Direct emission of I2 molecule and IO radical from the heterogeneous reactions of gaseous ozone with aqueous potassium iodide solution

Article abstract of DOI:10.1021/jp903486u

Recent studies indicated that gaseous halogens mediate key tropospheric chemical processes. The inclusion of halogen-ozone chemistry in atmospheric box models actually closes the ～50% gap between estimated and measured ozone losses in the marine boundary

Full text of DOI:10.1021/jp903486u

7
707
2
009, 113, 7707–7713
Published on Web 06/16/2009
Direct Emission of I Molecule and IO Radical from the Heterogeneous Reactions of
2
Gaseous Ozone with Aqueous Potassium Iodide Solution
Yosuke Sakamoto, Akihiro Yabushita, and Masahiro Kawasaki*
Department of Molecular Engineering, Kyoto UniVersity, Kyoto 615-8510, Japan
Shinichi Enami
W. M. Keck Laboratories, California Institute of Technology, Pasadena, California 91125
ReceiVed: April 16, 2009; ReVised Manuscript ReceiVed: May 21, 2009
Recent studies indicated that gaseous halogens mediate key tropospheric chemical processes. The inclusion
of halogen-ozone chemistry in atmospheric box models actually closes the ∼50% gap between estimated and
measured ozone losses in the marine boundary layer. The additional source of gaseous halogens is deemed
to involve previously unaccounted for reactions of O
we report that molecular iodine, I (g), and iodine monoxide radical, IO(g), are released ([I
during the heterogeneous reaction of gaseous ozone, O (g), with aqueous potassium iodide, KI(aq). It was
found that (1) the amounts of I (g) and IO(g) produced are directly proportional to [KI(aq)] up to 5 mM and
2) IO(g) yields are independent of bulk pH between 2 and 11, whereas I (g) production is markedly enhanced
(g) reacts with I at the air/water interface to produce I (g) and IO(g) Via HOI
3
(g) with sea surface water and marine aerosols. Here,
2
2
(g)] > 100[IO(g)])
3
2
(
2
-
at pH < 4. We propose that O
and IOOO intermediates, respectively.
3
2
-
Introduction
The observed ozone deposition velocities over seawater range
-
1 17-20
from 0.01 to 0.12 cm s .
Garland and co-workers first
(g)
(g) is emitted
Recent field observations in the marine boundary layer
MBL), atmospheric box models, and laboratory experiments
proposed that halides at the sea surface can enhance O
3
(
21
uptake, and experimentally demonstrated that I
2
have revealed that activated halogens play various essential roles
in the global environment. Halogens deplete tropospheric ozone,
perturb the HO
nuclei (CCN), thereby influencing climate change. Unexpect-
edly high concentrations of the iodine monoxide radical, IO(g)
22
even in the absence of biogenic activity. More recently, it has
-
been reported that I in the sea surface microsublayer enhances
x
/NO
x
cycle, and generate cloud condensation
-
O
3
(g) uptake, even though I concentrations in bulk seawater
1-4
-
-
-6 23,24
are extremely low ([I ]/[Cl ] ∼ 10 ).
Fine sea salt aerosol
-
particles, which are significantly enriched in I relative to
(
up to 5-10 times higher than previous measurements) were
25-28
seawater (by 2-4 orders of magnitude),
to release photoactive inorganic halogen compounds, such as
Br
, into the MBL.29 Brown et al. demonstrated that solid
potassium iodide, KI(s), reacts with O (g) much faster than
3
react with O (g)
5
,6
recently determined over the ocean. The detection of IO(g)
and OIO(g) at nighttime suggests the existence of a dark
source.4 Significantly, IO(g) was detected across the tropical
Atlantic ocean, i.e., far away from biogenic coastal sources.8
These observations suggest a hitherto unknown but ubiquitous
2
,7
3
30-32
-
KBr(s) to produce KIO (s).
3
These results imply that I at
both seawater and dehydrated/aqueous aerosol interfaces may
mediate or catalyze gaseous reactive halogen production in the
MBL.
IO(g) source, which could significantly affect the global O
budget. Enhanced halogen-ozone chemistry is actually able to
correct atmospheric box models that underestimated the extent
3
Previous reports have suggested that the ozonation of aqueous
8
-
of tropospheric ozone loss by 50%.
aerosols containing I occurs primarily at the air/water
interface.2
9,33,34
The gas/liquid interface is a unique media that
It is generally believed that the primary sources of reactive
3
5-39
9
-11
should be distinguished from the gas or liquid phases.
For
iodine compounds in the MBL are biogenic I
2
(g)
and alkyl
-
iodide emissions.12
example, I is more abundant at the air/water interface than in
I
2
(g) is rapidly photolyzed into I atoms,
-
40,41
9
the bulk phase, in contrast with Cl .
Furthermore, surface-
which are rapidly oxidized to IO(g). Alkyl iodides are photo-
specific intermediates, which have not been previously detected
lyzed and/or oxidized by OH, Cl, and NO
3
, ultimately leading
3
6,42,43
to IO(g).2,13-16
in homogeneous reactions, have been recently identified.
Recent field measurements show, however, that
-
Finlayson-Pitts and co-workers proposed that Br at the air/
aerosol interface is converted by O (g) to a surface-specific
atmospheric box models based on biogenic iodine emissions
cannot account for the observed levels of gas-phase iodine
species, suggesting the existence of additional sources of reactive
3
-
44
2
BrOOO intermediate, which ultimately yields Br (g). It is
8
apparent that further studies of rates and mechanisms of
reactions at air/water interfaces are required to establish their
possible relevance to atmospheric chemistry. Here, we report
iodine over and across the oceans.
