Synthesis and Mid-IR Absorption Cross Sections of BrNO2

Article abstract of DOI:10.1021/jp960319z

A method for the synthesis of gaseous BrNO2 at concentrations of the order of 1E-9 mol/cm³ at atmospheric pressure is described, based on the heterogeneous reaction of ClNO2 with aqueous bromide solutions.We measured the gas-phase infrared absorption cross sections in the range 700-1800 cm^-1.The matrix isolation spectrum was recorded for identification by comparison with literature data.

Full text of DOI:10.1021/jp960319z

©
Copyright 1996 by the American Chemical Society
VOLUME 100, NUMBER 41, OCTOBER 10, 1996
LETTERS
Synthesis and Mid-IR Absorption Cross Sections of BrNO2
Andreas Frenzel, Volker Scheer, Wolfgang Behnke, and Cornelius Zetzsch*
Fraunhofer Institut f u¨ r Toxikologie und Aerosolforschung, Nikolai-Fuchs-Strasse 1,
D-30625 HannoVer, Germany
X
ReceiVed: January 29, 1996; In Final Form: August 2, 1996
9
3
A method for the synthesis of gaseous BrNO
pressure is described, based on the heterogeneous reaction of ClNO
2
at concentrations of the order of 10- mol/cm at atmospheric
2
with aqueous bromide solutions. We
measured the gas-phase infrared absorption cross sections in the range 700-1800 cm . The matrix isolation
-1
spectrum was recorded for identification by comparison with literature data.
Introduction
BrNO2 is of interest in atmospheric chemistry because it is
in solid oxygen employing a matrix-isolation apparatus, depicted
in Figure 1, which allows observation of IR spectra in the range
-1
4
000-600 cm . Details of this apparatus are given else-
1
,2
produced by reaction of N2O5 with sea-salt aerosol.
It is
9
,10
where;
here we give a brief description:
assumed to be photolyzed easily into Br and NO2 and may
A gold-coated hollow copper cylinder is mounted in a vacuum
chamber on a rotating vertical axis driven by a stepper motor.
The copper cylinder is cooled to 20 K by a cryorefrigerator
placed below. From one side of the vacuum chamber a capillary
ending in a nozzle is pointing horizontally at the surface of the
cylinder. The sample gas is injected through this nozzle on
the cold rotating cylinder thereby freezing out as a continuous
trace. At right angle to the gas inlet the IR beam is focused by
a Cassegrain objective through a KBr window on the gold
surface and reflected back via a beam splitter on the narrow
band MCT detector.
thereby contribute to sudden ozone-depletion events occurring
in the Arctic troposphere.3 UV and IR absorption cross sections
and atmospheric concentrations are still unknown. The atmo-
spheric lifetime is very uncertain, there is only an estimate of
4,5
the lifetime limited by unimolecular decomposition.
Previously, BrNO2 was produced in an argon matrix by
6
7
Tevault et al. and by Feuerhahn et al. via reaction of atomic
8
bromine with NO2. Finlayson-Pitts et al. identified BrNO2 in
-11
3
the gas phase at a level of about 10 mol/cm (estimated from
their data using our IR-absorption cross section) after hetero-
geneous reaction of N2O5 with NaBr salt and also after gas-
phase reaction of N2O5 with BrNO. Here we describe a
continuous flow technique for the synthesis of BrNO2 in the
gas phase by heterogeneous reaction of ClNO2 with a dilute
solution of NaBr. Infrared spectra of gaseous and matrix
isolated BrNO2 are presented.
For UV detection of Br impurities we used a Kontron
2
UVIKON 860 double-beam spectrometer with a 10 cm quartz
glass cell.
Synthesis of BrNO2. The flow scheme of the experimental
setup is displayed in Figure 2. We produced BrNO2 in a
continuous flow at atmospheric pressure by the following
steps: N2O5 is produced by the gas-phase reaction of O3 (750
ppm O3 in O2, prepared by photolyzing O2 with low-pressure
Hg lamps) with NO2 (2000 ppmv NO2 + 2% O2 in He) at a
resulting flow of typically 80 mL/min. By varying the ratio of
the mass flows of the two components we worked with an excess
of either NO2 or O3. N2O5 was passed through a vertically
Experimental Section
Spectroscopic Methods. IR absorption spectra were re-
corded by a Bruker IFS48 FTIR spectrometer, in the gas phase
employing a 20.5 cm glass cell equipped with AgCl windows,
X
Abstract published in AdVance ACS Abstracts, September 15, 1996.
S0022-3654(96)00319-X CCC: $12.00 © 1996 American Chemical Society

