Heterogeneous uptake of gaseous N2O5 by (NH4)2SO4, NH4HSO4, and H2SO4 aerosols

Article abstract of DOI:10.1021/jp010490x

The heterogeneous uptake of gaseous N₂O₅ by ammonium sulfate [(NH₄)₂SO₄], ammonium bisulfate [NH₄HSO₄], and sulfuric acid [H₂SO₄] aerosols as a function of relative humidity has been investigated at room temperature and atmospheric pressure. Ammonium-containing aerosols were generated by a constant-output atomizer and conditioned by passing through a diffusion dryer. Sulfuric acid aerosols were produced by the homogeneous reaction of SO₃ and H₂O in a borosilicate vessel. Addition of a dry or wet N₂ flow controlled the relative humidity (RH) of these aerosol flows. Using a chemical ionization mass spectrometer (CIMS) for N₂O₅ concentration monitoring and a scanning mobility particle spectrometer (SMPS) for aerosol characterization, reaction probabilities (γ) in the range of 0.001 to 0.1 for the uptake of N₂O₅ were determined as a function of RH. The results are expressed as follows: γ[(NH₄)₂SO₄] = 2.79 × 10^-4 + 1.30 × 10^-4 × (RH) - 3.43 × 10^-6 × (RH)² + 7.52 × 10^-8 × (RH)³, γ[NH₄HSO₄] = 2.07 × 10^-3 - 1.48 × 10^-4 × (RH) + 8.26 × 10^-6 × (RH)², and γ[H₂SO₄] = 0.052 - 2.79 × 10^-4 × (RH). We suggest that the water content and phase in the ammonium-containing aerosols control the reactivity of N₂O₅ while liquid-phase ionic reactions primarily dominate the uptake in sulfuric acid aerosols.

Full text of DOI:10.1021/jp010490x

J. Phys. Chem. A 2001, 105, 6465-6470
6465
Heterogeneous Uptake of Gaseous N2O5 by (NH4)2SO4, NH4HSO4, and H2SO4 Aerosols
Sean M. Kane, Francois Caloz, and Ming-Taun Leu*
Earth and Space Sciences DiVision, Jet Propulsion Laboratory, California Institute of Technology,
Pasadena, California 91109
ReceiVed: February 7, 2001; In Final Form: April 27, 2001
The heterogeneous uptake of gaseous N2O5 by ammonium sulfate [(NH4)2SO4], ammonium bisulfate
[NH4HSO4], and sulfuric acid [H2SO4] aerosols as a function of relative humidity has been investigated at
room temperature and atmospheric pressure. Ammonium-containing aerosols were generated by a constant-
output atomizer and conditioned by passing through a diffusion dryer. Sulfuric acid aerosols were produced
by the homogeneous reaction of SO3 and H2O in a borosilicate vessel. Addition of a dry or wet N2 flow
controlled the relative humidity (RH) of these aerosol flows. Using a chemical ionization mass spectrometer
(CIMS) for N2O5 concentration monitoring and a scanning mobility particle spectrometer (SMPS) for aerosol
characterization, reaction probabilities (γ) in the range of 0.001 to 0.1 for the uptake of N2O5 were determined
as a function of RH. The results are expressed as follows: γ[(NH4)2SO4] ) 2.79 × 10-4 + 1.30 × 10-4
×
(RH) - 3.43 × 10-6 × (RH)2 + 7.52 × 10-8 × (RH)3, γ[NH4HSO4] ) 2.07 × 10-3 - 1.48 × 10-4 × (RH)
+ 8.26 × 10-6 × (RH)2, and γ[H2SO4] ) 0.052 - 2.79 × 10-4 × (RH). We suggest that the water content
and phase in the ammonium-containing aerosols control the reactivity of N2O5 while liquid-phase ionic reactions
primarily dominate the uptake in sulfuric acid aerosols.
