QUEENSLAND UNIVERSITY OF TECHNOLOGY

SCHOOL OF CHEMISTRY, PHYSICS AND MECHANICAL ENGINEERING

STUDY OF NEW PARTICLE FORMATION IN SUBTROPICAL URBAN ENVIRONMENT IN BRISBANE,

AUSTRALIA

HING CHO CHEUNG

A THESIS SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS OF THE DEGREE OF DOCTOR OF PHILOSOPHY

September 2012

i

ABSTRACT

Atmospheric ultrafine particles play an important role in affecting human health,

altering climate and degrading visibility. Numerous studies have been conducted to

better understand the formation process of these particles, including field

measurements, laboratory chamber studies and mathematical modeling approaches.

Field studies on new particle formation found that formation processes were

significantly affected by atmospheric conditions, such as the availability of particle

precursors and meteorological conditions. However, those studies were mainly

carried out in rural areas of the northern hemisphere and information on new particle

formation in urban areas, especially those in subtropical regions, is limited. In

general, subtropical regions display a higher level of solar radiation, along with

stronger photochemical reactivity, than those regions investigated in previous studies.

However, based on the results of these studies, the mechanisms involved in the new

particle formation process remain unclear, particularly in the Southern Hemisphere.

Therefore, in order to fill this gap in knowledge, a new particle formation study was

conducted in a subtropical urban area in the Southern Hemisphere during 2009,

which measured particle size distribution in different locations in Brisbane, Australia.

Characterisation of nucleation events was conducted at the campus building of the

Queensland University of Technology (QUT), located in an urban area of Brisbane.

Overall, the annual average number concentrations of ultrafine, Aitken and

nucleation mode particles were found to be 9.3 x 103, 3.7 x 103 and 5.6 x 103 cm-3,

respectively. This was comparable to levels measured in urban areas of northern

ii

Europe, but lower than those from polluted urban areas such as the Yangtze River

Delta, China and Huelva and Santa Cruz de Tenerife, Spain. Average particle number

concentration (PNC) in the Brisbane region did not show significant seasonal

variation, however a relatively large variation was observed during the warmer

season. Diurnal variation of Aitken and nucleation mode particles displayed different

patterns, which suggested that direct vehicle exhaust emissions were a major

contributor of Aitken mode particles, while nucleation mode particles originated

from vehicle exhaust emissions in the morning and photochemical production at

around noon. A total of 65 nucleation events were observed during 2009, in which 40

events were classified as nucleation growth events and the remainder were nucleation

burst events. An interesting observation in this study was that all nucleation growth

events were associated with vehicle exhaust emission plumes, while the nucleation

burst events were associated with industrial emission plumes from an industrial area.

The average particle growth rate for nucleation events was found to be 4.6 nm hr-1

(ranging from 1.79-7.78 nm hr-1), which is comparable to other urban studies

conducted in the United States, while monthly particle growth rates were found to be

positively related to monthly solar radiation (r = 0.76, p <0.05). The particle growth

rate values reported in this work are the first of their kind to be reported for the

subtropical urban area of Australia.

Furthermore, the influence of nucleation events on PNC within the urban airshed was

also investigated. PNC was simultaneously measured at urban (QUT), roadside

(Woolloongabba) and semi-urban (Rocklea) sites in Brisbane during 2009. Total

PNC at these sites was found to be significantly affected by regional nucleation

iii

events. The relative fractions of PNC to total daily PNC observed at QUT,

Woolloongabba and Rocklea were found to be 12%, 9% and 14%, respectively,

during regional nucleation events. These values were higher than those observed as a

result of vehicle exhaust emissions during weekday mornings, which ranged from

5.1-5.5% at QUT and Woolloongabba. In addition, PNC in the semi-urban area of

Rocklea increased by a factor of 15.4 when it was upwind from urban pollution

sources under the influence of nucleation burst events.

Finally, we investigated the influence of sulfuric acid on new particle formation in

the study region. A H2SO4 proxy was calculated by using [SO2], solar radiation and

particle condensation sink data to represent the new particle production strength for

the urban, roadside and semi-urban areas of Brisbane during the period June-July of

2009. The temporal variations of the H2SO4 proxies and the nucleation mode particle

concentration were found to be in phase during nucleation events in the urban and

roadside areas. In contrast, the peak of proxy concentration occurred 1-2 hr prior to

the observed peak in nucleation mode particle concentration at the downwind semi-

urban area of Brisbane. A moderate to strong linear relationship was found between

the proxy and the freshly formed particles, with r2 values of 0.26-0.77 during the

nucleation events. In addition, the log[H2SO4 proxy] required to produce new

particles was found to be ~1.0 ppb Wm-2 s and below 0.5 ppb Wm-2 s for the urban

and semi-urban areas, respectively. The particle growth rates were similar during

nucleation events at the three study locations, with an average value of 2.7 ± 0.5 nm

hr-1. This result suggested that a similar nucleation mechanism dominated in the

study region, which was strongly related to sulphuric acid concentration, however the

iv

relationship between the proxy and PNC was poor in the semi-urban area of Rocklea.

This can be explained by the fact that the nucleation process was initiated upwind of

the site and the resultant particles were transported via the wind to Rocklea. This

explanation is also supported by the higher geometric mean diameter value observed

for particles during the nucleation event and the time lag relationship between the

H2SO4 proxy and PNC observed at Rocklea.

In summary, particle size distribution was continuously measured in a subtropical

urban area of southern hemisphere during 2009, the findings from which formed the

first particle size distribution dataset in the study region. The characteristics of

nucleation events in the Brisbane region were quantified and the properties of the

nucleation growth and burst events are discussed in detail using a case studies

approach. To further investigate the influence of nucleation events on PNC in the

study region, PNC was simultaneously measured at three locations to examine the

spatial variation of PNC during the regional nucleation events. In addition, the

impact of upwind urban pollution on the downwind semi-urban area was quantified

during these nucleation events. Sulphuric acid was found to be an important factor

influencing new particle formation in the urban and roadside areas of the study

region, however, a direct relationship with nucleation events at the semi-urban site

was not observed. This study provided an overview of new particle formation in the

Brisbane region, and its influence on PNC in the surrounding area. The findings of

this work are the first of their kind for an urban area in the southern hemisphere.

v

KEYWORDS

Particle size distribution, nucleation, particle formation, ultrafine particles,

subtropical urban.

vi

LIST OF PUBLICATIONS

Cheung, H.C., Johnson, G.R., Morawska, L. and Ristovski, Z.D. (2011a). Particle

detection efficiency for CPC’s depends on ambient aerosol composition and

condensation medium. Submitted for publication in Atmospheric Environment.

Cheung, H.C., Morawska, L. and Ristovski, Z.D. (2010). Observation of new

particle formation in subtropical urban environment. Atmospheric Chemistry

and Physics Discussions, 10, 22623-22652.

Cheung, H.C., Morawska, L. and Ristovski, Z.D. (2012). Influence of sulphuric acid

on nucleation in subtropical urban area of Australia. Subumitted for publication

in Atmospheric Environment.

Cheung, H.C., Morawska, L. and Ristovski, Z.D. (2011b). Observation of new

particle formation in subtropical urban environment. Atmospheric Chemistry

and Physics, 11, 1-11.

Cheung, H.C., Morawska, L., Ristovski, Z.D. and Wainwright, D. (2011c).

Influence of medium range transport of particles from nucleation burst on

particle number concentration within the urban airshed. Atmospheric Chemistry

and Physics Discussions, 11, 32965-32992.

Cheung, H.C., Morawska, L., Ristovski, Z.D. and Wainwright, D. (2012). Influence

of medium range transport of particles from nucleation burst on particle number

concentration within the urban airshed. Atmospheric Chemistry and Physics, 12,

4951-4962.

Jayaratne, E.R., Johnson, G.R., McGarry, P., Cheung, H.C. and Morawska L. (2011).

Characteristics of airborne ultrafine and coarse particles during the Australian

dust storm of 23 September 2009. Atmospheric Environment, 45, 3996-4001.

Morawska, L., Wang, H., Ristovski, Z., Jayaratne, E.R., Johnson, G., Cheung, H.C.,

Ling, X. and He, C. (2009). JEM Spotlight: Environmental monitoring of

airborne nanoparticles. Journal of Environmental Monitoring, 11, 1758-1773.

1

TABLE OF CONTENTS

ABSTRACT………………………………………………...…………….…...……..i

KEYWORDS………………………………………………...………….……….…..v

LIST OF PUBLICATIONS…………………………………..….………………...vi

STATEMENT OF ORIGNAL AUTHORSHIP…………………...….…………...3

ACKNOWLEDGEMENTS………………………………………...………………4

CHAPTER1. INTRODUCTION………………………………….………………..5

CHAPTER 2. LITERATURE REVIEW…………….……………...……………18

CHAPTER 3. ENVIRONMENTAL MONITORING OF AIRBORNE

NANOPARTICLES……………………………………………………...………...68

CHAPTER 4. PARTICLE DETECTION EFFICIENCY FOR CPC’S

DEPENDS ON AMBIENT AEROSOL COMPOSITION AND

CONDENSATION MEDIUM……………………………………………...…....137

CHAPTER 5. OBSERVATION OF NEW PARTICLE FORMATION IN

SUBTROPICAL URBAN ENVIRONMENT……………………………..…….159

CHAPTER 6. INFLUENCE OF MEDIUM RANGE TRANSPORT OF

PARTICLES FORM NUCLEATION BURST ON PARTICLE NUMBER

CONCENTRATION WITHIN THE URBAN AIRSHED………………….….196

2

CHAPTER 7. INFLUENCE OF SUFLURIC ACID ON THE NEW PARTICLE

FORMATION IN SUBTROPICAL URBAN SOUTHERN HEMISPHERE...232

CHAPTER 8. GENERAL DISCUSSION……………………………...…….….258

APPENDIXES…………………………………………………………………….265

STATEMENT OF ORIGINAL AUTHORSHIP

The work contained in this thesis has not been previously submitted for a degree or

diploma at 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.

Signed:

Date: .... 3 .... .f.�p.t ...... -?.!!.l.v

3

QUT Verified Signature

4

ACKNOWLEDGEMENTS

I would like to express my sincere thanks to my supervisor, Lidia Morawska, for her

supervision in my scientific research and encouragement. I would like to thank my

co-supervisor, Zoran Ristovski, for his inspirational advice in my research and

guidance in field measurement. My gratitude also goes to my co-supervisor, David

Wainwright, who has provided me with valuable advice on the field measurement. I

would like to thank my teammates in International Laboratory for Air Quality and

Health (ILAQH) for their friendship and help throughout my study. Last but not

least, I would like to thank my wife, Celine, and family in their continuous support

and encouragement.

5

CHAPTER 1. INTRODUCTION

1.1 Description of Scientific Problem Investigated

The formation processes of ultrafine particles (UFP) in the atmosphere are difficult

to fully understand due to the multiplicity of sources and mechanisms involved. Also,

the scientific knowledge behind particle formation processes is still not fully

developed. Nucleation processes, such as binary nucleation (involving sulphuric acid

and water) and ternary nucleation (involving with sulphuric acid, water and ammonia)

are the major processes by which UFP are formed in the atmosphere. Other processes

which contribute to the formation of UFP are organics condensation and ion-induced

nucleation (IIN) (Kulmala and Kerinen 2008), however it is important to note that

UFP formation in the atmosphere may not solely governed by one of the above

mechanism, but involves other processes. Urban pollution sources, such as fossil fuel

combustion (including vehicles, industrial and power generation) and biomass

burning, significantly contribute to UFP concentrations (Morawska et al., 2008).

Furthermore, pollution plumes observed in the urban environment are often a

mixture originating from different emission sources. Thus, UFP formation in the

urban atmosphere is a complex process.

Due to the significant impact the UFPs have on human health and the environment,

many studies have been conducted on the various aspects and characteristics of these

particles (Kulmala et al., 2004, Morawska et al. 2004). A more detailed review of

these studies can be found in the Chapter 2. In summary, particle formation in the

atmosphere is highly depending on the existing chemical species in the atmosphere,

6

as well as local meteorological conditions. In general, discrepancies between the

observed and modelled results for nucleation rate also imply that different processes

and factors are involved (Weber et al., 1996, 1997 and 1998). Previous observational

studies focused on the physical properties of UFP, such as size and number

distributions and their temporal variation (e.g. Morawska et al., 1999), where the

formation of UFP was attributed to different processes depending on the

environmental conditions (Kulmala et al., 2004). The limitations of measurement

techniques for investigating the chemical composition of UFPs has impeded the

progress of research on health impact assessment, source sectors/ regions

apportionment and the species involved in the initial stages of the nucleation

processes (McMurry 2000; Kulmala and Kerinen 2008).

Particle formation processes have been widely studied in rural areas, boreal forests

and along coastal and marine boundaries (Kulmala et al. 2004). New particle

formation is often observed in clean environments, whereas the presence of a

coagulation sink was limited, indicating that the cluster molecule undergoes

nucleation process rather coagulating/condensing on pre-existing particles (Kulmala

2003). The majority of urban studies focused on roadside measurements, since

vehicle emissions are the major source of pollution in these areas. These studies

concluded that UFP concentration was associated with traffic density, especially

during the morning traffic peak period. Furthermore, new particle formation

associated with photochemical reactions was suggested to be another significant

source of the UFP concentrations during noontime, when the strongest solar radiation

was observed (Cheung et al., 2011c; Pey et al., 2008). However, identification of the

7

specific pollution sources which contribute to nucleation in the urban environment is

challenging due to the complexity of emission sources within the urban environment,

including vehicle, industrial and domestic emissions etc.

Therefore, the knowledge on UFP formation in the atmosphere is limited in urban

areas, particularly in the southern hemisphere, where its impact on regional

atmospheric quality is still unclear. Thus, more investigations are needed on the

influence of urban pollution on UFP formation in the atmosphere.

The major scientific topics that have been addressed by previous studies can be

summarised as follows:

- Intensive field measurements of neutral/charged clusters were mainly

conducted in clean environments, such as continental rural areas/boreal

forests, as well as arctic, marine environments and the upper troposphere.

These studies were mostly conducted in the northern hemisphere and they

showed that nucleation processes varied according to different environmental

settings.

- Several parameterisations of particle formation mechanisms were developed

for homogeneous binary/ternary nucleation, ion-induced nucleation etc.

Although some parameterisation studies showed a good agreement between

modelling results and the observation data, the modelling results of particle

8

formation rate were usually lower than the observation data, suggesting that

new particle formation may not be the result of a single mechanism.

However, to fully understand the atmospheric aerosol nucleation process, several

scientific gaps need to be further investigated, which include:

- Improvements in the instrumentation techniques used to measure the

chemical composition of nucleation particles;

- Characterisation of particle formation processes in the urban environment,

especially in the south hemisphere;

- Investigation of the key mechanisms involved in particle formation process;

and

- The inclusion of these findings in climate models for climatological

applications.

This work focuses on the characterisation of particle formation processes in the

subtropical urban areas of the southern hemisphere, which generally has stronger

solar radiation and relatively lower ambient particle concentrations compared to

other continental regions that provided favourable conditions for new particle

formation. Thus, a field measurement based study was conducted in urban area of

Brisbane, in South-East Queensland, Australia.

9

1.2 Aims of the Study

The aim of this study was to characterise the new particle formation process in a

subtropical urban environment as follows:

- To characterise the temporal and spatial variations of UFP;

- To quantify new particle formation events and their impact on particle

number concentrations (PNCs) within the urban airshed of Brisbane;

- To investigate the influence of sulphuric acid on nucleation in the study

region; and

- To summarise and draw conclusion based on the above findings and to

provide recommendations for the further study of nucleation processes in the

study region.

1.3 Specific Objectives of the Study

There were four main objectives in this work:

i) To investigate the capability of particle monitoring instruments, in order to select

the best particle counting instrument for use in this work.

- A literature review was conducted on particle measurement techniques, in

order to summarise the advantages and disadvantages of each technique. The

review discussed methods for the measurement of particle properties such as

number, concentration, size distribution and surface area. In addition, the

measurement of particle composition by direct and indirect techniques was

also discussed (Chapter 3).

10

- Laboratory testing for water-based and butanol-based condensation particle

counters (WCPC and BCPC) was conducted. The tests compared the WCPC

and BCPC with regard to their detection efficiencies for aerosols of different

composition. Furthermore, the effects of coincidence error on the

measurement of particle number were discussed and based on these findings,

a particle counting instrument was recommended for ambient particle

measurements (Chapter 4).

ii) To quantify UFP concentration, and examine their temporal and spatial

variations in South-East Queensland.

- The measurement of UFP concentrations was conducted at a campus building

of the Queensland University of Technology (QUT), which is located in the

central business district area of Brisbane, during January 2009 to December

2009. Furthermore, an intensive study was carried out at Woolloongabba and

Rocklea monitoring sites, representing roadside and semi-urban environments,

respectively, during the winter season of 2009. The temporal and spatial

variation of UFP concentrations within the Brisbane region were studied and

are presented in Chapter 5 and 6.

iii) To evaluate the impact of regional nucleation events on UFP concentrations

in the Brisbane region.

11

- The impact of regional nucleation events on UFP concentrations was studied

by conducting simultaneous measurements of UFP at three study locations.

The variation in UFP concentrations at three stations during regional

nucleation events provided evidence that the UFP concentration downwind of

the event was affected by nucleation process in the upwind area. This impact

was described and quantified in Chapter 6.

iv) To explain new particle formation processes in South-East Queensland.

- New particle formation (i.e. a nucleation) events are identified based on the

measured particle size distribution data. The relationship between nucleation

processes and other factors, including meteorological conditions and particle

precursors, were studied and described in Chapter 5. Furthermore, the

influence of sulphuric acid, an important contributor to the nucleation process,

on nucleation at the three study locations was investigated and is reported in

Chapter 7.

1.4 Account of Scientific Process Linking the Research Papers

As mentioned in the above sections, several manuscripts were published or submitted

to peer-reviewed journals based on the present work, which mainly focused on the

study of new particle formation in the sub-tropical urban area of Brisbane.

1. To better understand the capability of the instruments used for monitoring

UFP, a review of the current instruments used for particle monitoring was

carried and published in the Journal of Environmental Monitoring, with a title

12

of “JEM Spotlight: Environmental monitoring of airborne nanoparticles”.

This paper reviewed and concluded on the performance of the instruments

currently used to measure the physical and chemical properties of UFP.

Furthermore, a study of the particle detection efficiency of Condensation

Particle Counters (CPCs) was conducted to compare the detection

performance of butanol (BCPC) and water (WCPC) based CPCs (Cheung et

al. 2011a). Aerosols with different chemical composition were measured by

the CPCs and the results showed that the detected concentration of water

insoluble particles was underestimated by the WCPC. The findings of this

work allowed the selection of an appropriate instrument for measuring

ambient particle concentrations in the current work. The two manuscripts

which resulted from this work are presented in Chapter 3 and 4.

2. A year-long measurement campaign for measuring UFP concentration was

carried out at QUT campus during 2009, which developed a comprehensive

database for research on the issue of new particle formation (Cheung et al.,

2010, 2011c). This dataset was used to investigate the temporal variation of

UFP and its relationship with meteorological parameters and trace gaseous

pollutants. Although there have been other studies on new particle formation

conducted in the coastal, rural and forest areas of Australia (Johnson et al.,

2005, Guo et al., 2008, Modini et al., 2009 and Ristovski et al., 2010), studies

on new particle formation focusing on urban environments is limited. This

study provides the only database that records the continuous measurement

13

particle size distribution in a subtropical urban area of Australia. It also was

the first study which reported on the particle growth rate of nucleation events

in an urban area of Australia.

3. An intensive study of regional nucleation was carried out at different

environments around Brisbane, including urban (QUT), roadside

(Woolloongabba) and semi-urban (Rocklea) areas, during the winter season

of 2009. In this study, regional nucleation events were observed, and the

temporal and spatial variations of UFP concentration were investigated.

Condensation sink was found to be a factor which occasionally suppressed

the nucleation process at the roadside site. The enrichment factor of particle

number concentration at the semi-urban site was also calculated according to

the influence of upwind urban pollution during nucleation burst events

(Cheung et al., 2011c, 2012).

4. Sulphuric acid was proposed as a key component which contributes to the

nucleation process. Particle production strength could be represented by a

H2SO4 proxy, which is a function of sulphur dioxide, solar radiation and

condensation sink. The influence of the H2SO4 proxy on nucleation mode

particles was investigated (Cheung et al. 2011b) and a moderate to strong

linear relationship was found between H2SO4 proxies and freshly formed

particles (with a particle size 4-6 nm) during the nucleation growth events.

14

This result implied that the observed nucleation process was significantly

affected by the presence of sulphuric acid in the ambient atmosphere.

5. Furthermore, the database was used to study the dust storm event in relation

to UFPs which occurred in South-East Queensland in 2009, and the findings

were published in Atmospheric Environment, with a title of “Characteristics

of airborne ultrafine and coarse particles during the Australian dust storm of

23 September 2009” (Jayaratne et al., 2011). This manuscript is included in

the appendices of this thesis.

15

1.5 References

Cheung, H.C., Johnson, G.R., Morawska, L. and Ristovski, Z.D. (2011a). Particle detection efficiency for CPC’s depends on ambient aerosol composition and condensation medium. Submitted for publication in Atmospheric Environment.

Cheung, H.C., Morawska, L. and Ristovski, Z.D. (2010). Observation of new particle formation in subtropical urban environment. Atmospheric Chemistry and Physics Discussions, 10, 22623-22652.

Cheung, H.C., Morawska, L. and Ristovski, Z.D. (2012). Influence of sulphuric acid on nucleation in subtropical urban area of Australia. Submitted for publication in Atmospheric Environment.

Cheung, H.C., Morawska, L. and Ristovski, Z.D. (2011b). Observation of new particle formation in subtropical urban environment. Atmospheric Chemistry and Physics, 11, 1-11.

Cheung, H.C., Morawska, L., Ristovski, Z.D. and Wainwright, D. (2011c). Influence of medium range transport of particles from nucleation burst on particle number concentration within the urban airshed. Atmospheric Chemistry and Physics, 12, 4951-4962.

Guo, H., Ding, A., Morawska, L., He, C., Ayoko, G., Li, Y., Hung, W. (2008). Size

distribution and new particle formation in subtropical eastern Australia,

Environmental Chemistry, 5, 382-390.

Jayaratne, E.R., Johnson, G.R., McGarry, P., Cheung, H.C. and Morawska L. (2011). Characteristics of airborne ultrafine and coarse particles during the Australian dust storm of 23 September 2009. Atmospheric Environment, 45, 3996-4001.

Johnson, G.R., Ristovski, Z.D., Anna, B.D., Morawska, L. (2005). The hygroscopic behaviour of partially volatilized coastal marine aerosols using the VH-TDMA technique, Journal of Geophysical Research, 110, (D20203), doi:10.1029/2004JD005657.

Kulmala, M. (2003). How Particles Nucleate and Grow. Science, 302, 1000-1001.

Kulmala, M. and Kerminen, V-M. (2008). On the formation and growth of atmospheric nanoparticles. Atmospheric Research, 90, 132 - 150.

Kulmala, M., Vehkamäki, H. Petäjä, T., Dal Maso, M., Lauri, A., Kerminen, V.-M., Birmili, W. and McMurry, P.H. (2004). Formation and growth rates of ultrafine

16

atmospheric particle: a review of observations. Journal of Aerosol Science, 35, 3729-3739.

McMurry, P.H. (2000). A review of atmospheric aerosol measurements. Atmospheric Environment 34, 1959-1999.

Modini, R., Ristovski, Z., Johnson, G.R., Congrong, H., Surawski, N., Morawska, L.,

Suni, T., Kulmala, M. (2009). New particle formation and growth at a remote,

sub-tropical coastal location, Atmospheric Chemistry and Physics, 9 (19), 7607-

7621.

Morawska, L., Moore, M.R. and Ristovski, Z. (2004). Health Impacts of Ultrafine Particles: Desktop Literature Review and Analysis. Department of the Environment and Heritage, Australian Government.

Morawska, L., Ristovski, Z., Jayaratne, E.R., Keogh, D.U. and Ling X. (2008). Ambient nano and ultrafine particles from motor vehicle emissions: Characteristics, ambient processing and implications on human exposure. Atmospheric Environment, 42, 8113-8138.

Morawska, L., Wang, H., Ristovski, Z., Jayaratne, E.R., Johnson, G., Cheung, H.C., Ling, X. and He, C. (2009). JEM Spotlight: Environmental monitoring of airborne nanoparticles. Journal of Environmental Monitoring, 11, 1758-1773.

Pey, J., Rodiguez, S., Querol, X., Alastuey, A., Moreno, T., Pataud, J. P. and Van Dingenen, R. (2008). Variations of urban aerosols in the western Mediterranean. Atmospheric Environment, 42, 9052-9062.

Ristovski, Z.D., Suni, T., Kulmala, M., Boy, M., Meyer, N.K., Duplissy, J., Turnipseed, A., Morawska, L., Baltensperger, U. (2012). The role of sulphates and organic vapours in growth of newly formed particles in a eucalypt forest, Atmospheric Chemistry and Physics, 10, 2919-2926.

Weber, R.J., Marti, J.J., McMurry, P.H., Eisele, F.L., Tanner, D.J. and Jefferson, A. (1997). Measurements of new particle formation and ultrafine particle growth rates at a clean continental site. Journal of Geophysical Research, 102, 4375 – 4385.

Weber, R.J. and McMurry, P.H. (1996). Fine particle size distributions at the Mauna Loa observatory, Hawaii. Journal of Geophysical Research – Atmospheres, 101 (D9), 14767 – 14775.

17

Weber, R.J., McMurry, P.H., Mauldin, L., Tanner, D., Eisele, F., Brechtel, F., Kreidenweis, S., Kok, G., Schilawski, R. and Baumgardner, D. (1998). A study of new particle formation and growth involving biogenic trace gas species measured during ACE-1. Journal of Geophysical Research, 103, 16385 – 16396.

18

CHAPTER 2. LITERATURE REVIEW

This literature review introduces the topic of atmospheric particles, how they form

and their sources, followed by the environmental and human health impacts of urban

ultrafine particles (UFPs). The proposed nucleation mechanisms for atmospheric

particles and the controlling factors of nucleation are also discussed, together with

the findings of previous new particle formation studies conducted in urban

environments. Finally, the approaches to data analysis and the instrumentation

applied for new particle formation measurements are reviewed and the current gaps

in knowledge identified.

2.1 Introduction to atmospheric particles

Atmospheric particles (or particulate matter, PM) are defined as a mixture of liquid

and solid particles which are suspended in the atmosphere. They can originate from

natural activities such as volcanic eruptions, wind blown dust, sea spray and bushfire.

They can also be generated through anthropogenic sources such as vehicle exhaust

emissions, fuel combustion, domestic cooking and tobacco smoking. Particles can be

emitted directly from the above sources, and in that case they are referred to primary

particles, or they can be formed through a series of chemical reactions and gas to

particle partitioning where a gaseous precursor is emitted from the above sources,

and these particles are called secondary particles. Both primary and secondary

particles, once emitted into the atmosphere, will undergo physical processes of

growth, evaporation, condensation, coagulation, deposition and different chemical

reactions. Atmospheric particles can be classified by their size or formation

19

processes. Particles with a diameter less than 0.1 µm, 1 µm, 2.5 µm and 10 µm are

defined as ultrafine particles (UFP) (quantified in units of number concentration) and

PM1, PM2.5 and PM10 (quantified in units of mass concentration), respectively. Due

to the different measurement techniques for number and mass concentrations the

referred diameters are also different. The particle diameter referred to in number

measurement is the mobility diameter which is defined as the charged particles with

the same velocity as the spherical charged particles moving in an electric field; and

aerodynamic diameter for can be defined as the diameter of the particles with the

same settling velocity as spherical particles.

They can also be classified according to their formation processes, such as nucleation,

Aitken, accumulation and coarse modes etc. Nucleation mode particles, which range

in size from a few nanometres to tens of nanometres, are formed by nucleation

processes which start with formation of a molecular cluster and its subsequent

growth. Aitken mode particles, ranging from tens to hundreds nanometres, are

generated directly from primary sources or growth from nucleation mode particles.

Accumulation mode particles, ranging from tens of nanometres up to a few

micrometers in size, are mainly formed by the condensation of other gaseous species,

such as organics and sulphuric acid vapour on pre-existing particles, as well as by

coagulation with other existing particles. Coarse mode particles include wind blown

dust, sea salt particles and other larger particles generated by mechanical processes,

such as tire and engine wear particles. The above particle types also have different

physiochemical properties, such as growth, condensation and coagulation properties,

with different impacts on environment and human health.

20

2.2 Impact of atmospheric urban particles

Atmospheric particles, which originate from natural and anthropogenic sources, can

be found in both remote and urban areas. The emission rates of these sources in

terms of mass are listed in Table 2-1, which shows that natural source emissions

contributed almost 80-90% of total particle mass emissions. However, these natural

sources are widely distributed around the world, while the anthropogenic sources are

mainly located in densely populated areas. In urban areas, vehicle exhaust emissions

are the major contributor of total particle number concentration mainly in UFP size

range (Morawska et al., 2008).

Amount, Tg/yr [106 metric tons/yr] Source Range Best Estimate Natural

Soil dust 1000-3000 1500 Sea salt 1000-10000 1300 Botanical debris 26-80 50 Volcanic dust 4-10000 30 Forest fires 3-150 20 Gas-to-particle conversion 100-260 180 Photochemical 40-200 60

Total for natural sources 2200-24000 3100 Anthropogenic

Direct emissions 50-160 120 Gas-to-particle conversion 260-460 330 Photochemical 5-25 10

Total for anthropogenic sources 320-640 460 Table 2-1. Sources and Estimate of Global Emissions of Atmosphere Aerosol (Hinds 1999).

21

2.2.1 Impact on human health

Atmospheric particles have been found to have an adverse impact on human health.

According to a report by the United Nation Environment Program (UNEP) and

World Health Organisation (WHO), more than 500,000 people per year die from

diseases related to particulate air pollution (UNEP 2002). The health risk of these

particles was found to be related to their size, surface area and chemical composition.

A positive relationship was observed between mortality rate and the mass

concentration of PM2.5 and particle sulphate, but not total suspended particles (TSP)

(Pope 2000). Donaldson and Tran (2002) claimed that particles with a higher reactive

surface will induce a greater inflammatory response in the human lung. Nel (2005)

found that a number of illnesses on pulmonary were related to PM and that its ability

to obstruct the airway and decrease gas exchange can lead to the exacerbation of

asthma and chronic bronchitis. Furthermore, PM was also found to be related to

cardiovascular diseases, including heart attacks, stroke, heart rhythm disturbances

and sudden death.

In terms of the health implications of different PM size fractions, Pekkanen et al.

(1997) showed that the number concentration of UFP was more closely associated

with variations in peak expiratory flow than coarse particles. Similar findings were

obtained by Peters et al. (1997), who found that UFP were associated with a decrease

in peak expiratory flow and an increase in cough and feeling ill during the day.

Churg and Brauer (2000) indicated that UFP can penetrate deeper into the human

lung than fine and coarse particles, while the review by Morawska et al. (2004) on

the impact of UFP on human health has shown that the health impacts of particles

22

varied according to their size. In addition, Oberdörster and Utell (2002) suggested

that UFPs may cross the blood-brain and alveolar-capillary barriers and enter the

central nervous system. However, the nature and extent of the impact of UFP on the

central nervous system remains unclear, and further investigation is needed to fill the

scientific gaps in knowledge on this issue.

More recent epidemiological studies have shown the short-term and long-term

effects of UFP on human health. Belleudi et al. (2010) estimated the short-term

effects of UFP on hospital admissions for cardiac and respiratory diseases, using

case-crossover analysis with a time-stratified approach. The study found that the

PNC was associated with hospital admissions for heart failure (2.4% [0.2-4.7%]) and

chronic obstructive pulmonary disease (COPD) (1.6% [0.0-3.2%]) and those effects

were generally stronger for elderly people and during the winter season. Knol et al.

(2009) investigated the likelihood of health effects due to UFPs and suggested that

short term health effects, such as hospital admissions for cardiovascular and

respiratory diseases, aggravation of asthma symptoms and lung function decrements,

were related to the exposure of UFP. In contrast, the likelihood of an effect due to

long term UFP exposure was rated low to medium for those health effects.

2.2.2 Impact on environment

In addition to the impact on human health, atmospheric particles also influence the

environment, either directly or indirectly. For example, the interaction between

atmospheric particles and solar radiation can alter climate forcing by scattering light

back into the space which reduces the amount light reaching the earth and thus

23

results in a negative climate forcing (cooling) effect. At the same time, atmospheric

particles can also absorb solar radiation which causes a positive climate forcing

(warming) effect, with the overall effect on climate forcing depending on the

chemical composition of the atmospheric particles (Charlson et al., 1992). The

Intergovernmental Panel on Climate Change (IPCC) reviewed the estimation of

climate forcing for different major particulate matter components (IPCC 2007), the

findings of which are presented in Figure 2-1. The overall climate forcing from total

airborne particulate matter is estimated to be -0.5 ± 0.4 W m-2, while for individual

particle species, the estimates are as follows: sulphate, -0.4 ± 0.2 W m-2; fossil fuel

organic carbon, -0.05 ± 0.05 W m-2; fossil fuel black carbon, +0.2 ± 0.15 W m-2;

biomass burning, +0.2 ± 0.15 W m-2; nitrate, -0.1 ± 0.1 W m-2; and mineral dust, -

0.1±0.2 W m-2.

24

Figure 2-1. Radiative forcing estimated for different atmospheric components (IPCC 2007).

Besides the direct effect of atmospheric particles on climate forcing, atmospheric

particles can also influence the climate forcing indirectly by acting as cloud

condensation nuclei (CCN), hence altering the cloud formation (Twomey 1977).

Atmospheric particulates can alter the radiative budget of the Earth by modifying the

cloud droplet size spectrum and precipitation at the surface (Ramanathan et al., 2001).

Several studies investigated the indirect impact of urban particles on climate forcing.

For example, Rosenfeld (2000) found that anthropogenic CCN, which nucleate many

25

small cloud droplets, are highly inefficient at producing rain droplets, resulting in the

suppression of rain over urban areas. Kaufman and Koren (2006) estimated the

effect of urban pollution on cloud cover and found that an increase in cloud cover

followed increases in particle concentration in the aerosol column. The impact of

anthropogenic particles on temperature variation was also simulated in China and it

was found that during the daytime, temperature decreased by -0.7 ºC over the

industrial parts of China, due to the blockage of solar radiation by the clouds. In

contrast, the temperature increased by +0.7 ºC at night during the winter, due to the

long-wave surface warming effect (Huang et al., 2006).

In addition to the impact on climate forcing, atmospheric particles also impair

visibility. Particles with a diameter around 0.4-0.7 µm are the most effective at

scattering visible light, thus resulting in reduced visibility (Watson 2002).

Hygroscopic species, such as sulphate and organics, were found to be the major

components affecting visibility in urban environments (Dzubay et al., 1982). In

addition, urban pollution does not only affect local air quality, but it can also

significantly reduce visibility in downwind areas. For example, Cheung et al. (2005)

estimated that about 90 % of total light extinction in a downwind rural area was

attributable to upwind urban pollution plumes. Moreover, aerosols were also found to

be indirectly affecting the radiative forcing through the biogeochemical feedbacks

process (Mahowald 2011). Mahowald (2011) found that the biogeochemical cycle

had been modified by the deposition of aerosols which could supply either nutrients

or toxins suppressing growth.

26

2.3 New particle formation

With regard to the significant impact of atmospheric particles on human health and

the environment, scientists are interested on the formation processes of these

atmospheric particles. The new particle formation consists of two major processes: i)

nucleation and ii) growth. Firstly, the vapour molecules (e.g. sulphuric acid and

water) undergo a nucleation process to form a thermodynamically stable cluster

which can be in a neutral or charged state. Then the condensing vapours, such as low

volatile organics and inorganics (e.g. iodide, sulphuric and nitric acids), can

condense on those newly formed clusters to become larger particles. When combined,

the nucleation and growth processes can result in new particle formation, as

illustrated in Figure 2-2 (Kulmala 2003).

Figure 2-2. Illustration of particle formation process (Kulmala 2003).

27

Although the first particle measurements were conducted by Aitken in the late 19th

century, our understanding of particle formation processes was limited by lack of

adequate measurement techniques at the time. In the past decades, the advancement

of particle measurement techniques has allowed for measurements to be conducted to

an accuracy as low as 2.5 nm diameter (Kulmala et al. 2004). This improvement in

measurement techniques has enabled scientists to study the formation processes of

nucleation mode particles.

Kulmala and Kerminen (2008) summarised the various formation mechanisms which

have been proposed by previous studies, including:

i) Homogenous binary nucleation involving mixtures of water and sulphuric

acid;

ii) Homogeneous ternary nucleation involving water, sulphuric acid and

ammonia in the continental boundary layer;

iii) Ion-induced nucleation of binary, ternary or organic vapours which are found

in the upper troposphere and lower stratosphere; and

iv) Homogeneous nucleation involving iodide species in coastal environments.

2.3.1 Homogeneous binary nucleation

Homogenous binary nucleation consists of two substances, i) sulphuric acid (H2SO4)

and water (H2O). Sulphuric acid is formed by the oxidation of sulphur dioxide (SO2)

in the atmosphere. The product of H2SO4 and water has a lower vapour pressure than

28

the H2SO4 itself (Marti et al., 1997). In addition, water vapour and sulphuric acid are

widely present in the Earth’s atmosphere and therefore, the reaction between H2SO4

and water is likely to occur in the atmosphere. When the H2SO4 molecules collide

with other H2SO4 and H2O molecules, they will form a cluster and this cluster

continues to grow and overcomes the nucleation barrier. Then a thermodynamically

stable cluster is formed from the gas phase. This mechanism is called homogeneous

binary nucleation, because no other catalyst like a foreign surface is involved in the

formation.

Vehkamäki et al. (2002) developed a binary nucleation model which showed a good

agreement with the theoretical rate for nucleation within the temperature range 230-

300 K, at a relative humidity between 0.01-100 % and with a total sulphuric acid

concentration between 104-1011 cm-3. Brock et al. (1995) tested the volatility of

particles collected in the upper troposphere (~ 10 km altitude) and found that 90 % of

the particles with diameter below 40 nm were composed of H2SO4 and H2O. Based

on this finding, they went on to suggest that binary nucleation was a possible

mechanism for the new particle formation in upper troposphere.

However, nucleation rates for the binary nucleation of H2SO4 and H2O were often

underestimated compared to observational results, whereby the predicted nucleation

rate for H2SO4 and H2O nucleation was lower than the observed nucleation rate by

factors of 107 at low altitudes. Similar findings were obtained by Weber et al. (1996,

1997 and 1998), which showed that the binary nucleation model underestimated the

nucleation rate of nano-sized particles (with diameter of 3-10 nm). This discrepancy

29

between the modelling and observational results implies that an additional formation

mechanism must exist.

2.3.2 Homogeneous ternary nucleation

Homogenous ternary nucleation, which involves sulphuric acid, water and ammonia

(NH3), is proposed to explain the particle formation process, since NH3 is in

abundance in troposphere and is able to decrease the vapour pressure of sulphuric

acid, and therefore enhance the nucleation rate (Scott and Cattell 1979, Coffman and

Hegg 1995).

This hypothesis is supported by a number of modelling studies which have shown

that higher particle nucleation rates were predicted based on ternary nucleation, as

opposed to a binary nucleation mechanism. For example, Ball et al. (1999) reported a

modelling result which showed that the homogenous nucleation rate was enhanced

by several orders of magnitude with the presence of ammonia, which was in a good

agreement with the experimental results conducted at a temperature of 298 K. Napari

et al. (2002) also constructed a modified parameterisation of ternary nucleation and

showed that an increase of one order of magnitude in the ammonia mixing ratio can

increase the nucleation rate by several orders of magnitude. Recently, Jung et al.

(2008) compared six nucleation parameterisations, including the binary and ternary

nucleation, ion-induced nucleation, semi-empirical first order nucleation and barrier-

less nucleation, with observational data in a sulphur rich environment. The results

showed that only ternary nucleation correctly simulated the particle burst in all

nucleation events. However, the input value of the ammonia concentration for the

30

modelling was usually higher than that in the actual concentration in the atmosphere.

Larsen et al. (1997) estimated the particle formation rate by ternary nucleation using

an ammonia concentration of 500 ppb, however the general atmospheric ammonia

concentration was in a range 10-1000 ppt. Therefore, other mechanisms were

proposed to explain the particle formation process.

2.3.3 Ion induced nucleation

Ion-induced nucleation (IIN) is another possible mechanism for particle formation.

IIN has higher particle growth rates due to the presence of electrostatic forces, which

enhances the stability of the electrically charged cluster. Using ion-induced

nucleation model simulation, Yu and Turco (2000, 2001) showed that charged

clusters have a faster growth rate than neutral clusters in a range of different

environments, including the upper troposphere, lower stratosphere and boreal forests.

Laakso et al. (2007) conducted measurements of charged and neutral particles in the

boreal forests of Finland and found that charged particles outnumbered the neutral

particles during some nucleation events, therefore indicating that IIN was the

possible formation mechanism contributing these events. In addition, it was

suggested that IIN was also occurring in upper troposphere and lower stratosphere.

Lee et al. (2003) simulated nucleation events by using in-situ measurement data of

ion-clusters and other particle precursors. The results showed that IIN was observed

under low surface area and low temperature conditions, however, the new particle

formation attributable to IIN only represented a small fraction of overall new particle

formation. Kulmala et al. (2007) also conducted measurements of neutral and ion

31

clusters down to a size below 1 nm in the boreal forests of Hyytiälä, Finland. In this

study, the results showed that both neutral and ion clusters (with a mean diameter ~

1.5-1.8 nm) were present almost all of the time during nucleation events. The

concentration of air ion clusters was found to be around 10-100 cm-3, which was

significantly lower than the concentration of neutral clusters, which ranged in order

of 1000 cm-3 during the study period. Therefore, it was suggested that less than 10 %

of the formation rate of particles with diameter of 2 nm was attributable to ion-

induced mechanisms under boreal forest conditions.

2.3.4 Organics involved nucleation

As mentioned in Section 2.3.1, although homogeneous binary nucleation is thought

to be one of the primary mechanisms involved in new particle formation, the particle

formation rate simulated by the nucleation model could not explain the observational

results. Therefore, it has been suggested that organic species may contribute to the

particle growth process, in order to enhance the particle production rate. One

possible pathway for particle formation involving organic species is that organic

species condense onto the stable cluster formed by H2SO4 and water, thus enhancing

the particle growth. Several studies have shown a positive relationship between the

mixing ratio of organic species and the particle formation rate (O’Dowd et al., 2002;

Tunved et al., 2006; Laaksonen et al., 2008). For example, O’Dowd et al. (2002)

found that, in a forest environment, most nucleation events occurred when an

elevated biogenic organic acid mixing ratio was also observed. Tunved et al. (2006)

also found that higher particle formation rates were linked to higher monoterpene

emissions. The evolution of organic aerosols have been studied by Jimenez et al.

32

(2009), using the high-time resolution measurement of aerosol chemical composition.

The result of the study shown that the organic aerosols in a downwind area were

increasingly oxidized, less volatile and more hygroscopic compared to the aerosols

measured in an upwind area. Therefore, both of these studies indicated that organic

species may have contributed to the particle formation process.

2.3.5 Iodide species involved nucleation in coastal environment

In coastal environments, iodide species were generated from macroalgal iodocarbon

emissions and iodide oxides were observed in the marine particles. Therefore, iodide

species were proposed as a possible contributor to the particle formation process in

marine environments. O’Dowd et al. (2002) suggested that particle formation rate

can be enhanced when condensation iodine vapours (CIVs) are present during

homogeneous nucleation. The research simulated homogenous binary nucleation of

H2SO4 and water with different concentrations of CIVs, the results of which showed

that the higher the CIVs concentration applied, the higher the particle formation rate

that was observed. The results suggested that CIVs could be one of the possible

candidates involved in marine particle formation.

2.3.6 Other possible nucleation mechanisms

In addition to the mechanisms mentioned above, several other species were found to

enhance particle formation rates, including amino acids and oxidation products of

isoprene (Froyd et al., 2009; Paulot et al., 2009; Loukonen et al., 2010).

33

In more recent studies, ammines were found in freshly formed particles (Makela et

al., 2001) and its presence was found to enhance particle formation rates in both

laboratory and modelling studies (Barsanti et al., 2009; Wang et al., 2010).

Furthermore, dimethylamine was found to have a greater effect (by lowering the

nucleation energy barrier) than ammonia at enhancing particle formation rates when

coupled with sulphuric acid when modelled using quantum chemistry (Loukonen et

al., 2010). However, the accurate measurement method of ammines close to the

cluster scale is still under development and more effort is needed to help understand

the impact of ammines on particle formation processes (Kurtén et al., 2011).

Similarly, the role of isoprene in new particle formation also remains unclear.

Although isoprene is a major component of biogenic organics, its oxidation products

were found in the composition of secondary organic aerosol (Froyd et al., 2009).

More recent studies on the effect of isoprene on particle formation using plant

chamber (a chamber with different kind of trees growing in it) found that isoprene

suppressed the particle formation rate (Kiendler-Scharr et al., 2009). In this plant

chamber experiment, the authors found that PNC decreased while the isoprene

mixing ratio increased. To further support this finding, Kanawade et al., (2011)

conducted field measurements in a Michigan forest (United States) to study the

relationship between particle nucleation and isoprene. They found that no nucleation

events were observed while the ratio of isoprene/monoterprene was higher than 10.

In contrast, a low isoprene/monoterprene ratio (< 0.5) was observed in the boreal

forests of Finland, where nucleation events frequently occurred (Spirig et al., 2004).

These findings indicate that isoprene has a negative effect on particle formation rate.

34

However, further research is required to fully understand these recently proposed

nucleation mechanisms.

In the above discussion, it was shown that either binary or ternary nucleation

mechanisms cannot fully explain the observed particle formation rates under ambient

atmospheric conditions. Whilst the IIN mechanism has been observed in the

atmosphere, it only contributed a small fraction to total particle formation rates. Also,

the role of organics species in the nucleation process remains unclear, and therefore,

more effort is needed to investigate fundamental nucleation mechanisms.

2.4 Field studies of new particle formation

In Section 2.1 and 2.2, the emission sources of atmospheric particles in different

urban and rural environments were discussed and although total particle emissions

from anthropogenic activities were lower than that of natural sources, anthropogenic

emissions were more concentrated in populated areas, where people are more

exposed to the polluted atmosphere. Also, urban and rural atmospheric conditions are

usually different in terms of particle formation precursors and the diurnal profile of

the emission sources etc. These differences have a direct impact on the particle

formation processes which occur in these environments. For example, Kulmala et al

(2004) found that the level of pre-existing particles was higher in urban areas

compared to rural areas, which served to suppress the new particle formation process.

In addition, the dominant precursors are often different in urban and rural areas,

which lead to different formation and growth properties of the atmospheric particles.

Therefore, the investigation of urban particle formation is important for gaining a

35

better understanding of the specific nucleation processes which occur in these

environments.

This section will discuss the classification scheme used for nucleation events,

according to changes in particle size distribution, as well as how meteorological,

physio-chemical and other factors affect the nucleation process in urban areas.

Furthermore, the analysis techniques applied in the study of particle nucleation,

including data analysis and instrumentation approaches will also be discussed.

