no /n - reed college€¦ · mellon environmental studies summer experience fellowship and reed...

Acknowledgments

Correlations of MS Products with RONO2, Cumulative

0.15

0.10

0.05

0.00

Diur

nal N

O3 S

peci

es ( p

pbv )

2420161284Hour of Day

MARGA NO3- (aero)

TD-LIF total ANs TD-LIF particle ANs

6/10/13 6/20/13 6/30/13 7/10/13Central Daylight Time

0.6

0.5

0.4

0.3

0.2

0.1

0.0

NO2 L

oss

Rate

(pp b

/hr)

Predicted production of Organonitrate Predicted production of Inorganic nitrate

1.0

0.8

0.6

0.4

0.2

0.0

µg m

-3

6/10/13 6/20/13 6/30/13 7/10/13Central Daylight Time

16

12

8

4

0

µg m

-3

nss - Na+

Ca2+

Mg2+

K+

NO3-

< PM1PM1 - PM2.5

a 06/09/2013 16:00 CDT b 06/12/2013 13:00 CDT35

30

25

20

15

latit

ude

50

45

40

35

30

latit

ude

-100 -95 -90 -85 -80 -75 longitude

-115 -110 -105 -100 -95 -90 longitude

4000

3000

2000

1000

0altit

ude

(mAG

L)

6000

4000

2000

0altit

ude

(mAG

L)

c 06/24/2013 19:00 CDT d 06/28/2013 19:00 CDT

36

34

32

30

28

26

24

latit

ude

45

40

35

30

25

20

latit

ude

-94 -92 -90 -88 -86 -84 -82 longitude

-120 -110 -100 -90 longitude

2000

1500

1000

500

0

altit

ude

(mAG

L)

0 -10 -20 -30 -40 -50 -60 -70 hours since release0 -10 -20 -30 -40 -50 -60 -70 hours since release

12000

8000

4000

0

altit

ude

(mAG

L)




air mass age (h)

0 24 48 72

mass-weighted trajectorycluster centroid (scaled by fraction of particles)

NO3/N2O5 and BVOC Concentrations at SOAS

1400

1200

1000

800

600

400

200Gas

pha

se C

5H9N

O5 (

c oun

ts/m

in)

0.40.30.20.1Predicted loss to Isoprene (ppb)

R2 = 0.702

NOy fate at SOAS 2013: Organonitrate Formation via NO3 + BVOC and Inorganic Nitrate Formation via Heterogeneous Uptake of HNO3

- NO3 loss during the day is due to photolysis, reaction with NO and reaction with BVOCs- Almost 25% of daytime NO3 loss is from C10 terpenes (Figure 6)- Almost 15% of nighttime NO3 loss is from Isoprene- Close to half of nighttime NO3 loss comes from α-pinene- Modeled photolysis is normalized to solar radiation values (W m-2) to account for cloud cover- Predicted NO3 loss is calculated using published NO3 + BVOC values

(NO3, loss)pred = kNO3 + VOC [VOC][NO3]SS

RONO2, Cumulative = Σ(NO3, loss)pred x Δt

- RONO2, Cumulative is correlated to measured mass spectrometry data (Figure 7)

0.6

0.5

0.4

0.3

0.2

0.1

0.0

Aero

s ol O

rga n

onitr

a te

(AM

S , µ

g/m

3 )

12:00 AM6/28/13

12:00 AM6/29/13

12:00 AM6/30/13

12:00 AM7/1/13

12:00 AM7/2/13

12:00 AM7/3/13

Central Daylight Time

1.5

1.0

0.5

0.0

Predicted NO3 loss to Terpenes (µg/m

3)

Monoterpene Loss Isoprene Loss

600

400

200

0

C 10H

17NO

5 (g,

cou

nts /

s ec )

0.60.40.2Predicted loss to monoterpene (ppb)

600

400

200

0

C 10H

15NO

5 (g,

cou

nts /

s ec )


800600400200

0

C 10H

15NO

5 (p,

cou

nts /

s ec )


600

400

200

0

C 10H

17NO

5 (p,

cou

nts /

s ec )


aR2 = 0.198

cR2 = 0.436

bR2 = 0.419

dR2 = 0.674

0.25

0.20

0.15

0.10

0.05

0.00

Aero

sol O

rgan

oni tr

a te

( AM

S, p

pb)

