SciELO - Scientific Electronic Library Online

 
vol.15 issue2SELECTIVE HYDROGENATION OF CINNAMALDEHYDE WITH Pt AND Pt-Fe CATALYSTS: EFFECTS OF THE SUPPORTHydroisomerization of Ethylbenzene on Mordenite-Based Bifunctional Catalysts with Different Platinum Contents author indexsubject indexarticles search
Home Pagealphabetic serial listing  

Services on Demand

Journal

Article

Indicators

Related links

Share


Brazilian Journal of Chemical Engineering

Print version ISSN 0104-6632On-line version ISSN 1678-4383

Braz. J. Chem. Eng. vol. 15 no. 2 São Paulo June 1998

https://doi.org/10.1590/S0104-66321998000200008 

Epoxidation of Cyclohexene on Heterogenized Molybdenum Compounds

 

E.F.M. Barradas, A.R. Cestari, C. Airoldi and R. Buffon*
Instituto de Química - Unicamp, P.O. Box 6154, 13083-970 Campinas - SP, Brazil
fax: 019 788-3023 e-mail: rbuffon@iqm.unicamp.br

(Received: ; Accepted: March 4, 1998)

 

Abstract - Heterogenization of molybdenum species, starting either with Mo(CO)6 or MoO2(acac)2, on the surface of functionalized silicas bearing one (Si-Et1) or two (Si-Et2) ethylenediamine ligands was studied. The resulting systems are active in the catalytic epoxidation of cyclohexene by tert-butyl hydroperoxide. Using fresh catalysts, Si-Et2 results in higher selectivity, regardless of the Mo precursor. However, MoO2(acac)2-based systems are far more active. Formation of diols was never detected. Although XPS analyses point mainly to Mo(VI) species on the surface of all systems, UV-vis spectra suggest different ligands in their coordination sphere.
Keywords: Epoxidation, molybdenum, functionalized silica.

 

 

INTRODUCTION

Soluble Mo(VI) compounds are the most versatile catalysts for the epoxidation of olefins (Parshall and Ittel, 1992). Homogeneous catalysis, however, is plagued with several drawbacks, warranting a search for supported molybdenum catalysts. Research done so far has focused on the use of organic polymers as supports. Preparation of such polymer-supported catalysts has usually been based on anion exchange (Sobczak and Ziólkowski, 1978), cation exchange (Ivanov et al. 1979; Boeva et al. 1984), and chelating ion exchange resins (Bhaduri and Khwaja, 1983; Yokoyama et al. 1985; Sherrington and Simpson, 1991). Of these catalytic systems, the one based on polybenzoimidazol (PBI) showed the best performance and stability (Miller and Sherrington, 1995,a,b). Nevertheless, several drawbacks such as the following have limited the utilization of these systems: i) the instability of the catalyst due to molybdenum leaching; ii) the thermo-oxidative instability of the polymer under reaction conditions; and iii) the poor mechanical properties of the organic support. Therefore, utilization of an organic functionalized silica might be a means of by-passing such problems, since functionalized silicas bearing organic groups such as amine (Allum et al. 1975; Gonçalves and Airoldi, 1989), acetylacetone (Espínola et al. 1992), etc. may be used as coordinating supports for molybdenum complexes, resulting in systems suitable for the epoxidation of olefins.

Interestingly, both zero- and hexavalent molybdenum compounds, viz. Mo(CO)6 and MoO2(acac)2, show similar catalytic activity and are considered to be precursors of the actual catalyst. In view of these facts, the aim of this work was to study heterogenization of Mo(CO)6 and MoO2(acac)2 on functionalized silicas bearing ethylenediamine ligands. Cychohexene was used as the substrate in the catalytic tests since internal olefins are more reactive towards epoxidation (Parshall and Ittel, 1992).

 

EXPERIMENTAL

The organofunctionalized silicas, bearing one (Si-Et1) or two (Si-Et2) ethylenediamine groups, were prepared via a sol-gel process, according to the literature (Cestari et al. 1996), from tetraethoxysilane and 3-trimethoxysilylpropylethylenediamine in an acidic medium. Reaction of Si-Et1 with glutaraldehyde, followed by reduction, produced Si-Et2. Nitrogen loading was determined by the Kjeldahl method.

Catalyst Preparation

The Mo/support catalysts were prepared by addition of a toluene solution of the molybdenum compound to the support, followed by heating under reflux for 72 h (MoO2(acac)2) or irradiation (Hg lamp, 125 W) for 30 min (Mo(CO)6). After the solution was filtered off, the catalysts were washed with toluene and dried in vacuo. Molybdenum loadings were determined using X-ray fluorescence.

