University of Southern California Dissertations and Theses

Investigating the temporal trends, sources, and toxicity of ambient particulate matter (PM) in different metropolitan environments, and development of a novel aerosol generation setup for inhalat...

(USC Thesis Other)

Investigating the temporal trends, sources, and toxicity of ambient particulate matter (PM) in different metropolitan environments, and development of a novel aerosol generation setup for inhalat...

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content Investigating the temporal trends, sources, and toxicity of ambient
particulate matter (PM) in different metropolitan environments, and
development of a novel aerosol generation setup for inhalation exposure
studies
by
Sina Taghvaee
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the Requirements for the Degree
DOCTOR OF PHILOSOPHY
Engineering (Environmental Engineering)
August 2020
Copyright 2020 Sina Taghvaee
ii
Dedication

To my parents for their love and unconditional support throughout my life. Thank you both for
giving me strength to reach for the stars and chase my dreams ….

iii
Acknowledgements

This research was financially supported by the Institute for Environmental Research (IER),
Tehran University of Medical Science (grant number 90-03-46-15705), the United States
National Institutes of Health (grant numbers: 2R01AI065617-17A1, 5R01ES024936-04,
1RF1AG051521-01 and 1R01ES024936-01), and US Department of Defense (US-Army
Medical Research Acquisition Activity, grant numberW81XWH-17-1-0535). The authors are
also grateful to Viterbi School of Engineering, University of Southern California Ph.D.
fellowship award.

I would like to express my deepest gratitude to my supervisor, Professor Constantinos Sioutas,
whose mentorship and support helped me significantly for preparing this work.

I would also like to sincerely thank the following people and groups that helped us during
experimental measurements, and chemical analysis.

Wisconsin State Lab of Hygiene (WSLH):
Dr. James J. Schauer
Dr. Martin M. Shafer

Institute of Nuclear and Radiological Sciences & Technology, Energy & Safety, N.C.S.R.
Demokritos:
Dr. Evangelia Diapouli
Dr. Manousos Ioannis Manousakas
Dr. Vasiliki Vasilatou
Dr. Kostas Eleftheriadis

Center for Air Pollution Research (CAPR), Institute for Environmental Research (IER), Tehran
University of Medical Sciences:
Dr. Mohammad Sadegh Hassanvand
iv
Department of Environmental Health Engineering, School of Public Health, Tehran University
of Medical Sciences:
Dr. Masud Yunesian
Dr. Kazem Naddafi

I would also like to extend my gratitude to my former and current colleagues and group mates
because of their sincere help and support in the research projects that I have been involved with
during my almost three years as a PhD student at USC:

Dr. Mohammad H. Sowlat
Dr. Amirhosein Mousavi
Ehsan Soleimanian
Milad Pirhadi
Abdulmalik Altuwayjiri

And, finally, I am grateful of my qualifying and defense exam committee members:

Dr. Constantinos Sioutas (Chair)
Dr. George Ban-Weiss
Dr. William Mack
Dr. Kelly Sanders
Dr. Amy Childress
Dr. Caleb Finch

v

Table of Contents

Dedication
Acknowledgements
List of Tables
List of Figures
Abstract

Chapter 1: Introduction

1.1. Background
1.2. Overview

Chapter 2: Source apportionment of ambient PM 2.5 in two locations in central Tehran using the
Positive Matrix Factorization (PMF) model

2.1. Introduction
2.2. Methodology
2.2.1. Sampling site
2.2.2. Sampling period and instrumentation
2.2.3. PMF model
2.3. Results and discussion
2.3.1. Overview of the data
2.3.2. Number of Factors
2.3.3. Factor identification
2.3.3.1. Vehicular emissions
2.3.3.2. Secondary aerosol
2.3.3.3. Industrial emissions
2.3.3.4. Biomass burning
2.3.3.5. Soil
2.3.3.6. Road dust
2.4. Summary and conclusions

Chapter 3: Source-specific lung cancer risk assessment of ambient PM 2.5-bound Polycyclic
Aromatic Hydrocarbons (PAHs) in central Tehran
3.1. Introduction
3.2. Methodology
3.2.1. Sampling Locations
3.2.2. Sampling schedule, methodology and instrumentation
3.2.3. PMF model
3.2.4. Cancer risk characterization
3.3. Results and discussion
3.3.1. Overview of the data
3.3.2. Number of Factors
3.3.3. Factor identification
3.3.3.1. Petrogenic sources and petroleum residue
3.3.3.2. Natural gas and biomass burning
3.3.3.3. Industrial emissions
3.3.3.4. Diesel exhaust emissions
ii
iii
vii
viii
x

1

1
3

8

8
12
12
14
15
17
17
19
29
29
30
31
33
34
35
36

38

38
42
42
42
44
48
51
51
53
58
58
59
59
60
vi

3.3.3.5. Gasoline exhaust emissions
3.3.4. Source-specific lung cancer risk assessment
3.3.5. Limitations of the study
3.4. Summary and conclusions

Chapter 4: Source apportionment of the oxidative potential of fine ambient particulate matter
(PM 2.5) in Athens, Greece

4.1. Introduction
4.2. Experimental methodology
4.2.1. Sampling site
4.2.2. Sample collection and analysis
4.2.3. Determination of PM oxidative potential via the DCFH assay
4.2.4. Source Apportionment of PM 2.5 and its associated oxidative potential
4.3. Results and Discussion
4.3.1. Mass concentration and chemical composition of PM 2.5
4.3.1.1. Concentrations of PM 2.5 mass and carbonaceous species
4.3.1.2. Levoglucosan Concentrations
4.3.1.3. Metals and trace elements
4.3.2. Oxidative potential of PM 2.5
4.3.3. Source apportionment of ambient PM 2.5 and its associated oxidative
Potential
4.3.3.1. Source apportionment of PM 2.5 mass concentration using the PCA
approach
4.3.3.2. Correlation analysis between individual chemical species and PM
oxidative potential
4.3.3.3. Source apportionment of PM oxidative potential using MLR approach
4.4. Summary and conclusions

Chapter 5: Development of a novel aerosol generation system for conducting inhalation
exposures to ambient particulate matter (PM)

5.1. Introduction
5.2. Methodology
5.2.1. PM collection
5.2.1.1. High-volume sampler PM collection and filter extraction
5.2.1.2. PM collection using VACES/aerosol-into-liquid tandem
5.2.2. Aerosol generation and inhalation exposure
5.2.3. Chemical analysis
5.3. Results and discussion
5.3.1. Physical properties of the ambient vs. re-aerosolized PM
5.3.2. Chemical composition of the ambient vs. re-aerosolized PM
5.3.2.1. Mass balance for bulk chemical components
5.3.2.2. Metals and trace elements
5.3.2.3. Polycyclic Aromatic Hydrocarbons (PAHs)
5.4. Summary and conclusion
Chapter 6: Conclusions and future research
6.1. Conclusions
6.2. Future research
References
61
62
66
68

70

70
73
73
73
75
76
78
78
78
82
83
85

88

88

89
90
94

96

96
99
99
99
100
101
104
104
104
110
110
113
117
118
120

120
122
125
vii

List of Tables

Table 2.1. Summary statistics and mass fractions of the measured chemical components in Tohid
retirement home

Table 2.2. Summary statistics and mass fractions of the measured chemical components in the
school dormitory

Table 3.1. Statistical characteristics of chemically analyzed ambient PAHs concentrations in both
sampling sites

Table 3.2. Unit cancer risk factors for BaPeq lifetime and outdoor exposure (Bandowe et al., 2014;
OEHHA, 1994)

Table 3.3. Statistical characteristics of the chemically analyzed ambient particle-bound PAHs
concentration

Table 3.4. Seasonal average of meteorological parameters in central Tehran during the study period
(Errors correspond to one standard deviation (SD))

Table 3.5. Total BaP eq and lung cancer risk associated with outdoor exposure (Errors correspond
to one standard deviation (SD))

Table 3.6. Total BaPeq and lung cancer risk associated with lifetime exposure (Errors correspond
to one standard deviation (SD))

Table 3.7. Comparison of total BaP(eq) results with previous studies (Errors correspond to one
standard deviation (SD))

Table 4.1. The overall and seasonal averages (± standard deviation) of PM 2.5 mass concentrations
as well as its associated oxidative potential and chemical components

Table 4.2. Seasonal averages (± standard deviation) of meteorological parameters during the warm
and cold period

Table 4.3. Loadings of chemical species in the factors resolved by the principal component analysis
(PCA). Loadings> 0.6 are bolded
Table 4.4. Spearman bivariate correlation analyses between volume-based oxidative potential (µg
Zymosan/m
3
) and weekly sampled concentrations of individual chemical species. Highly
correlated species are bolded, and significant values (P<0.05) are marked with an asterisk (*)
Table 4.5. Results of the multiple linear regression (MLR) analysis between a) PM 2.5 oxidative
potential (as the dependent variable) and selected chemical species (as independent variables); b)
OC (as the dependent variable) and selected chemical species (as independent variables)

18

19

47

49

42

57

63

63

66

80

80

88

89

91

viii

List of Figures

Fig. 2.1. The location of sampling sites in central Tehran
Fig. 2.2. Correlations between the predicted and measured PM 2.5 concentration for: a) Tohid
retirement home; and b) school dormitory
Fig. 2.3. PMF-resolved factor profiles and their temporal trends in Tohid retirement home
Fig. 2.4. PMF-resolved factor profiles and their temporal trends in school dormitory
Fig. 2.5. Relative concentrations of sources to PM 2.5 mass concentrations in Tohid retirement
home: a) overall; c) warm phase; and e) cold phases and school dormitory: b) overall; d) warm
phase; and f) cold phase
Fig.2.6. Temporal trends of factor contributions to PM 2.5 mass concentrations in the cold and warm
phases for: a) Tohid retirement home; and b) school dormitory (The box plots show the inter
quartile range along with upper and lower quartile; dots show the minimum and maximum values;
and median of data are also presented as a horizontal line in the box)
Fig. 2.7. Metrological parameters including: a) monthly average temperature; and b) monthly
average wind speed in Tehran during the study period
Fig. 2.8. Correlation analysis between: a) contributions of vehicular emissions and EC in Tohid
retirement home; b) contributions of vehicular emissions and OC in Tohid retirement home; c)
contributions of vehicular emissions and EC in school dormitory; d) contributions of vehicular
emissions and OC in school dormitory
Fig. 2.9. Correlation analysis between: a) contributions of secondary aerosols and OC in Tohid
retirement home; and b) contributions of secondary aerosols and OC in school dormitory
Fig. 3.1. Sampling locations in Tehran
Fig. 3.2. Correlation between the chemically analyzed and predicted total ambient PAHs mass
concentrations by the PMF model
Fig. 3.3. Q robust values for different PMF solutions
Fig. 3.4. PMF source profiles for 5-factor solution
Fig. 3.5. Relative contribution of PMF resolved factor profiles to total ambient PAHs mass
concentration for the whole study period
Fig. 3.6. Temporal variation in relative contribution of the PMF resolved factor profiles to the total
PAHs
Fig. 3.7. The relative contribution of different PAH sources to lung cancer risk

13

17

21

23

26

27

27

28

29

44

53

54

55

56

57

64

ix

Fig. 4.1. The average mass concentrations of a) PM 2.5 during the entire study period, and by season;
b) carbonaceous components (elemental carbon (EC), organic carbon (OC), and water soluble
organic carbon (WSOC)) by season. Error bars correspond to one standard deviation (SD)

Fig. 4.2. Average levoglucosan concentrations during the entire study period, and by season. Error
bars correspond to one standard deviation (SD)

Fig.4.3. Average concentrations of metal and trace elements by season. Error bars correspond to
one standard deviation (SD)

Fig. 4.4. PM 2.5 oxidative potential for warm and cold phases: a) Mass-normalized oxidative
potential; b) Volume-based oxidative potential. Error bars correspond to one standard deviation
(SD)

Fig.4.5.Relative contribution of: a) EC and OC to PM 2.5 oxidative potential b) different sources to
OC mass concentrations; and c) different sources to PM 2.5 associated oxidative potential

Figure 5.1. Schematic of: a) high-volume sampler with PM 10 inlet, and ultrafine (UFP) impactor;
b) the VACES, coupled with high-volume aerosol-into-liquid collector; and c) aerosol generation
setup for filter collection and inhalation exposure

Figure 5,2. Number size distributions of re-aerosolized particles as a function of a) nebulizer’s
compressed air pressure; and b) dilution flow rate

Figure 5.3. Total a) number; and b) mass concentrations of re-aerosolized particles as a function
of nebulizer's compressed air pressure

Figure 5.4. Total a) number; and b) mass concentrations of nebulized particles as a function of
dilution flow rate

Figure 5.5. Typical number size distribution of ambient PM at central Los Angeles obtained during
our field tests

Figure 5.6. Chemical composition of re-aerosolized versus ambient a) PM 0.18; and b) PM 2.5

Figure 5.7. Correlation analysis between the nebulized and ambient mass ratios of metals and trace
elements in a) PM 0.18; and b) PM 2.5

Figure 5.8. Comparison of the redox active metals ratio in (a) PM0.18; and (b) PM2.5
Figure 5.9. Comparison of selected PAHs mass ratios in the ambient versus re-aerosolized: a)
PM 0.18; and b) PM 2.5

79

83

84

87

93

102

106

107

108

109

112

115

116

118

x

Abstract
Many metropolitan environments around the globes have been facing serious air pollution
challenges during the last decades due to rapid rate of industrialization and urbanization, resulting in a wide
variety of adverse health consequences, including neurodegenerative disorder, cardiovascular diseases,
respiratory problems and inflammation. While ambient particulate matters (PM) mass concentration are
regulated by the air quality agencies and health officials as one of the major criteria air pollutants, several
previous studies have indicated that ambient PM is consisted of various chemical components, each of them
being emitted from different pollution sources. In addition, there are a plethora of studies in the literature
documenting the higher toxicity of some PM compounds (e.g., redox active metals) in comparison to the
other chemical constituents. Therefore, it is vital to examine the association of PM toxicity with individual
components of PM (and in turn their emitting sources) rather than the total PM bulk mass. This would
definitely help the air quality authorities to develop more targeted control schemes in protecting people
from detrimental health impacts of exposure to ambient PM.
Consequently, the main objective of this dissertation is to investigate the major contributing sources
to PM mass concentration as well as its associated toxicity and health impacts. To this end, experimental
measurements of PM were conducted in Tehran, Athens, and Los Angeles as examples of highly polluted
and populated urban environments in three different continents (i.e., Asia, Europe, and America). The
collected samples were then analyzed for their chemical and toxicological compounds; followed by the
implementation of source apportionment techniques such as positive matrix factorization (PMF) receptor
model, principal component analysis (PCA), and multiple linear regression analysis to identify the sources
and quantify their relative contributions to the PM mass concentration, oxidative potential and health
consequences. Eventually, a novel aerosol generation setup was developed as the last part of this
dissertation to provide inhalation exposure studies with chemically and physically stable sources of aerosol
that perfectly resemble the complex physio-chemical characteristics of ambient PM. Hence, the outcomes
xi

of these researches are of great importance in prioritizing the necessary air quality policies to mitigate the
exposure of people to toxic PM compounds.

1
Chapter 1: Introduction
1.1. Background
Increased numbers of motor vehicles, rapid urbanization, and industrialization have
caused serious air pollution challenges in many urban environments around the globe during last
decades. More than 90% of the world population is exposed to air pollutant concentrations
exceeding WHO guideline limits (WHO, 2018). Among different air pollutants, ambient
particulate matter (PM) is of great importance due to its complex chemical composition as well
detrimental health impacts. Ambient PM is comprised of various chemical constituents,
including carbonaceous aerosols, metals and trace elements, inorganic ions, and organic species.
In addition based on several source apportionment studies, ambient PM can be originated from a
wide variety of sources, such as traffic related emissions, secondary aerosols, residential biomass
burning, ship emissions, soil, road dust, industrial emissions, and sea salt (Chen et al., 2018;
Gugamsetty et al., 2012; Guo et al., 2017; Hasheminassab et al., 2014a; Khodeir et al., 2012; Liu
et al., 2017; Mantas et al., 2014; Port et al., 2017). Therefore, the abovementioned characteristics
have motivated plenty of researchers to investigate different size ranges of ambient PM as one of
the most important criteria pollutants in metropolitan environments.
Several toxicological and epidemiological studies have confirmed the adverse health
consequences of exposure to ambient PM, including daily mortality, cardiovascular diseases,
neurodevelopmental and neurodegenerative disorders, and cardiopulmonary diseases (Davis et
al., 2013; Dockery and Stone, 2007; Gauderman et al., 2015; Mabahwi et al., 2014; Miller et al.,
2007; Morgan et al., 2011; Pope et al., 2004; Rich et al., 2013; Wai et al., 2015; Zhou et al.,
2014). Moreover, according to the most recent global burden of disease (GBD) study, 6.5 million
premature deaths were estimated to be due to exposure to various air pollutants in 2015, ranking
2
this risk factor above road accidents, malnutrition, AIDS, malaria, and tuberculosis as other
important causes of death worldwide (H. Wang et al., 2016). In addition among different air
pollutants, exposure to PM was associated with majority of air pollution related deaths, causing
4.5 million of premature deaths in 2015. This in fact corroborates the importance of further
researches regarding ambient PM physico-chemical characteristics, chemical composition,
sources, and toxicity.
However, it is interesting to note that different PM components in different size ranges
are not equally toxic; indicating that the total PM mass concentration is not the best parameter
for investigating adverse health effects of exposure to ambient PM. For instance, redox active
metals (e.g., Fe, Cu, and Mn) (Akhtar et al., 2010; Gasser et al., 2009), elemental carbon (EC)
(Cho et al., 2005; Samara, 2017), organic carbon (OC) (Chirizzi et al., 2017; Samara, 2017),
water soluble organic carbon (WSOC) (Bae et al., 2017; Vreeland et al., 2017), and PAHs (Cho
et al., 2005; Shirmohammadi et al., 2016) are some of the toxic components of PM associated
with high oxidative potential levels. On the other hand, secondary inorganic ions (e.g., secondary
ammonium nitrate and sulfate) are considered as rather harmless PM species, while they
contribute significantly to total PM mass concentrations. Thus, source apportionment studies are
beneficial in identifying different contributing sources to ambient PM, each having their own
chemical compositions and therefore toxicity. Results from these studies can be used by policy
makers to design air pollution control schemes for mitigating ambient PM pollutions and their
associated adverse health impacts in urban areas. In this regard, the presented studies in this
dissertation will aim to investigate and determine the temporal trends and sources of ambient PM
constituents and their associated toxicity in some of the highly crowded and polluted parts of the
world.
3
1.2. Overview
Our work begins with a study, in which the main contributing sources to ambient fine PM
(i.e., Particulate matters with aerodynamic diameters <2.5 µm) mass concentration were
determined for two sampling sites in central Tehran (i.e. Tohid retirement home and school
dormitory) by the means of Positive Matrix Factorization (PMF) model. To achieve this goal,
PM2.5, water soluble ions, and metals mass concentrations were implemented along with other
auxiliary variables such as elemental and organic carbon (EC/OC), and meteorological data for
specification and quantification of PM2.5 sources in the area. Eventually, a 7 factor solution was
selected as the most acceptable result for both locations based on the evaluation of the resulting
PMF source profiles, temporal trends of each factor in cold phase (i.e. Fall and winter) and warm
phase (i.e. spring and summer), correlation analysis between EC/OC data and PMF resolved
factor profiles, analysis of bootstrap runs, and R
2
values of predicted vs. measured PM2.5
concentration. The resolved factors included vehicular emissions, secondary aerosol, industrial
emissions (i.e. industrial emissions 1 and 2), biomass burning, soil, and road dust in Tohid
retirement home. On the other hand, vehicular emissions, secondary aerosol, industrial
emissions, biomass burning, soil, brake wear particles, and tire dust were the 7 factors resolved
by PMF for another sampling site. Results indicated that vehicular emissions were the dominant
contributors to PM2.5 in both sampling sites, with slightly higher contribution in school dormitory
(49.3%) compared to Tohid retirement home (48.8%). Secondary aerosol had also the second
highest percentage of contribution in both locations, with slightly higher contribution in Tohid
retirement home (28%) rather than school dormitory (24%) which might be reasonable due to its
more distance from major traffic flows. In addition, while two industrial factors were identified
in Tohid retirement home (with totally more than 17% contribution), only one industrial factor
4
(less than 2 % contribution) was recognized in another sampling site which might be due to the
fact that the retirement home is more impacted by industrial activities. The other resolved factor
profiles for Tohid retirement home were biomass burning, soil, and road dust with relative
contributions of 3%, 2.8% and less than 1% respectively. Biomass burning, soil, and the
remaining non-tailpipe road emissions (including brake wear particle and tire dust) were also
accounting for 16%, 8.2%, and less than 1% of total PM2.5 mass concentration in school
dormitory. Results of this study can be used as a beneficial tool for policy making purposes
regarding air quality improvement and addressing adverse health effects of exposure to ambient
PM2.5 in highly crowded parts of Tehran.
In the next study, we focused on Polycyclic Aromatic Hydrocarbons (PAHs) as one of
the well-known toxic constituents of PM in the same area. Hence, the source-specific lung cancer
risk characterization of ambient PM2.5-bound polycyclic aromatic hydrocarbons (PAHs) was
performed in central Tehran. Similar with the first study, the PMF model was implemented again
to determine the contributing sources to these detrimental organic compounds. Five factors were
identified as the major sources of airborne PAHs in central Tehran, including petrogenic sources
and petroleum residue, natural gas and biomass burning, industrial emissions, diesel exhaust
emissions, and gasoline exhaust emissions, with approximately similar contributions of around
20% to total PAHs concentration from each factor. Results of the PMF source apportionment
(i.e., PAHs factor profiles and contributions) were then used to calculate the source-specific lung
cancer risks for outdoor and lifetime exposure, using the benzo[α]pyrene (BaP) equivalent
method. Our risk assessment analysis indicated that the lung cancer risk associated with each
specific source is within the range of 10
−6
–10
−5
, posing cancer risks exceeding the United States
Environmental Protection Agency's (USEPA) guideline safety values (10
−6
). Furthermore, the
5
epidemiological lung cancer risk for lifetime exposure to total ambient PAHs was found to be
(2.8 ± 0.78)×10
−5
. Diesel exhaust and industrial emissions were the two sources with major
contributions to the overall cancer risk, contributing respectively to 39% and 27% of the total
risk associated with exposure to ambient PAHs. Results from this study provided an estimate of
the cancer risk caused by exposure to ambient PAHs in highly crowded areas in central Tehran,
and can be used as a guide for the adoption of effective air quality policies in order to reduce the
human exposure to these harmful organic species.
Afterward, we then switched the gears to investigate the PM2.5 toxicity in an urban
background site in Athens, the capital of Greece as one of the air pollution hotspots in Europe.
To this end, weekly time-integrated PM2.5 samples were collected during summer phase (June-
September of 2017) as well as winter period (February-March of 2018). The collected samples
were then chemically analyzed for their elemental and organic carbon (EC/OC), water-soluble
organic carbon (WSOC), metals and trace elements, and tracer of biomass burning (i.e.,
Levoglucosan) components. Moreover, the oxidative potential of these samples were quantified
by the means of DCFH in vitro assay. At next step, we conducted spearman rank-order
correlation analysis, principal component analysis (PCA), and multiple linear regression analysis
(MLR) to identify the major sources contributing to PM2.5 associated oxidative potential.
According to our results, remarkably higher mass-based (intrinsic), and volume-based (extrinsic)
oxidative potentials were observed in urban background of Athens, compared to many other
metropolitan areas of the world. In addition, the MLR analysis came up with vehicular activities
(characterized by EC) (44%), secondary organic aerosol (SOA) formation (identified on the basis
of WSOC) (16%), and biomass burning (characterized by levoglucosan) (9%) as main sources of
PM induced toxicity. Therefore, these findings underscored the significant impact of traffic and
6
SOA on the oxidative potential of ambient PM2.5 in greater Athens area and can further be used
as beneficial guideline by public health authorities to prioritize the required policies to mitigate
the adverse health effects of exposure to ambient PM2.5.
Eventually, while the second and third studies were designed to evaluate the health
impacts of exposure to ambient PM in some of the crowded and populated locations in Middle
East and Europe, further inhalation exposure studies are of great importance to more accurately
assess the detrimental health outcomes of ambient PM. Thus, the main aim of the fourth study
was to develop an innovative method for generating physically and chemically stable sources of
aerosols that are well representative of ambient particulate matters (PM), with the ultimate
objective of using them for inhalation exposure studies. The protocol included collection of
ambient PM samples on filters using a high-volume sampler, which were then extracted with
ultrapure Milli-Q water using vortexing and sonication. As an alternative approach for collection,
ambient particles were directly captured into aqueous slurry samples using the versatile aerosol
concentration enrichment system (VACES)/aerosol-into-liquid collector tandem technology. The
aqueous samples from both collection protocols were then re-aerosolized using a commercially
available nebulizer. The physical characteristics (i.e., particle size distribution) of the generated
aerosols were examined by the means of a scanning mobility particle sizer (SMPS) connected to
a condensation particle counter (CPC) at different compressed air pressures of the nebulizer, and
dilution air flow rates. In addition, the collected PM samples (both ambient and re-aerosolized)
were chemically analyzed for water-soluble organic carbon (WSOC), elemental and organic
carbon (EC/OC), inorganic ions, polycyclic aromatic hydrocarbons (PAHs), and metals and trace
elements. Using the aqueous filter extracts, we were able to effectively recover the water-soluble
components of ambient PM (e.g., water-soluble organic matter, and water-soluble inorganic
7
ions); however, this method was deficient in recovering some of the important insoluble
components such as EC, PAHs, and many of the redox-active trace elements and metals. In
contrast, using the VACES/aerosol-into-liquid collector tandem technology for collecting
ambient PM directly into water slurry, we were able to preserve the water-soluble and water-
insoluble components very effectively. These results illustrate the superiority of the
VACES/aerosol-into liquid collector tandem technology to be used in conjunction with the re-
aerosolization setup to create aerosols that fully represent ambient PM, making it an attractive
choice for application in inhalation exposure studies.

8
Chapter 2: Source apportionment of ambient PM2.5 in two locations in central Tehran
using the Positive Matrix Factorization (PMF) model

2.1. Introduction
Industrial progress and urbanization in metropolitan areas have caused air pollution to
become a serious problem in these areas. Previous studies have confirmed the association
between exposures to ambient fine particulate matter (PM2.5) and excessive rates of health
outcomes, including neurodegenerative effects as well as respiratory disease (Davis et al., 2013;
Gauderman et al., 2015; Mabahwi et al., 2014; Miller et al., 2007; Pope et al., 2004; Wai et al.,
2015; Zhou et al., 2014). Additionally, studies have indicated that there will be a significant
increase (6-13% per 10 µg/m
3
of PM2.5) in cardiopulmonary diseases due to long-term exposure
to PM2.5 (Beelen et al., 2008; Krewski et al., 2009; Pope III et al., 2002). Such adverse health
effects are considerable particularly among children, the elderly, and other susceptible groups in
the community. For instance, there is compelling evidence suggesting that long-term exposure of
children to PM2.5 will lead to a decline in the lung growth rate and performance (WHO, 2011).
On the other hand, increased levels of PM2.5 are also associated with significant economic and
gross domestic product (GDP) losses due to the extra medical expenses and decreased working
hours (G. Wang et al., 2016).
Based on several PM2.5 source apportionment studies, ambient PM2.5 mass concentrations can be
linked to different sources, including traffic-related emissions, secondary aerosols, residential
biomass burning, ship emissions, soil, road dust, industrial emissions, and sea salt, each of which
having different chemical fingerprints (Chen et al., 2018; Gugamsetty et al., 2012;
Hasheminassab et al., 2014a; Mantas et al., 2014; Port et al., 2017). It is also shown that different
9
PM components have different levels of toxicity and health impacts, suggesting that the total PM
mass concentration may not be the best metric in terms of the associated health impacts (Chen
and Lippmann, 2009; Khodeir et al., 2012; Lippmann, 2010; Mauderly and Chow, 2008;
Schlesinger, 2007). Therefore, it is quite essential to perform PM source apportionment studies
to identify PM sources with different chemical composition, levels of toxicity, and therefore
health impacts. Results from such studies would be quite beneficial for policy makers in order to
mitigate ambient particulate pollution and its associated health impacts.
The positive matrix factorization (PMF) is a well-known receptor model that has been
used widely for source apportionment of ambient PM mass concentrations (Crilley et al., 2017;
Gugamsetty et al., 2012; Hasheminassab et al., 2014a; Kuang et al., 2015; Liu et al., 2017;
Pandolfi et al., 2011; Sharma et al., 2016; Zauli Sajani et al., 2015). More recently, source
apportionment of ambient PM number concentrations have also been performed using PMF,
aiming to elucidate sources that contribute to particle number rather than PM mass (Beddows et
al., 2015; Harrison et al., 2011; Kasumba et al., 2009; Sowlat et al., 2016a; Z. B. Wang et al.,
2013). It is, however, noteworthy that most of aforementioned source apportionment studies
have been conducted in the developed countries, while such studies seem to be of paramount
importance in developing countries (such as countries in the Middle East) as well, due to their
rapid growth and associated air quality concerns; there have only been a limited number of
studies in Middle eastern countries (for example (Al-Dabbous and Kumar, 2015; Nayebare et al.,
2016; Sowlat et al., 2013; Tecer et al., 2012)), due to limited available funding for research, and
limited access to advanced instruments and the required analytical methods.
Rapid urbanization and industrialization during the past few decades have caused Tehran,
the capital of Iran with a population of around 9 million, to place among the top 500 cities of the
10
world in terms of PM2.5 annual mean concentration (Amini et al., 2014; Naddafi et al., 2012a,
2012b; WHO, 2016), causing respiratory diseases, angina pectoris, and cardiovascular problems
(Akbarzadeh et al., 2018; Hosseinpoor et al., 2005; Mohsenibandpi et al., 2017). The particular
topography of the city, with mountains to the north and east, limits the horizontal dispersion of
ambient air pollutants in these directions. In addition, the prevailing westerly and southerly
winds transport pollutants from industrial units to the city center (Atash, 2007).
Historically, high concentrations of ambient carbon monoxide (CO) were regarded as the
main cause of unhealthy days in Tehran before the implementation of new regulations in 2002,
banning the use of carbureted vehicles (Hosseini and Shahbazi, 2016). However, according to the
Tehran Air Quality Control Company’s (AQCC) report, PM2.5 is currently the most important
criteria air pollutant in Tehran. In fact, the annual mean concentration of CO is 2.7±0.4 ppb,
while the annual mean PM2.5 concentration is at 32±4 µg/m
3
(Habibi et al., 2017). Previous
studies in the area have indicated that Tehran’s PM2.5 mainly comprised organic matter (35%),
elemental carbon (9%), dust (25%), sulfate (11%), ammonium (5%), and nitrate (2%) (Arhami et
al., 2017). Additionally, PM2.5 clusters have divided the city into northern (with lower PM
pollution) and southern (more polluted) zones. However, CO hotspots have covered an area from
northeast to southwest of Tehran and its cold spots have been spread over the other part of city
(Habibi et al., 2017). PM10 is another criteria pollutant in Tehran with very high annual average
concentrations (as high as 100.8 µg/m3). Previous studies in the area have reported annual mean
concentrations as high as 88, 53, and 38 ppb for NO, NO2, and SO2, respectively, with higher
levels in the winter and lower levels in the summer (Amini et al., 2016, 2014). The annual mean
concentrations of benzene (C6H6), alkylbenzenes (BTEX), and toluene (C7H8) have also been
reported to be 8.4±3.3, 62, and 25±11 µg/m
3
, respectively, which are far higher than the EU air
11
quality standards (5 µg/m
3
) and WHO guidelines (1.7 µg/m
3
) (Amini et al., 2017; Gemmer and
Bo, 2013).
The investigation of the temporal/seasonal variation of air pollutants in Tehran has
revealed that higher level of pollution is observed during September-December, due to stable
meteorological conditions caused by inversions during the colder months of the year (AQCC,
2013; Bahari et al., 2014). Previously conducted studies in Tehran have suggested that traffic and
industrial emissions are the two major sources in the area (Kamali et al., 2015; Naddafi et al.,
2012b; Sotoudeheian and Arhami, 2014). The vehicles fleet of Tehran comprises 4 million
mobile sources, including light-duty passenger cars (72%), motorcycles (18%), taxis (2%), buses
(<0.5%), minibuses (0.7%), pickups (5%) and trucks (2.5%), with light-duty passenger cars and
motorcycles contributing to 90% of the total fleet (Hosseini and Shahbazi, 2016; Shahbazi et al.,
2015). These mobile sources play a significant role in Tehran air quality, with major
contributions to PM, CO, VOC, and NOX levels. Shahbazi et al. (2016) have reported that mobile
sources are responsible for nearly 70% of total PM pollution, while the power sector (20%) and
industry (8%) are the next important sources. Nearly all of CO (about 98%) is also emitted by
vehicular emissions. Mobile sources also have major contributions to volatile organic
compounds (VOCS) emissions (88%), while the rest of VOCs emissions mostly comes from
fugitive losses at the gas stations (about 10%). NOx emission are dominated by vehicular
activities, power generation, and household and commercial sector accounting for 47%, 22%,
and 20% of total NOx emissions, respectively. Finally, SO2 is the only pollutant dominated by
emissions from power plants and oil refineries (68%) as well as industrial activities (21%)
(Shahbazi et al., 2016). To the best of our knowledge, no PMF source apportionment studies
have so far been performed to identify sources and quantify their contributions to ambient PM2.5
12
in central Tehran. The only relevant study is that of Arhami et. al. (2017), in which the authors
have used principal component analysis (PCA) to perform chemical speciation and source
identification for ambient PM2.5, but quantitative estimations of source contributions have not
been obtained in their work due to the inherent limitations of the PCA method (Arhami et al.,
2017).
In the present study, we performed source apportionment of ambient PM2.5 in two
locations in central Tehran from May 2012 through June 2013 using chemically-speciated data
and the PMF model. PM2.5 mass concentrations, water-soluble ions, and metals and trace
elements which were measured as part of the Health Effects of Air Pollution Panel Study
(HEAPPS) (Hassanvand et al., 2015, 2014), were included as inputs to the PMF to identify major
PM2.5 sources and quantify their contributions to ambient PM2.5 mass concentrations.
Additionally, elemental and organic carbon (EC and OC) measurements as well as
meteorological data were used for further characterization and identification of the resolved
factors by the PMF model. The results of this study can be used as a beneficial tool for
researchers to estimate source-specific health outcomes and for policy-makers to mitigate the
most detrimental sources with highest contributions.
2.2. Methodology
2.2.1. Sampling site
In this study, we collected samples in two distinct sites in central Tehran: a school
dormitory (35° 42′ 40.33″ N, 51° 22′ 49.75″ E) and a retirement home (35° 42′ 20.30″ N, 51° 22′
14.41″ E). The school dormitory was located in close proximity (200 m away) to a major
freeway (i.e., Chamran freeway) and 40 m away from a major street (Dr. Fatemi Street), while
the other sampling site was located in a residential area, about 650 m away from Chamran
13
freeway. There was also a 1.1 km distance between the two sampling sites (Hassanvand et al.,
2015, 2014). Figure 2.1 shows the location of sampling stations in central Tehran.

