Novel tungsten-based catalyst for epoxidation of cyclohexene

HU Hongding，ZHU Mingqiao，Umsa JAMEEL ，TONG Zhangfa



Novel tungsten-based catalyst for epoxidation of cyclohexene

HU Hongding1，ZHU Mingqiao1，Umsa JAMEEL1，TONG Zhangfa2

(College of Chemical and Biological EngineeringZhejiang UniversityHangzhouZhejiangChina；Guangxi Key Laboratory of Petrochemical Resource Processing and Process Intensification TechnologySchool of Chemistry and Chemical EngineeringGuangxi UniversityNanningGuangxiChina

Olefin epoxidation is of academic and industrial importance in modern chemistry. However, olefin epoxidation is still quite a challenge due to the difficulty of mass transfer when using the green oxidant hydrogen peroxide. A novel tungsten-based catalyst for epoxidation was prepared using an oxidative condensation method, and was applied to study the effect on epoxidation of cyclohexene with commercially available hydrogen peroxide as an oxidant. The conversion of cyclohexene (51.2%) and selectivity to cyclohexene oxide (69.2%) were obtained under a certain reaction condition. It exhibited high catalytic activity and good selectivity towards cyclohexene oxide under mild conditions without any solvent and with less environmental contaminations.

cyclohexene；tungsten-based catalyst；epoxidation；phase-transfer；catalysis

Introduction

Olefin epoxidation is a major industrial reaction for the production of a broad variety of chemicals. Epoxides are widely used as epoxy resins, paints, surfactant, and intermediates in various organic syntheses[1-2]. But olefin epoxidation is still a great challenge[3-6]. Although olefin is epoxidized efficiently using organic peroxide as an oxidant in the industrial process, it costs a lot as well as not easy to be stored. Hydrogen peroxide is probably the best terminal oxidant after oxygen[3]with respect to environmental and economic considerations[7]. There are three main kinds of catalysts used for cyclohexene epoxidation using H2O2as an oxidant: polyoxometalates (POMs)[8], supported transition metal[9]and molecular sieves[10]. In general, thepolyoxometalates (POMs) is much more effective than the others.

In recent years, tungsten-based polyoxometalates have received an increasing interest in selective oxidation of organic compounds with hydrogen peroxide[11-12]. Traditional polyoxometalates catalyst for olefin oxidation was heteropolyphosphatotungstate. Venturello[13], an earlier scholar in this research area studied the quaternary ammonium heteropolyphosphatotungstate for cyclohexene epoxidation and the resulting yield of cyclohexene oxide was 88% under mild reaction conditions. Still, plenty of drawbacks existed, for example, the use of toxic and carcinogenic chlorinated solvents leading to environmental contamination[14-16].

In 1997, Tézé[17]synthesized a new polyoxometalatestungstoborate. In 2007, Reinoso[11]also conducted a research on 13-tungstoborate stabilized by an organostannoxanehexamer. In 2010, Zhao[18]prepared tungstoborate catalyst with the same method used by Tézé and used it for oxidation of alcohols with hydrogen peroxide, but not for cyclohexene oxidation.

In this study, a novel tungsten-based catalyst containing tungsten and boron for epoxidation was prepared in a much simpler way than that used by Tézé’s and Reinoso, and was used for the epoxidation of cyclohexene with commercially available hydrogen peroxide (30%(mass) H2O2) as an oxidant. The formula and structure of the catalyst were studied with IR and TG.

1 Experimental

1.1 Catalyst preparation

The raw materials, sodium tungstate (AR, 99.7%, mass), sulphuric acid (AR, 98%, mass), hydrogen peroxide (AR, 30%, mass), boricacid (AR, 99.5%, mass) and cetyltrimethylammonium chloride (CTMA, AR, 97%, mass) were purchased from Sinopharm Chemical Reagent Co., Ltd.

Typically, an 8.25 g sample of sodium tungstate (Na2WO4·2H2O) was dissolved to configure a certain concentration of solution. An 8.17 g solution of 30%(mass) diluted sulphuric acid was added with stirring and the pale yellow precipitate appeared. Then 30%(mass) H2O2was added drop by drop with vigorous stirring until the precipitate disappeared. Boric acid 0.52 g (W/B=3) was added, and reacted for 30 min under 60℃. After this, 2.76 g of cetyltrimethyl- ammonium chloride was added, and reacted for another 60 min under 60℃with much more vigorous stirring. Then the product was cooled down to room temperature and filtered with a funnel. The solid was washed with a large number of deionized water and dried for 11 h under 40℃. Finally, the catalyst C19H44W3BNO18was obtained.

