global simulation of glyoxal and methylglyoxal, and implications for soa

25
1 Global simulation Global simulation of of glyoxal glyoxal and methylglyoxal, and methylglyoxal, and implications for and implications for SOA SOA Tzung-May Fu, Daniel J. Jacob Harvard University April 11, 2007 Work supported by EPRI Thomas Kurosu, Kelly Chance Harvard/SAO CfA CHOCHO CH 3 C(O)CHO

Upload: jadzia

Post on 15-Jan-2016

39 views

Category:

Documents


0 download

DESCRIPTION

CHOCHO. Global simulation of glyoxal and methylglyoxal, and implications for SOA. CH 3 C(O)CHO. Tzung-May Fu, Daniel J. Jacob Harvard University. Thomas Kurosu, Kelly Chance Harvard/SAO CfA. April 11, 2007 Work supported by EPRI. SOA formation through uptake of dicarbonyls. - PowerPoint PPT Presentation

TRANSCRIPT

Page 1: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

1

Global simulation Global simulation of glyoxalof glyoxal

and methylglyoxal,and methylglyoxal,

and implications for SOAand implications for SOA

Tzung-May Fu, Daniel J. Jacob

Harvard University

April 11, 2007Work supported by EPRI

Thomas Kurosu, Kelly Chance

Harvard/SAO CfA

CHOCHO

CH3C(O)CHO

Page 2: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

2

SOA formation through uptake of dicarbonylsSOA formation through uptake of dicarbonyls

CHOCHOSOA

Isoprene (350 Tg C yr-1), monoterpenes, acetone, MBO, C2H4, C3H6

Reversible?

Photolysis

Oxidation

Deposition

[OH]

RH

pH

nuclei

Oligomers?

organic acids?

H* ~ 105

C2H2, C2H4, C3H6, aromatics, acetone, glycolaldehyde, hydroxyacetone

CH3C(O)CHOH* ~ 103

OH, O3, NO3

Page 3: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

3

Is dicarbonyl uptake irreversible?Is dicarbonyl uptake irreversible?O

rgan

ic /

su

lfat

e [g

/g]

[Glyx]g = 5 ppb

[Liggio et al., 2005b]

Irreversible

Time

= 2. x 103

For [Glyx]g = 0.1 ppb,

∆[Glyx]particle = 3 g m-3 hr-1

[Kroll et al., 2005]

Reversible

[Glyx]g = 200 ppb

KH* = 2.6 x 107 M atm-1

For [Glyx]g = 0.1 ppb,

[Glyx]particle = 0.003 g m-3

Org

anic

/su

lfa

te [g

g

-1]

Page 4: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

4

What What are the irreversible processesare the irreversible processes in the aqueous in the aqueous phase?phase?

H2O

H2O

Oligomers

+ hydrates + H2O

Kalberer et al. [2004]Liggio et al. [2005] Hastings et al. [2005] Zhao et al. [2006]

1

Organic acids

+ OH ?

Ervens et al. [2004]Lim et al. [2005]Warneck et al. [2005]Sorooshian et al. [2006]

2

Altieri et al. [2006]

3

H* ~ 105

H* ~ 103

Page 5: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

5

Tracers emitted, non-standard

– ISOP, MONX, MBO, C2H4, PRPE, C2H2, ACET, HAC, BENZ, TOLU, XYLE, GLYX, MGLY, GLYC, MVK, MACR, PAN, PMN, ACRPAN, ENPAN, GPAN, GLPAN, MPAN, NIPAN

Chemistry– New chemical mechanism from MCM v3.1, University of Leeds– JPL 2006 rate constants and photolysis (p-dependent)– Standard SOA from BVOC– Reactive uptake of dicarbonyls by aqueous aerosol and cloud droplets

[Liggio et al., 2005b; Zhao et al., 2006]

Standard emissions– FF + BF: GEIA + regional– BB: GFED2– BG: MEGAN

Non-standard emissions– FF + BF for C2H2: Xiao et al. [2007]– FF + BF for C2H4, arom.: RETRO– BB: Scale GFED2 CO w/ EFs– BG: MEGAN

Dry/wet deposition: – GLYX, MGLY, GLYC, PANs

GEOS-Chem v736 4x5GEOS-Chem v736 4x5 2005 2005/12/12 – 2006 – 2006/11/11 (GEOS4) (GEOS4)

