Measurement of absolute absorption cross sections for nitrous acid (HONO) in the near-infrared region by the continuous wave cavity ring-down spectroscopy (cw-CRDS) technique coupled to laser photolysis

Article abstract of DOI:10.1021/jp203001y

Absolute absorption cross sections for selected lines of the OH stretch overtone 2ν₁ of the cis-isomer of nitrous acid HONO have been measured in the range 6623.6-6645.6 cm^-1 using the continuous wave cavity ring-down spectroscopy (cw-CRDS) technique. HONO has been generated by two different, complementary methods: in the first method, HONO has been produced by pulsed photolysis of H₂O₂/NO mixture at 248 nm, and in the second method HONO has been produced in a continuous manner by flowing humidified N₂ over 5.2 M HCl and 0.5 M NaNO₂ solutions. Laser photolysis synchronized with the cw-CRDS technique has been used to measure the absorption spectrum of HONO produced in the first method, and a simple cw-CRDS technique has been used in the second method. The first method, very time-consuming, allows for an absolute calibration of the absorption spectrum by comparison with the well-known HO₂ absorption cross section, while the second method is much faster and leads to a better signal-to-noise ratio. The strongest line in this wavelength range has been found at 6642.51 cm^-1 with σ = (5.8 ± 2.2) × 10^-21 cm².

Full text of DOI:10.1021/jp203001y

ARTICLE
pubs.acs.org/JPCA
Measurement of Absolute Absorption Cross Sections for Nitrous Acid
(HONO) in the Near-Infrared Region by the Continuous Wave Cavity
Ring-Down Spectroscopy (cw-CRDS) Technique Coupled to Laser
Photolysis
†
†
†
‡
,†
Chaithanya Jain, Pranay Morajkar, Coralie Schoemaecker, Bela Viskolcz, and Christa Fittschen*
†
0
PhysicoChimie des Processus de Combustion et de l Atmosph ꢀe re, Universit ꢁe Lille Nord de France, PC2A CNRS Universit ꢁe Lille 1,
UMR 8522, F-59650 Villeneuve d Ascq, France
0
‡
Department of Chemistry and Chemical Informatics, Faculty of Education, University of Szeged, Szeged, Boldogasszony sgt. 6, Hungary
6
725
S Supporting Information
b
ABSTRACT: Absolute absorption cross sections for selected lines of the
OH stretch overtone 2ν of the cis-isomer of nitrous acid HONO have
1
ꢀ
1
been measured in the range 6623.6ꢀ6645.6 cm using the continuous
wave cavity ring-down spectroscopy (cw-CRDS) technique. HONO has
been generated by two different, complementary methods: in the first
method, HONO has been produced by pulsed photolysis of H O /NO
2
2
mixture at 248 nm, and in the second method HONO has been produced
in a continuous manner by flowing humidified N over 5.2 M HCl and 0.5
2
M NaNO solutions. Laser photolysis synchronized with the cw-CRDS
2
technique has been used to measure the absorption spectrum of HONO
produced in the first method, and a simple cw-CRDS technique has been
used in the second method. The first method, very time-consuming,
allows for an absolute calibration of the absorption spectrum by compar-
ison with the well-known HO absorption cross section, while the second method is much faster and leads to a better signal-to-noise
2
ꢀ
1
ꢀ21
2
ratio. The strongest line in this wavelength range has been found at 6642.51 cm with σ = (5.8 ( 2.2) ꢁ 10
cm .
’
INTRODUCTION
transform infrared (FTIR) spectroscopy has also been used,
11
especially in laboratory studies. While the qualitative absorp-
tion spectra of HONO have been studied by many authors in the
Nitrous acid (HONO) is an important chemical species in the
atmosphere as well as in laboratory studies. In the atmosphere it
is, especially in polluted areas, a major photochemical precursor
for OH radicals in the early morning but has also been detected
in remote areas such as the south pole. HONO can be produced
in a simple gas phase reaction between OH radicals and NO, but
heterogeneous reactions seem to be more important, and
heterogeneous formation has been shown to occur on ice but
also on photocatalytic surfaces. In laboratory studies, it has
been identified as an heterogeneously formed, important OH
radical precursor in atmospheric simulation chambers, but also
as a byproduct in the photolysis of CH ONO, another impor-
tant OH-radical precursor for chamber studies. HONO can also
be an interesting OH-precursor for laser photolysis studies,
because it generates OH radicals after 351 nm excimer photo-
lysis, thus avoiding possible unwanted complications arising at
1
2ꢀ16
UVꢀvis and IR range,
the major difficulty using spectro-
1
scopic detection methods is the uncertainty linked to the
absolute absorption cross sections: HONO exists in an equilib-
rium with other components such as NO, NO , HNO , and H O
2
2
3
2
and is a rather unstable molecule, decomposing easily through
heterogeneous reactions. Therefore, determining the absolute
HONO concentration contained in an absorption cell is rather
3
4,5
17
difficult. Febo et al. have developed a method allowing the
generation of stable HONO flows with very high purity, and this
method has subsequently been used to determine absorption
6
7
3
18
19
cross sections in the UV and in the IR.
Cavity enhanced absorption spectroscopy is getting more and
more popular, especially in the near IR range, where cheap and
reliable components such as distributed feedback (DFB) lasers
and detectors are available. This makes this spectroscopic range
8,9
shorter, but more commonly used wavelengths.