Characterization

Diffuse reflectance infrared spectra were recorded on a Perkin-Elmer FT-IR spectrometer. UV-vis were obtained on a CARY 5 G spectrometer using pressed samples. XPS spectra were obtained with a Perkin-Elmer (PHI-1257) spectrometer, using a standard X-ray source with Mg as the anode (hn = 1253.6 eV), operating at 15 kV and 200 W. The binding energy values were corrected by fixing the Si 2p band at 103.5 eV.

Catalytic Tests

Epoxidation of cyclohexene (Riedel) was carried out in argon, in the liquid phase in a glass batch reactor, using tert-butyl hydroperoxide (TBHP) as the oxidant. In a typical experiment, 275 mg of catalyst (0.06 mmol of Mo), 7.0 ml of cyclohexene, 1.4 ml of a solution of tert-butyl hydroperoxide (86 wt.% in cyclohexane - 11.5 mmol), 0.5 ml bromobenzene (internal standard), and 0.2 ml 1,2-dichloroethane were stirred under reflux. The reaction was monitored by GC analysis of the liquid phase (HP 5890 gas chromatograph, equipped with an HP 1 capillary column and a flame ionization detector). Conversions are given as a function of the tert-butyl hydroperoxide consumption.

 

RESULTS AND DISCUSSION

The organofunctionalized silicas employed in this work can be formulated as

a7form1.JPG (16696 bytes)

When these materials were allowed to react with the molybdenum compounds, a color change from light yellow to a greenish tonality was observed. The maximum molybdenum loading, 0.25 mmol/g of support, might be due to the low specific area of these systems (ca. 30 m2/g). It should be noted that the increase in the ethylenediamine/silica ratio does not correspond to a net increase in the nitrogen content. Therefore, the molybdenum loadings on both silicas are nearly the same (Table 1).

FT-IR Spectroscopic Studies

According to the IR spectra of Mo(CO)6/support, all CO ligands were released during the surface reaction. In the case of MoO2(acac)2, it appeared that acetylacetone had also been replaced. Since all impregnation reactions were performed in air, the formation of oxo species was to be expected. However, in the region between 1000 and 850 cm-1, characteristic of n Mo=O, all spectra were similar (Si-Et1, SiEt-2, Si-Et1/Mo(CO)6, Si-Et1/MoO2(acac)2, Si-Et2/Mo(CO)6, and Si-Et2/MoO2(acac)2). The only exception was a sample of Si-Et1/MoO2(acac)2 that showed a shoulder at ca. 900 cm-1.

XPS Analyses

Analyses of the Mo 3d5/2 orbital energy (~ 232.4 eV) of both Si-Et1/Mo(CO)6 and Si-Et2/MoO2(acac)2 clearly show the presence of Mo(VI) on the surface. However, the line widths of these spectra (Fig. 1a), higher than 1.5 eV, suggest the existence of molybdenum species in oxidation states other than Mo(VI) and/or in different chemical environments. Interestingly, when the MoO2(acac)2 sample was treated with TBHP and 1,2-dichloroethane, its spectrum did not change; in the case of the Mo(CO)6-based sample only Mo(VI) species were observed (Fig. 1b).

 

Table 1: Catalyst composition

Catalyst Molybdenum loading
(mmol/g)
N/Mo molar ratio
chemical analysisa XPS
Si-Et1/Mo(CO)6 0.22 2.8 2.4
Si-Et1/MoO2(acac)2 0.25 2.5 n.d.b
Si-Et2/Mo(CO)6 0.21 3.0 n.d.
Si-Et2/MoO2(acac)2 0.25 2.5 1.9c

a Kjeldahl method (N) and X-ray fluorescence (Mo); b not determined; c poor spectrum.

a7fig1.JPG (41257 bytes)

Binding energy (eV)

Binding energy (eV)

Figure 1: XPS spectra of Si-Et1/Mo(CO)6: a) fresh sample; b) sample after interaction with tert-butyl hydroperoxide and 1,2-dichloroethane.

 

UV-Vis Spectroscopic Studies

The UV-vis spectrum of Si-Et1/MoO2(acac)2 (Fig. 2b) shows an absorption band at ca. 330 nm, which can not be assigned either to the support (Fig. 2a) or to MoO2(acac)2. The lack of this band (which is probably due to a ligand-metal charge transfer) in the spectrum of Si-Et1/Mo(CO)6 is the only spectroscopic evidence of the existence of different species when the molybdenum precursor is changed.

