Analysis and structure prediction of chlorinated polycyclic aromatic hydrocarbons released from combustion of polyvinylchloride

Article abstract of DOI:10.1016/S0045-6535(03)00507-1

Chlorinated polycyclic aromatic hydrocarbons (Cl-PAHs) released from combustion of polyvinylchloride (PVC) at different furnace temperatures were investigated. A laboratory-scale tube-type furnace with electric heating was utilized to control combustion conditions. Glass fabric filters and adsorbents were used to collect the combustion emissions. Following Soxhlet extraction, concentration and column chromatography purification, isomers separation, selective detection and identification of Cl-PAHs were performed on GC/MS system on the basis of retention data and mass spectra. Their quantification was accomplished by using external standard calibration technique. About 18 Cl-PAHs were determined, most of which were monochlorinated derivatives of naphthalene, biphenyl, fluorene, phenanthrene, anthracene, fluoranthene and pyrene. Only two dichlorophenanthrenes or anthracenes were identified. The possible positions of chlorine atoms attached to the aromatic rings are predicted by quantitative structure-property relationship. The levels of these compounds were in the range of 0.30-29.08 μg/g PVC. The relationship between the formation of Cl-PAHs and PAHs was discussed.

Full text of DOI:10.1016/S0045-6535(03)00507-1

Chemosphere 53 (2003) 495–503
www.elsevier.com/locate/chemosphere
Analysis and structure prediction of chlorinated
polycyclic aromatic hydrocarbons released from
combustion of polyvinylchloride
1
Dongli Wang , Xiaobai Xu , Shaogang Chu, Daren Zhang
*
Research Center for Eco-Environmental Sciences, Chinese Academy of Sciences, P.O. Box 2871, Beijing 100085, PR China
Received 20 November 2002; received in revised form 23 April 2003; accepted 29 April 2003
Abstract
Chlorinated polycyclic aromatic hydrocarbons (Cl-PAHs) released from combustion of polyvinylchloride (PVC) at
different furnace temperatures were investigated. A laboratory-scale tube-type furnace with electric heating was utilized
to control combustion conditions. Glass fabric filters and adsorbents were used to collect the combustion emissions.
Following Soxhlet extraction, concentration and column chromatography purification, isomers separation, selective
detection and identification of Cl-PAHs were performed on GC/MS system on the basis of retention data and mass
spectra. Their quantification was accomplished by using external standard calibration technique. About 18 Cl-PAHs
were determined, most of which were monochlorinated derivatives of naphthalene, biphenyl, fluorene, phenanthrene,
anthracene, fluoranthene and pyrene. Only two dichlorophenanthrenes or anthracenes were identified. The possible
positions of chlorine atoms attached to the aromatic rings are predicted by quantitative structure–property relationship.
The levels of these compounds were in the range of 0.30–29.08 lg/g PVC. The relationship between the formation of
Cl-PAHs and PAHs was discussed.
Ó 2003 Elsevier Ltd. All rights reserved.
Keywords: Polyvinylchloride; Combustion; Cl-PAHs; Structure prediction
1
. Introduction
thalenes (PCNs) and other halogenated polycyclic
aromatic hydrocarbons (Cl-PAHs and Br-PAHs) with
three or more fused aromatic rings have been detected in
samples of various origins. Shiraishi et al. (1985) dem-
onstrated their presence in chlorinated tap waters.
Rappe et al. (1982) reported the formation of poly-
chlorinated pyrenes during PCB fires. Occurrence of a
large number of Cl-PAHs was reported in automobile
exhausts, snow, urban air and wastewater of pulp and
During recent decades, a lot of interest has been
focused on the detection of chlorinated polycyclic aro-
matic hydrocarbons (Cl-PAHs) in environmental sam-
ples. Apart from polychlorinated dibenzo-p-dioxins and
polychlorinated dibenzofurans (PCDD/Fs) and poly-
chlorinated biphenyls (PCBs), polychlorinated naph-
€
paper mills (Haglund et al., 1987; Nilsson and Ostman,
*
1993; Brodskv et al., 1999). Tausch and Stehlik (1985)
detected Cl-PAHs, ranging from four to seven aromatic
rings, in fly ash from an incineration plant for radioac-
tive wastes. Polychlorinated PAHs were detected in
emissions from coal combustion and municipal waste
incineration by Eklund and Str o€ mberg (1983) and by
Corresponding author. Tel.: +86-10-629-19-177; fax: +86-
0-629-23-563.
