Kinetics of HCCL + NOx reactions

Article abstract of DOI:10.1002/kin.10013

The kinetics of reactions of HCCl with NO and NO₂ were investigated over the temperature ranges 298-572 k and 298-476 k, respectively, using laser-induced fluorescence spectroscopy to measure total rate constants and time-resolved infrared diode laser absorption spectroscopy to probe reaction products. Both reactions are fast, with k(HCCl + NO) = (2.75 ± 0.2) × 10^-11 cm³ molecule^-1 s^-1 and k(HCCl+NO₂) = (1.10 ± 0.2) × 10^-10 cm³ molecule^-1 s^-1 at 296 K. Both rate constants displayed only a slight temperature dependence. Detection of products in the HCCl + NO reaction at 296 K indicates that HCNO + Cl is the major product with a branching ratio of φ = 0.68 ± 0.06, and NCO + HCl is a minor channel with φ = 0.24 ± 0.04.

Full text of DOI:10.1002/kin.10013

Kinetics of HCCl + NOx
Reactions
RANDALL E. BAREN, MARK A. ERICKSON, JOHN F. HERSHBERGER
Department of Chemistry, North Dakota State University, Fargo, North Dakota 58105
Received 2 April 2001; accepted 9 July 2001
ABSTRACT: The kinetics of reactions of HCCl with NO and NO were investigated over the
2
temperature ranges 298–572 k and 298–476 k, respectively, using laser-induced fluorescence
spectroscopy to measure total rate constants and time-resolved infrared diode laser absorp-
tion spectroscopy to probe reaction products. Both reactions are fast, with k(HCCl + NO) =
�
11
3
� 1 � 1
� 10
3
(
2.75 � 0.2) � 10
cm molecule
s
and k(HCCl + NO ) = (1.10 � 0.2) � 10
cm mole-
2
�
1
� 1
cule
s at 296 K. Both rate constants displayed only a slight temperature dependence.
Detection of products in the HCCl + NO reaction at 296 K indicates that HCNO + Cl is the ma-
jor product with a branching ratio of � = 0.68 � 0.06, and NCO + HCl is a minor channel with
�
= 0.24 � 0.04. �C 2001 John Wiley & Sons, Inc. Int J Chem Kinet 34: 12–17, 2002
INTRODUCTION
is any information on product channels available for
either title reaction.
The detailed gas-phase chemistry of nitrogen oxides is
of great interest because of the role these compounds
play in pollutant emissions from fossil fuel combustion
We report here kinetic studies of the reactions of
˜
1
0
HCCl (X A ) with NO and NO2. HCCl is produced
by the 193-nm excimer laser photolysis of CHClBr2.
This precursor is known to produce HCCl from previ-
ous kinetic and spectroscopic studies [4–6], and cannot
produce CH or CCl radicals from a single 193-nm pho-
ton. HCCl is probed by laser-induced fluorescence near
603 nm to measure total rate constants [3]. Both title
reactions have several possible product channels:
[1]. Chlorinated hydrocarbons are frequently present in
hazardous waste incinerator emissions [2]. The kinetics
ofchlorinatedhydrocarbonfragmentssuchasHCClare
therefore of interest in efforts to model these processes.
Unlike the unsubstituted carbene CH2, the ground
electronic state of HCCl is a singlet. Although this
species has been characterized spectroscopically [3,4],
relatively little experimental data on HCCl kinetics are
available. In general, HCCl is known to be moderately
reactive with unsaturated hydrocarbons and NO, but
unreactive with O2 [5,6]. One previous study of the
HCCl + NO reaction has been reported, in which a
HCCl + NO → HCNO + Cl
0
� 1
� 1
� 1
� 1
1
r H = � 129.6 kJ mol
(1a)
(1b)
(1c)
(1d)
(1e)
(2a)
HCCl + NO → HNCO + Cl
0
1 H = � 405.7 kJ mol
r
�
11
room temperature rate constant of (1.5 � 0.5) � 10
HCCl + NO → NCO + HCl
3
� 1 � 1
cm molecule
s was reported [5]. No rate con-
0
1
r H = � 386.2 kJ mol
stant for HCCl + NO2 kinetics has been published, nor
HCCl + NO → HCN + ClO
0
1
r H = � 188.9 kJ mol
Correspondence to: John F. Hershberger; e-mail: john
hershberger@ndsu.nodak.edu.
