Photoionization Mass Spectrometry Experiments and Master Equation

Mar 4, 2014 - Master equation computations, constrained by experimental kinetic results, .... Environmental Science & Technology 2014 48 (16), 9935-99...
0 downloads 0 Views 2MB Size
Article pubs.acs.org/JPCA

CH2NH2 + O2 and CH3CHNH2 + O2 Reaction Kinetics: Photoionization Mass Spectrometry Experiments and Master Equation Calculations Matti P. Rissanen,†,‡ Arkke J. Eskola,†,# Thanh Lam Nguyen,§ John R. Barker,∥ Jingjing Liu,⊥ Jingyao Liu,⊥ Erkki Halme,† and Raimo S. Timonen*,† †

Laboratory of Physical Chemistry, Department of Chemistry, University of Helsinki, P.O. Box 55, FIN-00014 Helsinki, Finland Division of Atmospheric Sciences, Department of Physics, University of Helsinki, P.O. Box 64, FIN-00014 Helsinki, Finland § Department of Chemistry & Biochemistry, The University of Texas at Austin, Texas 78712-0165, United States ∥ Department of Atmospheric, Oceanic, and Space Sciences, University of Michigan, Ann Arbor, Michigan 48109-2143, United States ⊥ Institute of Theoretical Chemistry, State Key Laboratory of Theoretical and Computational Chemistry, Jilin University, Changchun 130023, China ‡

S Supporting Information *

ABSTRACT: Two carbon centered amino radical (CH2NH2 and CH3CHNH2) reactions with O2 were scrutinized by means of laboratory gas kinetics experiments together with quantum chemical computations and master equation modeling. In the experiments, laser photolysis of alkylamine compounds at 193 nm was used for the radical production and photoionization mass spectrometry was employed for the time-resolved detection of the reactants and products. The investigations were performed in a tubular, uncoated borosilicate glass flow reactor. The rate coefficients obtained were high, ranging from 2.4 × 10−11 to 3.5 × 10−11 cm3 molecule−1 s−1 in the CH2NH2 + O2 reaction and from 5.5 × 10−11 to 7.5 × 10−11 cm3 molecule−1 s−1 in the CH3CHNH2 + O2 reaction, showed negative temperature dependence with no dependence on the helium bath gas pressure (0.5 to 2.5 Torr He). The measured rate coefficients can be expressed as a function of temperature with: k(CH2NH2 + O2) = (2.89 ± 0.13) × 10−11 (T/300 K)−(1.10±0.47) cm3 molecule−1 s−1 (267−363 K) and k(CH3CHNH2 + O2) = (5.92 ± 0.23) × 10−11 (T/300 K)−(0.50±0.42) cm3 molecule−1 s−1 (241−363 K). The reaction paths and mechanisms were characterized using quantum chemical calculations and master equation modeling. Master equation computations, constrained by experimental kinetic results, were employed to model pressure-dependencies of the reactions. The constrained modeling results reproduce the experimentally observed negative temperature dependence and the dominant CH2NH imine production in the CH2NH2 + O2 reaction at the low pressures of the present laboratory investigation. In the CH3CHNH2 + O2 reaction, similar qualitative behavior was observed both in the rate coefficients and in the product formation, although the fine details of the mechanism were observed to change according to the different energetics in this system. In conclusion, the constrained modeling results predict significant imine + HO2 production for both reactions even at atmospheric pressure.

1. INTRODUCTION Amines are ubiquitous to the Earth’s environment and, after ammonia (NH3), are the most common atmospheric bases.1 They are emitted to the atmosphere mostly from biogenic and oceanic sources with substantial anthropogenic influences through, for example, food and energy production by animal husbandry and biomass burning, and many other activities.1,2 A significant, yet to be quantified, anthropogenic addition to this budget will commence if amines are adopted for scrubbing CO2 from power plant flue gases (pilot demonstrations are already using amines for this purpose).3,4 Once emitted to the atmosphere, amines are likely to react with ambient OH radicals and will produce mostly carboncentered amino radicals and lesser yields of nitrogen-centered © 2014 American Chemical Society

imino radicals. For example, the aminomethyl radical (•CH2NH2) is the dominant product of the methylamine degradation (CH 3 NH 2 ) and the methylimine radical [CH3N(•)H)] is a minor product.3 Amine inclusion in atmospheric (nano)particles will also be occurring, but the extent and efficiency of this process remains uncertain at the moment.5,6 Photolysis by solar UV-radiation is a potential sink for some amine compounds, leading to the formation of alkylamine radicals,7 but this is not expected to be a particularly strong source of radicals, because the flux of UV-photons in the Received: November 15, 2013 Revised: February 26, 2014 Published: March 4, 2014 2176

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

Table 1. Performed R + O2 Measurements (R = CH2NH2 and CH3CHNH2) and the Results Obtained T/K

p/ Torr

a k/10−11 cm3 molecule−1 s−1

[O2]/1012 molecules cm−3

k′/s−1

lamp/ window

kwall/s−1

k(CH2NH2 + O2) = (2.89 ± 0.13) × 10−11 (T/300 K)−(1.10±0.47) cm3 molecule−1 s−1b k(CH2NH2 + O2) = (9.02 ± 4.87) × 10−12 × exp([−2.83 ± 1.34]kJ mol−1/RT) cm3 molecule−1 s−1c 267 0.98 3.52 ± 0.30 2.70−7.07 155.7−292.2 44.4 ± 6.0 298 0.51 2.87 ± 0.21 3.34−8.02 144.1−247.9 22.8 ± 2.5 298 1.1 3.05 ± 0.08 2.54−8.94 110.8−310.0 27.5 ± 2.1 298 1.12 2.40 ± 0.07 4.30−11.92 134.8−321.7 30.8 ± 1.9 298 2.46 2.45 ± 0.12 2.87−10.70 132.5−298.1 36.9 ± 3.1 336 1.24 2.42 ± 0.16 2.62−11.73 150.8−347.1 44.0 ± 5.2 363 1.34 2.62 ± 0.19 3.00−9.21 127.2−276.2 33.1 ± 5.5 k(CH3CHNH2 + O2) = (5.92 ± 0.23) × 10−11 (T/300 K)−(0.50±0.42) cm3 molecule−1 s−1b k(CH3CHNH2 + O2) = (3.47 ± 1.40) × 10−11 × exp([−1.33 ± 0.97]kJ mol−1/RT) cm3 molecule−1 s−1c 241 0.93 7.48 ± 0.78 2.07−4.99 260.2−440.8 92.6 ± 9.2 267 1.05 6.28 ± 0.44 2.20−5.39 245.4−415.5 70.9 ± 6.5 298 1.1 5.90 ± 0.75 2.65−5.17 246.3−388.3 69.2 ± 8.4 298 0.52 5.88 ± 0.34 2.53−6.49 245.9−470.4 66.3 ± 3.0 298 2.58 5.46 ± 0.38 2.77−6.23 255.3−388.8 57.1 ± 2.5 336 1.28 6.93 ± 0.87 2.43−5.45 282.3−484.5 66.8 ± 6.8 363 1.37 6.23 ± 1.06 2.58−5.26 273.3−413.2 77.7 ± 10.5

