Laboratory evaluation of a prototype photochemical chamber designed to investigate the health effects of fresh and aged vehicular exhaust emissions

Abstract

Laboratory experiments simulating atmospheric aging of motor vehicle exhaust emissions were conducted using a single vehicle and a photochemical chamber. A compact automobile was used as a source of emissions. The vehicle exhaust was diluted with ambient air to achieve carbon monoxide (CO) concentrations similar to those observed in an urban highway tunnel. With the car engine idling, it is expected that the CO concentration is a reasonable surrogate for volatile organic compounds (VOCs) emissions. Varying the amount of dilution of the exhaust gas to produce different CO concentrations, allowed adjustment of the concentrations of VOCs in the chamber to optimize production of secondary organic aerosol (SOA) needed for animal toxicological exposures. Photochemical reactions in the chamber resulted in nitric oxide (NO) depletion, nitrogen dioxide (NO₂) formation, ozone (O₃) accumulation, and SOA formation. A stable SOA concentration of approximately 40 μg m⁻³ at a chamber mean residence time of 30 min was achieved. This relatively short mean residence time provided adequate chamber flow output for both particle characterization and animal exposures. The chamber was operated as a continuous flow reactor for animal toxicological tests. SOA mass generated from the car exhaust diluted with ambient air was almost entirely in the ultrafine mode. Chamber performance was improved by using different types of seed aerosol to provide a surface for condensation of semivolatile reaction products, thus increasing the yield of SOA. Toxicological studies using Sprague-Dawley rats found significant increases of in vivo chemiluminescence in lungs following exposure to SOA.

Full text

495
Introduction
Epidemiological studies have shown associations between
exposures to ambient particles and adverse health out-
comes, such as cardiopulmonary morbidity and mor-
tality (Dockery et al., 1993; Dockery & Pope, 1994; Pope
et al., 2002; Schwartz, 2004). Because ambient particles
originate from many dierent sources, with associated
dierences in composition and physicochemical char-
acteristics, it is likely that there are corresponding dier-
ences in toxicity. Mobile source emissions are a signicant
contributor to individual and community exposures, and
are of substantial interest in health eect studies. Laden
et al. (2000) showed that direct vehicular emissions were
associated with increased rate of cardiovascular deaths.
Another study showed an association between transient
exposure to trac emissions and an increase in the risk
of myocardial infarction (Peters et al., 2004).
Although epidemiological studies can examine asso-
ciations between population exposures and outcomes,
there is still a need to study the eects of exposure to
particulate pollutants from specic sources. So far, most
research has focused on measurements of the toxicity of
particles directly emitted from combustion sources. For
example, exposure to directly emitted diesel particles has
been associated with lung infection or inammation and
oxidative stress in mice (McDonald et al., 2004). In a toxi-
cological study, exposure to gasoline vehicle emissions
was associated with cell damage (Seagrave et al., 2003).
RESEARCH ARTICLE
Laboratory evaluation of a prototype photochemical chamber
designed to investigate the health eects of fresh and aged
vehicular exhaust emissions
Vasileios Papapostolou, Joy E. Lawrence, Edgar A. Diaz, Jack M. Wolfson, Stephen T. Ferguson,
Mark S. Long, John J. Godleski and Petros Koutrakis
Department of Environmental Health, Harvard School of Public Health, Boston, Massachusetts, USA
Abstract
Laboratory experiments simulating atmospheric aging of motor vehicle exhaust emissions were conducted using a
single vehicle and a photochemical chamber. A compact automobile was used as a source of emissions. The vehicle
exhaust was diluted with ambient air to achieve carbon monoxide (CO) concentrations similar to those observed
in an urban highway tunnel. With the car engine idling, it is expected that the CO concentration is a reasonable
surrogate for volatile organic compounds (VOCs) emissions. Varying the amount of dilution of the exhaust gas to
produce dierent CO concentrations, allowed adjustment of the concentrations of VOCs in the chamber to optimize
production of secondary organic aerosol (SOA) needed for animal toxicological exposures. Photochemical reactions
in the chamber resulted in nitric oxide (NO) depletion, nitrogen dioxide (NO2) formation, ozone (O3) accumulation,
and SOA formation. A stable SOA concentration of approximately 40 μg m−3 at a chamber mean residence time of
30 min was achieved. This relatively short mean residence time provided adequate chamber ow output for both
particle characterization and animal exposures. The chamber was operated as a continuous ow reactor for animal
toxicological tests. SOA mass generated from the car exhaust diluted with ambient air was almost entirely in the
ultrane mode. Chamber performance was improved by using dierent types of seed aerosol to provide a surface for
condensation of semivolatile reaction products, thus increasing the yield of SOA. Toxicological studies using Sprague-
Dawley rats found signicant increases of in vivo chemiluminescence in lungs following exposure to SOA.
Keywords: Photochemical chamber, secondary organic aerosol, vehicular emissions
Address for Correspondence: Department of Environmental Health, Harvard School of Public Health, 401 Park Drive, Boston, Massachusetts
02215, USA. E-mail: vpapapos@hsph.harvard.edu.
(Received 04 December 2010; revised 06 May 2011; accepted 06 May 2011)
Inhalation Toxicology, 2011; 23(8): 495–505
© 2011 Informa Healthcare USA, Inc.
ISSN 0895-8378 print/ISSN 1091-7691 online
DOI: 10.3109/08958378.2011.587034
Inhalation Toxicology
2011
23
8
495
505
04 December 2010
06 May 2011
06 May 2011
0895-8378
1091-7691
© 2011 Informa Healthcare USA, Inc.
10.3109/08958378.2011.587034
UIHT
587034
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

496 V. Papapostolou et al.
Inhalation Toxicology
Ambient aerosol consists of a complex mixture of both
directly emitted particles (primary) and those formed
in the atmosphere through photochemical reactions
(secondary) (Lim & Turpin, 2002). is raises questions
about how the toxicity of pollution emitted from dier-
ent sources is aected by atmospheric transformations.
Recently, there have been a few studies that have begun
to look at this question. e TERESA (Toxicological
Evaluation of Realistic Emission Source Aerosols) study
(Godleski et al., 2011; Godleski et al., 2011; Kang et al.,
2011) investigated the toxicity of primary and simulated
secondary emissions from coal-red power plants. Other
researchers investigated the toxicological properties of
secondary aerosol from a diesel generator (Zielinska
et al., 2010).
Vehicular primary emissions include elemental car-
bon (EC) particles (primarily from diesel vehicles) as
well as road and brake dust and other fugitive emissions.
However, vehicular pollutants also include pollutant
gases such as carbon monoxide (CO), oxides of nitrogen
(NOx), mostly in the form of nitric oxide (NO), volatile
organic compounds (VOCs), and semivolatile organic
compounds (SVOCs) as evaporative losses from the fuel
and from engine lubricants.
