The oxidative cleavage of C=C double bonds is of major importance in both the organic synthesis and industrial production of fine chemicals. It represents an important transformation of alkenes to carbonyl compounds. In general, the reaction is performed using ozone [1], stoichiometric oxidants such as OsO4 and NaIO4 or transition metal catalysts in combination with oxidants such as NaOCl, H2O2 [2], and tert-BuOOH [3]. However, most of the reported oxidation procedures use toxic, hazardous and expensive reagents and produce large amounts of waste. Therefore, the development of green methods for oxidative cleavage has received considerable attention. Recently, progress has been made in this area through the application of oxygen as an environmentally friendly, cheap and readily available oxidant [4]. The aerobic oxidative cleavage of α-aryl-substituted alkenes has been successfully performed by photoirradiation or in the presence of metal- and organocatalysts.

When the oxidation of styrenes was performed under UV irradiation in water solvent, high selectivities (ca. 90%) but low yields of the respective aldehydes (ca. 10–15%) were achieved [5]. Alternatively, a high yield of benzoic acid, up to 81%, was obtained when the aerobic photooxidation of β-methylstyrene was carried out in ethyl acetate solvent in the presence of catalytic amounts of CBr4 [6]. Recently, copper(II) chloride has been reported as a metal catalyst for the aerobic cleavage of aromatic alkenes. Reactions were performed in mixed THF/H2O (9:1) solvent under the pressure of oxygen (0.4 MPa), in which the in situ-formed 2-hydroperoxytetrahydrofuran was proposed to be the true oxidant [7]. For example, acetophenone was obtained from α-methylstyrene in a yield of 85% under these conditions (60 °C, 5 h). Aerobic cleavage of styrenes were carried out in presence of 2,2′-azobis(isobutyronitrile) (AIBN), tert-butyl nitrite (TBN), bis(4-methoxyphenyl)disulfide (DS) and N–hydroxyphthalimide (NHPI). For example, acetophenone was obtained from α-methylstyrene in a yield of 80–90% when AIBN (25 mol%) in nitromethane (0.1 MPa, 60 °C, 12 h) [8], TBN (2 mol%) in compressed CO2 (13 MPa, 80 °C, 24 h) [9], DS (5 mol%) in acetonitrile (0.1 MPa, 25 °C, 16 h, LED lamp) [10] or NHPI (10 mol%) in N,N-dimethylacetamide (0.1 MPa, 80 °C, 24 h) [11] was used.

NHPI is known to catalyze various oxidation reactions through a free-radical mechanism [12]. This catalyst has attracted considerable interest because it is non-toxic, can be easily prepared from phthalic anhydride and hydroxylamine, and demonstrates high activity even under mild conditions. The catalytic activity of NHPI results from its ability to be converted to the phthalimide-N-oxyl (PINO) radical. PINO formation is accelerated in the presence of transition metals, azo-compounds, peroxides and enzymes. Therefore, the combination of NHPI with such additives, mainly cobalt(II) salts, is typically used. Oxidation reactions carried out in the presence of NHPI often require polar solvent due to low solubility of NHPI in hydrocarbons. As solvent elimination is an important factor in green chemistry, to overcome this problem, some lipophilic derivatives of NHPI has been synthesized and applied in solvent-free oxidations of alkylaromatics [13,14,15]. In this paper, in contrast to previous studies, the catalytic aerobic cleavage of alkenes in solvent-free conditions has been performed. The catalytic activity of metal-free systems (NHPI, NHPI/AIBN, and NHPI/alkylammonium salts), metal-containing systems (NHPI in combination with Co(II), Mn(II) or Cu(II) salts) and lipophilic derivative of NHPI have been reported.


