Investigating the use of secondary organic aerosol as seed ... · compounds and identifying SOA components arising from a known precursor in such a complex matrix is extremely dif-ﬁcult

Atmos Chem Phys 11 5917ndash5929 2011wwwatmos-chem-physnet1159172011doi105194acp-11-5917-2011copy Author(s) 2011 CC Attribution 30 License

AtmosphericChemistry

and Physics

Investigating the use of secondary organic aerosol as seed particlesin simulation chamber experiments

J F Hamilton1 M Rami Alfarra 23 K P Wyche4 M W Ward 1 A C Lewis1 G B McFiggans3 N Good3P S Monks4 T Carr 4 I R White4 and R M Purvis1

1Department of Chemistry University of York Heslington York YO10 5DD UK2National Centre for Atmospheric Science (NCAS) School of Earth Atmospheric and Environmental Sciences University ofManchester Manchester M13 9PL UK3Centre for Atmospheric Science School of Earth Atmospheric and Environmental Sciences University of ManchesterManchester M13 9PL UK4Atmospheric Chemistry Group Department of Chemistry University of Leicester Leicester LE1 7RH UK now at Laboratoire de Meteorologie Physique Blaise Pascal Univ Clermont-Ferrand 63000 France

Received 13 October 2010 ndash Published in Atmos Chem Phys Discuss 27 October 2010Revised 28 March 2011 ndash Accepted 18 April 2011 ndash Published 23 June 2011

Abstract The use ofβ-caryophyllene secondary organicaerosol particles as seeds for smog chamber simulations hasbeen investigated A series of experiments were carriedout in the Manchester photochemical chamber as part ofthe Aerosol Coupling in the Earth System (ACES) projectto study the effect of seed particles on the formation ofsecondary organic aerosol (SOA) from limonene photo-oxidation Rather than use a conventional seed aerosol con-taining ammonium sulfate or diesel particles a method wasdeveloped to use in-situ chamber generated seed particlesfrom β-caryophyllene photo-oxidation which were then di-luted to a desired mass loading (in this case 4ndash13 microg mminus3)Limonene was then introduced into the chamber and oxi-dised with the formation of SOA seen as a growth in thesize of oxidised organic seed particles from 150 to 325 nmmean diameter The effect of the partitioning of limoneneoxidation products onto the seed aerosol was assessed us-ing aerosol mass spectrometry during the experiment and thepercentage ofmz 44 an indicator of degree of oxidationincreased from around 5 to 8 The hygroscopicity of theaerosol also changed with the growth factor for 200 nm par-ticles increasing from less than 105 to 125 at 90 RH Thedetailed chemical composition of the limonene SOA couldbe extracted from the complexβ-caryophyllene matrix usingtwo-dimensional gas chromatography (GCtimes GC) and liquidchromatography coupled to mass spectrometry High resolu-tion Fourier Transform Ion Cyclotron Resonance Mass Spec-trometry (FTICR-MS) was used to determine exact molecu-

Correspondence toJ F Hamilton(jfh2yorkacuk)

lar formulae of the seed and the limonene modified aerosolThe average OC ratio was seen to increase from 032 to 037after limonene oxidation products had condensed onto the or-ganic seed

1 Introduction

Atmospheric aerosols affect the climate through both directand indirect radiative processes however there remain sig-nificant uncertainties in the magnitude and variability of ef-fects arising from limited knowledge on sources compo-sition properties and the mechanisms of their formation(Fuzzi et al 2006) The gas phase oxidation of biogeniccompounds in the atmosphere can result in the formation oflower volatility species which can form biogenic secondaryorganic aerosol (BSOA) one of the most uncertain factorsin the global radiation budget (Intergovermental Panel forClimate Change 2007) Recent modelling studies of atmo-spheric aerosols using explicit chemical mechanisms basedon results from simulation chamber experiments have showna significant under-prediction of secondary organic aerosolformation (SOA) when compared to measurements For ex-ample in their respective studies Volkamer et al (2006) andCapes et al (2009) reported under predictions in modelledSOA by factors of the order 4ndash15 when compared to mea-surements One explanation for this effect is a lack of back-ground particles in traditional smog chamber experimentsIn chambers where particle formation occurs by nucleationthere is an induction time for SOA formation as condens-able products must exceed saturation vapour pressures In

Published by Copernicus Publications on behalf of the European Geosciences Union

5918 J F Hamilton et al Investigating the use of secondary organic aerosol as seed particles

general owing to the widespread presence of pre-existingcondensation surfaces in the real atmosphere organic nucle-ation is not widely observed the more common phenomenonis the propagation of organic aerosol mass via the condensa-tion of oxidized VOC products onto pre-existing aerosolsBetter simulation in chamber experiments can therefore beachieved by including background or ldquoseedrdquo particles for ox-idised products to partition onto

It is a criterion of absorptive partitioning theory that allcondensed components are miscible across the whole con-centration range and thus the greatest amount of interac-tion is likely to take place when seed particles have similarproperties as the compounds condensing onto them Histor-ically most chamber experiments have used inorganic aque-ous solutions of ammonium sulphate or sulphuric acid asseed aerosols to study SOA formation (Hallquist et al 2009)sometimes dried prior to use (eg Odum et al 1997) Suchstudies have shown the formation of organosulphate speciesin monoterpene and isoprene SOA which have also beenseen in ambient samples (Iinuma et al 2007 Liggio andLi 2006 Surratt et al 2007) Volkamer et al (2009) foundthat the yield of SOA from acetylene photo-oxidation wherethe main product is glyoxal strongly depends on the chem-ical composition of the seed particles In this case use ofan ammonium sulphate seed led to one of the lowest SOAyields However the yield was significantly increased by theaddition of fulvic acid to the ammonium sulphate Jang etal (2003) used seed particles from diesel engine exhaust inchamber studies to mimic primary organic aerosol How-ever diesel and fulvic acid are made up of thousands ofcompounds and identifying SOA components arising from aknown precursor in such a complex matrix is extremely dif-ficult In both cases there is a disparity between the chemicalproperties of the seed and the oxidised organic compoundspartitioning to form SOA In the atmosphere most organicaerosol (OA) is oxidised even in urban areas (Jimenez et al2009 Zhang et al 2007) Thus a better seed particle tosimulate many environments would be an oxidised organicaerosol (OOA) There are a number of difficulties involvedwith using OOA as a seed Firstly which organic compoundis most suitable Previous studies have used dioctylphthalate(Song et al 2007) and highly oxidised organic acids (Corri-gan et al 2008) as a surrogate of SOA but these are singlecompounds rather than the range of molecular weights andfunctionalities seen in the atmosphere Oxidised compoundsare also generally difficult to work with and stick to surfaces

A more appropriate seed for investigating the partitioningof a wide range of semi-volatile OVOCs would be SOA it-self Dommen et al (2009) usedα-pinene SOA as a seedaerosol for isoprene SOA although the isoprene was13C la-belled to allow the seed to be differentiated This methodrequired the use of plants to fix13CO2 which then emit-ted labelled isoprene and thus is limited to this particu-lar VOC Here we present the use of sesquiterpene SOAformed from β-caryophyllene photo-oxidation as a seed

particle for simulation chamber experiments Within thiswork β-caryophyllene SOA seed has been used to studythe generation of SOA from limonene photo-oxidationβ-Caryophyllene reacts rapidly with hydroxyl radicals andozone resulting in almost immediate nucleation owing to thehigh molecular weight and low volatility of its first genera-tion oxidation products The chamber was subsequently di-luted to obtain a low particle concentration and remove themajority of the gas phase products The VOC to be studiedin this case limonene was then introduced into the chamberand oxidised The gas phase loss of limonene and the for-mation of its oxidation products were followed by ChemicalIonisation Reaction Time of Flight Mass Spectrometry (CIR-TOF-MS) The partitioning of limonene oxidation productsto the seed was followed by monitoring the shift in the par-ticle size distribution The change in bulk chemical compo-sition was analysed in real time using Aerosol Mass Spec-trometry The difference in the detailed chemical composi-tion between the seed and the limonene SOA was analysedoffline using Comprehensive Two-dimensional Gas Chro-matography Time-of-Flight Mass Spectrometry (GCtimesGC-TOFMS) Liquid Chromatography Ion Trap Mass Spec-trometry (LC-MSn) and Fourier Transform Ion CyclotronResonance Mass Spectrometry (FTICR-MS)

