Uptake and reaction of ClONO2 on NaCl and synthetic sea salt

Article abstract of DOI:10.1021/jp0046290

The uptake and reaction of chlorine nitrate (ClONO₂) on NaCl and synthetic sea salt (SSS) powders was studied at 298 K using a Knudsen cell interfaced to a quadrupole mass spectrometer. A time-dependent uptake coefficient was observed, with a large initial uptake coefficient measured for most samples of γ_init > 0.1, followed by a smaller and slowly declining uptake coefficient at longer reaction times. This behavior is shown to be consistent with uptake into, and reaction in, water on the salt surface. The steady-state uptake coefficient on NaCl was invariant over a range of ClONO₂ concentrations from 10¹² to 10¹³ molecules cm^-3 but was dependent on the number of salt particle layers in a manner consistent with approximately two layers of particles being available for reaction. The results of experiments using monolayers and sub-monolayers of salts, where the available reactive surface is known, give an uptake coefficient at longer times after the rapid initial uptake for ClONO₂ on NaCl of γ_t = (6.5 ± 3.0) × 10^-3 (2σ). The larger uptake coefficient obtained initially compared to longer reaction times helps to reconcile different values reported earlier by other research groups for this reaction. The uptake coefficient on SSS, which holds more water, is much larger than that for NaCl, with γ_init = (0.42_-0.42^+0.46) and γ_t = (0.16_-0.16^+0.20 (2σ). At the higher uptake coefficients measured for SSS, this reaction could be a significant source of Cl₂ in the marine boundary layer if chlorine nitrate is available at a constant concentration of ～5 ppt.

Full text of DOI:10.1021/jp0046290

5178
J. Phys. Chem. A 2001, 105, 5178-5187
Uptake and Reaction of ClONO2 on NaCl and Synthetic Sea Salt
Michael E. Gebel† and Barbara J. Finlayson-Pitts*
Department of Chemistry, UniVersity of California, IrVine, California 92697-2025
ReceiVed: December 31, 2000; In Final Form: March 15, 2001
The uptake and reaction of chlorine nitrate (ClONO2) on NaCl and synthetic sea salt (SSS) powders was
studied at 298 K using a Knudsen cell interfaced to a quadrupole mass spectrometer. A time-dependent uptake
coefficient was observed, with a large initial uptake coefficient measured for most samples of γinit > 0.1,
followed by a smaller and slowly declining uptake coefficient at longer reaction times. This behavior is
shown to be consistent with uptake into, and reaction in, water on the salt surface. The steady-state uptake
coefficient on NaCl was invariant over a range of ClONO2 concentrations from 1012 to 1013 molecules cm-3
but was dependent on the number of salt particle layers in a manner consistent with approximately two layers
of particles being available for reaction. The results of experiments using monolayers and sub-monolayers of
salts, where the available reactive surface is known, give an uptake coefficient at longer times after the rapid
initial uptake for ClONO2 on NaCl of γt ) (6.5 ( 3.0) × 10-3 (2σ). The larger uptake coefficient obtained
initially compared to longer reaction times helps to reconcile different values reported earlier by other research
groups for this reaction. The uptake coefficient on SSS, which holds more water, is much larger than that for
NaCl, with γinit ) (0.42+0.46) and γt ) (0.16+0.20
-0.16) (2σ). At the higher uptake coefficients measured for SSS,
-0.42
this reaction could be a significant source of Cl2 in the marine boundary layer if chlorine nitrate is available
at a constant concentration of ∼5 ppt.
Introduction
salt such as NaCl is also possible:11-14
The uptake and reaction of gases with sea salt aerosol can
lead to the formation of gaseous inorganic chlorine compounds,
including Cl2, HCl, BrCl, ClNO2, and ClONO2. Some of these
compounds can be photolyzed, producing chlorine atoms. Such
a source of highly reactive atomic chlorine in the troposphere
has the potential to impact the oxidative balance and overall
air chemistry.1-10 Chlorine atoms in the troposphere can react
with either organics
ClONO2 + NaCl f Cl2 + NaNO3
(5)
Subsequent photolysis of Cl2 generates two chlorine atoms so
that another chloride ion, in addition to the one from which
ClONO2 was formed, is activated. Clearly, understanding the
kinetics of reaction 5 is important for assessing the role of
chlorine chemistry in the marine boundary layer.
The kinetics of the reaction of chlorine nitrate with NaCl
powders has been studied by two groups, with results differing
by ∼2 orders of magnitude. Using a flow reactor, Timonen et
al.12 measured uptake coefficients in the range of 0.009-0.07;
however, after correction for the effects of diffusion15,16 into
the salt layers, which accounts for the increased available surface
area beyond the top surface geometric area, the final value
reported was (4.6 ( 3.0) × 10-3 at 296 K. Rossi and
co-workers13,14 used a Knudsen cell to study this reaction and
reported the uptake coefficients to be 0.23 ( 0.06 and 0.10 (
0.05, respectively. Similar kinetics were observed using thin,
spray-deposited films of salt as well as bulk powder samples
with multiple grain layers; as a result, they chose not to apply
a correction for pore diffusion to their measured uptake
coefficients. In all studies, Cl2 was observed as the major
reaction product in 100% yield, within experimental error.
We report here experimental studies of the uptake of ClONO2
on powders of NaCl as well as on synthetic sea salt (SSS). We
show that a large initial uptake, followed by smaller, long-term
uptake, is consistent with water on the surface playing a major
role in the uptake and reaction of ClONO2. By using sub-
monolayers of particles with well-defined sizes in some experi-
ments, the role of diffusion into the multiple particle layers was
probed and its impact on measurements made using multiple
Cl + RH f HCl + R
(1)
or ozone:
Cl + O3 f ClO + O2
(2)
One possible subsequent reaction of the ClO formed in
reaction 2 is reaction with NO2 to form chlorine nitrate:
ClO + NO2
9
M8 ClONO2
(3)
Potential fates of ClONO2 in the troposphere include deposition
as well as hydrolysis on surfaces to form HOCl and nitric acid:
ClONO2 + H2O f HOCl + HNO3
(4)
Further reaction of chlorine nitrate with the components of sea
* Author to whom correspondence should be addressed [telephone (949)
824-7670; fax (949) 824-3168; e-mail bjfinlay@uci.edu].
† Present address: Monitoring and Laboratory Division, California Air
Resources Board, 9480 Telstar Ave., Ste. 4, El Monte, CA 91731.
10.1021/jp0046290 CCC: $20.00 © 2001 American Chemical Society
Published on Web 05/05/2001

Uptake and Reaction of ClONO2 on Salts
J. Phys. Chem. A, Vol. 105, No. 21, 2001 5179
layers assessed. The results of the previous studies12-14 are all
shown to be consistent with our observations. The atmospheric
implications are discussed.
Experimental Section
Experiments were performed in a glass Knudsen cell,
described in detail elsewhere,17 which is equipped with a
movable lid for covering the sample. Reactant gases were
introduced into the cell from a glass gas-handling manifold
through a stainless steel needle valve coated with halocarbon
wax (Halocarbon Products, series 1500). Salt powders, contained
in one of two shallow, cylindrical Teflon holders with diameters
of 3.1 and 4.9 cm, respectively, were placed on the sample
platform prior to system pump-down. The Knudsen cell was
interfaced to a quadrupole mass spectrometer (ABB Extrel, 150-
QC) through one of two removable glass apertures with hole
diameters of 3.9 and 5.8 mm, respectively.
