Tungstic Acid: A Simple and Effective Solid Catalyst in Terpene Alcohol Oxidation Reactions with Hydrogen Peroxide

In this work, we report for the first time, the tungstic acid-catalyzed oxidation of terpene alcohols with hydrogen peroxide. This simple, solid, and commercially available catalyst efficiently promoted the conversion of borneol, geraniol and nerol to camphor and epoxide products, respectively. Effects of main reaction parameters, such as catalyst load, the molar ratio of oxidant to the substrate, time, and reaction temperature were investigated. Conversions and selectivity greater than 90% were achieved using 1.0 mol % of H2WO4 after 2 h of reaction at 90 °C. The activation energy was equal to 66 kJmol−1. We propose a reaction mechanism based on the experimental results. This solid catalyst was easily recovered and reused without loss of activity. As far as we know, it is the first time that tungstic acid was used as the catalyst in the oxidation reactions of terpene alcohols.

The combination of tungsten catalysts and hydrogen peroxide has been used in the oxidation of terpene alcohols, which are renewable raw materials to produce fragrance ingredients, perfumes, flavours, and building blocks in the synthesis of drugs [12][13][14][15]. Recently, Keggin heteropolyacid salts derived from phosphotungstic acid were synthesized and successfully evaluated as catalysts in reactions to oxidize monoterpenes and their alcohols [16][17][18]. Heteropoly salts containing large cations with ionic radium higher than 1.3 Angstroms such as Cs + or K + are insoluble in polar solvents and have been used as heterogeneous catalysts [19,20]. Moreover, some modifications in the phosphotungstic anion, such as creating vacancies in their structure and or consecutively, filling them with transition metal cations have generated efficient catalysts toward epoxidation and oxidation reactions [21][22][23][24][25].
Besides the Keggin heteropolyacids, other tungsten compounds likewise their oxides, salts, and complexes have been widely used as catalysts in oxidation reactions with hydrogen peroxide [26][27][28]. Among these, the Venturello system, a two-component association consisting of tungstate and phosphate ions, under acidic conditions, represents a valuable catalytic combination to epoxidize terminal and cyclic olefins with hydrogen peroxide under phase-transfer conditions [29]. This system comprises soluble alkaline tungstate, alkaline salts of phosphoric acid, or even mixtures of both, used in a 1:2 molar ratio, which achieve high selectivities toward epoxide (80-90%) and a virtual complete conversion of hydrogen peroxide.
Noyori et al. have developed an efficient catalytic biphasic system based on the combination Na 2 WO 4 -Q + HSO 4 − to oxidize secondary alcohols to ketones and epoxidize olefins in organic systems with hydrogen peroxide acting as the oxidant [30]. Trunschke et al. demonstrated that the use of hydrothermal combustion techniques converts the inactive MnWO 4 salt into a highly active, selective, nanostructured, and crystalline MnWO 4 / MnO x catalyst, very efficient in the dehydrogenation of propane [31]. Verified that the reaction of [{WO(O 2 ) 2 } 2 (μ-O)] 2− dinuclear peroxotungstate with HNO 3 dramatically accelerates the epoxidation of cyclooctene and various types of olefins [2].
Alternatively, to overcome the drawbacks of the homogeneous and liquid bi-phasic systems, the salts or tungsten complexes can be supported on ion-exchange resins. For instance, a solid catalyst based on the {PO 4 [W(O)(O 2 ) 2 ] 4 }/ Amberlite IRA-90 resin was synthesized and was an efficient heterogeneous system to epoxidize a plethora of olefins in CH 3 CN at 311 K [32,33].
Tungstic acid is a commercially affordable solid used in acid-catalyzed or oxidative processes [34,35]. Even though Prat and Lett have described the use of H 2 WO 4 in olefin epoxidations with H 2 O 2 (aq) as an oxidant during the eighty's years, its use remains still scarcely explored [36]. In that work, regardless of the high yield achieved in the epoxidation of different unsaturated alcohols and olefins at room temperature, longer reaction times (5 o 40 h) and a rigid pH control (i.e., buffered solutions with sodium acetate and trimethylamine N-oxide) was required. Recently, an option for the use of a biphasic homogeneous system was developed combining tungstic acid (catalyst)/ ammonium-based ionic liquid (IL) (co-catalyst)/ hydrogen peroxide (oxidant) [37]. This system was highly effective in the oxidation of eight-membered ring cyclic alcohols to ketone under organic solvent-free conditions and at 363 K temperature.
In this present work, we described that the tungstic acid himself can act as a heterogeneous catalyst in oxidation reactions of terpene alcohols with hydrogen peroxide. This solid acid catalyst was used without any additives, transferphase agents, and pH controllers. Using borneol as a model molecule, the effects of the main reaction parameters were assessed such as time, oxidant: substrate molar ratio, catalyst load, and temperature. The reaction scope was extended to the other terpene alcohols. The reusability of the catalyst was successfully evaluated.