HCCl + NO → CN + HOCl
Contract grant sponsor: Division of Chemical Sciences, Office
of Basic Energy Sciences of the Department of Energy.
Contract grant number: DE-FG03-96ER14645.
�
c 2001 John Wiley & Sons, Inc.
DOI 10.1002/kin.10013
0
� 1
1
r H = � 64.7 kJ mol
HCCl + NO2 → HCNO + ClO
0
� 1
1r H = +92.5 kJ mol

KINETICS OF HCCL + NOx REACTIONS
13
HCCl + NO2 → HNCO + ClO
NO2 (Matheson) was purified by vacuum distilla-
tion at 77 and 220 K. NO (Matheson) was purified by
vacuum distillation at 180 K. CH2CO (ketene) was syn-
0
� 1
� 1
� 1
1
r H = � 368.5 kJ mol
(2b)
(2c)
(2d)
HCCl + NO2 → NCO + HOCl
�
thesized by pyrolysis of acetic anhydride at 700 C and
0
1
r H = � 311.2 kJ mol
purified by vacuum distillation at 77 K. SF6 (Matheson)
was purified by vacuum distillation at 77 K. CHClBr2
(Aldrich) was subjected to several freeze-pump-thaw
cycles to remove dissolved air.
HCCl + NO2 → HCN + ClO2
0
1
r H = � 128.4 kJ mol
Thermochemical data was taken from JANAF tables
Typical reaction conditions in the LIF experiments
were0.1TorrofCHClBr2 and0.005–0.06TorrofNOor
NO2. For the IR absorption experiments, typical condi-
tions were 0.05 Torr of CHClBr2, 0–1.0 of Torr NO, and
1.0 Torr of SF6 buffer gas. Based on an absorption co-
0
[
7], except for NCO, for which 1f H = 131.38 kJ
�
1
mol was used [8].
EXPERIMENTAL
�
1
� 1
efficient of 0.097 cm Torr for CHClBr2 at 193 nm,
13
� 3
The photolysis laser light was provided by an ex-
cimer laser (Lambda Physik COMPEX 200) operat-
ing at 193 nm. Typical photolysis pulse energies were
typical [HCCl]0 number densities of � 3 � 10 cm
were obtained (assuming a quantum yield of unity).
�
4 mJ. For HCCl detection, 602.53 nm probe light
� 0.1–0.5 mJ pulse ) was produced by a dye laser
RESULTS
�
1
(
(Continuum ND-6000) pumped by the second har-
Total Rate Constants
monic of an Nd:YAG laser (Continuum Surelite II-10).
The photolysis and probe beams were made collinear
using a dichroic mirror and copropagated down a resis-
tively heated Pyrex reaction cell. Fluorescence was de-
HCCl + NO. Laser-induced fluorescence signals at
6
1
02.53 nm with � 200 ns lifetimes were observed upon
93-nm photolysis of HCClBr2. Based on previous
spectroscopic data [3], we can confidently assign these
signals to the A(050) � X(000) transition of the HCCl
molecule. Figure 1 shows boxcar integrated signal am-
plitudes as a function of the delay time between ex-
cimer and dye laser pulses. As shown, the HCCl kinetic
signals were found to be well approximated by single
exponential decays over the range t = 10–120 ␮sec.
The decay rates were not significantly affected by
the addition of He, N2, CF4, or SF6 buffer gases. This
�
tected at 90 from the laser beams using an R508 photo-
multiplier tube. Longpass filters were used to suppress
detection of scattered laser light. The unamplified PMT
signal was recorded by a boxcar integrator (Stanford
Instruments Model 250) with a 500-ns delay and a
300-nsgatewidthandaveragedonapersonalcomputer.
Adigitaldelaygenerator(StanfordDG535)wasusedto
vary the delay between excimer and dye laser pulses in
order to produce a HCCl concentration vs. time profile.
Control experiments were performed to verify that flu-
orescence signals at 602.53 nm originated from HCCl.
Only upon photolysis of the precursor molecules were
such signals observed.