2.4.1 Classification of nucleation events

Statistical analysis of nucleation event/non-event data with other measured

parameters is a basic approach to study the factors governing the particle formation

process. Identifying nucleation events is the first step, whereby a nucleation event is

usually characterised by an increase in the number of nucleation particles which

gradually grow to larger particles. Based on the above system, different researchers

use different approach to the classification of events, and therefore, Dal Maso et al.

(2005) established a scheme for classifying nucleation events, which has been

applied in a number of nucleation studies (e.g. Lee et al., 2008; Sogacheva et al.,

2008; Vana et al., 2008). According to this scheme, a nucleation event is classified

by measuring the new nucleation mode particles (< 25 nm size) that are formed over

a one hour period, as well as quantifying the subsequent particle growth. When the

particle size distribution data is presented in a contour plot, the shape of the

propagation should look like a "banana" (see Figure 2-3 as an example). This

“banana” shape is a typical feature of the nucleation event. Furthermore, nucleation

36

events can be defined as Class I, Class II, a Non-Event or an undefined group,

depending on several criteria used. Further information about this event classification

procedure are provided by Dal Maso et al. (2005). In the urban environment,

nucleation events have been observed with and without particle growth (Park et al.,

2008; Gao et al., 2009). For example, in addition to nucleation events, observation of

increases in nucleation mode particle concentration during the daytime, where the

particles did not grow into larger particles (indicated by the near constant geometric

median diameter (GMD) value during the event period), which is classified as a

nucleation burst event.

Figure 2-3. Example of a contour plot of particle size distribution during a nucleation event (Bottom). The propagation of the particle size distribution displays a “banana” shape. (Dal Maso et al. 2005).

37

2.4.2 Factors favouring urban nucleation

2.4.2.1 Solar radiation

When investigating the meteorological conditions which affect the nucleation

process, solar radiation was directly linked to new particle formation due to its

positive relationship with photochemical reactions. For example, the amount of

hydroxyl radicals was positively related to the strength of solar radiation. In previous

studies, nucleation events were often observed during late mornings when precursors

were present in the atmosphere and solar radiation was close to reaching its daily

maximum (see Figure 2-4, Woo et al., 2001; Qian et al., 2007). These photochemical

activities also enhanced photo-oxidation reactions which led to a higher particle

growth rate. The seasonal variation of particle growth rate was also positively related

to the strength of solar radiation, with higher particle growth rates observed during

the warmer summer season (see Figure 2-5, Salma et al., 2011).

38

Figure 2-4. Time series plot of i) particle number concentration (size ranging from 3-10 nm), SO2, NO2 and O3 (bottom), and ii) solar radiation (upper) (Woo et al. 2001).

39

Figure 2-5. Seasonal variation of the particle growth rate observed during nucleation events in 2008-2009 at Budapest, Hungary. (Salma et al. 2011).

2.4.2.2 Temperature

Temperature affects the condensation and volatilisation of particles, such that cooler

conditions enhance the condensation process of organic vapours on the cluster

molecule, as well as the nucleation resulting from vehicle exhaust during morning

hours (Morawska et al., 2008). In contrast, warmer conditions may induce the

volatilisation of particles during the nucleation growth process. The volatile species

of a particle will be volatilised under high temperatures, reflected by a decrease in

particle size after the particles have been formed in the atmosphere (Yao et al., 2010).

This particle volatilisation process is called the particle shrinkage. Figure 2-6 is an

example of particle shrinkage observed during a late afternoon nucleation event

which suggests that the semi-volatile species of particles will undergo evaporation or

gas/ particle repartitioning (Yao et al., 2010). Although higher solar radiation is

40

supposed to be associated with higher temperatures, the principle of particle

shrinkage is still not well understood and needs further investigation.

Figure 2-6. Observation of particle growth followed by particle shrinkage. (Yao et al., 2010).

2.4.2.3 Relative humidity

Relative humidity is another meteorological parameter affecting nucleation events.

For example, condensation onto particles is enhanced under high relative humidity

conditions, which leads to an increase in accumulation mode particles, as well as

total particle volume (Yue et al., 2009). Therefore, high relative humidity suppresses

new particle formation by increasing the condensation sink of the atmosphere. Since

variation in relative humidity is usually inverse to variation in solar radiation, low

relative humidity is often observed during particle formation events, due to the

photochemical reactions which occur during high solar radiation conditions. This

phenomenon was also observed during nucleation growth events which occurred at

noon in urban areas of New York, United States, with minimum relative humidity

observed during these events (Jeong et al., 2004).

41

2.4.2.4 Wind direction and speed

The influence of wind on the particle formation process can be discussed in terms of

both wind direction and wind speed. Firstly, observed pollution plumes are often

related to wind direction, which can transport pollution from the emission source

regions, as well as clean air masses to the measurement site. In urban environments,

PNC is contributed by different sources, such as traffic emissions, industrial

activities and power generation, which are mixed together in the urban airshed. A

number of studies have investigated the impact of different pollution sources on

PNCs. For example, Bein et al. (2007) estimated the impact of industrial pollution

sources on an urban site in Pittsburgh, United States by using the unique source

markers of coal combustion with back-trajectory analysis. The results showed that

PNC could reach orders of magnitude of 1015-1017 particles/cm3 when the

measurement site was downwind from the coal burning pollution plume. This result

was higher than the PNC attributable to traffic emissions, with an order of magnitude

around 1012 particles/cm3.

In contrast to wind direction, wind speed mainly impacts the dispersion and

coagulation of particles. The relationship between wind speed and the PNC is often

presented by a “U” shaped curve (Charron and Harrison 2003; Hussein et al., 2006).

This phenomena can be explained by the fact that sub-micrometer particles consist of

two major components, i) UFPs which are generally diluted by increasing wind

speed, and ii) particles between 100 nm and 2.5 µm in diameter which are generally

proportional to wind speed due to the suspension and re-suspension of particles. This

42

relationship occasionally altered the condensation sink profile of the pre-existing

particles.

2.4.2.5 Sulphuric acid

Sulphuric acid (H2SO4) plays a critical role in binary and ternary nucleation, which is

mainly formed by oxidation reactions of SO2 in the atmosphere (Lovejoy et al., 1996;

Jayne et al., 1997). The heterogeneous chain reactions for SO2 to form H2SO4 are as

follows:

SO2 + OH + M -> HSO3 + M (R1)

HSO3 + O2 -> SO3 + HO2 (R2)

SO3 + H2O -> H2SO4 (R3) where M is catalyst

Although atmospheric H2SO4 is considered to play a major role in nucleation, the

measurement of H2SO4 is quite difficult and thus, SO2 is often used as an indirect

indication of the amount of H2SO4 in the atmosphere (Woo et al., 2001; Stanier et al.,

2004). Observational results have shown that nucleation bursts frequently occur at

the same time as increases in the SO2 mixing ratio (e.g. Woo et al., 2001; Park et al.,

2009; Salma et al., 2011). Figure 2-7 shows the temporal variation of SO2 during a

nucleation event in an urban area of Atlanta, United States and it can be seen that the

SO2 plume was usually observed at the same time as a nucleation burst. A product of

SO2 mixing ratio and solar radiation/UV can also be used to indicate the production

of H2SO4 (Gao et al. 2009), while a separate parameter, H2SO4 proxy, can also be

linked to the strength of particle precursors containing SO2, as well as to solar

43

radiation and the condensation sink (Petäjä et al., 2007; Salma et al., 2011). H2SO4

proxy is an approximation of the nucleation mode particles which have been formed

by H2SO4. It is calculated by the available H2SO4 (usually estimated by the ambient

SO2 and solar radiation) over the condensational sink of particles. Salma et al. (2011)

calculated the average H2SO4 proxy values for nucleation event days and non-event

days over four seasons. The results showed that SO2 values did not vary significantly

across the four seasons. By contrast, the H2SO4 proxy values showed higher values

during nucleation event days than non-event days.

Figure 2-7. Concentrations of selected gases during a 3 – 10 nm particle event on April 1, 1999 at Atlanta, United States. (Woo et al., 2001).

44

2.4.2.6 Volatile Organic Compounds (VOCs)

As mentioned in Section 2.3.3, organic species have been suggested as an important

factor for enhancing the particle formation process. A number of studies which were

conducted in the field, in the laboratory and by modelling, also found that particle

growth rates were related to the amount of organic species involved in the nucleation

(e.g. O’Dowd et al., 2002; Laaksonen et al., 2008). Laaksonen et al. (2008) analysed

the chemical composition of particles during and after nucleation events using a

variety of techniques. The mass spectra of the particles collected during and after the

nucleation were very similar, implying that similar secondary organic species were

condensing onto all particles. Furthermore, the particle growth rates during the

nucleation events were found to be related to the concentration of monoterpene

oxidation products (MTOP) and a linear fit was found based on the following

equation, [Growth rate] = 0.97 + 0.0043*[MTOP], with a coefficient of

determination (r2) of 0.965. Both these results provided evidence that VOC

oxidation products may play a key role in determining the strength and rate of the

nucleation process.

2.4.3 Formation and growth rates of atmospheric aerosol particles

Particle formation rate describes the amount of critical clusters formed during the

nucleation process, which can be used to estimate the activation coefficient (Kulmala

et al., 2006). Particle growth rate is defined as the change in particle size per unit of

time as a result of the particle growth process, which can be used to calculate the

concentration of the condensation vapour and its source rate (Kulmala et al., 2006).

Spracklen et al. (2008) also used particle growth rate to estimate the amount of cloud

45

condensation nuclei (CCN) in climate modelling. This section will present the

mathematical definitions of particle formation and growth rates, as well as their

observed values in urban environments.

2.4.3.1 Estimation of the particle formation and growth rates

Due to instrument limitations, it is not possible to quantitatively measure the critical

clusters which are formed during atmospheric nucleation. At present, it is only

possible to detect particles down to a size of ~ 2.5 nm. Therefore, the nucleation rate

of clusters cannot be measured, only estimated, based on the following equation

(Kulmala et al., 2004):

eq. (1)

where particle formation rate, JD, is defined by the flux of particles reaching the size

D as a result of the particle growth process, t is the time, and n(Dp, t) represents the

particle number size distribution. However, it is quite difficult to determine particle

number size distribution and particle growth rate at size D and therefore, instead of

estimating the instantaneous particle formation rate JD(t), the average particle

formation rate, JD, over time interval ∆𝑡 , which is the duration of the particle

formation event, is often used. After time averaging, the equation can be rewritten as

follows:

46

eq. (2)

In eq. (2), ND, Dmax is the total PNC in the size range [D, Dmax] and Dmax is the

maximum size the critical clusters can reach because of their growth during ∆𝑡. The

left hand side of eq. (2) is the observed change in ND, Dmax during ∆t and can be

obtained from particle size distribution or number concentration measurements. On

the right hand side of the equation, the first and second terms represent the loss of

particles by self-coagulation and coagulational scavenging to larger pre-existing

particles in the size range [D, Dmax]. The last term on the right hand side represents

the external influence of other air masses on ND, Dmax.

When the effect of transport are small, eq. (2) can be simplified to

JD = [∆ND, Dmax/ ∆t]observed + [∆ND, Dmax/ ∆t]self-coag + [∆ND, Dmax/ ∆t]coag-scav eq. (3)

However, it may significantly underestimate the actual particle production rate when

the nuclei PNC is very high (> 105 cm-3). Also, if the pre-existing particle

concentration is very high, JD values can be under-estimated.

47

To estimate the particle growth rate, GR, the time evolution (Δt) of the lowest

detectable particle diameter to the upper limit of the nucleation mode particle, ΔD

(e.g. 25 nm) is needed. The GR can be calculated by:

GR = ΔD/Δt eq. (4)

2.4.3.2 Observed particle formation and growth rates in urban environments

The observed particle formation and growth rates in urban areas varied depending on

atmospheric conditions. In Table 2-2, particle growth rates ranged from 0.5-16 nm h-

1 for urban sites and the reported values for regional and local nucleation events were

similar. These observed particle growth rates were lower than that in marine areas,

which reached values > 100 nm h-1 (O’Dowd et al., 2007; Ehn et al., 2010). Higher

particle formation rates have also been observed for regional nucleation events,

which can reach values of 70-80 cm-3 s-1 compared to values of 0.2-50 cm-3 s-1 for

local nucleation events.

48

Locations GR (nm h-1) J3 (cm-3 s-1) References

St. Louis, United States 0.5-9 1-80 (regional) Shi et al. (2007)

Atlanta, United States 2-6 20-70 (regional) McMurry et al. (2003)

Birmingham, United Kingdom 4 5-50 Shi et al. (2001)

Leipzig, Germany -a 13 (±1.2) Wehner and Wiedensohler (2002)

Luukki, Finland 4-6 -a Väkevä et al. (2000)

Manchester, United Kingdom 8 >>0.2 Williams et al. (1998)

Pittsburgh, United States 4-5 -a Stanier et al. (2002)

Vienna, Italy -a 2.5 Winklmayer (1987)

Atlanta, United States -a 10-15 Woo et al. (2001)

New Delhi, India 11.6-16 -a Kulmala et al. (2005)

Marseille, France 1.1-8.1 -a Kulmala et al. (2005)

Marseille, France 2-8 3-5.3 Petäjä et al. (2007)

Athens, Greece 2.3-11.8 -a Kulmala et al. (2005)

Athens, Greece 1.2-9.9 1.3-6.5 Petäjä et al. (2007)

Harrow, Canada 6.4 4.4 Jeong et al. (2010)

Ridgetown, Canada 4.7 3.3 Jeong et al. (2010)

Bear Creek, Canada 2.9 3.6 Jeong et al. (2010)

Egbert, Canada 5.1 4.7 Jeong et al. (2010)

Toronto, Canada 6.7 1.1 Jeong et al. (2010)

Budapest, Hungary 2.0-13.3 (7.7±2.4)

J6:1.65-12.5 (4.2±2.5) Salma et al. (2011)

Note: a) ‘-‘ Data not available

Table 2-2. Summary of particle growth and formation rates in urban areas.

49

Section 2.4 summarised the factors controlling the nucleation process, such as

meteorological parameters and the nucleation mode particle precursors. Nucleation

was found to be favoured under strong solar radiation, low temperature conditions,

with a wind direction that transported regional pollution plumes towards the

nucleation site. Sulphuric acid and H2SO4 proxy were also found to be positively

related to the level of PNC, but the contribution of organic species remains unclear.

The variation in particle growth and formation rates observed for different urban

areas is usually the result of the different atmospheric conditions in each area and

higher particle formation rates were generally observed during regional nucleation

events compared to local events.

2.5 Characterisation of particle formation processes

2.5.1 Temporal variation

Characterisation of the temporal variation of particle size distribution and other

relevant factors is a common approach to explain which factors control particle

formation. In past studies, researchers revealed that sulphuric acid plays an important

role in particle formation in rural areas, whereby an increase in particle number was

often observed about 1-2 hrs after an increase in sulphuric acid concentration (Weber

et al., 1995, 1997; Charron et al., 2007). The time lag between these two parameters

suggests that it took some time for the nucleation process to occur and for the

particles to grow to the minimum detectable size (Weber and McMurry 1996).

Statistical analysis of nucleation particles and meteorological parameters is another

technique to understand the favourable atmospheric conditions related to nucleation

burst events. Researchers have found that the temporal variation of nucleation events

50

did not show a clear seasonal pattern, with nucleation burst events occurring in

different seasons for different study areas (e.g. Boy et al., 2003; Hyvärinen et al.,

2008; Sogacheva et al., 2008). Based on the diurnal variation of observed parameters,

nucleation bursts usually started in late mornings when temperatures were relatively

high and humidity was relatively low.

2.5.2 Backward trajectory technique

To better understand the relationship between nucleation events and the origin of air

masses, a number of back trajectory studies were conducted (e.g. Hussein et al., 2009;

Hyvärinen et al., 2008; Kristensson et al., 2008). Several models were applied for

this trajectory analysis, including HYSPLIT and FLEXTRA. Detailed descriptions of

the models and their accuracy are described in Draxler and Hess (1998), Stohl et al.

(1995) and Stohl and Seibert (1998). Using backward trajectory analysis, Hyvärinen

et al. (2008) categorised the wind sectors corresponding to regional nucleation events

in Utö, Baltic Sea. The results showed that nucleation events were related to NW-

NNW wind sectors, which originated from the boreal forest areas in Finland. In

addition to the origin of the air masses, trajectory analysis provides the speed of the

air masses during transportation. Hyvärinen et al. (2008) also noticed that the air

masses had spent more than 10 hr over the Baltic Sea on only 19% of the nucleation

days. Although the authors did not find any clear correlation between these two

parameters, they demonstrated that after spending sufficient time over the sea, no

nucleation events were expected to take place when these air masses were

transported to the measurement site.

51

2.5.3 Modelling of particle formation mechanisms

Two types of modelling techniques were applied for the particle formation study: i)

the mathematical approach (i.e. parameterisation model) was used to explain the

relationship between the input variables by an empirical equation and the other is the

chemical model (i.e. Chemical Transport Model (CTM)/ Chemical Box Model

(CBM), which includes the consideration of physical, chemical and meteorological

processes in nucleation. The parameterisation of formation mechanisms has been

investigated by numerous groups, including ternary nucleation (Napari et al., 2002),

binary nucleation (Vehkamäki et al., 2002; Jaecker-Voirol and Mirabel 1989) and

ion-induced nucleation (Modgil et al., 2005), while others used parameterisation to

explain in-situ particle formation (e.g. Weber et al., 1999; Kulmala et al., 1998; Yu

2006; Jung et al., 2008). Binary nucleation is proposed to be a principle formation

mechanism, since H2SO4 and water are basically presented anywhere in the

atmosphere. Weber et al. (1999) predicted the binary nucleation rate in remote

troposphere regions at altitudes greater than ~ 4 km above sea level and found that

the concentration of sulphuric acid vapour observed in the atmosphere was sufficient

to reproduce the observed nucleation rate. However, this study also pointed out that

binary nucleation may not be valid for the warmer temperatures observed at lower

altitudes, where sulphuric acid concentration was lower due to the higher saturation

vapour pressure. This implied that ternary nucleation was a more significant

mechanism for particle formation. Further ground based nucleation studies were also

conducted which showed similar levels of particle formation due to ternary

nucleation (e.g. Birmili et al., 2000, Jung et al., 2008 Kulmala et al., 2000).

52

Therefore, it is clear that the dominant nucleation process varies under different

environmental conditions. Although, the parameterisation method can simulate the

new particle formation event, the calculated formation rates of particles were

underestimated by the input data which were observed in the ambient. Thus, the

CTM approach was applied to better simulate the nucleation including the physical,

chemical and meteorological processes. Zhang and Wexler (2002) conducted a

timescale analysis of aerosol dynamic under urban conditions. Their result have

shown that the process of condensation, coagulation, nucleation and emission were

the dominating factors affecting the particle size distribution in the urban

environment. A reasonably well matched result for the particle number concentration

and new particle formation events were obtained by the Weather Research and

Forecast model coupled with Chemistry (WRF-Chem), compared to the observation

data in lower troposphere of the eastern United States (Luo and Yu 2011). Although,

CTM needed a larger input dataset than that of the parameterisation model and

higher uncertainty was found for a large scale study (i.e. long-range transport from

outside the model domain) (Chang et al., 2009). This approach can provide however,

an insight to the nucleation mechanism in the local/ regional study. If the simulated

result does not fit with the observed data, this implies that there are some currently

unknown mechanisms/ theories involved in the nucleation.

Section 2.5 discussed several data analysis techniques applied in urban nucleation

studies, however there are still some areas of knowledge which need to be improved

53

in order to gain a better understanding of the nucleation process which occur in urban

environments. The backward trajectory technique is usually applied to trace the

medium and long range (from tens to hundreds of kilometres) transportation of air

masses, however the grid resolution is not accurate enough to identify the path of air

movement in a scale of 10 km (or lower). Usually urban emission sources are located

very close to each other, and the air masses contain a mixture of pollution from

different sources, which makes it hard to identify the primary pollution sources

affecting urban nucleation events. Further studies on particle source identification

during nucleation events could help to better understand the primary contributors to

new particle formation. Furthermore, the results of parameterisation studies showed

that different nucleation mechanisms can dominate in the same study location

depending on the atmospheric conditions. Therefore, more field studies are also

needed in different study environments, to better understand the nucleation process

in specific study areas.

2.6 Development of measurement techniques

The instrumental techniques for measuring UFP have improved significantly over the

past decades. The development of condensation particle counters (CPCs) was

initiated in the 1970s (Sinclair and Hoopes 1975; Bricard et al., 1976), and

commercialised in the 1980s (Kousaka et al., 1982). More recent CPCs, which can

detect particle as small as 2.5 nm, were developed by TSI Inc. (Models 3776 and

3786) and Scanning Mobility Particle Sizers (SMPS), which consisted of Differential

Mobility Analyser (DMA) and CPC, were also developed to measure the size

54

distribution of particles. After the DMA separates the particles in the selected size

range, the concentration of particles in the desired size range is measured by the CPC.

The complete size distribution of particles can be obtained by changing the size

window of the DMA. Since the development of the SMPS, numerous studies were

conducted to investigate the particle formation processes and many of the

observational based studies were reviewed by Kulmala et al. (2004).

A technique for measuring the size distribution of charged particles, which can detect

ion clusters down to ~2.5 nm in size, has only become available in recent years and

measurement results have provided insight into particle formation by ion-induced

nucleation (e.g. Kulmala et al., 2007; Vana et al., 2008). Three very recently

developed instruments have also been applied to measure neutral and charged ion

clusters with diameters down to 1 nm and 0.3 nm, respectively, namely a Balanced

Scanning Mobility Analyser (BSMA), an Air Ion Spectrometer (AIS), and a Neutral

Cluster Air Ion Spectrometer (NAIS). Using these instruments, Kulmala et al. (2007)

showed that both neutral and charged ion clusters (with a mean diameter ~1.5-1.8 nm)

were present almost all the time during nucleation events in the boreal forests of

Hyytiälä, Finland and suggested that atmospheric aerosol formation started with

particle ~1.5 nm in size. For < 3 nm clusters, the concentration of air ion clusters was

around 10-100 cm-3 which was significantly lower than the concentration of neutral

clusters, which was ranged from 1500-3000 cm-3 during the study period. Therefore,

the authors suggested that less than 10% of the particles formation rate was

contributed by ion-induced mechanisms under boreal forest conditions.

55

Chemical analysis of atmospheric UFP is limited, with Viana et al. (2008) pointing

out that only 1% of previous studies on airborne particulate pollution target UFPs

and hence, there is limited information available on the source profile of UFPs.

Traditional methods used to measure the chemical composition of UFP are based on

the offline technique of collecting the aerosol on the filter, followed by chemical

analysis in laboratory (McMurry 2000). More recently, several instruments have

been developed, based on the mass spectrometry technique, for measuring the real-

time chemical composition of UFPs (Suess and Prather 1999). Most of these mass

spectrometry instruments enable the measurement particles down to ~20 nm in size.

For the chemical characterisation of particles below 10 nm, the Thermal Desorption

Chemical Ionisation Mass Spectrometer (TDCIMS) is capable of measuring the

molecular composition of atmospheric aerosol in the range 4 -10 nm (Voisin et al.,

2003). The TDCIMS consists of an electrostatic precipitator for collecting charged

particles, which is equipped with a nanometre aerosol differential mobility analyser

to separate the sub-10 nm particles. After collection, the sample is analysed by the

evaporation-ionisation technique. A more detailed description of the system can be

found in Voisin et al. (2003).

Although several instruments are available to measure the chemical composition of

UFP, most of them are not commercially available. The Aerosol Mass Spectrometer

(AMS), which is available on the market, measures the chemical composition of UFP

down ~40 nm in size. Rhoads et al. (2003) used the AMS to measure the real-time

chemical composition of semi-volatile species in Atlanta, United States. In this study,

organic species contributed about two thirds of the total measured chemical mass of

56

UFP. It should be noted that sulphate was not measured in this study, which was also

an important component in atmospheric aerosols. Pakkanen et al. (2001) also

measured chemical composition of UFP in urban and rural areas in Helsinki, Finland

and found similar chemical compositions in both areas. In this study, the measured

components accounted for about 15-20% of total UFP mass. Sulphate and

ammonium were the major measured components which contributed about 6-8%,

and 4-5% of the total ultrafine masses (490-520 ng/m3), respectively. This suggested

that carbonaceous species contributed about 70% of the total ultrafine mass which

was not accounted in the composition measurement.

In conclusion, the techniques for measuring particle chemical composition have

improved significantly in recent years. This has allowed the study of the chemical

composition of particles down to 4 nm in size, however these instruments are still not

widely available, which limits the progress of the study of nucleation.

2.7 Research problems

Based on the above overview of previous research on particle formation, the major

scientific issues that have been identified can be summarised as follows:

- Intensive field measurements of neutral/charged clusters were mainly

conducted in clean environments, such as continental rural area/boreal forests,

arctic and marine environments, and the upper troposphere. These studies

were mostly conducted in the northern hemisphere and they showed that

nucleation processes were complex in different environmental settings; and

57

- Several parameterisations of particle formation were developed for

homogeneous binary/ternary nucleation, ion-induced nucleation etc.

Although some parameterisation studies showed a good agreement between

modelling results and the observed data, the modelling results for particle

formation rate were usually lower than the observed data, which suggested

that the particle formation may not be attributable to a sole mechanism.

In order to fully understand the atmospheric aerosol nucleation process, several gaps

in knowledge still need to be addressed, which include:

- Improvement of instrumentation techniques for measuring the chemical

composition of nucleation particles;

- Characterisation of the particle formation processes which occur in the urban

environment, especially in southern hemisphere, as there have been limited

nucleation studies conducted in this area. In addition, higher solar radiation

has been recorded in the southern hemisphere than in the northern hemisphere,

which can enhance photochemical activity;

- Investigation of the mechanisms involved in the particle formation process;

and

- The inclusion of findings from particle formation studies in climate models

for estimating climate change.

58

2.8 References

Ball, S.M., Hanson, D.R. and Eisele, F.L. (1999). Laboratory studies of particle

nucleation: initial results for H2SO4, H2O, and NH3 vapors. Journal of Geophysical Research, 104, 23709 - 23718.

Barsanti, K.C., McMurry, P.H. and Smith, J.N. (2009). The potential contribution of organic salts to new particle growth. Atmospheric Chemistry and Physics, 9, 2949 – 2957.

Bein, K.J., Zhao, Y., Johnston, M.V. and Wexler, A.S. (2007). Identification of sources of atmospheric PM at the Pittsburgh Supersite – Part III: Source characterization. Atmospheric Environment, 41, 3974 – 3992.

Belleudi, V., Faustini, A., Stafoggia, M., Cattani, G., Marconi, A., Perucci, C.A. and Forastiere, F. (2010). Impact of Fine and Ultrafine Particles on Emergency Hospital Admissions for Cardiac and Respiratory Diseases. Epidemiology, 21, 414-423.

Birmili, W., Wiedensohler, A., et al., 2000. Evolution of newly formed aerosol particles in the continental boundary layer: a case study including OH and H2SO4 measurements. Geophysical Research Letters 27 (15), 2205-2208.

Boy, M., Rannik, U., Lehtinen, K.E.J., Tarvainen, V., Hakola, H. and Kulmala, M. (2003). Nucleation events in the continental boundary layer: Long-term statistical analyses of aerosol relevant characteristics. Journal of Geophysical Research 108 (D21), 4667, doi:10.1029/2003JD003838.

Bricard, J., Delattre, P., Madelaine, G. and Pourprix, M. (1976). Detection of ultra-fine particles by means of a continuous flux condensation nuclei counter. In: Liu, B.Y.H. (Ed.), Fine Particles: Aerosol Generation, Measurement, Sampling, and Analysis. Academic Press, New York, 565-580.

Brock, C.A., Hamill, P., Wilson, J.C., Jonsson, H.H. and Chan, K.R. (1995). Particle formation in the upper tropical troposphere: A source of nuclei for the stratospheric aerosol. Science, 270, 1650 – 1653.

Chang L-S., Schwartz, S.E., McGraw, R., and Lewis, E.R. (2009). Sensitivity of aerosol properties to new particle formation mechanism and to primary emissions in a continental-scale chemical transport model. Journal of Geophysical Research 114, D07203, doi:10.1029/2008JD011019.

Charlson, R.J., Schwartz, S.E., Hales, J.M., Cess, R.D., Jr. Coakley, J.A., Hansen, J.E. and Hofmann, D.J. (1992). Climate forcing by anthropogenic aerosols. Science, 255, 423 - 430.

59

Charron, A. and Harrison, R.M. (2003). Primary particle formation from vehicle emissions during exhaust dilution in the roadside atmosphere. Atmospheric Environment 37, 4109-4119.

Charron, A., Birmili, W. and Harrison, R.M. (2007). Factors influencing new particle formation at the rural site, Harwell, United Kingdom. Journal Geophysical Research 112, D14210. Doi:10.1029/2007JD008425.

Cheung, H.C., Wang, T., Baumann, K. and Guo, H. (2005). Influence of regional pollution outflow on the concentrations of fine particulate matter and visibility in the coastal area of southern China. Atmospheric Environment, 39, 6463 – 6474.

Churg, A. and Brauer, M. (2000). Ambient Atmospheric Particles in the Airways of Human Lungs. Ultrastructural Pathology, 24:6, 353 - 361.

Coffman, D.J. and Hegg, D.A. (1995). A preliminary study of the effect of ammonia on particle nucleation in the marine boundary layer. Journal of Geophysical Research, 100, 7147 - 7160.

Covert, D.S., Kapustin, V.N., Quinn, P.K. and Bates, T.S. (1992). New particle formation in the marine boundary layer. Journal of Geophysical Research D, 97, 20581-20589.

Dal Maso, M., Kulmala, M., Riipinen, I., Wagner, R., Hussein, T., Aalto, P.P. and Lehtinen, K.E.J. (2005). Formation and growth of fresh atmospheric aerosols: eight years of aerosol size distribution data from SMEAR II, Hyytiälä, Finland. Boreal Environment Research, 10, 323 – 336.

Dzubay, T.G., Stevens, R.K., Lewis, C.W., Hern, D.H., Courtney, W.J., Tesch, J.W. and Mason, M.A. (1982). Visibility and Aerosol Composition in Houston, Texas. Environmental Science and Technology, 16, 514-525.

Donaldson, K. and Tran, C.L. (2002). Inflammation caused by particles and fibers. Inhalation Toxicology, 14, 5 - 27.

Draxler, R.R. and Hess, G.D. (1998). An overview of the HYSPLIT_4 modeling system for trajectories, dispersion and deposition. Australian Meteorological Magazine 47, 295-308.

Ehn, M., Vuollekoski, H., Petäjä, T., Kerminen, V-M., Vana, M., Aalto, P., de Leeuw, G., Ceburnic, D., Dupuy, R., O’Dowd, C.D. and Kulmala, M. (2010). Growth rates during coastal and marine new particle formation in western Ireland. Journal of Geophysical Research, 115, D18218, doi:10.1029/2010JD014292.

Froyd, K.D., Murphy, S.M., Murphy, D.M., de Gouw, J.A., Eddingsaas, N.C. and Wennberg, P.O. (2009). Contribution of isoprene-derived organosulfates to

60

free tropospheric aerosol mass. Proceedings of the National Academy of Sciences of the United States of America, 107, 21360 – 21365.

Gao, J., Wang, T., Zhou, X., Wu, W. and Wang, W. (2009). Measurement of aerosol number size distributions in the Yangtze River delta in China: Formation and growth of particles under polluted conditions. Atmospheric Environment, 43, 829 – 836.

Hinds, W.C. (1999). Aerosol technology: properties, behaviour, and measurement of airborne particles. New York, Wiley.

Hyvärinen, A.P., Komppula, M., Engler, C., Kivekäs, N., Kerminen, V.M., Dal Maso, M., Viisanen, Y. and Lihavainen, H. (2008). Atmospheric new particle formation at Utö, Baltic Sea 2003-2005. Tellus 60B, 345-352.

Huang, Y., Dickinson, R.E. and Chameides, W.L. (2006). Impact of aerosol indirect effect on surface temperature over East Asia. PNAS, 103, 4371 - 4376.

Hussein, T., Junninen, H., Tunved, P., Kristensson, A., Dal Maso, M., Aalto, P.P., Hansson, H.C., Swietlicki, E. and Kulmala, M. (2009). Time span and spatial scale of regional new particle formation events over Finland and Southern Sweden. Atmos. Chem. Phys. 9, 4699-4716.

Hussein, T., Karppinen, A., Kukkonen, J., Härkönen, J., Aalto, P.P., Hämeri, K., Kerminen, V.-M. and Kulmala, M. (2006). Meteorological dependence of size-fractionated number concentrations of urban aerosol particles. Atmospheric Environment, 40, 1427 – 1440.

IPCC. (2007). Climate Change 2007: The Physical Science Basis. Contribution of Working Group I to the Fourth Assessment Report of the Intergovernmental Panel on Climate Change [Solomon, S., Qin, D., Manning, M., Chen, Z., Marquis, M., Averyt, K.B., Tignor, M. an Miller, H.L. (eds.).]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, 996 pp.

Jaecker-Voirol, A. and Mirabel, P. (1989). Heteromolecular Nucleation in the Sulfuric Acid-Water System. Atmospheric Environment 23, 2053-2057.

Jeong, C.-H., Evans, G.J., McGuire, M.L., Chang, R.Y.-W., Abbatt, J.P.D., Zeromskiene, K., Mozurkewich, M., Li, S.-M. and Leaitch, W.R. (2010). Particle formation and growth at five rural and urban sites. Atmospheric Chemistry and Physics, 10, 7979-7995.

Jeong, C.-H., Hopke, P.K., Chalupa, D. and Utell, M. (2004). Characteristics of Nucleation and Growth Events of Ultrafine Particles Measured in Rochester, NY. Environmental Science and Technology, 38, 1933 – 1940.

Jayne, J.T., Pöschl, U., Chen, Y., Dai, D., Molina, L.T., Worsnop, D.R., Kolb, C.E. and Molina, M.J. (1997). Pressure and Temperature Dependence of the Gas-

61

Phase Reaction of SO3 with H2O and the Heterogeneous Reaction of SO3 with H2O/H2SO4 Surfaces. Journal of Physical Chemistry, 101, 10000 – 10011.

Jung, J.G., Pandis, S.N. and Adams, P.J. (2008). Evaluation of Nucleation Theories in a Sulfur-Rich Environment. Aerosol Science and Technology, 42:7, 495 - 504.

Kanawade, V.P., Jobson, B.T., Guenther, A.B., Erupe, M.E., Pressely, S.N., Tripathi, S.N. and Lee, S.-H. (2011). Isoprene suppression of new particle formation in mixed deciduous forest. Atmospheric Chemistry and Physics Discussions, 11, 11039 – 11075.

Kiendler-Scharr, A., Wildt, J., Dal Maso, M., Hohaus, T., Kleist, E., Mentel, T.F., Tillmann, R., Uerlings, R., Schurr, U. and Wahner, A. (2009). New particle formation in forests inhibited by isoprene emissions. Nature, 461, 381 – 384.

Kousaka, Y., Niida, T., Okuyama, K. and Tanaka, H. (1982). Development of a mixing type condensation nucleus counter. Journal of Aerosol Science 13 (3), 231-240.

Kristensson, A., Dal Maso, M., Swietlicki, E., Hussein, T., Zhou, J., Kerminen, V.M. and Kulmala, M. (2008). Charaterization of new particle formation events at a background site in Southern Sweden: relation to air mass history. Tellus 60B, 330-344.

Kulmala, M. (2003). How Particles Nucleate and Grow. Science, 32, 1000 - 1001.

Kulmala, M. and Kerminen, V-M. (2008). On the formation and growth of atmospheric nanoparticles. Atmospheric Research, 90, 132 - 150.

Kulmala, M., Laaksonen, A., et al. (1998). Parameterizations for sulfuric acid/water nucleation rates. Journal of Geophysical Research – Atmospheres 103 (D7), 8301-8307.

Kulmala, M., Lehtinen, K.E.J. and Laaksonen, A. (2006). Cluster activation theory as an explanation of the linear dependence between formation rate of 3 nm particles and sulphuric acid concentration. Atmospheric Chemistry and Physics, 6, 787-793.

Kulmala, M., Petäjä, T., Mönkkönen, P., Koponen, I.K., Dal Maso, M., Aalto, P.P., Lehitnen, K.E.J. and Kerminen, V.-M. (2005). On the growth of nucleation mode particles: source rates of condensable vapour in polluted and clean environments. Atmospheric Chemistry and Physics, 5, 409-416.

Kulmala, M., Pirjola, L. and Mäkelä, J.M. (2000). Stable Sulphate Clusters as a Source of New Atmospheric Particles. Nature 404, 66-69.

Kulmala, M., Riipinen, L., Sipilä, M., Manninen, H.E., Petäjä, T., Junninen, H., Dal Maso, M., Mordas, G., Mirme, A., Vana, M., Hirsikko, A., Laakso, L.,

62

Harrison, R.M., Hanson, I., Leung, Carl, Lehtinen, K.E.J. and Kerminen, V.-M. (2007). Toward Direct Measurement of Atmospheric Nucleation. Science, 318, 89 – 92.

Kulmala, M., Vehkamäki, H., Petäjä, T., Dal Maso, M., Lauri, A., Kerminen, V-M., Birmili, W. McMurry, P.H. (2004). Formation and growth rates of ultrafine atmospheric particles: a review of observations. Aerosol Science, 35, 143 - 176.

Kurtén, T., Petäjä, T., Smith, J., Ortega, I.K., Sippila, M., Junninen, H., Ehn, M., Vehkamäki, H., Mauldin, L., Worsnop, D.R. and Kulmala, M. (2011). The effect of H2SO4 – amine clustering on chemical ionization mass spectrometry (CIMS) measurements of gas-phase sulfuric acid. Atmospheric Chemistry and Physics, 11, 3007 – 3019.

Laakso, L., Gagne, S., Petäjä, T., Hirsikko, A., Aalto, P.P., Kulmala, M. and Kerminen, V.-M. (2007). Detecting charging state of ultra-fine particles: instrumental development and ambient measurements. Atmospheric chemistry and physics, 7, 1333 –

Laaksonen, A., Kulmala, M., O’Dowd, C.D., Joutsensaari, J., Vaattovaara, P., Mikkonen, S., Lehtinen, K.E.J., Sogacheva, L., Dal Maso, M., Aalto, P., Petäjä, T., Sogachev, A., Yoon, Y.J., Lihavainen, H., Nilsson, D., Facchini, M.C., Cavalli, F., Fuzzi, S., Hoffmann, T., Arnold, F., Hanke, M., Sellegri, K., Umann, B., Junkermann, W., Coe, H., Allan, J.D., Alfarra, M.R., Worsnop, D.R., Riekkola, M.-L., Hyötylainen, T. and Viisanen, Y. (2008). The role of VOC oxidation products in continental new particle formation. Atmospheric Chemistry and Physics, 8, 2657 – 2665.

Larsen, L., Roth, B., Van Dingenen, R. and Raes, F. (1997). Photolytic aerosol formation in SO2-HNO2-H2O-air mixtures, with and without NH3. Journal of Aerosol Science, 28, 719-720.

Lee, Y.-G., Lee, H.-W., Kim, M.-S., Choi, C.Y. and Kim, J. (2008). Characteristics of particle formation events in the coastal region of Korea in 2005. Atmospheric Environment, 42, 3729 – 3739.

Lee, S.H., Reeves, J.M., Wilson, J.C., Hunton, D.E., Viggiano, A.A., Miller, T.M., Ballenthin, J.O. and Lait, L.R. (2003). Particle Formation by Ion Nucleation in the Upper Troposphere and Lower Stratosphere. Science, 301, 1886 – 1889.

Loukonen, V., Kurtén, T., Ortega, I.K., Vehkamäki, H., Padua, A.A.H., Sellegri, K. and Kulmala, M. (2010). Enhancing effect of dimethylamine in sulfuric acid nucleation in the presence of water – a computational study. Atmospheric Chemistry and Physics, 10, 4961 – 4974.

Lovejoy, E.R., Hanson, D.R. and Huey, L.G. (1996). Kinetics and Products of the Gas-Phase Reaction of SO3 with Water. Journal of Physical Chemistry, 100, 19911 – 19916.

63

Luo, G. and Yu, F. (2011). Simulation of particle formation and number concentration over the Eastern United States with the WRF-Chem + APM model. Atmospheric Chemistry and Physics, 11, 11521-11533.

Mahowald, N. (2011), Aerosol indirect effect on biogeochemical cycles and climate. Science, 334, 794-796.

Mäkelä, J.M., Yli-Koivisto, S., Hiltunen, V., Seidl, W., Swietlicki, E., Teinilä, K., Sillanpä, M., Koponen, I.K., Paatero, J., Rosman, K. and Hämeri, K. (2001). Chemical composition of aerosol during particle formation events in boreal forest. Tellus B, 53, 380 – 393.

Marti, J.J., Jefferson, A., Cai, X.P., Richert, C., McMurry, P.H. and Eisels, F. (1997). H2SO4 vapor pressure of sulphuric acid and ammonium sulphate solutions. Journal of Geophysical Research-Atmospheres, 102, 3725 - 3735.

McMurry, P.H. (2000). A review of atmospheric aerosol measurements. Atmospheric Environment 34, 1959-1999.

Modgil, M.S., Kumar, S., Tripathi, S.N. and Lovejoy, E.R. (2005). A Parameterization of Ion-Induced Nucleation of Sulphuric Acid and Water for Atmospheric Conditions. Journal of Geophysical Research 110, D19205.

Morawska, L., Moore, M.R. and Ristovski, Z. (2004). Health Impacts of Ultrafine Particles: Desktop Literature Review and Analysis. Department of the Environment and Heritage, Australian Government.

Napari, I. Noppel, M. Vehkamaki, H. and Kulmala, M. (2002). An improved model for ternary nucleation of sulphuric acid – ammonia – water. Journal of Chemical Physics, 116, 4221 – 4227.

Nel, A. (2005). Air pollution-related illness: Effects of particles. Science, 308, 804 - 806.

Oberdörster, G. and Utell, M.J. (2002). Ultrafine Particle in the Urban Air: To the Respiratory Tract – And Beyond?. Environmental Health Perspectives, 110, A440 - A441.

O’Dowd, C.D., Aalto, P., Hämeri, K., Kulmala, M. and Hoffmann, T. (2002). Atmospheric particle from organic vapours. Science, 416, 497 – 498.

O’Dowd, C.D., Jimenez, J., Bahreini, Roya, Flagan, R.C., Seinfeld, J.H., Hämeri, K., Pirjola, L., Kulmala, M., Jennings, S.G. and Hoffmann, T. (2004). Marine aerosol formation from biogenic iodine emissions. Nature, 417, 632 – 636.

O’Dowd, C.D., Yoon, Y.J., Junkerman, W., Aalto, P., Kulmala, M., Lihavainen, H. and Viisanen, Y. (2007). Airborne measurements of nucleation mode particles I: coastal nucleation and growth rates. Atmospheric Chemistry and Physics, 7, 1491-1501.

64

Paulot, F., Crounse, J.D., Kjaergaard, H.G., Kürten, A., St. Clair, J.M., Seinfeld, J.H. and Wennberg, P.O. (2009). Unexpected Epoxide Formation in the Gas-Phase Photooxidation of Isoprene. Science, 325, 730 – 733.

Pakkanen, T., Kerminen, V., Korhonen, C., Hillamo, R., Aarino, P., Koskentalo, T. and Maenhaut, W. (2001). Urban and Rural Ultrafine (PM0.1) Particles in the Helsinki Area. Atmospheric Environment 35, 4593-4607.

Pekkanen, J., Timonen K.L., et al. (1997). Effects of ultrafine and fine particles in urban air on peak expiratory flow among children with asthmatic symptoms. Environmental Research, 74(1), 24 - 33.

Peters, A., Wichmann, H.E., et al. (1997). Respiratory effects are associated with the number of ultrafine particles. American Journal of Respiratory and Critical Care Medicine, 155(4), 1376 - 1383.

Petäjä, T., Kerminen, V.-M., Dal Maso, M., Junninen, H., Koponen, I.K., Hussein, T., Aalto, P.P., Andronopoulos, S., Robin, D., Hämeri, K., Bartzis, J.G. and Kulmala, M. (2007). Sub-micron atmospheric aerosols in the surroundings of Marseille and Athens: physical characterization and new particle formation. Atmospheric Chemistry and Physics, 7, 2705-2720.

Pope, C.A. III. (2000). Epidemiology of Fine Particulate Air Pollution and Human Health: Biologic Mechanisms and Who’s at Risk?. Environmental Health Perspectives, 108, 713 - 723.

Rhoads, K.P., Phares, D.J., Wexler, A.S. and Johnston, M.V. (2003). Size-resolved ultrafine particle composition analysis, 1. Atlanta. Journal of Geophysical Research 108 (D7), 8418, doi:10.1029/2001JD001211.

Rosenfeld, D. (2000). Suppression of Rain and Snow by Urban and Industrial Air Pollution. Science, 287, 1793 - 1796.

Qian, S., Sakurai, H. and McMurry, P.H. (2007). Characteristics of regional nucleation events in urban East St. Louis. Atmospheric Environment, 41, 4119 – 4127.

Salma, I., Borsόs, T., Weidinger, T., Aalto, P., Hussein, T., Dal Maso, M. and Kulmala, M. (2011). Production, growth and properties of ultrafine atmospheric aerosol particles in an urban environment. Atmospheric Chemistry and Physics, 11, 1339 – 1353.

Scott, W.D. and Cattell, F.C.R. (1979). Vapour pressure of ammonium sulfates. Atmospheric Environment, 13, 307 - 317.

Shi, J.P., Evans, D.E., Khan, A.A., and Harrison, R.M. (2001). Sources and concentration of nanoparticles (<10 nm in diameter) in the urban atmosphere. Atmospheric Environment, 35, 1193-1202.

65

Shi, Q., Sakurai, H., and McMurry, P.H. (2007). Characteristics of regional nucleation events in urban East St. Louis. Atmospheric Environment, 41, 4119-4127.

Sinclair, D. and Hoopes, G.S. (1975). A continuous flow condensation nucleus counter. Aerosol Science 6, 1-7.

Sogavcheva, L., Saukkonen, L., Nilsson, E.D., Dal Maso, M., Schultz, D.M., de Leeuw, G. and Kulmala, M. (2008). New aerosol particle formation in different synoptic situations at Hyytiälä, Southern Finland. Tellus, 60B, 485 – 494.

Spirig, C., Guenther, A., Greenberg, J.P., Calanca, P. and Tarvainen, V. (2004). Tethered balloon measurements of biogenic volatile organic compounds at a boreal forest site. Atmospheric Chemistry and Physics, 4, 215 – 229.

Spracklen, D.V., Carslaw, K.S., Kulmala, M., Kerminen, V-M., Sihto, S-L., Riipinen, I., Merikanto, J., Mann, G.W., Chipperfield, M.P., Wiedensohler, A., Birmili, W. and Lihavainen, H. (2008). Contribution of particle formation to global cloud condensation nuclei concentrations. Geophysical Research Letters, 35, L06808, doi:10.1029/2007GL033038.

Stanier, C.O., Khlystov, A.Y. and Pandis, S.N. (2002). Investigation of nucleation bursts during the Pittsburgh air quality study. Abstracts of the Sixth International Aerosol Conference (pp. 1291-1292).

