Research Collection

Doctoral Thesis

Investigating aerosol chemical composition and yields from α-pinene photooxidation derived products: The impact of relativehumidity, NOx/VOC and organic mass concentration

Author(s): Pfaffenberger, Lisa Maria

Publication Date: 2014

Permanent Link: https://doi.org/10.3929/ethz-a-010235287

Rights / License: In Copyright - Non-Commercial Use Permitted

This page was generated automatically upon download from the ETH Zurich Research Collection. For moreinformation please consult the Terms of use.

ETH Library

DISS. ETH NO. 21760

Investigating aerosol chemical composition and yields from α-pinene photooxidation derived products:

The impact of relative humidity, NOx/VOC and organic mass concentration

A thesis submitted to attain the degree of

DOCTOR OF SCIENCES of ETH ZURICH

(Dr. sc. ETH Zürich)

presented by

LISA MARIA PFAFFENBERGER

Dipl.-Meteorologin (Univ.), Ludwig-Maximilians-Universität München

born on 29.01.1985

citizen of Germany

examiners

Prof. Dr. Urs Baltensperger, examiner Prof. Dr. Thomas Peter, co-examiner

Dr. André Prévôt, co-examiner Dr. Harald Saathoff, co-examiner

2014

Contents

Summary ................................................................................................................................. I 

Zusammenfassung ................................................................................................................. V 

1  Introduction ....................................................................................................................... 1 

1.1  Aerosol definition and properties ................................................................................. 2 

1.2  Aerosol effects on the Earth’s climate ......................................................................... 4 

1.3  Aerosol effects on human health .................................................................................. 7 

2  The physical and chemical basis ...................................................................................... 9 

2.1  Tropospheric gas-phase chemistry: NOx–VOC catalytic cycles ................................ 10 

2.2  Secondary organic aerosol formation and aging ........................................................ 12 

2.3  Biogenic organic compounds: focus on α-pinene ...................................................... 15 

3  Motivation of the thesis ................................................................................................... 17 

4  Methodology ..................................................................................................................... 21 

4.1  Smog chamber ............................................................................................................ 22 

4.2  Instrumentation .......................................................................................................... 22 

4.2.1  Particle phase instruments .................................................................................. 22 

4.2.2  Gas-phase instruments ........................................................................................ 23 

5  The link between organic aerosol mass loading and degree of oxygenation: an α-

pinene photooxidation study .................................................................................................. 25 

Abstract .................................................................................................................................... 26 

5.1  Introduction ................................................................................................................ 26 

5.2  Method ....................................................................................................................... 29 

5.2.1  Experimental setup ............................................................................................. 29 

5.2.1.1  Introduction of particle and gas-phase reactants into the chamber ............. 30 

5.2.1.2  Instrumental setup ....................................................................................... 33 

5.2.2  Estimation of OH exposure ................................................................................. 34 

5.2.2.1  OH tracer method ........................................................................................ 34 

5.2.2.2  α-Pinene method .......................................................................................... 35 

5.2.2.3  Application of the methods to the dataset ................................................... 35 

5.2.3  Wall loss correction ............................................................................................ 37 

5.3  Results ........................................................................................................................ 37 

5.3.1  General reproducibility of the aerosol degree of oxygenation ........................... 37 

5.3.2  Dependence of degree of oxygenation on the organic mass concentration ........ 40 

5.3.3  Dependence of degree of oxygenation on the OH exposure .............................. 44 

5.3.4  Classification of chemical composition using reference mass spectra ............... 45 

5.4  Conclusions ................................................................................................................ 47 

5.5  Supplementary material ............................................................................................. 49 

6  Higher relative humidity and VOC/NOx increase α-pinene secondary organic

aerosol yields ........................................................................................................................... 67 

Abstract ................................................................................................................................... 68 

6.1  Introduction ................................................................................................................ 69 

6.2  Methods ...................................................................................................................... 71 

6.2.1  Experimental setup and instrumentation ............................................................ 71 

6.2.2  Smog chamber operation and aerosol seeding .................................................... 72 

6.2.3  Estimation of the hygroscopic growth factors and liquid water content ............ 75 

6.2.4  Determination of OH exposure and extent of α-pinene ozonolysis .................... 76 

6.2.5  Determination of suspended and wall-loss-corrected organic mass and yield ... 78 

6.3  Results ........................................................................................................................ 81 

6.3.1  SOA yield dependence on RH, NOx/α-pinene and aerosol seed composition ... 81 

6.3.2  SOA elemental composition ............................................................................... 83 

6.3.3  Particle size dependent uptake of organic mass ................................................. 85 

6.4  Discussion .................................................................................................................. 90 

6.5  Conclusions ................................................................................................................ 92 

6.6  Supplementary material ............................................................................................. 94 

6.6.1  Transmission and collection efficiency correction ............................................. 94 

6.6.2  Seed composition ................................................................................................ 96 

7  VIsible light Photosensitized Secondary Organic Aerosol evolution (VIPSOA) ..... 107 

7.1  Introduction and scientific objective ........................................................................ 108 

7.2  Method ..................................................................................................................... 109 

7.3  Results ...................................................................................................................... 111 

7.4  Measurement challenges and outlook ...................................................................... 116 

8  Final conclusions and outlook ...................................................................................... 119 

List of abbreviations and acronyms .................................................................................... 123 

Bibliography ....................................................................................................................... 125 

Acknowledgements ............................................................................................................. 134 

Curriculum vitae ................................................................................................................. 135 

I

___________________________________________________________________________

Summary

___________________________________________________________________________

Atmospheric aerosol particles are next to greenhouse gases main contributors to the Earth’s

radiative budget and thus important climate drivers, while their scientific understanding is still

low. An aerosol is a suspension of fine solid or liquid particles in a gas and can be of primary

(e.g. mechanical processes) or secondary (by gas-to-particle conversion) origin. Organic

aerosol (OA) represents a major fraction of the submicron mass loading, whereas sources,

formation processes and chemical transformations of secondary organic aerosol (SOA) are

still poorly known. SOA contributes more than primary organic aerosol (POA) to total OA in

most cities around the world. SOA is mainly formed by oxidation of volatile organic

compounds (VOC), resulting in products with lower saturation vapour pressures and

subsequent condensation on pre-existing particles or nucleation. The volatility and chemical

composition of OA, and in parallel the degree of oxygenation (atomic oxygen to carbon ratio

(O : C)) is changed by gas and particle phase reactions (functionalization, fragmentation and

oligomerisation). Current models have been able to reproduce the SOA loading for a few

urban measurement campaigns by allowing dynamic behaviour (evaporation, oxidation in the

gas phase and re-condensation) of POA and considering highly volatile vapours which remain

preferentially in the gas-phase. Even though the loading could be accurately predicted in those

cases, the models remain still underconstrained in terms of fragmentation, detailed chemical

compositions and volatility of compounds. Measurements of aerosol chemical composition

and loading in the ambient air and in controlled environments as smog chambers (SC) are of

great value to improve model predictions.

The presented work 1) identifies the organic aerosol mass concentration as important factor

responsible for the discrepancy between ambient and SC aerosol degree of oxygenation, 2)

presents the relative humidity and NOx/VOC as important drivers of the aerosol yield from α-

pinene photooxidation and 3) suggests direct irradiation to possibly influence the O : C of

wood burning aerosol. Three sets of systematic experiments were conducted in the stationary

SC of the Paul Scherrer institute (PSI) where the aerosol chemical composition and total mass

concentration was measured by means of Aerodyne high resolution time-of-flight aerosol

mass spectrometers (HR-ToF-AMS) and scanning mobility particle sizers (SMPS).

II

In the first part of the study, the composition of SOA at different organic mass concentrations

from α-pinene photooxidation was compared in terms of the fraction of particulate CO2+, a

surrogate for carboxylic acids, vs. the fraction of C2H3O+, a surrogate for aldehydes, alcohols

and ketones. In addition, the evolution of the atomic hydrogen-to-carbon ratio (H : C) vs. the

O : C was investigated in the Van Krevelen space. Low (near-ambient) organic mass

concentrations were found to be necessary to obtain oxygenation levels similar to those of

low-volatility oxygenated organic aerosol (LV-OOA) commonly identified in ambient

measurements. The effects of organic mass loading and oxidant exposure were decoupled by

inter-experiment comparisons at the same OH exposure (integrated hydroxyl radical (OH)

concentration). An OH exposure between 3 and 25 × 107 cm-3 h was required to increase the

O : C by 0.05 during aerosol aging. For the first time, LV-OOA-like aerosol from the

abundant biogenic precursor α-pinene was produced in a smog chamber by oxidation at

typical atmospheric OH concentrations. A significant correlation between measured SOA and

reference LV-OOA mass spectra is shown for experiments with low OA (< 18 µg m-3).

During the second set of experiments, SOA yields (organic mass formed-to-precursor mass

reacted) from the photooxidation of α-pinene were investigated at low (~25 %) and higher

(~ 60 %) relative humidity (RH), various NOx/VOC ratios (0.04–3.8) and with different seed

chemical compositions. Higher RH increased SOA yields up to six times compared to low

RH, with greater increases for the most hygroscopic/acidic seeds. The yields at NOx/VOC

ratios < 0.1 were 2–11 times higher compared to yields at NOx/VOC ratios of up to 3.8. This

NOx dependence follows the same trend as seen in previous studies for α-pinene SOA. The

chemical signature as observed in Van Krevelen diagrams is dominantly influenced by the

NOx/VOC ratio. While previous studies suggest a major effect of the relative humidity on

particle chemical composition via e. g. oligomerisation, we conclude, based on size-resolved

data, that most of the α-pinene reaction products are not absorbed by the prevailing organic

(nonpolar) and inorganic (polar) seeds, rather, the seed provides the surface for condensation.

In the last section of this study, the possible interaction of visible light irradiation and OA was

investigated. Radiation could turn aerosol compounds into photosensitizers which enhance

reactions by catalysing photochemistry within the particle or at the particle surface. First test

measurements indicated an influence of visible light (λ > 400 nm) on wood burning aerosol,

once SOA was formed and irradiated by UV light. Final conclusions are subject to more

sophisticated measurement techniques.

III

This work elucidates how sensitive SOA composition is to a variety of parameters, even when

mainly focusing on α-pinene as one single precursor. As the O : C influences the chemical,

volatility and hygroscopic properties of ambient aerosol, smog chamber studies must be

performed at near-ambient concentrations to accurately simulate ambient aerosol behaviour.

The SOA yields and chemical composition reported in this thesis is of great value for

improving model parameterizations.

IV

V

___________________________________________________________________________

Zusammenfassung

___________________________________________________________________________

Aerosolpartikel leisten neben Treibhausgasen einen Hauptbeitrag zum Strahlungshaushalt der

Erde und sind folglich wichtige klimawirksame Bestandteile in der Atmosphäre, wohingegen

deren wissenschaftliches Verständnis noch gering ist. Als Aerosol bezeichnet man eine

Suspension aus winzigen festen oder flüssigen Partikeln in einem sie umgebenden Gas, wobei

die Partikel von primärem (z.B. aus mechanischen Prozessen) oder sekundärem (durch Gas-

zu-Partikel Konversion) Ursprung sein können. Organisches Aerosol (OA) repräsentiert einen

Hauptbestandteil der Masse von Partikeln kleiner als einem Mikrometer, allerdings sind

Quellen, Entstehungsprozesse und chemische Umwandlungen von sekundärem organischem

Aerosol (SOA) noch wenig erforscht. SOA trägt in den meisten Städten der Welt mehr zum

organischen Gesamtaerosol bei als primäres organisches Aerosol (POA). Die Bildung von

SOA geschieht vor allem durch die Oxidation von flüchtigen organischen Verbindungen

(VOC), wodurch Produkte mit niedrigeren Sättigungsdampfdrücken entstehen, gefolgt von

anschließender Kondensation auf bereits vorhandenen Partikeln oder Nukleation. Gas- und

Partikelphasenreaktionen (Funktionalisierung, Fragmentierung und Oligomerisation)

verändern die Flüchtigkeit, die chemische Zusammensetzung, und zugleich den

Oxidierungsgrad von OA (atomares Verhältnis zwischen Sauerstoff und Kohlenstoff (O : C)).

Neueste Modelle haben die Aerosolbelastung im Vergleich zu direkten Messungen in einigen

Städten erfolgreich nachsimuliert, indem dynamisches Verhalten (Verdampfung, Oxidierung

in der Gasphase und nochmalige Kondensation) von POA zugelassen wurde und indem

Produkte berücksichtigt wurden, die extrem flüchtig sind und bevorzugt in der Gasphase

verweilen. Obwohl in diesem Fall die Aerosolbelastung korrekt vorhergesagt werden konnte,

sind Modelle nichtdestotrotz in Bezug auf die Fragmentierung, die detaillierte chemische

Zusammensetzung und die Flüchtigkeit der Substanzen nicht vollständig beschrieben.

Messungen der chemischen Zusammensetzung von Aerosolen und der Partikelbelastung in

Außenluft und in kontrollierten Räumen wie z. B. Smogkammern (SC) sind daher von großer

Bedeutung, um Modellvorhersagen zu verbessern.

Die hier vorliegende Arbeit 1) identifiziert die organische Massenkonzentration als einen

wichtigen Faktor für die Diskrepanz zwischen dem Oxidierungsgrad von atmosphärischem

VI

und von in SC generiertem Aerosol; 2) verdeutlicht die Wichtigkeit der relativen Feuchte und

des NOx/VOC Verhältnisses für die Aerosol-Ausbeute aus der Photooxidation von α-Pinen

und 3) deutet darauf hin, dass sichtbares Licht vermutlich das O : C von

Holzverbrennungsaerosol beeinflusst. In der stationären SC des Paul Scherrer Instituts (PSI)

wurden drei systematische Versuchsreihen durchgeführt, während derer die chemische

Zusammensetzung des Aerosols mittels eines hochauflösendem Aerodyne Flugzeit

Massenspektrometers (HR-ToF-AMS) und die Gesamtmassenkonzentration mittels eines auf

der Mobilität der Partikel basierenden Größenmessgeräts (SMPS) gemessen wurden. Im

ersten Teil der Studie wurde die Zusammensetzung von SOA, durch Photooxidation von α-

Pinen entstanden, bei verschiedenen Massenkonzentrationen in Bezug auf dessen Anteil an

CO2+ in der Partikelphase (einem Proxy für Karbonsäuren) gegenüber dem Anteil an C2H3O

+

(einem Proxy für Aldehyde, Alkohole und Ketone) verglichen. Zusätzlich wurde die

Veränderung des atomaren Wasserstoff-Kohlenstoff-Verhältnisses (H : C) gegenüber O : C

im Van Krevelen Diagramm untersucht. Es wurde herausgefunden, dass niedrige

(vergleichbar mit der realen Atmosphäre) organische Massenkonzentrationen nötig sind, um

Oxidierungsgrade zu erreichen, welche vergleichbar sind mit jenen von schwerflüchtigem

oxidiertem organischem Aerosol (LV-OOA), das üblicherweise in atmosphärischen

Messungen identifiziert wird. Die Effekte der organischen Massenkonzentration und der

„OH-Exposition“ (integrierte Hydroxyl-Radikal (OH) Konzentration) wurden entkoppelt,

indem Experimente bei gleicher OH-Exposition verglichen wurden. Eine OH-Exposition

zwischen 3 und 25 × 107 cm-3 h war erforderlich, um das O : C während der Aerosolalterung

um 0.05 zu erhöhen. Zum ersten Mal wurde in einer Smogkammer LV-OOA-ähnliches

Aerosol aus dem oft vorhandenen biogenen Vorläufergas α-Pinen durch Oxidation bei typisch

atmosphärischen OH-Konzentrationen produziert. Massenspektren von gemessenem SOA

zeigen eine signifikante Korrelation mit LV-OOA-Referenzspektren bei Experimenten mit

niedrigem OA (< 18 µg m-3).

Während der zweiten Experimentreihe wurden SOA-Ausbeuten (d.h. gebildete organische

Masse/ reagierte Masse eines Vorläufergases) während der Photooxidation von α-Pinen bei

niedriger (~ 25 %) und hoher (~ 60 %) relativer Feuchte (RH), verschiedenen NOx/VOC

Verhältnissen (0.04–3.8) und unterschiedlichen chemischen Zusammensetzungen eines

vorgelegten Aerosols („Seed-Aerosol“) untersucht. Eine höhere Feuchte erhöhte die SOA-

Ausbeute hierbei um bis zu sechs Mal im Vergleich zu einer niedrigen Feuchte, wobei höhere

Zunahmen mit denjenigen „Seed-Aerosolen“ erreicht wurden, die am hygroskopischsten und

sauersten waren. Die Ausbeuten bei einem NOx/VOC-Verhältnis < 0.1 waren 2–11-mal höher

VII

als diejenigen bei einem NOx/VOC-Verhältnis von bis zu 3.8. Die NOx-Abhängigkeit folgte

dem gleichen Trend, der in früheren Studien zu α-Pinen-SOA gefunden wurde. Die chemische

Signatur, die in Van Krevelen-Diagrammen untersucht wurde, wird in erster Linie durch das

NOx/VOC-Verhältnis beeinflusst. Frühere Studien schlugen einen großen Einfluss der

relativen Feuchte auf die chemische Zusammensetzung der Partikel vor, z. B. einhergehend

mit Oligomerisation. Dahingegen folgern wir, basierend auf größenaufgelösten Daten, dass

die meisten α-Pinen Reaktionsprodukte nicht von vorhandenen organischen (unpolar) und

anorganischen (polar) „Seed-Aerosolen“ absorbiert werden, sondern das „Seed-Aerosol“

vielmehr eine Oberfläche für Kondensation bietet.

Im letzten Teil der Studie wurde eine mögliche Wechselwirkung von direkter Licht-

Einstrahlung und OA untersucht. Licht könnte Bestandteile des Aerosols in Photosensitizer

umwandeln, welche durch Reaktionen innerhalb eines Partikels oder auf der

Partikeloberfläche photochemische Prozesse katalysieren. Erste Testexperimente zeigten

einen Einfluss von sichtbarem Licht (λ > 400 nm) auf Holzverbrennungsaerosol,

vorausgesetzt, SOA wurde vorher gebildet und mit UV-Licht bestrahlt. Für endgültige

Schlussfolgerungen sind jedoch noch ausgereiftere Messtechniken notwendig.

Diese Arbeit macht deutlich, wie stark die SOA-Zusammensetzung von einer Fülle an

Parametern abhängt, auch wenn wir uns mit α-Pinen auf ein einziges Vorläufergas

konzentriert haben. Um die Eigenschaften von atmosphärischem Aerosol wahrheitsgetreu zu

simulieren, müssen Smogkammerstudien bei Konzentrationen, die jenen in der Atmosphäre

ähnlich sind, durchgeführt werden. Dadurch werden die chemischen Eigenschaften wie das

O : C-Verhältnis, sowie die Flüchtigkeit und die Hygroskopizität des atmosphärischen

Aerosols beeinflusst. Die Ausbeuten und chemischen Zusammensetzungen von SOA, die in

dieser Dissertation präsentiert werden, sind von großer Wichtigkeit, um Parametrisierungen in

Modellen zu verbessern.

___________________________________________________________________________

1 Introduction

___________________________________________________________________________

2 1 Introduction

1.1 Aerosol definition and properties

Particles in the atmosphere originate from natural sources, such as windborne dust, sea spray

and volcanoes, and from anthropogenic activities, such as industrial processes or combustion

of fuels. Whereas an aerosol is technically defined as a suspension of fine solid or liquid

particles in a gas, common usage refers to the aerosol as the particulate component only.

Emitted directly as particles (primary aerosol) or formed in the atmosphere by gas-to-particle

conversion processes (secondary aerosol), atmospheric aerosols are generally considered to be

the particles that range in size from a few nanometers (nm) to tens of micrometers (µm) in

diameter. The term PMx, defined as particulate matter with an aerodynamic diameter smaller

than x µm, is often used to describe the aerosol mass concentration (e.g. PM1, PM2.5). Once

airborne, particles can change their size and composition by condensation of vapours or by

evaporation, by coagulating with other particles, by chemical reaction, or by activation in the

presence of water supersaturation to become fog or cloud droplets (Seinfeld and Pandis,

2006). There are several properties of particles that are important for their role in atmospheric

processes. These include their number concentration, their mass, size, chemical composition,

and aerodynamic and optical properties. Of all these properties, particle size is the most

important. It is not only related to the source or processing of the particles, but also to their

effects on health, visibility, and climate (Finlayson-Pitts and Pitts, 2000).

Atmospheric removal processes can be divided into two groups: dry deposition and wet

deposition. Dry deposition refers to the direct transfer of species to the earth’s surface without

the aid of precipitation. Wet deposition encompasses all processes by which species are

transferred to the earth’s surface in aqueous form like rain, snow or fog. Whereas atmospheric

trace gases have lifetimes ranging from less than a second to a century or more, residence

times of particles in the troposphere vary only from a few days to a few weeks, depending on

their shape, density and chemical composition, as well as precipitation frequency (Seinfeld

and Pandis, 2006). Figure 1.1 displays formation, growth, chemical transformation and

removal processes of atmospheric aerosols and gases as a function of particle diameter (Raes

et al., 2000).

Aerosols interact with radiation by absorption and scattering, the sum of which, the overall

attenuation, is defined as extinction. In addition to the size, the radiative effects of a

population of particles depend on their composition through compound and mixture specific

refractive indices. There are substances mainly absorbing (e.g. black carbon (BC)) substances

1.1 Aerosol definition and properties 3

Figure 1.1. Scheme of microphysical processes influencing the size distribution and chemical composition of atmospheric aerosol. The scheme highlights the large range of sizes that are involved in the formation and evolution of aerosol particles, and how aerosols participate in atmospheric chemical processes through homogeneous, heterogeneous and in-cloud reactions (Raes et al., 2000).

exclusively scattering (e.g. sulfate). The atmospheric aerosol is generally a mixture of species

from a number of sources. The aerosol mixing state depends on how all the components are

distributed among the particles (Seinfeld and Pandis, 2006). Figure 1.2 shows a schematic of

possible mixtures: fresh or aged externally mixed aerosol and fresh or aged internally mixed

aerosol.

Figure 1.2. Schematic of externally and internally mixed aerosol. Particles vary in size, chemical compositions and the fractions of species within one particle.

4 1 Introduction

In nature, a fresh externally mixed aerosol population could represent black and organic

carbon emissions mixed with sea salt background aerosol, the aged aerosol could be covered

with a layer of secondary organic aerosol. The internally mixed aerosol population could arise

from new particle formation in a clean background atmosphere, from coating of BC with

SOA or ammonium sulfate, or from droplet coalescence within clouds (Andreae et al., 1986).

An ideal internally mixed aerosol can be produced in laboratories.

When a light-absorbing compound is present, all particles in an internal mixture exhibit some

absorption, whereas in an external mixture, only part of the aerosol exhibits absorption. The

ability of the aerosol to absorb water, defined as the hygroscopicity, is also influenced by the

mixing state of the aerosol. Assuming that both hygroscopic and non-hygroscopic

components are present, in an internal mixture every particle (of a given size) exhibits

proportionally the same growth as RH is increased; in an external mixture, only the

hygroscopic particles grow (Seinfeld and Pandis, 2006).

1.2 Aerosol effects on the Earth’s climate

Aerosols can influence the Earth’s climate in two ways: by aerosol-radiation interactions (ari)

and aerosol-cloud interactions (aci). In the IPCC report (Intergovernmental Panel on Climate

Change, IPCC, 2013b) large efforts have been made to determine the effects of both types of

interactions on the Earth’s radiative budget. The commonly used parameter to quantify the

contribution of various drivers to climate change is the radiative forcing (RF), defined as the

net change in the energy balance of the Earth system due to some imposed perturbation, in

W m-2 and averaged over a particular period of time. Recent research has clarified the

importance of distinguishing forcing and rapid adjustments from feedbacks: Forcings

associated with agents such as greenhouse gases and aerosols act on the global mean surface

temperature through the instantaneous change in the global radiative (energy) budget. Rapid

adjustments arise when forcing agents, by altering flows of energy internal to the system,

affect cloud cover or other components of the climate system and thereby alter the global

radiative budget indirectly. In contrast, feedbacks are associated with changes in climate

variables that are mediated by a change in global-mean surface temperature. Furthermore one

can distinguish between the traditional concept of radiative forcing (RF), defined as the

change in the radiative budget over a given time period, and the relatively new concept of

effective radiative forcing (ERF) which includes in addition rapid adjustments (IPCC, 2013b).

1.2 Aerosol effects on the Earth’s climate 5

Figure 1.3 links the former terminology of aerosol direct, semi-direct and indirect effects with

the terminology used in the recent IPCC report (IPCC, 2013b). The radiative forcing from

aerosol-radiation interactions (RFari) encompasses radiative effects from anthropogenic

aerosols before any adjustment takes place, and corresponds to what is usually referred to as

the aerosol direct effect. Rapid adjustments induced by aerosol radiative effects on the surface

energy budget, the atmospheric profile and cloudiness contribute to the effective radiative

forcing from aerosol-radiation interactions (ERFari). They include what has earlier been

referred to as the semi-direct effect. The radiative forcing from aerosol-cloud interactions

(RFaci) refers to the instantaneous effect on cloud albedo due to changing concentrations of

cloud condensation and ice nuclei, also known as the Twomey effect. All subsequent changes

to the cloud lifetime and thermodynamics are rapid adjustments, which contribute to the

effective radiative forcing from aerosol-cloud interactions (ERFaci) (IPCC, 2013b).

Figure 1.3. Schematic of the new terminology used in Assessment Report AR5 for aerosol-radiation and aerosol- cloud interactions and how they relate to the terminology used in the earlier AR4 report (Forster et al., 2007). The blue arrows depict solar radiation, the grey arrows terrestrial radiation, and the brown arrow symbolises the importance of couplings between the surface and the cloud layer for rapid adjustments (IPCC, 2013b).

6 1 Introduction

Figure 1.4 shows the RF over the industrial era (2011 relative to 1750) by emitted

compounds, putting the aerosol climate effects into relation with the RFs of greenhouse gases,

surface albedo (defined as ratio of reflected to incident radiation) changes due to land use and

natural changes in solar irradiance (IPCC, 2013a). For all gases, the changes in emissions and

changes in abundance are considered. All changes leading to an increase of greenhouse gas

(CO2, H2O, CH4, N2O, O3 and halocarbons) concentrations have a positive effect on the RF.

CO2 represents the dominant positive RF by both, abundance and emitted compound. NOx

caused a positive RF by increased O3 production and a negative RF by influencing the

lifetime of CH4 and reducing its abundance (for a detailed description of the tropospheric gas-

phase chemistry, see Sect. 2.1) and through contributions to nitrate aerosol formation.

Figure 1.4. Radiative forcing estimates in 2011 relative to 1750 and aggregated uncertainties for the main drivers of climate change. Values are global average radiative forcing partitioned according to the emitted compounds or processes that result in a combination of drivers. Best estimates of the net radiative forcing are shown as black diamonds with corresponding uncertainty intervals (numerical values and confidence level in the net forcing on the right: VH – very high, H – high, M – medium, L – low, VL – very low). Total anthropogenic radiative forcing is provided for three different years relative to 1750 (IPCC, 2013a).

1.3 Aerosol effects on human health 7

Emissions of BC have a positive RF through aerosol-radiation interactions and BC on snow

and ice. Scattering aerosol comprising mineral dust, sulfate, nitrate and organic carbon show

negative RF (IPCC, 2013b).

The total effective radiative forcing due to aerosols, assessed to be –0.9 (–1.9 to –0.1) W m–2

with medium confidence [ERFari = –0.45 (–0.95 to +0.05) W m–2, ERFaci = –0.45 (–1.2 to

0.0) W m–2, excluding the effect of absorbing aerosol on snow and ice], is counteracting the

ERF of well-mixed greenhouse gases (CO2, CH4, N2O and halocarbons), assessed to be +2.83

(+2.26 to +3.40) W m–2 with very high confidence (IPCC, 2013b). Due to the high uncertainty

of the aerosol ERF, research on the global and regional aerosol budget and composition is

needed to better evaluate the present and future impact of aerosols on the climate.

1.3 Aerosol effects on human health

Particles can affect human health by inhalation and followed inflammation of the lung or

cardiopulmonary diseases. Three main particle deposition mechanisms in the respiratory tract

include inertial impaction, gravitational settling and diffusion. Particles larger than 5 µm in

diameter are deposited in the upper respiratory tract where also particles smaller than a few

tens of nm are efficiently removed by diffusion (Maynard and Kuempel, 2005). Particle

deposition modelling indicates that up to 90 % or more of the inhaled mass fraction of

particles smaller than 100 nm deposits in the respiratory tract, with up to approximately 50 %

in the alveolar region, the deepest part of the lung (ICRP, 1994). Ultrafine particles below

100 nm diameter are potentially the most dangerous, because they own the largest surface

area and highest content of potentially toxic hydrocarbons among all particular matter

sources. Finally they can penetrate deeper into the lung tissue than fine or coarse particles

(Oberdörster and Utell, 2002; Nel, 2005). A correlation between increased mortality and

PM2.5 levels as well as a decrease in mortality for decreased PM2.5 levels was found for six

cities in the US (Dockery et al., 1993; Laden et al., 2006; Lepeule et al., 2012). Pope III et al.

(2009) found that each 10 µg m-3 increase in PM2.5 concentration was associated with

approximately a 4 %, 6 % and 8 % increased risk of all-cause, cardiopulmonary, and lung

cancer mortality, respectively.

