Infrared emission from the CF3 + NO2 reaction

Article abstract of DOI:10.1021/jp962655l

The reaction of the CF₃ radical with NO₂ has been investigated by using time-resolved FTIR emission spectroscopy. Strong infrared emission has been attributed to products, CF₂O and FNO, excited in the ν₂ and ν₁ modes, respectively. The direct one-step production pathway 1a is suggested as a major reaction channel: CF₃ + NO₂ → CF₂O + FNO, ΔH°₂₉₈ = -267 kJ mol^-1 (1a); → CF₂O + F + NO, ΔH°₂₉₈ = -31 kJ mol^-1 (1b). The rate constant for reaction 1 was measured to be (2.4 ± 0.5) × 10^-11 cm³ molecule^-1 s^-1. The F atom formation pathway 1b is the minor channel: the relative branching ratio of reactions 1a and 1b was estimated as 1 : 0.015. The formation of the FON isomer formed via a five-center intermediate is discussed as a possible carrier of an unidentified emission band near 1880 cm^-1. The CF₃O + NO reaction was investigated by the same technique. Lower emission intensities from the same reaction products were observed, with proportionally less vibrational excitation in the CF₂O product.

Full text of DOI:10.1021/jp962655l

2634
J. Phys. Chem. A 1997, 101, 2634-2642
Infrared Emission from the CF3 + NO2 Reaction
K. W. Oum and G. Hancock*
Physical and Theoretical Chemistry Laboratory, Oxford UniVersity, South Parks Road, Oxford OX1 3QZ, U.K.
ReceiVed: August 30, 1996; In Final Form: January 10, 1997X
The reaction of the CF3 radical with NO2 has been investigated by using time-resolved FTIR emission
spectroscopy. Strong infrared emission has been attributed to products, CF2O and FNO, excited in the ν2
and ν1 modes, respectively. The direct one-step production pathway 1a is suggested as a major reaction
channel: CF3 + NO2 f CF2O + FNO, ∆H°298 ) -267 kJ mol-1 (1a); f CF2O + F + NO, ∆H°298 ) -31
kJ mol-1 (1b). The rate constant for reaction 1 was measured to be (2.4 ( 0.5) × 10-11 cm3 molecule-1 s-1.
The F atom formation pathway 1b is the minor channel: the relative branching ratio of reactions 1a and 1b
was estimated as 1 : 0.015. The formation of the FON isomer formed via a five-center intermediate is
discussed as a possible carrier of an unidentified emission band near 1880 cm-1. The CF3O + NO reaction
was investigated by the same technique. Lower emission intensities from the same reaction products were
observed, with proportionally less vibrational excitation in the CF2O product.
Introduction
CF3 + NO2 f CF3O + NO
CF3O f CF2O + F
(1c)
(2)
The role of the CF3 radical in stratospheric chemistry has
been the subject of much research, both experimental1 and
theoretical.2 The species is also of importance in the field of
plasma chemistry, particularly in etching processes, in which
free radical reactions are initiated by the dissociation of
halocarbon precursor molecules, and in the pyrolysis of
fluorocarbon polymer materials exposed to high temperature.3
Absolute rate coefficients and reaction mechanisms for these
elementary halocarbon radical reactions are therefore required
in the modeling of these environments.
Kinetic and molecular dynamic information has been reported
for the elementary reactions between CF3 and the atomic species,
O,4 N,4 H,5 F,6 and I,7 and a variety of molecules, such as O2,1,8
NO,9 NO2,10-15 and the self-recombination reaction of the CF3
radical has also been investigated.16 In particular, the mecha-
nisms and kinetics of the reactions of NO, NO2, O3, and HO2
with either the CF3 radical or the related CF3O and CF3O2
species have been emphasized by several authors because of
their possible roles in the chemistry of the stratosphere.13,17-24
Interest has been focused on the various CF3/CF3O + NOx
reactions, since these can reduce NOx in the stratosphere, but
substantial uncertainties remain in the nitration mechanisms. For
example, although the CF2O molecule has been identified as a
major end product of the reaction between CF3 and NO2,
estimates of the quantum yield of CF2O have been conflicting,
as the contribution of the addition reaction to form CF3NO2 (or
CF3ONO) is uncertain (see refs 14, 15, and 25). Furthermore,
the role of the CF3O radical in the CF3 + NO2 reaction is not
clear.
F + CF3I f CF3 + FI
(3)
In a later series of experiments, Sugawara et al. applied time-
resolved diode laser spectroscopy to follow directly the decay
of the CF3 radicals in the CF3 + NO2 reaction.11 The rate
coefficient obtained, (2.5 ( 0.3) × 10-11 cm3 molecule-1 s-1
at 300 K, was, however, a factor of 10 greater than that measured
by Rossi et al. One possible explanation advanced for the
discrepancy between the two measurements involved the known
tendency of NO2 to dissolve in Teflon and thus to affect the
concentration assumed in the VLPP technique.10 If such an
explanation were correct, it would suggest that the rate coef-
ficient measured by Sugawara et al. is the more reliable.
Furthermore, Sugawara et al. observed absorption lines for
transitions of the ν1 band of nitrosyl fluoride, FNO, at very early
reaction times (∼15 µs), an observation which prompted the
authors to propose mechanism 1a involving a four-centered
transition state:
CF3 + NO2 f CF2O + FNO
(1a)
A further investigation of the CF3 + NO2 reaction was
performed by Francisco and Li, using IRMPD of CF3I in the
presence of NO2.12 Strong but spectrally unresolved IR emission
was observed and assigned to CF2O, and the rate constant for
the reaction was taken to be that for the rising portion of this
emission, yielding a value identical to that of Sugawara et.al.11
Francisco and Li suggested that the major products of the CF3
+ NO2 reaction were CF2O‡ and vibrationally cold FNO, despite
FTIR analysis showing no evidence of FNO as an end product.12
The final products of the CF3 + NO2 reaction have been
investigated in a flow tube reactor coupled to a chemical
ionization mass spectrometric (CIMS) detector by Bevilacqua
et al.13 The various products observed were CF2O (major
product), FNO, and CF3NO2 or CF3ONO. CF3O was confirmed
by monitoring a decline of the peak by reaction with isobutane.