HONO is often detected and quantified by spectroscopic
methods: UVꢀvis absorption using an open path differential
optical absorption spectroscopy (DOAS) technique has been
Received: March 31, 2011
Revised:
August 29, 2011
10
very successful for atmospheric measurements, but Fourier
Published: August 29, 2011
r 2011 American Chemical Society
10720
dx.doi.org/10.1021/jp203001y
|J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
Figure 1. Schematic diagram of the experimental setup. APD = avalanche photo diode, DL = diode laser, OI = optical isolator, BS = beam splitter,
AOM = acousto-optical modulator, M = mirror, and L = lens.
attractive, because now the rather small absorption cross sections
expected for overtone transitions occurring in this wavelength
range can be compensated for by the high sensitivity of the cavity
enhanced methods.
continuous. The pulsed method has allowed determining abso-
lute absorption cross sections, while the continuous method led
to a better signal-to-noise ratio. The results presented in this
work will allow quantitative determination of HONO in labora-
tory experiments.
HONO is known to exist in an equilibrium in two different
forms, trans- and cis-HONO, in a ratio of approximately 2 to 1 in
20
favor of the trans form. The rovibrational parameters of both
’
EXPERIMENTS
isomers, including the 2ν overtone of the OH stretch, have been
1
1
4,15
published:
the band center of the 2ν overtone of the cis-
isomer has been located at around 6665 cm , but no individual
absorption lines have been assigned in this wavelength range.
There are four main components to the experimental setup:
1
ꢀ
1
the photolysis cell, the photolysis laser, the cw-CRDS system,
and the laser-induced fluorescence (LIF) system. A detailed
description of this setup has already been published,
2
3,24
The 2ν overtone of the cis-isomer lies close to well-known and
and a
1
21,22
relatively intense 2ν overtone features of the HO radical,
schematic view of the experimental setup is given in Figure 1. The
photolysis cell has three axes and is made of stainless steel,
internally coated with Teflon. Photolysis is achieved along the
longest axis with an excimer laser (Lambda Physik LPX 202i)
operating at 248 nm. The beam entering the cell through a quartz
window has a size of approximately 1.5 ꢁ 3.0 cm, with an energy
1
2
an important intermediate in the oxidation of volatile organic
compounds (VOCs). The knowledge of absolute absorption
cross sections for some HONO absorption lines in this wave-
length range would allow for future, simultaneous measurements
of HONO and HO in laboratory studies.
2
ꢀ
2
Only one fairly indirect measurement of absorption cross
of 52 mJ cm .
7
sections for HONO is available in this range: Djehiche et al. have
The LIF system has been used in the frame of this work for
determining OH decays to accurately quantify the absolute H O
concentration in the photolysis cell. The procedure has been
detected HONO by continuous wave cavity ring-down spectros-
2
2
ꢀ1
copy (cw-CRDS) at 6625.69 cm as a reaction product during
24
the photolysis of CH ONO in a simulation chamber at a total
described in detail in a earlier work and will not be repeated
here. The only change compared to the earlier work is that we
now use instead of the photon counting devise a simple photo-
multiplier (Hamatsu R928) connected to a boxcar integrator
(EG&G 4121B).
cw-CRDS, coupled to laser photolysis for some experiments,
was used for the measurement of the HONO absorption
spectrum. The principle of this technique has been published
3
pressure of 40 Torr of air. Through simultaneous measurements
of CH O and subsequent modeling they deduced the theoretical
2
HONO concentration, postulating that it is formed in a reaction
of OH radicals with the precursor CH ONO. From this fairly
3
indirect approach they have estimated the absorption cross
ꢀ
21
2
section for cis-HONO to σ6625.69cmꢀ1 = 4.2 ꢁ 10 cm , assum-
ing that only 1/3 of the formed HONO is in cis-configuration.
In this work we present the measurement of the HONO
23
previously, and recent improvements concerning the synchro-
nization and acquisition can be found in a more recent work.
The mirrors used in this work (Los Gatos) had a reflectivity
ꢀ1
24
spectrum in the range 6623.6ꢀ6645.6 cm using two different,
complementary methods for generating HONO: pulsed and
1
0721
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
ꢀ1
roughly 3 cm ; the acquisition time of each portion was around
h. The method of HONO generation, described above, leads to
1
a stable HONO concentration on a short time scale (1 h).
However, the acquisition time of the whole spectrum was too
long to assume a constant HONO concentration. Therefore, at
the end of each individual portion, we have measured the
ꢀ
1
absorption line at 6643.17 cm and have normalized each
portion of the spectrum with respect to this individual line.
Thereafter, the HONO flow has been replaced by pure N , and
2
the baseline for the individual portion was measured.
All data have been acquired using an acquisition card
National Instruments PC-6259) installed in a personal compu-
(
ter, and all programs were written in the Lab View software
version 8.2.1. The experiments for the full spectrum were
conducted at a total pressure of 40 Torr. For method (a)
3
ꢀ1
4
0 cm min standard temperature and pressure (STP) He
Figure 2. Typical time-resolved absorption signal for HONO formed
was bubbled through a 50% H O solution (Sigma Aldrich), and
from H
2
O
2
photolysis in the presence of NO. Red dots, 40 Torr helium;
2 2
3
ꢀ1
1
cm min STP of NO (99%, Air Liquide) was added and
3
blue dots, 74 Torr N . The inset shows a zoom of the first 300 ms for the
2
ꢀ1
experiment in 40 Torr helium.
mixed with the main flow of 200 cm min STP He. A 52
2
ꢀ
mJ cm photolysis energy leads to an initial OH concentration
13
ꢀ3
of 3.3 ꢁ 10 cm , leading to a HONO concentration of 3.1 ꢁ
13
ꢀ3
of R = 0.999954, leading to ring-down times in the empty cavity
of around 60 μs. Under our conditions ([HONO] = 3.1 ꢁ
10 cm (for details see next paragraph). For method (b)
3
ꢀ1
50 cm min STP N was flown over HCl (maintained at 12 °C
by a thermostat) and NaNO solution and mixed with a main
flow of 100 cm min STP He, leading to an average HONO
2
1
3
ꢀ3
1
0 cm , 37 cm absorption path) the strongest absorption lines
2
3
ꢀ1
led to a decrease in ring-down time of around 5ꢀ10 μs.