Materials and methods

α-Methylstyrene (AMS, Sigma-Aldrich) and 4-methyl-α-methylstyrene (4MeAMS, Sigma-Aldrich) were distilled under vacuum before oxidation. 2-Isopropenylnaphthalene (2-IPN, ICP Warsaw), 4-fluoro-α-methylstyrene (4FAMS, Aldrich), 1-dodecene (Acros Organics), cyclohexene (Acros Organics), acetophenone (AcPh, Alfa Aesar), N-hydroxyphthalimide (NHPI, Acros Organics), 2,2′-azobis(2-methylpropionitryl) (AIBN, Acros Organics), tetrabutylammonium bromide (TBAB, Merck), dimethyldioctadecylammonium bromide (DMDOAB, Sigma-Aldrich), cetylpyridinium chloride monohydrate (CPC, Aldrich), tetrabutylammonium hydrogensulfate (TBAHS, Merck), tetrabutylammonium acetate (TBAA, Sigma-Aldrich) and methyltrioctylammonium chloride (MTOAC; Aldrich) were used as received. The transition metal salts were obtained from Merck and used as received.

A mixture of cetylpyridinium hydrogensulfate/sulfate was obtained by treating an aqueous solution of CPC with sulfuric acid, followed by the evaporation of HCl and water. C16-NHPI was synthesized based on previously the described procedure [15]. 4-Methoxy-α-methylstyrene (4MeOAMS) was obtained by the reported two-step method [16] consisting of the alkaline cleavage of bisphenol A to phenol and 4-isopropenylphenol, followed by the alkylation of 4-isopropenylphenol using dimethyl sulfate. The product was purified by vacuum distillation.

4-Methoxy-α-methylstyrene: 1H NMR (400 MHz, acetone-d6) δ ppm 7.45–7.42 (dd, J = 2.2 Hz, 7.0 Hz, 2H), 6.89–6.87 (dd, J = 2.4 Hz, 9.2 Hz, 2H), 5.29–5.28 (dd, J = 0.8 Hz, 1.6 Hz, 1H), 4.97–4.96 (dd, J = 1.6 Hz, 2.8 Hz, 1H), 3.79 (s, 3H), 2.10–2.09 (dd, J = 0.8 Hz, 1.6 Hz, 3H); 13C NMR (100 MHz, acetone-d6) δ ppm 160.2, 143.4, 134.2, 127.3, 114.3, 110.7, 55.4, 21.9.

4-Hexadecyloxycarbonyl-N-hydroxyphthalimide: 1H NMR (300 MHz, DMSO-d6) δ ppm 11.02 (s, 1H), 8.35 (d, J = 7.7 Hz, 1H), 8.16 (s, 1H), 7.96 (d, J = 7.8 Hz, 1H), 4.31 (t, J = 6.3 Hz, 2H), 1.68–1.77 (m, 2H), 1.20–1.38 (m, 26H), 0.83 (t, J = 6.6 Hz, 3H); 13C NMR (75 MHz, DMSO-d6) δ ppm 164.9, 163.9, 135.9, 133.2, 130.1, 124.2, 123.3, 66.3, 32.0, 29.7, 29.4, 29.3, 28.7, 26.1, 22.8, 14.6; mp 110.7÷111.2 °C.

The qualitative analyses were performed by GC (Agilent Technologies 7890C) connected with an MS (Agilent Technologies 5975C) detector. The GC–MS was equipped with an HP 5MS column (30 m × 0.25 mm × 0.25 μm film). The GC quantitative analyses were performed on a Hewlett-Packard 5890 Series II gas chromatograph with a flame ionization detector (FID) using a (ZB-5HT) capillary column (30 m × 0.32 mm × 0.10 μm film) with helium as the carrier gas. The following conditions were used: column head pressure of 60 kPa, injection port temperature of 280 °C and detector temperature of 280 °C.

The 1H and 19F NMR spectra were recorded on an Agilent 400-NMR working at 400 MHz. Samples were prepared by mixing 50 μmol NHPI and 50 μmol TBAB or TBAF and dissolving the mixture in acetone-d6 or benzene-d6. For TBAF, a 1 M solution in THF was used. The THF was removed using a stream of N2, the NHPI was added, and the mixture was diluted with deuterated solvent.

