Reaction of gaseous nitric oxide with nitric acid on silica surfaces in the presence of water at room temperature

Article abstract of DOI:10.1021/jp012330r

The reaction of gaseous NO with HNO₃ on borosilicate glass in the presence of water was studied as a function of surface water coverage ar 298 K anda total pressure of one atm in N₂. The loss of gaseous NO and the formation of NO₂ were measured in a long path cell using FTIR. The glass walls of the cell provided the surface upon which the chemistry occurred. Water coverages on thin glass cover disks were determined in a separate apparatus by measuring the intensity of the infrared band of liquid water ar 3400 cm^-1. Approximately one monolayer was present on the surface at 20% RH and 12 monolayers at 100% RH. The rate of the reaction of NO with HNO₃ on the surface was the largest under conditions where approximately three surface monolayers of water were present on the surface. We propose a model for this reaction in which HNO₃, added first to the dry cell, hydrogen-bonds to the silanol groups on the surface. The first step in the reaction is believed to be HNO_3(surface) + NO_(g) → HONO_(surface) + NO_2(g). Subsequently, HONO on the surface reacts with HNO₃ to generate solvated N₂O₄ as a product. Dissociation of N₂O₄ generates NO₂ as the final gas phase product. This chemistry is potentially important in renoxification of the boundary layer of polluted urban atmospheres where silica surfaces are plentiful in particles, soils and building materiais, as well as globally in the free troposphere where dust particles are present.

Full text of DOI:10.1021/jp012330r

J. Phys. Chem. A 2001, 105, 10339-10346
10339
Reaction of Gaseous Nitric Oxide with Nitric Acid on Silica Surfaces in the Presence of
Water at Room Temperature
N. A. Saliba, H. Yang, and B. J. Finlayson-Pitts*
Department of Chemistry, UniVersity of California, IrVine, IrVine, California 92697-2025
ReceiVed: June 19, 2001; In Final Form: August 27, 2001
The reaction of gaseous NO with HNO
function of surface water coverage at 298 K and a total pressure of one atm in N
and the formation of NO were measured in a long path cell using FTIR. The glass walls of the cell provided
the surface upon which the chemistry occurred. Water coverages on thin glass cover disks were determined
3
on borosilicate glass in the presence of water was studied as a
2
. The loss of gaseous NO
2
-
1
in a separate apparatus by measuring the intensity of the infrared band of liquid water at 3400 cm .
Approximately one monolayer was present on the surface at 20% RH and 12 monolayers at 100% RH. The
rate of the reaction of NO with HNO
three surface monolayers of water were present on the surface. We propose a model for this reaction in
which HNO , added first to the dry cell, hydrogen-bonds to the silanol groups on the surface. The first step
3
on the surface was the largest under conditions where approximately
3
in the reaction is believed to be HNO3(surface) + NO(g) f HONO(surface) + NO2(g). Subsequently, HONO on the
surface reacts with HNO to generate solvated N O as a product. Dissociation of N O generates NO as the
3 2 4 2 4 2
final gas phase product. This chemistry is potentially important in “renoxification” of the boundary layer of
polluted urban atmospheres where silica surfaces are plentiful in particles, soils and building materials, as
well as globally in the free troposphere where dust particles are present.
I. Introduction
of the reaction was proposed in two steps as described below
More than five decades of laboratory studies have shown that
oxides of nitrogen react on surfaces in the presence of water.
For example, Smith noted during gas-phase studies of the
HNO3(surface) + NO f NO + HONO
(2)
(3)
(4)
(5)
2
(surface)
1
followed by subsequent reactions such as
reaction of NO with HNO3 that there appeared to be a surface
reaction dependent on water vapor. Such heterogeneous reac-
tions may be potentially important in the atmosphere where
oxides of nitrogen, present as air pollutants from combustion
processes, are in contact with many surfaces in the form of
suspended particles as soil, roads, buildings, and plants.
Despite the well-recognized occurrence of heterogeneous
chemistry for the oxides of nitrogen, the kinetics and mecha-
nisms of these reactions are not well understood. One difficulty
arises in simultaneously measuring surface and gas-phase
reaction species. However, recent infrared spectroscopic studies
of heterogeneous hydrolysis of NO2 on silica surfaces conducted
HNO3(surface) + HONO(surface) f 2NO + H O
2
2
(surface)
or
2
HONO(surface) f NO + NO + H O
(surface)
2
2
The net reaction is
HNO3(surface) + NO f 3 NO + H O
(surface)
2
2
2
2
3
by Grassian and co-workers as well as by this laboratory,
showed that this well-known4-16 reaction
If all reactants and products are in the gas phase, reaction 2 is
close to thermoneutral. However, the free energy change for
reaction 2 with typical atmospheric concentrations of the
reactants and products is negative, so that it has been proposed
to be potentially feasible on surfaces in the atmosphere.18
surface
2
NO + H O
8 HONO + HNO3
(1)
2
2
produces N2O4 as a key intermediate on the surface. HNO3 was
also observed spectroscopically on the surface as hypothesized
in the previous studies4 where only gas-phase measurements
could be made, or nitrate ions measured in washings from the
surface after reaction.