Introduction
The heterogeneous uptake of gaseous N2O5 by sulfuric acid
aerosols
The heterogeneous reaction of N2O5 with sulfuric acid has
been investigated extensively in aerosols,6-10 droplet,11 and bulk
surfaces12,13 in the temperature range of 200-300 K. The results
at room temperature have been reviewed recently14,15 and found
to be in excellent agreement among these investigations. For
this reason we have performed some experiments to check the
reliability of our recently developed aerosol reactor as shown
in the Experimental Section. For ammonium bisulfate aerosols,
Mozurkewich and Calvert6 have measured the γ values as a
function of relative humidity from 1 to 76% and demonstrated
the peak reactivity of ∼0.05 at ∼55%. In a previous investiga-
tion Hu and Abbatt9 have determined the reaction probability
being 0.04-0.05 at 50-69% and decreasing to a value of ∼0.02
at 83-94% for ammonium sulfate aerosol. Since the data basis
is rather limited, we feel the confirmation of these two sets of
data on ammonium-containing sulfate aerosols in the range of
relative humidity from 1 to 100% is necessary.
N2O5 + H2O f 2HNO3
(1)
is one of the most important atmospheric processes in controlling
OH and NOx concentrations, and thus determining the ozone
balance in the upper troposphere and lower stratosphere.1,2
Reaction 1 converts the reactive nitrogen oxides (NOx) into the
reservoir species HNO3, which is then recycled back to NOx
by the reaction with OH radicals and the photolytic process, or
removed from the atmosphere by deposition.
In a global troposphere model, Dentener and Crutzen3 have
adopted a reaction probability (γ) of 0.1 for ammonium bisulfate
aerosol as a typical aerosol in the atmosphere. They found the
decreases of yearly averaged NOx, O3, and OH levels of 49%,
9%, and 9% with respect to values obtained by considering gas-
phase reactions only. Dentener and Crutzen3 have also calculated
the impact by using a smaller reaction probability of 0.01 and
the decreases of NOx, O3, and OH are still very high, being
40%, 4%, and 3%, respectively. These calculations suggest that
the magnitude of γ for reaction 1 is also of fundamental
importance in global tropospheric chemistry.
Sulfuric acid aerosol is produced from gas-phase processes
of sulfur-bearing compounds and also from volcanic emissions.
Under lower stratospheric conditions, the aerosol may consist
entirely of sulfuric acid. In the middle and upper troposphere,
however, reactions of sulfuric acid aerosols with ammonia lead
to the neutralized forms of sulfate aerosols: ammonium bisulfate
and ammonium sulfate.4 The properties of these aerosols,
including deliquescence and efflorescence, under varying at-
mospheric conditions of relative humidity (RH) and temperature,
have been thought to be important in heterogeneous chemistry.4,5
Because of the importance of this reaction in the atmosphere,
we have undertaken a series of experiments to measure reaction
probabilities for these aerosols using an aerosol flow-reactor
that is interfaced to a chemical ionization mass spectrometer.
In the following sections we will describe the preparation and
measurement of aerosols, the procedures used to determine
reaction probability, and the potential impact of our results on
atmospheric chemistry.
Experimental Method
The apparatus used in these experiments is composed of two
main sections. The first section includes a brief discussion of
the generation and characterization of the various sulfate aerosols
under investigation. The second section describes the instruments
needed to introduce and measure N2O5. The components of each
section are discussed along with the treatment relevant to data
analysis.
Aerosol Generation and Characterization. Figure 1 shows
the components of the experimental setup used to control aerosol
* Author to whom correspondence should be addressed.
10.1021/jp010490x CCC: $20.00 © 2001 American Chemical Society
Published on Web 06/01/2001

6466 J. Phys. Chem. A, Vol. 105, No. 26, 2001
Kane et al.
Figure 2. Typical size distribution of ammonium sulfate aerosols. Data
shown are from 0.1 wt % ammonium sulfate with 92% RH and are
corrected for multiple charging of particles. The mean diameter is 0.12
µm and the geometric standard deviation is 1.965. The surface area
density is 0.0016 cm2/cm3.
Figure 1. Schematic diagram of generation and measurement of
ammonium-containing sulfate aerosols.
formation and measurement. Ammonium sulfate (J. T. Baker,
99.4%) and ammonium bisulfate (E M Science, 99.8%) aerosols
were formed by using a constant output atomizer (TSI, model
3075) filled with 0.1%-by-mass solutions in distilled water. This
dilution provided sufficient aerosol output for uptake measure-
ments without blocking such aerosol measurement instruments
such as the regulating valves and inlet impactor.
electrical mobility before being passed to a condensation particle
counter (TSI, model 3022A). By scanning the classifier voltage
and measuring the particle sizes at those voltages, a particle
distribution was obtained. Since particle separation depends on
electrical mobility and not particle size, a correction to the
particle distribution must be made to account for multiply
charged aerosols. A typical data of sulfate aerosol from this
setup is shown in Figure 2. From this information, a variety of
properties of the aerosol can be calculated, including total
particle density, surface area, and aerosol volume, with particular
interest in the surface area-to-volume ratio of the aerosol flow.
