Kinetics and thermodynamics of the gas phase reaction SO3 + NH3 + N2 ? H3NSO3 + N2

Article abstract of DOI:10.1021/jp970652i

The kinetics of the gas phase reaction SO₃ + NH₃ + N₂ ? H₃NSO₃ + N₂ were studied with a flow reactor coupled to a chemical ionization mass spectrometer for detection of SO₃ and H₃NSO₃. The rate coefficient for the association reaction SO₃ + NH₃ + N₂ was measured as a function of temperature (280-340 K) and pressure (20-80 Torr of N₂). Analysis of the SO₃ decay and H₃NSO₃ appearance at higher temperatures yielded rate coefficients for H₃NSO₃ decomposition and the enthalpy of the reaction SO₃ + NH₃ ? H₃NSO₃: ΔH°_298K = -24 ± 1 kcal mol^-1.

Full text of DOI:10.1021/jp970652i

4950
J. Phys. Chem. A 1997, 101, 4950-4953
Kinetics and Thermodynamics of the Gas Phase Reaction SO3 + NH3 + N2 T
H3NSO3 + N2
Edward R. Lovejoy
NOAA Aeronomy Laboratory, 325 Broadway, Boulder, Colorado 80303
ReceiVed: February 24, 1997; In Final Form: April 15, 1997X
The kinetics of the gas phase reaction SO3 + NH3 + N2 T H3NSO3 + N2 were studied with a flow reactor
coupled to a chemical ionization mass spectrometer for detection of SO3 and H3NSO3. The rate coefficient
for the association reaction SO3 + NH3 + N2 was measured as a function of temperature (280-340 K) and
pressure (20-80 Torr of N2). Analysis of the SO3 decay and H3NSO3 appearance at higher temperatures
yielded rate coefficients for H3NSO3 decomposition and the enthalpy of the reaction SO3 + NH3 T H3NSO3:
∆H° ) -24 ( 1 kcal mol-1.
298K
Introduction
of H3NSO3. The implications of these results to the understand-
ing of the atmospheric chemistry of H3NSO3 are discussed.
The dominant loss mechanism for SO3 in the atmosphere is
reaction with gas phase water to produce sulfuric acid.1-3 The
association reaction of SO3 with ammonia (eq 1) typically
Experimental Section
The kinetics of the reaction of SO3 with NH3 were studied
by monitoring the concentration of SO3 at the exit of a laminar
flow reactor as a function of the contact distance between SO3
and NH3. SO3 was monitored with chemical ionization mass
spectrometry.3 The reactor was a Pyrex cylinder 50 cm long
with a 1.62 cm internal diameter. The temperature of the reactor
was controlled by circulating temperature-regulated fluid through
copper tubing soldered to an insulated copper sleeve surrounding
the reactor. The reactor temperature was measured with a
thermocouple mounted on the end of a movable inlet. The
temperature gradient was less than 2 K along the reaction zone
for the conditions of the present work. SO3 was generated near
the end of the movable inlet by oxidizing SO2 on a hot nichrome
filament in the presence of O2, as described previously.3
Ammonia was added to the reactor at the upstream end along
with the main flow of N2. Mixtures of NH3 in N2 were prepared
in 12 L Pyrex bulbs by adding measured pressures of NH3 and
N2. The reactor pressure was measured with a capacitance
manometer and gas flows were measured with mass flow meters.
The mass flow meters were calibrated with a wet test meter or
by measuring the rate of change of pressure in a calibrated
volume.
SO3 + NH3 + M T H3NSO3 + M
(1)
accounts for less than 10-4 of the SO3 loss in the atmosphere.3,4
H3NSO3 appears to have an unusually strong affinity for H2-
SO4, and it has been proposed that the association of H3NSO3
and H2SO4 may be an important step in the formation of aerosol
in the atmosphere despite the low concentrations of H3NSO3.4
The stability of H3NSO3 is not well established. High-level
calculations by Wong et al. predict that H3NSO3 is bound by
about 19 kcal mol-1 relative to NH3 and SO3.5 Lovejoy and
Hanson4 report that the H3N-SO3 bond enthalpy is greater than
20 kcal mol-1, based on the observation of exponential SO3
decays in excess NH3 at room temperature.
