Heterogeneous chemistry and its potential impact on climate Jingqiu Mao (Princeton/GFDL) North Carolina State University, 11/25/2013.

Heterogeneous chemistry and its potential impact on climate

Jingqiu Mao (Princeton/GFDL)

North Carolina State University, 11/25/2013

Acknowledgement

• Songmiao Fan (GFDL)

• Daniel Jacob (Harvard)

• Larry Horowitz (GFDL)

• Vaishali Naik (GFDL)

Outline

1. A missing sink for radicals

2. Implications for radiative forcing from biomass burning

3. Aerosol Fe speciation sustained by gas-phase HO2

O3

O2

O3

OH HO2

hn, H2O

Deposition

NO

H2O2

CH4, CO, VOCs

NO2

STRATOSPHERE

TROPOSPHERE

8-18 km

hn

hn

hn

H2O2 is a radical reservoir.

(Levy, Science, 1971)

Models ONLY underestimate CO in Northern extratropics

(Shindell et al., JGR, 2006)

Cannot be explained by emissions:

Need to double current CO anthro emissions (Kopacz et al., ACP, 2010).

MOPITT satellite(500 hPa)

Multi-model mean (500 hPa)20-90 N

20 S – 20 N

20 – 90 S

Annual cycle of CO

The alternative explanation is that model OH is wrong, but how?

All models have more OH in NH than SH (N/S > 1) Obs-derived estimates show the opposite (N/S < 1), with 15-30% uncertainties

(Naik et al., ACP, 2013)

Obs-derivedModels

Present Day OH Inter-hemispheric (N/S) ratio

O3

OH HO2

hn, H2O

Deposition

NO

H2O2

CH4, CO, VOC

NO2

Clouds/Aerosolshn

Uniqueness of HO2 in heterogeneous chemistry:• lifetime long enough for het chem (~ 1-10 min vs ~1 s for OH).• high polarity in its molecular structure (very soluble compared to

OH/CH3O2/NO/NO2).• very reactive in aqueous phase (superoxide, a major reason for DNA

damage and cancer).

Gas: L[HO2] ~ [HO2]∙ [HO2]Uptake: L[HO2] ~ [HO2]

Aerosol uptake is only significant when gas-phase [HO2] is relatively low.

Gas phase HO2 uptake by particles

HO2

aerosol

HO2(aq)

NH4+

NH4+

NH4+

NH4+

SO42-

SO42-

SO42-

SO42-

HSO4-

HSO4-

HSO4-

Aqueous reactions

NH4+

HSO4-

④① ② ③

γ(HO2) defined as the fraction of HO2 collisions with aerosol surfaces resulting in reaction.

① ② ③ ④

Laboratory measured γ(HO2) on sulfate aerosols are generally low…

Except when they add copper in aerosols…

Cu-dopedAqueousSolid

(Mao et al., ACP, 2010)

HO2(aq)+O2-(aq)→ H2O2 (aq)

Cu(II) Cu(I)

HO2(g) H2O2(g)

Conventional HO2 uptake:HO2 → H2O2(g)

Arctic Research of the Composition of the Troposphere from Aircraft and Satellites (ARCTAS) Phase I: April 1st ~ April 20th,2008

ARCTAS-A DC-8 flight track

Conventional HO2 uptake does not work over Arctic!

(Mao et al., ACP, 2010)

Joint measurement of HO2 and H2O2 suggest that HO2 uptake by aerosols may in fact not produce H2O2 !

Median vertical profiles in Arctic spring (observations vs. GEOS-Chem model)

We hypothesized a bisulfate reaction to explain this:

But it is not catalytic and thereby inefficient to convert HO2 radical to water. There must be something else …

I took this picture

Cu is one of 47 transitional metals in periodic table…

Trace metals in urban aerosols (Heal et al., AE, 2005)

Transition metals have two or more oxidation states:

Fe(II) Fe(III)

Cu(I) Cu(II)

- e

+ e- e

+ e

reduction(+e) + oxidation(-e) = redox

Cu and Fe are ubiquitous in crustal and combustion aerosols

Cu/Fe ratio is between 0.01-0.1

IMPROVE

Cu is fully dissolved in aerosols.

