The application of a profluorescent nitroxide probe to detect reactive oxygen species derived from combustion generated particulate matter

Considering that this was the first time that a profluorescent nitroxide probe was applied in investigating the oxidative stress potential of PM, the proof of concept regarding the detec

Queensland University of Technology

Discipline of Physics and Chemistry

THE APPLICATION OF A PROFLUORESCENT NITROXIDE PROBE

TO DETECT REACTIVE OXYGEN SPECIES DERIVED FROM COMBUSTION-GENERATED PARTICULATE MATTER

Branka Miljevic

A THESIS SUBMITTED TO THE QUEENSLAND UNIVERSITY OF TECHNOLOGY IN FULFILMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

September 2010

ABSTRACT

Particulate pollution has been widely recognised as an important risk factor

to human health In addition to increases in respiratory and cardiovascular morbidity associated with exposure to particulate matter (PM), WHO estimates that urban PM causes 0.8 million premature deaths globally and that 1.5 million people die prematurely from exposure to indoor smoke generated from the combustion of solid fuels Despite the availability of a huge body of research, the underlying toxicological mechanisms by which particles induce adverse health effects are not yet entirely understood Oxidative stress caused by generation of free radicals and related reactive oxygen species (ROS) at the sites of deposition has been proposed

as a mechanism for many of the adverse health outcomes associated with exposure

to PM In addition to particle-induced generation of ROS in lung tissue cells, several recent studies have shown that particles may also contain ROS As such, they present a direct cause of oxidative stress and related adverse health effects

Cellular responses to oxidative stress have been widely investigated using various cell exposure assays However, for a rapid screening of the oxidative potential of PM, less time-consuming and less expensive, cell-free assays are needed The main aim of this research project was to investigate the application of

a novel profluorescent nitroxide probe, synthesised at QUT, as a rapid screening assay in assessing the oxidative potential of PM Considering that this was the first time that a profluorescent nitroxide probe was applied in investigating the oxidative stress potential of PM, the proof of concept regarding the detection of PM–derived ROS by using such probes needed to be demonstrated and a sampling methodology needed to be developed Sampling through an impinger containing profluorescent nitroxide solution was chosen as a means of particle collection as it allowed particles to react with the profluorescent nitroxide probe during sampling, avoiding

in that way any possible chemical changes resulting from delays between the sampling and the analysis of the PM Among several profluorescent nitroxide probes available at QUT, bis(phenylethynyl)anthracene-nitroxide (BPEAnit) was found to be the most suitable probe, mainly due to relatively long excitation and emission wavelengths (λex= 430 nm; λem= 485 and 513 nm) These wavelengths are

long enough to avoid overlap with the background fluorescence coming from light absorbing compounds which may be present in PM (e.g polycyclic aromatic hydrocarbons and their derivatives) Given that combustion, in general, is one of the major sources of ambient PM, this project aimed at getting an insight into the oxidative stress potential of combustion-generated PM, namely cigarette smoke, diesel exhaust and wood smoke PM

During the course of this research project, it was demonstrated that the BPEAnit probe based assay is sufficiently sensitive and robust enough to be applied

as a rapid screening test for PM-derived ROS detection Considering that for all three aerosol sources (i.e cigarette smoke, diesel exhaust and wood smoke) the same assay was applied, the results presented in this thesis allow direct comparison

of the oxidative potential measured for all three sources of PM In summary, it was found that there was a substantial difference between the amounts of ROS per unit

of PM mass (ROS concentration) for particles emitted by different combustion sources For example, particles from cigarette smoke were found to have up to 80 times less ROS per unit of mass than particles produced during logwood combustion For both diesel and wood combustion it has been demonstrated that the type of fuel significantly affects the oxidative potential of the particles emitted Similarly, the operating conditions of the combustion source were also found to affect the oxidative potential of particulate emissions Moreover, this project has demonstrated a strong link between semivolatile (i.e organic) species and ROS and therefore, clearly highlights the importance of semivolatile species in particle-induced toxicity

KEYWORDS

Combustion aerosols, combustion-generated particulate matter, cigarette smoke, wood smoke, diesel exhaust, health aspects of aerosol, health effects of particulate matter, radicals, reactive oxygen species, ROS, oxidative stress, oxidative

potential, inflammatory potential, in vitro, profuorescent nitroxides; BPEAnit;

fluorescence, impinger, bubbling, collection efficiency

STATEMENT OF ORIGINAL AUTHORSHIP

The work contained in this thesis has not been previously submitted to meet requirements for an award at this or any other higher education institution To the best of my knowledge and belief, the thesis contains no material previously published or written by another person except where due reference is made

Signature:

Date:

Table of Contents

ABSTRACT i

KEYWORDS iii

STATEMENT OF ORIGINAL AUTHORSHIP iv

LIST OF PUBLICATIONS ix

LIST OF TABLES x

LIST OF FIGURES x

ABBREVIATIONS xv

ACKNOWLEDGEMENTS xvii

Chapter 1 1

INTRODUCTION 1

1.1 Description of scientific problem investigated 1

1.2 Overall aims of the study 2

1.3 Specific objectives of the study 3

1.4 Account of scientific progress linking the scientific papers 5

Chapter 2 9

LITERATURE REVIEW 9

2.1 Introduction – particles and health effects 9

2.1.1 Summary of epidemiological findings 9

2.1.2 Summary of toxicological findings 10

2.2 Aerosol fundamentals and basic terminology 12

2.3 Combustion-generated PM 15

2.3.1 Cigarette smoke 16

2.3.2 Diesel exhaust PM 17

2.3.3 Wood smoke particles 20

2.4 Particle characteristics relevant for health effects 22

2.4.1 Size and surface area 22

2.4.2 Composition 24

2.5 Measurement of oxidative stress capacity of the PM 28

2.5.1 In vitro studies 28

2.5.2 Cell-free assays 30

2.6 Nitroxides 36

2.6.1 Profluorescent nitroxides 37

2.7 An overview of particle sampling approaches for toxicological studies 42

2.8 Summary of research needs 44

2.9 References 46

Chapter 3 63

ON THE EFFICIENCY OF IMPINGERS WITH FRITTED NOZZLE TIP FOR COLLECTION OF ULTRAFINE PARTICLES 63

Abstract 65

3.1 Introduction 66

3.2 Experimental 67

3.3 Results and discussion 70

3.4 Conclusion 76

3.5 References 77

Chapter 4 79

THE APPLICATION OF PROFLUORESCENT NITROXIDES TO DETECT REACTIVE OXYGEN SPECIES DERIVED FROM COMBUSTION-GENERATED PARTICULATE MATTER: CIGARETTE SMOKE – A CASE STUDY 79

