Investigation into the mechanism of polyoxotungstates-catalyzed cyclooctene epoxidation by ESI-MS

Abstract

The reaction process of cyclooctene epoxidation was real-time monitored by ESI-MS using H₃PW₁₂O₄₀ and H₂WO₄ as catalysts. By observing the species changes of the catalysts in the reaction solutions and the combination of catalysts with the substrate, a catalytic reaction mechanism of polyoxometalate clusters was proposed from the aspect of aggregation and dissociation. When H₃PW₁₂O₄₀ was used as the catalyst, it was degraded into minor peroxo-species [PW₄O_16−x(O₂)_x]³⁻ (x = 0, 2, 4, 6, 8) which performed as the active center of the epoxidation of cyclooctene. When H₂WO₄ was used as the catalyst, it reacted with H₂O₂ and generated mononuclear tungsten peroxo-species [HWO₂(O₂)₂]⁻ and [HWO₂(O₂)₃]⁻. After the addition of cyclooctene, the peroxo-species catalyzed the epoxidation of cyclooctene, generating 1,2-epoxycyclooctane and combining with 1,2-epoxycyclooctane to form the intermediate [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻.

---

Introduction

Polyoxometalates (POMs) consist of metal oxide clusters based on the early transition metals Mo, W, V, Nb and Ta, and have provoked significant interest in catalysis due to their redox and acid–base properties.^1,2 They have been used as homogeneous catalysts for a wide variety of organic substrate oxidations, such as alkane, olefins and alcohols etc.^3–5 Epoxides are products of olefin epoxidation, and are valuable compounds used as resins, paints, surfactants, and intermediates in various organic syntheses.⁶ In contrast to synthetic methods with low atom efficiencies that use stoichiometric reagents, catalytic epoxidation of alkenes with hydrogen peroxide (H₂O₂) has attracted much attention, because H₂O₂ has a high content of active oxygen species and only water is formed as a by-product. Therefore, a large number of H₂O₂-based epoxidation systems that use POM catalysts have been developed, such as H₃PW₁₂O₄₀ and TBA₄H₄[γ-SiW₁₀O₃₆] etc.^7–13

Tungsten-based epoxidation systems with hydrogen peroxide have attracted much attention because of their high reactivity for olefins and inherent poor activity for the decomposition of hydrogen peroxide.¹² In the catalytic process, polyoxotungstates act as catalyst precursors, which decompose into the monomeric, dimeric and tetrameric peroxo species by reaction with hydrogen peroxide.¹⁴ The [PO₄{WO(O₂)₂}₄]^{3− 7,15,16} and [W₂O₃(O₂)₄(H₂O)₂]^2− 17,18 peroxotungstates, which can catalyze the epoxidation, have been isolated and characterized crystallographically. Although the peroxo-species in the TBA₂[W₂O₃(O₂)₄(H₂O)₂]/H₃[PW₁₂O₄₀]-based epoxidation systems with H₂O₂ have been investigated by ³¹P/¹⁷O NMR and Density Functional Theory (DFT) calculations,^19–21 the intermolecular oxidation reaction providing further proof that peroxo-species are reactive species has not yet been probed.

Mass spectrometry, as a new characterization method for the observation of active species,^22,23 has been previously applied to the study of catalytic mechanisms. For example, Cronin observed the generation of an intermediate Fe(v)=O species within a synthetic non-haem complex and its reaction with an olefin by variable temperature mass spectrometry.²⁴ In this paper, we investigate the reaction mechanism for the epoxidation of cyclooctene with H₂O₂ catalyzed by H₃PW₁₂O₄₀ and H₂WO₄ with mass spectroscopy for the aspects of decomposition and aggregation and explored the effects of reaction time, ratio of catalyst in the catalytic system.

Results and discussion

Epoxidation of cyclooctene by H₂O₂ catalyzed by polyoxometalates

The oxidation of cyclooctene by H₂O₂ was examined in the presence of different amounts of H₃PW₁₂O₄₀ (catalyst 1) and H₂WO₄ (catalyst 2). The results of these reactions are summarized in Table 1 and Fig. S1 and S2.† The reaction cannot proceed without a catalyst (Entry 5). These two kinds of polyoxometalates were successful catalysts evaluated in terms of both conversion of cyclooctene and yield of 1,2-epoxycyclooctane when the amount of catalyst is low (Entry 1 and 3). However, increasing the amount of catalyst led to a significant decrease of 1,2-epoxycyclooctane yield (Entry 2 and 4) without any other by-products increased (Fig. S1 and S2†). This phenomenon of non-conservation in GC analysis strongly suggests that the catalyst and 1,2-epoxycyclooctane combined to form new complexes, which may be difficult to gasify in GC.

