Aluminum derivative peroxides in the (t-BuO)₃Al–2t-BuOOH catalytic system as a source of electron-excited dioxygen: a quantum chemical study on a model

Abstract

The quantum chemical study of the MeOOH/(MeO)₃Al model system has been carried out in order to predict the mechanism of the catalytic decomposition of t-BuOOH under mild conditions for the t-BuOOH/(t-BuO)₃Al system being a powerful synthetic tool for selective oxidation. To elucidate the chemical excitation of O₂ eliminated in the catalytic reaction and to predict the electronic state of O₂, the topology of the potential energy surface (PES), the structures of intermediates and transition states, the activation and reaction energies were obtained at the B3LYP/cc-pVTZ theory level. It was shown that the peroxide, (MeO)₂AlOOMe, corresponding to the experimentally obtained (t-BuO)₂AlOOBu-t, is formed in the first step of the reaction. After that, in the main pathway, the aluminum-containing peroxide reacts with the second MeOOH molecule through the nucleophilic substitution of the second methoxy group forming the MeOAl(OOMe)₂ diperoxide. The diperoxide rearranges to aluminum-containing ozonide MeOAlOOOMe. The ozonide isomerizes in the mononuclear-metal dioxygen intermediate (MeO)₃Al·O₂. The latter decomposes through the adiabatic ((MeO)₃Al + O₂(b¹Σ⁺_g)) and non-adiabatic ((MeO)₃Al + O₂(X³Σ⁻_g)) pathways, which corresponds to experimental data about the incomplete conversion of O₂ to O₂(b¹Σ⁺_g). The generation of O₂(b¹Σ⁺_g) was revealed by the analysis of the energy diagram calculated with the CCSD(T), CCSDT(Q), and CASSCF methods. It was suggested that the η¹-(MeO)₃Al·O₂ and, thus, (t-BuO)₃Al·O₂ complexes are new sources of O₂(b¹Σ⁺_g).

---

1. Introduction

Aluminum-containing peroxides have become of special interest for industrial organic synthesis, e.g., the original K. Ziegler's Alfol process and its modification called Epal.¹ It is known that the oxidation of organoaluminum compounds (OAC) by dioxygen to the alkoxides of higher alcohols passes through the intermediate steps of forming organoaluminum peroxides: R₂AlOOR, R(RO)AlOOR, and (RO)₂AlOOR. The organoaluminum peroxides can be an impurity in OACs used as the essential co-compounds of catalysts for the coordination polymerization of α-olefins and other monomers. However, details of the stepwise OAC oxidation mechanism are still unknown.^2,3 Chemiluminescence and, thus, chemical excitation are observed in the OAC oxidation.⁴ The high reactivity and low thermal stability of aluminum-containing peroxides are caused by the lability of the Al–OO fragment, which makes it difficult to isolate them individually and, thus, to study their reactivity.

The investigations by V. A. Dodonov et al.^5,6 of the early 1990s summarized in review⁷ point out that organoaluminum peroxides (RO)₂AlOOR′ can be synthesized individually only with tertiary alkyl (aryl) peroxo groups (see ref. 8 and ESI†). The nucleophilic substitution reactions of aluminum tri-tert-butoxide with tertiary hydroperoxides (t-BuOOH, PhC(CH₃)₂OOH, Ph₃COOH) have been studied further in ref. 5, 6 and 9–11 (Scheme 1). The corresponding organoaluminum peroxides have been obtained in the reactions with cumene hydroperoxide and trityl hydroperoxide in benzene at room temperature (Scheme 1).

Scheme 1 The high-yield synthesis of individual organoaluminum peroxides (t-BuO)₂AlOOR′ from ready available (t-BuO)₃Al.

Unlike aryl-containing hydroperoxides, tert-butyl hydroperoxide reacts with aluminum tert-butoxide yielding dioxygen and tert-butanol. The reaction of aluminum tert-butoxide with tert-butyl hydroperoxide at the mole ratio of 1 : 2 in benzene or CCl₄ yields 80–90% of O₂. The elimination of O₂ remains the main reaction^5,6 when the compound ratio is increased to 1 : 10. Thus, the catalytic reaction occurs in the (t-BuO)₃Al–2t-BuOOH system, where (t-BuO)₃Al is a catalyst and t-BuOOH is a reagent. The catalytic systems based on aluminum or titanium tert-butoxides and tert-butyl hydroperoxide oxidize organic sulfides to sulfones with the yield of 82–99% at room temperature and at the mole ratio of 1 : 2 for (t-BuO)₃Al and t-BuOOH^12–14 (see also ref. 15 and 16, review¹⁷ and references therein). Other examples of aluminum catalysts in a combination with H₂O₂ or alkyl hydroperoxide as oxidants are relatively scarce: BINOL-Al in the Bayer–Villiger reaction developed by C. Bolm et al.,^18,19 Al(salelen) for sulfides (chiral)oxidation developed by T. Katsuki.^20,21

In order to understand the processes in the (t-BuO)₃Al–2t-BuOOH catalytic system, reactions in non-reactive solvents (benzene and carbon tetrachloride) and in oxidizable solvents (various hydrocarbons) were studied in ref. 5, 6, 9 and 11. As was shown,^5,6,9,11 the dioxygen formed in the catalytic reaction decays practically completely oxidizing geminal C–H bonds in alkanes and alkylaromatic hydrocarbons. The selective oxidation of C–H bonds in alkanes and alkylarenes was studied⁸ by means of EPR spin trapping with 2-methyl-2-nitrosopropane (MNP) and C-phenyl-N-tert-butyl nitrone (PBN). The following oxygen-centered radicals were identified:⁸ t-BuO˙, t-BuOO˙, (t-BuO)₂AlOO˙, and (t-BuO)₂AlO˙. On the basis of the experimental results, the existence of di-tert-butoxy-tert-butyltrioxyaluminum, (t-BuO)₂AlOOOBu-t, and its thermolysis mechanism were suggested (Scheme 2).^6–8

Scheme 2 Tentative mononuclear aluminum–oxygen intermediates in the catalytic dissociation of t-BuOOH with singlet O₂ generation: (t-BuO)₃Al is a catalyst, t-BuOOH is a reagent.

Generation of singlet dioxygen is a subject of numerous reviews,^22–26 books,^27,28 and ongoing researches. Indispensable is the role of singlet oxygen in photodynamic therapy of cancer, photomedicine and photobiology. Reactive oxygen substances (ROS) in ground and excited electronic states are of general importance for biochemistry and biomedical applications^29–33 as well as environmental and radiochemical applications.^34–36 While O₂(¹Δ_g) photosynthesis and O₂ activation on transition metals' compounds are studied worldwide in many groups, the presented here catalytic chemical excitation of dioxygen up to O₂(¹Σ⁺_g) on non-transition metal centre, that is, as was proposed earlier, in reactions of aluminum-containing peroxides, represents an essential novelty.

However, the proposed earlier catalytic mechanism (Scheme 2) of t-BuOOH dissociation with the active O₂ generation cannot be considered as a proved mechanism because it is impossible to exclude high activation barriers for the specified reactions. It was not established earlier whether the assumed reactions are elementary ones and whether other short-lived and highly reactive intermediates containing aluminum dioxygen bond exist. Thus, the goal of the present study is to predict reaction mechanism in the (MeO)₃Al + 2MeOOH system modeling the (t-BuO)₃Al–2t-BuOOH system where the catalytic activation of dioxygen with the participation of a metal center was exhibited earlier. Here, we use quantum chemical methods to study structures, vibrational frequencies of intermediates and transition states and to predict reaction paths and sources of chemically excited and chemically activated O₂.

2. Computational details

The topology of the potential energy surface (PES) of the (MeO)₃Al + 2MeOOH system and the reaction mechanism are studied with the density functional theory (DFT) at the B3LYP/cc-pVTZ theory level.^37–41 The switch to the (MeO)₃Al + 2MeOOH model system obtained by substitution of the bulky tert-butyl groups by methyl groups makes it possible to study thoroughly various hypothetical reaction pathways without a considerable increase of computational time. Taking into account a conformational non-rigidity of molecules, the (MeO)₃Al conformation isomer characterized by the C_3h point symmetry group and two MeOOH (methylhydroperoxide) molecules were selected as the initial reagents.

To investigate the electronic structure of some intermediates that are primarily important for the generation of ¹O₂ and the determination of the electronic state of ¹O₂, state-of-the art quantum chemical methods were employed. On the model system, it becomes possible to use the coupled cluster methods (up to CCSDT(Q)) and the complete active space method (CASSCF).