1
E-mail address: xuxb@public.bta.net.cn (X. Xu).
Current address: Beijing Center for Disease Prevention and
1
Control, 16 Hepingli Middle Street, Dongcheng District,
Beijing 100013, PR China.
0
045-6535/$ - see front matter Ó 2003 Elsevier Ltd. All rights reserved.
doi:10.1016/S0045-6535(03)00507-1

496
D. Wang et al. / Chemosphere 53 (2003) 495–503
Oehme et al. (1987). According to Sovocool et al. (1989),
chlorobromo-substituted PAHs were also shown to oc-
cur in fly ash from a municipal waste incinerator. Re-
cently more detailed data on congeners of PCNs that
accumulated in wildlife and abiotic environmental ma-
trices become available (Falandysz, 1998).
hydrocarbon hydroxylase (AHH) inducer (Franzen
et al., 1988). Some of the 75 possible PCN congeners
exhibit toxic effects similar to those of 2,3,7,8-TCDD
(Table 1).
The chlorine in the emissions from municipal waste
incinerators most likely originates from chlorine-
containing waste materials like PVC. It has been con-
firmed that PVC can lead to the formation of PCDD/Fs
during incineration of wastes (Elomaa et al., 1997). The
formation of polychlorinated benzene has also been
determined during combustion of PVC (Ahling et al.,
1978). However, to the best of our knowledge, no ex-
tensive analysis and identification of Cl-PAHs released
from PVC combustion has been reported previously
except for PCDD/Fs. The aim of the present work was
mainly to determine Cl-PAHs in emissions from PVC
combustion and to investigate their formation mecha-
nism in the combustion process, as well as the relation-
ship between their formation and the furnace operation
Interest has been focused on Cl-PAHs since some of
them have also been shown to have stronger toxicity and
mutagenic properties than their corresponding PAHs
in test systems for biological activity (Table 1). For
instance, it was found by Colmsj o€ et al. (1984, 1988)
and Rannug et al. (1986) that some Cl-PAHs showed
significant mutagenic effects in the presence of a
metabolizing system. L o€ froth et al. (1985) found that
7
-chlorobenzo[a]anthracene, 9-chloroanthracene, and 6-
chlorochrysene showed strong direct mutagenic effects.
The latter compound has also been shown to have a high
affinity to the tetrachlorodibenzo-p-dioxin (TCDD)
receptor (T o€ ftg ꢀa rd et al., 1985) and to be a potent aryl
Table 1
Toxicity data of Cl-PAHs cited from previous publications (Registry of Toxic Effects of Chemical Substances, 1998; Kannan et al.,
998)
1
Compounds
Toxicity
3
3
2
-Chloronaphthalene
TWA 0.2 mg/m ; STEL 0.6 mg/m ; LD50 ¼ 2078 mg/kg (mouse);
LD50 ¼ 886 mg/kg (rat)
3
3
1-Chloronaphthalene
TWA 0.2 mg/m ; STEL 0.6 mg/m ;
LD50 ¼ 1540 mg/kg (mouse);
LD50 ¼ 1091 mg/kg (rat)
3
Trichloronaphthalene
Tetrachloronaphthalene
TWA 5 mg/m
TWA 2 mg/m
3
2,3,6,7-Tetrachloronaphthalene
Oral administration guinea pig
LD50 > 3 mg/kg
1
,2,6,8-Tetrachloronaphthalene
TEF 1.65E)05
TWA 0.5 mg/m
3
Pentachloronaphthalene
,2,3,6,7-Pentachloronaphthalene
Hexachloronaphthalene
1
TEF 1.7E)04
TWA 0.2 mg/m
3
1
1
1
1
1
1
,2,3,4,5,6-Hexachloronaphthalene
,2,3,5,6,8-Hexachloronaphthalene
,2,3,4,6,7-/1,2,3,5,6,7-Hexachloronaphthalene
,2,3,5,7,8-Hexachloronaphthalene
,2,3,6,7,8-Hexachloronaphthalene
,2,3,4,5,6,7-Heptachloronaphthalene
TEF 2E)03
TEF 1.5E)04
