Atmospheric Oxidation Mechanism of Furfural Initiated by Hydroxyl


by coupling the unimolecular rate theory and master equation for collisional energy transfer (RRKM-ME)(20) as being implemented in the Mesmer code...
0 downloads 0 Views 456KB Size


Subscriber access provided by University of Missouri-Columbia

Article

The Atmospheric Oxidation Mechanism of Furfural Initiated by Hydroxyl Radicals Xiaocan Zhao, and Liming Wang J. Phys. Chem. A, Just Accepted Manuscript • DOI: 10.1021/acs.jpca.7b00506 • Publication Date (Web): 10 Apr 2017 Downloaded from http://pubs.acs.org on April 10, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

The Journal of Physical Chemistry A is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Revision Submitted to JPCA

The Atmospheric Oxidation Mechanism of Furfural Initiated by Hydroxyl

Radicals

Xiaocan Zhao a and Liming Wang* a,b a

School of Chemistry & Chemical Engineering, South China University of Technology, Guangzhou 510640,

China;

b

Guangdong Provincial Key Laboratory of Atmospheric Environment and Pollution Control, South

China University of Technology, Guangzhou 510006, China.

Abstract Furfural is emitted into the atmosphere because of its potential applications as an

intermediate to alkane fuels from biomass, industrial usages, and biomass burning. The kinetic

and mechanistic information of the furfural chemistry is necessary to assess the fate of furfural in the atmosphere and its impact on the air quality. Here we studied the atmospheric oxidation mechanisms of furfural initiated by the OH radicals using quantum chemistry and kinetic

calculations. The reaction of OH and furfural was initiated mainly by OH additions to C2 and C5

positions, forming R2 and R5 adducts, which could undergo rapid ring-breakage to form R2B

and R5B, respectively. Our calculations showed that these intermediate radicals reacted rather

slowly with O2 under the atmospheric conditions because the additions of O2 to these radicals

are only slightly exothermic and highly reversible. Alternatively, these radicals would react

directly with O3, NO2, and/or HO2/RO2 etc. Namely, the atmospheric oxidation of furfural

would unlikely result in ozone formation. Under typical atmospheric conditions, the main products

in

OH-initiated

furfural

oxidation

include

2-oxo-3-pentene-1,5-dialdehyde,

5-hydroxy-2(5H)-furanone, 4-oxo-2- butenoic acid, and 2,5-furandione. These compounds will likely stay in the gas phase and are subject to further photo-oxidation.

1

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1. Introduction

Furan compounds, being obtainable through hydrolysis-dehydration of cellulose and

hemicellulose,1-2 are important platform compounds in producing renewable liquid fuels and chemicals. For example, furfural can be transformed by hydrogenation into 2-methylfuran, an

additional agent for gasoline, and the aldol condensation of furfural and acetone can be used to produce alkane fuels in large quantities.3 The rise of bio-based industry has made furfural

production from lignocellulosic biomass to be a point of focus of many recent studies. Furfural

is also a proverbial natural compound, and is applied widely in the vegetable oil, plastics and rubber industries, and pharmaceutical industry, etc. In this context, production of furfural from

renewable agricultural sources has reached an annual amount of ~300 kTon worldwide with

China being the largest producer (~70%).4 With its extensive usage, furfural is emitted into the

atmosphere, where its oxidation may lead to ozone and secondary organic aerosol formation.

Mechanistic information of furfural in the atmosphere is needed for assessing its fate and its impact on air quality. Field measurement had also identified the presence of furfural from biomass burning in the atmospheric aerosols.5

Atmospheric oxidation of furfural is likely initiated by photolysis, reactions with OH and

NO3, and to a less extent, with O3. Rate coefficients of (3.51 ± 0.01) × 10–11 cm3 molecule–1 s–1 and (1.20 ± 0.28) × 10–12 cm3 molecule–1 s–1 were reported for its reaction with OH radical

(296 K)

6

and NO3 radical (298 K),7 respectively, indicating an average lifetime of ~5 h for

furfural during daytime due to its reaction with the OH radicals or ~0.4h during nighttime due

to its reaction with NO3 radicals. No other information is available on the reaction kinetics and mechanism for the atmospheric oxidation of furfural. This makes it difficult to assess the impact of furfural emission on air quality. In this work, we report a theoretical study on the

atmospheric oxidation mechanism of furfural initiated by OH radicals. We have identified the

main oxidation reaction pathways and products using quantum chemistry and kinetic calculations.

