Direct probing of Criegee intermediates from gas-phase ozonolysis using chemical ionization mass spectrometry

Transcription

1 Supplementary Information Direct probing of Criegee intermediates from gas-phase ozonolysis using chemical ionization mass spectrometry Torsten Berndt 1 *, Hartmut Herrmann 1 and Theo Kurtén 2 1 Leibniz-Institute for Tropospheric Research, TROPOS, Leipzig, Germany. 2 Department of Chemistry, University of Helsinki, Helsinki, Finland. S1

2 1 Quantum-chemical calculations 1.1 Computational details: For configurational sampling, we applied a slightly modified version of the approach recently proposed and tested for hydrogen shifts of polyfunctional peroxy radicals [S1]. Systematic configurational sampling (by rotating over all dihedral bonds) was first carried out with the Spartan 14 program [S2] using the MMFF force-field [S3]. For the cations and anions formed by adding or removing protons from Criegee Intermediates, the default charge assignment by MMFF was found to be incorrect; the FFHINT keyword was then used to enforce the correct overall charge by changing the charge assigned to the carbon atom of the Criegee (carbonyl oxide) functional group. The conformer generation was followed by B3LYP/6-31+G(d) [S4] optimizations and frequency calculations on all unique structures identified in the MMFF conformer generation, including the initial MMFF runs with incorrect overall charges (the latter being included just in case they led to different unique minima at the B3LYP level). The conformational sampling routine automatically generated conformers for both syn and anti isomers of the Criegee intermediates, which possess this isomery. (Unlike conformers, the syn and anti isomers have large barriers preventing their interconversion test calculations indicate that this applies also to protonated or deprotonated Criegee Intermediates). After the B3LYP optimizations, the two conformer sets for each Criegee Intermediate with syn/anti isomery were separated, and treated as separate systems in all subsequent stages. The cis/trans isomers of the vinylhydroperoxide C6O3H10 (both formed from the syn isomer of the corresponding Criegee Intermediate, but via different transition states, and different orientations with respect to the C=C bond) were treated similarly, as were the syn/anti isomers of the C6O3H10 secondary ozonide. However, the anti-isomer of the secondary ozonide exhibited high ring strain, and was more than 50 kcal/mol above the syn-isomer in energy. Its formation can thus be considered unlikely, and it was omitted from further study. For each system, all conformers within 2 kcal/mol of the lowest-energy conformer at the B3LYP/6-31+G(d) level were selected for ωb97x-d/aug-cc-pvtz [S5] optimizations and frequency calculations using Gaussian 09 Rev. D.01 [S6], with the ultrafine integration grid. When the number of conformers below 2 kcal/mol in relative energy was less than 25, the lowest 25 were picked for the higher-level optimizations. For systems with less than 25 conformers overall, all conformers were selected. Next, all conformers within 2 kcal/mol of the lowest-enthalpy conformer at the ωb97x-d/aug-ccpvtz level was selected for CCSD(T)-F12a/VDZ-F12 [S7] single-point energy calculations using Molpro [S8]. When the number of conformers below 2 kcal/mol in relative enthalpy at the ωb97x-d/aug-cc-pvtz level was less than 10, the lowest 10 were picked for the CCSD(T)-F12a/VDZ-F12 calculations (and when there were less than 10 conformers overall, all conformers were selected). Finally, the singlepoint energy of the lowest-enthalpy conformer at the CCSD(T)-F12a/VDZ-F12//ωB97X- D/aug-cc-pVTZ level was recomputed at the computationally much more demanding CCSD(T)-F12a/VTZ-F12 level. (The difference between VDZ-F12 and VTZ-F12 proton affinities, as well as between the F12a and F12b variants of CCSD(T)-F12, were found to be minor, on the order of a few kj/mol or less). The final enthalpy was then obtained by combinding the thermal enthalpy contributions (which include the vibrational zeropoint energy) from the ωb97x-d/aug-cc-pvtz calculation with the CCSD(T)-F12a/VTZ- F12 electronic energy. S2

