FTIR study of the products of the reaction between HCCO and NO

Article abstract of DOI:10.1039/a807258b

The main pathways of the reaction HCCO + NO → products (1) were investigated at room temperature in the gas phase in Ar at 570 mbar pressure by employing FTIR spectroscopy. Mixtures of NO₂-C₂H₂-NO-Ar were photolysed under stationary conditions using a high-pressure Hg lamp at λ = 312 nm. HCCO was generated in the reaction system in the reactions NO₂ + hv → O(³p) + NO, O(³P) + C₂H₂ → H + HCCO. The main products of reaction (1) include CO₂, CO, and HCNO. The results can be rationalised with the reaction channels HCCO + NO → HCN + CO₂ (1a) and HCCO + NO → HCNO + CO (1d). The branching ratios were determined to be k(1a)/k₁ = (0.28 ± 0.10) and k(1d)/k₁ = (0.64 ± 0.12). Formation of HNCO could not be confirmed. Other species detected include H₂O and HONO. These are formed, together with additional CO, in consecutive reactions of H atoms and OH radicals.

Full text of DOI:10.1039/a807258b

View Article Online / Journal Homepage / Table of Contents for this issue
FTIR study of the products of the reaction between HCCO and NO
U. Eickho† and F. Temps*
Institut f ur Physikalische Chemie, Universitat Kiel, Olshausenstr. 40, D-24098 Kiel, Germany.
Ž Ž
E-mail: temps=phc.uni-kiel.de
Recei¿ed 17th September 1998, Accepted 16th No¿ember 1998
The main pathways of the reaction HCCO ] NO ] products (1) were investigated at room temperature in the
gas phase in Ar at 570 mbar pressure by employing FTIR spectroscopy. Mixtures of NO ÈC H ÈNOÈAr were
2
2 2
photolysed under stationary conditions using a high-pressure Hg lamp at j \ 312 nm. HCCO was generated
in the reaction system in the reactions NO ] hl ] O(3P) ] NO, O(3P) ] C H ] H ] HCCO. The main
2
2 2
products of reaction (1) include CO , CO, and HCNO. The results can be rationalised with the reaction
2
channels HCCO ] NO ] HCN ] CO (1a) and HCCO ] NO ] HCNO ] CO (1d). The branching ratios
2
were determined to be k /k \ (0.28 ^ 0.10) and k /k \ (0.64 ^ 0.12). Formation of HNCO could not be
1a 1
1d 1
conÐrmed. Other species detected include H O and HONO. These are formed, together with additional CO, in
2
consecutive reactions of H atoms and OH radicals.
and product analysis by FTIR spectroscopy. The HCCO rad-
icals were produced in the reaction system via the reaction
sequence:
1
Introduction
Elucidation of the mechanisms of NO reburning processes in
Ñames is of great interest because of their importance for
reducing the emission of NO from hydrocarbon combustion
NO ] hl ÈÈÈ
Õ
NO ] O
(2)
2
x
(
see e.g. ref. 1È4 and references therein). However, rather little
O ] C H ÈÈÈ
Õ
H ] HCCO
(3a)
is known about many of the elementary chemical reactions
2 2
which may participate.
The reaction
The paper is organized as follows: Following this Intro-
duction, Section 2 summarizes the experimental details.
Section 3 describes the experimental results which were
obtained. This section is divided into three subsections: The
Ðrst reports the results of the qualitative and quantitative
product measurements, the second describes the reaction
mechanism which provides a basis for rationalizing the data,
and the third presents the results for the branching ratio of the
reaction. Section 4 contains a discussion of the experimental
data, numerical simulations of the reactions system, compari-
sons with other experimental and theoretical investigations,
and a brief discussion of implications of the measurements for
the NO reburning mechanism. The main conclusions are
gathered in Section 5.
HCCO ] NO ] products
(1)
is assumed to play a key role in NO reburning, especially in
fuel-rich Ñames (see e.g. ref. 5 and 6). Reaction (1) is known to
be rather fast under room-temperature conditions (k \ 1.8
1
]
1013 cm3 mol~1 s~1).7h9 However, the product distribution
and the detailed mechanism are not well understood.9,10 In
particular, there are many thermodynamically allowed
product channels,11h14 including:
HCCO ] NO
ÈÈÈ
ÈÈÈ
ÈÈÈ
ÈÈÈ
ÈÈÈ
Õ HCN ] CO , * H¡ \ [516 kJ mol~1 (1a)
2
r 298
Õ
HNCO ] CO, * H¡ \ [487 kJ mol~1 (1b)
r 298
2 Experimental
The title reaction was studied in a 21.3 cm long, 4.6 cm id,
Duran glass reaction cell sealed with NaCl windows. The
set-up was similar to that described in our study of the reac-
Õ
HOCN ] CO, * H¡ \ [383 kJ mol~1 (1c)
r 298
Õ
HCNO ] CO, * H¡ \ [188 kJ mol~1 (1d)
r 298
Õ
HONC ] CO, * H¡ \ [135 kJ mol~1 (1e)
r 298
tion of CH with NO.18 The cell was situated in the sample
2
`
M
chamber of an FTIR spectrometer (IFS 66V, Bruker) and con-
ÈÈÈ
Õ
HC(NO)CO,
* H¡ \ [167 kJ mol~1 (1f)
r 298
nected to an all-glass vacuum system for gas handling, evac-
The reaction is generally assumed to proceed via complex for-
mation. HC(NO)CO, nitrosoketene, can be formed at high
pressures as a collisionally stabilized recombination product.
On the other hand, in the low-pressure regime, the initially
formed complex can dissociate via several exit channels. The
yields of HCN vs. the di†erent possible isomeric ““HCONÏÏ
species and the branching ratios between these isomers are of
uated by
a
turbomolecular pump. Gas mixtures,
NO ÈC H ÈNOÈAr, were prepared manometrically in a
2
2 2
storage bulb. Care was taken to allow sufficient time for
mixing. The mixtures were then transferred to the reaction
cell, where they were photolysed with light from a 200 W
high-pressure Hg lamp (Oriel). The Hg light spectrum was
narrowed by using a 300È340 nm, 45¡ reÑective dielectric
mirror as a bandpass Ðlter. The resulting photolysis spectrum
was dominated by the strong 312 nm Hg line. The Hg light
was collimated and imaged into the reaction cell with the
dielectric mirror in order to irradiate the entire reaction
volume as evenly as possible.
considerable relevance for modelling NO reburning.15h17
x
Here, we report results of a recent study of reaction (1)
which allowed us to identify the main product channels of the
reaction and to determine their branching ratios. The investi-
gation involved photolysis of mixtures of NO ÈC H ÈNOÈAr
2
2 2
Phys. Chem. Chem. Phys., 1999, 1, 243È251
243

View Article Online
FTIR spectra were taken with 0.25 cm~1 nominal
resolution by co-adding 100 scans. A weak NortonÈBeer apo-
dization function was used and a zero-Ðlling factor of 8.
Strong IR absorption lines were chosen for the quantitative
determinations of the di†erent molecules of interest, in order
to achieve high detection sensitivities with minimal inter-
ferences from other species in the reaction mixtures. The
absorption line proÐles were integrated using the standard
OPUS software supplied with the spectrometer. Reference
spectra of calibrated gas mixtures of the pure substances in Ar
were used for converting the integrated IR signals to corre-
sponding partial pressures. Gas mixtures for the calibrations
were prepared systematically for each species, with concentra-
tions in ranges similar to those found in the reaction to take
into account deviations from the LambertÈBeer law at the
spectrometer resolution used.
were taken up to 240 min after the photolysis in order to
check for possible dark reactions which could deplete reaction
products. No sign of a dark reaction was observed before the
start of the photolysis, even after storing a reactant mixture
for up to 24 h.