2. Theoretical Theoretical Methods

The molecular structures were optimized and the vibrational frequencies were calculated

at the M06-2X/6-311++G(2df,2p) level, which was found to be suitable for kinetics 2

ACS Paragon Plus Environment

Page 2 of 18

Page 3 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

calculations.8 The single-point electronic energies of the optimized structures were calculated

by using the ROCBS-QB3 model chemistry (QB3 in short),9 which has an uncertainty of ~4

kJ/mol in the predicted energy and barrier height, and by using the explicitly correlated method at RHF-UCCSD(T)-F12a/cc-pVDZ-F12 level (F12a in short), which can reach at least

the conventional CCSD(T)/cc-pVQZ quality of electron correlation with much less computational cost.10-11 It was noticed that thee T1-diagnostics in UCCSD calculations exceeded 0.02 for a few transition states of O2-addition, indicating the multireference nature of the wave

function and requesting multireference calculations,12-13 which are, however, difficult for these

conjugated systems. The M06-2X and QB3 calculations were performed using the Gaussian 09 package14 and the F12a ones using the Molpro 2015 package.15-16

The reaction rate constants at high-pressure limit were estimated by using the traditional

transition states,17-18

N =P∙R∙

bcd

NS T ∆Y GZ [T 10_ ∙ exp V− \∙V ° ∙ \ ℎ [T ] `a

where NS is the Boltzmann constant, ∆Y GZ is the Gibbs energy barrier, P is the reaction path

degeneracy, e is 1 or 2 for uni- or bi-molecular reaction, and R is the tunneling correction factor which was calculated using the asymmetric Eckart model.19

The pressure-dependence of the kinetics for several reaction steps was modeled by

coupling the unimolecular rate theory and master equation for collisional energy transfer (RRKM-ME)20 as being implemented in the Mesmer code,21 in which the tunneling correction

factor is evaluated according to Miller.22 Single exponential down model was used to simulate

approximate the collisional energy transfer with 〈∆g〉ijkl of 250 cm−1. The collision frequencies were estimated by the method of Gilbert and Smith.23

3. Results and Discussion 3.1 The Initial Initial Steps

In the atmosphere, reaction of furfural with the OH radicals proceeds as OH addition to the

unsaturated bonds, forming adducts R2~R5, or as H-abstraction from –CHO,

3

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

We have located the transition states for OH additions (TS2~TS5) and H-abstraction (TS-RCO) at M06-2X level, and calculated the energies at QB3 and F12a levels (Table 1). Figure 1 displays

the potential energy surfaces of the reaction. Transition states for additions to C2 and C5 and H-abstraction are lower than the separated reactants because of the existence of pre-reactive

complex (PRC), which is ~22 kJ/mol (QB3) below the separated reactants. Excellent

agreement can be observed between QB3 and F12a energies and barrier heights with differences all within 3 kJ/mol except for the relative energy of R5. Discussion below are all based on QB3 energies unless otherwise stated.

The high-pressure-limit rate coefficient at 298 K was determined as 3.48 × 10–11 cm3

molecule–1 s–1 with QB3 barrier heights, or as 2.47 × 10–11 cm3 molecule–1 s–1 with F12a ones

(Table 1), both agreeing quantitatively with the experimental value6 of (3.51 ± 0.11) × 10–11 cm3 molecule–1 s–1 and suggesting dominant additions to C2 and C5 while negligible additions to

C3 and C4. The discrepancy between QB3 and F12a was only found on the predicted branching ratio for H-abstraction, e.g., ~3% with the QB3 barriers heights versus ~11% with the F12a

ones.