In the presence of sunlight, primary pollutant gases
react with ozone (O3), hydroxyl radical (·OH), or other
radicals. Inorganic gases are converted into secondary
pollutant particles such as ammonium nitrate (NH4NO3)
and ammonium sulfate ([NH4]2SO4), while VOCs are
converted into secondary organic aerosol (SOA) through
photochemical reactions. SOA is formed mostly by oxi-
dation of VOC species with 7 or more carbons to produce
reaction products that subsequently partition between
gas and particle phase (Seinfeld & Pandis, 2006).
Although substantial amounts of both gas- and parti-
cle-phase gases in primary emissions are oxidized, they
produce relatively low yields of SOA. Particulate phase
organics such as polycyclic aromatic hydrocarbons
(PAHs) are products of both incomplete combustion and
pyrolysis of fuel and/or lubricant. In the atmosphere,
these PAHs are oxidized to form predominantly particle-
phase nitro- and oxy-PAH derivatives (Nikolaou et al.,
1984; Finlayson-Pitts, 1986; Seinfeld & Pandis, 2006).
Even those PAH reactions that do not result in particle-
phase products can have a signicant impact on the toxic
or genotoxic potency of particles (Feilberg et al., 2002).
A large number of laboratory chamber studies (Izumi
& Fukuyama, 1990; Odum et al., 1997a, 1997b; Cocker
et al., 2001a; Kleindienst et al., 1999; Kleindienst et al.,
2004; Ng et al., 2006) have investigated the SOA-forming
potential of individual aromatic hydrocarbons and mix-
tures typical of mobile source emissions, which com-
prise the major component of the precursors of SOA in
urban areas. Kleindienst et al. (1999) found that toluene
achieved the highest SOA yield, followed by p-xylene and
1-3-5 trimethylbenzene.
is paper presents the laboratory development and
optimization of a prototype dynamic photochemical
chamber that produces sucient amounts of SOA from
VOCs in car exhaust of a single vehicle under irradiation
with articial light. It was the purpose of this study to
demonstrate the ability to detect biological outcomes
of exposure to SOA produced by a single vehicle, in
animal toxicological studies. e aim was not to fully
simulate the whole spectrum of atmospheric photo-
chemical reaction products that are formed in the real
ambient atmosphere. e same degree of oxidation
that occurs in the atmosphere cannot be truly recreated
in a photochemical reaction chamber using articial
light because, as it is known, photochemical chambers
generally produce a less oxidized aerosol and smaller
yield than can be found in the atmosphere. However,
for toxicological testing, the reproducibility oered by
articial lighting compared to actual solar irradiation
(which can vary in intensity by hour and by day) is
desirable because it aords reproducible products from
reproducible gas-phase mixtures. is is especially true
for inhalation toxicology studies, where stable produc-
tion and output are required over an exposure period of
several hours, using a dynamic system. Results from the
prototype tests were used to design a planned follow-up
study of the toxicity of similarly aged realistic aerosol
derived from mobile source emissions from an urban
trac tunnel.
Methods
Photochemical chamber
e photochemical reaction chamber (Ruiz et al., 2007a)
previously used for a TERESA coal-red power plant
study (Ruiz et al., 2007b; Kang et al., 2010) was used to
simulate the atmospheric photochemical aging of motor
vehicle exhaust diluted with ambient air.
e design and manufacture of the chamber are
described in detail elsewhere (Ruiz et al., 2007a). is
rectangular shaped 0.575 m3 chamber has an FEP (uo-
rinated ethylene propylene) Teon-coated aluminum
framework. e chamber has a oor area of 1.52 × 0.31
m2 and a 1.22 m height achieving a relatively low sur-
face (S) to volume (V) ratio of S × V = 9.4 m−1 to minimize
particle losses. FEP Teon lm with thickness of 51 μm
(American Duralm, Holliston, MA) was used on the two
larger sides of the framework to allow irradiation to pen-
etrate into the chamber. PTFE (polytetrauoro-ethylene)
Teon was used on the four smaller sides to minimize
wall reactions. A ventilated wooden enclosure was used
to protect researchers from exposure to ultraviolet (UV)
light and to remove excess heat generated by the lamps.
A total of 48 UVA-340 20 W lamps (Q-Panel Lab Products,
Cleveland, OH) were mounted on two enclosure sides at
a distance of about 5 × 10−2 from the chamber walls.
e UVA-340 lamps, with a strong photon emis-
sion around 340 nm, provide an excellent simulation
of the ground level solar spectrum (between 295 nm
and 365 nm) (Carter, 1995). e stable UV radiation of
these lamps enhances the photolysis of O3, nitrous acid
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

Photochemical chamber for toxicity testing 497
© 2011 Informa Healthcare USA, Inc.
(HONO), and formaldehyde (HCHO), with moderate
heat release.
Emission generation and sampling
e laboratory experimental system is shown schemati-
cally in Figure 1. A compact automobile (1997 Toyota
RAV4, SUV) was used as a source of emissions. e engine
was run slightly fuel-rich by controlling the vacuum at the
Manifold Absolute Pressure sensor during experiments,
causing a slight decrease in the eciency of fuel combus-
tion and resulting in CO concentrations (after dilution
with ambient air) similar to those observed in a highway
tunnel (Rogak et al., 1998; Kirchstetter et al., 1996; Fraser
et al., 1999; Fraser et al., 1998; Pierson et al., 1996). We
considered it reasonable to adjust the levels of ineciency
of fuel combustion to achieve CO levels comparable to
CO levels observed in mobile source eet emissions at an
urban highway tunnel e.g. 2–12 ppm. Also, the CO would
be a qualitative indicator of the amount of VOCs concen-
tration in the mixture introduced in the chamber, e.g. low
CO–low VOCs, high CO–high VOCs. us, we expected
that the CO served as a reasonable surrogate for VOCs
emissions (Fraser et al., 1998).
e car was parked in between buildings of the Harvard
School of Public Health and the Harvard School of Dental
Medicine in the Longwood Medical Area in Boston, MA.
e tailpipe was inserted into a tube that captured all of
the exhaust from the tailpipe plus outdoor air (primary
dilution of the exhaust with cooling to near ambient tem-
perature). e mixture was pulled by a variable speed
fan through a duct (length about 30 ft) into the labora-
tory. Most of the mixture was vented to a fume hood in
the laboratory, with a small amount transferred by a
vacuum pump to the reaction chamber. Before entering
the chamber, a secondary dilution of the exhaust mixture
was made with outside ambient air.
Measurement and monitoring
Continuous gas monitors were used to monitor the
concentrations entering and exiting the chamber, with a
4-way to automatically switch at 5-min intervals between
upstream and downstream. CO was monitored by infra-
red absorption (Model 48 Analyzer; ermo Scientic,
Franklin, MA), NO and NO2 by chemiluminescence
(Model 42C Analyzer; ermo Scientic), O3 by UV pho-
tometry (Model 49C Analyzer; ermo Scientic).
