Experimental and theoretical understanding of the gas phase oxidation of atmospheric amides with OH radicals: Kinetics, products, and mechanisms

Article abstract of DOI:10.1021/jp503759f

Atmospheric amides have primary and secondary sources and are present in ambient air at low pptv levels. To better assess the fate of amides in the atmosphere, the room temperature (298 ± 3 K) rate coefficients of five different amides with OH radicals were determined in a 1 m³ smog chamber using online proton-transfer-reaction mass spectrometry (PTR-MS). Formamide, the simplest amide, has a rate coefficient of (4.44 ± 0.46) × 10^-12 cm³ molec^-1 s^-1 against OH, translating to an atmospheric lifetime of ～1 day. N-methylformamide, N-methylacetamide and propanamide, alkyl versions of formamide, have rate coefficients of (10.1 ± 0.6) × 10^-12, (5.42 ± 0.19) × 10^-12, and (1.78 ± 0.43) × 10^-12 cm³ molec^-1 s^-1, respectively. Acetamide was also investigated, but due to its slow oxidation kinetics, we report a range of (0.4-1.1) × 10^-12 cm³ molec^-1 s^-1 for its rate coefficient with OH radicals. Oxidation products were monitored and quantified and their time traces were fitted using a simple kinetic box model. To further probe the mechanism, ab initio calculations are used to identify the initial radical products of the amide reactions with OH. Our results indicate that N-H abstractions are negligible in all cases, in contrast to what is predicted by structure-activity relationships. Instead, the reactions proceed via C-H abstraction from alkyl groups and from formyl C(O)-H bonds when available. The latter process leads to radicals that can readily react with O₂ to form isocyanates, explaining the detection of toxic compounds such as isocyanic acid (HNCO) and methyl isocyanate (CH₃NCO). These contaminants of significant interest are primary oxidation products in the photochemical oxidation of formamide and N-methylformamide, respectively.

Full text of DOI:10.1021/jp503759f

Article
pubs.acs.org/JPCA
Experimental and Theoretical Understanding of the Gas Phase
Oxidation of Atmospheric Amides with OH Radicals: Kinetics,
Products, and Mechanisms
Nadine Borduas,† Gabriel da Silva,‡ Jennifer G. Murphy,*,† and Jonathan P. D. Abbatt*,†
†Department of Chemistry, University of Toronto, 80 St. George Street, Toronto, Ontario M5S 3H6, Canada
‡Department of Chemical and Biomolecular Engineering, The University of Melbourne, Victoria 3010, Australia
S
* Supporting Information
ABSTRACT: Atmospheric amides have primary and secondary sources and are present in
ambient air at low pptv levels. To better assess the fate of amides in the atmosphere, the
room temperature (298 3 K) rate coefficients of five different amides with OH radicals
were determined in a 1 m3 smog chamber using online proton-transfer-reaction mass
spectrometry (PTR-MS). Formamide, the simplest amide, has a rate coefficient of (4.44
0.46) × 10−12 cm3 molec−1 s−1 against OH, translating to an atmospheric lifetime of ∼1
day. N-methylformamide, N-methylacetamide and propanamide, alkyl versions of
formamide, have rate coefficients of (10.1 0.6) × 10−12, (5.42 0.19) × 10−12, and
(1.78 0.43) × 10−12 cm3 molec−1 s−1, respectively. Acetamide was also investigated, but
due to its slow oxidation kinetics, we report a range of (0.4−1.1) × 10−12 cm3 molec−1 s−1
for its rate coefficient with OH radicals. Oxidation products were monitored and quantified
and their time traces were fitted using a simple kinetic box model. To further probe the
mechanism, ab initio calculations are used to identify the initial radical products of the
amide reactions with OH. Our results indicate that N−H abstractions are negligible in all cases, in contrast to what is predicted
by structure−activity relationships. Instead, the reactions proceed via C−H abstraction from alkyl groups and from formyl
C(O)−H bonds when available. The latter process leads to radicals that can readily react with O2 to form isocyanates, explaining
the detection of toxic compounds such as isocyanic acid (HNCO) and methyl isocyanate (CH3NCO). These contaminants of
significant interest are primary oxidation products in the photochemical oxidation of formamide and N-methylformamide,
respectively.
carbon capture facility in Norway.10 The mechanism of gas-
INTRODUCTION
■
phase oxidation of amines to amides has also been investigated
Amides are emitted directly to the atmosphere from biological
through ab initio calculations.11,12
sources as well as from industrial processes and are present at
Only a few studies have looked at the atmospheric fate of
pptv levels in the atmosphere.1 In particular, tobacco smoke is a
amides. Chakir and co-workers measured the cross-section of
source of formamide, acetamide, and propanamide, presumably
N,N-dialkylated amides and determined that its value was
formed during the combustion process.2 Mixing ratios of over
smaller than or equal to 3 × 10−20 cm2 molec−1 beyond 270
400 pptv of N,N-dimethylformamide have been observed near
nm, indicating that photolysis is a negligible sink for amides.13
waste and sewage operations and an emission rate of 10 mg/kg
Koch et al. were the first to report rate coefficients for the
of N,N-dibutylformamide has been measured from charbroiling
reaction of OH radicals with amides, and they specifically
hamburgers.3,4 N,N-Dimethylformamide has also been detected
looked at N-alkylated acetamides and propanamides, determin-
from textile floor coverings from a polyamide composite.5
ing values between 5 and 14 × 10−12 cm3 molec−1 s−1.14 1-
Amides have been detected in ambient particles, biomass
Methyl-2-pyrrolidinone, 2-chloro-N-isopropylacetanilide, and
N-ethylperfluorobutyramide have been the subject of atmos-
burning aerosols and fogwater, as well.6−8
Amides have secondary as well as primary sources to the
pheric fate studies where rate coefficients with OH and NO3
atmosphere. For example, they can be formed via amine
radicals and with ozone were reported; OH reactivity was found
to dominate the fate of these amides.15−17 Solignac et al.
oxidation by OH radicals. Since amines also have biogenic and
anthropogenic sources to the atmosphere, including their use in
large quantities in carbon capture and storage (CCS)
studied the kinetics of OH and Cl radicals with N-alkylated
technology, they may represent a significant source of amides
Special Issue: Mario Molina Festschrift
to ambient air.9 This secondary source is of particular concern
since amides are typically longer-lived species than amines.9
Received: April 16, 2014
Indeed, a recent report by Zhu et al. detected formamide,
potentially from amine oxidation, from an industrial scale
Revised: July 11, 2014
© XXXX American Chemical Society
A
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
formamides and acetamides and later determined the rate
coefficients of OH radicals with cyclic amides.18,19 In addition,
Barnes et al. reviewed these results and completed a product
study of the smaller amides, identifying isocyanates as potential
oxidation products.20 NO3 radicals react with N,N-dimethylated
amides with rate coefficients on the order of 10−14 cm3 molec−1
s−1, making the sink of amides toward NO3 radicals competitive
with OH radicals only at night.21 These kinetics studies have
been recently reviewed by Nielsen et al. in the context of the
impact of amines on CCS.9 Furthermore, amides may partition
to aerosols but whether this sink is competitive with gas phase
oxidation is unknown.
