Article pubs.acs.org/JPCA
Theoretical Study of the Hydrogen Abstraction of Substituted Phenols by Nitrogen Dioxide as a Source of HONO Abraham Shenghur, Kevin H. Weber, Nhan D. Nguyen, Watit Sontising, and Fu-Ming Tao* Department of Chemistry and Biochemistry, California State University, Fullerton, California 92834, United States S Supporting Information *
ABSTRACT: The mild yet promiscuous reactions of nitrogen dioxide (NO2) and phenolic derivatives to produce nitrous acid (HONO) have been explored with density functional theory calculations. The reaction is found to occur via four distinct pathways with both proton coupled electron transfer (PCET) and hydrogen atom transfer (HAT) mechanisms available. While the parent reaction with phenol may not be significant in the gas phase, electron donating groups in the ortho and para positions facilitate the reduction of nitrogen dioxide by electronically stabilizing the product phenoxy radical. Hydrogen bonding groups in the ortho position may additionally stabilize the nascent resonantly stabilized radical product, thus enhancing the reaction. Catechol (ortho-hydroxy phenol) has a predicted overall free energy change ΔG0 = −0.8 kcal mol−1 and electronic activation energy Ea = 7.0 kcal mol−1. Free amines at the ortho and para positions have ΔG0 = −3.8 and −1.5 kcal mol−1; Ea = 2.3 and 2.1 kcal mol−1, respectively. The results indicate that the hydrogen abstraction reactions of these substituted phenols by NO2 are fast and spontaneous. Hammett constants produce a linear correlation with bond dissociation energy (BDE) demonstrating that the BDE is the main parameter controlling the dark abstraction reaction. The implications for atmospheric chemistry and ground-level nitrous acid production are discussed.
■
morning.8 Direct emission of HONO from combustion processes and biomass burning is also known to occur;6,9−11 nevertheless, they are believed to contribute a small amount to the total HONO production budget,2 up to approximately 10% of the HONO observed.12,13 A considerable amount of research has been conducted on the heterogeneous disproportionation reaction and the role of surfaces.14 Experiment12−15 and theory16 have clearly shown the extent of the homogeneous gas phase reaction to not be appreciable. A key feature of the accepted mechanism for this reaction is the formation of NO2 dimer on a polar surface which can then “activate” to a more reactive species that subsequently reacts. Water clearly plays an important role in this process. Theoretical work has recently indicated that the reactive asymmetric trans dimer can reasonably be formed through an asymmetric cis dimer that can readily isomerize to the active form.17 The well-known sources (i.e., heterogeneous disproportionation and direct emission) can reasonably account for nighttime HONO but still other sources must be in effect, especially during the daytime. HONO may be formed by the reaction of NO2 on the surface of soot;18−26 however, these soot particles appear to deactivate too quickly to account for the HONO observed in the atmosphere.27−29 Additionally, the results are unclear as it is thought that semivolatile and watersoluble organics generated in diesel exhaust may also be
INTRODUCTION Accurate measurements of nitrous acid (HONO) concentrations in the lower troposphere1 have revealed that there is a considerable amount of HONO being produced during the day that had been previously unaccounted for. HONO is now speculated to be a major, if not a dominant source of OH radical in the troposphere during the day, up to ∼60%.2,3 Part of the difficulty in correctly accounting for the HONO concentration during the day is because of the photolabile nature of this molecule (reaction 1).4 HONO(g) + hν(φ = 1, λ < 400 nm) → HO(g)+ NO(g) (R1) •
•
Once formed, OH and NO radicals can participate in a wide range of reactions, from acid formation to particulate matter and ozone formation.5 Importantly, the hydroxyl radical is the primary oxidant in the troposphere and as such its concentration determines much of the chemistry of our atmosphere. Despite the importance of HONO in the atmosphere, mainly as a result of complexity of the system, the sources of formation have yet to be fully understood.6 First detected in 1979,7 HONO has long been known to be a sink for OH radical and to build up during the night by the heterogeneous disproportionation reaction of nitrogen dioxide (NO2) and water to produce nitric acid and nitrous acid (reaction 2). 2NO2(ads)+H 2O(ads) → HNO3(ads) + HONO(g)
(R2)
Received: August 22, 2014 Revised: October 10, 2014
The accumulation of HONO is proportional to NO2 at night, and is photolyzed to emit hydroxyl radical in the early © XXXX American Chemical Society
A
dx.doi.org/10.1021/jp508516c | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
contributing to HONO production.26 It has been demonstrated that polycyclic aromatic hydrocarbons, the primary core of soot particles, can react with NO2 in a heterogeneous matter30 although the role of humidity is little understood. Another possible formation mechanism of HONO involves the participation of ammonia (NH3) in the heterogeneous disproportionation reaction. Field studies have observed a correlation between ammonia and HONO levels31,32 and theoretical work from this group has indicated the formation of HONO and ammonium nitrate should be spontaneous,33 and thus to be an important contributor to the total HONO budget (reaction 3).
