Development of semi-continuous chemo-enzymatic terpene epoxidation: combination of anthraquinone autooxidation and the lipase-mediated epoxidation process

Abstract

Lipase has been used for epoxidizing olefins such as monoterpenes for more than two decades. This epoxidation is accomplished by adding hydrogen peroxide (H₂O₂) to a carboxylic acid in the presence of a lipase such as Candida antartica lipase B (CALB) to produce percarboxylic acid, which then epoxidizes monoterpenes according to the Prilezhaev mechanism. One drawback of this process is the need for continuous addition of hydrogen peroxide to maintain maximum productivity. To overcome this hurdle, the industrial anthraquinone autooxidation process for hydrogen peroxide production was scaled down and coupled with lipase-mediated epoxidation in a semi-continuous manner. Palladium on alumina pellets (5% loading) was used as the catalyst for obtaining high yields of high-concentration hydrogen peroxide (50% weight by volume), followed by epoxidation of 3-carene, (+) limonene, and α-pinene. A total reaction time of 5 h was used for hydrogen peroxide production and 2–3 h for the epoxidation reactions. Pure 3-carene epoxide and α-pinene epoxide were obtained in isolated yields of 88.8 ± 2.8% and 83.8 ± 2.6%, respectively. Limonene epoxide was obtained as a mixture of mono- and di-epoxides in a ratio of 70% and 30%, respectively, with an isolated yield of 71.5 ± 3.1%.

---

1 Introduction

Epoxides, also known as oxiranes, are cyclic ethers that are industrially significant owing to their high reactivity to form intermediates, which in turn form products of high value. The process is straightforward and involves the addition of a free or substituted oxygen atom to an olefin.¹ Epoxides can be produced using pure oxygen or air,^2–4 hydroperoxides,^5,6 hydrogen peroxide,^7,8 and peroxy compounds.⁹ Industrially, oxygen- or ozone-based epoxidation is practiced as a gas-phase reaction in the presence of a metal catalyst for ethene,¹⁰ propene,¹¹ and butene;^12,13 however, other olefins in the liquid state are seldom epoxidized in this manner. An alternative to this method is the Prilezhaev reaction, which uses peroxycarboxylic acid in stoichiometric amounts to perform epoxidations. Note that meta-chloroperbenzoic acid (m-CPBA) is the most commonly used peroxycarboxylic acid in these syntheses.¹⁴ Generally, handling and cleaning issues coupled with the possibility of an explosion hazard make this process dangerous at industrial levels of production; therefore, in situ generation of peroxycarboxylic acid or slow addition of the compound is recommended. However, slow addition of peroxycarboxylic acid produces huge amounts of waste (equimolar to the amount of product); hence, in situ generation is preferred.¹⁵

Peroxycarboxylic acid can be generated in situ either chemically or enzymatically. Harsh reaction conditions, such as a strong mineral acid catalyst with carboxylic acid and hydrogen peroxide, are required to produce peroxycarboxylic acid. This leads to waste neutralization issues and unwanted side reactions that contribute to polluting processes, e.g., formation of performic acid. Therefore, the enzymatic method reported by Björkling et al. is preferred.¹⁶ The reaction schemes for both chemical and enzymatic means of epoxidation are depicted in Scheme 1. Ever since this report was published in 1992, Prilezhaev-based epoxidation that uses lipases has been the most preferred route for epoxidations^17–26 owing to its safer and simpler synthetic conditions.

Scheme 1 Epoxidation of alkene (black) using the procedure of Prilezhaev (red) and Björkling et al. (blue).¹⁶ The rectangular, dashed box implies that the compound was generated in situ. (R, R′, R′′, and R′′′ are functional groups present on the alkene).

However, as evident from the scheme above, lipase-mediated epoxidation has one major drawback, i.e., the exhaustion of hydrogen peroxide, which limits the synthesis to a batch process. This implies the need to add fresh hydrogen peroxide continuously or in situ generation using enzymatic,^27–30 electrochemical,^31–33 photocatalytic,^34–36 or chemical means (Schemes 3 and 4).^37–39 Ni et al. reported the use of an enzymatic cascade for producing H₂O₂ and CO₂ from methanol in combination with a peroxidase enzyme for the production of halogenated thymols.⁴⁰ Holtmann et al. used an electrochemical approach to generate H₂O₂ from the same reaction.⁴¹ Churakova et al. used EDTA in the presence of light to efficiently generate moderate amounts of H₂O₂ from an aromatic peroxygenase in order to obtain aromatic phenols from their corresponding precursors.⁴² In addition to these specific examples, several other works have used the combination of in situ H₂O₂ generation methods with reactions requiring H₂O₂.^43,44 One common trend in these processes is that they are excellent innovations for lab-scale applications when only low concentrations of hydrogen peroxide are required. The feasibility of the industrial use of these methods has not been reported so far or is pending investigation. In industrial applications, hydrogen peroxide is produced chemically using autooxidation⁴⁵ or direct H₂/O₂ (ref. 46) processes as well as using the 2-propanol oxidation process for a brief period.⁴⁷

The focus of this work is the functionalization of renewable resources for the production of value-added fine chemicals. Terpenes are one such resource as they are naturally occurring hydrocarbons predominantly obtained from plants as ethereal oils. Terpenes comprise repeating isoprene units, which are susceptible to biological degradation and can be obtained as waste products from the paper and pulp industry.^48–50 Currently, these hydrocarbon reserves are combusted for energy production or used in paints and varnishes; however, functionalizing these terpenes would be beneficial for the fragrance, flavor, fine chemical, and of late, the polymer industry.^51–54 Recently, our research group reported a combination of the chemical anthraquinone process and lipase-mediated epoxidation,⁵⁵ the very first report on using such a combination for producing epoxides. Although innovative, the process could only be used for a single run, i.e., a batch reaction. In this study, terpenes were chosen as the olefins to be epoxidized. This study also focuses on designing a semi-continuous epoxidation method for terpenes and developing a prototype process for the industry. We combined the processes of H₂O₂ production and epoxidation, which provided us better control on production as well as room for other reactions that require H₂O₂. H₂O₂ production was increased by optimizing the catalyst loading. The transfer of gaseous hydrogen to dissolved hydrogen in the working solution was enhanced using high mixing rates. Moreover, a stainless steel mesh container for the palladium catalyst was used to protect the palladium catalysts from shear forces associated with the high mixing rates. The combined effect of both should yield higher H₂O₂ yields. For epoxidation, ethyl acetate was used as the reaction solvent owing to its greenness and capability as a peroxy acid generator (Scheme 2).