Particle size distribution and concentration were
monitored using a Scanning Mobility Particle Sizer
(SMPS Model 3934; TSI Inc., Shoreview, MN) coupled
with a Condensation Particle Counter (CPC Model 3785;
TSI Inc.) and an Aerodynamic Particle Sizer (APS Model
3321; TSI Inc.). Particle count was monitored using a
Condensation Particle Counter (CPC Model 3007; TSI
Inc). Particle mass concentration was estimated using the
SMPS, assuming spherical particles and estimated parti-
cle density (ρp = 1.0 g cm−3) for SOA. Particle instruments
were connected directly to the outlet of the chamber with
short pieces of tubing to minimize particle losses.
Temperature and relative humidity were monitored
using a Vaisala transmitter (HMD-70; Vaisala Oyj,
Helsinki, Finland). e sensor was placed inside the
enclosure, just outside of the chamber. A ow of 0.3 L
min−1 was taken from the chamber and passed over the
sensor probe.
Figure 1. Single vehicle exhaust laboratory experimental system.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

498 V. Papapostolou et al.
Inhalation Toxicology
VOC samples were collected in a subset of experiments
on duplicate stainless steel TD (thermal desorption),
unconditioned Carbopack B tubes (Supelco Division;
Sigma-Aldrich Co) that were subsequently analyzed by
gas chromatography/mass spectrometry (GC/MS). e
tubes were sampled upstream and downstream of the
chamber, at a ow rate of 20 cm3 min−1 for 120 min. e
ow rate through each tube was adjusted by a valve and
ow stability was continuously monitored during each
experiment using a vacuum gauge to measure the pres-
sure drop.
Typical experimental protocol
Unless otherwise indicated, the experimental proto-
col was: (1) ush overnight with ambient air; (2) in the
morning of the day of the experiment, start car and
adjust CO concentration in exhaust by adjusting primary
and secondary dilutions; (3) connect diluted exhaust to
chamber and establish the desired mean residence time
by adjusting the ow through the chamber; (4) wait for
steady CO concentration in the chamber at target level
(typically 2 mean residence times); (5) turn on lamps,
and; (6) leave lamps on for duration of the experiment
(typically 5 mean residence times). In the seed aerosol
experiments, addition and subsequent stabilization of
the seed aerosol concentration preceded the irradiation
of the chamber.
EC/OC tests
ree sets of experiments were conducted (see
Table 2) by varying the diluted car exhaust CO concen-
tration and the mean residence time in the chamber.
When a stable CO concentration, exiting the chamber,
was reached, either 10 or 20 ppm, the lamps were turned
on and the photochemical reactions that involved oxi-
dation of vehicular precursor gas-phase compounds
took place. In combination with the dierent CO con-
centrations, two dierent ow rates were also tested,
either 20 or 10 L min−1, resulting in 30 or 60 min mean
residence time in the chamber, respectively. Samples
for EC/organic carbon (OC) analysis were collected
upstream and downstream of the chamber. e sample
ow through the lters was started when the lamps were
turned on and was stopped at the end of the experi-
ment. e chamber irradiation experiments lasted for
4 h. e lters used for EC/OC collection were 47 mm
pre-red quartz (Sunset Lab. Inc., Tigard, OR) placed
in PFA Teon lter holders (Savillex, Minnetonka, MN).
It should be noted that all samples for EC/OC analysis
were collected during experiments where only diluted
car exhaust was photochemically oxidized in the cham-
ber with no articial aerosol added. e lters were sub-
sequently stored in a freezer and later analyzed in our
laboratory facilities using the IMPROVE-TOR (ermal
Optical Reectance) method (Chow et al., 1993) using
our Sunset OCEC Dual-Optical Lab Analyzer (Sunset
Lab. Inc.).
Seed aerosol tests
Formation of SOA was also studied in the presence of dif-
ferent types of inert seed aerosol. e seed aerosol was
added for three reasons: (1) to provide sucient particle
surface area to allow rapid aggregation of the freshly
generated ultrane SOA particles and condensation of
gases, resulting in improved secondary PM yield; (2) to
decrease the time required to form a stable accumulation
mode size distribution; and (3) to achieve PM concentra-
tions similar to those found in an urban highway tunnel
in the northeastern United States.
e seed aerosol was generated (from an aque-
ous dispersion) with a HEART (High-output Extended
Aerosol Respiratory erapy; Westmed, Inc, Tucson,
AZ) nebulizer. e output of the nebulizer was diluted
with particle-free dry air in a 4-L vessel. e aerosol was
then mixed with the diluted car exhaust in the chamber.
Following stabilization of the CO concentration in the
chamber, addition of the seed aerosol was started. When
a steady-state baseline seed aerosol mass concentration
Table 1. Amounts of reacted VOCs.
Upstream of
chamber, mean
(±SD) (ppb)
Downstream of
chamber, mean
(±SD) (ppb) Reacted (%)
Benzene 60.7 ± 2.9 58.4 ± 2.0 3.8
Ethylbenzene 10.3 ± 1.1 9.7 ± 1.0 5.4
Toluene 49.8 ± 8.8 44.6 ± 5.0 10.3
o-Xylene 12.4 ± 1.4 10.9 ± 0.9 11.8
m/p-Xylene 32.1 ± 3.1 27.8 ± 2.9 13.4
During this experiment: CO: 20.6 ± 0.6 ppm, NOx–NO: 148.3 ± 12.9
ppb, O3: 325.5 ± 14.0 ppb, SMPS: 30.6 ± 2.1 μg m−3, CPC
55,000 ± 6,000 # cm−3. Tres = 30 min.
CO, carbon monoxide; CPC, Condensation Particle Counter;
SMPS, Scanning Mobility Particle Sizer; Tres, residence time;
VOC, volatile organic compound.
Table 2. Organic carbon analysis.
10 ppm CO and 30 min residence time,
mean (±SD) (μg m−3)
10 ppm CO and 60 min residence time,
mean (±SD) (μg m−3)
20 ppm CO and 30 min residence
time, mean (±SD) (μg m−3)
Upstream Downstream Upstream Downstream Upstream Downstream
Total OC 24.2 ± 7.0 77.7 ± 7.6 16.9 ± 3.9 539 ± 10.2 20.4 ± 4.7 50.6 ± 15.2
OC1 0.7 ± 0.3 2.4 ± 0.5 0.5 ± 0.1 2.5 ± 0.3 1.0 ± 0.4 2.0 ± 0.1
OC2 15.6 ± 5.9 37.5 ± 6.1 10.2 ± 3.4 22.3 ± 4.3 7.3 ± 0.6 21.8 ± 5.0
OC3 5.8 ± 0.6 25.6 ± 2.1 4.6 ± 0.6 20.3 ± 3.5 10.1 ± 2.9 18.1 ± 7.0
OC4 1.3 ± 0.1 7.4 ± 0.6 0.9 ± 0.1 5.7 ± 1.4 1.5 ± 0.5 5.0 ± 1.8
Temperature range for each OC fraction: ∼25–120°C (OC1); 120–250°C (OC2); 250–450°C (OC3); and 450–550°C (OC4).
