Evolution of products in the combustion of scrap tires in a horizontal, laboratory scale reactor

Article abstract of DOI:10.1021/es990883y

A horizontal laboratory reactor was used to study the evolution of byproducts from the combustion of scrap tires at five nominal temperatures (ranging from 650 to 1050 °C) and different oxygen:sample ratios A model was used to calculate the bulk air ratio (λ), and the oxygen consumption was discussed considering this ratio λ. More than 100 volatile and semivolatile compounds were identified and quantified by gas chromatography mass spectrometry, plotting their yields vs the bulk air ratio and temperature. Five different behaviors considering the bulk air ratio and the temperature were identified.

Full text of DOI:10.1021/es990883y

Environ. Sci. Technol. 2000, 34, 2092-2099
Evolution of Products in the
TABLE 1. Ultimate Analysis of the Tire Used in This Work
% wt
% wt
Combustion of Scrap Tires in a
C
H
N
S
83.5
7.8
0.39
1.5
O (by diff)
volatile m atter
fixed carbon
ash
6.8
61.2a
31.5a
7.3a
Horizontal, Laboratory Scale Reactor
A . F U L L A N A , * R . F O N T ,
J . A . C O N E S A , A N D P . B L A S C O
a
Calculated by Conesa et al. 1998 (18).
Department of Chemical Engineering, University of Alicante,
Apartado 99, Alicante, Spain
A horizontal laboratory reactor was used to study the
evolution of byproducts from the combustion of scrap tires
at five nominal temperatures (ranging from 650 to 1050
°C) and different oxygen:sample ratios. A model was used
to calculate the bulk air ratio (λ), and the oxygen
FIGURE 1. Diagram of the furnace used in combustion experiments.
consumption was discussed considering this ratio λ.
More than 100 volatile and semivolatile compounds were
identified and quantified by gas chromatography mass
spectrometry, plotting their yields vs the bulk air ratio and
temperature. Five different behaviors considering the
bulk air ratio and the temperature were identified.
They analyzed several volatile hydrocarbon compounds and
discussed the evolution of their yields with cracking severity.
Now the use of scrap tires for fuel offers the best alternative
for reusing rubber. Several industries use this waste to replace
traditional fuels. Thus, in 1996, 27 cements kilns and 12 paper
mills used tire-derived fuel (TDF) as a supplement fuel (8).
On the other hand, TDF was used in a dedicated tire-to-
energy plant.
The main problem of using TDF is the air contamination
effects. Although there are not many papers on the subject,
some have studied scrap tire combustion where the pollutants
are analyzed in great detail.
In this paper, the pollutants resulting from the combustion
of scrap tires have been analyzed at different conditions.
More than 100 compounds have been identified and quanti-
fied including light hydrocarbons (methane, acetylene, ...),
monoaromatic compounds (benzene, toluene, xylene, ...),
polyaromatic hydrocarbons (naphthalene, phenathrene,
pyrene, ...), partially oxygenated compounds (dibenzofuran,
naphthol, ...) and other heterocyclic compounds.
Introduction
As a result of the great increase in the use of automobiles,
the destination of scrap tires has been converted into a social
problem of great importance. The production of scrap tires
is valued at 2.5 million tons a year in the U.S.A., 2.0 millions
tons in the European Union, and 0.5 millions tons in Japan
(1).
Disposing of scrap tires in landfills is the most common
solution for this type of waste. The major problem of using
this method is that the scrap tires take up considerable space
because they are not easily compacted, and furthermore the
openings expose the landfills to insects, rodents, and birds
(2). On the other hand, the holes between pieces of tires
permit rainwater to pass that may produce an important
leaching of wastes. Another problem of landfills with scrap
tires is the high risk of fire. Uncontrolled fire may produce
large amounts of toxic (3) and mutagenicity compounds (4).
Recycling tires has been extended to use in asphalt, fuel,
and pyrolysis. The high net calorific power (about 8500 kcal/
kg) converts the tires in appropriate waste for obtaining fuel
by pyrolysis. Pyrolysis of scrap tires is an expensive process
and markets are limited. Another possibility of utilization of
the high net calorific power of tires is the use of the whole
tires as fuel.
Several authors have studied the pyrolysis of tires: (1)
Teng et al. (5) studied the pyrolysis of scrap tires from the
point of view of manufacturing active carbon. (2) Conesa et
al. (6) studied the primary decomposition by thermogravi-
metric analysis. They found that a tire is composed of three
decomposable fractions from the point of view of pyrolytic
decomposition. The same authors checked that the first
fraction could be a mineral oil, the second natural rubber,
and the third synthetic rubber (styrene-butadiene). (3) Conesa
et al. (7) also studied the composition of the gas obtained
from the pyrolysis of scrap tires in a fluidized bed reactor.
Equipment and Experimental Procedures
Material. The original tire (DUNLOP SP LE MANS) was
shredded into small pieces with a mean diameter of 5 mm.
These pieces were then immersed in liquid nitrogen and
were ground in a grinder to a diameter less than 1 mm.
The ultimate analysis of the tires is presented in Table 1.
The equipment used for this analysis was a Perkin-Elmer
2400 oven with a temperature of 1020 °C. (The combustion
was carried out in a pure oxygen atmosphere. Organic
compounds were totally decomposed to CO2, SO2, N2, and
water by a flash combustion and post catalytic oxidation.
