Cloud droplet activity changes of soot aerosol upon smog ......condensation nuclei (CCN) properties at the onset of UV ra-diation implies that the lifetime of soot particles in the

Atmos Chem Phys 14 9831ndash9854 2014wwwatmos-chem-physnet1498312014doi105194acp-14-9831-2014copy Author(s) 2014 CC Attribution 30 License

Cloud droplet activity changes of soot aerosol upon smog chamberageing

C Wittbom 1 A C Eriksson1 J Rissler2 J E Carlsson2 P Roldin13 E Z Nordin2 P T Nilsson2 E Swietlicki1J H Pagels2 and B Svenningsson1

1Department of Physics Lund University PO Box 118 221 00 Lund Sweden2Ergonomics and Aerosol Technology Lund University PO Box 118 221 00 Lund Sweden3Department of Physics PO Box 48 University of Helsinki 00014 Helsinki Finland

Correspondence toC Wittbom (cerinawittbomnuclearluse)

Received 14 February 2014 ndash Published in Atmos Chem Phys Discuss 2 April 2014Revised 9 August 2014 ndash Accepted 12 August 2014 ndash Published 17 September 2014

Abstract Particles containing soot or black carbon are gen-erally considered to contribute to global warming Howeverlarge uncertainties remain in the net climate forcing result-ing from anthropogenic emissions of black carbon (BC) to alarge extent due to the fact that BC is co-emitted with gasesand primary particles both organic and inorganic and sub-ject to atmospheric ageing processes In this study dieselexhaust particles and particles from a flame soot genera-tor spiked with light aromatic secondary organic aerosol(SOA) precursors were processed by UV radiation in a 6 m3

Teflon chamber in the presence of NOx The time-dependentchanges of the soot nanoparticle properties were charac-terised using a Cloud Condensation Nuclei Counter anAerosol Particle Mass Analyzer and a Soot Particle AerosolMass Spectrometer The results show that freshly emittedsoot particles do not activate into cloud droplets at supersat-urationsle 2 ie the BC core coated with primary organicaerosol (POA) from the exhaust is limited in hygroscopicityBefore the onset of UV radiation it is unlikely that any sub-stantial SOA formation is taking place An immediate changein cloud-activation properties occurs at the onset of UV ex-posure This change in hygroscopicity is likely attributed toSOA formed from intermediate volatility organic compounds(IVOCs) in the diesel engine exhaust The change of cloudcondensation nuclei (CCN) properties at the onset of UV ra-diation implies that the lifetime of soot particles in the at-mosphere is affected by the access to sunlight which dif-fers between latitudes The ageing of soot particles progres-sively enhances their ability to act as cloud condensation nu-clei due to changes in (I) organic fraction of the particle

(II) chemical properties of this fraction (eg primary or sec-ondary organic aerosol) (III) particle size and (IV) particlemorphology Applyingκ-Koumlhler theory using aκSOA valueof 013 (derived from independent input parameters describ-ing the organic material) showed good agreement with clouddroplet activation measurements for particles with a SOAmass fractionge 012 (slightly aged particles) The activa-tion properties are enhanced with only a slight increase in or-ganic material coating the soot particles (SOA mass fractionlt 012) however not as much as predicted by Koumlhler theoryThe discrepancy between theory and experiments during theearly stages of ageing might be due to solubility limitationsunevenly distributed organic material or hindering particlemorphology

The change in properties of soot nanoparticles upon pho-tochemical processing clearly increases their hygroscopicitywhich affects their behaviour both in the atmosphere and inthe human respiratory system

1 Introduction

Atmospheric aerosols are known to have a significant effecton visibility climate and human health Aerosol particles in-fluence the climate and hydrological cycle of Earth by act-ing as cloud condensation nuclei (CCN) referred to as theindirect aerosol effect (IPCC 2007) or the effective radia-tive forcing from aerosolndashcloud interactions (ERFaci IPCC2013) The ability of aerosol particles to act as CCN de-pends on the particle size and chemical composition as well

Published by Copernicus Publications on behalf of the European Geosciences Union

9832 C Wittbom et al Cloud droplet activity changes of soot aerosol upon smog chamber ageing

as the ambient water vapour supersaturation The indirectaerosol effect includes increase in cloud albedo due to ad-dition of cloud nuclei by pollution (Twomey 1974) reduc-tion of drizzle and increased cloud lifetime (Albrecht 1989)as well as an increase in cloud thickness (Pincus and Baker1994) Also soot aerosol can contribute to daytime clearingof clouds (Ackerman et al 2000)

Soot particles make up a large fraction by number of theatmospheric aerosol especially in urban locations in the sizerange lt 100 nm (Rose et al 2006 and references therein)Freshly emitted soot particles or black carbon (BC) areknown to have a predominantly warming effect on the cli-mate due to their ability to absorb light referred to as a di-rect aerosol effect (IPCC 2007) or the radiative forcing fromaerosolndashradiation interactions (RFari IPCC 2013) Also theabsorption may increase with photochemical ageing and wa-ter uptake (Zhang et al 2008 Cappa et al 2012) AfterCO2 BC is estimated as the strongest anthropogenic climate-forcing agent in the present-day atmosphere (Bond et al2013) with a total warming effect of about+11 W mminus2

(with 90 uncertainty bounds of+017 to 21 W mminus2)However BC also has an effect on the Earthrsquos hydrologicalcycle According to Bond et al (2013) the largest uncertain-ties in net climate forcing are related to the lack of knowl-edge about cloud interactions with BC when co-emitted withorganic carbon (OC)

Soot is not only present close to sources but can be trans-ported long distances The warming effect of soot in the Arc-tic has recently gained increased interest both due to trans-port of soot from urban areas and also due to possible fu-ture soot emissions in the Arctic area from ship traffic whichis made possible due to the reduction in ice coverage Theenhanced light absorption by soot due to condensing mate-rial during photochemical ageing may influence and boostthe warming of the surroundings even further Also soot de-posited on snow and ice may enhance surface heating andice melting by decreasing the surface albedo (eg Bond etal 2013 Tunved et al 2013 and references therein) Theparticle number concentration is in general low (about a cou-ple of hundred cmminus3) in the Arctic (Tunved et al 2013)Wet removal and photochemical ageing play central roles incontrolling the aerosol size distribution properties during theArctic summer As pointed out by Tunved et al (2013) it ishighly relevant to obtain a better understanding of the micro-physical properties of the different aerosol populations abun-dant in the Arctic environment in order to improve the under-standing of the direction and magnitude of a future climatechange in the region

According to the World Health Organization (WHO2012) emissions of diesel engine exhaust are classified ascarcinogenic in humans in addition to being linked withclimate change The unfavourable health effects caused byexposure to diesel exhaust are described in several studies(eg Sydbom et al 2001 Mills et al 2007 Hart et al2009) The deposited fraction in the human respiratory sys-

tem is well described by the mobility diameter (for particleslt 400 nm) whilst deposited dose by surface area and mass re-quire knowledge of the characteristics of the particles due totheir agglomerated structure (Rissler et al 2012) The parti-cle lung deposition is substantially altered by hygroscopicity(Loumlndahl et al 2009) A change towards more hygroscopicproperties will shift the deposited fraction in the respiratorytract towards larger sizes Photochemical processing of dieselexhaust particles thereby alters the uptake in the human res-piratory system due to enhanced hygroscopicity

Diesel exhaust aerosol is formed during combustion pro-cesses mainly of refractory carbonaceous material (BC) thatis highly agglomerated primary organic compounds in theparticle phase and volatile organic compounds (VOCs) in thegas phase There is no clear definition for soot formed fromincomplete combustion however in general soot consists ofroughly eight parts of carbon and one part of hydrogen byatoms (Tree and Svensson 2007) In this study the term sootparticles refers to the agglomerated particles emitted fromdiesel vehicles or by a soot generator including both BC andorganic carbon (Petzold et al 2013) Diesel exhaust particles(DEPs) and soot generator particles (FSPs) refer to particlesfrom the diesel vehicle and the flame soot generator respec-tively

Freshly emitted soot particles are hydrophobic or limitedin hygroscopicity and unlikely to contribute to the CCN pop-ulation in the atmosphere (Weingartner et al 1997 Meyerand Ristovski 2007 Zhang et al 2008 Tritscher et al2011) The size of these non-spherical particles will nor-mally differ between measurement techniques (DeCarlo etal 2004) for example showing a larger mobility thanvolume-equivalent diameter However as the soot particlesreside in the atmosphere they undergo physical and chem-ical changes during UV exposure ndash they age The agglom-erated soot particles are exposed to chemical gas-to-particleprocesses in the atmosphere resulting in condensation of or-ganic and inorganic (Rose et al 2006) vapours and coagu-lation of particles onto the agglomerates Due to this ageingprocess the soot particles collapse into a more compact struc-ture (Weingartner et al 1997 Pagels et al 2009 Tritscher etal 2011) Hygroscopicity is enhanced with increasing age-ing time affected by high sulfur content in the fuel highVOC levels in the emissions pre-treatment of the exhaustgas with ozone and UV radiation (eg Weingartner et al1997 Tritscher et al 2011) Weingartner et al (1997) pro-posed that diesel exhaust aerosol which was pre-treated withO3 and then subjected to UV radiation would show an en-hancement in hygroscopicity However the large scatter indata made it impossible to draw conclusions Soot agglomer-ates become more hygroscopic when coated partly or fullyby organic or inorganic material The coating material trans-forms the agglomerates to enable them to act as CCN

In the atmosphere in general aerosol particles are com-posed of both organic and inorganic compounds Organicmaterial makes up a significant fraction (20 to 90 ) of the

Atmos Chem Phys 14 9831ndash9854 2014 wwwatmos-chem-physnet1498312014

C Wittbom et al Cloud droplet activity changes of soot aerosol upon smog chamber ageing 9833

submicrometre aerosol mass in many locations (Kanakidouet al 2005 Jimenez et al 2009) Organic aerosol is ei-ther introduced into the atmosphere from primary sources(ie primary organic aerosol POA) or formed in the atmo-sphere via complex gasndashparticle conversion processes form-ing secondary organic aerosol (SOA) SOA can be formedwhen VOCs either from biogenic or anthropogenic sourcesare present Oxidation products from the VOCs condense andproduce SOA The condensation of oxidised VOCs alters theparticle properties which may lead to an increased abilityof the particle to act as CCN ndash though organic compoundswill typically not improve the water activity as much as in-organic salts Previous studies of biogenic SOA have shownthat aerosol yield and hygroscopicity of SOA are dependenton experimental conditions (eg Asa-Awuku et al 2009Frosch et al 2013) For example hygroscopicity stronglydepended on exposure to OH with less hygroscopic parti-cles in the presence of an OH scavenger than in experimentswith OH present

Koumlhler theory (Koumlhler 1936) is commonly used to pre-dict and describe the ability of the particles to activate intocloud droplets (further described in Sect 3) However theactivation of particles in complex mixtures in ambient air isproblematic to explain in theory (eg Laaksonen et al 1998Svenningsson et al 2006) Previous work regarding the ac-tivation properties and processes of organic aerosols in reac-tion chambers often studied SOA from biogenic precursorssuch as limoneneα-pinene andβ-caryophyllene ndash Prenni etal 2007 (α-pineneβ-pinene13-carene and toluene) Du-plissy et al 2008 (α-pinene) Kundu et al 2012 (limonene)Frosch et al 2013 (β-caryophyllene)

Rissler et al (2006) first introduced the hygroscopicity pa-rameterκ (H-TDMA-derived) describing the number of ionsor non-dissociating molecules per unit volume of the dry par-ticle A very similar parameter also denotedκ was later in-troduced by Petters and Kreidenweis (2007) Thisκ param-eter is in principle the same as theκ introduced by Rissleret al (2006) differing only due to different choice of units(Rissler et al 2010) Since the one introduced by Petters andKreidenweis (2007) is more broadly used and reported in theliterature this one will be reported here and is the one re-ferred to from now on This hygroscopicity parameter rangesfrom 05 to 14 for salts such as NaCl that are highly CCN-active Slightly to very hygroscopic organic species haveκ

values between 001 and 05 and non-hygroscopic compo-nents have aκ = 0 (such as untreated soot particles eg Hen-ning et al 2012)κ has previously been reported to be in therange 0ndash013 (apparentκ) for photochemically aged dieselsoot and for SOA from pure gas phase of the diesel vehiclein the range 009ndash014 (Tritscher et al 2011) These valuescould be compared toκ values for aged biogenic SOA forβ-caryophyllene SOAκ is 0002ndash016 (eg Huff Hartz et al2005 Asa-Awuku et al 2009 Frosch et al 2013) and forα-pineneκ = 01 (eg Prenni et al 2007 Duplissy et al2008) Dusek et al (2006) calculatedκ values for air masses

in the range 015ndash030 originating from four different placesbut arriving in the same area Not many studies have beenperformed concerning how the ageing process affects the ac-tivation properties of soot particles into cloud droplets andthe few previous studies performed have not been able to cap-ture the rapid change of the particle properties and improvedactivation (eg Tritscher et al 2011)

The aim of this study is to experimentally examine thechange in cloud activation properties of photochemicallyprocessed soot and evaluate the results usingκ-Koumlhler mod-elling We present results from scanning flow CCN analysis(SFCA) enabling high temporal and supersaturation resolu-tion Linked to the activation properties are results from on-line measurements of the chemical composition of the par-ticles (soot core and organic coating) evaluated through on-line mass spectrometry Also used in the evaluation is theknowledge about the change in shape and morphology of theparticles determined from the particle massndashmobility rela-tionship The change in particle properties (chemical compo-sition shape and morphology) during ageing has also beenused for modelling the critical supersaturation (sc) of thecoated soot particles

2 Experimental

Ageing experiments of soot aerosols and precursors werecarried out in the smog chamber in the aerosol laboratoryat Lund University (LU) In total six experiments were eval-uated and presented in this study chosen by coverage in dataThe photochemical ageing was induced using black lights(peak at 354 nm) in a 6 m3 TeflonFEP bag The experimen-tal set-up is described elsewhere (Nordin et al 2013) Anoverview of the experiments performed is given in Table 1

Here we present the results from experiments where twosources of primary aerosol were used soot agglomeratesfrom (1) a Euro II Diesel Passenger Vehicle (VW Passat1998) and (2) a diffusion flame soot generator described indetail elsewhere (Malik et al 2011) The soot nanoparti-cles and gases from the vehicle were transferred from thetailpipe via a heated inlet system using an ejector diluter (DI-1000 Dekati Ltd Finland) with a modified inlet nozzle toachieve a primary dilution ratio of 45 into the initially cleansmog chamber (particle number and volume concentrationslt 100 cmminus3 and lt 001 microg mminus3 respectively) For the pres-surised air supplying the ejector and all other pressurised airapplications in the chamber a special filter configuration wasused In short the pressurised air was preheated to 140Cand passed through multiple filter sets to remove particles or-ganic acids NOx SOx and O3 Total dilution ratios ofsim 350for the diesel vehicle and 50ndash100 for the flame soot gener-ator were finally achieved in the Teflon chamber dependingon the injection time of the exhausts This corresponds tonumber concentrations of 6000ndash12 000 cmminus3 and mass con-centrations of 3ndash14 microg mminus3

wwwatmos-chem-physnet1498312014 Atmos Chem Phys 14 9831ndash9854 2014


Table 1Experimental details of the dry particle mobility diameter (dmdry) the primary particle diameter (dpp) the critical supersaturationmeasured by the two different CCNC instruments (sc 52 andsc 53) and CCNC operational mode conventional (1T -stepwise) orscanning flow (SFCA)

