Kinetics of several oxygen-containing carbon-centered free radical reactions with nitric oxide

Article abstract of DOI:10.1021/acs.jpca.5b01027

Kinetics of four carbon-centered, oxygen-containing free radical reactions with nitric oxide (NO) were investigated as a function of temperature at a few Torr pressure of helium, employing flow tube reactors coupled to a laser-photolysis/resonance-gas-discharge-lamp photoionization mass spectrometer (LP-RPIMS). Rate coefficients were directly determined from radical (R) decay signals under pseudo-first-order conditions ([R]₀ [NO]). The obtained rate coefficients showed negative temperature dependences, typical for a radical-radical association process, and can be represented by the following parametrizations (all in units of cm³ molecule^-1 s^-1): k(CH₂OH + NO) = (4.76 × 10^-21) × (T/300 K)^15.92 × exp[50700/(RT)] (T = 266-363 K, p = 0.79-3.44 Torr); k(CH₃CHOH + NO) = (1.27 × 10^-16) × (T/300 K)^6.81 × exp[28700/(RT)] (T = 241-363 K, p = 0.52-3.43 Torr); k(CH₃OCH₂ + NO) = (3.58 ± 0.12) × 10^-12 × (T/300 K)^-3.17±0.14 (T = 221-363 K, p = 0.50-0.80 Torr); k(T)₃ = 9.62 × 10^-11 × (T/300 K)^-5.99 × exp[-7100/(RT)] (T = 221-473 K, p = 1.41-2.95 Torr), with the uncertainties given as standard errors of the fits and the overall uncertainties estimated as ±20%. The rate of CH₃OCH₂ + NO reaction was measured in two density ranges due to its observed considerable pressure dependence, which was not found in the studied hydroxyalkyl reactions. In addition, the CH₃CO + NO rate coefficient was determined at two temperatures resulting in k_298K(CH₃CO + NO) = (5.6 ± 2.8) × 10^-13 cm³ molecule^-1 s^-1. No products were found during these experiments, reasons for which are briefly discussed.

Full text of DOI:10.1021/acs.jpca.5b01027

Article
pubs.acs.org/JPCA
Kinetics of Several Oxygen-Containing Carbon-Centered Free Radical
Reactions with Nitric Oxide
Matti P. Rissanen,*,† Marvin Ihlenborg,‡ Timo T. Pekkanen,§ and Raimo S. Timonen§
†Department of Physics, University of Helsinki, P.O. Box 64, 00014 Helsinki, Finland
‡Department of Chemistry, Institute of Physical Chemistry, University of Kiel, Max-Eyth-Straße 1, 24118 Kiel, Germany
§Department of Chemistry, University of Helsinki, P.O. Box 55, 00014 Helsinki, Finland
ABSTRACT: Kinetics of four carbon-centered, oxygen-containing free radical reactions
with nitric oxide (NO) were investigated as a function of temperature at a few Torr
pressure of helium, employing flow tube reactors coupled to a laser-photolysis/resonance-
gas-discharge-lamp photoionization mass spectrometer (LP-RPIMS). Rate coefficients
were directly determined from radical (R) decay signals under pseudo-first-order
conditions ([R]0 ≪ [NO]). The obtained rate coefficients showed negative temperature
dependences, typical for a radical−radical association process, and can be represented by
the following parametrizations (all in units of cm3 molecule−1 s−1): k(CH2OH + NO) =
(4.76 × 10−21) × (T/300 K)15.92 × exp[50700/(RT)] (T = 266−363 K, p = 0.79−3.44
Torr); k(CH3CHOH + NO) = (1.27 × 10−16) × (T/300 K)6.81 × exp[28700/(RT)] (T = 241−363 K, p = 0.52−3.43 Torr);
k(CH3OCH2 + NO) = (3.58 0.12) × 10−12 × (T/300 K)−3.17 0.14 (T = 221−363 K, p = 0.50−0.80 Torr); k(T)3 = 9.62 × 10−11
× (T/300 K)−5.99 × exp[−7100/(RT)] (T = 221−473 K, p = 1.41−2.95 Torr), with the uncertainties given as standard errors of
the fits and the overall uncertainties estimated as 20%. The rate of CH3OCH2 + NO reaction was measured in two density
ranges due to its observed considerable pressure dependence, which was not found in the studied hydroxyalkyl reactions. In
addition, the CH3CO + NO rate coefficient was determined at two temperatures resulting in k298K(CH3CO + NO) = (5.6 2.8)
× 10−13 cm3 molecule−1 s−1. No products were found during these experiments, reasons for which are briefly discussed.
·CH2OH, reaction is investigated, and the acetyl radical
reaction [CH3·C(O)] with NO is briefly touched. The previous
investigations of these reactions are scarce, with only room
temperature rate coefficients reported for the ·CH2OH + NO,15
CH3·C(O) + NO16,17 and for the CH3·CHOH + NO reaction.7
To our knowledge there are no previous results for the
·CH2OCH3 + NO reaction.
The search and use of replacements for fossil fuels in
combustion engines have the potential to increase ethanol
(CH3CH2OH)18,19 (currently the most used biofuel18) and
dimethyl ether (CH3OCH3)12,20 emissions to the atmosphere.
The isomeric C2H5O radicals investigated in the current work
(i.e., CH3·CHOH and ·CH2OCH3) are their primary
degradation products and are formed from the parent
molecules by simple hydrogen abstraction reactions.7,9,14
Under atmospheric conditions, the reaction with molecular
oxygen (O2) is expected to dominate carbon-centered radical
removal due to the very high O2 concentration and relatively
fast rate coefficients.1,5−7 However, the reactions with NOx
could offer gas-phase synthetic routes to potentially harmful
INTRODUCTION
■
Oxygenated, carbon-centered free radical reactions with
nitrogen oxides (NOx) occur in atmospheric oxidation
chemistry1,2 and in low temperature combustion chemistry.3,4
The primary oxygenated radicals are formed in the gas phase
from volatile oxygen containing organic compounds, such as
alcohols and ketones, in hydrogen abstraction reactions by
radicals (e.g., ·OH and ·Cl), in thermal decompositions under
combustion conditions, and by solar UV photolysis.1−6 For
example, the 1-hydroxyethyl radical (CH3·CHOH) is the
dominant product of OH-initiated ethanol (CH3CH2OH)
oxidation as the presence of OH-group lowers the α-hydrogen
bond dissociation energy.7−9 The 2-hydroxyethyl radical
(·CH2CH2OH) is the minor product of this reaction, and the
abstraction from −OH group is generally found negligible.