A band between 500 and 600 nm, as attributed to the Mo(V) species by Garbowski and Praliaud (1988), was not observed in any of our systems.

Catalytic Activity

Preliminary experiments were performed under the standard conditions described in the literature (Miller and Sherrington, 1995,a) in order to evaluate the effects of 1,2-dichloroethane and/or of previous activation (treatment with TBHP and 1,2-dichloroethane under reflux) on the catalytic activity of Si-Et1/MoO2(acac)2 in the epoxidation of cyclohexene. It turned out that previous activation did not lead to an increase in catalytic activity (Table 2). However, the addition of 1,2-dichloroethane (DCE), in very small amounts ([DCE]:[Mo] = 0.04) appeared to substantially affect the reaction rate. Therefore, all subsequent catalytic experiments were carried out with DCE and without previous activation.

Table 3 presents the results obtained with the Mo(CO)6 and MoO2(acac)2 precursors, either supported on Si-Et1 and Si-Et2 or in homogeneous solution, along with some data from the literature. It appeared that the heterogenization of Mo(CO)6 and MoO2(acac)2 on the organofunctionalized silicas caused a decrease in the cyclohexene epoxidation rate: within 4 h reaction time, TBHP reached a maximum conversion of ca. 72%, corresponding to a turnover number ([cyclohexene epoxide]:[Mo]) of ca. 140.

In preliminary recycling experiments with Si-Et1/MoO2(acac)2, Mo leaching was not observed. Catalytic activity, however, decreased slightly.

Epoxidation of cyclohexene over molybdenum species heterogenized on functionalized silicas, as studied in this work, can be considered to be highly selective. The only by-products observed were cyclohexene and chlorocyclohexanol. The latter probably results from a reaction between epoxide and DCE.

Figure 2: UV-vis spectra of Si-Et1: a) support; b) sample a after reaction with MoO2(acac)2 ; c) sample a after reaction with Mo(CO)6.

 

Table 2: Preliminary catalytic tests with the Si-Et1/MoO2(acac)2 system in the epoxidation of cyclohexenea

Conditions conversion %
30 min 90 min 240 min
Without activation/without DCE 4.3 8.9 15.3
Without activation/with DCE (0.2 mL)   15.9 45.2 71.5
Without activation/with DCE (1.0 mL) traces   6.1 10.9
With activation/with DCE (0.2 mL) 17.1 32.9 56.9

a Using fresh catalyst samples; [TBHP]:[Mo] = 192.

 

Table 3: Effects of the support on the catalytic activity of MoO2(acac)2 and Mo(CO)6

Catalyst precursora conversion %
20 min 30 min 60 min 90 min 120 min 240 min
Si-Et1/MoO2(acac)2   15.9   45.2   71.5
Si-Et1/Mo(CO)6   20.0   28.0   34.7
Si-Et2/MoO2(acac)2   24.4   59.5   72.1
Si-Et2/Mo(CO)6   13.1   n.d.   32.1
MoO2(acac)2 73.9   88.7   93.1 100
MoO2(acac)2 b 90.2   92.2   95.9  
Mo(CO)6 64.8   93.8   97.5 100
PBI/MoO2(acac)2 b 91.0   98.5   100  

a Using fresh catalyst samples; [TBHP]:[Mo] = 192. b Miller and Sherrington (1995,a); PBI = polybenzimidazole.

 

Although the MoO2(acac)2-based systems were far more active (Table 3), the least and the most selective systems were based on Mo(CO)6: Si-Et1/Mo(CO)6 was the least selective of the systems studied, yielding 9% by-products; Si-Et2/Mo(CO)6 showed the highest selectivity (~100% epoxide). Formation of diols was never observed.

 

CONCLUSIONS

Heterogenization of molybdenum compounds on the surface of organofunctionalized silicas bearing one or two ethylenediamine ligands results in catalysts which are active in the epoxidation of cyclohexene with tert-butyl hydroperoxide. Although these systems show lower catalytic activity than those reported for analogous heterogeneous systems, molybdenum does not leach from the surface, allowing the catalyst to be recycled. The differences in activity and selectivity observed for the different systems analysed in this work can be ascribed to the existence of molybdenum in diverse chemical environments which originate from the different molybdenum precursors used, as detected by UV-vis spectroscopy.

 

ACKOWLEDGEMENTS

The financial support received from FAPESP and CNPq is gratefully acknowledged. The authors also thank Carlos André Perez (NUCAT-PEQ-COPPE-UFRJ) for the XPS analyses.

