optical properties of moderately absorbing organic and

Optical Properties of Moderately Absorbing Organic and Mixed Organic/Inorganic

Aerosols at Relative Humidities up to 95%

Benjamin Brem (Measurements) Francisco Mena (Closure Modeling)

Tami Bond and Mark Rood (PIs)

DoE Grant No. DE-FG02-08ER64533

Department of Civil and Environmental Engineering University of Illinois at Urbana-Champaign

3rd ASR Science Team Meeting, Crystal City, VA

March 14, 2012

Relative Humidity Effects on Aerosol Optical Properties

• Aerosol water uptake and aerosol size = function of aerosol chemistry and relative humidity (RH)

• Water uptake alters light scattering and light absorption

• RH effects on scattering previously characterized (RH < 90%)

• Limited understanding of RH effects on absorption 2

Primary Light Absorbing Organic Carbon (LAOC) Aerosols

• Emission Sources of Primary OC: – 91% from open and residential biomass burning (Bond et al., 2004)

• Chemical Characteristics: – Complex mixture of polycyclic aromatic hydrocarbons (PAHs),

humic-like substances and biopolymers (Pöschl , 2003)

– Up to 70% water soluble by mass (Chen and Bond, 2010)

• Optical Properties: – Weakly absorbing at visible wavelengths

– Wavelength dependent absorption with strongest absorption towards blue\ ultraviolet

– Few in situ measurements at multiple wavelengths and variable RH

3

LAOC on Filter

Research Motivations & Objectives

Instrumentation challenges limit light absorption measurement as a function of RH • Objective 1: Build and benchmark equipment to measure aerosol

absorption and scattering as a function of RH up to 95% in a laboratory setting (presented at previous ASR meetings)

4

Research Motivations & Objectives

Instrumentation challenges limit light absorption measurement as a function of RH • Objective 1: Build and benchmark equipment to measure aerosol

absorption and scattering as a function of RH up to 95% in a laboratory setting (presented at previous ASR meetings)

Limited knowledge of LAOC absorption and its response to RH • Objective 2: Measure absorption and scattering of LAOC as a function of

RH and evaluate optical closure (main focus of this presentation)

4

Research Motivations & Objectives

Instrumentation challenges limit light absorption measurement as a function of RH • Objective 1: Build and benchmark equipment to measure aerosol

absorption and scattering as a function of RH up to 95% in a laboratory setting (presented at previous ASR meetings)

Limited knowledge of LAOC absorption and its response to RH • Objective 2: Measure absorption and scattering of LAOC as a function of

RH and evaluate optical closure (main focus of this presentation)

Limited understanding of optical properties as a function of RH for internal LAOC/ inorganic solute mixtures • Objective 3: Perform closure study for LAOC mixed with inorganic solute

as a function of RH (in the works)

4

Approach Overview

5

LAOC generated by pyrolysis of wood @ 425°C and its mixtures with NaCl

Nephelometer Extinction Cell

H-TDMA SMPS Offline Analysis

Extinction Dry/Wet

Scattering Dry/Wet

Absorption Dry/Wet

Growth Factor

PSD Dry

Ionic Fraction

Approach Overview

5

LAOC generated by pyrolysis of wood @ 425°C and its mixtures with NaCl

Nephelometer Extinction Cell

H-TDMA SMPS Offline Analysis

Extinction Dry/Wet

Scattering Dry/Wet

Absorption Dry/Wet

Growth Factor

PSD Dry

Ionic Fraction

Optical (Mie) Model

Optical (Mie) Modeling Inputs

6

• Dry size distribution & hygroscopic size growth factors humidified size distribution and aerosol water content

• Dry LAOC refractive index: – 1.57+0.017i (467nm), 1.57+0.01i (530nm), 1.57+0.002i (660nm)

n from literature, ki from spectroscopy from filter extracts

• Humidified refractive index mixing models:

– Linear Volume Average (LVA) : commonly used for OC, infers uniform mixture (completely soluble LAOC)

– Dynamic Effective Medium Approximation (DEMA) Infers non-uniform, composite media with randomly distributed insoluble inclusions of absorbing material (Chylek, 1984)

