On or in?Where does atmospheric ice chemistry occur?

Tara Kahan Sumi Wren

On or in?Where does atmospheric ice chemistry occur?

Thanks to:

Ran Zhao

Klaudia Jumaa

Nana Kwamena

Funding:NSERCCFCAS

(Uncomfortable) conclusions

• Reactions within the ice matrix seem to be accurately modeled using aqueous parameters (if the correct concentrations are known)

• Photolysis at the air-ice surface shows different kinetics from that within the ice matrix

• Heterogeneous reaction kinetics at the air-ice interface may be quite different (or not!!) from those in the ice matrix

• Exclusion of salts to the air-ice interface may be different from bulk thermochemical predictions

• The air-ice interface presents a very different solvating environment from the liquid-air interface

I will discuss three kinds of experiment

• Photolysis experiments using ice samples• Bimolecular reactions using ice samples• Exclusion of solutes and nature of the ice

surface

The QLL

Ice matrix

Potentially two regions of ice where reactions can occur

Reactivity may not be similar in both regions

nm sc

ale

mm

scal

e

In most laboratory experiments, reagents are frozen from solution and samples are melted prior to analysis

Often the kinetics may be well predicted from aqueous-phase results

Are kinetics measured in bulk ice indicative of reactivity in the QLL?

Where does atmospheric ice chemistry occur?

Analyse for reagents & products using fluorescence,HPLC, etc.

Example 1: direct photolysis of aromatic compounds

Anthracene

Time (s)

0 1000 2000 3000 4000 5000 6000

ln(I/

I o)

-3

-2

-1

0 Aqueous solution

Ice cubes

Time (min)

0 20 40 60 80 100 120 140

ln(I/

I o)

-3

-2

-1

0

Naphthalene, waterAnthracene, water

Time (min)

0 20 40 60 80 100 120 140

ln(I/

I o)

-3

-2

-1

0

Naphthalene, iceAnthracene, iceNaphthalene, waterAnthracene, water

Anthracene and naphthalene photolysis on ice: In situ measurements

monochromator and PMT

liquid light guide

laserVarious wavelengths< 3

computer

oscilloscope

Raman/LIF

T.F. Kahan and D. J. Donaldson, J. Phys. Chem. A 111, 1277-1285 (2007)

Time (s)

0 1000 2000 3000 4000 5000 6000

ln(I/

I o)

-3

-2

-1

0

Aqueous solution

Ice cubes

Ice granules

Anthracene photolysis in bulk samples

Medium kobs (10-3 s-1)

Bulk water 0.25 0.06

Air-water interface 0.17 0.03

Ice cube 0.4 0.2

Ice granule 1.0 0.3

Air-ice interface 1.04 0.08

T. F. Kahan et al, Environ. Sci. Technol., 44, 1303-1306 (2010)

[NaX] in solution (mol L-1)

0.0 0.2 0.4 0.9 1.0

k obs

(s-1

)

0.0000

0.0002

0.0004

0.0006

0.0008

0.0010

0.0012

NaCl (296 K)NaCl (253 K)NaBr (253 K)NaCl (243 K)

Time (min)

0 10 20 30 40 50

ln(I/

I o)

-4

-3

-2

-1

0

N

N

CH3

H

O

CH3

In aqueous solution

On ice

A photolysis rate enhancement is observed for harmine on the ice surface as well.

But on frozen salt solutions the rate reverts tothat seen on the water surface

Exclusion of saltsduring freezing createsan aqueous brinelayer at the surface

T.F. Kahan et al.,Atmos. Chem. Phys., 10, 10917-10922 (2010).