TEF 2.27E)03
TEF 2E)03
TEF 5.9E)04
TEF 3.45E)03
TWA 0.1 mg/m ; STEL 0.3 mg/m
3
3
Octachloronaphthalene
9-Chlorofluorene
Intravenous injection mouse
LD50 ¼ 56 mg/kg
2
9
9
6
1
1
1
1
6
-Chloroanthracene
-Chlorophenanthrene
-Chloroanthracene
-Chlorochrysene
SAT 100 lg/plate, 0.4702 lmol/plate
SAT 10 lg/plate, 0.0470 lmol/plate
SAT 20 lg/plate, 0.0940 lmol/plate
SAT 20 lg/plate, 0.07612 lmol/plate
SAT 10 lmol/plate
-Chloropyrene
,3-Dichloropyrene
,6-Dichloropyrene
,8-Dichloropyrene
-Chlorobenzo[a]pyrene
SAT 3330 lmol/plate
SAT 3330 lmol/plate
SAT 3330 lmol/plate
0
Tumority-mouse TDL 80 mg/kg/8D-I
Note: TWA––time weighted average; STEL––short term expose limit; LD50––lethal dose 50 percent kill; TEF––toxic equivalent factor;
SAT––testing for mutagenicity on Salmonella typhimurium TA98 and TA100; TDLo––lowest published toxic concentration.

D. Wang et al. / Chemosphere 53 (2003) 495–503
parameters. The possible positions of chlorine atoms 2.3. Extraction and purification
497
attached to some of Cl-PAHs are predicted by quanti-
tative structure–property relationship (QSPR).
After cooling, the sample boat, glass wool, glass fiber
filter and XAD-2 adsorbents were collected and ex-
tracted with dichloromethane in Soxhlet apparatus for
1
8 h. The extracts were concentrated to a volume of 4 ml
2
. Experimental method
by K–D apparatus and a gentle stream of nitrogen. 0.5
ml of extract was cleaned up by a glass column (10 mm
i.d.) filled with 10 g of silica gel (deactivated by 3%
2
.1. Materials
H O), which was eluted with 80 ml of n-hexane, and 50
2
PVC, in powder form, free of additives, was the
ml of a mixture of n-hexane and dichloromethane (6:4,
v/v). The first fraction, which contained Cl-PAHs, was
concentrated to 1 ml for GC/MS analysis.
product of Beijing Second Chemical Factory. 1-Chloro-
naphthalene, 2-chloronaphthalene, 9-chlorophenanth-
rene, 9-chloroanthracene and PCB14 were purchased
from Fluka (Sweden), AccuStandard Inc., Acros, Ald-
rich Chem. Co. (Germany) and Chem Service, respec-
tively. Standard solutions of Cl-PAHs were prepared
using n-hexane as the solvent. The silica gel in particle
size from 100 to 200 mesh was activated in 130 °C for
2
.4. Gas chromatography and mass spectrometry
Analysis of Cl-PAHs was performed on a HP 6890
GC/5973 MSD system fitted with a 30 m HP-5ms col-
umn (0.25 mm i.d., 0.25 lm film). The oven temperature
of the gas chromatograph was programmed from 50 °C
1
3 h. Subsequent deactivation was performed by adding
3% (w/w) distilled water. Analytical grade solvents were
used after distillation in an all-glass apparatus.
(
2 min) to 290 °C at a rate of 4 °C/min, and the tem-
perature held on for 10 min. The temperature of the
injector was maintained at 290 °C. The temperature of
ion source of mass spectrometry was 280 °C, ionizing
voltage 70 eV, and scan range m=z 50–550 amu.
2
.2. Combustion procedure
The combustion tests were performed in a tube-type
furnace shown in Fig. 1 as described previously (Piao
et al., 1999). At the beginning of a test, the tube furnace
was first heated up to the test temperature (600–900 °C),
then air was introduced into the quartz tube at a 2 l/min
flow rate. The quartz boat containing about 1 g of PVC
was pushed into the combustion zone of the furnace.