In subsequent studies, Mochida and Finlayson-Pitts17 showed
that gaseous NO reacts with HNO3 on a “wet” porous glass to
generate nitrogen dioxide as the major gas-phase product, along
with small amounts of gas-phase HONO. Thus, the mechanism
Nitric acid is known to be readily taken up on a variety of
19-22
surfaces, e.g., soil and its components such as silica.
Model
suggest that this uptake could impact NOx and O3
in the troposphere. This process could also contribute to
-16
2
3,24
studies
“
renoxification” of the atmosphere and better reconcile field
2
5,26
and modeling experiments,
especially if HNO3 produces
photochemically reactive species such as NO2 and HONO. Thus,
a discrepancy has been reported between the measured ratio of
[
1
HNO3]/[NOx] ≈ 5 in the free troposphere and the values of
2
5,26
5-100 predicted by models.
Although there are several
*
To whom correspondence should be addressed: E-mail bjfinlay@uci.edu.
Phone: (949) 824-7670. Fax: (949) 824-3168.
hypotheses regarding the source of this discrepancy such as
1
0.1021/jp012330r CCC: $20.00 © 2001 American Chemical Society
Published on Web 10/20/2001

10340 J. Phys. Chem. A, Vol. 105, No. 45, 2001
Saliba et al.
Figure 1. Cell used to measure water coverage on the borosilicate cover glass disks.
liquid-phase reactions of HCHO with HNO3 in aerosols and
resolution with 150 co-added scans and a total scan time of 3.9
min. The cell consists of a borosilicate glass cylinder (10 cm
diameter × 91.4 cm length) and two stainless steel rods holding
the mirrors (Al with a silicon monoxide protective coating)
which are attached to two stainless steel plates at each end of
the cell. To avoid reactions of the gases with the stainless steel,
the metal surfaces were coated with halocarbon wax (Halocarbon
Products Corp., Series 1500). The optical base path length was
0.8 m, with a total path length of 38.4 m. The long path cell
was wrapped in a dark cloth to prevent photolysis of reactants
and products.
cloud droplets, or reactions on soot,2
unknown.
5-27
the cause remains
To better understand the uptake and potential reactions of
HNO3 on surfaces, we have carried out further studies of the
reaction of gaseous NO with HNO3 on a smooth borosilicate
glass surface as a function of varying amounts of surface-
adsorbed water. The relationship between the gas phase water
vapor concentration and the amount of water on the surface
was established in a newly designed experimental apparatus
using transmission FTIR and thin cover glass disks. These
studies provide insight into the reactive forms of nitric acid on
the surface and the reaction mechanism. The atmospheric
implications are discussed.
Dry, gaseous HNO3 obtained from the vapor above an HNO3/
H2SO4 mixture (1:2 v:v) was first admitted to the cell. HNO3
was allowed to adsorb onto the cell walls over five minutes.
The remaining gas-phase HNO3 was then pumped out, and this
conditioning/adsorbing process was repeated at least three times.
II. Experimental Section
1
5
-3
An NO concentration of (0.65-40) × 10 molecule cm was
then added into the cell as a mixture with N2. Initial relative
humidities of 0, 20, 30, 40, 50 and 70% were obtained by adding
a portion of dry N2 followed by bubbling N2 (100% humid)
through a fritted glass immersed in water to give a total pressure
of 1 atm. Gaseous reactants and products in the long path cell
were measured using FTIR starting immediately after the
addition of the reactants, for up to 350 min reaction time. Loss
of NO and formation of gaseous NO2 and HONO were
measured using their absorption bands at 1876, 2900, and 1264
A. Measurements of Water Coverage on Thin Borosilicate
Glass Disks. The amount of liquid water adsorbed on thin cover
glass disks at different relative humidities (RH) was determined
by transmission infrared spectroscopy using the cell shown in
Figure 1. The thin cover glass samples were thin Micro Cover
Glasses (VWR Scientific, Inc.) with 0.13 to 0.17 mm thickness
and 25 mm diameter. The cell, made of Pyrex glass, was 3.2
cm in diameter, 11 cm in length and capped with infrared-
transmitting ZnSe windows. To increase the weak adsorbed H2O
signal, five disks of cover glass were placed in thin slots along
a U-shaped glass rod, giving a total of 10 glass surfaces for
water uptake.
-1
cm , respectively.
Spectra of these species were quantitatively analyzed using
a least-squares fitting procedure described in detail by Gomer
A mixture of water vapor in N2 at various relative humidities
was generated by diluting a 100% RH stream, obtained by
bubbling N2 through Nanopure water (Barnstead, 18 MΩ cm),
with dry N2. The flow rates were controlled by calibrated
Matheson TF 1050 flowmeters. Spectra were collected at 0.5
28
et al. The concentration of each species is determined relative
to a reference spectrum of known concentration. Absolute
concentrations for NO and NO2 reference spectra were deter-
mined using calibrations of the pure gases. Nitrous acid was
-1
quantified using infrared cross sections for 1264 cm peak
-1
cm resolution with 1024 co-added scans and a total scan time
of 14.5 min. A background spectrum was obtained after the
cell and thin cover glass disks had been purged with dry N2 for
29,30
determined by Barney et al.
in this laboratory.
