Characteristics, sources and water-solubility of ambient submicron organic aerosol in springtime in Helsinki, Finland

Journal of Aerosol Science 56 (2013) 61–77

Contents lists available at SciVerse ScienceDirect

Journal of Aerosol Science journal homepage: www.elsevier.com/locate/jaerosci

Characteristics, sources and water-solubility of ambient submicron organic aerosol in springtime in Helsinki, Finland Hilkka Timonen a,b,n, Samara Carbone a, Minna Aurela a, Karri Saarnio a, Sanna Saarikoski a, Nga L. Ng c, Manjula R. Canagaratna d, Markku Kulmala e, Veli-Matti Kerminen e, Douglas R. Worsnop a,d,e, Risto Hillamo a a

Air Quality Research, Finnish Meteorological Institute, P.O. Box 503, FI-00101 Helsinki, Finland Department of Science and Technology, University of Washington, Bothell, US c School of Chemical and Biomolecular Engineering and School of Earth and Atmospheric Sciences, Georgia Institute of Technology, Atlanta, GA 30332, USA d Aerodyne Research, Inc. 45 Manning Road, Billerica, MA 01821-3976, USA e Department of Physics, University of Helsinki, P.O. Box 64, FI-00014 University of Helsinki, Finland b

a r t i c l e i n f o

abstract

Available online 21 June 2012

In this study the characteristics, sources and water-solubility of submicron organic aerosol (OA) were investigated in Helsinki, Finland. An Aerodyne high-resolution timeof-flight aerosol mass spectrometer (HR-ToF-AMS) was used to determine the submicron non-refractory aerosol components nitrate, sulfate, ammonium, chloride and organics between April 9 and May 8, 2009. The concentrations of the major watersoluble ions and water-soluble organic carbon (WSOC) were measured by a particleinto-liquid sampler (PILS) combined with a total organic carbon (TOC) analyzer and two ion chromatographs (IC) between April 25 and May 28, 2009. Parallel measurements of the submicron particulate matter (PM1), organic carbon (OC), black carbon (BC), meteorological quantities and trace gases were used to complement and validate the AMS and PILS-TOC-IC data. Sources or atmospheric processes affecting the organic aerosol were investigated by applying the Positive Matrix Factorization (PMF) analysis to the high-resolution mass spectra of the HR-ToF-AMS organics. All together seven factors were needed to describe the variation in the obtained dataset. The factors consisted of two different types of low-volatility oxygenated OA (LV-OOA), local and long-range-transported (LRT) biomass burning OA (BBOA), semi-volatile OA (SV-OOA), hydrocarbon-like OA (HOA), and one local source (coffee roastery). These factors were interpretable and could be connected to specific sources or chemical characteristics (biomass burning, traffic, biogenic emissions, oxidized long-range-transported aerosols, marine-processed aerosols and nearby industrial activity) of ambient aerosols. In order to study the organic fraction and PMF factors further, the elemental ratios OM:OC, O:C, H:C and N:C were calculated. The value of the OM:OC ratio varied between 1.4 and 2.1. A high OM:OC ratio (1.5–2.1) was observed for the highly-oxidized and water-soluble fraction, whereas this ratio was clearly lower (1.2–1.4) for local and fresh sources such as traffic. Two different factors representing local and long-range-transported biomass burning were observed. Local biomass burning emissions had a lower OM:OC ratio, indicating that this factor was less aged and had a different source area compared with the LRT BBOA. The water-solubilities of the OA factors were studied by investigating the

Keywords: Urban aerosol Aerosol mass spectrometer PMF Organic matter Water-solubility

n Corresponding author at: Finnish Meteorological Institute, P.O. Box 503, FI-00101 Helsinki, Finland. Tel.: þ 358 9 1929 5503; fax: þ 358 9 1929 5403. E-mail address: hilkka.timonen@fmi.fi (H. Timonen).

0021-8502/$ - see front matter & 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.jaerosci.2012.06.005

62

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

correlation between these factors and WSOC and by reconstructing the concentration of water-soluble particulate organic matter (WSPOM) from the OA factors. The reconstructed WSPOM had a good correlation with the measured concentration of WSPOM. & 2012 Elsevier Ltd. All rights reserved.

1. Introduction Organic aerosol (OA) comprises a large fraction of the ambient submicron aerosol mass and has therefore a significant impact on climate, visibility and human health (e.g. Kanakidou et al., 2005; Jimenez et al., 2009; Heald et al., 2011). Due to its complex nature, the knowledge about the sources, behavior and chemical composition of the organic fraction is limited. The compounds in the organic fraction are typically divided into primary and secondary organic aerosol (POA and SOA, respectively) according to their origin. POA refers to organic compounds that are directly emitted in a particulate form or vapors that condense onto particles without undergoing gas-phase chemistry, whereas SOA is formed in the atmosphere by gas-to-particle conversion. The fact that the OA consist of thousands of different compounds that continually change both chemically and physically depending on atmospheric conditions (Robinson et al., 2007; Hallquist et al., 2009; Jimenez et al., 2009) makes the characterization of OA challenging. In order to understand the atmospheric behavior of OA and to determine its health and climate effects, the chemical composition of OA must be known. During the last decade, significant progress has been made in characterizing the organic compounds of atmospheric aerosols on a molecular level. Individual compounds have been shown to be tracers of specific sources, so measuring such tracer compounds makes it possible to get detailed insights into aerosol precursors and aerosol formation processes (Hallquist et al., 2009). The major tracers for biomass burning, biogenic SOA or marine aerosols have been chemically characterized (e.g. Simoneit et al.; 1999; Phinney et al., 2006; Hallquist et al., 2009 and references therein; Yasmeen et al., 2011; Zhang et al., 2011 and references therein). However, a full and extensive chemical characterization of organic fraction based on offline sampling and chemical analyses is usually not feasible. Therefore, other approaches have been applied to gather information on the organic fraction of the atmospheric aerosol. A large number of high-time-resolution, online devices, such as the Aerodyne aerosol mass spectrometer (AMS), semi-continuous EC/OC carbon aerosol analyzer and particle-into-liquid sampler (PILS) coupled with either ion chromatographs (IC) or total organic carbon (TOC) analyzer, have been developed and used intensively during the last decade. Compared with offline sampling and subsequent chemical analyses, online methods typically offer artifact-free data with high time resolution. In addition, time-resolved data enable the chemical characterization of OA during fast processes, including different combustion processes (Liu et al., 2011) and evolution and aging of aerosols (Zhang et al., 2007; Jimenez et al., 2009; Heald et al., 2010; Morgan et al., 2010a,b; Ng et al., 2010; Hennigan et al., 2011; Ng et al., 2011a). The atmospheric organic aerosol can be divided into a water-soluble and water-insoluble fraction. The water-soluble fraction represents the highly-oxidized and typically long-range transported aged fraction of OA. The water-solubility affects the chemical and physical properties of aerosols, such as its acidity and radiative properties as well as its ability to act as cloud condensation nuclei (e.g. Jacobson et al., 2000; Saxena & Hildemann, 1996). The main sources of water-soluble OA are secondary organic aerosol formation and biomass burning (Saxena & Hildemann, 1996; Decesari et al., 2006; Hennigan et al., 2009; Sun et al., 2011). The water-insoluble fraction often represents local or regional fresh emissions, like traffic. The primary goal of this study was to explore the sources and characteristics of the submicron organic aerosol at an urban background station in Helsinki, Finland. The OA was measured with an Aerodyne high-resolution time-of-flight aerosol mass spectrometer (HR-ToF-AMS), and the measured data were analyzed using Positive Matrix Factorization (PMF). The results obtained from the HR-ToF-AMS measurements were compared with simultaneous high-resolution measurements of water-soluble organic carbon (WSOC), trace gas measurements and meteorological quantities. To our knowledge, this is the first study in Europe where highly time-resolved WSOC and inorganic ion data have been available concurrently with the HR-ToF-AMS data, making it possible to investigate the solubility of different fractions of OA. An additional aim of this study was to investigate whether it is possible to reconstruct the water-soluble organic fraction of the aerosol based on AMS data.

2. Experimental 2.1. Measurement site ¨ The measurements were conducted at an urban background station, SMEAR III (601 120 , 241 570 , 30 m a.s.l., Jarvi et al., 2009). The SMEAR III station has been established for conducting long-term measurements of chemical and physical properties of atmospheric aerosols, trace gas concentrations, meteorological quantities and turbulent fluxes. The SMEAR III station has an air-conditioned container for scientific instruments and a 31-m-high measurement tower for flux and meteorology measurements at different heights. The site is situated at the University of Helsinki campus area about 5 km northeast from the city center of Helsinki. Next to the station there are the buildings of the Finnish Meteorological Institute

