20 min) than the pyrolysis of3P/DDO-Sb(0) as evidenced by the decomposition tempera-ture from Figs. 2 and 3. The yield of liquid products instep 1 from 3P/DDE-Sb(5) is 5 wt% and from 3P/DDE-Sb(0) is 2.4 wt%. The yield of liquid products in step 2from 3P/DDE-Sb(5) is 63 wt% and from 3P/DDE-Sb(0) is78 wt% (Table 2). The presence of synergist with the DDEflame retardant has significant effect on the liquid, gas andresidue, which is different than the DDO flame retardantmixed plastics [13]. The yield of liquid products from 3P/DDE-Sb(5) is 15 wt% less than 3P/DDE-Sb(0). The yieldof gaseous and residue products are higher from 3P/DDE-Sb(5) than 3P/DDE-Sb(0) pyrolysis. The major portion of an-timony was found in the step 1 liquid products as antimonytribromide (SbBr3) and the remaining in residue. The residuewas about 3 times higher in 3P/DDE-Sb(5) pyrolysis residuethan 3P/DDE-Sb(0). In both the cases, the density of the step1 liquid products is lower than the step 2 liquid products dueto the low molecular weight hydrocarbons which are formedduring low temperature pyrolysis (330 �C for 2 h).


3P/DDE-Sb(5) step 1

0

1000000

2000000

3000000

4000000

5000000

6000000

7000000

0 10 20 30 40 50 60 70 80Time (min)

Resp

on

se

1

2

3

4

5

6

3P/DDE-Sb(5) step 2

0

500000

1000000

1500000

2000000

2500000

0 10 20 30 40 50 60 70 80Time (min)

Resp

on

se

1

2 3

45 6 7

8

910

(a)

(b) (d)

(c)

3P/DDE-Sb(0) step 1

0

500000

1000000

1500000

2000000

2500000

3000000

3500000

4000000

4500000

5000000

0 10 20 30 40 50 60 70 80

Time (min)

Resp

on

se

1

23

4

5

6

7

8

3P/DDE-Sb(0) step 2

0

500000

1000000

1500000

2000000

2500000

3000000

0 10 20 30 40 50 60 70 80Time (min)

Resp

on

se

1

2

3 4 5

6

7 8

9

1011

Fig. 6. GC/MSD analysis of liquid products obtained from 3P/DDE-Sb(5) and 3P/DDE-Sb(0). Total ion chromatogram of liquid product from: (a) step 1 of 3P/

DDE-Sb(5); (b) step 2 of 3P/DDE-Sb(5); (c) step 1 of 3P/DDE-Sb(0); (d) step 2 of 3P/DDE-Sb(0).

The volatility distribution of the liquid products was ana-lyzed by using CeNP gram (C stands for carbon and NP fornormal paraffin) proposed by Murata et al. [14]. We can deter-mine the retention times of normal paraffins by using a PEplastic derived oil as an index material. The yield of hydrocar-bons was estimated by applying GC/FID, which gives a pro-portional signal to the weight fraction of carbon content ofthe compounds. NP gram can be drawn by plotting the weightfraction of hydrocarbons, which are located within the rangeof retention values of two successive normal paraffins, Cn�1

and Cn, against the carbon number, n. Since the carbon num-ber corresponds to the boiling point range of a certain group ofhydrocarbons, the NP gram actually represents a volatility

Table 3

GC/MSD analysis of liquid products of step 1 from 3P/DDE-Sb(5) pyrolysis:

qualitative and semi-quantitative analysis of major compounds

Peak

no.

RT

(min)

SI

(%)

CAS # Name Conc. (%)

1 9.2 98 100-41-4 Ethylbenzene 21.3

2 15.7 98 98-82-8 Cumene 9.5

3 41.7 96 1081-75-0 1,3-Diphenylpropane 16.5

4 46.1 94 605-02-7 1-Phenylnaphthalene 1.0

5 50.8 92 723-98-8 2,3-Dihydro-1H-

cyclopenta[l]phenanthrene

3.6

6 60.4 92 1165-53-3 1,2,4-Triphenylbenzene 1.0

distribution of the liquid product. Fig. 4 shows the carbonnumber distribution of the liquid hydrocarbon products (CeNP gram) obtained from the controlled pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0). The liquid products obtainedfrom the step 1 pyrolysis have relatively high concentration(boiling point range of n-C9 and n-C17) than the step 2 pyrol-ysis oils in both the samples. The liquid products were distrib-uted in the range of n-C5 to n-C25 with major peaks at n-C9,n-C14 and n-C17. It is noted that the retention times of aromatichydrocarbons are higher than that of normal paraffins, thusstyrenic derivatives appears at n-C9.

