ion exclusion, ph, and halogen activation at the air- ice interface · 2014. 1. 30. · sumi n....

Ion Exclusion, pH, and Halogen Activation at the Air-

Ice Interface

By

Sumi N. Wren

A thesis submitted in conformity with the requirements

for the degree of Ph. D. of Environmental Chemistry

Graduate Department of Chemistry

University of Toronto

© Copyright by Sumi N. Wren (2013)

ii

Ion Exclusion, pH and Halogen Activation at the Air -Ice Interface

Sumi N. Wren

Ph. D. of Environmental Chemistry, Department of Chemistry, University of Toronto, 2013

A B S T R A C T

Although the air-ice interface is atmospherically important, it is difficult to model

accurately because exclusion and precipitation of solutes during freezing, deposition of

atmospheric species, and heterogeneous/photochemical processes all affect its properties. In this

thesis, glancing-angle spectroscopic methods were developed to study ice surfaces. Glancing-

angle Raman spectroscopy showed that nitrate is not strongly excluded to the ice surface during

freezing, in contradiction with expectations based on equilibrium thermodynamics. Glancing-

angle laser-induced fluorescence showed that hydronium ions are not strongly excluded when

dilute acidic solutions (HNO3 or HCl) are frozen. These results suggest that solutes are not

universally excluded and that care should be taken in modelling surface concentrations on ice.

Deposition of HCl(g) was found to result in different pH responses at the ‘pure’ vs.

‘salty’ ice surfaces. Changes at the ‘salty’ ice surface were consistent with the existence of a brine

layer at the air-ice interface while changes at the ‘pure’ ice surface were distinctly different,

indicating that it may not be appropriate to model it as a cold, liquid layer. Significantly, results

also suggest that the sea ice surface is buffered against pH changes, with important implications

for interpreting pH-dependent chemistry.

The conversion of sea-salt derived halides to reactive halogen species can lead to

dramatic changes in the oxidative capacity of the overlying atmosphere. At ambient pH and

naturally occurring halide concentrations, the dark ozonation of NaBr and NaI solutions was

found to proceed more quickly on frozen vs. aqueous substrates, consistent with a freeze-

concentration enhancement in halide concentration at the surface. A photochemical mechanism

for halogen release from artificial saline snow was evidenced. The presence of ozone and light

in the actinic region leads to accelerated production of Br2 and BrCl and the release of Cl2, in a

process enhanced by high surface area, acidity and additional gas phase Br2. The results provide

strong evidence for snowpack ‘halogen explosion’ chemistry in which gas phase halogens are

recycled back into a concentrated brine layer at the snow grain surface.

iii

AC K N O W L E DG EM EN T S

Many people in the Chemistry Department have helped me throughout my PhD, from the

Graduate Office to Chem Stores to the Electronics Shop. I am particularly indebted to Anna Liza

Villavelez and her administrative prowess. I would also like to thank the Chemistry Machine Shop

(particularly John Ford, Dave Heath and Ahmed Bobat) and the glass blower, Jack O`Donnell (for

patiently fixing all my broken glassware!). Thanks to Dan Mathers in Analest.

The various institutions that have provided funding throughout my studies also deserve my

thanks. Thanks to NSERC for providing postgraduate scholarships, IGAC (International Global

Atmospheric Chemistry) for providing funding to attend IGAC workshops, the Centre for Global Change

Science for the travel grant that allowed me the opportunity to perform research at the Paul Scherrer

Institute (PSI) and the Department of Chemistry for providing me with a special opportunity travel grant

to perform research in Lyon, France. Thanks also to the Department of Chemistry for a Chemistry

Teaching Fellowship.

Throughout my studies I have had the opportunity to work and collaborate with a number of

people, both within the Department and elsewhere. I am very grateful to Dr. Thorsten Bartels-Rausch and

Dr. Markus Ammann for hosting me at PSI. I would also like to thank Christian George for hosting me at

IRCELYON. I want to thank everyone in the Environmental Chemistry division for helping me

throughout my studies (and for creating such a positive, supportive, friendly environment). In particular,

I should thank Milos Markovic, Greg Wentworth and Phillip Gregoire for helping me with the occasional

IC analysis. I also want to thank Anne Myers and Hussain Masoom for being great co-organizers for

ECC. Of course a big thank you goes out to all past and present members of the Donaldson Group

(affectionately, the DDG`s). Thank you for putting up with me all these years, for listening to my

ramblings, for helping me both inside and outside the lab, and for being lovely people all around. I must

also thank Tara Kahan, the true ‘Ice Queen’, who taught me everything she knows about ice.

A great deal of thanks also goes to my committee members, Dr. Jon Abbatt and Dr. Jen Murphy

for providing me with invaluable advice throughout my PhD. Jon and Jen, I have a great deal of respect

for you both. Jen, you have been a great role model and a positive influence. Additionally, I want to

thank you for working on the Chemistry Teaching Fellowship with me. Jon, I really enjoyed working

with you on numerous projects (halogen review paper, chlorine activation project, ECC), thank you for

always taking the time to chat with me.

Certainly the person who deserves the greatest share of my thanks is my supervisor, Dr. Jamie

Donaldson. Jamie, thank you for being such a great, optimistic, enthusiastic, supportive (I could go on)

supervisor and mentor. I really enjoyed working with you over the past five years and respect the way

you think about science. Thank you for challenging me, and also for trusting me to make my own

decisions. I could not have imagined a better fit for a supervisor. Thank you.

Finally, thanks to my family (mom, dad and darling sister) and friends, for always being there

for me. I really do appreciate all the love, support and food (yum!) provided along the way.

iv

T AB L E O F CON T E N T S

Title Page…………………………………………………………………………………………………i

Abstract…………………………………………………………………………………………………..ii

Acknowledgements…………………………………………………………………………………….iii

List of Tables……………………………………………………………………………………………viii

List of Figures……………………………………………………………………………………………ix

Chapters

1. Introduction………………..…………………………………………………………………………1

1.1. Overview of Air-Ice Chemical Interactions (AICI)

1.1.1. Overview of Active Snow/Ice Chemistry

1.1.2. Halogen Activation and Polar Ozone Depletion

1.1.2.1. Background and Atmospheric Significance

1.1.2.2. Outstanding Questions from an AICI Perspective

1.1.3. Nitrate Snow Photochemistry


1.1.3.2. Outstanding Questions from an AICI Perspective

1.2. The Air-Ice Interface: Current Knowledge and Outstanding Questions

1.2.1. The ‘Pure’ Air-Ice Interface

1.2.1.1. Structure and Physical Properties

1.2.1.2. Chemical Reactivity

1.2.2. The ‘Impure’ Air-Ice Interface

1.2.2.1. Contaminant-Induced Surface Disordering

1.2.2.2. Salty Ice: Brine Layers and Eutectics

1.2.3. Processes Affecting the Air-Ice Interface

1.2.3.1. Exclusion of Solutes During Freezing

1.2.3.2. Precipitation of Solutes During Freezing

1.2.3.3. Deposition and Post-Freezing Processing

1.3. Techniques for Investigating the Air-Ice Interface

1.3.1. Overview and Challenges

1.3.2. Glancing-Angle Spectroscopy

1.3.2.1. Basic principles

1.3.2.2. Development of Glancing-Angle Raman for Solutes

1.3.2.3. Surface Sensitivity

1.4. Summary of thesis objectives

1.5. References

v

2. Exclusion of Nitrate to the Air-Ice Interface During Freezing……………………………….37

2.1. Introduction

2.2. Materials and Methods

2.2.1. Glancing-Angle Raman Spectroscopy

2.2.2. Aqueous Experiments

2.2.3. Ice Experiments

2.3. Results

2.4. Discussion

2.5. Atmospheric Implications

2.6. Conclusions

2.7. References

3. Laboratory Study of pH at the Air-Ice Interface………………………………………………53

3.1. Introduction


3.2.1. Experimental Overview

3.2.2. Apparatus

3.2.3. Fluorescence Spectra and Excitation Spectra

3.2.4. Sample Preparation

3.2.5. Bulk pH Measurements

3.2.6. Deposition of HCl(g)

3.2.7. Chemicals

3.3. Results

3.3.1. Development of pH Probes for the Liquid Surface

3.3.1.1. Acridine Photophysics and Acid-Base Behaviour

3.3.1.2. Acridine 430/470 Ratio

3.3.1.3. Acridine Fluorescence Decay Rate

3.3.1.4. Harmine Photophysics and Acid-Base Behaviour

3.3.1.5. Harmine 290/320 Ratio

3.3.1.6. Harmine Fluorescence Decay Rate

3.3.2. Interfacial pH of Frozen Acidic and Basic Solutions

3.3.2.1. Fluorescence pH Probes for the Frozen Surface

3.3.2.2. Freezing Acidic and Basic Solutions

3.3.2.3. Exclusion of Chloride

3.3.2.4. Exclusion and Self-Association of Probe Molecules

3.3.3. Interfacial pH Changes Due to Deposition of Gas Phase HCl

3.4. Discussion

3.4.1. Developing pH Probes for the Frozen Surface


3.4.3. Exclusion of Chloride

3.4.4. Interfacial pH Changes from Gas Phase Deposition of HCl

3.5. Conclusions

3.6. References

vi

4. How Does the Deposition of Gas Phase Species Affect pH at Salty Interfaces?................91

4.1. Introduction

4.2. Experimental Methods

4.2.1. Experiment Overview

4.2.2. Apparatus



4.2.5. Introduction of HCl(g) or NH3(g)

4.3. Results and Discussion

4.3.1. Harmine as a pH Indicator at the Frozen Salt Water Surface

4.3.2. pH Changes at the Frozen Freshwater vs. Frozen Salt Water Surface

4.3.3. pH Changes at the Frozen Salt Water vs. Frozen Seawater Surface


4.5. Conclusions

4.6. References

5. Spectroscopic Studies of the Heterogeneous Reaction Between O3(g) and Halides at the

Surface of Frozen Salt Solutions…………………………………….………………………….111

5.1. Introduction

5.2. Experiment Setup

5.2.1. Apparatus

5.2.2. Spectroscopic Measurements

5.2.3. Ozone Generation


5.2.5. Chemicals

5.2.6. Calibration of Raman Data

5.3. Results

5.3.1. Frozen NaBr + Ozone

5.3.2. Frozen NaI + Ozone

5.4. Discussion

5.5. Conclusions

5.6. References

6. Photochemical Bromine and Chlorine Activation from Artificial Saline Snow………..131

6.1. Introduction


6.2.1. Experimental Apparatus

6.2.2. Chemical Ionization Mass Spectrometer

6.2.3. Calibration

6.2.4. Snow Preparation

6.2.5. Snow Characterization


vii

6.3.1. Photochemical Chlorine Production

6.3.2. Direct Observation of a Snowpack Halogen Explosion

6.3.3. Influence of Snow Br‾/Cl‾

6.3.4. Wavelength Dependence

6.3.5. Surface Area Dependence

6.3.6. pH Dependence

6.3.7. HOx and O3

6.3.8. Brine Chemistry

6.4. Conclusions


6.6. Appendix: Supporting Information

6.7. References

7. Conclusion…………………………………………………………………………………………163

7.1. Summary

7.1.1. Development of Techniques for the Ice Surface

7.1.2. Concentration of Species at the Air-Ice Interface

7.1.3. The ‘Pure’ vs. ‘Impure’ Air-Ice Interface

7.1.4. What is the pH of Sea Ice?

7.1.5. Towards a Better Understanding of Halogen Activation Chemistry

7.2. Future Directions

7.3. References

viii

LIST OF TABLES

C H A P T E R S I X

Table 6.1. Parameters used in the experiments. The BASE scenario conditions are shown in bold

text.

ix

L I ST O F F I G UR E S

C H A P T E R O N E

Figure 1.1 Key reactions for gas phase ozone depletion (X = Cl, Br, I) (adapted from Simpson et

al.(3))

Figure 1.2 Sources of halides in the polar boundary layer (source: Abbatt et al. (11)). Bulk

salinity values are given in practical salinity units, PSU (PSU is roughly equivalent to the salt

mass fraction expressed in parts per thousand).

Figure 1.3. Representative phase diagram for a binary water-salt (NaX) system. The shaded

region represents the range of temperatures and total NaX mol fractions for which a liquid

brine is thermodynamically predicted to be in coexistence with solid water ice.

Figure 1.4. Raman spectra of a liquid D2O sample exposed to ambient air. The spectrum of a

bulk sample is the dashed red line and the glancing-angle spectrum from the surface region is

the solid blue line. The OD-stretch and OH-stretch appear in the spectrum around ~2400 and

~3400 cm-3, respectively. Source: Figure 1 from Wren and Donaldson (121).

Figure 1.5. DMSO adsorption isotherms: Wren and Donaldson (120) glancing angle Raman

using intensity of the ν-CSC (yellow triangles); Allen et al. (123) surface tension (green circles);

Tarbuck and Richmond (122) surface tension (red squares). All data have been normalized to

the high concentration glancing-angle Raman data. Source: Figure 2 from Wren and Donaldson

(121).

Figure 1.6. Nitrate symmetric stretching peak (ν -sym NO3¯ ) intensity as a function of

calculated [NO3¯]bulk (=[HNO3]tot × αbulk): bulk data (black circles) and glancing-angle Raman data

(red triangles) for HNO3, and glancing-angle Raman data for KNO3 (green squares). Each point

represents the average of 3–5 trials and the vertical error bars represent the standard deviation

of the average. Source: Figure 7 from Wren and Donaldson (121).

Figure 1.7. A cartoon of ice and its associated ‘liquid’ or disordered regions. Motivating

questions and their associated Chapters are shown.

C H A P T E R T W O

Figure 2.1. Glancing-angle Raman spectra acquired at the surface of aqueous Mg(NO3)2

solutions with bulk concentrations: 0.25 M (long red dash), 0.50 M (medium green dash), 0.75 M

(short blue dash), 1.25 M (orange dash-dot). The spectrum acquired at the surface of pure water

is shown as the solid black line.

x

Figure 2.2. Area under the -sym NO3ˉ peak as a function of Mg(NO3)2 concentration. Each

point represents the average of at least two trials and the error bars represent the associated

uncertainty.

Figure 2.3. Glancing-angle Raman spectra acquired at the surface of a 100 mM Mg(NO3)2(aq)

sample prior to freezing (red dashed line) and of the same sample after freezing to 268 K (black

solid line). Each plotted spectrum is an average of 4 individual spectra.

Figure 2.4. Glancing-angle Raman spectra acquired at the surface of frozen samples: 100 mM

Mg(NO3)2 cooled to 258 K (solid black line); 100 mM Mg(NO3)2 subsequently warmed to 268 K

(dashed red line); 75 mM Mg(NO3)2 cooled to 258 K (dashed blue line). Each plotted spectrum is

an average of 4 individual spectra. The area under the -sym NO3ˉ peak (in arbitrary area units)

is the same within experimental error for all three spectra.

C H A P T E R T H R E E

Figure 3.1. Acridine fluorescence spectra acquired at the liquid water surface with initial pH of

2.3 (dashed black line) and 8.8 (dashed red line) and at the frozen water surface with initial pre-

freezing pH of 3.4 (solid black line) and 9.4 (solid red line).

Figure 3.2. Acridine 430/470 intensity ratios as a function of initial pH. Data from spectra

acquired at the liquid surface are shown as open circles. A the frozen surface (solid symbols)

samples had pH adjusted with HNO3 (red triangles), HCl (black circles), and NaOH (green

squares). The liquid data has been fit to a 3-parameter sigmoidal function (solid black line) of

the form f= a/(1+exp(-(x-x0)/b)) with a = 2.2 ± 0.1, b = 0.7 ± 0.1 and x0 = 5.6 ± 0.2 and R2 = 0.92.

Figure 3.3. Single-exponential fits to the acridine fluorescence decay profile at 470 nm: a) Liquid

water surface pH 4.6 (R2 = 0.99) b) liquid water surface pH 10.9 (R2 = 0.96); c) frozen water

surface pH 4.6 (R2 = 0.96); d) frozen water surface pH 10.1 (R2 = 0.94). Raw data is plotted as

open circles (red for samples adjusted with HNO3 and green for samples adjusted with NaOH)

and the fit is plotted as the solid black line.

Figure 3.4. Acridine fluorescence decay rates at 430 nm (open symbols) and 470 nm (solid

symbols) measured at the liquid water surface plotted as a function of initial pH for samples

adjusted with HNO3 (red triangles), HCl (black circles), and NaOH (yellow diamonds);

unadjusted samples (green squares).

Figure 3.5. Acridine fluorescence decay rates following excitation at 337 nm plotted as a

function of bulk [Clˉ] (assuming complete ionization). Measured at the liquid water surface at

xi

470 nm for samples acidified with HCl (this study, black circles) and at the liquid salt water

(NaCl) surface at 495 nm (red triangles). The solid line is a linear fit to the HCl data. The NaCl

data was obtained by Dr. T. F. Kahan.

Figure 3.6. Harmine excitation spectra acquired at the liquid water surface with initial pH ~ 6

(dashed red line) and pH ~ 10 (dashed black line); and at the frozen water surface with initial

pH ~ 6 (solid red line) and pH ~ 10 (solid black line). Harmine fluorescence monitored at 430

nm.

Figure 3.7. Harmine 290/320 intensity ratios as a function of initial pH. Data from spectra

acquired in bulk liquid water (crosses), at the liquid water surface (open triangles), and at the

frozen water surface (red triangles).

Figure 3.8. Harmine fluorescence decay rates at 430 nm following excitation at 320 nm.

Measured at the liquid water surface (black circles) and at the liquid salt water (0.5 M NaCl)

surface (red triangles).

Figure 3.9. Acridine fluorescence decay rates at 470 nm measured at the frozen surface plotted

as a function of initial pre-freezing pH for samples adjusted with HNO3 (red triangles), HCl

(black circles) and NaOH (yellow diamonds); unadjusted samples (green squares). Note the

very different behaviour with HCl compared to HNO3. The purple dotted line is a guide to the

eye.

Figure 3.10. (a) The harmine 300/320 ratio acquired at the sample surface as a function of time

during the slow freezing of a water sample (initial pH adjusted with NaOH to pH = 8.5). (b)

Harmine fluorescence intensity at 320 nm (following excitation at 337 nm) during the freezing

of the same sample. The sample froze at time, t = 0 (indicated by the dashed grey line).

Figure 3.11. Acridine fluorescence decay rates at 470 nm acquired at the liquid water surface

(open circles) and frozen water surface (black circles) following excitation at 337 nm, plotted as

a function of initial pH for samples acidified with HCl. Decay rates were obtained by fitting

single exponential decays to the fluorescence vs. time signal.

Figure 3.12. Harmine fluorescence spectra acquired at the frozen water surface for pre-freezing

harmine concentrations of 2.5 × 10-6 M (medium dash red) and 1.0 × 10-7 M (short dash green).

Harmine fluorescence spectrum acquired at the liquid surface with a bulk harmine

concentration of 2.5 × 10-6 M (solid black). All samples in equilibrium with air (initial pH ~ 5.9).

The portion of the spectrum that interferes with the Raman OH-stretching mode (~ 380 nm) has

been removed.

xii

Figure 3.13. Acridine fluorescence spectra acquired at the frozen water surface for pre-freezing

acridine concentrations of 7.5 × 10-7 M (solid black), 1.5 × 10-6 M (medium dash red) and 5.0 × 10-6

M (short dash green). All samples in equilibrium with air (initial pH ~ 5.9).

Figure 3.14. (a) Acridine 430/470 intensity ratio as a function of time relative to the introduction

of HCl(g). Measured at the liquid surface (black circles) and at the frozen surface (two trials

shown, red squares and triangles). The purple line indicates the time at which HCl(g) was

introduced to the chamber (time = 0). (b) Acridine fluorescence decay rate at 470 nm as a

function to time relative to the introduction of HCl(g). Measured at the liquid surface (black

circles, left axis) and at the frozen surface (red squares, right axis). Time = 0 denotes the time at

which HCl(g) was introduced to the chamber. (c) Acridine fluorescence decay rates at 470 nm

measured at the liquid surface, plotted as a function of their corresponding 430/470 intensity

ratios. Shown are a sample prior to HCl(g) exposure (green triangle) and during exposure to

HCl(g) (i.e., the liquid surface data from Figures 3.14a and 3.14b) (solid black circles). Samples

whose initial pH values were either unadjusted or adjusted with HCl(aq) (i.e., the liquid surface

data from Figure 3.2) are illustrated as open circles. (d) Acridine fluorescence decay rates at 470

nm measured at the frozen surface, plotted as a function of their corresponding 430/470

intensity ratios. A sample prior to HCl(g) exposure (green triangles) and during exposure to

HCl(g) (i.e., the frozen surface data from Figures 3.14a and 3.14b) (solid red squares). Samples

whose initial pH values were either unadjusted or adjusted with HCl(aq) (i.e., the frozen surface

data from Figure 3.2) (open squares).

C H A P T E R F O U R

Figure 4.1. Harmine excitation spectra acquired at the frozen freshwater surface (red traces) and

at the frozen salt water (0.5 M NaCl) surface (green traces) for strongly basic pre-freezing pH

~9.8 (solid traces) and near-neutral pre-freezing pH (dashed traces). Spectra were collected by

scanning the excitation wavelength in 5 nm steps while monitoring harmine fluorescence at ~

430 nm.

Figure 4.2. The harmine 290/320 intensity ratio measured at the frozen freshwater surface (red

circles) and at the frozen salt water (0.5 M NaCl) surface (green triangles) as a function of pre-

freezing pH. Initial pre-freezing pH was adjusted with NaOH(aq) or HCl(aq) and measured

with a commercial pH electrode.

Figure 4.3. Harmine fluorescence intensity (in arbitrary units) measured at 430 nm following

excitation at 320 nm plotted as a function of time relative to the freezing of a freshwater sample

(red circles, left axis), salt water (0.5 M NaCl) sample (green triangles, inner right axis) and

xiii

artificial seawater sample (blue squares, outer right axis). The samples froze at t = 0 (indicated

by the dashed grey line).


circles) and at the frozen salt water (0.5 M NaCl) surface (green triangles) as a function of time.

The dashed line indicates the time (t = 0) at which a 0.5 SLPM flow of 100 ppm of HCl in N2 was

introduced to the chamber. The pre-freezing pH of the samples was adjusted with NaOH(aq) to

a pH ~ 9.8. The final melted pH of the freshwater sample was ~ 3 and the final melted pH of the

salt water sample was pH ~ 2.5.

Figure 4.5. The harmine 290/320 intensity ratio measured at the frozen salt water surface (green

symbols) and at the frozen artificial seawater (blue symbols) surface as a function of time. The

dashed line indicates the time (t = 0) at which a 50 sccm flow of N2 passing over a 12.75 wt%

NH4OH(aq) solution held at 253 K was introduced to the chamber. The pre-freezing pH of the

salt water samples was adjusted with NaOH(aq) to a pH ~ 8.1, to be the same as the equilibrium

pre-freezing pH of the seawater samples. The final melted pH of the salt water samples was > 10

while the final melted pH of the seawater samples was < 10. The different shapes of the symbols

represent two separate trials.

C H A P T E R F I V E

Figure 5.1. (a) Excitation spectra of harmine at an air- pure ice interface (blue circles) and of the

same sample after the deposition of gas-phase ammonia (red triangles); (b) Excitation spectra of

harmine at the surface of a frozen 10 mM NaBr solution (blue circles) and of the same sample

after exposure to 8 × 1015 molec cm-3 gas-phase ozone for ~20 min (red triangles). The excitation

spectra were all obtained by monitoring the emission intensity at 420 nm.

Figure 5.2. Fluorescence intensity of harmine excited at 322 nm and monitored at 420 nm at an

air-ice interface during exposure to 8 × 1015 molec cm-3 gas-phase ozone. The arrow indicates the

time at which ozone was introduced to the reaction chamber. The dashed red line is the average

fluorescence intensity prior to exposure while the solid red line is a linear fit to the data during

ozone exposure.

Figure 5.3. (a) Raman spectra of the surface of a frozen 50 mM NaBr solution prior to ozone

exposure (solid black trace) and after ~30 min exposure to 5 × 1015 molec cm-3 gas-phase ozone

(dashed red trace); (b) First-order decay of Raman OH-signal at 3450 cm-1 during the reaction of

5 × 1015 molec cm-3 gas-phase ozone with a frozen 100 mM NaBr solution. The solid line is a

linear fit to the semi-log data.

xiv

Figure 5.4. (a) Observed ozonation rate constants at the surface of frozen halide solutions

plotted as a function of initial halide concentration. (b) Comparison of ozonation rate constants

at the surface of frozen and aqueous bromide solutions. For both parts frozen bromide (LIF) is

given by black circles; frozen bromide (Raman, scaled): red triangles; frozen iodide (Raman,

scaled): green squares; aqueous bromide (calculated from Eq. 7 in Clifford and Donaldson (23)):

solid blue line. Error bars represent the standard deviation about the mean of at least three

trials.

Figure 5.5. Photographs of frozen 25 mM sodium iodide samples after ~5 min exposure to gas-

phase ozone. (a) illustrates the amber-coloured reaction product (triiodide), while (b) clearly

shows the product concentrated at the peak formed at the sample surface (indicated by the

arrow) due to non-uniform freezing.

Figure 5.6. Calculated uptake coefficients, γ, for ozone (8 × 1015 molec cm-3) reacting at the

surface of frozen Br- solutions as a function of initial [Br‾], plotted on a log-log scale. The solid

line is a linear fit to the log-log data. The error bars represent the uncertainty in the rate

constants shown in Figure 5.4.

C H A P T E R S I X

Figure 6.1. Schematic of the experimental apparatus.

Figure 6.2. Photographs of the artificial snow samples taken immediately after preparation.

Figure 6.3. Time evolution of the dihalogen concentrations during a typical experiment: Br2 (red

triangles), Cl2 (black circles) and BrCl (green squares). Each point represents the average of a 2.5

minute time bin (77 data points). The dashed line indicates the time at which the ozone

generator was turned on and the solid line (t = 0) indicates the time at which the samples were

illuminated. Panels: a) BASE scenario conditions as in Table 6.1; BASE scenario conditions

except with b) high purity NaCl(s) (Br¯ < 0.001 wt%) to prepare solution; c) 360 nm long-pass

filter; d) coarse mode snow, scaled up by a factor of 5; e) pre-freezing pH = 4.3; f) snowpack T =

252 ±1 K, scaled up by a factor of 5.

Figure 6.4. Time traces of a) dihalogen concentrations – Br2 (red triangles), Cl2 (black

circles) and BrCl (green squares); and b) ozone concentrations (light blue diamonds)

during an experiment in which an additional flow of Br2(g) was delivered to the sample

from a glass manifold (in the period of time indicated by the vertical blue dashed lines).

The yellow areas indicate time periods during which the sample was illuminated. The

dotted areas indicate time periods during which the ozone generator was switched on

xv

([O3] ~ 1 × 1014 molecules cm-3). The experiment was run under BASE scenario

conditions except high purity NaCl(s) (Br¯ < 0.001 wt%) was used to prepare the

solution. Each point represents the average of a 1 minute time bin (30 data points).

Figure 6.5. Dependence of the chlorine (Cl2) yield as a function of snowpack T. Vertical error

bars represent the range from two trials. Horizontal error bars represent the estimated

uncertainty in the snowpack T. The yields were calculated by integrating the Cl2 concentration-

time signal to 50 minutes and scaling by the flow rate.

Figure 6.6. Time evolution of dihalogen concentrations: Br2 (red triangles), Cl2 (black

circles) and BrCl (green squares) for the BASE scenario conditions with [O3] of a) 3 × 1013

molecules cm-3 and b) 9 × 1013 molecules cm-3. Each point represents the average of a 2.5

minute time bin (77 data points). In these experiments the samples were illuminated (λ

> 310 nm) prior to turning on the ozone generator (at t = 0).

Figure 6.7. Time evolution of a given dihalogen concentration as a fraction of the total

dihalogen concentration for the BASE scenario conditions and varying [O3] of 3 × 1013 molecules

cm-3 (circle), 6 × 1013 molecules cm-3 (triangle), 9 × 1013 molecules cm-3 (square) and 1.3 × 1015

molecules cm-3 at 252 K (crosses). Samples were illuminated (λ > 310 nm) prior to turning on the

ozone generator (at t = 0). Panel a) Cl2 (black symbols); b) Br2 (red symbols) and c) BrCl (green

symbols).


dihalogen concentration for the BASE scenario conditions and snow sample temperature of 252

K (triangles) and 257 K (circles). Samples were illuminated at t = 0. Panel a) Cl2 (black symbols);

b) Br2 (red symbols) and c) BrCl (green symbols).

Figure 6.9. Representative time evolution of dihalogen concentrations: Br2 (red triangles), Cl2

(black circles) and BrCl (green squares) for BASE scenario conditions but different pre-freezing

NaCl concentrations a) 0.1 M, b) 0.5 M (BASE) and c) 1.0 M. Each point represents the average

of a 2.5 minute time bin (77 data points). The dashed line indicates the time at which the ozone

generator was turned on and the solid line (t = 0) indicates the time at which the samples were

illuminated.


dihalogen concentration for the BASE scenario conditions and pre-freezing NaCl of 0.1 M

(circle), 0.5 M (triangle), 1.0 M (square) (the same data as in Figure 6.9). Samples were

illuminated at t = 0. Panel a) Cl2 (black symbols); b) Br2 (red symbols) and c) BrCl (green

symbols).

1

C H AP T E R O NE

INTRODUCTION

Figure 1.2 reprinted from J. P. D. Abbatt et al., Halogen activation via interactions with

environmental ice and snow in the polar lower troposphere and other regions, Atmospheric

Chemistry and Physics, 12, 6237 – 6271 (2012). DOI: 10.5194/acp-12-6237-2012 with permission

from the author.

Figures 1.4, 1.5 and 1.6 reprinted with permission from S. N. Wren and D. J. Donaldson,

Glancing-angle Raman study of nitrate and nitric acid at the air-aqueous interface, Chemical

Physics Letters, 522, 1 – 10 (2012). DOI: 10.1016/j.cplett.2011.10.019. Copyright (2012) Elsevier.

2

1.1. Overview of Air-Ice Chemical Interactions (AICI)

1.1.1. Overview of Active Snow and Ice Chemistry

The word ‘cryosphere’ is derived from the Greek words ‘kryos’, meaning cold and

‘sphaira’, meaning globe, and thus refers to the frozen, snow or ice covered regions of our

planet. Although the cryosphere covers a seasonal maximum of 40% of the earth’s surface,

historically, snow and ice were considered to be chemically inert and therefore unimportant as

reactive substrates (1). Furthermore, since the polar regions are remote and relatively pristine,

they were not initially expected to be regions of active oxidative chemistry. However, over the

past few decades, scientists have revealed that ice and snow do indeed represent important

multiphase reactors and a growing body of evidence is showing that heterogeneous and

photochemical reactions occurring on frozen media can significantly perturb the composition

and oxidative capacity of the overlying atmosphere. Measurements of higher than expected

concentrations of trace atmospheric species such as NOx (NO + NO2), HONO, HOx (OH + HO2),

reactive halogen species (RHS, such as Br2, BrO, BrCl and Cl2) and small organics (such as

HCHO and CH3CHO) have now been attributed to air-ice chemical interactions (AICI) (1-3).

Although the importance of snow and ice as reactive substrates has now been recognized, a

good mechanistic understanding of the multiphase chemistry that proceeds on frozen media is

lacking. This is largely due to a poor understanding of the physical and chemical nature of the

air-ice interface.

In the Introduction, I first discuss two important examples of ice/snow mediated

chemistry: halogen activation (which refers to the conversion of inert halides into reactive

halogen species) and its role in polar ozone depletion (Section 1.1.2) and nitrate snowpack

photochemistry (Section 1.1.3). After using these examples to motivate a need for a better

description of the air-ice interface, I provide a brief overview of the environmental air-ice

interface and the processes that affect it (Section 1.2). Subsequently I describe the main

technique used here to investigate the air-ice interface (glancing-angle spectroscopy) (Section

1.3). The Introduction concludes with a brief summary of the thesis objectives (Section 1.4).

3

1.1.2. Halogen Activation and Polar Ozone Depletion


In the mid-1980s, scientists made the surprising discovery that during the polar spring,

surface ozone concentrations episodically decrease from ambient (20 – 40 ppbv) to near-zero

levels (as low as < 1 ppb) (4-6). Soon after, Barrie et al. (7) showed that ground level ozone

concentrations at Alert, NWT are anti-correlated with measurements of filterable bromine (sum

of particulate Br and gaseous HBr) which led to the understanding that ozone depletion events

(ODEs) are intimately connected to elevated levels of reactive bromine species (8). The gas

phase photolysis of reactive bromine species such as BrO and Br2 leads to the formation of Br

radicals which catalytically destroy ozone as shown in Figure 1.1. The key step in the reaction

sequence is the BrO self-reaction which leads to the formation of two Br radicals, either directly

or via Br2.

Figure 1.1 Key reactions for gas phase ozone depletion (X = Cl, Br, I) (adapted from Simpson et

al.(3))

The ozone destruction cycle shown in Figure 1.1 requires a source of reactive bromine.

Although biogenic bromine sources were initially considered (7), it has now been established

that the main source of the reactive bromine is sea-salt derived bromide (3). A reaction sequence

4

known as the ‘bromine explosion’ is thought to be largely responsible for the conversion of this

sea-salt bromide to Br2 (9, 10):

HOBr + Br‾ + H+ Br2 + H2O (R1)

Br2 + hν → 2Br (R2)

Br + O3 → BrO + O2 (R3)

BrO + HO2 → HOBr + O2 (R4)

Net: Br‾ + H+ + O3 + HO2 → Br + 2O2 + H2O

In the first step (R1), mp denotes that the reaction is a multiphase process: gas phase

HOBr reacts heterogeneously with condensed phase Br‾. The net result of R1-R4 is the

conversion of one inactive bromine species (Br‾) to two reactive bromine species (2Br) via the

consumption of only one reactive bromine species (HOBr). Thus the ‘bromine explosion’

sequence is auto-catalytic and leads to the exponential growth of reactive bromine. Over the

years an increasingly complex picture of halogen activation and halogen-related ozone

depletion has emerged, as described in greater detail in Simpson et al. (3) and Abbatt et al. (11).

This picture includes chemistry involving chlorine, iodine, NOx, and organics (volatile organic

compounds – VOCs, dissolved organic matter –DOM) in multiphase processes that are highly

coupled (3). However, as briefly discussed below (in Section 1.1.2.2), our understanding of this

chemistry is far from complete.

Halogen activation and subsequent ozone depletion significantly influences the

composition of the troposphere (3). During ODEs (which typically persist for 1 – 3 days

depending on meteorology), ozone-dominated oxidation pathways are rendered less important,

which changes the fate and lifetime of trace atmospheric compounds (3). For example, during

ODEs, high Cl radical concentrations lead to the rapid depletion of small hydrocarbons ((12),

also see Simpson et al. (3)). The Cl-oxidation of VOCs in turn affects the budgets of oxidized

volatile organic compounds (OVOCs) and HOx. High BrO mixing ratios are also expected to

5

influence the sulphur cycle by rapidly oxidizing dimethyl sulfide (DMS) to dimethyl sulfoxide

(DMSO). The fate of DMS (which is otherwise governed by reactions with OH and NO3) has

important implications for new particle formation (13). Finally, elevated BrO mixing ratios are

now known to be responsible for atmospheric mercury depletion events – episodes during

which gas phase elemental mercury (Hg0) is rapidly oxidized to oxidized inorganic mercury

(collectively referred to as reactive gaseous mercury or RGM) (3, 14, 15). The conversion of Hg0

to RGM represents a serious toxic threat to polar ecosystems: RGM is readily scavenged from

the atmosphere; is more toxic and bioaccumulative than Hg0; and has the potential to be

converted into a methylated form, which is in turn more toxic and more bioaccumulative than

RGM (14).

1.1.2.2. Outstanding Questions From an AICI Perspective

Several open questions regarding halogen activation remain. Here I focus on the chief

knowledge gaps that have motivated a need to better understand the chemical and physical

properties of the air-ice interface. As mentioned above, a ‘bromine explosion’ mechanism is

thought to play an important role in driving ODEs. However, the mechanism shown in R1 – R4

requires a source of ‘seed’ reactive bromine. Where does this ‘seed’ reactive bromine originate? How

is it formed from sea-salt bromide? Several proposals exist including heterogeneous processes

involving ozone (16-18) or the hydroxyl radical (19); photochemical processes (involving for

example, NO3¯ or H2O2 photolysis) (20); photosensitized reactions (21); and freeze-induced

processes (22, 23). Furthermore, although the HOBr reaction (R1) has been studied in the

laboratory, with results confirming its key role in forming Br2 (24-26), many mechanistic details

of the chemistry are lacking or are based upon known aqueous phase chemistry (i.e., using

aqueous phase reactions rates extrapolated to lower temperature) (27). One important question

concerns the extent to which protons are required for R1 to proceed, especially given that sea-

salt bromide originates from seawater, which is naturally alkaline. Laboratory experiments (24)

suggest that Br2 formation from HOBr reactions on frozen halide solution surfaces occurs at

higher bulk pH values than might be expected based on aqueous phase chemistry (27). The

question of acidity becomes difficult since it requires knowledge of pH at the frozen halide

6

solution surface – no small challenge as discussed below in Section 1.2.3. Can halogen activation

be understood based on aqueous phase chemistry/parameters? What is the mechanism? How efficient is

the recycling of reactive bromine? In addition to pH, the importance of enriched Br‾ concentrations

(both absolute – with respect to sea water – and relative – with respect to Cl‾) and temperature

have not been well established. This is largely due to difficulties in measuring or predicting

local halide concentrations on ice/snow (which are enriched relative to their bulk, melted values

as discussed in Section 1.2.3). Thus a better understanding of concentration enrichment and pH

at the frozen sea ice surface is a prerequisite to better understanding halogen activation.

Figure 1.2 Sources of halides in the polar boundary layer (source: Abbatt et al. (11)). Bulk salinity

values are given in practical salinity units, PSU (PSU is roughly equivalent to the salt mass

fraction expressed in parts per thousand).

Since a good understanding of the processes and factors involved in initiating ‘bromine

explosion’ chemistry is lacking, it has been difficult to determine which frozen sea ice substrates

and environmental conditions are most favourable for halogen activation and ozone depletion.

Here I use the term ‘frozen sea ice substrates’ loosely to encompass all frozen substrates which

may have originated or been contaminated with a sea-salt source. As illustrated in Figure 1.2,

these include: sea ice (newly formed sea ice, first year sea ice and multiyear ice), brine covering

sea ice, re-freezing leads and polynyas, frost flowers, saline snow (lying on sea ice, or

contaminated by frost-flowers and/or sea-salt aerosol), blowing/subliming saline snow etc. The

7

potential role that these substrates play in halogen activation is discussed more thoroughly in

Abbatt et al (11). These substrates each differ in pH, total salinity (i.e., bulk halide

concentration), [Br‾]/[Cl‾] ratio, specific surface area and vary in their temporal and spatial

extent. Since bulk concentration and pH are not readily related to local concentration and pH

(see Section 1.2.3), a good understanding of how these substrates differ in a way that is relevant

to halogen activation is still lacking.

Finally, most of the discussion has thus far focused on the role of reactive bromine species.

Although iodide concentrations in seawater are quite low, high levels of boundary layer IO (~

20 pptv) have been measured in the Antarctic (28-30). Somewhat perplexingly, similar elevated

levels are not observed in the Arctic (31, 32). Thus questions remain as to the source (biogenic

vs. sea-salt derived) of this reactive iodine. Only small levels of IO are required to significantly

enhance ozone depletion (via direct reaction as well as via the IO + BrO cross reaction) (3). In

addition, iodine activation is expected to influence new particle formation via I2O4 and I2O5 (32).

Less is known about the role that reactive chlorine plays, in part due to a scarcity of field

chlorine measurements. Moreover, it is not known if or how sea-salt derived Cl‾ is activated on

frozen sea ice substrates. However, recent measurements (33) of unexpectedly high Cl2 mixing

ratios (> 400 pptv) correlated with light and ozone have motivated a need to understand its

source and impact. They have also given reason to study the role that actinic radiation may play

in halogen activation.

1.1.3. Nitrate Snowpack Photochemistry


NOx (= NO + NO2) is often thought of as an urban pollutant, thus it was expected that NOx

concentrations in the polar regions, which are remote and relatively pristine, would be quite

low (sub-pptv levels). However, in the late 1990’s, scientists began measuring high

concentrations of NOx and HONO over the Arctic (34-37) and Antarctic (38-40) snowpack.

Photochemical NOx production has also been observed over mid-latitude snow (41).

Understanding the formation of nitrogen oxides is important since they play an important role

8

in determining the oxidative capacity of the atmosphere (NO2 is an ozone precursor, NO alters

the OH/HO2 balance, and HONO is an important OH source). Additionally, the chemistry that

is thought to be responsible for NOx production (photolysis of snowpack nitrate, as described

below) also leads to snowpack OH formation. The presence of oxidants such as OH within the

snow interstitial air likely changes the fate of snowpack contaminants (i.e., organics) and may

lead to the formation of oxidized hydrocarbon compounds such as formaldehyde and acetone

(1, 2). Active nitrogen chemistry is also linked to halogen chemistry (20), as noted above.

Furthermore, since nitrate is ubiquitous in snow and ice, it is desirable to use the nitrate ice core

data to construct information about past atmospheric composition (42). Interpreting the nitrate

ice core data requires a good understanding of the (photo)chemical cycling of nitrogen oxides in

frozen media.

NOx release from surface snow has been the focus of many laboratory experiments (43-

51), as well as controlled experiments in the field (i.e., using snow blocks or piles and snow

chambers) (37, 39, 52, 53). Early on, field measurements showed that snowpack NOx exhibits a

diurnal cycle (34, 35, 37, 39, 54), implicating a photochemical source. ‘Shading’ experiments

further evidenced the photochemical nature of the high NOx (37, 39) and the measured

wavelength dependence for NOx production from sunlit snow (which generally tracks j-NO3‾

for aqueous solutions) confirmed that nitrate is the precursor species (43). And so, as first

proposed by Honrath et al. (34), the source of NOx has now been attributed to the photolysis of

snowpack nitrate (see Grannas et al. (1), and reference therein). The important role of the

snowpack has been demonstrated by field experiments which show that NOx concentrations are

higher in the snowpack interstitial air than in ambient air (34, 37, 39), and that NOx fluxes are

typically positive (from the snowpack) (54). Measurements indicate that NO2 levels are

generally higher than NO levels by a factor of 2 – 3 (35, 37), reaching up to 100s of pptv. HONO

levels are highly variable, but often similar to NO levels (37).

