Mesoporous Niobium Oxyhydroxide Catalysts for Cyclohexene Epoxidation Reactions

Abstract

Mesoporous catalysts were synthesized from the precursor NbCl₅ and surfactant CTAB (cetyltrimethylammonium bromide), using different synthesis routes of, in order to obtain materials with different properties which are capable of promoting the epoxidation of cyclohexene. The materials were characterized by X ray diffractometry (XRD), thermogravimetry (TG), acidity via pyridine adsorption, Hammet titration and N₂ adsorption/desorption. The characterization data indicate that the calcination process of the catalysts was efficient for elimination of the surfactant, but it caused a collapse of the structure, causing a Brunauer Emmett Teller (BET) specific area decrease (ClNbS-600, 44 and ClNbS-AC-600, 64 m² g⁻¹). The catalysts that have not been calcined showed high BET specific areas (ClNbS 198 and ClNbS-AC 153 m² g⁻¹). Catalytic studies have shown that mild reaction conditions promote high conversion. The catalysts ClNbS and ClNbS-AC showed high conversions of cyclohexene, 50 and 84%, respectively, while the calcined materials showed low conversion (<30%). The epoxide formation was confirmed by nuclear magnetic resonance (NMR).

1. Introduction

The use of molecular sieves has been of great industrial interest, since they direct the reactions due to their defined porous structure, being able to promote a greater selectivity for the product of interest. Two classes of well-defined and widely studied porous materials are the microporous zeolites [1] and the mesoporous silicates known as Mobil Composition of Mater (MCM) [2]. Moreover, the incorporation of transition metals, such as those of group V, in these porous materials can generate materials with high catalytic activity and good selectivity [1]. The presence of mesopores allows compounds of higher kinetic volume to access the active sites deposited in the cavities, which is difficult when using supports with smaller pores, such as the microporous. Niobium compounds stand out among those of neighboring elements because of their bifunctional properties, since they may have high acidity and ability to form highly oxidizing species in the presence of H₂O₂ by the formation of peroxo (Nb-O-OH) surface groups [3]. This allows their oxides to be widely used in heterogeneous catalysis in several types of reactions [4]. In addition, the bifunctional character of niobium oxides has been explored in reactions that require dehydration/oxidation properties of the substrate [5]. The niobium oxides commonly used have low specific area and low mesoporosity [6], which diminishes their activity as catalysts in reactions of total or partial oxidation involving molecules with higher kinetic volume. Selective oxidation reactions are less thermodynamically favored than those that promote total oxidation, necessitating a suitable catalyst to direct the formation of the product of kinetic interest [7].

Epoxidation of olefins is a selective oxidation reaction, especially in the production of various raw materials, as well as in the synthesis of intermediates for the production of chemical and pharmaceutical products [8,9,10,11]. Epoxidation may occur via homogeneous catalysis, but the greatest disadvantage of this process is the recovery of the catalyst, increasing interest in the development of heterogeneous catalysts capable of performing this reaction [12]. In the literature there are several studies on the epoxidation of olefin model molecules, and V and Nb compounds have been widely used but supported on porous materials. Using porous materials Nogueira et al., obtained near 100% conversion of the toluene oxidation reactions [13]. Rohit et al., employing vanadium-titania catalysts obtained conversion of 94% for the epoxidation of a variety of alkenes with organic solvent extracted TBHP as the oxidant [14]. In another work, Gallo et al. employing catalysts based on niobium metallocenes deposited into mesoporous silica showed conversion of 58% and high selectivity to limonene epoxide >98% [15]. Works using niobium compounds are scarce, mainly dealing with synthetic compounds where their textural properties can be modulated, in order to obtain high activity in epoxidation reactions.

Thus, it is understood that it is of great technological interest to obtain catalysts based on niobium compounds that can be efficiently used in selective oxidation reactions of olefins, aiming at the production of compounds of high added value that are of interest to the petrochemical industry. Different niobium catalyst synthesis methods are mentioned in the literature [16,17,18,19,20] as well as studies on the activity of these materials [21]. Therefore, we propose the synthesis of niobium oxyhydroxide from NbCl₅ using a surfactant to increase the hydrophobicity of the catalysts in order to improve their interaction with the non-polar substrate, and we test these materials for their ability to realize the epoxidation of cyclohexene used as a model molecule.