___________________________________________________________________________

2 The physical and chemical basis

___________________________________________________________________________

10 2 The physical and chemical basis

2.1 Tropospheric gas-phase chemistry: NOx–VOC catalytic cycles

Volatile organic compounds (VOCs) are precursors for secondary organic aerosol (SOA)

formation via the condensation of low volatility and semi-volatile oxidation products. Their

gas-phase reaction pathways are highly dependent on the ratio between nitrogen oxide (NOx)

and VOC concentration as shown in Figure 2.1. Tropospheric O3 concentrations are

dependent on the reaction chain lengths of NOx ([NOx] = [NO] + [NO2]) and HOx ([HOx] =

[OH] + [HO2] + [RO2]) catalytic cycles (Thornton et al., 2002). An important measure of

atmospheric ozone-forming oxidation cycles is the ozone production efficiency (OPE). NOx

can be viewed as the catalyst in O3 formation, because it gets cycled back and forth between

NO and NO2 in O3 generation. The OPE can be described by the formation of NO2 leading to

ozone production ( ) versus the loss of NOx ( ) which is mostly due to reaction of NO2

with OH, here as an example for HO2 (Seinfeld and Pandis, 2006):

OPE

(2.1)

where and are the reaction rates of HO2 with NO and OH with NO2,

respectively. Hydroxyl radicals (OH ̇ ) are mainly produced by the photolysis of ozone (next

to HONO, H2O2 and aldehyde photolysis):

O H O → 2OH ̇ O (R 2.1)

OH ̇ radicals react with VOCs to produce alkyl radicals which in turn react with oxygen to

give alkyl-peroxy radicals:

RCH OH∙ → RCH ∙ H O (R 2.2)

RCH ∙ O M → RCH O ∙ M (R 2.3)

RCH O ∙ NO → RCH O∙ NO (R 2.4)

NO →NO O (R 2.5)

O O → O (R 2.6)

In addition to O3 production and re-cycling of NOx, this reaction chain leads to reduction of

VOCs which are then able to turn into low-volatile products undergoing gas-to-particle

conversion. Three reaction regimes have been defined involving the OPE: NOx limited, VOC

limited and transition regime:

2.1 Tropospheric gas-phase chemistry: NOx–VOC catalytic cycles 11

NOx limited or low NOx regime: HO2˙ (hydroperoxyl radicals) react at low NOx

concentrations, instead of with NO, with alkyl-peroxy radicals (R 2.7) or do self-reaction to

form hydrogen peroxide (R 2.8). For a fixed HOx source at low NOx concentrations, O3

increases linearly with increases in NOx concentrations.

RCH O ∙ HO ∙ → RCH OOH O (R 2.7)

HO ∙ HO ∙ H O → H O O H O (R 2.8)

VOC limited or high NOx regime: At higher NOx concentrations, effective ozone production

rates decrease with increasing NOx due to the enhanced nitric acid (HNO3) formation, which

in turn reduces NOx concentration and consumes OH instead of reacting with RCH3 (R 2.2):

NO OH∙ → HNO (R 2.9)

Figure 2.1. Reaction scheme of coupled NOx and VOC cycle producing oxidized species and O3 that can be terminated by the formation of HNO3 (mainly high NOx regime) or the formation of peroxides (mainly low NOx regime) (Guderian, 2000).

Transition regime: Between the NOx limited and VOC limited regimes, the NOx/VOC ratio

is at its optimum to efficiently produce and recycle OH radicals and therefore maximize O3

production and degradation of VOCs (Seinfeld and Pandis, 2006; Thornton et al., 2002).

EKMA-diagrams (Empirical Kinetic Modeling Approach, EPA, 1977) show highest for

VOC (ppmC)/NOx(ppm) of approximately 8/1, a typical ratio in the transition regime. The

VOC-limited, transition and NOx limited regime can be observed when following an air

12 2 The physical and chemical basis

parcel from a city with high NOx emissions on its way to rural areas with less pollution and

more dominant VOC emissions compared to NOx or in SC experiments when starting with

high NOx.

2.2 Secondary organic aerosol formation and aging

The dynamic nature of the semi-volatile species resulting from the NOx-VOC cycling is

important for their absolute yields, defined as the ratio of aerosol mass formed to gas-phase

precursor mass reacted. Depending on their saturation vapor pressure and inferred volatility, a

higher or lower percentage of the total concentration of a compound is found in the particle

phase. Gas-phase oxidation reactions can reduce volatility by the addition of polar functional

groups or increase it by the cleavage of carbon–carbon bonds (Kroll and Seinfeld, 2008). As a

consequence, SOA yields and degree of oxygenation (described by the O : C ratio) may not

be constant for one precursor compound, but subject to the extent to which it was exposed to

oxidants.

Barley et al. (2009) comprehensively summarized the history of partitioning formulations:

The partitioning of semi-volatile organic components was originally thought to be dominated

by adsorption (Pankow, 1987), before Pankow et al. (1994) developed an equilibrium

partitioning model to distinguish between absorptive partitioning into a condensed phase and

adsorption onto the particle surface. Odum et al. (1996) described the partitioning by splitting

semi-volatile products into two volatility bins. Pankow et al. (2001) developed the absorptive

partitioning model to describe the gas/particle partitioning of each component in a complex

multicomponent system. The condensation of multiple organic compounds into an aerosol

needs to take account of interactions between molecules in the condensed phase (deviations

from Raoult’s Law) as well as the volatility of the components. The absorptive partitioning

model provides a mathematically simple method of predicting the condensed phase

composition in a multicomponent system at temperatures and pressures relevant to the

atmosphere. Donahue et al. (2006) further developed the model to consider a number of

condensable compounds with a broad range of volatility (volatility basis set (VBS)). This

approach allows use of large numbers of potentially condensable compounds by binning them

according to their saturation concentration ( ∗) value. The amount of condensed material is

calculated by summing all components i ensuring mass balance between the two phases for

each component considered.

2.2 Secondary organic aerosol formation and aging 13

Defining a partitioning coefficient ξi for compound i given its ∗ value:

,

,

,

, ,1 ,

,1

, ∙1

∗

(2.2)

With: ,,

, ∙ ,

∗ (2.3)

where caer,i and cg,i are the mass concentrations of compound i in the condensed and gas phase,

respectively, and ctot,i the sum of both, all in µg m-3. Kp,i is the partitioning constant of

compound i. coa is the total particle-phase concentration of compounds miscible with

compound i. R is the ideal gas constant, T (K) the temperature, Mi (g mol-1) molecular weight

of species i, ζ is a molality-based activity coefficient for compound i in the liquid phase and

, is the saturation vapour pressure of pure compound i at temperature T. Following Eq.

(2.3), the partitioning constant Kp,i can be increased by decreasing the molality-based activity

coefficient, the vapour pressure or the temperature.

The total mass of condensed miscible organic material, coa, is given by the sum of the

products of the individual total component concentrations in both phases and their partitioning

coefficients: ∑ ∙ . When ∗ is equal to coa, half of the semi-volatile mass resides

in the particle phase and the other half in the gas phase. Figure 2.2 (Kroll and Seinfeld, 2008)

shows the fractions ξ of semi-volatile compounds i in the particle phase using a 2-product

approach (Odum et al., 1996; Figure 2.2a–b) and a broader volatility distribution (Donahue et

al., 2006; Figure 2.2c–d). Partitioning plots at two organic mass loadings coa = 1 µg m-3

representative for remote areas (left panels) and coa = 10 µg m-3 representative for urban areas

(right panels), show the increased particulate fraction for semi-volatile compounds for

increased coa. Note that these plots show only the fraction ξ of each semi-volatile compound

in the particle phase; particle-phase concentrations are obtained by multiplying ξ by total mass

concentration of each semi-volatile compound. The figure visualizes the difficulty to

represent the whole volatility range with only two surrogate compounds.

14 2 The physical and chemical basis

The aerosol yield from a precursor gas-phase concentration is defined as the ratio between

organic aerosol mass formed (Δcoa) per mass of hydrocarbon reacted (ΔHC) (Kroll and

Seinfeld, 2008):

∆∆

∙ ∑ ,

, (2.4)

where αi is the mass yield of compound i. The SOA yield from a given precursor is therefore

not a stoichiometric quantity, but rather increases with increasing total organic particulate

loading.

Figure 2.2. Representation of gas–particle partitioning for a complex mixture of semi-volatiles using (a–b) the ‘two-product model’, and (c–d) the ‘volatility basis set’. Partitioning at two mass loadings of organic aerosol (1 and 10 mg m-3) is shown for each. (Kroll and Seinfeld, 2008, adapted).

Kroll and Seinfeld (2008) focused on three primary factors determining volatility of organic

compounds in the atmosphere: (1) oxidation reactions of gas-phase organic species, which

lower volatility by addition of functional groups but can also increase volatility by cleavage of

carbon-carbon bonds; (2) reactions in the particle (condensed) phase, which can change

volatility either by oxidation or formation of high-molecular-weight species; and (3) the

extent to which these reactions occur, as the volatility distribution of oxidation products will

2.3 Biogenic organic compounds: focus on α-pinene 15

continuously evolve as a result of ongoing chemistry. Those processes are illustrated in Figure

2.3 from Jimenez et al. (2009).

Figure 2.3. Fragmentation, functionalization or oligomerisation change volatility (and the degree of oxygenation). The branching ratio (β) determines if functionalization will reduce volatility, whereas fragmentation can generate more-volatile species, which are less likely to partition to the OA (Jimenez et al., 2009, adapted).

2.3 Biogenic organic compounds: focus on α-pinene

Biogenic volatile organic compounds (BVOC) are emitted mainly from vegetation and are

therefore underlying seasonal variations, where highest BVOC emissions occur at high

temperatures and high solar radiation. The annual global VOC flux was estimated to be 1150

Tg C, composed of 44 % isoprene, 11 % monoterpenes (such as α-/β-pinene), 22.5 % other

reactive VOC, and 22.5 % other VOC (Guenther et al., 1995). Globally, BVOC emissions, are

one order of magnitude more abundant than anthropogenic VOCs (Guenther et al., 2006).

Steinbrecher et al. (2009) presented detailed BVOC emission indices for a large range of tree

species. α-Pinene (29 %) and β-pinene (21 %) represent the highest emitted fractions of

monoterpenes (C10H16), which consist in general of two isoprene units (C5H8). The model

compound mainly used in this thesis was α-pinene (M = 136.2340 g mol-1 (NIST)) with its

structure shown next to isoprene in Figure 2.4.

16 2 The physical and chemical basis

Figure 2.4. Chemical structure of a) the monoterpene α-pinene (C10H16) and b) isoprene (C5H8).

BVOC play a key role in aerosol formation and yearly fluxes of 12–70 Tg yr-1 biogenic SOA

(BSOA) were estimated by “bottom-up” models (Hallquist et al., 2009). “Bottom-up”

methods combine known or inferred biogenic (most notably isoprene and terpenes) and/or

anthropogenic VOC precursor fluxes in global models with laboratory data from SOA

formation experiments to obtain a global organic aerosol field.

The global emission of monoterpenes is 4 times lower than that of isoprene, but its influence

on the aerosol burden is higher due to the high SOA formation potential. α-Pinene SOA yields

up to 20 % at ~50 µg m-3 organic mass concentration (Hao et al., 2011, and references

therein), while even 28.5 % for 281 µg m-3 OA (Cocker et al., 2001) were reported, whereas

isoprene yields span much lower values: < 1 % (Kleindienst et al., 2006), 1–3 % for 10–

100 µg m-3 OA (Dommen et al., 2009).

Radiocarbon dating (14C) methods indicate that in summer non-fossil, biogenic sources are

dominant in Zürich with 60 % of OC, where SOA prevails (Szidat et al., 2006). Also the high

non-fossil SOA fraction together with a high contribution of monoterpenes to SOA in

Marseille in summer time indicated that biogenic VOC are important SOA precursors (El

Haddad et al., 2013). In winter, a high fraction of non-fossil SOA can originate from wood

burning emissions, and the contribution of biogenic sources is more difficult to be quantified.

___________________________________________________________________________

3 Motivation of the thesis

___________________________________________________________________________

18 3 Motivation of the thesis

Organic aerosol represents a major fraction of the submicron mass loading, whereas sources

of SOA are still poorly known. Aerosol has a large effect on the Earth’s climate, attributed

with a large uncertainty as discussed in Section 1.2. Comparing modelled and observed SOA

(Figure 3.1) during measurement campaigns in different cities all over the world, showed that

the VBS approach (Sect. 2.2) represents measured SOA much better than the traditional

approach where POA is assumed non-volatile and SOA less dynamic (Hodzic et al., 2010).

Still, many processes influencing the chemical composition and amount of SOA are unknown.

Figure 3.1. Comparison of observed to predicted SOA ratios. Blue symbols represent estimates of the traditional SOA approach, while red refers to the volatility basis set. (Hodzic et al., 2010 and references therein)

This thesis is contributing to a better understanding of SOA by investigating chemical

composition and yields from the photooxidation of the atmospheric abundant aerosol

precursor α-pinene, when being exposed to various atmospherically relevant experimental

conditions. We studied BVOC because it is evident from the NOx-VOC catalytic cycle

(described in Sect. 2.1) that with reducing anthropogenic emissions of both VOC and NOx,

biogenic sources become increasingly important in atmospheric chemical processes

(Steinbrecher et al., 2009).

The aim of this study is to examine the impact of the mass concentration on SOA chemical

composition as well as the NOx/VOC ratios and particulate water content, contributing to the

total absorptive mass, on SOA chemical composition and yields. This is performed by varying

α-pinene initial concentrations, NOx/α-pinene ratios, aerosol seed composition

(hygroscopicity, acidity) and relative humidity. The relationship between O : C and organic

mass concentration as well as SOA yields reported here may aid the parameterization of the

3 Motivation of the thesis 19

α-pinene SOA dependence on total organic mass concentration, NOx and particulate water for

use in atmospheric models.

___________________________________________________________________________

4 Methodology

___________________________________________________________________________

22 4 Methodology

4.1 Smog chamber

Smog chambers provide a controlled isolated environment to investigate the formation and

aging of aerosols (Alfarra et al., 2006; Paulsen et al., 2005; Kalberer et al., 2004). The PSI

smog chamber is a 27-m3 bag made of a fluorocarbon film, enclosed in a temperature-

controlled housing. A schematic of the chamber is shown in Figure 4.1. The chamber can be

irradiated with four xenon arc lamps (16 kW total) to simulate the solar spectrum with the

intensity of a Swiss winter day at noon and with 80 recently installed black lights (Philips,

Cleo performance 100W/R) to accelerate oxidation by increased UV radiation, with emission

between 300–400 nm wavelength (light characterization in Platt et al. (2013).

Figure 4.1. Plan view of the reaction chamber and enclosure (Paulsen et al., 2005).

4.2 Instrumentation

The temperature (T) and RH measurement was optimized by passing sampling air through a

radiation shielded sensor.

4.2.1 Particle phase instruments

The Aerodyne high resolution time of flight aerosol mass spectrometer (HR-ToF-AMS,

Aerodyne Research, Inc., Billerica, MA, USA) provides online measurements of real-time

non-refractory chemical speciation (organics, ammonium, nitrate, sulfate, chloride) and mass

4.2 Instrumentation 23

loading as a function of particle size on the principle of thermal vaporization and electron

impact (DeCarlo et al., 2006). The HR-TOF-AMS contains a time-of-flight mass spectrometer

which allows distinguishing of different elemental compositions (ions) at the same unit mass

with a good resolution power for m/z <120. The high pressure lens PM2.5, increasing the upper

particle size cut off from ~ 1000 nm (standard lens) till ~ 2500 nm (Williams et al., 2013) is

an important improvement.

The scanning mobility particle sizer (SMPS) measures the size distribution of fine particles by

separation based on their electrical mobility. The sampled air is usually dried prior to entering

the particle-phase instruments, to minimize loss of large particles and the influence of water

on the HR-ToF-AMS measurement.

4.2.2 Gas-phase instruments

The proton transfer reaction mass spectrometer (PTR-MS) measures with unit mass resolution

molecules which have a proton affinity higher than that of water (165.2 kcal mol-1). The main

constituents of the air (nitrogen, oxygen, argon, carbon dioxide) have a lower proton affinity

than water and the measurement of volatile organic compounds of low concentration is

therefore facilitated (Lindinger et al., 1998; Hellén et al., 2008).

A chemiluminescence-based NOx instrument (Monitor Labs 9841A NOx analyzer) and a

modified NOx instrument including a photolytic NO2-to-NO converter (Thermo

Environmental Instruments 42C trace level NOx analyzer equipped with a blue light

converter) and two ozone monitors (Monitor Labs 8810 ozone analyzer, Environics S300

ozone analyzer) monitor the gas phase in the chamber.

___________________________________________________________________________

5 The link between organic aerosol mass

loading and degree of oxygenation:

an α-pinene photooxidation study

___________________________________________________________________________

L. Pfaffenberger1, P. Barmet1, J. G. Slowik1, A. P. Praplan1,+, J. Dommen1, A. S. H.

Prévôt1 and U. Baltensperger1

1Laboratory of Atmospheric Chemistry, Paul Scherrer Institute, 5232 Villigen, Switzerland

+Now at: Department of Physics, University of Helsinki, Helsinki, Finland

Published in Atmospheric Chemistry and Physics, 13, 6493-6506, doi:10.5194/acp-13-6493-2013

26 5 The link between organic aerosol mass loading and degree of oxygenation

Abstract

A series of smog chamber (SC) experiments was conducted to identify factors responsible for

the discrepancy between ambient and SC aerosol degree of oxygenation. An Aerodyne high-

resolution time-of-flight aerosol mass spectrometer is used to compare mass spectra from α-

pinene photooxidation with ambient aerosol. Composition is compared in terms of the

fraction of particulate CO2+, a surrogate for carboxylic acids, vs. the fraction of C2H3O

+, a

surrogate for aldehydes, alcohols and ketones, as well as in the Van Krevelen space, where

the evolution of the atomic hydrogen-to-carbon ratio (H : C) vs. the atomic oxygen-to-carbon

ratio (O : C) is investigated. Low (near-ambient) organic mass concentrations were found to

be necessary to obtain oxygenation levels similar to those of low-volatility oxygenated

organic aerosol (LV-OOA) commonly identified in ambient measurements. The effects of

organic mass loading and OH (hydroxyl radical) exposure were decoupled by inter-

experiment comparisons at the same integrated OH concentration. An OH exposure between3

and 25 × 107 cm-3 h is needed to increase O : C by 0.05 during aerosol aging. For the first

time, LV-OOA-like aerosol from the abundant biogenic precursor α-pinene was produced in a

smog chamber by oxidation at typical atmospheric OH concentrations. Significant correlation

between measured secondary organic aerosol (SOA) and reference LV-OOA mass spectra is

shown by Pearson’s R² values larger than 0.90 for experiments with low organic mass

concentrations between 1.2 and 18 µg m-3 at an OH exposure of 4 × 107 cm-3 h, corresponding

to about two days of oxidation time in the atmosphere, based on a global mean OH

concentration of ~ 1 × 106 cm-3. α-Pinene SOA is more oxygenated at low organic mass

loadings. Because the degree of oxygenation influences the chemical, volatility and

hygroscopic properties of ambient aerosol, smog chamber studies must be performed at near-

ambient concentrations to accurately simulate ambient aerosol properties.

5.1 Introduction

Organic aerosol (OA) represents 20 to 90 % of the submicron atmospheric aerosol (Jimenez et

al., 2009 and references therein) and has numerous sources. It can be introduced directly by

combustion and mechanical processes into the atmosphere as primary organic aerosol (POA)

or can be formed by condensation of gas-phase reaction products with low vapor pressures

resulting in secondary organic aerosol (SOA). Modeling of SOA formation and aging

5.1 Introduction 27

processes requires accurate prediction of both SOA mass concentration and composition.

While traditional models underestimated the SOA burden (Volkamer et al., 2006) prediction

of SOA mass has recently been improved through the use of the volatility basis set framework

(Donahue et al., 2006; Robinson et al., 2007) although the system remains significantly

underdetermined (Hodzic et al., 2010). In contrast, efforts to reproduce ambient SOA

oxygenation levels using atmospherically relevant OH concentrations have been largely

unsuccessful. Here we explore the factors governing the degree of oxygenation of α-pinene

SOA produced via photooxidation in smog chamber experiments.

Analysis of aerosol mass spectrometer (AMS) data using the positive matrix factorization

(PMF) source apportionment method allows the total OA to be represented as a linear

combination of factors representing various sources and/or processes. The SOA fraction is

often represented in terms of semi-volatile oxygenated OA (SV-OOA) and low-volatility

oxygenated OA (LV-OOA) (Ulbrich et al., 2009; Jimenez et al., 2009; Metzger et al., 2010).

In addition to their volatility difference, SV-OOA has a lower atomic oxygen-to-carbon

(O : C) ratio (~ 0.25–0.6) than LV-OOA and often represents fresh aerosol closer to the

source, while LV-OOA, with a higher O : C ratio (~ 0.6–1.0) represents more aged OA (Ng et

al., 2010; DeCarlo et al., 2010; Lanz et al., 2010; Jimenez et al., 2009).

Two AMS mass fragments previously shown to be useful in describing atmospheric SOA

occur at m/z 44 (mostly particulate CO2+, with a minor contribution of C2H4O

+) and m/z 43

(mostly C2H3O+ for SOA, with a lesser contribution from C3H7

+). The f44 value, being the

ratio of the organic fraction of m/z 44 to total organics, has been empirically related to the

atomic O : C ratio for ambient measurements and is in large part derived from carboxylic

acids, of which a larger fraction is found in more aged aerosols (Aiken et al., 2008; Duplissy

et al., 2011). The ratio of organic m/z 43 to total organics (f43) is more closely related to

fragmentation of aldehydes, ketones and alcohols. The organic mass fractions in unit mass

resolution f44and f43are approximations for those in high resolution, being the organic mass

fractions of particulate CO2+ (f pCO2

+) and of C2H3O+ (f C2H3O

+). The f44 and f43 values

observed in ambient LV-OOA and SV-OOA factors obtained by PMF analysis of 43 ambient

datasets (Ng et al., 2010) typically fall within a triangular space. LV-OOA has high f44 and

low f43, while SV-OOA has lower f44 and a range of f43 values. The same triangle has been

transferred to high-resolution data within the Van Krevelen space (H : C vs. O : C) (Van

Krevelen, 1950; Ng et al., 2011a; Heald et al., 2010). SOA produced in smog chamber (SC)

studies typically falls within the range of ambient SV-OOA, almost always showing lower

degrees of oxygenation than ambient LV-OOA (Ng et al., 2010), with only a few exceptions

28 5 The link between organic aerosol mass loading and degree of oxygenation

(Bahreini et al., 2005; Chhabra et al., 2011). These mass spectral differences suggest that SC

SOA evaporates at lower temperatures than ambient LV-OOA. Although volatility is not

directly measured in most studies, Huffman et al. (2009) showed that SC SOA is typically

more volatile than ambient SOA. SOA from various non-biogenic pure precursors can reach

high degrees of oxygenation, although having higher or lower f43 values than ambient. In an

α-pinene oxidation flow reactor experiment f44 values as high as 0.25 were reached; however

these required OH concentrations that were orders of magnitude higher than in smog

chambers (Ng et al., 2010). Not only volatility but also hygroscopicity was related to the

atomic O : C ratio (Duplissy et al., 2011; 2008). More oxygenated particles take up more

water at a given relative humidity, suggesting increased cloud formation potential.

Measurements of dark α-pinene ozonolysis in a continuous-flow chamber by Shilling et al.

(2009) demonstrated that the aerosol chemical composition depends on the total organic mass

concentration; however the generated aerosol was less oxygenated than ambient LV-OOA.For

non-wall-loss-corrected organic mass loadings between > 140 µg m-3 and 0.5 µg m-3, O : C

values increasing from 0.29 to 0.45 and H : C values decreasing from 1.51 to 1.38 were

reported. Recently, LV-OOA-like aerosol was obtained in a smog chamber, but mostly by

starting with oxygenated gas-phase precursors (Chhabra et al., 2011), whereas most primary

volatile organic compound emissions are thought to be more hydrocarbon-like. However,

these results indicate that the location of SOA in the f pCO2+ - f C2H3O

+ as well as in the Van

Krevelen space is affected by precursor identity. Chhabra et al. (2011) showed O : C ratios for

α-pinene photooxidation of approximately 0.3–0.4 for organic mass concentrations of 54–

64 µg m-3 and OH exposures of 1.7–3.3 × 107 cm-3 h.

Three recent studies performed in flow reactors found SOA from α-pinene photooxidation

falling to the range of LV-OOA (Lambe et al., 2011; Massoli et al., 2010; Kang et al., 2011).

When high OH concentrations (e.g., 2 × 109 to 2 × 1010 cm-3 in Lambe et al. (2011)) are

produced in flow reactors, the probability that a low-volatility reaction product collides with

an OH radical and reacts further before it collides with a particle and condenses as SOA is

enhanced. Thus, the accelerated chemistry may lead to a different set of products, while

inhibiting slower condensed-phase reactions. Therefore the importance of heterogeneous

reactions may also be enhanced; Slowik et al. (2012) found that heterogeneous reactions at

OH exposures of 3 × 108 cm-3 h are sufficient to transform ambient biogenic aerosol to an

LV-OOA-like composition.

5.2 Method 29

The inability of smog chambers to generate α-pinene SOA with oxygenation, volatility and

(potentially) hygroscopicity similar to ambient LV-OOA could be an important gap in the

current understanding of ambient SOA. In this paper, results from an experimental study

using α-pinene, being, among other biogenic organic gases, an atmospherically relevant

precursor for SOA, are presented. The high non-fossil SOA fraction together with a high

contribution of monoterpenes to SOA in Marseille in summer time indicates that biogenic

VOC are important SOA precursors (El Haddad et al., 2013). The aim of this study is to find

the main driving factors responsible for the inability of smog chamber studies to yield LV-

OOA-like aerosol from the biogenic precursor α-pinene, even after the equivalent of tens of

hours of atmospheric aging. OH exposures of 3–14 × 107 cm-3 h were reached, which is

higher compared to the studies of Ng et al. (2011a) and Chhabra et al. (2011), but less than in

typical flow reactor studies.

5.2 Method

5.2.1 Experimental setup

Nine experiments (see Table 5.1) were carried out in the smog chamber (SC) of the Paul

Scherrer Institute (PSI): a 27 m³ Teflon bag suspended in a temperature-controlled wooden

housing. Four xenon arc lamps (4 kW rated power, 1.55 × 105 lumens each, XBO 4000

W/HS, OSRAM) were used to simulate the solar light spectrum. In addition, in seven of the

nine experiments 80 UV lights (Philips, Cleo performance 100 W) located underneath the SC

bag were illuminated to accelerate the aging process. The smog chamber housing is covered

inside with a reflective aluminum foil to maintain light intensity and light diffusion. A

comprehensive description of the PSI smog chamber is available in Paulsen et al. (2005).

During all experiments, photooxidation of the biogenic precursor α-pinene led to secondary

aerosol formation and growth. In general, experimental conditions were adjusted using the

following sequence: humidification of the chamber, addition of seed aerosol if applicable,

introduction of SOA precursor, addition of OH precursor, addition of NOx if applicable,

mixing period, turning-on of xenon and/or UV lights to generate OH radicals followed by a

reaction time of 5 to 22 h. The procedure for each individual experiment is described in Sect.

5.2.2.1 and summarized in Table 5.1.

30 5 The link between organic aerosol mass loading and degree of oxygenation

Table 5.1. Overview of experiment conditions. Org max (wlc) is the wall-loss-corrected organic mass concentration at the end of the experiment. Radiation sources are UV or xenon (Xe) lights. The OH tracer butanol-d9 is abbreviated with but-d9. Black carbon (BC) and NH4HSO4 represent seed aerosol in experiment 7.

Expt. Initial Org max RH NO NO2 Radiation Added

no. α-pinene (wlc) av(sd) av(sd) av(sd)* source(s)

[ppb] [µg m-3] [%] [ppb] [ppb]

1 7 4.5 50(1) 3.9(0.6) 4.9(1.8) UV+Xe HONO 2 14 17.9 48(2) 3.0(0.4) 5.9(0.9) UV+Xe HONO

3 20 6.2 40(2) 0.3(0.1) 0.8(0.3) UV+Xe HONO but-d9 SO2

4 22 1.8 44(1) 4.2(0.9) 20.0(9.9) UV+Xe HONO but-d9 SO2 NO

5 23 9.2 47(2) 0.6(0.9) 6.9(0.3)* UV+Xe but-d9 NO2

6 44 42.1 44(2) 0.3(0.8) 19.0(0.4)* Xe but-d9 O3 NO2 7 45 97.2 28(2) 1.0(0.2) 3.6(1.3) UV+Xe HONO BC & NH4HSO4 8 46 78.3 49(1) 2.0(1.1) 22.0(12.0) Xe HONO 9 50 66.2 47(3) 1.8(0.3) 9.0(3.0) UV+Xe HONO

*Initial NO2 addition that decayed to zero during experiments 5 and 6.