Table 1 Oxidation of cyclooctene catalyzed by H₃PW₁₂O₄₀ and H₂WO₄ using 30% H₂O₂ as oxidant^a

Entry	Catalyst	Concentration [mM]		Conv.^b [% mol]	Yield^b [% mol]
		[POM]	[W]
a Reaction condition: refer to the Experimental section.b Determined by GC with area normalization.
1	H3PW12O40	0.005	0.06	86.71	76.30
2	H3PW12O40	0.1	1.2	84.63	20.41
3	H2WO4	0.06	0.06	80.52	71.21
4	H2WO4	1.2	1.2	80.37	35.29
5	None	0	0	0	0

---

MS spectra of the catalytic solution

To confirm our speculation, the final catalytic solution was characterized by electrospray ionization mass spectrometry (ESI-MS). In the ESI-MS spectra of the catalytic solution catalyzed by H₃PW₁₂O₄₀ (Fig. 1A) and H₂WO₄ (Fig. 1B), the most intense peak (m/z 765.1) with an isotopic distribution (Fig. S3A†) that agrees with the pattern calculated for [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ can be seen as a combination of peroxo-species {W₂} and 1,2-epoxycyclooctane (Fig. S3B†). Besides, [PW₁₂O₄₀]³⁻ (m/z 958.7) and [HPW₁₂O₄₀]²⁻ (m/z 1438.6) also exist in the catalytic system 1 (Fig. 1A), while species belonging to [WO₄]²⁻ cannot be measured in the system 2 (Fig. 1B) for the low solubility of H₂WO₄. In order to infer the formation of [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻, we replaced cyclooctene with 1,2-epoxycyclooctane in the catalytic solution, i.e. 1,2-epoxycyclooctane, H₂O₂ and catalysts dissolved in CH₃CN. The MS spectra of the final solution were the same as the system with cyclooctene as the substrate (Fig. S4†) and no other organic product was monitored in the GC spectra, which tends to indicate that [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ was assigned to a combination of peroxo-species with a product of 1,2-epoxycyclooctane.

Fig. 1 MS spectra of catalytic solution epoxidation reaction of cyclooctene after the epoxidation reaction. Reaction conditions: cyclooctene (1 mmol), H₂O₂ (1 mmol), CH₃CN solvent (2 mL), 50 °C, 4 h. (A) H₃PW₁₂O₄₀ (0.005 mmol) as catalyst; (B) H₂WO₄ (0.06 mmol) as catalyst.

To further probe the formula of the [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ anions generated from the different catalysts, CID experiments on [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ were performed (Fig. 2). The CID spectra of [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ with H₃PW₁₂O₄₀ as the catalyst (Fig. 2A) and H₂WO₄ as the catalyst (Fig. 2B) were similar, proving that [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ anions have the same structure in different catalytic systems. Besides, the CID spectra of the [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ from cyclooctene and 1,2-epoxycyclooctane with H₃PW₁₂O₄₀ (Fig. S5A and B†) and H₂WO₄ (Fig. S5C and D) as the catalyst are almost the same.

Fig. 2 CID mass spectra of [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ generated from different reaction solutions. (A) H₃PW₁₂O₄₀ (0.005 mmol) as the catalyst; (B) H₂WO₄ (0.06 mmol) as the catalyst. Collision energy = 12 eV, isolation width = 9. The parent ion (denoted by a diamond) is shown in a blue square box in each spectrum.

Real-time monitoring on the process of cyclooctene epoxidation

The reaction solutions of 1 and 2 (H₃PW₁₂O₄₀ and H₂WO₄ as catalyst respectively) were real-time monitored by ESI-MS to reveal the change of species during the reaction, identifying the species requisite for cyclooctene epoxidation and the complex species that catalytic active species combined 1,2-epoxycyclooctane.

