The effect of oils on PAH, PCDD, PCDF, and PCB emissions from a spark engine fueled with leaded gasoline

Article abstract of DOI:10.1016/S0045-6535(00)00053-9

The effect of synthetic and mineral oils on the formation of polyaromatic hydrocarbons (PAHs), polychlorinated dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and biphenyls (PCBs) in emissions from a spark ignition engine was studied on a Skoda Favorit engine fueled with leaded gasoline. The test cycle simulated urban traffic conditions on a chassis dynamometer, in accordance with the ECC 83.00 test. The data for selected PAHs as well as PCDDs, PCDFs, and PCBs congener profiles are presented. PCDD/Fs emissions for an unused oil and the oil after 10 000-km operation varied from 300 to 2000 fmol/m³, PCBs emissions from 75 to 178 pmol/m³, and PAHs emissions from 150 to 420 μg/m³. The content of PCBs in oils varied from 2 to 920 mg/kg.

Full text of DOI:10.1016/S0045-6535(00)00053-9

Chemosphere 41 (2000) 1905±1911
The e€ect of oils on PAH, PCDD, PCDF, and PCB emissions
from a spark engine fueled with leaded gasoline
e
Jan Broz , Roman Grabic , Juraj Kilian , Milan Lojkasek , Stellan Marklund ,
c
e
c
b
ꢁ
ꢁ
ꢀ
ꢁꢀ
c
a,
d
a
Tomas Ocelka , Vladimõr Pekarek , Josef Pribyl , Vratislav Tydlitat ,
*
ꢁ
ꢁ
ꢀ
ꢁ
d
ꢀ
ꢁ
Jan Vyska
a
Institute of Chemical Process Fundamentals, Rozvojovaꢁ 135, 16502 Praha 6, Czech Republic
b
Department of Environmental Chemistry, Umeꢂa University, S-90187 Sweden
District Public Health Department, Palackeꢁho 121, 73802 Fryꢁdek-Mꢁõstek, Czech Republic
c
d
Motor Vehicle Research Institute, Lihovarskaꢁ 12, 18068 Praha 9, Czech Republic
e
ꢁ
Research Institute of Inorganic Chemistry Ltd., Revolucꢀnꢁõ 84, 40001 Ustꢁõ nad Labem, Czech Republic
Received 3 December 1999; accepted 7 February 2000
Abstract
The e€ect of synthetic and mineral oils on the formation of polyaromatic hydrocarbons (PAHs), polychlorinated
dibenzo-p-dioxins (PCDDs), dibenzofurans (PCDFs), and biphenyls (PCBs) in emissions from a spark ignition engine
ꢀ
was studied on a Skoda Favorit engine fueled with leaded gasoline. The test cycle simulated urban trac conditions on
a chassis dynamometer, in accordance with the ECC 83.00 test. The data for selected PAHs as well as PCDDs, PCDFs,
and PCBs congener pro®les are presented. PCDD/Fs emissions for an unused oil and the oil after 10 000-km operation
varied from 300 to 2000 fmol/m3, PCBs emissions from 75 to 178 pmol/m3, and PAHs emissions from 150 to 420 lg/m3.
The content of PCBs in oils varied from 2 to 920 mg/kg. Ó 2000 Elsevier Science Ltd. All rights reserved.
Keywords: Spark engine; Oil; Emissions; PAH; PCB; PCDD/F
1. Introduction
1987, 1990; Bingham et al., 1989; Hagenmaier et al.,
1990; Oehme et al., 1991; Schwind et al., 1991) who
reported PCDD/Fs emission data. The conditions under
which the just mentioned compounds are formed during
In many industrial countries, PCDD/Fs as one of the
most toxic components in trac emissions make up only
a few percent in the inventory of all known dioxin
sources. However, in relation to environmental protec-
tion, the relevance of the emissions formed during gas-
oline combustion lies in the fact that a great number of
these small-scale generators of these pollutants operate
in regions with dense population and with the emissions
close to the recipient. For that reason, trac emissions
were studied by a number of authors (Marklund et al.,
ꢁ
combustion processes are well-known (Pekarek et al.,
1995). The content of halogens and lead in the gasoline
became the key aspect in these studies. In some countries
(e.g., in Germany), the use of leaded gasoline containing
halogenated scavengers has been banned, as PCDD/Fs
emissions attained the values ranging from 30 to 520 pg
I-TEQ PCDD/Fs per km (Marklund et al., 1987; Bing-
ham et al., 1989; Hagenmaier et al., 1990; Schwind et al.,
1991), while those for leaded and unleaded gasoline were
39±151 and 4±19 pg I-TEQ PCDD/Fs per km, respec-
tively (Schwind et al., 1991). Hence, with regard to the
environmental protection, replacement of leaded fuels
* Corresponding author.
