Selective continuous flow synthesis of hydroxy lactones from alkenoic acids

Abstract

A novel approach to synthesize hydroxy lactones using flow chemistry is reported. The system could be applied to a variety of functionalized alkenoic acids allowing the simple and eco-friendly generation of lactonic products within one hour (50 min). The reaction was optimized in terms of efficiency, productivity, resources and eco-sustainability using an integrated flow process assisted by in-line work-up and purification. In one example, the method was scaled to deliver 22 mmol of the product.

---

Introduction

The broad occurrence of lactone rings in natural and bioactive compounds¹ and in their precursors² makes their preparation an important target for synthesis. Diverse approaches have been reported,³ and although major advances have been made, further criteria need to be met to improve the efficiency and versatility, and to provide eco-friendly protocols.

To this aim, continuous flow chemistry represents an appealing area of research inspiring the development of green⁴ and modern methods.⁵ So far, only two examples have shown the profitable use of flow technology for the preparation of lactones. In particular, Park and collaborators described the palladium-catalyzed diacetoxylation of alkenes using peracetic acid in a microchemical device.⁶ Albeit the method was designed to provide vicinal diacetoxy compounds, in two cases, the reaction afforded five-membered acetoxy lactones. More recently, the photocatalytic/reductive reaction of in situ generated acyl succinate intermediates was explored to prepare γ-lactones in 51–68% yield and a moderate diastereoselectivity ratio.⁷ There is, therefore, still room for improvement in terms of substrate scope, yield and selectivity, as well as the flow set-up for integrated downstream operations.

Given that a batch method was recently reported for the bromolactonization of 4-pentenoic acid by using a selenide-based xerogel in the presence of H₂O₂ and NaBr,⁸ we were intrigued with the possibility of studying a similar tactic for preparing hydroxy lactones under flow conditions. It is worth noting that, to the best of our knowledge, selenium-mediated catalysis has never been investigated in flow systems. Thus, taking inspiration from the oxidative cycle of glutathione peroxidase (GPx), we sought to employ benzeneperseleninic acid (PhSeO₃H, 1) generated in situ from benzeneseleninic acid (PhSeO₂H, 2) and H₂O₂,⁹ to oxidize alkenoic acids into epoxide intermediates that can readily undergo intramolecular cyclization to form the corresponding hydroxy lactones (Scheme 1).

Scheme 1 GPx-inspired mechanism induced by selenium-organic reagents employed for the flow synthesis of hydroxy lactones.

Herein, reaction conditions for translating the process into a flow platform are developed to satisfy the specific requirements of continuous flow reactors in a sustainable fashion. Substrate scope and scaling-out are also described, demonstrating the versatility and robustness of the method.

Results and discussion

To realize the translation of the reaction into a flow process, we decided to divide the work into three steps: (a) definition of the model reaction and screening of key experimental parameters, (b) design of a convenient flow set-up, (c) substrate scope and scale-up. Decisions on the best conditions and improvements were made, taking into consideration the efficiency, resources and eco-sustainability.

Commercially available (E)-3-pentenoic acid (3a) was selected as the model substrate for initial reaction screenings, while PhSeO₂H (2) was employed as the pre-catalyst to form in situ PhSeO₃H (1) by reaction with H₂O₂.⁹ H₂O/acetone (5 : 1, v/v) and EtOAc were the solvents of choice because of their safety profile and ability to solubilize both reagents and products. Reaction screening was performed at room temperature in a modular flow system equipped with loop injection systems, two HPLC pumps, a reactor coil (10 mL), a back pressure regulator (BPR, 100 psi) and a fraction collector (Fig. 1). Thus, a 0.3 M solution of 3a in EtOAc and a solution of 2 (0.1–0.5 equiv.) and H₂O₂ (30% wt, 5 equiv.) in H₂O/acetone were injected into the loops, mixed in a T-piece and flowed through the reactor coil at 25 °C. The crude reaction mixture was collected and analyzed by ¹H-NMR analysis for the determination of the reaction yield.