The spectrum has been measured using two different methods
of HONO generation, one based on a pulsed photolytic genera-
tion (a) and one based on a continuous, chemical synthesis (b).
13
ꢀ3
concentration of 1.7 ꢁ 10 cm . The most intense line has
been measured also at different pressures (10, 40, and 73 Torr of
N and He).
2
(
a) In the first method, HONO was produced in situ by the
photolysis of hydrogen peroxide, H O , in presence of excess
’ RESULTS AND DISCUSSION
2
2
NO. Under these conditions, OH radicals initially formed from
the H O photolysis are converted nearly quantitatively to
The full HONO absorption spectrum has been measured in
2
2
4
0 Torr He using two different complementary methods of
HONO generation, a pulsed (a) and a continuous (b) manner.
a). Pulsed HONO Generation by Laser Photolysis. Photo-
lysis of H O at 248 nm leads to the formation of two OH
HONO in a fast reaction sequence (see Results and Discussion).
The ring-down events have to be measured relative to the photolysis
(
24
pulse as explained in the previous publication. A lab view
program started the data acquisition and triggered simulta-
neously a delay generator which in turn triggered after 2 s the
excimer laser. Ring-down events were recorded over a total
period of 4 s, that is, 2 s before and 2 s after the photolysis pulse
2
2
25,26
radicals,
which reacts in the absence of other reaction
partners in the following reaction sequence
H2O2 þ hν248nm f 2OH
ðR1Þ
ðR2Þ
ðR3Þ
ðR4Þ
(see Figure 2). To renew the gas mixture between two laser
pulses and to pump out most of the HONO produced during the
previous laser pulse, photolysis was carried out at a repetition rate
of maximal 0.08 Hz. To ensure that the data acquisition and the
fitting of the individual ring-down events had been completed
before the next photolysis pulse, the lab-view program was made
to verify that data acquisition and fitting of the ring-down events
is finished before starting the next data acquisition cycle. A
minimum of 500 ring-down events have been acquired over the
full 4 s measurement time (asking in average for 2ꢀ5 photolysis
pulses) before the cw-CRDS laser was tuned by the Lab View
program to the next wavenumber. The whole spectrum was
OH þ H O f HO þ H O
2
2
2
2
OH þ HO2 f O2 þ H2O
2
HO f H O þ O
2
2
2
2
In the beginning, OH radicals will predominantly react with
H O to form HO radicals R2. As the reaction advances and
2
2
2
HO radical concentration increases, reaction R3 becomes more
2
or less important, depending on initial H O concentration and
2
2
27
photolysis energy. Finally at longer reaction times, the HO2
radicals recombine R4. The H O concentration used for the
ꢀ
1
measured with a resolution of 0.005 cm ; that is, 4600 time-
2
2
resolved traces as shown in Figure 2 have been recorded.
HONO generation has been measured regularly by cw-CRDS
ꢀ1
(
b) In the second method, a stable concentration of HONO
through its absorption line at 6639.88 cm (σ = 2.23 ꢁ
ꢀ22
2
was produced in a continuous manner outside the cell by passing
10
cm , paper in preparation) and occasionally through
a calibrated flow of humidified N over 5.2 M HCl solution kept
OH decays (see Figure 3), obtained by LIF, and was [H O ] =
2
2 2
15
ꢀ3
at 12 °C followed by a 0.5 M solution of NaNO . The absorption
2.9 ꢁ 10 cm , while the time-resolved HO profile has been
ꢀ ꢀ19
2
2
1
2
spectrum was measured continuously; 25 ring-down events were
averaged before the cw-CRDS diode laser was tuned to the next
wavenumber. The whole spectrum was measured with a resolu-
measured at 6638.20 cm (σ40Torr = 2.93 ꢁ 10 cm ). A fit of
these HO -profiles using the well-known rate constants for
2
R2 ꢀ R4 leads to an initial OH concentration of [OH] = 3.3 ꢁ
0
ꢀ
1
13
ꢀ3
tion of 0.005 cm . The spectrum as shown in the Supporting
Information has been measured in seven individual portions of
10 cm . The calculation of the OH concentration using the
photolysis energy (52 mJ cm ) and H O absorption cross
ꢀ
2
2
2
1
0722
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
Figure 2: 7.2% of the initial OH radicals react through R3, and
4
0
.4% react through R7; we can thus calculate [HONO] =
13
ꢀ3
.884 ꢁ [OH] = 2.9 ꢁ 10 cm
.
