Atmosphericiodinelevelsinfluencedbysea ......1 Supplementary Information Atmospheric iodine levels influenced by sea surface emissions of inorganic iodine Lucy J. Carpenter1*, Samantha

Correction notice Nature Geosci. 6, 108–111 (2013) Atmospheric iodine levels influenced by sea surface emissions of inorganic iodine Lucy J. Carpenter, Samantha M. MacDonald, Marvin D. Shaw, Ravi Kumar, Russell W. Saunders, Rajendran Parthipan, Julie Wilson & John M. C. Plane In the version of this Supplementary Information originally posted online on 13 January 2013, equation (21) and the units for [O3] in the text after equation (19) were incorrect. These errors were corrected on 27 March 2013. In the version of this Supplementary Information originally posted online on 13 January 2013, and in the revised Supplementary Information posted online on 27 March 2013, equations (20) and (21) contained errors. These errors were corrected on 25 April 2013.

© 2013 Macmillan Publishers Limited. All rights reserved.

1

Supplementary Information

Atmospheric iodine levels influenced by sea surface emissions of inorganic

iodine

Lucy J. Carpenter1*

, Samantha M. MacDonald2, Marvin D. Shaw

1, Ravi Kumar

2, Russell W.

Saunders2, Rajendran Parthipan

1,3 , Julie Wilson

4,1, and John. M. C. Plane

2*

1Department of Chemistry, University of York, Heslington, York YO10 5DD, UK

2School of Chemistry, University of Leeds, Leeds, LS2 9JT, UK

3Now at College of Chemical Sciences, Institute of Chemistry, Ceylon, Sri Lanka

4Department of Mathematics, University of York, Heslington, York YO10 5DD, UK

*Correspondence to: [email protected] and [email protected]

Methods

Gaseous I2 measurements by spectrophotometry

Ozone was produced from dry hydrocarbon-free air by its exposure to a commercial ozone

generator (185 nm excitation, UVP) and monitored using a model 49i ozone analyzer (Thermo

Scientific). Ozonolysis of iodide solutions was carried out using a 2 L reaction vessel with a 49

cm2 surface area of solution (Fig. S1). Flow rates over the solution were maintained at 0.2 L

min-1

. 20 mL iodide solutions (10-6

- 10-5

M KI: AnalaR®

, BDH, ≥ 99.0 %) were prepared either

from phosphate (Fluka, 99.8 %) - buffered HPLC water at pH 8 or by spiking seawater collected

from the coastal waters of Cargese, Corsica (42.14º N, 8.60º E, filtered through 0.45µm filter

paper (Whatman) and stored in amberised glass bottles at 2- 8 °C) with iodide (0.2 mL, 1 x 10-3

M). Iodide solutions were administered into the reaction vessel using a gas tight syringe

(Samco) via a Luer lock tap.

Control experiments were conducted to establish whether ozone was deposited to the

experimental system in the absence of solution, to HPLC water or to the phosphate buffer

solution in the absence of iodide. The uptake of ozone in all control experiments was small (<

5% of the rate over buffered iodide solutions). The aerodynamic resistance (Γa) in the

experimental system was determined using very concentrated iodide solutions (0.02 M) at which

the surface resistance Γs is zero (1). Γs and hence k1 (Eq. (1) and (2) in main text) was

determined from observed pseudo first order ozone uptake rates at 70 ppbv O3(g) over buffered

iodide solutions (10-6

- 10-5

M [I-]) (1). Exponential curves fitted the observed data with r

2

values of 0.99 and the measured ozone uptake rates were very similar to the results of Garland et

al. (2) such that vd was fitted well using equations (3) and (4) with λ= k1[I-] and k1= 2.0 × 10

9

M-1

s-1

at 293 K – this value was used in the interfacial model (Table S1). Although there is

substantial evidence that iodide accumulates at the air/water interface compared to the bulk (3),

the derived rate constant k1 integrates such effects.

Prior to measurement of iodine emissions, ozone was passed through the system until a constant

concentration (± 2.5%) was observed. Moisture was removed from the gas stream using two

SUPPLEMENTARY INFORMATIONDOI: 10.1038/NGEO1687

NATURE GEOSCIENCE | www.nature.com/naturegeoscience 1


2

spiral condensers in series held at 0 and -10°C. Total gaseous iodine losses within the

experimental system were typically 40 ± 2.2% and these were accounted for.

Emissions of gaseous I2 were measured by trapping the gas flow in n-hexane at ≤ -50 °C for 60

minutes followed by spectrophotometric detection at 522 nm (4) using a dual beamed Perkin

Elmer Lambda 25 UV/VIS spectrophotometer fitted with 10 mm quartz sample and reference

cells. I2 calibration curves were constructed using standard solutions prepared from ground

molecular iodine (puriss, p.a., Riedel-de Haen) dissolved in hexane (HPLC grade, Fischer UK)

within the concentration range 0.5 – 5 x 10-3

M. Trapping efficiencies and losses of gaseous

iodine were determined using a commercial I2 permeation tube (Kin-Tek™, USA) maintained at

60°C.