Fig. 1 General flow set-up employed during the reaction optimization. BPR: back pressure regulator; FC: fraction collector; L: loop injector; P_1–2: pumps; R: 10 mL reactor coil.

Initially we have evaluated the effect of the amount of 2 on the reaction outcome (Table 1). The best result was obtained using 0.5 equiv. of 2 affording 5a in nearly quantitative conversion in 50 min. Intrigued by the high conversion of 3a into the epoxy intermediate 4a (Table 1, entry 2), we then envisaged the possibility of promoting the cyclization of 4a into 5a by acidic catalysis using a volatile acid such as formic acid (10% mol).¹⁰ This would allow us to reduce the amount of selenium reagent needed also to convert 4a into 5a. Accordingly, the desired lactone 5a was obtained in 77% yield with a conversion of 90% using 0.2 equiv. of PhSeO₂H (2) (Table 1, entry 7). Remarkably, a slight increase of PhSeO₂H (2) (0.3 equiv.) afforded 5a in quantitative yield (Table 1, entry 8),¹¹ while higher flow rates had a negative effect on the reaction performance (Table 1, entry 9).

Table 1 Optimization of the reaction conditions (2)^a

Entry	PhSeO2H (2) (equiv.)	Flow rate (mL min^−1)	Conversion^b (%)	4a/5a^b
a All reactions were conducted according to Fig. 1. Reagents and conditions: 3a [0.6 mmol, 0.3 M in EtOAc], 2 [0.1–0.5 equiv., 0.06–0.3 mmol, 0.03–0.15 M in H2O/acetone (5 : 1, v/v)] and H2O2 [30% wt, 5 equiv., 3 mmol, 1.5 M in H2O/acetone (5 : 1, v/v)], 10 mL coil reactor, 25 °C. b Determined by ^1H-NMR analysis of the crude reaction mixture.
1	0.1	0.2	85	100/0
2	0.2	0.2	94	100/0
3	0.3	0.2	100	55/45
4	0.4	0.2	100	16/84
5	0.5	0.2	100	0/100
6	0.1 + HCO2H	0.2	88	80/20
7	0.2 + HCO2H	0.2	90	23/77
8	0.3 + HCO2H	0.2	100	0/100
9	0.3 + HCO2H	0.3	86	0/100

---

Having established the best conditions (Table 1, entry 8), we looked at the downstream section of the process. Besides efficiency, we also sought to simplify as much as possible the flow protocol in order to keep the environmental factor, costs and handling as low as possible. Thus, a continuous liquid–liquid membrane-based separator (S) was employed to separate the aqueous phase from the organic solution (Fig. 2). The organic layer was then directed towards a glass column (C) (Omnifit Labware DIBA HIT column, L × ID 6.6 mm × 100 mm) packed with Amberlyst A21 and silica to catch the selenium-containing reagents and traces of the starting material (Fig. S1, ESI†). The outflow was collected, furnishing 5a in 90% isolated yield with high purity (Table 2, entry 1). Importantly, no traces of selenium and 3a were detected proving the efficiency of the catch system. After reaction, two valves (V₁ and V₂) ensured the washing of the A21 packed column with an ethanolic solution of NH₄OH (5%, v/v), and enabled the recovery of both the catalyst and the unreacted starting material that can be reused for other reactions (Fig. 2).

Fig. 2 Integrated flow set-up for the synthesis and purification of hydroxy lactones 5a–m. BPR: back pressure regulator; P_1–2: pumps; R: 10 mL reactor coil; S: liquid–liquid membrane separator; SC_1–2: scavenger column; V_1–4: valves.