0
The systematic error introduced by uncertainties in the
reaction mechanism is considered to be small: an error in the
initial OH-radical concentration would of course have a direct,
linear impact on the obtained HONO concentration; however,
the impact on the HONO yield itself through a changed impact
of R3 and R7 is negligible (see example further down). The initial
OH concentration is mainly obtained from the absolute HO2
concentrations, measured by cw-CRDS, while the H O con-
2
2
centration, measured through pseudo-first order OH decays and
by absorption in the near IR, has only a minor impact on the
extraction of the initial OH concentration. The absorption cross
2
1,22
section of HO is thought to be known to better than (20%,
2
we therefore estimate the initial OH concentration to be known
with the same accuracy. A possible systematic error source for the
conversion of OH radical into HONO is impurity of NO , arising
2
from an oxidation of NO by trace amounts of O : NO reacts
2
2
with both, OH and HO , in addition reactions and would
2
Figure 3. OH and HO
2
profiles for an experiment at 10 Torr N
2
with
= 6.55 ꢁ 10 cm . Upper
graph OH-decays: red dots, [NO] = 0; black dots, [NO] = 2.5 ꢁ
therefore decrease the HONO concentration. To illustrate the
impact of such an impurity, typical experimental traces, together
with model simulations, are shown for an experiment in 10 Torr
15
ꢀ3
12
ꢀ3
[H
2
O
2
] = 1.1 ꢁ 10 cm and [OH]
0
15
ꢀ3
10
cm ; full lines represent the simulation with rate constants as
of N . The upper graph in Figure 3 shows experimental OH
2
given in the text, dashed line the simulation with the additional impurity
1
3
ꢀ3
profiles together with simulations in the absence and presence of
of [NO
2
] = 2.5 ꢁ 10 cm ; lower graph experimental HO
2
profiles
15
ꢀ3
NO (2.5 ꢁ 10 cm ): in the presence of NO, the OH decay is,
under these conditions, slowed down due to the cycling of HO
2
with simulations: blue line, simulated HONO profile without NO
2
impurity; green line, HONO profile in presence of NO
2
impurity.
2 28
radicals by R6a. Impurities of NO however increase the decay
2
20
= 9.37 ꢁ 10ꢀ cm ) leads to
rate of OH radicals drastically due to a reaction of NO with both
section at 248 nm (σ
2
H O ,248nm
2
2
1
3
ꢀ3
OH and HO radicals: the dashed line in the upper graph
[
OH] = 3.52 ꢁ 10 cm , in very good agreement with the value
2
13
ꢀ3
represents an impurity of 2.5 ꢁ 10 cm , that is, 1% of the
NO concentration. An impurity on this order of magnitude
would be noticeable in our OH decays, and we therefore estimate
obtained from cw-CRDS measurements.
The addition of NO to the reaction mixture will lead to a
competition between R2 and
an upper limit of the NO impurity of 1% of the NO concentra-
2
OH þ NOð þ MÞ f HONOð þ MÞ
The reaction R5 is still in the falloff range at pressures of 40 Torr
ðR5Þ
tion. The lower graph in Figure 3 shows the corresponding HO2
profiles: the red curve in the absence of NO is the basis for the
calculation of the initial OH radical concentration, and it can be
29
ꢀ31
He with a low pressure rate constant of k5,0 = 6.0 ꢁ 10
ꢁ
6
ꢀ2 ꢀ1
seen from the black dots that after NO addition HO decreases
2
[
He] cm molecule
s . However, HO2 radicals formed
rapidly to concentrations below our detection limit. The upper
blue line shows the simulated HONO profile, which under these
conditions corresponds to a HONO yield of [HONO] = 0.943 ꢁ
through R2 even in the presence of NO in competition will
rapidly be transformed back to OH through the fast reaction
HO2 þ NO f OH þ NO2
The competing reaction channel
HO2 þ NOð þ MÞ f OH þ HNO3
ðR6aÞ
[OH] (a lower initial radical concentration compared to the
0
conditions used for the complete spectrum leads to a decrease of
the impact of R3). The green line shows the simulated HONO
profile obtained in the presence of a 1% NO impurity (2.5 ꢁ
ðR6bÞ
2
13
ꢀ3
1
0
cm ): under these conditions, the HONO yield drops to
has been found to be very minor at room temperature at 200 Torr
0
.89, that is, a decrease of 5%. From the combined uncertainties
3
0
N , so it can safely been neglected under our conditions
2
(σHO2, rate constants and NO impurity), we estimate the
2
(40 Torr He). Another possible competing reaction for OH
radicals is
uncertainty of the HONO concentration, and thus on the
obtained absorption cross sections, to be (30%.
From the time-resolved cw-CRDS signals (typical examples
are shown in Figure 2), the absorbance αt=0 and, now with
knowledge of the photolytically generated HONO concentra-
tion, the absorption cross section σ can be extracted from:
OH þ HONO f H O þ NO2
and will gain some importance at longer reaction times with a rate
ðR7Þ
2
3
1
ꢀ12
3
ꢀ1 ꢀ1
constant of k = 7.0 ꢁ 10
cm molecule
s . The rate
7
ꢀ
12
3
ꢀ1 ꢀ1 32
constants for R2 (k = 1.7 ꢁ 10
cm molecule
s
)
and
cm
2
ꢀ
ꢁ
ꢀ
31
6
ꢀ2 ꢀ1
ꢀ13
3
RL
c
1
1
R5 (k = 6.0 ꢁ 10
cm molecule
s
= 8.4 ꢁ 10
5
αt ¼0 ¼ ½HONOꢂ ꢁ σ ¼
ꢀ
ð1Þ
ꢀ
1
ꢀ1
29
molecule
s
at 40 Torr) are comparable, leading at con-
τ
τ0
t ¼0
1
5
ꢀ3
centrations of H O (2.9 ꢁ 10 cm ) and NO (5.1 ꢁ
2
2
1
5
ꢀ3
ꢀ
first
1
0
cm ) to similar pseudofirst-order rate constants of k
where αt=0 is the absorbance immediately after the photolysis
2,5
1
≈
5000 s . A simple simulation shows that under this condition
8.4% of initial OH radicals will be transformed into HONO
within around 1 ms, that is, immediately on the time scale of
pulse, R is the ratio between the cavity length L, that is, the
L
8
distance between the two cavity mirrors, to the length L over
A
which the absorber is present (in our case the overlap of
1
0723
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
Figure 4. HONO spectrum obtained by laser photolysis: upper, black line is baseline (τ ); middle, green line is τ obtained through fits of individual,
0
t=0
time-resolved HONO profiles to eq 2 (right Y-scale); and red line is absorbance α obtained from eq 1 (left Y-scale).
