Kinetic mechanism of ClONO2 uptake on polycrystalline film of NaCl

Article abstract of DOI:10.1021/jp056272b

Kinetic studies and the mechanism determination of ClONO₂ uptake on polycrystalline NaCl were carried out using a coated-insert flow tube reactor combined with high-resolution, low-energy electron-impact mass spectrometer under the following conditions: p =1-2 Torr, linear flow velocity v = 3.5-75 m s^-1, T = 293 and 387 K, [ClONO₂] = (0.5-25) × 10¹² molecules cm^-3. The salt was deposited as a film from nonsaturated aqueous solution on the sliding rod. The temporal dependences of the uptake coefficient and the partial uptake coefficients leading to a formation of the prime Cl₂ and HOCl products were measured for different ClONO₂ concentrations. These dependences are established to be described by γ = γ₀ exp(-t/τ) + γ_s, γ_0.s^-1 = a_0.S + b_0.s[ClONO₂]. In the framework of the proposed kinetic model, the data are explained and the main elementary kinetic parameters of the uptake are evaluated. The model is based on a combination of Langmuir adsorption, formation of surface complexes on initial active sites, Z _ch, followed by their unimolecular decomposition. Decomposition is proposed to proceed concurrently in two channels, one of which is a released surface site that conserves the properties of the initial site. In the other channel, the initial Z_ch transforms into Z_ph followed by steady-state uptake and reproduction of final Z_ph. The model gives an analytical expression for experimental parameters γ₀, γ_s, and τ in terms of elementary rate constants and the reactant volume concentration. The final objective of the proposed model is the extrapolation of γ₀, γ_s, and r parameters to real tropospheric conditions.

Full text of DOI:10.1021/jp056272b

J. Phys. Chem. A 2006, 110, 6771-6780
6771
Kinetic Mechanism of ClONO2 Uptake on Polycrystalline Film of NaCl†
‡
‡
‡
§
,§
V. V. Zelenov, E. V. Aparina, S. A. Kashtanov, D. V. Shestakov, and Yu. M. Gershenzon*
Institute for Energy Problems of Chemical Physics, Russian Academy of Sciences, ChernogoloVka 142432,
Russia, and N. N. SemionoV Institute of Chemical Physics, Russian Academy of Sciences, Moscow 119991,
Russia
ReceiVed: October 31, 2005; In Final Form: March 2, 2006
2
Kinetic studies and the mechanism determination of ClONO uptake on polycrystalline NaCl were carried
out using a coated-insert flow tube reactor combined with high-resolution, low-energy electron-impact mass
-
1
spectrometer under the following conditions: p ) 1-2 Torr, linear flow velocity V ) 3.5-75 m s , T ) 293
1
2
-3
and 387 K, [ClONO
2
] ) (0.5-25) × 10 molecules cm . The salt was deposited as a film from nonsaturated
aqueous solution on the sliding rod. The temporal dependences of the uptake coefficient and the partial uptake
coefficients leading to a formation of the prime Cl
2
and HOCl products were measured for different ClONO
2
-
1
concentrations. These dependences are established to be described by γ ) γ
0
s
exp(-t/τ) + γ , γ0,s ) a0,s
+
2
b0,s[ClONO ]. In the framework of the proposed kinetic model, the data are explained and the main elementary
kinetic parameters of the uptake are evaluated. The model is based on a combination of Langmuir adsorption,
formation of surface complexes on initial active sites, Zch, followed by their unimolecular decomposition.
Decomposition is proposed to proceed concurrently in two channels, one of which is a released surface site
that conserves the properties of the initial site. In the other channel, the initial Zch transforms into Zph followed
by steady-state uptake and reproduction of final Zph. The model gives an analytical expression for experimental
parameters γ
0
, γ , and τ in terms of elementary rate constants and the reactant volume concentration. The
s
final objective of the proposed model is the extrapolation of γ
conditions.
0
, γ , and τ parameters to real tropospheric
s
1
. Introduction
Against this background of stratospherically focused work,
the fact of natural generation of halogens in the troposphere,
designated “halogen activation in the troposphere”, has been
Heterogeneous reactions are of interest in many fields of
science and industry. They are important constituents of chain
reactions, catalysis, catalytic combustion, etching, ore dressing,
and so on. In atmospheric chemistry, these reactions have been
of particular interest since 1985, when Farman et al. revealed a
significant depletion of ozone layer in the Antarctic atmosphere
in spring. As early as 1974, Molina and Rowland had postulated
a possibility of the stratospheric ozone destruction via a chain
reaction mechanism:
3
studied less intensively. In 1986, Bottenheim et al. as well as
4
Oltmans and Komhyr observed extensive destruction of arctic
troposphere ozone almost as soon as the stratospheric ozone
hole was revealed. Platt et al. made a great contribution by
achieving real-time measurements of active bromine (Br, BrO)
1
5
in the troposphere. They observed BrO radicals not only in
the troposphere in polar regions but also in boundary zone over
interior continental seas such as the Dead Sea and Caspian Sea.
The main precursors of halogens in the troposphere are sea-
salt aerosol particles that arise because of oceanic wave action
and saline land erosion.
2
3
Cl + O f ClO + O2
ClO + O f Cl + O2
Precursors of the chlorine atoms were proposed to be
anthropogenic emissions of chlorofluorocarbon compounds that
are stable in the troposphere and photolyzed in the stratosphere.
Usually, active bromine concentration in the troposphere is
higher than that of active chlorine despite the fact that the
abundance of bromide ions in seawater is 600 times lower than
that of the chloride ions. There are two reasons for this contra-
dictory situation. First, atomic chlorine produced in the gas phase
is much more reactive with respect to hydrocarbons (HC’s),
which diminishes its concentration. Second, in the absence of
1
2
The works of Farman and Molina and Rowland initiated
intensive field observations of the atmospheric constitution in
polar and middle latitude regions as well as numerous laboratory
studies. The ozone destruction proved to proceed at relatively
low altitudes in the presence of polar stratospheric clouds or in
the middle latitude regions in the presence of sulfuric aerosols.
Since 1985, a range of laboratory studies on the atmospheric
chemistry kinetics of low-temperature heterogeneous reactions
capable of generating active chlorine in polar stratospheric
clouds and on sulfuric aerosols in the Junge layer have been
published.
NO in marine troposphere, bromine concentration rises sharply
x
(“bromine explosion”) because of two-phase branching-chain
processes, which include small atmospheric constituents such
-
+
as Br, Br , HOBr, and HO as well as Br and H ions in the
water interface:
2
2
6
H++Br-
hν
+
2HO2
+
HOBr(aq)
8 Br (g)
9
8 2Br + 2O f 2BrO
8
2
3
†
‡
§
2HOBr(g) f 2HOBr(aq) (I)
Part of the special issue “David M. Golden Festschrift”.
Institute for Energy Problems of Chemical Physics.
N. N. Semionov Institute of Chemical Physics.
The branching rate for similar chlorine process is much
1
0.1021/jp056272b CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/04/2006

6772 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Zelenov et al.
lower because the reaction7
+
-
HOCl + H + Cl f H O + Cl ,
2
2
k ) 2.1 × 10 mole s-1
4
-2
is 10 times slower compared to8
6
+
-
HOBr + H + Br f H O + Br ,
2
2
k ) 2.3 × 10 mole s-1
1
0
-2
Figure 1. Schematic diagram of the coated-insert flow tube reactor.
(
1) Sampler, (2) reactor tube, (3) heating/cooling jacket, (4) sliding
rod with the saline coating, (5) supporting tube, (6) leading magnet,
7) compensation volume, (8) cryostat, (9) teflon ampule with ClONO
frozen.