1.2 Catalyst characterization

The catalyst was characterized by infrared spectrography (IR) on the Nicolet 380 with the wave number range of 4000—400 cm-1. The pyrolysis characteristics were studied in an atmosphere of nitrogen by thermo gravimetric analysis (TG) on the Pyris 1 TGA. It was heated up from 50℃to 850℃at the speed of 10℃·min-1.

1.3 Cyclohexene oxidation

Oxidation of cyclohexene with 30%(mass) H2O2was carried out at 60℃under magnetic stirring in a 50 ml three-neck flask immersed in a thermostatted bath. Typically, the reaction mixture consisted of 0.50 g of catalyst, 0.53 g of co-catalyst (Na2HPO4), 4.10 g of cyclohexene, and 5.66 g of oxidant (30%(mass) H2O2). In all cases the molar ratio of oxidant to substrate was 1:1. The reaction products were analyzed by gas chromatography (GC-1690) using a capillary column SE-54 (column length, pore diameter and film thickness: 30 m×0.30 mm×0.5μm) and a flame ionization detector.

2 Results and discussion

2.1 Catalyst characterization

Owing to less information of the group W-O-B, the three kinds of pure raw material used for the preparation of catalysts were characterised at first and the results are shown in Fig. 1.

The IR spectra a, b and c in Fig.1 stand for H3BO3, Na2WO4·2H2O and Na2SO4, respectively, and spectrum d is the mixture of all three compounds. From Fig.1, it is clear that the spectrum d is a simple superposition of spectra a, b and c. Thus, we believed that there was no reaction triggered among H3BO3, Na2WO4·2H2O and Na2SO4.

Fig. 2 is the IR spectrum of H3BO3, the same as IR spectrum of Fig.1. Its characteristic peak wavenumbers are as following, B—OH: 3300—3200 cm-1, a strong absorption peak, the shape also is wide, 700—800 cm-1, a moderately strong and wide absorption peak, 1000 cm-1, a strong absorption peak; B—O: 1380—1310 cm-1, a very strong absorption peak. There might be a certain deviation with that in satandard IR spectra, but in general, Fig. 2 is basically identical.

Fig.3 shows the IR spectra of intermediates. Spectrum a is borotungstic acid and spectrum b is peroxy-tungstic acid. The difference between them was adding boric acid into the intermediate b. Compared with spectra b, there is no B—OH characteristic peak in spectra a, namely, there is no peak in 700—800 cm-1, and the peak in 3300—3200 cm-1is so weak that it does not conform with the manuals. However, the peak representing B—O in 1350 cm-1is quite strong. We presume that the addition of boric acid triggered a reaction, and it kept the functional groups B—O intact, but the bond between O and H was broken because of the dehydration with —OH between tungstic acid and boric acid.

The IR spectrum of catalyst with W/B=3 is shown in Fig.4 and analyzed as follows. 3500 cm-1: —OH or H2O; 2900 cm-1: —CH2—; 1480 cm-1: N+. The range of the fingerprint peaks is 1100—700 cm-1, indicating that the quaternary ammonium heteropolytungstoborate catalyst remained Keggin structure. IR[19]: 1300 cm-1(m,(B—Oa))，952 cm-1(s,(W—Od)), 870 cm-1(vw,(W—Ob))，722 cm-1(s,(W—Oc)).

According to the analysis above, we infer that a new catalyst has been synthesized.

Fig.5 shows the TGA of catalyst with W/B=3, including two curves. The upper one is a curve of thermo gravimetric analysis (TG), the remaining masspercentage of the catalyst as a function of temperature while the catalyst was heated under a program control temperature. The lower one is the derivative of the above TG curve to temperature, named the derivative thermo gravimetric analysis.

The TG curve is divided into six phases according to the rate of mass reduction, as shown in Fig.5. The temperature range of the first stage was 50—183.07℃, and the remaining mass was 97.25%, and also it was the stage of free water dehydration. The temperature of the second stage was between 183.07℃and 300.04℃with the remaining mass 90.57%, the stage of crystal water dehydration. The temperature range of the third, fourth and fifth stage was 278.94—427.30℃, and the remaining mass was 58.47%, and it was the stage of quaternary ammonium salt decomposition. The temperature of the last stage was from 427.30℃to 850℃, and the catalyst mass remained unchanged. The remaining mass percent except free and crystal water was 64.53%.

