University of Southern California Dissertations and Theses

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

(USC Thesis Other)

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

PDF

Download a page range

Download transcript

Copy asset link

Request this asset

Transcript (if available)

Content
TOXICOLOGICAL CHARACTERISTICS OF PARTICULATE MATTER IN AN
URBAN ENVIRONMENT AND THEIR LINKAGE TO THE SOURCE-SPECIFIC
CONSTITUENTS

by

Vishal Verma

___________________________________________________________________

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
(ENVIRONMENTAL ENGINEERING)

May 2011

Copyright 2011 Vishal Verma

ii

Dedication

To my Anjalu
for her persistent love, motivation, and faith in my abilities
&
To my mother
for her unconditional love and support throughout my life
iii

Acknowledgements

I would like to take this opportunity to express my sincere thanks to all those people who
made it possible for me to complete this thesis. It is my great pleasure to convey my
sincere gratitude to all of them with my humble acknowledgement.

To begin with, I am deeply indebted to my graduate advisor Prof. Constantinos Sioutas,
who gave me an opportunity of working under his guidance and also his whole hearted
support, interest and encouragement, that made this work possible. His thoughtful
discussions on technical aspects, deep interest and enthusiasm in aerosol research and
pragmatic approach towards problem solving have helped me to develop the scientific
acumen.

I express deep sense of gratitude to the members of my guidance committee, Prof. Ronald
C Henry and Prof. Nicos A Petasis for providing thoughtful and insightful suggestions
on my research work as well as dissertation.

I would also like to thank my previous and current lab mates and friends
both in the group at USC and in collaborative institutions. I want to thank the
post doctors in our group: Dr. Zhi Ning, Dr.Subhasis Biswas, Dr. Markus Sillanpaa, and
also Debbie and Emma from UCLA who have provided valuable advice in science
discussion and guidance in the papers writing. Special thanks also go to Dr. Andrea
iv

Polidori. It is his advice and exceptional experience that have helped me during the initial
days of my study at USC. I also extend my thanks to Payam Pakbin, Neelakshi Hudda,
Kalam Cheung, Winnie Kam, Nancy Daher, Zhen Xu, Diana, Niloofar, Shruthi, and
Risabh Jain and for their valuable support.

Last but not the least, I would like acknowledge my wife (Geetanjali), my parents (Savitri
Devi and Inder Kumar Verma), my friends (Satya Da, Babu, Harish Bhai) who helped me
directly or indirectly throughout the work.

v

Table of Contents

Dedication ii

Acknowledgements iii

List of Tables viii

List of Figures ix

Abstract xii

Chapter 1: Introduction 1
1.1 Background 1
1.1.1 Characteristics and Sources of Particulate Matter (PM) 1
1.1.2 Health Effects of Particulate Matter 3
1.2 Rationale for the Present Study 4
1.3 Thesis Layout 7

Chapter 2: Physicochemical and Toxicological Properties of PM
Emissions from Heavy-Duty Diesel Vehicles 10
2.1 Introduction 10
2.2 Experimental Methods 11
2.2.1 Vehicles 13
2.2.2 Equipment and Instruments 16
2.2.3 Sample Analysis 18
2.3 Results and Discussion 19
2.3.1 Size Distribution 19
2.3.2 Size Segregated Mass Emission Factors 25
2.3.3 Particle Volatility 28
2.3.4 Emission Factors (EF) of Inorganic Ions and Total
Carbon (TC) 30
2.3.5 Size Segregated Chemical Speciation 35
2.3.6 DTT Activity 38
2.3.7 Water Soluble Organic Carbon and DTT 45
2.3.8 Correlation between Chemical Species and Oxidative
Potential 46
2.4 Conclusions 47

vi

Chapter 3: Toxicity Profiles of Ambient Particles Influenced by Vehicular,
Woodsmoke and Secondary Photochemical Sources 49
3.1 Oxidative Activity Profiles of Ambient Particulate Matter from
Primary and Secondary Sources 49
3.1.1 Introduction 49
3.1.2 Experimental Methods 51
3.1.2.1 Sampling Location 51
3.1.2.2 Sampling Protocol 52
3.1.2.3 Sample Analysis 54
3.1.3 Results and Discussion 55
3.1.3.1 Physical Parameters 55
3.1.3.2 Chemical Parameters 58
3.1.3.3 Oxidative Activity 66
3.1.3.4 Correlation of Oxidative Activity with
Chemical Constituents of PM 71
3.1.4 Conclusions 77
3.2 Oxidative Activity Profiles of Ambient Particulate Matter
Influenced by Woodsmoke Emissions from Wildfires 78
3.2.1 Introduction 78
3.2.2 Experimental Methods 79
3.2.2.1 Sampling Location 79
3.2.2.2 Sampling Protocol 80
3.2.2.3 Sample Analysis 81
3.2.3 Results and Discussion 81
3.2.3.1 Particulate Matter Characteristics 81
3.2.3.2 Gaseous Pollutants 84
3.2.3.3 Water Soluble Trace Elements 85
3.2.3.4 Water Soluble Organic Carbon (WSOC) 87
3.2.3.5 Organic Constituents 88
3.2.3.6 PM Oxidative Activity 91
3.2.4 Conclusions 97

Chapter 4: Contribution of Semi-Volatile Organic Compounds and
Non-Volatile Transition Metals in the PM Oxidative Activity 98
4.1 Contribution of Semi-Volatile Organic Compounds in the
Oxidative (DTT) Activity of Ambient Particles 98
4.1.1 Introduction 98
4.1.2 Experimental Methods 100
4.1.2.1 Sampling Location 100
4.1.2.2 Sampling Protocol 100
4.1.2.3 Sample Analysis 104
4.1.3. Results and Discussion 104
4.1.3.1 Physical Parameters 104
4.1.3.2 Chemical Parameters 110
vii

4.1.3.3 Oxidative Potential 118
4.1.3.4 Association of Oxidative Potential with PM
Chemical Composition 120
4.1.4 Summary and Conclusions 122
4.2 Contribution of Transition Metals in the Reactive Oxygen Species
Activity of PM Emissions from Retrofitted Heavy-Duty Vehicles 123
4.2.1 Introduction 123
4.2.2 Experimental Methods 124
4.2.2.1 Sampling Protocol 124
4.2.2.2 Sample Analysis 125
4.2.3 Results and Discussion 127
4.2.3.1 Water Soluble Metals 127
4.2.3.2 Macrophage ROS Activity 132
4.2.3.3 Chelex Treatment 135
4.2.3.4 Statistical Association between ROS
Activity and Metals 138
4.2.4 Summary and Conclusions 145

Chapter 5: Conclusions and Future Research Directions 147
5.1 Summary and Conclusions 147
5.2 Recommendations 149

Bibliography 154

Appendix 170
viii

List of Tables

Table 2.1: Details of test fleet 14

Table 2.2: Size fractionated mass EFs of test fleet 26

Table 2.3: Correlation coefficient (R) and significance level (p)
for oxidative potential (measured by DTT) and selected
chemical species from test fleet 46

Table 3.1: Average concentrations of water-soluble elements in
ambient quasi-ultrafine particles at the USC sampling
site during morning and afternoon periods 61

Table 3.2: Summary of the regression analysis for select species
with DTT and ROS levels of ultrafine particles at
USC sampling site. 73

Table 3.3: Average fractional concentration of water soluble
elements (ICP-MS) in ambient particles collected at
the USC site during and after the fire event 86

Table 3.4: Coefficients of statistical determination (R
2
) and
associated levels of significance (p-value) for the
correlations between selected water soluble PM
constituents and redox activities (as measured by
the DTT and ROS assays) for PM samples collected
during and after the fire period. 96

Table 4.1: Summary of the regression analysis between
major PM chemical constituents and DTT activity
of the ambient and thermodenuded ultrafine particles. 121

Table 4.2: Pearson correlation coefficients and associated levels
of significance for the univariate regression between
ROS activity and major water-soluble metals of PM
from the test fleet 139

Table 4.3: Parameters of the multivariate regression analysis
between ROS activity and water-soluble metals of PM
from the test fleet 144

ix

List of Figures

Figure 2.1: Experimental set-up (Dynamometer) 12

Figure 2.2: Particle (exhaust) number size distribution of the test fleet 21

Figure 2.3: Number (DMS) and mass emission factors (nano-MOUDI)
for exhaust particles from the test fleet 27

Figure 2.4: Size distribution of thermo denuded aerosols for cruise
driving mode of the test fleet 29

Figure 2.5: Emission factors of integrated chemical species
(from the test fleet) measured by MOUDI-NanoMOUDI 31

Figure 2.6: EC, WSOC and WIOC emission rates from the test fleet 33

Figure 2.7: Emission rates of size-resolved PM chemical species from
the test fleet 36

Figure 2.8: Oxidative potential (DTT consumption in n-moles min
-1
μg
-1

of PM ) of thermo-denuded and undenuded PM from the
test fleet 40

Figure 2.9: Relationship between oxidative activity and semivolatile
PM fraction at 150 °C for the PM emitted from test fleet 42

Figure 2.10: DTT consumption rate (for the exhaust particles) per unit
distance traveled by vehicles 43

Figure 2.11: DTT consumption in relation to the water-soluble organic
carbon (WSOC) of PM from the test fleet 45

Figure 3.1: Diurnal profile of the average particle number concentration
of quasi-ultrafine particles measured at the USC sampling
site and the average particle number size distribution in
“morning” and “afternoon” periods 57

Figure 3.2: Average concentration of the major inorganic ions in
quasi-ultrafine particles collected in the morning and afternoon
periods at the USC sampling site 58

x

Figure 3.3: Average concentration of the various organic species
measured at the USC sampling site during the morning
and afternoon periods 62

Figure 3.4: Oxidative activity (DTT) of quasi-ultrafine particles collected
at the USC sampling site during the morning and afternoon
periods 68

Figure 3.5: Oxidative activity (ROS) of quasi-ultrafine particles
collected at the USC sampling site during the morning and
afternoon periods 70

Figure 3.6: Evolution of the particle number size distribution from days
characterized by wildfire emissions to periods not directly
affected by fire 83

Figure 3.7: 24-hr averaged concentration of NO, NO2, CO and O3 at
North Main Street (Los Angeles) and Particle Number
Concentration at the USC (also 24-hr average) before, during
and after the fire period 84

Figure 3.8: Fractional concentration of the organic constituents in
PM samples collected at the USC site, during and after
the fire events 89

Figure 3.9: Oxidative activity of PM samples collected at the USC
site during and after the fire period 92

Figure 4.1: Experimental setup for sampling ambient semi-volatile
particles 102

Figure 4.2: Particle size distributions (number and volume based) of
the concentrated ambient and thermodenuded particles at
50, 100 and 200
o
C 107

Figure 4.3: Concentration of water soluble inorganic ions in the
concentrated ambient and thermodenuded particles
at different TD temperature configurations – 50, 100, 200
o
C 112

Figure 4.4: Concentration of elemental and organic carbon in the
concentrated ambient and thermodenuded particles collected
at three TD temperature configurations (50, 100 and 200
o
C) 114

xi

Figure 4.5: Concentration of PAHs in concentrated ambient and
thermodenuded aerosols collected at three TD temperature
configurations – 50, 100 and 200
o
C 117

Figure 4.6: Percentage loss in oxidative potential, measured by DTT
activity on heating the ambient aerosols at 50, 100 and 200
o
C 118

Figure 4.7: Distribution of water-soluble metals in the exhaust PM
from various vehicle-configurations, under three driving
cycles, i.e. cruise, UDDS and idle 128

Figure 4.8: Water-soluble content (µg/g) of major redox active metals in
the exhaust PM from various vehicle-configurations 131

Figure 4.9: Reactive Oxygen Species (ROS) activity of PM from the
tested vehicles 133

Figure 4.10: Percent removal (average of all vehicle-configurations) of
major water-soluble elements using Chelex treatment of
the exhaust PM samples 136

Figure 4.11: Percent removal of ROS activity in relation to that of
aggregate water soluble metals after Chelex treatment of the
exhaust PM samples from test vehicle-configurations
under different driving cycles 137

Figure 4.12: Correlation of measured and reconstructed ROS activity
with successive inclusions of different species in
the multivariate regression model 142

xii

Abstract

A number of population based epidemiological studies as well as recent toxicological and
clinical studies indicate a strong association between particulate matter (PM) exposure
and adverse health outcomes. Despite commendable progress in particle-related
toxicological research for the last few decades, the exact mechanisms by which PM
inflicts health injuries are still largely unknown and constitute a subject of great interest
for the scientific community. An increase in the abundance of reactive oxygen species
(ROS) in biological systems after PM exposure and the resulting oxidative stress has
been hypothesized to be mostly responsible for the initiation of inflammatory cascades.

The core objective of this work is to determine the toxicological characteristics of
particulate matter in an urban environment and to investigate their associations with the
source-specific particle constituents. This objective has been accomplished by evaluating
the oxidative potential of particles collected from various sources such as exhaust tail
pipe of the heavy-duty diesel vehicles, wood-smoke and ambient particles in segregation
to their primary and secondary sources. The role of semi-volatile organic compounds in
the oxidative activity of PM was assessed by their removal using thermodenuder and
measuring the resultant oxidative potential by DTT assay. Similarly, the contribution of
transition metals was quantified by their chelation using Chelex
®
chromatography. The
use of statistical tools (bivariate and multivariate regression techniques) further supported
in identification of the specific PM constituents responsible for major variability in the
xiii

responses of toxicological assays. The results demonstrate the importance of both semi-
volatile (organic compounds) and non-volatile (transition metals) species of particulate
matter in stimulating the generation of different oxidizing species in biological systems,
measured by DTT and ROS assay. These findings are useful in elucidating the health
risks related to the PM exposure from different sources and ultimately in promulgating
the effective control strategies to protect public health.

1

Chapter 1: Introduction

1.1 Background
1.1.1 Characteristics and Sources of Particulate Matter (PM)
Particulate matter is a general name given to the solid particles and/or liquid droplets, and
when suspended in air, is also known as aerosol. Particulate matter is a mixture of many
subclasses of chemical species - both organic and inorganic, most of which are
potentially hazardous to human health. Particulate matter can be divided into three broad
categories based on their aerodynamic diameters. These modes are: coarse mode (2.5-10
µm), accumulation mode (0.1-2.5µm) and the ultrafine mode (0.01-0.1 µm). Particles
with aerodynamic diameter larger than 10 µm are usually not the point of interest in
environmental and health studies because of their short life time in the atmosphere. The
aerodynamic diameter of particles is an important property because it determines
majority of its behavior in the atmosphere such as residence time and deposition
characteristics in the human respiratory system.

The sources of airborne particulate matter can be divided into two broad categories:
primary and secondary. Primary sources include combustion sources (heavy and light
duty vehicles, woodsmoke, industries), and construction activities. The particles emitted
from these primary sources may undergo photochemical processing in the presence of
various atmospheric oxidants to yield secondary particles. The physical and chemical
characteristics of these secondarily formed particles are distinctly different compared to
2

their primary precursors. The secondary photochemical reactions in addition to the
vehicular emissions constitute the most prominent sources of ultrafine particles in an
urban metropolis like Los Angeles (Fine et al., 2004).

Another important property of the particulate matter which governs its existence either in
particle or gas phase is the volatility. Thus, particulate matter can be considered as a
complex mixture of semi-volatile (most organic compounds and the major inorganic ions
– ammonium, sulfate and nitrate) and non-volatile (metals, elemental carbon and few
organic compounds) species (Seinfeld and Pandis, 2006). The semi-volatile compounds
exist simultaneously in the gas and particle phases at equilibrium. The vapor pressure of
these compounds is a strong function of temperature, which determines the extent of their
partitioning into particulate and gaseous phase at a given concentration. Recent studies
have indicated that the organic compounds volatilized and subsequently photo-oxidized
from the semi-volatile fraction of the particles substantially contribute to the formation of
secondary organic aerosol (SOA) (Robinson et al., 2007). This contribution far exceeds
that from the photo-oxidation of primary volatile organic compounds (VOCs).

Heavy-duty diesel trucks constitute only a small fraction of the total fleet in California
but have an important contribution to the emissions of fine and ultrafine particles. Efforts
from both the regulatory and the manufacturing communities have been devoted to the
reduction of emissions from diesel-powered engines. US EPA 2007 emissions standards
enforce to reduce the diesel PM mass emissions from heavy-duty engines ten fold from
3

the old 0.1 g bhp
-1
h
-1
PM limit to 0.01 g bhp
-1
h
-1
(Merrion, 2003). These new emission
standards motivated the development of several after-treatment devices for in-use heavy-
duty diesel vehicles (HDDVs), e.g. diesel particulate filter (DPF) and selective catalytic
reduction (SCR) retrofits to reduce the PM and NO
x
emissions, respectively. Although,
these after-treatment devices significantly reduce the mass emission rates (McGeehan et
al., 2005) of DEPs, but not necessarily the number-based particle emissions. Several
studies have shown that under certain conditions, enhanced formation of ultrafine
particles occurs by heterogeneous nucleation for the vehicles equipped with after-
treatment devices (Kittelson et al., 2006; Vaaraslahti et al., 2004).

1.1.2 Health Effects of Particulate Matter
Both epidemiological and clinical research have demonstrated strong links between
atmospheric aerosols and adverse health effects, including premature deaths (Hoek et al.,
2002; Samet et al., 2000), respiratory and cardiovascular diseases (Gauderman et al.,
2007; Riediker et al., 2004; Stayner et al., 1998), and neurodegenerative disorders (Peters
et al., 2006). Particles deposit in different parts of the lung according to their
aerodynamic diameter, resulting in varying degree of toxic potency. In-vitro and in-vivo
experiments have linked particle exposure to airway inflammation, mitochondrial damage
and lung cancer (Castranova et al., 2001; Garshick et al., 2004; Kumagai et al., 1997; Li
et al., 2003; Saldiva et al., 2002; Yin et al., 2007). Chronic exposure of DEPs may lead to
exacerbation of pulmonary diseases such as asthma and bronchitis as well as lung cancer
(Garshick et al., 2004; Rusznak et al., 1994). Alveolar macrophages (AM), which are
4

responsible for the defense of susceptible cells, release pro-inflammatory cytokines in the
presence of PM constituents, such as polycyclic aromatic hydrocarbons (PAHs), and may
experience alteration in their gene expressions (Koike et al., 2002). A few studies have
also described negative impacts of PM on reproductive systems (Watanabe and Oonuki,
1999), liver functions (Folkmann et al., 2007) and brain activity (Cruts et al., 2008).

1.2 Rationale for the Present Study
The exact mechanisms by which particulate matter inflicts health injuries remain largely
unknown; however, PM-associated reactive oxygen species (ROS), and the resulting
oxidative stress induced in human tissues is believed to be implicated in inflammatory
effects. In healthy biological systems, natural generation of ROS as a result of aerobic
metabolism is balanced through scavenging by endogenous antioxidants (Halliwell and
Cross, 1994); oxidative stress results when ROS concentrations exceed the capacity of
the antioxidant systems to maintain ROS levels within physiologically normal ranges.
Toxicological studies have confirmed the oxidative activity of particulate matter and its
ability to stimulate cellular generation of ROS (Donaldson et al., 1996; Donaldson et al.,
2002; Donaldson et al., 2003; Kunzli et al., 2006). A few investigations have also linked
PM toxicological characteristics to their chemical composition, most notably to organic
species and transition metals (Chen and Lippmann, 2009; Cho et al., 2005; DiStefano et
al., 2009; Ntziachristos et al., 2007; Shafer et al., 2010). Most of these studies indicate a
consistently higher oxidative potential on a per PM mass basis of quasi-ultrafine particles
5

(e.g. ≤250 nm) compared to the fine and coarse particles (Cho et al., 2005; Donaldson et
al., 2002; Ntziachristos et al., 2007).

Both molecular and cellular assays have been used to examine PM-induced oxidative
stress. In molecular (cell-free) assays, chemical method such as consumption of
dithiothreitol [DTT; (Cho et al., 2005)] has been used. The DTT assay determines in
quantitative terms the capacity of a test sample to transfer electrons from dithiothreitol
(DTT) to oxygen – a reaction analogous to the cellular redox reaction involving NADPH
and oxygen. The electron transfer is monitored by the rate at which DTT is consumed
under a standardized set of conditions, and the rate is proportional to the concentration of
the redox-active species in the PM sample. To assess the relevance of this assay to
cellular effects, studies have been conducted in which the DTT based redox activity of a
sample has been compared with its capacity to induce the stress protein, hemeoxygenase-
1 in two cell lines – RAW264.7 and BEAS-2B. RAW 264.7 is a murine macrophage cell
line that mimics the oxidative stress response of pulmonary alveolar macrophages (Hiura
et al., 1999; Li et al., 2002), and BEAS-2B is a transformed human bronchial epithelial
cell line, which mimics the oxidative stress response of primary bronchial epithelial cells
(Li et al., 2002). DTT activities of PM samples were found to correlate with the induction
of hemeoxygenase-1 in both cell lines (Li et al., 2003). These data suggest that redox
activity measured by this procedure is a reasonable predictor of oxidative stress status in
cells.

6

Fluorescent probes such as 2 ′,7 ′- dichlorofluorescein diacetate (DCF) (Becker et al.,
2005) has been extensively employed to study the cellular generation of ROS. The
macrophage ROS assay is a fluorogenic cell-based method to examine the production of
reactive oxygen species and is performed in rat alveolar macrophages (NR8383,
American Type Culture Collection), which are exposed to aqueous suspensions of PM.
The ROS probe used in this method – DCFH-DA (2,7-dichlorodihydrofluorescein
diacetate) is a non-fluorescent, membrane permeable compound. Upon entering a cell, it
is de-acetylated by cellular house-keeping esterases, yielding 2,7-
dichlorodihydrofluorescein (DCFH). Non-fluorescent DCFH can then be converted by
free radicals and other oxidants within the cell into the highly fluorescent 2,7-
dichlorofluorescein (DCH) which can be easily monitored using a fluorescence plate
reader. Further details of this assay are given in Landreman et al., (2008).

While each of these assays was putatively considered a germane tool for measurement of
net PM oxidative activity in biochemical systems, an increasing body of evidence,
discussed in literature (Cho et al., 2005; DiStefano et al., 2009; Shinyashiki et al., 2009)
suggests that the various toxicity assays are selective in their response to different PM
constituents. It is thus generally believed that knowing particle chemical composition
along with their physical attributes and source of emissions may be the critical starting
point in understanding the fraction within PM that drives the health effects.

7

The primary objective of this study is to evaluate the toxicological characteristics of
particulate matter in an urban environment and their linkages to the source-specific
particle constituents. This objective is carried out by collecting both ultrafine and fine
particles from an array of diverse emission sources, i.e. primary (heavy-duty diesel
vehicles, and ambient PM influenced by vehicular and woodsmoke emissions), and
secondary photochemical sources. Both cell-free (DTT) and cell-based (macrophage
ROS) assays were used to evaluate the oxidative activity of the collected particles. In
addition, physicochemical characteristics of the sampled particles, such as particle
number distribution, elemental and organic carbon, water soluble organic carbon, water
soluble elements, inorganic ions and organic species were also analyzed. The associations
of PM chemical constituents with their oxidative characteristics were investigated by
mechanistic (physicochemical segregation of PM constituents) and statistical (bivariate
and multivariate regression) techniques. This study offers a novel and informative
perspective on the relationship between composition and sources of atmospheric particles
to their relative toxicity potential.

1.3 Thesis Layout
This thesis is divided into five chapters:

Chapter 1 is the introduction and gives an overview of the particulate matter along with
their sources, characteristics and health effects. It also sets up the rationale for the present
study.
8

Chapter 2 describes the physicochemical and toxicological characteristics (i.e. size
distribution, volatility, size-resolved chemical components, and DTT activity) of PM
emitted from heavy-duty diesel vehicles retrofitted with state-of-the-art after-treatment
devices. Apart from assessing the potency of semi-volatile species to induce oxidative
stress, the results are useful in assessing the efficacy of these control technologies and in
establishing the future needs for additional measures to control these PM species.

Chapter 3 discusses the physicochemical and toxicological characteristics of ambient
particles from different primary and secondary sources (e.g. vehicular emissions,
woodsmoke, and secondary photochemical reactions) at an urban site near downtown Los
Angeles.

Chapter 4 explores the linkages between chemical constituents of PM with their
toxicological profiles. Both mechanistic (physicochemical segregation of PM
constituents) and statistical (bivariate and multivariate regression) analyses have been
performed to quantify the contribution of these species in oxidative activity of the
particles.

Finally, Chapter 5 of the thesis is the concluding chapter which summarizes the findings
of the present investigation. It also identifies the limitations associated with the current
methodologies and provides recommendations to improve upon. The chapter thus paves
9

the road for future investigations aimed at a better assessment of the role of different PM
constituents in the observed health effects.

10

Chapter 2: Physicochemical and Toxicological Properties of
PM Emissions from Heavy-Duty Diesel Vehicles

2.1 Introduction
Diesel exhaust particles (DEPs) are normally agglomerates of hundreds of volatile/semi-
volatile species adsorbed onto its refractory carbonaceous core (Bayona et al., 1988). As
general perception has emerged towards the potential risks of diesel particulate matter,
policy makers are promulgating stricter emission control rules and regulations. To
effectively meet such stringent emission standards, various advanced engine design and
control technologies are being considered and rigorously evaluated for the newer fleet of
heavy duty trucks. While these after-treatment devices (such as DPF) have been highly
efficient in removing refractory solid particles (>50nm), some of the potentially harmful
volatile and semi volatile species (such as PAHs), originally emitted in the vapor phase at
high plume temperature, may penetrate through (Kittelson, 1998; Matter et al., 1999). As
the exhaust temperature decreases drastically at the tail pipe exit, these vapor phase
species condense and form fresh nucleation mode particles (Kittelson, 1998).

There has been a limited number of papers in the literature reporting the chemical
composition of PM emissions from diesel vehicles equipped with DPF and SCR retrofits
(Grose et al., 2006), especially for newer vehicles meeting the US 2007 emission
standards. To a certain degree, these studies are limited by the insufficient PM mass of
the exhaust that poses a challenge for detailed chemical analysis. Nevertheless, PM
emissions from a diesel engine with advanced after-treatment devices are fundamentally
11

different than that from an uncontrolled diesel engine with respect to particle size,
morphology, and chemical composition.

The primary objective of this collaborative study between the California Air Resources
Board (CARB) and the University of Southern California is to evaluate the
physicochemical (size distribution, volatility, size-resolved chemical composition) and
toxicological characteristics of the volatile and non-volatile fractions of particles emitted
from a variety of different engines, fuels and emissions control, each operating under
different driving conditions using a dynamometer set-up. Comparisons within heavy-duty
vehicle types and driving cycles and also with respect to a baseline vehicle (without any
control technology) are discussed.

2.2 Experimental Methods
Experiments were carried out at the California Air Resources Board’s (CARB) heavy-
duty diesel emission testing laboratory (HDETL) in downtown Los Angeles. Ayala et al.,
(2002) described the dynamometer specifications in details. Figure 2.1 shows the
schematic of the experimental setup. The sampling train includes heavy-duty chassis
dynamometer, constant volume sampling (CVS) dilution tunnel and aerosol samplers.
Diesel vehicle exhausts were transported by a stainless steel hose pipe and diluted with
filtered air through the CVS. Measurements were taken 18 diameter lengths downstream
of the exhaust introduction in the CVS. Three driving cycles, i.e. steady state cruise
(50mph), transient [EPA urban dynamometer driving schedule (UDDS)] and idle were
12

tested to simulate various real-world driving conditions. The fuel used to run the engines
was CARB ultra-low sulfur diesel (ULSD) with sulfur content less than 15 ppm. Tunnel
blank levels were measured and vehicles were conditioned (warmed up) everyday before
the start of official runs. The CVS was cleaned prior to starting the project.

Figure 2.1: Experimental setup

13

2.2.1 Vehicles
The test fleet comprised of four heavy-duty diesel vehicles in seven configurations (Table
2.1). A 1998 Kenworth truck served as a baseline vehicle, without any emission control
technology. The same Kenworth truck was also tested with three different control
technologies: a Continuously Regenerating Technology [CRT], consisting of a diesel
oxidation catalyst (DOC) followed by an uncatalyzed trap; CRT in combination with a
selective catalytic reduction system [Zeolite or vanadium based SCRTs]. The other three
test vehicles were a diesel hybrid electric bus, a school bus, and a Caltrans truck.
14

Table 2.1: Details of test fleet

Vehicle Engine After treatment (AT) Dilution
Make Nomenclature Model
Size
[L]
Type
Miles
on AT
Approximate
Dilution
Ratio in CVS
Kenworth V-SCRT
Cummins
M11,
reflashed
11
Vanadium
based
SCRT®
50,000
9.2 Cruise
6-30 UDDS
14 Idle
Kenworth Z-CERT
Cummins
M11,
reflashed
11
Zeolite based
SCRT®
0 on
SCR,
50,000
on
CRT
9.2 Cruise
6-30 UDDS
14 Idle
International DPX
International
DT466E
7.6
Engelhard
DPX
30,000
6.2 Cruise
5-25 UDDS
22 Idle
Gillig (35ft)
with Alison
Hybrid
Hybrid-
CCRT
Cummins 5.9 CCRT® 1000 5-50 UDDS
Thompson-
School Bus
EPF/School
Bus/Horizon
Cummins 5.9
Cleaire -
Horizon
32,000
8.3 Cruise
8-80 UDDS
25 Idle
Kenworth CRT
Cummins
M11,
reflashed
11
Continuously
Regenerating
Technology
64,000
9.2 Cruise
6-30 UDDS
14 Idle
Kenworth Baseline
Cummins
M11,
reflashed
11 None NA.
9.2 Cruise
6-30 UDDS
14 Idle

15

The two SCRT technologies consist of a wall-flow particulate trap (CRT) followed by a
SCR section. The CRT was the same in each configuration. The difference between them
lies in the choice of catalysts (vanadium or Zeolite) for the SCR to control oxides of
nitrogen (NOx). The Caltrans truck, with a smaller engine (7.6 L) than the Kenworth
truck (11 L, Table 2.1), is retrofitted with an Engelhard DPX filter. The DPX filter is
comprised of a diesel particulate trap with a catalytic wash-coat. The diesel hybrid
electric bus (San Joaquin Valley RTD) is equipped with a catalyzed continuously
regenerative trap, or CCRT, consisting of a DOC followed by a catalyzed trap, which was
virtually brand-new, with only 1000 miles on the odometer. The last test vehicle was an
Elk Grove school bus, equipped with an electric particle filter (EPF). The EPF (Horizon)
consists of a non-catalyzed silicon carbide substrate for PM control, coupled with an
electric heating element and a small blower. The trap is regenerated periodically using
electricity from the grid (plug in configuration) during non-operational periods—mostly
at night. Hereafter, the test fleet is referred as baseline, CRT, V-SCRT, Z-SCRT, DPX,
hybrid-CCRT and EPF/School Bus/Horizon.