The IR analyses were performed on an FT-IR Mettler-Toledo iC10 spectrometer equipped with an ATR probe. The solubility tests for the NHPI-TBAB complex were performed in small beaker. Benzene (5 g) and NHPI (0.10 g; 0.66 mmol) were added to the beaker in which the ATR probe was immersed. Then, TBAB was added in 15 portions (approximately 0.02 g; 0.06 mmol). The height of the peak corresponding to the carbonyl groups (1731 cm−1) of NHPI was recorded in real time.

General oxidation procedure

Oxidation was carried out in a 25 ml two-necked round-bottomed flask equipped with a magnetic stirrer and reflux condenser with an oxygen-filled balloon. The flask was filled with AMS (19.3–38.5 mmol); AcPh (0–17 mmol); NHPI (0–10 mol%) or C16-NHPI (0–1 mol%); and AIBN (0–0.1 mol%), transition metal salt (0–0.5 mol%) or ammonium salt (0–1 mol%) as the co-catalyst. The reaction apparatus was flushed with pure O2, sealed with a septum and immersed in an oil bath preheated to the desired temperature. The reaction mixture was stirred (750 rpm) at 60–90 °C for 24 h under 0.1 MPa O2. Samples were taken after the appropriate time, mixed with naphthalene or tert-butylbenzene as an internal GC standard and diluted with acetone.

Additional reactions were carried out using substrates besides AMS, such as 2-IPN (24 mmol), 4MeOAMS (14 mmol), 4MeAMS (14 mmol), 4FAMS (3.35 mmol), 1-dodecene (22.5 mmol) and cyclohexene (50 mmol).

Oxidation in bubble reactor: a 15 ml bubble reactor (jacked wide test tube connected to a thermostat with gas inlet at the bottom of test tube) was heated to 70 °C, filled with AMS (77 mmol); NHPI (1 mol%) and AIBN (0.1 mol%). The reactor was equipped with reflux condenser. The oxygen flow (2.5 l/h) was kept constant during process (10 h).

Results and discussion

Herein, the described studies were performed using α-methylstyrene (AMS) as the starting material. Its oxidation with pure dioxygen under the applied conditions led to the formation of the cleavage products acetophenone (AcPh) and formaldehyde. The formation of α-methylstyrene oxide (epoxide) as the main byproduct and a small amount of 2-phenylpropionic aldehyde (by epoxide isomerization [17]) was also confirmed by GC–MS analysis. Additional byproducts with high molecular weights that cannot be detected using GC may have also formed, e.g., polyperoxides.

α-Methylstyrene oxidation in the presence of NHPI and the NHPI/AIBN system

The reported NHPI-catalyzed reactions, including the oxidative cleavage of styrenes [11], are usually performed in polar solvent due to the low solubility of the catalyst (used in amounts of 10–20 mol%) in non-polar media. However, during oxidation reactions, more polar compounds are formed, and the solubility of NHPI in these mixtures is expected to increase. Therefore, in this paper, the NHPI-catalyzed aerobic cleavage of AMS without solvent has been studied. Reactions were performed in the presence of 0.1–5 mol% NHPI at temperatures of 60–90 °C under the atmospheric pressure of oxygen (Table 1; Fig. 1). For comparison, the oxidative cleavage of AMS was also performed in the polar solvents acetonitrile and N,N-dimethylacetamide (DMA).

Table 1 α-Methylstyrene oxidative cleavage with oxygen in the presence of NHPI
Fig. 1
figure 1

Oxidation of α-methylstyrene in the absence of solvent catalyzed by NHPI, C16-NHPI or NHPI + AIBN. Conditions: 38.5 mmol AMS, 1 mol% NHPI (or C16-NHPI), 0.1 mol% AIBN, 70 °C, 750 rpm, 0.1 MPa O2

The catalytic effect of NHPI was observed in the reactions performed at 70 and 90 °C. For example, when 1 mol% NHPI was used at 70 °C, the conversion of AMS doubled (Table 1, entry 7) compared to the blank reaction (Table 1, entry 4). Although this amount of NHPI dissolves quite fast (< 3 h) at 70 and 90 °C, it takes a long time to dissolve at 60 °C (approximately 7 h). Additionally, the presence of NHPI precipitate in the reaction mixture can increase the rate of radical termination and thus may be the reason why the catalytic effect was not observed at 60 °C. The results also showed that the selectivity for acetophenone decreases when the temperature increases, and at 60 and 70 °C, the lower AcPh selectivity leads to greater epoxide selectivity. At 90 °C, in addition to acetophenone and epoxide, heavy products, perhaps polyperoxides, appear (especially in the absence of NHPI).

