Characterization of polycyclic aromatic hydrocarbon particulate and gaseous emissions from polystyrene combustion

Article abstract of DOI:10.1021/es9709031

The partitioning of polycyclic aromatic hydrocarbons (PAHs) between the particulate and gaseous phases resulting from the combustion of polystyrene was studied. A vertical tubular flow furnace was used to incinerate polystyrene spheres (100-300 μm) at different combustion temperatures (800- 1200 °C) to determine the effect of temperature and polystyrene feed size on the particulate and gaseous emissions and their chemical composition. The furnace reactor exhaust was sampled using real-time instruments (differential mobility particle sizer and/or optical particle counter) to determine the particle size distribution. For chemical composition analyses, the particles were either collected on Teflon filters or split into eight size fractions using a cascade impactor with filter media substrates, while the gaseous products were collected on XAD-2 adsorbent. Gas chromatography/mass spectroscopy (GC/MS) was used to identify and quantify the specific PAH species, their partitioning between the gas and particulate phases, and their distribution as a function of emission particle size. The total mass and number of PAH species in both the particulate and gas phases were found to decrease with increasing incineration temperature and decreasing polystyrene feed size, while the mean diameter of the particles increases with increasing incineration temperature and decreasing feed size. In addition, the PAH species in the particulate phase were found to be concentrated in the smaller aerosol sizes. The experimental results have been analyzed to elucidate the formation mechanisms of PAHs and particles during polystyrene combustion. The implications of these results are also discussed with respect to the control of PAH emissions from municipal waste-to-energy incineration systems. The partitioning of polycyclic aromatic hydrocarbons (PAHs) between particulate and gaseous phases resulting from the combustion of polystyrene was studied. A vertical tubular flow furnace was used to incinerate polystyrene spheres to determine the effect of temperature and polystyrene feed size on the particulate and gaseous emissions and their chemical composition. The furnace reactor exhaust was sampled using real-time instruments to determine the particle size distribution. The total mass and number of PAH species in both the particulate and gas phases were found to decrease with increasing incineration temperature and decreasing polystyrene feed size, while the mean diameter of the particles increases with increasing incineration temperature and decreasing feed size. In addition, the PAH species in the particulate phase were found to be concentrated in the smaller aerosol sizes.

Full text of DOI:10.1021/es9709031

Environ. Sci. Technol. 1998, 32, 2301-2307
cadmium, lead, and mercury are also of significant concern
Characterization of Polycyclic
Aromatic Hydrocarbon Particulate
and Gaseous Emissions from
Polystyrene Combustion
in municipal solid waste (MSW) incineration (1, 2), trace
organic products such as polycyclic aromatic hydrocarbons
(PAHs) from incomplete combustion are of special concern
due to the potential carcinogenicity for some of these
compounds. A major source of PAHs from MSW incinerators
is high-volume plastics (3-8), which are primarily comprised
of polyethylene (65 wt %), polystyrene (22%), polypropylene
(9%), and PVC (4%) (3). Previous incineration studies have
indicated that combustion of polystyrene leads to a signifi-
cantly larger amount of PAHs than the other plastics,
probably due to its aromatic structure (4-8). A better
understanding of the chemical and physical processes that
are responsible for PAH emissions from polystyrene com-
bustion will lead to better methods for postcombustion
control and, potentially, methods to reduce the formation of
these compounds in combustion systems.
S U S A N K . D U R L A K A N D P R A T I M B I S W A S *
Aerosol and Air Quality Research Laboratory, Department of
Civil and Environmental Engineering, University of
Cincinnati, Cincinnati, Ohio 45221-0071
J I C H U N S H I A N D M A R Y J O B E R N H A R D
Environmental Science Department, The Procter & Gamble
Company, 5299 Spring Grove Avenue,
Cincinnati, Ohio 45217-1087
The combustion of polystyrene is a complex and not
thoroughly understood process. This is especially true of
the PAH and soot particle formation mechanisms. When a
polymer is combusted, there is generally a two-stage py-
rolysis/ combustion process. First the polymer is heated,
releasing low molecular weight gaseous pyrolysis products
followed by higher molecular weight gaseous products. This
forms a gas cloud of pyrolysis products around the poly-
styrene sphere. Next, oxygen diffuses into the pyrolysis
product cloud, and a diffusion flame is established (9). In
the diffusion flame, soot (fine particulate matter) is formed
consisting of agglomerates of particle chains of 105-106
carbon atoms. It has been reported that polystyrene
produces a significant amount of fine particles (7, 8). Elomaa
and Saharinen (5) have reported that as much as 30-50% of
the original mass remains as particulate matter. While the
exact formation mechanism of soot in polystyrene combus-
tion is not well understood, several studies have examined
the development of soot monomers and aggregates in
hydrocarbon flames (10, 11). The development of soot in
the flame has been linked to PAH formation and destruction
in the flame, with PAH compounds forming early in the flame,
via condensation reactions with acetylene (10) or reactive
coagulation of smaller PAHs (11).