Exp Source dmdry dapp db

pp sc sc CCNC[nm] by number by number (52) (53) operation

(by mass) [nm] (by mass) [nm] [] [] mode

DEP1 Diesel exhaust BTX 150 28plusmn 8 (35) 30 (37) 018ndash18 ndash 1T -stepwiseDEP2 Diesel exhaust TX 150 28plusmn 8 (35) 28 (35) 018ndash205 ndash SFCADEP3 Diesel exhaust TX Whole aerosol 28plusmn 8 (35) 29 (36) 054ndash179 050ndash069 SFCADEP4 Diesel exhaust TX 90 28plusmn 8 (35) 28 (35) 025ndash088 039-081 SFCA

150 28plusmn 8 (35) 28 (35) 019ndash150 019ndash041300 28plusmn 8 (35) 28 (35) 007ndash065 ndash400 28plusmn 8 (35) 28 (35) 007 ndash

FSP1 Soot generator TX 90 ndashc 27 (34) 140ndash179 039ndash069 SFCA150 ndashc 27 (34) 018ndash082 019ndash095300 ndashc 27 (34) 008ndash009 008

FSP2 Soot generator TX 60 ndashc 18 (22) 078ndash156 076-096 SFCA90 ndashc 18 (22) 061ndash203 053ndash063

150 ndashc 18 (22) 028ndash116 036ndash121

a TEM image analysis (Rissler et al 2013)b Calculated according to Rissler et al (2013)c TEM samples missing

By changing the air-to-fuel ratio in the flame soot genera-tor particles with primary particle sizes ofsim 18 andsim 27 nmwere generated (calculated according to Rissler et al (2013)see Sect 51 for more details) In all experiments selectedamounts of toluene andm-xylene (at a ratio of 2 1) wereadded to allow investigations of the early ageing as well asthe full transformation of BC from agglomerates to spheresToluene andm-xylene are anthropogenic SOA precursorscommonly found in diesel and gasoline exhaust The initialamount of VOCs was 300ndash900 ppb The NO level after ex-haust injection wassim 500ndash600 ppb Ozone was used to titrateNO down to a concentration of about 50 ppb before the on-set of UV radiation This also allowed us to investigate theeffects of addition of ozone on CCN properties of soot

21 Instrumentation

For monitoring the soot transformation during ageing a com-prehensive instrumental set-up was used shown schemati-cally in Fig 1

The cloud-activation properties were measured usingcontinuous-flow streamwise thermal-gradient CCN coun-ters (CFSTGC from DMT CCNC-100 DMT described byRoberts and Nenes 2005 and Lance et al 2006) In theCCNC a single supersaturated column and an optical particlecounter (OPC) are used for measurements of CCN Supersat-uration is achieved via a vertical thermal gradient through thecolumn (change of temperature) and can be set to a range of007ndash2 according to the manufacturer The column is con-tinually wetted and the instrument keeps a continuous flowA differential mobility analyser (DMA) was placed prior

to the CCNC enabling a size selection of the dry quasi-monodisperse aerosol particles according to their mobilitydiameter (dm)

In short the operation principle in the CCNC is that diffu-sion of water vapour is faster than diffusion of heat in air Thetemperature and water vapour concentration will travel fromdifferent distances upwind from the walls to points along thesymmetry axis in the centre of the column Water diffusesmore quickly than heat Hence there is more water vapouravailable than thermodynamically allowed at the points alongthe symmetry axis Sample air passes along this axis sur-rounded by sheath air

Experiments were performed during two separate cam-paigns During the first campaign (experiment DEP1 Ta-ble 1) the supersaturation change was induced in a ldquoconven-tionalrdquo manner (forsc lt 1 ) the flow is kept constant whilethe temperature gradient is varied in a stepwise way Also themobility diameter (dm) was kept constant Due to the slowtemperature stabilisation measurement of complete super-saturation spectra is time consuming To capture thesc gt 1 in the experiment DEP1 both the supersaturation (18 ) inthe CCNC as well asdm (150 nm) was kept constant allow-ing the ageing of the aerosol scan past the supersaturationto capture the point of activation The residence time in theCCNC issim 6ndash12 s

However during the second campaign the Scanning FlowCCN Analysis (SFCA introduced and described in detailby Moore and Nenes 2009) was introduced (DEP2ndash4 andFSP1ndash2 Table 1) This altered way of operation enablesrapid and continuous measurements of the supersaturationspectra By adding a small software change to the robust and



NOx CO amp O3Analyzers

SMPS

DMA-TD-APM

FIDTesto350

Ejector diluter

AC

Fluorescent tubes

STEEL CHAMBER

SMOG CHAMBER

Exhaust air out

6 m3

22 m3

Generator

SP-AMS

DM

A

CCNC

CCNCSFCA

Figure 1 Schematic illustration of the instrumental set-up Twosources of particles were examined soot from (1) a Euro II DieselPassenger Vehicle and (2) a Flame Soot Generator The photochem-ical ageing was induced using black lights (peak at 354 nm) in a6 m3 TeflonFEP bag inside the steel chamber Two cloud conden-sation nucleus counters (CCNCs DMT 100) running in parallelin SFCA operational mode measured the cloud-activation proper-ties (for all experiments except DEP1) The Aerosol Particle MassAnalyzer (APM Kanomax Japan 3600) characterised the parti-cle massndashmobility relationship Both instruments measured the ex-haust aerosol after mobility size selection by a differential mobilityanalyser (DMA) A scanning mobility particle sizer (SMPS) sys-tem monitored the particle number size distribution The chemicalcomposition of the particles was determined using a Soot ParticleAerosol Mass Spectrometer (SP-AMS Aerodyne Research)

well-established hardware of the CCNC the flow is insteadvaried through the column in a controlled manner while thestreamwise temperature gradient (1T ) and pressure (P) ismaintained constant A higher flow increases the differencein travel distances between water and heat and thereby in-creases the supersaturation The flow rate in the chamber isvarying in a scan cycle where the flow increasesdecreaseslinearly (for 120 s) as well as being briefly kept constant atmaximumminimum flow rates (for 20 s)

When calibrating the instrument a size-selected aerosol ofwell-known chemistry is used here ammonium sulfate andsucrose The flow and the corresponding activated fractionof the particles generate a supersaturation curve with a ldquocrit-ical flow raterdquoQ50 From knowledge of the particle dry di-ameter and chemical compositionQ50 is translated to a crit-ical supersaturation (sc) using Koumlhler theory Hence everyinstantaneous flow rate corresponds to a critical supersatu-ration (see calibration curves for one of the CCNCs Sup-plement Fig S1) However the calibration curves are spe-cific for the chosen1T scan time and pressure The calibra-tion curves are presented with error bars representing 95

confidence intervals SFCA enables measurements of manysupersaturation spectra during a short time period allowinga better temporal resolution The inlet temperature can bekept closer to ambient conditions therefore minimising bi-ases from volatilisation of semi-volatile compounds in theinstrument (Moore and Nenes 2009)

To capture the whole supersaturation spectra of the ageingsoot agglomerates three values of1T were used (1T = 1810 and 4 K) Two CCNC instruments were running in SFCAmode in parallel after a DMA Hence the two instrumentswere measuring the samedm but with overlapping1T Thesize selection in the DMA is given in Table 1 The rapidand continuous measurements of the supersaturation spectramade it possible to capture the change in activation due to thefast ageing of the soot agglomerates By running the CCNCinstruments in parallel with inverse scan cycles ie with oneinstrument at maximum flow rate and the other at minimumthe aerosol flow was kept constant (1 L minminus1)

The particle massndashmobility relationship of individual par-ticles was measured using an Aerosol Particle Mass Analyzerafter size selection with a Differential Mobility Analyser(DMA-APM McMurry et al 2002 Kanomax Japan 3600)The increasing mass fraction of condensed material on sootparticles undergoing changes in morphology was quantifiedusing the approach introduced by Pagels et al (2009) Athermodenuder was introduced between the DMA and APM(Malik et al 2011) hence a DMA-TD-APM The ther-modenuder operated at 300C By comparing measurementswith and without thermodenuder it was possible to quantifythe size-dependent mass fraction condensed onto the non-volatile soot cores For calculations of the volume-equivalentdiameter (dve) the peak of the mass distribution was usedAn example of operational details is found in the Supplement(Table S1) and a more detailed description of the operationalprocedure and calculations are found elsewhere (Rissler etal 2013 2014)

A custom-built scanning mobility particle sizer (SMPS)system (Loumlndahl et al 2008) was used for measuring the par-ticle number size distribution from about 10 to 600 nm TheSMPS system consists of a DMA (Vienna 028 cm long)a 63Ni bipolar charger and a condensation particle counter(CPC model 3010 TSI Inc USA) with a sheathaerosolflow relationship of 4907 dm3 minminus1

The chemical composition of the particles (soot core andorganic coating) was determined using an online Aerodynehigh-resolution time-of-flight mass spectrometer (HR-ToF-AMS Aerodyne Research) For detection of refractory blackcarbon (rBC) the instrument was equipped with a laser va-poriser which is referred to as a Soot Particle Aerosol MassSpectrometer (SP-AMS Aerodyne Research) The laser wasoperated in 5 min periods every hour while the tungsten va-poriser (used in an (non-SP) AMS) was engaged continu-ously The 5 min intervals were used to estimate the organicaerosol (OA) mass fraction (mfOA) further described in theSupplement The tungsten vaporiser data were only used to



Table 2 Input values for CCN modelling The values for SOA andPOA are independently retrieved from measurements modellingand previous studies (see references in Sect 4)

Water SOA POA BC

M(kg molminus1) 0018153 02 04ρ (kg mminus3) 9971 1400 800 1850i ndash 1 0 0σ (N mminus1) 0072 0072lowast 0072lowast

κ ndash 01265 0 0T (K) 29815

lowast In solution with water

study the chemical composition of the OA Both instrumentconfigurations are described elsewhere (DeCarlo et al 2006Onasch et al 2012)

High-resolution transmission electron microscopy (HR-TEM) image analysis of the soot agglomerates to determineprimary particle size and soot microstructure has been per-formed and is described elsewhere (Rissler et al 2013) Inshort soot was deposited onto lacey carbon coated copperTEM grids using an electrostatic precipitator (NAS model3089 TSI Inc operated at 96 kV 1 lpm) and then analysedusing an HR-TEM (JEOL 3000 F 300 kV) equipped with afield emission gun

For general monitoring as well as for detailed chemistrymodelling (not included in this study) NOx O3 CO RHtemperature differential pressure and UV intensity werecontinuously monitored throughout the experiment Fur-thermore in selected experiments a Proton Transfer Reac-tion Mass Spectrometer (PTR-MS Ionicon Analytik GmbHAustria) was used for monitoring of time-resolved VOCconcentration (light aromatic compounds and other selectedVOCs) The monitoring instruments as well as the SMPS andAMS are further described by Nordin et al (2013)

3 Theory

Theoretical calculations of the critical supersaturations forthe CCN activation of the soot particles coated with differentorganics have been performed using Koumlhler andκ-Koumlhlertheory The Koumlhler theory describes the saturation ratiosover an aqueous solution droplet as (Pruppacher and Klett1997 Seinfeld and Pandis 2006)

s =p

p0= aw times Ke (1)

The saturation ratio is defined as the ratio of the actual partialpressure of water (p) to the equilibrium pressure over a flatsurface of pure water (p0) at the same temperature The ac-tivity of water in solution is described by the termaw and Ke(the so-called Kelvin effect) determines the effect the surfacecurvature has on the equilibrium water vapour pressure The

Kelvin term is given by

Ke = exp

(4σsolMw

RTρwDwet

) (2)

whereσsol is the surface tension of the droplet solutionMwandρw is the molar weight and density of waterR is the uni-versal gas constantT is the absolute temperature andDwet isthe diameter of the spherical aqueous solution droplet (inputvalues are listed in Table 2) In the basic equation an ap-proximation is made of the partial molar volume of water bythe molar volume of pure water (Kreidenweis et al 2005)In this studyσsol is parameterised by the surface tension ofwater (σw) with a constant value of 0072 N mminus1 The activ-ity of water (aw) can be described by the following form ofRaoultrsquos law where the vanrsquot Hoff factor (is) represents theeffects of ion interactions and dissociation (Kreidenweis etal 2005 Rose et al 2008)

aw =

(nw

nw + isns

)=

(1+ is

ns

nw

)minus1

=

(1+

nsumMw

ρwπ6

(D3

wetminus d3

s

))minus1

(3)

in whichds is the diameter of the dry particlens andnw arethe amount of substance (number of moles) of solute and ofwater in the solution respectively andnsum is the sum of thedifferent contributing components in the particles calculatedas

nsum=

sumi

εi iiρid3s

Mi

π

6(4)

in which εi is the volume fraction of a componenti in thedry particle of diameterds ii is the vanrsquot Hoff factor andρi the density of the component andMi is the correspond-ing molecular mass for that particular component For hy-groscopic salts (strong electrolytes) such as ammonium sul-fate and sodium chloride the vanrsquot Hoff factor is similar(but not identical) to the stoichiometric dissociation number(νs) ie the number of ions per molecule or formula unit(ν(NH4)2SO4 =3νNaCl = 2) Deviations betweenis andνs canbe attributed to solution non-idealities The vanrsquot Hoff factorcan be calculated using (see eg Kreidenweis et al 2005Florence et al 2011)

is = νsφs (5)

If the solution is non-idealφs deviates from unity ieφsrepresents the molal or practical osmotic coefficient of thesolute in aqueous solution

The hygroscopicity parameterκ is a broadly used pa-rameter for direct comparison of hygroscopicity between H-TDMA and CCNC measurements (Petters and Kreidenweis2007) related toaw as follows

aw =

(1+ κ

Vs

Vw

)minus1

=D3

wetminus d3s

D3wetminus d3

s (1minus κ) (6)



whereVs andVw corresponds to the solute volume (assumedas the dry particle volume) and the water volume respec-tively In the same way asnsumaboveκ is the total contribu-tion from all volume fractions of components in the particleand is given by a simple mixing rule

κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)

andκ can alternatively be calculated from pairedscndashds val-ues derived from CCNC measurements with the followingapproximation (valid forκ gt 02)

κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)

For 0 ltκ lt 02 the contribution of the initial dry aerosol parti-cle volume to the total volume of the droplet is non-negligibleand the behaviour approaches that predicted by the Kelvinequation ie expected for an insoluble but wettable particle

4 Modelling

41 CCN activation modelling

As discussed by others (eg Khalizov et al 2009 Tritscheret al 2011 Henning et al 2012 Rissler et al 2012) thenon-sphericity and restructuring of the soot particles will in-troduce a systematic error assessing the volume from mea-surements of the mobility diameter (dm) In theκ model theparticles are assumed to be spherical Heredm is convertedinto volume-equivalent diameter (dve) to account for thenon-sphericity of the dry particles when usingκ-Koumlhler the-ory (dve is used in a similar way as in Tritscher et al 2011)The sizes of the dry particles (dve) as well as the soot andtotal organic mass fractions (mfBC(APM) and mforg(APM)respectively) in the particles were derived from direct mea-surements of the relationship between particle mass and mo-bility diameter using the DMA-TD-APM set-up (see detailsin Sect 21) Calculations are performed according to (Mc-Murry et al 2002)

dve =3

radic6

π

m

ρcorr (11)

wherem is the particle mass andρcorr is the corrected mate-rial density of the particles assessed from