In addition, the termination rate of the branching chain (I) is
the reaction of halogens with HC’s, which is higher in the case
of chlorine.
(
2
9
Nevertheless, Spicer et al. observed an abnormally high
9
-3
They measured γ ) 0.23 ( 0.06, which was independent of
chlorine concentration, [Cl2] > 10 particles cm , in the littoral
zone of Long Island near New York city, which is 25 times
higher than that of bromine measured under the same condi-
1
0
13
-3
both [ClONO2] (10 -10 molecules cm ) and the salt
substrate type. The only gas-phase product observed was Cl2
molecules. The Knudsen reactor coupled to a quadrupole mass
spectrometer was used in a number of additional studies.1
On the basis of direct measurement of the surface residence
time of ClONO2 on NaCl, the γ was determined to be of 0.1.19
To reproduce γ and to reveal a role of surface adsorbed water
1
0
tions. There is other evidence of abundant chlorine in the
troposphere, for example, observation of morning HC’s con-
9-20
1
1
sumption, that cannot be explained by the reactions of the
bromine-containing compounds because of their low reactivity.
This conclusion is confirmed by direct measurement of Cl2 +
1
0
12
-3
20
12
for [ClONO2] ) 10 -10 molecules cm , Aguzzi and Rossi
HOCl concentration varying in the range from 26 to 254 pptv.
1
8,19
obtained γ ) 0.1 ( 0.05 in agreement with previous results.
Gebel and Finlayson-Pitts were the first to observe two-steps
These observations have stimulated a search for mechanisms
that could be responsible for the tropospheric active chlorine
formation.
According to our concept, chlorine activation may be the
dominant halogen activation process in the MBL near industrial
regions where NOx concentration in the troposphere is high
enough, whereas the active bromine plays an important role in
remote locations. A simple two-phase branching-chain process
2
1
ClONO2 uptake, that is, initial uptake with γ0 = 0.1 followed
-
3
by steady-state interaction with γs ) (6.5 ( 3.0) × 10 for
12
13
-3
[
ClONO2] ) 10 -10 molecules cm and T ) 298 K. Before
these measurements, the NaCl powder was heated and pumped
out to delete surface-adsorbed water. The initial uptake was
interpreted as a diffusion of gas-phase reactant through the
liquidlike layer of adsorbed water with simultaneous reaction
is shown below, which provides active chlorine production in
Here, the heterogeneous
reaction of ClONO2 with NaCl plays a key role:
2
2
the MBL polluted with NOx.1
3-15
inside the layer. Hoffman et al. proposed a model for
estimating available surface areas to correct experimental uptake
2
1
coefficient. Application of this model to prior data gave a
-
2
2O3
2NO2
corrected γs ) (2.4 ( 1.2) × 10 .
NaCl
8 Cl2
hν
ClONO2
9
8 2Cl
8 2ClO
8 2ClONO ... (II)
2
In the present work, the kinetics of the ClONO2 consumption
and the product formation is studied to determine and define
the uptake coefficient in terms of rate constants for elementary
steps of reactive uptake. This approach enables us to reveal the
contribution of initial and steady-state steps in the total uptake,
to determine a characteristic transition time between them, and,
as a final objective, to extrapolate laboratory data to real
tropospheric conditions.
A number of authors have investigated the reaction of ClONO2
On the basis of the large dipole moment
of ClONO2 molecule, the reaction has been proposed to proceed
1
6-22
with solid NaCl.
1
6
via surface intermediate complex of two dipole particles:
2
. Experimental Section
Product Cl2 molecules were confirmed to be the main product
of the reaction.1
7-22
To diminish effects of coating porosity, thin films were
At the same time, there is a considerable
deposited on a quartz rod from nonsaturated aqueous solutions.
The roughness of the films was determined with a profilometer.
The experiments were carried out in a coated-insert flow tube
reactor coupled to a phase-sensitive high-resolution low-energy
electron-impact mass spectrometer. Ionization near the thresh-
old of the molecular ion formation allows the exclusion of
background from dissociative ionization and identification of
individual substances in a mixture of gases when their mass
spectra are similar. The last consideration is especially important
for the substances studied in the work.
difference in the ClONO2 uptake coefficients measured by
previous authors. The first quantitative measurement of ClONO2
1
7
uptake on NaCl was performed by Timonen et al., using a
flow reactor coupled to a quadrupole mass spectrometer at 296
23
8
11
-3
K and [ClONO2] ) 3 × 10 -3 × 10 molecules cm . The
reactive uptake coefficient, γ, was determined by monitoring
both ClONO2 consumption and Cl2 formation. For dry solid
-
3
NaCl, γ was determined to be (4.6 ( 3.0) × 10 at T ) 296
K. Additionally, a trace of HOCl was detected as a byproduct,
which could be formed by the reaction
The coated-insert flow tube reactor is shown schematically
in Figure 1. Helium carrier gas flow with the addition of the
ClONO2 molecules moves through the outer wide glass tube
(2), which is coated with a Teflon film. The thin quartz rod (4)
with a saline coating can be moved along the axis of the outer
tube to vary the exposure time of the salt to ClONO2. The quartz
rod is inserted into the reactor tube from a compensation volume
ClONO + (H O) f HOCl + (HNO )
3 ad
(IV)
2
2
ad
Caloz et al. investigated the same reaction, using a Teflon-
coated Knudsen reactor coupled to a quadrupole mass spec-
1
8
trometer. Uptake of ClONO2 was measured on powder, salt
grains, single-crystal surfaces, and thin deposited salt layers.

Kinetic Mechanism of ClONO2 Uptake
7) through the supporting tube (5). An additional helium flow
passes through the compensation volume to avoid uncontrolled
salt reaction by a diffusion flux of ClONO2 from the reactor
J. Phys. Chem. A, Vol. 110, No. 21, 2006 6773
(
2
stat
2
sys
σ )
σ
x
+ σ
To quantitatively evaluate a validity of the model description
of our experimental dependencies we have used the well-known
(2) into the compensation volume (7). A gas sampling into the
mass spectrometer is carried out through an orifice at the top
of the alumina inlet cone (1). The source of ClONO2 is a Teflon
ampule (9) filled with Teflon capillary pieces and frozen
ClONO2. The ampule is placed into a homemade cryostat (8).
The cryostat has a temperature range of 113-283 K with
accuracy of (0.01 K.
2
ø functional
exp
model 2
M
yj - yj
2
ø )
(2)
∑
(
)
j)1
σj
exp
exp
j
The ClONO2 absolute concentration in the reactor is calcu-
lated from eq 1
where yj is a value of experimental dependence y (x ), j )
j
model
1
,..M, and y
is the model approximation. At best, the value
j
2
of ø would be of the order of M that is the size of experimental
array.
pClONO2(Tc)
w1He
[
ClONO ] )
pnL (1)
2
w1He + w2He + w3He
pa
Synthesis of ClONO2. Chlorine nitrate is produced by the
reaction of ClF with anhydrous HNO3 in stainless steel autoclave
3
24
where w1He, w2He, and w3He are the measured He mass flows
through inlets 1, 2, and 3; pa is the absolute pressure in the
with an internal volume of 130 cm . The initial reactants, ClF
(0.05 M) and HNO3 (0.05 M), are placed into the autoclave at
-196 °C and heated slowly to 0 °C. After cooling to -78 °C,
the autoclave is attached to a vacuum line allowing rapid
ClONO2 distillation into a reservoir held at -95 °C. The ClF
reactant is produced by the reaction of Cl2 with F2. To prepare
ClF, F2 (0.1 M) and Cl2 (0.1 M) are placed into the autoclave
ampule (9) measured by pressure gauge PG1; pClONO (Tc) is the
2
partial pressure of ClONO2 at temperature, Tc, of the cryostat;
p is the measured pressure in the reactor; and nL is the Loschmidt
number.