Firstly, the reaction solutions were real-time monitored using ESI-MS by one-step (Fig. S6 and S7†). First, the catalyst, H₂O₂ (1 mmol) and cyclooctene (1 mmol) were dissolved in CH₃CN. Then the samples were taken from the solution at intervals. The real-time monitoring of the solution of catalyst 1 (Fig. S6†) showed that the peaks assigned to [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ appeared at the 5^th min and its abundance increased progressively. At last, it became the most intense peak at about the 60^th min. The corresponding GC spectra showed that cyclooctene was completely converted to 1,2-epoxycyclooctane at about the 60^th min (Fig. S9A†). The ESI-MS spectra of the real-time monitoring of the solution containing catalyst 2 (Fig. S7†) were similar to that with catalyst 1. The peaks assigned to [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ appeared at the 5^th min and became the most intense peaks for the low solubility of H₂WO₄. The corresponding GC spectra showed that cyclooctene was converted to 1,2-epoxycyclooctane completely at about the 3^rd hour (Fig. S10A†).

To further investigate how the [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ formed, the catalytic reactions were real-time monitored using ESI-MS by two-step. H₂O₂ and the catalyst were mixed previously and reacted for 4 h, and cyclooctene was added to the solution.

For the two-step catalytic system 1 (H₃PW₁₂O₄₀ as catalyst), the ESI-MS monitoring data (Fig. 3) was illustrated as follows. Firstly, when H₃PW₁₂O₄₀ (0.005 mmol) was dissolved in CH₃CN, all major peaks in the mass spectrum can be assigned to [PW₁₂O₄₀]³⁻ (m/z 958.7) and [HPW₁₂O₄₀]²⁻ (m/z 1438.6), which means the {PW₁₂} unit is stable in CH₃CN at this step (Fig. 3A). After adding 1 mmol H₂O₂ to the previous solution with stirring at 50 °C for 5 min, some new species formed during this process. As the spectrum (Fig. 3B) shows, the new peaks at m/z 340.9–383.6 can be assigned to [PW₄O_16−x(O₂)_x]³⁻ (x = 0, 2, 4, 6, 8), with m/z 340.9 for [PW₄O₁₆]³⁻, m/z 351.6 for [PW₄O₁₄(O₂)₂]³⁻, m/z 362.2 for [PW₄O₁₂(O₂)₄]³⁻, m/z 372.9 for [PW₄O₁₀(O₂)₆]³⁻, m/z 383.6 for [PW₄O₈(O₂)₈]³⁻, meaning that the parent {PW₁₂} can react with H₂O₂, yielding the degraded peroxo-species [PW₄O_16−x(O₂)_x]³⁻. Upon continuous heating of the solution for 4 h, the intensity of the peaks assigned to [PW₄O_16−x(O₂)_x]³⁻ (x = 0, 2, 4, 6, 8) gradually increased and became the main peaks (Fig. 3C). After adding cyclooctene (1 mmol) to the former reaction solution of {PW₁₂} + H₂O₂ with stirring at 50 °C for 5 min, the new species related to [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ (m/z 765.1) emerged accompanied with a decline of [PW₄O_16−x(O₂)_x]³⁻ (Fig. 3D). After continuous heating of the solution for 4 h, the intensity of [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ increased (Fig. 3E), indicating that the [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ intermediate was a combination of peroxo-species {W₂} and product 1,2-epoxycyclooctane. As 1,2-epoxycyclooctane increased in the system, the amount of [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ gradually increased. Besides, the corresponding GC spectra showed that cyclooctene quickly completed the conversion to 1,2-epoxycyclooctane at about the 10^th min (Fig. S9B†), which supported that [PW₄O_16−x(O₂)_x]³⁻ (x = 0, 2, 4, 6, 8) were active species in the catalytic system.

Fig. 3 Real-time monitoring on the cyclooctene epoxidation by two-step of system 1. (A) H₃PW₁₂O₄₀ (0.005 mmol) dissolved in CH₃CN; (B) stirring 5 min at 50 °C after the addition of H₂O₂ (1 mmol); (C) stirring 4 h; (D) stirring further 5 min after the addition of cyclooctene (1 mmol); (E) stirring further 4 h.