Stanier, C.O., Khlystov, A.Y. and Pandis, S.N. (2004). Nucleation Events During the Pittsburgh Air Quality Study: Description and Relation to Key Meteorological, Gas Phase, and Aerosol Parameters. Aerosol Science and Technology, 38(S1), 253 – 264.

Stohl, A. and Seibert, P. (1998). Accuracy of trajectories as determined from the conversation of meteorological tracers. Quarterly Journal of the Royal Meteorological Society 124, 1465-1484.

Stohl, A., Wotawa, G., Seibert, P. and Kromp-Kolb, H. (1995). Interpolation errors in wind fields as a function of spatial and temporal resolution and their impact on different types of kinematic trajectories. Journal of Applied Meteorology 34, 2149-2165.

Suess, D.T. and Prather, K.A. (1999). Mass Spectrometry of Aerosols. Chemical reviews 99(10), 3007-3036.

Tunved, P., Hansson, H.-C., Kerminen, V.-M., Strom, J., Dal Maso, M., Lihavainen, H., Viisanen, Y., Aalto, P.P., Komppula, M. and Kulmala, M. (2006). High Natural Aerosol Loadings over Boreal Forests. Science, 312, 261 – 263.

Twomey, S. (1977). The influence of Pollution on the Shortwave Albedo of Clouds. Journal of Atmospheric Science, 34, 1149 - 1152.

66

UNEP. (2002). Global Environment Outlook-3: Past, present and future perspectives. London, United Kingdom, Earthscan Publications Ltd.

Väkevä, M., Hämeri, K., Puhakka, T., Nilsson, E.D., Hohti, H. and Mäkelä, J.M. (2000). Effects of meteorological processes on aerosol particle size distribution in an urban background. Journal of Geophysical Research D, 100, 21057-21067.

Vana, M., Ehn, M., Petäjä, T., Vuollekoski, H., Aalto, P., de Leeuw, G., Cehurnis, D., O’Dowd, C. and Kulmala, M. (2008). Characteristic features of air ions at Mace Head on the west coast of Ireland. Atmospheric Research, 90, 278 – 286.

Vehkamäki, H., Kulmala, M., Napari, I., Lehtinen, K.E.J., Timmrech, C., Noppel, M. and Laaksonen, A. (2002). An improved parameterization for sulphuric acid – water nucleation rates for tropospheric and stratospheric conditions. Journal of Geophysical Research, 107(D22), 4622, doi:10.1029/2002JD002184.

Voisin, D., Smith, J.N., Sakurai, H., McMurry, P.H. and Eisele, F.L. (2003). Thermal Desorption Chemical Ionization Mass Spectrometer for Ultrafine Particle Chemical Composition. Aerosol Science and Technology 37, 471-475.

Wang, L., Khalizov, A.F., Zheng, J., Xu, W., Ma, Y., Lal, V. and Zhang, R. (2010). Atmospheric nanoparticles formed from heterogeneous reactions of organics. Nature Geoscience, 3, 238 – 242.

Watson, J.G. (2002). Critical review-visibility: science and regulation, Journal of The Air and Waste Management Association ,52.

Weber, R.J., Marti, J.J., McMurry, P.H., Eisele, F.L., Tanner, D.J. and Jefferson, A. (1997). Measurements of new particle formation and ultrafine particle growth rates at a clean continental site. Journal of Geophysical Research, 102, 4375 – 4385.

Weber, R.J. and McMurry, P.H. (1996). Fine particle size distributions at the Mauna Loa observatory, Hawaii. Journal of Geophysical Research – Atmospheres, 101 (D9), 14767 – 14775.

Weber, R.J., McMurry, P.H., Eisele, F.L. and Tanner, D.J. (1995). Measurement of expected nucleation precursor species and 3 to 500 nm diameter particles at Mauna Loa Observatory, Hawaii. Journal of Atmospheric Science 52, 2242-2257.

Weber, R.J., McMurry, P.H., et al. (1999). New particle formation in the remote troposphere: a comparison of observation at various sites. Geophysical Research Letters 26 (3), 307-310.

Weber, R.J., McMurry, P.H., Mauldin, L., Tanner, D., Eisele, F., Brechtel, F., Kreidenweis, S., Kok, G., Schilawski, R. and Baumgardner, D. (1998). A study of new particle formation and growth involving biogenic trace gas species

67

measured during ACE-1. Journal of Geophysical Research, 103, 16385 – 16396.

Wehner, B. and Wiedensohler, A. (2002). Long term measurements of submicrometer urban aerosols: Statistical analysis for correlations with meteorological conditions and trace gases. Atmospheric Chemistry and Physics Discussion, 2, 1699-1733.

Williams, P.I., Choularton, T.W., Gallagher, M.W., Coe, H., and Bower, K.N. (1998). Results of monitoring ultra-fine, fine and accumulation mode particles above an urban canyon within the city of Manchester. Journal of Aerosol Science, 29, S593-S594.

Winklmayer, W. (1987). Untersuchung des ultrafeinen aerosols in der urbanen atmosphäre von Wien. Ph.D. thesis, University of Vienna.

Woo, K.S., Chen, D.R., Pui, D.Y.H. and McMurry, P.H. (2001). Measurements of Atlanta aerosol size distributions: observation of ultrafine particle events. Aerosol Science and Technology, 34, 78 – 87.

Yao, X., Choi, M.Y., Lau, N.T., Lau, A.P.S., Chan, C.K. and Fang, M. (2010). Growth and Shrinkage of New Particles in the Atmosphere in Hong Kong. Journal of Aerosol Science and Technology, 44, 639-650.

Yu, F. (2006). Effect of Ammonia on New Particle Formation, A Kinetic H2SO4-H2O-NH3 Nucleation Model Constrained by Laboratory Measurements. Journal of Geophysical Research 106, 4797-4814.

Yu, F. and Turco, R.P. (2001). From Molecular Clusters to Nanoparticles, Role of Ambient Ionization in Tropospheric Aerosol Formation. Journal of Geophysical Research, 106, 4797 – 4814.

Yu, F. and Turco, R.P. (2000). Ultrafine aerosol formation via ion-mediated nucleation. Geophysical Research Letters, 27, 883 – 886.

Yue, D., Hu, M., Wu, Z., Wang, Z., Guo, S., Wehner, B., Nowak, A., Achtert, P., Wiedensohler, A., Jung, J., Kim, Y.J. and Liu, S. (2009). Characteristics of aerosol size distribution and new particle formation in the summer in Beijing. Journal of Geophysical Research, 114, D00G12, doi:10.1029/2008JD010894.

Zhang, K.M. and Wexler, A.S. (2002). Modeling the number distributions of urban and regional aerosols: theoretical foundations. Atmospheric Environment, 36, 1863-1874.

68

CHAPTER 3

Environmental monitoring of airborne nanoparticles

L. Morawska, H. Wang, Z. Ristovski, E.R. Jayaratne, G. Johnson, H.C. Cheung, X.

Ling and C. He

International Laboratory for Air Quality and Health, Queensland University of

Technology

GPO Box 2434, Brisbane QLD 4001, Australia

Published by the Journal of Environmental Monitoring

69

STATEMENT OF JOINT AUTHORSHIP

Title: JEM Spotlight: Environmental monitoring of airborne nanoparticles

Authors: L. Morawska*, H. Wang, Z. Ristovski, E.R. Jayaratne, G. Johnson, H.C.

Cheung, X. Ling, and C. He

Morawska L.

Developed the structure of the paper, complied most of its contents and reviewed the

measurement techniques and common locations for particle concentration and

chemical composition studies.

Wang H.

Reviewed particle elemental composition measurement techniques and compared the

most commonly used measurement instruments.

Ristovski Z.

Reviewed the measurement techniques for particle surface area and elemental

composition and assisted in writing the manuscript.

Jayaratne E.R.

Reviewed the measurement techniques for particle concentration and size

distribution and compared the difference between CPC and SMPS measurements in

different environmental particle number monitoring studies.

Johnson G.

Reviewed the measurement techniques for particle concentration and size

distribution and compared the difference in particle number concentrations for the

different types of CPCs in measuring particle number concentrations.

Cheung H.C.

Assisted with comparing particle number concentration for the different types of

CPCs and assisted in writing the manuscript.

70

Ling X.

Reviewed the measurement techniques for particle concentration and size

distribution and compared the difference CPC and SMPS in different environmental

particle number monitoring studies.

He C.

Reviewed particle elemental composition measurement techniques and compared the

most commonly used measurement instruments

71

CHAPTER 3. ENVIRONMENTAL MONITORING OF AIRBORNE

NANOPARTICES

L. Morawska*, H. Wang, Z. Ristovski, E.R. Jayaratne, G. Johnson, H.C. Cheung, X.

Ling, and C. He

International Laboratory for Air Quality and Health, Queensland University of

Technology, GPO Box 2434, Brisbane QLD 4001, Australia

Abstract

The aim of this work was to review the existing instrumental methods to monitor

airborne nanoparticles in different types of indoor and outdoor environments in order

to detect their presence and to characterise their properties. Firstly the terminology

and definitions used in this field are discussed, which is followed by a review of the

methods to measure particle physical characteristics including number, concentration,

size distribution and surface area. An extensive discussion is provided on the direct

methods for particle elemental composition measurements, as well as on indirect

methods providing information on particle volatility and solubility, and thus in turn

on volatile and semivolatile compounds of which the particle is composed. A brief

summary of broader considerations related to nanoparticle monitoring in different

environments concludes the paper.

72

3.1. Introduction

3.1.1 Sources of Nanoparticles

The beginning of the twenty-first century has witnessed an explosion of interest in

the science and technology of engineered nanoparticles - structures that range in size

from a few, up to about 100 nm. These particles can escape into the environment,

and along with an increasing demand for nanomaterials in terms of both quantity and

quality, there have also been growing concerns regarding their potential impacts on

human health. However, nanomaterial engineering is not the only source of

nanoparticles in ambient air. To the contrary, there are many natural and

anthropogenic processes which can lead to the formation of large quantities of

nanoparticles, and as a result, they are omnipresent in both indoor and outdoor air.

The most significant sources of nanoparticles are combustion processes, both natural

and anthropogenic (the later including vehicle and industrial emissions, biomass

burning and tobacco smoking), but also natural processes, in particular those leading

to secondary particle formation via nucleation. In the urban environment, motor

vehicle combustion is the main source of secondary airborne nanoparticles, which

are not emitted directly by the source but formed in the air from precursors

originating from one or more sources. For example, considerable progress in engine

combustion technologies has led to more complete combustion, whereby the size of

primary black carbon soot particles in vehicle exhaust has decreased substantially

from the micrometer into the nanometre size range. These smaller soot particles have

a reduced surface area for the volatile organic compounds in vehicular exhaust to

condense upon and as a result, instead of condensing onto soot particles, these

73

semivolatile species homogeneously nucleate to form high concentrations of

nanoparticles. Sometimes formation processes also involve photochemistry and in

such cases, light is an essential factor for the process to proceed. Examples of these

processes include the formation of secondary nanoparticles from biogenic emissions

in forest or marine environments, as well as from sources found in the indoor

environment, including modern office equipment (e.g. printer emissions) or

consumer products (e.g. detergents or paints).

3.1.2 Impacts of Nanoparticles

The potential hazards from the inhalation of nanoparticles by humans are very

different to those from the inhalation of larger particles because nanoparticles are not

readily removed from the airstream of inhaled air in the upper parts of the respiratory

tract and therefore, they are inhaled into much deeper regions of the lung1. When in

the small containments of the alveoli region, diffusional deposition of the particles on

the epithelium becomes an efficient physical mechanism, with an alveolar deposition

of about 40% for 50 nm particles compared to about 10% for 700 nm particles2. The

nanoparticles deposited in this oxygen/blood exchange region can penetrate very

quickly and efficiently into the blood stream. If these particles are charged, they pose

an added risk to human health, since inhaled charged particles have a five to six-fold

increased probability of depositing in the lung than uncharged particles of the same

size3.

To date, the toxicity of these nanoparticles, their penetration across the blood-brain

barrier and the pathways leading to nanoparticle-related cardiovascular diseases have

74

been demonstrated. In addition to health effects, these man made nanoparticles have

also been shown to have significant impacts on the environment, more specifically

on atmospheric properties and climate modification, by providing seeds for

atmospheric nucleation processes, as well as changing the optical properties of the

atmosphere.

Considering these potential risks to human health and the environment, it is of

critical importance to not only monitor the presence of nanoparticles in the air, but

also to obtain a good quantitative understanding of their physical and chemical

properties, as well as spatial and temporal trends in indoor and outdoor environments.

The instrumental methods which are now available to monitor the presence of these

particles in the air, as well as characterise their properties, are the main focus of this

review.

3.2. Definition of ‘Nanoparticles’

Many terms have been used in relation to particles in the nanosize range, which

extends from about 1 to over 100 nm, with the most common terms being ultrafine

particles and nanoparticles. Within the field of aerosol science the term “ultrafine

particle” has been used in relation to particles smaller than 100 nm 4, while

nanoparticles are generally referred to as those smaller than 50 nm 4. Both these

terms constitute a somewhat arbitrary classification of particles in terms of their size,

indicating the significant role of this physical characteristic on particle fate in the air.

75

Common to the various definitions of ultrafine particles was “at least one dimension

less than 100 nm” 5, 6. While there has not been universal agreement on these terms,

they have been used to differentiate between particles formed through different

mechanisms. In particular, in the field of vehicle emissions, primary particles, which

are generated during a combustion processes, are generally referred to as ultrafine

particles, while secondary particles (i.e. those that are not emitted from a source but

formed in the air) and those originating from homogenous or heterogeneous

nucleation are referred to as nanoparticles. In contrast to the fields of ambient aerosol

or combustion emissions science, the term “engineered nanoparticle” is the preferred

term when describing nanosize particles originating from various manufacturing or

engineering processes.

A more rigorous definition of these terms has been introduced by the International

Standards Organisation (ISO). In particular, ISO/TC 146/SC 2/WG1 N 320 defines a

nanoparticle as “A particle with a nominal diameter smaller than about 100 nm”, a

nanoaerosol as “An aerosol comprised of or consisting of nanoparticles and

nanostructured particles” and a nanostructured particle as “A particle with

structural features smaller than 100 nm, which may influence its physical, chemical

and/or biological properties”. This means that a nanostructured particle may have a

maximum dimension substantially larger than 100 nm, since a 500 nm diameter

agglomerate of nanoparticles would be considered a nanostructured particle. The

same document defines an ultrafine particle as “A particle sized about 100 nm in

diameter or less” and thus, an ultrafine aerosol would contain a majority of particles

of this diameter or less.

76

It should be mentioned here that the 100 nm cut-off for nanoparticles is not derived

from particle behaviour in the respiratory tract following deposition, and therefore it

is not a health based metric 7. A health based metric will need to consider the fact

that as particles become smaller, surface curvature, the arrangement and percentage

of atoms on the particle surface, and the size dependent quantum effects, such as

quantum confinement, play an increasingly significant role in determining behaviour

7.

When referring to nanoparticle measurements, an unspoken assumption is made that

the instrumental methods used provide information on particles in the specific size

range, which is below 100 nm. This is possible if the instrumental method enables

measurements of particle number size distribution, usually in a broader range, from

which the sections of data encompassing nanoparticles are then extracted. If, rather

than employing instrumentation for particle size distribution measurement, only a

particle counter is used, the outcome of the measurement is the total particle number

concentration in the detection size range of the instrument. This means that the

outcomes of the measurements are not specifically nanoparticle concentrations,

unless specific inlets are used which restrict the range of particles entering the

instrument’s sensing volume. While it is true that, in most typical environments,

particle number concentration is dominated by nanoparticles, it is important to keep

in mind that these are not the same and that there are environments where there are

significant particle modes outside the nanosize range.

77

In the view of the fact that the instruments detecting nanoparticles do not strictly

restrict particle size (as discussed above), when discussing the instrumental

techniques for nanoparticle monitoring, it is generally not essential to use a rigorous

definition of the particles, and therefore, in relation to the review, there is no need to

accept a particular definition. For simplicity, through the paper only the term

“nanoparticle” is used, unless refereeing to published data using other terms.

3.3. Particle Concentration and Size Distribution

3.3.1 Particle number concentration measurements

The particle detection and counting techniques used in environmental monitoring

primarily employ optical detection methods and this is also true of nanoparticle

sampling. However particles smaller than about 50 nm do not interact strongly with

electromagnetic radiation of optical or near optical wavelength, and so are not

detected efficiently by light blocking or scattering. To overcome this range limitation,

environmental nanoparticle number concentration measurements must employ

Condensation Particle Counters (CPC’s) which effectively enlarge the particles to

detectable sizes by condensing a low vapour pressure material onto the original

particles from the gas phase. Condensation Particle Counters (CPCs) typically

contain water or butanol as the condensable species used to grow the particles to a

detectable size, although a small number, such as TSI’s PTrak, use propenol. A list

of water-based and butanol-based CPCs are provided in Table 3-1. The mass

diffusivity of condensable species dictates the design of the instrument and therefore,

instruments using species’ with a lower diffusivity than air, such as the butanol based

78

instruments, rely on the greater diffusivity of air to carry heat away from a warm

vapour enriched aerosol stream as it passes though a cooler condensing tube, thereby

increasing the vapour concentration to the super-saturation levels needed for particle

growth. In contrast, instruments relying on highly diffusive species for particle

growth, such as water vapour, may achieve super-saturation by passing a cooler

aerosol stream through a warm tube coated with the condensable species, so that the

more mobile condensable species carry heat to the cooler aerosol, thereby reaching

super-saturation8. Alternatively, such instruments may rely on the rapid mixing of

two flows, each saturated at different temperatures, to produce super-saturation9, 10.

For an insoluble species, the predicted lower detection size limit is the Kelvin

diameter corresponding to the super-saturation ratio achieved in the aerosol.

However, a species which is soluble in the condensing species will have an

associated equilibrium vapour pressure for the condensing species lower than that for

an inert particle such that the particles may be detectable at smaller sizes. Therefore,

the solubility of the aerosol in the condensable species can affect the lower detection

limit achieved by an instrument.

Several studies have been conducted to compare the performance of butanol and

water based CPCs for different types of aerosol. These have examined the relative

response of the instruments to aerosols for different particle sizes, compositions and

concentrations. Biswas et al.11 conducted a study comparing a butanol based CPC

(BCPC) with a water based instrument (WCPC) for NH4SO4, NH4NO3, glutaric acid,

and adipic acid aerosols generated in the laboratory. The concentration ratio

79

(WCPC/BCPC) recorded by the two instruments ranged from 1.0-1.1 for particles

sizes in the range 10-50 nm, while a slightly higher WCPC/BCPC ratio was obtained

for particle sizes of less than 15 nm. In addition, Hering et al.8 compared an ultrafine

water-based CPC TSI 3785 with a BCPC TSI 3025. Comparable results of WCPC

and BCPC (within ±3% differences) were observed for 80 nm Oleic acid, and 50 nm

NaCl particles. They also showed that the water based instrument responded with

varying sensitivity near the lower size limit depending on the composition of the

aerosol with detection efficiency for the smallest particles being greater when water

soluble species were present in the particles.

Iida et al.12 also compared the performance of the WCPC TSI 3786 and BCPC TSI

3025 under field conditions using different particle sizes. The tests were conducted in

freeway tunnels and ambient environments. For ambient air, the WCPC and BCPC

values were comparable for particles > 5 nm and at 3 nm, the BCPC showed higher

detection efficiency than the WCPC. In contrast, the tunnel data showed that the

WCPC/BCPC ratio was larger than 1.0 for the smaller particles. The authors

suggested that the difference in performance between the WCPC and BCPC, in

ambient and freeway tunnel environments, may be due to differences in the

instrumentation or differences in particle composition. In addition, Biswas et al.11

observed that the WCPC/BCPC ratio varied with particle concentration, with the

WCPC/BCPC ratio being > 1.0 when the particle concentration was below 3 x 104

cm-3 but < 1.0 for higher concentrations (3 x 104 – 8 x 104 cm-3). Similarly, Mordas et

al.13 used 15 and 30 nm silver particles to compare the performance of a WCPC with

respect to an electrometer. The results showed that WCPC/Electrometer ratio was >

80

1.0 for particle concentrations below 3 x103 cm-3, close to 1 for particle

concentrations in the range 3 x 103 - 5 x 104 cm-3 and < 1 for larger concentrations.

The results of the above studies do not support any firm conclusions regarding the

reasons behind the observed differences between the WCPC and BCPC. In Biswas et

al.11, the chemical composition of the aerosol did not affect the relative particle

concentrations recorded by the water and butanol based instruments, however Iida et

al.12 suggested that the chemical properties of the aerosol play a role near the

instruments lower size limit, showing a difference in performance for water and

butanol based instruments with the same 3 nm nominal detection limit when

measuring 3 nm ambient and freeway tunnel particles. Higher concentrations were

recorded by the butanol based instrument for ambient air but lower concentrations

were recorded for vehicle emissions. Consistent differences in response were

observed for a WCPC and BCPC in both Biswas et al.11 and Mordas et al.13, where

the WCPC consistently counted more particles than the BCPC under lower

concentrations, but counted fewer particles at higher concentrations. Further

comparisons of WCPC and BCPC performance are needed to investigate the impact

of differing aerosol types near the lower cut-off sizes of the instrument.

81

Table 3-1. Commercial CPCs and their specifications.

Water-Based CPCs

Manufacturer Model Size range (nm) 1Conc. range (p/cm3) Response time to 95% conc. (sec)

2Flow rate (lpm) Working fluid

From To From To Aerosol flow 5Inlet flow

TSI 3781 6 > 3000 0 5 x 105 < 2 0.12 ± 0.012 0.60 ± 0.12 water

TSI 33782 10 > 3000 0 5 x 104 < 3 0.60 ± 0.06 water

TSI 43785 5 > 3000 0 1 x 107 1.0 ± 0.1 water

TSI 3786 2.5 > 3000 0 1 x 105 < 2 0.3 0.60 ± 0.03 water

Alcohol-Based CPCs

TSI 3 3010 10 > 3000 0.0001 1 x 104 1.0 ± 0.1 N-butyl alcohol

TSI 2, 3 3022A 7 0 9.99 x 106 < 13 0.3 ± 0.015 1.5 ± 0.15 (H) 0.3 ± 0.015 (L)

N-butyl alcohol

TSI 2, 3 3025A 3 0 9.99 x 104 < 1 (H), < 5 (L) 1.5 ± 0.15 (H) 0.3 ± 0.03 (L)

N-butyl alcohol

TSI 1, 3 3760A 11 > 3000 0.0001 1 x 104 < 3 (decreasing conc.) < 1.5 (increasing conc.)

1.5 ± 1.5 N-butyl alcohol

TSI 1, 3 3762 11 > 3000 0.0001 1 x 104 < 1.5 (decreasing conc.) < 1 (increasing conc.)

3.0 ± 0.3 N-butyl alcohol

TSI 3771 10 > 3000 0 1 x 104 3 1.0 ± 0.05 N-butyl alcohol

TSI 3772 10 > 3000 0 1 x 104 3 1.0 ± 0.05 N-butyl alcohol

TSI 23775 4 > 3000 0 1 x 107 4 (H), 5 (L) 0.3 ± 0.015 1.5 (H), 0.3 (L) N-butyl alcohol

TSI 2 3776 2.5 > 3000 0 3 x 105 < 0.8 (H), < 5.0 (L) 0.05 (with 0.25 lpm sheath flow)

1.5 (H), 0.3 (L) N-butyl alcohol

TSI 3790 23 > 3000 0 1 x 104 < 5 1.0 N-butyl alcohol

TSI 3007 10 >1000 0 1 x 105 < 9 0.1 0.7 isopropyl alcohol

GRIMM 2 5.401 4.5 > 3000 0 1 x 107 3.9 (at 90%) 0.3 1.5 (H), 0.3 (L) N-butyl alcohol

GRIMM 2 5.403 4.5 > 3000 0 1 x 107 3.9 (at 90%) 0.3 (with 3 lpm sheath flow)

1.5 (H), 0.3 (L) N-butyl alcohol

GRIMM 5.412 23 0 1.2 x 104 ≤ 4 0.6 N-butyl alcohol

Kanomax 13885 10 0 ~105 4.2 ± 0.4 Propylene Glyol

82

Notes: 1 External vacuum pump used in TSI 3010, 3760, 3762, 3771, 3772; Kanomax 3885.

2 For some CPCs, high and low flow modes available which stated with H or L in the bracket.

3 Discontinuous models

4 TSI 3785 also marketed as Quant Technologies 400

5 Inlet flow equals aerosol flow plus transport flow. 5

83

3.3.2 Particle size distribution

3.3.2.1 Differential Mobility Particle Sizer /Scanning Mobility Particle Sizer

Particle size distribution in the submicrometer size range is generally measured with

a differential mobility particle sizer (DMPS) or a scanning mobility particle sizer

(SMPS). Component systems feature an electrostatic classifier with a differential

mobility analyser (DMA) that selects the size bins and a condensation particle

counter (CPC) that counts the number of particles in each bin. The sample first

passes through a bipolar ion neutralizer which uses a radioactive source such as Kr-

85 or Po-210 to ionize the particles into positive and negative ions and brings their

charge level to a Fuch’s equilibrium charge distribution. The charged and neutral

aerosols next enter the DMA where they are deflected by an electric field. Only

particles within a narrow range of electrical mobility and, therefore size, are allowed

to pass through an open slit into the CPC. In the DMPS, the voltage giving the

electric field is increased in discrete steps to cover the entire particle size distribution

to be measured. In the SMPS, the voltage is continuously ramped over a user-

selected period of time. Associated software controls instrument operation and

calculates the number-size distributions, taking into account multiple charge effects

and detection efficiency. The versatility afforded by the individual components

enables the selection of a system that best fits the sizing requirements. For example,

the TSI SMPS’s are capable of measuring aerosols in a wide number concentration

ranging from 1-108 particles cm-3 in varying size windows between 3 nm and 1.0 μm.

Data may be acquired in up to 167 size channels. The lower size limit cannot be

84

smaller than the lower size detection level of the CPC which can range from 3 to 23

nm (Table 1) while the width of the size window is controlled by the ratio of the

sheath to the sample flow rate which is usually kept around 10:1 L min-1.

3.3.2.2 Fast Mobility Particle Sizer

The Model 3091 Fast Mobility Particle SizerTM Spectrometer or FMPS marketed by

TSI Inc, uses similar technology to the SMPS, namely electrostatic classification

within a laminar flow and dynamic scanning, but rather than using a single CPC to

detect particles as they exit from a fixed location at the base of the inner electrode,

and scanning the mobility diameter across the entire range as occurs in the SMPS,

the FMPS detects particles in situ as they reach the outer electrode and does so

simultaneously at 22 different locations, greatly reducing the time required to

examine the entire mobility range 14. The FMPS achieves this by locating particle

counting electrometers at multiple locations along the column. In this way the

instrument is able to acquire a full size distribution in as little as 1 second. The

instrument also differs from the SMPS/DMPS in that it uses an electrical unipolar ion

generator to produce a predictable positive charge distribution on the aerosol instead

of the predictable bipolar charge distribution applied in the SMPS/DMPS, which use

radioactive source based neutralizers.

The FMPS sensitivity is generally poorer than that of the SMPS and DMPS systems

because electrometer signal noise results in a minimum reading equivalent to an

apparent particle number concentration of 100 cm-3 at the smallest detectable particle

85

sizes, when a 1 second averaging time is used, however this improves to around 50

cm-3 for a ten second average. This is the worst case scenario and the sensitivity is

better for larger particle sizes. The maximum concentration measurable also depends

on diameter but can be as high as 107 cm-3 at the smallest sizes. The FMPS classifies

particles into a total of 32 channels at a size resolution of 16 channels per decade. A

total of 20 channels lie in the nanoparticle size range. The sensitivity of the

instrument is poorest at the 6nm lower size limit, being 100 cm-3 when a 1 s

averaging time is used and 50 cm-3 when a 10 s interval is used, however this

improves to 0.9 cm-3 and 0.2 cm-3 at 500 nm when a 1 s and 10 s intervals are used,

respectively.

3.3.2.3 Electrical Low Pressure Impactor (ELPI)

The Electrical Low Pressure Impactor (ELPI, Dekati Ltd., Tampere, Finland) 15

performs both real time size distribution measurement and the simultaneous

collection of the size classified material for chemical analysis. Like the FMPS, the

ELPI uses a unipolar corona charger to achieve a known aerosol charge distribution

and detects particle concentration by using multiple sensitive electrometers, but uses

inertial classification rather than electrical mobility to assess particle size. After

being charged, the aerosol particles are classified according to their aerodynamic

diameter onto a series of impaction stages and a final filter, while electrometers

measure the rate at which the particles deliver charge to each stage. The ELPI

classifies particles into a total of 12 channels at a size resolution of 4.3 channels per

86

decade. A total of 3 channels lie in the nanoparticle size range. Although it is unable

to measure particles in the coarse size range accurately, due to a low total charge

carried by these particles, the ELPI can measure the number, mass concentration and

size distribution of nanoparticles down to 30 nm, with a good time resolution (of a

few seconds) very effectively 16. The sensitivity is again limited at the lower size

limit by lower charge carried by smaller particles being 83 cm-3 for the 14 nm size

classification for the 30 Lpm flowrate of the outdoor ELPI.

3.3.3 Particle surface area, surface topography and morphology

Total particle surface area is a parameter of interest from a health effects point of

view since a good correlation has been found between particle surface area and

certain health effects17, 18. However, to date, there is no well established technique to

measure it. In this section we review the existing techniques and discuss the different

definitions of surface area that each technique measures.

In many cases, airborne particles are not smooth spheres and more commonly they

are agglomerates, as is the case for diesel soot particles, and therefore, it is not clear

what actually presents the particle surface. The definition of particle surface that is

measured also depends on the measurement method used. In principle, all

instruments are based on the attachment of molecules or atoms to the surface of

aerosol particles. Some of the definitions coming from different measurement

methods are: the BET surface, the active surface and the equivalent surface.

87

The BET method, named after Brunauer, Emmett and Teller, measures the number of

adsorption sites available on a particle. This is done by measuring the amount of a

gas (most commonly N2) that can be absorbed on the particle surface and in turn

using this value as a measure of surface area19. Although this method has shown to

correlate well with pulmonary inflammatory responses17, unfortunately it is very time

consuming and often difficult to apply.

The active surface area is based on the integral collision cross section of the particles.

The active surface of a single particle is approximately proportional to the inverse of

the particle mobility. The total active surface can be calculated by integrating the

measured mobility distribution or it can be measured directly via the adsorption of

labelled species onto the particles20. If the labelled species are radioactive atoms,

then the method is called the ‘epiphaniometer’, and if they are ions, then the method

is called ‘diffusion charger’.

The epiphaniometer was developed at the Paul Scherrer Institute21. The EPI is based

on the attachment of lead atoms (211Pb), produced by the radioactive decay of a long-

lived 227Ac source. The number of attached 211Pb lead atoms is then determined by

counting the ß-decay events of its progeny, 211Bi. It is a very sensitive, but slow

instrument, with an integration time of 30 min and a detection limit of 0.003 µm2 cm-

3. Therefore, it cannot be used to measure transient emissions such as those from

motor vehicles.

In the diffusion charging sensor (referred to as DC), positive ions from a corona

discharge diffuse onto the particles. After passing through the charging section of the

88

instrument, the aerosol passes by an ion trap electrode, to which a low voltage is

applied, and this removes the remaining ions. The charged particles are then

precipitated onto an electrically insulated filter. The filter current yields the ion

attachment rate (number of ions attached in unit time), which is proportional to the

active surface of the particle ensemble as shown by Konstandopoulos et al.22. The

DC yields the same information as the epiphaniometer, but it is much faster and

simpler to use, such that the response time is short enough to allow transient

measurements. On the other hand, the DC is also significantly less sensitive, with a

lower detection limit. This is because potential high particle charges have to be

avoided, in order to avoid artifacts which may result from the repelling Coulomb

force. The detection limit of around 1 µm2/cm3 is sufficient for direct measurement

of emissions from sources, such as vehicles and ambient air measurement in urban

areas.

As the most common method of characterizing particle size is the measurement of

number size distribution by mobility analysis, the surface area can be calculated from

the measured number size distributions. In this case, the calculated surface area

yields a mobility equivalent surface (i.e. the surface of spherical particles having the

same size distribution as the measured ones). This surface area is easy to determine

but in many cases has no physical meaning.

Particle morphology and surface topography can be examined directly using

Scanning Electron Microscopy (SEM). SEM uses a high energy electron beam to

scan the surface of a sample resulting in the production of secondary electrons as

89

well as characteristic X-Rays, as atoms in the sample surface are ionised. Back

scattered beam electrons can also be detected. Each of these signals can be

independently examined. When secondary electron imaging is used the surface of

particle and provides depth of field as well as resolution down to a few nanometres.

Hence SEM is very useful for examining the surface structure of nanoparticles and

the nanostructure of larger particles. Alternatively the back scattered electron signal

can be used to produce an image in which intensity is proportional to the atomic

numbers of element within the particle surface. Characteristic X-Rays representing

specific elements, are also produced when the electron beam removes inner shell

electrons from the atoms in the particle surface and these can be used to probe the

elemental composition within regions on the particle surface.

TEM can be used to image particles with much greater resolution than SEM,

achieving lateral resolutions of fractions of a nanometre, however the technique is

used for examining particle structure rather than surface topography. The sample is

illuminated with a beam of high energy electrons and an image formed on film from

the transmitted electrons. As with SEM characteristic X-Rays can be used to examine

elemental composition.

3.3.4 Particle Light Scattering and Absorption

Nanoparticles play an important role in the Earths radiation balance, directly through

their ability to absorb and scatter light, and indirectly when they act as cloud

condensation nuclei (CCN) and thereby alter cloud albedo and lifetime.

90

Measurement of the potential for particles of various sizes to act as CCN is usually

achieved through the use of CCN counters. These instruments determine the CCN

content of bulk aerosol without discrimination in terms of the initial particle size and

hence are not nanoparticle monitoring instruments as such.

Light scattering relevant to the direct effect is commonly measured non-size-

selectively using the integrating light scattering nephelometer. This is also a bulk

method which does not distinguish between scattering by nanoparticles and by larger

particles or by gases. Light scattering and absorption by bulk aerosol is typically

dominated by those particles larger than the wavelength of the light so that

nanoparticles typically make only a minor contribution to the overall signal. Hence

measurement of the direct effect for nanoparticles is not usually sought.

Light absorption can be measured by collecting nanoparticles on a filter as is the case

in black carbon measurements, and a number of instrument exist for this purpose,

however light scattering and absorption by nanoparticles accumulated on a filter may

not be representative of their behaviour when suspended in air23. Instruments which

assess optical properties for suspended particles typically become insensitive when

particles are smaller than the wavelength of the radiation in question, making them

ineffective at diameters of less than a few hundred nanometres.

91

3.3.5 Comparison between CPC and SMPS number concentration monitoring in the

environment

Over the last twenty years, the CPC and SMPS have emerged as the most popular

instruments for monitoring particle number concentration in the environment. While

the CPC is used to measure total particle number concentration, the SMPS provides

number-size distributions within given size ranges. Most of the published reports on

environmental particle number and number-size distribution have used instruments

manufactured by TSI Incorporated. In interpreting the results from these instruments,

a parameter that is of crucial importance is the lower end of the measurement size

range. This lower size limit is determined both by instrumental factors and operator

decisions. For example, the lower size limit of the CPC is determined by the

capability of the instrument and ranges from 3-23 nm (Table 1). However, the lower

end of the detection window of the SMPS is set to a value above this, in the range

from 10 – 30 nm, which is up to 10 nm higher than the achievable lower limit. This

is done to achieve a compromise as to the overall size of the window. The loss of a

few nanometres at the lower end enables a significant extension of the window at the

upper end. This is generally a desirable option when monitoring submicrometer

particles in the environment. However, when a study specifically focuses on

nanoparticles and formation of secondary particles through nucleation, it is

preferable to use a narrower window with a lower size cut-off. The TSI Model 3936

SMPS, for example, can be used with a 3085 nano differential mobility analyser

(DMA) and a Model 3025 CPC to monitor particles in the size range 3-150 nm.

Where a number concentration within a specific size range is required (for example

92

below 100 nm), it is sometimes possible to extract this information from a

measurement of particle number size distribution in a broader size range.

Figure 3-1. Particle number-size scans obtained by the authors with two different

SMPS’s in a rural outdoor environment. While the SMPS 3936 is able to detect the

full nucleation mode centred at 18 nm, the SMPS 3934 is not able to do this due to

the higher lower size cut-off. The lower and upper size cut-offs of the two

instruments are shown as pairs of broken lines (3936) and solid lines (3934).

Figure 3-1 shows the particle number size distributions obtained simultaneously by

the authors in a rural outdoor environment using two different SMPS systems. The

3934 operated on a size window of 15-698 nm while the 3936, using a nano DMA,

0.E+00

2.E+03

4.E+03

6.E+03

8.E+03

1.E+04

1 10 100 1000

Particle Size D (nm)

dlog

N/lo

gD

SMPS 3936SMPS 3934

4 nm 15 nm 160 nm 698 nm

93

was set at 3-160 nm. Note that, while it counts particles up to a much larger size, the

3934 loses its detection efficiency at the lower end of the window and is unable to

see the complete nucleation mode centred at about 18 nm, which is clearly detected

by the SMPS 3936 with the nano DMA. We believe that this nucleation mode

occurred due to the photo-oxidation of biogenic precursors as it lasted only for a few

hours close to mid-day and was shown to consist of volatile substance that

evaporated completely when passed through a thermodenuder heated to 200°C. The

total ambient particle number concentration was about 3800 cm-3. The count median

diameters reported were 59.5 nm and 24.9 nm, for the SMPS 3934 and 3936,

respectively. This large difference is a direct result of the different size detecting

ranges employed and illustrates the importance of caution when comparing results

from various instruments.

Where only a CPC is used, the outcome of the measurement is the total particle

number concentration in the detection size range of the instrument. In interpreting

this result, there are two important aspects to be considered. Firstly, the result does

not reflect the nanoparticle concentrations, unless specific inlets are employed to

restrict the range of particles entering the sensing volume of the instrument. Secondly,

as stated earlier, the detection range of the CPC often extends to lower sizes than the

window set by the SMPS. Therefore, CPC’s are able to detect particles in the earlier

stages of nucleation, and the presence of the nucleation mode which is below the size

detection limit set by the SMPS. As a result, CPC’s would detect more particles than

the SMPS, the difference being significant in environments where a nucleation mode

is frequently present.

94

This fact is often overlooked when comparing particle number concentrations,

specifically nanoparticles, reported in different studies using instruments with

different size range windows. In order to assess the impact which these differences

have on the reported particle number concentrations, data from 60 studies reporting

total particle number concentrations in a wide range of environments was compiled

and the results grouped according to the instrumentation used, that is CPC or SMPS.

The CPC and SMPS results were extracted from the following papers: Aalto et al. 24,

Harrison et al. 25, Kittelson et al. 26 and Shi et al. 27 who used both the CPC and

SMPS; Vakeva et al. 28, Zhu et al. 29, Imhof et al. 30, Paatero et al. 31 and Westerdahl

et al. 32, Schnieder et al. 33, Lechowicz et al. 34, Asmi et al. 35, Bergmann et al. 36 and

Weimer et al. 37 who used only the CPC and McMurry and Woo 38, Tuch et al. 39,

Morawska et al. 40, Hitchins et al. 41, Junker et al. 42, Jamriska and Morawska 43, Pitz

et al. 44, Ruuskanen et al. 45, Cheng and Tanner 46, Molnar et al. 47, Morawska et al. 48,

Thomas and Morawska 49, Wehner et al. 50, Zhu et al. 51, Zhu and Hinds 52, Ketzel et

al. 53, Longley et al. 54, Tunved et al. 55, Wehner and Wiedensohler 56, Gidhagen et al.

57, Gramotnev and Ristovski 58, Gramotnev et al. 59, Hussein et al. 60, Jamriska et al.

61, Janhall et al. 62, Jeong et al. 63, Ketzel et al. 64, Morawska et al. 65, Stanier et al. 66,

Gidhagen et al. 67, Holmes et al. 68, Imhof et al. 69, Rodriguez et al. 70, Janhall et al. 71,

Virtanen et al. 72, Wahlin et al. 73, Woo et al. 74, Abu-Allaban et al. 75, Laakso et al. 76,

Hussein et al. 77, Mejia et al. 78, Pey et al. 79, Barone and Zhu 80, Yue et al. 81, Wu et

al. 82, Westerdhal et al. 83, Buonanno et al. 84 and Minoura et al. 85 who used only the

SMPS. Other studies, such as Hameri et al. 86 and Kaur et al. 87, which measured

particle concentration without using a CPC or SMPS (e.g. P-trak etc.) were not

95

included in the analysis, nor were the four tunnel studies, Abu-Allaban et al. 75,

Gpuriou et al. 88, Jamriska et al. 61 and Imhof et al. 69.

The mean concentrations measured by the CPC's and SMPS's were 58.7×103/cm3 and

50×103/cm3, respectively, and the median concentrations were 28.3×103/cm3 and

21.4×103/cm3, respectively. In other words, the mean and the median CPC

measurements were 17% and 32% higher than the SMPS measurements, respectively.

The difference in median concentrations was analysed using a Students t-test and

found to be statistically significant at a confidence level of over 99%. These

differences are expected to be larger for environments where a nucleation mode is

present and smaller where aged aerosol dominates. It is, therefore, important to take

these differences into consideration when attempting to establish quantitative

understanding of variation in particle concentrations between different environments,

which is of significance for human exposure and epidemiological studies.

Both CPC and SMPS have been also employed in engineered airborne particle

characterisation. There have been a handful of studies conducted in the work

environment of several nanotechnology facilities where engineered nanoparticles are

formed, which included measurements of size distribution, concentration, mass, and

physico-chemical characterisation of the particles 89-93. Most of these studies, which

were conducted in the ambient air of the facility (not just inside the particle reactor),

showed an increase in particle concentration, which was usually higher closer to the

production site. This elevation in concentration - even if it is not large - is of

significant concern, due to unknown health effects of these particles. However, as

96

concluded by the Royal Academy of Engineering 94, experience with occupational

monitoring during nanoparticle production is still in its infancy, with many

uncertainties regarding an appropriate metric and empirical method.

3.4. Measurements of Nanoparticle Elemental Composition

Chemical characterization of airborne nanoparticles is important, since, in addition to

size particle chemical properties further influence the impacts of the particles on

human health and global climate 95-97. Several recent reviews have discussed the

methods used for chemical characterization and their application to atmospheric

chemistry 98-104, however these methods often are difficult to implement in the

nanometre size range due to the small mass of the particles involved. The purpose of

this section is not to describe in detail the operation principles of the different

available techniques capable of measuring nanoparticle composition, but to compare

and discuss their strengths and limitations.

3.4.1 Offline Measurements

The most direct method for the chemical analysis of aerosols is to pass a sample flow

through a filter and analyse the collected material in the laboratory using well-

established analytical procedures. A common method of analysing the sample

collected on the filter is to dissolve the material in water or another solvent, and

analyse the solution using regular analytical methods, including gas

chromatography–mass spectrometry (GC-MS), ion chromatography (IC) and/or

97

proton nuclear magnetic resonance (HNMR). The sample can also be analysed in situ,

without transferring it to water or other solvents, by using X-ray fluorescence,

scanning electron microscopy (SEM), transmission electron microscopy (TEM)

and/or secondary-ion mass spectrometry (SIMS). Chow 105 discussed suitable filter

materials for various analytical methods, species sampling artefacts and analytical

techniques that can be used for various species.

Although filters are inexpensive and easy to use, and they have been widely used in

traditional atmospheric chemistry in the past, unfortunately, the very small mass of

nanoparticles has posed a new challenge when using this sampling technology to

determine their size-dependent chemical composition. As a result, when collecting

size-classified nanoparticles for chemical analysis, impactors are the most frequently

used devices 106-109. For example, the NanoMOUDI (Nano-micro-orifice Uniform

Deposit Impactor, MSP Corp, Shoreview, MN, USA) can collect particles classified

by aerodynamic diameter down to 10 nm 110. The second generation of MOUDI

includes 10-13 stages covering the size range 0.01 – 18 µm and up to 6000 micro-

orifice nozzles as small as 50 µm diameter are used in the 30 L/min Model 122 to

reduce pressure drop, jet velocity, particle bounce and re-entrainment 111. To achieve

enough mass for laboratory analysis, the NanoMOUDI ambient sampling process

generally need to continue for days. In order to shorten this sampling time, Geller et

al. 112 utilized a USC Ultrafine Concentrator to concentrate nanoparticles by a factor

of 20-22 before the NanoMOUDI sampling. This system was employed to collect

enough nanoparticles in 3 consecutive 3 h time intervals (i.e. morning, midday, and

98

afternoon) for examining diurnal variations of size-fractionated ultrafine particle

chemistry in the Los Angeles Basin.

Advantages of impactors include their relatively high sampling rate, simplicity of

operation and compatibility of sampling substrates with commonly used analytical

methods. Bounce is a major disadvantage of using impactors to collect nanoparticles,

as they rely on the particles sticking to the substrates when they impact, and after

impaction, the dislodged particles become re-entrained in the air flow, such that they

are then able to deposit on subsequent stages.

In addition to impaction, electrostatic precipitation is another effective sampling

method for collecting nanoparticles. For example, the Nanometer Aerosol Sampler,

consisting of a grounded cylindrical sampling chamber with an electrode at the

bottom of the chamber, can be used to sample aerosols that have been conditioned

and positively charged, like those from the output of a DMA, onto sample substrates

for further analysis, such as SEM/TEM and GC-MS 113-115. The instrument is

designed to collect 2-100 nm particles and the size of the captured particles can be

controlled using two electrode sizes, in order to get a uniform deposition size that is

optimal for the particular analysis system. The electrostatic precipitation technique

can also be used as a sampler in some online analysis instruments (discussed below).

One limitation of these off-line methods is the poor time resolution associated with

collecting a large enough sample for bulk analysis. Gas-phase partitioning after

collection and contamination during handling are other issues that also need to be

considered.

99 

3.4.2 Online measurements

The on-line measurement of particles avoids potentially significant sampling

artefacts caused by the evaporation, adsorption and chemical reaction of particulate

species which can result when aerosols remain on substrates for long periods of time

100, and also from gas-particle partitioning after collection 116.

3.4.2.1 On-line bulk sampling and analysis methods

A more general approach has also been used for the automated and in-situ sampling

of aerosols by absorption into water, using the Particle Into Liquid Sampler (PILS)

117, the Steam Jet Aerosol Collector (SJAC) 118 or the Semi-Continuous Measurement

System for Ionic Species (SCMSIS) 119. Particles grown by supersaturated water are

impacted on a surface, over which water continuously flows and an ion

chromatographer (IC), with a suitable detector, is employed to periodically analyse

the collected water sample to give the quantitative soluble ion composition of the

aerosol with a high time resolution (minutes). NH4+, K+, Na+, Mg2+, Cl-, NO3

-, SO42-,

as well as some short-chain organic acids, can be measured in this way and the

collection efficiency for particle diameters between 0.03 - 10 m is greater than 97%

117. Recent improvements in the design have enabled measurements down to 10 nm

diameter 120.

Thermal desorption aerosol GC-MS/FID (TGA) is another novel continuous

measurement method 121-123. Nanoparticles are collected in bulk, by impaction onto a

100

cooled surface, which is then isolated after collection and heated, vaporizing the

components either for total mass concentration by GC-FID (flame ionization detector)

or for compound identification by GC-MS. This method can be used to measure the

organic fraction of aerosols with a one hour time resolution. Overall, these automated

instruments are less labour-intensive and more suitable for continuous monitoring.