Elaborating in a simple manner, the C, H and N of catalyst would be disappeared after calcination at high temperature. However, the W and B remained in the form of WO3and B2O3. It points out that the unchanged mass of catalyst belonged to WO3and B2O3in the sixth stage. With the raw material feeding molar ratio of(Na2WO4):(H3BO3):(CTMA) 1:3:1, the molecular formula C19H44W3BNO18was gotten according to the element conservation principle, and the remaining percent was 64.24% theoretically.

Surprisingly, it was found that the mass of C, H, N, W, B and O calculated from TG analysis coincided with the molecular formula suggested by the getting from element conservation principle and the catalyst structure was deduced, shown in Fig.6, based on the reaction process.

2.2 Pre-experiment

There were some of the typical products of cyclohexene oxidation. The target product was cyclohexene oxide (1), generated by the epoxidation of the cyclohexene double bond[20]. The 1,2- cyclohexanediol (2) was the main side product, formed by the hydrolysis of the cyclohexene oxide. Another two side products, cyclohexene-1-one (3) and cyclohexene-1-ol (4) were the products of the allylic oxidation.

Owing to the quaternary ammonium containing hydrophilic and lipophilic group, the catalyst was lying in the middle layer between water phase and oil phase. The initially insoluble catalyst could form soluble active species under the effect of hydrogen peroxide and then transfer the active oxygen atom into cyclohexene, leading to cyclohexene oxide production. When hydrogen peroxide was used up, the catalyst returned to its original structure and precipitated from the reaction medium. Therefore, the catalyst possessed both the advantages of homogeneous catalysis and heterogeneous catalysis.

Several blank experiments were carried out with the purpose of meeting the requirement of industrial catalyst and determining the active component as well seen from Table 1. It showed that the conversion of cyclohexene was almost zero when boricacid or cetyltrimethylammonium chloride was used as catalyst, respectively. However, when sodium tungstate was used as catalyst the selectivity for cyclohexene oxide was higher than 50% in spite of only a small conversion of 3.53%. If the catalysts were a mixture, like Entry 4, 5, the conversion was obviously increased.

Table 1 Effect of different catalysts on catalytic performance

①A: Na2WO4, B: H3BO3, C: C16H33(CH3)3NCl, A+B/A+B+C: mixture.

Note: Reaction conditions : 4.10 g cyclohexene, 5.66 g 30%(mass) H2O2, 0.50 g catalyst, 0.53 g co-catalyst Na2HPO4, 60℃, 5h.

Among the three raw materials, we found that sodium tungstate had higher catalytic activity than the other two. We believed that the catalytic active component was tungsten on the basis of literatures. The reason why the result was not desired when sodium tungstate used directly as the catalyst was that sodium tungstate, a kind of inorganic metal salts, dissolved well in water phase but with much lower solubility in oil phase. Thus, the mass transfer problem between two phases still existed. Although the magnetic stirring forced them to mix, reaction was limited to the water phase which eventually affected the overall reaction rate. So we believed that a novel tungsten-based catalyst for cyclohexene epoxidation with hydrogen peroxide as an oxidant was prepared, which was not only efficient but also possessed the necessity with friendly phase transfer function, like Entry 6 in Table 1.

2.3 Effect of different W/B molar ratio on catalyst performance

With the determination of catalyst molecularformula above, a series of catalysts with different W/B molar ratio were prepared, with general formula C19H44WBNO(6+4n)whilerepresented the different W/B molar ratio. Here, the different W/B ratio catalyst was assigned as W/B=:1 for convenience except the catalyst without B.

The catalytic performance of catalysts with different W/B molar ratio was investigated for cyclohexene oxidation using 30%(mass) H2O2as an oxidant in a solvent-free systemand the results are given in Table 2.

Table 2 Effect of different W/B molar ratio on catalytic performance

Note: Reaction conditions: cyclohexene 4.10 g(0.05 mol), 30%(mass) H2O25.66 g(0.05 mol), catalyst 0.5 g, 60℃, 0.1 MPa.

It was found that the conversion of cyclohexene zigzag changed with the decrease of W/B molar ratio, but it did not change much overall. The highest conversion of cyclohexene was 16.0% with W/B=3:1 while the highest selectivity to cyclohexene oxide was 66.3% with W/B=12:1. Generally speaking, the yields of cyclohexene oxide with different W/B molar ratio catalyst were approximate within experimental and analytical error, and we believed that the W/B molar ratio of catalyst was not the main factor to influence the catalytic activity. Table 2 also shows that the catalysts exhibited a high selectivity to cyclohexene oxide, and the main by-product was cyclohexanediol. We believed that the selectivity to cyclohexene oxide could be promoted by reducing hydrolysis of the cyclohexene oxide in further study.