Increasing RH

Increasing RH

Measured LAOC Optical Properties

7

40 50 60 70 80 90 1001000

1250

1500

1750

2000

40 50 60 70 80 90 1001000

1250

1500

1750

2000

40 50 60 70 80 90 100-50

0

50

100

150

200

250

40 50 60 70 80 90 100

0.90

0.95

1.00

1.05

Ex

tinct

ion

b ep(M

m-1)

RH (%)

Measured

467, 530, 660 λ (nm)

Scat

terin

g b sp

(Mm

-1)

RH (%)

Abso

rptio

n b ap

(Mm

-1)

RH (%)

Sing

le S

catte

ring

Albe

do ω

RH (%)

Error bars dotted for clarity Error bars dotted for clarity

FINDINGS: • Absorption increases by a factor of 2 at High RH (λ = 467 and 530 nm) • Absorption at λ = 660 nm is below detection limit of instrumentation

40 50 60 70 80 90 1001000

1250

1500

1750

2000

40 50 60 70 80 90 1001000

1250

1500

1750

2000

40 50 60 70 80 90 100

0

50

100

150

200

250

40 50 60 70 80 90 100

0.88

0.92

0.96

1.00

1.04

Ex

tinct

ion

b ep(M

m-1)

RH (%)

Measured

LVA Model

467, 530, 660 λ (nm)

Scat

terin

g b sp

(Mm

-1)

RH (%)

Abso

rptio

n b ap

(Mm

-1)

RH (%)

Sing

le S

catte

ring

Albe

do ω

RH (%)

Optical Closure Results (1): Linear Volume Average (LVA) RI Mixing

8

Error bars removed for clarity Error bars removed for clarity

FINDING: LVA predicts extinction and scattering, but not absorption

Optical Closure Results (2) : Dynamic Effective Medium Approximation (DEMA) RI Mixing

9

40 50 60 70 80 90 1001000

1250

1500

1750

2000

40 50 60 70 80 90 1001000

1250

1500

1750

2000

40 50 60 70 80 90 100-50

0

50

100

150

200

250

40 50 60 70 80 90 100

0.88

0.92

0.96

1.00

1.04

Ex

tinct

ion

b ep(M

m-1)

RH (%)

Measured

LVA Model

DEMA Model

467, 530, 660 λ (nm)

Scat

terin

g b sp

(Mm

-1)

RH (%)

Abso

rptio

n b ap

(Mm

-1)

RH (%)

Sing

le S

catte

ring

Albe

do ω

RH (%)

Error bars removed for clarity Error bars removed for clarity

FINDING: DEMA captures trends in absorption and SSA indicating that configuration of insoluble absorbing material might be responsible for absorption enhancement

30 40 50 60 70 80 90 100

1.0

1.1

1.2

1.3

1.4

1.5

30 40 50 60 70 80 90 100

1.0

1.1

1.2

1.3

1.4

1.5

30 40 50 60 70 80 90 100

1.0

1.5

2.0

2.5

30 40 50 60 70 80 90 100

0.88

0.92

0.96

Ext

inct

ion

Gro

wth

Fac

tor f

(RH

) bep

λ = 467 nm Experiment 1 Experiment 2 Experiment 3 Experiment 4 LVA DEMA

RH (%)

Sca

tterin

g G

row

th F

acto

r f(R

H) b

sp

RH (%)

Abs

orpt

ion

Gro

wth

Fac

tor f

(RH

) bap

RH (%)

Sing

le S

catte

ring

Albe

do ω

RH (%)

f RH bsp =bsp RH

bsp(RH<45%)

Normalized LAOC Optical Properties for Multiple Experiments (λ = 467nm)

10

FINDINGS: • Low Variability in f(RH) ext. and scat. particles have same hygroscopicity • Same trend in absorption growth and SSA reduction, represented best with DEMA

11

Normalized LAOC Optical Properties for LAOC- NaCl Mixtures (λ = 467nm)