Three competing reaction channels were suggested:
There have been several studies on the CF3 + NO2 reaction.
Using a very-low-pressure-photolysis (VLPP) molecular beam
sampling apparatus combined with mass spectroscopy, Rossi
et al. observed the production of CF2O from infrared multiple
photon dissociation (IRMPD) of CF3I in the presence of NO2.10
The rate coefficient of the CF3 + NO2 reaction was found to
be (2.7 ( 0.5) × 10-12 cm3molecule-1 s-1 at 298 K, and the
following reaction mechanism was suggested:
X Abstract published in AdVance ACS Abstracts, February 15, 1997.
S1089-5639(96)02655-2 CCC: $14.00 © 1997 American Chemical Society

IR Emission from the CF3 + NO2 Reaction
J. Phys. Chem. A, Vol. 101, No. 14, 1997 2635
of the rate coefficient, the establishment of the reaction
mechanism, and the measurement of the relative branching ratio
for reactions 1a and 1b. The related reaction of CF3O + NO
was also investigated to compare two different reaction channels
which yield the same products. The comparison of the two
reactions CF3 + NO2 and CF3O + NO support the conclusion
that both CF2O and FNO from the CF3 + NO2 reaction are
vibrationally excited.
CF3 + NO2 f CF2O + FNO
f CF3O + NO
(1a)
(1c)
(1d)
f CF3ONO
Pathway 1a was found to be the major channel. The overall
rate coefficient for reaction 1 was found to be k1 ) (1.0 ( 0.7)
× 10-11 cm3 molecule-1 s-1. O’Sullivan et al. also reported
that reaction of CF3 radicals with NO2 resulted in the formation
of CF2O (∼75% yield) and CF3NO2 (or CF3ONO) (∼25%),
and no other C-containing compounds were detected in the IR
spectra.14 Furthermore, O’Sullivan et al. inferred the possibility
that CF2O and FNO could also be dissociative products of CF3-
ONO. Very recently, the title reaction was studied at 296 ( 2
K using IR fluorescence and UV absorption spectroscopy and
a branching ratio of (reaction 1a):(reaction 1c) was estimated
to be (70 ( 12)%:(30 ( 12)%.15 In contrast, other mass
spectroscopic analyses showed that the quantum yield of CF2O
as a final product of the CF3 + NO2 reaction is unity.25
The rate coefficients and products of the reaction of the related
CF3O radical with NO and NO2 have been reported.13,17-24 The
CF3O radical is an important haloalkoxy radical formed during
the tropospheric oxidation of CFC substitutes which contain a
CF3 group. The reaction products and rate coefficients are
Experimental Section
The time-resolved FTIR emission technique and the experi-
mental procedures have been described previously in detail.26
A brief summary is presented here. The CF3 radical was
produced from IRMPD of either CF3I or CF3Br using radiation
at 1074.6 cm-1 from a pulsed CO2 laser (the 9R(14) line). At
low CO2 laser fluence, the majority of CF3I (or CF3Br) precursor
molecules are dissociated in an IRMPD scheme:
CF3I + nhν f CF3 + I
(7a)
The fluence is insufficient to promote the less favorable
dissociation (7b):
CF3I + nhν f F + products
(7b)
CF3O + NO f CF2O + FNO
k4 ) (5.2 ( 2.7) ×
10-11 cm3 molecule-1 s-1 (4)
k5 ) (5.1 ( 0.6) ×
At high fluence, F atoms are formed by process 7b, and thus
in order to avoid the side reactions caused by F atoms, all time-
resolved FTIR emission spectra for the CF3 + NO2 reaction
were carefully taken at low fluence (0.5-3.0 J cm-2). IR
emission was passed through a modified Michelson interfer-
ometer operating in stop-scan mode, and recorded either with
a HgCdTe detector (detection range 1100-2400 cm-1, with an
optical filter used to block scattered CO2 laser radiation below
1100 cm-1) or with an InSb detector (detection range 1850-
4500 cm-1, with a filter combination). Time-resolved signals
were captured with a custom built transient recorder (1 µs time
resolution). In some cases, narrow band IR filters were used
to isolate emission from the products of interest. The gases
used in this study were CF3I, Fluorochem, >99%; CF3Br,
Fluorochem, >99%; H2, BOC, 99.7%; Ar, BOC, 99.998%; NO2,
BOC. All thermodynamic data in this study are taken from
the literature.27
CF3O + NO2 f CF3ONO2
10-12 cm3 molecule-1 s-1 (5a)
f CF2O + FNO2
(5b)
The rate coefficient of reaction 4 is taken from a recent study
investigated by a pulsed radiolysis UV absorption technique at
254 and 276 nm by Sehested et al.18 Studies of reaction (4)
have been used to show that CF2O is the only C-atom-containing
product.19 However, the FNO product was not detected,
possibly as a result of its efficient destruction by collision with
the walls of the reactor: the failure to detect FNO may not be
used as evidence to exclude its production in the CF3O + NO
reaction. For reaction 5, the rate coefficient is taken from recent
studies by photolysis/FTIR absorption methods19 and laser
photolysis/LIF.20 Chen et al. used photolysis of CF3NO to
prepare CF3O2 and subsequently CF3O in 700 Torr of air at
297 ( 2 K and found that the formation of CF3ONO2 is the
dominant channel (k(5a)/[k(5a) + k(5b)] > 90%).21
Notwithstanding the improved accuracy in recent measure-
ments of the rate coefficient for the CF3 + NO2 reaction, the
mechanism of the reaction, the branching ratio of the product
channels, and the role of the CF3O radical in the mechanism,
remain unclear. In this study, IR emission assigned to the CF2O,
FNO, and HF (with H2 added) products of the CF3 + NO2
reaction was observed and interpreted to show that the emitting
products are formed via the mechanism:
Results
1. Time-Resolved FTIR Emission Studies of the CF3 +
NO2 Reaction. IR emission was observed over the range
1100-4500 cm-1 when CF3I was irradiated in the presence of
NO2. Figure 1 shows a survey of the infrared emission spectrum
obtained over this range summed between 0-300 µs after the
dissociating laser pulse and taken with both the HgCdTe and
InSb detectors. In the range covered by the HgCdTe detector,
1100-2400 cm-1, we identify five emission bands marked as
a-f on the upper panel of Figure 1, namely at 1184, 1250, 1625,
1844, 1880, and 1950 cm-1. We first deal with feature a at
1184 cm-1, which is assigned to the precursor molecule CF3I‡
(ν4) excited by IR multiple-photon excitation (IRMPE),28,29 as