E-mail address: pekarek@icpf.cas.cz (V. Pekarek).
Â
0045-6535/00/$ - see front matter Ó 2000 Elsevier Science Ltd. All rights reserved.
PII: S 0 0 4 5 - 6 5 3 5 ( 0 0 ) 0 0 0 5 3 - 9

1906
J. Broꢀz et al. / Chemosphere 41 (2000) 1905±1911
by lead-free fuels is well-founded on both theoretical
and experimental grounds.
distance) consisted of twelve engine testing runs. The
consumption of gasoline was expressed by weight. For
each oil type used, the engine was washed with toluene
(total of 2 l) and conditioned by 6±20 min/running.
However, reported PCDD/Fs emission data di€er
greatly, likely due to di€erent conditions of the experi-
ments and their dicult reproducibility (Schwind et al.,
1991).
2.2. Sampling
Nevertheless, the great di€erence in PCDD/Fs emis-
sions reported in studies of Marklund (1987, 1990)
dealing with the combustion of leaded gasoline con-
taining scavengers cannot be explained solely on this
ground. On the other hand, it might be probably stres-
sed that the values by one order of magnitude lower
were obtained (Marklund, 1990) on an engine lubricated
by a semisynthetic oil.
The sampling (national standard CSN 124070) was
made with the tube placed at 150 mm far from the
prolonged (é 50 mm) end of the tail pipe. The mea-
surements of emissions, ¯ow rates, temperature (90±
110°C), and pressure were realized at the same sampling
point. An absorption apparatus with heated probe was
used. The coarse particles were collected on glass ®bre
®lter Whatmann GF/F. The aerosols were absorbed in
two impingers with 100 ml of ice-cooled 2-ethoxyetha-
nol. The gas volumes were measured after a dry-tower
by a gasometer. The volume of the withdrawn gas
sample varied from 3.4 to 4.0 m3 at 0°C at the pressure
101.3 kPa.
Polyaromatic hydrocarbons (PAHs) in trac emis-
ꢁ
sions have been studied in detail by Parma and Novak
ꢀ
(1994), Sebor et al. (1994), and by Mi et al. (1996).
The above facts prompted us to study the e€ect of
synthetic and mineral oils on PAH, PCB, PCDD, and
PCDF emissions generated during combustion of leaded
gasoline. Determination of these noxious substances
under identical experimental conditions using two dif-
ferent analytical methods and two successive experi-
mental runs could render the mutual comparison of
formation of these compounds for the oils under study
possible. This would enable us to discuss factors that
a€ect formation of these persistent organic pollutants
(POP). Along with these experiments, the accumulation
of PAH and PCB in motor oils has also been examined.
2.3. Oils and chemicals
The synthetic Pennzoil Performax SAE 5W-40 oil
and the mineral Mogul Forte GX SAE 15W-40 were
used.
2.4. Analytical methods
A1 and A2. The puri®cation steps in Table 1 were
used both in the Czech and Swedish laboratory for
PAH, PCB, and PCDD/F analysis for treating exhaust
gas samples.
2. Experimental
A3 and A4. The oil samples for PAH and PCB
analysis in both laboratories were pretreated in accor-
dance with the standard operational procedure (semi-
permeable membrane device) to remove substantial part
of the oil and then analyzed as described above (A1).
2.1. Experimental arrangement
ꢀ
A Skoda Favorit passenger car (driven for about
118 000 km) with carburettor spark ignition engine of
1.3 l displacement without catalytic converter was used
for these experiments. The emissions according to the
EEC 83.00 test for CO, CH, and NOx were below the
EEC standard limits.
2.5. Data presentation
The experiments simulated urban trac regime ac-
ꢁ
To make comparison of our results with reported
data, the outgas was calculated from the fuel con-
sumption at its theoretical molar composition
(C:H ˆ 1:2). The volume of dry exhaust gases was es-
timated to 9.41 m3/kg fuel. The formation of 1 m3 of dry
cording to the EEC15 and 83 tests (Bratsky and Stacho,
1993) on dynamometer rolls. The maximum, minimum,
and mean velocities were 50, 0, and 19 km/h, respec-
tively. One experiment (duration 2 h 44 min, ca 48-km
Table 1
The puri®cation steps
Extraction (glass ®lter with toluene-24 h, 2-ethoxyethanol/H2O by hexane)
#
10 ml volume adjustment
!