2 Experimental

21 Chamber

A comprehensive and detailed description of the Universityof Manchester Aerosol Chamber will be presented in a futurepublication (McFiggans et al 2011) hence only an introduc-tion and general overview is provided here The Manchesteraerosol chamber is an 18 m3 (3 m (H)times 3 m (L) m (H)times 3 m2 m (W)) FEP Teflon bag mounted on three rectangular alu-minium frames The central rigid rectangular frame is fixedwith the upper and lower frames free to move vertically caus-ing the bag to expand and collapse as sample air is intro-duced and extracted A bank of halogen lamps and a 6 kWXenon arc lamp are mounted on the enclosure housing thebag which is coated with reflective ldquospace blanketrdquo servingto render the enclosure as an integrating sphere maximisingthe irradiance in the bag and ensuring even illumination Thecombination of illumination has been tuned and evaluatedto mimic the atmospheric actinic spectrum over the wave-length range 290ndash800 nm and has a maximum total actinicflux of 07times 1018 photos sminus1 mminus2 nmminus1 over the region 460ndash500 nm The calculatedj (O1D) value during the reported ex-periments was 36times 10minus5 sminus1 (290ndash340 nm) Air condition-ing between the enclosure and the bag removes additionalunwanted heat generated by the lamps The air charge inthe bag is dried and filtered for gaseous impurities and par-ticles using a combination of Purafil (Purafil SP media fromPurafil Inc USA) charcoal and HEPA filters (Donaldson

Atmos Chem Phys 11 5917ndash5929 2011 wwwatmos-chem-physnet1159172011

J F Hamilton et al Investigating the use of secondary organic aerosol as seed particles 5919

Filtration (GB) Ltd) prior to humidification with ultrapuredeionised water This setup was found to be efficient at re-moving VOCs NO2 and O3 but was not fully able to removeNO

Parent VOCs are introduced into the chamber through in-jection into a heated glass bulb fed with a flow of filterednitrogen NOx levels are controlled by injection from a cylin-der into the charge line and a high capacity O3 generatorwas employed to control initial O3 concentrations as well asserving as a cleaning agent during flushing between exper-iments Cycling between experiments is facilitated by fullcomputer control as is the monitoring of key chamber con-ditions (eg temperature and relative humidity) Automaticfillflush cycling to clean the chamber before and after eachexperiment with 3 m3 minminus1 flow of scrubbed dried rehu-midified (and optionally ozonised) air is achieved by controlof electro-pneumatic valves Each cycle takes approximately12 min and cleaning is normally achieved after 5 or 6 cyclesCoupling morning and evening cycling with overnight soak-ing at ppm levels of O3 allows experiments to be carriedout on subsequent days Longer experiments are possible on2-day cycles Relative humidity (RH) and Temperature(T )

are measured at several points throughout the chamber (bydewpoint hygrometer and a series of thermocouples and re-sistance probes respectively) and are controlled by divertingair through the inlet humidification circuit and by controllingthe air conditioning setpoint respectively

22 Experimental methodology

The secondary organic aerosol seed was generated from thephoto-oxidation ofβ-caryophyllene in the presence of NOx(Initial VOCNOx sim1ndash2) using a conventional nucleationand growth experiment The experimental details are sum-marised in Table 1 Approximately 98ndash99 of the precursorβ-caryophyllene had reacted after about 3 h This was veri-fied using the CIR-TOF-MS After this time chamber lightswere turned off and a fraction of the chamber content wasflushed out and replaced by clean air This resulted in di-lution of the seed concentration down to approximately 4 to13 microg mminus3 depending on the experiment and also the dilu-tion of the gas phase reactants During the dilution phaselimonene was injected into the chamber and if requiredthe NOx level was adjusted by the injection of an additionalquantity of NO2 gas to keep the VOCNOx ratio at around 2(a fixed ratio was chosen to reduce system variables pertain-ing to the presence of NOx and hence to enable comparisonsto be made between experiments) Following this flushrefillphase chamber lights were turned back on again to allow forthe photooxidation of limonene

Filters were collected in a specially constructed holder po-sitioned in the chamber vent line Aerosol samples were col-lected onto 47 mm quartz fibre filters (Whatman) at a flowrate of 3 m3 minminus1 Filters were collected during the dilutionstage to obtain aβ-caryophyllene seed sample which was

used as a background for the limonene SOA Limonene SOAsamples were collected at the end of the experiment Af-ter sampling filters were immediately placed in pre-cleanedglass vials and stored belowminus20C until analysis

Experiments were carried out at an averageT of257plusmn 07C and nominal relative humidity of 71plusmn 16 Chamber humidity was controlled using vapour from heatedultra pure water (Purelab Ultra System Elga) Knownamounts ofβ-caryophyllene (C15H24 Sigma Aldrich) andlimonene (C10H16 Flukage 990) were evaporated from aheated glass bulb and continuously flushed into the chamberusing a flow of nitrogen During each experiment the initialchamber NOx concentration was controlled by introducingNO2 gas into the chamber from a cylinder containing 10 NO2 in nitrogen

23 Gas phase measurements

NO and NO2 mixing ratios were measured using a chemi-luminescence detector (Model 42i Thermo Scientific MAUSA) Ozone was measured using a UV photometric gasanalyser (Model 49C Thermo Scientific MA USA) Thegas phase organic compounds within the chamber were mea-sured using Chemical Ionisation Reaction Time-of-FlightMass Spectrometry (CIR-TOF-MS) This technique has beendescribed in detail elsewhere (Blake et al 2004 Wyche etal 2007) hence here it is only described in brief alongsideexperiment specific details

The CIR-TOF-MS instrument comprises a bespoke tem-perature controlled (40Cplusmn 05) radioactive (241Am) ionsourcedrift cell assembly coupled to an orthogonal time-of-flight mass spectrometer equipped with a reflectron ar-ray (Kore Technolgy Ltd Ely UK) During these exper-iments proton transfer reaction ionisation was employedas the chemical ionisation technique utilising hydronium(H3O+) as the primary reagent ion (Lindinger et al 1993)In this instance the hydronium ions were generated from ahumidified N2 carrier gas (purity = 99998 ) Providing cer-tain reaction conditions are met ion molecule reaction be-tween H3O+ and the organic analyte will yield a protonatedproduct ion (RH+) and neutral water

H3O++Rminusrarr RH+

+H2O (R1)

Despite being considered a relatively soft ionisation methodthe nascent RH+ ion may be sufficiently energetic so as todissociate and produce one or more fragment ions (Blake etal 2009)

Chamber air containing the analyte was delivered in acontinuous stream (approximately 230 sccm) to the drift cellvia a 05 m long14rdquo (internal diameter) Teflon sample lineheated to 40 (plusmn1)C The centre drift cell was operated at anEN (ie electric fieldgas number density) ratio of approxi-mately 90 Td An energy ramp was applied at the base of thecell to remove potential water-cluster ions (eg RH+H2O)

wwwatmos-chem-physnet1159172011 Atmos Chem Phys 11 5917ndash5929 2011


Table 1 Experimental details and SOA yields

Experiment Phase 1 Phase 2 1[VOCβminusc] 1[VOC lim ] Phase 1 Phase 3 Phase 1 Phase 3 Phase 3YSOA YSOA Phase 3Date [VOCβminusc]0

a [VOClim ]0a ppbV ppbV VOCNOx

a VOCNOxa Peak Seed Peak Peak Phase 1 (Condensation)e

ppbV ppbV SOA massb Measured Condensed (Seed) microgmminus3 SOA massc SOA massd

microgmminus3 microgmminus3

25 Jun 2008 4877 9282 4480 8539 142 200 3558 6960 6533 952 1375(plusmn871) (plusmn1019)


10 Jul 2008 4810 5261 4184 4839 064 188 6580 4476 3128 1886 1162(plusmn860) (plusmn669)

a [VOC]0 and VOCNOx based on 5 min integrated CIR-TOF-MS and NOx datab Maximum measured SOA mass during Phase 1 of the experimentc Maximum measured SOAmass during Phase 3 of the experimentd Maximum SOA formed from condensation during Phase 3e SOA yield from condensation during Phase 3

The CIR-TOF-MS was calibrated via three separate meth-ods (i) using step-wise dilution of a gravimetrically pre-pared gas standard (BOC Special Gases UK) containing avariety of VOCs and oxygenated (O)VOCs (ii) using cali-bration material produced in-house via the injection of liquidsamples (prepared by volumetric dilution of the pure materialin hexane) into 10 l Tedlar bags (SKC Inc USA) contain-ing both humidified and dry pure nitrogen and (iii) usinggaseous standards derived from permeation tubes (Vici IncUS Eco-scientific UK) diluted humidified and delivered tothe CIR-TOF-MS by a commercial calibration unit (Kintecmodel 491)