The inherent limitation of the online bulk sampling instruments, however, is that they

cannot provide any information regarding size-dependent distribution of chemical

compounds because size-resolved data are not available 116.

3.4.2.2 Real-time mass spectrometry of aerosols

In order to improve the source apportionment of atmospheric particles and determine

causality between potential toxic particles and increased human morbidity, a

comprehensive approach that simultaneously characterizes single particles for

aerodynamic diameter and chemical composition is needed in place of bulk chemical

analysis 124. Due to its high sensitivity, mass spectrometers are often used when

studying single particles in the atmosphere 125. Although it has been recently

demonstrated that laser induced breakdown spectroscopy, coupled with an

aerodynamic lens, is able to determine the elemental composition of metal

nanoparticles 126, real-time aerosol mass spectrometry (RTAMS) is the most widely

used technique capable of simultaneously sizing and speciating single aerosol

particles in situ. Since Davis 127 began the online chemical analysis of single particles

using a Surface/Thermal Ionization Mass Spectrometry over 35 years ago, scores of

101

RTAMS with different approaches have been developed and applied to various

laboratory and field measurements. Two reviews provided a chronological survey of

the RTAMSs and discussed their design principles and operation 128, 129. However,

two main challenges emerge when these methods are extended to nanoparticles. One

is how to achieve efficient sampling of particles from air under ambient atmospheric

pressure into the high-vacuum environment of a mass spectrometer source, the other

is the small particle mass 130. Due to these limitations some existing RTAMS cannot

be used for nanoparticle measurement. For example, the Aerosol Composition Mass

Spectrometer (ACMS), developed by Schreiner, can only detect particles with a

diameter greater than 300 nm 131. Table 2 lists the name, measurement technology,

detection limit and selected publications relating to those RTAMS with the capacity

to measure nanoparticles. This list is not exhaustive, but represents a wide spectrum

of the novel techniques used in the nanoparticles field. All of the abbreviations for

the instruments and methods mentioned in this paper (including those listed in Table

3-2) can be found in Table 3-3.

102

Table 3-2. Instruments capable of measuring nanoparticles composition using mass spectrometry

a SI, surface ionization; EI, electron impact ionization; CI, chemical ionization;

b NA, not available; PA, Polydisperse aerodynamic; POA, Polydisperse opto-aerodynamic;

Instrument Vaporization & Ionisationa

Mass spectrometer Sizingb Optimum size range Species Detection limit Selected literature

AMS Thermal-EI Quadruple PA 40-1000nm SO4, NH4, NO3 & Organics (nonrefractory)

~ 2 µg/m3 132, 133

ATOFMS Single laser Bipolar RETOF POA 30-3000 nm SO4, NH4, NO3 & Organics < 1 µg/m3 134, 135

IT-AMS Thermal-EI 3D Quadruple Ion Trap

PA 60 – 600 nm Nitrate, sulphate 0.16 µg/m3 for nitrate

0.65 µg/m3 for sulphate

136

NAMS Single laser RETOF NA 7-25 nm Atomic composition 105 particles/cm3 130, 137, 138

Particle Blaster Single laser RETOF NA 17 – 900 nm Atomic composition NA 139

PIAMS Single laser RETOF POA < 300 nm Organic 50-500ng/m3 140

RSMS III Single laser Bipolar Linear TOF MA 50 – 750 nm Nitrate, sulphate, carbon, metal

NA 141-143

SPLAT I/II Dual laser RETOF POA 50-3500/125-600 nm Atomic composition NA 144, 145

SI-PBMS Thermal-SI Quadruple NA 14 - 1000 nm Alkali metal 103 atoms 146

TD-CIMS Thermal-CI Triple quadruple NA 6 - 20 nm Molecular composition 50 pg/m3 147-151

TDPBMS Programmable thermal-EI

Quadruple NA 20-500 nm Molecular composition ~ 0.1 µg/m3 152, 153

103

Table 3-3. List of Instrumental and Method Abbreviations

ACMS Aerosol composition mass spectrometer

ADL Aerodynamic lens

AMS Aerosol mass spectrometry

ATOFMS Aerosol time-of-flight mass spectrometry

B/WCPC Butanol / water based condensation particle counter

CPC Condensation particle counter

DMA Differential mobility analyser

DMPS Differential mobility particle sizer

ELPI Electrical low pressure impactor

FMPS Fast mobility particle sizer

GC-FID Gas chromatography-flame ionization detector

GC-MS Gas chromatography–mass spectrometry

HNMR Proton nuclear magnetic resonance

HR-TOF High resolution TOF mass spectrometer

IC Ion chromatography

ICP-MS Inductively coupled plasma mass spectrometer

INAA Instrumental neutron activation analysis

LDI Laser desorption/ionization

LIF Laser-induced fluorescence

MALDI Matrix-assisted laser desorption ionization

NAMS Nanoaerosol mass spectrometer

NanoMOUDI Nano-micro-orifice uniform deposit impactor

PERCI Photoelectron resonance capture ionization

PIAMS Photoionisation aerosol mass spectrometry

REMPI Resonance enhanced multiphoton ionization

RETOF Reflecting time of flight

104

RSMS Rapid single-particle mass spectrometer

RTAMS Real-time aerosol mass spectrometry

SEM Scanning electron microscopy

SIMS Secondary-ion mass spectrometry

SI-PBMS Surface ionization particle beam mass spectrometer

SMPS Scanning mobility particle sizer

SPLAT-MS Single particle laser ablation time-of-flight MS

TDCIMS Thermal desorption chemical ionization MS

TDLIBS Laser-induced breakdown spectroscopy

TDPBMS Thermal desorption particle beam MS

TEM Transmission electron microscopy

TGA Thermal desorption aerosol GC-MS/FID

TD-GC×GC-TOF/MS Thermal desorption coupled to comprehensive gas chromatography-time of flight mass spectrometry

VUV Vacuum ultraviolet

There are currently two transportable aerosol mass spectrometers that are

commercially available and able to be used for field measurements: the TSI model

3800 Aerosol Time of Flight Mass Spectrometer (ATOFMS) 135, which is based on

the desorption/ionization of single particles by the 260 nm light from NdYAG laser,

and the Aerodyne Research Inc. Aerosol Mass Spectrometry (AMS) 132, 133, which

focuses on a broader range of particle sizes using the continuous electron impact (EI )

ionization method. ATOFMS uses a reflecting time of flight (RETOF) spectrometer

for collecting both positive and negative ions, and delivers quantitative size and

largely qualitative composition information on individual particles. AMS can give

quantitative data on both the size and composition of the entire aerosol ensemble, but

gives only limited data on specific particles and cannot measure refractory

105

components 116. The three versions of the AMS (Q-AMS, TOF-AMS and HR-TOF-

AMS) currently in use vary in the type of mass spectrometric detector used, being

either a quadruple mass spectrometer (Q), a TOF mass spectrometer or a high

resolution TOF mass spectrometer (HR-TOF).

A typical RTAMS configuration includes three component blocks: (1) sample

introduction system; (2) vaporization/ionization source region; and (3) mass analyser.

The sample introduction system is one of the most important factors determining

whether a RTAMS can be used to measure nanoparticles because if insufficient

nanoparticles enter the instrument, no measurements can be conducted in this size

range. Vaporization and ionization methods influence not only its capacity to analyse

nanoparticles but also the resulting data formats (atomic or molecular mass spectra).

This is a very important factor to consider when analysing the composition of

organic compounds and/or other refractory matters (salts and metals), or if the

number concentration of particles is quite low. All of the three major mass analysers

in aerosol mass spectrometers (TOF, quadruple mass filter and the quadruple ion trap)

can be employed for nanoparticles analysis. However, only the TOF and the ion trap

instruments can be used as a true single particle mass spectrometer 99.

Two methods have been employed to introduce particles into aerosol mass

spectrometry. Electrostatic 154 and electrodynamic 130 fields can be used to transmit

and select particles at the lower end of the nanometer size range, less than 50 nm and

10 nm, respectively, which are designated a charge prior to entering the inlet. This

method becomes less effective as the particle size increases because the particle

kinetic energy grows too large to handle with electrostatics or electrodynamic alone

103. Instead, a so-called aerodynamic lens (ADL) developed by McMurry et al 155,

106

tends to be more effective for particles at the higher end of the nanometer size range.

Generally, ADL consists of a 100 µm flow limiting orifice attached to a 1 cm inner

diameter, 30 cm long tube. The particles are gently forced to the centre of the tube

while passing through a series of carefully designed and machined apertures before

they reach the end of the lens where a 2 mm nozzle accelerates the particles into the

next vacuum chamber for measurements 156, 157. It is advertised that TSI 3800-030

ATOFMS, with an ADL, has a transmission of the inlet system from 30-300 nm. The

major difficulties in relation to focusing nanoparticles arise from their low inertia and

high diffusivity. McMurry et al 155 also developed a tool to design and evaluate ADL

systems. After optimizing certain parameters, the ADL system transfers particles 3–

30 nm in diameter with 50-80% efficiency, respectively 158, 159. This technique

bridges the gap between particles less than 3 nm in diameter, which can only be

focused by an electrodynamic lens, and larger particles which are not easy to focus

using electrodynamics 99.

There are many more choices than the particle inlet technique when using

vaporization and ionization methods in RTAMS. The most popular method is laser

desorption/ionization (LDI), which has been employed by ATOFMS. LDI is a single

laser technique and has been found to be highly sensitive. It is ideal for measuring

salt and metal containing particles, which are very hard to be ionized by other ways.

Strong ion signals can be achieved from most types of materials (refractory and

semivolatile, organic and inorganic etc), however considerable fragments induced by

LDI, can make it difficult to identify the organic compounds in the particles. Some

techniques, such as lower pulse energy laser beam and matrix-assisted laser

desorption ionization (MALDI), have been used to reduce the fragmentation.

107

Laser induced plasma is a similar ionization source to LDI except that a much higher

laser irradiance is used. It can be used to quantitatively convert particles to atomic

cations. Reents et al. 139, 160 developed an aerosol mass spectrometer, the Particle

Blaster, using this ionization technique to measure both the complete elemental

composition and particle size of individual particles. The instrument can measure

particles in the size range of 17 - 900 nm diameter. Zachariah et al 161, 162 extended

this work and discussed quantitative measurements using the laser induced plasma

method.

EI as a universal ionization method has also been used in RTAMS. There are three

obvious advantages associated with EI ionization: universal detection of all

vaporized molecules with similar sensitivity; easy identification of the compounds

based on the well-established standard spectra database; and quantification of a

molecules concentration relying on the proportionality between the total ion intensity

and the total number of electrons in the molecule. However, EI ionization still leads

to extensive fragmentation and complex mass spectra and several groups have

coupled aerosol mass spectrometers with soft ionization sources to develop methods

that reduce this fragmentation, including resonance enhanced multiphoton ionization

(REMPI) 163, photoelectron resonance capture ionization (PERCI) 164, Li+ ion

attachment 132, chemical ionization (CI) 165, 166 and vacuum ultraviolet (VUV) single

photo-ionization 167. Although these softer ionization techniques have been shown to

simplify the complexity of organic mass spectra in both gas-phase and aerosol phase

mixtures, the detection limits of most instruments using these methods are not

sufficiently low enough to measure single nanoparticles in the ambient environment.

108

Another method for the real-time chemical analysis of individual aerosol particles is

to combine a surface ionization with mass spectrometry (e.g. SI-PBMS) 127, 146.

Elements with sufficiently low ionization potentials are ionized in contact with a hot

metal surface, and the emitted positive ions are analysed by mass spectrometry. This

technique may provide high sensitivity for certain elements in individual aerosol

particles and has the potential to quantitatively analyse single nanoparticles down to

the size of individual molecules146. This instrument is very robust and suitable for

field measurement applications.

Recently a nanoaerosol mass spectrometer (NAMS) has been designed especially for

the real-time characterization of individual airborne sub-10-nm nanoparticles 130, 137,

138. The NAMS consists of an aerodynamic inlet, a quadruple ion guide, a quadruple

ion trap and a time-of-flight mass analyser. Charged particles in the aerosol are

drawn through the aerodynamic inlet, focused through the ion guide and captured in

the ion trap. A high-energy laser pulse is employed to completely disintegrate the

trapped particles into atomic ions, so that the atomic composition of the particle is

attained from the relative signal intensities of the atomic ions.

3.4.3 Tandem measurements

As summarized by Park et al.168, more complete information on particle transport and

physicochemical properties can be obtained when using multiple instruments in

tandem. For example, Cai et al. 169 used a technique that combined MALDI, laser-

induced fluorescence (LIF) and a dual quadruple ion trap mass spectrometer. It was

demonstrated that the mass spectra of fluorescently labelled nanoparticles with a size

109

of 27 nm in diameter can be acquired by utilizing the unique tandem trap

arrangement along with a frequency scan scheme.

In the field of chemical composition of nanoparticles, several instruments have been

developed and applied in laboratory and field studies. For example, Neville et al. 170

analysed the combustion-generated particles (down to 10 nm) selected with

Differential Mobility Analyser (DMA), using an Instrumental Neutron Activation

Analysis (INAA), to determine the size dependence of elemental concentration (e.g.

Mg, Ca, Fe, Al et al.). Smith et al. 148-150 also successfully measured the chemical

composition of atmospheric aerosols in the 6-33 nm diameter range using a nano-

DMA with a 20-min-resolution, in tandem with TDCIMS. Average ion molar ratios

for nitrate, organics and sulphur species were measured, including nitrogen-

containing organic compounds, organic acids and hydroxyl organic acids.

DMA can also be used with aerosol mass spectrometry to investigate the

composition-dependent mixing characteristics. Park et al. 168 used ATOFMS to

sample mobility-classified particles and observed that the particles with same

mobility diameter were clearly separated into two groups with different vacuum

aerodynamic diameter and with different chemical composition. Okada et al. 171 also

designed a spectrometer, which consisted of a DMA and inductively coupled plasma

mass spectrometer (ICP-MS), to simultaneously measure the size-dependent

concentration and chemical composition of nanometer-sized metal particles. For a

particle size of 30 nm, the lower detection limit of the spectrometer for particle

concentration was about 1 × 105 particles/cm3.

Recently, a novel nanoparticle sampler, which includes up to three UPC-Nano DMA

systems in tandem with one electrostatic precipitator, was developed by McMurry et

110

al 108 to provide sufficient nanoparticle mass for chemical analysis over sampling

periods of about 10 min. It was demonstrated that this sampler can collect 150pg of

particles with the diameter of about 8 nm per hour, which is comparable to the

amount collected by impactors and 100-300 times higher than mass spectrometers

that collect particles at low pressure.

Several other instruments can also be used in tandem to measure particle transport

and physicochemical properties, without using a DMA. For example, Hamilton et al.

used online thermal desorption coupled with comprehensive gas chromatography-

time of flight mass spectrometry (TD-GC×GC-TOF/MS) to analyse ambient and

laboratory aerosols. Over 10, 000 individual organic components were isolated from

around 10mg of aerosol material in a single procedure 172, 173.

In many circumstances, the tandem measurement methods are powerful and quite

effective. For example, integrating DMA and different RTAMS has the potential to

be quite useful when monitoring nanoparticles. At present, the biggest challenge for

nanoparticle elemental composition measurement is the specification of individual

organic compounds on single particles. The further development of softer ionization

techniques, as well as tandem mass spectrometry (MS X MS), will almost certainly

provide a better solution to this issue.

3.4.4 Indirect methods

Accurate time and size resolved data on the chemical composition of ambient

submicrometer aerosols using standard methods of chemical analysis has limitations

due to the difficulty of collecting sufficient material for analysis, as discussed above.

111

Alternative physicochemical methods have therefore been developed that infer the

composition and structure of the particles through indirect measurements. A typical

example of such an indirect measurement method is based on the Tandem

Differential Mobility Analyser technique (TDMA). The method was first introduced

by Liu et al. 174 with the aim of studying the change in particle electrical mobility

(diameter) due to a specific aerosol process with which the particles were

conditioned. The authors named the system an aerosol mobility chromatograph. The

term that is used today to refer to this measurement technique (TDMA) was

introduced years later by Rader and McMurry 175.

The basic design of any TDMA consists of 2 DMA’s, a conditioning device and one

or two CPC’s. The first differential mobility analyser (DMA) selects a narrow size

distribution of particles centred on a set electrical mobility. This quasi-monodisperse

particle size distribution is then “processed” within a conditioner and the change of

the particle diameter and/or size distribution due to the conditioning is measured by a

second DMA in conjunction with a CPC. The second DMA with the CPC can

operate either as a Scanning or Differential Mobility Particle Sizer S(D)MPS. The

term TDMA comes from the fact that there are two DMA’s used in tandem. In

general the second DMA, as well as the conditioning process, can be replaced by

another device that measures a different particle property such as mass, morphology,

composition, aerodynamic size, etc. A detailed review of these tandem techniques

has been recently published 176.

There are a number of different ways the particles can be conditioned/processed in

between the first size classification and the second size measurement. By far the

most common way is to expose the dry selected particles to a well-defined relative

112

humidity (RH) and to measure the water uptake of the particles, a set-up which is

known as hygroscopicity TDMA or H-TDMA. The water uptake is measured

through the distribution of growth factors (Gf) or sometimes called the growth factor

probability density function (Gf-PDF). The hygroscopic growth factor, Gf, is the

ratio of humidified (dw) to dry particle diameter (dd) at a well-defined relative

humidity RH. The H-TDMA’s have been extensively used in various environments.

A detailed review of the H-TDMA techniques and their applications in various

environments has been recently published by Swietlicki et al. 177. Duplissy et al. 178

give a more technical overview with an intercomparison of several H-TDMA models.

The authors also propose a more standardised application of the H-TDMA

techniques aiming at high data quality and data comparability. The importance of

the inversion procedure to obtain the full information from the Gf-PDF has been

recently shown by Gysel et al. 179. They also propose a standardization of the H-

TDMA inversion scheme and provide an excellent discussion of the measurement

errors in TDMA systems. Although there algorithm has been developed for H-

TDMA systems it can be used in other TDMA arrangements.

The Organic TDMA (O-TDMA) has a similar setup to the H-TDMA but instead of

water vapour the particles are exposed to an organic vapour, such as ethanol 180. The

uptake of ethanol is proportional to the organic content enabling estimates of the

organic fraction in even the ultrafine newly formed particles181, 182. The Volatility

TDMA, on the other hand, exposes the size selected particles to a controlled elevated

temperature. The volatilisation (reduction in the size of the particle after heating) of

particles is measured as a function of temperature. The technique provides a suitable

tool to measure online physico-chemical properties of aerosol particles with a high

size resolution by heating the aerosol (typically up to 300°C). Where the aerosol

113

particles are composed of a mixture of several chemical species with differing

volatilisation or decomposition temperatures, the volatilisation curve shows distinct

features associated with the removal of various components and the volume fraction

of these components can be calculated. Comparing with laboratory measurements of

“pure” species the chemical composition of the aerosols can be inferred. The

technique has been extensively used for atmospheric measurements from

characterising the presence of sulphuric acid in marine aerosols 183, nucleation events

in the urban 184 and marine 185 environments as well as diesel exhaust 186.

The hygroscopic and volatile measurements have been recently combined into one

instrument the volatile and hygroscopic TDMA (VH-TDMA) 187. Although several

versions of the technique have been developed 188, 189 the main goal is to measure the

change of the hygroscopic properties as the more volatile components are evaporated

from the particles. In this case more information on the composition of the particles

can be inferred. The technique has been successfully applied in various environments

such as marine environments 188, 190, 191, free troposphere192, smog chamber

measurements 193 as well as diesel exhaust 194.

It is important to point out that TDMA measurements can only infer the chemical

composition of the aerosols and should be used in conjunction with other online

techniques for aerosol composition measurements such as the Aerodyne Aerosol

Mass Spectrometer (AMS).

114

3.5. Monitoring design considerations

There are many different settings, as well as types of environments, where airborne

nanoparticles are present and monitored. The settings include nanoparticle source

emission characterisation, examples of which are vehicle emission testing or

monitoring processes in a reactor where particles are engineered. The environments

include the outdoors, ranging from clean background to very polluted air as well as

indoor environments, including residential buildings, non-industrial workplaces and

occupational workplaces where the engineering of nanoparticles takes place. When

developing an experimental design for nanoparticle monitoring under these different

conditions, several factors need to be taken into consideration, including the fact that:

1. Particle concentrations could be very high, exceeding the range of the

instruments used and thus, requiring the application of a dilution system

195;

2. Variation in some operation parameters may lead to very rapid variations

in source emissions 196, 197;

3. Spatial variation of particle concentration could be very high, under

certain circumstances exceeding an order of magnitude within a few

meters of distance 25, 27, 51, 198;

4. Temporal variation in particle concentration may reach several orders of

magnitude in seconds 74, 197;

5. Rapid formation of secondary nanoparticles may occur and the sampling

process itself may actually affect this process 199, 200;

6. Initial size of the newly formed nanoparticles could be below the size

detection limit of the instrument used and therefore, they may pass

115

undetected or they may be detected only if they grow to sufficiently large

sizes; and

7. Losses of particles could occur within the sampling system and sampling

lines 201, 202.

While an in depth discussion of the above aspects of nanoparticle monitoring is

outside the scope of this review, several comments are made below in relation to

the impact of monitoring design on nanoparticle formation, as well as on the

monitoring requirements stemming from spatial and temporal variation in particle

concentration. For further information, the reader is also referred to Baron &

Willeke 201.

3.5.1 Impact of monitoring design on particle formation and losses

An example of a monitoring process where design may have a significant impact on

particle formation is motor vehicle exhaust sampling. The exhaust plume, composed

of thousands of particle, gaseous and vapour pollutants, often exits the exhaust pipe

at the temperature of up to several hundred degrees Celsius. As it cools down and it

is diluted by the ambient air and/or the clean, filtered air of a dilution system, rapid

particle nucleation and/or condensation of vapours on the surface of existing particles

(condensation seeds) may occur. Therefore, large discrepancies have been observed

when comparing the results of particle number concentrations measured directly

from vehicle exhaust. While particle volume/mass showed reasonable reproducibility

between different studies, results of particle number measurements (with the vast

majority of them being nanoparticles) were difficult to reproduce, even in the same

study. Some artefacts and poor reproducibility in vehicle emission measurements

were due not only to the different instruments used but also to the fact that the

116

majority of particles (in terms of number) belonged to the nucleation mode and were

formed in the process of dilution. The number of particles formed in the nucleation

mode is very sensitive to the dilution conditions and small changes (of the dilution

temperature, for example) can result in a significant change in particle number

concentration. A detailed discussion on the effects of dilution conditions on sampling

and measurements of particle numbers in vehicle emissions can be found in Kasper

et al 203. In order to develop a reproducible and comparable method that could be

used in laboratories around the world, the UNECE-GRPE Particulate Measurement

Program (PMP) was formed, with a focus on the future regulation of nanoparticle

emissions from light duty vehicles and heavy duty engines, with the goal of

amending existing approval legislation to stipulate an extensive reduction of particle

emissions from mobile sources 204. Based upon the recommendation of the PMP, the

European Commission has added a particle number limit to its Euro 5/6 proposed

emission standards for light-duty vehicles. Only solid particles are counted, as

volatile material is removed from the sample, according to the PMP procedure.

The monitoring design can also affect the measured concentration and size

distribution by contributing to particle losses. Impaction losses, and gravitational

settling which are important considerations in handling larger particles are less

important for nanoparticles, however as particle size decreases, diffusion becomes

more important, increasing the rate of particle deposition on the walls of the sample

inlet tubing. Hence the sample inlet must be designed to minimise the sample

residence time. Hence the tubing used to carry the sample to the instrument must be

kept as short as possible. The flow rate may also have to be increased through the use

of an auxiliary flow pump at the instrument end to augment the flow generated by the

instrument.

117

3.5.2 Spatial and temporal variation in particle characteristitcs A recent review 4 showed that the average levels of particle number concentration in

clean outdoor environments (not affected by anthropogenic activities) are of the

order of 2.67 ± 1.79 x 103 cm-3, while levels at urban sites are 4 times higher and

levels at street canyons, roadside, road and tunnel sites are 27, 18, 16 and 64 times

higher, respectively. Thus the range of concentrations between clean and vehicle

effected environments spans over two orders of magnitude. Of importance in relation

to particle monitoring is that it is often only a very small shift in time or space

between these very different environments. For example, Kaur et al. 87 showed that

there is a considerable variability in UF particle exposure, of up to an order of

magnitude above background, within a few seconds and over a few metres as people

move through the polluted microenvironments. The study investigated exposures of

volunteers walking or travelling by bus, car or taxi, along two busy roads and

carrying P-Trak Ultrafine Particle Counters (TSI Model 8525). Similar conclusions

were derived by Gouriou et al. 88 who showed that particle concentration encountered

by car passengers may present high peaks, up to 106 particles cm−3. An example of

spatial variation in particle concentration in indoor environments is given by Ning et

al. 205, who showed that significant spatial variation in indoor particle number

concentration was present even before the air was well mixed.

To account for temporal variation, monitoring should be conducted for a

sufficiently long period of time to include changes in source operation (indoor or

outdoor), meteorological parameters variation (outdoor), and operation of

ventilation system (indoor). To account for spatial variation in particle

concentration, two design options are available: either employing several sets of

instruments, which could be prohibitive due to the costs, or the designation of a

118

reference site, which would enable measurements with only one or two sets of

instruments. In the later case, particle concentration is measured at the reference

site either during the whole monitoring campaign, or the monitoring alternates

between the reference site and the other sites. In the former case, two sets of

instruments are required 206, and in the latter case, it is done with only one set of

instruments 207.

3.6. Summary

The demand for quantitative methods to characterise airborne nanoparticles has led

to significant progress in the design and manufacture of the fast response instruments

required for detecting individual nanoparticles and their characteristics. This review

showed that these instruments are now capable of counting and size classifying the

particles in real time, and also of providing insight into particle structure and

chemical composition. The new generation of condensation particle counters, time-

of-flight mass spectrometers or systems such as Volatilization/Humidification

Tandem Differential Mobility Analysers, enable sufficient insight into the

physicochemical nature of nanoparticles, so that the science of nanoparticle

formation, as well as their post-formation dynamics can be revealed.

Data from the 60 studies reporting total particle number concentrations in a wide

range of environments found a statistically significant difference between the median

concentrations measured by the CPC’s and SMPS’s. These differences are expected

to be larger for environments where a nucleation mode is present and smaller where

aged aerosols dominate. It is, therefore, important to take these differences into

consideration when attempting to establish quantitative understanding of variation in

119

particle concentrations between different environments, which is of significance for

human exposure and epidemiological studies.

Application of many of these techniques is still in the research domain rather than in

everyday use, and requires knowledge and experience to provide meaningful data. It

is expected that the coming years will bring further development of the techniques to

make them cheaper, more robust and applicable to the diversity of the environments

when nanoparticle monitoring is required.

120

3.7 References

1. Y. Song, X. Li and X. Du, European Respiratory Journal, 2009, In Press.

2. A. D. Maynard, Philosophical Transactions of the Royal Society of London A,

2000, 358, 2593-2610.

3. B. S. Cohen and J. Q. Xiong, Health Physics, 1998, 74, 554-560.

4. L. Morawska, Z. Ristovski, E. R. Jayaratne, D. Keogh and X. Ling,

Atmospheric Environment 2008, 42, 8113-8138.

5. R. Aitken, K. Creely and C. Tran, Nanoparticles: An Occupational hygiene

review 274, Health and Safety Executive, 2004.

6. Australian Safety and Compensation Council, ed. D. o. E. a. W. Relations,

Editon edn., 2006.

7. D. Mark, in Nanotechnology: Consequences for Human Health and the

Environment, eds. R. Hester and R. Harrison, RSC Publishing, Cambridge,

Editon edn., 2007.

8. S. V. Hering, M. R. Stolzenburg, F. R. Quant, D. R. Oberreit and P. B. Keady,

Aerosol Science and Technology, 2005, 39, 659-672.

9. C. S. Kim, K. Okuyama and J. F. Fern´andez de la Mora, Aerosol Science and

Technology, 2003, 37, 791-803.

10. J. Wang, V. F. McNeill, D. R. Collins and R. C. Flagan, Aerosol Science and

Technology 2002, 36, 678-689.

11. S. Biswas, P. Fine, M. D. Geller, S. V. Hering and C. Sioutas, Aerosol

Science and Technology, 2005, 39, 419-427.

12. K. Iida, M. R. Stolzenburg, P. McMurry, J. N. Smith, F. R. Quant, D. R.

Oberreit, P. B. Keady, A. Eiguren-Fernandez, G. S. Lewis, N. M. Kreisberg

and S. V. Hering, Aerosol Science and Technology, 2008, 42, 862-871.

13. G. Mordas, H. E. Manninen, T. Petaja, P. P. Aalto, K. Hameri and M.

Kulmala, Aerosol Science and Technology, 2008, 42, 152-158.

121

14. H. Tammet, A. Mirme and E. Tamm, Atmospheric Research, 2002, 62, 315-

324.

15. J. Keskinen, K. Pietarinen and M. Lehtimaki, Journal of Aerosol Science,

1992, 23, 353-360.

16. R. Hillamo, T. Makela and V. M. Kerminen, Electrical low pressure

impactor (ELPI) in atmospheric aerosol studies, Finnish Meteorological

Institute, Helsinki, Finland, 2002.

17. G. Oberdörster, J. N. Finkelstein, C. Johnston, R. Gelein, C. Cox, R. Baggs

and A. C. P. Elder, Acute pulmonary effects of ultrafine particles in rats and

mice, Health Effects Institute, 2000.

18. D. M. Brown, M. R. Wilson, W. MacNee, V. Stone and K. Donaldson,

Toxicology and Applied Pharmacology, 2001, 175, 191-199.

19. A. S. J. Gregg and K. S. W. Sing, Adsorption, surface area and porosity,

Academic Press, London, 1995.

20. H. Burtscher, Journal of Aerosol Science, 2005, 36, 896-932.

21. H. W. Gäggeler, U. Baltensperger, M. Emmenegger, D. T. Jost, A. Schmidt-

Ott, P. Haller and M. Hofmann, J. Aerosol Sci., 1989, 20, 557.

22. A. G. Konstandopoulos, D. Zarvalis, E. Papaioannou, N. D. Vlachos, G.

Boretto, M. F. Pidria, P. Faraldi, O. Piacenza, P. Prenninger, T. Cartus, H.

Schreier, W. Brandstätter, C. Wassermayr, G. Lepperhof, V. Scholz, B. Luers,

J. Schnitzler, M. Claussen, A. Wollmann, M. Maly, G. Tsotridis, B. M.

Vaglieco, S. S. Merola, D. Webster, D. Bergeal, C. Görsmann, H.

Obernosterer, D. Fino, N. Russo, G. Saracco, V. Specchia, N. Moral, A.

D'anna, A. D'alessio, R. Zahoransky, E. Laile, S. Schmidt and M. Ranalli,

The diesel exhaust aftertreatment (DEXA) cluster: a systematic approach to

diesel particulate emission control, SAE 2004-01-0694, 2004.

23. T. C. Bond and T. L. Anderson, Aerosol Science and Technology, 1999, 30,

582-600.

122

24. P. Aalto, K. Hameri, E. Becker, R. Weber, J. Salm, J. Makela, C. Hoell, C.

O'Dowd, H. Karlsson, H. Hansson, M. Vakeva, I. Koponen, G. Buzorious

and M. Kulmala, Tellus, 2001, B53, 344-358.

25. R. Harrison, M. Jones and G. Collins, Atmospheric Environment, 1999, 33,

309-321.

26. D. B. Kittelson, W. F. Watts and J. P. Johnson, Atmospheric Environment,

2004, 38, 9-19.

27. J. Shi, D. Evans, A. Khan and R. Harrison, Atmospheric Environment, 2001,

35, 1193-1202.

28. M. Vakeva, K. Hameri, M. Kulmala, R. Lahdes, J. Ruuskanen and T.

Laitinen, Atmospheric Environment, 1999, 33, 1385-1397.

29. Y. Zhu, W. Hinds, S. Shen and C. Sioutas, Aerosol Science and Technology,

2004, 38, 5-13.

30. D. Imhof, E. Weingartner, C. Ordonez, R. Gerhig, M. Hill, B. Buchmann and

U. Baltersperger, Environmental Science and Technology, 2005, 39, 8341-

8350.

31. P. Paatero, P. Aalto, S. Picciotto, T. Bellander, G. Castano, G. Cattani, J.

Cyrys, M. Kulmala, T. Lanki, F. Nyberg, J. Pekkanen, A. Peters, J. Sunyer

and F. Forastiere, Atmospheric Environment, 2005, 39, 2261-2273.

32. D. Westerdahl, S. Fruin, T. Sax, P. Fine and C. Sioutas, Atmospheric

Environment, 2005, 39, 3597-3610.

33. J. Schneider, U. Kirchner, S. Borrmanna, R. Vogt and V. Scheer,

Atmospheric Environment, 2008, 42, 4257-4268.

34. S. Lechowicz, R. Jataratne, L. Morawska and M. Jamriska, Atmospheric

Environment, 2008, 42, 8353-8357.

35. E. Asmi, M. Antolo, T. Yli-Tuomi, M. Jantunen, P. Aarnio, T. Mäkelä, R.

Hillamo and K. Hameri, Science of The Total Environment, 2009, In Press.

123

36. M. Bergmann, U. Kirchner, R. Vogt and T. Benter, Atmospheric Environment,

2009, In Press.

37. S. Weimer, C. Mohr, R. Richter, J. Keller, M. Mohr, A. S. H. Pre´voˆt and U.

Baltensperger, Atmospheric Environment, 2009, 43, 624-630.

38. P. McMurry and K. Woo, Journal of Aerosol Medicine, 2002, 15, 169-178.

39. T. Tuch, P. Brand, H. Wichmann and J. Heyder, Atmospheric Environment,

1997, 31, 4193-4197.

40. L. Morawska, G. Johnson, Z. D. Ristovski and V. Agranovski, Atmospheric

Environment, 1999, 33, 1983-1990.

41. J. Hitchins, L. Morawska, R. Wolff and D. Gilbert, Atmospheric Environment,

2000, 34, 51-59.

42. M. Junker, M. Kasper, M. Roosli, M. Camenzind, N. Kunzli, C. Monn, G.

Theis and C. Braun, Atmospheric Environment, 2000, 43, 3171-3181.

43. M. Jamriska and L. Morawska, The Science of The Total Environment, 2001,

264, 241-255.

44. M. Pitz, W. Kreyling, B. Holscher, J. Cyrys, H. Wichmann and J. Heinrich,

Atmospheric Environment, 2001, 35, 4357-4366.

45. J. Ruuskanen, T. Tuch, H. Brink, A. Peters, A. Khlystov, A. Mirme, G. Kos,

B. Brunekreef, H. Wickmann, Z. Buzorious, M. Vallius, W. Kreyling and J.

Pekkanen, Atmospheric Environment, 2001, 35, 3729-3738.

46. M. Cheng and R. Tanner, Atmospheric Environment, 2002, 36, 5795-5806.

47. P. Molnar, S. Janhall and M. Hallquist, Atmospheric Environment, 2002, 36,

4115-4123.

48. L. Morawska, E. R. Jayaratne and K. Mengersen, Atmospheric Environment,

2002, 36, 4375-4383.

49. S. Thomas and L. Morawska, Atmospheric Environment, 2002, 36, 4277-

4288.

124

50. B. Wehner, W. Birmili, T. Gnauk and A. Wiedensohler, Atmospheric

Environment, 2002, 36, 2215-2223.

51. Y. Zhu, W. Hinds, S. Kim, S. Shen and C. Sioutas, Atmospheric Environment,

2002, 36, 4323-4335.

52. Y. Zhu, W. C. Hinds, S. Kim and C. Sioutas, Journal of the Air & Waste

Management Assoc., 2002, 52, 1032-1042.

53. M. Ketzel, P. Wahlin, R. Berkowicz and F. Palmgren, Atmospheric

Environment, 2003, 37, 2735-2749.

54. I. Longley, M. Gallagher, J. Dorsey, M. Flynn, J. Allan, M. Alfarra and D.

Inglis, Atmospheric Environment, 2003, 37, 1563-1571.

55. P. Tunved, H. Hansson, M. Kulmala, P. Aalto, Y. Viisanen, H. Karlsson, A.

Kristensson, E. Swietlicki, M. Maso, J. Strom and M. Komppula,

Atmospheric Chemistry and Physics Discussions, 2003, 3, 2183-2205.

56. B. Wehner and A. Wiedensohler, Atmospheric Chemistry and Physics

Discussions, 2003, 3, 867-879.

57. L. Gidhagen, C. Johansson, J. Langner and G. Olivares, Atmospheric

Environment, 2004, 38, 2029-2044.

58. G. Gramotnev and Z. Ristovski, Atmospheric Environment, 2004, 38, 1767-

1776.

59. G. Gramotnev, Z. D. Ristovski, R. J. Brown and P. Madl, Atmospheric

Environment, 2004, 38, 2607-2610.

60. T. Hussein, A. Puustinen, P. Aalto, J. Makela, K. Hameri and M. Kulmala,

Atmospheric Chemistry and Physics Discussions, 2004, 4, 391-411.

61. M. Jamriska, L. Morawska, S. Thomas and C. He, Environmental Science &

Technology, 2004, 38, 6701-6709.

62. S. Janhall, A. Jonsson, P. Molnar, E. Svensson and M. Hallquist, Atmospheric

Environment, 2004, 38, 4331-4340.

125

63. C. Jeong, P. Hopke, D. Chalupa and M. Utell, Environmental Science and

Technology, 2004, 38, 1933-1940,.

64. M. Ketzel, P. Whalin, A. Kristensson, E. Swietlicki, R. Berkowicz, O.

Nielsen and F. Palmgren, Atmospheric Chemsitry and Physics, 2004, 4, 281-

292.

65. L. Morawska, W. Hofmann, S. Thomas, Z. D. Ristovski, M. Jamriska, T.

Rettenmoser and S. Kagerer, Atmospheric Environment, 2004, 38, 3529-3533.

66. C. Stanier, A. Khlystov and S. Pandis, Atmospheric Environment, 2004, 38,

3275-3284.

67. L. Gidhagen, C. Johansson, J. Langner and V. Foltescu, Atmospheric

Environment, 2005, 39, 1711-1725.

68. N. S. Holmes, L. Morawska, K. Mengersen and R. Jayaratne, Atmospheric

Environment, 2005, 39, 3977-3988.

69. D. Imhof, E. Weingartner, A. Prevot, C. Ordonez, R. Kurtenbach, P. Wiesen,

J. Rodler, P. Sturm, I. McCrae, A. Sjodin and U. Baltersperger, Atmospheric

Chemistry and Physics Discussions, 2005, 5, 5127-5166.

70. S. Rodriguez, R. van Dingenen, J. Putaud, S. Martins-Dos Santos and D.

Roselli, Atmospheric Environment, 2005, 39, 6734-6746.

71. S. Janhall, F. Olofson, P. Andersson, J. Pettersson and M. Hallquist,

Atmospheric Environment, 2006, 40, 5355-5366.

72. A. Virtanen, T. Ronkko, J. Kannosto, J. Ristimaki, J. Makela, J. Keskinen, T.

Pakkanen, R. Hillamo, L. Pirjola and K. Hameri, Atmospheric Chemistry and

Physics, 2006, 6, 2411-2421.

73. P. Wåhlin, F. Palmgren and R. Van Dingenen, Atmospheric Environment,

2001, 35, S63-S69.

74. K. S. Woo, D. R. Chen, D. Y. H. Pui and W. E. Wilson, Aerosol Science and

Technology, 2001, 34, 57-65.

126

75. M. Abu-Allaban, W. Coulomb, A. Gertler, J. Gillies, W. Pierson, C. Rogers, J.

Sagebiel and L. Tarnay, Aerosol Science and Technology, 2002, 36, 771-789.

76. L. Laakso, T. Hussein, P. Aarino, M. Komppula, V. Hiltunen, Y. Viisanen

and M. Kulmala, Atmospheric Environment, 2003, 37, 2629-2641.

77. T. Hussein, K. Hameri, P. Aalto, P. Paatero and M. Kulmala, Atmospheric

Environment, 2005, 39, 1655-1668.

78. J. Mejia, L. Morawska and K. Mengersen, Atmospheric Chemistry and

Physics, 2008, 8, 1127-1138.

79. J. Pey, S. Rodrı´guez, X. Querol, A. Alastuey, T. Moreno, J.-P. Putaud and R.

Van Dingenen, Atmospheric Environment 2008, 42, 9052-9062.

80. T. L. Barone and Y. Zhu, Atmospheric Environment, 2008, 42, 6749-6758.

81. W. Yue, M. Stölzel, J. Cyrys, M. Pitz, J. Heinrich, W. G. Kreyling, H. E.

Wichmann, A. Peters, S. Wang and P. K. Hopke, Science of the Total

Environment, 2008, 398, 133-144.

82. Z. Wu, M. Hu, P. Lin, S. Liu, B. Wehner and A. Wiedensohler, Atmospheric

Environment, 2008, 42, 7967-7980.

83. D. Westerdahl, X. Wang, X. Pan and K. M. Zhang, Atmospheric Environment,

2009, 43, 697-705.

84. G. Buonanno, A. A. Lall and L. Stabile, Atmospheric Environment, 2009, 43,

1100-1105.

85. H. Minoura, H. Takekawa and S. Terada, Atmospheric Environment, 2009, 43,

546-556.

86. K. Hameri, M. Kulmala, P. Aalto, K. Leszczynski, R. Visuri and K.

Hamekoski, Atmospheric Research, 1996, 41, 281-298.

87. S. Kaur, R. Clark, P. Walsh, S. Arnold, R. Colvile and M. Nieuwenhuijsen,

Atmospheric Environment, 2006, 40, 386-398.

127

88. F. Gouriou, J. Morin and M. Weill, Atmospheric Environment, 2004, 38,

2831-2840.

89. E. Demou, P. Perer and S. Hellwig, Annals of Occupational Hygiene, 2008,

52, 695-706.

90. J. H. Han, E. J. Lee, J. H. Lee, K. P. So, Y. H. Lee, G. N. Bae, S.-B. Lee, J. H.

Ji, M. H. Cho and I. J. Yu, Inhalation Toxicology, 2008, 20, 741-749.

91. A. D. Maynard, P. A. Baron, M. Foley, A. A. Shvendova, E. R. Kisin and V.

Castranova, Journal of Toxicology and Environmental Health, Part A, 2004,

67, 87-107.

92. M. Seipenbusch, A. Binder and G. Kasper, Annals of Occupational Hygiene,

2008, 52, 707-706.

93. B. Yeganeh, C. M. Kull, M. S. Hull and L. C. Marr, Environmental Science

and Technology, 2008, 42, 4600-4606.

94. RSRAE, Nanoscience and nanotechnologies: opportunities and uncertainties

RS policy document 19/04. , The Royal Society and Royal Academy of

Engineering, London, 2004.

95. M. Kulmala, H. Vehkamaki, T. Petaja, M. Dal Maso, A. Lauri, V. M.

Kerminen, W. Birmili and P. H. McMurry, Journal of Aerosol Science, 2004,

35, 143-176.

96. G. Oberdorster, E. Oberdorster and J. Oberdorster, Environmental Health

Perspectives, 2005, 113, 823-839.

97. P. Biswas and C. Y. Wu, Journal of the Air & Waste Management

Association, 2005, 55, 708-746.

98. K. A. Prather, C. D. Hatch and V. H. Grassian, Annu. Rev. Anal. Chem., 2008,

1, 485-514.

99. D. G. Nash, T. Baer and M. V. Johnston, Int. J. Mass Spectrom., 2006, 258,

2-12.

100. R. C. Sullivan and K. A. Prather, Analytical Chemistry, 2005, 77, 3861-3885.

128

101. D. M. Murphy, Science, 2005, 307, 1888-1890.

102. Y. C. Liu and H. He, Progress in Chemistry, 2007, 19, 1620-1631.

103. M. V. Johnston, S. Y. Wang and M. S. Reinard, Applied Spectroscopy, 2006,

60, 264A-272A.

104. P. H. McMurry, Atmospheric Environment, 2000, 34, 1959-1999.

105. J. C. Chow, Journal of the Air and Waste Management Association, 1995, 45,

320-382.

106. T. A. Pakkanen, V.-M. Kerminen, C. H. Korhonen, R. E. Hillamo, P. Aarnio,

T. Koskentalo and W. Maenhaut, Atmospheric Environment, 2001, 35, 4593-

4607.

107. S. B. Sardar, P. M. Fine, P. R. Mayo and C. Sioutas, Environmental Science

& Technology, 2005, 39, 932-944.

108. P. H. McMurry, A. Ghimire, H.-K. Ahn, H. Sakurai, K. Moore, M.

Stolzenburg and J. N. Smith, Environmental Science & Technology, 2009,

available online; doi:10.1021/es8029335.

109. A. Berner and C. Luerzer, The Journal of Physical Chemistry, 1980, 84,

2079-2083.

110. V. A. Marple and B. A. Olson, A micro-orifice impactor with cut sizes down

to 10 nanometer for diesel exhaust sampling G1135242, G1145242,

G1155242/#2782, Pennsylvania State University, University Park, PA, 1999.

111. MSP Corporation, Product information of Model 120, 122 and 125 -

MOUDI-II Impactors, http://www.mspcorp.com/images/products/MSP-

PI120_Rev%20A%20(PI)%20MOUDI%20II-002.pdf, Accessed April 26,

2009.

112. M. D. Geller, S. Kim, C. Misra, C. Sioutas, B. A. Olson and V. A. Marple,

Aerosol Science and Technology, 2002, 36, 748-762.

113. M. P. H. Bottger, Z. Bi, D. Adolph, K. A. Dick, L. S. Karlsson, M. N. A.

Karlsson, B. A. Wacaser and K. Deppert, Nanotechnology, 2007, 18, 105304.

129

114. J. T. Hinkley, H. A. Bridgman, B. J. P. Buhre, R. P. Gupta, P. F. Nelson and

T. F. Wall, Science of The Total Environment, 2008, 391, 104-113.

115. M. Wensing, T. Schripp, E. Uhde and T. Salthammer, Science of the Total

Environment, 2008, 407, 418-427.

116. D. E. Heard, ed., Analytical Techniques for Atmospheric Measurement,

Blackwell Publishing, 2006.

117. D. A. Orsini, Y. Ma, A. Sullivan, B. Sierau, K. Baumann and R. J. Weber,

Atmospheric Environment, 2003, 37, 1243-1259.

118. A. Khlystov, G. P. Wyers and J. Slanina, Atmospheric Environment, 1995, 29,

2229-2234.

119. S. B. Hong, D. S. Kim, S. Y. Ryu, K. W. Lee, Y. J. Kim and J. H. Lee, Part.

Part. Syst. Charact., 2009, 25, 444-453.

120. S. Orsini, S. Livi, K. Torkar, S. Barabash, A. Milillo, P. Wurz, A. di Lellis, E.

Kallio and SERENA Team1, Planetary and Space Science, In Press,

Available Online, doi:10.1016/j.pss.2008.1009.1012

121. J. C. Chow, J. Z. Yu, J. G. Watson, S. S. H. Ho, T. L. Bohannan, M. D. Hays

and K. K. Fung, J. Environ. Sci. Health Part A-Toxic/Hazard. Subst. Environ.