Ep /mJ cm−2

[amine]/1012 molecules cm−3

5.6 5.5 3.0 4.9 5.9 9.9 8.9

5.2 3.9 2.7 9.5 3.5 3.7 2.9

6.4 4.8 10.2 9.9 6.5 5.8 7.8

3.0 4.4 6.6 4.6 4.1 15.8 7.7

Br/Sap Br/Sap Br/Sap Br/Sap Br/Sap Br/Sap Br/Sap

N/Quartz N/Quartz Xe/sap N/Quartz N/Quartz Xe/sap Xe/sap

Statistical one standard error obtained from the fit. Overall uncertainty estimated as ±25% for the CH2NH2 + O2 reaction and ±30% for the CH3CHNH2 + O2 reaction (see text). bFit given by equation: kT = k300K (T/300 K)n. cFit according to Arrhenius equation kT = A × exp(−EA/RT).

a

CH3CHNH2 radicals (2). In addition to the title radicals, other species are formed as well (1b, 2b):

planetary boundary layer, where most amine emissions occur, is quite low. The high reaction rates with OH-radicals and the negative temperature dependences of the rate coefficients suggest that amines are possible players in reactive nitrogen chemistry in the colder regions above the boundary layer.8−11 Interestingly, amines have also been detected in the interstellar medium, and it has been proposed that they are linked to the formation of amino acids (potential first building blocks of life) in interstellar space.12,13 Due to the attention amines have received recently, several reviews have been published concerning their atmospheric sources and sinks, their known gas- and solid-phase chemistry, and their partitioning into aerosols.1−5 Previous kinetic studies of carbon centered amino radical reactions with O2 are scarce. Only the simplest of the alkylamine radicals, the CH2NH2 radical, has been studied prior to the present work. Jansen et al.14 studied the CH2NH2 + O2 reaction at 298 K and 1 atm of SF6, employing pulse radiolysis and UV-absorption detection. Also at 298 K, Masaki et al.15 investigated the same reaction at a few Torr pressure of N2 (1 Torr =1 mmHg = 133.3 N m−2) employing a photoionization mass spectrometer technique that is similar to the one used in the current study. In the present investigation we have measured the temperature dependences of the CH2NH2 + O2 and CH3CHNH2 + O2 reactions, as well as inspected the products, the intermediate species and the kinetics of the reaction by quantum chemical computations. In addition, the possible atmospheric implications of the results have been discussed briefly.

H 2NCH 2−CH 2NH 2 + hv(193 nm) → 2CH 2NH 2

→CH 2NH + other products

(1) (1b)

H 2NCH(CH3)−CH 2NH 2 + hv(193 nm) → CH3CHNH 2 + CH 2NH 2 →CH3CHNH + CH 2NH + other products

(2) (2b)

The initial radical concentrations produced were typically close to 1 × 1011 molecules cm−3 and were estimated from the measured gas flows and the precursor photodepletion signal, i.e., the decrease of the precursor signal due to the laser pulse photolyzing the gas mixture. The precursors ethylenediamine [(CH 2 NH 2 ) 2 , Aldrich, 99.5%] and 1,2-propanediamine (CH3CHNH2CH2NH2, Aldrich, 99%) were degassed prior to use with several freeze−pump−thaw cycles. The reactant oxygen (O2, Messer-Griesheim, 99.9995%) and the bath gas helium (He, Aga, 99.9996%) were used as supplied. The radicals and the products of the reactions studied were directly detected using a resonance gas lamp photoionization mass spectrometer, which has previously been described in detail by Eskola and Timonen.16 In ionizing the molecules and the radicals, high-cut filtered radiation from resonance gas lamps was applied, with discrete energies ranging from 7.11 eV (a nitrogen lamp with a quartz window) to 16.9 eV (a neon lamp with a multichannel plate filter). The CH2NH2 + O2 rate coefficient measurements were performed using a bromine lamp and a sapphire window (IE = 7.87 eV), whereas the CH3CHNH2 + O2 reaction kinetics were measured using a nitrogen lamp with a quartz window (IE = 7.11 eV) and a xenon lamp with a sapphire window (IE = 8.44 eV) (Table 1); no differences in the CH3CHNH2 radical signal profiles obtained were observed with these lamps. In the product studies, the higher ionization energies were also used.

2. EXPERIMENTAL INVESTIGATIONS The reactions were studied as a function of temperature and pressure in a 16.5 mm i.d. uncoated Pyrex flow tube reactor. The radicals for the experiments were produced by 193 nm ArF exciplex laser photolysis of two alkylamine precursors. Ethylenediamine [(CH2NH2)2] was used for producing the CH 2 NH 2 radicals (1) 1 5 and 1,2-p ropanediamine [CH3CHNH2CH2NH2] was employed for producing the 2177

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

were then determined. By plotting the rate coefficients against corresponding reactant concentrations, i.e., k′ versus [O2], the bimolecular rate coefficient kT at temperature T could be obtained from the slope of the plot (Figure 1 and 2) according

In the beginning and at the end of each experiment (i.e., an experiment at one temperature and one pressure), the radical decay rate in the reactor was measured without added O2 reactant. The radical decay rate obtained this way is a sum of the other significant loss processes for the radical in the reactor, except for the R + O2 reaction under study. It mainly contains the radical wall loss term and the reaction with the parent compound (i.e., photolysis precursor), but also other possible loss processes due to impurities or leaks, which were not monitored during the measurements. This loss rate was termed the wall loss rate (kwall) and is included in Table 1 together with the other experimental parameters. No systematic drift or any other irregularities were observed in the values obtained for kwall, which would have indicated interfering heterogeneous reactions obscuring the rate coefficient determinations. After the wall rate measurement, the reactant (O2) was added to the gas flow. The measurements were conducted under pseudo-first-order conditions with the O2 concentrations in large excess over the initial radical concentrations produced: [O2] ≫ [R]0. For the pseudo-first-order limit we applied the minimum requirement of: [O2]minimum ≥ 10 × [R]0, i.e., the smallest reactant O2 concentration used in deriving bimolecular rate coefficients was at least 10 times the estimated initial radical concentration [R]0 of the experiment. Under these conditions (and due to the low [R]0 produced), the radical decay rate could be described by a single-exponential function: [R]t = [R]0 exp(−k′t) + a0, where [R]t is the signal proportional to radical concentration at time t, k′ is the pseudo-first-order rate coefficient describing the time dependence of the radical signal decay, and a0 is the prephotolysis signal background level (Figure 1). The O2 concentrations were determined using the pressure change in a calibrated volume and the corresponding pseudo-first-order rate coefficients (k′)

Figure 2. A plot that was used to determine the bimolecular rate coefficient of the CH3CHNH2 + O2 reaction at 267 K and 1.1 Torr He. Included in the inset are the CH3CHNH (m/z = 43 Th) product signals obtained in the presence and absence of O2, respectively.

to an expression: kT(R + O2) = kwall + k′[O2]. From these individual experiments, the temperature dependences of the reactions were expressed according to kT = k300K(T/300 K)n, where k300K is the rate coefficient at T = 300 K, and n is the parameter describing temperature dependence. In addition, the more common Arrhenius equation [kT = A × exp(−EA/RT)] was used to obtain the Arrhenius activation energy (EA) and the freguency factor (A) of the reactions studied (Table 1). The overall uncertainty of the bimolecular rate coefficients obtained was estimated as ±25% for the CH2NH2 + O2 reaction and ±30% for the CH3CHNH2 + O2 reaction. These values include the estimated uncertainties in measured gas flows (i.e., reactant concentrations), as well as statistical uncertainties in the fitted parameters, and was determined by the propagation of errors method.