The activity of NHPI was compared with its lipophilic derivative 4-hexadecyloxycarbonyl-N-hydroxyphthalimide (C16-NHPI) (Table 1, entry 9). The improved solubility of C16-NHPI resulted in a higher reaction rate and therefore higher AMS conversion. The effect was especially pronounced at the beginning of the reaction, where the rate of oxidation in the presence of C16-NHPI was significantly higher than that in the presence of NHPI (Fig. 1).

Interestingly, a significant increase in the reaction rate was observed when AIBN was applied in combination with NHPI (Table 1, entry 10; Fig. 1). AIBN decomposed under the applied conditions (half-life temperature T1/2 = 59 °C within 10 h [18]) to radicals that can abstract hydrogen from NHPI to generate the PINO radical.

The results obtained using the polar solvents MeCN and DMA (Table 1, entries 15, 16, 17, and 18) showed that to achieve high AMS conversion, NHPI should be added in relatively large amounts of 10–20 mol%. The formation of a complex between NHPI and the solvent may make the formation of the PINO radical more difficult [19]. In contrast to the previously reported results [11], we obtained a higher yield of acetophenone in MeCN than in DMA. This could be a result of the reaction temperature (70 °C). Previous studies on solvent effects were performed at 80 °C, and the increased vapor pressure of MeCN at this temperature could limit the mixture’s contact with oxygen. It was observed that the presence of MeCN positively influenced the selectivity for acetophenone.

The effect of transition metal ions on α-methylstyrene oxidation in the presence of NHPI

According to the literature, NHPI-catalyzed reactions are often co-catalyzed with transition metal ions that accelerate the formation of the PINO radical from NHPI and the decomposition of the initially formed peroxides [20]. However, in a previous report [11], for the NHPI-catalyzed cleavage of styrenes performed in polar solvent, the presence of Co(II), Mn(II), Cu(II) and Fe(II) salts had a neutral or negative effect. We assume that this could be a result of the relatively high amount of metal used in the study, i.e., 5 mol%. We have observed a similar negative effect when increasing the amount of Cu(II) from 1 to 10 mol% in the NHPI-catalyzed oxidation of 1-methoxy-4-(1-methylethyl)benzene [21]. Most likely, the known reaction of transition metal salts in the lower state with the peroxyl radicals formed in the system could lead to a decrease in the reaction rate [22]. Therefore, we decided to investigate the influence of transition metal salts in the range of 0.05–0.5 mol% on the oxidative cleavage of AMS catalyzed by NHPI in the absence of solvent. The obtained results demonstrated that Co, Mn and Cu salts notably accelerated AMS oxidation (Table 2; Fig. 2).

Table 2 α-Methylstyrene oxidative cleavage with oxygen in the presence of NHPI and co-catalyst
Fig. 2
figure 2

Oxidation of α-methylstyrene in absence of solvent catalyzed by NHPI + metal ions. Conditions: 38.5 mmol AMS, 1 mol% NHPI, 0.5 mol% Me2+, 70 °C, 750 rpm, 0.1 MPa O2

The conversion of AMS is much greater than that in the reaction without catalyst and approximately 2–3 times greater than that in the reaction with only NHPI as the catalyst. The activity of metal ions increased in the order of Cu1+/2+ < Mn2+/3+ < Co2+/3+, which is typical for reactions catalyzed by NHPI/metal salt systems and in accordance to the order of increasing redox potential. The highest substrate conversion was achieved using 0.5 mol% Co(acac)2 (Table 2, entry 4), although the highest selectivity for AcPh was found in the reaction using Mn(OAc)2 (Table 2, entry 9). The elimination of NHPI from this catalytic system sharply decreased the conversion of AMS and diminished the selectivity for AcPh (Table 2, entry 3). This confirms the crucial role of NHPI in this catalytic system. Indeed, transition metal ions accelerate the oxidative cleavage reaction, but they are not as efficient as the combination of NHPI and metal ions.