A typical FTIR ““productÏÏ spectrum is depicted in Fig. 1.
This was obtained with a reaction mixture of 0.147 mbar
NO , 29.61 mbar C H , and 4.93 mbar NO in Ar at 569.1
2
2 2
mbar total pressure (mixture 11, see Table 1) after a photolysis
time of 60 min. A spectrum taken immediately before the pho-
tolysis was subtracted from the measured 60 min spectrum to
obtain the di†erence spectrum shown. The di†erent molecules
were identiÐed with the help of reference spectra, depicted in
Fig. 1 for comparison underneath the product spectrum. All
molecules which are depleted are shown with negative signals
All gases used were of the highest commercially available
purities: Ar (99.9999%), C H (99.6%), NO (98%), NO
2
2
2
(
99.5%), CO (99.997%), and CO (99.995%). Ar, CO, and CO
2
2
were taken as supplied. C H , NO , and NO were carefully
2
2
2
puriÐed until no impurities could be detected in their FTIR
spectra. C H , NO , and NO were condensed in traps at 77
2
2
2
K and puriÐed until no impurities could be detected in their
FTIR spectra. C H and NO were condensed in traps at 77 K
and puriÐed by fractional distillation. The Ðrst fraction of the
C H , which evaporated upon slowly warming, was free of
2
2
2
2
acetone; the Ðrst tenth of the evaporated NO was free of NO
and N O. HCNO for measuring FTIR reference spectra was
2
synthesised by pyrolysis of 3-phenyl-4-oxoiminoisoxazol-5-
(
4H)-one.19 Reference samples of HCN and HNCO were pre-
pared by heating mixtures of NaCN or NaNCO with stearic
acid. HONO was obtained as a 1 : 1 mixture with NH
equilibrium over solid NH NO .20
3 in
4
2
3
Results
3.1 Reaction products
The investigation of reaction (1) began with qualitative and
quantitative measurements of the products of the photolysis of
a series of di†erent NO ÈC H ÈNOÈAr gas mixtures in the
2
2 2
reaction cell. The partial pressures of the gases were adjusted
to maximize the e†ect of reaction (1) on the observable reac-
tion products. All mixtures contained ca. 0.15 mbar NO and
2
ca. 30 mbar C H . The NO was systematically varied from
2
2
0
O p /mbar O 20. A constant total pressure of ca. 570 mbar
Fig. 1 FTIR spectra. Top trace: Measured di†erence spectrum
NO
(
““product spectrumÏÏ) for mixture 11 (Table 1) after 60 min photolysis.
was reached by adding Ar. The exact mixture compositions
are listed in Table 1. The total photolysis times were 60 min.
FTIR spectra were taken before (reaction time ““zeroÏÏ) and,
with the Hg light blocked for the duration of the FTIR scans,
after 5, 10, 20, 40, and 60 min irradiation. Further spectra
A spectrum taken immediately before the photolysis was subtracted
from the measured 60 min spectrum to obtain the di†erence spectrum
shown. Species which are depleted are shown with negative signals,
while products exhibit positive signals. Below: Reference spectra of
the di†erent pure gases for comparison.
Table 1 Experimental conditions, product yields (60 min photolysis), and branching ratios C
NO ÈC H ÈNOÈAr mixtures. The quoted average values were derived from mixtures 5È13.
CO2 , CCO , and CHCNO
for the di†erent
2
2 2
Mixture
number
p
/
p
/
p
/
p
/
[*p
/
p
/
p
/
p
/
total
NO2
C2H2
NO
NO2
CO2
CO
HCNO
mbar
mbar
mbar
mbar
mbar
mbar
mbar
mbar
C
CCO
CHCNO
CO2
1
2
3
4
572.5
569.5
564.9
575.6
0.146
0.151
0.149
0.148
30.56
29.98
29.78
30.03
0.00
0.45
0.92
1.48
0.1339
0.1136
0.1040
0.0954
0.0158
0.0108
0.0122
0.0110
0.0322
0.0340
0.0360
0.0331
0.0131
0.0213
0.0243
0.0240
0.29
0.24
0.29
0.28
0.35
0.50
0.61
0.61
È
0.26
0.37
0.41
5
6
7
8
9
0
1
2
3
560.9
571.2
564.5
579.5
572.8
571.6
569.1
572.8
572.7
0.151
0.153
0.153
0.153
0.152
0.151
0.147
0.154
0.148
29.70
30.12
30.39
29.87
30.53
30.00
29.61
30.48
30.47
1.86
2.24
2.45
2.50
3.07
3.96
4.93
10.17
19.95
0.0934
0.0881
0.0958
0.0853
0.0835
0.0751
0.0770
0.0525
0.0352
0.0117
0.0099
0.0108
0.0100
0.0095
0.0095
0.0120
0.0078
0.0053
0.0323
0.0320
0.0334
0.0319
0.0301
0.0278
0.0314
0.0238
0.0136
0.0218
0.0233
0.0251
0.0233
0.0221
0.0205
0.0212
0.0180
0.0137
0.31
0.28
0.28
0.29
0.28
0.31
0.38
0.35
0.34
0.61
0.65
0.61
0.68
0.63
0.66
0.74
0.82
0.62
0.37
0.44
0.43
0.47
0.44
0.45
0.45
0.60
0.66
1
1
1
1
Av.
570.0
0.150
30.00
0È20
È
È
È
È
0.31 ^ 0.10
0.67 ^ 0.12
0.48`0.25
0.12
~
244
Phys. Chem. Chem. Phys., 1999, 1, 243È251

View Article Online
(
NO and NO), while products exhibit positive signals. The
2
spectra were integrated and converted to partial pressures and
concentrations to obtain reactant and product time proÐles.
The main reaction products were CO , CO, and HCNO.
2
These products are seen in Fig. 2 which shows an expanded
part (from 2400 to 2000 cm~1) of the spectrum of Fig. 1. The
CO and CO absorptions are immediately obvious. The pro-
2
duction of HCNO could be unambiguously seen from the
observed characteristic l , l ] l [ l , and l ] l [ l IR
2
2
5
5
2
4
4
band systems at 2125È2230 cm~1.21 The HCNO absorption
lines overlapped with the somewhat stronger CO R branch
lines in the same region, but single rovibrational lines could be
found to determine both species. Despite its strong IR absorp-
tion around 2275 cm~1, a region without other interfering
absorptions, HNCO could not be detected. Thus, HNCO
does not appear to be a product of reaction (1). HCN, the
assumed co-product of the observed CO from reaction (1a),
2
was also not detected in the FTIR spectra. However, the HCN
absorption band at 1400 cm~1 is relatively weak, while the
bands around 725 and 3300 cm~1 were obscured by the
strong C H absorptions. C H was always present in very
2
2
2 2
large excess; its pressure was too high to allow analysis of the
C H depletion because of the small change during the reac-
tion.
2
2
In addition to the above-mentioned main products, HONO
and H O were observed (cf. Fig. 1). However, these two mol-
2
ecules are not of major interest in the following. Very weak
additional FTIR absorption signals were observed at around
1
775 and 1795 cm~1 and still remain to be assigned; the
respective molecules can be only very minor products.
Measured reactant and product time proÐles for a selected
measurement (mixture 10; 0.151 mbar NO , 30.0 mbar C H ,
Fig. 3 Measured reactant and product partial pressures as a func-
tion of photolysis time (mixture 10, Table 1). Filled symbols show the
experimental data, full lines refer to the results of the numerical simu-
lations (see Discussion).
2
2 2
3
3
.96 mbar NO, 571.6 mbar total pressure) are plotted in Fig.