Gaseous HOI measurements by spectrophotometry

The experimental set up used was similar to that shown in Fig. S1 except no condenser was

used. Experiments were carried out by supplying dry ozone (150 ppbv, 100 sccm) to the 2 L

reaction vessel containing 1 L of phosphate buffer solution (PBS; 0.1 mol L-1

, pH 8, Fluka

Analytical) and KI (10 – 50 x 10-6

M; Fisher Scientific). Evolved gaseous HOI was collected

within a blacked-out midget bubbler (25 mL, Supelco) containing phenol red (PR;

phenolsulfonphthalein) solution (15 mL, 50 x 10-6

M; Sigma Aldrich) and PBS (at pH 7) at 5

°C. Experiments were conducted with (∼ 120 rpm) and without mixing of KI PBS to

investigate the effect bulk mixing had on gaseous HOI evolution (Fig. 1, lower panel).

The assay for the determination of evolved gaseous HOI was based upon the selective

iodination of aqueous PR to iodophenol blue (IPB; 3, 3, 5, 5 - tetraiodophenolsulfonphthalein)

(5), adapted from a previous HOI in-situ incubation study of seaweeds (6). IPB was measured

spectrophotometrically at 591 nm with ε = 47.4 mmol L-1

cm-1

(5) (Fig. S2). Since PR is in

excess and four halogenations of PR are required to form the final tetrahalo product, the molar

HOI: IPB ratio is 4. The ratio of PR loss to IPB generated was observed to be 1, confirming

that the tetrahalo- compound is formed from rapid halogenation of PR and is the only stable

product.

The selectivity of the IPB method for HOI over I2 was confirmed by the complete absence of

IPB after passing gaseous I2 (40 ppbv) for 150 min through PR solution, as shown by Figure S2.

The trapping efficiency of gaseous HOI by the PR trap utilized was investigated by using two

midget gas bubblers (25 mL, Supelco), each containing 20 mL of PR (50 x 10-6

M, pH 7, 0.1 M

PBS), in series under the experimental conditions. Gaseous HOI was produced by introducing

gaseous ozone (150 ppbv, 100 sccm) to a 20 x 10-6

M KI, 0.1 M PBS, pH 7 (1 L) solution

magnetically stirred at 60 – 120 rpm. Gaseous HOI enrichment was carried out in triplicate for

2.5 hrs and determined spectrophotometrically from IPB produced in each trap. IPB was not

detected within the second trap suggesting a HOI trapping efficiency of 100%.

Gaseous I2 and HOI measurements by iodine oxide particle (IOP) detection

Iodide solutions (10-7

– 10-3

M) were prepared using KI (Alfa Aesar, 99%) and deionised water

and the pH of the solution before and after the experiment recorded (Jenway 350 pH meter). 85

mL of iodide solution was added to a 133 mL cylindrical reaction vessel with a surface area of

54.7 cm2

(Fig. S3). Ozone was produced from O2 by exposure to a mercury UVP pen-ray lamp

(185 nm excitation) and was measured using a model 49c ozone analyzer (Thermo Scientific).

The ozone concentration was varied by changing the distance of the pen-ray lamp from the


3

photolysis cell. The ozone flow rate was maintained at 0.01 L min-1

with an additional 0.04 L

min-1

N2 flow giving a flow rate through the reaction vessel of 0.05 L min-1

. The gas flow

containing emitted I2 was then introduced to the iodine oxide particle generation cell where an

additional 0.04 L min-1

O3/O2 flow and 0.51 L min-1

N2 flow were added. I2 was detected by

conversion into iodine oxide particles (IOPs); this was achieved by photolysis using a small

tungsten lamp to produce I atoms and subsequent reaction with O3. The IOPs were then detected

using an electrical mobility spectrometer (EMS VIE10: Tapcon GmbH), consisting of a nano-

differential mobility analyser (nano-DMA) and a Faraday cup electrometer (FCE).

Detection of HOI was achieved by replacing the tungsten lamp with a Xenon lamp, in order to

increase the relative fraction of light output in the UV. Using a blue glass band-pass filter

(Schott UG-1, transmittance window 270-420 nm, and >670 nm), selective photolysis of HOI

was achieved. This technique is thus specific for an iodine-containing species that absorbs

strongly in the near-UV region, i.e., the technique is designed to discriminate specifically

against I2. In the experiments which involved just buffered I- solutions, HOI is therefore the only

possible candidate. A yellow glass long-pass filter (Schott GG495, transmittance > 480 nm)

was employed to selectively photolyse I2. The photolysis rates of HOI and I2 through each of the

filters were determined by convoluting the transmitted spectral intensity of the Xenon lamp

(measured using a grating spectrometer and CCD detector) with the respective molecular

absorption cross section (7). The IOP masses measured with each of the filters were then used to

determine the ratio [HOI]/[I2] from the following expression (note there is negligible photolysis

of HOI through the yellow filter):

€

HOI[ ]I2[ ]

=Mb * 2J

y (I2 )M y −2Jb (I2 )

Jb (HOI)

where Mb and My

are the IOP masses observed using the blue and yellow filters, respectively;

Jy(I2) is the photolysis rate of I2 through the yellow filter, and Jb(I2) and Jb(HOI) are the

photolysis rates for I2 and HOI through the blue filter.