Methanol oxidation (CH3OH) in the ambient atmosphere
leads mainly to ·CH2OH radicals for the same reasons. The
acetyl radical ([CH3·C(O)] is a common product of many
photochemical reactions of ketones, with acetone (CH3−
C(O)CH3) photolysis being a significant contributor.10,11 The
methoxymethyl radical (·CH2OCH3) is an important primary
radical in the atmospheric dimethyl ether (CH3OCH3)
oxidation,12−14 produced by H-abstraction from the ether.
This study includes the reactions of two isomeric but
structurally very different C2H5O species (CH3·CHOH and
·CH2OCH3) with nitric oxide and thus supplies important
structurally relevant information on the radical reactivity. In
addition, the reaction of the smallest hydroxyalkyl radical,
Special Issue: 100 Years of Combustion Kinetics at Argonne: A
Festschrift for Lawrence B. Harding, Joe V. Michael, and Albert F.
Wagner
Received: January 31, 2015
Revised: May 5, 2015
© XXXX American Chemical Society
A
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
and toxic substances and also could potentially affect fuels
ignition behavior by affecting the chain branching steps in the
reaction mechanism. Thus, the reactivity of these radicals
toward nitrogen oxides should be assessed. In a previous
publication we have studied reactions of these same radicals
with NO2 under similar conditions,21 and here we extend the
data set to include the NO reactions of these species.
EXPERIMENTAL SECTION
■
The details of the laser-photolysis/resonance-gas-discharge-
lamp photoionization mass spectrometer (LP-RPIMS) appara-
tus and its data acquisition and analysis have been described in
detail previously,22 and thus, only a brief description of the
most relevant details is reported here. The reactions were
studied in a 17 mm i.d. tubular stainless-steel flow reactor
coated with Halocarbonwax or polydimethylsiloxane (PDMS)
and in a 16.5 mm i.d. uncoated Pyrex reactor, under a few Torr
pressure (about 0.5−3.5 Torr) of helium bath gas. All the
experiments were performed under pseudo-first-order con-
ditions with the molecular reagent nitric oxide (NO) in large
excess over the initial radical concentrations produced (i.e.,
[R]0 ≪ [NO]).
The hydrocarbon radical precursor was introduced into the
gas flow by bubbling a small, variable part of the bath gas
helium through a temperature-controlled liquid precursor
reservoir. The concentration in the gas mixture was then
estimated from the gas flow rates and the vapor pressures of the
pure compounds. The chemicals 1-hydroxy-2-propanone
(hydroxyacetone, CH3C(O)CH2OH, 95%) and 3-hydroxy-2-
butanone (acetoine, CH3C(O)CHOHCH3, 97%) used to
produce the hydroxyalkyl radicals (·CH2OH and CH3·
CHOH, respectively) were obtained from Fluka, and 1-
methoxy-2-propanone (methoxyacetone, CH3C(O)CH2OCH3,
95%) and 3,3-dimethyl-2-butanone (pinacolone, CH3C(O)C-
(CH3)3, 98%) for the ·CH2OCH3 and CH3·C(O) radical
production, respectively, were obtained from Aldrich. The
precursor liquids were purified by freeze−pump−thaw cycles.
The reagent nitric oxide (NO, Linde, 99.5%) was prepared as
described previously23 and used as a pure (∼100%) gas. Bath
gas helium (He, Messer-Griesheim, 99.9996%) was used as
supplied.
Figure 1. Structures of the photolytic precursors and the
corresponding radicals produced.
radicals was varied in some experiments to ensure that the
results obtained were independent of this parameter.
The decay of the produced radical concentration as a
function of time in the flow reactor was measured under
varying NO concentrations, and a single-exponential decay was
subsequently fitted to the signals obtained, with the decaying
time-dependence giving the pseudo-first-order rate coefficient
(k′) of the investigated reaction. In the beginning and at the
end of each experiment, the wall loss rate (kW) of the radical in
the reactor was measured without added NO reagent. The kW
contains all other loss processes of the radical in the reactor
except the R + NO reaction and results in a zero-reagent rate
shown as one data point in the pseudo-first-order rate plot
(Figure 2), with the slope of the plot giving the bimolecular R +
The radicals were generated by pulsed ArF exciplex laser
photolysis at 193 nm, which for the 1-methoxy-2-propanone
(i.e., for the ·CH2OCH3 radical source) can be presented as
CH3C(O)CH2OCH3 + hv (193 nm)
→ CH3·C(O) + ·CH2OCH3
(1)
CH3C(O)CH2OCH3 + hv (193 nm) → other products
Figure 2. Pseudo-first-order plot of the ·CH2OCH3 + NO reaction,
showing the observed first-order radical decay rates (k′) as a function
of reagent NO concentration of the experiment at 363 K and 0.80
Torr. The insert illustrates the signal decays obtained in the absence
(orange) and presence (cyan) of the reagent NO.
(2)
The ketones used in the current study (see Figure 1) are known
to photodissociate by simple C−C bond fission producing
acetyl radicals and substituted alkyl radicals (e.g., for
hydroxyacetone, CH3C(O)CH2OH + hv → CH3·C(O) +
·CH2OH 10) but also potentially have more photolysis
pathways accessible at 193 nm than just channel 1. Due to
the low initial radical concentrations produced in the laser pulse
(i.e., generally close to 1011 molecules cm−3), together with the
rather high reagent concentration used in the experiments
([NO] > 1012 molecules cm−3), the importance of other species
produced in the photolysis reaction 2 is considered negligible to
the results obtained. The laser pulse energy generating the
NO rate coefficient under those experimental conditions. For
the ·CH2OCH3 radical detection, a xenon ionization lamp (8.44
eV) with a sapphire window was used, and for hydroxyalkyl and
acetyl radical detections, a chlorine lamp (8.9−9.1 eV) with a
calcium fluoride window was used.