Scheme 2 Lipase-mediated (CALB) epoxidation of 3-carene, limonene and α-pinene in ethyl acetate.

Scheme 3 Reduction of 2-ethyl anthraquinone (2-EAQ) in the presence of 2.5 mol% palladium at 60 °C to produce 2-ethyl anthraqhydroquinone (2-EAH₂Q).

Scheme 4 Auto oxidation of 2-EAH2Q to 2-EAQ and hydrogen peroxide (H₂O₂) production at 22 °C in the presence of air.

2 Results & discussion

First, the anthraquinone-based autooxidation process was optimized to obtain maximum H₂O₂ production. Lipase-mediated epoxidation has already been optimized in our previous studies^24,25 and will be used herein with a single change in the reaction medium. Ethyl acetate was used in this work as opposed to toluene and deep eutectic solvents, which were used previously.

2.1 Optimization of palladium loading

The first part of the developmental stage of the H₂O₂ production process was determination of the amount of palladium required for the hydrogenation reaction in the anthraquinone process. Previously, we had determined that 10 mol% was necessary to successfully convert 2-ethyl anthraquinone (2-EAQ) to 2-ethyl anthrahydroquinone (2-EAH₂Q). The catalyst used previously was a fine commercial powder of Pd/C as opposed to Pd/Al₂O₃, which was used in this study.⁵⁵ Hence, optimization was performed using the one variable at a time (OVAT) strategy, wherein one of the parameters is varied whilst keeping the others constant. Here, the parameter of interest was the percentage of palladium used with respect to the amount of 2-EAQ used. Tests were performed by mixing 2 × 10⁻² mol 2-EAQ with a 100 cm³ working solution (60% mesitlene and 40% tributyl phosphate) at 250 rpm, as explained in section 4.2.1. The H₂O₂ concentration obtained at the end of the reaction was calculated using the ABTS assay, which has been described in section 4.3.1. The isolated yields obtained at the end of the run for various palladium-to-2-EAQ ratios are given in Table 1, which shows that 2.5 mol% yielded the best results for hydrogen peroxide production.

Table 1 Isolated yields of H₂O₂ (50%) after one cycle of the anthraquinone process upon using 2 × 10⁻² mol 2-EAQ with 60 cm³ of mesitylene and 40 cm³ of tributyl phosphate

S. no.	% Pd/Al2O3 : 2-EAQ (mol : mol)	Isolated yield (%)
1	0.5	—
2	1.25	94.3
3	2.5	97.2
4	5	93.8
5	10	90

---

According to Table 1, a loading of 0.5 mol% did not produce any 2-EAH₂Q; as a result, no H₂O₂ was produced. Using 1.25 and 2.5 mol% resulted in yields of 94.3% and 97.2%, respectively; however, when using 5 and 10 mol%, the yields were less than 97%. This behavior could be the result of nonspecific hydrogenation of the aromatic rings of the 2-EAQ molecule, reduction of a single keto function of 2-EAQ, or dimerization of nonspecific products.³⁸

Interestingly, when using scaled-up conditions, i.e., 1.25 mol% Pd/Al₂O₃, 6.5 × 10⁻² mol 2-EAQ, 60% mesitylene, 40% tributyl phosphate (working-solution volume of 150 cm³), 60 °C temperature, and mixing at 250 rpm, there was a substantial reduction in the isolated yield of H₂O₂ (∼75%). The reason for the decrease in the H₂O₂ yield was poisoning of the catalyst, as confirmed when the catalysts were transferred from the working solution to a glass beaker and left overnight in the fume hood to dry after washing with 2-propanol. The following day, yellowish crystals were visible on both the beaker and the catalyst, which we hypothesize was 2-EAQ. Increasing the washing steps with 2-propanol did not help with the removal of this impurity; however, such a behavior was not observed when using 2.5 mol% Pd/Al₂O₃ under scaled-up conditions. We presume that this is because of an appropriate ratio of Pd-to-2-EAQ was being used for the hydrogenation reaction. At the end of the reaction, a similar H₂O₂ yield was obtained. Hence, in the interest of scaling up, the decision to use 2.5 mol% of Pd/Al₂O₃ with respect to 2-EAQ was made.

2.2 Reusability tests for palladium

Once the palladium loading for H₂O₂ production was finalized, the next step was to check for the number of cycles for which the catalyst could be used. All tests were performed in triplicate. The first tests were performed without washing between runs. The H₂O₂ concentration was tested using the ABTS assay (section 4.3.1), and the results are given in Table 2. The reaction conditions used in this test are given in section 4.2.1.

