Selective Catalytic Oxidation of Benzyl Alcohol by MoO 2 Nanoparticles

: Selective oxidation of benzyl alcohol to benzaldehyde was carried out with MoO 2 nanoparticles (MoO 2 NPs). MoO 2 NPs were synthesized by two di ﬀ erent approaches and characterized by several techniques. The synthesis was done by a hydrothermal procedure using ethylenediamine and either Fe 2 O 3 or hydroquinone. In the latter case, an additional calcination step under N 2 was performed to eliminate passivating agents at the surface of the nanoparticles. The synthesized nanocatalysts showed similar catalytic properties, being e ﬃ cient catalysts in the oxidation of benzyl alcohol. High substrate conversion and product selectivity were achieved under all tested conditions. Studies were conducted using two di ﬀ erent oxidants: tert -butyl hydroperoxide and hydrogen peroxide, in our continuous e ﬀ ort to obtain more e ﬃ cient catalysts for more sustainable catalytic processes. When H 2 O 2 was used as the oxidant, 94% yield was achieved with 100% selectivity for benzaldehyde, which was a very promising result to undergo other studies with this system. Moreover, elucidate aspects of the reaction mechanism, study it was to conclude that the reaction undergoes, to some extent, through a radical mechanism with both oxidants.


Introduction
Molybdenum complexes are largely known for their good catalytic activity and have been widely used, especially in oxidation reactions, as homogeneous catalysts. However, heterogeneous catalysts present a much easier way of separation and can be easily recyclable. Nanotechnology has been a major breakthrough in recent times, allowing the possibility of producing materials with a higher surface area. Hybrid and non-hybrid metal-oxide nanoparticles have proven to be very active heterogeneous catalysts [1,2], while some can be highly selective in oxidation reactions [3]. Metal and metal oxide-containing nanostructures with different morphologies (nanowires, nanobelts, nanofibers, etc.) are known to exhibit better adjustable properties in fields such as gas sensors, electrodes, and optoelectronics than their conventional counterparts [4].
Partial oxidation of primary alcohols to aldehydes is of great interest worldwide because of its wide application in the perfume and food industries. However, it remains difficult to avoid over oxidation and specially to make them without using harmful solvents in the process. The oxidation of benzyl alcohol (BzOH) to benzaldehyde (BzCHO) is an important reaction because of the wide applications of the latter in industry [5]. Numerous methods have been developed to achieve the desired transformation with the use of a stoichiometric amount of oxidant. The oxidation of benzyl alcohol to benzaldehyde continues to be studied, since this reaction has a high reactivity, but presents selectivity issues yielding some by-products, which are well-known [6]. In this context, both homogeneous and heterogeneous transition metal-based catalysts have been developed for aerobic alcohol oxidations [7].
The liquid phase catalytic oxidation method is more common for this process than the gas phase oxidation, whereas the control in selectivity is generally a problem for the latter. In traditional oxidation processes, large amounts of toxic and volatile organic solvents and inorganic oxidants (CrO 3 , KMnO 4 , NaClO, MnO 2 , etc,) are extensively used [8,9]. Therefore, these oxidants are toxic, difficult to handle, and expensive. They are also less selective to the desired product and generate a larger amount of waste. From an environmental point of view, such conditions are mandatorily unwanted, while searching for more sustainable alternatives is highly desirable.
In recent years, nanocatalysts have received great attention in nanoscale heterogeneous catalysis with the advantage of practical recovery by separation of the solid catalyst from the medium. They also offer higher surface areas, low-coordinating sites, and surface vacancies compared to bulk heterogeneous systems.
In continuation of our efforts to investigate more sustainable processes for the catalytic oxidation of several organic substrates, we report in this work the oxidation of benzyl alcohol catalyzed with two different types of synthetized MoO 2 nanoparticles (Scheme 1).
Catalysts 2020, 10, x FOR PEER REVIEW 2 of 15 alcohol oxidations [7]. The liquid phase catalytic oxidation method is more common for this process than the gas phase oxidation, whereas the control in selectivity is generally a problem for the latter.
