Enhancement of N2O4 on Porous Glass at Room Temperature: A Key Intermediate in the Heterogeneous Hydrolysis of NO2?

Article abstract of DOI:10.1021/jp993169b

The heterogeneous hydrolysis of NO2 at surfaces in the atmosphere is believed to be a significant source of HONO, a key OH precursor in urban areas. However, the mechanism of this reaction is not known. The uptake of 2.9 Torr of NO2 in N2 at a total pressure of 508 Torr on a porous glass surface with varying amounts of surface-adsorbed water was studied using FTIR at 294 K. The ratio of N2O4 to NO2 was enhanced on the glass surface relative to the gas phase. On a relatively dry surface, the formation of surface-adsorbed HNO3 was observed over a period of ca. 20 h, likely due to the reaction with small amounts of water on the surface. Gas-phase NO and N2O were also generated. When larger amounts of water were initially present on the surface, surface-adsorbed HNO3 was formed immediately, as well as gas-phase NO, N2O, and HONO. Although the NO2 concentrations used in the present studies are much larger than those found in the atmosphere, this work suggests that N2O4 should be considered as a key intermediate in the heterogeneous hydrolysis of NO2 to form HONO.

Full text of DOI:10.1021/jp993169b

© Copyright 2000 by the American Chemical Society
VOLUME 104, NUMBER 2, JANUARY 20, 2000
LETTERS
Enhancement of N2O4 on Porous Glass at Room Temperature: A Key Intermediate in the
Heterogeneous Hydrolysis of NO2?
W. S. Barney and B. J. Finlayson-Pitts*
Department of Chemistry, UniVersity of California, IrVine, IrVine, California 92697-2025
ReceiVed: September 7, 1999; In Final Form: NoVember 11, 1999
The heterogeneous hydrolysis of NO2 at surfaces in the atmosphere is believed to be a significant source of
HONO, a key OH precursor in urban areas. However, the mechanism of this reaction is not known. The
uptake of 2.9 Torr of NO2 in N2 at a total pressure of 508 Torr on a porous glass surface with varying
amounts of surface-adsorbed water was studied using FTIR at 294 K. The ratio of N2O4 to NO2 was enhanced
on the glass surface relative to the gas phase. On a relatively dry surface, the formation of surface-adsorbed
HNO3 was observed over a period of ∼20 h, likely due to the reaction with small amounts of water on the
surface. Gas-phase NO and N2O were also generated. When larger amounts of water were initially present on
the surface, surface-adsorbed HNO3 was formed immediately, as well as gas-phase NO, N2O, and HONO.
Although the NO2 concentrations used in the present studies are much larger than those found in the atmosphere,
this work suggests that N2O4 should be considered as a key intermediate in the heterogeneous hydrolysis of
NO2 to form HONO.
Introduction
not included in airshed models because the mechanisms and
kinetics are not sufficiently well understood. For example, while
HONO has been observed in the gas phase, the corresponding
amount of HNO3 shown in reaction 1 has not. The reason has
been proposed to be that HNO3, which is notoriously “sticky”,
remains adsorbed on the surface. This is consistent with the
observations of nitrate in washings of the surface postreaction,4
and with a very recent observation12 of the formation of
adsorbed HNO3 on the surface of hydrated silica particles.
Oxides of nitrogen play key roles in the chemistry of the
troposphere and stratosphere.1 While the gas-phase chemistry
of species such as NO and NO2 is reasonably well-known,
heterogeneous reactions occurring on surfaces are not well
understood. For example, the hydrolysis of NO2 on surfaces
2NO2 + H2O surface8 HONO + HNO3
(1)
Silica surfaces such as powders or porous glass are known
to hold adsorbed water readily because of their polar surface
-Si-OH groups.13 While this adsorbed water can be removed
by heating under vacuum, and indeed the surface can be
dehyroxylated at sufficiently high temperatures, heating at lower
temperatures leaves adsorbed water on the surface. As a result,
such surfaces can be used as supports for water to study surface
reactions involving water. For example, Grassian and co-
workers12 recently reported studies of reaction 1 using silica
powders.
was proposed some 16 years ago to explain the formation of
HONO observed when NO2 was injected into large environ-
mental chambers in the presence of water vapor,2,3 which was
confirmed in a number of subsequent studies.4-9 This is
important because HONO is believed to be the major source of
OH in many urban areas at dawn.10,11 However, despite almost
two decades of work in this area, its formation by reaction 1 is
* Author to whom correspondence should be addressed.