Figure 1. Image of coarse NaCl crystals at 120× magnification. (Figure
Ion current signals were obtained by phase sensitive detection
and processed with a lock-in amplifier (EG&G, model 5209)
interfaced to a PC with data acquisition software (ABB Extrel,
Merlin, rev. B). The species monitored by their parent ions
included Cl2 (m/z 70, 72, and 74), HOCl (m/z 52 and 54), HCl
(m/z 36 and 38), and H2O (m/z 18). In most cases, the parent
ion with the most abundant isotopic composition was monitored
after the species was identified by confirming the expected signal
is reproduced here at 50% of its original size.)
+
ratios. Chlorine nitrate was monitored by its fragments NO2
(m/z 46), ClO+ (m/z 51), and NO+ (m/z 30). Uptake measure-
ments were based on the NO2 fragment (m/z 46) because it
+
had the highest intensity and consequently the highest signal-
to-noise ratio.
In the Knudsen cell, reactant gas molecules exit through an
aperture into the mass spectrometer or are lost by uptake on
the reactive surface. A net uptake coefficient, γnet, can be
determined from the ratio of the loss through the exit aperture
to the loss on the surface. In our experiments, the measured
uptake coefficient was derived from the relationship
(I0 - Ir) Ahole
γnet
)
(I)
(
)
Ir
Asurf
Figure 2. SEM image of fine NaCl crystals.
Here, I0 is equal to the steady-state reactant signal at the mass
spectrometer with the surface covered, and Ir is equal to the
reactant signal when the reactive surface is exposed. Ahole is
the area of the exit aperture, and Asurf is the area of the reactive
surface. In the initial data analysis, γnet was calculated using
but passed through it with vigorous agitation; that is, these
particles were just below the size that the no. 40 sieve could
retain. Optical microscopy showed that these particles were
cubic with sides of 300 µm on average. These samples were
designated subcoarse.
the geometric surface area of the sample holder for Asurf
;
Fine salts were those that resulted either from extensive
grinding by mortar and pestle or by mechanical grinding in a
Wig-L-Bug amalgamator (Crescent Dental, model 3110-3A) for
5 min. In an investigation of the reaction of NO2 with NaCl
and SSS ground using the Wig-L-Bug, Langer et al.19 deter-
mined that fine NaCl powders had grain sizes between 1 and
10 µm and had a surface area of 2.9 × 103 cm2/250 mg pellet.
Assuming this powder consists of monodisperse nonporous
cubes, each grain will have sides of 2.4 µm. They also showed
that SSS powders generated by the Wig-L-Bug had grain sizes
between 0.1 and 5 µm and a surface area of 2.4 × 104 cm2/250
mg pellet. Again, assuming monodisperse, nonporous cubes,
the sides will be 0.3 µm.
however, as discussed in detail below, the true reactive surface
area was taken into account in the calculation of the final
reaction probabilities.
The salts in these studies were obtained commercially: NaCl
(Bicron, single crystals; or Fluka, powder), and synthetic sea
salt (Aquarium Systems, Instant Ocean). Salts were prepared
to yield two average grain sizes, fine and coarse. Coarse NaCl
powders were generated by sieving commercially obtained
powders or large single crystals (cuttings) that were ground by
mortar and pestle. The fraction that passed through a no. 20
sieve but was retained by a no. 40 sieve was designated coarse.
On the basis of the sieve sizes, these crystals should have
average diameters between 425 and 850 µm;18 however, they
were examined using an optical microscope and found to be
monodisperse cubic particles with average sides of 400 µm
(Figure 1). SSS was sieved in the same manner to obtain
particles of similar size. One experiment was performed using
coarse NaCl that had been initially retained by the no. 40 sieve
The NaCl finely ground by hand was shown by SEM to have
a roughly bimodal distribution of average diameters of ∼10 and
<1 µm. Figure 2 shows an SEM image of this fine NaCl. On
the basis of an estimate that there are <100 small particles for
every 10 µm particle, the surface area of the large particles will

5180 J. Phys. Chem. A, Vol. 105, No. 21, 2001
Gebel and Finlayson-Pitts
Figure 3. Uptake at 298 K of ClONO2 (left axis) on 0.28 g (31 layers) of undried, unheated, fine NaCl and production of Cl2 (right axis) and HOCl
(right axis). The initial ClONO2 concentration was 5.9 × 1012 molecules cm-3. The signal at m/e 36 (left axis) is shown to demonstrate that there
was no significant production of HCl. The diameter of the sample holder is 3.1 cm, and the aperture diameter is 5.8 mm.
account for >90% of the total surface area of the particles, and
therefore fine salts were designated as having average diameters
of 10 µm.
amounts of HF could be removed by warming the trap to room
temperature and allowing the HF to react with the glass, forming
SiF4 that could be pumped off of the collection trap after it
was recooled to -115 °C.
The number of salt layers for a given amount of salt in the
sample holder was calculated as (m/FbAh), where m is the mass
(g) of the salt sample, Fb is the bulk density, A is the geometric
area of the sample holder, and h is the height of one salt particle.
The bulk density is the density of the unpacked powder in the
sample holder. It was calculated from the measured mass that
filled the known volume of the sample holder. For NaCl, Fb )
1.2 g cm-3 for both grain sizes, and for SSS, Fb ) 1.2 g cm-3
for fine particles and Fb ) 1.0 g cm-3 for coarse particles. The
difference in the bulk densities for fine compared to coarse SSS
is due to the tendency of coarse SSS to form clumps that do
not stack as efficiently. After grinding, salts were either
immediately placed in the Knudsen cell or dried in an oven
(100 °C) overnight. These dried salts, however, were exposed
to ambient air for the short length of time required to weigh a
sample and place it within the Knudsen cell, during which time
water can be taken up from room air. The salts were pumped
on while uncovered in the Knudsen cell prior to reaction. Some
salt samples were heated during this pump down time by
maintaining the sample platform temperature at 75° C in order
to reduce the amount of water adsorbed to the powder.
ClONO2 was synthesized by the reaction of HNO3 with ClF
(giving ClONO2 and HF) using a method similar to that
described by Schack.20 Dry nitric acid was condensed in a glass
trap by successive freezing of the vapor over a mixture of
concentrated sulfuric and nitric acids (2:1 by volume). The trap
was held at -115 °C using an ethanol bath and exposed to a
molar equivalent of ClF. The mixture was allowed to warm to
-70 °C for a few minutes to maximize the extent of reaction.
Then the trap was brought back down to -115 °C, and any
volatile byproducts were pumped off. Distillation of the ClONO2
was performed by holding the reaction trap at -85 °C and the
collection trap at -105 °C and gently pumping from the reaction
trap into the collection trap. Brownish yellow chlorine nitrate
liquid condensed in the collection trap over a period of minutes.
At least 50% yield of ClONO2 could be collected without any
significant HF impurity condensed in the collection trap. Small
Concentrated nitric acid (Fisher, 69.7 wt %) and sulfuric acid
(Fisher, 96.0 wt %), Cl2 (Matheson, high purity, 99.5%), HCl
gas (Scott Specialty Gases, 99.995%), and ClF (Ozark-Mahon-
ing Co.) were obtained commercially.
Results
ClONO2 Reaction with NaCl Powders. Figure 3 shows the
loss of chlorine nitrate and the production of Cl2 in a typical
experiment using fine NaCl powder that had not been dried or
heated prior to exposure. A time-dependent uptake behavior is
clearly seen, particularly during the first ∼50 s. This uptake
behavior is similar to that observed in previous studies with
this system using HNO3 or SO2 as the reactant,17,21,22 which
was interpreted as due to initial uptake into water on the salt.