Table 4



Peak

no.

RT

(min)

SI

(%)

CAS # Name Conc.

(%)

1 7.8 94 19549-87-2 2,4-Dimethyl-1-heptene 2.9

2 9.2 98 100-41-4 Ethylbenzene 7.6

3 11.9 98 100-42-5 Styrene 10.4

4 15.7 98 98-82-8 Cumene 2.2

5 20.5 95 98-83-9 a-Methyl styrene 1.8

6 32.5 91 23609-46-3 1,2-Diethylcyclooctane 1.5

7 38.0 95 629-62-9 Pentadecane 1.6


9 44.8 92 593-45-3 Octadecane 1.3

10 50.8 91 723-98-8 2,3-dihydro-1H-


0.8


Fourier transform infrared (FT-IR) analysis of the pyrolysisoils (Fig. 5(a)e(d)) showed that there were bond vibrationsthat are typical of mono-substituted aromatic hydrocarbonsand aliphatic hydrocarbons. Both of the step 1 oils from 3P/DDE-Sb(5) (Fig. 5(a)) and 3P/DDE-Sb(0) (Fig. 5(c)) weresimilar to each other as were both the step 2 oils. TheFT-IR software identified the spectra as mono-substitutedaromatic hydrocarbons and aliphatic hydrocarbons for all ofthe oils. Irrespective of the plastic sample, whether it con-tained antimony or not, the step 1 oils have similar features(Fig. 5(a) and (c)). The three peaks at 3023, 3064, and3080 cm�1 are probably due to aromatic CeH bonds,although the peak at 3080 cm�1 could also be a methylenegroup in an olefin. The peak at 2962 cm�1 indicates the pres-ence of methyl groups and the peaks at 2925 and 2872 cm�1

indicate either methyl groups or possibly methylene groups.The band a 1450 cm�1 could also indicate the presence ofmethylene group in-phase deformation. Substituted benzenering absorption patterns are present between 1700 and2000 cm�1. The peaks at 1491 and 1605 cm�1 could beC]C stretches. The peak at 1372 cm�1 indicates a CeCH3

group (symmetrical deformation band), probably a singlemethyl group attached to a carbon atom due to the absenceof a peak at higher wavelengths that would indicate multiplemethyl groups on the same carbon atom. The peak at910 cm�1 is most likely caused by a CH]CH2 group. The

Table 5



Peak

no.

RT

(min)

SI CAS # Name Conc.

(%)

1 9.2 98 100-41-4 Ethylbenzene 14.4

2 11.9 98 100-42-5 Styrene 10.7

3 15.7 98 98-82-8 Cumene 6.0

18.5 90 103-65-1 Propylbenzene 0.2

4 20.5 95 98-83-9 a-Methyl styrene 5.4


6 44.9 92 7614-93-9 1,3-Diphenyl-1-butene 10.3

7 50.8 92 723-98-8 2,3-Dihydro-

1H-cyclopenta[l]phenanthrene

9.8


Table 6



Peak

no.

RT

(min)

SI

(%)

CAS # Name Conc.

(%)

