CH₂OO Criegee biradical yields following photolysis of CH₂I₂ in O₂

Abstract

Yields of CH₂OO and CH₂IO₂ from the reaction of CH₂I radicals with O₂ are reported as a function of total pressure, [N₂] and [O₂] at T = 295 K using three complementary methods. Results from the three methods are similar, with no observed additional dependence on [O₂]. The CH₂I + O₂ reaction has a yield of ∼18% CH₂OO at atmospheric pressure.

---

Criegee biradicals (CR₂OO) are key reaction intermediates in the ozonolysis of unsaturated organic compounds,¹ and their involvement in the atmospheric oxidation of alkenes has long been postulated.^2,3 Despite much effort, direct observations of Criegee biradicals have only recently been reported for CH₂OO^4–8 and CH₃CHOO.⁹ Photolysis of CH₂I₂ in the presence of O₂ has been shown to produce CH₂OO at low pressures through the reactions:^5,10

CH₂I₂ + hν → CH₂I + I (R1)

However, the reaction between CH₂I and O₂ proceeds via initial formation of an excited complex, CH₂IOO^#, which has the potential for collisional stabilisation to produce CH₂IO₂ peroxy radicals (R2b) in competition with production of the CH₂OO Criegee biradical (R2a) (Scheme 1):¹⁰

CH₂I + O₂ → CH₂IOO^#

CH₂IOO^# → CH₂OO + I (R2a)

CH₂IOO^# + M → CH₂IO₂ + M (R2b)

A number of investigations by Enami and co-workers^11–13 suggested production of HCHO and IO from CH₂I + O₂, but other studies,^14,15,33 including measurements in this laboratory,¹⁵ have demonstrated that the production of IO results from secondary processes, and that IO is not a direct reaction product of (R2).^14,15,33

Scheme 1 Chemical activation scheme to describe the reaction between CH₂I and O₂(R2), where the initially formed excited species CH₂IO₂^# either proceeds to produce CH₂OO + I (k_2a) or is collisionally stabilised to produce the CH₂IO₂ peroxy radical or (k_2b[M]).

The yields of CH₂OO and CH₂IO₂ from CH₂I + O₂ were recently measured by Huang et al.¹⁰ as a function of [N₂], [O₂] and [He] by monitoring the I atoms produced in (R1) and (R2a)via their infrared absorption owing to F′′ = 4 → F′ = 3 of the ²P_3/2 → ²P_1/2 spin–orbit transition at 7603.138 cm⁻¹. Given the stoichiometry between CH₂I and I in (R1), it is possible to infer the fraction of CH₂I radicals producing CH₂OO through comparison of the I atom yields from the instant photolytic production in (R1) and the slower production via(R2a). While there is potential for multi-photon dissociation of CH₂I₂ to produce CH₂ + 2I, it is expected that this is relatively minor compared to production of CH₂I + I.^16–19 Huang et al. showed that the yield of CH₂OO decreases with total pressure, consistent with collisional stabilisation of the CH₂IOO^# intermediate to CH₂IO₂.

However, Huang et al. also reported significant differences in the I atom yields from (R2a) (and thus in CH₂OO yields) between experiments performed in N₂ buffer gas and those performed in O₂, indicating a much greater efficiency of O₂ for stabilisation of CH₂IOO^# to CH₂IO₂ compared to N₂, and an unusual interaction between CH₂IOO^# and O₂.

In this work we report observations of the yields of CH₂OO and CH₂IO₂ from CH₂I + O₂ following laser flash photolysis of CH₂I₂–N₂–O₂ gas mixtures as a function of [N₂], [O₂] and total pressure using several complementary methods at total pressures between 25 and 450 Torr. Experiments were initially performed to monitor I atom fluorescence, thus enabling inference of the yields of CH₂OO and CH₂IO₂ in the manner described by Huang et al.¹⁰ Subsequent experiments monitored the yields of HCHO from reactions of CH₂OO–CH₂IO₂ in the presence of excess SO₂ or NO by laser-induced fluorescence (LIF) of HCHO at λ ∼ 353.1 nm. Full experimental details are given in the ESI.† All experiments were performed at T = 295 K unless stated otherwise.