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

63

(FMI) and University of Helsinki. About 200 m east from the station, there is a major road with heavy traffic (60,000 cars/day). A small forest and some buildings are also located west from the station. 2.2. Aerodyne high-resolution time-of-flight aerosol mass spectrometer An Aerodyne high-resolution time-of-flight aerosol mass spectrometer (AMS; Aerodyne Research Inc, USA) was used to determine the concentrations of major species of the submicron particulate matter (PM1): organics, sulfate, nitrate, chloride and ammonium. The measurement method has been described in detail by DeCarlo et al. (2006). Briefly, in the HR-ToF-AMS an aerodynamic lens is used to form a narrow beam of particles that is transmitted into the detection chamber, in which the non-refractory fraction of the aerosol is flash-vaporized upon impact on a hot surface (600 1C) under high vacuum (Jayne et al., 2000). Crustal material, sea salt and black carbon are not vaporized at this temperature and therefore cannot be detected using this technique. The vaporized compounds are ionized using electron ionization (70 eV) and the formed ions are guided to the time-of-flight chamber. The time-of-flight chamber has two ion optical modes: V-mode and W-mode. The V-mode is a single-reflection configuration, in which ions follow a V-shaped path from the extraction region to a reflector and then back to the detector. The W-mode is a two-reflection configuration, in which ions follow a W-shaped bath from the extraction region to a reflector, back to the hard mirror, a second time to the reflector, and finally to the detector. The W-mode has a higher resolving power (up to 4000 at m/z 200) than the V-mode (up to 2000 at m/z 200), but a lower sensitivity. A multi-channel plate is used as a detector. In this study, the time resolution of the AMS measurements was five minutes and the instrument was alternating between the V-mode (2.5 min) and W-mode (2.5 min). The high resolution of W-mode facilitates the separation and identification of isobaric ions, which are ions with same nominal mass but different exact mass (Gross 2004). The source apportionment accomplished in the W-mode dataset presented similar results to the V-mode, including the numbers of factors chosen in the solution and their mass spectra. However, due to the hard mirror malfunction there were several interruptions in the W-mode dataset and therefore the V-mode was chosen for being more continuous. The one-minute detection limits for the submicron aerosol are o0.04 mg m  3 for all the compounds in the V-mode (DeCarlo et al., 2006). The collection efficiency (CE) represents the fraction of sampled particle mass that is detected with the AMS. The value of CE is required for the estimation of aerosol mass concentration measured by the AMS. The CE of an aerosol mass spectrometer is affected, for example, by the incomplete collection of particles in the vaporizer and by properties (e.g. phase, composition and morphology) of vaporized particles (Huffman et al., 2005; Matthew et al., 2008; Cross et al., 2009). Based on previous studies (Canagaratna et al., 2007) and comparisons with the other instruments in this study (PILS, semicontinuous EC/OC carbon aerosol analyzer, and tapered element oscillating micro-balance), the value of 0.5 for the CE was used in this paper. 2.2.1. Positive matrix factorization analysis Positive Matrix Factorization (PMF; Paatero & Tapper, 1994; Paatero, 1997) is a multivariate factor analysis tool that decomposes a sample data matrix into two or more matrices with varying factor contributions (in this case the concentration at given time) and constant factor profiles (in this case the mass spectrum representing certain source or similar chemical composition). The IGOR 6.11 (Wavemetrics, Lake Oswego, OR), Squirrel 1.47 (Sueper, 2011) and a PMF Evaluation Tool (PET; described by Ulbrich et al., 2009) were used to analyze the AMS data. The APES (Analytic Procedure for Elemental Separation; Aiken et al., 2007, 2008) was used for the elemental analysis of the AMS data. 2.3. PILS-TOC-IC A particle-into-liquid sampler combined with a total organic carbon analyzer (TOC-VCPH, Shimadzu, Japan) and two ion þ chromatographs (Dionex, Sunnyvale, USA) were used to measure the concentrations of major ions (Na þ , NH4 , K þ , Mg2 þ ,  2 2þ  Ca , Cl , NO3 , SO4 , oxalate) and WSOC online. The method is described in detail by Timonen et al. (2010). Shortly, one TOC-VCPH and two IC instruments were connected to a PILS to enable simultaneous measurements of water-soluble organic carbon and major ions. A PM1 cyclone (sharp cut cyclone, SCC 1.829, BGI Inc. US) and two denuders (a parallel plate carbon filter denuder from Sunset Laboratory Inc., Portland, OR and an annular denuder URG-2000, 30  242 mm, Chapel Hill, NC coated with H3PO4) were used to cut off supermicron particles, gaseous organic compounds and ammonia prior to the PILS. One half of the sample liquid produced by the PILS was filtered by a polytetrafluoroethylene (PTFE) filter in order to remove the water-insoluble carbonaceous particles from the sample that was then fed to the TOC-VCPH analyzer for a subsequent online WSOC analysis. The other half of the collected PILS sample was further split into two equal fractions by a simple t-shape splitter. The split liquid flows were fed to the sample loops of the two IC systems (Dionex ICS-2000;  2 þ Dionex, Sunnyvale, USA) in order to measure anions (Cl–, NO3 , SO4 , oxalate) and cations (Na þ , NH4 , K þ , Mg2 þ , Ca2 þ ). The time resolution of the PILS-TOC-IC measurements depended on the sample volume needed for the subsequent analysis and sensitivity of the analytical instrument. Therefore, a six-minute time resolution was used for the WSOC and 15-min time resolution was used for the inorganic ions. Blank values were measured for the PILS-TOC-IC system by directing the air flow through a PTFE filter prior to the PILS. The limit of quantification for the WSOC in the TOC-VCPH was 4 mg l  1, which is equal to the concentration of 0.15 mg m  3 in air. Based on the analyzed test samples, the uncertainty of the IC

64

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

analysis was of the order of 10–15% for all analyzed ions. The quantification limit for the major ions was 2.5 ng ml  1, which is equal to the concentration of 0.1 mg m  3 in air. The measured WSOC concentrations were multiplied by the estimated average molecular weight per carbon weight in aerosols in order to estimate the amount of particulate organic matter (WSPOM). For WSOC, a constant organic matter to organic carbon (OM:OC) ratio of 1.6 was used based on Turpin & Lim (2001) and Saarnio et al. (2010). 2.4. Semi-continuous EC/OC carbon aerosol analyzer A semi-continuous EC/OC carbon aerosol analyzer (Sunset Laboratory Inc, Oregon) was used to measure the concentrations of elemental (EC) and organic carbon (OC) (Turpin et al., 1990). The method has been described in detail by Saarikoski et al. (2008). Shortly, during a measurement cycle the instrument collects a sample for 164 min. After the sampling period, particles deposited on a collection filter are heated in a quartz oven, in which elemental and organic carbon concentrations are individually quantified. During the first phase of the analysis, the sample is purged with helium and the temperature is raised in steps up to 650 1C in order to vaporize organic carbon. In the second phase, the sample is purged with a helium–oxygen mixture and the temperature is again raised in steps from 650 to 850 1C to oxidize all elemental carbon. In the first measurement phase, pyrolysis converts a part of organic carbon to a light-absorbing substance that resembles EC (Johnson & Huntzicker, 1979; Viidanoja et al., 2002). This part of OC is not measured until it has been oxidized in the second phase. The value of laser transmission is used to separate the pyrolyzed OC from EC. Vaporized carbon compounds formed in the oven are purged to MnO2 catalyst, in which they are further oxidized to carbon dioxide and quantified with a non-dispersive infrared detector. Since elemental carbon concentrations were rather low in Helsinki, ‘‘optical EC’’ concentrations instead of ‘‘thermal EC’’ ones were used in this study. ‘‘Optical EC’’ is determined by optical methods (light transmission) and is therefore referred to as black carbon (BC) hereafter. 2.5. Tapered element oscillating microbalance A tapered element oscillating microbalance (TEOM1400a, Rupprecht & Patashnick Co. Inc, USA) was used to measure PM1.3 mass. In the TEOM, the sample is collected to a filter that is placed on the top of an oscillating tapered cone. The mass concentration is calculated from changes in the oscillation frequency. When comparing the TEOM measurements with those of other instruments, 30-min average concentrations were used. A virtual impactor was used to cut off particles larger than 1.3 mm of aerodynamic diameter prior to the TEOM. The TEOM has been observed to underestimate ambient PM concentrations due to the evaporation of semi-volatile material from the TEOM collection filter (Grover et al., 2006; Wilson et al., 2006). The error caused by this evaporation was not corrected in this study, since the required correction factor was uncertain (probably less than 20% of the concentrations, Saarikoski et al., 2007) and should have no influence on our main conclusions. 2.6. Meteorological and gas data Local meteorological data were obtained from the Finnish Meteorological Institute Kumpula weather station (Vaisala, Milos 500) situated next to the SMEAR III station. The temperature was measured using a Pt100 (Pentronic Ab) sensor, the relative humidity was measured with a HMP45D (Vaisala Oyj) sensor, and the global radiation was measured with a CM11 (Kipp & Zonen) sensor. Carbon monoxide (CO) concentrations were measured at the SMEAR III station using a Horiba APMA 370 Analyzer (Horiba, Kyoto, Japan). To establish the potential source areas of the measured aerosol particles, 120-h air mass back trajectories were calculated for the sampling periods using FLEXTRA (Stohl & Wotawa, 1995). 3. Results and discussion 3.1. Description of the measurement period The online instruments, their time resolutions, measurement periods and measured components are summarized in Table 1. The HR-ToF-AMS measurements (major inorganic ions and organics) were conducted from April 9 to May 8, 2009. Table 1 Instruments, analyzed components and their time resolutions during the spring 2009 intensive campaign. Component/property

Instrument

Cutoff size (mm)

Time resolution

Measurement period

Ions, OC, mass Ions, WSOC

HR-ToF-AMS PILS-TOC-IC

1 1

April 9–May 8, 2009 April 29–May 29, 2009

OC, BC PM

semi-continuous EC/OC analyzer TEOM

2.5 1.3

2 min WSOC: 6 min Ions: 15 min 3h 1h

April 9–May 1, 2009 Continuous measurements

360 300 240 180 120 60 0

65

15 10

5 0 -5 BC OC WSOC

8 Concentration (µg m-3)

20

Temperature Wind direction

Temperature (°C)

WD (°)

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

6 4 2 0 30

Organics Sulfate Total mass

20

Nitrate Ammonium

10

0 4/11/2009

4/15/2009

4/19/2009

4/23/2009

4/27/2009

5/1/2009

5/5/2009

5/9/2009

Fig. 1. Wind direction (a), the concentrations of WSOC, black carbon and OC measured by the PILS-IC-TOC and semi-continuous EC/OC analyzer (b), and the concentrations of major ions and organics measured by the AMS (c) between April 9 and May 8, 2009.

The PILS-TOC-IC was used to measure the concentrations of major ions and WSOC from April 25 to May 29, 2011. Due to technical problems, the PILS-TOC-IC system was not working until April 25, and therefore the overlapping period of the AMS and PILS-TOC-IC measurements was only from April 25 to May 8, 2009. The concentrations of BC and OC were measured from April 9 to May 1, 2009, with the semi-continuous EC/OC carbon analyzer. The total mass of PM1.3 was continuously measured using the TEOM. Fig. 1 represents the time series of the concentrations of inorganic ions, BC, WSOC, OC and PM1.3 for the measurement period from April 9 to May 8, 2009. On average, the PM1.3 concentration was 6.6 mg m  3 during the measurement period. The major inorganic ions, black carbon and organic matter were responsible for 46%, 9% and 45%, respectively, of the analyzed particulate mass. During the PILS-TOC-IC measurements, 51% of the submicron particulate organic matter was water-soluble, on average. An excellent correlation was observed between the concentrations of all the high-time-resolution online measurement devices (Timonen et al., 2010). The temperature and wind direction during the intensive measurement period is presented in Fig. 1. The temperature was in the range 5 to 20 1C with a clear day-to-night variation representing typical springtime conditions in Finland. The prevailing wind direction was west/southwest (Fig. 1). Two short biomass burning episodes were observed during the measurement period. The backward trajectories showed that the air masses were long-range transported to Helsinki from forest fire areas in Russia from April 14 to 15 and again from April 26 to 29, 2009. During these periods, the measured PM1.3 concentrations (10–20 mg m  3) and concentrations of secondary ions (sulfate and nitrate), BC and OC were elevated. The AMS tracers of biomass burning (m/z 60.0211 and m/z 73.029) were also elevated.