It should be noted that the calculation using the SMPS software
is based on the assumption that the aerosols are spherical.
To ensure that aerosol size distribution measured after
exposure to the flow tube was indeed the same as those
generated during atomization or reaction, the aerosol size
distribution was measured either through the flow tube or
through a ∼0.5 cm diameter conductive silicon tube under
otherwise identical conditions. The silicon tube has a much
smaller volume (∼17 cm3) than the flow tube (∼1000 cm3),
thus residence time in the tubing was significantly reduced,
eliminating possible size changes occurring during transit in the
flow tube. Particle sizes were measured at 25, 60, and 90% RH.
In each case, the surface area-to-volume ratios of the aerosol
flow were less than 5% variation from each other, with no clear
trend in the difference. Thus, we conclude that the aerosol loss
in the reactor is negligible.
An additional concern in particle measurement was possible
size changes associated with the mixing of the aerosol stream
with sheath air in the electrostatic classifier. To test this,
experiments were performed using the usual room air sheath
inlet (RH 42%) or encasing the inlet in a dry N2 flow to achieve
∼0% RH (measured by hygrometer). Experimental data were
then collected for aerosols at RH of 17, 47, and 95%. As in the
case of adjusting aerosol residence time, there was little effect
observed by modifying the sheath gas humidity, indicating that
the size distribution data reported by our instruments is an
accurate reflection of particle sizes in the reaction cell.
Aerosols were produced by the expansion of compressed, dry
N2 through an orifice, which then atomized liquid drawn from
the reservoir. The N2 carrier passed through a molecular sieve
trap in order to remove the impurities. Large droplets impacted
upon the wall of the atomizer and returned to the reservoir, while
a fine spray of particles left the generator. This flow was then
passed through a diffusion dryer (TSI, model 3062), which was
used to partially control the relative humidity of the final aerosol
flow and to allow large aerosol particles settling to the wall,
providing a more repeatable particle size distribution. Further
change of the aerosol humidity was achieved by adding water
vapor or dry N2 flows to the aerosol stream. Relative humidity
was measured at the bottom of the flow reactor with a fast-
response, digital hygrometer (VWR, model 35519-041), which
was protected from irreversible aerosol coating on the sensor
surface by a tube filter (Balston, DFU-9930-05 DF), inserted
into the exit tube of the flow reactor. It is noted that the relative
humidity measurements were performed in the absence of N2O5
in order to prevent the contamination of the sensor element.
Sulfuric acid aerosols were produced by a different method
than that described above (not shown in Figure 1) and based
solely on SO3 hydrolysis. The first step in aerosol production
was the mixing of SO3 vapor and moisturized N2 streams into
a glass reactor, allowing reaction and nucleation. SO3 was
supplied from fuming sulfuric acid (Aldrich, 26-29.5% SO3)
in a N2 carrier. This aerosol stream then passed into a
condensation chamber containing sulfuric acid (prepared from
concentrated sulfuric acid, J. T. Baker, 96.2%) of known
composition. The residence time in this chamber was long
enough to allow aerosol particles reaching equilibrium in the
exchange of water vapor above the solution. The time required
to reach equilibrium varies greatly as a function of temperature
and composition, limiting the range of aerosols that could be
produced by this source.