Fig. 1 shows the typical I atom signal following photolysis of CH₂I₂–O₂–N₂. The instantaneous photolytic production of iodine atoms through (R1) can be clearly distinguished from the subsequent growth in (R2a). The I atom signals were analysed using eqn (1):

where [I]_t is the iodine atom signal at time t, S₀ is the amplitude of the instant photolytic signal resulting from (R1), S₁ is the amplitude of the iodine atom signal resulting from the slower growth process occurring after photolysis, k₂′ is the pseudo-first-order rate coefficient for (R2) (i.e. k₂′ = k₂[O₂]), and k_loss is the rate coefficient representing the slow loss of iodine atoms from the detection region via reaction or diffusion. A value of k₂ = (1.67 ± 0.04) × 10⁻¹² cm³ s⁻¹ was determined in this work (see ESI†), in agreement with previous measurements of (1.40 ± 0.35) × 10⁻¹² cm³ s⁻¹,²⁰ and (1.6 ± 0.2) × 10⁻¹² cm³ s⁻¹.²¹ All errors quoted for this work are statistical at the 1σ level unless stated otherwise.

Fig. 1 Iodine atom signal following photolysis of CH₂I₂ in the presence of O₂. For this plot total pressure (N/V) = 3.27 × 10¹⁷ cm⁻³ (∼10 Torr, predominantly N₂); [O₂] = 4.02 × 10¹⁵ cm⁻³; [CH₂I₂] = 5.03 × 10¹² cm⁻³. The time resolution is such that iodine production from photolysis and reaction can be identified. The fit to eqn (1) is shown by the solid line.

The absolute iodine atom yield from reaction (R2a) is given by the ratio S₁/S₀, and was observed to decrease with increasing total pressure of N₂, consistent with production of the CH₂OO Criegee biradical at low pressures and stabilisation of the chemically activated CH₂IO₂^# species to the CH₂IO₂ peroxy radical at higher pressures. Solution of the I atom yield from (R2) (Φ_I), and thus the CH₂OO yield, is given by the Stern–Volmer relationship in eqn (2):

where Φ_I(R2) is the iodine atom yield from (R2) (i.e. S₁/S₀), k_2b/k_2a is the Stern–Volmer quenching coefficient, and [M] is the total number density of the system.

Fig. 2 shows the Stern–Volmer plot for reaction (R2). The intercept of the iodine atom Stern–Volmer plot is 1.08 ± 0.12, consistent with channel 2a being the dominant bimolecular process. The slope of the Stern–Volmer plot gives the Stern–Volmer quenching coefficient (k_2b/k_2a), and is equal to (2.28 ± 0.11) × 10⁻¹⁹ cm³ for these experiments, similar to the value of k_2b/k_2a = (3.1 ± 0.2) × 10⁻¹⁹ cm³ reported by Huang et al.¹⁰ for experiments in N₂ buffer gas. For stabilisation of CH₂IO₂^# by O₂, Huang et al. report a value of k_2b/k_2a = (4.09 ± 0.32) × 10⁻¹⁸ cm³. The iodine atom experiments in this work were conducted at low [O₂] (∼4 × 10¹⁵ cm⁻³) to ensure (R2a) was sufficiently slow to provide confidence in the resolution of the photolytic I atom production from the chemical I atom production. The effects of CH₂IO₂^# stabilisation by O₂ were thus investigated in the HCHO yield experiments.