10.1021/jp993169b CCC: $19.00 © 2000 American Chemical Society
Published on Web 12/22/1999

172 J. Phys. Chem. A, Vol. 104, No. 2, 2000
Letters
Commercially available porous glass plates are too thick (∼1
mm) to transmit in the region of interest from 1800 to 1500
cm-1. To obtain sufficiently thin plates for these experiments,
a plate of porous glass (Corning) approximately 1 mm thick
was further etched by immersion in a 7.7% (v:v) HF solution
for 21 min to a thickness of ∼0.3 mm. This was found to be
the maximum thickness at which usable spectra in the 1800-
1500 cm-1 region could be obtained. The plate was then rinsed
in Nanopure water and dried in an oven for 1 h at 120 °C. The
plate was then mounted on the holder and inserted into the cell.
For those experiments in which the porous glass was heated
prior to exposure to NO2, the cell was evacuated and heated
from the outside for 35 min at 280 °C and then cooled to room
temperature before gases were added. When the plates are left
in air over a period of time, they adsorb gaseous organics, which
were removed by heating in air at 800 °C for 1 h.
NO2 was synthesized by mixing in a 5 L bulb excess O2
(Oxygen Service Company, 99.993%) with NO (Matheson,
99%) which had been passed through a trap at 196 K to remove
impurities such as HNO3. The N2 was Ultrahigh Purity Grade
(>99.999%, Oxygen Service Company). Gaseous HNO3 used
for reference spectra was obtained using the vapor above a 1:2
(v:v) mixture of HNO3 (Fisher, 70%, 15.8 M) and H2SO4
(Fisher, 95.7%, 18 M).
Results and Discussion
Figure 2a shows the gas-phase spectrum when 2.9 Torr of
NO2 is added to the cell, along with N2, to a total pressure of
508 Torr at room temperature. Strong peaks in the 1600-1650
cm-1 region are due to NO2. The smaller peak at ∼1750 cm-1
is due to the dimer, N2O4, which is in equilibrium with NO2.
Using the equilibrium constant14 Kp ) 2.5 × 10-19 cm3
molecule-1 for the reaction
2NO2 T N2O4
(2,-2)
Figure 1. Schematic diagram of the cell. The path length for the
infrared beam through the cell is 6.7 cm.
the concentrations of N2O4 and NO2 under these conditions are
2 × 1015 and 9 × 1016 molecules cm-3, respectively.
We report here FTIR studies of the uptake of NO2 on porous
glass at room temperature and show that adsorbed NO2 exists
preferentially in the form of the dimer, N2O4. In the presence
of water on the surface, surface-adsorbed HNO3 is formed and
detected directly by FTIR. Possible atmospheric implications
are discussed.
Figure 2b shows the spectrum of the surface-adsorbed plus
the gas-phase species. Figure 2c shows the difference between
Figure 2b and 2a, i.e., the spectrum of the adsorbed species
present at approximately the same time that the gas-phase
spectrum in Figure 2a was obtained. The band due to adsorbed
N2O4 at 1740 cm-1 is clearly seen, while that due to NO2 is
not evident. (Porous glass absorbs particularly strongly between
1610 and 1660 cm-1; imperfect subtractions in such regions
are common when such strong bands are ratioed or subtracted.)15
Figure 2d shows the spectrum of the adsorbed species 20 h
later. There is a new absorption centered at ∼1680 cm-1. The
assignment of this band to HNO3 is confirmed by a reference
spectrum (Figure 2e) in which a clean plate of porous glass
was exposed to 0.57 Torr of gaseous HNO3 for 5 min. The
gaseous HNO3 was pumped away before the reference spectrum
was taken. The relatively small change in the surface-adsorbed
N2O4 peak while that due to HNO3 is increasing is likely due
to replenishment of the N2O4 from the gas phase during the
run.
Experimental Section
Figure 1 is a schematic diagram of the reaction cell. It is
made of borosilicate glass and has ZnSe windows at each end,
providing a 6.7 cm path length for the infrared beam. The porous
glass is held at the end of a positioning rod and can be
withdrawn into the sidearm for heating and for obtaining the
infrared spectrum of the gases alone. With the porous glass
sample in the infrared beam, the spectrum of both the surface
species and the gases is recorded. Spectra of the surface species
are obtained by ratioing the single-beam spectra of the porous
glass before and after addition of the NO2 and subtracting a
gas-phase spectrum recorded with the glass raised. Gas-phase
spectra are recorded approximately 2 min before and after the
glass spectrum, and the amount subtracted is determined by
interpolation. Even during the fastest part of the NO2 decay,
the absorbance difference between the two gas-phase spectra
was never more than 5%. Spectra were recorded as 64 co-added
scans at 0.5 cm-1 resolution at pressures between 500 and 700
Torr of N2.