5% aliquot for PAH analysis, HPLC (A1), GC/MS (A2)
#
Puri®cation by H2SO4±SiO2 column-Al2O3 column
!
2% CH2Cl2 in hexane-PCB fraction,
GC/MS/MS analysis (A1), GC/HRMS analysis (A2)
50% CH2Cl2/hexane-``C'' column, toluene-PCDD/F (GC/MS/MS analysis-A1, GC/HRMS-A2)
#

J. Broꢀz et al. / Chemosphere 41 (2000) 1905±1911
1907
outgas (0°C, 101.3 kPa) thus corresponds to 1:73 Æ 0:04
where non detected values in Table 3 were neglected.
The higher values were calculated by the use of the half
value of presented detection limits. The concentration
range shows the uncertainty of these estimations.
Results obtained in experiment II prove that under
the same experimental arrangement, PCDD measure-
ments (Sa, Sb) in both runs yield well-comparable data.
This con®rms good reproducibility of the sampling
method, analyses and of the state of the engine as well.
The TEQs in emissions from synthetic oil (Sa) in ex-
periment II shows lower concentrations than in emis-
sions from mineral oils (M and MR).
ꢀ
km for the Skoda Favorit car.
3. Results and discussion
3.1. Explorative experiments with synthetic oil
In our ®rst trial experiment (experiment I), with the
synthetic oil (see Section 2) we used the leaded gasoline
91 octane, containing 111 mg Cl/kg, 54.5 wt.% aromatic
hydrocarbons (AH) and 0.13 g Pb/kg. The data obtained
in this work are compared in Table 2 with the mea-
surements by Marklund et al. (1990) of gasoline con-
taining scavengers.
The congener pro®les in emissions from mineral oils
in experiments II and III di€er. However, within the
experiment III similar pro®les were found for M and
MR oils. The content of PCDD/Fs in emissions for
synthetic oils di€er signi®cantly not only in concentra-
tions but in their congener pro®les as well.
It is worth mentioning that the reported data were
obtained with the GC/HRMS method (A2). Our ®nding
and older results (Marklund et al., 1990) are in agree-
ment with very low values of PCDD/Fs.
The di€erences in the experiments with the same car
and under comparable experimental arrangement dem-
onstrate well that how the emissions of PCDD/F might be
in¯uenced by the fuel, oil, and combustion conditions.
The concentration of PCDFs in exhaust gases is higher
than PCDDs in all experiments and the PCDD/PCDF
ratio in experiments II and III show di€erences in chem-
ical processes by fuel combustion in the same engine.
Our results present the possible concentration inter-
val for PCDD/Fs which might be expected for leaded
gasoline and engines of comparable type: 1.4±11 pg
TEQ/m3. It may be of general interest that our results of
urban air measurements in Prague during the winter
period of 1996 are scattered between the values of 0.047±
0.39 pg TEQ/m3. PCDD/Fs here determined come of
course from all urban sources.
3.2. E€ect of oil type on PCDDs and PCDFs emissions
The second experiment (experiment II) was carried
out with the leaded gasoline 91 octane (Czech made)
containing 0.13 g Pb/kg, 3.0 mg Cl/kg, 37.2 wt.% AH,
and 8% MTBE. Three motor oils were studied: synthetic
oil (S), mineral oil (M), and the same mineral oil after
10 000-km operation (MR). Emission samples were an-
alyzed in the Czech laboratory by the analytical method
A1. The reproducibility of sampling, analytical methods,
and PCDD determinations was evaluated from two
successive runs (Sa and Sb) using GC/MS/MS technique.
In the third experiment (experiment III), we used the
same gasoline of di€erent production charge which
contained 0.12 g Pb/kg, 1.0 mg Cl/kg, 34.4 wt.% AH,
and 2.8% MTBE. Here the preparation of the samples
for analysis and analytical determination of polychlori-
nated dibenzo-p-dioxins (PCDDs) and polychlorinated
dibenzofurans (PCDFs) congeners were performed both
in the Czech Republic (method A1) and in laboratories
in Sweden (method A2) and samples were taken in
two consecutive runs. Obtained data are presented in
Table 3.