In traditional oxidation processes, large amounts of toxic and volatile organic solvents and inorganic oxidants (CrO3, KMnO4, NaClO, MnO2, etc,) are extensively used [8,9]. Therefore, these oxidants are toxic, difficult to handle, and expensive. They are also less selective to the desired product and generate a larger amount of waste. From an environmental point of view, such conditions are mandatorily unwanted, while searching for more sustainable alternatives is highly desirable. In recent years, nanocatalysts have received great attention in nanoscale heterogeneous catalysis with the advantage of practical recovery by separation of the solid catalyst from the medium. They also offer higher surface areas, low-coordinating sites, and surface vacancies compared to bulk heterogeneous systems.
In continuation of our efforts to investigate more sustainable processes for the catalytic oxidation of several organic substrates, we report in this work the oxidation of benzyl alcohol catalyzed with two different types of synthetized MoO2 nanoparticles (Scheme 1). Scheme 1. Schematic representation of the reactions occurring with the different catalysts reported in this work.
In this work, we synthesized different MoO2 nanoparticles and explored their catalytic activity in the oxidation of benzyl alcohol to benzaldehyde in a solventless process, preferably. For MoO2 synthesis, we used a hydrothermal method using ethylenediamine (EDA) as the reducing agent and Fe2O3 [10] or hydroquinone [11] as assisting agents. In the final product, the assisting agent remains in residual amounts at the surface of the material [11]. Fe2O3 nanoparticles have also been produced by a co-precipitation method [12,13] to understand the influence of iron oxide in catalytic studies.
The catalytic experiments were conducted by testing the temperature of the reaction, amount of oxidant, and the different types of nanoparticles, in accordance with the synthesis method. Several reaction conditions were screened. Two different oxidants were used, tert-butyl hydroperoxide (TBHP) and hydrogen peroxide (H2O2), and different amounts of them were added to the reactions to study the influence of the oxidant on the catalytic oxidation of benzyl alcohol. Different reaction temperatures and solvents were tested to assess the optimized reaction conditions for this system.

Characterization of the Catalyst
MoO2 NPs were synthesized by two different methods as reported previously [10,11,14,15]. This involved merging the solvothermal synthesis with thermal reduction using commercial MoO3 in the presence of a mixture of ethylenediamine (EDA) as the reducing agent and Fe2O3 (first method) or hydroquinone (second method) as assisting agents, respectively [10]. In the former case, synthesis with Fe2O3 working as an assisting agent has two roles: the first is to act as a nucleation seed for the growth of the MoO2 NPs; the second is a role as a regenerator of the oxidized EDA during the process. These MoO2 NPs will be denoted hereafter as MoO2-Fe2O3 NPs.
In the second method, the synthesis of MoO2 NPs was done by replacing Fe2O3 by hydroquinone, using a reported procedure [11]. However, since it was described that hydroquinone was retained at In this work, we synthesized different MoO 2 nanoparticles and explored their catalytic activity in the oxidation of benzyl alcohol to benzaldehyde in a solventless process, preferably. For MoO 2 synthesis, we used a hydrothermal method using ethylenediamine (EDA) as the reducing agent and Fe 2 O 3 [10] or hydroquinone [11] as assisting agents. In the final product, the assisting agent remains in residual amounts at the surface of the material [11]. Fe 2 O 3 nanoparticles have also been produced by a co-precipitation method [12,13] to understand the influence of iron oxide in catalytic studies.
The catalytic experiments were conducted by testing the temperature of the reaction, amount of oxidant, and the different types of nanoparticles, in accordance with the synthesis method. Several reaction conditions were screened. Two different oxidants were used, tert-butyl hydroperoxide (TBHP) and hydrogen peroxide (H 2 O 2 ), and different amounts of them were added to the reactions to study the influence of the oxidant on the catalytic oxidation of benzyl alcohol. Different reaction temperatures and solvents were tested to assess the optimized reaction conditions for this system.

Characterization of the Catalyst
MoO 2 NPs were synthesized by two different methods as reported previously [10,11,14,15]. This involved merging the solvothermal synthesis with thermal reduction using commercial MoO 3 in the presence of a mixture of ethylenediamine (EDA) as the reducing agent and Fe 2 O 3 (first method) or hydroquinone (second method) as assisting agents, respectively [10]. In the former case, synthesis with Fe 2 O 3 working as an assisting agent has two roles: the first is to act as a nucleation seed for the growth of the MoO 2 NPs; the second is a role as a regenerator of the oxidized EDA during the process. These MoO 2 NPs will be denoted hereafter as MoO 2 -Fe 2 O 3 NPs.