The dilution air flow rate at CVS was 2600 cfm (74 m
3
min
-1
) for cruise and UDDS
cycles, and 1600 cfm (45.4 m
3
min
-1
) for the idle. For EPF and Hybrid-CCRT, the flow
rates were maintained at 1600 cfm for all cycles. These flow-rates result in approximate
dilution ratios of 6-9 for cruise, 5-80 for UDDS, and 15-25 for idle.

16

2.2.2 Equipment and Instruments
Specific descriptions of some of the instruments and equipment used for this study are
provided in this section.

Nano-MOUDI
Size-resolved samples were collected using a micro-orifice uniform deposited impactor
(MOUDI) upstream of a nano-MOUDI (MSP Corporation, Minneapolis, MN) loaded
with pre-cleaned aluminum foil substrates. Particles were classified in the following
aerodynamic size ranges: 10-18nm, 18-32nm, 32-56nm, 56-100nm, 100-180nm, 180nm-
2.5μm and >2.5μm. The MOUDI-Nano MOUDI tandem was operated for multiple runs
in order to accumulate sufficient mass for chemical analysis for each vehicle and driving
cycle.

DMS/ EEPS
Size distribution of engine exhausts from the CVS was monitored every second by two
multiple channel differential mobility spectrometers: a DMS500 (Cambustion) and an
engine exhaust particle sizer (EEPS 3090, TSI Inc.). Both DMS and EEPS classify
particles on the basis of their mobility diameter. The cut-off size ranges of EEPS and
DMS are 5.6-523 nm and 4.5- 1000 nm, respectively. With high time resolutions, they
are both capable of tracking transient particle behavior, especially during UDDS cycles.
With few exceptions (V-SCRT, Z-SCRT only), in which the DMS was placed

17

downstream of the particle measurement program (PMP) sampler, both instruments were
connected directly to CVS.

PMP
A particle measurement program (PMP) protocol was developed in Europe to measure
the solid particle emissions from light duty vehicles. The sampling train of PMP contains
a volatile particle remover (VPR) and a particle counter. The VPR provides two stage
dilution connected by an evaporation tube (ET). The temperature for the primary dilution,
ET and secondary dilution is 150, 300 and ~35
o
C, respectively. Detailed information for
the PMP can be seen in Herner et al.,(2007a).

Thermodenuder
Particle volatility was determined by two thermodenuders (Model ELA-230, Dekati Ltd)
sampling in parallel, each heating the entering aerosol to 150 and 230
o
C, respectively.
The thermodenuder consists of a heating section, followed by an adsorption/cooling unit.
As aerosol stream was drawn from the CVS and passed through the heating tube, part of
its volatile/semi-volatile components was sheared off. These labile species adsorb onto a
layer of activated charcoal placed on the walls of the thermodenuder, leaving the non-
volatile PM fraction to be collected on Teflon filters (47mm, PTFE, Gelman) placed
downstream of the thermodenuders. Multiple runs were integrated to achieve desired
sample mass loadings on these filters to perform various chemical and toxicological
analyses. Solid particle number concentrations and size distributions were monitored

18

intermittently by a condensation particle counter (CPC 3022 A, TSI Inc., MN) and a
differential mobility particle sizer (DMA, TSI 3085).

High Volume Sampler
The high volume sampler is a specially designed filter holder [Hi-Q Environmental
Products Co., CA; (Misra et al., 2002)] to collect particles on 20x25 cm filters at a
relatively high flow rate (450 l/min). Integrated PM samples from the CVS were
collected using this sampler on Teflon coated glass fiber (GF) filters (Pallflex, Fiberfilm
T60A20 – 8x10 inch, Pall Corp., East Hills, NY).

2.2.3 Sample Analysis
Aluminum substrates from MOUDI-NanoMOUDI stages were pre and post-weighed
during the sampling to determine the mass loadings. The elemental and organic carbon
(EC, OC) content of deposited PM was analyzed by thermal-optical method (Kleeman et
al., 1999). The ion concentrations were derived by ion chromatography (IC) technique.
The water-soluble organic carbon (WSOC) content extracted from Teflon coated GF
filters were analyzed by Shimadzu TOC-5000A liquid analyzer (Zhang et al., 2008). The
oxidative potential of PM was measured by the DTT assay. The DTT assays were
performed for denuded and undenuded PM collected on Teflon filters and undenuded PM
on Teflon-coated glass fiber filters.

19

2.3 Results and Discussion
2.3.1 Size Distribution
Figure 2.2 presents tunnel blank- subtracted mean particle size distributions for vehicles
at different operating conditions. These distributions derived from DMS/EEPS
measurements are grand averages of multiple runs for each driving cycle. The data are
reported in terms of number per vehicle kilometer traveled for cruise and UDDS cycles,
and number per hour for idle. The primary focus here is to give a brief overview in order
to provide meaningful insights and support to some of the findings described in
subsequent sections.

While performing preliminary quality assurance – quality control (QA/QC) of real time
data for SCRTs (V-cruise, UDDS, Z-Cruise), we noticed that few EEPS sizes channels
(10-20nm) were almost always saturated due to particle over-loadings. These size bins
were subsequently replaced with corresponding secondary dilution-corrected DMS data
(the DMS was placed downstream of the PMP secondary dilution with the ET off) to
obtain more accurate size distributions. For rest of the test fleet [except CRT], we used
data from the DMS, which was directly connected to the CVS. We have added a tunnel
blank distribution in Figure 2.2f (Baseline vehicle).

Size distribution patterns for both SCRTs are quite similar with sharp modes (Figure 2.2
a,b) at ~ 10nm. The distinguishable feature is the less prominent nucleation mode for
UDDS runs especially for the Z-SCRT-UDDS cycle. This may be due to the fact that the

20

Zeolite-SCRT requires much higher temperature to trigger and sustain nucleation than the
vanadium based SCRT catalysts (Herner et al., 2007b). Moreover, in general the zeolite
catalysts have lot more catalytic surface area than vanadium catalysts and the Z-SCRT
system used for this study is completely new (Table 2.1). These provide higher storage
sites for sulfate generated by the upstream DOC and DPF during transient and low
temperature testing.

21

Figure 2.2 a, b, c, d, e, f: Particle number size distribution

V-SCRT®
1.E+09
1.E+11
1.E+13
1.E+15
1.E+17
1 10 100 1000 Dp (nm)
dN/dlogDp
(Particles km
-1
)
1.E+09
1.E+11
1.E+13
1.E+15
dN/dlogDp
(Particles hr
-1
)
V-SCRT-Cruise
V-SCRT-UDDS
V-SCRT-Idle
Z-SCRT®
1.E+07
1.E+09
1.E+11
1.E+13
1.E+15
1.E+17
1 10 100 1000 Dp (nm)
dN/dlogDp
(Particles km
-1
)
1.E+09
1.E+11
1.E+13
1.E+15
dN/dlogDp
(Particles hr
-1
)
Z-SCRT-Cruise
Z-SCRT-UDDS
Z-SCRT-Idle
a)
b)

22

Figure 2.2 (Continued)

DPX®
1.E+07
1.E+09
1.E+11
1.E+13
1.E+15
1.E+17
1 10 100 1000 Dp (nm)
dN/dlogDp
(Particles km
-1
)
DPX_Cruise
DPX-UDDS
Non-Nucleating Vehicles
1.E+07
1.E+09
1.E+11
1.E+13
1 10 100 1000 Dp (nm)
dN/dlogDp
(Particles km
-1
)
EPF-Cruise
EPF-UDDS
Hybrid-CCRT-UDDS
d)
c)

23

Figure 2.2 (Continued)

Although the main purpose of SCR technologies is to reduce NOx by ammonia, at
elevated temperatures, their in-built catalysts may encourage the formation of sulfate, an
important component acting potentially as seed aerosol for particle formation by
CRT®
1.E+08
1.E+10
1.E+12
1.E+14
1.E+16
1 10 100 1000
Dp (nm)
dN/dlogDp
(Particles km
-1
)
CRT-Cruise
CRT-UDDS
Baseline
1E+11
1E+12
1E+13
1E+14
1E+15
1E+16
1 10 100 1000
Dp (nm)
dN/dlogDp
(Particles km
-1
)
1.E+10
1.E+11
1.E+12
1.E+13
1.E+14
dN/dlogDp
(Particles hr
-1
)
Baseline-Cruise
Baseline-UDDS
Tunnel Blank
f )
e)

24

condensation of semi-volatile organic vapors. For a brief sampling period, we bypassed
the SCR portion from exhaust after-treatment system (SCRT) just to investigate the
impact of the SCR catalysts. Although, this modification did not result in visible
alteration of the shape (on log scale) of the distributions (CRT, Figure 2.2e), number
concentration decreased by a factor of 2-3 from the V-SCRT and Z-SCRT cruise cycles -
suggesting SCR catalysts’s role on nucleation. Unlike cruise or UDDS cycles, idle runs
are characterized with remarkably low particulate number emission rates, coupled with
broad size distributions (Figure 2.2a, b). The second test engine, Engelherd-DPX (Figure
2.2c) displays a dominant nucleation mode, almost identical to the SCRTs/CRT. It has
been hypothesized that the catalyst wash-coat on DPF (if saturated) may be enhancing the
conversion of SO
2
to SO
3
/sulfate and partially stimulates the nucleation process (Hansen
et al., 2001).

Contrary to the general notion that particulate filters augment nucleation, the hybrid
vehicles [with a CCRT] and the school bus [with an EPF] were found to be highly
efficient in suppressing if not eliminating this PM mode. The Hybrid-CCRT vehicle
(Figure 2.2d) resulted in concentrations (CVS) in the range of ~10
4
particles cm
-3
, thus a
1,000-fold improvement over the previously tested vehicles (>10
7
particles cm
-3
). We
hypothesize that the initial capacity of its relatively new trap (with only 1000 miles on it)
to store sulfur has significantly suppressed the formation of nuclei mode particles. Once
all the storage sites are saturated, nucleated sulfate particles are expected. For this vehicle
only a few odd large particles are left downstream (Figure 2.2d). The school bus (Figure

25

2.2d), however, was the cleanest amongst the entire test fleet with number emissions less
than 1500 particles cm
-3
measured in the CVS. It is important to note here that this vehicle
is equipped only with an uncatalyzed filter (EPF) which is least likely to enhance the
nucleation process. Thus, nucleation is not only control device specific but also a
function of age and operating conditions e.g. temperature of the catalysts.

The baseline truck, on the other hand, represents the older genre of vehicles and was
found to emit substantial amounts of larger particles (Figure 2.2f) with modes in the 60-
100 nm range. Because of their large surface area, these accumulation mode particles act
as adsorption sites and thus perfect sinks for organic vapors, leading to suppression of
nucleation mode (Liu et al., 2007).

2.3.2 Size Segregated Mass Emission Factors
Size fractionated mass emissions factors (in mg km
-1
or mg hr
-1
) are calculated based on
the loadings on the MOUDI-nano-MOUDI impaction plates. The mass loadings on the
individual substrates are generally low for retrofitted vehicles and depending on size
ranges variation in the order of 20-40% (Standard Deviation/Mean; from few duplicate
measurements) are observed.

Although a direct correspondence of mobility and aerodynamic diameters is not accurate
without establishing some conversion factors, we can utilize the information from Table
2.2 to complement the mobility size distributions (Figure 2.2) described before. Number

26

and mass based size distributions are found to be in reasonable agreement with each other
in terms of their trends.

Table 2.2: Size fractionated mass EFs (mg km
-1
for Cruise and UDDS, mg hr
-1
for Idle)
Size (nm) 10--18 18-32 32-56 56-100 100-180
180-
2500 >2500
V-SCRT
UDDS 0.73 1.44 2.46 1.36 0.86 1.07 0.83
Idle 3.18 2.48 1.77 4.25 12.03 7.78 6.37
Z-SCRT
Cruise 0.9 1.48 1.6 1.15 0.53 0.97 0.5
UDDS 0.17 0.34 0.32 0.17 0.62 0.55 0.66
Idle 1.06 3.54 1.42 0.35 3.54 0 3.18
DPX
Cruise 0.65 0.29 0.3 0.11 0.27 0.21 0.2
UDDS 0.56 0.46 1.25 0.6 0.41 0.2 0.35
Idle 4.91 1.89 12.45 3.4 7.92 3.02 5.66
Hybrid
(CCRT)
UDDS 0.26 0.12 0.22 0.33 0.17 0.57 0.35
CRT
Cruise 1.35 0.86 0.98 0.49 0.57 1.15 0.37
UDDS 1.25 1.52 3.19 0.55 1.52 2.63 0.83
EPF
Cruise 0.04 0.22 0.12 0.14 0.07 0.15 0.12
UDDS 0.05 0.08 0.16 0.1 0.05 0.29 0.17
Idle 5.92 3.48 5.92 4.53 3.13 8.71 7.66
Baseline
Cruise 1.02 1.96 5.96 9.98 29.4 31.1 1.03
UDDS 1.39 2.54 10.7 19.6 86.8 192 3.41
Idle 19.363 60.241 565.83 600.26 2327.9 6049.9 346.39

27

Figure 2.3: Number (DMS) and mass emission factors (nano-MOUDI)

The emission factors are remarkably low (~1-12 mg km
-1
) for the fleet operated with
control technologies compared to the baseline vehicle (~80-316 mg km
-1
). While the
majority (>95%) of PM mass is concentrated between 100nm-2.5µm for the baseline
truck, nuclei modes are clearly visible for vehicles retrofitted with control devices.
Significant reduction (>90%) of the mass is achieved for vehicles retrofitted with control
devices.

Some general trends and inferences can be drawn from Figure 2.2 and Table 2.2.
Consistent with a previous study (Vaaraslahti et al., 2004), the majority of the control
technologies evaluated here have promoted bulk production of nano-size (nucleation)
particles during steady state and high speed segment of transient running cycles. The
EPF
V-SCRT
Z-SCRT
DPX
CRT
Baseline
EPF
DPX
CRT
Baseline
Hybrid-CCRT
Z-SCRT
V-SCRT
1.0E-01
1.0E+00
1.0E+01
1.0E+02
1.0E+03
1.0E+10 1.0E+11 1.0E+12 1.0E+13 1.0E+14 1.0E+15 1.0E+16
Number EF (Particles km
-1
)
Mass EF (mg km
-1
)
Cruise
UDDS
Non-Nucleating
Vehicles

28

cruise cycles on average generate higher nucleation and lesser accumulation mode
particles than the UDDS cycles. The differences between these driving cycles are even
more pronounced in mass distributions (Table 2.2): significant shifts towards larger sizes
are apparent in UDDS runs due to increased emission of accumulation mode particles
during the acceleration processes (Polidori et al., 2008). Aerosol formation mechanism,
poorly understood till date, seems to be a function of vehicle type and driving conditions
and retrofit design.

Figure 2.3 is an informative graph showing a plot of number EFs vs. Mass EFs (nano-
MOUDI, except V-SCRT-Cruise) for each vehicle and driving cycle. Each data point
corresponds to a cruise or an UDDS cycle for a given vehicle. The graph shows that for
several of the after-treatment devices tested, particle number emissions increase with
reduced mass emissions. There are some outliers or exceptions observed for Z-SCRT at
UDDS cycle, Hybrid-CCRT, EPF and Idle (not shown) for which both number and mass
EFs are relatively low. Thus, most of the new after-treatment devices appear to be highly
efficient in reducing mass emissions but some are not effective in controlling number
concentrations due to the formation of nuclei mode particles.

2.3.3 Particle Volatility
Particle volatility was measured at 150 and 230
o
C except in few occasions (Hybrid-
CCRT, EPF) when both the thermodenuders were operated at 150
o
C to maximize our
capability to collect mass of non-volatile particles. For the majority of the cases,

29

volatility at 230
o
C is at least an order of magnitude higher than their counterparts at
150
o
C, indicating complete disappearance of large fraction of particles within this
temperature window.

The significant particle number loss between 150 to 230
o
C, especially for cruise cycles,
may be elucidated by the evaporation profiles shown in Figure 2.4. The majority of
particles in the 7-20 nm size range for DPX and CRT have disappeared as the aerosol
stream is heated from 150
o
C to 230
o
C. This is in total contrast to the baseline vehicle,
where no noticeable shift in size spectrum is observed. Matter et al., (1999) reported very
similar thermal desorption trends between 172 and 204
o
C for particles sampled
downstream of a DPF.

Figure 2.4: Size distribution of thermo denuded aerosols for cruise mode
Note: Available only for DPX®, CRT® and Baseline

1.00E+01
1.00E+02
1.00E+03
1.00E+04
1.00E+05
1.00E+06
1.00E+07
1.00E+08
1 10 100 1000
Dp (nm)
dN/dlogDp (Number cm
-3
)
Baseline-TD 150
Baseline-TD 230
CRT-TD 150
CRT-TD 230
DPX-TD150
DPX-TD 230
Significant
loss of
nucleation
mode particles

30

2.3.4 Emission Factors (EF) of Inorganic Ions and Total Carbon (TC)
Figure 2.5 shows the EFs (expressed in mg of PM species per kilometer driven) of
chemical species integrated over the MOUDI-NanoMOUDI stages from 0.01- 2.5 µm.
The low emission vehicles pose a challenge for PM chemical analysis due to very low
mass loadings on individual substrates. The analytical uncertainties are shown by the
error bars.

31

Figure 2.5: Emission factors of integrated chemical species measured by MOUDI-
NanoMOUDI. a). Cruise 50 mph b). Transient UDDS

Cruise 50 mile Hr
-1
Species
Sodium Ammonium Potassium Chloride Nitrate Sulfate TC
EF (mg Km
-1
)
1e-5
1e-4
1e-3
1e-2
1e-1
1e+0
1e+1
1e+2
UDDS
Species
Sodium Ammonium Potassium Chloride Nitrate Sulfate TC
EF (mg Km
-1
)
1e-5
1e-4
1e-3
1e-2
1e-1
1e+0
1e+1
1e+2
1e+3

a)
b)

32

At both cruise and UDDS cycles, significant reductions in the emission of TC were
observed for the retrofitted vehicles compared to the baseline vehicle. However, elevated
emission rates of sulfate and ammonium were observed for vehicles and-or driving cycles
that have a substantial fraction of nucleation mode particles, i.e. CRT, V-SCRT
(measurements only available for UDDS cycle), Z-SCRT-cruise and DPX, compared to
the baseline vehicle. The school bus retrofitted with an electric particulate filter (Horizon)
was consistently the cleanest vehicle with the lowest emissions of inorganic ions and TC.
This is also in agreement with the emission rates of PM based on gravimetric mass
measurement.

The emission factors of TC were further split into EC, water-soluble and water-insoluble
OC (WSOC and WIOC, respectively) and shown in Figure 2.6. The baseline vehicle
(shown on secondary axis) not only emits considerable amount of EC (Cruise: 13.8 mg
km
-1
, UDDS: 65.9 mg km
-1
) but also elevated levels of WIOC (Cruise: 28.2 mg/km;
UDDS: 88 mg/km) and WSOC (Cruise: 2.7 mg km
-1
; UDDS: 14 mg km
-1
). The low EFs
of WIOCs (~0.1-0.4 mg km
-1
) and WSOC (0.05-0.95 mg km
-1
) for the retrofitted vehicles
suggest that these control devices are also effective (65-99%) in reducing the major
fraction of organic carbon in addition to EC. However, it is important to note here that
only the baseline, V-SCRT, Z-SCRT and CRT configurations (same vehicle) are directly
comparable and caution should be exercised while comparing with other vehicle
configurations (DPX, Horizon, CCRT).

33

Figure 2.6: EC, WSOC and WIOC emission rates at a) Cruise 50mph and b) Transient
UDDS c) Proportion of WSOC to OC.
Note: V-SCRT cruise and CCRT-Cruise cycles are not available

Cruise
Z-SCRT
DPX
Horizon
Baseline
EF (mg km
-1
)
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
EF (mg km
-1
)
0
10
20
30
40
50
EC
WSOC
WIOC
UDDS
V-SCRT
Z-SCRT
DPX
Horizon
CCRT
N.a.N.
Baseline
EF (mg km
-1
)
0.0
0.2
0.4
0.6
0.8
EF (mg km
-1
)
0
50
100
150
200
a)
b)

34

Figure 2.6 (Continued)

In addition to the overall reduction of OC, the proportion of WSOC to OC, illustrated in
Figure 2.6c varies with vehicle configurations and running cycles. The baseline vehicle
shows the lowest OC solubility (WSOC/OC =0.08-0.13) during both cruise and UDDS
cycles, suggesting freshly emitted primary OC. By comparison, OC emitted from
Horizon and SCRTs consists primarily of WSOC. DPX and CCRT WSOC contents are
intermediate between the baseline and other retrofitted vehicles. An earlier study by
Hameri et al., (2001) indicated higher water solubility of OC in the nucleation/UF mode
than the accumulation mode. This may explain, to a certain extent, the higher WSOC/OC
ratio of SCRTs and Horizon, where significant fraction of OC is observed in the
WSOC/OC
Configuration
Baseline V-SCRT Z-SCRT DPX Horizon CCRT
Fraction
0.0
0.2
0.4
0.6
0.8
1.0
Cruise
UDDS
c)

35

nucleation mode. In contrast, OC dominates in accumulation mode for the baseline
vehicle. In addition to size distribution of OC, control technologies seem to have major
influences on the percentage of WSOC emitted by vehicles. The DPF of Horizon and
SCRTs (higher WSOC/OC) are uncatalyzed, while DPX and CCRT (lower WSOC/OC)
have catalyzed filters. Although SCRTs and CCRT both have DOCs upstream of the
DPF, the results suggest that DOCs, in comparison to DPF types (catalyzed or
uncatalyzed) have a minimal effect on the WSOC/OC ratio. This conclusion, however,
does not imply that the DOCs have not contributed in reducing OC from the total vehicle
emissions.

2.3.5 Size Segregated Chemical Speciation
The ion and total carbon emission data reported in Figure 2.7 a,b,c are further segregated
into three size ranges, i.e. 10-56nm, 56-180nm and 180-2500nm. The V-SCRT-UDDS,
Z-SCRT cruise, CRT and DPX have displayed the highest content of sulfate in the
nucleation mode. This finding is consistent with the conclusions of recent emission
studies that nuclei mode particles from DPF equipped vehicles are predominantly sulfates
(Grose et al., 2006). However, significant levels of TC were also present in these
vehicles. During the analysis of the physical PM properties, we observed that nucleation
is suppressed in CCRT and Horizon, which have both a higher mass fraction of total
carbon concentrations in all size ranges and less sulfate in nucleation mode PM. Unlike
the SCRT cruise cycles, the UDDS cycles for Z-SCRT are unique in the sense that the
emission rates of TC dominate those of sulfate. This may be again due to higher

36

activation temperature of Zeolite catalysts, which may impede the conversion of SO
2
to
sulfate during transient cycles, for which the exhaust temperature fluctuates continuously.
Herner et al., (2007b) reported that the critical temperatures (post after-treatment
temperatures) required for Z-SCRT and V-SCRT to trigger nucleation are 373
o
C and 330
o
C, respectively.

Figure 2.7: Emission rates of size-resolved PM chemical species. a) Cruise 50 mph, b)
Transient UDDS c) Baseline vehicle
Note: Nucleation (Nuc, 10-56 nm), UF: 56-180 nm and Acc: 180 nm – 2.5 μm.

Cruise 50 mph
Particle Size Range
Emission Factor (mg km
-1
)
0.0
0.5
1.0
1.5
2.0
2.5
3.0
CRT
®
Z-SCRT
®
DPX
Horizon
Nuc UF Acc Nuc UF Acc Nuc UF Acc Nuc UF Acc

a)

37

Figure 2.7 (Continued)
UDDS
Particle Size Range
Emission Factor (mg km
-1
)
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
9.0
CRT
®
ZSCRT
®
DPX Horizon VSCRT
®
CCRT
®
Nuc UF Acc Nuc UF Acc Nuc UF Acc Nuc UF Acc Nuc UF Acc Nuc UF Acc

Baseline
Particle Size Range
Em ission Factor (m g km
-1
)
0
10
20
30
40
50
60
70
80
90
100
UDDS
Crusie 50 mph
Nuc UF Acc Nuc UF Acc

Sodium Ammonium Potassium Chloride
Nitrate Sulfate TC
b)
c)

38

Figure 2.7c shows the mass size distribution of the chemical components of PM
emissions from the baseline vehicle. Total carbon concentration outweighs any other
species for the baseline vehicle (>99% of total PM mass concentration). Further, the
UDDS cycle of Baseline vehicle emits almost 3-fold higher TC than the cruise cycle.
This can be explained by the shift of TC peaks during cruise cycle in the 100-180nm
(25mg km
-1
) size rage to the even larger accumulation mode (180-2500nm, 90mg km
-1
)
during UDDS cycle. The higher emission of TC during the transient cycles, however,
was not observed for Z-SCRT and Horizon (Figure 2.7a, b). The Z-SCRT-UDDS cycle
emitted the least amount of PM mass among all the retrofitted cycles of Kenworth truck
(CRT; SCRT cycles). This particular cycle also lacked a pronounced nucleation mode
which was observed in CRT and other SCRT cycles. Higher activation temperatures for
the Zeolite catalysts can inhibit the nucleation process (as mentioned before) especially
during the fluctuating exhaust temperature of the UDDS cycle. Since a considerable
fraction of TC (mostly OC for retrofitted vehicles) is associated with the nucleation
mode, it is likely that the lack of nucleation can retard condensation of organics on these
particles. Emissions of TC from both the Horizon cycles are low and comparable to the
analytical limits of detection; therefore caution should be exercised in making inferences
based on these data.

2.3.6 DTT Activity
Blank-subtracted (filter blank DTT~0-0.0005 n-mole µg
-1
min
-1
) DTT consumption rates,
normalized per unit mass of PM are reported in Figure 2.8 a,b,c for various vehicle

39

configurations. The vehicle with the Horizon trap had the highest per mass DTT activity
(0.16-0.19 n-mole µg
-1
min
-1
for cruise and UDDS) irrespective of driving conditions.
The DTT consumption rates from both Vanadium and Zeolite-based SCRTs are on the
same order of magnitude (0.01-0.02 n-mole µg
-1
min
-1
). It is interesting to note that when
the selective catalytic reduction (SCR) section from SCRT is removed from the exhaust
stream and the vehicle is operated with only CRT (i.e., the DOC+ uncatalyzed filter), the
DTT activity increased by a factor of almost 3, both for cruise and UDDS cycles. The
CCRT (Catalyzed filter + DOC) is the most efficient among the test fleet with DTT rates
as low as 0.006 n-mole µg
-1
min
-1
. Of particular note is the elevated level of activity for
the DPX and Horizon vehicles. The baseline vehicle emitted PM with similar oxidative
characteristics to those of newer vehicles with control technologies.

Figure 2.8 also reports the DTT values for the thermo denuded filters. There is a
significant reduction (50-100%) in the activity as particles are heated to 150
o
C and their
semi-volatile component is removed. The baseline vehicle, however, did not show any
alteration in DTT response between denuded and undenuded exhaust stream, except
while idling. This is because of the highly refractory nature (mostly soot) of these
particles.

40

Figure 2.8 a, b, c: Oxidative Potential (DTT consumption in n-moles min
-1
μg
-1
of PM )
of thermo-denuded and undenuded PM
Cruise
Baseline
Baseline-TD150
CRT
CRT-TD150
V-SCRT
V-SCRT-TD150
Z-SCRT
Z-SCRT-TD150
CCRT
CCRT-TD150
DPX
DPX-TD150
Horizon
Horizon-TD150
DTT (n-mole µg
-1
min
-1
)
0.00
0.02
0.04
0.16
0.18
0.20

UDDS
B aselin e
B aselin e-T D 150
CRT
CRT-TD 150
V-SCRT
V-SC R T -T D 1 50
Z -SCRT
Z-S C R T -T D 150
CCRT
CCR T -T D 1 5 0
DPX
DPX T D 1 50
Ho rizo n
H o rizo n-TD 150
DTT (n-m ole µg
-1
min
-1
)
0.00
0.02
0.04
0.06
0.08
0.16
0.18

NA
a)
b)
Undenuded
Undenuded
Thermo denuded (150
o
C)
Thermo denuded (150
o
C)

41

Figure 2.8 (Continued)
Idle
B aselin e-Id le
B aselin e-T D 150
B aselin e-T D 230
DP X
D P X -T D 230
Ho riz o n
H o riz o n -T D 150
D TT (n -m o le µg
-1
min
-1
)
0.00
0.02
0.04
0.06
0.08
0.10

The increased DTT activity of the semi-volatile PM fraction is further highlighted by
comparing with their corresponding semi-volatile PM mass and number fractions (Figure
2.9). For vehicles with (soot removing) control devices, the semi-volatile fraction
contributed roughly 70-100% of the net oxidative activity compared to 20-30% for
baseline vehicle. The sole exception is the CCRT vehicle, which hardly produced any
oxidative activity from its semi-volatile fraction, despite emitting 45% of semi-volatile
PM by mass and 37% by number. For this vehicle (CCRT) the relatively low level of OP
has originated from its residual non-volatile fractions. In general, particle number-based
c)
Undenuded
Thermo denuded (150
o
C)
Thermo denuded (230
o
C)

42

volatility, which is predominantly driven by the evaporation of the sub-50 nm, so-called
nucleation mode particles (Kittelson, 1998), followed the activity trends better than the
PM mass volatility.
Figure 2.9: Relationship between oxidative activity and semivolatile PM fraction at 150
°C (Showing good correlation of DTT (filled circles) with particle number (hollow
diamonds) and poor correlation with PM mass (bars)).