3.2. PMF analysis of organic aerosol High-resolution AMS data combined with Positive Matrix Factorization has proven to be very useful in the characterization of organic aerosol (e.g. Lanz et al., 2007; Aiken et al., 2009; Ulbrich et al., 2009; Allan et al., 2010). By using the PMF method, the OA can be divided into factors representing the contribution of different sources like cooking, traffic or biomass burning, or into factors that represent compounds with similar chemical characteristics, such as lowvolatility oxygenated organic aerosol (LV-OOA) and nitrogen containing OA (Aiken et al., 2009; Ulbrich et al., 2009). In this study, the PMF was conducted on high-resolution, V-mode organic mass spectra measured with the HR-ToF-AMS. The final number of factors in the PMF is defined by the user, which is typically the most subjective part of the PMF analysis (Ulbrich et al., 2009; Zhang et al., 2011). If the number of factors is too low, the PMF will lump organic species from distinct sources and processes into single factors. Using too many factors, on the other hand, can result in splitting of PMF factors with unrealistic factor time trends and mass spectra (Ulbrich et al., 2009). In order to find an optimal number of factors for this dataset, PMF solutions with a number of factors varying from one to ten were calculated. The known m/z tracers in mass spectra, correlations with external data (e.g. inorganic ions, BC, gases and meteorological data) and statistical parameters calculated by the program (e.g. Q-value, Q/Qexp), were subsequently used to determine the number of factors needed to describe the variation in the dataset and to interpret the origin of the factors. In this study, a solution with nine factors was needed to identify seven different probable sources, from the local to the regional scale. A summary of diagnostics and results from the different factor solutions are shown in supplement (Table 1, Fig. S7 contributions of factors in 7–10 factor solutions and Fig. S8 Q/Qexp). With the smaller number of factors ( o9) the factor containing fragments of methane sulfonic acid (LV-OOAþMSA) could not be separated, and with the larger number of factors ( 49) a further splitting of factors was observed. The nine-factor solution explains almost all variation in the data, but unfortunately, in this solution LV-OOA splits into three factors with similar time series and mass spectra. The three LV-OOA factors were summed up in the final solution. Fig. 2 represents the time series of the factors and Fig. 3

66

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

fraction of signal

Fig. 2. The time series of the PMF factors during the spring 2009 campaign between April 9 and May 8, 2009.

Cx, HO

0.15 0.10 0.05 0.00 0.15 0.10 0.05 0.00 0.15 0.10 0.05 0.00 0.15 0.10 0.05 0.00 0.15 0.10 0.05 0.00 0.12 0.08 0.04 0.00 0.15 0.10 0.05 0.00

CH

CHN

CHO1

CHO1N

CS

CHOgt1

CHOgt1N

HOA

SV-OOA

Local BBOA

Coffee Roastery

LRT BBOA

LV-OOA + MSA

LV-OOA

10

20

30

40

50

60

70

80

90

100

110

120

130

m/z Fig. 3. The mass spectra of the factors during the spring 2009 campaign between April 9 and May 8, 2009.

represents the mass spectra of these factors. In this paper, only the results of this recombined seven factor solution are shown, and the corresponding results of the nine-factor solution are shown in the supplement (Supplement Fig. S1–S6).

3.2.1. PMF analysis Seven distinct organic aerosol factors were identified from the AMS data using the PMF analysis (Figs. 2 and 3). The identified factors included hydrocarbon-like OA (HOA) and three oxygenated OA (OOA) factors, of which two were lowvolatility OOA and one semi-volatile OOA (SV-OOA). Two factors were observed to represent emissions from biomass burning, the first representing local emissions and the second one long-range-transported emissions. The seventh factor

67

2.0

0.8 LV-OOA

1.5

0.6

HOA 0.4

1.0 SV-OOA Local BBOA

0.2

0.5

LRT BBOA LV-OOA + MSA

LV-OOA (µg m-3)

Coffee Roastery, LV-OOA + MSA, LRT BBOA, Local BBOA, SV-OOA, HOA (µg m-3)

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

Coffee Roastery 0.0

0.0

4

8

12 Hour of day

16

20

24

Fig. 4. Diurnal cycles of the factors.

Table 2 The relative contribution of each factor to the total organic mass measured by the HR-ToF-AMS during the spring 2009 campaign (April 9 to May 8, 2011). Factor

Contribution to OA mass (%)

LV-OOA LV-OOA þ MSA Local BBOA Coffee roastery LRT BBOA SV-OOA HOA

56 5 4 1 9 14 11

was identified to represent emissions from a nearby coffee roastery because the mass spectrum resembled that of caffeine from the NIST Chemistry WebBook (http://webbook.nist.gov/chemistry). Diurnal cycles calculated for each factor are shown in Fig. 4. 3.2.1.1. Characterization of factors. The oxygenated organic aerosol fraction is typically divided into two types by the PMF, called the low-volatility oxygenated organic aerosol (LV-OOA) and semi-volatile oxygenated organic aerosol (SV-OOA). The LV-OOA is expected to be aged, highly oxidized and processed organic fraction, originating mainly from secondary aerosol þ formation (Jimenez et al., 2009; Ng et al., 2010). The LV-OOA is dominated by the fragment CO2 . This signal is assumed to originate mainly from acids or acid-derived compounds (Alfarra et al., 2004; Duplissy et al., 2011; Ng et al., 2011a,b) that are known to be mostly water-soluble (Decesari et al., 2007). In chamber studies, it has been observed that the þ hygroscopicity of SOA is correlated strongly with the relative abundance of the fragment CO2 (Duplissy et al., 2011). The þ þ SV-OOA has two main fragments, C2H3O and C3H7 , and it has been shown to be composed of more-volatile, lessoxygenated and less-chemically-processed secondary material (Jimenez et al., 2009; Ng et al., 2010). The fragment C2H3O þ is predominantly due to non-acid oxygenates (Ng et al., 2011a). The hygroscopicity of SV-OOA has been demonstrated to be very low (Raatikainen et al., 2010). The LV-OOA has been observed to correlate well with WSOC in previous studies (Kondo et al., 2007; Sun et al., 2011 and Xiao et al., 2011), whereas the SV-OOA is clearly less correlated with WSOC. In the combined seven-factor solution, the LV-OOA represented on average 56% of the total organic mass (Table 2). The LV-OOA correlated well with inorganic ions, sulfate, nitrate and ammonium (r¼0.75, r¼0.61 and r¼0.84, respectively). In the time series of LV-OOA, two likely long-range transport episodes, with elevated LV-OOA concentrations, were observed. The prevailing wind direction was from southeast to southwest (125–2251) during the episodes (Fig. 5). Long-range transported emissions from southern Europe are typically observed in Finland when the wind direction is from the southern sector (Niemi et al., 2004; Saarikoski et al., 2008). The LV-OOA correlated clearly with WSOC (r¼ 0.90), whereas there was no correlation between the SV-OOA and WSOC (Fig. 6). On average, the SV-OOA represented 14% of the total organic fraction during the measurement period. For SV-OOA, the largest concentrations were observed when the wind direction was from the north with wide forested areas (Fig. 5). The observed SV-OOA mass spectrum for this study is similar to that of the oxidation products of a-pinene observed in chamber experiments (Bahreini et al., 2005; Kiendler-Scharr et al., 2009). The fragments of oxidation products of secondary VOC’s typically observed in ambient measurements of biogenically-influenced þ aerosol (e.g. fragments of methyl furan (C5H6O þ (m/z 82) and C4H5 (m/z 53), Robinson et al., 2011), are also clearly seen in

68

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

315

270

8 6 4 2 0

0 315

45

2 4 6 8

90

270

135

225

270

4 3 2 1 0

0 315

45

90 1 2 3 4

270

135

315

270

270

0 8 6 4 2 0

135

0 2.5 2 1.5 1 0.5 0

45

0.511.522.5

90

135

225

180

315

0 3 2.5 45 2 1.5 1 0.5 0 90 0.511.522.53

180

0 3 2.5 45 2 1.5 1 0.5 0 90 0.511.522.53

225

90

135

225

135 180

270

0.20.40.60.81

180

225

315

45

225

180

315

0 1 0.8 0.6 0.4 0.2 0

180

45 5/1/2009

2 4 6 8

90

4/21/2009 4/11/2009

135

225 180

Fig. 5. Concentrations (mg m  3) of the individual factors as a function of the wind direction. The traces are colored by time. (a) LV-OOA, (b) LVOOA þ MSA, (c) LRT BBOA, (d) Coffee Roastery, (e) Local BBOA, (f) SV-OOA and (g) HOA. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

the SV-OOA spectrum. A factor that was associated with the biogenic emissions from boreal forests was also observed in a recent study made in Hyyti¨al¨a, Finland (Finessi et al., 2012). Previous studies have shown that boreal forests emit large amounts of biogenic volatile organic compounds (BVOCs) when the biogenic activity is high (Hakola et al., 2003; Kulmala et al., 2004). Tunved et al. (2006) showed that a substantial gas-to-particle formation of BVOC to SOA takes place over the boreal forests in northern Europe. The composition and water-solubility of SOA formed from BVOCs have been intensively studied during the last decade, and the major tracers of BVOCs have been chemically characterized (Yasmeen et al., 2011; Go´mez-Gonza´lez et al., 2012). Some oxidation products of BVOC’s typically observed in ambient aerosols in boreal areas, e.g. pinic and pinonic acid (Cavalli et al., 2006; Laaksonen et al., 2008; Go´mez-Gonza´lez et al., 2012), are known to be water-soluble. Cavalli et al. (2006) found that during nucleation events when the contribution of water-soluble biogenic oxidation products was high, the amount of water-insoluble compounds was also increased, indicating that the oxidation products are to a large extent water-insoluble, at least at the beginning of their atmospheric lifespan. þ In the spectra of LV-OOAþMSA, the fragments of MSA (fragments CHS þ and CH3SO2 , Phinney et al., 2006) can clearly be seen (Fig. 3). On average, the LV-OOAþMSA factor represents 4% of total OA. When the contribution of LV-OOA þMSA factor was high, the wind direction was typically from the Baltic Sea (SW, Fig. 5), indicating that this factor represents

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

69

8

LV-OOA, SV-OOA (µg m-3)

LV-OOA (slope =0.77, Intercept 0.48, R=0.91) SV-OOA (slope =0.01, Intercept 0.43, R=0.07) 6

4

2

0 0

2

4 WSPOM (µg m-3)

6

8

Fig. 6. The correlation of water-soluble particulate organic matter (WSPOM) with low-volatility OA (LV-OOA, r ¼0.91) and semi-volatile OA (SV-OOA, r ¼0.07).