Nitrate photochemistry in aqueous media has been studied extensively. The nitrate n *

absorption band is centred ~ 305 nm. Photolysis of aqueous nitrate solutions in the actinic

9

region (λ > 290 nm) is known to yield NO2 directly, and NO and HONO via secondary reactions

(55) :

NO3¯ + hν NO2 + O¯ (R5)

NO3¯ + hν NO2¯ + O(3P) (R6)

In the aqueous phase, the quantum yield of R5 is ~ 9 × 10-3 and the quantum yield of R6 is

approximately a factor of 9 lower (which partially explains why observed NO2 levels are higher

than NO levels) for λ = 305 nm (56). The total quantum yield for the primary processes (R5 and

R6) is ~ 1%. Additional NO2 is formed via the oxidation of nitrite (NO2¯) by the hydroxyl radical

(OH):

NO2¯ + OH NO2 + OH¯ (R7)

Alternatively, photolysis of nitrite (NO2¯) formed via R6 can yield NO:

NO2¯ + hν NO + O¯ (R8)

NO can also be formed via the photolysis of NO2 (R9) or HONO (R10):

NO2 + hν NO + O(3P) (R9)

HONO + hν NO + OH (R10)

The source of the HONO is the condensed phase reaction:

NO2¯ + H+ HONO (R11)

Importantly, the O¯ formed via R5 or R8 is a precursor for OH:

10

O¯ + H2O OH + OH¯ (R12)

The main sink for NOx in the troposphere is reaction of NO2 with OH to form HNO3 (1). If this

occurs within the snowpack, or near the snow surface, before the NOx escapes to the overlying

atmosphere, then the impact of snowpack NOx and HONO formation on the overall oxidizing

capacity of the polar boundary layer may be limited (i.e., if the HNO3 is simply recycled back to

the snowpack) (1). Other reactions have been proposed but it is not the goal here to provide a

comprehensive review, rather to outline the issue as it pertains to ice/snow microphysical

processes.

1.1.3.2. Outstanding Questions From an AICI Perspective

The exact mechanism of snowpack nitrate photolysis remains elusive. Several laboratory

studies suggest that nitrate photolysis occurs in a ‘liquid-like’ region, as opposed to within bulk

ice, where slow diffusion and solvent ‘cage’ effects are likely to lead to recombination of

photoproducts (46). For example, Dubowski et al. (45) studied nitrite production resulting from

the irradiation (λ = 313 nm) of 10 mM NaNO3(aq) ice pellets as a function of temperature. They

argue that the continuity in the temperature-dependence of the experimentally obtained nitrite

quantum yields (over the melting point of ice) evidences that the chemistry occurs in a liquid-

like region. A similar argument was made by Chu and Anastasio (46), who studied the

quantum yield for OH production (R5 + R12) from frozen nitrate solutions using benzoic acid

(BA) as an OH trap. The continuity in the temperature-dependence of the experimentally

obtained OH quantum yields, as well as the good agreement between calculated activation

energies (Ea) and entropy changes (ΔS) for ice and solution experiments were cited as evidence

of chemistry occurring in a liquid-like region.

The physical basis for the existence of a ‘liquid-like’ region or layer is given in Section 1.2.

However, as described in Section 1.2, the physical and chemical properties of the liquid-like

regions on ice/snow are not well known. Since at a microscopic level ice and snow are highly

heterogeneous, caution should be taken when interpreting results from freeze-illuminate-melt-

11

analyze experiments which are blind to this heterogeneity. Interpretation generally requires

making several questionable assumptions (regarding, for example, the location and

concentration of reagents as described below). They typically also rely on the validity of

extrapolating aqueous phase quantum yields or rates to lower temperatures or extrapolating

pH dependences to conditions existing on ice. Often ionic strength effects are ignored. Similarly,

models describing nitrate photolysis on snow often rely on aqueous phase rate constants

(extrapolating to sub-273 K temperatures only when temperature-dependent rate

parameterization is available). Since these models are usually optimized against experimental

or field data, it may not be surprising that in some situations they can adequately describe

nitrate photochemistry (49, 57). However, the credibility of extrapolating bulk or aqueous

parameters to ice/snow conditions deserves further investigation. This requires an improved

understanding of the liquid-like regions on ice/snow.

Field studies generally find that higher bulk nitrate concentrations lead to higher NOx

and HONO levels (37, 43). But what are the concentrations of the reagents (i.e., nitrate and nitrite) in

the liquid-like region(s)? A good understanding of nitrate photochemistry requires knowledge of

not the bulk nitrate concentration but the localized nitrate concentration. Arriving at this

knowledge is difficult (for reasons described in Section 1.2.3 below) and so interpretation or

modeling of experimental results and field observations has thus relied on estimations. As

described in Section 1.2.3, many groups (46, 49, 57, 58) have used an approach based on freezing

point depression to estimate the concentration of nitrate in the liquid-like regions (LRs) of

snow/ice. Using this approach, an LR fraction is first calculated. Then, all of the available NO3‾

is presumed to be concentrated in this fraction. This leads to a nitrate concentration enrichment

of many orders of magnitude, depending on the conditions (46, 49). For example, Chu and

Anastasio (46) report an enrichment factor of ~ 5600 for bulk [NO3‾] = 40 μM, pH = 5 and 263 K.

Further assumptions regarding the distribution of the LR fraction must be made in order to

arrive at its thickness (57) and its location (and so it is usually assumed to reside at the surface).

The validity of such assumption requires further scrutiny. Since the LR fraction is thought to be

temperature dependent (see Section 1.2.3), concentration enrichments should be a function of

12

temperature. However, Cotter et al. (43) performed laboratory experiments on natural

snowblocks containing an average [NO3‾] of 60 ± 40 μg kg-1 and found no difference in the

production rates for NO and NO2 at 243 K and 253 K. Inconsistencies such as this highlight the

need to better understand how local concentration in the liquid-like regions is related to

temperature and other factors (although these factors will also affect other aspects of the

chemistry, for example, as noted above, the quantum yields are temperature dependent).

The role of morphology and LR location is probably quite significant (1). Where are the

liquid-like regions located? Dubowski et al. (44) suggest that NO2 and NO2¯ production from

spray-frozen KNO3(aq) ice occurs in a LR present at the sample surface. The authors suggest

that only NO2 formed in the outermost layers can escape to the gas phase (although their

analysis makes no attempt to estimate nitrate or nitrite concentrations in the LR). Similarly,

Jacobi et al. (59) argue that nitrate must be present near the surface to allow for the escape of

products to the gas phase, and suggest that this is the case for both natural snow and artificial

snow used in their experiments. Differences between the results from the Dubowski et al. (44)

study and a later study (45) by the same group (concerning the absolute value of the quantum

yield) highlight the important role of ice/snow morphology, which differed for the two studies.

Beine et al. (60) also cite the importance of understanding nitrate localization in interpreting

field results. The location of the LR will likely play an important role in governing the fate of

photo-formed products (escape to the gas phase interstitial air, recombination etc.). Since NO2,

NO and HONO all have different solubilities (i.e., Henry’s law constants), they may be

differently affected. For example, Boxe et al. (51) find that NO is more weakly held on ice

compared to NO2 and suggest that as a result, NO2 formed near the surface escapes to the gas

phase while NO2 formed in deeper layers will lead to local NO production. Jacobi et al. (49)

argue that this is in keeping with the fact that the extrapolated Henry’s Law constants for NO

and NO2 differ by a factor of 15 (at 243 K). HONO is even more soluble. Abida and Ostoff (50)

observe HONO emissions from frozen nitrate solutions post-illumination and suggest that

HONO/NO2¯ are formed below the ice surface layer and only slowly migrate or diffuse to the

13

surface. The importance of particularly unstable intermediates (such as peroxynitrous acid,

HOONO) will also likely depend on where they are formed (1, 50).

Furthermore, the role that the chemical form of nitrate plays is currently not known. It is

generally thought that the source of snowpack nitrate is adsorbed or dissolved HNO3 (1).

However, in aged snows, nitrate may be associated with heavier cations such as Na+, Ca+ or

NH4+ (which may be especially prevalent in regions with a marine influence) (1, 60, 61). If NO3‾

is associated with H+, does it exist as dissolved NO3‾ or molecular HNO3? How does this influence the

mechanism and products of nitrate photolysis on snow?

Snowpack HONO formation is thought to be favoured at low pH, although the exact pH

dependence for HONO production is not well known (49, 50, 61-64). This is largely due to the

fact that we are unable to reliably estimate the pH of the liquid-like regions where the chemistry

is thought to occur (even most models do not attempt to do so!) (58). Local pH is important

since it will determine the speciation of N(III), and will thereby influence HONO chemistry. At

neutral and high pH, N(III) will be present as NO2¯, at lower pH it will be present as HONO

(R11) – which can escape to the gas phase, and at even lower pH, HONO will protonate to form

the nitroacidium ion (H2ONO+). It is probable that each species will have a very different fate

within the snowpack (62). Snow pH may also influence the air-ice partitioning of HONO

formed in the firn air (49). The failure to observe HONO over alkaline snow has been cited as

evidence that HONO production requires protons (61). However, low HONO levels have also

been measured under ‘favourable’ conditions (high snowpack [NO3‾], strong light intensity,

and fresh, acidic snow) indicating that our understanding of this chemistry is limited (60). The

role of pH may be lessened if HONO is formed via alternate routes. Significantly, it has not

been established whether HONO is produced in the condensed phase at the snow grain surface

or within the interstitial air (e.g., via a surface catalyzed reaction (36) or by the gas phase

recombination of OH + NO). In the last few years, other mechanisms for HONO production

from snow have been explored, for example involving the nitroacidium ion (63), or dark

reactions involving protonated nitrosamide (NH3NO+) (65) or humic substances (64). Clearly,

14

our understanding of snowpack HONO formation is quite poor. A better comprehension of

HONO formation within surface snow will require an improved understanding of pH in the

liquid-like regions in/on snow.

1.2. The Air-Ice Interface: Current Knowledge and Outstanding Questions

1.2.1. The ‘Pure’ Air-Ice Interface

1.2.1.1. Structure and Physical Properties

At temperatures well below the melting-point of ice, intrinsic premelting results in the

formation of a molecular-level disordered region at the air-ice interface (66-68). The existence of

a premelting layer on ice has been measured by various techniques such as atomic force

microscopy (69), ellipsometry (70, 71), x-ray diffraction (72), photoelectron spectroscopy (73),

sum-frequency generation spectroscopy (SFG) (74), glancing-angle Raman spectroscopy (75),

glancing-angle x-ray spectroscopy (76, 77) and attenuated total reflection (ATR) infrared

spectroscopy (78).The basis for the premelting phenomenon, which is not unique to water

surfaces, is the lowering in surface free energy that results from its formation. Although the

existence of a premelting layer at the ice surface was first proposed by Michael Faraday in 1842,

the properties of this layer (e.g., thickness, temperature dependence of thickness, temperature of

onset, molecular orientation, density, viscosity etc.) are still not well known (66). However,

since water molecules at the air-ice interface are more dynamic and disordered than those in the

bulk ice matrix (i.e., more ‘liquid-like’ in character) the premelting layer on ice is frequently

referred to as the ‘quasi-liquid layer’ (QLL) on ice. The notion that a ‘liquid-like’ region exists at

the air-ice interface provides the basis for using aqueous phase chemistry to interpret ice/snow

chemistry. However, evidence exists to suggest that the physical properties of the QLL are

unique. For example laboratory studies using glancing-angle Raman spectroscopy (75) and

sum-frequency generation spectroscopy (SFG) have shown that the hydrogen-bonding structure

of the QLL is intermediate to that of the air-water interface and bulk ice. As another example,

computational studies show that the mobility of ions within a QLL is dramatically reduced,

especially as the thickness of the QLL decreases (79). The idea that the QLL is a homogeneous

15

layer with constrainable properties is even debatable; it is more likely that a gradient in physical

properties exists from the air-QLL to the QLL-ice interfaces (80).

From an atmospheric chemistry perspective it may be imperative to understand whether

species exist at the surface of the QLL (i.e., adsorbed to the strict air-QLL interface), within the

QLL (i.e., following dissolution) or adsorbed on underlying ice. Additionally, knowledge of the

thickness of the disordered layer, as well as the temperature dependence of its thickness, is

required if surface concentrations are to be determined (in some models they are simply

determined using an arbitrarily set thickness) (81). Although it is generally accepted that QLL

thickness decreases with decreasing temperature (69), a large range in thickness has been

reported – from 1 – 100 nm at 272 K depending on the measurement technique (66). The

thickness of the QLL, or more accurately, the water activity in the QLL, may also play an

important role in determining solvation and speciation at the air-ice interface. It should also be

noted that molecular-level disordering is also present at ice-ice boundaries (a.k.a. grain

boundaries, sometimes also referred to as grain boundary melting) that may be located within a

bulk sample. In my work, I am most interested in the QLL that is located at the air-ice interface

since it is the one that is most relevant to atmospheric chemistry.

1.2.1.2. Chemical Reactivity

In addition to being physically unique, a growing body of evidence is showing that the

QLL is chemically distinct. Research from our group has shown that the photochemistry of

polycyclic aromatic hydrocarbons (PAHs) and the reactivity of hydroxyl radicals are different at

the pure air-ice interface than at the air-water interface (82-85). Laboratory and computational

studies show that large organic compounds such as PAHs readily self-associate at the air-ice

interface, following exclusion during freezing (see Section 1.2.3.1) or deposition from the gas

phase (82, 86-89), furthering the notion that the air-ice interface presents a chemically distinct

environment. In light of these findings, the applicability of aqueous phase parameters (e.g.,

reaction rates, mechanisms etc.) to the air-ice interface has been questioned. This has also

16

caused a recent shift in the community away from the term QLL and its associated implication

of a ‘liquid’ character at the ice surface (90).

1.2.2. The ‘Impure’ Air-Ice Interface

1.2.2.1. Contaminant-Induced Surface Disordering

The picture of the air-ice interface becomes increasingly complicated as impurities are

added to the system. Because impurities lower the melting point of water (i.e., they cause

freezing point depression), their presence influences the thickness of the disordered region (69).

Impurities are also expected to lower its onset temperature, that is, the temperature above

which a disordered region will exist (~240 K for pure ice, although this is also debated)(66).

Contaminant-induced surface disordering can also occur for multi-component systems, at

temperatures below their eutectic (91, 92). In addition to affecting its onset temperature and

thickness, impurities affect the physical and chemical properties of the disordered regions on

ice (75).

1.2.2.2. Salty Ice: Brine Layers and Eutectics

As the system moves from a ‘pure’ system to an ‘impure’ system, there is a regime shift

(90, 93). For multicomponent systems, at temperatures below the freezing point of water but

above the eutectic temperature (the lowest melting temperature for the system), two phases are

in coexistence (see Figure 1.3). For dilute salt solutions, these phases are a pure ice phase and a

liquid brine phase. The result is two fundamental differences between the ‘pure’ ice surface and

the ‘salty’ ice surface. The first is that while the ‘pure’ ice surface is ‘wetted’ by the

aforementioned QLL, which is not predicted thermodynamically by the Gibbs phase rule, and

which may not be appropriately described as a true liquid, the ‘salty’ ice surface is likely

covered by its thermodynamically predicted brine, which is a true liquid. The second concerns

scale. The QLL is quite thin (nm scale at 263 K) (66) whereas brine layers are expected to be

relatively thicker (although this will, in a bulk sense, depend on temperature and total solute

concentration). However, substrate morphology also plays an important role. For example, high

and low surface area substrates will have different brine thicknesses for the same total brine

17

fraction, assuming that the brine evenly wets only the surface. Brine may be located in the ice

interior as well (at grain boundaries, triple junctions or trapped liquid pockets). The location of

the brine, as well as its thickness will have important consequences for chemistry occurring at

the air-ice interface, for reasons that have been mentioned previously. Whether or not these

various regions can be adequately described as a true liquid remains highly debated (90). For

small brine fractions (which may not completely ‘wet’ the air-ice interface), or for molecular-

level disordering below the eutectic temperature (92, 94), the term quasi-brine layer (QBL) is

occasionally invoked.

Figure 1.3. Representative phase diagram for a binary water-salt (NaX) system. The shaded

region represents the range of temperatures and total NaX mol fractions for which a liquid brine

is thermodynamically predicted to be in coexistence with solid water ice.

1.2.3. Processes Affecting The Air-Ice Interface

1.2.3.1. Exclusion of Solutes During Freezing

Several processes affect the chemical and physical properties of the air-ice interface,

including the freezing process itself. During freezing, solutes and other contaminants are

generally excluded from the ice matrix to interfacial regions (located at the air-ice interface or

within the bulk ice). Here I will simply refer to these interfacial regions as liquid regions (LR)

18

and ignore the distinction between QLL/QBL/brine/liquid. The exclusion process leads to

concentration enrichment in the LR (the so-called freeze concentration effect) and has in some

cases been known to significantly accelerate or induce chemistry (22, 95-98). Predicting the

magnitude of the concentration enrichment (i.e. LR concentration) is crucial to understanding

chemistry. Different approaches have been taken to estimate concentrations in the LRs. In the

simplest approach, some fraction of the solutes (often 100%) is assumed to be excluded into an

LR of some predefined thickness (81, 99). Cho et al. (92) present a more sophisticated

formulation based on freezing point depression. Using their formulation, a temperature-

dependent LR fraction is calculated. Solutes are then assumed to be completely excluded into

this LR. The major drawback is that such a formulation does not predict LR thickness,

distribution or location. Similarly, the multicomponent phase diagram (such as the one shown

in Figure 1.3, which is a construction based on equilibrium thermodynamics) gives the overall

concentration within the LR (which is a function of temperature alone) as well as the overall LR

fraction (which is a function of both temperature and total concentration). However, the phase

diagram does not provide information regarding the location or distribution of the LR. In

addition to the air-ice interface, grain boundaries, triple junctions and liquid pockets represent

possible LR locations.

Since directly probing LR concentration and location is difficult, the validity of these

approaches has not been established – or at least has not been established for all classes of

compounds. Are all compounds equally excluded? Large, hydrophic organic compounds such as

polycyclic aromatic hydrocarbons (PAHs) may be totally excluded to grain boundaries or to the

air-ice interface (100). What about inorganic salts such as nitrate or halides? Several laboratory

studies do suggest that halides are excluded into LRs at concentrations that are well predicted

using an equilibrium thermodynamic approach (92, 101), but more work is required to

understand if this is also the case for other ions such as nitrate. There is evidence to suggest that

it is not. For example, a recently developed snow model obtained the best agreement with field

observations if 100% of the available chloride and bromide but only 6% of the available nitrate

was partitioned to the LR (81, 99). Furthermore, it has not been established whether an

19

equilibrium thermodynamic approach is valid over all temperatures and over a large

concentration range (i.e. spanning QLL and brine conditions). What about H3O+ and OH¯? Are

protons excluded to the air-ice interface? This question is particularly important because it relates to

surface pH. Since H3O+ and OH¯ ions are highly mobile, their behaviour may be expected to be

quite different. In addition to knowledge of LR concentration, knowledge of LR location(s) is

required to interpret chemistry occurring on ice and snow. How valid is it to assume that the LR

exists at the air-ice interface?

Another consequence of freezing is the development of a freezing potential. This occurs if

cations and anions are differentially incorporated into the bulk ice during freezing. In this case,

a freezing potential (also known as the Workman-Reynolds freezing potential) develops across

the growing ice front (102, 103). The magnitude of the freezing potential, which can be

measured using an electrometer, depends on the freezing rate and the total concentration (104).

This freezing potential is then relaxed via the migration of H3O+ or OH¯ to/from the LR which

can alter its pH. For example, when NH4Cl(aq) freezes, more NH4+ ions are incorporated into

the ice than Cl¯ ions, leading to a positive freezing potential (the charge of the ice is positive

relative to the unfrozen solution). The positive freezing potential is relaxed by the migration of

H3O+ to the LR, thereby lowering its pH. Freezing NaCl(aq) is expected to have the opposite

effect since more Cl¯ is incorporated into the ice than Na+ (102). Direct experimental evidence for

freeze-induced pH changes is difficult to obtain. However, several groups have attributed

accelerated reaction rates and enhanced yields to freeze-induced pH changes which are

consistent with the formation and relaxation of a freezing potential. (98, 105-107).

The influence of kinetic effects (related to the method and rate of freezing) on the resulting

LR concentration and location is not well known. One recent study suggests that dynamic

instabilities that are generated during freezing govern LR formation (rather than

thermodynamics) (108). It has also been suggested that rapid-freezing (such as occurs when a

solution is immersed in liquid N2) is more likely to result in trapped liquid pockets or a high

grain boundary density (59). Heger et al. (100) studied the aggregation of methylene-blue at

20

grain boundaries and found that concentrations were enriched by 3 – 4 orders of magnitude for

rapid freezing in liquid N2 (and then warming) but by over 6 orders of magnitude for slow

freezing to 243 K. However, Jacobi et al. (59) studied nitrate photolysis in artificial snow and

concluded that the distribution of nitrate in the artificial snow (which was prepared by rapid

freezing in liquid N2) was comparable to that in natural snows. Grannas et al. (95), found no

dependence on freezing method on the kinetics of the freeze-accelerated pyridine + PNA

reaction. The influence of the freezing method on the LR concentration and locations is not

captured by a phase diagram. The differences that have been highlighted here also draw

attention to the applicability of laboratory prepared snow and ice mimics to environmental

study.

1.2.3.2. Precipitation of Solutes During Freezing

During freezing and subsequent cooling, precipitation of solutes leads to chemical

fractionation in the remaining brine. For example, when seawater freezes, precipitation of

hydrohalite (NaCl·2H2O) at temperatures below 251 K leads to a relative depletion of brine

chloride and a relative enrichment of brine bromide. Calculations by Koop et al. (94) show that

the brine [Br¯] is a factor of ~38 greater at 240 K than at 273 K whereas the brine [Cl¯] is only a

factor of ~11 greater. The decrease in the brine [Cl¯]/[Br¯] from its seawater value of ~ 650 may

have important implications for halogen activation. Indeed Lopez-Hilfiker et al. (109) have

shown that changes in the products and yields of the reaction of N2O5 with halide-doped ice are

readily explained by temperature-related changes in brine [Br¯] and [Cl¯].

If precipitation occurs, physical separation of the brine from the ice will impart a chemical

signature that may be readily detected from a bulk analysis. For example, frost flowers typically

exhibit a relatively high [Br¯]/[Cl¯] ratio (110) due to the wicking of bromide-enriched brine up

the ice skeleton. If there is no separation, a bulk analysis of a melted sample will not reveal the

chemical fractionation that may exist in the frozen state.

21

Since seawater pH is naturally alkaline and buffered by the carbonate system, an

important consideration is the effect that the precipitation of carbonate salts has on brine pH,

especially given that R1 is thought to be shut-off at a seawater pH (27). Sander et al. (111) were

the first to propose that the precipitation of calcium carbonate could allow for the removal of

brine alkalinity, making it possible for sea ice surfaces to become acidified by atmospheric trace

acids and thereby triggering ‘bromine explosion’ chemistry. Subsequently, Morin et al. (112)

showed, via a modeling study, that the identity of the precipitating salt, either calcite (CaCO3 at

271 K) or ikaite (CaCO3·6H2O at 268.5 K), should determine the ultimate reduction in brine

alkalinity, with no reduction below sea water values for ikaite precipitation. Recently,

Dieckmann et al. (113, 114) identified ikaite as the dominant calcium carbonate polymorph in

both the Arctic and Antarctic. A better understanding of the role of carbonate precipitation in

pH buffering at frozen sea ice surfaces is still required.

1.2.3.3. Deposition and Post-Freezing Processing

As mentioned earlier, several different ice and snow substrates exist. These substrates

vary in specific surface area, density, total ionic content, specific ionic content and in their

spatial and temporal extent. They are also formed via different pathways. Contrast, for

example, sea ice, which is formed when high salinity seawater freezes vs. freshly falling snow,

which is formed in the atmosphere via the condensation of water vapour onto existing ice

particles (with the concurrent uptake up trace acids such as HCl or HNO3)(80). Once formed,

these substrates are subject to dry and wet deposition of atmospheric trace species and aerosols

(which will firstly affect the air-ice interface). How does deposition of trace atmospheric species affect

the air-ice interface? How do deposited species contribute to surface disordering? Do they remain at the

air-ice interface, diffuse into the bulk, or migrate along grain boundaries? Furthermore, ice present as

snow in a snowpack is subject to temperature gradients, which leads to snow metamorphism

(local melting and re-freezing, sublimation and condensation etc.)(115). Thus contaminants and

solutes will likely be remobilised following deposition. More work is required to understand

how deposition vs. exclusion processes influence concentration and pH at the air-ice interface

and the LRs in/on ice.

22

1.3. Techniques for Investigating the Air -Ice Interface

1.3.1. Overview and Challenges

At this point it should be clear that, on a microscopic level, snow and ice are highly

heterogeneous. As described above, this heterogeneity is largely due to the presence of LRs

which may be present at the air-ice interface or within the bulk ice sample (grain boundaries,

triple junctions, liquid pockets etc). Since surface properties (concentration of reagents, pH, etc.)

may be significantly different from bulk properties, and because we currently lack the ability to

confidently predict surface properties from (readily measured) bulk properties, it is desirable to

measure surface properties directly. This requires the development of surface-sensitive

techniques that are appropriate for ice surfaces.

Directly investigating the air-ice interface is difficult since the QLL/QBL/LR that is

present is microscopically thin. The relatively high vapour pressure of water over ice makes the

use of traditional surface-science techniques such as x-ray techniques (XPS and NEXAFS),

molecular beam scattering and photoelectron spectroscopy difficult, especially at warmer

temperatures (although XPS and NEXAFS have recently been used to study the ice surface at

pressures up to a few mbar, and at temperatures nearing the melting point of ice (101)). Other

techniques such as ellipsometry and atomic-force microscopy (AFM), which have been used to

probe the air-ice interface, lack molecular specificity. Sum-frequency generation spectroscopy

(SFG) is very surface-sensitive, but it is also experimentally challenging. Furthermore, it has

thus far only been used to study the water structure at the air-ice interface (74, 116). Glancing-

angle spectroscopic techniques (glancing-angle Raman spectroscopy and glancing-angle laser-

induced fluorescence (LIF)) were pioneered in the Donaldson group for the study of the air-

aqueous and air-ice interface (75, 82, 117). A large focus of the current thesis was the novel

application of these techniques to the air-ice interface. In section 1.3.2.1 I outline the basic

principles of glancing-angle spectroscopy, focusing particularly on glancing-angle Raman

spectroscopy since it is relatively new technique. In Section 1.3.2.2 I highlight its potential for

the quantification of solutes at the air-aqueous interface and in Section 1.3.2.3 I address its

23

surface-sensitivity. Glancing-angle LIF has already been used extensively for the air-ice

interface (82, 85, 118, 119) and so is not described in detail here.

1.3.2. Glancing-Angle Spectroscopy

1.3.2.1. Basic Principles

The glancing-angle LIF approach was first described by Mmereki and Donaldson (117)

and the glancing-angle Raman approach was first described by Kahan et al. (75). In explaining

this technique, the terms ‘surface’ or ‘surface region’ refer to the surface region to which the

glancing-angle probe is sensitive (~50 – 100 nm, see Section 1.3.2.3). In brief, fluorescence or

Raman scattering is induced at the sample surface by a linearly polarized laser beam. By

impinging the surface at a very shallow, glancing angle (> 87 from the surface normal) the

outermost surface region is selectively interrogated. The technique relies on the fact that at these

high angles of incidence, the majority of the input radiation will be reflected at the interface

(e.g., at an incident angle of 87° from the surface normal, 67% of p-polarized (polarized in the

plane of incidence) and 80% of s-polarized (polarized perpendicular to the plane of incidence)

input radiation will be reflected) (75). In the case of glancing-angle LIF, a high surface

sensitivity is additionally achieved through the use a surface-active fluorescent probe

molecules.

Early work using glancing-angle Raman focused on studying the shape of the OH-

stretching band of water, both at the liquid water surface and at the frozen water surface (75).

Glancing-angle Raman spectra acquired at the air-liquid interface show an OH-stretching

signature indicative of reduced hydrogen bonding there. Furthermore, the shape of the OH-

stretching band can be tuned from more ‘surface-like’ to more ‘bulk-like’ in nature by

decreasing the angle of incidence (less grazing)(75). In general, we find that the OH-stretching

peak intensity is significantly enhanced at the surface (with respect to solute peaks). As shown

in Figure 1.4, glancing-angle Raman spectra acquired at the liquid D2O surface exposed to the

ambient atmosphere have a significant OH-stretching peak in addition to the OD-stretch, due to

isotopic exchange with gas-phase H2O at the surface; bulk Raman spectra of the same sample do

24

not show this OH feature. Wren and Donaldson (120) also show that the intensity of the OH-

stretching band of water, measured using glancing-angle Raman spectroscopy, is sensitive to

the presence of other species there, although the exact mechanism responsible for the observed

intensity decrease is not well understood. Nevertheless, this phenomenon was exploited to

investigate the heterogeneous ozonation of frozen halide solutions (Chapter Five), which leads

to the formation of surface active halogens/trihalide ions (18).

Figure 1.4. Raman spectra of a liquid D2O sample exposed to ambient air. The spectrum of a bulk

sample is the dashed red line and the glancing-angle spectrum from the surface region is the

solid blue line. The OD-stretch and OH-stretch appear in the spectrum around ~2400 and ~3400

cm-3, respectively. Source: Figure 1 from Wren and Donaldson (121).

1.3.2.2. Glancing-Angle Raman for Solutes

Recently, the glancing-angle Raman technique was extended to the study of solutes at the

air-aqueous interface. Although it is difficult to quantify solute concentrations at surfaces using

absolute Raman intensities, surface concentration profiles can be obtained as a function of bulk

concentrations. In this approach, water features (the OH-stretching mode, or the bending mode)

are used for normalization. In an early study (120), an adsorption isotherm for dimethyl

sulfoxide (DMSO), which is known to exhibit surface activity to the water surface was obtained.

In that study the intensity of the -CSC bending mode (~700 cm-1) was tracked as a function of

bulk DMSO concentration. The DMSO adsorption isotherm obtained using glancing-angle

25

Raman spectroscopy is plotted with adsorption isotherms obtained by two other studies

Figure 1.5. DMSO adsorption isotherms: Wren and Donaldson (120) glancing angle Raman using

intensity of the ν-CSC (yellow triangles); Allen et al. (123) surface tension (green circles); Tarbuck

and Richmond (122) surface tension (red squares). All data have been normalized to the high

concentration glancing-angle Raman data. Source: Figure 2 from Wren and Donaldson (121).

Figure 1.6. Nitrate symmetric stretching peak (ν -sym NO3¯ ) intensity as a function of

calculated [NO3¯]bulk (=[HNO3]tot × αbulk): bulk data (black circles) and glancing-angle Raman data

(red triangles) for HNO3, and glancing-angle Raman data for KNO3 (green squares). Each point

represents the average of 3–5 trials and the vertical error bars represent the standard deviation of

the average. Source: Figure 7 from Wren and Donaldson (121).

26

(122, 123) using surface tension methods in Figure 1.5. The good agreement among the datasets

suggests that the technique is sensitive to an interfacial region similar to that probed by surface

tension measurements. Glancing-angle Raman spectroscopy was also used to study the extent

of nitric acid (HNO3) dissociation (α) at the air-aqueous interface as well as to study the surface

affinity of nitrate (NO3¯) (121). Figure 1.6 shows the Raman intensity for the ν-sym (NO3¯) mode

(the nitrate symmetric stretch) as a function of calculated bulk [NO3¯] for HNO3(aq) and

KNO3(aq) solutions. In the bulk (black circles), the relationship is linear as expected. The

isotherm obtained at the surface (red triangles – HNO3 and green squares – KNO3) agrees with

the bulk isotherm for low bulk concentrations but deviates at high bulk concentrations; the

observed leveling off at higher bulk concentrations is indicative of a saturation in the surface

concentration. This result is presented here to provide further evidence that glancing-angle

Raman spectroscopy can be used to probe solutes at the air-aqueous interface. In Chapter Two,

this approach was used to study the exclusion of nitrate to the air-ice interface during freezing.

1.3.2.3. Surface Sensitivity

Studies from the Donaldson group suggest that the glancing-angle Raman technique has a

probe depth of ~ 50 – 100 nm. This estimate is based on the observations mentioned above, as

well as on observations which are mentioned in greater detail in Kahan et al. (75). In particular,

Kahan et al. (75) measured the water OH-spectrum at the surface of liquid dodecane under

conditions where there are ≤50 monolayers of water present (124). As well, Kahan et al. (75)

report a decrease in the OD:OH signal in glancing-angle Raman spectra acquired at the D2O ice

surface due to the condensation of water vapor. Figure 1.4 illustrates the situation at the D2O

surface exposed to ambient conditions. Clearly, this technique is less surface sensitive than

many of the surface sensitive techniques mentioned previously (e.g., SFG, XPS). Nonetheless, as

illustrated by Figure 1.4 and 1.6, it is sensitive to a surface region which is distinctly different

from the bulk, and thus can be used to gain insight into the nature of atmospheric surfaces.

Note that for experiments on ice, the LR is expected to be thicker than our probe depth (i.e., we

do not expect to be probing significant contributions from bulk ice).

27

1.4. Summary of Thesis Objectives

Air-ice chemical interactions can have a significant influence on the composition and

oxidative capacity of the overlying boundary layer. However our ability to understand and

model air-ice chemical interactions is impeded by our incomplete understanding of the air-ice

interface itself. The specific knowledge gaps addressed in this thesis are shown in Figure 1.7

and summarized briefly here.

Figure 1.7. A cartoon of ice and its associated ‘liquid’ or disordered regions. Motivating questions

and their associated Chapters are shown.

As described above, the environmental air-ice interface is ‘wetted’ by a disordered or

liquid-like region (LR) whose properties are not well known but are additionally dependent on

conditions such as temperature and total ionic content. However, as summarized in Figure 1.7,

the LR at the air-ice interface represents only one of the possible LRs on ice (the others being

grain boundaries, triple junctions and liquid pockets). Chemical reagents are thought to reside

in/at the LRs due to exclusion during freezing, or deposition and subsequent migration. Of

28

particular relevance to atmospheric chemistry is the concentration and speciation of chemical

reagents at the air-ice interface. This information is required in order to interpret, model or

predict chemistry at the ice surface (it will also help establish the validity of applying aqueous

phase parameters to the ice surface). To what extent can equilibrium thermodynamic models be used

to predict the concentration of species at the air-ice interface? Can we relate surface concentrations to

bulk concentrations? Is it valid to assume that solutes are totally excluded to the LR that exists at the air-

ice interface? Specifically, is the nitrate ion (whose photochemistry on ice/snow is poorly understood)

excluded to the air-ice interface? These questions are addressed in Chapter Two (Exclusion of

Nitrate to the Air-Ice Interface).

Knowledge of the proton concentration (or more specifically the proton activity, or pH)

at the air-ice interface is crucial to understanding the pH dependencies of chemical reactions

occurring on ice surfaces (such as halogen activation or HONO production). Are protons excluded

during freezing (i.e. how does surface pH change during freezing)? How do trace atmospheric acids affect

surface pH? These questions are addressed in Chapter Three (Laboratory Study of pH at the Air-

Ice Interface) and Chapter Four (How Does the Deposition of Gas Phase Species Affect pH at

Salty Interfaces?). Additionally, these chapters address the potential influence of deposition vs.

exclusion on surface concentrations. The difference between the air-ice interface found at the

‘pure’ ice and ‘salty’ ice surface is also discussed, along with the validity of describing the LR at

the air-ice interface as a thin, cold, liquid layer. The question of pH buffering at the sea ice

surface (which may or may not be influenced by carbonate precipitation and has significant

implications for halogen activation chemistry), is addressed in Chapter Four.

The final two Chapters focus on better understanding halogen activation chemistry from

frozen sea ice substrates. In Chapter Five (Spectroscopic Studies of the Heterogeneous Reaction

Between O3(g) and Halides at the Surface of Frozen Salt Solutions), surface-sensitive techniques

are used to directly investigate the dark ozonation of frozen halide solutions as a source of seed

halogen. The influence of interfacial chemistry and freeze-concentration effects are addressed.

The significance of the ozonation reaction for iodine activation is also discussed. The research

29

presented in Chapter Six (Photochemical Bromine and Chlorine Activation from Artificial Saline

Snow) does not focus specifically on the air-ice interface. Rather the role of actinic radiation on

halogen activation, particularly chlorine activation, is studied in controlled laboratory

experiments with gas phase detection of dihalogens. The observed influence of substrate surface

area, pre-freezing pH, pre-freezing concentration and experiment temperature (spanning the

NaCl-water eutectic) is used to infer information about the local environment and the chemical

mechanism. The thesis concludes with a summary of the main results and a discussion of future

research directions.

1.5. References

1. A. M. Grannas, A. E. Jones, J. Dibb, M. Ammann, C. Anastasio, H. J. Beine, M. Bergin, J.

Bottenheim, C. S. Boxe, G. Carver, G. Chen, J. H. Crawford, F. Domine, M. M. Frey, M. I.

Guzman, D. E. Heard, D. Helmig, M. R. Hoffmann, R. E. Honrath, L. G. Huey, M. Hutterli, H. W.

Jacobi, P. Klan, B. Lefer, J. McConnell, J. Plane, R. Sander, J. Savarino, P. B. Shepson, W. R.

Simpson, J. R. Sodeau, R. von Glasow, R. Weller, E. W. Wolff, T. Zhu, An overview of snow

photochemistry: evidence, mechanisms and impacts. Atmos. Chem. Phys. 7, 4329-4373 (2007)

2. F. Domine, P. B. Shepson, Air-snow interactions and atmospheric chemistry. Science 297, 1506-

1510 (2002)

3. W. R. Simpson, R. von Glasow, K. Riedel, P. Anderson, P. Ariya, J. Bottenheim, J. Burrows, L. J.

Carpenter, U. Friess, M. E. Goodsite, D. Heard, M. Hutterli, H. W. Jacobi, L. Kaleschke, B. Neff, J.

Plane, U. Platt, A. Richter, H. Roscoe, R. Sander, P. Shepson, J. Sodeau, A. Steffen, T. Wagner, E.

Wolff, Halogens and their role in polar boundary-layer ozone depletion. Atmos. Chem. Phys. 7,

4375-4418 (2007)

4. S. J. Oltmans, Surface Ozone Measurements In Clean-Air. J. Geophys. Res. Oc.Atm. 86, 1174-1180

(1981)

5. J. W. Bottenheim, A. G. Gallant, K. A. Brice, Measurements Of NOy Species And O3 At 82-

Degrees-N Latitude. Geophys. Res. Lett. 13, 113-116 (1986)

6. L. A. Barrie, G. Denhartog, J. W. Bottenheim, S. Landsberger, Anthropogenic Aerosols And Gases

In The Lower Troposphere At Alert Canada In April 1986. J. Atmos. Chem. 9, 101-127 (1989)

7. L. A. Barrie, J. W. Bottenheim, R. C. Schnell, P. J. Crutzen, R. A. Rasmussen, Ozone Destruction

And Photochemical-Reactions At Polar Sunrise In The Lower Arctic Atmosphere. Nature 334, 138-

141 (1988)

8. J. C. McConnell, G. S. Henderson, L. Barrie, J. Bottenheim, H. Niki, C. H. Langford, E. M. J.

Templeton, Photochemical Bromine Production Implicated In Arctic Boundary-Layer Ozone

Depletion. Nature 355, 150-152 (1992)

9. S. M. Fan, D. J. Jacob, Surface Ozone Depletion In Arctic Spring Sustained By Bromine Reactions

On Aerosols. Nature 359, 522-524 (1992)

10. M. Mozurkewich, Mechanisms For The Release Of Halogens From Sea-Salt Particles By Free-

Radical Reactions. J. Geophys. Res. Atmos. 100, 14199-14207 (1995)

11. J. P. D. Abbatt, J. L. Thomas, K. Abrahamsson, C. Boxe, A. Granfors, A. E. Jones, M. D. King, A.

30

Saiz-Lopez, P. B. Shepson, J. Sodeau, W. Toohey, C. Toubin, R. von Glasow, S. N. Wren, X. Yang,

Halogen activation via interactions with environmental ice and snow. Atmos. Chem. Phys. 12, 6237

- 6271 (2012)

12. B. T. Jobson, H. Niki, Y. Yokouchi, J. Bottenheim, F. Hopper, R. Leaitch, Measurements Of C2-C6

Hydrocarbons During The Polar Sunrise 1992 Experiment - Evidence For Cl Atom And Br Atom

Chemistry. J. Geophys. Res. Atmos. 99, 25355-25368 (1994)

13. R. von Glasow, P. J. Crutzen, Model study of multiphase DMS oxidation with a focus on

halogens. Atmos. Chem. Phys. 4, 589-608 (2004)

14. A. Steffen, T. Douglas, M. Amyot, P. Ariya, K. Aspmo, T. Berg, J. Bottenheim, S. Brooks, F.

Cobbett, A. Dastoor, A. Dommergue, R. Ebinghaus, C. Ferrari, K. Gardfeldt, M. E. Goodsite, D.

Lean, A. J. Poulain, C. Scherz, H. Skov, J. Sommar, C. Temme, A synthesis of atmospheric

mercury depletion event chemistry in the atmosphere and snow. Atmos. Chem. Phys. 8, 1445-1482

(2008)

15. C. R. Stephens, P. B. Shepson, A. Steffen, J. W. Bottenheim, J. Liao, L. G. Huey, E. Apel, A.

Weinheimer, S. R. Hall, C. Cantrell, B. C. Sive, D. J. Knapp, D. D. Montzka, R. S. Hornbrook, The

relative importance of chlorine and bromine radicals in the oxidation of atmospheric mercury at

Barrow, Alaska. J. Geophys. Res. Atmos. 117, D00R11 (16 pp) (2012)

16. K. W. Oum, M. J. Lakin, B. J. Finlayson-Pitts, Bromine activation in the troposphere by the dark

reaction of O3 with seawater ice. Geophys. Res. Lett. 25, 3923-3926 (1998)

17. N. W. Oldridge, J. P. D. Abbatt, Formation of Gas-Phase Bromine from Interaction of Ozone with

Frozen and Liquid NaCl/NaBr Solutions: Quantitative Separation of Surficial Chemistry from

Bulk-Phase Reaction. J. Phys. Chem. A 115, 2590-2598 (2011)

18. S. N. Wren, T. F. Kahan, K. B. Jumaa, D. J. Donaldson, Spectroscopic studies of the heterogeneous

reaction between O3(g) and halides at the surface of frozen salt solutions. J. Geophys. Res.

Atmos.115, D16309 (8 pp) (2010)

19. S. J. Sjostedt, J. P. D. Abbatt, Release of gas-phase halogens from sodium halide substrates:

heterogeneous oxidation of frozen solutions and desiccated salts by hydroxyl radicals. Environ.