Table 2 Isolated yields of 50% H₂O₂ when performing the reusability test (in triplicate)

% Isolated H2O2 yield
Run 1	Run 2	Run 3
87.5 ± 2.2	58.7 ± 4.3	18.5 ± 3.8

---

According to Table 2, the amount of H₂O₂ decreased drastically over three cycles possibly because of catalyst inactivation, leaching of the palladium on account of the high shear and grinding forces associated with mixing, or residual reactants sticking to the catalyst. First, the catalyst was washed in an attempt to increase reusability. For this purpose, several solvents were screened, such as acetone, acetone : water (1 : 1 v/v), water, 2-propanol, and n-hexane : acetone (1 : 1 v/v). None of the solvents improved the H₂O₂ yield and are therefore not discussed here. A literature survey was conducted, and, finally, the washing protocol reported by Wang et al. in 2004 (ref. 56) was selected. The palladium on alumina catalyst was first washed with 15 cm³ of ethanol and then with 15 cm³ water; this process was repeated two times. A detailed account of the procedure can be found in section 4.2.2.

2.3 Use of a stainless steel mesh container to enhance Pd/Al₂O₃ reusability

Once the washing protocol was finalized, the washing process was performed again; consequently, the reusability of the catalyst was slightly higher (Fig. 2) than the previous tests performed (Table 2) without washing. However, the loss of activity over four cycles was still high. Catalyst pellets were recovered via filtration. During this step, the filter paper showed black spots of palladium particles, possibly resulting from leaching of the catalyst caused by shear forces or grinding operation of the magnetic stirrer.

To overcome this drawback, the hydrogenation step was repeated using a stainless-steel mesh container for Pd (Fig. 1). In this step, the catalyst was loaded into a stainless steel mesh designed in the form of a pouch. Upon performing hydrogenation using this setup, the loss of activity was minimal (Fig. 2). As shown in the figure, the experiment with the container that houses the catalyst yielded better results than the reaction without it. Unfortunately, there was a considerable loss in the activity of the palladium catalyst despite the washing and protection from shear and grinding forces; the reason for this could be due to the leaching of the catalyst from the support or the formation of intermediates during hydrogenations that could not produce H₂O₂. The exact reason needs to be investigated further.

Fig. 1 Use of a stainless-steel container to reduce catalyst leaching and to increase the miscibility of hydrogen with 2-EAQ. a – 2-EAQ in the working solution before hydrogenation. b – 2-EAH2Q, the product of hydrogenation of 2-EAQ.

Fig. 2 Isolated H₂O₂ yields obtained using a stainless-steel container (squares) and that obtained without it (diamonds). Reaction conditions: 0.02 mol 2-EAQ, 65 cm³ mesitylene, 35 cm³ tributyl phosphate, and 2.5 mol% Pd/Al₂O₃ (5% loading).

2.4 Comparison of H₂O₂ production in the industry with that achieved this work

H₂O₂ is produced worldwide on an industrial scale exclusively via the anthraquinone autooxidation process.⁵⁷ One cycle of the optimized anthraquinone process comprises hydrogenation, catalyst separation, oxidation, and extraction/concentration steps.^37–39 Traditionally, a slurry-type reactor is used for the hydrogenation step,⁵⁷ as in the case of our previous study.⁵⁵ This leads to leaching of the palladium catalyst into the working solution, thereby catalyzing the decomposition of H₂O₂. As a result, fixed-bed reactors are now used in the industry to avoid the tedious unit operation of filtration.⁵⁸ This study combines the advantages of both reactor types and presents an initial case of a “hybrid reactor” for this purpose. Pd/Al₂O₃ pellets are placed in a self-assembled stainless-steel container with a mesh size of 0.2 mm, which is advantageous for preventing any shear loss due to mixing whilst ensuring proper diffusion of reactants to the catalytic surface. The mixing of reactants in this work was performed at 1000 rpm, which ensures maximum mass transfer.

Step 1: reduction and inert gas sparging. Herein, 6.5 × 10⁻² mol 2-EAQ was dissolved in 150 cm³ of the working solvent in a 250 cm³ round-bottom flask (hydrogenation chamber). The working solvent was yellow in color at this point. Then, 2.5 mol% Pd/Al₂O₃ (5%) pellets enclosed in a stainless-steel mesh container were added to this mixture. The reaction vessel was sparged with an inert gas to remove traces of air locked within the reaction vessel. Hence, argon was used for 2–4 min; alternatively, nitrogen gas could be used. Post sparging, a hydrogen atmosphere was maintained and the reactants were mixed at 1000 rpm and 60 °C in an oil bath for 5 h. In this study, a hydrogen balloon was used for this purpose; however, at a pilot-plant scale, a hydrogen feed line with a maximum pressure of 1–2 bar is required to avoid any nonselective hydrogenation of 2-EAQ.

Step 2: catalyst separation. After completion of hydrogenation, the palladium catalyst was separated from the reaction mixture by removing the stainless-steel mesh container containing the catalyst. The catalyst, along with the stainless-steel container, was washed according to the process explained in section 4.2.2. The working solution, which was dark red to brown in color at this stage, was transferred to a fresh 250 cm³ round-bottom flask (sparge chamber). The liquid contents were sparged once again with argon for 2–4 min and then transferred into a fresh 250 cm³ round-bottom flask (oxygenation chamber).

Step 3: oxidation. The oxidation step was performed by pumping air (maximum amount of 250 cm³) through the dark red to brown working solution 15–30 min using a commercially available aquarium pump. The color of the solution was inspected visually. The oxidation step was prolonged until a yellow color was observed, after which the solution was transferred to a separating funnel. No catalyst was needed for the oxidation step.

Step 4: extraction with water. Following the oxidation step, H₂O₂ was extracted from the organic phase using a separating funnel with 4.4 cm³ water to obtain a ca. 50% (weight by volume) mixture of H₂O₂. The organic phase was then pumped into the first vessel for the second round of reduction. The aqueous phase with hydrogen peroxide was stored in a reservoir vessel at 4 °C prior to use in the epoxidation process.

2.5 Development of the combined semi-continuous approach for the epoxidation of terpenes

A semi-continuous approach based on the combined anthraquinone process and a lipase process was used to epoxidize monoterpenes. The scheme of this setup is shown in Fig. 3.