Fig. 2.1. The location of sampling sites in central Tehran.

14
2.2.2. Sampling period and instrumentation
Twenty-four-hour PM2.5 samples were collected on PTFE (47 μm diameter, SKC Inc.)
and quartz filters (47 μm diameter, Whatman Inc.) using low-volume air samplers (FRM
OMNU
TM
air Sampler, multi-cut inlet; BGI, USA) from May 2012 through June 2013 in both
sites. Meteorological data, including monthly-averaged wind speed and temperature, were
obtained for a local weather station (i.e., Geophysics station) from Islamic republic of Iran
Meteorological Organization’s (IRIMO) website. Additionally, an ion chromatography (IC)
instrument was used to determine the levels of water-soluble ions (i.e., NH4
+
, K
+
, Mg
2+
, Ca
+2
,
NO3
-
, and SO4
2-
) in the samples. An Inductively Coupled Plasma Optical Emission Spectrometer
(ICP-OES) instrument was also used for determining the concentrations of metals and trace
elements (i.e., Al, Ti, Ba, Se, Li, Sr, Fe, Ni, Cd, Zn, Sn, Cu, Mn, Cr, Si) in the PM2.5 samples. To
determine the composition of metal (loid)s in PM2.5, one quarter of the filters was acid-digested
at 170 °C for 4 h in a closed Teflon vessel with a mixture of acids (i.e., 4 mL HNO3, 0.2 mL HF,
and 2 mL HClO4). The solution was dried on a hotplate and then made up to 10 mL with 2%
HNO3 prior to analysis by ICP-OES (Hassanvand et al., 2015). More detailed information
regarding the sampling procedure and chemical analysis can be found in (Hassanvand et al.,
2015, 2014). Thirty-nine PM2.5 samples were chemically analyzed for each site and were used in
our PMF analysis. Furthermore, quartz filters were also analyzed for their EC and OC content
with a EC and OC analyzer using thermo-optical transmittance (TOT) according to the NIOSH
5040 protocol. However, since the number of data points for EC and OC were smaller than the
available data points for ions, metals, and elements, the EC and OC data were not used as inputs
to the model; rather, these data were used in the correlation analysis between factor contributions
and EC and OC concentrations to corroborate the factors resolved by the PMF model.
15
2.2.3. PMF model
PMF is a receptor model which can be used for identification of sources and
quantifications of their contributions to the target variable (here, ambient PM2.5) (Paatero and
Tapper, 1994; P Paatero et al., 2014; Paatero, 1997). This multivariate model is used for solving
the chemical mass balance (CMB) equation:

Where, refers to the concentration for the sample and the specie; P represents
number of factors; is the relative contribution of factor to sample; refers to the
profile factor of each source for the specie; and is the PMF residual error for the
sample and the specie.

Then PMF intends to find out the most appropriate factor profile and contribution by
minimizing the objective function, :

Where, refers to the number of samples; is the number of species; and is the
uncertainty of the measured concentration for the sample and the specie. Such
minimization is performed with the constraint of non-negative values for source profiles and
contributions (Norris et al., 2014; Reff et al., 2007).
The PMF model is able to employ user-provided uncertainty for each of the measured
samples, which provides the model with the level of confidence in the environmental
16
measurements (P. Paatero et al., 2014). In this study, since laboratory uncertainties were not
reported for the chemical components, the uncertainty was estimated using the method proposed
by Ito et al. (Ito et al., 2004). This method requires the species limit of detection (LOD) as input
for uncertainty estimation according to the following equation:

Where, is the estimated uncertainty for the sample and the specie; refers to
the concentration of sample and specie ; and is the limit of detection associated with the
sample and the specie.

We employed the USEPA’s PMF version 5.0 for our PM2.5 source apportionment. We
ran the PMF in the robust mode and PM2.5 concentration was selected as total variable. Extra
modeling uncertainty was another important parameter which accounts for unconsidered errors
and uncertainties. A 10% extra modeling uncertainty was chosen for this study to obtain the best
solution in terms of physical interpretations and statistical considerations. Interpolation
techniques were also used for substituting missing values or outliers in the concentration matrix.
Then, these interpolated values were set as “weak” by the model (i.e., their uncertainty was
tripled) so that their effect on final results would be lowered (Norris et al., 2014; Reff et al.,
2007). We also used the signal-to-noise ratio as a determinant of whether the source of
variability in the measurements is the actual data or the data noise. Species with S/N ratios
greater than 1 are considered as strong species with reliable measurement signals (Norris et al.,
2014)
17
Bootstrap (BS), displacement (DISP), and BS-DISP (bootstrap + displacement) analysis
were also used to estimate the uncertainties associated with the model outputs. For the BS
analysis, 100 runs were conducted and the output was acceptable in case mapping was available
for more than 80% of the factors. For the DISP method, the decrease associated with the Q value
was less than 1% and no factor swap was observed for the . Thus, the solution
seemed to be valid in terms of rotational ambiguity. Finally, for the BS-DISP analysis, results
were considered as reliable if there was less than 0.5% decrease in the Q value (Norris et al.,
2014)
2.3. Results and discussion
2.3.1. Overview of data
Summary statistics of the measured species are presented in Tables 1 and 2 for Tohid
retirement home and school dormitory, respectively. These statistics include mean, minimum
(min), maximum (max), standard deviation (SD), and signal-to-noise (S/N) ratio. All of the
species had signal-to-noise ratios well above unity, indicating strong signals for all of the
species. In addition, the mean PM2.5 levels were 30.9 and 33.2 µg/m
3
in Tohid retirement home
and school dormitory, respectively, which are consistent with the 37 µg/m
3
average value
reported by the World Health Organization (WHO) (WHO, 2016). Figure 2.2 shows the
correlations between the predicted and measured PM2.5 mass concentrations for both sampling
sites. There is an obvious strong association between the predicted and measured PM2.5 mass
concentrations in the school dormitory (slope=1.17, R²=0.95) and Tohid retirement home
(slope=0.87, R²=0.95), further corroborating the PMF model’s ability to accurately estimate the
PM2.5 concentrations and source contributions.

18
Fig. 2.2. Correlations between the predicted and measured PM 2.5 concentration for: a) Tohid retirement
home; and b) school dormitory.

(a)

(b)

Table 2.1. Summary statistics and mass fractions of the measured chemical components in Tohid
retirement home
Species Mean Max Min SD S/N
PM 2.5 Mass (µg/m
3
) 30.90 87.14 18.19 13.57 9.00
NH 4 (ng/µg PM 2.5) 38.52 61.64 8.94 15.28 10.00
K (ng/µg PM 2.5) 3.66 9.43 0.81 1.93 10.00
Mg (ng/µg PM 2.5) 3.40 6.39 0.77 1.51 9.57
Ca (ng/µg PM 2.5) 33.98 63.93 7.70 15.13 10.00
NO 3 (ng/µg PM 2.5) 129.64 230.11 30.81 54.51 10.00
SO 4 (ng/µg PM 2.5) 112.00 195.59 26.19 48.02 10.00
Al (ng/µg PM 2.5) 2.98 15.44 0.30 2.80 9.86
Ti (ng/µg PM 2.5) 0.79 2.21 0.14 0.51 9.39
Ba (ng/µg PM 2.5) 0.44 0.85 0.14 0.23 7.81
Li (ng/µg PM 2.5) 0.03 0.11 0.01 0.02 4.90
Sr (ng/µg PM 2.5) 1.64 3.96 0.18 0.92 10.00
Fe (ng/µg PM 2.5) 8.77 30.95 1.60 6.01 9.91
Ni (ng/µg PM 2.5) 0.18 0.37 0.05 0.07 4.77
Cd (ng/µg PM 2.5) 0.01 0.03 0.00 0.01 2.30
Zn (ng/µg PM 2.5) 4.86 14.99 0.84 3.28 9.52
Sn (ng/µg PM 2.5) 0.06 0.14 0.01 0.03 7.76
Cu (ng/µg PM 2.5) 0.95 2.08 0.24 0.38 9.97
Mn (ng/µg PM 2.5) 1.83 5.36 0.35 1.06 10.00
Cr (ng/µg PM 2.5) 0.15 0.41 0.03 0.08 10.00
Si (ng/µg PM 2.5) 9.82 18.92 0.55 4.73 10.00
Se (ng/µg PM 2.5) 0.01 0.02 0.00 0.00 3.07

19
Table 2.2. Summary statistics and mass fractions of the measured chemical components in the school
dormitory
Species Mean Max Min SD S/N
PM 2.5 Mass (µg/m
3
) 33.21 118.00 10.01 20.07 9.00
NH 4 (ng/µg PM 2.5) 39.86 111.60 10.55 20.87 10.00
K (ng/µg PM 2.5) 2.91 7.48 0.36 1.95 9.98
Mg (ng/µg PM 2.5) 12.22 33.82 3.05 7.46 10.00
Ca (ng/µg PM 2.5) 26.28 70.16 6.71 13.52 10.00
NO 3 (ng/µg PM 2.5) 132.42 363.57 40.65 68.49 10.00
SO 4 (ng/µg PM 2.5) 119.22 302.34 27.46 65.61 10.00
Al (ng/µg PM 2.5) 3.50 14.16 0.05 2.55 9.81
Ti (ng/µg PM 2.5) 1.00 4.62 0.09 0.88 9.65
Ba (ng/µg PM 2.5) 0.64 2.04 0.18 0.39 9.12
Se (ng/µg PM 2.5) 0.01 0.03 0.00 0.01 4.77
Li (ng/µg PM 2.5) 0.04 0.16 0.01 0.03 6.36
Sr (ng/µg PM 2.5) 1.00 3.99 0.02 0.74 9.84
Fe (ng/µg PM 2.5) 10.02 35.94 0.75 6.95 9.92
Ni (ng/µg PM 2.5) 0.23 0.61 0.05 0.14 5.47
Cd (ng/µg PM 2.5) 0.01 0.06 0.00 0.01 2.93
Zn (ng/µg PM 2.5) 16.72 64.00 0.98 13.85 9.82
Sn (ng/µg PM 2.5) 0.05 0.15 0.00 0.03 6.29
Cu (ng/µg PM 2.5) 1.17 3.93 0.22 0.81 9.60
Mn (ng/µg PM 2.5) 1.59 5.31 0.28 1.03 10.00
Cr (ng/µg PM 2.5) 0.25 1.50 0.04 0.26 10.00
Si (ng/µg PM 2.5) 11.17 45.07 0.11 8.88 9.80

2.3.2. Number of factors
Different number of factors and various extra modeling uncertainties were used to
identify the most plausible, physically and statistically, solutions. The final solutions were
chosen based on several criteria, including (i) evaluation of the resulting source profiles; (ii)
temporal trends for each factor in the cold (fall and winter) and warm phases (spring and
summer); (iii) analysis of error estimation methods (i.e., BS, DISP, and BS-DISP runs); (iv) R
2

values of the predicted vs. measured PM2.5 concentrations; (v) acceptable ranges for Q/Qexp; and
(vi) correlation between EC and OC concentrations and source contributions of the resolved
factors. It should be noted that PMF runs were performed under two scenarios; in the first
20
scenario, the data matrices from the two sites were combined and the PMF model was run on this
unit matrix, while in the second scenario, the PMF was individually run on the data matrices of
each site. However, since the results of combined dataset scenario were not satisfactory in terms
of model statistics and physical interpretation of factor profiles and contributions, we opted to
run the PMF model using a separate matrix for each sampling location.
The seven PMF-resolved factor profiles and temporal trends of each factor are shown in
Figures 2.3 and 2.4 for Tohid retirement home and school dormitory, respectively. Figure 2.5
also illustrates the relative contribution of each of the factors resolved by the PMF model to
PM2.5 mass concentrations for overall data, as well as for cold and warm phases in both sampling
locations. Figure 2.6 indicates the temporal trends of factor contributions to PM2.5 mass
concentrations in cold and warm phases for Tohid retirement home and the school dormitory.
The temporal variations of meteorological parameters, including monthly-averaged wind speed
and temperature, are also presented in Figure 2.7. Finally, results of the correlation analysis
between EC and OC concentrations and the contributions of the vehicular emissions factor are
shown in Figure 2.8 for both sampling sites, whereas Figure 2.9 depicts the correlation analysis
between OC and contributions of the secondary aerosol factor for Tohid retirement home and
school dormitory.

21
Fig. 2.3. PMF-resolved factor profiles and their temporal trends in Tohid retirement home.

22

23

Fig. 2.4. PMF-resolved factor profiles and their temporal trends in school dormitory.

24

25

26
Fig. 2.5. Relative concentrations of sources to PM 2.5 mass concentrations in Tohid retirement home: a)
overall; c) warm phase; and e) cold phases and school dormitory: b) overall; d) warm phase; and f) cold
phases.
Tohid retirment home

(a)overall
School dormitory

(b)overall

(c)warm phase

(d)warm phase

(e)cold phase (f)cold phase
27
Fig.2.6. Temporal trends of factor contributions to PM 2.5 mass concentrations in the cold and warm
phases for: a) Tohid retirement home; and b) school dormitory (The box plots show the inter quartile
range along with upper and lower quartile; dots show the minimum and maximum values; and median of
data are also presented as a horizontal line in the box).

(a) (b)

Fig. 2.7. Metrological parameters including: a) monthly average temperature; and b) monthly average
wind speed in Tehran during the study period.

(a)

28

(b)
Fig. 2.8. Correlation analysis between: a) contributions of vehicular emissions and EC in retirement
home; b) contributions of vehicular emissions and OC in retirement home; c) contributions of vehicular
emissions and EC in dormitory; d) contributions of vehicular emissions and OC in dormitory.

(a)

(b)

(c)

(d)
29
Fig. 2.9. Correlation analysis between: a) contributions of secondary aerosols and OC in Tohid retirement
home; and b) contributions of secondary aerosols and OC in school dormitory.

(a)

(b)

2.3.3. Factor identification
2.3.3.1. Vehicular emissions
This factor is mainly characterized by high loadings of Ti, Ba, Fe, Cu, Zn, and Mn in
Tohid retirement home. Additionally, the dominant elements for this factor are Ba, Cd, Zn, Cu,
Ni and Mn in the school dormitory site. This factor also has a significant contribution to total
PM2.5 concentrations, accounting for 48.8% and 49.3% of total PM2.5 in Tohid retirement home
and school dormitory, respectively (Figure 2.5). Previous studies have indicated that Fe, Cu, Zn,
Ba, Cd, Ni, and Mn are elemental tracers and reliable markers of vehicular emissions (Lim et al.,
2010). Investigation of the temporal trends for this factor also revealed higher contributions in
the cold season than in the warm season for both sampling locations, which is most probably due
to the stable metrological conditions as a result of lower mixing heights, lower wind speeds, and
the resulting inversions that prevail in colder months of the year (Sowlat et al., 2016a) (Figure
2.6).
30
We also found relatively high correlations between EC and OC concentrations and
contributions of this factor to total PM2.5 concentrations at both sites (Figure 2.8). This
corroborates the vehicular origin of this factor, as EC and OC together are well known tracers of
exhaust emissions (Zhang et al., 2013; Zong et al., 2016). We should point out that there were a
few slightly negative values in the source contribution data, which were set to zero since
negative source contributions do not have physical meaning. All of the above discussion points
to "vehicular emissions" as the logical title for this factor, in-line with the results of the previous
emission inventory study performed in Tehran, indicating vehicular (exhaust and non-exhaust)
emissions as the most significant contributor to ambient PM (Shahbazi et al., 2016). The
observed difference in the estimated contribution of vehicular emissions to ambient PM between
this study (roughly 50%) and Tehran emission inventory study (70%) can also be attributed to
the different PM size ranges studied, the different approaches used to estimate these
contributions, and the fact that the emission inventory methodology tends to overestimate
primary emissions as it does not take into account the impact of secondarily formed PM and non-
anthropogenic PM (soil, sea salt, etc.).

2.3.3.2. Secondary Aerosol
This factor is represented by high loadings of NO3
-
, NH4
+
, SO4
2-
, and has a major
contribution to total PM2.5 concentrations in Tohid retirement home (28%) and school dormitory
(24%) (Figures 2.3, 2.4, and 2.5). NO3
-
, SO4
2-
, and NH4
+
are the components of ammonium
sulfate and nitrate formed by gas phase reactions of acidic gaseous precursors (i.e., HNO3 and
H2SO4) with ammonia (NH3). Analysis of the seasonal contributions also indicates some levels
of contribution in the warm period, which can be related to higher photochemical activities,
31
leading to the formation of secondary organic aerosol (SOA) and secondary ammonium sulfate.
Conversely, contributions in cold seasons can be attributed to higher dissociation constant of
ammonium nitrate leading to secondary ammonium nitrate formation, and also formation of
nighttime SOA (Hasheminassab et al., 2014a, 2014b; Heo et al., 2009; Mozurkevich, 1993;
Saffari et al., 2016). There are also moderate–to-strong correlations (with R
2
values ranging from
0.3-0.7) between OC concentrations and the contributions of this factor to total PM2.5
concentrations for school dormitory and Tohid retirement home (Figure 2.9). We, therefore,
elected "secondary aerosol" as the most suitable title for this factor, which encompasses
secondary ammonium nitrate and sulfate, and SOA.

2.3.3.3. Industrial Emissions
This factor indicated high loadings of Cu, Cd, Sn, and Cr in school dormitory and
contributed 1.8% to total PM2.5 concentrations (Figure 2.5). Previous studies have indicated that
electrical wire industry greatly relies on Cu as an important element. In addition, Cu is also used
in machinery industries and electro motors (Günter, 1999). Cd is used in battery industry (e.g.,
rechargeable nickel cadmium batteries), and for electroplating purposes (e.g., reducing corrosion
of steels) (Krishnamurthy et al., 2014; Scoullos et al., 2001). Sn is also used in solder and tin
plating (Black, 2005). Cr is another element used in metallurgy, refractory, and foundry industry
(e.g., producing metal casting). Cr plating, for example on cars and bicycles, produces a smooth,
silver finish that is highly resistant to corrosion (Edwards, 1997; Morrison and Murphy, 2010;
Papp and Lipin, 2006; Zhao et al., 2001). Therefore, as suggested by previous studies, these
metals can be used as tracers of industrial emissions.
32
High loadings of Se, Ni, Cu, and Cr are also associated with a PMF-resolved factor
profile in Tohid retirement home. In addition, Se is also a heavy metal which is used in the
electronics, plastic, glass, and paints industry (Risher et al., 2003). Loadings of Ni can also be
attributed to local Ni-industry emissions such as cement, glass, stainless steel, and brick
production (Hasheminassab et al., 2014b; Tian et al., 2012). Another factor was also resolved in
the retirement home with high levels of Sr, Ni, Cd, Sn, Cu (Figure 2.3), which probably has an
industrial origin according to aforementioned studies. The temporal trend of this factor indicated
higher contributions of this source factor to ambient PM2.5 in the cold phase at both sites, which
can be explained by stable metrological conditions in this season (Figure 2.6), as discussed
earlier. These results are in agreement with previous studies, reporting that there are
approximately 3100 industrial units in Tehran, which can lead to high levels of various tracers of
industrial emissions (Shahbazi et al., 2016). Both industry factors (i.e., industrial emissions 1 and
2) in the retirement home are combined and shown as one factor with the title of "industrial
emissions" in Figures 2.5 and 2.6. Besides, the aggregated contribution from these two factors is
more than 17%, which is well above the contribution of industrial emissions in the school
dormitory. It is noteworthy that most of the industrial units are scattered in the western and
southwestern parts of Tehran (Shahbazi et al., 2016). Therefore, the higher contribution of
industrial emissions in Tohid retirement home could be attributed to the closer proximity of this
site to industrial units scattered in the that area. For example, there are numerous auto repair
shops located in close proximity to this sampling site, which could be another reason for
observing the higher contribution of this source factor in this site. Additionally, the average
contribution of industry factor of both sampling sites is around 9%, which is consistent with the
7% contribution of industrial emissions to ambient PM reported by previous studies. We should
33
point out that the industrial sector is comprised of different subsectors including combustion of
oil and other fossil fuels (Shahbazi et al., 2016).

2.3.3.4. Biomass burning
This factor is mainly characterized by significant levels of K
+
and accounts for 3% of
total PM2.5 concentrations in Tohid retirement home and 16% of total PM2.5 concentrations in the
school dormitory (Figures 2.3, 2.4, and 2.5). Previous studies have stated that high loadings of
K
+
is a clear indicator of biomass burning in the area (Lee et al., 2008; Santoso et al., 2008).
Open burning of agricultural straw in the suburban areas, as well as restaurants and Kebab shops
(that use charcoal) in the city are the most important sources of biomass burning in Tehran. In
addition, analysis of temporal patterns for this factor indicates higher contributions in the cold
season than in the warm season (Figure 2.6), which is likely due to more stable metrological
conditions prevailing in the cold period. We also observed relatively high loadings of water-
soluble ions (i.e., SO4
2-
, NO3
-
, NH4
+
, Ca
2+
, and Mg
2+
) in this factor at the school dormitory. This
does not necessarily mean that these ions are in fact part of the factor, as it is known that biomass
burning is not usually associated with high loadings of such ions, except for small loadings of
SO4
2-
and NO3
-
(Lee et al., 2008; Zhang et al., 2013); rather, this could be attributed to the partial
portioning of these ions in this factor as a result of the limited number of samples in the study,
causing the PMF model to fail to cleanly partition these ions into the secondary aerosol factor, as
would have been expected.

34
2.3.3.5. Soil
This factor contributed 2.8% to total PM2.5 concentrations in Tohid retirement home and
8% to total PM2.5 concentrations in the school dormitory. In addition, relatively high loadings of
Al, Ti, Li, and Fe are associated with this factor in both sampling sites. Based on the previous
studies, crustal elements, such as Al, Ti, Fe, Ca, and Si, are well known tracers of soil
(Hasheminassab et al., 2014b; Heo et al., 2009; Sowlat et al., 2012; Zong et al., 2016). It is
noteworthy that elevated loadings of Cu, Zn, Mn, and Cr in the Tohid retirement home indicated
that this factor is also affected by brake and tire wear, road surface abrasion, and the
resuspension of road surface dust (Dall’Osto et al., 2008; Pant and Harrison, 2013). This is also
the case for the other sampling site in which elevated levels of Mn and Cr are observed due
mainly to the aforementioned activities. Additionally, very high loadings of Ca and Si, as crustal
elements that are soil tracers, were observed in the school dormitory, which corroborates the soil-
related origin of this factor. The temporal trends of this factor also exhibited higher contributions
to the total PM2.5 concentrations in the warm season as compared to the cold seasons (Figure
2.6), which might be due to more unstable atmospheres as a result of higher wind speed and
lower relative humidity. In addition, previous studies performed in the area have indicated that
dust storms/events that originate from deserts in Iraq, Kuwait, and Saudi Arabia (Givehchi et al.,
2013; Shahsavani et al., 2012) are frequently observed during the warm season (spring and
summer) and can drastically elevate the concentrations of PM2.5 and its crustal elements (as soil
tracers). It should be noted that there is an outlier in our data, observed in the warm season, as
shown in the box plots of the temporal variations for both of these factors (Figure 2.6). This data
point was measured on May 25
th
, a day on which Tehran was experiencing a severe dust storm
35
event that had greatly elevated PM2.5 concentrations (Hassanvand et al., 2015, 2014). The
partitioning of this data point in this factor further corroborates the soil origin of this factor.
2.3.3.6. Road dust
High levels of Al, Ti, Ba, Fe, Zn, Mn, and Si are associated with this factor in Tohid
retirement home, contributing to a very small portion (less than 1%) of total PM2.5 concentrations
(Figures 2.3 and 2.5). According to the previous studies, Al and Si are important chemical tracers
of particle resuspension (road dust and non-tail pipe emission). In addition, significant loads of
Zn are associated with tire dust, while Ba is linked with brake wear (Harrison et al., 2012).
Finally, Ti is also an important tracer of road dust (Harrison et al., 2012; Kotchenruther, 2016;
Vossler et al., 2016). Temporal variation of this factor also revealed higher contributions in the
cold phase as compared to the warm phase, due mainly to stable meteorological conditions (i.e.,
low wind speed and inversion) (Figure 2.6). All of this evidence points to road dust as the most
likely origin for this factor.
Similarly, there was a factor resolved by the model for the school dormitory,
(contributing less than 1% to PM2.5 concentrations) with relatively high loadings of Al, Ti, Ba,
Li, Fe, Mn, and Si (Figure 2.4). Previous studies have indicated that Ba is a tracer for brake wear.
Additionally, Al, Ti, Li, Fe, and Si are also well-known tracers of resuspended soil dust (Cohen
et al., 2010; Harrison et al., 2012; Jeong et al., 2016). As can be seen in Figure 2.4, this factor
contributed to ambient PM2.5 concentrations to a higher degree in the warm season, which is the
typical seasonal trend of a soil factor. This similarity could be explained by the dominance of
soil and crustal elements in this factor. All of this discussion suggests that "brake wear particles"
may be the most plausible origin for this factor due to significant levels of Ba. The negligible
36
contribution of this factor to PM2.5 is explained by the fact that road dust (and soil) are expected
to partition mostly in the coarse PM size fraction.
Finally, another source profile with a rather negligible contribution (less than 1%) to total PM2.5
was observed in the school dormitory. This factor was called tire dust because of its very high
loading of Zn (a well-known tracer of tire wear particles(Harrison et al., 2012)), in addition to
minor loadings of Cu, Mn, Cr, Fe, Si, Al, and Cd (tracers of road dust). This factor contributed to
PM2.5 mass concentrations to a higher degree in the cold phase, possibly as a result of stable
atmospheres in this season (Figure 2.4). Since tire dust and brake wear particles are subsets of
road dust, these two factors were combined in Figures 2.5 and 2.6 and their contributions were
reported as that of a single road dust factor.

2.4. Summary and conclusions
In this study, we used the PMF model for source apportionment of ambient PM2.5 in
central Tehran. Vehicular emissions, secondary aerosol, industrial emissions, biomass burning,
soil, and road dust were the source factors resolved in both sampling sites. The results indicated
that vehicular emissions were the most significant contributor to PM2.5, with a slightly higher
contribution in the school dormitory (49.3%) compared to Tohid retirement home (48.8%).
Secondary aerosols were the next important contributor to ambient PM2.5 levels in both locations,
with a slightly higher contribution in Tohid retirement home (28.1%) as compared to the school
dormitory (24%). All other factors exhibited relatively minor contributions to PM2.5
concentrations in both sites, together making up to around 20% of PM2.5 mass concentrations.
Results from the present study highlight the significant impacts of traffic-related emissions (both
tailpipe and non-tailpipe) on ambient PM2.5 levels, especially since the precursors of secondary
organic and inorganic aerosols also come from vehicular emissions. This has important
37
implications for policy makers in promulgating legislation to control vehicular emissions in the
city of Tehran in an effort to significantly improve air quality in terms of PM2.5 levels.

38
Chapter 3: Source-specific lung cancer risk assessment of ambient PM2.5-bound Polycyclic
Aromatic Hydrocarbons (PAHs) in central Tehran

3.1. Introduction
Increased number of motor vehicles, industrial units, and rate of urbanization have led to
significant deterioration of ambient air quality, particularly in highly populated urban areas of the
world. Polycyclic aromatic hydrocarbons (PAHs) are organic materials that are composed of
multiple aromatic rings. Previous studies have indicated that these organic compounds mainly
come from incomplete combustion of fossil fuel in industries, fossil fuel combustion at high
temperature (e.g., in vehicle’s engines), and wood and biomass burning (Alves et al., 2015;
Galarneau, 2008; Guo et al., 2003; Harvey, 1997; Lima et al., 2005; Miller et al., 2010).
Additionally, while low molecular weight PAHs mainly exist in the vapor phase, higher
molecular weight PAHs (i.e., PAHs having four or more rings) are mostly found in particle phase
due to their low vapor pressure (Baek et al., 1991; Park et al., 2002). PAHs are ubiquitous
species and can contaminate not only the areas adjacent to their sources, but can also transport
long distances and contaminate receptor sites located far away from their sources (Kaya et al.,
2012; Motelay-massei et al., 2003; Wang et al., 2014).
Due to their mutagenic and carcinogenic characteristics, several PAHs are considered as
priority pollutants by the United States Environmental Protection Agency (USEPA) and the
International Agency for Research on Cancer (IARC) (Delgado-saborit et al., 2011; Luch, 2005;
US-EPA, 2003; Vendrame et al., 2001; Wang et al., 2010). Additionally, among different PAHs,
benzo(α)pyrene (BaP) is used as an index for risk assessment of exposure to PAHs due to its
well-established carcinogenic effects (Callén et al., 2014a). One of the most important adverse
39
health effects of exposure to ambient PAHs is lung cancer; several studies have evaluated and
confirmed this association in different areas of the world (Armstrong et al., 2004; Miller et al.,
2013; Rengarajan et al., 2015; Shen et al., 2014). Lung cancer is considered as one of the most
prominent causes of mortality due to cancer worldwide, leading to more than 1 million deaths
annually (Almasi et al., 2015; Ferlay et al., 2015; Zhang et al., 2014). Although mortality
attributable to lung cancer is mostly dominant in developing countries, such as Iran,
epidemiological data on morbidity and mortality rates due to lung cancer are quite scarce in these
areas (Didkowska et al., 2016; Hassanipour et al., 2017). In Iran, lung cancer incidence has been
steadily increasing during the past few decades, leading to 4361 reported deaths in 2012
(Hassanipour et al., 2017; Hosseini et al., 2009; Khorasani et al., 2015). In Tehran, in particular,
the annual percentage increase in the age-standardized lung cancer incidence was around 14%
from 2003 to 2008 for both men and women (Almasi et al., 2015). Moreover, the highest average
standardized rate (ASR) of lung cancer occurrence was observed among females in Tehran, as
compared to other cities in Iran, due mainly to Tehran’s highly polluted ambient air combined
with physical inactivity and smoking (Hassanipour et al., 2017; Keramatinia et al., 2016;
Mohagheghi et al., 2009).
Considering these important health impacts, evaluation of the levels, identification and
apportionment of the specific pollution sources, and lung cancer risk assessment of exposure to
PAHs in the atmosphere are essential guiding tools to policy makers in adopting effective
policies for mitigating the levels and, in turn, the adverse health effects of exposure to PAHs.
The positive matrix factorization (PMF) model is one of the most commonly accepted
multivariate receptor models used for source apportionment studies all around the world. Several
studies have so far used this model for source apportionment of PAHs mass concentrations in
40
different compartments of environment, including ambient air, soil, and aqueous sediments
(Aydin et al., 2014; Callén et al., 2014a, 2014b; Jang et al., 2013; Tian et al., 2013; Zhang et al.,
2012; Zuo et al., 2007). Results from the ambient air studies indicate that vehicular emissions
(including gasoline and diesel exhaust emissions), unburned petroleum and crude oil residues,
industrial emissions (e.g., iron and steel production), coal combustion, biomass burning, and
natural gas combustion are the most important sources that release PAHs into the atmosphere
(Aydin et al., 2014; Callén et al., 2014a; Jang et al., 2013).
Tehran, the capital of Iran, has been facing serious air pollution challenges over the past few
decades, mainly due to the substantially increased number of motor vehicles, significant
industrial development, and remarkably increased urbanization rate (Halek et al., 2007; Hosseini
and Shahbazi, 2016; Naddafi et al., 2012b). Asthma, pulmonary, respiratory, and cardiovascular
diseases are among the most significant health consequences of deteriorated ambient air quality
in this mega city that has more than 9 million residents (Akbarzadeh et al., 2018; Faridi et al.,
2018; Hosseinpoor et al., 2005; Naddafi et al., 2012a; Shahi et al., 2014). In addition, the city is
surrounded by Alborz Mountains to the north and east, which prevent the horizontal dispersion
of pollutants in these directions. The predominant westerly and southerly wind directions also
exacerbate the situation by transporting industrial pollutants to the central and highly crowded
parts of the city (Atash, 2007). In Tehran, the lowest air quality index (AQI) values are usually
observed during the cold season of the year, due to the stable meteorological condition caused by
lower wind speeds and the occurrence of the temperature inversion phenomenon (AQCC, 2013;
Bahari et al., 2014).
Previous studies have indicated that vehicular and industrial emissions are the major
sources of air pollution in Tehran. Traffic emissions are major contributors to CO (98%), volatile
41
organic compounds (VOCs) (88%), and NOx (47%), while energy conversion activities account
for 22% of NOx; and most of SO2 emissions are due to power plants (68%), and industrial
activities (21%) (Askariyeh and Arhami, 2013; Kamali et al., 2015; Naddafi et al., 2012b;
Shahbazi et al., 2016). Moreover, mobile sources and secondary aerosols are responsible for
more than 70% of PM2.5 (Arhami et al., 2018; Taghvaee et al., 2018), which is known as the
criteria air pollutant according to the Tehran Air Quality Control Company (AQCC) report
(AQCC, 2013). There have been only a limited number of studies aiming to apportion the
sources of PAHs in Tehran, in which the authors have used diagnostic ratios and principal
component analysis (PCA) approaches (Halek et al., 2010; Hoseini et al., 2016). Nevertheless,
none of these approaches are able to provide quantitative estimates of source factor contributions
to the concentrations of the target pollutant (PAHs in this case). The main limitation of
diagnostic ratios is that they can only suggest probable sources of PAHs, without the capability
of providing source profiles and contributions (Halek et al., 2010). In addition, the PCA outputs
are associated with drawbacks, including the inability of this method to provide source
contributions, and the existence of negative factor loadings in the outputs of the analysis, which
are not physically interpretable in source apportionment studies (Jang et al., 2013). The only
study of inhalation risk assessment of exposure to total PM 10-bound PAHs in Tehran was done
by Hoseini et al. (2016). However, the authors of that study only calculated the risks associated
with exposure to total PAHs levels without apportioning the estimated cancer risks to specific
sources of PAH (Hoseini et al., 2016).
The main objective of this study was to perform source-specific risk assessment of
ambient air PM2.5-bound PAHs in central Tehran, by employing the PMF model to apportion the
sources of PM2.5-bound PAHs in central Tehran in the period of May 2012 through May 2013.
42
Chemically speciated PM2.5-bound PAHs mass concentrations were used as the input to the PMF
model. The PMF source apportionment results (i.e., PAHs source profiles and contributions)
were then used to perform source-specific cancer risk assessment, using the BaP equivalent
method.