Fe solubility is 80% in combustion aerosols, but much less in dust.

Cu(II) + HO2 → Cu(I) + O2 + H+

Cu(I) + HO2 Cu(II) + H2O2

What we thought was happening in aerosols…

As Fe(III) + HO2 is 300 times slower than Cu(II) + HO2, so we thought Fe was unimportant…

Net: HO2 +HO2 → H2O2 + O2

Cu(II) + HO2 → Cu(I) + O2 + H+

Cu(I) + HO2 Cu(II) + H2O2

What we thought was happening in aerosols…

As Fe(III) + HO2 is 300 times slower than Cu(II) + HO2, so we thought Fe was unimportant…

But we missed one electron transfer reaction (very fast)

Cu(I) + Fe(III) → Cu(II) + Fe(II)

Net: HO2 +HO2 → H2O2 + O2

Cu(II) + HO2 → Cu(I) + O2 + H+

Cu(I) + HO2 Cu(II) + H2O2

What we thought was happening in aerosols…

As Fe(III) + HO2 is 300 times slower than Cu(II) + HO2, so we thought Fe was unimportant…

But we missed one electron transfer reaction (very fast)

Cu(I) + Fe(III) → Cu(II) + Fe(II)

Fe(II) + HO2 Fe(III) + H2O2

With three reactions to close the cycle…

Fe(II) + H2O2 → Fe(III) + OH + OH−

Fe(II) + OH → Fe(III) + OH−

The product from HO2 uptake depends on the fate of Fe(II).

Net: HO2 +HO2 → H2O2 + O2

Net: HO2 + H2O2 → OH + O2 + H2O

Net: HO2 +HO2 → H2O2 + O2

Net: HO2 + OH → O2 + H2O

Cu-Fe redox coupling in aqueous aerosols

Cu only: HO2 → H2O2

Cu + Fe : HO2 → H2O or H2O2

and may also catalytically consume H2O2.

Conversion of HO2 to H2O is much more efficient as a radical loss.

In gas phase, H2O2 can photolyze to regenerate OH and HO2.

(Mao et al., 2013, ACP)

Modeling framework for HO2 aerosol uptake

HO2

aerosol

[HO2]surf

2

1)4

( HOg

in AnvD

aR

*

21][

)4

(H

HOA

vD

aR surf

gout

Rin

[HO2]surf

[HO2]bulk

outinbulk RR

dt

HOd

][ 2

2HOn Rout

[HO2]surf is higher than [HO2]bulk because of its short lifetime.

0][)][

(1

2222

2 HO

Iaq PHOk

dt

HOdr

dr

d

rD

provides a relationship between [HO2]surf

and [HO2]bulk.

The diffusion equation with chemical loss (kI[HO2]) and production (PHO2)

Aqueous chemistry include Cu, Fe, Cu-Fe coupling, odd hydrogen and photolysis.

Uptake rate

Volatilization rate

Chemical loss rate

Ionic strength correction for aerosol aqueous chemistry

Non-ideal behavior due to the electrostatic interactions between the ions.

1. Use Aerosol Inorganic Model (AIM) to calculate the ionic strength and activity coefficients for major ions (i.e. NH4

+, H+, HSO4-, SO4

2-).2. Calculate activity coefficients for trace metal ions and neutral

species based on specific ion interaction theory.3. Account for salting-out effect on Henry’s law constant.

iii cAa Ai is activity coefficient for any species and also a function of ionic strength. + -

+-

Ideal solution(cloud droplets)

Non-ideal solution (aqueous aerosol)

+

++ +

++ --

---

--- --

--- -

- --

-

----

---

--

Chemical budget for NH4HSO4 aerosols at RH=85%, T=298 KCu/Fe = 0.05, HO2(g) = 10 pptv, H2O2(g) = 1 ppb

70% of HO2 gas uptake is lost in aerosols (γ(HO2) = 0.7) no H2O2 is net produced. Fe(III) reduction is dominated by Fe(III) + Cu(I), instead of

photoreduction (implications for Fe speciation)

γ(HO2) dependence on aerosol pH and Cu concentrations

• γ(HO2) is high at typical rural conditions (0.4-1 at 298 K), even higher at low T.