Flow Reactor and N2O5 Measurement. The experimental
setup for the reaction probability measurement is shown
schematically in Figure 3. The central component is the
atmospheric pressure flow reactor. The reactor is a 130-cm long
Pyrex tube with an inner diameter of 2.5 cm suspended
vertically. Although equipped with a jacket to allow cooling of
The characterization of aerosols was identical regardless of
aerosol type and composition. Following passage through the
flow reactor, the aerosol flow was channeled into an electrostatic
classifier (TSI, model 3071A). After exposure to a 85Kr bipolar
charger, aerosol particles were selected according to their

Uptake of N2O5 by (NH4)2SO4, NH4HSO4, and H2SO4
J. Phys. Chem. A, Vol. 105, No. 26, 2001 6467
The change of N2O5 concentrations is given in the following
equation:
d[N2O5]/dt ) -k[N2O5]A
(2)
where A is the surface area density (or the surface-to-volume
ratio) for sulfate aerosol. All measurements began 25 cm from
the bottom of the flow reactor to ensure a laminar flow being
fully developed (without the effects of turbulent mixing). The
slope aquired by the change in concentration vs time is the first-
order decay rate for a given experiement. Reaction times are
determined from the injector position and flow velocity accord-
ing to the relationship t ) l/V where V is the flow velocity (cm/
s) and l is the length of the stainless steel injector position
relative to the starting position (cm).
The calculated decay rates, k, are then corrected for axial
and radial diffusion effects in the flow tube according to the
treatment reported in the literature.17-19 The corrected decay
rate is designated to be kc. In the data analysis the gas-phase
20
diffusion coefficient of N2O5 in N2 is estimated using the
relationship
Dg ) (63.92/p) × (T/293)1.94
(3)
Figure 3. Schematic diagram of aerosol flow reactor. The reactor
interfaces to both an aerosol measurement apparatus shown in Figure
1 and a chemical ionization mass spectrometer.
where p is the pressure (Torr) and T the temperature (K) of the
reactor. Typically, the correction for diffusion effects is less
than 10% and hence the estimation of diffusion coefficient does
not cause any significant error in the determination of reaction
probability.
From the corrected decay rates, kc, the reaction probability
for a selected aerosol can be determined. A simple and
straightforward estimate of the uptake coefficient can be made
by using the aerosol size distribution
the tube, all experiments reported herein were conducted at room
temperature (295 ( 1 K).
N2O5 was synthesized at 295 K by the reaction NO2 with O3
in an O2 carrier and was collected in a Pyrex vessel at 196 K.
The sample was further purified with distillations in order to
remove impurities. O3 was prepared by using an electrical
discharge tube. During the experimental run, N2O5 was kept at
213 to 237 K in a methanol chiller (Haake, model A80). A
movable stainless steel inlet ran through the central axis of the
flow tube, allowing addition of N2O5 at variable distances. A
separate inlet carried the aerosol and N2 mixture into the tube.
All flow sources had bypass valves to remove chemicals as
needed without changing the total flow conditions.
Total flow rates in the tube varied between 1.0 and 2.5 L/min,
allowing the reaction to take place under laminar flow conditions
(Reynolds number Re ) 60-150). After passing through the
flow tube, the reactant stream was channeled either into the
electrostatic classifier or through a glass capillary for N2O5
detection by mass spectroscopy, with excess gas vented into
the exhaust hood.
Detection of N2O5 was accomplished by chemical ionization
mass spectrometry using CF3I as the ionizing gas.12,16 In the
ion flow-tube, CF3I passed over a rhenium ribbon, which was
heated by a direct current of 8-10 A, to produce I- ions. The
subsequent reaction of I- with N2O5 produced NO3-. The
detection sensitivity is about 1 × 1011 molecule cm-3 (S/N )
1, 1 s integration in the aerosol reactor). The product for reaction
1, HNO3, was also monitored by the reaction I- + HNO3 + M
f I-‚(HNO3) + M. The advantage in this experimental method
is that it can distinguish between N2O5 and HNO3 through the
selective ion-molecule reactions. Flow rates of CF3I and N2
carrier gas were regulated to reach a final pressure of ∼0.15
Torr in the ion flow-tube. Initial N2O5 concentrations in the
aerosol reactor were varied in the range of ∼5 × 1012 to ∼2 ×
1014 molecule cm-3 in the uptake experiments. However, there
was no observed difference in reaction probability.
4kc
γ )
(4)
Aω
where A is the surface area density of aerosols measured by
SMPS and ω the mean molecular velocity of N2O5. However,
all of the data shown in the next section are analyzed by a more
complete method developed directly from the SMPS data,21
γNiωπri2
k ) k(r ) )
(5)
∑
∑
i
c
i
0.75 + 0.283Kni
Kni(Kni + 1)
i
1 + γ
(
)
and the Knudsen number
Kni ) 3Dg/ωri
(6)
is calculated for various particle radii (ri) of aerosols shown in
Figure 2. Typically, the difference in γ values determined by
eqs 4 and 5 is less than 10-20% for the results shown in the
next section. It should be noted that we assume the shape of
aerosol is spherical in the data treatment. While the assumption
is true for liquid-phase aerosol, but it may underestimate the
surface area density of dry, solid sulfate aerosols. Further
discussion on this point will be given in the next section.