3.3. E€ect of oils on PCBs spark ignition engine emissions
The data of the content of polychlorinated biphenyls
(PCBs) in emissions of spark ignition engines are rare in
the literature. We found only the data of Parma and
ꢁ
Novak (1994), however, information under which con-
ditions these measurements were done is not available.
During experiment III, PCBs content in emissions was
measured in two laboratories ꢀA1; A2† on di€erent in-
struments. The results are summarized in Table 4.
The results are in fairly good agreement with respect
to di€erent analytical laboratories. The emissions do not
contain remarkable amounts of octa- to deca-chlori-
nated biphenyls. Higher values of tetra- and hexachlo-
rinated biphenyl congeners were found in emissions
from synthetic oil.
The I-TEQ values in Table 3 were calculated using
WHO 25±29.5.1998 recommended TEFs. The sum of
the PCDD/Fs is presented as the TEQ concentration
interval. The lower values correspond to calculations
Table 2
The data obtained in comparison with the measurements by
Marklund et al. (1990) of gasoline containing scavengers
The intake air for the engine was also analyzed on the
PCBs content. Its concentration equaled to 0.3 pmol
PCB/m3, in agreement with measurements in other parts
of the Prague territory. In the exhaust gases, the PCB
concentration is more than two orders of magnitude
pg I-TEQ PCDD/Fs per km
Mineral oil (Marklund et al., 1987)
Semisynthetic oil (Marklund et al., 1990)
Synthetic oil (this work)
30±520
1.1±2.8
0.4

1908
J. Broꢀz et al. / Chemosphere 41 (2000) 1905±1911
Table 3
E€ect of di€erent oils on PCDD/Fs spark ignition engine emissions (experiments II and III)a
Experiment II
Method A1
Experiment III
Method A1
Method A2
Sa
Sb
M
MR
S
M
MR
S
M
MR
PCDD/Fs pg/m3
2378TCDD
<1.2
<1.2
2.0
<1.3
<1.3
2.9
<1.1
<1.4
1.1
1.7
<1.4
3.0
2.9
2.2
7.2
18
<2.2
<2.5
<2.4
<2.2
2.2
<1.7
<1.4
<1.3
<1.3
<1.0
<0.9
47
<0.8
<0.9
<0.8
<0.8
<0.6
0.7
2.1
11
8.2
14
11
21
56
7.0
15
14
18
11
16
5.3
41
15
76
<1.7
<4.0
<8.7
<7.2
<5.8
<9.2
20
<1.2
<6.5
<18
<9.2
<12
8.9
12378PeCDD
123478HxCDD
123678HxCDD
123789HxCDD
1234678HpCDD
OCDD
1.2
1.6
1.6
2.6
1.2
<1.3
5.6
3.5
11
6.9
27
15
57
15
28
4.0
39
7.7
2378TCDF
<2.7
<2.2
<2.4
2.3
NA
NA
NA
NA
NA
NA
NA
NA
NA
NA
<4.0
<4.0
<6.4
6.3
4.9
5.3
5.3
8
5.7
5.6
2.8
4.3
2.9
12378PeCDF
23478PeCDF
123478HxCDF
123678HxCDF
234678HxCDF
123789HxCDF
1234678HpCDF
1234789HpCDF
OCDF
<1.1
2.1
1.0
1.7
5.0
3.9
7.2
7.8
2.6
4.3
1.0
1.8
0.4
0.9
5.0
2.4
2.7
3.0
6.8
11
4.7
8.7
<1.8
39
8.7
8.6
1.7
3.5
4.0
1.2
4.5
0.7
6.8
<1.2
8.8
<2.5
24.0
5.0
0.5
69
<4.3
6.6
<3.9
19
19
2.3
8.8
1.5
<1.3
<1.8
10
19
6.6
66.0
<12
<66
<17
59
<4.7
14.0
3.0
TEQ pg/m3
PCDD=Fs
P
1.4±3.5
±
3.3±6.5
6.9±8.6
8.2±11
2.4±4.1
2.1±3.0
30
2.9±7.2
4.7±11
Total PCDD/Fs fmol/m3
P
TCDD
50
19
62
17
81
36
93
15
150
130
84
21
20
43
23
100
85
91
45
61
32
520
11
11
22
11
61
110
38
19
29
7
140
280
280
100
120
72
<120
<160
<190
<40
43
<78
<260
<310
66
P
P
P
PeCDD
HxCDD
HpCDD
64
21
61
33
95
42
72
42
73
OCDD
24
33
59
39
120
420
270
250
240
150
1900
85
98
P
TCDF
150
28
NA
NA
NA
NA
NA
±
300
320
250
69
420
210
200
170
43
52
94
P
P
P
PeCDF
HxCDF
HpCDF
320
370
250
<150
2000
110
100
110
130
700
16
22
86
23
OCDF
<4.1
390
<11
1300
<150
300
P
PCDD=Fs
1300
320
a S: synthetic oil; M: mineral oil; MR: mineral oil after 10 000 km-operation; NA: not analyzed.