Following the high initial uptake is a period of relatively stable,
slowly declining, uptake.
Molecular chlorine was the major gaseous product observed;
as seen in Figure 3, its production was anticorrelated with the
loss of the reactant ClONO2. Small amounts of HOCl, indicated
by the signal at m/z 52, were produced initially but declined
toward zero when the uptake of ClONO2 appeared to reach a
steady-state value. There was no evidence of production of
gaseous HCl.
In several experiments, salts were heated for ∼2 h while in
the Knudsen cell before reaction. The loss of ClONO2 and
release of Cl2 are shown for one of these experiments using
dried, coarse, heated NaCl in Figure 4. The signal at m/z 52
due to HOCl was very small throughout the experiment, and
there was no significant HCl production. For these samples,
the difference between the initial uptake and the subsequent
uptake was much less severe than for the finer, unheated salts.
As shown in Table 1, the average yield of Cl2, ∆[Cl2]/
∆[ClONO2], was measured in five experiments to be 81 ( 21%
(2σ). The reported error is based on the calibrations for Cl2 and
ClONO2 and is dominated by the relative error in ∆[ClONO2].

Uptake and Reaction of ClONO2 on Salts
J. Phys. Chem. A, Vol. 105, No. 21, 2001 5181
Figure 4. Uptake of ClONO2 on 3.1 g (9 layers) of dried, coarse, heated NaCl. The initial ClONO2 concentration was 4.6 × 1012 molecules cm-3
.
The aperture is 3.9 mm, and the diameter of sample holder is 3.1 cm.
TABLE 1: Summary of Uptake Experiments and Generation of Cl2 for ClONO2 on NaCl
size and
pretreatment
of salta
sample
holder
no. of aperture
av particle
total
particle diameter initial [ClONO2]
Cl2 yield
(%) ((2σ)
expt
1
2a fine
2b fine
3
4
5
6
7
8
9
sizeb (µm) mass (g) area (cm2) layersc
(mm) (molecules cm-3
)
measured γinit
measured γt
fine
10
10
10
10
10
10
10
400
400
10
10
10
400
400
400
400
10
0.843
1.425
1.426
0.394
0.907
1.341
0.281
2.913
7.657
7.614
5.062
9.359
1.413
1.338
3.257
3.133
0.788
2.982
7.54
7.54
7.54
7.54
7.54
7.54
7.54
7.54
93
158
158
44
100
148
31
8
8
336
223
413
3.9
0.11
0.13
f
0.035
0.056
0.025
0.049
0.042
0.050
0.050
0.030
0.038
0.049
0.039
0.038
0.035
0.031
0.033
0.029
0.043
0.033
3.9
3.9
3.9
3.9
5.8
5.8
3.9
3.9
5.8
5.8
5.8
3.9
5.8
3.9
3.9
3.9
3.9
1.4 × 1013
3.2 × 1013
1.4 × 1013
1.1 × 1013
5.5 × 1012
5.9 × 1012
2.9 × 1012
3.3 × 1012
1.2 × 1013
3.3 × 1012
1.3 × 1013
1.1 × 1013
8.8 × 1012
2.3 × 1012
4.6 × 1012
8.2 × 1012
2.4 × 1013
fine
fine
fine
fine
0.15
0.15
0.10
0.18
0.07
0.17
0.14
0.26
h
0.06
>0.07
h
h
h
86 ( 68
78 ( 40
coarse, dried
coarse, dried
fine
18.9
93 ( 23
18.9
18.9
18.9
7.54
7.54
18.9
7.54
7.54
7.54
10 fine
11 fine
12 coarse
3.9
13 coarse, dried
14 coarse, dried, heated
15 coarse, dried, heated
16 fine, dried, heated
17 fine, heated
3.7
3.6
8.7
87
329
83 ( 30
65 ( 57
10
h
averagei ((2σ):
0.14 ( 0.11
0.039 ( 0.018 81 ( 21
18 subcoarse
19 coarse
20 coarse
300
400
400
400
2500
10
0.211
0.101
0.0895
0.0594
2.650
0.921
7.54
7.54
7.54
7.54
7.54
7.54
0.8
5.8
5.8
5.8
5.8
5.8
1.3
6.5 × 1011
2.0 × 1012
1.3 × 1012
1.6 × 1012
5.9 × 1012
3.9 × 1013
0.13
0.24
0.02
0.007
0.17
d
0.019
0.0072
0.0063
0.0024
0.021
0.3
0.2
0.2
1.2
21 coarse
22 large crystalsg
23 fine
102
g0.020e
a See text for a description of the method of preparation of each type of sample. b Crystal size was determined by microscopy (see text). c Determined
using the bulk density and average crystal height; see text for method of calculation. d Ir was reduced to baseline level making measurement of γinit
impossible. e Represents a lower limit to γ. f This salt was previously exposed to ClONO2. g These crystals were large enough to measure by hand
with calipers. h The initial uptake for these samples showed little enhancement relative to the longer term uptake. i Average (( 2σ) for γinit was
calculated for multilayer experiments with measured values only.
The small amounts of HOCl produced were not quantified, but
from the weak signals (Figures 3 and 4) and high yields of Cl2,
HOCl clearly must be a minor product.
The uptake of ClONO2 was measured using samples of
various sizes, numbers of salt layers in the sample holder, and
methods of sample treatment (e.g., heating and pumping).
Although most experiments were carried out with multilayer
samples, five involved the use of less than a monolayer of
crystals, the size of which had been measured using optical
microscopy or, in one case, manually using calipers. The results
for multi- versus sub-monolayer coverage of the sample holder
are presented in the following sections; the implications for
diffusion and reaction of ClONO2 beyond the top layer of thick
samples are treated in detail subsequently.
Effects of Multiple Salt Layers on Uptake. Table 1 shows
the measured uptake coefficients calculated by assuming Asurf

5182 J. Phys. Chem. A, Vol. 105, No. 21, 2001
Gebel and Finlayson-Pitts
Figure 6. Uptake at 298 K of ClONO2 on 0.39 g (43 layers) of dried
but unheated, fine SSS, and production of Cl2. The initial ClONO2
concentration was 8.6 × 1012 molecules cm-3. The diameter of the
sample holder is 3.1 cm, and the aperture diameter is 5.8 mm.
Figure 5. Measured γt as a function of the number of particle layers
for coarse NaCl samples.
Unlike the experiments with NaCl, SSS samples would
routinely degas quantities of water when the lid was uncovered
sufficiently to measurably raise the total pressure to between 1
and 9 mTorr in the Knudsen cell during the first seconds of
uptake. If released in an instantaneous burst, this would be
equivalent to ∼(0.2-2) × 1017 molecules of water released from
salt samples ranging from 0.3 to 2.1 g in mass; a direct
correlation between sample mass and pressure increase cannot
be made as the burst is more dependent on drying conditions.
is the geometric area of the sample holder. Two different values
are given for most experiments: (1) that for the initial rapid
uptake, γinit, and (2) that at slightly longer times where the
uptake coefficient is smaller and changes less rapidly, γt.
