Lu-Jian
Zhou
a,
Xiao-Hui
Liu
*a,
Han-Wen
Zhang
a,
Can
Xue
a,
Han-Kang
Zhong
a and
Xian-Tai
Zhou
*ab
aSchool of Chemical Engineering and Technology, Sun Yat-sen University, Zhuhai, 519082, P.R. China. E-mail: zhouxtai@mail.sysu.edu.cn
bHuizhou Research Institute, Sun Yat-sen University, Huizhou, 516081, P.R. China
First published on 18th October 2024
Selective oxidation of cyclohexene is rather challenging due to its two bond reactive centers: allylic C–H and CC. It is of great significance to achieve highly selective oxidation towards the two reactive centers by regulating the catalytic system. Herein, the selective oxidation towards the allylic C–H and CC bond reactive sites of cyclohexene was achieved by using CuCl2 and VCl3 as the catalyst, respectively, in the presence of tert-butyl hydroperoxide (TBHP). The CuCl2-catalyzed oxidation of cyclohexene mainly yielded oxidation products at the allylic position, while the VCl3-catalyzed oxidation of cyclohexene mainly yielded epoxidation products at the CC bond. The two different reaction mechanisms are mainly due to the different roles of the t-BuOO˙ radical in the respective catalytic systems. Electron paramagnetic resonance characterizations showed that the amount of t-BuOO˙ radicals in the CuCl2 catalytic system is much lower than that in the VCl3 system. The two different mechanisms were proposed by the means of 18O2 experiments, KIE kinetics and EPR. These mechanisms revealed that the CuCl2 catalyst could rapidly generate the t-BuOO˙ radical, which undergoes bimolecular decay to produce O2. The oxygen then combines with the allyl radical of cyclohexene to produce the oxidized product at the allylic reactive center. In contrast, the VCl3 catalyst promoted the generation of the t-BuOO˙ radical, which reacted with the CC bond of cyclohexene to directly yield the epoxide.
Many efforts have been made in recent years to improve the selectivity of some desired products from cyclohexene oxidation by developing various catalytic protocols with environmentally friendly oxidants, such as molecular oxygen, H2O2 or TBHP (tert-butyl hydroperoxide).11–14 In general, cyclohexene was oxidized with molecular oxygen catalyzed by various metal catalysts, such as copper,15 cobalt,16 manganese salts,17 and the main obtained products were generated from the allylic position, such as 2-cyclohexen-1-ol and 2-cyclohexen-1-one (Path A). This could be ascribed to the cyclohexenyl radical, which could be generated smoothly using a metal catalyst or via thermal dehydrogenation. Subsequently, an allylic hydroperoxide radical (ROO˙) intermediate generated easily from the reaction between the cyclohexenyl radical and O2.18 The oxidation of cyclohexene with O2 is subject to radical chain reactions, with ROO˙ radicals being the main chain carriers.19
Different product distributions in cyclohexene oxidation could be obtained when H2O2 was used as the oxidant.20 The decomposition mechanism of H2O2 is closely related to the type of catalyst. Efficient catalysts for the oxidation of cyclohexene using H2O2, such as Nb silicates,21 Mn/Al2O3,22 Ag/TiO2/SiO2,23 achieved a high selectivity for the epoxide products. Such catalysts promote the heterolytic cleavage of H2O2 to HOO˙ radicals, which can coordinate at the catalytic metal center to selectively transfer one oxygen atom to the CC bond of cyclohexene, forming epoxides. However, various catalysts, such as Cu-zeolite-Y,24 Fe-based MOF MIL-88A (Fe),25 CoW composite oxides,26 promoted the homolytic cleavage of H2O2 to ˙OH, which resulted in products from the allylic oxidation pathway (Path A).
Compared with H2O2, TBHP exhibits a higher stability and is widely used in the large-scale production of propylene oxide, such as the Halcon processes, which simultaneously produce tert-butanol as a byproduct.27–29 The type of catalyst also plays a key role in the selectivity of products for cyclohexene oxidation with TBHP as the oxidant. For example, VO2-SiO230 and Mo-based catalysts31 showed an excellent epoxidation selectivity for cyclohexene oxidation in the presence of TBHP; most other metal–base catalysts, such as Cu, Co and Mn, exhibit dominant selectivity for allylic products.32–34 The homolytic decomposition of the O–O bond in the TBHP molecule could generate t-BuO˙ and ˙OH radicals, while ROO˙ radicals are the main radicals in the heterolytic cleavage of TBHP. The difference in product selectivity could be ascribed to the generation of different types of reactive oxygen species by various catalysts.