The theoretical work by Wong et al.5 predicts that gas phase
H3NSO3 is a zwitterion with significant electron donation from
the N to S leading to a large dipole moment of 6.6 D. Wong et
al.5 also predict that gas phase H3NSO3 has a significantly longer
(about 0.1 Å) N-S bond than the solid form. Canagaratna et
al.6 have measured the microwave spectrum of gas phase H3-
NSO3 and find that the structure agrees very well with the
theoretical predictions of Wong et al.5 They also confirm that
H3NSO3 is a zwitterion with about 0.4 electrons transferred from
N to S.
Shen et al.7 reported the first kinetic measurements for the
reaction SO3 + NH3. They measured a rate coefficient of (6.9
( 1.5) × 10-11 cm3 molecule-1 s-1 in 1-2 Torr of He with a
flow reactor and photofragment-emission detection of SO3.
Lovejoy and Hanson4 have recently reported measurements of
the pressure dependence of the association rate coefficient and
observation of the H3NSO3 product.
The CIMS ion-molecule reactor was operated at 5 Torr with
about a 0.1 s contact time with the SO3 reactor effluent. SO3
was detected by using the following reactions3,8
-
-
NO3 HNO3 + SO3 f NO3 SO3 + HNO3
(2)
and
SiF5- + SO3 f FSO3- + SiF4
(3)
In the present work, the kinetics of the loss of SO3 in the
presence of NH3 are measured as a function of temperature
(280-343 K) and pressure (20-80 Torr of N2). At elevated
temperatures (383-402 K) the decomposition of H3NSO3 is
observed, and equilibrium constants for reaction 1 are measured.
These data yield the temperature dependence of the SO3 + NH3
reaction near the low-pressure limit and the heat of formation
H3NSO3 was detected by using
-
-
NO3 HNO3 + H3NSO3 f H2NSO3 HNO3 + HNO3 (4)
Initial SO3 concentrations in the neutral flow reactor ranged
from about (1-10) × 109 molecule cm-3. The NH3 concentra-
tion was typically at least 50 times larger than [SO3 ] so that
X Abstract published in AdVance ACS Abstracts, June 15, 1997.
S1089-5639(97)00652-X This article not subject to U.S. Copyright. Published 1997 by the American Chemical Society

SO3/NH3/N2 Gas Phase Reaction
J. Phys. Chem. A, Vol. 101, No. 27, 1997 4951
k0(T)[M]
k(M, T) )
0.6y
(5)
k0(T)[M]
k∞(T)
1 +
(
)
2
-1
k0(T)[M]
k∞(T)
y ) 1 + log
(
(
) )
]
[
where k0(T) ) k300(T/300)-n and k∞(T) ) k300(T/300)-m. A fit
0
∞
of the k(M,T) data (Table 1) to eq 5, including the previous
measurements4 at 295 K, gives k3000 ) (3.6 ( 0.4) × 10-30 cm6
molecule-2 s-1, k∞300 ) (4.3 ( 1.2) × 10-11 cm3 molecule-1
s-1, and n ) 6.1 ( 1.0, with m fixed equal to zero. The errors
reflect an estimated 20% uncertainty in the measured rate
coefficients. The fit results changed by less than 10% for 0 <
m < 3. The assumption that the high-pressure limiting rate
coefficient is independent of temperature (i.e. m ) 0) is
reasonable considering that the high-pressure limit rate coef-
ficient is large at room temperature and the temperature
dependence of k∞ is generally weak (see, for example, ref 10).