emissions from sea areas. The LV-OOAþ MSA factor did not correlate with WSOC although MSA itself is water-soluble. Bubble bursting, especially when the biogenic activity is high, has been proposed to have an important role in the transfer of organic matter from the sea surface to the atmosphere (Facchini et al., 2008 and references therein; Sciare et al., 2009; Ovadnevaite et al., 2011). The submicron organic aerosol generated by bubble bursting in marine areas has been found to be almost completely water-insoluble (O’Dowd et al., 2004; Facchini et al., 2008; Sciare et al., 2009). The lack of correlation between LV-OOAþMSA and WSOC was not surprising, since the WSOC represents highly-oxidized, aged and long-range transported fraction of aerosols, whereas a big fraction of MSA is likely associated with primary marine aerosols produced by bubble bursting. One PMF factor was observed to represent the emissions originating from a local coffee roastery. When the wind came from the direction of the roastery (southwest 2101), the smell of roasted coffee could clearly be noticed around the measurement station. On average, the contribution of the coffee roastery factor was only 1% of OA, but during the plumes the concentration of this factor was up to 3 mg m  3. Typically factors with less than 5% contribution to the mass cannot be retrieved by PMF (Ulbrich et al., 2009). Although this factor had a very low contribution in terms of mass, it could be retrieved likely due to its distinct mass spectrum. The mass spectra of coffee roastery factor resembled closely that of þ þ caffeine, having all the main peaks formed in the ionization of caffeine (m/z ’s C2H4N þ 42.0344, C4H7 55.0548, C3H3N2 þ þ þ 67.0296, C6H9 81.0704, C6H10 82.0783, C7H9O 109.065: NIST Chemistry WebBook, http://webbook.nist.gov/chemistry), including the molecular ion C8H10N4O2 at m/z 194.0804. High concentrations of the coffee roastery factor were observed mostly in the mornings between about 09:00 and 12:00 (Fig. 3) when the wind direction was from the coffee roastery located 2 km from the SMEAR III station (Fig. 5). Most likely the reasons for the higher concentration in the mornings are due to the diurnal industrial processes. In addition, the detection of this source depends on the wind direction, but there are only limited number of episodes for a more detailed analysis. Emissions from this atypical source have also been indentified during the winter of 2009 at the same site (Carbone et al., in preparation). In terms of the aerosol mass size distribution, the organic matter emitted by the coffee roastery was found to be mainly below 200 nm of vacuum aerodynamic diameter (Carbone et al., in preparation). Two PMF factors were related to biomass burning, the first one for local burning and the other one for long-rangetransported (LRT) material from biomass burning. Both factors had signals at m/z values previously associated with þ þ biomass burning (C2H4O2 at m/z 60.0211 and C3H5O2 at m/z 73.029; Aiken et al., 2009). Biomass burning associated with domestic heating and operation of saunas are quite typical in Finland, especially in winter and spring (Frey et al., 2009, and references therein). The measurement site used in this study was located close to the residential area, so local biomass burning emissions were expected to be seen. LRT biomass burning emission episodes are typically observed in Finland in springtime (Niemi et al., 2004; Saarikoski et al., 2008). Based on satellite observations deploying MODIS sensor on board of NASA EOS Terra satellite and on the backward trajectories of air masses, two long-range-transported biomass burning episodes were observed during the measurement campaign (April 14–15 and April 26–29, 2009). Fig. 5 represents the polar graphs in which the concentrations of local and LRT BBOA are shown as a function of the wind direction. High LRT BBOA emissions were observed when the wind direction was between south and east. Emissions of LRT BBOA seemed to come in distinct plumes when the wind direction was from south, and between the plumes the LRT BBOA concentration was close to zero. Local biomass burning emissions originated from all the directions, and the concentration of the local BBOA seemed to be less variable with only some changes due to wind direction. The local and LRT BBOA concentrations correlated only slightly with each other (r ¼0.47), indicating that they did not represent same emissions. Both the local and LRT BBOA were strongly correlated with the WSOC (Fig. 7, local r ¼0.80, LRT r ¼0.80). This is expected as biomass burning is a major source of WSOC in northern Europe (Saarikoski et al., 2008; Yttri et al., 2009). Both local and LRT BBOA had an intercept in Fig. 7, suggesting that there was always some WSOC that was not related to BBOA. Compared with the

70

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

WSPOM (µg m-3)

8

6

4

2 LRT-BBOA (slope=5.8, intercept=1.5, R=0.76) Local BBOA (slope=3.8, intercept=0.5, R=0.80)

0 0.0

0.5

1.0 1.5 -3 LRT-BBOA, Local BBOA (µg m )

2.0

2.5

Fig. 7. The correlation of water-soluble particulate organic matter with the local (r ¼0.80) and long-range transported BBOA (r ¼0.76).

Slope=0.42, Intercept -0.24 R=0.94

5/7/2009 3.0

5/3/2009

4/29/2009

LRT BBOA (µg m-3)

2.5

4/25/2009 2.0

4/21/2009 4/17/2009

1.5

Slope=0.11, Intercept=-0.01 R=0.69

4/13/2009

1.0 0.5 0.0

0

1

2

3

4

5

6

7

LV-OOA (µg m-3)

Fig. 8. The correlation of the long-range-transported BBOA with the low-volatile OA (LV-OOA). The two biomass burning episodes (April 14–15 and April 26–29, 2009) are clearly separated.

LRT BBOA, the high-resolution mass spectra of the local BBOA showed a stronger fragmentation pattern of aliphatic hydrocarbons indicative of fresher and less oxidized fraction of OA. Fig. 8 shows the correlation between the LRT BBOA and LV-OOA. The two biomass burning episodes (April 14–15 and April 26–29, 2009) had clearly different relations between these two OA types. The wind direction was slightly different during the episodes, and also the backward trajectories indicated that the plumes had slightly different source areas. The first plume was long-range transported from the middle parts of Russia to the east of the station, whereas the second plume originated from southern Russia. In a previous study by Saarnio et al. (2010), it was observed that the emissions from forest fires were clearly affected by the burning type (flaming versus smoldering), burning material (wood type, hay etc.), and the distance from which the emissions were transported before arriving in the measurement site. With the limited data available for this study, the reasons for the observed differences between the two plumes cannot be identified in detail. In general, the sum of BBOAs (local and LRT) had slightly lower contribution in this study (13%) than the annual-average in Helsinki in 2006–2007 (19%; Saarikoski et al., 2008). However, the contribution was similar to that obtained in Helsinki in spring 2006 (12%). Hydrocarbon-like OA factor represents typically primary emissions associated strongly with vehicle exhaust emissions. þ þ þ HOA has a characteristic hydrocarbon pattern (main fragments C2H3O þ , C4H7 and C4H9 ) with little or no signal for CO2 (Canagaratna et al., 2004; Zhang et al., 2005; Ng et al., 2011b). In this study, HOA represented 11% of total organic fraction and it had a clear diurnal cycle with the strongest peak during the morning rush hour (Fig. 4). The diurnal cycle and the correlation with BC (further discussed in Chapter 3.3.2) indicated that the traffic was the main source of HOA in this study. 3.2.1.2. Oxidation state of PMF factors. In order to further characterize the PMF factors, a triangle plot (f44 against f43, where f44 ¼m/z 44/org and f43 ¼m/z 43/org) was constructed (Ng et al., 2010). Ng et al. (2010) has shown that the OOA components typically cluster within a well-defined triangular region in the f44 vs. f43 space. Since photochemical aging leads to an increase in the fraction of m/z 44, the f44 axis can be considered as an indicator of atmospheric aging (Ng et al., 2010). Differences in f43 arise from different sources and chemical formation pathways of OOA components. In the f44 against f43 plot, the factors were clearly separated from each other (Fig. 9). The most oxygenated factors, LRT BBOA, LV-OOA and LV-OOAþMSA, were situated at the top of the triangle. The HOA, SV-OOA and coffee roastery factors

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

71

0.30

0.25

5/6/2009 5/1/2009

0.20

4/26/2009

f44_HR

LV-OOA 0.15

4/21/2009 4/16/2009

LV-OOA + MSA

4/11/2009

LRT BBOA 0.10 Local BBOA

0.05

Coffee Roastery

HOA

0.00 0.00

0.05

0.10

SV-OOA 0.15

0.20

f43_HR Fig. 9. The f44 against f43 plot derived from the spring 2009 campaign data

2.5

H:C-Ratio N:C -Ratio

O:C -Ratio OM:OC -Ratio

2.0

Ratio

1.5

1.0

0.5

0.0 4/11/2009

4/16/2009

4/21/2009

4/26/2009

5/1/2009

5/6/2009

Fig. 10. The time series of the ratios OM:OC, O:C, H:C and N:C calculated from the HR-ToF-AMS data.

were located at the bottom of the triangle, indicating that they represented less oxidized fraction of the OA. The local BBOA was situated in the middle of the triangle, indicating that it was more aged than e.g. HOA but not as processed as the longrange transported fractions. The LRT BBOA had a clearly higher f44 value than local BBOA, which shows that this factor was more aged fraction of biomass burning emissions. The SV-OOA had a very low value of f44 and quite high value of f43, suggesting that it originates from relatively fresh local emissions. Similar to our SV-OOA that we speculate to originate from biogenic emissions, there is some evidence that the compounds with a biogenic influence are often located on the lower right corner of the triangle (Ng et al., 2010). 3.2.2. Elemental analyses Since the speciation of organic aerosol at the molecular level cannot be reached using the AMS measurement technique, elemental ratios can provide insight into the composition and changes due to the aging of OA in the atmosphere (Heald et al., 2010; Ng et al., 2011a). Fig. 10 represents the time series for the ratios OM:OC, O:C, H:C and N:C. The OM:OC ratio varied between 1.4 and 2.1 during the measurement period. Similar values of this ratio have also been found in previous studies (Turpin & Lim, 2001; Aiken et al., 2008; Saarnio et al., 2010). The atomic oxygen to carbon (O:C) ratio, characterizing the oxidation state of the ambient OA, ranged from 0.03 to 0.7 in line with the previous studies (Aiken et al., 2008). The H:C and N:C ratios were in the range 1.2–1.8 and 0.03–0.084, respectively. The elemental ratios were also calculated for the PMF factors (Table 3). The highest values of the OM:OC ratio were observed for the LV-OOA (2.00), LV-OOAþMSA (2.02) and LRT BBOA (1.97). The LV-OOA and LRT BBOA were aged and

72

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

Table 3 The elemental ratios OM:OC, H:C and N:C for each factor.

0.55

Factor

OM:OC

O:C

H:C

N:C

LV-OOA LV-OOA þ MSA Local BBOA Coffee Roastery LRT BBOA SV-OOA HOA

2.03 2.02 1.55 1.44 1.97 1.39 1.20

0.68 0.63 0.32 0.15 0.62 0.20 0.03

1.20 1.47 1.34 1.60 1.33 1.52 1.81

0.018 0.012 0.012 0.084 0.024 0.002 0.004

1.42

0.50

1.75 OM:OC

O:C

1.38 1.36

1.65

21 20

N:C

0.30

1.70

H:C

0.35

22

1.40

0.45 0.40

23x10-3

1.44

1.80

1.34 1.60

19

1.32

18

1.30

0.25 4

8

12 16 Hour of day

20

24

Fig. 11. Diurnal cycles of the ratios O:C, N:C, H:C and OM:OC.