photolysis beam and absorption path), c is the speed of light, τ0
and τt=0 being the ring-down times before and immediately after
We therefore are confident that the time-resolved evolution of
the absorbance is due to diffusion phenomena only. One can
notice that the renewal of the gas mixture is probably not totally
completed just before the next photolysis pulse; that is, some
HONO is still left in the photolysis cell (the baseline is not
completely flat). This however has no influence on the calibra-
tion of the absorption measurement, because the laser photolysis
technique accounts only for HONO that is formed from the
actual photolysis pulse, all possible residual HONO will also
contribute with the same intensity to the baseline. The same is
true for any absorption due to other species present in the
the photolysis pulse, respectively. To extract τt=0 and τ , the time-
0
resolved signals as shown in Figure 2 have been analyzed for each
wavelength by a data analysis program written in Lab View. τ has
0
been obtained as the intercept of the linear regression of all of the
ring-down events that occurred during the 2 s before the
photolysis pulse. For extracting τt=0, all ring-down events
occurred after the photolysis pulse have been fitted by a two
phase association
τt ¼ τt ¼0 þ að1 ꢀ eꢀkfasttÞ þ bð1 ꢀ eꢀkslowt
Þ
ð2Þ
photolysis cell such as H O: these absorptions will not be visible,
2
because it affects τ and τ in the same way (see Figure 7).
0
t=0
HONO is a stable product; therefore the two empirical rate
constants kfast and kslow can be interpreted by two different loss
To illustrate the measurement principle, a small portion of the
obtained spectrum (although not typical for the whole spectrum,
processes. The rapid decay with an average rate of k
≈
ꢀ1
fast,40TorrHe
because measured with a higher resolution of 0.001 cm ) is
ꢀ
1
1
2 s for the red dots (40 Torr Helium) can be linked to the
presented in Figure 4: each set of dots (black, green, and red) at
one given wavelength is the result of one time-resolved HONO
signal as shown in Figure 2. The upper black line is the baseline,
as obtained from ring down events acquired just before the
photolysis pulse. The middle green dots have been obtained as
τt=0 from fits of the individual, time-resolved HONO profiles to
eq 2. The lower red line represents the absorbance α as calculated
from eq 1. In this manner the spectrum accessible with our DFB
diffusion of HONO out of the photolysis volume (to illustrate
this process, the same experiment in 74 Torr of N is also shown
2
in Figure 2 (blue dots): now, the rate is only k
≈
fast,74TorrN2
ꢀ1
2
s ). This process is completed after roughly 200 ms (more than
1
s for 74 Torr N ); that is, after this delay the HONO, generated
2
initially in the volume limited by the size of the photolysis beam,
has diffused into the entire volume. The following slow loss k
slow
ꢀ1
with an average rate of kslow ≈ 0.16 s can possibly be
approximated as the loss of HONO through processes such as
the renewal of the gas mixture by pumping and heterogeneous
losses on the reactor wall. To evaluate if this assumption is
reasonable, one can carry out a simple estimation: the photolysis
ꢀ1
laser (6623.6ꢀ6645.6 cm ) has been measured with a resolu-
ꢀ
1
tion of 0.005 cm and is presented in Figure 5. The repetition
rate of the experiment was very low (0.08 Hz maximal); therefore
the total measurement time for the spectrum shown in Figure 5
spread over 2 weeks. The generated HONO concentration
2
cell has an inner diameter of 6 cm leading to a surface of 28.3 cm ,
2
depends directly on the H O concentration and on the photo-
while the photolysis beam has a surface of 4.5 cm , that is, a ratio
2 2
lysis energy. To ensure that the HONO concentration was
constant over the whole measurement period, the H O con-
of 6.2 between photolyzed over total volume (for simplicity the
side arms of the reaction cell have not been taken into account,
which would increase this ratio). Taking a typical time-resolved
signal such as shown in Figure 2, one can calculate the ratio of the
2
2
centration has regularly been checked by measuring its absorp-
ꢀ
1
tion line at 6639.88 cm . The photolysis energy has also
regularly been checked by a calibrated power meter. Occasion-
absorbencies α: taking into account that R (see eq 1) evolves
L
from 2.22 just after the laser pulse (0.82/0.37) to 1.05 (0.82/
ally, HO
2
concentrationꢀtime profiles have been recorded in the
0
.78) at long reaction times, one finds a ratio of αt=0 over αt=200ms
absence of NO, verifying simultaneously the H
and the photolysis energy.
O concentration
2 2
of 7.2, in good agreement with the above calculated volume ratio.
1
0724
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
Table 1. Line Strengths of Some Selected HONO Absorption
a
Lines in the Near-Infrared Region
wavenumber
absorption cross-section
background
ꢀ
1
ꢀ21
2
(cm
)
(10
cm )
absorption (%)
6
6
6
6
6
6
625.01
637.36
638.26
642.51
643.17
644.00
1.6 ( 1.0
3.5 ( 1.6
3.8 ( 1.6
5.8 ( 2.2
4.2 ( 1.7
4.8 ( 1.9
26
38
30
37
33
52
a
Errors have been estimated from the signal-to-noise ratio (Δα = 1.5 ꢁ
ꢀ
8
ꢀ1
1
0
cm ) plus 30% for other errors such as drift in H O concentra-
2 2
tion, photolysis energy, and uncertainties in the retrieval of the HONO
concentration. In the third column is added the percentage of this
absorption cross section that is due to the congested background
absorption.
Figure 5. Full spectrum of HONO accessible with the DFB diode,
obtained by pulsed HONO generation and calibrated to the six
individual absorption lines indicated by an arrow.
