Two-step continuous-flow synthesis of -terpineol

-Terpineol is a monoterpene naturally present in essential oils, of high value on the market as it is a compound widely used as a flavoring, aromatic substance in the cosmetics and food industry. This study aims to produce α-terpineol by two different synthetic strategies, using both batch and continuous flow systems, focusing on the optimization of the process, improving the reaction conversion and selectivity. The first strategy adopted was a one-stage hydration reaction of α-pinene by an aqueous solution of chloroacetic acid (molar ratio 1:1 between pinene and the acid) in continuous flow conditions. This reaction was carried out at 80 oC with a residence time of 15 min, obtaining good values of conversion (72 %) and selectivity (76 %), and productivity of 0.67 Kg.day. The second strategy accomplished was a two-step cascade reaction with limonene as starting material, where the first step is a chemo specific double bond addition using trifluoroacetic acid, and the second step is the basic hydrolysis of the ester promoted by a solution of sodium hydroxide (2.25 M) in methanol (1:1). This reaction was adapted to a continuous flow condition, where all steps happen in a residence time of 40 min, at 25 oC, with no quenching between steps required, giving a conversion of 97 % and selectivity of 81 %, with productivity of 0.12 Kg.day.


Introduction
Over the years flavor and fragrances sector has been growing in all its applications and nowadays it represents a multi-billionaire global market. The growing global industrialization has led to the massive production of processed food, beverages, personal care products, detergents, cleaning products and soaps, which shows the industry's necessity to produce scented or flavored products. Thereby, such a high demand for natural products in this particular area could be seen as a disadvantage because of the fluctuating prices of raw materials. To outline the problem and continue expanding the market, scientific innovations where needed to deliver synthetic fragrances and flavors. 1-3 α-Terpineol is a high value monoterpene naturally present in essential oils widely used as a flavoring aromatic substance. Likewise, it is also used as an anti-fungal agent, as a disinfectant in cleaning commodities, 4 as a fine chemical building block, 5 and has antibacterial 6 and antitumoral activities. 7,8 Consequently, there is an intense search for more effective synthetic ways to obtain α-terpineol. 4,9 Different methods have been described in the literature, using both monoterpenes and oxygenated terpenes as starting materials under acidic conditions (Scheme1). 4,[9][10][11][12][13] Scheme 1: Different starting materials for -terpineol synthesis. 3 In this context, the development of chemo selective synthesis of -terpineol has, as main challenge, to avoid degradation and isomerization products which current leads to low yield and selectivities. 14 On the other hand, the adoption of continuous flow technology for the synthesis of natural product on safer and more efficient conditions has become popular.
Providing better control on reaction parameters such as mixing, mass and heat transfer and online purification, downstream processing help minimize solvent usage, waste and manual handling. [15][16][17] In this context, our research group 16,18,19 have been involved in the development of flow chemistry methodologies for organic synthesis and biocatalysis 15,16 and here in we report our effort on optimizing -terpineol synthesis starting from readily available starting materials (limonene and -pinene) by the use of continuous flow technology. 10,20,21

Materials
Reagents were purchased from different sources and used without further purification: Limonene 98 % from ER do Brasil and -pinene 98 % from Alfa Aesar. Chloroacetic acidfrom Vetec, trifluoracetic acid from Sigma-Aldrich. Cyclohexane and methanol were purchased from Tedia. 1 H-NMR was recorded on a Bruker Advance 500 MHz spectrometer. Reported chemical shifts (δ) are expressed in parts per million (ppm) down field from tetra methyl silane (TMS).

Chromatography Analysis
Samples were prepared by stirring 15 μL of reaction crude and 985 μL of ethyl acetate.

Batch synthesis of -terpineol from -pinene
To a 4 ml flask was added a mixture of -pinene (1.58 mL, 10 mmol) in water (0.4 ml), the mixture was stirred and heated to 70 °C. Then, chloroacetic acid (0.94 mL, 10 mmol) was added and the reaction lasted 4 h. The reaction was monitored using thin layer chromatography with a mixture of ethyl acetate/ hexane 30 % as eluent. Then, the reaction mixture was diluted in 10 ml of ethyl acetate, washed with a 20 % K2CO3 solution (10 mL). The aqueous phase was re-extracted with ethyl acetate

Continuous flow synthesis of -terpineol from -pinene
In a flow line, -pinene flow through backflow regulator (Swagelok SS-4C-1/3) and is mixed with a second stream of an aqueous solution chloroacetic acid (27 mol. L -1 ) into a Tmixer. The combined stream then flows through homemade static mixer (stainless column filled 5 with glass wool 100 mg) and reacted into PTFE reactor coil 16 mL (diameter 0.01 mm) externally heated to 80 ºC during 15 min. The reagents were pumped to maintain a 1:1 molar reaction between acid and substrate the total flow was 1.12 mL/min, being 60 % of the flow rate from the -pinene ( 0.68 mL/min) and 40 % of the flow rate from the 27 M acid solution ( 0.44 mL/min), resulting in conversion values of 72 % and 76 % selectivity.