3.2. Methodology
3.2.1. Sampling Locations
PM2.5 samples were collected in a school dormitory (35° 42′ 40.33″ N, 51° 22′ 49.75″ E)
and a retirement home (35° 42′ 20.30″ N, 51° 22′ 14.41″ E), both located in central Tehran and
only 1.1 km away from each other. Both sampling sites were relatively close to traffic activities
in a major freeway (Chamran) and crowded streets. Figure 3.1 illustrates the location of sampling
sites. More information regarding the sampling locations can be found in Hassanvand et al
(2015, 2014).

3.2.2. Sampling schedule, methodology and instrumentation
Particle sampling was carried out by employing low-volume air samplers (FRM
OMNU™ air Sampler, multi-cut inlet; BGI, USA) to collect 24-h PM2.5 samples on PTFE filters
(47 μm diameter, SKC Inc.) from May 2012 through May 2013. A total of 144 and 132 24-hr
samples were collected in the retirement home and the school dormitory, respectively. 24-hour
filter samples collected in three or four consecutive days (depending on mass loadings) were then
composited together for chemical analysis (Hassanvand et al., 2015). A gas chromatography-
mass spectrometry (GC-MC) instrument was used to chemically analyze fourteen particle-bound
PAHs, including naphthalene (Nap), acenaphthylene (Acy), phenanthrene (Phen), anthracene
43
(Anth), fluoranthene (Flt), pyrene (Pyr), benzo[a]anthracene (BaA), chrysene (Chr),
benzo[b]fluoroanthene (BbF), benzo[k]fluoroanthene (BkF), benzo[α]pyrene (BaP),
dibenzo[a,h]anthracene (DahA), benzo[ghi]perylene (BghiP), and indeno[123-cd]pyrene (Ind) in
the PM2.5 samples (gas-phase PAHs were not measured in this study). We excluded samples for
which the concentrations of individual PAHs were below the detection limit (BDL), and a total
of 74 chemically analyzed samples (38 samples for school dormitory and 36 samples for the
retirement home) were used as input to the PMF model. The percentage of BDL values was
7.69% for the retirement home and 2.56% for the school dormitory. More detailed information
regarding the sampling methodology and chemical analysis can be found in Hassanvand et al
(2015). Data for meteorological parameters, including ambient temperature and wind speed were
also available for a nearby weather station (i.e., Geophysics Weather station) from the website of
Islamic republic of Iran Meteorological Organization (IRIMO).

44

Fig. 3.2. Sampling locations in Tehran.

3.2.3. PMF model
PMF is a multivariate receptor model, developed by the USEPA, for source
apportionment of a physically- or chemically-speciated target pollutant using the least-squares
regression method. The chemical mass balance (CMB) equation is solved as the governing
equation in this statistical model:
45

Where, represents the concentration of the th specie in the th sample; P refers to the
number of contributing factors to the target variable (here, the total PAHs concentrations); is
the mass concentration contributed by the th factor to the th sample; indicates the total
PAH mass concentration of the th specie in the th sample; and is the residual associated
with the th sample and th specie.
This receptor model employs the least-squares technique for minimization of the
objective function (Q) to resolve the best non-negative values for source profiles and
contributions:

Where, is the measured or estimated uncertainty value of a speciated data set
containing samples and species.
In addition to the concentration matrix, user-provided uncertainty values are also
included as input to the PMF model. These uncertainty values can be derived from laboratory
measurements or analytical methods (Hopke, 2003; Norris et al., 2014; Paatero and Tapper,
1994; P. Paatero et al., 2014; Paatero, 1997; Reff et al., 2007). Since these values were not
determined experimentally for our chemically analyzed species, the uncertainty values were
estimated for each sample using the following analytical equation suggested by Chueinta et al
(2000).
46

Where, is the estimated uncertainty of the th specie in the th sample; refers to
species concentration; is the limit of detection; and is also a constant which is derived as
following:

PAH source apportionment was performed using USEPA's PMF model version 5.0 in the
robust mode to minimize the impact of extreme values on the PMF final solutions. When the
robust mode is used in the model, the objective function (Q) is calculated by excluding the
samples with high uncertainty (i.e., values with scaled residuals > 4) (Brown et al., 2015; Norris
et al., 2014). Additionally, the PAH concentration matrix for both sites were combined as a unit
matrix (with 74 samples) and used as a unit input to the PMF model. This was done to increase
the number of samples for the PMF runs and, therefore, the statistical power of the model. It
should be noted that there is only 1.1 km distance between these two sampling sites, which
makes it quite reasonable to assume that the source profiles are consistent across the sites. This
assumption is also in agreement with the results of our previous PM2.5 source apportionment
study performed at the same sampling sites in Tehran, in which we found similar source profiles
and contributions across the two sites (Taghvaee et al., 2018). In addition, based on Table 3.1,
the PAH/PM2.5 ratios were quite similar for all individual PAHs at the two sampling sites
(Hassanvand et al., 2015), further supporting the fact that the two sites are exposed to similar
sources to the same extent, and substantiating the choice of a pooled analysis/PMF run on the
data from the two sites.
47
Table 3.1. Statistical characteristics of chemically analyzed ambient PAHs concentrations in both
sampling sites
Species
Tohid retirement home School dormitory
Mean Standard deviation Mean Standard deviation

Nap (ng/µg of PM 2.5) 0.05 0.03 0.05 0.04
Acy (ng/µg of PM 2.5) 0.05 0.02 0.05 0.04
Phen (ng/µg of PM 2.5) 2.11 0.91 2.28 1.32
Anth (ng/µg of PM 2.5) 0.78 0.32 0.76 0.55
Flrt (ng/µg of PM 2.5) 0.94 0.50 0.91 0.78
Pyr (ng/µg of PM 2.5) 0.95 0.59 0.90 0.79
BaA (ng/µg of PM 2.5) 1.10 0.55 1.05 0.77
Chr (ng/µg of PM 2.5) 1.71 0.73 1.68 1.04
BbF (ng/µg of PM 2.5) 1.07 0.63 1.01 0.85
BkF (ng/µg of PM 2.5) 0.53 0.31 0.53 0.55
BaP (ng/µg of PM 2.5) 0.23 0.13 0.24 0.21
DBahA (ng/µg of PM 2.5) 0.58 0.34 0.60 0.51
BghiP (ng/µg of PM 2.5) 0.46 0.22 0.45 0.30
Ind (ng/µg of PM 2.5) 0.09 0.04 0.09 0.05

A 10% extra modeling uncertainty was also used to achieve the most stable and
physically interpretable solution. This feature allows the model to take into consideration the
errors and uncertainties not accounted for in uncertainty values that are either reported by
laboratories or calculated using the abovementioned equations. Signal-to-noise ratio was used as
another important parameter for determining if the noise of data is the main source of variability
in our experimental measurements. It is noteworthy that S/N ratios greater than 1 indicate strong
species with good signals.
We also used the displacement (DISP) method, bootstrap (BS) analysis, and a
combination of DISP and BS (BS-DISP) as other statistical tools for estimation of uncertainties
in the PMF resolved solution. The DISP method explores the validity of the PMF solution in
terms of rotational ambiguity. In this regard, the PMF solution was valid if there were no factor
48
swaps for smallest dQmax (dQmax=4) and if the decrement of Q was less than 1%. For the BS
analysis, 100 runs were performed, and the results were regarded as reliable if more than 80% of
the factors were mapped (Norris et al., 2014). Eventually for the BS-DISP analysis, the final
solution was considered to be valid if less than 0.5% drop was associated with the Q value
(Norris et al., 2014; Reff et al., 2007).

3.2.4. Cancer risk characterization
Lung cancer risk assessment was performed using the BaP equivalent method. Several
previous studies have employed BaP as an index for determining the relative potency of different
PAHs using the Potency Equivalent Factor (PEF). It is assumed that BaP has a PEF of 1 and
other species are compared to BaP based on their toxicity (Kam et al., 2013; Sauvain et al., 2003;
Stayner et al., 1998; Tsai et al., 2001). PEF can be obtained for various groups of PAHs from the
Office of Environmental Health Hazard Assessment (OEHHA) database as well as from other
sources (Boström et al., 2002; Collins et al., 1998; Lagoy and Nisbet, 1992; Sauvain et al., 2003).
In this method, concentrations of individual PAHs are multiplied by their corresponding
PEF to obtain corresponding BaPeq concentration for each species. Total BaPeq for source “j” is
then calculated from the following equation:

Where, is the total BaPeq of the jth source; refers to the BaP eq of
the ith species in the th source; and n refers to the number of species.
Total BaPeq for each source is then multiplied by unit risk factors to obtain the lung
cancer risk associated with that individual source. It should be noted that the California
49
Environmental Protection Agency (CalEPA) unit risk factors used in this study are based on
epidemiological lifetime exposure (Bandowe et al., 2014; OEHHA, 1994) (Table 3.2). Therefore,
in order to calculate the cancer risk associated with outdoor exposure, the lifetime unit risk factor
was modified by multiplying it by a factor of 0.2, i.e. the estimated average fraction of time
people spend outdoors (Al horr et al., 2016; Choi and Spengler, 2014; Samet and Spengler,
2003).

Table 3.2. Unit cancer risk factors for BaP eq lifetime and outdoor exposure (Bandowe et al., 2014;
OEHHA, 1994)
CalEPA unit risk factor (m
3
/ng)
(Lifetime exposure)
CalEPA unit risk factor (m
3
/ng)
(Outdoor exposure)

However, when calculating the lifetime cancer risk, indoor and outdoor exposures to
BaPeq concentrations should be considered separately, since the indoor/outdoor PAHs levels as
well as the time fractions spent indoors and outdoors differ. For the outdoor exposure fraction of
lifetime cancer risk calculation (corresponding to 20% of exposure during lifetime), source-
specific ambient BaPeq concentrations derived from the PMF model can be directly applied.
Since people spend typically about 80% of their time indoors, the source-specific ambient BaPeq
concentrations resolved by the PMF model should be modified to account for the infiltration of
PAHs emitted by outdoor sources indoors. Hassanvand et al.(2014) reported no smoking or
cooking activity, i.e., the two major sources of indoor PAHs (Masih et al., 2012a; Sadiktsis et
al., 2016; Yassin et al., 2016), in the specific study sites. We, therefore, assumed that the indoor
PAH concentrations are a result of infiltration indoors of PAHs emitted by outdoor sources. This
assumption is further supported by the similar composition of the PAHs measured indoors and
outdoors in the study by Hassanvand et al. (2015). Based on the Hassanvand et al. (2015) study,
50
the average ( indoor/outdoor ratio for PAHs was 0.85 ( 0.37) in central Tehran. This ratio
is consistent and within the range of values reported in the literature in other cities (Arhami et al.,
2010; Hu et al., 2018). We, therefore, calculated source-specific indoor BaPeq concentrations
using the following equation:

As mentioned above, it was assumed that people spend approximately 80% of their time
indoors and 20% of their time outdoors. Thus, the lifetime lung cancer risk for indoor and
outdoor exposure can be determined using following equations:

The uncertainty associated with the source-specific BaPeq values were estimated using the
standard deviations (SD) for the absolute concentrations of individual PAHs in the PMF-resolved
factor profiles, according to the following equation (Farrance and Frenkel, 2012):

Where, is the uncertainty associated with the calculated BaPeq values for
source j; is the standard deviation of ith species in the th source; and is the potency
equivalent factor of ith species.
51
Also, the uncertainty of indoor BaP eq concentrations was estimated using the following
equation (Farrance and Frenkel, 2012):

Where, refers to the uncertainty associated with the indoor BaPeq levels
for the jth source; is the calculated indoor BaP(eq) concentration of jth source;
is the average indoor/outdoor ratio of PAHs concentrations; and is the
standard deviation (SD) of the indoor/outdoor ratio of PAHs concentrations.
Then, the uncertainty associated with the estimated cancer risks for each individual
source was calculated according to the following equation (Farrance and Frenkel, 2012):

Where, is the uncertainty associated with the estimated lung cancer risk (LCR)
for source j; and URF refers to the CalEPA’s unit risk factors (URF) used for outdoor and
lifetime exposure LCR estimation.

3.3. Results and discussion
3.3.1. Overview of the data
Table 3.3 shows the statistical characteristics of measured and chemically analyzed
species included in the PMF analyses. The arithmetic average, minimum (min), maximum (max),
standard deviation (SD), and signal to noise ratio (S/N) are presented in this table. In addition, it
52
should be noted that the S/N ratio is significantly higher than 1 for all of the species, suggesting
strong signals for all species. Figure 3.2 illustrates the correlations between the measured and
model-predicted total PAHs mass concentrations. Based on the figure, the PMF model was able
to perfectly predict the total PAHs mass concentrations, judging by the strong association
(slope=1.05, R²=0.99) between the predicted and measured PAHs concentrations.

Table 3.3. Statistical characteristics of the chemically analyzed ambient particle-bound PAHs
concentrations
Species Mean Max Min Standard Deviation

S/N
PM 2.5 Mass (µg/m
3
) 32.44 118.00 10.01 17.38 9.00
Nap (ng/µg of PM 2.5) 0.05 0.15 0.00 0.03 8.73
Acy (ng/µg of PM 2.5) 0.05 0.16 0.01 0.03 8.78
Ace (ng/µg of PM 2.5) 0.02 0.07 0.00 0.01 6.39
Flu (ng/µg of PM 2.5) 1.59 5.28 0.39 0.77 8.92
Phen (ng/µg of PM 2.5) 2.20 5.89 0.28 1.12 8.95
Anth (ng/µg of PM 2.5) 0.77 3.06 0.09 0.44 8.81
Flrt (ng/µg of PM 2.5) 0.93 4.10 0.08 0.65 8.77
Pyr (ng/µg of PM 2.5) 0.92 3.44 0.06 0.69 8.67
BaA (ng/µg of PM 2.5) 1.08 4.10 0.16 0.66 8.87
Chr (ng/µg of PM 2.5) 1.69 4.47 0.31 0.89 8.81
BbF (ng/µg of PM 2.5) 1.04 3.92 0.09 0.74 8.78
BkF (ng/µg of PM 2.5) 0.53 2.89 0.05 0.44 8.73
BaP (ng/µg of PM 2.5) 0.23 1.11 0.01 0.17 8.48
DBahA (ng/µg of PM 2.5) 0.59 2.65 0.08 0.43 8.85
BghiP (ng/µg of PM 2.5) 0.45 1.56 0.09 0.26 8.78
Ind (ng/µg of PM 2.5) 0.09 0.24 0.01 0.05 8.29

53
Fig. 3.2. Correlation between the chemically analyzed and predicted total ambient PAHs mass
concentrations by the PMF model

3.3.2. Number of Factors
Several PMF runs were performed to determine the optimal number of factors and most
suitable extra modeling uncertainty. The best solution was selected on the basis of factor profile
evaluation, the investigation of temporal variations in the cold (i.e., fall and winter) and warm
(i.e., spring and summer) seasons of the year, examination of Qrobust plot of PMF solutions with
different number of factors (Figure 3.3), results of the uncertainty analyses (i.e., BS, DISP, and
BS-DISP runs), and correlations between predicted and measured PAHs concentrations. Figure
3.3 shows the Qrobust value for different PMF runs with different number of factors. It can be seen
in the figure that increasing the number of factors up to 5 led to a major decrease in the Qrobust
value (i.e., a 50% decrease when increasing the number of factors from 3 to 5). However, further
increase in the number of factors did not change the Qrobust value significantly. For example,
Qrobust only decreased by around 10% when increasing the number of factors from 5 to 6. This,
combined with the evaluation of all other criteria that were mentioned above led to the selection
54
of five-factor as the best solution for the PMF run in this study (Brown et al., 2015). Figure 3.4
shows the PMF-resolved source profiles for the 5-factor solution. The relative contributions of
PMF-resolved factor profiles to total PAHs mass concentrations are also presented in Figure 3.5.
The seasonal averages for meteorological parameters, including ambient temperature, and wind
speed are shown in Table.3.4. Finally, the temporal variations of the normalized contribution of
the PMF-resolved sources to the total PAHs mass concentrations in the cold and warm phases of
the year are illustrated in Figure 3.6.

Fig. 3.3. Q robust values for different PMF solutions

55
Fig. 3.4. PMF source profiles for 5-factor solution

56

Fig. 3.5. Relative contribution of PMF resolved factor profiles to total ambient PAHs mass concentration
for the whole study period

57
Table 3.4. Seasonal average of meteorological parameters in central Tehran during the study period
(Errors correspond to one standard deviation (SD)).

Average prevailing
wind direction
Average wind speed
(m/s)
Average temperature (
0
C) Season
W/SW
3.8 0.4 23.0 5.6
Spring
W/SW
3.1 0.3 29.3 2.8
Summer
W/SW
2.5 0.3 12.6 6.8
Fall
W/SW
2.6 0.4 9.5 3.6 Winter

Fig. 3.6. Temporal variation in relative contribution of the PMF resolved factor profiles to the total PAHs

58
3.3.3. Factor identification
3.3.3.1. Petrogenic sources and petroleum residue
The first factor is characterized by high levels of Nap, Acy, Phen, Anth, Flrt, Chr, and has
a significant contribution (23%) to total PAH mass concentrations (Figures 3.4, and 3.5).
Previous studies have indicated that 2- and 3-ring PAHs (i.e., lower molecular weight PAHs),
such as Nap, Acy, Phen, and Anth, are important tracers for petrogenic sources (Aydin et al.,
2014; Yunker et al., 2002). It should be noted that these PAHs can also be formed during the
pyrolysis of unburned fossil fuels (Dachs et al., 2002; Wang et al., 2015). Phen and 3-4 ring
PAHs (such as Flrt and Chr) are also known as chemical markers of crude oil leakage and
petroleum residue (Wang et al., 2014; Zakaria et al., 2002). In addition, presence of Phen and
Chr are indicative of a factor with volatile PAHs (Liu et al., 2009). Analyzing the temporal
trends for this factor also indicated slightly higher mean contributions in the warm season in
comparison to the cold season, although the difference is not statistically significant (P value=
0.33) (Figure 3.6). This, combined with the dominance of low molecular weight PAHs in this
factor, suggest that this factor likely originates from unburned petroleum (and its other
derivatives) due to evaporation (Jang et al., 2013; Meijer et al., 2008; Park et al., 2011). We
therefore elected “petrogenic sources and petroleum residue” as the most suitable title for this
factor. It is noteworthy that some of the main petrogenic sources of PAH are leakages from
underground and ground-level storage tanks. In Tehran, these losses are mainly occurring in gas
stations, due to the uncontrolled loading operation of underground storage tanks and vehicle
refueling activities (Shahbazi et al., 2016).

59
3.3.3.2. Natural gas and biomass burning
Significant loadings of Acy, Phen, Anth, Pyr, BaA, and Chr are associated with this
factor, which contributes 20% to total PAH concentrations (Figures 3.4 and 3.5). According to
the literature, BaA, Chr, Phen, Pyr, and Flrt are well-established chemical tracers for natural gas
combustion, which is the dominant method of residential and commercial heating and cooking in
Tehran (Callén et al., 2014a; Simcik et al., 1999). Wood combustion emission is also
characterized by high levels of BaP, Acy, and Anth (Jang et al., 2013; Mcdonald et al., 2000).
Additionally, it should be noted that BkF, BbF, and BghiP have also been associated with
biomass burning profile in previous studies (Gupta et al., 2011; Wang et al., 2014). Therefore,
this factor was labeled “natural gas and biomass burning”. Biomass burning occurs in Tehran in
the form of agricultural burning in the outskirts of the city and charcoal burning in Kebab shops
scattered in the city (Taghvaee et al., 2018). Finally, the investigation of the temporal trends for
this factor revealed higher contributions in the cold season compared to the warm season,
although the difference is not statistically significant (Pvalue= 0.25). This relatively higher
contribution can be explained by the higher occurrence of residential heating in the cold season
(Figure 3.6) (Taghvaee et al., 2018).

3.3.3.3. Industrial emissions
Factor 3 has high levels of BaA, Flrt, Pyr, and NaP, and contributes 17% to total PM2.5-
bound PAHs concentration (Figures 3.4 and 3.5). According to the previous studies, BaA, Pyr,
Flrt, and Phen are tracers of iron and steel production (Aydin et al., 2014; Odabasi et al., 2009),
as previously identified sources of pollution in Tehran (Shahbazi et al., 2016). In addition, Pyr
and Chr are chemical markers for industrial oil burning (Rajput and Lakhani, 2009). NaP can
60
also be regarded as a tracer for fugitive loss of petroleum (Khairy and Lohmann, 2013; Wang et
al., 2014). Based on the above discussion, we named this factor “industrial emissions”.
Previous studies in the area have indicated that there are approximately 3100 industrial
units in Tehran, contributing to ambient concentrations of gaseous and PM2.5 pollutants. High
levels of NaP (as a tracer of fugitive loss of petroleum) can be attributed to the presence of an oil
and petroleum refinery unit located to the south of Tehran (Shahbazi et al., 2016; Taghvaee et al.,
2018). DbahA, BaP, BkF, and Ind were also previously observed as dominant species present in
the ambient air of an industrial site (Rajput and Lakhani, 2009). Moderate loadings of Anth, and
BaP in this factor have also been previously associated with industrial profiles (Jang et al., 2013).
In agreement with the results of the study of Taghvaee et al. (2018), higher contributions are
observed for this factor in the cold season compared to the warm season, and the difference is
approaching statistical significance (Pvalue= 0.16). This temporal trend can be attributed to the
lower mixing heights and wind speeds, leading to more stable meteorological conditions, in the
cold season (Figure 3.6) (Taghvaee et al., 2018).

3.3.3.4. Diesel exhaust emissions
Factor 4 has high loadings of Flrt, Pyr, Chr, BbF, BaP, DBahA, BghiP, Ind and
contributes 22% to total PAH mass concentrations (Figures 3.4 and 3.5). According to the results
from previous studies, PAHs with more than four rings (i.e., BbF, BkF, BaP, DBahA, BghiP, and
Ind) mainly originate from combustion (pyrogenic sources) at high temperatures and vehicular
emissions (Dachs and Eisenreich, 2000; Harrison et al., 1996; Simcik et al., 1999; Wang et al.,
2015). In fact, BbF, BkF, DahA, and BghiP are specific tracers of traffic emissions (Li and
Kamens, 1993; Simcik et al., 1999; Venkataraman and Friedlander, 1994). Pyr and Phen are also
61
used as a diesel indicator to distinguish between diesel- and gasoline- powered vehicle emissions
(Alsberg et al., 1989; Harrison et al., 1996; Jang et al., 2013; Larsen and Baker, 2003).
Additionally, diesel emissions are identified on the basis of Chr as another important tracer (Tian
et al., 2013), which showed a high loading in this factor. This factor also had higher loadings of
Pyr and BkF than factor 5 (i.e., gasoline exhaust emissions), which is another indicator for
distinguishing between that diesel exhaust and gasoline exhaust emissions (Lee et al., 2004;
Wang et al., 2009, 2015). The above discussion corroborates the selection of “diesel exhaust
emissions” as the best title for this factor. This is consistent with previous studies identifying
diesel exhaust emissions with similar factor profiles (Aydin et al., 2014; Wang et al., 2015).
Finally, by analyzing the trends for temporal variations of this factor, higher contributions are
observed in the cold season compared to the warm season, although the difference is not
statistically significant (Pvalue= 0.42). As noted earlier for other sources/factors, this temporal
variation might be due to more stable meteorological conditions caused by lower wind speeds
and reduced atmospheric dispersion caused by temperature inversions in wintertime (Figure 3.6).

3.3.3.5. Gasoline exhaust emissions
This factor has very high loadings of higher molecular weight PAHs such as Ind, BghiP,
and significant loadings of BaA, and Chr, contributing to 18% of total PAH concentrations
(Figures 3.4 and 3.5). According to previous studies in the literature, BghiP and Ind are strong
indicators of vehicular emissions, particularly gasoline exhaust emissions (Andreou and
Rapsomanikis, 2009; Guo et al., 2003; Harrison et al., 1996; Li et al., 2011; Marr et al., 2006;
Simcik et al., 1999; Venkataraman and Friedlander, 1994). A few other studies have also
considered BaA, in addition to BghiP and Ind, as a significant indicator of exhaust emissions
62
from gasoline powered vehicles (Aydin et al., 2014; Esen et al., 2008; Tian et al., 2013), which
indicated a relatively high loading in this factor. Additionally, high loading of Chr in this factor
also corroborates the traffic-related origin of this factor (Wang et al., 2015). The temporal trend
for the contributions of this factor to total PAHs concentrations indicates higher contributions in
the cold season than in the warm season (Figure 3.6), although the difference, similarly to other
factors identified by our PMF, was not statistically significant (Pvalue= 0.37). This temporal
variation is consistent with the results of the study of Taghvaee et al. (2018), who found higher
contributions of traffic-related sources to PM2.5 concentrations in the cold season in Tehran
(Taghvaee et al., 2018). Similarly to factor 4, the higher level of contribution in cold seasons is
originated from stable meteorological conditions, lower wind speed, lower mixing height, and
inversion. Therefore, “gasoline exhaust emissions” was selected as the best title for this factor.

3.3.4. Source-specific lung cancer risk assessment
Tables 3.5 and 3.6 present the total BaPeq and lung cancer risk for each of the PAHs
sources as well as for the total PAHs concentrations associated with the outdoor and lifetime
exposure scenarios, respectively. According to these tables, diesel exhaust and industrial
emissions are the major sources of PAHs with significantly higher cancer risks than the others. In
addition, the epidemiological lung cancer risks for outdoor and lifetime exposure to total ambient
PAHs are (6.4 10
-6
and (2.8 10
-5
, respectively. It is also noteworthy that the
lifetime lung cancer risk for each individual PAHs source is in the range of 10
-6
-10
-5
, exceeding
the USEPA-recommended guideline value (10
-6
) (Bai et al., 2009; Tongo et al., 2017).
Therefore, it is vital to improve the ambient air quality in Tehran to minimize the risk of lung
cancer attributable to exposure to ambient PM-bound PAHs.
63
Table 3.5. Total BaP eq and lung cancer risk associated with outdoor exposure (Errors correspond to one
standard deviation (SD)).

Source Total
BaP(eq) outdoor
(ng/m
3
)
Lung cancer risk (*10
-6
)
outdoor exposure

Petrogenic sources and
petroleum residue
4.91 2.40 1.08 0.53
Natural gas and
biomass burning
2.99 1.09 0.66 0.24
Industrial emissions
7.94 3.77 1.75 0.83
Diesel exhaust emissions
11.19 3.40 2.46 0.75
Gasoline exhaust
emissions
2.10 0.53 0.46 0.12
Total PAHs
29.13 5.74 6.41 1.27

Table 3.6. Total BaP eq and lung cancer risk associated with lifetime exposure (Errors correspond to one
standard deviation (SD)).
Source BaP(eq) outdoor
(ng/m
3
)
BaP(eq) indoor
(ng/m
3
)
Lung cancer risk (*10
-6
)
Lifetime exposure
Petrogenic sources and
petroleum residue
4.91 2.40 4.17 2.73 4.75 2.93
Natural gas and biomass
burning
2.99 1.09 2.54 1.44 2.89 1.51
Industrial emissions 7.94 3.77 6.75 4.34 7.69 4.65
Diesel exhaust emissions
11.19 3.40 9.51 5.04 10.83 5.18
Gasoline exhaust
emissions
2.10 0.53 1.79 0.89 2.03 0.90
Total PAHs
29.13 5.74 24.76 7.39 28.20 7.77

The relative contribution of different PAHs sources to cancer risk are also shown in
Figure 3.7. Although the results of the PMF model indicated comparable contributions (of
around 20% for each of the identified PAHs sources) to total PAHs mass concentrations, the
cancer risk estimates do not follow the same trend. Diesel exhaust and industrial emissions have
significantly higher cancer risks than the other identified PAHs sources, together contributing to
approximately 70% of the lung cancer risk associated with exposure to ambient PM 2.5-bound
64
PAHs. In addition, while gasoline exhaust emissions contributed 18% to total PAHs
concentrations, the cancer risk associated with exposure to PAHs from gasoline powered
vehicles is much lower (only 7%) compared to that of other factors.

Fig. 3.7. The relative contribution of different PAH sources to lung cancer risk.

Finally, Table 3.7 presents the comparison of lung cancer risks due to exposure to total
PAHs concentrations estimated in this study with those reported in previous studies in the
literature (Akyüz and Çabuk, 2008; Bandowe et al., 2014; Kam et al., 2013; Manoli et al., 2016;
Martellini et al., 2012; Masih et al., 2012b; Masiol et al., 2012). The total BaP eq was selected as
the best parameter for this comparison, mainly because BaPeq is directly derived from mass
concentrations of different PAHs using PEFs, so it can be directly compared and contrasted
across studies, while the unit risk factor and lung cancer risk are parameters mainly dependent on
our assumptions for exposure scenarios (i.e., lifetime, occupational, or other scenarios of
exposure), which are different across studies. Our results indicate that the total BaP eq is higher
(29 ng/m
3
) in central Tehran compared to other cities around the world. Of particular interest is
65
the total BaP eq level (0.85 0.41 ng/m
3
) reported for the city of Thessaloniki, Greece, which is
significantly lower than what we observed in this study, even though Thessaloniki is one of the
most polluted European cities in terms of ambient particulate pollution (Argyropoulos et al.,
2016; Moussiopoulos et al., 2008). Similarly, the reported BaPeq levels in two other European
cities, i.e. Venice and Florence, were as low as 1.9 2.6 ng/m
3
and 0.79 ng/m
3
, respectively, both
being quite lower than the BaPeq level observed in Tehran

(Martellini et al., 2012; Masiol et al.,
2012). The city that came closest to Tehran in terms of BaPeq levels was Zonguldak, which is a
center of industrial activities in Turkey, for which winter time BaP eq levels as high as 22.5 ng/m
3

are reported (Akyüz and Çabuk, 2008). Of particular interest are the levels reported inside the I-
710 freeway in Los Angeles, California, in which the level of PAHs are significantly higher than
background levels and other regional freeways due to higher volume of heavy duty vehicles
(HDVs). As can be seen from this table, the levels observed in central Tehran are even higher
than levels measured inside this freeway that is characterized by a high percentage of HDVs
(Kam et al., 2013; Liacos et al., 2012; Lovett et al., 2017; Ning et al., 2008), stressing the
necessity of regulatory actions to reduce levels of PAHs in the city to mitigate the associated
adverse health effects, including lung cancer incidence, associated with exposure to these air
pollutants.

66
Table 3.7. Comparison of total BaP (eq) results with previous studies (Errors correspond to one standard
deviation (SD)).