The effect of alkylammonium salts on α-methylstyrene oxidation in the presence of NHPI

To increase the solubility of NHPI in non-polar media such as AMS, additives that form lipophilic complexes with NHPI may be applied. In our study, ammonium salts that differed in carbon chain length and counterion (Table 3; Fig. 3) were selected. Combining 1 mol% NHPI and 10 mmol/l (0.125 mol%) ammonium salt led to a higher conversion compared with NHPI alone (Table 3, entries 1–8).

Table 3 α-Methylstyrene oxidative cleavage with oxygen in the presence of NHPI and alkylammonium salts
Fig. 3
figure 3

Oxidation of α-methylstyrene in the absence of solvent catalyzed by NHPI and different alkylammonium salts. Conditions: 38.5 mmol AMS, 1 mol% NHPI, 10 mmol/l ≈ 0.125 mol% alkylammonium salt, 70 °C, 750 rpm, 0.1 MPa O2. DMDOAB dimethyldioctadecylammonium bromide, TBAHS tetrabutylammonium hydrogensulfate, TBAB tetrabutylammonium bromide, TBAA tetrabutylammonium acetate, MTOAC methyltrioctylammonium chloride

The results indicate that the type of both cation and anion plays a crucial role in the catalytic performance. A comparison of reactions with the same ammonium cation but different anions shows that the conversion of AMS is greater when sulfate, hydrogen sulfate or acetate are used rather than chloride or bromide. This result may be related to the acidity of the effective acids, which can be formed from the listed anions to affect the interaction between NHPI and the ammonium salt. The acidity increases in the order of HOAc < HSO4  < H2SO4 < HCl < HBr. Acetic acid has the highest pKa value (approximately 4.7), which is the closest to the pKa of NHPI (approximately 7) [23]. In the reaction with bromides that differ in chain length of the alkylammonium cation, a higher conversion of AMS was achieved with the addition of lipophilic salts (Table 3, entries 2 and 8). According to previous investigations [24], it is very likely that formed complex is better soluble in AMS if alkyl chains are longer.

Based on the obtained conversions and yields, we chose TBAA and MTOAC for further research. The investigation of the influence of temperature indicated that the reaction conversion of AMS is generally greater at higher temperatures, but the highest selectivity for AcPh was achieved at 70 °C. These results are in accordance with results obtained in the presence of NHPI alone (Table 1). Tests with increased amounts of ammonium salts (40 and 80 mmol/l, which is approximately 0.5 and 1 mol%, respectively) led to a decrease in the conversion of the substrate and the selectivity for AcPh. A similar decrease in alkylammonium salt activity at high concentration was observed previously [25] in oxidation reactions as a result of micelle shape changes. GC analysis of the products of the reactions using alkylammonium salts indicated that the peak areas of α-methylstyrene epoxide were always notably smaller than the peak areas of the ketone. This suggests that the selectivity for epoxide was lower than that for ketone and that additional products undetectable on GC were formed, e.g., polyperoxides of AMS.

To test the solubility of NHPI in the presence of alkylammonium salts, an ATR probe (FT-IR spectrometer) was immersed in a suspension of NHPI in benzene, and the height of the peak corresponding to the carbonyl group (1731 cm−1) was measured during the addition of TBAB at room temperature (Fig. 4). After the introduction of more than about 80 mol% TBAB (vs NHPI), the peak height increased sufficiently. A clear solution was obtained after the introduction of approximately 150 mol% TBAB to NHPI (reaching 2 wt% NHPI in benzene).