Figure 3 shows the results of a similar experiment but where
the porous glass plate had been left exposed to room air
overnight and was then was used without prior heating. In room
air, the porous glass takes up large amounts of water. For
example, a peak at 5260 cm-1 was observed due to surface-
adsorbed water13 for this sample, but not for that used for the
experiments in Figure 2. Figure 3a shows the spectrum of the
Porous glass transmits infrared radiation well above ∼2000
cm-1; however, it drops off rapidly at lower wavenumbers.13

Letters
J. Phys. Chem. A, Vol. 104, No. 2, 2000 173
Figure 2. FTIR spectra of 2.9 Torr of NO2 in N2 (total pressure 508
Torr) (a) gas-phase only, (b) gas-phase plus surface-adsorbed species,
(c) subtracted spectrum (b - a) showing the spectrum due to the
adsorbed species, and (d) same as (c) but 20 h later; (e) reference
spectrum of HNO3 adsorbed on porous glass; 0.57 Torr of HNO3 was
added to the cell for 5 min and then pumped away before recording
the spectrum. The porous glass had been heated under vacuum before
the exposure to NO2. All spectra are on the same scale.
Figure 3. FTIR spectra of 3.3 Torr of NO2 in N2 (total pressure 573
Torr) (a) gas-phase only, (b) gas-phase plus surface-adsorbed species,
(c) subtracted spectrum (b - a) showing the spectrum due to the
adsorbed species. The porous glass was not heated to remove adsorbed
water prior to exposure to NO2.
gas, Figure 3b shows that of the gas plus adsorbed species, and
Figure 3c shows the difference, all recorded immediately after
introduction of NO2 into the cell. In contrast to the relatively
“dry” porous glass in Figure 2c, the HNO3 absorption is much
larger than that of N2O4 on the surface, suggesting that reaction
with the water on the surface to form HNO3 has occurred even
at the shortest reaction times.
In the gas phase, N2O and NO were observed as reaction
products under all conditions and in the “wet” case, HONO
was also observed. For example, Figure 4 shows the gas-phase
spectrum in the 2200 cm-1 region taken at about the same time
as that in Figure 2d. Superimposed on the narrow rotational
bands of CO2 (due to changes in purging of the air in the light
beam) is a set of broad peaks, which can be seen by comparison
to a reference spectrum (shown by the heavy line) to be due to
gaseous N2O, which increases linearly with time. The concen-
tration after 1200 min is 2 × 1016 molecules cm-3, compared
to a loss of NO2 of ∼7 × 1016 molecules cm-3 during this
period. NO first increases and then decreases, with a peak
concentration of ∼2.5 × 1016 molecules cm-3 after 200 min
and 1 × 1016 molecules cm-3 at 1200 min.
Figure 4. Gas-phase spectrum taken under the same experimental
conditions as Figure 2d. The heavy line is a reference spectrum of gas-
phase N2O. Other absorptions are due to CO2 outside the cell.
at a concentration of ∼1 × 1015 molecules cm-3. In experiments
where the porous glass had been heated under vacuum to remove
surface water prior to introducing NO2, gas-phase HONO was
not observed, even at similar levels of NO2 loss. HONO has
been observed in other studies17,18 to be a precursor to N2O.
However, the yield of N2O we observed was not significantly
affected by the amount of water on the porous glass, i.e., whether
HONO was observed in the gas phase. This suggests that either
HONO is not formed on the “dry” surface and there is more
Figure 5 shows the gas-phase spectrum in the 1250 cm-1
region during the experiment using “wet” glass in Figure 3.
When the peak due to gaseous N2O4 is subtracted, a small band
at 1263 cm-1 is detectable, which by comparison to a reference
spectrum16 (shown by the heavy line) is due to gaseous HONO

174 J. Phys. Chem. A, Vol. 104, No. 2, 2000
Letters
+ O2 may be responsible for the observed N2O production.
∆Hrxn for the gaseous decomposition of N2O3 to NO + NO2 is
+9.7 kcal mol-1, compared to -0.2 kcal mol-1 to produce N2O
+ O2; such reactions which are not highly exothermic are often
found to be surface-catalyzed.