2MeOOH → 2MeOH + ³O₂ (1)

2MeOOH → 2MeOH + O₂(¹Δ_g) (2)

2MeOOH → 2MeOH + O₂(b¹Σ⁺_g) (2a)

(MeO)₃Al·O₂ → (MeO)₃Al + O₂(¹Δ_g) (3)

(MeO)₃Al·O₂ → (MeO)₃Al + O₂(b¹Σ⁺_g) (3a)

(MeO)₃Al·O₂ → (MeO)₃Al + ³O₂ (4)

The energies of the reactions (1) and (2) were estimated at the CCSDT(Q,fc)/cc-pVTZ theory level using the focal point analysis based on the CCSD(T,fc)/cc-pVTZ and CCSDT(Q,fc)/cc-pVDZ calculations. The energy of the reaction (2a) was calculated on the basis of the estimated reaction energy (2) and the experimental energy of the O₂(b¹Σ⁺_g) ← O₂(a¹Δ_g) transition.

As will be shown in the present study, the (MeO)₃Al·O₂ intermediate is of especial importance for O₂ generation. Thus, to establish the electronic state of ¹O₂ eliminated in the (MeO)₃Al·O₂ decomposition and to reveal the possibility of chemical electron excitation, i.e., to distinguish between the reactions (3) and (3a), the PES profile corresponding to the reaction ((MeO)₃Al·O₂ → (MeO)₃Al + ¹O₂) was determined while the energy of the reaction was calculated using the CCSD(T)/6-31+G(2df,p)//B3LYP/cc-pVTZ, CCSD(T)/cc-pVTZ//B3LYP/cc-pVTZ composite approaches, and the CAS(14,10)/6-311G(d) theory level. The energies of the quasi-isolated system [(MeO)₃Al + ¹O₂] were obtained by relaxed scans at the CAS(14,10)/6-311G(d) theory level. These reaction paths correspond to the adiabatic reaction. In these calculations, the aluminum–adjacent oxygen interatomic distance in the Al–O₂ moiety was elongated to 1.5 Å above the equilibrium bond length in the (MeO)₃Al·O₂. Moreover, to distinguish between the reactions (3) and (3a) and, thus, to reveal an electronic term of O₂, the energy of the reaction (4) was calculated at the CCSD(T)/cc-pVTZ theory level while the energies of the asymptotes corresponding to the reactions (3) and (3a), i.e., the levels of products, were estimated on the basis of the CCSDT(Q)/cc-pVTZ data for the electronic states of O₂ and the coupled cluster data for the reaction (4). This procedure was applied since the calculations of (MeO)₃Al·O₂ with the CCSDT(Q) method are prohibitive time- and resource demanding.

To take into account the possibility of chemically activated O₂ generation (reaction (4)) and, thus, to predict non-adiabatic reaction pathways in the MeOOH decomposition of the MeOOH/(MeO)₃Al catalytic system, intersections between PESs were searched at the B3LYP/cc-pVTZ theory level.

The calculations were conducted with the Gaussian 03 software⁴² (for the B3LYP and the CASSCF methods), the GAMESS^43–47 and CFOUR programs⁴⁸ (for the CCSD(T) method). For the CCSDT(Q,fc)/cc-pVTZ energy estimations, the focal point analysis was carried out on the results obtained with the CFOUR and MRCC^49–51 suites of programs. Minimum energy crossing point (MECP) search was conducted with ORCA suite of programs.^52–54 The Moltran,⁵⁵ GaussView03,⁵⁶ and ChemCraft⁵⁷ molecular editors were used to analyze the results.

3. Results

3.1. Reaction in the (MeO)₃Al + MeOOH system

The association (MeO)₃Al + MeOOH → (MeO)₃Al⋯HOOMe occurs barrierlessly forming quasi-cyclic isomeric complexes 1 and 1a (Fig. 1 and 2) stabilized by the hydrogen bond (the distances between the O and H atoms are about 2.131 Å and 1.702 Å for 1 and 1a, respectively) and by the Al–O coordination bond (the distances between the Al and O atoms are about 2.031 Å and 1.991 Å for 1 and 1a, respectively). The corresponding reaction energies are −93.8 kJ mol⁻¹ and −83.4 kJ mol⁻¹. 1a isomerizes in the structure 1 with the activation energy of 63.5 kJ mol⁻¹. Complex 1 is rearranged via saddle point TS1 to structure 2 that is the complex of (MeO)₂AlOOMe and MeOH stabilized by the hydrogen and coordination bonds. For 1a, similar reaction leading to the monomethylperoxide was not determined despite of special efforts. The activation (E_a) and reaction (E_r) energies for 1 → 2 are 56.9 kJ mol⁻¹ and 31.1 kJ mol⁻¹, respectively. The dissociation of 2 occurs with no transition state with E_a of 77.9 kJ mol⁻¹; the energy level of isolated reactants, (MeO)₂AlOOMe (3), MeOH, and MeOOH, is located by 15.2 kJ mol⁻¹ higher than the initial reagents ((MeO)₃Al + 2MeOOH). For peroxide 3, there is an isomer 3a (Fig. 1 and 2), which differs in the formation of the intramolecular donor–acceptor bond between Al and O atoms stabilizing the three-membered quasi-cycle of AlOO. The relative energy (E_rel) of mono-η²-methyl peroxide 3a is −0.3 kJ mol⁻¹.

Fig. 1 Profile of the singlet PES for the reaction 2MeOOH + (MeO)₃Al → (MeO)Al(OOMe)₂ + 2MeOH obtained at the B3LYP/cc-pVTZ theory level. The dotted and solid lines indicate the minimum energy pathways. The relative energies are in kJ mol⁻¹.

Fig. 2 Structures calculated by full geometry optimization at the B3LYP/cc-pVTZ theory level for the reaction 2MeOOH + (MeO)₃Al → (MeO)Al(OOMe)₂ + 2MeOH. Bond lengths are in Å.

3.2. Elimination in the (MeO)₂AlOOMe + MeOOH system

Complex 2 (Fig. 2) is coordinatively unsaturated, which allows of the attachment of the MeOOH molecule from the side opposite to the MeOH molecule bonded in the donor–acceptor complex. Structure 4 formed without activation with E_r of −62.1 kJ mol⁻¹ decomposes with the activation energy of 48.5 kJ mol⁻¹ with no transition state. The released complex 5 is stabilized by the donor–acceptor and hydrogen bonds with a six-membered quasi-cycle. The isomerization of 5 in 6 (Fig. 1) occurs through transition state TS2 with the activation energy of (E_a) 133.4 kJ mol⁻¹ and with E_r of 48.8 kJ mol⁻¹. 5 is the complex of bis-η²-methyl peroxide 7 and MeOH (Fig. 2). The cleavage of the hydrogen bond stabilizing the complex requires the energy of 27.6 kJ mol⁻¹.

3.3. Substitution of the methoxy group in the (MeO)₂AlOOMe + MeOOH system

Contrary to the association with the formation of 4 (Fig. 1), the MeOOH molecule can attack from the side of MeOH bound in the complex. In this case, 4a (Fig. 2) is formed. Through the saddle point TS1b, it can rearrange to 4b with the activation energy (E_a) of 4.7 kJ mol⁻¹. The generation of the prereaction complexes of several structural types results in the nonuniqueness of the pathway for the reaction in the (MeO)₂AlOOMe + MeOOH + MeOH system (Fig. 2) with the generation of bis-η²-methylperoxide (structure 7, Fig. 2), i.e., with the cleavage of the peroxo bond in the O–OMe group and the concerted closing of AlOO rings with the two η²-methylperoxo moieties formation.

The reaction of the second MeOOH molecule with 3 or 3a or their complexes with the MeOH molecule can occur similarly to the interaction of the first MeOOH molecule with (MeO)₃Al, i.e., according to the nucleophilic substitution mechanism contrary to the rearrangement 5 → TS2 → 6 (Fig. 1) that includes the cleavage of the O–O bond in the MeOO group and successive elimination of MeOH. Another reaction pathway is determined by 4c which an isomer of 4. Complex 4c with the relative energy (E_rel = −125.5 kJ mol⁻¹) that is similar to E_rel of 4 (E_rel = −124.8 kJ mol⁻¹) eliminates MeOH. In this no transition state reaction with the activation energy of 63.5 kJ mol⁻¹ 8 is formed that is the isomer of 5. Structure 8 (Fig. 1 and 3) with a four-membered quasi-cycle consisting of the coordination and hydrogen bonds is 14.3 kJ mol⁻¹ less stable than 5 (Fig. 1 and 2) formed by the coordination of the MeOOH molecule which is involved in the six-membered quasi-cycle. The isomerization 8 → TS3 → 9 occurs with the activation energy (E_a) of 34.9 kJ mol⁻¹ and is a weakly exothermic reaction (E_r = −29.9 kJ mol⁻¹). Then, during the elimination of MeOH from 9, one or two peroxo groups may close in three-membered cycle or two cycles, respectively, with simultaneous formation the hydrogen-bonded complexes similar to 6.