HONO or from heterogeneous reactions at longer reaction times
will not be detected: all absorption features appearing directly
after the photolysis pulse are inherent to HONO. To obtain a
spectrum with better signal-to-noise ratio, we have therefore
remeasured the whole spectrum using a continuous source of
It has been mentioned that the spectrum shown in Figure 5 has
been measured for practical reasons at a low spectral resolution
with a poor signal-to-noise ratio. However, this spectrum has
been used to choose six of the most intense absorption lines to be
remeasured very carefully with high spectral resolution to obtain
a significant calibration. Six absorption lines, indicated by an
arrow in Figure 5, have therefore been measured using a higher
HONO: humidified N was flown first over a 5.2 M HCl solution
2
(kept at 12 °C) and then over a 0.5 M NaNO solution. First tests
2
1
7
using the method developed by Febo et al. did not lead to any
signal, probably due to very low concentrations. We then tried a
ꢀ
1
resolution (0.001 cm , the same as shown in Figure 4) and
taking special care on the verification of the absolute HONO
concentration: before starting the acquisition of each line, the
HO and OH concentrationꢀtime profiles as well as the H O
19
variant published by Barney et al. , in which humidified N is
2
flown over 5.2 M HCl (kept at 0.0 °C) and solid NaNO .
2
2
2
2
However, the HONO concentration that could be generated by
this method was still too low to obtain a spectrum with a
satisfying signal-to-noise ratio. Therefore, a slight modification
ꢀ
1
absorption around 6639.88 cm have been measured in the
absence of NO. The six individual lines have been submitted as
Supporting Information, and the peak absorption cross sections
of theses lines in 40 Torr helium are listed in Table 1. It can be
seen from Figure 4 (and from the figures in the Supporting
Information) that a broad absorption underlies the individual
absorption lines: this background absorption is unequivocally
HONO, because it is generated by the photolysis pulse and is
probably due to an unresolved, dense spectrum. Sometimes it is
not possible to record a baseline in the absence of HONO;
therefore we have added in Table 1 the percentage of the peak
absorption cross section that is due this background absorption.
Please keep in mind that HONO is in equilibrium between its cis-
and trans-isomer; that is, only around 1/3 of the photolytically
generated HONO is present in form of the cis-isomer. This fact
has not been taken into account in the calculation of the
absorption cross sections, because the result of this work should
be seen as a tool enabling the simple quantification of HONO in
laboratory experiments. Under this point of view it seems idle to
take into account that the cisꢀtrans equilibrium and absorption
cross sections given in this work can therefore be directly used to
obtain total HONO concentrations and not only the concentra-
tion of the cis-population.
using NaNO in solution has been employed: this way, much
2
higher concentrations of HONO could be obtained. A drawback
of this modification is (a) the high water concentration in the
reactor (H O has strong absorption lines in this wavelength
2
region) and (b) the instability of the HONO concentration on
the longer time scale: a slow increase in the HONO concentra-
tion on the hour-time scale was observed, possibly due to an
increase in NaNO concentration following the slow evaporation
2
of water. This change in HONO concentration was too fast to
allow measuring the whole spectrum considering identical HONO
concentrations. Therefore the spectrum has been measured in
seven individual portions, and each portion has been calibrated to
ꢀ
1
a reference: one of the HONO lines (6643.17 cm ) has been
chosen as reference line and has been regularly measured before
and after the acquisition of each portion of the spectrum; the
absorbance of the entire portion was calibrated subsequently to
the intensity of this line.
Once the whole spectrum has been measured and all individual
portions have been calibrated to the reference line, the absorbance α
of the six HONO lines shown as Supporting Information were
plotted against the absolute absorption coefficients σ of the same
lines obtained by the pulsed method and tabulated in Table 1. This
should present a linear relationship if the reference line calibration
for the continuous spectrum is valid: as can be seen in Figure 6, this
seems to be the case. The slope of a linear regression of this plot
represents the HONO concentration generated by the continuous
(
b). Continuous HONO Generation from HCl and NaNO . It
2
has been explained that the spectrum obtained by generating
HONO in a pulsed manner could, for practical reasons, be
obtained with a poor signal-to-noise ratio only. The pulsed
method has other advantages, noticeably that (a) the absolute
HONO concentration can be known from a calibration relative
1
3
ꢀ3
method: (1.58 ( 0.2) ꢁ 10 cm .
to HO profiles and that (b) other stable species present in the
A comparison of the spectra, obtained with both methods, is
shown for a small portion in Figure 7; the entire spectrum has
2
cell such as H O or products formed in an equilibrium with
2
1
0725
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
Figure 6. Plot of absorbance α obtained by continuous HONO
production against the absorption cross sections σ obtained by pulsed
HONO generation for the six lines shown as Supporting Information.
ꢀ
1
Figure 8. HONO absorption line at 6642.51 cm : upper graph at 10
Figure 7. Small portion of HONO spectrum, obtained by both
methods: the green line shows the spectrum obtained by laser photolysis
method; the black line shows the spectrum obtained by continuous
Torr N , lower graph at 74 Torr N . The black lines represent the
2
2
individual absorption lines identified by the Fityk program through a fit
of the experimental data (black dots) to a Voigt profile.
HONO production, and red spikes show the H O spectrum, as obtained
2
by Macko et al.
measured in 40 Torr N /O , while the spectrum in this work has
2
2
been obtained in 40 Torr He; therefore, the value of Djehiche
et al. would be expected to be slightly smaller than the value
obtained in this work due to pressure broadening (see next
paragraph). However, the 30% higher absorption cross section
from Djehiche et al. can still be considered in good agreement,
looking at the very indirect method they have used to extract the
absorption cross section.
been submitted as Supporting Information to this publication.