According to the reactionprocess, the bond W—O—B was formed by dehydration synthesis between the hydroxyl group of boric acid and tungsten acid peroxide. So the different amounts of boric acid may influence the number of W—O—B, and further influence catalyst structure. The effects of catalyst structure on catalyst activity mainly reflected on the diffusion of the reactants on its surface. However, the organic chain introduced by quaternary ammonium which would increase steric hindrance was on one side of W. That was to say, the active component W was almost at the end of the large molecules, hence the experimental results were approximate, no matter what the W/B was or even there was no B. To make full use of the raw materials, the W/B=3 was chosen.

2.4 Effect of lower temperature and longer reaction time on catalytic performance

Owing to the cyclohexene oxide hydrolysis and hydrogen peroxide decomposition, the yield of cyclohexene oxide would decrease with a higher temperature or longer reaction time. So the quaternary ammonium heteropolytungstoborate catalyst with W/B=3 was applied in a mild condition. The conversion of cyclohexene 51.24% and selectivity to cyclohexene oxide 69.18% were obtained under the reaction conditions of temperature 35℃, molar ratio of H2O2to cyclohexene 1:1, catalyst dosage 0.5 g and reaction time 24 h.

3 Conclusions

A novel tungsten-based catalyst was prepared for cyclohexene epoxidation, which exhibited a good catalytic activity. The yields of cyclohexene oxide with different W/B ratio were approximate within experimental error but varied with the different acidic intensity during preparation. TG characterization showed that the catalyst formula is C19H44W3BNO18(W/B=3), which is consistent with the ratio of raw materials used. The conversion of cyclohexene 51.2% and selectivity to cyclohexene oxide 69.2% were obtained under the conditions of reaction temperature 35℃, molar ratio of H2O2to cyclohexene 1:1, catalyst dosage 0.5 g and reaction time 24 h. It demonstrated a high catalytic activity and good selectivity to cyclohexene oxide without any solvents or environmental contamination.

Acknowledgements

This work was financially supported by the Zhejiang Province Natural Science Foundation (Y4080247), and Opening Project of Guangxi Key Laboratory of Petrochemical Resource Processing and Process Intensification Technology (K002).

[1] Hoegaerts D, Sels B F, Vosde D E, Verpoort F, Jacobs P A. Heterogeneous tungsten-based catalysts for the epoxidation of bulky olefins [J]., 2000, 60(3/4): 209-218.

[2] Keigo K, Mirki K, Kazuya Y, Shiro H, Noritaka M. Olefin epoxidation with hydrogen peroxide catalyzed by lacunary polyoxometalate [gamma-SiW10O34(H2O)2]4-[J].—, 2007, 13(2): 639-648.

[3] Hughes M D, Xu Y J, Jenkins P, Mcmorn P, Landon P, Enache D I, Carley A F, Attard G A, Hutchings G J, King F, Stitt E H, Johnston P, Griffin K, Kiely C J. Tunable gold catalysts for selective hydrocarbon oxidation under mild conditions [J]., 2005, 437(7062): 1132-1135.

[4] Li M Q, Jian X G, Yang Y Q. Lacunary Keggin polyoxotungstate as reaction-controlled phase-transfer catalyst for catalytic epoxidation of olefins [J]., 2004,22(8): 874-876.

[5] Maiti S K, Dinda S, Gharah N, Bhattacharyya R. Highly facile homogeneous epoxidation of olefins using oxo-diperoxo tungstate(Ⅵ) complex as catalyst, bicarbonate as co-catalyst and hydrogen peroxide as a terminal oxidant [J]., 2006, 30(3): 479-489.

[6] Prabhakar R, Morokuma K, Hill C L, Musaev D G. Insights into the mechanism of selective olefin epoxidation catalyzed by [gamma- (SiO4)W10O32H4]4-—a computational study [J]., 2006, 45(14): 5703-5709.

[7] Lane B S, Burgess K. Metal-catalyzed epoxidations of alkenes with hydrogen peroxide [J]., 2003, 103(7): 2457-2473.

[8] Ishii Y, Yamawaki K, Yoshida T, Ura T, Ogawa M. Oxidation of olefins and alcohols by preoxomolybdenum complex derived from tris(cetylpyridinium) 12-molybdophosphate and hydrogen-peroxide [J]., 1987, 52(9): 1868-1870.