Generally, for multiple layers of salts that are expected to hold
significant amounts of water, for example, fine, unheated
samples,17,19,21 a high initial uptake is observed, with γinit in
the range of ∼0.1-0.3. For these samples, γt falls in the range
from 0.025 to 0.056. The value of γt fell slowly with time. For
example, the measurement of γt ) 0.025 for experiment 2b was
performed on the salt sample used in experiment 2a (γt ) 0.056)
in which it had been repeatedly exposed to ClONO2. With these
multilayer samples, the value of γt was, within experimental
error, independent of the number of layers of the salt.
Discussion
These experiments show that there are three regimes for
reaction: (1) a rapid initial uptake and reaction which is
controlled by water on the salt surface, (2) a smaller and slowly
declining uptake into the water layer which occurs over longer
periods of time after the initial rapid uptake, and (3) a slow
passivation of the surface by reaction. Reaction to generate Cl2
occurs in all regimes.
Monolayer and Sub-monolayer Samples. Five experiments
(18-22) were performed using coarse salt samples of mass and
size calculated to give approximately a monolayer or less than
a monolayer coverage of the salt on the sample holder. This
was confirmed by visual inspection; the individual salt crystals
could be seen to be distributed over the bottom of the sample
holder. Because the particle size in these experiments had been
well characterized, the total reactive surface area is known and
diffusion into, and reaction with, additional layers of the sample
is not relevant.
Rapid Initial Uptake. That the rapid initial uptake is
controlled by water on the salt surface is indicated by three
observations. First, this initial “spike” in the uptake is seen
primarily with finely ground salts that have not been heated
and pumped on prior to the reaction; such samples have been
shown in previous studies to hold significant amounts of water,
which control the reactivity.17 Second, this high initial uptake
is exacerbated in the case of reaction with SSS, which holds
much larger amounts of water than NaCl.19,21 Third, a spike in
the production of gaseous HOCl is observed initially (Figure
3), as expected from the hydrolysis, reaction 4 above, of
ClONO2.
Because the available surface area is less than the geometric
area of the sample holder used to calculate γt and γinit, the
measured values of the reaction probabilities are expected to
be lower than for the multilayer experiments. The data in Table
1 show that this is indeed the case, with γt ranging from 0.002
to 0.02. Figure 5 shows the values of γt as a function of the
number of particle layers for coarse NaCl samples, where the
particle sizes and shapes are relatively uniform and well
characterized. Whereas the measured uptake coefficient is
invariant with the number of particle layers for more than two
layers, it drops off at monolayer and sub-monolayer coverage.
The reaction of gaseous HNO3 with NaCl is known to
generate gaseous HCl:17,21,23-31
HNO3 + NaCl f HCl + NaNO3
(6)
ClONO2 Uptake on Synthetic Sea Salt Powders. Figure 6
shows the loss of ClONO2 and the production of Cl2 for a typical
experiment using SSS. The experiments performed on SSS are
summarized in Table 2. The values of γinit are again large,
ranging from ∼0.1 to 0.8, larger than those for NaCl.
In earlier studies of the reaction of gaseous HNO3 with NaCl
in this laboratory,17 we interpreted the production of HCl as
being due to the uptake of HNO3 into water on the salt, causing
the pH to fall until HCl degassed. Thus, the HNO3 generated
in the hydrolysis of ClONO2 will first acidify the water layer.
If sufficient acidification occurs, the generation of gas-phase
HCl is expected.17 However, as seen in Figure 3, in the present
experiments no significant amounts of HCl were observed. This
may reflect the fact that there is another mechanism besides
degassing of HCl for consuming acid in this system, the acid-
Cl2 was again the major gaseous reaction product observed.
The yields of Cl2 during the steady-state uptake were measured
for three experiments and are shown in Table 2. The average is
78 ( 13% (2σ), which is within experimental error of the yield
for the reaction with NaCl. The generation of small amounts of
HCl and HOCl was observed in some experiments.

Uptake and Reaction of ClONO2 on Salts
J. Phys. Chem. A, Vol. 105, No. 21, 2001 5183
TABLE 2: Summary of Uptake Experiments and Generation of Cl2 for ClONO2 on Synthetic Sea Salt (SSS)
size and
pretreatment av particle
sample
holder
no. of
aperture
total
particle diameter initial [ClONO2]
Cl2 yield
expt
of salt
sizeb (µm) mass (g) area (cm2) layersc
(mm)
(molecules cm-3
)
measured γinit measured γt (%) ((2σ)
24
25
26
27
28
29
30
31
32
33
fine
fine
coarse, dried
coarse, dried
coarse, dried
coarse, heated
coarse
fine, dried
fine, dried
fine, dried
10
10
0.578
2.135
0.983
1.200
1.170
1.273
1.048
0.335
0.388
0.295
7.54
7.54
7.54
7.54
7.54
7.54
7.54
7.54
7.54
7.54
64
236
3.3
4.0
3.9
4.3
3.5
37
43
33
5.8
5.8
5.8
3.9
3.9
3.9
3.9
5.8
5.8
5.8
4.9 × 1012
3.6 × 1012
5.2 × 1012
8.7 × 1012
1.7 × 1013
0.45
0.28
0.37
0.22
0.76
0.12
0.11
0.60
0.44
0.80
0.08
0.06
0.15
0.14
0.16
0.09
0.05
0.32
0.16
0.36
400
400
400
400
400
10
7.8 × 1012
8.7 × 1012
9.7 × 1012
74 ( 11
74 ( 12
85 ( 15
10
10
average ((2σ):
0.42 ( 0.46
0.16 ( 0.20
78 ( 13
a See text for a description of the method of preparation of each type of sample. b Crystal size was determined by microscopy (see text). c Determined
using the bulk density and average crystal height; see text for method of calculation.
catalyzed reaction of HOCl with chloride ions:32
HOCl + Cl- + H+ f Cl2 + H2O
(7)
Thus, HCl may be consumed in the secondary reaction 7, and
the water layer may not become sufficiently acidic to expel HCl
into the gas phase.
If the initial uptake of ClONO2 on NaCl and SSS is due to
uptake into, and reaction in, water on the salt surface, it may
be described at least semiquantitatively in terms of a liquid
resistance model often applied to the uptake of gases into
liquids.10,33,34 This behavior was observed previously using this
system to study the uptake of SO2 on SSS.22 Given that we are
dealing in the present case with small amounts of water on solid
salts, rather than pure liquids, such an approach is semiquan-
titative at best. However, it supports interpretation of the rapid
initial uptake as being due to uptake into a liquid-like surface.
According to this model, if diffusion of the gas to the surface
is fast, the net uptake coefficient is approximated by
Figure 7. 1/γobs vs t1/2 for undried, unheated, fine samples of NaCl
(upper plot) and SSS (lower plot).
for comparison to this value; however, the recommended value36
of the reaction probability for ClONO2 on ice over the range
180-200 K is 0.3, similar to the initial uptake coefficient on
water at room temperature measured here.
If the subsequent decrease in the uptake after the very high
initial values is due to the surface water becoming saturated
with ClONO2, 1/γ should show a dependence on the square
root of time. As seen in Figure 7, the time dependence of the
uptake on NaCl and SSS samples that were not dried or heated
does indeed follow a t1/2 dependence in the initial stages of the
reactions.