Abstract 81

4.1 Introduction 82

4.2 Experimental 86

4.2.1 Materials 86

4.2.2 Cigarette smoke sampling 86

4.2.3 Fluorescence measurements 89

4.2.4 Calibration curve 90

4.3 Results 90

4.3.1 Mainstream cigarette smoke - linearity 90

4.3 2 Sidestream cigarette smoke - sensitivity 92

4.4 Discussion 96

4.5 Conclusion 98

4.6 References 99

4.7 Supplementary Information 103

4.7.1 Reaction of BPEAnit with peroxyl radicals 103

Chapter 5 105

PARTICLE EMISSIONS, VOLATILITY AND TOXICITY FROM AN ETHANOL FUMIGATED COMPRESSION IGNITION ENGINE 105

Abstract 108

5.1 Introduction 109

5.2 Methodology 111

5.2.1 Engine, fuel and testing specifications 111

5.2.2 Particle measurement methodology 112

5.2.3 Particle volatility methodology 114

5.2.4 ROS concentration measurement – BPEAnit assay 115

5.3 Results 116

5.3.1 Particle size distributions 116

5.3.2 Particle volatility 119

5.3.3 ROS concentration results 122

5.4 Discussion 124

5.5 References 127

5.6 Supporting Information 130

Chapter 6 135

OXIDATIVE POTENTIAL OF LOGWOOD AND PELLET BURNING PARTICLES ASSESSED BY A NOVEL PROFLUORESCENT NITROXIDE PROBE 135

Abstract 138

6.1 Introduction 139

6.2 Experimental 141

6.2.1 Wood burners 141

6.2.2 Sampling setup and instrumentation 142

6.2.3 BPEAnit assay 144

6.3 Results 145

6.3.1 Particle emissions 145

6.3.2 ROS from logwood burning particles 147

6.3.3 Correlation between ROS and organics 148

6.3.4 ROS from pellet burning particles 152

6.4 Discussion 153

6.5 References 157

6.6 Supporting information 160

Chapter 7 163

CONCLUSIONS 163

7.1 Principal significance of findings 164

7.2 Directions for future research 171

7.3 References 173

LIST OF PUBLICATIONS

1 Miljevic, B., Modini, R.L., Bottle, S.E., Ristovski, Z.D., 2009 On the efficiency

of impingers with fritted nozzle tip for collection of ultrafine particles

Atmospheric Environment 43, 1372-1376

2 Miljevic, B., Fairfull-Smith, K.E., Bottle, S.E., Ristovski, Z.D., 2009 The

application of profluorescent nitroxides to detect reactive oxygen species derived from combustion-generated particulate matter: Cigarette smoke – a

case study Atmospheric Environment 44, 2224-2230

3 Surawski, N.C., Miljevic, B., Roberts, B.A., Modini, R.L., Situ, R., Brown, R.J.,

Bottle, S.E., Ristovski, Z.D., 2009 Particle emissions, volatility and toxicity

from an ethanol fumigated compression ignition engine Environmental Science & Technology 44, 229-235

4 Miljevic, B., Heringa, M.F., Keller, A., Meyer, N.K., Good, J., Lauber, A.,

DeCarlo, P.F., Fairfull-Smith, K.E., Nussbaumer, T., Burtscher, H., Prevot, A.S.H., Baltensperger, U., Bottle, S.E., Ristovski, Z.D., 2010 Oxidative potential of logwood and pellet burning particles assessed by a novel

profluorescent nitroxide probe Environmental Science & Technology 44,

6601-6607

LIST OF TABLES

Table 4-1 The amount of ROS per a) cigarette; b) puff (nsamples = 4) 92

Table 5-1 Speed, load and fuel settings used for both experimental campaigns 112

Table S 5-1 Test engine specifications 130

LIST OF FIGURES Figure 2-1 Typical urban PM number (A), surface (B) and volume (C) size distributions Taken from (Seinfeld and Pandis, 2006) 13

Figure 2-2 Typical engine exhaust size distribution; both mass and number weightings are shown (Kittelson, 1998) 18

Figure 2-3 The fractional deposition and mechanism of deposition of inhaled particles of different sizes in each region of the human respiratory tract (TB – tracheobronchial, A – alveolar, NPL – Nasal Pharyngeal Laryngeal; TOTAL – sum of all three curves) based on the ICRP Model (1994) assuming nose breathing The image is adopted from Oberdörster et al., (2007) 23

Figure 2-4 Simplified mechanism of quinoid redox cycling (QH2 – catechol) (Squadrito et al., 2001) 27

Figure 2-5 Scheme of particulate matter (PM) catalysed DTT oxidation 32

Figure 2-6 Hydrolysis of DCFH-DA and ROS-induced oxidation of DCFH 34

Figure 2-7 SOD–mimetic activity of 2,2,6,6-tetramethyl-piperidinoxyl (TPO) 37

Figure 2-8 Reaction of trapping carbon-centred radiclas with nitroxide (3AP) and derivatisation with NDA (Flicker and Green, 2001) 40

Figure 2-9 Structures of some of the profluorescent nitroxides synthesised at QUT In these examples five membered nitroxide ring is covalently fused to: A) 9,10-bis(phenylethynyl)anthracene (BPEA); B)9,10-diphenylanthracene and C) phenanthrene 41

Figure 3-1 Schematic representation of the experimental set-up 69

Figure 3-2 Droplet size distribution generated by bubbling particle-free air at 1 L min-1 through 40 ml of solvent (water, 50% DMSO or cell media) in impinger with frit porosity 1 71

Figure 3-3 Particle losses on the fritted nozzle tip for the impingers used in this study 72

Figure 3-4 Removal efficiency (A) and solvent capture efficiency (B) of the impinger with the porosity grade 1 fritted nozzle tip operating at 1 L min-1 for two different volumes (40 and 20 mL) of heptane and water 74

Figure 3-5 Removal efficiency (A) and solvent capture efficiency (B) of the impingers with the porosity grade 1 and 2 operating at 1 L min-1 for two different volumes (40 and 20 mL) of heptane 75