 

REFERENCES

Allum, K. G.; Hancock, R. D.; Howell, I. V.; McKenzie, S.; Pitkethly, R. C. and Robinson, P. J., Supported Transition Metal Complexes. II. Silica as the Support, Journal of Organometallic Chemistry, 87, 203 (1975).         [ Links ]

Bhaduri, S. and Khwaja, H., Polymer-Supported Complexes. Part 3. Synthesis of a Polystyrene-Anchored Molybdenum(V) Dithiocarbamate-Derivative and its Applications in Reactions Involving t-Butyl Hydroperoxide, Journal of Chemical Society, Dalton Transactions, No 3, 415 (1983).         [ Links ]

Boeva, R.; Kotov, S. and Jordanov, N. I., Epoxidation of Propylene with tert-Butyl Hydroperoxide in the Presence of Heterogeneous Polymeric Catalysts, Reaction Kinetics and Catalysis Letters, 24, No 3-4, 239 (1984).         [ Links ]

Brock, T. and Sherrington, D. C., Preparation of Spherical Polybenzimidazole Particulates using a Non-Aqueous Suspension Methodology, Polymer, 33, No 8, 1773 (1992).         [ Links ]

Cestari, A. R.; Bruns, R. E. and Airoldi, C., A Fractional Factorial Design Applied to Organofunctionalized Silicas for Adsorption Optimization, Colloids and Surfaces. A.: Physicochemical and Engineering Aspects, 117, No 1-2, 7 (1996).         [ Links ]

Espínola, J. G. P.; Freitas, J. M. P.; Oliveira, S. F. and Airoldi, C., Chemisorption of some Divalent Cations on Silica Modified Through the Reaction of Pentane-2,4-dione with 3-propylethylenediamine Silica Gel, Colloids and Surfaces, 68, No 4, 261 (1992).         [ Links ]

Garbowski, E. and Praliaud, H., in Les Techniques Physiques d’Étude des Catalyseurs, Imélik, B. and Védrine, J. C. , eds., Technip, Paris (1988).         [ Links ]

Gonçalves, A. S. and Airoldi, C., Some Features of 2- and 4-aminopyridines Anchored on Silica Gel: Properties and Chemisorption of Metal(II) Chlorides in Non-Aqueous Solutions, Polyhedron, 8, No 24, 2901 (1989).         [ Links ]

Ivanov, S., Boeva, R. and Tanielyau, S., Catalytic Epoxidation of Propylene with tert-Butyl Hydroperoxide in the Presence of Modified Cation-Exchange Resin" Amberlite" IRC-50, Journal of Catalysis, 56, No 2, 150 (1979).         [ Links ]

Miller, M. M. and Sherrington, D. C., Alkene Epoxidation Catalysed by Mo(VI) Supported on Imidazole-Containing Polymers. I. Synthesis, Characterisation, and Activity of Catalysts in the Epoxidation of Cyclohexene, Journal of Catalysis, 152, No 2, 368 (1995a).         [ Links ]

Miller, M. M. and Sherrington, D. C., Alkene Epoxidation Catalysed by Mo(VI) Supported on Imidazole-containing Polymers. II. Recycling of Polybenzimidazole-Supported Mo(VI) in the Epoxidation of Cyclohexene, Journal of Catalysis, 152, No 2, 377 (1995b).         [ Links ]

Parshall, G. W. and Ittel, S. D., Homogeneous Catalysis: The Applications and Chemistry of Catalysis by Soluble Transition Metal Complexes. Wiley, New York (1992).         [ Links ]

Sherrington, D. C. and Simpson, S., Polymer-Supported Mo and V Cyclohexene Epoxidation Catalysts: Activation, Activity, and Stability, Journal of Catalysis, 131, No 1, 115 (1991).         [ Links ]

Sobczak, J. and Ziólkowski, J., The Molybdenum(V) Complexes as the Homogeneous and Heterogenized Catalysts in Epoxidation Reactions of Olefins with the Organic Hydroperoxides, Journal of Molecular Catalysis, 3, No 1-3, 165 (1978).         [ Links ]

Yokoyama, T.; Nishizawa, M.; Kimura, T. and Suzuki, T. M., Catalytic Epoxidation of Olefins with tert-Butyl Hydroperoxide in the Presence of Polymer-Supported Vanadium(V) and Molybdenum(VI) Complexes, Bulletin of the Chemical Society of Japan., 58, No 11, 3271 (1985).         [ Links ]

 

* To whom correspondence should be addressed.

Creative Commons License All the contents of this journal, except where otherwise noted, is licensed under a Creative Commons Attribution License