1.46

1.2 BC HOA H:C -ratio, weekdays

-3

1.44

0.8

1.42

0.6

1.40

0.4

1.38

0.2

1.36

H:C

BC, HOA (µg m )

1.0

1.34

0.0 4

8

12

16

20

24

Hour of the day

Fig. 12. Diurnal cycles of the black carbon and HOA concentrations and the H:C ratio.

highly-oxidized long-range transported fractions with good correlations with WSOC, so large values of OM:OC were expected. The local BBOA had a significantly lower OM:OC ratio (1.55) than the long-range transported BBOA (1.97), as can be expected for less-processed local emissions. The lowest values of OM:OC were observed for the HOA (1.2), coffee roastery (1.39) and SV-OOA (1.39). The lowest N:C ratios were observed for HOA and SV-OOA, whereas the coffee roastery factor had the highest N:C ratio. The caffeine (C8H10N4O2) molecule has four nitrogen atoms, which likely explained the high value of N:C for the corresponding factor. The OM:OC and O:C ratios had similar diurnal cycles, with a maximum in the afternoon between 15:00 and 19:00 and a minimum in the morning between 05:00 and 08:00 (Fig. 11). In several earlier studies, the concentration of organic carbon has been found to peak in the daytime due to the SOA formation (e.g. Plaza et al., 2006; Takegawa et al., 2006; Aiken et al., 2008). The value of H:C had a maximum between 06:00 and 09:00 in the morning and a minimum in the afternoon. In previous studies, a peak in black carbon concentrations, caused by a morning rush hour, has been typically observed ¨ between about 06:00 and 09:00 (Jarvi et al., 2008; Timonen et al., 2011). In order to evaluate the effect of traffic to the H:C ratio, the diurnal cycle of the H:C ratio was calculated separately for weekdays, Saturdays and Sundays (Fig. S9). The peak in value of H:C during the morning rush hour was seen only during weekdays (Fig. 12). In terms of black carbon concentrations, the same morning peak was observed for the BC concentrations. Based on the diurnal cycles similar to that

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

73

of BC and the fact that the morning peak was seen during the weekdays only, it seems likely that the peak in the value H:C during the morning was caused by the emissions from traffic. The N:C ratio had a large peak in the morning between 09:00 and 11:00 due to the high contribution from the coffee roastery factor in the morning. 3.2.3. Van Krevelen diagram The Van Krevelen diagram represents the hydrogen to carbon atomic ratio (H:C) as a function of the oxygen to carbon atomic ratio (O:C). The diagram, developed by Van Krevelen (1950), was originally used to illustrate how the elemental composition changes during coal formation. The Van Krevelen diagram has been found to be useful to illustrate the changes due to aging in AMS data (Heald et al., 2010; Ng et al., 2011a). Heald et al. (2010) observed that the bulk composition of the total OA occupies a narrow range in the space of a Van Krevelen diagram characterized by a slope of  1 (i.e. one hydrogen is lost due for each oxygen added due to oxidation). The atmospheric aging observed in the total OA from ambient studies results from a range of processes including volatilization, oxidation, mixing of air masses and condensation of further products. In this study, the organic aerosol data points were scattered along an average Van Krevelen slope of 0.8 (Fig. S10) but a steeper slope was observed for the values of O:C o0.2 and a shallower slope for the values of O:C 40.2. The observed trend of a shallower slope at the higher O:C values is consistent with observations of Ng et al. (2011a) who extracted an average Van Krevelen slope of around  0.5 from a compilation of OOA component observed in multiple ambient environments. The location of the PMF factors in the Van Krevelen diagram is shown in Fig. S4. HOA, representing mostly primary emissions with the highest H:C and lowest O:C ratios, was located on the upper left corner of the Van Krevelen diagram. The factors representing most likely the water-soluble fraction of OA (high O:C and low H:C ratios, LV-OOA and LRT BBOA) were located on the lower right side of the Van Krevelen diagram. Again, the local and LRT BBOA had significantly different elemental compositions indicating they did not represent the same emissions. 3.2.4. Reconstruction of WSOC Only a few studies have been published in which parallel high-time-resolution WSOC and AMS measurements were conducted. Kondo et al. (2007) measured submicron aerosol properties with a quadruple AMS and PILS-TOC during the winter and summer of 2004. They found that the signal at m/z 44 and the derived OOA mass concentrations were highly correlated with WSOC (r2 ¼ 0.78–0.91). The average OOA/WSOC ratio in their study was 3.2470.08 mg/mgC. Xiao et al. (2011) measured the properties of submicron aerosols in the Pearl River Delta, China, in 2006 using a quadruple AMS and PILS-TOC. They observed a good correlation between the WSOC and OOA (r2 ¼0.79) and estimated that approximately 86% of the LV-OOA and 61% of the SV-OOA was water-soluble on the basis of carbon content comparison. The concentration of the WSOC cannot be directly derived from AMS measurements. Simultaneous measurements of the WSOC with the PILS-TOC-IC and organic matter with the AMS provided a possibility to evaluate the water-solubility of different organic factors obtained by the PMF. Based on the recent studies, water-soluble organic carbon consists of secondary and highly-oxidized compounds (Hennigan et al., 2008; Salma & Lang, 2008). In this study, the LV-OOA and LRT BBOA had a good correlation with WSOC. As shown before, both these PMF factors correspond to oxidized and aged fractions of the OA, indicating that they probably represent a similar material as the WSOC. In the concentration time series, the sum of the LV-OOA and LRT BBOA was very close to that of WSOC with a strong correlation (r ¼0.91) between them. However, it seems likely that each factor contains both water-soluble and water-insoluble compounds. Therefore, the Multiple Linear Regression (IGOR 6.11) was used to estimate the coefficients a, b, c, d, e, f and g in the following equation: Y ¼ a  X1 þ b  X2 þc  X3 þ d  X4 þ e  X5þ f  X6 þ g  X7, where Y¼ WSOC, X1¼LV  OOA, X2 ¼(LV OOA þMSA), X3¼LRT BBOA, X4¼coffee roastery, X5¼BBOA, X6¼ SV OOA and X7¼ HOA. In the first solution the multipliers for the factors LV  OOA þMSA and Coffee Roastery were slightly negative, but for the final solution these two multipliers were forced to zero. By forcing multipliers to zero did not change the multipliers of the other factors significantly. Since the TOC-VCPH analyzer measured the concentration of water-soluble carbon and the AMS measured the concentration of organic matter, the results of elemental analysis (number of atoms of carbon present in the organic fraction) were used to convert the organic matter from the AMS to the organic carbon concentration. The following multipliers were obtained for the factors to reconstruct the WSOC: WSOC ¼ 0:88  LVOOA þ0  ðLVOOA þMSAÞ þ 1:63  LRT BBOA þ0  Coffee roastery þ 0:96  localBBOAþ 0:14  SVOOA þ 0:11  HOA: The time series of the reconstructed WSOC and WSOC measured by the PILS-TOC-IC had very similar temporal patterns (Fig. 13). Also correlation between the reconstructed WSOC and the measured WSOC was good (r ¼0.93, Fig. S11). Based on the strong correlation of WSOC with both LV-OOA and BBOA (both LRT BBOA and local BBOA), it seems likely that the BBOA and LV-OOA were mainly water-soluble. Also the high contributions of the LV-OOA and BBOA to the reconstructed water-soluble fraction from the AMS data indicate that the factors LV-OOA and BBOA were highly watersoluble. The most aged fractions, LV-OOA and LRT BBOA, had high values of O:C (0.68 and 0.62), whereas the O:C ratio of the local BBOA was clearly lower (0.32). The values of O:C for the factors representing aged OA were close to that of

74

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

6

Concentration ( µg m-3)

5 4

WSOC reconstructed WSOC LV-OOA 0.88 ± 0.021 LRT BBOA 1.63 ± 0.131 Local BBOA 0.96 ± 0.073 SV-OOA 0.14 ± 0.038 HOA 0.11 ± 0.029

3 2 1 0 4/25/09 4/26/09 4/27/09 4/28/09 4/29/09 4/30/09 5/1/09 5/2/09 5/3/09 5/4/09 5/5/09 5/6/09 5/7/09 5/8/09

Fig. 13. Measured time series of the WSOC concentration (PILS-TOC-IC) and the WSOC time series reconstructed from the PMF factors derived from the AMS organics data.

water-soluble organic matter measured by Sun et al. (2011) (0.56). A similar approach to reconstruct the WSOC has been made for the Pearl River Delta in the southern China (Xiao et al., 2011). Xiao et al. (2011) found that the WSOC concentrations could be predicted with the following equation WSOC ¼0.42  LV  OOA þ0.38  SV OOA þ0.29. They also found a good correlation between the measured and predicted values of WSOC (r2 ¼ 0.79).

4. Conclusions The data collected during the spring 2009 measurement campaign in Helsinki, Finland, represents a unique dataset obtained using novel high-resolution measurement devices in an urban background area. The sources of organic aerosols were studied using Positive Matrix Factorization. During the measurement period, a variety of different sources were observed to contribute to the total aerosol loading. Altogether seven factors were needed to describe the variation in the data set. The results of the PMF were validated by comparing them to parallel measurements of meteorological parameters, trace gases and aerosol components. The factors found by the PMF described well the expected sources or chemical characteristics of ambient aerosols in a boreal forest environment in springtime. The factor LV-OOA was found to represent long-range-transported emissions from southern Europe. The LV-OOA correlated with secondary ions and WSOC, which is typical for long-range-transported emissions. The LV-OOA þMSA had a clear MSA signal and representing aerosols originating from marine areas. Although this factor was found to represent highly-oxidized and aged aerosols, it did not correlate with the WSOC. Using conventional filter sampling, it is possible to measure biomass burning emissions but it is almost impossible to separate local biomass burning emissions from longrange-transported biomass burning emissions. By using PMF, it was possible to separate local and long-range-transported biomass burning. The local BBOA included local biomass burning associated with e.g. domestic heating, whereas the longrange-transported BBOA represented emissions from forest fires in Russia. Both factors had a clear biomass burning signal þ þ at m/z 60 and m/z 73 (C2H4O2 and C3H5O2 ), but the elemental composition and oxidation state of the factors were different as well as their origin. The SV-OOA was probably associated with biogenic emissions originating from the large forest areas in northern Finland. The coffee roastery factor represented emissions from a small local source. The HOA correlated with BC and had a diurnal pattern similar to the H:C ratio, indicating that it was related to local traffic emissions from the Helsinki metropolitan area. Simultaneous measurements of WSOC and organic matter from the AMS provided a possibility to evaluate the solubility of organic fractions recognized by PMF. The LV-OOA and LRT BBOA, typical of highly-oxidized and aged fractions of OA, correlated well with WSOC, thus indicating that they were mostly water-soluble. Also, it was found that it is possible to reconstruct WSOC from the AMS data. The PMF combined with the highly-time-resolved measurements provided an important tool for the source apportionment. By using PMF it was possible to distinguish emissions due to small local sources from emissions that were aged, regional and long-range transported.

Acknowledgments Financial support from the Graduate School in Physics, Chemistry, Biology and Meteorology of Atmospheric Composition and Climate Change (University of Helsinki), European Union (EUCAARI, Contract no.: 036833), Helsinki Energy and Ministry of Transport and Communications Finland (project number 20117) is gratefully acknowledged. The research was also supported by the Academy of Finland Center of Excellence program (project number 1118615) and by the Finnish Funding Agency for Technology and Innovation, Grant number 40209/08 (KASTU).