As for the reaction between furan and OH radical,24 the chemically activated adducts R2*

and R5* from addition might break the ring before being de-activated by collision,

We have identified transition states for the two ring-opening processes, and obtained barrier

heights of 53.1 kJ/mol and 96.6 kJ/mol for R2 and R5, respectively, at QB3 level. Comparatively for the ring-opening of similar radical in furan, we obtained a barrier height of 78.7 kJ/mol at

QB3 level and Mousavipour et al. reported a value of 77 kJ/mol at CCSD/6-311+G(3df,2p)

level.24 The barrier heights indicate slow ring-opening of R2 and R5, i.e., at high-pressure-limit 4

ACS Paragon Plus Environment

Page 4 of 18

Page 5 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

rates of 1.9 × 103 s–1 and 0.4 s–1, respectively, at 298 K. However, chemically activated R2* and R5* can decompose promptly. RRKM-ME simulations showed that the relative fractions of R2 and R2B reached thermodynamic equilibrium within 10–8 s of reaction time, which is too short

for R2 to recombine with the atmospheric oxygen (Figure S1). So are the relative fractions of R5 and R5B (Figure S2). At 76o Torr and in the temperature range of 243~313 K, calculations

showed that R2B was the dominant product for addition of OH to C2-position (> 98%); while

for addition to C5-position, production of R5B increased with temperature, from 5% of the

C5-addition at 243 K, 7.5% at 273 K, 10% at 298 K, and to 13% at 313 K (Figure S2). The fate of these radicals was examined below. 3.2 Reactions Reactions of R2 and R5 with O2

Because of the electron delocalization in R2, O2 can add to C3 and C5 positions, forming four

peroxy radicals R2-3OO-a/s and R2-5OO-a/s (a/s = anti/syn represents the relative directions

of –OH and -OO with respect to the ring). So do the additions of O2 to R5. In the atmosphere, these peroxy radicals can possibly decompose back to R2/R5 + O2, isomerize unimolecularly,

and react bimolecularly with NO/HO2/RO2 as

where NSn is the bimolecular removal rate, and Noln is the rate for possible unimolecular processes. Tables 2 lists the reaction energy, barrier heights, and high-pressure-limited rate

constants for these additions (see Tables S1 and S2 for temperature dependence), and Figure 2 shows the potential energy surfaces for reaction of R5 with O2 and the ensuing steps (Figure S3

for reaction of R2). The barrier heights for O2 additions might have high uncertainty as being

inferred from T1-diagnostic of higher than 0.035 in UCCSD calculations. Oxygen addition localizes the delocalized electron in R2 and R5; therefore the stabilization energy of ~70

kJ/mol for additions to R2 and of ~40 kJ/mol for R5 are much lower than the stabilization energy of >100 kJ/mol for the ‘normal’ alkyl radicals. Energies of the transition states for O2 5

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

additions are again lower than the separated reactants due to the formation of pre-reaction complexes between R2/R5 and O2.

We have explored possible unimolecular processes for these peroxy radicals and estimated

their rates (Table S3). The rates obtained here might be about one order of magnitude higher

or lower due to the uncertainties in barrier height and tunneling correction factor. For

H-migrations, the tunneling correction factors are highly sensitive to the imaginary vibrational

frequency associated with the transition mode. Table 2 lists the rates at 298 K for the fastest unimolecular process of the corresponding peroxy radical for O2 additions. Assuming steady state for the peroxy radicals, we could obtain the effective bimolecular removal rates for R2 and R5 via O2 addition in the atmosphere as Nrss =

Nt (NSn + Noln ) Nu + NSn + Noln

For all peroxy radicals except for R2-3OOa, Nu ’s were orders of magnitude higher than