The R + NO rate coefficients were determined as a function
of temperature and pressure, and the temperature dependences
observed were fitted with three commonly applied equations:24
B
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Table 1. Results and Conditions of the Current R· + NO Experiments [R = ·CH2OH, CH3·CHOH, ·CH2OCH3, and CH3·C(O)]
a
b
c
d
T, K
p/Torr ([He]/1016 molecules cm−3
)
[NO]/1013 molecules cm−3 k /10−12 cm3 molecule−1 s−1
kW /s−1
kW /s−1
laser fluence/mJ cm−2
e
R = CH2OH (CH2OH + NO → Products)
Below 300 K: k(T)2 = [(1.25 0.22) × 10−14] × exp[(13800 400)/(RT)] cm3 molecule−1 s−1
0.14
Above 295 K: k(T)1 = [(3.19 0.04) × 10−11] × (T/300 K)−2.59
cm3 molecule−1 s−1
k(T)2 = [(1.92 0.29) × 10−13] × exp[(7000 400)/(RT)] cm3 molecule−1 s−1
266−363 K: k(T)3 = (4.76 × 10−21) × (T/300 K)15.92 × exp[50700/(RT)] cm3 molecule−1 s−1
266
266
267
282
298
298
298
336
0.74 (2.69)
3.44 (12.49)
2.76 (9.98)
2.80 (9.58)
0.83 (2.70)
3.10 (10.00)
3.12 (10.11)
3.51 (10.01)
0.62 (1.65)
3.78 (10.01)
0.39−1.27
0.48−2.63
1.12−4.57
1.14−5.57
0.47−3.99
0.62−3.08
1.03−5.66
1.68−7.14
1.22−9.47
1.08−7.21
8.01 0.78
6.33 0.23
6.27 0.39
4.60 0.12
3.27 0.22
3.32 0.23
3.14 0.14
2.39 0.15
1.90 0.06
1.99 0.05
25.3 3.1
31.3 1.8
33.3 1.8
29.8 0.8
24.1 1.1
22.1 0.7
22.9 1.3
22.9 1.5
26.2 1.8
18.1 1.0
28.9 5.2
32.3 3.2
39.8 9.7
34.0 3.6
24.4 4.0
26.6 4.3
23.3 4.3
28.3 5.8
26.8 3.1
18.7 1.9
11.0
13.0
17.0
17.0
12.6
17.6
9.9
11.2
f
363
363
15.4
e
R = CH3CHOH (CH3CHOH + NO → Products)
Below 280 K: k(T)2 = [(1.88 1.10) × 10−15] × exp[(20800 1200)/(RT)] cm3 molecule−1 s−1
Above 295 K: k(T)1 = [(1.31 0.04) × 10−11] × (T/300 K)−3.76
cm3 molecule−1 s−1
0.35
k(T)2 = [(2.21 0.78) × 10−13] × exp[(10200 900)/(RT)] cm3 molecule−1 s−1
241−363 K: k(T)3 = (1.27 × 10−16) × (T/300 K)6.81 × exp[28700/(RT) cm3 molecule−1 s−1
241
254
266
279
298
2.30 (9.23)
2.42 (9.19)
2.49 (9.04)
2.55 (8.84)
0.56 (1.80)
2.82 (9.14)
3.17 (9.10)
0.52 (1.39)
3.43 (9.12)
0.16−1.03
0.46−1.97
0.45−1.78
0.43−1.90
1.01−2.75
0.85−2.72
1.01−4.78
1.94−5.80
1.26−5.04
62.53 3.25
36.46 2.51
21.83 2.52
16.93 0.46
12.94 0.40
14.03 0.55
8.32 0.17
6.06 0.13
6.96 0.21
109.6 5.0
70.4 3.0
90.0 21.8
49.7 1.2
55.8 1.8
59.1 1.3
40.1 0.9
39.1 1.7
45.6 1.4
126.5 17.4
59.2 26.8
93.2 25.3
52.7 5.1
58.4 6.9
59.4 8.7
37.8 4.5
36.5 4.6
42.5 6.2
18.8
20.9
5.7
12.3
18.9
f
298
336
363
363
15.6
18.5
18.5
g
R = CH3OCH2, (CH3OCH2 + NO → Products)
At [He] = 2.1 × 1016 molecules cm−3
:
k(T)1 = [(3.58 0.12) × 10−12] × (T/300 K)−3.17
cm3 molecule−1 s−1
0.14
k(T)2 = [(2.22 0.49) × 10−13] × exp[(6900 500)/(RT)] cm3 molecule−1 s−1
221−363 K: k(T)3 = (5.72 × 10−11) × (T/300 K)−6.33 × exp[−6900/(RT)] cm3 molecule−1 s−1
At [He] = 6.2 × 1016 molecules cm−3
:
221−473 K: k(T)3 = (9.62 × 10−11) × (T/300 K)−5.99 × exp[−7100/(RT)] cm3 molecule−1 s−1
f
221
221
241
241
0.49 (2.12)
1.41 (6.16)
0.53 (2.12)
1.54 (6.15)
0.59 (2.14)
1.75 (6.34)
0.65 (2.12)
0.59 (1.92)
2.67 (8.58)
1.94 (6.20)
0.74 (2.13)
2.08 (5.95)
2.25 (6.03)
0.80 (2.13)
2.45 (6.02)
2.64 (6.03)
2.95 (6.02)
0.23−2.32
0.47−1.43
0.75−1.99
0.43−1.34
0.49−1.96
0.79−1.61
0.84−4.69
1.01−2.54
1.06−2.11
0.88−3.73
1.35−6.84
1.49−4.60
0.78−5.48
2.24−8.80
1.91−7.38
2.51−9.27
3.79−15.4
9.13 0.70
12.40 0.25
7.56 0.46
10.78 0.64
5.31 0.44
7.34 0.13
3.45 0.11
3.47 0.36
5.98 0.15
5.50 0.28
2.49 0.16
3.83 0.19
3.09 0.14
1.87 0.05
2.35 0.03
1.53 0.06
0.83 0.07
8.3 1.1
4.2 1.1
6.5 1.4
5.6 1.0
8.3 1.6
3.2 1.2
10.6 1.0
6.1 1.2
5.2 0.7
22.7 1.1
17.4 1.0
23.5 0.9
18.3 1.2
12.8 1.1
19.6 0.8
15.7 1.1
19.0 1.4
13.7 8.9
6.3 2.1
4.4 5.80
8.0 5.1
5.3 4.3
3.3 1.2
13.1 2.7
4.5 5.6
6.4 1.9
29.4 5.4
19.7 6.1
27.7 5.4
21.3 4.5
13.3 2.6
19.2 1.2
16.0 2.9
19.5 6.2
h
h
9.3
12.1
12.1
hi
,
fh
267 ,
fh
267 ,
fh
298 ,
fh
300 ,
300 ,
fh
j
302
336 ,
12.1
fh
j
338
9.0
8.8
j
361
363 ,
fh
j
393
9.0
7.5
9.7
j
423
j
473
R = CH3CO (CH3CO + NO2 → Products)
f
k
298
2.67 (8.65)
3.13−11.8
0.48 0.02
0.56 0.05
0.31 0.02
0.44 0.03
16.27 0.84
15.35 0.91
16.03 1.74
16.24 3.46
15.64 1.76
12.64 3.59
l
f
k
l
336
3.26 (9.37)
6.60−15.4
a
b
1 Torr = 133.3 N m−2. Uncertainty stated as 1 standard error of the fit; estimated overall uncertainty in the determined bimolecular rate
c
coefficients is about 20% (see Figures 3 and 4) except in the CH3CO + NO reaction, where it was estimated to be about 50%. Average of
measured wall rates. Wall rate determined from y-axis intercept. 17 mm i.d. HalocarbonWax coated reactor and chlorine-lamp with CaF2 window
for photoionization (9.1 eV). Laser pulse energy was not determined during the experiment. Xe/sapphire for photoionization (8.44 eV) unless
otherwise stated. 16.5 mm uncoated Pyrex reactor. N-lamp with quartz window (7.1 eV). 17 mm reactor with PDMS coating. The rate coefficient
d
e
f
g
h
i
j
k
C
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Table 1. continued
l
obtained from a direct fit to the obtained radical signal. The result when the signal is processed taking into account the apparent production signal at
the radical mass (see text).
k(T)1 = k300K(T/300 K)n
(3)
k(T)2 = A exp[−EA/(RT)]
(4)
k(T)3 = A′(T/300 K)m exp[−Em/(RT)]
(5)
where T is temperature in kelvin, R is the gas constant, and
k
300K, A, A′, n, m, EA, and Em are empirical parameters obtained
from the fits.
The flow tube rate determinations contain several
uncertainty sources, which need to be taken into consideration
in analyzing the results. The uncertainties mainly arise from
uncertainties in reagent concentrations (e.g., uncertainties in
flow rates, vapor pressures of the pure compounds, calibration
of pressure sensors, etc.) but potentially from other sources too,