3 Proton affinities (enthalpies of protonation reactions) were computed by assigning the free proton an enthalpy of 2.5RT, consisting of a translational enthalpy of 1.5RT, and an extra 1RT corresponding to the pv term added by Gaussian 09 to all the computed thermal enthalpy contributions (which include vibrational zero-point energies in addition to the genuine thermal contributions). This definition of the proton enthalpy allows the enthalpies obtained from the quantum chemical calculations to be directly subtracted from each other. The conformational sampling of the protonated forms of the dioxirane, vinylhydroperoxide and secondary ozonide structural isomers of C6O3H10 sometimes led to artefact structures. For the two former cases, this occurred especially when other protonation sites than the aldehydic oxygen where tested. For example, protonating the dioxirane group often led to protonated Criegee intermediates, or to even lower-energy ether-like structures with hydroxyl and carbonyl functionalities. While such structures may be real in the sense of being distinct minima on the C 6O 3H 11 +potential energy surface, their formation from any of the isomers studied were would require breaking multiple covalent bonds, and thus likely have prohibitively high barriers. Such artefacts were therefore discarded from further study, and were not included in the proton affinity calculations. Note that the C O or C C interactions present in the lowest-energy conformers of the protonated/deprotonated C 6O 3H 10 Criegee Intermediates (see Scheme 1 below) do not require breaking covalent bonds in order to form. The lowest-energy conformer found for the cis and trans isomers of the protonated C 6O 3H 10 vinyl hydroperoxides (see Scheme 3 below) are borderline cases, as they contain a very strong and close interaction between the C2 (vinylic) and C6 (protonated aldehydic) carbon atoms, which could be considered a new bond (as the C2-C6 distances are as low as 1.55 Å). If the C2-C6 interaction is classified as a bond, then the structure can be considered a protonated cyclic hydroxyl Criegee Intermediate. According to this definition, the π-bond of the vinyl hydroperoxide would have been broken (though the distance between the hydroperoxide C1 carbon and the C2 carbon is still below 1.50Å, indicating some remaining double bonding character). 1.2 Proton affinities: We computed proton affinities for the charger species tetrahydrofuran, diethylether, diethylamine, n-butylamine and tert-butylamine, as well as for the Criegee Intermediates H 2COO, CH 3CHOO, (CH 3) 2COO, CH 3CH 2CHOO and OHC(CH 2) 4CHOO. We also considered alternative structural isomers for 1, 3 and 6-carbon Criegee Intermediates (with acid and dioxirane functionalities for the two former cases, and acid, dioxirane, vinyl hydroperoxide and secondary ozonide functionalities for the latter case). OHC(CH 2) 4CHOO is the Criegee species formed in the ozonolysis of cyclohexene. CH 3CHOO and OHC(CH 2) 4CHOO have both syn and anti isomers, with the former facing a carbon atom with an α-hydrogen, and the latter not. By this definition, H2COO is formally anti and (CH3)2COO formally syn. Based on the results by Nguyen et al. [S9], Criegee Intermediates were assumed to protonate at the terminal oxygen atom of the Criegee (carbonyl oxide) functional group. We also computed proton affinities for the anion formed by removing the aldehydic hydrogen of OHC(CH2)4CHOO (this hydrogen is the weakest bound based both on general chemical understanding as well as test calculations). The C6 dioxirane and vinyl hydroperoxide species all preferred to protonate at the aldehyde group, while the C6 acid preferred to protonate at the acid group. The computed proton affinities are presented in Table 1 and Table S1, and the lowest-energy conformers of the neutral, protonated and deprotonated S3

4 OHC(CH2)4CHOO Criegee Intermediates are shown in Scheme S1. The lowest-energy neutral and protonated conformers for the various structural isomers of OHC(CH2)4CHOO are shown in Schemes S2 and S3. Scheme S1. Lowest-energy structures of the neutral, protonated and deprotonated OHC(CH 2) 4CHOO Criegee intermediates formed in cyclohexene ozonolysis, optimized at the ωb97x-d/aug-cc-pvtz level, with single-point calculations at the CCSD(T)- F12a/VDZ-F12 level used to determine the lowest-energy conformer. Left column: synisomers. Right column: anti-isomers. Top row: neutral Criegee Intermediates. Middle row: protonated Criegee Intermediates. Bottom row: deprotonated Criegee Intermediates. The distance between the oxygen or carbon of the aldehyde group and the Criegee (carbonyl oxide) carbon is given in Ångström for the cationic and anionic species, respectively. Color coding: gray=c, white=h, red=o. S4