Fig. 4
figure 4

Infrared peak height of carbonyl group (1731 cm−1) in NHPI molecule during addition of TBAB in portions (black dot represents clear homogenous solution)

The obtained results demonstrated that the mixture of NHPI and TBAB dissolves in non-polar hydrocarbon much better than NHPI alone. This indicates that interactions similar to those that occur in phase transfer catalysis are present. The interactions between the alkylammonium salt and NHPI were studied by Taha and Sasson [24]. By examining the UV absorbance, they suggested that a hydrogen bond complex is formed between NHPI’s hydroxyl group and the anion from the ammonium salt (Scheme 1).

Scheme 1
scheme 1

Interactions between NHPI molecule and alkylammonium salt

We have extended that study to gain a better understanding of this matter. 1H NMR spectroscopy revealed changes in the chemical shift of the proton in the NOH group in NHPI in the presence of ammonium salts. The NOH proton chemical shift in the TBAB-NHPI mixture has an increased value of approximately 1 ppm compared to pure NHPI. For the TBAF-NHPI complex, this peak has a chemical shift above 12 ppm. Additionally, we used 19F NMR to measure signals for the fluoride ion, and we found the characteristic peak for Bu4N+ FHF at − 144 ppm (see supplement). The results seam to confirm previous studies, as well as suggest that this proton can also be abstracted from NHPI to create the PINO anion, although the anion cannot directly serve as a catalyst for radical oxidation. The abstraction of a proton from NHPI by different anions depends on their strength as bases.

The effect of AcPh on α-methylstyrene oxidation in the presence of NHPI

We observed that NHPI suspension was present in the reaction mixture at the beginning of the reaction. Then, it dissolved to form a clear solution after a specified time, depending on the quantity of NHPI and the temperature. This suggests that an increase in the NHPI solubility relates to an increase in the polarity of the reaction mixture due to the formation of oxidized products. The formed AcPh is not oxidized under the applied conditions and thus can be used as a diluent in AMS oxidation (Table 4). Furthermore, the addition of AcPh does not complicate the separation of the reaction mixture, which is important from an industrial point of view.

Table 4 α-Methylstyrene oxidative cleavage with oxygen in the presence of NHPI and AcPh

As expected, in the studied reaction, NHPI dissolves much faster when AcPh is used as a diluent (AMS:AcPh = 3:2 v/v). It took approximately 10 min to obtain a transparent solution when 1 mol% NHPI was used. Even at 10 mol%, a transparent solution was formed, although it took few hours. Finally, the addition of AcPh to the reaction mixture containing 1 mol% NHPI sufficiently increased AMS conversion (Table 4, entry 3). The increase in catalyst amount from 1 to 10 mol% led to a further increase in substrate conversion but had a negative influence on the selectivity for AcPh.

Oxidation of different alkenes in the presence of the NHPI/AIBN or NHPI/Co(II) systems

Substrates such as 4MeAMS, 4MeOAMS, 4FAMS and 2-IPN were oxidized under solvent-free conditions using NHPI and AIBN or Co(acac)2 (Table 5). As expected, the main products were the respective ketones, and the byproducts were mostly the respective epoxides.

Table 5 Oxidative cleavage of alkenes with oxygen in the presence of NHPI and co-catalyst

Generally, higher conversions of unsaturated substrate were achieved in the reactions carried out using Co2+ salt rather than AIBN, but the selectivity for C=C bond cleavage products was higher in systems containing azo-initiator. The type of aryl group (phenyl or naphthyl) and the type of substituent at the para position of the phenyl group can influence the reactivity of alkenes, as well as the polarity of the reaction mixture, which is associated with NHPI solubility. For example, the oxidation of 2-isopropenylnaphthalene proceeded much slower than that of AMS (Table 5, entry 3). Perhaps this is due to the lower polarity of 2-acetonaphthone compared with that of AcPh and hence the reaction mixture during the oxidation of 2-IPN. The presence of methyl or fluorine substituents positively influences the oxidation rate in contrast to the presence of methoxyl groups. It was observed that for methoxy derivatives, more epoxide is generated, and hence the ketone selectivity is lower.