Results and Discussion
Data Collection and Treatment. The first step in determin-
ing the reaction probability of N2O5 by sulfate aerosols is the
calculation of the first-order decay rate, k. A typical experiment
measures the concentration of N2O5 at several injector positions.
Figure 4 shows typical changes in N2O5 signal as a function
of reaction time. As described above in the Experimental
Section, reaction time is calculated from changes in the injector
position and flow rates. The plot allows the calculation of the

6468 J. Phys. Chem. A, Vol. 105, No. 26, 2001
Kane et al.
TABLE 1: Reaction Probability Determinations for the
Uptake of N2O5 on Ammonium Sulfate Aerosols (The
uncertainty represents one standard deviation.)
RH (%)
γ
uncertainty
8
32
37
38
39
42
45
48
51
55
58
62
65
72
76
86
92
9.4 × 10-4
4.0 × 10-3
3.8 × 10-3
5.8 × 10-3
6.2 × 10-3
3.9 × 10-3
2.5 × 10-3
4.1 × 10-3
0.013
5.9 × 10-4
9.6 × 10-4
8.5 × 10-4
2.5 × 10-4
1.2 × 10-3
3.4 × 10-3
1.2 × 10-3
2.3 × 10-3
2.4 × 10-3
8.7 × 10-4
1.1 × 10-3
1.0 × 10-3
2.1 × 10-3
2.3 × 10-3
9.6 × 10-4
1.3 × 10-3
2.2 × 10-3
0.012
0.011
6.0 × 10-3
0.018
0.020
0.025
0.033
0.042
Figure 4. Typical N2O5 decay profiles with and without aerosol flow.
Aerosols are generated from 0.1 wt % ammonium bisulfate at 39%
RH. The experimental conditions are: p(total) ) 723 Torr, T ) 294
K, V ) 7.6 cm/s, surface area density ) 8.6 × 10-4 cm2/cm3, and [N2O5]
∼ 1.5 × 1014 molecules/cm3.
increases above 50% RH, N2O5 uptake increases from ∼0.01
to ∼0.04 at 92% RH. The polynomial fit of the data gives
γ[(NH4)2SO4] ) 2.79 × 10-4 + 1.30 × 10-4
×
(RH) - 3.43 × 10-6 × (RH)2 + 7.52 × 10-8 × (RH)3 (7)
Our results are supported by comparison to the phase transition
of ammonium sulfate aerosols described by Tang and Munkel-
witz22 and Xu et al.23 The aerosol has an efflorescence point
near 40% RH at 25 °C, above which the water content increases
steadily and below which particles are completely dry. As our
particles are formed in a wet environment and subsequently
dried, the expected water content is expected to match these
results. The closeness of the uptake data to the evaporation
profile of ammonium sulfate suggests that the water content of
the aerosol controls the N2O5 uptake. It is further noted that
the reaction primarily takes place with thin liquid water on the
surface of solid aerosols (e40% RH). On the other hand, the
hydrolysis may occur throughout the volume of liquid-phase
aerosols (>40% RH).
In the data treatment using eq 5 we assume the ammonium
sulfate aerosols are spherical. At humidity greater than 40%
RH, the sulfate aerosols are in liquid form and the assumption
is correct. But, under dry conditions (<40% RH) the aerosols
become solid and the crystal structure of ammonium sulfate is
orthorhombic. Since the surface area density on the basis of
this crystal structure is greater than that using the spherical
assumption, the γ values we report in Table 1 and Figure 5
may be smaller.
Data for uptake on ammonium sulfate collected by Hu and
Abbatt9 are represented with square symbols in Figure 5. Our
data does not match the trend of these results with RH, but
does fall within the same range of γ value (0.02-0.05) at high
RH. A limited set of data reported by Mozurkewich and Calvert6
are also shown in Figure 5 (solid triangles). Their result at 25%
RH is in excellent agreement with out data at low RH. However,
their data at 60% RH, needing to be reduced by ∼15% because
of a small error in the data treatment,6,7 is still substantially
greater than our data. The source of the discrepancy is unknown,
but probably associated with the preparation and conditions of
the aerosols.