In the second method, the synthesis of MoO 2 NPs was done by replacing Fe 2 O 3 by hydroquinone, using a reported procedure [11]. However, since it was described that hydroquinone was retained at the surface of the NPs [14], the final product was calcined under nitrogen to prevent the oxidation of MoO 2 back to MoO 3 , in order to obtain MoO 2 NPs without any trace of (in)organic agents. These will be denoted afterward as MoO 2 . A synthesis procedure test without EDA and Fe 2 O 3 (or hydroquinone) was also performed without any success [14].
These MoO 2 NPs synthesized by the different protocols were characterized by powder X-ray diffraction (XRD), Scanning Electron Microscopy (SEM), and Fourier-Transform Infrared Spectroscopy (FTIR). The XRD powder pattern profile of the MoO 2 -Fe 2 O 3 NPs (Figure 1) displayed Bragg diffraction peaks corresponding to a typical pseudo-tetragonal rutile MoO 2 phase in agreement with other literature reports [10] (P63/mmc, JCPDS card, 50-0739, a = 2.838 Å, c = 4.720 Å). Analysis of the XRD powder pattern also showed that there was the presence of both Mo-and Fe-containing oxide phases. However, specifically concerning the Mo oxide, the pattern also showed that this was a pseudo-tetragonal rutile MoO 2 sample (i.e., without any other detectable Mo-containing phase) [14]. The average crystallite size of the NPs was 16.7 nm in the case of MoO 2 -Fe 2 O 3 , which agreed with the published data [11,14].
Catalysts 2020, 10, x FOR PEER REVIEW 3 of 15 the surface of the NPs [14], the final product was calcined under nitrogen to prevent the oxidation of MoO2 back to MoO3, in order to obtain MoO2 NPs without any trace of (in)organic agents. These will be denoted afterward as MoO2. A synthesis procedure test without EDA and Fe2O3 (or hydroquinone) was also performed without any success [14]. These MoO2 NPs synthesized by the different protocols were characterized by powder X-ray diffraction (XRD), Scanning Electron Microscopy (SEM), and Fourier-Transform Infrared Spectroscopy (FTIR). The XRD powder pattern profile of the MoO2-Fe2O3 NPs ( Figure 1) displayed Bragg diffraction peaks corresponding to a typical pseudo-tetragonal rutile MoO2 phase in agreement with other literature reports [10] (P63/mmc, JCPDS card, 50-0739, a = 2.838 Å, c = 4.720 Å). Analysis of the XRD powder pattern also showed that there was the presence of both Mo-and Fe-containing oxide phases. However, specifically concerning the Mo oxide, the pattern also showed that this was a pseudo-tetragonal rutile MoO2 sample (i.e., without any other detectable Mo-containing phase) [14]. The average crystallite size of the NPs was 16.7 nm in the case of MoO2-Fe2O3, which agreed with the published data [11,14]. For the (calcined) MoO2 NPs, the XRD powder pattern profile depicted in Figure 1 shows the Bragg diffraction peaks corresponding only to the pseudo-tetragonal rutile MoO2 phase [10] (P63/mmc, JCPDS card, 50-0739, a = 2.838 Å, c = 4.720 Å). The obtained X-ray diffraction pattern shows three reflections at 2θ = 36.73°, 53.49°, and 66.16° indexed to 100, 102, and 110 planes, respectively. The XRD powder diffraction pattern also showed a wide peak at 2θ = 23.5° that could be attributed to the silica from the support glass slide. The average crystallite size of the nanoparticles, estimated using the Scherrer equation, was 9.7 nm, approximately.
SEM microscopy was used to assess the morphology of both types of MoO2 nanoparticles. Figure  2 shows the SEM images from both MoO2 nanoparticles confirming that the structures were formed by aggregated ultrathin nanosheets, as reported in previous studies. The aggregated ultrathin nanosheets formed hierarchical uniform-sized tremella-like structures, displaying an almost spherical morphology. For the (calcined) MoO 2 NPs, the XRD powder pattern profile depicted in Figure 1 shows the Bragg diffraction peaks corresponding only to the pseudo-tetragonal rutile MoO 2 phase [10] (P63/mmc, JCPDS card, 50-0739, a = 2.838 Å, c = 4.720 Å). The obtained X-ray diffraction pattern shows three reflections at 2θ = 36.73 • , 53.49 • , and 66.16 • indexed to 100, 102, and 110 planes, respectively. The XRD powder diffraction pattern also showed a wide peak at 2θ = 23.5 • that could be attributed to the silica from the support glass slide. The average crystallite size of the nanoparticles, estimated using the Scherrer equation, was 9.7 nm, approximately.