. The NO depletion reaches ca. 50% after 60 min photolysis.
2
Correspondingly, the products CO , CO, and HCNO as well
as HONO and H O increased with time, as expected. The
C H and NO, which were present in large excess, remained
2 2
2
practically constant. Analogous plots were obtained for the
2
other reaction mixtures.
Comparing the data from all experimental runs, the total
NO consumption was found to decrease from ca. 90% after
2
6
0 min photolysis in the measurement without added NO to
ca. 25% at p \ 20 mbar. The production of CO and CO
NO
2
was seen to decrease correspondingly. Despite the drop in the
total NO consumption, the absolute yield of HCNO was
2
found to exhibit (at Ðrst) a pronounced increase upon addition
of increasing NO, before decreasing again due to the smaller
total NO consumption. A maximum of HCNO formation
2
was observed in the measurement with p \ 2.45 mbar. This
NO
trend shows the importance of reaction (1d) for the product
formation.
After completion of the photolysis, the NO , NO, CO ,
2
2
and CO remained constant. Measurements of the HCNO
after photolysis showed a weak decay under the conditions in
the reaction cell, due to a slow (presumably heterogeneous)
dark reaction. However, an e†ect on the branching ratio for
reaction (1) can be eliminated by considering the initial
product formation rates at short photolysis times (see below).
A slow decay was also seen for HONO, while H O, being a
2
product of the HCNO and HONO decomposition steps, was
seen to increase slowly.
The measured absolute NO depletions and the corre-
2
sponding CO , CO, and HCNO product yields after 60 min
2
photolysis are compiled in Table 1. In addition, Table 2 lists
the initial NO depletion rates and the corresponding initial
2
product formation rates determined from the observed time
proÐles towards short reactions times.
Fig. 2 Part of Fig. 1 (mixture 11) on an expanded scale, showing the
3.2 Reaction mechanism
formation of the reaction products CO , CO, and HCNO. The
2
The reaction mechanism which provides a basis for rational-
ising the observed product yields consists of (i) the NO pho-
tolysis system, (ii) the reactions involving HCCO radicals,
product spectrum is given by the top trace, di†erent reference spectra
are shown underneath. Note the absence of HCNO in the product
spectrum.
2
Phys. Chem. Chem. Phys., 1999, 1, 243È251
245

View Article Online
Table 2 Initial product formation rates and resulting branching ratios C
mixtures. The quoted average values were derived from mixtures 5È13.
CO2 , CCO , and CHCNO
determined for the di†erent NO ÈC H ÈNOÈAr
2
2
2
Mixture
p
/
p
/
p
/
p
/
[r0
/
r0
/
r0 /
r0
/
total
NO2
C2H2
NO
NO2
CO2
CO
HCNO
number mbar mbar mbar mbar mbar min~1 mbar min~1 mbar min~1 mbar min~1
C
CCO
CHCNO
CO2
1
2
3
4
572.5 0.146 30.56
569.5 0.151 29.98
564.9 0.149 29.78
575.6 0.148 30.03
0.00 0.008 37
0.45 0.003 50
0.92 0.003 03
1.48 0.002 50
0.000 90
0.000 30
0.000 33
0.000 29
0.001 43
0.000 93
0.000 97
0.000 90
0.000 50
0.000 60
0.000 73
0.000 77
0.27
0.21
0.27
0.29
0.17
0.41
0.55
0.64
È
0.22
0.39
0.55
5
6
7
8
9
0
1
2
3
560.9 0.151 29.70
571.2 0.153 30.12
564.5 0.153 30.39
579.5 0.153 29.87
572.8 0.152 30.53
571.6 0.151 30.00
569.1 0.147 29.61
1.86 0.002 43
2.24 0.002 38
2.45 0.002 55
2.50 0.002 32
3.07 0.002 07
3.96 0.001 87
4.93 0.001 72
0.000 29
0.000 22
0.000 26
0.000 22
0.000 17
0.000 19
0.000 23
0.000 14
0.000 09
0.000 90
0.000 83
0.000 83
0.000 77
0.000 77
0.000 63
0.000 69
0.000 53
0.000 25
0.000 79
0.000 71
0.000 75
0.000 80
0.000 68
0.000 65
0.000 62
0.000 45
0.000 28
0.30
0.23
0.25
0.24
0.20
0.25
0.32
0.27
0.29
0.67
0.61
0.55
0.57
0.66
0.57
0.72
0.77
0.55
0.59
0.53
0.51
0.64
0.59
0.64
0.66
0.65
0.69
1
1
1
1
572.8 0.154 30.48 10.17 0.001 23
572.7 0.148 30.47 19.95 0.000 70
Av.
570.0 0.150 30.00 0È20
È
È
È
È
0.26 ^ 0.10 0.63 ^ 0.10 0.61 ^ 0.14
including their formation in the reaction O ] C H and their
300È340 nm wavelength region leads to O(3P) atoms,
2
2
subsequent depletion, especially via reaction (1), (iii) reactions
NO ] hl ÈÈÈ
Õ
NO ] O(3P)
(2)
of CH radicals which are formed in the second channel of the
2
2
O ] C H reaction, and (iv) reactions of H atoms and OH
with a quantum yield of unity.12,22,23 However, recombi-
nation of the O atoms with NO and their reaction with NO
have to be taken into account,
2
2
radicals and additional consecutive reactions. The di†erent
elementary reactions and the rate constants taken into
account are compiled in Table 3. The main reactants for the
various radical species are NO , NO, and C H , which are
present in the reaction mixtures in much higher concentra-
tions than all other species.
2
O ] NO ] M ÈÈÈ
Õ
NO ] M
(4)
(5a)
(5b)
2
2 2
2
O ] NO ÈÈÈ
Õ
Õ
NO ] O
2
2
O ] NO ] M ÈÈÈ
NO ] M
3
.2.1 Photolysis of NO . The photolysis of NO has been
2
3
2
2
studied in considerable detail in a number of laboratories (see
e.g. ref. 12 and 22È24). The photodissociation of NO2 in the
Reactions (4) and (5b) are in the low-pressure regime under
the conditions used. The respective rate constants are well
Table 3 Elementary reactions in the detailed reaction mechanism
Reaction
k/cm3 mol~1 s~1 ahc
Ref.