Figure 3-6 Removal efficiency (A) and solvent capture efficiency (B) of the SKC midget impinger with the coarse pore size fritted nozzle tip operating at 1 L min-1 and containing 10 mL of either heptane or water 76

Figure 4-1 9-(1,1,3,3-tetramethylisoindolin-2-yloxyl-5-ethynyl)-10-(phenylethynyl) anthracene (BPEAnit) 85

Figure 4-2 Experimental set-up for the cigarette mainstream smoke sampling 87

Figure 4-3 Experimental set up for sidestream cigarette smoke sampling 88

Figure 4-4 A) Fluorescence emission of BPEAnit when exposed to puffs of mainstream cigarette smoke coming from one 3R4F cigarette (λex= 430 nm); B) fluorescence intensity at 485 nm plotted against the number of puffs 91

Figure 4-5 A) fluorescence emission of BPEAnit when exposed to the gas phase of sidestream smoke; B) fluorescence emission of BPEAnit when exposed to total sidestream smoke; C) fluorescence intensity at 485 nm plotted against sampling time (λex= 430 nm) 93

Figure 4-6 Left y-axis: An example of size distribution of particles entering the impinger and particles trapped in the impinger; Right y-axis: Collection

efficiency of the impinger used in this study (pore 1) when bubbling particles from 50 to 380 nm in diameter through 20 mL of DMSO at 1 L min-1 94

Figure 4-7 ROS concentration related to sidestream smoke particles Error bars present one standard deviation 95

Figure S 4-1 Time course for the reaction of BPEAnit with peroxyl radicals 104 Figure 5-1 Schematic representation of the experimental configuration used in this study 113

Figure 5-2 SMPS derived particle number distributions at intermediate speed (1700 rpm), full load, for neat diesel (E0) and 40% ethanol (E40) engine operation Error bars denote ± one standard error 117

Figure 5-3 Correlation of particle size (CMD) with the ethanol substitution percentage for tests conducted at 2000 rpm, full load ( =-0.939) Error bars denote ± one standard error 118

Figure 5-4 Brake-specific PM2.5 emissions at intermediate speed (1700 rpm) with various load settings and ethanol substitutions Error bars denote ± one standard error 119

Figure 5-5 Volume fraction remaining (VFR) versus thermodenuder temperature at intermediate speed (1700 rpm) (a) 100% load E0 and E40 (b) 25% load E0 and E20 Note well the linear scale on the ordinate for (a) and the logarithmic scale

on the ordinate for (b) Error bars are calculated using the uncertainties in the diameter measurement 120

Figure 5-6 Percentage of volatile particles at intermediate speed (1700 rpm) and 50%, 25% load and idle mode for various ethanol substitutions Error bars have been calculated using the statistical uncertainty in the counts 121

Figure 5-7 Fluorescence spectra of BPEAnit control (HEPA filtered) and test samples for neat diesel (E0) and 40% ethanol (E40) at intermediate speed (1700 rpm) and full load 122

Figure 5-8 ROS concentrations at intermediate speed (1700 rpm) with various load settings and ethanol substitutions Error bars denote ± one standard error 123

Figure S 5-1 Particle size distribution for different thermodenuder temperatures ( (a) 100% load E0 (b) 100% load E40 (c) 50% load E0 (d) 50% load E40 131

Figure S 5-2 The volume percentage of volatile material that coats non-volatile particles at intermediate speed (1700 rpm) and for various load settings and ethanol substitutions Error bars are calculated using the uncertainties in the diameter measurement 132

Figure S 5-3 A calculation of air available for combustion at intermediate speed (1700 rpm), half load, for various ethanol substitutions Error bars are calculated based using the uncertainties in the fuel consumption and gaseous emissions measurements 132

Figure S 5-4 OH radical emissions from an AVL Boost simulation conducted at 2000 rpm, full load 133

Figure 6-1 A schematic representation of the experimental setup 144

Figure 6-2 Examples of mass concentration (A), number concentration (B) and geometric mean diameter (D(GMD); C) for particles from logwood burning in the traditional logwood stove (black) and the automatic pellet boiler (grey) The graphs show the values after dilution Dotted lines denote the time periods at which samples for each phase of logwood burning were normally taken 147

Figure 6-3 Average ROS concentrations of logwood burning particle emissions for cold and warm start Warm start 1 and 2 present sampling after refilling the stove for the first and second time Error bars present one standard error (n=4) 148

Figure 6-4 Fluorescence emission of BPEAnit when sampling with and without a thermodenuder 149

Figure 6-5 Correlation between the amount of ROS and the amount of organics for start-up phase of cold-start (A), stable phase of cold-start (B) and warm-start (C) logwood burning 151

Figure 6-6 Fluorescence intensity of BPEAnit after sampling emissions from the automatic pellet boiler 152

Figure 6-7 Examples of the temperature in the logwood stove (black) and pellet boiler (grey) during combustion The dips in the logwood stove temperature are due to refilling of the stove 154

Figure 6-8 Correlation between the combustion chamber temperature and ROS concentration for logwood burning Error bars present one standard deviation 155

Figure S 6-1 Size dependent collection efficiency of the impinger used in this study when bubbling aerosol through 20 ml of DMSO and at flow rate of 1 L min-1(adopted from Miljevic et al., 2010 ) 161