The partitioning of polycyclic aromatic hydrocarbons
(PAHs) between the particulate and gaseous phases resulting
from the combustion of polystyrene was studied. A vertical
tubular flow furnace was used to incinerate polystyrene
spheres (100-300 µm) at different combustion temper-
atures (800-1200 °C) to determine the effect of temperature
and polystyrene feed size on the particulate and gaseous
emissions and their chemical composition. The furnace
reactor exhaust was sampled using real-time instruments
(differential mobility particle sizer and/or optical particle
counter) to determine the particle size distribution. For
chemical composition analyses, the particles were either
collected on Teflon filters or split into eight size fractions
using a cascade impactor with filter media substrates, while
the gaseous products were collected on XAD-2 adsorbent.
Gas chromatography/mass spectroscopy (GC/MS) was
used to identify and quantify the specific PAH species, their
partitioning between the gas and particulate phases, and
their distribution as a function of emission particle size.
The total mass and number of PAH species in both the
particulate and gas phases were found to decrease with
increasing incineration temperature and decreasing poly-
styrene feed size, while the mean diameter of the particles
increases with increasing incineration temperature and
decreasing feed size. In addition, the PAH species in the
particulate phase were found to be concentrated in the
smaller aerosol sizes. The experimental results have been
analyzed to elucidate the formation mechanisms of PAHs
and particles during polystyrene combustion. The
When the PAHs have been built up to molecules with an
approximate molecular weight of 1000, or 100 carbon atoms,
soot inception occurs (10). Once soot nuclei are formed,
condensation reactions continue to occur at the nuclei
surface, resulting in growth to monomers between 0.01 and
0.05 µm (12). Once the soot reaches the monomer size,
further growth is by Brownian coagulation.
Previous combustion studies on polystyrene (4, 5, 7, 8)
have examined primarily the range of organic compounds,
including PAHs, that are formed at various combustion
temperatures. Panagiotou and Levendis (7) studied the
combustion of large (300 µm) and small (60 µm) polystyrene
spheres at 1000 °C using high-speed cinematography, and
they found that the small polystyrene spheres were consumed
much more quickly (6-7 ms) than the larger ones (120-160
ms). Wheatley et al. (4) examined the PAH emissions resulting
from combustion of several plastics and found significant
variation with temperature in the range of 750-1150 °C.
Although Wheatley et al. (4) did not differentiate between
the PAHs collected in the particulate and gas phases, other
studies (5, 8) have attempted to study partitioning, but
variations in the collection techniques introduced consider-
able uncertainties in the data. Hawley-Fedder et al. (8)
differentiated between the PAHs in the particulate and gas
implications of these results are also discussed with respect
to the control of PAH emissions from municipal waste-
to-energy incineration systems.
Introduction
Incineration is currently the focus of much research and
regulatory effort, due in part to the potentially toxic trace
organic compounds generated as products of incomplete
combustion. While emissions of heavy metals such as
* To whom correspondence should be addressed. Telephone:
(513)556-3697; fax: (513)556-2599; e-mail: pratim.biswas@uc.edu.
9
S0013-936X(97)00903-6 CCC: $15.00
Published on Web 06/04/1998
1998 Am erican Chem ical Society
VOL. 32, NO. 15, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 3 0 1

into particle size ranges of 250-300 and 104-175 µm.
Experiments were performed with these two size ranges of
polystyrene feed. The feed rate, on a mass basis, and airflow
rate were equivalent in both experiments, maintaining the
same fuel-to-air ratio. A microscope photograph of the sieved
particles is shown in Figure 2A, and a transmission electron
microscope (TEM) picture of the soot agglomerates collected
from the furnace exhaust is shown in Figure 2B.