Study Location BaP(eq) (ng/m
3
)
Present study central Tehran, Iran 29.1 5.7
Akyüz and Çabuk, 2008 Zonguldak, Turkey
22.5 4.2
Kam et al., 2013 I-710 freeway in Los
Angeles
23.3 4.4
Bandowe et al., 2014 Xi'an, China 17
Masih et al., 2011 Northern India 14.03

Kam et al., 2013 I-110 freeway in Los
Angeles
12.7 2.1
Kam et al., 2013 Surface streets (Los
Angeles)
8.6 1.5
Kam et al., 2013 Background
(USC metro)
7.4 1.1
Kam et al., 2013 Los Angeles Metro
gold line (light rail)
6.3 1.0
Kam et al., 2013 Los Angeles Metro red
line (subway)
5.1 0.8
Masiol et al., 2012 Venice-Mestre,
Northern Itally
1.9 2.6
Manoli et al., 2016 Urban Traffic site at
Thessaloniki, Greece
0.85 0.41
Martellini et al., 2012 Urban traffic site at
Florence, Italy
0.79

3.3.5. Limitations of the study
Although the results of this study provide insights into the source-specific lung cancer
risk associated with exposure to ambient PAHs in Tehran, we acknowledge that there might be
some limitations associated with our methodology and analysis. First, we recognize that the
differences in the number of rings between the indoor- and outdoor-generated PAHs might lead
to different toxicity and carcinogenicity. We should note, however, that, as the title of our study
indicates, the risk assessment in our investigation was based entirely on exposure to outdoor
emission sources, since we have no means of determining individual exposures to PAHs of
67
indoor origin in a highly populated and very diverse area like central Tehran. Therefore, any
other risk due to indoor-generated PAHs is above and beyond the cancer risk that was estimated
based on exposure to outdoor-generated PAHs, which affects the entire population of that city
and was the focus of our study.
In addition, as mentioned in the Methodology section, smoking and cooking are the two
major indoor sources of PAH (Masih et al., 2012a; Sadiktsis et al., 2016; Yassin et al., 2016).
However, we did not observe such activities in the school dormitory and the retirement home
during the entire campaign (Hassanvand et al., 2014). Furthermore, as reported in the study of
Hassanvand et al. (2015), the composition of the PAHs measured indoors and outdoors in both
sampling locations were quite similar. This evidence further supports our approach in basing our
entire lung cancer risk calculation on PAHs of outdoor origin.
Another limitation of this study is that, although the BaPeq method takes into account the
carcinogenicity of all PAHs, recent studies have indicated that a few PAHs, including
dibenzo(a,h)anthracene, may have higher carcinogenicity than BaP (USEPA, 2010), therefore the
BaPeq method may not be fully representative of the carcinogenic potential of exposure to such
compounds. These caveats notwithstanding, the BaPeq is well-established approach in estimating
cancer risk attributable to airborne PAH, and has been used in many recent publications to assess
the cancer risk associated with exposure to PAHs (Alves et al., 2017; Bandowe et al., 2014; Bian
et al., 2016; Castaneda et al., 2017; Kam et al., 2013; O.Ogbonnaya et al., 2017; Zhuo et al.,
2017). The selection of the BaP eq method in our study made it easier to put our data in
perspective by comparing the risk values in Tehran to those in other cities in the world.
Finally, we acknowledge that the BaPeq method is a calculation-based approach for lung
cancer risk assessment, rather than an epidemiological evaluation of that risk, and as such, the
68
accuracy of this method may be lower than that of epidemiological approaches. Ideally, our
BaPeq approach could be used in conjunction with epidemiological lung cancer data for a more
rigorous cancer risk assessment in central Tehran. However, as mentioned in the Introduction
section, epidemiological lung cancer data in Tehran are quite scarce. In the absence of adequate
epidemiological data, the BaPeq method was implemented as the next best approach to derive an
approximate estimate of the lung cancer risk associated with exposure to ambient PAHs in this
city.

3.4. Summary and conclusions
The main objective of this research was to carry out a source-specific lung cancer risk
assessment for PM2.5-bound PAHs in central Tehran. For this purpose, we used the PMF model
for source apportionment of airborne PAHs. Petrogenic sources and petroleum residue, natural
gas and biomass burning, industrial emissions, diesel exhaust emissions, and gasoline exhaust
emissions were the 5 sources with approximately equal contribution to the total PAHs
concentrations. The BaPeq method was used along with the PMF resolved source contributions to
calculate the cancer risk associated with exposure to each individual source of PAHs. The
calculated source-specific lifetime cancer risk levels were higher than the guideline value of 10
-6

set by the USEPA. Diesel exhaust and industrial emissions were found to produce higher cancer
risks compared to other sources (together contributing to approximately 70% of cancer risk),
while a negligible cancer risk (only 7%) was related to gasoline exhaust emissions, even though
this source contributed to almost 20% of PAHs concentrations. The results from the present
study have important implications in terms of public health. While not based on epidemiological
data, our results indicate significant health risks due to exposure to PAHs in the central part of
69
Tehran, an area with a high level of PAH pollution and very high population density, together
maximizing the impact of deteriorating air quality on human health. Risk assessment
methodologies, such as the BaPeq method employed in this study, provide us with approximate
estimates of the magnitude of cancer risk due to exposure to PAHs, and could be combined with
more robust epidemiological studies to develop a fundamental understanding of the excessive
lung cancer risk caused by exposure to PAHs. This information is crucial for policy makers to
develop effective air pollution control strategies that can mitigate adverse health outcomes of
exposure to these hazardous organic compounds.

70
Chapter 4: Source apportionment of the oxidative potential of fine ambient particulate
matter (PM2.5) in Athens, Greece

4.1 Introduction
Particulate matter (PM) is one of the most critical pollutants in the atmosphere,
particularly in metropolitan areas that are heavily impacted by emissions from vehicles,
industrial facilities, and other sources of urban air pollution. Among the different size fractions,
ambient fine PM (i.e., air suspended particles with an aerodynamic diameter smaller than 2.5
µm, PM2.5) is of great importance due to its distinct physico-chemical characteristics and diverse
sources, as well as its adverse human health impacts due to acute and chronic exposures.
Previous studies have indicated that exposure to PM2.5 is associated with several detrimental
health impacts, including increased daily mortality as well as numerous cardiovascular,
neurological, and respiratory diseases, such as chronic obstructive pulmonary disease (COPD)
(Dockery and Stone, 2007; Gauderman et al., 2015). One of the main biological mechanisms
believed to be contributing to the etiology of many of these detrimental health effects is the
generation of reactive oxygen species (ROS) in cells, which triggers an oxidative stress response
involving several proinflammatory cascades that ultimately result in pathology (Ayres et al.,
2008; Jiang et al., 2016). When the intracellular ROS concentration exceeds an equilibrium
threshold determined by available antioxidants within the cell, cellular oxidative stress and, in
turn, adverse health outcomes ensue (Delfino et al., 2013; Li et al., 2009). Thus, several
researchers have sought to develop biological and chemical assays to quantify the oxidative
potential of particulate matter as indexed by this cellular oxidative stress response (Verma et al.,
2015). The in vitro 2’,7’-dichlorodihydrofluorescein (DCFH) assay, which is conducted using
71
alveolar macrophage cell cultures derived from rat lung tissue, is one of the most widely used
measures of oxidative potential (Landreman et al., 2008; Shirmohammadi et al., 2017), which
has also been linked to markers of airway and systemic inflammation (Delfino et al., 2010;
Zhang et al., 2016).
It should be noted, however, that when evaluating the oxidative potential and adverse
health effects of ambient PM, not all of the chemical components and size ranges could be
considered equally toxic. Numerous studies have shown that some PM components, including
redox active metals such as Fe, Cu, Mn and V (Akhtar et al., 2010; Charrier and Anastasio, 2012;
Gasser et al., 2009), water-soluble organic carbon (WSOC) (Bae et al., 2017; Samara, 2017;
Verma et al., 2012, 2009; Vreeland et al., 2017), elemental carbon (EC) (Cho et al., 2005;
Kleinman et al., 2012; Samara, 2017; Shirmohammadi et al., 2017), organic carbon (OC)
(Chirizzi et al., 2017; Samara, 2017; Styszko et al., 2017), and polycyclic aromatic hydrocarbons
(PAHS) (Cho et al., 2005; Lundstedt et al., 2007; Shirmohammadi et al., 2016), are significantly
associated with the oxidative potential of PM, while many other components, including
secondary inorganic ions, are rather innocuous species, and while they may contribute
significantly to PM mass concentrations, they do not have an appreciable impact on PM toxicity
(Fang et al., 2016). In addition, particles emitted from various sources each have distinct
chemical fingerprints, and many common chemical PM species are released from a wide variety
of sources (Hasheminassab et al., 2014b; Manousakas et al., 2017b; Mousavi et al., 2018b; Port
et al., 2017; Taghvaee et al., 2018). Therefore, it has become increasingly important to link the
chemical species and sources of ambient PM with oxidative potential, which provides us with
critical information to effectively reduce emissions from sources that release PM with greater
toxicity.
72
Few studies have evaluated the adverse health effects and air quality deterioration
associated with ambient PM in Athens, the capital of Greece, and one of Europe’s largest
metropolitan areas, with approximately 5 million residents (Economopoulou and
Economopoulos, 2002; Karakatsani et al., 2003; Mantas et al., 2014). The economic crisis in
Greece, starting in 2010, caused a great increase in the price of residential heating fuel oil, which
in turn led to extensive and widespread burning of biomass, particularly wood, as an alternative
fuel (Argyropoulos et al., 2016; Diapouli et al., 2017b). As a result, ambient wintertime PM2.5
and BC concentrations have increased remarkably over the past few years in Athens
(Athanasopoulou et al., 2017; Paraskevopoulou et al., 2014) and are observed to be also spatially
widespread across the Athens metropolitan area (Kalogridis et al., 2018). Previous studies in this
city have indicated that local emission sources, including vehicular activities, secondary aerosols,
and biomass burning, are the major contributors to PM2.5. Additionally, the chemical
composition of PM2.5 is highly dominated by secondary organic and inorganic compounds in
Athens (Amato et al., 2016; Diapouli et al., 2017b; Pateraki et al., 2019). However, none of the
studies have so far linked the chemical components and sources of ambient PM 2.5 to its oxidative
potential in this city.
The main objective of this study was to characterize chemically and toxicologically
ambient PM2.5, and to link individual chemical components and sources of PM2.5 to the measured
oxidative potential in a suburban sampling site, approximately 7 km to the northeast of the city
center of Athens, Greece. Ambient PM2.5 samples were collected in two seasons of the year,
including a summer, or warm season (July-September), and a winter season (February-March).
Collected PM samples were then analyzed for their chemical components, and the PM oxidative
potential was determined using the DCFH in vitro assay. Principal component analysis (PCA)
73
coupled with multiple linear regression (MLR) were used to link sources of ambient PM 2.5 to the
measured oxidative potential.

4.2. Experimental methodology
4.2.1. Sampling site
PM2.5 samples were collected at the Global AtmosphereWatch (GAW) Demokritos
station (DEM_Athens) in Athens, Greece. The DEM_Athens station is located inside the
National Center for Scientific Research “Demokritos” campus (37° 59' 25'' N, 23° 48' 34'' E, 270
meters above sea level (m.a.s.l.)). The campus is situated in a suburban region, approximately 7
km northeast of the metropolitan center of the city, and covers an area of 600 acres in a forest of
pine trees, at the foot of Mount Hymettus. The site is not directly impacted by fresh urban PM
emissions and is thus considered representative of the urban background air quality in Athens
(Eleftheriadis et al., 2014; Triantafyllou et al., 2016).
4.2.2. Sample collection and analysis
Ambient PM2.5 samples were collected during two seasons of the year, including summer
(June 5 to September 15, 2017), and winter (February 2 to March 16, 2018) campaigns. Weekly
time-integrated PM2.5 samples were collected on 37 mm PTFE (Teflon) and Quartz filters (Pall
Life Sciences, 2-mm pore, Ann Arbor, MI) using Sioutas Personal Cascade Impactor Samplers
(PCISs, SKC Inc., Eighty-Four, PA, USA)(Misra et al., 2002b; Singh et al., 2003) operating at a
flow rate of 9 L/min. Therefore, the sampled air volume for each weekly time-integrated sample
was 90.72 m
3
. In this study, since we aimed to collect the whole PM2.5 range without size
segregation, we only used stage A of the PCIS to remove particles > 2.5 µm, and used 37-mm
74
Teflon and Quartz filters in the “after-filter” stage of the PCIS to collect PM2.5. In addition, the
collection plate of stage A was covered with a thin layer of grease to prevent bouncing of coarse
particles from stage A onto the after-filter stage. The Teflon filters collected with the PCISs were
used for the DCFH in vitro assay, and water-soluble organic carbon (WSOC), while organic
species (i.e., levoglucosan) analyses was performed on Quartz filters. The mass concentrations of
PM2.5 samples on 37mm PTFE and Quartz filters were quantified gravimetrically by weighing
these filters before and after sampling. This procedure was done using a high precision (± 0.001
mg) microbalance (MT5, Mettler Toledo Inc., Columbus, OH), following filter equilibration
under controlled temperature (22-24 °C) and relative humidity (40-50%) conditions. Chemical
analysis of the samples was carried out at the Wisconsin State Lab of Hygiene (WSLH) for
levoglucosan and water-soluble organic carbon (WSOC). Gas Chromatography/Mass
Spectrometry (GC/MS) was employed to determine levoglucosan concentrations (Schauer et al.,
1999). Finally, a Sievers 900 Total Organic Carbon Analyzer was used to analyze the samples
for water-soluble organic carbon (WSOC) following extraction and filtration (0.22 μm pore size)
of the filters using ultrapure water (Stone et al., 2008).
In addition to PCIS sampling, a sequential low volume reference sampler equipped with a
PM2.5 inlet (PM10/2.5 SEQ 47/50-CD with Peltier cooler, Sven Leckel GmbH) (2.3 m
3
/h) was
used to collect daily 24-h samples on 47-mm Teflon filters (PTFE Watman, 1 μm pore size). 24-
h PM2.5 concentrations were determined gravimetrically by a microbalance (Sartorius Model BP
211 D) of 0.01 mg accuracy. Prior to weighing, the filters were conditioned for 48 h at 19-21 °C
and 45-50% RH. These filters were also analyzed by ED-XRF (Epsilon 5, PANalytical) for
major and trace elements. Details on the analysis and QA/QC procedures may be found in
Manousakas et. al (2017a). Furthermore, a model-4 semi-continuous OC/EC field analyzer
75
(Sunset Laboratory Inc, USA) was used for the quantification of EC/OC by thermo-optical
transmittance (TOT) analysis on a 3-hour basis. The instrument was sampling at a flow rate of 8
lpm, from a PM2.5 cut-off inlet and was equipped with an in-line parallel carbon denuder for the
removal of organic gases. The EUSAAR2 protocol was applied for sample analysis (Cavalli et
al., 2010).

4.2.3. Determination of PM oxidative potential via the DCFH assay
To determine the oxidative potential of the collected PM samples, the fluorogenic DCFH
in vitro assay, which uses rodent alveolar macrophage cells (i.e., scavenging cells in the inner
epithelial lining of the lung), was employed. For this assay, 1.00 ml sterilized Milli-Q water was
used to extract PM from the Teflon filters while subjecting them to 16 hours of agitation at room
temperature in the absence of light, followed by 30 minutes of sonication. During the treatment
phase of this assay, cultures of the NR8383 rat alveolar cell line (American Type Culture
Collection) were exposed to the suspended PM aqueous phase slurries derived from filter
extraction along with 15 µM concentrations of 2’,7’-dichlorodihydrofluorescein diacetate
(DCFH-DA) in 96-well cell culture plates. As soon as it enters a cell, DCFH-DA is converted to
the non-fluorescent DCFH through a process of de-acetylation by intracellular esterase enzymes.
The non-fluorescent DCFH is then converted to the highly fluorescent dichlorofluoroscein
(DCF) due to oxidation by reactive species produced in the cytoplasm of the cells. DCF
production was monitored spectrophotometrically using a microplate reader, which provides a
reliable index of PM oxidative potential in fluorescence units per mass of PM (FU•μg
−1
PM).
Moreover, since the Toll-like receptors (TLR-2) of macrophage cells are able to identify
Zymosan, this glucan was used in the assay as a positive control for observation of a potent
76
response. The intrinsic PM oxidative potential can be obtained in units of μg Zymosan.μg
−1
PM
by normalizing the fluorescence data (reported in FU.μg
−1
PM) to the response of a unit of
Zymosan. The intrinsic PM oxidative potential can then be multiplied by the ambient PM mass
concentration for the corresponding samples to obtain the extrinsic PM induced toxicity (i.e., in
units of μg Zymosan.m
−3
air) (Landreman et al., 2008).

4.2.4. Source Apportionment of PM2.5 and its associated oxidative potential
In this study, the principal component analysis (PCA) was done using daily 24-h samples
of EC, OC, trace, and metal elements to identify the major source factors that contribute to PM2.5
concentrations. Assuming that correlated chemical species have similar emission sources, PCA
was performed to find out the clusters of chemical components, each representing specific PM 2.5
sources. Varimax orthogonal rotation method was utilized to better fit the data and provide
optimized factor loadings, allowing for the matching of each component with a single factor. In
addition, source factors having eigenvalues greater than unity were considered to be significant
sources (Argyropoulos et al., 2016). The Kaiser-Meyer-Olkin (KMO) value we also set to above
0.5 to ensure the suitablity of the data for the PCA procedure (Pechenizkiy et al., 2004).
Volumetric concentrations of weekly sampled WSOC, OC, EC, levoglucosan, trace and
metal elements, and the corresponding PM2.5 oxidative potential were used to perform bivariate
correlation analysis by calculating Spearman’s rho (non-parametric) coefficients. The correlation
coefficients provide us with preliminary data, allowing us to identify species that are highly
correlated with measured PM oxidative potential. We then employed the MLR analysis between
these species (as independent variables and chemical markers of various PM2.5 sources) and PM
oxidative potential (as dependent variable) to determine the source factors most responsible for
77
driving the PM2.5 oxidative potential. According to the approach suggested by previous studies
(Argyropoulos et al., 2016; Shirmohammadi et al., 2018), we ran MLR using various
combinations of source markers, and kept the statistically significant (p-value < 0.05) species
that led to the highest R
2
regression values. Then, the standardized regression coefficient (Beta)
of each chemical marker was used along with the derived R
2
value of the regression line to
determine the relative contribution of each source to PM2.5 oxidative potential. Additionally, the
standard error (std. error) for both unstandardized (B) and standardized coefficient (Beta) was
calculated. The following equation was used for deriving the standard error for unstandardized
coefficients of a multiple linear regression estimate:
(1)
) ' (
N
Y Y
σ
est

−
=
2

where,
est
σ refers to the standard error of the regression estimate; Y is the actual score
(sample); ' Y is the predicted score from the MLR estimation; and N is the number of samples
(Everitt and Skrondal, 2010). In order to calculate the standard error for standardized coefficient
(Beta), samples should be standardized by converting the values into the Z-score, using the
equation below:
(2)
Y
Y σ
Z
−
=
where, Z-score is the standardized format of score (sample), σ is the standard deviation of
score (sample), andY refers to the mean of samples (Kreyszig, 2011). Finally, the standard error
of Beta can be obtained by substituting the derived Z-score and their corresponding mean
(instead of Y and ' Y ) in equation (1).

78
4.3. Results and Discussion
4.3.1. Mass concentration and chemical composition of PM2.5
4.3.1.1. Concentrations of PM2.5 mass and carbonaceous species
Figure 4.1(a) and Table 4.1 presents the average PM2.5 mass concentrations of collected
samples using PCIS during the summer and winter seasons as well as during the whole study
period (i.e., average of warm and cold seasons). Based on this figure, comparable PM2.5 mass
concentrations (P-value = 0.890) were observed in the cold (11.8 ± 6.0 μg/m
3
) and summer
seasons (11.2 ± 3.3 μg/m
3
). Typically, stable meteorological conditions, lower wind speed, and
lower mixing heights result in higher PM2.5 levels in cold periods of the year (Sowlat et al.,
2016a). However, as can be seen in Table 4.2, in the present study, the wind speeds are
comparable during the cold and warm season, leading to similar degrees of atmospheric
instability and horizontal dispersion of the emissions, which could be one of the reasons for
observing comparable PM2.5 concentrations in cold and warm seasons. Consistent with our
results, previous studies in suburban areas of Athens have also indicated comparable PM2.5
concentrations in both seasons (Diapouli et al., 2017b; Vasilatou et al., 2017). In addition, the
significant contribution of SOA to PM2.5 mass concentrations during warm period
counterbalances the impact of meteorological conditions that increase PM2.5 concentrations
during cold period, leading to comparable PM2.5 levels in both phases of the sampling.
Furthermore, the average PM2.5 mass concentration during the whole study period (11.4± 4.4
μg/m
3
) is in agreement with the results from the recent studies of Amato et al.(2016) (at the same
sampling site) and Diapouli et al. (2017b). This agreement also corroborates the fact that our
sampling has been representative of a whole-year sampling, and that the average PM2.5 level
measured in this study can be considered as the annual average of PM2.5 in the study area.
79
Fig. 4.1. The average mass concentrations of a) PM 2.5 during the entire study period, and by season; b)
carbonaceous components (elemental carbon (EC), organic carbon (OC), and water soluble organic
carbon (WSOC)) by season. Error bars correspond to one standard deviation (SD).

(a)

(b)
80
Table 4.1. The overall and seasonal averages (± standard deviation) of PM 2.5 mass concentrations as well
as its associated oxidative potential and chemical components.

Species Warm phase Cold phase Whole study period
PM 2.5 (µg/m
3
) 11.2  3.3 11.8  6.0 11.4  4.4
EC (µg/m
3
) 0.44  0.1 0.26  0.06 0.36  0.12
OC (µg/m
3
) 2.0  0.4 1.8  0.3 1.93  0.39
WSOC (µg/m
3
) 1.4  0.5 0.7  0.3 1.08  0.5
Levoglucosan (ng/m
3
) 5.5  0.8 44.8  39.0 24.8  33.54
S (ng/m
3
) 855.9  244.9 617.1  247.6 736.5  265.8
Al (ng/m
3
) 319.7  250.5 251.5  405.7 274.2  335.4
Si (ng/m
3
) 275.8  146.9 87.5  36.0 213.0  150.6
Ca (ng/m
3
) 166.5  52.4 68.4  44.7 121.9  69.2
Fe (ng/m
3
) 133.3  43.4 68.3  42.1 103.8  52.9
Ti (ng/m
3
) 9.6  4.7 5.5  6.3 7.7  5.6
Mn (ng/m
3
) 4.8  1.8 4.2  2.0 4.5  1.87
Na (ng/m
3
) 171.9  86.5 157.8  80.5 165.5  79.9
K (ng/m
3
) 107.4  27.3 111.5  39.7 109.4  32.2
Zn (ng/m
3
) 12.6  2.7 9.8  2.4 11.2  2.9
Cu (ng/m
3
) 9.9  2.0 9.0  2.3 9.5  2.1
Pb (ng/m
3
) 7.9  0.9 8.2  1.2 8.0  1.0
Ni (ng/m
3
) 2.0  0.4 1.4  0.2 1.75  0.5
Br (ng/m
3
) 3.3  0.7 3.4  0.7 3.3  0.7
Intrinsic PM OP 12910  4898.6 8549.3  4388.1 11041.6  5034.3
Extrinsic PM OP 150.6  84.4 83.3  35.8 120.5  80.7

Table 4.2. Seasonal averages (± standard deviation) of meteorological parameters during the warm and
cold period.

Temperature (°C) Relative Humidity (%) Wind Speed (m/s)
Warm period 26.5 ± 3.5 45.6 ± 12.0 1.9 ± 0.7
Cold period 11.9 ± 2.3 72.4 ± 12.0 2.2 ± 0.8

The seasonal average concentrations of carbonaceous compounds, including elemental
(EC) and organic carbon (OC), measured on a 3-h basis, and water soluble organic carbon
(WSOC), measured on a weekly basis, are illustrated in Figure 4.1(b). According to this figure,
higher concentrations were observed for EC in warm season (0.44 ± 0.1 μg/m
3
) compared to the
81
cold season (0.26 ±0.06 μg/m
3
), and the difference was statistically significant (P-value = 0.004)
according to the independent sample T-test. This could be attributed to the fact that our sampling
site is downwind of the city center, so vehicular emissions from the abundant traffic sources in
the city center reach the study area to a higher extent in the summertime due to the enhanced
atmospheric dispersion and mixing during that period (Toro Araya et al., 2014). The average EC
concentration for the whole study period (0.36±0.12 μg/m
3
) was also within the range of values
reported by previous studies carried out in the area (Amato et al., 2016; Paraskevopoulou et al.,
2014).
Investigating the seasonal trend of OC concentrations also revealed comparable
concentrations (P-value = 0.403) in cold (1.8 ± 0.3 μg/m
3
) and warm (2.0 ± 0.4 μg/m
3
) periods,
with summertime values being only slightly higher. In addition to primary sources (i.e., biomass
burning and traffic activities), OC can be originated from secondary sources and photochemical
activities, leading to the formation of SOA. Therefore, while the contribution of some primary
sources (such as biomass burning for residential heating) to OC concentrations are expected to be
higher during the cold season, their impact is counterbalanced by SOA contributions due to peak
photochemical activities in the summer, which increase OC concentrations to levels higher than
those observed in the cold season (Paraskevopoulou et al., 2014). In addition, the average OC
concentration observed for the whole study period (1.93±0.39 μg/m
3
) is in good agreement with
the OC values reported in the previous studies performed in urban background sites near Athens
(Amato et al., 2016; Paraskevopoulou et al., 2014). WSOC (a chemical marker of SOA (Arhami
et al., 2017; Snyder et al., 2009)) levels were enhanced during the summer (1.4 ± 0.5 μg/m
3
)
compared to the winter (0.7 ± 0.3 μg/m
3
), due to the photochemical formation of secondary
82
organic compounds (Hasheminassab et al., 2014b; Heo et al., 2009) and the winter vs. summer
difference was statistically significant (P-value = 0.012).

4.3.1.2. Levoglucosan Concentrations
The average concentrations of levoglucosan, one of the most widely used tracers of
biomass burning (Simoneit et al., 1999), during the whole study period, as well as by season, are
illustrated in Figure 4.2. According to the figure, the average concentration of levoglucosan was
significantly higher during the cold season (44.8 ± 39.8 ng/m
3
) as opposed to the warm season
(5.5 ± 0.8 ng/m
3
) (P-value = 0.039). This trend clearly reveals the higher wood burning
activities, particularly for domestic heating purposes, in colder months of the year, as observed in
previous studies (Diapouli et al., 2017a, 2017b; Gratsea et al., 2017). In addition, the average
levoglucosan concentrations for the whole sampling period (24.8 ±33.54 ng/m
3
) is quite
comparable with previously reported value by Amato et al. (2016) (37 ng/m
3
).

83
Fig. 4.2. Average levoglucosan concentrations during the entire study period, and by season. Error bars
correspond to one standard deviation (SD).

4.3.1.3. Metals and trace elements
Previous studies have indicated that major and trace elements can be originated from a
wide variety of sources such as resuspension of road dust, tire and break ware, road abrasion, and
soil dust emissions (Adamiec et al., 2016; Harrison et al., 2012). Figure 4.3 presents the average
concentrations of metal and trace elements during the warm and cold seasons. As can be seen in
the figure, crustal elements including Ca, Si, Ti, and Fe, that are typical chemical markers of soil
dust activities (Hasheminassab et al., 2014b; Zhang et al., 2013), had significantly higher
concentrations (P-value < 0.05) during the summer, due mainly to drier atmospheric conditions
(i.e., lower relative humidity in the summer (Table 4.2)) that facilitate particle resuspension from
the loose soil open surfaces in the city of Athens and the surrounding region (Athanasopoulou et
al., 2010). Furthermore, comparable concentrations were observed for Cu, and Pb (with P-values
84
of 0.351, and 0.605 respectively), that are used as tracers of road dust and non-tail pipe traffic
emissions (Adamiec et al., 2016; Hjortenkrans et al., 2007) during both phases of the sampling
campaign. Comparable concentrations of Na (P-value = 0.64) a chemical tracer of sea salt (Zong
et al., 2016), were observed during the summer and winter period, with slightly higher
concentrations in the warm phase. Generally, Na concentrations are expected to be higher in the
summer time due to higher wind speeds that facilitate the dispersion of sea salt particles further
inland. However, strong northeasterly winds experienced during the winter time in the Athens
area (Diapouli et al., 2017b) also bring sea salt to the area, which lead to observing comparable
Na concentrations during the warm and cold seasons. This is consistent with the meteorological
data presented in Table 4.2, showing comparable wind speeds during both seasons.

Fig.4.3. Average concentrations of metal and trace elements by season. Error bars correspond to
one standard deviation (SD).

85
Overall, S, Al, Si, Ca, and Fe were among the most abundant elemental species in the
collected PM2.5 samples. Additionally, concentrations of the measured trace elements and metals
were consistently within the range of reported values in previous studies (Amato et al., 2016;
Diapouli et al., 2017b; Paraskevopoulou et al., 2014). For example, in this study, the average
concentration of Fe and Si were 103.8 ng/m
3
and 213.0 ng/m
3
, respectively. These levels are in
good agreement with the results from the study of Amato et al. (2016) at the same sampling
location, reporting Fe and Si concentrations of 112.0 ng/m
3
and 234.0 ng/m
3
, respectively.

4.3.2. Oxidative potential of PM2.5
Figure 4.4 presents the intrinsic (per PM mass) and extrinsic (per air volume) levels of
PM2.5 oxidative potential in the warm and cold seasons. In addition, the average values of
intrinsic and extrinsic PM2.5 oxidative potential are shown in Table 4.1 during the two sampling
campaigns as well as the whole study period. While the mass-normalized oxidative potential (in
units of µg Zymosan/mg PM) is indicative of the intrinsic PM-induced toxicity and used in
conventional toxicology, the volumetric oxidative potential (in units of µg Zymosan/m
3
of air) is
a more suitable measure for PM population exposure and inhalation assessment studies
(Shirmohammadi et al., 2018, 2017). As can be seen in Figure 4.4(a), the mass-based oxidative
potential of the collected ambient PM2.5 samples was higher in summer (12910.7± 4898.6 µg
Zymosan/mg PM) as compared to winter (8549.3±4388.1 µg Zymosan/mg PM), although the
differences were marginally significant (P-value = 0.151). Similar temporal trends were observed
for the volumetric PM oxidative potential, indicating higher per-volume oxidative potential in
the summer (150.6± 84.4 µg Zymosan/m
3
) than in the winter (83.3± 35.8 µg Zymosan/m
3
)
(Figure 4.4(b)) with marginally significant statistical difference between seasonal levels (P-value
86
=0.113). Comparison of the mass-based oxidative potential levels measured in this study
(11041.6±5034.3 µg Zymosan/mg PM) with those of earlier studies carried out in different
metropolitan areas around the world indicated higher oxidative potential levels compared to
those of many other cities, including Denver (with the average of 2006.6 µg Zymosan/mg PM),
Thessaloniki (around 738.0 µg Zymosan/mg PM), and Los Angeles (with the average of 748.5
µg Zymosan/mg PM). It is noteworthy that the mass-normalized oxidative potential of the
ambient PM2.5 samples in this study are even higher than those measured inside major freeways
in the Los Angeles County, including I-110 (3660±1743 µg Zymosan/mg PM) and I-710
freeways (3439±3058 µg Zymosan/mg PM) (Shirmohammadi et al., 2017). In addition, the
mass-based oxidative potential of the ambient PM2.5 samples collected in this study are also
higher than that of the ultrafine PM (UFP) samples collected in central Los Angeles (9424±3703
µg Zymosan/mg PM) (Mousavi et al., 2018a), and near the Los Angeles international airport
(i.e., LAX) (4600.93 ± 1516.98 µg Zymosan/mg PM) (Shirmohammadi et al., 2018). These
results underscore the very high oxidative potential of the ambient PM2.5 in the metropolitan area
of Athens and its nearby suburban areas, which we believe is one of the most notable
observations of our study. Similar trends were also observed for the volume-based oxidative
potential of PM2.5, indicating much higher (by 3-5 fold) levels in our sampling site (120.5±80.7
µg Zymosan/m
3
) compared to other urban environments, including Denver (with the average of
20.9 µg Zymosan/m
3
), Thessaloniki (around 28.0 µg Zymosan/m
3
), and Los Angeles (with the
average of 35.8 µg Zymosan/m
3
) (Saffari et al., 2014). The higher oxidative potential level
observed in our sampling site is of particular significance because the elevated PM toxicity
pertains to an urban background area in Athens, compared with the heavily impacted traffic and
metropolitan areas investigated in previous studies.
87
Fig. 4.4. PM 2.5 oxidative potential for warm and cold phases: a) Mass-normalized oxidative potential; b)
Volume-based oxidative potential. Error bars correspond to one standard deviation (SD).

(a)

(b)

88
4.3.3. Source apportionment of ambient PM2.5 and its associated oxidative potential
4.3.3.1. Source apportionment of PM2.5 mass concentration using the PCA approach
Table 4.3 illustrates the results of the PCA analysis performed on the daily EC, OC, as
well as trace elements and metal concentrations during the whole study period. Our analysis
indicates that the PCA was able to identify four source factors, explaining about 95% of total
variance in the data. The identified factors were named as soil dust emissions (due to significant
loadings of Ca, Si, and Ti as crustal elements (Sowlat et al., 2012; Taghvaee et al., 2018)),
vehicular emissions (due to high loadings of EC, and Ni (Hasheminassab et al., 2014b; Zong et
al., 2016)), road dust and non-tail pipe traffic emissions (due to high loadings of Cu, and Pb
(Adamiec et al., 2016; Pant and Harrison, 2013)), and a fourth factor with a very high loadings of
OC (and to a lesser extent EC). The fourth factor cannot be attributed to a unique source, as OC
is originates from a multitude of sources, including both primary (i.e., traffic and biomass
burning), and secondary (i.e., SOA) sources. Therefore, as will be discussed in Section 4.3.3.3.,
we performed an additional MLR analysis to apportion OC concentrations to its contributing
sources.
Table 4.3. Loadings of chemical species in the factors resolved by the principal component analysis
(PCA). Loadings> 0.6 are bolded.