ABBREVIATIONS

AAPH – 2,2’-azo-bis-(2-amidinopropane) dihydrochloride

3AP – 3-amino-2,2,5,5,-tetramethyl-1-pyrrolidinyloxy

ATP – adenosine triphosphate

BPEAnit – BPEA-nitroxide or 9,10-bis(phenylethynyl)anthracene-nitroxide

DCFH – 2’,7’– dichlorodihydrofluorescein

DCFH-DA – 2’,7’– dichlorodihydrofluorescein diacetate

DEP – diesel exhaust particles

DHR-6G – dihydrorhodamine–6G

DMPO – 5,5-dimethyl-1-pyrroline-N-oxide

DMSO – dimethyl sulphoxide

DNA – deoxyribonucleic acid

DTNB – 5,5’-dithiobis-2-nitrobenzoic acid

DTT – dithiothreitol

EPR – electron paramagnetic resonance

ETS – environmental tobacco smoke

GC-MS – gas chromatography-mass spectrometry

GM-CSF – granulocyte-macrophage colony-stimulating factor

LC-MS – liquid cromatography-mass spectrometry

MAPK – mitogen-activated protein kinase

NQO1 – NADPH quinine oxidoreductase

OH-1 – heme oxygenase-1

POHPAA – p-hydroxyphenylacetic acid

ROS – reactive oxygen species

SOA – secondary organic aerosol

SOD – superoxide dismutase

SS – sidestream smoke

TNF-α – tumor necrosis factor α

TPO – 2,2,6,6- tetramethyl-piperidinoxyl

VOC – volatile organic compounds

WHO – World Health Organisation

ACKNOWLEDGEMENTS

My deepest gratitude goes to my supervisors, A/Prof Zoran Ristovski and Prof Steven Bottle for providing me with an opportunity to work under their supervision and learn from them Zoran and Steve, thank you for giving me the freedom to explore my own ideas, make decisions and work independently, while still giving me all the supervision and support I needed Thank you for your patience, friendship and continuous encouragement during my PhD journey

I would like to thank Queensland University of Technology for awarding me with the Australian government funded International Postgraduate Research Scholarship It was an honour to be a recipient of this scholarship

I would also like to thank the European Comission project EUROCHAMP for the financial support during the wood burning campaign in Horw, Switzerland I am grateful to Prof Urs Baltensperger for providing me with an opportunity to participate in this campaign

I am thankful to Dr Kathryn Fairfull-Smith for always having the nitroxide ready for me

I would also like to acknowledge the support, friendship and invaluable advices of my colleagues from ILAQH and Bottle group

I thank my parents, sister and grandparents for their love, encouragement and support; I thank my friends back home in Croatia for constantly reminding me

that there's no such place as far away A special thanks goes to all the members of

Sundac and Osterman family for all the help, support and friendship they have given

me during my time in Australia

At last, but far from least, I would like to express my innermost gratefulness

to my husband Senad Dear Seni, if it wasn’t for your love, support and persistent faith in me, my passion for research, for this project, would be drowned by loneliness and nostalgia I dedicate this thesis to you

Chapter 1

INTRODUCTION

1.1 Description of scientific problem investigated

A great number of epidemiological and laboratory studies have shown strong associations between levels of ambient particulate matter (PM) and increased respiratory and cardiovascular disease morbidity and mortality, particularly among individuals with pre-existing cardiopulmonary diseases (Englert, 2004) To develop methods that could help to mitigate the adverse health outcomes induced by PM, it is important to know the PM properties and the mechanism(s) that are responsible for PM toxicity Identification of the PM properties that are the most relevant for promoting adverse health effects is crucial not only for our mechanistic understanding but also for the implementation of strategies for improving air quality Despite the availability of a huge body of research, the underlying toxicological mechanisms by which particles induce adverse health effects are not yet entirely understood Recently, it has become evident that those particles have the ability to generate free radicals and related reactive oxygen species (ROS) These species are responsible for driving oxidative stress at sites of deposition and thereby triggering a cascade of events associated with inflammation and, at higher concentrations, cell death

One of the important aspects of environmental sciences in the last decade was to indentify the physical and chemical characteristics of ambient PM responsible for its health effects and within that scope, particle size, surface area and chemical components, such as metals and certain classes of organics (e.g quinones) have been implicated in PM-induced health effects and more specifically,

in the generation of ROS

ROS can be formed endogenously, by the lung tissue cells, during the phagocytic processes initiated by the presence of PM in the lungs, or by particle-

related chemical species that have the potential to generate ROS In addition to the particle-induced generation of ROS, several recent studies have shown that particles may also contain ROS (so called, exogenous ROS) As such, they present a direct cause of oxidative stress and related adverse health effects and the hypothesis that particles contain or produce ROS is the driving force for this research project

It is a reasonable assumption that exogenous ROS can cause the same responses in the cell as endogenously formed ROS Therefore, a rapid screening assay able to evaluate PM oxidative potential in terms of their inherent ROS would

be beneficial for gaining better understanding about the nature of the particles most relevant for their negative health impact Such a screen would also provide a helpful tool in efforts to further improve air quality and protect public health

Cellular responses to oxidative stress have been widely investigated using various cell exposure assays However, in order to provide a rapid screening test for the oxidative potential of PM, less time-consuming and cheaper, cell-free (or acellular) assays are necessary Several cell-free approaches have been used to explore oxidative potential of PM in a quantitative manner They all have certain limitations, do not provide directly comparable results and, to date, none of these assays has been acknowledged as the best acellular assay and none have yet been widely adopted for investigation of potential PM toxicity

1.2 Overall aims of the study

Taking into account the research problem introduced in the previous section, the main aim of this research project was to develop a methodology that would enable us to apply a novel profluorescent nitroxide probe, synthesised at Queensland University of Technology (QUT), for the detection and quantification of ROS present on and derived from the surface of combustion-generated PM

QUT’s profluorescent nitroxide probes have been used in monitoring polymer degradation and investigating cellular redox status, but have never been applied in

investigating airborne particulate matter Profluorescent nitroxides are a type of compound consisting of a fluorophore linked to a nitroxide-containing ring They have a very low fluorescence emission due to inherent quenching of fluoreophore’s fluorescence by the nitroxide group, but upon radical trapping or redox activity, a strong fluorescence is observed Previously published literature regarding nitroxide chemistry led to the idea of applying profluorescent nitroxides in assessing the oxidative stress potential of PM Using this methodology ROS are not measured directly; their amount in the sample is expressed as the amount of the nitroxide probe that has been transformed to a fluorescent product (calculated using calibration curve; for description see page 88) It should be noted that, in addition

to ROS and other free radicals, profluorescent nitroxides can react with strong oxidants and reductants and if such species are present in PM sample, they might contribute to the overall rise of fluorescence

This was the first time that profluorescent nitroxide probes were applied as

a tool for the assessment of PM’s potential to cause oxidative stress and the key task was primarily to demonstrate the proof of concept regarding the detection of PM–derived ROS by using a profluorescent nitroxide probe

Given that combustion, in general, is one of the major sources of ambient

PM affecting in this manner both indoor (e.g tobacco smoking, wood burning stoves and fireplaces) and outdoor (e.g vehicle emissions, biomass burning) air quality, this project aimed at getting an insight into oxidative stress potential of PM coming from different combustion sources