these emission bands were also detected at early times (<30
µs) without NO2 present. Feature b also appeared in the absence
of NO2 and is again in the position expected for emission from
CF3I (ν2 + ν5), but with NO2 present it persisted for longer
time periods than before. Emission near 1625 cm-1, feature c,
was attributed to the asymmetric stretch of NO2, which appears
to be excited through an energy transfer process.30 Strong
emission in the features d, e, and f are in the regions expected
CF3 + NO2 f CF2O + FNO
∆H°298 ) -267 kJ
mol-1 (1a)
∆H°298 ) -31 kJ
mol-1 (1b)
f CF2O + F + NO
F + H2 f HF(v) + H ∆H°298 ) -135 kJ mol-1 (6)
The discussion in this report is concerned with the determination
of the products of the CF3 + NO2 reaction, the measurement
for FNO at 1844 cm-1 31-34 FON near 1880 cm-1 32-34
, , NO at

2636 J. Phys. Chem. A, Vol. 101, No. 14, 1997
Oum and Hancock
Figure 1. The infrared emission spectrum obtained following the CF3 + NO2 reaction over the detection range of 1100-4500 cm-1, summed
between 0-300 µs. Instrumental resolution is 6 cm-1. Conditions were 50 mTorr CF3I, 100 mTorr NO2, and 5 Torr Ar. Bands a-f shown in the
figure are identifed in the text.
Figure 2. Time-resolved FTIR emission spectra of the CF3 + NO2 reaction in the 1100-2100 cm-1 region, taken with the HgCdTe detector, at
low fluence with a 6 cm-1 resolution: 50 mTorr CF3I, 100 mTorr NO2, and 5 Torr Ar. Emission intensity is in arbitrary units. Each spectrum in
the result of averaging over a 2 µs window at the times shown following the CO2 laser pulse. Features a-f are as in Figure 1.
1876 cm-1, and CF2O at 1944 cm-1 (ν2),17,35 all potential
products of reaction 1. The major additional feature seen in
the InSb spectrum (Figure 1) is emission at 4000 cm-1 from
vibrationally excited HF formed when H2 was added to the
reaction mixture and indicating the presence of F atoms which
react in process 6.
Figure 2 shows time-resolved FTIR spectra in the range
1100-2100 cm-1 following the reaction of CF3 + NO2 taken
at 2 µs time intervals and at a resolution of 6 cm-1. The spectra
clearly show that the emission near 1200 cm-1, band a, from
excited CF3I‡ peaks at very early time and disappears quickly,
while the emission bands d-f attributed to reaction products

IR Emission from the CF3 + NO2 Reaction
J. Phys. Chem. A, Vol. 101, No. 14, 1997 2637
Figure 3. The intensity versus time traces of each emission extracted
from Figure 2. Emission from excited precursor (emission a) appears
earlier than that assigned to FNO (emission d) and CF2O (emission f),
and both CF2O and FNO are formed with the same production rate.
The evolution rate of NO2 emission (emission c) is compared with the
relaxation rate of the CF3I‡ emission on the top right-hand side.
appear later and last longer. The CF3I‡ band near 1250 cm-1
is overlapped by potential emission from CF2O (ν4). After the
excited precursor emission near 1184 cm-1 has disappeared (for
example, at 80 µs), the emission at 1250 cm-1 remains and then
decays with the same kinetics as for band f near 1944 cm-1
.
Figure 3 shows the intensity versus time traces of the peaks
of each emission feature extracted from Figure 2. What is clear
from these traces is that the rise times for features (a) and (c)
are markedly different from those of the major emissions d, e,
and f. For feature a, attributed to vibrationally excited precursor,
rapid formation during the CO2 laser pulse is followed by rapid
quenching, as has been observed in many previous studies of
IRMPE.29 Bands d and f are seen to rise with the same rate,
but decay somewhat differently. Band e shows the same kinetic
Figure 4. Comparison of the emission bands near 1800-2000 cm-1
following the CF3 + NO2 reaction (i) and with the emission band of
CF2O itself (ii). CF2O in Figure 4ii was generated from the oxidation
reaction of the CF3 radical with O3 in the absence of NO/NO2. Band d
is identified as FNO, band e discussed in the text, and band f identified
as CF2O are marked in the Figure 4i. Also shown in Figure 4i are the
absorption bands of FNO and CF2O taken from ref 17.
observed emission bands, d, e, and f, detected from the CF3 +
behavior as band d. The emission from NO2 at 1625 cm-1
,
NO2 reaction, together with the infrared absorption spectra of
band c, appeared to be excited by slower energy transfer from
excited products of the reaction of CF3 + NO2 or excited
precursor molecules. Figure 3 also compares the evolution rate
of NO2 emission with the relaxation rate of the CF3I‡ emission.
FNO (ν1 band origin at 1844 cm-1 31-34
and CF2O (ν2 band
)
origin at 1944 cm-1).17,35 The resolution of the present
observations (6 cm-1) is such that positive identification through
rotational structure of an emission band is not possible, and
thus the assignments of the bands in this region necessarily
remain tentative. Bands d and f are in the positions expected
for the strong ∆ν1 and ∆ν2 ) -1 transitions in FNO and CF2O,
respectively, and are identified as such. The source for band e
is not immediately obvious. It could be argued that band e could
come from highly vibrationally excited CF2O emitting in a red-
shifted region. In order to test this, the emission from CF2O
only was generated from the reaction of the CF3 radical with
O3 reaction 8 in the absence of NO/NO2:
Infrared emission between 1850-4500 cm-1 was investigated
using the InSb detector. Compared to the HgCdTe detector,
the InSb detector had a higher detection sensitivity in this region,
and thus spectra near 1944 cm-1 were obtained with a better
signal to noise ratio, as can be seen in Figure 1, but, owing to
the InSb detector cutoff (∼1850 cm-1), only emission band f
and a part of band e were observed. Two other weak emissions
were seen, one assigned to excited precursor CF3I‡ near 2100-
2400 cm-1 (2ν1 at 2146 cm-1, ν4 + ν1 at 2258 cm-1, and 2ν4
at 2370 cm-1) and one near 3851 cm-1, in the position expected
for the overtone transition 2ν2 of CF2O. The different time
evolutions of the bands near 2100-2400 and 3851 cm-1 were
consistent with them being from excited precursor and from
the same source as feature f, respectively.