Example 2: oxidation at the air-ice interface

Water surface result

Kinetics of O3(g) + Br-

S. N. Wren et al.,J. Geophys. Res. 115  Article Number: D16309 (2010).

ice matrix

brine layer

brine in liquid pockets

X-

X-

X-

X -X-X-X- X- X-X-

Apparent saturation in kinetics, may be consequence of the [Xˉ] in the brine being independent of the initial solution [Xˉ]

Ion Exclusion

NaBr(aq) + NaBr(s)

NaBr(aq) + H2O(aq)

ice and NaBr(s)

NaBr(aq) + ice eutectic

temperature

0

-28

-10

T(C)

mol fraction NaBr

The much faster reaction rates on ice are best understood as a consequence of salt exclusion during freezing, yielding highly concentrated brines on the surface

Medium (days, 50 ppb O3)Gas phase1 22

Organic film (surf)2 44Water (surf) 44

Ice (surf, -11 ºC) 5Ice (surf, -30 ºC) 5

[Ozone] (1015 molec.cm-3)

0 2 4 6 8 10 12

k obs

(s-1

)

0.000

0.002

0.004

0.006

0.008

0.010

0.012

Ice, -30 oC

Ice, -11 oC

Phenanthrene ozonation kinetics on ice

1. Kwok et al. Environ. Sci. Technol. 1994 28: 5212. Kahan et al. Atmos. Environ. 2006 40: 3448

The loss kinetics of phenanthrene by ozonation at the air-ice interface are (a) faster than in solution; (b) first order in

phenanthreneT. F. Kahan and D. J. DonaldsonEnviron. Res. Lett. 3 045006  doi: 10.1088/1748-9326/3/4/045006 , (2008)

Water surface (upper limit)

Reactions of OH with aromatics at the air-ice interface

• OH formed from photolysis of H2O2, NO3¯, or NO2

¯

Excitation Wavelength (nm)

250 260 270 280

Inte

nsity

(arb

itrar

y un

its)

0

1

2

3

4

5

Excitation spectra in aqueous solution

Benzene

Phenol

• Reagents frozen from solution or deposited from gas

• Anthracene + OH• Benzene + OH Phenol

Excitation Wavelength (nm)

250 260 270 280

Inte

nsity

(arb

itrar

y un

its)

0

1

2

3

4

5

Excitation spectra in aqueous solution

Benzene + NaNO2

Irradiated sample

Phenol

Phenol formation observed from the photolysis of OH-precursors in water, but not on ice

T.F. Kahan et al., Atmos. Chem. Phys, 10, 843-854, (2010)

Heterogeneous reactions of aromatics with OH(g) formed from HONO photolysis on water and ice surfaces

Time (min)

0 5 10 15 20 25 30 35

Phe

nol E

mis

sion

Inte

nsity

(arb

itrar

y un

its)

0

5

10

15

20

25

30

35Lamp on

Benzene

Phenol formation on water

No phenol formation on ice

Time (min)

0 10 20 30 40

ln(I/

I o)

-0.6

-0.5

-0.4

-0.3

-0.2

-0.1

0.0

0.1

Anthracene

Water surface

Ice surface

T.F. Kahan et al., Atmos. Chem. Phys, 10, 843-854, (2010)

Once again, we do the ice cube and crushed ice experiments …

Time (s)

0 500 1000 1500 2000 2500

Phe

nol C

once

ntra

tion

(mol

L-1)

0.0

2.0e-7

4.0e-7

6.0e-7

8.0e-7

1.0e-6

1.2e-6

1.4e-6

Aqueous solution

Ice cubes

Ice granules

Phenol formation rates from OH + benzene in bulk samples

How well are solutes excluded to the air-ice surface?

[NaX] in solution (mol L-1)

0.0 0.2 0.4 0.9 1.0

k obs

(s-1

)

0.0000

0.0002

0.0004

0.0006

0.0008

0.0010

0.0012

NaCl (296 K)NaCl (253 K)NaBr (253 K)NaCl (243 K)

Water surface result

Kinetics of O3 + Br-

Raman Shift (cm-1)

600 800 1000 1200 1400 1600

Nor

mal

ized

Inte

nsity

-0.1

0.0

0.1

0.2

0.3

0.4

0.5 M 1.5 M 2.5 M

For nitric acid, NaNO3 and Mg(NO3)2, we see a good Ramansignal from the nitrate sym. str. at the air-water interface

H-O-H bend

nitrate

Raman Shift (cm-1)

800 1000 1200 1400 1600

Inte

nsity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

liquidfrozen

Raman Shift (cm-1)