Thus, the oxidation of cyclohexene is highly complicated, which involves a complex reaction network of various free radicals and chain reactions. Especially for TBHP, three kinds of reactive oxygen radicals (t-BuO˙, ˙OH and t-BuOO˙) could be produced via the cleavage of the TBHP molecule, which directly affect the selectivity of the products. Therefore, achieving selective generation of reactive oxygen radicals in the cleavage of TBHP is of great significance for selective control of cyclohexene oxidation products.
With this idea in mind, the selective oxidation of cyclohexene was achieved by using CuCl2 and VCl3 as catalysts in the presence of TBHP. Two distinct mechanisms, namely the allylic oxidation pathway and the epoxidation pathway, were obtained, respectively. A series of characterizations, such as EPR, KIE kinetics and 18O2 experiments, showed that the different roles of the t-BuOO˙ radical in the respective catalytic systems yielded two distinct mechanisms. This work would provide the potential application prospects for improving the product selectivity of the alkene with the allyl site and CC bond as two active sites.
As shown in Table S1,† only 9% cyclohexene was converted in the absence of catalyst. It is known that there are significant differences in the catalytic performance of different metal chlorides. This is mainly related to the different abilities of various metal chlorides to activate TBHP for generating reactive oxygen species. Among these, CuCl2 and VCl3 exhibit excellent catalytic performances, with conversion rates of 93% and 90%, respectively. The effect of reaction conditions, such as the solvent and reaction temperature, on cyclohexene oxidation was explored (Tables S2–S4†). Upon closer observation, it was found that when copper chloride was used as the catalyst, the products obtained from cyclohexene oxidation were mainly via pathway A, while using vanadium chloride mainly produced pathway B products (Fig. 1 and Fig. S2†). However, for other metal chlorides, there seems to be no significant difference in the products that are generated from the two pathways. Therefore, we attempted to explore the different mechanisms for catalytic oxidation of cyclohexene catalyzed by CuCl2 and VCl3.
The allyl position of cyclohexene is an active reaction site, which can be determined by kinetic isotope effect (KIE) experiments (Fig. S3†).35 A KIE value of 6.73 (KH/KD) was obtained for the oxidation of cyclohexene-H10 and cyclohexene-D10 with TBHP catalyzed by CuCl2, which indicates that the allyl-H activation step is the rate-determining step. However, the KIE value of 0.87 demonstrates that the oxidation mechanism of cyclohexene when using VCl3 as a catalyst is completely different from that when using CuCl2 as the catalyst (Fig. 2). The findings obtained from the KIE tests are consistent with the above catalytic performances.
Subsequently, the effect of TBHP oxidant dosage on the reaction was further investigated by using CuCl2 as the catalyst. As a result, it was found that with the increase of TBHP dosage, the oxidation products at the allyl position gradually increased, while the products from epoxidation gradually decreased (Fig. 3a). Meanwhile, numerous gas bubbles were observed in the reaction solution during the experiments (Fig. S4b†). After collection and analysis, the generated gas in the solution was confirmed to be molecular oxygen (Fig. S5a†). The generated dioxygen content is much higher than the oxygen content in air and a standard gas containing 21% oxygen (Fig. S5b†). Therefore, it could be deduced that the produced dioxygen originates from TBHP.36
Afterwards, while using different amounts of TBHP as the oxidant, a certain amount of oxygen (flow rate: 5 mL min−1) was introduced into the reaction system. It was found that, regardless of the amount of TBHP used, the selectivity of Path A products remained unchanged (Fig. 3b). This indicates that the role of TBHP is to promote the generation of dioxygen, which is an important participant in this catalytic process. Cyclohexene oxidation catalyzed by CuCl2 with TBHP was then carried out under an 18O2 atmosphere. Apart from products containing 16O, products containing 18O at the allyl position were observed by CSI-MS (18O-labeled 2-cyclohexen-1-one, m/z = 98.0618; 18O-labeled 2-cyclohexen-1,4-dione, m/z = 114.0453) (Fig. S6†).
Similarly, the effect of TBHP oxidant dosage on the oxidation of cyclohexene was also investigated with VCl3 as the catalyst. However, the selectivity of the product hardly changed with the amount of TBHP, with the main generated products corresponding to the CC bond position (Path B) (Fig. 4a). Meanwhile, no gas bubbles were observed in the reaction solution. After introducing oxygen at a 5 mL min−1 flow rate, the selectivity of the product changed significantly. The oxidized products were transformed to those generated at the allyl position (Path A) (Fig. 4b). This also indicates that the reaction rate between allyl radicals and oxygen is much higher than that of epoxidation.