Fig. 3 Structures calculated by full geometry optimization at the B3LYP/cc-pVTZ theory level for the reactions in the (MeO)₂AlOOMe + MeOOH + MeOH system. Bond lengths are in Å.

3.4. Reactions of isomeric mono- and bis-η²-methyl peroxides MeOAl(OOMe)₂

Let us dwell on the reactivity of 7 (Fig. 2). The intramolecular transfer of O atom from one η²-methyl peroxo group to another η²-methyl peroxo group with the formation of ozonide-1,3 (10) is a weakly exothermic (E_r = −40.5 kJ mol⁻¹) elementary reaction with high activation energy (E_a = 182.2 kJ mol⁻¹). The four-membered cycle in 10 (Fig. 4) is opened over the longest O–O bond (r_e(O–O) = 2.264 Å) via transition state TS5 with the activation energy of 37.4 kJ mol⁻¹. The reaction 10 → 11 (Fig. 4 and 5) is practically athermic (E_r = 2.9 kJ mol⁻¹). For the fully optimized structure 11, the O atom of the η¹-O₂ group is not located exactly above the Al atom, but it is slightly shifted towards one of the methoxy groups, the valence angle ∠OAlO is 78° in this case. 11 is a η¹-complex of the O₂ and (MeO)₃Al monomers, i.e., its monomolecular decay on the isolated monomers proceeds without saddle point as was exhibited by the relaxed scan at the B3LYP/cc-pVTZ theory level where smooth potential energy curve corresponding to the Al–OO bond cleavage was obtained. In other words, the association reaction O₂ + (MeO)₃Al → 11 is barrierless.

Fig. 4 Profile of the singlet PES corresponding to the reaction in the (MeO)Al(OOMe)₂ + 2MeOH system obtained at the B3LYP/cc-pVTZ theory level. The dotted and solid lines indicate the minimum energy pathways. The relative energies are in kJ mol⁻¹.

Fig. 5 Structures calculated by full geometry optimization at the B3LYP/cc-pVTZ theory level for the reaction of (MeO)₂AlOOMe. Bond lengths are in Å.

For 7 (Fig. 2), several isomers can be obtained by opening one or two three-membered quasi-cycles AlOO (Fig. 4). Mono-η²-methyl peroxides, except for conformation isomers formed by methyl group rotation, have conformation isomers 12 and 12a (Fig. 4) which are different in the dihedral angle ∠O–Al–O–O, where O–O is a peroxo group corresponding to the opened cycle. This angle is −11.6° for 12 and −179.0° for 12a (Fig. 4). The 12 and 12a conformers are characterized by a comparable energetic stability. Indeed, the relative energy (E_rel) of 12 and two MeOH monomers is 7.5 kJ mol⁻¹ and E_rel = 4.6 kJ mol⁻¹ was calculated for 12a and two MeOH monomers (Fig. 4). The two OOAl cycles (7 → 12b) are opened with E_r = 29.3 kJ mol⁻¹ (Fig. 4). In all the cases (7.4 kJ mol⁻¹ for 7 → 12 and 4.5 kJ mol⁻¹ for 7 → 12a), the activation energy is many times less than E_a of the reaction 7 → TS4 → 10 (Fig. 4). If two peroxo groups in MeOAl(OOMe)₂ are not closed in a cycle, the conformation isomer 12b with two MeOH monomers (Fig. 4) is 29.4 kJ mol⁻¹ less energetically stable than the initial reagents ((MeO)₃Al + 2MeOOH).

Mono-η²-methyl peroxide 12 (E_rel = 7.5 kJ mol⁻¹) is isomerized with E_r = 13.5 kJ mol⁻¹ to ozonide-1,3 (13) via transition state TS6 with the relative energy (E_rel) of 129.3 kJ mol⁻¹ (Fig. 3 and 4). Ozonide-1,3 (10) is more energetically favorable than 13. Indeed, E_r calculated for the reaction 13 → 10 is 61.4 kJ mol⁻¹; however, isomerization through TS7 (Fig. 4) with the methyl group shift is a reaction with a rather high barrier (E_a = 185.9 kJ mol⁻¹). This activation energy value is insignificantly higher than E_a = 182.0 kJ mol⁻¹ for the reaction pathway to 7 → 10.

12a is slightly different in reactivity from the conformer 12 (Fig. 5). In this case, the transfer of the O atom of the OO-η² group to the peroxo group and the increase of the oxygen chain length occurs with the activation energy of 159.2 kJ mol⁻¹ for TS6a, i.e., 12 is more reactive because E_a for 7 → 10 is 22.8 kJ mol⁻¹ more. The OOOCH₃ chain is stabilized in the form of the η²-donor–acceptor coordination complex 13a that is 41.4 kJ mol⁻¹ more energetically stable than 13. Ozonide-1,2 (structure 13a, Fig. 4 and 5) turns out to be 20 kJ mol⁻¹ less stable than ozonide-1,3 (10, Fig. 5). Contrary to the reaction 12 → 13, the isomerization 12a → 13a is an exothermic (E_r = 25.0 kJ mol⁻¹).

3.5. Reaction of mono-η²-methyl peroxide with MeOOH

Contrary to 5 and 8, where the coordination of the oxygen atoms of the OH and OOH groups to the Al atom and the formation of 6- and 4-membered quasi-cycles stabilize the structures, another structure can be predicted, namely, a mono-η²-methyl peroxo complex with the hydrogen bonded MeOOH. Indeed, 14 with MeOH (Fig. 6 and 7) is energetically more stable than the initial reagents (E_rel = −29.3 kJ mol⁻¹). The energy of the hydrogen bond in the complex 14 is 35.7 kJ mol⁻¹ as calculated at the B3LYP/cc-pVTZ theory level. The isomerization 14 → 15 (Fig. 6 and 7) occurs with E_a = 173.7 kJ mol⁻¹ through transition state TS8 (Fig. 7). Structure 15 is a complex of the MeOH molecule and (MeO)₂AlOOOMe. 15 is stabilized by the hydrogen bond which occurs between the middle O atom in the AlOOOCH₃ chain and HO-group of the MeOH molecule. The hydrogen bond can be formed both with the O atom of the OCH₃ group and with the O atom of the OAl group. The isomerization 15 → 15a (Fig. 6 and 7) is an exothermic process (E_r = −22.5 kJ mol⁻¹); the second reaction (15 → 15b) is athermic (E_r = −0.1 kJ mol⁻¹). When MeOH is eliminated from 15 and isomeric 15a or 15b, quasi-cycles are closed with the formation of 10 and 13a.

Fig. 6 Profile of the singlet PES for the reaction in the (MeO)₂AlOOMe⋯MeOOH system obtained at the B3LYP/cc-pVTZ theory level. The dotted and solid lines indicate the minimum energy pathways. The relative energies are in kJ mol⁻¹.

Fig. 7 Structures calculated by full geometry optimization at the B3LYP/cc-pVTZ theory level for the reaction of (MeO)₂AlOOMe with MeOOH. Bond lengths are in Å.