5 Scheme S2. Lowest-energy structures of neutral (left-hand column) and protonated (right-hand column) C 6H 10O 3 structural isomers: acid (top row), dioxirane (middle row) and secondary ozonide (bottom row). Structures have been optimized at the ωb97x- D/aug-cc-pVTZ level, with single-point calculations at the CCSD(T)-F12a/VDZ-F12 level used to determine the lowest-energy conformer. Color coding as in Scheme 1. Scheme S3. Lowest-energy structures of neutral and protonated cis (top row) and trans (bottom row) isomers of the C6H10O3 vinyl hydroperoxides. The left-hand column shows the neutral vinylhydroperoxides, the middle column the lowest-energy structures found in the conformational sampling, and the right-hand column the lowest-energy protonated structures with the vinylhydroperoxide group intact. Key C-C distances in S5

6 the lowest-energy cationic structures are given in Ångström. Structures have been optimized at the ωb97x-d/aug-cc-pvtz level, with single-point calculations at the CCSD(T)-F12a/VDZ-F12 level used to determine the lowest-energy conformer. Color coding as in Scheme 1. Table S1. CCSD(T)-F12a/VTZ-F12//ωB97X-D/aug-cc-pVTZ proton affinities ( K) for the studied species. Criegee intermediate species indicated in bold. Species Proton affinity, kj/mol Tetrahydrofuran Diethylether CH2OO Formic acid C-dioxirane syn-ch 3CHOO anti-ch3choo syn-ch3ch2choo anti-ch3ch2choo (CH 3) 2COO Propionic acid C3 secondary dioxirane C3 primary dioxirane n-butylamine tert-butylamine Diethylamine syn-ohc(ch2)4choo anti-ohc(ch 2) 4CHOO OHC(CH 2) 4COOH acid C6 dioxirane syn-c6 secondary ozonide cis-c 6 vinyl hydroperoxide a / b trans- C 6 vinyl hydroperoxide a / b syn-oc(ch2)4choo anti-oc(ch 2) 4CHOO a)including the protonated structures with a C2-C6 bond (see Scheme S3) b)excluding the protonated structures with a C2-C6 bond (see Scheme S3) The computed proton affinities for the charger species are within a few kj/mol of the experimental values. The proton affinity for H 2COO is about 5 kj/mol larger than that given by Nguyen et al. [S9]. As our CCSD(T)-F12a/VTZ-F12 values are likely very close to the CCSD(T) basis-set limit (as indicated by a difference of less than 1.5 kj/mol between CCSD(T)-F12a/VDZ-F12 and CCSD(T)-F12a/VTZ-F12 proton affinities for all neutral species), and the core-valence, scalar-relativistic and spin-orbit corrections computed by Nguyen et al. mostly cancel out in the proton affinity calculation, the difference is likely caused by the use of DFT geometries and zero-point energies in our study, as opposed to coupled-cluster ones in theirs. Given the relatively large size of the cyclohexene-derived Criegee Intermediates, which form the main focus of our study, the S6