NSn + Noln at all temperatures, resulting in Nrss ≈ Nt (NSn + Noln )⁄Nu = xry (NSn + Noln ) , in

which xry ’s are the equilibrium constants (Tables 2, S1, and S2) and NSn is proportional to [NO] and/or [HO2]. When NSn ≥ 1 scd (with [NO] of a few ppbv or higher), the rates in Table 2

suggested that R2 and R5 were removed via the stable peroxy radicals R2-5OO-a/s and R5-2OO-s, respectively, as

Nrss,u} ≈ ~xEq,R2−5OOa + xEq,R2−5OOs (NBi + NUni ) Nrss,u€ ≈ xEq,R5−2OOs (NBi + NUni )

When NSn is negligibly small (with extremely low level of [NO] and [HO2]), we have, Nrss,u} ≈ xEq,R2−3OOa NUni,R2−3OOa + xEq,R2−5OOa NUni,R2−5OOa Nrss,u€ ≈ xEq,R5−2OOa NUni,R5−2OOa + xEq,R5−2OOs NUni,R5−2OOs

Typical atmospheric conditions are close to the high [NO] scenarios.

For radical R2, at relatively high NSn (> 1 s–1), we obtained Nrss,}‚ƒ of ~NSn × 10–13 cm3

molecule–1 s–1 (via R2-5OOa and R2-5OOs), or an effective removal rate of ~5 × 105 s–1 with

[O2] of ~5 × 1018 molecule cm–3 and NSn of 1 s–1. The reaction of R2-5OOa/-5OOs with NO would form R2-5Oa/-5Os (ΔE0K ~ –20 kJ/mol) radicals, which will react with O2 as

6

ACS Paragon Plus Environment

Page 6 of 18

Page 7 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

The ring breakages of R2-5Oa/-5Os, with barrier heights of more than 80 kJ/mol, are too slow

to compete with the bimolecular reaction with O2. At relatively low NSn (< 10–2 s–1), we obtained Nrss,}‚ƒ of ~3 × 10–14 cm3 molecule–1 s–1 (via R2-5OOa) as

Note that formation of R2 radical is a rather minor channel. For R5, at relatively high NSn (> 1 s–1), we obtained Nrss,}‚ƒ of ~NSn × 10–19 cm3

molecule–1 s–1 (via R5-2OOs), or an effective removal rate of ~0.5 s–1 with [O2] of ~5 × 1018 molecule cm–3 and NSn of 1 s–1. Under high [NO] conditions, we had

in which we could not obtain a structure for the intermediate radical because optimization

indeed resulted in a complex between the two fragments. At relatively low NSn (< 10–2 s–1), we

obtained Nrss,}‚ƒ of ~10–20 cm3 molecule–1 s–1 (via R5-2OOa and R5-2OOs), or an effective

removal rate of ~0.05 s–1. Under these conditions (pristine atmosphere in remote area), R5 would be a relatively long-lived radical. The main unimolecular process in R5-2OO-s/a might be the H-migrations as

where the elimination of CO is extremely fast with a barrier less than 20 kJ/mol. This would be the dominant product channel when NSn is low.

The slow removal of R5 in the atmosphere by O2 suggests direct reactions of R5 with other

atmospheric trace species as

where XO could possibly be ONO, HOO, or O3, for which the reactions are exothermic with ΔE0K of –73.5, –110.5, and –294.0 kJ/mol, respectively. Assuming the similar rate constants as the 7

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

reactions of CH3 radical (in cm3 molecule–1 s–1, N†‡ ~ 2 × 10–12 and Nˆ†} ~ 2 × 10–11),25 radical

R5 reacts more likely with O3 and/or NO2 under typical atmospheric conditions of [O3] ~ 1012 molecule cm–3 (40 ppbv) or [NO2] ~ 2.5 ×1011 molecule cm–3 (10 ppbv) when NO is low. The process also formed 5-hydroxy-2(5H)-furanone (C4H4O3).