CF + O3 f CF O‡ + FO
2
∆H°298 ) -284 kJ mol-1
3
2
(8a)
f CF2O‡ + F + O2
∆H°298 ) -231 kJ
When CF3Br was used as a source for the CF3 radical, the
same infrared emissions in the 1800-2000 cm-1 region, bands
d, e, and f were detected. The similarity of the spectra with
those generated in the experiments which used CF3I as the
precursor for the CF3 radical show that other photolytic products
from the precursor, for example, I or Br, do not interfere
significantly with the CF3 + NO2 reaction.
mol-1 (8b)
Time-resolved FTIR emission spectra of reaction 8 showed
that the emission from high vibrational levels of CF2O pre-
dominated, with a cascade to lower levels shifting the spectrum
to higher wavenumbers as time progressed, eventually leading
to emission from the ground vibrational state of CF2O near 1944
cm-1. Figure 4 illustrates a comparison of the emission bands
near 1800-2000 cm-1 following the CF3 + NO2 reaction:
Figure 4i, with the emission band of CF2O itself formed from
2. Emission Bands near 1800-2000 cm-1
. The main
problem of this study is to identify the emission bands from
reaction products near 1800-2000 cm-1. Figure 4i shows the

2638 J. Phys. Chem. A, Vol. 101, No. 14, 1997
Oum and Hancock
the CF3 + O3 reaction, Figure 4ii, summed over the same time
interval and under similar precursor pressures. The time
evolution of these two spectra shows that band e behaves very
differently from the low-wavenumber emission from vibra-
tionally excited CF2O shown in Figure 4ii, the former retaining
its separation from the main CF2O peak at 1944 cm-1 over all
times later than ∼20 µs from reaction initiation. Considering
that the available energy in the vibrationally excited CF2O in
the two reactions 1 and 8, are similar, 267 and 231 kJ mol-1 27
,
respectively, it appears that band e in Figure 1 cannot be from
highly vibrationally excited CF2O.
NO has an emission band near 1874 cm-1, exactly in the
position observed for band e. To test this, a cold gas filter
experiment was performed to quantify the amount of NO (v )
1 f 0) emission. When a low pressure of NO (2 Torr) was
present in the cold gas filter cell, the intensity of the emission
between 1840-1920 cm-1 isolated by a filter and reaching the
detector was not significantly decreased, a result which implied
that the only a small portion of the emission was from NO (v)1)
(if the emission is entirely from the excited NO, ∼75% of the
emission would be resonantly absorbed in a 10 cm cell under
these conditions). The result of cold gas filter experiments gave
an estimate of the fraction of NO (v)1f0) in the emission
observed over 10-100 µs near 1840-1920 cm-1 as 1.65%.
Figure 5. Plots of the rising (b) and falling (O) rates of the CF2O
emission against NO2 pressure. Conditions were 7 mTorr CF3I, 5 Torr
Ar, and 16, 20, 25, 35, 40, 45, 50, 60, 70, 80, 89, 100, 110, 120, 130,
and 150 mTorr NO2; the emissions were observed through an
interference filter FWHM 50 cm-1 centered at 1880 cm-1. The straight-
line-least-squares-fits to the data give rising and falling rate constants
of (2.4 ( 0.5) × 10-11 and (6.4 ( 0.4) × 10-12 cm3 molecule-1 s-1
,
respectively.
separate these rate processes, a comparison of the quenching
effect of the emission by N2 and O2 was investigated. N2 or
O2 (0-2.5 Torr) were added in turn to a gas sample consisting
of 20 mTorr CF3I, 100 mTorr NO2, and 5.8 Torr Ar. When N2
was added to the gas sample, the rising rates were unchanged
but the falling rates were increased with N2 pressure with a
slope of 1.24 × 10-13 cm3 molecule-1 s-1, which is identified
with the quenching rate of the vibrationally excited products
by N2. However, when O2 was added to the gas sample, both
rising and falling rates were increased, with slopes of 1.08 ×
10-12 and 2.14 × 10-13 cm3 molecule-1 s-1, respectively: the
slope for the falling rates corresponds to quenching of the
emission by O2 and the observation of increased rising rates at
higher O2 pressure was consistent with the loss rate of the CF3
radical by reaction with O2 under the present experimental
conditions.1,8 The precursor pressure dependence of the emis-
sion was also investigated. The rising rates were again found
to be independent of CF3I pressure, and the amplitude of the
emission increased linearly with CF3I pressure. It seems that
the production of CF2O is not affected by any subsequent
reactions with other products from the CF3 + NO2 reaction.
The time dependence of the emission on the pressure of added
H2 was also investigated. The emission showed a simple
quenching response with increased H2; rising rates were found
to be independent of H2 pressure and falling rates were increased
at higher H2 pressures. This result implies that the production
mechanism for CF2O is independent of H2 and that the
subsequent H2 + CF2O reaction 9 is not significant on the
timescale of the current experiments:
Comparison of the spectra in Figure 4 and the result of NO
cold gas filter experiment indicate that the emission band e is
from neither CF2O nor NO. A possible candidate for band e is
FON; this will be discussed later in detail.
3. CO2 Laser Fluence Dependence of the Emitting
Species. In the low CO2 laser fluence region (<3 J cm-2), the
intensities of the emissions identified as from CF2O, FNO, and
CF3I‡ were monitored as a function of laser fluence. The CF2O
and FNO emissions showed the same fluence dependencies with
a threshold for the formation of the CF3 radical near 0.6 J cm-2
.