3. COMPUTATIONAL METHODS In addition to experimental investigation, both reactions were further characterized by quantum chemical calculations. The reaction paths and mechanisms were obtained by quantum chemical computations and the channel specific yields, under various experimental conditions, were obtained using master equation modeling. 3.1. Quantum Chemical Calculations. For both the CH2NH2 + O2 (RS-A) and the NH2CHCH3 + O2 (RS-B) reaction systems, equilibrium geometries of all stationary points on the potential energy surface (PES) were first optimized using the UB3LYP/6-311++G(d,p) level of theory.17,18 To verify the nature (minimum or transition state) of each stationary point on each PES, a harmonic vibrational analysis was carried out; harmonic vibrational frequencies were obtained as a result. Where necessary, intrinsic reaction coordinate (IRC)19−22 calculations were performed at the same level of theory for each transition state to identify the

Figure 1. An example of experimental data used to determine the bimolecular rate coefficient of the CH2NH2 + O2 reaction at 298 K and 1.1 Torr He. Included are the signals measured for the CH2NH2 radical decay (upper inset; k′decay(CH2NH2) = 113.5 ± 2 s−1, m/z = 30 Th) and for the CH2NH product (m/z = 29 Th) formation in the presence and absence of O2 (lower inset). All of the product profiles were recorded under identical experimental conditions. The line in the CH2NH product signal has been drawn with the formation rate that was measured for the radical decay rate, i.e, k′decay(CH2NH2) = k′formation(CH2NH) = 113.5 s−1. 2178

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

Figure 3. Potential energy surface for reaction system RS-A. The color code goes as follows: red color represents oxygen atom, blue color represents nitrogen atom, gray color stands for carbon atom and white color represents hydrogen atom. In addition, structures identified by horizontal parallel lines represent averages of multiple conformers; see text for details.

Figure 4. Potential energy surface for reaction system RS-B. (Note that although the species names are similar to those in Figure 3, the chemical species are not necessarily related.).

For the CH2NH2 + O2 reaction (RS-A): the direct hydrogenabstraction reaction producing HO2 + CH2NH was neglected, because its barrier is about 10 kcal mol−1 and the reaction is thus too slow to be significant in the atmosphere (see Figure S1, Supporting Information). For the reaction Int1

corresponding reactants and products. For improved accuracy, single point energies were computed at the B3LYP optimized geometries using the CBS-QB3 method,23,24 which is expected to produce results accurate to within about ±2 kcal mol−1. The PESs are shown in Figure 3 for RS-A and Figure 4 for RS-B. 2179

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

→ Int-3, which is the most important pathway, IRCMax(CBSQB3 // B3LYP)25 calculations were used to obtain a more reliable barrier height for TS-13 and raised the barrier from 22.0 kcal mol−1 to 23.1 kcal mol−1. The computed molecular properties for all species in the two reaction systems are summarized in the Supporting Information. All of the quantum chemistry calculations were carried out using the Gaussian 09 suite of software.26 3.2. Kinetics Calculations. The kinetics calculations were carried out using the MultiWell Program Suite.27−29 For pressure-dependent reactions and those that consist of multiple isomers and isomerization pathways, a master equation code must be used. The MultiWell master equation code employs the Rice−Ramsperger−Kassel−Marcus (RRKM) theory for computing energy-dependent microcanonical rate constants of reactions with intrinsic energy barriers.30,31 For reactions that lack intrinsic barriers (i.e., “barrierless” reactions), the semiempirical Inverse Laplace Transform (ILT) approach of Forst31,32 was used to compute microcanonical rate constants. RRKM theory30,31 was used for reactions with intrinsic energy barriers; the energy-dependent specific unimolecular rate constant k(E) is given by: ⎧ m‡ σ ⎫ 1 G‡(E − E ) ext 0 ⎬ k(E) = ⎨ ‡ ρ m h ( E ) σ ⎩ ext ⎭ ⎪







geometries, and relative energies. The Arrhenius parameters for kuni were obtained by assuming that −E∞ is equal to the computed enthalpy of reaction at 0 K for the exothermic recombination reaction: E∞ = −ΔfHr(0 K). This is the same as assuming that E∞ = E0, the critical energy for reaction, an assumption that tends to give more accurate values for k(E) near the reaction threshold at the expense of small errors in the computed k∞. The A-factor was then computed as A∞ = krec(T) exp(−ΔfHr(0 K)/RT). The final ILT expression for the microcanonical rate constants is given by k(E) = A∞

ρ (E − E 0 ) ρ(E + Δf Hr(0 K)) = A∞ ρ (E ) ρ (E )

(T-2)

where ρ(E) is the density of states of the dissociating intermediate. Further empirical adjustments were made to A∞ and energies to obtain better quantitative agreement with measurements of RS-A. In particular, A∞ for the Re → Int1 reaction step was adjusted to obtain krec ≈ 9 × 10−11 cm3 molecule−1 s−1 at the high pressure limit. This value is a little greater than the rate constant measured by Jansen et al.14 in 1 bar of SF6 at 298 K and is consistent with the experimental rate constant for analogous reaction O2 + C2H5.37 The adjusted value for the redissociation of Int1 is A∞ = 1.1 × 1016 s−1, which is typical for “barrierless” bond dissociation into two free radicals. The values of A∞ that were used in the final model for RS-A ranged from 4 × 1015 s−1 to 1.1 × 1016 s−1, which are reasonable for barrierless reactions. In a final empirical adjustment, the energy of TS-13 was raised by an additional 1.6 kcal mol−1 (from the IRCmax value, see above) in order to fit the experimental rate data in about 1 mbar of helium (this work) and in 1 bar of SF6,14 both at 298 K. No adjustments were made to RS-B, because of the limited extent of the available experimental information and because high-level quantum chemical calculations for this system are very expensive. We confined our attention to determining whether the behavior of RS-B is predicted to be qualitatively similar to that of RS-A. For RS-A, the master equation simulations employed “double arrays”29 consisting of 1500 array elements: 500 elements at a 10 cm−1 energy spacing, followed by energy grains of 85 cm−1 to a maximum energy of 85000 cm−1. For RS-B, the “double arrays consisted of 1500 array elements: 700 elements at a 10 cm−1 energy spacing, followed by energy grains of 106 cm−1 to a maximum energy of 85000 cm−1. The kinetics calculations were carried out for the temperature range 220−320 K and pressure range 0.001−10 bar. The master equation simulations were initiated using a chemical activation energy distribution of the excited adduct and terminated after the simulation time needed for several hundred collisions, when essentially all of the initially excited adduct had reacted, or had been stabilized by collisions. The pressure-dependent product branching ratios were obtained in simulations using estimated energy transfer parameters for the intermediates and for SF6, N2, and He buffer gases. Energy transfer was described by using the “exponential-down” model with energy transfer parameter α, which is approximately equal to the average energy transferred in deactivation collisions: ⟨ΔE⟩down. Based on prior experience,38 the values of α were roughly estimated to be 150 cm−1 for He, 300 cm−1 for N2, and 1200 cm−1 for SF6. The following Lennard-Jones parameters for collider gases were obtained