For the catalytic system 2 (H₂WO₄ as catalyst), the real-time ESI-MS monitoring data (Fig. 4) by two-step was similar to system 1. Firstly, when the catalyst H₂WO₄ (0.06 mmol) mixed with CH₃CN, there are only weak peaks assigned to [HWO₄]⁻ as a result of the low solubility of H₂WO₄ in CH₃CN (Fig. 4A). After adding 1 mmol H₂O₂ to the H₂WO₄ solution with stirring at 50 °C for 5 min, a remarkable change occurred in the mixed solution. As the spectrum (Fig. 4B) shows, the new peaks at m/z 280.9–296.9 can be assigned to [HWO₂(O₂)₂]⁻ (m/z 280.9) and [HWO(O₂)₃]⁻ (m/z 296.9). This result indicates that H₂WO₄ can react with H₂O₂, producing soluble peroxo-species [HWO_4−x(O₂)_x]⁻ (x = 2, 3). Besides, the reaction was performed under heating, which also increases the solubility of H₂WO₄. Upon continuous heating of the solution for 4 h, the intensity of the peaks assigned to [HWO(O₂)₃]⁻ gradually increased and became the main peak (Fig. 4C). Then, after adding cyclooctene (1 mmol) to the mixed reaction solution of H₂WO₄ + H₂O₂ with stirring at 50 °C for 5 min, a new species belonging to [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ (m/z 765.1) emerged accompanied with a decrease of [HWO_4−x(O₂)_x]⁻ (Fig. 4D). After continuous heating of the solution for 4 h, the intensity of [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ increased (Fig. 4E). The corresponding GC spectra showed that cyclooctene converted to 1,2-epoxycyclooctane completely at about the 30^th min (Fig. S10B†), which supported the speculation that [HWO_4−x(O₂)_x]⁻ (x = 2, 3) were active species in the catalytic system.

Fig. 4 Real-time monitoring of the cyclooctene epoxidation by a two-step of system 2. (A) H₂WO₄ (0.06 mmol) mixed with CH₃CN; (B) stirring for 5 min at 50 °C after the addition of H₂O₂ (1 mmol); (C) stirring for 4 h; (D) stirring for a further 5 min after the addition of cyclooctene (1 mmol); (E) stirring for a further 4 h.

In summary, using Keggin type heterpolyoxotungstates H₃PW₁₂O₄₀ and mononuclear tungstate H₂WO₄ as catalysts, the reaction process of cyclooctene epoxidation was real-time monitored by ESI-MS (Scheme 1). When H₃PW₁₂O₄₀ was used as catalyst, it was degraded into minor peroxo-species [PW₄O_16−x(O₂)_x]³⁻ (x = 0, 2, 4, 6, 8) which performed as the active center of the epoxidation of cyclooctene. After the addition of cyclooctene, the peroxo-species and 1,2-epoxycyclooctane formed intermediate [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻. When H₂WO₄ was used as the catalyst, it reacted with H₂O₂ and generated mononuclear tungsten peroxo-species [HWO₂(O₂)₂]⁻ and [HWO₂(O₂)₃]⁻, which catalyzed the epoxidation of cyclooctene generating 1,2-epoxycyclooctane. After the addition of cyclooctene, the peroxo-species and 1,2-epoxycyclooctane formed intermediate [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻. In the presence of a large amount of catalyst, the non-conservation of olefin-epoxide in GC analysis also confirmed the existence of a stable intermediate. Future work will be focused on the structure and effect of the intermediate [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ in the catalytic system.

Scheme 1 Speculated mechanism of cyclooctene epoxidation using H₃PW₁₂O₄₀ and H₂WO₄ as catalysts.

Experimental

Materials

All chemicals were purchased from commercial sources and used without further purification.

Catalytic oxidation of cyclooctene

Oxidations of cyclooctene were carried out in a thermostated sealed glass tube equipped with a magnetic stirrer. The initial ratio of cyclooctene (1 mmol) oxidation by H₂O₂ (1 mmol, added as 30 wt% aqueous solution) at 50 °C for 4 h in 2 mL CH₃CN was used to assess the relative catalytic activities of the different POMs. Products were identified from their mass (GC-MS) spectra. The yields of the product 1,2-epoxycyclooctane as well as the conversion of the initial substrate cyclooctene were quantified by GC with area normalization.

Mass spectrometry

All MS data were obtained in the negative ion mode using an Agilent 6520 Q-TOF LC/MS. An electrospray source was used to collect data under the conditions specified below. The detector was a time-of-flight and all data were processed using the Agilent qualitative analysis B.04.00 software, while simulated isotope patterns were investigated using Agilent Isotope Pattern software and Molecular Weight Calculator. The following parameters were consistent for all ESI-MS scans given below. The calibration solution used was Agilent ESI tuning mix solution, enabling calibration between ∼100 m/z and 3000 m/z. This solution was diluted 100 : 1 with acetonitrile. Samples were introduced into the MS via an automatic sampler. The electrospray settings were set with the drying nitrogen gas temperature at 300 °C. The ion polarity for all MS scans recorded was negative, with the voltage of the capillary tip set at 3500 V, skimmer voltage at 65 V, RF at 750 V, fragmentor at 80 V. All theoretical peak assignments were determined via comparison of the experimentally determined isotopic patterns for each peak, with simulated isotopic patterns. CID experiments were performed using N₂ as the target gas. Collisional energy voltage was 12 V for [HW₂(O₂)₂O₅(C₈H₁₄O)₂]⁻ precursors.