The emissions were collected with glass wool, glass fiber
filter (pore diameter 0.2 lm), and an adsorption car-
tridge filled with 7 g of XAD-2.
Extraction ion chromatograph peaks of Cl-PAHs
þ
were determined by selecting molecular ion (M ), its
þ
þ
isotopic ion (M + 2) and fragment ion (M –Cl). Cl-
PAHs were identified by comparing their GC retention
data and mass spectra with those of authentic com-
pounds. In cases when the authentic standards were not
available, identification was performed by checking re-
ported relative retention data (Shiraishi et al., 1985) and
library mass spectra against an isotopic ratio within
Fig. 1. A schematic diagram of the experimental apparatus for PVC combustion.

498
D. Wang et al. / Chemosphere 53 (2003) 495–503
þ
ꢀ
10% of the theoretical value. The possible positions of
chlorine atoms attached to the aromatic rings were
predicted by QSPR.
ion (M ) of the two peaks occurs at m=z 200, and a
fragment ion occurs at m=z 165, indicating that a
chlorine atom is attached to fluorene molecule. The
compound corresponding to Peak 6 was identified as
monochlorofluorene on the basis of mass spectrum. The
compound corresponding to Peak 7 was tentatively
identified as 2-chlorofluorene from the mass spectrum
and relative retention time.
Cl-PAHs were quantified from the peak area of mass
chromatograms using an external standard technique.
Monochlorobiphenyl was determined against PCB14, of
which relative response factor was estimated to be
1
.638 (Ericksson, 1992). Semiquantitative estimation of
Cl-PAHs of which authentic samples were not available
was tentatively performed according to 9-chlorophen-
anthrene.
Fig. 3D shows the typical mass spectra of Peak 8, 9, 10,
11, 12, eluting at RT 39.73, 39.89, 40.12, 40.30, 40.55 min,
þ
respectively, as shown in Fig. 2. The molecular ion (M )
Quality assurance criteria for Cl-PAHs analysis were
based on the first measure of a blank sample covering
the complete analytical procedure. The limit of detection
of these peaks occurs at m=z 212, and a fragment ion
occurs at m=z 176. Compounds corresponding to these
peaks were tentatively identified as monochlorophen-
anthrene or monochloroanthracene. The mass spectra
and retention times of the compounds corresponding to
Peak 10 and 12 were in agreement with those of 9-chloro-
phenanthrene and 9-chloroanthracene, respectively.
The compounds corresponding to Peak 13 (RT 44.81
min) in Fig. 2 gave a molecular ion at m=z 246 (Fig. 3E),
indicating two chlorine atoms, and two fragment ions at
(
LOD) for chlorinated naphthalene and chlorinated
phenanthrene were estimated to be 0.061 and 0.072 lg/g
PVC respectively using a signal-to-noise ratio of 3:1.
Recovery tests were carried out by spiking standard
solution containing Cl-PAHs into glass fabric filter,
which was then sequentially subjected to the entire
sample preparation process. The recoveries of Cl-PAHs
were within the range of 83–113%. Reproducibility for
the determination of Cl-PAHs by external standard
technique was within 12%.
þ
þ
m=z 212 (M –Cl) and 176 (M –Cl–HCl). Peak 14 (RT
45.06 min) in Fig. 2 gave a similar fragmentation pat-
tern, with two chlorine atoms, to that of Peak 13. The
compound corresponding to this peak was tentatively
identified as 9,10-dichlorophenathrene on the basis of
relative retention time (Shiraishi et al., 1985) (Table 2).
The molecular ion and its fragmentation pattern indi-
cate the occurrence of dichloroanthracene or dichloro-
phenanthrene during combustion of PVC.
3
. Results and discussion
The extraction ion chromatograms of Cl-PAHs re-
leased from PVC combustion at 900 °C are shown in
Fig. 2. At least 18 chlorinated compounds can be seen
on the chromatogram. Fig. 3 shows the typical mass
spectra of these compounds.
Fig. 3F shows the typical mass spectra of Peak 15, 16,
17, 18, eluting at RT 46.33, 46.54, 46.97, 47.82 min,
þ
respectively. The molecular ion (M ) of these peaks
occurs at m=z 236, indicating one chlorine atom, and a
fragment ion occurs at m=z 200, which arises from the
loss of one chlorine atom from the molecular ion.