800 1000 1200 1400 1600

Inte

nsity

0.0

0.2

0.4

0.6

0.8

1.0

liquid

Now with some understanding of nitrate intensities, we freeze 100 mM solutions of Mg(NO3)2

Mg(NO3)2·H2O Phase Diagram

-10 C

Mg(NO3)2 wt %

T (C)

21

Liquid Mg(NO3)2(aq) + H2O(aq)

Two PhasesMg(NO3)2(aq) + H2O(s)

SolidMg(NO3)2·9H2O(s)

Brine [Mg(NO3)2]19 wt% 1.3 M Mg(NO3)2

2.6 M NO3ˉ

Eutectic T

Mg(NO3)2·H2O Phase Diagram

-10

Mg(NO3)2 wt %

T (C)

Brine [Mg(NO3)2]19 wt% 1.3 M Mg(NO3)2

2.6 M NO3ˉ

Raman Shift (cm-1)

800 1000 1200 1400 1600

Nor

mal

ized

Inte

nsity

0.00

0.05

0.10

0.15

0.20

1.5 M

0.5 M

Raman Shift (cm-1)

800 1000 1200 1400 1600

Inte

nsity

0.0

0.2

0.4

0.6

0.8

1.0

1.2

1.4

1.6

frozen

• Surface [NO3 ] ˉ not predicted by equilibrium phase diagram

• Nitrate must be excluded to liquid pockets or incorporated into ice matrix

• Not consistent with our previous work on halide ozonation at the ice surface

Does the absorption spectrum and/or the photolysis quantum yield change in the QLL?

Wavelength (nm)

300 350 400 450 500

Nor

mal

ized

Inte

nsity

0.0

0.2

0.4

0.6

0.8

1.0

Naphthalene Emission

IceWater

Naphthalene fluorescence in hexanes at 77 K

Kawakubo et al. J. Phys. Soc. Japan 1966 21: 1469

Red-shifts in emission spectra on ice indicate self-association:-This is observed for naphthalene, anthracene, phenanthrene, benzene and phenol ... Whether aromatic is frozen from solution or deposited from the gas phase and at all concentrations studied

Molecular dynamics simulations show that aromatics on ice surfaces are not as well solvated by the water molecules presentthere as on the liquid surface dueto the fewer “free” OH at ice surface. This feature is observedalso in the Raman spectrum ofsurface water vs. ice.Thus the aromatics tend to self-associate at the ice surface to lower their energies there.

2800 3000 3200 3400 3600 38000.0

0.2

0.4

0.6

0.8

1.0

For naphthalene, the “self-associated” absorption is shifted to the red ... into the actinic region

Time (min)

0 20 40 60 80 100 120 140

ln(I/

I o)

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0 Aqueous solution

Air-ice interface

Thus self-association on ice could contribute to enhanced naphthalene photolysis kinetics

Solar output

Bree and Thirunamachandran Molec. Phys. 1962 5: 397

Crystalline

Monomer

Naphthalene absorption

Excitation wavelength (nm)

260 280 300 320

Nor

mal

ized

Inte

nsity

0.0

0.2

0.4

0.6

0.8

1.0

Benzene photolysis on ice shows a similar enhancement ... and a similar red shift in absorption

Excitation wavelength (nm)

260 280 300 320

Nor

mal

ized

Inte

nsity

0.0

0.2

0.4

0.6

0.8

1.0

Water Ice

Benzene excitation spectra

Solar output

Time (min)

0 10 20 30 40 50

ln(I/

I o)

-1.2

-1.0

-0.8

-0.6

-0.4

-0.2

0.0

0.2

0.4

Benzene photolysis kinetics

Water

Ice

(Uncomfortable) conclusions

• Reactions within the ice matrix seem to be accurately modeled using aqueous parameters (if the correct concentrations are known)

• Photolysis at the air-ice surface shows different kinetics from that within the ice matrix

• Heterogeneous reaction kinetics at the air-ice interface may be quite different (or not!!) from those in the ice matrix

• Exclusion of salts to the air-ice interface may be different from bulk thermochemical predictions

• The air-ice interface presents a very different solvating environment from the liquid-air interface