Res. Lett. 3, 045007 (7 pp) (2008)

20. J. Abbatt, N. Oldridge, A. Symington, V. Chukalovskiy, R. D. McWhinney, S. Sjostedt, R. A. Cox,

Release of Gas-Phase Halogens by Photolytic Generation of OH in Frozen Halide-Nitrate

Solutions: An Active Halogen Formation Mechanism? J. Phys. Chem. A 114, 6527-6533 (2010)

21. D. I. Reeser, C. George, D. J. Donaldson, Photooxidation of Halides by Chlorophyll at the Air-Salt

Water Interface. J. Phys. Chem. A 113, 8591-8595 (2009)

22. P. O' Driscoll, K. Lang, N. Minogue, J. Sodeau, Freezing halide ion solutions and the release of

interhalogens to the atmosphere. J. Phys. Chem. A 110, 4615-4618 (2006)

23. D. O'Sullivan, J. R. Sodeau, Freeze-Induced Reactions: Formation of Iodine-Bromine Interhalogen

Species from Aqueous Halide Ion Solutions. J. Phys. Chem. A 114, 12208-12215 (2010)

24. J. W. Adams, N. S. Holmes, J. N. Crowley, Uptake and reaction of HOBr on frozen and dry

NaCl/NaBr surfaces between 253 and 233 K. Atmos. Chem. Phys. 2, 79-91 (2002)

25. U. Kirchner, T. Benter, R. N. Schindler, Experimental verification of gas phase bromine

enrichment in reactions of HOBr with sea salt doped ice surfaces. Phys. Chem. Chem. Phys. 101,

975-977 (1997)

26. A. K. Huff, J. P. D. Abbatt, Gas-phase Br2 production in heterogeneous reactions of Cl2, HOCl,

and BrCl with halide-ice surfaces. J. Phys. Chem. A 104, 7284-7293 (2000)

27. S. Fickert, J. W. Adams, J. N. Crowley, Activation of Br2 and BrCl via uptake of HOBr onto

aqueous salt solutions. J. Geophys. Res. Atmos. 104, 23719-23727 (1999)

28. A. Saiz-Lopez, A. S. Mahajan, R. A. Salmon, S. J. B. Bauguitte, A. E. Jones, H. K. Roscoe, J. M. C.

Plane, Boundary layer halogens in coastal Antarctica. Science 317, 348-351 (2007)

31

29. U. Friess, T. Wagner, I. Pundt, K. Pfeilsticker, U. Platt, Spectroscopic measurements of

tropospheric iodine oxide at Neumayer Station, Antarctica. Geophys. Res. Lett. 28, 1941-1944 (2001)

30. U. Friess, T. Deutschmann, B. S. Gilfedder, R. Weller, U. Platt, Iodine monoxide in the Antarctic

snowpack. Atmos. Chem. Phys. 10, 2439-2456 (2010)

31. G. Honninger, U. Platt, Observations of BrO and its vertical distribution during surface ozone

depletion at Alert. Atmos. Environ. 36, 2481-2489 (2002)

32. A. Saiz-Lopez, J. M. C. Plane, A. R. Baker, L. J. Carpenter, R. von Glasow, J. C. G. Martin, G.

McFiggans, R. W. Saunders, Atmospheric Chemistry of Iodine. Chem. Rev. 112, 1773-1804 (2012)

33. J. Liao, L. G. Huey, Z. Liu, D. J. Tanner, C. A. Cantrell, J. J. Orlando, F. M. Flocke, P. B. Shepson,

A. J. Weinheimer, S. R. Hall, H. J. B. Beine, Y. Wang, E. D. Ingall, C. R. S. Stephens, R. S. H.

Hornbrook, E. Apel, A. F. Fried, L. M. Mauldin, J. N. Smith, R. M. S. Staebler, J. A. Neuman, J. B.

N. Nowak, Observations of very high levels of molecular chlorine at Barrow, AK. Nat. Geosci.

Submitted, (2013)

34. R. E. Honrath, M. C. Peterson, S. Guo, J. E. Dibb, P. B. Shepson, B. Campbell, Evidence of NOx

production within or upon ice particles in the Greenland snowpack. Geophys. Res. Lett. 26, 695-698

(1999)

35. H. J. Beine, R. E. Honrath, F. Domine, W. R. Simpson, J. D. Fuentes, NOx during background and

ozone depletion periods at Alert: Fluxes above the snow surface. J. Geophys. Res. Atmos. 107, 4584

(12 pp) (2002)

36. X. L. Zhou, H. J. Beine, R. E. Honrath, J. D. Fuentes, W. Simpson, P. B. Shepson, J. W. Bottenheim,

Snowpack photochemical production of HONO: a major source of OH in the Arctic boundary

layer in springtime. Geophys. Res. Lett. 28, 4087-4090 (2001)

37. J. E. Dibb, M. Arsenault, M. C. Peterson, R. E. Honrath, Fast nitrogen oxide photochemistry in

Summit, Greenland snow. Atmos. Environ. 36, 2501-2511 (2002)

38. A. E. Jones, R. Weller, A. Minikin, E. W. Wolff, W. T. Sturges, H. P. McIntyre, S. R. Leonard, O.

Schrems, S. Bauguitte, Oxidized nitrogen chemistry and speciation in the Antarctic troposphere.

J. Geophys. Res. Atmos. 104, 21355-21366 (1999)

39. A. E. Jones, R. Weller, E. W. Wolff, H. W. Jacobi, Speciation and rate of photochemical NO and

NO2 production in Antarctic snow. Geophys. Res. Lett. 27, 345-348 (2000)

40. D. Davis, J. B. Nowak, G. Chen, M. Buhr, R. Arimoto, A. Hogan, F. Eisele, L. Mauldin, D. Tanner,

R. Shetter, B. Lefer, P. McMurry, Unexpected high levels of NO observed at South Pole. Geophys.

Res. Lett. 28, 3625-3628 (2001)

41. R. E. Honrath, M. C. Peterson, M. P. Dziobak, J. E. Dibb, M. A. Arsenault, S. A. Green, Release of

NOx from sunlight-irradiated midlatitude snow. Geophys. Res. Lett. 27, 2237-2240 (2000)

42. M. Legrand, P. Mayewski, Glaciochemistry of polar ice cores: A review. Rev. Geophys. 35, 219-243

(1997)

43. E. S. N. Cotter, A. E. Jones, E. W. Wolff, S. J. B. Bauguitte, What controls photochemical NO and

NO2 production from Antarctic snow? Laboratory investigation assessing the wavelength and

temperature dependence. J. Geophys. Res. Atmos. 108, 4147 (10 pp) (2003)

44. Y. Dubowski, A. J. Colussi, M. R. Hoffmann, Nitrogen dioxide release in the 302 nm band

photolysis of spray-frozen aqueous nitrate solutions. Atmospheric implications. J. Phys. Chem. A

105, 4928-4932 (2001)

45. Y. Dubowski, A. J. Colussi, C. Boxe, M. R. Hoffmann, Monotonic increase of nitrite yields in the

photolysis of nitrate in ice and water between 238 and 294 K. J. Phys. Chem. A 106, 6967-6971

(2002)

46. L. Chu, C. Anastasio, Quantum yields of hydroxyl radical and nitrogen dioxide from the

photolysis of nitrate on ice. J. Phys. Chem. A 107, 9594-9602 (2003)

47. L. Chu, C. Anastasio, Temperature and wavelength dependence of nitrite photolysis in frozen

32

and aqueous solutions. Environ. Sci. Technol. 41, 3626-3632 (2007)

48. C. S. Boxe, A. J. Colussi, M. R. Hoffmann, J. G. Murphy, P. J. Wooldridge, T. H. Bertram, R. C.

Cohen, Photochemical production and release of gaseous NO2 from nitrate-doped water ice. J.

Phys. Chem. A 109, 8520-8525 (2005)

49. H. W. Jacobi, B. Hilker, A mechanism for the photochemical transformation of nitrate in snow. J.

Photochem. Photobiol. A. 185, 371-382 (2007)

50. O. Abida, H. D. Osthoff, On the pH dependence of photo-induced volatilization of nitrogen

oxides from frozen solutions containing nitrate. Geophys. Res. Lett. 38, L16808 (5 pp) (2011)

51. C. S. Boxe, A. J. Colussi, M. R. Hoffmann, I. M. Perez, J. G. Murphy, R. C. Cohen, Kinetics of NO

and NO2 evolution from illuminated frozen nitrate solutions. J. Phys. Chem. A 110, 3578-3583

(2006)

52. H. J. Beine, F. Domine, W. Simpson, R. E. Honrath, R. Sparapani, X. L. Zhou, M. King, Snow-pile

and chamber experiments during the Polar Sunrise Experiment 'Alert 2000': exploration of

nitrogen chemistry. Atmos. Environ. 36, 2707-2719 (2002)

53. R. E. Honrath, S. Guo, M. C. Peterson, M. P. Dziobak, J. E. Dibb, M. A. Arsenault, Photochemical

production of gas phase NOx from ice crystal NO3‾. J. Geophys. Res. Atmos. 105, 24183-24190 (2000)

54. A. E. Jones, R. Weller, P. S. Anderson, H. W. Jacobi, E. W. Wolff, O. Schrems, H. Miller,

Measurements of NOx emissions from the Antarctic snowpack. Geophys. Res. Lett. 28, 1499-1502

(2001)

55. J. Mack, J. R. Bolton, Photochemistry of nitrite and nitrate in aqueous solution: a review. J.

Photochem. Photobiol. A 128, 1-13 (1999)

56. J. J. Jankowski, D. J. Kieber, K. Mopper, Nitrate and nitrite ultraviolet actinometers. Photochem.

Photobiol. 70, 319-328 (1999)

57. C. S. Boxe, A. Saiz-Lopez, Multiphase modeling of nitrate photochemistry in the quasi-liquid

layer (QLL): implications for NOx release from the Arctic and coastal Antarctic snowpack. Atmos.

Chem. Phys. 8, 4855-4864 (2008)

58. J. Bock, H. W. Jacobi, Development of a Mechanism for Nitrate Photochemistry in Snow. J. Phys.

Chem. A 114, 1790-1796 (2010)

59. H. W. Jacobi, T. Annor, E. Quansah, Investigation of the photochemical decomposition of nitrate,

hydrogen peroxide, and formaldehyde in artificial snow. J. Photochem. Photobiol. A 179, 330-338

(2006)

60. H. J. Beine, A. Amoroso, F. Domine, M. D. King, M. Nardino, A. Ianniello, J. L. France,

Surprisingly small HONO emissions from snow surfaces at Browning Pass, Antarctica. Atmos.

Chem. Phys. 6, 2569-2580 (2006)

61. H. J. Beine, F. Domine, A. Ianniello, M. Nardino, I. Allegrini, K. Teinila, R. Hillamo, Fluxes of

nitrates between snow surfaces and the atmosphere in the European high Arctic. Atmos. Chem.

Phys. 3, 335-346 (2003)

62. C. Anastasio, L. Chu, Photochemistry of Nitrous Acid (HONO) and Nitrous Acidium Ion

(H2ONO+) in Aqueous Solution and Ice. Environ. Sci. Technol. 43, 1108-1114 (2009)

63. S. Hellebust, T. Roddis, J. R. Sodeau, Potential role of the nitroacidium ion on HONO emissions

from the snowpack. J. Phys. Chem. A 111, 1167-1171 (2007)

64. H. Beine, A. J. Colussi, A. Amoroso, G. Esposito, M. Montagnoli, M. R. Hoffmann, HONO

emissions from snow surfaces. Environ. Res. Lett. 3, 045005 (6 pp) (2008)

65. S. Hellebust, D. O'Sullivan, J. R. Sodeau, Protonated Nitrosamide and Its Potential Role in the

Release of HONO from Snow and Ice in the Dark. J. Phys. Chem. A 114, 11632-11637 (2010)

66. Y. Li, G. A. Somorjai, Surface premelting of ice. J. Phys. Chem. C 111, 9631-9637 (2007)

67. J. G. Dash, A. W. Rempel, J. S. Wettlaufer, The physics of premelted ice and its geophysical

consequences. Rev. Mod. Phys. 78, 695-741 (2006)

33

68. R. Rosenberg, Why is ice slippery? Phys. Today 58, 50-55 (2005)

69. A. Doppenschmidt, H. J. Butt, Measuring the thickness of the liquid-like layer on ice surfaces

with atomic force microscopy. Langmuir 16, 6709-6714 (2000)

70. D. Beaglehole, D. Nason, Transition Layer On The Surface On Ice. Surf. Sci. 96, 357-363 (1980)

71. Y. Furukawa, M. Yamamoto, T. Kuroda, Ellipsometric Study Of The Transition Layer On The

Surface Of An Ice Crystal. J. Cryst. Growth 82, 665-677 (1987)

72. A. Kouchi, Y. Furukawa, T. Kuroda, X-Ray-Diffraction Pattern Of Quasi-Liquid Layer On Ice

Crystal-Surface. J. Phys. Paris. 48, 675-677 (1987)

73. H. Bluhm, D. F. Ogletree, C. S. Fadley, Z. Hussain, N. Salmeron, The premelting of ice studied

with photoelectron spectroscopy. J. Phys. -Condens. Mat. 14, L227-L233 (2002)

74. X. Wei, P. B. Miranda, Y. R. Shen, Surface vibrational spectroscopic study of surface melting of

ice. Phys. Rev. Lett. 86, 1554-1557 (2001)

75. T. F. Kahan, J. P. Reid, D. J. Donaldson, Spectroscopic probes of the quasi-liquid layer on ice. J.

Phys. Chem. A 111, 11006-11012 (2007)

76. H. Dosch, A. Lied, J. H. Bilgram, Glancing-Angle X-Ray-Scattering Studies Of The Premelting Of

Ice Surfaces. Surf. Sci. 327, 145-164 (1995)

77. A. Lied, H. Dosch, J. H. Bilgram, Surface Melting Of Ice I(H) Single-Crystals Revealed By

Glancing Angle X-Ray-Scattering. Phys. Rev. Lett. 72, 3554-3557 (1994)

78. V. Sadtchenko, G. E. Ewing, A new approach to the study of interfacial melting of ice: infrared

spectroscopy. Can. J. Phys. 81, 333-341 (2003)

79. M. A. Carignano, P. B. Shepson, I. Szleifer, Ions at the ice/vapor interface. Chem. Phys. Lett. 436,

99-103 (2007)

80. F. Domine, J. Bock, D. Voisin, D. J. Donaldson, Can We Model Snow Photochemistry? Problems

With Current Approaches. J. Phys. Chem. A, DOI: 10.1021/jp3123314 (2013)

81. J. L. Thomas, J. Stutz, B. Lefer, L. G. Huey, K. Toyota, J. E. Dibb, R. von Glasow, Modeling

chemistry in and above snow at Summit, Greenland - Part 1: Model description and results.

Atmos. Chem. Phys. 11, 4899-4914 (2011)

82. T. F. Kahan, D. J. Donaldson, Photolysis of polycyclic aromatic hydrocarbons on water and ice

surfaces. J. Phys. Chem. A 111, 1277-1285 (2007)

83. T. F. Kahan, R. Zhao, K. B. Jumaa, D. J. Donaldson, Anthracene Photolysis in Aqueous Solution

and Ice: Photon Flux Dependence and Comparison of Kinetics in Bulk Ice and at the Air-Ice

Interface. Environ. Sci. Technol. 44, 1302-1306 (2010)

84. T. F. Kahan, D. J. Donaldson, Benzene Photolysis on Ice: Implications for the Fate of Organic

Contaminants in the Winter. Environ. Sci. Technol. 44, 3819-3824 (2010)

85. T. F. Kahan, R. Zhao, D. J. Donaldson, Hydroxyl radical reactivity at the air-ice interface. Atmos.

Chem. Phys. 10, 843-854 (2010).

86. T. P. Liyana-Arachchi, K. T. Valsaraj, F. R. Hung, Molecular Simulation Study of the Adsorption

of Naphthalene and Ozone on Atmospheric Air/Ice Interfaces. J. Phys. Chem. A 115, 9226-9236

(2011)

87. V. F. McNeill, A. M. Grannas, J. P. D. Abbatt, M. Ammann, P. Ariya, T. Bartels-Rausch, F.

Domine, D. J. Donaldson, M. I. Guzman, D. Heger, T. F. Kahan, P. Klan, S. Masclin, C. Toubin, D.

Voisin, Organics in environmental ices: sources, chemistry, and impacts. Atmos. Chem. Phys. 12,

9653-9678 (2012)

88. D. Ardura, T. F. Kahan, D. J. Donaldson, Self-Association of Naphthalene at the Air-Ice Interface.

J. Phys. Chem. A 113, 7353-7359 (2009)

89. D. Heger, P. Klan, Interactions of organic molecules at grain boundaries in ice: A solvatochromic

analysis. J. Photochem. Photobiol. A 187, 275-284 (2007)

90. T. Bartels-Rausch, H.-W. Jacobi, T. F. Kahan, J. L. Thomas, E. S. Thomson, J. P. D. Abbatt, M.

34

Ammann, J. R. Blackford, H. Bluhm, C. Boxe, F. Domine, M. M. Frey, I. Gladich, M. I. Guzman, D.

Heger, T. Huthwelker, P. Klan, W. F. Kuhs, M. H. Kuo, S. Maus, S. G. Moussa, V. F. McNeil, J. T.

Newberg, J. B. C. Pettersson, M. Roeselova, J. R. Sodeau, Relationship between snow

microstructure and physical and chemical processes. Atmos. Chem. Phys. Discuss.12, 30409-30541

(2012)

91. V. F. McNeill, T. Loerting, F. M. Geiger, B. L. Trout, M. J. Molina, Hydrogen chloride-induced

surface disordering on ice. Proc. Natl. Acad. Sci. U. S. A. 103, 9422-9427 (2006)

92. H. Cho, P. B. Shepson, L. A. Barrie, J. P. Cowin, R. Zaveri, NMR investigation of the quasi-brine

layer in ice/brine mixtures. J. Phys. Chem. B 106, 11226-11232 (2002)

93. M. H. Kuo, S. G. Moussa, V. F. McNeill, Modeling interfacial liquid layers on environmental ices.

Atmos. Chem. Phys. 11, 9971-9982 (2011)

94. T. Koop, A. Kapilashrami, L. T. Molina, M. J. Molina, Phase transitions of sea-salt/water mixtures

at low temperatures: Implications for ozone chemistry in the polar marine boundary layer. J.

Geophys. Res. Atmos. 105, 26393-26402 (2000)

95. A. M. Grannas, A. R. Bausch, K. M. Mahanna, Enhanced aqueous photochemical reaction rates

after freezing. J. Phys. Chem. A 111, 11043-11049 (2007)

96. P. O'Driscoll, N. Minogue, N. Takenaka, J. Sodeau, Release of nitric oxide and iodine to the

atmosphere from the freezing of sea-salt aerosol components. J. Phys. Chem. A 112, 1677-1682

(2008)

97. N. Takenaka, A. Ueda, Y. Maeda, Acceleration Of The Rate Of Nitrite Oxidation By Freezing In

Aqueous-Solution. Nature 358, 736-738 (1992)

98. N. Takenaka, A. Ueda, T. Daimon, H. Bandow, T. Dohmaru, Y. Maeda, Acceleration mechanism

of chemical reaction by freezing: The reaction of nitrous acid with dissolved oxygen. J. Phys.

Chem. 100, 13874-13884 (1996)

99. J. L. Thomas, J. E. Dibb, L. G. Huey, J. Liao, D. Tanner, B. Lefer, R. von Glasow, J. Stutz, Modeling

chemistry in and above snow at Summit, Greenland - Part 2: Impact of snowpack chemistry on

the oxidation capacity of the boundary layer. Atmos. Chem. Phys. 12, 6537-6554 (2012)

100. D. Heger, J. Jirkovsky, P. Klan, Aggregation of methylene blue in frozen aqueous solutions

studied by absorption spectroscopy. J. Phys. Chem. A 109, 6702-6709 (2005)

101. A. Krepelova, T. Huthwelker, H. Bluhm, M. Ammann, Surface Chemical Properties of Eutectic

and Frozen NaCl Solutions Probed by XPS and NEXAFS. Chemphyschem 11, 3859-3866 (2010)

102. V. L. Bronshteyn, A. A. Chernov, Freezing Potentials Arising On Solidification Of Dilute

Aqueous-Solutions Of Electrolytes. J. Cryst. Growth. 112, 129-145 (1991)

103. E. J. Workman, S. E. Reynolds, A Suggested Mechanism For The Generation Of Thunderstorm

Electricity. Phys. Rev. 74, 709-709 (1948)

104. P. W. Wilson, A. D. J. Haymet, Effect of Ice Growth Rate on the Measured Workman-Reynolds

Freezing Potential between Ice and Dilute NaCl Solutions. J. Phys. Chem. B 114, 12585-12588

(2010)

105. D. Heger, J. Klanova, P. Klan, Enhanced protonation of cresol red in acidic aqueous solutions

caused by freezing. J. Phys. Chem. B 110, 1277-1287 (2006)

106. N. Takenaka, M. Tanaka, K. Okitsu, H. Bandow, Rise in the pH of an unfrozen solution in ice due

to the presence of NaCl and promotion of decomposition of gallic acids owing to a change in the

pH. J. Phys. Chem.A 110, 10628-10632 (2006)

107. C. Robinson, C. S. Boxe, M. I. Guzman, A. J. Colussi, M. R. Hoffmann, Acidity of frozen

electrolyte solutions. J. Phys. Chem. B 110, 7613-7616 (2006)

108. J. Cheng, C. Soetjipto, M. R. Hoffmann, A. J. Colussi, Confocal Fluorescence Microscopy of the

Morphology and Composition of Interstitial Fluids in Freezing Electrolyte Solutions. J. Phys.

Chem. Lett. 1, 374-378 (2010)

35

109. F. D. Lopez-Hilfiker, K. Constantin, J. P. Kercher, J. A. Thornton, Temperature dependent

halogen activation by N2O5 reactions on halide-doped ice surfaces. Atmos. Chem. Phys. 12, 5237-

5247 (2012)

110. L. E. Kalnajs, L. M. Avallone, Frost flower influence on springtime boundary-layer ozone

depletion events and atmospheric bromine levels. Geophys. Res. Lett. 33, L10810 (3 pp) (2006)

111. R. Sander, J. Burrows, L. Kaleschke, Carbonate precipitation in brine - a potential trigger for

tropospheric ozone depletion events. Atmos. Chem. Phys. 6, 4653-4658 (2006)

112. S. Morin, G. M. Marion, R. von Glasow, D. Voisin, J. Bouchez, J. Savarino, Precipitation of salts in

freezing seawater and ozone depletion events: a status report. Atmos. Chem. Phys. 8, 7317-7324

(2008)

113. G. S. Dieckmann, G. Nehrke, S. Papadimitriou, J. Goettlicher, R. Steininger, H. Kennedy, D. Wolf-

Gladrow, D. N. Thomas, Calcium carbonate as ikaite crystals in Antarctic sea ice. Geophys. Res.

Lett. 35, L08501 (3 pp) (2008)

114. G. S. Dieckmann, G. Nehrke, C. Uhlig, J. Goettlicher, S. Gerland, M. A. Granskog, D. N. Thomas,

Brief Communication: Ikaite (CaCO3·6H2O) discovered in Arctic sea ice. Cryosphere 4, 227-230

(2010)

115. F. Domine, M. Albert, T. Huthwelker, H. W. Jacobi, A. A. Kokhanovsky, M. Lehning, G. Picard,

W. R. Simpson, Snow physics as relevant to snow photochemistry. Atmos. Chem. Phys. 8, 171-208

(2008)

116. X. Wei, P. B. Miranda, C. Zhang, Y. R. Shen, Sum-frequency spectroscopic studies of ice

interfaces. Phys. Rev. B 66, 085401 (13 pp) (2002)

117. B. T. Mmereki, D. J. Donaldson, Laser induced fluorescence of pyrene at an organic coated air-

water interface. Phys. Chem. Chem. Phys. 4, 4186-4191 (2002)

118. T. F. Kahan, D. J. Donaldson, Heterogeneous ozonation kinetics of phenanthrene at the air-ice

interface. Environ. Res. Lett. 3, 045006 (6 pp) (2008)

119. T. F. Kahan, N. O. A. Kwamena, D. J. Donaldson, Different photolysis kinetics at the surface of

frozen freshwater vs. frozen salt solutions. Atmos. Chem. Phys. 10, 10917-10922 (2010)

120. S. N. Wren, D. J. Donaldson, Glancing-angle Raman spectroscopic probe for kinetics at water

surfaces. Phys. Chem. Chem. Phys. 12, 2648-2654 (2010)

121. S. N. Wren, D. J. Donaldson, Glancing-angle Raman study of nitrate and nitric acid at the air-

aqueous interface. Chem. Phys. Lett. 522, 1-10 (2012)

122. T. L. Tarbuck, G. L. Richmond, Adsorption of Organosulfur Species at Aqueous Surfaces:

Molecular Bonding and Orientation. J. Phys. Chem. B 109, 20868 - 20877 (2005)

123. H. C. Allen, D. E. Gragson, G. L. Richmond, Molecular Structure and Adsorption of Dimethyl

Sulfoxide at the Surface of Aqueous Solutions. J. Phys. Chem. B 103, 660 - 666 (1999)

124. E. Demou, H. Visram, D. J. Donaldson, P. A. Makar, Uptake of water by organic films: the

dependence on the film oxidation state. Atmos. Environ. 37, 3529-3537 (2003)

37

C H AP T E R T WO

EXCLUSION OF NITRATE TO THE AIR-ICE INTERFACE DURING

FREEZING

S. N. Wren and D. J. Donaldson

Adapted with permission from S. N. Wren and D. J. Donaldson, Exclusion of nitrate to the air-

ice interface during freezing, Journal of Physical Chemistry Letters, 2, 1967 – 1971 (2011).

DOI: 10.1021/jz2007484. Copyright (2011) American Chemical Society.

Contributions: S. N. Wren performed the experiments, interpreted the results and wrote the

paper. D. J. Donaldson provided critical review and guidance.

38

2.1. Introduction

Chemistry occurring in the snowpack can significantly influence the composition of the

overlying boundary layer (1). Fluxes of NOx (NOx = NO + NO2) and HONO have been

measured over Arctic, Antarctic, and mid-latitude snowpacks and it is now well established

that chemical processes initiated by the photolysis of snowpack nitrate are responsible (see

Grannas et al. (2) and references therein). Although nitrate snowpack photochemistry has been

the subject of several recent laboratory studies (3-10), the mechanism of NOx and HONO

production from frozen media is not well understood. The uptake and release of nitrogen

oxides from sunlit snow is important since these compounds play an important role in

determining the oxidative capacity of the overlying polar boundary layer. Furthermore, since

nitrate (NO3¯ ) is ubiquitous in snow, the ice-core nitrate record is used to reconstruct past

climate and atmospheric composition (11). A good understanding of the chemical cycling of

nitrogen species between the air and ice is therefore required for the accurate interpretation of

the ice-core nitrate data.

During freezing, the majority of solutes are rejected from the growing ice lattice and

become concentrated either at grain boundaries or nodes within the ice crystal, or at the air-ice

interface. Solute exclusion into these liquid regions (LR) may significantly increase their

concentrations over those in the bulk sample, with important consequences for reactions

occurring there. Several aqueous phase bimolecular and heterogeneous reactions have been

shown to be accelerated upon freezing (12-15) with enhanced reaction rates attributed to freeze-

concentration effects. The specific location of rejected solutes (at the surface or in liquid

pockets/grain boundaries) is important for understanding chemistry in or on frozen media since

this will determine the availability of these solutes for reaction and/or exchange with the

atmosphere. Furthermore, there is some evidence to suggest that reactivity at the air-ice

interface is distinctly different from reactivity in the liquid-like pockets that may exist within an

ice crystal (16, 17).

39

Nitrate at the ice surface has been the subject of only a few spectroscopic studies to date

(18, 19). In the present experiments we have used glancing-angle Raman spectroscopy to probe

the exclusion of nitrate anions to the ice surface. This technique (which is sensitive to the upper

~ 50-100 nm of the sample surface (20)) is less sensitive than interface-specific techniques such

as sum frequency generation (SFG) spectroscopy or x-ray photoelectron spectroscopy (XPS)

which have resolution on the order of a few nm, but is nevertheless appropriate for this study

given the large liquid fractions which are expected (vide infra) . This study represents the first

use of glancing-angle Raman spectroscopy to probe solutes at the air-ice interfacial region.


2.2.1. Glancing-Angle Raman Spectroscopy

The reaction chamber and glancing-angle Raman technique have been described in detail

elsewhere (14, 16, 20). The reaction chamber consists of a Teflon box sitting on a cooled copper

base. Quartz windows at the front and back of the chamber allow a laser beam to enter and exit

the chamber. Side ports allow gases to be introduced and ventilated. A stainless steel base-plate

covered the chamber floor and samples were placed on this plate. Raman scattering was

induced at the sample surface using the 355 nm output of a frequency-tripled Nd:YAG laser

(pulse repetition 10 Hz, pulse energy ~0.6 mJ). The laser beam impinged the sample surface at a

glancing angle (> 85 from the surface normal). Raman scattering was collected perpendicular to

the surface using a 7 mm diameter liquid light guide suspended ~5 mm above the sample. The

collected light was imaged onto the entrance slit (0.5 mm diameter) of a ¼ m monochromator

and the transmitted intensity was detected and amplified by a photomultiplier tube. A 355 nm

laser-line long pass filter was placed between the exit slit of the monochromator and the

photomultiplier tube to reduce interference from Rayleigh scattering. The electronic signal was

read out by a digital oscilloscope which averaged the intensity vs. time signal over 64 laser

shots. A LabView program sampled a 50 ns window centred on the intensity vs. time peak, and

this value was stored for subsequent analysis. Raman spectra were collected by scanning the

monochromator in steps (~ 25 cm-1).

40

2.2.2. Aqueous Experiments

Glancing-angle Raman spectroscopy was used to study the surface of Mg(NO3)2(aq)

solutions (0.25 – 1.5 M). Aqueous Mg(NO3)2 (0.05 – 1.50 M) solutions were prepared by

dissolving Mg(NO3)2·6H2O(s) in 18 M deionized water. For these studies, 4 mL of solution

were spread onto a thin piece of stainless steel shimstock resting on the chamber floor such that

the impinging laser beam cleared the front lip of the sample surface (to avoid directly

penetrating into the bulk).

The incident laser beam was vertically polarized. In this geometry we only collect

perpendicular components of the scattered light, allowing us to detect antisymmetric vibrations

and a ‘depolarized’ spectrum (20). The bending mode of water appears in the Raman spectra as

a peak centred around 1600 cm-1(21). All spectra were first baseline corrected by subtracting the

minimum intensity in the 500 – 1600 cm-1 range and then normalized to the OH-bending feature

of water. The area under the Raman peak due to the nitrate symmetric stretch was calculated

using the trapezoidal rule. The area was calculated over a 100 cm-1 window centred around the

peak, and the baseline was re-defined by the boundaries of this window.

The symmetric (-sym) and asymmetric (-asym) nitrate stretching modes appear in the

surface Raman spectra of liquid solutions at ~1000 cm-1 and 1350 cm-1 respectively. The peaks

are slightly red-shifted compared to their reported literature values of 1043 cm-1 and 1370 cm-1

respectively (22). The intensities of the nitrate Raman peaks increase with increasing Mg(NO3)2

concentration as displayed in Figure 2.1. The area under the Raman peak arising from the

nitrate symmetric stretch was determined for each spectrum. Figure 2.2 shows the peak areas as

a function of Mg(NO3)2 concentration. Peak area was found to depend linearly on

[Mg(NO3)2](aq) over this concentration range, consistent with (unpublished) results for

KNO3(aq) obtained by this group. The Raman signal from the water bend is very weak, and so

the pure water spectrum and spectra for [Mg(NO3)2] < 0.25 M suffer from poor signal-noise

ratios. The calibration plot (Figure 2.2) was used to interpret the results acquired at the frozen

surface assuming that the intensity of the nitrate symmetric stretch scales to the intensity of the

41

water bending vibration in the same way on the ice sample surface as on the liquid aqueous

surface. This assumption as addressed in further detail in the Discussion.

Figure 2.1. Glancing-angle Raman spectra acquired at the surface of aqueous Mg(NO3)2 solutions

with bulk concentrations: 0.25 M (long red dash), 0.50 M (medium green dash), 0.75 M (short blue

dash), 1.25 M (orange dash-dot). The spectrum acquired at the surface of pure water is shown as

the solid black line.

Figure 2.2. Area under the -sym NO3¯ peak as a function of Mg(NO3)2 concentration. Each point

represents the average of at least two trials and the error bars represent the associated

uncertainty.

42

2.2.3. Ice Experiments

An aqueous Mg(NO3)2 (0.05 – 1.50 M) solution was prepared by dissolving

Mg(NO3)2·6H2O(s) in 18 M deionized water. A 4 mL aliquot of solution was then spread onto a

thin piece of stainless steel shimstock resting on the chamber floor and the sample was frozen

either rapidly (i.e. within 5 min, with the chamber already cooled to the desired temperature) or

slowly (i.e. over ~ 1 hr, at a rate of ~0.2 K min-1 by gradually reducing the chamber temperature

from room temperature to the desired temperature). Frozen samples were fairly smooth and flat

on top and had rounded edges (surface area ~ 9 - 10 cm2 and thickness ~ 0.3 – 0.5 cm). Samples

that were frozen rapidly sometimes formed two small bumps at the sample surface with visible

grain boundaries spreading from these peaks. Samples that were frozen slowly were more

homogeneous and macroscopically smooth. Both slow and fast freezing resulted in

polycrystalline ice samples that were slightly whitish in colour with small pockets and grain

boundaries clearly visible throughout. Samples were positioned below the liquid light guide

such that the impinging laser beam cleared the front edge of the sample.

The relative humidity of the chamber was not controlled in these experiments and so

condensation of water vapour within the chamber could result in a growing ice surface

(potentially covering rejected nitrate molecules and thereby reducing their Raman intensity). To

test for this possibility, in a few experiments, a slow flow of dry nitrogen gas (30 sccm) was

introduced to the chamber after the sample was frozen. Introducing N2 should cause the air in

the chamber to become sub-saturated with water with respect to the saturation vapour pressure

above the ice sample, and result in an evaporating ice surface. However, we observe neither an

absolute decrease in Raman intensity (due to a shrinking ice volume) nor a change in the

relative Raman intensities (due to a change in surface composition) on the timescale of these

experiments. Therefore we conclude that the low nitrate signal we observed was not simply due

to the nitrate being buried by a growing ice surface. In several experiments the position of the

ice sample on the chamber floor (and hence the area probed by the laser beam) was adjusted to

test for the possibility of regions of higher concentration existing on the ice surface. Moving the

sample did not significantly change the shape or intensity of the spectra.

43

2.3. Results

Figure 2.3 shows glancing-angle Raman spectra acquired at the surface of a 100 mM

Mg(NO3)2(aq) sample prior to (dashed line) and after (solid line) freezing it from room

temperature to 268 K. The nitrate symmetric stretching band (-sym NO3¯), located at ~1000 cm-1

(22) and indicated by an arrow in the Figure, is barely distinguishable in the spectrum acquired

prior to freezing, as expected based on the dependence of the nitrate intensity on [Mg(NO3)2] at

the liquid surface (Figure 2.2). As illustrated in Figure 2.3, the frozen sample shows a much

more clearly defined -sym NO3¯ peak. The larger peak area at the frozen surface is indicative of

an enriched nitrate concentration there, suggesting that nitrate is excluded to the ice surface

during freezing.

Figure 2.4 shows that cooling the same sample to 258 K (solid trace) did not affect the

spectrum, nor did warming the sample back to 268 K as displayed by the dashed red trace.

Spectra acquired at the frozen surface also did not depend on the rate of freezing between 5 and

60 minutes. Within our experimental uncertainty, the area under the -sym NO3¯peak was found

to be independent of both temperature and the rate of cooling or warming for all frozen 100

mM Mg(NO3)2 samples. As illustrated by the dashed blue trace in Figure 2.4 it is also difficult to

distinguish the spectrum acquired at the surface of a frozen 75 mM Mg(NO3)2 sample from the

ones acquired at the surface of frozen 100 mM samples. Spectra acquired at the surface of 50

mM samples were also very similar. We made a few attempts to measure nitrate at the surface

of frozen KNO3(aq) (50 – 500 mM) and frozen HNO3(aq) (100 mM) samples. Preliminary results

showed little to no increase in the -sym NO3¯ feature upon freezing and so further experiments

were not undertaken.

44

Figure 2.3. Glancing-angle Raman spectra acquired at the surface of a 100 mM Mg(NO3)2(aq)

sample prior to freezing (red dashed line) and of the same sample after freezing to 268 K (black

solid line). Each plotted spectrum is an average of 4 individual spectra.

Figure 2.4. Glancing-angle Raman spectra acquired at the surface of frozen samples: 100 mM

Mg(NO3)2 cooled to 258 K (solid black line); 100 mM Mg(NO3)2 subsequently warmed to 268 K

(dashed red line); 75 mM Mg(NO3)2 cooled to 258 K (dashed blue line). Each plotted spectrum is

an average of 4 individual spectra. The area under the -sym NO3¯ peak (in arbitrary area units)

is the same within experimental error for all three spectra.

-sym NO3ˉ

45

2.4. Discussion

It is experimentally difficult to quantify solute concentrations in the liquid regions of ice;

typically they are calculated using measured pre-freezing or post-melting concentrations and an

equilibrium thermodynamic analysis. Cho et al. (23) present a formulation derived from ideal

solution thermodynamics for calculating the fraction of water in the LR , φH2O(T):

[1]

Here Cbulk is the bulk molal solute concentration, MH2O is the molecular weight of water, R is the

gas constant, Hf0 is the melting enthalpy of water, T is the temperature, and Tf is the freezing

temperature of water. In Equations [1], all impurities are assumed to be located in the LR, that

is, the fraction of solutes in the LR is set to 1. The molal concentration of solutes in the unfrozen

liquid region, CLR, is then given by:

CC

TLR

bulk

H O

2

( ) [2]

This formulation has been adopted by several groups in order to model the kinetics of nitrate

photolysis in frozen media (10, 24).

Alternatively, the concentration of Mg(NO3)2 in the unfrozen solution (CLR) may be

estimated from the Mg(NO3)2-H2O phase diagram (25). At 268 K this is ~0.8 mol L-1 (~12 wt%).

According to the calibration plot for -sym NO3¯ peak area versus [Mg(NO3)2] (Figure 2.2), a 0.8

mol L-1 Mg(NO3)2 solution should give rise to a peak area over an order of magnitude larger

than the one we observe. The fact that we only observe a small increase in peak area upon

freezing suggests that nitrate is not being excluded to the ice surface to the extent predicted by

the phase diagram. Under the same conditions, equations [1] and [2] predict a CLR of ~2.7 mol L-

1, which should give rise to an even larger nitrate peak area.

46

In both approaches outlined above, CLR’s are predicted to be a strong function of

temperature: at 258 K the phase diagram predicts a CLR of ~1.6 mol L-1 and equations [1] and [2]

predict a CLR of ~8.5 mol L-1. However, Figure 2.4 shows the -sym NO3¯ peak area to be

independent of temperature between 258 and 268 K. The lack of temperature dependence

further emphasizes the failure of the thermodynamic approaches to predict nitrate exclusion.

Our conclusion that the nitrate concentration at the ice surface is less than expected relies

on the relationship between the nitrate peak intensity and concentration being the same at the

ice and water surfaces. Note that the Raman signal we detect is proportional to the density of

scatterers (the CLR), the scattering cross section and the path length. Although it is possible that

the scattering cross section could be different in the QLL than in solution, we expect a true

liquid to exist at the air-ice interface under our conditions. In the limiting case for these

experiments (258 K, pre-freezing [Mg(NO3)2] = 50 mM) the phase diagram predicts φH2O(T) =

0.03. Assuming complete exclusion to the surface, and given a total ice volume of 4 cm3 and a

total ice surface area ~10 cm2, this φH2O(T) translates to a liquid layer at the surface with

thickness ~0.1 mm. Under the same conditions, equation [1] predicts φH2O(T) = 0.04 which

translates to a slightly thicker liquid layer. Therefore, if exclusion is primarily to the ice surface,

both approaches predict that the sample should be covered with a true brine layer whose

thickness is greater than the probe depth of our technique (~50 – 100 nm). Even if only 1% of the

total liquid fraction lies at the surface, we expect to be probing a liquid environment at the

frozen salt surfaces under our conditions (and hence the path length is the same under all

conditions). If this is the case, comparison of spectra acquired at the ice surface to spectra

acquired at the liquid surface should be valid. If this is not the case, then direct comparison of

ice spectra to liquid spectra may not be possible, but the same conclusion is arrived at: a

thermodynamic approach does not sufficiently predict nitrate exclusion to the surface.

Our results suggest two possibilities: either nitrate preferentially resides in liquid pockets

or grain boundaries within the ice sample rather than to the ice surface, or some of the nitrate is

47

effectively incorporated into the ice matrix. However, given the low solubility of nitrate in ice

(26) we favour the first possibility. These findings suggest that caution should be taken in a)

using an equilibrium thermodynamic approach to calculate nitrate enrichment factors for the

LR’s and/or b) assuming that the LR is located exclusively at the ice surface. A recent study by

Cheng et al. (27) proposes that factors other than thermodynamics (such as the instability

created during freezing by the rejected solute) may govern the location and nature of the

unfrozen regions in ice.

The behaviour we observe for nitrate differs from that observed for halides. An NMR

spectroscopic investigation by Cho et al. (23) showed that during the freezing of NaCl(aq)

solutions, there is a preferential migration of ions to liquid regions located at the top and

bottom of the NMR tube. A molecular dynamics simulation by Carignano et al. (28) also shows

that Na+ and Cl¯ ions are rejected from freezing ice into a liquid region located at both air-ice

interfaces. Furthermore, in the case of halides, there is evidence to suggest that enriched surface

concentration can be sufficiently predicted using an equilibrium thermodynamic analysis. An

XPS study by Křepelová et al. (29) shows that at temperatures above the eutectic, an unfrozen

NaCl brine phase is present at the outermost surface of the sample (probe depth ~1.1 nm).

According to their analysis, the composition of the unfrozen solution reflects the expected bulk

composition based on the NaCl-water phase diagram. We recently studied the heterogeneous

ozonation of bromide at the surface of frozen NaBr samples prepared in the same fashion as in

the present experiments (frozen from the bottom up) (14). The reaction kinetics we observed are

consistent with formation of a brine layer at the sample surface with a bromide concentration

predicted by the NaBr-H2O phase diagram (30).

Fewer studies have directly investigated the exclusion of nitrogen oxides during freezing.

Takenaka et al. (15) froze solutions of nitrous acid in the presence of oxygen from the bottom

up. The unfrozen solution was separated from the ice during freezing and the authors

monitored the concentration-time profiles of nitrite and nitrate (the nitrite oxidation product) in

each phase. They found that the concentration of nitrite in the unfrozen solution increased by

only around 20%. This result is consistent with the behaviour we observe for nitrate in this

48

experiment: there is exclusion to the surface, but not to a large extent. Křepelová et al. (18) used

NEXAFS to interrogate nitrate formed by NO2 hydrolysis at a 230 K ice surface. Although this is

a very different situation than nitrate exclusion from solution, the authors concluded that, at

low amounts, nitrate was not uniformly distributed on the surface, and was not in equilibrium

with gas phase nitric acid.