1
γnet
1
R
1
)
+
(II)
Γsol + Γrxn
Here, R is the mass accommodation coefficient and the solubility
and reaction terms are given by eqs III and IV, respectively:
This effect of water on the uptake and reaction of ClONO2
is not surprising, given similar effects on the reaction of HNO3
with NaCl and sea salt. For example, Beichert and Finlayson-
Pitts17 described the reactive uptake of HNO3 as effectively
uptake into, and reaction in, an aqueous phase. De Haan and
Finlayson-Pitts21 reported even larger amounts of water on SSS
powder and hence chemistry similar to that for NaCl. Allen et
al.29 showed that water exposure after the reaction of HNO3
with NaCl could effectively reactivate the surface. Davies and
Cox26 reported that the uptake of HNO3 increased with
increasing water vapor pressure and proposed a surface mech-
anism involving water. Ghosal and Hemminger30 subsequently
showed their data and those of Davies and Cox are both26
consistent with a water-induced mobilization of the NaNO3
product on the surface.
4H*RT
Γsol
Γrxn
)
)
D /πt
(III)
(IV)
x
x
l
cj
4H*RT
cj
D k
l
rxn
H* is the effective Henry’s law coefficient; R is the gas constant;
T is the temperature; cj is the mean speed of the gas (for ClONO2
at 298 K, 2.5 × 104 cm s-1); Dl is the liquid-phase diffusion
coefficient (for ClONO2 in water at 298 K, Dl is estimated35 to
be on the order of 1 × 10-5 cm2 s-1); and krxn (s-1) is a pseudo-
first-order reaction rate to account for reaction in the bulk
aqueous phase, which affects the net uptake. The behavior of
the ClONO2 upon first exposing the salt (t f 0) is controlled
by physical transport across the gas-liquid interface, that is,
tf0
net
In summary, the rapid initial uptake of ClONO2 both on NaCl
and on SSS is controlled by water on the surface of the salts.
Both the production of HOCl from the reaction with undried
NaCl during this initial uptake and the t-1/2 dependence of γinit
are consistent with the role of water in this initial uptake. Note,
however, that during the initial rapid uptake, Cl2 is also being
generated. Hence, although water is a controlling factor, a
γ
) γinit f R. The values of the uptake coefficient
measured immediately after exposure, γinit, for the multiple
layers of salts that showed distinct time-dependent behavior are
seen in Table 1 to be in the range from 0.06 to 0.26 for NaCl
and in Table 2 to be in the range from 0.11 to 0.80 for SSS.
The mass accommodation coefficient, R, for ClONO2 on water
does not appear to have been measured directly in other studies

5184 J. Phys. Chem. A, Vol. 105, No. 21, 2001
Gebel and Finlayson-Pitts
TABLE 3: Summary of Experiments at Monolayer and Sub-monolayers of Salt
particle
size (µm)
total
mass (g)
sample holder
area (cm2)
no. of
particle layers
total surface
measured
corrected for
expt
areaa Aspec (cm2)
steady-stateb γobs
true surface areac γtrue
18
19
20
21
22
300
400
400
400
2500
0.2107
0.1011
0.0895
0.0594
2.6501
7.54
7.54
7.54
7.54
7.54
0.8
0.3
0.2
0.2
1.2
19.5
7.0
6.2
4.1
29.4
0.019
0.0074
0.0075
0.0077
0.0044
0.0054
0.0072
0.0063
0.0024
0.021
average ((2σ):
0.0065 ( 0.0030
a Calculated assuming monodisperse cubic crystals (see text). b Measured as described in text using the geometric surface area, Asurf ) 7.54 cm2.
c Calculated using Asurf ) total surface area in column 6.
combination of dissolution and reaction of ClONO2 must be
occurring even in this initial stage.
question. Table 3 summarizes the data for the experiments at
approximately a monolayer or less where the available surface
area, Aspec, is well-defined. Aspec can be unambiguously deter-
mined for samples consisting of a single species with known
particle size and geometry. Previous studies12,19,37 have estab-
lished that the salts are nonporous, that is, their BET surface
area is effectively given by the surface area of the crystal faces,
so the total surface area for these samples is Aspec ) 6d2N, where
d is the length of a side and N is the total number of particles
(assuming that the portion of the crystals making contact with
the sample holder is also available for reaction). Values for γt
Uptake at Longer Times. The slower and continuously
declining uptake of ClONO2 must also involve water on the
surface. For example, the observation that the steady-state uptake
can continue for several hundred seconds or more suggests that
this uptake does not occur by a simple gas-solid interaction
mechanism. If ClONO2 reacted with a dry, solid NaCl surface,
the surface would quickly become covered with the solid
reaction product, NaNO3. At 0.1 mTorr, ClONO2 has a collision
rate of 1.5 × 1017 s-1 with a 7.54 cm2 NaCl geometric surface
area. Assuming γ ) 0.005,12 the time required for a number of
reactiVe collisions equal to the number of atoms on the surface,
∼5 × 1015 surface chloride (based on 6.4 × 1014 Cl atoms
cm-2), is ∼6 s. However, our experiments show the uptake
occurs for more than several hundred seconds, even with much
higher chlorine nitrate concentrations, without showing signifi-
cant signs of surface deactivation. Clearly, the uptake of
ClONO2 on NaCl (and SSS) powders is not a simple gas-solid
interaction. This is not surprising considering that there is water
associated with the powders and that numerous previous studies
have indicated that such water affects the uptake of other
reactive gases on NaCl (and SSS) powder.14,17,19,21,26-30,37-39
If this reaction system was truly uptake of ClONO2 into a
bulk aqueous phase as described by eqs II-IV, the net observed
uptake, γnet, would approach a constant value. At this point,
the solution would be saturated with ClONO2 and its time-
independent uptake would be due to removal in the liquid phase
by reaction, eq IV. However, in the present studies γnet was
observed to slowly decline with time in the second regime
following the rapid, initial uptake. This is not surprising given
that there is no bulk aqueous phase but, rather, small amounts
of water on some as yet unknown sites on the solid salt surfaces.
The continuing slow decline of γnet with time may be due to
passivation of the salt surface as the reaction proceeds and forms
nitrate on the surface. Thus, the analogy with uptake into bulk
aqueous solutions represented by eqs II-IV is primarily useful
in providing a semiquantitative approach to interpreting the
initial rapid uptake when γnet approaches R.
calculated using the specific surface area of the sample, Aspec
,
are given in Table 3 as γtrue. It can be seen that values of the
uptake coefficient for these experiments calculated using the
known surface area are significantly smaller than the values
measured in the experiments using multiple layers of salt, with
an average corrected value of (6.5 ( 3.0) × 10-3 (2σ). This
suggests that although the measured, larger, average value of
γt of 0.039 ( 0.018 for the multilayer experiments is invariant
with the number of layers from 4 to 413, it must reflect the
availability of more than the top layer of particles.
The packing of the salt into the sample holder leaves the top
surface structure quite porous. Knowing the total volume of the
sample holder when filled and the salt mass and solid density,
the porosity of the surface can be calculated, that is, the fraction
of the geometric surface area that is open. For 500 µm particles,
the porosity is 0.35; that is, for every two salt particles on the
surface, there is an opening equivalent in size to one salt particle.
Given this porosity, it is reasonable to assume that all six sides
of the crystals in the top layer are available for reaction, rather
than just one as would be the case for a smooth surface of zero
porosity.