*
To whom correspondence should be addressed. E-mail: kawasaki@
moleng.kyoto-u.ac.jp.
the detection of I
2 3
(g) and IO(g) during the reaction of O (g)
1
0.1021/jp903486u CCC: $40.75  2009 American Chemical Society

7
708 J. Phys. Chem. A, Vol. 113, No. 27, 2009
Letters
Figure 1. Schematic diagram of the present experimental setup
combined with a CRDS and a gas/liquid interaction cell. HRM and
PMT stand for high reflective mirror and photomultiplier tube,
respectively. Synthetic resin putties are used for a solution stopper.
with aqueous KI(aq) in the dark by cavity ring-down spectros-
copy (CRDS).4
5-47
Experimental Section
Figure 1 shows a schematic diagram of the experimental
setup. The principle of CRDS and pertinent experimental details
are presented in the Supporting Information and in previous
publications.4
8-50
3
O (g) was produced by 1 slm (standard liter
per minute) O
2
flow through a high pressure discharge ozonizer
and monitored by UV absorption by a 253.7 nm Hg lamp prior
to the gas/liquid interaction cell. Gas flow rates were controlled
by mass flow controllers, and the total flow rate was maintained
Figure 2. CRD spectra of gaseous IO (A) and I
2
(B) formed during
1
5
16
-3
the reaction of 7.3 × 10 and 1.3 × 10 molecules cm
3
O (g),
respectively, with 5 mM KI solution at 100 Torr and at room
temperature. The high [O (g)] for I spectra was used for clarification
of the spectrum, but other measurements were performed below 7.3 ×
3
2
3
at 3 slm. The concentrations of O (g) were in the range from
1
3
15
-3
4
O
.8 × 10 to 7.3 × 10 molecules cm (2-298 ppmv). At
15
-3
1
0 molecules cm .
1
6
-3
3
(g) concentrations greater than 1.0 × 10 molecules cm ,
aerosol formation from iodine oxides, I
To reduce the complication, [O
n m
O
,51,52 was observed.
15
flow of nitrogen gas (0.05 slm) was introduced at the ends of
the ring-down cavity, close to the mirrors to minimize mirror
deterioration caused by exposure to the reactants and products
in the cell. The total flow rate was adjusted to 3 slm so that the
gas in the cell was completely replaced within a 0.70 s time
3
(g)] was kept below 7.3 × 10
-3
molecules cm , so that no aerosols were formed. The product
concentrations were monitored with an OPO laser (Spectra-
Physics, MOPO-SL, spectral resolution 0.2 cm ) at 435.6 nm
for the IO(g) band head of the A Π3/2 r X Π3/2 (V′ ) 3, V′′ )
) transition. The absorption cross section of IO(g) at 435.63
-1
2
2
3
interval. Hence, the average contact time of O (g) with the
0
-17
2
-1
KI(aq) solution was ∼0.70 s. To minimize possible secondary
reactions, a freshly prepared solution was used for the measure-
ment of each data point. The present application of CRDS
enabled us to monitor primary gaseous products released from
the reaction of O (g) with KI(aq) solutions with an adequate
3
wavelength resolution and sensitivity in less than 1 s. Control
experiments confirmed that dark heterogeneous reactions are
nm was previously measured to be 5.9 × 10 cm molecule
53
with the same wavelength resolution. The IO(g) signal baseline
was taken at 435.0 nm, a region in which there was no IO(g)
absorption. [I
known I (g) into the reaction cell with spectral fitting at
30-455 nm for the B-X band of I
spectra were also measured at 532-540 nm. The observed IO(g)
and I (g) concentrations in the present experiments were (0.03
2
(g)] was calibrated by introducing a concentration-
2
5
4
4
2 2
(g). The I (g) absorption
2
exclusively responsible for I (g) and IO(g) production. See the
Supporting Information for further details.
2
11
14
-3
-
4.0) × 10 and (0.02 - 1.1) × 10 molecules cm ,
respectively.
Results
The gas/liquid interaction cell consisted of a Pyrex glass
container (21 or 96 mm i.d. and 60 cm length) fully covered
with aluminum foil. The cell was maintained at 100 Torr by
means of a rotary pump, a mechanical booster pump, and a
2
Figure 2 shows typical CRD spectra for IO(g) and I (g) during
O
3
(g) flow over the KI(aq) solution. We verified that neither
absorption nor scattering signals appeared within the 420-450
and 530-540 nm detection ranges in the absence of O (g). There
was no difference of the IO(g) and I (g) yields between NaI(aq)
N
2
(l) trap in tandem, and monitored by an absolute pressure
3
gauge. Since there was no appreciable difference in the results
between the 21 and 96 mm reaction cells, all experiments were
conducted at room temperature with the 21 mm cell. KI(aq)
solution (0.01-50 mM and 50 mL) was filled 0.8 ( 0.1 cm
above the bottom of the reaction cell (see Figure 1). The CRDS
detection region is 2 mm above the solution surface. A slow
2
+
+
and KI(aq) solution (within 2%) because K and Na cations
do not play a role in the interfacial reaction, which is consistent
with the previous reports on the ozonolysis of NaI(s) and
KI(s).3
0-32
2
These results suggest that IO(g) and I (g) are
-
3
produced from the reaction of O (g) with I (aq).