Chemicals
All solvents and reactants were acquired from commercial sources. H 2 WO 4 , dimethylacetamide (DMA), and all the terpene alcohols were purchased from Sigma-Aldrich (99 wt. %). The 34 wt. % aqueous hydrogen peroxide (Vetec) was the oxidant in all reactions.

Products Identification
The main products were identified through GC-MS analyses (Shimadzu MS-QP 2010 ultra, mass spectrometer, electronic impact mode at 70 eV, coupled to a Shimadzu 2010 plus, GC). Additionally, reaction products were identified by coinjection with authentic samples in a gas chromatograph (Shimadzu, GC 2010, capillary column, FID).

Catalytic Runs
Borneol, a secondary bicyclic terpenic alcohol, whose carbonylic derived are used in the synthesis of various fine chemicals was the model molecule. The reactions were performed in a three-necked glass flask (25 mL), fitted with a sampling system and a reflux condenser. In a typical run, a solution of DMA (10 mL) containing borneol (1.0 mmol) was heated to reaction temperature (393 K). After adding an adequate amount of H 2 WO 4 , hydrogen peroxide (2.0 mmol) was gently dropped and the reaction started, being carried out over 2 h.
The reaction monitoring was done by analyzing periodically collected aliquots in a gas chromatograph (Shimadzu, GC 2010, capillary column, FID). The reaction conversion was calculated following the Eq. 1: In this equation, A 0 = initial area of substrate GC peak, and Ai = remaining area of substrate GC peak.
The product selectivity (Pi) was calculated as follows: Herein, A Pi = Corrected GC peak area of product and A consumed = Consumed GC peak area of the substrate, calculated as A 0 -Ai.
The selectivity of alkyl peroxides, which were not detected by GC analyses, was calculated from the mass balance of reaction as described in Eq. 3: In this equation, A consumed = consumed area of GC peak of the substrate (A 0 −Ai) and A products = Σ GC peak area of the products.

Effect of Catalyst Load
Terpene alcohols are raw materials renewable and very useful in the synthesis of various fine chemicals, and their oxidation has been a research goal of our group [16-19, 23-25, 27]. Therefore, the H 2 WO 4 -catalyzed borneol oxidation reactions with hydrogen peroxide were carried in according to previous work [19]. In that work, it was verified that due to the lower reactivity of borneol if compared to the terpene alcohols (i.e., allylic, primary, acyclic), the reactions should be carried out at higher temperatures, greater than 333 K, as performed in the oxidation of geraniol or nerol [38]. Moreover, the DMA solvent stabilizes intermediate peroxides which are normally formed in tungsten-catalyzed oxidation reactions with H 2 O 2 [26]. It motivated the replacement of CH 3 CN solvent for DMA.
Initially, we evaluated the influence of catalyst load in conversion and selectivity of oxidation reaction of borneol aiming to find what the minimum load led to the higher conversion. The kinetic curves are presented in Fig. 1.
Regardless of oxidant excess, only a poor conversion was achieved (< 5%). Conversely, even in the presence of minimum catalyst load, the reaction had a noticeable increase in conversion, which jumped from 34% in the presence of 0.10 mol % of H 2 WO 4 to 94% of conversion when 0.50 mol % was the load used (Fig. 1). The maximum conversion (99%) was attained with 2.5 mol %. In addition, the initial rate reaction was gradually enhanced when the load was increased.
The performance of tungstic acid was superior to that achieved by other solid tungsten catalysts reported in the literature. In Table 1, the main reaction conditions such as catalyst load, molar ratio substrate to oxidant and reaction time are summarized. The H 2 WO 4 -catalyzed reactions were faster than others and required a smaller catalyst load.
The camphor was the only reaction product formed in all the runs (Scheme 1). The only exception was the uncatalyzed reaction, where borneol peroxide was also obtained with 30% of selectivity. In all runs the camphor selectivity was > 97%) 1 3 The literature has described that the oxidation reactions using tungsten-containing compounds (i.e., meta tungstate, heteropolyacids, tungsten complexes) as catalysts, and hydrogen peroxide as oxidant, commonly involve the formation of peroxo-intermediates (i.e., mono, di, and or tetra peroxides). These intermediates act as transfer agents of an oxygen atom from the peroxide oxidant to the substrate (i.e., alcohol, aldehyde, or olefin), generating a carbonylic product or an epoxide, mainly when the substrate is unsaturated allylic alcohol [25][26][27][28][29][30].
In Scheme 2, a probable reaction pathway is proposed for the oxidation of borneol to camphor in an H 2 WO 4 -catalyzed oxidation reaction with hydrogen peroxide. Noyori et al. suggested that peroxide complexes are the most probable intermediates in olefin oxidations in biphasic systems [30]. On the other hand, some peroxide-tungsten complexes were isolated and characterized through crystal XRD diffraction analyses [28,29,32]. Hida and Nogusa suggested that the DMA solvent stabilizes the formation of peroxide intermediates involving alcohol substrate and tungsten catalyst, nonetheless, we judge that there is no evidence that in solution this specie survives a significant time, consequently, it was not included in Scheme 2 [26].