Figure 5. Reaction probability measurements for N2O5 hydrolysis on
ammonium sulfate aerosol as a function of relative humidity. The
uncertainty represents one standard deviation of our data.
first-order decay of N2O5 under given conditions (k ) slope).
The upper curve (solid circles) indicates the decay of N2O5
without aerosol present, while the open circles represent the
uptake of N2O5 by sulfate (in this case, ammonium bisulfate).
To properly correct for the increased decay of N2O5 in various
RH conditions, the first-order decay rates as a function of RH
in the absence of aerosol are determined throughout the
experiment. The reaction probabilities on the wall of the reactor,
γw ) 2rkc/ω (r was defined as the radius of the flow reactor,
1.25 cm), were determined to be in the range of ∼2 × 10-6 at
low RH to ∼1 × 10-5 at high RH. The overall decay rate was
then corrected for the wall contributions, yielding only decay
rate caused by sulfate aerosol. This value is then corrected for
axial and radial diffusion as described above. From this value,
the reaction probability can be determined. This procedure is
adopted in every experiment discussed in the following sections.
Ammonium Sulfate. The uptake of N2O5 on ammonium
sulfate aerosols is shown in Figure 5 and also listed in Table 1.
The uptake on dry aerosols (under the condition of less than
40% RH) is found to be relatively small, <0.01. As the humidity
Ammonium Bisulfate. Experiments similar to the above were
also performed for ammonium bisulfate aerosols. The results
are shown in Figure 6 and also tabulated in Table 2. The uptake
of N2O5 increases steadily as the water content increases. Again

Uptake of N2O5 by (NH4)2SO4, NH4HSO4, and H2SO4
J. Phys. Chem. A, Vol. 105, No. 26, 2001 6469
Figure 7. Reaction probability measurements for N2O5 hydrolysis on
sulfuric acid aerosol as a function of relative humidity.
Figure 6. Reaction probability measurements for N2O5 hydrolysis on
ammonium hydrogen sulfate aerosol as a function of relative humidity.
The uncertainty represents one standard deviation of our data.
TABLE 3: Reaction Probability Determinations for the
Uptake of N2O5 on Sulfuric Acid Aerosols
TABLE 2: Reaction Probability Determinations for the
Uptake of N2O5 on Ammonium Bisulfate Aerosols (The
uncertainty represents one standard deviation.)
RH (%)
H2SO4 (wt %)
g
0.7
0.7
2.9
2.9
7.7
16.0
28.4
81
81
76
76
70
63
55
0.044
0.047
0.058
0.035
0.033
0.075
0.061
RH (%)
γ
uncertainty
13
15
21
29
32
34
36
37
38
43
48
56
63
67
74
83
89
99
1.5 × 10-3
1.8 × 10-3
4.8 × 10-3
1.1 × 10-3
3.1 × 10-3
0.014
4.2 × 10-4
1.4 × 10-3
1.8 × 10-3
1.0 × 10-3
2.3 × 10-3
5.2 × 10-3
2.4 × 10-3
1.1 × 10-3
3.2 × 10-3
1.3 × 10-3
2.7 × 10-3
3.1 × 10-3
1.9 × 10-3
2.9 × 10-3
0.011
value6-11 of N2O5 uptake at lower relative humidity (RH <
25%). If we exclude the data (shown in open circles) by
Mozurkewich and Calvert6 because the data analysis is in
error,6,7 we derive a fit for all data by linear regression:
3.2 × 10-3
1.9 × 10-3
0.012
0.015
0.018
0.018
0.022
0.028
0.035
0.060
0.046
γ[H2SO4] ) 0.052 - 2.79 × 10-4 × (RH)
(9)
The reaction probability on H2SO4 aerosol shows only a weak
dependence upon the relative humidity, in contrast to those on
ammonium-containing aerosols as shown in the previous
sections. The results reported here suggest that the reaction
mechanism in the uptake of N2O5 by these sulfate aerosols is
substantially different. Detailed discussion is given in the
following section.