The mass flow of ClONO2 molecules from the ampule (9)
into the reactor is controlled by varying the cryostat temperature.
Under conditions where the transport time of the flux w1He
through the ampule exceeds the time of diffusion mixing
3
of 1000 cm at -196 °C, heated to room temperature, and held
for 4 h, followed by heating to 200 °C for 3 h. Anhydrous HNO3
is produced by the reaction of KNO3 with concentrated H2SO4.
The main impurity in synthesized ClONO2 is Cl2. To purify
ClONO2, the initial ClONO2/Cl2 mixture is placed in the ampule
(9) at -120 °C. Upon heating to -98 °C, the ampule is pumped
through the reactor while monitoring the intensity on m/z 70
until Cl2 vanishes from the ampule.
considerably, pClONO (Tc) corresponds to the saturated vapor
2
pressure of ClONO2 in the ampule. Thus, exploitation of eq 1
allows the calibration of the mass spectrometer for ClONO2
and measurement at the ClONO2 saturated vapor pressure as a
function of temperature.
Main Parameters of the Flow Reactor. The inner diameter,
dR, of the reactor tube (2) is 1.3 cm; outer diameter, dr, of the
rod (4) is 0.13 cm; the maximum length of the rod, Lr , is 50
cm; the linear flow velocity, V, along the reactor is (0.35-7.5)
Possible contribution of NO2 to ClONO2 mass-spectral
intensity is taken into account. The corrected intensity ∆I46(30
eV) of ClONO2 is determined by the difference between a full
ion current I46(30 eV) and the contribution I46(11.8 eV) due to
3
-1
×
10 cm s ; and the characteristic pressure, p, is 1-2 Torr.
Salt Sample Preparation. Salt is deposited on the rod by
NO2, multiplied by relative sensitivity ηNO of mass spectrometer
2
to NO2 at the two ionization energies:
dipping it into unsaturated aqueous solution of NaCl followed
by drying at 200 °C. The surface covered is equal to the
geometric surface of the insert rod.
∆
I (30 eV) ) I (30 eV) - I (11.8 eV)η
;
NO2
46
46
46
ηNO2 ) 14.85
Measurement Procedure. Ion current intensities are mea-
sured at a fixed m/z without magnetic field scans using phase-
sensitive ion counts. The minimum acquisition time is 2.5 s
per cycle, with up to 10 cycles per experimental point. A
calibration of mass spectrometer for the stable substances NO2,
Cl2, ClONO2, and HCl is obtained by introducing the measured
fluxes into the reactor because the pressure in the reactor is
known. Characteristic ionization energies are: Ee ) 30 eV for
ClONO2 (m/z 46), 11.8 eV for NO2 (m/z 46), 30 eV for HCl
Data Treatment. As was shown earlier,25,26 the kinetics of
consumption of a gas-phase reactant on the insert rod is
described by first-order eq 3
-dn/dt ) kn
(3)
k
where n is a concentration of the gas-phase ClONO2 reactant;
tk ) [0, ∆L/V] is a contact time of ClONO2 with the rod inserted
into the reactor for the length ∆L; and V is a mean linear velocity
of He carrier gas. The rate constant k for heterogeneous uptake
(m/z 36), 30 eV for Cl2 (m/z 70), 30 eV for HOCl (m/z 52), and
3
0 eV for Cl2O (m/z 86). The uncertainty, σ, for every
k
experimental intensity, I, measured N times was calculated by
is shown to be expressed in terms of its kinetic (k ) and diffusion
d
27
(
k ) limits
N
1
2
1
k
1
1
σ )
(I - Imean)
)
+
k d
∑
i
x
N(N - 1) i)1
k
k
where25,26
after the acquisition procedure using soft hardware. Any
intensity, Ii, that differs by 3σ from Imean was rejected. The
reported experimental parameters obtained by data handling have
the uncertainties consisted of two components. The fist is
statistical standard error, σstat, of the measured data. The second
one is determined by systematic bias, σsys, in the measured
temperature, pressure, and reactant fluxes. The latter was of the
order of 5%. The overall error is
k
γc
dR
F
d
D
k )
k ) 4K(F)
(4)
2
2
1 - F
d
R
Here, F )dr /dR; K(F ) 0.1) ) 1.3 is a dimensionless diffusion
rate constant; c ) 2.52 × 10 cm s is a mean molecular
velocity of ClONO2 at 293 K; and D is a diffusion coefficient
2
5
4
-1

6774 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Zelenov et al.
d
k
k
of ClONO2 in He. In our case k . k ; therefore, k = k . On
the basis of eqs 3 and 4, the dependence of γ on exposure time,
t, of the saline covering to ClONO2 flux is given by
0
ln(I
/IClONO2(t))
2
ClONO2
dR
c
1 - F
γ(t) )
(5)
tk
F
0
where I
and IClONO (t) are the intensities of ClONO2
2
ClONO2
measured without a rod and with an inserted rod, respectively.
The partial uptake coefficient, γP1, describes the ClONO2
uptake leading to a formation of some gas-phase product P1.
The analytic expression for γP1 follows from the hypothetic
kinetic mechanism
Figure 2. Temperature dependence of the saturated vapor pressure
for ClONO . The symbols are experimental data. Dotted line shows
2
k(t)
linear regression by formula ln(p,Torr) ) 30.254 - (5444 ( 150)/T.
ClONO2
8 products
Bold solid line is taken from ref 28 data expressed by ln(p,Torr) )
1
8.396 - 3475.5/T for gas/liquid equilibrium.
k1(t)
ClONO2
8 P1
correspond to saturated vapor pressure at the vapor/solid
equilibrium. The temperature dependence of the saturated vapor
pressure was measured using a linear relationship between
The mechanism is described by eqs 6a and b
pressure pClONO and concentration [ClONO2] according to eq
1. The relative saturated vapor pressure is determined by well-
known eq 9
∂
[ClONO (t ,t)]
2
2
k
-
) k(t) [ClONO (t ,t)]
(6a)
(6b)
2
k
∂
tk
∂
[P (t ,t)]
1
k
ln(p/p ) ) A - ∆ H°(ClONO )/RT
(9)
0
S
2
)
k (t) [ClONO (t ,t)]
1
2 k
∂tk
where ∆SH°(ClONO2) is the enthalpy of sublimation, A is a
where γP1 ) γ‚k1/k is proposed by definition.
calibration constant, and p is a reference pressure.
0
On the basis of eqs 6a and b, the concentration of the product
P1 can be determined at two positions of the rod: (i) the rod is
inserted for the length ∆L, and the signal of product P1 is
recorded after some time, the uptake being steady-state and
characterized by k(t f ∞), k1(t f ∞); (ii) the rod is further
inserted for the length ∆L, and the signal of product P1 is
recorded after insertion at once, the values of k(t) and k1(t) being
time-dependent
A conversion of the relative dependence (eq 9) to the absolute
2
8
one uses reference data, where the temperature dependence
of saturated vapor pressure is given in the range of 193-299 K
for the vapor/liquid equilibrium. Additionally, a melting point,
2
8
Tm ) 166 K, was determined. The reference data are shown
in Figure 2 by a bold solid line. The intercept of reference and
our data at the point of the phase change at T ) 166 K enabled
us to convert the relative dependence into the absolute values.
