The semi-volatile nature of secondary organic aerosol (SOA) in the Mexico City Metropolitan Area

November 2, 2007EAS Graduate Student Symposium

Christopher J. Hennigan

SOA background

Carbonaceous particulate matter

Elemental carbon

(EC, Black carbon, soot)

Organic carbon (OC)

Primary organic aerosol (POA)

Secondary organic aerosol (SOA)

SOA importance

• Can be a major fraction of PM2.5

– 10-80% of OC• Still poorly understood

– Sources– Formation mechanisms– Chemical composition– Impact on climate change (e.g., role as CCN)

MIRAGE field study• MIRAGE – part of multi-agency field study in 2006

•T1 ground site: located at Universidad Tecnológica de Tecámac (Tecamac University), approximately 30 km northeast and downwind of Mexico City center

PM2.5 online measurements• Inorganic ions: Na+, NH4

+, Ca2+, Mg2+, Cl-, NO3-, SO4

2-, (PILS-IC)• Water Soluble Organic Carbon (PILS-WSOC)

Gas-phase measurements included CO, NOx, HNO3, NH3, OH, VOCs

Meteorology parameters

PILS-WSOC measurement • Measures water soluble fraction of OC aerosol on-line

• PILS-WSOC vs. filter-extracted WSOC compare favorably (slopes = 0.88 – 1.35)

• WSOC and SOA by EC tracer method highly correlated (R2 = 0.70 - 0.79) observed WSOC/SOA 0.67 – 0.75

• Excellent agreement (R2 = 0.86 - 0.93) between oxygenated organic carbon (OOC) aerosol via AMS and WSOC; 88% of OOC water soluble

• Less than 10% of primary OC is water soluble (summer, fall, and winter in Tokyo)

• WSOC is an approximate measure of SOA

[Sullivan et al., 2004; Zhang et al., 2005; Miyazaki et al., 2006; Kondo et al., 2007]

March 27-29 chemical composition

•Typical diurnal patterns

•Correlation suggests similar sources and atmospheric processing

• Use nitrate behavior to investigate SOA?

7654321

WSO

C (µg C

m-3)

12:00 AM3/27/2006

12:00 PM 12:00 AM3/28/2006

12:00 PM 12:00 AM3/29/2006

12:00 PM

Time (CST)

20

15

10

5

0

Nitr

ate

(µg

m-3

)

WSOC Nitrate

5

4

3

2

WSO

C (µ

g C

m-3

)

2015105Nitrate (µg m-3)

R2 = 0.80N = 162

Nitrate production

1000

800

600

400

200

0

Sola

r rad

iatio

n (W

m-2

)

12:00 AM3/27/2006

12:00 PM 12:00 AM3/28/2006

12:00 PM 12:00 AM3/29/2006

12:00 PM

Time (CST)

7

6

5

4

3

2

1

WSO

C (µg C

m-3)

20

15

10

5

0

Nitr

ate

(µg

m-3

)

WSOC Nitrate solar radiation

Average NO3- concentration increase (7:00am – 11:00am) = 13.8 μg m-3

∫11am

7amk * [OH] * [NO2] * dt = 15 μg m-3Average morning HNO3 (g) production

ISORROPIA thermodynamic equilibrium model prediction: NO3

-

(NO3- + HNO3)

> 0.90

NO3- from secondary photochemical production; WSOC from SOA formation

Avg. increase (7:00am – 11:00am): NO3- = 13.8 μg m-3 (300%); WSOC = 1.6 μg C m-3 (50%)

Nitrate, WSOC concentration decrease

1000

800

600

400

200

0

Sola

r rad

iatio

n (W

m-2

)

12:00 AM3/27/2006

12:00 PM 12:00 AM3/28/2006

12:00 PM 12:00 AM3/29/2006

12:00 PM

Time (CST)

7

6

5

4

3

2

1

WSO

C (µg C

m-3)

20

15

10

5

0

Nitr

ate

(µg

m-3

)