[9] Fraile J M, Garcǐa J I, Mayoral J A, Salvatella L, Vispe E, Brown D R, Fuller G. Experimental and theoretical studies on structure-reactivity relationships of titanium-modified silicas in the hydrogen peroxide-promoted oxidation of cyclohexene [J]., 2003, 107(2): 519-526.

[10] Torres J C, Cardoso D, Pereira R. The influence of Si/Al and Si/Ti molar ratios of different [Ti, Al]-beta catalysts in the partial oxidation of cyclohexene with hydrogen peroxide [J]., 2010, 136(1/2/3): 97-105.

[11] Reinoso S, Dickman M H, Matei M F, Kortz U. 13-Tungstoborate stabilized by an organostannoxane hexamer [J]., 2007, 46(11): 4383-4385.

[12] Liu L L, Chen C C, Hu X F, Mohamood T, Ma W H, Lin J, Zhao J C. A role of ionic liquid as an activator for efficient olefin epoxidation catalyzed by polyoxometalate [J]., 2008, 32(2): 283-289.

[13] Venturello C, D’AloisioR. Quaternary ammonium tetrakis (diperoxotungsto) phosphates(3-) as a new class of catalysts for efficient alkene epoxidation with hydrogen-peroxide [J]., 1988, 53(7): 1553-1557.

[14] Kuznetsov A E, Geletii Y V, Hill C L, Morokuma K, Musaev D G. Mechanism of the divanadium-substituted polyoxotungstate [gamma-1,2-H2SiV2W10O40]4-catalyzed olefin epoxidation by H2O2: acomputational study [J]., 2009, 48(5): 1871- 1878.

[15] Sun Y, Xi Z W, Cao G Y. Epoxidation of olefins catalyzed by [pi-C5H5NC16H33]3[PW4O16] with molecular oxygen and a recyclable reductant 2-ethylanthrahydroquinone [J].:, 2001, 166(2): 219-224.

[16] Cai Z Y, Zhu M Q, Tang Y, Liu Y, Dai H, Chen X Z, He C H. Carbon-supported gold catalyst modified by doping with Ag for cyclohexene oxidation [J]., 2011, 236/237/238: 3046-3050.

[17] Tézé A, Michelon M, Hervé G. Syntheses and structures of the tungstoborate anions [J]., 1997, 36(4): 505-509.

[18] Zhao W, Ma B C, Hua H, Zhang Y S, Ding Y. Environmentally friendly and highly efficient alkenes epoxidation system consisting of [pi-C5H5N(CH2)11CH3]3PW4O32/H2O2/ethyl acetate/olefin [J]., 2008, 9(14): 2455-2459.

[19] Fu N, Lu G X. Hydrogen evolution over heteropoly blue-sensitized Pt/TiO2under visible light irradiation [J]., 2009,127(3/4): 319-322.

[20] Elías V R, Crivello M E, Herrero E R, Casuscelli S G, Eimer G A. Synthesis of titanium-containing mesoporous silicas as catalysts for cyclohexene epoxidation [J]., 2009, 48(20): 9076-9082.

新型钨基催化剂上环己烯环氧化反应

胡红定1，朱明乔1，Umsa JAMEEL1，童张法2

（1浙江大学化学工程与生物工程学院，浙江杭州310027；2广西大学化学化工学院，广西石化加工及过程强化技术重点实验室，广西 南宁 530004）

烯烃环氧化在现代化学中具有重要的学术和工业价值，但采用绿色氧化剂过氧化氢仍面临传质的挑战。采用过氧缩合法制备了一种新型钨基催化剂，用于过氧化氢与环己烯环氧化研究，环己烯的转化率为51.2%，环氧环己烷的选择性为69.2%。结果显示该催化剂催化活性高，环氧环己烷的选择性好，反应条件温和，工艺无溶剂且绿色环保。

环己烯；含钨催化剂；环氧化；相转移；催化

10.11949/j.issn.0438-1157.20150679

TQ 231.4；O 643.32

A

0438—1157（2015）08—3007—07

朱明乔。

胡红定（1989—），女，硕士研究生。

浙江省自然科学基金项目（Y4080247）；广西石化资源加工及过程强化技术重点实验室开放课题基金项目（K002）。

2015-05-25.

ZHU Mingqiao, zhumingqiao@zju.edu.cn

supported by the Natural Science Foundation of Zhejiang Province(Y4080247) and the Opening Project of Guangxi Key Laboratory of Petrochemical Resource Processing and Process Intensification Technology (K002).

2015-05-25收到初稿，2015-06-10收到修改稿。