M. Bressi, F. Cavalli, C. Belis, J-P. Putaud, S. Dos Santos, K. Douglas, R. Passarella R. Fröhlich, A. Prévôt E. Petralia, M. Berico, A. Malaguti, M. Stracquadanio.

M. Bressi, F. Cavalli, C. Belis, J-P. Putaud, S. Dos Santos, K. Douglas, R. Passarella

R. Fröhlich, A. Prévôt

E. Petralia, M. Berico, A. Malaguti, M. Stracquadanio

Chemical composition of submicron particles with an Aerosol Chemical Speciation Monitor

at the JRC-Ispra site

1. IntroductionWhy to study the chemical composition of submicron particles (PM1)?

2Aerosol SourcesAerosol Size

« Aerosols: collection of airborne solid or liquid particles » (IPCC, 2007)

Adapted from Putaud et al. (2010)

Comparison of annual PM2.5 levels at different sites in Europe

2020 EU Annual PM2.5 limit value

= 20 µg/m3

Why to study the chemical composition of PM1 at Ispra?

Decesari et al. (2001)

Po Valley region

3

Why to use an Aerosol Chemical Speciation Monitor (ACSM)?

1. Scientific reasons 30 min time resolutionApportionment of organic aerosols

2. Operational reasons Long-term, unattended and stable field measurements

3. Specificity of the JRC

4

http://www.psi.ch/acsm-stations/ acsm-and-emep-stations

European ACSM network

Objectives of this talk

Present the assessment of the precision and the accuracy of ACSM measurements

Document the levels and the chemical composition of PM1 at Ispra

Focus on the organic fraction of PM1 at Ispra

5

2. Material and methods

2.1 Sampling site

6

PM1020/06/2013 On-going

20/12/2012 20/08/2013

Continuous measurements since 20 December 2012

Ng et al., 2011

Q-MS

Chemical composition of non-refractory (NR) PM1: Organics (Org), NO3, SO4, NH4, Cl

Time resolution: 30 min

2.2 ACSM description

Aerodyne Research Inc.

7

4 calibrationsRFNO3

=3.03E-11 amp/(µg/m3)

RIENH4=5.3

Ionization Efficiency (IE)Ammonium Nitrate

calibrations

1 calibrationRFNO3

=4.46E-11 amp/(µg/m3)

RIENH4=6.0

Collection Efficiency (CE) Relative IE (RIE)

Time series correction

Ion transmission

correction (ITC)

CE=0.5 for every compound.RIE: Org, SO4, NO3, Cl from literature

(Canagaratna et al., 2007; Jimenez et al., 2003; Ng et al., 2011)

Very stable N2 signal SEM voltage modified every ~2-3 days

Close to the default correction Substantial discrepancies

ACSMs

8

2.3 Calibration, tuning and correction factors

3. Results3.1 Metrology

9

OR

GA

NIC

SA

MM

ON

IUM

Comparison between 2 ACSMs: 20/06/2013 18/08/2013

Conclusion: Despite discrepancies between both instruments (ages, calibrations, etc.) Good agreement between both ACSMs

0.98<r2<0.99 and 0.8<slopes<1.2 for Org, SO4, NO3, NH4

12h averaged data(n=116)

Conclusion: Good agreement between ACSM and other analytical techniques

10

Comparison between ACSM and independent analytical techniques: 20/12/2013 28/02/2013O

RG

AN

ICS

NIT

RA

TE

AC

SM

C

OM

PO

UN

DS

+

EC

3.2 Chemistry

11

Temporal variation of NR-PM1 during the winter season

(20/12/2012 21/03/2013)

Strong daily variations atmospheric dilution and

condensation

Example of an atmospheric process visible with the

ACSM

Temporal variation of PM1 chemical composition

12

Average PM1 chemical composition during the winter season (20/12/2012 21/03/2013)

[PM1]ACSM+EC= 31.9 ± 18.1 µg/m3

Average (µg/m3, %)

PM1 levels during winter at Ispra >> + EU: PM2.5 Annual limit value 2020 = 20 µg/m3

+ WHO: PM2.5 24-hour mean daily max = 25 µg/m3

High proportion of Organics (54%)SIA (37%)

Very high PM1 levels at Ispra Comparable NR-PM1 chemical composition

Zhang et al. (2007), Jimenez et al. (2009)

3.3 Organic Aerosol (OA) apportionment

13Hallquist et al. (2009)

Classical classes of OA in AMS/ACSM studies

OA

OOA

HOA

LV-OOA

SV-OOA

COA BBOA

Organic compounds: Definition: any molecules made of C and H that can also contain O, N, P.Low level of understanding regarding their effects on human health, biogeochemical cycles and Earth’s climate.