Species Source factors
Soil dust emissions Vehicular emissions Road dust emissions OC factor
OC 0.03 0.12 0.01 0.97
EC -0.29 0.75 -0.08 0.47
Ni -0.04 0.95 -0.03 0.01
Cu -0.38 0.51 0.63 -0.38
Pb 0.21 -0.16 0.95 0.07
Ca 0.98 -0.06 0.04 -0.08
Si 0.98 -0.16 0.06 0.04
Ti 0.98 -0.13 0.04 0.07
%Variance 44.56 20.81 18.12 11.99
%Cumulative 44.56 65.37 83.50 95.49
89
4.3.3.2. Correlation analysis between individual chemical species and PM oxidative potential
Bivariate correlation analysis was carried out between the volume-based PM oxidative
potential and weekly sampled concentrations of WSOC, OC, EC, levoglucosan, and trace
elements and metals to find species that are highly correlated with PM-induced toxicity. Table
4.4 shows the Spearman’s rho coefficient (R); in which the highly (with R>0.7) and relatively
highly (with R> 0.6) correlated species are bolded. Based on the table, OC was strongly
correlated with PM oxidative potential (with R>0.7). This result was well in agreement with the
results of previous studies, introducing OC as a significant fraction of PM with notable toxicity
(Samara, 2017; Styszko et al., 2017). Moderate-to-strong associations were also observed
between oxidative potential and EC, as well as with WSOC, in accordance with results from
previous studies demonstrating the oxidative potential of these toxic species (Shirmohammadi et
al., 2017; Vreeland et al., 2017). In addition, a moderate association was also observed between
PM-induced oxidative potential and Ni as tracer of traffic emissions (Querol et al., 2007).
Table 4.4. Spearman bivariate correlation analyses between volume-based oxidative potential (µg
Zymosan/m
3
) and weekly sampled concentrations of individual chemical species. Highly correlated
species are bolded, and significant values (P<0.05) are marked with an asterisk (*).

Species R Species R
WSOC 0.66
*
Ti -0.01
OC 0.78
*
Mn -0.24
EC 0.66
*
Na -0.25
Levoglucosan 0.03 K 0.20
S 0.48 Zn 0.06
Al -0.03 Cu -0.17
Si 0.03 Pb -0.11
Ca -0.14 Ni 0.47
Fe -0.07 Br 0.31

90
4.3.3.3. Source apportionment of PM oxidative potential using MLR approach
As mentioned in the Methodology section (Section 4.2.4), based on the results of the
correlation analysis between individual chemical components and the oxidative potential of
ambient PM2.5 samples, we performed an MLR analysis to determine the most significant
sources responsible for the PM-induced toxicity, using oxidative potential as the dependent
variable and weekly sampled individual species as source markers and independent variables; the
results of which are presented in Table 4.5(b). As can be seen in the table, EC and OC were the
most important species contributing to PM oxidative potential, with the standardized regression
coefficients (Beta) of 0.497 and 0.451, respectively. Finally, the coefficient of statistical
determination for the model indicates that 73% of variation in the oxidative potential of ambient
PM2.5 samples was explained by EC (a tracer of vehicular emissions) and OC.
Although EC is a tracer of vehicular emissions (Hasheminassab et al., 2014b; Heo et al.,
2009; Zong et al., 2016), OC is not a unique tracer of a specific pollution source. Previous
studies have indicated that a combination of primary combustion sources (i.e., biomass burning,
industrial emissions, airport, port and vehicular activities) as well as secondary formation
mechanisms (e.g., photochemistry producing SOA) contribute to OC concentrations (Arhami et
al., 2018; Paraskevopoulou et al., 2014). As a result, it is vital to identify the relative contribution
of these primary and secondary sources to OC and, in turn, to PM2.5 oxidative potential.
Therefore, we performed another set of MLR analysis using OC as the dependent variable and
individual marker species as independent variables. According to the standardized regression
coefficients presented in Table 4.5, the majority of OC was associated with WSOC (Beta=0.75)
which is a tracer of SOA (Arhami et al., 2017; Liu et al., 2018) and levoglucosan (Beta=0.42)
which is a tracer of biomass burning (Simoneit et al., 1999). The remaining loading was also due
91
to OC from traffic sources (Beta=0.27), represented by EC (Zong et al., 2016). Finally, this
model was able to explain 88% of total variance in the OC concentrations. It should be noted that
since WSOC also comes from primary biomass burning sources, we estimated the biomass
burning fraction of WSOC (i.e., WSOCbb) to support our choice of WSOC as a tracer for SOA in
this study, employing a technique used in previous studies. Based on this approach, the
concentration of levoglucosan, a well-known tracer of biomass burning (Simoneit et al., 1999),
was divided by 0.135, the ratio of levoglucosan/OC derived from the wood smoke profiles (Fine
et al., 2004), to derive the concentration of biomass burning originated OC. Furthermore,
assuming that 71% of biomass burning fraction of OC is water soluble (Sannigrahi et al., 2005),
the OC originating from biomass burning was multiplied by 0.71 to estimate the WSOC bb. The
WSOC from non-biomass burning sources (i.e., WSOCnb) was then derived by subtracting
WSOCbb from total WSOC. The results from this analysis indicated that WSOCbb (with the
average of 0.12 µg/m
3
) is a very small fraction (i.e. 11%) of total WSOC (with the average of 1.1
µg/m
3
), while majority of WSOC is composed of WSOCnb which can be regarded as a metric of
SOA formation (Snyder et al., 2009). These results confirmed the suitability of using WSOC as a
tracer of SOA in this study.
Table 4.5. Results of the multiple linear regression (MLR) analysis between a) PM 2.5 oxidative potential
(as the dependent variable) and selected chemical species (as independent variables); b) OC (as the
dependent variable) and selected chemical species (as independent variables).

(a)

92

(b)

Using the standardized coefficients and R
2
values obtained from the MLR models
described above, we calculated the relative contributions of EC (i.e., of vehicular emissions) and
OC to PM induced toxicity, the results of which are shown in Figure 4.5(a). As illustrated in the
figure, EC and OC contributed to 38% and 35% of oxidative potential, respectively. Similarly,
Figure 4.5(b) illustrates the relative contribution of different sources to OC mass concentrations.
According to this figure, WSOC (a tracer of SOA) had the highest contribution to OC
concentrations, accounting for 45% of its concentrations. Vehicular emissions (as represented by
EC) and biomass burning (as represented by levoglucosan) contributed to 26% and 17% of OC
concentrations, respectively. The relative contribution of each source to OC can then be
multiplied by the relative contribution of OC to PM oxidative potential (i.e., 35%) to obtain an
estimate of the relative contribution of that individual source to PM induced toxicity. Combining
the data presented in Figures 4.5(a) and 4.5(b), the relative contributions of vehicular emissions,
secondary sources, and biomass burning to PM oxidative potential were calculated, the results of
which are presented in Figure 4.5(c). As shown in the figure, traffic activities were responsible
for approximately 44% of PM-induced toxicity, followed by SOA (16%), and biomass burning
(9%). These results also explain the higher oxidative potential of the ambient PM2.5 samples that
were collected during the summer, which can be attributed to the higher concentrations of EC
93
and WSOC during that sampling period. Finally, according to previous studies in suburban
Athens, road dust (i.e., brake and tire wear) and oil combustion were also identified as sources of
PM2.5 (Amato et al., 2016; Diapouli et al., 2017b). The undetermined portion of PM oxidative
potential might therefore be attributable to these sources.

Fig. 4.5. Relative contribution of: a) EC and OC to PM 2.5 oxidative potential b) different sources to OC
mass concentrations; and c) different sources to PM 2.5 oxidative potential.

(a)

94

(b)

(c)

4.4. Summary and Conclusions
This study sought to identify pollution sources that contribute to the oxidative potential of
ambient PM2.5 in the metropolitan area of Athens, Greece. Our results indicated significantly
95
higher mass-normalized and volumetric oxidative potential of the ambient PM2.5 collected in the
study area as compared to those collected in many urban areas around the world. It is of
particular note that the oxidative potential of the PM2.5 samples in Athens was even higher than
that measured inside major freeways in Los Angeles, California. In addition, the significantly
higher PM toxicity in our sampling site is one of the most notable observations of this study,
since the measured oxidative potential pertains to a site which is not directly impacted by fresh
urban PM emissions. Our attempt to apportion sources of PM2.5 indicated that vehicular
emissions were responsible for approximately 44% of PM-induced oxidative potential of the
samples, followed by secondary organic aerosol (SOA) (16%) and biomass burning (9%) as the
next major contributors. Overall, our findings highlight the importance of traffic emissions and
photochemical reactions in deriving PM-induced toxicity in Athens, Greece, and can be used by
policy makers to prioritize the required public health actions and policies to mitigate the adverse
health effects of PM by controlling the emissions from sources that are causing most of the PM
toxicity in the area of Athens.

96
Chapter 5: Development of a novel aerosol generation system for conducting inhalation
exposures to ambient particulate matter (PM)

5.1. Introduction
Many metropolitan areas of the world suffer from severe air pollution as a by-product of
rapid urbanization, industrial, and technological advancements during the past few decades.
More than 90% of the world population is exposed to air pollutant concentrations exceeding
WHO guideline limits (WHO, 2018). Several epidemiological as well as toxicological studies
have well documented the harmful health impacts of exposure to ambient particulate matters
(PM), including neurological, cardiovascular, respiratory, and pulmonary diseases (Delfino et al.,
2010; Dockery and Stone, 2007; Gauderman et al., 2015; Morgan et al., 2011; Rich et al., 2013;
Wai et al., 2015). There is, therefore, a need to conduct in vitro and in vivo toxicological studies
to further evaluate the health effects of exposure to ambient PM, which requires the use of
aerosols that are representative of real-world PM. Ambient PM is comprised of several different
chemical constituents and varies in size by 5 orders of magnitude from a few nanometers to tens
of micrometers; this makes PM a complex pollutant to reproduce in the laboratory (Bladt et al.,
2012; Filep et al., 2016; Jacoby et al., 2011; Keskinen and Rönkkö, 2010). There are several
aerosol generators that are used for producing particles in the laboratory (Arefin et al., 2017;
Hahn et al., 2001; Polk et al., 2016; Shimada et al., 2009; Steiner et al., 2017). For example,
nano-objects and their aggregates and agglomerates (NOAA) are being used for nano-aerosol
generation (Ahn et al., 2017). Clemente et al. (2018) developed a novel aerosol generator to re-
aerosolize target nanomaterials (e.g., TiO and ZuO) with specific chemical composition.
However, these laboratory-generated particles are of unique physicochemical characteristics and,
97
therefore, do not adequately represent the physical and chemical characteristics of ambient real-
world aerosols (Lippmann and Chen, 2009). In addition, direct use of ambient PM in the lab is
not necessarily a viable alternative, as ambient PM concentrations are generally not sufficiently
high to induce acute adverse health effects in toxicological studies (Jung et al., 2010; Lippmann
and Chen, 2009; Liu et al., 2014).
The development of particle concentrators has resolved several of the above issues, as
these instruments are capable of significantly increasing the PM concentration in the inlet of the
exposure chamber (Chang et al., 2002; Demokritou et al., 2003, 2002; Gupta et al., 2004a,
2004b; Sioutas et al., 1999, 1997, 1995), reaching levels that can cause responses in
toxicological studies. The versatile aerosol concentration enrichment system (VACES) is one of
the most widely used aerosol concentrators which employs condensational growth followed by
virtual impaction to enrich PM concentrations in the air flow (Kim et al., 2001a, 2001b, 2000;
Ning et al., 2006; Pakbin et al., 2011; Sioutas et al., 1999). The VACES is able to effectively
concentrate different size fractions of PM (i.e., ultrafine, fine, and coarse PM) by 20-30 times,
while preserving the physical and chemical characteristics of ambient particles (Kim et al.,
2001a, 2001b; Saarikoski et al., 2014). In addition, VACES can also be used in tandem with a
Biosampler
TM
(SKC Inc, Eighty-Four, PA; Willeke et al., 1998) or the high-volume aerosol-into-
liquid collector to provide highly concentrated liquid suspension, which corroborates its
versatility and ability for simultaneous in vivo and in vitro exposure assessment experiments
(Daher et al., 2011; Kim et al., 2001a, 2001b; Wang et al., 2013). Due to the abovementioned
advantages, the VACES has been used in several toxicological (both in vivo and in vitro)
exposure studies to supply either of concentrated ambient particles (CAPs) or PM liquid
98
suspensions (Budinger et al., 2011; Maciejczyk and Chen, 2005; Mills et al., 2011; Qiu et al.,
2017; Quan et al., 2010; Wang et al., 2018; Xu et al., 2011; Zheng et al., 2015).
In spite of the abovementioned advantages, there are some major drawbacks associated
with employing aerosol concentrators in exposure assessment studies. For instance, inhalation
exposure studies usually span several weeks and require physically and chemically stable PM for
the exposure (Cheng et al., 2016a, 2016b; He et al., 2017; Shimada et al., 2009). However, the
concentration and chemical composition of the aerosol provided by these concentrators during
the inhalation exposure is inevitably variable, as it depends on the variations in the ambient PM.
Furthermore, inhalation exposure studies of animals and/or humans to specific pollution sources
(e.g., vehicular emissions, biomass burning, power plants, airport or ship emissions) require
sophisticated and expensive facilities to house animals in a stationary or mobile laboratory.
These are costly logistical issues, which may limit the applicability of aerosol concentrators in
exposure assessment studies.
A potential solution to these issues is to decouple the PM collection from the inhalation
exposure process, mainly because the former is a lot simpler to conduct in the field alone than in
combination with the latter. Therefore, in this study, we developed a new technique for
generation of stable aerosols that are representative of real-world ambient PM. In this method,
the ambient PM samples are collected on filters, followed by extraction of the filters in Milli-Q
ultrapure water to produce a concentrated PM slurry. The PM slurries are then re-aerosolized
using a commercially available nebulizer to produce a constant flow and concentration of
aerosols that are physically and chemically representative of ambient PM to be used for
inhalation studies. In order to ensure the representativeness of the re-aerosolized PM, their
physical and chemical characteristics are compared with those of ambient PM collected in
99
parallel. In addition, the results from this method are also compared with those from a potentially
superior method, in which ambient PM is captured directly into a suspension of Milli-Q ultrapure
water using the high-volume aerosol-into-liquid collector, developed by Wang et al. (2013),
followed by re-aerosolization of the suspension to produce the concentrated exposure PM.

5.2. Methodology
5.2.1. PM collection
5.2.1.1. High-volume sampler PM collection and filter extraction
In the first approach, ambient PM samples were collected on PTFE membrane filters (20
x 25 cm, 3.0 µm pore size, PALL Life Sciences, USA) using a high-volume PM sampler; the
schematic of the set-up for high-volume sampler is shown in Figure 5.1(a). As shown in the
figure, the high-volume sampler setup is composed of a 180 ° bend, which serves as a PM10 inlet,
and a very high flow rate (i.e., 400 lpm), low pressure drop (i.e., 2 kPa) ultrafine (UFP) slit
nozzle impactor for segregation of accumulation and ultrafine PM modes with a cut point
diameter of 0.18 µm (Misra et al., 2002). Prior to the aqueous extraction of samples, the mass
loading of collected ambient PM on the filters was determined gravimetrically using a high
precision (± 0.001 mg) microbalance (MT5, Mettler Toledo Inc., Columbus, OH), following the
equilibration of filters at proper temperature (22-24 °C) and relative humidity (40-50%) for 24-
48 h. The filters were then divided into half and one half was cut into 16 pieces; the pre-
extraction weights of these small filter pieces were quantified using the abovementioned
microbalance. Subsequently, vortexing was done for 5 minutes to effectively soak each filter
piece in 10 ml of ultrapure Milli-Q water, followed by sonication of the solution for 30 minutes
to extract the PM (i.e., water-soluble fraction of PM) from the filter pieces. Following sonication,
the filter pieces were dried up and weighed to estimate the extracted PM mass in the slurry. The
100
extracted PM mass was quantified as the difference between the total pre-extraction and post-
extraction weight of filters. The efficiency of the extraction was also determined as the ratio of
the actual extracted PM mass to the total collected PM mass on the filters before extraction.
Finally, it is noteworthy that the unextracted half of the filter was used as the reference ambient
PM0.18 sample for chemical analysis, to be compared with the chemical composition of the re-
aerosolized PM from filter extractions (see Section 5.2.2).

5.2.1.2. PM collection using the VACES/aerosol-into-liquid-collector tandem
In the second approach, we collected ambient PM2.5 in slurry samples using the
VACES/aerosol-into-liquid collector tandem (Wang et al., 2013), the schematic of which is
presented in Figure 5.1(b). As shown in the figure, after passing through the PM2.5 inlet, the
ambient air is drawn into a saturation tank that is filled with ultrapure Milli-Q water using a
vacuum pump (Model 2067, GAST Manufacturing, USA) working at a flow rate of 300 lpm.
The incoming ambient particles are mixed with saturated water vapor inside the tank that is
operated at 30 °C. The mixture is then drawn through a condensation unit (each line operating at
100 lpm) that is connected to a chiller operating at -5 °C. This drops the temperature of the
particle-vapor mixture to around 20 °C. The supersaturation created by the temperature drop
causes the water vapor to condense on the particles, growing them to 3-4 µm droplets. The
airstreams from two of the condensation lines are merged, making a total flow rate of 200 lpm,
which passes through the impaction nozzles of the aerosol-into-liquid collector, causing the
liquid droplets to impact and get collected on the lateral cylindrical surfaces of the collector.
These droplets are then drained into the bottom section (i.e., collection stage) of the collector and
form the concentrated slurry sample.
101
The airstream in the third condensation unit (100 lpm) passes through a virtual impactor
with a cut point diameter of 1.5 μm, causing the grown droplets to enter into the minor flow (i.e.,
5 lpm) and become enriched in concentration by approximately 20 times. Subsequently, the
airstream passes through a diffusion dryer (Model 3620, TSI Inc., USA), filled with silica gel, to
remove the excess water from the concentrated grown particles, which causes them to return to
their original size (Kim et al., 2001a, 2001b). The air stream then passes through filters, on
which the particles are collected for chemical analysis. These samples are used as the reference
ambient samples, to which the chemical characteristics of the re-aerosolized PM from slurry
collection are compared (Section 5.2.2). Both filter and slurry PM collections were done at the
University of Southern California’s particle instrumentation unit (PIU), which is an urban site
exposed to a mixture of primary (coming mostly from the I-110 freeway located 150 m upwind
of the site) and secondary PM (Shirmohammadi et al., 2018; Sowlat et al., 2016b; Wang et al.,
2016).

5.2.2. Aerosol generation and inhalation exposure
To provide a stable source of concentrated PM for in vivo exposure studies, the collected
PM slurries (i.e., the PM liquid suspension from both methods) were re-aerosolized using the
commercially available HOPE nebulizers (B&B Medical Technologies, USA). As presented in
Figure 5.1(c), a pump (Model VP0625-V1014-P2-0511, Medo Inc., USA) is generating HEPA-
filtered compressed air introduced into the nebulizer’s suspension to re-aerosolize the PM liquid
suspension and produce the concentrated aerosol. Using a vacuum pump (Model VP0625-
V1014-P2-0511, Medo Inc., USA) at various dilution flow rates, the re-aerosolized PM is then
drawn through a diffusion dryer (Model 3620, TSI Inc., USA) filled with silica gel to remove
102
excess water, and then through Po-210 neutralizers (Model 2U500, NRD Inc., USA) to remove
their electrical charges. The particles are then collected in parallel on 37 mm PTFE (Teflon) and
Quartz (Pall Life Sciences, 2-µm pore, Ann Arbor, MI) filters for chemical analysis and compare
their chemical composition to that of the corresponding ambient samples. In addition, a scanning
mobility particle sizer (SMPS 3936, TSI Inc., USA) connected to a condensation particle counter
(CPC 3022A, TSI Inc., USA) was used to evaluate the physical properties (i.e., number and mass
size distribution) of the re-aerosolized PM at different pressures (i.e., 80-140 inch H2O), and
dilution air flow rates (5, 10, and 15 lpm). Finally, as shown in Figure 5.1(c), part of the air flow
can also enter the animal exposure chambers for in vivo inhalation exposure assessments.
Figure 5.1. Schematic of: a) high-volume sampler with PM 10 inlet, and ultrafine (UFP) impactor; b) the
VACES, coupled with high-volume aerosol-into-liquid collector; and c) aerosol generation setup for filter
collection and inhalation exposure

(a)

103

(b)

(c)
104
5.2.3. Chemical analysis
The quartz filters (i.e., ambient and re-aerosolized) were analyzed for polycyclic aromatic
hydrocarbons (PAHs), elemental (EC), and organic carbon (OC), while the Teflon filters (i.e.,
ambient and re-aerosolized) were chemically analyzed for inorganic ions, and metals and trace
elements. EC and OC content of the samples was analyzed using a Model-4 semi-continuous
OC/EC field analyzer (Sunset Laboratory Inc, USA) following the National Institute for
Occupational Safety and Health (NIOSH) Thermal Optical Transmission (TOT) method (Birch
and Cary, 1996). The PAHs were also determined using Gas Chromatography/Mass
Spectrometry (GC/MS) (Schauer et al., 1999). Ion chromatography (IC) was employed to
quantify the concentrations of inorganic ions. Finally, particle-bound metals and trace elements
were determined using magnetic sector inductively coupled plasma mass spectrometry (IC-PMS)
(Herner et al., 2006). In order to perform multiple chemical analyses on each filter, punches of
known surface areas were taken from each filter, and the pertinent chemical analyses were
performed on individual filter punches. The results were then extrapolated to the whole filter,
based on the ratio of the surface area of the punch to the surface area of the filter.

5.3. Results and discussion
5.3.1. Physical properties of the ambient vs. re-aerosolized PM
As mentioned in the Methodology section, a scanning mobility particle sizer (SMPS
3936, TSI Inc.) was used for analyzing the physical characteristics of the re-aerosolized PM at
different pressures, and dilution flow rates. The PM slurries used for these experiments were
obtained from the filter extractions, with a mass concentration of 40±2 µg/ml. Figure 5.2(a)
indicates the changes in the number size distribution of the re-aerosolized particles as a function
105
of compressed air pressure at a constant dilution air flow rate (i.e., 15 lpm). According to this
figure, applying higher pressure increases the re-aerosolization of the suspended PM, leading to
higher number (and in turn mass concentrations) in the system. For example, increasing the
pressure from 80 to 140 inches of H2O increased the total number and mass concentrations from
nearly 44700 particles/cm
3
and 46.9 µg/m
3
to almost 262000 particles/cm
3
and 114 µg/m
3
,
respectively (Figure 5.3). It should be noted, however, that increasing the pressure to levels
higher than 130-140 inches of H2O produces considerable vapor condensation in the mixing
chamber, which leads to greater rate of particles losses. Additionally, as presented in Figure
5.2(a), the generated aerosols have number mode at a diameter of around 50 nm, without
significant changes in the mode diameter as the pressure changes.
Figure 5.2(b) also illustrates the dependency of the number size distribution of the re-
aerosolized PM on the flow rate at a constant compressed air pressure (i.e., 80 inches of H2O).
As shown in the figure, a decrease in the flow rate leads to higher number and mass
concentrations of the re-aerosolized PM in the system; for instance, the total number and mass
concentrations increased from 44100 particles/cm
3
and 46.9 µg/m
3
to 133000 particles/cm
3
and
141 µg/m
3
, respectively, when the flow rate decreased from 15 to 5 lpm (Figure 5.4). This
indicates that in addition to pressure, the number and mass concentration of the re-aerosolized
PM can be controlled by changing the flow rate. However, according to figure 5.2(b), in spite of
the changes in the maximum and total number concentrations of the re-aerosolized PM at
different flow rates, the number modes were observed at around 50 nm, at all different flow
rates. It should be noted that the above discussion on the aerosol concentrations and size
distributions pertains to the specific nebulizer characteristics as well as the aqueous PM slurry
106
concentrations employed; for example, a more concentrated slurry suspension would result in a
higher aerosol concentration with a larger mean diameter, and vice versa (Hinds, 1999).

Figure 5,2. Number size distributions of re-aerosolized particles as a function of a) nebulizer’s
compressed air pressure; and b) dilution flow rate.

(a)

(b)
107
Figure 5.3. Total a) number; and b) mass concentrations of re-aerosolized particles as a function of
nebulizer's compressed air pressure.

(a)

(b)
108
Figure 5.4. Total a) number; and b) mass concentrations of nebulized particles as a function of dilution
flow rate

(a)

(b)

109
In order to evaluate the representativeness of the re-aerosolized PM in terms of the
physical characteristics, we measured number size distributions of ambient PM at the same
sampling site using the SMPS (connected to CPC) instrument, the results of which are presented
in Figure 5.5. As can be seen in the figure, the measured number size distributions are quite
consistent with those reported previously for the same sampling site in year-long sampling
campaigns (Mousavi et al., 2018b; Sowlat et al., 2016a), showing a number mode diameter in the
ultrafine PM size range (i.e., at 40 nm). In addition, comparison of the ambient number size
distributions with those of the re-aerosolized PM indicates that the size distributions of re-
aerosolized PM particles are quite consistent with those of the ambient PM in terms of the shape
of the distribution and the mode diameters. This further corroborates the fact that the re-
aerosolized PM is quite representative of the ambient PM in terms of the physical characteristics.

Figure 5.5. Typical number size distribution of ambient PM at central Los Angeles obtained during our
field tests.
0
1000
2000
3000
4000
5000
6000
7000
8000
9000
10 100 1000
dN/dLogdp (Number/cm
3
)
diameter (nm)

110
5.3.2. Chemical composition of the ambient vs. re-aerosolized PM
5.3.2.1. Mass balance for bulk chemical components
Figure 5.6(a) illustrates the mass balance for the bulk chemical components of the
ambient PM0.18 (collected by high-volume sampler) versus re-aerosolized PM0.18 from the filter-
extracted slurry. According to the figure, there is general agreement between the chemical mass
balance of ambient and re-aerosolized PM. Water-soluble organic matter (WSOM), sulfate, and
metals and trace elements were the major chemical constituents of both ambient and re-
aerosolized particles. In this study, a conversion factor of 1.8 was used to calculate water-soluble
organic matter (WSOM) from WSOC concentration. This ratio accounts for the contribution of
non-carbon atoms (i.e., hydrogen, and oxygen) to the bulk mass of organic matter (Turpin and
Lim, 2001).
For the water-soluble ions (i.e., chloride, nitrate, ammonium, potassium, and sulfate),
there was a strong association between the mass fractions of the re-aerosolized and ambient
PM0.18 samples. For instance, the mass fraction of chloride was almost the same in the ambient
(i.e., 8.26 µg/mg of PM) and re-aerosolized (i.e., 8.07 µg/mg of PM) samples. Similarly, the
WSOM mass fraction was also perfectly comparable between ambient (i.e., 355 µg/mg of PM)
and re-aerosolized samples (i.e., 372 µg/mg of PM). However, due to the water-insolubility of
some PM components (e.g., EC, and some metals and elements), these insoluble fractions of PM
were not extracted efficiently during the aqueous extraction, which artificially increased the
relative contribution of soluble components, such as WSOM, to total PM0.18 mass in re-
aerosolized (i.e., 53%) versus ambient PM (i.e., 42%) samples. Unlike water-soluble species, the
mass fraction of EC was dropped by 75% upon water extraction, since EC is an insoluble species
and remains mostly on the filter even after applying the extraction protocol (Azeem et al., 2017;
111
Wallén et al., 2010). Similarly, lower mass fractions were observed for several metals and trace
elements in the re-aerosolized samples compared to the ambient samples, which will be
discussed in detail in section 5.3.2.1.
The chemical mass balance for ambient PM2.5 (collected on VACES filters) versus re-
aerosolized PM2.5 (from slurries collected using the VACES/aerosol-into-liquid collector) is
shown in figure 5.6(b). Based on the figure, a virtually excellent agreement was observed
between chemical composition of ambient and re-aerosolized samples. Organic matter (OM),
sulfate, nitrate, ammonium, and metals were the major contributors to PM2.5 composition in both
set of filters, substantiating the similarity of ambient and re-aerosolized samples in terms of
chemical characteristics. In addition, similar to filter extraction protocol, the mass fractions and
relative contributions of water-soluble ions (i.e., chloride, nitrate, ammonium, potassium, and
sulfate) were almost the same in ambient and re-aerosolized filters. For example, the ambient and
re-aerosolized mass fractions of sulfate were 181 µg/mg of PM and 161 µg/mg of PM,
respectively. The actual mass fraction of OM was also quite comparable between ambient (i.e.,
437 µg/mg of PM) and re-aerosolized (i.e., 482 µg/mg of PM) samples. However, unlike
aqueous filter extraction, the water-insoluble constituents of PM were almost completely
reconstructed using the VACES/aerosol-into-liquid collector tandem. For instance, the ambient
mass fraction of EC was 34.3 µg/mg of PM which was similar to that of re-aerosolized one (i.e.,
31.0 µg/mg of PM). This shows that VACES/aerosol-into-liquid collector tandem is able to
retrieve nearly 90% of ambient EC in collected PM slurries, which is a significant advantage
over Particle-Into-Liquid Sampler (PILS), capturing 20% of soot particles (i.e., EC) into pure
water (Wonaschuetz et al., 2018). Similar mass fractions were also observed between metal
112
species (i.e., water-soluble and water-insoluble) for both set of samples, which will be discussed
in further detail in the following section.
Figure 5.6. Chemical composition of re-aerosolized versus ambient a) PM 0.18; and b) PM 2.5
Re-aerosolized PM0.18

(a)

Ambient PM0.18

(a)
Re-aerosolized PM2.5 Ambient PM2.5

(b)
113
5.3.2.2. Metals and trace elements
Due to the importance of metals and organic species in driving the toxicity of ambient
PM (Daher et al., 2011; Saffari et al., 2015; Samara, 2017; Yang et al., 2014), in this and the
following section, we have compared and contrasted the chemical composition of ambient versus
re-aerosolized PM samples (from filter-extracted slurries as well as slurries collected using the
VACES/aerosol-into-liquid collector) in terms of metals/elements and organic (i.e., PAHs)
components in more detail. Figure 5.7(a) indicates the correlation line between the mass
fractions of metals and trace elements in the ambient PM0.18 samples and those in the re-
aerosolized PM from filter-extracted samples. As can be seen in the figure, the slope of the
correlation line is significantly higher than one (i.e., 1.8). This indicates that metals and trace
elements that have lower water solubility (e.g., Fe, V, Cr, Ba and Mn) (Birmili et al., 2006;
Saffari et al., 2015) are not effectively extracted into the slurry samples, leading to significantly
higher mass fractions of these metals and elements in the ambient PM0.18 samples compared to
those in the re-aerosolized PM samples. This point is more clearly illustrated in Figure 5.8 (a) as
bar charts comparing the mass fraction of redox-active metals, such as V, Cr, Mn, and Fe
(Akhtar et al., 2010; Argyropoulos et al., 2016; Charrier and Anastasio, 2012; Decesari et al.,
2017; Gasser et al., 2009; Lovett et al., 2018; Mousavi et al., 2018a, 2018b) in the ambient
PM0.18 samples versus re-aerosolized PM samples from filter extracts. Results of the independent
t-test also confirmed the significant differences in the mass fractions of V (P value=0.062), Cr
(Pvalue=0.039), Mn (Pvalue=0.007), and Fe (Pvalue=0.063) between the re-aerosolized and ambient
PM0.18 samples. Since the aerosol generation system is mainly designed to supply concentrated
ambient PM for the in vivo health exposure studies, the re-aerosolized samples should preserve
the toxicological characteristics of ambient PM. However, as can be seen in the figure, the mass
114
fractions of many of these redox-active species are significantly higher in the ambient PM0.18
samples, compared to the re-aerosolized PM samples. In case of Fe, for instance, the pertinent
mass fraction was 2460 ng/mg of PM in the ambient PM0.18 samples compared to 307 ng/mg of
PM in the re-aerosolized PM samples. This clearly indicates the inability of this filter extraction
protocol in preserving the concentration of many of the ambient metals and elements, including
the redox-active species, in the re-aerosolized samples.
A similar correlation analysis was also performed between the metal and trace element
mass fractions of the ambient PM2.5 samples versus the re-aerosolized PM2.5 from the slurry
samples collected using VACES/aerosol-into-liquid collector (Figure 5.7(b)). As can be seen in
the figure, a high correlation was observed between the mass fractions of the ambient versus re-
aerosolized PM in the VACES-collected filter and slurry samples with an R
2
value of 0.99 and a
slope of nearly 1 (i.e., 1.15). This is shown more clearly in Figure 5.8 (b), which presents the
comparison of the mass fractions of redox-active metals (i.e, V, Cr, Mn, Fe, Ni, Cu, and Ti) for
the ambient versus re-aerosolized PM samples collected using VACES as bar charts. For
instance, the mass fraction of Fe was 17600 ng/mg of PM in the ambient samples, which is quite
consistent with that of re-aerosolized sample (i.e., 15100 ng/mg of PM). In addition, according to
the independent sample t-test, there were no statistically significant differences (i.e.,
Pvalue>>0.05) between the mass fractions of the metals in re-aerosolized versus ambient PM
samples. This further corroborates the advantages of using the VACES/aerosol-into-liquid
collector over the aqueous extraction of filters in preserving the chemical composition (and
particularly metal elements) of ambient PM samples, which leads to generating PM for
inhalation studies that are far more representative of real-world aerosols.