7 agreement between our results and literature values can be considered satisfactory. We note that while the CCSD(T)/F12 calculations with the larger VTZ-F12 basis set changed the results very little compared to those obtained using the computationally much cheaper VDZ-F12 basis, the CCSD(T)-F12 and ωb97x-d/aug-cc-pvtz proton affinities for the Criegee Intermediates were significantly different, with the latter being some tens of kj/mol larger. A comparison of the proton affinities of the different Criegee Intermediates reveals some interesting trends. As expected, increasing the size of the system increases the proton affinity, likely due to increased stabilization and/or delocalization of the charge in the protonated cation. This explains the increase of about kj/mol in going from H 2COO to the two CH 3CHOO isomers, kj/mol in going from H 2COO to the two CH 3CH 2CHOO isomers, and 83 kj/mol in going from H 2COO to (CH 3) 2COO. However, the very high proton affinities found for the OHC(CH 2) 4CHOO isomers are somewhat surprising. As the carbonyl oxide group in this structure located on a terminal carbon atom, the proton affinity could be expected to be fairly close to that of CH 3CH 2CHOO, with some modest extra stabilization due to the additional carbon atoms beyond the β-carbon. Instead, the differences in proton affinities between the corresponding isomers of CH3CH2CHOO and OHC(CH2)4CHOO are over 60 kj/mol. The reason for this is evident from Scheme 1: In the lowest-energy conformers of the protonated OHC(CH2)4CHOOH + cations, the oxygen atom of the aldehyde group forms a strong interaction with the partially positively charged carbon atom of the protonated carbonyl oxide group. While the distance between the C and O atoms is not quite short enough for the interaction to be considered a covalent bond, it very likely explains the anomalously high proton affinity, as conformers lacking the interaction have relative energies tens of kj/mol higher. Note that even these conformers still have proton affinities considerably higher than CH 3CH 2CHOO due to hydrogen bonding between the aldehyde and the protonated carbonyl oxide groups. An analogous interaction between the carbon atoms of the deprotonated aldehyde group and the carbonyl oxide group explains the similarly anomalously high proton affinity of the OC(CH 2) 4CHOO - anion. For all three of the Criegee Intermediates possessing syn/anti isomery, the anti isomer has a proton affinity about 15 kj/mol higher than the syn isomer. This is caused by the fact that while the neutral syn isomer is lower in energy than the neutral anti isomer, the opposite is true for the protonated cations. This may be related to slightly stronger repulsive interactions and ensuing steric strain in the syn isomer of the cation, where the -OOH + proton and the C-H hydrogens are closer to each other. Interestingly, the syn/anti difference is almost identical for the CH 3CHOO, CH 3CH 2CHOO and OHC(CH 2) 4CHOO systems, despite the strong aldehyde-carbonyl oxide interaction in the protonated forms of the latter. The large difference in proton affinities between the Criegee Intermediate and its structural isomers persists as the system size and complexity is increased from the CH2O2 to C3H6O2 to C6H10O3. The protonated C6 dioxirane and acid are both significantly stabilized by hydrogen bonding (as evident from Scheme 2 and Table 1), but the difference in proton affinities between the dioxirane and acid on one hand and the two Criegee Intermediates on the other is still on the order of 100 kj/mol. The C6 vinyl hydroperoxide has a proton affinity comparable to the C6 acid if the cyclic structure illustrated in the middle column of Scheme 3 can form without a significant barrier (otherwise the proton affinity of the vinyl hydroperoxide is comparable to that of the dioxirane). Clearly, no other C6H10O3 isomer than the Criegee Intermediate is capable of being protonated by the studied amines. S7

8 1.3 Clustering enthalpies with (THF)H + : We computed clustering enthalpies at the CCSD(T)-F12a/VTZ-F12//ωB97X- D/aug-cc-pVTZ level for clusters of protonated tetrahydrofuran (THF)H + for three different isomers of CH 2O 2: the Criegee Intermediate, the dioxirane, and formic acid. For the Criegee Intermediate, for which the computed proton affinity is higher than that of (THF)H +, we also compute the enthalpy of the (CH 2OO)H + + THF reaction, as the reverse of this pathway is the more favorable cluster break-up channel. The results are shown in Table S2, with the cluster structures illustrated in Scheme S4. It can be seen that the (CH2OO)(THF)H + cluster is significantly more strongly bound (and thus stable) than the other structural isomers. This strong binding is caused by the proton being almost equally shared between the CH2OO and THF moieties due to their similar proton affinities, as shown in Scheme S4. Scheme S4. ωb97x-d/aug-cc-pvtz structures of the protonated clusters of tetrahydrofuran with three CH 2O 2 structural isomers: Criegee Intermediate (top), dioxirane (middle), formic acid (bottom). Color coding as in Scheme S1. Table S2. CCSD(T)-F12a/VTZ-F12//ωB97X-D/aug-cc-pVTZ reaction enthalpies ( K) for the dissociation of the clusters of three CH 2 O 2 structural isomers with protonated tetrahydrofuran. Reaction Enthalpy, kj/mol (CH 2OO)(THF)H + => CH 2OO + (THF)H (CH 2OO)(THF)H + => (CH 2OO)H + + THF (dioxirane)(thf)h + => dioxirane + (THF)H (HCOOH)(THF)H + => HCOOH + (THF)H S8