The slope of our data in Figure 2 gives the enthalpy of
[
P1(∆L/V,t f ∞)]
-1
sublimation, ∆SH°(ClONO2) ) 45.26 ( 1.25 kJ mole . The
)
[ClONO2]0
difference between the literature enthalpy of evaporation ∆vH°-
2
8
(
ClONO2) and our ∆SH°(ClONO2) determines the enthalpy
a â (t f ∞) (1 - exp{-k(t f ∞) ∆L/V} ) (7a)
P1
-
1
of melting ∆mH°(ClONO2) ) 16.36 ( 1.35 kJ mole . The
absolute dependence at the vapor/solid equilibrium is given by
[
P1(2∆L/V,t)]
)
a â (t) (1 - exp{-k(t) ∆L/V}) +
P1
[ClONO2]0
ln p(Torr) ) 30.254 - (5444 ( 150)/T, 133 < T <163 K
a â (t f ∞) exp{-k(t) ∆L/V} (1 - exp{-k(t f ∞) ∆L/V})
P1
Cl2O Impurity in the ClONO2 Flux. The Cl2O impurity in
the ClONO2 flux was detected on m/z 86 at Ee ) 30 eV. Figure
(7b)
3
shows temporal dependences of ClONO2 and Cl2O concentra-
where a is a number of molecules of the product P1 formed
from one ClONO2 molecule; âP1(t) ) k1(t)/k(t).
The expression for γP1(t) follows from the ratio of 7b to 7a
tions after an initial flux of ClONO2 into the reactor. The sum
of their concentrations is seen to be constant over all times of
observation. This means that Cl2O impurity is a product of
partial decomposition of ClONO2 in the delivery line. After
passivating the delivery line by ClONO2 flow for 3 h, the content
of Cl2O in ClONO2 becomes less than 5%. Nevertheless, to
correct the variation in the ClONO2 concentration during the
experiment, the contribution of [Cl2O] was taken into account.
Calibration of the mass spectrometer for Cl2O was carried out
using the product balance [ClONO2(t ) 0)] ) [ClONO2(t)] +
γ (t)
γ(t)
[P (2∆L/V,t)]
1
P1
)
-
(
γP1(t f ∞) γ(t f ∞) [P (∆L/V,t f ∞)]
1
1
- exp{-k(t f ∞) ∆L/V}
exp{-k(t) ∆L/V}
(8)
)
1
- exp{-k(t) ∆L/V}
Here, γ(t) and γ(t f ∞) are determined from consumption of
ClONO2. The value γP1(t f ∞) is calculated from eq 7a at the
detection of the product P1(∆L/V, t f ∞) from steady-state
uptake ClONO2 on the first part of the rod.
2[Cl2O(t)], as well as a relationship between ClONO2 and Cl2O
concentrations and their mass-spectral intensities I86 ) [Cl2O]‚
η(Cl2O), ∆I46 ) [ClONO2]‚η(ClONO2). A relative sensitivity
η(ClONO2)/η(Cl2O) ) 0.14 for the mass spectrometer was
determined, taking into account all continuum of experimental
points I86(ti), ∆I46(ti) and the analytic relations given above. The
relative sensitivity obtained enables us to convert the intensity
3
. Experimental Results
ClONO2 Saturated Vapor Pressure Measurements. For
kinetic studies, we needed characteristic ClONO2 fluxes, which

Kinetic Mechanism of ClONO2 Uptake
J. Phys. Chem. A, Vol. 110, No. 21, 2006 6775
Figure 3. Temporal variation in the initial mixture of the reactants at
Figure 4. Temporal variations in the mass-spectral signals for ClONO
reactant (1) and the products Cl (2), HOCl (3, on an enlarged scale)
at two positions of the rod: (I), without rod and (II), with rod inserted.
Procedure was being carried out under conditions [ClONO
] ) 2.5 ×
2
1
5
-1
ClONO
concentration [ClONO
by eq 1. (1), experimental [ClONO
Cl O(t)]/[ClONO . Solid curves show their approximations by 0.23
2
flux from the ampule of 1 × 10 molecules s . The
2
1
2
-3
2
]
0
) 5.5 × 10 molecules cm is calculated
(t)]/[ClONO ; (2), experimental
2
2
]
0
2
12
-3
[
2
2
]
0
10 molecules cm , p ) 2 Torr, T ) 293 K, ∆L ) 10 cm, tres ) 4.9
2
-
+
0.67(1 - exp(-t/τ)) and 0.035 + 0.36 exp(-t/τ), respectively, with
× 10 s.
τ ) 36 min.
ratio I86/I46 into the [Cl2O]/[ ClONO2] ratio. Figure 3 shows
recalculated data. The corrected ClONO2 concentration in the
reactor was calculated by the formula
[
ClONO2]0
[
ClONO ] )
2
2
I86 η(ClONO2)
1
+
∆
I46 η(Cl O)
2
where [ClONO2]0 is the ClONO2 concentration determined by
eq 1.
An estimated upper limit value of the Cl2O uptake coefficient
on NaCl was obtained from measurement of the Cl2O intensities
at two positions of the rod, that is, I86(L ) 0) ) 566 ( 10 Hz
and I86(L ) 10 cm) ) 576 ( 11 Hz under the following
experimental conditions: T ) 293 K, p ) 2.02 Torr, V ) 220
Figure 5. Temporal variation in the total and partial uptake coefficients
for ClONO
to Figure 4. The symbols are experimental data for ClONO
tion (1), Cl (2) and HOCl (3) formation. Solid curves show approxima-
tion by eq 25 with parameters γch, γ
2
uptake on fresh NaCl under conditions given in the capture
2
consump-
2
-
3
s
, γini, τch, τini of 6.7 × 10 , 1.01
-
3
-3
-
1
-2
12
× 10 , 0.2, 15.3 s, 1.7 s for ClONO
2
; of 2.6 × 10 , 0, 0.2, 17.7 s,
cm s , tk ) ∆L/V ) 4.5 × 10 s, [ClONO2]0 ) 1.9 × 10
-
3
-3
1
.6 s for HOCl; of 7.1 × 10 , 1.01 × 10 , 0, 11 s, 0 s for Cl
2
.
-
3
11
-3
molecules cm , and [Cl2O] ) 3.5 × 10 molecules cm .
-
4
Evaluated γCl O e 2.7 × 10 follows from the formula
2
δI0 2
δI10 2
dR
γ
e
+
Cl O
2
( )
(
)
(
d /d ) c
tk
x
I
I10
r
R
Cl O
0
2
Thus, the upper estimation of γCl O is 1 order of magnitude less
2
than that of ClONO2 and the Cl2O impurity of 5% does not
affect the results for ClONO2 uptake.
Total and Partial Uptake Coefficients for ClONO2/NaCl.
Temporal consumption of the ClONO2 reactant and formation
of Cl2 and HOCl products were investigated for [ClONO2] )
11
13
-3
(
5 × 10 - 2.5 × 10 ) molecules cm . Figure 4 shows typical
temporal dependences of ClONO2, Cl2, and HOCl signals at
initial insertion of the rod covered with NaCl into the flux of
gas-phase reactant and after reinsertions of the same part of the
rod. Figure 5 shows the total and partial uptake coefficients
calculated from the data in Figure 4, using eqs 5 and 8. At the
initial exposure, a sharp drop in the uptake coefficient is seen.
Further diminishing the coefficient down to some steady-state
γs occurs with a time constant of the order of 10 s, the γs being
stable for a long time. At the beginning, HOCl is the main
product of the reaction. HOCl formation is sure to result from
reaction IV due to chemisorbed water on the surface of saline
covering. In the course of initial reaction of fresh salt coating,
the H2O molecules are consumed. Therefore, the HOCl product
is not present for repetitive rod insertions. Similar temporal
behavior of γ(t) on fresh saline coverings was observed
Figure 6. Dependence of the uptake coefficient γch on ClONO
concentration for the initial uptake of ClONO on NaCl at T ) 293 K.