Table 2 Flow synthesis of hydroxy lactones 5a–m^a

Entry	Substrate	Products	Yield^b (%)	d.r.^c
a All reactions were conducted according to Fig. 2. b Isolated yield. c Diastereomeric ratio (d.r.) was determined by ^1H-NMR analysis. d 4-Hexenoic acid (3h) was used in the commercially available mixture constituted by (E)- and (Z)-isomers in an 8 : 2 ratio; the E and Z ratio of 3h was maintained during the formation of the corresponding S- and R-product 5h. n.a.: not applicable.
1	3a	5a	90	>99 : 1
2	3b	5b	80	n.a.
3	3c	5c	74	n.a.
4	3d	5d	79	n.a.
5	3e	5e	75	95 : 5
6	3f	5f	92	>99 : 1
7	3g	5g	75	>99 : 1
8^d	3h	5h	78	80 : 20^d
9	3i	5i	67	n.a.
10	3j	5j	82	>99 : 1
11	3k	5k	78	n.a.
12	3l	5l	77	n.a.
13	3m	5m	76	n.a.

---

We next evaluated the substrate scope of the reaction using different alkenoic acids 3b–m (Table 2). Reactions were performed according to the improved flow set-up depicted in Fig. 2. As a result, the method achieved efficient and versatile furnishing pure hydroxy lactones 5a–m in good to high isolated yields. Both alkyl, cycloalkyl and aromatic substituents, and more constrained substrates were well tolerated.

While the transient epoxyde intermediate is responsible for the exclusive formation of trans stereoisomers, the regiochemistry of the reaction can be explained according to previous reports.¹² Indeed, substrates 3a, 3d and 3e (ref. 13) (Table 2, entries 1, 4 and 5) reacted by an anti-Baldwin like mechanism leading to 5-endo products, while 5-exo products were favored over 6-endo analogs (Table 2, entries 2, 8 and 10–13). Analogously, 6-exo ring closure was preferred over the 7-endo path (Table 2, entry 3), and 7-exo ring closure was favored over the 8-endo (Table 2, entry 9). Furthermore, endocyclic compounds 3f and 3g (Table 2, entries 6 and 7) furnished stereoselectively the corresponding cis-bicyclic lactones 5f and 5g.¹³ It is interesting to note that no by-products were detected (for instance, the epoxyde) in any of the examples, with water being the only by-product generated during the reaction.

The method was also easily scalable, since we were able to continuously process 22 mmol of 3-hydroxy-γ-butyrolactone (3-BHL) (5d), a useful building block for synthetic transformations and the precursor of cholesterol-reducing drugs such as Crestor™ and Lipitor™, the antibiotic Zyvox™, and the anti-hyperlipidemic Zetia™.^1c A continuous flow preparation of 5d employing high pressure hydrogenation of malic acid over a ruthenium-based catalyst has been recently disclosed;¹⁴ however, this process employed hazardous conditions, expensive catalyst and purifications. Other chemical and chemo-enzymatic approaches also suffer from several limitations,¹⁵ resulting in the high cost of the product. More recently, a biosynthetic pathway to 3-BHL (5d) using acyl-CoA enzymes has been described;¹⁶ although fashionable, the efficacy of the route for large scale operations needs to be proven. In our case, we have used the integrated flow system illustrated in Fig. 2 using four reactor coils operating in parallel and an additional A21 packed column for catalyst scavenging. To avoid resin saturation, the column was washed after 3.5 h of continuous pumping (when about half of the reservoir solution of 3d was reacted). Switch to the second column then guaranteed the purification of the rest of the reaction mixture. Under these conditions, 22 mmol of 5d were continuously synthesized in high purity with a productivity of 2.45 mmol h⁻¹ (corresponding to 6 g d⁻¹).