HCCO ] NO ] HCN ] CO (1a)
5.0 ] 1012
1.3 ] 1013
This work
This work
See Text
2
HCCO ] NO ] HCNO ] CO (1d)
NO ] hl ] NO ] O (2)
1.3 ] 10~4 a,c
6.4 ] 1010
2
O ] C H ] H ] HCCO (3a)
26È29
2
2
O ] C H ] 3CH ] CO (3b)
1.6 ] 1010
26È29
2
2
2
O ] NO ] M ] NO ] M (4)
5.5 ] 1011 b
5.8 ] 1012
12, 25
2
O ] NO ] NO ] O (5a)
12
12
12
12
12
12
12
12
12
12
8
8
2
2
O ] NO ] M ] NO ] M (5b)
7.4 ] 1011 b
6.3 ] 1011 b
4.2 ] 10~2 a
1.6 ] 1013
2
3
NO ] NO ] M ] N O ] M (6)
3
2
2 5
N O ] M ] NO ] NO ] M ([6)
2
5
3
2
NO ] NO ] NO ] NO (7)
3
2
2
NO ] O ] NO ] O (8)
1.0 ] 1013
8.3 ] 109 b
6.0 ] 104 a
1.9 ] 109 b
1.5 ] 105 a
8.0 ] 1012 b,d
8.0 ] 1012 b,d
O6 ] 109
3
2
2
NO ] NO ] M ] N O ] M (9)
2
2
2 4
N O ] M ] NO ] NO ] M ([9)
2
4
2
2
NO ] NO ] M ] N O ] M (10)
2
2 3
N O ] M ] NO ] NO ] M ([10)
2
3
2
HCCO ] NO ] HCNO ] CO (11a)
2
2
HCCO ] NO ] CO ] other products (11b)
2
HCCO ] C H ] products (12)
8
2
2
HCNO ] products (13)
1.0 ] 10~4 a
1.9 ] 1013
See text
18, 30
18, 30
30
31
29, 32
28
28
12
12
3
CH ] NO ] HCNO ] H (14a)
2
3
CH ] NO ] HCN ] OH (14b)
3.3 ] 1012
2
3
CH ] NO ] products (15)
5.9 ] 1013
1.7 ] 108
2
2
3
CH ] C H ] products (16)
2
2 2
H ] NO ] M ] HNO ] M (17)
3.2 ] 1011 b
8.0 ] 1013
1.0 ] 1011 b
3.6 ] 1012 b
3.5 ] 1012 b
4.7 ] 1011 b
5.0 ] 1012
5.0 ] 1012
H ] NO ] OH ] NO (18)
2
H ] C H ] M ] C H ] M (19)
2
2
2 3
OH ] NO ] M ] HONO ] M (20)
OH ] NO ] M ] HNO ] M (21)
2
3
OH ] C H ] M ] C H OH ] M (22)
12, 28, 29
Estimated
Estimated
2
2
2 2
C H OH ] NO ] products (23)
2
2
C H OH ] NO ] products (24)
2
3
2
a First-order rate constant in s~1. b Second-order rate constant in low pressure or fall-o† regime in cm3 mol~1 s~1 (M \ Ar, p \ 570 mbar).
c Fitted e†ective Ðrst-order rate constant describing the NO photolysis. d Branching ratio adjusted to describe the measured formation of
2
HCNO without added NO.
246
Phys. Chem. Chem. Phys., 1999, 1, 243È251

View Article Online
known (ref. 12, 24 and 25; see Table 3). Other consecutive
reactions, e.g.
The slow HCNO decay reaction observed after photolysis
could be formally described by a Ðrst-order reaction
NO ] NO ] M E8F N O ] M
(6)
(7)
HCNO ÈÈÈ
Õ
products
(13)
3
2
2 5
NO ] NO ÈÈÈ
Õ
NO ] NO
with an e†ective rate coefficient of k \ (5.3 ^ 1.0) ] 10~5
s~1 for all experimental runs.
3
2
2
13
NO ] O ÈÈÈ
Õ
NO ] O
(8)
3
2
2
NO ] NO ] M E8F N O ] M
(9)
3.2.3 Reactions of CH . Additional reactions which have
2
2
2 4
2
to be taken into account arise because of the formation of
NO ] NO ] M E8F N O ] M
(10)
2
2 3
3CH radicals in reaction (3b). The fast reaction
2
cannot have a signiÐcant e†ect on the observed Ðnal product
distribution.12,23,24 The small equilibrium amount (O1% of
the NO at the highest NO) of N O in the NO ÈNO mix-
3
CH ] NO ÈÈÈ
Õ
HCNO ] H
(14a)
(14b)
2
ÈÈÈ
Õ HCN ] OH
2
2 3
2
tures can be taken into account.
contributes to the production of HCNO (see ref. 18 and 30).
However, the branching ratios for the two channels are
known (k /k \ 0.84 and k /k \ 0.15).18 The reaction
3
.2.2 Formation and depletion reactions of HCCO. The
HCCO radicals which are the subject of this work are produc-
ed in the reaction
1
4a 14
14b 14
3
CH ] NO ÈÈÈ
Õ
products
(15)
2
2
O ] C H ÈÈÈ
Õ
HCCO ] H
(3a)
(3b)
again can be important only at very low NO,30 while the reac-
2
2
tion
ÈÈÈ
Õ
CH ] CO
2
3
CH ] C H ÈÈÈ
2 2
Õ
products
(16)
The branching ratios for (3a) and (3b) are known from mea-
2
surements of the HCCO channel by Peeters and co-workers26
is very slow31 at room temperature.
and from a determination of the CH channel from our
The production of HCNO via reaction (14a) in addition to
reaction (1d) has to be taken into account to determine the
branching ratio of reaction (1).
2
laboratory27 to be k /k \ 0.8 and k /k \ 0.2 (each with
3a
3
3b 3
estimated uncertainties of ^0.1; k \ 8.0 ] 1010 cm3 mol~1
3
s~1 28,29). Hence, reaction (3a), i.e. production of HCCO, con-
stitutes the main pathway, while CH formation is only of
2
3.2.4 Reactions of H atoms and OH radicals. Further
attention has to be directed to additional radical species
which are produced in the reaction system. In particular, reac-
tion (3a) leads to a corresponding amount of H atoms. Under
the conditions used, the H atoms undergo consecutive reac-
tions with NO, NO , and C H :
secondary importance. CH is generated in its X
3
3B
1 ground
2
electronic state (43CH ).
2
The fraction of the O atoms leading to HCCO in the
present work is given by k [C H ]/(k [C H ] ] k [NO]
3
a
2 2
3
2 2
4
]
k [NO ]). Thus, under the conditions used, only between
5
2
5
9% (at p \ 0) and 13% (p \ 20 mbar) of the O atoms
2
2 2
NO
NO
from the NO photolysis produce HCCO. A rather large frac-
H ] NO ] M ÈÈÈ
Õ
HNO ] M
OH ] NO
(17)
(18)
(19)
2
tion of the
O
atoms (k [NO]/(k [C H ] ] k [NO]
4
3
2 2
4
H ] NO ÈÈÈ
Õ
Õ
]
k [NO ]); up to ca. 76% at p \ 20 mbar, the highest
2
2
5
2
NO
p
\ 20 mbar, the highest p ) reacts via recombination with
H ] C H ] M ÈÈÈ
C H ] M
NO
NO
2 2
2 3
NO [reaction (4)]. However, reaction (4) is e†ectively a ““zero
The respective reaction rate constants are well known (ref. 28,
cycleÏÏ reaction because it leads back to the NO . The reac-
tion is only responsible for the observed reduction of the NO
consumption in the experiments with increasing p . Reaction
5) plays only a small role in the O atom balance. Hence, reac-
tions (2) and (3a) constitute a suitable source for HCCO for
the present purpose.
The subsequent HCCO depletion in the reaction system in
the presence of a large excess of NO is dominated by the title
2
2
9 and 32; see Table 3). With the rate constants as listed, the
fractions of the H atoms reacting via the di†erent routes
2
NO
Mk [NO]/(k [NO] ] k [NO ] ] k [C H ]) etc.N under
(
1
7
17
18
2
19 2 2
the experimental conditions with p in the range 0È20 mbar
NO
are 0È30%, 78È55%, and 22È15%, respectively, for the reac-
tions with NO, NO , and C H .