5.1 × 10-3
0.012
0.069
9.5 × 10-3
we use the polynomial fit of our data, and the result gives
γ[NH4HSO4] ) 2.07 × 10-3 - 1.48 × 10-4
×
Reaction Mechanisms. The comparison of N2O5 uptake in
ammonium-containing sulfate aerosols with that of water is
shown in Figure 8. It appears that γ[NH4HSO4]is slightly greater
than γ[(NH4)2SO4] over the wide range of RH from 10 to100%.
Within experimental error, each set of data approaches the
uptake value of water (γ ∼ 0.05).14,25 The previous evaluation14
quoted the error limit at a factor of 2. However, the most recent
evaluation15 suggests that the γ value for water is ∼0.022 at
295 K. It is noted that the water content of aerosol strongly
affects the activity of N2O5 uptake as described in the previous
section. This is consistent with the suggestion of Mentell et
al.,26,27 who studied N2O5 uptake on Na+-neutralized sulfate
aerosols. In addition, the γ value for ammonium bisulfate is
slightly greater than that of ammonium sulfate, possibly due to
the phase change in ammonium sulfate aerosol at 40-80% RH
or the presence of a small quantity of H+ in the ammonium
bisulfate aerosol.
(RH) + 8.26 × 10-6 × (RH)2 (8)
The trend of reaction probability versus RH follows the water
activity of the ammonium bisulfate at room temperature.22-24
The observation suggests that hydrolysis is the dominant step
in the uptake of N2O5. A kinetic study was previously
investigated at 293 K by Mozurkewich and Calvert.6 The trend
in their data matches ours fairly well, but with somewhat larger
γ values overall. It should be noted, however, that subsequent
work by this group finds these values should be ∼15%
smaller,6,7 which would bring the two studies somewhat closer
to agreement.
Sulfuric Acid. The reactive uptake on sulfuric acid aerosols
has been examined extensively by a number of groups
previously,6-13 thus our examination was kept quite limited,
mainly for the confirmation of these previous results using our
apparatus. The data are shown in Figure 7 and also listed in
Table 3. Our data corresponds reasonably well with published
Recently Robinson et al.11 postulated a simple reaction
mechanism for the uptake of N2O5 on sulfuric acid aerosols

6470 J. Phys. Chem. A, Vol. 105, No. 26, 2001
Kane et al.
relative humidity (or the water content of aerosols) at room
temperature. Hydrolysis is believed to be the dominant reaction
step. For sulfuric acid aerosols, however, the reaction probability
decreases slightly as the relative humidity increases and we
believe that ionic reactions apparently play an important role.
At 50-100% RH, the γ values for ammonium bisulfate
aerosols are in the range 0.02 to 0.06. Thus, reaction 1 is very
efficient in converting NOx to HNO3 in the atmosphere. Using
a γ value of 0.1 after adopting the temperature dependence of
reaction probability, reaction 1 has an important effect on the
concentrations of NOx, O3, and OH in the global troposphere
as suggested by Dentener and Crutzen3.
Acknowledgment. This research was performed at the Jet
Propulsion Laboratory, California Institute of Technology, under
a contract with the National Aeronautics and Space Administra-
tion (NASA). Francois Caloz is grateful to the financial support
by the National Foundation of Switzerland.
Figure 8. Comparison of reaction probability measurements for N2O5
on water, ammonium bisulfate, and ammonium sulfate aerosols as a
function of relative humidity.
References and Notes
(1) Scientific Assessment of Ozone Depletion: 1998; Global Ozone
Research and Monitoring Project-Report No. 44; World Meteorological
Organization, Geneva, Switzerland; Chapter 3.
(2) AViation and the Global Atmosphere, IntergoVernmental Panel on
Climate Change; Cambridge University Press: New York, 1999; Chapter
2.
(3) Dentener, F. J.; Crutzen, P. J. J. Geophys. Res. 1993, 98, 7149.
(4) Huebert, B. J.; Lazrus, A. L. J. Geophys. Res. 1980, 85, 7337.
(5) Cziczo, D. J.; Abbatt, J. P. D. J. Geophys. Res. 1999, 104, 13781.
(6) Mozurkewich, M.; Calvert, J. G. J. Geophys. Res. 1988, 93, 15889.
(7) Fried, A.; Henry, B.; Calvert, J. G.; Mozurkewich, M. J. Geophys.