Fig. 3 Semi-continuous approach for combining the industrial anthraquinone process (blue-box contents) for hydrogen peroxide production and lipase-mediated epoxidation (red-box contents) of monoterpenes.

First, the anthraquinone process was performed for hydrogen peroxide formation. It was followed by lipase-mediated epoxidation of monoterpenes. After 5 h, a membrane pump was used to transfer the working solution (dark red to brown) from the hydrogenation chamber to a sparging chamber. A flow rate of 100 cm³ min⁻¹ (maximum capacity of the pump) was used throughout this process for transferring liquids. The reduced working solution was then transferred from the sparging chamber to the oxygenation chamber, and, after oxidation for 15–30 min, the working solution (oxidized, yellow) was transferred to a separating funnel to which water was added for extraction. The working solution was then transferred to the hydrogenation chamber and the reaction was repeated.

The hydrogen peroxide solution produced using the anthraquinone autooxidation process (50% w/v) was collected in a reservoir until further use (Fig. 3). The epoxidation chamber comprised two parts: a reaction chamber and a purification chamber. The reaction chamber contained ethyl acetate, terpene, and CALB. To this mixture, an appropriate amount (with respect to terpene) of H₂O₂ was added to start the reaction, which was monitored by sampling at regular intervals and subjecting the samples to chromatography-mass spectrometry (GC-MS) analysis. After confirming 100% conversion of the starting material, the reaction components were transferred to a purification chamber and pure epoxide was obtained according to a protocol published previously.²⁴

2.6 Epoxidation results

Lipase-mediated epoxidation of monoterpenes, namely 3-carene, limonene, and α-pinene, was performed according to the procedure mentioned in section 4.2.4. Samples were withdrawn at regular intervals, and the conversion profile was followed using gas GC-MS, as reported previously.⁵⁵ The kinetics of the epoxidation of the three reactants is shown in the Fig. 4.

Fig. 4 Conversion profile of 5 × 10⁻³ mol reactant (3-carene – diamond; limonene – square; α-pinene – triangle), 7.5 × 10⁻³ mol H₂O₂ (50% w/v), and 0.1 g CALB in 25 × 10⁻³ L ethyl acetate at 45 °C.

From the figure, it can be inferred that limonene (square) converted to its corresponding mono- and di-epoxide within 2 h and that 3-carene (diamond) and α-pinene (triangle) underwent full conversion within 3 h. It is to be noted that the limonene reaction was ca. 67% selective for the mono epoxide and 30% selective for the di-epoxide. Additionally, all three reactions yielded diols (≤3%) because of epoxide ring opening (ESI;† GC chromatograms and mass spectra). Comparing this result with that obtained using our previous system with toluene,²⁴ the reaction time reduced by a factor of 4 for limonene and 3.33 for 3-carene and α-pinene. The GC chromatograms and mass spectra are attached in the supplementary information. Comparing the results of the epoxidation in this study with the choline chloride : urea·H₂O₂ DES system²⁵ demonstrates that the reaction shows similar results of total turnover of reactants within 3 h.²⁵ The epoxides produced were then purified using a procedure developed previously.²⁴ Isolated yields of 88.8 ± 2.8, 71.5 ± 3.1, and 83.8 ± 2.6% were obtained for 3-carene, limonene, and α-pinene, respectively.

A prototype of this process was recently published by our research group.⁵⁵ Previously, a working solution of toluene and ethyl acetate (3 : 2 v/v) was used with a catalyst loading of 10 mol%. Additionally, a fine powder of commercially available palladium on carbon (Pd/C) was used as the catalyst. The use of Pd/C was problematic during the filtration step as fine particles passed through the filter, leading to multiple filtration steps. The oxidation step was coupled with epoxidation using CALB as a catalyst. Lipase converted ethyl acetate into ethanol and acetic acid, thereby restricting the process to be operated strictly in the batch mode. The process used in the present study overcomes all the challenges discussed above and is capable of being operated in a semi-continuous mode. The working solution was replaced by mesitylene : tributyl in a ratio of 1 : 1 (v/v). Although not essential, this was done to facilitate better solubility of 2-EAQ, 2-EAH₂Q, and hydrogen gas. Moreover, this liquid combination is one of the preferred solvents in the industry.³⁹ To overcome the issues related with filtration, palladium on alumina pellets was used so that the filtration step could be avoided; however, there were issues regarding catalyst leaching that need to be solved. To this end, a stainless-steel mesh container designed in-house was used. A minor change was made in the epoxidation part of the process by using an increased amount of terpene compared with the previous work.⁵⁵

3 Conclusion

This study describes the coupling of chemical and enzymatic processes, i.e., the anthraquinone autooxidation synthesis of hydrogen peroxide and lipase-mediated epoxidation. To the best of our knowledge, such a design is the first of its kind. We used a stainless-steel mesh container to prevent the shear and mechanical grinding forces, thereby enhancing hydrogenation reactions by employing high mixing rates. In other words, this setup is a combination of the continuous stirred tank reactor (CSTR) and a fixed-bed reactor, making it a hybrid reactor that incorporates the advantages of both. High mixing rates ensure maximum mass transfer, catalyst reuse of up to five cycles with minimal loss of activity, and low-temperature operation, which are innovations in hydrogen peroxide production. Owing to the combination of the two processes, there is the opportunity to use the hydrogen peroxide reservoir as feed for other reactions that require H₂O₂. Additionally, this combination gives the option of diluting H₂O₂ according to demand. Since lipase-mediated epoxidation has been studied exclusively for a variety of reactants, the range of this combined process is broad. Moreover, the conversion profiles of the three tested compounds suggest that compared with toluene, the time taken to achieve complete conversion in ethyl acetate was 4 times lower for limonene and 3.33 times lower for 3-carene and α-pinene. To summarize, we believe that this new semi-continuous approach can be scaled up to industrial standards with relative ease and extended to other olefins as well.