24 Aerosol characterisation

The total SOA particle number concentration was measuredusing an ultrafine water-based condensation particle counter(wCPC 3786 TSI Inc) This model has a minimum sizecut-off of 25 nm The size distribution of the generatedSOA particles was measured using a Differential MobilityParticle Sizer (DMPS) The DMPS instrument consists oftwo Vienna design Differential Mobility Analysers (DMAs)(Williams 1999) an ultrafine DMA for particles in the sizerange 34ndash34 nm and a standard DMA for particle sizes from20ndash500 nm After transmission through the DMAs particlesare counted using condensation particle counters (CPC) Theultrafine DMA was attached to a TSI 3025A CPC and thestandard DMA to a TSI 3010 CPC The DMAs were oper-ated in parallel and utilised a custom built sealed recirculat-ing sheath air system which was dried and filtered Parti-cle size distributions in the diameter size range from 3 nm to500 nm were obtained every 10 min

A hygroscopicity tandem differential mobility analyser(HTDMA) was used to measure on-line size resolved wateruptake at 90 RH The HTDMA dries the sample aerosol tolt10 RH using a Nafionreg drier (Perma Pure MD-110-12Toms River NJ USA) A Strontium-90 aerosol neutraliserthen brings the sample into charge equilibrium A DMA(BMI Haywood CA USA) operated with a re-circulating

sheath flow of 5 l minminus1 and sample flow of 05 l minminus1 se-lects particles of a single mobility Diameters chosen werelarger than the mode of the number size distribution thusavoiding the sampling of a significant fraction of multi-charged particles The sample is then humidified to 90 RHusing a Gore-Texreg humidifier Cubison et al 2006) Thehumidified size selected sample is then passed through a res-idence coil for 15 s A second DMA (BMI) and CPC (TSI3782) is then used to measure the size distribution of the hu-midified sample using the DMPS technique The temperatureof the DMAs and humidification system is controlled usingpeltier units (Supercoolreg AA-040-12-22 Sweden) The hu-midity is measured using a dew point hygrometer (EdgeTechDewmaster MA USA) The operation of the HTDMA wasvalidated by sampling ammonium sulphate and sodium chlo-ride test aerosols using the operating procedures describedby Good et al (2010) The data was inverted using themethod described by Gysel et al (2009) The data from theHTDMA is reported in terms of the hygroscopic (GFD0RH)the wet particle diameter at a given RH divided by the par-ticlersquos dry diameter (D0) The aerosol particles measuredappeared internally mixed hence the reported growth factorsare the number weighted means of the measured growth fac-tor probability density distributions unless otherwise stated(Gysel et al 2009)

Measurements of the secondary organic aerosol composi-tion were made using a compact Time-of-flight Aerosol MassSpectrometer (Drewnick et al 2005) (cToF-AMS Aero-dyne Research Inc USA) The instrument operated in thestandard configuration taking both mass spectrum (MS) andparticle time-of-flight (pToF) data and was calibrated using350 nm monodisperse ammonium nitrate particles A moredetailed description of the instrument and its calibration wasprovided by Drewnick et al (2005)

The detailed chemical composition of the SOA was deter-mined from filter samples using two complementary tech-niques The semi-volatile fraction of the SOA was anal-ysed using direct thermal desorption coupled to GCtimes GC-TOFMS A small piece of the filter (weight between



10ndash20 mg) was cut out placed inside a thermal desorp-tion glass vial and injected using an autosampler Twostage thermal desorption was used to ensure a narrowband of analytes was injected onto the column (MPS2 au-tosampler and TDUCIS4 Gerstel Germany) GCtimes GC-TOFMS was carried out using a cryo-jet Pegasus 4D (LecoSt Joseph MI USA) incorporating an Agilent 6890N GasChromatograph and a Pegasus III reflectron time-of-flightmass spectrometer A liquid nitrogen cooled gas jet mid-point modulator at approximatelyminus160C was used to en-able comprehensive two-dimensional separations A sec-ondary oven housed within the GC oven allowed indepen-dent column temperature control The column set usedin these experiments provides a primary volatility basedseparation followed by a secondary polarity based separa-tion The first dimension was a 5 phenyl- 95 methyl-polysiloxane 30 mtimes 320 micromtimes 10 microm HP-5 (JampW Scien-tific Wilmington DE USA) The second column wasa 50 phenyl-polysiloxane 2 mtimes 100 micromtimes 01 microm BP10(SGE Melbourne Australia) The GC was held at 70C fortwo minutes and then raised at 25C minminus1 to 250C andheld at this temperature for a further sixteen minutes Thecarrier gas was helium (999999 BOC Gases GuildfordUK) supplied at 15 ml minminus1 The modulator and secondaryoven were operated at +30C and +15C above GC oventemperature respectively

The remainder of the filter sample was extracted into highpurity water filtered and reduced to 1 ml using a vacuum sol-vent evaporator (Biotage Sweden) The water soluble andhigh molecular weight compounds were analysed using liq-uid chromatography-ion trap mass spectrometry (LC-MSn)LC-MS was carried out using an HCT Plus ion trap massspectrometer (Bruker Daltonics GmbH Bremen Germany)equipped with an Eclipse ODS-C18 column with 5 microm par-ticle size (Agilent 46 mmtimes 150 mm) Both positive andnegative ionisation modes were used Full details of themethodology and instrument have been published previously(Hamilton et al 2008)

Samples were also analysed at high resolution using aBruker APEX 94 T Fourier Transform Ion Cyclotron Res-onance Mass Spectrometer Extracts were sprayed at a flowrate of 2 microL minminus1 into an Apollo II electrospray interfacewith ion funnelling technology Spectra were acquired inboth positive and negative ion mode over the scan rangemz100ndash3000 using the following MS parameters nebulis-ing gas flow 09 l minminus1 drying gas flow 5 l minminus1 dry-ing temperature 190C collision cell accumulation 005ndash05 s and data acquisition size 2 MB (yielding a target res-olution of 130 000 atmz 400) Data were analysed usingDataAnalysis 40 software (Bruker Daltonics Bremen Ger-many) The instrument was calibrated using protonated (pos-itive ion mode) or deprotonated (negative ion mode) argi-nine clusters The mass spectra were internally recalibratedwith a series of prominent peaks positive ndash C10H16NaO3C12H20NaO3 C13H22NaO3 C14H22NaO3 C14H22NaO4

C16H28NaO4 negative ndash C10H15O5 C9H14O4 C10H15O4C10H16O5 C14H22O4 Background contaminants also seenin pure water and blank extracted filters were removed andintensity weighted OC and HC ratios were calculatedThree dimensional Van Krevelen plots were created in IDLv63 (ITTVIS) with data displayed in contour plots over10 contour levels Data were interpolated between sam-ple points and the intensity scaled to 70 of the maximumvalue

3 Results and discussion

The experimental details are given in Table 1 A typical ex-periment is split into three phases

ndash In phase 1 theβ-caryophyllene was introduced thelights turned on and SOA formed

ndash In phase 2 the lights were turned off the chamber wasflushed to reduce seed and VOC concentrations thendiluted with clean air while the limonene was injectedinto the chamber

ndash In phase 3 the lights were turned on and limonenephoto-oxidation occurred followed by product conden-sation onto the organic seed

The evolution of certain key parameters is shown in Fig 1for a typical example of the organic seed experiments(25 June 2008) In Fig 1 the beginning of phase 1 occursat time = 0 and the vertical red dotted lines show the begin-ning and end of phase 2 In this experiment 49 (plusmn9) ppbVof β-caryophyllene was introduced into the chamber alongwith sufficient NO2 gas to give an initial NOx (ie the sumof NO and NO2) concentration of 34 (plusmn1) ppbV This VOCand NOx combination thus gave an initial VOCNOx ratio ofapproximately 14 It should be noted that no additional NOgas was added to the chamber matrix at any time and the ini-tial quantity of NO observed in Fig 1 at the beginning of eachexperiment phase results from incomplete scrubbing of thecharge gas for NO Following injection of the reactants thechamber lights were turned on (at time = 0) and almost im-mediately the SOA mass in the chamber was observed to in-crease The SOA number density peaked after 10 min and themass concentration peaked at around 36 microg mminus3 after 60 minA density of 13 was used for both theβ-caryophyllene andlimonene SOA based on the most recent literature to calcu-late the SOA mass from the DMPS volume data (Bahreiniet al 2005 Varutbangkul et al 2006) Very little ozonewas observed to accumulate within the chamber during phase1 (maximum of 8 (plusmn2) ppbV at the end of phase 1 ow-ing to its rapid reaction with theβ-caryophyllene (lifetimeof β-caryophyllene with 30 ppbV O3 is around 2 min) Incomparison with other VOC chamber photo-oxidation sys-tems (Odum et al 1997)β-caryophyllene chamber photo-oxidation results in relatively unusual NOx chemistry which