The symbols are experimental data. The solid line shows linear
2
2
regression by 1/γch ) ach + bch[ClONO
2
] at ach ) 74.7 ( 7.7, bch
)
-
12
-1
3
(
18.7 ( 0.8) × 10 molecules cm .
previously for NO3 uptake on NaCl and NaBr.3
0-33
Figures 6-8
will be discussed in the framework of the kinetic mechanism
presented in the next section.
4
. Discussion
Morphology of the NaCl Surface Exploited. There are two
uncertainty factors that have to be taken into account to
determine trace gas uptake on saline coverings. These are the
porosity and roughness of the surface. The first was analyzed

6776 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Zelenov et al.
the resulting two-dimensional surface product is not transformed
3
9
into a three-dimensional crystalline structure.
Uptake on the a sites results in the highest energy release.
Uptake is likely to proceed without regeneration of the sites.
Uptake on the b sites appeared to be similar to that on crystal
defects containing chemisorbed water. Water molecules are
trapped by the b sites during coating preparation and cannot be
removed during surface drying because of high heat of adsorp-
tion. While participating in the reaction with ClONO2, the b
sites can regenerate or convert to the c sites with lower heat of
adsorption, the total density of the sites being constant.
Transformation of the b sites into the c sites during reactive
uptake is supposed to lead to a transition between the uptake
mechanisms, that is, from chemisorption at the beginning to
the physisorption responsible for the steady-state process.
Reproduction of the c sites can be explained only by the
presence of surface-adsorbed water. According to Seisel et al.,40
even dry NaCl monocrystals contain adsorbed water, which
corresponds to ∼10-20 formal monolayers. During uptake of
a ClONO2 molecule on the c site, water molecules give a
mobility to the NaNO3 product. Surface mobility of the products
enables them to be rearranged and form a new three-dimensional
crystalline phase. Thus, although not converted in the reaction,
the water remains and reproduces the c defects.
Figure 7. Dependence of the characteristic time τch on ClONO
concentration for the initial uptake of ClONO on NaCl at T ) 293 K.
The symbols are experimental data. The solid curve shows approxima-
2
2
12
tion by τch ) τ
0
3
(1 + nth/[ClONO
2
]) at τ
0
) 4.1 ( 0.6 s, nth ) 4 × 10
-
molecules cm
.
Kinetic Mechanism of ClONO2 Initial Uptake. Interaction
between ClONO2 molecules and the NaCl surface is proposed
to be described by the reaction mechanism
Figure 8. Dependence of the uptake coefficient γ
concentration for steady-state uptake of ClONO on NaCl. The symbols
are experimental data at 293 K (1) and 387 K (2, on an enlarged scale).
The solid lines show linear regressions by 1/γ (293 K) ) a
[ClONO ] at a ) 521 ( 61, b
) (14.8 ( 0.6) × 10 molecules
(387 K) ) 179.6 ( 21.5.
s 2
on ClONO
2
s
s
+
Mechanism V includes reversible chemisorption of ClONO2
molecules on the Zch sites followed by unimolecular decomposi-
tion of the surface complex. The decomposition proceeds via
two competitive channels. One of them reproduces initial Zch,
and another one transforms it into the Zph site that has different
reactive and adsorptive properties. Here, ka ) R c/4 is the
adsorption rate constant; kd ) Ad exp(-Q/RT) is the desorption
rate constant; R is a sticking coefficient; c is the mean molecular
-
11
-1
b
s
2
s
s
3
cm ; 1/γ
s
previously by Rossi.34 In fact, porosity increases chemically
active surfaces as compared to a geometric surface area because
of diffusional gas penetration. This factor is most relevant in
experiments on powders. The reactive surface of coarse grains
is shown by electron microscopy to approach their geometric
surface. Gas-phase species react on the first surface layer and
do not have the opportunity to diffuse inside when γ g 0.1.
1
3 -1
velocity of ClONO2; Ad = 10
s
is the preexponential factor;
kr is the rate constant of unimolecular decomposition of the
surface complex ClONO2‚‚‚Zch; and æ is a probability to
transform site Zch into site Zph in the course of unimolecular
decomposition.
According to this scheme, the decomposition with the rate
constant ækr transforms initial active sites Zch into sites Zph
responsible for steady-state uptake. The latter reproduces itself
and the surface density of Zph stays unaltered. The decomposi-
tion with the rate constant (1-æ)kr reproduces initial Zch and
gives some product P2. The contribution of Zch sites in a total
surface density of active sites diminishes because a part of them
transforms into Zph.
A prerequisite for the proposed chain mechanism of the initial
uptake is based on the large number of NaCl monolayers that
are spent in the course of the uptake. The evaluation follows
from equality of a number of spent ClONO2 molecules to a
number of the reacted surface sites under assumption that the
uptake of one molecule leads to consumption of one surface
site. This equality is expressed by eq 10
34
Film coatings used in this study have the least internal surface.
The second uncertainty factor is roughness of the surface.
Using a profilometer with linear resolution of 50 Å, a roughness
of the film coatings deposited from aqueous solutions turns out
to increase the surface area no more than 15%.35
Adsorption Sites of the NaCl Crystallites. Two steps for
ClONO2 uptake on fresh NaCl are observed in the experiments.
We associate the difference between them with the difference
in the properties of the adsorption sites where the uptake occurs.
According to literature data, there are three basic types of
adsorption sites, each of them having its own energy distribu-
tion: (a) sites capable of chemisorption, even up to dissociative
chemisorption; their initial surface density for the fresh NaCl
36
surface is of ∼1% with respect to the total surface site density
1
4
-2 37
Z0 ) 6.4 × 10 sites cm ; (b) the defects of crystalline
38
structure with the surface density of ∼20%; and (c) the vacant
sites generated on conditionally defectless surface. Their action
is controlled by physisorption of gas-phase reactants, the heat
of adsorption being lower compared to those of a and b sites.
Reactive uptake on the c sites is shown to stop at once after the
first interaction between the gas-phase molecule and the site;
[
ClONO ]cS
2
∞
∫
γ(t) dt ) n
S
S
(10)
0
4

Kinetic Mechanism of ClONO2 Uptake
J. Phys. Chem. A, Vol. 110, No. 21, 2006 6777
where [ClONO2] is the ClONO2 volume concentration; c ) 2.53
k Z Sf
ækr t
d
dt
r
0
4
-1
V
nV ) -
exp -
(19)
×
10 cm s is a mean molecular velocity; S is a geometric
(
)
1 + (n /n )
1 + (n /n )
th
V
th
V
surface of the saline covering; nS is the surface density of reacted
sites; γ(t) is the approximation of the data in Figure 5 by
Alternatively, a formal definition of γ is given by
3
function 10 γ(t) ) 6.67 exp(-t/15.25) + 200 exp(-t/1.757)
without the contribution of the steady-state uptake (γs ) 1.01
×
d
dt
γc
-
3
V
nV ) - n S
(20)
10 ).
V
4
A number of spent monolayers are determined by the relation
nS/(f Z0) because only f ) 0.2 part of the total density Z0 of
Comparison of eqs 19 and 20 gives explicit expression for γ
active sites are capable of reactive uptake. The evaluation gives
4
Z0 f
kr
ækr t
56 monolayers. It means that initial uptake proceeds via a chain
t
τch
(
γ )
exp -
) γ exp -
ch
mechanism with partial reproduction of the initial sites.