Conclusions

We have developed an efficient and eco-friendly strategy for the regio- and diastereo-selective continuous synthesis and purification of hydroxy lactones. To the best of our knowledge, the present study is the only case reported and carried out that uses selenium-mediated catalysis under flow conditions. The results clearly show that our method is robust, reliable and versatile and permits the large scale preparation of important synthons and natural products such as 3-BHL (5d) and 5-hydroxy-4-decanolide (5j), a sex pheromone of the parasitic wasp genus Nasonia. In particular, 3-BHL (5d) was continuously produced (isolated yield: 74%) with high purity without the need of human intervention for work-up and purifications. Overall, this approach was superior to previously reported synthesis^3a,9 and more environmentally friendly due to the intrinsic atom economy of the reaction, the use of green solvents and mild reaction conditions, it being safe and amenable for automation, the formation of water as the only by-product of the reaction, and the possibility of recovering and reusing unreacted starting materials and selenium-containing reagents. Current efforts are directed towards the implementation of this method to self-controlled multistep synthesis and bio-enzymatic transformations whose results will be reported in due course.

Acknowledgements

Fondazione Cassa di Risparmio di Perugia (2014.01000.021) is gratefully acknowledged for the financial support.

Notes and references

1. (a) Y.-M. Ham, Y.-J. Ko, S.-M. Song, J. Kim, K.-N. Kim, J.-H. Yun, J.-H. Cho, G. Ahn and W.-J. Yoon, J. Funct. Foods, 2015, 13, 80; (b) N. P. Lopes, M. J. Kato and M. Yoshida, Phytochemistry, 1999, 51, 29; (c) M. K. Wikström and J. A. Berden, Biochim. Biophys. Acta, 1972, 283, 403; (d) S.-H. Lee and O.-J. Park, Appl. Microbiol. Biotechnol., 2009, 84, 817.
2. (a) M. Seitz and O. Reiser, Curr. Opin. Chem. Biol., 2005, 9, 285; (b) S. Gil, M. Parra, P. Rodriguez and J. Segura, Mini-Rev. Org. Chem., 2009, 6, 345.
3. (a) C. Le Floch, E. Le Gall, E. Léonel, T. Martens and T. Cresteil, Bioorg. Med. Chem. Lett., 2011, 21, 7054; (b) M. Mondal, H. J. Ho, N. J. Peraino, M. A. Gary, K. A. Wheeler and N. J. Kerrigan, J. Org. Chem., 2013, 78, 4587; (c) C. Shu, M.-Q. Liu, Y.-Z. Sun and L.-W. Ye, Org. Lett., 2012, 14, 4958; (d) E. Howard and K. A. Woerpel, Tetrahedron, 2009, 65, 6447; (e) H. S. Park, D. W. Kwon, K. Lee and Y. H. Kim, Tetrahedron Lett., 2008, 49, 1616; (f) G. Kokotos, Org. Lett., 2013, 15, 2406; (g) O. Anaç, F. S. Güngör and G. Merey, Helv. Chim. Acta, 2006, 89, 1231; (h) O. V. Turova, E. V. Starodubtseva, M. G. Vinogradov, V. A. Ferapontov and M. I. Struchkova, Tetrahedron: Asymmetry, 2009, 20, 2121; (i) A. Comini, C. Forzato, P. Nitti, G. Pitacco and E. Valentin, Tetrahedron: Asymmetry, 2004, 15, 617.
4. For recent reviews see: (a) L. Vaccaro, D. Lanari, A. Marrocchi and G. Strappaveccia, Green Chem., 2014, 16, 3680; (b) C. Wiles and P. Watts, Green Chem., 2014, 16, 55; (c) S. G. Newman and K. F. Jensen, Green Chem., 2013, 15, 1456.