1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)

where mfBC(APM) mfSOA(APM) and mfPOA(APM) are thecontributing mass fractions in the particles of the differentcomponents approximated as described below

The initial mass fraction of POA (mfPOA(APM)) has beenapproximated with the measured values of mforg(APM)before the onset of UV The mass ratio of soot to POAis assumed to be constant for particles of a specific sizethroughout the experiment Therefore after the onset of UVmfPOA(APM) is assumed to decrease at the same rate asmfBC(APM) This assumption of constant proportions be-tween soot and POA is consistent with measurements by theSP-AMS Hence after the onset of UV mfPOA(APM) is sub-tracted from mforg(APM) to calculate the mass fraction ofSOA (mfSOA(APM)) Volume fractions (εi) for the differ-ent components have been calculated from the mass fractionsand densities to be used in theκ-Koumlhler modelling as wellas in calculatingρcorr above

In this study a high correlation is observed betweenthe volume-equivalent diameter (dve) the mobility diame-ter (dm) and the SOA mass fraction (mfSOA(APM)) in theparticle (Supplement Fig S2) Hence by applying a fit tothe empirical data an estimated volume-equivalent diameter(dvefit) has been derived fromdm and mfSOA(APM) to gaina higher resolution in theκ-Koumlhler modelling (see equationsin Sect 53)

The water activity is calculated from knowledge of the mo-lar masses of the components in the particle The mean molarmass (M) is a difficult property to measure as pointed outby Hallquist et al (2009) In the literature a variety of aver-age molar mass (M) for both SOA (sim 015ndash0480 kg molminus1Hallquist et al 2009 Kuwata et al 2013) and POA (025ndash07 kg molminus1 for lubrication oil eg Stubington et al 1995)are reported The difference in molar mass among studies isprobably due to different experimental conditions (eg com-bustion conditions NOx regime aerosol loading oxidantsand precursors used) In this study the molar masses havebeen set to values of 02 kg molminus1 and 04 kg molminus1 for SOAand POA respectively (Table 2)MSOA is a mean value de-rived from model runs (see Sect 42) whileMPOA is ap-proximated with the molecular weight of octacosane (C28minusM = 0395 kg molminus1 Lide 2005) a representative compo-nent in lubrication oil

In the same way the material densities for POA and SOAis set to constant values of 800 and 1400 kg mminus3 respectivelyndash 800 kg mminus3 is typical for hydrocarbons (Ristimaki et al2007) The density for SOA is derived from measurements ofthe aged particles in this study in agreement with previouslyreported densities of anthropogenic SOA formed fromm-xylene and toluene (eg 1330ndash1480 and 1240ndash1450 kg mminus3

respectively Ng et al 2007) For the soot particles a pri-mary particle density of 1850 kg mminus3 is used in the modelin good agreement with previous studies A density of 1800ndash2000 kg mminus3 is often reported for the compacted soot coreeg 1840 kg mminus3 (Choi et al 1994) 1800 kg mminus3 (Ristimaki



et al 2007) and 2000 kg mminus3 (Park et al 2003 Cross et al2007)

In this study SOA is treated as a water-soluble compound(iSOA = 1 Svenningsson et al 2006) and POA as a water in-soluble compound (iPOA = 0 for example octacosane is notwater soluble Lide 2005κ asymp 0ndash002 for SOA formed fromlubrication oil Lambe et al 2011) Using the above valuesfor molecular weight density and vanrsquot Hoff factor as in-put the estimated values ofκi becomesκSOA = 01265 forSOA κPOA = 0 for POA andκBC = 0 for the soot core re-spectively The calculatedκSOA value is similar as reportedin the literature for isoprene-derived secondary organic ma-terial particles (κ = 012plusmn 006 Kuwata et al 2013 andreferences therein) laboratory smog chamber SOA fromtrimethylbenzene (κ asymp 004ndash015 Jimenez et al 2009)κ

values of SOA formed fromm-xylene and toluene (κ asymp 01ndash027 Lambe et al 2011) and photochemically aged dieselsoot particles (κ = 0ndash013 Tritscher et al 2011)

For fresh POA aκ value below 01 is expected due to thelow O C ratio (lt 02 not shown) and highMPOA of the ma-terial as pointed out by others (Tritscher et al 2011 Kuwataet al 2013) If impurities (such as sulfur) are present in thefuel κ values greater than 01 may be found (Gysel et al2003) However Tritscher et al (2011) concluded that theκ

value for the fresh emissions (soot+ POA) should be close tozero due to low sulfur content in the fuel used in the study Inthis study low-sulfur fuel was used (lt 10 ppm) and the AMSdetected no particulate sulfate (lt 4 ng mminus3 during 1 min ofdetection) Therefore POA is treated as a water-insolublecompound Due to the insolubility of BC the soot particleas a whole is treated as water insoluble

The input parameters for the modelling exercises are thusindependently derived and have not been varied to fit the em-pirical results

42 Modelling of gas-phase chemistry organic aerosolformation and composition ADCHAM

In order to better understand the mechanisms responsible forthe observed changes in SOA coating chemical composi-tion and hygroscopic properties and to bring the discussionfurther detailed modelling was performed of the gas-phasechemistry SOA formation and composition during one ofthe experiments (DEP2 see Table 1) DEP2 was chosen dueto the good cover in empirical results of cloud activationproperties massndashmobility relationship and particle chemicalcomposition For this we used the Aerosol Dynamics gas-and particle-phase chemistry model for laboratory CHAM-ber studies (ADCHAM Roldin et al 2014) ADCHAM isa model primarily intended to be used to recreate labora-tory chamber experiments on SOA The model explicitlysimulates the deposition and re-evaporation of organic com-pounds from the chamber walls all fundamental aerosol dy-namics processes detailed gas- and particle-phase chemistryand the mass transfer limited mixing of compounds in the

particle phase It uses the detailed gas phase Master Chem-ical Mechanism version 32 (Jenkin et al 2003) an aerosoldynamics and particle phase chemistry module and a kineticmultilayer module for diffusion-limited transport of com-pounds between the gas phase particle surface and particlebulk phase In the online Supplement we describe in detailhow the ADCHAM model was set up

5 Results and discussion

51 The overall picture

No activation into cloud droplets was observed for the freshsoot particles at supersaturation below 2 neither from thediesel exhaust (DEP) nor from the flame soot generator (FSP)experiments (further described in Sect 52) This is in agree-ment with previous observations eg Tritscher et al (2011)and Henning et al (2012) The newly emitted diesel exhaustaerosol shows a carbon mean oxidation state in the rangeminus19 tominus16 (Fig 2a) obtained from AMS results (wherethe carbon mean oxidation stateasymp 2 O CndashH C Kroll et al2011) The results are similar to previous studies (Kroll et al2011) Furthermore the mass spectrum of the newly emitteddiesel exhaust aerosol shows an organic (POA) content ofmostly hydrocarbon species (Fig 2b) consistent with dieselexhaust measurements by Canagaratna et al (2007) An in-crease in carbon mean oxidation state is visible at the onsetof UV exposure (Fig 2a) At the same time measurements(AMS) as well as modelling (ADCHAM) of the HC ratioshows a decrease (Supplement Fig S4d) The mass spectraof the organic content (POA) of the aerosol emitted from thesoot generator are very similar to the mass spectra of the POAfrom the diesel exhaust aerosol According to measurementsof the chemical composition inorganic salts do not form dur-ing the early stage of the experiments

The morphology of the fresh emissions from both thediesel vehicle and the flame soot generator has been charac-terised in detail elsewhere (Rissler et al 2013) The primaryparticle diameters are consistent between experiments for theDEP but differing for the FSP depending on how the genera-tor was operated In summary the count median diameter ofthe primary particle diameter (dpp) in the agglomerates fromthe diesel vehicle (experiment DEP1-4) wassim 28 nm bynumber see Table 1 From the flame soot generator agglom-erates with two differentdpp were generated ndash agglomerateswith a smallerdpp of about 18 nm by number (experimentFSP2) and agglomerates with a largerdpp asymp 27 nm by num-ber (FSP1) TEM samples for experiments FSP1 and FSP2are missing However the primary particle diameters havebeen calculated from the measurements of the particle massand mobility diameter according to Rissler et al (2013)Results from these measurements show that the volume-equivalent diameters (dve) of FSP1 are closer to thedve ofthe DEP (Supplement Fig S2 ndash FSP1 (circles) and DEP1-3



Figure 2 Evolution of the carbon mean oxidation state of the or-ganic material coating the diesel soot(a) (carbon mean oxidationstateasymp 2 O CndashH C Kroll et al 2011) during DEP2 Also shownare the high-resolution mass spectra for the POA(b) and SOA(c)sim 1 h before andsim 5 h after the onset of UV exposure respectivelyThe characteristic for urban POA is the high amount of hydrocar-bons especiallymz 43 and 57 More oxygenated species wheremz 44 is dominating overmz 43 are typical for more rural SOAAlso the latter show a higher carbon mean oxidation state than theformer

(triangles) fall on the same line) implying that the primaryparticle diameter (dpp) also is similar for FSP1 and DEP butdifferent for FSP2 The differences between FSP1 and FSP2are reflected in the CCN measurements where FSP1 andDEP require a more similar supersaturation for activationfurther discussed in Sect 52 The number size distributionshowed a geometric mean mobility diameter (GMD) averagefor all DEP experiments ofsim 82 nm (with standard devia-tion σg asymp18) for the fresh soot aerosols ranging from 76to 92 nm (178 ltσg lt 189) between experiments The freshaerosols from experiments FSP1 and FSP2 showed a GMDof sim 117 nm (σg asymp 163) andsim 79 nm (σg asymp 188) respec-tively The parameters (GMD andσg) were determined fromthe fitted lognormal number size distributions Primary parti-cle diameters (dpp) from calculations and TEM picture anal-ysis as well as selected sizes (dm) for the CCN measurementsfor the different experiments are listed in Table 1

Figure 3 The evolution of the activation properties for thewhole aerosol during the ageing event of experiment DEP3Measurements from two CCNCs are overlapping in the range025 lt supersaturation lt 075 thereby are two lines visible for thelow supersaturations Before and just after the onset of UV expo-sure (UV on= minus075 and 003 h respectively) almost no particles(lt 1 ) are activated (dark blue) and the SOA mass ratio (mfSOA) isclose to zero mfSOA(AMS) is estimated from measurements fromthe SP-AMS As the soot particle acquires more organic material(mfSOA increase) the CCN properties improve By the end of theexperiment (UV on= 337 h) all particles are activated (dark red)and mfSOA asymp 04

At the onset of UV exposure an immediate (within 5 min)enhancement of the activation properties of the coated sootcores is seen Typically the first full supersaturation spectrawere measured within the first 10ndash15 min after onset of UVexposure for the particle size-resolved aerosol (see Sect 52for details) In the early stage of the experiments a mobilitydiameter (dm) of 150 nm was selected in most experimentsThe selected size was based on the knowledge that particlesof a smallerdm were not activated in the beginning of the ex-periments and particles with largerdm were too few in num-ber Furthermore focusing on one size improved the time res-olution of the measurements substantially During one exper-iment (DEP3 Fig 3) the whole aerosol was measured con-tinuously without any size selection upstream of the CCNCIn the beginning of this experiment a small number fractionof particles of the whole aerosol (lt 1 ) activates at high su-persaturation (gt 2 ) This was the only experiment show-ing this early activation which could either be activation ofexhaust aerosol or impurities As the aerosol ages the par-ticles become better CCN and by the end of the experimentalmost the whole aerosol is activated at high supersaturations(gt 16 ) (Fig 3)

The minimumdm that was able to activate at a certain su-persaturation at a certain SOA mass fraction (mfSOA(AMS))was retrieved by integrating number size distributions fromlarger to smaller sizes (Fig 4) Here the SOA mass frac-tion (mfSOA(AMS)) is estimated from measurements fromthe SP-AMS in the same way as the SOA mass fraction(mfSOA(APM)) is estimated from the APM measurementsbut for the whole size distribution (information on the pro-cedure to derive the masses of BC and organic material from



Figure 4 The minimum mobility diameter (dmmin) required foractivation of the diesel exhaust particles (DEP3) at different super-saturations (supersaturation colour coded) with respect to the SOAmass fraction (mfSOA(AMS)) for the whole aerosol measured bySP-AMS As the mfSOA(AMS) increases the activation propertiesof the diesel exhaust aerosol particles improve ie there is an in-crease of activated particles of smaller sizes for a certain supersatu-ration Also shown are the theoretical values for ammonium sulfateparticles (ASκAS asymp 056 organic fraction= 0) and organic parti-cles (κSOA asymp 013 organic fraction= 1) Note that before and justafter the UV exposure only about lt 1 of the particles are acti-vated

SP-AMS data is found in the Supplement) Hence size distri-bution measurements from the SMPS along with the changein SOA mass fraction derived from the SP-AMS are linkedto the activation properties measured by the CCNC For ex-ample at the onset of UV exposure if the particles are ex-posed to a supersaturation of 1 a minimumdm of 480 nmis required to activate 02 permil exhaust particles (Figs 3 and4) The mfSOA(AMS) has increased slightly from 0 tosim 001at this point As the exhaust particles acquire more SOA theybecome better CCN By the end of DEP3 the mfSOA(AMS) isabout 04 and the minimumdm has decreased to about 60 nmfor a supersaturation of 1

The observation of an immediate change in activation atthe onset of UV exposure is consistent in all experiments in-dependent of amounts of added precursors oxidants and sootparticle source Due to the new way of operating the CCNCusing scanning flow (SFCA) with a higher supersaturationand time resolution it was possible to cover the developmentof the activation properties more thoroughly than ever donebefore (Figs 5ndash7 ldquoscrdquo) Hence it was possible to captureboth the onset of activation covering the very first changesof the particles becoming better CCN as well as followingthe evolution of the particles

The evolution of a decreasingsc along with an increas-ing SOA mass fraction (mfSOA(APM)) coating the soot coresand a change in volume-equivalent diameter is illustrated inFig 6 (DEP2) The decline ofsc continues throughout the

Figure 5 Comparison of the two measurement techniquescontinuous-flow streamwise thermal-gradient CCNC (CFSTGC or-ange) and Scanning Flow CCN Analysis (SFCA blue red andgreen) with error bars (black) representing 95 confidence inter-vals from experiments DEP1 and DEP2 respectively The changein sc over time for diesel exhaust particles is captured in both ex-periments though with a better resolution while using SFCA Aslower decrease ofsc is seen in DEP2 (SFCA) due to altered ex-perimental conditions with a slower ageing process During DEP2the temperature gradient (1T ) was changed three times (4 K blue10 K red and 18 K green) Note that the first measurement pointusing CFSTGC was possible due to letting the experiment scan pasta constant supersaturation anddm (see Sect 21 for more informa-tion)

ageing process although the effect is more prominent at thebeginning of the process For the mfSOA(APM) the trend mir-rors thesc with a higher increase in mass fraction at the be-ginning which levels out by the end The rate of decreasingscand increase of coating mfSOA(APM) differ between experi-ments (compare Figs 6 and 7) ie the rates are dependent onamount of SOA precursors and ozone added and NOx lev-els The time of photochemical ageing in the smog chamberuntil the soot particles become CCN active at a supersatu-ration of 02 (equivalent to the supersaturation in a stra-tocumulus cloud) range from 15 to gt 45 h With respect toorganic condensational growth this corresponds to an atmo-spheric ageing time at the mid-latitudes of between 4 h and afew days (for details see Supplement) However in the atmo-sphere other compounds such as biogenic organics and inor-ganics are present which also will affect the hygroscopicityand the lifetime of the soot particles hence the calculatedatmospheric ageing time is an approximation