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

Photochemical chamber for toxicity testing 499
© 2011 Informa Healthcare USA, Inc.
was reached the lamps were turned on to initiate the
photochemical reactions.
One set of experiments used hollow glass spheres
(HGS; diameter, d = 2–20 μm; density, ρp = 1.1 g cm−3;
Polysciences Inc., Warrington, PA), which is chemically
rather inert. Another set of experiments used Mt Saint
Helens Ash (MSHA; density, ρp = 2.6 g cm−3), which is both
chemically and toxicologically inert (Savage et al., 2003).
Well-mixed chamber− Particle losses tests
e chamber was lled at a ow of 5 L min−1 with car
exhaust diluted with ambient air at 20 ppm of CO. Particle
and gas concentrations exiting the chamber were stabi-
lized after a 2-h period and the lamps were turned on.
After 4 h of elapsed irradiation time (2 mean residence
times) the lamps were turned o. By that time there was
a stable chamber concentration of secondary aerosol.
At that point, the ow into the chamber was lowered
from 5 L min−1 to roughly 2.2 L min−1, substituting clean,
particle-free air for diluted car exhaust. is ow of 2.2 L
min−1 was sucient to replace the total ow sampled
downstream of the chamber through the gas/particle
instruments. e CO monitor, at 1.1 L min−1, was used
to measure the CO concentration. e SMPS, at 0.3 L
min−1, was used to measure the particle mass concentra-
tion assuming spherical particles and particle density
of ρp = 1.0 g cm−3. e CPC, at 0.77 L min−1, was used to
measure particle number concentrations. e clean air
ow (2.2 L min−1) going into the chamber (slightly higher
than the sum of the instrumental ows, 2.17 L min−1) was
used to slightly pressurize the chamber and prevent con-
vex distortion of the Teon lm walls (with the very small
excess ow released through small leaks in the chamber).
is ow through the chamber approximated static mode
conditions (negligible ow relative to chamber volume).
Measurements were taken at 5-min intervals with the
three instruments. is experimental setup was used to
test whether the reaction chamber behaved as a well-
mixed chamber and also to estimate the particle losses
in the chamber.
Determination of NO2 photolysis rate
We determined the NO2 photolysis rate in the chamber
by steady-state actinometry (Cocker et al., 2001b). A
mixture of NO and O3 in clean, particle-free air was intro-
duced continuously into the chamber. When a stable
NO2 concentration was reached, the lamps were turned
on and the NO2 photolysis rate (j1) was calculated using
the photostationary state relationship of NO, NO2, and O3
(Equation (1)):
jk
12
3
2
=⋅[][]
[]
NO O
NO
(1)
where k2 = 3.0 × 10−12 × exp(−1500 × T−1) cm3 molecule−1
s−1 (with gas concentrations in molecules−1 s−1), the tem-
perature (T in degrees K) dependent rate constant for the
reaction of NO with O3 (Sander et al., 2003).
Equation (1) was derived from reactions R1–R3:
NO NO O
2
1
+→+hv j
(R1)
Reaction R1 describes the photolysis of NO2 at wave-
lengths λ < 424 nm
OO MO++→
23
1
k
(R2)
Reaction R2 describes the reaction (with rate constant
k1) of the oxygen atom with the oxygen molecule to form
ozone (M is O2 or N2 that absorbs excess energy to stabi-
lize O3), and:
NO ONOO+→+
322
2
k
(R3)
Reaction R3 describes the reaction (with rate constant k2)
of O3 with NO to regenerate NO2.
At the photostationary state,
d
d
OOO MNOO
3
12 23
0
[]
== ⋅
[]
⋅
[]
⋅
[]
−⋅
[]
⋅
[]
tkk
(2)
and
d
d
ONO OO M
[]
== ⋅
[]
−⋅
[]
⋅
[]
⋅
[]
tjk01212
(3)
us,
OOO M
NO
SS
3
12
2
[]
=⋅
[]
⋅
[]
⋅
[]
{}
⋅
[]
{}
k
k
(4)
and
ONO
OM
[]
=⋅
[]
{}
⋅
[]
⋅
[]
{}
SS
j
k
12
22
(5)
Substituting,
ONO
NO
SS
3
12
2
[]
=⋅
[]
{}
⋅
[]
{}
j
k
(6)
and rearranging, equation 1 is developed.
Toxicological studies
Animals
Male Sprague-Dawley (SD-CD) rats 250–350g were
obtained from Taconic Farms (Rensselaer, NY); housed
and managed according to NIH guidelines for the care
and use of laboratory animals. Upon arrival, animals
were randomly assigned a unique identication number
which determined the exposure group.
Experimental design
Two dierent exposure scenarios were evaluated: (1)
MSHA; (2) MSHA+SOA and these were compared to
ltered air (Sham) exposures. For any given day of expo-
sures, 2 SD-CD rats were exposed to aerosol and 2 to
ltered room air in the individual whole-body exposure
chambers for 6 h. At the end of the exposure, animals
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

500 V. Papapostolou et al.
Inhalation Toxicology
from each group had chemiluminescence studies of the
heart and lung (Gurgueira et al., 2002).
Exposure
Tailpipe emissions from the car were collected and
diluted with ambient air as described above. e cham-
ber was operated dynamically, with residence time in the
chamber optimized for the production of secondary PM
from the exhaust. We also veried that the production of
secondary particle mass was consistent for the specic
concentration of CO in the diluted exhaust. MSHA was
used as a seed aerosol (see Section Seed aerosol eect).
Excess primary and photochemical gases from the pho-
tochemical chamber were removed before exposure
using a countercurrent membrane diusion denuder
(Ruiz et al., 2006).
Organ chemiluminescence
Spontaneous chemiluminescence of the surface of
lung and heart was measured as previously described
(Gurgueira et al., 2002). Briey, a orn EMI CT1
single-photon counting apparatus with an EMI 9816B
photomultiplier cooled at −20°C was used. Rats were
anesthetized with sodium pentobarbital (50 mg kg−1 i.p.)
and connected to an animal ventilator (5 ml breath−1,
60 breaths min−1; Harvard Apparatus, Cambridge, MA),
the chest was opened and the animals were placed in
the measurement compartment. Body temperature was
kept at 37°C using isothermal pads (Braintree Scientic,
Braintree, MA). Emission data were expressed as counts
per second per unit of tissue surface (cps cm−2). Statistical
analyses used analysis of variance models to assess an
exposure eect on the mean and to compare the expo-
sure eects across dierent scenarios.