2
2 2
The HNO from reaction (17) can undergo reactions with
itself33 or with NO . Likewise, the C H radicals from reac-
reaction (k \ 1.8 ] 1013 cm3 mol~1 s~1).7h9 From the
2
2 3
1
tion (19) are likely to react with NO, either to give C H
experimentally observed formation of CO , CO, and HCNO,
2
2
2
]
HNO or via recombination to H CCHNO (or
H CCNOH, after subsequent isomerization). However, these
consecutive reactions cannot a†ect the observed product dis-
tribution for reaction (1). The implications of the formation of
OH radicals in reaction (18), on the other hand, need to be
examined in a little more detail. The OH radicals undergo
corresponding consecutive reactions with NO, NO
C H :
the reaction channels (1a) and (1d) are seen to constitute the
2
main pathways,
2
HCCO ] NO ÈÈÈ
ÈÈÈ
Õ
HCN ] CO
(1a)
(1d)
2
Õ
HCNO ] CO
Other HCCO reactions cannot compete with reaction (1)
under all experimental conditions except at very low NO,
where the reaction
2 and
2
2
OH ] NO ] M ÈÈÈ
Õ
HONO ] M
(20)
(21)
(22)
HCCO ] NO ÈÈÈ
Õ
products
(11)
2
OH ] NO ] M ÈÈÈ
Õ
Õ
HNO ] M
2
3
(
with k \ 1.6 ] 1013 cm3 mol~1 s~1)8 contributes to the
11
OH ] C H M ÈÈÈ
C H OH ] M
product formation. It is worth noticing here that the product
measurements in the absence of added NO indicate that reac-
tion (11) can, at least in part, also produce HCNO. The reac-
tion
2 2
2 2
Reactions (20)È(22) are in their fall-o† regimes under the con-
ditions of interest. Taking the known rate constants (ref. 12, 28
and 29; see Table 3), the fractions of the OH reacting via the
three di†erent routes under the experimental conditions
HCCO ] C H ÈÈÈ
Õ
products
(12)
2
2
(p \ 0È20 mbar) are 0È83%, 4È1%, and 96È16% with NO,
NO , and C H , respectively. Thus, while HONO has indeed
NO
(
with k O 6 ] 109 cm3 mol~1 s~1)8 is too slow to play a
12
2
2 2
role.
been measured, HNO formation can be neglected. On the
3
Phys. Chem. Chem. Phys., 1999, 1, 243È251
247

View Article Online
r
\ d[CO]/dt \ k [NO][HCCO] ] k [C H ][O] (Ic)
CO
1d
3b 2 2
r
\ d[HCNO]/dt \ k [NO][HCCO]
HCNO
1d
]
k
[NO][3CH ] (Id)
14a
2
where k I [NO ] is the e†ective NO photolysis rate under
the experimental conditions. With the steady-state approx-
2
P
2
2
imations for [O], [H], [HCCO], and [3CH ], and with the
2
abbreviations
k [C H ]
A \
B \
3a 2 2
k [C H ] ] k [NO ]
3
2 2
5a
2
k
[NO ]
18
18
2
2
k
[NO] ] k [NO ] ] k [C H ]
1
7
19 2 2
and
A
] AB
C \ 1
the branching ratios k /k and k /k
1d 1 are related to the
1
a 1
experimental product yields according to
d[CO ]
A
k
k
B
B
2
\ C 1a
(IIa)
(IIb)
(IIc)
[
d[NO ]
2
1
d[CO]
A
A
k
k
k
k
\
C
1d ] 3b
[
d[NO ]
2
1
3a
d[HCNO]
k
k
B
k
k
k
k
Fig. 4 Measured yields of CO , CO, and HCNO from reaction (1)
1
d ] 3b 14a
2
\ C
as a function of the partial pressures of NO in the reaction mixtures.
[d[NO ]
2
1
3a 14
Here, the factors C \ A/(1 ] AB) in eqn. (IIa)È(IIc) simply
account for the extra NO depletion routes via the reactions
other hand, it is seen that reaction (22) is the main OH deple-
tion route at low NO, but it has only a small e†ect at high
NO. This conclusion is important, since the C H OH
complex from reaction (22) might react further with NO.
Plausible reaction channels could include pathways
2
O ] NO (5a) and H ] NO (14), while the extra terms on
2
2
the righthand sides in eqn. (IIb) and (IIc) allow for the addi-
2
2
tional CO and HCNO production routes O ] C H ] 3CH
2 2
2
]
CO (3b) and 3CH ] NO ] HCNO ] H (14a).
2
Eqn. (IIa)È(IIc) are readily solved and evaluated. Thus,
taking the experimental data as listed in Table 1, branching
C H OH ] NO ÈÈÈ
Õ
HCNO ] H CO
(23a)
(23b)
2
2
2
ratios C , C , and C
(where, ideally, C
is identiÐed
`M
CO2 CO
HCNO
CO2
ÈÈÈ
Õ
other products
with C and C , C
with C ) were found for the pro-
1
a
CO HCNO
1d
duction of CO , CO, and HCNO, respectively. The data for
A recombination reaction (23b) would lead to stable molecules
e.g. HOCHxCHNO, HOCHxCNOH, OCHwCHNOH,
c-C H NO ), but reaction (23a) would be an unwanted addi-
2
the di†erent reaction mixtures are given in Table 1. Taking
(
the points from the experimental runs with p P 1.5 mbar,
NO
2
3
2
the average values from the observed Ðnal product yields were
tional HCNO source. As seen above, this extra source does
not play an important role at high NO. Furthermore, it is
found to be:
C
\ 0.31 ^ 0.10,
C
\ 0.67 ^ 0.12 and
CO2
CO
C
\ 0.48`0.25 . The error limits include the statistical and
pointed out that H CO, which is the second product of reac-
HCNO
~0.12
2
estimated systematical uncertainties.
tion (23a), could not be found, despite a detection sensitivity
Considering the slow HCNO decay reaction (13) observed
comparable to that for HCNO. Thus, reaction (23a) was
neglected. A reaction according to
after completion of the photolysis, the above C
HCNO
must, in fact, be considered as a lower limit to C . This is
value
1
d
C H OH ] NO ÈÈÈ
Õ
C H O ] HONO
(24a)
(24b)
reÑected by the fact that C
\ C above. Obviously, reac-
2
2
2
2 2
HCNO
CO
tion (13) has to be taken into account. This was accomplished
`M
ÈÈÈ
Õ
other products
by analysis of the initial NO depletion rates and the respec-
2
tive initial product formation rates determined at short reac-
was taken into account to explain the observed HONO in the
measurements without added NO.
tion times. The initial rates which were determined from the
di†erent experimental time proÐles (e.g. Fig. 3) have been col-
lected in Table 2. The individual measurements were evalu-
ated by using eqn. (IIa)È(IIc) above. The results, which are
listed in Table 2, are plotted as a function of the NO pressures
in the reaction mixtures in Fig. 4. The product branching
ratios from the initial reaction rates including estimated sys-
tematical errors were found to be
3.3 Branching ratios for reactions (1a) and (1d)
The experimental data show that reaction (1) proceeds via
the channels HCCO ] NO ] HCN ] CO
(1a) and
2
HCCO ] NO ] HCNO ] CO (1d). It is the task, in the fol-
lowing, to determine the branching ratios C \ k /k and
1
a
1a 1
C1
d
\ k /k for these two pathways.
1d 1
C
\ 0.26 ^ 0.10
\ 0.63 ^ 0.12
\ 0.61 ^ 0.16
Considering the above mechanism, at high NO, the reaction
system is dominated by reactions (1)È(5a), (14), and (17)È(19).
Under these conditions, NO depletion and CO , CO, and
CO2
C
CO
2
2
C
HCNO production can be described with the rate equations
HCNO
4
Discussion
[
r
\ [d[NO ]/dt \ k I [NO ]
2 P
NO2
2
2
4
.1 Branching ratios for reactions (1a) and (1d)
[
k [NO][O] ] k [NO ][H] (Ia)
4
18
2
According to the present measurements, the reaction between
HCCO and NO has two main channels, i.e.
r
\ d[CO ]/dt \ k [NO][HCCO]
1a
(Ib)
CO2
48
2
2
Phys. Chem. Chem. Phys., 1999, 1, 243È251

View Article Online
The importance of the individual elementary steps in the
full reaction mechanism was further examined with a sensi-
tivity analysis. The results of the sensitivity analysis for
mixture 10 (time proÐles for Fig. 3) are depicted in Fig. 5.