2. Materials and Methods

2.1. Synthesis of Catalysts

The synthesis of ClNbS and ClNbS-600 catalysts was performed by dissolving 41.2 mmols of CTAB in 25 mL of 1-butanol (99% VETEC), 69 mL of 1-hexanol (99% VETEC) and 19 mL of deionized water in a beaker at 60 °C with constant stirring. After dissolution, 46.2 mmol of NbCl5 were added. After the total solubilization of the salt, NH₄OH (5 mol L⁻¹) was dripped to pH 7. The volume of the formed solution was completed to 1000 mL with deionized water at 60 °C and the system was left under constant magnetic stirring for two (2) days for aging to yield the hydrophobized niobium oxyhydroxide. The solid formed was then macerated in an agate mortar. Part of the macerated solid was subjected to a heat treatment, remaining at 600 °C for 3 h, following a heating ramp of 10 °C min⁻¹. The catalysts ClNbS and ClNbS-600, respectively, were thus obtained.

The ClNbS-AC and ClNbS-AC-600 catalysts were synthesized as described above, however after pH correction the formed solution was poured into a Bergof ^® BR100 stainless steel autoclave containing one teflon beaker; then 20 mL of deionized water were added. The system was left at 60 °C under constant magnetic stirring for 2 (two) days for aging to produce the hydrophobized niobium oxyhydroxide. The formed product was transferred to a beaker with 100 mL of deionized water and a white solid suspension formed on top. The suspension was washed with deionized water and oven dried at 60 °C for 12 h. The ClNbS-AC-600 catalyst underwent the same calcination process as ClNbS-600.

2.2. Catalysts Characterization

The textural properties of the materials were analyzed through a N₂ adsorption/desorption isotherm that was obtained at a temperature of 77 K using a Quantachrome Autosorb IQ₂ equipment and pore volumes were calculated based on the adsorption branch of the isotherm using the Barrett Joyner Halenda (BJH) method. The specific surface area value was obtained using the BET method. X ray diffraction analysis was performed using the SHIMADZU model XRD-7000 X ray diffractometer, equipped with a copper tube and graphite monochromator. Analyses were performed under 30 mA current and 30 kV voltage. The velocity used was 4 degrees min⁻¹ for scanning between the angles 10° < 2θ < 70° by applying time constants of 5 s per increment. Infrared spectra (FTIR) were obtained using a Perkin Elmer FTIR RXI spectrophotometer in the region of 4000 to 500 cm⁻¹ with a resolution of 4 cm⁻¹ and a mean scan of the signal equal to 64 scans. The thermal analyzes were performed on a Shimadzu-TGA50H thermocouple, up to a maximum temperature of 700 °C, in air atmosphere, under a flow of 100 mL min⁻¹ at a heating rate of 10 °C min⁻¹, using approximately 3 mg of sample. The acidic properties of the catalyst were determined by the pyridine adsorption method. In this test, 10 mg of catalyst was placed in crucibles inside a quartz tube in a furnace and heated to 100 °C, for 2 h, under air flow (80 mL min⁻¹). After the cleaning of the surface, the pyridine was then introduced, at 50 °C. After the adsorption step, the temperature was raised at 100 °C under air flow (80 mL min⁻¹) to remove the pyridine physisorbed on the catalyst surface. Discs of 1 cm of diameter were made under vacuum and 6 ton cm⁻². The infrared spectrum was acquired using a Spectrm RX, 64 scans were recorded, in a region of 1800–1400 cm⁻¹. Quantitative determination of the acid sites present in all catalysts was performed by Hammett titration as described by Zhao et al. [22].

2.3. Catalytic Tests

Cyclohexene epoxidation reactions were performed in the presence of cyclohexene (VETEC, 99%) 50% hydrogen peroxide (SYNTH) and acetonitrile (J.T. BAKER, 99%).

Batch type glass reactors with a capacity of 15 mL and autogenous pressure were used. The reuse for the best catalyst was studied.

The initial condition studied was adapted according to experiments indicated in the literature [23] using 20 mmol of cyclohexene, 2 mL of 50% hydrogen peroxide, 50 mg of catalyst in 10 mL of acetonitrile, at 25 °C for 1 h and constant stirring.