1
6448 J. Phys. Chem., Vol. 100, No. 41, 1996
Letters
2
from the
TABLE 1: Calibration of IR Spectra of BrNO
Reaction of ClNO
2
(≈250 ppm) with NaBr Solutions
-
4
a
(
5 × 10 mol/L) by Analysis of Wet Alkaline Denuders
-
[Br-]
[
NO
x
]
[HNO ]
3 g
[BrNO ]
2 g
N
meas
2 ref
[BrNO ]
b,c
1.9
2.9
1.6
2.9
0.59
1.5
0.37
1.3
1.3
1.5
1.2
1.6
3.5
3.6
2.5
2.3
3
5
7
1
1.2
1.1
1.3
1.5
b
d
a
Concentrations are given in units of 10-9 mol/cm , where data from
3
ion chromatography are expressed as gas-phase equivalents.
N
meas
denotes the number of similar runs averaged in each line. [BrNO ]ref is
2
the resulting concentration of the spectrum given in Figure 3. The
average of 16 runs leads to a value of [BrNO ]ref ( 2σstat ( δcal ) (1.2
2
-9
3 b
(
0.4 ( 0.1) × 10 mol/cm . Cold trap (-16 or -24 °C,
respectively) inserted behind the wetted-wall flowtube. N concen-
2 5
O
c
-4
trations under these conditions were neglected. [NaBr] ) 8 × 10
-4
(
instead of 5 × 10 ) mol/L employed for the synthesis of BrNO
2
.
d
Wetted-wall flowtube (length 80 cm) with continuous flow of 0.1 M
NaOH instead of a static wet denuder.
Figure 1. Scheme of the matrix isolation apparatus.
ClNO2 at -20 °C. By analogy our observation may be
explained by the formation of N2O5 from NO2, formed from
BrNO2 in the cold trap, with ozone. In the absence of ozone
NOBr might be formed in the reaction sequence
-
-
+
BrNO + H O f Br + NO + 2H
(3)
(4)
(5)
2
2
3
-
+
BrNO + Br + H f Br + HONO
2
2
+
-
HONO + H + Br f NOBr + H O
2
Quantitative Determination of BrNO2. BrNO2 concentra-
tion was measured in experiments with a small excess of ozone
by trapping BrNO2 in a wet denuder of 90 cm length, coated
with 500 µL of 0.1 M NaOH and analyzing the liquid by ion
2-
Figure 2. Schematic diagram of the wetted-wall flowtube setup.
chromatography with CO3 /HCO3- eluent and AS4 column
and a membrane suppressor (Dionex) with a conductivity
detector. By attaching a second denuder behind the first,
employing 0.1 M NaOH or sulfite solution we checked and
confirmed the efficiency of the first denuder. Impinger
measurements with methyl orange confirmed that oxidizing
species, such as Br2 or HOBr, are also scavenged by the first
denuder quantitatively. Table 1 summarizes our ion chromato-
aligned thermostated flowtube (inner diameter 6 mm, length
3
0 cm) whose inner walls were completely wetted with a film
of 3 M NaCl flowing downward. ClNO2 is produced almost
quantitatively by the net reaction
-
-
N O + Cl f ClNO + NO
3
(1)
2
5
2
-
graphic determinations of BrNO2. From [NOx ], the sum of
details of which were given by Behnke et al.11 The effluent
gas was passed through a second wetted-wall flowtube of similar
construction (see also Figure 2). This flowtube was wetted with
nitrate and nitrite concentrations, expressed as gas phase
equivalents, we subtracted [HNO3]g (determined from IR
spectra), to derive [BrNO ] , the BrNO gas-phase concentration.
2
g
2
-
4
-4
-11
3
8
× 10 or 5 × 10 M NaBr solution. ClNO2 reacts with
N O concentrations were low (5 × 10 mol/cm at maximum)
2
5
-
Br according to the net reaction
and therefore neglected. For each single measurement the
concentration of the reference spectrum given in Figure 3,
[BrNO2]ref, was achieved by spectrum subtraction. The preci-
sion of the measurement is meant to be reflected by the statistical
error. HNO3 calibration and wall losses have to be considered
as additional systematic errors. A loss of NOy-species on the
walls of the system in and behind the IR cell might lead to an
underestimation of the BrNO2 concentration in the IR cell and
thus result in a too high absorption cross section. To reduce
these errors all measurements were done at steady state. For
HNO3 we estimated the calibration error to 15%.
-
-
ClNO + Br f BrNO + Cl
(2)
2
2
To reduce the partial pressure of water vapor we thermostated
both flowtubes to 2 °C. The ClNO2 mixing ratio after the first
flowtube was typically 250 ppmv, measured by FTIR absorption.
Calibration of the IR absorption spectra of ClNO2 and HNO3
was done by alkaline denuders similar to the procedure described
below. After the second flowtube with a reaction length of 34
cm (corresponding to a contact time of 6.4 s), the gas was fed
first (in some experiments) through a cold trap to further reduce
water concentration, then through the absorption cell of the IR
spectrometer or into the matrix isolation apparatus. If a cold
trap was inserted we observed elevated HNO3. N2O5 was
emerging from the cold trap when ozone was in excess (<1
ppm at -16, -24 °C to 4 ppm at -60, -70 °C). Only with
excess of NO2 a small amount of NOBr was produced in the
-
The [Br ] values are always larger than the [BrNO2]g values,
indicating a contribution of other bromine-containing com-
pounds not visible in the IR. [Br2]g was determined in separate
experiments by UV absorption. Under conditions similar to
those of the experiments summarized in the first line of Table
-
9
3
1, it was observed to be 2.0 × 10 mol/cm (with an estimated
error of 5%), and with a small excess of NO2 and a cold trap
(-60 °C) inserted to be 1.8 × 10 mol/cm . The efficiency
12
-9
3
cold trap. George et. al. reported similar reactions when drying