It is unclear why halides may be strongly excluded to the ice surface while nitrate and

nitrite are not. The different behaviour may be related to the aqueous surface affinity of the

ions. There is a wealth of experimental (31-33) and theoretical (34) evidence to suggest that the

heavy halide anions (Br¯ and I¯) exhibit a strong surface affinity. On the other hand, several

studies suggest that the NO3¯ anion exhibits a near-neutral surface affinity (35, 36) or is even

depleted at the air-aqueous interface (37, 38). However, a correlation of exclusion to the air-ice

interface with ion aqueous surface affinity is only speculative and further work is required to

understand the mechanism of solute exclusion during freezing.


Laboratory investigations have attempted to elucidate where nitrate photolysis occurs in

ice. Several of these studies suggest that nitrate photolysis occurs in a liquid-like region (3, 10).

However, the location of this region, whether it be at the air-ice interface or within the bulk ice

matrix has not yet been determined, although yields of gas-phase products (NO2) may suggest

that nitrate photolysis is occurring at the surface (3). Field and experimental observations also

suggest that ice/snow morphology and the location of NO3¯ is crucial to reactivity (4, 39).

In the summer months, polar temperatures can exceed 0 C, resulting in surface melting

(40). In addition, temperature gradients within the snowpack can lead to snowpack

metamorphism (41). When wet metamorphism occurs, liquid water is remobilized, the

morphology of the ice grains changes, and the distribution of solutes within the ice grains may

be affected. Thus, although nitrate may initially be present at the ice grain surface (e.g., from the

adsorption and dissolution of gas-phase HNO3), its location is likely to change with time.

49

The average snowpack nitrate concentration measured at Summit, Greenland (2000) was

4.4 μM and a typical summertime temperature was -20C (42). Jacobi and Hilker (10) used [3]

and [4] to calculate a LR concentration of 230 mM under these conditions, which represents an

enrichment of over four orders of magnitude. However, our results suggest that surface

concentrations of nitrate may be significantly lower than expected from application of [3], [4]

and equilibrium phase diagrams. Given the magnitude of the calculated enrichment, assuming

that this LR resides at the surface of ice crystals may result in gross misinterpretation of the

mechanism of snowpack nitrate photochemistry.

2.6. Conclusions

In conclusion, we have directly measured nitrate anion at the ice surface using glancing-

angle Raman spectroscopy. Our results suggest that nitrate is excluded to the ice surface during

freezing but that its concentration there is much lower than that predicted using an equilibrium

thermodynamic analysis; the concentration of solutes in the unfrozen solution existing at the ice

surface may be significantly different from the concentration of solutes found within liquid

pockets or grain boundaries. Care should be taken in assuming that all solutes are excluded into

a liquid layer existing at the ice surface. Predicting surface concentrations may be non-trivial,

especially given that nitrate and halides appear to show different affinities for exclusion to the

ice surface. Further work is needed to connect the mechanism of snowpack nitrate photolysis

with nitrate exclusion and where the reaction is taking place.

50

2.7. References


1510 (2002)







3. Y. Dubowski, A. J. Colussi, M. R. Hoffmann, Nitrogen dioxide release in the 302 nm band

photolysis of spray-frozen aqueous nitrate solutions. Atmospheric implications. J. Phys. Chem. A

105, 4928-4932 (2001)

4. Y. Dubowski, A. J. Colussi, C. Boxe, M. R. Hoffmann, Monotonic increase of nitrite yields in the

photolysis of nitrate in ice and water between 238 and 294 K. J. Phys. Chem. A 106, 6967-6971

(2002)

5. L. Chu, C. Anastasio, Quantum yields of hydroxyl radical and nitrogen dioxide from the

photolysis of nitrate on ice. J. Phys. Chem. A 107, 9594-9602 (2003)

6. E. S. N. Cotter, A. E. Jones, E. W. Wolff, S. J. B. Bauguitte, What controls photochemical NO and

NO2 production from Antarctic snow? Laboratory investigation assessing the wavelength and

temperature dependence. J. Geophys. Res. Atmos. 108, 4147 (8 pp), (2003)

7. C. S. Boxe, A. J. Colussi, M. R. Hoffmann, D. Tan, J. Mastromarino, A. T. Case, S. T. Sandholm, D.

D. Davis, Multiscale ice fluidity in NOx photodesorption from frozen nitrate solutions. J. Phys.

Chem. A 107, 11409-11413 (2003)

8. C. S. Boxe, A. J. Colussi, M. R. Hoffmann, J. G. Murphy, P. J. Wooldridge, T. H. Bertram, R. C.

Cohen, Photochemical production and release of gaseous NO2 from nitrate-doped water ice. J.

Phys. Chem. A 109, 8520-8525 (2005)

9. C. S. Boxe, A. J. Colussi, M. R. Hoffmann, I. M. Perez, J. G. Murphy, R. C. Cohen, Kinetics of NO

and NO2 evolution from illuminated frozen nitrate solutions. J. Phys. Chem. A 110, 3578-3583

(2006)

10. H. W. Jacobi, B. Hilker, A mechanism for the photochemical transformation of nitrate in snow. J.

Photochem. Photobiol. A. 185, 371-382 (2007)

11. R. Rothlisberger, M. A. Hutterli, S. Sommer, Factors controlling nitrate in ice cores: Evidence from

the Dome C deep ice core. J. Geophys. Res. Atmos. 105, 20565-20572 (2000)





(2008)


reaction between O3(g) and halides at the surface of frozen salt solutions. J. Geophys. Res. Atmos.

115, D16309 (8 pp) (2010)



Chem. 100, 13874-13884 (1996)



51

Interface. Environ. Sci. Technol. 44, 1302-1306 (2010);



18. A. Krepelova, J. T. Newberg, T. Huthwelker, H. Bluhm, M. Ammann, The nature of nitrate at the

ice surface studied by XPS and NEXAFS. Phys. Chem. Chem. Phys. 12, 8870-8880 (2010)

19. C. J. Pursell, M. A. Everest, M. E. Falgout, D. D. Sanchez, Ionization of nitric acid on ice. J. Phys.

Chem. A 106, 7764-7768 (2002)


Phys. Chem. A 111, 11006-11012 (2007)

21. M. Pavlovic, G. Baranovic, D. Lovrekovic, Raman study of the bending band of water.

Spectrochimica Acta 47A, 897-906 (1991)

22. M. R. Waterland, A. M. Kelley, Far-ultraviolet resonance Raman spectroscopy of nitrate ion in

solution. J. Chem. Phys. 113, 6760-6773 (2000)



24. C. S. Boxe, A. Saiz-Lopez, Multiphase modeling of nitrate photochemistry in the quasi-liquid

layer (QLL): implications for NOx release from the Arctic and coastal Antarctic snowpack. Atmos.

Chem. Phys. 8, 4855-4864 (2008)

25. K. Thomsen. Aqueous Solutions Magnesium Nitrate Phase Diagram.

http://www.phasediagram.dk/binary/magnesium_nitrate.htm (accessed May 27, 2011)

26. E. Thibert, F. Domine, Thermodynamics and kinetics of the Solid Solution of HNO3 in Ice. J. Phys.

Chem. B 102, 4432-4439 (1998)

27. J. Cheng, C. Soetjipto, M. R. Hoffmann, A. J. Colussi, Confocal Fluorescence Microscopy of the

Morphology and Composition of Interstitial Fluids in Freezing Electrolyte Solutions. J. Phys.

Chem. Lett. 1, 374-378 (2010)


99-103 (2007)



30. M. Tang, W. H. Tao, C. C. Wu, Y. P. Chen, Solid-liquid equilibrium and heat capacity

measurements of water plus sodium bromide and water plus sodium carbonate binary mixtures.

J. Chin. Inst. Chem. Eng. 34, 599-603 (2003).

31. S. Ghosal, J. C. Hemminger, H. Bluhm, B. S. Mun, E. L. D. Hebenstreit, G. Ketteler, D. F. Ogletree,

F. G. Requejo, M. Salmeron, Electron Spectroscopy of Aqueous Solution Interfaces Reveals

Surface Enhancement of Halides. Science 307, 563-566 (2005)

32. P. B. Petersen, J. C. Johnson, K. P. Knutsen, R. J. Saykally, Direct experimental validation of the

Jones-Ray effect. Chem. Phys. Lett. 397, 46-50 (2004)

33. D. Clifford, D. J. Donaldson, Direct Experimental Evidence for a Heterogeneous Reaction of

Ozone with Bromide a the Air-Aqueous Interface. J. Phys. Chem. A 111, 9809-9814 (2007)

34. P. Jungwirth, D. J. Tobias, Specific Ion Effects at the Air/Water Interface. Chem. Rev. 106, 1259 -

1281 (2006).

35. D. E. Otten, P. B. Petersen, R. J. Saykally, Observation of nitrate ions at the air/water interface by

UV-second harmonic generation. Chem. Phys. Lett. 449, 261-265 (2007).

36. L. M. Pegram, J. Record, M. Thomas, Partitioning of atmospherically relevant ions between bulk

water and water/vapor interface. Proc. Natl. Acad. Sci. U. S. A. 103, 14278-14281 (2006)

37. M. A. Brown, B. Winter, M. Faubel, J. C. Hemminger, Spatial Distribution of Nitrate and Nitrite

Anions at the Liquid/Vapor Interface of Aqueous Solutions. J. Am. Chem. Soc. 131, 8354-8355

(2009)

52

38. L. X. Dang, T.-M. Chang, M. Roeslova, B. C. Garrett, D. J. Tobias, On NO3¯-H2O interactions in

aqueous solutions and at interfaces. J. Chem. Phys. 124, 066101 (3 pp) (2006).

39. H. J. Beine, A. Amoroso, F. Domine, M. D. King, M. Nardino, A. Ianniello, J. L. France,

Surprisingly small HONO emissions from snow surfaces at Browning Pass, Antarctica. Atmos.

Chem. Phys. 6, 2569-2580 (2006)

40. H. J. Beine, F. Domine, A. Ianniello, M. Nardino, I. Allegrini, K. Teinila, R. Hillamo, Fluxes of

nitrates between snow surfaces and the atmosphere in the European high Arctic. Atmos. Chem.

Phys. 3, 335-346 (2003)

41. F. Domine, M. Albert, T. Huthwelker, H. W. Jacobi, A. A. Kokhanovsky, M. Lehning, G. Picard,

W. R. Simpson, Snow physics as relevant to snow photochemistry. Atmos. Chem. Phys. 8, 171-208

(2008)

42. R. E. Honrath, Y. Lu, M. C. Peterson, J. E. Dibb, M. A. Arsenault, N. J. Cullen, K. Steffen, Vertical

fluxes of NOx, HONO, and HNO3 above the snowpack at Summit, Greenland. Atmos. Environ. 36,

2629-2640 (2002)

53

C H AP T E R T H R E E

LABORATORY STUDY OF pH AT THE AIR-ICE INTERFACE


Adapted with permission from S. N. Wren and D. J. Donaldson, Laboratory study of pH at the

air-ice interface, Journal of Physical Chemistry C, 116, 10171 – 10180 (2012).

DOI: 10.1021/jp3021936. Copyright (2012) American Chemical Society.



54

3.1. Introduction

Chemical reactions occurring in or on frozen media are known to have a significant

influence on the composition of the overlying atmosphere (1). Some air-ice chemical

interactions, such as snowpack HONO production (2) or heterogeneous conversion of bromide

to bromine on sea ice (3), are thought to be pH-sensitive in keeping with the pH-sensitivity of

their liquid phase analogs. However, a good, mechanistic understanding of heterogeneous

chemistry on ice is still lacking, with efforts to gain a better understanding in this field

hindered, in part, by a poor understanding of pH at the ice surface. In this paper we seek to

address two questions regarding pH at the air-ice interface: When water containing trace

amounts of acid or base freezes, how is surface pH affected? And, when acids are introduced

from the gas phase, how is the pH at the ice surface affected?

Understanding the frozen water surface is complicated by the presence of what is

commonly referred to as the quasi-liquid layer (QLL) on ice. The QLL is the disordered region

present at the surface of (relatively) pure water ice at temperatures as low as ~240 K (4).

Spectroscopic studies have shown that properties of this layer (such as the extent of hydrogen

bonding) are intermediate to those of the liquid water surface and bulk ice, which is why it is

referred to as ‘quasi-liquid’ or ‘liquid-like’(5, 6). The exact thickness of this layer, as well as the

temperature dependence of its thickness, are still not well known (4). However, it is generally

accepted that QLL thickness increases with increasing temperature (7) and in the presence of

impurities (7-9). Since the QLL itself is not well-defined, constraining QLL concentrations is

difficult at best. Moreover, as impurity concentrations increase, frozen systems shift to a regime

in which a true liquid brine layer may exist in equilibrium with pure ice, as predicted by multi-

component phase behaviour (10).

The issue of surface pH is further complicated by the mechanism of freezing. During

freezing, impurities (solutes) which are excluded from the growing ice matrix may become

concentrated in the interfacial layer at the ice surface. Impurities are also segregated to

interfacial layers found within the bulk ice (grain boundaries or triple junctions, also called

55

nodes) or to liquid micropockets within the frozen matrix. Quantifying concentration

enrichments in these unfrozen regions is difficult and not always easily predicted from the bulk

composition (i.e., pre-freezing or melted composition)(11). However, it is well known that

freeze-concentration effects have important consequences for chemistry occurring in/on frozen

media (12-16). When the impurity is an acid or a base, it may also influence the surface pH.

In addition, as water freezes, certain ions may be selectively incorporated into the ice

crystal lattice. If cations and anions are differentially incorporated, a freezing potential develops

at the growing ice front (first discovered by Workman and Reynolds (17) in 1948 and discussed

in greater detail by Bronshteyn and Chernov (18)). The magnitude of the freezing potential,

which is also referred to as the Workman-Reynolds Freezing Potential (WRFP), depends on

several factors such as ion identity and concentration, as well as ice growth rate (18, 19). The

freezing potential is relaxed by the migration of H3O+ and OH¯ between the ice and the

unfrozen region. Several groups have attributed pH changes inferred during freezing to this

phenomenon (20-22).

Surface properties can also be influenced by the deposition of trace species from the gas

phase. Previous studies (references (23, 24) and reference therein) have investigated the uptake

of trace inorganic acids such as HCl(g) to the ice surface, with most studies performed at very

low temperatures (< 243 K) and low HCl partial pressures (< mTorr) (i.e., conditions relevant to

the upper troposphere/lower stratosphere). These experiments show that HCl is readily taken

up by the ice surface from the gas phase (even at very low temperatures), with a mass

accommodation coefficient ~ 0.2 (25). McNeill et al. (9, 26) have also shown that deposition of

gas phase HCl can induce surface disorder. However, these studies did not focus on the

resulting acidity at the ice surface.

Since ice in the environment is inhomogeneous on a microscopic scale, studying

concentration distributions is experimentally difficult. Only a few groups have attempted to

study the local acidity in the unfrozen solution of ice. Robinson et al. (20) used 3-fluorobenzoic

56

acid (FBH) as a pH probe and used the effective 19F NMR chemical shift to determine the degree

of FBH dissociation in the unfrozen portion of ice. Although this probe can differentiate

between the unfrozen and frozen environments, it is not surface selective. Heger et al. (21) used

UV-VIS spectroscopy and the diacid cresol red (CR) as a pH probe. The CR is assumed to be

segregated at the grain boundaries, and in equilibrium with protons due to their high

diffusivity in ice. This probe is not surface selective either. Other groups (22) have inferred pH

changes from measuring reaction kinetics.

In this study we use glancing-angle laser-induced fluorescence (LIF) with surface-

sensitive fluorescent probes to investigate pH at the air-ice interface. For the fluorescent probes

we use the pH-sensitive dyes acridine and harmine and adopt an approach similar to the one

described by Sayer et al. (27). Such an approach has been combined with LIF by our group to

study pH at the air-aqueous surface (28). After validating the technique for the frozen surface,

we use it to investigate the two main questions outlined above: When water containing trace

amounts of acid or base freezes, how is interfacial pH affected? And, when acids are introduced

from the gas phase, how is interfacial pH affected?


3.2.1. Experimental Overview

In this work we study surface pH using glancing-angle laser-induced fluorescence (LIF) in

conjunction with interface-sensitive pH probes. The glancing-angle approach relies on the fact

that at very high angles of incidence (with respect to the surface normal) the majority of the

incident laser light is reflected at the air-sample interface(6). As a result, the glancing-angle

technique preferentially interrogates the surface region. Moreover, both probes are surface-

active species with low-solubility in water that have been shown to partition to the air-aqueous

interface of liquid samples (28, 29). The fluorescent dyes acridine (dibenzopyridine, C13H9N )

and harmine (7-methoxy-1methyl-9H-pyrido[3,4-b]indole) were used as interface-sensitive pH

probes. Use of the probes relies on the fact that the neutral and protonated forms of each dye

exhibit different absorption and fluorescence maxima.

57

3.2.2. Apparatus

The reaction chamber consists of a Teflon box covering a copper base on which the sample

sits. Quartz windows (2” diameter) at the front and back of the box allow a laser beam to enter

and exit the chamber and third window on the side of the box facilitates alignment and

observation. Side ports allow gases to be introduced and ventilated. A liquid light guide is

suspended vertically from a hole at the top of the chamber. Samples were placed on a stainless

steel plate covering the chamber floor and the liquid light guide was fixed a few mm above the

sample surface.

3.2.3. Fluorescence and Excitation Spectra

Fluorescence at the sample surface was induced using the frequency-doubled output of a

tunable optical parametric oscillator (OPO) (pulse energy = 0.5 – 2.5 mJ, pulse frequency = 10

Hz, wavelength range = 245 – 355 nm). The OPO was pumped by a Q-switched, frequency-

tripled Nd:YAG laser (355 nm). The vertically polarized laser beam passed through a

collimating system consisting of three lenses and a pinhole aperture, and was then directed into

the chamber using a series of quartz prisms. The system was aligned such that the laser beam

impinged the sample surface directly below the liquid light guide at a glancing-angle (> 85

from the surface normal). The laser spot size on the surface was ~ 7 mm diameter. Laser-

induced fluorescence was collected by the liquid light guide and imaged onto the entrance slit

of a ¼ m grating monochromator (slit widths varied from 1 – 3 mm depending on the

experiment). The wavelength-separated light passed through a 355 nm laser line long pass filter

(in order to reduce interference from Rayleigh scattering) and was detected by a

photomultiplier tube. The intensity vs. time signal was averaged over 64 laser shots by a digital

oscilloscope and a LabVIEW program was used to sample a portion of the averaged intensity

vs. time signal. The effective time resolution of the experiments was ~ 5 ns.

Acridine fluorescence spectra were acquired by exciting acridine at 337 nm and scanning

the monochromator from 390 nm – 600 nm in steps of ~6 nm. For spectra acquired at the liquid

surface, a 50 ns time slice from the intensity vs. time trace was used. For spectra acquired at the

58

frozen surface, a 25 ns time slice from the intensity vs. time trace was used. Fluorescence spectra

were baseline corrected by subtracting the minimum intensity data point collected at a

wavelength > 550 nm.

Harmine excitation spectra were acquired by monitoring fluorescence at ~430 nm while

scanning the excitation wavelength from 265 – 355 nm in steps of 5 nm. Since the fluorescence

decay rate was very fast and did not show a strong dependence on physical state (frozen vs.

liquid) or excitation wavelength, the same time slice of 10 ns was used for experiments on liquid

and frozen surfaces. Excitation spectra were not baseline corrected since a suitable wavelength

region was not available.

For bulk experiments, a few mL of sample was placed in a 1 cm path-length quartz

cuvette. Fluorescence was collected with the liquid light guide oriented perpendicular to the

incident laser beam (in the horizontal plane)(6).


A stock saturated solution of acridine in distilled water (~1.5 × 10-4 M) was prepared and

allowed to stir for 24 hours. Experimental solutions were prepared daily by diluting the stock

solution in distilled water to a final acridine concentration of 7.5 × 10-7 M. Experiments involving

harmine were prepared in a similar manner. A stock saturated solution (~2.5 × 10-5 M) of

harmine in deionized water was prepared. Experimental solutions were diluted to a final

harmine concentration of 1.0 × 10-7 M.

Initial solution pH was measured using a pH electrode after the solution had reached

equilibrium with atmospheric CO2 (giving a pH ~5.9). Initial solution pH was adjusted by

adding one of HCl(aq), HNO3(aq) or NaOH(aq) dropwise until the solution reached the desired

pH; a buffer was not used. Solutions were not degassed.

59

Prior to sample introduction, the chamber was purged with dry N2. Approximately 4 mL

of solution was then spread onto a stainless steel plate sitting on the chamber floor. During this

time, the chamber air was briefly in contact with the laboratory air. Liquid samples were

prepared at room temperature. Unless stated otherwise, frozen samples were prepared by

bottom-up freezing with the chamber held at 263 K (time to freezing < 5 min). In a few cases

bottom-up freezing was induced by slowly lowering the chamber temperature (time to freezing

> 30 min). Typically the top of the sample was the last region to freeze. Both liquid and frozen

samples were flat on top with rounded edges (thickness ~0.5 cm, total surface area ~ 6 cm2). A

flow of 0.5 SLPM dry N2 through the reaction chamber was maintained throughout the

experiment. This was done in order to a) minimize the influence of atmospheric CO2 and b)

prevent new-ice growth at the air-ice interface due to condensation of water. As in Chapter

Two, we did not observe evidence for sublimation of ice.


Bulk pH was measured before each experiment using a commercial pH electrode (Orion

Model 520 A). The pH electrode was calibrated daily using a three-point calibration (pH 4, 7,

10). In some cases the final bulk (melted) pH was measured after the experiment. For the sake of

clarity, we will use the term ‘initial pH’ to refer to pre-freezing bulk pH for the frozen case or

bulk pH for the liquid case. We will often use the terms ‘final bulk pH’ or ‘final melted pH’ to

refer to the pH of the melted sample.

3.2.6. Deposition of HCl(g)

Both the acridine 430/470 ratio and the harmine 290/320 ratio were used to infer pH.

Samples were frozen under a steady 0.5 SLPM flow of dry N2 as described above. HCl(g) was

then delivered to the samples by switching the flow to a 0.5 SLPM flow of 100 ppm HCl(g) in

balance N2. The same flow conditions were used for liquid and frozen experiments. In most

cases, the samples were exposed to the HCl(g) flow a fixed period in order to better compare

final pH measurements. From the absolute change in bulk pH we estimate that ~5% of the HCl

was taken up by the 4 mL sample.

60

For the acridine experiments, samples were initially in equilibrium with air and thus had

an initial pH ~ 5.9. A stable 430/470 baseline was first established prior to the introduction of

HCl(g), the 430/470 ratio was then monitored during HCl(g) deposition. For the harmine

experiments, initial sample pH was adjusted with NaOH (to initial pH ~ 9.5). The 290/320 ratio

was monitored during freezing until a stable 290/320 baseline was established. The 290/320 ratio

was then monitored during HCl(g) deposition

3.2.7. Chemicals

Acridine (Fluka, min 97.0%), Harmine (Aldrich, 98%), Sodium Chloride (ACP Chemicals,

min 99.0%), Sodium Hydroxide (ACP Chemicals, min 97.0%), Ammonium Hydroxide (Fisher

Scientific, 28.0 – 30.0%), Hydrochloric Acid (Fisher Scientific, 36.5 – 38.0%), Nitric Acid (Fisher

Scientific, 68.0 – 70.0%), Compressed Nitrogen (Linde, Grade 4.8), 100 ppm HCl(g) balance

N2(g) (Linde, Lot # KAL-HCL-100-2). All chemicals were used without further purification.

3.3. Results

3.3.1. Development of pH Probes for the Liquid Surface

3.3.1.1. Acridine Photophysics and Acid-Base Behaviour

The photophysics of acridine (Ac, dibenzopyridine, C13H9N) have been described in detail

elsewhere (27). Here we use it as an interface-sensitive pH indicator (27, 28). In aqueous

solution, Ac absorbs in the near-UV (337 nm in this study) and is excited to the S1(21A1) state.

Excited acridine (Ac*) emits a broad fluorescence spectrum with a peak maximum at ca. 430 nm.

In the ground state, the nitrogen on neutral acridine (Ac) protonates to form the acridinium

cation (AcH+) with a pKa ~ 5.4 at 298 K(27).

Ac + H+ AcH+

Therefore, in aqueous solutions with pH < 5.4, AcH+ will be the dominant species.

Protonated acridine also absorbs strongly in the near-UV (337 nm); AcH+* emits a broad

61

fluorescence spectrum with a peak maximum around 470 – 480 nm. The first excited state (Ac*)

is more basic than the ground state (Ac), with a pKa of 9.2 (27). Thus, excitation of aqueous

solutions with pH 5.4 – 9.2 will result in the formation of both Ac* and AcH+* and the proportion

of each species will be reflected by the spectral profile of the fluorescence band. In this work, the

relative contributions of Ac* and AcH+* to acridine’s fluorescence spectrum were used to infer

acridine’s local pH.

3.3.1.2. Acridine 430/470 Ratio

Laser-induced fluorescence spectra acquired at the liquid surface following excitation at

337 nm are shown for basic and acidic bulk pH values in Figure 3.1. At a bulk pH of 8.8 (shown

in Figure 3.1), the dominant species in solution is neutral acridine (Ac) and hence emission at

430 nm due to Ac* dominates the fluorescence spectrum. As the bulk pH is lowered, Ac is

protonated to form AcH+. As a result, at a bulk pH of 2.3, the fluorescence spectrum exhibits a

smaller contribution from Ac* at 430 nm and a larger contribution from AcH+* at 470 nm.

Previous studies (27, 28) have shown that the relative intensities of the features appearing

at 430 nm and 470 nm in acridine’s fluorescence spectrum can be used to infer acridine’s local

pH (we will refer to the intensity ratio as the 430/470 ratio hereafter). We used this approach to

obtain a calibration curve for pH at the liquid water surface. For the calibration curve samples,

the bulk pH of the acridine solution was adjusted with HCl(aq), HNO3(aq) or NaOH(aq) as

required. The 430/470 ratios obtained from liquid surface spectra have been plotted as a

function of their corresponding bulk pH values as open symbols in Figure 3.2. The 430/470

calibration curve obtained in this study is in good agreement with those obtained previously

(27, 28). The data is well-fit by a 3-parameter sigmoidal curve (R2 = 0.92). The effective

equivalence point obtained from our calibration (~5.6) is slightly higher than the reported bulk

pKa (5.4) (27). Clifford et al. (28) also observed a higher effective equivalence point (~6.5) and

suggested that the shift is due to the fact that protonated acridine partitions less well to the

surface.

62

Figure 3.1. Acridine fluorescence spectra acquired at the liquid water surface with initial pH of

2.3 (dashed black line) and 8.8 (dashed red line) and at the frozen water surface with initial pre-

freezing pH of 3.4 (solid black line) and 9.4 (solid red line).

Figure 3.2. Acridine 430/470 intensity ratios as a function of initial pH. Data from spectra

acquired at the liquid surface are shown as open circles. A the frozen surface (solid symbols)

samples had pH adjusted with HNO3 (red triangles), HCl (black circles), and NaOH (green

squares). The liquid data has been fit to a 3-parameter sigmoidal function (solid black line) of the

form f= a/(1+exp(-(x-x0)/b)) with a = 2.2 ± 0.1, b = 0.7 ± 0.1 and x0 = 5.6 ± 0.2 and R2 = 0.92.

63

3.3.1.3. Acridine Fluorescence Decay Rate

Fluorescence decay rates were obtained by fitting single-exponential functions to the

acridine fluorescence decay profiles at 470 nm and 430 nm. Recall that the emissions at 470 nm

and 430 nm are due to the protonated (AcH+*) and neutral (Ac*) forms respectively. Typical fits

to decay profiles at 470 nm for acidic and basic conditions are shown for the liquid surface in

Figures 3.3a and 3.3b respectively, and for acidic and basic conditions at the frozen surface in

Figures 3.3c and 3.3d respectively. The inset in each figure shows the linearized data with

intensity on a logarithmic scale. The inset shows that the decays are well-fit by a single

exponential over ~ 3 lifetimes. The "blip" seen in the decays near 20 ns (and also ~ 150 ns in the

liquid cases) is an artifact of the detection electronics. Deviations from the fit at longer times

may be due to an offset in the ‘zero’ signal. Overall the fits were good (R2 > 0.9), although the

fits to the 430 nm decays at low pH values were significantly poorer (occasionally as low as R2

~0.7).

The fluorescence decay rates at 430 nm and 470 nm at the liquid surface are plotted as a

function of bulk pH in Figure 3.4. On the liquid surface, the fluorescence decay rates at 430 nm

and 470 nm exhibit a very similar dependence on bulk pH, with an overall trend of decreasing

decay rate with decreasing pH over the range pH 3 - 10. Below pH ~3, the decay rate increases

rapidly with acidity for samples acidified with HCl but not for samples acidified with HNO3.

The increase in the fluorescence decay rate at these low pH’s (and associated high HCl

concentrations) may be attributed to quenching by chloride (30). The decay rates at 470 nm have

been plotted as a function of bulk [Cl¯] (assuming complete ionization of HCl) in Figure 3.5.

Also plotted on Figure 3.5 are fluorescence decay rates measured at the NaCl(aq) surface. The

datasets show the same linear dependence on the bulk chloride concentration, which further

suggests that chloride is the species responsible for the observed quenching.

64

Figure 3.3. Single-exponential fits to the acridine fluorescence decay profile at 470 nm: (a)

Liquid water surface pH 4.6 (R2 = 0.99) (b) liquid water surface pH 10.9 (R2 = 0.96); (c) frozen

water surface pH 4.6 (R2 = 0.96); (d) frozen water surface pH 10.1 (R2 = 0.94). Raw data is

plotted as open circles (red for samples adjusted with HNO3 and green for samples adjusted

with NaOH) and the fit is plotted as the solid black line.

(a) (b)

(c) (d)

65

Figure 3.4. Acridine fluorescence decay rates at 430 nm (open symbols) and 470 nm (solid

symbols) measured at the liquid water surface plotted as a function of initial pH for samples

adjusted with HNO3 (red triangles), HCl (black circles), and NaOH (yellow diamonds);

unadjusted samples (green squares).

Figure 3.5. Acridine fluorescence decay rates following excitation at 337 nm plotted as a function

of bulk [Cl¯] (assuming complete ionization). Measured at the liquid water surface at 470 nm for

samples acidified with HCl (this study, black circles) and at the liquid salt water (NaCl) surface at

495 nm (red triangles). The solid line is a linear fit to the HCl data. The NaCl data was obtained

by Dr. T. F. Kahan.

66

3.3.1.4. Harmine Photophysics and Acid-Base Behaviour

The photophysics and acid-base behaviour of harmine (7-methoxy-1methyl-9H-

pyrido[3,4-b]indole), a member of the β-carbolines group, has been described in detail

elsewhere (31-34). The acid-base behaviour of harmine is more complicated than that of

acridine, owing to the fact that neutral, cationic and zwitterionic forms can form in solution.

However, harmine’s fluorescence is not readily quenched in higher ionic strength environments

and so it is ideal for salt water/seawater studies. In aqueous solution, the more basic pyridinic

nitrogen on neutral harmine (N) will accept a proton to form a cation (C). The pKa of the

pyridinic nitrogen at 295 K has been reported as 7.73 ± 0.02 (32). Thus in the ground state, in

aqueous solutions with pH < 7, the cation is the dominant species in solution. In the pH range 7

– 9.5, both the neutral and cationic forms may be present, and above pH 9.5 the neutral form

dominates. Both the neutral and cationic forms absorb in the UV. The excited cation (C*)

fluoresces strongly with a peak maximum centred ~ 400 nm while neutral harmine (N*)

fluoresces much more weakly, with a peak maximum ~350 nm. In the excited state, the pyrrolic

nitrogen can be deprotonated to yield a zwitterion (Z*) which emits at ~500 nm. Below pH ~ 9,

Z* is formed by a double proton transfer mechanism from N* and above pH ~ 9, Z* can also be

formed from C*; Z* is not expected to form directly from the ground state. Therefore, as in the

case with acridine, the spectral profile of the fluorescence band reflects the relative abundance

of each species in solution. However, due to the low intensity of the neutral emission vs. cation

emission, changes in the fluorescence spectra due to changes in solution pH are too small to be

useful in this experiment. Furthermore, the harmine fluorescence spectrum, following excitation

at 337 nm, overlaps with the Raman OH-stretching band from water (centred around ~3450 cm-1

which appears at ~380 nm). However, the spectral profile of harmine’s absorption spectrum is

also pH dependent (31). Below pH ~ 7, the absorption is due to the cation form, with an

absorption maximum ~ 320 nm. Above pH ~ 9.5, the absorption is due to the neutral form, with

the absorption maximum blue-shifted to ~ 300 nm. In this work, the relative contribution of the

neutral and cation forms to harmine’s absorption spectrum were used to infer harmine’s local

pH.

67

3.3.1.5. Harmine 290/320 Ratio

Excitation spectra of neutral and protonated harmine at the liquid water surface are

shown in Figure 3.6 (emission monitored ~430 nm). The spectra have been normalized to the

cation absorption maximum at 320 nm. The liquid surface spectra are in good agreement with

the bulk spectra obtained by this group and others (31). The maximum absorption shifts from ~

300 nm for the neutral form which is present at basic pH values, to ~ 320 nm for the cationic

form which is present at neutral and acidic pH values. Since fluorescence is monitored at 430

nm (the emission maximum for the cationic form), fluorescence from basic samples is generally

weaker, giving rise to noisier spectra.

The shift in harmine’s absorption spectrum with changing pH has previously been

exploited (29, 35): Clifford and Donaldson (29) monitored the decrease in harmine fluorescence

intensity at 410 nm which is commensurate with the decrease in harmine absorption at 337 nm

accompanying increasing pH. Here we adopt a similar approach to that described above for

acridine and use a ratio of intensities from harmine’s excitation spectrum. Specifically, we use

the relative intensities of the 290 nm and 320 nm absorptions to infer harmine’s local pH

(hereafter the 290/320 ratio). Since this work represents the first time this approach has been

used for harmine, we first obtained calibration curves for both bulk and surface liquid samples.

Bulk sample pH was adjusted with NaOH(aq) or HCl(aq). Unadjusted samples had an initial

pH ~ 5.9 (water in equilibrium with air). The 290/320 ratios are plotted as a function of bulk pH

for bulk and surface liquid samples in Figure 3.7. The bulk liquid and liquid surface data are in

very good agreement. This is consistent with previous work involving acridine (28). As in the

case with acridine, the apparent inflection point ~ 8.1 is slightly higher than the reported pKa

(7.7)(32).

68

Figure 3.6. Harmine excitation spectra acquired at the liquid water surface with initial pH ~ 6

(dashed red line) and pH ~ 10 (dashed black line); and at the frozen water surface with initial pH

~ 6 (solid red line) and pH ~ 10 (solid black line). Harmine fluorescence monitored at 430 nm.

Figure 3.7. Harmine 290/320 intensity ratios as a function of initial pH. Data from spectra

acquired in bulk liquid water (crosses), at the liquid water surface (open triangles), and at the

frozen water surface (red triangles).

69

Figure 3.8. Harmine fluorescence decay rates at 430 nm following excitation at 320 nm. Measured

at the liquid water surface (black circles) and at the liquid salt water (0.5 M NaCl) surface (red

triangles).

Figure 3.9. Acridine fluorescence decay rates at 470 nm measured at the frozen surface plotted as

a function of initial pre-freezing pH for samples adjusted with HNO3 (red triangles), HCl (black

circles) and NaOH (yellow diamonds); unadjusted samples (green squares). Note the very

different behaviour with HCl compared to HNO3. The purple dotted line is a guide to the eye.

70

3.3.1.6. Harmine Fluorescence Decay Rate

Fluorescence decay rates in water were obtained by fitting single-exponential functions to

the harmine fluorescence decay profiles at 430 nm following excitation at 320 nm. The decay

rates are plotted as a function of bulk pH in Figure 3.8. Note that the absolute fluorescence

decay rates for harmine are significantly faster than those for acridine. Harmine’s fluorescence

decay rate also shows a slight pH dependence, although the strength of this dependence is

considerably weaker than that obtained for acridine (by a factor of 2, approximately).

Fluorescence decay rates for harmine in 0.5 M NaCl(aq) are also shown in Figure 3.8 as a

function of bulk pH. From Figure 3.8 it is clear that harmine’s fluorescence lifetime is slightly

quenched by the presence of chloride. The presence of chloride overrides any pH dependence

and thus the fluorescence decay rates are no longer a function of bulk pH (but would seem to be

a function of the chloride concentration). However, it should be noted that harmine’s

fluorescence is less readily quenched than is acridine’s (by comparison with Figure 3.3 and

extrapolation to 0.5 M Cl¯). Since harmine’s fluorescence is not extinguished in higher ionic

strength environments, it is suitable for studies on salt water systems.


3.3.2.1. Fluorescent pH Probes for the Frozen Surface

Acridine fluorescence spectra acquired at the frozen water surface following excitation at

337 nm are shown in Figure 3.1 for basic and acidic pre-freezing pH values. The fluorescence

signal at the frozen surface was observed to be much weaker than the fluorescence signal at the

liquid surface. At a more basic pre-freezing pH, the fluorescence spectrum shows significant

relative intensity at 430 nm due to emission from neutral acridine (Ac*). At a more acidic pre-

freezing pH, the fluorescence spectrum exhibits a relative decrease in intensity at 430 nm due to

Ac* and a relative increase in intensity at 470 nm due to protonated acridine (AcH+*)(27). Thus,

the ratio of intensities at 430 nm and 470 nm (the 430/470 ratio) reflects the relative proportion of

Ac and AcH+ at the frozen surface, which is in turn related to the local pH there. In the following

we use the acridine 430/470 ratio, as displayed in Figure 3.2, to infer local pH, as we (28) and

others (27) have done at the liquid surface.

71

Acridine fluorescence decay rates at 470 nm obtained at the frozen surface are plotted as a

function of initial pH in Figure 3.9. The acridine fluorescence decay rates show a dependence on

initial pH and on chloride concentration (on both the liquid water surface and frozen surface)

and thus can provide additional information about acridine’s local surface environment.

Harmine excitation spectra acquired at the frozen water surface are shown in Figure 3.6

for basic and neutral pre-freezing pH values. Again, the fluorescence signal at the frozen surface

was observed to be much weaker than the fluorescence signal at the liquid surface. The frozen

spectra exhibit similar features to the liquid spectra: the maximum absorption shifts from ~ 290

nm for the neutral form which is present at basic pH values, to ~ 320 nm for the cationic form

which is present at pH < 8 (31). The ratio of intensities at 290 nm and 320 nm (the 290/320 ratio)

reflects the relative proportion of the neutral and cationic forms, which are related to harmine’s

local pH. In the following we use the harmine 290/320 ratio to infer local pH as illustrated in

Figure 3.7. In some experiments we followed the harmine 300/320 ratio, changes to which also

reflect changes in pH.

Thus we have established the three measures with which we may infer interfacial pH: the

acridine 430/470 ratio, the harmine 290/320 ratio and the acridine fluorescence decay rate at 470

nm. To study pH changes at the frozen surface using our pH probes we adopted the following

approach:

a) start with water (i.e., deionized water) containing the pH probe (acridine or harmine), as

described above

b) adjust the pre-freezing pH with HCl(aq), HNO3(aq) or NaOH(aq) as required

c) freeze 4 mL of sample at 263 K in the cooled reaction chamber

d) acquire laser-induced fluorescence spectrum (acridine) or excitation spectrum (harmine)

at the frozen surface

e) obtain the 430/470 ratio (acridine) or 290/320 ratio (harmine) from the spectra and infer

surface pH by comparing to the liquid surface result

72

f) for acridine experiments, obtain the acridine fluorescence decay rate at 470 nm and

compare to the liquid surface result

3.3.2.2. Freezing Acidic and Basic Solutions

When water freezes, how is interfacial pH affected? In this first section, we report on

results obtained when water containing acid or base (prepared as described in Section 3.2.4) is

frozen.

The 430/470 ratios obtained from acridine fluorescence spectra acquired at the ice surface

are plotted as a function of initial pH in Figure 3.2 as solid symbols. The 290/320 ratios obtained

from harmine excitation spectra acquired at the ice surface are plotted as a function of initial pH

in Figure 3.7 as solid red triangles. For comparison, the ratios obtained at the liquid surface are

shown as hollow symbols in both figures.

We used acridine to infer pH at the surface of initially acidic samples. As illustrated by

Figure 3.2, the 430/470 ratio obtained at the frozen surface is similar to that measured at the

liquid surface for the same initial pH. In other words, the 430/470 ratio measured at the frozen

surface is suggestive of a surface pH that is unchanged with respect to the initial pH. This result

is found to be independent of the acid identity: HNO3 or HCl.

Additional insight can be gained upon examination of the acridine fluorescence decay

rates. As illustrated by Figure 3.9, at the frozen surface (excluding samples acidified with HCl),

the acridine fluorescence decay rates at 470 nm show an overall trend of decreasing decay rate

with decreasing pH. This trend is similar to that observed for the liquid surface (Figure 3.4).

Within our ability to extract fluorescence lifetimes, single-exponential decays were obtained for

the frozen samples at all pre-freezing pH values. In other words, the relationship between

fluorescence decay rate and initial pH is largely preserved during freezing. This observation is

consistent with there being little change in surface pH during freezing and supports the results

presented above. Note however that the fluorescence quenching is much stronger at the frozen

73

than the liquid surface as evidenced by the decreased fluorescence intensity measured there.

Consistent with this, the decay rates measured at the frozen surface are generally about a factor

of 2 faster than those measured at the liquid surface.

We used both acridine and harmine to infer pH at the surface of initially basic samples

(note that the pH probes are sensitive over different pH ranges). In the former case (Figure 3.2),

the acridine 430/470 ratios obtained at the frozen surface show considerable scatter, but tend to

be similar to or slightly lower than the corresponding liquid surface ratios. In the latter case

(Figure 3.7), the harmine 290/320 ratios are consistently lower on the frozen surface than on the

liquid surface for the same initial pH. This decrease could be suggestive of a relative decrease in

surface pH due to freezing for these initially basic samples.

To further explore the apparent decrease in surface pH upon freezing initially basic

solutions, the harmine 300/320 ratio was followed during the freezing of samples with initial pH

> 7. In these experiments a stable ratio was first obtained at the liquid surface. The ratio was

then monitored as a function of time as the temperature in the chamber was lowered. As can be

seen in Figure 3.10a, the ratio remained stable prior to freezing, then decreased rapidly during

the freezing stage (t = 0), and was stable again post-freezing. As noted above, freezing was also

accompanied by a sudden decrease (by a factor of ~ 2 or more) in the absolute fluorescence

intensity as illustrated by Figure 3.10b. This resulted in a decrease in the signal to noise in the

spectra acquired at the frozen surface.