5.2.1.1 Introduction of particle and gas-phase reactants into the chamber

Liquid-phase α-pinene (98 %, Aldrich) and optionally the hydroxyl radical (OH) tracer (9-

fold deuterated butanol, 98 %, D9, Cambridge Isotope Laboratories), hereafter referred to as

butanol-d9, were sequentially injected with a syringe into an evaporation glass bulb heated to

80 °C. The two gas-phase compounds were carried with dilution and flush flows into the bag

(each 15 L min-1, maintained for 15 min) from an air purifier (737-250 series, AADCO

Instruments, Inc., USA), further referred to as “pure air”. The experiments were carried out at

40–50 % relative humidity (RH), except experiment 6, where the prevailing RH was

28 ± 2 %. The temperature (T) varied within a range of 21 °C to 24 °C between experiments.

Three NOx sources (HONO, NO and NO2) were added either in combination or separately

during the different experiments. Table 5.1 provides an overview of the initial α-pinene

concentrations, maximum wall-loss-corrected (wlc) organic mass concentrations, RH, initial

NO2 concentration before lights were switched on, average NO and NO2 concentrations

during the experiments, the radiation source(s) and the seed added.

HONO was used as a source of both NO and OH, except for experiments 5 and 6. It was

produced by continuous mixing of sodium nitrite (NaNO2, 1–3 mmol L-1) and sulfuric acid

(H2SO4, 10 mmol L-1) solutions in a reaction vessel by means of a peristaltic pump (Taira and

Kanda, 1990). The HONO product was entrained in 2.3–2.7 L min-1 pure air depending on the

experiment. Before lights on, 1–5 ppbv ( ± 10 %) of HONO was injected to provide NOx and

5.2 Method 31

enable photochemistry immediately after lights on (Initial concentrations: See supplementary

material Table S 5.1). The injection continued throughout the experiments to maintain a

similar NO concentration level.

In experiments 5 and 6, NO2 served as the NOx source, with 6.9 ± 0.7 ppbv and

19.0 ± 1.9 ppbv NO2 (purity: 98 %; 1005 ppmv ± 3 %), respectively, injected before lights on.

During experiment 4, NO (99.8 %; 1005 ppmv ± 2 %) was continuously injected in addition

to HONO, and the flow was decreased stepwise from 10.0 to 3.8 mL min-1 (0.37 to

0.14 ppbv min-1), which resulted in a slower increase of NO2 and a constant level of NO in the

chamber.

Before switching on the lights during experiment 7, a suspension of 50 µL L−1 black carbon

(printer ink, Tokai carbon, Japan) containing black carbon and a solution of 4 g L−1

ammonium hydrogen sulfate (NH4HSO4), both in water, were sequentially nebulized and

introduced into the chamber with 0.6 L min−1 and a dilution flow of 10 L min−1 to act as seed

particles. During experiments 3 and 4, 50 pptv SO2 (99.98 %; 502 ppmv ± 2 %) was injected

to photochemically produce H2SO4 nucleation, providing an aerosol surface rapidly and thus

accelerating SOA formation (Metzger et al., 2010). In experiment 6, eight hours after lights

on, 123 ± 4 ppbv of O3 were added to the existing concentration of 41 ± 1 ppbv to investigate

whether this would accelerate oxidation compared to the earlier period of the experiment. To

minimize SOA formation by -pinene ozonolysis, the injection was performed only after

99 % of the α-pinene had reacted.

Once all initial gas- and aerosol-phase components were present in the bag, a 20 to 30-min

mixing period was allowed before the lights were switched on. After each experiment the

smog chamber was cleaned by the injection of several ppmv of ozone for 5 h and irradiation

for 10 h with UV lights at 20 °C, followed by a flushing period with pure air and high relative

humidity (~ 70 %) at 30 °C for at least 20 h.

To ensure that the organic matter (OM) formed during the experiments is not significantly

influenced by background contamination in the smog chamber, blank experiments before,

during and after the campaign were carried out. The conditions of the five blank experiments

are listed in Table S 5.2 in the Supplement. During two blank experiments (B1 and B2) at RH

≈ 50 % with neither HONO nor seed aerosol present, the maximum mass concentration

measured by the SMPS was 0.16 µg m-3 and 0.03 µg m-3 after 5 h and 8 h exposure to UV and

xenon lights, respectively. A blank experiment (B3) with 6.6 ± 0.2 µg m-3 ammonium sulfate

((NH4)2SO4) seed, 10 ppbv HONO and a relative humidity of 61 ± 6 % yielded a peak organic

32 5 The link between organic aerosol mass loading and degree of oxygenation

mass concentration of 1.7 ± 0.1 µg m-3 30 min after lights on. A second and third seeding

period 1 h and 2 h after lights on resulted in a slight increase in the organic mass

concentration. Two additional seeded blank experiments (B4 with NO and B5 with HONO)

with a lower (NH4)2SO4 concentration yielded only 0.1–0.16 µg m-3 organic mass

concentration. The significantly higher background concentration in blank experiment B3 can

be explained by the contamination of the (NH4)2SO4 seed solution with organic compounds.

There is evidence that primary organic aerosol (~ 0.3 µg m-3) was injected with the seed. In

addition, it is likely that water-soluble organic compounds are injected with the aqueous seed,

which are then oxidized in the gas phase and form additional SOA. Moreover, the yield of

secondary organic mass is enhanced by the surface provided by the seed. Condensation occurs

earlier in the seeded blank experiments than for the unseeded blanks, where higher vapor

pressures are required for nucleation. Experiment 7 (which is most similar to the seeded blank

experiment) included a concentration of 69.2 µg m-3 and thus an organic mass concentration

well above the blank value. The organic mass concentration of the unseeded experiments is

well above that of the blank experiments B1, B2, B4, and B5.

In general, it is difficult to quantify the contribution of contaminants in the smog chamber to

the total organic aerosol mass over the course of an experiment. We estimate the time

evolution of contaminant mass using the multi-linear engine (ME-2, Paatero et al. (1999)), as

implemented in Igor Pro (Wavemetrics, Inc.) by Canonaco et al. (2013) ME-2 represents the

mass spectral time series recorded during the experiment as a linear combination of statistic

factor mass spectra and their time-dependent intensities, while allowing some or all of the

factor spectra to be constrained using a priori information. In this case, contaminant mass

spectra were constrained (a-value: 0) using one of the three selected spectra collected during

the blank experiment B3, while two additional left free. One hundred twenty iterations (40 for

each mass spectrum after the first, second and third seeding period, representing three

different aging times and chemical compositions) of the model using different randomly

distributed initial values resulted in a contribution of the constrained blank MS between 6.6

and 9.9 % in the first two hours to between 10 and 20 % in the last three hours of experiment

4 with the lowest organic mass concentration. A detailed description of the model runs and

results can be found in the Supplement (Figure S 5.1).

Another indication that the low-concentration experiments are not dominated by the

background contamination is shown in Figure S 5.2 of the Supplement. The chemical

composition of SOA in blank experiment B3 measured by the AMS differs from that

measured during the low-concentration experiments of the study as shown by their different

5.2 Method 33

location in the Van Krevelen space (H : C vs. O : C). We conclude that only experiment 4

with the lowest organic mass concentration in this study is impacted by a non-negligible

amount of contaminants from the smog chamber system.

5.2.1.2 Instrumental setup

Various instruments were used to monitor gas and aerosol properties in the PSI smog

chamber. A high-resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS,

Aerodyne Research, Inc., Billerica, MA, USA) was operated online to measure the chemical

composition (organics, ammonium, nitrate, sulfate, chloride) of non-refractory submicron

particles (DeCarlo et al., 2006). Because this instrument samples at a low flow rate

(0.1 L min-1), a supporting flow of ~ 3 L min-1 was maintained parallel to the AMS inlet to

minimize diffusive losses in the sampling lines. Gas-phase compounds with a higher proton

affinity than water (166.5 kcal mol-1) were measured with a proton transfer reaction mass

spectrometer (PTR-MS, Ionicon). Mass-to-charge ratios related to α-pinene (m/z 81 and m/z

137) and the OH tracer butanol-d9 (m/z 66) were analyzed in detail (see Sect. 5.2.2). A

chemiluminescence-based NOx instrument (Monitor Labs 9841A NOx analyzer) was attached

to the HONO source to monitor the injected concentration throughout the experiment. A

modified NOx instrument including a photolytic NO2-to-NO converter (Thermo

Environmental Instruments 42C trace level NOx analyzer) and two ozone monitors (Monitor

Labs 8810 ozone analyzer, Environics S300 ozone analyzer) monitored the gas phase in the

chamber. A scanning mobility particle sizer (SMPS, consisting of a TSI condensation particle

counter (CPC) 3022A and a TSI differential mobility analyzer (DMA) 3081) measured the

aerosol size distribution and a condensation particle counter (TSI CPC 3025A) the total

particle number concentration (diameter d > 3nm), respectively.

The unit mass and high-resolution data from the HR-ToF-AMS were corrected for

contributions from the gas phase. The contributions of multiple species to the same integer

m/z were deconvolved using a fragmentation-table-based approach (Allan et al., (2004);

Aiken et al., (2008)). The fragmentation tables were optimized for the current dataset, which

required changes at m/z 14, 36, 39, 40, 46, 47, 48, 64, 65, 80, 81, and 98, as discussed in the

Supplement (see Table S 5.3).

The total organic mass concentration was derived using the chemical composition (organics,

ammonium, nitrate, sulfate, chloride) measurements from the HR-ToF-AMS and the total

34 5 The link between organic aerosol mass loading and degree of oxygenation

volume measured with the SMPS. The ratio of SMPS to HR-ToF-AMSvolume, after applying

compound-specific densities, results in a correction factor applied to the measured HR-ToF-

AMS organic mass concentration (details in the Supplement). The data were further corrected

for wall losses, as described in Sect. 5.2.3.

5.2.2 Estimation of OH exposure

Depending on the design of smog chamber experiments, the OH concentration and thus the

photochemical age of the reaction system can vary considerably between experiments.

Furthermore, within a single experiment the photochemical age is not necessarily directly

proportional to the light exposure time. Therefore we here discuss reaction time in terms of

OH exposure (unit: cm-3 h), defined as the OH concentration integrated over time. This scale

is preferable when comparing the evolution of experiments, especially when observing

oxidation processes, as it is the chemically relevant aging time. In this study we determine

OH exposures combining two methods: an OH tracer method introduced by Barmet et al.

(2012) and a method based on the initial decay of the gas-phase precursor α-pinene.

5.2.2.1 OH tracer method

Via a heated glass bulb, 1.5 µL (corresponding to ~ 14.5 ppbv) of the OH tracer butanol-d9

(detected as M+H+-H2O at m/z 66) was injected into the chamber (Sect. 5.2.1.1). The OH

tracer was present during the last hour of experiment 5, the second half of experiment 6, and

for the entire lengths of experiments 3 and 4. Butanol-d9 is consumed by reaction with OH

and diluted by the HONO input described in Sect. 5.2.1.1 (we assume as a first approximation

a constant chamber volume (V)) following the reaction kinetics governed by Eq. (5.1):

butanol‐d9OH,butanol‐d9 ∙ OH ∙ butanol‐d9 ∙ butanol‐d9 , (5.1)

where kOH, butanol-d9 = 3.4 × 10-12 cm3 molecules-1 s-1, V represents the chamber volume

(27 m3), and fdil the dilution flow [m3 s-1].

OHexposure,

∙ ∆

∆d (5.2)

The slope of the temporal decay of ln(butanol-d9) at each time step was divided by the

reaction rate constant kOH, butanol-d9 and corrected for the dilution to derive the OH

5.2 Method 35

concentration at each time step. Integrating this term with time (Eq. (5.2)) results in the OH

exposure (Barmet et al., 2012).

5.2.2.2 α-Pinene method

Similar to the OH tracer decay, the decay of α-pinene can be utilized to derive the OH

concentration. Compared to the OH tracer method, limitations of this method are the more

rapid consumption of α-pinene and its simultaneous reaction with both OH and O3 as shown

by Eq. (5.3). The OH exposure was therefore derived from the decay of ln(α-pinene), while

correcting for the time-dependent reaction with O3 and, if present, for the dilution of the

chamber air due to the HONO input (described in Sect. 5.2.2.1) as shown in Eq. (5.4).

α‐pineneOH ∙ OH ∙ α‐pinene ∙ O ∙ α‐pinene ∙ α‐pinene , (5.3)

where kOH = 5.3 × 10-11 cm3 molecules-1 s-1 and kO3 = 8.9 × 10-17 cm3 molecules-1 s-1,

OHexposure ∙ ∆

∆∙ O d (5.4)

Only data fulfilling the criterion α-pinene ≥ 1 ppbv were used to ensure all measurements fell

well above the PTR-MS detection limit. After α-pinene decayed below 1 ppbv, the OH

concentration was assumed to be constant for the rest of the experiment time for experiments

with continuous HONO addition.

5.2.2.3 Application of the methods to the dataset

The two estimates of OH exposure were compared in experiments 3 and 4, where both α-

pinene and butanol-d9 were above detection limit throughout the experiment (Figure S 5.3,

Supplement). The methods agree within 20 % for later in the experiment, which shows that

the α-pinene method is valid at least for experiments with constant HONO input and thus a

constant source of OH radicals.

The α-pinene method was utilized to derive the OH concentration of experiments 7, 8 and 9

with constant HONO input and α-pinene above detection limit for at least 1.3 h. The OH

exposure from a repeat experiment using the butanol-d9 method was applied to experiments 1

and 2, conducted at substantially higher HONO/α-pinene ratios (0.2 and 0.4, respectively).

The fast decay of α-pinene within the first 30 min during those two experiments due to the

36 5 The link between organic aerosol mass loading and degree of oxygenation

high accumulated HONO concentration before lights on and off-gassing from the walls after

lights on, as previously measured by Metzger et al. (2008), gives rise to an overestimation of

the OH exposure when using the α-pinene method. This is due to the assumption of a constant

OH concentration once the α-pinene is consumed. The repeat experiment shows the same

characteristics in the α-pinene decay, with the advantage of having the OH tracer butanol-d9

present for the whole experiment. A comparison of the three experiments is shown in Figure

S 5.4 of the Supplement. By applying the OH exposure of the repeat experiment to

experiments 1 and 2, an overestimation of the OH concentration later in the experiment could

be avoided.

In the course of experiments 5 and 6, the initially added NOx concentration decayed to very

low concentrations, causing a substantial change in the gas-phase chemistry including the

production rate of ozone and thus OH and the fate of peroxy radicals. An indication of

changes in chemistry is the change in the steepness of the ozone increase during the

experiments 5 (1.5 h after lights on) and 6 (5.5 h after lights on), once the NOx decayed to low

concentrations. One may assume that RO2 radicals react less with NO in this “low NOx”

regime and more with HO2 and RO2 forming e.g., peroxides. However there was no direct

measure of RO2 or peroxides available.

In experiment 6, the ozone leveled off before being increased artificially by the addition of

123 ± 4 ppbv of O3 eight hours after lights on. For those two experiments, OH exposures

derived from the α-pinene decay in the beginning and the decay of butanol-d9, added towards

the end of the experiment, were merged. The mean OH concentration in the smog chamber

was indirectly determined by dividing the OH exposure by the respective time period. The

highest OH concentration was 5 × 107 cm-3 and occurred at the beginning of experiments 1

and 2. On average, the OH concentrations during the experiments were in the range of 0.12–

2.7 × 107 cm-3. A compilation of the derived OH exposures is shown in Figure S 5.5 in the

Supplement. When comparing experiments utilizing the α-pinene-derived OH exposure, note

that the OH concentration is assumed to be constant after the drop below the threshold of

1 ppbv α-pinene. This extrapolation leads to uncertainties of the α-pinene derived OH

exposure, which increases with experiment time. Nevertheless these uncertainties do not alter

the main outcome of the study.

5.3 Results 37

5.2.3 Wall loss correction

The measured aerosol mass concentration in the smog chamber is the net result of mass

produced in the chamber and mass lost to the walls. Wall loss rates were determined from a

period near the end of the experiment where chemical changes to the aerosol have slowed to

the point that wall losses are expected to be the dominant factor controlling changes in the

measured particle mass. The measured (AMS) aerosol mass during this period was fitted with

an exponential. The data were then corrected by dividing the measured organic mass

concentration by the exponential decay (see Figure S 5.6 in the Supplement). The estimated

half-lives ranged between 2.8 and 10 h, the maximum derived from experiment 8, with a very

short period to fit. Selecting a period when minor mass production is still possible leads to a

lower limit for the wall loss correction assuming no major fragmentation due to the UV light

exposure and no major loss of vapors to the walls. This may lead to some underestimation of

the aerosol loading, which however is difficult to quantify.

5.3 Results

5.3.1 General reproducibility of the aerosol degree of oxygenation

The vaporization and ionization processes in the AMS fragment the sampled aerosol

compounds. As described in the Introduction, the fraction of particulate CO2+ (f pCO2

+) and

the corresponding organic mass fraction f44 (defined as the organic signal at m/z 44

normalized to the total organic mass) are related to carboxylic acids and proportional to the

atomic O : C ratio, whereas the fraction of fragment C2H3O+ (f C2H3O

+) and the corresponding

organic mass fraction f43 are related to aldehydes, ketones and alcohols. The higher the O : C,

the more oxygenated the aerosol. The f44-f43 space was used to separate the two factors LV-

OOA and SV-OOA (Ng et al., 2010) and is used here in Figure 5.1a and Figure 5.1b to

describe the chemical evolution of aerosol derived from α-pinene photooxidation in this

study. Figure 5.1a shows 30-min averages of f pCO2+ vs. f C2H3O

+ from high-resolution data,

while Figure 5.1b compares 30-min averaged data of this study with other studies (Chhabra et

al., 2011; Kang et al., 2011; Lambe et al., 2011) in the f44-f43 space with each of the

experiments marked by a different symbol. The legend in Figure 5.1a gives the final organic

mass concentration reached in each experiment after wall loss correction. Despite varying

38 5 The link between organic aerosol mass loading and degree of oxygenation

organic mass concentrations, OH concentrations, NOx concentrations and RH in the smog

chamber (see Table 5.1), all experiments are located within the same region on the right-hand

side of the triangle originating from ambient data (Ng et al., 2010). Experiment 6, 7 and 8

show an initial decrease in f pCO2+due to the condensation of early-generation semi-volatile

organic compounds, as shown previously in Baltensperger et al. (2005). During the ongoing

oxidation reactions in the smog chamber more highly oxygenated products are formed,

resulting in an increased f pCO2+ together with a decreased f C2H3O

+. This causes movement

from the lower right to the upper left in Figure 5.1a and Figure 5.1b with increasing OH

exposures, indicated by the arrows. Together with e.g., SOA from wood burning, which is

found on the left-hand side (Heringa et al., 2011), the mixture of aerosol from several sources

can explain a wide range of the empirically derived triangle. The three arrows indicating

f pCO2+ - f C2H3O

+ pairs at the same OH exposure are explained in Sect. 5.3.2.

5.3 Results 39

Figure 5.1. (a) Thirty-minute averages (except expt. 4: 60min) of f pCO2+ vs. f C2H3O

+ color coded for the nine experiments. The legend shows the final wall-loss-corrected organic mass concentration at the end of each experiment. The three arrows indicate the f pCO2

+ - f C2H3O+ pair when an OH exposure of

3.5 × 107 cm-3 h is reached for selected experiments (4, 3, 6). Dashed lines represent the observed range of ambient SOA (Ng et al., 2010). (b) Unit mass resolution organic mass fraction f44 vs. f43 color coded for the nine experiments. For comparison, environmental chamber (Chhabra et al., 2011) and flow reactor data (Kang et al., 2011; Lambe et al., 2011) of α-pinene photooxidation experiments are shown and color coded by the OH exposure.

Figure 5.1b shows the organic mass fractions f44 vs. f43 in unit mass resolution for the nine

experiments of this study compared to α-pinene photooxidation studies in three other systems.

Next to the color code by experiment number of this study, the data of Kang et al. (2011),

Lambe et al. (2011) and Chhabra et al. (2011) are color coded with the OH exposure. Chhabra

et al. (2011) measured in an environmental chamber (Caltech dual 28 m³ Teflon laboratory

chambers), while Lambe et al. (2011) and Kang et al. (2011) utilized flow reactors (Potential

aerosol mass chamber, Kang et al., 2007) for their experiments. Chhabra et al. (2011) show

slightly lower f44 and higher f43 at organic mass concentrations of 64 and 54 µg m-3,

respectively, while the experiments were performed at much lower RH (4.2 %; 4.9 %) and

with different OH precursors (H2O2; CH3ONO) and NOx concentrations (0; 800 ppbv)

present. In the study of Kang et al. (2011) the organic mass concentrations (90–110 µg m-3)

40 5 The link between organic aerosol mass loading and degree of oxygenation

are similar to the high-concentration experiments in this study and for low OH exposures, and

the location in the f44-f43 space is comparable. For higher OH exposures, the oxidation process

continues on a similar slope as in this study, indicating further aging. Lambe et al. (2011)

showed comparable pairs of f44-f43 for low OH exposures, while for higher OH exposures

much higher f44 was found, which the authors suggested to be due to the loss of mass after

fragmentation in the particle takes place, seen by shrinking particles for higher OH exposures

(Fig. S2, Lambe et al., 2011).

Thus, for the first time, an organic mass fraction f44 as high as many ambient LV-OOA and

well above the SV-OOA range is obtained from an α-pinene photooxidation smog chamber

study at ambient-like OH concentrations.

Figure 5.2 shows the Van Krevelen diagram (H : C vs. O : C) of the nine experiments (30-min

averaged data, except expt. 4: 60-min average). The relationship between f44and O : C, as well

as f43 and H : C is shown in Figure S 5.7a and Figure S 5.7b for data of this study. The

diagram allows for interpretation of changes in functionality (slope -2: aliphatic → carbonyl

functionality; slope 0: aliphatic → alcohol or peroxide functionality; slope -1: addition of

carboxylic acid functionality). Heald et al. (2010) found an average slope of -1 for bulk

atmospheric and laboratory OA. The ongoing oxidation in the nine experiments results in an

increased O : C together with a decrease in H : C, seen in a movement from the upper left to

the lower right in the figure. All experiments fall within the range of ambient data indicated

by the dashed lines (Ng et al., 2011a) and follow a slope of roughly between -1 and -2,

indicated by the solid lines.

5.3.2 Dependence of degree of oxygenation on the organic mass concentration

Figure 5.1a shows three arrows pointing to f pCO2+ - f C2H3O

+ pairs at the same OH exposure

of 3.5 × 107 cm-3 h (averaged ± 15 min), isolating the effect of organic mass concentration for

three selected experiments (expt. no. 3, 4, 6) during which the OH tracer butanol-d9 was

present in the chamber for most of the experiment. The respective OH exposures were derived

according to the methods described in Sect. 5.2.2. The organic mass concentrations (wlc,

averages and standard deviation for ± 15 min, except expt. 4: 60 min) at this OH exposure in

experiments 6, 3 and 4 were 42.3 ± 0.6 µg m-3, 5.8 ± 0.2 µg m-3 and 1.7 ± 0.1 µg m-3,

respectively.

5.3 Results 41

Figure 5.2. Van Krevelen diagram (H : C vs. O : C ratio) of the nine experiments together with the observed range of ambient SOA represented by dashed lines. The data are averaged over 30 min (except expt. 4: 60 min).

The corresponding averaged f pCO2+ values were 0.106 ± 0.001, 0.131 ± 0.002 and

0.140 ± 0.014, respectively. Seen by the differences in f pCO2+ for the three different mass

loadings, the data clearly show that at the same OH exposure, the degree of oxygenation is

dependent on the total organic mass concentration present in the chamber.

For a comprehensive comparison of all experiments described in this study, the OH exposure

was utilized as the aging scale, rather than the time after lights on. The evolution of O : C as a

function of organic mass concentration (wlc), both averaged for 10 min, is shown in Figure

5.3 with each of the nine experiments marked by a different symbol and color-coded by the

OH exposure. Initially, there is a period of rapid mass increase with a slow increase in O : C,

followed by a later period of a continuous O : C increase with only a slight further increase in

mass. The precise rate and termination point of the late-experiment mass increase are

somewhat uncertain due to corresponding uncertainties in the applied minimum wall loss

correction method (Figure S 5.6, Supplement).

42 5 The link between organic aerosol mass loading and degree of oxygenation

Figure 5.3. O : C ratio as a function of the organic mass concentration (µg m-3) in the chamber for the nine experiments. The color code indicates the OH exposure. Thirty-minute mean values at selected OH exposures are represented by colored squares (except expt. 4: 60-min average). The colored lines represent linear least orthogonal distance fits between O : C and org (wlc) at three selected OH exposures with a mean slope of (0.07 ± 0.01) % m3 µg-1.

The initial period is likely governed by condensation of early-generation, semi-volatile

products and their condensation therefore yields a low increase in O : C. As aging proceeds,

the gas-phase organics become more oxidized and thus more oxygenated (higher O : C)

compounds condense. Additional aging of the particles is possible through repartitioning of

semi-volatile condensed-phase OA to the gas phase, followed by gas-phase oxidation and re-

condensation of the oxygenated products. This process also increases O : C, while only

slightly affecting the total OA mass. Therefore, partitioning is more strongly affected by the

significantly varying mass concentrations between the nine experiments than the slight mass

increase within a given experiment. Higher OA mass correspond to increased condensation of

semi-volatile compounds, while low-volatility compounds occur in the particle phase even at

low concentrations (Donahue et al., 2006). Ng et al. (2010) also suggest that more oxygenated

5.3 Results 43

compounds tend to be less volatile. The reaction of less oxygenated, semi-volatile compounds

with OH proceeds in the gas phase until either (1) they become sufficiently oxygenated to

condense (which requires a higher level of oxygenation than at high mass concentrations) or

(2) they fragment sufficiently such that they will not enter the particle phase. Both pathways

lead to an increase in the net O : C for experiments conducted at low OA concentrations.

In Figure 5.3 the color-coded squares correspond to average O : C ratios and OA mass

concentration (wlc) at OH exposures of 2.5, 4, 6, 8 and 11 × 107 cm-3 h. Averages were

calculated as ± 15 min from the given OH exposure, except for expt. 4, where a ± 30-min

average was calculated. Linear (least orthogonal distance) fits of O : C as a function of OA

mass concentration at 2.5, 4 and 6 × 107 cm-3 h OH exposure are shown by the lines in the

corresponding color. Table 5.2 presents 30-min averages (except expt. 4: 60-min) of O : C

and organic mass concentration (wlc) for the five different OH exposures. The significant

dependence of O : C on the organic mass is shown by the slope ∆O : C/∆org (wlc), on average

(0.070.01) % m3 µg-1, with slopes and intercepts listed in Table 5.3.

Table 5.2. Thirty-minute averages of O : C and wall-loss-corrected organic mass concentration in µg m-3, org (wlc) at the given OH exposures. The data are displayed in Figure 5.3.

OH exposure [× 107 cm-3 h]

2.5 4 6 8 11 Expt. no.

O:C Org

(wlc) O:C

Org (wlc)

O:C Org (wlc)

O:COrg (wlc)

O:COrg (wlc)

1 0.54 3.2 0.60 4.4 0.62 4.5 0.64 4.6 0.66 4.5 2 0.54 15.6 0.56 18.1 0.59 18.0 0.61 18.0 0.65 18.0 3 0.52 6.0 0.54 5.7 0.53 5.5 0.54 5.7

4* 0.51* 1.5* 0.55* 1.6* 0.54* 1.7* 5 0.52 9.1 6 0.46 41.5 7 0.49 92.9 0.51 97.3 8 0.45 77.8 0.47 78.2 9 0.49 62.2 0.51 65.3 0.53 65.3

*Data are 60-min averages

44 5 The link between organic aerosol mass loading and degree of oxygenation

Table 5.3. Slopes and intercepts of linear (least orthogonal distance) fits between O : C and org (wlc). Fit lines are shown in Figure 5.3 with the OH exposure as color code.

OH exposure [× 107 cm-3 h]

O:C = a+b·org (wlc) 2.5 4 6

a 0.53 ± 0.01 0.57 ± 0.01 0.57 ± 0.02

b [m3 µg-1] -0.00071 ± 0.00022 -0.00084 ± 0.00027 -0.00056 ± 0.00079

This implies that at typical atmospheric concentration levels, partitioning effects on SOA

oxygenation are highly sensitive to the organic mass concentration. Using typical atmospheric

organic mass concentrations in smog chamber studies is not only crucial to reproduce the

volatility and degree of oxygenation of ambient SOA but also to accurately determine cloud

condensation nuclei activity due to the dependence of the hygroscopicity on the O : C ratio

(Duplissy et al., 2011). Ng et al. (2011b) found an even stronger dependence of O : C on the

organic mass concentration (0.55 % m3 µg-1) of ambient data with an unknown distribution of

OH exposures (possibility of low OH exposure for high OA mass and high OH exposure for

low OA mass).

5.3.3 Dependence of degree of oxygenation on the OH exposure

The degree of oxygenation reached by α-pinene SOA depends on both the organic mass

concentration (discussed above) and the OH exposure. For each experiment, the slope of

∆O : C/∆(OH exposure) was fitted (least orthogonal distance fit) for the part of the

experiment dominated by aging (rather than mass production), defined as the period after the

peak of suspended organic mass concentration was reached in the chamber. The individual

slopes are shown in Figure S 5.8 and summarized in Table S 5.4 in the Supplement. An OH

exposure between 3 and 25 × 107 cm-3 h is needed to increase O : C by 0.05.