Letters
BrO(g) and Br
J. Phys. Chem. A, Vol. 113, No. 27, 2009 7709
2
(g) were not observed during the exposure of
16
-3
2
.4 × 10 molecules cm
3
O (g) to a 50 mM KBr(aq) solution.
Under the present conditions, the limit value of the detection is
11
-3
11
2
.7 × 10 molecules cm for Br
2
(g) and 1.1 × 10 molecules
-3
cm for BrO(g). These results are consistent with the fact that
-
-1 -1
the reaction rate constant of Br (aq) + O
3
(aq), 248 M s , in
6
-
the aqueous phase is 5 × 10 times lower than that of I (aq) +
9
-1 -1 55
O
3
(aq), 1.2 × 10 M s . The lower air/water interface
-
-
affinity of Br as compared to I will also lower its reaction
probability.4
consistent with previous studies.
Other possible IO(g) sources were considered. For example,
0,41
-
3
The inertness of Br (aq) toward O (g) is
2
9
5
6
IO(g) could arise from the I
from the reported rate constant, k ) 4.3 × 10 cm molecule
at 300 K and [O
2
3
(g) + O (g) reaction. However,
-18
3
-1
-
1
56
15
-3
s
3
(g)] ∼ 7.3 × 10 molecules cm , the
maximum concentrations used in the present study, the estimated
reaction half-life is ∼30 s, which is ∼45 times longer than the
present reaction time window, 0.70 s. Thus, the gaseous reaction
of I
in our system. Furthermore, a saturated aqueous iodine I
solution in 50 mM KI (aq) did not enhance the total IO(g) yields
2
(g) + O
3
(g) is too slow to explain the formation of IO(g)
2
(aq)
1
4
-3
even though ∼10 molecules cm
ozonation due to the evaporation. The ozonolysis of saturated
(aq) solution also did not produce any IO(g). We also
confirmed that no IO(g) was formed from dry KI(s) that were
2
I (g) is present before/after
I
2
15
fully bedded in the cell and exposed up to 7.5 × 10 molecules
cm (g). The results on KI(s) are consistent with previous
reports on the ozonation of KI(s), which predominantly produced
Thus, no significant IO(g) is formed by ozonation
(g/aq) and KI(s) under the present conditions. These
experimental results therefore indicate that I (aq) is required
-3
O
3
KIO
of I
3
(s).3
0-32
2
-
to produce IO(g) in our setup. The results also suggest that IO(g)
-
is produced directly from the reaction of O
3
(g) + I (aq);
however, other IO(g) production mechanisms involving the
reaction of O (g) with product(s) from the reaction O (g) with
3
3
-
I (aq) cannot be ruled out.
2
Figure 3 shows the IO(g) and I (g) formation as a function
of [O
3
(g)] at [KI(aq)] ) 5 mM. The solid lines are the fitting
(g)] plateaus at [O
(g)] ∼
.0 × 10 molecules cm , while [IO(g)] plateaus later at
(g)] ∼ 5.0 × 10 molecules cm . The [IO(g)]/[I (g)] ratio
is plotted as a function of [O (g)] in Figure 3C, and has a
positive [O (g)] dependence. This behavior implies that the IO(g)
and I (g) formation mechanisms involve different pathways and
stoichiometries (see below). Figure 4 shows the [IO(g)] and
curves to be used as eye guides. [I
2
3
15
-3
1
15
-3
Figure 3. [IO(g)] (A) and [I (g)] (B) formed during the reaction of
[
O
3
2
2
O
3
(g) with 5 mM KI(aq) solution as a function of [O
has an uncertainty ∼10% for IO and ∼15% for I . (C) The [IO(g)]/
(g)] ratio plotted as a function of [O (g)] at [KI(aq)] ) 5 mM obtained
from the fitting curves of parts A and B.
3
(g)]. Every plot
3
2
3
[I
2
3
2
-
15
[
I
2
(g)] as a function of [I (aq)] when [O
3
(g)] ) 7.3 × 10
-3
-
plot correlates with IO(g)/I
cf. Figure 3).
Figure 6 shows the IO(g) and I
on the bulk aqueous KI(aq) solution pH between 2 and 13 during
the reaction with O (g). There is a significant difference of the
pH profiles between the IO(g) and I (g) formation. The IO(g)
formation was independent of the bulk pH from 2 to 11, while
production significantly decreased (by a factor of ∼7) from
pH 2 to 4 and then remained approximately constant from pH
to 11. The enhancement of the I (g) production below pH 4
2 3
(g) yields as functions of [O (g)]
molecules cm . Both products plateau when [I (aq)] > 5 mM,
likely due to the mass transfer limitations. That is, the
equilibrium between reactions and diffusion from the bulk was
established at [I (aq)] > 5 mM when [O
molecules cm .
(
2
(g) production dependence
-
15
3
(g)] ) 7.3 × 10
-3
3
2
The effective uptake coefficient, γeff, for O
3
(g) on the KI(aq)
solution was obtained from the difference between inflow and
outflow [O (g)] (see the Supporting Information for details). γeff
decreased with increasing [O
I
2
3
-4
3
(g)] from 1.7 × 10 to 0.8 ×
4
2
-
4
+
-
1
0 , leveling off at higher [O (g)], as shown in Figure 5A.
3
may be the result of the reaction H + HOI + I f I
Above pH 11, both the IO(g) and I (g) production decrease
2
2
+ H O.