The green line represents the spectrum obtained using the pulsed
method, and the black line represents the spectrum obtained by
the continuous method. For comparison, the position of H O
2
33
absorption lines as published by Macko et al. have been added
to the graph (red spikes). It can be seen that the H O lines are
2
clearly visible in the continuous HONO spectrum but do not
(c). Pressure Broadening of HONO Absorption Lines. The
goal of this work is to provide a tool for measuring absolute
HONO concentrations in laboratory experiments under reduced
pressures. For this reason, we have investigated the pressure
appear in the pulsed HONO spectrum. The HONO absorption
ꢀ
1
features on the other hand (for example, 6638.26 or 6639.36 cm )
are clearly visible in both spectra, continuous and pulsed. There
is, however, a limitation even to the pulsed method: very strong
ꢀ
1
broadening of the most intense line at 6642.51 cm in the
ꢀ
1
H O-absorption lines (for example, 6626.5 cm , see Supporting
pressure range 10ꢀ74 Torr with He and N as bath gas. In
2
2
Information) can decrease the intensity of the ring-down events
below the trigger threshold and preventing measuring HONO
absorption features.
Figure 8 are shown the two lines obtained at 10 Torr N (upper
2
graph) and at 74 Torr N (lower graph). It is apparent (and has
2
already been discussed) that the background absorption is a
convolution of nonresolved absorption lines, leading to broad
background absorption. In Figure 8 we have illustrated this by
adding the result of a least-squares fit of the experimental data to
individual, Voigt-shaped absorption lines, obtained by the free
The result of this work can be compared to the only published
7
absorption cross section, deduced by Djehiche et al. for the line
ꢀ
1
ꢀ21
2
at 6625.69 cm : σ = 4.2 ꢁ 10
cm . From looking at the
whole spectrum, published as Supporting Information, it can be
seen that this is a rather small line with an absorption cross
34
software fitting program Fityk (Version 0.9.8). From this figure
it is obvious that it will not be useful to extract line strengths and
pressure broadening coefficients for individual lines to propose a
reliable tool to extrapolate absorption cross sections under various
experimental conditions. Instead, we present the absorption cross
ꢀ21
2
7
section of σ = 3.2 ꢁ 10
cm (Djehiche et al. have taken into
account the cisꢀtrans equilibrium, while in this work it has not
been taken into account; hence σ from the Supporting Informa-
tion has been multiplied by 3 for comparison). Djehiche et al.
1
0726
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
ꢀ
1a
Table 2. Pressure-Dependent Absorption Cross Sections at the Center of the Strongest Line at 6642.51 cm
He
N
2
pressure (Torr)
ꢀ21
6642.51cmꢀ1 (10
2
ꢀ21
6642.46cmꢀ1 (10
2
ꢀ21
6642.51cmꢀ1 (10
2
ꢀ21
2
σ
cm )
σ
cm )
σ
cm )
σ
6642.46cmꢀ1 (10 cm )
1
4
7
0
0
4
7.0 ( 2.6
5.8 ( 2.2
4.6 ( 1.8
1.8 ( 1.0
2.1 ( 1.1
6.6 ( 2.5
5.1 ( 2.0
4.3 ( 1.8
1.8 ( 1.0
2.3 ( 1.2
2.4 ( 1.2
2.3 ( 1.2
ꢀ1
a
For practical purposes, the absorption cross section at 6642.46 cm is also given (see text).
sections in the center of the strongest absorption line in Table 2
for the three individual pressures and for both bath gases, He and
N . It is not always possible to measure the baseline in the
2
’ REFERENCES
(1) Volkamer, R.; Sheehy, P.; Molina, L. T.; Molina, M. J. Atmos.
Chem. Phys. 2010, 10, 6969–6991.
absence of HONO, for example, in atmospheric simulation
chambers. For this reason, we have also added the pressure-
(2) Liao, W.; Case, A. T.; Mastromarino, J.; Tan, D.; Dibb, J. E.
Geophys. Res. Lett. 2006, 33, L09810.
ꢀ
1
(
3) Hellebust, S.; Roddis, T.; Sodeau, J. R. J. Phys. Chem. A 2007,
11, 1167–1171.
4) Monge, M. E.; D'Anna, B.; George, C. Phys. Chem. Chem. Phys.
010, 12, 8991–8998.
5) Langridge, J. M.; Gustafsson, R. J.; Griffiths, P. T.; Cox, R. A.;
dependent minimum absorption cross section at 6642.46 cm
.
1
(
2
’
CONCLUSION
The near-IR spectrum of a portion of the 2ν absorption band
(
1
Lambert, R. M.; Jones, R. L. Atmos. Environ. 2009, 43, 5128–5131.
(6) Rohrer, F.; Bohn, B.; Brauers, T.; Br €o ning, D.; Johnen, F.-J.;
Wahner, A.; Kleffmann, J. Atmos. Chem. Phys. 2005, 5, 2189–2201.
of the cis-isomer of HONO has been measured at a total pressure
ꢀ
1
of 40 Torr helium in the range 6623.6ꢀ6645.6 cm , using two
different methods for generating HONO. Absolute absorption
cross sections for selected nitrous acid lines have been deter-
mined through calibration of the generated HONO concentra-
(
7) Djehiche, M.; Tomas, A.; Fittschen, C.; Coddeville, P. Environ.
Sci. Technol. 2011, 45, 608–614.
8) Jain, C.; Parker, A. E.; Schoemaecker, C.; Fittschen, C. Chem-
PhysChem 2010, 11, 3867–3873.
9) Aluculesei, A.; Tomas, A.; Schoemaecker, C.; Fittschen, C. Appl.
Phys. B: Lasers Opt. 2008, 92, 379–385.