Despite the work summarized above, the atmospheric
behavior and reactivity of formamide, the smallest and most
volatile of amides, has yet to be investigated. To address this
gap, five different amides, (i) formamide, (ii) N-methylforma-
mide, (iii) acetamide, (iv) N-methylacetamide, and (v)
propanamide, were investigated to elucidate their reactivity
trends toward the OH radical, their atmospheric lifetimes and
ultimately their atmospheric fate. Relative rate coefficients were
measured in a smog chamber using online mass spectrometry
and the oxidation products’ traces were fitted using simple box
model kinetics. The mechanism was further evaluated using ab
initio calculations to determine the location of the H-
abstraction and fate of the resultant radicals.
setup and on the time scale of the oxidation experiment. Most
amides, with the exception of acetamide, experienced no decay
under lit conditions, in the absence of hydrogen peroxide and
on the time scale of the oxidation experiments. Following the
establishment of a stable amide signal inside the chamber, pure
air from an AADCO 737-series generator was bubbled through
an aqueous solution of 30% hydrogen peroxide at a rate of 500
sccm. Approximately 15 min later, the UVB lights surrounding
the chamber were turned on to generate OH radicals in
concentrations of approximately 2 × 107 molecules cm−3,
inferred from the loss rate of n-butanol. Hydrogen peroxide was
continuously added to the chamber to keep the supply of OH
radical precursor approximately constant during the experi-
ment. Experiments were conducted under dry (<1% RH) and
NOx-free conditions.
HNCO was detected as an oxidation product and was
quantified using PTR-MS (detected at m/z 44), ion
chromatography (IC), and a homemade HNCO source. A
flow of HNCO was generated by heating cyanuric acid inside a
stainless steel tube to 250 °C, controlled by a thermocouple,
while passing a flow of 200 sccm of nitrogen over the headspace
of the tube. A dilution flow was added to the headspace flow to
avoid high gas-phase concentrations of HNCO and, thus,
polymerization of HNCO in the tubing. The source showed
variability in its HNCO permeation rate from week to week, so
the PTR-MS was independently calibrated by bubbling HNCO
through bubblers filled with deionized water at a known rate
and for a known amount of time and quantified by IC (see
Figure S1). Further experimental details are available in the
Supporting Information. We ruled out any interference from
EXPERIMENTAL AND THEORETICAL METHODS
■
1. Laboratory Study. The relative rate kinetics of five
amides were measured using a previously described exper-
imental setup.22 Briefly, a proton-transfer-reaction mass
spectrometer (PTR-MS) (Ionicon Analytik GmbH, Innsbruck
Austria)23 was connected to a 1 m3 Teflon FEP film (American
Durafilm) bag (Ingeniven) mounted on a Teflon-coated frame,
surrounded by a total of 12 UVB lamps (Microlites Scientific).
The chamber did not contain a fan. The PTR-MS was
connected to the chamber by a ∼1 m long 1/16″ ID silco-
coated steel inlet (Ionicon’s standard transfer line) heated to 50
°C linked to an FEP tubing that sampled from the middle of
the chamber. The flow through the inlet of the PTR-MS was
100 sccm. These inlet conditions are similar to conditions
optimized by Gylestam et al. for the detection of isocyanates by
PTR-MS.24 The signals of the compounds of interest were
normalized to m/z 21 corresponding to the H318O+ ion.
The amides were purchased from Sigma-Aldrich and used as
is. The amide calibrations on the PTR-MS were accomplished
by injecting known amounts of an aqueous solution of the
analyte into a glass tube connected to the chamber. A flow of
pure air passed through the glass tube adding the amide to the
chamber and resulting in mixing ratios ranging from 40 to 300
ppbv. Amides were all monitored at their respective protonated
molecular ion m/z and all had linear calibrations with
sensitivities ranging from 14 to 18 ncps/ppbv (see Table S1
and Equation S1).
+
CO2 (also detected at m/z 44), since ambient CO2
concentrations of ∼380 ppmv do not produce any significant
signal at m/z 44 in the PTR-MS. Control experiments involving
H2O2 and UVB light, but in the absence of a gas-phase amide,
showed production of HNCO, implying a background signal
for this weak acid in the chamber. This background was
accounted for in the yield calculation by averaging the final
HNCO signal of more than 10 control experiments and
subtracting this value (ranging from 8 to 40% of the total
signal) from the final HNCO signal in the amide oxidation
experiments. The yield reported is the quotient of moles of
product produced (corrected by its respective background
signal) divided by the moles of amide that decayed away at the
end of the reaction, that is, when the lights surrounding the
chamber were turned off. This method of calculating yields was
verified by plotting the appearance of product versus the loss of
starting material and taking the slope of that plot as the yield as
well as verified by the kinetic model fits. The yields reported
represent the range measured during different experiments.
Methyl isocyanate was also detected as an oxidation product
at m/z 58. To correct for its decay, the methyl isocyanate time
trace was subtracted by 4% of the n-butanol signal at m/z 57, a
percentage that represents the natural abundance of 13C in n-
butanol, which has an isobaric interference with methyl
isocyanate. Formamide (m/z 46) was corrected by subtracting
2% of the acetaldehyde signal at m/z 45 to account for its 13C
contribution to m/z 46.
A typical oxidation experiment consisted of injecting a
known amount of an aqueous solution of amide into the dark 1
m3 chamber. n-Butanol was added as the reference compound
and detected by the PTR-MS at m/z 57, corresponding to the
loss of water from the protonated molecular ion. Compared to
amines,22 amides were better behaved in the smog chamber;
little to no conditioning of the walls was required and the
amides were detected by the PTR-MS with quick response
times. Stable amide signals could also be monitored for hours,
further demonstrating their stability within the experimental
Following each oxidation experiment, the chamber was
purged at least overnight with pure air. The chamber was also
cleaned on a weekly basis by bubbling 30% hydrogen peroxide
into an amide-free chamber with the UVB lights turned on for
several hours. A small background signal of formamide and
HNCO was observed during these control experiments,
B
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Figure 1. Plot of eq 1 for formamide (A), N-methylformamide (B), N-methylacetamide (C), and propanamide (D), where the linear regression is
forced through the origin.
suggesting a contribution of wall chemistry within this
experimental set up. The product yields reported have been
corrected using these background signals.
D). Note that n-butanol serves as a good reference compound
as it does not react with amides, has no observable wall loss,
and is detected at m/z 57 by the PTR-MS, which does not
interfere with other signals of interest.
Rate coefficients were determined based on a minimum of
three reproducible experiments for formamide (4.44 0.46) ×
2. Ab Initio Study. The mechanism of OH radicals reacting
with the five studied amides, as well as the reaction of select
product radicals with O2, was investigated using computational
ab initio methods. Structures were optimized using the M06-2X
density functional with the 6-31G(2df,p) basis set.25 Sub-
sequent higher-level energies were obtained using the G3X-K
composite theoretical method.26 This method has been
designed specifically for accurate thermochemical kinetics,
and utilizes the M06-2X/6-31G(2df,p) structures in a series
of single point energy calculations with basis sets of
incrementally decreasing size, from Hartree−Fock theory to
perturbation theory (MP2, MP3, MP4) and, ultimately, to
coupled cluster theory (CCSD(T)). Energies calculated at
these levels are combined with empirical scaling terms to arrive
at the final G3X-K energy. Energies quoted here are 0 K
enthalpies (i.e., electronic energy + zero point energy) and are
expected to be accurate to within 1 kcal mol−1 on average.26 All
calculations were performed using the Gaussian 09 code.27
10−12 cm3 molec−1 s−1, N-methylformamide (10.1
10−12 cm3 molec−1 s−1, N-methylacetamide (5.42
0.6) ×
0.19) ×
10−12 cm3 molec−1 s−1, and propanamide (1.78 0.43) × 10−12
cm3 molec−1 s−1. See Figure 1A−D, for typical relative rate
plots for the four amides studied. Furthermore, forcing the
linear regression through the origin does not change the slope
by more than 5%, a value consistently less than the reported
standard deviation of the experiments.