To calibrate the iodine detection system a cell containing iodine crystals was added to the

experimental set-up in place of the solution cell, and maintained at 273 K using an ice bath. N2

was flowed over the cell at varying flow rates from 5 mL min-1

to 25 mL min-1

and a linear trend

in IOP mass observed. The vapour pressure of I2 at 273 K was taken from Baxter and Grose (8).

The concentration of I2 entrained in the N2 flow was further diluted by the addition of the O3/O2

and N2 flows in the IOP generation cell. By comparing the concentration of I2 flowing through

the IOP generation cell, and the amount of I2 in the IOPs measured (assuming a composition of

I2O5 as deduced by Saunders and Plane (9), and a bulk I2O5 density of 5.0 g cm-3

), a percentage

efficiency for the system was calculated for each flow rate. The average efficiency was

4.2 × 10-2

%, and this conversion factor was used to convert the measured IOP mass into I2

number density in order to estimate the I2 flux from the solution. The conversion efficiency is a

function of the photolysis rate of I2 (or HOI) and the residence time of the gas flow in the

photolysis cell. Both of these parameters can of course be changed to increase the efficiency.

However, it is important that the IOP mass determined from the measured particle size

distribution is a linear function of the I2 in the flow, and the particles do not grow to have

diameters in excess of 40 nm (the cut-off of the Tapcon EMS instrument). These constraints

were met by photolysing only a small fraction of the I2. The linearity of the detection system


4

was checked using flows of known I2 concentration, as described above. Finally, even though

the conversion efficiency was kept low, the IOP method has an excellent detection limit of ~ 2

ppt for I2.

Modelling iodine emissions: The sea-surface model

Model set up and initiation The commercial modeling program FACSIMILE (MCPA Software Ltd, UK) was used to solve

the chemical reactions involved in interfacial aqueous iodine dynamics, including iodide

oxidation, iodine disproportionation, and iodine reduction. The complete reaction scheme is

shown in Table S1. Separate experiments were performed to verify the model could

successfully simulate iodine disproportionation at different pH levels (Fig. S4).

Iodine production is initiated by the gas phase flux of O3 into the interfacial layer, FO3

FO3 = vD [O3(g)] (7)

The change in O3 concentration in the interfacial layer, [O3int], is thus:

d[O3int] /dt = A/VvD [O3(g)] (8)

where A/V is equal to the inverse of the depth of the interfacial layer, defined as the

reactodiffusive length δ.

Rapid production of I2 follows the reaction of deposited O3 from the atmosphere with iodide at

the sea surface (Eq. 1 and 2 in main text). We note that Sakamoto et al. (10), also measured IO

production, inferred as a by-product of an IOOO- intermediate formed in reaction (1), at about

1% of the gaseous I2 formed. Because of the minor importance of this species compared to I2

and HOI, and the lack of kinetic data, we ignore its production in this study.

The interfacial layer was treated as a box, assuming no horizontal advection (i.e. assuming horizontal gradients to be small) but mixing vertically with bulk mixed layer water at a fixed

interfacial layer turnover time (for laboratory studies) or by a wind speed-dependent expression

for transfer velocity for ambient conditions (11) calculated from the water-side resistance to

HOI and I2 transfer. Calculation of mass transfer velocities for both laboratory and environmental conditions are described below.

Concentrations of [I-], [H+] and [OH-] were fixed for each model run. For all computations, all iodine species except IO3

- and HIO2 (neither of which effect the gaseous iodine evolution) were

in steady state after a few seconds. For modelling surface seawater iodine emissions, we

simulated an open ocean scenario with pseudo first order rate constants for “O3 + DOM”

interfacial reactions of 100 s-1

(12) and “I2/HOI + DOM” of 5 x 10-5

s-1

(13) and a coastal ocean

scenario with the same reactions at 500 s-1

(12) and 7 x 10-3

s-1

(13), respectively. Thus we

assume that organic reactions dominate chemical control of O3 deposition in coastal waters (12,

14, 15) and are competitive with the iodide reaction in open ocean waters. Total ozone

deposition velocities vD were calculated from equations (3) and (4) in the main text (where λ

includes “O3 + DOM” reactions), using aerodynamic resistances typical for the marine boundary

layer (12).


5

Mass flux calculations According to the two-resistance model for air-water partitioning (16, 17), the mass flux F (mol

m-2

s-1

) of a trace gas species is described by:

F = Kt (cw – cg/H) (9)

l/ Kt= 1/Kw+ 1/HKa (10)

where cw and cg are the respective water and gaseous concentrations, H is the dimensionless gas-

over-liquid form of the Henry’s law constant and Kt, Kw and Ka are the total, liquid and air mass

transfer coefficients (m s-1

), respectively. The inverse of Kw and Ka are the respective water-side

and air-side resistances.

For water-soluble molecules such as HOI, the rate of mass transfer is dominated by the air-side

resistance thus Kt = HKa. For sparingly soluble molecules such as I2, the rate of mass transfer is

dominated by the water-side resistance, though the air-side resistance can reduce the total mass

transfer by several percent.

Calculating mass transfer coefficients for laboratory conditions To calculate Ka for our laboratory experiments, we use an empirical formulation relevant for

laminar flow conditions/indoor environments (18) based upon the dimensionless Sherwood

number, Sh:

Ka = (ShDa)/L (11)

Here Ka is in m h-1

, L is the characteristic length (m) calculated from the square root of the

source area, and Da is the diffusivity in air (m2 h

-1).