4 Materials and methods

4.1 Materials

2-Ethyl anthraquinone (2-EAQ) (Lot#: E12206-100G, 97% purity), mesitylene (Lot#: M7200-500ML, 98% purity), tributyl phosphate (Lot#: 158615-1L, 97% purity), 2,2′-azino-bis(3-ethylbenzothiazoline-6-sulfonsäure diammonium-salz (ABTS) (Lot#: A1888-2G, 98% purity), peroxidase from horseradish (HRP) (Lot#: P6782-50MG), 950–2000 units per mg by ABTS assay), α-pinene (Lot#: 147524-250ML, 98% purity), and (+) 3-carene (Lot#: 115576-1L, 90% purity) were purchased from Sigma-Aldrich. Palladium on alumina (Pd/Al₂O₃) pellets were obtained from VWR Chemicals (Lot#: 41825.06, 5% loading). (+) Limonene (Lot#: 179395000, 96% purity) was purchased from Fischer Scientific. Potassium phosphate buffer (KPi) (1 × 10⁻⁴ mol L⁻¹, pH 5.0) was prepared in the laboratory according to standard buffering procedure and used as such. Ethanol (Lot#: T171.2-25L, 96% purity) was purchased from Carl Roth. Candida antartica lipase B (CALB) was obtained from c-LEcta (Lot#: 20606-4, 17 000 PLU g⁻¹). LC-MS grade ethyl acetate (EtoAc) was obtained from Th. Geyer (Lot: 2278-1L, 99.95% purity). Disposable cuvettes (Ref. #: 67.742, polystyrene material, 10 × 4 × 45 mm³) were bought from Sarstedt. Membrane pumps (KNF SIMDOS®10) fitted with a PTFE membrane to specifically handle organic solvents were used. Tygon® F-4040-A tubing (Lot#: 224-0525, inner diameter 3.2 × 10⁻³ m, outer diameter 6.4 × 10⁻³ m, thickness 1.6 × 10⁻³ m) was purchased from VWR Chemicals and used for liquid transfer. A UV-Vis spectrophotometer from Shimadzu (UV-1800) was used to measure the absorbance during the ABTS assay. A stainless-steel mesh (wire diameter 0.12 mm; mesh size 0.20 mm) was purchased from Metallwaren-Riffert, Austria.

4.2 Synthetic methods

4.2.1 Optimization of the palladium catalyst for hydrogenation of 2-EAQ. First, the amount of palladium for the reduction of 2-EAQ was determined. For this purpose, the following parameters were kept constant: 2 × 10⁻² mol 2-EAQ, 0.1 L yellow working solution (three volume equivalents mesitylene and two volume equivalents tributyl phosphate), hydrogen atmosphere, 60 °C temperature, and mixing at 250 rpm. For the palladium catalyst, prior experience suggested using 10 mol% of 2-EAQ for optimum results.⁵⁵ Nevertheless, tests were conducted using 0.5, 1.25, 2.5, 5, and 10 mol%. The reaction was run for 5 h, after which the catalyst was removed via filtration. The working solution, which was dark red to brown at this stage, was oxidized using air from an aquarium pump at maximum capacity for 15–30 min at 20–22 °C; 10 cm³ double deionized distilled water was used to extract H₂O₂. For scaled-up reactions, water was added accordingly to obtain a 50% (w/v) solution of hydrogen peroxide. The H₂O₂ obtained was quantified using the ABTS assay.

4.2.2 Washing protocol for palladium catalysts. The catalysts were washed according to the procedure reported by Wang et al. in 2004.⁵⁶ That is, 15 cm³ of ethanol was added to the catalysts in a freshly washed and cleaned beaker. Then, the catalysts were immersed in this beaker and mixed for 30–60 s. There should be no spillage of contents during this time. Ethanol was then discarded and replaced with 15 cm³ of double deionized distilled water, and the solution was remixed for 30–60 s. The water was then discarded. This procedure was repeated two times. The wet catalyst was then dried using an inert gas (argon). Alternatively, nitrogen gas could be used. The dry and clean catalyst was then used for the hydrogenation of 2-EAQ. The same procedure was followed for the stainless-steel mesh container.

4.2.3 Application of the in-house-designed stainless-steel mesh container to enhance catalyst lifetime. After the optimization step, the reusability of the Pd/Al2O3 catalyst was tested. To establish the reusability, two tests were conducted. The first test involved adding the catalyst directly to the working solution (i.e., 60% mesitylene, 40% tributyl phosphate, and 2-EAQ). In the second test, a stainless steel mesh container designed in-house was used to shield the catalysts from any shear or grinding forces associated with mixing. The reaction was run until hydrogen peroxide was produced and quantified. The catalysts were then washed properly, as mentioned in the previous section (washing protocol). The washed catalysts were then used for a second time and the process was repeated. The peroxide content was measured and compared with that obtained in the previous run for the reactions with and without the stainless-steel mesh container. Reaction conditions: 0.02 mol 2-EAQ, 100 cm3 working solution (60% mesitylene and 40% tributyl phosphate), 60 °C reduction temperature, and 22–23 °C oxidation temperature. A hydrogen atmosphere was maintained in the vessel using a balloon filled with hydrogen gas. The working solution was oxidized using an aquarium pump.

4.2.4 Lipase-mediated terpene epoxidation. Lipase-mediated epoxidation of monoterpenes was performed using 25 cm³ EtAc. 5 × 10⁻³ mol monoterpene (3-carene, limonene, and α-pinene), 7.5 × 10⁻³ mol H₂O₂ (50%) from the reservoir, and 0.1 g CALB for the reaction. The reaction temperature was set at 45 °C, and mixing was controlled at 250 rpm using a magnetic stirrer. Sampling (0.002 cm³ of the reaction mixture was dissolved in 0.998 cm³ EtoAc) was performed regularly at 15, 30, 45, 60, 90, 120, and 180 min. Shortly before hydrogen peroxide was added, a sample was taken. Conversion was performed using GC-MS, as explained previously.^24,55 The products were purified using a procedure described previously;²⁴ subsequently, isolated yields were calculated.