Table 5.4 shows OH exposures (and corresponding organic mass concentrations (wlc))

required to obtain O : C > 0.6, a typical lower limit value for LV-OOA. For experiments

during which this threshold of 0.6 was not reached, the derived slopes of ∆O : C/∆(OH

exposure) and intercepts were utilized to estimate the total OH exposure needed to do so. The

OH exposure observed for oxidation of SOA to an O : C value of ~ 0.64 in the Mexico City

5.3 Results 45

plume (~ 4–5 × 107 cm-3 h, 6 h air transport time) as estimated from aircraft measurements

(DeCarlo et al., 2010; Dusanter et al., 2009) is, despite different precursors, on the lower edge

of the range of OH exposures required in this study.

Table 5.4. The OH exposure required to increase the O : C ratio above 0.6 at the corresponding organic mass concentration org (wlc). For experiments that do not exceed the threshold of 0.6, the OH exposures are extrapolated and the maximum measured org (wlc) is given, both indicated by the asterisk.

Expt. no. OH exposure org (wlc) [107 cm-3 h] [µg m-3] 1 4.7 4.4 2 7.0 18.1 5 7.2* 9.2* 7 9.7* 97.2* 6 11.6* 42.1* 9 12.5* 66.2* 8 19.8* 78.3* 3 23.7* 6.2* 4 35.2* 1.8*

5.3.4 Classification of chemical composition using reference mass spectra

To relate the aerosol chemical composition found in this study to ambient data, the SC

organic mass spectra were compared to reference mass spectra retrieved from ambient

measurements. These reference spectra were obtained from PMF analysis of a number of

ambient datasets and compiled by Ng et al. (2011a) and are available on the AMS Spectral

Database (Ulbrich et al., 2009). Figure 5.4 shows R² values (Pearson correlation) of the LV-

OOA (filled circles) and SV-OOA (empty circles) unit mass spectra from (Ng et al., 2011a)

with unit mass spectra from this study as a function of wall-loss-corrected organic mass

concentration. For this comparison, all organic mass spectra are converted to be consistent

with the fragmentation table of Aiken et al. (2008). The mass spectra for each of the nine

experiments were averaged for 30 min surrounding the OH exposures of 2 × 107 cm-3 h and

4 × 107 cm-3 h, respectively (i.e., the selected exposure ± 15 min). All organic m/z calculated

as a constant fraction of m/z 44 in the fragmentation table (i.e., m/z 16, 17, 18, 19, 20 and 28)

were excluded from the correlation test to avoid overweighting of these ions. The general

trend for both exposures shows increased correlation with LV-OOA and decreased correlation

with SV-OOA for decreasing mass concentrations. For all mass concentrations, an increasing

46 5 The link between organic aerosol mass loading and degree of oxygenation

Figure 5.4. Squares of the Pearson correlation coefficients, R², of measured mass spectra in comparison with LV-OOA (filled symbols) and SV-OOA (empty symbols) reference spectra as a function of the organic mass concentration (wlc). The color code represents the OH exposure at the midpoint of a 30-min mass spectral average. The corresponding mass spectra are shown in Figure S 5.10 in the Supplement.

R² (LV-OOA) and a decreasing R² (SV-OOA) with increasing OH exposure is found. The

dependence on mass concentration occurs for the partitioning-related reasons discussed above

in conjunction with increased O : C for low mass concentrations, while the dependence on

OH exposure occurs because SOA spectra become more LV-OOA-like with photochemical

age. For the high-concentration experiments, the difference between R² (SV-OOA) and

R² (LV-OOA) is more pronounced for the lower OH exposure than for the higher OH

exposure. Nevertheless, the lower R² (LV-OOA) of the high-concentration experiments

compared to the low-concentration experiments even for the enhanced OH exposure suggest a

substantial difference in the chemical composition as a function of the organic aerosol

loading. The low-concentration experiments (organic mass < 20 µg m-³ in Figure 5.4)

correlate strongly with LV-OOA (R² > 0.9) for an OH exposure of 4 × 107 cm-3 h. This

indicates that for low organic mass concentrations, SOA from smog chamber α-pinene

photooxidation yields LV-OOA-like mass spectra after a sufficient oxidation time. Also for

5.4 Conclusions 47

higher OH exposures, the correlation to LV-OOA is very high (R² > 0.89) for the low-

concentration experiments, while the R² values with SV-OOA decrease dramatically (Figure

S 5.9, Supplement).

5.4 Conclusions

A series of smog chamber experiments were conducted to identify main reasons for the

discrepancy between ambient and smog chamber aerosol degree of oxygenation, which in

turn strongly affects aerosol volatility and hygroscopicity. In this study, the aerosol products

from α-pinene photooxidation are located on the right-hand side of the f44-f43 space bounding

the range of PMF factors previously identified in ambient SOA (Ng et al., 2010). The O : C

ratio, a proxy for the degree of oxygenation, is increasing with decreasing organic mass

loadings. This has been shown in a comparison of nine smog chamber experiments where the

OH exposure was estimated to isolate the effects of mass loading and photochemical age.

Similar dependencies of f44 on the precursor concentration have been shown previously for

other compounds than α-pinene, such asβ-caryophyllene (Alfarra et al., 2012).

We conclude that low (near-ambient level) organic mass loadings are required in order to

obtain oxygenation levels comparable to those of the low-volatility oxygenated organic

aerosol (LV-OOA) PMF factors retrieved in numerous field campaigns. An OH exposure

between 3 and 25 × 107 cm-3 h is needed to increase O : C by 0.05 during aerosol aging. This

study marks the first smog chamber production of LV-OOA-like SOA from the

atmospherically relevant precursor α-pinene. For ozonolysis of α-pinene, the degree of

oxygenation falls in the range of SV-OOA even at low concentrations (Shilling et al., 2009;

Ng et al., 2010). This is consistent with the hypothesis of Donahue et al. (2012) that further

aging occurs with OH after/together with ozonolysis. A correlation test between measured

SOA and reference LV-OOA (Ng et al., 2011a) aerosol mass spectra shows Pearson’s R²

values larger than 90 % for a sufficient OH exposure in the chamber in low mass

concentration experiments (~ 1.2–20 µg m-3). We suggest that smog chamber studies must not

only be performed at reasonable oxidant concentrations but also at near-ambient mass

concentrations to accurately simulate the chemical properties, volatility and hygroscopicity of

ambient aerosol.

48 5 The link between organic aerosol mass loading and degree of oxygenation

Acknowledgements. This work has been supported by the EU 7th Framework projects

EUROCHAMP-2 and PEGASOS, as well as the Swiss National Science Foundation. We

thank Rene Richter and Günther Wehrle for their technical support at the smog chamber, and

in addition Imad El-Haddad, Stephen Platt and Robert Wolf for discussions and their support

during the experiments.

5.5 Supplementary material 49

5.5 Supplementary material

Table S 5.1. Overview of HONO input into the smog chamber before switching on the lights. The last column contains the ratio between HONO and α-pinene initial concentrations.

Expt. initial HONO α-pinene HONO/α-pinene

no. ppbv ( ± 10 % instrument

accuracy) ppbv

1 1.6 7 0.2

2 4.9 14 0.4

3 1.0 20 0.05

4 1.0 22 0.05

7 1.9 45 0.04

8 2.8 46 0.06

9 5.1 50 0.1

50 5 The link between organic aerosol mass loading and degree of oxygenation

Table S 5.2. Conditions of blank experiments B1-B5. The SMPS mass concentration is only given for blank experiments without seed. Detection limit is abbreviated with DL.

Blank

no.

Maximum

Organic

mass

RH T NO NO2 radiation added Previous

av(sd) av(sd) av(sd) av(sd) source seed SC use

µg·m-3 % °C ppb ppb ppb µg·m-3

(B1)

SMPS: 0.03

(Suspended)

AMS: below

DL

Ca. 50* Ca. 22* 0.9(0.3) no

data UV+Xe - -

moped emissions

(up to

100 µg·m-3)

(B2)

SMPS: 0.16

(Suspended)

AMS: below

DL

49(2) 21.8(0.6) 1.0(0.3) Below

det.lim. UV+Xe - -

α-pin photooxid.

(2–73 µg/m³)

moped emissions

(up to

100 µg·m-3)

(B3) AMS: 1.7

(Suspended) 61(6)** 25.2(1.4) 3.3(0.6)

Below

det.lim. UV+Xe

HONO

10

(NH4)2SO4

(6.6 ± 0.2)

α-pin photooxid.

(1.4–80 µg·m-3)

(B4) AMS: 0.16

(wlc) 80-85 UV+Xe

NO

40***

(NH4)2SO4

(1.8)

α-pin photooxid.

(1.4–80 µg·m-3)

(B5) AMS: 0.11

(wlc) 80-85 UV+Xe

HONO

~ 2

(NH4)2SO4

(1.2)

α-pin photooxid.

(1.4–80 µg·m-3)

* During blank experiment B1, no radiation shielded T/RH measurement existed. 55 % RH

and 19.5 °C were measured in darkness before lights were switched on. Assuming the

temperature increases by 3 °C after lights are switched on, leads to a RH of 50 %.

** Before lights were switched on, 85 % RH and 19.2 °C were measured. After lights were

switched on, the temperature increased to 25 °C, resulting in a decreased RH of 60.9 %.

*** 40 ppbv of NO added to the blank before lights were switched on.

5.5 Supplementary material 51

Table S 5.3. Changes in unit mass resolution fragmentation table compared to Aiken et al. (2008).

m/z Expt. 7

(with NH4HSO4 seed)

Expt. 1-6; 8-9

(without NH4HSO4 seed)

Aiken et al. (2008)

(standard frag table)

Organics

14 3.84·frag_organic[13] (0.14-4.17)·

frag_organic[13]

16 0.107·frag_organic[17] 0.107·frag_organic[17] 0.25·frag_organic[17]

28 0.93·frag_organic[44] 0.93·frag_organic[44] 1·frag_organic[44]

30 0.16·frag_organic[29] 0.16·frag_organic[29] 0.022·frag_organic[29]

36 36,-frag_air[36] 36,-frag_air[36]

37 37 37 37,-frag_chloride[37]

38 38,-frag_air[38] 38,-frag_air[38] 38,-frag_chloride[38],-

frag_air[38]

39 5·frag_organic[38] 5·frag_organic[38]

40 40,-frag_air[40] 40,-frag_air[40]

46 0.025·frag_organic[44] 0.025·frag_organic[44]

47 47 47

48 4·frag_organic[62] 0.19·frag_organic[62] 0.5·frag_organic[62]

64 0.35·frag_organic[50]+

0·frag_organic[78]

0.07·frag_organic[50]+

0.03·frag_organic[78]

0.5·frag_organic[50]+

0.5·frag_organic[78]

65 0.55·frag_organic[51]+

0.15·frag_organic[79]

0.55·frag_organic[51]+

0.1·frag_organic[79]

0.5·frag_organic[51]+

0.5·frag_organic[79]

80 2·frag_organic[94] 1.1·frag_organic[94] 0.75·frag_organic[94]

81 0.4·frag_organic[67]+

0.25·frag_organic[95]

0.3·frag_organic[67]+

0.35·frag_organic[95]

0.5·frag_organic[67]+

0.5·frag_organic[95]

98 0.65·frag_organic[84]+

0.55·frag_organic[112]

0.5·frag_organic[84]+

0.7·frag_organic[112]

0.5·frag_organic[84]+

0.5·frag_organic[112]

air

14 14,-frag_nitrate[14],-

frag_organic[14]

14,-frag_nitrate[14],-

frag_organic[14]

14,-frag_nitrate[14]

52 5 The link between organic aerosol mass loading and degree of oxygenation

NH4

16 0.77·frag_NH4[17] 0.77·frag_NH4[17] 16,-frag_water[16],-

frag_air[16],-

frag_sulfate[16],-

frag_organic[16]

NO3

46 46,-frag_organic[46] 46,-frag_organic[46] 46

K

39 39,-frag_organic[39] 39,-frag_organic[39] 39

Correction for collection and transmission efficiency

The volumes derived from the two instruments SMPS and AMS were compared, applying the

following densities in g/cm³ to the AMS species: ρOrg=1.4; ρSO4=1.78; ρNO3=1.72; ρNH4=1.75;

ρChl=1.4. The AMS data were corrected by applying the following correction factors (CF) to

the organic mass concentration based on the ratio volumeSMPS/volumeAMS. Possible reasons

for a disagreement between SMPS and AMS are given in brackets.

CF=2.7 for experiment 1 (volume weighted dm = 60-70nm)

CF=1.5 for experiment 2 (volume weighted dm = 70nm)

CF= (1.0-1.2) for experiments 3, 4, 5, 6, 8, 9 (volume weighted dm=150-200nm)

CF=1.5 for experiment 7 (bouncing of sulfate)

For experiments 1 and 2, the volume weighted mean diameter is on the lower edge of the

AMS measurement range and therefore the AMS samples significantly less than the SMPS.

As all experiments except expt. 7 were nucleation experiments, the chemical composition is

not expected to vary substantially as a function of diameter, relevant for expt. 1.

5.5 Supplementary material 53

Table S 5.4. Slope of O : C/∆(OH exposure) for the period where aging dominates (see Figure S 5.8) and the OH exposure required to increase O : C by 0.05.

Expt. no. ∆O:C/∆(OH exposure) ∆OH exposure/∆O:C [10-9 cm3 h-1] [107 cm-3 h/0.05]

4 0.20 24.50

3 0.35 14.43

8 0.84 5.96

1 0.93 5.37

2 1.10 4.55

9 1.11 4.50

6 1.52 3.29

7 1.57 3.18

5 1.63 3.07

54 5 The link between organic aerosol mass loading and degree of oxygenation

Contribution of blank experiment B3 to experiment 4

We used the mass spectra (MS) measured after the first, second and third seeding period

during blank experiment B3, representing different aging times and thus chemical

composition, as input for the statistical tool ME-2 (multi-linear engine: model by Paatero et

al. (1999), analysis interface by Canonaco et al. (2013) to estimate its contribution to the total

organic mass concentration. For the ME-2 runs, the blank experiment mass spectrum was

fixed (a-value: 0) and two more free components were allowed (similar to the approach of

Lanz et al. (2007) for ambient measurements). 120 iterations (40 for each blank experiment

MS) of the model using different randomly distributed initial values resulted in an average

contribution of the constrained blank MS between 6.6–9.9 % in the first two hours up to 10–

20 % in the last three hours of experiment 4 with the lowest organic mass concentration. For

the model runs, the m/z range of 12–250 was used. 5 out of 40 iterations resulted in the

following time series and mass spectra. The spectrum of blank experiment B3 was fixed as

factor 1, shown in Figure 5.1a and Figure 5.1b. Figure 5.1a shows a representative time series

of the three factors found for experiment 4. Factor 2 can be interpreted as early SOA product,

while factor 3 represents the later SOA product. Figure 5.1b includes the mass spectra input

(factor 1) and output from the ME-2 model runs.

5.5 Supplementary material 55

Figure S 5.1. a) Time series of the concentration of the three factors (factor 1: fixed blank B3 spectrum, factor 2+3: two free spectra) and b) corresponding normalized mass spectra.

56 5 The link between organic aerosol mass loading and degree of oxygenation

Figure S 5.2. Van Krevelen diagram (H : C vs. O : C ratio) of the nine experiments together with the observed range of ambient SOA represented by dashed lines (Ng et al., 2011a) and blank experiment B3. Data points represent 30 min averages, except for Expt. 4 (60-min average). Arrows indicate the time when mass spectra for ME-2 model runs were taken.

Figure S 5.3. Comparison of OH exposures derived from α-pinene decay and butanol-d9 decay for experiments where butanol-d9 was above detection limit throughout the entire experiment. The dashed line represents the 1:1 line.

5.5 Supplementary material 57

Retrieval of the OH exposure of experiment 1 and 2 from a repeat experiment

As the decay of α-pinene in the beginning of experiment 1 and 2 was very rapid, using the α-

pinene method including the extrapolation to the whole experiment time leads to a possibly

strong overestimation of the OH exposure. For this reason a repeat experiment was conducted

which showed the same characteristics in α-pinene decay, but with the OH tracer butanol-d9

present for the whole experiment time (See Figure S 5.4). The repeat experiment resembles

strongly experiment 2, which has the same initial α-pinene concentration (14 ppbv). During

experiment 1 (with an initial α-pinene concentration of 7 ppbv), the reactant decays within the

same time. This lower initial α-pinene concentration is also the reason for the lower O3

production. The replaced OH exposures derived from the α-pinene decay for experiments 1

(black line) and 2 (turquoise line) are shown in the lower panel together with the OH exposure

of the repeat experiment (purple line).

Figure S 5.4. The α-pinene and O3 concentrations of experiment 1, 2 and the repeat experiment are shown as a function of light exposure time (upper panel). The lower panel shows the OH concentration and exposure retrieved from the decay of the tracer butanol-d9, present during the repeat experiment as well as the OH exposures derived from the α-pinene method, which were replaced by the repeat experiment for analysis.

58 5 The link between organic aerosol mass loading and degree of oxygenation

Figure S 5.5. OH exposures for the nine different experiments (color code) derived from the decay of α-pinene, butanol-d9 or a combination of both. The OH exposure of experiment 1 and 2 was derived from a repeat experiment.

Figure S 5.6. The measured organic mass concentration (green) was fitted exponentially (black) for the last three hours of experiment 5 where wall loss dominates over organic mass production. This procedure results in a lower limit of the wall loss corrected organic mass concentration (purple).

5.5 Supplementary material 59

Figure S 5.7. Thirty-minute averages (except expt. 4: 60-min) of O : C ratio vs. organic mass fraction f44 (a) and of hydrogen-to-carbon ratio vs. organic mass fraction f43 (b) color coded for the nine different experiments; 2-min data is represented by the grey dots. The linear regressions are compared to the fit of Aiken et al. (2008) and to the linear and polynomial fit in Ng et al. (2011a).

60 5 The link between organic aerosol mass loading and degree of oxygenation

Figure S 5.8. O : C ratio as a function of OH exposure for the nine smog chamber experiments. The data was fitted with a line for the period when aging dominates, i.e. after the peak of suspended organic mass is reached. The slopes of ∆O : C/∆(OH exposure) are shown in Table S 5.4. An OH exposure between 3 and 25 × 107 cm-3 h is required to increase O : C by 0.05.

Figure S 5.9. Squares of the Pearson correlation coefficients, R², of measured mass spectra in comparison with LV-OOA (filled symbols) and SV-OOA (empty symbols) reference spectra (Ng et al., 2011b) as a function of the organic mass concentration (wlc). The correlation was performed on 30-min averaged MS at specific OH exposures ( ± 15 min) indicated by the color code. The corresponding mass spectra are presented in Figure S 5.10.

5.5 Supplementary material 61

62 5 The link between organic aerosol mass loading and degree of oxygenation

5.5 Supplementary material 63

64 5 The link between organic aerosol mass loading and degree of oxygenation

5.5 Supplementary material 65

Figure S 5.10. Thirty-minute-averaged organic mass spectra of the nine experiments at OH exposures (if reached) of 2·, 4·, 6·, 8·, and 11 × 107 cm-3 h ( ± 15 min), together with reference LV-OOA and SV-OOA spectra from Ng et al. (2011b). The reference spectra were converted to the fragmentation table of Aiken et al. (2008) and normalized. Correlation tests of each spectrum with both reference spectra were performed, while m/z's directly proportional to m/z 44 and m/z's present in only one, the measured or reference spectrum, were excluded.

___________________________________________________________________________

6 Higher relative humidity and VOC/NOx

increase α-pinene secondary organic aerosol

yields

___________________________________________________________________________

L. Pfaffenberger1, I. El-Haddad1, J. Dommen1, C. Marcolli2, P. Barmet1,+, C. Frege1, S.

M. Platt1, E. A. Bruns1, M. Krapf1, J. G. Slowik1, R. Wolf1, A. S. H. Prévôt1 and U.

Baltensperger1

1 Laboratory of Atmospheric Chemistry, Paul Scherrer Institute, 5232 Villigen, Switzerland

2 Institute for Atmospheric and Climate Science, ETH Zurich, 8092 Zurich, Switzerland

+ Now at: (Department Construction, Traffic and Environment, Canton of Aargau, 5001 Aarau, Switzerland)

in preparation for Atmospheric Chemistry and Physics

68 6 The link between organic aerosol mass loading and degree of oxygenation

Abstract

Secondary organic aerosol (SOA) yields from the photooxidation of the volatile organic

compound (VOC) α-pinene were investigated in smog chamber (SC) experiments at low

(~25 %) and high (~60 %) relative humidity (RH), various NOx/VOC ratios (0.04–3.8) and

with different seed chemical compositions (sulfate-containing or hydrophobic organic). A

combination of scanning mobility particle sizer and Aerodyne high resolution time-of-flight

aerosol mass spectrometer data was used to determine SOA mass concentration and chemical

composition. High RH increased yields (SOA formed-to-precursor reacted) up to six times

(1.5–6.4) compared to low RH, with greater increases for the most hygroscopic/acidic seeds.

The yields at low NOx/VOC (< 0.1) ratios were in general higher compared to yields at high

NOx/VOC (1.2–3.8) ratios. This NOx dependence follows the same trend as seen in previous

studies for α-pinene SOA.

Using Van Krevelen diagrams (plotting H : C vs. O : C), we show that the NOx/VOC ratio has

a strong effect on the aerosol chemical signature. For the same equivalent OH (hydroxyl

radical) exposure, experiments at low NOx/VOC ratios formed products with higher H : C

while experiments with higher NOx/VOC ratios showed lower H : C. This is expected from

higher formation yields of hydroxyl functional groups and peroxides at low NOx, compared to

organonitrates and carbonyls at high NOx.

While previous studies suggest a major effect of the relative humidity on particle chemical

composition via e. g. oligomerisation, we conclude, based on size-resolved data in this study,

that most of the α-pinene reaction products are not absorbed by the chosen inorganic and

organic seeds, rather, the seeds provide a surface for condensation. Our results are consistent

with literature observation indicating that there is no inorganic-organic mixing at different

RH. This study clearly presents varying yields dependent on the experimental conditions, for

the same α-pinene concentration reacted. We present wall-loss-corrected yields as a function

of the absorptive mass concentration consisting of organics and the associated liquid water

contents. We parameterized the volatility distributions for the use in modelling studies. It

might occur in the ambient atmosphere, that SOA loadings are sensitive to slight changes in

the absorptive mass concentration. To accurately predict SOA yields, the NOx/VOC ratio and

relative humidity have to be considered in models.

6.1 Introduction 69

6.1 Introduction

Organic aerosol (OA) accounts for 20 to 90 % of the submicron ambient aerosol (Jimenez et

al., 2009 and references therein), a great part of which is secondary organic aerosol (SOA)

formed via the condensation of low volatile oxidation products of gas-phase precursors.

Several direct (e.g. radiocarbon dating) and indirect observations underline the key role of

biogenic volatile organic compounds (VOC) in SOA formation (El Haddad et al., 2013 and

references therein). Current state-of-the-art models are unable to predict the burden of

biogenic SOA, especially in urban atmospheres (Hoyle et al., 2011), highlighting a

fundamental deficit in our knowledge of the chemical pathways and phase partitioning

mechanisms by which SOA accumulates and evolves in the atmosphere.

The ensemble of gaseous and particulate phase species involved in SOA formation is

immensely complex. The compounds relevant for SOA formation are often a minor gas phase

fraction, resulting in yields (SOA formed-to-precursor reacted) of only a few percent. Their

chemical composition and volatility distribution strongly depend on the oxidation conditions,

most notably on the fate of organic peroxy radicals (RO2), which either react with nitrogen

oxides (NOx) or other peroxy radicals (RO2 and HO2). The influence of NOx on the oxidation

mechanisms and SOA formation is commonly described by the NOx/VOC cycle and has been

under close scrutiny lately. For most light precursors, such as isoprene and monoterpenes

(including α-pinene), SOA yields appear to be strongly influenced by the NOx/VOC ratio,

with a general enhancement observed under low NOx conditions (Ng et al., 2007; Presto et al.,

2005). Another factor influencing SOA yields is the dynamic nature of the semi-volatile

species involved in SOA formation. These are subject to ongoing chemical degradation,

which may lead to compounds of either lower or greater volatility (Kroll and Seinfeld, 2008).

As a consequence, SOA yields and degrees of oxygenation (described by the atomic oxygen

to carbon ratio O : C) may not be constant, depending on the extent to which these species

were exposed to oxidants.

SOA yields are generally described by the absorptive equilibrium partitioning of condensable

species to a well-mixed liquid phase (Odum et al., 1996), which depends upon the chemical

species’ saturation vapour pressures (classically two model products are considered) and their

liquid-phase activities (modified Raoult’s law). Donahue and co-workers have proposed the

use of a “volatility basis set” (VBS) for a better representation of the wide range of OA in the

atmosphere and the ongoing oxidation of semi-volatile organics (Donahue et al., 2011; 2012

and references therein). However, several difficulties remain in estimating or measuring the

70 6 The link between organic aerosol mass loading and degree of oxygenation

saturation vapor pressures, molar activity coefficients and the mean molecular weight of the

condensing species (e.g. Clegg et al., 2008b, a).

Difficulties increase when considering the role of relative humidity (RH) and of inorganic

particles on organic partitioning (Zuend and Seinfeld, 2013 and references therein). In

addition to its potential influence on the kinetics of gas-phase reactions (e.g. the increased

formation rate of hydroxyl radicals (OH) with increasing H2O), from a thermodynamics point

of view, water interacts with SOA components by altering the water content of aerosol

particles (also at sub saturation conditions) and hence the equilibrium concentration of water

soluble organic compounds. An increase in aerosol-phase organics can be achieved by 1)

increasing the absorptive particulate mass, by 2) decreasing the average molecular weight of

condensed species or by 3) decreasing the activity coefficients of organic species (Pankow,

1994). Therefore, it is expected that an increase in particulate water content would in principle

enhance SOA yields of water-miscible species. Model calculations predict a pronounced

effect at low organic mass loadings (Pankow, 2010). An RH dependent yield has been

experimentally reported for a number of precursors including limonene, α-pinene and ∆3-

carene (Jonsson et al., 2006). Jonsson et al. (2006 and references therein) include previous

results from ozonolysis of the three aforementioned precursors. A substantial effect of RH on

SOA yields was observed only for studies at precursor concentration <1000 ppb (effect seen

with 124 ppb) and with large variation in RH (0.01 % & 31 % RH (Bonn et al., 2002); < 2–

58 % RH (Cocker et al., 2001); < 2–85 % RH (Jonsson et al., 2006)).

This RH effect is potentially enhanced by an acid catalysed reactive uptake of SOA products

onto the wet surface of particles (Kroll and Seinfeld, 2008; Kleindienst et al., 2006; Jang et

al., 2002; Iinuma et al., 2007). While this has been shown to play a key role for particular

oxygenated VOCs, e.g. glyoxal (Clegg and Seinfeld, 2006), SOA from a number of

precursors was found to be marginally affected by aqueous phase reactions (e.g. Ng et al.,

2007). However, the impact of RH on organic partitioning can be severely suppressed by

considerable deviations from ideal mixing between the condensing organic species and the

prevailing condensed phase. A growing number of studies shows that internally mixed

organic/ inorganic/ water aerosol particles may undergo a salting out effect, developing two

stable liquid phases: an aqueous electrolyte solution and an organic solution. The miscibility

of an organic analyte in an aqueous solution of electrolytes depends on numerous factors

including temperature, relative humidity, organic compound polarity commonly predicted by

its O : C, species relative contributions in bulk particulate matter and the chemical nature of

the electrolytes (Zuend and Seinfeld, 2013; You et al., 2013). For example, phase separation

6.2 Methods 71

was observed to always occur for organic compounds with an O : C < 0.5 and at low relative

humidity (You et al., 2013), and is especially pronounced in the case of ammonium sulfate

(compared to ammonium hydrogen sulfate and nitrate). Neglecting phase separation of

organic compounds and the effect of RH thereon bears the potential for invalid yield

predictions.

In this study, we investigate through systematic smog chamber experiments SOA formation

from the oxidation of α-pinene, a major biogenic precursor. The aim of this study is to

examine the impact of NOx/VOC ratios and particulate water content on SOA chemical

composition and yields. This is performed by varying relative humidity, NOx/α-pinene ratios

and aerosol seed composition (hygroscopicity, acidity). SOA yields and volatility

distributions reported here may aid the parameterization of the NOx and particulate water

dependence of α-pinene SOA for the use in atmospheric models.

6.2 Methods

6.2.1 Experimental setup and instrumentation

20 experiments (see Table 6.1) were carried out in the Paul Scherrer Institute (PSI) smog

chamber (SC): a 27 m³ Teflon bag suspended in a temperature-controlled wooden housing

(Paulsen et al., 2005). Photochemistry was initiated by four xenon arc lamps (4 kW rated

power, 1.55×105 lumens each, XBO 4000 W/HS, OSRAM), facing parallel to the SC bag,

emitting a light spectrum similar to the solar spectrum, and 80 black lights (Philips, Cleo

performance 100 W) to accelerate the aging process, located underneath the SC bag, with

emission between 300–400 nm wavelength (light characterisation in Platt et al. (2013)). A

reflecting aluminum foil surrounds the SC bag to maintain light intensity and light diffusion.

Various parameters were monitored in the SC. The temperature (T) and RH measurement was

optimized by passing SC air through a radiation shielded sensor. One of two different high

resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS, Aerodyne Research, Inc.,

Billerica, MA, USA) was operated during three different campaigns to measure online size-

resolved chemical composition (organics, ammonium, nitrate, sulfate, chloride) of non-

refractory particles (DeCarlo et al., 2006). The HR-ToF-AMS were equipped with two

different PM2.5 lenses (Williams et al., 2013) to sample particles up to large diameters above

1 µm. The sampled aerosol was dried (~10 % RH) before entering the HR-ToF-AMS. A

72 6 The link between organic aerosol mass loading and degree of oxygenation

supporting flow of ~1.5 L min-1 was maintained parallel to the HR-ToF-AMS to minimize

diffusive losses in the sampling lines. The HR-ToF-AMS data was corrected for gas-phase

contributions.