Indeed, fertilized soils with low pH have been identified as strong sources of HONO and OH.43 The photolysis of nitrates and nitrites to form nitrogen dioxide is a process that may significantly increase the rate of nitrogen dioxide reduction during the day in natural places such as seawaters or at the surface of soils.48 This “HONO from below” is proving to be an important missing puzzle piece for understanding the daytime production of nitrous acid. In the atmosphere, phenols are typically found as a result of combustion processes or the secondary oxidation of aromatic pollution. During the daytime, when OH levels are significant, phenol is rapidly oxidized to compounds such as catechol. In this reaction, the addition of the OH radical to the aromatic ring is believed to be the dominant process, ∼95% of the reaction branch by one estimate.49 During the night, phenols are believed to react primarily with NO3 radical to form selectively ortho-nitrophenol in the gas phase, and both the ortho- and para-nitrophenols in aqueous solutions.50 The observed regioselectivity is the result of a lone pair of the phenolic oxygen donating electron density to the aromatic ring, thus directing to the ortho- and para- positions through resonance. The initiation step for the gas phase reaction with NO2 is believed to be the same as with NO3, the abstraction of hydrogen from the phenol forming a phenoxy resonance stabilized radical (RSR). This process is depicted in Scheme 1. Depending on the nature of phenol (and also added chemical moieties) this reaction can proceed even in the dark and can be greatly accelerated in the presence of light.
2NO2(g)+H 2O(g) + NH3(g) → NH4 +NO3−(s) + HONO(g) (R3)
Alternatively, HONO can be formed from the photolysis of gas phase ortho-nitrophenols34 and possibly methyl-substituted nitroaromatics.35 Aerosols, fog droplets, and additionally ice containing humiclike substances (HumLiS) have been found to reduce NO2 to HONO.36−39 However, the formation of HONO of the surfaces of these atmospheric aerosols either have a very small uptake coefficients or rapidly deactivate as found with the soot particles and are thus expected to contribute little to daytime HONO production.40 Instead, vertical profiling40 and gradient measurements41 have pointed toward ground sources as being strong emitters of HONO. An experimental study by Wong et al. indicated that a considerable amount of HONO can be produced close to the ground during the day42 and additionally equilibrium exchange between the soil-air interfaces has been shown to contribute to a large amount of HONO released.43 It was suspected that humic acid might be involved in HONO production since it is ubiquitous on the ground, representing a large fraction of water-soluble organics in the soil. Additional field observations43−45 have supported the conclusion that humic acid is a key component in the missing HONO source puzzle. Humic acids are formed naturally from decomposing organic matter such as lignin. These organic acids have no definite structure, but what all humic substances, including the smaller and more highly oxidized fulvic acids, and even smaller still HumLiS residues, have in common is that they are primarily composed of oligomeric phenol residues which, along with the considerable amount of carboxylic acid substituents, determine their chemical reactivity. Indeed, substituted phenols have been shown to reduce NO2 to HONO46 and furthermore, it was subsequently demonstrated that humic acid films exhibit enhanced activity toward NO2 when illuminated with visible light, a photochemical reaction.44 More recently, Sosedova et al.47 employed poly phenolic films to replicate humic acid that also produced a comparable yield of HONO as applied to an upper atmospheric boundary layer. A key for understanding this reaction on the molecular level is that the active reagent is the NO2 molecule. Much work has been done in the field of organic chemistry on the nitration of aromatic molecules, such as phenol; however, becuase of the difficulties in preparing and utilizing NO2 gas as a reagent, little work had been accomplished in this particular area. NO2, and NOx in general, is mainly associated with anthropogenic pollution but can also be formed from the photolysis of nitrites and nitrates which are constantly being produced biologically by microbial bacteria as part of the nitrogen cycle, and additionally as introduced to soils by humankind as fertilizers.