0

5

10

15

20

25

30

35

iso

pre

ne

mo

no

terp

enes

acet

on

e

PR

PE

hyd

roxy

acet

on

e

eth

ene

tolu

ene

acet

ylen

e

xyle

nes

ben

zen

e

gly

oxa

l

gly

cola

ldeh

yde

met

hyl

gly

oxa

l

Em

issi

on

s [

Tg

/ y

r]

Biogenic

Biomass burning

Anthropogenic

40

9

11

0

64

Page 6: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

6

New isoprene oxidation – adapted from MCM v3.1New isoprene oxidation – adapted from MCM v3.1

ISOP

IALD

GLYC MGLY

OH, O3

OH

HACGLYX

MVKOH NIALD MVK MACR

NO3

9.8h1h1.5h 2.7h

1.5h 0.7h

High NOx, no RO2 recycling

ISOP + OH 0.045 GLYX + 0.508 GLYC + 0.233 MGLY + 0.197 HAC + 1.033 CH2O

2h 1h0.7h

0.3h

Production of glyoxal

Larger yield of methgylglyoxal, GLYC, HAC

Larger yield of CH2O

Page 7: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

7

Are the two Isoprene Are the two Isoprene SOA pathways additive? SOA pathways additive?

SOA via partitioning of SOA via partitioning of semi-volatile products semi-volatile products

from isoprenefrom isoprene

SOA via irreversible SOA via irreversible uptake of glyoxal uptake of glyoxal

from isoprenefrom isoprene

Y = 1~2 % at high [NOx]

Y = 3 % at low [NOx]

YGLYX ~ 10 % at high [NOx]

YGLYX < 5 % at low [NOx]

Experiments by [Kroll et al., 2006] Mechanism from MCM v3.1 (U of Leeds)

Methacrolein is an important intermediate

Methyl vinyl ketone is an important intermediate

Two pathways of SOA formation from isoprene are additive

Page 8: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

8

C2H2 + OH 0.636 GLYX

MONX + O3 0.05 GLYX + 0.05 MGLY

MBO + OH 0.63 GLYC + (0.63 ACET)

BENZ + OH 0.252 GLYX

TOLU + OH 0.162 GLYX + 0.124 MGLY

XYLE + OH 0.156 GLYX + 0.230 MGLY

Parameterized chemistryParameterized chemistry

(0.16 - 0.29)

(0.08 - 0.39)

(0.03 – 0.40)

(0.03 - 0.18)

(0.11 - 0.42)

High NOx

C2H4 + OH 0.995 [ GLYC + (1-) ∙ 2 HCHO + HO2], = 0.3 ~ 1

CC22HH44 oxidation – from MCM v3.1 oxidation – from MCM v3.1

Page 9: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

9

@ sfc @ 2 km

[ppb] [ppb]

Monthly mean [GLYX], Jul 2006Monthly mean [GLYX], Jul 2006

10-1 ~ 10-2 ppb

~ 10-2 ppb

CHOCHO Inventory 9.3 [Gg]

Production Emission [Tg yr-1] Molar yield [%] 51 [Tg yr-1] Isoprene 408 7.1 + 2.8 21 C2H2 6.9 64 9.8 Glyoxal 5.3 100 5.2 C2H4 12 6.3 1.9 Monoterpenes 110 2.6 1.2 Benzene 6.0 25 1.1 Toluene 7.7 16 0.79 Xylenes 5.9 16 0.51 Glycolaldehyde 3.9 11 0.41 MBO 2.0 6.1 0.08 Loss 51. [Tg yr-1] Photolysis 24 SOA formation 12+6.0 Oxidation 5.9 Dry deposition 1.2 Wet deposition 0.90

Page 10: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

10

@ sfc @ 2 km

[ppb] [ppb]

Monthly mean [MGLY], Jul 2006Monthly mean [MGLY], Jul 2006

10-1 ~ 10-2 ppb

~ 10-2 ppb

CH3COCHO Inventory 17 [Gg]

Production Emission [Tg yr-1] Molar yield [%] 159 [Tg yr-1] Isoprene 408 19 + 12 69 Acetone 65 12 + 0.008 9.7 PRPE