Letters
J. Phys. Chem., Vol. 100, No. 41, 1996 16449
TABLE 2: Comparison of Observed Absorption Frequencies (cm-1) of BrNO
2
matrix isolated
Ar matrix
gas phase
2
O matrix
this work
assignment
Tevault et al.6
Feuerhahn et al.7
this work
Finlayson-Pitts et al.8
NO
NO
NO
2
2
2
sym str
scissor
asym str
1291 (1295, 1287)
784
1655 (1652)
1200
1289
784
1660
1203
1289
588
1723
1294
787
1667
1200
1292
787
≈1660
combination band
2
Figure 3. Gas-phase IR spectrum of BrNO after subtraction of water,
nitric acid and N O .
2 5
of ion chromatographic detection of 2Br- from Br2 was
2
Figure 4. IR spectrum of BrNO in an oxygen matrix.
determined by denuder measurements of pure Br2 vapor (≈290
-
reaction. First attempts to produce BrNO2 directly by the
reaction of N2O5 with NaBr solution were not successful. This
ppmv) to be (72 ( 10)%, where Br and small amounts of
-
-
BrO3 (<1% of Br ) were found. This low detection efficiency
-
may be explained by effective hydrolysis of N2O5 at low Br
is caused by the inability of the ion chromatographic system to
detect BrO , due to the low dissociation constant of HOBr. The
difference of the [Br ] value and the [BrNO2]g value in the first
-
-
concentrations as expected, wereas at high Br concentration
-
-
BrNO2 is lost by reaction with Br . ClNO2 is known to be
1
1,13
-9
more stable against hydrolysis than N2O5.
line of Table 1 then corresponds to (1.5 ( 0.4) × 10 mol/
3
The Br2 that was detected by UV spectroscopy is produced
probably by
cm Br2 in the gas phase, just overlapping with the UV
determination under similar conditions (slightly less Br2 is
expected at lower NaBr concentrations). BrCl (and ClNO2)
-
-
2
-
BrNO
2
+ Br f Br
2
+ NO
(6)
would result in a Cl signal in the denuder measurements and
were always below 5 ppmv.
+
(
transfer of Br ). By analogy to the reaction of ClNO2 with
1
2
Infrared Absorption Spectra
iodide the reactions
ClNO + Br f BrCl + NO
2
Figure 3 shows a gas-phase IR absorption spectrum of BrNO2
-
-
-
1
(7)
(
resolution 0.5 cm , 64 scans, 20.5 cm cell). The gas was
2
produced as described above and afterward dried in a cold trap
+
(
transfer of Cl ) and
-
10
3
-10
(
-70 °C). 1.8 × 10
mol/cm of HNO3, 1.6 × 10
mol/
3
cm of N2O5 and water were subtracted from this spectrum.
The BrNO2 concentration for this spectrum is determined to be
-
-
BrCl + Br f Br + Cl
(8)
2
-9
3
(
1.2 ( 0.4 ( 0.1) × 10 mol/cm (average of 16 measurements
may occur as well. In that case reaction 8 or other loss reactions
for BrCl must be effective so that BrCl does not escape into
the gas phase in substantial amounts. Assuming that ClNO2 is
lost only via reactions 2 and 7, which is justified by the strongly
increased uptake of ClNO2 on Br- solution, the branching
ratio(reaction 2)/(reaction 7) must at least be higher than the
yield [BrNO2]out/[ClNO2]in ≈ 0.12.
with 2σ statistical error and estimated calibration error δcal).
The corresponding molecular absorption cross section is given
on the right-hand axis of Figure 3. The observed absorption
-
1
bands are at 1667, 1294, 1200, and 787 cm
.
11
Figure 4 shows a spectrum of the cryogenically trapped gas
-1
where the matrix consists of oxygen (resolution 0.5 cm , 128
scans, T ) 20 K). The absorption frequencies in an O2 matrix
A part of the BrNO2 is lost in the flowtube by reaction 6
during its production. Higher BrNO2 concentrations can be
achieved if the reaction time is made shorter than given above.
On the other hand, this causes a fraction of ClNO2 to break
through the flowtube.
-1
are 1655 (1652), 1291 (1295, 1287), 1200, and 784 cm . The
frequencies in brackets are those of the smaller absorption for
the split bands. There is only a small frequency shift compared
6
to Ar matrix data given by Tevault et al.; the data of Feuerhahn
7
et al. differ significantly. Table 2 compares our data with the
-1
previous ones. The band at 1200 cm , which was not assigned
Discussion
6
by Tevault et al., always appears together with the other three
The method to produce BrNO2 in a continuous flow of carrier
gas allows us now to examine chemical and physical properties
of the molecule. Further increase of the yield should be possible
by variation of concentrations and reaction time. The secondary
product Br2 and excess ozone or NO2 may cause problems. The
bands and is probably a combination band or overtone.
Reaction Mechanisms
Up to now there are no distinctive data to decide whether
+
the net reaction 2 proceeds via the intermediate NO2 or in direct