Materials. HNO3 was 70.1 wt % (Fisher) and H2SO4 was
9
5.8 wt % (Fisher). Nitric oxide (Matheson 99%) was purified
2
4 h. Reference spectra of gas-phase water at different relative
by passing it rapidly through a liquid nitrogen trap. The N2 was
9.999% (Oxygen Services Company) and used as received.
humidities were measured without the cover glass disks and
subtracted before integration.
9
B. Reaction of Gaseous NO with HNO3 On a Borosilicate
Glass Surface. These experiments were performed in a long
path infrared cell mounted vertically in the sample compartment
of an FTIR spectrometer (Mattson, Cygnus) and equipped with
an MCT detector. All experiments were carried out at 1 cm-
III. Results and Discussion
A. Water Coverage on Glass. Figure 2a shows typical
infrared spectra in the 3800 to 2800 cm region where
absorptions due to the stretching vibrations of water occur. The
-
1
1

Reaction of Gaseous Nitric Oxide with Nitric Acid
J. Phys. Chem. A, Vol. 105, No. 45, 2001 10341
Figure 2. (a) Typical absorption spectra of water adsorbed on thin cover glass disks at different relative humidities and at room temperature; (b)
-
1
spectra from (a) plotted at lower resolution (4 cm ) with gas-phase water subtracted and smoothing of the spectra.
-
1
ν1 stretch of gas-phase water is centered at 3652 cm and the
toward lower wavenumbers in the presence of more surface
water reflects a trend in which water changes from strong
interactions with the surface with some hydrogen-bonding to
adjacent water molecules, to three-dimensional water hydrogen-
bonding as is the case of the bulk liquid. This is similar to the
-
1 31
asymmetric ν3 stretch at 3756 cm . These bands appear as a
series of sharp rotational lines superimposed on a broad band
-
1
centered at ∼3400 cm at the highest water coverages. The
broad band is due to liquid water, and is red-shifted by up to
3
3,34
2
00 wavenumbers compared to the gas phase due to intermo-
effects observed by Ewing and co-workers
on the infrared
32
lecular hydrogen bonding; the shift in band position from the
gas to the liquid is also accompanied by an increase in the
absorption coefficient.32
spectrum of water adsorbed on NaCl crystals, in which the center
-
1
-1
of the 3400 cm band was red-shifted to 3500 cm at
submonolayer coverages.
Figure 2b shows more clearly the surface water band. These
spectra were obtained by subtracting from Figure 2a the
contribution from gas-phase water, converting the spectra to a
lower resolution (4 cm-1 instead of 0.5 cm which is adequate
for this broad band), and smoothing them. These show that the
surface water peak shifts from ∼3600 cm to ∼3400 cm as
the water coverage increases, and at 100% RH, the spectrum
becomes indistinguishable from that of liquid water. The shift
The number of monolayers (ML) of adsorbed H2O on glass
as a function of relative humidity was calculated from the
-
1
integrated absorbance, A˜ (cm ), and the known integrated
-1
33-35
-17
absorption coefficient for liquid water,
cm molecule (base 10)
σj ) 6.1 × 10
-
1
-
1
-1
1
5
ML ) A˜ /(1.0 × 10 N σj )
(I)
where N ) 10 is the number of thin cover glass surfaces and

10342 J. Phys. Chem. A, Vol. 105, No. 45, 2001
Saliba et al.
2
Figure 3. Number of monolayers of adsorbed H O on cover glass as
a function of relative humidity. The solid line is a fit to the data and
the dotted line shows a BET isotherm for multilayer adsorption.
15
1
× 10 molecule cm-2 is the surface density of one monolayer
2
36
of water, based on an area per water molecule of 10 Å . An
integrated absorption coefficient of σj ) 6.1 × 10
-
17
cm
-
1
molecule was used for all water coverages on the surface. In
order to avoid systematic errors in determining the number of
water layers that might be introduced by the smoothing
procedure, the spectra used for quantification were the 0.5 cm-
spectra (Figure 2a) but with the contribution of gas-phase water
subtracted out. Figure 2 shows the blue shift in the absorption
spectrum due to a strong interaction between water and the
surface at low coverages; it is therefore expected that the
absorption coefficient will also be smaller than that for bulk
liquid water at these lower coverages. However, given the
uncertainty inherent in estimating the correction factor for the
absorption coefficient for such a perturbed liquid-surface system,
we have used the bulk liquid water value at all coverages.
Figure 3 summarizes the number of monolayers of water on
the glass surface as a function of the relative humidity. The
data suggest a Type II isotherm37 characteristic of multilayer
absorption. The dotted line shown in Figure 3 represents a fit
for a BET isotherm of the form37
1
Figure 4. Decay of gas-phase NO and formation of NO in the long
2
3
path cell whose walls had first been exposed to HNO . (a) 0% RH and
1
6
-3
[
1
NO]
0
) 1.6 × 10 molecule cm (b) 30% RH and [NO] ) 1.4 ×
0
1
6
-3
16
0
molecule cm ; (c) 50% RH and [NO]
0
) 1.4 × 10 molecule
-
3
16
-3
cm ; (d) 70% RH and [NO]
pressure was 1 atm in N at room temperature.