The amounts of these compounds were determined by gas
chromatography; the standard used was sulfanilamide, and
the sample size was 2 mg.)
Details on the pyrolytic decomposition can be found
elsewhere (6, 7).
Pyrolysis Furnace. A horizontal quartz reactor was used
(Figure 1), where the sample was introduced inside a boat
that moved toward the inside of the reactor. The volatiles
evolved were cracked and burnt inside the reactor at the
second part of the reactor. The sample in the small boat was
introduced at constant velocity (0.5 mm/ s) into the furnace,
whose temperature profile was previously determined. Figure
2 shows the temperature profile corresponding to the nominal
temperature 850 °C.
* Corresponding author phone: +34-965903400 ext. 3003; fax:
+34-965903828; e-mail: andres.fullana@ua.es.
9
2 0 9 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 11, 2000
10.1021/es990883y CCC: $19.00
2000 Am erican Chem ical Society
Published on Web 04/21/2000

FIGURE 2. Temperature profile in furnace (nominal temperature
850 °C).
FIGURE 3. CO and CO yields vs the bulk air ratio.
2
TABLE 2. Experimental Conditions in the Combustion Scrap
Tire Using a Horizontal Laboratory Scale Reactor
The volatiles evolved were collected in a Tedlar bag for
analysis of light hydrocarbon compounds and a small tube
containing the resin (XAD-2) was used for collecting the
semivolatile compounds.
nominal
temp
mass of scrap
tire (mg)
mass flow
(mg/s)
bulk air
ratio (λ)
run no.
Analytical Method. The gases collected in the Tedlar bag
were analyzed by gas chromatography. The capillary Alumina-
KCL PLOT column with the FID detector was used to analyze
the light hydrocarbon compounds: methane, ethane, eth-
ylene, propylene, butane, acetylene, benzene, toluene, etc.
On the other hand, the Carbosieve S2 and Mol Sieve 5
columns with TCD were used for the analysis of O2, N2, CO2,
and CO.
The semivolatiles adsorbed in the XAD-2 resin (around
3 g) were extracted with 100 mL of dichloromethane, following
the EPA 3540c method. An internal deuterate standard (1,4-
dichlorobenzene-d4, naphthalene-d8, acenaphthene-d10, phen-
athrene-d10, chrysene-d12, and pyrene-d12) was added before
the extraction. The deuterate compounds were used to
estimate the evolution yields of the byproducts, and then
the extract was concentrated to 2 mL with a micro-Kuderma-
Danish equipment.
Semivolatile organic compounds were analyzed by gas
chromatography/ mass spectrometry (GC/ MS). The capillary
column DB-5 MS was used, and the compounds were
identified by MS (15-450 amu, 70 eV). The qualitative
identification of compounds was performed comparing the
mass spectrums with those of the NIST database. The index
from Lee and Kovad (9, 10) was also used to identify
compounds with a similar spectrum. A semiquantitative
estimation of the yields was obtained using factors obtained
for the deuterate internal standards (EPA method 8270c
section 7.7.3). Note that the software of GC/ MS equipment
suggests a list of the possible compounds. In the identification
of the peaks of the chromatogram, the most probable
compound is taken except when the other indicators (Lee
and Kovad index, boiling point, and logical correlation in
chromatograms of different runs) also indicate that another
compound is the most probable. For a precise identification
and quantification of every single compound using the EPA
method 8270c, a standard for each compound would have
been required. This level of work was however considered
to be beyond the scope of this research.
p850-100
c850-100
c850-61
c850-40
c850-30
c850-21
c850-15
c650-100
c750-100
c950-100
c1050-100
850
850
850
850
850
850
850
650
750
950
1050
100
100
61
40
30
21
15
100
100
100
100
2.0
2.0
0
0.24
0.40
0.61
0.82
1.19
1.65
0.24
0.24
0.24
0.24
1.21
0.80
0.60
0.41
0.29
2.00
2.00
2.00
2.00
run with excess of oxygen). This relation is the inverse of
bulk equivalence ratio φ used elsewhere (11).
Table 2 presents the operating conditions of the 11 runs
performed at different temperatures and oxygen ratios.
The mass flow of waste was calculated considering that
the sample, which was uniformly distributed in a short
distance and burnt in a fixed front, so the rate of advance
of the boat can be related with the burning rate of the sample.
Under this assumption and using the ultimate analysis data,
the mass flow in the combustion process can be calculated;
these values are shown in Table 2.
The gas flow (air or nitrogen) used was 280-mL/ min at 25
°C and 1 atm. For this flow, the residence time estimated is
5-7 s inside the central part of the furnace, where the
temperature is above the nominal temperature minus 50 °C.
Results
The results are presented and discussed from two viewpoints.
On one hand, runs from 1 to 7 were programmed to estimate
the effect of the bulk air ratio λ. In this way, this ratio λ was
varied between 0 (pyrolysis experiment) and 1.39, while the
temperature was maintained at 850 °C. On the other hand,
in the experiments 7-12, the temperature is the variable
operating conditions, and the bulk air ratio is approximately
constant (λ ) 0.25).