Baseline-Cruise
Baseline-UDDS
Baseline-Idle
CRT-Cruise
CRT-UDDS
V-SCRT-Cruise
V-SCRT-UDDS
Z-SCRT-Cruise
Z-SCRT-UDDS
CCRT-UDDS
DPX-Cruise
DPX-UDDS
DPX-Idle (230 oC)
Horizon-Cruise
Horizon-UDDS
Horizon-Idle
Oxidative Activity from Semi-Volatile Fraction
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Semi-Volatile PM Mass /Number Fraction
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Oxidative Activity from Semi-Volatile Fraction
Semi-Volatile PM Mass Fraction
Semi-Volatile PM Number Fraction

Total Oxidative Activity (1): Sum of oxidative activity from semi-volatile
and non-volatile fraction

43

The DTT results (Figure 2.10 a,b,c) expressed per unit vehicle distance traveled provide a
better idea of the total oxidative load imparted on the environment by these vehicles.
They also provide a quantitative assessment of the effectiveness of these after treatment
devices. While we observed a comparable or lower oxidative response of the baseline
vehicle per unit mass of PM, its net DTT consumption (per km) is substantially higher
(>3 times) than retrofitted vehicles.

Figure 2.10 a,b,c: DTT consumption per unit distance ( in n-moles min
-1
km
-1
) traveled
by vehicles for Cruise and UDDS cycles and per hour( in n-moles min
-1
hr
-1
) for Idle

Cruise
Baseline
CRT
V-SCRT
Z-SCRT
CCRT
DPX
Horizon
DTT(n-mole min
-1
km
-1
)
0
300
600
900
1200
1500
DTT(n-mole min
-1
km
-1
)
0
100
200
300
400
500

NA
a)

44

Figure 2.10 (Continued)
UDDS
Baseline
CRT
V-SCRT
Z-SCRT
CCRT
DPX
Horizon
DTT(n-m ole m in
-1
km
-1
)
0
1500
3000
4500
6000
7500
DTT(n-m ole m in
-1
km
-1
)
0
500
1000
1500
2000
2500

Idle
Baseline
DPX
Horizon
DTT(n-mole min
-1
hr
-1
)
0
30000
60000
90000
120000
150000
DTT(n-mole min
-1
hr
-1
)
0
6000
12000
18000
24000
30000

b)
c)

45

2.3.7 Water Soluble Organic Carbon and DTT
Figure 2.11 shows the WSOC (% of PM mass) and DTT consumption (per unit PM mass)
for vehicle-configurations. DTT and WSOC followed the same trend, irrespective of
driving condition or control devices. The school bus (Horizon), which exhibited the
highest oxidative activity per unit mass, also emitted the highest fraction of soluble
organics (35-83%). Similarly higher WSOC content was found for the higher DTT values
of PM produced by the DPX-idle cycle.

Figure 2.11: DTT consumption (n-moles min
-1
μg
-1
of PM) in relation to the water-
soluble organic carbon (WSOC) of PM from the exhaust

0
0.04
0.08
0.12
0.16
0.2
Baseline-
Cruise
Baseline-
Cruise-TD150*
Baseline-UDDS
Baseline-
UDDS-TD150*
V-SCRT-Cruise
V-SCRT-UDDS
Z-SCRT-Cruise
Z-SCRT-
Cruise-TD150*
Z-SCRT-UDDS
CCRT-UDDS
DPX-Cruise
DPX-UDDS
Horizon-Cruise
Horizon-UDDS
DTT (n-mole min
-1
µg
-1
)
0
10
20
30
40
50
60
70
80
90
WSOC (% of PM mass)
DTT
WSOC
y = 0.0026x
R
2
= 0.9
0
0.05
0.1
0.15
0.2
0.25
0 20 40 60 80 100
WSOC (% of PM mass)
DTT(n-mole min
-1
µg
-1
)
46

2.3.8 Correlation between Chemical Species and DTT Activity
To assess the contribution of various species to DTT activity, the Pearson’s coefficient
(R) and the associated significance levels (p-value) have been calculated and are shown
in Table 2.3 for the correlation of different chemical constituents with DTT values. The
DTT was significantly correlated with WSOC (R=0.94, p<0.01) and organic acids (sum,
R=0.91; p<0.01) and moderately correlated with organic carbon (R=0.76, P=0.02). All
other species, including inorganic ions, PAHs, EC, alkanes either have low or negative
correlation. No correlation analysis was performed between metals and DTT. Although
some of the transition metals (such as Cu) can actively participate in DHBA assay, their
contributions are insignificant for DTT tests (Cho et al., 2005).

Table 2.3: Correlation coefficient (R) and significance level (p) for the DTT activity and
selected chemical species

Species R P
EC
-0.35
0.37
OC
0.76
0.02
NO
3
-

-0.09
0.77
SO
4
2-

-0.32
0.27
NH
4
-

-0.25
0.26
K
+

0.43
0.20
Cl
-

0.34
0.15
WSOC
0.94
<0.01
Alkanes (Alk.)
0.03
0.54
PAHs
-0.26
0.75
Organic Acids (OA)
0.91
<0.01

47

2.4 Conclusions
This investigation presents some of the first detailed particle characterizations for
advanced NOx and PM retrofits in heavy duty diesel vehicles. The test fleet includes a
diesel hybrid electric with a DPF (Hybrid-CCRT), a continuously regenerating
technology (CRT), a vanadium-based SCR catalyst with a CRT

(V-SCRT), a Zeolite-
based SCR catalyst with a CRT (Z-SCRT), a DPX, and an EPF (Horizon/School Bus). A
HDDV vehicle without emission controls served as the baseline vehicle. The fleet was
tested under three driving cycles: cruise at 50 mph, UDDS and idle.

While comparing with a baseline vehicle, significant reduction (>90%) in mass, EC, OC,
WSOC emissions is achieved for vehicles with retrofits. However, enhanced nucleation
mode particles were observed for some of the vehicles especially during cruise cycles.
The Hybrid-CCRT and EPF vehicles were efficient in controlling both mass and number
emissions. The vehicles with significant nucleation (CRT, V-SCRT, Z-SCRT and DPX)
mode particles produced considerable amount of sulfates especially during steady state
operations. On the contrary, the non-nucleating configurations (Horizon, CCRT, Z-
SCRT-UDDS) were associated with higher amount of total carbon in the form of OC.
Soluble fraction of OC was highest for Horizon followed by SCRTs, DPX and baseline.

Despite a general increase in the intrinsic DTT activity (per mass basis) of exhaust PM
with use of most control technologies, the overall activity (expressed per km or per hr)
was substantially reduced for retrofitted configurations compared to the baseline vehicle.

48

The semi-volatile fraction of the exhaust particles was observed to be highly oxidative in
nature as demonstrated by a significant reduction in DTT activity (by 50-100%) observed
for thermally-denuded PM. Correlation analysis performed between all the species,
showed that DTT is moderately associated (R=0.76) with organic carbon (OC) and
strongly associated (R=0.94) with the water soluble organic carbon (WSOC).

49

Chapter 3: Toxicity Profiles of Ambient Particles Influenced
by Vehicular, Woodsmoke and Secondary Photochemical
Sources

This chapter is divided into two sections, describing the oxidative activity profiles of
ambient PM from vehicular and secondary photochemical sources (section 3.1), and from
woodsmoke sources (wildfires; section 3.2). The major findings of both the studies have
been summarized in the respective sections.

3.1 Oxidative Activity Profiles of Ambient Particulate Matter from Primary and
Secondary Sources

3.1.1 Introduction
Atmospheric quasi-ultrafine particles (quasi-UFP) originate from two broad categories of
sources: primary and secondary. Primary particles are directly emitted from combustion
sources, including heavy and light duty vehicles, woodsmoke, and industries including,
for example, power plants. In the presence of various atmospheric oxidants, primary
particles may undergo photochemical processing yielding secondary particles with
distinctly different physical and chemical characteristics compared to their precursor
primary particles. Diurnal changes in physical properties (particle number size
distribution and volatility) of the quasi-ultrafine particles at an urban location near
downtown Los Angeles have been discussed in detail by Moore et al., (2007). In a
companion paper, the chemical composition of ambient samples collected in two
different time periods (morning and afternoon) at the same site was also presented (Ning

50

et al., 2007). Results from both studies indicated a strong influence of commute traffic
emissions during the morning hours as evidenced by elevated concentrations of nitrogen
oxides (NO
x
), carbon monoxide (CO), black carbon (BC) (Moore et al., 2007), alkanes,
PAHs and hopanes (Ning et al., 2007). By contrast, the afternoon concentrations of
oxygenated organic acids and sulfate rose, while other species were diluted by the
increased mixing height or lost due to increasing temperature. These are clear indicators
that secondary photochemical reactions are a major source of quasi-ultrafine aerosols in
the afternoon.

The reaction mechanisms leading to the generation of secondary organic aerosols (SOA)
have not been fully characterized due to the complex nature of organic compounds in the
atmosphere. The differences in the physico-chemical characteristics of primary and
secondary particles will likely lead to differences in their oxidative activities including
their ability to induce cellular oxidative stress (Baltensperger et al., 2008; Delfino et al.,
2009; Ntziachristos et al., 2007). Recent laboratory experiments also indicate that
combining primary diesel particles with ozone (in order to mimic atmospheric
photochemical transformation processes) substantially enhances the oxidative activity of
the PM (Li et al., 2009); however, there are very few studies investigating this important
issue using real world atmospheric aerosols.

The present section of the chapter focuses on comparing the oxidative activity profiles of
the quasi-ultrafine particles (<180 nm in this study) at an urban site near downtown Los

51

Angeles collected during two different time periods of the day – morning and afternoon,
representing particles characterized by primary emissions and secondary aerosols
formation, respectively. The oxidative activity of the collected particles has been
measured by two independent assays: 1) the DTT assay, and 2) the macrophage ROS
assay. Detailed chemical analyses of the collected samples including water soluble
elements, inorganic ions, organic species and water soluble organic carbon have been
conducted to distinguish the chemical properties of the source specific particles, and also
to investigate their correlation with the measured oxidative activities. Our goal was to
contrast the toxicity profiles of ambient quasi-ultrafine particles emitted from primary
vehicular sources with those produced from secondary processes and atmospheric aging.

3.1.2 Experimental Methods
3.1.2.1 Sampling Location
This study was conducted at the Particle Instrumentation Unit (PIU) of Southern
California Particle Center (SCPC) on the campus of University of Southern California
(USC) near downtown Los Angeles. The sampling site is located approximately 130 m to
the NE of the I-110 freeway, near construction, parking facilities and other nominal
sources of pollutant emissions. Pollutant sources near this site are predominantly from
vehicular sources with traffic patterns consistent with a mixed-use urban environment.
Light duty gasoline vehicles constitute the major fraction (>93%) of traffic at the stretch
of freeway near the site, with the rest being heavy duty diesel vehicles. Studies indicate
that the average monthly PM chemical composition in the quasi-ultrafine mode at the

52

USC and other similar Los Angeles-area sites is dominated by organic carbon, which
varies from 40 – 90% depending upon the season and location (Fine et al., 2004; Hughes
et al., 1998; Sardar et al., 2005). Additional sampling site information including the
results of earlier investigations of the physical and chemical properties of the PM
observed at the USC site can be found in Ning et al., (2007) and Fine et al., (2004).

3.1.2.2 Sampling Protocol
Sampling was conducted between June and August, 2008 over a period of 10 consecutive
weeks (excluding weekends). Quasi-ultrafine particles (D
p
<180 nm) were collected using
a low pressure drop impactor on Zefluor PTFE Membrane Filters (Pall Life Sciences, 3
µm, 8"x10", 28139-597) seated in a specially designed 20x25 cm high volume sampler
(HIQ Environmental Products Co., CA). The impactor has been designed primarily as a
separator of quasi-ultrafine particles from accumulation mode (0.18 µm ≤ Dp ≤2.5 µm)
particles at a flow rate of 400 L/min. The design and operating performance of the multi-
slit impactor have been described in detail previously (Misra et al., 2002). Time
integrated samples were collected during both “morning” (6:00-9:00 PDT) and
“afternoon” (11:00-14:00 PDT) time periods and were grouped into three sample sets
(S1, S2, S3). There were 18 days total for the S1 sampling set, while both the S2 and S3
sets were collected for 16 days each. As the samples were composited for each grouping
period (i.e. particles were collected on the same filter); the total number of samples for
the entire campaign was 6 - two samples (morning and afternoon) for each grouping
period (S1, S2 and S3). Thus, each sample of set S1 encompassed 54 hrs (18 days

53

multiplied by 3 hrs per day), while 48 hrs (16 days multiplied by 3 hrs per day) for each
sample of sets S2 and S3. The “morning” period corresponds to rush hour traffic when
the ambient aerosols at the sampling site are dominated by primary particles freshly
emitted from vehicles on the nearby freeway (Moore et al., 2007). The “afternoon” period
represents the mixture of primary and secondary particles undergoing physical and
chemical changes (i.e., photo-oxidation, volatilization, dilution and possibly re-
suspension) (Moore et al., 2007). The strong influence of secondary photochemical
sources during the afternoon at the sampling site is evidenced by the concentration of
ozone [(O
3
); averaged over three time periods, i.e., S1, S2 and S3], which increased from
9±4 ppb in the morning period to 50±10 ppb in the afternoon period. Meteorological
conditions were quite consistent during the entire study period, with higher temperature
and lower humidity in afternoon (28 ±2
o
C and 44 ±5 %) compared to the morning period
(20 ±1
o
C and 70 ±5 %). The wind direction was mostly from south-west at the site. A
Scanning Mobility Particle Sizer (SMPS Model 3080 and DMA Model 3081,
condensation particle counter Model 3022A, TSI Inc., St. Paul., MN) was also set up
concurrently to measure the particle size distribution. The sampling flow rate was 1.5 L
/min, with a scan up time of 150 s. The minimum and maximum sizes detectable at these
settings are 7.64 and 225 nm, respectively. The mass concentrations of quasi-ultrafine
particles in both the morning and afternoon periods were obtained using a parallel
MOUDI (Micro-Orifice Uniform Deposit Impactor; MSP Corporation, Minneapolis,
MN) measurement. The mass concentration of quasi-ultrafine particles measured by the

54

MOUDI and the multi-slit impactor has been demonstrated to be in agreement (R
2
>0.85)
(Misra et al., 2002).

3.1.2.3 Sample Analysis
All filter samples, along with field blanks and lab blanks, were extracted in high purity
deionized water at ambient temperature using a shaking table. The water extracts were
analyzed for water soluble organic carbon (WSOC), water soluble ions, and water soluble
trace elements. Water soluble ions were analyzed using two Ion Chromatography (IC)
analyses: 1) analysis for cations; and 2) analysis for anions (Zhang et al., 2008). Trace
elements were analyzed from the water extracts using a high resolution Inductively
Coupled Plasma-Mass Spectroscopy (ICP-MS) as described by Zhang et al., (2008).
Water soluble organic carbon (WSOC) for all the samples was analyzed using a
Shimadzu TOC-5000A liquid analyzer (Decesari et al., 2001). A separate section of the
sample, field blank and lab blank filters was extracted with a solvent mixture of
methylene chloride and methanol using Soxhelet extraction. The extracts were
concentrated before gas chromatography/mass spectrometry analysis as described by
Stone et al., (2008). All chemical analyses were conducted with check standards and
multipoint calibration curves to assure accurate quantification. All results were blank
subtracted using the field blanks processed and analyzed with the samples.

The oxidative activity of the collected particles was quantified by two different assays: 1)
consumption of dithiothreitol in a cell-free system (DTT assay), and 2) macrophage ROS

55

- in vitro exposure to rat Alveolar Macrophage (AM) cells using Dichlorofluorescin
Diacetate (DCFH-DA) as the ROS probe. High purity deionized water was used for the
filter extraction technique for both the DTT and ROS assays.

3.1.3 Results and Discussion
3.1.3.1 Physical Parameters
Figure 3.1a shows the hourly average particle number concentration in the size range
from 7.64 nm to 180 nm, measured by the SMPS during the sampling campaign. As
shown in the figure, the diurnal profile displays a shoulder in the morning (6:00 to 9:00
AM) to the distinguishable peak in the afternoon (11:00 AM to 2:00 PM). This represents
emissions from morning rush hour traffic and secondary photo-chemical reactions
occurring in the afternoon at the site, respectively (Moore et al., 2007). Average particle
number concentrations in the afternoon are 1.6±0.1 times those in the morning,
suggesting secondary particle formation consistent with the observations by Moore et al.,
(2007) at the same site. Figure 3.1b shows the average particle number size distribution
(dN/dlogDp) measured during the morning and afternoon periods, respectively. The size
distributions in both the morning and afternoon are mono-modal; however, the afternoon
period is distinguished by a smaller peak mode diameter (28.9±1.1 nm) compared to the
morning period (peak mode diameter = 41.5±1.5 nm). The larger peak mode diameter in
the morning period may be due to the combined effects of condensation of organic vapors
co-emitted from traffic sources onto freshly emitted ultrafine PM, which would be result
of the lower temperatures and atmospheric dilution that prevail during this time period,

56

and possibly to coagulation of highly concentrated nanoparticles. Both of these processes
will enhance the growth of the particles emitted from vehicular sources compared to the
relatively freshly formed secondary aerosols in the afternoon period (Zhu et al., 2002).
Average mass concentrations of quasi-ultrafine particles were also higher in the afternoon
period (5.7±2.6 μg/m
3
) compared to 3.8 ±1.7 μg/m
3
in the morning period, indicating the
additional contribution of secondary aerosol formation to the total PM0.18 (particles with
aerodynamic diameter ≤ 180 nm) mass.

57

Figure 3.1: a) Diurnal profile of the average particle number concentration of quasi-
ultrafine particles measured at the USC sampling site and (b) the average particle number
size distribution of quasi-ultrafine particles in “morning” and “afternoon” periods.

Figure 3.1a

Figure 3.1b

Afternoon
Peak Mode Diameter = 28.9 nm
Total Conc. = 17.2x10
3
particles/cm
3

Morning
Peak Mode Diameter =
41.5 nm
Total Conc. = 11.0x10
3
particles/cm
3

58

3.1.3.2 Chemical Parameters
Figure 3.2 shows the average mass concentration of major water soluble inorganic ions
[sulfate (SO
4
2-
), ammonium (NH
4
+
), and nitrate (NO
3
-
)] in quasi-ultrafine particles
measured at the sampling site in the morning and afternoon periods.

Figure 3.2: Average concentration of the major inorganic ions in quasi-ultrafine particles
collected in the morning and afternoon periods at the USC sampling site.

Concentrations of these ionic species are higher in the afternoon period compared to
those in the morning, with afternoon-to-morning ratios of 1.76±0.38, 1.60±0.04 and
Nitrate Sulfate Ammonium
μg/m
3
0.0
0.5
1.0
1.5
2.0
2.5
3.0
Morning
Afternoon

59

2.11±0.61 for NH
4
+
, NO
3
-
and SO
4
2-
, respectively. This increase in concentrations in the
afternoon period may be attributed to the oxidation of precursor gases (e.g. ammonia,
sulfur dioxide and nitric oxide) by photo-chemical byproducts [e.g., ozone (O
3
), hydroxyl
free radical (HO
.
), hydroperoxyl radical (HO
2
.
)] followed by nucleation and/or
condensation onto pre-existing particles (Ning et al., 2007). Following the approach by
Zhang et al., (2004a), particle acidity can be expressed as the ratio of sum of measured
ammonium and sodium concentrations to the concentrations needed to fully neutralize
the measured sulfate, nitrate and chloride, with a value of unity suggesting the full
neutralization and a value close to zero suggesting an acidic aerosol. The ratios measured
during the morning and afternoon periods are 0.93±0.03 and 0.82±0.01, respectively,
thereby indicating an 11±3 % increase in particle acidity in the afternoon period. This is
consistent with the previous studies demonstrating higher acidity of the ultrafine particles
[measured to fully neutralizing ammonium ratio of ~0.80; (Zhang et al., 2004a)] in the
initial stage of photochemically induced nucleation.

Table 3.1 shows the comparison of the mass concentrations (average and standard
deviation) of water-soluble trace elements and metals in the quasi-ultrafine particles in
the morning and afternoon. A total of twenty elements have been quantified as shown in
the table. Sulfur (S), Aluminum (Al), Calcium (Ca), Iron (Fe) and Sodium (Na) are the
dominant species in the ambient atmosphere both in the morning and afternoon periods.
The concentration of certain metals [Vanadium (V), Chromium (Cr), Nickel (Ni), Copper
(Cu), Arsenic (As), Cadmium (Cd) and Zinc (Zn)] was higher in the morning period.

60

These heavy metals are known to be emitted from vehicles from both fuel combustion
and lube oil emissions (Geller et al., 2006). Cu originates from vehicle brake abrasion
(Garg et al., 2000; Sanders et al., 2003; Sternbeck et al., 2002) and Zn is mostly a product
of tire attrition (Singh et al., 2002). Elevated levels of these metals in the morning period
suggest traffic as the primary source of particles during the morning hours at the
sampling site. Other metals such as Na, Magnesium (Mg), Phosphorus (P), Al, Ca,
Potassium (K), Titanium (Ti), Manganese (Mn), Fe, Cobalt (Co) and Barium (Ba)
displayed a different pattern with higher values in the afternoon period. These metals also
exhibited a high degree of correlation amongst themselves (R>0.70, not shown) and
probably originate from re-suspension of road dust, which is higher in the afternoon due
to the higher wind speed (3.5±0.3 m/s compared to 0.9±0.4 m/s in the morning period),
lower relative humidity (44±5 % compared to 70±5% in the morning period) and
possibly higher vehicle speeds on the nearby streets and freeways, compared to the
congested morning traffic hours. Although, the majority of these species are expected to
be partitioned in the coarse PM (PM
10-2.5
), there is a significant fraction in the sub-200
nm range, as demonstrated by earlier investigations in urban areas (Garg et al., 2000).

61

Table 3.1: Average concentrations of water-soluble elements in ambient quasi-ultrafine
particles measured by ICP-MS at the USC sampling site during morning and afternoon
periods

Metal
Morning Afternoon Ratio
(Morning/
Afternoon)
Average
Conc., ng/m
3
S.D.
Average
Conc., ng/m
3
S.D.
Na 57.82 24.98 195.59 47.54 0.30
Mg 17.65 8.16 64.36 6.93 0.27
Al 112.55 42.20 311.37 34.95 0.36
P 3.81 6.00 10.44 3.36 0.36
Ca 110.31 127.09 198.15 50.86 0.56
S 472.53 311.01 578.42 130.74 0.82
K 22.39 17.77 82.59 34.39 0.27
Ti 7.02 3.36 24.43 6.77 0.29
V 5.78 4.72 5.50 1.43 1.05
Cr 1.35 1.24 0.55 0.49 2.46
Mn 2.34 1.30 4.73 1.19 0.49
Fe 106.94 62.39 286.21 78.88 0.37
Co 0.09 0.05 0.18 0.03 0.49
Ni 3.09 2.46 2.12 0.75 1.46
Cu 16.26 18.61 10.75 6.13 1.51
Zn 20.26 13.06 14.87 4.42 1.36
As 0.22 0.13 0.19 0.07 1.15
Cd 0.06 0.04 0.05 0.01 1.21
Ba 3.06 1.68 10.49 0.87 0.29
Pb 3.72 2.83 4.38 1.61 0.85

Figure 3.3a shows the average mass concentration of quasi-UFP WSOC and CO in the
morning and afternoon periods. CO is an inert gas species often used as a marker of
vehicular emissions (Zhu et al., 2002) and atmospheric dilution (Westerdahl et al., 2009),
while WSOC consists of a variety of polar organic compounds and is regarded as an
indicator of secondary organic aerosols (SOA) (Docherty et al., 2008). As shown in the

62

figure, the CO concentration decreased by a factor of 2.0±0.1 in the afternoon, as a result
of a decrease in vehicular emissions coupled with an increase in the mixing height and
atmospheric dilution in the afternoon period. However, despite the 2-fold increase in
atmospheric dilution suggested by the CO concentrations, the average concentration of
WSOC was elevated by 2.5±0.9 times compared to the morning period. We interpret this
to indicate the formation of SOA driven by the oxygenation of particles with aging in the
atmosphere during photochemical reactions in the afternoon.

Figure 3.3: Average concentration of the various organic species measured at the USC
sampling site during the morning and afternoon periods a) Water Soluble Organic Carbon
and Carbon Monoxide, b) Polycyclic Aromatic Hydrocarbons (PAHs), c) Hopanes and
Steranes, and d) Alkanes and Organic Acids.

WSOC CO
μg/m
3
(WSOC), ppm (CO)
0.0
0.2
0.4
0.6
0.8
1.0
a)

63

Figure 3.3 (Continued)

Phenanthrene
Anthracene
Fluoranthene
Acephenanthrylene
Pyrene
Benzo(ghi)fluoranthene
Cyclopenta(cd)pyrene
Benz(a)anthracene
Chrysene
1-Methylchrysene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(j)fluoranthene
Benzo(e)pyrene
Benzo(a)pyrene
Perylene
Indeno(1,2,3-cd)pyrene
Benzo(ghi)perylene
Dibenz(ah)anthracene
Coronene
ng/m
3
0.00
0.05
0.10
0.15
0.20
0.25
0.30
0.35
b)
c)

64

Figure 3.3 (Continued)

Figure 3.3 also shows the average quasi-UFP mass concentrations of speciated organic
compounds [PAHs, (Figure 3.3b); hopanes and steranes, (Figure 3.3c); and alkanes and
organic acids, (Figure 3.3d)] in the morning and afternoon periods. Most of the PAHs,
hopanes, steranes, and alkanes were higher in the morning period compared to the
afternoon period, attributed to their origin from vehicular sources (Ning et al., 2007).
However, morning-to-afternoon ratio for certain species (particularly PAHs and alkanes)
is relatively higher [up to 3.5±0.9 as for benzo (GHI) perylene], compared to that for CO
(morning-to-afternoon ratio = 2.0±0.1; Figure 3a). The relatively larger decrease in the
Pentadecane
Hexadecane
Heptadecane
Octadecane
Nonadecane
Eicosane
Heneicosane
Docosane
Tricosane
Tetracosane
Pentacosane
Hexacosane
Heptacosane
Octacosane
Nonacosane
Triacontane
Hentriacontane
Dotriacontane
Tritriacontane
Tetratriacontane
Pentatriacontane
Hexatriacontane
Pentadecylcyclohexane
Hexadecylcyclohexane
Heptadecylcyclohexane
Octadecylcyclohexane
Nonadecylcyclohexane
Octanoic acid
Decanoic acid
Dodecanoic acid
Tetradecanoic acid
Pentadecanoic acid
Hexadecanoic acid
Heptadecanoic acid
Octadecanoic acid
ng/m
3
0
2
4
6
8
10
12
20
40
60
80
100
Morning
Afternoon
Alkanes
Organic
Acids
d)

65

concentration of PAHs and alkanes compared to CO in the afternoon indicates their
possible volatilization and photo-oxidation in addition to dilution. Interestingly, in the
fine size range (PM
2.5
), concentrations of these organic species (PAHs and alkanes) were
higher in the afternoon than the morning period. We hypothesize that volatilization of
these species from the ultrafine particles in afternoon period and subsequent re-
condensation onto the larger particles might be responsible for this trend (Venkataraman
et al., 1999). Similar observations on the partitioning of PAHs between ultrafine and
accumulation PM modes with increasing ambient temperature have also been made in
other studies (Miguel and Friedlander, 1978; Pierce and Katz, 1975).

By contrast, concentrations of most of the organic acids were higher in the afternoon
period (afternoon-to-morning ratio ranges from 1.2 to 3.0). These acids are probably the
products of photo-chemical oxidation of either organic gases or semi-volatile species
evaporating from the primary particles (Zhang et al., 2004b). Food cooking and
vegetative detritus emissions might be included into other sources of some organic acids,
but the sampling site is not substantially impacted by these sources, which supports the
notion that atmospheric chemistry is likely responsible for the increased concentrations of
these acids in the afternoon. The afternoon-to-morning ratio was higher (up to 3) for high
carbon number organic acids (C
15
-C
29,
e.g., Octadecanoic Acid, Pentadecanoic Acid)
compared to the low carbon number acids (e.g. Octanoic Acid, Decanoic Acid; ratio
~1.5), probably due to the higher volatility of organic species associated with decrease in
carbon number (Ning et al., 2007).

66

3.1.3.3 Oxidative Activity
The oxidative activity of the quasi-ultrafine particles was measured by two different
methods as described earlier - the DTT and macrophage ROS assays. Figure 3.4 shows
the results of the DTT assay for the three samples sets (S1, S2, and S3) and their average
values in the morning and afternoon periods. The results have been expressed as mass
(of PM) based activity (Figure 3.4a) in nmol.min
-1
.µg
-1
and also volume (of air) based
activity (Figure 3.4b) in nmol.min
-1
.m
-3
, respectively. As shown in Figure 3.4a, the mass
based DTT activity of the quasi-UFP is consistently higher in the afternoon period for all
of the sample sets with an average afternoon-to-morning ratio of 1.8±0.7. The DTT assay
measures the capability of particles to generate superoxide radicals (O
2
*-
) in their
interaction with the thiol group compounds, which are the main antioxidants present in
endothelial cells (Foresti et al., 1997). Although this assay is insensitive to hydroxyl free
radicals (HO
.
), O
2
*-
produced in the above reaction can play a role in the generation of
HO
.
also, by regenerating the precursors of Fenton’s reaction (e.g. Fe
+2
and H
2
O
2
)
(Schoonen et al., 2006). In a recent experimental study, Li et al., (2009) demonstrated
higher oxidative activity measured by the DTT assay of diesel exhaust particles (DEPs)
aged with O
3
compared to the freshly emitted diesel PM. Our results showing the
increased DTT activity of the quasi-ultrafine particles in afternoon period are very
consistent with the work described earlier by Li et al., (2009), and lead to one of the most
important findings of this study – compounds associated with secondary aerosols
produced by photochemical processes in the afternoon period possess higher oxidative

67

activity, and thus are more capable of generating free radicals and causing cell damage in
biological systems.