• Effective γ(HO2) can be higher than 1, due to the reactive uptake of H2O2.

• γ(HO2) uptake is still higher than 0.1 when Cu is diluted by a factor of 10.

Cu/Fe=0.1

Cu/Fe=0.01typical rural site

(Mao et al., 2013, ACP)

Test this mechanism in two global models

GFDL AM3 chemistry-climate model (nudge)GEOS-Chem chemical transport model

In both models, we assume γ(HO2) = 1 producing H2O for all aerosol surfaces (based on effective radius and hygroscopic growth).

number

area

volume

Aerosol surface area is mainly contributed by submicron aerosols (sulfate, organic carbon, black carbon)

Typical aerosol distribution

Improvement on modeled CO in Northern extratropicsBlack: NOAA GMD Observations at remote surface sites Green: GEOS-Chem with (γ(HO2) = 1 producing H2O) Red: GEOS-Chem with (γ(HO2) = 0)

(Mao et al., 2013, ACP)

CO at 500 hPa

AM3 with het chem off

MOPITT

AM3 with het chem on

MOPITT (2000-2004) AM3(2001-2005)

OH ratio (NH/SH)

(Mao et al., 2013, GRL)

Improvement in AM3 model

Conclusions• The product of HO2 uptake is likely to be H2O, not the

radical reservoir H2O2.

• γ(HO2) is somewhere between 0.1 and >1.0. This remains largely uncertain.

• We find that the model results are largely improved when γ(HO2) set to 1 (both GEOS-Chem and AM3).

• Further experimental work is needed, particularly at low T (< room temperature 298 K).

Outline

1. A missing sink for radicals

2. Implications for radiative forcing from biomass burning

3. Aerosol Fe speciation sustained by gas-phase HO2

The impact of biomass burning emissions on oxidants and radiative forcing

IPCC AR4 only estimates the direct forcing from biomass burning aerosols (+0.03 ±0.12 W m-2).

Cooling or warming?

Perturbation tests of biomass burning emissions on global OH and ozone

Computed change of global mean OH is 6.3% for doubling 2000 bb emissions.

(Prinn et al., 2005).

1997 Indonesian fires, 6%

AM3 model with different magnitude of biomass burning emissions (for year 2000).

Estimated global OH from CH3CCl3

(Mao et al., 2013, GRL)

(Mao et al., 2013, GRL)

Model suggests a net warming effect when bb emissions increased by more than a factor of 2.

Outline

1. A missing sink for radicals

2. Implications for radiative forcing from biomass burning

3. Aerosol Fe speciation sustained by HO2 uptake

We want to test our model for Cu-Fe-HOx chemistry.

A dominant source of nutrient iron to open ocean, critical for plankton in surface waters.

Oxidative stress and health impact of ambient aerosols

Why do we care about aerosol Fe speciation?

“Give me a half a tanker of iron and I'll give you the next ice age”- John Martin

Phytoplankton blooms in the South Atlantic Ocean. (MODIS)

Ocean Fe is mainly supplied by dust (95%)

Crystal structure of hematite

Fe(II)

solubilities ~ 0.1%

Solubilization of dust Fe by atmospheric processing

(Shi et al., 2012)

Solubilization of dust Fe

Fe(III)= Fe3+ + Fe(OH)2+ + Fe(OH)2+ + Fe(SO4)+ + …

Fe(II)= Fe2+ + Fe(OH)+ + Fe(SO4) + …Fe(II) is more bioavailable

What do we know about Fe redox chemistry?