Barnes et al. used an FTIR technique to study the kinetics of
other amides but reported interference in the method when
looking at formamide.20 Consequently, this study represents
the first reported experimental rate coefficient for formamide
and compares well with Barnes’ predicted value of 4 × 10−12
cm3 molec−1 s−1. In addition, our value for N-methylformamide
lies within the uncertainty of Barnes et al.’s value of (8.6 2.4)
× 10−12 cm3 molec−1 s−1.20 N-Methylacetamide’s reaction with
OH radicals has been previously investigated, once by Koch et
al., (5.2 0.19) × 10−12 cm3 molec−1 s−1, and again by Barnes
RESULTS AND DISCUSSION
■
1. Kinetics Results. Upon exposure to OH radicals inside
et al., (11
3) × 10−12 cm3 molec−1 s−1.14,20 The value
the 1 m3 smog chamber, the amides under study exhibited
obtained in this relative rate study closely matches Koch et al.’s
value, measured by absolute rate kinetics using flash photolysis
coupled to resonance fluorescence. This study is also the first to
report the rate coefficient of propanamide with OH radicals.
Since its rate coefficient is significantly slower than the
reference compound, a slightly more scattered rate plot was
observed (Figure 1D).
The uncertainty reported with the rate coefficients represents
the standard deviation of multiple experiments and is less than
a 10% error for formamide, N-methylformamide, and N-
pseudo-first order decay kinetics at room temperature (298
3
K) and atmospheric pressure. The relative rate method28
allowed for the determination of their respective bimolecular
rate coefficients using n-butanol as the reference compound.
Accurate knowledge of the rate coefficient of n-butanol with
OH radicals, (8.86
0.85) × 10−12 cm3 molec−1 s−1, is
necessary to employ eq 1.29 A plot of the natural logarithm of
the normalized amide decay signal as a function of the natural
logarithm of the normalized n-butanol decay signal yields the
ratio of both rate coefficients, kamide+OH/kbutanol+OH (Figure 1A−
C
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
methylacetamide. An additional uncertainty arises from the rate
coefficient of the reference compound n-butanol with OH
radicals, and consequently, a total uncertainty closer to 20%
should be considered when using these rate coefficients in
model evaluations. Propanamide’s rate coefficient is approx-
imately five times slower than the reference compound n-
butanol, which leads to an uncertainty of 24%.
Methyl isocyanate was detected solely in the oxidation of N-
methylformamide by OH radicals. This product is of major
health concern and was responsible for the death of thousands
of residents in Bhopal, India during a methyl isocyanate leak
from a nearby pesticide plant in 1984.33 Methyl isocyanate was
previously detected by FTIR from the oxidation of N-
methylformamide but the product yield was not quantified.20
In this study, we quantify methyl isocyanate by assuming it has
the same sensitivity in the PTR-MS as the average of the
amides quantified (17.1 npcs/ppbv; see Table S1).
We employ ab initio calculations to assist in understanding
the kinetics of the OH radical reactions with the amides
investigated here, as well as the mechanism of isocyanic acid
and methyl isocyanate formation. Results are presented for the
five amide + OH radical systems, so as to first understand the
radical products arising from these reactions and to gain insight
into the differences measured in their kinetics. Following this
first step, the secondary reactions of key radical products with
O2 are examined, shedding light on the first generation
products arising from the photochemical oxidation of these
amides.
a. Formamide. Figure 2A shows the oxidative decay of
formamide and the production of HNCO. The formamide
decay is well captured by the pseudo-first-order fit (red line in
Figure 2A). The yield of HNCO was measured to be 17−19%
in a set of six reproducible experiments. When fitting the
HNCO trace with Equation S8 in a kinetic box model, we find
that a yield (γ2) of 25% best fits the data, similar to the
experimental value. Note that HNCO has no appreciable gas-
phase loss rate to oxidants within the time frame of these
oxidation experiments. Indeed, its ambient temperature lifetime
with OH radicals, extrapolated from high-temperature data, is
on the order of decades.34 We consequently do not consider a
sink for HNCO when fitting its time trace. To further explore
the site of H-abstraction on formamide and the mechanism of
HNCO formation from formamide, we turn to ab initio
calculations.
Acetamide was also investigated under identical conditions
but since its oxidative decay was difficult to reproduce, in part
due to a slow reaction and an unsteady signal inside the
chamber, a large uncertainty on the rate coefficient was
obtained. Based on three experiments, an average rate
coefficient of (0.77
0.35) × 10−12 cm3 molec−1 s−1 was
measured for acetamide (see Figure S3 for a typical plot). The
previously published value by Barnes et al. of (3.5 1.0) ×
10−12 cm3 molec−1 s−1 is quite high compared to our
experimental results.20 We choose to report a range of (0.4−
1.1) × 10−12 cm3 molec−1 s−1 for acetamide to take into
consideration the large uncertainty associated with this
measurement.
⎛
⎜
⎝
⎞
⎟
⎠
⎛
⎜
⎝
⎞
⎟
⎠
kamide+OH
kbutanol+OH
[amide]0
[butanol]0
ln
=
ln
[amide]
[butanol]
(1)
t
t
2. Reaction Products and Mechanisms. The exper-
imental decays of formamide, N-methylformamide, N-methyl-
acetamide, and propanamide, and the production of their
respective oxidation products were fitted using simple kinetic
box model equations (see Supporting Information for full
description of equations). The yields of the oxidation products
were also quantified (see Scheme 1). Because the reaction of
Scheme 1. Product Identification and Quantification of
Amide Oxidation by OH Radicals in a 1 m3 Teflon FEP Film
Bag and in the Absence of NOx Based on Experimental and
a
Theoretical Evidence
In Figure 3, an energy diagram is provided for the formamide
+ OH radical reaction, where abstraction can take place from
either the formyl C−H site (right-hand side, Figure 3) or from
the amide N−H site (left-hand side). In both cases, reaction
initially proceeds via the formation of a prereaction complex,
followed by an H-abstraction transition state that exothermi-
cally leads to water and the corresponding CNOH2 radical
product. For N−H abstraction, the overall reaction is only
mildly exothermic (ca. 3.08 kcal mol−1), reflecting the very high
dissociation energy for this bond (116 kcal mol−1).
Accordingly, the transition state for N−H abstraction is high
in energy, 6.26 kcal mol−1 above the reactants, and abstraction
from this site will proceed very slowly and is unlikely to
contribute in any substantial way to the formamide + OH
reaction. The C−H bond in formamide is considerably weaker
than the N−H bond (bond dissociation energy of about 93 kcal
mol−1), and abstraction at this site is therefore much more
facile. Here, the abstraction transition state lies just above the
energy of the reactants (0.16 kcal mol−1), and we suggest that
this reaction almost exclusively accounts for the formamide +
OH products, which would be the NH2CO radical (+H2O).