Excited precursor emissions, at 1184 and near 2250 cm-1, show
in contrast a much lower fluence threshold, well below that for
the production of the CF3 radical. Thus, it was confirmed that
the emissions from CF2O and FNO originated in the CF3 radical
pathway 1a, and not from the reaction of CF3I‡ + NO2.
4. Kinetics of the Product Emission. The time depend-
encies of the emission features were investigated at various NO2
concentrations. These experiments were performed at low
fluence in order to avoid any side reactions involving F atoms
formed from the precursor.29 A narrow band-pass optical filter
(1840-1920 cm-1) was placed in front of the HgCdTe detector
to measure the emission from bands d, e, and f: this wave-
number range was chosen to encompass emission from the most
highly vibrationally excited CF2O formed to avoid the cascade
of vibrational population from higher levels. It should be
emphasized, however, that the peaks of these emission bands
showed very similar rising rates, as shown in Figure 3. The
time dependence of the emission showed a characteristic double-
exponential rising and falling behavior (one for reaction and
one for relaxation), the rates of which both increased with
increasing NO2 pressures. A plot of the rising rates versus NO2
pressure was linear, as shown in Figure 5, with a slope of (2.4
( 0.5) × 10-11 cm3 molecule-1 s-1, which was consistent with
two previous measurements of the rate coefficient of the CF3
+ NO2 reaction within error range.11,12 The falling rates were
also increased with a slope of (6.4 ( 0.4) × 10-12 cm3
molecule-1 s-1 as a function of the NO2 pressure. From such
behavior, however, it is not possible unequivocally to identify
the rising rate with process 1 (as was done in the IR emission
study of Francisco and Li12), as fast relaxation and slow reaction
would produce the same double-exponential kinetics. To
H2 + CF2O f products
(9)
The above results confirm that the rising rates are consistent
with the loss of the CF3 radical, i.e., the CF2O production by
reaction 1, and falling rates are the relaxation of vibrationally
excited products. Therefore, the observed slope of the rising
rates as a function of NO2 pressure, (2.4 ( 0.5) × 10-11 cm3
molecule-1 s-1, represents the rate constant for the CF3 + NO2
reaction.
Further investigation of the emission was carried out with a
filter in the range 1915-1980 cm-1, i.e., covering all of band
f. Again double-exponential rising and falling rates were seen,

IR Emission from the CF3 + NO2 Reaction
J. Phys. Chem. A, Vol. 101, No. 14, 1997 2639
coefficients of HF(v) are known to be larger than those for
CF2O,38 the intensity of HF(v) emission observed was much
smaller than that of CF2O.
5.b. Measurement of the CF2O:F Ratio. A series of
experiments based on the IRMPD of CF4 were carried out to
quantify the relative amounts of F and CF2O produced from
the CF3 + NO2 reaction. CF4 was chosen as the “reference
precursor” because the IRMPD of CF4 produces CF3 and F in
a 1:1 yield:
CF4 + nhν f CF3 + F
∆H°298 ) 545 kJ mol-1 (10)
In the present system both NO2 and H2 are present. The CF3
radicals produced by the IRMPD of CF4 react with NO2 via
one of two reaction channels (the numbers in square brackets
are the branching ratios for the two channels):
Since the CF2O species is produced in the reaction of CF3 +
NO2, its emission may be used as a measure of the CF3
concentration. The emission from HF excited by the F + H2
reaction is related to the F atom concentration: note, however,
that F atoms are produced by both the IRMPD of CF4 and by
one branch of the CF3 + NO2 reaction. Therefore, the ratio of
the emission intensities from CF2O and HF in the presence of
NO2 and H2 following the IRMPD of CF4 corresponds to a ratio
of CF2O : F ) 1:(1 + y). The concentration ratio of CF2O:HF
from the CF3 + NO2 reaction is 1:y, when CF3 is produced
from the IRMPD of the CF3I precursor. The intensity ratio of
the CF2O and HF emissions produced from the CF3I precursor
was measured at low fluence (to avoid IRMPD directly forming
F atoms), then compared with that from CF4 precursor under
the same experimental conditions except at high fluence
(conditions required for the IRMPD of the precursor). The
comparison yielded a value of y ) 0.015 ( 0.002, which means
that the relative branching ratio of two reaction pathways
(reaction 1a):(reaction 1b) is 1:0.015, and is a result which is
in good agreement with the estimation of a small amount of
NO (v ) 1f0) made from the NO cold gas filter experiment.
Error limits are estimated from uncertainties in the response of
the IR detector over the dynamic range studied. In conclusion,
our results are consistent with the products of the CF3 + NO2
reaction being predominantly CF2O + FNO with a far smaller
quantity of CF2O + NO + F.
Figure 6. (a) Time evolution of normalized vibrational populations
of HF(v) formed in the CF3 + NO2 reaction in the presence of H2.
Conditions were 7 mTorr CF3I, 75 mTorr NO2, 50 mTorr H2, and 5.2
Torr Ar with a 10 cm-1 resolution. Extrapolated values to zero time
represent the nascent vibrational distribution of the HF(v) and are plotted
in (b). The nascent distribution of HF(v) is compared with the literature
values for the F + H2 reaction.
but with a rising rate constant a factor of 2 smaller than that
shown in Figure 3. We interpret this lower rate constant as
demonstrating the effects of vibrational cascade from higher
levels of CF2O slowing down the production process. Figure
2 shows that as time progresses the peak of feature f is not
developed until some 20 µs after reaction initiation, in contrast
with the earlier appearance of the resolved peak of feature d.
Vibrational cascade will thus affect any kinetic observations
which include the high-wavenumber part of peak f and will lead
to an underestimation of the true rate constant.
5. F-Atom Formation in the CF3 + NO2 Reaction. 5.a.
Source of Emission from Vibrationally Excited HF. Further
results were obtained when H2 was added to the reaction system.
No HF emission was seen in the absence of NO2, but when
NO2 was added, HF emission appeared. The fluence depend-
ence of this emission was found to be similar to that from CF2O
and FNO, a result which implied that the source of HF was
from reaction of the CF3 radical.