Effect of Bulk Air Ratio (λ). Table 3 shows the yields of
volatile and semivolatile organic compounds detected in the
experiments at different bulk air ratio. A logical result of yields
can be observed, and an estimated error around 5% can be
considered as acceptable for these runs.
Index annotations with the method used in the quanti-
fication and identification have been included in the tables
where the compounds are shown.
Experim ental Procedure. The bulk air ratio λ (λ ) (mfuel
/
mair stoichiometric/ (mfuel/ mair actual) is an important parameter in
)
)
Glass wool in the second part of the reactor was used in
runs with λ g 0.4 to obtain a good air mixing.
Figure 3 shows the variation of the CO2 yield, observing
that this increases progressively with λ values as far as λ )
1, whereas the CO yield increases for λ less than 0.45, while
the combustion process, as it is an indicator of the quantity
of oxygen present in the process (λ ) 0 in the pyrolysis
processes, λ)1 when the oxygen present is the stoichiometric
necessary one for a complete combustion, and λ > 1 for the
9
VOL. 34, NO. 11, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 0 9 3

TABLE 3. Yields and Behavior of Volatile and Semivolatile Compounds for the Runs at Different Bulk Air Ratioe
run no.
p850-100
c850-100
c850-61
c850-40
c850-30
c850-21
c850-15
tem p (°C)
sam ple m ass (g)
850
0.1
120
0
850
0.1
120
0.24
0.559
850
0.060
73
0.40
0.633
850
0.040
48
0.61
0.734
850
0.030
36
0.82
0.768
850
0.020
25
1.19
1.01
850
0.015
18
1.65
1.04
m ass flow (m g/m in)
bulk air ratio (λ)
total out bulk ratio (λT
)
0
out
pyrolysis
combustion
p850-100
c850-100
c850-61
c850-40
c850-30
c850-21
c850-15
behavior type
Yields of Volatile Compounds (mg/100 mg Scrap Tire)
a
CO2
3.7
150
170
200
210
310
nd
0.005
nd
0.002
nd
0.005
nd
nd
0.0000
0.001
nd
320
nd
0.008
nd
0.007
nd
0.002
nd
nd
0.0000
0.001
0.001
nd
0.42
0.064
0.56
COa
1.9
8.3
25.8
6.6
0.032
4.2
0.045
0.84
0.007
nd
0.022
0.061
0.001
0.042
8.2
22.8
3.8
0.012
2.4
0.021
0.73
0.005
nd
20.3
2.5
15.2
1.6
m ethanea
ethanea
ethylenea
propanea
acetylenea
I
I
I
I
I
I
I
I
I
I
I
I
I
0.28
6.3
0.10
1.07
nd
0.012
nd
0.155
0.037
0.040
15
0.002
1.16
0.010
0.54
0.003
nd
0.012
0.012
0.019
0.014
2.6
0.003
0.89
0.006
0.38
nd
a
CS2
butanea
pentanea
propinea
butinea
nd
nd
nd
0.042
0.024
0.020
4.9
0.19
12.2
0.016
0.014
0.009
1.93
0.062
4.9
hexanea
benzenea
toluenea
nd
0.170
0.010
0.18
0.92
32.5
0.40
20.5
0.084
6.9
total light hydrocarbons
Yields of Semivolatile Organic Compounds (mg/kg Scrap Tire)
phenylethyneb
styreneb
940
4700
nd
250
1400
nd
260
910
nd
8
100
60
nd
14
350
13
13
300
650
5
36
110
52
88
82
550
33
63
63
170
250
20
75
59
560
81
80
22
64
nd
nd
nd
18
62
1200
77
33
nd
22
nd
nd
90
3
nd
110
nd
17
nd
15
480
270
21
120
150
nd
nd
120
16
73
6
nd
64
nd
nd
5100
nd
nd
nd
3
nd
190
64
49
nd
nd
250
550
nd
7
140
nd
nd
50
18
106
19
nd
67
nd
nd
nd
nd
nd
nd
5
I
I
II
2,3-butanediolc
benzaldehydec
benzonitrileb
nd
nd
420
290
nd
18
3000
nd
32
nd
87
nd
420
240
nd
36
1800
nd
32
nd
24
nd
nd
37
3900
500
17
50
39
nd
400
1300
130
160
630
2600
6
310
59
59
I
I
benzene, ethenylm ethyl-b
octane, 2,4,6-trim ethyl-b
1-hexanol, 2-ethylb
benzene, 1-propynyl-b
acetophenonec
II
III
I
II
III
II
I
O-isopropenyltolueneb
benzofuran, 2-m ethyl-c
benzene, 1,3-diethenyl-b
heptanoic acid, 2-ethylc
1H-indene, 1-m ethylenec
benzaldehyde, ethylb
naphthalenea,d
12
9
nd
16
29
6
nd
19
II
nd
56
16
49
III
I
I
III
I
II
II
I
5600
1600
nd
110
68
4600
320
18
13
140
nd
250
450
43
100
170
1300
nd
210
21
22
4000
180