Impact of Molar Ratio Oxidant to Substrate on the H 2 WO 4 -Catalyzed Borneol Oxidation with Hydrogen Peroxide
Some metal catalysts have their performance compromised when a higher oxidant load is used, due to the presence of water in the reaction media. Moreover, the water can lead to the formation of undesirable products, due to nucleophilic addition reactions and or rearrangement of carbon skeletal.
To investigate these possibilities, reactions were performed using the oxidant at three concentration loads (Fig. 2). An increase in oxidant load had a slight impact on the initial rate of the reactions. In the presence of oxidant excess, the formation of catalytically active species (diperoxo-tungstate (VI) complex anion, Scheme 2) is favoured and consequently, the reaction is accelerated [40]. Therefore, the conversion gradually enhanced when a higher oxidant load was used. Conversely, camphor was always the major product regardless of oxidant amount (> 97%). Compared to the unsaturated terpene alcohols such as geraniol or nerol that can be converted to epoxides and carbonylic products, the borneol oxidation gives only camphor [39,41,42].

Effect of Temperature on the H 2 WO 4 -Catalyzed Borneol Oxidation with Hydrogen Peroxide
The influence of temperature on kinetic curves of H 2 WO 4 -catalyzed borneol oxidation with H 2 O 2 was assessed and the main results are in Fig. 3. An increase in temperature led to a higher initial rate of reaction as well as a greater conversion. It can be assigned to increase the number of effective collisions as well as its energy, resulting in a quicker conversion rate. As greater the reaction temperature higher the initial rate and conversion of reactions (Fig. 3a). This is evidence of the endothermic character of the reaction. Similarly, the camphor selectivity was also gradually enhanced when the reaction temperature was increased (Fig. 3b).
From the data in Fig. 3, the order of the reaction concerning the borneol concentration was obtained. It was possible to determine the reaction rate constant (k) using the pseudofirst-order rate law. The curves presented high linearity, suggesting the first-order dependence concerning borneol concentration. Figure 4a shows the curves. From these curves, kinetic constants were calculated ( Table 2).
From Table 2, an Arrhenius plot was obtained (Fig. 4b). The activation energy was equal to 66 kJmol −1 , while the reaction was of first order concerning the borneol concentration (Fig. 4a). Literature has scarce data about the tungstic acid-catalyzed oxidations with hydrogen peroxide [43]. Conversely, there was a considerable number of works using stoichiometric oxidants such as chromium oxide or salts oxidant [44][45][46]. However, the literature survey on the kinetics of oxidation of cyclic alcohols with different oxidants reveals that the reactivity of alcohols varies with the type of oxidant or solvent used, and a direct comparison with borneol oxidation seems inadequate.