The chemical analysis procedure used in this work is
similar to that used by Wheatley et al. (4). A 47-µm Teflon
filter (Gelman) was used to capture the particles, and XAD-2
adsorbent (housed in a Teflon tube approximately 10 cm ×
1 cm diameter and cooled with an ice bath) was used to
adsorb the gaseous combustion products downstream of the
Teflon filter. While there could be considerable uncertainty
in accurately assigning the PAH yield to the gas and particulate
phases, due to sampling artifacts, studies (15, 16) indicate
that using a Teflon filter can minimize the absorption of
gaseous compounds onto the filter. Gas-particle partitioning
results should be considered a rough estimate, but uncer-
tainty in the partitioning does not affect the accuracy of the
total PAH yield, which is the sum of the two phases.
Both the Teflon filters and XAD-2 adsorbent were spiked
with two internal standards phenanthrene-d10 and benzo-
[a]pyrene-d12; extracted for 17 h in a Soxhlet apparatus with
methylene chloride; concentrated in a rotary evaporator to
1 mL; and stored in sealed, refrigerated vials for analysis. The
Soxhlet extracts were analyzed using an HP5890 gas chro-
matograph with an HP5971 mass spectroscopy detector
(using electron ionization). The GC column was a DB-5 MS
column with 0.25 mm i.d. × 30 m length and 0.25 µm film
thickness (J&W Scientific). The initial column temperature
was set at 50 °C with a heating rate of 15 °C/ min and a final
column temperature of 300 °C. The mass selective detector
was operated in the electron impact mode. The trace
combustion products were identified by matching with a
mass spectra library (17) and comparison with known
retention indices of PAHs from the literature (18, 19). The
quantification of the individual combustion products was
done by comparison with the two perdeuterated PAH internal
standards. As in previous studies (8, 20), a linear interpolation
between the two internal deuterated standards was used to
estimate the relative response of the identified PAHs. Ideally
every compound should be identified and quantified indi-
vidually using a known standard of that compound; however,
this is impractical due to the large number of PAH species
observed in polystyrene combustion systems.
FIGURE 1. Schematic diagram of experimental setup.
phases at 800-950 °C, finding that most of the PAHs (88%)
were associated with the soot at 800 °C. As the combustion
temperature was increased, however, an increasing fraction
of PAHs was found in the gas phase. Despite the high
variability in these previous studies, a clear trend of decreasing
total PAH yield (in both mass and number of species) with
increasing combustion temperature has been reported in
the literature. The majority of the PAHs detected have been
in the particulate phase.
To the best of our knowledge, no studies have reported
the size distributions of the particles formed during poly-
styrene combustion or determined the PAH chemical com-
position as a function of particle size. The distribution of
the PAHs in different size ranges is an important parameter,
as the deposition of particles in the respiratory tract and
lung is size dependent. In this study, the trace PAHs in the
particulate and gaseous emissions from polystyrene com-
bustion were investigated. Combustion parameters such as
furnace temperature and the size of polystyrene feed spheres
were varied to determine the effect on resultant particle size,
composition as a function of size, and overall partitioning
between the gaseous and particulate phases.
The particle size distribution in the furnace exhaust was
measured by three complementary instruments, due in part
to the large particle size range (from submicrometer to 20
µm and above): an optical particle counter (OPC) for 0.1-
3.0 µm, a differential mobility particle sizer (DMPS, TSI) in
combination with a condensation nuclei counter (CNC, TSI)
for 0.01-1.0 µm, and an eight-stage cascade impactor (Mark
3, Andersen) covering sizes from ∼0.28 to >20 µm. The
cascade impactor was primarily used to collect particles in
various size ranges for chemical analyses.
Experimental Procedures
The polystyrene combustion experiments were conducted
using a vertical, laminar flow, tubular furnace with
a
maximum temperature of 1200 °C (Figure 1). Dry, particle-
free air was mixed with small polystyrene spheres and fed
through the furnace reactor. An average of about 130% (by
mass) excess air was used for the combustion. The gas
residence time in the furnace was approximately 0.5 s. Most
MSW incinerators have gas residence times of about 2 s;
however, the residence time in this study was long enough
to ensure the complete combustion of the polystyrene spheres
(combustion time of ∼6 to 120 ms) (7). The aerosol dynamics
and chemistry were “frozen” at the exit of the reactor by
adding dilution air (2:1) at approximately 20 cm downstream
of the furnace, resulting in diluted exhaust temperatures
approximately half of the initial exhaust temperature, similar
to other studies (13, 14). At the sampling location, the diluted
exhaust had cooled to less than 90 °C.