When considering the first generation products for
formamide, the NH2CO radical appears to be the exclusive
reaction product, and in the atmosphere it will readily associate
with O2 to yield a peroxyl radical, NH2C(O)O2. An energy
diagram for this reaction is provided in Figure 4. We see that
a
The yields represent the range measured during different experi-
ments.
acetamide was slower, we were unable to reproducibly quantify
the yields of its oxidation products. HNCO was detected under
all experimental conditions as a primary product from
formamide and possibly as a primary or secondary product
from the other amides investigated. HNCO is toxic through
protein carbamylation and so exposure levels through biomass
burning and diesel exhaust are currently of concern.30,31
A
recent report evaluating HNCO concentrations at three
different urban and regional sites in the U.S. found mixing
ratios ranging from <0.003 to 1.2 ppbv and suggest the
presence of both primary and secondary sources to explain
observed diurnal profiles.32
D
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Figure 2. (A) Oxidative decay of formamide and formation of its major primary product, HNCO. The dots represent the 10 s PTR-MS data and the
solid red lines represent the exponential fits (Equations S4 and S8 (where γ2 = 25%) for formamide and HNCO, respectively). (B) Oxidative decay
of N-methylformamide and its products, methyl isocyanate, formamide, and HNCO. The dots represent 10 s PTR-MS data and the solid red lines
represent the exponential fits (Equations S4 and S8 (where γ2 = 50%), S6 (where γ1 = 15%), S9 (where γ1γ2 = 15%), for N-methylformamide, methyl
isocyanate, formamide, and HNCO, respectively).
Figure 3. Theoretical energy diagram for formamide + OH. Energies are 0 K enthalpies in kcal mol−1, at the G3X-K level of theory.
O2 addition to NH2CO is exothermic by 36.89 kcal mol−1,
providing considerable excess vibrational excitation to the
peroxyl radical intermediate. Bath gas collisions would be
expected to remove this excess energy, providing a thermalized
peroxyl radical that could then be lost by bimolecular reactions
with NO and other species. Instead, however, the α-amino
group provides a low-energy avenue to unimolecular reaction
before collisional deactivation can proceed, as recently
discovered for α-aminoalkylperoxyl radicals.35,36 Here, the
HO2 radical is expelled in a concerted process via the transition
state shown as an inset in Figure 4. This transition state energy
is 19 kcal mol−1 above the peroxyl radical and a substantial
17.92 kcal mol−1 below the reactants; the vibrationally excited
peroxyl radical adduct will initially have more than enough
energy to surmount this barrier. Following HO2 loss, a weak H-
bonded adduct is formed, which can dissociate along a
barrierless potential to yield isocyanic acid (+HO2). This
process therefore explains the experimental observation of
E
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
isocyanate in Figure 2B does not consider any sink terms, as its
reactivity toward OH radicals is estimated to be ∼10−13 cm3
molec−1 s−1 by the EPI Suite model and would account for <1%
difference in the fit.40 Formamide (fit of 15%) and HNCO were
also detected as oxidation products (Figure 2B). Note that in
Figure 2B, HNCO is treated as a secondary product arising
solely from the oxidation of formamide (Equation S9).
Figure 5 depicts a theoretical energy diagram for the N-
methylformamide reaction with OH, proceeding via abstraction
at either of the two unique C−H sites. Abstraction of an N−H
hydrogen is again energetically uncompetitive (with transition
state 4.03 kcal mol−1 above the reactants) and is shown only in
the Supporting Information (Figure S6). We observe in Figure
5 that the OH radical can abstract a H atom from either carbon,
leading to the CH2NHCHO (left-hand side) or CH3NHCO
(right-hand side) radicals. In both cases, the transition states lie
below the reactants in energy, at −1.30 kcal mol−1 and −0.67
kcal mol−1, for the respective alkyl and formyl abstractions. This
result is consistent with the rate coefficient of N-methyl-
formamide being more than twice that of formamide.
Considering the uncertainty of the calculations reported here,
we cannot definitively discriminate between the CH2NHCHO
and the CH3NHCO radical products. As such, we assign
theoretical yields of 50% to both products as a rough initial
estimate (see Scheme 1).
Energy diagrams for the reaction of O2 with the respective
CH2NHCHO and CH3NHCO radicals are shown in Figure 6.
For the CH2NHCHO radical (top of Figure 6), the resultant
peroxyl radical can eliminate HO2 to yield an imine, but the
required barrier heights are at around the energy of the
entrance channel, and the initially excited peroxyl radical will be
rapidly deactivated to below the required thresholds.
Interestingly, the high barrier for this process arises from the
reaction thermodynamics, with the conjugated imine product
CH2NCHO being comparatively unfavorable. Note also from
Figure 6 that an intramolecular abstraction from the formyl
group can take place, at 4.44 kcal mol−1 below the reactants,
forming a radical intermediate that can ultimately dissociate to
HNCO and CH2OOH (although again at above the reactant
energy). Indeed, HNCO was detected experimentally in <5%
yield, and may have two gas phase sources within these
experiments: a primary product from the reaction of
CH2NHCHO + OH (as discussed above) and a secondary
Figure 4. Theoretical energy diagram for formamide’s formyl radical +
O2. Energies are 0 K enthalpies in kcal mol−1, at the G3X-K level of
theory.
isocyanic acid as the major primary oxidation product of
formamide. Based on the reaction energetics and the known
chemically activated reaction kinetics of similar peroxyl
radicals,11,37,38 we expect almost quantitative production of
HNCO + HO2 upon reaction of formamide with OH at
tropospheric conditions and, therefore, suggest a theoretical
yield of 100% (see Scheme 1).
The experimentally measured HNCO yield of 17−19% only
accounts for a fraction of the mass balance (Scheme 1). It is
therefore possible that surface interactions of HNCO may be
affecting its concentration in the gas phase. Perhaps HNCO
uptake onto particles or surfaces is also taking place in the
chamber. Alternatively, quenching of the NH2C(O)O2 radical
could be an appreciable process, in which case bimolecular
chemistry is expected to lead to the carbonyloxyl radical
NH2C(O)O, which would likely dissociate rapidly to CO2 and
the aminyl radical NH2.39 Further investigation of HNCO’s
heterogeneous chemistry is currently ongoing in our laboratory
to address this discrepancy.
b. N-Methylformamide. The progress of the reaction of N-
methylformamide with OH radicals is illustrated in Figure 2B.
Methyl isocyanate appears to be the major product and was
measured in 36−40% yield. The fit using Equation S8 with a γ2
= 50% seems to reproduce the data well. The fit for methyl
Figure 5. Theoretical energy diagram for N-methylformamide + OH. Energies are 0 K enthalpies in kcal mol−1, at the G3X-K level of theory. The
N−H abstraction requires a significant barrier (4.03 kcal mol−1) and is shown in Figure S6.
F
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Figure 6. Theoretical energy diagrams for N-methylformamide’s methyl radical + O2 (top) and formyl radical + O2 (bottom) reactions. Energies are
0 K enthalpies in kcal mol−1, at the G3X-K level of theory.
product from the oxidation of formamide, which was also
detected in the oxidation of N-methylformamide.
estimated 50% yield.20 In our oxidation experiments, the
diformamide protonated molecular ion (m/z 74) was detected
by the PTR-MS but as a trace product with signal intensities
consistently less than the signal of HNCO (see Figure 2B), and
so we did not quantify its yield.