Whether this mechanism involving enhanced N2O4 concen-
trations on the surface and its isomerization and autoionization
followed by reaction with water also applies in the atmosphere
at the much smaller concentrations is not clear. Certainly the
gas-phase N2O4 concentrations in equilibrium with atmospheric
levels of NO2 (<0.1 ppm generally) are much smaller than those
in the present studies. Previous studies2-8 of the formation of
HONO in laboratory systems from the heterogeneous reaction
of NO2 at ppm concentrations have established that the reaction
is first-order in NO2. This is inconsistent with a direct reaction
of gas-phase N2O4, which would be second-order. Various
reaction schemes have been proposed. For example, a stepwise
reaction involving a slow adsorption of NO2 at the surface
followed by a rapid reaction of two surface-adsorbed NO2 with
water has been proposed.3 The uptake of NO2 on liquid water29
is known to be slow. Alternatively, a fast adsorption of NO2 on
the surface followed by slower reactions of intermediate species
have been suggested.3,9 Surface-adsorbed N2O4 may be a key
short-lived intermediate in such mechanisms.
Figure 5. Gas-phase spectrum taken under the same experimental
conditions as in Figure 3. The open circles show the spectrum remaining
after the contribution of N2O4 has been subtracted out, and the heavy
line is a reference spectrum of gaseous HONO.
than one path to N2O formation, or that HONO remains
adsorbed on the dry surface. The NO and N2O concentrations
measured at these short reaction times were ∼3 × 1016
molecules cm-3 and 5 × 1015 molecules cm-3, respectively,
and hence were again the major gas-phase products.
In short, when water is present on the porous glass surface,
NO, N2O, and HONO are formed at the same time that HNO3
is generated on the surface. It should be noted that this chemistry
is likely occurring on all of the surfaces of the reaction cell,
not just the porous glass surface. However, “blank” runs in
which NO2 was followed with time in the cell in the absence
of the porous glass showed that the loss of NO2 was much
smaller (∼7% compared to 52% loss over 200 min), indicating
that most of the chemistry observed was heterogeneous. In the
“blank” runs, the only gas-phase product observed was NO.
Grassian and co-workers12 recently reported studies of the
interaction of NO2 with dry and hydrated silica particles,
respectively. Adsorbed HNO3 was observed on the hydrated
powders and N2O4 on both dry and hydrated powders, similar
to the observations on porous glass reported here. Gaseous N2O
was not reported.
N2O4 has also been observed as an intermediate on the surface
of ice at 86 K after exposure to NO2,19 and upon heating,
generates HONO and HNO3 among other products. The
formation of N2O4 on the ice surface is not unexpected, given
that the equilibrium (2, -2) will shift to the right at lower
temperatures. However, the results presented here suggest that
N2O4 is also an important intermediate at room temperature on
other types of surfaces such as glass.
Summary
The ratio of N2O4 to NO2 is enhanced on porous glass
surfaces at room temperature, relative to that in the gas phase.
Surface-adsorbed HNO3 is also generated, at short reaction times
on a “wet” surface and at longer reaction times on relatively
dry surfaces. Production of HONO is observed simultaneously
in the gas phase when there is water initially present on the
surface, and NO and N2O are also produced in both cases. These
results suggest that N2O4 at the interface may be a key
intermediate in the heterogeneous reaction 1 of NO2 to form
gaseous HONO and surface-adsorbed HNO3.
Acknowledgment. The authors are grateful to the California
Air Resources Board (Contract No. 97-311) for support of this
work. We also thank B. E. Koel for providing preprints prior
to publication, Ian D. Chapman for providing some of the porous
glass samples, Jorg Meyer for experimental assistance, and J.
N. Pitts, Jr., Vicki Grassian, and A. Apkarian for helpful
discussions.
References and Notes
N2O4 in solution and at low temperatures is known20-24 to
isomerize and autoionize to NO+NO3-. Reaction of this ionic
form with water may then generate HONO + HNO3, in a
manner similar to that proposed for the N2O5 hydrolysis on ice
surfaces.25 For example, Choi et al.26 have reported that NO+
in clusters with gas-phase water containing more than four water
molecules reacts to form HONO. HONO undergoes a self-
reaction27 to form N2O3:
(1) Finlayson-Pitts, B. J.; Pitts, J. N., Jr. Chemistry of the Upper and
Lower Atmosphere: Theory, Experiments and Application; Academic
Press: San Diego, CA, 2000.