Oxidation of non-aromatic alkenes such as cyclohexene and 1-dodecene in the presence of NHPI/AIBN system did not lead to cleavage of double bond. In the case of 1-dodecene as substrate, only trace amounts of dodecenol was detected. When cyclohexene was oxidized, allylic hydrogen was abstracted. That led to formation of cyclohex-2-en-1-ol and cyclohex-2-en-1-one. Other identified products were cyclohexene oxide and trace amount of adipic dialdehyde. The results suggest that the presence of aromatic ring in the vinyl position is crucial for C=C bond effective cleavage. The aromatic ring increases electron density in C=C bond.

Oxidation of α-methylstyrene in bubble reactor

In the industry, large scale oxidation processes are carried out in flow reactors in which oxygen is supplied continuously. Therefore, in our work we decided to investigate the oxidation of AMS in a bubble reactor. Continuous flow of O2 was maintained during the reaction (2.5 l/h O2). We found that both conversion of AMS (60%) and selectivity of AcPh (78%) are greater in the process carried out in bubble reactor than in process carried out under pure O2 atmosphere performed under the same conditions (1 mol% NHPI, 0.1 mol% AIBN, 10 h, 70 °C), which were respectively 49% (conversion) and 47% (selectivity). This is probably due to better contact between substrate and oxygen, although we met some problems. The NHPI crystals disturb O2 barbotage (at the beginning of reaction) and formaldehyde undergoes polymerization in condenser attached to reactor.

Proposed mechanism of α-methylstyrene oxidation in the presence of NHPI

The previous studies demonstrated that the NHPI-catalyzed oxidative cleavage of AMS and other benzylic olefins with oxygen proceeds via a radical mechanism [11]. In the presence of NHPI, the PINO radical is first generated by the abstraction of a hydrogen atom. This can be accelerated by the application of an azo-initiator (such as AIBN) or transition metal ions [20]. Next, the PINO radical, which is a good electrophile, reacts with the C=C double bond of the benzylic olefin. The formed alkyl radical is trapped by O2 (a biradical molecule), which leads to the formation of a peroxyradical. The latter dissociates, the dioxetane molecule is formed and the PINO radical is regenerated. Dioxetanes are not stable and further decompose to form ketones and/or aldehydes. It seems that the peroxyradical intermediate can also participate in the formation of the main byproduct, the respective epoxide (Scheme 2). It is also possible that radicals from AIBN decompositions reacts with O2 and then adds to AMS to form intermediate that participates in formation of polyperoxides. The polyperoxide is known to decompose to AcPh and formaldehyde.

Scheme 2
scheme 2

Proposed mechanism for oxidative cleavage of α-methylstyrene catalyzed by NHPI

It is known that depending on the structure of the olefin, peroxyradicals can either abstract hydrogen or be added to the double bond. In AMS, practically only the addition reaction occurs [26]. When NHPI is present, peroxyradicals can also abstract hydrogen to form PINO and the respective hydroperoxide. The presence of metal salts accelerates PINO formation, as well as the decomposition of the formed peroxide compounds.


In summary, we proved that the oxidative cleavage of benzylic α-olefins can be effectively catalyzed by NHPI in the absence of solvent. Even low quantities of NHPI (1 mol%) sufficiently accelerate the reaction at low temperature. The reaction proceeds faster in the presence of a co-catalyst, such as metal ions or azo-initiator, in combination with NHPI. We confirmed the presence of interactions between NHPI and the ammonium quaternary salt by means of NMR. The alkylammonium salts help dissolve NHPI in non-polar media, and a higher conversion is achieved when they are applied in small amounts. Increased amounts of ammonium salt have a negative effect on the selectivity for ketone. The NHPI low solubility problem in non-polar AMS is greatly diminished by application of the product (AcPh) as a diluent. The latter solution seems to be most promising and is worthy of further research.