According to our theoretical results, the O2CH2NHCHO
peroxyl radical can be collisionally deactivated under tropo-
spheric conditions, allowing it to be removed through
bimolecular chemistry (Scheme S1). If this peroxyl radical is
converted to the corresponding alkoxyl radical,
OCH2NHCHO, through reaction with NO (and other
radicals), β-scission at the C−N bond could then lead to
formaldehyde and the NHCHO radical, whereas C−H scission
(or H abstraction by O2) would yield NH(CHO)2,
diformamide (Scheme S1). The NHCHO aminyl radical
formation is of interest as they are generally thought to be
removed from the atmosphere through reaction with radical
species such as NO and NO2, potentially leading to toxic
compounds that include nitrosamines and nitramines.9 If the
NHCHO aminyl radical is formed in the N-methylformamide
chamber experiments, it could be long-lived enough to abstract
a H from another compound in the chamber and produce
formamide, which was detected as an oxidation product, albeit
with low yields (<30%). However, the production of
formamide in these oxidation experiments does not show
purely first order kinetics, underlining the possibility that
heterogeneous chemistry may also be occurring on the chamber
walls (Figure 2B). On the other hand, Barnes et al. detected
diformamide as an oxidation of N-methylformamide in an
Upon reaction of the CH3NHCO radical with O2, a facile
unimolecular reaction pathway is available, as illustrated in the
bottom half of Figure 6. Similar to the formamide NH2CO
radical, O2 addition produces a peroxyl radical that can
eliminate HO2 via a barrier 18.00 kcal mol−1 below the
reactants. This reaction leads to the experimentally observed
reaction product methyl isocyanate. A second mechanism that
also proceeds with a transition state energy below the reactants
is available with a relative barrier height of −6.62 kcal mol−1. In
this case, the peroxyl group abstracts an H atom from the N-
methyl moiety, ultimately yielding CH2NH and CO2 and
regenerating the OH radical in the process. This mechanism
has been identified in secondary and tertiary amines and is
known to be uncompetitive with the facile HO2 elimination
reaction, when available.41 The experimental methyl isocyanate
yield of 36−40% is commensurate with similar yields of both
CH2NHCHO and CH3NHCO evolving from the reaction of
OH radicals with N-methylformamide, followed by almost
quantitative conversion of CH3NHCO to CH3NCO + HO2.
These results are also consistent with methyl isocyanate
G
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
detected by Barnes et al. from the oxidation of N-
methylformamide.20
Formamide may arise as a secondary oxidation product of the
CH2NHC(O)CH3 radical (by analogy to N-methylformamide).
Since this radical is the major reaction product, it may be
consistent with formamide yields in the range of 62−87%.
Formamide may also be produced as an oxidation product of
N-methylacetamide originating from the CH3NHC(O)CH2
radical. We also cannot rule out the possibility that formamide
is being produced via heterogeneous chemistry occurring on
the walls of the chamber and the exact mechanism of
formamide production remains unclear. Comparatively, Barnes
et al. state that they were unable to detect formamide in their
experimental setup due to interferences with the OH radical
precursor, CH3ONO, and potential oxidation products.20
e. Propanamide. Time traces of the evolution of the
products of propanamide are presented in Figure S5. The major
oxidation product detected is formamide but its experimentally
measured yield fell within the uncertainty associated with the
formamide production during amide-free control experiments
so we only report a < 30% yield. On the other hand, using
Equation S6 in a kinetic box model to fit the formamide trace,
we obtain a yield of 60%. The discrepancy between the
measured and fitted yields is likely due to the presence of a sink
for formamide at long reaction times (see Figure S5). HNCO
was detected during the experimental oxidation and again may
be a primary or secondary product.
When the methyl group in acetamide is extended to an ethyl
group, as in the case of propanamide, we find that N−H
abstraction still requires a significant barrier, but that
abstraction from the now secondary C−H carbon has a slightly
lower (but still positive) transition state barrier of +0.34 kcal
mol−1 (Figure S9). On the other hand, abstraction from the
primary C−H site proceeds with a barrier just below the
reactants (−0.30 kcal mol−1). Within the accuracy of the
reported calculations, we therefore suggest that both the
CH2CH2C(O)NH2 and CH3CHC(O)NH2 radicals will arise as
products of the propanamide + OH reaction. These results also
support a rate coefficient for the reaction of propanamide with
OH that is toward the lower end of the amides measured in this
study.
Even though we are not quantitative about the yield of
formamide from the oxidation of propanamide, its production
could originate from the minor NH2C(O)CHCH3 radical
product, relative to the NH2C(O)CH2CH2 radical (which we
then hypothesize cannot lead to formamide production).
Irrespective of these speculations, the specific pathways that
result in relatively high yields of formamide from oxidation of
all of the substituted amides studied here remain to be resolved
and the role of heterogeneous chemistry should also be
considered.
3. Reactivity Trends. Figure 7 provides the structures of
nine simple amides and includes a summary of the rate
coefficients measured in this work and in previous work. The
compounds are arranged in rows with increasing degree of N-
methylation from left to right, and in columns with increasing
length of alkyl chain from top to bottom. The rate coefficients
increase as a function of increasing number of methyl groups
on the nitrogen. This observation may be explained by the
addition of a methyl group which adds a competitive reactive
pathway for H-abstraction from the methyl C−H. Yet, the
ability of the methyl group to donate electron density to the
amide bond facilitates the H-abstraction of the acetyl C−H by
the electrophilic OH radical. There is a decreasing reactivity
trend from formamide to propanamide to acetamide toward
c. Acetamide. Acetamide reacted away too slowly within our
experimental setup to analyze its products. We show through ab
initio calculations that acetamide is indeed expected to react
more slowly than the other amides considered in this study. A
theoretical energy diagram for the acetamide + OH reaction is
provided in Figure S7. Here, an H-abstraction can again
transpire at the nitrogen, but with a barrier significantly above
the reactant energies (>5 kcal mol−1), and this process is
expected to be negligible. Instead, abstraction from the methyl
group is thought to dominate, proceeding via a transition state
at 2.04 kcal mol−1 above the reactants, leading to the
CH2C(O)NH2 radical as the major product. The relatively
large barrier for this reaction can explain the slow kinetics
reported here, where the rate coefficient is measured to be a
fraction of that of the formamide + OH reaction. The slow rate
of reaction also likely contributes to the difficulties associated
with obtaining reproducible measurements of the acetamide +
OH rate coefficient in the chamber experiments. According to
these calculations, the N−H abstraction is uncompetitive, in
contrast to the proposed mechanism by Barnes et al.,20 and we
therefore suggest that the abstraction occurs solely on the
methyl group. The products have yet to be quantified, although
we did detect formamide and HNCO, but we believe that the
proposed 100% yield for HNCO by Barnes et al. is unlikely.20
d. N-Methylacetamide. Time traces of the evolution of the
products of N-methylacetamide are presented in Figure S4.
Formamide was detected as the major product with measured
yields between 62 and 87% and a fit of 80% yield best
reproduced its experimental trace. A product at m/z 60 was also
observed and could be either N-methylformamide or
acetamide. Its trace was fitted with a yield of 10% using
Equation S6 under the assumption that m/z 60 is a primary
product but has a sink with a rate coefficient similar to
acetamide’s with OH. HNCO was also detected as a product
and was fitted with Equation S9.