29
nd
100
nd
100
310
40
77
95
1100
nd
280
13
18
290
76
benzo[b]thiopheneb,d
decanalc
isoquinolineb,d
benzo[b]thiophene, 6-m ethyl-b,d
decanoic acid, m ethyl esterc,d
naphthalene, m ethylb,d
biphenylb,d
nd
1100
5600
250
490
2700
12000
nd
I
I
I
acenaphthyleneb,d
acenaphthenea,d
18
21
390
nd
170
7
nd
67
150
3
naphthalene, ethenylb,d
biphenyleneb
nd
nd
41
nd
nd
nd
nd
230
22
nd
nd
nd
nd
nd
nd
6
nd
4
I
I
II
II
I
2-naphthalenolc,d
17
nd
nd
nd
nd
230
nd
nd
30
nd
nd
nd
nd
23
19
nd
dibenzofuranb,d
nd
naphthalene, 1,4,5-trim ethylb,d
phenaleneb,d
120
260
3600
250
nd
I
I
fluoreneb,d
1000
110
nd
370
31
nd
aliphatic hydrocarbon <20Cd
phenol, 4-allyl-2,6-dim ethoxy-c
biphenyl, 4-vinyl-b,d
benzophenoneb
5
II
I
I
I
I
II
I
I
I
I
260
220
260
250
nd
1200
4000
4700
2100
60
160
nd
66
nd
340
2900
890
440
nd
nd
6
16
nd
nd
9
12
16
88
1300
260
64
nd
nd
nd
nd
nd
49
440
85
37
1H-phenaleneb,d
1,2-dihydrophenanthreneb,d
fluoren-9-oneb,d
14
dibenzothiopheneb,d
phenanthrenea
170
1900
420
100
anthracenec
11
nd
naphthalene, 1-phenyl-b,d
9
2 0 9 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 11, 2000

TABLE 3 (Continued)
pyrolysis
combustion
p850-100 c850-100 c850-61 c850-40 c850-30 c850-21 c850-15 behavior type
Yields of Semivolatile Organic Compounds (mg/kg Scrap Tire)
1-heptadecanolb,d
nd
83
27
370
41
nd
630
17
50
450
140
4200
84
3
nd
130
nd
nd
170
190
28
160
4000
4400
30
550
nd
100
nd
60
93
270
nd
52
nd
280
nd
2100
nd
91
360
500
520
nd
520
80
110
nd
7
nd
II
I
1H-phenale-1-onec,d
180
1800
240
nd
2900
17
270
2100
270
10600
nd
benzo[def]fluoreneb
nd
nd
I
phenanthrene, 9-m ethyl-b,d
hexadecanoic acidc,d
nd
nd
I
57
nd
54
nd
II
I
2-phenylnaphthaleneb,d
aliphatic hydrocarbon 20-30b,d
9,10-dim ethylanthraceneb
fluorantheneb,d
2000
nd
2800
nd
5500
nd
II
I
150
1100
1400
260
180
78
20
23
I
2-m ethoxyphenantreneb
pyreneb,d
20
80
28
27
II
I
M-terphenylb
590
nd
120
140
nd
230
nd
26
160
nd
II
II
II
III
I
benzonaphthofuranb,d
decanedioic acid, dibutyl esterb
benzo[b]naphtho 2,3-furanb
benzo[c]phenantreneb
acetic acid, octadecyl esterc
11H-benzo[b]fluoreneb,d
2-eicosanoxyethanolb
9-octadecenam idec
nd
nd
nd
40
nd
nd
81
nd
59
60
230
nd
1800
nd
180
nd
730
nd
nd
190
nd
460
1900
nd
140
nd
580
49
nd
163
nd
455
1300
nd
nd
nd
81
59
760
1600
nd
nd
59
110
300
1500
6100
nd
790
nd
36
nd
II
I
69
7700
nd
6
II
II
I
nd
5100
nd
benzo[b]naphtho-2,3-d thiopheneb,d
fluorene, 9-phenyl-b,d
benzo[ghi]fluorantheneb
chrysenea,d
310
150
996
3800
nd
nd
nd
nd
I
321
130
nd
nd
nd
I
180
nd
1000
430
nd
nd
610
nd
dodecanoic acid, tetradecil esterc
2-m ethylchryseneb
III
I
110
650
nd
nd
270
48
nd
nd
nd
nd
nd
nd
nd
nd
nd
1,2′-binaphthaleneb
nd
nd
I
anthracene, 9-phenyl-b
benzo[j]fluorantheneb
perylenea
nd
nd
I
12000
5100
580
500
nd
2100
640
63
nd
nd
170
270
nd
nd
24
110
210
nd
nd
44
nd
nd
nd
nd
nd
nd
I
nd
nd
I
benzo[j]aceanthylene, 3-m ethylb
1,2:7,8-dibenzphenanthreneb
aliphatic hydrocarbon >)30b
benzo[ghi]peryleneb
nd
nd
I
nd
nd
I
31
1800
nd
8100
nd
III
I
3200
210
11000
380
nd
33000
77
nd
25000
nd
nd
25000
nd
nd
11900
dibenzo DEF, p-chrysenec
total sem ivolatile com pds
nd
nd
I
23000
22000
a
b
Authentic quantitative standard. Forward values (forward value ) 100 * ∑[(ILib‚IUnk)2]1/2/∑ILib‚∑IUnk where I ) refers to the intensity of the
lib
m ethod spectrum at a given m ass and IUnk) refers to the intensity of the “unknown” sam ple spectrum at a given m ass) larger than 90 and
quantification using internal standard. Forward values (see definition in footnote b) between 80 and 90 and quantification using internal standard.