0.70.60.50.40.30.20.1Predicted loss to Monoterpene (ppb)

y = 0.22754x + 0.03547R2 = 0.6316

Monoterpene Isoprene

- AMS NO3 mass loading in organonitrates is correlated with RONO2, Cumulative to monoterpenes with a slope of 0.23 giving a 23% molar yield of organonitrate from NO3 (Figure 8)- Gas phase C5H9NO5 correlates well with NO3 loss to isoprene (Figure 9)- Aerosol phase C5H9NO4 & C5H9NO5 were not detected during SOAS- Individual CIMS products correlated against predicted NO3 show which products are from NO3 radical reaction and which are not (ex. RO2 + NO) using the best �t analysis- Gas phase is not well correlated suggesting another type of reaction is taking place- Aerosol phase shows C10H17NO5 appears to be a product of NO3 reaction with an R2 value of 0.674

0.12

0.10

0.08

0.06

0.04

0.02

0.00

Mon

oter

pene

Oxi

datio

n Ra

te (p

pbv

hr-1

)

2420161284Hour of Day

HO NO3 O3

- OH dominates daytime oxidation rates with a peak of 0.6 ppbv/hr- NO3 rates are most dominant in the early morning hours when O3 concentrations are at their lowest (blue and red trace respectively)- Nighttime oxidation of total monoterpenes proceed at approximately equal rates from NO3 and O3 (blue and red traces respectively)- Monoterpenes used in calculations are α-pinene, β-pinene, camphene, myrcene and limonene

- Two observed periods of high NO3

- are correlated with high PM2.5 aerosol mass loading- High NO3

- is also correlated with high sea salt and non sea salt minerals- Elemental analysis of the high sea salt events shows higher Na+ content and contains Cl- (not shown)- The second dust event shows no Cl- as well as higher Ca2+ and Si loading- Wind trajectory modelling shows wind coming from the Gulf of Mexico during the �rst event could explain the high Cl- content

- Modelling also shows continental United States sources of mineral origin during the two events- High Si and Ca2+ loading during the second event leads us to believe the source is terrestrial in origin

Benjamin R. Ayres1, Hannah M. Allen1, Danielle C. Draper1, Robert Wild2, Steven S. Brown2, Abigail Koss3,4, Joost A. De Gouw3,4, Douglas A. Day3,4,Pedro Campuzano-Jost3,4, Brett B. Palm3,4, Weiwei Hu3,4, Jose L. Jimenez3,4, Kevin F. Olson5, Allen H. Goldstein5, Paul Romer5, Ronald C. Cohen5,

Karsten Baumann6, Eric Edgerton6, Benjamin Lee7, Claudia Mohr8, Joel Thornton7, Juliane L. Fry1

1Reed College, Portland, OR; 2NOAA-ESRL, Boulder, CO; 3Cooperative Institute for Research in Environmental Sciences, Boulder, CO; 4University of Colorado Boulder, Boulder, CO; 5University of California Berkeley, Berkeley, CA; 6Atmospheric Research and Analysis, Inc., 7University of Washington, Seattle, WA, 8Karlsruhe Institute of Technology, Karlsruhe, Germany

Figure 12 - PM and Ion Chromatography spectra of the SOAS campaign

Figure 13 - Back trajectories of wind over the SOAS campaign

Predicted NO3 + BVOC

Mellon Environmental Studies Summer Experience Fellowship and Reed College Undergraduate Research Opportunity GrantEPA # RD-83539901Christoph Knote for back trajectory mapsA special thank you Anne Marie Carlton, Jim Moore and all of the folks who made the SOAS campaign possible.

Comparison of Oxidants

- Sum of individual components of NOy and measured NOy match within 86%- Steady state predicted N2O5 and measured N2O5 correlate well; steady state NO3 was used in rate calculations due to ambient concentrations being below the detection limit of the instrument (Cavity Ring Down, RONALD)- NO3 + BVOC reactions account for almost 50% of NO3 loss during the day- Predicted NO3 + BVOC loss rates correlated with aerosol mass spectrometry (AMS) peak data provide a molar yield of 23% NO3 reaction to aerosol phase organic nitrate- Chemical Ionization Mass Spectrometry (CIMS) coupled to a FIGAERO analyzes m/z of gas and aerosol phases: - Correlations against predicted NO3 + isoprene show only gas phase C5H9NO5 production; no aerosol phase isoprene nitrate product is measured - Correlations against NO3 + monoterpene show that products with the formula C10H17NO5 partition to the aerosol phase - Inorganic NO3 analysis by ion chromatography shows dust events catalyze heterogeneous uptake of HNO3

6

5

4

3

2

1

0

Conc

entra

tion

(ppb

v)