which were not observed during the experiments but which
would not be easily separable from the data obtained. For
example, possible contaminants in the reagent NO flow and the
generation of other radical products in the photolysis reaction 2
constitute additional potential, though minor, uncertainty
factors. By adding the different uncertainty contributions and
using the propagation of errors method, the overall
uncertainties of the determined reaction rate coefficients were
estimated (Table 1).
Figure 3. Determined rate coefficients of the ·CH2OH, CH3·CHOH,
and CH3·C(O) reactions with NO shown as a function of
temperature. The overall uncertainty of the obtained rate coefficients
was estimated as 20%, and the dashed lines indicate the variation of k
between these limits. Also included are the previous results by
Pagsberg et al.15 for the ·CH2OH + NO reaction, by Nesbitt et al.27 for
the closely related ·CD2OH + NO reaction, by Miyoshi et al.7 for the
CH3·CHOH + NO reaction, and by McDade et al.,17 Christey and
Voisey,28 and Sehested et al.16 for the CH3·C(O) + NO reaction.
RESULTS AND DISCUSSION
■
The determined R + NO rate coefficients and their derived
temperature parametrizations are presented in Table 1. The
rate coefficients increase with decreasing temperature (see
Figure 3 and Figure 4), common to a radical association
reaction of the type R + NO → R−NO* (+M) → RNO, with a
prereactive complex (R−NO*) on the reaction coordinate.25,26
Three different parametrizations of the temperature depend-
ence were used (eqs 3−5) in order to adequately account for
the observed dependence as well as to supply information for
different purposes. The CH3·C(O) + NO reaction was
investigated only at two different temperatures, and thus, the
temperature parametrizations were not derived.
parametrizations would be needed to describe the data
obtained. Furthermore, the three-parameter fitting procedure
returned unrealistic error limits (e.g., bigger uncertainty value
than the corresponding parameter value, which would imply a
negative rate coefficient), and thus, only the estimated overall
uncertainties for these fits are reported (Table 1). In addition,
the variation of the rate coefficients within these estimated
limits are indicated by dashed lines in Figures 3 and 4.
The activation energy (EA) of the Arrhenius equation (eq 4)
is not generally applicable for reactions with submerged barriers
on the potential energy surface (i.e., reactions with apparent
negative activation energies), but the Arrhenius parameters
were nevertheless determined in order to enable easier
comparison with previous results. As noted above, using only
two parameters resulted in poor expression for the determined
temperature dependences especially in the case of the
hydroxyalkyl reactions, with significant deviation from the
measured rate coefficients in the extremes of the experimental
temperature range. However, the temperature dependences can
be better described in the Arrhenius form by dividing the
experimental temperature range into two different regimes,
each with different parametrization. Thus, the obtained
hydroxyalkyl rate coefficients were arbitrarily divided into two
different temperature ranges which resulted in the para-
metrizations given in Table 1. The curvature and change in
the apparent activation energy potentially indicate another
reaction channel becoming important as the temperature
Due to significant curvature observed in the temperature
dependences of the determined rate coefficients [k(T)], the
commonly applied two-parameter expressions 3 and 4 cannot
fully describe the data obtained. In the hydroxyalkyl reactions
only the three-parameter eq 5 is able to return a good
description of the measured rate coefficients, whereas in the
·CH2OCH3 + NO reaction also the two-parameter expressions
produce reasonably good fits for the low pressure data. It
should be noted that although there is no real physical reason
for using the three-parameter expression [except the curvature
in the k(T)], and even by inspection it is clear that the
parameters obtained vary much for results that are comparable
in magnitude, yet the given parametrizations were found to give
the best description of the current results. Only the
parametrizations that were found to give a reasonable fit are
given in Table 1. For example, one Arrhenius parametrization is
able to describe the temperature dependence of the ·CH2OCH3
+ NO reaction at the lower density range quite well, and thus,
only one parametrization is given, whereas for the results
obtained at the higher density range with a broader
experimental temperature range, two different Arrhenius
D
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
photolysis wavelength, it is not possible to calculate the initial
radical concentrations used in these experiments. However, it is
possible to use indirect indicators of the concentration
produced, the concentration range being crucial for the rate
determination by the pseudo-first-order rate method. These
indicators include, for example, the linearity of the obtained
pseudo-first-order rate plots and the single-exponential decay
behavior of the radical signals observed, especially the radical
decays observed in absence of NO (see Figure 2). In addition,
the variation of the laser power in the experiments did not
cause noticeable changes in the obtained rate coefficients, and
thus, the results obtained in the current study are not affected
by second-order radical processes to any significant extent.