higher: 75±178 pmol PCB/m3. It is worthy to note that
the concentrations of PCB congeners reported by Parma
benzo(a)pyrene with some less toxic PAHs. Ben-
zo(a)anthracene was present in signi®cant amounts. Our
results agree with the reported data. The results of
ꢁ
and Novak (1994) are of about one to two orders greater
ꢁ
Parma and Novak (1994) are also in a good agreement
with our ®ndings.
then our data. Because of the still great number of
leaded gasoline-fed cars as emission resources in the
highly populated urban regions in many countries, the
PCBs values thus found are warning.
Used mineral oils generate greater total amounts of
PAHs.
3.4. E€ect of oils on PAHs spark ignition engine emissions
(experiment III)
3.5. Contents of PCBs and PAHs in motor oils
Biphenyls and polycyclic aromatic hydrocarbons
generated during fuel combustion are characterized by
their high solubility in lipophilic media. The oil on
cooler walls of the combustion space contacts with
formed persistent organic compounds. Preliminary an-
alyses of PCBs and PAHs in oils (Table 6), support the
presumptions. The relatively high PAHs and PCBs
In experiment III, the selected PAHs were also de-
termined in both laboratories ꢀA1; A2†. Results pre-
sented in Table 5 are compared with data reported in the
literature for leaded gasoline.
In our experiments, we have observed in emissions
the formation of the toxic dibenzo(a,h)anthracene and

J. Broꢀz et al. / Chemosphere 41 (2000) 1905±1911
1909
Table 4
E€ect of oils on PCBs content in spark ignition engine
emissions (experiment III)a
aerosol particles, passed through the air cleaner. These
aerosol particles are very important for the formation of
PCB and PCDD/F. This fact is in an agreement with
Miyabara et al. (1999) ®ndings, where the amounts of 26
ng PCDD/Fs/g on exhaust particles were found.
PCB pmol/m3
Method A1
Method A2
On the other hand, oxygen-rich fuel mixtures en-
hance the extent of de novo synthetic reactions (Grabic
et al., 1999). Therefore, optimization of the oxygen
content becomes also a successful approach to emission
reductions. The oxygenous substances added to gasoline
increase the content of oxygen utilizable for fuel com-
bustion. Thus, 2.7 mass% of these compounds lead to a
decrease of 15% of the hydrocarbons and a 25% decrease
S
M
MR
S
M
MR
TriCB
28
66
28
28
62
27
NA
30
NA
24
NA
15
TetraCB
PentaCB
HexaCB
HeptaCB
OctaCB
NonaCB
27
58
18
29
25
39
22
40
20
27
16
23
24
13
12
17
28
21
2.2
0.3
<0.1
178
1.3
0.1
<0.1
89
1.0
0.1
<0.1
104
2.8
0.2
<0.1
112
4.5
<0.7
<0.1
104
NA
<0.9
<0.1
75
ꢀ
of CO in the exhaust gas (Sebor et al., 1995). However,
DecaCB
of PCB
ꢀtetra±
P
the e€ect of oxygen and oxygenous compounds on
PCDD/F distribution and their congener pro®les has
not as yet been determined, eventhough the presence of
oxygen seems to be an important factor in the present
knowledge about de novo synthetic reactions.
decaCB†
a S: synthetic oil; M: mineral oil; MR: mineral oil after 10 000-
km operation; NA: not analyzed.
The technical state of some other parts of the car
might a€ect the results too. Probably, the technical state
of the engine including exhaust system was the reason of
extremely high values of PAH and PCB which we met in
one of our experiments. The same conclusion can be
probably attributed for extremely high values of PCB
contents in used oil show that we deal with hazardous
waste which requires proper care during its combustion
or recycling. A potential presence of PCDD/F cannot be
excluded from the theoretical standpoint.