Two scanning mobility particle sizers (SMPS) were additionally deployed for the

measurement of the aerosol size distributions. The first SMPS (condensation particle counter,

CPC 3022 (TSI), and custom built differential mobility analyser, DMA: extended length Leff =

0.93 cm, dm max=1000 nm, recirculating sheath flow) was connected to the HR-ToF-AMS

sampling line to analyze the dried particles. A second SMPS (SMPSwet, TSI: CPC 3022A,

DMA classifier 3081, recirculating sheath flow) and a CPC (TSI: CPC 3025A) measured the

wet particle number size distribution and total number concentration (d > 3 nm), respectively.

Gas-phase compounds with a higher proton affinity than water (166.5 kcal mol-1) were

measured with a proton transfer reaction mass spectrometer (PTR-MS, Ionicon). The PTR-

MS was calibrated before each experiment for α-pinene detected at m/z 137 and m/z 81; the

accuracy of these measurements was estimated to be ~5 %, based on the purity indicated on

the calibration gas cylinder.

A chemiluminescence-based NOx instrument (Monitor Labs 9841A NOx analyzer) was

attached to the HONO source to monitor the injected concentration throughout the

experiment. A modified NOx instrument including a blue-light photolytic NO2-to-NO

converter (Thermo Environmental Instruments 42C trace level NOx analyzer) and two ozone

monitors (Monitor Labs 8810 ozone analyzer, Environics S300 ozone analyzer) monitored

NOx and O3 in the SC.

6.2.2 Smog chamber operation and aerosol seeding

SOA formation and growth from the precursor α-pinene was induced by the following SC

operation sequence: (1) humidification of the chamber, (2) addition of seed aerosol, (3)

introduction of VOCs, (4) addition of nitrous acid (HONO), (5) addition of nitrogen oxides

(equal amounts of NO + NO2) if applicable, (6) an equilibration period (30–45 min), (7)

switching on of xenon and black lights to generate OH radicals, and (8) a reaction time of 5 to

20 h (corresponding to 0–2 × 108 cm-3 h OH exposure, see Sect. 6.2.4). Experimental

conditions for each individual experiment are summarized in Table 6.1. Before each

experiment, the SC was cleaned by the injection of several ppmv of ozone for 5 h and

irradiation with black lights for 10 h at 20 °C, followed by a flushing period with pure air and

6.2 Methods 73

high relative humidity (~60 %) at ~30 °C for at least 20 h. To ensure that organic matter (OM)

formed during the experiments is not significantly influenced by background contamination in

the SC, three blank experiments were carried out (high RH, seed aerosol, lights switched on,

but without addition of α-pinene). The organic mass concentration formed was substantially

lower (< 0.1 up to 2.8 µg m-3) than during comparable experiments (similar NOx/VOC and

RH).

In the chamber, the temperature varied within a range of 21 °C to 26 °C between experiments.

Due to heat from the xenon and black lights, the temperature increased, stabilizing only ~1 h

after experiment start. The increase in temperature of 1–4 °C led to an absolute decrease in

RH of ~2-20 %, thus the range given in Table 1, 23–75 % RH, is representative for the

temperature-stable period afterwards. Liquid-phase α-pinene (98 %, Aldrich) and an OH

tracer (9-times deuterated butanol, 98 %, D9, Cambridge Isotope Laboratories), hereafter

referred to as butanol-d9 (1 µL injected ≈ 10 ppb in the SC), were sequentially injected into

an evaporation glass bulb heated to 80 °C. The two VOCs were carried into the bag by a

dilution and flush flow (each 15 L min-1, maintained for 15 min) from an air purifier (737-250

series, AADCO Instruments, Inc., USA), further referred to as “pure air”. Initial α-pinene

concentrations were 16.1–31.7 ppbv.

HONO was used as a source of both NO and OH, produced in a reaction vessel by continuous

mixing of the reagents sodium nitrite (NaNO2, 1 mmol L-1 in milli-Q H2O) and three different

concentrations of sulfuric acid solutions (H2SO4, 1 mmol L-1 (expt. 1-6), 10 mmol L-1 (expt. 7,

8, 11, 13) and 100 mmol L-1 (expt. 9, 10, 12, 14) in milli-Q H2O) (Taira and Kanda, 1990).

The HONO product was carried by 2.5 ± 0.2 L min-1 pure air flow into the SC and particulate

H2SO4 was minimised by a Teflon filter applied between HONO source and SC. 2 ppbv

(± 10 %) of HONO were injected before lights on to initiate photochemistry and the injection

was continued throughout all experiments. In addition, equal concentrations of NO (99.8 %;

1005 ppmv ± 2 %) and NO2 (purity: 98 %; 1005 ppmv ± 3 %), resulting in 19.6–75.1 ppbv

initial NOx, were added during experiments with NOx/α-pinene > 1. Within the results section,

the two terms “low NOx” and “high NOx” refer to the following conditions:

- low NOx = NOx/α-pinene < 0.1: HONO injected continuously (indicated by the

asterisks (*) in Table 6.1 and figures).

- high NOx = NOx/α-pinene > 1: Initial injection of NO + NO2 with continuous HONO

injection

74 6 The link between organic aerosol mass loading and degree of oxygenation

For high NOx experiments the initial NOx concentration, which decays with time to lower

values, and for low NOx experiments the mean NOx concentration is given in Table 6.1.

During 14 experiments (no. 1–14) and two blank experiments, a solution (1 g L-1) of

ammonium hydrogen sulfate (NH4HSO4, Aldrich) in ultrapure milli-Q water was nebulized

(0.6 L min−1) and introduced into the SC with a pure air dilution flow of 10 L min−1 to act as

seed particles. To keep the seed aerosol as far as possible in a liquid state, no drier was used

behind the nebulizer. We determine the acidity of the seed particles, here defined by the

NH4/SO4 ratio measured by the HR-ToF-AMS. A NH4/SO4 ratio between 1 and 2 indicates a

neutral seed composition, a mixture of NH4HSO4, (NH4)3(SO4)2 (letovicite) and (NH4)2SO4.

For simplicity, we replaced (NH4)3(SO4)2 by an equal mix of NH4HSO4 and (NH4)2SO4 for the

assumption of density and growth factors. By contrast, a NH4/SO4 ≤ 1 indicates an acidic

seed, including a mixture of NH4HSO4 and H2SO4.

During experiments 7–14, the nebulized NH4HSO4 solution was partly neutralized to

(NH4)2SO4 presumably by background NH3 released from the walls, as indicated by the molar

NH4/SO4 ratio (Figure S 6.3, Supplement). During experiments 1–6 the seed was composed of

a mixture of NH4HSO4 and H2SO4, due to a significant addition of gas-phase concentration of

H2SO4 via the HONO injection, suppressing the seed neutralization by NH3.

Additionally, three α-pinene experiments (no. 15–17) and one blank experiment were

conducted using an inert hydrophobic fluorinated hydrocarbon as seed (further referred to as

CF-seed; CF3CF2CF2O-[CF(CF3)CF2-O-]nCF2CF3; Krytox® 1525). These experiments were

conducted to isolate the humidity effect on α-pinene SOA from possible water-phase reactions

happening during the hydrophilic seed experiments in an inorganic-seed-SOA-mixture. The

CF-seed was generated via evaporation of the fluorinated hydrocarbon in the liquid phase at a

temperature of 125–145 °C and subsequent homogeneous nucleation in a pure air flow of

2.4 ± 0.1 L min-1. The CF-seed concentration was 6.7–10.0 µg m-3 when lights were switched

on, and decayed to values below detection limit of the HR-ToF-AMS (0.1 µg m-3), after α-

pinene SOA condensation. The CF-seed mass spectrum in the HR-ToF-AMS is clearly

distinct from α-pinene SOA, with main contributions at m/z 69 (CF3), m/z 169 (C3F7) and

m/z 119 (C2F5) (Figure S 6.4 and Table S 6.1, Supplement). Table 6.1 lists the seed

composition and estimated physical state of each experiment dependent on RH (acidic =

ammonium hydrogen sulfate (NH4HSO4) & sulfuric acid (H2SO4); neutral = ammonium

sulfate ((NH4)2SO4) & ammonium hydrogen sulfate (NH4HSO4); CF = fluorinated carbon).

6.2 Methods 75

There is no measurable efflorescence for NH4HSO4; efflorescence of (NH4)2SO4 happens at

35 % RH (Martin, 2000).

6.2.3 Estimation of the hygroscopic growth factors and liquid water content

The absolute liquid water content (LWC) of the aerosol particles was derived for the bulk

aerosol mass and the size resolved mass distribution based on literature growth factors, the

measured RH and chemical composition. The growth factor GF (RH) of a particle is defined

as the ratio of wet diameter at a given RH to the dry diameter: ⁄ .

Inorganic GFs were taken from the Aerosol Diameter Dependent Equilibrium Model

(ADDEM, Topping et al., 2005) for diameters of 360 nm. GFs of organics were derived using

the relationship between the hygroscopicity parameter κ and the degree of oxygenation

[κ = 0.29×(O : C)] from Chang et al. (2010), well representing the hygroscopicity of α-pinene

SOA measured by Massoli et al. (2010). The measured degree of oxygenation at an OH

exposure of (2.0 ± 0.5) ×107 cm-3 h was used to derive κ, which in turn was converted to GF,

assuming a negligible curvature (Kelvin) effect (Kreidenweis et al., 2005):

1 κ %

%

(6.1)

The mixed GFs for aerosol containing inorganic and organic species were determined using

as a first approximation the Zdanovskii-Stokes-Robinson (ZSR) volume mixing rule (Stokes

and Robinson, 1966):

∑ ∙ (6.2)

Where εi and GFi(RH) denote the volume fraction and GF (RH) of species i, respectively. The

H2O volume ( was calculated using the definition of GF (RH) and the dry volume (Vdry):

∙ 1 (6.3)

multiplied by the density of water (1 g cm-3) results in the LWC.

The dry (Sdry) and wet (Swet) surfaces were calculated with Eq. (6.4) and Eq. (6.5),

respectively:

6 ∙ / (6.4)

6 ∙ / (6.5)

76 6 The link between organic aerosol mass loading and degree of oxygenation

The LWC and surface distributions were derived using size resolved pToF (particle time-of-

flight) data. Due to the low pToF signal of NH4, its surface distribution was estimated based

on SO4 pToF measurements.

6.2.4 Determination of OH exposure and extent of α-pinene ozonolysis

The gas-phase composition, the resulting OH concentration and thus the photochemical age of

a reaction system may considerably vary between experiments. Furthermore, within a single

experiment the photochemical age is not necessarily directly proportional to the light

exposure time. Therefore, we here discuss reaction time in terms of OH exposure (cm-3 h),

defined as the OH concentration integrated over time. OH exposures were derived based on

the decay of the OH tracer butanol-d9, detected by the PTR-MS as M+H+-H2O at m/z 66,

following the methodology introduced by Barmet et al. (2012):

OHexposure,

∙ ∆

∆dt (6.6)

Where butanol-d9 and kOH, butanol-d9 (= 3.4 × 10-12 cm3 molecules-1 s-1) are the butanol-d9

concentration and its reaction rate with OH, respectively. t is the time after lights on, V the SC

volume (we assume as a first approximation a constant chamber volume), and fdil the dilution

flow due to HONO input (Sect. 6.2.2).

The percentage of α-pinene reacted with OH was derived based on Eq. (6.7):

, ∙ , ∙ (6.7)

Where [α-pin], [O3] and [OH] denote the concentrations of α-pinene, O3 and OH, and kOH, α-pin

(= 5.3 × 10-11 cm3 molecules-1 s-1) and kO3, α-pin (= 8.9 × 10-17 cm3 molecules-1 s-1) the reaction

constants of α-pinene with OH and O3, respectively. The production of O3 was faster under

high NOx compared to low NOx, due to a more efficient VOC-NOx catalytic cycle. Therefore,

the lowest percentage (~65 %) of α-pinene reacted with OH was achieved during experiment

no. 9, under high NOx, while for all other experiments 67.9–87.8 % of α-pinene reacted with

OH (Table 6.1). We also expect that the further processing of the first generation products

formed via α-pinene reaction with OH or O3 – that don’t contain C=C bonds – to

6.2 Methods 77

predominantly proceed through reactions with OH. Accordingly, we conclude that SOA

compounds detected under these conditions are mainly from OH chemistry.

Table 6.1. Overview of experimental conditions. Seed types {acidic = ammonium hydrogen sulfate (NH4HSO4) & sulfuric acid (H2SO4); neutral = ammonium sulfate ((NH4)2SO4) & ammonium hydrogen sulfate (NH4HSO4); CF = fluorinated carbon (see Sect.6.2.2)} and their estimated phase states {(L) = liquid; (-) = liquid and/or solid}. Initial seed mass concentrations; relative humidity (RH); measured mean (*) NOx concentrations during low NOx experiments and measured initial NOx concentrations during high NOx experiments; measured initial α-pinene concentrations (which reacted before an OH exposure of (2.0 ± 0.5) ×107 cm-3 h) and their estimated fractions reacted with OH radicals in %, rest reacted with O3). NOx/α-pinene ratios; wall-loss corrected organic mass concentrations and corresponding yields averaged over the OH exposure of (2 ± 0.5) ×107 cm-3 h. Standard deviations (1sd) given in brackets are measurement variability. Horizontal lines separate experiments with different i) seed composition, ii) RH, iii) NOx/α-pinene. Blank experiments (B1, B2 and B3) are listed at the very bottom.

No seed RH NOx α-pin α-pin NOx/ org (wlc) yield (wlc)

type (phase) initial initial or initial = decay α-pin at OH exp.:

mean(*) reacted by OH 2(±0.5) × 107 cm-3 h

µg m-3 % ppbv ppbv % µg m-3 %

1 acidic (L) 8.0(0.5) 69(2) 44.4(0.8) 20.7 80.0 2.1 13.4(0.2) 11.5

2 acidic (L) 12.3(0.5) 67(2) 70.4(1.3) 18.7 80.3 3.8 8.6(0.1) 8.1

3 acidic (L) 4.9(0.3) 66(2) 19.6(0.7) 16.1 81.6 1.2 12.6(0.6) 13.8

4 acidic (L) 4.7(0.2) 29(1) 23.6(0.6) 19.9 81.3 1.2 3.9(0.0) 3.5

5 acidic (L) 8.0(0.3) 28(1) 52.1(0.6) 20.3 74.6 2.6 2.1(0.1) 1.8

6 acidic (L) 5.2(0.3) 27(1) 1.3(0.4)* 18.3 87.8 0.071 12.0(0.5) 11.6

7 neutral (L) 8.8(0.4) 67(1) 0.7(0.2)* 20 81.9 0.037 16.2(0.4) 14.3

8 neutral (L) 4.3(0.6) 60(1) 1.0(0.2)* 18.7 86.5 0.052 12.3(0.4) 11.6

9 neutral (L) 5.5(0.2) 56(2) 65.8(0.8) 30.9 65.0 2.1 11.1(0.1) 6.4

10 neutral (L) 4.4(0.2) 50(1) 1.3(0.2)* 30.6 79.2 0.041 29.6(1.1) 17.1

11 neutral (-) 4.1(0.2) 26(1) 0.7(0.3)* 18.9 81.3 0.039 5.5(0.2) 5.1

12 neutral (-) 3.6(0.1) 26(1) 1.9(0.4)* 30.5 78.4 0.062 20.3(1.3) 11.8

13 neutral (-) 8.2(0.3) 25(1) 1.1(0.3)* 19.6 87.4 0.055 5.3(0.1) 4.8

14 neutral (-) 3.2(0.2) 23(1) 75.1(0.7) 27.8 69.4 2.7 2.4(0.1) 1.5

15 CF (L) 7.1(0.3) 58(1) 56.2(0.7) 31.7 67.9 1.8 9.8(0.1) 5.4

16 CF (L) 10.0(0.5) 58(2) 58.3(0.5) 31.3 73.9 1.9 10.7(0.1) 6.1

17 CF (L) 6.7(0.1) 26(1) 53.3(0.6) 30.5 69.7 1.7 4.7(0.1) 2.1

B1 CF (L) 0.3(0.1) 58(2) 53.0(0.6) - - - < 0.1 -

B2 neutral (L) 4.5(0.7) 68(2) 0.9(0.5)* - - - 2.8(1.0) -

B3 acidic (L) 3.8(0.2) 75(3) 1.4(0.3)* - - - 0.5(0.1) -

78 6 The link between organic aerosol mass loading and degree of oxygenation

6.2.5 Determination of suspended and wall-loss-corrected organic mass and yield

Suspended OA mass. Determining accurate measures of the total organic aerosol

concentration, [OA], is crucial for obtaining accurate aerosol mass yields, Y. Nevertheless, we

note that our main objective is determining the effects of changes in the environmental

conditions (i.e. NOx concentration, RH and seed composition) on SOA yields. Therefore,

systematic biases in the yield determination, thoroughly discussed below, may marginally

influence our conclusions. The suspended organic mass concentration, was derived by

utilizing the chemical composition measurements from the HR-ToF-AMS scaled to the total

volume measured by the SMPS, using compound-specific densities (ρorg = 1.4 g/cm³, ρNH4HSO4

= 1.79 g/cm³, ρ(NH4)2SO4 = 1.77 g/cm³, ρH2SO4 = 1.83 g/cm³).

For some experiments (9, 10, 12 and 14-17), the organic mass concentrations determined by

the HR-ToF-AMS were corrected for a below unity transmission efficiency at the lower edge

cut-off of one of the two PM2.5 lenses employed (Figure S 6.1 and Figure S 6.2, Supplement).

Additionally, for organic mass calculation, we have considered that the measured NO3 signals

are entirely related to organonitrates, rather than NH4NO3. This assumption mainly stems

from (1) simultaneous increases of the NO3 signal in the same particle size region as OA (see

pToF-data, Sect. 6.3.3), while inorganic nitrate would be expected to mix within the

electrolyte rich aerosol at larger diameters and (2) the occurrence of NO3 under acidic

conditions, thermodynamically unfavorable for NH4NO3. This is also supported by the higher

NO+/NO2+ ratios measured in the SC compared to ratios recorded during NH4NO3

nebulization (on average 1–2.8 times higher, Supplement Figure S 6.10), typically expected

from organonitrates (Farmer et al., 2010). We cannot exclude the presence of NH4NO3, which

would increase the calculated LWC by only 1–13 % and decrease the calculated OA mass by

only 2–7 %, assuming all detected nitrate to be NH4NO3. Finally, for the yield calculations,

we consider that the accuracy of the aerosol phase measurements to be around 30 %.

Wall-loss-corrected-OA mass. To correct the total OA concentration for losses of particles

and vapours to the chamber walls, we use Eq. (6.8), introduced by Hildebrandt et al. (2011),

based on the mass balances of the suspended organic aerosol mass, , and the mass of the

organic aerosols on the walls, (summed up to derive [OA]):

ω ∙ ∙ (6.8)

Here, represents the loss rate of organic particles to the walls, derived by fitting

between 5–8 h after lights were switched on, when SOA production is expected to be

6.2 Methods 79

negligible (α-pinene concentration < 1 ppbv, Figure S 6.6 and Table S 5.2, Supplement). We

determine an average loss rate of 0.13 h-1, corresponding to a particle half-life of 5.3 h. This is

consistent with theoretical values calculated for the loss of 100-300 nm particles by

diffusional deposition and gravitational settling under our conditions. This average was

used in the case of insufficient data point statistics to perform accurate fitting. In Eq. (6.8), ω

is a dimensionless proportionality coefficient between the mass of organic vapors that

partition into the wall-deposited particles and the mass of organic vapors that partition into the

suspended particles, with ω ranging between 0 and 1. Here, for the base case calculations, we

neglect the condensation of organic vapors onto the wall-deposited particles (i.e. ω 0),

consistent with previous studies of α-pinene SOA production (Hildebrandt et al., 2011 and

references therein). This assumption gives a lowest estimate of SOA yields. Considering the

second limiting case: ω 1, which characterizes an equal partitioning of organic vapors

between the wall-deposited and suspended particles, would increase the determined [OA] and

Y.

We note that Eq. (6.8) does not take the loss of vapors onto the Teflon walls of the chamber

into consideration, which was recently reported to depend on the wall-to-seed surface ratio

and to greatly suppress SOA yields from laboratory chambers under certain conditions (Zhang

et al., 2014). We are convinced that gas-wall partitioning does not significantly influence the

interpretation of our results, focused on relative differences in the yields determined under

different conditions. This is because: (1) this effect was found to be minor for the α-pinene

SOA system (2) SOA mass formation rapidly evolved after lights were switched on and (3)

we have maintained a relatively constant wall-to-seed surface ratio for all experiments, crucial

for vapor-wall-losses; hence losses of organic vapors may lead to a systematic negative bias

in the determined yields, but do not influence the comparison between the experiments.

SOA yield determination and parameterization. SOA mass yields, Y, are defined in Eq. (6.9)

as the organic mass concentration formed, [OA], per precursor mass consumed, ΔCα-pinene:

(6.9)

For comparison purposes, Y are reported in Figure 6.2 and (Table 6.1) at an OH exposure of

(2.0 ± 0.5) ×107 cm-3 h, reached during all experiments. The parameterization of SC SOA

yields measured is based on the absorptive partitioning theory of Pankow (Eq. (6.10),

(Pankow, 1994)), which expresses the production of a set of semi-volatile surrogate products

(total number N) as a function of the mass yield of these products, αi, and their partitioning

80 6 The link between organic aerosol mass loading and degree of oxygenation

coefficients, ξi (a dimensionless quantity reflecting the condensed-phase mass fraction of

these products). The critical parameters driving the partitioning of these products are their

effective saturation concentration, Ci*, and the total concentration of the absorptive organic

phase, COA. As discussed below, we consider the absorptive organic mass as the sum of the

total OA concentration including organonitrates, [OA], and the liquid water content in this

phase (Sect. 6.3.3).

∑ ∑ 1∗

(6.10)

In Eq. (6.10), Ci* is a semi-empirical property (inverse of the Pankow-partitioning coefficient,

KP) reflecting the vapor pressure of the pure constituents, ,° , and their interaction with the

organic mixture (effectively including liquid phase activities, γi), as expressed in Eq. (6.11):

∗ ,°

(6.11)

Where Mi denotes the molecular weight of compound i, R the ideal gas constant and T the

temperature. Here, SC yields from single experiments are fitted as a function of COA using the

volatility basis set (VBS) (Donahue et al., 2006), which separates semi-volatile organics into

logarithmically spaced bins of effective saturation concentrations Ci*. In practice, a total

number of 4 bins spanning our measurement range were assumed: N = 4, Ci* = 0.1, 1, 10 and

100. To solve Eq. (6.10) for the parameters αi, we have introduced a novel approach using a

Monte Carlo simulation. This approach provides the best estimate of αi values representing

the measured yields, Y, from single experiments vs. the measured absorptive masses, COA. It

also offers a measure for the uncertainties related to the determination of the volatility

distributions (αi parameters) from SC experiments. The method proceeds as follows:

(1) A possible yield domain is inscribed against COA, based on our best estimates of the

measurement accuracy of SOA and α-pinene mass;

(2) A range of possible inputs is defined for the parameters αi;

(3) αi parameters are randomly generated over the range;

(4) Deterministic computations of Y vs. COA are performed using Eq. (6.10) and the generated

αi inputs.

(5) Yield curves that fall within the domain defined in step (1) are saved, aggregated and

presented as probability distribution functions (PDF).

6.3 Results 81

6.3 Results

6.3.1 SOA yield dependence on RH, NOx/α-pinene and aerosol seed composition

Figure 6.1 shows the suspended and wall-loss-corrected organic mass concentrations as a

function of OH exposure for experiments listed in Table 6.1. Two panels separate

experiments with ~20 ppbv (top panel) and ~30 ppbv (bottom panel) initial α-pinene

concentration. Asterisks indicate low NOx experiments and the same symbol shapes are used

for the same seed compositions. SOA mass is rapidly formed after lights were switched on

and the wall-loss-corrected mass reaches a certain plateau. This happens in general later for

experiments with higher SOA formed than for less SOA formed, at an approximate OH

exposure of 2 ×107 cm-3 h.

Figure 6.1. 20-min-averaged wall-loss-corrected (symbols & lines) and suspended (lines) organic mass concentrations as a function of OH exposure. Data is separated according to similar initial α-pinene concentration (20ppbv - top panel, 30ppbv - bottom panel). Seed composition is given in the legend and asterisks indicate low NOx experiments.

82 6 The link between organic aerosol mass loading and degree of oxygenation

In Figure 6.2, we examine the relationship between the yields at the same OH exposure of

(2.0 ± 0.5) ×107 cm-3 h and the prevailing experimental conditions: RH, NOx/α-pinene and

seed composition. Lines connect experiments with all parameters approximately constant

except RH (dashed) or NOx/α-pinene (solid). The NOx/VOC of experiment 5 lies between

expts. 1 and 2, the reason for the line split. We observe a clear influence of the RH on the

yields, increasing on average by 0.15 ± 0.06 % per 1 % RH, for the range investigated.

Despite the limited statistics, the extent of the RH effect was found to be highly dependent on

the seed chemical nature, with a greater influence for the acidic seed (0.22 ± 0.03 % per 1 %

RH, p < 0.001), the most hygroscopic aerosol, compared to the non-acidic seed

(0.15 ± 0.03 % per 1 % RH, p < 0.001) and the hydrophobic seed (0.09 ± 0.04 % per 1 % RH,

p = 0.05).

Figure 6.2. Average wall-loss-corrected yields at (2.0 ± 0.5) ×107 cm-3 h OH exposure as a function of RH. Symbol sizes represent α-pinene reacted, symbol colors represent NOx/α-pinene and symbol shapes represent seed composition according to Table 6.1. Experiments with similar NOx/α-pinene, seed composition and α-pinene reacted are connected with dashed lines showing the increase in yield for increased RH Experiments with similar RH, α-pinene reacted and seed composition are connected with solid lines showing the increased yield with decreasing NOx/α-pinene ratio.

6.3 Results 83

This indicates that the particulate water content plays a central role in the partitioning of SOA

compounds, either by altering the thermodynamic properties of the bulk phase (e.g. increasing

the absorptive mass or decreasing the compounds activity coefficients, non-reactive uptake)

or by providing a reactive sink for SOA semi-volatile species (e.g. formation of lower

volatility compounds/oligomers in the bulk phase, reactive uptake). These two processes will

be investigated below.

SOA yields are also significantly enhanced under low NOx conditions, in agreement with

literature data (Ng et al., 2007). The decrease in SOA yields against NOx/α-pinene ratio is

estimated at 3.3 ± 0.6 % (p < 0.001). Such decrease indicates that SOA compounds formed

under low NOx conditions are less volatile than those formed under high NOx conditions.

Overall, these results highlight the sensitivity of SOA yields to the prevailing oxidation

conditions and the particle bulk-phase composition. Therefore considering such conditions for

accurate predictions of the SOA burden in the atmosphere is needed.

6.3.2 SOA elemental composition

The effect of the experimental conditions on SOA chemical composition are investigated in

the Van Krevelen space in Figure 6.3a. The mean and standard deviation (1sd) of the

elemental O : C and H : C (atomic hydrogen to carbon ratio) at an OH exposure of

(2.0 ± 0.5) ×107 cm-3 h are shown by markers, while 1-h-averages for OH exposures >

2 ×106 cm-3 h are represented by solid lines. The overall region of H : C to O : C in the Van

Krevelen space is very comparable for all experimental conditions and to ambient data,

represented by the triangular shape (Ng et al., 2011a). Data of experiments 1-14, with

inorganic seed, follows a similar slope (-0.86, least orthogonal distance fit) with aging: An

increase in O : C during aerosol aging takes place under all conditions. Experiments 15-17

show lower H:C and O:C especially in the beginning due to some contribution of the

fluorinated hydrocarbon seed to these ratios. Experiments with very low organic mass loading

follow a similar slope as the other experiments, but with lower H : C and O : C.

84 6 The link between organic aerosol mass loading and degree of oxygenation

Figure 6.3. Van Krevelen diagrams: Mean (± 1 sd measurement variability) H : C versus O : C at OH exposure (2.0 ± 0.5) ×107 cm-3 h (symbols). Symbol colors indicate RH, symbol shapes the seed composition and the asterisk low NOx experiments. The triangular shape (solid black lines) represents the range of ambient SOA (Ng et al., 2011a). (a) Additionally included 1 h-averaged (lines) H : C versus O : C (for OH exposure > 2 × 106 cm-3 h) show similar region and trend for all conditions. The dashed line represents the (least orthogonal distance) fitted slope of expts. (1-14) with inorganic seed: -0.86. Expts. 15-17 show influence of CF seed compounds (data shown from suspended organic mass > 0.3 µg m-3). (b) Data separated by wall-loss-corrected organic mass concentrations to exclude concentration effects (left panel: 2-6 µg m-3; middle panel: 8-14 µg m-3; right panel: 16-30 µg m-3). The grey shaded areas include all low NOx experiments.