Eng., 2007, 42, 1521-1541.

122. B. J. Williams, A. H. Goldstein, N. M. Kreisberg and S. V. Hering, Aerosol

Science and Technology, 2006, 40, 627-638.

123. N. M. Kreisberg, S. V. Hering, B. J. Williams, D. R. Worton and A. H.

Goldstein, Aerosol Science and Technology, 2009, 43, 38-52.

124. A. L. Hunt and G. A. Petrucci, Trac-Trends in Analytical Chemistry, 2002, 21,

74-81.

125. D. M. Murphy, Mass Spectrometry Reviews, 2007, 26, 150-165.

126. K. Park, G. Cho and J. H. Kwak, Aerosol Science and Technology, 2009, 43,

375-386.

130

127. W. D. Davis, The Journal of Vacuum Science and Technology, 1973, 10, 278.

128. D. T. Suess and K. A. Prather, Chemical Reviews, 1999, 99, 3007-3036.

129. C. A. Noble and K. A. Prather, Mass Spectrometry Reviews, 2000, 19, 248-

274.

130. S. Y. Wang, C. A. Zordan and M. V. Johnston, Analytical Chemistry, 2006,

78, 1750-1754.

131. J. Schreiner, C. Voigt, P. Zink, A. Kohlmann, D. Knopf, C. Weisser, P. Budz

and K. Mauersberger, Review of Scientific Instruments, 2002, 73, 446-452.

132. M. R. Canagaratna, J. T. Jayne, J. L. Jimenez, J. D. Allan, M. R. Alfarra, Q.

Zhang, T. B. Onasch, F. Drewnick, H. Coe, A. Middlebrook, A. Delia, L. R.

Williams, A. M. Trimborn, M. J. Northway, P. F. DeCarlo, C. E. Kolb, P.

Davidovits and D. R. Worsnop, Mass Spectrometry Reviews, 2007, 26, 185-

222.

133. J. T. Jayne, D. C. Leard, X. F. Zhang, P. Davidovits, K. A. Smith, C. E. Kolb

and D. R. Worsnop, Aerosol Science and Technology, 2000, 33, 49-70.

134. E. Gard, J. E. Mayer, B. D. Morrical, T. Dienes, D. P. Fergenson and K. A.

Prather, Analytical Chemistry, 1997, 69, 4083-4091.

135. K. A. Prather, T. Nordmeyer and K. Salt, Analytical Chemistry, 1994, 66,

1403-1407.

136. A. Kurten, J. Curtius, F. Helleis, E. R. Lovejoy and S. Borrmann, Int. J. Mass

Spectrom., 2007, 265, 30-39.

137. S. Wang and M. V. Johnston, Int. J. Mass Spectrom., 2006, 258, 50-57.

138. C. A. Zordan, S. Wang and M. V. Johnston, Environmental Science &

Technology, 2008, 42, 6631-6636.

139. W. D. Reents and Z. Ge, Aerosol Science and Technology, 2000, 33, 122 -

134.

131

140. B. Oktem, M. P. Tolocka and M. V. Johnston, Analytical Chemistry, 2004, 76,

253-261.

141. D. A. Lake, M. P. Tolocka, M. V. Johnston and A. S. Wexler, Environmental

Science & Technology, 2003, 37, 3268-3274.

142. M. S. Reinard, K. Adoua, J. M. Martini and M. V. Johnston, Atmospheric

Environment, 2007, 41, 9397-9409.

143. M. P. Tolocka, D. A. Lake, M. V. Johnston and A. S. Wexler, Atmospheric

Environment, 2004, 38, 3215-3223.

144. A. Zelenyuk and D. Imre, Aerosol Science and Technology, 2005, 39, 554-

568.

145. A. Zelenyuk, J. Yang, E. Choi and D. Imre, Aerosol Science and Technology,

2009, 43, 411-424.

146. M. Svane, M. Hagstrom and J. B. C. Pettersson, Aerosol Science and

Technology, 2004, 38, 655 - 663.

147. D. Voisin, J. N. Smith, H. Sakurai, P. H. McMurry and F. L. Eisele, Aerosol

Science and Technology, 2003, 37, 471-475.

148. J. N. Smith, K. F. Moore, P. H. McMurry and F. L. Eisele, Aerosol Science

and Technology, 2004, 38, 100 - 110.

149. J. N. Smith, M. J. Dunn, T. M. VanReken, K. Iida, M. R. Stolzenburg, P. H.

McMurry and L. G. Huey, Geophysical Research Letters, 2008, 35.

150. J. N. Smith, K. F. Moore, F. L. Eisele, D. Voisin, A. K. Ghimire, H. Sakurai

and P. H. McMurry, J. Geophys. Res.-Atmos., 2005, 110.

151. J. N. Smith and G. J. Rathbone, International Journal of Mass Spectrometry,

2008, 274, 8-13.

152. H. J. Tobias, P. M. Kooiman, K. S. Docherty and P. J. Ziemann, Aerosol

Science and Technology, 2000, 33, 170 - 190.

153. H. J. Tobias and P. J. Ziemann, Analytical Chemistry, 1999, 71, 3428-3435.

132

154. D. B. Kane, B. Oktem and M. V. Johnston, Aerosol Science and Technology,

2001, 35, 990 - 997.

155. X. Wang and P. H. McMurry, Aerosol Science and Technology, 2006, 40,

320-334.

156. P. Liu, P. J. Ziemann, D. B. Kittelson and P. H. McMurry, Aerosol Science

and Technology, 1995, 22, 293-313.

157. P. Liu, P. J. Ziemann, D. B. Kittelson and P. H. McMurry, Aerosol Science

and Technology, 1995, 22, 314-324.

158. X. Wang, A. Gidwani, S. L. Girshick and P. H. McMurry, Aerosol Science

and Technology, 2005, 39, 624 - 636.

159. X. Wang, F. E. Kruis and P. H. McMurry, Aerosol Science and Technology,

2005, 39, 611 - 623.

160. W. D. Reents and M. J. Schabel, Analytical Chemistry, 2001, 73, 5403-5414.

161. R. Mahadevan, D. Lee, H. Sakurai and M. R. Zachariah, The Journal of

Physical Chemistry A, 2002, 106, 11083-11092.

162. D. Lee, K. Park and M. R. Zachariah, Aerosol Science and Technology, 2005,

39, 162-169.

163. M. Bente, T. Adam, T. Ferge, S. Gallavardin, M. Sklorz, T. Streibel and R.

Zimmermann, Int. J. Mass Spectrom., 2006, 258, 86-94.

164. B. W. LaFranchi and G. A. Petrucci, Int. J. Mass Spectrom., 2006, 258, 120-

133.

165. J. D. Hearn and G. D. Smith, Analytical Chemistry, 2004, 76, 2820-2826.

166. W. A. Harris, P. T. A. Reilly and W. B. Whitten, Analytical Chemistry, 2007,

79, 2354-2358.

167. S. J. Hanna, P. Campuzano-Jost, E. A. Simpson, D. B. Robb, I. Burak, M. W.

Blades, J. W. Hepburn and A. K. Bertram, Int. J. Mass Spectrom., 2009, 279,

134-146.

133

168. K. Park, D. Dutcher, M. Emery, J. Pagels, H. Sakurai, J. Scheckman, S. Qian,

M. R. Stolzenburg, X. Wang, J. Yang and P. H. McMurry, Aerosol Science

and Technology, 2008, 42, 801-816.

169. Y. Cai, W. P. Peng and H. C. Chang, Analytical Chemistry, 2003, 75, 1805-

1811.

170. M. Neville, J. F. McCarthy and A. F. sarofim, Atmospheric Environment

(1967), 1983, 17, 2599-2604.

171. Y. Okada, J. Yabumoto and K. Takeuchi, Journal of Aerosol Science, 2002,

33, 961-965.

172. J. F. Hamilton, P. J. Webb, A. C. Lewis, J. R. Hopkins, S. Smith and P. Davy,

Atmospheric Chemistry and Physics, 2004, 4, 1279-1290.

173. J. F. Hamilton, P. J. Webb, A. C. Lewis and M. M. Reviejo, Atmospheric

Environment, 2005, 39, 7263-7275.

174. B. Y. H. Liu, D. Y. H. Pui, K. T. Whitey, D. B. Kittelson and Y. Kousaka,

Atmospheric Environment, 1978, 12, 99-104.

175. D. J. Rader and P. H. McMurry, Journal of Aerosol Science, 1986, 17, 771-

787.

176. K. Park, D. Dutcher, M. Emery, J. Pagels, H. Sakurai, J. Scheckman, M. Qian,

M. C. Stoltzenburg, X. Wang, J. Yang and P. H. McMurry, Aerosol Science

and Technology, 2008, 42, 801-816.

177. E. Swietlicki, H. C. Hansson, K. Hameri, B. Svenningsson, A. Massling, G.

McFiggans, P. H. McMurry, T. Petaja, P. Tunved, M. Gysel, D. Topping, E.

Weingartner, U. Baltensperger, J. Rissler, A. Wiedensohler and M. Kulmala,

Tellus Series B-Chemical and Physical Meteorology, 2008, 60, 432-469.

178. J. Duplissy, M. Gysel, S. Sjogren, N. Meyer, N. Good, L. Kammermann, V.

Michaud, R. Weigel, S. Martins dos Santos, C. Gruening, P. Villani, P. Laj, K.

Sellegri, A. Metzger, G. McFiggans, G. Wehrle, R. Ritcher, J. Dommen, Z. D.

Ristovski, U. Baltensperger and E. Weingartner, Atmospheric Measurment

Techniques Discussions, 2008, 1, 127-168.

134

179. M. Gysel, G. B. McFiggans and H. Coe, Journal of Aerosol Science, 2009, 40,

134-151.

180. J. Joutsensaari, P. Vaattovaara, M. Vesterinen, K. Hameri and A. Laaksonen,

Atmospheric Chemistry and Physics, 2001, 1, 51-60.

181. P. Vaattovaara, P. E. Huttunen, Y. J. Yoon, J. Joutsensaari, K. E. J. Lehtinen,

C. D. O'Dowd and A. Laaksonen, Atmospheric Chemistry and Physics, 2006,

6, 4601-4616.

182. P. Vaattovaara, M. Rasanen, T. Kuhn, J. Joutsensaari and A. Laaksonen,

Atmospheric Chemistry and Physics, 2005, 5, 3277-3287.

183. D. A. Orsini, A. Wiedensohler, F. Stratmann and D. S. Covert, Journal of

Atmospheric and Oceanic Technology, 1999, 16, 760-772.

184. H. Sakurai, M. Fink, P. H. McMurry, L. Maudlin, K. Moore, J. Smith and F.

L. Eisele, Journal of Geophysical Research, 2005, 110, D22S04

185. G. R. Johnson, Z. D. Ristovski, B. D'Anna and L. Morawska, Journal of

Geophysical Research, 2005, 110, D20203.

186. H. Sakurai, K. Park, P. H. McMurry, D. D. Zarling, D. B. Kittelson and P. J.

Ziemann, Environmental Science & Technology, 2003, 37, 5487-5495.

187. G. R. Johnson, Z. Ristovski and L. Morawska, Journal of Aerosol Science,

2004, 35, 443-455.

188. J. M. Tomlinson, R. J. Li and D. R. Collins, Journal of Geophysical

Research-Atmospheres, 2007, 112.

189. P. Villani, D. Picard, V. Michaud, P. Laj and A. Wiedensohler, Aerosol

Science and Technology, 2008, 42, 729-741.

190. C. A. Fletcher, G. R. Johnson, Z. D. Ristovski and M. Harvey, Environmental

Chemistry, 2007, 4, 162-171.

191. G. R. Johnson, Z. D. Ristovski, B. D'Anna and L. Morawska, Journal of

Geophysical Research D: Atmospheres, 2005, 110, 1-14.

135

192. H. Venzac, K. Sellegri, P. Villani, D. Picard and P. Laj, Atmospheric

Chemistry and Physics, 2009, 9, 1465-1478.

193. N. K. Meyer, J. Duplissy, M. Gysel, A. Metzger, J. Dommen, E. Weingartner,

M. R. Alfarra, C. Fletcher, N. Good, G. McFiggans, A. M. Jonsson, M.

Hallquist, U. Baltensperger and Z. D. Ristovski, Atmospheric Chemistry and

Physics 2009, 9, 721-732.

194. N. K. Meyer and Z. D. Ristovski, Environmental Science and Technology,

2007, 41, 7309 - 7314.

195. L. Knibbs, R. de Dear, L. Morawska and P. Coote, Atmospheric Environment,

2007, 41, 4553-4557.

196. Z. Ristovski, E. R. Jayaratne, M. Lim, G. A. Ayoko and L. Morawska,

Environmental Science & Technology, 2006, 40, 1314-1320.

197. L. Morawska, C. He, J. Hitchins, D. Gilbert, S. Parappukkaran and K.

Mengersen, Atmospheric Environment, 2003.

198. D. Brouwer, J. Gijsbers and M. Lurvink, Annals of Occupational Hygiene,

2004, 48, 439-453.

199. T. Kawai, Y. Goto and M. Odaka, Influence of dilution process on engine

exhaust nanoparticles, Society of Automobile Engineers No. SAE-2004-01-

0963, 2004.

200. I. S. Abdul-Khalek and D. B. Kittelson, The Influence of Dilution Conditions

on Diesel Exhaust Particle Size Distribution Measurements, Society of

Automobile Engineers No. SAE-1999-01-1142, 1999.

201. P. A. Baron and K. Willeke, eds., Aerosol Measurement: Principles,

Techniques and Applications, 2nd edn., van Nostrand Reinhold, New York,

2001.

202. L. Morawska, C. R. He, J. Hitchins, D. Gilbert and S. Parappukkaran,

Atmospheric Environment, 2001, 35, 3463-3473.

136

203. M. Kasper, Sampling and measurement of nanoparticle emissions for type

approval and field control, Society of Automobile Engineers, 2005.

204. M. Mohr and U. Lehmann, Comparison Study of Particle Measurement

Systems for Future Type Approval Application, Swiss Contribution to GRPE

Particulate Measurement Program (GRPE-PMP CH5), 2003.

205. Z. Ning, C. Cheung, J. Fu, M. Liu and M. Schnell, Science of the Total

Environment, 2006, 367, 822-830.

206. N. Holmes, L. Morawska and K. Mengersen, Atmospheric Environment, 2006,

40, 5902-5928.

207. N. Bukowiecki, J. Dommen, A. Prevot, E. Weingartner and U. Baltensperger,

Atmospheric Chemistry and Physics, 2003, 3.

137

CHAPTER 4

Particle detection efficiency for CPC’s depends on ambient aerosol composition and

condensation medium.

H.C. CHEUNG, G.R. JOHNSON, L. MORAWSKA*, and Z.D. RISTOVSKI

International Laboratory of Air Quality and Health, Queensland University of

Technology

GPO Box 2434, Brisbane QLD 4001, Australia

Submitted to Atmospheric Environment in March 2011

*Corresponding Author:

Tel: (617) 3138 2616; Fax: (617) 3138 9079

E-mail - l.morawska@qut.edu.au

138

STATEMENT OF JOINT AUTHORSHIP

Title: Observation of new particle formation in subtropical urban environment.

Authors: H.C. Cheung, L. Morawska*, Z.D. Ristovski

H.C. Cheung

Designed and developed the methodology, conducted the field measurement,

analysed and interpreted the data, and wrote the manuscript.

L. Morawska

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

Z.D. Ristovski

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

139

CHAPTER 4. PARTICLE DETECTION EFFICIENCY FOR CPC'S

DEPENDS ON AMBIENT AEROSOL COMPOSITION AND

CONDENSATION MEDIUM.

H.C. CHEUNG, G.R. JOHNSON, L. MORAWSKA*, and Z.D. RISTOVSKI

International Laboratory of Air Quality and Health, Queensland University of

Technology

GPO Box 2434, Brisbane QLD 4001, Australia

Abstract

The detection efficiency of butanol-based (BCPC) and water-based (WCPC)

CPCs was investigated in response to particles of different composition found in

indoor and outdoor environments. Number concentrations of NaCl, DOS, laser toner

and citric acid particles were measured using both classes of CPC. The relative

response ratios of the CPC’s to a given particle type was used as a measure of the

relative detection efficiencies of the CPCs with respect to those particles. Particle

detection at small sizes was strongly dependant on the solubility of the target aerosol

in the CPC condensation medium especially near the lower nominal diameter limit of

the CPC. Strongly water soluble aerosols resulted in more efficient particle detection

in water based CPC’s relative to butanol based instruments while water insoluble

aerosols such as DOS and laser toner particles which are soluble in alcohols, were

more readily detected at small sizes by the butanol based instrument. Citric acid, a

material displaying good solubility in both water and butanol yielded comparable

140

response in both types of instrument within the nominal detection size range of both

instruments.

Water insoluble but butanol soluble DOS and laser toner particles smaller

than 20 nm were undercounted by the WCPC. This undercounting was less

pronounced for particles larger than 20 nm. Water soluble NaCl particles, were

undercounted by the BCPC at very small sizes but the BCPC/WCPC ratio converged

to unity at size larger than 12 nm. Citric acid particles tended to be undercounted by

the WCPC compared to the BCPC, yielding BCPC/WCPC ratios in a narrow range

(1.07 - 1.20) for particles larger than 8 nm.

It was concluded that the nominal lower detection size limit specification for

a CPC can be very composition dependant with soluble compositions appearing to

significantly enhance the activation of vapour condensation in the CPC. This fact

should be given careful consideration when comparing data from CPCs using

different condensation media or when choosing a CPC for ambient measurements.

Keywords: Condensation Particle Counters; detection efficiency; ultrafine particles.

141

4.1 Introduction

Condensation Particle Counters (CPCs) are widely used for measuring the

number concentration of nanoparticles in outdoor and indoor environments and in

emissions from specific sources such as combustion engines industrial process and

laser printers (He et al., 2007; Morawska et al., 2008). They are also commonly used

in size distribution measurement systems such the Differential Mobility Analyser and

Scanning Mobility Particle Sizer (Hämeri et al., 2002; Cheung et al., 2010) as well as

physicochemical characterisation systems based on the tandem differential mobility

analysis technique (Rader and McMurry, 1986; Johnson et al., 2004). Although there

have been a number of CPCs available commercially with different specifications for

particle size range, concentration range, response time and sampling flow rate, the

basic working principle is similar in each case. Nanoparticles too small to be

detected directly by optical means are enlarged to detectable sizes by condensing a

low vapour pressure material onto their surfaces from the gas phase (Morawska et al.,

2009b). The two major types of condensable species commonly used are alcohols

(mostly butanol and isopropyl alcohol) and water. The lower detection size limit of

insoluble particles depends on the Kelvin diameter corresponding to the super-

saturation ratio achieved in the aerosol as it passes through the instrument. It is also

known that the condensation tube design of a CPC affects the minimum detectable

size at which a particle can be detected by the CPC. The detectable particles sizes are

governed by the temperature difference between the saturator and condenser inside

the CPCs so that the minimum detectable size can be varied to some extent by

changing that temperature difference (Metres et al., 1995; Hering and Stolzenburg,

2005).

142

Soluble particles, however, effectively lower the saturation vapour pressure of

the condensing species above the particle surface, resulting in a smaller activation

diameter and hence a lower detection size limit compared to insoluble materials.

Therefore, divergence of detection efficiencies of different aerosols by water based

CPC (WCPC) and butanol based CPC (BCPC) are to be expected.

Several studies have demonstrated variations in CPC performance with

different aerosols. These studies have investigated the detection efficiencies of

WCPC and BCPC for different types of aerosol including lab-generated aerosols and

outdoor ambient aerosols. Biswas et al. (2005) found that the WCPC/BCPC

concentration ratio ranged from 1.0 to 1.1 for particle sizes in the range 10 to 50 nm

for ammonium sulphate ((NH4)2SO4), ammonium nitrate (NH4NO3), glutaric acid

and adipic acid. Hering et al. (2005) examined the detection efficiency of a WCPC

with respect to hydrophobic species including oleic acid and dioctyl sebacate (DOS)

as well as hydrophilic species including sodium chloride (NaCl), (NH4)2SO4 and

NH4NO3. The investigation showed that a lower D50 (diameter detected with 50 %

efficiency) was obtained for water-soluble species (3.6 nm < D50 < 4.8 nm) but

higher values for hydrophobic species (8 nm < D50 < 30 nm). Moreover, Iida et al.

(2008) conducted a comparison between a WCPC and a BCPC by measuring the

particle number concentrations at an urban location and in a freeway tunnel. The

results showed a WCPC/BCPC ratio < 1.0 for ambient urban particles with diameters

< 5 nm. In contrast, a WCPC/BCPC ratio > 1.0 was obtained for freeway tunnel

particles. The authors concluded that the differences observed between these two

types of particles could be due either to differences in the instrumentation used, or to

differences in particles composition. Franklin et al. (2010) compared the responses of

two different WCPC’s (WCPC1 and WCPC2) and one BCPC to diesel combustion

143

aerosols. They found that WCPC1 (Model: TSI 3782, with nominal D50 = 10 nm)

significantly undercounted particles compared to the BCPC (Model: TSI 3010, with

nominal D50 = 10 nm) while WCPC2 (Model: TSI 3786, with nominal D50 = 2.5 nm)

closely agreed with the BCPC and only undercounted for particle diameters

approaching the lower cut off size of the instrument. The differences in detection

efficiency between different CPC’s for different aerosol species were less

pronounced for particles with diameters larger than 30 nm (Hering et al., 2005;

Mordas et al., 2008). These previous studies indicate that the particle number

concentration measured by a WCPC depends on the aerosol composition and its

solubilities to the condensation medium. This understanding of the varying detection

ability of CPCs with different particulate species is especially important for source

specific measurement where the aerosol may be of a specific chemical composition

and size.

In previous studies, hydrophilic or hydrophobic aerosols were selected for the

CPC detection performance testing. However, little information is currently available

for species which are soluble both in water and alcohol. Therefore in this study, we

undertook a comprehensive comparison of the detection performance of WCPC’s

and BCPC’s for hydrophobic species as well as for citric acid which can dissolve in

both water and alcohol in order to develop an in-depth understanding of the influence

of particle composition CPC performance.

4.2 EXPERIMENTAL SETUP

In this study, four different types of particle were tested including hygroscopic

(NaCl), hydrophobic (DOS and Laser Printer Toner) and citric acid which can

144

dissolve in water and alcohol (Haynes ed., 2011). DOS and Laser printer toner

particles have been shown to be hydrophobic through VH-TDMA measurements

where the particles absorb no moisture when exposed to very high humidity (Johnson

et al., 2004; Morawska et al., 2009a). The response of different classes of CPC to

diesel combustion engine exhaust aerosols has previously been examined in some

detail by Franklin et al. (2000), so that aerosol was not included in the current study.

A list of the test particle composition and the equipment used in the experiment is

shown in Table 4-1.

Aerosols Solubility

Sodium Chloride (NaCl) Soluble in water; insoluble in alcohol

Dioctyl sebacate (DOS) Insoluble in water; soluble in alcohol

Laser Printer Toner Insoluble in water; slightly soluble in alcohol

Citric acid Soluble in water and alcohol

Instruments Manufacture/model Remarks

Electrostatic classifier TSI 3080 Size range coverage: 4 – 60 nm

TSI 3071 Size range coverage: > 60 nm

Water based CPC TSI 3781

TSI 3782

D50 ≥ 6 nm

D50 ≥ 10 nm

Butanol based CPC TSI 3010

TSI 3010*

D50 ≥ 10 nm

D50 ≥ 7.6 nm

TSI 3025 D50 ≥ 3 nm

Table 4-1. A list of testing aerosols and instrumentation applied in this study. Note: TSI 3010* The asterisk (*) indicates that the temperature difference between saturator and condenser of the CPC was increased into 21˚C to lower the D50 value to 7.6 nm (Mertes et al. 1995).

145

4.2.1 Aerosol generation

Polydisperse NaCl, printer toner, citric acid and dioctyl sebacate (DOS)

particles were generated by three different methods (see Figure 4-1 a-c).

a) The NaCl, printer toner (note that an additional method was also used to

generate toner particles, see below) and citric acid particles were generated using the

tube furnace by heating up the materials to 680 ˚C for NaCl, 390 ˚C for toner and

175 ˚C for citric acid, respectively. The vaporized materials then condense to form

nanoparticles which follow the air stream to the particle measurement modules.

b) DOS aerosol was generated by a Condensation Monodisperse Aerosol

Generator (CMAG), Model: TSI 3475, operating in homogenous nucleation mode by

heating up the DOS inside the saturator to vaporise it and then passing through a

condensation tube to form homogeneous DOS nano-particles.

c) In addition to the toner aerosol generated by the tube furnace, toner aerosol

generated during laser printing was also examined. The commercial laser printer was

installed inside the air chamber and the CPC measurements were conducted during

the printing process. The laser toner particle generated during the printing process

was used to simulate the actual situation of indoor emission by the laser printer.

146

Figure 4-1. Systematic diagram of the experimental. Left: aerosol generation; right: aerosol detection.

4.2.2 Particle measurement

After the particles were generated, the polydisperse particles were passed

through an Electrostatic Classifier (EC), Model: TSI 3080 (with Nano-DMA TSI

3085)/ TSI 3071 (with Long-DMA 3081). The polydisperse aerosol size was

classified by the EC according to the electrical mobility of the neutralised singly

charged particles to form a monodisperse aerosol for which the count median

diameter (CMD) could be selected by choosing the corresponding EC voltage. The

aerosol was then split into two streams, one going to the WCPC (Model: TSI 3781 or

TSI 3782), the other to the BCPC (Model: TSI 3010, TSI 3010* or TSI 3025), to

measure the particle number concentrations. Before the experiment, all the flow

147

paths of EC and the CPCs were calibrated using a bubble flow meter to ensure that

the flow rate was within instrument specification and the particle number

concentrations were corrected according to the actual flow rate of the CPCs. The

lengths of tubing used between the instruments were minimised to reduce deposition

losses and the tubing between the EC and each CPC was in each case the same (~ 40

cm).

This investigation focuses on the relative performance of the CPC’s. Hence the

comparison between the BCPC and WCPC reflects their relative responses and

neither instrument is assumed to be 100% efficient in detecting particles of a given

type. The sampling period for each sample was 5 mins (with data being logged every

second this yielded a total of 300 data points for each test) and a further 5 mins was

allowed for the monodisperse aerosol concentration to stabilize after changing the

EC voltage to a new particle size.

4.3 RESULTS AND DISCUSSION

4.3.1 BCPC/WCPC ratios for different aerosol compositions and particle sizes

Number concentrations of different particle types were measured by BCPC and

WCPC and the ratios BCPCi/WCPCj (where the indices i and j: represent models of

CPC) were then calculated. These BCPC/WCPC ratios were used to represent the

relative detection abilities of the CPCs. For example a higher BCPC/WCPC value

indicated that the BCPC had a higher detection efficiency than the WCPC for the

tested aerosol. In the following section, the results will be presented in two parts. In

part i) we consider the effect of particle composition and in part ii) the influence of

particle concentrations on the relative detection efficiency of the CPC’s.

148

Figure 4-2 shows the BCPC/WCPC ratios for different aerosols. The

combination of the BCPC and WCPC used for NaCl, DOS and toner were TSI 3010*

(BCPC) and TSI 3781 (WCPC). For Citric acid, TSI 3025 (BCPC) and TSI 3781

(WCPC) were used. Standard deviation calculated by the 300 data points was

represented by the error bar. The toner aerosols generated by two methods are

denoted TonerF (when generated by the tube furnace) and TonerC (when generated by

the printer in the chamber).

Figure 4-2. BCPC/WCPC ratios when exposed to different sizes of NaCl, DOS, toner and citric acid particles.

4.3.1.1 Hygroscopic - NaCl

BCPC3010*/WCPC3781 ratios for particles smaller than 8 nm were all below 0.79

(± 0.01–0.06). The discrepancy between the BCPC and WCPC was less for particles

in the range 12 – 100 nm. BCPC3010*/WCPC3781 ratios for these larger NaCl particles

ranged from 0.92 to 0.96 (±0.01-0.02). This result is comparable to that by Hering et

al. (2005) who found an enhanced response for NaCl in the WCPC for particles

smaller than 12 nm. The authors suggest that the water uptake effects of salt aerosol

149

in the conditioner reduced the diffusional loss of the aerosol which is greater for

smaller particles. In addition, the authors also noted that the reduction of equilibrium

vapour pressure of water above salt particle lowers the activation diameter for

droplet growth (Hering et al., 2005).

4.3.1.2 Hydrophobic - DOS and toner

For DOS particles, the WCPC3781 significantly undercounted at 10 nm

(BCPC3010*/WCPC3781: 2.7 ±1.7). The overall trend of the ratio was to approach unity

for particle sizes ≥ 12 nm. A similar result was found by Hering et al. (2005) who

showed that the D50 for a WCPC (Model: TSI 3785) for DOS particles was

approximately 10 nm for a contaminated DOS sample and 30 nm for a new sample.

The “contaminated” sample referred to a DOS sample which had been stored in the

nebuliser overnight and the “new” sample referred to one freshly decanted into a

clean nebuliser. The major difference of the DOS generation methods is the size

distribution of the aerosol that the DOS particles produced by the evaporator tends to

have smaller sizes while larger DOS particles will be generated by nebuliser method.

For toner particles, the result presented in Figure 4-2 shows that the

WCPC3781 is incapable of detecting toner particles at diameter below 18 nm. The

BCPC3010*/WCPC3781 ratios for the tube furnace and chamber methods of toner

particle generation were 4.2±0.5 at 18 nm and 9.6±2.6 at 15 nm. These ratios

decreased for particles size larger than 20 nm, ranging from 1.70 down to 1.19 for

TonerF and from 1.31 down to 1.21 to for TonerC. This result was not unexpected as

the particles emitted by the printing process of the laser printer have been shown to

be hydrophobic (Morawska et al., 2009b) and are believed to contain aromatic

150

organic compounds such as toluene, ethylbenzene, xylene and styrene oligomers

which are water insoluble compounds. These results are consistent with water

insoluble species having difficulty activating the condensation process with water

vapour.

4.3.1.3 Citric Acid

Citric acid is an organic acid which is a solid at room temperature. It dissolves

both in water and in butanol. Less variability of the BCPC3025/WCPC3781 ratio was

obtained for citric acid compared to that for the other aerosols. The ratio for 8 nm

particles was 1.20±0.27 and this showed larger variability than that for particles

larger than 20 nm (1.07 – 1.19 ± 0.06-0.14). Since citric acid dissolves readily both

in water and butanol, the particles should readily undergo condensational growth by

both water and butanol. However these CPC’s have somewhat different nominal D50

values. The BCPC3025 has a lower D50 (3 nm) value than WCPC3781 (D50 = 6 nm) so

that a higher count for smaller particles is to be expected. The slightly higher

BCPC/WCPC ratios observed for particles large than 20 nm is probably due to

differences in the slope of the detection efficiency curves of the two instruments. For

example the WCPC3781 is claimed to have a detection efficiency of 50% at 6 nm

rising to 90% at 20 nm while the WCPC3025 detection efficiency is claimed to rise

rapidly from 50% at 3 nm to 90% at 5 nm.

From the results of NaCl and citric acid, the WCPC showed a capability to

detect those water soluble particles even for particles smaller than 10nm. These

results could be due to the hygroscopic properties of the aerosols as this would

enhance water uptake and so increase the size of otherwise undetectable small

151

particles beyond the Kelvin diameter in the CPC’s condenser so that they are then

able to grow to detectable sizes.

4.4 The effect of particle concentration

Figure 4-3 shows the BCPC/WCPC ratios obtained with different sizes of

NaCl particles in three concentration ranges, a) low: Cn < 1000 cm-3, b) medium:

1000 < Cn < 10000 cm-3 and c) high: Cn > 10000 cm-3. In this experiment, the TSI

3010 and TSI 3782 were used to represent the BCPC and WCPC because these two

CPCs have identical nominal D50 values of 10 nm. Irrespective of the concentration,

the BCPC/WCPC ratios decline sharply as the particle size is reduced below 14 nm.

This result is consistent with the explanation given in Section 4.3.1.1 which shows

that the BCPC/WCPC ratios were lower for small NaCl particles due to the enhanced

response to hygroscopic particles at very small sizes in the WCPC.

152

Figure 4-3. BCPC/WCPC ratio as a function of NaCl particle concentration.

Average BCPC/WCPC ratios for particles sizes in the range 6.7 – 100 nm

were 0.96 (±0.15), 0.90 (±0.15) and 0.78 (±0.14) for the low, medium and high

concentrations ranges respectively. These differences could probably be due to

coincidence errors which tend to occur in all CPCs at high concentrations. Under

normal operation, no more than one particle is assumed to be present in the

measuring volume at any given time. However a coincidence error will occur under

conditions of high particle concentration whenever more than one particle enters the

measuring volume at the same time and this will result in undercounting (Sachweh et

al., 1998). According to the instrument manuals, the upper concentration limit of the

BCPC3010 is 1 x 104 cm-3 (±10% accuracy) while that of WCPC3782 is 5 x 104 cm-3

(±10% accuracy). The lower upper limit in the case of the BCPC3010 is therefore the

most likely reason for the decreasing BCPC/WCPC ratio at high concentrations. The

dependence on particle size of the BCPC/WCPC ratio for low and medium

concentration ranges was similar, but for the high concentration range the

BCPC/WCPC ratio fluctuated when the particle sizes was in the range 10-20 nm.

When comparing the low and medium concentration range results, the variation in

the differences for particles in the size range from 6.7 to 100 nm ranged from 2% to

10%. Because the relationship between coincidence error and particle number

concentration is non-linear (Hermann and Wiedensohler, 2001), higher rates of

coincidence error occurred close to the CPC’s upper concentration limit. Therefore,

the coincidence error rate would have been greater for the BCPC3010 than for the

WCPC3782 when exposed to the same particle concentration. The observed variation

153

of the WCPC/BCPC ratios at low and medium concentrations therefore appears to be

consistent with these expected different rates of coincidence error.

To further evaluate the influences of particle number concentration on the

detection performance differences between BCPC and WCPC we calculated the

Pearson’s correlation coefficient (r) for a possible correlation between the

BCPC/WCPC ratio and the challenge aerosol particle number concentration for

various particle sizes of NaCl (see Figure 4-4). Particle number concentrations

below 1000 cm-3 were used in this analysis in order to minimize the effects of

coincidence error. Weak correlations between BCPC/WCPC and number

concentration were found for various particle sizes. The |r| values ranged from 0.03

to 0.77 with an averaged value of 0.31±0.21 (with p values < 0.01). The signs of

slopes of the ratio versus particle concentration curves (these curves are not shown)

varied in an apparently random manner between positive and negative values. This

lack of any clear consistent relationship between the correlation coefficient and

particle size implies that the relative detection performance of the CPCs does not

depend strongly on particle number concentration provided the concentration is not

so great as to produce significant coincidence error in either CPC.

154

Figure 4-4. Product of Pearson’s correlation coefficient, r, between BCPC/WCPC ratios and particle concentration varies with particles sizes.

4.5 CONCLUSION

In this study, number concentration measurements of NaCl, DOS, laser toner

and citric acid particles were tested by using BCPC and WCPC. The BCPC/WCPC

ratios for each type of particles were used to examine the relative detection

efficiencies of the CPCs.

The effect of water insoluble composition seems to be, to suppress the activation of

condensation for water vapour in the WCPC. Hydrophobic aerosol smaller than 20

nm were significantly undercounted by the WCPC and this undercounting was

reduced for particles larger than 20 nm as indicated by a decrease of the

BCPC/WCPC ratio for DOS from 1.2 down to 0.98, and for laser toner particles

from 1.70 down to 1.19 (TonerF) and from 1.31 down to 1.21 (TonerC). For NaCl,

the BCPC/WCPC ratio was close to unity for particle with diameters larger than 12

nm. This result suggests that the hygroscopic nature of NaCl, by reducing the

equilibrium vapour pressure above the particle surface, effectively acts to lower the

155

activation diameter for droplet growth in such particles and this produces a

corresponding improvement in the lower detection size limit.

For citric acid particles larger than 8 nm, the BCPC/WCPC ratio varied within a

narrow range (1.07 – 1.20). Although citric acid particles were only slightly

undercounted by the WCPC compared to the BCPC, the WCPC was unable to detect

smaller citric acid particles close to the instruments nominal lower diameter limit.

The results presented showed that the detection capability of CPCs is strongly

dependant on the solubility of the target aerosol in the condensation medium. In

addition, BCPC/WCPC ratios were obtained for a range of particle number

concentrations. Apart from the effects of coincidence errors which occur in all CPCs,

the results did not show a strong correlation between the BCPC/WCPC ratio and the

particle number concentration. According to the findings of this study, the BCPC

should be chosen when for measuring ambient concentrations of water insoluble

particles. Both the BCPC and WCPC appear to be suitable for ambient particle

concentration measurements provided that the particles are known to be at least

somewhat hygroscopic. WCPC’s should be used with care in ambient measurements

and only after first verifying that the aerosol is not hydrophobic. In addition, the

discrepancies of the detection abilities between both the BCPC and WCPC can be

applied to the study of the composition of particles.

Acknowledgements

This project was supported by the Australian Research Council and Queensland

Transport through Linkage Grant LP0882544.

156

4.6 References

Biswas, S., Fine, P., Geller, M., Hering, S., Sioutas, C., 2005. Performance

Evaluation of Recently Developed Water-Based Condensation Particle Counter.

Aerosol Science and Technology, 39:5, 419-427.

Cheung, H.C., Morawska, L., Ristovski, Z.D., 2010. Observation of new particle

formation in subtropical urban environment. Atmospheric Chemistry and Physics

Discussion, 10, 22623-22652.

Hämeri, K., O’Dowd, C.D., Hoell, C., 2002. Evaluating measurements of new

particle concentrations, source rates, and spatial scales during coastal nucleation

events using condensation particle counters. Journal of Geophysical Research,

107:D19, 8101-8111.

Haynes, W.M., ed., 2011. CRC Handbook of Chemistry and Physics, 91th Edition

(Internet Version 2011), CRC Press/ Taylor and Francis, Boca Raton, FL.

He, C., Morawska, L., Taplin, L., 2007. Particle Emission Characteristics of Office

Printers. Environmental Science and Technology, 41:17, 6039-6045.

Hermann, M., Wiedensohler, A., 2001. Counting efficiency of condensation particle

counters at low-pressures with illustrative data from the upper troposphere.

Journal of Aerosol Science, 32, 975-991.

Hering, S.V., Stolzenburg, M.R., 2005. A Method for Particle Size Amplification by

Water Condensation in a Laminar, Thermally Diffusive Flow. Aerosol Science

and Technology, 39, 428-436.

Hering, S.V., Stolzenburg, M.R., Quant, F.R., Oberreit, D.R., Keady, P.B., 2005. A

Laminar-Flow, Water-Based Condensation Particle Counter (WCPC). Aerosol

Science and Technology, 39, 659-672.

Iida, K., Stolzenburg, M.R., McMurry, P.H., Smith, J.N., Quant, F.R., Oberrit, D.R.,

Keady, P.B., Eiguren-Femandez, A., Lewis, G.S., Kreisberg, N.M., Hering, S.V.,

2008. An Ultrafine, Water-Based Condensation Particle Counter and its

Evaluation under Field Conditions. Aerosol Science and Technology, 42:10, 862-

871.

157

Johnson, G.R., Ristovski , Z., Morawska, L., 2004. Method for measuring the

hygroscopic behaviour of lower volatility fractions in an internally mixed aerosol.

Journal of Aerosolo Science, 35, 443-455.

Lee, S.C., Lam, S., Ho, K.F., 2001. Characterization of VOCs, ozone, and PM10

emissions from office equipment in an environmental chamber. Building and

Environment, 36, 837-842.

Mertes, S., Schröder, F., Wiedensohler, A., 1995. The Particle Detection Efficiency

Curve of the TSI-3010 CPC as a Function of the Temperature Difference between

Saturator and Condenser. Aerosol Science and Technology, 23, 257-261.

Morawska, L., He., C, Johnson, G., Jayaratne, R., Salthammer, T., Wang, H., Uhde,

E., Bostrom, T., Modini, R., Ayoko, G., McGarry, P., Wensing, M., 2009a. An

investigation into the characteristics and formation mechanisms of particles

originating from the operation of laser printers. Environmental Science and

Technology, 43, 1015-1022.

Morawska, L., Ristovski, Z., Jayaratne, E.R., Keogh, D.U., Ling, X., 2008. Ambient

nano and ultrafine particles from motor vehicle emissions: Characteristics,

ambient processing and implications on human exposure. Atmospheric

Environment, 42, 8113-8138.

Morawska, L., Wang, H., Ristovski, Z., Jayaratne, E.R., Johnson, G., Cheung, H.C.,

Ling, X., He, C., 2009b. Environmental monitoring of airborne nanoparticles.

Journal of Environmental Monitoring, 11:1758-1773.

Mordas, G., Manninen, H.E., Petäjä, T., Aalto, P.P., Hämeri, K., Kulmala, M., 2008.

On Operation of the Ultra-Fine Water-Based CPC TSI 3786 and Comparison with

Other TSI Models (TSI 3776, TSI 3772, TSI 3025, TSI 3010, TSI 3007). Aerosol

Science and Technology, 40, 1090-1097.

Rader, D.J., McMurry, P.H., 1986. Application of the tandem differential mobility

analyzer to studies of droplet growth or evaporation. Journal of Aerosol Science,

17:5, 771-787.

158

Sachweh, B., Umhauer, H., Ebert, F., Buttner, H., Friehmelt, R., 1998. In situ optical

particle counter with improved coincidence error correction for number

concentration up to 107 particle cm-3. Journal of Aerosol Science, 29, 1075-1086.

159

CHAPTER 5

Observation of new particle formation in subtropical urban environment

H.C. Cheung, L. Morawska and Z.D. Ristovski

International Laboratory for Air Quality and Health, Queensland University of

Technology

GPO Box 2434, Brisbane QLD 4001, Australia

Published by the Journal of Atmospheric Chemistry and Physics

160

STATEMENT OF JOINT AUTHORSHIP

Title: Observation of new particle formation in subtropical urban environment.

Authors: H.C. Cheung, L. Morawska*, Z.D. Ristovski

H.C. Cheung

Designed and developed the methodology, conducted the field measurement,

analysed and interpreted the data, and wrote the manuscript.

L. Morawska

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

Z.D. Ristovski

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

161

CHAPTER 5. OBSERVATION OF NEW PARTICLE FORMATION IN

SUBTROPICAL URBAN ENVIRONMENT.

H.C. Cheung, L. Morawska and Z.D. Ristovski

International Laboratory for Air Quality and Health, Queensland University of

Technology, GPO Box 2434, Brisbane QLD 4001, Australia

Abstract

The aim of this study was to characterise the new particle formation events in a

subtropical urban environment in the southern hemisphere. The study measured the

number concentration of particles and its size distribution in Brisbane, Australia

during 2009. The variation of particle number concentration and nucleation burst

events were characterised as well as the particle growth rate which was first reported

in urban environment of Australia. The annual average NUFP, NAitken and NNuc were

9.3 x 103, 3.7 x 103 and 5.6 x 103 cm-3, respectively. Weak seasonal variation in

number concentration was observed. Local traffic exhaust emissions were a major

contributor of the pollution (NUFP) observed in morning which was dominated by the

Aitken mode particles, while particles formed by secondary formation processes

contributed to the particle number concentration during afternoon. Overall, 65

nucleation burst events were identified during the study period. Nucleation burst

events were classified into two groups, with and without particles growth after the

burst of nucleation mode particles observed. The average particle growth rate of the

nucleation events was 4.6 nm hr-1 (ranged from 1.79 – 7.78 nm hr-1). Case studies of

162

the nucleation burst events were characterised including i) the nucleation burst with

particle growth which is associated with the particle precursor emitted from local

traffic exhaust emission, ii) the nucleation burst without particle growth which is due

to the transport of industrial emissions from the coast to Brisbane city or other

possible sources with unfavourable conditions which suppressed particle growth and

iii) interplay between the above two cases which demonstrated the impact of the

vehicle and industrial emissions on the variation of particle number concentration

and its size distribution during the same day.

163

5.1 Introduction

Understanding the formation process of atmospheric particles is vital because of the

significant impact of particulate matter on human health and climate change

(Charlson et al., 1992, Donaldson et al., 1998). Atmospheric particles can be formed

by nucleation process via a number of different mechanisms (e.g. Kulmala 2003;

Kulmala et al., 2004), such as binary nucleation (involving H2SO4 and water vapour),

ternary nucleation (involving NH3, H2SO4 and water vapour) and ion-induced

nucleation for charged particles, depending on the environmental conditions. To date,

numerous studies have been conducted in different locations, in order to investigate

particle formation processes in different environmental settings, including the free

troposphere (e.g. Weber et al., 2001), boreal forests (e.g. Vehkamäki et al., 2004) and

coastal areas (e.g. O’Dowd et al., 1999; Lee et al., 2008). However, most of these

studies focused on particle formation in rural settings and in colder climates, with

very few studies conducted in urban environments, especially in the southern

hemisphere (Kulmala et al., 2004). A limited number of studies were conducted in

continental (e.g. Woo et al., 2001; Moore et al., 2007; Wu et al., 2008) and coastal

(Pey et al., 2008; Rodríguez et al., 2008; Fernández-Camacho et al., 2010; Pérez et

al., 2010) urban areas. These studies examined the variation of particle number

concentration in urban environments. The major influence on particle number

concentration was vehicle exhaust emissions during the traffic peak hours (e.g. Pey

et al., 2008; Pérez et al., 2010) and new particle formation by photochemical

reactions (e.g. Pey et al., 2009), as well as the influence of power plant and industrial

emissions from an area upwind from the urban site (Gao et al., 2009). The few

examples include studies on particle formation associated with natural emissions

from a Eucalypt forest in South-East Australia (Ristovski et al., 2010; Suni et al.,

2008), which concluded that natural emissions were in fact a source of particle

164

formation. In addition, new particle formation was observed in the coastal area of

Eastern Australia (Johnson et al., 2005; Modini et al., 2009), the result of which

showed that new particles were formed by the condensation of sulphate and/or

organic vapours onto sulphate clusters to form an observable particle. Guo et al.

(2008) conducted a short-term intensive study on particle formation in the rural

environment of Eastern Australia, in which particle formation was suggested to be

influenced by the photochemical processes of the urban air plume. The findings of

Guo et al. (2008) provided an insight to the impact of urban pollution on nucleation

processes. For the urban environment, Mejía et al. (2009) characterised the

favourable atmospheric conditions for nucleation burst events in a coastal urban area

in Brisbane, which is the only nucleation study conducted in an urban area in the

southern hemisphere to date. The study showed that the nucleation events mostly

occurred during the summer and it also suggested cleaner air masses of a local origin

mixing with traffic exhaust emission after the events. However, Mejía et al. (2009)

did not investigate the nucleation growth process after the nucleation burst events,

and thus particle formation parameters, such as particle growth rate, are not available

for the urban environment in southern hemisphere. The particle growth rate is an

important factor for the calculation of climate forcing.

To further investigate the characteristics of the particle formation processes in a

subtropical urban environment, we conducted a one year-long measurement of the

size distribution of particles in the size range 4 – 110 nm, at an urban area of

Brisbane, Australia. The aim of this study was to characterise the temporal variation

of particle number concentration, and to explain the controlling factors that

influenced new particle formation processes in the subtropical urban environment.