0
) 1.4 × 10 molecule cm . The total
2
are used in the context of “equivalent numbers of layers”
because the water is unevenly distributed.
We assume in the experiments with HNO3 on the surface
that preadsorbing HNO3 on the glass does not alter the
44
subsequent uptake of water. Although Bogdan and Kulmala
reported that HNO3 and HCl do affect the uptake of water on
silica powder, we did not observe an increase in the 5275 cm
cBRH
-1
fractional coverage )
(II)
(
1 - RH)[1 + (c - 1)RH]
combination infrared band of water on silica powder when it
had been “dosed” with HNO3 before exposure to water vapor.45
(This band was followed to avoid interfering absorptions in the
B
where cB ) 100 is a constant. We understand that although the
fit could be improved with a multi-parameter model, the data
in Figure 3 are adequate for determining the number of water
layers under our experimental conditions.
-
1
3000-3500 cm region by HNO3 itself).
B. Reaction of Gaseous NO with HNO3-H2O Thin
Surface Films. Figure 4a shows the results of a typical
experiment in which a mixture of NO in N2 was added to the
cell in the absence of water after dosing with HNO3. There was
no detectable loss of NO and only a slow formation of NO2.
This could be due to some thermal oxidation of NO by
molecular oxygen impurities from small amounts of air leakage
into the cell during these long experiments, or to a very slow
reaction between NO and HNO3 on the cell surface. With
approximately one monolayer of water, a slightly higher rate
of NO2 formation was observed (e.g., Figure 4b). With three
monolayers of water, the reaction was much faster, with
measurable losses of NO and rapid formation of NO2 (e.g.,
Figure 4c). However, with a further increase in water coverage
to five layers (Figure 4d), the rate of formation of NO2 again
decreased.
Our data are consistent with literature reports of the uptake
3
8
of water on glass, particularly given the different analytical
methods that were used and the different treatments under which
the glass surfaces were prepared. It is interesting, for example,
to note that in 1918 Langmuir reported that 4.5 layers of water
were adsorbed on glass in air.39 This would be consistent with
∼
70% RH in their laboratory.
There is a great deal of evidence that the first few layers of
water on silica surfaces interact strongly with the surface and
do not behave like bulk liquid water. At least the first three
layers of water are known to be strongly perturbed.
addition, water is known to form clusters on the surface at low
coverages, rather than forming a uniform thin film.
result, in our experimental system, the water on the surface is
better thought of as clusters at relative humidities at or below
4
0-42
In
4
0-43
As a
The stoichiometry ∆[NO2]/∆[NO] was calculated from the
slopes of the lines obtained when NO and NO2 were plotted as
5
0%. Therefore, one, two or three layers of water on the surface

Reaction of Gaseous Nitric Oxide with Nitric Acid
J. Phys. Chem. A, Vol. 105, No. 45, 2001 10343
TABLE 1: Summary of Long Path Cell FTIR Measurements of the Decay of NO and Formation of NO
2
in the Reaction of NO
with HNO Adsorbed on the Cell Walls at Different Water Coverages on the Surface
3
number of
[NO]0 (1015
- d[NO]/dt
d[NO2]/dt
surface water layers (%RH) experiment # molecule cm-3) (1013 molecule cm min-1
-3
)
(1013 molecule cm min-1
-3
)
∆NO2/∆NOa average ( 2σ
1.9 (30)
2.5 (40)
3 (50)
1
2
3
4
4.0
7.2
9.5
0.12
0.14
0.46
0.63
0.35
0.47
2.5
2.9
3.3
5.4
3.5
15.0
2.2
3
.8 ( 1.1
.2 ( 0.6
1
2
3
4
2.6
8.0
11.0
15.0
0.08
0.42
0.59
1.1
0.2
1.5
2.2
3.3
2.5
3.6
3.7
3.0
3
1
2
3
4
5
6
7
0.65
0.095
0.36
0.91
2.0
4.3
7.3
0.35
1.1
2.6
7.1
15
3.7
3.0
2.8
3.6
3.5
2.7
3.6
3.2
7.0
8.2
14
22
40
20
40
11
3
.3 ( 0.4
a
From the ratio of {d[NO
2
]/dt}/{d[NO]/dt}.
These experiments clearly show that the rate of the hetero-
geneous reaction of gaseous NO with HNO3 on borosilicate glass
depends strongly on the presence of water on the surface. The
reaction was so slow as to be undetectable in the absence of
water, but accelerated as the number of surface water layers
approached three. With further increases in water, however, the
rate again decreased.
Nitric acid is expected to hydrogen bond with the polar silanol
4
0,46,47
groups (-Si-OH) at the silica surface.
Independent
evidence for this HNO3-silica interaction was obtained45 from
the absorption spectrum of porous glass and silica before and
-
1
after dosing with HNO3. The sharp peak at ∼3750 cm due to
the O-H stretch of free (i.e., not hydrogen-bonded) -SiOH
3
2
surface groups decreased upon adsorption of HNO3 but
recovered when HNO3 was removed by extensive pumping.