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

75

Appendix A. Supplementary material Supplementary data associated with this article can be found in the online version at http://dx.doi.org/10.1016/j.jaerosci. 2012.06.005.

References Aiken, A.C., DeCarlo, P.F., & Jimenez, J.L. (2007). Elemental analysis of organic species with electron ionization high-resolution mass spectrometry. Analytical Chemistry, 79, 8350–8358. Aiken, A.C., DeCarlo, P.F., Kroll, J.H., Worsnop, D.R., Huffman, J.A., Docherty, K., Ulbrich, I.M., Mohr, C., Kimmel, J.R., Sueper, D., Zhang, Q., Sun, Y., Trimborn, A., Northway, M., Ziemann, P.J., Canagaratna, M.R., Onasch, T.B., Alfarra, R., Pre´vˆot, A.S.H., Dommen, J., Duplissy, J., Metzger, A., Baltensperger, U., & Jimenez., J.L. (2008). O/C and OM/OC ratios of primary, secondary, and ambient organic aerosols with high resolution time-of-flight aerosol mass spectrometry. Environmental Science & Technology, 42, 4478–4485. Aiken, A.C., Salcedo, D., Cubison, M.J., Huffman, J.A., DeCarlo, P.F., Ulbrich, I.M., Docherty, K.S., Sueper, D., Kimmel, J.R., Worsnop, D.R., Trimborn, A., Northway, M., Stone, E.A., Schauer, J.J., Volkamer, R.M., Fortner, E., de Foy, B., Wang, J., Laskin, A., Shutthanandan, V., Zheng, J., Zhang, R., Gaffney, J., Marley, N.A., Paredes-Miranda, G., Arnott, W.P., Molina, L.T., Sosa, G., & Jimenez, J.L. (2009). Mexico City aerosol analysis during MILAGRO using high resolution aerosol mass spectrometry at the urban supersite (T0)—Part 1: fine particle composition and organic source apportionment. Atmospheric Chemistry and Physics, 9, 6633–6653. Allan, J.D., Williams, P.I., Morgan, W.T., Martin, C.L., Flynn, M.J., Lee, J., Nemitz, E., Phillips, G.J., Gallagher, M.W., & Coe, H. (2010). Contributions from transport, solid fuel burning and cooking to primary organic aerosols in two UK cities. Atmospheric Chemistry and Physics, 10, 647–668. Alfarra, M.R., Coe, H., Allan, J.D., Bower, K.N., Boudries, H., Canagaratna, M.R., Jimenez, J.L., Jayne, J.T., Garforth, A.A., Li, S.M., & Worsnop, D.R. (2004). Characterization of urban and rural organic particulate in the Lower Fraser valley using two Aerodyne aerosol mass spectrometers. Atmospheric Environment, 38, 5745–5758. Bahreini, R., Keywood, M.D., Ng, N.L., Varutbangkul, V., Gao, S., Flagan, R.C., Seinfeld, J.H., Worsnop, D.R., & Jimenez, J.L. (2005). Measurements of secondary organic aerosol (SOA) from oxidation of cycloalkenes, terpenes, and M-xylene using an Aerodyne aerosol mass spectrometer. Environmental Science & Technology, 39, 5674–5688. Canagaratna, M.R., Jayne, J.T., Ghertner, D.A., Herndon, S., Shi, Q., Jimenez, J.L., Silva, P.J., Williams, P., Lanni, T., Drewnick, F., Demerjian, K.L., Kolb, C.E., & Worsnop, D.R. (2004). Chase studies of particulate emissions from in-use New York City vehicles. Aerosol Science and Technology, 38, 555–573. Canagaratna, M.R., Jayne, J.T., Jimenez, J.L., Allan, J.D., Alfarra, M.R., Zhang, Q., Onasch, T.B., Drewnick, F., Coe, H., Middlebrook, A., Delia, A., Williams, L.R., Trimborn, A.M., Northway, M.J., DeCarlo, P.F., Kolb, C.E., Davidovits, P., & Worsnop, D.R. (2007). Chemical and microphysical characterization of ambient aerosols with the Aerodyne aerosol mass spectrometer. Mass Spectrometry Reviews, 26, 185–222. Carbone et al., in preparation. Cavalli, F., Facchini, M.C., Decesari, S., Emblico, L., Mircea, M., Jensen, N.R., & Fuzzi, S. (2006). Size-segregated aerosol chemical composition at a boreal site in southern Finland, during the QUEST project. Atmospheric Chemistry and Physics, 6, 993–1002. Cross, E.S., Onasch, T.B., Canagaratna, M., Jayne, J.T., Kimmel, J., Yu, X.-Y., Alexander, M.L., Worsnop, D.R., & Davidovits, P. (2009). Single particle characterization using a light scattering module coupled to a time-of-flight aerosol mass spectrometer. Atmospheric Chemistry and Physics, 9, 7769–7793, http://dx.doi.org/10.5194/acp-9-7769-2009. DeCarlo, P.F., Kimmel, J.R., Trimborn, A., Northway, M., Jayne, J.T., Aiken, A.C., Gonin, M., Fuhrer, K., Horvath, T., Docherty, K.S., Worsnop, D.R., & Jimenez, J.L. (2006). Field-deployable, high-Resolution, time-of-flight aerosol mass spectrometer. Analytical Chemistry, 78, 8281–8289. Decesari, S., Fuzzi, S., Facchini, M.C., Mircea, M., Emblico, L., Cavalli, F., Maenhaut, W., Chi, X., Schkolnik, G., Falkovich, A., Rudich, Y., Claeys, M., Pashynska, V., Vas, G., Kourtchev, I., Vermeylen, R., Hoffer, A., Andreae, M.O., Tagliavini, E., Moretti, F., & Artaxo, P. (2006). Characterization of the organic composition of aerosols from Rondoˆ nia, Brazil, during the LBA-SMOCC 2002 experiment and its representation through model compounds. Atmospheric Chemistry and Physics, 6, 375–402. Decesari, S., Mircea, M., Cavalli, F., Fuzzi, S., Moretti, F., Tagliavini, E., & Facchini, M.C. (2007). Source attribution of water-soluble organic aerosol by nuclear magnetic resonance spectroscopy. Environmental Science and Technology, 41, 2479–2484. Duplissy, J., DeCarlo, P.F., Dommen, J., Alfarra, M.R., Metzger, A., Barmpadimos, I., Pre´vˆot, A.S.H., Weingartner, E., Tritscher, T., Gysel, M., Aiken, A.C., Jimenez, J.L., Canagaratna, M.R., Worsnop, D.R., Collins, D.R., Tomlinson, J., & Baltensperger, U. (2011). Relating hygroscopicity and composition of organic aerosol particulate matter. Atmospheric Chemistry and Physics, 11, 1155–1165. Facchini, M.C., Rinaldi, M., Decesari, S., Carbone, C., Finessi, E., Mircea, M., Fuzzi, S., Ceburnis, D., Flanagan, R., Nilsson, E.D., de Leeuw, G., Martino, M., Woeltjen, J., & O’Dowd, C.D. (2008). Primary submicron marine aerosol dominated by insoluble organic colloids and aggregates. Geophysical Research Letters, 35, L17814, http://dx.doi.org/10.1029/2008GL034210. Finessi, E., Decesari, S., Paglione, M., Giulianelli, L., Carbone, C., Gilardoni, S., Fuzzi, S., Saarikoski, S., Raatikainen, T., Hillamo, R., Allan, J., Mentel, Th.F., ¨ a, ¨ T., Kulmala, M., Worsnop, D.R., & Facchini, M.C. (2012). Determination of the biogenic secondary organic aerosol Tiitta, P., Laaksonen, A., Petaj fraction in the boreal forest by NMR spectroscopy. Atmospheric Chemistry and Physics, 12, 941–959, http://dx.doi.org/10.5194/acp-12-941-2012. ¨ ¨ Frey, A.K., Tissari, J., Saarnio, K.M., Timonen, H.J., Tolonen-Kivimaki, O., Aurela, M.A., Saarikoski, S.K., Makkonen, U., Hytonen, K., Jokiniemi, J., Salonen, R.O, & Hillamo, R.E.J. (2009). Chemical composition and mass size distribution of fine particulate matter emitted by a small masonry heater. Boreal Environment Research, 14, 255–271. Go´mez-Gonza´lez, Y., Wang, W., Vermeylen, R., Chi, X., Neirynck, J., Janssens, I.A., Maenhaut, W., & Claeys, M. (2012). Chemical characterization of atmospheric aerosols during a 2007 summer field campaign at Brasschaat, Belgium: sources and source processes of biogenic secondary organic aerosol. Atmospheric Chemistry and Physics, 12, 125–138. Gross, J.H. (2004). Mass Spectrometry—A Textbook (1st Ed.). Springer-Verlag: Heidelberg. Grover, B.D., Eatough, N.L., Eatough, D.J., Chow, J.C., Watson, J.G., Ambs, J.L., Meyer, M.B., Hopke, P.K., Al-Horr, R., Later, D.W., & Wilson, W.E. (2006). Measurement of both nonvolatile and semi-volatile fractions of fine particulate matter in Fresno, CA. Aerosol Science and Technology, 40, 811–826. Hakola, H., Tarvainen, V., Laurila, T., Hiltunen, V., Hellen, H., & Keronen, P. (2003). Seasonal variation of VOC concentrations above a boreal coniferous forest. Atmospheric Environment, 37, 1623–1634. Hallquist, M., Wenger, J.C., Baltensperger, U., Rudich, Y., Simpson, D., Claeys, M., Dommen, J., Donahue, N.M., George, C., Goldstein, A.H., Hamilton, J.F., Herrmann, H., Hoffmann, T., Iinuma, Y., Jang, M., Jenkin, M.E., Jimenez, J.L., Kiendler-Scharr, A., Maenhaut, W., McFiggans, G., Mentel, Th.F., Monod, A., Pre´vˆot, A.S.H., Seinfeld, J.H., Surratt, J.D., Szmigielski, R., & Wildt, J. (2009). The formation, properties and impact of secondary organic aerosol: current and emerging issues. Atmospheric Chemistry and Physics, 9, 5155–5236. Heald, C.L., Kroll, J.H., Jimenez, J.L., Docherty, K.S., DeCarlo, P.F., Aiken, A.C., Chen, Q., Martin, S.T., Farmer, D.K., Artaxo, P., & Weinheimer, A.J. (2010). A simplified description of organic aerosol composition and implications for atmospheric aging. Geophysical Research Letters, 37(L08803), 2010, http://dx.doi.org/ 10.1029/2010GL042737. Heald, C.L., Coe, H., Jimenez, J.L., Weber, R.J., Bahreini, R., Middlebrook, A.M., Russell, L.M., Jolleys, M., Fu, T.-M., Allan, J.D., Bower, K.N., Capes, G., Crosier, J., Morgan, W.T., Robinson, N.H., Williams, P.I., Cubison, M.J., DeCarlo, P.F., & Dunlea, E.J. (2011). Exploring the vertical profile of atmospheric organic aerosol: comparing 17 aircraft field campaigns with a global model. Atmospheric Chemistry and Physics, 11, 12673–12696.