165

5.2 Methodology

5.2.1 The topography and meteorology of Brisbane region

Brisbane is the capital city of the state of Queensland, Australia, located at 27’30oS

and 153oE. Brisbane city is surrounded by mountains from south to north, and faces

the Pacific Ocean to the East. It is the fastest growing urban region in Australia (2

million inhabitants). The major pollution sources affecting the CBD region are traffic

exhaust emissions generated in the inner city, and aircrafts, ships and industrial

emissions transported from the lower reaches of Brisbane River, approximately 15-

18 kms NE of the CBD. The Brisbane River meanders through the Brisbane region.

Morawska et al. (1998) provided a description of the wind patterns in the Brisbane

region, which are mostly governed by synoptic flows from the SE. A NE sea breeze

is also a daily feature throughout the year. In addition, an overnight SW drainage

flow from the mountain range to the West carries air parcels from the plateau region

and the Western coastal plain towards the city. On the rare occasion when gradient

winds are blowing from the NW, the combination of the light synoptic NW flow and

the overnight SW drainage flow can sufficiently delay the onset of the sea breeze to

cause recirculation of the city emissions, leading to photochemical smog events.

5.2.2 The QUT study site

The measurements were conducted at the International Laboratory of Air Quality and

Health (ILAQH), Queensland University of Technology (QUT), which is within the

CBD of Brisbane (Figure 5-1). The monitoring site was about 10 m above ground

level on the top floor of a QUT campus building, located to the SE of the city centre,

with a major highway (the Pacific Motorway) situated along the SW side of the

campus. Therefore, the pollution associated NE winds could be attributed to

166

industrial emissions (from the airport, oil refinery and Port of Brisbane), while the

pollution associated with S to NW winds could be attributed to local traffic exhaust

emissions.

5.2.3 Measurement techniques

The size distribution of ultrafine particles (UFPs) was measured at the QUT

monitoring site from 1st January to 31st December 2009. Particle size distribution in

the range 4 - 110 nm was measured by a Scanning Mobility Particle Sizer (SMPS)

system, which consisted of two parts, an Electrostatic Classifier (EC) (TSI 3080) and

a Condensation Particle Counter (CPC) (TSI 3781). The EC was equipped with a

nano-differential mobility analyser, which can separate the poly-disperse particles

into selected mono-disperse particles according to their particle mobility. The

number concentration of the mono-disperse particles was then counted by the CPC.

Each ambient sample was drawn into the SMPS system from outside the building

through a 0.635 cm (inner diameter) conductive tube and a sampling duration of 5

mins was adopted for each particle size distribution sample. Multiple charge

correction was applied to the particle size distribution measurements by using an

internal algorithm from the Aerosol Instrument Manager Software.

In this study, the size distribution data was classified into three groups, i) UFPs,

including particles ranging from 4 - 110 nm (NUFP); ii) Aitken mode with particles

(NAitken), which ranged from 30 - 110 nm; and iii) nucleation mode particles, which

were < 30 nm (Nnuc). In addition to the above particle measurements, meteorological

parameters, including wind speed and direction, temperature and relative humidity

(RH) were monitored at Kangaroo Point (1 km East of QUT) by the Queensland

Bureau of Meteorology. The QUT and Kangaroo Point sites were not blocked by

167

high rise buildings and therefore the use of wind data measured at Kangaroo Point

was representative of the synoptic wind direction of the study region. It should be

noted that global solar radiation was measured at the Queensland Environmental

Protection Agency site (Rocklea), about 10 km south of QUT.

Figure 5-1: Map of Brisbane.

5.2.4 Data processing and analysis

In this study, the raw particle size distribution and meteorological data were

synchronised into 10 min averaged data for data analysis and figure plotting.

According to Mejía et al. (2007) the lower limit of the particle size distribution

dataset was set to 1 cm-3. The upper limit was set to 5 x 105 cm-3. Some data were

removed from the database based on several criteria such as i) zero value for particle

concentration; ii) particle concentration higher than 5 x 105 cm-3 and iii) data

collected during instrument malfunction. During the one year measurement

campaign, 28 % of the data was removed based on the above data reduction

procedures and due to instrument maintenance. Correlations between the parameters

were tested using the Pearson-product moment correlations test with a 95%

168

confidence level (p < 0.05). The linearity of the tested parameters was indicated by

Pearson’s coefficient, r, with a perfect linear correlation between two parameters

indicated by an r value close to 1.

5.3 Results and Discussion

5.3.1 Overall results

The overall average concentration of ultrafine particles (NUFP), Aitken mode (NAitken)

and nucleation mode (Nnuc) measured in this study were 9.3 x 103 (±15.3 x 103), 3.7 x

103 (±5.1 x 103) and 5.6 x 103 cm-3 (±12.6 x 103), respectively. The values obtained in

this study are similar to those observed in similar environments in Northern Europe

(Hussein et al., 2004). The few studies conducted in Southern European countries

showed much higher concentrations than those which were reported in this study

(Pey et al., 2008, Rodríguez et al., 2008; Fernández-Camacho et al., 2010). Similar

values of NAitken and Nnuc were obtained in the urban areas of Helsinki, Finland,

which were 4.0 x 103 – 6.5 x 103 cm-3 and 5.5 x 103 – 7.0 x 103 cm-3, respectively

(Hussein et al., 2004). The NUFP measured in Brisbane was relatively lower than that

in other coastal urban areas, including the Yangtze River Delta, China (Gao et al.,

2009), Barcelona (Pey et al., 2008) and Huelva and Santa Cruz de Tenerife, Spain

(Rodríguez et al., 2008; Fernández-Camacho et al., 2010), which were 28.5 x 103,

14.2 x 103 and 22.0 – 26.3 x 103 cm-3, respectively.

The results of this study were also compared to those of a previous study conducted

in the Brisbane urban region from 1995 to 2000 (Mejía et al., 2007). The NUFP and

Nnuc measured in this study were about 8% and 60% higher than those measured by

Mejía et al. (2007), being 8.6 x 103 cm-3 (for particles in the size range 15 - 100 nm)

and 3.5 x 103 cm-3 (for particles in the size range 15 - 30 nm), respectively. In

169

relation to Nnuc, it should be noted that the nucleation mode particle concentration in

this study covered particles in the size range 4 - 30 nm, and therefore, it is expected

to be higher than the earlier result reported by Mejía et al. (2007).

The seasonal variation of particle number concentrations is depicted in Figure 5-2.

The Pearson’s coefficients, r, between particle number concentration of different

modes and temperature were calculated which ranged from 0.00 - 0.03, which

indicates that there was no statistical seasonal variation in particle concentrations.

This result is similar to that presented by Mejía et al. (2007) for the same study

region, however larger variations in particle number concentrations were observed in

each mode during the summer season, as reflected by the large interquartile range

(see Figure 5-2).

Figure 5-2: Monthly variations of (a) mean temperature, and particle number concentration of ultrafine (UFP) (b), Aitken mode (c) and nucleation modes particles (d). The median number concentrations and the 1st and 3rd quartiles are presented.

170

Figure 5-3 shows the diurnal variation of particle number concentration for different

modes with the diurnal variations of temperature and relative humidity also plotted.

Two peaks were observed for UFP during the day, the first of which occurred from

around 6 am to 8 am, possibly due to traffic exhaust emissions during the morning

peak hours (from around 6 am to 8 am) in Brisbane urban region (Mejía et al., 2007).

The second peak, which is much more important, was observed from around 12 noon

to 3 pm, and this may be due to the formation of new particles. During the period of

UFP morning peak, it was suggested that the Atiken mode particles contributed by

the direct diesel and petrol engine emissions, which produce particles in the size

range of about 20 – 130 nm and 20 – 60 nm, respectively (Morawska et al., 2008).

Also the nucleation mode particles could be formed during the dilution and cooling

of engine exhausted sulphuric and organic vapours by condensation onto sulphur

clusters (Meyer and Ristovski 2007). During the period of the second UFP peak, a

nucleation mode peak was also observed associated with highest level of solar

radiation, which implies that new particles were produced during the early afternoon

by photochemical reactions.

171

Figure 5-3: Diurnal variation of (a) averaged solar radiaiton, (b) averaged wind direction/speed, (c) averaged temperature and RH, and (d) averaged UFP, nucleation mode and Aitken mode particle concentrations.

5.3.2 Relationship between particle number concentrations and meteorological parameters

5.3.2.1 Temperature, relative humidity and solar radiation

From Figure 5-3 it can be seen that temperature and relative humidity display an

anti-correlation, whereby increases in temperature were associated with decreases in

relative humidity. The influences of temperature on particle number concentration

were not confirmed in previous studies. Some studies found that high particle

concentration was related to relatively high temperatures (e.g. Kim et al., 2002),

whilst others found that they were associated with relatively low temperatures (e.g.

Olivares et al., 2007). In this study, a weak correlation between particle number

concentrations and temperature was observed (r = 0.36 - 0.53; p < 0.01), however

higher variations in NUFP, NAitken and Nnuc were observed during warmer and lower

humidity conditions (Figures 5-4 and 5-5). The highest number concentrations of all

172

particle sizes were associated with temperature around 32˚C. The peaks of NAitken and

Nnuc observed in the early afternoon (under high temperature conditions) suggested

that the contribution by new particle formation processes was the greatest, followed

by particle growth to larger particles. In addition, another peak of NAitken was

observed with temperature around 10˚C (see Figure 5-4). Also higher NAitken

concentrations were observed under humid conditions (see Figure 5-5). This result

may be due to enhanced coagulation and condensation effects under high humidity

conditions.

In some cases, temperature data can not directly reflect the strength of photochemical

activities which occurred on warm cloudy days. In addition, condensation vapour

H2SO4 production was related to the solar radiation (Ristovski et al., 2010).

Therefore, solar radiation was used to indicate the reactivity of photochemical

reactions. The particle number concentration did not show a clear relationship with

the ambient temperature. In contrast, a positive relationship between particle number

concentration and solar radiation data was observed (r = 0.92-0.98; p < 0.01). This

result showed that the Nnuc was related to the photochemical reactions.

173

Figure 5-4: Particle number concentrations of (a) ultrafine (UFP), (b) Aitken mode and (c) nucleation mode particles and their variation as a function of temperature. The median number concentrations and the 1st and 3rd quartiles are presented.

Figure 5-5: Number concentrations of (a) ultrafine (UFP), (b) Aitken mode and (c) nucleation mode particles and their variation as a function of relative humidity. The median number concentrations and the 1st and 3rd quartiles are presented.

174

5.3.2.2 Wind direction and speed

Figure 5-6 shows the particle number concentration for different particle sizes under

different wind directions. For UFP, a sharp peak was associated with ENE wind

directions, and a lower broad peak was associated with SSE to WNW wind

directions. An interesting result was also obtained when dividing the data into Aitken

and nucleation modes. The sharp peak was observed in both of these two modes,

however the broad peak was only observed in the Aitken mode. From Figure 5-1, it

can be seen that the Brisbane Airport, oil refinery and Port of Brisbane were all

located to the NE of the monitoring site, whilst the CBD and Pacific Motorway were

located to the NW and SW of the monitoring site. Therefore, it is likely that the

nucleation mode particles were contributed by the industrial sources located to the

NE, while the Aitken mode particles were emitted from both industrial and vehicle

emission sources, as well as the coagulation/condensation of smaller particles under

humid conditions (see Section 5.2.1), which will contribute to the accumulation

mode. In addition, air masses blowing from the marine boundary (NE to SE

directions) were relatively clean. However, the inland air mass from the NE direction

was contaminated by industrial emissions. This interpretation can be supported by

higher NUFP in north-easterly air masses and lower NUFP in easterly or south-easterly

air masses (clean maritime air masses, which are thought to be much less loaded in

gaseous precursors). To better illustrate the directional dependence of the NUFP,

NAitken and Nnuc a wind rose plot of particle number concentration superimposed over

the location map is shown in the supplementary materials section (Figure 5-S1).

175

Figure 5-6: Number concentration of (a) ultrafine (UFP), (b) Aitken mode and (c) nucleation mode particles and their variation as a function of wind direction. The median number concentrations and the 1st and 3rd quartiles are presented.

In general, a negative correlation was observed between UFP concentration and wind

speed, indicated by a Pearson coefficient of r = -0.97 (p < 0.01). Higher particle

number concentration was associated with lower wind speeds (see Figure 5-7),

which can be explained by the stronger dispersion associated with high wind speeds

(Hussein et al., 2006). Similar results were also observed for Aitken and nucleation

mode particles. In addition, a larger variation of Nnuc was associated with the

moderate wind speed (~ 4 ms-1). Nnuc usually reached it’s daily peak value during

early afternoon and the corresponding wind speed was ~ 4 ms-1 (see Figure 5-3).

176

Figure 5-7: Number concentration of (a) ultrafine (UFP), (b) Aitken mode and (c) nucleation mode particles and their variation as a function of wind speed. The median number concentrations and the 1st and 3rd quartiles are presented.

5.3.3 Particle formation in subtropical urban atmosphere

5.3.3.1 Classification of nucleation events

The general definition of a nucleation event is a two-phase process involving the

burst of observed nucleation particles and the growth of these particles into

accumulation mode by condensation and/or coagulation (Kulmala et al., 2004). To

illustrate the nucleation event, a “banana” shape should be observed in the contour

plot of particle size distribution. The example of a nucleation event is shown in

Figure 5-8. Usually the lowest Nnuc is observed in the early morning, and then it

begins to increase at around 9 am. The geometric median diameter (GMD) of the

measured particles grows into an Accumulation mode during the day and the

evolvement of the particle size curve is often compared to a banana shape.

177

Figure 5-8: The time series of nucleation events observed at QUT on 21 October 2009. From bottom to top, the parameters are: i) Geometric median diameter (GMD) and contour plot of size distribution; ii) Particle number concentration of nucleation and Aitken mode particles; iii) Particle number concentration of ultrafine particles (UFP); iv) Temperature and relative humidity; and v) wind direction and speed.

A number of nucleation events were observed during this study, which were

classified into different groups (Class Ia/b and II) according to the classification

scheme developed by Dal Maso et al. (2005). Class Ia/b events are defined as those

events where the particle growth rate can be determined. A typical Class Ia event

demonstrates clear and strong particle formation events with little or no pre-existing

particles obscuring the observation of the newly formed mode, while a Class Ib event

is any other event where the particle growth rate can be determined. Class II events

are defined as events where the banana shape still observable, but the data fluctuates

to such an extent that formation rate calculation is impractical.

178

In the urban environment, nucleation burst events have been observed with and

without particle growth (Park et al., 2008; Gao et al., 2009). For example, in addition

to nucleation events, we occasionally observed increases in nucleation mode particle

concentration during the daytime, where the particles did not grow into larger

particles (indicated by the near constant GMD value during the event period). We

defined this kind of event as a nucleation burst event, and a total of 65 burst events

were identified. A more detailed discussion about the occurrence of nucleation

events, both with and without particle growth, is provided in Section 5.3.2 and 5.4

below.

5.3.3.2 Growth rate during nucleation events

During this study, there were several gaps in the dataset due to instrument

malfunction/maintenance. For example, if there were more than 3 hrs of missed data

between 8 am to 6 pm (the period during which the nucleation usually occurs), this

was not counted as a valid daily dataset. After the removal of invalid daily datasets, a

total of 252 days of data were counted.

Figure 5-9 shows the monthly averaged particle growth rate and solar radiation, as

well as the monthly occurrence of nucleation events. This data provided information

regarding the influence of photochemical activity on particle formation in Brisbane.

Higher particle growth rate was found to be associated with higher solar radiation

and the results showed a positive relationship (r = 0.76, p < 0.05) between the

particle growth rate of nucleation events and solar radiation. Similar findings were

obtained in previous studies, which showed that particle growth rates were associated

with the strength of solar radiation (e.g. Kulmala et al., 2004; Vehkamäki et al.,

2004). The number of nucleation events classified as Class Ia, Ib, and II were 4, 13,

179

and 23, respectively. The nucleation events (Class I and II) occurred throughout most

of the year, except in November, and only one Class II event was observed in

December. Although infrequent nucleation events were observed during November

and December, the nucleation bursts (without particle growth) were found to be

closely associated with NE wind directions. In addition to the seasonal variation of

temperature, the dominant wind direction measured during November and December

was different to other months. NE winds dominated during these warmer months,

while the main wind direction was from the SE-SW during other months. The

influence of wind direction on the nucleation events will be discussed in the case

studies below. The mean growth rate for the nucleation events was calculated by the

slope of GMD against time during the period of particle growth under 30 nm. The

growth rates of Class I events measured in this study ranged from 1.79 - 7.78 nm hr-1

(average 4.6 nm hr-1), which are comparable to other urban studies such as those

conducted in Atlanta (2.86 - 22.0 nm hr-1) (Woo et al., 2001) and East St. Louis

(average 6.7 nm hr-1) (Qian et al., 2007).

180

Figure 5-9: Seasonal variation in (a) particle growth rates and solar radiation and (b) number of class I event and the precentage ratio of class I event to total sampling days.

5.3.4 Case studies of nucleation burst events

In the above sections, it was shown that nucleation mode particle concentrations

were strongly associated with NE/SW winds. Further analysis of the daily variation

of particles and wind direction showed that nucleation events with particle growth

were usually associated with SW winds, while the nucleation burst events without

particle growth were associated with NW winds. In this section, case studies relating

to types of three nucleation events are discussed, including: i) new particle formation

by nucleation processes; ii) a nucleation burst without particle growth; and iii) the

interplay between these two situations.

5.3.4.1 Case I - Photochemical formation of nucleation particles

Case I nucleation events were observed during 28 - 29 April 2009. Significant strong

nucleation bursts were observed consecutively during these two days (peak 10 min

181

and hourly averaged Nnuc during the nucleation events were 47 x 104 and 18 x 104

cm-3, respectively). The time series plots of the particle size distribution and

meteorological parameters are illustrated in Figure 5-10.

Figure 5-10: The nucleation events observed on 28-29 April 2009. From bottom to top, the parameters are: i) Geometric median diameter (GMD); ii) Particle number concentration of nucleation and Aitken mode particles; iii) Particle number concentration of ultrafine particles (UFP); iv) Temperature and relative humidity; v) Solar radiation; and vi) wind direction and speed.

During these two days, the highest temperature was about 30 ºC and relative

humidity was around 20-40 % at noon. Land and sea breeze wind circulation was

observed on both days, with a moderate (~ 4 ms-1) SW wind (from inland)

dominating in the morning and a moderate NE wind (from the coast) dominating in

the afternoon. The variation in concentration of nucleation and Aitken mode particles

clearly showed the influence of the nucleation burst on particle number

182

concentrations. In the early morning (6 - 9 am) of 28 April 2009, an Aitken mode

peak was observed, which could be attributed to traffic exhaust emissions during the

morning peak hours. At around 10 am, a sharp peak of nucleation mode particles was

observed and the GMD reached the lowest value of the day (~ 8 nm). The wind

direction changed to NE at around 1 pm, and turned back to the SW again after

midnight. In terms of GMD, the nucleation mode particles were growing into larger

particles (GMD increased from 8 to 57 nm) until around midnight (~3 am on 29

April 2009). Another nucleation growth event was observed the following day, on 29

April 2009, with similar meteorological conditions to those which were observed on

the previous day. However, the concentration of nucleation particles during the

nucleation burst was lower than that observed on the previous day. This result

indicated that the higher number of pre-existing Aitken mode particles in the

atmosphere served to diminish the nucleation processes.

5.3.4.2 Case II - nucleation burst without growth into larger particles

A Case II nucleation event was observed on 11 November 2009. As shown in Figure

5-11, the variation in wind direction was similar to that observed during the Case I

nucleation events, whereby land and sea breeze circulation was observed, however,

the burst of nucleation particles did not appear with the SW wind (associated with

local traffic exhaust emissions). Instead, a nucleation burst was observed with the NE

wind at ~ 10 am. The GMD dropped from ~ 30 nm to 10 nm and Nnuc increased from

~ 7.0 x 103 to 10.0 x 104 cm-3 during the nucleation burst, while NAitken did not show

any significant variation, ranging from 2 - 5 x 103 cm-3. This plume disappeared at

around 6 pm and GMD rose to ~ 25 nm. Based on these findings, it was suggested

that the plume was not directly emitted from the local traffic exhaust emission or

183

ship emission from the Port of Brisbane, since the particles from vehicle and ship

emissions are in the range 20 -130 nm (Morawska et al., 2008) and 60 – 120 nm

(Sinha et al., 2003), respectively. However, the emissions of SO2 and VOCs from the

industrial sources located at the coast could be possible precursors to the formation

of new particles by nucleation process. Another possible source of this plume was

aircraft emissions from the Brisbane Airport, which was located in a NE direction

from the study site. Mazaheri et al. (2009) measured the particle size distribution

produced by commercial aircraft at Brisbane Airport and a very distinct peak of

nucleation mode particles was observed at around 15 nm. This result was comparable

to the average GMD measured during the nucleation burst events in this study, which

was 14 nm (ranging from 8 - 32 nm). In addition, the nucleation burst could be due to

the precursors of local emissions which were similar to that in the nucleation growth

event or the re-circulated aged plumes belonging to land and sea breeze; however the

particle growth process could have been suppressed due to unfavourable conditions,

the exact nature of which is not known.

184

Figure 5-11: The nucleation bursts measured on 11 November 2009. From bottom to top, the parameters are: i) Geometric median diameter (GMD); ii) Particle number concentration of nucleation and Aitken mode particles; iii) Particle number concentration of ultrafine particles (UFP); iv) Temperature and relative humidity; v) Solar radiation; and vi) wind direction and speed.

5.3.4.3 Case III - Interplay between new particle formation and nucleation burst events

A Case III nucleation event was observed on 15 March 2009. During this study, it

was found that particle formation via nucleation processes was associated with SW

winds (Figure 5-10) and the subsequent particle growth (banana shape of particle

size distribution) was usually followed by the presence of a NE wind. In contrast, the

nucleation burst events without particle growth were most commonly related to

emission sources from the NE (Figure 5-11). In some events, a partial banana shape

was observed with a SW wind in the morning, but the observation of particle growth

was interrupted by a nucleation burst plume from the airport region (Figure 5-12).

185

Figure 5-12: Contour plot of particle size distribution observed on 15 March 2009. From bottom to top, the parameters are: i) Geometric median diameter (GMD) and contour plot of size distribution; ii) Particle number concentration of nucleation and Aitken mode particles; iii) Particle number concentration of ultrafine particles (UFP); iv) Temperature and relative humidity; v) Solar radiation; and vi)wind direction and speed.

It can be seen that the nucleation process commenced at 9 am and the particles kept

growing into Aitken mode particles until 12 pm (GMD rose from 20 to 35 nm). After

12 pm, the wind direction changed to a NE wind and an air plume enriched with

nucleation mode particles was observed, which interrupted the observation of a

banana shaped progression of the particle size distribution curve. After the

interruption, the GMD dropped suddenly to below 20 nm and several similar cases

were also observed during the one year study period at QUT. Overall, the results

showed that the nucleation mode particles originated from a variety of sources such

as traffic exhaust emission in Brisbane CBD and industrial emissions located NE to

Brisbane. Although the observation of a banana shape was interrupted by another air

mass, the particle growth process could continue in other regions.

186

5.3.4.4 Source identification

Gaseous data measured at Pinkenba, which is located near the lower reaches of the

Brisbane River (close to the airport, oil refinery and port of Brisbane) and South

Brisbane (about 1km south to QUT) were used to conduct source analysis. These

gaseous measurements were conducted by the Queensland Environmental Protection

Agency. Also back-trajectories of the nucleation growth/ burst events were

calculated using the HYSPLIT model (Hybrid Single Particle Lagrangian Integrated

Trajectory, Version 4.9), in order to trace the origin of the air masses. It should be

noted that the grid resolution of the meteorological data used for back-trajectories

calculation was 1 x 1 degrees in latitude and longitude. The data resolution is not

accurate enough to trace the detailed air mass passage over the scale of this study

region, and therefore, it only provides an indication from which region the air mass

comes from.

The gaseous data available for Pinkenba included CO, NO2, and SO2, while only CO

and NO2 data were available for South Brisbane. The emission ratios of CO/NO2 and

SO2/NO2 were calculated. On average, the daily minimum of each gaseous species,

representing the background value, almost reached zero in our study region.

Therefore we did not subtract the background data for the emission ratio calculations,

as it was negligible. 48-h back trajectories were calculated for the first two sampling

hours of each event (see supplementary figures 5-S2 and 5-S3) and the average

CO/NO2 and SO2/NO2 concentrations measured at Pinkenba during the event period

were 89.7 and 0.57, respectively. Overall, the CO/NO2 ratio exceeded the ratios

reported in the 2008/2009 National Pollution Inventory (from www.npi.gov.au,

accessed on 15 January 2011) for other sources, such as vehicles (9.7), oil refineries

(6.4), ships (0.69) and wildfires (24.6). If the pollution plume was contributed by

187

single source, it was possible to identify the emission source by comparing these

emission ratios. For example, the ratio for SO2/NO2 (0.57) was very close to the ship

emission ratio of 0.69. Although back-trajectory analysis found that almost all

trajectories originated from the NE sector during the nucleation burst events, air

masses from the NE were influenced by a number of different sources, such as ship,

aircraft, oil refinery and the local vehicle emissions. Therefore, it was difficult to

identify the specific source/s which contributed to the nucleation burst events. In

addition, primary pollution plumes (e.g. CO and NO2) were observed at Pinkenba 1-3

hrs prior to the start of the nucleation burst events. From these results, we can

conclude that the nucleation burst events were most likely influenced by industrial

emissions from the area NE of the sampling site. As mentioned in Section 5.4.2, the

nucleation burst event could be associated to other possible sources, however the

particle growth process could have been suppressed by unfavourable conditions, the

exact nature of which is not known

For nucleation growth events, the CO/NO2 ratio obtained from South Brisbane was

10.2, which is close to the emission inventory data for vehicles (9.7). Back-trajectory

analysis also showed that the air masses originated from S-SW directions, except on

21/10/2009, which suggests that vehicle exhaust emissions contributed to the

nucleation growth event.

5.4 Conclusion A year long measurement campaign of the size distribution of ultrafine particles was

conducted at subtropical urban area of Brisbane, Australia during 2009. The annual

average NUFP, NAitken and Nnuc were 9.3 x 103, 3.7 x 103 and 5.6 x 103 cm-3,

respectively. Small seasonal variation in number concentration was observed, with

188

higher particle concentrations observed during the warmer months. Diurnal variation

of NUFP, NAitken and Nnuc showed the influence of local traffic exhaust emissions on

the particle number concentration during morning peak hours, and elevated

nucleation mode particle levels suggested the contribution of new particle formation

during the early afternoon. In relation to wind direction, NAitken and Nnuc were

associated with NE winds, which pointed to the emission sources present at the lower

reaches of Brisbane river (such as Brisbane Airport, the oil refinery and the Port of

Brisbane). A broad peak of NAitken particles was also observed during SSE to WNW

winds, which suggested the influence of local traffic exhaust emissions, with particle

size ranging from 30 - 70 nm. Overall, two major sources of Nnuc were identified in

this study, which were new particle formation by nucleation and primary nucleation

mode particles emitted from aircraft at Brisbane Airport. New particle formation via

nucleation process was frequently observed in this study, and the particle growth rate

(average 4.6 nm hr-1) was positively related to the strength of global solar radiation.

The nucleation events with particle growth were associated with SW winds which

suggested the influence of precursors emitted from traffic exhaust emission in

Brisbane city. An interesting question arose during the course of this study regarding

the absence of nucleation particle growth in air masses originating from the coast. It

may due to lack of nucleation particles precursor associated with coastal air mass. To

tackle this question, a further study on the new particle formation, with parallel

measurements of particle chemical composition and gaseous pollutants, is needed.

Acknowledgements This project was supported by the Australian Research Council and Queensland

Transport through Linkage Grant LP0882544. We would also like to thank the

Queensland Bureau of Meteorology for providing the meteorological data.

189

Supplmentary materials

Figure 5-S1. Wind rose plot of NUFP, NAitken and Nnuc superimposed over the location map. The origin of the wind rose is located at QUT. NUFP - red line; NAitken – blue line; Nnuc – green line. Unit in cm-3.

Figure 5-S2. Back-trajectories calculated during the nucleation burst events.

190

Figure 5-S3. Back-trajectories calculated during the class I nucleation growth events

191

5.5 References

Charlson, R.J., Schwartz, S.E., Hales, J.M., Cess, R.D., Jr. Coakley, J.A., Hansen,

J.E., Hofmann, D.J.: Climate forcing by anthropogenic aerosols, Science, 255,

423-430, 1992.

Dal Maso, M., Kulmala, M., Riipinen, I., Wagner, R., Hussein, T., Aalto, P.P.,

Lehtinen, K.E.J.: Formation and growth of fresh atmospheric aerosols eight

years of aerosol size distribution data from SMEAR II, Hyytiälä, Finland, Boreal

Environment Research, 10, 323-336, 2005.

Donaldson, K., Li, X.Y., MacNee, W.: Ultrafine (nanometre) particle mediated lung

injury, Journal of Aerosol Science 29, 553-560, 1998.

Fernández-Camacho, R., Rodríguez, S., de la Rosa, J., Sánchez de la Campa, A.M.,

Viana, M., Alastuey, A., Querol, X.: Ultrafine particle formation in the inland

sea breeze airflow in Southwest Europe, Atmospheric Chemistry and Physics

Discussions, 10, 17753-17788, 2010.

Gao, J., Wang, T., Zhou, X., Wu, W., Wang, W.: Measurement of aerosol number

size distributions in the Yangtze River Delta in China: Formation and growth of

particles under polluted conditions, Atmospheric Environment, 43, 829-836,

2009.

Guo, H., Ding, A., Morawska, L., He, C., Ayoko, G., Li, Y., Hung, W.: Size

distribution and new particle formation in subtropical eastern Australia,

Environmental Chemistry, 5, 382-390, 2008.

Hussein, T., Karppinen, A., Kukkonen, J., Härkönen, J., Aalto, P.P., Hämeri, K.,

Kerminen, V-M., Kulmala, M.: Meteorological dependence of size-fractionated

number concentrations of urban aerosol particles, Atmospheric Environment, 40,

1427-1440, 2006.

Hussein, T., Puustinen, A., Aalto, P.P., Mäkelä, J.M., Hämeri, K., Kulmala, M.:

Urban aerosol number size distribution, Atmospheric Chemistry and Physics, 4,

391-411, 2004.

Johnson, G.R., Ristovski, Z.D., Anna, B.D., Morawska, L.: The hygroscopic

behaviour of partially volatilized coastal marine aerosols using the VH-TDMA

192

technique, Journal of Geophysical Research, 110, (D20203),

doi:10.1029/2004JD005657, 2005.

Kim, S., Shen, S., Sioutas, C., Zhu, Y., Hinds, W.C.: Size Distribution and Diurnal

and Seasonal Trends of Ultrafine Particles in Source and Receptor Sites of the

Los Angeles Basin, Air and Waste Management Association, 52:33, 2955-2968,

2002.

Kulmala, M.: How Particles Nucleate and Grow, Science, 302, 1000 – 1001, 2003.

Kulmala, M., Vehkamäki, H., Petäjä, T., Dal Maso, M., Lauri, A., Kerminen, V.-M.,

Birmili, W., McMurry, P.H.: Formation and growth rates of ultrafine

atmospheric particles: a review of observations, Journal of Aerosol Science, 35,

143-176, 2004.

Lee, Y., Lee, H., Kim, M., Choi, C.Y., Kim, J.: Characteristics of particle formation

events in the coastal region of Korea in 2005, Atmospheric Environment, 42,

3729-3739, 2008.

Mazaheri, M., Johnson, G.R., Morawska, L.: Particle and Gaseous Emissions from

Commercial Aircraft at Each Stage of the Landing and Takeoff Cycle,

Environmental Science and Technology, 43, 441 – 446, 2009.

Mejía, J.F., Morawska, L.: An investigation of nucleation events in a coastal urban

environment in the Southern Hemisphere, Atmospheric Chemistry and Physics,

9, 7877-7888, 2009.

Mejíía, J.F., Wraith, D., Mengersen, K., Morawska, L.: Trends in size classified

particle number concentration in subtropical Brisbane, Australia, based on a 5

year study, Atmospheric Environment, 41, 1064-1079, 2007.

Meyer N.K., Ristovski, Z.: Ternary nucleation as a mechanism for the production of

diesel nanoparticles: Experimental analysis of the volatile and hygroscopic

properties of diesel exhaust using the volatilization and humidification tandem

differential mobility analyzer, Environmental Science and Technology, 41,

7309-7314, 2007.

193

Modini, R., Ristovski, Z., Johnson, G.R., Congrong, H., Surawski, N., Morawska, L.,

Suni, T., Kulmala, M.: New particle formation and growth at a remote, sub-

tropical coastal location, Atmospheric Chemistry and Physics, 9 (19), 7607-7621,

2009.

Moore, K.F., Ning, Z., Ntziachristos, L., Schauer, J.J., Sioutas, C.: Daily variation in

the properties of urban ultrafine aerosol – Part I: Physical characterization and

volatility, Atmospheric Environment, 41, 8633-8646, 2007.

Morawska, L., Thomas, S., Bofinger, N., Wainwright, D., Neale, D.: Comprehensive

characterization of aerosols in a subtropical urban atmosphere: Particle size

distribution and correlation with gaseous pollutants, Atmospheric Environment,

32, 2467-2478, 1998.

Morawaska, L., Ristovski, Z., Jayaratne, E.R., Keogh, D.U., Ling, X.: Ambient nano

and ultrafine particles from motor vehicle emissions: Characteristics, ambient

processing and implications on human exposure, Atmospheric Environment, 42,

8113-8138, 2008.

O’Dowd, C.D., McFiggans, G., Greasey, D.J., Pirjola, L., Hoell, C., Smith, M.H.,

Allan, B.J., Plane, J.M.C., Heard, D.E., Lee, J.D., Pilling, M.J., Kulmala, M.: On

the photochemical production of new particles in the coastal boundary layer,

Geophysical Research Letters, 26, 1707-1710, 1999.

Olivares, G., Johansson, C., Strom, J., Hasson, H.-C.: The role of ambient

temperature for particle number concentrations in a street canyon, Atmospheric

Environment, 41 (10), 2145-2155, 2007.

Park, K., Park, J.Y., Kwak, J-H., Cho, G.N., Kim, J.-S.: Seasonal and diurnal

variations of ultrafine particle concentration in urban Gwangju, Korea:

Observation of ultrafine particle events, Atmospheric Environment, 42, 788-799,

2008.

Pérez., N., Pey, J., Cusack, M., Reche, C., Querol, X., Alastuey, A., Viana, M.:

Variability of Particle Number, Black Carbon, and PM10, PM2.5, and PM1

Levels and Speciation: Influence of Road Traffic Emissions on Urban Air

Quality, Aerosol Science and Technology, 44, 487 – 499, 2010.

194

Pey, J., Querol, X., Alastuey, A., Rodŕiguez, S., Putaud, J.P., Van Dingenen., R.:

Source apportionment of urban fine and ultra-fine particle number concentration

in a Western Mediterranean city, Atmospheric Environment, 43, 4407-4415,

2009.

Pey, J., Rodŕiguez, S., Querol, X., Alastuey, A., Moreno, T., Putaud J.P., Van

Dingenen, R.: Variations of urban aerosols in the western Mediterranean,

Atmospheric Environment, 42, 9052-9062, 2008.

Ristovski, Z.D., Suni, T., Kulmala, M., Boy, M., Meyer, N.K., Duplissy, J.,

Turnipseed, A., Morawska, L., Baltensperger, U.: The role of sulphates and

organic vapours in growth of newly formed particles in a eucalypt forest,

Atmospheric Chemistry and Physics, 10, 2919-2926, 2010.

Rodríguez, S., Cuevas, E., González, Y., Ramos, R., Romero, P.M., Pérez, N.,

Querol, X., Alastuey, A.: Influence of sea breeze circulation and road traffic

emissions on the relationship between particle number, black carbon, PM1,

PM2.5 and PM2.5-10 concentrations in a coastal city, Atmospheric Environment,

42, 6523-6534, 2008.

Qian, S., Sakurai, H., McMurry, P.H.: Characteristics of regional nucleation events

in urban East St. Louis, Atmospheric Environment, 41, 4119-4127, 2007.

Sinha, P., Hobbs, P.V., Yokelson, R.J., Christian, T.J., Kirchstetter, T.W., Bruintjes,

R.: Emissions of trace gases and particles from two ships in the southern

Atlantic Ocean, Atmospheric Environment, 37, 2139-2148, 2003.

Suni, T., Kulmala, M., Hirsikko, A., Bergman, T., Laakso, L., Aalto, P.P., Leuning,

R., Cleugh, H., Zegelin, S., Hughes, D., van Gorsel, E., Kitchen, M., Vana, M.,

Hõrrak, U., Mirme, S., Mirme, A., Sevanto, S., Twining, J., Tadros, C.:

Formation and characteristics of ions and charged aerosol particles in a native

Australian Eucalypt forest, Atmospheric Chemistry and Physics, 8, 129 – 139,

2008.

Vehkamäki, H., Dal Maso, M., Hussein, T., Flanagan, R., Hyvärinen, A., Lauros, J.,

Merikanto, J., Mönkkönen, P., Pihlatie, M., Salminen, K., Sogacheva, L., Thum,

T., Ruuskanen, T.M., Keronen, P., Aalto, P.P., Hari, P., Lehtinen, K.E.J., Rannik,

195

Ü., Kulmala, M.: Atmospheric particle formation events at Värriö measurement

station in Finnish Lapland 1998-2002, Atmospheric Chemistry and Physics, 4,

2015-2023, 2004.

Weber, R.J., Moore, K., Kapustin, V., Clarke, A., Mauldin, R.L., Kosciuch, E.,

Cantrell, C., Eisele, F., Anderson, B., Thornhill, L.: Nucleation in the equatorial

pacific during PEM-tropics B: Enhanced boundary layer H2SO4 but no particle

production: NASA global tropospheric experiment Pacific Exploratory Mission

in the tropics phase B, Part 1: Measurement and analyses (PEM-Tropics B),

Journal of Geophysical Research, 106, 32767 – 32776, 2001.

Woo, K.S., Chen, D.R., Pui, D.Y.H., McMurry, P.H.: Measurements of Atlanta

aerosol size distributions: observations of ultrafine particle events, Aerosol

Science and Technology, 34, 75-87, 2001.

Wu, Z., Hu, M., Lin, P., Liu, S., Wehner, B., Wiedensohler, A.: Particle number size

distribution in the urban atmosphere of Beijing, China, Atmospheric

Environment, 42, 7967,7980, 2008.

196

CHAPTER 6

Influence of medium range transport of particles from nucleation burst on particle

number concentration within the urban airshed

H.C. Cheung, L. Morawska, Z.D. Ristovski and D. Wainwright

International Laboratory for Air Quality and Health, Queensland University of

Technology

GPO Box 2434, Brisbane QLD 4001, Australia

Published to the Journal of Atmospheric Chemistry and Physics

197

STATEMENT OF JOINT AUTHORSHIP

Title: Influence of medium range transport of particles from nucleation burst on

particle number concentration within the urban airshed.

Authors: H.C. Cheung, L. Morawska*, Z.D. Ristovski and D. Wainwright

H.C. Cheung

Designed and developed the methodology, conducted the field measurement,

analysed and interpreted the data, and wrote the manuscript.

L. Morawska

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

Z.D. Ristovski

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

D. Wainwright

Contributed to the development of the methodology and assess of the gaseous and

meteorological data.

198

CHAPTER 6. INFLUENCE OF MEDIUM RANGE TRANSPORT OF

PARTICLES FROM NUCLEATION BURST ON PARTICLE NUMBER

CONCENTRATION WITHIN THE URBAN AIRSHED.

H.C. Cheunga, L. Morawska*,a, Z.D. Ristovskia and D. Wainwrightb

aInternational Laboratory for Air Quality and Health, Queensland University of

Technology, Brisbane, QLD 4001, Australia

bQueensland Department of Environmental Resource and Management

Abstract

An elevated particle number concentration (PNC) observed during nucleation events

could play a significant contribution to the total particle load and therefore to air

pollution in urban environments. Therefore, a field measurement study of PNC began

to investigate the temporal and spatial variations of PNC within the urban airshed of

Brisbane, Australia. In 2009, PNC was monitored at urban (QUT), roadside (WOO)

and semi-urban (ROC) areas around the Brisbane region. During the morning traffic

peak period, the highest relative fraction of PNC reached about 5% at QUT and

WOO on weekdays. PNC peaks were observed around noon, which correlated with

the highest solar radiation levels at all three stations, thus suggesting that high PNC

levels were likely to be associated with new particle formation caused by

photochemical reactions. Wind rose plots showed relatively higher PNC for the NE

direction, which was associated with industrial pollution, accounting for 12%, 9%

and 14% of overall PNC at QUT, WOO and ROC, respectively. Although there was

199

no significant correlation between PNC at each station, the variation of PNC was

well correlated among three stations during regional nucleation events. In addition,

PNC at ROC was significantly influenced by upwind urban pollution during the

nucleation burst events, with the average enrichment factor of 15.4. This study

provides an insight into the influence of regional nucleation events on PNC in the

Brisbane region and it the first study to quantify the effect of urban pollution on

semi-urban PNC throughout nucleation events.

200

6.1 Introduction

Atmospheric aerosols have been reported to be significantly associated with the

alteration of climate forcing and the degradation of visibility, as well as the

deterioration of human respiratory and cardiovascular systems (Charlson et al., 1992,

Donaldson et al., 1998, Watson 2002). Due to their small size (< 0.1 µm), ultrafine

particles (UFPs) only contribute a very small amount to the total mass of atmospheric

particles, however they are most abundant by number (~70-90%) and potentially

have a greater impact on human health than the larger particles (< 2.5 µm)

(Morawska et al., 2008).

In urban environments, vehicle exhaust emissions are the most significant source of

UFP and variations in particle number concentration (PNC) are strongly associated

with local urban traffic activity (Morawska et al., 1998, 2008). Aircraft/ship

emissions also contribute to elevated PNCs at a magnitude of 105 – 106 cm-3 (Sinha

et al., 2003, Mazaheri et al., 2009). In addition to direct emissions from the above

sources, new particles formed by nucleation processes are another source of UFPs in

the urban environment, with PNC reaching magnitudes as high as 104 - 105 cm-3

during nucleation events (Qian et al., 2007, Pey et al., 2009, Cheung et al., 2011). In

previous studies, particle mass concentration has been studied with regard to long

range transport in an intercontinental scale (Jaffe et al., 2003), however the size

distribution of and temporal-spatial variations in PNC have only been investigated on

a local scale (Morawska et al., 1998; Hussein et al., 2004). For example, although

regional nucleation has been observed in Helsinki, Finland (Hussein et al., 2008),

Atlanta and Pittsburgh, United States of America (Stolzenburg et al., 2005, Stanier et

al., 2004), spatial variations in PNC have been studied in urban areas in Australia

(Mejia et al., 2008) and in the United States (Hudda et al., 2010). These studies have

201

not examined the impact of regional pollution on PNCs or the influence upwind

urban pollution has on PNC downwind during nucleation events.

This study aims to examine the effect of regional pollution on PNC in different

environments in the Brisbane region. After characterising the spatial variation of

PNC in three different urban locations, we investigated the influence of regional

nucleation on PNC in the same regions. Furthermore, the impact of urban pollution

on PNC downwind from a semi-urban area during a nucleation burst event is also

quantified. The results of this study are valuable for assessing the impact of

nucleation on PNC in an urban environment.

6.2 Methods and Techniques

6.2.1 Study design

Field measurements of particles and gaseous pollutants were conducted at three

locations in Brisbane in 2009, to represent the urban (1 January to 31 December

2009), roadside (21 May to 31 December 2009) and semi-urban environments (5

February to 31 December 2009).

6.2.2 The topography and meteorology of the Brisbane region

Brisbane is located at 27’30oS and 153oE, in Queensland, Australia. The Brisbane

city is surrounded by mountains from south to north, and faces the Pacific Ocean to

the East. Traffic exhaust emissions are the major pollution source affecting the

central business district (CBD). In addition, the aircraft, ship and industrial emissions

are occasionally transported from the lower reaches of the Brisbane River,

approximately 15-18 km NE of the CBD, by inland sea breezes. General wind

patterns in the Brisbane region are governed by land and sea breezes, which are

described in more detail by Morawska et al. (1998).

202

6.2.2.1 Brisbane CBD (urban general)

The measurements were conducted at the International Laboratory of Air Quality and

Health (ILAQH), Queensland University of Technology (QUT), which is within the

Brisbane CBD (Figure 6-1). The monitoring site is on the sixth floor of a QUT

campus building, located SE of the city centre, with a major highway (the Pacific

Motorway) situated along the SW side of the campus. Therefore, the pollution

associated with NE winds could be attributed to industrial emissions (from the

airport, oil refinery and Port of Brisbane), while the pollution associated with S to

NW winds could be attributed to local traffic exhaust emissions.

Figure 6-1. Map of monitoring sites.

6.2.2.2 Woolloongabba (roadside)

The Woolloongabba (WOO) monitoring station is located 3 km south of the Brisbane

CBD and is a part of the South-East Queensland air monitoring network of the

Department of Environmental Resource and Management (DERM). The monitoring

203

station is situated about 5 meters from the kerb of Ipswich Road, a road with a heavy

traffic flow volume of over 40,000, connecting the Southern Brisbane suburbs to the

CBD. A relatively higher PNC level was expected at this site due to the significant

impact of vehicle emission on PNC. In addition, a mutli-storey car park located 10

meters to the West of the station, and large scale road works surrounding the station,

could also influence particle pollution levels.

6.2.2.3 Rocklea (semi-urban)

The Rocklea (ROC) monitoring station is located around 10 km south of the

Brisbane CBD and is also operated by the DERM. This station is surrounded by an

open area, and the particle concentration was deemed to be free from the influence of

local emissions. The major emission sources are from light industrial (Brisbane

farmers markets) and residential (domestic cooking) sources in the Rocklea area.

6.2.3 Measurement techniques

UFP size distribution in the range 4-110 nm was measured at the QUT monitoring

site using a Scanning Mobility Particle Sizer (SMPS), which consists of two parts, an

Electrostatic Classifier (EC) (TSI 3080) equipped with a nano-Differential Mobility

Analyser (nano-DMA) and a Condensation Particle Counter (CPC) (TSI 3781).

Ambient air was drawn through a ~1m long conductive tubing connected to the EC.

The ratio of the aerosol/sheath air flow for the EC was kept at 1/10 (0.6 to 6L min-1),

and the scan time was five minutes. The size distribution data is then used to

calculate PNC for the QUT site. At the WOO and ROC stations, PNC is

continuously measured by a water-based CPC (TSI 3781) with a size-cut inlet of 1

nm, while particle mass concentrations of PM2.5 and PM10 are measured by a

204

Tapered Element Oscillating Microbalance (TEOM) in 30-minute intervals at each

site.

Gaseous pollutants, such as carbon monoxide (CO) and nitrogen oxide (NOx), were

measured at WOO; and ozone (O3) and CO were measured at ROC using real-time

gaseous analysers (Ecotech ML9830 for CO; Ecotech ML9841/ API 200A for NOx;

Ecotech ML9812 for O3). Meteorological parameters including wind direction/speed,

temperature, relative humidity and solar radiation, have also been measured. The

data was collected and validated by the DERM.