Scheme 1. Depiction of the Initiation Step in the Dark Reaction of Phenols and Nitrogen Dioxide
Here, in this work, we focus on the dark hydrogen abstraction reaction of phenols and nitrogen dioxide directly forming HONO in one kinetic step, and the effects of nuclear substitutions at the ortho, meta, and para positions, and combinations thereof, to gain further insight on the nature of this reaction.
■
THEORETICAL METHODS The abstraction of hydrogen from phenols by nitrogen dioxide was investigated in this work by density functional theory utilizing the Gaussian 2009 program package suite.51 The equilibrium geometries of reactant complexes, transition states, and product complexes were calculated with Becke’s three parameter exchange function combined with Lee, Yang, and Parr’s correlation functional (B3LYP)52−55 which has been successfully applied to the study of organic systems.56 The split valence Pople basis set 6-311++G(d),57 which is reasonably accurate and computationally affordable, was employed for all calculations. Molecular clusters, including reactant and product complexes (RC and PC, respectively), were assembled from the optimized monomers. Vibrational frequency calculations were performed (without scaling) to verify the molecular clusters to be real local minima (possessing all positive real frequencies) or B
dx.doi.org/10.1021/jp508516c | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
Figure 1. Hydrogen bound reaction complexes (RCs) and product clusters (PCs) in the hydrogen abstraction reaction of PhOH by NO2 to produce trans-HONO.
Figure 2. Geometries and singly occupied molecular orbitals (SOMOs) for the hydrogen atom transfer (HAT) [TS-A] and the proton coupled electron transfer (PCET) [TS-B] mechanisms that lead to trans-HONO in the hydrogen abstraction reaction between phenol and nitrogen dioxide.
while at the same time being computationally more affordable.60 Gibbs free energies of reaction, ΔG0, are reported at the standard state of 1 atm and 298.15 K and are calculated as the difference between infinitely separated products and reactants. Electronic activation energies, Ea, are taken to be the single point energy of the cluster at the transition state relative to the infinitely separated reactants. Zero-point-energycorrected electronic energies changes are denoted as ΔE0.
transition states (having a single imaginary frequency corresponding to the reaction coordinate). The transition state (TS) of the parent phenol reaction was additionally verified by an internal reaction coordinate calculation. All the calculations performed in this work were found to be free from spin-contamination with open-shell systems having S2 values very close to, if not exactly, 0.7500. The reported activation energies, Ea, were determined as the energy difference between the transition state and the isolated reactants. Bond dissociation energies (BDEs) were evaluated for the phenols considered in this body of work. The B3LYP/6311++G(d) differences in electronic energies predict an OH bond strength of 85.2 kcal mol−1 for phenol compared to an experimentally determined value of 87.0 kcal mol−1.58 As the energy barrier is closely related to bond dissociation enthalpy59 it was the method employed to produce the reported BDE in this work. Although the B3LYP method may slightly underestimate bond energies, it is known to be relatively free of spin contamination and can produce results similar to CASSCF
■
RESULTS AND DISCUSSION Phenol and NO2. The initial approach of phenol (PhOH) and NO2 prior to reaction can occur with several different configurations. Figure 1 depicts the hydrogen bound reactant complexes (or clusters) RC-A and RC-B, which are both precursor to trans-HONO. The zero point corrected electronic interaction energies at 0 K [ΔE0(RC-A) = −1.2 kcal mol−1, ΔE0(RC-B) = −1.2 kcal mol−1] and interaction free energy changes at 1 atm and 298.15 K [ΔG0298 K(A) = 5.0 kcal mol−1, C
dx.doi.org/10.1021/jp508516c | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
Figure 3. Hydrogen bound reaction complexes (RCs) and product clusters (PCs) in the hydrogen abstraction reaction of PhOH by NO2 to produce cis-HONO.