1 9.2 98 100-41-4 Ethylbenzene 4.1

2 11.9 98 100-42-5 Styrene 12.1

3 20.5 95 98-83-9 a-Methyl styrene 1.6

4 32.5 90 23609-46-3 1,2-diethylcyclooctane 1.3

5 38.0 95 629-62-9 Pentadecane 1.3


8 48.4 95 35465-71-5 2-Phenylnaphthalene 1.4

9 50.8 92 723-98-8 2,3-Dihydro-1H-


4.2



peaks at 700 and 750 cm�1 are indicators of mono-substitutedbenzene rings, although other substituted benzenes cannot beruled out. The observations from the FT-IR analysis of thestep 2 oils from the controlled pyrolysis of 3P/DDE-Sb(5)(Fig. 5(b)) and 3P/DDE-Sb(0) (Fig. 5(d)) were as follows.The three peaks at 3031, 3064, and 3080 cm�1 are typicalof CeH groups in benzene rings. The three peaks at 2854,2925, and 2956 cm�1 are indicators of methyl groups (CeCH3) and the peaks at 2925 and 2956 cm�1 could also bemethylene group CeH stretches, although the CH2 deforma-tion band at 1460 cm�1 is absent. Benzene ring substitutionpatterns are present between 1700 and 2000 cm�1. The peaksat 1495 and 1602 cm�1 are probably C]C stretches. Thepeak at 1377 cm�1 indicates a CeCH3 symmetrical deforma-tion band, probably a single methyl group attached to a carbonatom due to the absence of a peak at higher wavelengths thatwould indicate multiple methyl groups on the same carbonatom. The peaks at 700 and 750 cm�1 indicate the presenceof mono-substituted benzene rings, although the complex na-ture of this region of the spectra would indicate the presenceof other substituted benzenes. In the step 1 oils, the ratio ofmethyl and methylene groups (w2900 cm�1) to aromaticCeH groups (w3050 cm�1) were much smaller than in thestep 2 oils.

The GC/MSD analysis of oils from the controlled pyroly-sis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0) samples were per-formed and the general observations are as follows. Overallthe step 1 oils contained much fewer components than thestep 2 oils and significant numbers of aliphatic hydrocarbonswere present in the step 2 oils, along with the usual aromaticcompounds. A greater percentage of the oils could be char-acterised using the GC/MSD for the antimony free plasticsthan for the antimony containing plastics, especially in thestep 1 oils.

The controlled pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0) and analysis of liquid hydrocarbons collected at differ-ent steps showed some interesting observations as revealedby GC/MSD. The GC/MSD total ion chromatograms for the3P/DDE-Sb(5) oils from step 1 and step 2 are presented inFig. 6(a) and (b) and their corresponding chemical compoundsare given in Tables 3 and 4. Styrene and a-methyl styrene werecompletely absent in the step 1 oil. The main components ofthe step 1 oil are ethylbenzene, cumene, and 1,3-diphenylpro-pane. The main components of the step 2 oil are ethylbenzene,styrene, and 1,3-diphenylpropane. The total ion chromato-grams for the 3P/DDE-Sb(0) oils from step 1 and step 2 arepresented in Fig. 6(c) and (d) and their corresponding majorchemical compounds are given in Tables 5 and 6. Styreneand a-methyl styrene were present in the 3P/DDE-Sb(0) step1 oil along with significant quantities of ethylbenzene, cu-mene, 1,3-diphenylpropane, and 1,3-diphenyl-1-butene. Themain components of the step 2 oils are ethylbenzene, styrene,and 1,3-diphenylpropane.

The major compounds from 3P/DDE-Sb(5) and 3P/DDE-Sb(0) from the step 1 and step 2 oils are compared inFig. 7(a) and (b), respectively. Flame-retardant additive con-taining antimony [(DDE-Sb(5)] samples’ pyrolysis oils


0

5

10

15

20

25

4,4,5-

Trimeth

yl-2-h

exen

e

Ethylbe

nzen

e

Styren

e

Cumen

e

α-Meh

tylsty

rene

α-Meh

tylsty

rene

3,3-D

imeth

ylocta

ne

1,2-D

iethy

lcyclo

octan

e

1-Pen

tadec

ene

Tetrad

ecan

e

Pentad

ecan

e

Hexad

ecan

e

1,3-D

iphen

ylprop

ane

(3-Phe

nylbu

tyl)be

nzen

e

Heptad

ecan

e

1,3-D

iphen

yl-1-b

utene

Octade

cane

2-Phe

nylna

phtha

lene

2,3-D

ihydro

-1H-cy

clope

nta

[l]phe

nanth

rene

1,2,4-

Triphe

nylbe

nzen

e

Co

ncen

tratio

n,%

3P/DDE-Sb(5) Step 13P/DDE-Sb(5) Step 2

(a)