The observed γeff is well fitted by a Langmuir-Hinshelwood
2
mechanism rather than an Eley-Rideal mechanism, which is
significantly to levels below the limit of the detection. A couple
(g) uptake.5
7-69
-
consistent with a number of recent results of O
3
of explanations will be evaluated. First, the interfacial I
-
γ
eff increased with increasing [I ] from 0 to 50 mM, reaching
a plateau at higher [I ], as shown in Figure 5B. A fitting curve
concentration may be decreased due to the competition for the
-
-
surface sites with [OH ] >10 mM, hindering the interfacial
-
is obtained from a Langmuir-Hinshelwood mechanism (see the
reaction rate of interfacial I with O (g). A second explanation
3
-
Supporting Information for details). The observed γeff vs [O
3
(g)]
could be that, at pH > 11, IO is the dominant species (HOI a

7
710 J. Phys. Chem. A, Vol. 113, No. 27, 2009
Letters
Figure 4. [IO(g)] (A) and [I
2
(g)] (B) formed during the reaction of
(g) as a function of [KI(aq)]. Every plot
has an uncertainty of ∼10% for IO and ∼15% for I
1
5
-3
7
.3 × 10 molecules cm
O
3
2
.
Figure 5. (A) Effective uptake coefficient, γeff, of O
3
(g) on 5 mM
(g)]. (B) Effective uptake coefficient
KI(aq) solution as a function of [O
3
+
-
H + IO ; pK
a
) 10.8) and has a different reactivity than HOI.
15
-3
of 1.4 × 10 molecules cm
O
3
(g) as a function of [KI(aq)]. Every
-
-
For example, unlike HOI, IO may not react with I to yield
plot has an ∼30% uncertainty.
2
I .
The apparent bulk pH effects on the product formation
exclude the possibility that the observed IO(g) is generated from
-
1
0.8. HOI can subsequently react with another I (aq), eq 4, to
2 2
produce molecular iodine, I (aq). The produced I (aq) will either
partition to the gas phase, eq 5, or complex with I (aq), to
produce triiodide, eq 6.
5
6
the gas-phase reaction of I
2
with O
3
.
We also monitored the
-
15
pH change of 5 mM KI(aq) after 5 min of 5.0 × 10 molecules
-3
cm ozonation. The pH was observed to increase from 7.0 to
+
1
1.3. The significant increase in pH (decrease in [H ]) can be
-
-
I (aq) + O (g or interface) f IOOO (interface)
explained by the reaction mechanism proposed below.
3
(1)
Discussion
-
-
From the Henry’s law constant for O
3
(g) in water at 298 K,
(aq)]sat ) ∼20 µM under
(g)] ) 5 × 10 molecules cm . From the calculation to
IOOO (interface) f IO (aq) + O (aq)
(2a)
(3)
2
-1
H ) 0.012 M atm , we calculate [O
3
15
-3
[O
3
-
+
3
6
IO (aq) + H a HOI(aq)
(pK ) 10.8)
a
derive the diffusion depth with the 0.7 s reaction time and a
diffusion coefficient of 10- cm s , I only in the top ∼50
5
2
-1
-
µm of aqueous solution (in 8 mm depth) will be available for
-
+
-
HOI(aq) + I (aq) + H a I
2
(aq) + H
2
O
(4)
(5)
the reaction with dissolved O
partitions to the air/water interface.
3
. Furthermore, I preferentially
3
8-40
Also, the reaction of
-
9
I (aq) with O
3
(aq) is diffusion controlled with k ) 1.2 × 10
I (aq) f I (g)
2
2
-1
-1 55
M
s . Considering these facts, it may be reasonable to
assume that the reaction proceeds predominantly at the air/water
interface rather than in the bulk phase.
Enami et al. showed that the reaction of I at the air/water
-
-
I (aq) + I (aq) a I (aq)
(6)
2
3
-
3
interface of microdroplets with a ppmv level of O (g) can
-
-
+
-
-
HOI/IO (aq) + 2O (aq) f IO (aq) + 2O + H
3 3 2
produce triiodide, I
3
(aq) (in the equilibrium between I (aq)
-
(7)
and I (aq)), and iodate, IO
2
3
(aq), in a short reaction time of ∼1
(g) evolution agrees with the reported
formation. The initial ozonation of I- produces a hy-
poiodous acid intermediate (eqs 1, 2a, and 3), with a pK of
29
ms. The observation of I
2
-
29
I
3
O
3
(g) may first dissolve in the air/water interface Via a
Langmuir-Hinshelwood mechanism (see Figure 5 and the
a

Letters
J. Phys. Chem. A, Vol. 113, No. 27, 2009 7711
The fact that IO(g) is emitted from the KI(aq) solution surface
is inconsistent with the production of IO in bulk aqueous
solution if one apply a diffusion-controlled rate constant of 1.5
9
-1 -1
72
×
10 M
s
for IO(aq) + IO(aq). All IO(aq) would have
-
been self-consumed and converted to I
Therefore, O (g) or dissolved O
to produce IOOO (interface) and subsequently IO and HOI at
the air/water interface rather than in the bulk. The produced
IO(interface) is then rapidly emitted into the gas phase prior to
2
and IO
3
within 1 s.
-
3
3
(interface) reacts with I (aq)
-
+
consumption by reactions. Considering the absence of [H ]
effects over a pH range of 2-10, the IO formation is expected
to occur shortly after the initial reaction step, eq 1. IO formation
at the air/water interface could be explained by a number of
-
unimolecular and/or bimolecular reactions involving the IOOO
intermediate, eqs 2b, 9, 10, and 11.