The Sherwood number is function of the temperature-dependant Schmidt number of the gas in

question in air (Sca, dimensionless), and the Reynold’s number Re:

Sh = 0.664 Sca1/3Re

1/2 (12)

Sca = µ/(ρDa) (13)

Re = (Luρ)/µ (14)

Where u is the air velocity (m h-1

), ρ is the density of air (g m-3

) and µ is the viscosity of air (g

m-1

h-1

). Thus, at 20 oC, we calculate Ka= 4.2 x 10

-4 m s

-1 for HOI and 3.7 x 10

-4 m s

-1for I2.


6

For calculating the liquid mass transfer coefficients Kw from still water, we again use the

empirical approach of (18) where:

Kw = 2.99 Dw (15)

In this case Kw and Dw (the liquid phase diffusion coefficient) are in m h-1

and m2 h, respectively.

Total mass transfer coefficients Kt for laboratory conditions were calculated from Eq. 10 using

temperature - dependant Henry’s law constants for HOI (19) and I2 (20).

Calculating mass transfer coefficients for environmental conditions Various empirical parameterisations for the air side mass transfer coefficient Ka applicable to the

environment and for experiments in wind tunnels have been derived as a function of wind speed,

u, friction velocity, u*, the Schmidt number of the gas, and the drag coefficient, CD. For

calculating Ka (m s-1

) for environmental conditions, we use the parameterization suggested by

Johnson (21):

Ka = (16)

where κ is the von Karman constant (commonly taken to be 0.4 in seawater) and u* and CD are

calculated from equations given in (21) for a wind speed of 7 m s-1

.

We used the Nightingale et al. (11) parameterization for the waterside transfer velocity. Total

mass transfer coefficients Kt for environmental conditions were calculated assuming a seawater

temperature of 15 oC, air temperature of 20

oC and 10 m wind speed of 7 m s

-1.

The change in aqueous phase concentration due to volatilisation (assuming cg/H to be

negligible) is thus:

dcw/dt = Kt/ δ .cw (17)

where δ is the interfacial layer thickness, defined here as the reacto-diffusive length for O3.

Multiple linear regression model for defining marine emissions of HOI and I2

In order to derive algorithms for marine emissions of HOI and I2, the relationships between each

covariate, iodide concentration (I-), ozone concentration (O3) and wind speed (ws), and the

response variable in question (as computed from the sea-surface model) were investigated.

Ozone levels have a simple multiplicative effect on the response in both cases, i.e. increasing the

ozone level by a factor, k, increases both HOI and I2 emissions by the same factor. Therefore,

models were initially developed for a constant ozone level and later multiplied by the

appropriate factor. Ozone and wind speed were considered separately for fixed values of the

other covariate in each case. Although both responses show clear association with iodide and

wind speed individually, none of the relationships are linear and the covariates were therefore

transformed before fitting a linear regression model. The functions applied to each covariate to


7

achieve a linear relationship with the response are shown in Table S2, together with the

correlation in each case.

A multiple linear regression model of the form

€

ˆ y = β0 +β1x1 +β2x2 +β3x1x2 (18)

where

€

x1 and

€

x2 denote the transformed covariates and the

€

βi , for i = 0,…,3 are coefficients to

be determined, was fitted to allow for interaction between the variables. For both responses, the

intercept was not significant in the regression and models were subsequently fitted without the

€

β0 term.

In the case of I2, the coefficient of

€

(I − )1.3was also not significant, but the coefficients of ln(ws)

and the interaction term,

€

ln(ws)*(I − )1.3, were both highly significant (p < 0.0001). The F-

statistic for the comparison of the two models, with and without the

€

(I − )1.3term, showed that the

reduced model is preferable. The final model to predict I2 flux is therefore

€

FluxI2 = O3(g)[ ]∗ I−(aq)[ ]1.3∗ 1.74×109 −6.54×108 ∗ ln(ws)( )

(19)

where the flux is in nmol m-2

d-1

, [O3] is in nmol mol-1

, [I-] in mol dm

-3 and wind speed in m s

-1

This fit resulted in a correlation of 0.9991 between the calculated and predicted I2 values.

For HOI, the coefficients of both covariates and the interaction term were all found to be highly

significant, leading to the following model with a correlation of 0.9986 between calculated and

predicted values:

€

FluxHOI = O3(g)[ ]∗ 4.15×105 ∗I−(aq)[ ]ws

−20.6ws

−2.36×104 ∗ I−(aq)[ ]

(20)

The least significant term in this model is that involving the wind speed alone, although the p-

value of 1.02 x 10-08

for the corresponding coefficient shows that this is highly significant. With

a p-value < 2.2 x 10-16

, the partial F-test also shows that this variable is significant in the model.

However, fitting a model without this variable leads to a simpler model that still has a

correlation of 0.9945 between calculated and predicted values:

€

FluxHOI = O3(g)[ ]∗ I −(aq)[ ] * 3.56×105

ws− 2.16×104

. (21)


8

Tropospheric HAlogen chemistry MOdel (THAMO)

THAMO is a 1-D chemistry transport model with 200 stacked boxes at a vertical resolution of 5

m (total height 1 km). Wind speed measurements collected at three heights (4, 10 and 30 m)

were used to construct an eddy diffusion coefficient (Kz) profile (22, 23). Aircraft measurements

of temperature in the BL at Cape Verde show a strong temperature inversion about 1 km from

the surface (22), indicating that the BL at Cape Verde is decoupled from the free troposphere.

