Hemisphere Using Offline Aerosol Mass Spectrometry I

Size-segregated long-term chemical analysis and source apportionment of the organic aerosol in the Northern hemisphere using offline aerosol mass spectrometry I. EI Haddad, K. R. Daellenbach, C. Bozzetti, A. Vlachou, and A.S.H. Prevot Main highlights

> 2500 samples analyzed at > Kohtla-Järve Tartu Tallin 25 sites Rugsteliskes Frauenfeld Long term, specially resolved datasets at urban and rural Bern Vilnius Preila sites in Switzerland, France, Basel Italy, Lithuania, Estonia and Lebanon

St. Gallen Size resolved data at rural Paris locations in central Europe Vaduz Payerne and Amazon San Vittore PM10 Investigation of haze events in PM2.5 Magadino Megacities in China and PM1 Europe Zurich Marseille Milano Beirut

Beijing South America: Xi’an Shanghai Brazil: Amazon Guangzhou PM10 PM2.5

Daellenbach et al., 2016 and in prep; Huang et al., 2014 Asia: Bozzetti et al., 2016a, 2016b and 2016c.; Vlachou et al. in prep China: 4 sites, heavy pollution episode El Haddad et al., in prep Main highlights

Payerne Basel Bern Frauenfeld Traffic 2013 Rugsteliskes Cooking Biomass burning Vilnius PM10 (S/W) Primary biological OA Preila Local OA (waste treat) PM2.5 (S/W) Summer OOA Winter OOA St. Gallen PM1 (S/W) Sulfur-containing OA Vaduz Coal burning Zurich Secondary organic rich San Vittore Secondary inorganic rich 2011-12

2013 Magadino Marseille Milano Beijing

Xi’an

Shanghai

Asia: Guangzhou Daellenbach et al., 2016 and in prep; Huang et al., 2014 Bozzetti et al., 2016a, 2016b and 2016c.; Vlachou et al. in prep China: 4 sites, heavy pollution episode El Haddad et al., in prep Main highlights Case of Switzerland (similar to other European sites)

1.0 50 Daellenbach

0.8 40 

0.6 30 g/m bas

0.4 20 3 rel. contr. rel.

0.2 10 et al., in prep 0.0 0

Jul Jul Jul Jul JanFebMar AprMaiJun AugSepOctNovDec JanFebMar AprMaiJun AugSepOctNovDec JanFebMar AprMaiJun AugSepOctNovDec JanFebMar AprMaiJun AugSepOctNovDec Basel Bern Zurich Frauenfeld 1.0 50

0.8 40 

0.6 30 g/m pay

0.4 20 3 rel. contr. rel. 0.2 10 0.0 0

Jul Jul JanFebMar AprMaiJun AugSepOctNovDec JanFebMar AprMaiJun AugSepOctNovDec Payerne St. Gallen 1.0 50 HOA 0.8 COA 40 SC-OA  0.6 30 g/m WOOA 0.4 SOOA 20 3 rel. contr. rel. BBOA 0.2 10 OAexpl 0.0 0

Jul Jul Jul JanFebMar AprMaiJun AugSepOctNovDec JanFebMar AprMaiJun AugSepOctNovDec JanFebMar AprMaiJun AugSepOctNovDec Magadino San Vittore Vaduz Main highlights

2013 Magadino Marseille Milano  BBOA is the main source in Europe during winter (only PM1) Main highlights

2013 Magadino Marseille Milano  BBOA is the main source in Europe during winter (only PM1) Leaitch (2011)  Summer OOA, main source during Canadian Forest summer, is from terpene emissions Fit Switzerland Main highlights

2013 Magadino Marseille Milano

160

 BBOA is the main source in Europe ] -3 140 during winter (only PM1)  Summer OOA, main source during 120 summer, is from terpene emissions 100  Winter OOA dominated by biomass 80 60 burning SOA (in PM2.5) 40

20 Nitrocatecholsm [ng 0

01.11.2011 01.03.2012 01.07.2012 Main highlights

2013 Marseille Milano

 BBOA is the main source in Europe 0.10 Cx 0.08 Milled oak leaves 0.06 PMF-WSPBOA CH during winter (only PM1) 0.04 0.02 CHO1  Summer OOA, main source during 0.00 CHO>1