Histograms of the reduced sensitivity coefficients L ln c /L ln k
HCCO ] NO ÈÈÈ
ÈÈÈ
Õ
HCN ] CO
(1a)
(1d)
2
Õ
HCNO ] CO
The production of CO, HCNO, and CO was directly
observed in the FTIR spectra. HCN was not detected because
of the presence of high amounts of C H , but it is expected to
be produced in conjunction with the CO in reaction (1a).
2
i
j
for HCNO, CO, and CO for the main reaction steps are
2
shown. As seen, the absolute yields of these products are most
2
2
2
strongly a†ected by the NO photolysis rate (2), the reaction
2
The branching ratios for reactions (1a) and (1d) could be
O ] C H (3), and the reaction O ] NO ] M (4), as well as
2
2
determined from the Ðnal product yields after completion of
photolysis and from the initial product formation rates deter-
mined at short reaction times. The respective results are found
to be in good agreement, if allowance is made for the slow
heterogeneous decay of HCNO, which cannot be avoided.
the title reaction (1). The sensitivity coefficients for the three
products with respect to reactions (2)È(5a) are very similar,
which means that these reactions are of high importance only
for the absolute product yields, not for the relative product
amounts. The relative yields of CO and CO and HCNO are
2
The Ðnal values for C and C
agree with their experi-
seen to be determined by the ratio of the rate constants
CO
HCNO
mental accuracies. This observation is very important, since
one could imagine other sources for CO in the reaction
system, whereas the possibilities for production of HCNO are
k /k
a 1d .
1
4.3 Comparison with previous work
very limited. The reaction 3CH ] NO (14a) gives only a
2
small contribution, which was taken into account, and an
The overall rate constant for reaction (1) has been determined
in three previous investigations. The reaction was Ðrst mea-
sured by direct means with detection of HCCO by IR diode
laser absorption spectroscopy by Unfried et al.7 HCCO rad-
icals were generated in their study by photodissociation of
e†ect of a reaction like C H OH ] NO ] HCNO ] H CO
2
2
2
(
23a) could be ruled out by the absence of H CO.
2
Taking the average of the two C
ratio for channel (1a) at room temperature and pressures
around 570 mbar is found to be k /k \ (0.28 ^ 0.10). Like-
wise, taking the average of the two C values and the C
values, the branching
CO2
CH CO with an excimer laser at j \ 193 nm. From the expo-
1a
1
CO
2
nential HCCO decay curves at di†erent NO concentrations, a
HCNO
value from the initial HCNO rise, the branching ratio for
rate constant value of k \ 2.3 ] 1013 cm3 mol~1 s~1 was
1
channel (1d) for the same conditions is k /k \ (0.64 ^ 0.12).
found. The second investigation, by Temps et al.8 employed
1
d 1
According to eqn. (IIb) and (IIc), this value is somewhat
dependent on the branching ratio for reactions (3a) and (3b).
The above value was derived by adopting k /k \ 0.8.26,27 A
the discharge Ñow technique with HCCO detection by laser
magnetic resonance spectroscopy. The radical source was
reaction (3), as in the present work. The measured rate con-
3
a 3
smaller value for k /k would lead to a smaller result for
stant value of k \ 1.3 ] 1013 cm3 mol~1 s~1 was somewhat
3a
3
1
k /k and a correspondingly higher value for k /k . The
lower than that of Unfried et al.7 In any case, however, the
1d 1
1a 1
quoted uncertainty limits refer to the 95% conÐdence level
and include estimated systematic errors.
reaction was conÐrmed to be very fast. In the third study,
Boullart et al.9 investigated the C H ÈOÈNO system using
2
2
The above branching ratios add up to (k ] k )/k \ 0.92.
the discharge Ñow-molecular beam mass spectrometry tech-
nique. These authors derived a room temperature value of
1a
1a
1
Thus, within the experimental accuracy (^0.10), reactions (1a)
and (1d) can account almost fully for the reaction. Some room
k \ 1.8 ] 1013 cm3 mol~1 s~1 by following the change in
1
(
O0.08) is still left for small contributions from other reaction
steady-state mass signals of HCCO upon adding increasing
channels, including collisional stabilization of the
a
amounts of NO. Further measurements up to 700 K did give
indications for a small, but statistically signiÐcant, non-zero
activation energy.9
HCCO ] NO association complex. Furthermore, one has to
note that formation, via reaction (1e), of HONC, which would
rapidly isomerize to HCNO, may not be excluded from the
The product branching ratio of reaction (1) has, to the best
of our knowledge, been the subject of only one previous
experimental study. In their mass spectrometric work cited
above,9 Boullart et al. examined the formation of CO and
experimental data (it would be included in C
). However,
HCNO
HOCN formation appears less likely considering the available
information on the potential-energy hypersurface for the reac-
tion (see below and ref. 10).
CO produced in channels (1a)È(1e) at 700 K and 2 mbar.
2
4.2 Simulations of the reaction system and sensitivity analysis
Numerical simulations of the reaction system were carried out
in order to check the data analysis and investigate the e†ects
of the di†erent elementary steps which can play a role. The
elementary reactions and rate constants taken into account
are listed in Table 3. Fig. 3 includes the results of the simula-
tions for one experimental run (p \ 3.96 mbar; mixture 10,
NO
cf. Table 1). As seen, the experimental NO , CO , CO, and
2
2
HCNO proÐles are described very well. Similar agreement
was observed for the di†erent other reaction mixtures.
Also shown in Fig. 3 are simulated proÐles obtained with k
1
(
and k ) set to zero (dashed curve). Setting k \ 0, the pro-
1
1
1
duction of CO is turned o†. Likewise, reactions (3b) and (14a)
2
would remain as the sole sources for CO and HCNO, respec-
tively. Neither reaction can account for the observed products.
Thus, the calculated proÐles give a clear demonstration of the
importance of reactions (1a) and (1d) for the observed pro-
ducts. The simulations were less satisfactory for HONO and
Fig. 5 Sensitivity analysis for reaction mixture 10 (Table 1) after 40
min photolysis. The histograms show the reduced sensitivity coeffi-
cients L ln c /L ln k for the products HCNO, CO, and CO with
H O, even though the trends in these proÐles were predicted.
2
However, these species did not enter into the analysis. They
i
j
2
respect to variations of the rate constants of the main reactions in the
mechanism of Table 3. Reactions omitted were found to have negligi-
ble sensitivity.
are consecutive products which could also be a†ected by het-
erogeneous reactions.
Phys. Chem. Chem. Phys., 1999, 1, 243È251
249

View Article Online
They also found mass signals at m/z \ 43, consistent with a
tion between HCCO and NO (1), as well as the reactions of
CH , CH , and CH with NO, are of fundamental importance
because they lead to recycling of the NO back into the reac-
tion mechanism. These reactions have found technical impor-
tance in the so-called reburn stages in large-scale power
generation (e.g. ref. 35).
Reaction (1) has been proposed as the main NO reburning
reaction under some fuel-rich conditions.5,6 The detailed NO
reduction routes, and thus all NO reburn models, are,
however, highly critically a†ected by branching between the
product with an ““HCONÏÏ sum formula, which they attrib-
3
2
uted to HNCO. From the observed CO /CO ratio, and the
2
alleged HNCO detection, reaction (1) was concluded to
proceed according to channels (1a) and (1b) with k /k \ 0.23
1
a 1
and k /k \ 0.77, respectively.