Difference in Lee and Kovad’s index lesser than 1% respect index published in refs 9 and 10). Not detected: nd.
c
d
e
it decreases for λ above 0.45. CO was not detected for values
of λ above 1. These yields of CO2 and CO show that the
assumption made in the calculation of the bulk air ratio was
acceptable.
is formed (see Figure 4c) by aliphatic hydrocarbons and other
amide compounds. The formation of these compounds is
favored by the presence of air, and it seems that they are
quite stable vs oxygen at 850.
Effect of Tem perature. Table 4 shows the total ap-
proximate yields of volatile and semivolatile organic com-
pounds detected in the runs at different temperatures with
λ equal to 0.25.
Table 3 also shows the behavior of the different com-
pounds detected. The behavior of the volatiles or light
hydrocarbons is the expected one, with yields decreasing as
λ increases. A fact to note is that the benzene and the toluene
are more resistant to oxygen atmosphere than the other
semivolatile compounds.
Two different behaviors when varying the temperature
can be found: Type A or compounds whose yields decrease
with temperature. This group is formed principally by
monoaromatic hydrocarbons and partiality oxidized com-
pounds. Figure 5a shows the yields of some compounds type
A vs the temperature. Type B or compounds with a maximum
yield. This group is formed by the light hydrocarbons with
greater yields (benzene, toluene, methane, ethylene, and
acetylene) and the PAHs. It can be observed (see Figure 5b)
that the yields of these compounds have maximums at
intermediate temperatures (750-850 °C).
Semivolatile compounds have been classified according
to their behavior against λ. Three types were found (see Table
3): I. Compounds whose yields decrease as the bulk air ratio
(λ) increases (see Figure 4a): monoaromatic compounds and
polyaromatic hydrocarbons (PAHs). Note that all compounds
of this group are formed by pyrolysis (λ ) 0). II. Compounds
whose yields increase as λ increases for low λ values and
decrease for very high λ values (see Figure 4b). These
compounds are partially oxidized: alcohols, organic acids,
furans and some amides; the total amount of compounds in
this group is lower than the amount in group I. The
compounds of this group are not formed or one is formed
with very low yields by pyrolysis, but their presence is favored
at intermediate bulk air ratio; the yields decrease at high
values of λ due to the oxidation or combustion reactions. III.
Compounds whose yields increase as λ increases. This group
Discussion
The total oxygen consumed in the combustion runs at
different λ was calculated using the yields of the compounds
analyzed (Table 3). Water and sulfur dioxide yields were
estimated also using the ultimate scrap tire analysis. A new
parameter for batch combustion can be defined: the total
9
VOL. 34, NO. 11, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 0 9 5

TABLE 4. Yields and Behavior of Volatile and Semivolatile Compounds for the Runs at Different Temperaturese
run no.
c650-100
c750-100
c850-100
c950-100
c1050-100
tem p (°C)
650
750
850
950
1050
m ass (g)
0.100
0.100
0.100
0.100
0.100
c650-100
c750-100
c850-100
c950-100
c1050-100
behavior type
Yields of Volatile Compounds (mg/100 mg Scrap Tire)
a
CO2
COa
125
119
149
24
3.7
0.37
3.9
0.044
1.1
0.51
0.053
nd
0.52
0.005
0.30
0.34
0.029
0.47
0.006
4.5
15
4.7
0.25
4.2
0.005
0.44
0.45
0.040
0.006
0.06
0.002
0.370
nd
26
6.6
m ethanea
ethanea
5.5
nd
0.94
nd
0.008
1.2
0.003
nd
nd
1.3
nd
B
A
B
A
A
B
A
A
A
A
A
A
A
A
A
B
A
0.032
4.20
nd
0.047
0.84
0.007
nd
0.001
nd
0.022
0.061
nd
0.001
0.042
8.2
0.40
20.5
ethylenea
propanea
propylenea
acetylenea
0.06
nd
0.001
0.38
0.001
nd
nd
nd
0.004
nd
nd
0.003
0.001
0.60
0.005
2.3
a
CS2
butanea
butylenea
1,3-butadienea
pentanea
nd
0.015
0.009
nd
0.020
0.013
5.5
propynea
pentenea
nd
butynea
0.22
0.004
5.6
0.90
17.2
hexanea
benzenea
toluenea
1.0
16.7
0.026
13.2
total light hydrocarbons
Yields of Semivolatile Organic Compounds (mg/kg Scrap Tire)
ethylbenzeneb
phenylethyneb
styreneb
800
1600
4900
140
1800
1000
4800
240
2400
630
65
45
nd
nd
300
320
nd
nd
nd
nd
120
nd
nd
65
14
77
84
nd
nd
84
nd
nd
25
34
52
81
nd
0
nd
26
nd
nd
nd
nd
139
840
70
20
nd
nd
0
nd
nd
nd
5
nd
38
0
nd
nd
500