On the basis of the reaction mechanism (V) and on the
simplest Langmuir adsorption model, a variation in the total
number of ClONO2 molecules on the salt surface and in the
gas phase is determined by eqs
c
nV + nth
(
1 + (n /n )
)
(
)
th
V
21)
where
1
ækr
nth
-
1
ch
γ
) a + b n , τ )
1 +
(22,23)
d
dt
ch
ch
V
ch
(
nV)
V
n ) -(J - J )Sf
(11)
(12)
(13)
V
a
d
with
dθch
dt
SZ0
) (J - J - J )S
a d r
1
R f
kd
1
c
ach
a )
1 + , b )
, n )
th
ch
ch
(
)
kr
f 4k Z
bch
r
0
dθph
dt
SZ0
) æ J S
r
Note that combined eqs 12 and 13 can be resolved relative to
the sought quantities θch(t) and θph(t) without assumption (eq
where nV is the ClONO2 concentration in the gas phase; θch ×
Z0 and θph × Z0 are surface density of Zch and Zph sites; S is a
geometric surface of the rod covered with NaCl and introduced
for the length L into the ClONO2 flux; V is a reactor volume
corresponding to the length L; f ) 0.2 is a fraction of Z0 capable
of reactive uptake; Ja, Jd, and Jr are the mass fluxes of ClONO2
onto unit surface corresponding to adsorption, desorption, and
reaction, respectively
1
5) of quasi stationarity on θch. In this case, the expression for
γ follows from the equating of right parts of eqs 11 and 20,
taking into account terms Ja and Jd from eq 14
t
t
t
γ ) γ exp -
1 - exp -
+ γini exp -
38
ch
(
τch) {
(
τini)}
(
)
τini
(
24)
where
J ) k n (1 - θ - θ ), J ) k θ Z , J ) k θ Z
r ch 0
a
a
V
ch
ph
d
d
ch
0
r
1
1
(14)
γini ) R f, τini
)
k + k 1 + (n /n )
d
r
V
th
On the assumption of dynamic equilibrium between the fluxes,
that is, for
The values of γch and τch are the same as those in eq 22 under
dynamic equilibrium. The time τini characterizes a transition
from empty surface to dynamic equilibrium between the
adsorption, desorption, and reaction fluxes. In one of the extreme
cases, that is, at nV , nth, this time is determined by the
J - J - J = 0
(15)
a
d
r
a relation (eq 16) between θch and θph follows from eq 15
- θph
+ (n /n )
desorption rate and approximately equals a residence time k-d
1
1
θ )
(16)
for adsorbed molecule on the surface. In the other extreme case,
that is, at unimolecular covering, the time τini is determined by
the adsorption rate and equals 4Z0/(cnV).
ch
1
th
V
NaCl
where nth ) (kd + kr)(Z0 /ka) is some threshold volume
concentration of ClONO2, which corresponds to unimolecular
covering in Langmuir model.
Equations 21 and 24 for γ describe the initial uptake on the
Zch sites only. In fact, in the course of the initial uptake the Zph
sites interact with ClONO2 as well. Formally, it means that Jr
from eq 14 must contain an additional term kr 2θphZ0. Here kr 2
is the rate constant of unimolecular decomposition of the surface
complexes ClONO2‚‚‚Zph. Substitution of Jr 2 ) kr2θphZ0 into
eq 19 gives additional contribution in γ, that is, γs(1 - exp(-
t/τch)), where γs is the uptake coefficient of the steady-state
uptake that occurs on the Zph sites only. Thus, in summary, an
uptake coefficient that describes both initial and steady-state
uptake is presented by formula
Taking into account initial condition θph(t ) 0) ) 0, after
substitution of θch from eq 16 into eq 13, we find an explicit
form for θph
ækr t
1
- θ ) exp -
(17)
ph
(
)
1
+ (n /n )
th V
On the basis of eq 16, expression for θch is given by
1
ækr t
γ ) γ exp(-t/τ ){1 - exp(-t/τ )} +
ch ch ini
θ )
exp -
(18)
ch
(
)
1
+ (n /n )
1 + (n /n )
γ (1 - exp(-t/τ )) + γ exp(-t/τ ) (25)
s ch ini ini
th
V
th
V
On the basis of definition 14 and assumption 15, eq 11 now
transforms into
Elementary Kinetic Parameters of the Initial Uptake.
Figures 6 and 7 show summarized data on dependences of γch

6778 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Zelenov et al.
-
1
TABLE 1: Summarized Elementary Constants for Reactive
in terms of rate constants kd and kr gives kr ) 2.6 ( 0.1 s and
ClONO Uptake on Polycrystalline Film of NaCl at 293 K
2
-1
kd ) 37 ( 4 s . The ratio ach/bch gives the threshold volume
1
2
-3
n
th
,
k
s
d
-1
,
k
s
r
-1
,
concentration nth ) (4.0 ( 0.4) × 10 molecules cm . For
calculation, the following parameters were used: R ) 1, f )
1
2
-3
stage
initial
steady-state
10 cm
1/æ
3
8
4
-1
NaCl
0
14
-2
4.0 ( 0.4
3.5 ( 0.4
37 ( 4
34.7 ( 4.3
2.6 ( 0.1
0.3 ( 0.2
11 ( 1.6
0.2, c ) 2.53 × 10 cm s , Z
) 6.4 × 10 sites cm .
On the basis of eq 23, the data in Figure 7 were approximated
by function τch ) τ0(1 + nth/[ClONO2]) where the sole sought
parameter was τ , the threshold concentration n being taken
and τch on the ClONO2 volume concentration. The data were
obtained by handling the temporal dependences similar to the
ones shown in Figure 5. The treatment was produced by eq 25.
The last of the terms in eq 25 is responsible for a nonstationary
uptake and gives an essential contribution in only several first
points of the experimental γ(t). Besides, an uncertainty in the
start of measurement is of 1 s. Therefore, precision of the τini
determination is low. To exclude effect of this stage of the
uptake, several first points of experimental γ(t) were not taken
into account at the first step of the treatment. All assemblage
of the residuary experimental points are described now by the
simplified expression
0
th
from the data in Figure 6. The result of this approximation is
τ ) 4.1 ( 0.6(1σ) s with ø ) 16 for size ) 7. On the basis
of its definition, τ ) 1/æ k , and the calculated k , the parameter
2
0
0
r
r
1/æ is found to be 11 ( 1.6 s. The parameter æ represents the
probability of transforming site Zch into site Zch during unimo-
lecular decomposition of the surface complex ClONO ‚‚‚Z .
2
ch
The reciprocal value 1/æ characterizes the number of catalytic
cycles of the reaction before transformation of Zch into Zph.
Steady-State ClONO /NaCl Uptake. Figure 8 shows de-
pendences of the uptake coefficient on the ClONO volumetric
2
2
concentration for steady-state uptake at two temperatures, 293
and 387 K. The regression form is likely to be valid because ø2
corresponds exactly to the number of the experimental points
for both temperatures. The dependences agree well with the
model proposed by us earlier for the NO3/NaCl and NO3/NaBr
γ(t) ) γ exp(-t/τ ) + γ (1 - exp(-t/τ )) (26)
ch
ch
s
ch
Parameters γch and τch were obtained by fitting of the experi-
mental γ(t) with eq 26, γs being taken independently from the
stationary part of the experimental γ(t). After that, all of the
experimental points, including the first ones deleted earlier, were
approximated by eq 25, where the sole sought value was τini.
The value of γini was taken as 0.2 based on its expression in
terms of parameters R and f given in eq 24. Note that our model
predicts a sharp initial drop in γ(t), which is likely explained
by the initial nonstationary uptake.
3
0-33
systems.
scheme
The steady-state uptake can be described by the
ka
kr
ClONO + Z {\ } ClONO ‚‚‚Z 98 Z + products (VI)
2 ph ph
kd
2
ph
On the basis of the proposed mechanism and precisely described
in the previous section, the uptake coefficient is expressed by
Elementary parameters were determined from approximation
of the dependences in Figures 6 and 7 by analytical expressions
-1
1
R fph
kd + k
r
nV
γs
)
1 +
(27)
(
)
2
2 and 23. The parameters ach ) 74.7 ( 7.7(1σ) and bch )
k
r
n
th
-12
-1
3
(
18.68 ( 0.8(1σ)) × 10 molecules cm were obtained from
linear regression of data in Figure 6 by eq 22. Here, σ is one
where fph is a relative portion of the Zph surface sites in the
2
NaCl
NaCl
standard error. Note that ø ) 8 calculated by eq 2 is in good
agreement with the number of experimental points in Figure 6.
total surface density Z0 ; and nth ) Z0 (kd + kr)/ka.