4. Discussion

4.1. Comparison of reaction pathways

The reaction pathways in the (MeO)₃Al + 2MeOOH system predicted in the quantum chemical study of the PES at the B3LYP/cc-pVTZ theory level can be divided into two groups: (1) with the formation of bis-methyl peroxides (bis-η²-, mono-η²- and acyclic (MeO)₂AlO₂Me) that are rearranged to isomeric ozonides (MeO)₂AlO₃Me; (2) with the attack of MeOOH onto the AlO₂Me three-membered ring of the mono-methyl peroxide, (MeO)₂AlO₂Me, forming the isomeric ozonides, (MeO)₂AlO₃Me. The ozonides are precursors for the reactive η¹-O₂-complex, (MeO)₃Al·O₂, that decomposes forming ¹O₂ and regenerating the catalyst, (MeO)₃Al. It is worth noting that these pathways have in common the formation of isomeric ozonides or their donor–acceptor complexes with MeOH and the formation of η²-methyl peroxides ((MeO)₂AlO₂Me or MeOAl(O₂Me)₂ depicted in Fig. 1). Therefore, mono-η²- and bis-η²-methyl peroxides are formed in different stages of different reaction pathways (Fig. 1 and 6). These η²-methyl peroxides differ from acyclic isomers in energetic stability. Indeed, closing one ring in the molecule stabilizes the structure by about 15 kJ mol⁻¹. Besides structure stabilization, closing the AlOO three-membered ring activates the O-atom of methyl peroxide. In this case, the O atom inserts in the Al–O bond of the another η²-methyl peroxo group, AlO₂Me.

However, the activation energy (182.0 kJ mol⁻¹) for the intramolecular transfer of the O atom in 7 with the formation of ozonide-1,3 (10) is by 48.6 kJ mol⁻¹ higher than the highest barrier for the net reaction 5 → 7 (Fig. 1) while the synchronous formation of two η²-groups (5 → 6, Fig. 1) is the reaction with the highest activation barrier in the PES region corresponding to the pathway of the successive substitution of two MeO in (MeO)₃Al (Fig. 1). Indeed, the activation energy (E_a = 133.4 kJ mol⁻¹) of the reaction 5 → 6 (Fig. 1) is more than two times higher than E_a = 56.9 kJ mol⁻¹ of the reaction 1 → 2 (Fig. 1) and the total activation energy of the stepwise reaction 4c → 9 (Fig. 1).

Another reaction path to ozonide to be compared is a path with the intermediate 9. Intermediate 9 (Fig. 1) is a precursor of conformers 12 and 12a (Fig. 4) that are formed from 9 after the elimination of the coordinative bonded MeOH and the closing of one AlO₂Me ring with the activation energy of 96.5 kJ mol⁻¹ with no transition state. The reaction 12 → 13 results in the formation of the OOO chain stabilized by including in the four-membered cycle AlO₃ with almost equivalent lengths of the corresponding bonds (the difference is not more than 0.0004 Å, Fig. 5). Structure 13 (Fig. 5) insignificantly differs in energy (0.1 kJ mol⁻¹ more stable) and in geometric parameters (the value of root mean square deviation for atomic positions is 0.039 Å) from a more symmetric structure that is characterized by the C_s symmetry point group. However, the symmetric structure has one (soft) imaginary vibrational frequency which breaks the symmetry. 13 (Fig. 4) is less energetically stable than 10 (E_r = −61.4 kJ mol⁻¹ for this rearrangement), the methyl group shift requires overcoming of the barrier of 185.9 kJ mol⁻¹ (Fig. 4). This E_a value is insignificantly higher than E_a (182.0 kJ mol⁻¹) of the rate-determining step for the minor pathway with TS4 (Fig. 4). While the barrier of 185.9 kJ mol⁻¹ for the two-step reaction 12 → 13 → 10 is comparable to the barrier height of 182.0 kJ mol⁻¹ in the single-step reaction 7 → TS4 → 10 determined relative to 7, the 12 → 13 reaction is endothermic and, thus, less thermodynamically favorable. Therefore, the pathway through TS4 is more probable than the two-step reaction with TS6 and TS7.

The ozonide-1,2 (structure 13a, Fig. 4 and 5) is formed with the activation energy (E_a) of 159.2 kJ mol⁻¹, which exceeds the activation energies in the reaction pathways of the successive (two-step) nucleophilic substitution of one methoxy group and the subsequent O–O bond cleavage closing the AlOOOMe ring. The activation barrier (7.8 kJ mol⁻¹) of the ozonide-1,2 isomerization, (13a) → ozonide-1,3 (10), is negligible in comparison with the activation energies calculated for 12a → TS6a → 13a, the isomerization 10 → TS5 → 11, and the ¹O₂ elimination from the (MeO)₃Al·O₂ complex (11) (Fig. 3). Therefore, the pathway 12a → TS6a → 13a → TS7a → 10 → TS5 → 11 (Fig. 3) is a more probable one among the three pathways of the ozonide synthesis. Indeed, the activation energy of 159.2 kJ mol⁻¹ for the rate-determining step (12a → TS6a → 13a), and the reaction pathway 7 → TS4 → 10 → TS5 → 11 → ¹O₂ + (MeO)₃Al with the rate-determining step 7 → TS4 → 10 for which E_a = 182.0 kJ mol⁻¹ is a minor channel (Fig. 4).

The reaction 14 → TS8 → 15 (Fig. 6) is a pathway to the ozonide 15 through the direct attack of MeOOH onto the mono-η²-methyl peroxo group with building the OOO chain. The activation energy (E_a) of the reaction 14 → 15 is 173.7 kJ mol⁻¹ (Fig. 6). This value is 14.5 kJ mol⁻¹ higher than E_a for the reaction pathway with TS6a, but it is 8.3 kJ mol⁻¹ less than E_a of the reaction through TS4, that is, the minor pathway with TS4 is less effective than this one with TS6a (Fig. 4).

Thus, the effectively competing reaction pathways to the ozonides and their complexes were predicted at the B3LYP/cc-pVTZ theory level. In each case, it was shown that the formation of a three-oxygen-atom chain (OOO) is the rate-determining step of the net reaction in the 2MeOOH + (MeO)₃Al catalytic system. It is worth noting that the reaction pathway including the isomerization of mono-η²-peroxo aluminum (12a) to ozonide-1,2 (13a) is characterized by the lowest barrier height (E_a = 159.2 kJ mol⁻¹, Fig. 4) and, thus, it is a major adiabatic channel.

4.2. On the possibility of the chemically excited O₂ formation. Whether it is O₂(¹Δ_g) or O₂(b¹Σ⁺_g)?

In order to establish the electronic configuration of O₂, i.e., the possibility of chemical electron excitation or chemical activation, in the decomposition of MeOOH in the MeOOH/(MeO)₃Al catalytic system which initiated the sequence of elementary reactions with the participation of the mono-η²-, bis-η²-methyl peroxo aluminum compounds, the ozonides, and (MeO)₃Al·O₂ (η¹-complex of O₂), the energies of the reactions (1) and (2) were calculated with high accuracy and the singlet PES region corresponding to the decomposition (MeO)₃Al·O₂ → (MeO)₃Al + ¹O₂ was scanned.

As was shown in the scrupulous study of oxygen allotropes in the form of molecules and complexes,⁵⁸ the energy of the transition O₂(a¹Δ_g) ← O₂(X³Σ⁻_g) is 99.5 ± 6.9 kJ mol⁻¹ as calculated with the CCSDT(Q) method that agrees well with the experimental value⁵⁹ of 94.7 kJ mol⁻¹. In this case, it is necessary to use the CCSDT(Q) method to calculate the energy of the reaction with O₂(¹Δ_g) formation, i.e., for the reaction (2). The energy of the reaction (2a) will be obtained if we know the high accurate energy estimations of the reactions (1) and (2) and also the experimental energy of O₂(b¹Σ⁺_g) ← O₂(X³Σ⁻_g).

In the present study, the energies of the reactions (1) and (2) were estimated at the CCSDT(Q,fc)/cc-pVTZ theory level using the focal point analysis based on the CCSD(T,fc)/cc-pVTZ and CCSDT(Q,fc)/cc-pVDZ calculations (ESI, Table 1S†). The energies of the reactions (1) and (2) are given as following −144.9 kJ mol⁻¹ and −45.4 kJ mol⁻¹ at the CCSDT(Q,fc)/cc-pVTZ theory level. The energy of the reaction (2a) was calculated to be 13.0 kJ mol⁻¹ taking into account the energy of the reaction (1) obtained here and the experimental value of the O₂(b¹Σ⁺_g) ← O₂(X³Σ⁻_g) transition energy. Thus, MeOOH is a high energetic molecule. Both pathways (2) and (2a) are energetically accessible because the decomposition of MeOOH with the O₂(¹Δ_g) elimination is weakly exothermic (E_r = −45.4 kJ mol⁻¹) while, in the generation of O₂(b¹Σ⁺_g), the decomposition reaction is weakly endothermic (E_r = 13.0 kJ mol⁻¹).