WSOC Nitrate solar radiation

Possibly due to:– Boundary layer dilution– Advection– Atmospheric processing (thermodynamics)– Combination of the above

Avg. decrease (11:00am – 12:45pm): NO3- = 14.9 μg m-3 (82%); WSOC = 2.68 μg C m-3 (56%)

Modeling nitrate thermodynamics

• Semi-volatile NH4NO3 well understood– Exists in gas (HNO3, NH3) or particle (NH4NO3) phase at

ambient conditions– NH4NO3: T, RH, pH, aerosol composition (SO4

2-) dependent

• ISORROPIA-II thermodynamic equilibrium model– Predicts aerosol H2O content

– Predicts equilibrium gas-particle partitioning of inorganic

species (including NH3(g)/NH4+, HNO3(g)/NO3

-)

[Nenes et al., 1998; Fountoukis et al., 2007]

ISORROPIA model output

20

15

10

5

0

Nitrate (µg m

-3 )

00:003/27/2006

12:00 00:003/28/2006

12:00 00:003/29/2006

12:00

Central Standard Time

30

25

20

15

10

5Tem

pera

ture

(ºC

) 80

60

40

20

RH

(%)

100

80

60

40

20

Per

cent

aer

osol

nitr

ate

Temperature RH Aerosol nitrate fraction Nitrate

[Fountoukis et al., 2007]

ISORROPIA predicts observed drop in NO3- just before noon

Correlation to SOA?

• Dramatic decrease in semi-volatile NO3

- concentration was due at least in part to thermodynamics (NO3

- evaporation)

• Was observed WSOC concentration decrease also due to the semi-volatile nature of SOA?

• Assess the possible impacts of BL dilution and advection by using a conservative tracer (CO in this case)

12:00 PM3/27/2006

6:00 PM

Time (CST)

20

15

10

5

Nitr

ate

(µg

m-3

)

35

30

25

20

15

10

5

NO

3 :CO

(µg m-3 ppm

-1)

Nitrate Nitrate:CO ratio

WSOC:CO and NO3-:CO ratios

5

4

3

2

WS

OC

(µg

C m

-3)

12:00 PM3/27/2006

6:00 PM

Time (CST)

20

15

10

5

Nitr

ate

(µg

m-3

)

35

30

25

20

15

10

5

NO

3 :CO

(µg m-3 ppm

-1)

12

10

8

6

4

2

WS

OC

:CO

(µg C m

-3 ppm-1)

Nitrate Nitrate:CO ratio

WSOC WSOC:CO ratio

6

5

4

3

2

1

Wind speed (m

s-1 )

12:00 PM3/27/2006

6:00 PM

Time (CST)

35

30

25

20

15

10

5

NO

3:C

O ra

tio (µ

g m

-3 p

pm-1

)

12

10

8

6

4

2

WS

OC

:CO

ratio (µg C m

-3 ppm-1)

Wind speed NO3:CO ratio WSOC:CO ratio

WSOC:CO and NO3-:CO

R2 = 0.46

Appears as if meteorology is driving the afternoon WSOC concentration, not thermodynamics (observed on all three days)

Summary

• Nitrate and WSOC (SOA) highly correlated (R2 = 0.80) in Mexico City Metropolitan Area (MCMA), indicating similar sources and atmospheric processing

• Nitrate experienced a rapid phase shift from the particle to the gas phase around noon, with observations and model results agreeing well; not so for SOA

• Fresh SOA with strong anthropogenic influence in the MCMA was less volatile than NH4NO3

– Semi-volatile intermediates more stable products– Oligomer formation?– Thermodynamic threshold not reached?

Fast (2 - 3hrs)

Acknowledgements

NSF grant: ATM 0513035

Amy P. Sullivan, Arsineh Hecobian, Rodney J. Weber, Athanasios Nenes, Christos I. Fountoukis, Oscar Vargas, Anne T. Case Hanks, L. Gregory Huey Georgia Institute of Technology

Barry L. LeferUniversity of Houston