nd
nd
nd
nd
nd
nd
23
39
A
A
A
A
A
A
A
A
A
A
- -
A
A
A
A
A
A
A
A
A
970
5200
nd
150
120
2400
580
1700
510
53
250
1400
nd
nd
nd
benzene, ethylm ethyl-b
benzaldehydec
m esityleneb
phenolb
nd
benzonitrileb
420
240
nd
benzene, ethenylm ethyl-b
benzene, 1-ethenyl-3-m ethylb
1-hexanol, 2-ethylb
36
benzene, 1-propynyl-b
benzene, 1-ethyl-2,3-dim ethyl-b
acetophenonec
1700
130
360
2400
240
84
123
100
410
26
2800
nd
28
55
170
34
54
58
150
8
1800
nd
0
nd
32
nd
nd
24
nd
290
nd
0
nd
72
nd
nd
nd
nd
23
phenol, 2-m ethyl-b
O-isopropenyltolueneb
styrene, 2,5-dim ethyl-b
benzofuran, 2-m ethyl-c
benzene, 1,3-diethenyl-b
1H-indene, 1-m ethylenec
benzaldehyde, ethylb
naphthalenea,d
37
600
140
21
44
38
680
200
21
18
14
3900
500
17
nd
nd
2200
230
17
nd
nd
0
B
B
benzo[b]thiopheneb,d
decanalc
naphthalene, 1,2-dihydro-6-m ethyl-b
thieno[3,2-c]pyridinec
isoquinolineb,d
A
A
B
A
B
A
A
A
B
B
B
B
B
A
B
A
A
A
B
B
B
3
16
50
naphthalene, 1,2-dihydro-2-m ethyl-b
benzo[b]thiophene, 6-m ethyl-b,d
naphthalene, m ethylb,d
naphthalene, 1-m ethylb,d
1H-indene, 1-ethylidene-
biphenylb,d
5
13
470
340
10
340
120
110
290
409
140
140
200
310
190
70
nd
40
nd
39
nd
400
nd
1300
130
nd
630
2600
160
nd
6
310
59
nd
nd
nd
27
nd
370
24
500
350
12
510
200
480
38
890
58
5900
79
340
100
150
820
21
acenaphthyleneb,d
naphthalene,1,7-dim ethylb,d
naphthalene, ethenyl-b,d
biphenyleneb
nd
34
1300
nd
nd
nd
26
acenaphthenea,d
biphenyl, 4-m ethyl-b,d
2-naphthalenolc,d
dibenzofuranb,d
naphthalene, 1,4,5-trim ethyl-b,d
phenaleneb,d
12
59
1000
110
nd
140
17
fluoreneb,d
390
23
aliphatic hydrocarbon <20
9
2 0 9 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 11, 2000

TABLE 4 (Continued)
c650-100
c750-100
c850-100
c950-100
c1050-100
behavior type
Yields of Semivolatile Organic Compounds (mg/kg Scrap Tire)
biphenyl, 4-vinyl-b,d
nd
42
nd
nd
60
nd
nd
nd
nd
nd
nd
nd
nd
nd
6
nd
nd
18
320
24
8
nd
nd
nd
nd
nd
nd
nd
29
0
nd
34
26
840
nd
nd
nd
nd
7
nd
nd
nd
nd
120
53
nd
nd
nd
nd
nd
nd
nd
44
77
nd
nd
66
3800
B
A
B
A
B
A
B
A
A
B
B
B
B
B
A
B
B
A
A
B
A
A
A
B
B
B
B
B
A
A
B
B
A
A
B
B
B
A
A
A
B
A
B
B
B
B
B
B
B
acridine, 9,10-dihydro-c
benzophenoneb
nd
85
nd
39
160
nd
nd
nd
9H-xantheneb
1H-phenaleneb,d
nd
10
11
nd
nd
nd
nd
nd
1,1′-biphenyl, 2,4′-dim ethyl-
1,2-dihydrophenanthreneb,d
9H-fluorene, 1-m ethyl-b
fluoren-9-oneb,d
41
13
66
16
76
95
nd
nd
140
130
560
170
270
13
74
nd
17
dibenzothiopheneb,d
phenanthrenea,d
160
470
370
640
17
340
2900
890
440
83
140
1900
nd
45
49
anthracenec,d
naphthalene, 1-phenyl-b,d
1-heptadecanolb,d
1H-phenale-1-onec,d
190
49
50
150
16
nd
79
180
170
19
27
370
nd
41
nd
nd
180
nd
nd
nd
benzo[def]fluoreneb
4H-cyclopenta[def]phenenthreneb
phenanthrene, 9-m ethyl-b,d
benzylnaphthalenec
2-phenylnaphthaleneb,d
aliphatic hydrocarbon 20-30
9,10-dim ethylanthraceneb
naphthalene, 1-8-di-1-propynylc
fluorantheneb,d
330
78
290
60
630
17
190
79
67
98
50
13
0
17
nd
nd
150
82
300
50
230
110
600
68
450
140
4200
84
210
71
3100
nd
2-m ethoxyphenantreneb
pyreneb,d
M-terphenylb,d
benzonaphthofuranb,d
naphthalene, 2-benzyllc
benzo[b]naphtho 2,3-furanb
benzo[c]phenantrenec
11H-benzo[b]fluoreneb,d
2-eicosanoxyethanolb
5,6-dihydrochryseneb
benzo[b]naphtho-2,3-d thiopheneb,d
benzo[ghi]fluorantheneb
chrysenea,b
25
8
82
109
260
35
33
5
94
100
56
69
15
nd
85
29
13
500
60
32
5
14
240
700
18
43
93
320
250
41
63
24
490
64
130
61
630
220
63
40
nd
nd
180
730
nd
nd
nd
nd
31
130
nd
nd
nd
190
460
1900
nd
nd
nd
56
330
590
nd
nd
nd
nd
nd
46
nd
740
370
nd
nd
370
14000
benzanphrenoneb
bis(ethylhexyl)phatalateb
triphenylene, 1-m ethyl-c
dodecanoic acid, tetradecil esterc
2-m ethylchryseneb
nd
nd
1,2′-binaphthaleneb
260
48
2100
640
63
nd
380
33000
anthracene, 9-phenyl-b
benzo[j]fluorantheneb
perylenea
benzo[j]aceanthylene, 3-m ethylb
1,2:7,8-dibenzphenanthreneb
benzo[ghi]peryleneb
total sem ivolatile com pounds
17
nd
nd
33000
39
nd
34000
a
b
Authentic quantitative standard. Forward values (forward value ) 100 * ∑[(ILib‚IUnk)2]1/2/∑ILib‚∑IUnk where I ) refers to the intensity of the
lib
m ethod spectrum at a given m ass and IUnk) refers to the intensity of the “unknown” sam ple spectrum at a given m ass) larger than 90 and
quantification using internal standard. cForward values (see definition in footnote b) between 80 and 90 and quantification using internal standard.