Fe(II) + HO2 Fe(III) + H2O2

Fe(II) + H2O2 → Fe(III) + OH + OH−

Fe(II) + OH → Fe(III) + OH−

Fe(II) Fe(III)

???Fe(III) + H2O2 /HO2 are too slow to be important

Fe(III)= Fe3+ + Fe(OH)2+ + Fe(OH)2+ + Fe(SO4)+ + Fe(C2O4) + …

Fe(II)= Fe2+ + Fe(OH)+ + Fe(SO4) + Fe(C2O4) 0 …

Current mechanisms for Fe(III)→Fe(II)

(Zhuang et al., 1992, Nature)

(1) Enhanced photolysis of Fe(III) by cloud processing

Cloud: pH~4Aerosols: pH<3

Cannot maintain the steady state of Fe(II)/Fe(III) after clouds evaporate.

Most time they are still in aerosol form.

cloudaerosol

?Fe(OH)2+ + hv

(2) Enhanced photolysis by organic acids

(photolysis rate ~ 10-2 s-1)

(Zuo and Hoigné, 1992, Johnson et al., 2013)

Limitation: need continuous supply of oxalic acid in aerosols (still unidentified yet).

Current mechanisms for Fe(III)→Fe(II)

Fe2+ + CO2

Fe(II) + H2O2 → Fe(III) + OH + OH−

Lifetime of Fe(II)< 1hr for 1ppb H2O2

(Zhu et al., 1997)

N-nighttime D-daytime

Current mechanisms cannot explain nighttime Fe(II) measurements!!

Significant amount of nighttime Fe(II) found in marine boundary layer!

A new driver for aqueous Fe(II) production

Fe(II) is sustained by gas-phase HO2!!!!

Diurnal cycle of HO2 over remote ocean

(Kanaya et al., 2000)

HO2 >0

Nighttime Fe(II) can be supplied by nighttime HO2

Nighttime HO2 can be produced from O3/NO3 + VOCs.

Future measurements to test such mechanism

][

][

][][][

][

)]([)]([

)]([

522423

21

total

total

Fe

Cu

OHkOHkHOk

HOk

IIIFeIIFe

IIFe

This mechanism can be easily tested by concurrent measurements of Cu and Fe in aerosols.

Extra slides

Solubilization of dust Fe by atmospheric processing

Soil has low Fe solubilities ~ 0.1% Solubilities of aerosol Fe in remote regions: up to 80%

Fine aerosols (<2.5 µm) tendto yield larger iron solubilities than coarse aerosols (Siefert et al., 1999; Baker et al., 2006)

(Baker et al., 2006)

Aerosol mass

Sol

ubili

ty

Sensitivity of tropospheric oxidants to biomass burning emissions

Global OH decreases with larger bb emissions.

Global ozone increases with larger bb emissions.

Other applications for aerosol TMI chemistry driven by HO2 uptake

• A major aqueous OH source (converted from gas-phase HO2 and H2O2), critical for SOA formation aerosol aging (O/C ratio).

• Oxidative stress and health (sustain soluble form of transitional metals in aerosols).

Aerosol uptake has large impact on ozone production efficiency

ΔO3/ ΔCO is a measure of ozone production efficiency.

Observations based on Jaffe et al. (2012)

(Mao et al., 2013, GRL)

Fe(II)/Fe ratio modulated by gas-phase HO2 concentrations

Field measurements of Fe(II)/Fe_total in MBL

Higher Cu/Fe ratio leads to higher Fe(II)/Fe_total

Fe(

II)/

Fe_

tota

l

What else in dust aerosols?Measurements from dust aerosols

There are tens of transitional metals in dust aerosols. We don’t know chemical kinetics for most of them. (Sun et al., 2005)

We only explored two transitional metals here…

Manganese (Mn)Chromium (Cr) ?Cobalt (Co) ?Vanadium (V) ?Zinc (Zn)?Titanium (Ti)??They may be all redox-coupled !

The theory is well established… For contributions on electron transfer reactions between metal complexes.

Rudolph A. MarcusNobel Prize in 1992

Henry TaubeNobel Prize in 1983