To probe how many layers of salt are involved in the ClONO2
reaction, we applied the model developed by Keyser, Moore,
and Leu15,16 (KML model). According to the KML model, the
true uptake coefficient, γtrue, is related to the measured value,
γobs, in the following manner
Ae + ηAi
γobs ) γtrue
(VI)
(
)
Ageo
The measured values of the uptake coefficients are based on
the assumption that only the geometric surface area of the
sample is available for reaction, that is, that there is no
significant diffusion of ClONO2 into the salt when multiple
layers are present. For both the fine and coarse NaCl with more
than approximately four layers of salt, there is no significant
dependence of the measured uptake coefficient on the number
of layers (Figure 5), showing that the ClONO2 does not probe
the salt below these top four layers. However, whether it
penetrates into the top two or three layers cannot be determined
from these multilayer experiments.
where Ae is the external surface area (i.e., top of the surface
layer), Ai is the internal surface area of the sample, Ageo is the
geometric surface area of the sample holder, and η is an
effectiveness factor. The effectiveness factor is the fraction of
the internal surface area that is available to the reactant gas that
diffuses into the pores of the sample. That is, there is a finite
depth of the sample that the reactant gas will be able to penetrate
during exposure. A large effectiveness factor indicates that the
gas can penetrate through many internal particle layers and thus
be exposed to a higher effective surface area during the uptake
period. In the limit of very slow reactions, the gas can probe
The data obtained using coarse NaCl powders at coverages
at and below a monolayer provide some insight into this

Uptake and Reaction of ClONO2 on Salts
J. Phys. Chem. A, Vol. 105, No. 21, 2001 5185
TABLE 4: Estimated Effectiveness Factors for Multilayer Experiments with NaCl
size and
pretreatment
of NaCl
no. of
particle
layers
sample
holder
no. of particle
layers probed
by ClONO2
particle
size (µm)
effectiveness
expt
area (cm2)
measured γt
factora η
1
fine
fine
fine
fine
fine
fine
fine
coarse, dried
coarse, dried
fine
fine
fine
coarse
coarse, dried
coarse, dried, heated
coarse, dried, heated
fine, dried, heated
fine, heated
10
10
10
10
10
10
10
400
400
10
10
10
400
400
400
400
10
93
157
157
44
100
148
31
7.54
7.54
7.54
7.54
7.54
7.54
7.54
7.54
18.9
18.9
18.9
18.9
7.54
7.54
18.9
7.54
7.54
7.54
0.035
0.056
0.025
0.049
0.042
0.050
0.050
0.030
0.038
0.049
0.039
0.038
0.035
0.031
0.033
0.029
0.043
0.033
0.007
0.01
1.6
2.6
1.2
2.3
2.0
2.3
2.4
1.5
1.9
2.3
1.8
1.8
1.8
1.6
1.7
1.4
2.0
1.5
2a
2b
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
0.001
0.030
0.010
0.009
0.044
0.056
0.103
0.004
0.004
0.002
0.216
0.163
0.206
0.045
0.012
0.002
8
8
336
223
413
3.9
3.7
3.6
8.7
87
10
329
average:
1.9 ( 0.8
a Calculated using eq VII and γtrue ) 6.5 × 10-3 from the monolayer and sub-monolayer experiments.
all of the salt sample surface and the BET surface area of the
salt is the appropriate reactive area. A small effectiveness factor
reflects the fact that the reaction is sufficiently fast that the gas
reacts at the surface before being able to penetrate into the salt.
As η f 0 and Ae f Ageo, then γobs f γtrue. When γobs > 0.1,
corrections to the measured uptake coefficient for increased
available surface area beyond the geometric area are usually
ignored.15
the final reported value of (4.6 ( 3.0) × 10-3 (2σ). Our value
based on the monolayer and sub-monolayer experiments in
which the available reactive surface areas are reasonably well-
known, γtrue ) (6.5 ( 3.0) × 10-3 (2σ), is within experimental
error of their value corrected for surface area.
Rossi and co-workers13,14 used a Knudsen cell similar to that
in the present studies and reported values for the uptake
coefficient of 0.23 ( 0.06 and 0.10 ( 0.05, respectively, in
reasonably good agreement with each other. They did not correct
for diffusion into lower layers of the salt because the measured
uptake coefficients were independent of the nature of the salt
sample, that is, ground powders and grains as well as single
crystals and thin films where diffusion into lower layers is not
possible. They measured these values using two different
experimental methods. In the steady-state approach, the uptake
coefficients were determined from the drop in the reactant signal
when the plunger covering the salt sample was lifted. In a second
approach, a pulse of gas was introduced into the cell and its
decay followed with time, out to ∼2 s; this was done both with
the salt covered and with it exposed to the gas. The exponential
decays of ClONO2 were fit to obtain the additional uptake due
to reaction with the salt when it was exposed. In these pulsed
valve experiments where the fit is over approximately the first
second of the reaction, the uptake would likely reflect predomi-
nantly the initial uptake which in our experiments is larger than
the steady-state values (Table 1). The values they obtained in
the pulsed experiments were in excellent agreement with those
from the steady-state approach, which occur over much longer
periods of time. However, if the steady-state values were
calculated on the basis of the first opening of the lid, these would
also correspond to our values of γinit. Our measured values for
Equation VI is equivalent16,40 to
γobs ) γtrueFbSBET(he + ηhi)
(VII)
where Fb is the bulk density, SBET is the BET specific surface
area (cm2 g-1), he is the height of the first layer, and hi is the
height of all the internal layers. The effective number of layers
probed by ClONO2 in the multilayer experiments can be
estimated using γtrue ) (6.5 ( 3.0) × 10-3 from the sub-
monolayer experiments. Knowing the particle size, the height
of particles in the top layer (he) and of the internal layers (hi),
the bulk density (Fb) and the BET surface area SBET, the
effectiveness factor, η, can be calculated from the measured
values of γobs. In previous studies, the salts were shown by BET
measurements to be nonporous,12,19,37 and hence the geometric
surface area of the crystals is also their BET surface area. Thus,
SBET can be calculated from the measured dimensions of the
cubic crystals and the true density of the salt. Table 4 shows
the values of the effectiveness factor calculated in this manner.
The number of particle layers probed will be equal to one (the
top layer) plus this effectiveness factor times the total number
of internal layers. On average, about two layers of NaCl react
with ClONO2.
Comparison with Previous Studies. Leu and co-workers12
used a flow tube holding salt samples on the bottom and
followed the loss of ClONO2 and formation of Cl2 with time.
The measured values of the uptake coefficient were in the range
of 0.009-0.07, which is similar in magnitude to the values
measured here using multiple salt layers. They argue that the
effective reactive surface area of the salt is larger than the top
geometric area due to the fact that diffusion and reaction of
ClONO2 in the subsurface layers can occur. To correct for this,
they applied their previously developed KML model15,16 to
correct their measured values of the uptake coefficients to obtain
γinit for the multilayer salt experiments ranged from 0.06 to 0.26,
with an average of 0.14 ( 0.11 (2σ), in excellent agreement
with the measurements of Rossi and co-workers.13,14
Both sets of previous studies also reported that Cl2 was
formed in 100% yield, within experimental error. Timonen et
al.12 reported HOCl formation when high water vapor (3 × 10-3
Torr) concentrations were present during the reaction. Aguzzi
and Rossi14 did not observe HOCl production in the reaction
with NaCl, but did when ClONO2 hydrolyzed on nonreactive
salt surfaces such as Na2SO4. These results are in agreement

5186 J. Phys. Chem. A, Vol. 105, No. 21, 2001
Gebel and Finlayson-Pitts
with the results presented here, where only small signals due
to HOCl were observed in some experiments at short reaction
times.