As shown in Figure 3.4b, the DTT activity per m
3
of air is also consistently higher for the
afternoon samples, with an average afternoon-to-morning ratio of 2.8±1.1 (compared to
1.8±0.7 for the mass based activity). The difference between the volume based and mass
based DTT activity may have important implications with respect to the assessment of
public exposure and the relative PM toxicity. While the mass-based oxidative activity is
appropriate for comparing the relative PM toxicity from different sources, the volume-
based activities become important in the context of the overall public exposure and the
associated risks of individual contributions from various sources to ambient PM
concentrations.

68

Figure 3.4: Oxidative activity (expressed as nmol of DTT/min) of quasi-ultrafine particles
collected at the USC sampling site during the morning and afternoon periods; a) per µg of
particulate mass and b) per m
3
of volume of air.

Figure 3.4a

Figure 3.4b

69

Figure 3.5 shows the results of macrophage ROS assay in the morning and afternoon
periods, expressed as µg of Zymosan units per mg of PM mass (a), and per m
3
of air (b)
on the collected quasi-UFP. There was significant variation in the ROS activity for the
three sample sets (S1, S2 and S3) and the afternoon-to-morning ratio averaged over the
three grouped periods was 1.2±1.1. Although the ROS activity of sample sets, S2 and S3
was higher in afternoon period, very high activity was observed for the morning period of
sample set S1. The macrophage ROS assay measures the generation of reactive oxygen
species by its reaction with DCFH. The ROS probe used in this method is sensitive to a
broad spectrum of oxidants including HO
.
, hydrogen peroxide (H
2
O
2
)

and nitric oxides
(NO
.
) (Schoonen et al., 2006). The inconsistent trend in ROS activity for the morning and
afternoon samples indicates that, although secondary organic compounds formed in
afternoon period may contribute to the formation of HO
.
in biological cells, the overall
capability of PM to generate free radicals is also dependent on species emitted from
primary sources in the morning period. The following section will further elaborate on
the relationship between the PM chemical constituents and free radicals generation.

70

Figure 3.5: Oxidative activity (expressed as the ROS response, or µg of Zymosan units)
of quasi-ultrafine particles collected at the USC sampling site during the morning and
afternoon periods; a) per mg of particulate mass and b) per m
3
of volume of air.

Figure 3.5a

Figure 3.5b

71

3.1.3.4 Correlation of Oxidative Activity with Chemical Constituents of PM
The DTT and ROS assays measure the PM oxidative activity based on the generation of
different reactive oxygen species. The dissimilar diurnal profiles of PM activity obtained
by these assays indicate different pathways of redox active chemical PM species in
generating cellular oxidative stress. Linear regression analysis was carried out to
investigate the association of the DTT and ROS activities with quasi-UFP chemical
composition. Table 3.2 shows a summary of the regression analysis [i.e. slope, intercept,
correlation coefficients (R
2
) and the associated levels of significance (p value)] for select
species which were correlated (R
2
>0.60) with either the DTT or ROS rates.

As shown in the table, we observed high correlation between WSOC and the DTT
activity of quasi-UFP (R
2
= 0.82; p = 0.01), despite the rather limited number of data
points. Previous studies have also documented the positive correlation of DTT activity
and WSOC content of the PM (Biswas et al., 2009; Hu et al., 2008; Verma et al., 2009).
This supports our hypothesis that the increased solubility of organic compounds mediated
by photo-chemical oxidants during secondary organic aerosols formation enhances the
capability of PM to generate free radicals.

The measured organic compounds were further categorized as per different functional
groups (PAHs, hopanes and steranes, alkanes, high and low carbon number organic
acids) to investigate their correlation with oxidative activity. As shown in Table 3.2, DTT
activity is positively correlated with high carbon number organic acids (C
15
-C
29
) (R
2
=

72

0.71; slope = 0.01). Many of these organic acids are formed by the photo-oxidation of
semi-volatile organic species, producing polar oxygenated compounds. Thus their
positive correlation with DTT suggests their partial contribution to the association of
DTT with WSOC discussed earlier.

73

Table 3.2: Summary of the regression analysis [slope, intercept, correlation coefficients
(R
2
) and the associated levels of significance (p value)] for select species with DTT and
ROS levels. Species included in this table are correlated (R
2
>0.60, shown bolded) with at
least one of the assays.

Species
DTT Activity Macrophage ROS
R
2
p Slope* Intercept
ψ
R
2
p Slope* Intercept
ψ
NO
3
-
0.68 0.04 0.82 -8.12 0.20 0.37 79.97 4.57
SO
4
-2
0.69 0.04 0.20 0.01 0.08 0.59 11.88 28.79
NH
4
+
0.62 0.07 0.58 -0.03 0.14 0.47 48.41 18.32
WSOC 0.82 0.01 0.70 0.00 0.09 0.58 40.11 29.14
Mg 0.80 0.02 0.01 0.07 0.15 0.44 0.48 27.75
Al 0.64 0.06 1.00E-03 0.05 0.14 0.47 0.10 25.36
Ca 0.17 0.41 0.00 0.20 0.84 0.01 0.30 1.45
S 0.04 0.72 0.00 0.24 0.91 0.01 0.14 -25.79
K 0.71 0.04 3.00E-03 0.12 0.22 0.35 0.37 28.11
Ti 0.79 0.02 0.02 0.08 0.15 0.44 1.18 28.72
V 0.01 0.84 -0.01 0.35 0.91 0.00 9.83 -8.16
Fe 0.69 0.04 1.00E-03 0.07 0.28 0.28 0.15 18.76
Ni 0.13 0.49 -0.04 0.42 0.73 0.03 16.12 5.38
As 0.01 0.85 -0.20 0.36 0.66 0.05 281.00 -10.08
Cd 0.08 0.60 -1.82 0.42 0.77 0.02 1024.00 -9.31
Ba 0.75 0.03 0.04 0.07 0.19 0.38 3.35 24.64
PAHs 0.68 0.04 -0.33 0.59 0.02 0.79 11.02 38.33
Organic
Acids
(C
15
-
C
29
) 0.71 0.15 0.01 0.07 0.07 0.74 0.45 44.02

Note: Sample size for the regression analysis is N = 6.
*expressed as oxidative activity (nmol/min for DTT and µg Zymosan units for ROS) per
µg or ng of species (µg for inorganic ions and WSOC; ng for metals, PAHs and organic
acids).
Ψ
expressed as nmol/min/m
3
for DTT and µg of Zymosan units/m
3
for macrophage ROS

74

In the present study, the DTT activity of the ultrafine particles is observed to be
negatively correlated with measured PAHs (R
2
= 0.68, slope = -0.33) as shown in Table
3.2. This is in contrast to previous studies showing rather modest to strong positive
associations between DTT activity and PAHs (Cho et al., 2005; Ntziachristos et al.,
2007). Although PAHs themselves do not contain the functional groups capable of
catalyzing the oxidation of DTT, their possible photo-oxidation during photochemical
episodes in the afternoon period may convert them to oxy-PAHs, quinones and nitro-
PAHs, which are all active in DTT assay (Cho et al., 2005). However, we did not
quantify these oxygenated products in the present study. Nevertheless, the negative
correlation of the measured PAHs with DTT activity may be explained by their lowered
concentrations in the afternoon due to possible volatilization. It is important to note that
this volatilization may be followed by photo-oxidation to form oxygenated PAHs, which
contribute to the increase of DTT activity in the afternoon. The difference in the sampling
protocols between our study and previous studies (Cho et al., 2005; Ntziachristos et al.,
2007) might explain the contrasting correlations. In the present study, the sampling
protocol was intentionally designed to distinguish the impacts of primary and secondary
sources on PM toxicity. This differs from the uninterrupted time integrated sampling of
previous studies, ranging from several hours to several days, without discriminating the
morning and afternoon periods, which would obscure the impact of PAHs chemical
transformation on redox properties of PM. Thus, the positive correlation in those studies
may be attributed to the confounding relationship between PAHs and oxygenated PAHs.
Moreover we did not observe any significant associations between DTT activity and the

75

concentrations of hopanes, steranes and alkanes, which again suggests that organic
compounds present in relatively fresh traffic PM emissions are not as redox active as
their aged photo-oxidized products.

Inorganic ions such as nitrate (NO
3
-
), sulfate (SO
4
-2
) and ammonium (NH
4
+
) and certain
water soluble trace elements (Mg, Al, K, Ti, Fe and Ba) are also correlated with DTT
activity as shown in Table 3.2, although none of these species is mechanistically active in
this assay (Cho et al., 2005). These inorganic species were correlated with WSOC
(R>0.75), suggesting their co-linearity with the redox active water soluble organic
compounds, rather than actual contribution to DTT activity.

Table 3.2 shows that ROS measured PM activity is correlated (R
2
>0.65; p<0.05) with
certain water soluble elements such as V, Ni, Cd (transition metals) and As. V and Ni are
the metals which participate in Fenton chemistry and are capable of generating hydroxyl
radicals once coming in contact with the biological cells (Schoonen et al., 2006). These
hydroxyl radicals can effectively oxidize the DCFH probe used in the ROS assay. The
ROS activity was significantly correlated with V concentrations (R
2
=0.91; p=0.01), and
also Ni (R
2
=0.73, p=0.03). Both of these species are mostly emitted from bunker fuel
(heavy fuel oil) combustion, although V is emitted at a proportionally higher rate
(Arhami et al., 2009). Hu et al., (2008) also observed an association between V and ROS
activity in a comparison of the oxidative activity profiles of PM in the Los Angeles-Long
Beach Harbor area and attributed this association to the influence of ship emissions on

76

ambient PM. The significant correlation of V with the ROS activity measured in this
study, in agreement with previous studies (Hu et al., 2008; Verma et al., 2009), is of
particular note because it confirms the impact of heavy fuel oil combustion on particle
oxidative activity. The remaining correlated metals (Cd and As) are associated with
primary sources such as vehicular exhaust (Arhami et al., 2009; Thomaidis et al., 2003),
and are unaffected by the secondary photo-chemical reactions occurring in afternoon.

We did not observe any significant correlation between WSOC and ROS activity
(R
2
=0.09; p=0.58), in contrast to the case of DTT activity. It indicates that despite the
possible contribution of secondary organic compounds on HO
.
production, the cellular
generation of oxidants (and thus ROS activity) is predominantly limited by the transition
metals, which are the active participants in the ROS assay and are mostly emitted from
primary sources. Consequently, the relative fractions of metals and WSOC in PM
determine the net change of ROS activity of particles attributable to primary and
secondary sources and probably explain its inconsistent trend in the morning and
afternoon period, as mentioned in previous section.

There are no studies in recent literature to document the oxidative activity of species such
as Ca and S, which are also significantly correlated with the ROS assay results, as shown
in the Table 3.2. Their correlation with ROS activity is probably due to their correlation
with redox active transition metals.

77

3.1.4 Conclusions
The ROS measured activity of PM is mostly driven by transition metals and possibly
amplified by WSOC, its overall value depends on the relative PM mass properties of both
groups of species in the morning and afternoon periods. On the other hand, the DTT
activity, determined primarily by WSOC, clearly increases in the afternoon period.
Although primary particles have the capability of generating free radicals in cells, their
photo-oxidation products - secondary particles appear to be more potent in terms of
generating oxidative stress, which underscores the importance of chemical
transformations of primary emissions with atmospheric aging on the overall PM toxicity
and associated health risks.

78

3.2 Oxidative Activity Profiles of Ambient Particulate Matter Influenced by
Woodsmoke Emissions from Wildfires

3.2.1 Introduction
The semi-arid Mediterranean climate of Southern California, characterized by dry and hot
autumns and unusually fast (>85mph) Santa Ana winds, contributes to the occurrence of
wildfires in the region, often during the fall period. Ambient levels of air pollutants
including carbon monoxide (CO), nitrogen oxides (NOx), volatile organic compounds
(VOCs) and particulate matter (PM) increase in the affected areas during these events
(Artaxo et al., 1994; Phuleria et al., 2005). Wood smoke generated from wildfires
contains a large number of organic/inorganic chemicals such as aldehydes, carboxylic
acids, polycyclic aromatic hydrocarbons (PAHs), resin acids, potassium, and magnesium.
Many of these chemicals (e.g. PAHs and aldehydes) are toxic and potential human
carcinogens (Naeher et al., 2007), while some of these species (e.g. potassium and
levoglucosan) are often used as tracers for the presence of wood smoke in the ambient air
(Schauer et al., 2001; Simoneit et al., 1999). Several in-vivo studies on laboratory
animals found a positive correlation between wood smoke exposure and adverse health
effects, such as pulmonary lung cell injury (Thorning et al., 1982), tracheal erosion and
oxidative stress (Dubick et al., 2002), and allergic airway inflammation (Barrett et al.,
2006).

On October 20, 2007, a series of wildfires began in Southern California, ranging from
Santa Barbara County to the U.S.–Mexico border, which burned more than 500,000 acres

79

of land causing the largest mandatory evacuation in the state's history. The counties of
San Diego and Los Angeles alone accounted for 91.5% of the total burnt area. Although
the last fire was fully contained by November 9, all of the fires in the Los Angeles
County were under control by October 30, 2007.

This paper presents measurements of various particulate and gaseous pollutants at a site
near the University of Southern California (USC), in downtown Los Angeles, during the
October 2007 fire event. Detailed chemical and toxicological characteristics of PM
samples collected during the fire episode, including water-soluble organic carbon,
inorganic elements, organic compounds and reactive oxygen species, were evaluated and
compared to those after the fire event, when outdoor PM
2.5
levels were mostly impacted
by traffic emissions. One of our main goals was to determine the oxidative activity
profiles of PM during these two periods and correlate them with source-specific PM
constituents, in order to evaluate the potential contributions of the two major PM sources
during this campaign (wildfires and vehicular traffic) to the overall PM oxidative
properties.

3.2.2 Experimental Methods
3.2.2.1 Sampling Location
Sampling was conducted at the University Park Campus of USC, 2.5 km south of
downtown Los Angeles. Further details about the sampling site are provided in section
3.1. Gaseous pollutant data were taken from SCAQMD (South Coast Air Quality

80

Management District) measurements at North Main Street (Los Angeles), about 2 km
northeast from the site.

3.2.2.2 Sampling Protocol
Integrated PM
2.5
samples were collected using a high volume sampler (HIQ
Environmental Products Co., CA; flow rate = 450 l/min) (Misra et al., 2002) on five
different days (October 24, October 25, October 27, November 1, November 14).
Particles were collected using Teflon coated glass fiber filters (Pallflex, Fiberfilm
T60A20 – 8x10 IN, Pall Corp., East Hills, NY) seated in a specially designed 20x25 cm
high-volume filter holder. Since the Los Angeles wildfires started on October 20 and
continued until October 30, samples collected from October 24 to October 28 represent
conditions impacted by both vehicular traffic and wildfire emissions, while the November
1 and 14 samples were considered representative of typical ambient conditions in that
area, affected mostly by vehicular traffic emissions.

A Scanning Mobility Particle Sizer (SMPS Model 3080 and DMA Model 3081, TSI Inc.,
St. Paul., MN) was set up concurrently to measure the particle size distribution. The
aerosol sampling flow rate was 0.3 L /min, with a scan up time of 255 s. The minimum
and maximum sizes detectable at these settings are 14.1 and 736.5 nm, respectively.

81

3.2.2.3 Sample Analysis
All Teflon coated glass fiber filter samples were extracted in water and then analyzed for
trace elements by Inductively Coupled Plasma-Mass Spectroscopy (ICP-MS) (Lough et
al., 2005) and water soluble organic carbon (WSOC) using a Shimadzu TOC-5000A
liquid analyzer (Decesari et al., 2001). For quantification of the organic compounds
present in the particulate phase during and after the fire event, 3 during-fire samples and
2 post-fire samples were composited, respectively, and analyzed by gas
chromatography/mass spectrometry (GC/MS). The oxidative activity of the collected
particles was assessed by two different assays: 1) macrophage ROS - in vitro exposure to
rat Alveolar Macrophage (AM) cells using Dichlorofluorescin Diacetate (DCFH-DA) as
the fluorescent probe (Landreman et al., 2008) and 2) consumption of Dithiothreitol
(DTT) in a cell-free system (DTT assay) (Cho et al., 2005).

3.2.3 Results and Discussion
3.2.3.1 Particulate Matter Characteristics
Figure 3.6 shows the evolution of the particle number size distribution from days during
wildfires (October 24-25, 26-27, and 27-28) to the post-fire period (November 01-02 and
14-15). All scans were taken from 04:30 pm until 10:30 am of the following day to
capture the effects of evening and morning traffic rush hours, which clearly caused an
increase in particle number concentration between 07:00 and 11:00 pm, and from 06:00
to 10:00 am. Scans collected between October 24 and 28 are characterized by a higher
number of larger particles (mobility diameter ≥ 100 nm). The mode diameter was

82

generally higher for most of the sampling time on days during fire (around 70-110 nm)
than that recorded on November 01-02 and 14-15 (peak value at 60 nm). Similar results
were observed during the 2003 wildfire event in Southern California (Phuleria et al.,
2005), and previous studies also indicated that the particle number size distribution of
freshly emitted wood-smoke is unimodal, with average mode diameter around 100 nm
(Hays et al., 2002). In addition, the particle number concentration increased during the
fire event by an approximate factor of 2 (Figure 3.7) and, although it was also notably
high on November 14-15, the mode diameter remained below 60 nm, which is typical of
particles emitted mainly from vehicular sources in Los Angeles (Kuhn et al., 2005).

83

Figure 3.6: Evolution of the particle number size distribution from days characterized by
wildfire emissions (October 24-25, 26-27, and 27-28) to periods not directly affected by
fire (November 01-02, 14-15). Data were collected using a Scanning Mobility Particle
Sizer (SMPS, Classifier Model 3080 and DMA Model 3081, TSI Inc., St. Paul., MN)

84

Figure 3.7: 24-hr averaged concentration of NO, NO
2
, CO and O
3
at North Main Street
(Los Angeles) and Particle Number Concentration at the USC (also 24-hr average)
before, during and after the fire period.

3.2.3.2 Gaseous Pollutants
Figure 3.7 presents 24-hr average concentrations of CO, nitrogen oxide (NO), nitrogen
dioxide (NO
2
), and Ozone (O
3
) in downtown Los Angeles for pre-, during- and post-fire
periods. While the concentrations of both CO and NO increased almost 3-fold during the
fire event, the O
3
and NO
2
levels were marginally affected by wildfire emissions.
Phuleria et al. (2005) observed a similar trend for these gases during the October 2003
0
20
40
60
80
100
120
140
160
13-Oct
15-Oct
17-Oct
19-Oct
21-Oct
23-Oct
25-Oct
27-Oct
29-Oct
31-Oct
2-Nov
4-Nov
6-Nov
8-Nov
10-Nov
12-Nov
14-Nov
16-Nov
18-Nov
NO, O 3,NO 2 , Partcle no. Conc.
0
200
400
600
800
1000
1200
1400
CO
Particle No.(x1.0 E-3) (cm-3) NO2,ppb NO,ppb O3, ppm CO,ppb
Pre-fire
During fire
Post-fire

85

Southern California wildfire. Reduced photochemical activity due to the smoke blanket
over the region might explain these trends in O
3
and NO
2
.

3.2.3.3 Water Soluble Trace Elements
Table 3.3 shows a comparison of the average fractional concentration (expressed in µg/g
of PM) of water-soluble elements in ambient particles measured at the USC site during
and after the fire event. Ambient levels of these trace elements from a previous study
conducted few months before at the same location are also provided (Arhami et al.,
2009). It should be noted that the average meteorological conditions during that study
were very similar to our post-fire period (temperature ranged from 59 to 63
o
F, and
relative humidity was between 60 and 71%). The fractional concentrations of certain
elements increased by a factor of 2 or more between October 24 and 28. The results of
one-sample t-test revealed a statistical difference (at the 95% confidence interval, or CI)
between the mean fractional concentrations of Mg, P, K and Mn observed during and
after the fire event. When a CI of 90% was considered, statistical difference was also
observed for Co, Ti, and other less abundant elements (e.g. Ga,Sr, Ru and others, not
shown). Potassium, often used as a signature species for biomass combustion and a tracer
of forest fires (Artaxo et al., 1994), was one of the elements whose levels were
consistently elevated during the fire episode. Interestingly, the fractional concentration of
K, even after the fire event, was higher (5447 µg/g) than its typical level (2245 µg/g) at
the site, which indicates a likely persistence of residual wood smoke even after the fire
was ceased. This may be particularly true for the November 1 sample, whose PM mass

86

may contain a considerable fraction of residual wood smoke constituents, including re-
suspended smoke-related particles from the paved roads.

Table 3.3 Average fractional concentration of water soluble elements (ICP-MS) in
ambient particles collected at the USC site during (October 24, 26 and 27, 2007) and after
(November 1 and 14, 2007) the fire event. Ratios between during- and post-fire
concentrations are also reported. Statistically significant levels and ratios (at p=0.10) are
shown in bold letters.

Element
During Fire Post Fire Elevation
Ratio
(during
fire / post
fire)
Typical ambient
levels at the sampling
site (µg/g of PM)
Average
Value,
µg/g of PM
S.D.
Average
Value,
µg/g of PM
S.D.
Average
Value,
µg/g of
PM
S.D.
Ti 7.11 0.94 2.13 1.11 3.34 2.89 0.65
Cu 631.73 546.24 236.10 80.42 2.68 275.23 47.44
Mn 334.75 134.77 153.49 69.72 2.18 102.52 2.00
Cd 4.00 2.58 1.85 0.71 2.16 3.89 1.45
Mg 5860.12 1328.00 2857.33 1422.36 2.05 2672.45 982.73
P 954.06 470.36 476.40 320.41 2.00 178.81 0.18
Al 577.21 464.83 298.70 414.18 1.93 256.40 24.53
K 9887.85 2102.38 5446.63 2594.56 1.82 2244.88 77.58
Zn 1513.01 1101.37 845.08 462.17 1.79 540.77 69.90
Na 28411.30 18607.50 16205.20 1421.05 1.75 14298.66 2762.341
Co 3.50 0.56 2.23 0.07 1.57 2.42 0.23
Ca 33829.90 20785.60 26472.00 27208.10 1.28 3228.07 1981.39
Fe 639.61 360.28 524.81 713.10 1.22 590.49 34.45
Cr 7.90 0.42 6.92 0.46 1.14 4.24 2.76
Ni 29.89 17.40 28.04 5.37 1.07 35.48 9.09
Pb 13.81 11.54 13.10 16.16 1.05 21.04 5.24
Ba 968.64 499.50 944.71 1199.34 1.03 404.67 152.45
S 23087.70 9865.16 28102.50 9808.47 0.82 20158.39 4923.36
V 57.08 36.38 77.50 15.20 0.74 105.54 5.86

87

Several studies have reported increased concentrations of K, P, Cl, Zn, Br, Si, Ca
(Echalar et al., 1995) and S (Artaxo et al., 1994) in ambient air affected by biomass
combustion or wildfires. These particle-bound elements traversed the Los Angeles basin
due to the prevailing westerly wind direction. An analysis of size-segregated biomass
burning samples showed that most of these elements dominate the fine PM mode
(aerodynamic diameter less than 2.5 µm), which can be transported by winds over
hundreds of kilometers (Artaxo et al., 1994). By contrast, the concentrations of other
elements (e.g. V, Fe, Ni, Cu, Zn, Ba, Pb) that typically originate from vehicular emissions
(Ning et al., 2007) or other combustion sources (residual fuel oil combustion; (Arhami et
al., 2009)) and road dust (Wang et al., 2005) were similar between samples collected
during and after the fire period.

3.2.3.4 Water Soluble Organic Carbon (WSOC)
The WSOC content of ambient particles was higher during the fire event (170 ± 80
μg/mg of PM) compared to the post-fire period (78±35 μg/mg). A one-sample t-test
confirmed a statistically significant difference (95% CI) between the mean WSOC levels
observed during and after the fire event. Similar increases in the WSOC concentration
were observed by Saarikoski et al. (2007) in ambient particles collected in Helsinki
during a major biomass-burning episode in April-May 2006. Wood smoke is rich in
organic compounds {56 to 90% by mass, depending on the wood type and combustion
conditions (Schauer et al., 2001)}, the majority of which (up to 99%) is oxygenated in the
atmosphere and becomes water-soluble (Novakov and Corrigan, 1996).

88

These studies (including our own) confirm that biomass burning is an important source of
water-soluble organic compounds in ambient aerosols. This is of particular note in the
context of global climate change, as WSOC plays an important role in the formation of
cloud condensation nuclei by modifying the hygroscopicity of atmospheric aerosols
(Cruz and Pandis, 1998).

3.2.3.5 Organic Constituents
Figure 3.8 presents the fractional concentrations of the speciated organic compounds in
ambient particles collected at USC (expressed as µg/g of PM), during and after the fire
period, and also from a previous study conducted at the same site few months before
(Arhami et al., 2009).

The concentration of levoglucosan increased by 2-fold during the fire episode compared
with the post-fire period (Figure 3.8a). However, post-fire levels of levoglucosan were
also significantly higher than those typically present at that site. This may be due to the
fact that levoglucosan is an extremely stable compound emitted in large quantities from
the pyrolysis of wood cellulose (Jordan et al., 2006), and has a long residence time in the
atmosphere. It is also possible that re-suspended smoke-related particles contribute to the
elevated levels of levoglucosan in the post fire samples.

The concentrations of most PAHs were not affected by the fire episode, with the
exception of retene, which was elevated during the fire event (6.68 µg/g) but was not

89

detected in the post-fire period (Figure 3.8b). Retene is known to be present in wood
smoke in relatively higher amounts compared to other PAHs and is commonly used as an
indicator of softwood burning (Villalobos-Pietrini et al., 2006).

Figure 3.8: Fractional concentration of the organic constituents in PM samples (expressed
in µg/g of PM) collected at the USC site, during and after the fire events: a)
Levoglucosan, b) Polycyclic Aromatic Hydrocarbons (PAHs), c) n-Alkanes and Fatty
acids

Levoglucosan
0
2000
4000
6000
During Fire After Fire Arhami et al,
2008
Fine et al, 2004
µg /g
a)

90

Figure 3.8 (Continued)

Figure 3.8c shows an increase in some of the n-alkanes concentrations measured in PM
samples collected from October 24 to 28. Although this increase was not substantial (up
to a factor of 1.4) compared with the corresponding post-fire levels, it was quite marked
PAHs
0
3
6
9
12
15
Phenanthrene
Anthracene
Fluoranthene
Acephenanthrylene *
Pyrene
Benzo(GHI)fluoranthen
Cyclopenta(cd)pyrene
Benz(a)anthracene
Chrysene
1-Methylchrysene
Retene
Benzo(b)fluoranthene
Benzo(k)fluoranthene
Benzo(j)fluoranthene *
Benzo(e)pyrene
Benzo(a)pyrene
Indeno(1,2,3-cd)pyrene
Benzo(GHI)perylene
Dibenz(ah)anthracene
Picene
Coronene
Dibenzo(ae)pyrene
µg/g
During Fire After Fire Arhami et al,2008
n-Alkanes and Fatty acids
1
100
10000
1000000
Tricosane
Tetracosane
Pentacosane
Hexacosane
Heptacosane
O ctacosane
Nonacosane
Heptadecane
Heneicosane
Triacontane
Hentriacontan
Dotriacontane
Tritriacontane
Pentatriaconta
Hexatriacontan
Heptatriaconta
O ctanoic acid
D ecanoic acid
Dodecanoic
Tetradecanoic
Pentadecanoic
Hexadecanoic
Heptadecanoic
Octadecanoic
Nonadecanoic
Palmitoleic
O leic acid
Linoleic acid
Linolenic acid
Eicosanoic
Heneicosanoic
Docosanoic
Tricosanoic
Tetracosanoic
µg/g
During Fire After Fire Arhami et al, 2008
b)
c)

91

(up to a factor of 3.5) with respect to the typical levels reported by Arhami et al., (2009)
at the same site. Fatty acids were present in significant amounts in the “during-fire”
samples, and were not quantifiable in samples collected on November 01 and November
14 (Figure 3.8c). Arhami et al., (2009) also showed negligible concentrations of fatty
acids present in the ambient air at the same site under similar metrological conditions.
These hydrocarbons (n-alkanes and fatty acids) are generally emitted by the volatilization
of plant waxes from the leaf surface at high temperatures, or from fossil fuel combustion
(Yan, 2007).

3.2.3.6 PM Oxidative Activity
Figure 3.9a shows the results of the DTT assay conducted on samples collected during
and after the fire event. The outcomes were reported both as nmol of DTT consumption
per µg of PM (nmol DTT / min × µg), as well as per m
3
of air (nmol DTT / min ×m
3
).
Particles collected between October 24 and 28 produced higher DTT activities per unit of
PM mass than those collected on November 01 and 14. DTT activity was the highest on
October 24 (0.024 nmol DTT / min × µg) and decreased gradually with time to 0.005
nmol DTT / min × µg on November 14. Data by Li et al., (2003) previously obtained at
the same site in the absence of wildfire emissions and under similar meteorological
conditions (average temperature 55 ± 4
o
F, relative humidity 61 ± 12 %) is also shown in
Figure 3.9 for comparison. Although the DTT activity may vary depending upon a
number of factors that influence the PM chemical composition (such as season and traffic
intensity), the levels reached at the peak of the fire event on October 24 are remarkably

92

higher than those observed in previous studies at the same site (Cho et al., 2005; Li et al.,
2003; Ntziachristos et al., 2007).