20 30 40 50 60 70 80 90 100

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

20 30 40 50 60 70 80 90 100

1.0

1.2

1.4

1.6

1.8

2.0

2.2

2.4

20 30 40 50 60 70 80 90 100

1.0

1.5

2.0

2.5

20 30 40 50 60 70 80 90 1000.88

0.90

0.92

0.94

0.96

Ext

inct

ion

Gro

wth

Fac

tor f

(RH

)bep

λ = 467nm 17% vol. NaCl 23% vol. NaCl 27% vol. NaCl

RH (%)

Sca

tterin

g G

row

th F

acto

r f(R

H)b

sp

RH (%)

Abso

rptio

n G

row

th F

acto

r f(R

H)b

ap

RH (%)

Sing

le S

catte

ring

Albe

do ω

RH (%)

FINDINGS • Observable deliquescence ~ 76% RH • Absorption increase is in the same range as for pure LAOC

Summary & Outlook • LAOC absorption↑ and SSA↓ with RH↑ at the 467 and 530

nm wavelengths. (660 nm below detection limit)

• Change in LAOC absorption and SSA is currently not implemented in models

• Widely used LVA RI mixing rule is unable to catch trends in absorption and SSA with RH

• DEMA captures trends in absorption and SSA indicating that configuration of insoluble absorbing material might be responsible for absorption enhancement

• Addition of NaCl affects scattering but no specific trend (in comparison to pure LAOC) was observed for absorption

12

QUESTIONS? CONTACT INFORMATION: BBREM@ILLINOIS.EDU WWW.HIWATER.ORG HTTP://AQES.CEE.ILLINOIS.EDU/

Thank you for your attention

14

Acknowledgements

• Bond / Rood Research Groups • DoE-ASP Grant No. DE-FG02-08ER64533 • DoE-ASR Grant No. DE-SC0006689

(current ASR support)

13

Appendix: Normalized LAOC Optical Properties for Multiple Experiments (λ = 530nm)

18

30 40 50 60 70 80 90 100

1.0

1.1

1.2

1.3

1.4

1.5

30 40 50 60 70 80 90 100

1.0

1.1

1.2

1.3

1.4

1.5

30 40 50 60 70 80 90 100

1.0

1.5

2.0

2.5

30 40 50 60 70 80 90 1000.92

0.94

0.96

0.98

1.00

f(R

H)b

ep

a) f(RH)bep 530nm 11/22/2010 12/03/2010 12/04/2010 11/11/2011 LVA DEMA

RH (%)

f(R

H)b

sp

b) f(RH)bsp 530nm

RH (%)

f(R

H)b

ap

c) f(RH)bap 530nm

RH (%)

ω

d) ω 530nm

RH (%)

Appendix: Normalized LAOC Optical Properties for Multiple Experiments (λ = 660nm)

19

30 40 50 60 70 80 90 100

1.0

1.1

1.2

1.3

1.4

1.5

30 40 50 60 70 80 90 100

1.0

1.1

1.2

1.3

1.4

1.5

30 40 50 60 70 80 90 100

-6-4-202468

10

30 40 50 60 70 80 90 100

0.96

0.98

1.00

1.02

f(R

H)b

ep

a) f(RH)bep 660nm 11/22/2010 12/03/2010 12/04/2010 11/11/2011 LVA DEMA

RH (%)

f(R

H)b

sp

b) f(RH)bsp 660nm

RH (%)

f(R

H)b

ap

c) f(RH)bap 660nm

RH (%)

ω

d) ω 660nm

RH (%)

Appendix: Methods: Aerosol Generation and Treatment

• LAOC aerosol was generated by pyrolyzing wood in an electrically-heated combustor at 425 ± 10 ºC

• Flaming (formation of BC) was avoided by using nitrogen sheath flow

• Aerosol was sampled at a 10:1 dilution into a 220 L storage chamber

• After pyrolysis event (~4-6 min), chamber was disconnected from sampling system and connected to optical instrumentation

• For mixing experiments inorganic solute was atomize into diluent

20

Appendix: Instrumentation • Optical Properties:

– bep Short path extinction cell (SPEC) λ= 467, 530 and 660 nm – bsp λ and temperature modified nephelometer – bap Difference between bep and bsp

• Physical Properties: – Dp SMPS (TSI) – Gf(Dp) H-TDMA

• Environmental Variables – RH / dew point temperature 2 RH (Vaisala), 2 Dew point (GE) – Temperature 4 Thermocouples