6/1/13 6/6/13 6/11/13 6/16/13 6/21/13 6/26/13 7/1/13Central Daylight Time

NO NO2 ! Alkyl NO3 ! Peroxy NO3 HNO3 HONO NO3

-

NOy

35

34

33

32

31

30

Latit

ude

(o N)

-88 -87 -86 -85Longitude(oW)

CTR SOAS ground site

SO2 emissions

NOx emissions(sqrt tons per day, only sources above 50 stpd)

Alabama Power Company Gaston plant

Georgia Power Bowen plant

8

6

4

2

0

[N2O

5]M

easu

red

( pp t

v )

6/6/13 6/11/13 6/16/13 6/21/13 6/26/13 7/1/13Central Daylight Time

543210

[N2O

5]S

S (p

ptv)

5

0[N2O5]SS (pptv)

6:00 AM6/13/13

12:00 PM6/14/13

6:00 PM6/15/13

Central Daylight Time

5

0[N2O5]Measured (pptv)

- N2O5 and NO3 were measured using a cavity ring down spectrometer for nitrate radical (RONALD, NOAA ESRL)- Steady state [NO3]SS was calculated using NO3 production rates and loss rates of NO3 to BVOC, NO and photolysis- Steady state loss rates of [N2O5]SS was calculated using [NO3]SS, [NO2] and Keq NO3+ NO2

- Measured [N2O5]SS correlates well with [N2O5]SS values for the campaign (Figure 3)- [NO3] was consistently below detection limits during SOAS so values of [NO3]SS are used for all following calculations- BVOC analysis from cryostat gas chromatography (TACOH, CIRES) shows high concentrations of isoprene followed by α-pinene and β-pinene (Figure 4)- [NO3] loss to BVOC can be calculated from this data- [NO3]SS tracks ΣANs, most of which is in the aerosol phase as seen in TD-LIF (Figure 5)

6/1/13 6/11/13 6/21/13 7/1/13 7/11/13Central Daylight Time

12

10

8

6

4

2

0

[BVO

C] (

ppb)

Isoprene !-Pinene !-Pinene Limonene Camphene

0.40

0.35

0.30

0.25

0.20

0.15

0.10

0.05

NO3

loss

(s-1

)

222018161412108642Hour of Day

NO3 photolysis NO Isoprene !-pinene !-pinene Limonene Other terpenes N2O5 heterogeneous uptake

Photolysis !"pinene

!"pinene

Isoprene

Limonene Minor Terpenes N2O5

NO !"pinene

!"pinene

Isoprene Limonene Minor Terpenes

N2O5 NO

1.2

0.8

0.4

0.0

NO3

SS (p

pt)

7/1/13 7/6/13 7/11/13Central Daylight Time

0.8

0.6

0.4

0.2

0.0

TD-L

IF !

AN ( p

pbv)

Total !ANs Aerosol phase !ANs

Table 1. NO 3 Rate Coefficients used to determine losses

Reaction A-Factor E/R k(298 K)

nitrate loss to terpenes is calculated from the steady state nitrate and terpene concentrations and then

cumulatively summed over time periods of aerosol buildup. These are correlated and show good

agreement between(NO 3, loss )pred (for both isoprene and monoterpene) and aerosol mass spectrum

maxima. Isoprene’s (NO 3, loss )pred correlation was found to be spurious as the concentrations are

very low, but correlate well with monoterpene. The main structure found in the aerosol’s makeup260

as measured by CIMS is C 10H17NO 5 and also correlates well to cumulative(NO 3, loss )pred when

compared to particle phase data.

Acknowledgements.We would like to acknowledge Anne Marie Carlton, Jim Moore, Karsten Baumann and all

of the colleagues that helped to set up this study. We thank Metrohm for the use of the MARGA system. We

would also like to acknowledge the National Center for Environmental Research (NCER) STAR Program, EPA265

#RD-83539901.

Scheme 1Generalized reaction fate forNO 2 in the troposphere. Oxidation ofNO 2 from atmospheric

oxidants leads to two paths.

9

Product

(molecules cm-3

s-1)

O3 + NO2 O2 + NO3

NO2 + NO3 N2O5

NO + NO3 2NO2

Isoprene + NO3 Productα-pinene + NO3 Productβ-pinene + NO3 Product

Camphene + NO3 ProductMyrcene + NO3 Product

Limonene + NO3 Product

1.2 x 10-13

2.7 x 10-27

1.5 x 10-11

3.03 x 10-12

1.19 x 10-12

245011000-170446-490

2.51 x 10-12

6.6 x 10-13

1.1 x 10-11

1.22 x 10-11

- Rate of NO3- uptake is driven by PM2.5 via heterogeneous uptake of HNO3 onto the surface of mineral aerosols