Results and Discussion
A baseline case with 250-300 µm polystyrene feed spheres
combusted at 900 °C was chosen to investigate the PAH
partitioning between the gas and particulate phases as well
as the PAH distribution as a function of the exhaust particle
size. The 900 °C furnace temperature is close to the nominal
operating temperature of many state-of-the-art municipal
solid waste incinerators (2, 3). Experiments were also
conducted to investigate the partitioning and particle size
distribution as a function of furnace temperature (800-1200
°C) and the impact of the polystyrene feed sphere size.
The polystyrene sample was purchased as pellet stock
polymer (Aldrich, Inc.), frozen with liquid nitrogen, and
crushed using a grinder. The polystyrene powder was sieved
9
2 3 0 2 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 15, 1998

FIGURE 2. (A) Photograph of the polystyrene feed spheres, with size cuts 104-175 and 250-300 µm. (B) Transmission electron microscope
(TEM) image of the soot particles collected downstream of the combustor at two magnifications.
Baseline Experim ent (T ) 900 °C). In the baseline case,
about 87% (by total ion abundance) of the eluted compounds
from the Teflon filter extract and 75% of the gaseous
compounds were identified. The mass yields at 900 °C of
individual PAH compounds are summarized in Table 1,
Section A, for the filter extract and in Table 1, Section B, for
the XAD-2 sorbent extract, respectively.
found by Wheatley et al. (4) under similar experimental
conditions.
The most abundant PAH species in the gaseous phase
were biphenyl, naphthalene, acenaphthylene, and styrene
(see Table 1). Again, there were several compounds that
were found in the present work but not by Hawley-Fedder
et al. (8) or Wheatley et al. (4). As shown in Table 1, the
partitioning between the gas and particulate phases was
about 65% and 35%, respectively at 900 °C. Though previous
researchers have not reported a detailed distribution of PAH
or their yield, Hawley-Fedder et al. (8) reported a partitioning
of total PAH mass (26% gas and 74% particulate matter) that
was very different from that reported here at the same
combustion temperature. This difference may be due to the
variation in the PAH collection methods used in the two
experimental systems. Hawley-Fedder et al. (8) used glass
wool to trap particles (vs Teflon filters used in this study) and
cold traps to trap gas phase PAHs (vs XAD-2 absorbent used
here). The large amount of surface area in the glass wool
The most abundant PAH species in the particulate phase
were benzo[j]fluoranthene, 2,7-dimethylpyrene, pyrene, and
dihydronaphthacene, followed by several other, generally
higher molecular weight, PAHs. Some of these compounds
were observed by Hawley-Fedder et al. (8) in the particulate
phase at 900 °C and by Wheatley et al. (4) in both particulate
and gas phases at 950 °C (see Table 1). Several other species
that were identified in the present filter extract were not
found by Hawley-Fedder et al. (8) or by Wheatley et al. (4).
The total number of PAH species identified in this work at
900 °C is 36, which is smaller than the 41 species found by
Hawley-Fedder et al. (8) but much larger that the eight species
9
VOL. 32, NO. 15, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 3 0 3

TABLE 1. PAH Yields under Baseline Conditions (900 °C)a
retention index
(Lee Index)
PAH yield
(mg/g of PS)
Hawley-Fedder
(10) (900 °C)
Wheatley et al.