Fig. 2 Stern–Volmer analyses for CH₂OO yields from CH₂I + O₂ as a function of total pressure. Main panel shows results from this work monitoring iodine atom production (squares; intercept = 1.08 ± 0.12; slope = (2.28 ± 0.11) × 10⁻¹⁹ cm³), and HCHO production in the presence of SO₂ (triangles; intercept = 1.46 ± 0.25; slope = (0.95 ± 0.24) × 10⁻¹⁹ cm³) and NO (circles; intercept = 1.41 ± 0.30; slope = (1.33 ± 0.31) × 10⁻¹⁹ cm³). Constraining the intercepts to unity for fits to SO₂ and NO data gives slopes of (1.37 ± 0.10) × 10⁻¹⁹ cm³ and (1.71 ± 0.16) × 10⁻¹⁹ cm³, respectively. Data shown for SO₂ and NO were taken over a range of [O₂] ((0.1–7.8) × 10¹⁸ cm⁻³). A fit to all data reported in this work gives an intercept of 1.10 ± 0.23 and a slope of (1.90 ± 0.22) × 10⁻¹⁹ cm³ (shown by the solid line). Error bars shown on the plot and those given for the fits are 1σ, with fits weighted to the experimental errors. Separate lines of best fit for results from the different methods are not shown for clarity but are given in the ESI.† The inset plot shows results from this work together with parameterisations given by Huang et al. for N₂ (solid light grey line), O₂ (broken black line) and air (broken dark grey line).

Fig. 3a shows a typical kinetic trace for HCHO following photolysis of CH₂I₂–O₂–N₂ in the absence of any additional co-reagent (i.e. SO₂ or NO), in which HCHO is produced in the system by reactions (R3–R6):

CH₂OO + I → HCHO + IO (R3)

CH₂IO₂ + I → CH₂IO + IO (R4)

CH₂IO₂ + CH₂IO₂ → 2CH₂IO + O₂ (R5)

CH₂IO → HCHO + I (R6)

Although the HCHO production in this system is not strictly pseudo-first-order, Gravestock et al.¹⁵ have shown that the growth from reactions (R3–R6) can be approximated to pseudo-first-order behaviour, and thus the data can be fitted using eqn (3):^14,15where [HCHO]_t is the HCHO signal at time t, S₀ is the amplitude of the HCHO signal at time zero, S₁ is the maximum HCHO signal, k_g′ is the pseudo-first-order rate coefficient for HCHO growth, and k_loss is the rate coefficient representing the slow loss of HCHO from the detection region via diffusion. Some initial HCHO production was observed owing to multi-photon photolysis of CH₂I₂ and the subsequent rapid reaction of ³CH₂ with O₂,^16–19 with S₀ typically no greater than 5–10% of S₁. In the present experiments k_g′ was typically ∼500 s⁻¹, which is one to two orders of magnitude slower than the reactions occurring when SO₂ or NO were added to the system. Simulations performed with the numerical integration package Kintecus²² (provided in the ESI†) indicate that eqn (3) faithfully describes the yields of HCHO (i.e. S₁) in this system. Reactions (R3–R6) imply that all the CH₂OO and CH₂IO₂ react to form formaldehyde, i.e. all the CH₂I radicals are converted to HCHO. The recent study by Huang et al.¹⁰ has demonstrated that the Criegee radical, CH₂OO, is formed with or near unity yields at low pressures from reaction (R2). The validity of 100% production of HCHO in the system can be tested at low pressures with Criegee reactions that produce formaldehyde. At low pressures the reaction between CH₂OO with SO₂ is known to produce 100% HCHO,²³ and below we demonstrate that the total HCHO yield in the system is the same with and without the addition of SO₂, only the timescale for its formation varies.

Fig. 3 HCHO fluorescence signals following photolysis of CH₂I₂ in the presence of O₂. Panel (a) shows HCHO signals at 150 Torr in the absence of any co-reagent with the fit to eqn (3). Panel (b) shows HCHO signals at 250 Torr in the presence of SO₂, with the fit to eqn (4). Panel (c) shows HCHO signals at 250 Torr in the presence of NO, with the fit to eqn (4). The inset plots in (b) and (c) show the evolution of the signals to longer times.