(2) Sakamaki, F.; Hatkayema, S.; Akimoto, H. Int. J. Chem. Kinetic.
1983, 15, 1013.
(3) Pitts, J. N., Jr.; Sanhueza, E.; Atkinson, R.; Carter, W. P. L.; Winer,
A. M.; Harris, G. W.; Plum, C. N. Int J. Chem. Kinet. 1984, 16, 919.
(4) Svensson, R.; Ljungstro¨m, E.; Lindqvist, O. Atmos. EnViron. 1987,
21, 1529.
(5) Bambauer, A.; Brantner, B.; Paige, M.; Novakov, T. Atmos.
EnViron. 1994, 28, 3225.
(6) Mertes, S.; Wahner, A. J. Phys. Chem. 1995, 99, 14000.
(7) Kleffmann, J.; Becker, K. H.; Wiesen, P. Atmos. EnViron. 1998,
32, 272.
(8) Jenkin, M. E.; Cox, R. A.; Williams, D. J. Atmos. EnViron. 1988,
22, 487.
(9) Febo, A.; Perrino, C. Atmos. EnViron. 1991, 25A, 1055.
(10) Lammel, G.; Cape, J. N. Chem. Soc. ReV. 1996, 25, 361, and
references therein.
(11) Winer, A. M.; Biermann, H. Res. Chem. Intermed. 1994, 20, 423.
(12) Goodman, A. L.; Underwood, G. M.; Grassian, V. H. J. Chem.
Phys. A. 1999, 103, 7217.
2HONO T N2O3 + H2O
(3,-3)
N2O3 is known to decompose to NO + NO2. Hence,
formation of HONO followed by reaction 3 on the porous glass
surface and decomposition of the N2O3 may be at least partially
responsible for the NO observed in our system. Consistent with
this, HONO has been observed to decompose into NO + NO2
in a glass smog chamber.28 We propose that a minor reaction
path in the N2O3 decomposition on the surface producing N2O

Letters
J. Phys. Chem. A, Vol. 104, No. 2, 2000 175
(13) Kiselev, A. V.; Lygin, V. I. Infrared Spectra of Surface Compounds;
Wiley: New York, 1975.
(21) Parts, L.; Miller, J. T., Jr. J. Chem. Phys. 1965, 43, 136.
(22) Goulden, J. D. S.; Millen, D. J. J. Chem. Soc. 1950, 2620.
(23) Givan, A.; Loewenschuss, A. J. Chem. Phys. 1990, 93, 7592.
(24) Cotton, F. A.; Wilkinson, G. AdVanced Inorganic Chemistry, 5th
ed.; Wiley: New York, 1988.
(25) Koch, K. G.; Banham, S. F.; Sodeau, J.; Horn, A. B.; McCoustra,
M. R. S.; Chesters, M. A. J. Geophys. Res. 1997, 102, 1513.
(26) Choi, J.-H.; Kuwata, K. T.; Haas, B.-M.; Cao, Y.; Johnson, M. S.;
Okumura, M. J. Chem. Phys. 1994, 100, 7153.
(27) Markovits, G. Y.; Schwartz, S. E.; Newman, L. Inorg. Chem. 1981,
20, 445.
(28) Ten Brink, H. M.; Spoelstra, H. Atmos. EnViron. 1981, 32, 247.
(29) Ponche, J. L.; George, Ch.; Mirabel, Ph. J. Atmos. Chem. 1993,
16, 1.
(14) DeMore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Kurylo, M. J.; Howard, C. J.; Ravishankara, A. R.; Kolb, C. E.; Molina,
M. J. Chemical Kinetics and Photochemical Data for Use in Stratospheric
Modeling. In JPL Publication 97-4, Jet Propulsion Laboratory, Pasadena,
CA, January 15, 1997.
(15) Weis, D.; Ewing, G. Anal. Chem. 1998, 70, 3175.
(16) Barney, W. S.; Wingen, L. M.; Lakin, M. J.; Brauers, T.; Stutz, J.;
Finlayson-Pitts, B. J. J. Phys. Chem., submitted for publication.
(17) Wiesen, P.; Kleffmann, J.; Kurtenbach, R.; Becker, K. H. Faraday
Discuss. 1995, 100, 121.
(18) Pires, M.; Rossi, M. J. Int. J. Chem. Kinet. 1997, 29, 869.
(19) Wang, J.; Koel, B. E. J. Phys. Chem. A 1998, 102, 8573.
(20) Addison, C. C. Angew. Chem. 1960, 72, 193.