Figure 6a shows the normalized vibrational populations of
HF as a function of delay time after the CO2 laser pulse. The
nascent HF(v) distribution was obtained from an extrapolation
of the data shown in Figure 6a to a delay of zero. The HF(v)
distribution so obtained compared well with literature values36,37
for the F + H2 reaction as shown in Figure 6b, a result which
implies strongly the reaction as the source of the majority of
HF in the present system. The kinetics of formation of HF were
found to be consistent with its formation from the reaction
sequence 1b and 6. The reaction must, however, be dominated
by pathway 1a because, despite the fact that the emission
Discussion
1. Band e. We conclude from the evidence presented above
that CF2O and FNO are both formed as major products of the
CF3 + NO2 reaction. However, there is still an unresolved
problem in identifying of band e, which emits near 1880 cm-1
.
The production and relaxation rates of band e were very similar
to those of FNO, band d, in Figure 3. A possibility of the
formation of the isomer FON via a five-center intermediate was
considered. Smardzewski and Fox in 1974 reported the three
fundamentals of FON from Ar- and N2- matrix isolation
experiments: 1886.6 (ν1), 735.1 (ν2), and 492.2 (ν3) cm-1 in
an Ar-matrix; 1904.1 (ν1), 724.6 (ν2), and 485.4 (ν3) cm-1 in a
N2-matrix.32 In each matrix, the peak assigned to FON was
blue shifted with respect to that from FNO by ∼34 cm-1
,
identical to that observed between bands d and e in the present

2640 J. Phys. Chem. A, Vol. 101, No. 14, 1997
Oum and Hancock
formed in excess NO2. The CF3O radical was produced by the
following mechanisms when O2 and NO were added to the
system:
CF3 + O2 + M f CF3OO + M
CF3OO + NO f CF3O + NO2
CF3O + NO f CF2O + F(NO)
CF3O + NO2 f CF3ONO2
(12)
(13)
(4)
Figure 7. Two possible reaction intermediates in the CF3+NO2 reaction
leading to CF2O and F(ON) products.
experiments. No observation of the gas phase absorption
spectrum of FON has been reported, but a similar isomeric shift
would be expected. It should however be noted that Smard-
zewski and Fox’s results have been questioned : the lack of
observation of a peak near 1887 cm-1 in a later Ar matrix
experiment33 (in which the isomeric FON was believed to be
present) and ab initio studies of the predicted isomeric and
isotopic shifts34 have raised the possibility that the two bands
observed in the 1850-1887 cm-1 range32 might be from FNO
present in different matrix sites. A simple idea of two possible
intermediates (namely, four-center and five-center) of the CF3
+ NO2 reaction is illustrated in Figure 7. If reaction proceeds
via a five-center intermediate, then FON can be formed, reaction
1e.
(5a)
(5b)
CF3O + NO2 f CF2O + FNO2
Although the rate coefficient for reaction of CF3O with NO (4)
is ∼10 times faster than that with NO2 (reaction 5), under
conditions of excess NO2 (and O2), reaction 5 necessarily
dominates reaction 4. Our conclusions from these studies are
that the mechanism including the CF3O radical is unlikely.
A brief investigation was made of the F + NO2 reaction as
a possible source of the FNO species as a small fraction of F
atoms are produced from the CF3 + NO2 reaction. F atoms
were produced by IRMPD of SF6 with the 10P(20) line of the
CO2 laser. When 50 mTorr of NO2 was added to the sample
of 30 mTorr SF6 and 5 Torr Ar, a weak emission near 1800-
1900 cm-1 was detected, the source of which could have been
FNO. However, the intensity of the emission was very small
when compared to the FNO emission from the CF3 +NO2
reaction, even at high SF6 pressures. Therefore, it seems that
the reaction of F atoms with NO2 is almost certainly not the
source of the FNO emission observed near 1800-1900 cm-1
in the CF3 + NO2 reaction.
CF3 + NO2 f CF2O + FNO
f CF2O + FON
∆H°298 ) -267 kJ
mol-1 (1a)
∆H°298 ) -112 kJ
mol-1 (1e)
The FON isomer is calculated to be the less stable by some
155 kJ mol-1, but has a barrier to isomerisation of some 35 kJ
3. Comparison with the CF3O + NO Reaction. The CF3O
+ NO and CF3 + NO2 reactions can yield the same products,
CF2O + F(NO).13,17-24 Emissions were observed in the 1850-
4500 cm-1 region following the IRMPD of CF3I in the presence
of excess NO and O2. CF3O is produced via processes 12 and
mol-1 34,39
Even if this interpretation is correct, the data are
.
not of sufficient quality to enable any quantitative conclusions
to be drawn concerning the relative quantum yields of channels
1a and 1e, owing to the overlap of the spectral features d, e,
and f and their unknown vibrational distributions and emission
coefficients. Ab initio calculations of transition state structures
similar to those illustrated in Figure 7 would help in deciding
the feasibility of forming the (less strained) five-membered ring
system leading to FON formation. However, at present such
formation must be regarded as speculative.
13, and C‡ O then reacts with NO to give the observed products
3
of reaction 4. Both NO and O2 were required to produce
emission from CF2O. The rate coefficient for reaction 13 has
been reported as (1.57 ( 0.31) × 10-11 cm3 molecule-1 s-1
from observations of the removal rate of the CF3OO radical,23,24
and the reaction of CF3OO + NO is thought to proceed solely
by this path.24 Another possible source for the CF3O radical is
the self-reaction of CF3OO, but under the present conditions
its contribution to CF2O production is low.
There is a well known but slow reaction between O2 and
NO to give NO2 or NO3.40 Since both O2 (600 mTorr) and
NO (70 mTorr) are present in the reaction system at relatively
high pressures, the reaction between these species could be a
significant source of extra NO2 and could thus react with CF3
as demonstrated in the earlier part of this study. An experiment
was performed in order to quantify this. Both O2 and NO were
introduced to the flow cell through the side arms of the flow
tube. The reagent O2 was introduced from a fixed position,
and NO injection point was varied, thus changing the contact
time between O2 and NO. There was no difference in intensity
of the CF2O emission when the injection position of NO was
changed. Therefore, the O2 + NO reaction could not be a source
of extra NO2.