Results and discussion
Determination of NO2 photolysis rate
e magnitude of light irradiation by the UVA-340 lamps
surrounding the chamber was estimated by measuring
the NO2 photolysis rate (j1), which we determined experi-
mentally to be j1 = 0.18 min−1 (using Equation (1)). Using
banks of UVB-313 lamps (in contrast with the UVA-340
lamps for these experiments) with this same chamber,
Ruiz et al. (2006) measured j1 = 0.096 min−1. As indicated
by Carter et al. (1995), the ideal light irradiation for a
photochemical chamber would correspond to approxi-
mately j1 = 0.3 min−1, which is approximately 50% of the
maximum NO2 photolysis rate observed at ground level
(at mid-US latitudes) on a clear day with direct overhead
sunlight. For example, Zafonte et al. (1977) estimated
j1 = 0.45 min−1 in Los Angeles at noon time.
Well-mixed chamber—particle losses test
We introduced a small ow of clean particle-free air into
the chamber, containing a stable concentration of CO
and particles and measured the concentration of CO and
particles downstream of the chamber for several hours.
Equation (8) was used to calculate the theoretical decay
curve due to dilution only in the chamber.
C
Ckt
t
()
()
=⋅−
()
0
exp
(7)
where the ratio C(t)/C(0) is the remaining fraction as a
dimensionless parameter in the chamber; k is the loss
rate constant or the reciprocal of the mean residence
time in the chamber, where k = F × V−1, in min−1 (F: ow
rate; V: volume).
Using the nominal chamber volume (V = 0.575 m3)
and the dilution clean air ow rate (F = 2.2 L min−1) the
theoretical decay curve k value was calculated to be
k = 3.8 × 10−3 min−1. Figure 2 also shows the experimental
CO decay curve over time in the chamber. Because CO
is relatively inert, losses in the chamber may be due to
dilution only. e experimental CO decay curve was
exponential, the k value was calculated to be k = 4.0 × 10−3
min−1. e excellent agreement between the experi-
mental CO decay curve and the theoretical curve due to
dilution only conrmed that this chamber behaved as a
well-mixed ow reactor.
e decay of the particle mass and particle count as
measured by the SMPS and the CPC, respectively, are
also shown in Figure 2. e experimental particle mass
decay curve was exponential, with a value of k = 5.1 × 10−3
min−1. is larger value of k indicates that there are par-
ticle mass losses onto the chamber wall surfaces, in addi-
tion to dilution. e maximum observed losses of particle
mass in the chamber were around 22%. e experimental
particle count decay curve was also exponential, with a
value of k = 7.0 × 10−3 min−1. is even larger value of k
indicates that in addition to dilution and losses onto the
chamber wall surfaces, particle count losses may also be
Figure 2. Particle and CO decay curves in the chamber.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

Photochemical chamber for toxicity testing 501
© 2011 Informa Healthcare USA, Inc.
due coagulation of freshly formed ultrane particles to
form larger particles.
e particle count CPC predominantly measures
ultrane particles (below 100 nm), which can be lost by
Brownian diusion to chamber walls and by coagula-
tion with other particles, and by electrostatic eects.
e particle mass SMPS measurements are dominated
by somewhat larger particles (since mass is propor-
tional to the cube of the diameter). Like the ultrane
particles, these larger particles, having a diameter
between 100 nm and 1 μm, can be lost be electrostatic
eects (McMurry & Rader, 1985), but with lower losses
by diusion and coagulation, and, in addition, they
can be lost due to gravitational settling (Crump &
Seinfeld, 1981; Pierce et al., 2008). ese mechanisms
can explain why the count and mass decay curves show
more losses than by dilution alone. Ruiz et al. (2007a)
used results from his pilot chamber particle losses tests
using an articial monodisperse aerosol to mathemati-
cally estimate the particle losses in the chamber using
a simple box model and assuming a well-mixed ow
reactor. Using this approach, Ruiz et al. (2007a) esti-
mated losses of 32, 47, and 23% for particles around 50,
100, and 500 nm in size, respectively.
Gas-phase reactions
Tests with car exhaust diluted to 10 and 20 ppm of CO
were conducted at 20 L min−1 (corresponding to a cham-
ber mean residence time of approximately 30 min).
Once a steady-state CO concentration was reached in
the chamber, the lamps were turned on to initiate pho-
tochemical reactions, quickly followed by formation of
SOA in the chamber.
Photochemical oxidation of the diluted car exhaust
followed a fairly typical pattern, with rapid conversion
of NO to NO2, followed by accumulation of O3. Typical
results from an experiment with 20 ppm of CO introduced
in the chamber are presented in Figure 3.
Only a few minutes after the lamps were turned on,
the initial amount of NO was completely titrated and
NO2 (measured by the chemiluminescence analyzer as
NOx–NO) started forming rapidly. O3 accumulated rap-
idly in the chamber reaching an average of 350 ppb after
240 min. e NOx–NO concentration that included a high
fraction in the form of NO2 was on average 150 ppb over
the course of the experiment.
Note that free radical formation in our chamber was
initiated without addition of any accelerant. In the Ruiz
et al. (2007a) study, as there were virtually no VOCs in
the diluted power plant stack gas, it was necessary to add
O3 to rst titrate the relatively high amount of NO. Also,
O3 was added to produce high concentrations of ·OH (by
photolysis of O3 using UVB-313 lamps) in order to oxidize
sulfur dioxide (SO2) in coal-red power plant emissions.
e reactions in our chamber resulted in a relatively
high O3 accumulation without any accelerant because
the amounts of aromatic hydrocarbons, among the total
VOCs, were relatively high. ese compounds, as well as
their major photochemical oxidation products, undergo
photolysis to generate free radicals (Seinfeld & Pandis,
2006).
SOA formation
Under the same conditions discussed above for the
gas-phase reactions (20 ppm CO of car exhaust diluted
with outdoor air and a chamber mean residence time
of 30 min, the chamber achieved a relatively stable SOA
mass concentration of approximately 40 μg m−3 after
5–6 h of irradiation time. Figure 4 shows the average SOA
mass concentration and its variability (standard error)
Figure 3. Gas-phase reactions in the well-mixed ow reactor.
Figure 4. Secondary organic aerosol formation at 20 ppm of CO
and a mean residence time of 30 min. SOA, secondary organic
aerosol.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

502 V. Papapostolou et al.
Inhalation Toxicology
from 5 experiments conducted under the same experi-
mental conditions.
In a related set of experiments, the CO introduced into
the chamber was adjusted to be approximately 10 ppm,
while the mean residence time in the chamber was
maintained at 30 min. With CO concentration reduced
to one half as much, the SOA yield was on average 10 μg
m−3 lower than the SOA formed at 20 ppm CO, reaching
only 30.2 ± 6.1 μg m−3. We observed that a decrease in the
amount of CO resulted in a decrease in the amount of
SOA formed in the chamber. is may be due to the cor-
responding decreased amount of VOCs concentrations
in the chamber. In such case, this is conrmation of our
assumption (made above) that CO serves as a surrogate
for VOCs.