6.2.4 Data processing and analysis

The raw particle size distribution measurements were transformed into 10 min

averaged data for figure plotting. The total PNC for QUT was calculated by adding

all of the particle counts in each size bin, which had a lower and upper limit of 1 cm-3

and 5 x 105 cm-3, respectively (Mejía et al., 2007). Approximately 28% of the data

removed from the database was based on the following criteria (the contribution of

each quality control is shown in brackets): i) if the particle concentration has a zero

value (~2%); ii) if the particle concentration is higher than 5 x 105 cm-3 (<1%); iii)

and if data has been collected during instrument malfunction (~26%). Since the time

resolutions of the particle mass concentration, gaseous and meteorological data

provided by the DERM were in 30 min intervals, all measurements were transformed

into 30 min averaged data for correlation analysis (Section 6.3.2). Since the PNCs

measured at three sites using SMPS and CPC, to remove the discrepancy of these

measurement methods, a relative PN contribution to total PNC has been used in

temporal and correlation analysis. Inter-comparison between the PNCs measured by

SMPS and CPCs has been shown in Figure-S1; moderate correlations have been

205

obtained (r2 = 0.47- 0.81) with slopes of 0.55-0.65. This implies that the method of

measuring PNCs by SMPS and CPCs for correlation analysis is justifiable and the

ultrafine particles accounted for more than 50% of the PNC (by using CPC).

Correlations between the parameters were tested using the Pearson correlation test,

with a 95% confidence level (p < 0.05). The linearity of the tested parameters was

indicated by the product of Pearson’s coefficient, r2, with a perfect linear correlation

between two parameters indicated by an r2 value close to 1. It should be noted that

the PNC data for WOO is missing for the months from January to April due to

instrument malfunction.

The back-trajectory of various air masses was calculated by using the HYSPLIT

model (Hybrid Single Particle Lagrangian Integrated Trajectory, Version 4.9), in

order to trace their origin. The meteorological data used for back-trajectory

calculations was 1˚ x 1˚ in latitude and longitude. The calculated trajectory analysis

provided an indication of which region the air mass came from. Further details about

the principle and operation of the HYSPLIT model are referenced in these articles

(Draxler and Hess 1997, 1998; Draxler 1999).

6.3 Results and Discussion

Firstly, the variation in PNC within each location was investigated by analysing the

diurnal variation together with other measured parameters. Secondly, correlations

between the PNC for the different locations were examined, along with the influence

of wind direction on PNC. Finally, two cases which represented typical regional

nucleation events and the transport of urban pollution to the downwind semi-urban

site were investigated.

206

6.3.1 Diurnal variation

Diurnal variations of the PNC measured at the three locations, which have been

classified into (a) weekdays and (b) weekends by all measured data, are illustrated in

Figure 6-2. It should be noted that the measurement periods at each site did not

overlap. The general meteorological conditions for weekdays and weekends were

similar, with SE-winds observed in the morning and NE-winds observed around

noon, while solar radiation reached a maximum at noon on all days. In contrast,

traffic volumes differed during the weekdays and on weekends, such that i) traffic

volumes were higher during weekdays than on weekends; ii) the daily traffic volume

pattern consisted of two peaks during weekdays, one in the morning (~ 6-7 am) and

one in the afternoon (around 3-6 pm); and iii) the daily pattern during weekends

consisted of a wider, broader peak.

In Figure 6-2a, it can be seen that the morning PNC peaks were observed both at

QUT and WOO. During that period, the measured relative fraction of PNC was

found to be nearly 5% for both sites however, they were not found at ROC. This

result suggests that the observed peaks are related to morning traffic activity on

nearby roads. Around noon, PNC peaks were observed at all three locations, as well

as the maximum solar radiation. The highest relative fraction of total PNC at noon is

7.6%, 6.0% and 8.9% for QUT, WOO and ROC locations, respectively. These peaks

are likely to be the result of new particle formation caused by photochemical

reactions (Cheung et al., 2011). It should be noted that the relative fraction of the

total PNC is affected by background PNC, traffic emissions and photochemical

particle production during the morning and noon periods. The maximum PNC

observed at QUT and WOO is at 12:00, and at 13:00 for ROC. The time lag at ROC

could be the result of the time the pollution plume requires to be transported from the

207

upwind area (CBD area), to the downwind area (ROC). This is discussed in more

detail in Section 6.3.5.

One interesting observation is that the influence of traffic activity was weak during

the afternoon period (~15:00-18:00) for both weekdays and weekends, with the PNC

found to decrease gradually between 14:00-15:00, even though traffic volume

remained relatively unchanged. Similar observations were made in the urban area of

Helsinki, Finland (Hussein et al., 2004).

208

Figure 6-2. Average diurnal variation of parameters measured for (a) weekdays and (b) weekends. From bottom to top: i) relative PNC measured at three sites, together with traffic volumes recorded at QUT; ii) wind vectors measured at the three sites; and iii) solar radiation (SR) measured at ROC.

209

6.3.2 Correlation among measured parameters

A summary of the correlation coefficients (r2) for the measured parameters from the

entire period is provided in Table 6-1. The low r values, 0.05 < r2 < 0.19, showed

that the PNC at the three sites were not correlated, however PM2.5 and PM10 at WOO

and ROC were well correlated (0.60 < r2 < 0.88). These results imply that the PNC at

each site were generally influenced by local sources, such as vehicle exhaust

emissions (Morawska et al., 2008), while PM2.5 and PM10 were influenced by intra-

city pollution.

Although there was no correlation between the PNC between the three sites, the PNC

appeared to be influenced by regional pollution during the nucleation event

(discussed in Section 6.3.4). PNC at QUT and ROC did not show a significant

correlation with primary gaseous pollutants such as CO and NOx, but PNC did show

a moderate correlation with CO (r = 0.35) and NOx (r = 0.47) at WOO. The results

observed at QUT are in contrast to those reported by Morawska et al. (1998), where

PNC (5-1000 nm) at QUT was reasonably well correlated with CO (r = 0.45) and

NOx (r = 0.40), and was also influenced by vehicle exhaust emissions. This

discrepancy may be due to the different measurement periods, as the measurements

were only conducted during the morning and afternoon peak traffic hours in

Morawska et al. (1998). However, a continuous measurement approach was used in

the present study, which included a more complex mixture of emissions and a

significant contribution to the PNC by nucleation process, which may have masked

the influence of vehicle exhaust emissions on PNC.

210

NQUT NWOO NROC PM10 (QUT)

NOx

(WOO) CO (WOO)

PM2.5

(WOO) PM10

(WOO) NOx

(ROC) O3 (ROC)

PM2.5

(ROC) PM10 (ROC)

RAD

NQUT 1.00 0.05 0.19 0.00 0.02 0.00 0.00 0.00 0.00 0.02 0.00 0.00 0.08

NWOO 1.00 0.12 0.00 0.22 0.12 0.01 0.00 0.03 0.01 0.00 0.00 0.07

NROC 1.00 0.00 0.04 0.03 0.00 0.00 0.01 0.08 0.00 0.00 0.12

PM10 (QUT) 1.00 0.00 0.00 0.12 0.17 0.00 0.01 0.13 0.14 0.00

NOx (WOO) 1.00 0.65 0.05 0.00 0.31 0.02 0.01 0.00 0.00

CO(WOO) 1.00 0.02 0.00 0.20 0.01 0.00 0.00 0.01

PM2.5 (WOO) 1.00 0.80 0.03 0.00 0.79 0.60 0.01

PM10 (WOO) 1.00 0.00 0.00 0.88 0.80 0.00

NOx (ROC) 1.00 0.27 0.01 0.00 0.06

O3 (ROC) 1.00 0.01 0.00 0.38

PM2.5 (ROC) 1.00 0.83 0.00

PM10 (ROC) 1.00 0.00

RAD 1.00

Table 6-1. r2 calculated between the parameters.

211

Furthermore, scatter plots (Figure 6-3) for carbon monoxide and nitrogen oxide at

WOO show the influence of two pollution plumes with different CO/NOx ratios as a

function of wind direction. The CO/NOx ratios for these two groups were ~ 25 and

7.5, respectively. The first group (ratio 25) was associated with winds from the SW

to NW, while the second group (ratio 7.5) was mainly associated with NE winds.

Given that the vehicle exhaust emission ratio in SE Queensland is about 10 (Cheung

et al., 2011), it is likely that the second group was affected by vehicle exhaust

emissions. Since low speed driving induces higher CO/NOx ratios than faster driving

modes (Holmén and Niemeier 1997), this could be a result of vehicle emissions from

the hospital car park, located 10 meters W of WOO. It should be noted that, in the

absence of in-situ measurements, this explanation is only speculative.

Figure 6-3. Scatter plots of COWOO versus NOx,WOO

212

6.3.3 Dependence of the particle number concentration on wind direction

The wind rose plots for PNC showed a similar pattern for the three sites (see Figure

6-4a), with all three sites affected by both land and sea breezes, which blew from the

SW and NE, respectively (see Figure 6-4b). In general, relatively higher

concentrations were observed in the NE quadrant, being 12%, 9% and 14% for QUT,

WOO and ROC, respectively, compared to around 3-6% for the other three quadrants.

This result implies that, in addition to the local sources at each site, there was a

significant source located in a NE direction of the sites. This is most likely the result

of the numerous industrial activities (i.e. Port of Brisbane, Oil Refinery, and

Domestic and International Airports) taking place NE of all three sites, as well as

from traffic emissions from the CBD, which is also upwind from WOO and ROC.

213

Figure 6-4. Wind rose plots of (a) relative PNC contribution at QUT, WOO and

ROC; (b) relative frequencies of wind direction.

6.3.4 Influences of regional pollution on the particle number concentration

During 2009, a total of 40 nucleation growth events were observed based on particle

size distribution data for QUT in Brisbane (Cheung et al. 2011) and more detailed

information regarding the classification of nucleation events can be found in this

214

paper. In order to further investigate the influence of nucleation growth events on the

PNC of the Brisbane region, we chose two nucleation events as case studies, one

which occurred on 17 July 2009 and the other on 9 September 2009. The first case

was a region wide event, while the second case was a local event. A Pearson’s

correlation coefficient was calculated for each event, using data obtained between

08:00-16:00, since the majority of nucleation events were initiated during this period.

From Table 6-2 it can be seen that the PNCs at all three sites showed significant

correlations during some nucleation events. For example, the r2 values for QUT-ROC,

QUT-WOO and WOO-ROC were 0.95, 0.71 and 0.75 on 17 July 2009, respectively

(with p < 0.05). To better illustrate the correlation of the PNCs from different

locations (urban and downwind semi-urban areas) during nucleation and non-

nucleation events, an example of the PNC scatterplot obtained during the event and

non-event days at QUT and ROC is shown in Figure 6-S2.

As shown in Figure 6-5a, similar temporal variations were observed on 17 July 2009,

when a nucleation growth event was observed at 10:30. At this time the relative PNC

at QUT, WOO and ROC increased from 1.5 % to 8 %, 2 % to 5 % and 1.5 % to 10 %

at each station, respectively, before returning to around 1.5 % for each site at

approximately 16:00.

215

Event types Date r2 (QUT-ROC) r2 (QUT-WOO) r2 (WOO-ROC)

Class Ia 15-Feb-2009 0.42 16-Feb-2009 0.11 28-Apr-2009 0.52 29-Apr-2009 0.17

Class Ib

30-Mar-2009 0.00 17-May-2009 0.74 6-Jun-2009 0.02 0.42 0.10 18-Jul-2009 0.10 0.58 0.03 28-Jul-2009 0.19 0.23 0.01 1-Aug-2009 0.00 0.46 0.01 2-Aug-2009 0.13 0.02 0.16 8-Aug-2009 0.53 0.25 0.09 9-Aug-2009 0.39 0.11 0.15 18-Aug-2009 0.82 0.66 0.41 9-Sep-2009 0.79 0.00 0.01 8-Oct-2009 0.16 0.33 0.59 21-Oct-2009 0.08 0.12 0.68

Class II

9-Feb-2009 0.09 26-Feb-2009 0.00 27-Feb-2009 0.19 28-Feb-2009 0.00 15-Mar-2009 0.64 16-Mar-2009 0.25 17-Mar-2009 0.54 22-Mar-2009 0.26 5-Apr-2009 0.00 12-May-2009 0.25 21-May-2009 0.04 0.01 0.38 30-May-2009 0.71 0.02 0.04 8-Jun-2009 0.53 0.08 0.05 17-Jul-2009 0.95 0.71 0.75 15-Aug-2009 0.28 0.77 0.20 10-Sep-2009 0.65 0.56 0.45 14-Sep-2009 0.16 0.17 0.20 16-Oct-2009 0.46 0.05 0.10 24-Oct-2009 0.15 0.51 0.67 3-Dec-2009 0.05 0.02 0.21

Table 6-2. The r2 values for PNC at QUT, WOO and ROC during nucleation growth

events (r2 values > 0.4 are bolded). Nucleation events were classified into Class Ia/b

where the particle growth rate can be determined and Class II where the banana

shape still observable, but the data fluctuates to such an extent that formation rate

calculation is impractical. More detailed explanation of class type can be found in

Cheung et al. (2011).

216

In contrast, there were some instances where the r2 was higher than 0.5 for QUT-

ROC (r2 = 0.79), but lower for QUT-WOO (r2 < 0.01) (e.g. 9 September 2009). In

addition, the temporal variation for PNQUT and PNROC were closely correlated during

the period between 10:30-15:30 (see Figure 6-5b), while PNWOO did not follow the

same trend. This indicates that there was a regional pollution plume that affected

QUT and ROC, but not WOO, which was located between the two sites. This may be

due to local atmospheric conditions at WOO which suppressed the nucleation, or it

may be the result of two individual nucleation events that occurred at QUT and ROC

simultaneously. Further analysis was required in order to explain this phenomenon,

including data for gaseous pollutants, meteorological conditions, PM2.5 and PM10

which will be discussed in following section.

217

Figure 6-5. Time series plot of parameters measured on (a) 17 July 2009 and (b) 9

September 2009. From bottom to top: i) relative PNC measured at QUT, WOO and

ROC; ii) wind vectors for QUT, WOO and ROC; and iii) solar radiation (SR) at

ROC.

218

6.3.4.1 Comparison of the two case studies

The back-trajectories for the 17 July 2009 and 9 September 2009 events were

calculated for the first two hours of the nucleation events (see Figure 6-6a-b). The

results suggest that the air masses at each site originated from the same region on

both occasions, which means that the absence of a nucleation event at WOO on 9

September 2009 is likely to be the result of other variable factors such as local

emission sources and/or meteorological conditions.

Figure 6-6. Back-trajectories for the first two hours of each event: (a) 17 Jul 2009

and (b) 9 Sep 2009.

In order to further investigate the similarities and differences between these two

cases, average PNC and PM10, as well as gaseous pollutants such as CO and NOx,

and meteorological conditions including temperature, relative humidity and wind

speed, were compared for WOO and ROC (see Table 6-3). Overall, no significant

219

differences in temperature or relative humidity were observed, although a relatively

lower wind speed was observed at WOO (~ 1 ms-1) compared to that of ROC (~ 3-4

ms-1) during both events. The temperature differences during those events at WOO

and ROC were small, implying that the impacts of height of mixing layer on both

locations were similar. This observation with high r2 (QUT-ROC) and relatively low

r2 (QUT-WOO), r2 (WOO-ROC) was also found on 30 May, 8 Jun, 8 Aug and 16

October 2009. For gaseous pollutants, the mixing ratios for NOx at WOO and ROC

remained the same for both events, while a relatively higher mixing ratio of 0.29

ppm for CO was observed at WOO on 9 September 2009, compared to 0.19 ppm on

17 July 2009. A relatively higher CO/NOx ratio of ~5.8 was also observed at WOO

on 9 September 2009 compared to ~ 3.8 on 17 July 2009.

In terms of particle mass, the PM10 at WOO was 13.6 mg cm-3 on 17 July 2009,

compared to 17.8 mg cm-3 on 9 September 2009 (an increase of 30.9%). On the other

hand the PM10 at ROC was 7.0 mg cm-3 on 17 July 2009, compared to 6.8 mg cm-3

on 9 September 2009 (an decrease of 2.9%). A higher number of pre-existing

particles in the atmosphere can act a strong sink for condensation nuclei, therefore

suppressing new particle formation (Kerminen et al., 2001). In our previous study

(Jayaratne et al, 2011), it has been shown that an increase of PM10 concentration in

the environment leads to a sharp decrease in the number of ultrafine particles.

220

Therefore, the relatively higher particle mass concentration of WOO on 9 September

2009, is indicative of more pre-existing particles. This explains the suppression of

the nucleation process at WOO on this day. Beside the influence of condensation

sink on the suppression of nucleation, the coagulation scavenging may be another

factor which removed the freshly formed particles in this case.

PN (103 cm-

3) PM10

(µg m-3) CO

(ppm) NOx

(ppm) Temp (ºC)

RH (%)

Wind Speed (ms-1)

WOO 8.3 (3.9)

13.6 (3.5)

0.19 (0.15)

0.05 (0.02)

17.7 (1.8)

47.8 (5.8) 1.0 (0.5)

ROC 3.0 (3.2) 7.0 (0.9) n/a

0.00 (0.00)

15.6 (2.8)

48.6 (10.1) 3.8 (0.4)

Table 6-3a. Average values of measured parameters of WOO and ROC from 08:00-

16:00 on 17 Jul 2009. A standard deviation showed in bracket.

PN (103 cm-

3) PM10

(µg m-3) CO

(ppm) NOx

(ppm) Temp (ºC)

RH (%)

Wind Speed (ms-1)

WOO 5.6 (0.7)

17.8 (7.2)

0.29 (0.36)

0.05 (0.02)

24.0 (1.5)

33.0 (6.8) 0.5 (0.1)

ROC 2.6 (1.6) 6.8 (3.3) n/a

0.00 (0.00)

22.5 (2.1)

32.3 (9.9) 3.4 (1.0)

Table 6-3b. Average values of measured parameters of WOO and ROC from 08:00-

16:00 on 9 Sep 2009. A standard deviation showed in bracket.

221

6.3.5 Influence of upwind pollution on the particle number concentration in the downwind area

In addition to the influence of regional pollution on the PNC at each site, upwind

pollution from a NE direction was also found to affect PNC. In order to characterise

the influence of nucleation burst events on air quality downwind from larger

pollution sources, we analysed the data from Cheung et al. (2011), based on 22

nucleation burst events that occurred at QUT during 2009, all of which were

associated with NE winds that originated from the same direction as the Brisbane

Airport, Oil Refinery and Port of Brisbane. Table 6-4 shows the Pearson’s

coefficient for PNC at QUT and ROC during the nucleation burst events. Since it

takes about 30 minutes for air masses to move from QUT to ROC (assuming an

average wind speed of ~ 5 ms-1 during the event period), the r2 values were also

calculated based on data taken 30 minutes later at ROC (e.g. 12:30 data from QUT

was compared with 13:00 data for ROC).

222

Date r2 r2 (ROC data shifted 30 mins)

8 Feb 2009 0.88 0.94 24 Feb 2009 0.04 0.02 15 Mar 2009 0.64 0.82 14 Apr 2009 0.52 0.87 3 Sep 2009 0.75 0.87 16 Sep 2009 0.79 0.93 17 Sep 2009 0.40 0.64 20 Oct 2009 0.56 0.40 22 Oct 2009 0.66 0.65 25 Oct 2009 0.33 0.82 28 Oct 2009 0.33 0.73 29 Oct 2009 0.21 0.66 31 Oct 2009 0.30 0.09 2 Nov 2009 0.51 0.81 5 Nov 2009 0.56 0.28 7 Nov 2009 0.31 0.00 11 Nov 2009 0.58 0.54 12 Nov 2009 0.90 0.81 14 Nov 2009 0.00 0.02 24 Nov 2009 0.37 0.23 26 Nov 2009 0.82 0.98 24 Dec 2009 0.00 0.08

Average 0.48 0.55

Table 6-4. The r2 values for PNC at QUT and ROC during nucleation burst events.

Only data observed between 08:00-16:00 has been used (shifted r2 values larger than

original r2, and the values ≥ 0.4 are bolded).

From this table it can be seen that variations in PNC at QUT and ROC were

correlated during most of the nucleation burst events. In 11 out of the 22 cases, the

“time shifted” r2 values were higher than the original values (only the cases with

shifted r2 ≥ 0.4 were counted), indicating that the variation in PNC in the downwind

area was associated with the air masses from the NE. In order to investigate this

phenomenon further, we constructed a time series plot of the PNC and wind vectors

223

at QUT and ROC for 25 October 2009 (see Figure 6-7). On this day, the original r2

calculated was 0.33, and the “time shifted” r2 significantly improved to 0.82. It can

be seen that the highest PNC levels (12 % contribution to total PN, ~ 80 x 103 cm-3)

were associated with a pollution plume that blew in from the NE, which reached the

CBD at around 11:00. As indicated by a particle burst observed at ROC around 12:00,

this plume reached ROC about 30-60 minutes later. The r2 value of 0.82 indicates

that the increase in PNC at QUT and ROC were similar and back-trajectories

calculated for QUT and ROC during the burst event confirmed that the air masses

originated from same region (see Figure 6-8).

224

Figure 6-7. Time series plot of parameters measured on 25 Oct 2009. From bottom

to top: i) number concentration of nucleation and Aitken modes particles; ii) relative

particle number concentration measured at QUT and ROC; iii) wind vectors at QUT

and ROC; and iv) solar radiation (SR) at Rocklea.

Figure 6-8. Back-trajectories calculated for the burst events on 25 Oct 2009.

225

To further evaluate the impact of the nucleation burst events on the PNC in the semi-

urban area of ROC, we determined the enrichment factor of PNC by calculating the

ratio of maximum to minimum PNC during the period 08:00-16:00. On 25 Oct 2009,

the PNC increased to about 11.6 times higher than the minimum PNC, which

indicates that the urban/industrial pollution plume had a significant effect on PNC in

the semi-urban area. The enrichment factor for other nucleation burst event days are

listed in Table 6-S1, in the supplementary section. The average enrichment factor for

the PNC at ROC is a result of upwind urban pollution during the nucleation burst

events and was found to be 15.4.

6.4 Conclusion

PNC variation was investigated in urban, roadside and semi-urban areas of Brisbane

during 2009. Overall, significant diurnal variation was observed at QUT and WOO

during weekdays, and PNC peaks were observed at all three sites around noon, which

corresponded with the highest solar radiation levels and suggested that the PNC is

associated with new particle formation by photochemical reactions. Wind rose plots

of the PNC showed that highly polluted air plumes blew in from a NE direction,

which is consistent with the direction of major urban and industrial areas, and

although PNC did not show a significant correlation between the three sites,

226

concentrations were found to be closely correlated during nucleation growth events.

In addition, PNC at the semi-urban site of ROC was significantly influenced by the

upwind urban pollution coming from the NE, with an average enrichment factor of

10.8. This study provides an insight into the influence of regional nucleation events

on the PNC in the Brisbane region and is the first study to quantify the effect of

urban pollution on semi-urban PNC. The findings of this study are useful for

environmental management and assessment in regard to PNC.

Acknowledgments

This project was supported by the Australian Research Council and Queensland

Transport through Linkage Grant LP0882544. We would also like to thank the

Department of Environmental Resource and Management, Queensland for providing

the air monitoring data; and the Queensland Bureau of Meteorology for providing the

meteorological data.

227

Supplementary materials

Date Daily maximum of relative PN contribution (%)

Daily minimum of relative PN contribution (%)

Enrichment factor

8 Feb 2009 12.1 0.3 40.3 15 Mar 2009 8.8 0.5 17.6 14 Apr 2009 8.5 0.4 21.3 3 Sep 2009 12.1 0.6 20.2 16 Sep 2009 7.5 0.7 10.7 17 Sep 2009 6.9 1.0 6.9 25 Oct 2009 13.9 1.2

11.6 28 Oct 2009 24.4 1.4 17.4 29 Oct 2009 9.0 1.8 5.0 2 Nov 2009 9.3 0.8

11.6 16 Nov 2009 11.7 1.6 7.3

Average 15.4

Table 6-S1. Enrichment factor of the relative PN contribution for the semi-urban

area, Rocklea under the influence of upwind pollution during the nucleation burst

events. Only data between the period of 08:00-16:00 have been used in the

calculation.

228

Figure 6-S1. Scatter plots of PNCs measured by SMPS (PNCSMPS) and CPC

(PNCCPC) at ROC and WOO.

Figure 6-S2. Scatter plots of relative PN contribution between QUT and ROC during

nucleation event (9 Sep 2009) and non-event (8 Mar 2009).

229

6.5 References

Stanier, C.O., Khlystov, A.Y. and Pandis, S.N. (2004). Nucleation events during the

Pittsburgh air quality study: description and relation to key meteorological, gas

phase and aerosol parameters. Aerosol Science and Technology, 38, 253-264.

Charlson, R.J., Schwartz, S.E., Hales, J.M., Cess, R.D., Jr. Coaklev, J.A., Hansen,

J.E. and Hofmann, D.J. (1992). Climate forcing by anthropogenic aerosols.

Science, 255, 423-430.

Cheung, H.C., Morawska, L., and Ristovski, Z.D. (2011). Observation of new

particle formation in subtropical urban environment. Atmospheric Chemistry and

Physics, 11, 3823-3833.

Donaldson, K., Li, X.Y. and MacNee, W. (1998). Ultrafine (nanometre) particle

mediated lung injury. Journal of Aerosol Science, 29, 553-560.

Draxler, R.R. (1999). HYSPLIT4 user’s guide. NOAA Tech. Memo. ERLARL-230,

NOAA Air Resources Laboratory, Silver Spring, MD.

Draxler, R.R. and Hess, G.D. (1997). Description of the HYSPLIT_4 modeling

system. NOAA Tech. Memo. ERLARL-224, NOAA Air Resources Laboratory,

Silver Spring, MD.

Draxler, R.R. and Hess, G.D. (1998). An overview of the HYSPLIT_4 modeling

system of trajectories, dispersion, and deposition. Australian Meteorological

Magazine, 47, 295-308.

Holmén, B.A. and Niemeier, D.A. (1997). Characterizing the effects of driver

variability on real-world vehicle emissions. Transportation Research Part D, vol. 3,

issue 2, 117.128.

Hudda, N., Cheung, K., Moore, K.F. and Sioutas, C. (2010). Inter-community

variability in total particle number concentrations in the eastern Los Angeles air

basin. Atmospheric Chemistry and Physics, 10, 11385–11399.

Hussein, T., Martikainen, J., Junninen, H., Sogacheva, L., Wagner, R., Dal Maso, M.,

Riipinen, I., Aalto, P.P. and Kulmala, M. (2008). Observation of regional new

particle formation in the urban atmosphere. Tellus, 60B, 509-521.

230

Hussein, T., Puustinen, A., Aalto, P.P., Mäkelä, J.M., Hämeri, K. and Kulmala, M.

(2004). Urban aerosol number size distribution. Atmospheric Chemistry and

Physics, 4, 391-411.

Jaffe, D., Mckendry, I., Anderson, T. and Price, H. (2003). Six “new” episodes of

trans-Pacific transport of air pollutants, Atmospheric Environment, 37, 391-401.

Jayaratne, E.R., Johnson, G.R., McGarry, P., Cheung, H.C. and Morawska, L. (2011).

Characteristics of airborne ultrafine and coarse particles during the Australian dust

storm of 23 September 2009. Atmospheric Environment, 45, 3996-4001.

Kerminen, V-M., Pirjola, L. and Kulmala, M. (2001). How significantly does

coagulational scavenging limit atmospheric particle production?. Journal of

Geophysical Research, 106, 24119-24125, doi:10.1029/2001JD000322.

Mazaheri, M., Johnson, G.R. and Morawska, L. (2009). Particle and Gaseous

Emissions from Commercial Aircraft at Each Stage of the Landing and Takeoff

Cycle. Environmental Science and Technology, 43, 441-446.

Mejía, J.F., Morawska, L. and Megersen, K. (2008). Spatial variation in particle

number size distributions in a large metropolitan area. Atmospheric Chemistry

and Physics, 8, 1127-1138.

Mejía, J.F., Wraith, D., Mengersen, K. and Morawska, L. (2007). Trends in size

classified particle number concentration in subtropical Brisbane, Australia, based

on a 5 year study, Atmospheric Environment, 41, 1064-1079.

Morawska, L., Thomas, S., Bofinger, N., Wainwright, D. and Neale, D. (1998).

Comprehensive characterization of aerosols in a subtropical urban atmosphere:

Particle size distribution and correlation with gaseous pollutants. Atmospheric

Environment, 32, 2467-2478.

Morawska, L., Ristovski, Z., Jayaratne, E.R., Keogh, D.U. and Ling, X. (2008).

Ambient nano and ultrafine particles from motor vehicle emissions:

Characteristics, ambient processing and implications on human exposure.

Atmospheric Environment, 42, 8113-8138.

231

Pey, J., Querol, X., Alastuey, A., Rodŕiguez, S., Putaud, J.P. and Van Dingenen. R.

(2009). Source apportionment of urban fine and ultra-fine particle number

concentration in a Western Mediterranean city. Atmospheric Environment, 43,

4407-4415.

Qian, S., Sakurai, H. and McMurry, P.H. (2007). Characteristics of regional

nucleation events in urban East St. Louis. Atmospheric Environment, 41, 4119-

4127.

Sinha, P., Hobbs, P.V., Yokelson, R.J., Christian, T.J., Kirchstetter, T.W. and

Bruintjes, R. (2003). Emissions of trace gases and particles from two ships in the

southern Atlantic Ocean. Atmospheric Environment, 37, 2139-2149.

Stolzenburg, M.R., McMurry, P.H. and Sakurai, H. (2005). Growth rates of freshly

nucleated atmospheric particles in Atlanta. Journal of Geophysical Research, 110,

D22S05, doi:10.1029/2005JD05935.

Watson, J.G. (2002). Visibility: Science and Regulation. Journal of Air and Waste

Management, 52, 628-713.

232

CHAPTER 7

Influence of sulfuric acid on the new particle formation in subtropical urban area in

the Southern Hemisphere.

H.C. Cheung, L. Morawska, Z.D. Ristovski and D. Wainwright

International Laboratory for Air Quality and Health, Queensland University of

Technology

GPO Box 2434, Brisbane QLD 4001, Australia

Submitted to Atmospheric Environment

233

STATEMENT OF JOINT AUTHORSHIP

Title: Influence of sulfuric acid on the new particle formation in a subtropical urban

area in the Southern Hemisphere.

Authors: H.C. Cheung, L. Morawska*, Z.D. Ristovski and D. Wainwright

H.C. Cheung

Designed and developed the methodology, conducted the field measurement,

analysed and interpreted the data, and wrote the manuscript.

L. Morawska

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

Z.D. Ristovski

Contributed to the development of the methodology, analysed and interpreted the

data, and the manuscript writing.

D. Wainwright

Contributed to the development of the methodology and assess of the gaseous and

meteorological data.

234

CHAPTER 7. INFLUENCE OF SULFURIC ACID ON THE NEW PARTICLE

FORMATION IN SUBTROPICAL URBAN AREA IN THE SOUTHERN

HEMISPHERE.

H.C. Cheunga, L. Morawska*,a, Z.D. Ristovskia and D. Wainwrightb

aInternational Laboratory for Air Quality and Health, Queensland University of

Technology, Brisbane, QLD 4001, Australia

bQueensland Department of Environmental Resource and Management

Abstract

The aim of this study was to investigate the influence of sulphuric acid on new

particle formation and the number concentration of ultrafine particles in a subtropical

urban area in the Southern Hemisphere. Measurements of particle size distribution

and sulphur dioxide (SO2) were conducted from June to July 2009 in urban (QUT),

urban-roadside (Woolloongabba) and semi-urban (Rocklea) areas of Brisbane,

Australia. A sulphuric acid (H2SO4) proxy was used to study its influence on new

particle formation and the variation of particle number concentration (PNC).

235

Temporal variation of the proxy was similar to variations in number concentration of

nucleation mode particles at the urban and urban-roadside stations, while the peak for

the observed proxy occurred 1-2 hours prior to the peak in nucleation mode particles

at the semi-urban site. The concentration of H2SO4 proxy during nucleation event

days was found to be 12% and 88% higher than on non-event days at QUT and

Woolloongabba, respectively. In contrast, the proxy was found to be 4.4 times lower

than on non-event days for the semi-urban Rocklea site. A moderate to strong linear

relationship was found between the H2SO4 proxy and freshly formed particles, as

shown by moderate r2 values 0.26-0.77 during the nucleation events at both the urban

and urban-roadside sites. The amount of the log[H2SO4 proxy] required to produce

new particles was found to be ~ 1.0 ppb Wm-2 s and below 0.5 ppb Wm-2 s, for the

urban and urban-roadside sites, respectively. Similar particle growth rates were

observed at the three study locations, with an overall average value of 2.7±0.5 nm hr-

1. A weak relationship was found between the proxy and particle number

concentration at the semi-urban site of Rocklea. This was because the nucleation

process was initiated in the upwind area and the mew particles were transported to

Rocklea. Differences in the relationship between the proxy and particle number

concentrations can be attributed to the fact that local nucleation occurred at the

236

urban/urban-roadside sites, while the nucleation process began upwind of the semi-

urban site and the new particles were transported downwind to the measurement site.

237

7.1 Introduction

The presence of particles in the atmosphere has raised scientific concern, due to their

potential impact on climate and human health. Atmospheric particles react directly

with solar radiation through absorption and scattering mechanisms (Myhre 2009),

and the also modify cloud formation processes by changing the profile of cloud

condensation nuclei, thus altering the radiative climate forcing of the Earth (Wang

and Penner 2009). Besides their impact on climate forcing, these particles can also

reduce the visibility and impair human respiratory and cardiovascular systems (Pope

and Dockery 2006; Cheung et al., 2005). To evaluate the effects of atmospheric

particles, it is necessary to improve our understanding of their physical and chemical

properties, as well as their formation process, and spatial and temporal variations in

the atmosphere.

Sulfuric acid has been identified as a critical component of the nucleation process,

which has been found to be positively related to the number concentration of

nucleation mode particles in situ (Kulmala and Kerminen 2008; Ristovski et al., 2010;

Kulmala et al., 2011). Sulfur dioxide (SO2) has been commonly used to indicate the

influence of sulfuric acid on the nucleation and for field measurement studies, the

production of atmospheric particles has estimated by measuring sulfuric acid

together with solar radiation (Passonen et al., 2009; Petäjä et al., 2009). Elevated

238

nucleation mode particle concentrations have also been observed along with SO2

pollution plumes in the urban environment (Woo et al., 2001), which implies that

new particle formation occurs in SO2 enriched plumes. Petäjä et al. (2009) used a

sulfuric acid proxy to represent the particle production strength by sulfuric acid,

which is a function of SO2 and solar radiation, as well as the condensation sinks of

pre-existing atmospheric particles. In general, higher proxy values were obtained

during nucleation event days compared to non-nucleation event days and seasonal

variation of the proxy was found to coincide with the occurrence of nucleation events

in the urban environment (Salma et al., 2011). The H2SO4 proxy has also been used

to estimate the concentration of global nucleation mode particles based on satellite

data (Kulmala et al., 2011). The study found a qualitative agreement between

observed and simulated global nucleation mode particle number concentrations,

however this approach is not applicable in highly polluted or very clean

environments, due to the limited spatial resolution of satellite data, which is

inadequate to distinguish SO2 column concentrations.

New particle formation has been observed in subtropical urban areas of Brisbane, in

the Southern Hemisphere, which were significantly influenced by photochemical

activities (Cheung et al., 2011a). In addition, particle number concentration in the

semi-urban area around Brisbane was influenced by regional nucleation events which

239

were initiated at upwind urban/industrial areas (Cheung et al., 2011b). However, the

mechanisms of nucleation observed in this region are still not clear. In this study, we

investigated a number of nucleation mechanisms by using a H2SO4 proxy as an

indication of the level of sulfuric acid associated with nucleation. The aim of this

study was to examine the influence of the sulfuric acid proxy on nucleation mode

particle concentrations during nucleation events in the subtropical urban area of

Brisbane, as well as to investigate temporal variations of the proxy at urban, roadside

and semi-urban locations around Brisbane.

7.2 Methods and Techniques

7.2.1 Study design

A field measurement campaign for particles and gaseous pollutants was conducted at

three locations around Brisbane during 11 June to 30 July 2009. Particle size

distributions measurements were conducted for about two weeks at each location.

7.2.2 The topography and meteorology of the Brisbane region

The city of Brisbane (27’30ºS and 153ºE), located in Queensland Australia, is

surrounded by mountains from South to North, and faces the Pacific Ocean to the

East. Two major pollution sources affecting the Brisbane region are traffic exhaust

240

emissions and industrial emissions from aircraft, port and factories which are located

at the lower reaches of the Brisbane River. Synoptic wind patterns in the Brisbane

region are dominated by land and sea breezes.

The measurements were conducted at three locations in the Brisbane region, as

illustrated in Figure 7-1. The monitoring sites are representative of the urban (QUT),

urban-roadside (Woolloongabba) and semi-urban (Rocklea) environments of

Brisbane. A more detailed description of the study locations can be found in Cheung

et al. (2011b).

Figure 7-1. Map of monitoring sites.

7.2.3 Measurement techniques

Ultrafine particle (UFP) size distributions in the range 4-110 nm were measured

using a Scanning Mobility Particle Sizer (SMPS), which consists of an Electrostatic

241

Classifier (EC) (TSI 3080) and a Condensation Particle Counter (CPC) (TSI 3781).

The size distribution data were then used to calculate the number concentration of

nucleation mode particles (Nnuc, 4-25 nm) and freshly formed particles (N4-6, 4-6nm).

Sulphur dioxides (SO2) was measured by using a real-time gaseous analyser with a

lower detection limit of 0.5 ppb and a resolution of 1 ppb (Ecotech ML9850).

Meteorological parameters, including wind direction/speed, temperature, relative

humidity and solar radiation were also measured. Those data were collected and

validated by the Department of Environmental Resources and Management,

Queensland, Australia (DERM).

7.2.4 Data processing and analysis

Particle number concentrations were calculated by adding all of the particle counts in

the size bins corresponding to the size range investigated. The condensation sink (CS)

of ultrafine particles was calculated using particle size distribution data. Since the

time resolution of gaseous and meteorological data provided by DERM were given

in 30 min intervals, all raw particle size distribution data were transformed into 30

min averaged data for figure plotting, data processing and correlation analysis.

Correlations between the parameters were tested using the Pearson correlation test,

with a 95% confidence level (p < 0.05). The linearity of the tested parameters was

242

indicated by the coefficient of determination, r2, with a perfect linear correlation

between two parameters indicated by an r2 value close to 1.

A nucleation event is classified by the development of particle size distribution

where an increase of Nnuc is observed and followed by a particle growth process

indicated by the increase of geometric mean diameter (GMD) from the nucleation

mode (≤ 25 nm to an accumulation mode (> 25 nm). Furhter detail about the

classification of the nucleation event is referred to in Cheung et al. (2011a).

The [H2SO4] proxy was calculated using the following equation (Petäjä et al., 2009):

[H2SO4] proxy = k * [SO2] [SR] / [CS]

Where [SO2] is ambient sulphur dioxides, [SR] is solar radiation and [CS] is the

condensation sink of atmospheric particles. The [SO2] and [CS] were measured at the

study locations, while [SR] was measured at Rocklea and used for proxy calculation

at three sites. The pre-factor, k, need an actual measurement of H2SO4 to solve it, and

which is not available in this study.

243

7.3 Results and discussion

During the study period, total of 7 nucleation events were observed at three sites. The

events were observed on 11/6, 12/6 and 16/6 at Woolloongabba; on 26/6 and 5/7 at

Rocklea; and 17/7 and 18/7 at QUT. This section discusses the temporal variation of

Nnuc and H2SO4 proxy for the three sites, as well as on nucleation and non-nucleation

event days. The Nnuc and proxy values are then calculated for nucleation and non-

nucleation event days, followed by an examination of the relationship between the

two parameters during the nucleation events. Finally, the characteristics of the

nucleation events between three sites are discussed.

7.3.1 Temporal variations of Nnuc and H2SO4 proxy

Time series plots of Nnuc and the H2SO4 proxy observed at a) QUT, b)

Woolloongabba and c) Rocklea are illustrated in Figure 7-2a-c. It can be seen that

variations in the proxy generally coincide with the diurnal variation of solar radiation,

whose highest values were generally found around noontime at the three sites. It

should be noted that the observed SO2 levels at the semi-urban site of Rocklea were

lower than the detection limit of the instrument during some of the sampling period

and therefore, the H2SO4 proxy was not calculated for these sampling days (see

Figure 7-2c).

244

Figure 7-2. Time series plot of Nnuc particles and H2SO4 proxy measured at (a)

QUT, (b) Woolloongabba and (c) Rocklea.

245

During nucleation events days, the temporal variation of Nnuc and the proxy were

similar for QUT and Woolloongabba sites. Figure 7-3a is an example of a time

series plot of Nnuc and the proxy observed on 11/6 at Woolloongabba, which shows

that both the measured Nnuc and proxy peaked at 12:00 during the day. In contrast,

the peaks for Nnuc and the proxy did not coincide for the 5/7/2009 event at Rocklea,

where the peak of the proxy was observed at about 16:00, while the peak of Nnuc was

observed about 2 hours later (see Figure 7-3b). A similar trend was been seen in

Atlanta where the SO2 plume was observed about 1 hr prior to the peak of nucleation

mode particles (Woo et al., 2001). The author suggested that the SO2 plume took

time to undergo new particle formation, therefore accounting for the time lag

between peaks in SO2 and nucleation mode particles. However, it is unclear if the

peak in Nnuc observed at Rocklea was caused by a local or upwind nucleation event.

The observed GMD during the initial stages of nucleation should be close to a lower

detectable particle size limit of the particle counter (~ 4nm), if the particles are

formed close to the study region. The GMD of particles during the initial stage of

nucleation events at three sites were are presented in Table 7-1, from which it can be

seen that relatively higher initial GMD values were found at Rocklea (averaged

~19.6 nm) compared to QUT (averaged ~12 nm) and Woolloongabba (averaged ~ 16

nm). The higher initial GMD at Rocklea indicates that the Nnuc particles are likely to

246

be the result of a nucleation event initiated upwind Rocklea, which were later

transported to the measurement site. Furthermore, particle growth rates (GR) of the

nucleation events were also calculated and the average GR at the QUT,

Woolloongabba and Rocklea sites were found to be 3.2 nm h-1, 2.6 nm h-1 and 2.4 nm

h-1, respectively. The GR values were relatively similar for the three sites, which

imply that the nucleation events were governed by similar formation process in the

study region.

Figure 7-3. Time series plot of the Nnuc and the H2SO4 proxy measured on (a)

11/6/2009 at Woolloongabba and (b) on 5/7/2009 at Rocklea.

247

Sites Dates GR (nm h-1) GMD (nm)

QUT 17/7/2009 2.9 12.7

18/7/2009 3.5 11.3

Woolloongabba 11/6/2009 3.0 14.7

12/6/2009 2.9 18.5

16/6/2009 1.9 14.7

Rocklea 26/6/2009 2.3 20.7

5/7/2009 2.5 18.4

Table 7-1. Particle growth rates (GR) and geometric mean diameter (GMD)

calculated on the nucleation event days at three study sites.

7.3.2 H2SO4 proxy for nucleation event and non-event days

The daily median Nnuc and H2SO4 proxy concentrations were calculated for

nucleation and non-nucleation event days at three sites (Table 7-2). At all three sites,

the Nnuc was higher on nucleation event days, being 2.3 times, 3.0 times and 2.5

times higher for QUT, Woolloongabba and Rocklea, respectively. Similar results

were observed for the H2SO4 proxy at QUT, Woolloongabba and Rocklea, where

higher proxy values were found on nucleation event days. Although there has been a

longer time lag between the observation of Nnuc and proxy at Rocklea compared to

QUT and Woolloongabba, these observations have been commonly seen in previous

urban studies (Erupe et al., 2010). Those findings have suggested that the nucleation

events observed in this study region were affected by sulfuric acid.

248

Sites Period All days Event days Non-event days

QUT 13-30 July 2009 Proxy (ppb Wm-2 s) 10.2 (±6.2) 10.4 (±4.0) 10.2 (±6.5)

Nnuc (cm-3) 4.7 x 103 (±3.0 x 103) 10.4 x 103 (±0.9 x 103) 4.6 x 103 (±2.3 x 103)

Woolloongabba 11-25 June 2009 Proxy (ppb Wm-2 s) 11.2 (±5.9) 15.9 (±4.8) 9.6 (±4.9)

Nnuc (cm-3) 4.4 x 103 (±2.0 x 103) 10.8 x 103 (±1.4 x 103) 3.7 x 103 (±1.9 x 103)

Rocklea 26 June – 12 July 2009

Proxy (ppb Wm-2 s) 2.4 (±100.5) 7.7 (±3.3) 1.8 (±108.1)

Nnuc (cm-3) 0.7 x 103 (±1.3 x 103) 1.5 x 103 (±0.2 x 103) 0.6 x 103 (±1.3 x 103)

Table 7-2. Median of H2SO4 proxy and N4-6 particles at three sites (standard deviation in parenthese).

249

7.3.3 Correlation between N4-6 and H2SO4 proxy during nucleation events

Figure 7-4. Scatter plots for N4-6 and H2SO4 proxy during nucleation events at (a) QUT, (b) Woolloongabba and (b) Rocklea.

To further investigate the relationship between the H2SO4 proxy and the number

concentration of freshly formed particles, the linear relationship between N4-6 and the

proxy during a nucleation event was found using the coefficient of determination (r2),

which ranged from 0.69-0.77 and 0.26-0.51 for QUT and Woolloongabba,

respectively (see Figures 7-4 (a-b)). The results showed a strong to moderate linear

relationship between these two parameters during nucleation events, with the average

slope of the linear fittings being 1.6 and 0.7 for QUT and Woolloongabba,

respectively. The slope of the fitting was affected by the nucleation mechanism, such

that a slope approaching to 2 suggests a particle formation process by the kinetic

nucleation, while a slope near to 1 indicates that the process was controlled by

activation nucleation (Sipilä et al., 2010). Variations in the slope are affected by

atmospheric conditions, such as meteorological conditions and the availability of

precursor particles. The scatter plots also provide information on the H2SO4 proxy

threshold which is needed to produce the freshly formed particles (particles 4-6 nm

250

in size are used to represent freshly formed particles). As seen in Figure 7-4 (a – b),

approximately 1.0 ppb Wm-2 s and 0.5 ppb Wm-2 s of log[H2SO4 proxy] are required

to produce new particles at QUT and Woolloongabba, respectively.

7.3.4 Directional dependence of H2SO4 proxy

In the above sections, elevated H2SO4 proxies were found to be associated with the

new particle formation process at both urban and roadside sites. In order to

investigate the source region of the H2SO4 proxy, wind rose plots for QUT and

Woolloongabba are shown in Figure 7-5a and in Figure 7-5b for Rocklea. The

results showed that the H2SO4 proxy was relatively higher in S and NW directions

from QUT, and SW from Woolloongabba, while they were significantly higher in

NW direction from Rocklea.

Figure 7-5. Windrose plots of H2SO4 proxies observed at (a) QUT and Woolloongabba, and (b) Rocklea.