-
-
IOOO f IO + O
(2b)
2
-
-
IOOO + I f f IO + products
(9)
-
-
IOOO + IOOO f f IO + products
(10)
-
IOOO + HOI f f IO + products
(11)
(12)
IO(aq) f IO(g)
-
Liu et al. theoretically reported that Br (aq) may form a stable
2
Figure 6. [IO(g)] (A) and [I (g)] (B) formed during the reaction of
-
-
O
3
(g) as a function of bulk pH. Every plot has an uncertainty of ∼10%
BrOOO intermediate during the reaction of Br (aq) with
-
55
for IO and ∼15% for I
2
.
O
(aq), which decomposes to yield BrO + O
.
The O-O
3
2
-
bond nearest to the bromine BrO-OO is elongated to 2.033
Å in length and thus more likely to be broken, while the BrO
5
9
Supporting Information). Here, we will propose the initial
5
5
bond is rather strong with a computed length of 1.813 Å. As
-
ozonation of I (aq) proceeds through a trioxide intermediate,
-
mentioned above, Br at the air/aerosol interface may react with
-
IOOO (interface), which is assumed to have a similar structure
-
O
3
(g) to form a surface-specific BrOOO intermediate, which
-
44,55
and reactivity as the reported BrOOO intermediate.
This
3
-
further reacts to form Br
2
(g).44 Similarly, the fate of IOOO
will be favorably IO + O from reaction 2a by way of
thermochemical considerations, as it is very exothermic, but IO
-
unstable intermediate is different from the stable iodate, IO
-
-
-1
2
(
∆
f
G°(IO
An enhancement of I
) KI(aq) is explained by reaction 4. The observation of the
(g) formation from the ozonolysis of dried KI(s)
3
(aq)) ) -134.9 kJ mol ).
2
(g) formation when using acidic (pH <
-
+
O
2
(eq 2b) might also be possible. Even if it occurred, its
contribution would be minor because of the observed results
which give a product ratio of [I (g)]/[IO(g)] > 100. In addition,
4
absence of I
is consistent with a reaction mechanism where H is necessary
to generate I . The observation of the significant pH increase
2
+
2
thermodynamic data also imply this reaction will be unfavorable:
2
-
73
73
∆
f
G°(I (aq)) ) -51.7,
∆
f
H°(O
3
(g)) ) 142.7,
f
∆ H°(IO(g))
from 7.0 to 11.3 during the ozonation also supports this
mechanism.
-
73
-1
)
121.5,74 and ∆
f
G°(O
2
(aq)) ) 31.8 (kJ mol ). Thus, one
may expect that the bimolecular reactions such as eqs 9-11
are more responsible for the IO formation than the unimolecular
reaction, eq 2b. Theoretical calculations for the thermodynamic
A Dushman mechanism (eq 8) might also explain the present
pH enhancement.7
0,71
-
and kinetic data of the IOOO intermediate are desirable. The
proposed reaction mechanism is summarized in Scheme 1.
The present observation that [I (g)] > [IO(g)] may be
explained by some effects other than the reaction pathways of
IO3- + 5I + 6H f 3I + 3H O
-
+
(8)
2
2
2
However, under low pH conditions, the IO -
expected to be a minor product compared with the I
under the present high I (aq) conditions due to the fact that
reaction 4 is faster than reaction 7. Thus, the Dushman
(aq) product is only
product
IOOO discussed above.
-
3
2
(1) The emissions of IO(g) and I
2
(g) are largely controlled
and
-
by the Henry’s law constants. The values 3.0 for I
2
2
-1
72
4.5 × 10 M atm for HOI were reported, respectively.
2
mechanism itself cannot produce sizable I (g) in our experiments.
Although the value for IO has not been determined yet,
7
2
As far as we know, the direct formation of IO(g) from the
the same value for IO and HOI can be assumed. If so,
-
heterogeneous reaction of I (aq) with O
3
(g) has not been
one can expect that I
rather than IO(g).
2
(g) is preferably in the gas phase
reported before. We have confirmed this overall reaction
mechanism by attempting a number of control ozonation
(2) Reactions of IO may occur in the gas phase or at the
2
interface, since IO has a much greater reactivity than I .
experiments. The reaction of O
not produce significant IO(g), which suggests that O
not react with I to produce IO(g) under the present conditions.
3
(g) with saturated I
2
(aq) does
3
(g) does
IO(g or interface) may have been consumed by the self-
reaction or reactions with other species prior to reaching
2

7
712 J. Phys. Chem. A, Vol. 113, No. 27, 2009
Letters
SCHEME 1: Proposed Reaction Mechanism for the
References and Notes
Formation of I (g) and IO(g) from the Interaction of
(g) with I (aq)
2
(
1) von Glasow, R.; Crutzen, P. J., Tropospheric Halogen Chemistry
-
O
3
in Treatise on Geochemistry ed. by Holland, H. D. Turekian, K. K. 2007.
(
(
(
2) Carpenter, L. J. Chem. ReV. 2003, 103, 4953–4962.
3) von Glasow, R. Nature 2008, 453, 1195–1196.
4) Saiz-Lopez, A.; Plane, J. M. C. Geophys. Res. Lett. 2004, 31,
L03111.
(
5) Wada, R.; Beames, J. M.; Orr-Ewing, A. J. J. Atmos. Chem. 2007,
5
8, 69–87.
(
6) Saiz-Lopez, A.; Mahajan, A. S.; Salmon, R. A.; Bauguitte, S. J. B.;
Jones, A. E.; Roscoe, H. K.; Plane, J. M. C. Science 2007, 317, 348–351.