4.3 Analytical methods

4.3.1 ABTS assay for H₂O₂ detection. The amount of H₂O₂ produced by the process was determined using the ABTS assay.⁵⁹ For this purpose, the reagents needed for the assay were first prepared; 2 × 10⁻³ mol L⁻¹ ABTS (in 0.1 mol L⁻¹ potassium phosphate buffer, KPi, pH 5.0) and 5 × 10⁻³ g L⁻¹ HRP was prepared fresh in appropriate amounts prior to use. The ABTS assay was performed as follows: a 1 × 10⁻³ L ABTS (colorless) solution was pipetted into a standard cuvette, followed by the addition of 0.1 cm³ of the sample (typically H₂O₂; water for blank) and 0.1 cm³ of the HRP enzyme. This mixture was pipetted up and down several times to ensure sufficient mixing of reactants. This mixture was left undisturbed at 22 °C for 10 min. The absorbance of this (green) solution was measured at 405 nm using a UV-Vis spectrophotometer. The concentration of H₂O₂ was determined based on the calibration curves obtained prior to the analyses.

4.3.2 Analysis of terpenes and terpene epoxides by gas chromatography-mass spectrometry (GC-MS). For the detection of the target compounds, i.e. terpenes and their corresponding epoxides, a GC-MS fitted with an autoinjector was used. Details of the equipment are:

• GC: QP 2010, Shimadzu.

• Autoinjector: AOC-5000 by Jain, Compipal.

• MS: GC-MS QP2010 Plus, Shimadzu).

A 30 m long BPX5 with dimensions of 0.25 mm diameter and 0.25 μm thickness was used as the GC column. Helium was used as the carrier gas at a flowrate of 13.2 ml min⁻¹. Temperature profile used for GC and MS for optimal separation of compounds was:

i. G̲a̲s̲ ch̲r̲o̲m̲a̲t̲o̲g̲r̲a̲p̲h̲: start at 60 °C and hold the temperature for 1 minute. Increase the temperature at the rate of 10 °C min⁻¹ until 170 °C, after which the temperature was further increased to 270 °C at the rate of 70 °C min⁻¹. This temperature was then held for 3 minutes.

ii. M̲a̲s̲s̲ sp̲e̲c̲t̲r̲o̲m̲e̲t̲e̲r̲: the ion source temperature was 200 °C and the interface temperature was maintained at 250 °C.

The software program “GC-MS Postrun Analysis” from Shimadzu was used to analyze the reaction components and the mass to charge ratio (m/Q) ratio were compared to the database of National Institute of Standard and Technology (NIST) library; version 14. Ethyl acetate was used as the solvent and to avoid huge signals form this compound, a solvent cut was introduced at 3.9 min with the help of the software. All GC-MS chromatograms obtained for 3-carene/epoxide, limonene/mono- and di-epoxide, α-pinene/epoxide and octanoic acid are given in the supplemental information.

Conflicts of interest

There are no conflicts to declare.