After the completion of each reaction, the catalyst was separated by centrifugation. Sample preparation consisted of the addition of 20 μL of heptane (VETEC, P.A) as internal standard (PI) in 0.98 mL of sample. They were then analyzed by gas chromatography with flame ionization detector (GC-FID), GC-Shimadzu, DB-5 ((5% -phenyl) -methylpolysiloxane column), 30 m × 0.32 mm × 0.25 μm. The parameters used were column temperature: 40 °C for 10 min; injection temperature: 200 °C with Split (1:25); injected volume: 0.4 μL; FID detector temperature: 220 °C; and entrainment gas N₂ with flow of 0.6 mL min⁻¹. The CLASS CR-10 software was used for data acquisition. The conversion was determined from a calibration curve with internal standard.

To determine the formation of the reaction products, solutions containing 300 μL of reaction were prepared in 300 μL of D₂O. These solutions were analyzed in a Bruker AVANCE DPX 200 Nuclear Magnetic Resonance Spectrometer. ¹H and ¹³C spectra were acquired, in which 16 and 128 scans were used, respectively. Analyzes were performed for the calibration curve, in which the preparation of the sample used the CDCl₃ as solvent.

3. Results and Discussion

3.1. Characterization of Catalysts

The results obtained through thermogravimetric analysis (Figure 1a) show that materials that were not calcined (ClNbS and ClNbS-AC) present three main mass loss events. The first two must be related to external and bulk hydroxylates of the catalysts. The mass loss above 350 °C is attributed to the decomposition of the surfactant group anchored to the surface of the material [24]. No significant mass loss events were observed for calcined catalysts (ClNbS-600 and ClNbS-AC-600); which means that the calcination temperature was sufficient to remove all surfactant and hydroxyls, leaving only Nb₂O₅. The X ray diffractograms (Figure 1b) showed higher crystallinity for the calcined catalysts, since those that were not thermally treated presented as totally amorphous. The catalysts ClNbS-600 and ClNbS-AC-600 presented a typical diffraction profile of niobium oxide (Nb₂O₅), consistent with the standard JCPDS 27-1003.

The N₂ adsorption/desorption analyzes were performed and are presented in Figure 2. The data presented in

Table 1. BET specific area values and pore volume by the BJH method for the synthesized catalysts.

Catalyst	BET Surface Area (m2 g−1)	Pore Volume (cm3 g−1)
ClNbS	198	0.17
ClNbS-600	44	0.10
ClNbS-AC	153	0.15
ClNbS-AC-600	64	0.09

show that the calcination process decreases the specific area in relation to non-calcined catalysts, is attributed to the occurrence of particle agglomeration and crystallization, reducing the likely porous structure of the material. Another relationship observed in the surface area data is in relation to the form of aging: the ClNbS catalyst presents a larger surface area than ClNbS-AC, which may be associated with larger organization of the ClNbS-AC pores, since there was a control pressure and temperature in the synthesis.

The isotherms of the catalysts ClNbS-AC and ClNbS-600 indicate the absence of pores, being type II, according to the classification suggested by IUPAC, typical of non-porous materials. The hysteresis formed by the desorption isotherm of these catalysts can be classified as H3, as it does not reach a saturation level and is reported for aggregate materials, forming pores of the crack type associated with a portion of interparticle pores [25,26].

In the case of the ClNbS and ClNbS-AC-600 samples, the profile of the type IV isotherm present in the same proportion of materials suggests the presence of mesopores, confirmed by the pore size distribution (not shown here). When reaching a saturation level in the adsorption of N_2(g), during the desorption, the ClNbS forms a hysteresis classified as H2, considered as bottle-type pores [26]. For the ClNbS-AC-600 the formed hysteresis is classified as H3. Table 1 shows the BET specific area values and pore volume of the materials.

The calcined catalysts were subjected to acidity analysis by adsorption of pyridine to qualitatively evaluate the type of acid sites present in these materials. For the non-calcined catalysts the analyzes were not carried out because the surfactant had adsorption bands in the same region where the pyridine adsorbed at the Lewis sites (1447 cm⁻¹) and adsorbed at the Bronsted sites (1550 cm⁻¹) were observed. The ClNbS-600 catalyst exhibits acidity Bronsted, while ClNbS-AC-600 catalyst presented Bronsted acid and Lewis acid sites (Figure 3), indicating that the aging process in an autoclave provides a difference in the type of acidic material. It is important to consider the acidity of the materials, since it is believed that the oxidant groups are formed in these acid sites by the decomposition of H₂O₂.