(>C3 alkene) 31 14 7.4

Methylglyoxal 3.4 100. 3.4 Monoterpenes 110 2.6 1.5 Xylenes 5.9 23 0.93 Toluene 7.7 12 0.75 Loss 159 [Tg yr-1] Photolysis 103 SOA formation 21+17 Oxidation 16 Dry deposition 1.3 Wet deposition 1.1

Page 11: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

11

]p

pt

[

Hourly mean [MGLY], [GLYX]

Local time [h]

Blodgett Forest, CA Aug-Sep 2000

Spaulding et al. [2003]

Tomakomai Forest, Japan Sep 2003

[pp

t]

Ieda et al. [2006]

Page 12: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

12

Comparison with SCIAMACHY [Wittrock et al., 2006]Comparison with SCIAMACHY [Wittrock et al., 2006]

SCIAMACHY GLYX VC (10LT)

SCIAMACHY HCHO VC

HCHO VC Dec05-Nov06

GLYX VC Dec05-Nov06 (12-15LT)

Page 13: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

13

@ surface

@ 4.2 km

SOASOAglyxglyx+SOA+SOAmglymgly

SOASOA from isoprene from isoprene (Y (Yisopisop = 1~3%) = 1~3%)

SOASOA from from monoterpenes etcmonoterpenes etc

P ~ 6 Tg yr-1 Pglyx ~ 12 + 6 Tg yr-1

Pmgly ~ 21 + 17 Tg yr-1

[ug m-3]

P ~ 10 Tg yr-1

[0.1 ug m-3]

30% 12% 58%

27% 23% 50%

Page 14: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

14

ConclusionsConclusions

I. Global simulation of glyoxal and methylglyoxalI. Global simulation of glyoxal and methylglyoxal

– Model sfc concentration same order as rural measurements

– Model glyoxal VC <~ satellite VC measurements

– Isoprene oxidation products well simulated w/ new mechanism

– Isoprene is the largest source dicarbonyls in the free troposphere

– Direct emissions from biomass burning large surface concentrations

II. Reactive uptake of dicarbonyls is significant SOA sourceII. Reactive uptake of dicarbonyls is significant SOA source

– Produces > 18 Tg yr-1 SOA, comparable to other biogenic SOA sources

– SOA from isoprene via glyoxal is independent of SOA from isoprene via

semi-volatile gaseous products

– Anthropogenic and biomass burning emissions produce significant SOA via

uptake of dicarbonyls

Page 15: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

15

WHAT DON’T WE UNDERSTAND ABOUT SOA FORMATION?WHAT DON’T WE UNDERSTAND ABOUT SOA FORMATION?

dicarbonyls

Oxidation by OH, O3, NO3

Direct Emission

Terpenes

Nucleation or Condensation

Aromatics

OC

Isoprene

CloudProcessing

FF: 45-80 TgC/yrBB: 10-30 TgC/yr

SOA: >20 Tg/yr

Fossil Fuel Biomass Burning

ANTHROPOGENIC SOURCESBIOGENIC SOURCES

Heterogeneous ReactionsPRECURSORS

CHEMISTRY1. NOx, SO2/acidity2. Multi-step oxidation

FORMATION PATHWAYS

C/o Colette Heald

Page 16: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

16

““Reactive uptake of dicarbonyl” Reactive uptake of dicarbonyl” is an attractive mechanism, because it …is an attractive mechanism, because it …

• is a sustained SOA source in aged air masses and free troposphere

• works for both biogenic and anthropogenic precursors

• takes place rapidly in clouds, consistent with field evidence

• produces SOA quickly near the source region. Diurnal variations of SOA concentration more similar to measurements with maximum concentration in the afternoon

• may explain large, heterogeneous source of oxalic acid in aerosols

• may explain observed oligomers in aerosols

Page 17: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

17

Experiments needed to determine …Experiments needed to determine …

• Photolysis quantum yield in the visible band

• Reversibility of uptake

• Uptake sensitivity to pH

• Uptake sensitivity to ionic strength

• Total mass contribution of oligomers

• Presence/ID of organic acid hydrate oligomers

• Consistency with ambient aerosol mass spectra

Page 18: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

18

Pabstthum, Germany Jul-Aug 1998

Grossmann et al. [2003]

