Crossed-Beams and Theoretical Studies of the Multichannel Reaction O(3P) + 1,2-Butadiene (Methylallene): Product Branching Fractions and Role of Intersystem Crossing

Abstract

The reactions of ground state oxygen atoms, O(³P), with unsaturated hydrocarbons (UHs) are relevant in the oxidation in different environments. They are usually multichannel reactions that exhibit a variety of competing product channels, some of which occur adiabatically on the entrance triplet potential energy surface (PES), while others occur nonadiabatically on the singlet PES that can be accessed via intersystem crossing (ISC). ISC plays a key role on the mechanism of these reactions, impacting greatly the product yields. Identification of all primary reaction products, determination of their branching fractions (BFs), and assessment of the role of ISC is central for understanding the mechanism of these reactions. This goal can be best achieved combining crossed-molecular-beam (CMB) experiments with universal, soft ionization, mass-spectrometric detection and time-of-flight analysis to high-level ab initio electronic structure calculations of triplet/singlet PESs and Rice-Ramsperger-Kassel-Marcus/Master Equation (RRKM/ME) computations of product BFs with inclusion of ISC effects. Over the years this combined approach was found to be rewarding and successful for O(³P) reactions with the simplest alkynes, alkenes, and dienes containing two, three, or four carbon atoms. Here, we report the full experimental and theoretical work on the reaction O(³P) + 1,2-butadiene that permits us to explore how the mechanism and product distribution vary when moving from O(³P) + allene (propadiene) to O(³P) + methylallene (1,2-butadiene) and when comparing this system to related C4 unsaturated systems, namely O(³P) + 1-butene and O(³P) + 1,3-butadiene. In the present CMB experiments at the collision energy of 41.8 kJ/mol we have observed and characterized nine different product channels. Synergistic ab initio transition-state theory-based master equation simulations coupled with nonadiabatic transition-state theory on the coupled triplet/singlet PESs were used for computing the product BFs and assisting the interpretation of the experimental results. Theoretical predictions and experimental results were found to be in overall good agreement. The finding of this work can be useful for the kinetic modeling of the oxidation of 1,2-butadiene and of systems involving 1,2-butadiene as an important intermediate.

1

List of the 32 energetically allowed isomeric product channels of the O­(³P) + 1,2-butadiene reaction.

1. 1,2-Butadiene Compared with the Simplest Diene, Allene, and with the Corresponding 4C Alkene, 1-Butene, and Conjugated Diene, 1-3-Butadiene

2

Newton (velocity vector) diagram for the O­(³P) + 1,2-butadiene reaction (E _c = 41.8 kJ/mol) with superimposed circles delimiting the maximum speed in the CM reference frame (the velocity vector of the CM is labeled, with Θ_CM = 44°) and the LAB scattering angular range of the various, indicated primary products. The velocity vectors of the two reactants species, which cross each other at 90° in the LAB frame, and the velocity vector of the CM of the system are indicated. Nine different reactive channels were identified: C₄H₅O + H (blue line), C₄H₄O + H₂ (red line), CH₃CO + C₂H₃ (green line), CH₂CHO + C₂H₃, CH₂CO + CHCH₃ (orange line), CO + C₃H₆ (magenta), HCO + C₃H₅ (cyan line), H₂CO + C₃H₄ (navy line) and C₃H₃O + CH₃ (black line). The O-containing primary products are represented by solid lines, while the dashed lines correspond to the hydrocarbon coproducts. Circles of the two coproducts are depicted only for the channels of CO, CH₃, HCO, and CH₂CO formation. Continuous circles refer to oxygenated coproducts, while dashed circles to hydrocarbon molecule/radical coproducts.

3

Product LAB angular distributions measured at m/z = 68, 43, 42, 41­(lhs) and 30, 29, 27, and 15 (rhs) for the O­(³P) + 1,2- butadiene reaction at E _c = 41.8 kJ/mol. The black solid curve superimposed on the experimental data (black dots) corresponds to the calculated global best-fit using the CM functions depicted in the Figures and . The separate contributions to the calculated global LAB product angular distributions are color-coded and indicated with the formula of the corresponding product (color coding as in Figure ).

4

(lhs) TOF distributions measured at Θ_CM = 44° for m/z = 68 (70 eV) and m/z = 43, 42, and 41 exploiting soft ionization (17 eV). (rhs) TOF distributions measured for m/z = 29, 28, 27, and 15 at Θ_CM = 44° exploiting soft ionization (17 eV). In each panel the black solid curve superimposed on the experimental data (empty dots) corresponds to the calculated global best-fit using the CM functions depicted in Figures and . The distinct contributions to the calculated global LAB TOF distributions are color-coded and indicated with the formula of the corresponding product (color coding as in Figure ).

5

TOF distributions for m/z = 68 at two different angles (Θ = 28° and 44°) at 70 eV (hard ionization). The black solid curve represents the calculated global best-fit using the CM functions shown in the Figures and . At this m/z value the broad slow peak is related to the heavy coproduct of the H channel 1 (blue), while the faster shoulder corresponds to C₄H₄O from the H₂ elimination channel 2 (red).

6

TOF distributions measured for m/z = 43 at Θ = 28° (corresponding to the forward direction) and Θ = 44° (corresponding to the CM) exploiting soft ionization (17 eV). The solid black curve represents the calculated best-fit by using the CM functions shown in the Figure ). At this m/z value the slower peaks are due to the two products already characterized at m/z = 68, namely the heavy coproduct of the H channel (blue) and the H₂ channel (red), while the faster peak is related unambiguously to acetyl formation (green).