26

Figure 1 Evolution of basic chamber measurements during 25062008 868

experiment 869

870

871 872

873

874

875

876

877

878

879

880

Fig 1 Evolution of basic chamber measurements during25 June 2008 experiment

will be discussed in more detail in a future publication (Al-farra et al 2011) In brief however the NOx behaviour ob-served in Fig 1 may be a result of the fast reaction rate ofβ-caryophyllene with ozone and resultant low radical yields

Following near complete reaction of theβ-caryophyllene(98ndash99 oxidation) the chamber was then diluted to leavea β-caryophyllene organic seed aerosol background of ap-proximately 4 microg mminus3 No change in particle size (due torevolatilisation) was seen during dilution indicating that theβ-caryophyllene SOA was mostly composed of low volatilitycompounds Figure 2 shows a comparison of SOA size distri-butions measured immediately before the lights were turnedoff at the end of phase 1 (before dilution) and the immedi-ately before lights were turned back on at the start of phase 3(after dilution) Grieshop et al (2007) showed that SOA fromα-pinene ozonolysis repartitions reversibly upon dilution buton a much longer time scale than has been observed in singlecomponent aerosols of similar sizes Our dilution time scalewas much shorter and did not lead to any volatilisation ofSOA It should be noted however that volatilisation of someof the SOA material may occur over longer time scales butthis was not explored here During phase 2 ozone levels inthe chamber dropped below the detection limit of the instru-ment (2 ppbV) indicating that the ozone mixing ratio wasnear zero before the start of phase 3 Following dilution 93(plusmn10) ppbV of limonene was injected into the chamber be-fore an additional ldquotop-uprdquo injection of NOx was added (in

27

Figure 2 Particle size distribution of the ββββ-caryophyllene seed particles Black 881

line indicates size distribution at the end of Phase 1 (pre-dilution) Red line 882

indicates size distribution at the end of Phase 2 (after flushing and dilution) 883

884

60x103

50

40

30

20

10

0

dN

dlo

gD

p

30025020015010050

Dp (nm)

10000

8000

6000

4000

2000

0

dN

dlo

gD

p

Before dilution After dilution

885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903

Fig 2 Particle size distribution of theβ-caryophyllene seed parti-cles Black line indicates size distribution at the end of Phase 1 (pre-dilution) Red line indicates size distribution at the end of Phase 2(after flushing and dilution)

the form of NO2) to ensure an initial VOCNOx ratio similarto that of phase 1 (iesim20) On average for the three experi-ments less than 05 (plusmn1) ppbV ofβ-caryophyllene remainedin the chamber following dilution Also from the applica-tion of a pseudo calibration sensitivity to the data (obtainedfor the C10 multifunctional monoterpene oxidation productpinonaldehyde) it is estimated that an average of only 2ndash3(plusmn1) ppbV ofβ-caryophyllene oxidation products were car-ried over to phase 3 of the experiment In addition completeevaporation of the seed particle post-dilution could only havecontributed to a maximum of 05 ppbV of the caryophyllenicVOC observed in practice this would be very much lower

Limonene oxidation was observed by the CIR-TOF-MSduring phase 3 of the experiment after chamber lights hadbeen switched back on For all experiments limonene wasquantified from the spectral signal corresponding to the pro-tonated limonene parent ion (mz137) During phase 3 thedecay of limonene was matched by the concomitant produc-tion of its gas phase reaction products including limonalde-hyde (mz 169 151 and 107) and limonaketone (mz 139)The oxidation of limonene and the production of limonalde-hyde are shown in Fig 1 Limonene oxidation and limon-aldehyde formation was reproducible between experimentswith approximately all of the limonene reacted by the end ofphase 3 (ie after 240 min) in each instance and limonalde-hyde formed in molar yields ofsim7ndash14 The formation ofozone and more conventional chamber NOx chemistry ob-served during phase 3 can also be seen in Fig 1 with a max-imum of 35 (plusmn2) ppb of ozone produced The SOA mass(as well as seed particle size) increased steadily from lightson until 240 min where a second nucleation event was seenwith a steep jump in the particle number concentrations Thiscan be seen in the particle size distribution shown in Fig 3

It is clear from Fig 1 that limonene oxidation products arepartitioning onto the organic seed aerosol increasing aerosolmass from 4 microg mminus3 to 70 microg mminus3 giving an average massgrowth of 345 microg mminus3 hminus1 The particles grew from 150 to325 nm mean diameter during the first 50 min after lights on



28

Figure 3 Particle size distribution during 25062008 experiment and f44 () 904

from the AMS measurements 905

906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929

Fig 3 Particle size distribution during 25 June 2008 experiment andf44 () from the AMS measurements

in phase 3 as shown in Fig 3 giving an average size growthof 35 nm minminus1 The formation of small particles in the sec-ond nucleation event effectively reduces further uptake ontothe seed particles In order to suppress the limonene nucle-ation a series of experiments was carried out under differentconditions including reducing the amount of limonene to 21(plusmn4) ppb in one experiment and increasing the amount of or-ganic seed to 13 microg mminus3 in another In all cases limonenenucleation could not be suppressed Potential reasons forthis may include the following (i) there is insufficient parti-cle mass onto which the limonene oxidation products can ab-sorb (ii) the loss of such limonene oxidation products to seedsurface area is slower than the production of condensable ma-terial meaning the partial pressure increases sufficiently highto exceed the nucleation threshold (iii) then-th generationoxidation product is so condensable that its partial pressureincreases to relatively low values which are above the nucle-ation threshold and iv) one or more of the limonene OVOCsare not fully miscible in theβ-caryophyllene SOA productssuch that they would prefer to be in the gas phase until theybuild up to concentrations above the nucleation thresholdTo allow comparison of limonene SOA composition formedfrom nucleation only and from an organic seeded experimentthe experiment could be stopped immediately as new parti-cle formation was observed and a filter was collected Thiswould minimise the mass of limonene nucleation SOA col-lected in seeded experiments and will be carried out in futurestudies

β-caryophyllene photo-oxidation during phase 1 of theorganic seed experiments resulted in the production of apeak SOA mass of the order 36ndash66 microg mminus3 between 40and 60 min after initial lights on (see Table 1) Substitut-ing these masses into Eq (2) along with the fraction ofβ-caryophyllene reacted allows the SOA yield(YSOA) to bedetermined

YSOA=Mp

1VOC(1)

For phase 1 of the seed experimentsYSOA was determined tobe of the order 10ndash19 During phase 3 of the experimentsthe condensation of limonene photo-oxidation products ontotheβ-caryophyllene seed resulted in the production of an ad-ditional 7ndash65 microg mminus3 of SOA mass on top of the seed suchthat the peak mass was measured to be 13ndash70 microg mminus3 be-tween 110 and 135 min after second lights on givingYSOAfor the condensed mass of the order 6ndash14 Owing to thelack of appropriate data these mass and yield values havenot been wall loss corrected Recent work has shown thatwall loses of SVOCs may be significant in chamber studies(Matsunaga and Ziemann 2010) which could lead to an ad-ditional factor of uncertainty in the yield values quoted hereHowever high molecular weight and low volatility oxida-tion products (such as those derived fromβ-caryophylleneand limonene) are believed to partition quickly to the aerosolphase and hence their gas phase concentrations will be lessinfluenced by the presence of chamber walls As the mea-sured SOA mass loading of the chamber does not accountfor loss of particle mass to the chamber walls the yields pre-sented here are expected to constitute a lower limit for theseexperiments