52 Detailed picture

In general cloud droplet formation ability of aerosol parti-cles can be described as a function of the number of dissolvedmolecules and ions in the activating droplet parameterisedby for example theκ value In the case of particles dom-inated by soot the volume occupied by insoluble material



Figure 6 Changes insc (blue) mfSOA (green markers empiricalgreen line fit) anddve (black) over time for diesel exhaust particlesduring experiment DEP2 No activation of particles was seen beforethe onset of UV

influences the Kelvin effect and therefore has to be taken intoaccount In present experiments the time-dependent changesof activation properties of the coated soot cores are sum-marised by four factors (I) particle organic fraction (II) typeof organic coating (POA or SOA) (III) particle size and (IV)morphology Here the results with respect to each of thesefactors will be discussed

(I) The ability of the soot cores to act as CCN increaseswith increasing amount of organic coating material ieincreases in SOA mass fraction (mfSOA(APM)) (Figs 6and 7) At the end of the ageing process when theSOA mass fraction makes up most of the particles(mfSOA(APM) ge 07 size dependent) thesc levels outIn all experiments all measured sizes show the sametrend The condensation rate of SOA was different indifferent experiments However independent of the ratesc is more or less the same for a chosen size with a cer-tain mfSOA(APM)

(II) No activation was observed in the early stage of theageing process ie before UV exposure although anincrease is visible in carbon mean oxidation state(Fig 2a) At this stage the soot core makes up mostof the particle mass (mfBC(APM) sim 09) and the re-maining fraction is dominated by POA The mass spec-tral signature measured by the AMS corresponds wellwith hydrocarbon like OA (HOA) commonly found inurban environments (Jimenez et al 2009) and previ-ous diesel exhaust emission (Canagaratna et al 2007)studies of POA (an example from DEP2 is shown inFig 2b) POA originates from the combustion process(presumably from lubrication oil) and even as the POAreacts (change in carbon mean oxidation state Fig 2a)before the onset of UV exposure it is not hygroscopicenough to suppress thesc below 2 According to sim-

ulations with the ADCHAM model (Roldin et al 2014Sect 42) no SOA is formed before the UV light isturned on and according to the simulations the con-centrations of ozone hydroxyl radicals (OH) and nitrateradicals (NO3) at this stage are insignificant (see Fig S3in the Supplement) Also no detectable particle mass in-crease or fundamental changes in the mass spectra areobserved before the onset of UV radiation other thana slight increase inmz 44 due to CO+2 Hence it isunlikely that any substantial SOA formation is takingplace during dark conditions before the onset of UVexposure (DSOA in Fig 8) Instead the increase in car-bon mean oxidation state can be explained by hetero-geneous oxidation of POA by NO2 at the surface of thesoot core (illustrated in Fig 8 as OPOA) Such reactionshave previously primarily been studied because of theirpotential importance for HONO formation in the atmo-sphere (eg Arens et al 2001 Han et al 2013) UsingATR-IR spectra Han et al (2013) observed a great in-crease in several absorbance bands associated with ni-tro (RndashNO2) and nitrate (RndashOndashNO2) organic functionalgroups after NO2 exposure However in our experi-ments the altered chemical composition of the organiccoating material ie a change in carbon mean oxida-tion state does not affect the activation properties ofthe particles at supersaturationssim 2 As concludedby Tritscher et al (2011) it cannot be ruled out that thehydrophobicity of the particlesrsquo surfaces could be a hin-dering effect of the particles to act as CCN All experi-ments show an initial O C ratio lt 02 which accordingto Kuwata et al (2013) classifies non-CCN-active com-pounds Also lubrication oil (for example octacosaneor C28) which POA originates from is not water solu-ble (Lide 2005) In general it should be pointed out thatthe organic mass fraction (mforg(APM)) is low through-out the period before UV exposure The initial total or-ganic aerosol (OA) fraction is 2ndash8 and just after theonset of UV radiation (up to 30 min in some experi-ments) it is still low (typically lt 9 ) ie the amountof formed oxidised material in the early stage of exper-iments is very small compared to the SOA fraction ofthe aged soot in this study

As mentioned in Sect 51 the first enhancement inhygroscopicity of the particles is observed when theexhaust aerosol and precursors are subjected to UVWeingartner et al (1997 and references therein) pro-posed that this change could be attributed to eg pho-tolysis of Polycyclic aromatic hydrocarbon (PAH) oran oxidation of the particle surface from photochemi-cally produced hydroxyl radicals (OH) or condensationof photochemically produced compounds At the onsetof UV exposure there is often (but not always) a de-tectable increase in the organonitrate (RndashONO2) levelin the chamber Also more oxidised organic compounds



are produced (increase in carbon mean oxidation stateillustrated in Fig 2a) Condensation of organonitratesand other oxidised organic compounds at the soot sur-face would improve the hygroscopicity of the particle(illustrated in Fig 8) Another possibility of the changein CCN properties at the onset of UV is oxidation ofPOA by OH or ozone Probably both mechanisms takeplace and contribute to improved CCN activity

According to the ADCHAM model simulations of theDEP2 experiment it takes approximately 1 hour duringthis experiment before oxidation products of the addedprecursorsm-xylene or toluene start to contribute to theSOA formation Instead the initial SOA formation afterthe onset of UV exposure is likely to be condensationof low-volatile organic compounds formed when OHreacts with intermediate volatility organic compounds(IVOCs Donahue et al 2009) present in the diesel ex-hausts Naphthalene alkylnaphthalenes and other PAHsare present in diesel exhausts (although at lower levelsthan light aromatic compounds and alkanes) (Schauer etal 1999) Still because of their high reactivity towardsOH and large fraction of low-volatile oxidation prod-ucts (high SOA yields) they may dominate the initialSOA formation in fresh diesel exhausts (see eg Chan etal 2009) However with the best estimate of the PAH(naphthalene) concentrations in the chamber the AD-CHAM model still substantially underestimates the ini-tial SOA formation (Supplement)

During the following hours of experimentsc is declin-ing as the amount of condensed organics increases Bythe end of the ageing process the organic coating con-sists mostly of oxygenated compounds with anmz sig-nature similar to the semi-volatile oxygenated organicaerosol factor commonly representing relatively freshSOA in the atmosphere (Fig 2c) Also the carbon meanoxidation state is aboutminus04 (illustrated in Fig 2a) inthe range of what is reported for alkanealkene photo-oxidised SOA (Kroll et al 2011) According to ourADCHAM model simulations of the DEP2 experimentby the end of the experiment more than 80 of theSOA mass can be attributed to the oxidation productsfrom the added SOA precursorsm-xylene and toluene(Fig S5 in the Supplement) The change in compositionof the organic coating into more oxygenated species ismore pronounced in the beginning of the ageing processwhen the POA fraction of OA is still substantial Forexample the carbon mean oxidation state drastically in-creases in the beginning and then flattens out towardsthe end (illustrated in Fig 2a)

(III) A size dependency is evident for both the activationproperties as well as for the condensation of organicmaterial For particles of a specific mfSOA(APM) thesmaller particles require a higher supersaturation for ac-tivation than the larger ones This is mainly explained by

00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)

Time from UV exposure (h)

mfSOA (dm=60 nm) mfSOA (dm=90 nm) mfSOA (dm=150) sc (dm=60 nm) sc (dm=90 nm) sc (dm=150 nm)

Figure 7 Decreasing critical supersaturation (sc blue) and increas-ing mfSOA (green markers empirical green line fit) for three dif-ferent mobility diameters (dm = 60 90 and 150 nm) from measure-ments of flame soot generator particles over time during experimentFSP2

the fewer amounts of water-soluble ions or moleculesand thus larger Kelvin effect which is more pronouncedfor particles with higher curvature However smallerparticles are more efficient in relative mass acquisi-tion by SOA condensation (Fig 7) as pointed out byTritscher et al (2011) They found that small particlesshow the highest hygroscopic growth although a moreefficient acquisition of SOA does not lower thesc to adegree where it would compensate for a smaller particlesize (Fig 7) It should be pointed out that the size depen-dence of mfSOA(APM) at a given time is much smallerthan that predicted for spherical particles according totransition regime mass transfer models This is a directconsequence of the highly agglomerated particle mor-phology (Pagels et al 2009)

(IV) Fresh soot and particles dominated by soot have an ag-glomerate structure (Fig 8) implying that the parti-cle volume is smaller than expected for spherical par-ticles A restructuring of the agglomerated soot parti-cles is seen during ageing (Fig 9andashc) in agreementwith previous reported work (eg Weingartner et al1997 Tritscher et al 2011) Tritscher et al (2011) ar-gue that shape effects are the main reason for the ab-sence of CCN activation of the fresh emissions An-other possibility is that the acquisition of water-solublematerial is not evenly distributed around the soot coreleading to specific activation sites (Fig 9a) If mostof the particlesrsquo soluble material is excluded from thedroplet at activation (Fig 9a solid blue lines) due toformation of several smaller separate droplets the criti-cal supersaturation will be higher than expected assum-ing that all soluble material is in the same droplet and



Figure 8A conceptual model illustrating different processes whichmight affect the activation properties of the soot particles in theearly stage of ageing Here POA is evenly distributed over the ag-glomerate The volume-equivalent diameter is used to account forthe non-sphericity of the particles In dark conditions (before theonset of UV radiation) POA on the soot surface might be oxidisedie oxidised primary organic aerosol (OPOA) Another possibil-ity is that gas-to-particle conversion processes may occur ie sec-ondary organic aerosol produced in dark conditions (DSOA) con-denses onto the soot particles At the onset of UV the organic mate-rial on the soot surface (OPOA andor DSOA) might be further ox-idised andor condensation of oxidised material (traditional SOA)andor condensation of small amounts of organonitrates transformsthe particles from non-activating into activating CCN atsim 2 su-persaturation

involved in the activation In contrast if droplet growthstarts at different sites successively incorporating moreand more of the particle and growing to form one sin-gle droplet (one of these sites is illustrated in Fig 9adashed blue lines) all soluble material in the particle isinvolved in the activation

The particles become more compact and spherical-likewith increasing amount of organic (SOA) coating material(Fig 9c) This is due to a combination of two effects SOAmaterial filling the void spaces (Nakao et al 2011) and re-structuring of the agglomerates The restructuring is due tocondensation of organic (SOA) material onto the soot par-ticles in the smog chamber Additionally condensing wa-ter in the CCNC may enhance the effect of restructuringThe changed morphology due to for example SOA conden-sation is illustrated in Fig 6 (increase indve during experi-ment DEP2) Fig 9 (TEM pictures of processed DEP) andFig 10 (sc decrease with increasingdve experiments DEP2DEP4 and FSP1 FSP2) The fresh soot particles show a size-dependent morphology with higher dynamic shape factorvalues for the larger particles than for the smaller ones Thedynamic shape factor (χ = (dmdve)(Cc(dve))(Cc(dm))DeCarlo et al 2004) typically increases with increasing mo-bility size and with decreasing primary particle size It variedbetween 17 and 25 for the freshly emitted highly agglomer-ated soot particles for all sizes measured here At the end ofthe experiments the particles are more spherical-like with ashape factor close to 1 As the particles grow by coagulationin the combustion process they develop larger branches with

Figure 9 TEM pictures of diesel exhaust particles (DEP1) a freshdiesel exhaust particle(a) after 1 h(b) and 4 h of ageing in thechamber(c) In (a) an activation site excluding most of the particlesrsquosoluble material in the droplet at activation is illustrated (solid bluecircles) as well as a site with more material included (dashed bluecircles) for the diameters 35 100 and 250 nm

a highly agglomerated structure When larger more branchedparticles are subjected to the ageing process they will ex-perience a more pronounced restructuring Weingartner etal (1997) found that the DEP exhibit a smaller restructur-ing combined with condensational growth than the soot ag-gregates from the soot generator This might be an effect ofthe primary particle size More-branched (less dense) aggre-gates are formed when smaller primary particles are presentin the combustion process from the soot generator (FSP2)Hence the FSP with the smallerdpp (by number)asymp 18 nmare able to restructure to a larger extent The empirically de-riveddve are smaller for the agglomerates with a smallerdpp(FSP2) compared to the agglomerates with a largerdpp (bynumber) of 28 nm (asymp FSP1 and all DEP experiments) at thesame mfSOA(APM) Although the activation properties showvery similar results for the two different kinds of soot thereis a slight difference Particles with the smallestdpp (FSP2)are slightly better CCN for a givendve (compare diamondswith circles andor trianglesdm = 150 nm in Fig 10) Theseparticles (FSP2) are probably activating at a lower supersatu-ration due to a smaller mass of soot for a givendm and hencea larger SOA mass than a particle with largerdpp (DEP andFSP1) for a certaindve However for a given mfSOA(APM)the same particles (FSP2) need a slightly higher supersatu-ration for activation than the soot agglomerates with largerprimary particles (compare diamonds with circles andor tri-anglesdm = 150 nm in Fig 11) For the same mfSOA(APM)a particle with largerdpp will activate at a lower supersatu-ration due to a larger size of the agglomerate (dve) therebyalso containing more water-soluble molecules than a particlewith smallerdpp

53 Modelled vs empirical results

Simple Koumlhler theory as well asκ-Koumlhler theory was usedin this study showing the same results (described in Sect 3)Therefore theκ-Koumlhler model represents results from both



Figure 10 Empirical sc for different volume-equivalent diameters (dve) size-selected according to their mobility diameter (dm colourcoded) in comparison with calculated values ofsc from κ-Koumlhler theory (greyscaling) Modelled values ranges from insoluble but wettableparticles (κwettable= 0 black line) to AS particles (κAS asymp 056 lightest grey) in the size range 40ndash300 nm Also plotted are modelled valuesfor theκ value derived for the SOA appearing at the end of the experiments in this study (κSOA asymp 013) Triangles denote coated diesel sootparticles (DEP1 2 and 4) Coated soot generator particles are represented by diamonds (FSP2) and circles (FSP1) differentiated due to theirdifferent primary particle sizes