Both in CuCl2- and VCl3-catalyzed oxidations of cyclohexene with TBHP, EPR signals corresponding to radical adducts of dimethyl pyridine N-oxide (DMPO) were observed. The adducts determined by hyperfine coupling constant (HFCC) values (αN = 14.5 G, αH(β) = 11 G, and g = 2.0061) can be attributed to DMPO spin adducts of the peroxyl radical (t-BuOO˙) (Fig. 5a and b).37 Comparing the peak intensity of peroxide radicals in the two catalytic systems, it can be seen that the amount of peroxide radicals in the CuCl2 catalytic system is much lower than that in the VCl3 system (Fig. 5c). The kinetics of the consumption of the peroxide radical indicated that the consumption of t-BuOO˙ radicals in the CuCl2 reaction system was much faster than that in the VCl3 reaction system (Fig. S7†). Combining the TBHP consumption rate curves of the two catalytic systems, it is evident that the role of the peroxide radicals is significantly different in the two systems (Fig. S8†). To further investigate whether the epoxidation of olefins is a mechanism of peroxyacids or the peroxyl radical, the epoxidations of cis-styrene catalyzed by CuCl2 and VCl3 were investigated. The results showed that both in the VCl3 and CuCl2 systems, the peroxide radical is the main species for olefin epoxidation. However, the efficiency of the CuCl2 catalyst for catalytic cis-styrene epoxidation was much lower that of VCl3. In addition, peroxyacids were the main species for epoxidation in the absence of catalyst (Table S5†).38
Three types of oxygen-containing radicals, including the tert-butylperoxyl radical (t-BuOO˙), tert-butoxyl radical (t-BuO˙) and hydroxyl radical (˙OH), could be generated from the decomposition of TBHP. The adducts determined by hyperfine coupling constant (HFCC) values (αN = 13.19 G, αH(β) = 8.16 G, and g = 2.0059) could be attributed to DMPO spin adducts of t-BuO˙ (Fig. S9a†).39 The adducts determined by hyperfine coupling constant (HFCC) values (αN = 14.8 G, αH(β) = 14.8 G, and g = 2.0059) can be attributed to DMPO spin adducts of ˙OH (Fig. S9b†).40 On the basis of the literature, TBHP is induced by Cu2+ to generate t-BuOO˙ and simultaneously Cu2+ is reduced to Cu+. Conversely, t-BuO˙ can be formed via the reductive decomposition of TBHP, which is accompanied by the regeneration of the Cu2+ catalyst from Cu+.41 Certainly, the bimolecular decay of t-BuOO˙ is another way to provide t-BuO˙, and more significantly, the oxygen produced in this process plays a key role in this reaction as the unique oxygen source.
In addition, the adducts of the cyclohexenyl radical and cyclohexenyl peroxyl radical by DMPO were detected at a mass to charge ratio (m/z) of 194.1535 and 226.1429, respectively (Fig. S10d and Fig. S10f†). Therefore, the t-BuOO˙ radical undergoes a self-polymerization reaction catalyzed by CuCl2 to produce O2. The generated dioxygen could combine with the cyclohexenyl radical to generate the corresponding oxidation products (Scheme 2).42
Scheme 2 Pathway for dioxygen and 2-cyclohexen-1-one generation in the CuCl2 catalytic oxidation of cyclohexene. |
According to DFT calculations, the decomposition of O–O bond homolysis to ˙OH is favored over heterolysis to the ROO˙ radical, as it requires less energy.43 The O–O bond homolysis of di-tert-butyl peroxide (DTBP) produces the tert-butoxyl radical (t-BuO˙), while the O–O bond homolysis of hydrogen peroxide (H2O2) produces the hydroxyl radical (˙OH). Using DTBP and H2O2 as oxidants, some control experiments catalyzed by CuCl2 were carried out to determine the roles of the t-BuO˙ and ˙OH radicals.
As a substrate, cyclohexane epoxide was not converted at all, regardless of the oxidant used among TBHP, DTBP and H2O2. This indicates that the generation of 1,2-cyclohexane-diol (5) is not obtained via the deep oxidation or hydration of epoxide (entries 1–3 in Table S6†). In contrast, cyclohexene could be converted with various conversions when using any of the three oxidants (entries 4–6 in Table S6†). The conversion was 35% and the main products were generated through Path A when H2O2 was used as the oxidant (entry 6 in Table S6†). However, the conversion was only 10% and the only product was the combination of the cyclohexenyl radical and t-BuO˙ when DTBP was used as the oxidant (entry 5 in Table S6†). These results indicate that both t-BuO˙ and ˙OH radicals can abstract the hydrogen atom from the α-position of cyclohexene to generate the cyclohexene radical.