(
tion against the well-known HO absorption cross section. The
2
(
strongest line in this wavelength range has been found at
ꢀ
1
ꢀ21
2
6
642.50 cm with σ = (5.8 ( 2.2) ꢁ 10
cm . Using current
(10) Perner, D.; Platt, U. Geophys. Res. Lett. 1979, 6, 917–920.
cw-CRDS setups, HONO concentrations of as low as 1 ꢁ
(11) Karlsson, R. S.; Ljungstrom, E. B. Environ. Sci. Technol. 1996,
1
2
ꢀ3
ꢀ9
ꢀ1
1
0
cm (corresponding to α = 5 ꢁ 10 cm ) can easily
30, 2008–2013.
be quantified at 40 Torr He. Pressure broadening up to 74 Torr
(12) Yamano, D.; Yabushita, A.; Kawasaki, M.; Perrin, A. J. Quant.
Spectrosc. Radiat. Transfer 2010, 111, 45–51.
has been measured of the most intense line for He and N as bath
2
(
13) Sironneau, V.; Orphal, J.; Demaison, J.; Chelin, P. J. Phys. Chem.
A 2008, 112, 10697–10702.
14) Guilmot, J. M.; Godefroid, M.; Herman, M. J. Mol. Spectrosc.
993, 160, 387–400.
15) Guilmot, J. M.; Melen, F.; Herman, M. J. Mol. Spectrosc. 1993,
60, 401–410.
gas. The results show that absorption spectroscopy in the near-IR
range is not suitable for atmospheric measurements due to a too
small absorption cross section combined with strong pressure
broadening. However, it can provide an interesting tool for
laboratory studies.
(
1
(
1
(16) Stockwell, W. R.; Calvert, J. G. J. Photochem. 1978, 8, 193–203.
(
17) Febo, A.; Perrino, C.; Gherardi, M.; Sparapani, R. Environ. Sci.
’
ASSOCIATED CONTENT
Technol. 1995, 29, 2390–2395.
18) Stutz, J.; Kim, E. S.; Platt, U.; Bruno, P.; Perrino, C.; Febo, A.
J. Geophys. Res. 2000, 105, 14585–14592.
19) Barney, W. S.; Wingen, L. M.; Lakin, M. J.; Brauers, T.; Stutz, J.;
(
S
Supporting Information. Six individual lines measured
b
at high resolution as well as the full spectrum as obtained with
both methods. This material is available free of charge via the
Internet at http://pubs.acs.org.
(
Finlayson-Pitts, B. J. J. Phys. Chem. A 2000, 104, 1692–1699.
(20) Varma, R.; Curl, R. F. J. Phys. Chem. 1976, 80, 402–409.
(
21) Thiebaud, J.; Crunaire, S.; Fittschen, C. J. Phys. Chem. A 2007,
11, 6959–6966.
22) Tang, Y.; Tyndall, G. S.; Orlando, J. J. J. Phys. Chem. A 2010,
1
’
AUTHOR INFORMATION
(
Corresponding Author
114, 369–378.
*E-mail: christa.fittschen@univ-lille1.fr.
(23) Thiebaud, J.; Fittschen, C. Appl. Phys. B: Lasers Opt. 2006,
5, 383–389.
8
(24) Parker, A.; Jain, C.; Schoemaecker, C.; Szriftgiser, P.; Votava,
O.; Fittschen, C. Appl. Phys. B: Lasers Opt. 2011, 103, 725–733.
(25) Thiebaud, J.; Aluculesei, A.; Fittschen, C. J. Chem. Phys. 2007,
’
ACKNOWLEDGMENT
1
26, 186101.
26) Vaghjiani, G. L.; Ravishankara, A. R. J. Chem. Phys. 1990,
2, 996–1003.
27) Johansson, O.; Bood, J.; Alden, M.; Lindblad, U. Appl. Phys. B:
Lasers Opt. 2009, 97, 515–522.
28) Vaghjiani, G.; Ravishankara, A. J. Geophys. Res. 1989,
The laboratory participates in the Institut de Recherche en
(
ENvironnement Industriel (IRENI), which is financed by R ꢁe gion
Nord Pas-de-Calais, the Minist ꢀe re de l'Enseignement Sup ꢁe rieur
et de la Recherche, the CNRS, and European Regional Devel-
opment Fund (ERDF). The authors are thankful for financial
support through the “Program Hubert Curien” Balaton. C.J.
thanks the European Union for financial support through project
MEST-CT-2005-020659.
9
(
(
9
4, 3487–3492.
(29) Forster, R.; Frost, M.; Fulle, D.; Hamann, H. F.; Hippler, H.;
Schlepegrell, A.; Troe, J. J. Chem. Phys. 1995, 103, 2949–2958.
1
0727
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728

The Journal of Physical Chemistry A
ARTICLE
(
30) Butkovskaya, N. I.; Kukui, A.; Pouvesle, N.; Bras, G. L. J. Phys.
Chem. A 2005, 109, 6509–6520.
31) Burkholder, J. B.; Mellouki, A.; Talukdar, R.; Ravishankara,
A. R. Int. J. Chem. Kinetics 1992, 24, 711–725.
32) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.;
(
(
Hampson, R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J.
Atmos. Chem. Phys. 2004, 4, 1461–1738.
(33) Macko, P.; Romanini, D.; Mikhailenko, S. N.; Naumenko,
O. V.; Kassi, S.; Jenouvrier, A.; Tyuterev, G.; Campargue, A. J. Mol.
Spectrosc. 2004, 227, 90–108.
(34) Wojdyr, M. J. Appl. Cryst. 2010, 43, 1126–1128.
1
0728
dx.doi.org/10.1021/jp203001y |J. Phys. Chem. A 2011, 115, 10720–10728