Hence, the Kz profile is assumed to increase up to a height of 30 m (where it peaks at 3 x 104

cm2 s

-1), after which it decreases at a constant rate to a value of 2 cm

2 s

-1 at the top of the BL.

The model treats iodine, bromine, O3, NOx and HOx chemistry using over 210 reactions. The

chemical scheme is from Saiz-Lopez et al. (25), updated by Mahajan et al. (22). The model is

constrained with typical measured values of the following species: [NOx] = 25 ppt; [CO] = 110

ppb; [DMS] = 30 ppt; [CH4] = 1820 ppb; [ethane] = 925 ppt; [CH3CHO] = 970 ppt; [HCHO] =

500 ppt; [isoprene] = 10 ppt; [propane] = 60 ppt; [propene] = 20 ppt (24, 26, 27, 28, 29) . The

average background aerosol surface area used is 1.0 × 10-6

cm2 cm

-3, an average value measured

at Cape Verde by Allan et al. (30). The modelled HOx concentrations are in sensible accord with

measured values at Cape Verde (31).

The sea-air flux of HOI and I2 from the interfacial model were computed for an average wind

speed of 7 m s-1

, temperature of 296 K, O3 mixing ratio of 30 ppbv and a sea-surface iodide

concentration of 100 x 10-9

M. Under these conditions the model predicted surface

concentrations of [HOI] = 5.7 x 10-9

M and [I2] = 6.6 x 10-12

M, with sea-air transfer velocities of

KtHOI

= 4.9 × 10-5

and KtI2

1.9 × 10-3

cm s-1

. The sea-air fluxes of HOI and I2 were calculated both

as purely evasive terms and also as equilibrated with their surface atmospheric concentrations.

Figure S5 illustrates the height-time profiles for IO, OIO, I2 and HOI over a diurnal cycle, where

the iodine source comprises the measured iodocarbon flux (22, 24) together with the fluxes of

HOI and I2 calculated (taking account of the partial pressure difference across the interface).

This figure illustrates a number of features: IO is only present at significant levels during

daytime, and extends from the surface up to about 300 m (~50% of the near surface mixing

ratio); OIO is only present at significant levels during the night because of its rapid photolysis

during daytime (32) although the mixing ratio is always relatively low; daytime levels of I2 are

very low because of rapid photolysis (33), and after an initial increase following sunset the I2

mixing ratio decreases during the night because of its reaction with NO3; HOI is present during

the day because of HOx chemistry, and increases towards sunset as its photolysis rate decreases;

HOI is present at low levels during the night because any HOI which evades from the ocean

reacts quite rapidly on sea-salt aerosol.

References

1. I. E. Galbally, C. R. Roy. Destruction of ozone at the Earth’s surface. Q. J. R. Meteorol. Soc. 106, 599–620 (1980).

2. J. A. Garland, A. W. Elzerman, S. A. Penkett. The mechanism for dry deposition of ozone

to seawater surfaces. J. Geophys. Res. 85, 7488–7492 (1980).

3. S. Ghosal et al. Electron spectroscopy of aqeous solution interfaces reveals surface

enhancement of halides. Science. 307, 563–566 (2005).

4. R. Chance et al. A comparison of spectrophotometric and denuder based approaches for the

determination of gaseous molecular iodine. Atmos. Meas. Tech .3, 177–185 (2010).


9

5. V. L. Hill, S. L. Manley. Release of reactive bromine and iodine from diatoms and its

possible role in halogen transfer in polar and tropical oceans. Limnol. Oceangr. 54, 812–822

(2009).

6. R. M. Wever et al. Brominating activity of the seaweed Ascophyllum nodosum: Impact on

the biosphere. Environ. Sci. Technol. 25, 446 – 449 (1991).

7. R. Atkinson et al. Evaluated kinetic and photochemical data for atmospheric chemistry:

Volume III – gas phase reactions of inorganic halogens. Atmos. Chem. Phys. 7, 981–1191

(2007).

8. G. P. Baxter, M. R. Grose. The vapor pressure of iodine between 50 degrees and 95

degrees. J. Am. Chem. Soc. 37, 1061–1072 (1915).

9. R. Saunders, J. M. C. Plane. Formation pathways and composition of iodine oxide ultra-fine

particles. Environ. Chem. 2, 299–303 (2005).

10. Y. Sakamoto, A. Yabushita, M. Kawasaki, S. Enami. Direct emission of I2 molecule and IO

radical from the heterogeneous reactions of gaseous ozone with aqueous potassium iodide

solution. J. Phys.Chem. A. 113, 7707–7713 (2009).

11. P. D. Nightingale et al. In situ evaluation of air-sea gas exchange parameterizations using

novel conservative and volatile tracers. Global. Biogeochem. Cycles. 14, 373–387 (2000).