3.3 Reactions of RingRing-Opening Radicals R2B and R5B with O2

Our IRC (Intrinsic Reaction Coordinate) calculation showed that the ring opening of R2

formed radical conformer R2B-Z1, which could connect to other conformers via internal

rotations along the C2-C3, C3-C4, and C4-C5 bonds (Scheme 1). Inter-conversion between

conformers is expected. In the atmosphere, R2B recombines reversibly with O2, forming peroxy

radicals. Using R2B-Z1 as the prototype, the recombination of R2B with O2 can be viewed as

Tables 2 and S4 list the energetic of the reaction, the barrier heights, and the rate constants, and Figure 3 shows the potential energy surface for reaction between R2B-Z1 and O2, and Figure S4 for reaction between R5B and O2. Because of the electronic resonance in R2B, O2

additions are thermodynamically disfavored with ΔE0K of only –25 and –17 kJ/mol and with barrier height of 3.9 and 8.9 kJ/mol for additions to C2 and C4, respectively. These addition

processes are highly reversible because of the slow rates of ~102 s–1 at 298 K for the possible

forward unimolecular steps in R2B-2OO/-4OO radicals and the rapid decomposition back from R2B-2OO/-4OO to R2B-Z1 + O2 at high-pressure-limit rates of > 107 s–1 at 298 K (Tables 2 and

S4). An effective bimolecular removal rate constants by reactions with O2 were estimated as

~7 ×10–21 cm3 molecule–1 s–1 at 298 K (via R2B-2OO dominantly) or an effective removal rate of ~0.03 s–1 under typical atmospheric conditions (or ~5 ×10–20 cm3 molecule–1 s–1 and ~0.2

s–1 with F12a energies), forming 2-oxo-3-pentene-1,5-dialdehyde (C5H4O3) via R2B-2OO. Note

that the HO2-elimination from R2B-2OO forms a hydrogen-bonded complex and is highly

endothermic. The slow removal of R2B by reactions with O2 suggests that R2B will have an 8

ACS Paragon Plus Environment

Page 8 of 18

Page 9 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

equilibrium distribution within its isomers either via unimolecular inter-conversions as shown in Scheme 1 or via repeated addition-dissociation of O2 as in isoprene-OH-O2 system.26

Again, the slow removal of R2B by its reactions with O2 promotes the importance of its

reactions with NO2, HO2, O3, and other trace radicals in the atmosphere as

where the intermediate radical is indeed a weak complex between 4-oxo-2-butenoic acid (C4H4O3) and HCO. This might be the main product for R2B in the atmosphere. Similarly for the reaction between R5B and O2, we had

Table 2 lists the reaction energies and barrier heights for O2 additions and the unimolecular

rates for the peroxy radicals. Addition to C5 is distinctly different from other additions because

it is more exothermic, and for R5B-5OO, its backward decomposition to R5 + O2 and forward decomposition are comparable. The effective bimolecular reaction with O2 is ~2 × 10–14 cm3

molecule–1 s–1 at 298 K, being orders of magnitude higher than other additions to R2B and R5B;

therefore,

R5B

in

the

atmosphere

2-oxo-3-pentene-1,5-dialdehyde (C5H4O3).

would

react

with

O2

and

form

product

3.4 The Reaction Pathway after H-Abstraction

The reaction pathway after H-abstraction is relatively simple as shown below,

where the intermediate peroxy radicals would react with NO and/or HO2/RO2 while their

unimolecular reaction is unlikely. The product would be 2,5-furandione (C4H2O3) and RONO2 9

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

or ROOH formed in the reactions of various RO2 intermediates and NO/HO2. Villanueva et al. proposed a similar mechanism for the formation of 2,5-furandione (C4H2O3) in the reaction of

furan and Cl-atom.27

3.5 A Note on the Accuracy of the Predicted Reaction Pathways

No doubt the reliability of predicted rate coefficients and branching ratios depends highly

on the accuracy and reliability of the reaction energies and barrier heights. Generally, ROCBS-QB3 has an uncertainty of ~5 kJ/mol in the predicted enthalpies of formation, or the reaction energy in this study.9 The uncertainty in barrier heights might be higher because of

the multi-reference nature of the wavefunction as being assessed by values of T1 diagnostic in

ROCCSD calculations,12-13 particularly for the transition states for O2 additions to R2B, R5, and

R5B which are the dominant radicals formed after the OH addition.