Thus, we attribute the decrease in the peak to a reversible
hydrogen-bonding of HNO3 to the silanol group. A similar
2
1
change has been observed by Goodman et al. when silica
powders were exposed to gaseous HNO3. The strength of this
hydrogen-bond can be estimated from ab initio calculations by
Figure 5. Rate of NO
concentration at 1 atm pressure in N
2
formation as a function of initial NO
and 50% RH.
2
4
8
Tao et al. of the binding of nitric acid to water in the gas
a function of time between 0 and 300 min for the runs where
-3 layers of water were on the surface. Table 1 summarizes
these data. The weighted average is ∆[NO2]/∆[NO] ) 3.3 (
-
1
phase. The binding energy was estimated to be ∼30 kJ mol ,
with two hydrogen bonds formed between the molecules. A
reasonable value for one hydrogen bond between water and
nitric acid is therefore 15 kJ mol , which lies in the range of
12-24 kJ mol reported for a variety of hydrogen bonds.
2
1
4
-3
1
.0 (2 σ). Small concentrations of HONO (∼10 cm ) were
-
1
detected at larger reaction times; for example, with an initial
-
1
32
16
-3
NO concentration of 2.2 × 10 molecule cm and three layers
14
of surface water, HONO at 260 min was ∼2 × 10 molecule
When small amounts of water are adsorbed on silica surfaces,
it is believed to cluster on the surface rather than forming a
uniformly distributed layer. This is attributed to an enthalpy of
adsorption of water on water clusters that is greater than that
-
3
16
-3
cm compared to NO2 at 4 × 10 molecule cm . Because
HONO was detectable when significant amounts of NO2 had
been formed, it may have been generated at least in part by the
surface NO2 hydrolysis reaction.2
-16
-1
for adsorption on an isolated silanol group (44 kJ mol vs 25
-
1 40
Figure 5 shows the rate of NO2 formation in the long path
cell as a function of the initial NO concentrations under
conditions where three layers of surface water were present,
indicating that the reaction generating NO2 is first-order in NO.
To ensure that NO2 formation was due to reaction 5, blank
runs were also carried out in which NO was introduced alone
into the clean cell whose walls had been cleaned by rinsing
with Nanopure water. Spectra of NO in the cell at 0 and 50%
RH were collected as a function of time; no significant formation
of NO2 was observed. Similarly, blank runs in which nitric acid
alone was introduced into the cell at various RH also gave no
reaction.
kJ mol ). When nitric acid has been preadsorbed on the
surface as in these experiments, water may cluster around the
surface HNO3. An alternate possibility is that water displaces
HNO3 from the silanol group onto the adjacent surface, but that
HNO3 remains in close proximity to the water now clustered
around the -SiOH group; the latter is suggested by the greater
strength of the hydrogen bond between water and the -SiOH
-
1
group (25 kJ mol ) compared to that between nitric acid and
-
1
water, estimated to be ∼15 kJ mol . When both water and
nitric acid are present, water stabilizes HNO3 by as much as 30
-
1
kJ mol relative to the gas phase, assuming two hydrogen-
bonds to nitric acid are involved.

10344 J. Phys. Chem. A, Vol. 105, No. 45, 2001
Saliba et al.
3
Figure 6. Model of reaction of HNO with NO on a silica surface in the presence of water.
It is relevant that HNO3 readily desorbs back into the gas
phase in a dry cell but most of it remains on the surface where
water is present. In a dry cell, the amount of HNO3 desorbing
into the gas-phase varies, depending on the condition of the
hence, solvation of HONO as it is formed should be particularly
important. The difference between the enthalpy of formation50
of HONO in the gas phase compared to solution (undissociated
-
1
HONO) is 40 kJ mol . This is more than sufficient to make
the reaction between NO and surface hydrogen-bonded HNO3
exothermic. The product NO2 will be formed initially in the
water cluster and solvation of this product will further increase
the reaction exothermicity. Because NO2 is much less soluble,
it will be released to the gas phase as shown in Figure 6, whereas
HONO remains on the surface to undergo further reaction with
adsorbed nitric acid.
-1
cell walls. A typical peak absorbance of ∼0.3-0.6 at 896 cm
is observed after dosing HNO3 in a dry cell, compared to ∼0.1
after water vapor is added. Figure 6 summarizes this model of
nitric acid and water on the surface.
Gaseous NO introduced in the cell reacts with the adsorbed
HNO3 surrounded by water molecules to produce NO2 and
HONO
As discussed earlier, the subsequent chemistry of HONO on
the surface may either be the reaction with another surface
HNO3 or the bimolecular reaction between two HONO mol-
ecules on the surface. Although neither can be firmly ruled out
based on our experiments, the former seems more likely
HNO + NO f NO + HONO
(2)
3
2
If the reactants and products in reaction 2 are in the gas phase,
0
-1
the standard enthalpy of reaction is ∆H 298K ) -1.4 kJ mol .