12. L. Ganzeveld et al. Atmosphere-ocean ozone exchange: A global modeling study of

biogeochemical, atmospheric, and waterside turbulence dependencies. Global. Biogeochem. Cycles. 23, GB4021 (2009).

13. V. W. Truesdale, G. W. Luther, C. E. Canosa-Mas. Molecular iodine reduction in seawater: An

improved rate equation considering organic compounds. Mar. Chem. 48, 143–150 (1995).

14. D. Clifford, D. J. Donaldson, M. Brigante, B. D’Anna, and C. George. Reactive uptake of ozone by

chlorophyll at aqueous surfaces, Environ. Sci. Technol. 42, 1138–1143 (2008).

15. M. Martino, B. Lézé, A. R. Baker, and P. S. Liss. Chemical controls on ozone deposition to water

Geophys Res. Lett. 39, doi:10.1029/2011GL050282 (2012).

16. W. G. Whitman. The two-film theory of gas absorption. Chem. Metal. Eng. 29, 146–148

(1923).

17. P. S. Liss, P. G. Slater. Flux of gases across the air-sea interface. Nature. 247, 181–184

(1974).

18. Z. Guo, N. F. Roache. Overall Mass Transfer Coefficient for Pollutant Emissions from

Small Water Pools under Simulated Indoor Environmental Conditions. Ann. Occup. Hyg. 47, 279–286 (2003).

19. R. von Glasow, R. Sander, A. Bott, P. J. Crutzen. Modelling halogen chemistry in the

marine boundary layer. 1. Cloud-free MBL, J. Geophys. Res.107, 4341–4356 (2002).

20. D. A. Palmer, R. W. Ramette, R. E. Mesmer. The hydrolysis of iodine – equilibria at high-

temperatures. J. Nucl. Mater. 130, 280–286 (1985).

21. M. T Johnson. A numerical scheme to calculate temperature and salinity dependent air-

water transfer velocities for any gas. Ocean Sci. 6, 913–932 (2010).

22. A. S. Mahajan et al. Measurement and modelling of tropospheric reactive halogen

species over the tropical Atlantic Ocean. Atmos. Chem. Phys. 10, 4611–4624 (2010).


10

23. R. B Stull. An introduction to boundary layer meteorology, Kluwer Academic Publishers, London.

(1988).

24. K. A. Read et al. Extensive halogen-mediated ozone destruction over the tropical Atlantic Ocean,

Nature, 453, 1232–1235 (2008).

25. A. Saiz-Lopez et al. On the vertical distribution of boundary layer halogens over coastal

Antarctica: implications for O3, HOx, NOx and the Hg lifetime, Atmos. Chem. Phys., 8, 887–

900 (2008).

26. L. J. Carpenter et al. Seasonal characteristics of tropical marine boundary layer air measured at the

Cape Verde Atmospheric Observatory. J. Atmos. Chem. 67, 87–140 (2010).

27. J. D. Lee et al. Year-round measurements of nitrogen oxides and ozone in the tropical North

Atlantic marine boundary layer. J. Geophys. Res. 114, 1–14 (2009).

28. A. S. Mahajan et al. DOAS observations of formaldehyde and its impact on the HO(x) balance in

the tropical Atlantic marine boundary layer. J. Atmos. Chem. 66, 3, 167–178(2008).

29. K. A. Read et al. Intra-annual cycles of NMVOC in the tropical marine boundary layer and their

use for interpreting seasonal variability in CO. J. Geophys. Res. 114, D21303 (2009).

30. J. D. Allan et al. Composition and properties of atmospheric particles in the eastern Atlantic and

impacts on gas phase uptake rates. Atmos. Chem. Phys. Discuss. 9, 18331–18374 (2009).

31. L. K. Whalley et al. The chemistry of OH and HO2 radicals in the boundary layer over the tropical

Atlantic Ocean. Atmospheric Chemistry and Physics, 10, 1555–1576 (2010).

32. J. C. Gómez Martin, S. H. Ashworth, A. S. Mahajan, and J. M. C. Plane. Photochemistry of OIO:

Laboratory study and atmospheric implications. Geophys. Res. Lett. 36, L09802 (2009).

33. A. Saiz-Lopez et al.Absolute absorption cross-section and photolysis rate of I2, Atmos. Chem. Phys,

4, 1443–1450(2004).


11

Figures S1-S5

3 way adjustable tap

Gas transfer lines (PFA, covered with black tape)

MFC Mass flow controller

Figure S1. Experimental set up used to investigate ozone uptake and quantify iodine emissions

by/from iodide solutions and seawater by spectrophotometry. The glass reaction vessel and

traps were covered in foil during all experiments to avoid photolysis of iodine.


12

Figure S2. Iodination of phenol red (PR, 50 µM) by gaseous HOI, showing the IPB peak at

about 590 nm and the PR “dip” at 430 nm. The blue trace shows the result of trapping HOI

evolved on exposure of 50 µM KI to 500 ppbv O3 at a flow rate of 100 sccm for 1 hr in PR

solution (pH 7) at 5 oC. The red trace is a 2 hr trap of I2 from a permeation oven (250 ppbv I2 at

100 sccm) in PR solution, showing the complete absence of IPB. The green trace is a 50 µM

PR blank. The spectra are all autozeroed against 50 µM PR.