However, as we can find out in Table 2, all the additions to R2B, R5, and R5B are highly

reversible with the back-decomposition rate Nu of > 104 s–1 for the RO2 radicals formed. The

rate Nu is orders of magnitude higher than any other possible unimolecular and bimolecular reactions for RO2 radicals in the atmosphere, except for R5B via R5B-5OO. Therefore, the

effective bimolecular rate coefficients for removal R2B/R5/R5B via RO2 radicals can be

simplified as Nrss ≈ Nt (NSn + Noln )⁄Nu = xry (NSn + Noln ) , namely, the Nrss ’s essentially

depend on the reaction energies, for which ROCBS-QB3 and RHF-UCCSD(T)-F12a perform reasonably well. For R5B via R5B-5OO, Nrss ≈ Nt .

The effective bimolecular rate coefficients for the removal of R2B, R5, and R5B radicals via

RO2 radicals thus obtained are either < 10–20 cm3 molecule–1 s–1 or > 10–14 cm3 molecule–1 s–1, rendering the effective rates of < 0.05 s–1 or > 5 × 104 s–1 in the atmosphere, except for R5 via

R5-2OOs (Table 2). On the other hand, the bimolecular removal rates of R2B, R5, and R5B are ~5 s–1 by reacting with NO2 of ~10 ppbv or ~0.5 s–1 and O3. Therefore, changing rate

coefficients by one order of magnitude due to the uncertainty on the reaction energies would unlikely alter the oxidation routes in the atmosphere when NO is low. Increased NO level

would not alter the route for R2B and R5B because of the fast unimolecular HO2 elimination processes in their peroxy radicals; while for R5, route via R5-2OOs might compete with the reactions of R5 with XO.

10

ACS Paragon Plus Environment

Page 10 of 18

Page 11 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

4. Conclusions

We have investigated the atmospheric oxidation mechanisms of furfural initiated by the

OH radicals using quantum chemistry and kinetic calculations. Scheme 2 shows the proposed mechanism at 298 K based on the calculated results. Calculations suggested that the

intermediate radicals R2, R5, R2B, and R5B would react rather slowly with O2 under the

atmospheric conditions because the additions of O2 were only slightly exothermic and therefore highly reversible due to electron delocalization in these radicals, e.g., the effective

rates via O2 additions are ~0.03, ~105, and ~1.0 s–1 for R2B, R5B, and R5, respectively.

Alternatively, R2B and R5 would react directly with the atmospheric trace species such as O3,

NO2, and HO2/RO2 etc. by transferring one O-atom. By depleting NO2 and O3, furfural oxidation

in the atmosphere may have a negligible or even negative ozone formation potential. Under typical atmospheric conditions, the main products include 2-oxo-3-pentene-1,5-dialdehyde

(C5H4O3) (via R2B and R5B), 5-hydroxy-2(5H)-furanone (C4H4O3) (via R5), 4-oxo-2-butenoic acid (C4H4O3) (via R2B and XO), and 2,5-furandione (C4H2O3) (via H-abstraction channel). In

the atmosphere, these compounds will likely stay in the gas phase, and further studies are desired for the photo-oxidation mechanism of these compounds. ASSOCIATED CONTENT

Supporting Information. Information Figures S1-S4 for the potential energy diagrams and Table S1-S4 for

the barrier heights and rates of various unimolecular reactions of intermediate radicals. The Supporting Information is available free of charge on ACS Publications website at DOI: xx. AUTHOR INFORMATION Corresponding Authors

* Liming Wang, E-mail: [email protected] ORCID Liming Wang: 0000-0002-8953-250X

Notes Notes

The authors declare no competing financial interests. 11

ACS Paragon Plus Environment

The Journal of Physical Chemistry

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACKNOWLEDGMENT

This work was supported by National Natural Science Foundation of China (No. 21477038) and Natural Science Foundation of Guangdong Province (No. 2016A030311005).