The hygroscopic growth at 90 RH was measured by theHTDMA In order to measure the effect of partitioning tothe seed a series of nucleation only experiments were car-ried out for bothβ-caryophyllene and limonene These sin-gle precursor experiments were for nominal mixing ratios of50 ppbV and 250 ppbV and selected dry particle diametersbetween 50 nm and 500 nm The growth factor of the pureβ-caryophyllene SOA ranged from 100 to 106 in the 5th to95th percentile range when binned at 60 min intervals from30 min after lights on (illustrated in Fig 4rsquos upper panel) over8 separate experiments Ageing resulted in a small increasein the growth factor over 270 min The growth factor of the



29

Figure 4 Hygroscopic growth factor measurements at 90RH The upper panel 930

shows the growth factors of un-seeded limonene and β-caryophyllene 931

experiments The data is averaged over 60 minute intervals box (25th and 75th 932

percentile) and whisker (5th and 95th percentile) plots show the range of pure 933

precursor growth factors measured The lower panel shows the evolution of the 934

growth factor with time of the β-caryophyllene seed aerosol which is diluted 935

(after 180 minutes) to act as seed for the subsequent limonene injection (at 205 936

minutes) 937

938

939

940

941

942

943

944

945

946

Fig 4 Hygroscopic growth factor measurements at 90 RH Theupper panel shows the growth factors of un-seeded limonene andβ-caryophyllene experiments The data is averaged over 60 minintervals box (25th and 75th percentile) and whisker (5th and95th percentile) plots show the range of pure precursor growth fac-tors measured The lower panel shows the evolution of the growthfactor with time of theβ-caryophyllene seed aerosol which is di-luted (after 180 min) to act as seed for the subsequent limonene in-jection (at 205 min)

pure limonene SOA ranged from 111 to 127 (also binnedat 60 min intervals) in the 5th to 95th percentile range (il-lustrated in Fig 4rsquos upper panel) over 6 experiments withan increase mainly within the first 120 min due to ageingThe lower panel in Fig 4 shows the growth factor measuredduring the 25 June 2008 limonene onβ-caryophyllene seedexperiment When theβ-caryophyllene aerosol was dilutedfor use as a seed its hygroscopic growth did not change sig-nificantly from its undiluted behaviour After the limoneneinjection and lights on condensation of limonene oxidationproducts on to the seed results in a rapid increase in thegrowth factor Several hours after the limonene injectionthe growth factors were comparable to those measured in thelimonene nucleation experiments

The bulk composition of theβ-caryophyllene seed wasinvestigated using a c-ToF-AMS The relative abundance ofthe ion signalmz44 expressed as a fraction of total organicsignal (f44) has been widely used as a proxy for the levelof oxygenation of organic aerosols (Alfarra et al 20042006 Zhang et al 2007) Mass fragment 44 is due mostlyto the ion fragment CO+2 and has previously been shown tostrongly correlate with organic OC for ambient and chamberOA (Aiken et al 2008) For theβ-caryophyllene seedf44 was found to be 46 at the end of phase 1 indicatinga low degree of oxidation as expected if first generation

products are nucleating to form the seed within 5 min oflights on (see Fig 3) Compositional analysis indicated thatthe mainβ-caryophyllene oxidation species in the seed wereβ-caryophyllonic acid andβ-caryophyllinic acid whichare first generation products This finding is in agreementwith the work or Jaoui et al (2003 2007) who reportedthe first measurements ofβ-caryophyllinic acid in SOAgenerated from the photo-oxidation ofβ-caryophyllene butis in slight contrast to the recent work of Li et al (2011)who reported the presence of several other compoundsamongst the most abundant species in theirβ-caryophylleneSOA formed from the dark ozonolysis ofβ-caryophylleneBesidesβ-caryophyllinic acid (and its structural isomers)the most abundant species observed by Li et al (2011)included 3-(4-acetyl-77-dimethyl-3-oxabicyclo[420]oct-4-en-2-yl) propanoic acid 23-dihydroxy-4-[2-(4-hydroxy-3-oxobutyl)-33-dimethylcyclobutyl]-4-oxobutanoic acidand 4-[33-dimethyl-2-(3-oxobutyl) cyclobutyl]-3-hydroxy-4-oxobutanoic acid (and structural isomers) FurthermoreLi et al (2011) state that these latter three compoundsamongst the most abundant components of their SOA con-stitute second-generation species When limonene oxidationproducts condense onto the seed in phase 3f44 increasesrapidly to 74 showing that limonene SOA is much moreoxidised than the seed It can be seen in Fig 3 that theseed ages slowly with a steady increase inf44 during phase1 to a final value of around 5 after 3 h The addition oflimonene SOA resulted in a much steeper jump in the degreeof oxidation of the aerosol confirming this increase is notdue to seed processing

The detailed composition of the seed and the limoneneSOA + seed were investigated using two offline techniquesGCtimes GC-TOFMS and LC-MSn GCtimes GC has been suc-cessfully used to study chamber generated SOA for a rangeof precursors (Hamilton et al 2005 2010) The GCtimes GCchromatograms obtained by thermal desorption of both seedand limonene filter samples are shown in Fig 5 where thex- and y-axis represents retention time on column 1 and 2respectively and response is indicated by a coloured contourThe chromatograms shown are total ion chromatograms Inthe GCtimes GC set up used increasing retention time on col-umn 1 indicates decreasing volatility and increasing reten-tion time on column 2 indicates increasing polarity Theβ-caryophyllene seed chromatogram in Fig 5a has a com-plex composition with the highest concentration of spots cir-cled in yellow The position of the seed spots on the con-tour plot indicates that this is low volatility and low polaritymaterial MS fragmentation patterns can be used to identifya small number of components but there are very few ap-propriate EI spectra available for comparison As shown inFig 5 there is a clear difference between the chromatogramof the β-caryophyllene seed and the limonene SOA + seedsample with an increased number of species present in thelatter The majority of the limonene SOA shown within thepink circle of Fig 5b has a lower retention time on column



30

Figure 5 GCtimesGC extracted ion contour plots (mz 93+91) ndash A ββββ-caryophyllene 947

seed B Limonene SOA + seed Yellow circle indicates highest concentration of 948

seed compounds Pink circle indicates limonene SOA compounds 949

950

951

952

953

954

955

956

957

A

B

Fig 5 GCtimes GC total ion contour plots ndash(A) β-caryophylleneseed (B) Limonene SOA + seed Yellow circle indicates highestconcentration of seed compounds Pink circle indicates limoneneSOA compounds

1 and an increased retention time on column 2 indicatingthat it is more volatile and more polar than the seed respec-tively This qualitatively correlates well with the increase inf44 seen in the AMS spectraβ-caryophyllene is therefore auseful material to use as the seed for studying monoterpeneSOA owing to its higher molecular weight which allows aclear distinction from the condensing organics The high po-lar resolving power of GCtimes GC has a clear advantage forthese samples over conventional GC-MS However the des-orption temperature of the GC technique (350C) may be in-sufficient to volatilise the least volatileβ-caryophyllene SOAcomponents (based on vapour pressure estimates of knownspecies and those identified by FTICR-MS in this study) andthus a complementary technique is required

LC-MS has also been used to investigate the compositionof the water soluble extracts of the two samples Positiveionisation where all peaks exist as their [M + Na]+ adductswas used to provide a response for the widest range of func-tionalities All mass spectra across the entire LC retentiontime were averaged (reducing artefact formation from direct

31

Figure 6 Average mass spectrum of filter extracts Upper ββββ-caryophyllene 958

seed Middle limonene SOA + seed Lower Difference mass spectra 959

960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350

Fig 6 Average mass spectrum of filter extracts Upperβ-caryophyllene seed Middle limonene SOA + seed Lower Dif-ference mass spectra

ESI analysis) to obtain a molecular weight distribution repre-senting the seed and the limonene SOA + seed and these areshown in Fig 6 Themzaxis has been shifted by 23 Da togive the molecular weight rather than the adduct ion massUsing this sytem sodium attaches itself to nearly every oxy-genated molecule This is true for standard mixtures andprevious SOA samples This is confirmed by the FTICR-MS wheregt95 of molecular formulae identified in posi-tive mode contain a Na ion Theβ-caryophyllene seed massspectrum shown in Fig 6a is dominated by compoundsin the range 238 to 300 Da Using LC-MSn these specieswere isolated and subjected to fragmentation to give dis-tinct compound specific mass spectra from which it waspossible to assign respective identities using known library