Absorbance

nm


13

KI solution

IOP Generation Cell

Tungsten lamp

System Control

EPC

FC

E

Nano-DMA + FCE

N2 + O3

O2 + O3 N2

O2 Hg lamp

HV

1 S

enso

r mod

ule

N2

Figure S3. Schematic diagram of the IOP detection system with black arrows showing direction

of air flow. The reaction vessel and IOP generation cell were covered with black cloth to prevent

unwanted photolysis of I2.


14

Figure S4. Comparison of measured (symbols with error bars representing 1σ variability) and

modelled (lines) iodine disproportionation at a range of pHs in phosphate buffer (NaH2PO4,

Fisher, 99%) prepared in N2-sparged HPLC water and adjusted for pH by titrating with freshly

prepared sodium hydroxide (NaOH, Fisher, 99%). I2 solutions were prepared by overnight

stirring of the required amount of I2 flakes (Analytical grade, Fisher) in degassed HPLC water,

covered by aluminium foil. Disproportionation was studied from an initial [I2] concentration of

1.6 x 10-4

M. [I2](aq) was monitored by spectrophotometric detection at 460 nm (32) and

calibrated as described in section 1.2, where degassed HPLC water was utilised instead of

hexane to dissolve ground I2(s).


15

Figure S5. Modelled iodine chemistry at Cape Verde using the 1-D model THAMO. The panels

show the diurnal variations of IO, OIO, I2 and HOI in the 1 km high marine boundary layer.


16

Tables S1-2

Table S1. Kinetic data used in the interfacial model

Aqueous or heterogeneous reaction Rate constant of

forward reaction

Rate constant of

reverse reaction

Reference(s)

Iodide solution and seawater reactions:

1. O3(g)→ O3 (interface, l) 1vd,O3 x (A/V) s-1 See text

2. O3 (interface, l) + I- (+ H+) → HOI 2.0 x 109 M-1 s-1 (pH 8)

Garland et al. 1980; Magi et

al., 1997

3. I2 + I-↔ I3- 6.2 x 109 M-1 s-1 8.9 x 106 s-1 Forward: Lengyel et al., 1993 Back: Palmer et al. 1984

4. I2 (+ H2O) ↔ I2OH- + H+ 3.2 s-1

2.0 x 1010M-1s-1 Lengyel et al., 1993

5. I2 (+ H2O) ↔ H2OI+ + I- 1.2 x 10-1 s-1 1.0 x 1010M-1s-1 Lengyel et al., 1993

6. I2 + OH- ↔HOI + I- 7.0 x 104 M-1 s-1 2.1 x 103 M-1 s-1 Sebők-Nagy and Körtvélyesi 2004

7. I2OH- ↔ HOI + I- 1.34 x 106s-1 4.0 x 108 M-1 s-1 Lengyel et al., 1993

8. H2OI+↔HOI + H+ 9.0 x 108s-1 2.0 x 1010 M-1 s-1 Lengyel et al., 1993

9. HOI + HOI ↔H+ + I- + HIO2 2.5 x 101 M-1 s-1 2.0 x 1010 M-2 s-1 Forward: Schmitz, 2004 Back: Edblom et al. 1987

10. HOI ↔ IO- + H+ 1.0 x 10-1 s-1 1.0 x 1010 M-1 s-1 Paquette et al., 1986

11. HIO2 + HOI ↔IO3- + I- + 2H+ 2.4 x 102 M-1 s-1 1.2 x 103 M-3 s-1

Forward: Furrow, 1987 Back: Schmitz, 2000

12. HOI + IO-↔HIO2 + I- 1.5 x 101 M-1 s-1 Negligible Bischel and von Gunten, 2000

13. I2→ bulk 2kmix

14. HOI → bulk 2kmix

Additional reactions in seawater only3:

15. O3 (interface, g) (+ DOM) →

products 500 s-1 (coastal) 100s-1 (open ocean)

Ganzeveld et al. 2009

16. I2 (+ DOM) → products

7.0 x 10-3 s-1 (coastal) 5.0 x 10-5s-1 (open ocean)

Truesdale et al., 1995a and 1995b

17. 4HOI (+ DOM) → products

7.0 x 10-3 s-1 (coastal) 5.0 x 10-5s-1 (open ocean)

Assumed analogously to R16

18. HOI + Br- + H+↔IBr 4.1 x 1012 M-2 s-1 8.0 x 105s-1 Faria et al.1993

19. HOI + Cl- + H+↔ICl 2.9 x 1010 M-2 s-1 2.4 x 106s-1 Wang et al., 1989

20. I2 + Br-↔ I- + IBr 4.74 x 103 M-1 s-1 2.0 x 109 M-1 s-1 Faria et al.1993

21. I2 + Cl- ↔ I2Cl- 8.33 x 104 M-1 s-1 5.0x 104s-1 5Margerum et al. 1986

22. ICl2-↔ICl + Cl- 6.0 x 105s-1 1.0 x 108 M-1 s-1 6Margerum et al. 1986

23. I- + ICl↔ I2Cl- 1.1 x 109 M-1 s-1 1.5s-1 Margerum et al. 1986

1 For defining d[O3interface(liquid)]/dt, A is surface area of liquid, V is volume of interfacial layer.

2For the stirred laboratory experiments,kmix was fixed at 0.4 s-1. For the unstirred experiments, kmix was

set to zero. In simulating marine conditions we calculate a wind-speed dependent water transfer velocity.