76

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

Hennigan, C.J., Bergin, M.H., Dibb, J.E., & Weber, R.J. (2008). Enhanced secondary organic aerosol formation due to water uptake by fine particles. Geophysical Research Letters, 35, L18801, http://dx.doi.org/10.1029/2008GL035046. Hennigan, C.J., Bergin, M.H., Russell, A.G., Nenes, A., & Weber, R.J. (2009). Gas/particle partitioning of water-soluble organic aerosol in Atlanta. Atmospheric Chemistry and Physics, 9, 3613–3628, http://dx.doi.org/10.5194/acp-9-3613-2009. Hennigan, C.J., Miracolo, M.A., Engelhart, G.J., May, A.A., Presto, A.A., Lee, T., Sullivan, A.P., McMeeking, G.R., Coe, H., Wold, C.E., Hao, W.-M., Gilman, J.B., Kuster, W.C., de Gouw, J., Schichtel, B.A., Collett, J.L., Jr., Kreidenweis, S.M., & Robinson, A.L. (2011). Chemical and physical transformations of organic aerosol from the photo-oxidation of open biomass burning emissions in an environmental chamber. Atmospheric Chemistry and Physics, 11, 7669–7686. Huffman, J.A., Jayne, J.T., Drewnick, F., Aiken, A.C., Onasch, T., Worsnop, D.R., & Jimenez, J.L. (2005). Design, modeling, optimization, and experimental tests of a particle beam width probe for the aerodyne aerosol mass spectrometer. Aerosol Science and Technology, 39, 1143–1163. Jacobson, M.C., Hansson, H.-C., Noone, K.L., & Charlson, R.J. (2000). Organic atmospheric aerosols: review and state of the science. Reviews of Geophysics, 38, 267–294. Jayne, J.T., Leard, D.C., Zhang, X., Davidovits, P., Smith, K.A., Kolb, C.E., & Worsnop, D.R. (2000). Development of an aerosol mass spectrometer for size and composition analysis of submicron particles. Aerosol Science and Technology, 33, 49–70. Jimenez, J.L., Canagaratna, M.R., Donahue, N.M., Prevot, A.S.H., Zhang, Q., Kroll, J.H., DeCarlo, P.F., Allan, J.D., Coe, H., Ng, N.L., Aiken, A.C., Docherty, K.S., Ulbrich, I.M., Grieshop, A.P., Robinson, A.L., Duplissy, J., Smith, J.D., Wilson, K.R., Lanz, V.A., Hueglin, C., Sun, Y.L., Tian, J., Laaksonen, A., Raatikainen, T., Rautiainen, J., Vaattovaara, P., Ehn, M., Kulmala, M., Tomlinson, J.M., Collins, D.R., Cubison, M.J., Dunlea, E.J., Huffman, J.A., Onasch, T.B., Alfarra, M.R., Williams, P.I., Bower, K., Kondo, Y., Schneider, J., Drewnick, F., Borrmann, S., Weimer, S., Demerjian, K., Salcedo, D., Cottrell, L., Griffin, R., Takami, A., Miyoshi, T., Hatakeyama, S., Shimono, A., Sun, J.Y., Zhang, Y.M., Dzepina, K., Kimmel, J.R., Sueper, D., Jayne, J.T., Herndon, S.C., Trimborn, A.M., Williams, L.R., Wood, D.R., Middlebrook, A.M., Kolb, C.E., Baltensperger, U., & Worsnop, D.R. (2009). Evolution of Organic Aerosols in the Atmosphere. Science, 326, 1525–1529. Johnson, R.L., & Huntzicker, J.J. (1979). Analysis of volatilizable and elemental carbon in ambient aerosols. In: T. Novakov (Ed.), Proceedings, Carbonaceous Particles in the Atmosphere. National Science Foundation & Lawrence Berkeley Laboratory: Berkeley, CA, pp. 10–13. ¨ ¨ ¨ T., Pakkanen, T., & Kulmala, M. (2008). Temporal variations in black carbon Jarvi, L., Junninen, H., Karppinen, A., Hillamo, R., Virkkula, A., Makel a, concentrations with different time scales in Helsinki during 1996–2005. Atmospheric Chemistry and Physics, 8, 1017–1027. ¨ ¨ ¨ T., Keronen, P., Siivola, E., Vesala, T., & Kulmala, M. (2009). The urban Jarvi, L., Hannuniemi, H., Hussein, T., Junninen, H., Aalto, P.P., Hillamo, R., Makel a, measurement station SMEAR III: Continuous monitoring of air pollution and surface–atmosphere interactions in Helsinki, Finland. Boreal Environment Research, 14(suppl. A), 86–109. Kanakidou, M., Seinfeld, J.H., Pandis, S.N., Barnes, I., Dentener, F.J., Facchini, M.C., Van Dingenen, R., Ervens, B., Nenes, A., Nielsen, C.J., Swietlicki, E., Putaud, J.P., Balkanski, Y., Fuzzi, S., Horth, J., Moortgat, G.K., Winterhalter, R., Myhre, C.E.L., Tsigaridis, K., Vignati, E., Stephanou, E.G., & Wilson, J. (2005). Organic aerosol and global climate modelling: a review. Atmospheric Chemistry and Physics, 5, 1053–1123. Kiendler-Scharr, A., Zhang, Q., Hohaus, T., Kleist, E., Mensah, A., Mentel, T.F., Spindler, C., Uerlings, R., Tillmann, R., & Wildt, J. (2009). Aerosol mass spectrometric features of biogenic SOA: Observations from a plant chamber and in rural atmospheric environments. Environmental Science & Technology, 43, 8166–8172. Kondo, Y., Miyazaki, Y., Takegawa, N., Miyakawa, T., Weber, R.J., Jimenez, J.L., Zhang, Q., & Worsnop, D.R. (2007). Oxygenated and water-soluble organic aerosols in Tokyo. Journal of Geophysical Research, 112, D011203, http://dx.doi.org/10.1029/2006JD007056. ¨ Kulmala, M., Suni, T., Lehtinen, K.E.J., Dal Maso, M., Boy, M., Reissell, A., Rannik, U., Aalto, P., Keronen, P., Hakola, H., Back, J., Hoffmann, T., Vesala, T., & Hari, P. (2004). A new feedback mechanism linking forests, aerosols and climate. Atmospheric Chemistry and Physics, 4, 557–562. ¨ a, ¨ T., Laaksonen, A., Kulmala, M., O’Dowd, C.D., Joutsensaari, J., Vaattovaara, P., Mikkonen, S., Lehtinen, K.E.J., Sogacheva, L., Dal Maso, M., Aalto, P., Petaj Sogachev, A., Yoon, Y.J., Lihavainen, H., Nilsson, D., Facchini, M.C., Cavalli, F., Fuzzi, S., Hoffmann, T., Arnold, F., Hanke, M., Sellegri, K., Umann, B., ¨ ainen, ¨ Junkermann, W., Coe, H., Allan, J.D., Alfarra, M.R., Worsnop, D.R., Riekkola, M.-L., Hyotyl T., & Viisanen, Y. (2008). The role of VOC oxidation products in continental new particle formation. Atmospheric Chemistry and Physics, 8, 2657–2665. Lanz, V.A., Alfarra, M.R., Baltensberger, U., Buchmann, B., Hueglin, C., & Pre´vˆot, A.S.H. (2007). Source apportionment of submicron organic aerosols at an urban site by factor analytical modelling of aerosol mass spectra. Atmospheric Chemistry and Physics, 7, 1503–1522. Liu, D., Allan, J., Corris, B., Flynn, M., Andrews, E., Ogren, J., Beswick, K., Bower, K., Burgess, R., Choularton, T., Dorsey, J., Morgan, W., Williams, P.I., & Coe, H. (2011). Carbonaceous aerosols contributed by traffic and solid fuel burning at a polluted rural site in Northwestern England. Atmospheric Chemistry and Physics, 11, 1603–1619. Matthew, B.M., Middlebrook, A.M., & Onasch, T.B. (2008). Collection efficiencies in an Aerodyne aerosol mass spectrometer as a Function of particle phase for laboratory generated aerosols. Aerosol Science and Technology, 42, 884–898. Morgan, W.T., Allan, J.D., Bower, K.N., Esselborn, M., Harris, B., Henzing, J.S., Highwood, E.J., Kiendler-Scharr, A., McMeeking, G.R., Mensah, A.A., Northway, M.J., Osborne, S., Williams, P.I., Krejci, R., & Coe, H. (2010a). Enhancement of the aerosol direct radiative effect by semi-volatile aerosol components: airborne measurements in North-Western Europe. Atmospheric Chemistry and Physics, 10, 8151–8171. Morgan, W.T., Allan, J.D., Bower, K.N., Highwood, E.J., Liu, D., McMeeking, G.R., Northway, M.J., Williams, P.I., Krejci, R., & Coe, H. (2010b). Airborne measurements of the spatial distribution of aerosol chemical composition across Europe and evolution of the organic fraction. Atmospheric Chemistry and Physics, 10, 4065–4083. Ng, N.L., Canagaratna, M.R., Zhang, Q., Jimenez, J.L., Tian, J., Ulbrich, I.M., Kroll, J.H., Docherty, K.S., Chhabra, P.S., Bahreini, R., Murphy, S.M., Seinfeld, J.H., Hildebrandt, L., Donahue, N.M., DeCarlo, P.F., Lanz, V.A., Pre´vˆot, A.S.H., Dinar, E., Rudich, Y., & Worsnop, D.R. (2010). Organic aerosol components observed in northern hemispheric datasets measured with aerosol mass spectrometry. Atmospheric Chemistry and Physics, 10, 4625–4641. Ng, N.L., Canagaratna, M.R., Jimenez, J.L., Chhabra, P.S., Seinfeld, J.H., & Worsnop, D.R. (2011a). Changes in organic aerosol composition with aging inferred from aerosol mass spectra. Atmospheric Chemistry and Physics, 11, 6465–6474. Ng, N.L., Canagaratna, M.R., Jimenez, J.L., Zhang, Q., Ulbrich, I.M., & Worsnop, D.R. (2011b). Real-Time methods for estimating organic component mass concentrations from aerosol mass spectrometer data. Environmental Science & Technology, 45, 910–916. ¨ H., Kulmala, M., Koskentalo, T., Sillanpa¨ a, ¨ M., & Rantamaki, ¨ Niemi, J.V., Tervahattu, H., Vehkamaki, M. (2004). Characterization and source identification of a fine particle episode in Finland. Atmospheric Environment, 38, 5003–5012. O’Dowd, C.D., Facchini, M.C., Cavalli, F., Ceburnis, D., Mircea, M., Decesari, S., Fuzzi, S., Yoon, Y.J., & Putaud, J.-P. (2004). Biogenically driven organic contribution to marine aerosol. Nature, 431, 676–680. Ovadnevaite, J., O’Dowd, C., Dall’Osto, M., Ceburnis, D., Worsnop, D.R., & Berresheim, H. (2011). Detecting high contributions of primary organic matter to marine aerosol: A case study. Geophysical Research Letters, 38, L02807, http://dx.doi.org/10.1029/2010GL046083. Paatero, P., & Tapper, U. (1994). Positive matrix factorization: a nonnegative factor model with optimal utilization of error estimates of data values. Environmetrics, 5, 111–126. Paatero, P. (1997). Least squares formulation of robust non-negative factor analysis. Chemometrics and Intelligent Laboratory, 37, 23–35. Phinney, L., Leaitch, W.R., Lohmann, U., Boudries, H., Worsnop, D.R., Jayne, J.T., Toom-Sauntry, D., Wadleigh, M., Sharma, S., & Shantz, N. (2006). Characterization of the aerosol over the sub-arctic north east Pacific Ocean. Deep-Sea Research Part II, 53, 2410–2433. Plaza, J., Gomez-Moreno, F.J., Nunez, L., Pujadas, M., & Artinano, B. (2006). Estimation of secondary organic aerosol formation from semicontinuous OC–EC measurements in a Madrid suburban area. Atmospheric Environment, 40, 1134–1147. Raatikainen, T., Vaattovaara, P., Tiitta, P., Miettinen, P., Rautiainen, J., Ehn, M., Kulmala, M., Laaksonen, A., & Worsnop, D.R. (2010). Physicochemical properties and origin of organic groups detected in boreal forest using an aerosol mass spectrometer. Atmospheric Chemistry and Physics, 10, 2063–2077. Robinson, A.L., Donahue, N.M., Shrivastava, M.K., Weitkamp, E.A., Sage, A.M., Grieshop, A.P., Lane, T.E., Pierce, J.R., & Pandis, S.N. (2007). Rethinking Organic Aerosols: Semivolatile Emissions and Photochemical Aging. Science, 315, 1259–1262.