Mexico City Apr 2003

Volkamer et al. [2005b]

]p

pt

[

Hourly mean [MGLY], [GLYX]

Local time [h]

Page 19: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

19

Blodget Forest, CABlodget Forest, CAAug-Sep 2000Aug-Sep 2000

Spaulding et al. [2003]

GEOS-ChemGEOS-ChemAug-Sep 2006Aug-Sep 2006

Local time [h]

Page 20: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

20

Isoprene oxidation – GEOS-ChemIsoprene oxidation – GEOS-Chem

ISOP

IALD MVK MACR

GLYC MGLY

OH, O3, NO3OH

HACGLYX

High NOx, no RO2 recycling

ISOP + OH 0.317 GLYC + 0.198 MGLY + 0.158 HAC + 0.969 CH2O

2h 1h

9.8h1h1.5h 2.7h

0.8h

No glyoxal production from isoprene

0.3h

Page 21: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

21

ORGANIC CARBON AEROSOLORGANIC CARBON AEROSOL

ReactiveOrganicGases

Oxidation by OH, O3, NO3

Direct Emission

Fossil Fuel Biomass Burning

Monoterpenes, etc

Nucleation or Condensation

Aromatics

ANTHROPOGENIC SOURCESBIOGENIC SOURCES

OC

FF: 45-80 TgC/yrBB: 10-30 TgC/yr

Secondary Organic Aerosol (SOA): 8-40 TgC/yrGoldstein and Galbally [2007] 510-910 TgC/yr

*Numbers from IPCC [2001]

Global Model Representation of SOA:1. “Effective primary” yield of semivolatile gas2. Two-product empirical fit to smog chamber data

Isoprene 350 TgC/yr

Partitioning of semivolatile gas?

Heterogeneous rxn of soluble gas?

Other mechanisms?

Volkamer et al. [2006]

C/o Colette Heald

Page 22: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

22

[Glyoxal] in ambient air[Glyoxal] in ambient air

Rural

10-1 ~10-2 ppb

Remote and FT

10-1 ~ 10-2 ppb

Location Time [Glyoxal], ppb Statistics Reference Rural Pinnacles, VA

(1037 m) 38º32’N, 78º21’W

Sep 1990 0.044 (<0.02 – 0.35)

Diurnal mean Diurnal range

Munger et al. [1995]

Metter, GA Jul-Aug 1991 0.018 (* - 0.091) 0.016

9-18LT mean 9-18LT maximum midafternoon mean

Lee et al. [1995]

Metter, GA Jun 1992 0.083 (* - 0.19) 0.085

9-18LT mean 9-18LT maximum midafternoon mean

Lee et al. [1995]

Blodgett Forest, CA (1315 m)

38º53’N, 120º37’W

15-19 Aug, 11-15 Sep, 2000

0.020 – 0.054 9-17LT range Spaulding et al. [2002]

Blodgett Forest, CA (1315 m)

38º53’N, 120º37’W

15-19 Aug, 11-15 Sep, 2000

0.027±0.015 (0.006 – 0.083)

Diurnal mean±S.D. Diurnal range

Spaulding et al. [2003]

Tábua, Portugal 40º19’N, 8º3’W

12-24 Aug, 1996 (0 – 0.6) Diurnal range Cerqueira et al. [2003]

Anadia, Portugal 40º25’N, 8º24’W

12-24 Aug, 1996 (0 – 0.2) Diurnal range Cerqueira et al. [2003]

Hokkaido, Japan 44º21’N, 142º15’E

22-29 Aug, 2002 0.018 (<0.001 – 0.065)

Diurnal mean Diurnal range

Matsunaga et al. [2004]

Thuringian Forest, Germany (605 m)

50º37’N, 10º43’E

Oct 2001, Oct 2002

(<0.005 – 0.03) Diurnal range Müller et al. [2005]

Tomakomai Flux Research Site, Japan (140 m)

42º44’N, 141º31’E

3-5 Sep, 2003 0.025 (0.012 – 0.0525) b

Diurnal mean Diurnal range

Ieda et al. [2006]