ꢁ
presented by Parma and Novak (1994).
3.6. In¯uences on POPs formation in spark engines
3.6.1. The state of tested engine
3.6.2. The e€ect of chlorine
De novo synthetic reactions are known to take place
on carbon particles and/or particles with silicate struc-
ture. The incomplete combustion (i.e., cold start, tech-
nical state of the engine) leads to formation of
carboniferous components, which will mix with dust
The restriction of PCB and PCDD/F in emissions can
be achieved by exclusion of the presence of halogens,
especially of chlorine in the reaction system, regardless
of whether it is organic- or inorganic-bond. The chlorine
content of 3±4 mg Cl/kg in leaded gasoline and in oils of
Table 5
E€ect of di€erent oils on PAH spark ignition engine emissions (experiment III)a
PAHs lg/m3
ꢀ
(Mi et al., (Sebor et al.,
I-TEF
(Nisbet and
LaGoy, 1992)
Anal. method A1
Anal. method A2
1996)
1994)
Test
ECE
Test at
90 km/h
S
M
MR
55.3
S
M
MR
Anthracene
Fenantrene
0.01
0.001
0.1
54
7
19
9
46
8
26.5
19.9
21.0
14.0
77.0
1.3
36.0
131.0
2.9
117.1 100.6 215.9 92.0
Benzo(a)anthracene
Chrysene
Pyrene
4
2
20
20
24
NA
3
18.1
1.6
20.4
1.6
46.4
3.8
2.3
2.3
0.01
0.001
5
5
2
1.4
3.0
11
0.4
73
16
2
5
37.2
ND
NA
13.1
0.6
37.8
ND
NA
10.8
ND
ND
0.5
74.5
ND
NA
21.6
1.0
24.0
1.1
19.0
0.15
26.0
12.0
1.3
31.0
0.062
44.0
21.0
2.1
Dibenzo (a,h)-anthracene
Fluorene
NA
3
0.001
0.001
0.1
31.0
15.0
2.8
Fluoroanthene
Benzo(b)¯uoranthene
Benzo(k)¯uoranthene
Benzo(a)pyrene
Benzo(g,h,i)perylene
4
45
ND
ND
9
0.7
ND
ND
1.4
0.5
46.6
0.1
1
2
ND
0.6
0.4
0.7
ND
0.88
1.6
ND
0.25
0.46
0.24
ND
0.4
2
0.01
0.1
2
8
1.6
0.5
1.4
0.6
1.3
0.5
0.55
0.21
Indeno(1,2,3-c,d) pyrene
PAHs
2
180.4
3
186.0
1.2
P
216.9 193.6 421.4 195.2 153.1 272.2
a S: synthetic oil; M: mineral oil; MR: mineral oil after 10 000-km operation; NA: not analyzed; ND: not detected.

1910
J. Broꢀz et al. / Chemosphere 41 (2000) 1905±1911
Table 6
PCBs and PAHs in oilsa
P
P
PCBs mg/kgb
PAHs mg/kgc
Fresh oil
A3
After 50 km run
A3
Used oil
A3
Fresh oil
A3
After 50 km run
Used oil
A3
A4
A3
A4
A4
Synthetic oil
Semisynthetic oild
Mineral oil
22
20
NA
47
2.8
NA
56
2.0
3.7
1.8
51
25
NA
NA
926
NA
361
605
NA
9.9
NA
23
NA
77
NA
63
NA
a A3 and A4 analytical method in the Czech Republic and Sweden, respectively.
b PCBs No. 28, 52, 101, 138,153, 180.
c Sum of PAHs content of compounds from column 1 in Table 5, excluding ¯uorene.
d SAE 10 W/40, operated 15 000±20 000 km, Mazda.
15±40 mg Cl/kg in Mogul Forte and of 50 mg Cl/kg
in Pennzoil Performax and at oil consumption of
0.4±1 g/km, results in 200±300 ng Cl/km levels. We have
found that emissions amounts of 6±25 ng PCB/km and
0.1±0.7 ng PCDD/F/km. It is evident that the content of
chlorine in the fuel and oil is high enough to form the
experimentally found concentrations of PCB and PCDD
in emissions.
enes, aromatic compounds such as benzene, chloro-
benzenes, phenols, diphenyl ethers, and also PCB, the
ꢁ
latter forming PCDF (Pekarek et al., 1995). The pre-
cursors formed by gasoline combustion involve car-
bonmonoxide, 1,3-butadiene, acrolein, and PAH.