115
Figure 5.7. Correlation analysis between the nebulized and ambient mass ratios of metals and trace
elements in a) PM 0.18; and b) PM 2.5

(a)

(b)
116
Figure 5.8. Comparison of the redox active metals mass ratio in (a) PM 0.18; and (b) PM 2.5

(a)

(b)

117
5.3.2.3. Polycyclic Aromatic Hydrocarbons (PAHs)
Figure 5.9 (a) illustrates the mass fractions of PAHs in the ambient PM0.18 samples
against the re-aerosolized PM from filter extracts. Previous studies have indicated that PAHs
result from incomplete combustion of fossil fuels and they induce detrimental health impacts
(e.g., lung cancer) due to mutagenic and carcinogenic characteristics (Hesterberg et al., 2012;
Kam et al., 2013; Lovett et al., 2017; Sauvain et al., 2003; Taghvaee et al., 2018). However, as
shown in the figure, there is no recovery of PAHs in the re-aerosolized PM samples from filter
extracts, due mainly to the water-insolubility of these organic compounds (Kim et al., 2013;
Miller et al., 1998; Tang et al., 2005; Tarafdar and Sinha, 2017). Therefore, given the toxicity of
these species and the importance of their inclusion in inhalation studies, this can be considered as
another major disadvantage of the filter extraction protocol in reconstructing PM that is well
representative of ambient PM.
Figure 5.9(b) also indicates the comparison of the mass fractions of organic species (i.e.,
PAHs) in the ambient versus re-aerosolized PM2.5 samples collected using the VACES/aerosol-
into-liquid collector tandem technology. As can be seen in the figure, unlike the filter extraction
protocol, this method was able to recover almost all PAHs in the re-aerosolized PM. For
example, similar mass ratios were observed for total PAHs in the ambient (i.e, 105 ng/mg of PM)
versus re-aerosolized samples (i.e., 100 ng/mg of PM). Given the toxic properties of PAHs, the
almost complete recovery of these species is considered as one of the crucial advantages of using
the VACES/aerosol-into-liquid collector tandem technology in collecting ambient PM directly
into slurry samples over the aqueous extraction of filters to be used in toxicological inhalation
exposure studies.

118
Figure 5.9. Comparison of selected PAHs mass ratios in the ambient versus re-aerosolized: a) PM 0.18; and
b) PM 2.5

5.4. Summary and conclusion
The main objective of this study was to develop a new protocol for generating chemically
and physically stable sources of aerosols to be used in inhalation exposure studies that are
representative of real-world ambient PM. Results from the present study indicated that the re-
aerosolized PM are quite representative of ambient PM in terms of the physical characteristics
(i.e., size distributions). The re-aerosolized PM from aqueous filter extracts also showed a rather
consistent mass balance to that of ambient samples for the bulk chemical components, especially
in terms of the water-soluble fractions of ambient PM (i.e., WSOM, and inorganic ions), which
were recovered efficiently in the re-aerosolized PM samples. However, the major drawback of
this protocol was that it produced slurries that were deficient in important redox-active species
such as EC, PAHs, and some toxic metals and elements, due mainly to their low solubility in
water, leading to less-than-ideal recovery of these species in the re-aerosolized PM. On the other
hand, the VACES/aerosol-into-liquid collector tandem technology showed superiority in
119
capturing not only the water-soluble, but also the water-insoluble components of PM directly
into aqueous slurries, leading to a very efficient recovery of all components of ambient PM in the
re-aerosolized PM. This makes the latter protocol an ideal choice to generate particles for
toxicological studies that are well representative of real-world ambient PM.

120
Chapter 6: Conclusions and future research
6.1. Conclusion
While we often do not realize it, the air we breathe is full of toxins and chemicals that
affect our overall health. There are a wide range of health effects of air pollution on humans,
such as decreased lung function, damaged cells in a person’s respiratory system, and the
development of diseases like emphysema, asthma, or even cancer. Therefore, our understanding
of the contributing sources to PM, its chemical components (i.e., PAHs), and its associated
toxicity are of great importance for public health policy makers and health authorities to adopt
appropriate policies for mitigating the adverse health effects of exposure to ambient PM. In this
regard, the presented researches in this dissertation can provide us with beneficial information
regarding the sources of ambient PM2.5 and their associated toxicity and health impacts in highly
crowded urban environments located at Middle East and Europe.
In the first study, the Positive Matrix Factorization (PMF) model was implemented to
derive the major sources of ambient PM2.5 in central Tehran. Our statistical analysis revealed
vehicular emissions, secondary aerosol, industrial emissions, biomass burning, soil, and road
dust (including tire and brake wear particles) as the main PM sources in both sampling sites.
Results indicated that almost half of PM2.5 mass concentration can be attributed to vehicular
emissions; followed by secondary aerosol sources with the 25% contribution to total PM2.5 mass
at both sampling locations. These finding clearly revealed the major role of traffic-related
emissions (both tailpipe and non-tailpipe) on ambient PM2.5 concentrations in central Tehran as
one the most crowded and polluted areas within the Greater Tehran Area.

121
In the next study, we performed the source-specific cancer risk characterization of
ambient PM2.5-bound polycyclic aromatic hydrocarbons (PAHs) in central Tehran. Five factors
were identified as the major sources of airborne PAHs in the area, including petrogenic sources
and petroleum residue, natural gas and biomass burning, industrial emissions, diesel exhaust
emissions, and gasoline exhaust emissions, with approximately similar contributions of around
20% to total PAHs concentration from each factor. Our risk assessment analysis also indicated
that diesel exhaust and industrial emissions were the two sources with major contributions to the
overall cancer risk, contributing respectively to 39% and 27% of the total risk associated with
exposure to ambient PAHs. In addition, the lung cancer risk associated with each specific source
was within the range of 10
-6
–10
−5
, posing cancer risks exceeding the United States
Environmental Protection Agency's (USEPA) guideline safety values (10
−6
). Results from this
study provide an estimate of the cancer risk caused by exposure to ambient PAHs in highly
crowded areas in central Tehran, and can be used as a guide for the adoption of effective air
quality policies in order to reduce the human exposure to these harmful organic species.
The main objective of the third study was chemical characterization and source
apportionment of the oxidative potential of ambient PM2.5 samples collected in an urban
background area in Athens, Greece. Our findings revealed that the intrinsic (per PM mass) and
extrinsic (per m
3
of air volume) oxidative potentials of the collected ambient PM2.5 samples were
significantly higher than those measured in many urban areas around the world. The results of
the MLR analyses indicated that the major pollution sources contributing to the oxidative
potential of ambient PM2.5 were vehicular emissions (characterized by EC) (44%), followed by
secondary organic aerosol (SOA) formation (characterized by WSOC) (16%), and biomass
burning (characterized by levoglucosan) (9%). Results from this study corroborate the impact of
122
traffic and SOA on the oxidative potential of ambient PM2.5 in greater Athens area, and can be
helpful in adopting appropriate public health policies regarding detrimental outcomes of
exposure to PM2.5.
Finally, the last research discussed in chapter 5 aimed to develop a novel system for
providing stable sources of aerosols that are well representative of real-world ambient particulate
matter (PM) in terms of both physical and chemical characteristics, with the ultimate objective of
using them for inhalation exposure studies. Our finding revealed that the water soluble
constituents of ambient PM (e.g., water-soluble organic matter, and water-soluble inorganic ions)
can be recovered effectively by re-aerosolizing the aqueous extracted filters. However, this
protocol was deficient in reconstructing EC, PAHs, and some of the redox-active metals and
trace elements as important insoluble components of ambient PM. On the other hand, employing
the VACES/aerosol-into-liquid collector tandem technology for collecting ambient PM directly
into ultrapure water enabled us to effectively recover all components (i.e., water soluble, and
water insoluble) of ambient PM. Therefore, outcomes of this study corroborated the superiority
of implementing VACES/aerosol-into liquid tandem technology for producing PM solutions;
followed by re-aerosolization procedure to generate stable aerosols that are fully representative
of ambient PM in terms of physical and chemical compositions. Ultimately, this protocol can be
implemented to simulate the inhalation exposure to real world ambient PM in in-vitro
toxicological studies.

6.2. Future research
Although our findings have important implications in understanding the ambient PM
sources and its adverse health impacts in some of the crowded and polluted urban environments
123
around the globe, below are several research ideas which can be followed in the future based on
the discussed research and methodology in this dissertation:

➢ While our filter based PM measurement and source apportionment analysis were
conducted in central Tehran for a period of one year, it might be very imperative to
investigate the temporal trends in the PM mass concentrations as well as contribution of
various pollution sources to ambient PM during larger periods of time to better evaluate
the impact of several ongoing emission reduction scenarios in Tehran on the overall air
quality in the area.

➢ In addition to characterizing and identifying the sources of PAHs as one of the well-
known toxic PM constituents, similar statistical methodology can be implemented to
determine the sources of redox active metals in the area that are highly toxic according to
the literature.

➢ Since there are growing number of evidences underscoring the critical role of particle
size and PM number concentrations from a human health perspective, the implemented
methodology in the second and third chapter (i.e., PMF model) can be utilized to
perform the source apportionment of PM number concentration in highly polluted
metropolitan areas such as Tehran. It should be noted that despite the well-known
association of PM number concentration with cardiovascular diseases, such studies have
not yet been performed in developing countries due to the expensive nature of these
measurements and analysis.
124
➢ In the absence of the comprehensive epidemiological findings for the adverse health
consequences of severe air pollution phenomenon on Tehran residents, the single
pollutant or multi pollutant linear regression model can be used to find out the
association of various inflammation and coagulation blood biomarkers with the PMF-
resolved source specific PM2.5 mass concentration in both retirement home (i.e., among
the seniors) and school dormitory (i.e., among the young populations).

➢ Finally, considering the unprecedented situation with Corona Virus leading to the major
lockdown and quarantine in different parts of the world, the developed aerosol
generation setup and PM slurry collection protocol can be used to investigate the PM
toxicity during the quarantine and normal situation in in-vivo and in-vitro exposure
studies.

125
References
Adamiec, E., Jarosz-Krzemińska, E., Wieszała, R., 2016. Heavy metals from non-exhaust
vehicle emissions in urban and motorway road dusts. Environ. Monit. Assess. 188, 1–11.
doi:10.1007/s10661-016-5377-1
Akbarzadeh, M.A., Khaheshi, I., Sharifi, A., Yousefi, N., Naderian, M., Namazi, M.H., Safi, M.,
Vakili, H., Saadat, H., Alipour Parsaa, S., Nickdoost, N., 2018. The association between
exposure to air pollutants including PM 10 , PM 2 . 5 , ozone , carbon monoxide , sulfur
dioxide , and nitrogen dioxide concentration and the relative risk of developing STEMI : A
case-crossover design. Environ. Res. 161, 299–303. doi:10.1016/j.envres.2017.11.020
Akhtar, U.S., Mcwhinney, R.D., Rastogi, N., Abbatt, J.P.D., Evans, G.J., Scott, J.A., 2010.
Cytotoxic and proinflammatory effects of ambient and source-related particulate matter
(PM) in relation to the production of reactive oxygen species (ROS) and cytokine
adsorption by particles. Inhal. Toxicol. 22, 37–47. doi:10.3109/08958378.2010.518377
Akyüz, M., Çabuk, H., 2008. Particle-associated polycyclic aromatic hydrocarbons in the
atmospheric environment of Zonguldak , Turkey. Sci. Total Environ. 405, 2–10.
doi:10.1016/j.scitotenv.2008.07.026
Al horr, Y., Arif, M., Katafygiotou, M., Mazroei, A., Kaushik, A., Elsarrag, E., 2016. Impact of
indoor environmental quality on occupant well-being and comfort : A review of the
literature. Int. J. Sustain. Built Environ. 5, 1–11. doi:10.1016/j.ijsbe.2016.03.006
126
Al-Dabbous, A.N., Kumar, P., 2015. Source apportionment of airborne nanoparticles in a Middle
Eastern city using positive matrix factorization. Environ. Sci. Process. Impacts 17, 802–12.
doi:10.1039/c5em00027k
Almasi, Z., Salehiniya, H., Amoori, N., Enayatrad, M., 2015. Epidemiology Characteristics and
Trends of Lung Cancer Incidence in Iran. Asian Pacific J. Cancer Prev. 17, 557–562.
doi:10.7314/APJCP.2016.17.2.557
Alsberg, T., Hakansson, S., Strandell, M., Westerholm, R., 1989. Profile Analysis of Urban Air
Pollution. Chemom. Intell. Lab. Syst. 7, 143–152.
Alves, C.A., Barbosa, C., Rocha, S., Calvo, A., Nunes, T., Cerqueira, M., Pio, C., Karanasiou,
A., Querol, X., 2015. Elements and polycyclic aromatic hydrocarbons in exhaust particles
emitted by light-duty vehicles. Environ. Sci. Pollut. Res. 22, 11526–11542.
doi:10.1007/s11356-015-4394-x
Alves, C.A., Vicente, A.M., Custódio, D., Cerqueira, M., Nunes, T., Pio, C., Lucarelli, F.,
Calzolai, G., Nava, S., Diapouli, E., Eleftheriadis, K., Querol, X., Musa, B.A., 2017.
Polycyclic aromatic hydrocarbons and their derivatives (nitro-PAHs, oxygenated PAHs, and
azaarenes) in PM 2.5 from Southern European cities. Sci. Total Environ. 595, 494–504.
doi:10.1016/j.scitotenv.2017.03.256

127
Amato, F., Alastuey, A., Karanasiou, A., Lucarelli, F., Nava, S., Calzolai, G., Severi, M.,
Becagli, S., Gianelle, V.L., Colombi, C., Alves, C., Custódio, D., Nunes, T., Cerqueira, M.,
Pio, C., Eleftheriadis, K., Diapouli, E., Reche, C., Minguillón, M.C., Manousakas, M.,
Maggos, T., Vratolis, S., Harrison, R.M., Querol, X., 2016. AIRUSE-LIFE + : a harmonized
PM speciation and source apportionment in five southern European cities. Atmos. Chem.
Phys. 16, 3289–3309. doi:10.5194/acp-16-3289-2016
Amini, H., Hosseini, V., Schindler, C., Hassankhany, H., Yunesian, M., Henderson, S.B., Künzli,
N., 2017. Spatiotemporal description of BTEX volatile organic compounds in a Middle
Eastern megacity : Tehran Study of Exposure Prediction for Environmental Health Research
( Tehran SEPEHR ) *. Environ. Pollut. 226, 219–229. doi:10.1016/j.envpol.2017.04.027
Amini, H., Taghavi-Shahri, S.M., Henderson, S.B., Hosseini, V., Hassankhany, H., Naderi, M.,
Ahadi, S., Schindler, C., Künzli, N., Yunesian, M., 2016. Annual and seasonal spatial
models for nitrogen oxides in Tehran , Iran. Sci. Rep. 1–11. doi:10.1038/srep32970
Amini, H., Taghavi-Shahri, S.M., Henderson, S.B., Naddafi, K., Nabizadeh, R., Yunesian, M.,
2014. Land use regression models to estimate the annual and seasonal spatial variability of
sulfur dioxide and particulate matter in Tehran, Iran. Sci. Total Environ. 488-489, 343–53.
Andreou, G., Rapsomanikis, S., 2009. Polycyclic aromatic hydrocarbons and their oxygenated
derivatives in the urban atmosphere of Athens. J. Hazard. Mater. 172, 363–373.
doi:10.1016/j.jhazmat.2009.07.023
AQCC, 2013. Air Quality Control Company [WWW Document]. URL
http://air.tehran.ir/portals/0/ReportFiles/AirPollution/New/24.pdf
128
Argyropoulos, G., Besis, A., Voutsa, D., Samara, C., Sowlat, M.H., Hasheminassab, S., Sioutas,
C., 2016. Source apportionment of the redox activity of urban quasi-ultra fi ne particles (PM
0.49) in Thessaloniki following the increased biomass burning due to the economic crisis in
Greece. Sci. Total Environ. 568, 124–136. doi:10.1016/j.scitotenv.2016.05.217
Arhami, M., Hosseini, V., Zare Shahne, M., Bigdeli, M., Lai, A., Schauer, J.J., 2017. Seasonal
trends, chemical speciation and source apportionment of fine PM in Tehran. Atmos.
Environ. 153, 70–82. doi:10.1016/j.atmosenv.2016.12.046
Arhami, M., Minguilion, M.C., Polidori, A., Schauer, J.J., Delfino, R.J., Sioutas, C., 2010.
Organic compound characterization and source apportionment of indoor and outdoor quasi-
ultrafine particulate matter in retirement homes of the Los Angeles Basin. Indoor Air 20,
17–30. doi:10.1111/j.1600-0668.2009.00620.x
Arhami, M., Zare, M., Hosseini, V., Rou, N., Lai, A.M., Schauer, J.J., 2018. Seasonal trends in
the composition and sources of PM2.5 and carbonaceous aerosol in Tehran, Iran. Environ.
Pollut. 239, 69–81. doi:10.1016/j.envpol.2018.03.111
Armstrong, B., Hutchinson, E., Unwin, J., Fletcher, T., 2004. Lung Cancer Risk after Exposure
to Polycyclic Aromatic Hydrocarbons : A Review and Meta-Analysis. Environ. Health
Perspect. 112, 970–978. doi:10.1289/ehp.6895
Askariyeh, M.H., Arhami, M., 2013. Projecting emission reductions from prospective mobile
sources policies by road link-based modelling. Int. J. Environ. Pollut. 53.
129
Atash, F., 2007. The deterioration of urban environments in developing countries: Mitigating the
air pollution crisis in Tehran, Iran. Cities 24, 399–409. doi:10.1016/j.cities.2007.04.001
Athanasopoulou, E., Speyer, O., Brunner, D., Vogel, H., Vogel, B., Mihalopoulos, N.,
Gerasopoulos, E., 2017. Changes in domestic heating fuel use in Greece : effects on
atmospheric chemistry and radiation. Atmos. Chem. Phys. 17, 10597–10618.
Athanasopoulou, E., Tombrou, M., Russell, A.G., Karanasiou, A., Eleftheriadis, K., Dandou, A.,
2010. Implementation of road and soil dust emission parameterizations in the aerosol model
CAMx: Applications over the greater Athens urban area affected by natural sources. J.
Geophys. Res. 115, 1–21. doi:10.1029/2009JD013207
Aydin, Y.M., Kara, M., Dumanoglu, Y., Odabasi, M., Elbir, T., 2014. Source apportionment of
polycyclic aromatic hydrocarbons ( PAHs ) and polychlorinated biphenyls ( PCBs ) in
ambient air of an industrial region in Turkey. Atmos. Environ. 97, 271–285.
doi:10.1016/j.atmosenv.2014.08.032
Ayres, J.G., Borm, P., Cassee, F.R., Castranova, V., Donaldson, K., Ghio, A., Harrison, R.M.,
Hider, R., Kelly, F., Kooter, I.M., Marano, F., Maynard, R.L., Mudway, I., Nel, A., Sioutas,
C., Smith, S., Baeza-squiban, A., Cho, A., Duggan, S., Froines, J., 2008. Evaluating the
Toxicity of Airborne Particulate Matter and Nanoparticles by Measuring Oxidative Stress
Potential — A Workshop Report and Consensus Statement. Inhal. Toxicol. 20, 75–99.
doi:10.1080/08958370701665517

130
Bae, M., Schauer, J.J., Lee, T., Jeong, J., Kim, Y., Ro, C., Song, S., Shon, Z., 2017. Relationship
between reactive oxygen species and water-soluble organic compounds : Time-resolved
benzene carboxylic acids measurement in the coastal area during the KORUS-AQ
campaign. Environ. Pollut. 231, 1–12. doi:10.1016/j.envpol.2017.07.100
Baek, S.O., Goldstone, M.E., Kirk, P.W.W., Lester, J.N., Perry, R., 1991. Phase distribution and
particle size dependency of polycyclic aromatic hydrocarbons in the urban atmosphere.
Chemosphere 22, 503–520.
Bahari, R.A., Abbaspour, R.A., Pahlavani, P., 2014. Prediction of PM2.5 Concentrations Using
Temperature Inversion Effects Based on AN Artificial Neural Network, in: The 1st ISPRS
International Conference on Geospatial Information Research. doi:10.5194/isprsarchives-
XL-2-W3-73-2014
Bai, Z., Hu, Y., Yu, H., Wu, N., You, Y., 2009. Quantitative Health Risk Assessment of
Inhalation Exposure to Polycyclic Aromatic Hydrocarbons on Citizens in Tianjin , China.
Bull. Environ. Contam. Toxicol. 83, 151–154. doi:10.1007/s00128-009-9686-8
Bandowe, B.A.M., Meusel, H., Huang, R., Ho, K., Cao, J., Hoffmann, T., Wilcke, W., 2014.
PM2.5-bound oxygenated PAHs, nitro-PAHs and parent-PAHs from the atmosphere of a
Chinese megacity: Seasonal variation, sources and cancer risk assessment. Sci. Total
Environ. 473-474, 77–87. doi:10.1016/j.scitotenv.2013.11.108
Beddows, D.C.S., Harrison, R.M., Green, D.C., Fuller, G.W., 2015. Receptor modelling of both
particle composition and size distribution from a background site in London, UK. Atmos.
Chem. Phys. 15, 10107–10125. doi:10.5194/acp-15-10107-2015
131
Beelen, R., Hoek, G., van den Brandt, P. a., Goldbohm, R.A., Fischer, P., Schouten, L.J., Jerrett,
M., Hughes, E., Armstrong, B., Brunekreef, B., 2008. Long-term effects of traffic-related
air pollution on mortality in a Dutch cohort (NLCS-AIR study). Environ. Health Perspect.
116, 196–202. doi:10.1289/ehp.10767
Bian, Q., Alharbi, B., Collett, J., Kreidenweis, S., Pasha, M.J., 2016. Measurements and source
apportionment of particle associated polycyclic aromatic hydrocarbons in ambient air in
Riyadh, Saudi Arabia. Atmos. Environ. 137, 186–198. doi:10.1016/j.atmosenv.2016.04.025
Birch, M.E., Cary, R.A., 1996. Elemental Carbon-Based Method for Monitoring Occupational
Exposures to Particulate Diesel Exhaust. Aerosol Sci. Technol. 25.
doi:10.1080/02786829608965393
Black, H., 2005. Getting the lead out of electronics. Environ. Health Perspect. 113, A682–5.
doi:ehp0113-a00682 [pii]
Boström, C.-E., Gerde, P., Hanberg, A., Jernström, B., Johansson, C., Kyrklund, T., Rannug, A.,
Törnqvist, M., Victorin, K., Westerholm, R., 2002. Cancer Risk Assessment , Indicators ,
and Guidelines for Polycyclic Aromatic Hydrocarbons in the Ambient Air. Environ. Health
Perspect. 110, 451–488.
Brown, S.G., Eberly, S., Paatero, P., Norris, G.A., 2015. Methods for estimating uncertainty in
PMF solutions : Examples with ambient air and water quality data and guidance on
reporting PMF results. Sci. Total Environ. 518-519, 626–635.
doi:10.1016/j.scitotenv.2015.01.022
132
Callén, M.S., Iturmendi, A., L opez, J. e M., 2014a. Source apportionment of atmospheric PM2 .
5-bound polycyclic aromatic hydrocarbons by a PMF receptor model . Assessment of
potential risk for human health pez. Environ. Pollut. 195. doi:10.1016/j.envpol.2014.08.025
Callén, M.S., Iturmendi, A., López, J.M., Mastral, A.M., 2014b. Source apportionment of the
carcinogenic potential of polycyclic aromatic hydrocarbons ( PAH ) associated to airborne
PM10 by a PMF model. Environ. Sci. Pollut. Res. 21, 2064–2076. doi:10.1007/s11356-013-
2116-9
Castaneda, D.M.A., Teixeira, E.C., Schneider, I.L., Lara, S.R. on, Silva, L.F.O., 2017. Exposure
to polycyclic aromatic hydrocarbons in atmospheric PM 1.0 of urban environments:
Carcinogenic and mutagenic respiratory health risk by age groups. Environ. Pollut. 224,
158–170. doi:10.1016/j.envpol.2017.01.075
Cavalli, F., Viana, M., Yttri, K.E., Genberg, J., Putaud, J.-P., 2010. Toward a standardised
thermal-optical protocol for measuring atmospheric organic and elemental carbon : the
EUSAAR protocol. Atmos. Meas. Tech. 3, 79–89.
Charrier, J.G., Anastasio, C., 2012. On dithiothreitol (DTT) as a measure of oxidative potential
for ambient particles : evidence for the importance of soluble transition metals. Atmos.
Chem. Physics12 9321–9333. doi:10.5194/acp-12-9321-2012
Chen, D., Zhao, N., Lang, J., Zhou, Y., Wang, X., Li, Y., 2018. Science of the Total
Environment Contribution of ship emissions to the concentration of PM 2 . 5 : A
comprehensive study using AIS data and WRF / Chem model in Bohai Rim. Sci. Total
Environ. 610-611, 1476–1486. doi:10.1016/j.scitotenv.2017.07.255
133
Chen, L.C., Lippmann, M., 2009. Effects of metals within ambient air particulate matter (PM) on
human health, Inhalation Toxicology. doi:10.1080/08958370802105405
Chirizzi, D., Cesari, D., Rachele, M., Dinoi, A., Giotta, L., Donateo, A., Contini, D., 2017. In
fluence of Saharan dust outbreaks and carbon content on oxidative potential of water
soluble fractions of PM2.5 and PM 10. Atmos. Environ. 163, 1–8.
doi:10.1016/j.atmosenv.2017.05.021
Cho, A.K., Sioutas, C., Miguel, A.H., Kumagai, Y., Schmitz, D.A., Singh, M., Eiguren-
fernandez, A., Froines, J.R., 2005. Redox activity of airborne particulate matter at different
sites in the Los Angeles Basin. Environ. re 99, 40–47. doi:10.1016/j.envres.2005.01.003
Choi, H., Spengler, J., 2014. Source attribution of personal exposure to airborne polycyclic
aromatic hydrocarbon mixture using concurrent personal , indoor , and outdoor
measurements. Environ. Int. 63, 173–181. doi:10.1016/j.envint.2013.11.007
Chueinta, W., Hopke, P.K., Paatero, P., 2000. Investigation of sources of atmospheric aerosol at
urban and suburban residential areas in Thailand by positive matrix factorization. Atmos.
Environ. 34, 3319–3329.
Cohen, D.D., Crawford, J., Stelcer, E., Bac, V.T., 2010. Characterisation and source
apportionment of fine particulate sources at Hanoi from 2001 to 2008. Atmos. Environ. 44,
320–328. doi:10.1016/j.atmosenv.2009.10.037
134
Collins, J.F., Brown, J.P., Alexeeff, G. V, Salmon, A.G., 1998. Potency Equivalency Factors for
Some Polycyclic Aromatic Hydrocarbons and Polycyclic Aromatic Hydrocarbon
Derivatives 1. Regul. Toxicol. Pharmacol. 28, 45–54.
Crilley, L.R., Lucarelli, F., Bloss, W.J., Harrison, R.M., Beddows, D.C., Calzolai, G., Nava, S.,
Valli, G., Bernardoni, V., Vecchi, R., 2017. Source apportionment of fine and coarse
particles at a roadside and urban background site in London during the 2012 summer
ClearfLo campaign. Environ. Pollut. 220, 766–778. doi:10.1016/j.envpol.2016.06.002
Dachs, J., Eisenreich, S.J., 2000. Adsorption onto Aerosol Soot Carbon Dominates Gas-Particle
Partitioning of Polycyclic Aromatic Hydrocarbons. Environ. Sci. Technol. 34, 3690–3697.
Dachs, J., Glenn, T.R., Gigliotti, C.L., Brunciak, P., Totten, L.A., Nelson, E.D., Franz, T.P.,
Eisenreich, S.J., 2002. Processes driving the short-term variability of polycyclic aromatic
hydrocarbons in the Baltimore and northern Chesapeake Bay atmosphere , USA. Atmos.
Environ. 36, 2281–2295.
Dall’Osto, M., Booth, M.J., Smith, W., Fisher, R., Harrison, R.M., 2008. A study of the size
distributions and the chemical characterization of airborne particles in the vicinity of a large
integrated steelworks. Aerosol Sci. Technol. 42, 981–991.
doi:10.1080/02786820802339587
Davis, D. a., Bortolato, M., Godar, S.C., Sander, T.K., Iwata, N., Pakbin, P., Shih, J.C., Berhane,
K., McConnell, R., Sioutas, C., Finch, C.E., Morgan, T.E., 2013. Prenatal Exposure to
Urban Air Nanoparticles in Mice Causes Altered Neuronal Differentiation and Depression-
Like Responses 8, 1–7. doi:10.1371/journal.pone.0064128
135
Delfino, R.J., Staimer, N., Tjoa, T., Arhami, M., Polidori, A., Gillen, D.L., George, S.C., Shafer,
M.M., Schauerf, J.J., Sioutas, C., 2010. Associations of Primary and Secondary Organic
Aerosols With Airway and Systemic Inflammation in an Elderly Panel Cohort.
Epidemiology 21. doi:10.1097/EDE.0b013e3181f20e6c.Associations
Delfino, R.J., Staimer, N., Tjoa, T., Gillen, D.L., Schauer, J.J., Shafer, M.M., 2013. Airway
inflammation and oxidative potential of air pollutant particles in a pediatric asthma panel. J.
Expo. Sci. Environ. Epidemiol. 23, 466–473. doi:10.1038/jes.2013.25
Delgado-saborit, J.M., Stark, C., Harrison, R.M., 2011. Carcinogenic potential , levels and
sources of polycyclic aromatic hydrocarbon mixtures in indoor and outdoor environments
and their implications for air quality standards. Environ. Int. 37, 383–392.
doi:10.1016/j.envint.2010.10.011
Diapouli, E., Kalogridis, A.-C., Markantonaki, C., Vratolis, S., Fetfatzis, P., Colombi, C.,
Eleftheriadis, K., 2017a. Annual Variability of Black Carbon Concentrations Originating
from Biomass and Fossil Fuel Combustion for the Suburban Aerosol in Athens, Greece.
Atmosphere (Basel). 8. doi:10.3390/atmos8120234
Diapouli, E., Manousakas, M., Vratolis, S., Vasilatou, V., Maggos, T., Saraga, D., Grigoratos, T.,
Argyropoulos, G., Voutsa, D., Samara, C., Eleftheriadis, K., 2017b. Evolution of air
pollution source contributions over one decade, derived by PM10and PM2.5source
apportionment in two metropolitan urban areas in Greece. Atmos. Environ. 164, 416–430.
doi:10.1016/j.atmosenv.2017.06.016
136
Didkowska, J., Wojciechowska, U., Ma, M., Lobaszewski, J., 2016. Lung cancer epidemiology:
contemporary and future challenges worldwide. Ann. Transl. Med. 4, 1–11.
doi:10.21037/atm.2016.03.11
Dockery, D.W., Stone, P.H., 2007. Cardiovascular Risks from Fine Particulate Air Pollution. N.
Engl. J. Med. 356, 511–513.
Economopoulou, A.A., Economopoulos, A.P., 2002. Air Pollution in Athens Basin and Health
Risk Assessment. Environ. Monit. Assess. 80, 277–299.
Edwards, J., 1997. Coating and Surface Treatment Systems for Metals: a comprehensive guide to
selection. ASM International, Materials Park, Ohio,.
Eleftheriadis, K., Ochsenkuhn, K.M., Lymperopoulou, T., Karanasiou, A., Razos, P.,
Ochsenkuhn-petropoulou, M., 2014. Influence of local and regional sources on the observed
spatial and temporal variability of size resolved atmospheric aerosol mass concentrations
and water-soluble species in the Athens metropolitan area. Atmos. Environ. 97, 252–261.
doi:10.1016/j.atmosenv.2014.08.013
Esen, F., Tasdemir, Y., Vardar, N., 2008. Atmospheric concentrations of PAHs , their possible
sources and gas-to-particle partitioning at a residential site of Bursa , Turkey. Atmos. Res.
88, 243–255. doi:10.1016/j.atmosres.2007.11.022
Everitt, B.S., Skrondal, A., 2010. The Cambridge Dictionary of Statistics, 4th ed.