For product molecule detection, infrared absorption
spectroscopy was used as described in previous publi-
cations [9,10]. A tunable lead-salt diode laser provided
the infrared probe radiation. The excimer and probe
beams were made collinear with a dichroic mirror and
copropagated through a 1.46-m absorption cell. The
UV light was then removed with a second dichroic
mirror, and the infrared light passed through a 0.25-m
monochromator and was focused onto a 1-mm InSb de-
tector (� 1-␮sec response). Transient infrared absorp-
tion signals were collected on a LeCroy 9310A digital
oscilloscope and transferred to a computer for anal-
ysis. The HITRAN molecular infrared database [11]
was used as an aid in locating N2O absorption lines
and calibrating absolute yields. HCNO and HNCO ab-
sorption lines were found with the aid of published
infrared spectral data [12,13].
Figure 1 Laser-induced fluorescence signals of HCCl as
a function of excimer-dye delay time. Reaction conditions:
PHCClBr = 0.10 Torr, PNO = variable, T = 296 K.
2

1
4
BAREN, ERICKSON, AND HERSHBERGER
observation suggests that vibrational relaxation effects
are insignificant in these experiments. Either HCCl was
produced vibrationally cold, or vibrational relaxation
was accomplished within � 10 ␮sec even in the absence
of buffer gas (if this were not true, we would expect to
see slower rise times in Fig. 1). The magnitude of the
LIF signals was lower in the presence of these buffer
gases, presumably because of fluorescence quenching.
As a result, most of the experiments were performed
without any buffer gas.
Upon addition of nitric oxide to the reactant mixture,
an increase in the HCCl decay rates was observed. In
the presence of significant NO concentrations, pseudo-
first-order conditions apply and a standard kinetic anal-
ysis shows that
0
[
HCCl] = [HCCl] exp(� k t)
Figure 3 Arrhenius plot of the HCCl + NO rate constant.
0
0
k = k1[NO] + kD
0
Arrhenius plot of k over this range. The data are well
fit to an Arrhenius expression:
where k is the observed pseudo-first-order decay rate
constant, k1 is the desired bimolecular rate constant for
the HCCl + NO reaction, and kD represents the decay
rate in the absence of added NO. Contributions to kD in-
clude self-reaction, reaction with precursor molecules,
and diffusion of detected species out of the probed
volume. Although some of these processes are not
strictly first order, the signals were found to be reason-
ably fit to pseudo-first-order kinetics even at low [NO].
1
k1 = (2.225 � 0.05) � 10�
11
3
� 1 � 1
s
� exp[(62.05 � 6.9)/T ] cm molecule
where the error bars represent one standard deviation.
�
11
3
At 296 K, a value of k1 = (2.75 � 0.2) � 10
cm
�
1
� 1
s was obtained.
molecule
0
Figure 2 shows k as a function of [NO] at several dif-
ferent temperatures. The slope of these plots represents
the bimolecular rate constant k1.
HCCl + NO . The addition of NO2 to the reaction
mixture caused a significant increase in HCCl pseudo-
2
Experiments on reaction (1) were conducted over
the temperature range 298–572 K. Figure 3 shows an
first-order decay rates, as shown in Fig. 4. Experiments
Figure 2 Pseudo-first-order decay rates for HCCl as a
Figure 4 Pseudo-first-order decay rates for HCCl as a
function of NO pressure. Reaction conditions: PHCClBr =
function of NO2 pressure. Reaction conditions: PHCClBr =
2
2
0
.10 Torr, PNO = variable.
0.10 Torr, PNO = variable.

KINETICS OF HCCL + NOx REACTIONS
15
Figure 5 Arrhenius plot of the HCCl + NO2 rate constant.
Figure 6 Transient infrared absorption signals for the
products of the HCCl + NO and associated secondary re-
0
0
actions (see text). Lines probed: N2O (00 1) � (00 0)
�
1
0
0
over the temperature range 298–476 K were performed,
as shown in Fig. 5. The data were well fit to an Arrhe-
nius expression:
R(3) at 2227.039 cm , CO2 (00 1) � (00 0) R(8) at
�
1
0
0
0 0
2
2
355.890 cm , HCNO (0100 0 ) � (0000 0 ) P(14) at
�
1
184.608 cm . Reaction conditions: PHCClBr = 0.05 Torr,
2
PNO = 0.2 Torr, PSF = 1.0 Torr.