Figure 3.9: Oxidative activity of PM samples collected at the USC site during and after
the fire period a) expressed as nmol of DTT consumption /min, b) expressed as the ROS
response, or µg of Zymosan units

a)

b)

0
200
400
600
800
1000
1200
1400
1600
1800
24-Oct. 26 -Oct. 27-Oct. 01-Nov. 14- Nov.
µg Zymosan Units/mg of PM
0
20
40
60
80
100
120
µg Zymosan Units/m
3
of air
µg Zymosan Units/mg PM
µg Zymosan Units/m3 Air
During Fire
Post-Fire
0.000
0.005
0.010
0.015
0.020
0.025
0.030
24-Oct 26 -Oct. 27-Oct. 01-Nov. 14- Nov. Li et al ,
2003
nmol DTT/minxug
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1
nmol DTT/minxm
3
DTT Consumption,
nmol/minxug of PM
DTT Consumption,
nmolDTT/minxm3 of air
During Fire
Post Fire

93

DTT expressed in units of nmol DTT / min × m
3
of air is also important in the context of
population exposure to ambient PM. This parameter (DTT activity per m
3
of air) peaked
on October 26 (0.80 nmol DTT / min × m
3
) because of a very high PM
2.5
concentration
(58.85 μg/m
3
; not shown) coupled with an elevated per-mass DTT activity (0.014 nmol
DTT / min × μg). Although on November 1, the per-mass activity was quite low (0.009
nmol DTT / min × μg), the high PM
2.5
concentration (63.47 µg/m
3
; not shown) led to a
substantial increase in DTT activity per m
3
of air (0.60 nmol DTT / min × m
3
). The lower
temperature and higher relative humidity on November 1 (the 24-hr average values were
60
o
F and 80%, respectively, compared to 78
o
F and 31% on October 24) might have
resulted into limited atmospheric mixing and, thus, higher PM
2.5
concentration. The DTT
activity per m
3
of air during the fire event was consistently higher (average, 0.62±0.21
nmol/min x m
3
) compared to that reported by Li et al., (2003) (0.27 nmol/min x m
3
. This
discussion illustrates an important point about the relative oxidative activity of PM
emitted from various sources; comparable activity on a per-PM mass basis may not
necessarily translate into an equivalent public risk if the overall exposure varies due to
changes in total PM concentration.

Figure 3.9b shows the results of the macrophage ROS assays on particles collected during
and after the fire period. In contrast to DTT, the highest ROS activity was observed for
the November 1 sample, both on per-mass and on per-volume of air basis. Although the
per-mass activity was relatively high on October 26, the lower values obtained on
October 24 and 27 make it difficult to associate the overall results with the occurrence of

94

wildfires. It should be noted that the DTT and macrophage ROS assays are probably
influenced by different PM species. While the former reflects the oxidative potential of a
number of polar (water soluble) organic compounds, such as oxygenated PAH, quinones,
and nitro-PAHs (Cho et al., 2005), the latter is mostly influenced by water soluble
transition metals, such as Fe, Cu, Cr, Ni, and Co (Arredouani et al., 2005). Transition
metals, which generate hydroxyl radicals through Fenton chemistry, are not active in the
DTT assay. Even PAHs have to be oxidized to water soluble compounds, such as
quinones and oxy-PAH before becoming redox active (Ntziachristos et al., 2007) in the
DTT assay.

Based on the above discussion, we attempted a correlation of PM oxidative activity with
particle constituents such as WSOC and water-soluble elements. We performed a
qualitative analysis of the impact of fire emissions on those components that seem to
affect the oxidative activity of the collected samples. Table 3.4 shows the coefficients of
statistical determination (R
2
) and the associated levels of significance (p values) for the
correlation of different selected PM species with DTT and ROS. The analysis showed a
significant association between DTT and WSOC (R
2
= 0.65, p= 0.102), confirming
previous results by Cho et al. (2005), where DTT was well correlated with the organic
carbon content of PM (R
2
= 0.53). Although our statistical analysis included only the
water-soluble fraction of organic carbon, the majority of organic compounds in wood
smoke are oxygenated and polar. The possible oxidation of PAHs during long distance
transport might also enhance their solubility in water. Table 3.4 also illustrates that some

95

of the measured elements (e.g. Ti, Mn, K, Mg and Co) are well-correlated with DTT.
However, most of these species (in particular Mg, K and Mn) were strongly associated
with WSOC, which implies that their relationship with DTT may not reflect their
oxidative potential, but rather their common origin with water soluble organic
compounds. Similarly, Ntziachristos et al. (2007) attributed the high correlation between
DTT and several transition metals (e.g. Mn, Cu, Fe and Zn) to their common source with
PAHs (i.e., vehicular emissions).

96

Table 3.4 Coefficients of statistical determination (R
2
) and associated levels of
significance (p-value) for the correlations between selected water soluble PM constituents
and oxidative activities (as measured by the DTT and ROS assays) for PM samples
collected during and after the fire period. Statistically significant correlations (R
2
≥0.60;
p≤0.10) are shown in bold.

Conversely, we observed strong associations between macrophage ROS and most of the
water soluble transition metals (Cr, V, Fe, Ni, Pb,) (Table 3.4), consistent with earlier
studies (Arredouani et al., 2005; Ciapetti et al., 1998) showing the capability of these
trace elements to induce oxidative stress in cells through the formation of ROS. The
correlation between the ROS assay and elements that are not redox active (e.g., Al, Ba
Chemical
Constituents
Correlation with DTT Correlation with ROS
R
2
p value R
2
p value
WSOC 0.65 0.10 0.00 0.96
Ti 0.85 0.03 0.24 0.40
Mn 0.75 0.06 0.01 0.85
K 0.73 0.06 0.08 0.61
Mg 0.71 0.07 0.03 0.75
Co 0.71 0.08 0.56 0.14
Cr 0.61 0.12 0.84 0.03
P 0.56 0.15 0.01 0.90
Al 0.43 0.23 0.71 0.07
Ba 0.43 0.77 0.95 0.04
Pb 0.40 0.25 0.94 0.00
Fe 0.39 0.26 0.96 0.00
Cd 0.39 0.26 0.42 0.23
Ni 0.36 0.28 0.80 0.04
Zn 0.31 0.32 0.54 0.15
Na 0.29 0.36 0.33 0.31
V 0.25 0.39 0.79 0.04
S 0.22 0.42 0.93 0.08
Ca 0.15 0.49 0.21 0.47
Cu 0.08 0.65 0.04 0.75

97

and S) is probably due to their common origin with transition metals. However, no
consistent increase in ROS activity was observed for any of these species during the fire
episode.

3.2.4 Conclusions
The opportunistic nature of this study constrained our sampling and data collection to a
small number of measurement days, which is one of our major limitations. Although
results from DTT and ROS assays may not provide definitive answers on the relative
health risks associated with PM emitted from motor-vehicles and wildfires (the two main
PM sources during the fire episode), the study presents an effective approach to link the
chemical composition of ambient PM with their toxicity response and thus to infer their
distinct sources. While both assays revealed considerable oxidative activity of PM from
traffic sources, which are ubiquitous in the Los Angeles area, the DTT assay showed
additional increase in oxidative activity due to the contribution of the wildfires.

98

Chapter 4: Contribution of Semi-Volatile Organic Compounds
and Non-Volatile Transition Metals in the PM Oxidative
Activity

This chapter is divided in two sections: section 4.1- evaluating the contribution of semi-
volatile organic compounds in the oxidative activity (measured by DTT assay) of ambient
particles; and section 4.2 - quantifying the role of transition metals in the ROS activity
(measured by macrophage ROS assay) of diesel exhaust particles. The major findings of
both studies have been summarized in the respective sections.

4.1 Contribution of Semi-Volatile Organic Compounds in the Oxidative (DTT)
Activity of Ambient Particles

4.1.1 Introduction
Atmospheric PM is a complex mixture of semi-volatile and non-volatile species
(Seinfeld and Pandis, 2006). Semi-volatile organic compounds (SVOC) exist
simultaneously in the gas and particle phases at equilibrium; partitioning between the two
phases depends upon their vapor pressure, which is a function of temperature. Since
chemical reactions are generally faster in gaseous phase, the oxidation of SVOC vapors
yields secondary organic aerosols (SOA), shifting them again into particulate phase
(Robinson et al., 2007). Thus, the cycling of SVOC between gas and particle phase
ultimately leads to a net increase in the oxygenated organic aerosols (Miracolo et al.,
2010) with modified chemical and likely, toxicological, characteristics. Certain SVOC,
such as polycyclic aromatic hydrocarbons (PAHs) and their derivatives, are known to be
99

genotoxic and carcinogenic (EPA, 2002). In Chapter 2, we have shown that the semi-
volatile fraction of diesel exhaust particles (DEPs), evaluated in chassis dynamometer
facilities, is highly redox-active and constitutes a dominant fraction (up to almost 100 %
in few vehicles emissions) of the PM oxidative potential. However, freshly emitted PM in
the low dilution environment of the dynamometer facilities, in which most of the SVOC
are bound to the particle phase, is not directly comparable with the ambient PM. In an
ambient urban environment, various atmospheric processes, including dilution and photo-
chemical reactions significantly alter the physicochemical and oxidative characteristics of
semi-volatile compounds. An improved understanding of these characteristics and the
overall contribution of semi-volatile compounds to the atmospheric aerosol toxicity is
essential in elucidating the health risks related to PM exposure.

The core objective of this study is to evaluate the contribution of semi-volatile PM
constituents in the oxidative potential of ambient quasi-ultrafine particles. A
thermodenuder (TD) was used in three temperature configurations - 50, 100 and 200
o
C
to shift the gas-particle partitioning of semi-volatile components of ambient aerosols
sampled at an urban site. The oxidative potential of collected particles is measured by the
DTT (dithiothreitol) assay. Detailed chemical analyses of the collected particles,
including organic and elemental carbon, water soluble elements, inorganic ions and
PAHs, are conducted to quantify the volatility profiles of different PM components, and
also to investigate their effect on oxidative potential.

100

4.1.2 Experimental Methods
4.1.2.1 Sampling Location
The sampling for this study was conducted at the Particle Instrumentation Unit (PIU) of
Southern California Particle Center (SCPC) near downtown Los Angeles. Meteorological
conditions at the site were mostly consistent during the sampling campaign, with 24-hour
average temperature values of 19 ± 2
o
C and relative humidity of 66 ± 7 %. Additional
information including the results of earlier PM investigations at the sampling site have
been previously reported (Chapter 3).

4.1.2.2 Sampling Protocol
Sampling was conducted in the fall of 2009, over a period of 6 consecutive weeks
(excluding weekends), for 5-6 hrs each day. Figure 4.1 shows the schematic of the
experimental setup used. The thermodenuder was operated at three different
temperatures, i.e. 50, 100 and 200
o
C, and both ambient and denuded particles were
collected for each temperature configuration. Thus, three sets of integrated samples,
corresponding to the three TD temperature configurations (50, 100 and 200
o
C), each
consisting of two samples – ambient and thermodenuded, were collected in succession.
The sampling duration of each set of samples varied from 50-80 hrs and was primarily
determined by the sufficient mass loading on each filter (~ 600 µg) required to conduct
chemical and toxicological analyses. All samples were stored in a freezer at -20
o
C
between sampling intervals. A low pressure drop impactor (Misra et al., 2002) was used
to select the quasi-ultrafine particles (Dp ≤180 nm in the present study). Particles were
101

concentrated using the versatile aerosol concentration enrichment system (VACES) to
enhance the rate of mass collection. Detailed laboratory and field characterization of the
VACES are described elsewhere (Khlystov et al., 2005; Kim et al., 2000; Kim et al.,
2001a; Kim et al., 2001b; Ning et al., 2006; Zhao et al., 2005), demonstrating its
capability to preserve aerosol chemical composition (including semi-volatile species) in
both the fine (Khlystov et al., 2005; Kim et al., 2001b; Ning et al., 2006) and the ultrafine
(Kim et al., 2000; Kim et al., 2001b; Zhao et al., 2005) PM modes during the
concentration enrichment process. In brief, aerosol enrichment is achieved by growing
the ambient particles (to 2-3 µm droplets) through a saturation–condensation system and
their subsequent concentration by virtual impaction. A diffusion dryer system desiccates
the aerosol stream exiting the virtual impactor to relative humidity <50 %. Concentrated
particles after VACES were segregated into two airstreams: one stream was fed into the
thermodenuder (described below) and the denuded stream was further split for separate
collection of particles onto 37 mm quartz (Whatman, QMA, 3.7 cm, 1851-037) and
Teflon (PALL Life Sciences, PTFE membrane, 2.0 um, R2PJO37) filters; particles from
the stream by-passing the thermodenuder (representing the ambient PM) were similarly
collected onto Teflon and quartz filters. The particle size distributions of ambient and
denuded aerosols were concurrently measured using two Scanning Mobility Particle
Sizers (SMPS Model 3080 and DMA Model 3081, condensation particle counter Model
3022A, TSI Inc., St. Paul., MN) set up at inlet and exit of the thermodenuder, throughout
the sampling period. The particle size range detectable by SMPS is 14.6-685.4 nm at an
aerosol flow rate of 0.3 lpm and with a scan time of 300 s. Within the context of this
102

study, ultrafine particles <50 nm in diameter are designated as the ‘‘nucleation’’ mode,
whereas those in the 50–180 nm range are called the ‘‘Aitken’’ mode (Hinds, 1999). This
terminology helps to a better description of the size-segregated particle losses described
later in the chapter.

Figure 4.1: Experimental setup

Ambient Air
Ultrafine
Impactor
(Dp
50
<180 nm
Heater
Thermodenuder
CPC
Electrostatic
Classifier
CPC
Teflon filter
Adsorption
/Cooling
Pump
Quartz filter
SaturationTank
Virtual Impactor
Diffusion Dryer
300 lpm
285
lpm
5 lpm
15 lpm
Heater
Pump
Teflon filter
Quartz filter
Electrostatic
Classifier
10 lpm
5 lpm
4.7 lpm
4.7 lpm
2.5 lpm
2.5 lpm
VACES
0.3 lpm

103

The thermodenuder (TD; Model ELA-230, Dekati Ltd. Osuusmyllynkatu 13, FIN-33700
Tampere, Finland) consists of a heating section, followed by an adsorption tube, which is
actively cooled. As the ambient aerosol is drawn and passed through the heating section,
part of its volatile/semi-volatile component is evaporated. The aerosol then enters the
adsorption/cooling tube, where the evaporated compounds are adsorbed onto activated
charcoal placed on the walls of this section. The remaining PM (not volatilized) is
collected onto the filters placed downstream of TD. The Dekati thermodenuder used in
this study allows a relatively small residence time of 0.7s in the heating section at 10 lpm.
The aerosol mass fraction remaining (MFR) after heating is quite sensitive to TD
residence time. Recent studies have shown that thermal equilibration time scales of
ambient particles vary over several orders of magnitude (from less than a second to
hours), depending on the factors including vaporization enthalpies and mass
accommodation coefficients of the evaporating species (An et al., 2007; Riipinen et al.,
2010). Although the condition of equilibrium is necessary to calculate thermodynamic
properties of aerosol species (e.g. enthalpy and vapor pressure) based on the TD data, our
goal was not to investigate these relationships. Instead, we successively removed semi-
volatile inorganic and organic species at different temperatures and determined the
degree to which their removal affects PM oxidative potential. In that respect, deviation
from the thermal equilibrium of aerosols should not pose a serious limitation in the
present case.
104

4.1.2.3 Sample Analysis
Sections of all PM samples were extracted in high purity deionized (18 mega-ohm) water
at ambient temperature. The water extracts from the Teflon filters were analyzed for trace
elements and major ions. Trace elements in the 0.45 µm filtered water extracts were
analyzed using a magnetic-sector (high resolution) Inductively Coupled Plasma-Mass
Spectroscopy (SF-ICP-MS) following the approach of Zhang et al. (2008). Ion
Chromatography (IC) was used to analyze the water soluble ions (Zhang et al., 2008).
The elemental and organic carbon (EC and OC) contents of deposited PM on the quartz
filters were analyzed by the ACE (aerosol characterization experiments)-Asia Method
(Schauer et al., 2003). A separate section of the quartz filters was analyzed for PAHs by
TD-GCMS (Thermal Desorption- Gas chromatography/mass spectrometry) method
(Sheesley et al., 2009) using Markes International Thermal Desorption Unit (model M-
10140; Foster City, CA, USA) coupled with an Agilent Technologies GCMS system (a
6890 GC interfaced with a 5973 MSD). Several field blanks and laboratory blanks
collected during the sampling were similarly processed and all analytical results were
blank corrected. The oxidative potential of PM collected on the Teflon filters was
quantified by consumption of dithiothreitol (DTT) in a chemical system (DTT assay).

4.1.3. Results and Discussion
4.1.3.1 Physical Parameters
Particle losses in the thermodenuder were evaluated with ambient and ammonium sulfate
[(NH
4
)
2
SO
4
] particles without heating the aerosols. Penetration efficiency was
105

determined by concurrent measurements of the particle size distributions at inlet and exit
of the thermodenuder. For 20 nm size and at a flow-rate of 10 lpm, the particle number
losses were less than 9 % for both ambient and (NH
4
)
2
SO
4
aerosols, with even lower
losses for larger particles (~2 % for 30 nm). This is consistent with theoretical diffusional
loss calculations for laminar flow (Reynolds number <1200) at 25
o
C (Baron and
Willeke, 2001). Although diffusional loss increases with a rise in temperature, it does not
substantially affect the denuder’s penetration efficiency in our experimental range (25-
200
o
C). Theoretical calculations showed a slight (<3 %) increase in diffusional loss for
20 nm particles on increasing the temperature from 25 to 200
o
C. The size-segregated
particle diffusion losses were used to correct the size distributions of denuded aerosols.

Figure 4.2 shows the number and volume size distributions of ambient as well as
thermodenuded aerosols. In general, these measurements show that heating the aerosols
causes substantial particle shrinkage due to the evaporation of semi-volatile components.
However, a small amount of water associated with the particles also evaporates on
heating, leading to a slight overestimation of the measured loss of semi-volatile species.
The relative magnitude of total number and volume losses provides an approximate
indication of the differential evaporation profiles of nucleation (<50 nm) and Aitken
mode (50 - 180 nm) particles. While number loss is mostly governed by the loss of
nucleation mode particles, shrinkage of Aitken mode particles is reflected in the volume
reduction. It should be noted that particle loss might imply shrinkage below the detection
limit of SMPS, i.e. 14.6 nm in our case. Thus, the substantial loss of nucleation mode
106

particles at 50
o
C, as evident from Figure 4.2 is pronounced in the higher number loss
(47±2 %) compared to the corresponding volume loss (41±7 %). Due to the Kelvin effect
(Hinds, 1999), nucleation mode particles tend to be more volatile at a given temperature
than larger particles. Earlier observations at the sampling site have shown that organic
carbon dominates (>40 %) the mass fraction of particles below 50 nm (Sardar et al.,
2005). Although size-segregated PM chemical speciation was not conducted in the
present study, bulk loss of the nucleation mode at 50
o
C is most likely attributed to the
evaporation of organic compounds.

107

Figure 4.2: Particle size distributions (number and volume based) of the concentrated
ambient and thermodenuded particles at 50, 100 and 200
o
C.
Note: Error bars show the standard deviation of multiple scans.
Dp, nm
550 500
dN/dlog(Dp), #/cm
3
0.0
4.0e+4
8.0e+4
1.2e+5
1.6e+5
2.0e+5
Ambient
Denuded
50
o
C, Number - 47± 2 % Loss

Dp, nm
5 50 500
dV/dlog(Dp), nm
3
/cm
3
0.0
2.0e+10
4.0e+10
6.0e+10
8.0e+10
1.0e+11
1.2e+11
Ambient
Denuded
50
o
C, Volume - 41± 7 % Loss

108

Figure 4.2 (Continued)
Dp, nm
550 500
dN/dlog(Dp), #/cm
3
0.0
4.0e+4
8.0e+4
1.2e+5
1.6e+5
2.0e+5
Ambient
Denuded
100
o
C, Number - 52± 6 % Loss

Dp, nm
5 50 500
dV/dlog(Dp), nm
3
/cm
3
0.0
2.0e+10
4.0e+10
6.0e+10
8.0e+10
1.0e+11
1.2e+11
Ambient
Denuded
100
o
C, Volume - 52± 12 % Loss

109

Figure 4.2 (Continued)

Dp, nm
550 500
dN/dlog(Dp), #/cm
3
0.0
4.0e+4
8.0e+4
1.2e+5
1.6e+5
2.0e+5
Ambient
Denuded
200
o
C, Number - 67± 5 % Loss

Dp, nm
5 50 500
dV/dlog(Dp), nm
3
/cm
3
0.0
2.0e+10
4.0e+10
6.0e+10
8.0e+10
1.0e+11
1.2e+11
Ambient
Denuded
200
o
C, Volume - 70 ± 9 % Loss

110

As the temperature is increased, shrinkage of the Aitken mode particles also becomes
prominent, leading to an equivalent loss in both number and volume concentration at 100
o
C (~52 %). The particle size distribution of thermodenuded aerosols remains bimodal at
100
o
C, with a slight shift in mode diameter of the first peak (~32 nm in ambient particle
size distribution) towards lower size (~20 nm) due to particle shrinkage. Further increase
in temperature to 200
o
C causes loss of smaller particles to the extent that nucleation
mode is almost eliminated, resulting in a unimodal size distribution. These observations
are consistent with previous findings showing that the fraction of ambient particles below
50 nm is internally mixed, consisting mostly of volatile materials, while particles in the
Aitken mode are externally mixed (Moore et al., 2007). These externally mixed particles
shrink upon heating by evaporation of semi-volatile material from the surface, leaving
behind the solid core, with a mode at around 80 nm. The fractional losses in volume
concentration of the denuded aerosols at 50, 100 and 200
o
C are consistent with the
gravimetrically measured mass reduction ratios at these temperatures (42 %, 52 % and 67
%, respectively).

4.1.3.2 Chemical Parameters
Water Soluble Ions
Figure 4.3 shows the concentrations of major water soluble inorganic ions in
concentrated ambient and denuded particles at three TD temperatures: 50, 100 and 200
o
C. Significant loss of all of these inorganic ions occurs with heating. At 50
o
C, a higher
loss was observed for nitrate (NO
3
-
; 49±7 %) and ammonium (NH
4
+
; 44±5 %), than
111

sulfate (SO
4
-2
; 33±4 %). The general trends in the loss of these ions are consistent with
their predominant molecular speciation [ammonium nitrate (NH
4
NO
3
)
,
(NH
4
)
2
SO
4,
and
sulfuric acid (H
2
SO
4
)] in an urban atmosphere (Sardar et al., 2005). However, the small
fraction remaining at 200
o
C is likely explained by the presence of refractory species,
such as NaNO
3
and Na
2
SO
4
, which are probably formed by the heterogeneous chemical
reactions of sea-salt with nitric and sulfuric acid (Engler et al., 2007).

Elemental and Organic Carbon (EC-OC)
Figure 4.4 shows the comparison of elemental (EC) and organic carbon (OC) in
concentrated ambient and denuded particles collected at different TD temperature
settings, i.e. 50, 100 and 200
o
C. These two species dominate the quasi-ultrafine
particulate mass, accounting for up to 70 % at the sampling site (Sardar et al., 2005). EC
is often considered as an indicator of diesel vehicular emissions, which is one of the
predominant sources of carbonaceous compounds in the Los Angeles atmosphere
(Schauer, 2003). In addition to traffic sources, OC is also contributed by photo-chemical
processes producing secondary organic compounds (Ning et al., 2007). Despite the
divergence in individual levels of both of these species, apparent in Figure 4.4, which is
attributed to the daily variations in emission source strengths, the ratio [(OC/EC)
ambient
] is
relatively stable, with an average value of 3.4±0.5 (not shown). This consistency
primarily reflects the similarity in meteorological conditions prevailing throughout the
study period, which also strengthens our comparison of the various parameters at
different TD temperature configurations, as discussed in the paper.
112

Figure 4.3: Concentration of water soluble inorganic ions (NH
4
+
; Figure 4.3a, NO
3
-
;
Figure 4.3b and SO
4
-2
; Figure 4.3c) in the concentrated ambient and thermodenuded
particles at different TD temperature configurations – 50, 100, 200
o
C.
Note:
1. Error bars show analytical uncertainties in the chemical analysis.
2. Percentages on top of the bars show loss in concentration on heating the aerosols.

0
1
2
3
4
5
6
7
50C 100 C 200 C
Concentration, µg/m
3
Temperature,
o
C
NH
4
+
(a)
Ambient
Denuded
44±5 %
94±
26%
49±12 %

0
2
4
6
8
10
12
14
50C 100 C200 C
Concentration, µg/m
3
Temperature,
o
C
NO
3
‐
(b)
Ambient
Denuded
49±7 %
93±16 %
92±
16%

113

Figure 4.3 (Continued)

0
2
4
6
8
10
12
14
50C 100 C 200 C
Concentration, µg/m
3
Temperature,
o
C
SO
4
‐2
(c)
Ambient
Denuded
33±4%
69
±9 %
49±8%

114

Figure 4.4: Concentration of elemental (Figure 4.4a) and organic (Figure 4.4b) carbon in
the concentrated ambient and thermodenuded particles collected at three TD temperature
configurations (50, 100 and 200
o
C).

Note:
1. Error bars show analytical uncertainties in the chemical analysis.
2. Percentages on top of the bars show loss in concentration on heating the aerosols.

0
1
2
3
4
5
6
7
8
9
50 C 100 C 200 C
EC, µg/m
3
(a)
Ambient
Denuded

0
5
10
15
20
25
30
50 C 100 C 200 C
OC, µg/m
3
(b)
Ambient
Denuded
35±2 %
40±3 %
74±5 %

115

As defined by the measurement methodology (ACE-Asia method), EC is considered
refractory even up to a temperature of 870
o
C (Schauer et al., 2003). As expected, this
species is only marginally affected [(within the measurement uncertainty (±5-13 %)] by
heating. In contrast, OC is progressively decreased with an increase in TD temperature
(35±2 %, 40±3 % and 74±5 % loss at 50, 100 and 200 oC, respectively). OC consists of a
variety of organic compounds such as PAHs, hopanes, steranes, alkanes and organic
acids. Most of these compounds are semi-volatile at ambient temperatures
(Subramanyam et al., 1994), which explains the substantial loss of OC with heating.

Polycyclic Aromatic Hydrocarbons (PAHs)
The concentration of PAHs in concentrated ambient and denuded aerosols as a function
of TD temperature is shown in Figure 4.5. In general, a substantial evaporative loss (up to
80 %) upon heating is evident. PAHs are semi-volatile organic compounds and traffic
emissions from the nearby freeway (I-110) is their most obvious source at our sampling
site (Ning et al., 2007). Twenty-one PAHs were quantified and for a more illustrative
interpretation of the loss profiles, these are divided into two groups, based on their
molecular weights, i.e. low molecular weight (LMW- Phenanthrene, Anthracene,
Fluoranthene, Pyrene, Benzo[ghi]fluoranthene, Cyclopenta[cd]pyrene,
Benzo[a]anthracene, Chrysene, Retene,; <252 g/mol), and high molecular weight (HMW-
Benzo[b]fluoranthene, Benzo[k]fluoranthene, Benzo[j]fluoranthene, Benzo[e]pyrene,
Benzo[a]pyrene, Perylene, Indeno[1,2,3-cd]pyrene, Benzo[ghi]perylene,
Dibenzo[ah]anthracene, Picene, Coronene, Dibenzo[ae]pyrene; ≥252 g/mol). The
116

fractions of LMW PAHs lost are 31±6 %, 66±7 % and 78±21 % at 50, 100 and 200
o
C,
respectively (Figure 4.5a). Recent studies have indicated that more than 80% of the PAH
mass in an urban atmosphere may be adsorbed onto the surface of soot particles rather
than absorbed into the organic matter (Dachs and Eisenreich, 2000; Perraudin et al.,
2005; Tran-Duc et al., 2010). The relatively lower enthalpy of adsorption compared to
that of absorption might further facilitate the volatilization of these compounds from the
particle surface. The loss in concentration of HMW PAHs is smaller at 50 and 100
o
C
(14±2% and 53±4%, respectively; Figure 4.5b), compared to the LMW PAHs. This is
consistent with the decreasing volatility of PAHs with an increase in number of carbon
atoms (Zheng and Fang, 2000). However, a difference in the evaporation profiles is not
apparent at 200
o
C, when most of the PAH mass is volatilized (~80%), leading to a
similar loss for both LMW and HMW PAHs.

117

Figure 4.5: Concentration of PAHs [categorized based on their molecular weights, i.e.
low molecular weight (LMW; Figure 4.5a) and high molecular weight (HMW; Figure
4.5b)], in concentrated ambient and thermodenuded aerosols collected at three TD
temperature configurations – 50, 100 and 200
o
C.
Note:
1. Error bars show analytical uncertainties in the chemical analysis.
2. Percentages on top of the bars show loss in concentration on heating the aerosols.

0
500
1000
1500
2000
2500
3000
50 C100 C200 C
Concentraton, pg/m
3
Temperature,
o
C
LMW PAHs
Ambient
Denuded
31±6%
66±7%
78±21%
(a)

0
500
1000
1500
2000
2500
3000
3500
50 C100 C200 C
Concentraton, pg/m
3
Temperature,
o
C
HMW PAHs
Ambient
Denuded
14±2%
53±4%
81±16 %
(b)

118

4.1.3.3 Oxidative Potential
Figure 4.6 shows the oxidative potential of concentrated ambient and thermodenuded
particles at three TD temperatures (50, 100 and 200
o
C), measured by the DTT assay. The
oxidative potential is decreased by 42±5 %, 47±8 % and 66±6 % as the particles are
heated at 50, 100 and 200
o
C, respectively. The marked decline in DTT activity as the
aerosols are denuded of their semi-volatile components highlights the significant
contribution of semi-volatile compounds to the oxidative load of ambient quasi-ultrafine
particles in an urban atmosphere.

Figure 4.6: Percentage loss in oxidative potential, measured by DTT activity on heating
the ambient aerosols at 50, 100 and 200
o
C. The DTT activity of the concentrated
ambient (undenuded) PM, corresponding to different temperature settings of the
thermodenuder is also shown on secondary Y-axis as a reference.
Note: Error bars in DTT activity represent 95 % confidence intervals.