5. For recent reviews see: (a) M. B. Plutschack, B. Pieber, K. Gilmore and P. H. Seeberger, Chem. Rev., 2017 DOI:10.1021/acs.chemrev.7b00183 , in press; (b) B. J. Reizman and K. F. Jensen, Acc. Chem. Res., 2016, 49, 1786; (c) A. Adamo, R. L. Beingessner, M. Behnam, J. Chen, T. F. Jamison, K. F. Jensen, J. C. Monbaliu, A. S. Myerson, E. M. Revalor, D. R. Snead, T. Stelzer, N. Weeranoppanant, S. Y. Wong and P. Zhang, Science, 2016, 352, 61; (d) H. P. L. Gemoets, Y. Su, M. Shang, V. Hessel, R. Luqueb and T. Noël, Chem. Soc. Rev., 2016, 45, 83; (e) A. Gioiello, V. Mancino, P. Filipponi, S. Mostarda and B. Cerra, J. Flow Chem., 2016, 6, 167; (f) I. M. Mándity, S. B. Ötvös, G. Szőlősi and F. Fülöp, Chem. Rec., 2016, 16, 1018; (g) I. M. Mandity, S. B. Oetvoes and F. Fueloep, ChemistryOpen, 2015, 4, 212; (h) B. Gutmann, D. Cantillo and C. O. Kappe, Angew. Chem., Int. Ed., 2015, 54, 6688; (i) R. J. Ingham, C. Battilocchio, D. E. Fitzpatrick, E. Sliwinski, J. M. Hawkins and S. V. Ley, Angew. Chem., Int. Ed., 2015, 54, 144; (j) M. Baumann and I. Baxendale, Beilstein J. Org. Chem., 2015, 11, 1194; (k) Y. Su, N. J. W. Straathof, V. Hessel and T. Noel, Chem. – Eur. J., 2014, 20, 10562; (l) J. C. Pastre, D. L. Browne and S. V. Ley, Chem. Soc. Rev., 2013, 42, 8849.
6. J. H. Park, C. Y. Park, H. S. Song, Y. S. Huh, G. H. Kim and C. P. Park, Org. Lett., 2013, 15, 752.
7. M. Fagnoni, F. Bonassi, A. Palmieri, S. Protti, D. Ravelli and R. Ballini, Adv. Synth. Catal., 2014, 356, 753.
8. C. M. Gatley, L. M. Muller, M. A. Lang, E. E. Alberto and M. R. Detty, Molecules, 2015, 20, 9616.
9. L. Sancineto, F. Mangiavacchi, C. Tidei, L. Bagnoli, F. Marini, A. Gioiello, J. Scianowski and C. Santi, Asian J. Org. Chem., 2017 DOI:10.1002/ajoc.201700193 , in press.
10. Other acids including trifluoroacetic acid, solid and supported p-toluensulfonic acid gave similar results. Hydrochloric acid and acetic acid were less effective and formed by-products.
11. When conducted in batch the conversion of 3a to 5a was achieved with a yield of 90% in 8 h using 4 equiv. of H₂O₂ and 5% mol of diphenyl diselenide (PhSe)₂ at room temperature.⁹ On the gram scale a higher amount of (PhSe)₂ or a longer reaction time is however needed to complete the reaction. Moreover, reactions conducted in batch mode requires tedious chromatographic purifications and catalyst recovery is not easily achievable.
12. H. Tan and J. H. Expenson, J. Mol. Catal. A: Chem., 1999, 142, 333.
13. When prepared in batch mode, 5e and 5g are formed in 85% yield with a lower diastereoselectivity ratio (85 : 15).⁹.
14. (a) A. M. Rouhi, Chem. Eng. News, 2003, 81, 55; (b) B. S. Kwak, Chim. Oggi, 2003, 21, 23.
15. (a) P. Kumar, A. N. Deshmukh, R. K. Upadhyay and M. K. Gurjar, Tetrahedron: Asymmetry, 2005, 16, 2717; (b) R. I. Hollingsworth, J. Org. Chem., 1999, 64, 7633; (c) A. Nakagawa, H. Idogaki, K. Kato, A. Shinmyo and T. Suzuki, J. Biosci. Bioeng., 2006, 101, 97; (d) T. Suzuki, H. Idogaki and N. Kasai, Enzyme Microb. Technol., 1999, 24, 13; (e) S. H. Lee, O. J. Park and H. S. Uh, Appl. Microbiol. Biotechnol., 2008, 79, 355.
16. C. H. Martin, H. Dhamankar, H.-C. Tseng, M. J. Sheppard, C. R. Reisch and K. L. J. Prather, Nat. Commun., 2013, 4, 1.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c7re00083a

---