Residence/irradiation time effect
Figure 5 shows the eect of residence/irradiation time
on the evolution of the secondary particle size distribu-
tion. After a steady-state CO concentration of 20 ppm was
reached, the chamber was capped to simulate a static
condition and the lamps were turned on. e SMPS was
used to sample every hour at 0.3 L min−1. In the absence
of seed aerosol, the initial burst in the rst hour with
particles forming in the ultrane region around 30 nm is
mainly due to nucleation (Cocker et al., 2001b).
e evolution of the particle count size distribution
indicates that there was a slow coagulation process, as
well as particle growth due to condensation/partitioning
of freshly formed oxidized (semivolatile) organics onto
surfaces, during the course of 0–7 h of elapsed residence/
irradiation time.
During the same static chamber experiment, we also
examined the eect of residence/irradiation time on
secondary aerosol mass formation. Figure 6 shows that
the highest SOA mass yield occurs approximately 2 h
after irradiation. As suggested by Figure 5, the dominant
SOA particle size formation mechanisms in the labora-
tory chamber were most likely nucleation and subse-
quent slow coagulation and condensation. ese results
indicate that a longer residence/irradiation time (longer
than one hour) might be needed to achieve total mass
concentrations that would be adequate for toxicological
testing.
VOCs consumption
e fractions of selected gasoline-associated VOCs that
reacted in the chamber with the lamps on ranged from
3.8 to 19.9% (see Table 1). It has been shown that a given
aromatic compound may have higher reactivity than
some other aromatics, but lower SOA-forming poten-
tial. For example, Carter (1994) has shown that ben-
zene has the lowest relative reactivity in the adjusted
NOx scales (3.8% reacted in our chamber), while
m-xylene has the highest (almost 20% reacted in our
chamber). Although our results show that more of the
m-xylene than toluene reacted in the chamber, Odum
et al. (1997b) have shown that toluene results in higher
SOA yields than m-xylene. e chamber was operated
in a dynamic mode with a short mean residence time.
Because fresh VOCs were being added constantly and
the rate of VOCs reactions was lower than the rate of
VOCs replenishment, the concentrations downstream
would be relatively high.
EC/OC formation
Results from the EC/OC analysis are presented in Table 2.
Four OC fractions at temperature ramping between
∼25°C and 550°C are reported. e average dierences
in total OC concentrations between downstream and
upstream concentrations for all three sets of experi-
ments clearly indicate the there is SOA formation in the
Figure 5. e eect of residence/irradiation time on particle size. Figure 6. e eect of residence/irradiation time on particle
mass.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

Photochemical chamber for toxicity testing 503
© 2011 Informa Healthcare USA, Inc.
chamber during irradiation of vehicular primary gas-
phase precursors. Shorter mean residence time (30 min)
in the chamber results in higher particle mass formation.
e substantially greater amounts downstream of the
chamber of OC2 and OC3 fractions, for temperatures
between 120°C and 4500°C, show that the secondary
aerosol has signicantly more very low-volatility organic
species than the primary aerosol. is material generally
has higher molecular weight and is more polar than the
higher volatility species that are released at lower tem-
peratures, consistent with the expected properties of
SOA compounds.
Seed aerosol effect
To test for the eect of seed aerosol surface on the forma-
tion of SOA, 12 experiments were conducted using HGS
as the seed aerosol and another 12 experiments were
conducted using MSHA. For this purpose, the experi-
ments were conducted by varying the seed aerosol con-
centration between low and high in combination with
variation in the mean residence time in the chamber and
in the CO concentration from the diluted car exhaust.
e primary particle mass concentration emitted by the
car was negligible.
In the seed aerosol experiments (see Table 3), following
stabilization of the CO in the chamber (ranged between
2.0 ppm and 26.4 ppm), addition of the seed aerosol was
started. When a steady-state baseline seed aerosol mass
concentration was reached (ranged between 42.4 μg m−3
and 804.1 μg m−3), the lamps were turned on to initiate
the photochemical reactions. After a 180-min irradia-
tion with the lamps on, the SOA formed in the chamber
was measured (ranged between 21.5 μg m−3 and 437.8 μg
m−3). e mean residence time in the chamber was set at
either 50 or 100 min.
e relationship between the steady-state SOA yield
concentration (Cs) formed in our chamber with the lamps
on and the varying parameters in our tests was examined.
We developed an empirical equation by linear regression
(SAS Institute Inc., Cary, NC) that shows a logarithmic
relationship between SOA and (1) the seed aerosol mass
concentration, (2) the CO concentration (surrogate for
VOCs concentration), with an R2 = 0.84. e type of seed
aerosol (hollow glass spheres or MSHA) was not signi-
cant when included in the model. It should be noted that
both types of seed aerosol used are inert aerosols. A reac-
tive seed aerosol might behave dierently. In this work,
the SOA yield formed was only dependent on the amount
of seed aerosol used. e empirical model is shown in
Equation (9):
CCC
SOAseed CO
=+⋅⋅ +⋅⋅
−−
exp( .. .)347305 10 28610
32
(8)
where CSOA: mass concentration (μg m−3); Cseed: seed aero-
sol mass concentration (μg m−3); CCO: carbon monoxide
concentration (ppm).
Seed aerosol addition had a signicant eect on the
SOA yield formed. As shown earlier, at an average of
10 ppm of CO in the chamber, the experiments with-
out seed aerosol yielded an average 30 μg m−3 of SOA
formed, while the seed aerosol experiments with the
same CO concentration yielded approximately an aver-
age of 70 μg m −3. e same eect was observed at a higher
concentration of approximately 20 ppm of CO. At that
concentration, the experiments without seed aerosol
yielded an average of 40 μg m−3 of SOA formed, while the
seed aerosol experiments yielded an average of 105 μg
m−3 of SOA formed. is was as expected. In the experi-
ments with no seed aerosol, the primary mechanisms
for particle formation are nucleation and coagulation of
the ultrane particles (Cocker et al., 2001a), consistent
with the evolution of the particle size distribution with
increasing residence/irradiation time shown in Figure 5.
e seed particle surface area competes with the cham-
ber walls for vehicular gases to condense on particles or
become lost on the walls, respectively. e higher the
seed particle concentration, the greater the surface area
and consequently the higher the gas molecules concen-
tration will condense on the particles. e mean resi-
dence time did not seem to play an important role in the
formation of SOA for the experiments with either type of
seed aerosol. Having the seed aerosol provide a surface
for condensation of gas-phase species minimizes the
roles of nucleation and coagulation eects that make
the mean residence time an important parameter in the
chamber tests described above with the car exhaust only
and without seed aerosol.
Table 3. Seed aerosol experimental data.