SEM microscopy was used to assess the morphology of both types of MoO 2 nanoparticles. Figure 2 shows the SEM images from both MoO 2 nanoparticles confirming that the structures were formed by aggregated ultrathin nanosheets, as reported in previous studies. The aggregated ultrathin nanosheets formed hierarchical uniform-sized tremella-like structures, displaying an almost spherical morphology.
Fourier-transform infrared spectroscopy (FTIR) was used to characterize the calcined MoO 2 NPs, as shown in Figure 3. Given that ethylenediamine and hydroquinone were removed by a calcination step, this spectrum was simpler than when the calcination step was not carried out [11]. In this way, the FTIR spectrum of the MoO 2 NPs was typical of a Mo oxide, showing a broad band at 3417 cm −1 ascribed to the stretching modes of the hydroxyl groups at the surface of the MoO 2 nanoparticles. The absorption band at 1626 cm −1 was due to the O-H bending modes [16,17]. Other bands observed at 894 and 711 cm −1 could be ascribed to the vibration of the Mo-O-Mo bonds [18]. Fourier-transform infrared spectroscopy (FTIR) was used to characterize the calcined MoO2 NPs, as shown in Figure 3. Given that ethylenediamine and hydroquinone were removed by a calcination step, this spectrum was simpler than when the calcination step was not carried out [11]. In this way, the FTIR spectrum of the MoO2 NPs was typical of a Mo oxide, showing a broad band at 3417 cm −1 ascribed to the stretching modes of the hydroxyl groups at the surface of the MoO2 nanoparticles. The absorption band at 1626 cm −1 was due to the O-H bending modes [16,17]. Other bands observed at 894 and 711 cm −1 coulc be ascribed to the vibration of the Mo-O-Mo bonds [18].

Catalytic Tests and Optimization of Alcohol Oxidation Conditions
Liquid phase oxidation of benzyl alcohol (BzOH) to benzaldehyde (BzCHO) was selected as the model reaction to assess the performance of the catalysts herein reported and to find the optimized reaction conditions using the MoO2-Fe2O3 and MoO2 calcined nanoparticles (MoO2 NPs) as catalysts. First, the optimization of the reaction conditions was achieved by testing the conversion of BzOH

Catalytic Tests and Optimization of Alcohol Oxidation Conditions
Liquid phase oxidation of benzyl alcohol (BzOH) to benzaldehyde (BzCHO) was selected as the model reaction to assess the performance of the catalysts herein reported and to find the optimized reaction conditions using the MoO 2 -Fe 2 O 3 and MoO 2 calcined nanoparticles (MoO 2 NPs) as catalysts. First, the optimization of the reaction conditions was achieved by testing the conversion of BzOH into BzCHO under solvent-free conditions. Following that, a series of experiments was conducted to investigate crucial factors such as different types of solvents and the influence of the temperature on the reaction, namely the solvent acetonitrile at different temperatures, 353 or 383 K, and decane at 383 K. The catalysts were tested with different oxidants. Tert-butyl hydroperoxide (TBHP) and hydrogen peroxide (H 2 O 2 ) were used as oxygen donors and to evaluate the oxidant effects on the catalytic system's performance, the TBHP:substrate ratio was varied using 100, 150, and 200 mol%.
Blank runs (without the catalyst and in the presence of an oxidizing agent) gave virtually no conversion (<3%) of the starting material.