1.3 Specific objectives of the study

The specific objectives of the study can be summarised as follows:

• To develop a sampling approach suitable for toxicological assessment The main requirement for this type of analysis was to conserve the chemical and surface properties of PM (for example, to avoid aggregation of particles during sampling) Considering the reactivity of ROS, there was also a need to

avoid any possible chemical changes resulting from delays between the sampling and the analysis of PM and to efficiently capture as much of the

particle-related ROS as possible during the sampling It was also important

to have a sampling system that is appropriate for field use

• To select a model combustion-generated aerosol (i.e PM source) needed to demonstrate proof of concept regarding the detection of PM–derived ROS

by using a profluorescent nitroxide probe The main requirement for this objective was to have an aerosol source convenient for laboratory use

• Among several profluorescent probes available at QUT, to select the one that is the most appropriate in the assessment of combustion aerosols The key aspect for this objective was to select a probe that has the excitation and emission wavelength long enough to avoid overlapping with the background fluorescence coming from optically active compounds which may be present in PM

• To perform validation of the probe, especially in terms of linearity of fluorescence response To obtain meaningful quantitative measurements of aerosol activity, the linearity of the fluorescence response of the probe is of crucial importance

• To apply profluorescent nitroxide probe in assessing the oxidative potential

of particles produced by other combustion sources and:

- to investigate oxidative stress potential in relation to different combustion conditions and different fuel characteristics

- to investigate oxidative stress potential in relation to the organic content

of PM

- to provide a comparison of oxidative stress potential of particles coming from different combustion sources

1.4 Account of scientific progress linking the scientific papers

This thesis contains a collection of papers in which the specific aims of the project were addressed These papers have been published or submitted for publication in refereed journals

As stated in the previous section, to perform PM sampling for toxicological assays, such as oxidative stress assays, it is important to choose the appropriate sampling approach For this study, a liquid impingement was thought to be a good sampling approach as it would allow particles to react with the profluorescent

nitroxide probe in a sampling liquid during sampling This avoids any possible

chemical changes resulting from delays between the sampling and the analysis of

PM Therefore, it was decided to sample PM into an impinger containing a profluorescent nitroxide solution Since liquid impingement is not 100% efficient in capturing all of the PM, it was anticipated to increase collection efficiency somehow and for that purpose an impinger with a fritted nozzle tip was chosen A fritted nozzle tip increases the contact surface between the aerosol and the liquid and should, therefore, increase the collection efficiency of the impinger In order to conduct quantitative chemical analysis on the particles collected by the impingers,

it is important to know the portion of the particles being collected in the liquid (i.e collection efficiency) A paper focused on the determination of the efficiency of impingers with fritted nozzle tip for collection of ultrafine and near-ultrafine (diameter < 220 nm) particles and factors influencing the collection efficiency is presented in Chapter 3 It is entitled “On the efficiency of impingers with fritted nozzle tip for collection of ultrafine particles” and has been published in the journal

“Atmospheric Environment” as a technical paper Although the collection

efficiencies presented in this paper are only for particles smaller than 220 nm, the same experimental procedure was used to determine collection efficiency for particles larger than 220 nm (up to 600 nm) The collection efficiency of particles larger than 220 nm in the impinger was needed for other studies presented in this thesis

The second paper in this thesis (presented in Chapter 4) aimed to validate the application of the profluorescent nitroxide probe in the detection and quantification of ROS derived from combustion-generated PM As a model aerosol for this study, a cigarette smoke was chosen Cigarette smoke is one of the most common indoor combustion-generated aerosols and it is easy to generate in a laboratory Among several profluorescent nitroxide probes available at QUT, BPEA-nitroxide (BPEAnit) was selected as the most suitable probe BPEAnit (Figure 1-1) has a fluorescence excitation maximum at 430 nm, and emission maxima at 485 and 513 nm and these wavelengths are long enough to avoid overlap with the background fluorescence coming from optically active compounds which may be present in PM (e.g polycyclic aromatic hydrocarbons and their derivatives) The approach used to test BPEAnit response linearity was to expose the probe to an increasing amount of an aerosol of relatively stable chemical and physical

properties For that purpose, sampling equal volumes of mainstream cigarette

smoke puffs was used To test the sensitivity of this approach when working with much lower concentrations of aerosol than in the case of mainstream cigarette smoke, sidestream cigarette smoke previously introduced into an environmental test chamber was also sampled Finally, the amount of ROS per mg of PM was calculated Overall, BPEAnit was shown to provide a sensitive and quantitative

response related to the oxidative capacity of the PM and these findings presented a

good basis for employing the new BPEAnit probe for the investigation of PM’s oxidative stress potential from other combustion sources This paper is entitled

“The application of profluorescent nitroxides to detect reactive oxygen species derived from combustion-generated particulate matter: Cigarette smoke – a case

study” and has been published in the journal “Atmospheric Environment” as a full

research paper

Figure 1-1 BPEAnit

The third paper in the thesis (Chapter 5) presents the application of the BPEAnit in assessing the oxidative stress potential of PM produced by an ethanol fumigated diesel engine Apart from the results of the BPEAnit assay, this paper presents emissions and volatility data for particles produced by a pre-Euro I diesel engine employing different percentages of ethanol fumigation The title of this paper is “Particle emissions, volatility and toxicity from an ethanol fumigated compression ignition engine” and it has been published in the journal

“Environmental Science and Technology” as a full research paper

Finally, the novel BPEAnit assay was applied to assess oxidative stress potential related to particles produced during biomass combustion by an automatic pellet boiler and a traditional logwood stove under various combustion conditions This work was performed in Horw, Switzerland, in collaboration with Paul Scherrer Institute (Villigen, Switzerland), University of Applied Sciences (Windisch, Switzerland) and Lucerne University of Applied Sciences and Arts (Horw, Switzerland) The role of the organic fraction of PM in its oxidative stress potential was explored and an overall comparison was made with the results obtained in previous two studies The paper presenting these results was entitled “Oxidative potential of log wood and pellet burning particles assessed by a novel profluorescent nitroxide probe” and has been published in the journal