6
�
9
k2 = (1.775 � 0.27) � 10
3
� 1 � 1
�
exp[(� 879.112 � 57.3)/T ] cm molecule
s
reaction (1c) with nitric oxide:
�
10
cm3
At 296 K, a value of k2 = (1.10 � 0.2) � 10
NCO + NO → N2O + CO
NCO + NO → CO2 + N2
(3a)
(3b)
�
1
� 1
s was obtained.
molecule
This reaction has been studied in several laborato-
Product Branching Ratios
ries [9,17–20]. It has a rate constant of (3.2–3.4) �
�
11
3
� 1 � 1
The product branching ratio of reaction (1) was mea-
sured at 296 K using infrared absorption spectroscopy.
HCNO, N2O, and CO2 molecules were detected, as
shown in the transient absorption signals of Fig. 6. Al-
though the rise times for the various product molecules
differ, this is probably because of vibrational relax-
ation kinetics rather than reaction kinetics. Since we
probe only the ground vibrational state, the nascent vi-
brational distribution must be relaxed to a Boltzmann
distribution in order for the molecular number densi-
ties to be reliably measured. SF6 has previously been
shown to be an efficient relaxer of vibrational exci-
tation for N2O and CO2 [14–16], and probably also
relaxes HCNO as well. Under the experimental condi-
tions of 1.0 Torr of SF6, this relaxation still takes about
10
cm molecule s at 296 K [17–19], with a
moderate negative temperature dependence. The two
product channels are formed in roughly equal yield
[9,20], with �3a = 0.44 � 0.07 and �3b = 0.56 � 0.07
at 296 K. These branching ratios are virtually indepen-
dent of temperature over the range 296–623 K [9].
The absolute concentrations of the detected prod-
ucts were determined as follows. For N2O and CO2,
tabulated infrared absorption linestrengths are avail-
able [11], making absolute calibration straightforward,
using formulae described in earlier publications [9,10].
These linestrengths have been checked by perform-
ing absorption measurements on static N2O and CO2
gas, and found to be accurate to within 10% or better.
For calibration of the transient signals, peak amplitues
wereused, andDopplerlinewidthswereassumed(pres-
sure broadening is insignificant under the experimental
conditions used). For fulminic acid (HCNO), no pub-
lished linestrengths are available. HCNO, although not
very stable, is not a transient species, so in principle
absorption coefficients could be measured in a static
50 ␮sec for N2O and CO2, but is essentially complete
on a timescale short compared to the slow decay, which
is attributed to diffusion of molecules out of the probed
volume.
The N2O and CO2 products are believed to originate
from a secondary reaction of NCO radicals formed in

1
6
BAREN, ERICKSON, AND HERSHBERGER
cell. The synthesis of HCNO of known purity is
rather difficult, however. An alternative approach was
therefore used: Ketene (CH2CO) molecules were pho-
tolyzed at 193 nm. Although the dissociation of ketene
at this wavelength is somewhat complicated, two prod-
ucts dominate [21,22]:
CH2CO + h␯(193 nm) → CH2 + CO
(4a)
CH2CO + h␯(193 nm) → HCCO + H (4b)
Channel (4a) is the major product, with �4a = 0.81,
and �4b = 0.11 [22]. Most of the methylene radicals
are formed in the triplet ground state, and a small
(<0.1) yield of CCO radicals was also observed [22].
In our calibration, typical conditions were 70 mTorr
of CH2CO, 0.2–1.0 Torr of NO, and 1.0 Torr of SF6
buffer gas. Under these conditions, [NO] > [CH2] and
Figure 7 Product yields as a function of reagent NO
[
NO] > [HCCO], so that essentially all of the CH2 and
pressure. Reaction conditions: PHCClBr = 0.05 Torr, PNO =
2
HCCO radicals react with NO:
variable, PSF6 = 1.0 Torr.
CH2 + NO → HCNO + H
(5a)
(5b)
(6a)
(6b)
peak amplitude of the HCNO transient infrared signal
was measured upon photolysis of a CH2CO/NO/SF6
mixture. Comparison of this signal to the calculated
CH2 + NO → HCN + OH
HCCO + NO → HCNO + CO
HCCO + NO → HCN + CO2
[HCNO]providedthecalibrationfactorwhichwasthen
applied to the signals obtained from the title reaction.