Compounds corresponding to these peaks were tenta-
tively identified as monochloropyrene and monochlo-
rofluoranthene on the basis of mass spectra.
The mass spectrum of Peak 1 (RT 22.95 min) in Fig.
þ
2
is shown in Fig. 3A. The molecular ion (M ) of the
peak occurs at m=z 162, indicating one chlorine atom,
and a fragment ion occurs at m=z 127, which arises out
of the loss of one chlorine atom from the molecular ion.
The mass spectrum and retention time of this compound
was in agreement with those of 2-chloronaphthalene.
Peak 2 (RT 23.17 min) in Fig. 2 gave a similar molecular
ion and a similar fragment ion. This compound was
identified as 1-chloronaphthalene on the basis of com-
paring its retention time and mass spectrum with those
of authentic compounds.
For the compounds, in which positions of chlorine
atoms attached to the aromatic rings cannot be deter-
mined, QSPR was used to predict the possible struc-
tures. Eight parameters obtained by MNDO quantum
chemical computations were used as molecular descrip-
tors, including total energy (TENG), heat of formation
(HF), ionization potential (IP), dipole moment (DM),
molecular surface area (MSA), LUMO, surface area of
Cl atoms (ClSA) and net atomic charge of Cl (ClNC).
For Cl-PAHs substituted by two chlorine atoms, the
larger absolute value is adopted. A better regression
model is obtained:
Fig. 3B shows the mass spectrum of Peak 4 (RT 29.00
þ
min) in Fig. 2. The molecular ion (M ) of the peak oc-
curs at m=z 188, and a fragment ion occurs at m=z 152,
indicating that this compound contains one chlorine
atom. Peak 3 (RT 26.82 min) and Peak 5 (RT 29.22 min)
have similar mass spectra to that of Peak 4. Compounds
corresponding to these peaks were identified as mono-
chlorobiphenyl from the mass spectra of each peak.
Fig. 3C shows the typical mass spectrum of Peak 6
RRT ¼ ꢁ2:0459 ꢁ 0:000906TENG ꢁ 8:7340ClNC
þ 0:06175DM
2
(
RT 34.40 min) and 7 (RT 35.32 min). The molecular
R ¼ 0:9937 F ¼ 735:458 ð> F0:0001
Þ

D. Wang et al. / Chemosphere 53 (2003) 495–503
499
Fig. 2. Extraction ion chromatograms of Cl-PAHs from PVC combustion at the furnace temperature of 900 °C.
T
T
0
3
¼ ꢁ16:732;
T
1
¼ ꢁ46:850;
T
2
¼ 10:482;
A summary of Cl-PAHs identified and predicted is
presented in Table 2. About 18 Cl-PAHs were deter-
mined, among which most were monochlorinated
¼ 6:240 ðjT0;1;2;3j > jTj0:0001
Þ

500
D. Wang et al. / Chemosphere 53 (2003) 495–503
Fig. 3. Representative mass spectra of some chromatographic peaks in Fig. 2.
PAHs, and only two dichlorophenanthrenes or an-
thracenes were identified.
900 °C. Dichlorinated phenanthrene or anthracene in
particular was detected only at 800 and 900 °C. The total
amount of these Cl-PAHs also gradually increased from
13.58 lg/g PVC at 600 °C to 101.95 lg/g PVC at 900 °C.
This tendency of increasing formation of Cl-PAHs
with increasing furnace temperature was also observed
in PAHs formation in the emissions during combustion
The types and the yields of Cl-PAHs released from
PVC combustion at different furnace temperature are
listed in Table 3. The result shows that the species and the
levels of these Cl-PAHs increased with increasing furnace
temperatures within the experimental range of 600–

D. Wang et al. / Chemosphere 53 (2003) 495–503
501
Table 2
Cl-PAHs identified by GC retention time, mass spectra and QSPR prediction in the emissions from PVC combustion
a
No.