“Environmental Science and Technology” as a full research paper

Chapter 2

LITERATURE REVIEW

2.1 Introduction – particles and health effects

Particulate matter is a mixture of solid and/or liquid particles suspended in the air, which originate from various sources PM generally can be divided into primary PM that is directly emitted into the atmosphere from a source, and secondary PM that is formed from precursor gases by photochemically induced generation of low volatility species that nucleate to form new particles or condense onto pre-existing particles PM comes from both natural and anthropogenic sources and varies in size, shape and composition Natural PM sources include sea/ocean spray, windblown dust, volcanic activity, naturally occurring biomass burning and gas-to-particle conversion of biogenic reactive gases, while anthropogenic PM sources and precursor gases include combustion processes and mechanical processes, such as production of materials (nanotechnology) or abrasion of surfaces (tires and roads) The majority of PM in urban atmosphere comes from combustion sources

2.1.1 Summary of epidemiological findings

The adverse health effects of PM have been extensively studied epidemiologically, in animals and in cells Epidemiological studies have demonstrated that exposure to ambient PM is associated with adverse health outcomes (Dockery et al., 1993; Pope et al., 2002) A great number of epidemiological studies have shown a correlation between short term exposure to elevated ambient PM and daily mortality and hospital admissions due to exacerbation of respiratory diseases like asthma, chronic bronchitis, as well as ischemic heart diseases and strokes (Brunekreef and Holgate, 2002; Pope and

Dockery, 2006; Pope et al., 2006; Pope et al., 1992; Schwartz, 1994) This was observed mainly among elderly and chronically ill persons Although most of the epidemiological research has focused on the effects of short term exposures, several studies have reported that long term exposure to particulate air pollution common to many metropolitan areas is an important risk factor for cardiopulmonary (including lung cancer) mortality (Dockery et al., 1993; Pope, 2007; Pope et al., 2002; Pope et al., 2004)

2.1.2 Summary of toxicological findings

To develop methods that could be employed to mitigate the adverse health effects induced by PM, it is important to understand the mechanisms that are involved in PM interactions with lung or other tissues Toxicological studies examine how voluntary individuals in exposure chambers, animal test subjects or cultured cells respond to controlled PM exposures and aim to pinpoint causative agents which are responsible for observed health outcomes resulting from exposure to

PM Because of the heterogeneity in PM composition, it is often difficult to identify

a single causative mechanism

Despite a huge body of research on the subject, the underlying toxicological mechanisms by which PM induces adverse health effects are not yet entirely understood There is a great number of studies indicating that inflammation via oxidative stress is the route through which PM may exert toxicity Oxidative stress is defined as an imbalance between the level of reactive oxygen species (ROS) and the cell’s (or body’s) natural antioxidant defence ROS is a collective term that refers to free radicals such as hydroxyl (HO.) and peroxyl (HOO., ROO.), ions such as superoxide (O2-.) and peroxynitrite (ONOO-) and molecules such as hydrogen peroxide (H2O2) and hydroperoxides (ROOH) Precursors of ROS such as carbon-centered radicals can also be considered as ROS ROS can be generated on PM’s surface (directly) or by cells, upon interaction with PM (indirectly) Oxidative stress initiates redox-sensitive transcription factors (for example, MAPK and the NF-κB cascade), which work synergistically to activate expression of proinflammatory

cytokines (IL-4, -6, -8, TNF-α, etc.), chemokines, and adhesion receptors Many in vivo and in vitro studies have reported increased expression of these transcription

factors and proinflammatory genes after exposure to PM (reviewed in Dreher, 2000; Gonzalez-Flecha, 2004; Riedl and Diaz-Sanchez, 2005), implying that an increased amount of ROS is generated in the cells upon exposure to PM Alveolar macrophages and airway epithelial cells play a central role in the body’s defence against inhaled foreign materials, by generating vigorous, localized inflammatory responses in the respiratory tract However, excessive inflammation may cause localized tissue damage through apoptosis and necrosis of cells, or, if sustained, may induce progression to a more wide spread disease

Specific biological indicators of oxidative stress, as well as particle properties relevant for the observed adverse health outcomes will be reviewed in more detail further in the text

Relevant methods for establishing toxicity of PM include a variety of approaches including: (1) human inhalation studies; (2) animal inhalation studies;

(3) animal in vivo instillation studies; (4) in vitro (cell exposure) studies and (5) in vitro acellular (cell-free) assays In vivo studies are of high importance for assessing

the health impact of PM as they allow experimental design that will most realistically reflect real-world conditions However, challenges such as the relatively high cost as well as the logistics of the whole experimental protocol and design (for

example, ethical approvals), make this type of studies impractical In this regard, in vitro assays (both cell exposure and acellular) have been recognised as cheaper,

simpler and less time-consuming and, therefore, among many researchers, are the

preferred approach for assessing PM toxicity Since an in vitro approach is more relevant to the problem being investigated in this research than in vivo studies, this literature review will not go into details of the in vivo methodology

2.2 Aerosol fundamentals and basic terminology

Ambient air particles range from 0.001 up to 100 µm in aerodynamic diameter1, spanning over five orders of magnitude That is why the physical characteristics of PM are most simply described in terms of size distributions, which can be represented in terms of a number, a surface area, or a volume/mass distribution (Figure 2-1) Log-normal distribution2 has been found to provide a good fit and has been used for a long time to describe size distribution of ambient PM The 0th moment of the particle size distribution, expressed as the number of particles per logarithmic size interval, can be formulated with the log-normal function as follows:

- (1) where D is particle diameter, DCMD is the count median diameter3, is standard deviation of the distribution and N is the total particle number

The surface area and volume (2nd and 3rd moment, respectively) size distributions can be defined similarly with respect to D:

3

Count median diameter (CMD) – particle diameter at which one half of the particle population lies

A large number of aerosol instrumentation is based on particle-counting sensors, which yield number size distribution From such data it is often necessary

to calculate the particles’ total surface and volume (or mass) Using these relationships, the integral transformation from total particle number (N) to surface (S) and/or volume (V) can be performed This has been outlined in more detail in Heintzenberg, (1994) Assuming spherical particles, the total surface area and volume can be written as a function of parameters of number size distribution:

* + '%& , - ( /0 123 3 ,)- ( /4- ( 4 5 (6)

6 #+ '%& ,$- ( /0$ 123 3 ,7- ( /4- ( 45 (7)

Figure 2-1 Typical urban PM number (A), surface (B) and volume (C) size distributions Taken from (Seinfeld and Pandis, 2006)