Products of the reactions were sought with different
photoionization energies, but no unambiguous product
identification could be made. The R + NO radical association
reaction is thought of leading to nitroso compounds (i.e.,
RNO), provided that the bath gas collisions are enough to
stabilize the nascent R−NO* collision complexes at the low
pressures of the current experiments. Indeed it seems that these
association products were formed in the studied reactions with
significant yields but were potentially photodissociated during
photoionization. This was indicated by the primary radical
signals (e.g., ·CH2OCH3 signals in the ·CH2OCH3 + NO
investigation), when higher ionization energies were applied
(i.e., higher than in normal rate measurements performed with
xenon and chlorine lamps, 8.44 and 9.1 eV, respectively),
resulting in “an equilibrium type of signal” that does not decay
to the prephotolysis background signal level during the reaction
time (cf. ref 23, a0 in Figure 5). This strongly suggests that a
Figure 4. Determined ·CH2OCH3 + NO reaction rate coefficients
shown as a function of temperature at two different bath gas densities
(Table 1). The density change of about a factor of 3 increases the
reaction rate coefficient roughly about 50%. The single value
determined at the highest density and shown with a star symbol has
been omitted from the fits. The overall uncertainty of the obtained rate
coefficients was estimated as 20%, and the dashed lines indicate the
variation of k between these limits for the results obtained at the
higher density range. Also included are the fits obtained for the
·CH2OH + NO and CH3·CHOH + NO reactions.
changes, but unfortunately with no products found (as will be
dealt more thoroughly below), this hypothesis cannot be
proven or rejected.
The thing that is striking from the obtained k(T) parameters
is the very steep temperature dependences, much stronger than
what were previously obtained for the same radicals in their
corresponding R + NO2 reactions.21 Furthermore, in the
hydroxyalkyl reactions the observed temperature dependences
are significantly stronger at colder temperatures, an observation
that could indicate pronounced reactivity on the reactor
surfaces in a cooler reactor. Though somewhat higher wall
rates were measured at cooler conditions, especially in the
CH3C·HOH + NO reaction where also the steepest temper-
ature dependence was observed (Table 1 and Figure 3), the
determined bimolecular rate coefficients cannot be explained by
increasing surface reactions as the wall rate was measured in
each experiment and is included in the rate coefficient analysis
(see Figure 2). Thus, the increase must be due to a gas-phase
process. Moreover, the same combination of radical precursor,
reactor, and photolysis production was used in our previous R
+ NO2 study21 with no indication of enhanced rate coefficients
at colder temperatures.
Figure 5. Schematic of the obtained radical signal explaining the fitting
routines used to analyze the CH3·C(O) radical decays. In the figure, a0
is the prephotolysis background signal level, [R]0 is the initial radical
concentration produced in the laser pulse, [R]t is the radical
concentration after reaction time t and is dependent on the pseudo-
first-order rate coefficient k(x)′ of the reaction. In the experiments a
baseline shift a2 was observed and thus a double-exponential function
was fitted to the CH3·C(O) radical signal (see text).
fragment with a similar mass as NO (i.e., m/z = 30 Th) is lost
from the parent ion (RNO+) subsequent to photoionization.
This type of result has been obtained for a few other R + NO
reactions measured in our setup, and the radical decay has been
returned to “a normal type of behavior” (i.e., R decay to the
prephotolysis background; Figure 2) by using lower ionization
energies. The nitroso compounds are generally expected to
have relatively low R−NO bond energies, which may result in
an equilibrium in the reaction (i.e., R + NO ↔ RNO) at
correspondingly low temperatures,3,23 though for the saturated
In the ·CH2OCH3 + NO reaction also a significant pressure
dependence was observed (see Figure 4), with the determined
rate coefficients increasing by about 35−65% with a factor of 3
increase in bath gas density, the relative increase being more at
higher temperatures. The investigated hydroxyalkyl reactions
did not show significant changes in the rates when the bath gas
helium pressure was varied at around 0.5−3.5 Torr.
Due to the missing absorption coefficients and quantum
yields for the radical precursor compounds at the 193 nm laser
E
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
radicals of the current study the equilibrium behavior is not
expected at least below 450 K.
order of microseconds, which could result in contribution of
excited species to the observed kinetics. In the current work the
reactions were studied in tens of milliseconds reaction time (see
Figure 2), with the first millisecond always omitted from the
analysis to exclude possible excited species contribution. Thus,
it is concluded that the discrepancy between the reported
values could result from the different radical production
method in the work of Pagsberg et al.15 in comparison with
the current investigation.