Figure 4. Geometries and singly occupied molecular orbitals (SOMOs) for the hydrogen atom transfers (HAT mechanism) transitions states, [TSC] and [TS-D], that lead to the production of cis-HONO in the hydrogen abstraction reaction between phenol and nitrogen dioxide.
ΔG0298 K(B) = 5.8 kcal mol−1] indicate a weak hydrogen bond is formed, however, there is a considerable entropic barrier to formation at elevated temperatures, which is consistent with the known negative temperature dependence of these reactions with NO2. In both RC-A and RC-B, the NO2 molecule is coplanar with the phenol. RC-A has a phenolic O atom to the ON of NO2 angle of 115° and in RC-B that same angle is 142°. The geometric difference between these two clusters is simply the orientation of the NO2 molecule as relatively “up” or “down” precursor complexes for trans-HONO production. Figure 1 also depicts the product clusters PC-A and PC-B that result from the abstraction of hydrogen from PhOH by NO2 in the reactant clusters RC-A and RC-B, respectively. In both PC-A and PC-B, trans-HONO is hydrogen bound to the
semiquinone radical in which the spin density is mainly located at the para position (directly opposing the carbon attached to the OH group). The zero point corrected electronic energies [ΔE0(PC-A) = −0.8 kcal mol−1, ΔE0(PC-B) = −0.8 kcal mol−1] indicate that this reaction is fundamentally favorable electronically, yet the predicted free energy changes at 298 K [ΔG0 (PC-A) = 7.8 kcal mol−1, ΔG0(PC-B) = 8.3 kcal mol−1] indicate that entropic effects play an important role and the reaction is not significant in the gas phase at the room temperatures. Figure 2 displays the transition state geometries of the transition states (TS-A and TS-B) located connecting the molecular clusters RC-A and PC-A and, likewise, RC-B and PCB. The calculated zero-point-corrected activation energies D
dx.doi.org/10.1021/jp508516c | J. Phys. Chem. A XXXX, XXX, XXX−XXX
The Journal of Physical Chemistry A
Article
Table 1. Thermodynamic Energy Changes (kcal mol−1), Dipole Moments (debye), OH Stretching Vibrations (cm−1), Associated Infrared Intensities (in Parentheses km mol−1), v(OH) Redshifts (cm−1) with IR Enhancements in Parentheses, and Interatomic Distances for the Hydrogen Bonds and Transition State Separations (Å) for the Molecular Clusters Considered in This Investigation of the Hydrogen Abstraction from Phenol by Nitrogen Dioxide complexes
ΔE
ΔE0
ΔH
ΔG
μ
ν(OH)
RC-A RC-B RC-C RC-D TS-A TS-B TS-C TS-D PC-A PC-B PC-C PC-D
−1.88 −1.46 −1.50 −1.24 13.40 12.49 12.08 12.46 −0.77 −1.25 1.10 0.29
−1.22 −1.15 −0.76 −0.69 11.13 10.34 9.54 9.95 −0.50 −0.72 1.20 0.65
−0.46 −0.86 −0.04 −0.11 11.06 10.16 9.41 9.75 −0.09 −0.40 1.59 0.94
4.96 5.82 5.72 5.48 20.69 20.42 19.38 20.26 7.76 8.27 9.40 9.64
1.55 2.03 1.92 2.05 7.44 6.71 6.29 6.61 7.65 7.49 7.19 6.82
3764 (240) 3773 (202) 3767 (117) 3769 (130) −1874 (13297) −1788 (6761) −1757 (6684) −1791 (8110) 3385 (1447) 3364 (1425) 3262(1000) 3219 (1099)