Experiments conducted in excess SO₂ or NO did not result in a decrease in the HCHO yield on addition of the co-reagent, indicating complete titration of both CH₂OO and CH₂IO₂ to HCHO. In both cases biexponential growth of HCHO was observed, as shown in Fig. 3b and c, with the observed HCHO signal in both cases described by eqn (4):

where [HCHO]_t is the HCHO signal at time t, S₀ is the amplitude of the HCHO signal at time zero, S₁ is the maximum HCHO signal, k_g1′ is the pseudo-first-order rate coefficient for the fast HCHO growth, k_g2′ is the pseudo-first-order rate coefficient for the slower HCHO growth, f is the fractional contribution of the fast growth process to the total HCHO yield (and hence (1−f) is the fractional contribution of the slower growth process to the total HCHO yield), and k_loss is the rate coefficient representing the slow loss of HCHO from the detection region via diffusion. f is therefore related to the yield of the Criegee, eqn (2), as we demonstrate below. Again, in experiments conducted with a photolysis wavelength of 248 nm, some initial HCHO production was observed owing to multi-photon photolysis of CH₂I₂ and the subsequent rapid reaction of ³CH₂ with O₂,^16–19 with S₀ typically no greater than 5–10% of S₁. In experiments using a photolysis wavelength of 355 nm, S₀ = 0.

For both SO₂ and NO experiments, the rate of the initial fast HCHO growth displayed a linear dependence on [SO₂] or [NO], respectively, with k_g1′ determined over the range 5000–60 000 s⁻¹. The rate of the slower secondary growth was independent of [SO₂] or [NO], and occurred at a similar rate to the HCHO growth observed in the absence of any additional co-reagent, and thus attributed to HCHO production viareaction (R3) or (R4–R6). The fact that k_g1′ ≫ k_g2′ means that f is reliably determined, and that k_g1′ is determined without any significant influence from the more complicated kinetics associated with the slower kinetics, k_g2′.

The fast HCHO in the presence of SO₂ is consistent with production from CH₂OO + SO₂:

CH₂OO + SO₂ → HCHO + SO₃ (R7)

The slower growth of HCHO occurs as a result of reactions (R4–R6). Reaction of the peroxy radical with SO₂ is unlikely,²⁴ and the slower HCHO growth is not dependent on [SO₂]. As [SO₂] is in large excess over radicals in the system, k_g1′ in eqn (4) is given by k_g1′ = k₇[SO₂], while k_g2′ approximates the growth of HCHO through reactions (R4–R6). The returned value of f in this case is equal to the CH₂OO yield.

Reactions of peroxy radicals with NO are typically fast (for example, k_CH3O2+NO = 7.7 × 10⁻¹² cm³ s⁻¹ at 298 K²⁵), while Welz et al.⁵ reported an upper limit for the rate coefficient for reaction of CH₂OO with NO of <6 × 10⁻¹⁴ cm³ s⁻¹. Therefore, we propose that the fast HCHO growth in experiments with NO results from the reaction of CH₂IO₂ with NO (R8) followed by the rapid decomposition of CH₂IO to HCHO and I in (R6),¹⁵ with the slower growth resulting from (R3):

CH₂IO₂ + NO → CH₂IO + NO₂ (R8)

CH₂IO → HCHO + I (R6)

As [NO] is in large excess over the other radicals in the system, k_g1′ in eqn (4) is therefore given by k_g1′ = k₈[NO], while k_g2′ approximates the growth of HCHO through reaction (R3). The returned values f in this case are thus equal to the yields of CH₂IO₂.

While the slower growth of HCHO is not strictly pseudo-first-order, but is treated as such by eqn (4), simulations (described in the ESI†) show that the yields of HCHO from the two growth processes are well described by eqn (4) and the yields from the two processes (i.e. S₁ and f) are faithfully determined by fitting to eqn (4). In both systems, the rate of the fast growth process (6000–60 000 for SO₂; 5000–20 000 s⁻¹ for NO) is significantly faster than that of the slower growth process (∼300–500 s⁻¹), ensuring that the two growth processes are essentially decoupled and the HCHO yields from the two growth processes can be distinguished, and that the rate coefficient describing the fast growth is equal to that for the pseudo-first-order reactions, (R7) or (R8). We assign no kinetic information to k_g2′ for either system, and as shown in the ESI,† the approximation of the slower growth process to pseudo-first-order kinetics leads to uncertainties in the yields of only 2–3%.