Effect of Terpene Substrate on H 2 WO 4 -Catalyzed Oxidation Reactions with Hydrogen Peroxide
The reaction scope was extended to the other terpenic alcohols (Fig. 5). Allylic alcohols (i.e., geraniol and nerol) tertiary (linalool), and terminal alcohol (β-citronellol) were the substrates. Besides the carbonylic products, the reactions of The kinetic curves of oxidation reactions are displayed in Fig. 6. The reactions with geraniol and nerol were faster than borneol, even having achieved the same conversion after 2 h. Conversely, linalool and β-citronellol s were the substrates less reactive. Different from borneol, these substrates have at minimum two types of sites to be oxidized: carbon with a hydroxyl group and double bonds (Fig. 7). The double bonds containing the hydroxyl group at the allylic position were preferentially epoxidized. Likewise demonstrated in previous works, when molybdenum or niobium were the catalysts, the epoxidation of the double bond of these monoterpenes is a hydroxy-assisted reaction [39,41,42].

Reusability of H 2 WO 4 Catalyst in the Oxidation Reactions with Hydrogen Peroxide
The reusability of tungstic acid was evaluated (Fig. 8). The catalyst was recovered through filtration and reused three times without loss activity. Camphor was always the major product in all the runs.
To verify the integrity of the catalyst, FT-IR analyses were recorded and are displayed in Fig. 9. A comparison of the infrared spectrum of the fresh catalyst with that obtained after the first and fourth cycles of recovery/reuse shows that the typical absorption bands of W=O bonds remained at the same wave number and no new absorption band was noticed, suggesting that the catalyst remained intact.
To verify the catalyst leaching, it was performed a qualitative/ quantitative test, using SnCl 2 solution (0.20 mol L −1 ) as the reduction agent. Figure 10 allows comparing the results obtained.
In the presence of stannous chloride (SnCl 2 ), the W +6 cations of the tungstate anions are reduced to W 5+ , giving a blue precipitate (Fig. 10). The UV-Vis spectra were analyzed but the typical band at 460-470 nm was absent in the sample collected from the reaction, evidence that no detectable amounts of tungstate anions were present.

Effect of Solvent on the H 2 WO 4 -Catalyzed Oxidation Reactions of Borneol with Hydrogen Peroxide
Hida   [26]. Those authors tested CH 3 CN and CH 3 OH at the reflux temperature and DMA and DMF at 363 K and verified that two firsts were almost inactive, while two later achieved a yield of 50 and 74% of 2-octanone, respectively. Herein, all the reactions with these solvents were carried out at 363 K in a sealed tube reactor (Fig. 11). The literature has ascribed the superior results achieved in reactions carried out in nitrogenated solvents to greater ability in stabilising peroxide-tungsten intermediates formed in the reactions where tungsten catalysts and hydrogen peroxide are used [26,27]. The solvents interact with the peroxide intermediates through the hydrogen bonds, allowing the reactions to achieve higher conversions.
In all the reactions, regardless of solvent, borneol was the major product. The best performance of nitrogenated solvents can be assigned to their ability to stabilize the peroxidized tungsten intermediates (i.e., [W(O 2 ) 2 ] 2− ) formed during the reaction [26,27]. Selectivity of H 2 WO 4 -catalyzed terpene alcohols oxidation reactions of terpene alcohols with H 2 O 2 , (a Alky peroxides, the probable reaction intermediates were the major products in the reaction of linalool and β-citronellol, evidence that the H 2 WO 4 catalyst was less efficient to promote the oxidation of these substrates)

Conclusions
Tungstic acid efficiently catalyzed the oxidation of terpene alcohols (borneol, geraniol, nerol) to carbonylic (ketone, aldehydes) or epoxide products in the presence of hydrogen peroxide. The impacts of catalyst load, temperature, and the molar ratio of oxidant to the substrate were assessed. An increase in catalyst load increased the reaction rate and the conversion of borneol. A reaction pathway was proposed where a peroxide-tungsten complex is the most probable reaction intermediate. The tungstic acid catalyst was efficiently recovered and reused without loss of activity. The use of a commercially affordable, solid, and reusable catalyst, besides the green oxidant, are the positive aspect of this protocol.