2. Mechanism for Formation of CF2O and FNO. Rossi
et al. suggested that the production of CF2O in the CF3 + NO2
reaction occurs via a two step reaction mechanism via reactions
1c and 2 involving the sequence CF3 f CF3O f CF2O.10
Under our conditions, process 2 can only take place for CF3O
species having considerable vibrational energy, and from the
kinetic evidence presented above, it would seem that process
1c is rate limiting. This two-step scheme, however, suggests
the formation of equal quantities of F and CF2O products, with
no FNO produced. Evidence for the F/CF2O ratio being far
smaller (0.015) is not consistent with this scheme. Furthermore,
features assigned to vibrationally excited CF2O and FNO appear
with the same production rate, as can be seen in Figure 3, and
is further good evidence to support the direct formation of both
products such as process 1a. We can discount the possibility
of formation of CF3O by reaction 1c followed by production
of vibrationally excited CF2O by process 5b, its reaction with
NO2. Under our conditions, reaction 5b would be rate deter-
mining (k5 = 0.2k4) and would thus result in slower formation
rates than those experimentally observed. No emission could
be detected in the CF3 + NO2 system from possible products
of the CF3O + NO2 reaction.19,21 In a separate experiment,
weak infrared emission near 1750 cm-1, which could be from
CF3ONO2, or more likely FNO2, was detected when CF3O was
The infrared emission from CF2O formed in the CF3O + NO
reaction was about four times smaller than that from the CF3 +
NO2 reaction at similar reagent concentrations. The slower
production rate of CF2O and the lower exothermicity of the
CF3O + NO reaction account for this. Figure 8 compares the
infrared emission spectra recorded following the CF3O + NO
reaction with that recorded following the CF3 + NO2 reaction.

IR Emission from the CF3 + NO2 Reaction
J. Phys. Chem. A, Vol. 101, No. 14, 1997 2641
Figure 8. Comparison of the infrared emission spectra recorded following the CF3O + NO reaction at 50 µs with that recorded following the CF3
+ NO2 reaction at 30 µs. A closer inspection, marked with circle, shows that the ratio of the F(NO) emission intensity to that of the CF2O emission
is greater in the case of the CF3O + NO reaction.
At first sight, the spectra appear similar, although closer
inspection shows that the ratio of the emission intensities of
F(NO):CF2O was greater in the case of the CF3O + NO reaction.
Both reactions are highly exothermic. However, the CF3 + NO2
reaction involves the formation of two new bonds, the CF2dO
and F-(NO) bonds. As a consequence, both products are highly
vibrationally excited and produce strong infrared emissions. In
contrast, the F-atom transfer pathway in the reaction of CF3O
with NO to give CF2O and F(NO) involves the formation of
only one new bond, the F-(NO) bond, and thus it would be
expected that F(NO) is more excited than CF2O.
Acknowledgment. The authors thank Dr. Carlos Canosa-
Mas for helpful discussions. K.W.O. is grateful to the British
Council for financial support.
References and Notes
(1) (a) Ryan, K. R.; Plumb, I. C. J. Phys. Chem. 1982, 86, 4678. (b)
Caralp, F.; Lesclaux, R.; Dognon, A. M. Chem. Phys. Lett. 1986, 129, 433.
(c) Withnall, R.; Sodeau, J. R. J. Photochem. 1986, 33, 1. (d) Clemitshaw,
K. C.; Sodeau, J. R. J. Phys. Chem. 1989, 93, 3552.
(2) (a) Francisco, J. S.; Williams, I. H. Int. J. Chem. Kin. 1988, 20,
455. (b) Francisco, J. S.; Li, Z.; Williams, I. H. Chem. Phys. Lett. 1987,
140, 531.
(3) Smolinsky, G.; Flamm, D. L. J. Appl. Phys. 1979, 50, 4982.
(4) Tsai, C.; Belanger, S. M.; Kim, J. T.; Lord, J. R.; McFadden, D.
L. J. Phys. Chem. 1989, 93, 1916.
Conclusions
(5) Tsai, C.; McFadden, D. L. J. Phys. Chem. 1989, 93, 2471.
(6) Iyer, R. S.; Rowland, F. S. J. Phys. Chem. 1981, 85, 2493.
(7) Kuznetsova, S. V.; Maslov, A. I. SoV. J. Quantum Electron. 1989,
19, 637.
(8) (a) Clemitshaw, K. C.; Sodeau, J. R. J. Photochem. Photobiol. A
1995, 86, 9. (b) Vedeneev, V. I.; Teitelboim, M. A.; Shoikhet A. A. IzV.
Akad. Nauk SSSr, Ser. Khim. 1977, 8, 1757.
(9) (a) Vakhtin, A. B.; Petrov, A. K. Spectrochim. Acta Part A 1990,
46, 603. (b) Orlando, J. J.; Smith, D. R. J. Phys. Chem. 1988, 92, 5147. (c)
Kaiser, E. W.; Wallington, T. J.; Hurley, M. D. Int. J. Chem. Kinet. 1995,
27, 205.
The reaction of the CF3 radical with NO2 has been investi-
gated by using time-resolved FTIR emission spectroscopy.
Emission features near 1850 and 1940 cm-1 have been
tentatively assigned as originating from the vibrationally excited
reaction products FNO and CF2O respectively, and the pos-
sibility of an isomeric FON species emitting at 1880 cm-1 is
discussed. Two reaction channels 1a and 1b were suggested
and the relative branching ratio of reactions 1a and 1b was
estimated as 1:0.015.
(10) Rossi, M. J.; Barker, J. R.; Golden, D. M. J. Chem. Phys. 1979,
71, 3722.
(11) Sugawara, K.; Nakanaga, T.; Takeo, H.; Matsumura, C. J. Phys.
Chem. 1989, 93, 1894.
(12) Francisco, J. S.; Li, Z. Chem. Phys. Lett. 1989, 162, 528.
(13) Bevilacqua, T. J.; Hanson, D. R.; Howard, C. J. J. Phys. Chem.
1993, 97, 3750.