(6) (950 °C)
PAH species
% of total
Section A: In the Particulate Phase
naphthalene
isoquinoline
biphenyl
acenaphthylene
4-m ethyl-9H-fluorene
9H-fluorene
200.0
212.1
231.1
245.4
256.3
268.2
270.5
274.5
279.1
285.6
286.8
295.6
301.7
309.2
319.4
325.7
327.1
333.9
335.8
339.1
342.3
344.8
346.6
352.6
357.6
359.8
363.5
380.4
387.0
390.0
406.8
412.0
432.9
0.105
0.059
0.066
0.036
0.081
0.243
0.105
0.120
0.134
0.119
0.500
0.156
0.650
0.436
0.428
0.828
0.092
0.122
0.130
1.114
0.639
0.185
0.980
0.088
0.374
0.270
0.283
0.978
0.906
0.294
0.395
0.316
1.289
0.84
0.47
0.53
0.29
0.64
1.94
0.84
0.96
1.07
0.95
3.99
1.25
5.19
3.48
3.42
6.61
0.73
0.97
1.04
8.90
5.10
1.48
7.82
0.70
2.99
2.16
2.26
7.81
7.24
2.35
3.15
2.52
10.29
x
x
x
x
x
x
4-ethenyl-1,1′-biphenyl
4,4′-dim ethyl-biphenyl
xanthene
9,10-dihydroanthracene
9,10-dihydrophenanthrene
dibenzothiophene
anthracene
1-phenylnaphthalene
3-m ethylphenanthrene
4,5,9,10-tetrahydropyrene
4,5-dihydropyrene
2-phenyl-naphthalene
1,4-dihydro-1,4-ethenoanthracene
2,7-dim ethylphenanthrene
1,2,3,6,7,8-hexahydropyrene
fluoranthene
pyrene
m -terphenyl
p-terphenyl
benzo[a]fluorene
4,5,6-trihydrobenz[de]anthracene
dihydronaphthacene
2,7-dim ethylpyrene
1,1′-binaphthyl
1,2′-binaphthyl
2,2′-binaphthyl
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
benzo[j]fluoranthene
total PAH yield (m g/g of PS)
% of PAHs in particulate
12.52
35.1
Section B: In the Gaseous Phase
styrene
3,5-diphenyl-1,2,4-trioxolane
indane
1-ethynyl-4-m ethylbenzene
1-phenylethanone
naphthalene
quinoline
biphenyl
acenaphthylene
4-m ethyl-9H-fluorene
1,1′-(1,2-dibrom o-1,2-ethanediyl)bisbenzene
4-ethenyl-1,1′-biphenyl
9-m ethylfluorene
9,10-dihydrophenanthrene
dihydronaphthacene
135.5
151.5
169.1
169.8
173.9
200.0
206.0
235.6
249.0
257.6
259.4
270.6
272.9
287.0
381.2
2.078
2.019
1.262
1.401
0.843
3.955
0.482
4.249
2.136
0.353
0.610
1.642
0.653
0.289
1.212
8.96
8.71
5.44
6.04
3.64
17.06
2.08
18.33
9.21
1.52
2.63
7.08
2.81
1.25
5.23
x
x
x
x
x
x
x
total PAH yield (m g/g of PS)
23.18
% of PAH′s in gas phase
64.9%
a
x indicates detected but not quantified in the literature studies.
was likely to absorb gas phase PAHs and wrongly attribute
them to the particulate phase. The setup used in this
experiment is more likely to correctly attribute PAHs to the
two phases.
basis of the instruments, this is most likely due to assumptions
inherent in each method.