Compounds
RRT (Obs.) Compounds
RRT (Pred.) Ions selected
b
b
b
b
1
2
3
4
5
6
2-Chloronaphthalene
1-Chloronaphthalene
Monochlorobiphenyl
Monochlorobiphenyl
Monochlorobiphenyl
Monochlorofluorene
Phenanthrene
0.658
0.663
0.770
0.832
0.838
0.987
1.000
1.014
1.140
2-Chloronaphthalene
0.6293
0.6557
0.8270
0.8383
0.8424
0.9764
1.0000
1.0133
1.1262
164, 162, 127
1-Chloronaphthalene
4-Chlorobiphenyl
3-Chlorobiphenyl
1-Chlorobiphenyl
1-Chlorofluorene
Phenanthrene
164, 162, 127
190, 188, 152
190, 188, 152
190, 188, 152
202, 200, 165
c
c
7
8
2-Chlorofluorene
Monochlorophenanthrene/monochloro-
anthracene
2-Chlorofluorene
2-Chloroanthracene
1-Chloroanthracene
202, 200, 165
214, 212, 176
9
Monochlorophenanthrene/monochloro-
anthracene
b;c
9-Chlorophenanthrene
1.145
1.1471
214, 212, 176
b;c
1
0
1
1.152
1.157
9-Chlorophenanthrene
1-Chlorophenanthrene
1.1527
1.1539
214, 212, 176
214, 212, 176
1
Monochlorophenanthrene/monochloro-
anthracene
b;c
9-Chloroanthracene
b;c
1
2
3
1.163
1.286
9-Chloroanthracene
9,10-Dichloroanthracene
1.1689
1.2813
214, 212, 176
248, 246, 176
1
Dichlorophenanthrene/dichloroanthra-
cene
c
c
14
15
16
17
18
a
9,10-Dichlorophenanthrene
c
3-Chlorofluorancene
1.293
1.329
1.337
1.349
1.373
9,10-Dichlorophenanthrene
c
3-Chlorofluorancene
1.2922
1.3117
1.3124
1.3595
1.3977
248, 246, 176
238, 236, 200
238, 236, 200
238, 236, 200
238, 236, 200
Monochloropyrene
Monochloropyrene
Monochloropyrene
2-Chloropyrene
1-Chloropyrene
4-Chloropyrene
Relative retention time calculated against phenanthrene.
Comparison with authentic standard.
Comparison with reference retention data (Shiraishi et al., 1985).
b
c
Table 3
Types and yields of Cl-PAHs in the emissions from PVC combustion at the temperature range of 600–900 °C
Compounds Yields (lg/g PVC)
00 °C
6
700 °C
800 °C
900 °C
2
1
4
3
1
1
2
2
1
9
1
9
9
9
3
2
1
4
-Chloronaphthalene
-Chloronaphthalene
-Chlorobiphenyl
1.76
2.68
0.59
1.19
1.02
3.90
7.56
0.99
2.48
0.22
0.27
0.70
0.47
2.92
6.63
7.06
1.18
nd
10.71
22.86
1.36
4.06
3.47
0.41
0.74
0.14
4.39
9.64
11.90
0.70
0.49
0.68
0.96
1.98
1.14
4.49
80.12
11.50
29.08
1.32
-Chlorobiphenyl
-Chlorobiphenyl
3.86
3.38
-Chlorofluorene
-Chlorofluorene
nd
0.50
0.96
0.73
0.074
0.78
1.55
1.39
0.36
-Chloroanthracene
-Chloroanthracene
-Chlorophenanthrene
-Chlorophenanthrene
-Chloroanthracene
,10-Dichloroanthracene
,10-Dichlorophenanthrene
-Chlorofluoranthene
-Chloropyrene
0.062
5.83
14.01
16.68
1.80
nd
nd
0.30
0.41
nd
0.36
0.43
nd
1.03
2.03
1.39
3.66
42.49
1.25
2.91
-Chloropyrene
-Chloropyrene
1.40
6.62
0.75
13.58
Total
101.95
‘
‘nd’’ refers to lower than detection limit.
of PVC by other researchers. Hawley-Fedder et al.