Both theory levels (composite and multi-reference) are in agreement with the results of the relaxed scan of the Al–O(O) bond in 11 at the B3LYP/cc-pVTZ theory level: (1) there is no maximum on the potential curve, i.e., the decomposition of 11 on the singlet PES occurs with no transition state; (2) the barrier height is 19.0 kJ mol⁻¹ (CCSD(T)//B3LYP with cc-pVTZ basis set), 34.7 kJ mol⁻¹ (CASSCF). It is worthwhile to note that no intruder states were detected during the orbital update for the CASSCF calculation because the potential curve for the relaxed scan was smooth. The optimized structure of (MeO)₃Al·O₂ was not distorted, i.e., it was similar to the structure 11 optimized with the B3LYP method, but the Al–O-bond was longer. No η¹-superoxo-η²-peroxo isomerism reported^60–65 for the compounds of transition metals was revealed for (MeO)₃Al·O₂.

Next, to distinguish between the reactions (3) and (3a), the energy of the reaction (4) was calculated with the composite CCSD(T,fc)/cc-pVTZ//B3LYP/cc-pVTZ approach: E_r is −104.6 (−115.1) kJ mol⁻¹, the value for B3LYP is given in parentheses for comparison. One can conclude that the electronic configuration of the ³O₂ molecule is correctly described by the CCSD(T) method (see the comparison of the results in ref. 58). Here, one can note that, for 11, the non-dynamic electron correlation is taken into account by the CCSD(T) method⁶⁶ as was revealed by the T₁ diagnostic (0.019) and the largest T₂ amplitude with absolute value of 0.15 calculated at the CCSD(T)/6-31+G(2df,p) theory level. The use of the flexible cc-pVTZ basis set for coupled cluster method provides the result which is nearly free from the basis set incompleteness error. Thus, one can believe that the E_r value of −104.6 kJ mol⁻¹ is a reliable estimation.

However, for reactions with the (MeO)₃Al·O₂ intermediate, due to the especial time and resource consumption of calculations with the CCSDT(Q) method even with the moderate cc-pVDZ basis set, it is impossible to use this method for estimations with the focal point analysis. The energy of the intermediate 11 was calculated at the CCSD(T)/cc-pVTZ theory level as well as the energy of the reaction (4). On the basis of these results, the relative energy of the (MeO)₃Al + O₂(¹Δ_g) level is estimated with the CCSDT(Q)/cc-pVTZ approach for the O₂(¹Δ_g)/O₂(X³Σ⁻_g) system. The relative energy of the (MeO)₃Al + O₂(b¹Σ⁺_g) level is obtained on the basis of the experimental energy of the excitation O₂(b¹Σ⁺_g) ← O₂(a¹Δ_g) and the calculated energy of the reaction (3).

One can believe that the chosen combination of the CCSD(T) and CCSDT(Q) methods provides reliable results competing with the experimental data. Thus, the calculated reaction energies for (MeO)₃Al·O₂ and experimentally obtained values for the excitation O₂(b¹Σ⁺_g) ← O₂(¹Δ_g) will be assembled on one energetic diagram (Fig. 8) and analyzed jointly. The level of isolated (MeO)₃Al + O₂(X³Σ⁻_g) is selected as the initial point.

Fig. 8 Energetic levels of (MeO)₃Al·O₂ (the structure 11) and different asymptotics, i.e., (MeO)₃Al + O₂(³Σ⁻_g), (MeO)₃Al + O₂(¹Δ_g), (MeO)₃Al + O₂(¹Σ⁺_g), and quasi-isolated [(MeO)₃Al + ¹O₂]. The (MeO)₃Al·O₂ and (MeO)₃Al + O₂(³Σ⁻_g) levels were calculated with the CCSD(T)/cc-pVTZ//B3LYP/cc-pVTZ composite method; the energies of the levels (MeO)₃Al + O₂(³Σ⁻_g) and (MeO)₃Al + O₂(¹Δ_g) were estimated at the CCSDT(Q)/cc-pVTZ theory level, [(MeO)₃Al + ¹O₂] was obtained by the relaxed scan at the CAS(14,10)/6-311G(d) theory level. The relative energies are in kJ mol⁻¹. The dotted line indicates the relaxed scan. The arrows connect pair of energy levels calculated by the depicted computational procedure. The crossed arrows correspond to assuming reaction pathways which contradict with the obtained data.

The main features of the determined location of the levels are: (1) the energy of the (MeO)₃Al + O₂(¹Δ_g) system is lower than the level of (MeO)₃Al·O₂ (complex 11); (2) the quasi-isolated system [(MeO)₃Al + ¹O₂] is energetically higher than 11; (3) the energy of the [(MeO)₃Al + ¹O₂] coincides with the energy of (MeO)₃Al + O₂(b¹Σ⁺_g). Assuming that the (MeO)₃Al·O₂ dissociation corresponds to the (MeO)₃Al + O₂(¹Δ_g), it is necessary to conclude that if the dissociation proceeds without a transition state (Fig. 8), the potential energy curve of the relaxed PES will exhibit decrease of the energy or if a saddle point or a conical intersection is located in the reaction pathway (Fig. 8), the curve will show an increase with subsequent decrease of the energy. However, as was calculated with quantum chemical methods, the energy curve smoothly increases, that is, the features of the relaxed scan curve for the assuming reactions are totally opposite. Therefore, one can conclude that the dissociation asymptotics of (MeO)₃Al·O₂ corresponds to (MeO)₃Al + O₂(b¹Σ⁺_g) rather than to (MeO)₃Al + O₂(¹Δ_g).

The position of the η¹-(MeO)₃Al·OO level (Fig. 8) in the system of energy levels (MeO)₃Al + O₂(b¹Σ⁺_g), (MeO)₃Al + O₂(¹Δ_g), and (MeO)₃Al + O₂(X³Σ⁻_g) calculated with the CCSD(T) and CCSDT(Q) methods and the position of the quasi-isolated [(MeO)₃Al + ¹O₂] level determined by the CASSCF method is of crucial importance for the generation of O₂(b¹Σ⁺_g). The energy of chemical bonds of two MeOOH is transferred through η¹-(MeO)₃Al·OO into the energy of the excited O₂ eliminated in the adiabatic decomposition reaction. The key factor to implement the quite unusual non-photoinduced process, i.e., the chemical excitation O₂(b¹Σ⁺_g) ← O₂(X³Σ⁻_g), is the multi-pathway sequence of elementary reactions leading for the model system to (MeO)₂AlOOOMe and, next, to η¹-(MeO)₃Al·OO. The (MeO)₃Al·OO complex is the key intermediate for the O₂(b¹Σ⁺_g) generation. On the basis of the conducted energy analysis for the diagram (Fig. 8), the hypothesis about the elimination of O₂(b¹Σ⁺_g) in adiabatic reaction on the singlet PES and in the catalytic reaction of the MeOOH decomposition, where (MeO)₃Al is the catalyst, can be considered to be justified with quantum chemical methods.

To our knowledge, the established in the present study the electronically excited state of the dioxygen, O₂(b¹Σ⁺_g), eliminated in the catalytic reaction of the MeOOH decomposition (the model system) is the first report about the chemical excitation of O₂ to the so-called second singlet state, O₂(b¹Σ⁺_g). Here, it should be noted that the O₂(¹Δ_g) state is doubly degenerate. This phenomenon is new for the chemistry of peroxides and for catalysis on transition and non-transition metal centers. The catalytic oxidation of various substrates by the system t-BuOOH-transition metal compound^67–73 conducted in the atmospheric O₂, the activation of O₂(X³Σ⁻_g) and the formation of 1 : 1 adducts of catalyst with O₂(¹Δ_g) as well as the O₂(¹Δ_g) photoinduced elimination^74,75 have been shown previously. While the reactivity of O₂(¹Δ_g) has been studied for about 80 years,^76–80 the chemistry of O₂(b¹Σ⁺_g) is not essentially defined. Indeed, firstly, the lifetime of O₂(b¹Σ⁺_g) determined by the spin-allowed transition O₂(b¹Σ⁺_g) → O₂(¹Δ_g)^75,79,81,82 is considerably shorter than the lifetime of O₂(¹Δ_g), i.e., the quenching^79,83–86 is quantitative. Second, O₂(b¹Σ⁺_g) is quantitatively deactivated through the electron–vibrational (e–v) mechanism unambiguously determined in ref. 83, 84 and 87. Moreover, as was shown in the earlier studies,^83,84,87 the oxidation of substrates by O₂(b¹Σ⁺_g) with C–H bonds becomes evading in the presence of the highly effective pathway of quenching.