The chemical composition as a function of particle size
has been investigated using a cascade impactor separating
the particles into a total of eight size fractions. The results
of the chemical analyses are summarized in Table 2. On
average, about 81% (by total ion abundance) of the observed
total ion chomatogram peaks have been identified. Several
species that were identified in the cascade impactor samples
were not found in the “total particulate matter” Teflon filter
sample and vice versa. In general, there were more, lower
molecular weight PAHs identified in the Teflon filter extract,
In addition to chemical analyses, the aerosol size distri-
bution was measured. The exhaust particle size distribution
at 900 °C is shown in Figure 3, which indicates that most of
the particles are smaller than 1 µm, with the peak at 0.3-0.4
µm. In addition, there is a general consistency in the
measurements of the three instruments over the particle size
range measured. Although some discrepancy does exist
between the actual number of particles determined on the
9
2 3 0 4 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 15, 1998

TABLE 2. Particulate Size Distribution of the PAHs under Baseline Conditions (900 °C)
mean impactor bin size (µm)
retention index total impactor
PAH species
(Lee Index)
PAH yield (mg/g) 14.4 6.1
2.5 1.24 0.66 0.36 0.14
acenaphthylene
2,3,6-trim ethylnaphthalene
9H-fluorene
4-ethenyl-1,1′-biphenyl
4,4′-dim ethyl-biphenyl
4,4a,9,10-tetrahydro-4a-m ethyl-2(3H)-phenanthrenone
1,2,3,4,5,6,7,8-octahydrophenanthrene
dibenzothiophene
anthracene
acridine
benzo[f]quinoline
phenanthridine
1-phenylnaphthalene
1,2,3,10b-tetrahydrofluoranthene
3-m ethylphenanthrene
o-terphenyl
4,5,9,10-tetrahydropyrene
thianthrene
2-phenylnaphthalene
2,7-dim ethylphenanthrene
1,2,3,6,7,8-texahydropyrene
fluoranthene
245.4
265.3
268.2
270.5
274.5
288.3
292.3
295.6
301.7
303.3
307.3
308.1
309.2
314.6
319.4
321.7
325.7
327.9
333.9
339.1
342.3
344.8
346.6
349.1
357.6
359.8
363.5
380.4
387.0
394.3
395.2
403.5
407.3
432.9
437.8
452.0
0.05
0.04
0.05
0.09
0.02
0.69
0.03
0.09
0.08
0.19
0.12
0.04
0.17
0.16
0.11
0.19
0.32
0.26
0.03
0.72
0.26
0.02
1.67
0.38
0.03
0.13
0.02
3.73
0.35
0.44
0.03
0.21
0.45
0.13
0.04
0.12
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
pyrene
9,10-dim ethylanthracene
p-terphenyl
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
x
benzo[a]fluorene
x
x
x
4,5,6-trihydrobenz[de]anthracene
5,12-dihydronaphthacene
2,7-dim ethylpyrene
benzo[c]phenanthrene
cyclopenta[cd]pyrene
benzo[a]carbazole
x
x
x
x
x
x
x
x
x
x
x
x
x
naphthacene
x
benzo[j]fluoranthene
benzo[k]fluoranthene
perylene
x
total PAH yield (m g/g of PS)
11.7
2.1 1.2 0.6 0.7 2.3 1.7 3.2
a
X indicates detected in this im pactor size range.
FIGURE 4. Particulate size distribution of the total PAH yield and
the number of PAH species under baseline condition (900 °C).
FIGURE 3. Size distribution of the particulates formed in the
polystyrene combustion exhaust under baseline conditions (900
°C): (b) Optical particle counter (OPC). (9) Cascade impactor. (Solid
line) Differential mobility particle sizer (DMPS).
The total mass of the aerosol-associated PAH species was
found to vary significantly with the size of the particles. The
variation of the number of PAH species and total PAH mass
as a function of particle size is shown in Figure 4. There is
a bimodal distribution in both the total PAH mass yield and
the number of species as a function of particle size, with the
minimum between 1 and 2 µm. In addition, there is also
significant difference in the molecular weights of the PAH
species on the particles above and below 1 µm. As shown
in Table 2, the PAHs associated with particles larger than
while more higher molecular weight PAHs were identified in
the cascade impactor sample extract. This may be due to
the Teflon filter trapping some of the lower molecular weight
compounds as the gas flows through the filter. In the cascade
impactor, the compounds deposit on the filters by impaction,
and the lighter compounds can flow around, not through,
the filter material.
9
VOL. 32, NO. 15, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 3 0 5

FIGURE 5. Temperature dependence of the PAH yield and the number
of PAH species in the gas and particulate phases. (9) PAH yield
in the particulate phase. (0) PAH yield in the gas phase. (b) Number
of PAH species on the particulates. (9) Number of PAH species in
gas phase.
FIGURE 6. Size distribution of particles in the exhaust as measured
by the optical particle counter (OPC) for two different feed particle
sizes and at two furnace temperatures.
sion is consistent with previous studies by Panagiotou and
Levendis (7), who reported that the smaller polystyrene
spheres burn faster in the furnace and thus have longer time
to burn more completely, although the overall combustor
residence time is approximately the same with both feed
sphere sizes.
The more complete combustion of smaller polystyrene
feed spheres is reinforced by the resultant exhaust particle
size distribution. As shown in Figure 6, the exhaust particle
size distribution for the smaller feed spheres at 900 °C is
similar to that for the larger feed spheres at 1200 °C. The
significant impact of the feed sphere size suggests that the
total PAH mass yield is highly correlated with combustion
condition even at the same furnace temperature. Therefore,
care must be taken in extrapolating the PAH mass yield data
from the laboratory experiments to full-scale municipal solid
waste incinerators.