9 The clusters of the different C3H6O2 isomers (secondary, primary anti and primary syn Criegee Intermediates, propionic acid, secondary and primary dioxirane) with (THF)H + were also investigated. For computational reasons this was done at the CCSD(T)-F12a/VDZ-F12//ωB97X-D/aug-cc-pVTZ level. Limited configurational sampling was carried out by first optimizing an arbitrary conformer at the ωb97x- D/aug-cc-pVTZ level for each cluster, freezing the bond distances and the angle of the O H + O moiety connecting the two components of the cluster, and then rotating over the dihedral angles of the molecules (using the conformational sampling tool of the Spartan program, and the MMFF force field). The produced conformers were then optimized at the ωb97x-d/6-31+g(d) level without constraints, and the lowest-enthalpy conformer was selected for the final CCSD(T)-F12a/VDZ-F12//ωB97X-D/aug-cc-pVTZ calculation. (For the secondary Criegee Intermediate and dioxirane, only one conformer was found.) For all the Criegee intermediates, dissociation into protonated Criegee Intermediates and neutral THF was by far the most favorable cluster breakup channel, so only data for this channel is shown. It is evident from Table S3 that the differences in stabilities between the different (C3H6O2)(THF)H + clusters are relatively small compared to the (CH2O2)(THF)H + clusters. This is likely due to the increase in proton affinities with system size, which on one hand makes the C3H6O2 acid and dioxiranes better proton acceptors, and thus stronger ligands for (THF)H +, while on the other hand making neutral THF a less competitive ligand with the protonated Criegee Intermediates, as the proton transfer to the Criegee Intermediate in the cluster becomes complete (see Scheme S5). Scheme S5. ωb97x-d/aug-cc-pvtz structures of the protonated clusters of tetrahydrofuran with six C3H6O2 structural isomers: Left top: primary syn-criegee Intermediate; Left middle: primary anti-criegee Intermediate; Left bottom: secondary Criegee Intermediate; Right top: primary dioxirane ; Right middle: secondary dioxirane, Right bottom: propionic acid. Color coding as in Scheme 1. S9

10 Table S3. CCSD(T)-F12a/VDZ-F12//ωB97X-D/aug-cc-pVTZ reaction enthalpies ( K) for the dissociation of the clusters of six C 3 H 6 CO 2 structural isomers with protonated tetrahydrofuran. Reaction Enthalpy, kj/mol (syn-ch 3CH 2CHOO)(THF)H + => syn-ch 3CH 2CHOOH + + THF (anti-ch3ch2choo)(thf)h + => anti-ch3ch2chooh + + THF (CH3COOCH3)(THF)H + => CH3COOH + CH3 + THF (prim-c 3H 6O 2-dioxirane)(THF)H => prim-c 3H 6O 2-dioxirane + (THF)H + (sec-c 3H 6O 2-dioxirane)(THF)H => sec- C 3H 6O 2-dioxirane + (THF)H + (CH3CH2COOH)(THF)H + => CH3CH2COOH + (THF)H S10