50 HO summer, is from terpene emissions -3 40 Amylopectin from maze 30 CHN

x10 20 CS  Winter OOA dominated by biomass 10 Air 0.100 Milled oak leaves Cx

burning SOA (in PM2.5) -3 0.0850 Milled oak leaves 0.0640 Starch from weat CH 0.0430 x10 CHO  PBOA, identified in PM , is 0.0220 1 10 0.0010 CHO 0 >1

dominated sugar like structures,

Relative signal Relative 50-3 HO normalized intensity normalized 60x10-3 40 Amylopectin from maze D-mannitol CHN from plant debris (~SOOA) 3040 Mannitol x10 20 CS 1020 0 Air

-3 50 40 Starch from weat 30

x10 20 40 60 80 100 20 10 m/z 0 -3 normalized intensity normalized 60x10 D-mannitol 40 20 0 20 40 60 80 100 m/z Main highlights

2013 Magadino Marseille Milano Beijing  BBOA is the main source in Europe during winter (only PM1) Xi’an  Summer OOA, main source during summer, is from terpene emissions Shanghai  Winter OOA dominated by biomass burning SOA (in PM2.5)  PBOA, identified in PM10, is dominated sugar like structures, Guangzhou from plant debris (~SOOA)  SOA formation is important during China Chinese haze events Better characterizing the AMS using the offline AMS technique

 Identifying particle-vaporizer interactions (e.g. Pieber effect)  Probing the instrument/vaporizor status (e.g. potential drifts):

 frequent injections of NH4NO3 and (NH4)2SO4 (NO/NO2 and NH4 and SO4 RIEs)  repeated injections of real samples  Examining the OA fragmentation patterns  Detecting new species OA fragmentation pattern

Analysis in argon for 3 sites in Lithuania Bozzetti et al., ACP Vilnius 0.12 Preila + Rugsteliskes fCO 0.08 + + 0.04 CO /CO2 ratio according to -3 the standard fragmentation table 25x10 (Aiken et al., 2008) 20 15 + fC2H4O2 10 5 0 0.25 0.20 + 0.15 fCO2 0.10 0.05 0.12 0.10 + 0.08 fC2H3O 0.06 0.04 1.2

+ + 0.8 C2H3O / CO2 0.4

0.0 2.5 2.0 + + CO2 / CO 1.5 1.0

01.11.2013 01.01.2014 01.03.2014 01.05.2014 01.07.2014 01.09.2014 OA fragmentation pattern

0.20 Bozzetti et al., ACP

1:1 line 0.15 p = Paris PM2.5 m = Magadino PM2.5 M = Magadino PM10

WB = Wood Burning smoke chamber experiments + 2 Lithuania PM1 0.10 + + CO2 /CO Lithuania = 1.78med fCO + + CO2 /CO Paris, Magadino, BB = 1.80med

0.05

0.00 0.00 0.04 0.08 0.12 + fCO OA fragmentation pattern

0.20 Bozzetti et al., ACP

Small Multi-functionalized 1:1 line 0.15 p = Paris PM2.5 m = Magadino PM2.5 M = Magadino PM10

WB = Wood Burning smoke chamber experiments + 2 Lithuania PM1 0.10 + + CO2 /CO Lithuania = 1.78med fCO + + CO2 /CO Paris, Magadino, BB = 1.80med

0.05

Most ROH Canagaratna, compounds 0.00 2015 0.00 0.04 0.08 0.12 + fCO Better characterizing the AMS using the offline AMS technique

 Identifying particle-vaporizer interactions (e.g. Pieber effect)  Probing the instrument/vaporizor status (e.g. potential drifts):

 frequent injections of NH4NO3 and (NH4)2SO4 (NO/NO2 and NH4 and SO4 RIEs)  repeated injections of real samples  Examining the OA fragmentation patterns  Detecting new species Carbonates

Identification and quantification El Haddad et El al., Haddad in prep Carbonates

Identification and quantification El Haddad et El al., Haddad in prep Carbonates Identification and quantification

Bozzetti et al., in prep 0.6

3 0.5 15x10

m/z 44 )

0.4 -3 10

0.3 (µg m

0.2 PToF

Normalized spectra Normalized

M d 0.1 0

0.0 20 40 60 80 100 1 2 3 4 5 -3 m/z 6x10 PToF time (s)

RIE=1.4 using NaHCO3 Carbonates When does that matter?

 Sampling PM2.5  Sampling biomass burning  Offline AMS Carbonates How to deal with it?

 For offline AMS: HCl fumigation (tested and works)  For non-AMS: PMF/ME-2?