(14) O’Sullivan, N.; Treacy, J.; Sidebottom, H. Homogeneous and
Heterogeneous Chemical Processes in the Troposphere. Presented at the
Joint-CEC-EUROTRAC Workshop, HALIPP/LACTOZ Working Group,
Strasbourg, France, 1995.
CF3 + NO2 f CF2O + FNO
f CF2O + F + NO
∆H°298 ) -267 kJ
mol-1 (1a)
∆H°298 ) -31 kJ
mol-1 (1b)
The rising rate of the product emission was identified with
the reaction of CF3 with NO2, and the rate constant for reaction
1 was determined to be (2.4 ( 0.5) × 10-11 cm3 molecule-1
s-1. The CF3O + NO reaction was investigated by the same
technique. Emission was again observed from the same
products, although the emission from CF2O relative to that from
F(NO) was somewhat smaller than the relative emission
intensities from the products of the CF3 + NO2 reaction. This
was explained by the fact that two new bonds are formed in
both product species in the CF3 + NO2 reaction.
(15) Sehested, J.; Nielsen, O. J.; Rinaldi, C. A.; Lane, S. I.; Ferrero, J.
C. Int. J. of Chem. Kinet. 1996, 28, 579.
(16) Selamoglu, N.; Rossi, M. J.; Golden, D. M. Chem. Phys. Lett. 1986,
24, 68.
(17) Chen, J.; Zhu, T.; Niki, H. J. Phys. Chem. 1992, 96, 6115.
(18) Sehested, J.; Nielsen, O. J. Chem. Phys. Lett. 1993, 206, 369.
(19) Fockenberg, Ch.; Bednarek, G.; Zellner, R. Abstracts in 13th
International Symposium on Gas Kinetics; Proceedings of the 13th
International Symposium on Gas Kinetics, Dublin, September 1994; Abstract
No. D35, p 316.
(20) Bourbon, C.; Brioukov, M.; Hanoune, B.; Sawerysyn, J. P.;
Devolder, P. Chem. Phys. Lett. 1996, 254, 203.

2642 J. Phys. Chem. A, Vol. 101, No. 14, 1997
Oum and Hancock
(21) Chen, J.; Young, V.; Zhu, T.; Niki, H. J. Phys. Chem. 1993, 97,
11696.
(22) (a) Dognon, A. M.; Carlap, F.; Lesclaux, R. J. Chim. Phys.-Chim.
Biol. 1985, 82, 349. (b) Dibble, T. S.; Maricq, M. M.; Szente, J. J.; Francisco,
J. S. J. Phys. Chem. 1995, 99, 17394.
(28) (a) Edgell, W. F.; May, C. E. J. Chem. Phys. 1954, 22, 1808. (b)
Fuss, W. Spectrochim. Acta, Part A 1982, 38, 829.
(29) Oum, K. W. D. Phil. Thesis, University of Oxford, 1996.
(30) Hartland, G. V.; Qin, D.; Dai, H. L. J. Chem. Phys. 1994, 100,
7832.
(23) Bhatnagar, A.; Carr, R. W. Chem. Phys. Lett. 1994, 231, 454.
(24) Turnipseed, A. A.; Barone, S. B.; Ravishankara, A. R. J. Phys.
Chem. 1994, 98, 4594.
(25) Fracheboud, J. M. D. Phil. Thesis, University of Oxford, 1995.
(26) (a) Heard, D. E.; Brownsword, R. A.; Weston, D. G.; Hancock, G.
Appl. Spectrosc. 1993, 47, 1438. (b) Hancock, G.; Heard, D. E. AdV.
Photochem. 1993, 18, 1. (c) Biggs, P.; Hancock, G.; Heard, D. E.; Wayne,
R. P. Meas. Sci. Technol. 1990, 1, 630.
(31) Magnuson, D. W. J. Chem. Phys. 1952, 20, 380.
(32) (a) Smardzewski, R. R.; Fox, W. B. J. Chem. Phys. 1974, 60, 2104.
(b) Smardzewski, R. R.; Fox, W. B. J. Am. Chem. Soc. 1974, 96, 304.
(33) Jacox, M. E. J. Phys. Chem. 1983, 87, 4940.
(34) (a) Lee, T. J. Chem. Phys. Lett. 1994, 223, 431. (b) Curtiss, L. A.;
Maroni, V. A. J. Phys. Chem. 1986, 90, 56.
(35) Nielsen, A. H.; Burke, T. G.; Woltz, P. J. H.; Jones, E. A. J. Chem.
Phys. 1952, 20, 596.
(27) (a) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.;
Kerr, J. A.; Troe, J. J. Phys. Chem. Ref. Data 1992, 21, 1566. (b) Okabe,
H. Photochemistry of Small Molecules; John Wiley & Sons: New York,
1978. (c) DeMoore, W. B.; Sander, S. P.; Golden, D. M.; Hampson, R. F.;
Howard, C. J.; Kurylo, M. J.; Molina, M. J.; Ravishankara, A. R.; Kolb, C.
E. Chemical Kinetics and Photochemical Data for Use in Stratospheric
Modelling EValuation. Technical Report No. 11 (JPL Publication 94-26),
Jet Propulsion Laboratory: Pasadena, 1994. (d) Benson, S. W. J. Phys.
Chem. 1994, 98, 2216.
(36) Berry, M. J. J. Chem. Phys. 1973, 59, 6229.
(37) Polanyi, J. C.; Woodall, K. B. J. Chem. Phys. 1972, 57, 1674.
(38) Herbelin, J. M.; Emanuel, G. J. Chem. Phys. 1974, 60, 689.
(39) (a) Alberts, I. L.; Handy, N. C.; Palmieri, P. Chem. Phys. Lett.
1986, 129, 176. (b) Smolyar, A. E.; Zaretskii, N. P.; Klimenko, N. M.;
Charkin,O. P. Russ. J. Inorg. Chem. (Transl. of Zh. Neorg. Khim.) 1979,
24, 1761.
(40) (a) Olbregts, J. Int. J. Chem. Kinet. 1985, 17, 835. (b) Ashmore, P.
G.; Burnett, M. G. J. Chem. Soc., Faraday Trans. 2. 1962, 58, 253.