(T-1)

where m‡ and m are the number of optical isomers, σ‡ext and σext are the external rotation symmetry numbers, and g‡e and ge are the electronic state degeneracies of the transition state and reactant, respectively; h is Planck’s constant, G‡(E − E0) is the sum of states of the transition state, E0 is the reaction critical energy, and ρ(E) is the density of states of the reactant molecule. The sums and densities of states were computed using the Stein-Rabinovitch extension33 of the Beyer− Swinehart algorithm,34 which was implemented in DenSum (part of the MultiWell Program Suite). For the RS-A, centrifugal corrections were not applied, but quantum mechanical tunneling through unsymmetrical Eckart barriers was included.35 For the RS-B, quantum mechanical tunneling was neglected and centrifugal corrections29,36 were applied. However, centrifugal corrections are not expected to be significant in either reaction system. This is because the total rate constant is dominated by the barrierless association reactions, which were adjusted empirically (as described below), and because the isomerization reactions have tight transition states, which are not affected much by centrifugal effects.30,31 The barrierless reactions are not subject to quantum mechanical tunneling, but isomerization reactions involving Htransfer may be affected significantly, and tunneling corrections were included in RS-A, where the back-reaction to regenerate reactants is significant. In RS-B, tunneling was neglected because back reaction is negligible, and including tunneling would have had little influence on the total rate constant. The ILT semiempirical method (a built-in option in MultiWell) requires the density of states of the reacting intermediate and two empirical parameters: A∞ and E∞, the Arrhenius parameters of the high pressure limiting rate constants. These constants were obtained by first assuming that the recombination rate constant krec(T) measured in the present experiments is near the high pressure limit krec,∞. The rate constant for the corresponding unimolecular reaction kuni is then given by kuni = Keq/krec, where the equilibrium constant Keq was calculated from the computed harmonic frequencies, 2180

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

from Hippler et al.,39 which are tabulated in the MultiWell User Manual: σ(He) = 2.55 Å, ε/kB(He) =10 K, σ(N2) = 3.74 Å, ε/ kB(N2) = 82 K, and σ(SF6) = 5.20 Å, ε/kB(SF6) = 212 K. The Lennard-Jones parameters of NH2CH2O2 and NH2CHCH3O2 (the primary adducts in the two reaction systems) were assumed to be approximated by those tabulated for n-Butane and n-Pentane, respectively, which have the same number of heavy atoms as the two primary adducts; the parameters for the other intermediates and complexes were assumed to be the same as the primary adduct in each specific reaction system. It should be emphasized that for pressures below about 1 bar, the results presented below show little pressure dependence, suggesting that errors associated with these crudely estimated energy transfer parameters are probably not important at pressures less than 1 bar. All of the parameters needed to reproduce the calculations are presented in the Supporting Information. All unimolecular rate constants and recombination rate constants in this study are expressed in units of s−1 and cm3 molecule−1 s−1, respectively; energies are expressed in units of kcal mol−1, and in the case of energy transfer parameters, in units of cm−1.

Torr pressure rate coefficient values for the R + O2 reactions, where R is the corresponding unsubstituted carbon centered alkyl radical, have been determined as, 1.7 × 10−14 cm3 molecule−1 s−1 (CH3 + O2 reaction),40 1.0 × 10−12 cm3 molecule−1 s−1 (C2H5 + O2)41 and 5.5 × 10−12 cm3 molecule−1 s−1 (n-C3H7 + O2),42 and illustrate the rate enhancing effect of the amino group attached to the radical center (likely due to the electron donating nature of the N-atom). It should be noted that for the CH3 + O2 reaction, there is no facile exit channel, and at a few Torr, the reaction will be far from the high pressure limit; hence the comparison might not be entirely applicable. Nevertheless, it exemplifies how large a change in rate coefficients results from exchanging one of the H atoms in methyl radical with an NH2-group. Common Arrhenius equation (kT=A × exp(−EA/RT) is not well-suited for reactions with submerged barriers on the PES (e.g., the lowest energy paths in Figures 3 and 4), as the activation barrier has little meaning when the coefficient EA is negative. In addition, the kT = k300K(T/300 K)n equation provides more useful information with a glance (i.e, the rate coefficient at room temperature, k300K, and the dependence on temperature with one parameter; n), and hence, was chosen. Nevertheless, due to common practice, i.e., to enable easier comparison, the common Arrhenius parameters were also determined (Table 1). The high rate coefficients obtained here indicate that the carbon centered amino radicals produced in the ambient atmosphere will quickly react with O2 to produce the corresponding ro-vibrationally excited amino peroxy radicals. The excited peroxy species can then, depending on the prevailing conditions, either, stabilize in collisions with the bath gas, dissociate back to the radical and O2 reactants, or react further by uni- and bimolecular reaction pathways. In the low pressure experiments of the current investigation, the amino radical reactions with O2 are observed to form imine products; this finding is supported by the present quantum chemical calculations (see below). In the CH2NH2 + O2 reaction, methylene imine CH2NH was observed as a main product, and the formation signal was shown above in the inset of Figure 1. However, the CH2NH is also produced in the photolysis of the ethylenediamine parent compound (1b), potentially through dissociation of the excited CH2NH2 to CH2NH and H-atom, and hence, the CH2NH formation kinetics could not be reliably deduced. Nevertheless, there is also a significantly slower formation component in the methyleneimine product profile recorded (Figure 1), which is evident as a slower rise in the signal after the nearly instantaneous photolysis production at t = 0. The coproduct HO2 was not observed in the product measurements, despite multiple attempts. This was not surprising, since we have not previously been able to directly detect alkoxy or peroxy species with the resonance gas lamp photoionization method. At higher pressures, the CH2NH2O2 peroxy radical stabilization is expected to compete, as shown below by quantum chemical calculations and master equation modeling. A small signal possibly corresponding to imine production (CH3CHNH, N-methylmethaneimine) was also observed in the CH3CHNH2 + O2 reaction, but with much lower intensity than that observed in the CH2NH2 + O2 reaction system. Moreover, the yield of imine produced promptly by photolysis In the CH3CHNH2 + O2 system was quite large and interfered with identifying and measuring the small reaction yields. However, the quantum chemical calculations show the