Figure 11Measured critical supersaturation (sc) by estimated mass fraction SOA (mfSOA) for experiments FSP1 (circles) FSP2 (diamonds)and DEP2 (triangles) Also shown areκ-Koumlhler modelled results usingdvefit as input fords (filled markerslines) The measured mobilitydiameter of the particles are colour coded separating only the FSP from DEP measurements fordm = 150 nm (dm = 60 nm dark bluedm = 90 nm light bluedmFSP= 150 nm greendmDEP= 150 nm pinkdm = 300 nm orange) The measurements are well representedby the model for mfSOAgt 012 For lower organic fraction the particle properties are hindering the activation into cloud droplets ie a highersupersaturation is needed for activation Also the model is not as representative for FSP2dm = 90 and 150 nm (green diamonds) as for theother experiments



these models To improve the performance of theκ-Koumlhlermodel the input parameters have been tested as follows(Fig 12) Firstly a model taking only the Kelvin effectinto account was used In this model the Raoult term is ne-glected by settingiSOA = 0 and the dry diameter (ds) of theparticle is equal to the measured volume-equivalent diam-eter (dvemeasured) The particles are assumed to be insolu-ble but wettable hence the model is named Wettable Sec-ondly both the Kelvin effect and the chemical compositionwas taken into account but neglecting the shape effect (iedsequals the mobility diameterdm) Hence the model is calledκ-Koumlhler (dm) In the third modelκ-Koumlhler (dvemeasured)both the chemical composition size and shape are accountedfor (discussed in more detail further down) As input pa-rameter fords the volume-equivalent diameter (denoted asdvemeasured) is used estimated from the measurements of themassndashmobility relationship (DMA-TD-APM) For the fourthmodelds is exchanged by the volume-equivalent diameter(dve) derived from an empirical fit function with input param-eters of the mobility diameter (dm) and mass fraction SOA(mfSOA(APM)) denoted asdvefit (Eqs (13) and (14) belowFig S2 Supplement) This model is calledκ-Koumlhler (dvefit)

and uses an approximation of the volume-equivalent diame-ter performed to gain better time resolution of the modelledresults for comparison with empirical results (Figs S2 andS7andashb in the Supplement)

In the model the porosity and structure of the coated sootparticles as well as the size is partly represented by the num-ber of water-soluble molecules (Raoultrsquos law) and partly bythe effect of surface curvature of the solution droplet (theKelvin effect) The volume-equivalent diameter (dve) ac-counts for the change in morphology in the model and canfor a DEP with a primary particle diameter (dpp) of sim 28 nmfrom a Euro II vehicle be approximated by (in nm) (Supple-ment Fig S2)

dvefitDEP = (0540897times mfSOA + 0414766)

times dm minus 249877times mfSOA + 3153536 (13)

anddve for a flame soot generator particle (FSP) with a pri-mary particle diameter (dpp) of sim 18 nm can be approxi-mated by

dvefitFSP= (0358919times mfSOA + 0490475)

times dm + 0764891times mfSOA + 118527 (14)

The massndashmobility relationship anddpp for the diesel sootinvestigated here is similar to a number of emission studies inthe literature (eg Park et al 2003 Maricq and Ning 2004Rissler et al 2013) and the parameterisation may thus beof relevance for diesel exhaust in general As discussed inSect 52 the primary particle size of the soot might influencethe activation properties This is partly accounted for bydvein the model (Fig 11 FSP2 ndash diamonds compared to lines)iedve is smaller for smallerdpp at a givendm

The κ-Koumlhler (dvemeasured) and κ-Koumlhler (dvefit) mod-els capture the evolution of the decreasingsc well exceptat the beginning of the ageing process whensc is underes-timated (Fig 12 turquoise and orange lines respectively)When the measured mobility diameter is used as input fords in the modelκ-Koumlhler (dm) the results deviate far fromexperimental results (Fig 12 pink line) Only at the end ofthe ageing process when the particles are more spherical-like anddm asymp dve does theκ-Koumlhler (dm) model agree withthe empirical results (Fig 12 pink line vs blue triangles) Inthe early stage of the experiments the Wettable model bestexplains the observedsc (Fig 12 green line) Hence theslightly coated soot particles activate at a higher supersatu-ration than expected with the assumption that the organicfraction is SOA with aκ = 013 However the agreement be-tween Wettable model and the empirical results in the begin-ning of the experiments might be a misleading coincidenceActivation was clearly occurring and visible in the CCNC atthe onset of UV radiation though measurements of wholeactivation steps were not possible above 2 supersaturationThere are many processes and possibilities to explain thechanged behaviour of the soot particles from non-activatinginto activating CCN as discussed in Sect 52 (Fig 8)

The deviation between modelled (κ-Koumlhler (dvemeasured)

andκ-Koumlhler (dvefit)) and empirical results in the early age-ing process (mfSOA(APM) lt 012) might be due to hinderingshape effects (Tritscher et al 2011) or possibly only partsof the soot particle serve as activation sites due to the highlyagglomerated structure (Figs 9a and S7a in the Supplement)If activation sites were considered the critical diameter cor-responding to activation would be much smaller than the di-ameter of the whole particle (eitherdve or dm is used in themodel) A plausible range ofds would be 30ndash50 nm and thenthe activation diameter would besim 100 nm (the same as thevolume-equivalent diameter in the beginning of the ageingprocess) In any case this effect would possibly be more pro-nounced for the larger particles

Another possibility (to explain the deviation between mod-elled and empirical results for mfSOA(APM) lt 012) wouldbe a transformation of the hydrophobic organic material to asemi-hydrophilic or wettable material As suggested by oth-ers (Bilde and Svenningsson 2004 Petters and Kreidenweis2008 Kuwata et al 2013) knowledge about the solubilityof the particle material as well as about the particle phasecan be important parameters when interpreting experimentaldata and modelling Petters and Kreidenweis (2008) arguethat thesc of certain mixtures and solubilities are more sen-sitive to dry particle diameter and also noticeably affected bysmall amounts of moderately soluble and hygroscopic com-pounds Kuwata et al (2013) suggested that organic com-pounds acting as CCN could be divided into three differ-ent regimes depending on the O C ratio of the materialThey found that for the insoluble regime there is no acti-vation into cloud droplets (O C lt 02)κ = 0 Compoundsthat are highly CCN active are in the highly soluble regime



Figure 12 Empirical (blue triangles) vs modelled (lines) results for experiment DEP2 Four different models are compared (1) Wettabletaking only the Kelvin effect into account (ieiSOA = 0) and the dry diameter of the particle (ds) equals the measured volume-equivalentdiameter (dvemeasured) (2) κ-Koumlhler(dm) accounts for the chemical composition but neglects the shape effect (ieds equals the mobilitydiameterdm) (3) in κ-Koumlhler (dvemeasured) both the chemical composition and shape effects are accounted for andds = dvemeasured and(4) for the modelκ-Koumlhler (dvefit) the particle diameter has been calculated by a fit function ofdvemeasuredand mfSOA (Eqs 13 14)

with a κ gt 01 (O C gt 06) In between (02 lt O C lt 06)most compounds are in the slightly soluble regime withlow κ values Kuwata et al (2013) also derived a mod-ified κ-Koumlhler equation accounting for sparingly solublecompounds The discrepancy in this study between modeland empirical results in the early stage of ageing when theamount of organic material in the particle and volume of wa-ter in the droplet is small (mfSOA(APM) lt 012 water volumelt 50 of the droplet) could possibly be explained by limita-tions in solubility (Fig 12) The investigation and modellingof semi-hygroscopic material to explain the diverging resultsis not within the scope of this study but should be a futurefocus

At the beginning of the experiments (before the onsetof UV exposure) only POA (andor processedtransformedPOA) is assumed to constitute the organic fraction consis-tent with the absence of CCN activation atsim 2 NeitherPOA nor soot is considered soluble (κ = 0) in this study(as discussed before) and therefore no activation of freshlyemitted particles is visible At the onset of UV exposure asmall quantity of SOA is produced reflected in the measure-ments as an immediate change of the particles into becom-ing better CCN Also the mfSOA(APM) was not negligibleeven if the organic material was only slightly hydrophilic iewater-soluble molecules were probably present at this pointThe organic material in the early ageing process is proba-bly either oxidised POA (heterogeneously before the onsetof UV or by OH andor ozone just after the onset of UV) orSOA formed from condensation of low-volatile organic com-

pounds (probably naphthalene) in the diesel exhaust (dis-cussed in Sect 52) By the end of the ageing process the for-mation of SOA fromm-xylene and toluene is dominating Ei-ther way the assumptions made for theκ-Koumlhler modellingin this study are too simple to explain the activation of theslightly coated soot particles (mfSOA(APM) lt 012) Eitherthe POA (just after UV onset) should be treated as slightlysoluble or SOA should be treated as less soluble

As the ageing proceeds more SOA condenses onto theparticles The SOA is considered hydrophilic in the model(κSOA = 013) which will enhance the ability of the soot par-ticles to activateκ values for ambient particles show highvariation depending on content assumed in calculations andinstrument used for observations egκ = 004ndash047 over theAmerican continent (Shinozuka et al 2009) 016ndash046 inGermany (Wu et al 2013) 022 for the oxygenated organiccomponent (Chang et al 2010) and 010ndash020 for Ama-zonian background aerosol (Rissler et al 2004 Gunthe etal 2009) An averageκ value ofsim 03 has been observedfor many continental locations (eg Andreae and Rosenfeld2008 Poumlschl et al 2009 Shinozuka et al 2009 Rose et al2010 Hersey et al 2013) The lower range ofκ values in theliterature corresponds to a higher content of organics whilethe higher values correspond to less organics and a highercontent of salts The low value ofκSOA(= 013) in this studycompared to urban aerosol hygroscopicity (κ asymp 03) could beattributed to the lack of salts which are not as present in thechamber experiments as in the atmosphere



However the calculatedκsum(= ε times κSOA) is in goodagreement withκCCN (derived from the CCN measurements)for all experiments (Supplement Fig S7b) In the literatureκ values for SOA formed from lubrication oil (κ asymp 0ndash002Lambe et al 2011) are in the same range asκsum andκCCNat the beginning of the ageing process By the end of theageing processκsum andκCCN are more similar toκ valuesof SOA formed fromm-xylene and toluene (κ asymp 01ndash027Lambe et al 2011)κsum is in the same range asκ val-ues from previous chamber studies of diesel exhaust (κ = 0ndash013 Tritscher et al 2011)κCCN andκsum deviate the mostfor the smallest and highestκ values (Fig S7b Supplement)Calculations ofκCCN are more uncertain at the beginningof the ageing process when theκ values are lowκSOA onthe other hand shows larger uncertainties at the end of theageing process due to uncertainties in fitted mfSOA(APM)values As mfSOA(APM) increasessc decreases in agree-ment with the models (κ-Koumlhler (dvemeasured) andκ-Koumlhler(dvefit)) and experimental results for mfSOA(APM) gt 012(Figs 11 and 12) Theκ value varies with the molar mass ofSOA For example a change ofplusmn0020 kg molminus1 in MSOAwould change theκ value withplusmn001 This would result ina change in critical supersaturation ofplusmn001ndash003 (depend-ing on mfSOA(APM) and size of the particle) still close toempirical results

To summarise in general the number of ions or water-soluble molecules in the particle determines the point of ac-tivation at a certain saturation However for the freshly emit-ted soot agglomerates this approach is only partly true Forthese particles where the un- or semi-soluble material of thedry particle makes up a large part of the volume fraction ofthe droplet during activation the material can have a pro-nounced effect on the activation properties The organic frac-tion the properties of this fraction and to some degree thesize of the particle (numbers I II and III) are represented intheory byκ ie representing the number of ions or water-soluble molecules in the particle Shape effects such as sizeand extent of agglomeration (numbers III and IV) also af-fect the CCN behaviour of the fresh and slightly processedsoot However this effect will ebb away as the water-soluble(organic) material coating the particles increases

54 Uncertainties

In the CCNC the high supersaturations required for mea-surements of fresh soot or early aged soot are hard to achieveFurthermore the first activation scans are difficult to evalu-ate and do not show full supersaturation spectra These scanshave therefore been excluded here but they still bring valu-able information of the early activation properties Also wecannot rule out biases from volatilisation of semi-volatilecompounds in the instrument even though the operationmode of SFCA minimises this effect As discussed by others(Asa-Awuku et al 2009 Frosch et al 2013) part of the ma-terial may volatilise inside the column of the instrument due

to the large temperatures required for high supersaturationThe effect would be a smaller size of the particle with lesshydrophilic content which would become less CCN activethough the loss of SOA becomes less probable as the dropletsform and gets diluted inside the instrument However the re-sults from this study sometimes show the opposite effect (egin Fig 11 there is a discrepancy between empirical resultsfor FSP2 from the two instruments) The two instrumentsmeasure at different temperature differences (1T ) and whenhigher temperatures are used the particles are more CCN ac-tive A possible explanation for this could be that the mea-surements are performed at the end of each calibration curve(ie low vs high flow for different1T ) and are thereforeassociated with larger error bars Another explanation couldbe that the activation occurs at different positions inside thecolumn this is something that should be investigated furtherand hence is not accounted for in this study

During measurements using the newly developed SFCAthe aerosol is subjected to the same temperature differencefor each supersaturation scan in each instrument (sometimesoverlapping) and thereby also more or less subjected to thesame volatilisation loss of the semi-volatile compounds Inthe original CFSTGC mode the temperature difference in-crease with increased supersaturation leading to differentvolatilisation losses during one supersaturation scan (whichcan be avoided by instead altering the particle diameter) Insummary using the SFCA mode in the CCNC during mea-surement minimises the volatilisation effect although wecannot rule out biases totally

A DMA-TD-APM set-up measures the massndashmobility re-lationship of the particles From these measurements thevolume-equivalent diameters (dvemeasured) as well as the or-ganic and soot mass fractions (mforg(APM) and mfBC(APM)respectively) are derived Thereafter the POA and SOA massfractions (mfPOA(APM) and mfSOA(APM) respectively) ofthe particles are estimated from mforg(APM) as described inthe Sect 41 Measurements of the organic fraction (POA)at the beginning of the experiments (before and just af-ter the onset of UV) are stable over size with an aver-age mforg(APM) = 3ndash4 In general the OA fraction (bothPOA and SOA) from DMA-TD-APM measurements differsslightly from that measured by SP-AMS with the assump-tions made in this study (for more information regarding theSP-AMS assumptions see the Supplement) Likely the dif-ference is attributed to the OA-properties in relation to thetwo measurement techniques and the quantification of BCin the SP-AMS For example POA may be strongly boundto or within the soot core and may thus be incompletely re-moved with the thermodenuder Further changes in Collec-tion Efficiency (for example laser ndash particle beam overlap) ofthe SP-AMS when transforming the soot from aggregated tospherical structure upon aging require further investigationThis should not have a significant effect on the modelling norwhen comparing empirical results with modelled ones In themodel the POA fraction is considered hydrophobic and the



same corrections are made for both the empirical and mod-elled results (as described below)

Uncertainties from the difficulties of measurements ofCCN properties and mass fractions of the particles are in-herited in the calculations and models Firstly the point ofactivation of particles inside the CCNC column is unknownTherefore the temperature at activation is uncertain Forthe largest used1T (= 18 K) the temperature ranges from29615 to 31415 K in the column As described before theabsolute temperature (T = 29815 K) and the surface tensionof water (σwater= 0072 N mminus1) are used for the calibrationand the model calculations These values show good agree-ment with empirical results and when compared to the tem-perature range in the column Secondly the SOA mass frac-tion is an approximation as described earlier An inaccuracyof mfSOA(APM) will be larger in the early part of the age-ing process than in the end and will have the same effect inthe model Thirdly for better resolution mfSOA(APM) datawere fitted with sigmoidal functions which has been usedin the model This approach could cause errors while mod-elling On the other hand the same functions have been usedfor plotting the cloud-activation data

6 Conclusions

Diesel exhaust aerosol and soot from a flame soot gener-ator spiked with light aromatic SOA precursors (m-xyleneand toluene) was photochemically aged The time-dependentchanges of the soot particle were characterised with respectto hygroscopic properties massndashmobility relationship andchemical composition with the main focus on CCN prop-erties