At temperatures above 380 K the unimolecular decomposition
of H3NSO3 into SO3 + NH3 competed efficiently with the
association reaction, and the SO3 decays were not simple
exponentials (see Figure 2). For these conditions the SO3 decay
and H3NSO3 appearance were modeled to extract the rate
coefficient for decomposition of H3NSO3. The kinetics were
modeled with the following mechanism
Figure 1. First-order rate coefficient for SO3 loss as a function of
[NH3] and temperature. Circles ) 280 K, 41.4 Torr. Squares ) 301 K,
41.0 Torr. Triangles ) 334 K, 40.6 Torr.
TABLE 1: Summary of SO3 + NH3 Rate Coefficients
[NH3]max
kfb (10-12 cm3
p(N2)
(Torr) T (K) (cm s-1
Va
(1012 molecules no. of kI
molecule-1
)
cm-3
)
measmnts
s-1
)
20.5
20.1
21.1
20.2
20.5
41.4
41.0
41.0
40.6
40.0
81.3
79.8
80.4
280
298
317
317
341
280
301
319
334
343
280
300
320
391
423
445
441
472
197
214
237
240
247
100
107
113
53
81
124
118
133
19
43
72
68
88
7
8
9
6
8
8
8
7
6
6
6
7
8
2.76 (0.06)
1.90 (0.10)
1.34 (0.14)
1.35c (0.08)
0.87c (0.05)
4.58 (0.16)
3.16 (0.06)
2.30 (0.14)
1.70 (0.08)
1.47 (0.10)
8.2 (0.3)
SO3 + NH3 + M T H3NSO3 + M
(kf, kr)
(6)
SO3
SO3 + wall f (kw
)
(7)
(8)
4.9
7.1
13
5.04 (0.34)
3.47 (0.22)
SA
H3NSO3 + wall f (kw
)
a Average reactor flow velocity. b 95% confidence levels for precision
For this mechanism [SO3] and [H3NSO3] are described by
-
are indicated in parentheses. c SiF5 reagent ion.
[SO ]
3 0[(a + c)exp(at) - (b + c)exp(bt)] (9)
the SO3 loss was pseudo first order in SO3. The Reynolds
number for the neutral reactor flow was about 75.
[SO3] )
a - b
and
Results and Discussion
kf[NH3][SO ]
Rate coefficients were measured by monitoring the concen-
tration of SO3 at the exit of the flow reactor as a function of
the contact distance with NH3 for a range of NH3 concentrations.
The NO3-HNO3 reagent ion was used for the majority of the
measurements. At temperatures below 350 K the SO3 decays
were exponential, as expected for a first-order process, and as
observed previously.4 First-order SO3 loss rate coefficients were
extracted from the decays by using the Brown algorithm.9 First-
order rate coefficients are plotted as a function of the concentra-
tion of NH3 for a range of temperatures in 40 Torr of N2 in
Figure 1. The slopes of these plots yield the effective second-
order rate coefficients for SO3 + NH3. All of the measured
association rate coefficients are listed in Table 1. The quoted
errors are the 95% confidence levels for precision only. The
overall uncertainty in the rate coefficients is estimated to be
(20%. The room temperature rate coefficients measured in
this work are in excellent agreement with the previous measure-
ments of Lovejoy and Hanson,4 but are about 100 times smaller
than the original measurement by Shen et al.7 Heterogeneous
loss of SO3, possibly related to the relatively high levels of SO3,
may have influenced the Shen et al.7 measurement.
[H3NSO3] )
3 0[exp(at) - exp(bt)] (10)
a - b
where
a ) 0.5[(d2 - 4e)1/2 - d]
b ) -0.5[(d2 - 4e)1/2 + d]
c ) kr + kSwA
d ) kf[NH3] + kr + kSwO3 + kwSA
e ) kSwAkf[NH3] + krkwSO3 + kwSAkSwO3
The SO3 decay and H3NSO3 appearance were fit simulta-
neously to these equations with kf fixed at the value extrapolated
from the lower temperature data. The reaction time was
calculated using the relationship t ) z/(1.7V) where z is the
reaction distance and V is the average flow velocity.4 The rate
coefficient for SO3 wall loss was fixed at the value measured
in the absence of NH3. This value was always within 10% of
the diffusion-limited value, implying that the SO3 reaction
probability with the reactor wall was greater than 10-3. The
nonlinear regression variables included the first-order H3NSO3
The pressure and temperature dependence of the rate coef-
ficient for a three-body reaction may be described by the Troe
formalism10

4952 J. Phys. Chem. A, Vol. 101, No. 27, 1997
Lovejoy
TABLE 3: Predicted H3NSO3 Decomposition Lifetimes for
Tropospheric Conditions
pressure
(Torr)
kf (10-11 cm3
Ta (K) molecule-1 s-1
H3NSO3 decomp
lifetime (s)
)
kr (s-1
)
760
405
195
281
254
222
2.2
2.3
2.5
9 × 10-3
110
9100
1.1 × 10-4
1.4 × 10-7
7 × 106
a On the basis of a model atmosphere for 40° N in March (ref 15).
rameters and scaled vibrational frequencies.5 The entropies of
SO3 and NH3 were taken from the literature.11 These values
yielded an entropy change for the formation of H3NSO3 (reaction
1) of -0.036 kcal mol-1 K-1 for 250 K < T < 400 K.