Gas chromatography

GC analyses were performed on Shimadzu GC-2014C with a FID detector equipped with an Rtx-1701 Sil capillary column.

Acknowledgements

The authors thank the National Natural Science Foundation of China (21173021, 21231002, 21276026, 21371025), 973 Program (2014CB932103), the 111 Project (B07012) and the Fundamental Research Grant (20121942006) by Beijing Institute of Technology.

Notes and references

1. M. T. Pope, Heteropoly and Isopoly Oxometalates, Springer-Verlag, Berlin, 1983.
2. C. L. Hill, Chem. Rev., 1998, 98, 1–390.
3. K. Kamata, K. Yonehara, Y. Sumida, K. Yamaguchi, S. Hikichi and N. Mizuno, Science, 2003, 300, 964–970.
4. R. Ben-Daniel, A. M. Khenkin and R. Neumann, Chem.–Eur. J., 2000, 6, 3722–3728.
5. K. Yamaguchi, S. Shinachi and N. Mizuno, Chem. Commun., 2004, 424–425.
6. J. E. Bäckvall, Modern Oxidation Methods, Wiley-VCH, Weinheim, Germany, 2004.
7. R. B. Brown and C. L. Hill, J. Org. Chem., 1988, 53, 5762–5768.
8. A. C. Dengel, W. P. Griffith and B. C. Parkin, J. Chem. Soc., Dalton Trans., 1993, 2683–2688.
9. R. Ishimoto, K. Kamata and N. Mizuno, Angew. Chem., Int. Ed., 2012, 51, 4662–4665.
10. N. Mizuno, C. Nozaki, I. Kiyoto and M. Misono, J. Catal., 1999, 182, 285–288.
11. R. Neumann and M. Gara, J. Am. Chem. Soc., 1995, 117, 5066–5074.
12. T. M. Anderson, X. Zhang, K. I. Hardcastle and C. L. Hill, Inorg. Chem., 2002, 41, 2477–2488.
13. D. C. Duncan, R. C. Chambers, E. Hecht and C. L. Hill, J. Am. Chem. Soc., 1995, 117, 681–691.
14. N. Mizuno, K. Yamaguchi and K. Kamata, Coord. Chem. Rev., 2005, 249, 1944–1956.
15. Y. Ding, W. Zhao, H. Hua and B. C. Ma, Green Chem., 2008, 10, 910–913.
16. Y. Ding, B. C. Ma, D. J. Tong, H. Han and W. Zhao, Aust. J. Chem., 2009, 62, 739–746.
17. K. Kamata, K. Yamaguchi, S. Hikichi and N. Mizuno, Adv. Synth. Catal., 2003, 345, 1193–1196.
18. K. Kamata, K. Yamaguchi and N. Mizuno, Chem.–Eur. J., 2004, 10, 4728–4734.
19. C. Venturello and R. D’Alosio, J. Org. Chem., 1988, 53, 1553–1557.
20. Y. Ishii, K. Yamawaki, T. Ura, H. Yamada, T. Yoshida and M. Ogawa, J. Org. Chem., 1988, 53, 3587–3593.
21. R. Ishimoto, K. Kamata and N. Mizuno, Eur. J. Inorg. Chem., 2013, 10–11, 1943–1950.
22. L. Y. Fan, Z. G. Lin, J. Cao and C. W. Hu, Inorg. Chem., 2016, 55(6), 2900–2908.
23. Q. D. Jia, J. Cao, Y. P. Duan and C. W. Hu, Dalton Trans., 2015, 44, 553–559.
24. I. Prat, J. S. Mathieson, M. Güell, X. Ribas, J. M. Luis, L. Cronin and M. Costas, Nat. Chem., 2011, 3, 788–793.

---

Footnote

† Electronic supplementary information (ESI) available: Additional GC spectra of the epoxidation reaction of cyclooctene; comparison of calculated and observed mass spectra; CID mass spectra; real-time monitoring on the cyclooctene epoxidation by one-step; conversion-time curve of cyclooctene epoxidation. See DOI: 10.1039/c6ra09561e

---