Mechanistic Im plications. The experimental results
summarized above can be used to make some observations
about the mechanisms of incineration of polystyrene under
municipal solid waste incinerator conditions. First, the
decrease in the PAH yield with increasing combustion
temperature and smaller feed sphere size (as shown in Figures
5 and 6) indicate that these conditions lead to both faster
pyrolysis and gas-phase combustion reaction rates in the
diffusion flame zone and thus more complete combustion.
Second, there is a shift to heavier PAHs and larger particles
with more complete combustion, either at higher furnace
temperature or with smaller feed spheres. These observations
are consistent with the mechanism that includes the forma-
tion of larger PAH molecules through the agglomeration of
smaller PAH molecules or fragments at higher combustion
temperatures. The faster the flame reactions, the more
complete the combustion and the faster the agglomeration
processes, which in turn lead to higher fractions of larger
PAH species. Third, the peak in the bimodal distribution of
the PAH yield below 1.0 µm (Figure 4) may be due to
condensation/ deposition of PAHs, whereas the PAHs as-
sociated with the larger particles (i.e., the peak between 6
and 14 µm) may be partly due to surface chemical reactions
between smaller PAH fragments in the gas phase and those
adsorbed on the particle surface. Condensation will occur
preferentially on smaller particles because of the larger
amount of surface area, but chemical reaction may be driven
by surface area or by the reactivity of the particle surface. A
range of low and high molecular weight species will condense,
but chemical reaction will tend to create higher molecular
weight species. It is hypothesized that condensation domi-
nates in the smaller particles and that chemical reaction
dominates in the larger particles. This hypothesis is con-
1 µm are primarily those with retention indices larger than
fluoranthene (RI ) 344.8), while the smaller particles contain
PAHs with a wide range of retention indices and presumably
molecular weights as well. The total mass of the particles
does not have the bimodal distribution observed in the PAH
yields; the number size distribution of the particles has a
log-normal size distribution, as is normally associated with
combustion aerosol, with a mass mean diameter around 1.24
µm. The number mean diameter (Figures 3 and 6) is less
than 1.0 µm. The relatively large amount of PAHs associated
with the smaller particles (less than 1.0 µm) provides the
need for control of fine particulate emissions from incinera-
tors.
Variation with Com bustion Tem perature. The impact
of furnace temperature on polystyrene combustion was
investigated at 800, 900, 1000, 1100, and 1200 °C. The PAH
yields in the gas and particulate phases, as well as the number
of PAH species in each phase, are shown in Figure 5. There
was an exponential decrease in the total number of PAH
species and total mass yields when the combustion tem-
perature was increased from 800 to 1200 °C, with virtually
no PAHs observed at 1200 °C. At lower temperatures, the
percentage of PAH mass in the gas phase was substantially
higher than in the particulate phase (70% vs 30%), but at the
higher temperatures, the reverse was true: at 1100 °C, about
70% of the PAH mass is bound with the particles, while 30%
was in the gas phase. The size distribution of the particles
also varies significantly with the furnace temperature and is
shown in Figure 6 for 900 and 1200 °C. There is a significant
shift to larger particles as the furnace temperature increases.
At 900 °C, the average particle size is about 0.3 µm, but at
1200 °C it increases to about 1 µm. This increase in particle
size may be due to faster combustion at the higher tem-
peratures, allowing more time for the nucleated particles to
grow by coagulation and/ or condensation into larger sizes
in the furnace (as residence time was kept constant).
Variation with Polystyrene Feed Size. To determine the
impact of polystyrene feed sphere size, a smaller feed sphere
size cut (104-175 µm) was combusted at a furnace tem-
perature of 900 °C. The PAH mass yield with the smaller
feed spheres is about seven times smaller than with the larger
ones (250-300 µm). The yield for the smaller feed spheres
is about the same as the larger ones at a much higher furnace
temperature (1200 °C). In addition, the PAH partitioning is
about the same for the smaller feed spheres at 900 °C as for
the larger ones at 1200 °C. These observations suggest that
smaller polystyrene feed spheres lead to more complete
combustion at the same furnace temperature. This conclu-
9
2 3 0 6 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 32, NO. 15, 1998

(6) World Health Organization. J. Air Pollut. Control Assoc. 1989,
39, 792.