d
e
Difference in Lee and Kovad’s index lesser than 1% with respect to index published in refs 9 and 10. Not detected: nd.
out bulk air ratio (λTout) that equals the ratio between the
total oxygen consumed and total stoichiometric oxygen. A
value of λTout ) 1 would mean that only stoichiometric oxygen
was consumed in the run. Note that λ represents the
instantaneous ratio between the stoichiometric and actual
only on the total oxygen fed during the run and does not
depend on the instantaneous amount of oxygen (related to
λ) because the nonvolatile matter does not leave the solid if
no reaction with the oxygen occurs. A theoretical λT was
out
calculated from a mass balance for the two ways assuming
that all oxygen available reacted with the volatile and
semivolatile compounds and that the nonvolatile fuel is
composed only by carbon.
oxygen and that λT is related to the total oxygen fed.
out
In the batch combustion two possible ways can be
considered. One way is that the combustion takes place in
one step only with the volatile and nonvolatile matter burning
at the same time as the sample is introduced into the furnace.
Another possibility is a process with two steps: in the first
step the volatile matter (volatile and semivolatile compounds)
burns at the same velocity as the sample is introduced into
the furnace, and in the second step the nonvolatile matter
is burned. The combustion of nonvolatile matter depends
Figure 6 shows the relation λT vs λ. It can be seen that
out
the experimental values present an intermediate situation
between the two ways: at low values of λ, the values λT
out
approach the first way because the high production of volatile
compounds does not allow the oxygen to reach the non-
volatile matter. When λ is close to 1, oxygen is available for
nonvolatile fuel combustion and the second mechanism is
9
VOL. 34, NO. 11, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 0 9 7

FIGURE 5. (a) Some type A compounds and (b) some type B
compounds.
FIGURE 4. (a) Some type I compounds, (b) some type II compounds,
and (c) some type III compounds.
more important. For λ over 1 the total oxygen consumed is
equal to the stoichiometric oxygen; therefore λTout ) 1. Values
of λTout greater than unity are only due to experimental errors.
Figure 4a shows the yield vs bulk air ratio λ for some type
I compounds. A strong decreasing behavior was also observed
for monoaromatic and polyaromatic compounds for 0.5 bulk
air ratio, and the yields decrease more than five times in
these types of compounds. The strong decreasing behavior
indicates that this group is very reactive with the oxygen.
Only the most resistant compounds (naphthalene, phenan-
threne, ..., etc.) maintain their yields, due probably in part
to the fact that the oxygen is limited and reacts with the less
stable compounds. All volatile compounds (or light hydro-
carbons) have this behavior similar to the type I semivolatile
compounds. Note that these compounds, formed by pyrolysis
and cracking of tars, logically compress from light hydro-
carbons and to structures of high molecular weight as a
consequence of cracking, addition, and cyclation reactions.
FIGURE 6. Relation between oxygen fed and oxygen consumption.
Levendis et al. (12) analyzed the effects on φ in PAH yields
from tire waste. They found that PAH production showed an
asymptotic behavior as φ approached zero (i.e. λ ) ∞). They
suggested an exponential relationship between φ and the
level of PAH detected. This result is in accordance with the
behavior of type I compounds described here.
With respect to the type II compounds (Figure 4b), we
have not found any paper describing this behavior. Only a
few compounds (dibenzofuran, naphthol) were identified in
some papers but without describing the behavior (3, 11-13).
The relative high yield of type II compounds found in this
work could be due to slow cooling in the furnace (see Figure
2) that can promote the formation of the oxygenated
compounds. The presence of dibenzofuran in this group must
be noted, although in low yields, around 280 mg/ kg.
9
2 0 9 8 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 11, 2000

Group III (see Figure 4c) has the most stable compounds
vs the oxygen ratio. The compounds forming this group are
low toxicity ones. The behavior at very high bulk air ratio
cannot be deduced. It is emphasized that these compounds
were detected when operating at 850 °C, and it is possible
that at other temperatures their behavior could be different.