In all tested conditions using the studied catalysts, conversions were high and benzaldehyde was obtained as a major product, achieving excellent yields. As seen in Table 1, and as a general rule, in all cases, benzyl alcohol was converted to benzaldehyde with high selectivity. The only exception was when γ-Fe 2 O 3 was used as the catalyst for control purposes to understand the influence of Fe 2 O 3 in the MoO 2 -Fe 2 O 3 catalyst.  entries 1-3). These results showed that 150% of TBHP was the optimum amount of oxidant, reaching 95% conversion after 24 h of reaction time. The selectivity for benzaldehyde when the MoO 2 -Fe 2 O 3 catalyst was used with different ratios of oxidant/catalyst, did not suffer almost any change, reaching ca. 90%, and benzoic acid was the only by-product detected, although with very low selectivity. On the other hand, when the MoO 2 calcined catalyst was used, the achieved results were also very high, reaching 89-74% conversion ( Table 1, entries 5-7). The selectivity for the desired product, benzaldehyde, was also excellent, being 100% under the tested conditions. Aside from the study of different oxidant:substrate ratios, recyclability is also a crucial parameter to evaluate a heterogeneous catalyst. The recycling experiments were conducted using MoO 2 calcined nanoparticles as the catalyst, 150% of TBHP as the oxidant, and benzyl alcohol as the substrate in two successive runs (Table 1, entry 6). The catalyst was filtered after the first run, washed with dichloromethane, and dried prior to the recycling experiment. This recycling experiment was done with the MoO 2 calcined catalyst and 150% of TBHP, since it was the condition that presented higher selectivity. This reached 89% conversion and 100% selectivity for benzaldehyde, which was a very promising result. After the second cycle, 65% conversion was obtained and once again, the selectivity remained very high, achieving 100%. This result showed that this catalyst was very active for the oxidation of benzyl alcohol.
When assessing the catalytic performance of heterogeneous systems, it was important to evaluate the stability of the catalyst in terms of active center leaching and to ensure that the catalyst was robust and heterogeneous in nature.
To test active site leaching, the oxidation of benzyl alcohol was performed using the MoO 2 -Fe 2 O 3 nanoparticles as the catalyst and 200 mol% TBHP. During a normal reaction, the catalyst was removed by hot filtration after 2 h, and the reaction was followed until the end of the 24 h period. After filtering upon 2 h (at 22% conversion), the reaction was continued only with the homogeneous part. At the end of the 24 h period, the reaction reached only 27% conversion ( This clearly showed that the reaction almost stopped when the heterogeneous catalyst was filtered off from the reaction slurry, and therefore there was little or no leaching to the homogeneous phase. Thus, it was possible to conclude that the catalysts were truly heterogeneous in nature. Analyzing Figure 4, which shows the curves for the use of different amounts of TBHP oxidant, it was possible to observe that the kinetics at the beginning of the reaction was not affected by the different amounts of added TBHP from 100-200 mol%. However, they presented large differences after a 4 h reaction time, particularly for the catalytic reaction done with 150% of TBHP, where it clearly detached from the remaining reactions with different oxidant:substrate ratios.  The reason why the 200 mol% ratio was not the highest performing has to do with the fact that such a high amount of TBHP yielded larger amounts of t-butanol, which competes with TBHP for the active sites [19]. Therefore, using a lower ratio of TBHP at 150 mol% (relative to the substrate) was found to be an optimal balance for the catalyst performance, while the results reported here matched those in the literature for related systems [20].

Mechanistic Study
To determine whether the reaction proceeded through radical intermediates, their presence in the reaction mixture was analyzed. To a typical reaction mixture, a substrate equimolar amount (4.32 mmol) of the radical scavenger 2,6-di-tert-butyl-4-methylphenol was added and the reaction was followed for 24 h. This was tested for both the MoO2-Fe2O3 and MoO2 calcined catalysts. The reason why the 200 mol% ratio was not the highest performing has to do with the fact that such a high amount of TBHP yielded larger amounts of t-butanol, which competes with TBHP for the active sites [19]. Therefore, using a lower ratio of TBHP at 150 mol% (relative to the substrate) was found to be an optimal balance for the catalyst performance, while the results reported here matched those in the literature for related systems [20].

Mechanistic Study
To determine whether the reaction proceeded through radical intermediates, their presence in the reaction mixture was analyzed. To a typical reaction mixture, a substrate equimolar amount (4.32 mmol) of the radical scavenger 2,6-di-tert-butyl-4-methylphenol was added and the reaction was followed for 24 h. This was tested for both the MoO 2 -Fe 2 O 3 and MoO 2 calcined catalysts.