The size distribution of particles in the urban atmosphere is typically

characterised by three modes: nucleation (or nuclei) mode, accumulation mode and coarse mode The smallest of these, below 0.1 μm in diameter, are called the nucleation mode and are formed by rapid nucleation of low vapour pressure

compounds (mainly produced by combustion sources) and from chemical conversion of gases to particles in the atmosphere They are relatively short-lived and grow into larger particles between 0.1 and about 2 μm in diameter, known as

the accumulation mode Accumulation mode particles can remain suspended for up

to several weeks in the air, and are not readily removed by rain They are formed by coagulation of particles in the nucleation mode and by condensation onto existing particles For example, soot formed in flames grows by coagulation and condensation of semivolatile compounds to form particles that belong to the

accumulation mode size range The third, coarse mode comprises particles greater

than about 2 μm in diameter These are generally formed by break- up of larger matter, and include wind-blown dust and soil, particles from construction and from sea spray Due to their size they remain in the air for relatively short periods, but they make (in relation to their numbers) a disproportionate contribution to PM10

mass when this is measured close to a source The majority of generated PM lies in the submicrometer range (D < 1 µm)

combustion-While aforementioned size segregation of ambient PM is mainly related to particle origin and formation mechanism(s), size classification of ambient PM in wording of air quality standards includes PM10 and PM2.5, which are size fractions with an aerodynamic diameter smaller than 10 and 2.5 µm, respectively These particle size definitions are often used in studies related to the health effects of PM

PM2.5 is called fine particulate matter, while particles with aerodynamic diameter bigger than 2.5 µm are called coarse particles These size fractions are measured as mass Ultrafine PM is an unregulated subset of PM2.5 and it typically refers to particles with an aerodynamic diameter smaller than 0.1 μm Ultrafine particles contribute significantly to urban PM by number, yet very little by mass (Figure 2-1 A and C) Whereas fine and ultrafine PM is primarily formed during the combustion

processes, urban coarse PM is derived from abraded soil, road dust (for example, brake and tire dust) and construction debris

2.3 Combustion-generated PM

Combustion, in general, is the major source of PM in the urban atmosphere, affecting in this manner both indoor (e.g tobacco smoking, wood burning stoves and fireplaces) and outdoor (e.g vehicle emissions, biomass burning) air quality

Combustion (burning) is defined as an exothermic chemical reaction of a particular substance (gas, liquid, or solid) and an oxidant (generally atmospheric oxygen) In a complete combustion reaction, the total amount of each element present in the burning substance should completely react with oxygen, producing a limited number of products For example, complete combustion of hydrocarbons would result only in generation of carbon dioxide and water vapour In reality, combustion processes are never perfect or complete Apart from CO2, H2O, SO2 and

NOx as expected products of combustion, incomplete combustion also results in the formation of a great number of different compounds present both in the gaseous phase and as particulates Combustion sources can be mobile or stationary, outdoor or indoor and they range from transport sources and industrial and power plants to open fire burning and tobacco smoking As already mentioned, most of the combustion-generated PM is smaller than 0.1 µm in diameter in terms of number, while most of the particle mass is larger than 0.1 µm

Cigarette smoke is in some countries still a major source of indoor PM; diesel exhaust presents the largest single source of PM in urban areas, while (residential) wood combustion is an important source of both indoor and outdoor

PM in developing countries, where it is used for heating and cooking, but also in developed countries with colder climates, where it is regaining its importance in residential heating as being a source of renewable energy Particles coming from these three sources were investigated as a part of this project and, therefore, a

concise description of main physiochemical characteristics of each is given in this literature review

2.3.1 Cigarette smoke

General considerations Mainstream cigarette smoke (MS) is the smoke

drawn through the cigarette filter which would enter the smoker’s mouth during a puff Sidestream smoke (SS) is the smoke released directly into the environment from the burning end of the cigarette As sidestream and exhaled mainstream smoke become a part of an indoor atmosphere, they are mixed and diluted to form environmental tobacco smoke (ETS) Apart from the characteristics of the cigarette itself, the manner in which a cigarette is smoked greatly influences smoke yield and composition Therefore, in order to obtain comparable results, it was necessary to standardize smoking conditions for testing and experimental purposes In 1967 the

US Federal Trade Commission proposed a standard cigarette testing protocol which was widely accepted and it is still in use The defined smoking conditions included

35 mL puff volume in duration of 2 s, taken every 60 s to 3 mm from the filter overwrap for filter cigarettes, 23 or 30 mm of butt length for non-filter cigarettes Due to the fact that these smoking conditions poorly reflect today's human smoking, Canada and the US states of Texas and Massachusetts have introduced new, more intense testing protocols which require using larger puff volumes (55 mL and 45 mL, respectively) and taking puffs every 30 s instead of every 60 s (Borgerding and Klus, 2005) A standard method for the generation and collection

of the sidestream smoke has not yet been clearly defined

Physical properties and chemical composition Cigarette smoke (both

mainstream and sidestream) is formed through a combination of combustion and pyrolysis (in oxygen deficient regions) at temperatures of ~800oC (during smouldering) to ~900oC (during a puff) (Jenkins et al., 2000) and produces a highly complex aerosol consisted of mainly electrically charged liquid droplets (particles)

in a vapour phase

To date, there have been around 4700 different compounds identified in mainstream cigarette smoke, at levels ranging from milligrams to picograms Out of these, about 69 are known or probable carcinogens (Hoffmann et al., 2001) Kabir

et al., (2007) reported that approximately 88% of the lung cancer deaths in men and 71% in women occurring in the US are attributable to cigarette smoking SS is qualitatively similar to MS; however, it is quantitatively different as the amounts of many substances found in SS are different from those found in MS While MS is slightly acidic, SS is slightly alkaline, as a consequence of higher amount of ammonia Higher pH causes nicotine to be present in its free base form As such, sidestream nicotine is predominantly present in the vapour phase, while mainstream nicotine mainly comes as a part of the smoke's PM (Jenkins et al., 2000)

Fresh, unaged MS contains approximately 5 × 109 particles /cm3 (Jenkins et al., 2000) with the count median diameter within the range of 0.18 and 0.44 µm (Bernstein, 2004; Morawska et al., 1997) As a result of such a high number concentration, a rapid coagulation and, consequently, an increase in the average particle diameter, occurs as the smoke ages Values for the count median diameter

of the SS and ETS are slightly lower, in the range of 0.1 – 0.25 µm (Borgerding and Klus, 2005; Jenkins et al., 2000; Morawska et al., 1997)