7

TOF distributions measured at m/z = 42 at two different angles (Θ = 28° and 44°) by exploiting soft ionization at 17 eV. The black solid curve indicates the calculated best-fit using the CM functions depicted in Figures and . Besides the products characterized at m/z = 68 and 43, at this m/z value also propene (magenta), ketene (purple), and vinoxy products have been identified. The best-fit at this mass is a significant refinement with respect to that presented in Figure 3 of ref .

8

TOF distributions measured for m/z = 41 at Θ = 28° (forward direction) and Θ = 44° (CM angle) exploiting soft ionization at 17 eV. The black solid curve represents the calculated best-fit by using the best-fit CM functions shown in the Figures and . At this m/z value, in addition to the reaction products characterized at m/z = 68, 43, and 42, the heavy coproduct C₃H₅ (allyl radical) (cyano line) of the HCO radical (channel (7)) was identified. Note that also this best-fit is a significant refinement with respect to that presented initially in Figure 4 of ref .

9

TOF distributions registered for m/z = 29 at Θ = 28° and Θ = 44° exploiting soft ionization at 17 eV. The black solid curve represents the calculated best-fit using the best-fit CM functions reported in the Figures and . At this m/z value CHCH₃ and HCO products are momentum matched to CH₂CO and C₃H₅ (channels (10) and (7), respectively) that were characterized at m/z = 42 and 41, respectively (see Figures and ).

10

TOF distributions registered for m/z = 27 at Θ = 28° and Θ = 44° exploiting soft ionization at 17 eV. The black solid curve superimposed to the experimental data (empty dots) represents the calculated best-fit using the best-fit CM functions shown in Figures and . The various contributions at this mass are indicated (color coding is as in the Figures –).

11

TOF distributions registered for m/z = 15 at Θ = 28° and Θ = 44° exploiting soft ionization at 17 eV. The black solid curve represents the calculated best-fit using the best-fit CM functions shown in the Figures and . At this mass-to-charge ratio CH₃ formation was identified via its parent ion CH₃ ⁺, corroborating the identification of its heavy coproduct (C₃H₃O) at higher m/z values (see Figures (rhs), , and ). Other contributions at this mass are also indicated (color coding as in Figure ).

12

(lhs): Best-fit CM angular distributions determined for the C₄H₅O (blue), C₄H₄O (red), CH₃CO (green), CH₂CHO (violet) and CH₂CO (orange) products. (rhs): Best-fit CM translational energy distributions determined for channels (1a,c,f), (2a,b), (8), and (9), whose total available energy (E _TOT) (given by E _c – ΔH₀ ⁰) and average translational energy fractions (⟨f _T⟩) are also shown. The shaded areas represent the error bounds of the derived best-fit CM functions.

13

(lhs): Best-fit CM angular distributions determined for the indicated CH₂CO (orange), C₃H₆ (magenta), CH₃ (black), H₂CO (navy), and HCO (cyan) products. (rhs): Best-fit CM translational energy distributions determined for channels (10a–c), (3c), (4a,b), (5a–c), and (7b,c), whose total energy and translational energy fraction are also shown. The shaded areas represent the error bounds of the derived best-fit CM functions.

14

Potential energy surface (simplified; for full PES see triplet and singlet PESs for C1 attack depicted in the Figures S1 and S2) of the O­(³P) to 1,2-butadiene system deriving from the electrophilic O­(³P) addition to C1. Additional ¹W1 and ¹W6 pathways are described in the PES of Figure . The PES is calculated at the CCSD­(T)/aug-cc-pVTZ//ωB97XD/aug-cc-pVTZ level of theory, with values in parentheses obtained at the CASPT2 level. The reaction pathways highlighted in bold red and blue solid lines refer to the main reaction pathways accessed by the reactive flux.

15

Potential energy surface (simplified; for full PES see triplet and singlet PESs for C2 attack depicted in the Figures S3 and S4) of the O­(³P) + 1,2-butadiene system deriving from the electrophilic O­(³P) addition to C2. The PES is calculated at the CCSD­(T)/aug-cc-pVTZ//ωB97XD/aug-cc-pVTZ level of theory, with values in parentheses obtained at the CASPT2 level. The reaction pathways highlighted in bold red and blue solid lines refer to the main reaction pathways accessed by the reactive flux.

16

Potential energy surface (simplified; for full PES see triplet and singlet PESs for C3 attack depicted in the Figures S5 and S6) of the O­(³P) + 1,2-butadiene system deriving from the electrophilic O­(³P) addition to C3. Additional ³W2 pathways are described in the potential energy surface of Figure . The PES is calculated at the CCSD­(T)/aug-cc-pVTZ//ωB97XD/aug-cc-pVTZ level of theory, with values in parentheses obtained at the CASPT2 level. The reaction pathways highlighted in bold red and blue solid lines refer to the main reaction pathways accessed by the reactive flux.

17

Theoretical total and attack site-specific rate constants for the O­(³P) + 1,2-butadiene addition and H-abstraction from methyl reactions as a function of temperature, compared with the experimental data of Deslauries et al. at 293 K (the experimental uncertainty does not exceed the symbol size).

18

Schematization of the reaction fluxes that follow O­(³P) addition to the C1, C2, and C3 sites, leading respectively to the formation of the ³W₁, ³W₂, and ³W₃ entrance wells with 0.157, 0.555, and 0.288 BFs, respectively. The main reaction pathways that are followed after O­(³P) addition are connected by bold lines, while the minor pathways considered in the calculations are shaded. The pathways followed after addition to C1, C2, and C3 are highlighted in green, orange, and blue, respectively.