For fresh soot particles no activation into cloud droplets atsupersaturations lt 2 is observed It is unlikely that any sub-stantial SOA formation is taking place before the onset of UVradiation in dark conditions At the onset of UV exposure animmediate change in activation properties occurs with onlya small increase of the organic fraction coating the soot parti-cles At this point more hydrophilic (oxidised) organic com-pounds containing eg carbonyl alcohol carboxylic acidhydrogen peroxide nitrate and nitro functional groups areproduced Initially the SOA formation is probably domi-nated by low-volatile oxidation products formed from the re-actions between OH and IVOCs in the diesel exhausts How-ever within an hour after the onset of UV exposure morevolatile oxidation products formed from the addedm-xyleneand toluene also start to condense onto the soot particles Bythe end of the experiments (after 4ndash5 h of photochemical age-ing) the SOA is dominated bym-xylene and toluene oxida-tion products The instantaneous change in CCN propertiesat the onset of UV-radiation could be attributed to conden-sation of SOA or it could be an effect of oxidation of theorganic material already in the particle phase

Ageing of soot particles progressively enhances their abil-ity to act as CCN In summary the time-dependent changesof activation properties are attributed to the (I) organic frac-tion of the particle and (II) chemical properties of this frac-tion as well as the (III) size and (IV) morphology of the par-ticle Information of these four parameters (IndashIV) is highlyrelevant when predicting the activation point of the slightlyprocessed soot However as the soot ages the shape (IV) ef-fects diminishes

As expected there is a size dependency of the activationproperties as well as for the mass acquisition of secondaryorganic material ndash two effects affecting the activation in thisexperiment Smaller particles are harder to activate into clouddroplets although slightly higher relative mass acquisition ofSOA are observed for smaller sizes The results also indicatethat the size of the primary particles and morphology of thefresh soot core might be of importance Aggregates consist-ing of primary particles with a smaller diameter (dpp) requirea higher supersaturation than aggregates made of larger pri-mary particles at the same SOA mass fraction On the otherhand aggregates with smallerdpp activate at lower super-saturation compared to particles with largerdpp at certainvolume-equivalent diameter (dve)

POA has been treated as hydrophobic in the CCN mod-elling not contributing to any CCN activity SOA on theother hand enhances the ability of the soot particles to act as aCCN with increasing amount of condensing material Theseassumptions seem to be to modest for modelling the clouddroplet formation in the early ageing process As discussedin Sect 52 the POA may undergo heterogeneous oxidationbefore the onset of UV exposure or be oxidised by OH orozone which may increase the hygroscopicity of the materialndash in which case POA (or OPOA) should be treated as slightlyhygroscopic The initial SOA formation might be from con-densation of low-volatile organic compounds in the dieselexhaust The chemical composition would then differ withslightly different hygroscopic properties than SOA formedfrom the added precursors as a result Hence the early SOAmight not be as hygroscopic as the aged one However theinstantaneous change in CCN properties can most likely beattributed to condensation of SOA A strong increase of SOAis seen at the beginning of the experiments accompanied bya similar trend for the increasing volume-equivalent diameter(dve) and a decrease in critical water vapour supersaturationsc

The decline insc required to activate the organic coatedsoot particles can be modelled byκ-Koumlhler theory account-ing for the agglomerated structure of the particles Informa-tion of the volume-equivalent diameter (dve) or shape factor(χ) is necessary for a good representation Also needed isthe chemical composition (egM ρ i) and amount of theorganic material (mf orε) If dm and mfSOA are knowndvecan be calculated from the empirically fitted functions Themodel captures the evolution of the activation properties wellfor mfSOA(APM) gt 012 Modelling ofκsum (Eq 7) based



on κSOA = 013 shows good agreement withκCCN (Eqs 9and 10) with largest deviations for the lowesthighest values(κ lt 0015 andκ gt 0085 in Fig S7b Supplement) Modelledsc whenκSOA is used for the calculations corresponds wellto thesc measured at the end of the experiments for all sizes(sc lt 1 )

The model does not capture the early steep decrease ofthesc(for mfSOA(APM) lt 012) In reality the slightly coatedsoot particles are not as good CCN as in the model whichcould be explained by a semi-hydrophilic organic layer (witha vanrsquot Hoff factor (i) andorκSOA value equal or close tozero) hindering effects by shape andor unevenly distributedorganic material A limitation in solubility could be an im-portant parameter for CCN activity of atmospheric aerosolparticles

The immediate change in CCN activation at the onset ofUV exposure implies that the lifetime of soot in the atmo-sphere is affected by access to sunlight Reduced photochem-istry as in wintertime in the Northern Hemisphere couldmean a longer residence time in the atmosphere for soot par-ticles due to prolonged hydrophobicity That is soot particlesmay perturb the radiation budget in the Arctic region due toa longer residence time of the aged particles with enhancedadsorption properties In the summer when sunlight is plen-tiful the same soot particles may age rapidly and due to en-hanced hygroscopicity of the particles the cloud droplet num-ber concentration may increase changing the cloud coverSuch a change could influence the radiation budget of the re-gion On the other hand an increase in hygroscopicity couldresult in enhanced deposition During summer photochem-ical ageing as well as wet removal have been pointed outas having central roles in controlling the properties of theaerosol size distribution Deposition of soot on the snow sur-face may in turn change the albedo of the snow surface andthereby influence the climate of the Arctic region

Change in hygroscopicity and morphology of the ageingsoot particles will also affect the deposition of the parti-cles in the human respiratory tract according to Loumlndahl etal (2009)



Appendix A

Table A1 Nomenclature

Nomenclature

aw Water activityADCHAM Aerosol Dynamics gas- and particle-phase chemistry model for laboratory CHAMber studiesBC Black CarbonCc Cunningham correction factorCCN Cloud condensation nucleiCCNC Cloud condensation nuclei counterCFSTGC Continuous-flow streamwise thermal-gradient counterCPC Condensational particle counterDEP Diesel exhaust particledm Mobility diameterdmdry Mobility diameter of the dry particleDMA-TD-APM Differential mobility analyserndashaerosol particle mass analyser with a thermodenuderdpp Primary particle diameterds Diameter of the dry particleDSOA Dark Secondary organic aerosoldve Volume-equivalent diameterdvefit Calculated volume-equivalent diameter from empirical fitdvemeasured Volume-equivalent diameter from measurementdwet Diameter of the spherical aqueous solution dropletFSP Flame soot generator particleGMD Geometric mean diameterHR-ToF-AMS High-resolution time-of-flight mass spectrometeri vanrsquot Hoff factor represents the effects of ion interactions and dissociationIVOC Intermediate volatility organic compoundKe Kelvin effectLV-OOA Low-volatility oxygenated organic aerosolmforg(APM) Mass fraction organic from APM measurements (massndashmobility relationship)mfPOA(APM) Mass fraction POA from APM measurements (massndashmobility relationship)mfSOA(AMS) Mass fraction SOA from AMS measurements (chemical composition)mfSOA(APM) Mass fraction SOA from APM measurements (massndashmobility relationship)mfBC(APM) Mass fraction soot from APM measurements (massndashmobility relationship)Mi Molecular mass for a specific componentnsum Sum of the different contributing components in the particlesO C ratio Oxygen to carbon ratioOA Organic aerosolOC Organic carbonOPC Optical particle counterOPOA Oxygenated primary organic aerosolp Actual partial pressureP Pressurepo Equilibrium pressure over a flat surface of pure waterPOA Primary organic aerosolQ50 Critical flow rateR Universal gas constants Saturation ratiosc Critical supersaturationsc 52 Critical supersaturation measured by instrument no 52SFCA Scanning flow CCN analysisSMPS Scanning mobility particle sizerSOA Secondary organic aerosolSP-AMS Soot particle aerosol mass spectrometerSV-OOA Semi-volatile oxygenated organic aerosolT Absolute temperatureTEM Transmission electron microscopeVOC Volatile organic compoundWettable Model taking only the Kelvin effect into account1T Streamwise temperature gradientεi Volume fraction of a specific component in the dry particleκ Hygroscopicity parameter describing the number of ions or non-dissociating molecules per unit volume of the dry particleκ-Koumlhler Model usingκ-Koumlhler theoryκ-Koumlhler(dm) Model usingκ-Koumlhler theory with thedm as input parameter fordsκ-Koumlhler(dvefit) Model usingκ-Koumlhler theory with the fitteddve as input parameter fordsκ-Koumlhler(dvemeasured) Model usingκ-Koumlhler theory with the empirically deriveddve as input parameter fordsκCCN Hygroscopicity parameter value derived from the measured critical supersaturationκSOA Hygroscopicity parameter value calculated from chemical composition of SOAν Dissociation numberρcorr Corrected material density of the particlesρi Density of a specific componentρw Density of waterσg Standard deviation of the geometric mean diameter from the lognormal number size distributionσsol Surface tension of the solution dropletσw Surface tension of waterφ Molal or practical osmotic coefficient of the solute in aqueous solutionχ Shape factor



The Supplement related to this article is available onlineat doi105194acp-14-9831-2014-supplement

AcknowledgementsThe Swedish Research Council (VR) TheSwedish Research Council Formas ModElling the Regionaland Global Earth system ndash MERGE ClimBEco research school(Lund University) Aerosols Clouds and Trace gases ResearchInfraStructure Network ndash ACTRIS and Cryosphere-atmosphereinteractions in a changing Arctic climate ndash CRAICC supported thiswork

Edited by H Skov

References

Ackerman A S Toon O Stevens D Heymsfield A Ra-manathan V and Welton E Reduction of tropical cloudinessby soot Science 288 1042ndash1047 2000

Albrecht B A Aerosols cloud microphysics and fractionalcloudiness Science 245 1227ndash1230 1989

Andreae M O and Rosenfeld D Aerosol-cloud-precipitationinteractions Part 1 The nature and sources of cloud-activeaerosols Earth-Sci Rev 89 13ndash41 2008

Arens F Gutzwiller L Baltensperger U Gaggeler H W andAmmann M Heterogeneous reaction of NO2 on diesel soot par-ticles Environ Sci Technol 35 2191ndash2199 2001

Asa-Awuku A Engelhart G J Lee B H Pandis S N andNenes A Relating CCN activity volatility and droplet growthkinetics ofβ-caryophyllene secondary organic aerosol AtmosChem Phys 9 795ndash812 doi105194acp-9-795-2009 2009

Bilde M and Svenningsson B CCN activation of slightly solubleorganics the importance of small amounts of inorganic salt andparticle phase Tellus B 56 128ndash134 2004

Bond T C Doherty S J Fahey D W Forster P M BerntsenT DeAngelo B J Flanner M G Ghan S Karcher B KochD Kinne S Kondo Y Quinn P K Sarofim M C SchultzM G Schulz M Venkataraman C Zhang H Zhang S Bel-louin N Guttikunda S K Hopke P K Jacobson M ZKaiser J W Klimont Z Lohmann U Schwarz J P Shin-dell D Storelvmo T Warren S G and Zender C S Bound-ing the role of black carbon in the climate system A scientificassessment J Geophys Res-Atmos 118 5380ndash5552 2013

Canagaratna M R Jayne J T Jimenez J L Allan J D Al-farra M R Zhang Q Onasch T B Drewnick F Coe HMiddlebrook A Delia A Williams L R Trimborn A MNorthway M J DeCarlo P F Kolb C E Davidovits P andWorsnop D R Chemical and microphysical characterization ofambient aerosols with the aerodyne aerosol mass spectrometerMass Spectrom Rev 26 185ndash222 2007

Cappa C D Onasch T B Massoli P Worsnop D R Bates TS Cross E S Davidovits P Hakala J Hayden K L Job-son B T Kolesar K R Lack D A Lerner B M Li S MMellon D Nuaaman I Olfert J S Petaja T Quinn P KSong C Subramanian R Williams E J and Zaveri R ARadiative Absorption Enhancements Due to the Mixing State ofAtmospheric Black Carbon Science 337 1078ndash1081 2012

Chan A W H Kautzman K E Chhabra P S Surratt J DChan M N Crounse J D Kuumlrten A Wennberg P OFlagan R C and Seinfeld J H Secondary organic aerosolformation from photooxidation of naphthalene and alkylnaph-thalenes implications for oxidation of intermediate volatility or-ganic compounds (IVOCs) Atmos Chem Phys 9 3049ndash3060doi105194acp-9-3049-2009 2009

Chang R Y-W Slowik J G Shantz N C Vlasenko A LiggioJ Sjostedt S J Leaitch W R and Abbatt J P D The hy-groscopicity parameter (κ) of ambient organic aerosol at a fieldsite subject to biogenic and anthropogenic influences relation-ship to degree of aerosol oxidation Atmos Chem Phys 105047ndash5064 doi105194acp-10-5047-2010 2010

Choi M Y A Hamins A Mulholland G W and KashiwagiT Simultaneous Optical Measurements of Soot Volume Fractionand Temperature in Premixed Flames Combustion Flame 99174ndash186 1994

Cross E S Slowik J G Davidovits P Allan J D Worsnop DR Jayne J T Lewis D K Canagaratna M and Onasch TB Laboratory and ambient particle density determinations usinglight scattering in conjunction with aerosol mass spectrometryAerosol Sci Tech 41 343ndash359 2007

DeCarlo P F Slowik J G Worsnop D R Davidovits P andJimenez J L Particle morphology and density characterizationby combined mobility and aerodynamic diameter measurementsPart 1 Theory Aerosol Sci Tech 38 1185ndash1205 2004

DeCarlo P F Kimmel J R Trimborn A Northway M J JayneJ T Aiken A C Gonin M Fuhrer K Horvath T andDocherty K S Field-deployable high-resolution time-of-flightaerosol mass spectrometer Anal Chem 78 8281ndash8289 2006

Duplissy J Gysel M Alfarra M R Dommen J Metzger APrevot A S H Weingartner E Laaksonen A RaatikainenT Good N Turner S F McFiggans G and BaltenspergerU Cloud forming potential of secondary organic aerosol undernear atmospheric conditions Geophys Res Lett 35 L03818doi1010292007gl031075 2008

Dusek U Frank G P Hildebrandt L Curtius J Schneider JWalter S Chand D Drewnick F Hings S Jung D Bor-rmann S and Andreae M O Size matters more than chem-istry for cloud-nucleating ability of aerosol particles Science312 1375ndash1378 2006

Donahue N M Robinson A L and Pandis S NAtmospheric organic particulate matter From smoke tosecondary organic aerosol Atmos Environ 43 94ndash106doi101016jatmosenv200809055 2009

Florence A T Attwood D and Attwood D Physicochemicalprinciples of pharmacy Pharmaceutical Press TJ InternationalPadstow Cornwall UK 2011

Frosch M Bilde M Nenes A Praplan A P Juraacutenyi Z Dom-men J Gysel M Weingartner E and Baltensperger U CCNactivity and volatility of β-caryophyllene secondary organicaerosol Atmos Chem Phys 13 2283ndash2297 doi105194acp-13-2283-2013 2013

Gunthe S S King S M Rose D Chen Q Roldin P FarmerD K Jimenez J L Artaxo P Andreae M O Martin ST and Poumlschl U Cloud condensation nuclei in pristine tropi-cal rainforest air of Amazonia size-resolved measurements andmodeling of atmospheric aerosol composition and CCN activity



Atmos Chem Phys 9 7551ndash7575 doi105194acp-9-7551-2009 2009

Gysel M Nyeki S Weingartner E Baltensperger U Giebl HHitzenberger R Petzold A and Wilson C W Properties ofjet engine combustion particles during the PartEmis experimentHygroscopicity at subsaturated conditions Geophys Res Lett30 1566 doi1010292003gl016896 2003