Carbonmonoxide is an ``excellent'' precursor in de novo
synthetic reactions, since the formation of chlorinated
POPs correlates well with its concentration. The addi-
tion of oxygenous components (e.g., MTBE or tert-butyl
amyl ether) to gasoline causes also reduction of CO
content in combustion gases. Analogously, benzene is
formed during fuel combustion via dealkylation of
aromatic C7±C12 hydrocarbons as well.
Varying chlorine content in the materials of the same
brand could be the reason of di€erent results when de-
termining the content of these substances in engine
emissions. It is worth to mention that in the period
following the dispersion of inorganic chlorides on roads
in winter the entry of halogen rich aerosol by air suction
into the combustion process is highly probable.
The important factor which a€ects POPs formation is
gasoline composition. Gasoline contains among others
alcohols, ethers, dienes, and AH. The AH content to
adjust the corresponding octane number ranges from 30
to 50 mass%, while benzene content should not exceed
5 vol%. A recommendation for gradual replacement of
leaded gasoline by reformulated unleaded fuels according
to the American Clean Air Act ± 1992 containing at
maximum of 25 vol% AH, 1 vol% benzene, and 2.7 mass%
oxygen (in the form of oxygenous components) is there-
fore the right step to a signi®cant reduction of POPs
precursors. The reproducibility of the experiments carried
out with the gasoline of the same brand but of di€erent
production charges can be in¯uenced not only by di€erent
chlorine and corrosion copper contents but also, and that
mainly, by chemical composition and amount of AH.
3.6.3. E€ect of heavy metals
The presence of heavy metals is fundamental for de
novo synthetic reactions to proceed. From this view-
point, the restricted use of leaded gasoline is the right
step leading to the reduction of PCB and PCDD/F
emissions.
In the combustion processes taking place in motor
vehicles, the e€ect of copper and its compounds has not
yet been discussed. These compounds are known to be
the most ecient catalysts of de novo synthetic reactions
ꢁ
(Pekarek et al., 1995). The presence of copper in motor
oils is due to the wear of metallic motor components.
Hence, it seems likely that copper content in oils will
increase with increasing degree of their wear, contrib-
uting thus to the formation of the above mentioned
pollutants in the combustion space. Gasoline producers
provide fuel analyses, which include data on the e€ect of
fuel on copper corrosion. However, the e€ect of copper
content in engine input materials on PCB and PCDD/F
formation has not been examined.
4. E€ect of oils
As to POPs formation by fuel combustion, the e€ect of
oil seems to be absolutely unknown. Its composition, so
important in regard to the components a€ecting de novo
synthetic reactions, has not been as yet unambiguously
resolved. Thus, for example Fourier transform of IR
spectra of semisynthetic and synthetic oils which we used
exhibit, according to our analytical results, similarity to
the spectrum of mineral oil used. Furthermore, synthetic
oils can contain chemically unknown additives. The dif-
3.6.4. E€ect of precursors
The course of de novo synthetic reactions in the
combustion system is signi®cantly a€ected by the pres-
ence of PCBs precursors. The main input precursors
which might be present at gasoline combustion are di-

J. Broꢀz et al. / Chemosphere 41 (2000) 1905±1911
1911
ꢁ
Bratsky, D., Stacho, D., 1993. Examination of car gasolines.
Ropa a uhlie 35 (2) 122±131 (in Slovak).
ferences between studied oils found by us could thus not
be expected to be large, because all the oils under study
contain the same mineral component. In our view, the
emission changes observed are mainly due to the di€erent
degree of the lubricating ability decrease. The formula-
tion of the oil composition such that it would minimize its
e€ect on de novo synthetic reactions is one way to im-
prove spark engine emissions. On this topic we did not
®nd any information in available literature sources.
ꢁ
ꢀ
ꢁꢀ
Grabic, R., Marklund, S., Pekarek, V., Puncochar, M., Ullrich,
J., 1999. E€ects of oxygen, sulphuric acid and peroxides on
formation of PCBs, PCDDs and PCDFs on the MWI ¯y
ash. Organohalogen Compounds 41, 251±254.
Hagenmaier, H., Dawidowski, N., Weberuss, U., Hutzinger,
O., Schwind, K.H., Thoma, H., Essers, U., Buehler, U.,
Greiner, R., 1990. Emission of polyhalogenated dibenzodi-
oxins and dibenzofurans from combustion-engines.