In short, the results presented here are in good agreement
with both sets of previous determinations of the uptake
coefficient for ClONO2 on NaCl if the two different reaction
regimes are separated; that is, there is a rapid initial uptake
represented by γinit , followed by a slower uptake represented
by γt. In addition, there is agreement that Cl2 is the major
gaseous product.
Determination of γ for ClONO2 on Synthetic Sea Salt.
The uptake coefficients for ClONO2 on SSS were almost always
>0.1. Keyser, Moore, and Leu15 show that under these condi-
tions of rapid uptake, diffusion into the salt is not important
because the gas reacts before it can diffuse into lower layers.
Under these conditions, the correction to the measured values
of the uptake coefficients is less than a factor of 3, which is on
the order of the scatter in the data (Table 2). In addition, any
such correction would be highly uncertain due to the hetero-
geneous chemical composition of individual salt crystals. We
have therefore not corrected our measured values of the uptake
coefficients for the reaction of ClONO2 with SSS for diffusion
into the salt.
initially in some experiments is consistent with its hydrolysis
(reaction 4). However, this reaction simultaneously produces
nitric acid, which acidifies the water. As a result, reaction 7
becomes more important, and the generation of HOCl in the
gas phase falls with time, as was observed. No induction time
was observed for the production of Cl2 within the limits of our
experiments. We cannot distinguish whether a direct reaction
of ClONO2 also occurs with chloride ions to generate Cl2 or
whether HOCl is an essential intermediate in the production of
Cl2 by reaction 4 followed by reaction 7. However, the fact
that HCl is a very minor product shows that the water layer
does not become sufficiently acidic to degas significant amounts
of HCl, as occurs in the direct HNO3-NaCl reaction.17,23-27,30
This suggests that there is a reaction or reactions that lead to
acid consumption, which is the case for reaction 7.
Atmospheric Implications. Using γ ) 6.5 × 10-3, the amount
of Cl2 generated by this reaction for a typical sea salt aerosol
concentration of 1 cm-3 can be estimated if we assume 2.5 µm
cubic NaCl particles and a background ClONO2 concentration
of 5 ppt.4,47 Under these conditions, 2 × 103 molecules of Cl2
are generated per second. This would lead to a concentration
of 3.5 ppt of Cl2 after 12 h of darkness. This estimate assumes
a source of ClONO2 that maintains a steady-state concentration
of 5 ppt despite the consumption of approximately that amount
in one night and that the steady-state kinetics characteristic of
“dry” NaCl is the appropriate value to use.
However, in the marine boundary layer, the relative humidity
is generally above the deliquescence point and, hence, the use
of the larger value of the initial uptake coefficient for sea salt
may be more appropriate. Using γinit ) 0.42 measured using
SSS, the rate of production of Cl2 could be as high as 1.3 ×
106 molecules cm-3 s-1, which would generate ∼200 ppt of
Cl2 in 12 h. This is clearly an overestimate because it again
assumes a constant concentration of ClONO2 and does not take
into account the possible production of HOCl. However, it is
noteworthy that this is of the same order of magnitude as
measurements of Cl2 in the marine boundary layer by Spicer
and co-workers,48 which peaked at ∼150 ppt, and of non-HCl
inorganic chlorine compounds by Keene, Pszenny, and co-
workers.49,50
The average values we measured here were γinit ) 0.42 (
0.46 and γt ) 0.16 ( 0.20, where the errors are the 2σ statistical
errors. Because a negative value of the uptake coefficient is
not physically possible, the lower error bounds must be adjusted,
giving γinit ) (0.42+0.46) and γt ) (0.16+0.20
-0.16). Thus, the uptake
-0.42
both initially and at longer reaction times is high and variable.
This is not surprising, given the large amounts of water
associated with SSS and the inhomogeneity of the crystal
composition. For example, Langer et al.19 and Weis and Ewing38
have shown that sea salt particles contain pockets of water in
microcrystalline and amorphous regions in addition to water
from crystalline hydrates such as MgCl2‚6H2O.
Reaction Mechanism. On the basis of the evidence that water
is strongly affecting the uptake of ClONO2 on both NaCl and
SSS, the mechanism of reaction is likely to be similar to that
proposed by Beichert and Finlayson-Pitts17 for the uptake and
reaction of HNO3 with NaCl. Thus, the ClONO2 is first taken
up into water on the surface and then reacts within a concen-
trated solution on the salt particle. The mechanism is expected
to be similar for NaCl and SSS, with the possible exception of
the presence of buffers such as CaCO3 in the case of SSS.
The reaction of ClONO2 with HCl on ice at low temperatures
(probably in a quasi-liquid layer on the ice surface) is
well-known with respect to polar stratospheric cloud
chemistry:10,36,41-44
Conclusions
The uptake coefficient for ClONO2 on NaCl powder is
initially large (>0.1) due to the effect of water adsorbed to the
surfaces of the powder. These results are consistent with the
high uptake observed by Rossi and co-workers13,14 measured
during initial exposure of the salt. The initial uptake follows a
t-1/2 dependence under “wet” conditions, as would be expected
due to the effect of solubility on the uptake of ClONO2 into
the water associated with the powders. The uptake of ClONO2
on SSS follows a similar time dependence but is even greater
than that on NaCl due to the effect of higher surface water,
with γinit ) (0.42+-00..4426).
At longer reaction times, there is a smaller uptake coefficient
for ClONO2 on NaCl powders with a measured value of 0.039
( 0.018 (2σ). Whereas this value does not depend on the
number of layers of salt particles over the range from 4 to 413,
experiments at monolayer and sub-monolayer coverages show
that approximately two particle layers react and that the true
steady-state uptake coefficient is (6.5 ( 3.0) × 10-3 (2σ). This
is consistent with the value of (4.3 ( 3.0) × 10-3 reported by
Leu and co-workers.12
ClONO2 + HCl
9
ice8 Cl2 + HNO3
(6)
The reaction is believed to involve an ionic mechanism, with
attack of the Cl- formed from dissociation of HCl on the
ClONO2.10,41-46 Reaction 6 may also occur via the initial
formation of HOCl formed by hydrolysis of ClONO2
ClONO2 + H2O f HOCl + HNO3
(4)
followed by the well-known acid-catalyzed reaction of HOCl
with the chloride ion:32
HOCl + Cl- + H+ f Cl2 + H2O
(7)
The chemistries of ClONO2 on both NaCl and SSS holding
water at room temperature are similar. The production of HOCl
The yield of Cl2 was measured to be close to unity for both
the NaCl and SSS reactions. In most cases, small amounts of

Uptake and Reaction of ClONO2 on Salts
J. Phys. Chem. A, Vol. 105, No. 21, 2001 5187
(22) Gebel, M. E.; Finlayson-Pitts, B. J.; Ganske, J. S. Geophys. Res.
Lett. 2000, 27, 887.
(23) Fenter, F. F.; Caloz, F.; Rossi, M. J. J. Phys. Chem. 1994, 98, 9801.
(24) Leu, M.-T.; Timonen, R. S.; Keyser, L. F.; Yung, Y. L. J. Phys.
Chem. 1995, 99, 13203.
(25) Fenter, F. F.; Caloz, F.; Rossi, M. J. J. Phys. Chem. 1996, 100,
1008.
(26) Davies, J. A.; Cox, R. A. J. Phys. Chem. A 1998, 102, 7631.
(27) Laux, J. M.; Hemminger, J. C.; Finlayson-Pitts, B. J. Geophys. Res.
Lett. 1994, 21, 1623.