6.3 Results 85

In Figure 6.3b the data was separated according to the organic mass concentrations to isolate

the NOx and RH effects on the chemical composition from the possible influence of enhanced

partitioning of semi-volatile organic species to the aerosol phase due to a higher SOA loading

(Pfaffenberger et al., 2013). We observe that NOx levels have the highest influence on SOA

elemental composition. Generally, low NOx experiments (marked with asterisks) show a

higher H : C than high NOx experiments, indicated by the grey shaded areas at the upper end

of the figure.

This is even more pronounced at an early reaction stage (OH exposure < 2 × 106 cm-3 h, not

shown in Figure 6.3), where fresh SOA products formed during all low NOx experiments have

a higher H : C and lower O : C. Such influence of NOx levels on SOA elemental composition

is consistent with our general understanding of SOA chemistry: under low NOx, substantial

amounts of hydroperoxides would result in a higher H : C ratio than e.g. aldehydes/ketones

formed under high NOx conditions (Valorso et al., 2011). NOx levels also influence the

amount of organonitrate formed: considering that the entire nitrate signal arises from

organonitrates (i.e. a higher estimate of organonitrate contribution), we estimate molar ratios

of NO3 to carbon of ~1:30 for high NOx and ~1:100 for low NOx (Figure S 6.5, Supplement).

Assuming molecules with 10 carbon atoms, these ratios imply that every 3rd molecule

contains one NO3 functional group for high NOx vs. every 10th molecule contains one NO3

functional group for low NOx.

Conversely, we did not observe a significant effect of RH and seed acidity on SOA elemental

composition, despite of their significant influence on SOA yields. Furthermore, we could not

reveal any significant dependence between these parameters and the ratio of organic

fragments larger than m/z 150 to the total organic mass, a proxy for oligomers measured with

the HR-ToF-AMS (Figure S 6.8, Supplement).

6.3.3 Particle size dependent uptake of organic mass

As the seed composition and RH appear to greatly influence SOA yields, we examine in this

section the interaction between these parameters and SOA, through the investigation of the

aerosol size-resolved chemical composition. This information is used later to deduce the

behavior of the absorptive organic phase, while modeling the aerosol dynamics in the SC is

beyond the scope of this study.

a)

b)

Figure 6.4. Evolution of size distributions. Measured organic, SO4, NH4, NO3 mass distributions for OH exposures of 0×, (0.5 ± 0.2)×, (1.0 ± 0.3)× and (2.0 ± 0.5) ×107 cm-3 h. Black lines represent estimated liquid water content (method: Sect. 6.2.3; RH and individual GFs given in the legend, percentage in brackets: fractions of SO4). The calculated dry and wet surface distributions are shown as dashed lines on the right axes. (a) High NOx with more acidic seed: Experiments 3 and 4 with NOx/α-pinene = 1.2 and exp. 2 with NOx/α-pinene = 3.8. (b) Low NOx with less acidic seed: Experiments 8 and 13 with NOx/α-pinene ≈ 0.05. Additional figures in Supplement (Figure S 6.9).

88 6 The link between organic aerosol mass loading and degree of oxygenation

Figure 6.4 shows the aerosol size resolved chemical composition at 0 ×107 cm-3 h,

(0.5 ± 0.2) ×107 cm-3 h, (1.0 ± 0.3) ×107 cm-3 h and (2.0 ± 0.5) ×107 cm-3 h OH exposure for

five experimental conditions. The liquid water mass distributions (black lines) and the dry and

wet surface distributions (dashed lines) were estimated as described in Sect. 6.2.3. Figure 6.4a

includes high NOx experiments 3 and 4 (NOx/α-pinene = 1.2) and experiment 2 (NOx/α-

pinene = 3.8). Figure 6.4b includes low NOx experiments 8 and 13 (NOx/α-pinene

ratio ≈ 0.05). Additional figures for experiments 1, 5 and 6 are available in the Supplement

(Figure S 6.9).

For all experiments, the aerosol size distributions show two externally mixed aerosol

populations, with a mode at lower diameters (~200 nm, mode1) mostly containing SOA and a

second mode at larger diameters (~400 nm, mode2) mostly consisting of the seed. We note

that the particle size distribution evolved consistently under different conditions, with the

smallest seed particles growing with SOA condensation. For higher yields, the main SOA

mass occurs in mode1. For example, despite the sizeable increase of the yield with particulate

water content, we did not observe a significant enhancement of SOA in mode2 for acidic

conditions or with RH. In addition, we did not note any statistically significant correlation

between the initial seed volume and SOA yields (Table 1), but instead the aerosol uptake

seems to be driven by the wet particle surface (Figure 6.4). Such behavior would imply that

SOA semi-volatile compounds do not additionally partition or react in the electrolyte rich

phase, but rather the reactive or non-reactive uptake of these products onto the particle surface

is enhanced with the increase of the initial particulate water content. Homogeneous nucleation

events are excluded, as there is no rapid particle number concentration increase after lights

were switched on (Figure S 6.11, Supplement) and no mode evolving from very small to

larger diameters, seen from SMPS data.

We assume uptake of α-pinene products by the total organic mass (including organonitrates)

and derived the corresponding LWC as described in Sect. 6.2.3. Due to the strong dependence

of the SOA yield on the NOx/α-pinene ratio, the data presented in Figure 6.5 was separated

accordingly (upper panel: NOx/α-pinene < 0.1; lower panel: NOx/α-pinene > 1) and shown for

absorptive masses larger than 1 µg m-3. If the LWC was only contributing to the liquid

absorptive mass concentration, the yield curves for low and high RH experiments should

overlay. This is however only the case for experiments 10 and 12 at absorptive mass

concentrations > 6 µg m-3. For lower absorptive mass concentrations all experiments that are

comparable in their seed composition, NOx/α-pinene and α-pinene concentration reacted,

6.3 Results 89

show an enhancement in the yield for higher RH. This means that a process to

comprehensively describe the yield is missing, when only considering organics and H2O. We

suggest that this increase is due to the direct influence of gaseous and particulate water to the

partitioning coefficient which is increased when the average molecular weight is decreased by

adding H2O to the system. Models show that this effect is more pronounced for experiments

at low organic mass loading (Hallquist et al., 2009; Pankow, 2011 and references therein).

Another increase is observed with increasing seed acidity: Yields of expts. 1, 2 and 3 (acidic)

are higher than the yields of exp. 9 (neutral) and expts. 15 and 16 (CF-seed). This might be

related to enhanced acid-catalyzed heterogeneous reactions on the seed surface (Kroll and

Seinfeld, 2008) or to the higher hygroscopic growth factor for H2SO4 compared to (NH4)2SO4

and NH4HSO4 (Topping et al., 2005). Aqueous phase chemistry with a potential to form

highly oxidized organic acids or oligomers plays a minor role for enhanced yields found in

this study as no higher O : C in Van Krevelen diagrams (Figure 6.3) and no increased

oligomer fraction (Figure S 6.8, Supplement) is correlated with the increased yields.

Figure 6.5. 20-min-averaged measured (symbols) and fitted (lines) wall-loss-corrected SOA yield as a function of wall-loss-corrected absorptive mass concentration (organics + NO3 + H2O) for low NOx experiments (upper panel) and high NOx experiments (lower panel). Data was limited to 2 ×107 cm-3 h OH exposure. Symbol colors indicate RH, symbol shapes the seed composition.

90 6 The link between organic aerosol mass loading and degree of oxygenation

Data in Figure 6.5 was used to estimate the mass yields per volatility bin ai for each

experiment using Eq. (6.10). Figure 6.6 shows the probability density functions (PDF) of ai

values of 100 possible solutions for one low RH (No. 4) and the corresponding high RH (No.

3) experiment. The higher intensity of mass yields in the Ci* bins 10 and 100 for high RH

shows the increased partitioning of more volatile products during experiments with higher RH

and yield. More PDFs are provided in the Supplement, Figure S 6.12.

Figure 6.6. Probability density functions (PDF) of mass yields ai for volatility bins (Ci* = 0.1, 1, 10,

100) for one low RH (No. 4, left panel) and one high RH (No. 3, right panel) experiment.

6.4 Discussion

A key parameter for yield parameterization, shown in Figure 6.5, is the estimation of the

absorptive mass, COA. Traditionally, COA is considered equal to the total OA mass and the

inorganic fraction is not considered to take part in the partitioning, which is consistent with

our observations (Sect. 6.3). As mentioned above, a great part of the organic mass occurs

within the particle population with a small inorganic fraction and the yields do not exhibit a

significant correlation with the inorganic aerosol mass. We suggest that even organic

compounds present in the same particles as the inorganic seed (mode2, Figure 6.4) are likely

to form 2 separate phases: an organic-rich phase and an electrolyte-rich phase. We recognize

that experimental determinations of liquid-liquid phase separation are lacking, but as shown

in the literature, mixing between sulfate and organic compounds only occur at very high RH

and for highly oxygenated compounds (O:C > 0.8; You et al., 2013). Therefore, we will

consider the organic phase, i.e. OA and the bounded water, to represent the absorptive mass.

6.4 Discussion 91

For the hydrophobic seed, its mixing properties with of SOA component are unknown and

hence experiments using this seed will not be included in the yield parameterization.

Products formed early after the oxidation started own two properties due to their high O : C:

1) They have a very low volatility and ability to stick at their first collision and thus

heterogeneous nucleation onto the surface occurs in the beginning of an experiment. 2) They

have a high miscibility with inorganic solutions and thus mixing with the prevailing seed can

occur in the beginning of an experiment. As these products stay in the aerosol phase and do

not evaporate they provide more absorptive mass for further products to partition in, which is

larger for higher relative humidity due to the hygroscopic behaviour.

The influence of the seed on the yield due to varying acidity or LWC remains to be fully

elucidated (i.e., to what extent: y(acidic seed) > y(neutral seed) > y(CF-seed)]. For higher

yields under acidic conditions no increasing effect on organics in mode2 is observed, while

this could have been related to a better mixing behavior of NH4HSO4 (You et al., 2013) and

H2SO4 at high RH. If absorptive partitioning is dominant, organics dilute efficiently in the

liquid phase; their concentration in the particle phase is lower for high LWC than for lower

LWC and their low partial pressure causes more uptake of organics, which in turn causes

more uptake of water. This feed-back reaction drives the uptake and provides even for the

small hygroscopic growth factors of organics a significant water volume. Cocker et al. (2001)

found a small RH effect studying α-pinene ozonolysis reaching up to 300 µg m-3 organic and

water concentrations, while few data points at mass concentrations 0-35µg m-3 and NH4HSO4

seed might even indicate an increase in yield compared to (NH4)2SO4. Also in Cocker et al.

(2001) little RH effect on very high SOA loadings was seen, following Pankow et al. (2010).

At high SOA loadings, all condensable products might already have condensed. We conclude

that low, atmospherically relevant organic mass concentrations were necessary to elucidate

the RH effect seen in this study.

Concerning miscibility, the mixing state of the aerosol might be a function of particle size.

When condensing α-pinene reaction products onto the CF-seed, there’s no measurable change

in the exponential decay of CF-related ions (e.g. CF3) in the HR-ToF-AMS measurement.

This indicates that the CF-seed doesn’t mix with α-pinene SOA, because otherwise it would

slow down the decrease of the CF-seed or even increase its suspended mass concentration by

partitioning effects.

At longer exposures to OH, SOA products increase in degree of oxygenation and their ability

to mix with the inorganic solutions might increase (You et al., 2013). Liquid-liquid phase

92 6 The link between organic aerosol mass loading and degree of oxygenation

separation could decrease organic partitioning because organic layers of different viscosity

inhibit fast diffusion and limit feedback uptake due to poorly mixed organic layers.

6.5 Conclusions

SOA yields from the photooxidation of α-pinene at low (~ 25 %) and high (~ 60 %) relative

humidity (RH), various NOx/VOC ratios (0.04–3.8) and with different seed chemical

compositions (sulfate-containing or hydrophobic carbon) were studied in controlled SC

experiments. High RH increased SOA yields up to five times compared to low RH, with

greater increases for the most hygroscopic/acidic seeds. While many aqueous phase studies

suggest a major effect of the relative humidity on the chemical composition via e. g.

oligomerisation, in this study, most of the organic products are unlikely to mix with the

prevailing seeds (nonpolar organic or polar inorganic) and the chemical composition of SOA

is mainly influenced by the NOx/α-pinene ratio via gas-phase reactions.

Yields presented here compare well with literature yields from α-pinene ozonolysis ranging

up to ~0.18 for 30 µg m-3 organic mass concentration summarized by Hao et al. (2011),

measured at typical laboratory temperatures (~25 ± 5 °C). In this study, yields at NOx/VOC

ratios < 0.1 were 2–11 times higher compared to yields at NOx/VOC ratios of up to 3.8. This

NOx dependence follows the same trend as seen in previous studies for α-pinene SOA (Ng et

al., 2007). Presto et al. (2005) found a factor up to 4 for ratios above 0.67 NOx/α-pinene. The

chemical signature, investigated in Van Krevelen diagrams, showed a main influence from the

NOx/VOC ratio. For the same equivalent OH (hydroxyl radical) exposure, experiments at low

NOx/VOC ( < 0.1) ratios formed products with higher H : C while experiments with higher

NOx/VOC ( > 1) showed lower H : C. This is expected from higher formation yields of

hydroxyl functional groups and peroxides at low NOx. We present wall-loss-corrected yields

as a function of absorptive masses combining organics and the bound liquid water content.

We parameterize the volatility distributions for the use in modelling studies. This study

clearly presents different yields for the same reacted α-pinene concentration for varying

experimental conditions. This might also occur in the ambient atmosphere, such that SOA

loadings are sensitive to slight changes in the absorptive mass concentration. The NOx/VOC

ratio and relative humidity have to be considered in models to accurately predict SOA yields.

6.5 Conclusions 93

Acknowledgements. This work has been supported by the EU 7th Framework projects

EUROCHAMP-2 and PEGASOS, as well as the Swiss National Science Foundation

(Ambizione PZ00P2_131673, SAPMAV 200021_13016), the EU commission (FP7,

COFUND: PSI-Fellow, grant agreement n.° 290605). We thank Rene Richter and Günther

Wehrle for their technical support at the smog chamber, and in addition Michel J. Rossi,

Martin Gysel, Andy Zünd, Neil Donahue and Barbara Turpin for the helpful discussions.

94 6 The link between organic aerosol mass loading and degree of oxygenation

6.6 Supplementary material

6.6.1 Transmission and collection efficiency correction

During experiments 9, 10, 12 and 14-17, the transmission efficiency of one of the two PM2.5

lenses used in the HR-ToF-AMS was weak for particles up to vacuum aerodynamic diameters

(dva) of 230 nm, the region were the organic : sulfate ratio is highest. The not-transmitted

aerosol was accounted for by means of HR-ToF-AMS and SMPS comparison. The sulfate

seed and CF-seed mass distributions were captured well within the measurement region by

both, AMS and SMPS. The seed volume Vseed was determined by applying the loss rate of the

CF-seed (kCF) or sulfate seed (kSO4), respectively, to the measured initial SMPS volume Vinitial:

∙ exp or ∙ exp (S 6.1)

The additionally formed organic volume ΔVSMPS was derived from the difference of total and

seed volume (∆VSMPS = Vtotal - Vseed). The method is displayed in Figure S 6.1. The collection

efficiency of organics (CEorg) in the AMS in turn was determined by the ratio of Vorg and

∆VSMPS shown in Figure S 6.2. HR-ToF-AMS mass concentrations divided by the

corresponding densities of the species yielded the AMS volume. CEorg was ~ 0.55 for

experiments 12 and 14 with lower RH and ~ 0.75 for experiments 9 and 10 with higher RH.

The CF-seed experiments showed CEorg of 1.

The remaining experiments were conducted with the second PM2.5 lens with good

transmission efficiency at lower dva. The SMPS upper diameter cut off for expts. 1–6 was set

to 600 nm mobility diameter instead of the standard 1000 nm in this study and thus the AMS

volume even exceeded the SMPS volume. Therefore, a CEorg of 1 was assumed for

experiments 1–6. The CEorg for experiments 7, 8, 11 and 13 were between 0.7–0.9. All

organic mass concentrations given in this study are corrected by dividing by the collection

efficiency, summarized in Table S 6.2.

6.6 Supplementary material 95

Figure S 6.1. Measured SMPS volume (thin line), estimated CF-seed volume (dashed line, with decay rate kCF determined from AMS data) and the difference of both, ΔVSMPS (thick line), as a function of time after lights on. ΔVSMPS corresponds to the secondary organic volume formed.

Figure S 6.2. Comparison of 20-minutes averaged volumes ΔVSMPS and VAMS (=Vorg). ΔVSMPS (= difference between total SMPS volume Vtotal and estimated SO4 volume). Legend contains information on which instrument was taken as reference.

96 6 The link between organic aerosol mass loading and degree of oxygenation

6.6.2 Seed composition

Figure S 6.3 shows the temporal evolution of NH4/SO4 in (moles/moles). Seen by the

different ratios and temporal evolution, the experiments were performed with different seed

compositions even when nebulizing always the same concentration (1g L-1) of NH4HSO4

solution in ultrapure milli-Q water. We estimated that 1 ppb of gaseous NH3 was needed to

neutralize 4 µg m-3 NH4HSO4 to (NH4)2SO4 using the following formula:

∙

∙ ∙ . ∙

/1ppb (S 6.2)

Where cNH3 is the concentration of NH3, Vmol is the molar volume, mSO4 the mass

concentration of SO4 and MSO4 the molar mass of SO4.

During experiments 1-6, the molar NH4/SO4 ratios represent acidic seed conditions. We claim

that this is due to a higher concentrated H2SO4 solution (0.1 M) serving the HONO source. A

volume as low as 0.1 µL of the 0.1 M H2SO4 solution would be sufficient to convert even the

highest seed concentration of liquid (NH4)2SO4 in the 27 m³ SC to NH4HSO4:

Assuming the highest seed concentration added to the chamber: 8 µg m-3 (NH4)2SO4 (=

6 × 10-8 mol m-³ (NH4)2SO4), scaled to the 27 m³ PSI smog chamber: 1.6 × 10-6 mol

(NH4)2SO4. To form NH4HSO4 from this, 1.6 × 10-6 mol H2SO4 = 1.6 × 10-4 g= 0.16 mg ≈

0.1 µL of pure H2SO4 would have to be injected to the chamber. 1 µL of our 0.1 M H2SO4

was most probably injected into the chamber.

The neutralization was confirmed by nebulization tests of (NH4)2SO4 and NH4HSO4 resulting

in apparent relative ionization efficiencies (RIE) of 1.1 and 0.55 for SO4, respectively, if no

chemical transformation is assumed. The RIE is defined as ionization efficiency (IE) of a

compound normalized to the IE of NO3 in the HR-ToF-AMS. Different RIEs for one

inorganic compound in the same HR-ToF-AMS are unlikely, rather point towards

neutralization of NH4HSO4 to (NH4)2SO4 between nebulizer and measurement.

6.6 Supplementary material 97

Figure S 6.3. Molar ratio of NH4 and SO4 as a function of time after lights on. Seed composition during experiments 1–6: SA & AHS due to significant concentration of H2SO4 added. During expts 7–14 the NH4HSO4 solution is partly neutralized to (NH4)2SO4 after nebulization.

98 6 The link between organic aerosol mass loading and degree of oxygenation

Figure S 6.4 shows the distinct mass spectrum of the fluorinated carbon seed (CF-seed)

aerosol. Table S 6.1 list the corresponding relative intensities compared to CF3 of each major

ion. The HR-ToF-AMS high resolution analysis is a good tool to distinguish between α-

pinene SOA and the CF-seed ion peaks in the mass spectrum. The decay rates of the CF-seed

concentration in the smog chamber could thus easily be estimated by fitting the sum of three

main ions (CF2, CF3, C2F3O).

Figure S 6.4. Mass spectrum of CF-seed measured with HR-ToF-AMS normalized to CF3 (ion with highest intensity).

Table S 6.1. Mass spectrum of CF-seed measured with HR-ToF-AMS normalized to CF3 (ion with highest intensity).

m/z Ion formula Normalized intensity

28 CO 0.130

31 CF 0.097

44 CO2 0.130

50 CF2 0.041

69 CF3 1.000

97 C2F3O 0.110

100 C2F4 0.094

119 C2F5 0.506

131 C3F5 0.113

m/z Ion formula Normalized intensity

147 C3F5O 0.130

150 C3F5 0.085

169 C3F7 0.762

185 C3F7O 0.013

235 C4F9O 0.010

285 C5F11O 0.041

297 C6F11O 0.010

335 C6F13O 0.032

6.6 Supplementary material 99

6.6.3 Additional figures and table

Table S 6.2. List of wall-loss-correction parameters and collection efficiencies applied to each experiment.

No 1/τ CForg

1 0.17 1

2 0.11 1

3 0.13 1

4 0.16 1

5 0.11 1

6 0.17 1

7 0.13 0.78

8 0.13 0.70

9 0.11 0.75

10 0.13 0.75

11 0.13 0.89

12 0.16 0.55

13 0.13 0.80

14 0.10 0.55

15 0.18 1

16 0.18 1

17 0.18 1

100 6 The link between organic aerosol mass loading and degree of oxygenation

Figure S 6.5. Molar concentrations of nitrate (NO3) as a function of molar concentration of carbon (C) for all experiments with hygroscopic seed. High NOx experiments follow a ratio of approximately 1:30, low NOx experiments a ratio of approximately 1:100. The two arrows indicate that NO3:C changes its slope when the initially high NOx concentration decayed to low values in the end of high NOx experiments 1, 2 and 3.

Figure S 6.6. 20-min-averaged organic mass concentration as a function of time after lights on. Traces were fit with an exponential between 5 and 8 h after lights on to determine wall loss rate. Wall loss decay rates for experiments with too short fit period were replaced by the mean 1/τ-value of 0.13 (dashed lines) and 0.18 (dotted lines).

6.6 Supplementary material 101

Figure S 6.7. 20-min-averaged wall-loss-corrected yield as a function of reacted α-pinene. At the same concentration of α-pinene reacted, different yields are obtained due to various NOx/α-pinene ratios, relative humidities and seed compositions. Traces are color coded for each experiment according to Table 1 in the main text. Asterisks indicate low NOx experiments.

Figure S 6.8. 20-min-averaged fraction of oligomers (m/z > 150) of the total organic mass concentration as a function of OH exposure. Symbols indicate the seed composition, colors indicate the respective experiment as given in the legend

Figure S 6.9. Evolution of size distributions. Measured organic, SO4, NH4, NO3 mass distributions for OH exposures of 0×, (0.5 ± 0.2)×, (1.0 ± 0.3)× and (2.0 ± 0.5) × 107 cm-3 h. Black lines represent estimated liquid water content (RH and individual GFs given for each expt., method described in main text). The calculated dry and wet surface distributions are shown on the right axes. Number of each experiment is given on the left edge (1, 5, 6).

6.6 Supplementary material 103

Figure S 6.10. 20-min averaged NO+/NO2+ ratio as a function of OH exposure compared to measured

NO+/NO2+ from calibration of NH4NO3 for each set of experiments.

Figure S 6.11. Total particle number concentration as a function of time after lights on. The increase after time = 0 is assumed due to an increased transmission efficiency of larger particles in the lines before the condensation particle counter measurement.

104 6 The link between organic aerosol mass loading and degree of oxygenation

6.6 Supplementary material 105

Figure S 6.12. Probability density functions (PDF) of mass yields ai for volatility bins (c*=0.1, 1, 10, 100) to solve Eq. (6.10) in the main text. Experiment number is given in the legend.

___________________________________________________________________________

7 VIsible light Photosensitized Secondary

Organic Aerosol evolution (VIPSOA)

___________________________________________________________________________

108 7 VIsible light Photosensitized Secondary Organic Aerosol evolution (VIPSOA)

7.1 Introduction and scientific objective

The direct interaction of radiation with particles might have an effect on their chemical

composition. Stemmler et al. (2006) found a conversion of nitrogen dioxide (NO2) to nitrous

acid (HONO) induced by visible light (400–700 nm) irradiation of humic acid films covering

the surface of a flow reactor. D'Anna et al. (2009) showed that near-ultraviolet and visible

irradiation initiated ozone removal on the surface of humic acid aerosol and coated wall flow-

tube systems. Monge et al. (2012) found that UV-A (315–380 nm) irradiation of gas-phase

isoprene and the internally mixed seed containing humic acid (HA), succinic acid and

NH4NO3(1∶10∶1 in weight) in pure air led to particle growth and control experiments showed

that the photosensitizer HA is required for this process. Those newly identified pathways of

photoinduced SOA formation and growth are suggested to be caused by the formation of

radicals and singlet oxygen at the surface of the seed aerosol that can react with VOC

impinging on the surface (Monge et al., 2012).

The oxidation state of a particle can be increased in a number of ways. First, the oxidation of

volatile organic compounds (VOC) by OH radicals in the gas-phase results in products with

lower saturation vapour pressures, which can condense on pre-existing particles (Figure 7.1a).

This process increases the total aerosol mass as well as the average O : C ratio, while in the

following processes mainly the O : C ratio is affected. Second, semi-volatile compounds may

evaporate from the surface, oxidize in the gas phase and re-condense again after the reaction,

while the carbon mass stay constant (Figure 7.1b). Third, OH radicals may react directly on

the surface of the particles (Figure 7.1c). Forth, an O : C increase may happen by radiation

entering the particle and initiating formation of OH radicals and oxidation within the particle

(Figure 7.1d). This last mentioned process is the subject of the present study. Photosensitizers,

chemical substances that absorb light, cause photochemical reactions via energy transfer and

exchange of electrons or protons, but don’t react themselves might be important in this

respect. Compounds included in the particle can turn into photosensitizers causing

photochemical reactions within the particle. Visible light (wavelength λ > 400 nm), which

excludes photolysis of gases and thus ozone and OH radical formation, on the chemical

composition of organic aerosol might play a non-negligible role. These experiments were

performed in collaboration with the group of Dr. Markus Ammann at PSI.

7.2 Method 109

Figure 7.1. Schematic of possible processes leading to an increase of the oxidation state of organic aerosols in the atmosphere. (VOC = volatile organic compound, OH = hydroxyl radical)

7.2 Method

During the ‘VIsible light Photosensitized Secondary Organic Aerosol evolution’ (VIPSOA)

study, wood burning (WB) emissions and known pure chemical compounds served as “real”

and “model” system, respectively.

For the introduction of wood burning emissions into the PSI smog chamber (SC; see Sect.

4.1.) the identical setup as described in detail in Heringa et al. (2011) was used involving a

modern logwood burner Attika Avant (year of manufacture: 2009), a heated ejector diluter

(Dekati Ltd., Tampere, Finland) and heated inlet lines (150 °C) to prevent condensation

(Figure 7.2). The gas phase and aerosol phase were monitored with a number of instruments

described in detail in Sect. 4.2: a CPC and an SMPS for the physical aerosol properties, and

an HR-ToF-AMS for the aerosol chemical composition, and O3 and NOx monitors for the gas-

phase composition.

In the “model” system, 112 µL (≈1000 ppb) liquid-phase cyclohexanol (double-bound

alcohol) was injected via the heated (T = 80 °C) glass bulb with 10 L min-1 flush and

10 L min-1 dilution flow into the smog chamber to serve as SOA precursor. A suspension of

the photosensitizer methylene blue (0.05 g L-1; light absorption between λ = 550–700 nm with

the maximum at λ = 656 nm) and ammonium sulfate (2 g L-1) in ultrapure milli-Q H2O was

nebulized with a pure air flow of 0.6 L min-1 and introduced in the smog chamber by addition

of a 10 L min-1 flush flow.

VOC+OH•

a) Exposure of gas phase VOC to OH radicals

OH• VOC

b) Oxidation and re-condensation of VOC

OH• VOC

b) Oxidation and re-condensation of VOC

OH•

c) Heterogeneous reaction on aerosol surface d) Solar radiation initiates reactions

110 7 VIsible light Photosensitized Secondary Organic Aerosol evolution (VIPSOA)

Figure 7.2. Schematic representation of the inlet system and smog chamber setup (Heringa et al., 2011, adapted).

All experiments were conducted at 50 ± 10 % relative humidity, 20–25 °C temperature and

after the standard smog chamber cleaning procedure described in Sect. 4.1. Four xenon arc

lamps installed in the chamber were covered with a UV blocking foil (METOLIGHT SFC-

10), resulting in a light spectrum with minor emission for λ < 400 nm. This setup is referred to

as ‘visible light’ hereafter. During experiments for the “model” system visible light was used

as radiation source, while during the wood burning experiments standard xenon arc lamps and

black lights mounted underneath the smog chamber bag to accelerate reactions were used to

activate the primary wood burning aerosol ahead of visible light and dark periods. A

comparison of the visible (xenon arc lamps covered with UV filter foil METOLIGHT SFC-

10), standard xenon (covered with fluorinated ethylene propylene (FEP) films on glass

filters), black (Cleo performance, 100W) and direct sun light spectra is shown in Figure 7.3,

together with the transmission efficiency of the METOLIGHT foil.

7.3 Results 111

Figure 7.3. Spectral resolved light intensity of visible (xenon arc lamps covered with UV filter foil METOLIGHT SFC-10), standard xenon (covered with fluorinated ethylene propylene (FEP) films on glass filters), black (Cleo performance, 100W) and direct sun light.

7.3 Results

Experiments using the model compounds gas-phase precursor cyclohexanol (C6H10O), the

photosensitizer methylene blue (MB) and ammonium sulfate (AS) seed did not show an effect

of visible light on SOA, at least when switching approximately every two hours between dark

and light exposure times. Table 7.1 lists the procedures of one blank (10.5.2010) and one real

experiment (15.5.2010).