0

5

10

15

20

25

4,4,5-

trimeth

yl-2-h

exen

e

2,4-D

imeth

yl-1-h

epten

e

Ethylbe

nzen

e

Styren

e

Cumen

e

1,2-di

ethylc

ycloo

ctane

Tridec

ane

1-Pen

tadec

ene

Pentad

ecan

e

Hexad

ecan

e

1,3-D

iphen

ylprop

ane

(3-Phe

nylbu

tyl)be

nzen

e

1-Non

adec

ene

Heptad

ecan

e

1-Phe

nylte

tralin

Octade

cane

1,3-D

iphen

yl-1-b

utene

Heneic

osan

e

2-Phe

nylna

phtha

lene

9-Phe

nyl-5

H-benz

ocyc

lohep

tene

2,3-D

ihydro

-1H-cy

clope

nta[l]p

hena

ntren

e

1,3-D

iphen

ylben

zene

1,2,4-

Triphe

nylbe

nzen

e

Co

ncen

tratio

n,%

3P/DDE-Sb(0) Step 13P/DDE-Sb(0) Step 2

(b)

Fig. 7. Comparison of major compounds observed in the liquid products from (a) 3P/DDE-Sb(5) and (b) 3P/DDE-Sb(0).

showing the major compounds as ethylbenzene, cumene and1,3-diphenylpropane were observed in the step 1 oils and nostyrene and a-methyl styrene was observed in step 1. Severalhydrocarbons observed in step 2 were not observed in step 1.The similar observation was found with the DDE-Sb(0) andthe main difference was that the styrene was also foundin the step 1 oils. It is clear that it is possible to control thedecomposition of main polymer matrix from flame-retardantadditives and to avoid the formation of bromine rich hydrocar-bons in the liquefaction process.

Highly brominated organic compounds are still among themost widely used flame retardants for plastics although therehas been recently some concern about them, namely in con-nection with health and environmental issues. The functionof brominated flame retardants is based upon hydrobromicacid (HBr) evolution through the thermal decomposition ofthe retardant during the course of combustion (often atmore than 1000 �C) [15]. It follows from the bond values[16] that the CeBr bond is often the weakest one in the

flame-retardant molecule and its splitting thus representsthe primary reaction step. The mechanism of thermal decom-position of brominated flame retardants will depend on theirstructure, namely the availability of hydrogen. Thus, aliphaticbromo-compounds undergo thermally initiated dehydro-bro-mination involving intramolecular hydrogen abstraction[17], which gives rise to HBr and unsaturated lower bromi-nated derivatives. Aromatic brominated compounds undergoradical debromination, in the presence of hydrogen donors(polymer, impurities, etc.), the bromine radical releasedfrom the aromatic bromine compound forms HBr. The debro-mination of brominated aromatics leads to the formation oflower brominated congeners which are derived from the par-ent structure [18].

The most common brominated aromatic flame retardantsare polybrominated diphenyl ethers (PBDEs) and the flameretardant in the plastic used in this work was decabromodiphenyl ether; it can therefore be expected that a large vari-ety of PBDE congeners will be present in the pyrolysis oils.


GC/ECD was used to detect the presence of any PBDEs in thepyrolysis oils, ECDs only respond to halogenated compoundsand some forms of oxygen and nitrogen. The GC was cali-brated using a Cambridge Isotopes Laboratories’ (USA) pre-dominant congener mixture that contained the congenerslisted in Table 7. GC/ECD chromatograms for the 3P/DDE-Sb(5) oils are given in Fig. 8(a) and (b). The GC/ECD anal-ysis of the pyrolysis oils from step 1 and step 2 of 3P/DDE-Sb(5) shows that many more peaks were present inthe GC/ECD trace of the step 1 oil than the step 2 oil (202peaks for step 1 and 58 for step 2). Heavier compounds, orcompounds with longer retention times, were much moreprevalent in the step 1 than in the step 2 oil.

A list of PBDEs that were either possibly present in the oilor defiantly missing from the oil is shown in Table 7 andcompared with the 3P/DDE-Sb(5) and 3P/DDE-Sb(0) oils.More PBDEs were missing from the 3P/DDE-Sb(0) oils

Table 7

GC/ECD analysis of liquid products for the polybrominated diphenyl ether

(PBDE) compounds from the controlled pyrolysis of 3P/DDE-Sb(5) and 3P/

DDE-Sb(0) samples

3P/DDE-Sb(5) 3P/DDE-Sb(0) RT (min)