Organic Aerosol

Oxygenated OA

Hydrocarbon-likeOA

Low-Volatility OOA

Semi-Volatile OOA

CookingOA

Biomass BurningOA

14

X = G * F Mass spectral Matrix

Factor contribution Factor profile

= *

+ EResidual matrix

Method: Positive Matrix Factorization (PMF)

m/z 12 m/z 13 m/z 14 … n8:00 0.01 0.10 0.058:30 0.00 0.27 0.309:00 0.22 0.00 0.00

…m

m/z 12 m/z 13 m/z 14 … n08:00am 0.01 0.10 0.0508:30am 0.00 0.27 0.3009:00am 0.22 0.00 0.00

…m

+Na Cl EC … n

11/09/2009 0,00 0,00 0,0112/09/2009 0,00 0,00 0,0213/09/2009 0,00 0,01 0,02

…m

m/z 12 m/z 13 m/z 14 … nm/z 12 m/z 13 m/z 14 … n08:00am 0.02 0.01 0.0508:30am 0.00 0.04 0.0609:00am 0.05 0.00 0.00

…m

m/z 12 m/z 13 m/z 14 … nm/z 12 m/z 13 m/z 14 … nF1 0.01 0.02 0.30F2 0.00 0.27 0.01F3 0.22 0.00 0.00…p

F1 F2 F3 … p08:00am 0.25 0.01 0.0508:30am 0.00 0.30 0.0609:00am 0.05 0.00 0.44

…m

15

m/z 57m/z 60

m/z 44

Factor identification (1/2)

Facto

r p

rofi

les

20/12/2012 21/04/2013

16

LV-OOA 1

LV-OOA 2

BBOA 1

BBOA 2

HOA

OOA other

Factor identification (2/2)

Correlation between my factor profiles and factor profiles found in the literature(field studies or chamber experiment)

17

Factor contribution (1/2)

Abso

lute

contr

ibuti

on

(µg/m

3)

Rela

tive c

ontr

ibuti

on

Diu

rnal co

ntr

ibuti

on

(µg/m

3)

Avera

ged c

ontr

ibu

tion

(µg

/m3;

%) Period

20/12/2012 21/04/2013

LV-OOA 1LV-OOA 2 BBOA 2HOAOOA other BBOA 1

18

Factor contribution (2/2)

Comparison of OA apportionment with other European studies (adapted from Crippa et al., 2013)

Note: Sampling periods Ispra study: December 2012 – April 2013 Other studies: February/March 2009

4. Conclusions & Perspectives

4.1 Conclusions

19

4.2 Perspectives

20

Metrology• Comparison between 13 ACSMs in Paris in November (ENEA instrument + data

treatment tool developed by Claudio Belis)• ACSM ? Chemical composition of PM in the European air quality network ?

Chemistry• A European aerosol phenomenology: real-time measurements of PM1 chemical

composition (JRC leadership under discussion)

OA apportionment• Use of external tracers, use of different methodologies (e.g. ME-2), etc.• Define a common methodology for OA apportionment at the European level

Source apportionment• Real sources behind this chemical composition??

APPENDICES

21

Why to study the chemical composition of submicron particles (PM1)?

22Forster et al. (2007)

23

Canagaratna et al. (2007)

24

AER

OS

OL

CH

EM

ISTR

Y

GA

S P

HA

SE

CH

EM

ISTR

Y

24-h daily sampling of PM2.5

20 Dec 2012 28 Feb 2013

Elemental Carbon (EC)

Organic Carbon (OC)

Organic Matter (OM):OM=1.4*OC (Putaud et al., 2010)Sunset Lab. OCEC Analyzer

(EUSAAR2 protocol)

Major ions including:SO4

NO3

NH4

ClIon ChromatographyPartisol Plus

AER

OS

OL

PH

YS

ICS

Condensation Particle Counter

Size distribution10 to 800 nm

Differential Mobility Analyzer

CO

NO, NO2

O3

20 Dec 2012 7 Feb 201320 Dec 2012 22 Feb 2013

SO2

OR

GA

NIC

S

Note: Org ENEA plotted with the default Ion Transmission Correction (ITC) on this graph

Defa

ult

IT

C

Exp

eri

men

tal ITC

12h averaged data(n=116)

25

ENEA versus PSI Q-ACSM

SU

LFA

TE

Defa

ult

IT

C

Exp

eri

men

tal ITC

12h averaged data(n=116)

Note: SO4 ENEA plotted with the default Ion Transmission Correction (ITC) on this graph

26

NIT

RA

TE

12h averaged data(n=116)

Incl

ud

ing

all

data

poin

ts

Excl

ud

ing

28

Ju

ne 1

2p

m

27

AM

MO

NIU

MC

HLO

RID

E

12h averaged data(n=116)

28

PSI Q-ACSM versus other analytical techniquesN

ITR

ATE

Good agreement for Secondary Inorganic Aerosol (SIA): r²=0.8-0.9, slope=0.9-1.0

SU

LFA

TE

AM

MO

NIU

M

29

CH

LO

RID

EO

RG

AN

ICS

AC

SM

CO

MP

OU

ND

S

+ E

C

Off-line Chloride large uncertainties

Surprisingly good despite expected

filter sampling artifacts

Good agreement

30

Zhang et al. (2007)

31

Jimenez et al. (2009)

32

Zhang et al. (2007)

33

Zhang et al. (2011)

34

Jimenez et al. (2009)

35

36

Kroll et al. (2011)

37

Kroll et al. (2011)

Courtesy of A. Prevot

38

39

Comparison between BBOA1 (Factor 5) and BBOA2 (Factor 4) factor profiles

40

Comparison between LVOOA1 (Factor 6) and LVOOA2 (Factor 2) factor profiles

41