0
1
2
3
4
5
6
7
8
0
10
20
30
40
50
60
70
80
90
100
50 C 100 C 200 C
DTT activity of Ambient
(Undenuded) PM, nmol/min/m
3
% loss in DTT activity
Temperature ,
o
C
% loss in DTT activity
DTT activity of ambient particles

119

The sizable decrease in DTT activity at 50
o
C (42 %) implies that a large fraction of
oxidative activity may be associated with constituents that are volatile at 50
o
C. Similarly,
compounds volatilizing in the temperature window of 100-200
o
C also appear to be fairly
redox active, resulting in an additional 19 % loss of DTT activity in this temperature
range. The relationship between volatility temperature and decrease in oxidative potential
may provide some insights in identifying the PM-bound chemical species with
considerable oxidative activity. For example, the loss profile of DTT activity with
temperature closely follows the thermal desorption profile of OC, with substantial loss in
the ambient-50
o
C (42 % and 35 % for DTT and OC, respectively) and 100-200
o
C (19 %
and 34 % for DTT and OC, respectively) compared to 50-100
o
C (5 % for both DTT and
OC). The temperature window of 50-100
o
C, although associated with a major loss of
inorganic ions [(NO
3
-
(44 %), SO
4
-2
(16%)], does not impact the DTT activity
significantly. The association of oxidative potential with various chemical constituents is
further investigated by regression analysis in the following section.

The overall decrease in oxidative potential with heating the ambient PM is relatively
lower compared to that of DEPs evaluated in the dynamometer study (Chapter 2). That
study showed a roughly 70-100 % decrease at 220
o
C depending on the vehicle type,
after-treatment technology used and the driving cycle. However, the dilution ratio of the
vehicular exhaust in that study was much lower (i.e. ~ 6-80) compared to the atmospheric
dilution (on the order of thousands) in the present case. The increase in dilution ratio
shifts the gas-to-particle phase equilibrium, with a lower fraction of semi-volatile species
120

in the particulate phase under high dilution conditions. Other physicochemical processes,
including photo-chemical reactions, might also modify the net content of SVOC and their
contribution to the oxidative potential of ambient particles. These differences in the loss
profiles of DTT activity between freshly emitted PM and aged ambient PM underline the
importance of atmospheric processes, in characterizing the overall toxicity of inhaled
aerosols and assessing their health impacts.

4.1.3.4 Association of Oxidative Potential with PM Chemical Composition
To further examine the linkage between oxidative potential and chemical PM species,
linear regression between the DTT activity and PM constituents was performed. Both
ambient and denuded samples were included in regression and the results are summarized
in Table 4.1. The analysis shows a strong association between DTT activity and OC
concentration (R = 0.92; p = 0.01), which is consistent with earlier investigations
demonstrating the capability of organic compounds to catalyze the oxidation of DTT
(Cho et al., 2005; Obrien, 1991). In the context of present study, the excellent correlation
of DTT activity with OC concentration further emphasizes the significant role of semi-
volatile organic compounds in the PM oxidative potential.

121

Table 4.1: Summary of the regression analysis [slope, intercept, Pearson coefficient (R)
and the associated level of significance (p value)] between major PM chemical
constituents and DTT activity.

Species
DTT Activity
R p
Slope* Intercept
ψ

Value Unc. Value Unc.
NO
3
-
0.85 0.03 0.15 0.04 1.16 0.25
SO
4
-2
0.79 0.06 0.15 0.06 1.01 0.36
NH
4
+
0.77 0.08 0.32 0.13 1.15 0.33
OC 0.92 0.01 0.09 0.02 0.35 0.34
EC 0.35 0.49 0.24 0.32 0.32 1.89
LMW PAHs 0.88 0.02 937.84 250.43 0.47 0.38
HMW PAHs 0.80 0.05 645.04 305.34 0.63 0.58

Note: Sample size for the regression analysis is N = 6.
*expressed as DTT activity (nmol/min) per µg of species.
Ψ
expressed as nmol of DTT/min/m
3
.

The DTT activity is also well correlated with both groups of PAHs, i.e. LMW (R = 0.88,
p = 0.02) and HMW PAHs (R = 0.80, p = 0.05). PAHs themselves do not contain
functional groups capable of catalyzing the oxidation of DTT. However, atmospheric
chemistry e.g. photo-oxidation, may transform PAHs to quinones and nitro-PAHs
(Ntziachristos et al., 2007), both of which are major classes of redox cycling agents.
Thus, the positive correlation of PAHs with DTT activity is probably due to their
correlation with quinones and nitro-PAHs, which have been shown to be active in the
DTT assay (Cho et al., 2005). In addition to the cellular oxidation of thiols, some
quinones (e.g. 1, 4-benzoquinone) have been shown to also inactivate the antioxidants
such as glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Rodrigueza et al., 2005).
Although quinones are not analyzed in our study, the correlation with DTT underscores
122

the importance of PAHs as possible surrogates or precursors of redox active organic
species that augments the PM oxidative potential.

Some inorganic ions, such as SO
4
-2
, NO
3
-
and NH
4
+
are also correlated with DTT activity
as shown in Table 4.1. To the best of our knowledge, there is no evidence in literature
that these species are redox active in the aerobic environment. Their association with
DTT may be due to co-linearity with redox active organic compounds, instead of actual
contribution to the PM oxidative potential. Species, minimally affected by heating (EC
and water soluble trace elements) were not correlated with the DTT activity.

4.1.4 Summary and Conclusions
We assessed the role of semi-volatile compounds in the oxidative potential of ambient
quasi-ultrafine particles at an urban site. The investigation follows the dynamometer
study (Chapter 2), which demonstrated a dominant contribution of semi-volatile
components in the DTT activity of PM emitted from heavy-duty diesel vehicles in a
dynamometer facility. Compared to that study, a lower, although still considerable
fraction (42-66 %) of the oxidative potential of ambient PM, is contributed by its semi-
volatile compounds. The difference in DTT activity profiles between the two studies is
attributed to the lower content of semi-volatile species in ambient PM, as a result of the
much higher atmospheric dilution, compared to the freshly emitted DEPs in
dynamometer studies. Species that are volatile at ~ 50
o
C and between 100 and 200
o
C
(e.g. OC) appear to be strongly associated with the DTT activity.
123

4.2 Contribution of Transition Metals in the Reactive Oxygen Species Activity of
PM Emissions from Retrofitted Heavy-Duty Vehicles

4.2.1 Introduction
Recent toxicological studies indicate that PM-associated reactive oxygen species (ROS;
particle-bound and/or their cellular generation upon PM exposure) and the resulting
oxidative stress may be responsible for the initiation of inflammatory cascades in
biological systems (Donaldson et al., 1996; Donaldson et al., 2005; Kunzli et al., 2006).
The macrophage ROS assay is a fluorogenic cellular method to measure the generation of
reactive oxygen species in alveolar macrophages on exposure to aqueous suspensions of
PM. Due to the location of alveolar macrophages on inner epithelial surface of the lung,
this assay represents an excellent model of pulmonary inflammation in response to PM
exposure (Landreman et al., 2008). Recent investigations have also indicated that ROS
activity measured by this assay is highly associated with the transition metal content of
PM samples [Fe, V, Ni, Cr; (Hu et al., 2008; Verma et al., 2009)]. However, most of
these associations have been inferred based on statistical analysis - either simple or
multiple linear regressions between PM chemical constituents and ROS activity. While
these statistical tools can provide valuable insights, the inherent limitations of any
regression technique, i.e. lack of an explicit causal relationship irrespective of the
strength of correlation, and multi-colinearity among independent variables, pose
challenges to comprehend the underlying mechanisms linking particle composition to
toxicity. For an improved understanding of these associations, mechanistic studies must
be conducted, which include successive removal of various PM components from the
124

exposure samples to analyze their individual impacts on the measured activity. One of
such approaches could be the use of a chelation technique for segregating the
contribution of metals to ROS activity of PM as demonstrated by Shafer et al. (2010).

The objective of this study is to assess the role of water-soluble metals in ROS activity of
DEPs. Particles were collected from heavy-duty vehicles with and without retrofits (e.g.
DPF, SCR, etc.) in a dynamometer set-up. The contribution of water-soluble transition
metals in the measured activity was quantified by their complexation and subsequent
removal using chelation method. The significant variations in metal composition of PM
collected from different vehicle-configurations provided insights on the effect of
emission control technologies on exhaust PM toxicity as well as demonstrated the
robustness of the analytical approach. Observations are further supported by uni- and
multi-variate regression analyses between ROS activity and the water-soluble metal
content of PM.

4.2.2 Experimental Methods
4.2.2.1 Sampling Protocol
DEPs were collected from heavy-duty vehicles tested on a chassis dynamometer at the
California Air Resources Board’s laboratory in Los Angeles. The details of the test
vehicle-configurations including their efficiency to control mass and number based PM
emissions have been reported previously (Chapter 2).
125

All vehicles were tested under three standard simulated driving conditions, i.e. steady
state cruise (80 km/h), transient [EPA urban dynamometer driving schedule (UDDS)
(Ayala et al., 2002)] and idling. Vehicle exhaust was transported by a stainless steel hose
pipe and was mixed with filtered (high efficiency particulate air filter) air in a constant
volume sampler (CVS) for primary dilution. Integrated PM samples from the CVS were
collected using a high volume sampler [HI-Q Environmental Products Co., San Diego,
CA; flow rate = 450 l/min; (Misra et al., 2002)] on Teflon coated glass fiber filters
(Pallflex, Fiberfilm T60A20 – 8x10 inch, Pall Corp., East Hills, NY). Both tunnel (from
CVS dilution tunnel) and field blanks (from high volume sampler) were similarly
collected during sampling.

4.2.2.2 Sample Analysis
Sub-sections of the filter-collected PM were extracted with 15 mL of high purity (18
mega-ohm) water. Extractions were performed in 20 mL acid-leached, capped,
polypropylene tubes, with continuous agitation for 16 hours at room temperature in the
dark. Extracts were filtered through acid-leached 0.22 µm polypropylene syringe filters
prior to analysis by magnetic sector inductively-coupled plasma mass spectrometry (SF-
ICPMS). Numerous field/tunnel filter blanks and laboratory method blanks were
similarly processed (Zhang et al., 2008). Estimates of uncertainty for each measurement
were derived by propagating the uncertainty (1 standard deviation, stdev) components
from the SF-ICPMS analysis (3-analytical replicates) and blank subtraction (1 stdev of
multiple blanks).
126

The ROS-activity of the PM extracts was measured by in vitro exposure to rat alveolar
macrophage (NR8383, ATCC# CRL-2192) cells using 2’,7’-dichlorofluorescin diacetate
(DCFH-DA) as the fluorescent probe. All samples, as well as positive and negative
controls were analyzed in triplicate (3-wells). A minimum of six dilutions (each of them
in triplicate) of every sample extract was run to ensure that a linear dose-response region
could be identified. Uncertainty in ROS method was estimated by propagating the
standard deviation of triplicate ROS measurements with standard deviation of the applied
method blank. These values ranged from 5 to 78% (median = 20%, n=24).

Sub-samples of the primary water extracts (samples and controls) were further processed
by chelation with the immobilized ligand iminodiacetate (Chelex chromatography) to
remove the metal ions. For the Chelex treatment, mini-columns (1 mL polypropylene
with Teflon frits) of Chelex were prepared with 0.2 g of 18 mega-ohm water-slurried Na-
Chelex. Columns were rinsed with high purity water and buffered with 0.1 M sodium
acetate. PM extracts (~1.75 mL) were processed through the columns, under gravity flow,
in a solid phase extraction manifold, at a flow rate of ~1.0 mL/min. The Chelex-
processed extract was collected and immediately assayed for ROS-activity and
subsequent elemental analysis by SF-ICPMS. Control samples were processed through
the Chelex columns to ensure that the treatment did not produce ROS active species or
inhibit the activity of Zymosan (a β-1,3 polysaccharide of D-glucose). Zymosan was used
as a positive control as it is recognized by TLR-2 (Toll-like receptors) on macrophage
cells, activating a strong immuno-chemical and ROS response (Ciapetti et al., 1998).
127

Blank-corrected ROS fluorescence data were normalized to the response of a
standardized unit of Zymosan to correct for minor variations in method sensitivity
between assay batches. ROS activity is therefore reported in terms of Zymosan units to
facilitate the comparison of various data sets. None of the Chelex method blanks
exhibited detectable ROS activity, and Chelex column blanks (high-purity water eluants)
produced no measurable (<5%) suppression of ROS activity in Zymosan positive
controls.

4.2.3 Results and Discussion
4.2.3.1 Water Soluble Metals
Figure 4.7 shows the distribution of metals in the exhaust PM collected from various
vehicle-configurations, under three driving cycles, i.e. cruise, UDDS and idle. For a more
illustrative interpretation, these metals are categorized into different groups based on their
periodic properties.

128

Figure 4.7: Distribution of water-soluble metals, categorized into different groups based
on their periodic properties, in the exhaust PM from various vehicle-configurations,
under three driving cycles, i.e. cruise, UDDS and idle.
Redox Active Transition metals: Mn, V, Ni, Cu, Fe, Cr; Divalent Transition metals: Zn, Cd, Co: Heavy
metals: Tl, Pb, U, Th, W; Alkaline Earth Metals: Sr, Mg, Ba, Ca; PGE: Rh, Pt, Pd; Rare Earth: Eu, Pr, Ce,
Nd, La, Sm, Y, Dy, Ho, Yb, Lu; Semi Metals: As, Sb; Higher valent/hydrolyzed metals: Ti, Sc, Al
Notes: NA implies “not available”.
Idle
Baseline
V-SCRT
Z-SCRT
DPX
School Bus
Hybrid
Fraction
0.0
0.2
0.4
0.6
0.8
1.0
1.2
NA
NA
UDDS
Fraction
0.0
0.2
0.4
0.6
0.8
1.0
1.2
Redox Active Transition Metals
Divalent Transition Metals
Heavy Metals
Alkaline Earth Metals
PGE
Rare Earth
Semi-metals
Higher Valent/Hydrolyzed Metals
Cruise
Fraction
0.0
0.2
0.4
0.6
0.8
1.0
1.2
NA

129

As shown, the differential removal of these metals by control technologies and also their
probable release in some systems [e.g. V, Ni and Cr in SCRTs (Hu et al., 2009)] lead to
significant changes in the emitted PM elemental composition. In the cruise cycle,
baseline vehicle elemental matrix is composed largely of alkaline earth metals (Sr, Mg,
Ba and Ca; 75%) and divalent transition metals (Zn, Cd and Co; 23 %). Most of these
metals primarily originate from the lube oil and its additives (Liu et al., 2008; McGeehan
et al., 2005). A reduction in the fraction of alkaline earth metals is observed in the PM
downstream of all retrofits, along with a corresponding increase in other metal fractions,
such as transition metals and hydrolyzed metals (mostly Al, not shown). The most
notable increase is in the relative abundance of redox active transition metals (Mn, V, Ni,
Cu, Fe and Cr), which are negligible in baseline vehicle (0.4 %), but constitute a
significant fraction in almost all of the retrofitted vehicles, with V-SCRT and school bus
among the highest (9.4 %, both), followed by Z-SCRT (8.6 %). The other groups of
metals such as heavy metals (Tl, Pb, U, Th, W), platinum group elements (Rh, Pt, Pd;
PGE), rare earth metals (Eu, Pr, Ce, Nd, La, Sm, Y, Dy, Ho, Yb, Lu) and semi-metals
(As, Sb) account for less than 3 % of the total metals emissions.

The PM metals distribution in UDDS cycle of the baseline vehicle is similar to that of the
cruise cycle, with a comparable fraction of both alkaline earth and divalent transition
metals. Although the fraction of divalent transition metals is decreased by retrofits in
UDDS cycle, an increase in the fraction of redox active metals is evident here as well.
PM samples from the baseline truck in the idle cycle were not available for assessing the
130

impact of retrofits on metals distribution during idling. However, a comparison with
cruise and UDDS cycles of the same vehicle-configuration shows a significantly smaller
fraction of redox active transition metals, as well as lower contribution from higher valent
metals in the idle cycle.

The increase in the fraction of redox active metals by retrofitted vehicles is of particular
interest and is better elucidated by plotting these species in terms of their individual
contents (µg/g of exhaust PM) as shown in Figure 4.8. The highest content of V in PM
samples from V-SCRT (both cruise and UDDS cycles) can be attributed to the possible
leaching of wash-coat catalyst from SCR system at high temperatures (Hu et al., 2009).
The most prominent increase is in the content of Fe by retrofitted configurations, i.e.
several (2-4) orders of magnitude difference, compared to the baseline vehicle. The Fe
content is the highest in school bus PM both in cruise and UDDS cycles. Remaining
metals (Cr, Ni, Mn and Cu) were also generally escalated (elevation factor reaching up to
~80 for Cr in VSCRT-cruise) in the PM samples from retrofitted vehicles, compared to
the baseline truck.

131

Figure 4.8: Water-soluble content (µg/g) of major redox active metals in the exhaust PM
from various vehicle-configurations, under (a) cruise, and (b) UDDS driving cycles
Note: BDL indicates “below detection levels” [of Ni (baseline-both cruise and UDDS), V
(baseline-UDDS) and Cu (school bus-cruise)].
Cruise
Redox Active Metals
VCr Mn Fe Ni Cu
Water-soluble Content, μg/g
0.01
0.1
1
10
100
1000
10000
BDL
(a)
BDL

UDDS
Redox Active Metals
VCr Mn Fe Ni Cu
Water-soluble Content, μg/g
0.1
1
10
100
1000
10000
Baseline
V-SCRT
Z-SCRT
DPX
School Bus
Hybrid
BDL
BDL
(b)

132

4.2.3.2 Macrophage ROS Activity
Figure 4.9 shows the results of macrophage ROS assay conducted on the water extracts
of exhaust PM samples from all test vehicles. The results have been expressed both on a
per mass of PM emitted basis (µg ROS activity/mg of PM) (Figure 4.9a) as well as ROS
activity per km (or per hr for idle) of vehicle driven (Figure 4.9b). To place our results in
a broader perspective, ROS activity measurements from other studies, both for ambient
aerosols in central Los Angeles at USC as well as engine exhaust PM (Cheung et al.,
2010), have also been presented as an insert in Figure 4.9a. As shown, a large variation
in intrinsic ROS activity of PM from various vehicle-configurations – over 100 fold - was
observed, although, excluding the very high activity for the school bus, measurements are
in the range of typical levels observed for ambient and diesel exhaust particles.

133

Figure 4.9: Reactive Oxygen Species (ROS) activity of PM from the tested vehicles,
expressed as (a) per mass of PM, and (b) per km (or hr for idle) of vehicle driven
Vehicles
Baseline V-SCRT Z-SCRT DPX School Bus Hybrid
μg of Zymosan/mg of PM
0
400
800
1200
5000
10000
15000
20000
25000
Cruise
UDDS
Idle
µg of zymosan/mg
Spring 4700 ± 1500
Summer 9500 ± 2400
Fall 2900 ± 820
Euro I (Diesel) 270 ± 32
Euro II (Biodiesel) 290 ± 45
Euro III (Gasoline)
8600 ± 540
Euro IV (Diesel) 29 ± 48
Euro V (Diesel) 1200 ± 840
Ambient
PM at
USC site
Engine
exhaust
PM (UDDS
cycle)
Typical ROS Levels
PM
(a)

Vehicles
Baseline V-SCRT Z-SCRT DPX School Bus Hybrid
μg of Zymosan/km (cruise and UDDS)
0.0
5.0e+3
1.0e+4
1.5e+4
2.0e+4
2.5e+4
3.0e+4
3.5e+4
6.0e+4
8.0e+4
1.0e+5
μg of Zymosan/hr (Idle)
0
5e+3
1e+4
2e+4
2e+4
3e+4
3e+4
4e+4
4e+4
Cruise
UDDS
Idle
(b)

134

The ROS activity per mass of particles emitted from all retrofitted configurations is
higher than the baseline truck, with a sole exception of the hybrid vehicle. The highest
activity is observed for the school bus, which has been shown to be one of the most
efficient vehicles in controlling both mass and number emissions (>99 %; Chapter 2). It
is interesting to note that the hybrid vehicle, with similar particle mass and number
emissions as that of the school bus (Chapter 2), yields the lowest ROS activity. The DPX
vehicle, which is also quite efficient in reducing PM mass emissions (up to 95 %; Chapter
2), generates particles with high ROS activity (up to ~ 4600 µg of Zymosan units/mg of
PM). Of particular note are also the elevated ROS levels observed for both vanadium and
Zeolite based SCRTs compared to the baseline vehicle. Thus, although SCRTs
effectively remove the targeted pollutants for which they were designed (i.e,
hydrocarbons and NOx) in the exhaust emissions (Herner et al., 2009; Pakbin et al.,
2009), their in-built catalysts appear to increase the fraction of certain redox active
species in DEPs.

The ROS results expressed on per km (or per hr for idle) basis (Figure 4.9b) provide an
overall assessment of the effectiveness of control technologies on total oxidative load
imparted by the exhaust PM. It is of particular note that, although per PM mass activity is
increased by most of the control technologies, all of the retrofitted configurations are
efficient in reducing the per km (or per hr) ROS activity, compared to the baseline
vehicle. This reduction is a result of the lower particulate mass emissions (by more than
one order of magnitude) from retrofitted configurations compared to the baseline truck
135

(Chapter 2). The decrease in ROS activity is, however, not linear with the overall PM
mass reduction in retrofitted vehicles. For example, an approximate 10-times reduction in
PM mass emission achieved by Z-SCRT (cruise) lowers the ROS activity (per km) by
only 5 times. The contrast between distance and mass based activity is consistent with the
measurements using DTT assay on the same suite of vehicles (Chapter 2). These findings
have potential implications from the perspective of human exposure and risk assessment,
and underscore the importance of both the intrinsic toxicity of PM, and its mass emission
rate, while evaluating the net environmental impacts of an after-treatment technology.

4.2.3.3 Chelex Treatment
Figure 4.10 shows the percent removal (average of all vehicle-configurations) of major
water-soluble elements from the exhaust PM extracts after Chelex treatment. The figure
illustrates that chelation is highly efficient in removing most cationic di- and tri-valent
metals from the DEP extracts. Except for Cr and V (which exhibited removal percentages
of 57 % and 26 %, respectively), the removal efficiency for most of the transition metals
was greater than 70 %. The extent of metal chelation from the water extracts depends on
a number of factors including their oxidation states and aqueous complexation (Haas and
Northup, 2004; Thompson et al., 2002). The high removal efficiency for most of the
transition metals indicate minimal complexation of these species in solution, making
them biochemically available. Of the redox-active metals, Mn was most effectively
removed (93%) on Chelex, followed by Cu (79%), Fe and Ni (73%). The tendency of Cr
ions to form hydrates (octahedral coordination with water molecules), probably impedes
136

its attachment and removal with the iminodiacetate ligands (Fackler and Holah, 1965).
Further, the soluble Cr likely has a significant Cr (VI) component, which is an oxyanion
and is not effectively removed on Chelex (Gad, 1989). The formation of oxyanions by
vanadium (vanadate ion; VO
4
−3
) might be responsible for its poor removal efficiency
(Tischer et al., 2004). Other metals exhibiting very high removal efficiency include Mg,
Al, Ca, Ba, Cd and Pb. The non-metals (S and P) and semi-metals (As), mostly present as
oxyanions (Shafer et al., 2010) exhibit low removal efficiency.

Figure 4.10: Percent removal (average of all vehicle-configurations) of major water-
soluble elements using Chelex treatment of the exhaust PM samples

0
20
40
60
80
100
120
140
Mg Al P S Ca V Cr MnFe CoNi CuZn As CdBa Pb
% Removal on Chelex

Figure 4.11 shows the percent removal of ROS activity after Chelex treatment of the
exhaust PM extracts from all test vehicles under different driving cycles. To emphasize
the correspondence between water-soluble metals and ROS activity, the percent removal
137

of the sum of major metals from respective PM samples is also shown (Figure 4.11).
Chelex treatment of DEPs water extracts removed a very large and fairly consistent
fraction of ROS activity across most of the vehicles. The reduction in ROS activity
averages at 77 (±20) % (range: 36% for DPX-idle to 100 % for baseline-UDDS and
hybrid vehicles). The substantial reduction in ROS activity after metals chelation and
removal highlights the dominant contribution of water soluble metals to redox properties
of PM and is one of the important findings of this study.

Figure 4.11: Percent removal of ROS activity in relation to that of aggregate water
soluble metals after Chelex treatment of the exhaust PM samples from test vehicle-
configurations under different driving cycles.

0
20
40
60
80
100
120
Baseline
V‐SCRT
Z‐SCRT
DPX
School Bus
Baseline
V‐SCRT
Z‐SCRT
DPX
School Bus
Hybrid
V‐SCRT
Z‐SCRT
DPX
School Bus
Cruise UDDS Idle
% Removal
ROS activity
Metals

138

Under idling conditions, the V-SCRT, DPX and school bus show significantly lower
ROS removal (<50 %), despite comparable reductions in the aggregate metals content of
their PM samples. The results of one-sample t-test revealed that the mean fractional
removal of ROS in idle cycle is statistically different from that in cruise (p=0.002) and
UDDS (p=0.001) cycles. A vehicle-specific analysis of the percent removal of individual
metals illustrated minimal chelation (<40 %) of transition metals for these vehicles [V,
Cr, Fe for VSCRT and V, Cr, Fe and Cu for DPX and school bus] in idle cycle. Thus, it is
likely that the chemical speciation of many relevant metals is substantially different in the
PM emitted under idling conditions, which is reflected in both Chelex and ROS removal
efficiencies. In addition, these data imply that not all measured elements are capable of
generating PM related oxidative stress and only a few species – mostly redox active
transition metals appear to play a central role. The identification of major contributors to
PM toxicity is further investigated in the regression section discussed below.

4.2.3.4 Statistical Association between ROS Activity and Metals
Univariate Regression
To further support our observations and provide indications on the specific metals driving
ROS activity, a regression analysis was conducted investigating the associations between
ROS activity and major water-soluble elements in PM. As a preliminary step, univariate
linear regression was performed to screen the elements that are significantly correlated
with ROS activity. This selection process was conducted in three steps: 1) including only
untreated (i.e. not Chelexed) samples; 2) including only Chelexed samples; and 3) using
139

untreated minus Chelexed levels (designated as “delta” hereafter), in the regression
analysis. Table 4.2 summarizes all of these correlations along with the associated levels
of significance.

Table 4.2: Pearson correlation coefficients and associated levels of significance for the
univariate regression between ROS activity and major water-soluble metals of PM

Elements
Untreated
samples
Chelexed
Samples
Untreated-
Chelexed (delta)
R p R p R p
Mg -0.19 0.36 0.26 0.22 -0.18 0.39
Al 0.38 0.07 0.13 0.56 0.39 0.06
P 0.27 0.21 0.66 0.00 0.12 0.56
S 0.13 0.53 -0.32 0.12 0.22 0.30
Ca -0.01 0.95 0.36 0.09 -0.02 0.94
V -0.02 0.92 0.03 0.88 0.16 0.47
Cr 0.67 0.00 0.30 0.15 0.65 0.00
Mn 0.62 0.01 0.33 0.11 0.61 0.01
Fe 0.93 0.00 0.58 0.00 0.91 0.00
Co 0.61 0.04 -0.08 0.70 0.63 0.04
Ni 0.22 0.31 0.20 0.36 0.21 0.32
Cu 0.47 0.02 0.17 0.43 0.48 0.02
Zn 0.46 0.02 0.12 0.56 0.47 0.02
As 0.03 0.89 0.42 0.04 -0.29 0.17
Cd 0.50 0.01 0.59 0.00 0.47 0.02
Ba 0.31 0.15 0.09 0.67 0.27 0.20
Pb 0.47 0.02 0.11 0.59 0.47 0.02

As shown in Table 4.2, for the untreated sample set, ROS activity is reasonably
correlated with several transition metals such as Fe, Cr, Mn and Co (R>0.60; p<0.05).
The highest correlation was observed for Fe (R=0.93; p=0.00), followed by Cr (R=0.67;
p=0.00), Mn (R=0.62, p=0.01) and Co (R=0.61; p=0.04). By contrast, none of the
140

transition metals are significantly correlated with ROS activity in the Chelexed sample
set, which is not surprising, given the very low levels of these species (most of which
were not significantly different than zero) remaining after chelation. Thus, regression
analysis on Chelexed sample set was deemed inappropriate. Interestingly, the correlation
coefficients obtained for delta levels are very similar to that for the untreated samples,
which is again due to the near complete removal of most metals and ROS after chelation.
Other metals (Mg, Al, Ca, V, Ni and Ba), semi-metals (As) and non-metals (P and S) are
not significantly correlated with ROS activity in either untreated or delta set.

Multivariate Regression
The univariate regression employed in previous section may not allow us to discriminate
and quantify the impact of individual elements on PM toxicity. We therefore applied
multivariate regressions in order to better understand the variance of ROS activity as a
function of transition metals content. The metals exhibiting highest correlations with
ROS activity in the untreated dataset (Fe, Cr, Mn, Co, Cd, Cu, Pb and Zn) were selected
and stepped through one-by-one to generate multiple linear models (Sigma Plot 11.0,
Systat Software Inc).

As a first step, the regression model included only Fe as the independent variable. The
predicted ROS activity based on this model is plotted as a function of the measured ROS
activity in Figure 4.12a. The regression analysis yielded a high coefficient of
determination (R
2
= 0.84) with a slope and intercept of 0.84 and 992.4 µg of Zymosan
141

units/mg of PM, respectively, which is a very satisfactory prediction of the ROS activity.
The dominant influence of Fe on ROS activity has been reported in earlier investigations
also, evaluating the toxicity of ambient particles (Hu et al., 2008; Shafer et al., 2010;
Verma et al., 2009; Zhang et al., 2008). These statistical findings support our chemical
fractionation measurements and suggest that the soluble Fe content of the PM is a very
robust indicator of its ROS activity.