74

Figure 3.10.(a) The harmine 300/320 ratio acquired at the sample surface as a function of time

during the slow freezing of a water sample (initial pH adjusted with NaOH to pH = 8.5). (b)

Harmine’s absolute fluorescence intensity at 430 nm (following excitation at 320 nm) during the

freezing of the same sample. The sample froze at time, t = 0 (indicated by the dashed grey line).

It is possible that some of the apparent pH lowering is due to the dissolution of

atmospheric CO2. To test for this, we monitored the 290/320 ratio with time in the presence and

absence of a N2 flow. In the absence of a N2 flow through the chamber, liquid samples with pH >

7 showed a decrease in their 290/320 ratio with time, indicating a decrease in surface pH. The

rate of decrease was accelerated when the chamber windows were removed and the sample

was completely exposed to the ambient air. Such a decrease was considerably slowed or not

observed in the presence of N2. In the absence of an N2 flow, frozen samples also showed a

decrease in 290/320 ratio with time; this decrease was not observed in the presence of N2.

(a)

(b)

75

However, in some cases (especially the harmine experiments), final melted pH values (as

measured with the pH electrode) indicated a ‘true’ lowering in the bulk pH, which we attribute

to the uptake of atmospheric CO2.

3.3.2.3. Exclusion of Chloride

The fluorescence decay rates obtained at the liquid and frozen surface of water samples

acidified with HCl have been plotted together in Figure 3.11 as a function of initial pH.

Fluorescence decay rates obtained at the frozen surface increase monotonically with decreasing

pH from pH ~6. At the liquid surface, an increase in fluorescence decay rate is also observed,

but only once the initial pH < 3. In Figure 3.5, the liquid surface decay rates for have been re-

plotted as a function of [Cl¯] (assuming complete ionization of HCl). That figure shows a linear

dependence of fluorescence decay rate on [Cl¯], which is indicative of quenching by chloride.

We observe a similar increase in the decay rate when HCl is replaced by NaCl (Figure 3.5),

which provides additional evidence that chloride is the species responsible for the observed

fluorescence quenching. Other groups have also observed quenching of acridine fluorescence

due to the presence of chloride (30). A similar quenching effect is not observed when HCl is

replaced with HNO3 (on the liquid or frozen surface, see Figure 3.4 and 3.9). As in the liquid

case, the increase in fluorescence decay rate with increasing HCl (i.e., decreasing pH) observed

at the frozen surface is likely related to quenching by chloride. From Figure 3.11 it is clear that

quenching occurs at a higher initial pH (i.e., lower initial [Cl¯]) for frozen samples than liquid

samples. This observation is highly suggestive of exclusion of chloride ions to the ice surface,

resulting in their enriched concentrations there.

76

Figure 3.11. Acridine fluorescence decay rates at 470 nm acquired at the liquid water surface

(open circles) and frozen water surface (black circles) following excitation at 337 nm, plotted as a

function of initial pH for samples acidified with HCl. Decay rates were obtained by fitting single

exponential decays to the fluorescence vs. time signal.

3.3.2.4. Exclusion and Self-Association of Probe Molecules

Low probe concentrations (7.5 × 10-7 M acridine and 1.0 × 10-7 M harmine) were used in

these experiments. As the harmine concentration is increased to 2.5 × 10-6 M, the fluorescence

maximum shifts from ~420 nm (the location of the cation emission) to ~ 445 nm in spectra

acquired at the frozen surface (see Figure 3.12). If the initial acridine concentration is increased

to 1.5 × 10-6 M, a significant feature at ~530 nm appears in fluorescence spectra acquired at the

frozen surface (see Figure 3.13). These red-shifted features are not observed in spectra acquired

at the surface of liquid samples under the same conditions and can likely be attributed to self-

associated acridine at the frozen surface (28). Red-shifted features have previously been

observed in glancing-angle LIF spectra of polycyclic aromatic hydrocarbons (PAHs) at the ice

surface and have been attributed to PAH self-association (36, 37). These results indicate that the

probe molecules are excluded during freezing, at least in part, to the ice surface (and possibly to

grain boundaries or nodes as well).

77

Figure 3.12. Harmine fluorescence spectra acquired at the frozen water surface for pre-freezing

harmine concentrations of 2.5 × 10-6 M (medium dash red) and 1.0 × 10-7 M (short dash green).

Harmine fluorescence spectrum acquired at the liquid surface with a bulk harmine concentration

of 2.5 × 10-6 M (solid black). All samples in equilibrium with air (initial pH ~ 5.9). The portion of

the spectrum that interferes with the Raman OH-stretching mode (~ 380 nm) has been removed.

Figure 3.13. Acridine fluorescence spectra acquired at the frozen water surface for pre-freezing

acridine concentrations of 7.5 × 10-7 M (solid black), 1.5 × 10-6 M (medium dash red) and 5.0 × 10-6

M (short dash green). All samples in equilibrium with air (initial pH ~ 5.9).

78

3.3.3. Interfacial pH Changes due to Gas Phase Deposition of HCl

When HCl is deposited from the gas phase, does the interfacial pH change at the frozen

water surface differ from that at the liquid water surface? We investigated changes in surface

pH due to the uptake of HCl(g) by monitoring the acridine 430/470 ratio and the acridine

fluorescence decay rate at 470 nm. The initial sample pH was unadjusted (i.e., samples had a

near-neutral initial pH); this pH was chosen so as to ensure an observable change in the acridine

430/470 ratio upon acidification. The 430/470 ratio is plotted as a function of time relative to the

introduction of HCl(g) in Figure 3.14a and the acridine fluorescence decay rate at 470 nm is

plotted as a function of time relative to the introduction of HCl(g) in Figure 3.14b.

As illustrated by the black circles in Figure 3.14a, when HCl(g) is introduced to the

chamber containing a liquid water sample with near-neutral initial pH, the 430/470 ratio

decreases steadily with time and then levels off at a final 430/470 ratio ~ 0.1. The final bulk pH

as measured by the pH electrode is acidic (pH ~ 3). The decrease in 430/470 ratio can be related

to a decrease in pH and the final 430/470 ratio of 0.1 is consistent with a pH < 4 (according to the

relationship between 430/470 ratio and pH given by Figure 3.2). The fluorescence decay rates

measured at the liquid surface show a similar response, as illustrated by the black circles in

Figure 3.14b. The observed decrease in the fluorescence decay rate is consistent with decreasing

pH (according to the relationship between decay rate and pH given by Figure 3.14c).

However, as shown by the red symbols in Figure 3.14a, when HCl(g) is introduced to

frozen water samples with near-neutral initial pH, the 430/470 ratio decreases very slowly or

not at all, with a final 430/470 ratio ~ 0.8 – 0.9. The final melted pH as measured by the pH

electrode is again acidic (pH 3 – 3.5 for all experiments). According to the relationship between

430/470 ratio and initial pH on frozen surfaces given by Figure 3.2, a sample with surface pH ~ 3

should give also rise to a 430/470 ratio < 0.4. Indeed, when the melted sample is re-frozen, the

resulting acridine fluorescence spectrum is characteristically ‘acidic’, giving rise to a 430/470

ratio < 0.4. This is consistent with the results shown in Figure 3.2, in which the liquid and ice

surface pH is the same. The same result is obtained when harmine is used as the pH probe:

79

when a frozen water sample whose initial pH has been adjusted with NaOH to be basic (pH > 9)

is exposed to HCl(g), the 290/320 ratio decreases very slowly with time. Thus, although the

same amount of HCl is delivered to the liquid and frozen samples (in a bulk sense), the resulting

pH at the surface is quite different for the two cases.

The acridine fluorescence decay rates also provide information about the surface

environment. The fluorescence decay rates measured at the liquid surface during HCl(g)

deposition are plotted as a function of their corresponding 430/470 intensity ratios in Figure

3.14c. Also on this plot are the decay rates obtained from liquid samples whose initial pH was

adjusted with HCl(aq) (i.e., the calibration points from Figure 3.2 and Figure 3.4). From this

figure it is clear that the correlation between the 430/470 intensity ratio and acridine’s

fluorescence decay rate is independent of whether the acid is already present in the sample

(hollow circles) or added from the gas phase (solid circles).

From Figure 3.14b it is can be seen that introducing HCl(g) to the ice surface results in a

very sudden, step-wise increase in the acridine fluorescence decay rate at 470 nm. The fast

decay rate (~ 0.15 ns-1) observed upon addition of HCl(g) to the neutral frozen sample is similar

to that obtained for a sample acidified with HCl(aq) to an initial pre-freezing pH ~ 3 – 4 (see

Figure 3.9). The fast decay rate is likely due to quenching by deposited chloride at the surface.

The fluorescence decay rates observed upon addition of HCl(g) are plotted as a function of their

corresponding 430/470 intensity ratios as solid symbols in Figure 3.14d, along with the

fluorescence decay rates from samples whose pH was adjusted prior to freezing (i.e. the

calibration points from Figure 3.2 and Figure 3.9) shown as open squares. Contrary to the liquid

case given in Figure 3.14c, here the relationship between the 430/470 ratio and the fluorescence

decay rate is different depending on whether HCl(g) is added from the gas phase or obtained

by freezing solutions of given pH.

At the high partial pressures of HCl used in this experiment (10 Pa) or ~75 mTorr, HCl is

expected to induce surface melting(38). However, the acridine fluorescence decay rates that we

80

obtained during the uptake experiments to the ice surface (Figures 3.14b and 3.14d) are

representative of a ‘frozen’ acridine environment, indicating that over the course of the

experiment, the surface did not melt significantly.

Figure 3.14. (a) Acridine 430/470 intensity ratio as a function of time relative to the introduction of HCl(g).

Measured at the liquid surface (black circles) and at the frozen surface (two trials shown, red squares and

triangles). The purple line indicates the time at which HCl(g) was introduced to the chamber (time = 0).

(b) Acridine fluorescence decay rate at 470 nm as a function to time relative to the introduction of HCl(g).

Measured at the liquid surface (black circles, left axis) and at the frozen surface (red squares, right axis).

Time = 0 denotes the time at which HCl(g) was introduced to the chamber. (c) Acridine fluorescence

decay rates at 470 nm measured at the liquid surface, plotted as a function of their corresponding 430/470

intensity ratios. Shown are a sample prior to HCl(g) exposure (green triangle) and during exposure to

HCl(g) (i.e., the liquid surface data from Figures 3.14a and 3.14b) (solid black circles). Samples whose

initial pH values were either unadjusted or adjusted with HCl(aq) (i.e., the liquid surface data from

Figure 3.2) are illustrated as open circles. (d) Acridine fluorescence decay rates at 470 nm measured at the

frozen surface, plotted as a function of their corresponding 430/470 intensity ratios. A sample prior to

HCl(g) exposure (green triangles) and during exposure to HCl(g) (i.e., the frozen surface data from

Figures 3.14a and 3.14b) (solid red squares). Samples whose initial pH values were either unadjusted or

adjusted with HCl(aq) (i.e., the frozen surface data from Figure 3.2) (open squares).

(a) (b)

(c) (d)

81

3.4. Discussion

3.4.1. Developing pH Probes for the Frozen Surface

We used glancing-angle LIF and acridine and harmine as surface-sensitive, fluorescent pH

probes. Our results show that the locations of the probe molecule spectral features are

unchanged upon freezing (i.e., neutral (Ac) acridine exhibits its fluorescence maximum at 430

nm on both the liquid and frozen surface, and so on). As a result, the measures which reflect pH

at the liquid water surface (the acridine 430/470 ratio and the harmine 290/320 ratio) most likely

also reflect pH at the frozen water surface. The different probe molecules were chosen for two

reasons. First, they are responsive over different pH ranges, with acridine sensitive from pH 4 –

7, and harmine sensitive from pH 7 – 10. Second, harmine’s fluorescence is not readily

quenched in high ionic strength environments. This second factor is important since it is our

goal to extend this technique to the study of frozen salt water and frozen seawater samples.

Fluorescence spectra acquired at the frozen surface show red-shifted features which are

likely due to self-associated probe molecule. These features are not observed at the liquid

surface for the same pre-freezing probe molecule concentration, consistent with both laboratory

(36, 39, 40) and theoretical (40-42) studies which suggest that polycyclic aromatic compounds

readily self-associate at the ice surface at submonolayer coverage. The aggregation of the dye

molecules is thought to be due to a) their enriched surface concentrations due to exclusion (for

example, Heger et al. (39) estimate a concentration enhancement of six orders of magnitude on

slow freezing methylene blue at 243 K) and b) favourable intermolecular interactions within the

QLL (39). The fact that we observe contributions from self-associated acridine (or harmine) in

our glancing-angle laser-induced fluorescence spectra gives us confidence that the probe

molecules are present at the ice surface. Molecular dynamics studies have also shown that

polycyclic aromatic compounds deposited from the gas phase show an affinity for the air-ice

interface (naphthalene in this study) (42).


As illustrated by Figure 3.2, freezing water containing trace amounts of HCl or HNO3

82

results in a surface pH (as inferred from the acridine 430/470 ratio) which is largely unchanged

with respect to the initial pH. These results suggest that protons (H3O+) are not strongly

excluded to the ice surface during freezing. Freezing samples containing trace amounts of

NaOH may result in a surface pH which is slightly lowered with respect to the initial pH, as

shown in Figures 3.2 and 3.7. Figure 3.7 suggests a maximum decrease of 2 pH units during

freezing. Experiments in which the surface pH was monitored during the freezing process

revealed that the observed pH decrease occurs rapidly only during the final stages of freezing.

However, other explanations for the apparent pH lowering may exist. For example, the

apparent decrease in surface pH on freezing may be, in part, due to the temperature

dependence of the pKa of the probe molecule. Such a temperature dependence has been

reported for acridine by Huh et al. (43), who find that its pKa increases with decreasing

temperature (43). Thus, for the same proton concentration, a colder solution will have relatively

more protonated acridine (AcH+) than the warmer solution. Using the temperature dependence

given by Huh et al. (43) (which they calculated from change in acridine’s absorption spectrum),

the pKa increases from ~ 5.4 at room temperature, to ~ 5.8 at 273 K. If their formulation remains

correct at T < 273 K, we would expect pKa ~ 6 at 263 K. Some of the apparent pH lowering may

also be due to the dissolution of atmospheric CO2 (either during sample preparation, freezing,

or melting) as discussed in Section 3.3.2.

These results indicate that freezing water containing small concentrations of acids does

not result in a large change in surface pH. Since HCl and HNO3 have a low solubility in bulk ice

(44, 45), they are likely rejected during freezing. Indeed, the rejection of HCl from the ice matrix

is evidenced by the increase in the acridine fluorescence decay rate with increasing chloride

content, as illustrated by Figure 3.11. In spite of this, the rejection of the acids does not lead to

an enhanced surface concentration of protons; in other words, predicting concentration

enrichments at the surface is not trivial. Wolff and Mulvaney (46) used SEM with x-ray

microanalysis system to investigate the solubility of HCl in ice and the diffusion of HCl through

ice. They rapidly froze 0.1 M HCl solutions, and found that within crystals and at most grain

83

boundaries, the concentration of HCl was very low (below the limit of detection ~0.01 M) but at

triple junctions the concentration was ~ 8 M (close to the eutectic concentration for HCl/water

mixtures in equilibrium with ice). Their result also indicates that although HCl is excluded

during freezing, the concentration of HCl may be different within the different ‘unfrozen

regions’. Previously we studied the exclusion of the nitrate anion on freezing Mg(NO3)2(aq)

solutions and found that surface nitrate concentrations were also lower than predicted (11).

Only a few groups have attempted to study the pH in the unfrozen regions of ice, and

none of them at the air-ice interface. Moreover, in many of these studies (20-22), an inorganic

salt was present in addition to the acid, and inferred pH changes were attributed to the

formation of a freezing potential. The observed pH change was found to depend on the sign of

the freezing potential, which in turn depends on the identities of the salt ions (i.e., on which ion

will be more effectively incorporated into the ice lattice). In the absence of salts, as in the present

experiments, the high mobility of H+ and OH¯ will prevent the formation of a freezing potential.

3.4.3. Exclusion of Chloride

Analysis of the acridine fluorescence decay rates indicates that chloride is excluded to the

ice surface during the freezing of samples containing HCl. Quantification of the concentration

enhancement is difficult for two reasons. Firstly, the quenching effect (chloride concentration

dependence) is convolved with the pH dependence. Secondly, the overall fluorescence is

quenched on ice. However, decay rates begin to increase at an initial pH ~ 3 on the liquid

surface and at an initial pH ~ 5 on the ice surface, which may suggest a concentration

enrichment of around two orders of magnitude. Although such a freeze-induced concentration

enrichment is a well accepted phenomenon which has also been predicted by molecular

dynamics simulations (47), only one laboratory study to date (48) has been able to directly

monitor surface concentrations of Cl¯ on frozen media. An NMR study by Cho et al. (8)

indirectly measured enhanced concentrations of Cl¯ in the unfrozen portion of a frozen sample.

In both cases, the solute was NaCl and not HCl. In the Cho et al. (8) paper, the authors present a

formulation for calculating QLL concentrations. Using their formulation and an initial HCl

84

concentration of 10-5 M (i.e., initial pH = 5), we calculate a concentration enrichment of over 5

orders of magnitude at 263 K. The liquid fraction that is calculated using their formulation is

very small, on the order of ~ 2 × 10-7. Recall that we are unable to directly quantify the extent of

chloride enrichment at the surface from the fluorescence decay rates so the discrepancy may not

be significant. However, we do note that although this formulation may hold for some

inorganic salts, it is not universal (11) and acids may behave differently.

3.4.4. Interfacial pH Changes from Gas Phase Deposition of HCl

When HCl(g) is deposited to the liquid water surface with near-neutral initial pH, the

surface pH – as inferred from the acridine 430/470 ratios and the fluorescence decay rates –

decreases rapidly. Such a decrease is expected: hydrolysis of HCl(g) should readily occur in the

interfacial region and diffusion of H3O+(aq) and Cl¯(aq) into the bulk should follow (28). In

consequence, the inferred pH at the liquid surface is independent of whether the acid is already

present in the sample or added from the gas phase.

However, when HCl(g) is introduced to a frozen sample with near-neutral initial pH, the

surface pH – as inferred from the acridine 430/470 ratio - decreases very slowly despite the fact

that the final melted pH indicates that the same amount of HCl is delivered to the frozen sample as to the

liquid sample. Exposing ice to HCl(g) gives rise to a final 430/470 ratio which is representative of

a mildly acidic surface pH while melting and re-freezing the same sample gives rise to a 430/470

ratio which is representative of a very acidic surface pH. Thus, pH at the frozen surface

depends on whether the acid is already present in the sample or added from the gas phase. The

very slow change in surface pH upon deposition of gas phase HCl(g) was also inferred from

measurements made using the harmine 290/320 ratio. The fact that both harmine and acridine

show a similar response gives us confidence that the phenomenon we observe is real.

The slow change we observe in surface pH as acid is deposited from the gas phase was

unexpected. The low final melted pH shows that HCl(g) is indeed taken up by the ice sample.

This is consistent with previous studies which show that the uptake of HCl(g) to ice surface is

85

very efficient(24). If HCl is being deposited, why does the pH decrease so slowly? One

possibility is that HCl ionization is inhibited at the ice surface and that HCl exists at least partly

in a molecular state, adsorbed at the surface. Several groups have studied HCl present on ice

surfaces (both theoretically and experimentally) and find that ionic and molecular HCl can

coexist at very low temperatures (< 150 K) (49-52). Note that these experiments were performed

at very low HCl partial pressures. It is possible that in our experiments, decreased water

availability for solvation in the QLL may also play an important role (53).

Where does the HCl go? The solubility of HCl in bulk ice is low (46) (for example, Hanson

and Mauersberger (44) place an upper limit of 0.01 mol % for the solubility of HCl in single

crystal ice under stratospheric conditions) and the diffusion coefficient of HCl is slow (for single

crystals of ice at 258 K has been reported (45) as 10-12 cm2 s-1). Thus, solid state diffusion of HCl

away from the ice surface is not likely to account for our observations. Rather, it is likely that

the deposited HCl migrated away from the surface along grain boundaries (and into liquid

pockets within the bulk ice). Molina et al. (54) found that HCl was readily taken up to the ice

surface. Later it was suggested that the observed uptake was due to high HCl partial pressures

and melting along grain boundaries. As noted earlier, Wolff and Mulvaney (46) also found that

HCl showed an affinity for nodes within the bulk ice. These observations are consistent with

those of the present experiments. Interestingly, we have observed a decrease in surface pH due

to the deposition of HCl(g) to a frozen salt water sample, whose surface is expected to be wetted

by a ‘true’ brine layer (55).

3.5. Conclusion

In conclusion, we have shown that glancing-angle LIF can be used in conjunction with

interface-sensitive, pH-sensitive, fluorescent probes to study pH at the air-ice interface. This

work represents the first use of this approach to investigate acidity on frozen media.

Fluorescence spectra acquired at the ice surface showing contributions from aggregated probe

molecules point to the surface-selectivity of this technique. Via the quenching of the acridine

fluorescence lifetime, we found strong evidence for a freeze-induced enrichment in chloride

86

concentration. This result also points to the surface-selectivity of the technique and constitutes

an interesting experimental observation of the freeze-concentration phenomenon.

In this study we have found that pH at the air-ice interface is largely unchanged with

respect to the pre-freezing pH. In other words, when slightly acidic solutions (containing HCl or

HNO3) are frozen, protons are not strongly excluded to the ice surface. When basic solutions

(containing NaOH) are frozen, surface pH decreases by no more than two pH units. This work

indicates that predicting the concentration of species at the air-ice interface is not trivial, and

that care should be taken in assuming total exclusion of solutes there.

We monitored pH at the frozen surface during the deposition of HCl from the gas phase.

The results show that proton concentrations increases more slowly than expected. It is likely

that in these experiments, uptake leads to the diffusion of HCl along grain boundaries, away

from the ice surface. Future work should focus on better understanding how trace acids in the

atmosphere will affect the acidity of ice and snow surfaces. These results may have significant

implications for interpreting multiphase chemistry occurring on frozen media.

3.6. References


1510 (2002)











4375-4418 (2007)

4. Y. Li, G. A. Somorjai, Surface premelting of ice. J. Phys. Chem. C 111, 9631-9637 (2007)


interfaces. Phys. Rev. B: Condens. Matter 66, 085401 (13 pp) (2002)

87


Phys. Chem. A 111, 11006-11012 (2007)




layer in ice/brine mixtures. J. Phys. Chem. C 106, 11226-11232 (2002)


surface disordering on ice. Proc. Natl. Acad. Sci. U. S. A. 103, 9422-9427 (2006)


Atmos. Chem. Phys. 11, 9971-9982 (2011)

11. S. N. Wren, D. J. Donaldson, Exclusion of Nitrate to the Air - Ice Interface During Freezing. J.

Phys. Chem. Lett. 2, 1967 - 1971 (2011)



13. N. Takenaka, H. Bandow, Chemical kinetics of reactions in the unfrozen solution of ice. J. Phys.

Chem. A 111, 8780-8786 (2007)



Chem. 100, 13874-13884 (1996)



(2008)

16. D. O'Sullivan, J. R. Sodeau, Freeze-Induced Reactions: Formation of Iodine-Bromine Interhalogen

Species from Aqueous Halide Ion Solutions. J. Phys. Chem. A 114, 12208-12215 (2010)

17. E. J. Workman, S. E. Reynolds, A Suggested Mechanism for the Generation of Thunderstorm

Electricity. Phys. Rev. 74, 709-709 (1948)

18. V. L. Bronshteyn, A. A. Chernov, Freezing Potentials Arising on Solidification of Dilute Aqueous-

Solutions of Electrolytes. J. Cryst. Growth 112, 129-145 (1991)

19. P. W. Wilson, A. D. J. Haymet, Effect of Ice Growth Rate on the Measured Workman-Reynolds

Freezing Potential between Ice and Dilute NaCl Solutions. J. Phys. Chem. C 114, 12585-12588

(2010)


electrolyte solutions. J. Phys. Chem. C 110, 7613-7616 (2006)

21. D. Heger, J. Klanova, P. Klan, Enhanced protonation of cresol red in acidic aqueous solutions

caused by freezing. J. Phys. Chem. C 110, 1277-1287 (2006)

22. N. Takenaka, M. Tanaka, K. Okitsu, H. Bandow, Rise in the pH of an unfrozen solution in ice due

to the presence of NaCl and promotion of decomposition of gallic acids owing to a change in the

pH. J. Phys. Chem. A 110, 10628-10632 (2006)

23. J. P. D. Abbatt, Interactions of atmospheric trace gases with ice surfaces: Adsorption and reaction.

Chem. Rev. 103, 4783-4800 (2003)

24. T. Huthwelker, M. Ammann, T. Peter, The uptake of acidic gases on ice. Chem. Rev. 106, 1375-1444

(2006)

25. F. Domine, E. Thibert, E. Silvente, M. Legrand, J. L. Jaffrezo, Determining Past Atmospheric Hcl

Mixing Ratios from Ice Core Analyses. J. Atmos. Chem. 21, 165-186 (1995)

26. V. F. McNeill, F. M. Geiger, T. Loerting, B. L. Trout, L. T. Molina, M. J. Molina, Interaction of

hydrogen chloride with ice surfaces: The effects of grain size, surface roughness, and surface

disorder. J. Phys. Chem. A 111, 6274-6284 (2007)

27. R. M. Sayer, R. D. B. Gatherer, J. P. Reid, A laser induced fluorescence technique for determining

88

the pH of water droplets and probing uptake dynamics. Phys. Chem. Chem. Phys. 5, 3740-3747

(2003)

28. D. Clifford, T. Bartels-Rausch, D. J. Donaldson, Suppression of aqueous surface hydrolysis by

monolayers of short chain organic amphiphiles. Phys. Chem. Chem. Phys. 9, 1362-1369 (2007)

29. D. Clifford, D. J. Donaldson, Direct Experimental Evidence for a Heterogeneous Reaction of

Ozone with Bromide a the Air-Aqueous Interface. J. Phys. Chem. A 111, 9809-9814 (2007)

30. A. Martin, R. Narayanaswamy, Studies on quenching of fluorescence of reagents in aqueous

solution leading to an optical chloride-ion sensor. Sens. Actuators, B 39, 330-333 (1997)

31. A. Dias, A. P. Varela, M. D. Miguel, R. S. Becker, H. D. Burrows, A. L. Macanita, Beta-carbolines.

2. Rate constants of proton transfer from multiexponential decays in the lowest singlet excited

state of harmine in water as a function of pH. J. Phys. Chem. 100, 17970-17977 (1996)

32. M. Balon, J. Hidalgo, P. Guardado, M. A. Munoz, C. Carmona, Acid-Base and Spectral Properties

of Beta-Carbolines .1. Tetrahydro-Beta-Carbolines. J. Chem. Soc., Perkin Trans. 2, 91-97 (1993)

33. M. J. Tapia, D. Reyman, M. H. Vinas, C. Carcedo, J. M. L. Poyato, Acid-base equilibria of methyl

beta-carboline-3-carboxylate in aqueous solution. J. Lumin. 101, 227-234 (2003)

34. M. M. Gonzalez, J. Arnbjerg, M. P. Denofrio, R. Erra-Balsells, P. R. Ogilby, F. M. Cabrerizo, One-

and Two-Photon Excitation of beta-Carbolines in Aqueous Solution: pH-Dependent

Spectroscopy, Photochemistry, and Photophysics. J. Phys. Chem. A 113, 6648-6656 (2009)


reaction between O3(g) and halides at the surface of frozen salt solutions. J. Geophys. Res. D:

Atmos. 115, D16309 (8 pp) (2010)


surfaces. J. Phys. Chem. A 111, 1277-1285 (2007)

37. T. F. Kahan, D. J. Donaldson, Heterogeneous ozonation kinetics of phenanthrene at the air-ice

interface. Environ. Res. Lett. 3, (2008)

38. M. J. Molina, in The Chemistry of the Atmosphere: Its Impact on Global Change, J. G. Calvert, Ed.

(Blackwell Scientific, Boston, MA, 1994), pp. 27



40. D. Heger, D. Nachtigallova, F. Surman, J. Krausko, B. Magyarova, M. Brumovsky, M. Rubes, I.

Gladich, P. Klan, Self-Organization of 1-Methylnaphthalene on the Surface of Artificial Snow

Grains: A Combined Experimental-Computational Approach. J. Phys. Chem. A 115, 11412-11422

(2011)

41. D. Ardura, T. F. Kahan, D. J. Donaldson, Self-Association of Naphthalene at the Air-Ice Interface.

J. Phys. Chem. A 113, 7353-7359 (2009)



(2011)

43. Y. Huh, J. G. Lee, D. C. McPhail, K. Kim, Measurement of pH at Elevated-Temperatures Using

the Optical Indicator Acridine. J. Solution Chem. 22, 651-661 (1993)

44. D. R. Hanson, K. Mauersberger, HCl/H2O Solid-Phase Vapor-Pressures and HCl Solubility in Ice.

J. Phys. Chem. 94, 4700-4705 (1990)

45. E. Thibert, F. Domine, Thermodynamics and Kinetics of the Solid Solution of HNO3 in Ice. J. Phys.

Chem. C 102, 4432-4439 (1998)

46. E. W. Wolff, R. Mulvaney, K. Oates, Diffusion and Location of Hydrochloric-Acid in Ice -

Implications for Polar Stratospheric Clouds and Ozone Depletion. Geophys. Res. Lett. 16, 487-490

(1989)


89

99-103 (2007)


and Frozen NaCl Solutions Probed by XPS and NEXAFS. ChemPhysChem 11, 3859-3866 (2010)

49. L. Delzeit, B. Rowland, J. P. Devlin, Infrared-Spectra of Hcl Complexed Ionized in Amorphous

Hydrates and at Ice Surfaces in the 15-90-K Range. J. Phys. Chem. 97, 10312-10318 (1993)

50. S. C. Park, H. Kang, Adsorption, ionization, and migration of hydrogen chloride on ice films at

temperatures between 100 and 140 K. J. Phys. Chem. C 109, 5124-5132 (2005)

51. H. Kang, T. H. Shin, S. C. Park, I. K. Kim, S. J. Han, Acidity of hydrogen chloride on ice. J. Am.

Chem. Soc. 122, 9842-9843 (2000)

52. P. Parent, C. Laffon, Adsorption of HCl on the water ice surface studied by x-ray absorption

spectroscopy. J. Phys. Chem. C 109, 1547-1553 (2005)

53. D. Ardura, D. J. Donaldson, Where does acid hydrolysis take place? Phys. Chem. Chem. Phys. 11,

857-863 (2009)

54. M. J. Molina, T. L. Tso, L. T. Molina, F. C. Y. Wang, Antarctic Stratospheric Chemistry of Chlorine

Nitrate, Hydrogen-Chloride and Ice - Release of Active Chlorine. Science 238, 1253-1257 (1987)

55. S. N. Wren, D. J. Donaldson, How does deposition of gas phase species affect pH at frozen salty

interfaces? Atmos. Chem. Phys. 12, 10065-10073 (2012)

91

C H AP T E R F OU R

HOW DOES THE DEPOSITION OF GAS PHASE SPECIES AFFECT pH AT SALTY

INTERFACES?


Reprinted from S. N. Wren and D. J. Donaldson, How does the deposition of gas phase species

affect pH at salty interfaces?, Atmospheric Chemistry and Physics, 12, 10065 – 10073 (2012).

DOI:10.5194/acp-12-10065-2012.



92

4.1 Introduction

Natural ice and snow were once considered to be chemically inert, but a growing body of

evidence shows that air-ice chemical interactions can significantly influence the composition of

the overlying atmosphere (1-3). For example, the photolysis of snowpack nitrate has been

shown to lead to upward fluxes of NOx (= NO + NO2) and HONO (2, 4-6) and heterogeneous

reactions occurring on sea ice surfaces are thought to be responsible for the activation and

release of reactive halogen species to the gas phase (3, 7). Although the importance of snow and

ice as (photo)chemical reactors is now recognized, a molecular-level understanding of chemical

reactions occurring on these substrates is still lacking. This is due, in part, to an inadequate

characterization of the surfaces on which air-ice chemical processes occur. As a result, and in

lieu of a better option, aqueous-phase physical-chemical mechanisms are often used to interpret

or model heterogeneous chemistry occurring on snow and ice substrates (8).

Accordingly, in keeping with the pH-sensitivity of their liquid phase counterparts, many

reactions occurring on ice are thought to be pH-sensitive. For example, a key reaction for

bromine activation is the heterogeneous oxidation of sea salt bromide by hypobromous acid:

HOBr + Br¯ + H+ → Br2 + H2O (R1)

The heterogeneous ozonation of bromide leads to the formation of OBr¯ which is in equilibrium

with HOBr. Reaction R1, which consumes one reactive bromine species but produces two

(leading to the ‘bromine explosion’) is acid-catalyzed in the aqueous phase (9). Based on this

pH-dependence, it has been thought (10, 11) that the medium on which bromine explosion

chemistry occurs should achieve a low enough pH for this reaction to proceed at an appreciable

rate. Similarly, the photolytic release of HONO from the snowpack (via R2 or R3 followed by

R4) is thought to be proton mediated:

NO3¯ + hν → NO2¯ + O(3P) (R2)

NO3¯ + O(3P) → NO2¯ + O2 (R3)

93

NO2¯ + H+ HONO (R4)

However, some recent kinetic and photochemical measurements indicate that reactions

occurring at the air-ice interface are not well represented by parameters measured at the air-

aqueous interface (12-14). A better understanding of the pH-sensitivities of ice surface processes

such as bromine activation and snowpack HONO production is needed. Clearly, this in turn

requires a better knowledge of the local pH where the key reactions occur. This is difficult to

obtain experimentally because surface pH measurement techniques are lacking, and the bulk

pH measured upon melting of a sample may not be the same as that present at the air-ice

interface.

Our ability to measure, predict, or even define, properties such as pH at the ice surface is

complicated by the presence of a disordered interfacial layer there. At temperatures as low as ~

240 K, a relatively pure ice surface is covered by what is commonly referred to as the quasi-

liquid layer (QLL): a disordered layer some tens of nanometres thick that has properties

intermediate to that of liquid water and bulk ice (15-17). The surface of snow grains within a

relatively ‘clean’ snowpack may be covered by such a genuine QLL. The thickness of the liquid-

like region is temperature dependent and also increases with increasing concentration of

impurities, such as salts or acids, (18-20), until the system enters a regime in which a liquid

brine layer coexists with pure ice, in accordance with multi-component phase equilibria (21, 22).

Trace acids (e.g., HNO3, HCl or H2SO4) in the atmosphere can be deposited to the snowpack or

sea ice surface, potentially altering the local pH there. Currently it is not known how the nature

of the interfacial layer (QLL vs. brine layer) affects the surface pH following deposition of trace

atmospheric acids.

Critically, it is also not known how or whether the uptake of trace acids affects local pH at

sea ice surfaces. Seawater with pH ~ 8.3 is naturally buffered against pH changes by the

carbonate system (i.e., dissolved CO2, HCO3¯, and CO32¯ in equilibrium with atmospheric CO2).

It is important to determine whether this buffering is maintained at the frozen surface since

94

seawater is generally thought to be too alkaline to activate bromine via the acid-catalyzed

mechanism shown above (R1). Sander et al. (23) first proposed that the precipitation of

carbonate during the freezing of seawater could lead to a reduced buffering capacity in the

brine, allowing the pH to drop following acid deposition thereby triggering “bromine

explosion” chemistry. However, Morin et al. (11), using improved thermodynamic parameters,

showed that the identity of the calcium carbonate polymorph influences the total alkalinity (and

thus the pH) of the brine. In their model, when calcite (CaCO3) precipitates at ~ 271 K the total

alkalinity of the brine drops, significantly reducing the buffering potential, and allowing acid

deposition to lower the pH. However, if ikaite (CaCO3·6H2O) is the form to precipitate (at ~268

K), the total alkalinity of the brine does not drop below its initial value. Recent field studies in

both the Arctic (24) and Antarctic (25) have shown that it is ikaite, and not calcite, that

precipitates in freezing seawater brine.

In the present study, we use glancing-angle laser-induced fluorescence (LIF)(26) to

address two questions concerning pH at the air-ice interface. First, does the nature of the

interfacial layer (QLL vs. brine layer) affect pH changes at the surface caused by the deposition

of gas phase acid? Wren and Donaldson (26) (Chapter Three) showed clear evidence for

different pH behaviour following acid deposition at the frozen vs. liquid freshwater surfaces.

And second, is there a change in the buffering capacity of seawater at the air-ice interface upon

freezing? Answering these questions will provide further insights into the environmental air-ice

interface and help to constrain models of chemical reactions there.

4.2 Materials and Methods

4.2.1 Experiment Overview

Three types of frozen surface were investigated in this study: frozen freshwater (i.e.,

deionized water), frozen salt water (0.5 M NaCl in deionized water) and frozen “0.5 M” artificial

seawater (“Instant Ocean®” dissolved in deionized water). Glancing-angle laser-induced

fluorescence (LIF) was used in conjunction with a surface-active fluorescent pH indicator to

study pH at these frozen surfaces. Surface-selectivity is achieved through the use of a laser

95

beam which impinges the sample surface at a very shallow, glancing angle (> 85 from the

surface normal). This technique, which has been used extensively for the study of atmospheric

interfaces (27-30), has been shown to be sensitive to the upper ~ 100 monolayers (~ 50 nm) of an

aqueous surface(15). In the present experiments, we also expect to be probing a narrow

interfacial region because a) of the known aqueous phase surface activity of the pH indicator

(27); b) of the fact that this class of compounds is thought to be excluded to and self-aggregate at

the air-ice interface (following exclusion during freezing or deposition from the gas phase) (31-

33) and c) we are investigating changes due to deposition of gas phase species to the surface.

Here we used harmine (7-methoxy-1-methyl-9H-pyrido[3,4-b]indole) – a surface-active

fluorescent dye whose absorption and fluorescence properties depend on pH (34, 35) – as the

pH indicator. This technique was first developed for the frozen freshwater surface and has been

described in greater detail in Chapter Three and in Wren and Donaldson (26). In the present

study we extend this method to the study of frozen salt water and frozen sea water surfaces.

Harmine is particularly suited to this application since its fluorescence is not readily quenched

in the higher ionic strength environments presented by the salt solutions.

4.2.2 Apparatus

The experimental apparatus consisted of a Teflon reaction chamber outfitted with 2”

diameter quartz windows to allow the laser beam to enter and exit. Side ports allowed gases to

be introduced and ventilated from the chamber. A liquid light guide was suspended from the

top of the chamber and samples were placed on a piece of stainless steel shimstock resting on its

copper floor below the liquid light guide. The temperature of the coolant flowing through

copper coils beneath the floor was controlled such that the sample temperature was ~ 263 K.

Fluorescence at the sample surface was induced using the frequency-doubled output of an

optical parametric oscillator (OPO) (10 Hz, ~ 0.4 mJ per pulse) pumped by the frequency-triplet

output of a Nd:YAG laser. The laser beam impinged the surface at > 85 from the surface

normal. Fluorescence was collected by the liquid light guide and imaged onto the entrance slit

of a ¼ m monochromator. The emission was detected and amplified by a photomultiplier tube

96

and then read out by a digital oscilloscope averaging the fluorescence decay profile over 64

laser shots. A LabView program running on a PC sampled a 5 ns time slice from the

fluorescence decay.

Glancing-angle harmine excitation spectra were acquired by monitoring fluorescence at

~430 nm (near the cation emission maximum) while scanning the excitation wavelength in 5 nm

steps from 260 nm – 355 nm. The excitation spectra were normalized to 1 at an excitation

wavelength of 320 nm. Figure 4.1 shows examples of spectra acquired in this way. During

experiments where the pH was monitored during deposition of gas-phase species, the ratio of

the fluorescence intensities measured following excitation at 290 nm and 320 nm (the harmine

290/320 ratio) was recorded (vide infra).

4.2.3 Sample Preparation

A stock 2.5 × 10-5 M harmine solution was prepared volumetrically by dissolving harmine

(Aldrich, 98%) in 18 mΩ deionized water. Experimental solutions containing 1.0 × 10-7 M

harmine were prepared fresh daily by diluting the stock solution in deionized water (the

freshwater samples), 0.5 M NaCl (the salt water samples) or ‘0.5 M’ Instant Ocean® (the

seawater samples). The 0.5 M NaCl solution was prepared volumetrically by dissolving sodium

chloride crystals (ACP Chemicals, min 99.0%) in 18 mΩ deionized water (100 mL). The ‘0.5 M’

seawater solution was prepared volumetrically by dissolving the same mass of Instant Ocean®

(Spectrum Brands) in 18 mΩ deionized water (100 mL). The composition of the artificial sea salt

as reported by Langer et al. (36) is predominantly: chloride (47.53 wt%), sodium (26.45 wt%),

sulfate (6.41 wt%), magnesium (3.19 wt%), calcium (1.00 wt%), potassium (0.952 wt%),

bicarbonate (0.356 wt%) and bromide (0.16 wt%). Samples were prepared by spreading 4 mL of

solution onto a stainless steel plate resting on the chamber floor; the resulting samples had a

surface area of ~ 6 cm2. Prior to sample introduction, the chamber was purged with a dry N2

flow at 0.5 SLPM (Linde, Grade 4.8). Ice samples were frozen from the bottom-up under the dry

N2 flow with the floor of the chamber cooled to 263 K.

97

4.2.4 Bulk pH Measurements

The bulk pH of the experimental solutions was measured using a commercial pH

electrode (Orion Model 520 A) which was calibrated daily using a three-point calibration (pH

4.01, 7.00, 10.00). For simplicity, in the text we use the term ‘initial pH’ to refer to the bulk, pre-

freezing, solution pH. The freshwater and salt water solutions were initially slightly acidic (pH

~ 5.9), as governed by equilibrium with atmospheric CO2 while the seawater solution had an

initial pH of ~ 8.1. Initial sample pH was occasionally adjusted using NaOH(aq) (prepared from

ACP Chemicals, min 97.0%) or HCl(aq) (prepared from Fisher Scientific, 36.5 – 38.0%). In some

cases the final pH of the melted samples was also measured using the commercial pH electrode.

To do this, the frozen sample was removed from the chamber immediately after the experiment,

separated from the stainless steel shimstock, and allowed to melt in a covered beaker. The final

bulk pH was obtained in order to gain a rough sense of the total acid or base uptake.

4.2.5 Introduction of HCl(g) or NH3(g)

pH decreases at the frozen freshwater vs. frozen salt water surface were studied following

the deposition of gas phase HCl to the surface. In these experiments, the initial sample pH was

adjusted to be > 9 using dilute NaOH(aq). Samples were frozen and then glancing-angle LIF

was used to obtain the harmine 290/320 ratio. A stable 290/320 ratio was obtained under a 0.5

SLPM N2(g) flow. This ratio was then followed as a function of time under a 0.5 SLPM flow of

100 ppmv HCl(g) in N2(g). The same conditions were used in our previous study (Chapter

Three, ref (26)), where we calculated that ~5-10% of the total HCl flowed through the chamber

was taken up by the frozen samples. High gas-phase concentrations were used in order to

obtain an observable in our experiments over a reasonable period of time. HCl-induced

disordering at the surface is a concern and the HCl-water phase diagram (37) does indicate that

at equilibrium a liquid-phase should be present for HCl vapour pressures used in the

experiments. However, although very slight surface melting may occur in these experiments,

based on measurements of acridine’s fluorescence lifetime (26) which differ for liquid and

frozen environments, we do not believe a liquid-layer forms under our conditions.