11 2 Supplementary Figures Figure S1: Ion traces at the nominal mass of 47 Th, standing for the protonated species of a compound with the chemical formula CH2O2, and at nominal 119 Th, (THF)H + adduct of a CH2O2 species, for different formic acid concentrations in the sample gas. A measurement cycle comprises 60 s signal accumulation. Figure S2: CH 2OO adduct signal, (CH 2OO)(THF)H +, as a function of reacted ethylene, reagent ion: protonated tetrahydrofuran, (THF)H +. Open symbols stand for a measurement series for constant ethylene concentration of molecules cm -3 and variable ozone concentrations up to molecules cm -3. The closed symbols represent results for a constant ozone concentration of molecules cm -3 and ethylene concentrations of up to molecules cm -3. S11

12 Figure S3: CH2OO adduct signal, (CH2OO)(THF)H +, as a function of reacted ethylene for an ozone concentration of molecules cm -3 and ethylene concentrations of up to molecules cm -3, reagent ion: protonated tetrahydrofuran, (THF)H +. S12

13 References [S1] Møller, K. H.; Hyttinen, N.; Otkjær, R. V.; Kurtén, T.; Kjaergaard, H. G. J. Phys. Chem. A 2016, 120, [S2] Spartan 14; Wavefunction Inc.: Irivine CA, [S3] Halgren, T. A. J. Comput. Chem. 1999, 20, 730. [S4] (a) Becke, A. D. J. Chem. Phys. 1993, 98, (b) Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B: Condens. Matter Mater. Phys. 1988, 37, 785. [S5] (a) Chai, J. D.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2008, 10, (b) Dunning, T. H. J. Chem. Phys. 1989, 90, (c) Kendall, R. A.; Dunning, T. H.; Harrison, R. J. J. Chem. Phys. 1992, 96, [S6] 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.; Nakatsuji, H.; Caricato, M., Li, X.; Hratchian, H. P.; Izmaylov, A. F.; Bloino, J.; Zheng, G.; Sonnenberg, J. L.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.; Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai, H.; Vreven, T.; Montgomery, Jr., J. A.; Peralta, J. E.; Ogliaro, F.; Bearpark, M.; Heyd, J. J.; Brothers, E.; Kudin, K. N.; Staroverov, V. N.; Kobayashi, R.; Normand, J.; Raghavachari, K.; Rendell, A.; Burant, J. C.; Iyengar, S. S.; Tomasi, J.; Cossi, M.; Rega, N.; Millam, J. M.; Klene, M.; Knox, J. E.; Cross, J. B.; Bakken, V.; Adamo, C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin, A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Martin, R. L.; Morokuma, K.; Zakrzewski, V. G.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Dapprich, S.; Daniels, A. D.; Farkas, Ö.; Foresman, J. B.; Ortiz, J. V.; Cioslowski, J.; Fox, D. J. Gaussian 09, Revision D.01; Gaussian, Inc.: Wallingford, CT, [S7] Adler, T. B.; Knizia, G.; Werner, H.-J. J. Chem. Phys. 2007, 127, [S8] Werner, H.-J.; Knowles, P. J.; Knizia, G.; Manby, F. R.; Schu tz, M.; Celani, P.; Korona, T.; Lindt, R.; Mitrushenkov, A.; Rauhut, G.; Shamasundar, K. R.; Adler, T. B.; Amos, R. D.; Bernhardsson, A.; Berning, A.; Cooper, D. L.; Deegan, M. J. O.; Dobbyn, A. J.; Eckert, F.; Goll, E.; Hampel, C.; Hesselmann, A.; Hetzer, G.; Hrenar, T.; Jansen, G.; Köppl, C.; Liu, Y.; Lloyd, A. W.; Mata, R. A.; May, A. J.; McNicholas, S. J.; Meyer, W.; Mura, M. E.; Nicklass, A.; O'Neill, D. P.; Palmieri, P.; Peng, D.; Pflüger, K.; Pitzer, R.; Reiher, M.; Shiozaki, T. ; Stoll, H.; Stone, A. J.; Tarroni, R.; Thorsteinsson, T.; Wang, M. MOLPRO, version , a package of ab initio programs; 2012; see [S9] Nguyen, M. T.; Nguyen, T. L.; Ngan, V. T.; Nguyen, H. M. T. Chem. Phys. Lett. 2007, 448, 183. S13