17

3[Cl-] and [Br-] were fixed at 0.55M and 8.6 x 10-4 M, respectively. Iodate [IO3-] was fixed at

200 x 10-9M for the seawater model runs (this has no effect on volatile iodine emissions). Photooxidation

of I- by O2 and other potential oxidants in seawater including IO3- and nitrate (NO3

-) were reported to be

negligible by Truesdale (2007) and were therefore not included in the model. Pseudo first-order

interfacial O3 loss rates to DOM in coastal and open waters were based on Ganzeveld et al. (2009).

4It is likely that HOI reacts with DOM more rapidly than I2, however even if we assume a reaction rate

100 times faster than that of I2 with DOM, the reduction in the modelled I2 and HOI flux (compared with

using the same rates for I2 and HOI, as shown in the Table) is 0.17% and 0.16%, respectively.

5The forward reaction is an upper limit, the backward reaction is a lower limit. Changing these rate

constants by a factor of 100 (whilst maintaining the equilibrium constant K) has a negligible difference

on the simulated iodine emissions.

6Estimates from Margerum et al. (1986). Changing these rate constants by a factor of 100 (whilst

maintaining the equilibrium constant K) has a negligible difference on the simulated iodine emissions.

Table S1 references

1. Y. Bichsel, U. von Gunten. Hypoiodous acid: Kinetics of the buffer-catalyzed

disproportionation, Water Res. 34, 3197–3203 (2000).

2. E. C. Edblom, L. Gyorgyi, M. Orban, I. R. Epstein. A mechanism for dynamical behaviour in the

Landolt reaction with ferrocyanide, J. Am. Chem. Soc.109, 4876–4880 (1987).

3. T. B. Faria, I. Lengyel, I. R. Epstein, K. Kustin. Combined mechanism explaining nonlinear

dynamics in bromine(III) and bromine(IV) oxidations of iodide ion. J. Phys. Chem. 97, 1164–1171

(1993).

4. S. Furrow. Reactions of iodine intermediates in iodate-hydrogen peroxide oscillators, J. Phys. Chem. 91, 2129–2135 (1987).

5. L. Ganzeveld et al. Atmosphere-ocean ozone exchange: A global modeling study of

biogeochemical, atmospheric, and waterside turbulence dependencies. Global. Biogeochem. Cycles. 23, GB4021 (2009).

6. J. A. Garland, A. W. Elzerman, S. A. Penkett. The mechanism for dry deposition of ozone

to seawater surfaces. J. Geophys. Res. 85, 7488–7492 (1980).

7. I. Lengyel, I. R Epstein, K. Kustin. Kinetics of iodine hydrolysis, Inorg. Chem. 32, 5880–5882

(1993).

8. D. W. Margerum et al. Kinetics of the iodine monochloride reaction with iodide measured by the

pulsed-accelerated-flow method, Inorg. Chem. 25, 4900–4904 (1986).

9. J. Paquette, J. C. Wren, B. L. Ford. The Disproportionation of Iodine (1), in Proceedings of OECD Iodine Workshop, Harwell, AERE R 11974 pp. 29–45 (1986)

10. D. A. Palmer, R. W. Ramette, R. E. Mesmer. Triiodide ion formation equilibrium and activity-

coefficients in aqueous-solution. J. Solution Chem. 13, 673–683 (1984).

11. G. Schmitz. Kinetics of the Dushman reaction at low I− concentrations, Phys. Chem.

Chem.Phys. 2, 4041–4044 (2000).

12. G. Schmitz. Inorganic reactions of iodine(+1) in acidic solutions, Int. J. Chem. Kinet. 36, 480–493

(2004).

13. K. Sebők-Nagy, T. Körtvélyesi. Kinetics and mechanism of the hydrolytic disproportionation of

iodine, Int. J. Chem. Kin. 36, 596–602 (2004).


18

14. V. W. Truesdale, C. E. Canosa-Mas, G. W. Luther. Disproportionation and reduction of molecular

iodine added to seawater, Mar. Chem. 51, 55–60 (1995a).

15. V. W. Truesdale, G. W. Luther, C. E. Canosa-Mas. Molecular iodine reduction in seawater: An

improved rate equation considering organic compounds, Mar. Chem. 48, 143–150 (1995b).

16. V. W. Truesdale. On the feasibility of some photochemical reactions of iodide in seawater, Mar. Chem. 104, 266–281 (2007).

17. Y. L. Wang, J. C. Nagy D. W. Margerum. Kinetics of hydrolysis of iodine monochloride measured

by the pulsed-accelerated- flow method, J. Am. Chem. Soc.111, 7838–7844 (1989).

Table S2. Transformation of the covariates required to give a linear relationship with the

response (for fixed O3 levels).

Response(y) Covariate(x) Transformation

f(x)

ws 1/ws HOI

I- I !

ws ln(ws)

I2 I

-

€

I −( )1.3


Documents

Atmosphericiodinelevelsinfluencedbysea ......1 Supplementary Information Atmospheric iodine levels influenced by sea surface emissions of inorganic iodine Lucy J. Carpenter1*, Samantha