A similar type of “false-equilibrium” signal was observed in
the CH3·C(O) + NO radical reaction, again likely due to the
use of too high ionization energy (9.1 eV), which was
nevertheless necessary in order to obtain good enough signals
to perform the rate experiments. By analogy to other R + NO
reactions,3,23 the CH3·C(O) + NO reaction is not expected to
show equilibrium behavior at such low temperatures as used in
the current study. An example of the observed signal profile and
the two different methods of fitting are shown in Figure 5. In
the first method a single-exponential decay curve was fitted
directly to the raw ion signal without taking into consideration
the observed shift in the prephotolysis background (marked as
a2 in Figure 5). In the second method a double-exponential
function was fitted to the obtained signal by assuming that the
apparent formation rate (i.e., the reason why the signal does
not decay to the background) is due to a photofragmentation
process during ionization and thus has exactly the same time
behavior and can be fixed as k(1)′ = k(−1)′. The amplitudes of
the exponentials can be read from the obtained signal (i.e., [R]0
and a2 in Figure 5, for example), and by fixing the amplitudes
and constraining the reverse rate equal to the forward rate, the
corrected k(1)′ could be retrieved. This procedure led to about
25% higher rate coefficients. The rate coefficients according to
both procedures are given in Table 1.
A reaction mechanism analogous to that observed in the
·CH2OH + O2 reaction,29 with a prereactive complex on the
reaction coordinate and in the case of the ·CH2OH + NO
reaction leading to HCHO + HNO products (and to CH3CHO
+ HNO in CH3·CHOH + NO reaction), could explain the
apparent negative activation energies and the lack of pressure
dependence obtained. However, we sought for the correspond-
ing aldehyde species in each reaction by applying different
ionization energies but did not detect them. From previous
works we know that the RPIMS used in this work is fairly
sensitive to these carbonyl compounds, and thus, it is unlikely
that they were formed in these reactions with a significant yield.
We have not been able to detect previously the coproduct
HNO despite considerable effort invested, potentially due to
absence of suitable photoionization lines within our photo-
ionization setup, resulting in photodissociation in the photo-
ionizator (as was also discussed above for the potential organo
nitroso products). Another conceivable possibility for un-
successful detection could be that the aldehyde species were
fragmented due to excitation obtained from the exothermic R +
NO reactions. However, in the absence of detected reaction
products, the actual reaction mechanisms can only be
speculated.
The analogous reaction of the isotopologue ·CD2OH with
NO was studied as a function of pressure and temperature (p =
0.5−1.5 Torr of He, T = 230−373 K) by Nesbitt et al.27 using
discharge-flow/mass spectrometry apparatus and found a
pressure independent rate coefficient of k298K(·CD2OH +
NO) = (2.2
0.4) × 10−12 cm3 molecule−1 s−1. The value
determined at room temperature is in relatively good
agreement with the result obtained in this study and consistent
with the proposed complex association decomposition
mechanism, provided that the α-hydrogens do not take part
in the primary reaction step (i.e., only secondary kinetic isotope
effect affecting the rate coefficient). However, the temperature
dependence obtained by Nesbitt et al.27 is opposite to what is
found here (see Figure 3). Though the discrepancy between the
obtained results cannot be reconciled based on the available
documentation, it should be noted that a similar discrepancy
(i.e., positive vs negative apparent activation energy) was also
observed in the CH2OH + O2 reaction rate coefficients
determined by Nesbitt et al.,30 in contrast with more recent
experimental and theoretical work.29
The other hydroxyalkyl reaction investigated, CH3C·HOH +
NO, has been studied previously by Miyoshi et al.7 at 295 K
and at a few Torr of He. They used similar methods as in the
current study and obtained (2.41
0.60) × 10−11 cm3
molecule−1 s−1. The rate coefficient was determined only at
room temperature and is in reasonable agreement with the
k
298K(CH3·CHOH + NO) = (1.38
0.07) × 10−11 cm3
molecule−1 s−1 determined in the current work (see Table 1
and Figure 3).
The CH3·C(O) + NO rate coefficient has been determined
by Sehested et al.16 at 295 K and 1 atm of SF6, employing pulse
radiolysis/UV−vis absorption method, and by McDade et al.17
at 298 K and at a few Torr of He using similar LP-RPIMS
method as used in the current work. Sehested et al.16 obtained
k
295K(CH3·C(O) + NO) = (2.4 0.7) × 10−11 cm3 molecule−1
s−1, a much higher value than obtained in this work [k298K(CH3·
C(O) + NO) = (4.8 2.4) × 10−13 cm3 molecule−1 s−1 and
k
298K(CH3·C(O) + NO)CORRECTED = (5.6 2.8) × 10−13 cm3
molecule−1 s−1], whereas McDade et al.17 obtained k298K(CH3·
C(O) + NO) = (9.3 2.7) × 10−13 cm3 molecule−1 s−1, much
closer to the current result. In addition, a previous estimation of
the CH3·C(O) + NO rate coefficient was made by Christie and
Voisey28 by fitting to a complex reaction system and finding a
value of k298 K(CH3·C(O) + NO) = 3.32 × 10−13 cm3
molecule−1 s−1, in good agreement with the values determined
in the current study. By comparing the results obtained in the
current and previous works, it seems that the studies performed
with the pulse radiolysis radical generation method have
significantly overestimated the rate coefficients, potentially due
to the combined effects of complex reaction mixture and
relatively high radical concentration inherent with the
technique, as was already noted above in the case of ·
CH2OH + NO reaction. There are no previous results for the ·
CH2OCH3 + NO reaction, and thus, a similar comparison
Previous rate coefficient determinations for the reactions
investigated are scarce. The ·CH2OH + NO reaction was
studied previously by Pagsberg et al.15 at 298 K and at
atmospheric pressure of argon (Ar) using F· + CH3OH
reaction for radical generation and UV absorption for direct
detection of the ·CH2OH radical. They obtained (2.5 0.2) ×
10−11 cm3 molecule−1 s−1, a vastly different value than what is
determined in the current work (3.71
0.13) × 10−12 cm3
molecule−1 s−1. However, the pulse radiolysis technique used by
Pagsberg et al.15 creates free radicals in a less controllable and
unselective fashion than the direct laser photolysis used in this
work. Pagsberg et al. also employed much higher reagent (and
thus also radical) concentrations in their experiments and
correspondingly shorter reaction monitoring time-scale on the
F
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
cannot be performed. The determined rate coefficients are
presented in Figures 3 and 4 together with the previously
reported values.