98

A similar approach was used to investigate pH increases at the frozen salt water vs.

frozen seawater surface. In those experiments, the initial salt water pH was adjusted to bring it

close to the seawater pH (~8.1) and the artificial seawater pH was not adjusted (pH ~8.1). Gas

phase ammonia was delivered to the chamber by passing a 50 SCCM flow of dry N2(g) through

a round-bottom flask containing a 12.75 wt% NH3(aq) solution kept at 253 K. Because the

solution froze, we cannot properly calculate an ammonia partial pressure from the parameters

of Clegg and Brimblecombe (38). Had the solution remained liquid, the ammonia partial

pressure is estimated to be ~ 70 Pa

4.3 Results and Discussion

4.3.1 Harmine as a pH Indicator at the Frozen Salt Water Surface

In the ground state, the pyridinic nitrogen on neutral harmine can accept a proton to form

a cation (with reported pKa ~ 7.7 at 298 K (35)). Both the neutral and cationic forms absorb in

the UV with absorption maxima at ~300 nm and ~320 nm for the neutral and cationic forms,

respectively. Following excitation in the UV, the harmine cation fluoresces strongly at ~ 400 nm

while neutral harmine fluoresces more weakly at ~ 350 nm. Previously (26), we showed that for

the frozen freshwater surface, the shape of harmine’s excitation spectrum (i.e., the spectrum

obtained by monitoring harmine fluorescence at ~430 nm while scanning the excitation

wavelength over the neutral and cationic absorptions) reflects the relative abundances of the

two species. This is illustrated in Figure 4.1, which shows harmine excitation spectra acquired at

the frozen freshwater surface: as the initial solution pH is increased from neutral (dashed red

line) to strongly basic (solid red line), the peak maximum shifts from ~ 320 nm to ~ 290 nm. In

this study we used harmine to study frozen salt water/sea water surfaces for the first time. Also

shown on Figure 4.1 are harmine excitation spectra acquired at the frozen salt water surface for

near-neutral (dashed green line) and strongly basic (solid green line) initial pH’s. Figure 4.1

illustrates that the excitation spectra acquired at the frozen salt water surface are similar to

those acquired at the frozen freshwater surface for similar initial pH values. In particular, the

locations of the neutral and cation absorptions are not influence by the different environments

presented by the frozen salt water vs. frozen fresh water surfaces (e.g., by higher ionic strength

99

or greater liquid layer fraction at the air-ice interface).

Previously (26) we reported that the ratio of harmine’s fluorescence intensity following

excitation at 290 nm to that following excitation at 320 nm (hereinafter the harmine 290/320

ratio) provides a useful parameter for studying pH at the frozen freshwater surface. The

harmine 290/320 ratios are plotted as a function of pre-freezing pH for frozen freshwater and

frozen salt water samples in Figure 4.2. The pre-freezing pH of each sample was adjusted with

either HCl(aq) or NaOH(aq) as required. Figure 4.2 shows that the harmine 290/320 ratio

obtained at the ice surface using glancing-angle LIF a) increases monotonically with an increase

in pH for pre-freezing pH values > 8.5; and b) is the same at the frozen freshwater vs. frozen salt

water surface. The good agreement gives us confidence that harmine can be similarly used to

study pH changes at the frozen salt water surface.

As reported in Wren and Donaldson (26) and Chapter Three, and as illustrated by the red

symbols in Figure 4.3, the freezing of freshwater samples is accompanied by a marked decrease

in the harmine fluorescence intensity (by ~50% or more). We have noted such fluorescence

quenching previously, for other probe compounds present at frozen interfaces (26). However, as

illustrated by the blue and green symbols in Figure 4.3, freezing salt water and seawater

samples does not result in a decrease in overall fluorescence intensity. Indeed, a slight increase

in harmine fluorescence intensity (~ 10%) is observed upon freezing. The lack of quenching on

the frozen salt water surface is suggestive of a more liquid-like environment there, in accord

with thermodynamic models (22).

100

Figure 4.1. Harmine excitation spectra acquired at the frozen freshwater surface (red traces) and

at the frozen salt water (0.5 M NaCl) surface (green traces) for strongly basic pre-freezing pH ~9.8

(solid traces) and near-neutral pre-freezing pH (dashed traces). Spectra were collected by

scanning the excitation wavelength in 5 nm steps while monitoring harmine fluorescence at ~ 430

nm.


circles) and at the frozen salt water (0.5 M NaCl) surface (green triangles) as a function of pre-

freezing pH. Initial pre-freezing pH was adjusted with NaOH(aq) or HCl(aq) and measured with

a commercial pH electrode.

101

Figure 4.3. Harmine fluorescence intensity (in arbitrary units) measured at 430 nm following

excitation at 320 nm plotted as a function of time relative to the freezing of a freshwater sample

(red circles, left axis), salt water (0.5 M NaCl) sample (green triangles, inner right axis) and

artificial seawater sample (blue squares, outer right axis). The samples froze at t = 0 (indicated by

the dashed grey line).


circles) and at the frozen salt water (0.5 M NaCl) surface (green triangles) as a function of time.

The dashed line indicates the time (t = 0) at which a 0.5 SLPM flow of 100 ppm of HCl in N2 was

introduced to the chamber. The pre-freezing pH of the samples was adjusted with NaOH(aq) to a

pH ~ 9.8. The final melted pH of the freshwater sample was ~ 3 and the final melted pH of the salt

water sample was pH ~ 2.5.

102

4.3.2 pH Changes at the Frozen Freshwater vs. Frozen Salt Water Surface

First we report on changes in surface pH due to the deposition of gas-phase HCl to frozen

freshwater vs. frozen salt water surfaces. Figure 4.4 shows the harmine 290/320 ratio as a

function of time relative to the introduction of HCl(g). The results obtained at the frozen

freshwater surface are shown as red circles and at the frozen salt water surface as green

triangles. The initial pH of both samples was > 9 and the final melted pH of the samples was < 3.

Although a similar amount of HCl(g) was delivered to each sample in a bulk sense, the pH

response at the surface, as inferred from the harmine 290/320 ratio, was very different.

The red symbols in Figure 4.4 show that deposition of HCl(g) to the frozen freshwater

surface results in a very slow (or no) decrease in the harmine 290/320 ratio, which we interpret

as a very slow decrease in the surface pH. The same result – obtained using acridine as a pH

indicator – is reported in Wren and Donaldson (26) and Chapter Three, where the possible

reasons are discussed in detail. In the case of HCl(g) deposition to frozen freshwater, the high

harmine 290/320 ratio measured at the end of the experiment is not consistent with the low pH

measured upon melting the sample. Since a) the solubility of HCl in single crystal ice is low (39)

and b) the diffusion of HCl within single crystal ice has been shown to be very slow (40), it is

unlikely that the deposited HCl is incorporated into the bulk ice matrix. We have previously

suggested (Chapter Three, ref (26) that the deposited HCl may migrate away from the ice

surface along grain boundaries in the ice. Migration of HCl along grain boundaries has also

been inferred in other studies (41, 42). Another possibility may be that HCl remains in a

molecular form at the surface. A few groups have shown that molecular and ionic HCl can

coexist on ice surfaces at very low temperatures (< 150 K) (43-46). However, it is not well known

how the decreased water availability in the QLL may affect dissolution and dissociation of HCl

there.

In contrast to the frozen freshwater result, the green symbols in Figure 4.4 illustrate that

when HCl(g) is deposited to the frozen salt water surface the harmine 290/320 ratio decreases

much more rapidly, indicating a lowering in surface pH (c.f. Figure 4.2). This is consistent with

103

acid dissolution at a liquid brine surface, which is expected thermodynamically to be present for

the frozen salt solution and which is also inferred from the fluorescence intensity results

presented above.

Several studies point to the presence of brine layers at the surface of frozen salt solutions.

Koop et al. (22) showed that frozen NaCl and NaBr mixtures would be partially liquid down to

~ 230 K. Cho et al. (18) used NMR line widths to show that a liquid brine coexists with solid ice

for salt solutions at sub-eutectic temperatures. They found that the system could be well

described using an equilibrium thermodynamic approach. Using their formulation for the

liquid brine fraction, which is based on freezing-point depression, we estimate a total liquid

brine fraction on the order of ~0.1 for a 0.5 M NaCl solution at 263 K. For a ~ 4 mL sample, with

surface area ~ 6 cm2, this translates to a maximum brine thickness of roughly 60 μm, assuming

all the liquid is present exclusively at the air-ice interface (although the samples in this study

did not have the visual appearance of being ‘wet’). Noting that this represents an extreme upper

limit to the brine layer thickness, the fact that the frozen salt water surface responds to the

deposition of HCl(g) as if it were a liquid is not necessarily surprising. By contrast, a pure ice

surface near the freezing temperature, is thought to be covered by a QLL which is much thinner

(10s of nm)(17). As noted above, the properties of this interface are often quite different from

those of a true liquid surface.

Recently, Kahan et al. (30) studied the kinetics of harmine photolysis at the frozen

freshwater surface vs. frozen salt water surface and found that at pre-freezing salt

concentrations greater than 0.1 M, the photolysis kinetics were well described by the aqueous-

surface kinetics but at the ‘pure’ ice surface the photolysis rate was significantly enhanced over

its liquid value. Such a rate enhancement has also been observed for other aromatics (33). The

authors concluded that by increasing the salt concentration, the brine fraction at the surface was

increased until the frozen salt water surface behaved like a true liquid solution. Together with

the present results, these results indicate that a) the frozen freshwater and frozen salt water

surfaces are fundamentally different and that b) the frozen salt water surface provides a

104

chemical environment much like that of a liquid surface, consistent with its greater predicted

liquid fraction.

4.4 pH changes at the frozen salt water vs. frozen seawater surface

Now we address the question of whether the frozen seawater surface is buffered against

pH changes. Because the sensitivity of harmine towards pH changes is greatest in the pH 8 – 10

region (as shown in Figure 4.2), in these experiments we monitored increases in surface pH due

to the deposition of gas-phase NH3(g). The initial pH of the salt water solution was adjusted

using NaOH(aq) to bring it close to the initial pH of the artificial seawater (~ 8.1). The

experimental approach was the same as that described above except that in this case NH3(g)

was introduced to the chamber. The experiment was also performed on room temperature

liquid salt water and seawater samples; as expected, buffering by liquid seawater was observed

in these experiments.

Figure 4.5 shows the harmine 290/320 ratio measured at the frozen salt water surface

(green symbols) and at the frozen seawater surface (blue symbols), as a function of time relative

to the introduction of NH3(g). First, it should be noted that the harmine 290/320 ratio measured

at the frozen seawater surface prior to NH3(g) deposition is consistent with a surface pH ~ 8.1

(i.e., unchanged with respect to the pre-freezing pH). This result is consistent with a previous

study in which we demonstrated that freezing of water samples doped with trace amounts of

acid or base leaves the surface pH largely unchanged with respect to the pre-freezing pH (26).

However, since harmine’s usefulness as a pH indicator is limited to the pH 8 – 10 region, a large

acidification of the surface upon freezing would be impossible to detect.

Deposition of NH3(g) to the frozen salt water surface gives rise to an increase in the

harmine 290/320 ratio, indicative of an increase in surface pH. This result is expected given the

observations reported above which indicated that the frozen salt water surface is covered by a

brine layer which behaves much like a liquid surface. The final melted pH of the salt water

sample was > 10 and the high final 290/320 ratio reflects this strongly basic pH. However, under

105

the same experimental conditions, the harmine 290/320 ratio increases much more slowly, if at

all, at the frozen seawater surface. The molality of the salt in the seawater solution was the same

as that of the salt water solution and thus at the experimental temperature of 263 K, the liquid

layer fractions of the two substrates should have been very similar. As with freezing NaCl

solutions, there was no decrease in fluorescence intensity on freezing of seawater (see Figure

4.3), consistent with our interpretation of a more liquid-like layer on both higher ionic strength

substrates. The difference in the pH response is therefore not explained by a difference in liquid

layer fraction as it was for the freshwater ice. Rather, the results suggest that the frozen

seawater surface is resistant to pH changes, likely due to the brine present at the interface

maintaining its buffering capacity.

Figure 4.5. The harmine 290/320 intensity ratio measured at the frozen salt water surface (green

symbols) and at the frozen artificial seawater (blue symbols) surface as a function of time. The

dashed line indicates the time (t = 0) at which a 50 sccm flow of N2 passing over a 12.75 wt%

NH4OH(aq) solution held at 253 K was introduced to the chamber. The pre-freezing pH of the

salt water samples was adjusted with NaOH(aq) to a pH ~ 8.1, to be the same as the equilibrium

pre-freezing pH of the seawater samples. The final melted pH of the salt water samples was > 10

while the final melted pH of the seawater samples was < 10. The different shapes of the symbols

represent two separate trials.

106

4.5 Atmospheric Implications

Due to the high salinity of seawater (ca. 30‰ ≈ 30 g of salt per kg of solution), at

temperatures above the eutectic, the frozen surfaces on which halide activation chemistry is

expected to occur (sea ice, frost flowers, saline snow etc.) are thought to be wetted by a liquid

brine layer (3). Depending on the processes that accompany freezing, brine salinities may

exceed 100‰ (e.g., on frost flowers)(47). On the other hand, the surface of snow grains within a

relatively ‘clean’ snowpack may be covered by a more genuine QLL. The present work has

shown that deposition of trace atmospheric acids to different frozen substrates may affect the

surface pH quite differently.

Of particular interest to the air-ice community is the question of the pH at surfaces

important for bromine activation. As mentioned in the introduction, a key reaction for bromine

activation (R1), is acid-catalyzed in the aqueous phase (9). Based on this requirement for

hydronium ions (i.e., a low pH), it has been thought (10, 11) that the medium on which bromine

explosion chemistry occurs must have a low pH for this reaction to proceed at an appreciable

rate. Despite this expectation, a few laboratory studies have shown that the heterogeneous

ozonation of frozen seawater (48)or frozen saltwater solutions (49, 50) still leads to bromine

production at ambient (neutral) pH.

The various surfaces which have been proposed for bromine activation chemistry – first-

year sea ice, brines on sea ice, frost flowers, saline snow and sea-salt aerosol – derive their

salinity from seawater, which is naturally alkaline. The results presented above indicate that the

frozen seawater surface a) likely maintains its slightly basic pH and b) is resistant to changes in

surface pH due to deposition of gas phase species. These results strongly suggest that the

buffering capacity at the sea-ice interface is maintained upon freezing. This observed buffering

at the frozen sea water surface is consistent with field studies which show that ikaite is the

dominant form of carbonate to precipitate from freezing seawater (24, 25) and also a modelling

study which shows that precipitation of ikaite does not lead to a reduction in total brine

alkalinity(11). Given that brine alkalinity may not be reduced upon freezing, the ‘low pH

107

requirement’ of bromine activation may be overstated.

Indeed, as mentioned briefly above, laboratory studies of bromide activation on frozen

surfaces (by oxidants HOBr, OH and O3) indicate that although the chemistry is accelerated at

low pH, it is still operative at neutral or even basic pH values. For example, Adams et al. (51)

studied the uptake and reaction of HOBr on frozen NaCl/NaBr and found that the product

yields were independent of pre-freezing pH between pH 4 – 10. Sjostedt and Abbatt (52)

studied the heterogeneous oxidation of frozen salt solutions by OH(g) and found that halogens

were released from neutral samples, although to a lesser extent than from acidic samples.

Finally, Oldridge and Abbatt (49) and Wren et al. (50) have also observed evidence for bromine

release due to the heterogeneous ozonation of frozen salt solutions at ambient pH. Thus, there is

a growing consensus that low pH may not be a strong criterion for halogen activation (3).

Gaining a better understanding of surface parameters such as pH will help to elucidate the

mechanism and the important substrate characteristics of bromine activation.

4.6 Conclusion

In summary, we studied pH changes at frozen freshwater, frozen salt water and frozen

seawater surfaces using a surface-sensitive spectroscopic approach. At the frozen salt water

surface, changes in surface pH due to the deposition of either HCl(g) or NH3(g) could be

rationalized on the basis of a brine layer at the air-ice interface which behaved like a true liquid

layer. Changes in pH at the frozen freshwater surface were distinctly different, indicating that

when a ‘true QLL’ exists, it may not be appropriate to describe it as a cold, liquid layer. In the

cryosphere, impurity levels and temperature (two factors which should affect the liquid-layer

fraction) can vary widely, so care should be taken in ascribing a universal surface pH. We also

show, for the first time experimentally, that the frozen seawater surface is buffered against pH

changes. This work has significant implications for understanding and interpreting pH-

dependent processes occurring on ice surfaces, such as “bromine explosion” chemistry.

108

4.7 References


1510 (2002)









Halogen activation via interactions with environmental ice and snow. Atmos. Chem. Phys. 12,

6237-6271(2012).

4. R. E. Honrath, Y. Lu, M. C. Peterson, J. E. Dibb, M. A. Arsenault, N. J. Cullen, K. Steffen, Vertical

fluxes of NOx, HONO, and HNO3 above the snowpack at Summit, Greenland. Atmos. Environ. 36,

2629-2640 (2002)

5. J. E. Dibb, M. Arsenault, M. C. Peterson, R. E. Honrath, Fast nitrogen oxide photochemistry in

Summit, Greenland snow. Atmos. Environ. 36, 2501-2511 (2002)

6. X. L. Zhou, H. J. Beine, R. E. Honrath, J. D. Fuentes, W. Simpson, P. B. Shepson, J. W. Bottenheim,

Snowpack photochemical production of HONO: a major source of OH in the Arctic boundary

layer in springtime. Geophys. Res. Lett. 28, 4087-4090 (2001)





4375-4418 (2007)

8. J. Bock, H. W. Jacobi, Development of a Mechanism for Nitrate Photochemistry in Snow. J. Phys.

Chem. A 114, 1790-1796 (2010)

9. S. Fickert, J. W. Adams, J. N. Crowley, Activation of Br2 and BrCl via uptake of HOBr onto


10. S. M. Fan, D. J. Jacob, Surface ozone depletion in arctic spring sustained by bromine reactions on

aerosols. Nature 359, 522-524 (1992)



(2008)



Interface. Environ. Sci. Technol. 44, 1302-1306 (2010)



14. T. F. Kahan, R. Zhao, D. J. Donaldson, Hydroxyl radical reactivity at the air-ice interface. Atmos.

Chem. Phys. 10, 843-854 (2010).


Phys. Chem. A 111, 11006-11012 (2007)


109

interfaces. Physical Review B 66, 085401 (13 pp) (2002)

17. Y. Li, G. A. Somorjai, Surface premelting of ice. J. Phys. Chem. C. 111, 9631-9637 (2007)


layer in ice/brine mixtures. J. Phys. Chem. B. 106, 11226-11232 (2002)




surface disordering on ice. Proc. Natl. Acad. Sci. U.S.A. 103, 9422-9427 (2006)


Atmos. Chem. Phys. 11, 9971-9982 (2011)






24. G. S. Dieckmann, G. Nehrke, C. Uhlig, J. Goettlicher, S. Gerland, M. A. Granskog, D. N. Thomas,

Brief Communication: Ikaite (CaCO3·6H2O) discovered in Arctic sea ice. Cryosphere 4, 227-230

(2010)

25. G. S. Dieckmann, G. Nehrke, S. Papadimitriou, J. Goettlicher, R. Steininger, H. Kennedy, D. Wolf-

Gladrow, D. N. Thomas, Calcium carbonate as ikaite crystals in Antarctic sea ice. Geophys. Res.

Lett. 35, (2008)

26. S. N. Wren, D. J. Donaldson, Laboratory Study of pH at the Air-Ice Interface. J. Phys. Chem. C 116,

10171-10180 (2012)

27. D. Clifford, D. J. Donaldson, Direct experimental evidence for a heterogeneous reaction of ozone

with bromide at the air-aqueous interface. J. Phys. Chem. A 111, 9809-9814 (2007)

28. D. Clifford, T. Bartels-Rausch, D. J. Donaldson, Suppression of aqueous surface hydrolysis by

monolayers of short chain organic amphiphiles. Phys. Chem. Chem. Phys. 9, 1362-1369 (2007)

29. B. T. Mmereki, D. J. Donaldson, Laser induced fluorescence of pyrene at an organic coated air-

water interface. Phys. Chem. Chem. Phys. 4, 4186-4191 (2002)

30. T. F. Kahan, N. O. A. Kwamena, D. J. Donaldson, Different photolysis kinetics at the surface of

frozen freshwater vs. frozen salt solutions. Atmos. Chem. Phys. 10, 10917-10922 (2010)





(2011)


surfaces. J. Phys. Chem. A. 111, 1277-1285 (2007)

34. A. Dias, A. P. Varela, M. D. Miguel, R. S. Becker, H. D. Burrows, A. L. Macanita, beta-carbolines

.2. Rate constants of proton transfer from multiexponential decays in the lowest singlet excited

state of harmine in water as a function of pH. J. Phys. Chem. 100, 17970-17977 (1996)

35. M. Balon, J. Hidalgo, P. Guardado, M. A. Munoz, C. Carmona, Acid-base and spectral properties

of beta-carbolines .1. Tetrahydro-beta-carbolines. J. Chem. Soc.-Perk. Trans. 2, 91-97 (1993)

36. S. Langer, R. S. Pemberton, B. J. FinlaysonPitts, Diffuse reflectance infrared studies of the reaction

of synthetic sea salt mixtures with NO2: A key role for hydrates in the kinetics and mechanism. J.

Phys. Chem. A 101, 1277-1286 (1997)

37. M. J. Molina, The Chemistry of the Atmosphere: The Impact of Global Change. (Blackwell Scientific

Publications, Boston, 1994).

110

38. S. L. Clegg, P. Brimblecombe, Solubility of ammonia in pure aqueous and multicomponent

solutions. J. Phys. Chem. 93, 7237-7248 (1989)

39. D. R. Hanson, K. Mauersberger, HCl/H2O solid-phase vapor-pressures and HCl solubility in ice.

J. Phys. Chem. 94, 4700-4705 (1990)

40. E. Thibert, F. Domine, Thermodynamics and kinetics of the solid solution of HCl in ice. J. Phys.

Chem. B 101, 3554-3565 (1997)

41. E. W. Wolff, R. Mulvaney, K. Oates, Diffusion and location of hydrochloric-acid in ice -

implications for polar stratospheric clouds and ozone depletion. Geophys. Res. Lett. 16, 487-490

(1989)

42. M. J. Molina, T. L. Tso, L. T. Molina, F. C. Y. Wang, Antarctic stratospheric chemistry of chlorine

nitrate, hydrogen-chloride and ice - release of active chlorine. Science 238, 1253-1257 (1987)

43. L. Delzeit, B. Rowland, J. P. Devlin, Infrared-spectra of HCl complexed ionized in amorphous

hydrates and at ice surfaces in the 15-90-K range. J. Phys. Chem. 97, 10312-10318 (1993)

44. H. Kang, T. H. Shin, S. C. Park, I. K. Kim, S. J. Han, Acidity of hydrogen chloride on ice. J. Am.

Chem. Soc. 122, 9842-9843 (2000)

45. S. C. Park, H. Kang, Adsorption, ionization, and migration of hydrogen chloride on ice films at

temperatures between 100 and 140 K. J. Phys. Chem. B 109, 5124-5132 (2005)

46. P. Parent, C. Laffon, Adsorption of HCl on the water ice surface studied by x-ray absorption

spectroscopy. Journal of Physical Chemistry B 109, 1547-1553 (2005)

47. L. Alvarez-Aviles, W. R. Simpson, T. A. Douglas, M. Sturm, D. Perovich, F. Domine, Frost flower

chemical composition during growth and its implications for aerosol production and bromine

activation. J. Geophys. Res. Atmos. 13, D21304 (10 pp) (2008)







reaction between O3(g) and halides at the surface of frozen salt solutions. J. Geophys. Res. Atmos.

115, D16309 (8 pp) (2010)





Res. Lett. 3, 045007 (7 pp) (2008)

111

C H AP T E R F I VE

SPECTROSCOPIC STUDIES OF THE HETEROGENEOUS REACTION BETWEEN O 3(g) AND HALIDES AT THE SURFACE OF FROZEN SALT

SOLUTIONS

S. N. Wren, T. F. Kahan, K. B. Jumaa and D. J. Donaldson

Reprinted with permission from S. N. Wren, T. F. Kahan, K. B. Jumaa and D. J. Donaldson,

Spectroscopic studies of the heterogeneous reaction between O3(g) and halides at the surface of

frozen salt solutions, Journal of Geophysical Research – Atmospheres, 115, D16309 (8 pp) (2010).

DOI: 10.1029/2010JD013929. Copyright (2010) American Geophysical Union.

Contributions: T. F. Kahan and K. B. Jumaa performed the harmine/LIF experiments. T. F.

Kahan helped with the interpretation of the data, provided critical review and helped to edit the

manuscript.

112

5.1. Introduction

Dramatic springtime depletion of tropospheric ozone has been associated with elevated

levels of bromine and bromine oxide (Br2 and BrO) in the polar boundary layer. At polar

sunrise, these compounds can react to form Br atoms which go on to destroy ozone in an

autocatalytic cycle known as the ‘bromine explosion’ (see recent review paper by Simpson et al.

(1) and references therein). The source of reactive bromine is bromide from sea salt; although

the concentration of bromide in sea water is small, its exclusion from bulk ice during freezing is

expected to result in a concentration enhancement at the ice surface which makes it accessible

for reaction with atmospheric oxidants. However, the details of the chemistry that converts

bromide ions into reactive bromine species are not well known.

One possible pathway for halide activation is heterogeneous reaction with gas-phase

ozone. The observation of Br2, BrCl, and BrO above the Arctic snowpack prior to polar sunrise

(i.e. in complete darkness) (2-4) indicates that bromine activation does not require sunlight, and

so implicates a role for O3 (5). In addition, ozone measurements made in the dark during the

ALERT 2000 campaign (6) are strongly suggestive of a surface sink for ozone, consistent with an

ozone reaction there. The dark production of Br2 from sea ice surfaces exposed to O3 has also

been observed in laboratory studies (5, 7).

Two mechanisms have been proposed for the production of Br2 from the reaction between

snowpack Br- and gas-phase O3. The first of these proceeds via the formation of BrO‾; in the

presence of protons, Br2 and OH‾ are the ultimate products (8). The second proceeds via charge

transfer to form O3‾; again the ultimate products are Br2 and OH‾(9). The first mechanism is

clearly accelerated in acidic environments. However, the effective pH of sea ice surfaces is not

well known (1, 10). Seawater is somewhat basic, and although the precipitation of carbonate

during the freezing process could result in reduced buffer capacity (1, 11), a mechanism that

doesn’t involve protons may also be required to understand the mechanism of bromine

activation.

113

Reactive iodine is also expected to have an impact on polar boundary-layer ozone

depletion despite its low natural concentrations in sea water (~(1-4) × 10-7 M) (12). However,

neither its importance, nor whether it is produced from sea ice iodide is well known. High

levels of iodine monoxide (IO), from ppt to ppb, have been measured at Neumayer Station (13,

14) and Halley, Antarctica (15). Modelling studies have shown that such IO concentrations lead

to increased O3 destruction rates (16). Iodine monoxide has been detected (15) in the absence of

direct sunlight, which suggests that a dark reaction with ozone may also be possible. Recently,

Sakamoto et al. (17) measured gas-phase I2(g) and IO(g) production due to the heterogeneous

reaction of gas-phase ozone with aqueous potassium iodide solutions. To our knowledge, there

have been no laboratory studies of the ozonation of frozen iodide solutions.

Due to the very small volume fraction represented by the disordered region at air-ice

interfaces, it is very difficult to directly monitor chemistry taking place there. Our group has

been developing glancing-angle laser-induced fluorescence (LIF) and Raman spectroscopic

probes for the study of these interfaces (18, 19). Although not strictly surface-specific, the

results presented in Kahan et al. (18) suggest that the glancing-angle Raman probe is sensitive to

the upper tens to about one hundred molecular layers. Comparison of the glancing-angle

Raman results presented in Wren and Donaldson (19) to surface tension measurements (20, 21)

suggests that these methods probe the same interfacial zone. In the following, discussion of the

‘interface’ or ‘interfacial region’ will refer to this surfacial region to which we are sensitive.

Recently, we have used these techniques to monitor the heterogeneous reactions of gas-

phase ozone with halides present at the surface of aqueous NaBr and NaI solutions (19, 22). The

LIF results (22) followed the formation of hydroxide anions in the interfacial region. In Wren

and Donaldson (19), a decrease in the intensity in the Raman signal of the OH-stretching mode

of water at the air-aqueous interface occurred in conjunction with the formation of trihalide

anions (Br3‾ and I3‾) at the interface. We speculated that the presence of these anions may

reduce the Raman intensity either by coating the surface (as observed when thin organic layers

coat the water surface) or by absorbing the Raman-scattered light generated at the interface, or

114

both (19). The two techniques gave the same dependence of rate on the bulk bromide

concentration, strongly suggestive that the same chemical process was responsible for the

production of hydroxide anions and halogen molecules at the surface.

In the following, we report kinetics of the surface reaction between gas-phase ozone and

frozen sodium bromide and frozen sodium iodide solutions measured with glancing-angle LIF

and glancing-angle Raman spectroscopy. The fluorescence experiments measured a pH increase

at the ice surface and the Raman experiments monitored a decrease in the intensity of the OH-

stretching vibration as described above. Similar to the aqueous results described above,

trihalide is formed as a product at the ice surface indicating molecular halogen formation there.

5.2. Experimental

5.2.1. Apparatus

The reaction chamber used here has been described previously (18). It consists of a Teflon

box sitting on a cooled copper base. Quartz windows at the front and back of the box allow a

laser beam to enter and exit the chamber. Side ports allow gases to be introduced and

ventilated. A thin stainless steel base plate was placed over the copper base for all experiments;

samples were positioned on this plate. Raman experiments were carried out at ca. -10˚C while

LIF experiments were carried out at -20˚C.

Fluorescence at the sample surface was induced by the unfocussed output of an Nd:YAG-

pumped optical parametric oscillator (OPO) while Raman scattering was induced by the

unfocussed, horizontally polarized, 355 nm output of a pulsed, frequency-tripled Nd:YAG laser.

In each case, the laser beam impinged the sample at an angle >85˚ from the surface normal.

Fluorescence and Raman scattering were collected using a 7 mm diameter liquid light guide

suspended ~6 mm above the sample surface. The collected light was imaged onto the entrance

slit of a grating monochromator; the transmitted intensity was detected by a photomultiplier

tube and sent to a digital oscilloscope which averaged the intensity vs. time signal over 64 laser

shots. In the case of fast reactions (between I‾ and O3), the oscilloscope averaged 16 rather than

115

64 laser shots. For Raman experiments, the collected light passed through a 355 nm laser line

long-pass filter before entering the monochromator. A portion of the signal centred on the peak

of the decay curve was averaged and saved for analysis. Fluorescence was not observed during

the Raman experiments.

5.2.2. Spectroscopic Measurements

Harmine (7-methoxy-1-methyl-9H-pyrido[3,4-b]indole) was used as a fluorescent pH

probe. At neutral pH, harmine has a broad fluorescence spectrum peaking around 410 nm (23).

At higher pH, harmine’s absorption spectrum shifts to the blue; for samples excited near 320

nm, this results in a decrease in the fluorescence intensity at 410 nm. Harmine has proved to be

a good probe for pH in high ionic strength environments since its fluorescence signal is not

strongly quenched there (22). Additionally, as shown in Chapter Four, the shape of harmine’s

absorption spectrum is not affected by the presence of salts.

In this study, excitation spectra were obtained by monitoring intensity at 420 nm and

manually scanning the excitation wavelength in 2 nm steps between 245 nm and 355 nm. Figure

5.1a shows excitation spectra of harmine at a pure air-ice interface before and after deposition of

gas-phase ammonia. Harmine fluorescence measured at the ice surface is much weaker than

harmine fluorescence measured at the liquid surface, resulting in a poorer S/N ratio for spectra

acquired on ice than on water (22). Inspection of the figure shows that the pH increase results in

a clear decrease in harmine fluorescence intensity following excitation at 322 nm. Therefore,

reaction kinetics were obtained by exciting the sample at 322 nm and monitoring emission at

420 nm.

Raman spectra at the frozen surfaces were obtained by manually scanning the

monochromator in steps to give ca. 30 cm-1 separation between data points. The reaction kinetics

were obtained by monitoring Raman intensity at 404 nm (corresponding to the OH-stretch band

of water at 3450 cm-1).

116

5.2.3. Ozone Generation

Ozone was formed in pure O2 via 185 nm photolysis using a commercial ozone generator.

For the LIF study, ozone concentrations were monitored by the attenuation of the 254 nm

output of a mercury lamp by ozone flowing through a 10 cm pathlength absorption cell prior to

entering the reaction chamber; the concentration was ~8 × 1015 molec cm-3 for each experiment.

For the Raman experiments, a single ozone concentration of ~5 × 1015 molec cm-3 was used for all

experiments.


Aqueous solutions were prepared by dissolving the respective salts in 18 MΩ deionised

water. For the LIF study, solutions also contained 10-6 M harmine. Ice samples were prepared by

freezing ~1-2 mL of solution on a thin piece of stainless steel shimstock, which rested on the

stainless steel base plate of the chamber floor. Absorption spectra of some melted samples were

obtained using a commercial UV-VIS spectrophotometer over the range 190-500 nm.

5.2.5. Chemicals

Sodium bromide (ACP Chemicals, 99%), sodium iodide (ACP Chemicals, Reagent Grade),

harmine (Aldrich, 98%), ammonium hydroxide (Fisher, reagent grade) and oxygen gas (Linde

Grade 4.5, 99.995%) were used as purchased without further purification.

5.2.6. Calibration of Raman Data

A comparison of the glancing-angle LIF (22) and Raman (19) studies of heterogeneous Br¯

ozonation kinetics at air-water interfaces shows the same dependence of the observed rate on

bulk [Br‾]. However, at the same bulk [Br‾], the absolute rate constant obtained from the pH

increase measured via glancing-angle LIF was about 50 times greater than that obtained from

the loss-rate of the OH-stretch intensity measured using glancing-angle Raman spectroscopy. A

scaling factor of 50 has been applied to all Raman results obtained in this study. We show below

that using this factor brings the Raman and LIF data into near agreement on the ice surface as

well.

117

5.3. Results

5.3.1. Frozen NaBr + Ozone

The excitation spectra of harmine at the surface of a frozen 10 mM NaBr solution acquired

before and after exposure to 8 × 1015 molec cm-3 gas-phase ozone is shown in Figure 5.1b. Just as

happens following exposure to ammonia, exposure to ozone clearly results in a decrease in

intensity at 320 nm indicating an increase in interfacial pH similar to that shown in Figure 5.1a.

Figure 5.2 shows the time dependence of harmine fluorescence intensity during exposure to

ozone. Since the fluorescence intensity is inversely proportional to pH, the linear fit to the decay

shown in the figure yields the pseudo-first order rate constant, kobs, for the reaction (22).

Figure 5.1. (a) Excitation spectra of harmine at an air- pure ice interface (blue circles) and of the

same sample after the deposition of gas-phase ammonia (red triangles); (b) Excitation spectra of

harmine at the surface of a frozen 10 mM NaBr solution (blue circles) and of the same sample

after exposure to 8 × 1015 molec cm-3 gas-phase ozone for ~20 min (red triangles). The excitation

spectra were all obtained by monitoring the emission intensity at 420 nm.

(b) (a)

118

Figure 5.2. Fluorescence intensity of harmine excited at 322 nm and monitored at 420 nm at an

air-ice interface during exposure to 8 × 1015 molec cm-3 gas-phase ozone. The arrow indicates the

time at which ozone was introduced to the reaction chamber. The dashed red line is the average

fluorescence intensity prior to exposure while the solid red line is a linear fit to the data during

ozone exposure.

Figure 5.3a shows Raman spectra of the OH-stretching feature at the surface of a frozen 50

mM NaBr solution acquired prior to and after ~30 min exposure to 5×1015 molec cm-3 gas-phase

ozone. The figure clearly illustrates that ozonation results in a decrease in the intensity of the

OH-Raman signal. Figure 5.3b shows the time dependence of the loss in OH-Raman signal

measured at 3450 cm-1 during ozonation of a frozen 100 mM bromide solution. From the linear

fit to the semi-log plot we extracted a pseudo-first order rate constant for the OH-Raman signal

loss. Figures 5.4a and 5.4b show the rate constants obtained using LIF (black circles) and Raman

(red triangles), plotted as a function of the initial bulk bromide concentration. The Raman rates

have been scaled by a factor of 50, as discussed above. From Figure 5.4a it can be seen that the

use of this scaling factor brings the Raman data into close agreement with the LIF data, thereby

suggesting that the scaling factor of 50 is appropriate for ice surfaces as well as liquid water

surfaces. The combination of the LIF and Raman data shown in Figure 5.4a is strongly

indicative of a saturation in the kinetics at higher bromide concentrations.

119

Figure 5.3. (a) Raman spectra of the surface of a frozen 50 mM NaBr solution prior to ozone

exposure (solid black trace) and after ~30 min exposure to 5 × 1015 molec cm-3 gas-phase ozone

(dashed red trace); (b) First-order decay of Raman OH-signal at 3450 cm-1 during the reaction of 5

× 1015 molec cm-3 gas-phase ozone with a frozen 100 mM NaBr solution. The solid line is a linear

fit to the semi-log data.

(b)

(a)

120

Figure 5.4. (a) Observed ozonation rate constants at the surface of frozen halide solutions plotted

as a function of initial halide concentration. (b) Comparison of ozonation rate constants at the

surface of frozen and aqueous bromide solutions. For both parts frozen bromide (LIF) is given by

black circles; frozen bromide (Raman, scaled): red triangles; frozen iodide (Raman, scaled): green

squares; aqueous bromide (calculated from Eq. 7 in Clifford and Donaldson (22)): solid blue line.

Error bars represent the standard deviation about the mean of at least three trials.

(a)

(b)

121

5.3.2. Frozen NaI + Ozone

Ozonation experiments were also performed on 5 – 100 mM frozen aqueous NaI

solutions. A first order decay of the OH-Raman signal intensity was also observed here; the

scaled rate constants are also plotted as a function of initial iodide concentration in Figure 5.4a.

Over the range of concentrations studied here, the kinetics are independent of the initial bulk

iodide concentration. The average rate constant in this range is (0.4 ± 0.1) s-1, approximately 100

times greater than the rate constants observed for bromide in the saturated region.

Figures 5.5a and 5.5b show photographs of frozen samples of 25 mM sodium iodide taken

after exposure to gas-phase ozone for ~5 min. The surface is not flat; rather it shows ridges and

peaks, indicative of non-uniform freezing. Figure 5.5a shows that exposure to ozone results in

the formation of a dark amber product at the surface. UV-VIS analysis of a sample melt revealed

the amber-coloured product to be triiodide anion (which has a reported λmax at 351 nm (24, 25)).

The portion of the sample indicated by the arrow in Figure 5.5b clearly indicates that the amber

product is most concentrated at the peak and along the ridge of the ice sample. At low initial

iodide concentrations (≤ 25 mM), exposure to ozone always resulted in localized amber colour

at the peaks and ridges. However, at higher initial iodide concentrations (50, 100 mM), a larger

fraction of the ice surface was amber in colour.

Figure 5.5. Photographs of frozen 25 mM sodium iodide samples after ~5 min exposure to gas-

phase ozone. (a) illustrates the amber-coloured reaction product (triiodide), while (b) clearly

shows the product concentrated at the peak formed at the sample surface (indicated by the

arrow) due to non-uniform freezing.

(a) (b)

122

No decrease in the intensity of the OH-stretching band was seen when pure water ice

was exposed to ozone, indicating that the observed losses in intensity are due to reaction, and

not, for instance, due to adsorption of ozone from the gas phase to the ice surface. Likewise, no

change in the fluorescence of harmine is observed upon exposure to ozone in the absence of

halide salts.

5.4. Discussion

Figure 5.1b shows that pH at the air-ice interface of frozen sodium bromide solutions

increases after exposure to gas-phase ozone, due to hydroxide anion formation. Under the same

conditions we also observed a decrease in the intensity of the OH-stretching band of water at

the surface of frozen sodium bromide and sodium iodide solutions. We attribute the loss in OH-

stretching intensity to the build-up of molecular bromine/tribromide and molecular

iodine/triiodide products at the surface, respectively.

For studies on frozen sodium bromide solutions, the good agreement between the LIF-

derived rates and the scaled Raman-derived rates suggests that a) the use of such a scaling

factor allows a semi-quantitative comparison between the two techniques; and b) the loss of

OH-Raman intensity and the increase in pH arise from the same chemical mechanism. These

results further suggest that the glancing-angle Raman probe being developed in this lab is a

useful tool with which to study heterogeneous processes occurring at atmospheric interfaces.

In (19) we measure tribromide and triiodide as the reaction products for the ozonation of

respective halide solutions at room temperature. In this study we measure triiodide as the

reaction product for the ozonation of frozen iodide solutions and infer tribromide as the

reaction product for the ozonation of frozen bromide solutions. The presence of trihalide anions

confirms the production of molecular halogen since in the presence of halides, molecular

halogen (X2) produced by the reaction will be in equilibrium with the trihalide anion, X3‾ (with

reported Keq(X=Br) = (16.1±0.3) M-1 and Keq(X=I) = (687±2) M-1 at 298K (26)). The production of Br2

and I2 under these conditions is significant because bromide activation is thought to occur in the

123

dark. As noted in the Introduction, one proposed mechanism for bromide activation by ozone is

greatly enhanced in acidic conditions (8). However, there has been some debate as to the acidity

of ice surfaces, and in particular sea ice surfaces (10, 27). The present results are suggestive of a

mechanism that does not require protons. Hunt et al. (9) have recently proposed a mechanism

for the interfacial reaction of gas-phase ozone with aqueous bromide that proceeds via the

formation of a surface complex between Br‾ and O3 and is not acid-assisted. Such a mechanism

may also be important at the sea ice surface.

At the surface of aqueous sodium bromide solutions, ozonation kinetics have been

observed (19, 22) to become concentration-independent at bulk bromide concentrations greater

than 1 M. However, as shown in Figure 5.4a, at the surface of frozen bromide solutions the

kinetics become independent of initial [Br‾] at much lower pre-freezing concentrations.

Whereas on the liquid surface, we interpret the saturation in the reaction kinetics as indicative

of a surface reaction, governed by adsorption of the halide ions to the interface, at a frozen

interface the relationship between the surface concentration of halide and its concentration in

the bulk prior to freezing is not so straightforward.