Salt Point, CA Aug-Sep 2005 <0.020 b 11-14LT range Seaman et al. [2006] Lassen, CA Aug-Sep 2005 <0.035 b 11-14LT range Seaman et al. [2006] Remote Carribean Sea and

Sargasso Sea Oct 1988 – Mar

1989 0.08 Zhou and Mopper [1990]

Free troposphere Southern Nova

Scotia, Canada (1 – 3 km)

28 Aug 1993 (0 – 0.5) 12-14LT range Lee et al. [1996] (pollution event)

Nashville, TN (> 1.9 km)

July 1995 0.026±0.009 (0.01 – 0.06)

Mean±S.D. Range

Lee et al. [1998]

Page 23: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

23

[Methyglyoxal] in ambient air[Methyglyoxal] in ambient air

Rural

>1 ~10-2 ppb

Remote and FT

~ 10-2 ppb

Location Time [Methylglyoxal], ppb

Statistics Reference

Rural Pinnacles, VA

(1037 m) 38º32’N, 78º21’W

Sep 1990 <= 0.05 Diurnal range Munger et al. [1995]

Metter GA Jul-Aug 1991 0.031 (* - 0.16) 0.033

9-18LT mean 9-18LT maximum midafternoon mean

Lee et al. [1995]

Metter GA Jun-Jul 1992 0.088 (* - 0.26) 0.08

9-18LT mean 9-18LT maximum midafternoon mean

Lee et al. [1995]

Blodgett Forest, CA (1315 m)

38º53’N, 120º37’W

15-19 Aug, 11-15 Sep, 2000

0.069 – 0.39 9-17LT range Spaulding et al. [2002]

Blodgett Forest, CA (1315 m)

38º53’N, 120º37’W

15-19 Aug, 11-15 Sep, 2000

0.13±0.06 (0.032 – 0.32)

Diurnal mean±S.D. Diurnal range

Spaulding et al. [2003]

Tábua, Portugal 40º19’N, 8º3’W

12-24 Aug, 1996 (0 – 2.0) Diurnal range Cerqueira et al. [2003]

Anadia, Portugal 40º25’N, 8o24’W

12-24 Aug, 1996 (0 – 0.2) Diurnal range Cerqueira et al. [2003]

Hokkaido, Japan 44º21’N, 142º15’E

22-29 Aug, 2002 0.028 (0.004 – 0.088)

Diurnal mean Diurnal range

Matsunaga et al. [2004]

Thuringian Forest, Germany (605 m)

50º37’N, 10º43’E

Oct 2001, Oct 2002 0.015 – 0.15 Diurnal range Müeller et al. [2005]

Tomakomai Flux Research Site, Japan (140 m)

42º44’N, 141º31’E

3-5 Sep, 2003 0.0541 (0.020 – 0.129) b

Diurnal mean Diurnal range

Ieda et al. [2006]

Salt Point, CA Aug-Sep 2005 <0.012 b 11-14LT range Seaman et al. [2006] Lassen, CA Aug-Sep 2005 0.015±0.0056 b 11-14LT range Seaman et al. [2006] Remote Carribean Sea and

Sargasso Sea Oct 1988 – Mar

1989 ~ 0.01 Zhou and Mopper [1990]

Free troposphere Southern Nova

Scotia, Canada (1 – 3 km)

28 Aug 1993 (0 – 1) 12-14LT range Lee et al. [1996] (pollution event)

Nashville, TN (> 1.9 km)

July 1995 0.02 (0 – 0.07)

Mean±S.D. Range

Lee et al. [1998]

Page 24: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

24

What What are the irreversible processesare the irreversible processes in the aqueous phase? in the aqueous phase?

I. Hydrate + OH

glyoxylic acid, pyruvic acid

Oxalic acid

II. Hydrate + H2O

Oligomers

Ervens et al. [2004]; Lim et al. [2005]; Warneck et al. [2005]; Sorooshian et al. [2006]

Kalberer et al. [2004]; Liggio et al. [2005]; Hastings et al. [2005]; Zhao et al. [2006]

Page 25: Global simulation  of glyoxal and methylglyoxal, and implications for SOA

25

III. Hydrate + OH

glyoxylic acid, pyruvic acid oligomers

oxalic acid oligomers

Altieri et al. [2006]

T = 10 min

T = 202 min