The reactivity of the studied radicals toward NO is quite
different from their previously observed reactivity in the
corresponding R· + NO2 reactions21 (Table 2). All these
rate coefficients showed negative temperature dependence,
typical for a complex radical−radical association process. In the
·CH2OCH3 + NO reaction, also a pressure dependence was
observed, whereas in the CH3·C(O), CH3·CHOH, and
·CH2OH reactions the determined rate coefficients did not
indicate any significant pressure dependence. The reactivity
differences between the studied reactions were tentatively
proposed to result from different charge distributions at the
radical center carbon atom, based on the structures of the
radicals and their measured rate coefficients with NO. The
results of the study increase our knowledge on oxygenated
carbon-centered free radical reactions with NO, for which a
relatively scarce database exists up to date.
21
Table 2. Current R· + NO and Previous R· + NO2 Room
Temperature (T = 298 K) Rate Coefficients for R =
·CH2OH, CH3·CHOH, ·CH2OCH3, and CH3·C(O)
k298K(R· + NO)/10−12 cm3
k298K(R· + NO2)/10−12 cm3
radical R·
molecule−1s−1
molecule−1s−1
·CH2OH
3.3
89.5
74.8
78.5
28.7
CH3·CHOH
·CH2OCH3
CH3·C(O)
13.2
AUTHOR INFORMATION
Corresponding Author
*E-mail: matti.p.rissanen@helsinki.fi. Phone: +358458730170.
a
■
3.5
0.6
a
Value at about 0.6 Torr of helium bath gas; note that the ·CH2OCH3
Notes
+ NO reaction rate coefficient is pressure dependent.
The authors declare no competing financial interest.
radicals show less reactivity toward NO than NO2, only the
CH3·CHOH radical approaching the high reactivity observed in
its R· + NO2 reaction. Most notably the reactivity order is
different, being (in the R· + NO reactions) roughly k298K(CH3·
CHOH) > k298K(·CH2OCH3) ∼ k298K(·CH2OH) > k298K(CH3·
C(O)), whereas in the R· + NO2 reactions the reactivities at
298 K are21 k298K(·CH2OH) > k298K(·CH2OCH3) ∼ k298K(CH3·
CHOH) > k298K(CH3·C(O)). Especially the ·CH2OH and
·CH2OCH3 radicals are significantly less reactive toward NO
than NO2. As in the analogic R· + NO2 reactions, the CH3·
C(O) radical is the least reactive radical in the group, with
about 2 orders of magnitude smaller rate coefficient with NO
than with NO2.
As has been commonly found in previous carbon-centered
free radical reaction investigations, the reactivity differences
observed between the studied oxygenated radicals toward a
common reagent can likely be explained by the inductive effects
of the substituents connected to the radical center carbon atom,
enhancing or decreasing the rate of reaction by increasing or
decreasing the electron density in the radical center.31−33
Electropositive alkyl substitution at the radical carbon (e.g., the
extra methyl group in CH3·CHOH in comparison with
·CH2OH) enhances the R· + NO reaction rate, whereas with
the electronegative methoxy substitution in the radical carbon
(i.e., −OCH3 in ·CH2OCH3) the reaction rate does not change
much from the ·CH2OH + NO reaction, indicating that the
electronegative methoxy and hydroxyl substituents affect the
observed reactivity in R· + NO reaction about equally. The
CH3·C(O) radical is somewhat of an outlier as far as the
structure goes (Figure 1), and thus, the comparison might be
less meaningful. It is possible, however, that the much lower
reactivity is caused by the electronegative carbonyl group
pulling away the electron density from the radical center.
The measured R· + NO rate coefficients are comparable to
their corresponding R· + O2 rate coefficients,7,29,34 and thus,
the loss of these radicals from an atmospheric gas mixture is
dominated by reaction with O2, the reactions with NO and
NO2 being of minor importance.
REFERENCES
■
(1) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Crowley, J. N.; Hampson,
R. F.; Hynes, R. G.; Jenkin, M. E.; Rossi, M. J.; Troe, J. Evaluated
Kinetic and Photochemical Data for Atmospheric Chemistry: Volume
I - Gas Phase Reactions of Ox, HOx, NOx and SOx Species. Atmos.
Chem. Phys. 2004, 4, 1461−1738.
(2) Wayne, R. P. Chemistry of Atmospheres, 3rd ed.; Oxford University
Press, Inc.: New York, 2000.
(3) Miller, J. A.; Bowman, C. T. Mechanism and Modeling of
Nitrogen Chemistry in Combustion. Prog. Energy Combust. Sci. 1989,
15, 287−338.
(4) Miller, J. A.; Pilling, M. J.; Troe, J. Unravelling Combustion
Mechanisms through a Quantitative Understanding of Elementary
Reactions. Proc. Combust. Inst. 2005, 30, 43−88.
(5) Monks, P. S. Gas-phase Radical Chemistry in the Troposphere.
Chem. Soc. Rev. 2005, 34, 376−395.
(6) Sander, S. P.; Friedl, R. R.; Abbat, J. P. D.; Barker, J. R.;
Burkholder, J. B.; Golden, D. M.; Kolb, C. E.; Kurylo, M. J.; Moortgat,
G. K.; Wine, P. H.; et al. Chemical Kinetics and Photochemical Data
for Use in Atmospheric Studies: Evaluation Number 17. Publication
10-06; National Aeronautics and Space Administration, Jet Propulsion
Laboratory, California Institute of Technology: Pasadena, CA, 2011.
(7) Miyoshi, A.; Matsui, H.; Washida, N. Reactions of Hydroxyethyl
Radicals with Oxygen and Nitric Oxide. Chem. Phys. Lett. 1989, 160,
291−294.
(8) Grosjean, D. Atmospheric Chemistry of Alcohols. J. Braz. Chem.
Soc. 1997, 8, 433−442.
(9) Carr, S. A.; Blitz, M. A.; Seakins, P. W. Site-Specific Rate
Coefficients for Reaction of OH with Ethanol from 298 to 900 K. J.
Phys. Chem. A 2011, 115, 3335−3345.
(10) Wu, Y.; Xie, D.; Xue, Y. Ab Initio Studies for the
Photodissociation Mechanism of Hydroxyacetone. J. Comput. Chem.
2003, 24, 931−938.