No conversion was obtained in the oxidation of both cyclohexen-1-one and 2-cyclohexen-1-ol using DTBP as the oxidant (entry 8 and 10 in Table S6†). However, when using TBHP and H2O2 as the oxidant, the main products were 1-one and 1,4-dione in the oxidation of cyclohexen-1-ol and 2-cyclohexen-1-one, respectively. Such observations indicate that ˙OH radicals are active species that promote further oxidation of cyclohexen-1-ol and 2-cyclohexen-1-one.
From Table S2 (entry 1†), it can be seen that the main product of cyclohexene oxidation was diol in the absence of catalyst, with a selectivity of up to 43%. When adding VCl3 catalyst into the solution, the selectivity of the diols decreased from 43% to 7%, while the selectivity of epoxide increased from 5% to 51%. Combining the results of cis-styrene epoxidation (Table S5†), the promoted generation of t-BuOO˙ by VCl3 could be concluded, thereby enhancing the selectivity of epoxide. In addition, cyclohexanol radicals were observed in CSI-MS spectra of DMPO spin adducts at a mass to charge ratio (m/z) of 212.1640 (Fig. S10h†), which were the products of the addition reaction between the ˙OH radical and double bond of cyclohexene. Subsequently, diol was generated from the combination of the cyclohexanol radical and ˙OH radical, not the hydration of epoxide.
Based on above observations and discussions, the roles of the three kind radicals—t-BuOO˙, t-BuO˙ and ˙OH—can be summarized as shown in Fig. 6. When the oxidation of cyclohexene was catalyzed by CuCl2, the cyclohexenyl radical was promoted to generate through t-BuO˙ and ˙OH radicals. Therefore, the main products were generated from the allyl position of cyclohexene (Route 1 and Route 2 in Fig. 6). However, the t-BuOO˙ radical was the active species when cyclohexene oxidation was catalyzed by VCl3, which resulted in the main products being generated from the CC bond of cyclohexene (Route 4 in Fig. 6). In addition, the ˙OH radical was the active species for generating 1,2-cyclohexane-diol when the oxidation was conducted in the absence of catalyst (Route 3 in Fig. 6).44
The different pathways of cyclohexene oxidation catalyzed by CuCl2 and VCl3 could be well demonstrated, according to the above discussion. A tentative mechanism of cyclohexene oxidation via Path A in the CuCl2/TBHP system was proposed, as shown in Scheme 3. Initially, TBHP is induced by Cu2+ to generate t-BuOO˙ and the Cu2+ was reduced to Cu+ (eqn (1) in Scheme 3). Conversely, t-BuO˙ can be formed via the reductive decomposition of TBHP, which is accompanied by the regeneration of Cu2+ from Cu+ (eqn (2) in Scheme 3). Certainly, the oxygen was produced via the bimolecular decay of t-BuOO˙, also providing t-BuO˙ in the presence of CuCl2 (eqn (3) in Scheme 3). Next, the oxygen combines with the cyclohexenyl radical to generate the cyclohexenyl peroxyl radical, and then to provide the corresponding products at the allyl position of cyclohexene (Path A) (eqn (4) and (5) in Scheme 3).
As mentioned above (Table S5†), trans-stilbene oxide was the predominant product in the epoxidation of cis-styrene catalyzed by VCl3, which suggested that the t-BuOO˙ radical was the active species. Therefore, a plausible mechanism of cyclohexene oxidation via Path B in the VCl3/TBHP system was proposed as shown in Scheme 4. The peroxyl radical (t-BuOO˙) could be generated by both the oxidation and reduction of VCl3 (eqn (1) and (2) in Scheme 4). In addition, t-BuOO˙ can be formed via the decomposition of TBHP by t-BuO˙, which is accompanied by the generation of t-BuOH (eqn (3) in Scheme 4). Then, the t-BuOO˙ radical reacted with cyclohexene to yield the epoxide directly (eqn (4) in Scheme 4).45–47
Data for this paper, including reactions screening and optimization experiments, detailed experimental procedures, spectral data, and computational study details, are provided in the ESI.†
Footnote |
† Electronic supplementary information (ESI) available: Experimental procedures, spectral data, and computational study details. See DOI: https://doi.org/10.1039/d4qo01681e |
This journal is © the Partner Organisations 2025 |