CO
ppm
Seed aerosol SOA Residence time
type μg m−3 μg m−3 min
2.0 HGS 294.8 48.9 100
3.0 MSHA 603.0 222.7 50
3.1 MSHA 60.2 21.5 50
4.3 HGS 42.4 35.0 100
4.6 HGS 341.0 72.7 50
5.2 MSHA 129.0 70.7 50
5.5 MSHA 128.3 77.6 50
5.9 MSHA 636.0 378.3 50
6.2 HGS 62.0 46.8 50
6.9 MSHA 804.1 437.8 50
6.9 MSHA 119.9 67.9 50
7.3 MSHA 140.9 69.0 50
7.4 MSHA 132.2 77.2 50
9.0 MSHA 218.6 87.9 50
9.0 MSHA 156.0 92.2 50
9.3 MSHA 626.6 329.2 50
9.5 HGS 104.8 89.2 50
11.3 HGS 154.5 71.7 50
11.8 HGS 170.5 64.0 100
13.7 HGS 174.2 71.7 100
20.7 HGS 228.5 103.0 50
20.9 HGS 62.3 73.5 100
26.0 HGS 370.0 143.7 100
26.4 HGS 83.7 101.5 50
HGS, hollow glass sphere; MSHA, Mt Saint Helens Ash; SOA,
secondary organic aerosol.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

504 V. Papapostolou et al.
Inhalation Toxicology
In vivo chemiluminescence
Exposure doses for these toxicological studies were
MSHA = 363.0 ± 66.0 μg m−3 and MSHA+SOA = 211.5 ± 95.3
µg m−3. In vivo chemiluminescence was assessed at
the surface of the heart and lung. No dierences were
observed between ltered air controls and either primary
or secondary particle exposures for in vivo chemilumi-
nescence of the heart (data not shown). However, in vivo
chemiluminescence of the lung showed clear dierences
between exposures. ere was no signicant dierence
between ltered air controls and exposures with lamps
o in the MHSA scenario conrming the lack of toxicity
with MHSA aerosols. With lamps on and with the for-
mation of secondary particles on the MSHA particles, a
signicant increase in lung chemiluminescence versus
ltered air controls was observed as well as between sce-
narios as shown in Figure 7.
Conclusions
e laboratory chamber tests investigated the photo-
chemical reactivity of car exhaust from a single vehicle,
diluted with ambient air. Once the UV lamps were turned
on, photochemical reactions were initiated, without the
addition of any accelerant, resulting in a rapid NO deple-
tion, NO2 formation, and O3 accumulation. We deter-
mined the conditions needed to produce an adequate
amount of SOA mass for use with an in-laboratory ani-
mal toxicological study (complete manuscript in prepa-
ration), and in this report we show a signicant response
to SOA in the lungs of normal rats at a concentration
less than that of the MSHA without formation of SOA. A
stable SOA concentration of approximately 40 μg m−3 at
a chamber mean residence time of 30 min was achieved.
is relatively short mean residence time provided
adequate chamber ow output for both particle char-
acterization and animal exposures. e chamber was
operated as a continuous ow reactor for toxicological
tests with animal exposures. e information gathered
from this laboratory evaluation was used to inform the
design and development of a signicantly larger reac-
tion chamber that would generate SOA mass from urban
highway tunnel eet emissions in a real eld setting. e
eld project has been completed and the manuscript is
in preparation.
Acknowledgements
is publication was made possible by USEPA grants
R-832416 and RD 83479801. Its contents are solely the
responsibility of the grantee and do not necessarily rep-
resent the ocial views of the USEPA. Further, USEPA
does not endorse the purchase of any commercial prod-
ucts or services mentioned in the publication. is work
was also supported by the Harvard NIEHS Center for
Environmental Health (grant P30ES000002). We thank
Dr. Stephen Rudnick for reading the manuscript and
providing helpful suggestions. We also thank Dr. Choong-
Min Kang and Samuel Pueringer for the analytical sup-
port. Technical assistance was provided by John Tosti,
Vivian Hatakeyama, Brenno Gomes, and Yasser Calil.
Declaration of interest
e authors report no declarations of interest.
References
Carter WPL. 1994. Development of ozone reactivity scales for volatile
organic compounds. J Air Waste Manage Assoc 44:881–899.
Carter WPL, Luo D, Malkina IL, Pierce JA. 1995. Environmental
chamber studies of atmospheric reactivities of volatile organic
compounds. Eects of varying chamber and light source. Final
Report.
Chow JC, Watson JG, Pritchett LC, Pierson WR, Frazier CA, Purcell RG.
1993. e DRI thermal/optical reectance carbon analysis system:
description, evaluation and applications in U.S. air quality studies.
Atmos Environ 27A:1185–1201.
Cocker DR III, Mader BT, Kalberer M, Flagan RC, Seinfeld JC. 2001a.
e eect of water on gas-particle partitioning of secondary organic
aerosol, II m-Xylene and 1,3,5-trimethylbenzene photooxidation
systems. Atmos Environ 35:6073–6085.
Cocker DR 3rd , Flagan RC, Seinfeld JH. 2001b. State-of-the-art chamber
facility for studying atmospheric aerosol chemistry. Environ Sci
Technol 35:2594–2601.
Crump JG, Seinfeld JH. 1981. Turbulent deposition and gravitational
sedimentation of an aerosol in a vessel of arbitrary shape. J Aerosol
Sci 12:405–415.
Dockery DW, Pope CA 3rd. 1994. Acute respiratory eects of particulate
air pollution. Annu Rev Public Health 15:107–132.
Dockery DW, Pope CA 3rd, Xu X, Spengler JD, Ware JH, Fay ME, Ferris
BG Jr, Speizer FE. 1993. An association between air pollution and
mortality in six U.S. cities. N Engl J Med 329:1753–1759.
Feilberg A, Nielsen T, Binderup ML, Skov H, Poulsen MWB. 2002.
Observations of the eect of atmospheric processes on the
Figure 7. e dierence between MSHA+SOA and ltered air
controls and the dierences between scenarios are both p < 0.05.
MSHA, Mt Saint Helens Ash; SOA, secondary organic aerosol.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.

Photochemical chamber for toxicity testing 505
© 2011 Informa Healthcare USA, Inc.
genotoxic potency of airborne particulate matter. Atmos Environ
36:4617–4625.
Finlayson-Pitts BJ, Pitts JN. 1986. Atmospheric Chemistry. New York:
John Wiley & Sons.
Fraser MP, Cass GR, Simoneit BRT. 1998. Gas-phase and particle-
phase organic compounds emitted from motor vehicle trac in a
Los Angeles roadway tunnel. Environ Sci Technol 32:2051–2060.
Godleski JJ, Diaz EA, Lemos M, Long M, Ruiz P, Gupta T, Kang CM,
Coull B. 2011. Toxicological Evaluation of Realistic Emission
Source Aerosols (TERESA)-power plant studies: assessment of
cellular responses. Inhal Toxicol in press.
Godleski JJ, Koutrakis P, Kang CM, Diaz EA, Rohr AC. 2011.
Toxicological Evaluation of Realistic Emission Source Aerosols
(TERESA): Introduction and overview. Inhal Toxicol in press.
Gurgueira SA, Lawrence J, Coull B, Murthy GG, González-Flecha B.