4. EXPERIMENTAL RESULTS AND DISCUSSION The rate coefficients obtained for both reactions show negative temperature dependence (Table 1, Figure 5), with no apparent

Figure 5. Measured bimolecular rate coefficients of the CH2NH2 and CH3CHNH2 radical reactions with O2 shown as a function of temperature. Also included are the previous values of the CH2NH2 + O2 reaction measured at 298 K by Masaki et al.15 at a few Torr pressure (hollow star) and by Jansen et al.14 at atmospheric pressure (hollow diamond). The error bars shown for the current results contain only the statistical one standard error obtained from the fitting procedure.

pressure dependence under the low pressure conditions of the current experimental investigation (0.5−2.5 Torr He). The negative temperature dependence together with the absence of bath gas density dependence are common for radical−radical addition reactions, which have no intrinsic potential energy barrier along the entrance channel of the reaction; note that ground state O2 is also a radical species with its two unpaired electrons (i.e., a biradical species). However, the rate coefficients determined in the current investigation are unusually large for carbon centered free radical reactions with O2. To give an example, similar room temperature and few 2181

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

new C−O bond between the α-carbon and O2. This initial step is followed by isomerization and/or fragmentation of the energy-rich ROO adduct (labeled Int-1 and Int1, respectively, in Figures 3 and 4). In each system, the ROO adduct has two or more distinct geometrical conformations, which differ slightly in energy and which can interconvert to each other by the rotation around the C−O bond. The most accurate approach would be to treat the conformers as minima of a single hindered internal rotation. Although feasible, the expense of such a procedure is not warranted for present purposes. Because the conformers are similar in energy and the barriers separating them are low, rapid equilibration is expected. Thus the rotamers are treated in the present work as a single average molecule (Int1), which simplifies the PES and reduces the expense of the calculations without much loss of accuracy. In the present work, all intermediates and transition states with two or more isomers are treated in this way and are marked in the figures with parallel horizontal lines. The geometries and relative energies of the individual conformers (identified by names terminated with a lowercase a, b, c, etc.) are tabulated in the Supporting Information. As mentioned above, the energy-rich ROO adduct is a free radical with the unpaired electron residing on the terminal Oatom. ROO can react in several ways. In both systems, the lowest energy pathway is a “tail-biting” isomerization, in which the terminal O-atom attacks one of the H-atoms in the NH2 group via a five-member cyclic transition state (TS-13 and TS12, for the two reaction systems). The product of this attack is a hydrogen-bonded cyclic complex (Int-3 and Int2, respectively, in the two reaction systems), which can subsequently dissociate to form HO2 + CH2NH (product set P1) and HO2 + CH3CHNH (product set P3), respectively, in the two reaction systems. The master equation calculations described below predict that under atmospheric conditions the lowest energy path in each system is the most important by a large margin. For that reason, the other possible pathways in RS-B will not be discussed further, although the most important pathways were included in the master equation modeling and all of the associated parameters are tabulated in the Supporting Information and displayed in Figure 4. All of the potential products indicated by quantum chemical calculations, no matter how unlikely they were deemed, were also sought in the experimental laboratory investigation. No other products than the imines discussed above were detected during these studies. 5.2. Master Equation Simulations. Master equation calculations were carried out using reduced mechanisms. For RS-A, the reduced mechanism consisted of three intermediates (Int-1, Int-2, and Int-3) and all four sets of terminal products, connected by six reactions. For RS-B, the reduced mechanism consisted of five intermediates (Int1, Int2, Int3, Int4, and Int9) and all seven sets of terminal products, connected by a total of 14 reactions. The other intermediates were neglected because they are weakly bound and have extremely short lifetimes. All of the isomerization reactions were treated as reversible, while reactions to form bimolecular products were treated as irreversible. For chemical activation simulations such as these, the MultiWell master equation code reports the relative yields (“fractions” f i) of intermediates and bimolecular sets (including the bimolecular reactants and terminal products), at the end of the simulation. Calculations were carried out for various

CH3CHNH as the most likely product (Figure 4) and considerably strengthen our tentative experimental identification. Thus, we conclude that the N-methylmethaneimine is the most likely product of the CH3CHNH2 + O2 reaction at the low pressures employed in the current work. Similarly to the CH2NH2 + O2 reaction, the CH3CHNH2O2 peroxy radical stabilization becomes increasingly more important as the bath gas pressure approaches one atmosphere, which corresponds to Earth’s surface where most of the amine emissions are located and where they are subsequently converted to the first generation oxidation products. In previous studies, Masaki et al.15 measured the CH2NH2 + O2 rate coefficient at 298 K with a photoionization mass spectrometer similar to the one used here and obtained 3.5 × 10−11 cm3 molecule−1 s−1 in good agreement with the current result: 2.89 × 10−11 cm3 molecule−1 s−1. They observed no pressure dependence in the 0.6 to 6 Torr (N2) pressure range, which is in accord with the current direct detection of CH2NH in the reaction at these low pressures. In higher pressure, not achievable with the current experimental setup, the formation of CH2NH2OO is expected to have greater importance, as concluded by Jansen et al.14 who also investigated the CH2NH2 + O2 reaction at T = 298 K, but at atmospheric pressure of SF6. Jansen et al.14 employed pulse radiolysis for F-atom generation from SF6, and subsequent F-atom reaction with methylamine for CH2NH2 radical production. The UV-absorption technique was employed for direct detection of CH2NH2 at 319.5 nm and resulted in a rate coefficient of 7.8 × 10−11 cm3 molecule−1 s−1, significantly higher value than the rate constants measured in the current study. No products were detected in these previous experimental investigations, but by comparison to the value obtained by Masaki and co-workers at lower pressure,15 Jansen postulated that the reaction has a pressure-dependent channel producing the CH2NH2OO peroxy radical. The present quantum chemical calculations confirm this speculation (see below). No previous studies have been performed with the CH3CHNH2 radical, and hence, a straightforward comparison cannot be made. However, the results can be compared to the corresponding unsubstituted hydrocarbon radical, the ethyl radical, for which the O2 reaction rate coefficient at room temperature and low pressure is 1.0 × 10−12 cm3 molecule−1 s−1,41 as mentioned above. The rate of CH3CHNH2 + O2 obtained in the current study under similar conditions, 5.92 × 10−11 cm3 molecule−1 s−1, is over 50 times faster and demonstrate how subtle electronic effects can lead to a large enhancement in the observed rates. Because the unsubstituted carbon species cannot be used as good analogies for the amine radicals, additional detailed experimental studies are needed.