References

1. G. Sienel, R. Rieth and K. T. Rowbottom, Epoxides, In: Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2000, pp. 1–16.
2. J. W. Brown, Q. T. Nguyen, T. Otto, N. N. Jarenwattananon, S. Glöggler and L. S. Bouchard, Epoxidation of alkenes with molecular oxygen catalyzed by a manganese porphyrin-based metal-organic framework, Catal. Commun., 2015, 59, 50–54.
3. L. Vanoye, J. Wang, M. Pablos, C. de Bellefon and A. Favre-Réguillon, Epoxidation using molecular oxygen in flow: facts and questions on the mechanism of the Mukaiyama epoxidation, Catal. Sci. Technol., 2016, 6(13), 4724–4732.
4. J. Dou and F. Tao, Selective epoxidation of cyclohexene with molecular oxygen on catalyst of nanoporous Au integrated with MoO3 nanoparticles, Appl Catal A Gen., 2017, 529, 134–142.
5. Z. Wang, Sharpless Epoxidation, In: Comprehensive Organic Name Reactions and Reagents, John Wiley & Sons, Inc., 2010, pp. 2580–2585.
6. S. El-Korso, I. Khaldi, S. Bedrane, A. Choukchou-Braham, F. Thibault-Starzyk and R. Bachir, Liquid phase cyclohexene oxidation over vanadia based catalysts with tert-butyl hydroperoxide: Epoxidation versus allylic oxidation, J. Mol. Catal. A: Chem., 2014, 394, 89–96.
7. B. S. Lane and K. Burgess, Metal-catalyzed epoxidations of alkenes with hydrogen peroxide, Chem. Rev., 2003, 103(7), 2457–2473.
8. T. Michel, M. Cokoja, V. Sieber and F. E. Kühn, Selective epoxidation of (+)-limonene employing methyltrioxorhenium as catalyst, J. Mol. Catal. A: Chem., 2012, 358, 159–165.
9. H. Klenk, P. H. Götz, R. Siegmeier and W. Mayr, Peroxy Compounds, Organic, In: Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2000, pp. 325–360.
10. M. L. Bocquet, A. Michaelides, D. Loffreda, P. Sautet, A. Alavi and D. A. King, New insights into ethene epoxidation on two oxidized Ag{111} surfaces, J. Am. Chem. Soc., 2003, 125(19), 5620–5621.
11. D. M. Perez Ferrandez, M. H. J. M. De Croon, J. C. Schouten and T. A. Nijhuis, Gas-phase epoxidation of propene with hydrogen peroxide vapor, Ind. Eng. Chem. Res., 2013, 52(30), 10126–10132.
12. R. Atkinson, S. M. Aschmann, J. Arey and E. C. Tuazon, Formation yields of epoxides and O(3P) atoms from the gas-phase reactions of O3 with a series of alkenes, Int. J. Chem. Kinet., 1994, 26(9), 945–950.
13. T. Berndt, A. Kreipl and U. Lenski, Epoxidation of 2-Butene from Raffinate 2 Feedstock, Ind. Eng. Chem. Res., 2010, 49(8), 3957–3960.
14. Z. Wang, Prilezhaev Reaction, In: Comprehensive Organic Name Reactions and Reagents, John Wiley & Sons, Inc., 2010, pp. 2270–2274.
15. A. S. Rao, H. R. Mohan and A. Charette, m-Chloroperbenzoic Acid, In: Encyclopedia of Reagents for Organic Synthesis, John Wiley & Sons, Ltd, 2001.
16. F. Björkling, H. Frykman, S. E. Godtfredsen and O. Kirk, Lipase catalyzed synthesis of peroxycarboxylic acids and lipase mediated oxidations, Tetrahedron, 1992, 48(22), 4587–4592.
17. F. P. Cuperus, S. T. Bouwer, G. F. H. Kramer and J. T. P. Derksen, Lipases Used for the Production of Peroxycarboxylic Acids, Biocatal. Biotransform., 1994, 9(1–4), 89–96.
18. M. Rüsch gen Klaas and S. Warwel, Complete and partial epoxidation of plant oils by lipase-catalyzed perhydrolydsis, Ind. Crops Prod., 1999, 9, 125–132.
19. C. Orellana-Coca, J. M. Billakanti, B. Mattiasson and R. Hatti-Kaul, Lipase mediated simultaneous esterification and epoxidation of oleic acid for the production of alkylepoxystearates, J. Mol. Catal. B: Enzym., 2007, 44(3–4), 133–137.
20. M. A. Moreira and M. G. Nascimento, Chemo-enzymatic epoxidation of (+)-3-carene, Catal. Commun., 2007, 8(12), 2043–2047.
21. Y. Xu, N. R. B. J. Khaw and Z. Li, Efficient epoxidation of alkenes with hydrogen peroxide, lactone, and lipase, Green Chem., 2009, 11(12), 2047–2051.
22. A. A. Tzialla, E. Kalogeris, A. Enotiadis, A. A. Taha, D. Gournis and H. Stamatis, Effective immobilization of Candida antarctica lipase B in organic-modified clays: Application for the epoxidation of terpenes, Mater. Sci. Eng. B Solid State Mater. Adv. Technol., 2009, 165(3), 173–177.
23. F. D. A. Corrêa, F. K. Sutili, L. S. M. Miranda, S. G. F. Leite, R. O. M. A. De Souza and I. C. R. Leal, Epoxidation of oleic acid catalyzed by PSCI-Amano lipase optimized by experimental design, J. Mol. Catal. B: Enzym., 2012, 81, 7–11.
24. S. Ranganathan, J. Tebbe, L. O. Wiemann and V. Sieber, Optimization of the lipase mediated epoxidation of monoterpenes using the design of experiments — Taguchi method, Process Biochem., 2016, 51(10), 1479–1485.
25. S. Ranganathan, S. Zeitlhofer and V. Sieber, Development of a lipase-mediated epoxidation process for monoterpenes in choline chloride-based deep eutectic solvents, Green Chem., 2017, 19(11), 2576–2586.
26. P. Zhou, X. Wang, B. Yang, F. Hollmann and Y. Wang, Chemoenzymatic epoxidation of alkenes with Candida antarctica lipase B and hydrogen peroxide in deep eutectic solvents, RSC Adv., 2017, 7(21), 12518–12523.
27. L. P. Nelles, J. A. Arnold and D. S. Willman, Enzymatic production of hydrogen peroxide and acetaldehyde in a pressure reactor, Biotechnol. Bioeng., 1990, 36(8), 834–838.
28. C. López and A. Cavaco-Paulo, In-situ enzymatic generation of hydrogen peroxide for bleaching purposes, Eng. Life Sci., 2008, 8(3), 315–323.
29. T. Tzanov, S. A. Costa, G. M. Gübitz and A. Cavaco-Paulo, Hydrogen peroxide generation with immobilized glucose oxidase for textile bleaching, J. Biotechnol., 2002, 93(1), 87–94.