142

Figure 4.12: Correlation of measured and reconstructed ROS activity with successive
inclusions of different species in the multivariate regression model: a) only Fe; b) Fe and
Co; c) Fe, Co and Cr.

y = 0.84x + 992.4
R² = 0.84
0
10000
20000
30000
40000
0 10000 20000 30000 40000
Reconstructed ROS, µg of
Zymosan/mg of PM
Measured ROS, µg of Zymosan/mg of PM
ROS = f (Fe) (a)

y = 0.87x + 829.8
R² = 0.87
0
10000
20000
30000
40000
0 10000 20000 30000 40000
Reconstructed ROS, µg of
Zymosan/mg of PM
Measured ROS, µg of Zymosan/mg of PM
ROS = f (Fe, Co) (b)

y = 0.90x + 638
R² = 0.90
0
10000
20000
30000
40000
0 10000 20000 30000 40000
Reconstructed ROS, µg of
Zymosan/mg of PM
Measured ROS, µg of Zymosan/mg of PM
ROS = f (Fe, Co, Cr) (c)

143

Fe is among the most abundant transition metals in DEPs and is generally present in two
oxidation states, Fe (II) and Fe (III). The solubility of Fe (II) is much higher than that of
Fe (III); aqueous equilibrium concentrations of amorphous ferrous and ferric hydroxide
[am-Fe(OH)
2
and am-Fe(OH)
3
] are nearly 10
–5
M and 10
–10
M, respectively (Nico et al.,
2009). The redox cycling of Fe, generating an array of free radicals (e.g. OH
*
, O
2
*-
, and
HO
2
*
) through Fenton reaction, is well documented (Chen and Lippmann, 2009; Rose
and Waite, 2005; Valko et al., 2005; Welch et al., 2002). Although, the exact mechanism
of the reaction is still a matter of debate, Fe (II) catalyzed reduction of molecular oxygen
to release hydroxyl radicals [Haber Weiss reaction; (Winterbourn, 1987)] is considered to
be largely responsible in Fe mediated toxicity. The comparatively higher water solubility
of Fe (II) further makes it more bioavailable to probably participate in the reactions
yielding ROS activity of DEP extracts.

The next step in the multivariate regression modeling involved iterative inclusions of
each of the additional correlated metals (Cr, Mn, Co, Cd, Cu, Pb and Zn), and
examination of their impacts on the model’s ability to predict ROS levels. The
comparison of reconstructed and measured ROS activity, considering both Fe and Co into
the model, is shown in Figure 4.12b. Both correlation coefficient and slope of the
regression equation increased to 0.87, while the intercept decreased to 829.8 µg of
Zymosan units/mg of PM. Similarly, inclusion of Cr further improved the model by
increasing the correlation coefficient and slope to 0.90, and decreasing the intercept to
638 µg of Zymosan units/mg of PM (Figure 4.12c). The multivariate regression was
144

concluded at this final step as no additional species significantly contributed to improve
the correlation.

The parameters for the final regression equation are summarized in Table 4.3. As evident,
all metals included into the model contributed independently [variance inflation factor
(VIF) < 4] and significantly (p<0.05) to account for 90 % of the variance (R
2
=0.90) in
ROS activity. The standardized coefficients listed in the table may be used to quantify an
approximate contribution of the individual species in measured ROS activity. The highest
coefficient of Fe (0.94) implies that ROS activity is most sensitive to the changes in Fe
content of PM; a unit standard deviation change in Fe content would change the ROS
activity by 0.94 standard deviations, which is about 3 times higher compared to that by an
equivalent change in Cr content (standardized coefficient = 0.28).

Table 4.3: Parameters of the multivariate regression analysis between ROS activity and
water-soluble metals of PM

Species
Unstandardized Coefficient
(µg of Zymosan /ng of metal) Standardized
Coefficient
p VIF
Value error
Constant -539.24 985.84
Fe 17.77 1.84 0.94 0.00 1.86
Co -2468.18 766.22 -0.34 0.00 2.20
Cr 226.94 92.54 0.28 0.02 2.45

145

Interestingly, despite the positive correlation of Co as shown in Table 4.2, its coefficient
becomes negative after Fe is included into the model (Table 4.3). Perhaps the negative
association of Co in ROS predicting equation is a statistical artifact driven by the large
positive coefficient of Fe. Similar anomalies were observed in previous explorations
(Ntziachristos et al., 2007; Zhang et al., 2008), and illustrate one of the limitations of any
study attempting to link toxicological PM properties to their chemical constituents strictly
based on regression analysis. Accordingly, the apparent associations of different metals
with ROS activity as implied from their coefficients might not be representative of their
independent toxic potency, but rather in relation to other included redox species, and is
thus highly subjective to the variables selection in multiple regressions.

4.2.4 Summary and Conclusions
This investigation explores the linkage between the ROS activity of DEPs and their
water-soluble metals content. A macrophage-based in vitro ROS assay was coupled with
the elemental fractionation of exhaust PM from various heavy-duty diesel vehicles. The
test fleet included a baseline (without any control technology), two SCRTs (vanadium
and Zeolite based), a DPX (DPF), a school bus (EPF) and a hybrid (CCRT) vehicle. The
fleet was tested under three driving cycles: cruise, UDDS and idle. The study
demonstrates that despite generally similar reductions in PM mass emissions from diesel
vehicles by various control technologies, the intrinsic ROS activity of the emitted
particles may vary dramatically with retrofit types. Although, mass based ROS levels
(µg/mg of PM) increased with the application of after-treatment devices, a significant
146

reduction was observed in the overall ROS activity (per km for cruise and UDDS and per
hr for idle) for retrofitted configurations, compared to the baseline vehicle.

A substantial fraction of ROS activity of DEPs was associated with the water-soluble
metals as evident from the near complete attenuation of activity after chelation of PM
sample extracts. However, relatively lower removal of the activity in few vehicle-
configurations (V-SCRT, DPX and school bus -idle), despite a large aggregate metals
removal, indicated that the generation of ROS was driven by a select group of metal
species. A univariate regression analysis further supported that ROS activity is associated
with only few transition metals such as Fe, Cr, Co and Mn. Multivariate regressions
conducted on the selected transition metals and oxidative activity of PM revealed that Fe
accounts for most of the variability in ROS levels.

147

Chapter 5: Conclusions and Future Research Directions

5.1 Summary and Conclusions
Several recent toxicological studies have confirmed the oxidative properties of particulate
matter and their capability to generate reactive oxygen species in biological systems
(Donaldson et al., 2002; Xia et al., 2004). The production of these pro-oxidant species
and their subsequent damage to important macromolecules (e.g. DNA, proteins, lipids)
are thought to be implicated in numerous diseases including respiratory (Gauderman et
al., 2007; Stayner et al., 1998), cardiovascular (Delfino et al., 2005) and
neurodegenerative disorders (Peters et al., 2006). Diesel exhaust particles represent an
important component of fine and ultrafine PM in urban areas (Steerenberg et al., 1998).
The newer diesel engines with emission control devices are very efficient in reducing the
mass emissions of particulate matter. The emissions of major chemical constituents (e.g.
elemental and organic carbon) are also substantially reduced by these late model vehicles.
However, the enhanced formation of nucleation mode particles is observed in the
exhausts of some retrofitted configurations. Much of the mass of these nucleation mode
particles is semi-volatile in nature (e. g. sulfuric acid). Although, the retrofitted vehicles
emit less WSOC (per mile of vehicle driven), water soluble fraction of the organic carbon
(WSOC/OC) is increased for most of the configurations.

Despite an increase in the intrinsic oxidative activity (per mass basis) of exhaust PM with
the use of most control technologies, the overall activity (expressed per km or per hr) was

148

substantially reduced for retrofitted configurations compared to the uncontrolled vehicle.
The responses of two toxicity assays – DTT, and macrophage ROS assay, used in our
investigation to measure the oxidative activity of the particles, appear to be driven by
different PM components. The DTT activity was found to be correlated mostly with the
polar organic compounds. Secondary organic aerosol (SOA) formed by the photo-
chemical reactions during afternoon period and woodsmoke particles emitted during the
wildfire events, both are enriched in water soluble organic compounds, and show
increased DTT activity compared to the primary particles generated from vehicular
emissions. The dominant contribution of organic compounds in the DTT assay was
further confirmed by a mechanistic analysis using thermodenuder; in which, the
successive removal of semi-volatile organic compounds (SVOC) progressively declined
the DTT activity of the particles. These SVOC are found to be highly redox-active and
constitute a dominant fraction (up to almost 100 %) of PM oxidative potential for the
DEPs evaluated in chassis dynamometer facilities. Compared to that, in ambient PM, a
lower but still significant fraction (42-66 %) of the oxidative potential is contributed by
its semi-volatile compounds. The difference in DTT activity profiles between DEPs and
ambient PM is attributed to the lower content of semi-volatile species in ambient PM, as
a result of the much higher atmospheric dilution, compared to the freshly emitted DEPs
in dynamometer facilities, in which most of the SVOC are bound to the particle phase.

However, another important metric of oxidative activity – macrophage ROS assay,
doesn’t appear to be much affected by organic compounds, but was found to be

149

correlated mostly with the transition metals (Fe, V, Cr, Ni, Co, Cd, Cu and Zn). These
metals participate in the Fenton chemistry to generate hydroxyl radicals, which can
effectively oxidize the DCFH probe used in this assay. The key role of the transition
metals in ROS activity of PM was further quantified by their removal using a Chelex
®

complexation method; by which, chelation of the DEPs samples removed a substantial
( ≥70 %) fraction of the ROS activity. Identification of the major drivers of ROS activity
by means of a multivariate regression analysis revealed Fe as the most important metal,
accounting for the highest (84%) fraction of variability in the ROS levels.

The DTT and ROS assays used in this study are two independent and intrinsically
different types of analyses, likely driven by different particle components, and in that
respect, they should be considered complementary to each other. Collectively, these two
assays provide an important framework for designing an effective methodology to
understand the mechanism of PM toxicity and to infer the subsequent health risks. The
variation in the oxidative activity profiles of PM originating from different sources
measured by these assays is helpful in linking the PM toxicity to its source specific
composition.

5.2 Recommendations
This study offers an important perspective in analyzing and comparing the toxicological
characteristics of PM from different sources, which can be directly/indirectly useful in
formulating the policies for human health prevention from PM exposure. However,

150

considering the complexity of atmospheric reactions and their impacts on the PM
physicochemical properties, further investigations are required which should focus on
developing the appropriate sampling and measurement strategies to capture these
variations affecting the PM toxic potency. We will discuss some of these aspects in this
section with our recommendations to address these concerns.

Recent studies have indicated that the mechanisms by which transition metals cause
cellular oxidative stress are largely governed by the oxidation state of the metal, which
determines its bio-availability and thus subsequent toxicity (Shafer et al., 2010; Verma et
al., 2010). For example, two most common oxidation states of iron under
environmentally relevant conditions are: Fe (II) and Fe (III). Cr, like Fe, also has two
major oxidation states, Cr (III) and Cr (VI). Cr (III) compounds are generally insoluble
and non-toxic, whereas Cr (VI) is highly soluble, and potentially toxic (Gad, 1989).
Manganese has three readily available oxidation states: Mn (II), Mn (III), and Mn (IV).
Manganese appears to exert its toxic effects through a variety of pathways including ROS
formation by Mn (II) and direct oxidation of biological molecules by Mn (III). While Mn
(II) salts are more bioavailable (soluble), the cytotoxicity of Mn (III) compounds appears
to be significantly greater than those of Mn (II) (Aschner et al., 2005; Dorman et al.,
2001; Reaney and Smith, 2005). The oxidation state of these metals within a particle at
the time of inhalation is a function of the emission source and changes occurred during
atmospheric aging. Studies have indicated that photochemistry plays a major role in
determining the pertinent oxidation state of a metal by inducing either oxidation or

151

reduction of metals in the ambient atmosphere (Hoffmann et al., 1997; Siefert et al.,
1996). For example, particulate iron has been shown to undergo both atmospheric
oxidation and/or reduction depending on the factors such as presence of electron donors
(oxalate, formate, acetate) and pH (Siefert et al., 1994). Most of the previous
investigations studying the redox reactions of particulate metals are based on a simulated
atmosphere in laboratory, without sampling the real world aerosols. As ambient aerosols
are generally complex mixture of a variety of organic and inorganic species, it is
extremely difficult to predict how the redox active transition metals will behave in
atmospheric PM. The mechanisms of redox reactions involving transition metals are also
a function of size of the particles. It has been hypothesized that in ultrafine particles, Fe
(II) oxidation might be the dominating mechanism, while photo-reduction dominates in
larger particles (Majestic et al., 2007). Thus, an analysis of the diurnal variations in
oxidation states of major transition metals should be very useful in providing the insights
on bio-availability of these metals from the perspectives of ambient PM toxicity. In
addition, this analysis would help to assess the role of photo-chemistry and infer the net
mechanisms (among various oxidative and reductive reactions occurring simultaneously)
governing the changes in oxidation states of transition metals in a real world atmosphere.
In the present study, DTT activity of the particles (both ambient and DEPs) is found to be
strongly correlated with their WSOC content. However, WOSC itself is a complex
mixture of diverse group of species, including dicarboxylic acids, poly acidic compounds,
polyethers etc (Decesari et al., 2000). Our investigation measured only a few organic
acids, constituting a very small fraction (2-5% in most cases) of WSOC. Thus, it is

152

difficult to conclude what specific species contribute to the observed relationship between
WSOC and the DTT activity of PM. We believe that more research is needed to identify
the organic compounds present in water soluble components of the particles and to
quantify their individual contribution in the overall PM oxidative potential.

There have been issues also about the changes in chemical characteristics of organic
compounds by using thermodenuder for sampling the semi-volatile species. A detailed
discussion of these heating effects on the particle composition can be found in
Denkenberger et al., (2007). In general, accelerated kinetics at high temperatures may
lead to the formation of oligomers inside thermodenuder. This would mean that
thermodenuders might introduce an artifact by altering the chemical characteristics of the
organic compounds in addition to thermal stripping. Considering these issues, we
recommend to develop a vacuum based denuder for stripping the semi-volatile
compounds from PM matrix, such that the chemical characteristics of both volatile and
non-volatile species remain preserved.

Lastly, we recommend that in addition to the in-vitro evaluations as conducted in our
investigation, in-vivo studies will also be useful in addressing the importance of semi-
volatile and non-volatile PM fractions from the perspectives of their distinct toxicity
profiles. As previously mentioned, DTT assay measures the oxidative potential of organic
compounds by their capability to generate superoxide radicals in a reaction between DTT
and oxygen. Although attenuation of DTT activity with aerosol heating demonstrates the

153

significant contribution of semi-volatile organic compounds to the PM oxidative
potential; superoxide radicals formed in the oxidation reaction can interact with transition
metals, also present in PM, to generate hydroxyl radicals. Thus, for a comprehensive
assessment of the relative toxicity profiles of semi-volatile and non-volatile components
and their overall contributions to the PM toxicity, further mechanistic research is needed,
including in-vivo exposure studies focusing on these two categories of PM.

154

Bibliography

An, W.J., Pathak, R.K., Lee, B.H., Pandis, S.N., 2007. Aerosol volatility measurement
using an improved thermodenuder: application to secondary organic aerosol. Journal of
Aerosol Science 38 (3), 305-314.

Arhami, M., Sillanpaa, M., Hu, S.H., Olson, M.R., Schauer, J.J., Sioutas, C., 2009. Size-
Segregated Inorganic and Organic Components of PM in the Communities of the Los
Angeles Harbor. Aerosol Science and Technology 43 (2), 145-160.

Arredouani, M.S., Palecanda, A., Koziel, H., Huang, Y.C., Imrich, A., Sulahian, T.H.,
Ning, Y.Y., Yang, Z.P., Pikkarainen, T., Sankala, M., Vargas, S.O., Takeya, M.,
Tryggvason, K., Kobzik, L., 2005. MARCO is the major binding receptor for
unopsonized particles and bacteria on human alveolar macrophages. Journal of
Immunology 175 (9), 6058-6064.

Artaxo, P., Gerab, F., Yamasoe, M.A., Martins, J.V., 1994. Fine Mode Aerosol
Composition at 3 Long-Term Atmospheric Monitoring Sites in the Amazon Basin.
Journal of Geophysical Research-Atmospheres 99 (D11), 22857-22868.

Aschner, M., Erikson, K.M., Dorman, D.C., 2005. Manganese dosimetry: Species
differences and implications for neurotoxicity. Critical Reviews in Toxicology 35 (1), 1-
32.

Ayala, A., Kado, N.Y., Okamoto, R.A., Holmen, B.A., Kuzmicky, P.A., Kobayashi, R.,
Stiglitz, K.E., 2002. Diesel and heavy-duty transit bus emissions over multiple driving
schedules: regulated pollutants and project overview. Society of Automotive Engineers
Technical Paper Series No. 2002-01-1722.

Ayala, A.K.N., Okamoto, R.A., Holmen, B.A., Kuzmicky, P.A., Kobayashi, R., Stiglitz,
K.E., 2002. Diesel and CNG heavy-duty transit bus emissions over multiple driving
schedules: regulated pollutants and project overview. SAE Technical Paper No. 2002-01-
1722.

Baltensperger, U., Dommen, J., Alfarra, R., Duplissy, J., Gaeggeler, K., Metzger, A.,
Facchini, M.C., Decesari, S., Finessi, E., Reinnig, C., Schott, M., Warnke, J., Hoffmann,
T., Klatzer, B., Puxbaum, H., Geiser, M., Savi, M., Lang, D., Kalberer, M., Geiser, T.,
2008. Combined determination of the chemical composition and of health effects of
secondary organic aerosols: The POLYSOA project. Journal of Aerosol Medicine and
Pulmonary Drug Delivery 21 (1), 145-154.

155

Baron, P.A., Willeke, K., 2001. Aerosol measurement; principles, techniques, and
application. NewYork: Wiley Interscience Publication.
Barrett, E.G., Henson, R.D., Seilkop, S.K., McDonald, J.D., Reed, M.D., 2006. Effects of
hardwood smoke exposure on allergic airway inflammation in mice. Inhalation
Toxicology 18 (1), 33-43.

Bayona, J.M., Markides, K.E., Lee, M.L., 1988. Characterization of Polar Polycyclic
Aromatic-Compounds in a Heavy-Duty Diesel Exhaust Particulate by Capillary Column
Gas-Chromatography and High-Resolution Mass-Spectrometry. Environmental Science
& Technology 22 (12), 1440-1447.

Becker, S., Dailey, L.A., Soukup, J.M., Grambow, S.C., Devlin, R.B., Huang, Y.C.T.,
2005. Seasonal variations in air pollution particle-induced inflammatory mediator release
and oxidative stress. Environmental Health Perspectives 113 (8), 1032-1038.

Biswas, S., Verma, V., Schauer, J.J., Cassee, F.R., Cho, A.K., Sioutas, C., 2009.
Oxidative potential of semi-volatile and non-volatile particulate matter (PM) from heavy-
duty vehicles retrofitted with emission control technologies. Environmental Science &
Technology 43 (10), 3905-3912.

Castranova, V., Ma, J.Y.C., Yang, H.M., Antonini, J.M., Butterworth, L., Barger, M.W.,
Roberts, J., Ma, J.K.H., 2001. Effect of exposure to diesel exhaust particles on the
susceptibility of the lung to infection. Environmental Health Perspectives 109 609-612.

Chen, L.C., Lippmann, M., 2009. Effects of metals within ambient air particulate matter
(PM) on human health. Inhalation Toxicology 21 (1), 1-31.

Cheung, K.L., Ntziachristos, L., Tzamkiozis, T., Schauer, J.J., Samaras, Z., Moore, K.F.,
Sioutas, C., 2010. Emissions of particulate trace elements, metals and organic species
from gasoline, diesel, and biodiesel passenger vehicles and their relation to oxidative
potential. Aerosol Science and Technology 44 (7), 500-513.

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. Environmental Research 99 (1), 40-47.

Ciapetti, G., Granchi, D., Verri, E., Savarino, L., Cenni, E., Savioli, F., Pizzoferrato, A.,
1998. Fluorescent microplate assay for respiratory burst of PMNs challenged in vitro
with orthopedic metals. Journal of Biomedical Materials Research 41 (3), 455-460.

Cruts, B., van Etten, L., Toernqvist, H., Blomberg, A., Sandstrom, T., Mills, N.L., Borm,
P.J.A., 2008. Exposure to diesel exhaust induces changes in EEG in human volunteers.
Particle and Fibre Toxicology 5 (4), doi:10.1186/1743-8977-1185-1184.

156

Cruz, C.N., Pandis, S.N., 1998. The effect of organic coatings on the cloud condensation
nuclei activation of inorganic atmospheric aerosol. Journal of Geophysical Research-
Atmospheres 103 (D11), 13111-13123.

Dachs, J., Eisenreich, S.J., 2000. Adsorption onto aerosol soot carbon dominates gas-
particle partitioning of polycyclic aromatic hydrocarbons. Environmental Science &
Technology 34 (17), 3690-3697.

Decesari, S., Facchini, M.C., Fuzzi, S., Tagliavini, E., 2000. Characterization of water-
soluble organic compounds in atmospheric aerosol: A new approach. J. Geophys. Res-
Atmos. 105 (D1), 1481-1489.

Decesari, S., Facchini, M.C., Matta, E., Lettini, F., Mircea, M., Fuzzi, S., Tagliavini, E.,
Putaud, J.P., 2001. Chemical features and seasonal variation of fine aerosol water-soluble
organic compounds in the Po Valley, Italy. Atmospheric Environment 35 (21), 3691-
3699.

Delfino, R.J., Sioutas, C., Shaista, M., 2005. Potential Role of Ultrafine Particles in
Associations between Airborne Particle Mass and Cardiovascular Health. Environmental
Health Perspectives 113 (8 ), 934–946.

Delfino, R.J., Staimer, N., Tjoa, T., Gillen, D.L., Polidori, A., Arhami, M., Kleinman,
M.T., Vaziri, N.D., Longhurst, J., Sioutas, C., 2009. Air Pollution Exposures and
Circulating Biomarkers of Effect in a Susceptible Population: Clues to Potential Causal
Component Mixtures and Mechanisms. Environmental Health Perspectives 117 ( 8),
1232-1238.

Denkenberger, K.A., Moffet, R.C., Holecek, J.C., Rebotier, T.P., Prather, K.A., 2007.
Real-time, single-particle measurements of oligomers in aged ambient aerosol particles.
Environmental Science & Technology 41 (15), 5439-5446.

DiStefano, E., Eiguren-Fernandez, A., Delfino, R.J., Sioutas, C., Froines, J.R., Cho, A.K.,
2009. Determination of metal-based hydroxyl radical generating capacity of ambient and
diesel exhaust particles. Inhalation Toxicology 21 (9), 731-738.

Docherty, K.S., Stone, E.A., Ulbrich, I.M., DeCarlo, P.F., Snyder, D.C., Schauer, J.J.,
Peltier, R.E., Weber, R.J., Murphy, S.M., Seinfeld, J.H., Grover, B.D., Eatough, D.J.,
Jiimenez, J.L., 2008. Apportionment of Primary and Secondary Organic Aerosols in
Southern California during the 2005 Study of Organic Aerosols in Riverside (SOAR-1).
Environmental Science & Technology 42 (20), 7655-7662.

Donaldson, K., Beswick, P.H., Gilmour, P.S., 1996. Free radical activity associated with
the surface of particles: a unifying factor in determining biological activity? Toxicology
Letters 88 (1-3), 293-298.

157

Donaldson, K., Brown, D., Clouter, A., Duffin, R., MacNee, W., Renwick, L., Tran, L.,
Stone, V., 2002. The pulmonary toxicology of ultrafine particles. Journal of Aerosol
Medicine-Deposition Clearance and Effects in the Lung 15 (2), 213-220.

Donaldson, K., Stone, V., Borm, P.J.A., Jimenez, L.A., Gilmour, P.S., Schins, R.P.F.,
Knaapen, A.M., Rahman, I., Faux, S.P., Brown, D.M., MacNee, W., 2003. Oxidative
stress and calcium signaling in the adverse effects of environmental particles (PM10).
Free Radical Bio Med 34 (11), 1369-1382.

Donaldson, K., Tran, L., Jimenez, L.A., Duffin, R., Newby, D.E., Mills, N., MacNee, W.,
Stone, V., 2005. Combustion-derived nanoparticles: a review of their toxicology
following inhalation exposure. Particle and Fibre Toxicology 2 (10), doi:10.1186/1743-
8977-2-10.

Dorman, D.C., Struve, M.F., James, R.A., Marshall, M.W., Parkinson, C.U., Wong, B.A.,
2001. Influence of particle solubility on the delivery of inhaled manganese to the rat
brain: Manganese sulfate and manganese tetroxide pharmacokinetics following repeated
(14-day) exposure. Toxicology and Applied Pharmacology 170 (2), 79-87.

Dubick, M.A., Carden, S.C., Jordan, B.S., Langlinais, P.C., Mozingo, D.W., 2002.
Indices of antioxidant status in rats subjected to wood smoke inhalation and/or thermal
injury. Toxicology 176 (1-2), 145-157.

Echalar, F., Gaudichet, A., Cachier, H., Artaxo, P., 1995. Aerosol Emissions by Tropical
Forest and Savanna Biomass Burning - Characteristic Trace-Elements and Fluxes.
Geophysical Research Letters 22 (22), 3039-3042.

Engler, C., Rose, D., Wehner, B., Wiedensohler, A., Bruggemann, E., Gnauk, T.,
Spindler, G., Tuch, T., Birmili, W., 2007. Size distributions of non-volatile particle
residuals (Dp < 800 nm) at a rural site in Germany and relation to air mass origin.
Atmospheric Chemistry and Physics 7 (22), 5785-5802.

EPA, 2002. Health assessment document for diesel engine exhaust. EPA/600/8-90/057F;
U.S. Environmental Protection Agency: Washington, DC.

Fackler, J.P., Holah, D.G., 1965. Properties of chromium (II) complexes. I. electronic
spectra of simple salt hydrates. Inorganic Chemistry 4 (7), 954-958.

Fine, P.M., Chakrabarti, B., Krudysz, M., Schauer, J.J., Sioutas, C., 2004. Diurnal
variations of individual organic compound constituents of ultrafine and accumulation
mode particulate matter in the Los Angeles basin. Environ Sci Technol 38 (5), 1296-
1304.

158

Folkmann, J.K., Risom, L., Hansen, C.S., Loft, S., Moller, P., 2007. Oxidatively damaged
DNA and inflammation in the liver of dyslipidemic ApoE(-/-) mice exposed to diesel
exhaust particles. Toxicology 237 (1-3), 134-144.

Foresti, R., Clark, J.E., Green, C.J., Motterlini, R., 1997. Thiol compounds interact with
nitric oxide in regulating heme oxygenase-1 induction in endothelial cells - Involvement
of superoxide and peroxynitrite anions. Journal of Biological Chemistry 272 (29), 18411-
18417.

Gad, S.C., 1989. Acute and chronic systemic chromium toxicity. Science of the Total
Environment 86 (1-2), 149-157.

Garg, B.D., Cadle, S.H., Mulawa, P.A., Groblicki, P.J., Laroo, C., Parr, G.A., 2000.
Brake wear particulate matter emissions. Environmental Science & Technology 34 (21),
4463-4469.

Garshick, E., Laden, F., Hart, J.E., Rosner, B., Smith, T.J., Dockery, D.W., Speizer, F.E.,
2004. Lung cancer in railroad workers exposed to diesel exhaust. Environmental Health
Perspectives 112 (15), 1539-1543.

Gauderman, W.J., Vora, H., McConnell, R., Berhane, K., Gilliland, F., Thomas, D.,
Lurmann, F., Avoli, E., Kunzli, N., Jerrett, M., Peters, J., 2007. Effect of exposure to
traffic on lung development from 10 to 18 years of age: a cohort study. Lancet 369
(9561), 571-577.

Geller, M., Biswas, S., Sioutas, C., 2006. Determination of particle effective density in
urban environments with a differential mobility analyzer and aerosol particle mass
analyzer. Aerosol Science and Technology 40 (9), 709-723.

Grose, M., Sakurai, H., Savstrom, J., Stolzenburg, M.R., Watts, W.F., Morgan, C.G.,
Murray, I.P., Twigg, M.V., Kittelson, D.B., Mcmurry, P.H., 2006. Chemical and physical
properties of ultrafine diesel exhaust particles sampled downstream of a catalytic trap.
Environmental Science & Technology 40 (17), 5502-5507.

Haas, J.R., Northup, A., 2004. Effects of aqueous complexation on reductive
precipitation of uranium by Shewanella putrefaciens. Geochemical Transactions 5 (3),
41-48.

Halliwell, B., Cross, C.E., 1994. Oxygen-derived species - their relation to human-
disease and environmental stress. Environmental Health Perspectives 102, 5-12.

Hameri, K., Vakeva, M., Aalto, P.P., Kulmala, M., Swietlicki, E., Zhou, J., Seidl, W.,
Becker, E., O'Dowd, C.D., 2001. Hygroscopic and CCN properties of aerosol particles in
boreal forests. Tellus Series B-Chemical and Physical Meteorology 53 (4), 359-379.

159

Hansen, K.F., Jensen, M.G., Ezerman, N., 2001. Measurement of the ‘‘true’’ efficiency
of a CRT filter. Report to Danish Road Safety and Transport Agency
/http://www.fstyr.dk/sw24048.aspS.

Hays, M.D., Geron, C.D., Linna, K.J., Smith, N.D., Schauer, J.J., 2002. Speciation of
gas-phase and fine particle emissions from burning of foliar fuels. Environmental Science
& Technology 36 (11), 2281-2295.

Herner, J.H., Hu, S., Robertson, W.H., Huai, T., Collins, J.F., Dwyer, H., Ayala, A.,
2009. Effect of advanced aftertreatment for PM and NOx control on heavy-duty diesel
truck emissions. Environmental Science & Technology 43 (15), 5928–5933.

Herner, J.H., Robertson, W.H., Ayala, A., 2007a. Investigation of ultrafine particle
number measurements from a clean diesel truck using the European PMP protocol.
Society of Automotive Engineers Technical Paper Series No. 2007-01-1114.

Herner, J.H., Robertson, W.H., Ayala, A., Chang, O., Biswas, S., Sioutas, C., 2007b.
Nucleation mode particle emissions from in-use heavy duty vehicles equipped with DPF
and SCR retrofits. American Association for Aerosol Research. 26th Annual Aerosol
Conference, September 24–18, 2007 Reno, NV.

Hinds, W.C., 1999. Aerosol technology: properties, behavior, and measurement of
airborne particles New York: Wiley Interscience Publication.

Hiura, T.S., Kaszubowski, M.P., Li, N., Nel, A.E., 1999. Chemicals in diesel exhaust
particles generate reactive oxygen radicals and induce apoptosis in macrophages. J
Immunol 163 (10), 5582-5591.