A theoretical energy diagram for N-methylacetamide’s
reaction with OH is provided as Figure S8. The ab initio
calculations once again indicate that abstraction from the N−H
site requires a large barrier (3.35 kcal mol−1) and it is not
discussed further. There are two unique C−H bonds in N-
methylacetamide, analogous to the respective methyl group
abstractions in acetamide and in N-methylformamide. The
lowest barrier to reaction corresponds to H atom abstraction
from the N-methyl group, producing the CH2NHC(O)CH3
radical. This radical should be the dominant reaction product,
accompanied by lesser quantities of CH3NHC(O)CH2. The
calculated overall barrier height for N-methyl H-abstraction
(−1.96 kcal mol−1) is similar to, but somewhat less than, the
equivalent H-abstraction in N-methyformamide (−1.30 kcal
mol−1) due to the introduction of the acetyl functional group.
The substantially higher experimental rate coefficient for the N-
methylformamide versus N-methylacetamide reaction with OH
therefore provides further support for our finding that formyl
C−H abstractions are primary processes in the photochemical
oxidation of amides, when available. The barrier for H-
abstraction from the acetyl group in N-methylacetamide (1.34
kcal mol−1) is also reduced somewhat from that for the
analogous process in acetamide (2.04 kcal mol−1) by
introduction of the N-methyl substituent, although it remains
above the reactant energies.
H
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
Figure 8. Plot of predicted rate coefficients using EPA’s EPI Suite
model as a function of experimentally measured rate coefficients for
the nine amides in Figure 7. The numbers represent the level of N-
methylation on the amide, so that 1° represents primary amides, 2°
represents secondary amides, and 3° represents tertiary amides. The
color coding matches Figure 7 so that the blue series represents
formamide and its alkylated analogues, the red series represents
acetamide and its alkylated analogues, and the green series represents
propanamide and its alkylated analogues. The solid line is the 1:1
relationship, and the dotted lines have slopes differing by a factor of 2
from the 1:1 relationship.
Figure 7. Summary of room temperature rate coefficients with OH
radicals of nine amides based on their alkyl chain length (grouped by
colors) and degree of N-methylation (grouped by column).
OH radicals which also seems to hold true for their methylated
analogues. This trend is not straightforward since the H-
abstraction occurs at different locations on the H/alkyl
functional group attached to the carbonyl, dictating the
preferred mechanistic pathway and hence influencing the
experimental rate coefficient.
The rate coefficients of amides with OH radicals are slower
than with their homologous amine. For example, the rate
coefficient of methylamine, the homologue of formamide has a
rate coefficient of (17.3 1.1) × 10−12 cm3 molec−1 s−1, almost
four times the rate of formamide.42 The carbonyl moiety draws
electron density from the functional groups on either side of its
carbon atom, rendering them less reactive toward an OH
radical. This chemical property explains why amides are longer
lived in the atmosphere than amines. An incomplete
homologous summary for amines is depicted in Figure S10.
The U.S. Environmental Protection Agency’s (EPA)
Estimation Program Interface (EPI) Suite model was used to
compare the experimentally determined amide rate coefficients
shown in Figure 7 with their empirical estimates.40 The
AOPWIN module (v1.92) within the EPI suite model employs
structure−activity relationships (SARs) for estimating OH
reaction rate coefficients based on those developed by Atkinson
and co-workers.28,43,44 Figure 8 shows the correlation between
the predicted rate coefficients using EPI Suite and the measured
rate coefficients; the numbers (1°, 2°, 3°) represent the degree
of N-methylation on the amide, and the colors match the amide
groups in Figure 7. The model appears to correctly predict the
influence of the changing degree of N-methylation, but
struggles to represent the primary amides (formamide,
acetamide, and propanamide). These primary amides are the
slowest to react with OH and, consequently, have the longest
atmospheric lifetimes.
N−H abstractions, when available, should in all cases be
negligible. Attempts to determine a value for the substituent
factor F(−C(O)N< ) for amides in the context of the SAR
model of Kwok and Atkinson44 proved to be difficult with the
limited data set of rate coefficients presently available, and
further kinetic studies are needed. However, our experiments
on formamide suggest that the rate coefficient term k(HC(O)
N<) should be around 4 × 10−12 cm3 molec−1 s−1 and this
number is also compatible with the available data for N-methyl
and N,N-dimethylformamide. In comparison, the analogous
aldehyde C−H abstraction rate coefficient term is 17 × 10−12
cm3 molec−1 s−1.44 It is therefore apparent that the −NH2
group is deactivating the formyl C−H abstraction, although not
to the extent where it can be neglected altogether (as is
presently the case). The rate coefficient component attributable
to N-methyl abstractions should be about 5 × 10−12 cm3
molec−1 s−1, consistent with most of the reactivity occurring
on the methyl group for N-methylacetamide for example.
Furthermore, a total rate coefficient on the order of 10 × 10−12
cm3 molec−1 s−1 for N-methylformamide agrees well with the
reaction proceeding at almost equivalent rates at the alkyl and
formyl substituents. It does appear that N-methylation of
amides has a minor impact upon the reactivity of the acetyl
functionality, and vice versa.
This analysis prompted us to consider the analogous series of
amines shown in Figure S10, and their predicted rate
coefficients are plotted against the experimentally measured
rate coefficients of amines with OH radicals in Figure S11. The
AOPWIN output predicts a major contribution from the N−H
abstraction mechanism. Nonetheless, the N−H abstraction
channel appears to be of more significance for amines than for
amides.11,45 This study’s mechanistic work suggests a
reevaluation of the way that H-abstraction is treated for
organo-nitrogen compounds within the AOPWIN model and
ongoing work in our laboratory aims to address this issue.
Upon consultation of the AOPWIN user guide, we can better
assess the origin of the predicted rate coefficients. Interestingly,
the AOPWIN model does not consider formyl C−H
abstractions in amides, but does include N−H abstractions,
along with reaction at saturated carbon atoms (see Supporting
Information for relevant AOPWIN notes). However, our ab
initio and product detection results demonstrate that the formyl
C−H and the alpha C−H abstractions are likely to be major
reaction channels for the amides. Furthermore, we show that
I
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
(2) Schmeltz, I.; Hoffmann, D. Nitrogen-Containing Compounds in
Tobacco and Tobacco Smoke. Chem. Rev. 1977, 77, 295−311.
(3) Leach, J.; Blanch, A.; Bianchi, A. C. Volatile Organic Compounds
in an Urban Airborne Environment Adjacent to a Municipal
Incinerator, Waste Collection Centre and Sewage Treatment Plant.
Atmos. Environ. 1999, 33, 4309−4325.
(4) Rogge, W. F.; Hildemann, L. M.; Mazurek, M. A.; Cass, G. R.;
Simoneit, B. R. T. Sources of Fine Organic Aerosol. 1. Charbroilers
and Meat Cooking Operations. Environ. Sci. Technol. 1991, 25, 1112−
1125.