Figure 7 shows the obtained number densities of
N2O, CO2, and HCNO as a function of reagent NO
pressure. Several conclusions can be obtained from
this result. The product yields quickly level off at
PNO > 0.2 Torr, indicating that at these NO pressures,
essentially every HCCl radical produced in the disso-
ciation reacts with NO. The ratio of [HCNO] in the
high pressure limit of Fig. 7 to the initially produced
The product branching ratios of these reactions have
been previously studied. Several measurements of �6a
have been reported, with values of �6a = 0.88 � 0.04
[
[
23] and 0.64 � 0.12 [24] at 296 K, and 0.77 � 0.09
25] at 700 K. Only one group has reported a measure-
ment of �5a, with a value of 0.84 at 296 K [26]. What
makes the calibration reliable is that fact the both of
these reactions produce HCNO in similar yields. Thus
every CH2CO molecule that is dissociated will produce
approximately 0.8 HCNO molecules, essentially inde-
pendent of uncertainties in the photolysis yields �4a and
[HCCl]0 concentration may therefore be taken as an
accurate estimate of the branching ratio �1a. Further-
more, the fact that the observed N2O and CO2 yields are
roughly equivalent provides evidence that these prod-
ucts do indeed originate from secondary reactions of
NCO radicals, reaction (3), which produces N2O and
CO2 in similar yields. The product branching ratio for
NCO formation, �1c, can therefore be determined by
�
4b. The number density of [CH2]0 + [HCCO]0 radi-
cals produced in the ketene photolysis was obtained by
measuring the 193-nm absorption coefficient of ketene
sample (the values obtained varied substantially be-
tween different ketene samples because of different
purity levels; CO2 impurity levels of � 30% are typ-
ical in our synthesized ketene samples) as well as the
absolute excimer pulse energy incident upon the reac-
tion cell, and assuming a total dissociation quantum
yield of unity. The number density of HCNO was then
determined by
�1c = [N2O]/(0.44 � [HCCl]0)
where we have assumed that every NCO radical cre-
ated in reaction (1) reacts with NO. Using the above
techniques, we obtain the following branching ra-
tios for reaction (1) at 296 K: �1a = 0.68 � 0.06 and
�
1c = 0.24 � 0.04, where the error bars represent one
[
HCNO] = ([CH ]0 + [HCCO] ) � 0.84
standard deviation.
2
0
Several possible product channels were not detected
in this experiment. Attempts to detect HNCO transient
signals were unsuccessful. We estimate �1b < 0.1. No
where we used the value of �5a, since channel (4a) is the
dominant pathway for ketene dissociation. The peak–

KINETICS OF HCCL + NOx REACTIONS
17
attempts were made to detect the products of channels
1d) or (1e). Our branching ratios for (1a) and (1c)
sum to near unity, however, suggesting that any other
channels present must be quite minor.
HCNO + Cl is the major channel, with a smaller con-
tribution from NCO + HCl.
(
BIBLIOGRAPHY
1
2
3
. Miller, J. A.; Bowman, C. T. Prog Energy Combust Sci
989, 15, 287.
. McEnally, C. S.; Sawyer, R. F.; Koshland, C. P.; Lucas,
D. Symp (Int) Combust, [Proc] 1994, 25, 325.
DISCUSSION
1
Our value for the HCCl + NO rate constant at 296 K is
somewhat higher than that of the one previous study
of this reaction, which reported k1 = (1.5 � 0.5) �
. Kakimoto, M.; Saito, S.; Hirota, E. J Mol Spectrosc
1
983, 97, 194.
�
11
3
� 1 � 1
10
cm molecule s [5]. We do not have any ex-
4. Chang, R.-C.; Fei, R.; Sears, T. J. J Mol Spectrosc 1997,
183, 341.
5. Wagener, R.; Wagner, H. Gg. Z Phys Chem 1992,
planation for the discrepancy, except to note that HCCl
fluorescence signals are quite weak and careful filtering
is required to eliminate background signals from scat-
tered laser light. Our work represents the first reported
value of k2 as well as the branching ratios of reaction
1
75, 9.
6
7
. Carstensen, H.-H.; Rehbein, C.; Wagner, H. Gg. Ber
Bunsen-Ges Phys Chem 1997, 101, 1429.