5. THEORETICAL RESULTS AND DISCUSSION 5.1. Reaction Mechanisms. The potential energy surfaces (PESs) for CH2NH2 + O2 (reaction system RS-A) and NH2CHCH3 + O2 (reaction system RS-B) are depicted in Figures 3 and 4, respectively. The optimized structures of the intermediates and transition states are shown in both figures. The energies (including zero point energies) of the intermediates, products, and transition states are given, relative to the reactants. Molecular parameters for all species are tabulated in the Supporting Information. In both reaction systems, an unpaired electron resides on the α-carbon (adjacent to the NH2 group) of CH2NH2 and NH2CHCH3. In both systems, the initial step is formation of a 2182

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

reaction conditions (T = 220−320 K and p = 0.001−10 bar), and the results are tabulated in the Supporting Information. Results obtained at 298.15 K for RS-A and RS-B as functions of pressure are shown in Figure 6a,b, and Figure 7, respectively.

N2 in RS-A (Figure 6b), but less than half in RS-B (Figure 7). Normally, the fact that the adduct in RS-B is a larger species than the analogous one in RS-A would favor more adduct formation. In this case, however, TS1-2 in RS-B is more than 5 kcal mol−1 lower in energy than TS-13, the analogous transition state in RS-A, which leads to enhanced formation of bimolecular products. Total bimolecular rate constants were obtained by using the following equation, which expresses the product of the highpressure limit rate constant for the initial step multiplied by the net fraction of reactants (Re) consumed:27 k tot = k rec, ∞(1 − fRe )

(T-3)

where krec,∞ is the high pressure limit recombination rate constant of the reaction Re → Int1 and f Re is the calculated fraction of Re regenerated by the back-reaction of Int1. Total rate constants for RS-A are shown as functions of temperature and pressure in Figure S8 (similar information is presented for RS-B) in the Supporting Information. The results for three collider gases (He, N2, and SF6) at 298 K are shown in comparison with experimental data for RS-A in Figure 8 (see Figure S11 for RS-B). The experimental data obtained in the present low-pressure experiments (∼1 mbar) and the model calculations for RS-A are displayed in Figure 9.

Figure 6. Fractions (reaction system RS-A) computed at 298.15 K at various pressures of (a) helium bath gas, and (b) nitrogen bath gas.

Figure 8. Total rate constants (reaction system RS-A) computed as a function of pressure for three collider gases, computed at 298.15 K. Experimental data are shown as points with error-bars.

Figure 7. Fractions (reaction system RS-B) computed at 298.15 K at various pressures of N2.

These plots show that most of net reaction in both systems proceeds to bimolecular products at p < 0.1 bar and the stabilized adduct in each system (Int-1 and Int1, respectively) is the major product at high pressures. Comparison of the two plots shows that back-reaction to regenerate reactants is prominent in RS-A and negligible in RS-B. In addition, adduct formation accounts for about half of the net reaction at 1 bar of

Figure 9. Calculated total rate constants (reaction system RS-A) as a function of temperature, computed for 1 mbar of He bath gas. 2183

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

measurements and also show a weak negative temperature dependence (i.e., the rate constant decreases with increasing T). Minor adjustments could almost certainly bring the computed rate constant into quantitative agreement with the experiments, but this was not pursued, since the uncertainties are significant (see preceding discussion), and nothing further would be learned from such an exercise.

The results show that with judicious adjustments, the computed total rate constant for RS-A can be brought into quantitative agreement with both the pressure and temperature dependences of the experimental data. This satisfactory level of agreement lends weight to the product distributions predicted by the calculations. At low pressures, bimolecular products dominate (HO2 + CH2NH, P1), while the RO2 adduct (Int-1) is dominant at high pressures. Under atmospherically relevant conditions, at 1 bar of N2, 298 K and with large abundance of oxygen, the calculations predict that the net reaction yield of RO2 is about 1/2 and that of the CH2NH + HO2 channel about 1/5 (see Figure 6b). Without any adjustments to the computed energies, the results of the master equation simulations are in qualitative, but not quantitative, agreement with the experimental rate constants. The CBS-QB3 method computes energies with associated errors of up to approximately ±2 kcal mol−1. Potentially, errors of this order can affect all of the rate constants and product ratios. In our modeling, all energies except for that of TS-13 were held at the computed values. Thus parameter adjustments are certainly compensating for a multitude of small errors. A full sensitivity analysis would be a very elaborate undertaking, but insight into the robustness of the present results can be obtained more easily. For the RS-A reaction system, the total rate constant and the product ratios depend mostly on competition among the processes affecting Int-1: redissociation, collisional stabilization, and reaction via transition state TS-13. This competition depends on both temperature and pressure. The good agreement between the present theoretical results and experiments thus relies only on a few parameters. In particular, it mostly relies on the assumed A∞ for the redissociation of Int-1 to regenerate CH2NH2 + O2 and on the energy of transition state TS-13, which was adjusted by +1.6 kcal mol−1, as discussed above. Parameter A∞ was set so that the highpressure limit of the recombination rate constant (which is related to the dissociation rate via the equilibrium constant) is slightly higher than the experimental highest experimental rate constants (see Figure 8). The uncertainty in A∞ associated with using the experimental data in this manner is probably of the order a power of 10. The effect of adjusting the energy of TS-13 is shown in Figure S9, which compares the results obtained with and without the adjustment to TS-13. Qualitatively, the results are similar, but the quantitative yields are affected, especially at low pressures. At a 1 bar of N2, the total rate constant changes by about a factor of x2, as do the relative product yields; at higher pressures, the uncertainties are reduced. The results for reaction system RS-B are in good qualitative agreement with those for RS-A. Although RS-B is more complicated, the calculations predict that most of the reaction also proceeds via the lowest energy path, which is about 5 kcal mol−1 lower in energy than in RS-A. This difference leads to the most striking difference between the two systems: because there is essentially no redissociation of the ROO adduct in this case, the total rate constant is almost independent of pressure (Figure S10), while the rate constant for producing the adduct still shows a substantial pressure-dependence (Figure S11). Also for the CH3CHNH2 + O2 reaction, the net yield of imine + HO2 channel is predicted to be high: about 0.6 at 298 K and 1 atm of N2 (Figure 7). At about 1 mbar pressure (close to the present experimental conditions), the computed total rate constants are within a factor of 2 of the experimental