Table 7.1. System model for experiments with and without methylene blue (MB), ammonium sulfate (AS) and cyclohexanone (C6H10O) with visible light irradiation.

date AS MB C6H10O visible lights comments

10.05.2010 blank without methylene blue

15.05.2010 no mass production, no effect on f44

112 7 VIsible light Photosensitized Secondary Organic Aerosol evolution (VIPSOA)

The more realistic system of wood burning emissions was investigated during six experiments

of which light exposure and dark periods are chronologically listed in Table 7.2. After

injection of primary wood burning emissions to the PSI smog chamber, 2–7 h exposure of UV

light radiation (except exp. 6) led to formation of secondary organic aerosol, increased

concentration of ozone and decreased concentration of nitrogen oxide (NO) stemming from

the burning process. During experiments 1 and 2, after UV lights were switched off, the

existing ozone (O3) concentration was titrated away by the addition of a slightly higher

concentration of NO [O3+NO→NO2+O2]. Thus the oxidation with ozone is excluded for these

experiments, but NO3 chemistry could still be possible by the back reaction of N2O5 formed

during the dark period [NO3+NO2↔N2O5]. A slight decrease of NO and NO2 during visible

light exposure was observed. The evolutions of O3, NO and NO2 concentrations during each

period are listed as the start and end concentrations in Table 7.2.

Table 7.2. Observed changes in the temporal evolution of the organic mass fractionf44 of wood burning (WB) aerosol during darkness and different light exposure periods. The experiment description indicates whether titration of O3 by NO addition took place. The radiation types Xe = xenon lights (+BL = black lights), vis = visible lights (λ > 400 nm) and dark = no radiation are listed chronologically together with the exposure time. Experiment 6 is the only experiment without SOA formation by UV activation (visible light only). NO, NO2 and O3 concentration changes are given as start and end concentration per exposure time.

date No experiment radiation time slope f44 (fit) NO NO2 O3

description type h [% h-1] ppb ppb ppb

13.9.2010

WB with NO titration

Xe+BL 2 -0.6 ∆f44/∆t* 35-2 0-28 2.8-38

1 dark 20 0.04 ± 0.01 16-10 47-46 0

vis 5 0.15 ± 0.03 9-7 50 0

27.9.2010 2 WB with NO titration Xe+BL 2 -2.5 ∆f44/∆t* 84-10 16-66 0-23

dark 8 0.08 ± 0.01 10-8 123-119 0

26.8.2010 3 WB w/o NO titration

Xe+BL 5 0.47 ± 0.04 14-0.3 6-7 8-100

vis 14 0.12 ± 0.01 0 6-1.5 96-66

dark 9 0.04 ± 0.04 0 1.3-1.2 65-57

27.5.2010 4 WB w/o NO titration Xe 2 0.9 ∆f44/∆t* 86-11 8-57 7-22

vis 16 0.11 ± 0.01 0 70-40 6-7

23.9.2009 5+ WB w/o NO titration Xe 7 1.0 ∆f44/∆t* - - -

22.9.2010 6 WB visible light only vis 19 0.04 ± 0.02 0 6 ± 1 22-49

+) personal communication M. Heringa, PSI

*) ∆f44/∆t instead of fit due to a varying slope over time

7.3 Results 113

The evolution of the gas and aerosol phase for experiment 1 is displayed in Figure 7.4 (lower

panel), where the grey shaded areas represent light exposure periods (2 h xenon and black

lights, 20 h darkness, 5 h visible lights). Aerosol species organics, nitrate and ammonium and

gas-phase NO and O3 were injected with the WB emissions. SOA was formed, while O3 (and

NO2, not shown) increased and NO decreased during the UV light irradiation. NO injection

titrated all O3 and NO2 formed. During the dark period, aerosol species decreased due to wall

loss and a slight NO and NO2 decrease for unknown reason is observed. Once the visible

lights were switched on, the decrease in NO was steeper, while NO2 stayed constant. The

upper panel of Figure 7.4 shows the corresponding evolutions of the organic mass fractions

f44, f43 and their ratio. The chemical composition of primary wood burning emissions is quite

variable (Heringa et al., 2011) and in this case represented by a percentage of ~ 8 % for

organic mass 44 of total organics (f44) and 2 % f43. UV radiation led to reactions of the fresh

emissions and decreases f44 and f43 by condensation of newly formed semi-volatile organic

compounds. During the dark period, f44 increased with 0.04 % h-1 together with a loss of the

suspended organic mass concentration. Once the visible lights were switched on, the increase

in f44 was steeper, 0.15 % h-1, but also accompanied by a decrease in the NO concentration.

Figure 7.4. Temporal evolution of organic mass fractions f44, f43, their ratio org44/org43 (upper panel) and organics, NO3, SO4, NH4, with the NO and O3 concentrations (lower panel) as a function of time (expt. 1). Experimental procedure: SC filling with wood burning emissions, UV light irradiation, O3 titration with NO, dark period, visible light exposure.

114 7 VIsible light Photosensitized Secondary Organic Aerosol evolution (VIPSOA)

A comparison of the increase in organic mass fraction f44 during experiments 1 (UV, dark,

visible), 5 (UV only) and 6 (visible only) is shown in Figure 7.5 where a similarity of dark

(red) and visible light without prior UV ‘activation’ (green) periods and the less and less steep

increase in f44 during UV light exposure (xe, violet) can be seen.

Figure 7.5. Organic mass fraction f44 as a function of time after lights on for the three different radiation types: UV light only (violet), UV activation,

Furthermore, the change in the organic chemical composition during UV light exposure and

visible light exposure was investigated for the entire mass spectrum measured by the AMS

(see instrument Sect. 4.2). Figure 7.6a shows normalized organic mass spectra measured after

2 h and 5 h of UV light exposure (upper panel) and their difference mass spectrum (lower

panel) with main increases in m/z 18, 28, 44 and 58, and main decreases in m/z 29, 60 during

experiment 5. Figure 7.6b shows normalized organic mass spectra measured in the beginning

and after 15.5 h of visible light exposure (upper panel) and their difference mass spectrum

(lower panel) with main increases in m/z 18, 28, 44 and 58, and main decreases in m/z 15, 29,

31, 60 during experiment 4. All in all, very similar mass spectra and increases/decreases in the

same m/z were found. The chemical composition was similar, but the change in chemical

composition was slower under visible light and in the dark compared to UV light.

7.3 Results 115

Figure 7.6. (a) Normalized organic mass spectra measured after 2 h and 5 h of UV light exposure during experiment 5 and their difference mass spectrum (main increases in m/z 18, 28, 44 and 58, main decreases in m/z 29, 60). (b) Normalized organic mass spectra measured in the beginning and after 15.5 h of visible light exposure during experiment 4 and their difference mass spectrum (main increases in m/z 18, 28, 44 and 58, main decreases in m/z 15, 29, 31, 60).

116 7 VIsible light Photosensitized Secondary Organic Aerosol evolution (VIPSOA)

To summarize, first results indicate a measurable impact on particle oxidation, but no major

SOA production enhanced by visible light. The slope of the organic mass fraction f44 versus

time in the HR-ToF-AMS which is correlated to the O : C ratio showed significant differences

in the darkness and during visible light exposure. There are indications that an activation of

the wood burning aerosol by UV light exposure is needed to start the process that is prevailing

with visible light.

7.4 Measurement challenges and outlook

Despite promising results a comprehensive study of the effects of radiation on aerosol was

beyond the scope of this thesis. Additional parameters need to be investigated to be able to

draw final conclusions:

- The presence of OH radicals needs to be excluded by means of PTRMS (described in

Sect. 4.2) measurements. The slightest photolysis of e.g. HONO due to a small

fraction of short wavelength radiation in the light spectrum would lead to OH radical

formation.

- A possible temperature increase during visible light exposure periods could lead to

increased evaporation of particle phase organic compounds or could enhance

oxidizing reactions in the aerosol and thus lead to an increase in O : C. Conditions

could be kept more stable by decreasing the temperature setting of the air conditioning

system of the PSI smog chamber immediately prior to the visible light exposure

periods.

- Wall losses could be different due to increased turbulence initiated by the heat transfer

from the covered xenon lights to the smog chamber during light exposure compared to

darkness.

- The possible formation of N2O5 can be omitted if UV light exposure takes places

during O3 titration by NO.

- The measurement duration of 20 h and more is prone to possible contamination of air

entering the smog chamber through leaks. With the small scale changes observed,

contamination has to be minimized and well characterized.

7.4 Measurement challenges and outlook 117

The following technical ideas could minimize the OH radical concentration:

1. CO could be used as OH scavenger

2. Aerosol produced in the smog chamber could be isolated from the gas-phase by a

denuder and the effect of radiation onto the aerosol suspended in pure air could be

investigated by means of the PAM (potential aerosol mass, Kang et al., 2007)

chamber. Vice versa, aerosol could be produced in the PAM chamber, isolated from

the gas phase and then exposed to visible light in the smog chamber.

Finally, the matrix of the study could be extended to other systems such as α-pinene, diesel

emissions or suitable model compounds to draw general conclusions and suggest processes

that may change the chemical state of the aerosol by radiation.

___________________________________________________________________________

8 Final conclusions and outlook

___________________________________________________________________________

120 8 Final conclusions and outlook

The organic aerosol burden predicted by global “bottom-up” models used to be significantly

underestimated compared to ambient measurements, before a larger volatility range of organic

species and a more dynamic nature of POA and SOA in terms of volatility were allowed. This

meant a big step forward in understanding and predicting the aerosol burden and OA

formation processes. Smog chamber studies increase the knowledge of aerosol properties by

providing measurement data of isolated, well-controlled processes. SC experiments used to be

performed with very high mass loadings of hundreds of µg m-3, thought valid to be

extrapolated to low ambient concentrations. The presented work shows that this is however

not the case, because the chemical composition of α-pinene SOA is dependent on the organic

mass concentration, being more oxygenated at low organic mass loadings. For the first time,

LV-OOA-like aerosol from α-pinene was produced in a smog chamber by oxidation at typical

atmospheric OH concentrations. Based on these findings smog chamber studies must be

performed at near-ambient concentrations to accurately simulate ambient aerosol.

SOA is mainly formed by the oxidation of gas-phase precursors where SOA yield and

composition depend on the surrounding gas-phase mixture. Reactions in VOC-limited or

NOx-limited regimes yield products of very different chemical signature, and thus SOA

composition. In the atmosphere, NOx levels close to emission sources are high, being mainly

of anthropogenic origin. For this reason, we present data of α-pinene photooxidation SOA

formed under different NOx/α-pinene conditions and find a negative correlation of yield as a

function of NOx/α-pinene, given the same seed composition. This is in line with previous

studies.

Atmospheric water vapour is of special interest, given the fact that it is a greenhouse gas with

highly variable concentration in the atmosphere. Present as particulate water in the aerosol

phase, it changes refractive indices and influences the uptake of water-soluble or water-

insoluble organic compounds. For example, an increased uptake of water-soluble organic

compounds was found with increasing RH in the ambient atmosphere (Hennigan et al., 2009).

In this study, the impact of relative humidity on the yield and chemical composition of SOA

products from α-pinene photooxidation was studied in a set of SC experiments. SOA yields

were up to 6 times higher during experiments at high RH (~60 %) compared to low RH

(~25 %), while not significantly affecting the chemical signature, at least when using spectra

obtained from the Aerodyne aerosol mass spectrometer. This work elucidates how sensitive

SOA compositions and yields are to a variety of parameters, even when mainly focusing on α-

pinene as one single precursor.

7.4 Measurement challenges and outlook 121

A possible impact of direct radiation on the chemical composition of SOA formed by

oxidation of wood burning emissions in the PSI SC was suggested. This may be investigated

with a better controlled experimental setup, that could reveal if the radiation effects

previously measured in flow reactors are applicable to an environmental chamber and thus

more comparable to the atmosphere. In addition, well-chosen model systems to better pin

down the responsible mechanisms and systems, where no effect is expected, would complete

the study.

The impact of relative humidity on SOA from many more precursors than α-pinene should be

studied in the future. This study presents yields up to an organic mass concentration of 30

µg m-3. More experiments with similar seed composition, relative humidity and NOx/VOC

with an organic mass range extended to higher concentrations could help to clearly

distinguish yield curves for each condition as a function of absorptive mass concentration.

These parameterizations would show higher statistical robustness, advantageous also for

modellers.

Another question is, how the laboratory produced α-pinene SOA would behave in the

atmosphere when mixed with a constant miscible organic background aerosol. The

experiments include to some extent a self-enhancing feedback mechanism: when SOA is

condensed, more SOA will partition in return. This effect might be minimized in future

experiments by 1) performing experiments with constant, discriminable background COA and

observing if the newly formed SOA fraction still varies as a function of gas-phase

composition or 2) by aiming for similar SOA yields with well-balanced initial conditions.

___________________________________________________________________________

List of abbreviations and acronyms

___________________________________________________________________________

aci = aerosol-cloud interactions

AHS = ammonium hydrogen sulfate

ari = aerosol-radiation interactions

AS = ammonium sulfate

BC = black carbon

BVOC = biogenic volatile organic

compound

CF-seed = fluorinated hydrocarbon as seed

CF = fluorinated carbon

CPC = condensation particle counter

DMA = differential mobility analyzer

dva = vacuum aerodynamic diameter

ERF = effective radiative forcing

GF = growth factor

H : C = atomic hydrogen: carbon ratio

HONO = nitrous acid

HR-ToF-AMS = Aerodyne high resolution

time-of-flight aerosol mass

spectrometer

IPCC = Intergovernmental Panel on

Climate Change

LV-OOA = low-volatility oxygenated

organic aerosol

LWC = liquid water content

OA = organic aerosol

O : C = atomic oxygen: carbon ratio

OM = organic matter

PMx = particulate matter with aerodynamic

diameter smaller than x µm

POA = primary organic aerosol

PSI = Paul Scherrer Institute

RH = relative humidity

RIE = relative ionization efficiency

S = surface

SA = sulfuric acid

SC = smog chamber

SMPS = scanning mobility particle sizer

SOA = secondary organic aerosol

T = temperature

VBS = “volatility basis set”

VOC = volatile organic compound

V = volume

wlc = wall-loss-corrected

ZSR = Zdanovskii-Stokes-Robinson

___________________________________________________________________________

Bibliography

___________________________________________________________________________

Aiken, A. C., DeCarlo, P. F., Kroll, J. H., Worsnop, D. R., Huffman, J. A., Docherty, K. S., Ulbrich, I. M., Mohr, C., Kimmel, J. R., Sueper, D., Sun, Y., Zhang, Q., Trimborn, A., Northway, M., Ziemann, P. J., Canagaratna, M. R., Onasch, T. B., Alfarra, M. R., Prevot, A. S. H., Dommen, J., Duplissy, J., Metzger, A., Baltensperger, U., and Jimenez, J. L.: O/C and OM/OC ratios of primary, secondary, and ambient organic aerosols with high-resolution time-of-flight aerosol mass spectrometry, Environ. Sci. Technol., 42, 4478-4485, 2008.

Alfarra, M. R., Paulsen, D., Gysel, M., Garforth, A. A., Dommen, J., Prevot, A. S. H., Worsnop, D. R., Baltensperger, U., and Coe, H.: A mass spectrometric study of secondary organic aerosols formed from the photooxidation of anthropogenic and biogenic precursors in a reaction chamber, Atmos. Chem. Phys., 6, 5279-5293, doi: 10.5194/acp-12-6417-2012, 2006.

Alfarra, M. R., Hamilton, J. F., Wyche, K. P., Good, N., Ward, M. W., Carr, T., Barley, M. H., Monks, P. S., Jenkin, M. E., Lewis, A. C., and McFiggans, G. B.: The effect of photochemical ageing and initial precursor concentration on the composition and hygroscopic properties of β-caryophyllene secondary organic aerosol, Atmos. Chem. Phys., 12, 6417-6436, 2012.

Allan, J. D., Delia, A. E., Coe, H., Bower, K. N., Alfarra, M. R., Jimenez, J. L., Middlebrook, A. M., Drewnick, F., Onasch, T. B., Canagaratna, M. R., Jayne, J. T., and Worsnop, D. R.: A generalised method for the extraction of chemically resolved mass spectra from Aerodyne aerosol mass spectrometer data, J. Aerosol Sci., 35, 909-922, 2004.

Andreae, M. O., Charlson, R. J., Bruynseels, F., Storms, H., van Grieken, R., and Maenhaut, W.: Internal mixture of sea salt, silicates, and excess sulfate in marine aerosols, Science, 232, 1620-1623, doi: 10.1126/science.232.4758.1620, 1986.

Bahreini, R., Keywood, M. D., Ng, N. L., Varutbangkul, V., Gao, S., Flagan, R. C., Seinfeld, J. H., Worsnop, D. R., and Jimenez, J. L.: Measurements of secondary organic aerosol from oxidation of cycloalkenes, terpenes, and m-xylene using an Aerodyne aerosol mass spectrometer, Environ. Sci. Technol., 39, 5674-5688, 2005.

Barley, M., Topping, D. O., Jenkin, M. E., and McFiggans, G.: Sensitivities of the absorptive partitioning model of secondary organic aerosol formation to the inclusion of water, Atmos. Chem. Phys., 9, 2919-2932, doi: 10.5194/acp-9-2919-2009, 2009.

Barmet, P., Dommen, J., DeCarlo, P. F., Tritscher, T., Praplan, A. P., Platt, S. M., Prevot, A. S. H., Donahue, N. M., and Baltensperger, U.: OH clock determination by proton transfer reaction mass spectrometry at an environmental chamber, Atmos. Meas. Tech., 5, 647-656, doi: 10.5194/amt-5-647-2012, 2012.

Bonn, B., Schuster, G., and Moortgat, G. K.: Influence of water vapor on the process of new particle formation during monoterpene ozonolysis, J. Phys. Chem. A, 106, 2869-2881, 2002.

Canonaco, F., Crippa, M., Slowik, J. G., Baltensperger, U., and Prévôt, A. S. H.: SoFi, an IGOR-based interface for the efficient use of the generalized multilinear engine (ME-2) for the source apportionment: ME-2 application to aerosol mass spectrometer data, Atmos. Meas. Tech., 6, 3649-3661, doi: 10.5194/amt-6-3649-2013, 2013.

126 Bibliography

Chang, R. Y. W., Slowik, J. G., Shantz, N. C., Vlasenko, A., Liggio, J., Sjostedt, S. J., Leaitch, W. R., and Abbatt, J. P. D.: The hygroscopicity parameter (κ) of ambient organic aerosol at a field site subject to biogenic and anthropogenic influences: relationship to degree of aerosol oxidation, Atmos. Chem. Phys., 10, 5047-5064, doi: 10.5194/acp-10-5047-2010, 2010.

Chhabra, P. S., Ng, N. L., Canagaratna, M. R., Corrigan, A. L., Russell, L. M., Worsnop, D. R., Flagan, R. C., and Seinfeld, J. H.: Elemental composition and oxidation of chamber organic aerosol, Atmos. Chem. Phys., 11, 8827-8845, doi: 10.5194/acp-11-8827-2011, 2011.

Clegg, S., Kleeman, M., Griffin, R., and Seinfeld, J.: Effects of uncertainties in the thermodynamic properties of aerosol components in an air quality model–Part 1: Treatment of inorganic electrolytes and organic compounds in the condensed phase, Atmos. Chem. Phys., 8, 1057-1085, 2008a.

Clegg, S., Kleeman, M., Griffin, R., and Seinfeld, J.: Effects of uncertainties in the thermodynamic properties of aerosol components in an air quality model–Part 2: Predictions of the vapour pressures of organic compounds, Atmos. Chem. Phys., 8, 1087-1103, 2008b.

Clegg, S. L., and Seinfeld, J. H.: Thermodynamic models of aqueous solutions containing inorganic electrolytes and dicarboxylic acids at 298.15 K. 1. The acids as nondissociating components, J. Phys. Chem. A, 110, 5692-5717, doi: 10.1021/jp056149k, 2006.

Cocker, D. R. I., Clegg, S. L., Flagan, R. C., and Seinfeld, J. H.: The effect of water on gas–particle partitioning of secondary organic aerosol. Part I: α-pinene/ozone system, Atmos. Environ., 35, 6049-6072, 2001.

D'Anna, B., Jammoul, A., George, C., Stemmler, K., Fahrni, S., Ammann, M., and Wisthaler, A.: Light-induced ozone depletion by humic acid films and submicron aerosol particles, J. Geophys. Res., 114, 1-12, doi: 10.1029/2008JD011237, 2009.

DeCarlo, P. F., Kimmel, J. R., Trimborn, A., Northway, M. J., Jayne, J. T., Aiken, A. C., Gonin, M., Fuhrer, K., Horvath, T., Docherty, K. S., Worsnop, D. R., and Jimenez, J. L.: Field-deployable, high-resolution, time-of-flight aerosol mass spectrometer, Anal. Chem., 78, 8281-8289, 2006.

DeCarlo, P. F., Ulbrich, I. M., Crounse, J., de Foy, B., Dunlea, E. J., Aiken, A. C., Knapp, D., Weinheimer, A. J., Campos, T., Wennberg, P. O., and Jimenez, J. L.: Investigation of the sources and processing of organic aerosol over the Central Mexican Plateau from aircraft measurements during MILAGRO, Atmos. Chem. Phys., 10, 5257-5280, doi: 10.5194/acp-10-5257-2010, 2010.

Dockery, D. W., Pope, C. A., Xu, X., Spengler, J. D., Ware, J. H., Fay, M. E., Ferris, B. G., and Speizer, F. E.: An association between air pollution and mortality in six U.S. cities, N. Engl. J. Med., 329, 1753-1759, doi: 10.1056/NEJM199312093292401, 1993.

Dommen, J., Hellén, H., Saurer, M., Jaeggi, M., Siegwolf, R., Metzger, A., Duplissy, J., Fierz, M., and Baltensperger, U.: Determination of the aerosol yield of isoprene in the presence of an organic seed with carbon isotope analysis, Environ. Sci. Technol., 43, 6697-6702, doi: 10.1021/es9006959, 2009.

Donahue, N. M., Robinson, A. L., Stanier, C. O., and Pandis, S. N.: Coupled partitioning, dilution, and chemical aging of semivolatile organics, Environ. Sci. Technol., 40, 2635-2643, 2006.

Donahue, N. M., Epstein, S. A., Pandis, S. N., and Robinson, A. L.: A two-dimensional volatility basis set: 1. organic-aerosol mixing thermodynamics, Atmos. Chem. Phys., 11, 3303-3318, 2011.

Bibliography 127

Donahue, N. M., Henry, K. M., Mentel, T. F., Kiendler-Scharr, A., Spindler, C., Bohn, B., Brauers, T., Dorn, H. P., Fuchs, H., Tillmann, R., Wahner, A., Saathoff, H., Naumann, K.-H., Möhler, O., Leisner, T., Müller, L., Reinnig, M.-C., Hoffmann, T., Salo, K., Hallquist, M., Frosch, M., Bilde, M., Tritscher, T., Barmet, P., Praplan, A. P., DeCarlo, P. F., Dommen, J., Prévôt, A. S. H., and Baltensperger, U.: Aging of biogenic secondary organic aerosol via gas-phase OH radical reactions, Proc. Natl. Acad. Sci. U. S. A., 109, 13503-13508, doi: 10.1073/pnas.1115186109, 2012.

Duplissy, J., Gysel, M., Alfarra, M. R., Dommen, J., Metzger, A., Prevot, A. S. H., Weingartner, E., Laaksonen, A., Raatikainen, T., Good, N., Turner, S. F., McFiggans, G., and Baltensperger, U.: Cloud forming potential of secondary organic aerosol under near atmospheric conditions, Geophys. Res. Lett., 35, L03818, doi: 10.1029/2007GL031075, 2008.

Duplissy, J., DeCarlo, P. F., Dommen, J., Alfarra, M. R., Metzger, A., Barmpadimos, I., Prevot, A. S. H., Weingartner, E., Tritscher, T., Gysel, M., Aiken, A. C., Jimenez, J. L., Canagaratna, M. R., Worsnop, D. R., Collins, D. R., Tomlinson, J., and Baltensperger, U.: Relating hygroscopicity and composition of organic aerosol particulate matter, Atmos. Chem. Phys., 11, 1155-1165, doi: 10.5194/acp-11-1155-2011, 2011.

Dusanter, S., Vimal, D., Stevens, P. S., Volkamer, R., and Molina, L. T.: Measurements of OH and HO2 concentrations during the MCMA-2006 field campaign - Part 1: Deployment of the Indiana University laser-induced fluorescence instrument, Atmos. Chem. Phys., 9, 1665-1685, doi: 10.5194/acp-9-1665-2009, 2009.

El Haddad, I., D'Anna, B., Temime-Roussel, B., Nicolas, M., Boreave, A., Favez, O., Voisin, D., Sciare, J., George, C., Jaffrezo, J. L., Wortham, H., and Marchand, N.: Towards a better understanding of the origins, chemical composition and aging of oxygenated organic aerosols: case study of a Mediterranean industrialized environment, Marseille, Atmos. Chem. Phys., 13, 7875-7894, doi: 10.5194/acp-13-7875-2013, 2013.

EPA, U. S.: Uses, limitations and technical basis of procedures for quantifying relationships between photochemical oxidants and precursors, EPA-450/2-77-021a, 1977.

Farmer, D., Matsunaga, A., Docherty, K., Surratt, J., Seinfeld, J., Ziemann, P., and Jimenez, J.: Response of an aerosol mass spectrometer to organonitrates and organosulfates and implications for atmospheric chemistry, Proc. Natl. Acad. Sci. U. S. A., 107, 6670-6675, 2010.

Finlayson-Pitts, B. J., and Pitts, J. N.: Chemistry of the upper and lower atmosphere, Academic Press, San Diego, USA, 2000.

Forster, P., Ramaswamy, V., Artaxo, P., Berntsen, T., Betts, R., Fahey, D. W., Haywood, J., Lean, J., Lowe, D. C., Myhre, G., Nganga, J., Prinn, R., Raga, G., Schulz, M., and Dorland, R. V.: Changes in atmospheric constituents and in radiative forcing. in: 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., D. Qin, M. Manning, Z. Chen, M. Marquis, K.B. Averyt, M.Tignor and H.L. Miller (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA., 2007.

Guderian, R.: Handbuch der Umweltveränderungen und Ökotoxikologie, Band 1A: Atmosphäre, Springer-Verlag, Berlin-Heidelberg-New York, 2000.

Guenther, A., Hewitt, C. N., Erickson, D., Fall, R., Geron, C., Graedel, T., Harley, P., Klinger, L., and Lerdau, M.: A global model of natural volatile organic compound emissions, J. Geophys. Res., 100, 8873-8892, 1995.

128 Bibliography

Guenther, A., Karl, T., Harley, P., Wiedinmyer, C., Palmer, P., and Geron, C.: Estimates of global terrestrial isoprene emissions using MEGAN (Model of Emissions of Gases and Aerosols from Nature), Atmos. Chem. Phys., 6, 3181-3210, 2006.

Hallquist, M., Wenger, J. C., Baltensperger, U., Rudich, Y., Simpson, D., Claeys, M., Dommen, J., Donahue, N. M., George, C., Goldstein, A. H., Hamilton, J. F., Herrmann, H., Hoffmann, T., Iinuma, Y., Jang, M., Jenkin, M. E., Jimenez, J. L., Kiendler-Scharr, A., Maenhaut, W., McFiggans, G., Mentel, T. F., Monod, A., Prevot, A. S. H., Seinfeld, J. H., Surratt, J. D., Szmigielski, R., and Wildt, J.: The formation, properties and impact of secondary organic aerosol: current and emerging issues, Atmos. Chem. Phys., 9, 5155-5236, 2009.

Hao, L., Romakkaniemi, S., Yli-Pirilä, P., Joutsensaari, J., Kortelainen, A., Kroll, J., Miettinen, P., Vaattovaara, P., Tiitta, P., and Jaatinen, A.: Mass yields of secondary organic aerosols from the oxidation of α-pinene and real plant emissions, Atmos. Chem. Phys., 11, 1367-1378, 2011.

Heald, C. L., Kroll, J. H., Jimenez, J. L., Docherty, K. S., DeCarlo, P. F., Aiken, A. C., Chen, Q., Martin, S. T., Farmer, D. K., and Artaxo, P.: A simplified description of the evolution of organic aerosol composition in the atmosphere, Geophys. Res. Lett., 37, L08803, doi: 10.1029/2010GL042737, 2010.

Hellén, H., Dommen, J., Metzger, A., Gascho, A., Duplissy, J., Tritscher, T., Prevot, A. S. H., and Baltensperger, U.: Using proton transfer reaction mass spectrometry for online analysis of secondary organic aerosols, Environ. Sci. Technol., 42, 7347-7353, doi: 10.1021/es801279m, 2008.

Hennigan, C. J., Bergin, M. H., Russell, A. G., Nenes, A., and Weber, R. J.: Gas/particle partitioning of water-soluble organic aerosol in Atlanta, Atmos. Chem. Phys., 9, 3613-3628, 2009.

Heringa, M. F., DeCarlo, P. F., Chirico, R., Tritscher, T., Dommen, J., Weingartner, E., Richter, R., Wehrle, G., Prevot, A. S. H., and Baltensperger, U.: Investigations of primary and secondary particulate matter of different wood combustion appliances with a high-resolution time-of-flight aerosol mass spectrometer, Atmos. Chem. Phys., 11, 5945-5957, doi: 10.5194/acp-11-5945-2011, 2011.