Both pyrolysis behaviors (A and B) when considering the
runs with λ constant and different temperatures are a result
of the competitive process between pyrosynthesis and
cracking. If cracking is greater than pyrosynthesis at all
temperatures, the compound would have a type A behavior.
In type B, pyrosynthesis is more important than cracking at
low temperatures but not at high temperatures. Figure 5a
presents some compounds with type A behavior, and the
compounds with type B behavior are represented in Figure
5b.
industrial combustors, where “puffs” can be formed and do
not mix perfectly with excess air. The presence of some groups
of compounds identified in this work, in industrial combus-
tors, can be useful for characterizing the combustion
(temperature, residence time of volatiles, and mixing with
air).
Acknowledgments
Support for this work was provided by CYCIT-Spain, Research
project AMB96-1076. Support for a part of the work, corre-
sponding to the degree thesis of A. Fullana was also provided
by Instituto de Cultura JUAN GIL-ALBERT.
Literature Cited
(1) Cunliffe, A. M.; Williams P. T. J. Anal. Appl. Pyrol. 1998, 44, 131.
(2) EPA (Environmental Protection Agency). 530-SW-90-074B;
October 1991.
Using batch pyrolysis of tires, Cunliffe and Williams (1)
found that the concentration of PAHs increases from 1.5 to
4.4 wt % in pyrolytic oils when the temperature increases
from 450 to 600 °C. These values are in accordance with type
II behavior for low temperatures found in this paper.
Durlank et al. (14) analyzed the PAH production from
polystyrene incineration in a vertical tubular flow furnace.
They showed that the PAH yield in the gas and particulate
phases as well as the number of PAH species and total mass
yields increased when the temperature was increased from
800 to 1200 °C. These values are in accordance with type II
behavior for high temperatures.
Mastral et al. (15, 16) analyzed the temperature influence
on PAH emissions in the fluidized bed combustion of coal.
They found that the total PAH had a maximum yield at 800
°C according to values in group II detected in the present
work.
Bofanti et al. (17) found a behavior for the PAH yields
with the temperature in the pyrolysis of coal apparently
different to results found in this work. They showed a PAH
increase with the temperature in a pyroprobe between 250
and 1000 °C. This behavior could be due to the very slow
residence time in the pyroprobe reactor, and consequently
cracking did not occur significantly. Cracking is responsible
for the maximum yield PAHs in the runs of the present work.
Atal (11) demonstrated, adding deuterate PAH to scrap tires,
that the PAH yields are a consequence of pyrosynthesing
and cracking and/ or burning in the reactor.
(3) Lemieux, P.; Ryan, J. Air Waste Manage. Assoc. 1993, 43, 1106.
(4) DeMarinin, D.; Lemieux, P.; Ryan, J.; Brooks, L.; Williams, R.
Environ. Sci. Technol. 1994, 28, 136.
(5) Teng, H.; Serio, M. A.; Wo´ jtowicz, M. A.; Bassilakis, R.; Solomon,
P. R. Ind. Eng. Chem. Res. 1995, 34, 3102.
(6) Conesa, J. A.; Marcilla, A. J. Anal. Appl. Pyrol. 1996, 37, 95.
(7) Conesa, J. A.; Font, R.; Marcilla, A. Energy Fuel 1996, 10(1), 134.
(8) Snonagel, P. Tires. In Ullman’s Encyclopedia of Industrial
Chemistry; VCS Verlagsgesellschaft:1996.
(9) Lee, M.; Vassilaros, D.; White, C.; Novotny, M. Anal. Chem. 1976,
51 (6), 768.
(10) Pereira, V.; Kovad, A.; Lee, M. J. High Resolut. Chromatogr. 1986,
9, 328.
(11) Atal, A.; Levendis, Y. A.; Carlson, J.; Dunayevskiy, Y.; Vouros, P.
Comb. Flame 1997, 110, 462.
(12) Levendis, Y. A.; Atal, A.; Carlson, J.; Dunayevskiy, Y.; Vouros, P.
Environ. Sci. Technol. 1996, 30, 2742.
(13) Levendis, Y. A.; Atal, A.; Courtemanche, B.; Carlson, J. B.Combust.
Sci. Technol. 1998, 131, 147.
(14) Durkland, S. K.; Biswas, P.; Shi, J.; Bernhard, M. J. Environ. Sci.
Technol. 1998, 32, 2301.
(15) Mastral, A. M.; Calle´n, M.; Murillo M. Fuel 1996, 75, 1533.
(16) Mastral, A. M.; Calle´n, M.; Murillo, M.; Garc´ıa, T. Fuel 1998, 77,
1513.
(17) Bonfanti, L.; Comellas, Ll.; Lliberia, J. Ll.; Vallhonrat-Matalonga,
R.; Pich-Santacana, M.; Lo´pez-Pin˜ ol, D. J. Anal. Appl. Pyrol 1997,
44, 89.
(18) Conesa, J. A.; Font, R.; Fullana, A.; Caballero, J. A. Fuel 1998,
77(13), 1469.
Received for review August 2, 1999. Revised manuscript
received February 21, 2000. Accepted February 29, 2000.
The toxic products identified in this paper, considering
bulk air ratios less than unity, are probably present in
ES990883Y
9
VOL. 34, NO. 11, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 0 9 9