6. ATMOSPHERIC IMPLICATIONS Previously, Schade and Crutzen2 have implicated the formation of CH2NH in the CH2NH2 + O2 reaction and indeed the formation of CH2NH was confirmed here, by direct measurements (Figure 1) together with computational methods (Figure 3). Recently, imine formation in a NH2•CHCO2− + O2 reaction has been observed and was postulated to form through a concerted HO2 elimination pathway.8 In addition, degradation of many primary amine emissions is believed to lead to imine formation in the atmosphere.3,4 Thus, the current work adds to the mounting evidence that imines are among the primary degradation products of primary and secondary amine emissions in the atmosphere. The gas-phase reactions of amines are much more rapid than those of ammonia (NH3).2 Consequently, the amines studied here are generally removed by gas-phase reactions, whereas NH3 is more important in aqueous droplet chemistry. Larger amines and polyfunctional amines, which have lower vapor pressures, may still be removed primarily by uptake in aqueous aerosols and surface waters. In a recent study, Da Silva43 suggests that the imine products formed in these amino radical reactions are also prone to uptake by aqueous aerosols, and hence, will hydrolyze relatively easily under atmospheric conditions. If this would be the fate of the imine species formed in the reactions investigated here, then these imine products would have led to hydrolysis products of NH3 + H2CO, through CH2NH + H2O reaction, and to NH3 + CH3CHO, through CH3CHNH + H2O reaction, respectively. Since ammonia and amines are the most abundant atmospheric bases, their abundance and removal are of great importance, e.g., for particle formation through salt production in the atmosphere, and through maintenance of acid−base balance of the atmosphere and ecosystems. The amine lifetimes with respect to reactions with OHradical have been estimated to range from minutes in maritime environments to days in remote continental locations.4 Hence, in light of the current results, amines are expected to be quickly converted to the corresponding peroxy radicals, and subsequently, with a substantial yield, to imine + HO2 bimolecular products. The stabilized peroxy radicals obtained from the same reactions will most likely react with other peroxy radicals (i.e., with RO2, where R = H or an arbitrary organic radical) or NO, depending on their availability in the gas mixture. According to reasoning presented by Schade and Crutzen,2 the imine formation in these reactions is also a potential source of N2O and HCN to the atmosphere, although this speculation is yet to be verified. The imine formation may also lead to another type of reaction class becoming increasingly important: the CN double bond could be susceptible to an O3 attack, which is not possible for saturated amine compounds, and hence, may open up a new avenue for organic nitrogen chemistry that has not received much attention in an atmospheric context. 2184

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

Present Address

7. CONCLUSIONS Amines are common to Earth’s environment, but their fate in the biosphere−atmosphere system is yet uncertain. Here we have examined a part of the atmospheric degradation of two primary amino radical compounds (CH 2 NH 2 and CH3CHNH2) that are expected to form due to primary amine emission degradation in the atmosphere. The investigations were performed by means of laboratory kinetics measurements together with high-level quantum chemical computations and master equation modeling. It was found that these reactions are unusually fast for carbon centered radical reactions with O2 and hence present an important removal route for these amino radicals from the gas phase. The main product distribution was observed to change according to the prevailing pressure conditions. The imines were observed in the experimental study under a low few Torr bath gas helium pressure, whereas the peroxy radical stabilization was indicated at higher pressures by quantum chemical/master equation computations. According to the experimentally constrained master equation calculations, the net product yield of imine + HO2 reaction channel is significant for both reactions even at atmospheric pressure. At 298 K and 1 bar of N2, the yield of CH2NH from the CH2NH2 + O2 reaction was about 0.2 and the yield of CH3CHNH from the CH3CHNH2 + O2 reaction about 0.6, respectively. Peroxy radical and imine + HO2 reaction channels are expected to lead to a considerably different chemistry; the imine compounds contain a double bond which is likely susceptible to ozonolysis and OH addition reactions (and then to subsequent peroxy radical formation), whereas the primary produced peroxy species are generally expected to be consumed in the reactions with other peroxy species (HO2 and RO2) in more pristine environments and also with NO reactions in urban areas. Not much is known about the amine gas-phase reactivity, not to mention the imine gas-phase reactivity, and the gathering evidence suggests that the kinetics of the reactions of nitrogen containing organic compounds certainly warrant further investigation.



#

Combustion Research Facility, Mail Stop 9055, Sandia National Laboratories, Livermore, California 94551−0969, USA.

Notes

The authors declare no competing financial interest.



ACKNOWLEDGMENTS J.R.B. thanks the U.S. National Science Foundation (Atmospheric and Geospace Sciences) for funding; this material is based upon work supported by the National Science Foundation under Grant Number (1231842). R.S.T. thanks Claus. J. Nielsen for helpful discussions and the Laboratory of Physical Chemistry of the University of Helsinki for support.



(1) Ge, X.; Wexler, A. S.; Clegg, S. L. Atmospheric Amines − Part I. A Review. Atmos. Environ. 2011, 45, 524−546. (2) Schade, G. W.; Crutzen, P. J. Emission of Aliphatic Amines from Animal Husbandry and Their Reactions: Potential Source of N20 and HCN. J. Atm. Chem. 1995, 22, 319−346. (3) Nielsen, C. J.; D’Anna, B.; Karl, M.; Aursnes, M.; Boreave, A.; Bossi, R.; Bunkan, A. J. C.; Glasius, M.; Hansen, A. M. K.; Hallquist, M. et al. Atmospheric Degradation of Amines (ADA), Summary Report: Gas Phase Photo-Oxidation of Methylamine, Dimethylamine and Trimethylamine, CLIMIT project no. 201604, OR 2/2011; Norwegian Institute for Air Research: Kjeller, Norway, 2011; pp 1− 146. (4) Nielsen, C. J.; Hermann, H.; Weller, C. Atmospheric Chemistry and Environmental Impact of the Use of Amines in Carbon Capture and Storage (CCS). Chem. Soc. Rev. 2012, 41, 6684−6704. (5) Qiu, C.; Zhang, R. Multiphase Chemistry of Atmospheric Amines. Phys. Chem. Chem. Phys. 2013, 15, 5738−5752. (6) Almeida, J.; Schobesberger, S.; Kürten, A.; Ortega, I. K.; Kupiainen-Mäaẗ tä, O.; Praplan, A. P.; Adamov, A.; Amorim, A.; Bianchi, F.; Breitenlechner, M.; et al. Molecular Understanding of Sulphuric Acid−Amine Particle Nucleation in the Atmosphere. Nature 2013, 502, 359−363. (7) Ross, P. L.; Van Bramer, S. E.; Johnston, M. V. Ultraviolet Photodissociation of Gas-Phase Alcohols, Amines, and Nitroalkanes. Appl. Spectrosc. 1996, 50, 608−613. (8) 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. (9) 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. (10) 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, DOI: 10.1021/jp406522z. (11) Onel, L.; Blitz, M. A.; Seakins, P. W. Direct Determination of the Rate Coefficient for the Reaction of OH Radicals with Monoethanol Amine (MEA) from 296 to 510 K. J. Phys. Chem. Lett. 2012, 3, 853−856. (12) Martins, Z.; Price, M. C.; Goldman, N.; Sephton, M. A.; Burchell, M. J. Shock Synthesis of Amino Acids from Impacting Cometary and Icy Planet Surface Analogues. Nat. Geosci. 2013, 6, 1045−1049. (13) Halfen, D. T.; Ilyushin, V. V.; Ziurys, L. M. Insights into Surface Hydrogenation in the Interstellar Medium: Observations of Methaneimine and Methyl Amine in Sgr B2(N). Astrophys. J. 2013, 767, 66−77.