During freezing, halide ions are expected to be excluded from the ice matrix. This

exclusion process leads to the formation of a concentrated brine at the air-ice interface (10, 28-

31), a feature which is predicted from the equilibrium NaX-water phase diagrams. At

temperatures below the freezing point of water but above the eutectic temperature (NaBr-H2O

(-28.0 C) and NaI-H2O (-31.5C) (32)), there exists two phases: solid ice and a brine solution

whose concentration is set by the temperature alone. Some fraction of this solution will be

present at the top surface of the ice (where we are probing the kinetics); it will also be present in

liquid pockets within the ice matrix and other interfaces. The relative amount of the solid and

solution phases depends on the total salt concentration, with larger amounts of the brine being

present for higher pre-freezing salt amounts.

124

Thus the saturation in the kinetics we report in Figure 5.4a is likely a consequence of the

[Br‾] in the brine being independent of the initial solution [Br‾]. Interestingly, the value of kobs

obtained here for the NaBr case in the “saturated kinetics” region is only ~ 2 different from that

seen on the water surface at a bulk concentration of 2.5 M, corresponding to the expected

equilibrium brine concentration at -10 oC.

The fact that we observed darker amber product at the peaks of frozen iodide solutions

(as displayed in Figure 5.5) is consistent with halide exclusion, since these regions were most

likely to have been the last to freeze, and hence are expected to contain the highest iodide

concentrations.

In Figure 5.4b, we compare the low-bromide kinetic results measured at water and ice

surfaces. It is clear that at bulk (pre-freezing) bromide concentrations less than 10 mM, the

kinetics are much more rapid at the frozen vs. the liquid surface. For example, the reaction on a

frozen 1 mM bromide solution, which corresponds to the bromide concentration in sea water, is

~60 times faster than at the surface of an aqueous solution of the same concentration (as

calculated from the results of Clifford and Donaldson (22). At the lowest iodide concentrations

studied (2.5 - 10 mM) we also observed faster rates at ice surfaces compared to aqueous surfaces

(19). The higher reaction rates on ice vs. liquid water at low bulk halide concentrations are

probably a consequence of higher concentrations of halide ions at the frozen surface, due to

their exclusion from the ice matrix during the freezing process as discussed above.

Following Poschl et al. (33), we may transform the pseudo-first order rate constants to

apparent reactive uptake coefficients, γ, for the reaction at the frozen surface. The uptake

coefficient is defined as the fraction of collisions between gas-phase ozone and halide anions at

the surface that give rise to reaction. This was calculated using the LIF-derived rate constants

for bromide and the scaled Raman-derived rate constants for iodide. The uptake coefficient was

estimated using the relationship:

125

)]([

4

33gO

k

OX

obs

(1)

where is the collisional cross-section of the halide and is the average molecular speed

of ozone. Crystal ionic radii (34) were used to calculate the areas of the spherical halide ions,

which were then used as . For uptake to frozen sodium iodide solutions a linear

extrapolation from the origin to the lowest iodide concentration studied provides a lower limit

for the rate constant in the concentration-dependent region.

The uptake coefficient at a bromide concentration of 10 mM was calculated to be (4 ±2) ×

10-8. This value is in good agreement with the γ (~1 × 10-8 ) obtained in a flow-tube study on

frozen NaBr solutions (7) with the same bulk NaBr concentration and at the same temperature.

At sea water concentrations of 10-3 M, we calculate γ to be (1.3 ± 0.5) × 10-8. The uptake

coefficient at a seawater iodide concentration of 4 × 10-7 M (35) was similarly calculated to be

approximately (1.6 ± 0.5) × 10-9.

The dependence of on the bromide concentration can give some insight into the

conditions of the reaction. For reactive uptake into bulk media, γ depends on the square root of

the dissolved reagent concentration, whereas for reaction on the surface the dependence reflects

the relationship between the initial bulk concentration and that present at the air-ice interface.

Figure 5.6 shows a log-log plot of the uptake coefficient versus bulk initial [Br‾] for the bromide

data shown in Figure 5.4. The slope of ~0.3 to the linear fit may indicate a bulk-like reaction

taking place in a brine layer at the ice surface. However, given that at phase equilibrium the

concentration of Br‾ in the brine is established by the temperature alone and not its initial bulk

concentration (vide supra), this is probably coincidental.

126

Figure 5.6. Calculated uptake coefficients, γ, for ozone (8 × 1015 molec cm-3) reacting at the surface

of frozen Br- solutions as a function of initial [Br-], plotted on a log-log scale. The solid line is a

linear fit to the log-log data. The error bars represent the uncertainty in the rate constants shown

in Figure 5.4.

Our measurements were carried out at very high ozone concentrations. The study by

Oldridge (7) measured Br2 production from frozen 8.4 mM bromide solutions at pH 2 and -20˚C

as a function of ozone concentration over the range 10 – 240 × 1012 molec cm-3. In that study, γ

was observed to decrease as a function of ozone concentration up to [O3] ~60 × 1012 molec cm-3

and then be independent of [O3]. From Figure 3-10 of that work (7) it is clear that the uptake

coefficient at an environmentally relevant ozone concentration of 30 ppb (2) is about an order of

magnitude greater than the uptake coefficient at high ozone concentrations where γ is [O3]-

independent. Applying this factor to the uptake coefficients estimated above for relevant sea-ice

halide concentrations gives estimates of γ for frozen bromide and iodide solutions of ~1 × 10-7

and ~2 × 10-8 respectively.

127

5.5. Conclusions

We have measured ozonation kinetics of halides at frozen surfaces using two surface-

sensitive techniques. Our results suggest that halide ions that have been excluded to the ice

surface react heterogeneously with ozone to produce molecular halogen. Although the

production of Br2 from sea ice is not unexpected, our results are significant since they were

obtained under dark, non-acidic conditions. This shows that the formation of ‘seed’ Br2 can

occur in the absence of excess protons. This may be important, given the uncertainty of the pH

of sea ice surfaces. The observed production of I2 suggests that a surface reaction between gas-

phase ozone and frozen iodide could be responsible for the presence of IO in the polar

boundary layer. Moreover, we have directly probed the air-ice interfacial region, giving

concrete support to the notion that it is here that the key chemistry occurs. The good agreement

between the LIF- and Raman-derived rate constants supports our earlier work (19) and further

suggests that monitoring the Raman intensity of the OH-stretching band of water provides

meaningful kinetic data.

5.6. References





4375-4418 (2007)

2. K. L. Foster, R. A. Plastridge, J. W. Bottenheim, P. B. Shepson, B. J. Finlayson-Pitts, C. W. Spicer,

The role of Br2 and BrCl in surface ozone destruction at polar sunrise. Science 291, 471-474 (2001)

3. C. W. Spicer, R. A. Plastridge, K. L. Foster, B. J. Finlayson-Pitts, J. W. Bottenheim, A. M. Grannas,

P. B. Shepson, Molecular halogens before and during ozone depletion events in the Arctic at

polar sunrise: concentrations and sources. Atmos. Environ. 36, 2721-2731 (2002)

4. T. Wagner, C. Leue, M. Wenig, K. Pfeilsticker, U. Platt, Spatial and temporal distribution of

enhanced boundary layer BrO concentrations measured by the GOME instrument aboard ERS-2.

J. Geophys. Res. Atmos. 106, 24225-24235 (2001)



6. J. W. Bottenheim, J. D. Fuentes, D. W. Tarasick, K. G. Anlauf, Ozone in the Arctic lower

troposphere during winter and spring 2000 (ALERT2000). Atmos. Environ. 36, 2535-2544 (2002)

7. T. N. W. Oldridge, Halogen Activation from Sea Ice: Nitrate Photolysis and

Heterogeneous Reaction with Ozone, Master's Thesis, University of Toronto, Toronto (2009)

8. Q. Liu, L. M. Schurter, C. E. Muller, S. Aloisio, J. S. Francisco, D. W. Margerum, Kinetics and

128

mechanisms of aqueous ozone reactions with bromide, sulfite, hydrogen sulfite, iodide, and

nitrite ions. Inorg. Chem. 40, 4436-4442 (2001)

9. S. W. Hunt, M. Roeselova, W. Wang, L. M. Wingen, E. M. Knipping, D. J. Tobias, D. Dabdub, B. J.

Finlayson-Pitts, Formation of molecular bromine from the reaction of ozone with deliquesced

NaBr aerosol: Evidence for interface chemistry. J. Phys. Chem. A 108, 11559-11572 (2004)



(2008)



12. I. B. Oh, D. W. Byun, H. C. Kim, S. Kim, B. Cameron, Modeling the effect of iodide distribution on

ozone deposition to seawater surface. Atmos. Environ. 42, 4453-4466 (2008)

13. U. Friess, T. Wagner, I. Pundt, K. Pfeilsticker, U. Platt, Spectroscopic measurements of

tropospheric iodine oxide at Neumayer Station, Antarctica. Geophys. Res. Lett. 28, 1941-1944 (2001)

14. U. Friess, T. Deutschmann, B. Gilfedder, R. Weller, U. Platt, Iodine Monoxide in the Antarctic

snowpack. Atmos. Chem. Phys. Discuss. 9, 25361-25407 (2009)

15. A. Saiz-Lopez, A. S. Mahajan, R. A. Salmon, S. J. B. Bauguitte, A. E. Jones, H. K. Roscoe, J. M. C.

Plane, Boundary layer halogens in coastal Antarctica. Science 317, 348-351 (2007)

16. R. Sander, R. Vogt, G. W. Harris, P. J. Crutzen, Modeling the chemistry ozone, halogen

compounds, and hydrocarbons in the arctic troposphere during spring. Tellus Ser. B-Chem. Phys.

Meteorol. 49, 522-532 (1997)

17. Y. Sakamoto, A. Yabushita, M. Kawasaki, S. Enami, Direct Emission of 12 Molecule and IO

Radical from the Heterogeneous Reactions of Gaseous Ozone with Aqueous Potassium Iodide

Solution. J. Phys. Chem. A 113, 7707-7713 (2009)


Phys. Chem. A 111, 11006-11012 (2007)

19. S. N. Wren, D. J. Donaldson, Glancing-Angle Raman Spectroscopic Probe for Reaction Kinetics at

Water Surfaces. Phys. Chem. Chem. Phys. 12, 2648-2654 (2010)

20. H. C. Allen, D. E. Gragson, G. L. Richmond, Molecular structure and adsorption of dimethyl

sulfoxide at the surface of aqueous solutions. J. Phys. Chem. B 103, 660-666 (1999)

21. T. L. Tarbuck, G. L. Richmond, Adsorption of organosulfur species at aqueous surfaces:

Molecular bonding and orientation. J. Phys. Chem. B 109, 20868-20877 (2005)

22. D. Clifford, D. J. Donaldson, Direct experimental evidence for a heterogeneous reaction of ozone

with bromide at the air-aqueous interface. J. Phys. Chem. A 111, 9809-9814 (2007)

23. A. Dias, A. P. Varela, M. D. Miguel, R. S. Becker, H. D. Burrows, A. L. Macanita, Beta-carbolines.

2. Rate constants of proton transfer from multiexponential decays in the lowest singlet excited

state of harmine in water as a function of pH. J.Phys. Chem. 100, 17970-17977 (1996)

24. A. D. Awtrey, R. E. Connick, The Absorption Spectra of I2, I3-, I-, IO3-, S4O62- and S2O32-. Heat of the

Reaction I3- = I2 + I-. J. Am. Chem. Soc. 73, 1842-1843 (1951)

25. D. A. Palmer, R. W. Ramette, R. E. Mesmer, Triiodide Ion Formation Equilibrium and Activity-

Coefficients in Aqueous-Solution. J. Sol. Chem. 13, 673-683 (1984)

26. T. X. Wang, M. D. Kelley, J. N. Cooper, R. C. Beckwith, D. W. Margerum, Equilibrium, Kinetic,

and UV-Spectral Characteristics of Aqueous Bromine Chloride, Bromine, and Chlorine Species.

Inorg. Chem. 33, 5872-5878 (1994)


electrolyte solutions. J. Phys. Chem. B 110, 7613-7616 (2006)



129


layer in ice/brine mixtures. J.Phys. Chem. B 106, 11226-11232 (2002)


99-103 (2007)

31. H. Eicken, Salinity Profiles of Antarctic Sea Ice - Field Data and Model Results. J. Geophys. Res.

Oceans 97, 15545-15557 (1992)

32. P. Deluca, L. Lachman, Lyophilization of Pharmaceuticals .4. Determination of Eutectic

Temperatures of Inorganic Salts. J. Pharm. Sci. 54, 1411-1415 (1965)

33. U. Poschl, T. Letzel, C. Schauer, R. Niessner, Interaction of ozone and water vapor with spark

discharge soot aerosol particles coated with benzo[a]pyrene: O-3 and H2O adsorption,

benzo[a]pyrene degradation, and atmospheric implications. J. Phys. Chem. A 105, 4029-4041 (2001)

34. R. C. Weast, Ed., CRC Handbook of Chemistry and Physics, (Chemical Rubber Publishing Company,

1983).

35. W. N. Chang, B. G. Heikes, M. H. Lee, Ozone deposition to the sea surface: chemical

enhancement and wind speed dependence. Atmos. Environ. 38, 1053-1059 (2004)

131

C H AP T E R S I X

PHOTOCHEMICAL CHLORINE AND BROMINE ACTIVATION FROM

ARTIFICIAL SALINE SNOW

Sumi N. Wren, D. J. Donaldson, J. P. D. Abbatt

Reprinted with permission from S. N. Wren, D. J. Donaldson, J. P. D. Abbatt, Photochemical

chlorine and bromine activation from artificial saline snow, Atmospheric Chemistry and Physics

Discussions, 13, 14163 – 14193 (2013). DOI:10.5194/acpd-13-14163-2013.


paper. D. J. Donaldson and J. P. D. Abbatt provided critical review and guidance.

132

6.1. Introduction

In the polar regions, heterogeneous reactions occurring on various frozen sea ice surfaces

are thought to be responsible for the activation and release of reactive halogen species

(particularly Br2, BrO and HOBr) ((1) and references therein). At polar sunrise, reactive bromine

species readily dissociate to yield Br radicals which can catalytically destroy ozone or

participate in other atmospheric oxidation processes, thereby influencing the oxidative capacity

of the polar boundary layer. Although a strong role for bromine in both springtime ozone

depletion (1) and mercury oxidation (2) has now been well established, many of the mechanistic

details of bromine activation are still lacking, precluding a good understanding of the

environmental conditions and sea ice/snow substrates that lead to such events (3).

Moreover, the role that chlorine plays in these processes is less well known – in part due

to a scarcity of reliable field measurements of photolyzable chlorine species (4). Due to the high

reactivity of Cl radicals, elevated chlorine levels are also expected to perturb the chemical

composition and oxidative capacity of the troposphere, with their greatest impact likely being

on the oxidation of volatile organic compounds (1).

Laboratory studies have shown that oxidation reactions involving O3 (5-7), OH (8) and

HOBr (9-11) will liberate Br2 (and in some cases BrCl) but not Cl2 from frozen halide solutions.

Indeed, the dark ozonation of frozen Br¯ solutions is thought to play an important role in

forming ‘seed’ Br2. The proposed mechanism is based on known aqueous phase chemistry (see

(12) and references therein):

O3(g) + Br¯ OBr¯ (R1)

OBr¯ + H+ HOBr (R2)

HOBr + Br¯ + H+ Br2 + H2O (R3)

When the Br¯/Cl¯ is low, HOBr will also oxidize chloride to yield BrCl (10, 13):

HOBr + Cl¯ + H+ BrCl + H2O (R4)

133

BrCl formed via R4 has a lower volatility than Br2 and may undergo further reactions in the

condensed phase to yield Br2(g) (12). This, together with the faster rate constant for R3 relative

to R4 (12), leads to the preferential release of Br2(g). The analogous dark ozonation of Cl¯ (R5) is

too slow to be considered an important Cl2 source (12).

O3(g) + Cl¯ OCl¯ (R5)

Recent field measurements which have revealed higher-than-expected concentrations of

molecular chlorine in the polar boundary layer (up to 400 pptv) (14-16) have therefore left

researchers searching for a missing chlorine source. Field measurements by Impey et al. (14)

during the Polar Sunrise Experiment show that Cl2 follows a diurnal cycle with a daytime

maximum, indicating a photochemical source. Recent measurements of Cl2 at Barrow, AK

suggest that Cl2 is correlated with both ozone and actinic flux (16). These observations

motivated the current study, which seeks to investigate whether a reaction involving both light

and ozone is responsible for activating chlorine from frozen sea ice/snow surfaces.

The majority of laboratory studies of halogen activation from frozen NaCl/NaBr solutions

have been performed using low surface area ice samples; studies performed using high surface

area snow samples are completely lacking. Furthermore, motivated by the importance of

stratospheric halogen activation, much of the research has been carried out at very low

temperatures. There is a need for more studies to be performed at the milder temperatures that

are commonly found in the Arctic spring. Not only are such temperatures more relevant for the

polar boundary layer, they represent a regime within which sea ice substrates will remain

partially liquid and hence the chemistry may be quite different (17). Finally, few laboratory

studies have been performed in the presence of light, which has the potential to drive HOx (=

OH + HO2) chemistry and halogen radical reactions within the snowpack. Very recently, a field

study using natural snow and ice substrates demonstrated the importance of light-driven

snowpack chemistry to halogen activation (18). The specific goal of our study was to determine

whether a photochemical pathway exists for halogen activation from artificial snow in the

134

presence of gas phase O3 and radiation at actinic wavelengths. Significant to this study is the

choice of high surface area samples and temperatures spanning the NaCl-water eutectic of 252

K. Unlike many previous studies, we chose to focus particularly on chlorine activation by using

snow with a low bromide content. In controlled laboratory experiments, we studied the impact

of environmental factors such as temperature, acidity, salt content, Br¯/Cl¯ ratio, snow surface

area and [O3] on the observed dihalogen release.


6.2.1. Experimental Apparatus

The overall experimental setup is shown in Figure 6.1 and the experimental parameters

and ranges are given in Table 6.1. Compressed air (Linde, Air Grade 0.1) served as the carrier

gas for these experiments. The relative humidity (RH) of the carrier gas was controlled by

diverting the bulk of the flow (190 sccm) through a ~45 cm long, double-jacketed Pyrex glass

flow tube that was partially filled with crushed ice. Air passing through the RH conditioning

flow tube was saturated with the water vapour pressure over ice at the experimental

temperature, thereby minimizing the potential for evaporation or condensation of water from or

to the snow sample. Ozone (O3) was generated by passing the remaining flow of dry

compressed air (20 sccm) by a Hg Pen-Ray lamp. The ozone-rich air then passed through a 10

cm path-length quartz cell where the ozone concentration was determined from the attenuated

output of a Hg Pen-Ray lamp (λmax = 254 nm). The ozone-rich air re-connected with the RH-rich

air downstream of the RH conditioning flow tube.

135

Figure 6.1. Schematic of the experimental apparatus.

Table 6.1. Parameters used in the experiments. The BASE scenario conditions are shown in bold

text. Parameter Range

Illumination 310 nm long-pass filter, 360 nm long-pass filter (w/ Xe lamp)

[Ozone] 2.7 × 1013

– 1.3 × 1015

molecules cm-3

Snowpack T (248, 252, 254, 257, 263) ±1 K

Pre-freezing [NaCl] 0.1, 0.5, 1.0 M

Br¯ in NaCl(s) <0.01 wt% (reagent grade), <0.001 wt% (high purity)

Pre-freezing pH 2.3, 4.3, ambient (~5.9)

Size fraction Fine, Coarse

136

The snow reaction chamber consists of a double-jacketed Pyrex glass vessel. Artificial

snow was ‘poured’ into the opening at the top of the reaction chamber to create a cylindrical

snow sample ~ 7 cm tall and ~ 4 cm in diameter. Gases enter at the base of the snow sample

through a 6.35 mm diameter inlet located on the side of the reaction chamber and are vented

through a side-arm located just above the top of the snow sample. The reaction chamber was

held at atmospheric pressure and a critical orifice controls the flow of gases into the ion-

molecule reaction region of the chemical ionization mass spectrometer (CIMS).

The samples were illuminated through a window at the top of the reaction chamber by

the output of a 1 kW Xenon arc-lamp. A ~ 20 cm path length water filter was placed between the

arc-lamp and the sample to remove IR radiation (to prevent over-heating of the sample) and a

mirror directed the light onto the sample. In the majority of experiments a 310 nm long-pass

filter (10% transmission at 310 nm, measured using a commercial UV-VIS spectrometer) was

also placed over the quartz glass window. In some experiments, the 310 nm filter was replaced

by a 360 nm long-pass filter (10% transmission at 360 nm). Additional experiments were

performed in which Br2 or Cl2 were flowed from a glass bulb manifold through the empty

reaction chamber. Illumination did not lead to a measurable decrease in the Br2 or Cl2 signals in

the absence of ozone for a) the full spectral output of the Xe-lamp with water filter and b) the

same but with a 310 nm or 360 nm long-pass filter.

The temperature of the reaction chamber and the RH conditioning flow tube was controlled

by a Neslab chiller. The temperature in the reaction chamber was occasionally checked using a

thermocouple and was typically ~5 K warmer than the chiller set point temperature.

137

6.2.2. Chemical Ionization Mass Spectrometer

Gases were detected using a home-built chemical ionization mass spectrometer (CIMS). The

instrument, which is described in greater detail in Thornberry and Abbatt (19), consists of two

differentially pumped regions: the ion molecule region (IMR) and the multiplier chamber (MC).

During operation, the pressures are 2.5 Torr and 2.5 × 10-6 Torr in the IMR and the MC

respectively. The CIMS control box (Merlin QMS Controller) was interfaced to a PC operating

using Merlin software from EXTREL (Abb Inc.).

A trace flow of SF6 (BOC, Grade 3.0), seeded in a 2 slpm dry N2 flow (Linde, N2 liquid

Grade 4.8) is passed over a 210Po ion source to produce the reagent ion, SF6¯, via electron

attachment. A -112 V applied to the sheath surrounding the ion source helps direct the reagent

ions into the IMR. Gas phase species produced by the experiment (O3, Cl2, BrCl and Br2) are

swept into the IMR by the carrier gas flow; charge-transfer reactions with the reagent gas

occurred in the IMR:

(R6)

Thus O3, Cl2, BrCl and Br2 are detected at m/z 48, 70, 114 and 160 respectively. A scan time

of 2 s was used. Ions enter the mass spectrometer through a pinhole which is biased to -14 V.

SF6¯ was chosen as the reagent ion because it can be used to simultaneously measure both

O3 and dihalogens. The reaction of SF6¯ with O3 is very fast, near the collision rate (20). The

reactions of SF6¯ with Br2 and BrCl are also fast, but the reaction with Cl2 is slower (20).

Although the reaction between SF6¯ and water vapour is fairly slow, high mixing ratios of water

vapour often preclude the use of SF6¯ as a reagent ion. However, at the cold temperatures used

in these experiments and with the pressure drop to 2.5 Torr in the ion-molecule region, the

water vapour pressure is low enough that the reagent ion signal remains sufficiently high.

138

6.2.3. Calibration

Calibrations for Cl2 and Br2 were performed routinely. Calibrations were performed in the

absence of ozone with the reaction chamber bypassed. A small flow of either Cl2 or Br2 was

released from a fixed-volume manifold to the main carrier gas flow (downstream of the RH

conditioning flow tube as indicated in Figure 6.1) and the pressure drop in the manifold as a

function of time was converted to the equivalent concentration of either Br2 or Cl2 in molecules

cm-3.

The Cl2 calibration bulb was prepared by delivering pure Cl2(g) (Matheson) to an

evacuated glass bulb. The Cl2 in the bulb was then diluted with dry N2 to a final mixing ratio on

the order of 10-4. The Br2 calibration bulb was prepared by flash-freezing a small volume of

liquid Br2 in a round-bottom flask, pumping off the remaining vapour, then delivering Br2(g)

from the round-bottom flask into an evacuated glass bulb after warming the sample. The Br2 in

the bulb was then diluted with dry N2 to a final mixing ratio on the order of 10-5.

The intensity of the reagent ion (SF6¯) varied somewhat from day-to-day. All signals were

normalized to the SF6¯ signal. During Cl2 calibration, peaks at m/z 70, 72 and 74 corresponding

to 35Cl35Cl¯, 35Cl37Cl¯ and 37Cl37Cl¯ respectively were measured with the expected isotopic

distribution. Similarly, during Br2 calibration, peaks at m/z 158, 160 and 162 corresponding to

79Br79Br¯, 79Br81Br¯ and 81Br81Br¯ respectively were measured with the expected distribution.

Calibrations were not performed for BrCl, but we assume that the CIMS is equally sensitive to

BrCl as it is to Br2 (21). This assumption is based on their similar gas-phase electron affinities

and their fast, collision-rate limited reaction with SF6 (20).

The gas release during calibration led to a CIMS signal in counts per second (cps) from

which we determined sensitivity. Typically, the SF6¯ signal was between (6 – 8) × 104 counts. In

all cases we normalized the measured ion signals to the SF6¯ reagent ion signal which resulted

in a unit-less intensity. For Br2 and BrCl the sensitivity was 1.7 × 10-16 per (molecule cm-3) and for

Cl2 the sensitivity was 5.8 × 10-17 per (molecule cm-3). Using the calibrated sensitivities, we

139

determined the concentration of dihalogens being released from the snow (molecules cm3).

The limits of detection were determined from the signal-to-noise (S/N) ratio of the

background signal during calibration. The noise was taken as the standard deviation of the

background signal, measured for one minute (30 scans). For our operating conditions, our

detection limits were (2 – 3) × 1011 molecules cm-3 for Br2, (1 – 2) × 1011 molecules cm-3 for BrCl

and (3 – 5) × 1011 molecules cm-3 for Cl2.

6.2.4. Snow Preparation

Artificial snow was prepared from a saline solution of NaCl(s) in 18 mΩ deionized water.

The chloride to bromide ratio (Cl¯: Br¯) was varied by using different purities of NaCl(s); here

we used reagent grade NaCl(s) (ACP Chemicals, 0.01% bromide impurity by weight) and high

purity NaCl(s) (Fluka TraceSELECT®, <0.001% bromide impurity by weight). The pre-freezing

[NaCl(aq)] was 0.5 M for the majority of these experiments (and was varied between 0.1 and 1.0

M for select experiments). Initial solution pH was measured using a commercial pH electrode

and adjusted using dilute H2SO4(aq) as needed. The pre-freezing pH was ~2.3 for the majority of

experiments. Unadjusted samples had an ambient pH ~ 5.9. To prepare the snow, the nozzle of

a mister bottle was used to disperse the saline solution into fine droplets which were directed

into a dewar of liquid N2. The contents of the dewar were then separated using a mesh sieve

(hole diameter roughly 1 mm). The ice particles (i.e., the snow grains) captured by the sieve

were mostly non-spherical (probably due to coagulation of individual droplets) and were

estimated to be ~0.5 mm in diameter. The ice particles that fell through the sieve were very fine

and uniform to the eye (exhibiting a strong resemblance to icing sugar), with an estimated

diameter < 0.5 mm (see Figure 6.2, left panel). This fine mode fraction was used for the majority

of the experiments. All prepared snow was stored in amber glass jars in a 243 K freezer for at

least 1 day before use to allow the ice particles time to anneal. Coarse mode ice particles were

prepared for some experiments by dropping saline solution directly into the liquid N2 using a

glass pipette. The resultant ice spheres were very roughly spherical and had a relatively wide

140

size distribution (see Figure 6.2, right panel). The mass of snow used was roughly ~ 40 g for the

fine mode samples and ~ 50 g for the coarse mode samples. The snow samples filled a volume

in the reaction chamber of ~ 80 cm2.

Figure 6.2. Photographs of the artificial snow samples taken immediately after preparation.

6.2.5. Snow Characterization

To ensure that the artificial snow samples retained the same bulk composition as their

initial solution, some of the snow samples were melted and their ionic content was analyzed by

ion-chromatography (IC). This was done separately for the coarse and fine fraction to ensure

that the method of snow preparation and separation did not result in chemical fractionation.

The analysis showed that the chloride concentrations remained unchanged due to freezing (for

both the coarse and fine mode fractions).

1 cm

141


6.3.1. Photochemical Chlorine Production

Figure 6.3a shows results from a typical experiment under conditions that define the

‘BASE’ scenario (which are shown in bold text in Table 6.1). Figures 6.3b – 6.3f demonstrate the

effect of varying one parameter at a time from the BASE scenario. All the figures have the same

vertical and horizontal scaling for ease of comparison and in all cases Cl2 is shown as black

circles, Br2 as red triangles, and BrCl as green squares. To aid the interpretation of our results,

we occasionally refer to the contribution of a given dihalogen to the total dihalogen sum [Cl2 +

Br2 + BrCl], henceforward the ‘Cl2 fraction’, ‘Br2 fraction’ and ‘BrCl fraction’.

We observe that the dark exposure to ozone liberates Br2 and in some cases BrCl from the

snow (Figure 6.3). Prior to illumination, [Br2] is greater than [BrCl], although the contribution of

[Br2] to the total halogen release ([Br2] + [BrCl] + [Cl2]) decreases slightly with time due to the

slow depletion of snow bromide. As mentioned, the dark ozonation of frozen NaCl/NaBr

solutions has been the subject of previous laboratory studies (5-7). These studies (6) suggest that

the reaction proceeds in a liquid brine layer via a mechanism (R1-R4) that is similar to that

occurring on/in the aqueous phase. The dark production of Br2 and BrCl that we observe here is

consistent with those studies.

Under all conditions studied here (i.e., shown in Figures 6.3a-6.3f), illumination of the

snow sample (at t = 0) leads to a rapid increase in the concentration of all three dihalogens with

the rate of increase most rapid for Br2 and slower for BrCl followed by Cl2. No halogen

production was observed when snow samples were illuminated in the absence of ozone. Cl2

production (above the limits of detection) was only observed when both light and ozone were

present.

142

Figure 6.3. Time evolution of the dihalogen concentrations during a typical experiment: Br2 (red

triangles), Cl2 (black circles) and BrCl (green squares). Each point represents the average of a 2.5

minute time bin (77 data points). The dashed line indicates the time at which the ozone generator

was turned on and the solid line (t = 0) indicates the time at which the samples were illuminated.

Panels: a) BASE scenario conditions as in Table 6.1; BASE scenario conditions except with b) high

purity NaCl(s) (Br¯ < 0.001 wt%) to prepare solution; c) 360 nm long-pass filter; d) coarse mode

snow, scaled up by a factor of 5; e) pre-freezing pH = 4.3; f) snowpack T = 252 ±1 K, scaled up by a

factor of 5.

143

The focus of this study is the accelerated halogen activation (particularly the Cl2

activation) we observe when the halide-doped snow is exposed to ozone and light. In the

following, we show that our observations support a snowpack ‘halogen explosion’ - a chemical

mechanism that is initiated by the photolysis of Br2 or BrCl (formed via the dark ozonolysis of

bromide) and perpetuated by the recycling of HOBr or HOCl back into the snowpack.

X2 + hν X + X (R7)

X + O3 XO + O2 (R8)

XO + HO2 HOX + O2 (R9)

HOX + hν OH + X (R10)

The mechanism shown above (R7 – R10) has previously been proposed (1, 3, 13, 22-25) to

be important for activating bromine in the polar regions. Although laboratory studies have

shown that HOBr will react with chloride/bromide ice surfaces to give Br2 or BrCl under certain

conditions (9-11), the entire ‘halogen explosion’ has not been directly investigated in controlled

laboratory experiments using frozen substrates. Nor has it been pursued as an important

mechanism for activating chlorine in the form of Cl2.

HOBr formed via R9 can oxidize Br¯ in the condensed phase leading to Br2 and BrCl

through R3 and subsequent reactions. The HOBr can also oxidize Cl¯ directly to give BrCl at the

surface which can oxidize Br¯ to Br2 (10). Cl radicals formed via the photolysis of BrCl (or

eventually Cl2) provide a route for HOCl formation and the HOCl can oxidize Cl¯ to give Cl2

directly. In our experiments, we believe that [Cl2] grows more slowly because oxidation of Br¯

will always be favoured over Cl¯ when it is available and because Cl2 formation requires BrCl

formation (to provide Cl radicals via photolysis). Although molecular Cl2 formation always

occurs subsequent to BrCl formation, its observation contemporaneously to Br2 shows that Br¯

does not need to be exhausted for Cl2 to form. The [Br2] and [BrCl] decrease rapidly once they

have reached their respective maxima; this is presumably due to the fact that the snow is

prepared from NaCl(s) containing only a small bromide impurity (< 0.01% Br¯ by weight).

Hence there is a small and finite amount of Br¯ that can be liberated from the snow as either Br2

or BrCl.

144

6.3.2. Direct Observation of a Snowpack Halogen Explosion

A direct role for reactive halogen species in the photochemical activation of halogens from

the snow was explored in test experiments in which additional Br2 gas was introduced to the

reaction chamber via a fixed-volume glass bulb manifold coupled to the main gas flow. Figure

6.4a shows [Cl2], [Br2] and [BrCl] and Figure 6.4b shows [O3] (expressed as a normalized signal)

during such an experiment. The sample had been previously subjected to a typical experiment

so that the majority of the snow bromide had been depleted.

Consistent with the ‘halogen explosion’ mechanism, Figures 6.4a and 6.4b show that

exposing the halide-doped snow to light, ozone and excess Br2 leads to an increase in [BrCl] and

[Cl2] (with [BrCl] increasing more rapidly than [Cl2]) and a simultaneous decrease in [Br2] and

[O3]. The only exception is the increase in [Br2] as it is introduced from the bulb. Production of

BrCl and Cl2 is not observed (or ceases) when any one of light, ozone or excess Br2 is absent.

Note that the [Br2] between ~ 60 – 80 min slowly decreases due to the fact that the additional Br2

is being leaked from a glass bulb of decreasing pressure. Figure 6.4a illustrates that higher

[BrCl] and [Cl2] are observed when higher [Br2] is present. Interestingly, in the absence of ozone,

blocking and un-blocking the light has no noticeable impact on the dihalogen concentrations.

As mentioned in Section 6.2 (Materials and Methods), depletion of Br2 or Cl2 by photolysis alone

was not important under our conditions.

145

Figure 6.4. Time traces of a) dihalogen concentrations – Br2 (red triangles), Cl2 (black circles) and

BrCl (green squares); and b) ozone concentrations (light blue diamonds) during an experiment in

which an additional flow of Br2(g) was delivered to the sample from a glass manifold (in the period

of time indicated by the vertical blue dashed lines). The yellow areas indicate time periods during

which the sample was illuminated. The dotted areas indicate time periods during which the ozone

generator was switched on ([O3] ~ 1 × 1014 molecules cm-3). The experiment was run under BASE

scenario conditions except high purity NaCl(s) (Br¯ < 0.001 wt%) was used to prepare the solution.

Each point represents the average of a 1 minute time bin (30 data points).

6.3.3. Influence of Snow Br¯/Cl¯

The important role that bromine activation plays in initiating the ‘bromine explosion’ is

also shown in experiments in which we reduced the snow Br¯/Cl¯ ratio by an order of

magnitude (through the use of high purity NaCl(s)). Figure 6.3b shows that although the Cl2

146

fraction is clearly enhanced relative to the BASE scenario under these conditions, the overall

[Cl2] is lower, despite the fact that the pre-freezing [Cl¯] is effectively unchanged. The fact that

reducing the Br¯/Cl¯ ratio while keeping pre-freezing [Cl¯] constant leads to a decrease in Cl2 is

consistent with a snowpack ‘halogen explosion’ that requires the initial formation of Br2 and

BrCl.

6.3.4. Wavelength Dependence

Support for a snowpack ‘halogen explosion’ driven by halogen photolysis is also provided

by the observed wavelength dependence. Replacing the 310 nm long-pass filter with a 360 nm

long-pass filter reduces the rate constant for dihalogen photolysis (J7) in the order Cl2 >> BrCl >

Br2 (26). Indeed, comparison of Figure 6.3c with Figure 6.3a shows that use of the 360 nm long-

pass filter leads to lower yields of all three dihalogens which can be attributed to less HOBr or

HOCl formation via R7 –R9 under these conditions. Furthermore, the slower evolution of the

Cl2 fraction under these conditions is consistent with lower yields of Cl radical due to the fact

that the J7 for BrCl and Cl2 photolysis are lowered to a greater extent than the J7 for Br2. The fact

that we observe photochemical halogen activation under both illumination conditions also

allows us to conclude that OH(g) formed via O(1D) from gas phase ozone photolysis is not an

important oxidant in our experiments.

6.3.5. Surface Area Dependence

Central to the ‘halogen explosion’ mechanism is the recycling of gas phase products back

into the snow. Hence, halogen activation via this mechanism should be enhanced when a larger

surface area is available and Cl2 production should be most affected since it relies solely on the

photochemical pathway (i.e., Br2 and BrCl can both be formed in the dark). The influence of

available surface area was investigated by performing experiments using a coarse mode snow

(see Figure 6.2, right panel and Figure 6.3d), with all other conditions the same as in the BASE

scenario. Comparison of Figure 6.3a and 6.3d shows that the dihalogen production is indeed

suppressed when the available surface area is reduced. In particular, the Cl2 fraction evolves

147

very slowly for the coarse mode experiments; the BrCl fraction remains high for the duration of

the experiment. We note that the larger yields we observe when a greater surface area is present

could also support a reaction taking place at the frozen surface. Halide ions are known to have a

positive surface affinity (27) and thus have the potential to be involved in interfacial reactions

(28).

6.3.6. pH Dependence

Dihalogen formation was reduced when the pre-freezing pH was increased to ~4.3 (see

Figure 6.3e). The dark evolution of [Br2] was the least affected, while the production of [Cl2]

upon illumination was dramatically suppressed. At a pre-freezing pH of 5.9, any dihalogen

production lay below the detection limits. A strong acidity dependence has been reported for

the HOBr + Br¯ reaction (R3) in the aqueous phase (13) and this dependence is likely to be

stronger for the analogous reaction with chlorine. Although studies of HOBr reactions with

frozen halides substrates has shown that Br2 and BrCl will form under a wide range of acidities

(9, 10) we note that those studies were carried out at temperatures below the eutectic for which

the ice surface should be quite different than in the present experiment and for which we also

see little halogen production (vide infra).

6.4. HOx and O3

The ‘halogen explosion’ mechanism requires an HO2 source to be present in the snow

interstitial air. Snowpacks are now known to be sources of OH, HO2 and small organics such as

HCHO (29, 30). Indeed, using the same experimental set-up and similar illumination conditions

Gao et al. (31) measured the release of VOCs from natural snow samples collected from urban,

rural and remote sites. Although organics were not intentionally introduced to our samples, it is

well known that reagent grade salts and laboratory deionized water contain organic impurities

(32). It is also possible that gas phase organics or acids present in the room contaminated our

samples during preparation (fast-freezing in liquid N2). Thus, upon sample illumination, it is

possible that active snowpack chemistry leads to HOx formation. Additionally, the oxidation of

148

VOCs by Cl radicals leads to the formation of HO2 and small aldehydes (i.e., HCHO) which are

themselves precursors for HOx (1). The production of HONO and NOx via the photolysis of a

nitrate impurity in the snow could also play an important role in perturbing the HOx budget. A

few other possibilities for HOx formation, including the photodissociation of surface adsorbed

O3 or reactions involving thermally-hot O(3P), are considered in the Appendix.

Figure 6.6 in the Appendix shows that reducing the ozone concentrations leads to an

overall decrease in the halogen production. Figure 6.7 shows that the evolution of the

dihalogen fractions with time is independent of gas phase [O3] over this range, suggesting that

the chemical mechanism is not affected by changing the ozone concentrations over the range.

6.4.1. Brine Chemistry

An important question concerns the environment where the chemistry is occurring. When

salt solutions freeze, salt ions are excluded from the growing ice matrix into an increasingly

concentrated liquid brine. Thus it has been well-established that sodium halide solutions will

contain small liquid fractions down to temperatures at, or even slightly below, their eutectic

temperatures (17). Together, the temperature and salt concentration dependence that we

observe suggests that the photochemical halogen production takes place in such a concentrated

liquid brine located at the surface of the snow grains. The reasons for this are as follows. At

temperatures near or below the NaCl-water eutectic of 252 K (17), the brine fraction is

negligible. Under these conditions, illustrated in Figure 6.3f, we find dihalogen production to be

significantly suppressed indicating that chloride availability in a liquid brine is key to this

mechanism. At temperatures below the eutectic, chloride will be precipitated in any remaining

brine, leading to chemical fractionation. We observe that the Cl2 fraction grows more slowly at

colder temperatures (see Figure 6.8a in the Appendix) indicating that the bromide/chloride ratio

in the existing brine may be affected by the removal of chloride via precipitation. Fractionation

of bromide with respect to chloride has previously been shown to affect heterogeneous

chemistry on frozen halide surfaces (33). The influence of temperature on the Cl2 production is

149

further illustrated in Figure 6.5 which shows the total Cl2 yield in 50 minutes of illumination as

a function of snow sample temperature. The yields were calculated by integrating the Cl2

concentration-time signal to 50 minutes and scaling by the flow rate. Huff and Abbatt (34) have

reported the only laboratory study of HOCl interactions with chloride/bromide ice surfaces.

They also report no reaction for HOCl or Cl2 with chloride/bromide-ice films at temperatures

below the NaCl-water eutectic.

Figure 6.5. Dependence of the chlorine (Cl2) yield as a function of snowpack T. Vertical error bars

represent the range from two trials. Horizontal error bars represent the estimated uncertainty in

the snowpack T. The yields were calculated by integrating the Cl2 concentration-time signal to 50

minutes and scaling by the flow rate.

Changing the pre-freezing [NaCl] concentration was found to have an effect on the total

dihalogen concentrations (Figure 6.9 in the Appendix) but no effect on the time evolution of the

dihalogen fractions (Figure 6.10 in the Appendix) over the range 0.1 – 1.0 M. This likely

indicates that a) the Br‾/Cl‾ ratio in the brine is the same for the three concentrations studied,

suggesting that fractionation of the halides does not occur during freezing; b) the chemical

mechanism remains the same in all three cases; and c) the difference in the absolute dihalogen

production is due to differences in the total halide content (i.e., brine fraction).

150

6.5. Conclusions

We have observed accelerated halogen activation, in particular Cl2 activation, from

artificial saline snow in the presence of ozone and radiation of actinic wavelengths. Illumination

of the snow samples was found to lead to the rapid release of Br2, then BrCl, and ultimately the

release of Cl2. Our observations are consistent with chemistry occurring in a concentrated liquid

brine located at the surface of the snow grains, with Cl2 activation only occurring at

temperatures above the eutectic (252 K). We found the production of the dihalogens to be

favoured under acidic conditions, more so for Cl2 than for Br2. Finally our results show that

photochemical halogen activation, particularly chlorine activation, is enhanced when a larger

surface area is present. Provided a HOx source is present, the results are most consistent with a

‘halogen explosion’ mechanism in which HOBr and HOCl are formed via the gas phase reaction

of O3 by halogen radicals and are recycled back into the snow to oxidize Br¯ or Cl¯.