Analyzing the obtained results, it was possible to confirm that only 33% and 39% conversion after the 24 h reaction was obtained, respectively ( Table 2, entries 2 and 4). The results from these tests were compared with those obtained previously and found to be lower in comparison with the first experiments with 74% ( Table 2,   From the above results, it could be rationalized that the radical scavenger (2,6-di-tert-butyl-4-methylphenol) would compete for active sites at the surface of the catalysts, although this was hard to conceptualize. Given its structure with the bulky t-butyl groups ortho to the phenol group, it would be extremely difficult for this molecule to approach the surface and bind strongly. In this way, the results showed that there was good evidence that the mechanism involves a radical path. However, since a considerable level activity was also detected, it also demonstrated that there was an alternative path that was not radical. It was also possible to observe that the kinetics of the reactions presented differences from the previous ones (Figures 4 and 5), showing that the reaction did not proceed at the same rate, as expected. When the MoO 2 calcined catalyst was used with the radical scavenger, the reaction kinetics showed analogous trends, as observed for MoO 2 -Fe 2 O 3 . It is known from the literature that oxidation reactions catalyzed by metal oxides involve metal reduction/re-oxidation steps involving TBHP, while the migration of lattice oxygen also occurs [21,22]. These steps at some point may occur through radical species, which can be scavenged by the radical scavenger, leading to the observed rate decrease.
These results concerning the existence of a radical mechanism were the first to the best of our knowledge as we could not find evidence in the literature for similar findings in related systems [23][24][25]. However, proof of evidence for ruling in or ruling out radical mechanisms was also not found. In this way, these results are worth further investigation.
The shape of the kinetic curves should also be discussed. As observed in Figures 4 and 5, these followed stepwise profiles, which indicate a complex reaction system with different mechanisms involved. These observations may be related to the fact that these catalysts have different active sites where the reaction takes place and display different reactivities, as recently reported by us using neutron scattering [15]. The profiles also demonstrated that as proof of a complex reaction with different mechanisms occurring, radical pathway(s) could not be ruled out, giving credit to the results from the experiment with the radical scavenger.
Given that the MoO 2 -Fe 2 O 3 catalyst was synthesized using Fe 2 O 3 as the assisting agent and was incorporated as a dopant [14], we assessed if this contaminant had any influence on the catalytic reaction. In this way, a reaction with Fe 2 O 3 nanoparticles was run to infer its influence on the catalytic oxidation reaction of benzyl alcohol for control purposes. followed stepwise profiles, which indicate a complex reaction system with different mechanisms involved. These observations may be related to the fact that these catalysts have different active sites where the reaction takes place and display different reactivities, as recently reported by us using neutron scattering [15]. The profiles also demonstrated that as proof of a complex reaction with different mechanisms occurring, radical pathway(s) could not be ruled out, giving credit to the results from the experiment with the radical scavenger.   Figure 6 shows the kinetics of the catalytic reaction with Fe 2 O 3 , and it was possible to observe that γ-Fe 2 O 3 had an apparent influence in the obtained result of conversion and in the obtained product selectivity. At the beginning of the reaction, benzaldehyde was formed. However, after 1 h reaction time, the benzaldehyde concentration decreased, being converted into benzoic acid. After 24 h of reaction, benzoic acid was the sole product (Table 1, entry 9). After obtaining these results, it was possible to conclude that the γ-Fe 2 O 3 had an influence on the selectivity of benzaldehyde when MoO 2 -Fe 2 O 3 was the catalyst. In this catalyst, there was only a small amount of γ-Fe 2 O 3 , most likely at the surface of the catalyst, which led to the formation of small amounts of benzoic acid (Table 1, entries 1-4), thus supporting the observed selectivity. This was not observed for the MoO 2 calcined catalyst ( Table 1, entries 5-8) lacking Fe 2 O 3 and corroborated these findings.
In a continuous effort to obtain more efficient catalysts and green processes, we tested the MoO 2 calcined catalyst in a reaction without solvent. Analyzing Table 1, entry 8, it is possible to observe that the reaction reached 69% substrate conversion and achieved 100% selectivity for benzaldehyde. This was a remarkable result, which matched almost the same conversion levels observed with the use of a solvent (Table 1, entries 5-7). possible to conclude that the γ-Fe2O3 had an influence on the selectivity of benzaldehyde when MoO2-Fe2O3 was the catalyst. In this catalyst, there was only a small amount of γ-Fe2O3, most likely at the surface of the catalyst, which led to the formation of small amounts of benzoic acid (Table 1, entries 1-4), thus supporting the observed selectivity. This was not observed for the MoO2 calcined catalyst ( Table 1, entries 5-8) lacking Fe2O3 and corroborated these findings. In a continuous effort to obtain more efficient catalysts and green processes, we tested the MoO2 calcined catalyst in a reaction without solvent. Analyzing Table 1, entry 8, it is possible to observe that the reaction reached 69% substrate conversion and achieved 100% selectivity for benzaldehyde. This was a remarkable result, which matched almost the same conversion levels observed with the use of a solvent (Table 1, entries 5-7).