7) Cainey, J. M.; Keywood, M.; Grose, M. R.; Krummel, P.; Galbally,
(
I. E.; Johnston, P.; Gillett, R. W.; Meyer, M.; Fraser, P.; Steele, P.; Harvey,
M.; Kreher, K.; Stein, T.; Ibrahim, O.; Ristovski, Z. D.; Johnson, G.;
Fletcher, C. A.; Bigg, E. K.; Gras, J. L. EnViron. Chem. 2007, 4, 143–150.
(
8) Read, K. A.; Mahajan, A. S.; Carpenter, L. J.; Evans, M. J.; Faria,
B. V. E.; Heard, D. E.; Hopkins, J. R.; Lee, J. D.; Moller, S. J.; Lewis,
A. C.; Mendes, L.; McQuaid, J. B.; Oetjen, H.; Saiz-Lopez, A.; Pilling,
M. J.; Plane, J. M. C. Nature 2008, 453, 1232–1235.
the detection region. Hence, the observed [IO(g)] at 2
mm above the aqueous surface may be significantly
reduced from the initial concentration.
(
9) Saiz-Lopez, A.; Plane, J. M. C.; McFiggans, G.; Williams, P. I.;
Ball, S. M.; Bitter, M.; Jones, R. L.; Hongwei, C.; Hoffmann, T. Atmos.
Chem. Phys. 2006, 6, 883–895.
(
10) Küpper, F. C.; Carpenter, L. J.; McFiggans, G. B.; Palmer, C. J.;
Atmospheric Implications
The present results for I
the work by Garland et al. demonstrating I
Waite, T. J.; Boneberg, E. M.; Woitsch, S.; Weiller, M.; Abela, R.;
Grolimund, D.; Potin, P.; Butler, A.; Luther, G. W.; Kroneck, P. M. H.;
Meyer-Klaucke, W.; Feiters, M. C. Proc. Natl. Acad. Sci. U.S.A. 2008, 105,
6954–6958.
(11) Dixneuf, S.; Ruth, A. A.; Vaughan, S.; Varma, R. M.; Orphal, J.
Atmos. Chem. Phys. 2009, 9, 823–829.
2
(g) production are consistent with
(g) formation from
both artificial and real seawater. It is apparent that interaction
2
2
2
-
of O
3
(g) with 0.1-0.4 µM I at the seawater surface is a possible
(
12) Peters, C.; Pechtl, S.; Stutz, J.; Hebestreit, K.; Honninger, G.;
source of reactive iodine compounds. Recent field and modeling
Heumann, K. G.; Schwarz, A.; Winterlik, J.; Platt, U. Atmos. Chem. Phys.
2005, 5, 3357–3375.
(13) Enami, S.; Yamanaka, T.; Hashimoto, S.; Kawasaki, M.; Tonokura,
K.; Tachikawa, H. Chem. Phys. Lett. 2007, 445, 152–156.
-
studies have revealed that I at the sea surface enhances O
3
(g)
(g)
uptake.2
3,24
Thus, a large discrepancy in the observed O
3
deposition velocities over the seawater from 0.01 to 0.12 cm
(14) Cotter, E. S. N.; Canosa-Mas, C. E.; Manners, C. R.; Wayne, R. P.;
-
1
-
s
could be explained by I , one of the most active components
Shallcross, D. E. Atmos. EnViron. 2003, 37, 1125–1133.
(15) Nakano, Y.; Ishiwata, T.; Kawasaki, M. J. Phys. Chem. A 2005,
109, 6527–6531.
23,24
in the seawater.
phase products during the rapid O
Here, we directly observe I
2
and IO as gas-
-
3
(g)-I (aq) interactions.
(
16) Enami, S.; Hashimoto, S.; Kawasaki, M.; Nakano, Y.; Ishiwata,
The present findings can be applied not only to the air/
seawater interface but also to the air/aerosol interface. It should
be emphasized that I- in fine sea salt aerosol particles is
significantly concentrated as compared to seawater by (2-4
T.; Tonokura, K.; Wallington, T. J. J. Phys. Chem. A 2005, 109, 1587–
1593.
(17) Gallagher, M. W.; Beswick, K. M.; Coe, H. Q. J. R. Meteorol.
Soc. 2001, 127, 539–558.
(
18) Lenschow, D. H.; Pearson, R.; Stankov, B. B. J. Geophys. Res.-
2
5-28
orders of magnitude).
In actual aerosols, HOI(aq) would
Oceans Atmos. 1982, 87, 8833–8837.
(19) Galbally, I. E.; Roy, C. R. Q. J. R. Meteorol. Soc. 1980, 106, 599–
620.
(20) Kawa, S. R.; Pearson, R. J. Geophys. Res. -Atoms. 1989, 94, 9809–
817.
2
9
-
29
competitively react with O
3
(g),
X
(X ) I, Br, Cl),
The reaction of HOI with
Br /Cl yields IBr/ICl, which could be emitted as IBr/ICl or
34
75-77
sulfur(IV), and organic species.
-
-
9
2 2
/Cl
after subsequent processing in the gas phase.2 At an
9
Br
iodine-enriched area such as Mace Head at the West sea of
(
21) Garland, J. A.; Elzerman, A. W.; Penkett, S. A. J. Geophys. Res.
1980, 85, 7488–7492.
(22) Garland, J. A.; Curtis, H. J. Geophys. Res. 1981, 86, 3183–3186.
1
2
Ireland and Lilia at the French Atlantic Coast of Brittany,
(
23) Chang, W. N.; Heikes, B. G.; Lee, M. H. Atmos. EnViron. 2004,
8, 1053–1059.