Hallquist M Wenger J C Baltensperger U Rudich Y Simp-son D Claeys M Dommen J Donahue N M GeorgeC Goldstein A H Hamilton J F Herrmann H Hoff-mann T Iinuma Y Jang M Jenkin M E Jimenez J LKiendler-Scharr A Maenhaut W McFiggans G Mentel ThF Monod A Preacutevocirct A S H Seinfeld J H Surratt J DSzmigielski R and Wildt J The formation properties and im-pact of secondary organic aerosol current and emerging issuesAtmos Chem Phys 9 5155ndash5236 doi105194acp-9-5155-2009 2009

Han C Liu Y C and He H Role of Organic Carbon in Hetero-geneous Reaction of NO2 with Soot Environ Sci Technol 473174ndash3181 2013

Hart J E Laden F Eisen E A Smith T J and GarshickE Chronic obstructive pulmonary disease mortality in railroadworkers Occup Environ Med 66 221ndash226 2009

Henning S Ziese M Kiselev A Saathoff H Moumlhler OMentel T F Buchholz A Spindler C Michaud V MonierM Sellegri K and Stratmann F Hygroscopic growth anddroplet activation of soot particles uncoated succinic orsulfuric acid coated Atmos Chem Phys 12 4525ndash4537doi105194acp-12-4525-2012 2012

Hersey S P Craven J S Metcalf A R Lin J Lathem T SuskiK J Cahill J F Duong H T Sorooshian A Jonsson H HShiraiwa M Zuend A Nenes A Prather K A Flagan R Cand Seinfeld J H Composition and hygroscopicity of the LosAngeles Aerosol CalNex J Geophys Res-Atmos 118 3016ndash3036 2013

Huff Hartz K E Rosenoslashrn T Ferchak S R Raymond T MBilde M Donahue N M and Pandis S N Cloud conden-sation nuclei activation of monoterpene and sesquiterpene sec-ondary organic aerosol J Geophys Res-Atmos 110 D14208doi1010292004JD005754 2005

IPCC Climate Change 2007 The Physical Science Basis Contri-bution of Working Group I to the Fourth Assessment Report ofthe Intergovernmental Panel on Climate Change 2007

IPCC Climate Change 2013 The Physical Science Basis Work-ing Group I Contribution to the IPCC 5th Assessment Reportndash Changes to the Underlying ScientificTechnical Assessment(IPCC-XXVIDoc4) 2013

Jenkin M E Saunders S M Wagner V and Pilling M JProtocol for the development of the Master Chemical Mecha-nism MCM v3 (Part B) tropospheric degradation of aromaticvolatile organic compounds Atmos Chem Phys 3 181ndash193doi105194acp-3-181-2003 2003

Jimenez J L Canagaratna M R Donahue N M Prevot A SH Zhang Q Kroll J H DeCarlo P F Allan J D CoeH Ng N L Aiken A C Docherty K S Ulbrich I MGrieshop A P Robinson A L Duplissy J Smith J D Wil-son K R Lanz V A Hueglin C Sun Y L Tian J Laak-sonen A Raatikainen T Rautiainen J Vaattovaara P EhnM Kulmala M Tomlinson J M Collins D R Cubison M

J Dunlea E J Huffman J A Onasch T B Alfarra M RWilliams P I Bower K Kondo Y Schneider J DrewnickF Borrmann S Weimer S Demerjian K Salcedo D Cot-trell L Griffin R Takami A Miyoshi T Hatakeyama SShimono A Sun J Y Zhang Y M Dzepina K Kimmel JR Sueper D Jayne J T Herndon S C Trimborn A MWilliams L R Wood E C Middlebrook A M Kolb CE Baltensperger U and Worsnop D R Evolution of OrganicAerosols in the Atmosphere Science 326 1525ndash1529 2009

Kanakidou M Seinfeld J H Pandis S N Barnes I DentenerF J Facchini M C Van Dingenen R Ervens B Nenes ANielsen C J Swietlicki E Putaud J P Balkanski Y FuzziS Horth J Moortgat G K Winterhalter R Myhre C EL Tsigaridis K Vignati E Stephanou E G and WilsonJ Organic aerosol and global climate modelling a review At-mos Chem Phys 5 1053ndash1123 doi105194acp-5-1053-20052005

Khalizov A F Zhang R Y Zhang D Xue H X Pagels J andMcMurry P H Formation of highly hygroscopic soot aerosolsupon internal mixing with sulfuric acid vapor J Geophys Res-Atmos 114 doi1010292008JD010595 2009

Kreidenweis S M Koehler K DeMott P J Prenni A JCarrico C and Ervens B Water activity and activation di-ameters from hygroscopicity data ndash Part I Theory and appli-cation to inorganic salts Atmos Chem Phys 5 1357ndash1370doi105194acp-5-1357-2005 2005

Kroll J H Donahue N M Jimenez J L Kessler S H Cana-garatna M R Wilson K R Altieri K E Mazzoleni L RWozniak A S Bluhm H Mysak E R Smith J D KolbC E and Worsnop D R Carbon oxidation state as a metricfor describing the chemistry of atmospheric organic aerosol NatChem 3 133ndash139 2011

Kundu S Fisseha R Putman A L Rahn T A and Maz-zoleni L R High molecular weight SOA formation duringlimonene ozonolysis insights from ultrahigh-resolution FT-ICRmass spectrometry characterization Atmos Chem Phys 125523ndash5536 doi105194acp-12-5523-2012 2012

Kuwata M Shao W Lebouteiller R and Martin S T Classify-ing organic materials by oxygen-to-carbon elemental ratio to pre-dict the activation regime of Cloud Condensation Nuclei (CCN)Atmos Chem Phys 13 5309ndash5324 doi105194acp-13-5309-2013 2013

Koumlhler H The Nucleus in and the Growth of HygroscopicDroplets T Faraday Soc 32 1152ndash1161 1936

Laaksonen A Korhonen P Kulmala M and Charlson R JModification of the Kuhler equation to include soluble tracegases and slightly soluble substances J Atmos Sci 55 853ndash862 1998

Lambe A T Onasch T B Massoli P Croasdale D R WrightJ P Ahern A T Williams L R Worsnop D R Brune W Hand Davidovits P Laboratory studies of the chemical composi-tion and cloud condensation nuclei (CCN) activity of secondaryorganic aerosol (SOA) and oxidized primary organic aerosol(OPOA) Atmos Chem Phys 11 8913ndash8928 doi105194acp-11-8913-2011 2011

Lance S Medina J Smith J N and Nenes A Mapping theoperation of the DMT Continuous Flow CCN counter AerosolSci Tech 40 242ndash254 2006



Lide D R Physical Constants of Organic Compounds inCRC Handbook of Chemistry and Physics Internet Version2005 CRC Press Boca Raton FL available athttpwwwhbcpnetbasecom(last access 10 December 2013) 2005

Loumlndahl J Pagels J Boman C Swietlicki E Massling ARissler J Blomberg A Bohgard M and Sandstroumlm TDeposition of biomass combustion aerosol particles in the hu-man respiratory tract Inhal Toxicol 20 923ndash933 2008

Loumlndahl J Massling A Swietlicki E Braumluner E V Ketzel MPagels J and Loft S Experimentally determined human respi-ratory tract deposition of airborne particles at a busy street Env-iron Sci Technol 43 4659ndash4664 2009

Malik A Abdulhamid H Pagels J Rissler J Lindskog MNilsson P Bjorklund R Jozsa P Visser J Spetz A andSanati M A Potential Soot Mass Determination Method fromResistivity Measurement of Thermophoretically Deposited SootAerosol Sci Tech 45 284ndash294 2011

Maricq M M and Ning X The effective density and fractal di-mension of soot particles from premixed flames and motor vehi-cle exhaust J Aerosol Sci 35 1251ndash1274 2004

McMurry P H Wang X Park K and Ehara K The relation-ship between mass and mobility for atmospheric particles A newtechnique for measuring particle density Aerosol Sci Tech 36227ndash238 2002

Meyer N K and Ristovski Z D Ternary nucleation as a mech-anism for the production of diesel nanoparticles Experimentalanalysis of the volatile and hygroscopic properties of diesel ex-haust using the volatilization and humidification tandem differ-ential mobility analyzer Environ Sci Technol 41 7309ndash73142007

Mills N L Toumlrnqvist H Gonzalez M C Vink E RobinsonS D Soumlderberg S Boon N A Donaldson K SandstroumlmT and Blomberg A Ischemic and thrombotic effects of di-lute diesel-exhaust inhalation in men with coronary heart diseaseNew Engl J Med 357 1075ndash1082 2007

Moore R H and Nenes A Scanning Flow CCN Analysis ndash AMethod for Fast Measurements of CCN Spectra Aerosol SciTech 43 1192ndash1207 2009

Nakao S Shrivastava M Nguyen A Jung H and Cocker DInterpretation of Secondary Organic Aerosol Formation fromDiesel Exhaust Photooxidation in an Environmental ChamberAerosol Sci Tech 45 964ndash972 2011

Ng N L Kroll J H Chan A W H Chhabra P S FlaganR C and Seinfeld J H Secondary organic aerosol formationfrom m-xylene toluene and benzene Atmos Chem Phys 73909ndash3922 doi105194acp-7-3909-2007 2007

Nordin E Z Eriksson A C Roldin P Nilsson P T CarlssonJ E Kajos M K Helleacuten H Wittbom C Rissler J LoumlndahlJ Swietlicki E Svenningsson B Bohgard M Kulmala MHallquist M and Pagels J H Secondary organic aerosol for-mation from idling gasoline passenger vehicle emissions investi-gated in a smog chamber Atmos Chem Phys 13 6101ndash6116doi105194acp-13-6101-2013 2013

Onasch T Trimborn A Fortner E Jayne J Kok G WilliamsL Davidovits P and Worsnop D Soot particle aerosol massspectrometer development validation and initial applicationAerosol Sci Tech 46 804ndash817 2012

Pagels J Khalizov A F McMurry P H and Zhang R Y Pro-cessing of Soot by Controlled Sulphuric Acid and Water Conden-

sationMass and Mobility Relationship Aerosol Sci Tech 43629ndash640 2009

Park K Cao F Kittelson D B and McMurry P H Relation-ship between particle mass and mobility for diesel exhaust parti-cles Environ Sci Technol 37 577ndash583 2003

Petters M D and Kreidenweis S M A single parameter repre-sentation of hygroscopic growth and cloud condensation nucleusactivity Atmos Chem Phys 7 1961ndash1971 doi105194acp-7-1961-2007 2007

Petters M D and Kreidenweis S M A single parameter repre-sentation of hygroscopic growth and cloud condensation nucleusactivity ndash Part 2 Including solubility Atmos Chem Phys 86273ndash6279 doi105194acp-8-6273-2008 2008

Petzold A Ogren J A Fiebig M Laj P Li S-M Bal-tensperger U Holzer-Popp T Kinne S Pappalardo G Sug-imoto N Wehrli C Wiedensohler A and Zhang X-Y Rec-ommendations for reporting ldquoblack carbonrdquo measurements At-mos Chem Phys 13 8365ndash8379 doi105194acp-13-8365-2013 2013

Pincus R and Baker M B Effect of precipitation on the albedosusceptibility of clouds in the marine boundary layer Nature372 250ndash252 1994

Poumlschl U Rose D and Andreae M O Part 2 Particle Hygro-scopicity and Cloud Condensation Nucleus Activity Clouds inthe Perturbed Climate System Their Relationship to Energy Bal-ance Atmospheric Dynamics and Precipitation 58ndash72 2009

Prenni A J Petters M D Kreidenweis S M DeMottP J and Ziemann P J Cloud droplet activation of sec-ondary organic aerosol J Geophys Res-Atmos 112 D10223doi1010292006jd007963 2007

Pruppacher H R and Klett J D Microphysics of Clouds andPrecipitation Atmospheric and oceanographic sciences libraryKluwer Academic Puplishers Dordrecht the Netherlands 2ndEdn ISBN0-7923-4211-9 1997

Rissler J Swietlicki E Zhou J Roberts G Andreae M OGatti L V and Artaxo P Physical properties of the sub-micrometer aerosol over the Amazon rain forest during the wet-to-dry season transition ndash comparison of modeled and mea-sured CCN concentrations Atmos Chem Phys 4 2119ndash2143doi105194acp-4-2119-2004 2004

Rissler J Vestin A Swietlicki E Fisch G Zhou J ArtaxoP and Andreae M O Size distribution and hygroscopic prop-erties of aerosol particles from dry-season biomass burning inAmazonia Atmos Chem Phys 6 471ndash491 doi105194acp-6-471-2006 2006

Rissler J Svenningsson B Fors E O Bilde M and Swi-etlicki E An evaluation and comparison of cloud conden-sation nucleus activity models predicting particle critical sat-urationfrom growth at subsaturation J Geophys Res 115D22208 doi1010292010jd014391 2010

Rissler J Swietlicki E Bengtsson A Boman C Pagels JSandstroumlm T Blomberg A and Loumlndahl J Experimental de-termination of deposition of diesel exhaust particles in the humanrespiratory tract J Aerosol Sci 48 18ndash33 2012

Rissler J Messing M E Malik A I Nilsson P T Nordin EZ Bohgard M Sanati M and Pagels J H Effective DensityCharacterization of Soot Agglomerates from Various Sourcesand Comparison to Aggregation Theory Aerosol Sci Tech 47792ndash805 2013



Rissler J Nordin E Z Eriksson A C Nilsson P T Frosch MSporre M K Wierzbicka A Svenningsson B Loumlndahl JMessing M E Sjogren S Hemmingsen J G Loft S PagelsJ H and Swietlicki E Effective density and mixing state ofaerosol particles in a near-traffic urban environment EnvironSci Technol 48 6300ndash6308 doi101021es5000353 2014

Ristimaki J Vaaraslahti K Lappi M and Keskinen J Hydro-carbon condensation in heavy-duty diesel exhaust Environ SciTechnol 41 6397ndash6402 2007

Roberts G C and Nenes A A Continuous-Flow StreamwiseThermal-Gradient CCN Chamber for Atmospheric Measure-ments Aerosol Sci Tech 39 206ndash221 2005

Roldin P Eriksson A C Nordin E Z Hermansson E Mo-gensen D Rusanen A Boy M Swietlicki E Svennings-son B Zelenyuk A and Pagels J Modelling non-equilibriumsecondary organic aerosol formation and evaporation with theaerosol dynamics gas- and particle-phase chemistry kinetic mul-tilayer model ADCHAM Atmos Chem Phys 14 7953ndash7993doi105194acp-14-7953-2014 2014

Rose D Wehner B Ketzel M Engler C Voigtlaumlnder J TuchT and Wiedensohler A Atmospheric number size distribu-tions of soot particles and estimation of emission factors At-mos Chem Phys 6 1021ndash1031 doi105194acp-6-1021-20062006

Rose D Gunthe S S Mikhailov E Frank G P Dusek UAndreae M O and Poumlschl U Calibration and measurementuncertainties of a continuous-flow cloud condensation nucleicounter (DMT-CCNC) CCN activation of ammonium sulfateand sodium chloride aerosol particles in theory and experimentAtmos Chem Phys 8 1153ndash1179 doi105194acp-8-1153-2008 2008