32

Figure 7 3D Van Krevelen diagrams for the seed and limonene SOA + seed 964

965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987

Fig 7 3-D Van Krevelen diagrams for the seed and limonene SOA + seed

spectra obtained from previousβ-caryophyllene SOA exper-iments The distinct fragmentation patterns produced indi-cate that these species are known primarily first and second-generation products of OH and O3 chemistry The majorpeaks are 238 254 270 and 284 and peak assignment andmolecular identification will be provided in a companionmanuscript (Alfarra et al 2011) The limonene SOA + seedmass spectrum shown in Fig 5b shows the appearance oflower molecular weight limonene oxidation products in therange 140 to 234 Da A reduction in the intensity of theseed ions (ie 254 270 etc) was observed and this is tobe expected considering that the seed has been diluted inthe chamber and would have been further diluted by thelimonene SOA In order to pull out the mass channels cor-responding to the limonene SOA a difference spectrum wascreated as shown in Fig 6c where peaks more abundant inthe seed have negative values on the y-axis and peaks moreabundant in the limonene + seed SOA have positive valueson the y-axis The total ion intensity was normalised to theamount of aerosol collected to allow a comparison of rela-tive composition The major limonene peaks detected corre-spond to molecular weights of 170 186 200 and 216 DaThe limonene SOA mass channels were subsequently in-vestigated further using LC-MSn to allow identification oflimonene oxidation products in the aerosol Using fragmen-tation patterns a total of 14 limonene products were identifiedand these are shown in Table 2

In a previous study investigating SOA formation from thephoto-oxidation of limonene Jaoui et al (2006) were able toidentify twenty-eight different compounds in limonene SOAformed under laboratory conditions The majority of thesecompounds (in the molecular weight range 86ndash204 g molminus1)

were products of ring cleavage the most abundant of whichincluded maleic acid (116) ketonorlimonic acid (174)ketolimonic acid (186) 5-hydroxyketolimonic acid (202)

3-carboxyheptanedioic acid (204) and 7-hydroxylimononicacid (200) (molecular weight given in parentheses) Alsoidentified in significant abundance in the aerosol phasewas the ring retaining compound 4-isopropenyl-1-methyl-1-hydroxy-2-oxocyclohexane (168) Of these compoundsonly ketolimonic acid and 7-hydroxylimononic acid wereidentified in the present study with the latter compound be-ing the most abundant compound in this study A furthersix particle phase compounds were found to be common be-tween this study and that of Jaoui et al these include limon-aldheyde (168) ketolimononaldehyde (170) limonalic acid(170) limononic acid (184) 7-hydroxylimononaldehyde(184) and limonic acid (186) No ring retaining compoundswere identified in the SOA of the current study

FTICR-MS was used in negative mode to determinemolecular formulae for each peak with a relative intensitygreater than 03 of the base peak Peaks were assumedto contain only C H and O and molecular formulae wereconverted into HC and OC ratios Comparison of the re-sults with a chamber background indicated that a numberof peaks were due to contamination These were identifiedas saturated fatty acids from the solvents used (not seen inGCtimes GC) and these ions were removed from any furtheranalysis A further manual inspection of molecular formulaewas used to remove unrealistic assignments based on maxi-mum OC = 2 and the degree of unsaturation (rings + doublebonds) Rather than investigate the detailed composition abulk view of the degree of oxidation was used for compari-son by plotting 3 dimensional Van Krevelen diagrams wherethe x- and y- axis represent the OC and HC ratios respec-tively and a coloured contour is used to represent peak in-tensity Although peak intensity does not scale necessarilywith concentration a comparison of the relative intensity pat-terns is still informative The 3-D Van Krevelen diagrams forthe seed and limonene SOA + seed are shown in Fig 7 The



Table 2 Limonene oxidation products identified in seeded SOA

34

Table 2 Limonene oxidation products identified in seeded SOA 996

Mass of product [M+Na]+ Potential molecular formula Identified products

116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6

seed plot shows a narrow OC range between approximately01 and 04 with an intensity weighted mean value of 032The peaks take up a very narrow region of the Van Krevelenspace In contrast the limonene SOA + seed is more spreadout with some molecules having much higher OC ratiosThe OC range was between 01 and 08 and the intensityweight mean was 037 It is also apparent that the degree ofunsaturation increases with an increase in lower HC valuesA similar increase of 005 is also seen in positive ionisationmode Limonene oxidation products are therefore smallerand more polar which will in turn increase their water solu-bility This may lead to a further skewing of the OC ratioHowever it is clear that theβ-caryophyllene first generationproducts are still sufficiently water soluble to be extractedAn increase in OC and decrease in HC should result in an

increase in hygroscopicity which is indeed seen to be the casehere The data qualitatively falls within the OC vs GF rela-tionship shown in Jimenez et al 2009 however this may notbe the case for other precursors

There is some evidence for formation of limonene +β-caryophyllene oligomers For example a peak atmz4552654 is identified as C24H39O8 which could be a com-bination of a C10 (limonene) and a C14(β-caryophyllene) ox-idation product However high molecular weight species(mzgt 350 Da) represent less than 5 of the total peak in-tensity These species are not seen in the LC-MS mass spec-tra and it is impossible at this time to rule out the formationof adducts during electrospray ionisation

4 Conclusions

In order to improve current simulation of SOA formationatmospherically representative seed particles are requiredThis work has shown thatβ-caryophyllene SOA generatedin-situ in the chamber shows promise as a more realistic or-ganic seed particle It can be formed quickly is relativelystable cheap and easy to use The relatively high molecularweight of the seed constituents leads to components that canbe separated from the oxidation products of common VOCsuch as isoprene monoterpenes and aromatics without theneed for expensive isotopic labelling Limonene SOA for-mation was studied using this seed and a clear change in theaerosol properties was observed during photo-oxidation andsubsequent gas-particle partitioning Thef44 from the AMSand the OC ratio obtained from the FTICR-MS both indicatethat limonene SOA is more polar and more oxidised than theseed and this resulted in an increase in hygroscopicity of theparticles This type of experiment could be used in the futureto study the effect of the seed on limonene SOA compositionby comparison to nucleation type experiments

AcknowledgementsThe authors gratefully acknowledge the UKNatural Environment Research Council for funding of the ACESproject (NEE0111601) and the APPRAISE program FTICR-MSmeasurements were made by Dr Ed Bergstrom University ofYork using facilities available in the Centre of Excellence inMass Spectrometry The York Centre of Excellence in MassSpectrometry was created thanks to a major capital investmentthrough Science City York supported by Yorkshire Forward withfunds from the Northern Way Initiative

Edited by R MacKenzie



References

Aiken A C DeCarlo P F Kroll J H Worsnop D R Huff-man J A Docherty K S Ulbrich I M Mohr C KimmelJ R Sueper D Sun Y Zhang Q Trimborn A NorthwayM Ziemann P J Canagaratna M R Onasch T B AlfarraM R Prevot A S H Dommen J Duplissy J MetzgerA Baltensperger U and Jimenez J L OC and OMOC ra-tios of primary secondary and ambient organic aerosols withhigh-resolution time-of-flight aerosol mass spectrometry Envi-ron Sci Technol 42 4478ndash4485 2008

Alfarra M R Coe H Allan J D Bower K N Boudries HCanagaratna M R Jimenez J L Jayne J T Garforth A ALi S M and Worsnop D R Characterization of urban and ru-ral organic particulate in the lower Fraser valley using two aero-dyne aerosol mass spectrometers Atmos Environ 38 5745ndash5758 2004

Alfarra M R Paulsen D Gysel M Garforth A A DommenJ Prevot A S H Worsnop D R Baltensperger U and CoeH A mass spectrometric study of secondary organic aerosolsformed from the photooxidation of anthropogenic and biogenicprecursors in a reaction chamber Atmos Chem Phys 6 5279ndash5293doi105194acp-6-5279-2006 2006

Alfarra M R Hamilton J F Wyche K P Good N WardM W Carr T Lewis A C Monks P S and McFiggansG B The effect of photochemical ageing and initial precursorconcentration on the composition and hygroscopic properties ofβ-caryophyllene secondary organic aerosol Atmos Chem PhysDiscuss submitted 2011

Bahreini R Keywood M D Ng N L Varutbangkul V Gao SFlagan R C Seinfeld J H Worsnop D R and Jimenez J LMeasurements of secondary organic aerosol from oxidation ofcycloalkenes terpenes and m-xylene using an Aerodyne aerosolmass spectrometer Environ Sci Technol 39 5674ndash5688 2005

Blake R S Whyte C Hughes C O Ellis A M and MonksP S Demonstration of proton-transfer reaction time-of-flightmass spectrometry for real-time analysis of trace volatile organiccompounds Anal Chem 76 3841ndash3845 2004