Calculations also showed that the change in heat capacity for
reaction 1 is negligible for 300 K < T < 400 K (|∆Cp| < 1 cal
mol-1 K-1), so that ∆H° is essentially independent of temper-
ature between 300 and 400 K.
In the equilibrium experiments, data were taken in the
entrance and mixing region immediately downstream of the SO3
moveable source (see Figure 2). It should be noted that at lower
temperatures SO3 decays were linear in this region, even though
the radial velocity and concentration gradients are not expected
to be fully developed until about 5-10 cm downstream of the
inlet.12 On the basis of the poorly defined reaction time at short
reaction distances, the uncertainties in the measured equilibrium
constants are estimated to be about a factor of 2. This yields a
reaction enthalpy for SO3 + NH3 T H3NSO3 of ∆H2°98K ) -24
( 1 kcal mol-1 and a heat of formation for H3NSO3 of ∆Hf°,298K
) -129.6 ( 1.0 kcal mol-1, where the standard state is 1 atm.
This experimental H3N-SO3 bond enthalpy is about 4 kcal
mol-1 larger than the theoretical value.5
Figure 2. SO3 and H3NSO3 signals (NO3-SO3 and H2NSO3-HNO3)
as a function of reaction distance and [NH3]. Reactor conditions: V )
288 cm s-1, 393 K, 40.0 Torr of N2. Squares: [NH3] ) 9 × 1013
molecule cm-3. Circles: [NH3] ) 1.8 × 1014 molecule cm-3. Solid
lines are fits to the data as described in the text. Fit results are listed
in Table 2.
TABLE 2: Summary of SO3 + NH3 +N2 T H3NSO3 + N2
Equilibrium Measurements
kf (10-13 cm3
[NH3]
∆H°
p(N2)
(Torr) T (K)
molecule-1 (1013 molecule
V
kr
(kcal
s-1
)
cm-3
)
(cm s-1) (s-1) mol-1
)
40.8 393
40.8 393
40.8 393
40.4 402
40.4 402
40.4 402
40.0 393
40.0 393
40.3 383
40.3 383
40.3 383
40.6 393
40.6 393
40.6 393
40.6 393
79.1 383
79.1 383
79.1 383
79.1 383
41.0 393
41.0 393
41.0 393
7.4
7.4
7.4
6.5
6.5
6.5
7.3
7.3
8.3
8.3
8.3
7.4
7.4
7.4
7.4
5
9
281
281
281
295
295
295
288
288
279
279
279
280
280
280
280
138
138
138
138
272
272
272
44 -24.0
41 -24.1
38 -24.1
97 -23.8
94 -23.9
117 -23.7
41 -24.1
37 -24.2
22 -24.1
20 -24.1
21 -24.1
33 -24.3
38 -24.1
39 -24.1
40 -24.1
33 -24.1
48 -23.8
35 -24.0
47 -23.8
36 -24.2
41 -24.1
35 -24.2
18
12
12
25
18
9
9
6
2
10
14
19
13
9
19
5
14
9
14
5
Atmospheric Implications
The lifetime of H3NSO3 with respect to unimolecular
decomposition for several conditions of pressure and temperature
characteristic of the troposphere are listed in Table 3. Other
potential atmospheric loss processes for H3NSO3 include
scavenging by aerosol and clustering with H2SO4. The hetero-
geneous reaction probability for H3NSO3 is probably near unity.