Hildebrandt, L., Henry, K. M., Kroll, J. H., Worsnop, D. R., Pandis, S. N., and Donahue, N. M.: Evaluating the mixing of organic aerosol components using high-resolution aerosol mass spectrometry, Environ. Sci. Technol., 45, 6329-6335, 2011.

Hodzic, A., Jimenez, J. L., Madronich, S., Canagaratna, M. R., DeCarlo, P. F., Kleinman, L., and Fast, J.: Modeling organic aerosols in a megacity: potential contribution of semi-volatile and intermediate volatility primary organic compounds to secondary organic aerosol formation, Atmos. Chem. Phys., 10, 5491-5514, doi: 10.5194/acp-10-5491-2010, 2010.

Hoyle, C. R., Boy, M., Donahue, N. M., Fry, J. L., Glasius, M., Guenther, A., Hallar, A. G., Huff Hartz, K., Petters, M. D., Petäjä, T., Rosenoern, T., and Sullivan, A. P.: A review of the anthropogenic influence on biogenic secondary organic aerosol, Atmos. Chem. Phys., 11, 321-343, 2011.

Huffman, J. A., Docherty, K. S., Mohr, C., Cubison, M. J., Ulbrich, I. M., Ziemann, P. J., Onasch, T. B., and Jimenez, J. L.: Chemically-resolved volatility measurements of organic aerosol from different sources, Environ. Sci. Technol., 43, 5351-5357, 2009.

ICRP: Human respiratory tract model for radiological protection. ICRP Publication 66, Ann. ICRP, 24, 1-3, 1994.

Bibliography 129

Iinuma, Y., Müller, C., Böge, O., Gnauk, T., and Herrmann, H.: The formation of organic sulfate esters in the limonene ozonolysis secondary organic aerosol (SOA) under acidic conditions, Atmos. Environ., 41, 5571-5583, 2007.

IPCC: Summary for policymakers. in: Climate Change 2013: the physical science basis. Contribution of working group I to the fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T.F., D. Qin, G.-K. Plattner, M. Tignor, S.K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, in press., 2013a.

IPCC: Climate Change 2013: the physical science basis. Contribution of working group I to the fifth Assessment Report of the Intergovernmental Panel on Climate Change [Stocker, T. F., D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P. M. Midgley (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, in press. , 2013b.

Jang, M., Czoschke, N. M., Lee, S., and Kamens, R. M.: Heterogeneous atmospheric aerosol production by acid-catalyzed particle-phase reactions, Science, 298, 814-817, 2002.

Jimenez, J. L., Canagaratna, M. R., Donahue, N. M., Prevot, A. S. H., Zhang, Q., Kroll, J. H., DeCarlo, P. F., Allan, J. D., Coe, H., Ng, N. L., Aiken, A. C., Docherty, K. S., Ulbrich, I. M., Grieshop, A. P., Robinson, A. L., Duplissy, J., Smith, J. D., Wilson, K. R., Lanz, V. A., Hueglin, C., Sun, Y. L., Tian, J., Laaksonen, A., Raatikainen, T., Rautiainen, J., Vaattovaara, P., Ehn, M., Kulmala, M., Tomlinson, J. M., Collins, D. R., Cubison, M. J., E, Dunlea, J., Huffman, J. A., Onasch, T. B., Alfarra, M. R., Williams, P. I., Bower, K., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S., Weimer, S., Demerjian, K., Salcedo, D., Cottrell, L., Griffin, R., Takami, A., Miyoshi, T., Hatakeyama, S., Shimono, A., Sun, J. Y., Zhang, Y. M., Dzepina, K., Kimmel, J. R., Sueper, D., Jayne, J. T., Herndon, S. C., Trimborn, A. M., Williams, L. R., Wood, E. C., Middlebrook, A. M., Kolb, C. E., Baltensperger, U., and Worsnop, D. R.: Evolution of organic aerosols in the atmosphere, Science, 326, 1525-1529, doi: 10.1126/science.1180353, 2009.

Jonsson, Å. M., Hallquist, M., and Ljungström, E.: Impact of humidity on the ozone initiated oxidation of limonene, Δ3-carene, and α-pinene, Environ. Sci. Technol., 40, 188-194, doi: 10.1021/es051163w, 2006.

Kalberer, M., Paulsen, D., Sax, M., Steinbacher, M., Dommen, J., Prevot, A. S. H., Fisseha, R., Weingartner, E., Frankevich, V., Zenobi, R., and Baltensperger, U.: Identification of polymers as major components of atmospheric organic aerosols, Science, 303, 1659-1662, doi: 10.1126/science.1092185, 2004.

Kang, E., Root, M. J., Toohey, D. W., and Brune, W. H.: Introducing the concept of Potential Aerosol Mass (PAM), Atmos. Chem. Phys., 7, 5727-5744, doi: 10.5194/acp-7-5727-2007, 2007.

Kang, E., Toohey, D. W., and Brune, W. H.: Dependence of SOA oxidation on organic aerosol mass concentration and OH exposure: experimental PAM chamber studies, Atmos. Chem. Phys., 11, 1837-1852, doi: 10.5194/acp-11-1837-2011, 2011.

Kleindienst, T. E., Edney, E. O., Lewandowski, M., Offenberg, J. H., and Jaoui, M.: Secondary organic carbon and aerosol yields from the irradiations of isoprene and α-pinene in the presence of NOx and SO2, Environ. Sci. Technol., 40, 3807-3812, 2006.

Kreidenweis, S. M., Koehler, K., DeMott, P. J., Prenni, A. J., Carrico, C., and Ervens, B.: Water activity and activation diameters from hygroscopicity data – Part I: Theory and application to inorganic salts, Atmos. Chem. Phys., 5, 1357–1370, 2005.

130 Bibliography

Kroll, J. H., and Seinfeld, J. H.: Chemistry of secondary organic aerosol: Formation and evolution of low-volatility organics in the atmosphere, Atmos. Environ., 42, 3593-3624, 2008.

Laden, F., Schwartz, J., Speizer, F. E., and Dockery, D. W.: Reduction in fine particulate air pollution and mortality: extended follow-up of the Harvard Six Cities study, Am. J. Resp. Crit. Care, 173, 667, 2006.

Lambe, A. T., Onasch, T. B., Massoli, P., Croasdale, D. R., Wright, J. P., Ahern, A. T., Williams, L. R., Worsnop, D. R., Brune, W. H., and Davidovits, P.: Laboratory studies of the chemical composition and cloud condensation nuclei (CCN) activity of secondary organic aerosol (SOA) and oxidized primary organic aerosol (OPOA), Atmos. Chem. Phys., 11, 8913-8928, doi: 10.5194/acp-11-8913-2011, 2011.

Lanz, V. A., Alfarra, M. R., Baltensperger, U., Buchmann, B., Hueglin, C., and Prevot, A. S. H.: Source apportionment of submicron organic aerosols at an urban site by factor analytical modelling of aerosol mass spectra, Atmos. Chem. Phys., 7, 1503-1522, 2007.

Lanz, V. A., Prevot, A. S. H., Alfarra, M. R., Weimer, S., Mohr, C., DeCarlo, P. F., Gianini, M. F. D., Hueglin, C., Schneider, J., Favez, O., D'Anna, B., George, C., and Baltensperger, U.: Characterization of aerosol chemical composition with aerosol mass spectrometry in Central Europe: an overview, Atmos. Chem. Phys., 10, 10453-10471, doi: 10.5194/acp-10-10453-2010, 2010.

Lepeule, J., Laden, F., Dockery, D., and Schwartz, J.: Chronic exposure to fine particles and mortality: an extended follow-up of the Harvard Six Cities Study from 1974 to 2009, Environ. Health Perspect., 120, 965, 2012.

Lindinger, W., Hansel, A., and Jordan, A.: On-line monitoring of volatile organic compounds at pptv levels by means of proton-transfer-reaction mass spectrometry (PTR-MS) medical applications, food control and environmental research, Int. J. Mass Spectrom. Ion Processes, 173, 191-241, doi: 10.1016/S0168-1176(97)00281-4, 1998.

Martin, S. T.: Phase transitions of aqueous atmospheric particles, Chem. Rev. (Washington, DC, U. S.), 100, 3403-3454, 2000.

Massoli, P., Lambe, A. T., Ahern, A. T., Williams, L. R., Ehn, M., Mikkilä, J., Canagaratna, M. R., Brune, W. H., Onasch, T. B., Jayne, J. T., Petäjä, T., Kulmala, M., Laaksonen, A., Kolb, C. E., Davidovits, P., and Worsnop, D. R.: Relationship between aerosol oxidation level and hygroscopic properties of laboratory generated secondary organic aerosol (SOA) particles, Geophys. Res. Lett., 37, L24801, doi: 10.1029/2010GL045258, 2010.

Maynard, A. D., and Kuempel, E. D.: Airborne nanostructured particles and occupational health, J. Nanopart. Res., 7, 587-614, 2005.

Metzger, A., Dommen, J., Gaeggeler, K., Duplissy, J., Prevot, A. S. H., Kleffmann, J., Elshorbany, Y., Wisthaler, A., and Baltensperger, U.: Evaluation of 1,3,5 trimethylbenzene degradation in the detailed tropospheric chemistry mechanism, MCMv3.1, using environmental chamber data, Atmos. Chem. Phys., 8, 6453-6468, doi: 10.5194/acp-8-6453-2008, 2008.

Metzger, A., Verheggen, B., Dommen, J., Duplissy, J., Prevot, A. S. H., Weingartner, E., Riipinen, I., Kulmala, M., Spracklen, D. V., Carslaw, K. S., and Baltensperger, U.: Evidence for the role of organics in aerosol particle formation under atmospheric conditions, Proc. Natl. Acad. Sci. U. S. A., 107, 6646-6651, doi: 10.1073/pnas.0911330107, 2010.

Bibliography 131

Monge, M. E., Rosenørn, T., Favez, O., Müller, M., Adler, G., Abo Riziq, A., Rudich, Y., Herrmann, H., George, C., and D’Anna, B.: Alternative pathway for atmospheric particles growth, Proc. Natl. Acad. Sci. U. S. A., 109, 6840-6844, doi: 10.1073/pnas.1120593109, 2012.

Nel, A.: Air pollution-related illness: effects of particles, Science, 308, 804-806, doi: 10.1126/science.1108752, 2005.

Ng, N. L., Chhabra, P. S., Chan, A. W. H., Surratt, J. D., Kroll, J. H., Kwan, A. J., McCabe, D. C., Wennberg, P. O., Sorooshian, A., Murphy, S. M., Dalleska, N. F., Flagan, R. C., and Seinfeld, J. H.: Effect of NOx level on secondary organic aerosol (SOA) formation from the photooxidation of terpenes, Atmos. Chem. Phys., 7, 5159-5174, 2007.

Ng, N. L., Canagaratna, M. R., Zhang, Q., Jimenez, J. L., Tian, J., Ulbrich, I. M., Kroll, J. H., Docherty, K. S., Chhabra, P. S., Bahreini, R., Murphy, S. M., Seinfeld, J. H., Hildebrandt, L., Donahue, N. M., DeCarlo, P. F., Lanz, V. A., Prévôt, A. S. H., Dinar, E., Rudich, Y., and Worsnop, D. R.: Organic aerosol components observed in northern hemispheric datasets from aerosol mass spectrometry, Atmos. Chem. Phys., 10, 4625-4641, doi: 10.5194/acp-10-4625-2010, 2010.

Ng, N. L., Canagaratna, M. R., Jimenez, J. L., Chhabra, P. S., Seinfeld, J. H., and Worsnop, D. R.: Changes in organic aerosol composition with aging inferred from aerosol mass spectra, Atmos. Chem. Phys., 11, 6465-6474, doi: 10.5194/acp-11-6465-2011, 2011a.

Ng, N. L., Canagaratna, M. R., Jimenez, J. L., Zhang, Q., Ulbrich, I. M., and Worsnop, D. R.: Real-time methods for estimating organic component mass concentrations from aerosol mass spectrometer data, Environ. Sci. Technol., 45, 910-916, 2011b.

NIST: P.J. Linstrom and W.G. Mallard, eds.,NIST chemistry webbook, NIST standard reference database number 69, National Institute of Standards and Technology, Gaithersburg MD, 20899, available at: http://webbook.nist.gov, (retrieved January 5, 2014).

Oberdörster, G., and Utell, M. J.: Ultrafine particles in the urban air: to the respiratory tract--and beyond?, Environ. Health Perspect., 110, A440, 2002.

Odum, J. R., Hoffmann, T., Bowman, F., Collins, D., Flagan, R. C., and Seinfeld, J. H.: Gas/particle partitioning and secondary organic aerosol yields, Environ. Sci. Technol., 30, 2580-2585, doi: 10.1021/es950943+, 1996.

Paatero, P.: The multilinear engine - A table-driven, least squares program for solving multilinear problems, including the n-way parallel factor analysis model, J. Comput. Graph. Stat., 8, 854-888, doi: 10.2307/1390831, 1999.

Pankow, J. F.: Review and comparative analysis of the theories on partitioning between the gas and aerosol particulate phases in the atmosphere, Atmos. Environ. (1967), 21, 2275-2283, 1987.

Pankow, J. F.: An absorption model of the gas/aerosol partitioning involved in the formation of secondary organic aerosol, Atmos. Environ., 28, 189-193, 1994.

Pankow, J. F., Seinfeld, J. H., Asher, W. E., and Erdakos, G. B.: Modeling the formation of secondary organic aerosol. 1. Application of theoretical principles to measurements obtained in the α-pinene/, β-pinene/, sabinene/, Δ3-carene/, and cyclohexene/ozone systems, Environ. Sci. Technol., 35, 1164-1172, 2001.

Pankow, J. F.: Organic particulate material levels in the atmosphere: Conditions favoring sensitivity to varying relative humidity and temperature, Proc. Natl. Acad. Sci. U. S. A., 107, 6682-6686, doi: 10.1073/pnas.1001043107, 2010.

132 Bibliography

Pankow, J. F.: On the ability of the gas/particle partitioning constant Kp to consider the effects of mean MW and the presence of high MW compounds, Atmos. Environ., 45, 1213-1216, doi: 10.1016/j.atmosenv.2010.11.041, 2011.

Paulsen, D., Dommen, J., Kalberer, M., Prévôt, A. S. H., Richter, R., Sax, M., Steinbacher, M., Weingartner, E., and Baltensperger, U.: Secondary organic aerosol formation by irradiation of 1,3,5-trimethylbenzene-NOx-H2O in a new reaction chamber for atmospheric chemistry and physics, Environ. Sci. Technol., 39, 2668-2678, 2005.

Pfaffenberger, L., Barmet, P., Slowik, J. G., Praplan, A. P., Dommen, J., Prévôt, A. S. H., and Baltensperger, U.: The link between organic aerosol mass loading and degree of oxygenation: An α-pinene photooxidation study, Atmos. Chem. Phys., 13, 6493-6506, doi: 10.5194/acp-13-6493-2013, 2013.

Platt, S. M., El Haddad, I., Zardini, A. A., Clairotte, M., Astorga, C., Wolf, R., Slowik, J. G., Temime-Roussel, B., Marchand, N., Ježek, I., Drinovec, L., Močnik, G., Möhler, O., Richter, R., Barmet, P., Bianchi, F., Baltensperger, U., and Prévôt, A. S. H.: Secondary organic aerosol formation from gasoline vehicle emissions in a new mobile environmental reaction chamber, Atmos. Chem. Phys., 13, 9141-9158, doi: 10.5194/acp-13-9141-2013, 2013.

Pope III, C. A., Ezzati, M., and Dockery, D. W.: Fine-particulate air pollution and life expectancy in the United States, N. Engl. J. Med., 360, 376-386, 2009.

Presto, A. A., Huff Hartz, K. E., and Donahue, N. M.: Secondary organic aerosol production from terpene ozonolysis. 2. Effect of NOx concentration, Environ. Sci. Technol., 39, 7046-7054, doi: 10.1021/es050400s, 2005.

Raes, F., Dingenen, R. V., Vignati, E., Wilson, J., Putaud, J.-P., Seinfeld, J. H., and Adams, P.: Formation and cycling of aerosols in the global troposphere, Atmos. Environ., 34, 4215-4240, 2000.

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

Seinfeld, J. H., and Pandis, S. N.: Atmospheric chemistry and physics : from air pollution to climate change (2nd ed.), John Wiley & Sons, 2006.

Shilling, J. E., Chen, Q., King, S. M., Rosenoern, T., Kroll, J. H., Worsnop, D. R., DeCarlo, P. F., Aiken, A. C., Sueper, D., Jimenez, J. L., and Martin, S. T.: Loading-dependent elemental composition of alpha-pinene SOA particles, Atmos. Chem. Phys., 9, 771-782, doi: 10.5194/acp-9-771-2009, 2009.

Slowik, J. G., Wong, J. P. S., and Abbatt, J. P. D.: Real-time, controlled OH-initiated oxidation of biogenic secondary organic aerosol, Atmos. Chem. Phys., 12, 9775-9790, doi: 10.5194/acp-12-9775-2012, 2012.

Steinbrecher, R., Smiatek, G., Köble, R., Seufert, G., Theloke, J., Hauff, K., Ciccioli, P., Vautard, R., and Curci, G.: Intra- and inter-annual variability of VOC emissions from natural and semi-natural vegetation in Europe and neighbouring countries, Atmos. Environ., 43, 1380-1391, doi: 10.1016/j.atmosenv.2008.09.072, 2009.

Stemmler, K., Ammann, M., Donders, C., Kleffmann, J., and George, C.: Photosensitized reduction of nitrogen dioxide on humic acid as a source of nitrous acid, Nature, 440, 195-198, 2006.

Bibliography 133

Stokes, R., and Robinson, R.: Interactions in aqueous nonelectrolyte solutions. I. Solute-solvent equilibria, J. Phys. Chem., 70, 2126-2131, 1966.

Szidat, S., Jenk, T. M., Synal, H. A., Kalberer, M., Wacker, L., Hajdas, I., Kasper-Giebl, A., and Baltensperger, U.: Contributions of fossil fuel, biomass-burning, and biogenic emissions to carbonaceous aerosols in Zurich as traced by 14C, J. Geophys. Res., 111, 2006.

Taira, M., and Kanda, Y.: Continuous generation system for low-concentration gaseous nitrous acid, Anal. Chem., 62, 630-633, 1990.

Thornton, J., Wooldridge, P., Cohen, R., Martinez, M., Harder, H., Brune, W., Williams, E., Roberts, J., Fehsenfeld, F., and Hall, S.: Ozone production rates as a function of NOx abundances and HOx production rates in the Nashville urban plume, J. Geophys. Res., 107, 4146, 2002.

Topping, D. O., McFiggans, G. B., and Coe, H.: A curved multi-component aerosol hygroscopicity model framework: Part 1 – Inorganic compounds, Atmos. Chem. Phys., 5, 1205–1222, 2005.

Ulbrich, I. M., Canagaratna, M. R., Zhang, Q., Worsnop, D. R., and Jimenez, J. L.: Interpretation of organic components from positive matrix factorization of aerosol mass spectrometric data, Atmos. Chem. Phys., 9, 2891-2918, doi: 10.5194/acp-9-2891-2009, 2009.

Valorso, R., Aumont, B., Camredon, M., Raventos-Duran, T., Mouchel-Vallon, C., Ng, N. L., Seinfeld, J. H., Lee-Taylor, J., and Madronich, S.: Explicit modelling of SOA formation from α-pinene photooxidation: sensitivity to vapour pressure estimation, Atmos. Chem. Phys., 11, 6895-6910, doi: 10.5194/acp-11-6895-2011, 2011.

Van Krevelen, D. W.: Graphical-statistical method for the study of structure and reaction processes of coal, Fuel, 24, 269–284, 1950.

Volkamer, R., Jimenez, J. L., San Martini, F., Dzepina, K., Zhang, Q., Salcedo, D., Molina, L. T., Worsnop, D. R., and Molina, M. J.: Secondary organic aerosol formation from anthropogenic air pollution: Rapid and higher than expected, Geophys. Res. Lett., 33, L17811, doi: 10.1029/2006GL026899, 2006.

Williams, L., Gonzalez, L., Peck, J., Trimborn, D., McInnis, J., Farrar, M., Moore, K., Jayne, J., Robinson, W., and Lewis, D.: Characterization of an aerodynamic lens for transmitting particles greater than 1 micrometer in diameter into the Aerodyne aerosol mass spectrometer, Atmos. Meas. Tech., 6, 3271-3280, 2013.

You, Y., Renbaum-Wolff, L., and Bertram, A. K.: Liquid–liquid phase separation in particles containing organics mixed with ammonium sulfate, ammonium bisulfate, ammonium nitrate or sodium chloride, Atmos. Chem. Phys., 13, 11723-11734, doi: 10.5194/acp-13-11723-2013, 2013.

Zhang, X., Cappa, C. D., Jathar, S. H., McVay, R. C., Ensberg, J. J., Kleeman, M. J., and Seinfeld, J. H.: Influence of vapor wall loss in laboratory chambers on yields of secondary organic aerosol, Proceedings of the National Academy of Sciences, 111, 5802-5807, 10.1073/pnas.1404727111, 2014.

Zuend, A., and Seinfeld, J. H.: A practical method for the calculation of liquid–liquid equilibria in multicomponent organic–water–electrolyte systems using physicochemical constraints, Fluid Phase Equilib., 337, 201-213, doi: 10.1016/j.fluid.2012.09.034, 2013.

___________________________________________________________________________

Acknowledgements

___________________________________________________________________________

I would like to thank...

... my doctoral advisor Prof. Urs Baltensperger for giving me the opportunity to do high-level

research in the Laboratory of Atmospheric Chemistry at PSI. With your constructive ideas

during Monday seminars, Urs and smog chamber meetings the work was always brought a

step forward. You showed interest and always gave very good input. Also, your effort to hold

the group together with LAC events made it nice to be part of the LAC besides the lab.

... Prof. Tom Peter who gave his support from the side of the Institute of Atmospheric and

Climate Science at ETH. The discussions during ETH interviews with questions from a more

distant view on the topic were inspiring.

... Dr. Harald Saathoff for taking the effort of being my external co-referee, coming to Zürich

and to review the Thesis. Thanks!

... Dr. André Prevôt who brought me to the LAC even without real project plan which allowed

us to flexibly follow promising research directions. Your input during scientific discussions

helped proceeding. Your openness and collegiality built a familiar environment.

... Dr. Josef Dommen to always ask critical questions which increased research quality by

much. ... Peter B. to go through all the effort during smog chamber experiments with me. The

two-way support was exceptional. Also the support of Arnaud, Manu, Carla and Emily

facilitated the experiments. ... Martin Gysel and Michel J. Rossi for fruitful discussions.

...Rene and Günther for support with the smog chamber hard ware ... Maarten, Claudia,

Torsten and Marie for introducing me to the LAC soft- and hard(ware) and much more. ...

Imad and Stephen who are inspiring discussion partners. ... the corridor-group: Kathrin,

Monica, Francesco, Robert, Kaspar and Suzanne for their collegiality. ... the LAC-band

through which I started singing again! ... Pete Sampras, Börni (sorry), Fede and Emmi for

being just there. A special thank you goes to Dani and the rest of my family for being the

greatest supporters!

___________________________________________________________________________

Curriculum vitae

___________________________________________________________________________

Lisa Pfaffenberger, born on 29th of January 1985 in Bad Tölz, Germany

Education

04/2010 – 03/2014 PhD, Laboratory of Atmospheric Chemistry, Paul Scherrer Institute,

Villigen, Switzerland

10/2004 – 10/2009 Studies of Meteorology (Diploma, Ludwig-Maximilians-University

Munich, Germany and Universitetet i Bergen, Norway

Diploma Thesis, Deutsches Zentrum für Luft- und Raumfahrt,

Oberpfaffenhofen, Germany

Topic: “Particle emissions from shipping and their impact on the

aerosol of the marine boundary layer – Results of the QUANTIFY

SHIPS Study 2007“

09/1995 – 05/2004 High School, Gymnasium Unterhaching, Germany, Abitur (A-Level)

Platform Presentations

09/2013 European Aerosol Conference, Prague, Czech Republic

“Dependence of α-pinene secondary organic aerosol formation on relative

humidity and aerosol surface distribution”

09/2012 European Aerosol Conference, Granada, Spain

“Is the chemical composition of α-pinene secondary organic aerosol dependent

on particulate water content?”

10/2011 American Association For Aerosol Research Conference, Orlando, USA

“How to produce low-volatility oxygenated organic aerosol“

09/2009 European Aerosol Conference, Karlsruhe, Germany

“From fresh exhaust to CCN: Transformation of particles from shipping“

136 Curriculum vitae

Peer reviewed publications

Two-stroke scooters are a dominant source of air pollution in many cities

Platt, S. M., El Haddad, I., Pieber, S. M., Huang, R.-J., Zardini, A. A., Clairotte, M., Suarez-Bertoa,

R., Barmet, P., Pfaffenberger, L., Wolf, R., Slowik, J. G., Fuller, S. J., Kalberer, M., Chirico, R.,

Dommen, J., Astorga, C., Zimmermann, R., Marchand, N., Hellebust, S., Temime-Roussel, B.,

Baltensperger U. and Prévôt, A. S. H.

Nature Communications, 5, 3749, 2014, doi:10.1038/ncomms4749

Modeling the evolution of aerosol particles in a ship plume using PartMC-MOSAIC

Tian, J., Riemer, N., West, M., Pfaffenberger, L., Schlager, H., and Petzold, A., 2014

Atmos. Chem. Phys., 14, 5327-5347, 2014, doi:10.5194/acp-14-5327-2014

Online measurements of water-soluble organic acids in the gas and aerosol phase from the

photooxidation of 1,3,5-trimethylbenzene

Praplan A. P., Hegyi-Gaeggeler K., Barmet P., Pfaffenberger L., Dommen J. and Baltensperger U.

Atmos. Chem. Phys., 14, 8665-8677, 2014, doi:10.5194/acp-14-8665-2014

The link between organic aerosol mass loading and degree of oxygenation: An α-pinene

photooxidation study

Pfaffenberger, L., Barmet, P., Slowik, J. G., Praplan, A. P., Dommen, J., Prévôt, A. S. H., and

Baltensperger, U., 2013

Atmos. Chem. Phys., 13, 6493-6506, doi:10.5194/acp-13-6493-2013

Effective Henry’s Law Partitioning and the Salting Constant of Glyoxal in Aerosols Containing

Sulfate

Kampf, C. J., Waxman, E. M., Slowik, J. G., Dommen, J., Pfaffenberger, L., Praplan, A. P., Prévôt,

A. S. H., Baltensperger, U., Hoffmann, T. and Volkamer, R., 2013

Environ. Sci. Technol., 47, 4236-4244, doi:10.1021/es400083d

Similarities in STXM-NEXAFS Spectra of Atmospheric Particles and Secondary Organic

Aerosol Generated from Glyoxal, α-Pinene, Isoprene, 1,2,4-Trimethylbenzene, and d-Limonene

Shakya, K. M., Liu, S., Takahama, S., Russell, L. M., Keutsch, F. N., Galloway, M. M., Shilling, J. E.,

Hiranuma, N., Song, C., Kim, H., Paulson, S. E., Pfaffenberger, L., Barmet, P., Slowik, J., Prevot A.

S. H., Dommen, J. and Baltensperger, U., 2013

Aerosol Sci. Technol., 47, 543-555, doi:10.1080/02786826.2013.772950

Curriculum vitae 137

Presenting SAPUSS: Solving Aerosol Problem by Using Synergistic Strategies in Barcelona,

Spain

Dall'Osto, M., Querol, X., Alastuey, A., Minguillon, M. C., Alier, M., Amato, F., Brines, M., Cusack,

M., Grimalt, J. O., Karanasiou, A., Moreno, T., Pandolfi, M., Pey, J., Reche, C., Ripoll, A., Tauler, R.,

Van Drooge, B. L., Viana, M., Harrison, R. M., Gietl, J., Beddows, D., Bloss, W., O'Dowd, C.,

Ceburnis, D., Martucci, G., Ng, N. L., Worsnop, D., Wenger, J., Mc Gillicuddy, E., Sodeau, J., Healy,

R., Lucarelli, F., Nava, S., Jimenez, J. L., Gomez Moreno, F., Artinano, B., Prévôt, A. S. H.,

Pfaffenberger, L., Frey, S., Wilsenack, F., Casabona, D., Jiménez-Guerrero, P., Gross, D., and Cots,

N., 2013

Atmos. Chem. Phys.,13, 8991-9019, doi:10.5194/acp-13-8991-2013

Development of a sensitive long path absorption photometer to quantify peroxides in aerosol

particles (Peroxide-LOPAP)

Mertes, P., Pfaffenberger, L., Dommen, J., Kalberer, M., and Baltensperger, U., 2012

Atmos. Meas. Tech., 5, 2339-2348, doi:10.5194/amt-5-2339-2012

A new method to discriminate secondary organic aerosols from different sources using high-

resolution aerosol mass spectra

Heringa, M. F., DeCarlo, P. F., Chirico, R., Tritscher, T., Clairotte, M., Mohr, C., Crippa, M., Slowik,

J. G., Pfaffenberger, L., Dommen, J., Weingartner, E., Prévôt, A. S. H., and Baltensperger, U., 2012

Atmos. Chem. Phys., 12, 2189-2203, doi:10.5194/acp-12-2189-2012

___________________________________________

Villigen, September 12, 2014