However, our experiments show that in order for reaction 2 to
occur, (i) HNO3 must be on the surface, and (ii) water must be
present. As discussed above, nitric acid hydrogen-bonded to the
surface and to a water molecule is estimated to be stabilized
compared to the gas phase32 by ∼30 kJ mol , making the
HNO + HONO f 2 NO + H O
(3)
3
2
2
0
The enthalpy of this gas-phase reaction is ∆H 298 K ) + 39 kJ
-1
-1
mol . Assuming, as discussed previously, that HNO3 is
-
1
-1
-1
reaction endothermic by ∼29 kJ mol .
stabilized by 30 kJ mol , HONO is stabilized by 40 kJ mol
However, water is also capable of solvating the reaction
due to solvation, and the water is generated in the liquid state,
-
1
-1
-1
products. The Henry’s Law constants, 49 L mol atm for
the reaction enthalpy becomes +65 kJ mol . However, it is
-
2
19,49
HONO compared to 1.4 × 10 for NO2,
show that nitrous
mechanistically reasonable to suggest that the HONO - HNO3
acid interacts more strongly with water than does NO2, and
reaction initially generates N2O4 rather than 2 NO2, and the

Reaction of Gaseous Nitric Oxide with Nitric Acid
reaction then becomes
J. Phys. Chem. A, Vol. 105, No. 45, 2001 10345
directly applied to the atmosphere because the actual form of
nitric acid on the cell walls and how that relates to atmospheric
conditions is not known. Understanding the amounts of water
on surfaces in the lower atmosphere and the form of surface
nitric acid are key to assessing the importance of this chemistry
under various atmospheric conditions. Increased reactive surface
HNO3(surface) + HONO(surface) f N O
4(surface/g)
+ H O
(l)
(6)
2
2
If the N2O4 product is in the gas phase, the standard enthalpy
-1
change for reaction 4 becomes +8 kJ mol . However, solvation
of N2O4 as it is formed will stabilize this product, increasing
the reaction exothermicity. N2O4 is much more soluble than NO2
in water, with its Henry’s Law constant19 being 2 orders of
magnitude larger than that for NO2. As an upper limit, we
60
areas of soils compared to the geometric surface area at the
earth’s surface must also be taken into account. Finally, the
reaction kinetics of NO with the nitric acid-water clusters needs
to be directly assessed. Such studies are currently underway in
this laboratory.
-
1
estimate an additional gain of 29 kJ mol based on the
difference between gaseous and liquid N2O4. Dissociation of
N2O4 then releases NO2 into the gas phase. Given the critical
role played by water, the most accurate representation of the
overall reaction may be
IV. Conclusions
Gaseous nitric oxide reacts at room temperature with nitric
acid on a glass surface in the presence of water. The major
gaseous product is NO2, with the overall reaction stoichiometry
corresponding to 3 NO2 produced per NO reacted. The reaction
is first order with respect to NO. These experimental observa-
tions are consistent with the overall reaction 2 HNO3(surface) +
NO(g) f 3 NO2(g) + H2O(l). However, the presence of water on
the surface is critical. Its role is likely to solvate the HONO
and N2O4 products generated in two steps that make up the
overall reaction. We propose a model in which HNO3 is
hydrogen-bonded to the surface in close proximity to water.
As a result, the overall reaction may be better represented as
follows: 2 HNO3(surface) + 3H2O(surface) + NO(g) f NO2(g) +
N2O4(surface) + 4 H2O(l). This chemistry is potentially important
in “renoxification” of HNO3 in the boundary layer of polluted
urban atmospheres where silica surfaces are plentiful in particles,
soils and building materials, as well as globally in the free
troposphere where dust particles are present.
2
HNO3(surface) + 3H O
+ NO(g)
f
2
(surface)
NO2(g) + N O
+ 4 H O (7)
2 (l)
2
4(surface)
This model is consistent with the work of Bogdan and
44,51,52
co-workers
who studied the uptake of nitric acid and water
on silica powders. They reported that the concentration of nitric
acid was larger in the layers adjacent to the silica surface and
that the enthalpies of fusion of microdroplets of nitric acid and
5
1
water on these surfaces are lower than for the bulk acid-water
solutions.52 Thus, nitric acid and water on silica surfaces cannot
be treated as bulk aqueous systems.
The reaction with five layers of water present on the surface
(Figure 4) is much slower than that with two to three layers of
water. This is likely due to the fact that water behaves like a
40-42
bulk liquid at these higher coverages.
Thus, under these
conditions the surface water may more closely resemble a bulk
aqueous solution of nitric acid, rather than surface-adsorbed
clusters as proposed in the model in Figure 6. Nitric acid is
well-known to dissociate in dilute aqueous solutions, even on
surfaces. Supporting this possibility is the observation by
Acknowledgment. We are grateful to the California Air
Resources Board for support of this work. We also thank J. N.
Pitts, Jr. for helpful discussions and encouragement to study
these heterogeneous systems, V. H. Grassian, G. E. Ewing, D.
Tobias and P. Jungwirth for providing preprints prior to
publication and for helpful discussions, and J. Meyer for
technical assistance.
21
45
Goodman et al. and by this laboratory that addition of water
vapor at high relative humidities when HNO3 is adsorbed on
silica leads to a decrease in the molecular nitric acid peak and
an increase in nitrate ion peaks. At the other extreme, if the
film is highly concentrated in nitric acid, molecular HNO3
associates with water molecules to form hydrates.53-59 Under
these conditions, the vapor pressure of nitric acid in equilibrium
References and Notes
(
(
1) Smith, J. H. J. Am. Chem. Soc. 1947, 69, 1741.