(11) Blitz, M. A.; Heard, D. E.; Pilling, M. J. Study of Acetone
Photodissociation over the Wavelength Range 248−330 nm: Evidence
of a Mechanism Involving Both the Singlet and Triplet Excited States.
J. Phys. Chem. A 2006, 110, 6742−6756.
(12) Herrmann, F.; Oßwald, P.; Kohse-Hoinghaus, K. Mass
̈
Spectrometric Investigation of the Low-Temperature Dimethyl Ether
Oxidation in an Atmospheric Pressure Laminar Flow Reactor. Proc.
Combust. Inst. 2013, 34, 771−778.
(13) Good, D. A.; Francisco, J. S. Tropospheric Oxidation
Mechanism of Dimethyl Ether and Methyl Formate. J. Phys. Chem.
A 2000, 104, 1171−1185.
(14) Carr, S. A.; Still, T. J.; Blitz, M. A.; Eskola, A. J.; Pilling, M. J.;
Seakins, P. W.; Shannon, R. J.; Wang, B.; Robertson, S. H.
Experimental and Theoretical Study of the Kinetics and Mechanism
CONCLUSIONS
■
Several oxygenated free radical reactions with NO were studied
in low pressure flow tube experiments using photoionization
mass spectrometry for direct radical detection. The determined
G
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
of the Reaction of OH Radicals with Dimethyl Ether. J. Phys. Chem. A
2013, 117, 11142−11154.
(15) Pagsberg, P.; Munk, J.; Anastasi, C.; Simpson, V. J. Reaction of
CH2OH with O2, NO, and NO2 at Room Temperature. J. Phys. Chem.
1989, 93, 5162−5165.
(16) Sehested, J.; Christensen, L. K.; Nielsen, O. J.; Wallington, T. J.
Absolute Rate Constants for F + CH3CHO and CH3CO + O2,
Relative Rate Study of CH3CO + NO, and the Product Distribution of
the F + CH3CHO Reaction. Int. J. Chem. Kinet. 1998, 30, 913−921.
(17) McDade, C. E.; Lenhardt, T. M.; Bayes, K. D. The Rate of
Reaction of Acetyl and Benzoyl Radicals with O2. J. Photochem. 1982,
20, 1.
(18) Goldemberg, J.; Guardabassi, P. Are Biofuels a Feasible Option?
Energy Policy 2009, 37, 10−14.
(19) Kirstine, W. V.; Galbally, I. E. The Global Atmospheric Budget
of Ethanol Revisited. Atmos. Chem. Phys. 2012, 12, 545−555.
(20) Semelsberger, T. A.; Borup, R. L.; Greene, H. L. Dimethyl Ether
(DME) as an Alternative Fuel. J. Power Sources 2006, 156, 497−511.
(21) Rissanen, M. P.; Arppe, S. L.; Timonen, R. S. Kinetics of Several
Oxygenated Carbon-Centered Free Radical Reactions with NO2. J.
Phys. Chem. A 2013, 117, 3902−3908.
(22) Eskola, A. J.; Timonen, R. S. Kinetics of the Reactions of Vinyl
Radicals with Molecular Oxygen and Chlorine at Temperatures 200-
362 K. Phys. Chem. Chem. Phys. 2003, 5, 2557−2561.
(23) Rissanen, M. P.; Amedro, D.; Krasnoperov, L.; Marshall, P.;
Timonen, R. S. Gas Phase Kinetics and Equilibrium of Allyl Radical
Reactions with NO and NO2. J. Phys. Chem. A 2013, 117, 793−805.
(24) Smith, I. W. M. The Temperature-Dependence of Elementary
Reaction Rates: Beyond Arrhenius. Chem. Soc. Rev. 2008, 37, 812−
826.
(25) Mozurkewich, M.; Benson, S. W. Negative Activation Energies
and Curved Arrhenius Plots. 1. Theory of Reactions over Potential
Wells. J. Phys. Chem. 1984, 88, 6429−6435.
(26) Troe, J. The Colourful World of Complex-Forming Bimolecular
Reactions. J. Chem. Soc. Faraday Trans. 1994, 90, 2303−2317.
(27) Nesbitt, F. L.; Payne, W. A.; Stief, L. J. Kinetic Studies of the
Hydroxymethyl Radical with NO and NO2. J. Phys. Chem. 1989, 93,
5158−5161.
(28) Christie, M. I.; Voisey, M. A. Reactions of Acetyl Radicals. Part
2. Reaction of Acetaldehyde with Nitrogen Dioxide in Presence of
Nitric Oxide. Trans. Faraday Soc. 1967, 63, 2702.
(29) Schocker, A.; Uetake, M.; Kanno, N.; Koshi, M.; Tonokura, K.
Kinetics and Rate Constants of the Reaction CH2OH + O2 → CH2O
+ HO2 in the Temperature Range of 236-600 K. J. Phys. Chem. A 2007,
111, 6622−6627.
(30) Nesbitt, F. L.; Payne, W. A.; Stief, L. J. Temperature
Dependence for the Absolute Rate Constant for the Reaction
CH2OH + O2 → HO2 + H2CO from 215 to 300 K. J. Phys. Chem.
1988, 92, 4030−4032.
(31) Marston, G.; Monks, P. S.; Wayne, R. P. In General Aspects of the
Chemistry of Radicals; Alfassi, Z. B., Ed.; John Wiley & Sons Ltd.: New
York, 1999; pp 429−471.
(32) Paltenghi, R.; Ogryzlo, E. A.; Bayes, K. D. Rates of Reaction of
Alkyl Radicals with Ozone. J. Phys. Chem. 1984, 88, 2595−2599.
(33) Timonen, R. S.; Gutman, D. Kinetics of the Reactions of
Methyl, Ethyl, Isopropyl, and tert-Butyl Radicals with Molecular
Chlorine. J. Phys. Chem. 1986, 90, 2987−2991.
(34) Rosado-Reyes, C. M.; Francisco, J. S.; Szente, J. J.; Maricq, M.
M.; Frøsig Østergaard, L. Dimethyl ether oxidation at elevated
temperatures (295-600 K). J. Phys. Chem. A 2005, 109, 10940−10953.
H
DOI: 10.1021/acs.jpca.5b01027
J. Phys. Chem. A XXXX, XXX, XXX−XXX