24) Oh, I. B.; Byun, D. W.; Kim, H. C.; Kim, S.; Cameron, B. Atmos.
-
strong I
initiated by I (aq) + O
be expected. The reported Henry’s law constants, H ) 3.0 for
2
(g) emissions from the reaction of HOI(aq) with I (aq)
3
-
3
(g) at the air/aerosol interface would
(
EnViron. 2008, 42, 4453–4466.
2
-1
72
(25) Duce, R. A.; Wasson, J. T.; Winchester, J. W.; Burns, E. J. Geophys.
Res. 1963, 68, 3943–3947.
I
2
, 2.4 × 10 for IBr, and 1.1 × 10 M atm for ICl imply
that a large fraction of I (g) emissions into the gas phase are
2
(
26) Seto, F. Y. B.; Duce, R. A. J. Geophys. Res. 1972, 77, 5339–5349.
more favorable than those of IBr(g)/ICl(g). Thus, whether sea
salt aerosols release photoactive inorganic halogen compounds
(27) Duce, R. A.; Winchest, Jw; Vannahl, T. W. J. Geophys. Res. 1965,
70, 1775–1799.
(
28) Martens, C. S.; Wesolows, J. J.; Harriss, R. C.; Kaifer, R. J.
Geophys. Res. 1973, 78, 8778–8792.
29) Enami, S.; Vecitis, C. D.; Cheng, J.; Hoffmann, M. R.; Colussi,
into the MBL or not is largely dependent on the actual
compositions.2 Here, we have shown that I
9
2
(g) emissions from
(
-
the O
<
3
(g) + I (aq) reaction are enhanced up to 7 times at pH
A. J. J. Phys. Chem. A 2007, 111, 8749–8752.
(30) Brown, M. A.; Ashby, P. D.; Ogletree, D. F.; Salmeron, M.;
Hemminger, J. C. J. Phys. Chem. C 2008, 112, 8110–8113.
4 over those at pH 5-10, and that non-negligible amounts of
IO(g) should be also emitted even at nighttime.
(
31) Brown, M. A.; Liu, Z.; Ashby, P. D.; Mehta, A.; Grimm, R. L.;
Hemminger, J. C. J. Phys. Chem. C 2008, 112, 18287–18290.
32) Brown, M. A.; Newberg, J. T.; Krisch, M. J.; Mun, B. S.;
Hemminger, J. C. J. Phys. Chem. C 2008, 112, 5520–5525.
33) Enami, S.; Vecitis, C. D.; Cheng, J.; Hoffmann, M. R.; Colussi,
A. J. J. Phys. Chem. A 2007, 111, 13032–13037.
34) Enami, S.; Vecitis, C. D.; Cheng, J.; Hoffmann, M. R.; Colussi,
Acknowledgment. The authors appreciate Ms. S. Hayase for
help with experiments. We greatly appreciate Dr. A. J. Colussi
of California Institute of Technology and Dr. Chad. D. Vecitis
of Yale University for valuable discussions. S.E. thanks the JSPS
Research Fellowships for Young Scientists. This work is partly
supported by a Grant-in-Aid from JSPS (#20245005).
(
(
(
A. J. Chem. Phys. Lett. 2008, 455, 316–320.
(35) Buch, V.; Milet, A.; Vacha, R.; Jungwirth, P.; Devlin, J. P. Proc.
Natl. Acad. Sci. U. S. A. 2007, 104, 7342–7347.
(
36) Davidovits, P.; Kolb, C. E.; Williams, L. R.; Jayne, J. T.; Worsnop,
D. R. Chem. ReV. 2006, 106, 1323–1354.
37) Enami, S.; Hoffmann, M. R.; Colussi, A. J. Proc. Acad. Natl. Sci.
U.S.A. 2008, 105, 7365–7369.
Supporting Information Available: Experimental details.
This material is available free of charge via the Internet at http://
pubs.acs.org.
(

Letters
J. Phys. Chem. A, Vol. 113, No. 27, 2009 7713
(
38) Ghosal, S.; Hemminger, J. C.; Bluhm, H.; Mun, B. S.; Hebenstreit,
(58) Clifford, D.; Donaldson, D. J.; Brigante, M.; D’Anna, B.; George,
C. EnViron. Sci. Technol. 2008, 42, 1138–1143.
(59) McCabe, J.; Abbatt, J. P. D. J. Phys. Chem. C 2009, 113, 2120–
2127.
(60) Ammann, M.; Poschl, U.; Rudich, Y. Phys. Chem. Chem. Phys.
2003, 5, 351–356.
(61) Chang, R. Y. W.; Sullivan, R. C.; Abbatt, J. P. D. Geophys. Res.
Lett. 2005, 32.
(62) Dubowski, Y.; Vieceli, J.; Tobias, D. J.; Gomez, A.; Lin, A.;
Nizkorodov, S. A.; McIntire, T. M.; Finlayson-Pitts, B. J. J. Phys. Chem.
A 2004, 108, 10473–10485.
E. L. D.; Ketteler, G.; Ogletree, D. F.; Requejo, F. G.; Salmeron, M. Science
2
005, 307, 563–566.
39) Jungwirth, P.; Tobias, D. J. J. Phys. Chem. B 2001, 105, 10468–
0472.
40) Cheng, J.; Vecitis, C. D.; Hoffmann, M. R.; Colussi, A. J. J. Phys.
Chem. B 2006, 110, 25598–25602.