The H2SO4 proxy is a function of SO2, solar radiation and the condensation sink

of atmospheric particles. Therefore, to better understand the reason of the

directional dependence of the proxy values, wind rose plots of the individual

251

parameters are shown in Figure 7-6 (a-c). There was no strong dependence on

wind direction for SO2 and CS at the three sites (see Figure 7-6a), however

relatively higher SR was found to the S and NW of QUT and to the SW of

Woolloongabba, which resulted in a higher H2SO4 proxy in those directions. For

Rocklea, the observed SO2 mixing ratios were close to zero in most of the wind

directions, however an elevated value for the H2SO4 proxy was observed in the

NW direction. This finding implies that the ambient SO2 and CS of particles were

relatively uniform for each station. The directional dependence of solar radiation

was mainly influenced by daily wind circulation in the Brisbane region, with

westerly and southerly wind directions often associated with the highest solar

radiation during the day (see supplementary material Figure 7-S1 which shows

the temporal variation of wind vector and solar radiation during the selected

sampling days).

252

Figure 7-6. Windrose plots of (a) SO2, (b) Condensation sink and (c) Solar radiation observed at QUT, Woolloongabba and Rocklea.

253

7.4 Conclusion

In this study, the number concentration of ultrafine particles and sulphur dioxide

were measured in the Brisbane region during June-July 2009 and H2SO4 proxy

concentrations were calculated for the urban, urban-roadside and semi-urban sites. A

total of 7 nucleation events were observed at three stations during the sampling

period which provided enough data for the investigation of the influence of H2SO4

proxy in related to the nucleation event. The proxy was found to be a key factor

influencing nucleation growth events in the urban and urban-roadside sites,

displaying temporal variations similar to those for the number concentration of

nucleation mode particles at QUT and Woolloongabba (the peak in proxy

concentration occurred 1-2 hours prior to the peak in nucleation mode particles at

Rocklea). Statistical analysis showed that higher values for the H2SO4 proxy were

obtained during nucleation event days compared to non-event days at QUT and

Woolloongabba, and a moderate to strong linear relationship was found between the

H2SO4 proxy and freshly formed particles (although the same was not true for the

Rocklea site). The amount of the log[H2SO4 proxy] required to produce new particles

was found to be ~ 1.0 ppb Wm-2 s and below 0.5 ppb Wm-2 s, for the urban and

urban-roadside sites, respectively. Directional dependence of the proxy was found to

be related to the diurnal variation of solar radiation, which was associated with wind

circulation patterns in the study region. Similar particle growth rates were observed

at the three study locations, with an overall average value of 2.7±0.5 nm hr-1. This

result points to similar nucleation mechanisms dominating in the study region. A

relatively weak relationship was found between the proxy and particle number

concentration at the semi-urban site of Rocklea. This was likely to be because the

254

nucleation process was initiated upwind of the site and the new particles were

transported downwind to Rocklea.

Overall, this study found that the relationship between the proxy and particle number

concentrations varied across the urban, urban-roadside and semi-urban sites. Given

that no distinct relationship was observed between the proxy and particle number

concentration at the semi-urban site, the difference could be due to a time lag in the

nucleation process, which was initiated at the upwind area and later transported to

the downwind area. Further investigation on the new particle formation process by

other mechanisms, such as photochemical particle production by organic species and

ion-induced nucleation, are needed to better understand the new particle formation

contributed by different processes.

Acknowledgments

This project was supported by the Australian Research Council and Queensland

Transport through Linkage Grant LP0882544. We would also like to thank the

Department of Environmental Resource and Management, Queensland for providing

the air monitoring data; and the Queensland Bureau of Meteorology for providing the

meteorological data.

255

Supplementary Material

Figure 7-S1. Time series plots of wind vector (upper) and solar radiation (lower) observed on 27-29 July 2009 at QUT.

256

7.5 References

Cheung, H.C., Morawska, L. and Rostovski, Z.D.: Observation of new particle

formation in subtropical urban environment, Atmospheric Chemistry and Physics,

11, 3823-3833, doi:10.5194/acp-11-3823-2011, 2011a.

Cheung, H.C., Morawska, L., Ristovski, Z. and Wainwright, D.: Influence of

medium range transport of particles from nucleation burst on particle number

concentration within the urban airshed, submitted to Atmospheric Chemistry and

Physics Discussions, 2011b.

Cheung, H.-C., Wang, T., Baumann, K. And Guo, H.: Influence of regional pollution

outflow on the concentrations of fine particulate matter and visibility in the

coastal area of southern China, Atmospheric Environment, 39, 6463-6474, 2005.

Erupe, M.E., Benson, D.R., Li, J., Young, L.-H., Verheggen, B., Al-Refai, M.,

Tahboub, O., Cunningham, V., Frimpong, F., Viggiano, A.A. and Lee. S.-H.:

Correlation of aerosol nucleation rate with sulfuric acid and ammonia in Kent,

Ohio: An atmospheric observation, Journal of Geophysical Research, 115,

D23216, doi:10.1029/2010JD013942, 2010.

Kulmala, M., Arola, A., Nieminen, T., Riuttanen, L., Sogacheva, L., de Leeuw, G.,

Kerminen, V.-M. and Lehtinen, K.E.K.: The first estimates of global nucleation

mode aerosol concentrations based on satellite measurements, Atmospheric

Chemistry and Physics, 11, 10791-10801, doi:10.5194/acp-11-10791-2011, 2011.

Kulmala, M. and Kerminen, V.-M.: On the formation and growth of atmospheric

nanoparticles, Atmospheric Research, 90, 132-150, 2008.

Myhre, G.: Consistency between satellite-derived and modelled estimates of the

direct aerosol effect, Science, 325, 187-190, 2009.

Paasonen, P., Sihto, S-L., Nieminen, T., Vuollekoski, H., Riipinen, I., Plaß-Dülmer,

C., Berresheim, H., Birmili, W. and Kulmala, M.: Connection between new

particle formation and sulphuric acid at Hohenpeissenberg (Germany) including

the influence of organic compounds, Boreal Environment Research, 14, 616-

629,2009.

257

Petäjä, T., Mauldin, III, R.L., Kosciuch, E., McGrath, J., Nieminen, T., Paasonen, P.,

Boy, M., Adamov, A., Kotiaho, T. and Kulmala, M.: Sulfuric acid and OH

concentrations in a boreal forest site, Atmospheric Chemistry and Physics, 9,

7435-7448, doi:10.5194/acp-9-7435-2009, 2009.

Pope, C.A. and Dockery, D.W.: Health effects of fine particulate air pollution: lines

that connect, Journal of Air and Waste Management Association, 56, 709-742,

2006.

Ristovski, Z.D., Suni, T., Kulmala, M., Boy, M., Meyer, N.K., Duplissy, J.,

Turnipseed, A., Morawska, L., Baltensperger, U.: The role of sulphates and

organic vapours in growth of newly formed particles in a eucalypt forest,

Atmospheric Chemistry and Physics, 10, 2919-2926, 2010.

Salma, I., Borsós, T., Weidinger, T., Aalto, P., Hussein, T., Dal Maso, M. and

Kulmala, M.: Production, growth and properties of ultrafine atmospheric aerosol

particles in an urban environment, Atmospheric Chemistry and Physics, 11, 1339-

1353, doi:10.5194/acp-11-1339-2011, 2011.

Sipilä, M., Berndt, T., Petäjä, T., Brus, D., Vanhanen, J., Stratmann, F., Patokoski, J.,

Mauldin III, R.L., Hyvärinen, A.-P., Lihavainen, H. and Kulmala, M.: The role of

sulphuric acid in atmospheric nucleation, Science, 327, 1243-1246, 2010.

Wang, M. and Penner, J.E.: Aerosol indirect forcing in a global model with particle

nucleation, Atmospheric Chemistry and Physics, 9,239-260, 2009.

Woo, K.S., Chen, D.R., Pui, D.Y.H. and McMurry, P.H.: Measurements of Atlanta

aerosol size distributions: observations of ultrafine particle events, Aerosol

Science and Technology, 34, 75-87, 2001.

258

CHAPTER 8. GENERAL DISCUSSIONS.

8.1 Introduction

A study of the new particle formation was conducted in different locations of

Brisbane, Australia over the year of 2009. This study examined the temporal

variation of the ultrafine particle number concentration as well as its size distribution

in an urban area of Brisbane. The nucleation process observed in the urban area of

Brisbane was quantified and characterized with other atmospheric conditions such as

the meteorological and the precursor for new particle formation. We also evaluated

the impact of nucleation events on the number concentration of particles in

downwind area via air masses transportation. Furthermore, the influence of sulfuric

acid on the nucleation events was studied to characterize the relationship between the

sulfuric acid and the number concentration of nucleation mode particles. Overall, this

is the first study to examine new particle formation in a subtropical urban

environment in the southern hemisphere. This study has provided further insight to

understanding the nucleation processes in an urban environment with stronger

photochemical activity and raises concern on the impact of the transport of ultrafine

particles in rural areas associated with new particle formation events.

8.2 Principal significance of findings

The principal significance of the findings of this research is summarised below:

I) Particle detection efficiency of CPC’s for different ambient aerosol

composition and condensation medium.

- Relative detection efficiencies of the CPC’s were tested with different types

of particles including i) water soluble aerosols, NaCl; ii) water insoluble

259

aerosols, including DOS and laser toner particles; and iii) water and alcohol

soluble aerosols, citric acid. Particle detection at small sizes was strongly

dependant on the solubility of the target aerosol in the CPC condensation

medium especially near the lower nominal diameter limit of the CPC.

Strongly water soluble aerosols resulted in more efficient particle detection in

water based CPC’s compared to butanol based instruments; while water

insoluble aerosols, citric acid, a material displaying good solubility in both

water and butanol yielded comparable response in both types of instrument

within the nominal detection size range.

- BCPC/WCPC ratios were obtained for a range of particle number

concentrations. Apart from the effects of coincidence errors which occur in

all CPCs, the results did not show a strong correlation between the

BCPC/WCPC ratio and the particle number concentration. According to the

findings of this study, BCPC should be chosen for measuring ambient

concentrations of water insoluble particles. Both BCPC and WCPC appear to

be suitable for ambient particle concentration measurements provided that the

particles are known to be at least somewhat hygroscopic. WCPCs should be

used with care in ambient measurements and only after verifying that the

aerosol is not hydrophobic.

II) New particle formation in subtropical urban environment of Brisbane,

Australia.

260

- Annual average of the number concentrations of ultrafine, Aitken and

nucleation modes particles were 9.3 x 103, 3.7 x 103 and 5.6 x 103 cm-3,

respectively in urban area of Brisbane. The concentration level was

comparable to the urban environment of Northern Europe, while lower

than those polluted urban areas such as Yangtze River Delta, China and

Huelva and Santa Cruz de Tenerife, Spain.

- Statistical analysis did not show a significant seasonal variation of the

averaged particle number concentration in urban area of Brisbane,

however a relatively larger variation was observed in warmer season.

Diurnal variation of Aitken and nucleation modes particles showed

different patterns. The result suggested that the Aitken mode particles

were contributed by the direct diesel and petrol engine emissions while

the nucleation mode particles were mainly contributed by the vehicle

exhaust emissions during the morning and by photochemical production

during the noon time associated with high solar radiation.

- A total of 65 nucleation events were observed at CBD of Brisbane during

year 2009 in which 40 nucleation growth events and 25 nucleation burst

events were observed. The nucleation burst and growth events observed

in CBD of Brisbane were found to be strongly affected by the origin of

pollution plumes. All nucleation growth events were associated with

vehicle exhaust emissions, while the nucleation burst events were

associated with industrial emissions by comparison of the emission ratio

of gaseous pollutants.

261

- The mean particle growth rate of nucleation growth events was found to

be 4.6 nm hr-1 (ranged from 1.79 – 7.78 nm hr-1). This value was

comparable to other urban studies in United States. Furthermore, monthly

particle growth rates were found to be positively related to monthly solar

radiation (r = 0.76, p <0.05). These particle growth rate values were the

first reported for the subtropical urban area of Australia.

III) Influence of medium range transport of particles from nucleation burst on

particle number concentration within the urban airshed.

- Regional nucleation was the major contributing factor to the total amount

of the PNCs at the urban (QUT), roadside (Woolloongabba) and semi-

urban (Rocklea) areas of Brisbane. Under the influence of regional

nucleation events, the highest relative fraction of PNC observed at QUT,

Woolloongabba and Rocklea were found to be 12%, 9% and 14%,

respectively. These highest relative fractions of PNC observed during the

morning traffic peak period were approximately 5.1-5.5% at QUT and

Woolloongabba during the weekdays.

- The PNC level of semi-urban area was significantly affected by upwind

urban pollution during the nucleation burst events. The average

enrichment factor was found to be 15.4 for the PNC observed at Rocklea,

when this region receiving the upwind urban pollution.

262

IV) Influence of sulfuric acid on the new particle formation

- The temporal variations of the H2SO4 proxies and the Nnuc were found to

be in phase during the nucleation events for the urban (QUT) and

roadside (Woolloongabba) areas. In contrast, the peak of proxy was 1-2 hr

prior to the observation of peak of Nnuc at the downwind semi-urban

(Rocklea) area of Brisbane.

- Moderate to strong linear relationship were found between the H2SO4

proxy and the freshly formed particles, also the amount of the log[H2SO4

proxy] required to produce a new particle were found to be ~1.0 ppb Wm-

2 s and below 0.5 ppb Wm-2 s, respectively for the urban and semi-urban

areas.

- Similar particle growth rates were obtained at three study locations with

an average value of 2.7±0.5 nm hr-1 during the nucleation events. This

result implied that a similar nucleation mechanism dominated in the study

region which was strongly influenced by the H2SO4 proxy. Although,

poor relationship between the proxy and particle number concentration

was found in the semi-urban area, Rocklea. It can be explained that the

nucleation process was initiated at the upwind area, and transported to

Rocklea. This explanation was suggested by the higher geometric mean

diameter value observed during the nucleation event and the time lagged

relationship between the H2SO4 proxy and particle number concentration

observed at Rocklea.

263

8.3 Scientific Recommendations

I) To further understand the regional distribution of the ultrafine particle

number concentrations and the source of the regional nucleation in

Brisbane region, a long- term monitoring network of the size distribution

of the atmospheric particles, the gaseous pollutants including SO2, NH3

and VOCs at different locations of Brisbane is suggested. The air

monitoring sites should be selected at locations which can represent the

atmospheric settings of the background, urban area of Brisbane city and

downwind rural areas in the northeast and southeast Brisbane. The

background air monitoring site should be able to capture the clean marine

air masses, where we can estimate the contribution of the pollution from

Brisbane region compared to that from marine. Also the sites at rural

downwind areas of the northeast and southeast Brisbane are selected to

study the influences of the urban pollutions on the continental background.

Also the lowest detection limit of the SO2 analyzer should be capable to

measure the mixing ratio in ppt level, since lesser than 1 ppb mixing ratio

of SO2 was often observed in Brisbane region. This suggestion can

improve the data quality in the future study of regional nucleation process

in Brisbane region.

II) In this study, sulfuric acid was found to be significantly influencing the

number concentration of the nucleation mode particles and strongly

related to the new particle formation in urban and roadside environments

of Brisbane. In addition to the contribution of sulfuric acid on nucleation

process, the influences of other components such as organic species and

264

charged particles on nucleation are needed to complete the picture of the

new particle formation process in the study region. With the information

provided from this further study, we can estimate the contribution of the

particles formed by each mechanism.

III) The study of the new particle formation using the Chemical Transport

Model (CTM) is recommended. CTM considers the physical and

chemical theories of the nucleation process simulation rather than the

empirical mathematical model which only explains the input data by

using a mathematical approach. Comparison between the observation and

simulation results of the particle dynamics can assess the accuracy of

model results to simulate the nucleation process. It provides insight to the

unknown mechanisms/ knowledge about nucleation if the observed and

simulated results do not match. Also, the particle number concentrations

will be predicted by using the modelling with the meteorological

conditions after the validation of the modeling result.

265

APPENDIXES

A) Manuscript of “Jayaratne, E.R., Johnson, G.R., McGarry, P., Cheung, H.C.

and Morawska, L. (2011). Characteristics of Airborne Ultrafine and Coarse

Particles during the Australian Dust Storm of 23 September 2009.

Atmospheric Environment, 45, 3996 – 4001.”

266

Characteristics of Airborne Ultrafine and Coarse

Particles during the Australian Dust Storm of 23

September 2009

E.R. Jayaratne, G.R. Johnson, P. McGarry, H.C. Cheung and L. Morawska*

International Laboratory for Air Quality and Health

Queensland University of Technology

GPO Box 2434, Brisbane, QLD 4001, Australia

Published by Atmospheric Environment

* Corresponding author contact details:

Tel: (617) 3138 2616; Fax: (617) 3138 9079

Email: l.morawska@qut.edu.au

267

Abstract

Particle number concentrations and size distributions, visibility and particulate mass

concentrations and weather parameters were monitored in Brisbane, Australia, on 23

September 2009, during the passage of a dust storm that originated 1400 km away in

the dry continental interior. The dust concentration peaked at about mid-day when

the hourly average PM2.5 and PM10 values reached 814 and 6460 µg m-3, respectively,

with a sharp drop in atmospheric visibility. A linear regression analysis showed a

good correlation between the coefficient of light scattering by particles (Bsp) and

both PM10 and PM2.5. The particle number in the size range 0.5-20 µm exhibited a

lognormal size distribution with modal and geometrical mean diameters of 1.6 and

1.9 µm, respectively. The modal mass was around 10 µm with less than 10% of the

mass carried by particles smaller than 2.5 µm. The PM10 fraction accounted for about

68% of the total mass. By mid-day, as the dust began to increase sharply, the

ultrafine particle number concentration fell from about 6x103 cm-3 to 3x103 cm-3 and

then continued to decrease to less than 1x103 cm-3 by 14h, showing a power-law

decrease with Bsp with an R2 value of 0.77 (p<0.01). Ultrafine particle size

distributions also showed a significant decrease in number during the dust storm.

This is the first scientific study of particle size distributions in an Australian dust

storm.

Keywords: Dust storm, Particle Concentration, Particle Size, Visibility, Air Pollution

268

1. Introduction

Dust storms occur when high winds caused by pressure gradients whip up loose soil

over a large area and transport it across the land. When the wind speed reaches a

threshold value, sand and dust particles on the surface of the ground begin to vibrate

and are ejected into the air – a process known as ‘saltation’. The impact of these

windborne particles on the surface ejects yet more particles and causes a chain

reaction. Ejected sands and dust can be transported by wind to great distances and, In

addition to reduced visibility that affects air and road transport, dust storms cause

soil erosion and loss of organic matter and nutrients from the soil (Wang et al., 2006).

From the point of view of human health, people with breathing-related problems,

such as asthma and emphysema, have been known to experience difficulties during

severe dust storms. Lei et al (2004) demonstrated that particulate matter in an Asian

dust storm increased lung inflammation and injury in pulmonary hypertensive rats.

However, other studies have shown that human mortality rates were not elevated

during dust storm days and attributed this to the absence of toxicity in crustal

particles (Hefflin et al., 1994; Schwartz et al., 1999).

Fine particles in the air are scavenged by larger particles. This process of

coagulation can lead to a shift of average particle size to larger values, especially

when the number concentration of particles is high (Matsoukas and Friedlander,

1991). Urban environments are dominated by particles from motor vehicle exhaust,

with the large majority of them being in the ultrafine size range, that is - smaller than

0.1 µm (Shi et al., 1999). A detailed account of the characteristics of ultrafine

particles in urban environments may be found in the two recent reviews by

269

Morawska et al. (2008) and Kumar et al. (2010). Thus, the passage of a dust storm

across a major city offers an ideal opportunity to investigate the coagulation process

between two distinctly different particle size groups in the outdoor environment.

Air quality monitoring stations routinely measure particulate mass in accordance

with the respective national ambient air quality standard requirements and normally

record this quantity as PM10 or PM2.5 - the mass concentration of particles smaller

than 10 µm or 2.5 µm, respectively. This has enabled a considerable amount of

research addressing particle mass concentrations during the passage of dust storms.

For example, Zhang et al (2006) monitored particles in the 20 March 2002 dust storm

in Beijing, China, and showed that the peak total suspended particle concentration

reached 12,000 µg m-3 while the mass concentrations of coarse particles accounted

for 91% of the total, compared to 61% on non-dust storm days. Choi and Choi (2008)

measured particulate mass concentrations at the ground during a dust storm in

Kangnung, Korea on 8 March 2004 and showed that PM10 concentrations reached

340 µg m-3. They also found that most of the particles were in the range between

PM2.5 and PM10. Several other studies have confirmed that the average particle size

in a dust storm occur in the size range 2-6 µm (Abdulla et al, 1988; Mikami et al,

2005; Kobayashi et al, 2007). Wang et al (2008) used aircraft measurements during

the 2006 dust storms over the coastal areas in Northern China and reported that

number concentrations of ultrafine particles exceeded 105 cm-3. While, data on

particle number distributions in dust storms is sparse, measurements of ultrafine

particles during dust storm episodes is highly limited.

270

The continental interior of Australia is a major global source region for atmospheric

dust. However, unlike dust and sand storms that occurs regularly in many parts of the

world such as in Northern China and the Sahara, Australian dust storms require a

specific sequence of environmental conditions. During heavy rain episodes, flood

waters from Queensland flow south and deposit large quantities of fluvial sediments

over a large area of the continental interior in and around the Lake Eyre Basin (see

map in Fig 1). Such intense flood events followed by prolonged drought conditions

can then lead to a significant erosion of alluvial dust with the onset of strong winds

that generally occur around September-November (Mitchell et al., 2010). In contrast

to the composition of dust in other parts of the world, Australian desert dust is

particularly rich in iron which gives it its typical reddish hue, while halites from dry

salt lakes comprise about 0.5% by mass (Radhi et al, 2010).

2. Methods

2.1 Overview and Aims

On the 22nd and 23rd September 2009, a large amount of dust was swept up in

strong winds caused by an intense low-pressure zone near Lake Eyre and was very

quickly carried eastwards and northwards (Fig 1). The ensuing dust storm was

estimated to be the worst in 70 years (AGBM, 2010). At its peak, the dust plume was

more than 3400 km long and stretched from southern New South Wales to far north

Queensland. Airborne particle concentrations of over 15,000 µg m-3 were recorded

at many locations. It is estimated that 1.6 x 109 kg of dust was removed from the

continental interior which at one time was losing 7.5 x 107 kg h-1 of dust off its

eastern coastline (Leys et al., 2009). The region affected by the dust included the

state capitals of Sydney and Brisbane. The dust reached Brisbane at a distance of

271

about 1400 km from its source at around 11 am on the 23rd and by 12 noon, resulted

in a drop in visibility to a few metres. Dust in the air gave the environment an eerie

red-orange colour and the air temperature dropped by several degrees.

The International Laboratory for Air Quality and Health (ILAQH) at the Queensland

University of Technology was carrying out measurements of airborne particles at the

top of two buildings in the Central Business District of Brisbane when the dust storm

passed over the city. The aim of this paper is to use the results obtained to investigate

the physical characteristics of the dust particles as well as to assess the impact of the

dust on the regular ultrafine particle number and mass concentrations in an urban

environment.

2.2 Monitoring Sites

As this was not a planned experiment, not all instruments were operative right

through the dust storm and not all at the same location. The measurements were

carried out at two locations, to be denoted Site A and Site B, being two buildings in

the Brisbane Central Business District (CBD), separated by a distance of about 0.5

km. Site A was located in a six-floor building within the Gardens Point campus of

the Queensland University of Technology, approximately 100m away and midway

between a busy freeway and the City Botanical Gardens. The air was sampled from

outside a 6th floor window. This site also included an air monitoring station operated

by the Queensland Department of the Environment and Resource Management

(DERM). Site B was located in a five-floor building situated next to a city street with

272

the air sampled from outside the 5th floor. Therefore, both monitoring sites could be

regarded as urban environments, normally dominated by vehicular emissions.

2.3 Instrumentation

The following particle measuring instruments were used in this study:

The TSI 3320 Aerodynamic Particle Sizer (APS) is an optical time-of-flight

spectrometer that provides real time high-resolution particle sizing from 0.5 to 20 µm.

A complete size distribution, in 52 size bins within the detection range, was obtained

every 1 min.

The TSI 3782 water-based Condensation Particle Counter (CPC) measures ultrafine

particle number concentration down to a size of 10 nm at concentrations up to 5x104

cm-3. Readings were taken every 1 s and the software was programmed to log

average values at intervals of 5 s.

The TSI 8520 DustTrak Aerosol Monitor is a portable laser photometer that

measures and records airborne dust concentration from 1 to 105 µg m-3. The

DustTrak is calibrated against a gravimetric reference using the respirable fraction of

standard ISO Arizona test dust which has a wide size distribution covering the entire

size range of the DustTrak and is representative of a wide variety of ambient aerosols

(TSI, 1997). An inlet impactor was used to restrict the sampled particle mass to an

upper size of 2.5 µm (PM2.5). Readings were taken every 1 s and the instrument was

set to log average values at intervals of 30 s.

Using PM2.5 data obtained during the dust storm, the DustTrak was calibrated against

a tapered element oscillating microbalance (TEOM) located at Site A. The TEOM is

273

an instrument that is certified by the US-EPA for gravimetric measurements of

particulate matter in ambient air. These results are shown in the Supplementary

Material of this paper.

A TSI 3936 Scanning Mobility Particle Sizer (SMPS) comprising a 3080

electrostatic classifier and a 3010 CPC was used to obtain particle size distributions

in the size range 4 to 100 nm in 91 size bins. A complete scan was derived every 10

min in real time.

Table 1 shows the location and times of operation of the various instruments. While

the APS and SMPS were located at Site A, the CPC and DustTrak were located at

Site B. These locations were not selected but, with the exception of the APS, the

instruments happened to be operating there on other projects as the dust storm

approached. It is unfortunate that the APS was not switched on until 16h.

Meteorological, visibility and PM10 concentrations were monitored at the roof level

of the building at the DERM air monitoring station at Site A. The meteorological

parameters included air temperature, relative humidity, wind speed and direction.

Visibility was monitored with an integrating nephelometer that measured the

atmospheric light scattering coefficient of particles (Bsp) and expressed it in Mm-1.

Particulate matter concentration in the form of PM10 was recorded with a TEOM at

Site A. Hourly average data of the meteorological conditions, visibility and PM10

values were also obtained from several ground-level DERM monitoring stations

around the city of Brisbane.

274

2.4 Data Analysis

The data on the CPC, APS, DustTrak were logged in real time at 1 s intervals. The

DERM data were available as 30 min averages. Linear regression analysis was used

to determine the correlation coefficients between half-hourly PM10 and PM2.5 values

and the corresponding Bsp values. The SMPS and APS data were processed and

analysed using Aerosol Instrumentation Manager Software from TSI.

3. Results and Discussion

3.1 Overview of the dust episode

The morning of the 23rd September 2009 was a typical fine spring day in Brisbane.

At 8 am, the air temperature was 23ºC and the relative humidity just over 60%. A

steady gentle breeze of 0.4 m s-1 blew in from the west. Ambient particle

concentration was normal with a PM10 level of 21 µg m-3 and a Bsp of 22 Mm-1 at

Site A. However, by about 10 am, with dust being transported in by the westerly

winds, the PM10 had exceeded 100 µg m-3, while the visibility had deteriorated,

almost doubling the Bsp to 41 Mm-1. The Australian ambient air quality standard for

PM10 averaged over 24 hours is 50 µg m-3. By 11 am, the effects of the dust storm

were clearly visible. Conditions continued to deteriorate rapidly in the next hour.

Maximum dust levels were observed near mid-day with the Bsp exceeding 1000

Mm-1. Thereafter, the visibility continued to improve steadily with the Bsp dropping

sharply until 17h and at a slower rate thereafter. By midnight, there was still a

considerable amount of dust in the air, as evidenced by the Bsp value of 83 Mm-1.

The real time variation of Bsp in Figure 2 shows the passage of the dust storm over

Brisbane. It is instructive to note that the Bsp in Brisbane on a normal day is 10-30

Mm-1.

275

3.2 DustTrak Accuracy

Figure 3 shows the half-hourly averaged PM2.5 data from the DustTrak at Site B and

the TEOM at Site A, obtained between 12:30 h and 15:30 h which corresponds to the

time period when the dust was most concentrated on the day of the dust storm. The

two parameters are plotted against each other. Despite the separation of about 0.5 km

between the two sites, the slope of the best line is very close to 1 (0.99 with R2 =

0.99) showing excellent agreement between the two parameters. This result indicates

that the material of the dust was very similar to the Arizona Dust that is used to

calibrate the DustTrak (TSI, 1997) and provides confidence that the DustTrak data

may be used as a reasonably accurate measure of the PM2.5 particulate matter

concentration in the dust storm.

3.3 Particulate Mass Concentrations

Figure 4 shows the hourly average particulate mass concentrations between 10h and

16h. The time axis shows the end-hour of each data bin. Thus, for example, the

maximum average PM2.5 and PM10 values of 814 and 6460 µg m-3, respectively,

were observed during the hour between 12-13h. It is clear that most of the particulate

mass in the size range below 10 µm lay between 2.5 µm and 10 µm.

3.4 Larger Particles

Figure 5(a) shows the particle number size distribution as measured with the APS at

16h. Note that particles smaller than 0.5 µm are not included in this figure. The total

APS particle number concentration was 17.3 cm-3 and exhibited a near-lognormal

size distribution with modal and geometrical mean diameters of 1.6 and 1.9 µm,

276

respectively. The particle volume size distribution (Figure 5(b)) shows that most of

the particle volume and, hence, mass was contributed by particles larger than 2.5 µm.

The modal mass was around 10 µm. A cumulative analysis showed that less than

10% of the mass was carried by particles smaller than 2.5 µm, while the PM10

fraction accounted for about 68% of the total mass. However, these comparisons

should be treated with caution, as the APS software calculates the volume from the

size assuming that the particles are spherical.

3.5 Ultrafine Particle Number Concentration

Next, we look at the impact of the dust on the ultrafine particle number

concentrations. It has been shown that the large majority of ultrafine particles in

urban settings are combustion aerosols from vehicle emissions (Shi et al., 1999;

Wahlin et al., 2001). Most of these particles are smaller than 200 nm, which is less

than one-tenth the size of particles in the dust storm. This gives rise to a process of

polydisperse coagulation whereby smaller particles diffuse to the surface of larger

particles. A tenfold difference in particle size produces a threefold increase in

coagulation (Hinds, 1982). Coagulation generally leads to a reduction in small

particle number with no change to the particle mass concentration.

On dust-free days, prior to the event day, the ultrafine particle number concentrations

at both measurement sites were typically of the order of 1x104 cm-3. Average values

peaked between 1x104 and 3x104 cm-3 during the peak traffic hours and dropped to

about 5x103 cm-3 in the early hours of the morning. The mean daytime concentration

on the day immediately prior to the dust storm, 22nd September, was 1.2x104 cm-3.

On the 23rd, after a normal peak number concentration of about 1.5x104 cm-3 during

277

the morning traffic peak, the concentration as measured by the CPC at Site B

stabilised at around 6x103 cm-3 by mid-morning. Figure 6 shows the ultrafine particle

number concentration together with the PM2.5 measured at the same location between

10h and 15h. The PM2.5 curve clearly shows the arrival and passage of the dust. At

11h, as the PM2.5 value began to increase sharply, the particle number concentration

showed the expected decrease. Particle number concentration values fell from about

5x103 cm-3 at 11h to less than 3x103 cm-3 by 12h at the peak of the storm. This initial

sharp decrease in ultrafine concentration coincided with the arrival of the main dust

storm peak. However the ultrafine particle number concentration then continued to

decrease at a slower rate even after the dust concentration had begun to decline.

These observations suggest that coagulation scavenging of the ultrafine particles was

accompanied by a second unidentified process, and that both acted simultaneously to

reduce the ultrafine particle number concentration. Right through the time period

depicted in Figure 6, the wind remained WSW (250ºTN ± 15º) at a fairly steady 4.5

± 0.5 m s-1. Air temperature was 26º ± 1º while the relative humidity dropped

steadily from about 50% at 10h to about 16% at 15h. Thus, it is unlikely that any

change in particle number concentration or particle size was due to a changing air

mass other than for the dust from the south-west. In support of the modelling studies

that have shown that fine mode particles are scavenged by larger particles in the

environment (Ackermann et al., 1998; Jung et al., 2002), the present study

demonstrates that the effect can be very effective in a real life dust storm situation.

Next, we investigate the effect of PM10 on the ultrafine particle number concentration.

Figure 7 shows the hourly average ultrafine particle number concentration against the

corresponding PM10 concentration between 7h and 16h on the 23rd September. The

278

graph shows a sharp decrease in ultrafine particle number concentration as the PM10

concentration increased. The ultrafine particle number concentration showed a

power-law decrease with PM10 with an R2 value of 0.73 (p<0.01).

Bsp and PM Correlations

In Figure 8, we look at correlations between the light scattering coefficient of

particles (Bsp) and the particulate matter concentrations, PM10 and PM2.5. Figure 8(a)

shows the hourly particulate matter concentrations against Bsp, between 7h and 16h

on the 23rd September. A linear regression analysis showed a good correlation

between Bsp and both PM10 and PM2.5, with an R2 value of 0.98 (p<0.01) for each.

Fine particles in the size range 0.4 to 0.7 µm that corresponds to the visible spectral

wavelength are more efficient at scattering light than other sizes. As this range falls

within the size ranges of both PM10 and PM2.5, it is not surprising that they both

correlate well with Bsp.

Figure 8(b) shows the corresponding hourly average ultrafine particle number

concentration measured by the CPC as a function of Bsp. In accordance with Figure

4, this graph showed a sharp decrease in particle number concentration as the dust

arrived and the Bsp increased. The particle number concentration showed a power-

law decrease with Bsp with an R2 value of 0.77 (p<0.01).

As stated, the APS was switched on soon after 16h and continued to sample until 10h

on the next day. Figure 9 shows the particle number concentration and geometrical

mean diameter as measured between 16h and 24h. If we disregard the peak at around

279

18h, which was no doubt associated with increased road dust from vehicular traffic

in the evening rush hour, we see that the number concentration dropped steadily from

about 16 cm-3 to about 5 cm-3 during this period. Over the same period, the hourly

average Bsp and PM10 dropped from about 280 Mm-1 to 85 Mm-1 and from 1300 µg

m-3 to 400 µg m-3, respectively (Figures 2 and 4). It is interesting to note that, from

16h to 24h, all three parameters decreased by the same factor of just over 3. This was

only possible if the particle size did not show a significant difference in time and this

is substantiated by the time series graph of the particle diameter in Figure 9. There

are twelve outlier points seen in the graph close to 20.30h and 23.00h, which are

clearly due to spurious effects as they fall well above the expected variation of the

long term readings. Excluding these twelve points, the APS particle geometrical

mean diameter over this period was 1.9 ± 0.1 µm, where the uncertainty is the

standard deviation. Thus, we see that the variation of the particle diameter about its

mean value was no more than about 5%. The slight increase at night-time was

probably caused by hygroscopic growth and/or particle coagulation, both phenomena

that have been observed in the environment. Right through the time period depicted

in Figure 9, the wind remained WSW (255ºTN ± 10º) at a fairly steady 4.5 ± 1.0 m s-

1.

3.6 Ultrafine Particle Size Distribution

Figure 10 shows the ultrafine particle size distributions before and during the dust

storm as measured by the SMPS. Each curve is the average of 12 scans over a full

two-hour period. The upper curve reflects the size distribution prior to the arrival of

the dust between 8 and 10 am. During this time, the particle mass concentration

values were as on any other day, indicating no excess dust in the atmosphere. As the

280

dust arrived, the ultrafine particle numbers decreased. This reduction was very

pronounced for ultrafine particles in the size range close to 100 nm but decreased at

smaller sizes, with no significant drop in number being detected for particles smaller

than 20 nm. This latter range is generally occupied by nanoparticles produced by

nucleation of the gaseous products of motor vehicle emissions (Kittelson et al., 2004).

Given the urban location of the measurement site, it is probable that local traffic

emissions were responsible for maintaining concentrations in this size range and that

those emissions had been produced too recently to have been significantly affected

by the surrounding dust particles.

4. Conclusions

At the peak dust time, the hourly-averaged PM2.5 and PM10 values were 814 and

6460 µg m-3, respectively, with the light scattering coefficient of particles, Bsp,

exceeding 1000 Mm-1. A linear regression analysis showed a good correlation

between Bsp and both PM10 and PM2.5. The PM10 fraction accounted for about 68%

of the total mass. The particle number concentration measured by the APS exhibited

a lognormal size distribution with modal and geometrical mean diameters of 1.6 and

1.9 µm, respectively. The modal mass was around 10 µm with less than 10% of the

mass carried by particles smaller than 2.5 µm. The ultrafine particle number

concentrations fell sharply as the dust storm passed over - from about 6x103 cm-3 to

about 3x103 cm-3 as the dust peaked and then continued to decrease to less than

1x103 cm-3 over the next two hours. Through our observations, we have also shown

that the number concentration of ultrafine particles in the environment is

significantly suppressed due to scavenging by larger particles during a dust storm.

The ultrafine particle number concentration showed a power-law decrease with PM10

281

with an R2 value of 0.73 (p<0.01). We believe that this is the first report of the

particle size distribution in an Australian dust storm.

Acknowledgements

We would like to thank the Queensland Department of the Environment and

Resource Management (DERM) for providing the hourly average PM and

meteorological data for the period of the dust storm.

282

References

Abdulla, S.A., Al-Rizzo, H.M. and Cyril, M.M., 1988. Particle size distribution of

Iraqi sand and dust storms and their influence on microwave communication

systems. IEEE Trans Antennas and Propagation 36, 114-126.

Ackermann, I.J., Hass, H., Memmesheimer, M., Ebel, A., Binkowski, F.S. and

Shankar, U., 1998. Modal aerosol dynamics model for Europe: Development

and first applications. Atmos Environ 17, 2981-2999.

AGBM, 2010. Annual Australian Climate Statement, 2009. Australian Government

Bureau of Meteorology. Issued 5 Jan 2010.

Choi, H. and Choi, D.S., 2008. Concentrations of PM10, PM2.5 and PM1 influenced by

atmospheric circulation and atmospheric boundary layer in the Korean

mountainous coast during a duststorm. Atmos Res 89, 330-337.

Hefflin, B.J., Jalaludin, B., McClure, E., Cobb, N., Johnson, C.A., Jecha L. and Etzel,

R.A., 1994. Surveillance for dust storms and respiratory diseases in

Washington State, 1991. Arch Environ Health 49, 170-174.

Hinds, W.C., 1982. Aerosol Technology. New York, John Wiley.

Jung, C.H., Kim, Y.P. and Lee, K.W., 2002. Simulation of the influence of coarse

mode particles on the properties of fine mode particles. J Aerosol Sci 33, 1201-

1216.

Kittelson, D.B., Watts, W.F. and Johnson, J.P., 2004. Nanoparticle emissions on

Minnesota highways. Atmos Environ 38, 9-19.

283

Kobayashi, H., Arao, K., Murayama, T., Iokibe, K., Koga, R. and Shiobara, M., 2007.

High resolution measurement of size distributions of Asian dust using a

Coulter multisizer. J Atmos Oceanic Technol 24, 194-205.

Kumar, P., Robins, A., Vardoulakis, S., Britter, R., 2010. A review of the

characteristics of nanoparticles in the urban atmosphere and the prospects for

developing regulatory control. Atmos Environ 44, 5035-5052.

Lei, Y.C., Chan, C.C., Wang, P.Y., Lee, C.T. and Cheng, T.J., 2004. Effects of Asian

dust event particles on inflammation markers in peripheral blood and

bronchoalveolar lavage in pulmonary hypertensive rats. Environ Res 95, 71-76.

Leys, J., Heidenreich, S. and Case, M., 2009. DustWatch interim report 22-23rd

September. DustWatch, Department of Environment and Climate Change,

NSW and Griffith University, Queensland, Australia.

Matsoukas, T. and Friedlander, S.K., 1991. Dynamics of aerosol agglomerate

formation. J. Colloid Interface Sci 146, 495–506.

Mikami, M., Aoki, T., Ishizuka, M., Yabum, S., Yamada, Y., Gao, W. and Zeng, F.,

2005. Observation of number concentration of desert aerosols in the South of

the Taklimakan Desert, China. J Meteorol Soc Japan 83A, 31-43.

Mitchell, R.M., Campbell, S.K. and Qin, Y., 2010. Recent increase in aerosol loading

over the Australian arid zone. Atmos Chem Phys 10, 1689-1699.

Morawska, L., Ristovski, Z., Jayaratne, E.R., Keogh, D.U. and Ling, X., 2008.

Ambient nano and ultrafine particles from motor vehicle emissions:

characteristics, ambient processing and implications on human exposure.

Atmos Environ 42, 8113-8138.

284

Radhi, M., Box, M.A., Box, G.P., Mitchell, R.M., Cohen, D.D., Stelcer, E. and

Keywood, M.D., 2010. Size-resolved mass and chemical properties of dust

aerosols from Australia’s Lake Eyre Basin. Atmos Environ 44, 3519-3528.

Schwartz, J., Norris, G., Larson, T., Sheppard, L., Claiborne, C. and Koenig, J., 1999.

Episodes of high coarse particle concentrations are not associated with

increased mortality. Environ Health Persp 107. 339-342.

Shi, J. P., Khan, A. A. and Harrison, R. M., 1999. Measurements of ultrafine particle

concentrations and size distribution in the urban atmosphere. Sci Total Environ

235. 51-64.

TSI, 1997. DustTrak Aerosol Monitor, Theory of Operation. Health and Safety

Application Note ITI-036. Revised 11/14/97. TSI Incorporated, St Paul MN,

USA.

Wåhlin, P., Palmgren, F. and Van Dingenen, R., 2001. Experimental studies of

ultrafine particles in streets and the relationship to traffic. Atmos Environ 35,

S63-S69.

Wang, X., Oenema, O., Hoogmoed, W.B., Perdok, U.D. and Cai, D., 2006. Dust

storm erosion and its impact on soil carbon and nitrogen losses in northern

China. Catena 66, 221-227.

Wang, W., Ma, J., Hatakeyama, S., Liu, X., Chen, Y., Takami, A., Ren, L. and Geng,

C., 2008. Aircraft measurements of vertical ultrafine particles over Northern

China coastal areas during dust storms in 2006. Atmos Environ 42, 5715-5720.

285

Zhang, R., Wang, Z., Shen, Z., Yabuki, S. and Kanai, Y., 2006. Physicochemical

characterization and origin of the 20 March 2002 heavy dust storm in Beijing.

Aerosol and Air Quality Res 6, 268-280.

286

Tables

Instrument Site Period Operated on 23/09/2009 APS A 16h – 24h SMPS A All day CPC B 0h - 15h DustTrak B All day TEOM A All day Nephelometer A All day Met Parameters A All day

Table 1: The instruments, their locations and times of operation.

287

Figure Captions

1. Map of Australia, showing the source and dispersion of dust. All sampling

was carried out in Brisbane.

2. Light scattering coefficient of particles (Bsp) as a function of time (Site A).

3. Half-hourly averaged PM2.5 data from the DustTrak and the TEOM during the

dust storm, plotted against each other.

4. Hourly average PM10 and PM2.5 concentrations as a function of time (Site A).

5. Particle number (a) and volume (b) size distributions measured by the APS

(Site A).

6. Ultrafine particle number concentration together with the PM2.5 measured at

Site B.

7. Hourly average ultrafine particle number concentrations as a function of the

PM10 concentration.

8. Hourly average (a) particulate matter and (b) ultrafine particle number

concentrations shown as a function of the light scattering coefficient of

particles, Bsp.

9. Particle number concentration and geometrical mean diameter as measured

by the APS.

10. Ultrafine particle size distributions before and during the dust storm.

288

Figures

Fig 1

289

Fig 2

0

200

400

600

800

1000

1200

06:00 09:00 12:00 15:00 18:00 21:00 24:00

Bsp

(Mm

-1)

Time

290

Fig 3

y = 0.99x - 25.67R² = 0.99

0

200

400

600

800

1000

0 200 400 600 800 1000

Dust

Trak

PM

2.5

(µg

m-3

)

TEOM PM2.5 (µg m-3)

291

Fig 4

0

2000

4000

6000

10:00 11:00 12:00 13:00 14:00 15:00 16:00

Part

icul

ate

Mat

ter C

onc

(µg

m-3

)

Time

PM10

PM2.5

292

Fig 5

0

10

20

30

0.1 1 10

APS P

artic

le N

umbe

r Con

cent

ratio

n

(dN

/dlo

gD c

m-3

)

(a)

0

200

400

600

0.1 1 10

APS P

artic

le V

olum

e Co

ncen

trat

ion

(dV/

dlog

D µ

m3

cm-3

)

Particle Size D (µm)

(b)

293

Fig 6 (Colour)

0

300

600

900

1200

0.0E+00

3.0E+03

6.0E+03

9.0E+03

1.2E+04

09:00 10:30 12:00 13:30 15:00

Particulate Matter PM

2.5 Conc (µg m-3)

Ultr

afin

e Pa

rtic

le N

umbe

r Con

c (c

m-3

)

Time

PNC

PM2.5

294

Fig 6 (B & W)

0

300

600

900

1200

0.0E+00

3.0E+03

6.0E+03

9.0E+03

1.2E+04

09:00 10:30 12:00 13:30 15:00

Particulate Matter PM

2.5 Conc (µg m-3)

Ultr

afin

e Pa

rtic

le N

umbe

r Con

c (c

m-3

)

Time

PNC

PM2.5

295

Fig 7

0.0E+00

2.0E+03

4.0E+03

6.0E+03

8.0E+03

1.0E+04

1.2E+04

0 2000 4000 6000

Ultr

afin

e Pa

rticl

e N

umbe

r Con

cent

ratio

n (c

m-3

)

Particulate Matter Concentration PM10 (µg m-3)

296

Fig 8

0.E+00

2.E+03

4.E+03

6.E+03

0 200 400 600 800 1000

Part

icul

ate

Mat

ter C

onc

(µg

m-3

)

Bsp (Mm-1)

PM10

PM2.5

(a)

0.0E+00

2.0E+03

4.0E+03

6.0E+03

8.0E+03

1.0E+04

1.2E+04

0 200 400 600 800 1000

Ultr

afin

e Pa

rtic

le N

umbe

r Con

c (c

m-3

)

Bsp (Mm-1)

(b)

297

Fig 9 (Colour)

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

16:00 17:30 19:00 20:30 22:00 23:30

APS Particle Diameter (µm

)

APS

Part

icle

Num

ber C

onc

(cm

-3)

Time

Concentration

Diameter

298

Fig 9 (B & W)

1.0

1.5

2.0

2.5

3.0

0

5

10

15

20

16:00 17:30 19:00 20:30 22:00 23:30

APS Particle Diameter (µm

)AP

S Pa

rtic

le N

umbe

r Con

c (c

m-3

)

Time

Concentration

Diameter

299

Fig 10 (Colour)

0

2000

4000

6000

8000

10000

1 10 100

Part

icle

Num

ber C

once

ntra

tion

(dN

/dlo

gD)

Particle Diameter (D nm)

08:00-10:00

12:00-14:00

14:00-16:00

300

Fig 10 (B & W)

0

2000

4000

6000

8000

10000

1 10 100

Part

icle

Num

ber C

once

ntra

tion

(dN

/dlo

gD)

Particle Diameter (D nm)

08:00-10:00

12:00-14:00

14:00-16:00