Organohalogen Compounds 2, 329±334.
Marklund, S., Andersson, R., Tysklind, M., Rappe, Ch.,
Egebaeck, K.E., Bjoerkman, E., Grigoriadis, V., 1990.
Emissions PCDDs and PCDFs in gasoline and diesel fueled
cars. Chemosphere 20, 553±561.
5. Conclusions
·
·
·
·
The spark ignition engine emissions contain not only
PCDD/F and PAH but also PCB.
Marklund, S., Rappe, Ch., Tysklind, M., 1987. Identi®cation of
polychlorinated dibenzofurans and dioxins in exhausts from
cars on leaded gasoline. Chemosphere 16, 29±36.
The higher PAH emissions were found for 10 000-km
operated oils as compared with new oils.
Mi, H.H., Lee, W.-J., Wu, T.-L., Lin, T.Ch., Wang, L.-Ch.,
Chao, H.-R., 1996. PAH emissions from gasoline-powered
engine. J. Environ. Sci. Health A 3 (8), 1981±2003.
Miyabara, Y., Hashimoto, S., Sagai, M., Morita, M., 1999.
PCDD and PCDF in vehicle exhaust particles in Japan.
Chemosphere 39, 143±150.
The contents of PCB and PAH in oils increase with
the time of their operation.
The discussion about possible e€ects in¯uencing the
PCB, PCDD, and PCDF formation applied to the
combustion conditions in cars is outlined.
The e€ect of oil on PAH, PCB, and PCDD/F forma-
tion is of general interest and should be therefore
studied to obtain more data on simpli®ed systems un-
der more de®ned conditions.
Nisbet, J.C., La Goy, P.K., 1992. Toxic equivalency factors
(TEFs) for polycyclic aromatic hydrocarbons. Reg. Tox.
Pharm. 16, 290±300.
·
Oehme, M., Larssen, S., Brevik, E.M., 1991. Emission factors
of PCDD and PCDF for road vehicles obtained by tunnel
experiment. Chemosphere 23, 1699±1708.
ꢁ
Parma, Z., Novak, L., 1994. Monitoring of POPs in Prague
territory (in Czech). Report of Mega Ltd., Straz pod
Acknowledgements
ꢁꢀ
Ralskem and of the Institute of Municipal Informatics
(IMIP) Prague.
The authors thank Petr Hapala (District Public
ꢁ
ꢁ
ꢀ
ꢀ ꢁ
Pekarek, V., Het¯ejs, J., Kastanek, F., 1995. Polychlorinated
dibenzo-p-dioxins and dibenzofurans in combustion pro-
cesses. Chem. Listy 89, 343±353.
ꢁ
ꢁ
ꢁ
Health Department, Frydek-Mõstek), Vladimõr Volak
ꢀ
ꢀꢁ
(Motor Vehicle Research Institute, Prague), Jirõ Drahos
(Institute of Chemical Process Fundamentals, Prague),
Schwind, K.H., Thoma, H., Hutzinger, O., Dawidowski, N.,
Weberuss, U., Hagenmaier, H., Buehler, U., Greiner, R.,
Essers, U., Bessey, E., 1991. Emission halogenierter dib-
enzodioxine und dibenzofurane aus verbrennungsmotoren.
ꢁ
and Milan Petrak (Research Institute of Inorganic
ꢁ
ꢁ
Chemistry, Ustõ nad Labem), the Directors of the in-
stitutions for their support of the project.
}
Z. Umweltchem. Okotox. 3, 291±298.
ꢀ
ꢀ
Sebor, G., Pospõsil, M., Maxa, D., Blazek, J., Cerny, J.,
ꢁꢀ
Vanura, J., 1994. Emissions from combustion of gasolines.
ꢀ
ꢁ
References
ꢀ
Project MZP CR No. 1677±94, Czech Republic (in Czech).
ꢀ
ꢀ
ꢀ
ꢁ
ꢁꢀ
ꢀ
Sebor, G., Kozak, P., Pospõsil, M., Blazek, J., 1995. Properties
of gasolines and their e€ects on the environment. Chem.
Listy 89, 233±244 (in Czech).
Bingham, A.G., Edmunds, C.J., Graham, B.W.L., Jones, M.T.,
1989. Determination of PCDDs and PCDFs in car exhaust.
Chemosphere 19, 669±673.