Considerable debate has centred on where bromine activation occurs in the polar regions,

with proposals that first-year sea ice (35); frost flowers (36, 37); saline snowpack (38) and

blowing salty snow (39, 40) play an important role. These substrates have variously been

proposed for their high salinity, large surface area or enriched bromide content.

Given the strong acidity and surface area requirement we observe for chlorine production,

we propose that aged saline snow (i.e., at a coastal location) will be most important for chlorine

activation in the polar boundary layer. Although the pH at the air-ice interface has not been

well-constrained, the bulk pH of seawater and seawater-derived substrates (frost flowers, first-

year sea ice, brines) is alkaline (pH ~ 8.3) and is buffered against pH change by the carbonate

system. Furthermore, work from our group suggests that the pH at an air-ice interface is largely

unchanged during freezing (41) and that the frozen seawater surface maintains some buffering

capacity (42). Snowpacks at coastal locations can achieve much lower pH values (~ 4.5) if they

are aged (i.e., have been acidified by atmospheric trace acids) (43, 44). The pH of snow at inland

sites may be even lower, but these samples may also be limited in their halide content.

151

Acidification of natural snow samples may also lead to a very low local surface pH which is not

captured by the bulk, melted pH. The relatively low surface areas of first-year sea ice and even

frost flowers (45, 46) also preclude them from playing an important role in activating chlorine

via the process observed in this study. In our study, we provide strong evidence for active

chemistry occurring within the interstitial air of snow that is enhanced when a larger surface

area is present. The proposal that aged, acidic snow is most relevant to halogen activation is

consistent with very recent measurements by Pratt et al. (18) showing efficient bromine release

from natural snow samples but not from sea ice or brine icicles.

Halide concentrations in coastal snowpacks can vary widely, with average chloride

concentrations ranging from < 1000 - > 33000 μg L-1 and average bromide concentrations

ranging from 30 – 450 μg L-1 for snow over first-year ice (and lower for snow on land or multi-

year ice)(43, 47). Although the concentrations that were used in this study were much higher,

and we did see a dependence on bulk halide concentration, we propose that this chemistry

could still be relevant at these low concentrations. Due to freeze-concentration effects, the

concentration of halides available at the surface of real snow grains may be quite high (48).

Furthermore, snow metamorphism may help mobilize the halides within the snowpack

(whereas our artificial snow may contain trapped liquid pockets due to the very harsh freezing

conditions).

Observation of both Br¯-enriched and Br¯-depleted snow have been observed in the field,

with enrichments attributed to non-sea salt contribution from HOBr and depletions due to

bromine activation chemistry (43, 47). The dependence on the Br¯/Cl¯ ratio we observe is quite

interesting. On the one hand, when a large amount of bromide is available, it will preferentially

be oxidized over chloride, and so production of Br2 and BrCl should dominate. High bromide

concentrations will also react at the surface with Cl2 and BrCl to form BrCl or Br2 respectively, as

has been observed by laboratory experiments (9, 10, 34). However, our experiments do suggest

that molecular Cl2 is formed even when bromide is still present. On the other hand, we find that

Cl2 production decreases when the Br‾/Cl‾ decreases, which we explain via the involvement of

152

HOBr formed via prior bromine activation. Thus a certain amount of bromide may be necessary

to initiate this chemistry. The photochemical halogen activation we observe is very efficient and

should lead to rapid depletion of snow bromide. The high surface area of snow, together with

bromide’s strong surface affinity (27) should particularly favour rapid depletion of snow

bromide.

An important question concerns the role that BrCl plays in contributing to the Cl radical

budget and to further chlorine activation. Our results show that a process involving light and

ozone could liberate large amounts of BrCl from aged, surface snow, which will release Cl

radicals upon photolysis. BrCl is important here, not only as a direct Cl radical source, but as a

precursor to HOCl formation, which we suggest leads to Cl2 formation. Generally consistent

with our study, observations of Br2 and BrCl over snow at Alert show similar trends for both

dihalogens, with Br2 reaching as high as 25 ppt and BrCl reaching as high as 35 ppt (24, 49).

Notably, BrCl was not observed above the 2 ppt detection limit at the beginning of the

campaign (in total darkness) but was observed almost continuously midway through the

campaign and onwards; Cl2 was not observed above the 2 ppt detection limit. The later

appearance of BrCl may be related to both bromide depletion in the snow and the increasing

irradiance during the campaign.

Our results implicate the need for a ‘seed’ reactive halogen source as well as a sustained

HOx source within the snow. Although the relatively high ozone concentrations were likely

responsible for the formation of ‘seed’ Br2 and BrCl in this study, OH produced within natural

snowpacks may play an important role in the initial formation of these species in the field (18).

This indicates that nitrate snowpack photochemistry and organic release from the snow – both

of which affect the HOx budget – are likely intimately linked to halogen activation; further

research is required to improve our understanding of the highly coupled air-ice interactions

occurring within the snowpack.

153

The polar regions, and particularly the Arctic, are strongly affected by our changing

climate. In the Arctic, a large decrease in sea ice extent has been accompanied by an increase in

the relative area of first-year sea ice compared to multi-year ice (50). A long-term increase in the

frequency of ozone depletion events during early spring may be attributed to this change (51).

Given that snow lying on first-year sea ice has a higher salinity, halogen chemistry is expected

to be affected as the Arctic succumbs to further changes. Our study contributes significantly to

our growing understanding of halogen activation processes and how they might be impacted in

a changing world. Further research is required using authentic snow and ice substrates to

support the work presented here.

6.7. Appendix: Supporting Information

Ozone Concentration

Since the dark ozonation of frozen NaCl/NaBr solutions is known to exhibit a (non-linear)

[O3] dependence (6), and the observed photochemical dihalogen production may be sensitive to

the extent of bromide ion depletion and the gas phase concentration of the dihalogens

themselves, the effect of [O3] on the photochemical dihalogen production was investigated in

experiments in which the samples were illuminated first, and then subsequently exposed to

ozone. No halogen production was observed prior to ozone exposure.

Figure 6.6. Time evolution of dihalogen concentrations: Br2 (red triangles), Cl2 (black circles) and BrCl

(green squares) for the BASE scenario conditions with [O3] of a) 3 × 1013 molecules cm-3 and b) 9 × 1013

molecules cm-3. Each point represents the average of a 2.5 minute time bin (77 data points). In these

a) b)

154

experiments the samples were illuminated (λ > 310 nm) prior to turning on the ozone generator (at t =

0).

Figure 6.7. Time evolution of a given dihalogen concentration as a fraction of the total dihalogen

concentration for the BASE scenario conditions and varying [O3] of 3 × 1013 molecules cm-3 (circle), 6 ×

1013 molecules cm-3 (triangle), 9 × 1013 molecules cm-3 (square) and 1.3 × 1015 molecules cm-3 at 252 K

(crosses). Samples were illuminated (λ > 310 nm) prior to turning on the ozone generator (at t = 0).

Panel a) Cl2 (black symbols); b) Br2 (red symbols) and c) BrCl (green symbols).

a)

b)

c)

155

Temperature Dependence


concentration for the BASE scenario conditions and snow sample temperature of 252 K (triangles)

and 257 K (circles). Samples were illuminated at t = 0. Panel a) Cl2 (black symbols); b) Br2 (red

symbols) and c) BrCl (green symbols).

a)

b)

c)

156

Pre-freezing [NaCl] Dependence

Figure 6.9. Representative time evolution of dihalogen concentrations: Br2 (red triangles), Cl2 (black

circles) and BrCl (green squares) for BASE scenario conditions but different pre-freezing NaCl

concentrations a) 0.1 M, b) 0.5 M (BASE) and c) 1.0 M. Each point represents the average of a 2.5

minute time bin (77 data points). The dashed line indicates the time at which the ozone generator was

turned on and the solid line (t = 0) indicates the time at which the samples were illuminated.

a)

b)

c)

157


concentration for the BASE scenario conditions and pre-freezing NaCl of 0.1 M (circle), 0.5 M

(triangle), 1.0 M (square) (the same data as in Figure 6.9). Samples were illuminated at t = 0. Panel a)

Cl2 (black symbols); b) Br2 (red symbols) and c) BrCl (green symbols).

a)

c)

b)

158

Other Possible HOx Sources

Here we discuss possibilities for OH formation. These could represent possible HOx

sources or could act as oxidants themselves. Direct halide oxidation by OH radical could be

responsible for some of the observed halogen release, if a route to OH formation (in the

condensed or gas phase) exists. The mechanism for OH oxidation of halide solutions is

described in greater detail in (8, 52). One possibility is OH formation from impurities within the

snowpack, as mentioned in the main text. Another possibility is that OH or HO2 are formed

during the photolysis of surface-adsorbed ozone. The products of the photodissociation of

ozone adsorbed to an ice surface at relatively warm temperatures are not well known. Ignatov

et al. (53) studied the photolysis of ozone (λ = 320 nm) adsorbed to water ice films at 80 K

(which are bare compared to the ‘brine-covered’ snow grains in this experiment) and found

evidence (using reflection-absorption FTIR) for an H2O2 complex with the ice surface; they also

suggest the possibility of HO2 and OH complexes with the ice surface. Another possibility for

OH production is that translationally-excited O(3P) from O3 photolysis at longer wavelengths

could react with H2O (either in the gas phase or at the ice surface) to form OH. This reaction has

an activation energy ~17 kcal mol-1 and Braunstein et al. (54) have shown that the cross-section

for OH formation is non-zero for collision energies > ~20 kcal mol-1. Thus photolysis of ozone in

the visible Chappuis band could lead to O(3P) with sufficient energy to form OH (although such

a route would require the collision with H2O to occur without prior energy dissipation).

Oxidation by O(3P)

Finally, an alternative route to OCl¯ formation could be the oxidation of Cl¯ by O(3P)

formed via the photolysis of surface adsorbed or dissolved O3. Very little is known about

reactions of O(3P)aq under the cold, briny conditions of our experiment. However one study

suggests that O(3P)aq efficiently scavenges halides in solution to form OCl¯ or OBr¯ (55). The

authors report a relative scavenging rate for Cl¯ only two orders of magnitude smaller than that

for Br¯ (the difference is over six order of magnitude for the ozonation reaction, (12)). This could

imply a route to form OCl¯ which could go on to release Cl2.

159

6.8. References





4375-4418 (2007)

2. A. Steffen, T. Douglas, M. Amyot, P. Ariya, K. Aspmo, T. Berg, J. Bottenheim, S. Brooks, F.

Cobbett, A. Dastoor, A. Dommergue, R. Ebinghaus, C. Ferrari, K. Gardfeldt, M. E. Goodsite, D.

Lean, A. J. Poulain, C. Scherz, H. Skov, J. Sommar, C. Temme, A synthesis of atmospheric

mercury depletion event chemistry in the atmosphere and snow. Atmos. Chem. Phys. 8, 1445-1482

(2008)



Halogen activation via interactions with environmental ice and snow. Atmosp. Chem. Phys. 12,

6237-6271 (2012)

4. A. Saiz-Lopez, R. von Glasow, Reactive halogen chemistry in the troposphere. Chem. Soc. Rev. 41,

6448-6472 (2012)







reaction between O-3(g) and halides at the surface of frozen salt solutions. J. Geophys. Res. Atmos.

115, D16309 (8 pp) (2010)



Res. Lett. 3, 045007 (7 pp) (2008)

9. A. K. Huff, J. P. D. Abbatt, Kinetics and product yields in the heterogeneous reactions of HOBr

with ice surfaces containing NaBr and NaCl. J. Phys. Chem. A 106, 5279-5287 (2002)



11. U. Kirchner, T. Benter, R. N. Schindler, Experimental verification of gas phase bromine

enrichment in reactions of HOBr with sea salt doped ice surfaces. Phys. Chem. Chem. Phys. 101,

975-977 (1997)

12. B. J. Finlayson-Pitts, The tropospheric chemistry of sea salt: A molecular-level view of the

chemistry of NaCl and NaBr. Chem. Rev. 103, 4801-4822 (2003)

13. S. Fickert, J. W. Adams, J. N. Crowley, Activation of Br-2 and BrCl via uptake of HOBr onto


14. G. A. Impey, P. B. Shepson, D. R. Hastie, L. A. Barrie, K. G. Anlauf, Measurements of

photolyzable chlorine and bromine during the Polar sunrise experiment 1995. J.Geophys. Res.

Atmos. 102, 16005-16010 (1997)

15. C. R. Stephens, P. B. Shepson, A. Steffen, J. W. Bottenheim, J. Liao, L. G. Huey, E. Apel, A.

Weinheimer, S. R. Hall, C. Cantrell, B. C. Sive, D. J. Knapp, D. D. Montzka, R. S. Hornbrook, The

relative importance of chlorine and bromine radicals in the oxidation of atmospheric mercury at

160

Barrow, Alaska. J. Geophys. Res. Atmos. 117, D00R11 (16 pp) (2012)

16. J. Liao, L. G. Huey, Z. Liu, D. J. Tanner, C. A. Cantrell, J. J. Orlando, F. M. Flocke, P. B. Shepson,

A. J. Weinheimer, S. R. Hall, H. J. B. Beine, Y. Wang, E. D. Ingall, C. R. S. Stephens, R. S. H.

Hornbrook, E. Apel, A. F. Fried, L. M. Mauldin, J. N. Smith, R. M. S. Staebler, J. A. Neuman, J. B.

N. Nowak, Observations of very high levels of molecular chlorine at Barrow, AK. Nature Geosci.

Submitted, (2013).




18. K. A. Pratt, K. D. Custard, P. B. Shepson, D. A. Thomas, D. Pohler, S. General, J. Zielcke, W. R.

Simpson, U. Platt, D. J. Tanner, L. G. Huey, M. Carlson, B. H. Stirm, Photochemical Production of

Molecular Bromine in Arctic Surface Snowpacks. Nature Geosci. 6, 351 - 356 (2013)

19. T. Thornberry, J. P. D. Abbatt, Heterogeneous reaction of ozone with liquid unsaturated fatty

acids: detailed kinetics and gas-phase product studies. Phys. Chem. Chem. Phys. 6, 84-93 (2004)

20. L. G. Huey, D. R. Hanson, C. J. Howard, Reactions Of SF6‾ And I‾ With Atmospheric Trace

Gases. J. Phys. Chem. 99, 5001-5008 (1995)

21. J. Abbatt, N. Oldridge, A. Symington, V. Chukalovskiy, R. D. McWhinney, S. Sjostedt, R. A. Cox,

Release of Gas-Phase Halogens by Photolytic Generation of OH in Frozen Halide-Nitrate

Solutions: An Active Halogen Formation Mechanism? J. Phys. Chem. A 114, 6527-6533 (2010)

22. S. M. Fan, D. J. Jacob, Surface Ozone Depletion In Arctic Spring Sustained By Bromine Reactions

On Aerosols. Nature 359, 522-524 (1992)

23. R. Vogt, P. J. Crutzen, R. Sander, A mechanism for halogen release from sea-salt aerosol in the

remote marine boundary layer. Nature 383, 327-330 (1996)

24. K. L. Foster, R. A. Plastridge, J. W. Bottenheim, P. B. Shepson, B. J. Finlayson-Pitts, C. W. Spicer,

The role of Br2 and BrCl in surface ozone destruction at polar sunrise. Science 291, 471-474 (2001)

25. T. Tang, J. C. McConnell, Autocatalytic release of bromine from Arctic snow pack during polar

sunrise. Geophys. Res. Lett. 23, 2633-2636 (1996)

26. B. J. Finlayson-Pitts, J. Pitts, James N., Chemistry of the Upper and Lower Atmosphere. (Academic

Press, United States of America, 2000).

27. P. Jungwirth, D. J. Tobias, Specific ion effects at the air/water interface. Chem. Rev. 106, 1259-1281

(2006)

28. E. M. Knipping, M. J. Lakin, K. L. Foster, P. Jungwirth, D. J. Tobias, R. B. Gerber, D. Dabdub, B. J.

Finlayson-Pitts, Experiments and simulations of ion-enhanced interfacial chemistry on aqueous

NaCl aerosols. Science 288, 301-306 (2000)


1510 (2002)







31. S. S. Gao, S. J. Sjostedt, S. Sharma, S. R. Hall, K. Ullmann, J. P. D. Abbatt, PTR-MS observations of

photo-enhanced VOC release from Arctic and midlatitude snow. J. Geophys. Res. Atmos. 117,

D00R17 (10 pp) (2012)

32. S. S. Gao, J. P. D. Abbatt, Kinetics and Mechanism of OH Oxidation of Small Organic

Dicarboxylic Acids in Ice: Comparison to Behavior in Aqueous Solution. J.Phys. Chem. A 115,

9977-9986 (2011)

161



5247 (2012)

34. A. K. Huff, J. P. D. Abbatt, Gas-phase Br-2 production in heterogeneous reactions of Cl-2, HOCl,

and BrCl with halide-ice surfaces. J. Phys. Chem. A 104, 7284-7293 (2000)

35. W. R. Simpson, D. Carlson, G. Honninger, T. A. Douglas, M. Sturm, D. Perovich, U. Platt, First-

year sea-ice contact predicts bromine monoxide (BrO) levels at Barrow, Alaska better than

potential frost flower contact. Atmos. Chem. Phys. 7, 621-627 (2007)

36. A. M. Rankin, E. W. Wolff, S. Martin, Frost flowers: Implications for tropospheric chemistry and

ice core interpretation. J.Geophys. Res. Atmos. 107, 4683 (12 pp) (2002)

37. L. Kaleschke, A. Richter, J. Burrows, O. Afe, G. Heygster, J. Notholt, A. M. Rankin, H. K. Roscoe,

J. Hollwedel, T. Wagner, H. W. Jacobi, Frost flowers on sea ice as a source of sea salt and their

influence on tropospheric halogen chemistry. Geophys. Res. Lett. 31, L16114 (4 pp) (2004)



(2008)

39. X. Yang, J. A. Pyle, R. A. Cox, N. Theys, M. Van Roozendael, Snow-sourced bromine and its

implications for polar tropospheric ozone. Atmos. Chem. Phys. 10, 7763-7773 (2010)

40. A. E. Jones, P. S. Anderson, M. Begoin, N. Brough, M. A. Hutterli, G. J. Marshall, A. Richter, H. K.

Roscoe, E. W. Wolff, BrO, blizzards, and drivers of polar tropospheric ozone depletion events.

Atmos. Chem. Phys. 9, 4639-4652 (2009)

41. S. N. Wren, D. J. Donaldson, Laboratory Study of pH at the Air-Ice Interface. J. Phys. Chem. C 116,

10171-10180 (2012)

42. S. N. Wren, D. J. Donaldson, How does deposition of gas phase species affect pH at frozen salty

interfaces? Atmos. Chem. Phys. 12, 10065-10073 (2012)

43. H. W. Jacobi, D. Voisin, J. L. Jaffrezo, J. Cozic, T. A. Douglas, Chemical composition of the

snowpack during the OASIS spring campaign 2009 at Barrow, Alaska. J. Geophys. Res. Atmos. 117,

D00R13 (13 pp) (2012)

44. P. de Caritat, G. Hall, S. Gislason, W. Belsey, M. Braun, N. I. Goloubeva, H. K. Olsen, J. O. Scheie,

J. E. Vaive, Chemical composition of arctic snow: concentration levels and regional distribution of

major elements. Sci. Tot. Environ. 336, 183-199 (2005)

45. F. Domine, A. S. Taillandier, W. R. Simpson, K. Severin, Specific surface area, density and

microstructure of frost flowers. Geophys. Res. Lett. 32, L13502 (4 pp) (2005)

46. R. W. Obbard, H. K. Roscoe, E. W. Wolff, H. M. Atkinson, Frost flower surface area and

chemistry as a function of salinity and temperature. J. Geophys. Res. Atmos. 114, D20305 (14 pp)

(2009)

47. L. Krnavek, W. R. Simpson, D. Carlson, F. Domine, T. A. Douglas, M. Sturm, The chemical

composition of surface snow in the Arctic: Examining marine, terrestrial, and atmospheric

influences. Atmos. Environ. 50, 349-359 (2012)



49. C. W. Spicer, R. A. Plastridge, K. L. Foster, B. J. Finlayson-Pitts, J. W. Bottenheim, A. M. Grannas,

P. B. Shepson, Molecular halogens before and during ozone depletion events in the Arctic at

polar sunrise: concentrations and sources. Atmos. Environ. 36, 2721-2731 (2002)

50. J. A. Maslanik, C. Fowler, J. Stroeve, S. Drobot, J. Zwally, D. Yi, W. Emery, A younger, thinner

Arctic ice cover: Increased potential for rapid, extensive sea-ice loss. Geophys. Res. Lett. 34, L24501

(5 pp) (2007)

162

51. S. J. Oltmans, B. J. Johnson, J. M. Harris, Springtime boundary layer ozone depletion at Barrow,

Alaska: Meteorological influence, year-to-year variation, and long-term change. J. Geophys. Res.

Atmos. 117, D00R18 (18 pp) (2012)

52. E. K. Frinak, J. P. D. Abbatt, Br(2) production from the heterogeneous reaction of gas-phase OH

with aqueous salt solutions: Impacts of acidity, halide concentration, and organic surfactants. J.

Phys. Chem. A 110, 10456-10464 (2006)

53. S. K. Ignatov, P. G. Sennikov, H. W. Jacobi, A. G. Razuvaev, O. Schrems, Surface species formed

during UV photolysis of ozone adsorbed on water ice films at 80 K. A combined RA-FTIR and

DFT study. Phys. Chem. Chem. Phys. 5, 496-505 (2003)

54. M. Braunstein, R. Panfili, R. Shroll, L. Bernstein, Potential surfaces and dynamics of the

O(3P)+H2O(X(1)A(1))-> OH(X-2 Pi)+OH(X-2 Pi) reaction. J.Chem. Phys. 122, 184307 (13 pp) (2005)

55. O. Amchai, A. Treinin, Chemical reactivity of O(3P) atoms in aqeous solution. Chem. Phys. Lett. 3,

611-613 (1969)

163

C H AP T E R S EVE N

CONCLUSIONS

164

7.1. Summary

7.1.1. Development of Techniques for the Ice Surface

The development of techniques capable of directly interrogating processes occurring at the

air-ice interface was a large focus of this work. In this thesis, glancing-angle Raman

spectroscopy and glancing-angle laser-induced fluorescence (LIF) were used in novel ways to

study the ice surface. Glancing-angle Raman spectroscopy was used for the first time to monitor

ion exclusion to the air-ice interface (Chapter Three). Although this technique was used here to

investigate nitrate exclusion, in theory it could be used to investigate the exclusion of any

compound containing a Raman-active vibrational mode. Glancing-angle Raman spectroscopy

was also used for the first time to measure reaction kinetics at the air-ice interface (by following

the decrease in the intensity of the OH-stretching band with reaction time). Since the exact

reason for the observed intensity decrease was not well established, it is not clear how this

approach could be applied to other systems. However, glancing-angle Raman spectroscopy

could potentially be used to directly monitor the formation or loss of a compound containing a

Raman-active vibrational mode.

Glancing-angle LIF was used in conjunction with surface-active, pH-sensitive

fluorescent dyes to investigate pH at the air-ice interface. The technique exploits the fact that the

neutral and protonated forms of the dyes have different emission and excitation spectra. The

surface-sensitivity of the technique rests in the glancing-angle approach as well as the fact that

the dyes are a) surface-active and b) excluded to the ice surface during freezing. We showed

that the dye acridine can be used to study pH changes at the surface of relatively ‘pure’ ice.

Specifically we showed that the shape of acridine’s fluorescent spectrum (via the acridine

430/470 ratio, as described in Chapter Three) can be used to monitor pH at the air-ice interface.

Additionally, acridine fluorescent decay rates can be used to infer information about acridine’s

local environment at the ice surface; the observed quenching of acridine’s fluorescence in the

presence of Cl‾ provided unique experimental evidence for chloride exclusion to the air-ice

interface. We showed that harmine is a useful pH probe for ‘salty’ ice surfaces since its

fluorescence is not strongly quenched in high ionic strength environments. Specifically we

165

showed that the shape of harmine’s excitation spectrum (via the harmine 290/320 ratio, as

described in Chapter Three and Chapter Four) can be used to monitor pH at the air-ice

interface. Harmine was proven to be sensitive in the pH 8 – 10 range; the use of others dyes in

the β-carboline family (i.e. dyes with different pKa’s) would allow investigations outside this

range.

7.1.2. Concentrations of Species at the Air-Ice Interface

Being able to confidently measure or model the concentration of a reagent at the air-ice

interface is a prerequisite to being able to model its environmental fate. In particular, the degree

to which surface concentrations are enriched relative to bulk concentrations must be known.

One goal of this thesis was to test the applicability of current assumptions regarding the

exclusion of ions during freezing. Using glancing-angle Raman spectroscopy, we found that the

nitrate ion is not strongly excluded to the air-ice interface (Chapter Two). In particular, we

found that nitrate surface concentrations are not well predicted using an equilibrium

thermodynamic approach (i.e., phase diagram or formulation based on freezing point

depression). That our inferred nitrate surface concentrations failed to exhibit a strong

temperature dependence is also in disagreement with expectations based on equilibrium

thermodynamics. This result indicates that not all solutes are universally excluded; models that

assume nitrate is totally excluded to a surface liquid region (LR) may be grossly overestimating

surface concentrations (by many orders of magnitude). Our results suggest that nitrate prefers

to reside within the bulk ice (i.e., trapped in liquid micropockets, or along grain boundaries),

which may have significant implications for interpreting nitrate snowpack photochemistry.

The notion that all solutes are totally excluded is also questioned by the results presented

in Chapter Three. Using glancing-angle LIF and the pH-sensitive dye acridine we showed that

pH at the surface of slightly acidic solutions is largely unchanged upon freezing, which

suggests that hydronium ions (H3O+) are not strongly excluded (to the air-ice interface) during

freezing. This result was found to be independent of acid identity (HNO3 and HCl). Our

166

observations suggest that a possible lowering in pH may accompany the freezing of slightly

basic solutions. This result also indicates that hydroxide ions (OH¯) are not strongly excluded

during freezing. Interestingly, we also found that pH at the ‘pure’ ice surface depends on

whether hydronium ions (from HCl in our study) are deposited to the already formed ice from

the gas phase or frozen from solution. Significantly, this result indicates that surface properties

cannot readily be related to bulk properties. Interpreting field measurements may thus require

some knowledge of snow/ice history. Overall, the results presented in Chapter Three also

indicate that predicting surface concentrations is difficult, and that care should be taken in

applying models from equilibrium thermodynamics to predict LR concentrations for the air-ice

interface. A large drawback of these models is that they cannot predict LR location or

distribution.

Halides, on the other hand appear to behave quite differently. Previous studies have

suggested that halides are excluded to LRs located at the air-ice interface and in concentrations

that are well predicted by equilibrium thermodynamics (1-4). The research presented in this

thesis builds upon that evidence. In Chapter Three, the observed increase in acridine’s

fluorescence decay rate upon freezing dilute HCl(aq) solutions (but not HNO3(aq) solutions)

was highly suggestive of fluorescence quenching by excluded Cl¯ ions. In Chapter Five we

found that, for low bulk halide concentrations, the heterogeneous ozonation of halides proceeds

more rapidly on frozen solutions than on aqueous solutions with the same bulk halide

concentration. We attributed this rate enhancement to enriched surface concentrations due to

the freeze-concentration effect. The saturation in the ozonation kinetics at higher bulk

concentrations (i.e., conditions for which a ‘thick’ brine is expected to be present) is also

consistent with temperature dictating brine concentration and therefore chemistry. For bromide,

the ozonation rate that we obtained in the ‘saturation region’ was in fairly good agreement with

the aqueous phase rate corresponding to the predicted brine composition. The idea that a liquid

‘brine’ is present on frozen halide solutions is also supported by the photochemical halogen

activation study (Chapter Six). We found that the photochemical release of Cl2 from NaCl-

doped snow is significantly suppressed at temperatures below the NaCl-water eutectic

167

temperature. This result indicates that the availability of halides in a liquid brine is critical to the

reaction mechanism.

7.1.3. The ‘Pure’ vs. ‘Impure’ Air-Ice Interface

Over the past few years, the approximation that all ice surfaces can be described as being

‘wetted’ by a thin, cold liquid has been challenged. In this thesis, by investigating both

relatively ‘pure’ ice substrates as well as relatively ‘impure’ (or ‘salty’) ice substrates, we

provided further evidence that this approximation is certainly not valid over all conditions. In

Chapters Three and Four we studied the deposition of HCl(g) to the ‘salty’ ice surface (a frozen

0.5 M NaCl solution) and found that the inferred pH decrease was generally consistent with

HCl(g) dissolution into a liquid brine (i.e., the air-ice interface behaved as if it were a liquid).

However, HCl(g) deposition to the ‘pure’ ice surface led to a very different result, indicating

that the surface environment in this case was quite different (i.e., the air-ice interface exhibited

unique behaviour).

7.1.4. What is the pH of Sea Ice?

One goal of this work was to directly investigate the ability of sea ice surfaces to buffer

their pH against acidification by trace atmospheric acids (especially given that this ability may

be influenced by the precipitation of carbonate salts during freezing). Using glancing-angle LIF

and harmine as a pH indicator (Chapter Four), we showed that while deposition of NH3(g) to a

‘salty’ ice surface leads to an expected increase in surface pH (consistent with NH3(g)

dissolution into a liquid brine), deposition of NH3(g) to a synthetic sea ice surface leads to little

change in surface pH. Since the two ice surfaces should be covered by similar brine fractions,

the observed difference is highly suggestive of a maintained buffering capacity at the synthetic

sea ice surface. This result provides important experimental support for the study by Morin et

al. (5) which suggests that brine alkalinity will not be reduced below seawater values if

carbonate precipitates as ikaite. These results have important implications for understanding

the pH dependence of halogen activation and for constraining the substrates that may be most

168

relevant for halogen activation chemistry.

7.1.5. Towards a Better Understanding of Halogen Activation Chemistry

One of the objectives of this thesis was to gain a better understanding of halogen

activation chemistry. Using glancing-angle Raman and glancing-angle LIF we directly

investigated the dark ozonation of frozen halide surfaces. Importantly, we found that at

ambient pH, this reaction leads to Br2/I2 (i.e., a low bulk pH is not a prerequisite). We also

observed very fast kinetics for the ozonation of frozen iodide solutions, which indicates that

even though natural [I¯] are low, ozonation may play an important role in activating I2. We

studied the activation of molecular halogens from halide-doped snow in the presence of ozone

and light in the actinic radiation using a chemical ionization mass spectrometer (Chapter Six).

This study was novel in that it used a high surface area substrate (artificial snow), and focused

on chlorine rather than bromine activation. We found that, at temperatures above the NaCl-

water eutectic, illumination leads to accelerated Br2 and BrCl production (relative to dark

ozonation) and Cl2 production. The release of Cl2 was found to be particularly enhanced for low

bulk pH and high surface area snow. Importantly, we found that Cl2 is activated before snow

Br¯ is completely depleted. We obtained direct evidence for a snowpack ‘halogen explosion’ in

the interstitial air of the halide-doped snow and suggest that HOCl or HOBr formed via

‘halogen explosion’ chemistry is recycled back to the snow grain surface, where it oxidizes Br¯

or Cl¯ in a liquid brine. Our results highlight a) the importance of understanding ‘seed’ halogen

formation; b) the importance of chemistry occurring in snowpack interstitial air; and c) the

importance of performing laboratory experiments at moderate (warm) temperatures, in the

presence of light, and with more natural, high surface area snow samples. This study suggests

that chlorine activation will be most important from aged (i.e., acidic) snow.

7.2. Future Directions

More work is required to understand surface concentrations on ice. Ideally, we want to be

able to relate bulk concentrations, which are easily measured upon sample melting, to surface

169

concentrations. However, the results from this thesis indicate that we are far from being able to

do so confidently. In particular, further research should be undertaken to determine under

which conditions a given predictive approach (i.e., assumption of total exclusion, or estimation

based on the phase diagram or freezing-point depression) might apply, that is, for which

species or groups of species, and over which temperature and concentration ranges.

Furthermore, it would be advantageous to be able to attribute differences in freeze-

concentration behaviour to differences in chemical properties. For example, is the reason that

halides are excluded to LRs at the air-ice interface related to their aqueous surface activity? The

systems that were studied in this thesis were very simple. Future laboratory studies should

focus on investigating the concentration of species at the surface of more complex systems (i.e.,

multicomponent systems with many solutes) which are more representative of environmental

ices. For example, will the presence of halide salts, which are thought to be excluded to a LR at

the air-ice interface, affect nitrate surface concentrations? The presence of additional solutes

should cause the system to shift from a ‘pure’ ice system covered by a ‘unique QLL’ to an

‘impure’ ice system covered by a ‘liquid brine’. This shift in surface environment may have

important consequences for some solutes (especially if one environment is more favourable

than the other). The presence of additional solutes may have interesting chemical effects as well.

For example, Richards et al. (6) suggest that the presence of Br¯ enhances NO3¯ concentrations at

the surface of deliquesced sea-salt aerosol. In addition to studying more complex systems,

authentic snow and ice samples must be studied. The roles of exclusion vs. deposition need to

be established (and again, this needs to be determined as a function of species, temperature and

concentration), as well as the role of ice morphology.

Here I suggest three possible approaches to improving our knowledge of surface

concentrations on ice. Firstly, more surface-sensitive techniques should be developed or used to

directly probe surface concentrations (i.e., there is a need to probe surface concentrations in

situ). The development of such techniques would also allow the direct monitoring of chemistry

occurring at the ice surface. Secondly, more laboratory studies should be undertaken with the

specific goal of linking chemistry to the surface environment. For example, laboratory studies

170

could be carefully planned to ‘tune’ the surface environment or surface concentration (i.e. by

changing temperature, ionic content or surface area). Changes in reaction yields or products

could then be interpreted in the context of predicted changes in surface concentrations. This

would provide a good evaluation of current approaches to predicting surface concentrations.

Such an approach was cleverly used by Lopez-Hilfiker et al. (4) to investigate the reaction of

N2O5 on halide-doped ice. Finally, multiphase chemical models should continue to explore the

sensitivity of their results to the degree of concentration enrichment. This could be done in two

ways. First, models could identify the assumptions regarding concentration enrichment that are

required in order to best reproduce field or laboratory results. Sensitivity studies could also

reveal how important (or not) it is that the concentration enrichment is accurately captured. For

example, does it matter (to chemistry) if proton concentrations are enriched by many orders of

magnitude (leading to a significant lowering in surface pH) or only very slightly (not leading to

a lowering in surface pH)? This should also be used to get an idea regarding which species

deserve the most attention in the laboratory. Of course these last two approaches (‘bulk’

laboratory studies and modelling studies) will have to be applied carefully, since they will

likely require additional assumptions regarding the chemistry taking place.

The need to better understand pH-dependent chemistry on ice should motivate future

research with a focus on ice surface pH. One aspect that deserves further attention is the

speciation of acids at the air-ice interface. Little work has been done to study acid dissociation at

ice surfaces at temperatures nearing the melting point (conditions under which the presence of

a ‘disordered’ or ‘liquid’ region might complicate the matter). Additionally, a better

understanding of the degree of acid dissociation at the air-ice interface is required, especially for

moderate temperatures and high acid loadings (i.e., under conditions where limited water

availability might start playing a role). Work also needs to be undertaken to determine whether

and how acid speciation (molecular vs. dissociated) affects chemistry. Another topic that

requires further attention concerns the buffering of sea ice surfaces. Acidification of sea ice

surfaces is more atmospherically relevant (than basification, which was presented in this thesis).

The glancing-angle LIF technique could be used with a different dye in the β-carboline family

171

(such as norharmane, which has a pKa of 6.85 at 298 K (7)) which should allow the investigation

of pH changes in the required range. The specific role that carbonate precipitation plays in

determining buffering at the sea ice surface should also be further investigated (i.e., by

performing experiments using ‘home-made’ seawater with and without carbonate salts). The

buffering capacity of authentic sea ice should be investigated as well. Additionally, pH

buffering at the surface of lower salinity substrates (i.e., representative of snow or ice that has

been contaminated by sea-salt aerosol or underlying brine, which will have some carbonate

present, but not at seawater levels) needs to be examined. Future work in technique

development could focus on applying the glancing-angle LIF technique to measure surface pH

of natural snow and sea ice samples.

It continues to be imperative that we understand when the ice surface behaves as if it

were covered by a cold, thin liquid and when it does not. This transition could be explored by

tracking pH changes due to acid deposition as was done in Chapter Five, but for a greater range

of ice substrates (‘tuning’ the surface by varying temperature and salt concentration).

The work presented in this thesis demonstrates the importance of considering light-

induced chemistry. In Chapter Six, we suggested that the photolysis of molecular halogens in

the snow interstitial air leads to the formation of HOBr and HOCl, which become the main

oxidants of snow Br‾ and Cl‾. However, in natural environments, the role of light could be

much more complicated given that natural snow contains light-absorbing organics, nitrate, and

hydrogen peroxide (OH precursors). Thus, more laboratory studies are required to understand

the role that light plays in controlling snowpack emissions. It will likely be essential to perform

these experiments using high surface area substrates (i.e., natural or artificial snow with its

associated interstitial air), especially if recycling of photo-formed species (i.e., gas phase

oxidants) back to the snowpack is integral to the chemistry. In general, more laboratory studies

should be performed in the presence of actinic radiation and using natural and artificial snow

samples (for example, similar to those recently performed by Pratt et al. (8)). Furthermore, not

much is known about photosensitized reactions occurring on snow/ice. Snow is known to

172

contain dissolved organic matter (DOM), humic-like substances (HULIS) and other unknown

chromophores (9-11). For example, at Barrow, AK, HULIS and unknown chromophores were

estimated to account for ~50% of the total light absorption (9). The photochemistry of these

species in snow (either direct, or indirect via photosensitized reactions) has not been well

studied and is thus poorly understood.

Our understanding of the physical and chemical properties of the air-ice interface remains

limited. However, only with a good, fundamental understanding of air-ice chemical interactions

will we a) be able to understand their impact on a global scale and b) understand how they will

be affected as we experience global climate change. Indeed, much of the emphasis has thus far

been placed on air-ice chemical interactions occurring in polar regions. Further work is required

to understand the impact of air-ice chemical interactions on a global scale. What role does snow

and ice play outside of the polar regions? NOx has been measured over mid-latitude snow (12), and

nitrate concentrations in urban areas are quite high (13). However, the importance of nitrate

snowpack chemistry at lower latitudes has not been well established. How does the presence of

road salts (which are heavily used in North American cities in the winter) affect snow and ice

chemistry in urban regions? Since a good understanding of air-ice chemical interactions is

lacking, it is difficult to predict how atmospheric composition (local or global) will be affected

by the changes in ice and snow cover that are accompanying global climate change. The polar

regions have been particularly affected. Of relevance to halogen activation is the recent decline

in the Arctic sea ice extent (14). The decrease in multiyear sea ice has been accompanied by a

relative increase in first year sea ice which has a higher salinity (and therefore higher bromide

and chloride content) (14). Since it has been suggested that first year sea ice represents an

important substrate for halogen activation (15), owing to it high salinity, the frequency and

characteristics of ozone depletion events are expected to be strongly affected by perturbations to

first year sea ice cover (16). Clearly, it is imperative that we understand how various frozen

substrates participate in air-ice chemical interactions so that we can understand how the

atmosphere will be affected as the polar regions rapidly succumb to changes in global climate.

173

7.3. References






99-103 (2007)



5247 (2012)



(2008)

6. N. K. Richards, L. M. Wingen, K. M. Callahan, N. Nishino, M. T. Kleinman, D. J. Tobias, B. J.

Finlayson-Pitts, Nitrate Ion Photolysis in Thin Water Films in the Presence of Bromide Ions. J.

Phys. Chem. A 115, 5810-5821 (2011)

7. M. Balon, J. Hidalgo, P. Guardado, M. A. Munoz, C. Carmona, Acid-Base And Spectral Properties

Of Beta-Carbolines .2. Dehydro And Fully Aromatic Beta-Carbolines. J. Chem. Soc. Perk. Trans. 2,

99-104 (1993)

8. K. A. Pratt, K. D. Custard, P. B. Shepson, D. A. Thomas, D. Pohler, S. General, J. Zielcke, W. R.

Simpson, U. Platt, D. J. Tanner, L. G. Huey, M. Carlson, B. H. Stirm, Photochemical Production of

Molecular Bromine in Arctic Surface Snowpacks. Nature Geosci. 6, 351 - 356 (2013)

9. H. Beine, C. Anastasio, G. Esposito, K. Patten, E. Wilkening, F. Domine, D. Voisin, M. Barret, S.

Houdier, S. Hall, Soluble, light-absorbing species in snow at Barrow, Alaska. J. Geophys. Res.

Atmos. 116, D00R05 (15 pp) (2011)

10. D. Voisin, J. L. Jaffrezo, S. Houdier, M. Barret, J. Cozic, M. D. King, J. L. France, H. J. Reay, A.

Grannas, G. Kos, P. A. Ariya, H. J. Beine, F. Domine, Carbonaceous species and humic like

substances (HULIS) in Arctic snowpack during OASIS field campaign in Barrow. J. Geophys. Res.

Atmos.117, D00R19 (17 pp) (2012)

11. C. Anastasio, T. Robles, Light absorption by soluble chemical species in Arctic and Antarctic

snow. J. Geophys. Res. Atmos. 112, D24304 (14 pp) (2007)

12. R. E. Honrath, M. C. Peterson, M. P. Dziobak, J. E. Dibb, M. A. Arsenault, S. A. Green, Release of

NOx from sunlight-irradiated midlatitude snow. Geophys. Res. Lett. 27, 2237-2240 (2000)

13. S. S. Gao, S. J. Sjostedt, S. Sharma, S. R. Hall, K. Ullmann, J. P. D. Abbatt, PTR-MS observations of

photo-enhanced VOC release from Arctic and midlatitude snow. J. Geophys. Res. Atmos. 117,

D00R17 (10 pp) (2012)

14. J. A. Maslanik, C. Fowler, J. Stroeve, S. Drobot, J. Zwally, D. Yi, W. Emery, A younger, thinner

Arctic ice cover: Increased potential for rapid, extensive sea-ice loss. Geophys. Res. Lett. 34, L24501

(5 pp) (2007)

15. W. R. Simpson, D. Carlson, G. Honninger, T. A. Douglas, M. Sturm, D. Perovich, U. Platt, First-

year sea-ice contact predicts bromine monoxide (BrO) levels at Barrow, Alaska better than

potential frost flower contact. Atmos. Chem. Phys. 7, 621-627 (2007)

16. S. J. Oltmans, B. J. Johnson, J. M. Harris, Springtime boundary layer ozone depletion at Barrow,

Alaska: Meteorological influence, year-to-year variation, and long-term change. J. Geophys. Res.

Atmos. 117, D00R18 (18 pp) (2012)

ion exclusion, ph, and halogen activation at the air- ice interface · 2014. 1. 30. · sumi n....

Documents