ASSOCIATED CONTENT

S Supporting Information *

Reaction system RS-A, Tables S1−S4: Calculated heats of formation of important species, rovibrational parameters for various minima and the optimized geometries. Figures S1−S9: PES for H-abstraction pathway, IRC calculations for various reaction steps and the rate coefficients as a function of temperature and He pressure. Reaction system RS-B, Tables S5−S10: Calculated heats of formation of important species, the optimized geometries and the rovibrational parameters for various minima. Computed primary product fractions, total rate coefficients as a function of temperature and He pressure, and summary of Master Equation input parameters. Figures S10− S12: Calculated rate coefficients as a function of He pressure and temperature, and the rate coefficient for ROO adduct formation. Total rate coefficient as a function of temperature at 1 mbar of He. This information is available free of charge via the Internet at http://pubs.acs.org.



REFERENCES

AUTHOR INFORMATION

Corresponding Author

*Tel: +358919150282; E-mail: raimo.timonen@helsinki.fi. 2185

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186

The Journal of Physical Chemistry A

Article

(14) Jansen, T. C.; Trabjerg, I.; Rettrup, S.; Pagsberg, P.; Sillesen, A. Experimental and Theoretical Investigation of the UV Spectrum and Kinetics of the Aminomethyl Radical, CH2NH2. Acta Chem. Scand. 1999, 53, 1054−1058. (15) Masaki, A.; Tsunashima, S.; Washida, N. Rate Constants for Reactions of Substituted Methyl Radicals (CH2OCH3, CH2NH2, CH2I, and CH2CN) with O2. J. Phys. Chem. 1995, 99, 13126−13131. (16) Eskola, A. J.; Timonen, R. J. 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. (17) Becke, A. D. Density-Functional Thermochemistry. III. The Role of Exact Exchange. J. Chem. Phys. 1993, 98, 5648−5652. (18) Stephen, P. J.; Devlin, F. J.; Chabalowski; Frisch, M. J. Ab Initio Calculation of Vibrational Absorption and Circular Dichroism Spectra Using Density Functional Force Fields. J. Phys. Chem. 1994, 98, 11623−11627. (19) Ishida, K.; Morokuma, K.; Kormornicki, A. The intrinsic Reaction Coordinate. An Ab Initio Calculation for HNC→HCN and H−+CH4→CH4+H−. J. Chem. Phys. 1977, 66, 2153−2156. (20) Fukui, K. The Path of Chemical Reactions - The IRC Approach. Acc. Chem. Res. 1981, 14, 363−368. (21) Gonzalez, C.; Schlegel, H. B. Reaction Path Following in MassWeighted Internal Coordinates. J. Phys. Chem. 1990, 94, 5523−5527. (22) Hratchian, H. P.; Schlegel, H. B. Accurate Reaction Paths Using a Hessian Based Predictor−Corrector Integrator. J. Chem. Phys. 2004, 120, 9918−9924. (23) Montgomery, J. A., Jr.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A. A Complete Basis Set Model Chemistry. VI. Use of Density Functional Geometries and Frequencies. J. Chem. Phys. 1999, 110, 2822−2827. (24) Montgomery, J. A.; Frisch, M. J.; Ochterski, J. W.; Petersson, G. A. A Complete Basis Set Model Chemistry. VII. Use of the Minimum Population Localization Method. J. Chem. Phys. 2000, 112, 6532− 6542. (25) Malick, D. K.; Petersson, G. A.; Montgomery, J. A., Jr. Transition States for Chemical Reactions I. Geometry and Classical Barrier Height. J. Chem. Phys. 1998, 108, 5704−5713. (26) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb, M. A.; Cheeseman, J. R.; Scalmani, G.; Barone, V.; Mennucci, B.; Petersson, G. A. et al. Gaussian 09, revision A.1; Gaussian, Inc.: Wallingford, CT, 2009. (27) Barker, J. R. Multiple-Well, Multiple-Path Unimolecular Reaction Systems. I. MultiWell Computer Program Suite. Int. J. Chem. Kinetics 2001, 33, 232−245. (28) Barker, J. R. Energy Transfer in Master Equation Simulations: A New Approach. Int. J. Chem. Kinetics 2009, 41, 748−763. (29) Barker, J. R.; Ortiz, N. F.; Preses, J. M.; Lohr, L. L.; Maranzana, A.; Stimac, P. J.; Nguyen, T. L.; Kumar, T. J. D. MultiWell-2013 Software;University of Michigan: Ann Arbor, MI (http://aoss.engin. umich.edu/multiwell/), 2013. (30) Holbrook, K. A.; Pilling, M. J.; Robertson, S. H. Unimolecular Reactions, 2nd ed.; Wiley: Chichester, U.K., 1996. (31) Forst, W. Unimolecular Reactions. A Concise Introduction; Cambridge University Press: Cambridge, U.K., 2003. (32) Forst, W. Unimolecular Rate Theory Test in Thermal Reactions. J. Phys. Chem. 1972, 76, 342−348. (33) Stein, S. E.; Rabinovitch, B. S. Accurate Evaluation of Internal Energy Level Sums and Densities Including Anharmonic Oscillators and Hindered Rotors. J. Chem. Phys. 1973, 58, 2438−2445. (34) Beyer, T.; Swinehart, D. F. Number ofMultiply-Restricted Partitions. Commun. ACM 1973, 16, 379−379. (35) Miller, W. H. Tunneling Corrections to Unimolecular Rate Constants, with Application to Formaldehyde. J. Am. Chem. Soc. 1979, 101, 6810−6814. (36) Weston, R. E., Jr.; Nguyen, T. L.; Stanton, J. F.; Barker, J. R. HO + CO Reaction Rates and H/D Kinetic Isotope Effects: Master Equation Models with ab Initio SCTST Rate Constants. J. Phys. Chem. A 2013, 117, 821−835.

(37) 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 II − Gas Phase Reactions of Organic Species. Atmos. Chem. Phys. 2006, 6, 3625−4055. (38) Barker, J. R.; Yoder, L. M.; King, K. D. Feature Article: Vibrational Energy Transfer Modeling of Non-Equilibrium Polyatomic Reaction Systems. J. Phys. Chem. A 2001, 105, 796−809. (39) Hippler, H.; Troe, J.; Wendelken, H. J. Collisional Deactivation of Vibrationally Highly Excited Polyatomic Molecules. II. Direct Observations for Excited Toluene. J. Chem. Phys. 1983, 78, 6709− 6717. (40) Washida, N. Reaction of Methyl Radicals with O(3P), O2, NO and O3. Proc. Yamada Conf. Free Radicals 1979, 3, 271. (41) Slagle, I. R.; Feng, Q.; Gutman, D. Kinetics of the Reaction of Ethyl Radicals with Molecular Oxygen from 294 to 1002 K. J. Phys. Chem. 1984, 88, 3648−3653. (42) Ruiz, P. R.; Bayes, K. D. Rates of Reaction of Propyl Radicals with Molecular Oxygen. J. Phys. Chem. 1984, 88, 2592−2595. (43) Da Silva, G. Atmospheric Chemistry of 2-Aminoethanol (MEA): Reaction of the NH2·CHCH2OH Radical with O2. J. Phys. Chem. A 2013, 116, 10980−10986.

2186

dx.doi.org/10.1021/jp411238e | J. Phys. Chem. A 2014, 118, 2176−2186