(7) Panagiotou, T.; Levendis, Y. Combust. Flame 1994, 99, 53-74.
(8) Hawley-Fedder, R. A.; Parsons, M. L.; Karasek, F. W. J. Chro-
matogr. 1984, 315, 201-210.
(9) Cullis, C. F.; Hischler, M. M. The Combustion of Organic Polymers;
Clarendon Press: Oxford, 1981.
(10) Hepp, H.; Siegmann, K.; Sattler, K. Chem. Phys. Lett. 1995, 233,
16-22.
(11) McKinnon, J. T.; Meyer, E.; Howard, J. B. Combust. Flame 1996,
105, 161-166.
(12) Flagan, R. C.; Seinfeld, J. H. Fundamentals of Air Pollution
Engineering; Prentice-Hall: Englewood Cliffs, NJ, 1988.
(13) Biswas, P.; Li, X.; Pratsinis, S. E. J. Appl. Phys. 1989, 65, 2445-
2450.
sistent with the data presented in Table 2, which show that
the PAHs associated with the larger particles are mostly of
high molecular weight, while the ones on the smaller particles
have a wide range of molecular weights. Further verification
of the hypothesized surface or interfacial reaction of PAHs
is needed before definitive conclusions can be reached.
The present experimental results also provide important
insights into emission control strategies for PAHs from
municipal solid waste incinerators. As shown in Figure 4,
a relatively large amount of PAHs formed from polystyrene
incineration are associated with the fine particles less than
about 1.0 µm. The relatively large amount of PAHs associated
with the smaller particles re-emphasizes the need for control
of fine particulate emissions from incinerators.
(14) Biswas, P. In Aerosol Measurement: Principles, Techniques and
Applications; Willeke, K., Baron, P., Eds.; Van Nostrand Rein-
hold: New York, 1992; pp 705-720.
Acknowledgments
(15) Hart, K. M.; Pankow, J. F. Environ. Sci. Technol. 1994, 28, 655-
Financial support for this study was provided by The Procter
& Gamble Company. Special thanks are due to Tai-Gyu Lee
(Aerosol and Air Quality Research Laboratory) for his
assistance with the combustion experiments. We would also
like to thank Staci Simonich (Procter & Gamble) for her help
with the GC/ MS chemical analysis methodology.
661.
(16) Turpin, B. J.; Liu, S. P.; Podolske, K. S.; Gomes, M. S. P.; Eisenreich,
S. J.; McMurry, P. H. Environ. Sci. Technol. 1993, 27, 2441-
2449.
(17) Hewlet Packard. NIST/ EPA/ MSDC 54k Mass Spectral Database
(NBS54k.l). 1990.
(18) Rostad, C. E.; Pereira, W. E. J. High-Resolut. Chromatogr.
Chromatogr. Commun. 1986, 9, 328-334.
(19) Lee, M. L.; Vassllaros, D. L.; White, C. M.; Novotny, M. Anal.
Chem. 1979, 51, 768-773.
(20) Robertson, D. J.; Elwood, J. H.; Groth, R. H. Chemical Composition
of Exhaust Particles from Gas Turbine Engines; 1979; EPA 68-
02-2458.
Literature Cited
(1) Lin, W. Y.; and Biswas, P. Combust. Sci. Technol. 1994, 101,
29-43.
(2) Durlak, S. K.; Biswas, P.; Shi, J. J. Hazard. Mater. 1997, 56, 1-20.
(3) Tillman, D. A.; Rosii, A. J.; Vick, K. M. Incineration of Municipal
and Hazardous Waste; Academic Press: San Diego, 1989.
(4) Wheatley, L.; Levendis, Y. A.; Vouros, P. Environ. Sci. Technol.
1993, 27, 2885-2895.
Received for review October 15, 1997. Revised manuscript
received April 14, 1998. Accepted April 22, 1998.
(5) Elomaa, M.; Saharinen, E. J. Appl. Polym. Sci. 1991, 42, 2819-
2824.
ES9709031
9
VOL. 32, NO. 15, 1998 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 2 3 0 7