As noted above, there was no change in the total HCHO yield in the system upon the addition of either SO₂ or NO, and this was observed to be the case at all total pressures. At low pressures where the CH₂OO yield is close to unity, the addition of SO₂ leads to reaction (R7) and formation of HCHO with close to 100% yield.²³ In the ESI,† Fig. S3 compares the HCHO signal in the system with and without the addition of SO₂. The fact that both traces observe the same amount of HCHO in the system adds validity to the assumption that in the absence of reagents, reactions (R3–R6), lead to 100% HCHO formation. At higher total pressures where CH₂IO₂ formation is significant, the reason the HCHO yield is still 100% is because there is no reaction between the peroxy radical and SO₂, which is in agreement with the literature,²⁴ and the peroxy radical is titrated to HCHO via(R4–R6). In the case of NO, it is the peroxy radical that reacts rapidly with the NO (R8) to form HCHO, but there is no significant reaction between the CH₂OO and NO, in accord with the results from Welz et al.,⁵ and therefore HCHO is formed on a slow timescale viareaction (R3). This again leads to 100% yield of HCHO in the system independent of total pressure, in accord with the data.

Thus, the fractional contributions of the fast and slow growth processes to the total HCHO yields in the presence of NO and SO₂ can be used to identify the yields of CH₂OO and CH₂IO₂ from the reaction of CH₂I with O₂. The fractional contribution of the fast growth process to the total HCHO yield in the presence of NO thus reflects the yield of CH₂IO₂ from (R2) (i.e. Y_CH2IO2 = k_2b[M]/(k_2a + k_2b[M]) = f_NO and Y_CH2OO = k_2a/(k_2a + k_2b[M]) = 1 − f_NO), while the fractional contribution of the fast growth process to the total HCHO yield in the presence of SO₂ reflects the yield of CH₂OO from (R2) (i.e. Y_CH2OO = k_2a/(k_2a + k_2b[M]) = f_SO2 and Y_CH2IO2 = k_2b[M]/(k_2a + k_2b[M]) = 1 − f_SO2).

Fig. 2 also shows the Stern–Volmer analysis for CH₂OO yields determined by the SO₂ and NO experiments (i.e. Stern–Volmer plots for 1/f_SO2 and 1/(1−f_NO), respectively). Experiments with SO₂ (triangles) give k_2b/k_2a = (0.95 ± 0.24) × 10⁻¹⁹ cm³, while those with NO (circles) give k_2b/k_2a = (1.33 ± 0.31) × 10⁻¹⁹ cm³, with intercepts of 1.46 ± 0.25 and 1.41 ± 0.30, respectively. Constraining the intercepts to unity in the fits to data from the SO₂ and NO experiments gives k_2b/k_2a = (1.37 ± 0.10) × 10⁻¹⁹ cm³ and k_2b/k_2a = (1.71 ± 0.16) × 10⁻¹⁹ cm³, respectively. Further details can be found in the ESI.† The relative errors in the SO₂ and NO experiments are typically larger than those for the iodine atom experiments owing to the need to fit a greater number of parameters in eqn (4) compared to eqn (1), and the smaller range of pressures in the Stern–Volmer plot for which the yields can be determined by the SO₂ or NO method (at low and high pressures, where one of CH₂OO or CH₂IO₂ dominates the HCHO growth it is difficult to resolve the two growth components, and thus to retrieve the relative yields, in the fit to eqn (4)). From Fig. 2, the fact that within error there is reasonable agreement in the CH₂OO yields from the HCHO and the iodine atom experiments is further indication that all the sources of HCHO in each of the systems are understood and defined.