After these promising results, we tried to replace the oxidant TBHP by hydrogen peroxide with the MoO2-Fe2O3 catalyst. The first tests were carried out with some solvents and solventless condition (Table 3), since promising results were obtained without a solvent with TBHP.  After these promising results, we tried to replace the oxidant TBHP by hydrogen peroxide with the MoO 2 -Fe 2 O 3 catalyst. The first tests were carried out with some solvents and solventless condition (Table 3), since promising results were obtained without a solvent with TBHP. Regarding the use of H 2 O 2 , the results were not so promising at first sight. Conversions were found to range from medium to fair, between 12-48%. The effect of the temperature was also tested in the reaction performance. It was possible to say, after analyzing Table 3, that temperature had a slightly negative effect in the final conversion with H 2 O 2 as the oxidant (Table 3, entries 1-3). This demonstrated that in the reactions at 353 K and 383 K with acetonitrile as the solvent when the temperature was raised, although marginally, both the substrate conversion and benzaldehyde selectivity dropped. Changing the solvent to decane or using solventless conditions (Table 3, entries 3 and 4) had an even more dramatic negative impact on the reaction by lowering the substrate conversion, and by significantly affecting product selectivity as well.
It is known that H 2 O 2 decomposes inefficiently in parallel processes. To try to circumvent this, we then took the best conditions (Table 3, entry 1) and explored the effect of new additions of the oxidant during the catalytic reaction (Table 3, entries 5 and 6). The obtained results allowed us to investigate the effect of adding a similar amount of H 2 O 2 at the beginning of the reaction and after 4 h. We found that the obtained substrate conversion was twice the original one after 24 h (Table 3, entries 1 and 5). Taking this result into consideration and in the same trend, we conducted a new reaction and the oxidant was added in similar amounts at the beginning of the reaction, and then after 4 h and 8 h ( Table 3, entry 6). In this experiment, 94% substrate conversion was obtained with 100% selectivity for the desired product, benzaldehyde, at a relatively low temperature (353 K), which was remarkable. These results showed that it was possible to use hydrogen peroxide as an oxidant, adding it at known times and overcome the inefficient decomposition of H 2 O 2 . It also confirmed that degradation of H 2 O 2 occurred in parallel processes, making the reaction less efficient. Despite this, it also demonstrated that the process was feasible and, more importantly, highly selective toward benzaldehyde.
In Figure 7, it is possible to observe that the kinetics of the reaction had a higher rate when more oxidant was added after 4 h reaction time. At 4 h reaction time, 24% conversion was obtained and at 6 h (2 h after the addition of a new amount of oxidant), almost 60% conversion was reached. From 6 h to 8 h reaction, the kinetics was slower again and only 72% conversion was obtained. At 8 h, more H 2 O 2 was added again and the reaction proceeded from 72% to 94% after 24 h. A faster kinetics was not observed. However, the final conversion was improved, reaching 94% with 100% selectivity for benzaldehyde.  Catalyst stability was also evaluated for the MoO 2 calcined nanomaterial by FTIR analysis of the recovered catalyst at the end of the reaction. According to Figure 8, the FTIR spectrum seems to globally match that of the fresh MoO 2 calcined NP catalyst (Figure 3). It is possible to see that there was the presence now of low intensity νC-H modes at ca. 2900-3000 cm −1 , which are mostly due to surface-adsorbed species from the reaction. More importantly, the most striking feature is the presence of a pair of bands at 935 and 914 cm −1 due to the νMo=O modes, which is indicative that there are Mo VI =O species at the surface of the catalyst, as expected. This observation agrees with the mechanistic proposals and is also evidence that the MoO 2 NPs are preserved after catalysis [11,14].