Step 1 Step 2 Step 1 Step 2

PBDE #1 M M M PP 12.5

PBDE #2 M M M M 13.2

PBDE #3 M M M M 13.8

PBDE #10 M M M M 20.9

PBDE #7 PP PP M M 23.4

PBDE #8/11 PP M M M 24.3

PBDE #12 PP M PP M 24.4

PBDE #13 PP M M M 25

PBDE #15 PP M PP PP 25.7



PBDE #17 PP M M M 31.3

PBDE #25/33 PP M PP M 31.5



PBDE #37 PP PP PP PP 33.4

PBDE #75 PP PP PP M 36.5








PBDE #99 PP PP M M 42.6




PBDE #126/15 PP M PP M 44.2



PBDE #138/16 PP PP PP M 46.8



PBDE #190 M M PP M 49.6

PBDE #209 M M M M 71.1

PP¼ possibly present, M¼missing.

than the 3P/DDE-Sb(5) oils. ECD chromatograms for the3P/DDE-Sb(0) oils are given in Fig. 8(c) and (d). The PBDEsdetected by the ECD can only be said to be possibly presentin the oils because no clean-up was performed on the oil andno MSD was used to confirm the identification.

The GC/AED analysis confirmed the presence of antimonybromide in the step 1 oil and the absence of antimony bromidein step 2 oils indicating that the major quantity of antimonyalong with bromine was carried in the step 1 in the case of syn-ergist containing sample. The synergist antimony showed thesynergistic effect for the removal of bromine from the sampleat low pyrolysis temperatures. In the absence of synergist i.e.,DDE-Sb(0), the bromine removal effect was less as observedfrom the ECD analysis.

The pyrolysis oils from step 1 and step 2 of the 3P/DDE-Sb(0) plastic shows that these oils were noticeably differentto the 3P/DDE-Sb(5) oils. More peaks were present in thestep 1 oil than the step 2 oil (166 peaks for step 1 and 64peaks for step 2) but both oils contained fewer peaks thantheir equivalent 3P/DDE-Sb(5) oil. The responses of thepeaks in the 3P/DDE-Sb(0) oils was generally lower thanthe equivalent 3P/DDE-Sb(5) oil, but this could be due tothe response factors of the compounds present in the oil.Heavier compounds appeared to be present in the step 1oil than in the step 2 oil but some light compounds(w10 min) were present in the step 2 oil but were absentin the step 1 oil.

The bromine content in the water trap was analyzed by ion-chromatography. 3P/DDE-Sb(5) step 1 water trap showed thepresence of 2.4 mg of bromine and step 2 contained 0.4 mg.3P/DDE-Sb(0) step 1 water trap contains 3 mg and step 2 con-tained 42 mg of bromine. This clearly shows us that the pres-ence of synergist antimony carried the major portion ofbromine during the step 1 leaving very small portion of Brin the pyrolysis mixture and absence of synergist in the sam-ple could show very small amount of Br in step 1 water trapand 10 times higher HBr in step 2 water trap. The brominecontent in 3P/DDE-Sb(5) residue samples was 17 mg and3P/DDE-Sb(0) residue sample contains 7 mg. This showsthat the type of flame retardant has changed the bromine dis-tribution in the residue sample, which is in contrast with thepresence of decabromo diphenyl ether flame retardant. GC/TCD analysis (Fig. 9) of gaseous products collected once atthe end of the controlled pyrolysis study shows that the step1 gaseous products contain C1, C2 and C4 (iso-butene), C3eC5 hydrocarbons were absent in step 1 products from boththe samples and only iso-butene was observed in 3P/DDE-Sb(5) step 1 gaseous products. Powder X-ray diffraction anal-ysis of the solid residue from the 3P/DDE-Sb(5) confirmed thepresence of antimony bromide and antimony in the solidresidue.

4. Conclusions

The controlled pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0) showed that there are significant differences in the yields


3P/DDE-Sb(5) step 1

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0 10 20 30 40 50 60 70 80Time (min)

(a) (c)

(d)(b)

3P/DDE-Sb(5) step 2

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0 10 20 30 40 50 60 70 80

Time (min)

3P/DDE-Sb(0) step 1

0

0.002

0.004

0.006

0.008

0.01

0.012

0.014

0.016

0.018

0.02

0 10 20 30 40 50 60 70 80Time (min)

Time (min)

Resp

on

se

3P/DDE-Sb(0) step 2

0

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

0 10 20 30 40 50 60 70 80

Resp

on

se

Resp

on

se

Resp

on

se

Fig. 8. GC/ECD analysis of liquid products obtained from 3P/DDE-Sb(5) and 3P/DDE-Sb(0). ECD chromatogram of liquid product from: (a) step 1 of 3P/DDE-