Experiments in both SO₂ and NO were performed over a range of O₂ concentrations, with measurements taken using 100% O₂ buffer gas in both cases (see Fig. 4), in order to test if O₂ has a significant effect on the CH₂OO yield. In contrast to the work of Huang et al.,¹⁰ no dependence of k_2b/k_2a on [O₂] was observed in any of our measurements. A fit to all our data reported here gives k_2b/k_2a = (1.90 ± 0.22) × 10⁻¹⁹ cm⁻³, with an intercept of (1.10 ± 0.23). Huang et al. noted that their observed difference in CH₂IO₂^# stabilisation efficiency by N₂ and O₂ was an unusual result, with N₂ and O₂ often displaying similar collisional stabilisation efficiencies. It was proposed that O₂ may not be acting as a simple collision partner to remove excess energy in CH₂IO₂^#, but that there may be a reactive process occurring between O₂ and CH₂IO₂^#, potentially resulting in production of HCHO, IO and O₂. However, an investigation of CH₂I + O₂ by Gravestock et al.¹⁵ could not identify IO as a product of the reaction even when more than 10% of O₂ was present at 30 Torr total pressure, and our measurements of HCHO yields in this work are not consistent with the production of HCHO from this reaction.

Fig. 4 Inverse of CH₂OO yields from CH₂I + O₂ as a function of [O₂] for experiments with (a) SO₂ at 150 Torr (circles) and 350 Torr (squares); (b) NO at 50 Torr (open squares), 150 Torr (circles), 250 Torr (triangles), 350 Torr (filled squares) and 450 Torr (inverted triangles). Horizontal lines show the average inverse CH₂OO yield at each pressure for all experiments SO₂ (panel a) and NO (panel b). Error bars are 1σ in the fits to eqn (4).

At present, there does not appear to be any simple explanation as to the differences between this work and the work of Huang et al. in the apparent yields of CH₂OO and CH₂IO₂ from CH₂I + O₂ as a function of pressure. While the work of Huang et al. indicates a CH₂OO yield of only ∼4% in air at 760 Torr, our results indicate a yield of ∼18%, with potentially significant implications for the oxidation chemistry of halogen containing organic compounds and for our understanding of atmospheric chemistry in marine regions with high concentrations of species such as CH₂I₂.^26–32

In conclusion, we have measured the yields of CH₂OO and CH₂IO₂ from the reaction of CH₂I radicals with O₂ as a function of total pressure and as a function of [N₂] and [O₂] using three complementary methods. Results from the three methods are similar, with no observed dependence of the CH₂OO yield on [O₂]. We estimate that the reaction between CH₂I and O₂ reaction has a yield of ∼18% of the CH₂OO Criegee biradical at atmospheric pressure.

Acknowledgements

The authors are grateful to the National Centre for Atmospheric Science (NCAS) and the Engineering and Physical Sciences Research Council (EPSRC, grant reference EP/J010871/1) for funding.