Blake R S Monks P S and Ellis A M Proton-Transfer Reac-tion Mass Spectrometry Chem Rev 109 861ndash896 2009

Corrigan A L Hanley S W and Haan D O Uptake of gly-oxal by organic and inorganic aerosol Environ Sci Technol42 4428ndash4433 2008

Cubison M J Alfarra M R Allan J Bower K N Coe HMcFiggans G B Whitehead J D Williams P I Zhang QJimenez J L Hopkins J and Lee J The characterisation ofpollution aerosol in a changing photochemical environment At-mos Chem Phys 6 5573ndash5588doi105194acp-6-5573-20062006

Dommen J Helle|un H Saurer M Jaeggi M Siegwolf RMetzger A Duplissy J Fierz M and Baltensperger UDetermination of the Aerosol Yield of Isoprene in the Pres-ence of an Organic Seed with Carbon Isotope Analysis Envi-ronSciTechnol 43 6697ndash6702 2009

Drewnick F Hings S S DeCarlo P Jayne J T Gonin MFuhrer K Weimer S Jimenez J L Demerjian K L Bor-rmann S and Worsnop D R A new time-of-flight aerosolmass spectrometer (TOF-AMS) ndash Instrument description andfirst field deployment Aerosol Sci Technol 39 637ndash658 2005

Fuzzi S Andreae M O Huebert B J Kulmala M Bond

T C Boy M Doherty S J Guenther A Kanakidou MKawamura K Kerminen V-M Lohmann U Russell L Mand Poschl U Critical assessment of the current state of scien-tific knowledge terminology and research needs concerning therole of organic aerosols in the atmosphere climate and globalchange Atmos Chem Phys 6 2017ndash2038doi105194acp-6-2017-2006 2006

Good N Coe H and McFiggans G Instrumentational op-eration and analytical methodology for the reconciliation ofaerosol water uptake under sub- and supersaturated conditionsAtmos Meas Tech 3 1241ndash1254doi105194amt-3-1241-2010 2010

Grieshop A P Donahue N M and Robinson A L Isthe gas-particle partitioning in alpha-pinene secondary or-ganic aerosol reversible Geophys Res Lett 34 L14810doi1010292007GL029987 2007

Gysel M McFiggans G B and Coe H Inversion of tandemdifferential mobility analyser (TDMA) measurements J AerosolSci 40 134ndash151 2009

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 Prevot A S H Seinfeld J H Surratt J DSzmigielski R and Wildt J The formation properties andimpact of secondary organic aerosol current and emerging is-sues Atmos Chem Phys 9 5155ndash5236doi105194acp-9-5155-2009 2009

Hamilton J F Using Comprehensive Two-Dimensional Gas Chro-matography to Study the Atmosphere J Chromat Sci 48 274ndash282 2010

Hamilton J F Webb P J Lewis A C and Reviejo MM Quantifying small molecules in secondary organic aerosolformed during the photo-oxidation of toluene with hydroxyl rad-icals Atmos Environ 39 7263ndash7275 2005

Hamilton J F Lewis A C Carey T J and Wenger J C Char-acterization of polar compounds and oligomers in secondary or-ganic aerosol using liquid chromatography coupled to mass spec-trometry Anal Chem 80 474ndash480 2008

Iinuma Y Muller C Berndt T Boge O Claeys M and Her-rmann H Evidence for the existence of organosulfates frombeta-pinene ozonolysis in ambient secondary organic aerosolEnviron Sci Technol 41 6678ndash6683 2007

Intergovermental Panel for Climate Change Climate Change 20072007

Jang M S Carroll B Chandramouli B and Kamens R MParticle growth by acid-catalyzed heterogeneous reactions of or-ganic carbonyls on preexisting aerosols Environ Sci Technol37 3828ndash3837 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 MJ Dunlea E J Huffman J A Onasch T B Alfarra M RWilliams P I Bower K Kondo Y Schneider J Drewnick



F 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 C EBaltensperger U and Worsnop D R Evolution of OrganicAerosols in the Atmosphere Science 326 1525ndash1529 2009

Li Y J Chen Q Guzman M I Chan C K and MartinS T Second-generation products contribute substantially tothe particle-phase organic material produced byβ-caryophylleneozonolysis Atmos Chem Phys 11 121ndash132doi105194acp-11-121-2011 2011

Liggio J and Li S M Organosulfate formation during the uptakeof pinonaldehyde on acidic sulfate aerosols Geophys Res Lett33 L13808doi1010292006GL026079 2006

Lindinger W Hirber J and Paretzke H An IonMolecule-Reaction Mass-Spectrometer Used for Online Trace Gas-Analysis Int J Mass Spectrom 129 79ndash88 1993

Odum J R Jungkamp T P W Griffin R J Forstner H J LFlagan R C and Seinfeld J H Aromatics reformulated gaso-line and atmospheric organic aerosol formation Environ SciTechnol 31 1890ndash1897 1997

Song C Zaveri R A Alexander M L Thornton J AMadronich S Ortega J V Zelenyuk A Yu X Y LaskinA and Maughan D A Effect of hydrophobic primary or-ganic aerosols on secondary organic aerosol formation fromozonolysis of alpha-pinene Geophys Res Lett 34 L20803doi1010292007GL030720 2007

Surratt J D Kroll J H Kleindienst T E Edney E O ClaeysM Sorooshian A Ng N L Offenberg J H LewandowskiM Jaoui M Flagan R C and Seinfeld J H Evidence fororganosulfates in secondary organic aerosol Environ Sci Tech-nol 41 517ndash527 2007

Varutbangkul V Brechtel F J Bahreini R Ng N L KeywoodM D Kroll J H Flagan R C Seinfeld J H Lee A andGoldstein A H Hygroscopicity of secondary organic aerosolsformed by oxidation of cycloalkenes monoterpenes sesquiter-penes and related compounds Atmos Chem Phys 6 2367ndash2388doi105194acp-6-2367-2006 2006

Volkamer R Jimenez J L San Martini F Dzepina K ZhangQ Salcedo D Molina L T Worsnop D R and Molina MJ Secondary organic aerosol formation from anthropogenic airpollution Rapid and higher than expected Geophys Res Lett33 L17811doi1010292006GL026899 2006

Volkamer R Ziemann P J and Molina M J Secondary Or-ganic Aerosol Formation from Acetylene (C2H2) seed effect onSOA yields due to organic photochemistry in the aerosol aque-ous phase Atmos Chem Phys 9 1907ndash1928doi105194acp-9-1907-2009 2009

Williams P I Construction and Validation of a DMPS for AerosolCharacterisation 1999

Wyche K P Blake R S Ellis A M Monks P S Brauers TKoppmann R and Apel E C Technical Note Performance ofChemical Ionization Reaction Time-of-Flight Mass Spectrome-try (CIR-TOF-MS) for the measurement of atmospherically sig-nificant oxygenated volatile organic compounds Atmos ChemPhys 7 609ndash620doi105194acp-7-609-2007 2007

Zhang Q Jimenez J L Canagaratna M R Allan J D CoeH Ulbrich I Alfarra M R Takami A Middlebrook AM Sun Y L Dzepina K Dunlea E Docherty K De-Carlo P F Salcedo D Onasch T Jayne J T MiyoshiT Shimono A Hatakeyama S Takegawa N Kondo YSchneider J Drewnick F Borrmann S Weimer S Demer-jian K Williams P Bower K Bahreini R Cottrell LGriffin R J Rautiainen J Sun J Y Zhang Y M andWorsnop D R Ubiquity and dominance of oxygenated speciesin organic aerosols in anthropogenically-influenced NorthernHemisphere midlatitudes Geophys Res Lett 34 L13801doi1010292007GL029979 2007



general owing to the widespread presence of pre-existingcondensation surfaces in the real atmosphere organic nucle-ation is not widely observed the more common phenomenonis the propagation of organic aerosol mass via the condensa-tion of oxidized VOC products onto pre-existing aerosolsBetter simulation in chamber experiments can therefore beachieved by including background or ldquoseedrdquo particles for ox-idised products to partition onto