2) Goodman, A. L.; Underwood, G. M.; Grassian, V. H. J. Phys. Chem.
A 1999, 103, 7217.
5
6
with the solution is quite high (of the order of Torr) and less
nitric acid may remain on the silica surface for reaction.
Atmospheric Implications. Silica surfaces are ubiquitous in
the troposphere in the form of dust particles, soil and building
(
71.
(
3) Barney, W. S.; Finlayson-Pitts, B. J. J. Phys. Chem. A 2000, 104,
1
4) Sakamaki, F.; Hatakeyama, S.; Akimoto, H. Int. J. Chem. Kinet.
983, 15, 1013.
(5) Pitts, J. N.; Sanhueza, E.; Atkinson, R.; Carter, W. P. L.; Winer,
A. M.; Harris, G. W.; Plum, C. N. Int. J. Chem. Kinet. 1984, 16, 919.
materials. Nitric acid is well-known to be readily taken up by
1
such surfaces.1
9-22
The studies presented here suggest that the
sticking of nitric acid to such surfaces is more efficient in the
presence of surface water; given that water vapor is always
present in the lower atmosphere, this will not be a limiting factor
under atmospheric conditions. Our studies show that with the
appropriate amount of water on the surface, HNO3 can
potentially be converted into NO2. Such “renoxification” has
significant implications for the chemistry of both the free
troposphere and polluted urban areas. In order for such chemistry
to occur, there must be sufficient water on the surface to stabilize
the HNO3 and to solvate the reaction products. On the other
hand, if there is so much water on the surface that it behaves
like a bulk liquid and the nitric acid is largely dissociated, the
reaction does not occur. The data presented here with ap-
proximately three surface monolayers of water imply that the
reaction probability for loss of NO on the cell walls under these
(
6) Svensson, R.; Ljungstrom, E.; Lindqvist, O. Atmos. EnViron. 1987,
2
2
1, 1529.
(
2, 487.
7) Jenkin, M. E.; Cox, R. A.; Williams, D. J. Atmos. EnViron. 1988,
(8) Febo, A.; Perrino, C. Atmos. EnViron. 1991, 25A, 1055.
(9) Bambauer, A.; Brantner, B.; Paige, M.; Novakov, T. Atmos.
EnViron. 1994, 28, 3225.
10) Mertes, S.; Wahner, A. J. Phys. Chem. 1995, 99, 14 000.
11) Harrison, R. M.; Peak, J. D.; Collins, G. M. J. Geophys. Res. 1996,
01, 14 429.
12) Lammel, G.; Cape, J. N. Chem. Soc. ReV. 1996, 25, 361.
(13) Kleffmann, J.; Becker, K. H.; Wiesen, P. Atmos. EnViron. 1998,
2, 2721.
14) Kleffmann, J.; Becker, K. H.; Wiesen, P. J. Chem. Soc., Faraday
Trans. 1998, 94, 3289.
15) Harrison, R. M.; Collins, G. M. J. Atmos. Chem. 1998, 30, 397.
16) Wingen, L.; Yang, H.; Sumner, A. L.; Finlayson-Pitts, B. J., in
preparation, 2001.
(
(
1
(
3
(
(
(
-
8
conditions is of the order of 10 . However, this cannot be

10346 J. Phys. Chem. A, Vol. 105, No. 45, 2001
Saliba et al.
(
17) Mochida, M.; Finlayson-Pitts, B. J. J. Phys. Chem. A 2000, 104,
705.
18) Fairbrother, D. H.; Sullivan, D. J. D.; Johnston, H. S. J. Phys. Chem.
A 1997, 101, 7350.
19) Finlayson-Pitts, B. J.; Pitts, J. N. Chemistry of the Upper and Lower
(41) Icenhower, J. P.; Dove, P. M. Water Behaviour at Silica Surfaces.
In Adsorption on Silica Surfaces; Papirer, E., Ed.; Marcel Dekker: New
York, 2000; Ch. 9.
(42) Adsorption on Silica Surfaces; Papirer, E., Ed.; Marcel Dekker:
New York, 2000; Vol. 90.
9
(
(
Atmosphere: Theory, Experiments and Applications; Academic Press: New
York, 2000.
(43) Bogdan, A. Fumed Silica as a Host for Study of the Large Surface-
toVolume Ratio Problems in Finely Divided Aqueous Systems: Implications
for the Atmosphere. In Adsorption on Silica Surfaces; Papirer, E., Ed.;
Marcel Dekker: New York, 2000.
(
20) Padgett, P. E.; Bytnerowicz, A. Atmos. EnViron. 2001, 35, 2405.
(21) Goodman, A. L.; Bernard, E. T.; Grassian, V. H. J. Phys. Chem. A
2
001, 105, 6443.
(44) Bogdan, A.; Kulmala, M. J. Colloid Interfac. Sci. 1997, 191, 95.
(45) Sumner, A. L.; Ramazan, K.; Rivera, A.; Finlayson-Pitts, B. J. 2001,
unpublished results.
(
22) Hanisch, F.; Crowley, J. N. J. Phys. Chem. A 2001, 105, 3096.