41) Cheng, J.; Hoffmann, M. R.; Colussi, A. J. J. Phys. Chem. B 2008,
12, 7157–7161.
42) Hu, J. H.; Shi, Q.; Davidovits, P.; Worsnop, D. R.; Zahniser, M. S.;
Kolb, C. E. J. Phys. Chem. 1995, 99, 8768–8776.
43) Knipping, E. M.; Lakin, M. J.; Foster, K. L.; Jungwirth, P.; Tobias,
D. J.; Gerber, R. B.; Dabdub, D.; Finlayson-Pitts, B. J. Science 2000, 288,
01–306.
44) Hunt, S. W.; Roeselova, M.; Wang, W.; Wingen, L. M.; Knipping,
E. M.; Tobias, D. J.; Dabdub, D.; Finlayson-Pitts, B. J. J. Phys. Chem. A
(
1
(
(
1
(
(
30.
63) Hanisch, F.; Crowley, J. N. Atmos. Chem. Phys. 2003, 3, 119–
(
1
(
64) Kahan, T. F.; Kwamena, N. O. A.; Donaldson, D. J. Atmos. EnViron.
3
2
006, 40, 3448–3459.
(
(
65) Kwamena, N. O. A.; Staikova, M. G.; Donaldson, D. J.; George,
I. J.; Abbatt, J. P. D. J. Phys. Chem. A 2007, 111, 11050–11058.
66) Kwamena, N. O. A.; Thornton, J. A.; Abbatt, J. P. D. J. Phys. Chem.
A 2004, 108, 11626–11634.
67) McNeill, V. F.; Wolfe, G. M.; Thornton, J. A. J. Phys. Chem. A
007, 111, 1073–1083.
68) Mmereki, B. T.; Donaldson, D. J. J. Phys. Chem. A 2003, 107,
1038–11042.
69) Sullivan, R. C.; Thornberry, T.; Abbatt, J. P. D. Atmos. Chem. Phys.
2
004, 108, 11559–11572.
45) Okeefe, A.; Deacon, D. A. G. ReV. Sci. Instrum. 1988, 59, 2544–
551.
46) Wheeler, M. D.; Newman, S. M.; Orr-Ewing, A. J.; Ashfold,
M. N. R. J. Chem. Soc.-Faraday Trans. 1998, 94, 337–351.
(
(
2
(
(
2
1
2
(
(
47) Yu, T.; Lin, M. C. J. Am. Chem. Soc. 1993, 115, 4371–4372.
48) Ninomiya, Y.; Hashimoto, S.; Kawasaki, M.; Wallington, T. J. Int.
(
(
J. Chem. Kinet. 2000, 32, 125–130.
49) Enami, S.; Hoshino, Y.; Ito, Y.; Hashimoto, S.; Kawasaki, M.;
Wallington, T. J. J. Phys. Chem. A 2006, 110, 3546–3551.
50) Enami, S.; Ueda, J.; Nakano, Y.; Hashimoto, S.; Kawasaki, M. J.
Geophys. Res. 2004, 109.
51) Burkholder, J. B.; Curtius, J.; Ravishankara, A. R.; Lovejoy, E. R.
Atmos. Chem. Phys. 2004, 4, 19–34.
52) Martin, J. C. G.; Spietz, P.; Burrows, J. P. J. Phys. Chem. A 2007,
11, 306–320.
53) Nakano, Y.; Enami, S.; Nakamichi, S.; Aloisio, S.; Hashimoto, S.;
Kawasaki, M. J. Phys. Chem. A 2003, 107, 6381–6387.
54) Spietz, P.; Martin, J. C. G.; Burrows, J. P. J. Photoch. Photobio. A
005, 176, 50–67.
55) Liu, Q.; Schurter, L. M.; Muller, C. E.; Aloisio, S.; Francisco, J. S.;
Margerum, D. W. Inorg. Chem. 2001, 40, 4436–4442.
004, 4, 1301–1310.
(
(
(
(
70) Dushman, S. J. Phys. Chem. 1904, 8, 453–482.
71) Schmitz, G. Phys. Chem. Chem. Phys. 1999, 1, 1909–1914.
72) Vogt, R.; Sander, R.; Von Glasow, R.; Crutzen, P. J. J. Atmos.
(
Chem. 1999, 32, 375–395.
73) Bard, A. J.; Parsons, R.; Jordan, J., Standard Potentials in Aqueous
Solution 1985, Marcel Dekker, INC., New York, NY.
(
(
(
(
74) Kaltsoyannis, N.; Plane, J. M. C. Phys. Chem. Chem. Phys. 2008,
0, 1723–1733.
75) Bichsel, Y.; von Gunten, U. EnViron. Sci. Technol. 1999, 33, 4040–
045.
76) Carpenter, L. J.; Hopkins, J. R.; Jones, C. E.; Lewis, A. C.;
1
1
(
(
4
(
(
2
Parthipan, R.; Wevill, D. J.; Poissant, L.; Pilote, M.; Constant, P. EnViron.
(
Sci. Technol. 2005, 39, 8812–8816.
(
77) Martino, M.; Mills, G. P.; Woeltjen, J.; Liss, P. S. Geophys. Res.
(
(
56) Vikis, A. C.; Macfarlane, R. J. Phys. Chem. 1985, 89, 812–815.
57) Gonzalez-Labrada, E.; Schmidt, R.; DeWolf, C. E. Phys. Chem.
Lett. 2009, 36, L01609.
Chem. Phys. 2007, 9, 5814–5821.
JP903486U