Hoek, G., Brunekreef, B., Goldbohm, S., Fischer, P., van den Brandt, P.A., 2002.
Association between mortality and indicators of traffic-related air pollution in the
Netherlands: a cohort study. Lancet 360 (9341), 1203-1209.

Hoffmann, M.R., Webb, S.M., Siefert, R.S., 1997. Determination of photochemically
available iron in ambient aerosols. Abstracts of Papers of the American Chemical Society
213 115-GEOC.

Hu, S., Polidori, A., Arhami, M., Shafer, M.M., Schauer, J.J., Cho, A., Sioutas, C., 2008.
Redox activity and chemical speciation of size fractioned PM in the communities of the
Los Angeles-Long Beach harbor. Atmospheric Chemistry and Physics 8 (21), 6439-6451.

Hu, S.H., Herner, J.D., Shafer, M., Robertson, W., Schauer, J.J., Dwyer, H., Collins, J.,
Huai, T., Ayala, A., 2009. Metals emitted from heavy-duty diesel vehicles equipped with
advanced PM and NOx emission controls. Atmospheric Environment 43 (18), 2950-
2959.

160

Hughes, L.S., Cass, G.R., Gone, J., Ames, M., Olmez, I., 1998. Physical and chemical
characterization of atmospheric ultrafine particles in the Los Angeles area.
Environmental Science & Technology 32 (9), 1153-1161.

Jordan, T.B., Seen, A.J., Jacobsen, G.E., 2006. Levoglucosan as an atmospheric tracer for
woodsmoke. Atmospheric Environment 40 (27), 5316-5321.

Khlystov, A., Zhang, Q., Jimenez, J.L., Stanier, C., Pandis, S.N., Canagaratna, M.R.,
Fine, P., Misra, C., Sioutas, C., 2005. In situ concentration of semi-volatile aerosol using
water-condensation technology. Journal of Aerosol Science 36 (7), 866-880.

Kim, S., Chang, M.C., Kim, D., Sioutas, C., 2000. A new generation of portable coarse,
fine, and ultrafine particle concentrators for use in inhalation toxicology. Inhalation
Toxicology 12, 121-137.

Kim, S., Jaques, P.A., Chang, M.C., Barone, T., Xiong, C., Friedlander, S.K., 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 II: field evaluation. Journal of Aerosol Science 32 (11), 1299-1314.

Kim, S., Jaques, P.A., Chang, M.C., Froines, J.R., 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 I:
development and laboratory characterization. Journal of Aerosol Science 32 (11), 1281-
1297.

Kittelson, D.B., 1998. Engines and nanoparticles: A review. Journal of Aerosol Science
29 (5-6), 575-588.

Kittelson, D.B., Watts, W.F., Johnson, J.P., Rowntree, C., Payne, M., Goodier, S.,
Warrens, C., Preston, H., Zink, U., Ortiz, M., Goersmann, C., Twigg, M.V., Walker,
A.P., Caldow, R., 2006. On-road evaluation of two Diesel exhaust aftertreatment devices.
Journal of Aerosol Science 37 (9), 1140-1151.

Kleeman, M.J., Schauer, J.J., Cass, G.R., 1999. Size and composition distribution of fine
particulate matter emitted from wood burning, meat charbroiling, and cigarettes.
Environmental Science & Technology 33 (20), 3516-3523.

Koike, E., Hirano, S., Shimojo, N., Kobayashi, T., 2002. cDNA microarray analysis of
gene expression in rat alveolar macrophages in response to organic extract of diesel
exhaust particles. Toxicological Sciences 67 (2), 241-246.

161

Kuhn, T., Biswas, S., Fine, P.M., Geller, M., Sioutas, C., 2005. Physical and chemical
characteristics and volatility of PM in the proximity of a light-duty vehicle freeway.
Aerosol Science and Technology 39 (4), 347-357.

Kumagai, Y., Arimoto, T., Shinyashiki, M., Shimojo, N., Nakai, Y., Yoshikawa, T.,
Sagai, M., 1997. Generation of reactive oxygen species during interaction of diesel
exhaust particle components with NADPH-cytochrome P450 reductase and involvement
of the bioactivation in the DNA damage. Free Radical Biology and Medicine 22 (3), 479-
487.

Kunzli, N., Mudway, I.S., Gotschi, T., Shi, T.M., Kelly, F.J., Cook, S., Burney, P.,
Forsberg, B., Gauderman, J.W., Hazenkamp, M.E., Heinrich, J., Jarvis, D., Norback, D.,
Payo-Losa, F., Poli, A., Sunyer, J., Borm, P.J.A., 2006. Comparison of oxidative
properties, light absorbance, and total and elemental mass concentration of ambient
PM2.5 collected at 20 European sites. Environmental Health Perspectives 114 (5), 684-
690.

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 Science and Technology 42 (11), 946-957.

Li, N., Sioutas, C., Cho, A., Schmitz, D., Misra, C., Sempf, J., Wang, M.Y., Oberley, T.,
Froines, J., Nel, A., 2003. Ultrafine particulate pollutants induce oxidative stress and
mitochondrial damage. Environmental Health Perspectives 111 (4), 455-460.

Li, N., Wang, M.Y., Oberley, T.D., Sempf, J.M., Nel, A.E., 2002. Comparison of the pro-
oxidative and proinflammatory effects of organic diesel exhaust particle chemicals in
bronchial epithelial cells and macrophages. J Immunol 169 (8), 4531-4541.

Li, Q.F., Wyatt, A., Kamens, R.M., 2009. Oxidant generation and toxicity enhancement
of aged-diesel exhaust. Atmospheric Environment 43 (5), 1037-1042.

Liu, Z.G., Berg, D.R., Swor, T.A., Schauer, J.J., 2008. Comparative analysis on the
effects of diesel particulate filter and selective catalytic reduction systems on a wide
spectrum of chemical species emissions. Environmental Science & Technology 42 (16),
6080-6085.

Liu, Z.G., Victoria, N.V., Swor, T., Kittelson, D.B., 2007. Significance of fuel sulfur
content and dilution conditions on particle emissions from a heavily-used diesel engine
during transient operation. Society of Automotive Engineers Technical Paper Series No.
2007-01-0319.

162

Lough, G.C., Schauer, J.J., Park, J.S., Shafer, M.M., Deminter, J.T., Weinstein, J.P.,
2005. Emissions of metals associated with motor vehicle roadways. Environmental
Science & Technology 39 (3), 826-836.

Majestic, B.J., Schauer, J.J., Shafer, M.M., 2007. Application of synchrotron radiation for
measurement of iron red-ox speciation in atmospherically processed aerosols.
Atmospheric Chemistry and Physics 7 (10), 2475-2487.

Matter, U., Siegmann, H.C., Burtscher, H., 1999. Dynamic field measurements of
submicron particles from diesel engines. Environmental Science & Technology 33 (11),
1946-1952.

McGeehan, J.A., Yeh, S., Couch, M., Hinz, A., Otterholm, B., Walker, A.P., Blakeman,
P., 2005. On the road to 2010 emissions: field test results and analysis with DPF–SCR
system and ultra-low-sulfur diesel fuel. SAE Technical Paper 2005-01-3716.

Merrion, D., 2003. Heavy duty diesel emission regulations—past, present, and future.
Society of Automotive Engineers Technical Paper Series No. 2003-01-0040.

Miguel, A.H., Friedlander, S.K., 1978. Distribution of benzo[a]pyrene and coronene with
respect to particle size in Pasadena aerosols in the submicron range. Atmospheric
Environment 12 2407-2413.

Miracolo, M.A., Presto, A.A., Lambe, A.T., Hennigan, C.J., Donahue, N.M., Kroll, J.H.,
Worsnop, D.R., Robinson, A.L., 2010. Photo-oxidation of low-volatility organics found
in motor vehicle emissions: production and chemical evolution of organic aerosol mass.
Environmental Science and Technology 44 (5), 1638-1643.

Misra, C., Kim, S., Shen, S., Sioutas, C., 2002. A high flow rate, very low pressure drop
impactor for inertial separation of ultrafine from accumulation mode particles. Journal of
Aerosol Science 33 (5), 735-752.

Moore, K.F., Ning, Z., Ntziachristos, L., Schauer, J.J., Sioutas, C., 2007. Daily variation
in the properties of urban ultrafine aerosol - part I: physical characterization and
volatility. Atmospheric Environment 41 (38), 8633-8646.

Naeher, L.P., Brauer, M., Lipsett, M., Zelikoff, J.T., Simpson, C.D., Koenig, J.Q., Smith,
K.R., 2007. Woodsmoke health effects: A review. Inhalation Toxicology 19 (1), 67-106.

Nico, P.S., Kumfer, B.M., Kennedy, I.M., Anastasio, C., 2009. Redox dynamics of mixed
metal (Mn, Cr, and Fe) ultrafine particles. Aerosol Science and Technology 43 (1), 60-70.

163

Ning, Z., Geller, M.D., Moore, K.F., Sheesley, R., Schauer, J.J., Sioutas, C., 2007. Daily
variation in chemical characteristics of urban ultrafine aerosols and inference of their
sources. Environmental Science & Technology 41 (17), 6000-6006.

Ning, Z., Moore, K.F., Polidori, A., Sioutas, C., 2006. Field validation of the new
miniature Versatile Aerosol Concentration Enrichment System (mVACES). Aerosol
Science and Technology 40 (12), 1098-1110.

Novakov, T., Corrigan, C.E., 1996. Cloud condensation nucleus activity of the organic
component of biomass smoke particles. Geophysical Research Letters 23 (16), 2141-
2144.

Ntziachristos, L., Froines, J.R., Cho, A.K., Sioutas, C., 2007. Relationship between redox
activity and chemical speciation of size-fractionated particulate matter. Particle and Fibre
Toxicology 4 (5), doi:10.1186/1743-8977-4-5.

Obrien, P.J., 1991. Molecular mechanisms of quinone cytotoxicity. Chemico-Biological
Interactions 80 (1), 1-41.

Pakbin, P., Ning, Z., Schauer, J.J., Sioutas, C., 2009. Characterization of particle bound
organic carbon from diesel vehicles equipped with advanced emission control
technologies. Environmental Science & Technology 43 (13), 4679-4686.

Perraudin, E., Budzinski, H., Villenave, E., 2005. Analysis of polycyclic aromatic
hydrocarbons adsorbed on particles of atmospheric interest using pressurised fluid
extraction. Analytical and Bioanalytical Chemistry 383 (1), 122-131.

Peters, A., Veronesi, B., Calderón-Garcidueñas, L., Gehr, P., Chen, L.C., Geiser, M.,
Reed, W., Rothen-Rutishauser, B., Schürch, S., Schulz, H., 2006. Translocation and
potential neurological effects of fine and ultrafine particles a critical update. Particle and
Fibre Toxicology 3 (13), doi:10.1186/1743-8977-3-13.

Phuleria, H.C., Fine, P.M., Zhu, Y.F., Sioutas, C., 2005. Air quality impacts of the
October 2003 Southern California wildfires. Journal of Geophysical Research-
Atmospheres 110 (D7), D07S20, doi:10.1029/2004JD004626.

Pierce, R.C., Katz, M., 1975. Dependency of Polynuclear Aromatic Hydrocarbon Content
on Size Distribution of Atmospheric Aerosols. Environ. Sci. Technol. 9 (4), 347-353.

Polidori, A., Hu, S., Biswas, S., Delfino, R.J., Sioutas, C., 2008. Real-time
characterization of particle-bound polycyclic aromatic hydrocarbons in ambient aerosols
and from motor-vehicle exhaust. Atmospheric Chemistry and Physics 8 (5), 1277-1291.

164

Reaney, S.H., Smith, D.R., 2005. Manganese oxidation state mediates toxicity in PC12
cells. Toxicology and Applied Pharmacology 205 (3), 271-281.

Riediker, M., Cascio, W.E., Griggs, T.R., Herbst, M.C., Bromberg, P.A., Neas, L.,
Williams, R.W., Devlin, R.B., 2004. Particulate matter exposure in cars is associated with
cardiovascular effects in healthy young men. American Journal of Respiratory and
Critical Care Medicine 169 (8), 934-940.

Riipinen, I., Pierce, J.R., Donahue, N.M., Pandis, S.N., 2010. Equilibration time scales of
organic aerosol inside thermodenuders: evaporation kinetics versus thermodynamics.
Atmospheric Environment 44, 597- 607.

Robinson, A.L., Donahue, N.M., Shrivastava, M.K., Weitkamp, E.A., Sage, A.M.,
Grieshop, A.P., Lane, T.E., Pierce, J.R., Pandis, S.N., 2007. Rethinking organic aerosols:
semivolatile emissions and photochemical aging. Science 315 (5816), 1259-1262.

Rodrigueza, C.E., Fukutoa, J.M., Taguchic, K., Froines, J., Cho, A.K., 2005. The
interactions of 9,10-phenanthrenequinone with glyceraldehyde-3-phosphate
dehydrogenase (GAPDH), a potential site for toxic actions. Chemico-Biological
Interactions 155, 97-110.

Rose, A.L., Waite, T.D., 2005. Reduction of organically complexed ferric iron by
superoxide in a simulated natural water. Environmental Science & Technology 39 (8),
2645-2650.

Rusznak, C., Devalia, J.L., Davies, R.J., 1994. The Impact of Pollution on Allergic
Disease. Allergy 49 (18), 21-27.

Saarikoski, S., Sillanpaa, M., Sofiev, M., Timonen, H., Saarnio, K., Teinela, K.,
Karppinen, A., Kukkonen, J., Hillamo, R., 2007. Chemical composition of aerosols
during a major biomass burning episode over northern Europe in spring 2006:
Experimental and modelling assessments. Atmospheric Environment 41 (17), 3577-3589.

Saldiva, P.H.N., Clarke, R.W., Coull, B.A., Stearns, R.C., Lawrence, J., Murthy, G.G.K.,
Diaz, E., Koutrakis, P., Suh, H., Tsuda, A., Godleski, J.J., 2002. Lung inflammation
induced by concentrated ambient air particles is related to particle composition. American
Journal of Respiratory and Critical Care Medicine 165 (12), 1610-1617.

Samet, J.M., Dominici, F., Curriero, F.C., Coursac, I., Zeger, S.L., 2000. Fine particulate
air pollution and mortality in 20 US Cities, 1987-1994. New England Journal of
Medicine 343 (24), 1742-1749.

165

Sanders, P.G., Xu, N., Dalka, T.M., Maricq, M.M., 2003. Airborne brake wear debris:
Size distributions, composition, and a comparison of dynamometer and vehicle tests.
Environmental Science & Technology 37 (18), 4060-4069.

Sardar, S.B., Fine, P.M., Mayo, P.R., Sioutas, C., 2005. Size-fractionated measurements
of ambient ultrafine particle chemical composition in Los Angeles using the
NanoMOUDI. Environmental Science & Technology 39 (4), 932-944.

Schauer, J.J., 2003. Evaluation of elemental carbon as a marker for diesel particulate
matter. Journal of Exposure Analysis and Environmental Epidemiology 13 (6), 443-453.

Schauer, J.J., Kleeman, M.J., Cass, G.R., Simoneit, B.R.T., 2001. Measurement of
emissions from air pollution sources. 3. C-1-C-29 organic compounds from fireplace
combustion of wood. Environmental Science & Technology 35 (9), 1716-1728.

Schauer, J.J., Mader, B.T., Deminter, J.T., Heidemann, G., Bae, M.S., Seinfeld, J.H.,
Flagan, R.C., Cary, R.A., Smith, D., Huebert, B.J., Bertram, T., Howell, S., Kline, J.T.,
Quinn, P., Bates, T., Turpin, B., Lim, H.J., Yu, J.Z., Yang, H., Keywood, M.D., 2003.
ACE-Asia intercomparison of a thermal-optical method for the determination of particle-
phase organic and elemental carbon. Environmental Science & Technology 37 (5), 993-
1001.

Schoonen, M.A.A., Cohn, C.A., Roemer, E., Laffers, R., Simon, S.R., O'Riordan, T.,
2006. Mineral-induced formation of reactive oxygen species. Medical Mineraology and
Geochemistry 64 179-221.

Seinfeld, J.H., Pandis, S.N., 2006. Atmospheric chemistry and physics: from air pollution
to climate change. New Jersey: John Wiley and Sons, Inc.

Shafer, M.M., Perkins, D.A., Antkiewicz, D.S., Stone, E.A., Quraishi, T.A., Schauer, J.J.,
2010. Reactive oxygen species activity and chemical speciation of size-fractionated
atmospheric particulate matter from Lahore, Pakistan: an important role for transition
metals. Journal of Environmental Monitoring 12 (3), 704-715.

Sheesley, R.J., Schauer, J.J., Garshick, E., Laden, F., Smith, T.J., Blicharz, A.P.,
Deminter, J.T., 2009. Tracking personal exposure to particulate diesel exhaust in a diesel
freight terminal using organic tracer analysis. Journal of Exposure Science and
Environmental Epidemiology 19 (2), 172-186.

Shinyashiki, M., Eiguren-Fernandez, A., Schmitz, D.A., Di Stefano, E., Li, N., Linak,
W.P., Cho, S.H., Froines, J.R., Cho, A.K., 2009. Electrophilic and redox properties of
diesel exhaust particles. Environmental Research 109 (3), 239-244.

166

Siefert, R.L., Pehkonen, S.O., Erel, Y., Hoffmann, M.R., 1994. Iron Photochemistry of
Aqueous Suspensions of Ambient Aerosol with Added Organic-Acids. Geochimica Et
Cosmochimica Acta 58 (15), 3271-3279.

Siefert, R.L., Webb, S.M., Hoffmann, M.R., 1996. Determination of photochemically
available iron in ambient aerosols. Journal of Geophysical Research-Atmospheres 101
(D9), 14441-14449.

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. Atmospheric Environment 33 (2), 173-182.

Singh, M., Jaques, P.A., Sioutas, C., 2002. Size distribution and diurnal characteristics of
particle-bound metals in source and receptor sites of the Los Angeles Basin. Atmospheric
Environment 36 (10), 1675-1689.

Stayner, L., Dankovic, D., Smith, R., Steenland, K., 1998. Predicted lung cancer risk
among miners exposed to diesel exhaust particles. American Journal of Industrial
Medicine 34 (3), 207-219.

Steerenberg, P.A., Zonnenberg, J.A.J., Dormans, J.A.M.A., Joon, P.N.T., Wouters, I.M.,
van Bree, L., Scheepers, P.T.J., Van Loveren, H., 1998. Diesel exhaust particles induced
release of interleukin 6 and 8 by (primed) human bronchial epithelial cells (BEAS 2B) in
vitro. Experimental Lung Research 24 (1), 85-100.

Sternbeck, J., Sjodin, A., Andreasson, K., 2002. Metal emissions from road traffic and
the influence of resuspension - results from two tunnel studies. Atmospheric Environment
36 (30), 4735-4744.

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. Atmospheric Chemistry and Physics 8 (5), 1249-1259.

Subramanyam, V., Valsaraj, K.T., Thibodeaux, L.J., Reible, D.D., 1994. Gas-to-particle
partitioning of polycyclic aromatic hydrocarbons in an urban atmosphere. Atmospheric
Environment 28 (19), 3083-3091.

Thomaidis, N.S., Bakeasa, E.B., Siskos, P.A., 2003. Characterization of lead, cadmium,
arsenic and nickel in PM2.5 particles in the Athens atmosphere, Greece. Chemosphere 52
(6), 959-966.

167

Thompson, K.H., Tsukada, Y., Xu, Z.M., Battell, M., McNeill, J.H., Orvig, C., 2002.
Influence of chelation and oxidation state on vanadium bioavailability, and their effects
on tissue concentrations of zinc, copper, and iron. Biological Trace Element Research 86
(1), 31-44.

Thorning, D.R., Howard, M.L., Hudson, L.D., Schumacher, R.L., 1982. Pulmonary
Responses to Smoke Inhalation - Morphologic Changes in Rabbits Exposed to Pine
Wood Smoke. Human Pathology 13 (4), 355-364.

Tischer, C.A., Oliveira, M.B.M., Reicher, F., Iacomini, M., Petkowicz, C.L.O., Duarte,
M.E.R., Noseda, M.D., Gorin, P.A.J., 2004. Complexation of vanadium (V) oxyanions
with hexopyranose- and mannopyranoseuronic acid-containing polysaccharides:
stereochemical considerations. Carbohydrate Research 339 (4), 771-775.

Tran-Duc, T., Thamwattana, N., Cox, B.J., Hill, J.M., 2010. Adsorption of polycyclic
aromatic hydrocarbons on graphite surfaces. Computational Materials Science 49 (4),
S307-S312.

Vaaraslahti, K., Virtanen, A., Ristimaki, J., Keskinen, J., 2004. Nucleation mode
formation in heavy-duty diesel exhaust with and without a particulate filter. Environ Sci
Technol 38 (18), 4884-4890.

Valko, M., Morris, H., Cronin, M.T.D., 2005. Metals, toxicity and oxidative stress.
Current Medicinal Chemistry 12 (10), 1161-1208.

Venkataraman, C., Thomas, S., Kulkarni, P., 1999. Size distributions of polycyclic
aromatic hydrocarbons - Gas/particle partitioning to urban aerosols. Journal of Aerosol
Science 30 (6), 759-770.

Verma, V., Polidori, A., Schauer, J.J., Shafer, M.M., Cassee, F.R., Sioutas, C., 2009.
Physicochemical and toxicological profiles of particulate matter in Los Angeles during
the October 2007 Southern California wildfires. Environmental Science & Technology
43 (3), 954-960.

Verma, V., Shafer, M.M., Schauer, J.J., Sioutas, C., 2010. Contribution of transition
metals in the reactive oxygen species activity of PM emissions from retrofitted heavy-
duty vehicles. Atmospheric Environment 44 (39), 5165-5173.

Villalobos-Pietrini, R., Amador-Munoz, O., Waliszewski, S., Hernandez-Mena, L.,
Munive-Colin, Z., Gomez-Arroyo, S., Bravo-Cabrera, J.L., Frias-Villegas, A., 2006.
Mutagenicity and polycyclic aromatic hydrocarbons associated with extractable organic
matter from airborne particles <= 10 mu m in southwest Mexico City. Atmospheric
Environment 40 (30), 5845-5857.

168

Wang, C.F., Chang, C.Y., Tsai, S.F., Chiang, H.L., 2005. Characteristics of road dust
from different sampling sites in northern Taiwan. Journal of the Air & Waste
Management Association 55 (8), 1236-1244.

Watanabe, N., Oonuki, Y., 1999. Inhalation of diesel engine exhaust affects
spermatogenesis in growing male rats. Environmental Health Perspectives 107 (7), 539-
544.

Welch, K.D., Davis, T.Z., Van Eden, M.E., Aust, S.D., 2002. Deleterious iron-mediated
oxidation of biomolecules. Free Radical Biology and Medicine 32 (7), 577-583.

Westerdahl, D., Wang, X., Pan, X.C., Zhang, K.M., 2009. Characterization of on-road
vehicle emission factors and microenvironmental air quality in Beijing, China.
Atmospheric Environment 43 (3), 697-705.

Winterbourn, C.C., 1987. The ability of scavengers to distinguish OH. production in the
iron-catalyzed Haber-Weiss reaction - comparison of 4 assays for OH. Free Radical
Biology and Medicine 3 (1), 33-39.

Xia, T., Korge, P., Weiss, J.N., Li, N., Venkatesen, M.I., Sioutas, C., Nel, A., 2004.
Quinones and aromatic chemical compounds in particulate matter induce mitochondrial
dysfunction: Implications for ultrafine particle toxicity. Environmental Health
Perspectives 112 (14), 1347-1358.

Yan, B.Z., M.; Hu,Y. T.; Lee,S.; Kim,H. K.; Russell, A. G.; , 2007. Organic composition
of carbonaceous aerosols in an aged prescribed fire plume Atmos. Chem. Phys. Discuss.
7 18015–18042, 12007.

Yin, X.J.J., Dong, C.C., Ma, J.Y.C., Roberts, J.R., Antonini, J.M., Ma, J.K.H., 2007.
Suppression of phagocytic and bactericidal functions of rat alveolar macrophages by the
organic component of diesel exhaust particles. Journal of Toxicology and Environmental
Health-Part a-Current Issues 70 (9-10), 820-828.

Zhang, Q., Stanier, C.O., Canagaratna, M.R., Jayne, J.T., Worsnop, D.R., Pandis, S.N.,
Jimenez, J.L., 2004a. Insights into the chemistry of new particle formation and growth
events in Pittsburgh based on aerosol mass spectrometry. Environmental Science &
Technology 38 (18), 4797-4809.

Zhang, R.Y., Suh, I., Zhao, J., Zhang, D., Fortner, E.C., Tie, X.X., Molina, L.T., Molina,
M.J., 2004b. Atmospheric new particle formation enhanced by organic acids. Science 304
(5676), 1487-1490.

169

Zhang, Y.X., Schauer, J.J., Shafer, M.M., Hannigan, M.P., Dutton, S.J., 2008. Source
apportionment of in vitro reactive oxygen species bioassay activity from atmospheric
particulate matter. Environmental Science & Technology 42 (19), 7502-7509.

Zhao, Y., Bein, K.J., Wexler, A.S., Misra, C., Fine, P.M., Sioutas, C., 2005. Field
evaluation of the versatile aerosol concentration enrichment system (VACES) particle
concentrator coupled to the rapid single-particle mass spectrometer (RSMS-3). Journal of
Geophysical Research-Atmospheres 110 (D07S02), doi:10.1029/2004JD004644.

Zheng, M., Fang, M., 2000. Particle-associated polycyclic aromatic hydrocarbons in the
atmosphere of Hong Kong. Water Air and Soil Pollution 117, 175-189.

Zhu, Y.F., Hinds, W.C., Kim, S., Sioutas, C., 2002. Concentration and size distribution of
ultrafine particles near a major highway. Journal of the Air & Waste Management
Association 52 (9), 1032-1042.
170

Appendix

Peer-reviewed Publications List

1. Verma, V., Pakbin, P., Cheung, K.L., Cho, A.K., Schauer, J.J, Shafer, M.M,
Kleinman, M.T., and Sioutas, C., 2010. Physicochemical and oxidative
characteristics of semi-volatile components of quasi-ultrafine particles in an urban
atmosphere. Atmospheric Environment doi:10.1016/j.atmosenv.2010.10.044 (in
press).

2. Verma, V., Shafer, M.M., Schauer, J.J. and Sioutas, C., 2010. Contribution of
transition metals in the reactive oxygen species activity of PM emissions from
retrofitted heavy-duty vehicles. Atmospheric Environment, 44 (39), 5165-5173.

3. Moore, K.F., Verma, V., Minguillon, M.C., Sioutas, C., 2010. Inter- and intra-
community variability in continuous coarse particulate matter (PM
10-2.5
)
concentrations in the Los Angeles area. Aerosol Science and Technology 44 (7), 526-
540.

4. 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. Atmospheric Environment 43 (40), 6360-6368.

5. Biswas, S., Verma, V., Schauer, J.J., Cassee, F.R., Cho, A.K., Sioutas, C., 2009.
Oxidative potential of semi-volatile and non volatile particulate matter (PM) from
heavy-duty vehicles retrofitted with emission control technologies. Environmental
Science & Technology 43 (10), 3905-3912.

6. Verma, V., Polidori, A., Schauer, J.J., Shafer, M.M., Cassee, F.R., Sioutas, C., 2009.
Physicochemical and toxicological profiles of particulate matter in Los Angeles
during the October 2007 southern California wildfires. Environmental Science &
Technology 43 (3), 954-960.

7. Biswas, S., Verma, V., Schauer, J.J., Sioutas, C., 2009. Chemical speciation of PM
emissions from heavy-duty diesel vehicles equipped with diesel particulate filter
(DPF) and selective catalytic reduction (SCR) retrofits. Atmospheric Environment 43
(11), 1917-1925.

8. Biswas, S., Hu, S.H., Verma, V., Herner, J.D., Robertson, W.H., Ayala, A., Sioutas,
C., 2008. Physical properties of particulate matter (PM) from late model heavy-duty
diesel vehicles operating with advanced PM and NOx emission control technologies.
Atmospheric Environment 42 (22), 5622-5634.

Abstract (if available)

Abstract A number of population based epidemiological studies as well as recent toxicological and clinical studies indicate a strong association between particulate matter (PM) exposure and adverse health outcomes. Despite commendable progress in particle-related toxicological research for the last few decades, the exact mechanisms by which PM inflicts health injuries are still largely unknown and constitute a subject of great interest for the scientific community. An increase in the abundance of reactive oxygen species (ROS) in biological systems after PM exposure and the resulting oxidative stress has been hypothesized to be mostly responsible for the initiation of inflammatory cascades.

Linked assets

University of Southern California Dissertations and Theses

Conceptually similar

PDF

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

PDF

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

PDF

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

PDF

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

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

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

PDF

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

PDF

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...

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

PDF

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

PDF

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

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

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

PDF

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

PDF

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

PDF

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

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

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

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

Asset Metadata

Creator Verma, Vishal (author)

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

School Viterbi School of Engineering

Degree Doctor of Philosophy

Degree Program Environmental Engineering

Publication Date 01/20/2011

Defense Date 12/01/2010

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

Tag diesel exhaust particles,dithiothreitol (DTT) assay,OAI-PMH Harvest,organic compounds,reactive oxygen species,transition metals,urban aerosols

Place Name California (states), Los Angeles (city or populated place)

Language English

Contributor Electronically uploaded by the author (provenance)

Advisor Sioutas, Constantinos (committee chair), Henry, Ronald C. (committee member), Ning, Zhi (committee member), Petasis, Nicos A. (committee member)

Creator Email india.verma@gmail.com,vverma@usc.edu

Permanent Link (DOI) https://doi.org/10.25549/usctheses-m3617

Unique identifier UC194960

Identifier etd-Verma-4267 (filename),usctheses-m40 (legacy collection record id),usctheses-c127-434129 (legacy record id),usctheses-m3617 (legacy record id)

Legacy Identifier etd-Verma-4267.pdf

Dmrecord 434129

Document Type Dissertation

Rights Verma, Vishal

Type texts

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

Repository Name Libraries, University of Southern California

Repository Location Los Angeles, California

Repository Email cisadmin@lib.usc.edu