References

1. D. Johnson and G. Marston, Chem. Soc. Rev., 2008, 37, 699–716.
2. R. Criegee and G. Wenner, Justus Liebigs Ann. Chem., 1949, 564, 9–15.
3. R. Criegee, Angew. Chem., Int. Ed. Engl., 1975, 14, 745–752.
4. C. A. Taatjes, G. Meloni, T. M. Selby, A. J. Trevitt, D. L. Osborn, C. J. Percival and D. E. Shallcross, J. Am. Chem. Soc., 2008, 130, 11883–11885.
5. O. Welz, J. D. Savee, D. L. Osborn, S. S. Vasu, C. J. Percival, D. E. Shallcross and C. A. Taatjes, Science, 2012, 335, 204–207.
6. C. A. Taatjes, O. Welz, A. J. Eskola, J. D. Savee, D. L. Osborn, E. P. F. Lee, J. M. Dyke, D. W. K. Mok, D. E. Shallcross and C. J. Percival, Phys. Chem. Chem. Phys., 2012, 14, 10391–10400.
7. J. M. Beames, F. Liu, L. Lu and M. I. Lester, J. Am. Chem. Soc., 2012, 134, 20045–20048.
8. Y.-T. Su, Y.-H. Huang, H. A. Witek and Y.-P. Lee, Science, 2013, 340, 174–176.
9. C. A. Taatjes, O. Welz, A. J. Eskola, J. D. Savee, A. M. Scheer, D. E. Shallcross, B. Rotavera, E. P. F. Lee, J. M. Dyke, D. K. W. Mok, D. L. Osborn and C. J. Percival, Science, 2013, 340, 177–180.
10. H. Huang, A. J. Eskola and C. A. Taatjes, J. Phys. Chem. Lett., 2012, 3, 3399–3403.
11. S. Enami, T. Yamanaka, S. Hashimoto, M. Kawasaki, K. Tonokura and H. Tachikawa, Chem. Phys. Lett., 2007, 445, 152–156.
12. S. Enami, Y. Sakamoto, T. Yamanaka, S. Hashimoto, M. Kawasaki, K. Tonokura and H. Tachikawa, Bull. Chem. Soc. Jpn., 2008, 81, 1250–1257.
13. S. Enami, J. Ueda, M. Goto, Y. Nakano, S. Aloisio, S. Hashimoto and M. Kawasaki, J. Phys. Chem. A, 2004, 108, 6347–6350.
14. J. Sehested, T. Ellermann and O. J. Nielsen, Int. J. Chem. Kinet., 1994, 26, 259–272.
15. T. J. Gravestock, M. A. Blitz, W. J. Bloss and D. E. Heard, ChemPhysChem, 2010, 11, 3928–3941.
16. G. Hancock and V. Haverd, Chem. Phys. Lett., 2003, 372, 288–294.
17. H. M. Su, W. T. Mao and F. N. Kong, Chem. Phys. Lett., 2000, 322, 21–26.
18. U. Bley, F. Temps, H. G. Wagner and M. Wolf, Ber. Bunsen-Ges. Phys. Chem., 1992, 96, 1043–1048.
19. R. A. Alvarez and C. B. Moore, J. Phys. Chem., 1994, 98, 174–183.
20. A. J. Eskola, D. Wojcik-Pastuszka, E. Ratajczak and R. S. Timonen, Phys. Chem. Chem. Phys., 2006, 8, 1416–1424.
21. A. Masaki, S. Tsunashima and N. Washida, J. Phys. Chem., 1995, 99, 13126–13131.
22. J. C. Ianni, Kintecus, Windows Version 2.80, 2002, www.kintecus.com.
23. L. Vereecken, H. Harder and A. Novelli, Phys. Chem. Chem. Phys., 2012, 14, 14682–14695.
24. P. D. Lightfoot, R. A. Cox, J. N. Crowley, M. Destriau, G. D. Hayman, M. E. Jenkin, G. K. Moortgat and F. Zabel, Atmos. Environ., Part A, 1992, 26, 1805–1961.
25. R. Atkinson, D. L. Baulch, R. A. Cox, J. N. Crowley, R. F. Hampson, R. G. Hynes, M. E. Jenkin, M. J. Rossi and J. Troe, Atmos. Chem. Phys., 2006, 6, 3625–4055.
26. C. M. Roehl, J. B. Burkholder, G. K. Moortgat, A. R. Ravishankara and P. J. Crutzen, J. Geophys. Res., Atmos., 1997, 102, 12819–12829.
27. L. J. Carpenter, W. T. Sturges, S. A. Penkett, P. S. Liss, B. Alicke, K. Hebestreit and U. Platt, J. Geophys. Res., [Atmos.], 1999, 104, 1679–1689.
28. L. J. Carpenter, Chem. Rev., 2003, 103, 4953–4962.
29. C. Ordonez, J. F. Lamarque, S. Tilmes, D. E. Kinnison, E. L. Atlas, D. R. Blake, G. S. Santos, G. Brasseur and A. Saiz-Lopez, Atmos. Chem. Phys., 2012, 12, 1423–1447.
30. A. Saiz-Lopez, J. M. C. Plane, A. R. Baker, L. J. Carpenter, R. von Glasow, J. C. G. Martin, G. McFiggans and R. W. Saunders, Chem. Rev., 2012, 112, 1773–1804.
31. A. Saiz-Lopez and R. von Glasow, Chem. Soc. Rev., 2012, 41, 6448–6472.
32. L. J. Carpenter, S. D. Archer and R. Beale, Chem. Soc. Rev., 2012, 41, 6473–6506.
33. T. J. Dillon, M. E. Tucceri, R. Sander and J. N. Crowley, Phys. Chem. Chem. Phys., 2008, 10, 1540–1554.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c3cp52466c

---