137
Fang, T., Verma, V., Bates, J.T., Abrams, J., Klein, M., Strickland, M.J., Sarnat, S.E., Chang,
H.H., Mulholland, J.A., Tolbert, P.E., Russell, A.G., Weber, R.J., 2016. Oxidative potential
of ambient water-soluble PM 2.5 in the southeastern United States : contrasts in sources and
health associations between ascorbic acid (AA) and dithiothreitol (DTT) assays. Atmos.
Chem. Phys. 16, 3865–3879. doi:10.5194/acp-16-3865-2016
Faridi, S., Shamsipour, M., Krzyzanowski, M., Kunzli, N., Amini, H., Azimi, F., Malkawi, M.,
Momeniha, F., Gholampour, A., Hassanvand, M.S., Naddafi, K., 2018. Long-term trends
and health impact of PM2.5 and O3 in Tehran, Iran,2006-2015. Environ. Int. 114, 37–49.
doi:10.1016/j.envint.2018.02.026
Farrance, I., Frenkel, R., 2012. Uncertainty of Measurement : A Review of the Rules for
Calculating Uncertainty Components through Functional Relationships. Clin. Biochem.
Rev. 33, 49–75.
Ferlay, J., Soerjomataram, I., Dikshit, R., Eser, S., Mathers, C., Rebelo, M., Parkin, D.M.,
Forman, D., Bray, F., 2015. Cancer incidence and mortality worldwide : Sources, methods
and major patterns in GLOBOCAN 2012. Int. J. cancer 136, 359–386.
doi:10.1002/ijc.29210
Fine, P.M., Cass, G.R., Simoneit, B.R.T., 2004. Chemical Characterization of Fine Particle
Emissions from the Wood Stove Combustion of Prevalent United States Tree Species.
Environ. Eng. Sci. 21, 705–721. doi:10.1089/ees.2004.21.705
138
Galarneau, E., 2008. Source specificity and atmospheric processing of airborne PAHs :
Implications for source apportionment. Atmos. Environ. 42, 8139–8149.
doi:10.1016/j.atmosenv.2008.07.025
Gasser, M., Riediker, M., Mueller, L., Perrenoud, A., Blank, F., Gehr, P., Rothen-rutishauser, B.,
2009. Toxic effects of brake wear particles on epithelial lung cells in vitro. Part. Fibre
Toxicol. 13, 1–13. doi:10.1186/1743-8977-6-30
Gauderman, W.J., Urman, R., Avol, E., Berhane, K., McConnell, R., Rappaport, E., Chang, R.,
Lurmann, F., Gilliland, F., 2015. Association of Improved Air Quality with Lung
Development in Children. N. Engl. J. Med. 372, 905–913. doi:10.1056/NEJMoa1414123
Gemmer, M., Bo, X., 2013. Air Quality Legislation and Standards in the European Union :
Background , Status and Public Participation. Adv. Clim. Chang. Res. 4, 50–59.
doi:10.3724/SP.J.1248.2013.050
Givehchi, R., Arhami, M., Tajrishy, M., 2013. Contribution of the Middle Eastern dust source
areas to PM10 levels in urban receptors: Case study of Tehran, Iran. Atmos. Environ. 75,
287–295. doi:10.1016/j.atmosenv.2013.04.039
Gratsea, M., Liakakou, E., Mihalopoulos, N., Adamopoulos, A., Tsilibari, E., Gerasopoulos, E.,
2017. The combined effect of reduced fossil fuel consumption and increasing biomass
combustion on Athens’ air quality, as inferred from long term CO measurements. Sci. Total
Environ. 592, 115–123. doi:10.1016/j.scitotenv.2017.03.045
139
Gugamsetty, B., Wei, H., Liu, C.N., Awasthi, A., Tsai, C.J., Roam, G.D., Wu, Y.C., Chen, C.F.,
2012. Source Characterization and Apportionment of PM10, PM2.5 and PM0.1 by Using
Positive Matrix Factorization. Aerosol Air Qual. Res. 12, 476–491.
doi:10.4209/aaqr.2012.04.0084
Günter, J., 1999. Copper: Its Trade, Manufacture, Use, and Environmental Status, ASM
International.
Guo, H., Harsha, S., Kumar, S., Hu, J., Ying, Q., Gao, A., Zhang, H., 2017. Source
apportionment of PM 2 . 5 in North India using source-oriented air quality models *.
Environ. Pollut. 231, 426–436. doi:10.1016/j.envpol.2017.08.016
Guo, H., Lee, S.C., Ho, K.F., Wang, X.M., Zou, S.C., 2003. Particle-associated polycyclic
aromatic hydrocarbons in urban air of Hong Kong. Atmos. Environ. 37, 5307–5317.
doi:10.1016/j.atmosenv.2003.09.011
Gupta, S., Kumar, K., Srivastava, A., Srivastava, A., Jain, V.K., 2011. Size distribution and
source apportionment of polycyclic aromatic hydrocarbons ( PAHs ) in aerosol particle
samples from the atmospheric environment of Delhi , India. Sci. Total Environ. 409, 4674–
4680. doi:10.1016/j.scitotenv.2011.08.008
Habibi, R., Alesheikh, A.A., Mohammadinia, A., Sharif, M., 2017. An Assessment of Spatial
Pattern Characterization of Air Pollution : A Case Study of CO and PM 2 . 5 in Tehran ,
Iran. Int. J. Geo-information. doi:10.3390/ijgi6090270
140
Halek, F., Kavouci, A., Montehaie, H., 2007. Role of motor-vehicles and trend of air borne
particulate in the Great Tehran area , Iran Role of motor-vehicles and trend of air borne
particulate in the Great Tehran area , Iran. Int. J. Environ. Health Res. 3123.
doi:10.1080/09603120410001725649
Halek, F., Kianpour-rad, M., Kavousi, A., 2010. Characterization and source apportionment of
polycyclic aromatic hydrocarbons in the ambient air ( Tehran , Iran ). Environ. Chem. Lett.
8, 39–44. doi:10.1007/s10311-008-0188-4
Harrison, R.M., Beddows, D.C.S., Dall’Osto, M., 2011. PMF analysis of wide-range particle size
spectra collected on a major highway. Environ. Sci. Technol. 45, 5522–5528.
doi:10.1021/es2006622
Harrison, R.M., Jones, A.M., Gietl, J., Yin, J., Green, D.C., 2012. Estimation of the contributions
of brake dust, tire wear, and resuspension to nonexhaust traffic particles derived from
atmospheric measurements. Environ. Sci. Technol. 46, 6523–6529. doi:10.1021/es300894r
Harrison, R.M., Smith, D.J.T., Luhana, L., 1996. Source Apportionment of Atmospheric
Polycyclic Aromatic Hydrocarbons Collected from an Urban Location in. Environ. Sci.
Technol. 30, 825–832. doi:10.1021/es950252d
Harvey, R.G., 1997. Polycyclic Aromatic Hydrocarbons. John Wiley & Sons, New York.

141
Hasheminassab, S., Daher, N., Ostro, B.D., Sioutas, C., 2014a. Long-term source apportionment
of ambient fine particulate matter (PM2.5) in the Los Angeles Basin: A focus on emissions
reduction from vehicular sources. Environ. Pollut. 193, 54–64.
doi:10.1016/j.envpol.2014.06.012
Hasheminassab, S., Daher, N., Saffari, A., Wang, D., Ostro, B.D., Sioutas, C., 2014b. Spatial and
temporal variability of sources of ambient fine particulate matter (PM2.5) in California.
Atmos. Chem. Phys. 14, 12085–12097. doi:10.5194/acp-14-12085-2014
Hassanipour, S., Mokhtari, A.M., Fathalipour, M., Salehiniya, H., 2017. The incidence of lung
cancer in Iran: a systematic review and meta-analysis. Hassanipour S., Mokhtari A. M.,
Fathalipour M., Salehiniya H. 4, 1–7.
Hassanvand, M.S., Naddafi, K., Faridi, S., Arhami, M., Nabizadeh, R., Sowlat, M.H., Pourpak,
Z., Rastkari, N., Momeniha, F., Kashani, H., Gholampour, A., Nazmara, S.,
Alimohammadi, M., Goudarzi, G., Yunesian, M., 2014. Indoor/outdoor relationships of
PM10, PM2.5, and PM1 mass concentrations and their water-soluble ions in a retirement
home and a school dormitory. Atmos. Environ. 82, 375–382.
doi:10.1016/j.atmosenv.2013.10.048
Hassanvand, M.S., Naddafi, K., Faridi, S., Nabizadeh, R., Sowlat, M.H., Momeniha, F.,
Gholampour, A., Arhami, M., Kashani, H., Zare, A., Niazi, S., Rastkari, N., Nazmara, S.,
Ghani, M., Yunesian, M., 2015. Characterization of PAHs and metals in indoor/outdoor
PM10/PM2.5/PM1 in a retirement home and a school dormitory. Sci. Total Environ. 527-
528, 100–110. doi:10.1016/j.scitotenv.2015.05.001
142
Heo, J.-B., Hopke, P.K., Yi, S.-M., 2009. Source apportionment of PM2.5 in Seoul, Korea.
Atmos. Chem. Phys. 9, 4957–4971. doi:10.5194/acp-9-4957-2009
Herner, J.D., Green, P.G., Kleeman, M.J., 2006. Measuring the Trace Elemental Composition of
Size-Resolved Airborne Particles. Environ. Sci. Technol. 40, 1925–1933.
doi:10.1021/es052315q
Hjortenkrans, D.S.T., Bergbäck, B.G., Häggerud, A. V., 2007. Metal emissions from brake
linings and tires: Case studies of Stockholm, Sweden 1995/1998 and 2005. Environ. Sci.
Technol. 41, 5224–5230. doi:10.1021/es070198o
Hopke, P.K., 2003. Recent developments in receptor modeling. J. Chemom. 17, 255–265.
doi:10.1002/cem.796
Hoseini, M., Yunesian, M., Nabizadeh, R., Yaghmaeian, K., Ahmadkhaniha, R., Rastkari, N.,
Parmy, S., Faridi, S., Rafiee, A., Naddafi, K., 2016. Characterization and risk assessment of
polycyclic aromatic hydrocarbons ( PAHs ) in urban atmospheric Particulate of Tehran ,
Iran. Environ. Sci. Pollut. Res. 23, 1820–1832. doi:10.1007/s11356-015-5355-0

Hosseini, M., Naghan, P.A., Karimi, S., SeyedAlinaghi, S., Bahadori, M., Khodadad, K.,
Mohammadi, F., Keynama, K., Masjedi, M.R., 2009. Environmental risk factors for lung
cancer in Iran : a case - control study. Int. J. Epidemiol. 38, 989–996.
doi:10.1093/ije/dyp218
143
Hosseini, V., Shahbazi, H., 2016. Urban Air Pollution in Iran. Iran. Stud.
doi:10.1080/00210862.2016.1241587
Hosseinpoor, A.R., Forouzanfar, M.H., Yunesian, M., Asghari, F., Naieni, K.H., Farhood, D.,
2005. Air pollution and hospitalization due to angina pectoris in Tehran, Iran: A time-series
study. Environ. Res. 99, 126–131. doi:10.1016/j.envres.2004.12.004
Hu, Y., Bao, L., Huang, C., Li, S., Liu, P., Zeng, E.Y., 2018. Assessment of airborne polycyclic
aromatic hydrocarbons in a megacity of South China : Spatiotemporal variability , indoor-
outdoor interplay and potential human health risk *. Environ. Pollut. 238, 431–439.
doi:10.1016/j.envpol.2018.03.040
Ito, K., Xue, N., Thurston, G., 2004. Spatial variation of PM 2.5 chemical species and source-
apportioned mass concentrations in New York City. Atmos. Environ. 38, 5269–5282.
doi:10.1016/j.atmosenv.2004.02.063
Jang, E., Alam, M.S., Harrison, R.M., 2013. Source apportionment of polycyclic aromatic
hydrocarbons in urban air using positive matrix factorization and spatial distribution
analysis. Atmos. Environ. 79, 271–285. doi:10.1016/j.atmosenv.2013.06.056
Jeong, C.-H., Wang, J.M., Evans, G.J., 2016. Source Apportionment of Urban Particulate Matter
using Hourly Resolved Trace Metals, Organics, and Inorganic Aerosol Components. Atmos.
Chem. Phys. Discuss. 1–32. doi:10.5194/acp-2016-189

144
Jiang, H., Jang, M., Sabo-attwood, T., Robinson, S.E., 2016. Oxidative potential of secondary
organic aerosols produced from photooxidation of different hydrocarbons using outdoor
chamber under ambient sunlight. Atmos. Environ. 131, 382–389.
doi:10.1016/j.atmosenv.2016.02.016
Kalogridis, A., Vratolis, S., Liakakou, E., Gerasopoulos, E., Mihalopoulos, N., Eleftheriadis, K.,
2018. Assessment of wood burning versus fossil fuel contribution to wintertime black
carbon and carbon monoxide concentrations in Athens,Greece. Atmos. Chem. Phys. 18,
10219–10236.
Kam, W., Delfino, R.J., Schauer, J.J., Sioutas, C., 2013. A comparative assessment of PM2.5
exposures in lightrail, subway, freeway, and surface street environments in Los Angeles and
estimated lung cancer risk. Environ. Sci. Process. Impacts 15, 234–243.
doi:10.1039/c2em30495c
Kamali, N., Shahne, M.Z., Arhami, M., 2015. Implementing Spectral Decomposition of Time
Series Data in Artificial Neural Networks to Predict Air Pollutant Concentrations. Environ.
Eng. Sci. 32, 379–388. doi:10.1089/ees.2014.0350
Karakatsani, A., Andreadaki, S., Katsouyanni, K., Dimitroulis, I., Trichopoulos, D., Benetou, V.,
Trichopoulou, A., 2003. Air pollution in relation to manifestations of chronic pulmonary
disease : A nested case – control study in Athens , Greece. Eur. J. Epidemiol. 18, 45–53.
Kasumba, J., Hopke, P.K., Chalupa, D.C., Utell, M.J., 2009. Comparison of sources of
submicron particle number concentrations measured at two sites in Rochester, NY. Sci.
Total Environ. 407, 5071—5084. doi:10.1016/j.scitotenv.2009.05.040
145
Kaya, E., Dumanoglu, Y., Kara, M., Altiok, H., Bayram, A., Elbir, T., Odabasi, M., 2012. Spatial
and temporal variation and air – soil exchange of atmospheric PAHs and PCBs in an
industrial region. Atmos. Pollut. Res. 3, 435–449. doi:10.5094/APR.2012.050
Keramatinia, A., Hassanipour, S., Nazarzadeh, M., Wurtz, M., Monfared, A.B., Khayyamzadeh,
M., Bidel, Z., Mhrvar, N., Mosavi-Jarrahi, A., 2016. Correlation Between Nitrogen Dioxide
as an Air Pollution Indicator and Breast Cancer: a Systematic Review and Meta- Analysis.
Asian Pacific J. Cancer Prev. 17, 419–424.
Khairy, M.A., Lohmann, R., 2013. Source apportionment and risk assessment of polycyclic
aromatic hydrocarbons in the atmospheric environment of Alexandria , Egypt.
Chemosphere 91, 895–903. doi:10.1016/j.chemosphere.2013.02.018
Khodeir, M., Shamy, M., Alghamdi, M., Zhong, M., Sun, H., Costa, M., Chen, L.-C.,
Maciejczyk, P., 2012. Source apportionment and elemental composition of PM2.5 and
PM10 in Jeddah City, Saudi Arabia. Atmos. Pollut. Res. 3, 331–340.
doi:10.5094/APR.2012.037
Khorasani, S., Rezaei, S., Rashidian, H., Daroudi, R., 2015. Years of Potential Life Lost and
Productivity Costs Due to Premature Cancer-Related Mortality in Iran. Asian Pacific J.
Cancer Prev. 16, 1845–1850.
Kim, S., Jaques, P.A., Chang, M., Barone, T., Xiong, C., Friedlander, S.K., Sioutas, C., 2001a.
Versatile aerosol concentration enrichment system ( VACES ) for simultaneous in vivo and
in vitro evaluation of toxic effects of ultrafine ,fine and coarse ambient particles Part II :
Field evaluation. J. Aerosol Sci. 32, 1299–1314.
146
Kim, S., Jaques, P.A., Chang, M., Froines, J.R., Sioutas, C., 2001b. Versatile aerosol
concentration enrichment system (VACES) for simultaneous in vivo and in vitro evaluation
of toxic effects of ultrafine , fine, and coarse ambient particles Part I: Development and
laboratory characterization. J. Aerosol Sci. 32, 1281–1297.
Kleinman, M.T., Hamade, A., Meacher, D., Oldham, M., Sioutas, C., Chakrabarti, B., Stram, D.,
Froines, J.R., Cho, A.K., 2012. Inhalation of Concentrated Ambient Particulate Matter near
a Heavily Trafficked Road Stimulates Antigen-Induced Airway Responses in Mice. J. Air
Waste Manage. Assoc. 55, 1277–1288. doi:10.1080/10473289.2005.10464727
Kotchenruther, R. a., 2016. Source apportionment of PM2.5 at multiple Northwest U.S. sites:
Assessing regional winter wood smoke impacts from residential wood combustion. Atmos.
Environ. 142, 210–219. doi:10.1016/j.atmosenv.2016.07.048
Krewski, D., Jerrett, M., Burnett, R.T., Ma, R., Hughes, E., Shi, Y., Turner, M.C., Pope III, C. a,
Thurston, G., Calle, E.E., Thun, M.J., Beckerman, B., DeLuca, P., Finkelstein, N., Ito, K.,
Moore, D.K., Newbold, K.B., Ramsay, T., Ross, Z., Shin, H., Tempalski, B., 2009.
Extended follow-up and spatial analysis of the American Cancer Society study linking
particulate air pollution and mortality. Boston.
Kreyszig, E., 2011. Advanced Engineering Mathematics. John Wiley & Sons.
Krishnamurthy, N., Vallinayagam, P., Madhavan, D., 2014. ENGINEERING CHEMISTRY.
PHI Learning Pvt. Ltd.

147
Kuang, B.Y., Lin, P., Huang, X.H.H., Yu, J.Z., 2015. Sources of humic-like substances in the
Pearl River Delta, China: Positive matrix factorization analysis of PM2.5 major components
and source markers. Atmos. Chem. Phys. 15, 1995–2008. doi:10.5194/acp-15-1995-2015
Lagoy, P.K., Nisbet, I.C.T., 1992. Toxic Equivalency Factors ( TEFs ) for Polycyclic Aromatic
Hydrocarbons ( PAHs ). Regul. Toxicol. Pharmacol. 16, 290–300.
Landreman, A.P., Shafer, M.M., Hemming, J.C., Hannigan, M.P., Schauer, J.J., 2008. A
Macrophage-Based Method for the Assessment of the Reactive Oxygen Species ( ROS )
Activity of Atmospheric Particulate Matter ( PM ) and Application to Routine ( Daily-24 h )
Aerosol Monitoring Studies. Aerosol Sci. Technol. 42, 946–957.
doi:10.1080/02786820802363819
Larsen, R.K., Baker, J.E., 2003. Source Apportionment of Polycyclic Aromatic Hydrocarbons in
the Urban Atmosphere : A Comparison of Three Methods. Environ. Sci. Technol. 37, 1873–
1881. doi:10.1021/es0206184
Lee, J.H., Gigliotti, C.L., Offenberg, J.H., Eisenreich, S.J., Turpin, B.J., 2004. Sources of
polycyclic aromatic hydrocarbons to the Hudson River Airshed. Atmos. Environ. 38, 5971–
5981. doi:10.1016/j.atmosenv.2004.07.004
Lee, S., Liu, W., Wang, Y., Russell, A.G., Edgerton, E.S., 2008. Source apportionment of
PM2.5: Comparing PMF and CMB results for four ambient monitoring sites in the
southeastern United States. Atmos. Environ. 42, 4126–4137.
doi:10.1016/j.atmosenv.2008.01.025
148
Li, C.K., Kamens, R.M., 1993. The use of polycyclic aromatic hydrocarbons as source signatures
in receptor modeling. Atmos. Environ. 27, 523–532.
Li, N., Wang, M., Bramble, L.A., Schmitz, D.A., Schauer, J.J., Sioutas, C., 2009. The Adjuvant
Effect of Ambient Particulate Matter Is Closely Reflected by the Particulate Oxidant
Potential. Environ. Health Perspect. 117, 1116–1123. doi:10.1289/ehp.0800319
Li, W., Peng, Y., Shi, J., Qiu, W., Wang, J., Bai, Z., 2011. Particulate polycyclic aromatic
hydrocarbons in the urban Northeast Region of China : Pro fi les , distributions and sources.
Atmos. Environ. 45, 7664–7671. doi:10.1016/j.atmosenv.2011.04.004
Liacos, J.W., Kam, W., Del, R.J., Schauer, J.J., Sioutas, C., 2012. Characterization of organic ,
metal and trace element PM 2 . 5 species and derivation of freeway-based emission rates in
Los Angeles , CA. Sci. Total Environ. 435-436, 159–166.
doi:10.1016/j.scitotenv.2012.06.106
Lim, J.-M., Lee, J.-H., Moon, J.-H., Chung, Y.-S., Kim, K.-H., 2010. Source apportionment of
PM10 at a small industrial area using Positive Matrix Factorization. Atmos. Res. 95, 88–
100. doi:10.1016/j.atmosres.2009.08.009
Lima, A.L.C., Farrington, J.W., Reddy, C.M., 2005. Combustion-Derived Polycyclic Aromatic
Hydrocarbons in the Environment — A Review. Environ. Forensics 6, 109–131.
doi:10.1080/15275920590952739
Lippmann, M., 2010. Targeting the components most responsible for airborne particulate matter
health risks. J. Expo. Sci. Environ. Epidemiol. 20, 117–8. doi:10.1038/jes.2010.1
149
Liu, B., Wu, J., Zhang, J., Wang, L., Yang, J., Liang, D., Dai, Q., Bi, X., Feng, Y., Zhang, Y.,
Zhang, Q., 2017. Characterization and source apportionment of PM2.5 based on error
estimation from EPA PMF 5.0 model at a medium city in China. Environ. Pollut. 222, 10–
22. doi:10.1016/j.envpol.2017.01.005
Liu, W.J., Xu, Y.S., Liu, W.X., Liu, Q.Y., Yu, S.Y., Liu, Y., Wang, X., Tao, S., 2018. Oxidative
potential of ambient PM2.5in the coastal cities of the Bohai Sea, northern China: Seasonal
variation and source apportionment. Environ. Pollut. 236, 514–528.
doi:10.1016/j.envpol.2018.01.116
Liu, Y., Chen, L., Huang, Q., Li, W., Tang, Y., Zhao, J., 2009. Source apportionment of
polycyclic aromatic hydrocarbons ( PAHs ) in surface sediments of the Huangpu River
,Shanghai, China. Sci. Total Environ. 407, 2931–2938. doi:10.1016/j.scitotenv.2008.12.046
Lovett, C., Shirmohammadi, F., Sowlat, M.H., Sioutas, C., 2017. Commuting in Los Angeles :
Cancer and Non-Cancer Health Risks of Roadway , Light-Rail and Subway Transit Routes.
Aerosol Air Qual. Res. 1–12. doi:10.4209/aaqr.2017.09.0331
Luch, A., 2005. The Carcinogenic Effects of Polycyclic Aromatic Hydrocarbons. Imperial
College Press, London.
Lundstedt, S., White, P.A., Lemieux, C.L., Lynes, K.D., Lambert, I.B., Öberg, L., Haglund, P.,
Tysklind, M., 2007. Sources , Fate , and Toxic Hazards of Oxygenated Polycyclic Aromatic
Hydrocarbons (PAHs ) at PAH- contaminated Sites. AMBIO A J. Hum. Environ. 36, 475–
485.
150
Mabahwi, N.A.B., Leh, O.L.H., Omar, D., 2014. Human Health and Wellbeing: Human Health
Effect of Air Pollution. Procedia - Soc. Behav. Sci. 153, 221–229.
doi:10.1016/j.sbspro.2014.10.056
Manoli, E., Kouras, A., Karagkiozidou, O., Argyropoulos, G., Voutsa, D., Samara, C., 2016.
Polycyclic aromatic hydrocarbons ( PAHs ) at traffic and urban background sites of
northern Greece : source apportionment of ambient PAH levels and PAH-induced lung
cancer risk. Environ. Sci. Pollut. Res. 23, 3556–3568. doi:10.1007/s11356-015-5573-5
Manousakas, M., Diapouli, E., Papaefthymiou, H., Kantarelou, V., Zarkadas, C., Kalogridis, A.-
C., Karydas, A.-G., Eleftheriadis, K., 2017a. XRF characterization and source
apportionment of PM10 samples collected in a coastal city. X-ray Spectrom. 47, 190–200.
doi:10.1002/xrs.2817
Manousakas, M., Papaefthymiou, H., Diapouli, E., Migliori, a., Karydas, a. G., Bogdanovic-
Radovic, I., Eleftheriadis, K., 2017b. Assessment of PM2.5 sources and their corresponding
level of uncertainty in a coastal urban area using EPA PMF 5.0 enhanced diagnostics. Sci.
Total Environ. 574, 155–164. doi:10.1016/j.scitotenv.2016.09.047
Mantas, E., Remoundaki, E., Halari, I., Kassomenos, P., Theodosi, C., Hatzikioseyian, A.,
Mihalopoulos, N., 2014. Mass closure and source apportionment of PM2.5 by Positive
Matrix Factorization analysis in urban Mediterranean environment. Atmos. Environ. 94,
154–163. doi:10.1016/j.atmosenv.2014.05.002