Generally, the values of kr , kd, and nth differ from those
for the initial uptake. Similar to the initial uptake case, we can
On the basis of eq 22, the expression of parameters ach and bch
TABLE 2: Comparison with Literature Data
type of uptake
coefficient
value
substrate type
granules
T, K
ref
comments
-
3
1
γ
γ
s
(4.6 ( 3.1) × 10
296
17
18
independent of [ClONO
2
] )
8
-3
(
γ
3-3000) × 10 molecules cm
-
0
, γ
s
s
(2.3 ( 0.6) × 10
solid powder, grains,
crystals, thin films of
crystallites
polycrystalline film
solid NaCl
298
0
≈ γ
s
, independent of both
] )
substrate type and [ClONO
2
10
13
-3
10 -10 molecules cm
γ
γ
γ
γ
γ
γ
γ
γ
0
0
0
s
0.1
0.1 ( 0.05
g0.1
298
298
298
19
20
21
, γ
solid NaCl
-
3
2
(6.5 ( 3.0) × 10
0
s
0.42 ( 0.46
0.16 ( 0.2
synthetic
sea salt
solid NaCl
polycrystalline film
298
21
22
-
0
, γ
s
(2.4 ( 1.2) × 10
296
293
-
1
12
ch
γ
a
ch ) ach + bch[ClONO
2
];
this work
[ClONO
molecules cm
a mechanism of initial and
2
] ) (0.5-25) × 10
-3
ch ) 74.7 ( 7.7(1σ);
;
-
11
b
ch ) (1.9 ( 0.1(1σ)) × 10
-
1
3
molecules cm
steady-state uptake is proposed;
a number of elementary rate
constants are determined
-
1
γ
s
s
γ
s
) a
s
+ b
s
[ClONO
) 521 ( 61(1σ);
2
];
polycrystalline film
polycrystalline film
293
387
this work
this work
a
b
s
-
11
s
) (14.8 ( 0.6(1σ)) × 10
-
1
3
molecules cm
(6.0 ( 0.9(1σ)) × 10
-
3
γ
s
γ values are described at 293
and 387 K in the framework of
the common kinetic model

Kinetic Mechanism of ClONO2 Uptake
rewrite eq 27 in the form
J. Phys. Chem. A, Vol. 110, No. 21, 2006 6779
stages. The final objective of our approach is to extrapolate
laboratory data to real troposphere conditions. The value of γ
= 0.1 is proposed by Rossi to be a consistent value. Our γch
-
1
34
γs ) a + b n
(28)
(29)
s
s
V
and γs for “dry” salt at T ) 293 K are much lower compared
to his value. Maximum values of γch and γs are observed at
where
-
2
low [ClONO2], that is, γch ([ClONO2] f 0) = 10 and γs
-
3
1
R fph
kd
1
c
as
([ClONO2] f 0) = 2 × 10 . These values are closest to those
17 -3
as )
1 +
bs )
, n )
th
(
)
of Timonen et al., γs ) (4.6 ( 3.1) × 10 , and Hoffman et
22 -2
kr
f 4k Z
bs
ph
r
0
al., γ0,s ) (2.4 ( 1.2) × 10 . These lower values are unlikely
to explain high concentration of Cl2 observed in the MBL
The parameters as ) 521 ( 62(1σ) and bs ) (14.8 ( 0.55(1σ))
9
-
11
-1
3
regions by some field measurements. However, according to
×
10
molecules cm were determined from linear regres-
our preliminary data, γ increases in direct proportion to the
relative humidity (RH). It can reach value of ∼0.1 at RH of
20-30%, when NaCl particles will still be solid. At such a high
sion of the data in Figure 8 by eq 28. The ratio as/bs gives the
threshold volume concentration nth ) (3.5 ( 0.4) × 10
molecules cm at T ) 293 K. The rate constants kr ) 0.3 (
.2 s and kd ) 34.7 ( 4.3 s are evaluated from eq 29 at R
1 and fph ) 0.2. The value of fph follows from the model
assumption [Zch(t)] + [Zph(t)] ) f Z0 , where f ) 0.2 and
Zch(t f ∞)] ) 0. Note that γs has a maximum value at
ClONO2] , nth, that is, at low population of the surface by
1
2
-
3
-
1
-1
uptake coefficient, self-catalyzed extraction of Cl2 from NaCl
0
is possible in MBL regions polluted with nitric oxides.1
3-15
All
)
NaCl
of these points deserve to be weighed carefully. In the next
paper, the dependence of the ClONO2 uptake on humidified
NaCl will be considered and discussed.
[
[
adsorbed molecules
Acknowledgment. Yu.M.G. thanks Prof. B. J. Finlayson-
Pitts and her colleagues for interesting and useful discussions
during his visit to University of California, Irvine. Also,
Yu.M.G. and V.V.Z. thank all of the participants of the NSF
project "Atmospheric Integrated Research for Understanding
Chemistry of Interface (AIR UCI)" for active and useful
discussion on results of this paper during the 1st Workshop of
Environmental Molecular Science Institutes (June 25-30, 2005).
We understand that this work would be impossible without the
support of U.S. CRDF (award RC2-2521-MO-03) and RFBR
(
s
max)
γ
= R f (k /k ) ) 1/a
s
(30)
ph
r
d
Taking into account Arrhenius expressions for kr ) Ar exp-
-Er /RT) and kd ) Ad exp(-Q/RT), the γ
Er > Q when temperature rises. Otherwise, for Er < Q, the
(max)
(
will increase if
s
(
s
max)
γ
will decrease. Here Ar and Ad are preexponential factors
independent of temperature, Er is activation energy of unimo-
lecular decomposition, and Q is the heat of adsorption. It follows
from Figure 8 that the inequality Er > Q is realized. The ratio
(grants 00-03-32999 and 03-03-32711).
-
3
max
of γs (387 K) ) (6 ( 0.9(1σ)) × 10 to γs (293 K) ) (1.9
-
3
(
0.1(1σ)) × 10 gives
References and Notes
-1
(1) Farman, J. C.; Gardiner, B. C.; Shanklin, J. D. Nature 1985, 315,
207.
Er - Q ) (11.5 ( 1.6) kJ mole
(31)
(
2) Molina, M. J.; Rowland, F. S. Nature 1974, 249, 810.
(3) Bottenheim, J. W.; Galant, A. C.; Brice, K. A. Geophys. Res. Lett.
1986, 13, 113.
(4) Oltmans, S. J.; Komhyr, W. D. J. Geophys. Res. 1986, 91, 5229.
-
1
The value of Q ) (64.3 ( 0.3) kJ mole follows from kd (293
-1
K) ) (34.7 ( 4.3) s , supposing a normal preexponential factor
1
3
-1
Ad = 10 s . The sum of eq 31 and calculated Q gives the
(
(
(
5) Platt, U. Water, Air, Soil Pollut. 2000, 123, 229.
6) Sander, R.; Crutzen, P. J. J. Geophys. Res. D 1996, 101, 9121.
7) Wang, T. X.; Margerum, D. W. Inorg. Chem. 1994, 33, 1050.
-
1
activation energy Er ) (75.8 ( 1.6) kJ mole .