1984) found that under simulated incinerator condi-
tions, the total amounts of PAHs produced from PVC
combustion are greatest at 950 °C within the furnace
(

502
D. Wang et al. / Chemosphere 53 (2003) 495–503
temperature range of 800–950 °C. This suggested that
polychlorinated PAHs and their parent PAHs are easily
formed at higher furnace temperature, i.e. 800 and 900
References
Ahling, B., Bjørseth, A., Lunde, G., 1978. Formation of
chlorinated hydrocarbons during combustion of poly(vinyl
chloride). Chemosphere 7, 799–806.
°C. The possible reasons are as follows: These com-
pounds contain aromatic rings which are stable at higher
temperatures, and more favored cyclisation reaction can
occur by different gaseous hydrocarbons intermediates
released from incomplete combustion of PVC. Thus
increasing furnace temperature can enhance these PAHs
formation. Cl-PAHs detected in these tests are mainly
derivatives of naphthalene, phenanthrene, anthracene,
biphenyl, fluorene and pyrene. These chlorinated deriv-
atives might be formed mainly by chlorination reaction
of the PAHs with chlorine released from PVC combus-
tion, because direct scission of PVC chains to form
chlorine-containing compounds is a very minor decom-
position pathway (Lattimer and Kroenke, 1980).
Brodskv, E.S., Klyuev, N.A., Soyfer, V.S., Ibragimov, V.A.,
1
999. Chlorinated polycyclic aromatic hydrocarbons in the
waste water of pulp and paper mills. Organohalogen
Compd. 43, 127–130.
Colmsj o€ , A., Rannug, A., Rannug, U., 1984. Some chloro
derivatives of polynuclear aromatic hydrocarbons are
potent mutagens in Salmonella typhimurium. Mut. Res.
135, 21–29.
Colmsj o€ , A., Nilsson, U., Rannug, A., Rannug, U., 1988.
Polynuclear Aromatic Hydrocarbons: A Decade of Pro-
gress. Battelle Press, Columbus, OH, pp. 147–153.
Eklund, G., Str o€ mberg, B., 1983. Detection of polychlorinated
polynuclear aromatics in flue gases from coal combustion
and refuse incinerators. Chemosphere 12, 657–660.
Elomaa, M., Sawaranta, L., Mikkola, E., Lallonen, R., Zitting,
A., Zevenhoven, C.A.P., Hupa, M., 1997. Combustion of
polymeric materials. Crit. Rev. Anal. Chem. 27, 137–197.
Ericksson, M.D., 1992. Analytical Chemistry of PCBs. Lewis
Publishers, London, pp. 225–228.
Furnace temperature, airflow, the burning time and
the combustion duration have an effect on the levels of
Cl-PAHs in the emissions from PVC combustion. Dur-
ing combustion under high temperature, HCl is first
released from PVC, and then PAHs are produced in the
cooler zone of the furnace by cyclisation reaction of
gaseous hydrocarbon fragments formed by incomplete
combustion. But the PVC burning with a flame, which
can be observed, started approximately at 4 s (900 °C)
and 45 s (600 °C) after the boat containing PVC plastics
was positioned in the oven. Due to this delay of PVC
burning with decreasing furnace temperature, most of
HCl may leave along with the airflow, and a temperature
Falandysz, J., 1998. Polychlorinated naphthalenes: an environ-
mental update. Environ. Pollut. 101, 77–90.
Franzen, B., Haaparanta, T., Gustafsson, J.-ØbdA., Toftg ꢀa rd,
R., 1988. TCDD receptor ligands present in extracts of
urban air particulate matter induce aryl hydrocarbon
hydroxylase activity and cytochrome P-450c gene expression
in rat hepatoma cells. Carcinogenesis 9, 111–115.
ꢀ
Haglund, P., Alsberg, T., Bergman, A., Jansson, B., 1987.
Analysis of halogenated polycyclic aromatic hydrocarbons
in urban air, snow and automobile exhaust. Chemosphere
2
dependent equilibrium between HCl and Cl : 2HCl
M
1
6, 2441–2450.
(
g) + 1/2 O
2
ꢁ! Cl
2 2
(g) + H O (g) can shift to the reverse
Hawley-Fedder, R.A., Parsons, M.L., Karasek, F.W., 1984.
Products obtained during combustion of polymers under
simulated incinerator conditions III. Polyvinylchloride. J.