151
Marr, L.C., Dzepina, K., Jimenez, J.L., Reisen, F., Bethel, H.L., Arey, J., Gaffney, J.S., Marley,
N.A., Molina, L.T., Molina, M.J., 2006. and Physics Sources and transformations of
particle-bound polycyclic aromatic hydrocarbons in Mexico City. Atmos. Chem. Phys. 6,
1733–1745.
Martellini, T., Giannoni, M., Lepri, L., Katsoyiannis, A., Cincinelli, A., 2012. One year intensive
PM 2.5 bound polycyclic aromatic hydrocarbons monitoring in the area of Tuscany, Italy.
Concentrations, source understanding and implications. Environ. Pollut. 164, 252–258.
doi:10.1016/j.envpol.2011.12.040
Masih, J., Singhvi, R., Kumar, K., Jain, V.K., Taneja, A., 2012a. Seasonal Variation and Sources
of Polycyclic Aromatic Hydrocarbons (PAHs) in Indoor and Outdoor Air in a Semi Arid
Tract of Northern India. Aerosol Air Qual. Res. 12, 515–525.
doi:10.4209/aaqr.2011.11.0192
Masih, J., Singhvi, R., Kumar, K., Jain, V.K., Taneja, A., 2012b. Seasonal Variation and Sources
of Polycyclic Aromatic Hydrocarbons ( PAHs ) in Indoor and Outdoor Air in a Semi Arid
Tract of Northern India. Aerosol Air Qual. Res. 12, 515–525.
doi:10.4209/aaqr.2011.11.0192
Masiol, M., Hofer, A., Squizzato, S., Piazza, R., Rampazzo, G., Pavoni, B., 2012. Carcinogenic
and mutagenic risk associated to airborne particle-phase polycyclic aromatic hydrocarbons :
A source apportionment. Atmos. Environ. 60, 375–382.
doi:10.1016/j.atmosenv.2012.06.073
152
Mauderly, J.L., Chow, J.C., 2008. Health effects of organic aerosols, Inhalation Toxicology.
doi:10.1080/08958370701866008
Mcdonald, J.D., Zielinska, B., Fujita, E.M., Sagebiel, J.C., Chow, J.C., Watson, J.G., 2000. Fine
Particle and Gaseous Emission Rates from Residential Wood Combustion. Environ. Sci.
Technol. 34, 2080–2091. doi:10.1021/es9909632
Meijer, S.N., Sweetman, A.J., Halsall, C.J., Jones, K.C., 2008. Temporal Trends of Polycyclic
Aromatic Hydrocarbons in the U . K . Atmosphere : 1991 – 2005. Environ. Sci. Technol.
42, 3213–3218.
Miller, B.G., Doust, E., Cherrie, J.W., Hurley, J.F., 2013. Lung cancer mortality and exposure to
polycyclic aromatic hydrocarbons in British coke oven workers. BMC Public Health 13, 1–
8.
Miller, K.A., Siscovick, D.S., Sheppard, L., Shepherd, K., Sullivan, J.H., Anderson, G.L.,
Kaufman, J.D., 2007. Long-Term Exposure to Air Pollution and Incidence of
Cardiovascular Events in Women. N. Engl. J. Med. 350, 447–458.
doi:10.1056/NEJMoa1309753
Miller, L., Lemke, L.D., Xu, X., Molaroni, S.M., You, H., Wheeler, A.J., Booza, J., Grgicak-
mannion, A., Krajenta, R., Graniero, P., Krouse, H., Lamerato, L., Raymond, D., Reiners,
J., Weglicki, L., 2010. Intra-urban correlation and spatial variability of air toxics across an
international airshed in Detroit , Michigan ( USA ). Atmos. Environ. 44, 1162–1174.
doi:10.1016/j.atmosenv.2009.12.030
153
Misra, C., Kim, S., Shen, S., Sioutas, C., 2002a. A high flow rate, very low pressure drop
impactor for inertial separation of ultrafine from accumulation mode particles. J. Aerosol
Sci. 33, 735–752.
Misra, C., Singh, M., Shen, S., Sioutas, C., Hall, P.M., 2002b. Development and evaluation of a
personal cascade impactor sampler (PCIS). J. Aerosol Sci. 33, 1027–1047.
Mohagheghi, M.-A., Mosavi-Jarrahi, A., Malekzadeh, R., Parkin, M., 2009. Cancer Incidence in
Tehran Metropolis: The First Report from the Tehran Population-Based Cancer Registry,
1998 – 2001. Arch. Iran. Mediciine 12, 15–23.
Mohsenibandpi, A., Eslami, A., Shahsavani, A., Khodagholi, F., Alinejad, A., 2017.
Physicochemical characterization of ambient PM 2 . 5 in Tehran air and its potential
cytotoxicity in human lung epithelial cells ( A549 ) ☆. Sci. Total Environ. 593-594, 182–
190. doi:10.1016/j.scitotenv.2017.03.150
Morgan, T.E., Davis, D.A., Iwata, N., Tanner, J.A., Snyder, D., Ning, Z., Kam, W., Hsu, Y.-T.,
Winkler, J.W., Chen, J.-C., Petasis, N.A., Baudry, M., Sioutas, C., Finch, C.E., 2011.
Glutamatergic Neurons in Rodent Models Respond to Nanoscale Particulate Urban Air
Pollutants in Vivo and in Vitro. Environ. Health Perspect. 119, 1003–1009.
doi:10.1289/ehp.1002973
Morrison, R.D., Murphy, B.L., 2010. Environmental Forensics: Contaminant Specific Guide.
Academic Press.
154
Motelay-massei, A., Ollivon, D., Garban, B., Chevreuil, M., 2003. Polycyclic aromatic
hydrocarbons in bulk deposition at a suburban site : assessment by principal component
analysis of the influence of meteorological parameters. Atmos. Environ. 37, 3135–3146.
doi:10.1016/S1352-2310(03)00218-8
Mousavi, A., Sowlat, M.H., Hasheminassab, S., Polidori, A., Shafer, M.M., Schauer, J.J.,
Sioutas, C., 2018a. Impact of emissions from the Ports of Los Angeles and Long Beach on
the oxidative potential of ambient PM 0.25 measured across the Los Angeles County. Sci.
Total Environ. 651, 638–647.
Mousavi, A., Sowlat, M.H., Sioutas, C., 2018b. Diurnal and seasonal trends and source
apportionment of redox-active metals in Los Angeles using a novel online metal monitor
and Positive Matrix Factorization ( PMF ). Atmos. Environ. 174, 15–24.
doi:10.1016/j.atmosenv.2017.11.034
Moussiopoulos, Ν., Vlachokostas, C., Tsilingiridis, G., Douros, I., Hourdakis, E., Naneris, C.,
Sidiropoulos, C., 2008. Air quality status in Greater Thessaloniki Area and the emission
reductions needed for attaining the EU air quality legislation. Sci. Total Environ. 407,
1268–1285. doi:10.1016/j.scitotenv.2008.10.034
Mozurkevich, M., 1993. The dissociation constant of ammonium nitrate and its dependence on
temperature, relative humidity and particle size. Atmos. Environ. 27, 261–270.
doi:10.1016/0960-1686(93)90356-4
155
Naddafi, K., Hassanvand, M.S., Yunesian, M., Momeniha, F., Nabizadeh, R., Faridi, S.,
Gholampour, A., 2012a. Health impact assessment of air pollution in megacity of Tehran,
Iran. Iranian J. Environ. Health Sci. Eng. 9, 28. doi:10.1186/1735-2746-9-28
Naddafi, K., Sowlat, M., Safari, M., 2012b. Integrated assessment of air pollution in tehran, over
the period from september 2008 to september 2009. Iran. J. Public Health 41, 77–86.
Nayebare, S.R., Aburizaiza, O.S., Khwaja, H. a, Siddique, A., Hussain, M.M., Zeb, J., Khatib, F.,
Carpenter, D.O., Blake, D.R., 2016. Chemical Characterization and Source Apportionment
of PM2.5 in Rabigh, Saudi Arabia. Aerosol Air Qual. Res. 0658, 1–16.
doi:10.4209/aaqr.2015.11.0658
Ning, Z., Polidori, A., Schauer, J.J., Sioutas, C., 2008. Emission factors of PM species based on
freeway measurements and comparison with tunnel and dynamometer studies. Atmos.
Environ. 42, 3099–3114. doi:10.1016/j.atmosenv.2007.12.039
Norris, G., Duvall, R., Brown, S., Bai, S., 2014. EPA Positive M atrix Factorization (PMF) 5 . 0
Fundamentals and User Guide.
O.Ogbonnaya, U., O.Oyelami, A., O.Umego, M., V.Okere, U., Oluwadare, A.D., A.Fasina, S.,
Semple, K.T., 2017. Assessment of PAH contaminated land: Implementing a risk-based
approach. Environ. Technol. Innov. 8, 84–95. doi:10.1016/j.eti.2017.06.001
Odabasi, M., Bayram, A., Elbir, T., Seyfioglu, R., Dumanoglu, Y., Bozlaker, A., Demircioglu,
H., Altiok, H., Yatkin, S., Cetin, B., 2009. Electric Arc Furnaces for Steel-Making : Hot
Spots for Persistent Organic Pollutants. Environ. Sci. Technol. 43, 5205–5211.
156
OEHHA, 1994. Benzo[a]pyrene as a toxic air contaminant. Berkeley, California, USA:
California Environmental Protection Agency [WWW Document].
Paatero, P., 1997. Least squares formulation of robust non-negative factor analysis. Chemom.
Intell. Lab. Syst. 37, 23–35.
Paatero, P., Eberly, S., Brown, S.G., Norris, G. a., 2014. Methods for estimating uncertainty in
factor analytic solutions. Atmos. Meas. Tech. 7, 781–797. doi:10.5194/amt-7-781-2014
Paatero, P., Eberly, S., Brown, S.G., Norris, G.A., 2014. Methods for estimating uncertainty in
factor analytic solutions 781–797. doi:10.5194/amt-7-781-2014
Paatero, P., Tapper, U., 1994. Positive matrix factorization: A non-negative factor model with
optimal utilization of error estimates of data values. Environmetrics 5, 111–126.
Pandolfi, M., Gonzalez-Castanedo, Y., Alastuey, A., de la Rosa, J.D., Mantilla, E., de la Campa,
a. S., Querol, X., Pey, J., Amato, F., Moreno, T., 2011. Source apportionment of PM10 and
PM2.5 at multiple sites in the strait of Gibraltar by PMF: Impact of shipping emissions.
Environ. Sci. Pollut. Res. 18, 260–269. doi:10.1007/s11356-010-0373-4
Pant, P., Harrison, R.M., 2013. Estimation of the contribution of road traffic emissions to
particulate matter concentrations from field measurements: A review. Atmos. Environ. 77,
78–97. doi:10.1016/j.atmosenv.2013.04.028 Review
Papp, J.F., Lipin, B.R., 2006. “Chromite”. Industrial Minerals & Rocks: Commodities, Markets,
and Uses, 7th ed. SME.
157
Paraskevopoulou, D., Liakakou, E., Gerasopoulos, E., Theodosi, C., Mihalopoulos, N., 2014.
Long-term characterization of organic and elemental carbon in the PM2.5 fraction : the case
of Athens , Greece. Atmos. Chem. Phys. 14, 13313–13325. doi:10.5194/acp-14-13313-2014
Park, S., Kim, J., Jeong, M.-J., Song, B.-J., 2011. Source Identification of Atmospheric
Polycyclic Aromatic Hydrocarbons in Industrial Complex Using Diagnostic Ratios and
Multivariate Factor Analysis. Arch. Environ. Contam. Toxicol. 60, 576–589.
doi:10.1007/s00244-010-9567-5
Park, S.S., Kim, Y.J., Kang, C.H., 2002. Atmospheric polycyclic aromatic hydrocarbons in Seoul
, Korea. Atmos. Environ. 36, 2917–2924.
Pateraki, S., Manousakas, M., Bairachtari, K., Kantarelou, V., Eleftheriadis, K., Vasilakos, C.,
Assimakopoulos, V.D., Maggos, T., 2019. The traffic signature on the vertical PM profile :
Environmental and health risks within an urban roadside environment. Sci. Total Environ.
646, 448–459. doi:10.1016/j.scitotenv.2018.07.289
Pechenizkiy, M., Tsymbal, A., Puuronen, S., 2004. PCA-based feature transformation for
classification: issues in medical diagnostics. Proc. 17th IEEE Symp. Comput. Med. Syst.
535–540. doi:10.1109/CBMS.2004.1311770
Pope, C.A., Burnett, R.T., Thurston, G.D., Thun, M.J., Calle, E.E., Krewski, D., Godleski, J.J.,
2004. Cardiovascular Mortality and Long-Term Exposure to Particulate Air Pollution:
Epidemiological Evidence of General Pathophysiological Pathways of Disease. Circulation
109, 71–77. doi:10.1161/01.CIR.0000108927.80044.7F
158
Pope III, C.A., Burnett, R.T., Thun, M.J., Calle, E.E., Krewski, D., Thurston, G.D., 2002. to Fine
Particulate Air Pollution. J. Am. Med. Assoc. 287, 1132–1141.
doi:10.1001/jama.287.9.1132
Port, Q., Chen, D., Wang, X., Nelson, P., Li, Y., Zhao, N., Zhao, Y., Lang, J., Zhou, Y., Guo, X.,
2017. Ship emission inventory and its impact on the PM2.5 air pollution in Qingdao Port,
North China. Atmos. Environ. 166, 351–361. doi:10.1016/j.atmosenv.2017.07.021
Querol, X., Viana, M., Alastuey, a., Amato, F., Moreno, T., Castillo, S., Pey, J., de la Rosa, J.,
Sánchez de la Campa, a., Artíñano, B., Salvador, P., García Dos Santos, S., Fernández-
Patier, R., Moreno-Grau, S., Negral, L., Minguillón, M.C., Monfort, E., Gil, J.I., Inza, a.,
Ortega, L. a., Santamaría, J.M., Zabalza, J., 2007. Source origin of trace elements in PM
from regional background, urban and industrial sites of Spain. Atmos. Environ. 41, 7219–
7231. doi:10.1016/j.atmosenv.2007.05.022
Rajput, N., Lakhani, A., 2009. Measurements of polycyclic aromatic hydrocarbons at an
industrial site in India. Environ. Monit. Assess. 150, 273–284. doi:10.1007/s10661-008-
0229-2
Reff, A., Eberly, S.I., Bhave, P. V., 2007. Receptor Modeling of Ambient Particulate Matter
Data Using Positive Matrix Factorization: Review of Existing Methods. J. Air Waste
Manage. Assoc. 57, 146–154. doi:10.1080/10473289.2007.10465319
Rengarajan, T., Rajendran, P., Nandakumar, N., Lokeshkumar, B., Rajendran, P., Nishigaki, I.,
2015. Asian Pacific Journal of Tropical Biomedicine. Asian Pac. J. Trop. Biomed. 5, 182–
189. doi:10.1016/S2221-1691(15)30003-4
159
Rich, D.Q., Halu, O., Crooks, J., Baxter, L., Burke, J., Ohman-strickland, P., Thevenet-morrison,
K., Kipen, H.M., Zhang, J., Kostis, J.B., Lunden, M., Hodas, N., Turpin, B.J., 2013. The
Triggering of Myocardial Infarction by Fine Particles Is Enhanced When Particles Are
Enriched in Secondary Species. Environ. Sci. Technol. 47, 9414–9423.
doi:10.1021/es4027248
Risher, J., McDonald, A.R., Citra, M.J., Bosch, S., Amata, R.J., 2003. Toxicological Profile for
Selenium (Update). Atlanta, GA: U.S.Department of Health and Human Services, Public
Health Service.
Sadiktsis, I., Nilsson, G., Johansson, U., Rannug, U., Westerholm, R., 2016. Removal of
polycyclic aromatic hydrocarbons and genotoxic compounds in urban air using air filter
materials for mechanical ventilation in buildings. Sci. Technol. Built Environ. 22, 346–355.
doi:10.1080/23744731.2016.1152155
Saffari, A., Daher, N., Shafer, M.M., Schauer, J.J., Sioutas, C., 2014. Global Perspective on the
Oxidative Potential of Airborne Particulate Matter: A Synthesis of Research Findings.
Environ. Sci. Technol. 48, 7576–7583. doi:10.1021/es500937x
Saffari, A., Hasheminassab, S., Shafer, M.M., Schauer, J.J., Chatila, T. a., Sioutas, C., 2016.
Nighttime aqueous-phase secondary organic aerosols in Los Angeles and its implication for
fine particulate matter composition and oxidative potential. Atmos. Environ. 133, 112–122.
doi:10.1016/j.atmosenv.2016.03.022
160
Samara, C., 2017. On the Redox Activity of Urban Aerosol Particles : Implications for Size
Distribution and Relationships with Organic Aerosol Components. Atmosphere (Basel). 8,
1–16. doi:10.3390/atmos8100205
Samet, J.M., Spengler, J.D., 2003. Indoor Environments and Health : Moving Into the 21st
Century. Am. J. Public Health 93, 1489–1493.
Sannigrahi, P., Sullivan, A.P., Weber, R.J., Ingall, E.D., 2005. Characterization of Water-Soluble
Organic Carbon in Urban Atmospheric Aerosols Using Solid-State 13C NMR
Spectroscopy. Environ. Sci. Technol. 40, 666–672. doi:10.1021/es051150i
Santoso, M., Hopke, P.K., Hidayat, A., Diah Dwiana, L., 2008. Sources identification of the
atmospheric aerosol at urban and suburban sites in Indonesia by positive matrix
factorization. Sci. Total Environ. 397, 229–237. doi:10.1016/j.scitotenv.2008.01.057
Sauvain, J.-J., Vu Duc, T., Guillemin, M., 2003. Exposure to carcinogenic polycyclic aromatic
compounds and health risk assessment for diesel-exhaust exposed workers. Int. Arch.
Occup. Environ. Health 76, 443–455. doi:10.1007/s00420-003-0439-4
Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R.T., 1999. Measurement of Emissions
from Air Pollution Sources . 2. C 1 through C 30 Organic Compounds from Medium Duty
Diesel Trucks. Environ. Sci. Technol. 33, 1578–1587. doi:10.1021/es980081n
Schlesinger, R.B., 2007. The health impact of common inorganic components of fine particulate
matter (PM2.5) in ambient air: A critical review. Inhal. Toxicol. 19, 811–832.
doi:10.1080/08958370701402382
161
Scoullos, M.J., Vonkeman, G.H., Thornton, I., Makuch, Z., 2001. Mercury, Cadmium, Lead:
Handbook for Sustainable Heavy Metals Policy and Regulation. Springer.
Shahbazi, H., Reyhanian, M., Hosseini, V., Afshin, H., 2015. The relative contributions of
mobile sources to air pollutant emissions in Tehran, Iran: An emission inventory approache.
J. Emiss. Control Sci. Technol.
Shahbazi, H., Taghvaee, S., Hosseini, V., Afshin, H., 2016. A GIS based emission inventory
development for Tehran. Urban Clim. 17, 216–229. doi:10.1016/j.uclim.2016.08.005
Shahi, A.M., Omraninava, A., Goli, M., Soheilarezoomand, H.R., Mirzaei, N., 2014. The Effects
of Air Pollution on Cardiovascular and Respiratory Causes of Emergency Admission.
Emergency 2, 107–114.
Shahsavani, a., Naddafi, K., Jafarzade Haghighifard, N., Mesdaghinia, a., Yunesian, M.,
Nabizadeh, R., Arahami, M., Sowlat, M.H., Yarahmadi, M., Saki, H., Alimohamadi, M.,
Nazmara, S., Motevalian, S. a., Goudarzi, G., 2012. The evaluation of PM10, PM2.5, and
PM1 concentrations during the Middle Eastern Dust (MED) events in Ahvaz, Iran, from
april through september 2010. J. Arid Environ. 77, 72–83.
doi:10.1016/j.jaridenv.2011.09.007
Sharma, S.K., Mandal, T.K., Jain, S., Saraswati, Sharma, a., Saxena, M., 2016. Source
Apportionment of PM2.5 in Delhi, India Using PMF Model. Bull. Environ. Contam.
Toxicol. 97, 286–293. doi:10.1007/s00128-016-1836-1
162
Shen, H., Tao, S., Liu, J., Huang, Y., Chen, H., Li, W., Zhang, Y., Chen, Y., Su, S., Lin, N., Xu,
Y., Li, B., Wang, X., Liu, W., 2014. Global lung cancer risk from PAH exposure highly
depends on emission sources and individual susceptibility. Sci. Rep. 4, 1–8.
doi:10.1038/srep06561
Shirmohammadi, F., Hasheminassab, S., Wang, D., Shauer, james J., Shafer, M.M., Delfino,
R.J., Sioutas, C., 2016. The relative importance of tailpipe and non-tailpipe emissions on the
oxidative potential of ambient particles in Los Angeles , CA †. Faraday disscussions 189,
361–380. doi:10.1039/C5FD00166H
Shirmohammadi, F., Lovett, C., Hossein, M., Mousavi, A., Verma, V., Shafer, M.M., Schauer,
J.J., Sioutas, C., 2018. Chemical composition and redox activity of PM 0.25 near Los
Angeles International Airport and comparisons to an urban traf fi c site. Sci. Total Environ.
610-611, 1336–1346. doi:10.1016/j.scitotenv.2017.08.239
Shirmohammadi, F., Wang, D., Hasheminassab, S., Verma, V., Schauer, J.J., Shafer, M.M.,
Sioutas, C., 2017. Oxidative potential of on-road fine particulate matter (PM2.5) measured
on major freeways of Los Angeles, CA, and a 10-year comparison with earlier roadside
studies. Atmos. Environ. 148, 102–114. doi:10.1016/j.atmosenv.2016.10.042
Simcik, M.F., Eisenreich, S.J., Lioy, P.J., 1999. Source apportionment and source / sink
relationships of PAHs in the coastal atmosphere of Chicago and Lake Michigan. Atmos.
Environ. 33, 5071–5079.
163
Simoneit, B.R.T., Schauer, J.J., Nolte, C.G., Oros, D.R., Elias, V.O., Fraser, M.P., Rogge, W.F.,
Cass, G.R., 1999. Levoglucosan, a tracer for cellulose in biomass burning and atmospheric
particles. Atmos. Environ. 33, 173–182. doi:10.1016/S1352-2310(98)00145-9
Singh, M., Misra, C., Sioutas, C., 2003. Field evaluation of a personal cascade impactor sampler
(PCIS). Atmos. Environ. 37, 4781–4793. doi:10.1016/j.atmosenv.2003.08.013
Snyder, D.C., Rutter, A.P., Collins, R., Worley, C., Schauer, J.J., 2009. Insights into the origin of
water soluble organic carbon in atmospheric fine particulate matter. Aerosol Sci. Technol.
43, 1099–1107. doi:10.1080/02786820903188701
Sotoudeheian, S., Arhami, M., 2014. Estimating ground-level PM 10 using satellite remote
sensing and ground-based meteorological measurements over Tehran. J. Environ. Heal. Sci.
Eng. 1–13.
Sowlat, M.H., Hasheminassab, S., Sioutas, C., 2016a. Source apportionment of ambient particle
number concentrations in central Los Angeles using positive matrix factorization (PMF).
Atmos. Chem. Phys. 16, 4849–4866. doi:10.5194/acp-16-4849-2016
Sowlat, M.H., Naddafi, K., Yunesian, M., Jackson, P.L., Lotfi, S., Shahsavani, A., 2013. PM 10
Source Apportionment in Ahvaz, Iran, Using Positive Matrix Factorization. CLEAN - Soil,
Air, Water 41, 1143–1151. doi:10.1002/clen.201200131

164
Sowlat, M.H., Naddafi, K., Yunesian, M., Jackson, P.L., Shahsavani, A., 2012. Source
apportionment of total suspended particulates in an arid area in Southwestern Iran using
positive matrix factorization. Bull. Environ. Contam. Toxicol. 88, 735–740.
doi:10.1007/s00128-012-0560-8
Sowlat, M.H., Wang, D., Simonetti, G., Shafer, M.M., Schauer, J.J., Sioutas, C., 2016b.
Development and field evaluation of an online monitor for near-continuous measurement of
iron, manganese, and chromium in coarse airborne particulate matter (PM). Aerosol Sci.
Technol. 50, 1306–1319. doi:10.1080/02786826.2016.1221051
Stayner, L., Dankovic, D., Smith, R., Steenland, K., 1998. Predicted Lung Cancer Risk Among
Miners Exposed to Diesel Exhaust Particles. Am. J. Ind. Med. 34, 207–219.
Stone, E.A., Snyder, D.C., Sheesley, R.J., Sullivan, A.P., Weber, R.J., Schauer, J.J., 2008.
Source apportionment of fine organic aerosol in Mexico City during the MILAGRO
experiment 2006. Atmos. Chem. Phys. 8, 1249–1259.
Styszko, K., Samek, L., Szramowiat, K., Korzeniewska, A., Kubisty, K., Rakoczy-Lelek, R.,
Kistler, M., Giebl, A.K., 2017. Oxidative potential of PM10 and PM2.5 collected at high air
pollution site related to chemical composition: Krakow case study. Air Qual. Atmos. Heal.
10, 1123–1137. doi:10.1007/s11869-017-0499-3
Taghvaee, S., Sowlat, M.H., Mousavi, A., Hassanvand, M.S., Masud, Y., Naddafi, K., Sioutas,
C., 2018. Source apportionment of ambient PM 2.5 in two locations in central Tehran using
the Positive Matrix Factorization ( PMF ) model. Sci. Total Environ. 629, 672–686.
doi:10.1016/j.scitotenv.2018.02.096
165
Tecer, L.H., Tuncel, G., Karaca, F., Alagha, O., Süren, P., Zararsiz, A., Kirmaz, R., 2012.
Metallic composition and source apportionment of fine and coarse particles using positive
matrix factorization in the southern Black Sea atmosphere. Atmos. Res. 118, 153–169.
doi:10.1016/j.atmosres.2012.06.016
Tian, H.Z., Lu, L., Cheng, K., Hao, J.M., Zhao, D., Wang, Y., Jia, W.X., Qiu, P.P., 2012.
Anthropogenic atmospheric nickel emissions and its distribution characteristics in China.
Sci. Total Environ. 417-418, 148–157. doi:10.1016/j.scitotenv.2011.11.069
Tian, Y., Li, W., Shi, G., Feng, Y., Wang, Y., 2013. Relationships between PAHs and PCBs ,
and quantitative source apportionment of PAHs toxicity in sediments from Fenhe reservoir
and watershed. J. Hazard. Mater. 248-249, 89–96. doi:10.1016/j.jhazmat.2012.12.054
Tongo, I., Ogbeide, O., Ezemonye, L., 2017. Human health risk assessment of polycyclic
aromatic hydrocarbons ( PAHs ) in smoked fish species from markets in Southern Nigeria.
Toxicol. Reports 4, 55–61. doi:10.1016/j.toxrep.2016.12.006
Toro Araya, R., Flocchini, R., Morales Segura, R.G.E., Leiva Guzmán, M. a., 2014.
Carbonaceous aerosols in fine particulate matter of Santiago metropolitan Area, Chile. Sci.
World J. doi:10.1155/2014/794590
Triantafyllou, E., Diapouli, E., Tsilibari, E.M., Adamopoulos, A.D., Biskos, G., 2016.
Assessment of factors in fl uencing PM mass concentration measured by gravimetric & beta
attenuation techniques at a suburban site. Atmos. Environ. 131, 409–417.
doi:10.1016/j.atmosenv.2016.02.010
166
Tsai, P., Shieh, H., Lee, W., Lai, S., 2001. Health-risk assessment for workers exposed to
polycyclic aromatic hydrocarbons ž PAHs / in a carbon black manufacturing industry. Sci.
Total Environ. 278, 137–150.
USEPA, 2010. Development of A Relative Potency Factor (RPF) approach for 544 polycyclic
aromatic hydrocarbon (PAH) mixtures.
US-EPA, 2003. Appendix A to 40 CFR. Part 423e126 Priority Pollutants. [WWW Document].
URL http://www.epa.gov/
Vasilatou, V., Diapouli, E., Abatzoglou, D., Bakeas, E.B., Scoullos, M., Eleftheriadis, K., 2017.
Characterization of PM2.5 chemical composition at the Demokritos suburban station, in
Athens Greece. The influence of Saharan dust. Environ. Sci. Pollut. Res. 24, 11836–11846.
doi:10.1007/s11356-017-8684-3
Vendrame, R., Braga, R.S., Takahata, Y., Galva, D.S., 2001. Structure ± carcinogenic activity
relationship studies of polycyclic aromatic hydrocarbons ( PAHs ) with pattern-recognition
methods. J. Mol. Struct. 539, 253–265.
Venkataraman, C., Friedlander, S.K., 1994. Source Resolution of Fine Particulate Polycyclic
Aromatic Hydrocarbons Using a Receptor Model Modified for Reactivity. J. Air Waste
Manage. Assoc. 44, 1103–1108. doi:10.1080/10473289.1994.10467306
Verma, V., Fang, T., Xu, L., Peltier, R.E., Russell, A.G., Ng, N.L., Weber, R.J., 2015. Organic
Aerosols Associated with the Generation of Reactive Oxygen. Environ. Sci. Technol. 49,
4646–4656. doi:10.1021/es505577w
167
Verma, V., Ning, Z., Cho, A.K., Schauer, J.J., Shafer, M.M., Sioutas, C., 2009. Redox activity of
urban quasi-ultrafine particles from primary and secondary sources. Atmos. Environ. 43,
6360–6368. doi:10.1016/j.atmosenv.2009.09.019
Verma, V., Rico-martinez, R., Kotra, N., King, L., Liu, J., Snell, T.W., Weber, R.J., 2012.
Contribution of Water-Soluble and Insoluble Components and Their
Hydrophobic/Hydrophilic Subfractions to the Reactive Oxygen Species-Generating
Potential of Fine Ambient Aerosols. Environ. Sci. Technol. 46, 11384–11392.
doi:10.1021/es302484r
Vossler, T., Černikovský, L., Novák, J., Williams, R., 2016. Source apportionment with
uncertainty estimates of fine particulate matter in Ostrava, Czech Republic using Positive
Matrix Factorization. Atmos. Pollut. Res. 7, 503–512. doi:10.1016/j.apr.2015.12.004
Vreeland, H., Weber, R., Bergin, M., Greenwald, R., Golan, R., Russell, A.G., Verma, V.,
Sarnat, J.A., 2017. Oxidative potential of PM 2.5 during Atlanta rush hour : Measurements
of in-vehicle dithiothreitol ( DTT ) activity. Atmos. Environ. 165, 169–178.
doi:10.1016/j.atmosenv.2017.06.044
Wai, W., Tam, S., Wong, T.W., Wong, A.H.S., 2015. Association between air pollution and
daily mortality and hospital admission due to Ischaemic heart diseases in Hong Kong.
Atmos. Environ. doi:10.1016/j.atmosenv.2015.08.068.This

168
Wang, D., Pakbin, P., Saffari, A., Shafer, M.M., Shauer, J.J., Sioutas, C., 2013. Development
and Evaluation of a High-Volume Aerosol-into-Liquid Collector for Fine and Ultrafine
Particulate Matter. Aerosol Sci. Technol. 47, 1226–1238.
doi:10.1080/02786826.2013.830693
Wang, D., Sowlat, M.H., Shafer, M.M., Schauer, J.J., Sioutas, C., 2016. Development and
evaluation of a novel monitor for online measurement of iron, manganese, and chromium in
ambient particulate matter (PM). Sci. Total Environ. 565, 123–131.
doi:10.1016/j.scitotenv.2016.04.164
Wang, D., Tian, F., Yang, M., Liu, C., Li, Y., 2009. Application of positive matrix factorization
to identify potential sources of PAHs in soil of Dalian , China. Environ. Pollut. 157, 1559–
1564. doi:10.1016/j.envpol.2009.01.003
Wang, F., Lin, T., Feng, J., Fu, H., Guo, Z., 2015. Source apportionment of polycyclic aromatic
hydrocarbons in PM2.5 using positive matrix factorization modeling in Shanghai, China.
Environ. Sci. Process. Impacts 17. doi:10.1039/C4EM00570H
Wang, F., Lin, T., Li, Y., Ji, T., Ma, C., Guo, Z., 2014. Sources of polycyclic aromatic
hydrocarbons in PM2 . 5 over the East China Sea , a downwind domain of East Asian
continental out fl ow. Atmos. Environ. 92, 484–492. doi:10.1016/j.atmosenv.2014.05.003
Wang, G., Gu, S., Chen, J., Wu, X., Yu, J., 2016. Assessment of health and economic effects by
PM2.5 pollution in Beijing: a combined exposure-response and computable general
equilibrium analysis. Environ. Technol. 3330, 1–8. doi:10.1080/09593330.2016.1178332
169
Wang, H., Naghavi, M., Allen, C., Barber, R., Carter, A., Casey, D., Charlson, F., Chen, A.,
Coates, M., Coggeshall, M., 2016. Global, regional, and national life expectancy, all-cause
mortality, and cause-specific mortality for 249 causes of death, 1980–2015: a systematic
analysis for the Global Burden of Disease Study 2015. Lancet 388, 1459–1544.
doi:10.1016/S0140-6736(16)31012-1
Wang, W., Massey Simonich, S.L., Xue, M., Zhao, J., Zhang, N., Wang, R., Cao, J., Tao, S.,
2010. Concentrations , sources and spatial distribution of polycyclic aromatic hydrocarbons
in soils from Beijing , Tianjin and surrounding areas , North China. Environ. Pollut. 158,
1245–1251. doi:10.1016/j.envpol.2010.01.021
Wang, Z.B., Hu, M., Wu, Z.J., Yue, D.L., He, L.Y., Huang, X.F., Liu, X.G., Wiedensohler, a.,
2013. Long-term measurements of particle number size distributions and the relationships
with air mass history and source apportionment in the summer of Beijing. Atmos. Chem.
Phys. 13, 10159–10170. doi:10.5194/acp-13-10159-2013
WHO, 2018. Exposure to ambient air pollution from particulate matter for 2016.
WHO, 2016. WHO Global Urban Ambient Air Pollution Database.
WHO, 2011. Exposure to air pollution (particulate matter) in outdoor air. WHO Regional Office
for Europe, Copenhagen.
Yassin, M.F., Alhajeri, N.S., A.Kassem, M., 2016. Indoor and Built Polycyclic aromatic
hydrocarbons collected from indoor built environments on heating, ventilation and air
conditioning dust filters 25, 137–150. doi:10.1177/1420326X14541557
170
Yunker, M.B., Macdonald, R.W., Vingarzan, R., H. Mitchell, R., Goyette, D., Sylvestre, S.,
2002. PAHs in the Fraser River basin : a critical appraisal of PAH ratios as indicators of
PAH source and composition. Org. Geochem. 33, 489–515.
Zakaria, M.P., Takada, H., Tsutsumi, S., Ohno, K., Yamada, J., Kouno, E., Kumata, H., 2002.
Distribution of Polycyclic Aromatic Hydrocarbons ( PAHs ) in Rivers and Estuaries in
Malaysia : A Widespread Input of Petrogenic PAHs. Environ. Sci. Technol. 36, 1907–1918.
Zauli Sajani, S., Ricciardelli, I., Trentini, A., Bacco, D., Maccone, C., Castellazzi, S., Lauriola,
P., Poluzzi, V., Harrison, R.M., 2015. Spatial and indoor/outdoor gradients in urban
concentrations of ultrafine particles and PM2.5mass and chemical components. Atmos.
Environ. 103, 307–320. doi:10.1016/j.atmosenv.2014.12.064
Zhang, M., Wu, C., Zhu, X., Wang, Y., 2014. Loss of Imprinting of Insulin-Like Growth Factor
2 is Associated with Increased Risk of Primary Lung Cancer in the Central China Region.
Asian Pacific J. Cancer Prev. 15, 7799–7803.
Zhang, R., Jing, J., Tao, J., Hsu, S.C., Wang, G., Cao, J., Lee, C.S.L., Zhu, L., Chen, Z., Zhao,
Y., Shen, Z., 2013. Chemical characterization and source apportionment of PM2.5 in
Beijing: Seasonal perspective. Atmos. Chem. Phys. 13, 7053–7074. doi:10.5194/acp-13-
7053-2013
Zhang, X., Staimer, N., Gillen, D.L., Tjoa, T., Schauer, J.J., Shafer, M.M., Hasheminassab, S.,
Pakbin, P., Vaziri, N.D., Sioutas, C., Delfino, R.J., 2016. Associations of oxidative stress
and in fl ammatory biomarkers with chemically-characterized air pollutant exposures in an
elderly cohort 150, 306–319. doi:10.1016/j.envres.2016.06.019
171
Zhang, Y., Guo, C., Xu, J., Tian, Y., Shi, G., Feng, Y.-C., 2012. Potential source contributions
and risk assessment of PAHs in sediments from Taihu Lake , China : Comparison of three
receptor models. Water Res. 46, 3065–3073. doi:10.1016/j.watres.2012.03.006
Zhao, J., Xia, L., Sehgal, a., Lu, D., McCreery, R.L., Frankel, G.S., 2001. Effects of chromate
and chromate conversion coatings on corrosion of aluminum alloy 2024-T3. Surf. Coatings
Technol. 140, 51–57. doi:10.1016/S0257-8972(01)01003-9
Zhou, M., Liu, Y., Wang, L., Kuang, X., Xu, X., Kan, H., 2014. Particulate air pollution and
mortality in a cohort of Chinese men. Environ. Pollut. 186, 1–6.
doi:10.1016/j.envpol.2013.11.010
Zhuo, S., Du, W., Shen, G., Wang, R., Pan, X., Li, T., Han, Y., Li, Y., Pan, B., Peng, X., Cheng,
H., Wang, X., Shi, G., Xing, B., Tao, S., 2017. Urban air pollution and health risks of parent
and nitrated polycyclic aromatic hydrocarbons in two megacities, southwest China. Atmos.
Environ. 166, 441–453. doi:10.1016/j.atmosenv.2017.07.051
Zong, Z., Wang, X., Tian, C., Chen, Y., Qu, L., Ji, L., Zhi, G., Li, J., Zhang, G., 2016. Source
apportionment of PM2.5 at a regional background site in North China using PMF linked
with radiocarbon analysis: Insight into the contribution of biomass burning. Atmos. Chem.
Phys. 16, 11249–11265. doi:10.5194/acp-16-11249-2016
Zuo, Q., Duan, Y.H., Yang, Y., Wang, X.J., Tao, S., 2007. Source apportionment of polycyclic
aromatic hydrocarbons in surface soil in Tianjin , China. Environ. Pollut. 147, 303–310.

Abstract (if available)

Abstract Many metropolitan environments around the globes have been facing serious air pollution challenges during the last decades due to rapid rate of industrialization and urbanization, resulting in a wide variety of adverse health consequences, including neurodegenerative disorder, cardiovascular diseases, respiratory problems and inflammation. While ambient particulate matters (PM) mass concentration are regulated by the air quality agencies and health officials as one of the major criteria air pollutants, several previous studies have indicated that ambient PM is consisted of various chemical components, each of them being emitted from different pollution sources. In addition, there are a plethora of studies in the literature documenting the higher toxicity of some PM compounds (e.g., redox active metals) in comparison to the other chemical constituents. Therefore, it is vital to examine the association of PM toxicity with individual components of PM (and in turn their emitting sources) rather than the total PM bulk mass. This would definitely help the air quality authorities to develop more targeted control schemes in protecting people from detrimental health impacts of exposure to ambient PM. ❧ Consequently, the main objective of this dissertation is to investigate the major contributing sources to PM mass concentration as well as its associated toxicity and health impacts. To this end, experimental measurements of PM were conducted in Tehran, Athens, and Los Angeles as examples of highly polluted and populated urban environments in three different continents (i.e., Asia, Europe, and America). The collected samples were then analyzed for their chemical and toxicological compounds

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

Impact of urban source emissions on ambient particulate matter (PM) composition and associated toxicity in local and regional scales using source appointment models

PDF

Chemical and toxicological characteristics of particulate matter in urban environments with a focus on its sources, associated health impacts and mitigation policies

PDF

Physico-chemical characteristics and sources of ambient PM mass and number concentrations and their associated toxicity, and development of novel techniques for high time-resolution measurement o...

PDF

Analysis of sources and profiles of organic carbon in ambient particulate matter across fine and coarse sizes and introducing an optical technique for real-time urban dust measurement

PDF

Assessing indoor and outdoor air quality by characterizing the physicochemical and toxicological properties of particulate matter using…

PDF

Development of novel techniques for evaluating physical, chemical and toxicological properties of particulate matter in ambient air

$Physico-chemical properties and source apportionment of size-fractionated airborne particulate matter in urban areas with implications for public health$

PDF

Physico-chemical properties and source apportionment of size-fractionated airborne particulate matter in urban areas with implications for public health

PDF

Oxidative potential of urban atmospheric particles: spatiotemporal trends and associations with source-specific chemical components

$Chemical and toxicological characteristics and historical trends of size-fractioned particulate matter from traffic-related emissions in Los Angeles$

PDF

Chemical and toxicological characteristics and historical trends of size-fractioned particulate matter from traffic-related emissions in Los Angeles

PDF

Size-resolved particulate matter (PM) in urban areas: toxico-chemical characteristics, sources, trends and health implications

PDF

Toxicity of urban particulate matter: long-term health risks, influences of surrounding geography, and diurnal variation in chemical composition and the cellular oxidative stress response

$Exposure assessment and source apportionment of size fractions of airborne particulate matter$

PDF

Exposure assessment and source apportionment of size fractions of airborne particulate matter

PDF

Measurement and methods of assessing the impact of prevalent particulate matter sources on air quality in southern California

PDF

Temporal, spatial and toxicological characteristics of coarse particulate matter in an urban area and relation to sources and regulations

PDF

Investigating the role of urban emission sources on redox-active PM compounds and the chemical analysis of the standardized diesel exhaust particles

PDF

Characterization of black carbon: from source to evolution of physical and optical properties in the atmosphere

PDF

Toxicological characteristics of particulate matter in an urban environment and their linkage to the source-specific constituents

PDF

Investigation of physico-chemical characteristics of particulate matter from vehicular sources

PDF

Impacts of heat mitigation strategies and pollutant transport on climate and air quality from urban to global scales

PDF

Investigation of the physical and chemical characteristics of ambient coarse particulate matter in indoor and outdoor environments

Asset Metadata

Creator Taghvaee, Sina (author)

Core Title Investigating the temporal trends, sources, and toxicity of ambient particulate matter (PM) in different metropolitan environments, and development of a novel aerosol generation setup for inhalat...

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Engineering (Environmental Engineering)

Publication Date 07/23/2020

Defense Date 05/01/2020

Publisher University of Southern California (original), University of Southern California. Libraries (digital)

Tag aerosol generation,air quality,ambient particulate matter,OAI-PMH Harvest,oxidative potential,positive matrix factorization (PMF) model,source apportionment

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sioutas, Constantinos (committee chair), Ban-Weiss, George (committee member), Mack, Wiliam (committee member)

Creator Email sinataghvaee@gmail.com,taghvaee@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-c89-345071

Unique identifier UC11666069

Identifier etd-TaghvaeeSi-8740.pdf (filename),usctheses-c89-345071 (legacy record id)

Legacy Identifier etd-TaghvaeeSi-8740.pdf

Dmrecord 345071

Document Type Dissertation

Rights Taghvaee, Sina

Type texts

Source University of Southern California (contributing entity), University of Southern California Dissertations and Theses (collection)

Access Conditions The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...

Repository Name University of Southern California Digital Library

Repository Location USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA