Effects of Chemical Complexity on the Autoxidation Mechanisms of

Jan 23, 2015 - Formation of highly oxidized, multifunctional products in the ozonolysis of three endocyclic alkenes, 1- methylcyclohexene, 4-methylcyc...
24 downloads 3 Views 1MB Size
Article pubs.acs.org/JPCA

Effects of Chemical Complexity on the Autoxidation Mechanisms of Endocyclic Alkene Ozonolysis Products: From Methylcyclohexenes toward Understanding α‑Pinene Matti P. Rissanen,*,† Theo Kurtén,‡ Mikko Sipila,̈ † Joel A. Thornton,§ Oskari Kausiala,† Olga Garmash,† Henrik G. Kjaergaard,∥ Tuukka Petaj̈ a,̈ † Douglas R. Worsnop,†,⊥,# Mikael Ehn,† and Markku Kulmala† †

Department of Physics, University of Helsinki, P.O. Box 64, 00014 Helsinki, Finland Department of Chemistry, University of Helsinki, P.O. Box 55, 00014 Helsinki, Finland § Department of Atmospheric Sciences, University of Washington, Seattle, Washington 98195, United States ∥ Department of Chemistry, University of Copenhagen, Universitetsparken 5, 2100 Copenhagen Ø, Denmark ⊥ Department of Applied Physics, University of Eastern Finland, 70211 Kuopio, Finland # Aerodyne Research Inc., 45 Manning Road, Billerica, Massachusetts 01821, United States ‡

S Supporting Information *

ABSTRACT: Formation of highly oxidized, multifunctional products in the ozonolysis of three endocyclic alkenes, 1- methylcyclohexene, 4methylcyclohexene, and α-pinene, was investigated using a chemical ionization atmospheric pressure interface time-of-flight (CI-APi-TOF) mass spectrometer with a nitrate ion (NO3−) based ionization scheme. The experiments were performed in borosilicate glass flow tube reactors at room temperature (T = 293 ± 3 K) and at ambient pressure. An ensemble of oxidized monomer and dimer products was detected, with elemental compositions obtained from the high-resolution mass spectra. The monomer product distributions have O/C ratios from 0.8 to 1.6 and can be explained with an autocatalytic oxidation mechanism (=autoxidation) where the oxygen-centered peroxy radical (RO2) intermediates internally rearrange by intramolecular hydrogen shift reactions, enabling more oxygen molecules to attach to the carbon backbone. Dimer distributions are proposed to form by homogeneous peroxy radical recombination and cross combination reactions. These conclusions were supported by experiments where H atoms were exchanged to D atoms by addition of D2O to the carrier gas flow. Methylcyclohexenes were observed to autoxidize in accordance with our previous work on cyclohexene, whereas in α-pinene ozonolysis different mechanistic steps are needed to explain the products observed.



INTRODUCTION Understanding the details of gas-phase organic oxidation chemistry is essential for controlling and optimizing many physicochemical processes of fundamental scientific interest, with significant socio-economic importance. For example, the energy released upon combustion of a fossil fuel, or any other organic compound, is closely related to its oxygen content (which generally decreases the energy content), as the maximum energy obtainable from an organic compound is generated by breaking its hydrocarbon structure while oxidizing the fuel carbon to carbon dioxide (CO2) through series of radical reaction sequences. Gas-phase processing of volatile organic compounds (VOC) can lead to production of secondary organic aerosol (SOA),1,2 which is often the dominant form of fine mode particulate matter.3,4 For air quality, the gas-phase oxidation processes are crucial not only for SOA generation but also for cleansing the atmosphere from organic trace gases. By oxidation, the water insoluble organic © XXXX American Chemical Society

material is transformed into a more soluble form by inclusion of polar functional groups to the water insoluble hydrocarbon backbone,5 thus enabling their scavenging by liquid droplets and aerosols. In the atmosphere, virtually every hydrocarbon molecule will be converted to a peroxy radical before it is removed from the gas phase. Atmospheric gas-phase organic oxidation chemistry is thus largely a chemistry of peroxy radicals (RO2). Recently, RO2 isomerizations by intramolecular H-transfer reactions have received significant attention, due to being a mechanistic route to many species of potential interest.6,7 Very recently this autooxidation (i.e., autocatalytic radical oxidation mechanism that Special Issue: Mario Molina Festschrift Received: November 1, 2014 Revised: January 4, 2015

A

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

The Journal of Physical Chemistry A

Article



EXPERIMENTAL SECTION The alkene ozonolysis experiments were performed in two uncoated borosilicate glass flow tube reactors connected to a nitrate ion (NO3−) based chemical ionization atmospheric pressure interface time-of-flight (CI-APi-TOF) mass spectrometer. The deployment, basic measurement routines, and data analysis of this instrument have been described previously and will not be repeated here.22,23,29,30 A 205 cm long glass tube with a 4.7 cm i.d. and a 63 cm long with a 4.0 cm i.d. were employed, respectively. The oxidation reactions in the gas mixture were initiated by ozone. The influence of water on the product formation was inspected by adding H2O or D2O to the carrier gas flow. All the experiments were performed under laminar flow conditions, at room temperature (293 ± 3 K) and at ambient pressure. The total gas flow rate was set to about 11 lpm in all experiments, which resulted in around 0.15 and 0.11 m s−1 gas flow velocities, corresponding to about 4 and 19 s residence times in the shorter and longer flow tubes, respectively. No OH-scavengers were applied during these investigations. The amount of organic precursor in the gas flow was estimated from the measured gas flow rates and vapor pressures of the pure compounds. The chemicals 1-methylcyclohexene (97%), 4-methylcyclohexene (99%), α-pinene (98%), and deuterium oxide (D2O, 99.9%) were purchased from SigmaAldrich chemical company and were used without further purification. The bath gases nitrogen (N2, 99.9999% and 99.9996% purity) and synthetic air (N2 + O2, 99.999%) were obtained from Aga and used as supplied. Ozone was produced by an ozone generator (Dasibi 1008-PC) and the produced amount was quantified by an ozone analyzer (Thermo Scientific 49). A schematic of the experimental setup is shown in the Supporting Information. Details of the different experiments performed are given in Table 1.

proceeds by sequential O2 addition + peroxy radical hydrogentransfer isomerizations steps,8−10 a well-known process in food and oil product industry and also in low-temperature combustion chemistry7,11−21) was recognized as a viable gasphase pathway to highly oxygenated organic material under ambient atmospheric conditions.9,22 The oxidation processes were observed to lead to molecular weight growth chemistry by oxygen infusion into carbon structures, and thereby also to increasingly lower volatilities, ultimately resulting in extremely low volatility organic compounds (ELVOC).22,23 This process is postulated to be very important in the first steps of atmospheric new particle formation22,24,25 and potentially elucidates the first steps of the formation of highly oxidized organic material observed in atmospheric SOA3,4 and also in laboratory generated SOA.22,26,27 In addition, the proposed new scheme for atmospheric gas-to-particle conversion is partially based on recent understanding on low-volatility compounds, such as ELVOCs.28 In previous publications we illustrated the formation of highly oxidized ELVOC products from O3 and OH radical reactions with selected monoterpenes and unsaturated hydrocarbon precursors.22,23 For the cyclohexene + O3 system, a detailed formation mechanism could be constructed by performing high-level quantum chemical computations, allowed by the relatively small size and symmetric structure of the cyclohexene molecule.23 In the current paper, we use the knowledge gathered from these previous investigations to propose a mechanistic description of ELVOC formation from larger endocyclic alkenes 1-methylcyclohexene (1MCH) and 4methylcyclohexene (4MCH), surrogates for the more complex, abundant and atmospherically important biogenic monoterpenes such as α-pinene (structures in Figure 1), for which



COMPUTATIONAL DETAILS Quantum chemical calculations were performed to investigate the first steps of 1MCH and α-pinene oxidation, starting from the peroxy radicals generated in the ozonolysis reaction (which contain two carbonyl functionalities in addition to the peroxy radical group). Initial configurational sampling was carried out using the approach validated in Rissanen et al.23 Briefly, all possible conformers were first generated using the MMFF force field, followed by B3LYP/6-31+G(d) single-point energy evaluations and subsequent B3LYP/6-31+G(d) optimizations on structures within 5 kcal/mol of the lowest-energy conformer, all using the Spartan program.31 In the conformational sampling stage, transition states were treated using constrained minimizations as described in Rissanen et al. The lowest-energy structure for each reactant, transition state, or product was then selected for subsequent high-level treatment at the ωB97XD/ aug-cc-pVTZ level32 using the Gaussian 09 program suite,33 tight optimization criteria, and the ultrafine integration grid.

Figure 1. Structures of the alkene precursors investigated in this work: 1-methylcyclohexene (1MCH), 4-methylcyclohexene (4MCH), and αpinene.

ELVOC formation from ozonolysis was also investigated in the current work. Although the observed products in the α-pinene derived ELVOCs have clearly been formed by similar autoxidation reactions, they cannot be described purely by the same simple mechanistic principles as with 1MCH and 4MCH.

Table 1. Experimental Conditions of the Performed Measurements

a

reaction

bath gas

O2 percentage of the flow

volumetric flow, flow velocity, residence time/ dm3 min−1, m s−1, s

[O3] range/ 1011 cm−3

[alkene] range/ 1014 cm−3

added water

1MCH + O3 4MCH + O3 α-pinene + O3

N2/synthetic air N2/synthetic air N2/synthetic air

21 2−21 2−21

11.4, 0.11, 19 11.4−21.4, 0.11−0.2,a 4−19 5.3−30.0, 0.05−0.28,a 4−40

0.4−9 0.4−14.2 0.4−10.2

0.42−4.22 0.51−10.22 0.21−10.69

H2O/D2O H2O/D2O

4 s residence time obtained for the smaller flow tube: 63 cm long with a 4.0 cm i.d. B

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A For methylcyclohexene and for one of the H-shift reactions of α-pinene, a final single-point energy calculation was carried out at the ROHF-RCCSD(T)-F12a/VDZ-F12 level,34,35 using the Molpro 2012 program.36 The α-pinene RO2 system was too large to systematically treat at the CCSD(T)-F12 level, so only DFT-level energies could be calculated for the full set of transition states. Tunneling effects were computed with a onedimensional Eckart model.37 As shown in Rissanen et al.,23 the B3LYP/6-31+G(d) optimizations provide an excellent estimate of the energy ordering of the conformers, even though the absolute energies calculated at this level are less accurate. As the effects of including higher energy conformers (“local minima”) and/or torsional anharmonicity on the hydrogen shift rates at 298 K are both relatively small,23 the present approach (where only the lowest-energy conformer is selected for computationally demanding higher level calculations) should provide a good order-of-magnitude estimate of the hydrogen shift rates for 1MCH. For the α-pinene case, the lack of systematic CCSD(T)-F12 energy barriers prevents the reliable calculation of rates, but qualitative conclusions can still be drawn on the basis of the ωB97XD energetics and the single CCSD(T)-F12 barrier.

Figure 2. Nitrate CI-APi-TOF example mass spectra obtained from ozonolysis experiments of 1-methylcyclohexene, 4-methylcyclohexene, and α-pinene. The Y-axis shows the obtained ion counts relative to the reagent nitrate ion signals (i.e., reagent ions = NO3−, HNO3NO3−, and (HNO3)2NO3−).

other electronegative elements than oxygen are involved in the formation process. The D atom is one mass unit heavier than the H atom, and thus the peaks in the spectra were observed to shift by a mass unit with every exchanged hydrogen atom in the ELVOC product molecule (Figures 3 and 4). The relative molar yields of the highly oxidized products formed in the 1MCH and 4MCH ozonolysis systems (i.e., [ELVOC] produced in 1MCH and 4MCH experiments in similar alkene conversions) were determined by varying the organic precursor and ozone concentrations in the flow tube gas mixture and measuring the change in the product signal levels, with and without added reactants. Wall losses are inherent to flow tube studies, and therefore reagent concentrations were kept high enough to compensate for these losses and to obtain good enough signals for positive identifications. This resulted in a relative yield of Y4MCH+O3 = 3.6Y1MCH+O3 between these systems. The reference ozonolysis rate coefficients used for the yield determination were taken from a recent determination [1MCH (1.46 ± 0.10) × 10−16 cm3 molecule−1 s−1 and 4MCH (7.31 ± 0.36) × 10−17 cm3 molecule−1 s−1 41], and only the signals corresponding to peaks with the identified elemental compositions (Table 2) were used. At the highest reagent concentrations the yield curves started to bend significantly (Figure S4 in the Supporting Information), indicating uncertain second-order processes starting to influence the yield determination (e.g., potential condensation of generated low vapor pressure vapors), and thus only the low-concentration regime was used for the determination. No OH-scavengers were applied in any of the experiments.



RESULTS For the present purposes we have defined an ELVOC as a highly oxidized organic product species (at least six O atoms in the structure) capable of clustering with the nitrate ion. In other words, ELVOCs bind to NO3− stronger than HNO3 which in practice enables ELVOC detection with the nitrate ion based chemical ionization technique. Even though this rather arbitrary approach clearly takes into consideration species with varying volatilities, the definition is chosen for easier discussion of a group of gas-phase compounds with a common mechanistic origin and similar potential significance. The ELVOC dimers are defined as all the highly oxidized products detected with carbon numbers that are higher than the parent hydrocarbon; i.e., oxidized C18 species clustered with NO3− in α-pinene (C10H16) ozonolysis spectrum are considered ELVOC dimers. The dimers are named on the basis of their observed elemental compositions and are believed to be single covalently bound chemical compounds and should not be confused with van der Waals and hydrogen bonded dimers held together by electrostatic interactions. It is also worth emphasizing that these investigations were devised to study the formation mechanisms of the highly oxidized product molecules (i.e., ELVOC), and the potential of different unsaturated olefin structures to produce them. Thus, no rate parameters were determined during these experiments. However, the shortest residence time of about 4 s in the flow tube experiments constrains the whole autoxidation sequence to terminal closed-shell products in less than this time. CI-APiTOF example spectra of the three precursor compounds investigated are given in Figure 2 (example of the signal time behavior is given in the Supporting Information Figure S2). The elemental compositions of the peaks observed are gathered in Table 2. To gain insight into the product functionalities, we performed hydrogen to deuterium (H → D) exchange experiments by adding D2O to the bath gas flow. This supplied important information about the product structures, as acidic, labile hydrogens are the easiest to exchange,21,23,38−40 i.e., −OH and −OOH in the current oxidized product species when no



RESULTS OF QUANTUM CHEMICAL CALCULATIONS The zero-point corrected ROHF-RCCSD(T)-F12a/VDZF12//ωB97XD/aug-cc-pVTZ electronic energy barrier for the 1,6-hydrogen shift of the 1MCH derived peroxy radical (RO2(i); Figure S5 in the Supporting Information and Scheme 2 below) was 20.9 kcal/mol. The corresponding hydrogen shift rate at 298 K was computed to be 0.27 s−1, including an Eckart tunneling correction factor of 1030. As there are two similar H atoms at this abstraction site, the overall rate will be around 0.5 s−1. As the subsequent H-shifts in the autoxidation mechanism are likely to be faster than this (e.g., a 1,7-H-shift from an aldehyde group), the computed rate is consistent with ELVOC formation on a time scale of seconds. C

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A Table 2. Nitrate CI-APi-TOF (ELVOC) Peaks Observed in the Ozonolysis Experimentsa 286.0416 C7H12O8 0.06 286.0052 C6H8O9

1-Methylcyclohexene + O3 300.0209 334.1507 C7H10O9 C14H24O5

268.0310 C7H10O7

285.0338 C7H11O8

1.0b

0.06

380.1198 C14H22O8

382.0991 C13H20O9 (382.1355 C14H24O8)c

386.0940 C12H20O10

428.1046 C14H22O11

430.0838 C13H20O12 430.1202 (C14H24O11)

444.0995 C14H22O12 0.08

268.0310 C7H10O7 0.09

284.0259 C7H10O8 0.25

285.0338 C7H11O8

366.1406 C14H24O7

398.1304 C14H24O9

412.1097 C14H22O10

414.1253 C14H24O10 0.06

444.0995 C14H22O12

448.0944 C13H22O13

460.0944 C14H22O13

464.0893 C13H22O14

239.9997 C5H6O7 0.57

242.0154 C5H8O7 0.08

300.0572 C8H14O8

336.1300 C13H22O6

340.1249 C12H22O7

366.1406 C14H24O7 0.13

396.1147 C14H22O9

398.1305 C14H24O9

400.109 C13H22O10

412.1097 C14H22O10 0.12

414.0889 C13H20O11 0.10 (414.1253 C14H24O10)

316.0158 C7H10O10

317.0236 C7H11O10

318.0314 C7H12O10

416.1046 C13H22O11

428.1046 C14H22O11

430.1202 C14H24O11

432.0995 C13H22O12

492.0842 C14H22O15

508.0798 C14H22O16

524.0741 C14H22O17

257.0025 C5H7O8 0.21

476.0893 C14H22O14 0.07 α-Pinene + O3 266.0518 282.0467 C8H12O6 C8H12O7 0.09 0.35

284.0623 C8H14O7 0.12

298.0416 C8H12O8 0.07 298.0780 C9H16O7

299.0494 C8H13O8 0.25

330.0678 C9H16O9 0.09

308.0623 C10H14O7 0.63 340.0521 C10H14O9 0.40

310.0780 C10H16O7 0.29 341.0600 C10H15O9 0.23

324.0572 C10H14O8 0.15 342.0678 C10H16O9 0.41

325.0651 C10H15O8 1.0 356.0471 C10H14O10 0.20

326.0729 C10H16O8 0.28 357.0543 C10H15O10 0.57

328.0521 C9H14O9 0.10 358.0627 C10H16O10 0.35

329.0600 C9H15O9 0.06 372.0420 C10H14O11 0.17

373.0498 C10H15O11 0.11

374.0576 C10H16O11 0.14

388.0369 C10H14O12 0.17

389.0447 C10H15O12 0.09

448.1824 C19H30O8

462.1617 C19H28O9 0.05

464.1773 C19H30O9

466.1566 C18H28O10

478.1930 C20H32O9

480.1723 C19H30O10

482.1515 C18H28O11

492.1723 C20H30O10

494.1515 C19H28O11 0.08

498.1464 C18H28O12

508.1672 C20H30O11

510.1828 C20H32O11 0.06

524.1621 C20H30O12

526.1777 C20H32O12

540.1570 C20H30O13

542.1727 C20H32O13

556.1519 C20H30O14

562.1261 C18H28O16

572.1468 C20H30O15

574.1625 C20H32O15

4-Methylcyclohexene + O3 300.0209 301.0209 C7H10O9 C7H11O9 1.0 0.13

588.1418 C20H30O16

All peaks observed as clusters with NO3− in the spectra; no OH-scavengers were applied during the experiments. bThe intensity of the peak relative to the most intense peak in the spectrum; only peaks with over 5% relative intensity have been included. cPeaks in parentheses are minor peaks overlapping with the observed major peaks. a

For α-pinene, we were unable to calculate reliable absolute energy barriers for hydrogen abstractions as the crucial

CCSD(T)-F12 energy corrections could not be obtained for all transition states due to computational limitations. However, D

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Figure 4. (a) Peak shifts obtained for the 4-MCH ozonolysis derived monomer ELVOCs due to D2O addition to the gas stream. (b) Reagent nitrate ion clusters in the same experiment. Reagent peaks move similarly in the mass axis due to H → D exchange, i.e., HNO 3 *NO 3 − → DNO 3 *NO 3 − and (HNO 3 ) 2 *NO 3 − → HNO3DNO3*NO3−, (DNO3)2*NO3−. The inset illustrates the H/D exchange of the nitrate trimer reagent ion. A similar incomplete H → D shift is seen in reagent ions and in the product peaks (compare to Figure 3 for a complete H → D shift in reagent ions). Peaks with D atoms are shown in black.

Figure 3. α-Pinene monomer ELVOCs shift in mass when D2O is added to the bath gas flow. (a) Shown is a part of the ELVOC monomer spectrum from 320 to 370 Th, where the H → D shifts are most easily visible. The inset of the figure shows the same for all the observed C10 ELVOCs. (b) Reagent nitrate ion peaks with and without addition of D2O to the gas stream in the same experiment. The inset shows the almost complete conversion of the nitrate trimer peak: (HNO3)2*NO3− + D2O → (DNO3)2*NO3− + H2O.

frequencies of the transition states in the α-pinene system were also higher, between 2020 and 2100i cm−1 compared to 1300− 1940i cm−1 in the cyclohexene system, which will partially compensate for the higher barriers through increased tunneling correction factors. However, preliminary calculations indicate that the tunneling factors at 298 K are unlikely to exceed 3 × 104 (Supporting Information), which implies that the overall Habstraction rates will all be lower than those computed for the cyclohexene system, with the possible exception of the 1,4-Hshift from the aldehydic position in the secondary peroxy radical RO2(a) (denoted TSa3 in the Supporting Information). In terms of the overall reaction mechanism for α-pinene, Habstractions from the tertiary carbons on the cyclobutyl ring were calculated to be unfavorable despite the nascent carboncentered radical being stabilized by three carbon atoms, one of which is a keto group. This finding is likely due to ring strain in the product carbon-centered radical, as suggested by Vereecken et al.42 Ring strain is also a possible reason for the high barrier calculated at both DFT and CCSD(T)-F12 levels for the aldehydic H-abstraction by the primary peroxy radical RO2(b) (denoted TSb2 in the Supporting Information).

the relative ωB97XD/aug-cc-pVTZ barriers are likely reliable to within a few kcal/mol. We investigated the hydrogen abstractions from the most likely sites for the two peroxy radicals known to be formed in α-pinene ozonolysis. These were the aldehydic H, the tertiary H atom α to a keto group, and the secondary CH2 on the cyclobutyl ring for both radicals, and additionally the C(O)−CH3 hydrogens for the second peroxy radical RO2(b) (see Figure S6 and S7 in the Supporting Information for the structures of the peroxy radicals and the transition states). Other hydrogen atoms, such as the terminal methyl hydrogen atoms and the tertiary hydrogen with no αcarbonyls, should be more difficult to abstract than these, on the basis of previous studies of H-abstraction reactions on peroxy radicals (see for example refs 7−9, 12, and 23, and references therein). The barriers for all investigated H-abstractions by α-pinene peroxy radicals were quite high. In the cyclohexene system,23 zero-point corrected energy barriers at the ωB97XD/aug-ccpVTZ level varied between 14 and 20 kcal/mol. In contrast, the barriers computed for the α-pinene system at the same level ranged from 22 to 29 kcal/mol. The zero-point corrected CCSD(T)-F12 barrier calculated for the aldehydic H-shift across the cyclobutyl ring is 28.0 kcal/mol, which is within 0.3 kcal/mol of the DFT value (Table S2 in the Supporting Information), and thus strongly supports the results that Hshift barriers in the α-pinene system are high. The imaginary



DISCUSSION Autoxidation. Due to the close similarity between cyclohexene23 and methylcyclohexene ELVOC nitrate CI-APiTOF spectra obtained, and the similarity of their molecular structures, the oxidation processes are presumed to advance by E

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 1. 1-MCH Ozonolysis Leading to a Peroxy Radical with Seven Carbon Atoms and Four Oxygen Atoms through a Primary Ozonide (POZ) Decomposition, Criegee Intermediate (CI) Isomerization, and Vinyl Hydroperoxide (VHP) Splitting an OH Radical, Thereby Forming a Carbon-Centered Radical That Adds O2 To Form a Peroxy Radical [RO2(i)], Which Is the Starting Point for ELVOC Formation (for Example, Ref 46)

Autoxidation of Methylcyclohexenes. The guidelines presented above were then used to describe the autoxidation sequences shown in Schemes 1−4, forming the main ELVOC monomer products observed in the 1MCH ozonolysis mass spectra (Figure 2 and Table 2). These determinations were further supplemented by additional experiments with D2O or H2O addition to the gas stream (see example in Figure 3) and experiments where the concentrations of O3 and alkenes were varied consecutively. Scheme 1 shows the well-known ozonolysis chemistry in the presence of oxygen,46,47 which proceeds by O3 addition to the alkene double bond forming a primary ozonide (POZ), subsequent splitting of the carbon− carbon bond, and production of one of the four possible Criegee intermediate (CI) conformations (see Supporting Information Figure S3 for other possible structures). If a synconformer is formed (“syn” in this case simply meaning that the radical-center oxygen is oriented toward an alkyl group; Figure S3, Supporting Information), it can isomerize to a vinyl hydroperoxide (VHP)-type structure that is generally susceptible to dissociation to an OH and an oxy-radical, and to further isomerization into a carbon-centered radical that will add an O2 molecule, finally producing a four oxygen peroxy radical, which is the starting structure for ELVOC formation. Only the synconformer of the CI is expected to produce a VHP-type structure, whereas the anti-conformer leads to a dioxirane and further to a “hot acid”48,49 and is not expected to progress the radical chain reaction leading to an ELVOC. Using the above definition, three of the four CI conformers of 1MCH are “syn” and only one is “anti”. Schemes 2−4 show the continuation of 1MCH ozonolysis Scheme 1, for all three possible peroxy radicals formed from syn-conformers of the Criegee intermediates (for CI1−CI3 see Figure S3 (Supporting Information), and for the peroxy radicals in Schemes 3 and 4, see Scheme S2 (Supporting Information)), and illustrate the potential formation pathways for the monomer ELVOC products observed in the nitrate CI-APiTOF spectra (Figure 2). The schemes are drawn according to the general guidelines presented above. The first step of Scheme 2 was further investigated by quantum chemical calculations to determine whether it is fast enough to initiate autoxidation on the experimental time scale of seconds. (The initial steps of Schemes 3 and 4 involve aldehydic H-shifts,

similar autoxidative reaction pathways. The main difference between these ozonolysis systems is the formation of two types of Criegee intermediates (both with two different conformers) in methylcyclohexene ozonolysis, in comparison with only one in cyclohexene, which creates more potential autoxidative pathways. Based on our recent work on the cyclohexene ozonolysis system,23 previous literature concerning peroxy radical H-shift reactions,6−9,11 and hydrogen abstraction reactions in general,5 common guidelines for potential autoxidation sequences in methylcyclohexenes were generated and are briefly described next. First, abstraction from an aldehydic carbon is preferred, leading to an acylic radical and subsequent potential CO loss, which is generally not competitive with the very rapid O2 addition reaction.9,43 Next, irreversible abstraction from a carbon with hydroperoxide functionality leads to a prompt OH ejection and thus to a termination of the autoxidation. If neither of these two sites is available, then the third alternative Habstraction is either from a secondary carbon with adjacent oxygenated substituents, or from a tertiary carbon atom except when ring strain makes abstraction from a tertiary site unfavorable.42,44 As usual for saturated alkanes, abstraction from a tertiary position is assumed most favorable, followed by a secondary and, finally, a primary position. 1,5- to 1,7-H-shifts are considered the most likely, as according to previous studies their transition state structures are closest to optimal for the Hshift to occur due to the absence of ring strain. These are closely followed by 1,8- and 1,9-H-shifts (which are less likely due to being entropically disfavored), whereas 1,4-H-shifts are considered significantly less likely (this is according to common consensus, even though some fast 1,4-H-shifts from an aldehydic position have been illustrated9,45), and 1,3-H-shifts are assumed to be negligible. Oxygenated functional groups adjacent to the H-abstraction site are thought to facilitate Habstraction reactions and are crucial for the autoxidation process. HO2 loss through 1,4-H-shifts, concurrently creating a new double bond, are tentatively predicted on the basis of the detected product compositions (see Table 2 and discussion below). Though generally slow, these 1,4-H-shifts could become competitive by the weakening of C−H bonds due to adjacent oxygen-containing functional groups.9 F

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 2. First Part of the Proposed Autoxidative ELVOC Formation Pathways in the 1MCH Ozonolysis System, Starting from the Peroxy Radical [(RO2(i)] Formed in Scheme 1a

a

Potential formation routes for the isomeric products corresponding to the most intense peak in the 1MCH ozonolysis spectrum observed at 268.031 Th (C7H10O7*NO3−) are shown. The first H-shift step is calculated to proceed with an overall rate of around 0.5 s−1; subsequent steps are likely faster due to more favorable H-shift sites. The first H-shift reaction could happen to the aldehydic hydrogen as well (i.e., 1,4-H-shift), but the main oxidation route would still lead to product P2. All H-shifts, except those leading to prompt OH loss, are considered reversible, and the potential products detected in the ELVOC spectra have been marked with bold font.

for an alkyl + O2 association of 5 × 10−12 cm3 molecule−1 s−1 is used50). As was the case in our previous study on cyclohexene ELVOC formation,23 unimolecular termination reactions are proposed to form a large fraction of the closed-shell products. The role of bimolecular termination reactions, which will certainly be involved, are omitted from the schemes below and are discussed in more detail in a later section. The relative importance of the different reaction pathways presented in Schemes 2−4 is difficult to assess without detailed

which are known to be fast on the basis of previous studies of similar systems.) The bimolecular autoxidation steps in all of the current schemes, i.e., oxygen addition after formation of a carboncentered radical, are pseudounimolecular in a sense that O2 addition occurs practically instantly in ambient air, where about every fifth collision happens with an oxygen molecule, which leads to an effective first-order rate coefficient of 2.5 × 107 s−1 for the O2 addition reaction (when a common rate coefficient G

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 3. Second Part of the Proposed Autoxidative ELVOC Formation Pathways in the 1MCH Ozonolysis System, Starting from the Second Peroxy Radical [RO2(ii)] Shown in Scheme S2 (Supporting Information)a

a Formation of the heaviest monomer peak in the spectrum observed at 300.021 (C7H10O9*NO3−) is sketched. Note that also this route potentially leads to two distinct isomers.

energetics of the reaction potential energy surfaces. However, the end-product structures generated by the guidelines described above are all very similar due to common mechanistic steps forming them. Thus, due to their close structural similarity, by a first approximation, also their chemistry is expected to be similar too; a common property often exploited in structure−activity relationships (SAR).51−53 The same procedures were used to derive the autoxidative formation pathways in the 4MCH ozonolysis system, and the pathways proposed are shown below in Schemes 5 and 6. In 4MCH ozonolysis, only two potential peroxy radicals are formed through syn-CI-conformers, in contrast to the three possibilities in the 1MCH system (Schemes S2 and S3, Supporting Information).

There is a striking difference between the obtained 1- and 4methylcyclohexene ozonolysis mass spectra: identical monomer peaks, but with a reversed peak intensity distribution (Figure 2). Ketene loss (C2H2O) or CO loss can be ruled out, as the highest intensity ELVOC peaks observed in both MCH spectra are C7 species and thus have not lost any fragments containing carbon atoms. Thus, it seems that the differences in the experimental product distributions observed between the 1MCH and 4MCH systems result from different unimolecular termination channels. According to the mass spectra shown in Figure 2, the OH loss from the many O8 peroxy intermediates in the 1MCH system seems to be significantly more likely than from the analogous RO2 intermediates in the 4MCH case, and thus terminate to closed-shell ELVOC end products in fewer reaction steps, resulting in the “reversed” monomer ELVOC H

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 4. Third Autoxidative Reaction Pathway in 1MCH Ozonolysis, Starting from the Final Peroxy Radical Structure, RO2(iii), Shown in Scheme S2 (Supporting Information), Again Potentially Leading to Two Distinct Isomersa

a

Note that the initiating 1,9-H-shift is generally assumed less likely due to unfavorable entropic contribution in the transition state for the reaction, and thus this scheme is probably the least likely of Schemes 2−4.

RO2(y) after the first autoxidative O2 addition step (Scheme 6; for RO2(y) production see Scheme S3 in the Supporting Information), has three different, potentially close to equal importance H-shift reactions open: 1,5-H-shift to form a tertiary radical, 1,7-H-shift leading to OH loss and 1,8-H-shift to form a acyl-type radical, which can also dissociate, losing CO. The branching between these different channels may be dictated by subtle energetics in the system and thus the competition between channels may change with experimental conditions. These need to be studied in more detail in the future. Some of the RO2 radicals also potentially terminate the reaction chain by losing an HO2 radical,11−13,54−57 concurrently forming a double bond as shown in Schemes 5 and 6 for 4MCH. The formed double bond could then be attacked by OH or O3, creating additional carbon-centered radicals to

intensity pattern observed (Figure 2). In addition, also the CO loss from the 1MCH derived peroxy radicals seems significantly more likely in comparison with the peroxy radicals in the 4MCH system. In 1MCH, the first dimer products with less than C14 (i.e., C13 meaning that peroxy radicals with a composition of C7 and C6 have reacted, see next section below) are found with 6, 9, and 10 O atoms, whereas in the 4MCH the first product with less than C14 is found with 12 O atoms. Thus, it seems that the CO loss from the intermediate acyl-type radicals is significantly more competitive with the O2 additions in the 1MCH than in the 4MCH ozonolysis system (assuming similar detection sensitivity for all ELVOCs in both systems). Potentially important branching points in the determined autoxidation mechanism exist. For example, in the 4MCH ozonolysis system the C7H11O6 peroxy radical, formed through I

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 5. Proposed Autoxidation Sequence Producing ELVOCs in 4MCH Ozonolysis, Starting from the Peroxy Radical RO2(x) in Scheme S3 (Supporting Information)a

a

The main oxidation pathway is shown in the middle, from top to bottom. The proposed most likely termination reactions and branching points are shown on both sides of the main pathway. Potential products detected in the mass spectra have been marked with bold font.

products with this type of composition were detected in the current work (Table 2). Another imaginable route to products with the above-mentioned composition could be through an alkoxy radical (RO) isomerization reaction to a hydroxyalkyl species (R−OH) and a subsequent unimolecular termination reaction, and the potential involvement of alkoxy radicals in

continue the oxidation sequences. The HO2 loss product from ozonolysis-derived RO2 should be observed with a composition of CxHyOz, where x is the original amount of C atoms (minus the amount of occurred CO losses), y is the original amount of hydrogen atoms minus two (i.e., one H lost due to OH ejection from VHP and the second by HO2 ejection), and z is an even number (provided no CO was lost during the process), and few J

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 6. Second Part of the Proposed 4MCH Ozonolysis Autoxidation Reactions Shown, Starting from the RO2(y) in Scheme S3 (Supporting Information)

K

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

the autoxidation sequence advances, the peroxy functionalities, and thus also the subsequent alkoxy functionalities, become located in positions where there are no abstractable hydrogens available and render further oxidation through this pathway “impossible”. Moreover, reaction 5 would generate additional carbonyl functionality instead of hydroperoxy functionality, and thus could prevent the product detection by NO3− clustering in the CI-APi-TOF. Furthermore, if active, this route would be an HO2 source to the system leading to a competition between reactions 2 and 4, which might limit the importance of RO formation. Both reactions have very uncertain efficacies for the types of compounds we are studying and thus cannot be quantitatively assessed. Finally, the rate coefficients of the RO2 + RO2 reactions are known to be very structure specific, ranging from about 10−11 to 10−17 cm3 s−1, with electronegative substituents generally facilitating faster reactions and primary radicals showing higher reactivity than tertiary radicals.62−64 Additional unimolecular reaction channels competitive with the autoxidation sequence (e.g., CO loss from acyl-type radicals9,23,43) are most clearly seen in dimer product distributions. The dimers observed generally contain twice as many carbon atoms as the monomer species (i.e., in MCH ozonolysis C7 species will mainly form C14 dimers), but lower carbon numbers due to the fragmentation processes of the intermediate radicals, such as CO loss and potential ketene loss,9,65,66 mentioned above (e.g., CO loss derived C6 peroxy radical reacting with C7 will give C13 oxygenated compounds in methylcyclohexene systems). The dimers also generally have lower O:C ratios and more than double the amount of hydrogens than the monomer species (Table 2). Also, in D2O addition experiments, the dimers seem to exchange fewer hydrogens to deuteriums (relative to the oxygen number), which we interpret as being due to creation of an oxygen bridge from two peroxy radicals originally capable of transferring hydrogen to a labile, exchangeable position. These observations are consistent with our tentative identification of dimers as oxygen-bridged peroxide-type compounds formed by peroxy radical “head-to-head” recombination reactions. Another possibility for forming dimers could be through Criegee intermediate reactions with the autoxidized product species leading to secondary ozonides. However, the obtained elemental compositions (Table 2) would require much more oxidized reaction partners than Criegee intermediates to form the observed product distributions, and also, this type of reaction would likely favor the less oxidized species. Furthermore, the very short lifetimes of Criegee intermediates makes them much more improbable reaction partners than the rather long-lived peroxy radicals, under our low [NO] reaction conditions. OH Chemistry. It is notable that if OH chemistry was involved to a significant extent, the product with the same amount of H atoms as the parent alkene (i.e., y + 2) could be produced through this pathway too (e.g., C7H12O8 observed in 1MCH system). These types of products are not seen in the 4MCH spectrum but are observed in the 1MCH spectrum (Table 2 and Figure 2) and thus point to different reaction pathways being open. No OH-scavengers were applied during these experiments, and thus the products observed could be affected by OH-chemistry. The OH-yields of the 1MCH67 and α-pinene40 ozonolysis systems have been determined previously as 0.91 ± 0.20 and 0.86 ± 0.13, respectively, and the observed high, almost unity OH-yields strongly suggest that OH reactions should have a major impact on the oxidation

these reaction sequences is discussed in more detail in the next chapter. As with our previous work,23 the H → D exchange experiments with D2O addition provided valuable additional information on the product structures. Generally, the products containing less oxygen were also found to contain fewer exchangeable hydrogens (Figure 3 and 4) and support the autoxidation-type product formation where acidic hydrogens are created during molecular growth by sequential peroxy radical isomerization + O2 addition steps (R−O−O• → •R− O−O−H). In principle, all of the peroxy intermediates in these reaction sequences have a finite probability for terminating the reaction chain by a specific isomerization reaction followed by a unimolecular dissociation reaction (i.e., losing OH or HO2, Schemes 1−6), or in bimolecular reactions with other peroxy radicals (i.e., HO2 and RO2, see next section), and NO (negligible in the current reaction system).6,7,52,57 Thus, a product distribution with increasing amount of O atoms and exchangeable hydrogens should be obtained, as shown in Figures 2−4. Bimolecular Reactions and Dimers. It is important to emphasize that bimolecular reactions of the intermediate RO2 radicals in Schemes 1−6 can lead to products with the same elemental compositions, and also with very similar chemical structures (i.e., with the same number and type of functional groups, but positioned differently in the molecule), as the purely unimolecular framework presented above. Such multifunctional isomers will be very hard to distinguish from each other within a gas-phase reaction mixture using the existing detection methods and therefore pose serious challenges for experimental determinations of their relative importance. The most likely bimolecular reaction channels relevant for the current discussion are6,53,57,58 RO2 + RO2 → ROOR + O2

(1)

RO2 + HO2 → ROOH + O2

(2)

RO2 + RO2 → RCHO + ROH + O2

(3)

RO2 + RO2 → RO + RO + O2

(4)

Reaction 1 forms oxygen-bridged peroxides and is expected to be mainly responsible for the dimers observed. Reaction 2 leads to hydroperoxides, and a few of the detected products with even numbers of O atoms and with the same amount of H atoms as the parent molecule can be explained by reaction 2 together with the H-shift mechanism presented above. Reaction 3 could generate two ELVOC molecules providing that the parent RO2 radicals were oxidized enough before this termination reaction occurred. Reaction 4 is known to be a prominent route in peroxy radical reactions, but its effect is perhaps the most difficult to predict, as alkoxy radicals (RO) are common chain branching agents in organic oxidation mechanisms. RO can promptly decompose, undergo intramolecular H-shift reaction followed by O2 addition, or react directly with O2 in a hydrogen abstraction reaction generating an HO2 and a carbonyl compound: RO + O2 → RCO + HO2

(5)

The branching ratios for different RO reactions are heavily dependent on the actual alkoxy radical structure.52,53,59−61 This issue can cause ambiguity in the observed product structures, as similar compositions are expected for the products formed by reactions 4 and 5, as with unimolecular chemistry. However, as L

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Scheme 7. α-Pinene Ozonolysis Leads to Two Different Criegee Intermediates, Each with Two Different Conformersa

a

Three of them, shown here, are syn-isomers as described above and in the Supporting Information, and are thus potentially capable of isomerizing to a VHP, and subsequently to peroxy radicals. According to previous investigations, both of the Criegee intermediates make a significant contribution to reaction products observed (i.e., it does not matter on which end of the molecule the radical center forms), and for example in Leeds Master Chemical Mechanism (MCM), they are assumed to be equally important.70

α-Pinene. Similar accuracy as in the 1MCH and 4MCH ozonolysis autoxidation pathways cannot be obtained for αpinene due to its significantly more complicated nature, i.e., more structured C10H16 compound vs simpler C7H12 species (Figure 1). As may be expected, the more complicated structure of α-pinene also yields a correspondingly more complicated ELVOC spectrum (Figure 2 and Table 2). As a first guess, the rigid four-carbon-atom ring and the relatively isolated methyl groups connected to it in the α-pinene structure should make the autoxidation pathways more accessible to deduce than in many of the other abundant monoterpenes. However, after a closer look it is found that to produce some of the highly oxidized end products and some of the observed fragment species (i.e., C5 compounds in Table 2), the four-carbon ring is most likely broken during the oxidation sequence, which has been reported also, e.g., in OH-initiated α-pinene oxidation.44,68,69 Thus, to understand how the autoxidation could be advancing, we scrutinized the first steps of the ozonolysis reaction. The different Criegee intermediates and their subsequent reactions to “first-order” carbon-centered alkyltype radicals are shown in Scheme 7. The question now becomes, which of the three CIs in Scheme 7, if not all, lead to the formation of ELVOC. CI3 has

process (Table 2). However, only part of the OH addition to double bond leads to C−C bond rupture and thus to the opening of the ring, which seems to be a prerequisite for an efficient gas-phase autoxidation process. Furthermore, the carbonyl group added to the carbon structure in the ozonolysis oxidation was already previously postulated as important for the initiation of the autoxidation process.22 OH-addition to a double bond would create a carbon-centered radical with one more H atom than the parent species [e.g., 1-methylcyclohexene (C7H12) would be converted to a 1-methyl-2-hydroxycyclohexyl radical (C7H12OH)], in contrast to the ozonolysis oxidation where one H atom is lost by OH ejection from VHP decomposition (Scheme 1), creating a carbon-centered radical with two oxygen atoms and one H atom less than in the parent alkene (i.e., C7H11O2) to start the formation of an ELVOC by autoxidation. Thus, when the peroxy radicals from the OH- and O3-oxidation sequences mix, we obtain dimer products with twice the amount of H atoms as in the parent reagent, i.e., with 1MCH the expected composition is C14H24Ox (Table 2). If only OH-chemistry would be involved, it would be possible to obtain dimers with more than twice the amount of hydrogens, but as no such dimers are observed, the ozonolysis chemistry dominates under the conditions of the present experiments. M

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

The H → D exchanges observed in the α-pinene system seem to indicate differences in the autoxidation mechanism, as similar H → D shift occurs on products with higher oxygen content than in the corresponding 4MCH oxidation system (and in the previous cyclohexene system).23 Specifically, the cyclohexene and 4MCH ozonolysis leads to products with 2 and 3 exchanged H atoms in products with 7 and 9 O atoms, respectively, whereas in the α-pinene similar H-shifts are observed in products with 9 and 11 O atoms attached. This indicates that the autoxidation sequence for α-pinene is influenced by a reaction that was not considered in our ELVOC generation routine and which does not generate exchangeable hydrogen atoms during the oxidation process. A potential candidate could be a ring-closure reaction of peroxy radicals previously suggested in OH-initiated oxidation of monoterpenes,6,73,77 and postulated as a common reaction step creating bicyclic peroxy radicals in aromatic oxidation reactions.6,78,79 For example, Eddingsaas et al.73 have reported a 3% yield for a route involving a peroxy ring closure. This would tie one of the peroxy groups in the ring without giving the opportunity to create exchangeable hydrogen to the structure but would still be able to advance the autoxidation sequence. However, this type of reaction would need an intact double bond in the structure for the peroxy radical to bite into, concurrently creating another radical site to continue the oxidation process. Highly oxidized product species containing the same amount of H atoms as the parent alkene were also observed in the αpinene system (e.g., C10H16O9 and C10H16O11, Table 2). These products can most likely be explained by reactions of the type RO2 + HO2 → ROOH + O2 (2) [or possibly by reaction 3], which results in an additional exchangeable hydrogen atom in the product molecule in comparison with products formed through other proposed termination pathways (e.g., C10H14O9 and C10H14O11, Schemes 1−6). OH addition to the double bond and the similar subsequent autoxidation sequence could also lead to a product with one more exchangeable hydrogen (i.e., −O−H) and with H-amount equal to the parent alkene, but as no dimers with more than twice the amount of hydrogens as in the parent alkene are observed, this pathway seems less likely. Nevertheless, in the spectra these products are observed to exchange one H → D more than the other product species with the same amount of O atoms, consistent with this description. Previously postulated ketene (C2H2O) loss for αpinene at low pressures65,66 seems to be operating at higher pressures as well, as we see products with C8H12−14O7−8 compositions (Table 2) that could have been formed through this pathway, i.e., after C2H2O loss the α-pinene derived peroxy radical would have a composition of C8H13O3, which could then autoxidize similarly. Another route to produce species with such elemental compositions could be through CO loss followed by subsequent CH2O ejection, the aldehyde loss being a common decomposition pathway for alkoxy radicals.44,53,59 ELVOC Molar Yields. The nitrate CI-APi-TOF used in this work to detect the ELVOCs is blind to the majority of the products formed in these ozonolysis reactions. Fortunately, many of these products, including secondary ozonides, carboxylic acids, multifunctional aldehydes, and ketones and various moderately oxidized organic compounds, have been determined in previous investigations with more suitable detection methods and can thus be collected from the extensive literature covering the topic of α-pinene ozonolysis (see, for

been postulated to lead to a ring opening reaction with a yield of 14%, albeit via intermediate alkoxy radical formation.68,71 As the H-shifts from the cyclobutyl ring, and also across the ring, are predicted to be slow (Supporting Information), CI3 seems the most likely starting point for ELVOC formation by autoxidation. Generally, the abstraction from a tertiary carbon atom is favored over primary and secondary positions, as discussed above, but in α-pinene this is assumed to be a minor route.42,68,72 Moreover, even the 1,4 H-shift from an aldehydic position, previously indicated as surprisingly facile in related Hshift systems,9,45 is estimated to be too slow in the α-pinene case. The CI1 and CI2 derived peroxy radicals suffer from the same lack of preferable abstraction sites, due to ring abstraction being unlikely and the ring also preventing the bending of the structure and abstraction from the opposite side of the ring. As the H-shifts of the peroxy radicals commonly considered in literature to be formed in α-pinene ozonolysis do not seem rapid enough for efficient autoxidation on the experimental time scale, other mechanistic steps seem to be required for the autoxidation to proceed into such high oxygen contents as observed in the α-pinene ELVOC products. Currently, the best explanation seems to involve a ring-opening reaction, which has been proposed previously, for example, in the case of pinonaldehyde photo-oxidation, also through an alkoxy intermediate isomerization reaction.73 According to previous theoretical work, the most potential ring opening reactions should indeed occur through alkoxy radical rearrangement reactions, especially if the alkoxy is directly connected to the ring.44,53,68,74 Alkoxy radicals should form in the current reaction mixtures, at least via reaction 4. However, due to their rich and heavily structure related chemistry53,59,60 alkoxy radical reactions pose the biggest uncertainties in the mechanisms proposed, and to understand the details of their involvement in the autoxidation sequences is beyond the scope of the current study. Even though the mechanism for α-pinene is evidently harder to construct, it is clear that a similar type of autoxidative mechanism is operational, which is best demonstrated by the progression of H-shift products seen in experiments with D2O addition to the bath gas flow (Figure 3). The progression is seen from the products detected with increasing amount of exchangeable hydrogens. For example, the product observed at 325.0651 Th as a cluster with NO3− [m/z(NO3−) = 61.9884 Th] and identified as C10H15O8 peroxy radical, shifts by one mass unit (i.e., C10H15O8 → C10H14DO8). The products at 340.0521 Th and 342.0678 Th, closed-shell oxidation products C10H14O9 and C10H16O9, shift by 2 and 3 (i.e., C10H14O9 → C10H12D2O9 and C10H16O9 → C10H13D3O9), respectively, and the 357.0543 peroxy radical (C10H15O10) shifts by two mass units (i.e., C10H15O10 → C10H13D2O10). Note especially how the higher mass peroxy radical (C10H15O10) exchanges the same amount of hydrogen atoms to deuteriums as the lower mass closed-shell product (C10H14O9), in accordance with typical peroxy radical termination reactions. That is, the higher mass peroxy radical cannot make more H → D shifts than the lower mass closed-shell product it forms by unimolecular dissociation (see Schemes 1−6 for examples). The bimolecular termination reactions, however, could lead to these types of products through reaction 2 or 3. The observed H → D shift is thus another strong indication that the odd mass species observed in the mass spectra (Table 2) are really peroxy radicals, which are generally difficult to detect experimentally with mass spectrometry or other methods.6,55,75,76 N

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A example, refs 27, 73, and 80, and references therein), and 1methylcyclohexene ozonolysis,41,65,67,81−85 whereas for 4methylcyclohexene ozonolysis, apparently no gas-phase products have been determined previously. During the current experiments the calibration of the instrument could not be performed and thus the measured ion signals could not be converted into ELVOC product concentrations. However, the relative ELVOC product yields (YELVOC) of 1MCH and 4MCH ozonolysis systems at similar alkene conversions can still be reported by assuming equal sensitivities for all ELVOC in both reaction systems. By these assumptions the analysis resulted in YELVOC(4MCH) = 3.6YELVOC(1MCH).

intermediate conformations formed in 1-MCH ozonolysis, Figure S4 ELVOC yield plot for 1MCH + O3, Figure S5 species in 1MCH calculations, Figure S6 and S7 species in α-pinene calculation, Scheme S1 formation of dioxirane from an anti-CI, Scheme S2 first steps of 1MCH ozonolysis, Scheme S3 first steps of 4MCH ozonolysis, Table S1 energetics for the calculations, Table S2 calculated barriers and tunneling corrections. This material is available free of charge via the Internet at http://pubs.acs.org.

CONCLUSIONS The formation of extremely low-volatility organic compounds, ELVOC, from three selected alkene ozonolysis systems, 1methylcyclohexene, 4-methylcycohexene, and α-pinene, was studied in glass flow tubes using NO3− based chemical ionization mass spectrometry. The different structures of these alkenes were observed to lead to significantly different ELVOC spectra and thus supplied important structurally related information on the competition between autoxidation sequences (=increasing the level of oxidation) and different termination reactions. Common guidelines for ELVOC formation from cyclohexenes were generated and applied for methylcyclohexene autoxidation, whereas for α-pinene it was found that additional mechanistic steps are required to explain the product formation. Nevertheless, the α-pinene derived ELVOCs were observed to form through similar autoxidation sequences, indicated by experiments with D2O addition to the gas-flow. The first steps of the 1MCH and α-pinene oxidation were also investigated by quantum chemical calculations. The 1MCH ozonolysis-initiated oxidation was shown to proceed similarly as previously observed for cyclohexene system, and thus could be described by common rules generated around autoxidation. However, the H-abstraction barriers for the α-pinene-derived peroxy radicals were found to be quite high, necessitating different pathways to produce the ELVOCs observed. This was supported by surprisingly low amounts of labile hydrogens in the ELVOCs produced from α-pinene ozonolysis. These results illustrate the importance of studying different structural characteristics on ELVOC formation processes. The determination of a complex organic structure from a mass spectrum can rarely be unambiguous, and this ambiguity increases together with the amount of atoms in the molecule. In a previous study,23 we learned that the sites for internal H-shift reactions become more difficult to determine as the oxidation content of the molecule increases, because the H-abstraction sites become closer to equal in importance. The ambiguity especially increases with lengthening of the carbon chain as the amount of possible isomers quickly rises. This also implies that there is not one dominant autoxidation pathway, but instead, multiple isomers are likely to form, as was especially indicated for the α-pinene ozonolysis system investigated in the current work.

Notes



*M. P. Rissanen. E-mail address: matti.p.rissanen@helsinki.fi.





AUTHOR INFORMATION

Corresponding Author

The authors declare no competing financial interest.



ACKNOWLEDGMENTS This study was supported by Academy of Finland Centre of Excellence programs (project no. 272041 and 1118615), Academy Research Fellow grant (no. 266388), and PostDoctoral Researcher grant (no. 251427). H.G.K. thanks the support from The Danish Council for Independent Research Natural Sciences and the Danish Center for Scientific Computing. J.A.T. acknowledges support from the U.S. Department of Energy through grants DE-SC0011791 and DE-SC0006867. We are grateful to Prof. Solvejg Jorgensen and Miss Noora Hyttinen for their help in quantum chemical calculations, and we thank the TofTools team for providing us the machinery for analyzing the CI-APi-TOF spectra. The CSC IT center for Science in Espoo, Finland, is acknowledged for computer time.



REFERENCES

(1) Kroll, J. H.; Seinfeld, J. H. Chemistry of Secondary Organic Aerosol: Formation and Evolution of Low-volatility Organics in the Atmosphere. Atmos. Environ. 2008, 42, 3593−3624. (2) Ziemann, P. J.; Atkinson, R. Kinetics, Products, and Mechanisms of Secondary Organic Aerosol Formation. Chem. Soc. Rev. 2012, 41, 6582−6605. (3) Ng, N. L.; Canagaratna, M. R.; Zhang, Q.; Jimenez, J. L.; Tian, J.; Ulbrich, I. M.; Kroll, J. H.; Docherty, K. S.; Chhabra, P. S.; Bahreini, R.; Murphy, S. M.; Seinfeld, J. H.; Hildebrandt, L.; Donahue, N. M.; DeCarlo, P. F.; Lanz, V. A.; Prévôt, A. S. H.; Dinar, E.; Rudich, Y.; Worsnop, D. R. Organic Aerosol Components Observed in Northern Hemispheric Datasets from Aerosol Mass Spectrometry. Atmos. Chem. Phys. 2010, 10, 4625−4641. (4) Jimenez, J. L.; Canagaratna, M. R.; Donahue, N. M.; Prevot, A. S.; Zhang, Q.; Kroll, J. H.; DeCarlo, P. F.; Allan, J. D.; Coe, H.; Ng, N. L.; Aiken, A. C.; Docherty, K. S.; Ulbrich, I. M.; Grieshop, A. P.; Robinson, A. L.; Duplissy, J.; Smith, J. D.; Wilson, K. R.; Lanz, V. A.; Hueglin, C.; Sun, Y. L.; Tian, J.; Laaksonen, A.; Raatikainen, T.; Rautiainen, J.; Vaattovaara, P.; Ehn, M.; Kulmala, M.; Tomlinson, J. M.; Collins, D. R.; Cubison, M. J.; Dunlea, E. J.; Huffman, J. A.; Onasch, T. B.; Alfarra, M. R.; Williams, P. I.; Bower, K.; Kondo, Y.; Schneider, J.; Drewnick, F.; Borrmann, S.; Weimer, S.; Demerjian, K.; Salcedo, D.; Cottrell, L.; Griffin, R.; Takami, A.; Miyoshi, T.; Hatakeyama, S.; Shimono, A.; Sun, J. Y.; Zhang, Y. M.; Dzepina, K.; Kimmel, J. R.; Sueper, D.; Jayne, J. T.; Herndon, S. C.; Trimborn, A. M.; Williams, L. R.; Wood, E. C.; Middlebrook, A. M.; Kolb, C. E.; Baltensperger, U.; Worsnop, D. R. Evolution of Organic Aerosols in the Atmosphere. Science 2009, 326, 1525−1529. (5) Streitwieser, A.; Heathcock, C. H.; Kosower, E. M. Introduction to Organic Chemistry, 4th ed.; Prentice-Hall, Inc.: Hoboken, NJ, 1992. (6) Orlando, J. J.; Tyndall, G. S. Laboratory Studies of Organic Peroxy Radical Chemistry: an Overview with Emphasis on Recent

ASSOCIATED CONTENT

S Supporting Information *

Figure S1 schematic of the experimental setup, Figure S2 measured ELVOC ion signals as a function of time, Figure S3 the syn-CI and anti-CI notation and the different Criegee O

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A

Weight Gas Phase Products During Ozonolysis of α-pinene. Atmos. Chem.Phys. 2013, 13, 7631−7644. (28) Kulmala, M.; Petäjä, T.; Ehn, M.; Thornton, J.; Sipilä, M.; Worsnop, D. R.; Kerminen, V. M. Chemistry of Atmospheric Nucleation: On the Recent Advances on Precursor Characterization and Atmospheric Cluster Composition in Connection with Atmospheric New Particle Formation. Annu. Rev. Phys. Chem. 2014, 65, 21− 37. (29) Junninen, H.; Ehn, M.; Petäjä, T.; Luosujärvi, L.; Kotiaho, T.; Kostiainen, R.; Rohner, U.; Gonin, M.; Fuhrer, K.; Kulmala, M.; Worsnop, D. R. A High-resolution Mass Spectrometer to Measure Atmospheric Ion Composition. Atmos. Meas. Technol. 2010, 3, 1039− 1053. (30) Jokinen, T.; Sipilä, M.; Junninen, H.; Ehn, M.; Lönn, G.; Hakala, J.; Petäjä, T.; Mauldin, R. L., III; Kulmala, M.; Worsnop, D. R. Atmospheric Sulphuric Acid and Neutral Cluster Measurements Using CI-APi-TOF. Atmos. Chem. Phys. 2012, 12, 4117−4125. (31) Spartan’08; Wavefunction Inc.: Irivine CA, 2008. Spartan’10; Wavefunction Inc.: Irivine CA, 2011. Spartan’14; Wavefunction Inc.: Irivine CA, 2014. (32) Chai, J. D.; Head-Gordon, M. Long-Range Corrected Hybrid Density Functionals with Damped Atom-atom Dispersion Corrections. Phys. Chem. Chem. Phys. 2008, 10, 6615−6620. (33) 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 D.01; Gaussian, Inc.: Wallingford, CT, 2009. (34) Adler, T. B.; Knizia, G.; Werner, H.-J. A Simple and Efficient CCSD(T)-F12 Approximation. J. Chem. Phys. 2007, 127, 221106(1− 4). (35) Peterson, K. A.; Adler, T. B.; Werner, H. J. Systematically Convergent Basis Sets for Explicitly Correlated Wavefunctions: The Atoms H, He, B-Ne, and Al-Ar. J. Chem. Phys. 2008, 128, 084102(1− 12). (36) Werner, H.-J.; Knowles, P. J.; Manby, F. R.; Schütz, M.; Celani, P.; Knizia, G.; Korona, T.; Lindh, R.; Mitrushenkov, A.; Rauhut, G.; et al. MOLPRO, a Package of Ab Initio Programs, Version 2010.1; see http://www.molpro.net. (37) Eckart, C. The Penetration of a Potential Barrier by Electrons. Phys. Rev. 1930, 35, 1303−1309. (38) Wine, P. H.; Astalos, R. J.; Mauldin, R. L., III. Kinetic and Mechanistic Study of the OH + HCOOH Reaction. J. Phys. Chem. 1985, 89, 2620−2624. (39) Vaghjiani, G. L.; Ravishankara, A. R. Kinetics and Mechanism of OH Reaction with CH3OOH. J. Phys. Chem. 1989, 93, 1948−1959. (40) Aschmann, S. M.; Tuazon, E. C.; Arey, J.; Atkinson, R. Products of the Gas-Phase Reaction of O3 with Cyclohexene. J. Phys. Chem. A 2003, 107, 2247−2255. (41) Cusick, R. D.; Atkinson, R. Rate Constants for the Gas-phase Reactions of O3 with a Series of Cycloalkenes at 296 ± 2 K. Int. J. Chem. Kinet. 2005, 37, 183−190. (42) Vereecken, L.; Peeters, J. A Theoretical Study of the OHInitiated Gas-phase Oxidation Mechanism of β-Pinene (C10H16): First Generation Products. Phys. Chem. Chem. Phys. 2012, 14, 3802−3815. (43) Jagiella, S.; Libuda, H. G.; Zabel, F. Thermal Stability of Carbonyl Radicals Part I. Straight-Chain and Branched C4 and C5 Acyl Radicals. Phys. Chem. Chem. Phys. 2000, 2, 1175−1181. (44) Peeters, J.; Vereecken, L.; Fantechi, G. The Detailed Mechanism of the OH-initiated Atmospheric Oxidation of α-Pinene: a Theoretical Study. Phys. Chem. Chem. Phys. 2001, 3, 5489−5504. (45) da Silva, G. Hydroxyl Radical Regeneration in the Photochemical Oxidation of Glyoxal: Kinetics and Mechanism of the HC(O)CO + O2 Reaction. Phys. Chem. Chem. Phys. 2010, 12, 6698− 6705. (46) Johnson, D.; Marston, G. The Gas-Phase Ozonolysis of Unsaturated Volatile Organic Compounds in the Troposphere. Chem. Soc. Rev. 2008, 37, 699−716.

Issues of Atmospheric Significance. Chem. Soc. Rev. 2012, 41, 6294− 6317. (7) Vereecken, L.; Francisco, J. S. Theoretical Studies of Atmospheric Reaction Mechanisms in the Troposphere. Chem. Soc. Rev. 2012, 41, 6259−6293. (8) Crounse, J. D.; Nielsen, L. B.; Jørgensen, S.; Kjaergaard, H. G.; Wennberg, P. O. Autoxidation of Organic Compounds in the Atmosphere. J. Phys. Chem. Lett. 2013, 4, 3513−3520. (9) Crounse, J. D.; Knap, H. C.; Ørnsø, K. B.; Jørgensen, S.; Paulot, F.; Kjaergaard, H. G.; Wennberg, P. O. Atmospheric Fate of Methacrolein. 1. Peroxy Radical Isomerization Following Addition of OH and O2. J. Phys. Chem. A 2012, 116, 5756−5762. (10) Mattill, H. A. The Mechanism of the Autoxidation of Fats. Oil Soap 1941, 18, 73−76. (11) Glowacki, D. R.; Pilling, M. J. Unimolecular Reactions of Peroxy Radicals in Atmospheric Chemistry and Combustion. ChemPhysChem 2010, 11, 3836−3843. (12) Pilling, M. J. Interactions Between Theory and Experiment in the Investigation of Elementary Reactions of Importance in Combustion. Chem. Soc. Rev. 2008, 37, 676−685. (13) Taatjes, C. A. Uncovering the Fundamental Chemistry of Alkyl + O2 Reactions via Measurements of Product Formation. J. Phys. Chem. A 2006, 110, 4299−4312. (14) Mattill, H. A. Antioxidants and the Autoxidation of Fats. J. Biol. Chem. 1931, 90, 141−151. (15) Hamilton, L. A.; Olcott, H. S. Antioxidants and the Autoxidation of Fats. Oil Soap 1936, 5, 127−129. (16) Brodnitz, M. H. Autoxidation of Saturated Fatty Acids. A Review. J. Agric. Food Chem. 1968, 6, 994−999. (17) Amorati, R.; Foti, M. C.; Valgimigli, L. Antioxidant Activity of Essential Oils. J. Agric. Food Chem. 2013, 61, 10835−10847. (18) Bayston, J. H.; King, N. K.; Looney, F. D.; Winfield, M. E. Superoxocobalamin, the first intermediate in the autoxidation of vitamin B12r. J. Am. Chem. Soc. 1969, 91, 2775−2779. (19) Karlberg, A.-T.; Börje, A.; Johansen, J. D.; Lidén, C.; Rastogi, S.; Roberts, D.; Uter, W.; White, I. R. Activation of Non-sensitizing or Low-sensitizing Fragrance Substances into Potent Sensitizers − Prehaptens and Prohaptens. Contact Dermatitis 2013, 69, 323−334. (20) Hämäläinen, T. I.; Sundberg, S.; Mäkinen, M.; Kaltia, S.; Hase, T.; Hopia, A. Hydroperoxide Formation During Autoxidation of Conjugated Linoleic Acid Methyl Ester. Eur. J. Lipid Sci. Technol. 2001, 103, 588−593. (21) Perrin, O.; Heiss, A.; Sahetchian, K.; Kerhoas, L.; Einhorn, J. Determination of the Isomerization Rate Constant HOCH2CH2CH2CH(OO·)CH3 → HOC·HCH2CH2CH(OOH)CH3. Importance of Intramolecular Hydroperoxy Isomerization in Tropospheric Chemistry. Int. J. Chem. Kinet. 1998, 30, 875−887. (22) Ehn, M.; Thornton, J. A.; Kleist, E.; Sipilä, M.; Junninen, H.; Pullinen, I.; Springer, M.; Rubach, F.; Tillmann, R.; Lee, B.; et al. A Large Source of Low-Volatility Secondary Organic Aerosol. Nature 2014, 506, 476−479. (23) Rissanen, M. P.; Kurtén, T.; Sipilä, M.; Thornton, J. A.; Kangasluoma, J.; Sarnela, N.; Junninen, H.; Jørgensen, S.; Schallhart, S.; Kajos, M. K.; et al. The Formation of Highly Oxidized Multifunctional Products in the Ozonolysis of Cyclohexene. J. Am. Chem. Soc. 2014, 136, 15596−15606. (24) Donahue, N. M.; Kroll, J. H.; Pandis, S. N.; Robinson, A. L. A Two-dimensional Volatility Basis Set Part 2: Diagnostics of Organicaerosol Evolution. Atmos. Chem. Phys. 2012, 12, 615−634. (25) Kulmala, M.; Toivonen, A.; Mäkelä, J. M.; Laaksonen, A. Analysis of the Growth of Nucleation Mode Particles Observed in Boreal Forest. Tellus B 1998, 50, 449−462. (26) Schobesberger, S.; Junninen, H.; Bianchi, F.; Lönn, G.; Ehn, M.; Lehtipalo, K.; Dommen, J.; Ehrhart, S.; Ortega, I. K.; et al. Molecular Understanding of Atmospheric Particle Formation from Sulfuric Acid and Large Oxidized Organic Molecules. Proc. Natl. Acad. Sci. U. S. A. 2013, 110, 17223−17228. (27) Zhao, J.; Ortega, J.; Chen, M.; McMurry, P. H.; Smith, J. N. Dependence of Particle Nucleation and Growth on High Molecular P

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A (47) Epstein, S. A.; Donahue, N. M. Ozonolysis of Cyclic Alkenes as Surrogates for Biogenic Terpenes: Primary Ozonide Formation and Decomposition. J. Phys. Chem. A 2010, 114, 7509−7515. (48) Donahue, N. M.; Drozd, G. T.; Epstein, S. A.; Presto, A. A.; Kroll, J. H. Adventures in Ozoneland: Down the Rabbit-Hole. Phys. Chem. Chem. Phys. 2011, 13, 10848−10857. (49) Nguyen, T. L.; Winterhalter, R.; Moortgat, G.; Kanawati, B.; Peeters, J.; Vereecken, L. The Gas-Phase Ozonolysis of BetaCaryophyllene (C15H24). Part II: A Theoretical Study. Phys. Chem. Chem. Phys. 2009, 11, 4173−4183. (50) Atkinson, R.; Baulch, D. L.; Cox, R. A.; Hampson, R. F., Jr.; Kerr, J. A.; Rossi, M. J.; Troe, J. Evaluated Kinetic, Photochemical and Heterogeneous Data for Atmospheric Chemistry: Supplement V, IUPAC Subcommittee on Gas Kinetic Data Evaluation for Atmospheric Chemistry. J. Phys. Chem. Ref. Data 1997, 26, 521−1011. (51) Atkinson, R. Kinetics and Mechanisms of the Gas-Phase Reactions of the Hydroxyl Radical with Organic Compounds under Atmospheric Conditions. Chem. Rev. 1986, 86, 69−201. (52) Vereecken, L.; Peeters, J. A Structure-Activity Relationship for the Rate Coefficient of H-Migration in Substituted Alkoxy Radicals. Phys. Chem. Chem. Phys. 2010, 12, 12608−12620. (53) Vereecken, L.; Peeters, J. Decomposition of Substituted Alkoxy Radicals-Part I: A Generalized Structure-Activity Relationship for Reaction Barrier Heights. Phys. Chem. Chem. Phys. 2009, 11, 9062− 9074. (54) 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. (55) Rissanen, M. P.; Eskola, A. J.; Nguyen, T. L.; Barker, J. R.; Liu, J.; Liu, J.; Halme, E.; Timonen, R. S. CH2NH2 + O2 and CH3CHNH2 + O2 Reaction Kinetics: Photoionization Mass Spectrometry Experiments and Master Equation Calculations. J. Phys. Chem. A 2014, 118, 2176−2186. (56) Miller, J. A.; Pilling, M. J.; Troe, J. Unravelling Combustion Mechanisms Through a Quantitative Understanding of Elementary Reactions. Proc. Combust. Inst. 2005, 30, 43−88. (57) Hasson, A. S.; Tyndall, G. S.; Orlando, J. J. A Product Yield Study of the Reaction of HO2 Radicals with Ethyl Peroxy (C2H5O2), Acetyl Peroxy (CH3C(O)O2), and Acetonyl Peroxy (CH3C(O)CH2O2) Radicals. J. Phys. Chem. A 2004, 108, 5979−5989. (58) Kwan, A. J.; Chan, A. W. H.; Ng, N. L.; Kjaergaard, H. G.; Seinfeld, J. H.; Wennberg, P. O. Peroxy Radical Chemistry and OH Radical Production During the NO3-Initiated Oxidation of Isoprene. Atmos. Chem. Phys. 2010, 12, 7499−7515. (59) Orlando, J. J.; Tyndall, G. S.; Wallington, T. J. The Atmospheric Chemistry of Alkoxy Radicals. Chem. Rev. 2003, 103, 4657−4689. (60) Atkinson, R.; Kwok, E. S. C.; Arey, J.; Aschmann, S. M. Reactions of Alkoxyl Radicals in the Atmosphere. Faraday Discuss. 1995, 100, 23−37. (61) Hasson, A. S.; Tyndall, G. S.; Orlando, J. J.; Singh, S.; Hernandez, S. Q.; Campbell, S.; Ibarra, Y. Branching Ratios for the Reaction of Selected Carbonyl-Containing Peroxy Radicals with Hydroperoxy Radicals. J. Phys. Chem. A 2012, 116, 6264−6281. (62) Shallcross, D. E.; Teresa Raventos-Duran, M.; Bardwell, M. W.; Bacak, A.; Solman, Z.; Percival, C. J. A Semi-empirical Correlation for the Rate Coefficients for Cross- and Self-reactions of Peroxy Radicals in the Gas-Phase. Atmos. Environ. 2005, 39, 763−771. (63) Madronich, S.; Calvert, J. G. Permutation Reactions of Organic Peroxy Radicals in the Troposphere. J. Geophys. Res. 1990, 95, 5697− 5715. (64) Lightfoot, P. D.; Cox, R. A.; Crowley, J. N.; Destriau, M.; Hayman, G. D.; Jenkin, M. E.; Moortgat, G. K.; Zabel, F. Organic Peroxy Radicals: Kinetics, Spectroscopy and Tropospheric Chemistry. Atmos. Environ. 1992, 26, 1805−1964. (65) Wolf, J. L.; Richters, S.; Pecher, J.; Zeuch, T. Pressure Dependent Mechanistic Branching in the Formation Pathways of Secondary Organic Aerosol from Cyclic-alkene Gas-Phase Ozonolysis. Phys. Chem. Chem. Phys. 2011, 13, 10952−10964.

(66) Carlsson, P. T. M.; Dege, J. E.; Keunecke, C.; Krüger, B. C.; Wolf, J. L.; Zeuch, T. Pressure Dependent Aerosol Formation from the Cyclohexene Gas-Phase Ozonolysis in the Presence and Absence of Sulfur Dioxide: a New Perspective on the Stabilisation of the Initial Clusters. Phys. Chem. Chem. Phys. 2012, 14, 11695−11705. (67) Fenske, J. D.; Kuwata, K. T.; Houk, K. N.; Paulson, S. E. OH Radical Yields from the Ozone Reaction with Cycloalkenes. J. Phys. Chem. A 2000, 104, 7246−7254. (68) Jenkin, M. E.; Shallcross, D. E.; Harvey, J. N. Development and Application of a Possible Mechanism for the Generation of cis-pinic Acid from the Ozonolysis of Alpha and Beta Pinene. Atmos. Environ. 2000, 34, 2837−2850. (69) Fantechi, G.; Vereecken, L.; Peeters, J. The OH-initiated Atmospheric Oxidation of Pinonaldehyde: Detailed Theoretical Study and Mechanism Construction. Phys. Chem. Chem. Phys. 2002, 4, 5795−5805. (70) Master Chemical Mechanism, MCM v3.2;http://mcm.leeds.ac. uk/MCM. Jenkin.; et al. Atmos. Environ. 1997, 31, 81. Saunders.; et al. Atmos. Chem. Phys. 2003, 3, 161. (71) Capouet, M. Modeling the Oxidation of Alpha-Pinene and the Related Aerosol Formation in Laboratory and Atmospheric Conditions. Ph.D. Thesis, 2005; http://tropo.aeronomie.be/pdf/PhD_ Capouet.pdf. (72) Vereecken, L.; Peeters, J. H-atom Abstraction by OH-radicals from (Biogenic) (Poly)alkenes: C−H Bond Strengths and Abstraction Rates. Chem. Phys. Lett. 2001, 333, 162−168. (73) Eddingsaas, N. C.; Loza, C. L.; Yee, L. D.; Seinfeld, J. H.; Wennberg, P. O. α-Pinene Photooxidation under Controlled Chemical Conditions − Part 1: Gas-Phase Composition in Low- and High-NOx Environments. Atmos. Chem. Phys. 2012, 12, 6489−6504. (74) Müller, V. L.; Reinnig, M. C.; Naumann, K. H.; Saathoff, H.; Mentel, T. F.; Donahue, N. M.; Hoffmann, T. Formation of 3-Methyl1,2,3-butanetricarboxylic Acid via Gas phase Oxidation of Pinonic Acid − a Mass Spectrometric Study of SOA Aging. Atmos. Chem. Phys. 2012, 12, 1483−1496. (75) Meloni, G.; Peng Zou, P.; Klippenstein, S. J.; Ahmed, M.; Leone, S. L.; Taatjes, C. A.; Osborn, D. L. Energy-Resolved Photoionization of Alkylperoxy Radicals and the Stability of Their Cations. J. Am. Chem. Soc. 2006, 128, 13559−13567. (76) Sharp, E. N.; Rupper, P.; Miller, T. A. The Structure and Spectra of Organic Peroxy Radicals. Phys. Chem. Chem. Phys. 2008, 10, 3955− 3981. (77) Vereecken, L.; Peeters, J. Nontraditional (Per)oxy Ring-Closure Paths in the Atmospheric Oxidation of Isoprene and Monoterpenes. J. Phys. Chem. A 2004, 108, 5197−5204. (78) Birdsall, A. W.; Andreoni, J. F.; Elrod, M. J. Investigation of the Role of Bicyclic Peroxy Radicals in the Oxidation Mechanism of Toluene. J. Phys. Chem. A 2010, 114, 10655−10663. (79) Elrod, M. J. Kinetics Study of the Aromatic Bicyclic Peroxy Radical + NO Reaction: Overall Rate Constant and Nitrate Product Yield Measurements. J. Phys. Chem. A 2011, 115, 8125−8130. (80) Chhabra, P. S.; Lambe, A. T.; Canagaratna, M. R.; Stark, H.; Jayne, J. T.; Onasch, T. B.; Davidovits, P.; Kimmel, J. R.; Worsnop, D. R. Chemistry of α-Pinene and Naphthalene Oxidation Products Generated in a Potential Aerosol Mass (PAM) Chamber As Measured by Acetate Chemical Ionization Mass Spectrometry. Atmos. Meas. Technol. Discuss. 2014, 7, 6385−6429. (81) Nørgaard, A. W.; Nøjgaard, J. K.; Clausen, P. A.; Wolkof, P. Secondary Ozonides of Substituted Cyclohexenes: A New Class of Pollutants Characterized by Collision-induced Dissociation Mass Spectrometry Using Negative Chemical Ionization. Chemosphere 2008, 70, 2032−2038. (82) Treacy, J.; Curley, M.; Wenger, J.; Sidebottom, H. Determination of Arrhenius Parameters for the Reactions of Ozone with Cycloalkenes. J. Chem. Soc. Faraday Trans. 1997, 93, 2877−2881. (83) Gao, S.; Keywood, M.; Ng, N. L.; Surratt, J.; Varutbangkul, V.; Bahreini, R.; Flagan, R. C.; Seinfeld, J. H. Low-Molecular-Weight and Oligomeric Components in Secondary Organic Aerosol from the Q

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX

Article

The Journal of Physical Chemistry A Ozonolysis of Cycloalkenes and α-Pinene. J. Phys. Chem. A 2004, 108, 10147−10164. (84) Nøjgaard, J. K.; Nørgaard, A. W.; Wolkoff, P. Secondary Ozonides of Endo-cyclic Alkenes Analyzed by Atmospheric Sampling Townsend Discharge Ionization Mass Spectrometry. Int. J. Mass. Spectrom. 2007, 263, 88−93. (85) Orzechowska, G. E.; Nguyen, H. T.; Paulson, S. E. Photochemical Sources of Organic Acids. 2. Formation of C-5-C-9 Carboxylic Acids from Alkene Ozonolysis under Dry and Humid Conditions. J. Phys. Chem. A 2005, 109, 5366−5375.

R

DOI: 10.1021/jp510966g J. Phys. Chem. A XXXX, XXX, XXX−XXX