• Offline Filter Measurements (Chemical Properties) – OC/EC, Ions, OC functional groups 1H and 13C NMR

21

Appendix: Instrumentation

22

Humidifier

Cooler

Dew Point Meters

Extinction Cell

Nephelometer

RH Control Box

Impactor

9

Appendix: Benchmarking Ammonium Sulfate System Performance under purely Light Scattering Conditions

24

• Instrumentation capable of measuring optical properties up to 95% RH

• Measured data lie within uncertainty of model

• Shift in bsp deliquescence due to nephelometer heating of 0.5°C

Tang (1994)

Appendix PSL Results

25

Benchmarking with Nigrosin System Capable of Measuring Absorption as a Function of RH

• bap enhanced by a factor of 1.2 between 35 and 95% RH

• bap λ dependence similar to bulk measurements

• Closure modeling Poster Mena et al. Number: 2E8

26

Appendix LAOC PSD

27

Appendix LAOC Composition

28

Table 1 Composition analysis of OC aerosol generated by the pyrolysis of oak wood at 425°C. Elemental carbon (EC) and the ions of Br-, PO4

3- and Mg2+ were also determined but the results were below the detection limit of the analysis OM* OM*/ OC Cl- SO4

2- NO3- NH4

+ Ca2+ K+ Na+

Mean 97.25 1.68 0.74 0.28 0.21 0.07 0.21 0.87 0.17 Std. 1.13 0.06 0.23 0.04 0.20 0.08 0.15 0.25 0.06 *Corrected by forcing closure with mass analyzed by gravimetric analysis (OM/ OC describes the correction factor).

LAOC HTDMA Results

29

κ = 0.08 ± 0.015

Appendix: LAOC NMR Results

30

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

O=C-(R, H)

O=C-(OH/NH2)

C=C-(O,N)

C=C-(C,H)

O-C-O

C-O or C-N

C-C

High T water

High T methanol

0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70

O=C-(R, H)

O=C-(OH/NH2)

C=C-(O,N)

C=C-(C,H)

O-C-O

C-O or C-N

C-C

Low T water

Low T methanol

Percentage peak area

Percentage of total 13C NMR peak area under each spectral region for methanol and water extracts of OC generated at two 300 and 425C. Difference of contribution from two major functional groups is highlighted with green circles.

Appendix Global Relative Humidity (RH) Conditions

31

Observation-derived average global surface RH for the years between January 1960 and January 2011. Image provided by NOAA-ESRL, Boulder, Colorado

32

Appendix: Wood combustion process

Evans and Milne, 1986

Appendix Dry Data Processing

33

Appendix: Humidification System Details & Function

34

Dry Inflow with Aerosols

Humidified Outflow

HeatingTape Gore- Tex®

Membrane

RH Sensor

Controller

CirculatingWater

Appendix: RH Sensor Performance

V1, V2 = Capacitance based Vaisala Sensors DP1, DP2 = General Eastern Dew Point Meters (RH calculated with dry bulb temperatures )

35

32

33

34

35

36a) Setpoint = 34%

RH [%

]

48

49

50

51

52b) Setpoint = 50%

RH [%

]

V1 V2 DP1

DP2

88

89

90

91

92c) Setpoint = 90%

RH [%

]

V1 V2 DP1

DP2

96

97

98

99d) Setpoint = 97.5%

RH [%

]

Appendix: Modified Nephelometer Performance

36

• For all three wavelengths, the instruments differ less than 1.5%

Appendix: Extinction Cell and Nephelometer Sensitivity and Detection Limit

• Function of integration time • Best sensitivity is btw. 100

and 300s integration time • Corresponding absorption

detection limits: 57.3 Mm-1 (467 nm), 54.5 Mm-1 (530 nm) and 105 Mm-1 (660 nm) with a signal to noise ratio is assumed to be 3

37

Appendix POA Sources

38

Residential Biofuel

17% Residenti

al Coal 1%

Residential Other

0%

Transport: Non-road

1% Transport:

Road 4%

Industry 1%

Electricity Generatio

n 0%

Open vegetative

burning 74%

Other 2%

Total: 34 Tg/ yr (2000)