2.3.2 Diesel exhaust PM

Physical properties and chemical composition Diesel exhaust is a complex

mixture of combustion products of diesel fuel, with the composition depending on the type of the engine, operating conditions, the type of fuel and additives used and emission control system The primary emissions of diesel engines include gaseous precursors, like H2SO4, SO2, SO3, H2O, low-volatile and semivolatile organic compounds, agglomerated solid carbonaceous material (soot particles) of about 0.15 – 0.4 µm in diameter (Burtscher, 2005) and metallic ash During the dilution and cooling process, condensation on the surface of the existing particles, as well as

nucleation of low-volatile species occurs (Kittelson, 1998)

Figure 2-2 shows a ty

number and mass Nucleation

µm (Kittelson and Watts, 2002

sulphur compounds (Meyer an

dilution with ambient air an

nucleation mode particles may

soot or metallic compounds

Typically, 0.1-10 % of the part

number is found in the nuclea

roughly from 0.03 to 0.5 µm a

adsorbed materials like heavy h

Approximately 10 % of the pa

contained in the accumulation

diameters above 1µm and cont

makes no contribution to part

the diesel combustion process

material, which was deposited

1998)

Figure 2-2 Typical engine exhau

are shown (Kittelson, 1998)

a typical diesel particle size distribution weigion mode particles range in diameter from 0.003002), they mainly consist of semivolatile organi

r and Ristovski, 2007) and are formed during and the subsequent cooling A certain am

ay not be totally semivolatile and contain solid

ds (De Filippo and Maricq, 2008; Tobias et alparticle mass and often more than 90 % of thecleation mode The accumulation mode range

m and is composed mainly of carbon agglomeravy hydrocarbons, sulphur compounds, and met particle number and 80 - 90 % of the particletion mode The coarse mode consists of particcontains 5-20% of the total particle mass and esparticle number The coarse particles are not focess; they are formed by re-entrainment of paited on cylinder and exhaust system surfaces

haust size distribution; both mass and number w

weighted by 003 to 0.03 ganic and/or ring exhaust amount of olid core like

t al., 2001) the particle nges in size merates and metallic ash ticle mass is articles with

d essentially

ot formed in

f particulate (Kittelson,

r weightings

The type of fuel and fuel sulphur content, the combustion system and conditions, the exhaust treatment, diluting and cooling conditions affect mostly nucleation, but also coagulation and condensation of exhaust particles Number size distribution, which consists mainly of nucleation mode particles, varies strongly with different fuel, engine and driving conditions, whereas mass size distribution, in which accumulation mode is dominant fraction, is much less influenced by variations in these parameters Typically, engines operating under higher load and temperature and decreased air/fuel ratio produce more particulate matter (Zielinska et al., 2004)

Taking into account its complex nature, the chemical composition of diesel exhaust PM (and PM in general) is often expressed in terms of the organic carbon/elemental carbon ratio (OC/EC) (Alander et al., 2004; Shah et al., 2004; Sharma et al., 2005) The majority of the diesel PM mass is in the form of elemental carbon, which presents a core onto which various organic compounds may be adsorbed Organic compounds in diesel PM originate from unburned fuel and lubricating oil, partial combustion and pyrolysis products and include alkanes, cycloalkanes, alkylbenzenes and polycyclic aromatic hydrocarbons (PAHs) and their derivatives (Liang et al., 2005) PAHs are of special concern as they are considered

to be potential human carcinogens OC/EC ratio varies widely with the engine operating conditions, but generally there is a higher EC contribution in diesel PM emissions when the engine is running under higher load (Shah et al., 2004; Sharma

et al., 2005; Zielinska et al., 2004) Various metals, like Fe, Mg, Ca, Ba, Cr, Ni, Pb, Zn,

Cd, Cu, were also found in the diesel PM (Sharma et al., 2005)

The global fuel crisis and growing awareness regarding vehicle emission pollutants have created a need for alternative fuel sources Biomass has been recognised as a major world renewable energy source to supplement declining fossil fuel resources Biofuels are easily available from common biomass sources (for example crops, vegetable oil and animal fat) and they are being more and more pursued as a replacement for (or addition to) diesel (Demirbas, 2007) Ethanol is a liquid biofuel that has been widely investigated as a potential substitute (or supplement) for diesel A well documented advantage of ethanol usage in diesel

engines is the significant reduction in particulate mass emissions, especially at full load operation (Abu-Qudais et al., 2000; Di et al., 2009), while the effect on carbon monoxide (CO), hydrocarbons (HC) and oxides of nitrogen (NOx) is less clear Some studies have reported an increase of HC emissions with the use of ethanol-diesel blends (Lu et al., 2005; Rakopoulos et al., 2008) Recent studies have shown that the improvement in exhaust emissions provided by oxygenated fuels (such as ethanol) depend almost entirely on the oxygen content of the fuels, regardless of the oxygenate to diesel fuel blend ratios or the type of oxygenate (Wang et al., 2009) While there are many studies reporting emission characteristics of the engines fuelled with diesel-ethanol (or biofuel, in general) mixtures, there is little published data on the potential toxicity of PM produced by such sources

2.3.3 Wood smoke particles

Physical and chemical properties The main processes in wood combustion

are dehydratation, oxidation, pyrolysis and devolatilization, char combustion and gas-phase oxidation (Nussbaumer, 2003; Simoneit, 2002) The duration of each of these phases and, thus, physiochemical characteristics of wood smoke PM strongly depend on the combustion conditions (for example, excess air ratio4) and fuel size and characteristics (for example, moisture content) Fine particles emitted from residential wood combustion appliances can be divided into three typical classes based on composition and morphology: spherical organic carbon particles (“tar balls” (Posfai et al., 2003)), aggregated soot particles and inorganic ash particles In real combustion situations, where the combustion conditions change during a combustion cycle, these particle classes may co-exist and interact

Spherical organic particles are associated with lower combustion temperatures They have been observed during combustion under poor conditions, such as burning poor quality wood (for example, with high moisture content), insufficient air supply, or overloading the firebox (Kocbach Bolling et al., 2009)

4

Excess air ratio (λ) describes the ratio between the locally available and the stochiometric amount