. Chase, M. W., Jr.; Davies, C. A.; Downey, J. R., Jr.;
Frurip, D. J.; McDonald, R. A.; Syverud, A. N. J Phys
Chem Ref Data 1985, 14 (Suppl 1).
(1). Although neither the NCO nor HCl product from
channel (1c) was directly detected, we are quite con-
fident about the existence of this channel because the
NCO + NO reaction is the only reasonable source of
the detected N2O product molecules in this system. In
principle both of these products could be detected by
IR absorption, but their infrared linestrengths are sub-
stantially smaller than those of N2O, so that our indirect
method is the more sensitive approach.
In the product channel measurements, we assumed
that the CHClBr2 photolysis produced HCCl radicals
with a quantum yield of unity. No previous experi-
mental evidence for this assumption exists, however
our results strongly suggest that this assumption is ap-
proximately correct. If the quantum yield were much
less than one, then the sum of our observed branch-
ing ratios would also be less than one. Our value of
8
9
. East, A. L. L.; Allen, W. D. J Chem Phys 1993, 99, 4638.
. Cooper, W. F.; Park, J.; Hershberger, J. F. J Phys Chem
1
993, 97, 3283.
10. Quandt, R. W.; Hershberger, J. F. J Phys Chem 1996,
100, 9407.
11. Rothman, L. S.; Gamache, R. R.; Tipping, R. H.;
Rinsland, C. P.; Smith, M. A. H.; Benner, D. C.;
Devi, V. M.; Flaud, J.-M.; Camy-Peyret, C.; Perrin, A.;
Goldman, A.; Massie, S. T.; Brown, L. R.; Toth, R. A.
J Quant Spectrosc Radiat Transfer 1992, 48, 469.
2. Ferretti, E. L; Rao, K. N. J Mol Spectrosc 1974, 51, 97.
3. Steiner, D. A.; Polo, S. R.; McCubbin, T. K., Jr.; Wishah,
K. A. J Mol Spectrosc 1983, 98, 453.
1
1
1
4. Fakhr, A.; Bates, R. D., Jr. Chem Phys Lett 1980, 71,
3
81.
15. Stephenson, J. C.; Moore, C. B. J Chem Phys 1970, 52,
2333.
16. Cooper, W. F.; Hershberger, J. F. J Phys Chem 1992, 96,
�
1a + �1c = 0.92 is close to unity, within experimental
uncertainties, however.
7
71.
Formation of the major channel (1a) can proceed
by a straightforward mechanism, involving formation
of an HC(Cl)NO intermediate followed by C Cl bond
cleavage. A possible mechanism for NCO formation
in channel (1c) is formation of a three-member CNO
ring, followed by H atom migration to the Cl atom,
ring opening, and subsequent C Cl fission. Another
less likely possibility, which we cannot directly probe,
would be formation of the high energy isomer CNO
rather then NCO.
1
1
1
7. Perry, R. A. J Chem Phys 1985, 82, 5485.
8. Atakan, B.; Wolfrum, J. Chem Phys Lett 1991, 178, 157.
9. Juang, D. Y.; Lee, J.-S.; Wang, N. S. Int J Chem Kinet
1
995, 27, 1111.
2
2
0. Becker, K. H.; Kurtenbach, R.; Wiesen, P. Chem Phys
Lett 1992, 198, 424.
1. Unfried, K. G.; Glass, C. P.; Curl, R. F. Chem Phys Lett
1
991, 177, 33.
22. Glass, G. P.; Kumaran, S. S.; Michael, J. V. J Phys Chem
A 2000, 104, 8360.
2
2
2
3. Rim, K. T.; Hershberger, J. F. J Phys Chem A 2000, 104,
93.
4. Eickhoff, U.; Temps, F. Phys Chem Chem Phys 1999, 1,
43.
5. Boullart, W.; Nguyen, M. T.; Peeters, J. J Phys Chem
994, 98, 8036.
2
CONCLUSIONS
2
The reactions of HCCl with NO and NO2 are fast with
only a very slight temperature dependence. Analysis
of products of the HCCl + NO reaction indicates that
1
26. Grussdorf, J.; Temps, F.; Wagner, H. Gg. Ber Bunsen-
Ges Phys Chem 1997, 101, 134.