Sb(5); (b) step 2 of 3P/DDE-Sb(5); (c) step 1 of 3P/DDE-Sb(0); (d) step 2 of 3P/DDE-Sb(0).

and composition of pyrolysis oils from two different steps. The3P/DDE-Sb(5) oils show the absence of styrene and a-methylstyrene and they are observed in 3P/DDE-Sb(0) oils. ECDanalysis clearly showed that the major quantity of brominatedhydrocarbons was observed in step 1 oils rather than the step 2oils. The major portion of antimony was found in the form of

antimony tribromide in the step 1 oil and partly in residue.There is significant effect on the yield of pyrolysis productssuch as liquid, and residue. By controlling the pyrolysis pa-rameters, the formation of brominated hydrocarbons can beminimized to decrease the load on the debromination cata-lyst/sorbent.

0

20

40

60

80

C2H4 C2H6 C3H6 C3H8 iso-C4H8 iso-C4H10 n-C4H10 n-C5H10

Gas

Wei

ght,

%

3P/DDE-Sb(5) Step 13P/DDE-Sb(5) Step 23P/DDE-Sb(0) Step 13P/DDE-Sb(0) Step 2

Fig. 9. GC/TCD analysis of gaseous products from pyrolysis of 3P/DDE-Sb(5) and 3P/DDE-Sb(0).


Acknowledgements

One of the authors (YSS) thanks Science and Innovation,British Embassy, Tokyo for the Award of Global OpportunitiesFund under the UKeJapan Collaboration awards e Green andSustainable Chemistry and providing the travel grants to YSSand TBR to visit the University of Leeds, United Kingdom toinitiate the collaborative research program. WJH and PTWwould like to thank EPRSC for funding through researchGrant GR/S56801/01.

References

[1] Vehlow J, Bergfeldt B, Hunsinger H, Jay K, Mark FE, Tange L, et al.

Recycling of bromine from plastics containing brominated flame retar-

dants in state-of-the art combustion facilities. Brussels, Belgium:

BSEF; 2003.

[2] Jakab E, Uddin MA, Bhaskar T, Sakata Y. J Anal Appl Pyrolysis

2003;68/69:83.

[3] Hall WJ, Williams Paul T. Energy Fuels 2006;20:1536.

[4] Blazso M, Czegeny Zs, Csoma Cs. J Anal Appl Pyrolysis 2002;64:249.

[5] Hornung A, Balabanovich AI, Donner S, Seifert H. J Anal Appl Pyrolysis

2003;70:723.

[6] Antos K, Sedlar J. Polym Degrad Stab 2005;90:188.

[7] Hornung A, Donner S, Balabanovich A, Seifert H. J Clean Prod

2005;13:525.

[8] Mackenzie K, Kopinke FD. Chemosphere 1996;33:2423.

[9] More LS, Tremp J, Gloor R, Huber Y, Stengele M, Zennegg M. Environ

Sci Technol 2005;39:8691.

[10] Bhaskar T, Matsui T, Kaneko J, Uddin Md Azhar, Muto A, Sakata Y.

Green Chem 2002;4:372.

[11] Bhaskar T, Matsui T, Uddin Md Azhar, Kaneko J, Muto A, Sakata Y.

Appl Catal B Environ 2003;43:229.

[12] Bhaskar T, Negoro R, Muto A, Sakata Y. Green Chem 2006;8:697.

[13] Bhaskar T, Hall WJ, Mitan NMM, Muto A, Williams PT, Sakata Y,

unpublished results.

[14] Murata K, Hirano Y, Sakata Y, Uddin Md Azhar. J Anal Appl Pyrolysis

2002;65:71.

[15] Troitzsch J. The burning of plastics: international plastic flammability

handbook. 2nd ed. Hanser Publishers; 1990.

[16] Murov SL, Carmichael I, Huy G. Handbook of photochemistry. 2nd ed.

Marcel Dekker; 1993 [revised and expanded].

[17] Barontini F, Cozzani V, Cuzzola A, Petarca L. Rapid Commun Mass

Spectrom 2001;15:690e8.

[18] Weber R, Kuch B. Environ Int 2003;29:699e710.

controlled pyrolysis of polyethylene/polypropylene/polystyrene mixed plastics with high impact...

Documents