Microwave-assisted oxidation of benzyl alcohols using supported cobalt based nanomaterials under mild reaction conditions

2.1 Ball-milled assisted preparation of cobalt oxide NPs

In a typical synthesis of ball-milled materials, 0.2 g Al-SBA-15 (Si/Al molar ratio ~30) support was ground with the adequate amount of cobalt precursor (CoCl₂.6H₂O 98% Sigma-Aldrich) to reach a theoretical 0.5 wt% Co loading in a Retsch PM-100 planetary ball mill, using a 125 ml reaction chamber and 18 10 mm stainless steel balls, in a similar way to that previously reported [3]. Optimized milling conditions were 10 min at 350 rpm. The resulting materials (Co/Al-SBA-15) were then thoroughly washed with 50 ml water and subsequently calcined at 400°C under air for 4 h. Co content of the catalysts was estimated by energy-dispersive X-ray spectroscopy. We note that Al-MCM-41 was not selected as support in BM experiments due to its reduced stability under the BM preparation conditions as compared to Al-SBA-15. For comparison purposes, 1%, 5%, 10% and 20% Co/Al-SBA-15 materials were prepared in an analogous way in order to compare the activities in the investigated reactions.

2.2 Materials characterization

Powder X-ray diffraction patterns (XRD) were recorded on a Bruker AXS diffractometer with CuKα (λ=1.5418 Å) in the 5–85° 2θ range, using a 0.01° step size and a counting time per step of 4 s.

Nitrogen physisorption was measured with a Micromeritics instrument model ASAP 2000 at -196°C. The samples were outgassed for 24 h at 100°C under vacuum (p<10^-2 Pa) and subsequently analyzed. The linear part of the Brunauer-Emmett-Teller equation (relative pressure between 0.05 and 0.3) was used for the determination of the specific surface area. The pore size distribution was calculated from the adsorption branch of the N₂ physisorption isotherms and the BarretJoyner-Halenda (BJH) formula. The cumulative mesopore volume V_BJH was obtained from the pore size distribution curve.

Transmission electron microscopy (TEM) micrographs were recorded on a FEI Tecnai G² fitted with a CCD camera for ease and speed of use. The resolution is around 0.4 nm. Samples were suspended in ethanol and deposited straight away on a copper grid prior to analysis.

The metal content in the materials was also determined using inductively coupled plasma (ICP) in a Philips PU 70000 sequential spectrometer equipped with an Echelle monochromator (0.0075 nm resolution). Samples were digested in HNO₃ and subsequently analyzed by ICP.

2.3 MW-assisted reactions

In a typical reaction, 0.2 ml benzyl alcohol (≥99.0%, Sigma-Aldrich), 2 ml solvent [acetonitrile ≥99.9% (Sigma-Aldrich), ethyl acetate 99.8% (Sigma-Aldrich), toluene 99.9% (Sigma-Aldrich) or dimethylformamide (DMF) 99.8% (Sigma-Aldrich)], 0.3 ml H₂O₂ (50% v:v) and 0.05 g catalyst were added to a pyrex vial and microwaved in a pressure-controlled CEMDiscover MW reactor for a period of time, typically between 1 and 5 min at 300 W (70°C–80°C, maximum temperature reached) under continuous stirring. Samples were then withdrawn from the reaction mixture and analyzed by gas chromatography (GC) and GC/MS Agilent 6890 N fitted with a capillary column HP5 (30 m×0.32 mm×0.25 µm) and a flame ionization detector (FID). The identity of the products was confirmed by GC-MS and ¹H and ¹³C NMR. The blank reaction showed the thermal effects in the reaction were negligible (no conversion was obtained in the systems under the investigated conditions after 48 h).

Response factors of the reaction products were determined with respect to the substrates from GC analysis using standard compounds in calibration mixtures of specified compositions. The MW method was generally power controlled (by an infra-red probe) where the samples were irradiated with the required power output (settings at maximum power, 300 W) to achieve different temperatures in the range of 70–80°C.

3.1 Effect of reaction conditions

Upon characterization of the active phase, preliminary experiments to optimize reaction conditions were initially conducted with benzyl alcohol as substrate. Conversions were generally around 25–30% of starting material, with a complete selectivity to benzaldehyde typically in 2–5 min under MW irradiation (70–80°C, maximum temperature reached). A maximum of 40–45% conversion could only be achieved at longer times of reaction (>1 h) where the reaction did not seem to progress to any significant extent, in good agreement with previous reports from the group [3, 5]. Importantly, the analogous conventionally heated process at 100°C for 48 h provided negligible conversion in the systems even at higher catalyst loadings, which highlights the importance of the MW-assisted protocol. Selectivity to products was found to be quantitative to the target product (benzaldehyde), with no observed over-oxidation to benzoic acid, which was the case of the utilization of the more active supported iron oxide analogous systems (60% conversion, 82% selectivity to benzaldehyde, Table 1).

Table 1

Microwave-assisted oxidation of alcohols using 0.5% Co/Al-SBA-15 catalysts.

Entry	Substrate	Conversion (mol%)	Selectivity product (mol%)
Blank (no catalyst)		<5	>99
Blank (Al-SBA-15)		<5	>99
0.5% Fe-Al-SBA-15		60	82
1		27	>99
2		23	>99
3		17	>99
4		27	>99
5		24	>99
6		30	>99
7		33	95
8		–	–
9		<5	>99

Reaction conditions: 2 mmol substrate, 0.3 ml H₂O₂ 50%, 0.05 g catalyst, 2 ml acetonitrile, 300 W (70–80°C, maximum temperatures reached), 5 min reaction.