Probably, the η¹-(MeO)₃Al·OO complex can show reactivity which could mimic O₂(b¹Σ⁺_g), but in reaction pathway with bypass of O₂(b¹Σ⁺_g) quenching. The selective oxidation^12,13 of C–H bonds by the reactive forms of oxygen generated on the atoms of non-transition metals is a promising alternative to the developed catalytic systems on transition metals.⁸⁸

4.3. Non-adiabatic reactions in the MeOOH/(MeO)₃Al system and the possibility of the chemically activated O₂ formation

In order to find the possibility of “hot” O₂(³Σ⁻_g) generation in the strongly exothermic net reaction (1) and in the elimination reaction (4), which are not spin-conserving reactions, non-adiabatic reaction pathways, were searched and the triplet PES regions were studied with the B3LYP density functional. To determine electronic ground states for the located structures, the energies of the singlet–triplet vertical excitation of (MeO)₃Al and the intermediates were calculated. Then, the local minima and the energy degeneracy points (Minimum Energy Crossing Point, MECP) of the singlet and triplet PESs were searched.

The singlet spin coupling is the ground state for (MeO)₃Al since, as was found at the B3LYP/cc-pVTZ theory level, the energy of the singlet–triplet transition (T₁ ← S₀) is (564.5 kJ mol⁻¹). Thus, the asymptotics corresponding to the isolated monomers ³(MeO)₃Al + ³O₂ is not reachable from 11 with any remarkable thermodynamic probability. However, the singlet state is not the ground electronic state for 11 (Fig. 4), the triplet state is −79.0 kJ mol⁻¹ more stable energetically than the lowest lying singlet state of the structure 11. Structure 16 (ESI, Fig. 1S†) is obtained in the geometry optimization of 11 (Fig. 4) on the triplet PES. The singlet state is the ground state for mono-η²-, bis-η²-methyl peroxides, and ozonides as was predicted at the B3LYP/cc-pVTZ theory level.

The search for the degeneracy of the singlet and triplet PESs for 3a, 7, 10, and 11 (Fig. 2 and 4) revealed minimum energy crossing points and, thus, non-stationary points of PES which determine shape of PES and non-adiabatic pathways for the net catalytic reaction. Structures corresponding to the MECPs are denoted as 3a-MECP, 7-MECP, 10-MECP, and 11-MECP (Fig. 9). Their energies are given: 98.5, 91.5, 2.4, and 10.5 kJ mol⁻¹ at the B3LYP/cc-pVTZ theory level relative to 3a, 7, 10, and 11, respectively.

Fig. 9 Structures corresponding to the minimum energy crossing points for the singlet and triplet PESs optimized at the B3LYP/cc-pVTZ theory level. Bond lengths are in Å.

3a-MECP is a critical point which is not a stationary point of the singlet or triplet PESs for the AlOO cycle opening in 3a (Fig. 2) with the formation of structure 3b (Fig. 10). The structure 3b is 58.1 kJ mol⁻¹ less stable than 3a (the singlet ground state). The decomposition of 3b with the formation of O(³P) and (MeO)₃Al occurs with the reaction energy (E_r) of 50.5 kJ mol⁻¹. However, for the structure with two AlOO cycles, 7, the singlet–triplet spin recoupling does not lead to the cycle opening, i.e., in 7b (Fig. 10) with the energy of 75.8 kJ mol⁻¹ above the level of 7 (Fig. 1) was located. The O–O bond lengths optimized at the B3LYP/cc-pVTZ theory level are 2.301 Å and 1.501 Å while in 7 the bonds are of 2.020 Å and 1.506 Å. Spin conversions via 10-MECP (E_a,T←S = 2.4 kJ mol⁻¹) and 11-MECP (E_a,T←S = 10.5 kJ mol⁻¹) result in the strongly exothermic decomposition reaction 11 → (MeO)₃Al + ³O₂ (E_r = −119.8 kJ mol⁻¹). The stabilization of weakly bound complex 16 (ESI, Fig. 1S†) is unlikely under such conditions and the reaction leads to the elimination of the (MeO)₃Al and monomers. The asterisk denotes here and below a dioxygen molecule with the excess in energy. In this case, the O₂ molecule is in the ground electronic state, but it is also vibrationally and/or translationally excited, i.e., can show higher or probably uncommon reactivity as compared to O₂ thermalized under the (mild) conditions of the synthesis.^12,13,89 Moreover, these non-adiabatic reactions show the possibility of incomplete conversion of 11 in (MeO)₃Al and O₂(b¹Σ⁺_g); thus, in addition to physical quenching in O₂(b¹Σ⁺_g) → O₂(¹Δ_g) and O₂(¹Δ_g) → O₂(X³Σ⁻_g), it is a reaction pathway without the generation of ¹O₂. It agrees with the experimentally determined¹³ incomplete (about 50%) conversion of dioxygen in the reaction with singlet dioxygen trap, 9,10-methylanthracene, in weakly quenching solvent. The probability of low-energy singlet–triplet transition (the activation energies E_a,T←S(10-MECP) = 2.4 kJ mol⁻¹ and E_a,T←S(11-MECP) = 10.5 kJ mol⁻¹, and E_a(11 → (MeO)₃Al + O₂(b¹Σ⁺_g)) = 53.3 kJ mol⁻¹ are given for comparison) makes it possible to assume the possibility of the process controlling by selection of the solvent, i.e., stimulation of the adiabatic decomposition reaction and, thus, increase of the O₂(b¹Σ⁺_g) yield.

Fig. 10 Structures for triplet PES optimized at the B3LYP/cc-pVTZ theory level. Bond lengths are in Å.

The reaction η¹-(MeO)₃Al·OO → (MeO)₃Al + O₂ is a neither single-pathway nor product spin-pure process, so it is possible that the reactivity of O₂(b¹Σ⁺_g)/O₂(¹Δ_g) (the latter one is the product of the spin-allowed quenching of O₂(b¹Σ⁺_g)) and chemically activated (translationally–vibrationally–rotationally hot) ³O₂, which is the product of the spin-forbidden reaction, may be observed for the t-BuOOH/(t-BuO)₃Al system of experimental interest.

5. Conclusion

On the basis of the quantum chemical study of the MeOOH/(MeO)₃Al model system, it was exhibited for the first time that O₂ is generated chemically excited. One can conclude that the excited dioxygen, O₂(b¹Σ⁺_g), is produced in the liquid phase in the catalytic decomposition of t-BuOOH. O₂(b¹Σ⁺_g) is formed as a result of the successive reactions of aluminum tert-butoxide with tert-butyl hydroperoxide at the mole ratio of 1 : 2. In the first step, the aluminum-containing peroxide (t-BuO)₂AlOOt-Bu for the experimental system or (MeO)₂AlOOMe for the model system is produced. The latter one reacts with the second MeOOH molecule both through the pathway of the substitution of the methoxy group (main reaction pathway) and through the six-membered transition state (minor pathway) with the formation of metastable metal-containing mono-η²- and bis-η²-methyl peroxides and the elimination of the MeOH molecule. The formation of the isomeric ozonides (MeO)₂AlO₃Me is the rate-determining step. For the main reaction pathway, the activation energy (E_a) of the intermolecular O atom transfer was calculated to be 159.2 kJ mol⁻¹. From ozonide-1,3, (MeO)₂AlOOOMe, the corresponding metal dioxygen intermediate, η¹-(MeO)₃Al·O₂, is generated in the decomposition reaction with the elimination of dioxygen in the electron-excited state, O₂(¹Σ_g). The electronic state of O₂ was established by the analysis of the energy diagram calculated for the (MeO)₃Al/O₂ system at the CCSD(T,fc)/cc-pVTZ, CCSDT(Q)/cc-pVTZ, and CAS(14,10)/6-311G(d) theory levels. For the decomposition of η¹-(MeO)₃Al·O₂, a non-adiabatic pathway, that includes the singlet–triplet transition and leads to the generation of chemically activated ³O₂, was exhibited.