[ΔE0(TS-A) = +11.1 kcal mol−1, ΔE0(TS-B) = +10.3 kcal mol−1] indicate that both of these pathways are kinetically accessible and equally important, in the gas phase. This is somewhat surprising when comparing the singly occupied molecular orbitals (SOMOs) of these transition states. A comparison of the SOMO of TS-A and TS-B is also presented in Figure 2. Note, in the TS-A structure that the electronic activity of PhOH is localized on the lone pair of the PhOH oxygen and the aromatic ring is not involved in the SOMO. The orbitals appear to align in a manner consistent with a HAT mechanism.61 In TS-B the aromatic ring is extensively participating in the SOMO and the orbitals involved around the proton transfer are indicative of PCET.61 The change can be attributed to the orientation of the HOMO of NO2 as “up” or “down” in the formation of the reaction cluster. Despite of the relatively low values of the activation energy, because of the large positive change in free energy for the overall reaction, the reaction of phenol with NO2 would not be expected to proceed to an appreciable extent, consistent with experiment. Figure 3 presents the hydrogen bound reaction and product clusters that lead to production of cis-HONO. In the cisHONO pathways, reaction clusters are weakly bound [ΔE0(RC-C) = −0.8 kcal mol−1, ΔE0(RC-D) = −0.7 kcal mol−1] similar to trans-HONO precursor clusters and the free energy changes [ΔG0(TS-A) = 20.7 kcal mol−1, ΔG0(TS-B) = 20.4 kcal mol−1] again are substantially positive, indicating a negative temperature dependence for the reaction. Product clusters were located with the cis-HONO hydrogen bound to the semiquinone radical with the oxygen of HONO doubly bound to nitrogen either associating with an aromatic C−H group or orientated in the opposite direction, minimizing all steric interactions. These product clusters are found to have small, positive electronic energy change compared to the infinitely separated reactant monomers (i.e., they are slightly uphill compared to the trans-HONO pathway) with [ΔE0(PCA) = 1.2 kcal mol−1, ΔE0(PC-D) = 0.7 kcal mol−1]. The free energy changes are more positive than the trans-HONO precursors by approximately 1−1.5 kcal mol−1 [ΔG0(PC-C) = 9.4 kcal mol−1, ΔG0(PC-D) = 9.6 kcal mol−1]. The transition states connecting RC-C and RC-D to PC-D are shown in Figure 4. Despite extensive searching a reasonable transition state for RC-C to PC-C was not located, instead the NO2 molecule in PC-C prefers to “flip” into the PC-D avoiding a large energy barrier to reaction. The corrected electronic
Δν(OH)
R(OH···O)
−23 (202) −14 (194) −20 (79) −18 (92)
2.170 2.214 2.224 2.252 1.178, 1.168, 1.163, 1.164, 1.788 1.770 1.819 1.791
−340 (1395) −361 (1373) −279 (989) −322 (1088)
1.256 1.273 1.268 1.261
activation energies [ΔE0(TS-C) = 9.5 kcal mol−1, ΔE0(TS-D) = 10.0 kcal mol−1] are approximately 1 kcal mol−1 lower than the trans-HONO pathways, and free energy changes [ΔG0(TS-C) = 19.4 kcal mol−1, ΔG0(TS-D) = 20.3 kcal mol−1] are only slightly attenuated compared to the trans-HONO TSs. The SOMOs for TS-C and TS-D are included in Figure 4 and are clearly hydrogen atom transfers (HATs). Although the higher energy state of the cis-HONO compared to trans-HONO results in a more positive overall reaction free energy change, which appears to be reflected in the product clusters, the transition states for the cis-HONO formations appear to be more readily accessible (i.e., these reactions should occur somewhat faster, in both directions). Table 1 summarizes the electronic energy changes both uncorrected (ΔE) and corrected for zero point energy (ΔE0), enthalpy and free energy changes at standard conditions (ΔH and ΔG), dipole moments (μ), and OH stretching frequencies, ν(OH), with infrared intensities in parentheses, for the reactant, product, and transition state molecular assemblages. All reaction complexes are found to associate mildly (