HOCl were observed, particularly at short reaction times,
suggesting that hydrolysis of ClONO2 also occurs. It may be
that the mechanism of these reactions involves hydrolysis of
ClONO2 to generate HOCl, followed by the well-known acid-
catalyzed reaction with chloride ions to generate Cl2.
With these reaction probabilities, this reaction may be a
significant contributor to the observed48-50 levels of Cl2 at night
in the marine boundary layer.
(28) Laux, J. M.; Fister, T. F.; Finlayson-Pitts, B. J.; Hemminger, J. C.
J. Phys. Chem. 1996, 100, 19891.
Acknowledgment. We thank the National Science Founda-
tion and the Department of Energy for support of this research.
We also thank M. Zach and R. M. Penner for valuable
microscopy assistance, J. C. Hemminger, M.-T. Leu, M. J. Rossi
and J. N. Pitts, Jr., for helpful discussions, and M. Rossi and
V. Grassian for providing preprints prior to publication.
(29) Allen, H. C.; Laux, J. M.; Vogt, R.; Finlayson-Pitts, B. J.;
Hemminger, J. C. J. Phys. Chem. 1996, 100, 6371.
(30) Ghosal, S.; Hemminger, J. C. J. Phys. Chem. A 1999, 103, 4777.
(31) Hemminger, J. C. Int. ReV. Phys. Chem. 1999, 18, 387.
(32) Wang, T. X.; Margerum, D. W. Inorg. Chem. 1994, 33, 1050.
(33) Kolb, C. E.; Worsnop, D. R.; Zahniser, M. S.; Davidovits, P.;
Keyser, L. F.; Leu, M.-T.; Molina, M. J.; Hanson, D. R.; Ravishankara, A.
R.; Williams, L. R.; Tolbert, M. A. Laboratory Studies of Atmospheric
Heterogeneous Chemistry. In Progress and Problems in Atmospheric
Chemistry; Barker, J. R., Ed.; World Scientific: Singapore, 1995; Vol. 3;
Chapter 18.
References and Notes
(1) Keene, W. C.; Pszenny, A. A. P.; Jacob, D. J.; Duce, R. A.;
Galloway, J. N.; Schultz-Tokos, J. J.; Sievering, H.; Boatman, J. F. Global
Biogeochem. Cycles 1990, 4, 407.
(2) Finlayson-Pitts, B. J. Res. Chem. Intermed. 1993, 19, 235.
(3) Keene, W. C. Inorganic Cl Cycling in the Marine Boundary
Layer: A Review. In Naturally-Produced Organohalogens; Grimvall, A.,
Leer, E. W. B. D., Eds.; Kluwer Academic Publishers: Dordrecht, The
Netherlands, 1995; p 363.
(4) Sander, R.; Crutzen, P. J. J. Geophys. Res. 1996, 101, 9121.
(5) Vogt, R.; Crutzen, P. J.; Sander, R. Nature 1996, 383, 327.
(6) Finlayson-Pitts, B. J.; J. N. Pitts, J. Science 1997, 276, 1045.
(7) Ravishankara, A. R. Science 1997, 276, 1058.
(8) Andreae, M. O.; Crutzen, P. J. Science 1997, 276, 1052.
(9) DeHaan, D. O.; Brauers, T.; Oums, K.; Stutz, J.; Nordmeyer, T.;
Finlayson-Pitts, B. J. Int. ReV. Phys. Chem. 1999, 18, 343.
(10) Finlayson-Pitts, B. J.; Pitts, J. Chemistry of the Upper and Lower
AtmospheresTheory, Experiments, and Applications; Academic Press: San
Diego, CA, 2000.
(34) Kolb, C. E.; Jayne, J. T.; Worsnop, D. R.; Davidovits, P. Pure
Appl. Chem. 1997, 69, 959.
(35) Houghton, G. J. Chem. Phys. 1964, 40, 1628.
(36) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Chemical Kinetics and Photochemical Data for Use in Stratospheric
Modeling, Evaluation No. 12; JPL Publication 97-4, NASA Jet Propulsion
Laboratory: Pasadena, CA, January 15, 1997.
(37) Vogt, R.; Finlayson-Pitts, B. J. J. Phys. Chem. 1994, 98, 3747.
(38) Weis, D. D.; Ewing, G. E. J. Phys. Chem. A 1999, 103, 4865.
(39) Weis, D. D.; Ewing, G. E. J. Geophys. Res. 1999, 104, 21.
(40) Underwood, G. M.; Li, P.; Usher, C. R.; Grassian, V. H. J. Phys.
Chem. A 1999, 104, 819.
(41) Molina, M. J.; Tso, T.-L.; Molina, L. T.; Wang, F. C.-Y. Science
1987, 238, 1253.
(11) Finlayson-Pitts, B. J.; Ezell, M. J.; Pitts, J. Nature 1989, 337, 241.
(12) Timonen, R. S.; Chu, L. T.; Leu, M.-T.; Keyser, L. F. J. Phys.
Chem. 1994, 98, 9509.
(13) Caloz, F.; Fenter, F. F.; Rossi, M. J. J. Phys. Chem. 1996, 100,
7494.
(14) Aguzzi, A.; Rossi, M. J. Phys. Chem. Chem. Phys. 1999, 1, 4337.
(15) Keyser, L. F.; Moore, S. B.; Leu, M.-T. J. Phys. Chem. 1991, 95,
5496.
(16) Keyser, K. F.; Leu, M.-T.; Moore, S. B. J. Phys. Chem. 1993, 97,
2800.
(17) Beichert, P.; Finlayson-Pitts, B. J. J. Phys. Chem. 1996, 100, 15218.
(18) Lide, D. R. Handbook of Chemistry and Physics, 74th ed.; CRC
Press: Boca Raton, FL, 1994.
(19) Langer, S.; Pemberton, R. S.; Finlayson-Pitts, B. J. J. Phys. Chem.
A 1997, 101, 1277.
(42) Tolbert, M. A.; Rossi, M. J.; Malhotra, R.; Golden, D. M. Science
1987, 238, 1258.
(43) Tolbert, M. A.; Rossi, M. J.; Golden, D. M. Geophys. Res. Lett.
1988, 15, 847.
(44) Leu, M.-T. Geophys. Res. Lett. 1988, 15, 17.
(45) Gilligan, J. J.; Castleman, A. W., Jr. J. Phys. Chem. A 2001, 105,
1028.
(46) Oppliger, R.; Allanic, A.; Rossi, M. J. J. Phys. Chem. A 1997, 101,
1903.
(47) Singh, H. B.; Kasting, J. F. J. Atmos. Chem. 1988, 7, 261.
(48) Spicer, C. W.; Chapman, E. G.; Finlayson-Pitts, B. J.; Plastridge,
R. A.; Hubbe, J. M.; Fast, J. D.; Berkowitz, C. M. Nature 1998, 394, 353.
(49) Keene, W. C.; Maben, J. R.; Pszenny, A. A. P.; Galloway, J. N.
EnViron. Sci. Technol. 1993, 27, 866.
(20) Schack, C. J. Inorg. Chem. 1967, 6, 1938.
(21) DeHaan, D. O.; Finlayson-Pitts, B. J. J. Phys. Chem. A 1997, 101,
9993.
(50) Pszenny, A. A. P.; Keene, W. C.; Jacob, D. J.; Fan, S.; Maben, J.
R.; Zetwo, M. P.; Springer-Young, M.; Galloway, J. N. Geophys. Res. Lett.
1993, 20, 699.