Rose D Nowak A Achtert P Wiedensohler A Hu M ShaoM Zhang Y Andreae M O and Poumlschl U Cloud conden-sation nuclei in polluted air and biomass burning smoke near themega-city Guangzhou China ndash Part 1 Size-resolved measure-ments and implications for the modeling of aerosol particle hy-groscopicity and CCN activity Atmos Chem Phys 10 3365ndash3383 doi105194acp-10-3365-2010 2010

Schauer J J Kleeman M J Cass G R and Simoneit B RT Measurement of emissions from air pollution sources 2 C-1 through C-30 organic compounds from medium duty dieseltrucks Environ Sci Technol 33 1578ndash1587 1999

Seinfeld J H and Pandis S N Atmospheric chemistry andphysics From air pollution to climate change 2nd Edn JohnWiley amp Sons Inc Hoboken New Jersey 2006

Shinozuka Y Clarke A D DeCarlo P F Jimenez J L Dun-lea E J Roberts G C Tomlinson J M Collins D R How-ell S G Kapustin V N McNaughton C S and Zhou JAerosol optical properties relevant to regional remote sensing ofCCN activity and links to their organic mass fraction airborneobservations over Central Mexico and the US West Coast dur-ing MILAGROINTEX-B Atmos Chem Phys 9 6727ndash6742doi105194acp-9-6727-2009 2009

Stubington J F Sergeant G D Barrett D Do P T D Hand Raval K A Molecular-Weight Determination in the Studyof the Lubricating Oil Potential of Shale Oils Fuel 74 79ndash821995

Svenningsson B Rissler J Swietlicki E Mircea M Bilde MFacchini M C Decesari S Fuzzi S Zhou J Moslashnster J andRosenoslashrn T Hygroscopic growth and critical supersaturationsfor mixed aerosol particles of inorganic and organic compoundsof atmospheric relevance Atmos Chem Phys 6 1937ndash1952doi105194acp-6-1937-2006 2006

Sydbom A Blomberg A Parnia S Stenfors N Sandstroumlm Tand Dahlen S Health effects of diesel exhaust emissions Euro-pean Respiratory Journal 17 733ndash746 2001

Tree D R and Svensson K I Soot processes in compression ig-nition engines Prog Energ Combust 33 272ndash309 2007

Tritscher T Juraacutenyi Z Martin M Chirico R Gysel MHeringa M F DeCarlo P F Sierau B Preacutevocirct A S H Wein-gartner E and Baltensperger U Changes of hygroscopicityand morphology during ageing of diesel soot Environ Res Lett6 034026 doi1010881748-932663034026 2011

Tunved P Stroumlm J and Krejci R Arctic aerosol life cycle link-ing aerosol size distributions observed between 2000 and 2010with air mass transport and precipitation at Zeppelin stationNy-Aringlesund Svalbard Atmos Chem Phys 13 3643ndash3660doi105194acp-13-3643-2013 2013

Twomey S Pollution and the planetary albedo Atmos Environ8 1251ndash1256 1974

Weingartner E Burtscher H and Baltensperger U Hygroscopicproperties of carbon and diesel soot particles Atmos Environ31 2311ndash2327 1997

WHO Public Health Round-Up Diesel Exhaust CarcinogenicBulletin of the World Health Organization 477ndash556 availableathttpwwwncbinlmnihgovpmcarticlesPMC3397708(lastaccess 16 August 2012) 2012

Wu Z J Poulain L Henning S Dieckmann K Birmili WMerkel M van Pinxteren D Spindler G Muumlller K Strat-mann F Herrmann H and Wiedensohler A Relating particlehygroscopicity and CCN activity to chemical composition duringthe HCCT-2010 field campaign Atmos Chem Phys 13 7983ndash7996 doi105194acp-13-7983-2013 2013

Zhang R Khalizov A F Pagels J Zhang D Xue H and Mc-Murry P H Variability in morphology hygroscopicity and op-tical properties of soot aerosols during atmospheric processingP Natl Acad Sci USA 105 10291ndash10296 2008



as the ambient water vapour supersaturation The indirectaerosol effect includes increase in cloud albedo due to ad-dition of cloud nuclei by pollution (Twomey 1974) reduc-tion of drizzle and increased cloud lifetime (Albrecht 1989)as well as an increase in cloud thickness (Pincus and Baker1994) Also soot aerosol can contribute to daytime clearingof clouds (Ackerman et al 2000)

Soot particles make up a large fraction by number of theatmospheric aerosol especially in urban locations in the sizerange lt 100 nm (Rose et al 2006 and references therein)Freshly emitted soot particles or black carbon (BC) areknown to have a predominantly warming effect on the cli-mate due to their ability to absorb light referred to as a di-rect aerosol effect (IPCC 2007) or the radiative forcing fromaerosolndashradiation interactions (RFari IPCC 2013) Also theabsorption may increase with photochemical ageing and wa-ter uptake (Zhang et al 2008 Cappa et al 2012) AfterCO2 BC is estimated as the strongest anthropogenic climate-forcing agent in the present-day atmosphere (Bond et al2013) with a total warming effect of about+11 W mminus2

(with 90 uncertainty bounds of+017 to 21 W mminus2)However BC also has an effect on the Earthrsquos hydrologicalcycle According to Bond et al (2013) the largest uncertain-ties in net climate forcing are related to the lack of knowl-edge about cloud interactions with BC when co-emitted withorganic carbon (OC)

Soot is not only present close to sources but can be trans-ported long distances The warming effect of soot in the Arc-tic has recently gained increased interest both due to trans-port of soot from urban areas and also due to possible fu-ture soot emissions in the Arctic area from ship traffic whichis made possible due to the reduction in ice coverage Theenhanced light absorption by soot due to condensing mate-rial during photochemical ageing may influence and boostthe warming of the surroundings even further Also soot de-posited on snow and ice may enhance surface heating andice melting by decreasing the surface albedo (eg Bond etal 2013 Tunved et al 2013 and references therein) Theparticle number concentration is in general low (about a cou-ple of hundred cmminus3) in the Arctic (Tunved et al 2013)Wet removal and photochemical ageing play central roles incontrolling the aerosol size distribution properties during theArctic summer As pointed out by Tunved et al (2013) it ishighly relevant to obtain a better understanding of the micro-physical properties of the different aerosol populations abun-dant in the Arctic environment in order to improve the under-standing of the direction and magnitude of a future climatechange in the region

According to the World Health Organization (WHO2012) emissions of diesel engine exhaust are classified ascarcinogenic in humans in addition to being linked withclimate change The unfavourable health effects caused byexposure to diesel exhaust are described in several studies(eg Sydbom et al 2001 Mills et al 2007 Hart et al2009) The deposited fraction in the human respiratory sys-

tem is well described by the mobility diameter (for particleslt 400 nm) whilst deposited dose by surface area and mass re-quire knowledge of the characteristics of the particles due totheir agglomerated structure (Rissler et al 2012) The parti-cle lung deposition is substantially altered by hygroscopicity(Loumlndahl et al 2009) A change towards more hygroscopicproperties will shift the deposited fraction in the respiratorytract towards larger sizes Photochemical processing of dieselexhaust particles thereby alters the uptake in the human res-piratory system due to enhanced hygroscopicity

Diesel exhaust aerosol is formed during combustion pro-cesses mainly of refractory carbonaceous material (BC) thatis highly agglomerated primary organic compounds in theparticle phase and volatile organic compounds (VOCs) in thegas phase There is no clear definition for soot formed fromincomplete combustion however in general soot consists ofroughly eight parts of carbon and one part of hydrogen byatoms (Tree and Svensson 2007) In this study the term sootparticles refers to the agglomerated particles emitted fromdiesel vehicles or by a soot generator including both BC andorganic carbon (Petzold et al 2013) Diesel exhaust particles(DEPs) and soot generator particles (FSPs) refer to particlesfrom the diesel vehicle and the flame soot generator respec-tively

Freshly emitted soot particles are hydrophobic or limitedin hygroscopicity and unlikely to contribute to the CCN pop-ulation in the atmosphere (Weingartner et al 1997 Meyerand Ristovski 2007 Zhang et al 2008 Tritscher et al2011) The size of these non-spherical particles will nor-mally differ between measurement techniques (DeCarlo etal 2004) for example showing a larger mobility thanvolume-equivalent diameter However as the soot particlesreside in the atmosphere they undergo physical and chem-ical changes during UV exposure ndash they age The agglom-erated soot particles are exposed to chemical gas-to-particleprocesses in the atmosphere resulting in condensation of or-ganic and inorganic (Rose et al 2006) vapours and coagu-lation of particles onto the agglomerates Due to this ageingprocess the soot particles collapse into a more compact struc-ture (Weingartner et al 1997 Pagels et al 2009 Tritscher etal 2011) Hygroscopicity is enhanced with increasing age-ing time affected by high sulfur content in the fuel highVOC levels in the emissions pre-treatment of the exhaustgas with ozone and UV radiation (eg Weingartner et al1997 Tritscher et al 2011) Weingartner et al (1997) pro-posed that diesel exhaust aerosol which was pre-treated withO3 and then subjected to UV radiation would show an en-hancement in hygroscopicity However the large scatter indata made it impossible to draw conclusions Soot agglomer-ates become more hygroscopic when coated partly or fullyby organic or inorganic material The coating material trans-forms the agglomerates to enable them to act as CCN

In the atmosphere in general aerosol particles are com-posed of both organic and inorganic compounds Organicmaterial makes up a significant fraction (20 to 90 ) of the









2 Experimental
















21 Instrumentation










SMPS

DMA-TD-APM

FIDTesto350

Ejector diluter

AC

Fluorescent tubes

STEEL CHAMBER

SMOG CHAMBER

Exhaust air out

6 m3

22 m3

Generator

SP-AMS

DM

A

CCNC

CCNCSFCA












Water SOA POA BC


κ ndash 01265 0 0T (K) 29815





3 Theory


s =p

p0= aw times Ke (1)



Ke = exp

(4σsolMw

RTρwDwet

) (2)


aw =

(nw

nw + isns

)=

(1+ is

ns

nw

)minus1

=

(1+

nsumMw

ρwπ6

(D3

wetminus d3

s

))minus1

(3)


nsum=

sumi

εi iiρid3s

Mi

π

6(4)


is = νsφs (5)



aw =

(1+ κ

Vs

Vw

)minus1

=D3

wetminus d3s

D3wetminus d3

s (1minus κ) (6)




κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)


κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)


4 Modelling



dve =3

radic6

π

m

ρcorr (11)


1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)





































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References












































































































2 Experimental
















21 Instrumentation










SMPS

DMA-TD-APM

FIDTesto350

Ejector diluter

AC

Fluorescent tubes

STEEL CHAMBER

SMOG CHAMBER

Exhaust air out

6 m3

22 m3

Generator

SP-AMS

DM

A

CCNC

CCNCSFCA












Water SOA POA BC


κ ndash 01265 0 0T (K) 29815





3 Theory


s =p

p0= aw times Ke (1)



Ke = exp

(4σsolMw

RTρwDwet

) (2)


aw =

(nw

nw + isns

)=

(1+ is

ns

nw

)minus1

=

(1+

nsumMw

ρwπ6

(D3

wetminus d3

s

))minus1

(3)


nsum=

sumi

εi iiρid3s

Mi

π

6(4)


is = νsφs (5)



aw =

(1+ κ

Vs

Vw

)minus1

=D3

wetminus d3s

D3wetminus d3

s (1minus κ) (6)




κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)


κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)


4 Modelling



dve =3

radic6

π

m

ρcorr (11)


1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)





































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References

















































































































21 Instrumentation










SMPS

DMA-TD-APM

FIDTesto350

Ejector diluter

AC

Fluorescent tubes

STEEL CHAMBER

SMOG CHAMBER

Exhaust air out

6 m3

22 m3

Generator

SP-AMS

DM

A

CCNC

CCNCSFCA












Water SOA POA BC


κ ndash 01265 0 0T (K) 29815





3 Theory


s =p

p0= aw times Ke (1)



Ke = exp

(4σsolMw

RTρwDwet

) (2)


aw =

(nw

nw + isns

)=

(1+ is

ns

nw

)minus1

=

(1+

nsumMw

ρwπ6

(D3

wetminus d3

s

))minus1

(3)


nsum=

sumi

εi iiρid3s

Mi

π

6(4)


is = νsφs (5)



aw =

(1+ κ

Vs

Vw

)minus1

=D3

wetminus d3s

D3wetminus d3

s (1minus κ) (6)




κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)


κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)


4 Modelling



dve =3

radic6

π

m

ρcorr (11)


1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)





































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References







































































































SMPS

DMA-TD-APM

FIDTesto350

Ejector diluter

AC

Fluorescent tubes

STEEL CHAMBER

SMOG CHAMBER

Exhaust air out

6 m3

22 m3

Generator

SP-AMS

DM

A

CCNC

CCNCSFCA












Water SOA POA BC


κ ndash 01265 0 0T (K) 29815





3 Theory


s =p

p0= aw times Ke (1)



Ke = exp

(4σsolMw

RTρwDwet

) (2)


aw =

(nw

nw + isns

)=

(1+ is

ns

nw

)minus1

=

(1+

nsumMw

ρwπ6

(D3

wetminus d3

s

))minus1

(3)


nsum=

sumi

εi iiρid3s

Mi

π

6(4)


is = νsφs (5)



aw =

(1+ κ

Vs

Vw

)minus1

=D3

wetminus d3s

D3wetminus d3

s (1minus κ) (6)




κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)


κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)


4 Modelling



dve =3

radic6

π

m

ρcorr (11)


1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)





































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References







































































































Water SOA POA BC


κ ndash 01265 0 0T (K) 29815





3 Theory


s =p

p0= aw times Ke (1)



Ke = exp

(4σsolMw

RTρwDwet

) (2)


aw =

(nw

nw + isns

)=

(1+ is

ns

nw

)minus1

=

(1+

nsumMw

ρwπ6

(D3

wetminus d3

s

))minus1

(3)


nsum=

sumi

εi iiρid3s

Mi

π

6(4)


is = νsφs (5)



aw =

(1+ κ

Vs

Vw

)minus1

=D3

wetminus d3s

D3wetminus d3

s (1minus κ) (6)




κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)


κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)


4 Modelling



dve =3

radic6

π

m

ρcorr (11)


1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)





































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References







































































































κsum=

sumiεiκi (7)

κi = ii times

(ρiMw

ρwMi

)(8)


κCCN =

(4A3

27d3s ln2 sc

)(9)

A =

(4σsolMw

RTρw

) (10)


4 Modelling



dve =3

radic6

π

m

ρcorr (11)


1

ρcorr=

mfBC(APM)

ρBC+

mfSOA(APM)

ρSOA+

mfPOA(APM)

ρPOA (12)





































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References


































































































































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References



















































































































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References












































































































52 Detailed picture
















00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References


















































































































00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References










































































































00

01

02

03

04

05

06

07

08

09

10

00

02

04

06

08

10

12

0000 0100 0200 0300 0400

mf S

OA

s c (

)













































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References











































































































































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References


































































































































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References






























































































































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References















































































































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References








































































































54 Uncertainties









6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References








































































































6 Conclusions















Appendix A


Nomenclature






Edited by H Skov

References












































































































Appendix A


Nomenclature






Edited by H Skov

References






































































































Appendix A


Nomenclature






Edited by H Skov

References








































































































Edited by H Skov

References































































































































































































































































Documents

Cloud droplet activity changes of soot aerosol upon smog ......condensation nuclei (CCN) properties at the onset of UV ra-diation implies that the lifetime of soot particles in the