30. M. Aghbolaghy and A. Karimi, Simulation and optimization of enzymatic hydrogen peroxide production in a continuous stirred tank reactor using CFD-RSM combined method, J. Taiwan Inst. Chem. Eng., 2014, 45(1), 101–107.
31. T. Jeric, R. J. M. Bisselink, W. van Tongeren and A. M. Le Marechal, Decolorization and Mineralization of Reactive Dyes, by the H2O2/UV Process With Electrochemically Produced H2O2, Acta Chim. Slov., 2013, 60(3), 666–672.
32. J. M. Peralta-Hernández and L. A. Godínez, Electrochemical Hydrogen Peroxide Production in Acidic Medium Using a Tubular Photo-reactor: Application in Advanced Oxidation Processes, J. Mex. Chem. Soc., 2014, 58(3), 348–355.
33. V. Viswanathan, H. A. Hansen and J. K. Nørskov, Selective Electrochemical Generation of Hydrogen Peroxide from Water Oxidation, J. Phys. Chem. Lett., 2015, 6(21), 4224–4228.
34. Y. Shiraishi, S. Kanazawa, Y. Kofuji, H. Sakamoto, S. Ichikawa and S. Tanaka, et al., Sunlight-driven hydrogen peroxide production from water and molecular oxygen by metal-free photocatalysts, Angew. Chem., Int. Ed., 2014, 53(49), 13454–13459.
35. H. Zhuang, L. Yang, J. Xu, F. Li, Z. Zhang and H. Lin, et al., Robust Photocatalytic H2O2 Production by Octahedral Cd3(C3N3S3)2 Coordination Polymer under Visible Light, Sci. Rep., 2015, 5, 16947.
36. K. Mase, M. Yoneda, Y. Yamada and S. Fukuzumi, Efficient Photocatalytic Production of Hydrogen Peroxide from Water and Dioxygen with Bismuth Vanadate and a Cobalt(II) Chlorin Complex, ACS Energy Lett., 2016, 1(5), 913–919.
37. G. Goor, Hydrogen Peroxide - Manufacture and Industrial Use for Production of Organic Chemicals, In: Catalytic Oxidations with Hydrogen Peroxide as Oxidant, 1992, pp. 13–43.
38. G. Goor, J. Glenneberg and S. Jacobi, Hydrogen Peroxide, In: Ullmann's Encyclopedia of Industrial Chemistry, Wiley-VCH Verlag GmbH & Co. KGaA, 2000.
39. J. M. Campos-Martin, G. Blanco-Brieva and J. L. G. Fierro, Hydrogen peroxide synthesis: An outlook beyond the anthraquinone process, Angew. Chem., Int. Ed., 2006, 45(42), 6962–6984.
40. Y. Ni, E. Fernández-Fueyo, A. G. Baraibar, R. Ullrich, M. Hofrichter and H. Yanase, et al., Peroxygenase-Catalyzed Oxyfunctionalization Reactions Promoted by the Complete Oxidation of Methanol, Angew. Chem., Int. Ed., 2016, 55(2), 798–801.
41. D. Holtmann, T. Krieg, L. Getrey and J. Schrader, Electroenzymatic process to overcome enzyme instabilities, Catal. Commun., 2014, 51, 82–85.
42. E. Churakova, M. Kluge, R. Ullrich, I. Arends, M. Hofrichter and F. Hollmann, Specific photobiocatalytic oxyfunctionalization reactions, Angew. Chem., Int. Ed., 2011, 50(45), 10716–10719.
43. J. Rocha-Martin, S. Velasco-Lozano, J. M. Guisán and F. López-Gallego, Oxidation of phenolic compounds catalyzed by immobilized multi-enzyme systems with integrated hydrogen peroxide production, Green Chem., 2014, 16(1), 303–311.
44. K. Köninger, M. Grote, I. Zachos, F. Hollmann and R. Kourist, Light-driven Enzymatic Decarboxylation, J. Visualized Exp., 2016,(111), 5–9.
45. H.-J. Riedl and G. Pfleiderer, Production of Hydrogen Peroxide, US Pat., 2215883, 1940, p. 2.
46. J. K. Edwards, B. E. Solsona, P. Landon, A. F. Carley, A. Herzing and C. J. Kiely, et al., Direct synthesis of hydrogen peroxide from H2 and O2 using TiO2-supported Au-Pd catalysts, J. Catal., 2005, 236(1), 69–79.
47. F. F. Rust, Manufacture of Hydrogen Peroxide, 1959, pp. 1–14.
48. D. S. Seigler, Introduction to Terpenes, In: Plant Secondary Metabolism, 1998, pp. 312–323.
49. J. L. Bicas, A. P. Dionisio and G. M. Pastore, Bio-oxidation of terpenes: An approach for the flavor Industry, Chem. Rev., 2009, 109(9), 4518–4531.
50. R. Mewalal, D. K. Rai, D. Kainer, F. Chen, C. Külheim and G. F. Peter, et al., Plant-Derived Terpenes: A Feedstock for Specialty Biofuels, Trends Biotechnol., 2017, 35(3), 227–240.
51. J. L. F. Monteiro and C. O. Veloso, Catalytic Conversion of Terpenes into Fine Chemicals, Top. Catal., 2004, 27(1–4), 169–180.
52. L. Caputi and E. Aprea, Use of Terpenoids as Natural Flavouring Compounds in Food Industry, Recent Pat. Food, Nutr. Agric., 2011, 3(1), 9–16.
53. W. Schwab, C. Fuchs and F. C. Huang, Transformation of terpenes into fine chemicals, Eur. J. Lipid Sci. Technol., 2013, 115(1), 3–8.
54. S. Kandi, V. Godishala, P. Rao and K. V. Ramana, Biomedical Significance of Terpenes: An Insight, Biomed. Biotechnol., 2015, 3(1), 8–10.
55. S. Ranganathan, T. Gärtner, L. O. Wiemann and V. Sieber, A one pot reaction cascade of in situ hydrogen peroxide production and lipase mediated in situ production of peracids for the epoxidation of monoterpenes, J. Mol. Catal. B: Enzym., 2015, 114, 72–76.
56. Q. Wang, L. Wang, Y. Wang, F. He, Z. Li and Z. Mi, Study on deactivation and regeneration of Pd/Al 2 O 3 catalyst in hydrogen peroxide production by the anthraquinone process, React. Kinet. Catal. Lett., 2004, 81(2), 297–304.
57. J. H. Teles, I. Hermans, G. Franz and R. A. Sheldon, Oxidation, In: Ullmann's Encyclopedia of Industrial Chemistry, WILEY-VCH Verlag, 2015, pp. 1–103.
58. R. Ciriminna, L. Albanese, F. Meneguzzo and M. Pagliaro, Hydrogen Peroxide: A Key Chemical for Today's Sustainable Development, ChemSusChem, 2016, 9(24), 3374–3381.
59. H. Bisswanger, Enzyme Assays, In: Practical Enzymology, Wiley-VCH Verlag GmbH & Co. KGaA, 2011, pp. 93–264.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7re00112f

---