On the other hand, the XRD powder pattern from the recovered catalyst ( Figure 8) shows that, overall, there was a similar pattern to that in Figure 1 concerning the existence of the MoO 2 phase. However, diffraction peaks arising from the MoO 3 phase were found at the surface along with some other sharp diffraction peaks that were due to surface-adsorbed organics. These data corroborated the FTIR results discussed above, being supported by recent findings using neutron scattering [15], and explaining the considerable activity decrease in recycling experiments. globally match that of the fresh MoO2 calcined NP catalyst (Figure 3). It is possible to see that there was the presence now of low intensity νC-H modes at ca. 2900-3000 cm −1 , which are mostly due to surface-adsorbed species from the reaction. More importantly, the most striking feature is the presence of a pair of bands at 935 and 914 cm −1 due to the νMo=O modes, which is indicative that there are Mo VI =O species at the surface of the catalyst, as expected. This observation agrees with the mechanistic proposals and is also evidence that the MoO2 NPs are preserved after catalysis [11,14]. On the other hand, the XRD powder pattern from the recovered catalyst ( Figure 8) shows that, overall, there was a similar pattern to that in Figure 1 concerning the existence of the MoO2 phase. However, diffraction peaks arising from the MoO3 phase were found at the surface along with some other sharp diffraction peaks that were due to surface-adsorbed organics. These data corroborated the FTIR results discussed above, being supported by recent findings using neutron scattering [15], and explaining the considerable activity decrease in recycling experiments.
The catalytic performance demonstrated by the two different catalysts was relevant and their performances in the oxidation of BzOH to BzCHO using TBHP or H2O2 with or without solvent are compared in Table 4. As shown in entries 14-16, the overall catalytic performance of the MoO2-Fe2O3 and MoO2 calcined nanomaterials was comparable or better than other catalytic nanomaterials reported in the literature [6,[23][24][25][26][27][28][29][30][31][32][33][34]. The catalysts shown in Table 4 cover a wide selection of The catalytic performance demonstrated by the two different catalysts was relevant and their performances in the oxidation of BzOH to BzCHO using TBHP or H 2 O 2 with or without solvent are compared in Table 4. As shown in entries 14-16, the overall catalytic performance of the MoO 2 -Fe 2 O 3 and MoO 2 calcined nanomaterials was comparable or better than other catalytic nanomaterials reported in the literature [6,[23][24][25][26][27][28][29][30][31][32][33][34]. The catalysts shown in Table 4 cover a wide selection of homogeneous and heterogeneous catalysts and in terms of the BzCHO yield, it was higher than those of other catalysts. In some cases, the obtained results with other catalysts were better, however, the amount of the catalysts, oxidant, temperature, and pressure were higher. 30 min, 1, and 1.5 h, then at 2, 4, 6, 8, and 24 h of reaction. These samples were treated as described previously prior to injection in the GC column [35] using a Shimadzu QP2100-Plus GC/MS system with a capillary column Tecknokroma TRB-5MS (Barcelona, Spain) operating in the linear velocity mode. The presented results (conversion, yield, etc.) were the average of three replicates, each obtained by two different people to ensure their repeatability and reproducibility.

Conclusions
Normally, the most widely used catalysts for the reaction of the oxidation of benzyl alcohol to benzaldehyde described in the literature are nanoparticles of noble metals such as Pt, Au, Pt, or Ru. In this work, we developed heterogeneous based catalysts of MoO 2 NPs and demonstrated that they were efficient for the selective oxidation of benzyl alcohol to benzaldehyde using 30% aqueous H 2 O 2 or TBHP as the oxidant, which are considered as green oxidants, under solvent or solventless conditions. The results were very promising, exhibiting excellent catalytic efficiency, reaching 95% yield and almost 100% selectivity for benzaldehyde under the optimized tested conditions with TBHP. The existence of a radical pathway was supported by the experimental results and will be further addressed in the future.
Of even higher relevance were the studies with H 2 O 2 , which allowed us to achieve 94% substrate conversion with 100% selectivity, although it has the drawback of the cost of feeding additional oxidant to the reaction after 4 h and 8 h reaction time. This result showed that hydrogen peroxide was also consumed during the reaction in parallel inefficient processes. We are endeavoring further efforts to fine-tune the efficiency of the reaction by adjusting the initial H 2 O 2 amounts and the periodicity of additions to optimize the process. Additionally, a comparison with results found in a survey from the published literature for related catalytic systems showed that although the system reported here was not the best performing, it reached a similar performance under somewhat milder conditions, which is also advantageous.
Author Contributions: Conceptualization, methodology, formal analysis, validation, writing manuscript, and review, supervision, project management, C.D.N. Experimental work and draft writing, F.G. All authors have read and agreed to the published version of the manuscript.