The conducted simulation makes it possible to propose that the catalytic decomposition of t-BuOOH leads to a complicated oxidizer system being the source of the reactive oxygen species: mono-η²-, bis-η²-peroxides, isomeric ozonides, η¹-peroxide, chemically activated and chemically excited (O₂(b¹Σ⁺_g) and O₂(¹Δ_g)) dioxygen. O₂(¹Δ_g) is the product of the decay of the higher excited state. O₂(b¹Σ⁺_g) is characterized by the significantly shorter lifetime than O₂(¹Δ_g). The latter is the product of the spin-allowed transition decomposition of O₂(b¹Σ⁺_g). One can believe that quenching through the e–v mechanism for O₂(b¹Σ⁺_g) can create a pathway to exhibit an uncommon reactivity of O₂(¹Δ_g) which results from the chemical activation of the substrate by energy excess released in the O₂(¹Δ_g) ← O₂(b¹Σ⁺_g) relaxation. The formation of the (t-BuO)₃Al·O₂ complex analogous to (MeO)₃Al·O₂ predicted here on the basis of quantum chemical study can determine the reactivity of chemically immobilized O₂(b¹Σ⁺_g). Both pathways make the new chemistry of ¹O₂ possible, that is, the oxidizing activity results directly or indirectly from the reactivity of O₂(b¹Σ⁺_g). Thus, it may determine important enhancement to famous Fenton and Haber–Weiss chemistry.

Acknowledgements

A. I. P. is thankful to SFU Super - computer Centre for generous donation of computer time. O. B. G. is thankful for the partial support of this study provided by a grant under agreement between N.I. Lobachevsky State University of Nizhny Novgorod and the Ministry of Education and Science of the Russian Federation.