12

ACS Paragon Plus Environment

Page 12 of 18

Page 13 of 18

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

The Journal of Physical Chemistry

Table 1. Calculated relative energies (in kJ/mol) and rate constants (in cm3 molecule–1 s–1) for reaction between furfural and OH radicals (a) Species

Fur + OH PRC TS2 TS3 TS5

TS-RCO Expt. R2

∆g‹ƒ

∆Œ}‚ƒ

−21.9

9.9

0.0

0.0

27.8

0.676

7.4

43.5

0.000

38.7

R4 R5

TS2B R2B

TS5B R5B

0.008

25.8

−9.4 −4.5

28.8

−129.8 −90.1

ROCBS-QB3

∆g‹ƒ

∆Œ}‚ƒ

−21.8

11.0

0.0

26.2

1.311

6.0

41.8

0.002

−10.2

0.261

38.0

0.010

25.1 30.9

−91.6

−132.7

−94.5

−52.5

−92.8

−55.2

2.47

N}‚  × 10dd

−10.2 2.2

1.522

0.0

−2.1

3.51 ± 0.11

(b)

R3

N}‚  × 10

dd

−8.7 2.9

TS4

Total

RHF-UCCSD(T)-F12a

2.054 0.099 3.48

3.51 ± 0.11

−55.7

−19.5

−56.6

−20.4

−155.5

−117.5

−159.3

−121.3

−78.4

−39.0

−79.6

−40.3

−148.5

−116.2

−150.1

−117.8

−63.9

−26.4

−64.9

−27.3

−133.7

−99.8

−134.1

−100.2

(a) M06-2X = M06-2X/6-311++G(2df,2p), all at M06-2X geometries and ZPEs; (b) From Bierbach et al.6

Table 2. 2 Calculated reaction energies and barrier heights at ROCBS-QB3 level (in kJ/mol) and the predicted rate constants (Nt and Nrss in cm3 molecule–1 s–1, and Nu and Noln in s–1) (a) Reactions

R2 + O2 → R2-3OO-a R2 + O2 → R2-3OO-s

R2 + O2 → R2-5OO-a R2 + O2 → R2-5OO-s

R5 + O2 → R5-2OO-a R5 + O2 → R5-2OO-s

R5 + O2 → R5-4OO-a R5 + O2 → R5-4OO-s

R2B + O2 → R2B-2OO R2B + O2 → R2B-4OO R5B + O2 → R5B-3OO R5B + O2 → R5B-5OO

∆Ž

∆‘’“”

∆ŽZ

∆‘Z’“”

–63.0

–17.5

−0.8

43.2

–62.7 –77.1 –78.7 –36.6 –45.3

–17.3 –33.8

–16.3

6.0

–3.5

–33.7

–11.5

10.2

13.6

11.8

−25.0

21.8

−17.4

–13.0

–0.6

–31.6 –34.5

0.4

28.9

−18.5

24.9

−45.3

−3.0

10.8 3.9 8.9

26.8 5.1

•–,’“”

44.5

4.06

27.5

3.82 × 10–12

30.7 40.2 33.2 54.1 57.9 51.9 55.6 69.2 46.0

× 10–15

6.72 × 10–15 1.05

× 10–12

2.25 × 10–14 3.86 × 10–13 8.54

× 10–17

1.82 × 10–17 4.07 × 10–16 9.34

× 10–17

3.85 × 10–19 2.20 × 10–14

•—,’“”

•˜™š,’“”

1.45 × 102

~0.04

× 101

10–5

9.43

× 101

1.11 × 102 3.23

6.24 × 106 7.56 × 106 2.47

× 105

2.71 × 104

~10

~5 × 10–17

~1.7 × 10–13

~1.4

~5 × 10–21

~9 × 10–21