CONCLUSIONS
■
The rate coefficients of the oxidation of five different amides
with OH radicals were measured for formamide (4.44 0.46)
× 10−12 cm3 molec−1 s−1, N-methylformamide (10.1 0.6) ×
10−12 cm3 molec−1 s−1, acetamide (0.4−1.1) × 10−12 cm3
molec−1 s−1, N-methylacetamide (5.42
0.19) × 10−12 cm3
0.43) × 10−12 cm3
molec−1 s−1, and propanamide (1.78
molec−1 s−1 using relative rate experiments in a 1 m3 smog
chamber. These rate coefficients correspond to a wide range of
lifetimes against OH radicals on the order of 0.5−7 days using
an OH radical concentration of 2 × 106 molecules cm−3. It is
therefore likely that reactions with OH radicals are the
dominant daytime sink for amides.
(5) Sollinger, S.; Levsen, K.; Wunsch, G. Indoor Pollution by Organic
̈
Emissions from Textile Floor Coverings: Climate Test Chamber
Studies Under Static Conditions. Atmos. Environ. 1994, 28, 2369−
2378.
(6) Cheng, Y.; Li, S.; Leithead, A. Chemical Characteristics and
Origins of Nitrogen-Containing Organic Compounds in PM2.5
Aerosols in the Lower Fraser Valley. Environ. Sci. Technol. 2006, 40,
5846−5852.
(7) Laskin, A.; Smith, J. S.; Laskin, J. Molecular Characterization of
Nitrogen-Containing Organic Compounds in Biomass Burning
Aerosols Using High-Resolution Mass Spectrometry. Environ. Sci.
Technol. 2009, 43, 3764−3771.
(8) Raja, S.; Raghunathan, R.; Kommalapati, R. R.; Shen, X.; Collett,
J. L., Jr.; Valsaraj, K. T. Organic Composition of Fogwater in the
Texas−Louisiana Gulf Coast Corridor. Atmos. Environ. 2009, 43,
4214−4222.
(9) Nielsen, C. J.; Herrmann, H.; Weller, C. Atmospheric Chemistry
and Environmental Impact of the Use of Amines in Carbon Capture
and Storage (CCS). Chem. Soc. Rev. 2012, 6684−6704.
(10) Zhu, L.; Schade, G. W.; Nielsen, C. J. Real-Time Monitoring of
Emissions from Monoethanolamine-Based Industrial Scale Carbon
Capture Facilities. Environ. Sci. Technol. 2013, 47, 14306−14314.
(11) da Silva, G. Atmospheric Chemistry of 2-Aminoethanol (MEA):
Reaction of the NH2CHCH2OH Radical with O2. J. Phys. Chem. A
2012, 116, 10980−10986.
The evolution of the amides was monitored by PTR-MS and
so was the production of N-containing products, which
included toxic gas-phase molecules like HNCO and methyl
isocyanate. The time trace of the oxidation products were also
modeled using kinetic equations. The ab initio calculations
presented can explain the observation of HNCO and CH3NCO
upon the reaction of formamide and N-methylformamide,
respectively, with OH radicals. The peroxyl radical mechanisms
developed here are also consistent with no direct isocyanate
formation from any of the other amides investigated in this
study (although other isocyanates could be reasonably expected
to form in the OH radical reactions of other N-substituted
derivatives of formamide). The mechanism for formamide
production from N-methyl formamide and the other
substituted amides, however, remains unclear. Furthermore,
we expect that reactions with NO3 radicals or Cl radicals would
lead to similar mechanistic pathway assuming a H-abstraction
mechanism.
(12) Xie, H.; Li, C.; He, N.; Wang, C.; Zhang, S.; Chen, J.
Atmospheric Chemical Reactions of Monoethanolamine Initiated by
OH Radical: Mechanistic and Kinetic Study. Environ. Sci. Technol.
2014, 48, 1700−1706.
ASSOCIATED CONTENT
■
S
* Supporting Information
Further details on the experimental setup and HNCO
calibration are described. Experimental and theoretical details
of acetamide, N-methylacetamide, and propanamide are also
included. This material is available free of charge via the
Internet at http://pubs.acs.org.
(13) Chakir, A.; Solignac, G.; Mellouki, A.; Daumont, D.; Gas Phase,
U. V. Absorption Cross-Sections for a Series of Amides. Chem. Phys.
Lett. 2005, 404, 74−78.
(14) Koch, R.; Palm, W. -.; Zetzsch, C. First Rate Constants for
Reactions of OH Radicals with Amides. Int. J. Chem. Kinet. 1997, 29,
81−87.
AUTHOR INFORMATION
(15) Aschmann, S. M.; Atkinson, R. Atmospheric Chemistry of 1-
Methyl-2-pyrrolidinone. Atmos. Environ. 1999, 33, 591−599.
■
Corresponding Authors
*E-mail: jmurphy@chem.utoronto.ca.
*E-mail: jabbatt@chem.utoronto.ca.
́ ́
(16) Munoz, A.; Vera, T.; Sidebottom, H.; Rodenas, M.; Borras, E.;
̃
́
Vazquez, M.; Raro, M.; Mellouki, A. Studies on the Atmospheric Fate
of Propachlor (2-Chloro-N-isopropylacetanilide) in the Gas-Phase.
Atmos. Environ. 2012, 49, 33−40.
Notes
The authors declare no competing financial interest.
(17) Jackson, D. A.; Wallington, T. J.; Mabury, S. A. Atmospheric
Oxidation of Polyfluorinated Amides: Historical Source of Perfluori-
nated Carboxylic Acids to the Environment. Environ. Sci. Technol.
2013, 47, 4317−4324.
ACKNOWLEDGMENTS
■
The authors acknowledge the Canada Foundation for
Innovation and the Ontario Research Fund for infrastructure
support and NSERC and Environment Canada for operational
support. N.B. would like to thank Vanier Canada Graduate
Scholarships for funding. G.d.S. is grateful for support through
the Australian Research Council (DP110103889,
DP130100862, FT130101304). The authors would also like
to thank John Liggio for donating the HNCO source, as well as
Gregory Wentworth for help with the ion chromatograph.
(18) Solignac, G.; Mellouki, A.; Le Bras, G.; Barnes, I.; Benter, T.
Kinetics of the OH and Cl Reactions with N-Methylformamide, N,N-
Dimethylformamide and N,N-Dimethylacetamide. J. Photochem. Photo-
biol., A 2005, 176, 136−142.
(19) Solignac, G.; Magneron, I.; Mellouki, A.; Munoz, A.; Reviejo, M.
̃
M.; Wirtz, K. A Study of the Reaction of OH Radicals with N-Methyl
Pyrrolidinone, N-Methyl Succinimide and N-Formyl Pyrrolidinone. J.
Atmos. Chem. 2006, 54, 89−102.
(20) Barnes, I.; Solignac, G.; Mellouki, A.; Becker, K. H. Aspects of
the Atmospheric Chemistry of Amides. ChemPhysChem 2010, 11,
3844−3857.
REFERENCES
(21) El Dib, G.; Chakir, A. Temperature-Dependence Study of the
Gas-Phase Reactions of Atmospheric NO3 Radicals with a Series of
Amides. Atmos. Environ. 2007, 41, 5887−5896.
■
(1) Ge, X.; Wexler, A. S.; Clegg, S. L. Atmospheric Amines - Part I. A
Review. Atmos. Environ. 2011, 45, 524−546.
J
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A
Article
(22) Borduas, N.; Abbatt, J. P. D.; Murphy, J. G. Gas Phase Oxidation
of Monoethanolamine (MEA) with OH Radical and Ozone: Kinetics,
Products, and Particles. Environ. Sci. Technol. 2013, 47, 6377−6383.