1
b 1
In an accompanying ab initio quantum chemical study,
Nguyen et al.10 derived stationary points and transition states
on the HCCO ] NO potential-energy surface. Assuming the
reaction to proceed via an HC(NO)CO association complex
(
nitrosoketene), an exit channel was indeed found for the for-
CO ] HCN and the HCNO ] CO channels (1a) and (1d).
2
mation of HCN ] CO , following isomerization of the
HC(NO)CO complex via a four-centre cyclic transition state.
The conversion of NO to HCN is a key step in the NO reburn
2
mechanism.5,36,37 HCN is subsequently converted to NH and
i
However, the mechanism for the alleged production of HNCO
remained very unclear. On the other hand, CO bond rupture
in the HC(NO)CO complex to form HCNO ] CO was found
to be feasible. We mention here that di†erent isomeric
m/z \ 43 species, in particular HNCO and HCNO, cannot
easily be distinguished in mass spectrometric work such as
that of Boullart et al.9 A reassignment of their m/z \ 43 mass
signal to HCNO would bring their data into excellent agree-
ment with the present results.
N
under rich conditions.1,3 The HCNO could isomerize to
2
HNCO which would again lead to NH (HNCO has been sug-
gested as an additive for selective NO reduction according to
i
the Raprenox process38). On the other hand, the recent model
calculations by Miller et al. suggest that fast bimolecular reac-
tions of HCNO with OH and O lead to rapid regeneration of
the NO.17
A number of research groups have been involved in model-
ling laboratory-scale plug Ñow reactor measurements of NO
reburning (e.g. ref. 6, 15È17 and 39). Comprehensive mecha-
nisms, as proposed especially by Glarborg et al.15 (GADM),
Bockhorn et al.37 (DCWW), the Gas Research Institute40
(GRI), and Dean and Bozelli41 (DB), have formed the basis for
the modelling. Critical importance was attributed to the for-
mation of HCN, which was used to assess the di†erent
models. In a comparative study, Schelb et al.39 observed that
the HCN proÐles under rich conditions were described by the
GADM mechanism. However, the GADM mechanism
adopted reaction (1a) as a major NO reburn step. Further-
more, Glarborg and Miller15h17 were unsuccessful in model-
ling similar plug Ñow reactor NO reburn measurements of
HCN assuming HCNO ] CO as the main products of
HCCO ] NO. It appears that the formation of HCN in NO
reburning still requires further investigation.
Miller et al.17 very recently reported modelling calculations
of the product distribution of reaction (1). Taking the ab initio
potential-energy data for the system of Nguyen et al.,10 they
employed unimolecular rate theory to predict HCN ] CO to
2
be the major products (ca. 80%) at room temperature.
HCNO ] CO was calculated to account for the remaining ca.
2
HCN] CO was calculated to decrease and the production
of HCNO ] CO was found to increase. Other product chan-
nels were concluded to be negligible. These model calculations
contrast with the experimental results, both of the present
work and of Boullart et al.,9 which show the dominance of the
HCNO ] CO channel. The discrepancy may be traced back
to the calculated ca. 50 kJ mol~1 energy di†erence between
the rate-determining transition states for the two channels of
interest on the potential surface of Nguyen et al.10 According
to Nguyen et al., the HCN ] CO channel is the energetically
favoured one. It remains to be seen to what extent the calcu-
lated product branching ratio for reaction (1) is a†ected by
small errors in the calculated transition state energies10 and
existing small deÐciencies in the unimolecular rate theory
model calculations of ref. 17.
0% of the reaction. With increasing temperature, the yield of
2
Extrapolation of our data to higher temperatures depends
on the tightness and on the relative energies of the transition
states. On the one hand, reaction (1a) is thought to have a
tighter transition state than reaction (1d) which would favour
2
HCNO ] CO over HCN ] CO as the dominating higher-
2
temperature products. On the other hand, HCN ] CO
2
would be expected to increase compared to HCNO ] CO if
Preparation of nitrosoketene via pyrolysis of isonitroso
MeldrumÏs acid was reported by Matsui et al.34 Their identiÐ-
cation was based on FTIR spectra attributed to the
HC(NO)CO molecule, on the grounds of quantum chemical
IR frequency calculations. The assignment was supported by
mass spectrometric observations using chemical ionization of
an m/z \ 72 species, considered as HC(NO)CO É H`, and
detection of a nitrosoketene dimer. Considering, however, the
calculated low potential-energy barriers for the HC(NO)CO
isomerization and decomposition reactions of Nguyen et al.,10
the observation of a stable HC(NO)CO appears rather ques-
tionable. Several more stable isomers of HC(NO)CO exist10
which may exhibit similar IR transitions.
We note that Matsui et al. did Ðnd a thermal decomposi-
tion of their product to HCNO ] CO, which would be in
agreement with our product observations.34 On the other
hand, they did not detect HCN. Depending on the energy bar-
riers, the thermal decomposition of HC(NO)CO may yield
products in di†erent ratios than those from the energized
HCCO ] NO adduct.
the energy barrier for reaction (1a) was higher than that for
reaction (1d). At typical reburn temperatures, given the avail-
able data, both channels are likely to contribute. Further com-
putational work on reaction (1), in which we will address the
question of the temperature dependence by unimolecular rate
theory, is currently underway (see also ref. 17). In any case, the
title reaction is seen to have substantial e†ect on the NO to
HCN conversion rate in the NO reburn process.
x
It becomes obvious that major questions regarding the
detailed elementary reaction steps in the NO reburn mecha-
nism still remain to be answered.
5
Conclusions
In the present work, the reaction HCCO ] NO was found to
proceed according to two major channels,
HCCO ] NO ] HCN ] CO
(1a)
(1d)
2
]
HCNO ] CO
From FTIR measurements of the CO , CO, and HCNO, the
2
branching ratios are k /k \ (0.28 ^ 0.10) and k /k \ (0.64
4
.4 Implications for NO reburning
1a 1
1d 1
x
^ 0.12) for room temperature and up to ca. 1 bar pressure.
The advantage of the FTIR technique used is its unique
versatility, this was instrumental for the detection of HCNO,
but this versatility was bought at the expense of time
NO is generated as a pollutant in the combustion of hydro-
x
carbons according to the well known Zeldovich, Fenimore,
fuel-N, and N O mechanisms (see e.g. ref. 35). The fast reac-
2
250
Phys. Chem. Chem. Phys., 1999, 1, 243È251

View Article Online
1
1
2
8
9
0
J. Grudorf, F. Temps and H. Gg. Wagner, Ber. Bunsen-Ges.
Phys. Chem., 1997, 101, 134.
resolution. Time-resolved direct measurements of reaction (1)
would, therefore, be of great interest to support the present
Ðndings. With the recent detection of HCCO with laser
induced Ñuorescence,42 and with the advent of stable, low-cost
single-mode tunable diode lasers in the near-IR, real-time
experiments with sensitive direct detection of HCNO and
HCN do indeed appear feasible. HCNO is known to exhibit a
very rich near-IR spectrum.43 Furthermore, interest arises in
the unimolecular isomerization of HCNO to HNCO and in
the kinetics of bimolecular reactions of HCNO with small
radicals like OH, O, or H.
C. Wentrup, B. Gerecht and H. Briehl, Angew. Chem., 1979, 91,
5
03; G. Schulze, Diploma Thesis, Gieen, 1993.
Z. Vecera and P. N. Dasgupta, Environ. Sci. T echnol., 1991, 25,
255; P. Pagsberg, E. Ratajczak, A. Sillesen and Z. Latajka, Chem.
Phys. L ett., 1994, 227, 6.