23) Dentener, F. J.; Carmichael, G. R.; Zhang, Y.; Lelieveld, J.;.Crutzen,
(
P. J. J. Geophys. Res. 1996, 101, 22.
24) Underwood, G. M.; Song, C. H.; Phadnis, M.; Carmichael, G. R.;
Grassian, V. H. J. Geophys. Res. 2001, 106, 18055.
(46) Chuang, I.-S.; Kinney, D. R.; Bronnimann, C. E.; Ziegler, R. C.;
Maciel, G. E. J. Phys. Chem. 1992, 96, 4027.
(47) Bronnimann, C. E.; Zeigler, R. C.; Maciel, G. E. J. Am. Chem.
Soc. 1988, 110, 2023.
(
(25) Chatfield, R. B. Geophys. Res. Lett. 1994, 21, 2705.
(
26) Hauglustaine, D. A.; Ridley, B. A.; Solomon, S.; Hess, P. G.;
(48) Tao, F.-M.; Higgins, K.; Klemperer, W.; Nelson, D. D. Geophys.
Res. Lett. 1996, 23, 1797.
(49) Cheung, J. L.; Li, Y. Q.; Boniface, J.; Shi, Q.; Davidovits, P.;
Worsnop, D. R.; Jayne, J. T.; Kolb, C. E. J. Phys. Chem. A. 2000, 104,
2655.
(50) Wagman, D. D.; Evans, W. H.; Parker, V. B.; Schumm, R. H.;
Halow, I.; Balley, S. M.; Churney, K. L.; Nuttall, R. L. J. Phys. Chem.
Ref. Data 1982, 11, Suppl. No. 2.
(51) Bogdan, A.; Kulmala, M.; Gorbunov, B.; Kruppa, A. J. Colloid
Interfac. Sci. 1996, 177, 79.
(52) Bogdan, A.; Kulmala, M. Geophys. Res. Lett. 1999, 26, 1433.
(53) Redlich, O.; Bigeleisen, J. J. Am. Chem. Soc. 1943, 65, 1883.
(54) Redlich, O. Chem. ReV. 1946, 39, 333.
(55) Miles, F. D. Nitric Acid Manufacture and Uses; Oxford University
Press: London, 1961.
Madronich, S. Geophys. Res. Lett. 1996, 23, 2609.
27) Lary, D. J.; Lee, A. M.; Toumi, R.; Newchurch, M. J.; Pirre, M.;
Renard, J. B. J. Geophys. Res. 1997, 102, 3671.
28) Gomer, T.; Brauers, T.; Heintz, F.; Stutz, J.; Platt, U. UniVersity
of Heidelberg 1995.
29) Barney, W. S.; Wingen, L. M.; Lakin, M. J.; Brauers, T.; Stutz, J.;
Finlayson-Pitts, B. J. J. Phys. Chem. A 2000, 104, 1692.
30) Barney, W. S.; Wingen, L. M.; Lakin, M. J.; Brauers, T.; Stutz, J.;
Finlayson-Pitts, B. J. J. Phys. Chem A 2001, 105, 4166.
31) Herzberg, G. Molecular Spectra and Molecular Structure. II.
(
(
(
(
(
Infrared and Raman Spectra of Polyatomic Molecules; D. Van Nostrand
Company, Inc.: Princeton, N. J., 1945; Vol. II.
(
32) Pimentel, G. C.; McClellan, A. L. The Hydrogen Bond; W. H.
Freeman: San Francisco, 1960.
(
(
(
(
33) Foster, M.; Ewing, G. E. Surf. Sci. 1999, 427/428, 102.
34) Foster, M. C.; Ewing, G. E. J. Chem. Phys. 2000, 112, 6817.
35) Weis, D. D.; Ewing, G. E. J. Geophys. Res. 1996, 101, 18 709.
36) Chattoraj, D. K.; Birdi, K. S. Adsorption and the Gibbs Surface
(56) Davis, W. J.; DeBruin, H. J. J. Inorg. Nucl. Chem. 1964, 26, 1069.
(57) Schwartz, S. E.; White, W. H. Solubility Equilibria of the Nitrogen
Oxides and Oxyacids in Dilute Aqueous Solution. In AdVances in
EnVironmental Science and Engineering; Pfafflin, J. R., Ziegler, E. N., Eds.;
Gordon and Breach Science Publishers: New York, 1981; Vol. 4; p 1.
(58) Koller, J.; Hazdi, D. J. Mol. Structure 1991, 247, 225.
(59) Molina, M. J.; Zhang, R.; Wooldridge, P. J.; McMahon, J. R.; Kim,
J. E.; Chang, H. Y.; Beyer, K. D. Science 1993, 261, 1418.
(60) Hodson, M. E.; Langan, S. J.; Kennedy, F. M.; Bain, D. C.
Geoderma 1998, 85, 1.
Excess; Plenum Press: New York, 1984.
37) Masel, R. I. Principles of Adsorption and Reaction on Solid
Surfaces; Wiley: New York, 1996.
(
(
(
(
38) Frazer, J. H. Phys. ReV. 1929, 33, 97.
39) Langmuir, I. J. Am. Chem. Soc. 1918, 40, 1361.
40) Iler, R. K. The Chemistry of Silica; Wiley: New York, 1978.