It is a criterion of absorptive partitioning theory that allcondensed components are miscible across the whole con-centration range and thus the greatest amount of interac-tion is likely to take place when seed particles have similarproperties as the compounds condensing onto them Histor-ically most chamber experiments have used inorganic aque-ous solutions of ammonium sulphate or sulphuric acid asseed aerosols to study SOA formation (Hallquist et al 2009)sometimes dried prior to use (eg Odum et al 1997) Suchstudies have shown the formation of organosulphate speciesin monoterpene and isoprene SOA which have also beenseen in ambient samples (Iinuma et al 2007 Liggio andLi 2006 Surratt et al 2007) Volkamer et al (2009) foundthat the yield of SOA from acetylene photo-oxidation wherethe main product is glyoxal strongly depends on the chem-ical composition of the seed particles In this case use ofan ammonium sulphate seed led to one of the lowest SOAyields However the yield was significantly increased by theaddition of fulvic acid to the ammonium sulphate Jang etal (2003) used seed particles from diesel engine exhaust inchamber studies to mimic primary organic aerosol How-ever diesel and fulvic acid are made up of thousands ofcompounds and identifying SOA components arising from aknown precursor in such a complex matrix is extremely dif-ficult In both cases there is a disparity between the chemicalproperties of the seed and the oxidised organic compoundspartitioning to form SOA In the atmosphere most organicaerosol (OA) is oxidised even in urban areas (Jimenez et al2009 Zhang et al 2007) Thus a better seed particle tosimulate many environments would be an oxidised organicaerosol (OOA) There are a number of difficulties involvedwith using OOA as a seed Firstly which organic compoundis most suitable Previous studies have used dioctylphthalate(Song et al 2007) and highly oxidised organic acids (Corri-gan et al 2008) as a surrogate of SOA but these are singlecompounds rather than the range of molecular weights andfunctionalities seen in the atmosphere Oxidised compoundsare also generally difficult to work with and stick to surfaces

A more appropriate seed for investigating the partitioningof a wide range of semi-volatile OVOCs would be SOA it-self Dommen et al (2009) usedα-pinene SOA as a seedaerosol for isoprene SOA although the isoprene was13C la-belled to allow the seed to be differentiated This methodrequired the use of plants to fix13CO2 which then emit-ted labelled isoprene and thus is limited to this particu-lar VOC Here we present the use of sesquiterpene SOAformed from β-caryophyllene photo-oxidation as a seed

particle for simulation chamber experiments Within thiswork β-caryophyllene SOA seed has been used to studythe generation of SOA from limonene photo-oxidationβ-Caryophyllene reacts rapidly with hydroxyl radicals andozone resulting in almost immediate nucleation owing to thehigh molecular weight and low volatility of its first genera-tion oxidation products The chamber was subsequently di-luted to obtain a low particle concentration and remove themajority of the gas phase products The VOC to be studiedin this case limonene was then introduced into the chamberand oxidised The gas phase loss of limonene and the for-mation of its oxidation products were followed by ChemicalIonisation Reaction Time of Flight Mass Spectrometry (CIR-TOF-MS) The partitioning of limonene oxidation productsto the seed was followed by monitoring the shift in the par-ticle size distribution The change in bulk chemical compo-sition was analysed in real time using Aerosol Mass Spec-trometry The difference in the detailed chemical composi-tion between the seed and the limonene SOA was analysedoffline using Comprehensive Two-dimensional Gas Chro-matography Time-of-Flight Mass Spectrometry (GCtimesGC-TOFMS) Liquid Chromatography Ion Trap Mass Spec-trometry (LC-MSn) and Fourier Transform Ion CyclotronResonance Mass Spectrometry (FTICR-MS)

2 Experimental

21 Chamber

A comprehensive and detailed description of the Universityof Manchester Aerosol Chamber will be presented in a futurepublication (McFiggans et al 2011) hence only an introduc-tion and general overview is provided here The Manchesteraerosol chamber is an 18 m3 (3 m (H)times 3 m (L) m (H)times 3 m2 m (W)) FEP Teflon bag mounted on three rectangular alu-minium frames The central rigid rectangular frame is fixedwith the upper and lower frames free to move vertically caus-ing the bag to expand and collapse as sample air is intro-duced and extracted A bank of halogen lamps and a 6 kWXenon arc lamp are mounted on the enclosure housing thebag which is coated with reflective ldquospace blanketrdquo servingto render the enclosure as an integrating sphere maximisingthe irradiance in the bag and ensuring even illumination Thecombination of illumination has been tuned and evaluatedto mimic the atmospheric actinic spectrum over the wave-length range 290ndash800 nm and has a maximum total actinicflux of 07times 1018 photos sminus1 mminus2 nmminus1 over the region 460ndash500 nm The calculatedj (O1D) value during the reported ex-periments was 36times 10minus5 sminus1 (290ndash340 nm) Air condition-ing between the enclosure and the bag removes additionalunwanted heat generated by the lamps The air charge inthe bag is dried and filtered for gaseous impurities and par-ticles using a combination of Purafil (Purafil SP media fromPurafil Inc USA) charcoal and HEPA filters (Donaldson














H3O++Rminusrarr RH+

+H2O (R1)




































26


experiment 869

870

871 872

873

874

875

876

877

878

879

880




27




884

60x103

50

40

30

20

10

0

dN

dlo

gD

p

30025020015010050

Dp (nm)

10000

8000

6000

4000

2000

0

dN

dlo

gD

p


885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903







28



906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929




YSOA=Mp

1VOC(1)





29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References





















































H3O++Rminusrarr RH+

+H2O (R1)




































26


experiment 869

870

871 872

873

874

875

876

877

878

879

880




27




884

60x103

50

40

30

20

10

0

dN

dlo

gD

p

30025020015010050

Dp (nm)

10000

8000

6000

4000

2000

0

dN

dlo

gD

p


885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903







28



906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929




YSOA=Mp

1VOC(1)





29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References









































































26


experiment 869

870

871 872

873

874

875

876

877

878

879

880




27




884

60x103

50

40

30

20

10

0

dN

dlo

gD

p

30025020015010050

Dp (nm)

10000

8000

6000

4000

2000

0

dN

dlo

gD

p


885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903







28



906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929




YSOA=Mp

1VOC(1)





29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References






















































26


experiment 869

870

871 872

873

874

875

876

877

878

879

880




27




884

60x103

50

40

30

20

10

0

dN

dlo

gD

p

30025020015010050

Dp (nm)

10000

8000

6000

4000

2000

0

dN

dlo

gD

p


885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903







28



906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929




YSOA=Mp

1VOC(1)





29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References










































26


experiment 869

870

871 872

873

874

875

876

877

878

879

880




27




884

60x103

50

40

30

20

10

0

dN

dlo

gD

p

30025020015010050

Dp (nm)

10000

8000

6000

4000

2000

0

dN

dlo

gD

p


885

886

887

888

889

890

891

892

893

894

895

896

897

898

899

900

901

902

903







28



906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929




YSOA=Mp

1VOC(1)





29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References










































28



906

907

908

909

910

911

912

913

914

915

916

917

918

919

920

921

922

923

924

925

926

927

928

929




YSOA=Mp

1VOC(1)





29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References










































29








minutes) 937

938

939

940

941

942

943

944

945

946








30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References










































30




950

951

952

953

954

955

956

957

A

B




31



960

961

962

963

0

500000

1000000

1500000

2000000

2500000

3000000

50 100 150 200 250 300 350

mz

Inte

nsit

y

soa seed

0

200000

400000

600000

800000

1000000

1200000

1400000

50 100 150 200 250 300 350

limonene on soa

-1000000

-800000

-600000

-400000

-200000

0

200000

400000

600000

50 100 150 200 250 300 350





32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References










































32


965

966

967

968

969

970

971

972

973

974

975

976

977

978

979

980

981

982

983

984

985

986

987










34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References











































34



116 139 C5H8O3 C6H12O2

154 177 C9H14O2 O

O

O

156 179 C9H16O2 C8H12O3

O

O

HO

OR

O

O

OH

168 191 C10H16O2

O

O

170 193 C9H14O3

O

O

O

AND

O

OH

O

172 195 C8H12O4

O

O

OH

O

AND

O

O

HO

O

184 207 C10H16O3

O

O

OH

AND

O

O

HO

186 209 C9H14O4

O

OH

O

OH

188 211 C8H12O5

O

O

OH

O

OH

200 223 C10H16O4 O

O

OH

HO

216 239 C10H16O5 C9H12O6




4 Conclusions






References










































References
























































Documents

Investigating the use of secondary organic aerosol as seed ... · compounds and identifying SOA components arising from a known precursor in such a complex matrix is extremely dif-ﬁcult