1
6450 J. Phys. Chem., Vol. 100, No. 41, 1996
Letters
concentrations of ozone and NO2 can be reduced down to the
equilibrium NO2 + NO3 T N2O5 or possibly even lower by
using purified ClNO2 as reactant. The formation of Br2 probably
cannot be avoided; this complicates, e.g., mass spectrometric
investigations.
In the atmosphere the relatively high chloride concentration
in sea-salt aerosol, ranging from 0.5 mol/L (sea water) to >5
mol/L, should enhance the release of bromine species by
supplying the intermediate ClNO2. Because of its higher
stability against hydrolysis, ClNO2 can react with traces of Br-
at yields higher than N2O5 to form BrCl or BrNO2.
Further investigations of the reaction mechanisms are in
progress.
(2) Behnke, W.; Scheer, V.; Zetzsch, C. J. Aerosol Sci. 1994, 25 (Suppl.
), 277-278.
1
(
3) Barrie, L. A.; Bottenheim, J. W.; Hart, W. R. J. Geophys. Res.
994, 99, 25313-25314.
4) Wine, P. H.; Nicovich, J. M.; Stickel, R. E.; Zhao, Z.; Shackelford,
1
(
C. J.; Kreutter, K. D.; Daykin, E. P.; Wang, S. Halogen and sulphur reactions
relevant to polar chemistry. In The Tropospheric Chemistry of Ozone in
the Polar Regions; Niki, H.; Becker, K. H., Eds.; Springer: Berlin, 1993.
(5) Kreutter, K. D.; Nicovich, J. M.; Wine, P. H. J. Phys. Chem. 1991,
9
5, 4020-2028.
(6) Tevault, D. E. J. Phys. Chem. 1979, 83, 2217-2221.
(7) Feuerhahn, M.; Minkwitz, R.; Engelhardt, U. J. Mol. Spectrosc.
1979, 77, 429-439.
(8) Finlayson-Pitts, B. J.; Livingston, F. E.; Berko, H. N. J. Phys. Chem.
1
989, 93, 4397-4400.
9) Willner, H.; Simon, A. Patentschrift DE 3737515 C1, Deutsches
Patentamt, 1989.
(
Acknowledgment. We want to thank Prof. H. Willner for
the construction of the matrix-cryostat and BRUKER company
for a loan of the FTIR instrument and the design of the IR
microscope.
(10) Martin, K. Ph.D. Thesis, Universit a¨ t Hannover, 1993.
(11) Behnke, W.; George, C.; Scheer, V.; Zetzsch, C. J. Geophys. Res.,
in press.
(12) George, C.; Behnke, W.; Zetzsch, C.; Magi, L.; Ponche, J. L.;
Mirabel, P. Geophys. Res. Lett. 1995, 22, 1505-1508.
References and Notes
(13) George C.; Ponche, J. L.; Mirabel, P.; Behnke, W.; Scheer, V.;
Zetzsch, C. J. Phys. Chem. 1994, 98, 8780-8784.
(1) Finlayson-Pitts, B. J.; Livingston, F. E.; Berko, H. N. Nature 1990,
343, 622-625.
JP960319Z