By Hammet titration the total amount of acid sites of the catalysts were determined and the mean acidity values were 2.02, 2.78, 1.80 and 2.60 mmol g⁻¹ for ClNbS, ClNbS-600, ClNbS-AC and ClNbS-AC-600, respectively. The calcined catalysts had a higher amount of acid sites, possibly due to unclogging of the pores after surfactant elimination. The catalysts that passed through the aging process in the autoclave presented a smaller amount of acid sites, because this treatment caused a better organization of the catalysts structure, reducing the imperfections and faults on the surface of the catalysts, where possibly acidic sites would reside.

3.2. Epoxidation Reactions

Figure 4a shows the preliminary catalytic results for the catalysts before and after calcination. A first analysis identifies that the materials that did not undergo calcination presented a greater cyclohexene conversion capacity.

The catalysts ClNbS-AC and ClNbS are niobium oxyhydroxides (NbO₂OH) and as proposed by Souza et al., these materials are capable of forming peroxo groups in the presence of H₂O₂. These groups are highly oxidizing, which would explain the better catalytic efficiency of the ClNbS-AC (80%) and ClNbS (50%) catalysts in the oxidation of cyclohexene [27]. The calcined catalysts do not present these surface hydroxyls for this reason they are not able to form peroxo groups, so their catalytic activity in the oxidation of cyclohexene are lower. The ClNbS-AC catalyst showed the highest cyclohexene conversion (84%) under the conditions studied. Ziolek et al., also analyzed the conversion of cyclohexene using niobium compounds as a catalyst under similar reaction conditions and obtained a maximum conversion of 58% [28].

The results of the catalyst reuse tests are presented in Figure 4b. A small conversion decrease in the cycles of reaction was observed, which may be related to loss of catalyst mass during the washes and the deactivation of the catalytic sites after some reactions, however the catalyst showed a certain stability, with considerable cyclohexene conversions (~60%) after two reuse tests. However, in the third reuse of the catalyst there is a marked drop in the conversion capacity. The catalyst was recovered after this test and it was observed that the hydrophobizing group had been leached from the surface after the third use. The presence of the group that makes the catalyst have greater affinity for the solvent where the substrate is dissolved [23]. Other forms of synthesis are being studied in order to maintain a more stable structure and increase its reuse capacity.

¹H and ¹³C spectra were obtained to prove epoxide formation by comparing the ¹³C chemical shifts of the experimental and simulated. The values of the chemical displacements are presented in

Table 2. Experimental and simulated chemical displacement values for epoxide carbon.

	Experimental Chemical Shifts (ppm)	Simulated Chemical Shifts (ppm)
C1	51.22	52.07
C2	25.36	24.50
C3	20.53	19.47

.

The ¹³C spectrum is shown in Figure 5. The presence of the acetonitrile solvent, the cyclohexene substrate and the major product, epoxide, are observed. Other cyclohexene oxidation products, such as cyclohexanone, were not formed as no ketone carbonyl signal was observed.

4. Conclusions

The results presented an efficient catalytic process using niobium oxyhydroxide synthesized from the precursor NbCl₅. The TG analysis showed the incorporation of CTAB in the structure of the material, as well as the loss of this group in the calcination process. The ClNbS and ClNbS-AC materials presented high BET specific area, with values of 198 and 153 m² g⁻¹, respectively. The calcination process resulted in a decrease of specific area and an increase in the acidity of the material due to the acid site increase and greater crystallinity presented for these materials.

The use of niobium oxyhydroxide with mesoporosity is rarely described in the scientific literature. The acidity generated by the hydroxyls can be replaced by oxidizing groups, i.e., peroxisome species which can be formed with compounds of the V group in the presence of hydrogen peroxide. The mesoporosity and formation of these oxidizing groups appears to provide important properties in partial oxidation reactions, such as epoxidation of olefins. The epoxidation reactions using cyclohexene, hydrogen peroxide showed good conversion results for the reaction conditions studied, especially ClNbS-AC, which presented 84% conversion. The NMR results showed the formation of the epoxide as the major product of the reaction. The results were extremely promising and innovative, given the efficiency of the catalysts in the reactions, as well as the fact that there have been no high conversion value reported in the literature for mild reaction conditions.