(23) De Gouw, J.; Warneke, C. Measurements of Volatile Organic
Compounds in the Earth’s Atmosphere Using Proton-Transfer-
Reaction Mass Spectrometry. Mass Spectrom. Rev. 2007, 26, 223−257.
(24) Gylestam, D.; Karlsson, D.; Dalene, M.; Skarping, G.
Determination of Gas Phase Isocyanates Using Proton Transfer
Reaction Mass Spectrometry. Anal. Chem. Lett. 2011, 1, 261−271.
(25) Zhao, Y.; Truhlar, D. The M06 Suite of Density Functionals for
Main Group Thermochemistry, Thermochemical Kinetics, Non-
covalent Interactions, Excited States, and Transition Elements: Two
New Functionals and Systematic Testing of Four M06-Class
Functionals and 12 Other Functionals. Theor. Chem. Acc. 2008, 120,
215−241.
(26) da Silva, G. G3X-K Theory: A Composite Theoretical Method
for Thermochemical Kinetics. Chem. Phys. Lett. 2013, 558, 109−113.
(27) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. J. A.; Vreven, T.;
Kudin, K. N.; Burant, J. C. et al. Gaussian 09, revision B.01; Gaussian,
Inc.: Wallingford, CT, 2010.
(28) Atkinson, R. Kinetics and Mechanisms of the Gas-Phase
Reactions of the Hydroxyl Radical with Organic Compounds Under
Atmospheric Conditions. Chem. Rev. 1986, 86, 69−201.
(40) U.S. EPA Estimation Programs Interface Suite for Microsoft
Windows, version 4.11.; 2014.
(41) Onel, L.; Thonger, L.; Blitz, M. A.; Seakins, P. W.; Bunkan, A. J.
C.; Solimannejad, M.; Nielsen, C. J. Gas-Phase Reactions of OH with
Methyl Amines in the Presence or Absence of Molecular Oxygen. An
Experimental and Theoretical Study. J. Phys. Chem. A 2013, 117,
10736−10745.
(42) Carl, S. A.; Crowley, J. N. Sequential Two (Blue) Photon
Absorption by NO2 in the Presence of H2 as a Source of OH in Pulsed
Photolysis Kinetic Studies: Rate Constants for Reaction of OH with
CH3NH2, (CH3)2NH, (CH3)3N, and C2H5NH2 at 295 K. J. Phys.
Chem. A 1998, 102, 8131−8141.
(43) Atkinson, R. A Structure-Activity Relationship for the
Estimation of Rate Constants for the Gas-Phase Reactions of OH
Radicals with Organic Compounds. Int. J. Chem. Kinet. 1987, 19, 799−
828.
(44) Kwok, E. S. C.; Atkinson, R. Estimation of Hydroxyl Radical
Reaction Rate Constants for Gas-Phase Organic Compounds Using a
Structure-Reactivity Relationship: An Update. Atmos. Environ. 1995,
29, 1685−1695.
(45) Galano, A.; Alvarez-Idaboy, J. Branching Ratios of Aliphatic
Amines + OH Gas-Phase Reactions: A Variational Transition-State
Theory Study. J. Chem. Theory Comput. 2008, 4, 322−327.
(29) Hurley, M. D.; Wallington, T. J.; Laursen, L.; Javadi, M. S.;
Nielsen, O. J.; Yamanaka, T.; Kawasaki, M. Atmospheric Chemistry of
n-Butanol: Kinetics, Mechanisms, and Products of Cl Atom and OH
Radical Initiated Oxidation in the Presence and Absence of NOx. J.
Phys. Chem. A 2009, 113, 7011−7020.
(30) Roberts, J. M.; Veres, P. R.; Cochran, A. K.; Warneke, C.;
Burling, I. R.; Yokelson, R. J.; Lerner, B.; Gilman, J. B.; Kuster, W. C.;
Fall, R.; de Gouw, J. Isocyanic Acid in the Atmosphere and Its Possible
Link to Smoke-Related Health Effects. Proc. Natl. Acad. Sci. U.S.A.
2011, 108, 8966−8971.
(31) Wentzell, J. J. B.; Liggio, J.; Li, S.; Vlasenko, A.; Staebler, R.; Lu,
G.; Poitras, M.; Chan, T.; Brook, J. R. Measurements of Gas Phase
Acids in Diesel Exhaust: A Relevant Source of HNCO? Environ. Sci.
Technol. 2013, 47, 7663−7671.
(32) Roberts, J. M.; Veres, P. R.; VandenBoer, T. C.; Warneke, C.;
̈
Graus, M.; Williams, E. J.; Lefer, B.; Brock, C. A.; Bahreini, R.; Ozturk,
̈
́
F.; Middlebrook, A. M.; Wagner, N. L.; Dube, W. P.; de Gouw, J. A.
New Insights into Atmospheric Sources and Sinks of Isocyanic Acid,
HNCO, from Recent Urban and Regional Observations. J. Geophys.
Res. Atmos. 2014, DOI: 2013JD019931.
(33) Broughton, E. The Bhopal Disaster and its Aftermath: A Review.
Environ. Health 2005, 4, 6.
(34) Tsang, W. Chemical Kinetic Data Base for Propellant
Combustion. II. Reactions Involving CN, NCO, and HNCO. J. Phys.
Chem. Ref. Data 1992, 21, 753−791.
(35) Ly, T.; Kirk, B. B.; Hettiarachchi, P. I.; Poad, B. L. J.; Trevitt, A.
J.; da Silva, G.; Blanksby, S. J. Reactions of Simple and Peptidic Alpha-
Carboxylate Radical Anions with Dioxygen in the Gas Phase. Phys.
Chem. Chem. Phys. 2011, 13, 16314−16323.
(36) da Silva, G.; Kirk, B. B.; Lloyd, C.; Trevitt, A. J.; Blanksby, S. J.
Concerted HO2 Elimination from α-Aminoalkylperoxyl Free Radicals:
Experimental and Theoretical Evidence from the Gas-Phase
•
−
NH2 CHCO2 + O2 Reaction. J. Phys. Chem. Lett. 2012, 3, 805−811.
(37) da Silva, G.; Bozzelli, J. W.; Liang, L.; Farrell, J. T. Ethanol
Oxidation: Kinetics of the α-Hydroxyethyl Radical + O2 Reaction. J.
Phys. Chem. A 2009, 113, 8923−8933.
(38) Rissanen, M. P.; Eskola, A. J.; Nguyen, T. L.; Barker, J. R.; Liu,
J.; Liu, J.; Halme, E.; Timonen, R. S. CH2NH2 + O2 and CH3CHNH2
+ O2 Reaction Kinetics: Photoionization Mass Spectrometry Experi-
ments and Master Equation Calculations. J. Phys. Chem. A 2014, 118,
2176−2186.
(39) Poad, B. L. J.; Kirk, B. B.; Hettiarachchi, P. I.; Trevitt, A. J.;
Blanksby, S. J.; Clark, T. Direct Detection of a Persistent Carbonyloxyl
Radical in the Gas Phase. Angew. Chem., Int. Ed. 2013, 52, 9301−9304.
K
dx.doi.org/10.1021/jp503759f | J. Phys. Chem. A XXXX, XXX, XXX−XXX