Chromatogr. 315, 211–221.
reaction, thus reduce the opportunity of PAHs exposure
to Cl . So less Cl-PAHs formation was found at lower
2
furnace temperature of 600 °C. On the contrary, with
increasing furnace temperature, PAHs released from
PVC combustion have increasing chances of exposure to
Kannan, K., Imagawa, T., Blankenship, A.L., Giesy, J.P., 1998.
Isomer-specific analysis and toxic evaluation of polychlori-
nated naphthalenes in soil, sediment, and biota collected
near the site of a former chlor-alkali plant. Environ. Sci.
Technol. 32, 2507–2514.
Cl
2
, thus promoting the Cl-PAHs formation.
However, it should be noted that this experimental
system was but a simulated incinerator, in which
abatement of Cl-PAHs formation cannot be anticipated
only by changing the furnace temperature. In fact in a
combustion system, complicated factors have an effect
on Cl-PAHs formation. Further works on this subject
will be performed.
Lattimer, R.P., Kroenke, W.J., 1980. The formation of volatile
pyrolyzates from poly(vinyl chloride). J. Appl. Polym. Sci.
2
5, 101–110.
L o€ froth, G., Nilsson, L., Agurell, E., Sugiyama, T., 1985.
Salmonella/microsome mutagenicity of monochloro deriva-
tives of some di-, tri- and tetracyclic aromatic hydrocarbons.
Mut. Res. 155, 91–94.
€
Nilsson, U.L., Ostman, C.E., 1993. Chlorinated polycyclic
aromatic hydrocarbons: method of analysis and their occur-
rence in urban air. Environ. Sci. Technol. 27, 1826–1831.
Oehme, M., Manø, S., Mikaelsen, A., 1987. Formation and
presence of polyhalogenated and polycyclic compounds in
the emissions of small and large scale municipal waste
incinerators. Chemosphere 16, 143–153.
Acknowledgements
This work is supported by RCEES 9906 and NSFC
9837180. The authors are grateful to the Agilent Tech.
2
Co. for the apparatus supports and Dr. Yang Jian of
Institute of Plant Protection, Chinese Academy of Ag-
riculture Sciences for his valuable help.
Piao, M.J., Chu, S.G., Zheng, M.H., Xu, X.B., 1999. Charac-
terization of the combustion products of polyethylene.
Chemosphere 39, 1497–1512.

D. Wang et al. / Chemosphere 53 (2003) 495–503
503
Rannug, U., Nilsson, U., Colmsj o€ , A., 1986. In: Ramel, C.,
Lambert, B., Magnusson, J. (Eds.), Genetic Toxicology of
Environmental Chemicals. Part B: Genetic Effects and
Applied Mutagenesis. Alan Liss, New York, pp. 157–164.
Rappe, C., Marklund, S., Bergquist, P.A., Hansson, M., 1982.
Polychlorinated dioxins (PCDDs), dibenzofurans (PCDFs)
and other polynuclear aromatics (PCPNAs) formed during
PCB fires. Chem. Scr. 20, 56–61.
Sovocool, G.W., Donnelly, J.R., Munslow, W.D., Vonnahme,
T.L., Nunn, N.J., Tondeur, Y., Mitchum, R.K., 1989.
Analysis of municipal incinerator fly ash for bromo- and
bromochloro-dioxins, dibenzofurans, and related com-
pounds. Chemosphere 18, 193–200.
Tausch, H., Stehlik, G., 1985. Analysis of polycyclic
aromatic compounds in the fly ash of an incineration
plant for radioactive waste. J. High Res. Chromatogr. 8,
525–527.
Registry of Toxic Effects of Chemical Substances, 1998. CD-
ROM, NIOSH.
ꢀ
T o€ ftg ꢀa rd, R., Franz ꢁe n, B., Gustavsson, J.-A., L o€ froth, G., 1985.
Shiraishi, H., Pilington, N.H., Otsuki, A., Fuwa, K., 1985.
Occurrence of chlorinated polynuclear aromatic hydrocar-
bons in tap water. Environ. Sci. Technol. 19, 585–590.
Characterization of TCDD-receptor ligands present in
extracts of urban particulate matter. Environ. Int. 11, 369–
374.