H. Timonen et al. / Journal of Aerosol Science 56 (2013) 61–77

77

Robinson, N.H., Hamilton, J.F., Allan, J.D., Langford, B., Oram, D.E., Chen, Q., Docherty, K., Farmer, D.K., Jimenez, J.L., Ward, M.W., Hewitt, C.N., Barley, M.H., Jenkin, M.E., Rickard, A.R., Martin, S.T., McFiggans, G., & Coe, H. (2011). Evidence for a significant proportion of secondary organic aerosol from isoprene above a maritime tropical forest. Atmospheric Chemistry and Physics, 11, 1039–1050. ¨ M., Sofiev, M., Timonen, H., Saarnio, K., Teinila, ¨ K., Karppinen, A., Kukkonen, J., & Hillamo, R. (2007). Chemical composition of Saarikoski, S., Sillanpa¨ a, aerosols during a major biomass burning episode over northern Europe in spring 2006: experimental and modeling assessments. Atmospheric Environment, 41, 3577–3589. ¨ Saarikoski, S., Timonen, H., Saarnio, K., Aurela, M., Jarvi, L., Keronen, P., Kerminen, V.-M., & Hillamo, R. (2008). Sources of organic carbon in fine particulate matter in northern European urban air. Atmospheric Chemistry and Physics, 8, 6281–6295. ¨ K., Makel ¨ ¨ T., Sofiev, M., Koskinen, J., Aalto, P.P., Kulmala, M., Kukkonen, J., & Hillamo, R. (2010). Saarnio, K., Aurela, M., Timonen, H., Saarikoski, S., Teinila, a, Chemical composition of fine particles in fresh smoke plumes from boreal wild-land fires in Europe. Science of the Total Environment, 405, 2527–2542. Salma, I., & Lang, G.G. (2008). How many carboxyl groups does an average molecule of humic-like substances contain?. Atmospheric Chemistry and Physics, 8, 5997–6002. Saxena, P., & Hildemann, L.M. (1996). Water-soluble organics in atmospheric particles: a critical review of the literature and application of thermodynamics to identify candidate compounds. Journal of Atmospheric Chemistry, 24, 57–109. Sciare, J., Favez, O., Sarda-Este ve, R., Oikonomou, K., Cachier, H., & Kazan, V. (2009). Long-term observations of carbonaceous aerosols in the Austral Ocean atmosphere: Evidence of a biogenic marine organic source. Journal of Geophysical Research, 114, D15302, http://dx.doi.org/10.1029/2009JD011998. Simoneit, B.R.T., Schauer, J.J., Nolte, C.G., Oros, D.R., Elias, V.O., Fraser, M.P., Rogge, W.F., & Cass, G.R. (1999). Levoglucosan, a tracer for cellulose in biomass burning and atmospheric particles. Atmospheric Environment, 33, 173–182. Stohl, A., & Wotawa, G. (1995). A method for computing single trajectories representing boundary layer transport. Atmospheric Environment, 29, 3235–3239. Sueper, D., (2011). ToF-AMS High resolution analysis software. Available online at: /http://cires.colorado.edu/jimenez-group/ToFAMSResources/ToFSoftware/ index.htmlS. Sun, Y., Zhang, Q., Zheng, M., Ding, X., Edgerton, E.S., & Wan, X. (2011). Characterization and source apportionment of water-soluble organic matter in atmospheric fine particles (PM2.5) with high-resolution aerosol mass spectrometry and GC/MS. Environmental Science & Technology, 45, 4854–4861. Takegawa, N., Miyakawa, T., Kondo, Y., Jimenez, J.L., Zhang, Q., Worsnop, D.R., & Fukuda, M. (2006). Seasonal and diurnal variations of submicron organic aerosols in Tokyo observed using the Aerodyne aerosol mass spectrometer (AMS). Journal of Geophysical Research, 111, D11206, http://dx.doi.org/ 10.1029/2005JD006515. ¨ ¨ T., Kulmala, M., Kerminen, V.-M., Worsnop, D.R., & Hillamo, R. (2010). High timeTimonen, H., Aurela, M., Carbone, S., Saarnio, K., Saarikoski, S., Makel a, resolution chemical characterization of the water-soluble fraction of ambient aerosols with PILS-TOC-IC and AMS. Atmospheric Measurement Techniques, 3, 1063–1074. ¨ K., Kulmala, M., & Hillamo, R. (2011). Monitoring of inorganic ions, carbonaceous Timonen, H., Aurela, M., Saarnio, K., Frey, A., Saarikoski, S., Teinila, matter and mass with online and offline methods. Atmospheric Measurement Techniques Discussions, 4, 6577–6614. ¨ Tunved, P., Hansson, H.-C., Kerminen, V.-M., Strom, J., Dal Maso, M., Lihavainen, H., Viisanen, Y., Aalto, P.P., Komppula, M., & Kulmala, M. (2006). High natural aerosol loading over boreal forests. Science, 312, 261–263. Turpin, B.J., Cary, R.A., & Huntzicker, J.J. (1990). An insitu, time-resolved analyzer for aerosol organic and elemental carbon. Aerosol Science and Technology, 12, 161–171. Turpin, B.J., & Lim, H.-J. (2001). Species contributions to PM2.5 mass concentrations: revisiting common assumptions for estimating organic mass. Aerosol Science and Technology, 35, 602–610. Ulbrich, I.M., Canagaratna, M.R., Zhang, Q., Worsnop, D.R., & Jimenez, J.L. (2009). Interpretation of organics components from positive matrix factorization of aerosol mass spectrometric data. Atmospheric Chemistry and Physics, 9, 2891–2918. Van Krevelen, D.W. (1950). Graphical-statistical method for the study of structure and reaction processes of coal. Fuel, 24, 269–284. ¨ M., Laakia, J., Kerminen, V.-M., Hillamo, R., Aarnio, P., & Koskentalo, T. (2002). Organic and black carbon in PM2.5 and PM10: 1 year Viidanoja, J., Sillanpa¨ a, data from urban site in Helsinki, Finland. Atmospheric Environment, 26, 3183–3193. Wilson, W.E., Grover, B.D., Long, R.W., Eatough, N.L., & Eatough, D.J. (2006). The measurement of fine particulate semivolatile material in urban aerosols. Journal of the Air & Waste Management Association, 56, 384–397. Xiao, R., Takegawa, N., Zheng, M., Kondo, Y., Miyazaki, Y., Miyakawa, T., Hu, M., Shao, M., Zeng, L., Gong, Y., Lu, K., Deng, Z., Zhao, Y., & Zhang, Y.H. (2011). Characterization and source apportionment of submicron aerosol with aerosol mass spectrometer during the PRIDE-PRD 2006 campaign. Atmospheric Chemistry and Physics, 11, 6911–6929. Yasmeen, F., Szmigielski, R.R., Vermeylen, R., Go´mez-Gonza´lez, Y., Surratt, J.D., Chan, A.W.H., Seinfeld, J.H., Maenhaut, W., & Claeys, M. (2011). Mass spectrometric characterization of isomeric terpenoic acids from the oxidation of a-pinene, b-pinene, d-limonene, and D3-carene in fine forest aerosol. Journal of Mass Spectrometry, 46, 425–442. Yttri, K.E., Dye, C., Braathen, O.A., Simpson, D., & Steinnes, E. (2009). Carbonaceous aerosols in Norwegian urban areas. Atmospheric Chemistry and Physics, 9, 2007–2020. Zhang, Q., Alfarra, M.R., Worsnop, D.R., Allan, J.D., Coe, H., Canagaratna, M.R., & Jimenez, J.L. (2005). Deconvolution and quantification of hydrocarbon-like and oxygenated organic aerosols based on aerosol mass spectrometry. Environmental Science & Technology, 39, 4938–4952. Zhang, Q., Jimenez, J.L., Canagaratna, M.R., Allan, J.D., Coe, H., Ulbrich, I., Alfarra, M.R., Takami, A., Middlebrook, A.M., Sun, Y.L., Dzepina, K., Dunlea, E., Docherty, K., DeCarlo, P.F., Salcedo, D., Onasch, T., Jayne, J.T., Miyoshi, T., Shimono, A., Hatakeyama, S., Takegawa, N., Kondo, Y., Schneider, J., Drewnick, F., Weimer, S., Demerjian, K., Williams, P., Bower, K., Bahreini, R., Cottrell, L., Griffin, R.J., Rautiainen, J., & Worsnop, D.R. (2007). Ubiquity and dominance of oxygenated species in organic aerosols in anthropogenically-influenced Northern Hemisphere midlatitudes. Geophysical Research Letters, 34, L13801, http://dx.doi.org/10.1029/2007GL029979. Zhang, Q., Jimenez, J.L, Canagaratna, M.R., Ulbrich, I.M., Ng, N.L., Worsnop, D.R., & Sun, Y. (2011). Understanding atmospheric organic aerosols via factor analysis of aerosol mass spectrometry: a review. Analytical and Bioanalytical Chemistry, 401, 3045–3067, http://dx.doi.org/10.1007/s00216-011-5355-y.