3.2 Effect of the solvent

An investigation of a range of solvents in the MW-assisted process was conducted. Selected solvents ranged from nonpolar (toluene) to medium and polar (ethyl acetate, acetonitrile, DMF). The interaction of the solvent with MW irradiation (through the loss tangent) is in principle a critical factor to achieve a rapid and homogeneous heating of the reaction media [15]. This in turn leads to improved activities at reduced times of reaction. Results summarized in Figure 3 pointed out the optimum performance of DMF and acetonitrile as solvents in the reaction with respect to toluene and ethyl acetate. Interestingly, both solvents have been reported to generate peroxides under certain conditions [16], a fact which will support the relatively high conversion values for blank runs in the absence of catalyst (Table 1).

Figure 3

Solvent screening in the 0.5% Co/Al-SBA-15 catalyzed microwave-assisted oxidation of benzyl alcohol with hydrogen peroxide. Reaction conditions: 2 ml solvent, 0.2 ml benzyl alcohol, 0.3 ml H₂O₂ (50% v:v) and 0.050 g catalyst, microwaves, 300 W, 70–80°C, 5 min reaction.

3.3 Effect of the cobalt content

Having selected acetonitrile as the optimum solvent for the reaction, the effect of the cobalt content in the systems was investigated. Results showed a remarkable decrease in activity in the systems at increased Co contents (Figure 4). Although particle sizes could not be worked out for low loading materials (≤5 wt%) to establish a strict NP size/activity correlation, the observed reduced activities at increasing metal content have been previously attributed in similar chemistries to a significant agglomeration of NPs at higher metal loadings, in good agreement with previously reported results from the group [7, 17] and TEM images which clearly evidenced the presence of large Co₃O₄ clusters (Figure 2C).

Figure 4

Activities of Co-containing catalysts (with cobalt contents ranging from 0.5 to 20%) in the microwaveassisted oxidation of benzyl alcohol. Reaction conditions: 2 ml solvent, 0.2 ml benzyl alcohol, 0.3 ml H₂O₂ (50% v:v) and 0.050 g catalyst, microwaves, 300 W, 70–80°C, 5 min reaction.

3.4 Scope of the reaction

The scope of the reaction was subsequently investigated using different substrates. Results summarized in Table 1 show that the reaction was amenable to a range of substrates containing both electron-donating and electron-withdrawing substituents. The presence of electro-withdrawing substituents provided improved conversions in the systems, reaching a maximum of 33% to produce 4fluorobenzaldehyde (Table 1, entry 7). Conversions were generally low, but all reactions were carried out at temperatures in the 70–80°C range and after 5 min reaction.

For the sake of comparison, the conventionally heated oxidation of the best performing substrate (4fluorobenzyl alcohol) gave no conversion after 48 h reaction, even at increased temperatures of 100°C. However, the oxidation of more challenging secondary alcohols (entries 8 and 9) was not possible under the investigated conditions, with very low or trace amounts of ketone products after even 10–15 min reaction.

3.5 Catalyst reusability studies

The recyclability of the most active 0.5% Co/Al-SBA-15 catalyst was eventually studied under the investigated reaction conditions. The oxidation of 4-fluorobenzyl alcohol to 4fluorobenzaldehyde was selected as the test reaction, due to the higher activity observed for the cobalt nanocatalyst. Figure 5 clearly demonstrates both the conversion (black stripe bars, Figure 5) and selectivity (white bars, >95%) to 4-fluorobenzaldehyde as the main product of the low loaded cobalt material was maintained even after four uses in comparison with its initial catalytic activity and selectivity (first use). These findings support the recyclability of the material under the investigated reaction conditions. The structure and morphology of the catalyst was also preserved after four runs, as indicated by TEM and XRD (results not shown).

Figure 5

Reusability of 0.5% Co-Al-SBA-15 catalyst in the oxidation of 4-fluorobenzyl alcohol with H₂O₂. Black stripe and white bars represent conversion (<40%) and selectivity to 4fluorobenzaldehyde (>95%), respectively. Reaction conditions: 2 mmol 4-fluorobenzyl alcohol, 0.3 ml H₂O₂ 50% (v/v), 2 ml acetonitrile, 0.05 g catalyst, microwave irradiation, 300 W, 5 min reaction (each run).

3.6 Preliminary investigations of reaction kinetics

Following the observed moderate conversion and studies related to reaction conditions and catalyst reusability, a preliminary investigation on reaction kinetics was conducted in the particular case of the selective oxidation of 4-fluorobenzyl alcohol to 4fluorobenzaldehyde. Results summarized in Figure 6 show that the concentration of the benzyl alcohol derivative changes rapidly at the beginning of the reaction and becomes almost constant after a few min. This type of behavior is characteristic of reversible reactions. In this case, the equilibrium concentration of the reagent is around 66%.

Figure 6

Kinetic studies for the evolution of the concentration of reagent with time in the microwave-assisted selective oxidation of 4-fluorobenzyl alcohol using hydrogen peroxide.

Considering the reaction: A k1→ ←k2 B, the gradient equals k1+k2 and x_∞ is equal 100*K_eq/(1+K_eq); where K_eq is the constant of equilibrium. Therefore k1+k2=0.0108 and k1/k2=1.93. So k1=1.93k2 and 2.93k2=0.0108. The calculated gradient was 0.0108.

A perfect linear correlation between Ln x∞x∞-x (x_∞=65.9%) and time demonstrates that the reaction is of first order. The calculated reaction constants were found to be 0.0071 and 0.0037 s^-1 for k1 and k2, respectively. Importantly, a first order reaction with respect to the main reagent may imply that hydrogen peroxide follows zero order kinetics, meaning that it decomposes gradually under the reaction conditions, keeping atomic oxygen concentration fairly constant during all reaction times. Further detailed studies will follow in due course.