In this case the lifetime of H3NSO3 with respect to aerosol
scavenging will range from seconds in clouds to many hours
13
13
13
13
7.4
7.4
7.4
4
in clean air.13 Assuming a very efficient reaction with H2SO4
and ambient [H2SO4] ranging from 106 to 107 molecule cm-3 14
,
yields a lifetime of H3NSO3 with respect to clustering with
H2SO4 of about 103-104 s. This analysis shows that the
unimolecular decomposition of H3NSO3 is most important in
the lower troposphere for clean conditions. In the free
troposphere the dominant loss processes for H3NSO3 are
probably scavenging by aerosol and clustering with H2SO4. In
order to understand the role of H3NSO3 in the nucleation of
atmospheric particles, the kinetics of production and decomposi-
tion of clusters of the form H3NSO3(H2SO4)x are needed.
decomposition rate coefficient kr, the H3NSO3 wall loss
(kSwA), and the CIMS sensitivity for SO3 relative to H3NSO3. A
set of experimental SO3 and H3NSO3 profiles and the simul-
taneous fits to eqs 9 and 10 are presented in Figure 2. All the
equilibrium measurements are summarized in Table 2. The rate
coefficient for wall loss of H3NSO3 was consistently about 40%
less than the diffusion-limited value, implying that at these
elevated temperatures the wall reaction probability for H3NSO3
was reduced and/or evaporation of H3NSO3 was important. The
fitted wall loss rate coefficients for H3NSO3 varied by less than
about 20% for the range of NH3 concentrations used in this
work. The CIMS sensitivities for H3NSO3 and SO3 were a
function of the concentration of HNO3 in the ion-molecule
reactor. For the conditions of the present work, the CIMS was
typically 1-3 times more sensitive to H3NSO3 than SO3.
The equilibrium constant could only be measured over a
limited range of temperatures because of the very strong
temperature dependence of the rate coefficient for decomposition
of H3NSO3. Therefore, the reaction enthalpy was derived by
using a third-law analysis. The entropy of H3NSO3 was
calculated from experimental6 and theoretical5 structural pa-
Acknowledgment. This work was supported in part by the
NOAA Climate and Global Change Program.
References and Notes
(1) Reiner, T.; Arnold, F. J. Chem. Phys. 1994, 101, 7399.
(2) Kolb, C. E.; Jayne, J. T.; Worsnop, D. R.; Molina, M. J.; Meads,
R. F.; Viggiano, A. A. J. Am. Chem. Soc. 1994, 116, 10314.
(3) Lovejoy, E. R.; Hanson, D. R.; Huey, L. G. J. Phys. Chem. 1996,
100, 19911.
(4) Lovejoy, E. R.; Hanson, D. R. J. Phys. Chem. 1996, 100, 4465.
(5) Wong, M. W.; Wiberg, K. B.; Frisch, M. J. J. Am. Chem. Soc.
1992, 114, 523.

SO3/NH3/N2 Gas Phase Reaction
J. Phys. Chem. A, Vol. 101, No. 27, 1997 4953
(6) Canagaratna, M.; Phillips, J. A.; Goodfriend, H.; Leopold, K. R. J.
Am. Chem. Soc. 1996, 118, 5290.
(7) Shen, G.; Suto, M.; Lee, L. C. J. Geophys. Res. 1990, 95, 13981.
(8) Arnold, S. T.; Morris, R. A.; Viggiano, A. A.; Jayne, J. T. J.
Geophys. Res. 1995, 100, 14141.
(11) Chase, M. W., Jr.; Davies, C. A.; Downey, J. R., Jr.; Frurip, D. J.;
McDonald, R. A.; Syverud, A. N. JANAF Thermochemical Tables. J. Phys.
Chem. Ref. Data Suppl. 1 1985, 14.
(12) Keyser, L. F. J. Phys. Chem. 1984, 88, 4750.
(13) Lovejoy, E. R.; Huey, L. G.; Hanson, D. R. J. Geophys. Res. 1995,
100, 18775.
(9) Brown, R. L. J. Res. Natl. Bur. Stand. 1978, 83, 1.
(10) 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 Data for Use in Stratopheric Modeling; JPL
publication 94-26; Jet Propulsion Laboratory: Pasadena, CA, 1994.
(14) Weber, R. J.; McMurry, P. H.; Eisele, F. L.; Tanner, D. J. J. Atmos.
Sci. 1995, 52, 2242.
(15) Houghton, J. T. The Physics of Atmospheres; Cambridge University
Press: Cambridge, 1986.