Batch synthesis of -terpineol from limonene
The -terpineol synthesis using limonene as starting material was carried out in two steps.
1º step: -Terpenyl Trifluoroacetate 22 To a solution of limonene (1.62 mL, 10 mmol) in cyclohexane (10 mL), under constant stirring, trifluoroacetic acid (10 mmol, 0.76 mL) was added slowly at room temperature. After 1 h, the reaction mixture was diluted in 10 mL of ethyl acetate and washed with a 5 % NaHCO3 solution (10 mL). The organic phase was separated, dried over anhydrous sodium sulfate, filtered and the solvent was evaporated under reduced pressure. Crude α-terpenyl trifluoroacetate was obtained as light brown oil (1.6 g).

Results and Discussion
We began our work evaluating batch reactions already described in literature for terpineol synthesis starting from readily available monoterpenes -pinene and limonene. The reaction studied for -terpineol synthesis starting from -pinene was reproduced according to  As a second strategy we decided to evaluate the approach of starting from limonene for -terpineol synthesis, which requires a two-step methodology consisted in an oxidation reaction  Table S2). These results showed that after 30 min of reaction time no further 9 enhancement on conversion and selectivity values were observed. Moreover, other experiments evaluating molar ratio condition of limonene and trifluoroacetic acid were also tested on batch condition ( Table 2). Reaction using limonene (10 mmol) and CF3CO2H 10 mmol (1:1), at room temperature, for 30 min. *All conversion and selectivity values were determined by a GC/MS considering the substrate, limonene, consumption.
In our experiments a small decrease on selectivity was detected when 1:2 molar ratio (limonene: trifluoroacetic acid) was used (Entry 3, Table 1). It is important to highlight that with the aim of finding a better condition to continuous-flow process, we decided to evaluate a solvent free reaction, giving excellent conversions and good selectivity's (Entry 4, Table 2).
Considering preliminary results, this reactional condition has become a very interesting protocol for process intensification. Therefore, a study monitoring the reaction time on solventfree condition allowed us to observe that after only 5 min 93 % of conversion was already achieved with very good selectivity, 89 % (supporting information, Table S3).
As mentioned before, in order to arrive at the desired product, we need to run a two-step reaction. Since the ester intermediate is very unstable in acidic media, a cascade batch process is needed to fully understand the potential of this solvent free approach (Scheme 2). Therefore, the methanolic sodium hydroxide solution was added directly to the reaction media after the first step reaction time and samples were taken to follow product formation. After 40 min of total reaction time, the reaction has already reached maximum conversion (97 %) and excellent selectivity (93 %).

Scheme 2:
Two-step cascade batch reaction of the -terpineol synthesis.
With this results in hands we decided to move forward in order to translate batch protocol to a continuous-flow cascade process. Firstly, the reaction first step was study in flow conditions (supporting information, Table S4) and later on, the second step was assembled. The Residence time on the first step had a small change compared to the optimization protocol in order to have a flow rate where we could meet the second step requirements of residence time.
For the second reaction, mixing is a crucial step, so we decided to have an extended residence time in order to accomplish the hydrolysis reaction. Space time yield obtained for this cascade process is 0.12 Kg.day -1 , lower than the one obtained for the continuous-flow strategy starting from -pinene (4). The final compound can be easily purified by distillation from reaction crude mixture.

Conclusions
Based on the results presented, it was possible to develop two processes for the synthesis of -terpineol in continuous flow. It was possible to carry out the synthesis of a-terpineol in continuous flow using -pinene as starting material and chloroacetic acid in molar ratio 1:1, at 80 ºC with a total residence time of 15 min, obtaining good conversion values (72 % ± 2.45) and selectivity (76 % ± 1.25). These results proved to be much more interesting than those obtained in batch, where the reactions were carried out at 70 ºC for 4 h resulting in 88 % conversion and 67 % selectivity. Although the conversion value was higher for the batch reaction, in the continuous flow system the reaction time was reduced in 94 %, providing a huge increase in the efficiency of the reaction, resulting in a productivity of 0.67 Kg.day -1 under the best conditions found.
For the two-step cascade reaction to the obtainment of -terpineol starting from limonene, excellent conversion (97 % ± 0.47) and selectivity (80 % ± 1.25) results were presented. The advantages of this reaction system were: the first step was carried out without solvent, the second was carried in aqueous solution, and the hole processes could be done at room temperature, and the total residence time was of 40 min. As described, in batch, the total reaction time was of 2.5 h and resulted in 56 % conversion and 81 % selectivity. The productivity of this flow system was 0.12 Kg.day -1 .

Acknowledgment
Authors thanks CNPq, CAPES and FAPERJ for financial support