21
22
23
E. L. Feretti and K. N. Rao, J. Mol. Spectrosc., 1974, 51, 97.
A. B. Harker and H. J. Johnston, J. Phys. Chem., 1973, 77, 1153.
H. Gaedke and J. Troe, Ber. Bunsen-Ges. Phys. Chem., 1975, 79,
1
84.
2
4
J. Hahn, Diplomas Thesis, G oŽ ttingen, 1998.
25 J. V. Michael, W. A. Payne and D. A. Whytock, J. Chem. Phys.,
1
976, 65, 4830; G. Yarwood, J. W. Sutherland, M. A. Wick-
ramaaratchi and R. B. Klemm, J. Phys. Chem., 1991, 95, 8771.
J. Peeters, M. Schaekers and C. Vinckier, J. Phys. Chem., 1986,
9
9
Kinet., 1994, 26, 869; J. V. Michael and A. F. Wagner, J. Phys.
Chem., 1990, 94, 2453.
Acknowledgements
2
6
0, 6552; W. Boullart and J. Peeters, J. Phys. Chem., 1992, 96,
810; J. Peeters, W. Boullart and I. Langhans, Int. J. Chem.
This work has been supported by the Deutsche Forschungsge-
meinschaft within SFB 357 (Molekulare Mechanismen
Unimolekularer Prozesse) and by the Fonds der Chemischen
Industrie. We thank Dr. J. Gripp for the preparation of the
HCNO reference sample.
27 U. Bley, M. Koch, C. Oehlers, F. Temps and H. Gg. Wagner,
2
3rd International Symposium on Combustion, Abstract P-105,
Orleans, 1990.
2
8
D. L. Baulch, C. J. Cobos, R. A. Cox, C. Esser, P. Frank, Th. Just,
J. A. Kerr, M. J. Pilling, J. Troe, R. W. Walker and J. Warnatz, J.
Phys. Chem. Ref. Data, 1992, 21, 411; D. L. Baulch, C. J. Cobos,
R. A. Cox, P. Frank, G. Hayman, Th. Just, J. A. Kerr, T. Mur-
rells, M. J. Pilling, J. Troe, R. W. Walker and J. Warnatz, J. Phys.
Chem. Ref. Data, 1994, 23, 847.
References
1
2
3
4
J. O. L. Wendt, C. V. Sternling and M. A. Matovich, 14th Inter-
national Symposium on Combustion, The Combustion Institute,
Pittsburgh, 1972, p. 897.
S. L. Chen, J. M. McCarthy, W. D. Clark, M. P. Heap, W. R.
Seeker and D. W. Pershing, 21st International Symposium on
Combustion, The Combustion Institute, Pittsburgh, 1988, p. 1159.
T. Kolb, P. Jansohn and W. Leuckel, 22nd International Sympo-
sium on Combustion, The Combustion Institute, Pittsburgh, 1989,
p. 1193.
2
3
9
0
W. G. Mallard, F. Westley, J. T. Herron, R. F. Hampson and D.
H. Frizzell, NIST Chemical Kinetics Database, Ver. 5.0, NIST
Standard Reference Data, Gaithersburg, MD, 1993.
V. Seidler, F. Temps, H. Gg. Wagner and M. Wolf, J. Phys.
Chem., 1989, 93, 1070.
3
3
1
2
T. Bo
J. J. Ahumada, J. V. Michael and D. T. Osborne, J. Chem. Phys.,
972, 57, 3736.
Ž hland, F. Temps and H. Gg. Wagner, ref. 2, p. 841.
T. Etzkorn, S. Muris, J. Wolfrum, C. Dembny, H. Bockhorn, P.
F. Nelson, A. Attia-Shahin and J. Warnatz, 24th International
Symposium on Combustion, The Combustion Institute, Pittsburgh,
1
3
3
A. B. Callear and R. W. Carr, J. Chem. Soc., Faraday T rans. 2,
1
992, p. 925.
J. A. Miller and C. T. Bowman, Progr. Energy Sci. Combust.,
989, 15, 287.
1
975, 71, 1603; Y. He and M. C. Lin, Int. J. Chem. Kinet., 1992,
5
6
7
8
9
0
1
2
24, 743.
1
3
3
3
4
5
6
H. Matsui, E. J. Zu
Ž
ckerman, N. Katagiri, C. Kaneko, S. Ham and
D. Stapf and W. Leuckel, 26th International Symposium on Com-
bustion, The Combustion Institute, Pitsburgh, 1996, p. 2083.
K. G. Unfried, G. P. Glass and R. F. Curl, Chem. Phys. L ett.,
D. M. Birney, J. Phys. Chem., 1997, 101, 3936.
J. Warnatz, U. Maas and R. W. Dibble, Combustion, Springer,
Berlin, 1998.
P. Glarborg and S. Hadvig, Development and T est of a Kinetic
Model for Natural Gas Combustion, Nordic Gas Technology
Center, 1991.
H. Bockhorn, Ch. Chevalier, J. Warnatz and V. Weyrauch, V DI-
Berichte, 1991, 922, 171.
R. A. Perry and D. L. Siebers, Nature (L ondon), 1986, 342, 657.
D. Schelb, M. Braun-Unkho†, P. Frank, F. Lecomte, M. Cathon-
net and Ph. Dagaut, 27th International Symposium on Com-
bustion, Poster W2B17, Boulder, CO, 1998.
1
991, 177, 33.
F. Temps, H. Gg. Wagner and M. Wolf, Z. Phys. Chem., 1992,
76, 27.
W. Boullart, M. T. Nguyen and J. Peeters, J. Phys. Chem., 1994,
8, 8036.
M. T. Nguyen, W. Boullart and J. Peeters, J. Phys. Chem., 1994,
8, 8030.
Thermodynamic data from ref. 12 and 13, except for
1
3
7
9
1
1
1
3
3
8
9
9
*
H¡ (HCCO) \ ]177 kJ mol~1 from ref. 14.
a
298
R. Atkinson, D. L. Baulch, R. A. Cox, R. F. Hampson Jr., J. A.
4
0
C. T. Bowman, R. K. Hanson, D. F. Davidson, W. C. Gardiner,
V. Lisianski Jr., G. P. Smith, D. M. Golden, M. Frenklach and
M. Goldenberg, GRI-MECH 2.11, http://www.berkeley.me.edu/
Kerr, M. J. Rossi and J. Troe, J. Phys. Chem. Ref. Data, 1997, 26,
5
21.
1
1
1
3
4
5
C. F. Melius, BAC-MP4 Heats of Formation and Free Energies,
gri mech, 1996.
–
Sandia National Laboratories, Livermore, 1990.
4
4
4
1
2
3
A. M. Dean and J. W. Bozelli, in Combustion Chemistry II, ed. W.
D. L. Osborn, D. H. Mordaut, H. Choi, R. T. Bise, D. M.
Neumark and C. M. RohlÐng, J. Chem. Phys., 1997, 106, 10087.
P. Glarborg, M. U. Alzueta, K. Dam-Johansen and J. A. Miller,
Combust. Flame, 1998, in press.
C. Gardiner Jr., Springer, Berlin, 1998, in press.
L. R. Brock, B. Mischler, E. A. RohlÐng, R. T. Bise and D. M.
Neumark, J. Chem. Phys., 1997, 107, 665.
S. Albert, M. Winnewisser and B. P. Winnewisser, Ber. Bunsen-
Ges. Phys. Chem., 1997, 101, 1165.
1
1
6
7
L. Prada and J. A. Miller, Combust. Sci. T ech., 1998, in press.
J. A. Miller, J. L. Durant and P. Glarborg, 27th International
Symposium on Combustion. The Combustion Institute, Pittsburgh,
1
998, in press.
Paper 8/07258B
Phys. Chem. Chem. Phys., 1999, 1, 243È251
251