The threshold volume concentration nth(387 K) can be
evaluated by eq 29, taking into account the Arrhenius expres-
(8) Beckwith, R. C.; Wang, T. X.; Margerum, D. W. Inorg. Chem.
1996, 35, 995.
(9) 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.
10) Foste, K. L.; Plastridge, R. A.; Bottenheim, J. W.; Shepson, B.;
Finlayson-Pitts, B. J.; Spicer, C. W. Science 2001, 291, 471.
11) Singh, H. B.; Gregory, G. L.; Anderson, B.; Browell, E.; Sachse,
G. W.; Davis, D. D.; Crawford, J.; Bradshaw, J. D.; Talbot, R.; Blake, D.
R.; Thornton, D.; Newell, R.; Merrill, J. J. Geophys. Res. 1996, 101, 1907.
1
5
sions for kr and kd. This value is found to be 1.8 × 10
-
3
molecules cm for R ) 1. Thus, nth(387 K) . [ClONO2] in
the range of ClONO2 volume concentrations exploited. In this
case, the γs is expressed by eq 30 and must be independent of
(
(
[ClONO2].
Table 1 summarizes all of the elementary constants found
that characterize the initial and steady-state steps of the uptake.
The indicated uncertainties correspond to one standard error.
The values of nth and kd are seen to be identical, whereas the
rate constants of unimolecular decomposition differ by 1 order
of magnitude. This means that the difference between the surface
complexes ClONO2‚‚‚Zch and ClONO2‚‚‚Zph is likely to be
determined by their different reactivity rather than by their
adsorption/ desorption properties.
Comparison with Literature Results. Shown in Table 2,
the uptake coefficients determined in reviewed studies are
dramatically different in value. The partial uptake coefficients
leading to a formation of the prime gas-phase products were
not measured. Further, initial and steady-state steps of the uptake
were not defined in details.
(
12) Pszenny, A. A. P.; Keene, W. C.; Jacob, D. J.; Fran, S.; Maben, J.
R.; Zetwo, M. P.; Sringer-Young, M.; Galloway, J. N. Geophys. Res. Lett.
993, 20, 699.
13) Gershenzon, M. Yu.; Grigorieva, V. M.; Il’in, S. D.; Remorov, R.
1
(
G.; Shestakov, D. V.; Zelenov, V. V.; Aparina, E. V.; Gershenzon, Yu. M.
Global Atmospheric Change and Its Impact on Regional Air Quality. Nato
Science Series. IV. Earth and EnVironmental Science; Barnes, I., Ed.; Kluwer
Academic Publishers: Boston, MA, 2002; Vol. 16, p 109.
(14) Gershenzon, M. Yu.; Grigorieva, V. M.; Shestakov, D. V.; Zelenov,
V. V.; Gershenzon, Yu. M.; Zellner, R. The Mechanism of Heterogeneous
“Self-Cleaning” of the Coastal Troposphere. May it Have an “Explosion-
like” Autocatalytic Character? Combustion and Atmospheric Pollution; Roy,
G. D., Ed.; Frolov, S. M.; Starik, A. M. Torus-Press: Moscow, 2003; p
3
90.
(15) Gershenzon, M. Yu.; Grigorieva, V. M.; Il’in, S. D.; Shestakov,
D. V.; Gershenzon, Yu. M.; Zellner, R.; Finlayson-Pitts, B. J. Mechanism
of “Chlorine Explosion” in the NOx-Enriched Troposphere. Combustion
and Pollution: Atmospheric Impact; Roy, G. D., Ed.; Frolov, S. M.; Starik,
A. M. Torus-Press: Moscow, 2005; p 117.
Generally, γch and γs turn out to be dependent on ClONO2
concentration. This enables us to identify elementary stages of
the uptake process and to evaluate parameters of the two uptake
(
16) Finlayson-Pitts, B. J.; Ezell, M. J.; Pitts, J. N. Nature 1989, 337,
241.

6780 J. Phys. Chem. A, Vol. 110, No. 21, 2006
Zelenov et al.
(
17) Timonen, R. S.; Chu, L. T.; Leu, M.-T.; Keyser, L. F. J. Phys.
Chem. 1994, 98, 9509.
18) Caloz, F.; Fenter, F. F.; Rossi, M. J. J. Phys. Chem. 1996, 100,
494.
(29) Zelenov, V. V.; Aparina, E. V.; Shestakov, D. V.; Gershenzon,
Yu. M. Khim. Fiz. 2004, 23, 18.
(30) Gershenzon, M. Yu.; Il’in, S. D.; Fedotov, N. G.; Gershenzon, Yu.
M.; Aparina, E. V.; Zelenov, V. V. J. Atmos. Chem. 1999, 34, 119.
(31) Zelenov, V. V.; Aparina, E. V.; Gershenzon, M. Yu.; Il’in, S. D.;
Gershenzon, Yu. M. Khim. Fiz. 2002, 21, 42.
(
7
(
(
(
(
19) Koch, T. G.; Rossi, M. J. J. Phys. Chem. A 1998, 102, 9193.
20) Aguzzi, A.; Rossi, M. J. Phys. Chem. Chem. Phys. 1999, 1, 4337.
21) Gebel, M.; Finlayson-Pitts, B. J. J. Phys. Chem. A. 2001, 105, 5178.
22) Hoffman, R. C.; Gebel, M. E.; Fox, B. S.; Finlayson-Pitts, B. J.
(32) Zelenov, V. V.; Aparina, E. V.; Gershenzon, M. Yu.; Il’in, S. D.;
Gershenzon, Yu. M. Khim. Fiz. 2003, 22, 37.
Phys. Chem. Chem. Phys. 2003, 5, 1780.
23) Zelenov, V. V.; Loboda, A. V.; Aparina, E. V.; Dodonov, A. F.
(33) Zelenov, V. V.; Aparina, E. V.; Gershenzon, M. Yu.; Il’in, S. D.;
Gershenzon, Yu. M. Khim. Fiz. 2003, 22, 59.
(34) Rossi, M. J. Chem. ReV. 2003, 103, 4605.
(35) Dementiev, A. P.; Zelenov, V. V.; Aparina, E. V.; Shestakov, D.
V.; Il’in, S. D.; Gershenzon, Yu. M. Khim. Fiz. 2004, 23, 54.
(36) Dai, D. J.; Peters, S. J.; Ewing, G. E. J. Phys. Chem. 1995, 99,
10299.
(37) Kittel, C. Introduction to Solid State Physics, 6th Ed.; Wiley: New
York, 1986.
(38) Dai, D. J.; Ewing, G. E. J. Phys. Chem. 1993, 98, 5050.
(39) Finlayson-Pitts, B. J. Chem. ReV. 2003, 103, 4801.
(40) Seisel, S.; Fluckiger, B.; Caloz, F.; Rossi, M. J. Phys. Chem. Chem.
Phys. 1999, 1, 2257.
(
IzVestia RAS, Energ. 1997, 1, 70.
(
24) Brauer, G. Handbook on Inorganic Synthesis; MIR: Moscow, 1985.
(25) Gershenzon, Yu. M.; Grigorieva, V. M.; Zasypkin, A. Yu.;
Remorov, R. G. Kinetic resistance additivity for coaxial reactor. 13th Int.
Symp. on Gas Kinetics, Dublin, Ireland, Sept. 11-16, 1994. Book of
Abstracts, p 420.
(
26) Gershenzon, Yu. M.; Grigorieva, V. M.; Ivanov, A. V.; Remorov,
R. G. Faraday Discuss. 1995, 100, 83.
27) Franck-Kamenetskii, D. A. Diffusion and Heat Change in Chemical
Kinetics: Nauka, Moscow, 1967.
28) Schack, C. J. Inorg. Chem. 1967, 6, 1938.
(
(