Res. 1994, 99, 3517.
and bulk substrates for a wide range of acid composition and
temperature, using a single consistent set of accommodation
coefficient (R) and liquid-phase reaction parameters (H*, Dl,
and kl). The mechanism includes the following steps:
N2O5 + H2O f 2HNO3
(10)
(11a)
(11b)
+
N2O5 + H+ f HNO3 + NO2
NO2+ + H2O f HNO3 + H+
This model is highly successful in explaining the temperature
dependence from 200 to 300 K and also the composition
dependence from 40 to 80 wt %. However, the result for H2-
SO4 aerosol is in contrast with those for ammonium-containing
sulfate aerosols as shown in Figures 5-7. Perhaps, the shortage
of H+ in ammonium-containing sulfate aerosols inhibits the
reaction step (eq 11a). Thus the hydrolysis step (eq 10) provides
the limiting step for the reactive uptake.
(8) Lovejoy, E. R.; Hanson, D. R. J. Phys. Chem. 1995, 99, 2080.
(9) Hu, J. H.; Abbatt, J. P. D. J. Phys. Chem. A 1997, 101, 871.
(10) Hallquist, M.; Stewart, D. J.; Baker, J.; Cox, R. A. J. Phys. Chem.
A 2000, 104, 3984.
(11) Robinson, G. N.; Worsnop, D. R.; Jayne, J. T.; Kolb, C. E.;
Davidovits, P. J. Geophys. Res. 1997, 102, 3583.
(12) Hanson, D. R.; Ravishankara, A. R. J. Geophys. Res. 1991, 96,
17307.
(13) Zhang, R.; Leu, M.-T.; Keyser, L. F. Geophys. Res. Lett. 1995,
22, 1493.
Atmospheric Implications. On the basis of the result for
ammonium bisulfate at 50-100% RH, the γ value is about
0.02-0.06 at 294 K. We use the temperature dependence
reported for ammonium bisulfate aerosol6 and for sulfuric acid
aerosol10,11 and estimate that the γ value could increase up to a
factor of 3 at lower temperatures between 260 and 280 K if
ammonium bisulfate droplet remains as a supercooled liquid.
Thus, for the theoretical study of the global tropospheric
chemistry we recommend the uptake coefficient of 0.1 to be
used in the model calculation. According to the model of
Dentener and Crutzen3, the decreases of NOx, O3, and OH are
49%, 9%, and 9%, respectively, due to the effect of reaction 1
as compared to that using gas-phase reactions alone.
(14) DeMore, W. B.; et al. JPL Publication 97-4, 1998.
(15) Sander, S. P.; et al. JPL Publication 00-3, 2000.
(16) Leu, M.-T.; Timonen, R. S.; Keyser, L. F.; Yung, Y. L. J. Phys.
Chem. 1995, 99, 13203.
(17) Brown, R. L. J. Res. Natl. Bur. Stand. (U.S.) 1978, 83, 1.
(18) Keyser, L. F. J. Phys. Chem. 1984, 88, 4750.
(19) Leu, M.-T. Geophys. Res. Lett. 1988, 15, 17.
(20) Monchick, L.; Mason, E. A. J. Chem. Phys. 1961, 35, 1676.
(21) Fuchs, N. A.; Sutugin, A. G. Highly Dispersed Aerosols; Ann Arbor
Science: Ann Arbor, MI, 1970.
(22) Tang, I. N.; Munkelwitz, H. R. J. Geophys. Res. 1994, 99, 18801.
(23) Xu, J.; Imre, D.; McGraw, R.; Tang, I. J. Phys. Chem. B 1998,
102, 7462.
(24) Imre, D.; Xu, J.; Tang, I.; McGraw, R. J. Phys. Chem. A 1997,
101, 4191.
(25) Van Doren, J. M.; et al. J. Phys. Chem. 1990, 94, 3265.
(26) Wahner, A.; Mentel, T. F.; Sohn, M.; Stier, J. J. Geophys. Res.
1998, 103, 31103.
(27) Mentell, T. F.; Sohn, M.; Wahner, A. Phys. Chem. Chem. Phys.
1999, 1, 5451.
Conclusions
The reaction probabilities of N2O5 on ammonium sulfate and
ammonium bisulfate aerosols are strongly dependent on the