2002. Rapid increases in the steady-state concentration of reactive
oxygen species in the lungs and heart after particulate air pollution
inhalation. Environ Health Perspect 110:749–755.
Izumi K, Fukuyama T. 1990. Photochemical aerosol formation from
aromatic-hydrocarbons in the presence of NOx. Atmos Environ
24:1433–1441.
Kang CM, Gupta T, Ruiz PA, Wolfson JM, Ferguson ST, Lawrence JE,
Rohr AC, Godleski J, Koutrakis P. 2011. Aged particles derived from
emissions of coal-red power plants: the TERESA eld results.
Inhal Toxicol in press.
Kirchstetter TW, Singer BC, Harley RA, Kendall GR, Chan W. 1996.
Impact of oxygenated gasoline use on California light duty vehicle
emissions. Environ Sci Technol 30:661–670.
Kleindienst TE, Conver TS, McIver CD, Edney EO. 2004. Determination
of secondary organic aerosol products from the photooxidation of
toluene and their implications in ambient PM2.5. J Atmos Chem
47:79–100.
Kleindienst TE, Smith DF, Li W, Edney EO, Driscoll DJ, Speer RE,
Weathers WS. 1999. Secondary organic aerosol formation from
the oxidation of aromatic hydrocarbons in the presence of dry
submicron ammonium sulfate. Atmos Environ 33:3669–3681.
Laden F, Neas LM, Dockery DW, Schwartz J. 2000. Association of ne
particulate matter from dierent sources with daily mortality in six
U.S. cities. Environ Health Perspect 108:941–947.
Lim HJ, Turpin BJ. 2002. Origins of primary and secondary organic
aerosol in Atlanta: results of time-resolved measurements during the
Atlanta Supersite Experiment. Environ Sci Technol 36:4489–4496.
McDonald JD, Harrod KS, Seagrave J, Seilkop SK, Mauderly JL. 2004.
Eects of low sulfur fuel and a catalyzed particle trap on the
composition and toxicity of diesel emissions. Environ Health
Perspect 112:1307–1312.
Ng NL, Kroll JH, Keywood MD, Bahreini R, Varutbangkul V, Flagan
RC, Seinfeld JH, Lee A, Goldstein AH. 2006. Contribution of rst-
versus second-generation products to secondary organic aerosols
formed in the oxidation of biogenic hydrocarbons. Environ Sci
Technol 40:2283–2297.
Nikolaou K, Masclet P, Mouvier G. 1984. Sources and chemical reactivity
of polynuclear aromatic hydrocarbons in the atmosphere? A
critical review. Sci Total Environ 32:103–132.
Odum JR, Jungkamp TP, Grin RJ, Flagan RC, Seinfeld JH. 1997. e
atmospheric aerosol-forming potential of whole gasoline vapor.
Science 276:96–99.
Odum JR, Jungkamp TPW, Grin RJ, Forstner HJL, Flagan RC ,
Seinfeld JH. (1997b). Aromatics, reformulated gasoline, and
atmospheric organic aerosol formation. Environ Sci Technol
31:1890–1897.
Peters A, von Klot S, Heier M, Trentinaglia I, Hörmann A, Wichmann
HE, Löwel H; Cooperative Health Research in the Region of
Augsburg Study Group. 2004. Exposure to trac and the onset of
myocardial infarction. N Engl J Med 351:1721–1730.
Pierce JR, Engelhart GJ, Hidebrandt L, Weitkamp EA, Pathak
RK, Donahue NM, Robinson AL, Adams PJ, Pandis SN. 2008.
Constraining particle evolution from wall losses, coagulation and
condensation-evaporation in smog chamber experiments: optimal
estimation based on size distribution measurements. Aerosol Sci
Technol 42:1001–1015.
Pierson WR, Gertler AW, Robinson NF, Sagebiel JC, Zelinska B,
Bishop GA, Stedman DH, Zweidinger RB, Ray WR. 1996. Real-
world automotive emissions—summary of studies in the Fort
McHenry and Tuscarora mountain tunnels. Atmos Environ
30:2233–2256.
Pope CA 3rd, Burnett RT, un MJ, Calle EE, Krewski D, Ito K, urston
GD. 2002. Lung cancer, cardiopulmonary mortality, and long-
term exposure to ne particulate air pollution. JAMA 287:1132–
1141.
Rogak SN, Pott U, Dann T, Wang D. 1998. Gaseous emissions from
vehicles in a trac tunnel in Vancouver, British Columbia. J Air
Waste Manag Assoc 48:604–615.
Ruiz PA, Gupta T, Kang CM, Lawrence JE, Ferguson ST, Wolfson JM,
Rohr AC, Koutrakis P. 2007. Development of an exposure system
for the toxicological evaluation of particles derived from coal-red
power plants. Inhal Toxicol 19:607–619.
Ruiz PA, Lawrence JE, Ferguson ST, Wolfson JM, Koutrakis P. 2006. A
counter-current parallel-plate membrane denuder for the non-
specic removal of trace gases. Environ Sci Technol 40:5058–
5063.
Ruiz PA, Lawrence JE, Wolfson JM, Ferguson ST, Gupta T, Kang CM,
Koutrakis P. 2007. Development and evaluation of a photochemical
chamber to examine the toxicity of coal-red power plant
emissions. Inhal Toxicol 19:597–606.
Sander SP, Friedl RR, Golden DM, Kurylo MJ, Huie RE, Orkin VL,
Moortgat GK, Ravishankara AR, Kolb CE, Molina MJ, Finlayson-
Pitts BJ. 2003. Chemical kinetics and photochemical data for use in
atmospheric studies. Evaluation no. 14, Jet Propulsion Laboratory
publication 02-25.
Savage ST, Lawrence J, Katz T, Stearns RC, Coull BA, Godleski JJ. 2003.
Does the Harvard/U.S. Environmental Protection Agency Ambient
Particle Concentrator change the toxic potential of particles? J Air
Waste Manag Assoc 53:1088–1097.
Schwartz J. 2004. Air pollution and children’s health. Pediatrics
113:1037–1043.
Seagrave J, Mauderly JL, Seilkop SK. 2003. In vitro relative toxicity
screening of combined particulate and semivolatile organic
fractions of gasoline and diesel engine emissions. J Toxicol Environ
Health Part A 66:1113–1132.
Seinfeld JH, Pandis SN. 2006. Atmospheric chemistry and physics.
From air pollution to climate change. New York: John Wiley &
Sons.
Zafonte L , Rieer PL, Holmes Jr. 1977. Nitrogen-dioxide photolysis in
Los Angeles atmosphere. Environ Sci Technol 11:483–487.
Zielinska B, Samy S, McDonald JD, Seagrave J; HEI Health Review
Committee. 2010. Atmospheric transformation of diesel emissions.
Res Rep Health E Inst: 5–60.
Inhalation Toxicology Downloaded from informahealthcare.com by 71.174.46.132 on 06/20/11
For personal use only.