- An average value of 10% uptake (γHNO3

) for HNO3 was used to calculate the rate of uptake- Rate = γHNO3

SA vHNO3

[HNO3]- Organic and Inorganic rates are comparable in magnitude with di�erent peak times over the SOAS campaign (Figure 14)- Inorganic NO3 magnitudes are also comparable to alkyl nitrate magnitudes over the SOAS campaign (Figure 15)- Organic rates calculated using (NO3, loss)pred and time (see Predicted NO3 + BVOC)

14

Heterogeneous uptake of HNO3 on Dust

Table 1. NO 3 Rate Coefficients used to determine losses

Dust Event I Dust Event IIElement Composition (%) Si Ratio Composition (%) Si Ratio

nitrate loss to terpenes is calculated from the steady state nitrate and terpene concentrations and then

cumulatively summed over time periods of aerosol buildup. These are correlated and show good

agreement between(NO 3, loss )pred (for both isoprene and monoterpene) and aerosol mass spectrum

maxima. Isoprene’s (NO 3, loss )pred correlation was found to be spurious as the concentrations are

260

as measured by CIMS is C 10H17NO 5 and also correlates well to cumulative(NO 3, loss )pred when

compared to particle phase data.

Acknowledgements.We would like to acknowledge Anne Marie Carlton, Jim Moore, Karsten Baumann and all

of the colleagues that helped to set up this study. We thank Metrohm for the use of the MARGA system. We

would also like to acknowledge the National Center for Environmental Research (NCER) STAR Program, EPA265

#RD-83539901.

Scheme 1Generalized reaction fate forNO 2 in the troposphere. Oxidation ofNO 2 from atmospheric

oxidants leads to two paths.

9

Product

AlSiK

CaTi

MnFeNaMg

26.841.85.12.41.10.29.3

12.01.4

0.641

0.120.060.03

<0.010.220.310.02

26.245.25.53.31.10.19.47.51.8

0.591

0.110.070.02

<0.010.210.180.04

Figure 1 - Regional Map. SOAS site is denoted by + Figure 2 - NOy comparison. Stacked spectra shows that the sum of individual NOy components match total NOy measured

Figure 3 - Comparison of steady state and measured N2O5 correlate well

Figure 5 - ΣAN spectra show that almost all alkyl nitrates are in aerosol phase and steady state nitrate tracks well with alkyl nitrate production

Figure 4 - VOC concentrations over the SOAS campaign

Figure 6 - Diurnal averaged NO3 loss for June 1 - July 15. Pie graphs depict amount of loss and are sized by loading

Figure 7 - AMS data overlaid with predicted cumulative NO3 loss. Black dots denote start and stop of predicted buildup of NO3 loss

NO2

NO3

HNO3

BVOC

aerosol surface

organonitrates

inorganic nitrate aerosol

organonitrate aerosol

Inorganic and Organic NO3 fate

What we learned

SOAS Site

Figure 8 - AMS organonitrate is strongly correlated with predicted NO3 loss to monoterpenes

Figure 9 - Gas phase CIMS product C5H9NO5 correlate with with predicted NO3 loss to isoprene

Figure 10 - Individual CIMS m/z peak correlations show C10H17NO5 is the only well correlated product to NO3 loss to monoterpenes

Figure 11 - Oxidant reaction rates with monoterpenes.

Figure 14 - Loss rates of inorganic nitrate to dust via HNO3 and nitrate radical to BVOC

Figure 15 - Diurnally averaged inorganic nitrate concentration (IC) compared to alkyl nitrate concentrations (TD-LIF) show similar average magnitudes

- Central Alabama is a humid subtropical environment with an average daily high temperature of 28.4 oC- Emissions of SO2 and NOx come from regional power plants in Alabama and Georgia. The Gaston Plant emits 19.5 square tons of SO2 and 6.5 square tons of NOx per day (Figure 1)- Σ Alkyl NO3& Σ Peroxy NO3 measured by Thermal Dissociation - Laser Induced Fluorescence, NOx by cavity ring down spectroscopy and HNO3, HONO & NO3 by Monitor of AeRosol and Gas Analyzer (MARGA)- The individual NOy components show that NO2, sum alkyl nitrates and sum peroxyalkyl nitrates are the main contributors to NOy (Figure 2)- Organonitrates make a substantial contribution to the NOy budget (Figure 2)- Relatively constant contributions of NOy come from HNO3 and HONO

SOASSouthern

Oxidant &Aerosol

Study

no /n - reed college€¦ · mellon environmental studies summer experience fellowship and reed...

Documents