References

1. K. Weissermel and H.-J. Arpe, in Industrial Organic Chemistry, Wiley-VCH Verlag GmbH, Weinheim, Germany, 2008, pp. 193–215.
2. J. Lewiński, J. Zachara, P. Goś, E. Grabska, T. Kopeć, I. Madura, W. Marciniak and I. Prowotorow, Chem.–Eur. J., 2000, 6, 3215–3227.
3. J. Lewiński and A. H. Wheatley, in Modern Organoaluminum Reagents, ed. S. Woodward and S. Dagorne, Springer, Berlin-Heidelberg, Germany, 2013, vol. 41, ch. 55, pp. 1–58.
4. G. A. Tolstikov, S. K. Minsker, R. G. Bulgakov, G. Y. Maistrenko, A. V. Kuchin, U. M. Dzhemilev and V. P. Kazakov, Bull. Acad. Sci. USSR, Div. Chem. Sci., 1979, 28, 2471.
5. V. A. Dodonov, L. A. Stepovik and S. M. Safonova, Russ. J. Gen. Chem., 1990, 60, 1839–1848.
6. V. A. Dodonov, L. A. Stepovik, S. M. Safonova and V. A. Zinchenkov, Russ. J. Gen. Chem., 1990, 60, 1125–1136.
7. V. A. Dodonov, Bulletin of Lobachevsky Nyzhniy Novgorod State University, 2012, 2, 81–101.
8. L. P. Stepovik, I. M. Martinova, V. A. Dodonov and V. K. Cherkasov, Russ. Chem. Bull., 2002, 51, 638–644.
9. V. A. Dodonov, L. P. Stepovik, A. C. Soskova and E. A. Zaburdaeva, Russ. J. Gen. Chem., 1994, 6, 1715–1721.
10. L. P. Stepovik, V. A. Dodonov and E. A. Zaburdaeva, Russ. J. Gen. Chem., 1997, 67, 116–120.
11. V. A. Dodonov, E. A. Zaburdaeva, N. B. Dolganova, L. P. Stepovik and T. I. Zinov'eva, Russ. J. Gen. Chem., 1997, 67, 988–992.
12. V. A. Dodonov, E. A. Zaburdaeva and L. P. Stepovik, Russ. Chem. Bull., 2004, 53, 1729–1734.
13. E. A. Zaburdaeva, V. A. Dodonov and L. P. Stepovik, J. Organomet. Chem., 2007, 692, 1265–1268.
14. E. A. Zaburdaeva and V. A. Dodonov, Russ. Chem. Bull., 2011, 1, 177–179.
15. F. Naso, M. A. M. Capozzi, A. Bottoni, M. Calvaresi, V. Bertolasi, F. Capitelli and C. Cardellicchio, Chem.–Eur. J., 2009, 15, 13417–13426.
16. M. A. M. Capozzi, C. Centrone, G. Fracchiolla, F. Naso and C. Cardellicchio, Eur. J. Org. Chem., 2011, 4327–4334.
17. G. E. O'Mahony, A. Ford and A. R. Maguire, J. Sulfur Chem., 2012, 34, 301–341.
18. C. Bolm, O. Beckmann, T. Kühn, C. Palazzi, W. Adam, P. B. Rao and C. R. Saha-Möller, Tetrahedron: Asymmetry, 2001, 12, 2441–2446.
19. C. Bolm, O. Beckmann and C. Palazzi, Can. J. Chem., 2001, 79, 1593–1597.
20. T. Yamaguchi, K. Matsumoto, B. Saito and T. Katsuki, Angew. Chem., Int. Ed., 2007, 46, 4729–4731.
21. K. Matsumoto, T. Yamaguchi and T. Katsuki, Chem. Commun., 2008, 1704–1706.
22. C. S. Foote, Acc. Chem. Res., 1968, 1, 104–110.
23. F. Wilkinson, W. P. Helman and A. B. Ross, J. Phys. Chem. Ref. Data, 1993, 22, 113–262.
24. M. C. DeRosa and R. J. Crutchley, Coord. Chem. Rev., 2002, 233–234, 351–371.
25. W. Adam, D. V. Kazakov and V. P. Kazakov, Chem. Rev., 2005, 105, 3371–3387.
26. S. Miyamoto, G. E. Ronsein, F. M. Prado, M. Uemi, T. C. Corrêa, I. N. Toma, A. Bertolucci, M. C. B. Oliveira, F. D. Motta, M. H. G. Medeiros and P. D. Mascio, IUBMB Life, 2007, 59, 322–331.
27. A. A. Frimer, Singlet oxygen, CRC Press, Boca Raton, 1985.
28. H. H. Wasserman and R. W. Murray, Singlet oxygen, Academic Press, New York, 1979.
29. B. Halliwell, Am. J. Med., 1991, 91, S14–S22.
30. B. Halliwell and J. Gutteridge, in Free Radicals in Biology and Medicine, Clarendon Press, Oxford, England, United Kingdom, 1985, pp. 279–315.
31. A. Bast, G. R. M. M. Haenen and C. J. A. Doelman, Am. J. Med., 1991, 91, S2–S13.
32. I. E. Kochevar and R. W. Redmond, in Methods in Enzymology, ed. L. Packer and H. Sies, Academic Press, 2000, vol. 319, pp. 20–28.
33. C. Pierlot, J.-M. Aubry, K. Briviba, H. Sies and P. D. Mascio, in Methods in Enzymology, ed. L. Packer and H. Sies, Academic Press, New York, 2000, vol. 319, pp. 3–20.
34. F. Munoz, E. Mvula, S. E. Braslavsky and C. von Sonntag, J. Chem. Soc., Perkin Trans. 2, 2001, 1109–1116.
35. G. Merényi, J. Lind, S. Naumov and C. von Sonntag, Chem.–Eur. J., 2010, 16, 1372–1377.
36. S. Naumov, G. Mark, A. Jarocki and C. von Sonntag, Ozone: Sci. Eng., 2010, 32, 430–434.
37. R. H. Hertwig and W. Koch, Chem. Phys. Lett., 1997, 268, 345–351.
38. P. J. Stephens, F. J. Devlin, C. F. Chabalowski and M. J. Frisch, J. Phys. Chem., 1994, 98, 11623–11627.
39. A. D. Becke, J. Chem. Phys., 1993, 98, 5648–5652.
40. T. H. Dunning Jr, J. Chem. Phys., 1989, 90, 1007–1023.
41. D. E. Woon and T. H. Dunning Jr, J. Chem. Phys., 1993, 98, 1358–1371.
42. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery Jr, R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill, B. Johnson, W. Chen, M. W. Wong, J. L. Andres, C. Gonzalez, M. Head-Gordon, E. S. Replogle and J. A. Pople, Gaussian 03, Version D.02, Gaussian, Inc, Wallingford, CT, 2007.
43. M. W. Schmidt, K. K. Baldridge, J. A. Boatz, S. T. Elbert, M. S. Gordon, J. H. Jensen, S. Koseki, N. Matsunaga, K. A. Nguyen, S. Su, T. L. Windus, M. Dupuis and J. A. Montgomery, J. Comput. Chem., 1993, 14, 1347–1363.
44. B. Bolding and K. Baldridge, Comput. Phys. Commun., 2000, 128, 55–66.
45. M. S. Gordon and M. W. Schmidt, in Theory and Applications of Computational Chemistry, ed. C. E. Dykstra, G. Frenking, K. S. Kim and G. E. Scuseria, Elsevier, Amsterdam, The Netherlands, 2005, pp. 1167–1189.
46. J. L. Bentz, R. M. Olson, M. S. Gordon, M. W. Schmidt and R. A. Kendall, Comput. Phys. Commun., 2007, 176, 589–600.
47. P. Piecuch, S. A. Kucharski, K. Kowalski and M. Musiał, Comput. Phys. Commun., 2002, 149, 71–96.
48. J. F. Stanton, J. Gauss, M. E. Harding and P. G. Szalay, CFOUR (Coupled-Cluster techniques for Computational Chemistry) with contributions from A. A. Auer, R. J. Bartlett, U. Benedikt, C. Berger, D. E. Bernholdt, Y. J. Bomble, L. Cheng, O. Christiansen, M. Heckert, O. Heun, C. Huber, T.-C. Jagau, D. Jonsson, J. Jusélius, K. Klein, W. J. Lauderdale, D. A. Matthews, T. Metzroth, D. P. O'Neill, D. R. Price, E. Prochnow, K. Ruud, F. Schiffmann, W. Schwalbach, S. Stopkowicz, A. Tajti, J. Vázquez, F. Wang, J. D. Watts and the integral packages MOLECULE (J. Almlöf and P. R. Taylor) PROPS (P. R. Taylor), ABACUS (T. Helgaker, H. J. Aa. Jensen, P. Jørgensen and J. Olsen) and ECP routines by A. V. Mitin and C. van Wüllen, URL: http://slater.chemie.uni-mainz.de/cfour/index.php?n=Main.HomePage, accessed August 26th, 2016.
49. M. Kállay and P. R. Surján, J. Chem. Phys., 2001, 115, 2945–2954.
50. M. Kállay and J. Gauss, J. Chem. Phys., 2008, 129, 144101.
51. M. Kállay, Z. Rolik, J. Csontos, I. Ladjánszki, L. Szegedy, B. Ladóczki, G. Samu, K. Petrov, M. Farkas, P. Nagy, D. Mester and B. Hégely, Mrcc, a quantum chemical program suite, URL: http://www.mrcc.hu, accessed August 26th, 2016.
52. J. N. Harvey, M. Aschi, H. Schwarz and W. Koch, Theor. Chem. Acc., 1998, 99, 95–99.
53. F. Neese, Wiley Interdiscip. Rev.: Comput. Mol. Sci., 2012, 2, 73–78.
54. F. Neese, ORCA-An Ab Initio, DFT and Semiempirical electronic structure package, version 3.0.3, Max-Planck-Institute for Chemical EnergyConversion Stiftstr, Mülheim a. d. Ruhr, Germany, 34-36, 45470, accessed August 10th, 2016.
55. S. K. Ignatov, MOLTRAN: a program for molecular visualization and thermodynamic calculations, N. I. Lobachevsky State University of Nizhny Novgorod, Nizhny Novgorod, 2009, URL: http://www.unn.ru/chem/moltran/, accessed August 26th, 2016.
56. R. Dennington Roy II, K. Todd, M. John, E. Ken, H. W. Lee and G. Ray, GaussView, Version 03, Semichem Inc, Shawnee Mission, KS, 2003.
57. G. Zhurko, ChemCraft, Version 1.7 (build 365), URL: http://www.chemcraftprog.com, accessed July 20th, 2016.
58. O. B. Gadzhiev, S. K. Ignatov, M. Y. Kulikov, A. M. Feigin, A. G. Razuvaev, P. G. Sennikov and O. Schrems, J. Chem. Theory Comput., 2012, 9, 247–262.
59. K. P. Huber and G. Herzberg, Molecular Spectra and Molecular Structure IV. Constants of Diatomic Molecules, Van Nostrand Reinhold Co, New York, 1979.
60. Y. Gong, M. Zhou and L. Andrews, Chem. Rev., 2009, 109, 6765–6808.
61. C. J. Cramer and W. B. Tolman, Acc. Chem. Res., 2007, 40, 601–608.
62. D. M. Wagnerová and K. Lang, Coord. Chem. Rev., 2011, 255, 2904–2911.
63. A. Yokoyama, J. E. Han, J. Cho, M. Kubo, T. Ogura, M. A. Siegler, K. D. Karlin and W. Nam, J. Am. Chem. Soc., 2012, 134, 15269–15272.
64. J. Cho, H. Y. Kang, L. V. Liu, R. Sarangi, E. I. Solomon and W. Nam, Chem. Sci., 2013, 4, 1502–1508.
65. W. Nam, Acc. Chem. Res., 2007, 40, 522–531.
66. D. G. Liakos and F. Neese, J. Chem. Theory Comput., 2011, 7, 1511–1523.
67. T. Matsushita, D. T. Sawyer and A. Sobkowiak, J. Mol. Catal. A: Chem., 1999, 137, 127–133.
68. A. Sobkowiak, A. Qui, X. Liu, A. Llobet and D. T. Sawyer, J. Am. Chem. Soc., 1993, 115, 609–614.
69. D. T. Sawyer, A. Sobkowiak and T. Matsushita, Acc. Chem. Res., 1996, 29, 409–416.
70. D. A. Pantazis and J. E. McGrady, Inorg. Chem., 2003, 42, 7734–7736.
71. D. T. Sawyer, C. Sheu, H.-C. Tung and A. Sobkowiak, in Studies in Surface Science and Catalysis, ed. L. I. Simándi, Elsevier, Amsterdam, The Netherlands, 1991, vol. 66, pp. 285–296.
72. C. Kang, C. Redman, V. Cepak and D. T. Sawyer, Bioorg. Med. Chem., 1993, 1, 125–140.
73. M. Drees, S. A. Hauser, M. Cokoja and F. E. Kühn, J. Organomet. Chem., 2013, 748, 36–45.
74. A. A. Abdel-Shafi, M. D. Ward and R. Schmidt, Dalton Trans., 2007, 2517–2527.
75. R. Schmidt, Photochem. Photobiol., 2006, 82, 1161–1177.
76. P. R. Ogilby, Photochem. Photobiol. Sci., 2010, 9, 1543–1560.
77. A. A. Frimer, in Peroxides, ed. S. Patai, John Wiley & Sons, Ltd, Chichester, England, United Kingdom, 1983, pp. 201–234.
78. N. Hoffmann, Chem. Rev., 2008, 108, 1052–1103.
79. C. Schweitzer and R. Schmidt, Chem. Rev., 2003, 103, 1685–1758.
80. P. R. Ogilby, Chem. Soc. Rev., 2010, 39, 3181–3209.
81. R. Schmidt, J. Phys. Chem. A, 2006, 110, 5990–5997.
82. D. Weldor, T. D. Poulsen, K. V. Mikkelsen and P. R. Ogilby, Photochem. Photobiol., 1999, 70, 369–379.
83. R. Schmidt, J. Photochem. Photobiol., A, 1994, 80, 1–5.
84. M. Bodesheim and R. Schmidt, J. Phys. Chem. A, 1997, 101, 5672–5677.
85. R. Schmidt and M. Bodesheim, J. Phys. Chem. A, 1998, 102, 4769–4774.
86. R. D. Scurlock, B. Wang and P. R. Ogilby, J. Am. Chem. Soc., 1996, 118, 388–392.
87. R. Schmidt and M. Bodesheim, J. Phys. Chem., 1994, 98, 2874–2876.
88. E. Roduner, W. Kaim, B. Sarkar, V. B. Urlacher, J. Pleiss, R. Gläser, W.-D. Einicke, G. A. Sprenger, U. Beifuß, E. Klemm, C. Liebner, H. Hieronymus, S.-F. Hsu, B. Plietker and S. Laschat, ChemCatChem, 2013, 5, 82–112.
89. V. A. Dodonov, E. A. Zaburdaeva and L. P. Stepovik, Russ. J. Gen. Chem., 2000, 70, 1482–1483.

---

Footnote

† Electronic supplementary information (ESI) available: Optimized Cartesian coordinates for reported intermediates, transition states, and minimum energy crossing points; the data for the focal point analysis; the reaction pathways for the peroxides dissociation, and additional comments on chemical reactivity of the catalytic system. See DOI: 10.1039/c6ra21471a

---