Skip to main content
NCBI home page
JACS Au logoLink to JACS Au
. 2026 Feb 25;6(3):1676–1684. doi: 10.1021/jacsau.5c01570

Flow Chemistry Enabled Asymmetric Total Synthesis of (+)-Biotin through Pseudocatalytic Strategy

Xianjing Zheng †,§, Liang Gao †,§, Yajiao Zhang †,‡, Jiale Wu †,§, Dang Cheng †,‡, Li Wan †,§,*, Fener Chen †,‡,§,*
PMCID: PMC13014231  PMID: 41889743

Abstract

Continuous-flow chemistry has greatly expanded the capacity for tailored molecular construction, yet the multistep synthesis of natural products and small-molecule pharmaceuticals remains constrained by solvent and reagent incompatibilities, byproduct accumulation, reactor clogging, and mismatched time scales. Through the deliberate integration of multiple purpose-devised tubular continuous extractors, in-line liquid/liquid separators, fixed-bed drying units, and in-line waste-gas vent devices, we have established a fully coupled continuous-flow regime that enables the 13-step asymmetric total synthesis of (+)-biotin with a 43% overall yield and a total residence time of only 121.5 min. The process telescopes a newly developed chiral auxiliary-controlled desymmetrization to construct (3aS,6aR)-lactone in a pseudocatalytic closed-loop cycle through a customized continuous recycling system of the chiral auxiliary, enabling multiple equivalents of chiral product to be produced per molecule of auxiliary, together with a Pd-catalyzed Fukuyama coupling for rapid side-chain assembly, and a final one-flow debenzylation/ring-opening and ring-closing sequence. This work establishes a new paradigm in asymmetric flow chemistry by integrating closed-loop recycling of a chiral auxiliary with unified reaction-separation modules, transforming multistep synthesis into a seamless and sustainable platform for complex molecule production.

Keywords: biotin, pseudocatalytic, flow synthesis, multistep, closed-loop

graphic file with name au5c01570_0013.jpg

graphic file with name au5c01570_0012.jpg

1. Introduction

The natural vitamin (+)-biotin (1) was isolated by Kögl and co-workers from egg yolk in 1936. , This intriguing natural product possesses a highly functionalized and characteristic sulfur-containing bicyclic ureido framework, bearing three contiguous stereocenters in an all-cis configuration and a carboxybutyl side chain. − It serves as an essential coenzyme role in metabolic cycle, causing catalytic carbon dioxide fixation (carboxylation reaction) involved in gluconeogenesis, transcarboxylation-related processes and fatty acid synthesis. , Biotin deficiency in both humans and animals causes a series of severe symptoms that can be remedied through biotin supplementation as a dietary and feed additive. The current global demand for 1 exceeds 600 tons per year and serves diverse industrial end uses across pharmaceuticals, nutraceuticals, animal feed, cosmetics, personal care, and beverages.

The notable bioactivities in human nutrition and animal health, together with the structurally intriguing framework of 1, have stimulated extensive synthetic efforts for more than 70 years toward its total synthesis. , To date, over 60 elegant approaches for the construction of this vitamin have been established through the development of diverse and creative methodologies and strategies. , Among these synthetic endeavors, (3aS, 6aR)-lactone (Roche’s lactone 2), obtained from the chiral auxiliary-controlled desymmetrization of cyclic anhydride 4, has been employed as a key chiral building block for the commercial production of 1. , However, the current protocol for the chiral auxiliary-controlled ring-opening of cyclic anhydride 4 suffers from either modest diastereoselectivity, or insufficient recovery (<90%) of the expensive chiral auxiliary. − Furthermore, factors such as atom- and time- economy, safety, waste generation and operational efficiency remain significant concerns in the subsequent steps toward 1.

Continuous-flow synthesis is emerging as an attractive alternative to overcome some of the limitations associated with the batch process. − The main advantages of continuous-flow synthesis include faster and safer reaction processes, improved environmental sustainability, enhanced product quality control, and the ability to perform chemistry that is difficult or impossible under batch conditions. − Although the continuous-flow synthesis of complex natural products and active pharmaceutical ingredients holds great promise, only a limited number of successful examples have been reported due to the numerous challenging issues inherent in multistep continuous-flow processing platforms, such as solvent and reagent incompatibility, byproduct accumulation, reactor clogging and mismatched time-scales. − Herein, we report a 13-step continuous-flow asymmetric total synthesis of (+)-biotin 1 from commercially available meso-cyclic acid 3. This work represents the first example of the multistep continuous-flow synthesis of 1 and a benchmark case of multistep continuous-flow synthesis with respect to the number of integrated chemical transformations.

2. Results and Discussion

Our 13-step continuous-flow stereocontrolled total synthesis of 1 commenced with the dehydration of commercially available meso-cyclic acid 3 to form meso-cyclic anhydride 4. At the beginning of our investigation, we encountered difficulties in performing the dehydration of 3 under flow conditions using previously reported industrial batch protocol. This method removes water by azeotropic distillation with toluene in a Dean–Stark apparatus, which is incompatible with flow operation. To circumvent this problem, we first attempted to use acetic anhydride as an in-situ water trap. However, treatment of a suspension of 3 in toluene (0.11 M) with a solution of acetic anhydride in toluene (0.5 M) resulted in reactor clogging. Subsequently, a molecular-sieve based protocol was developed that successfully achieved the dehydration of 3 in flow (Table ). After extensive optimization, a solution of 3 in MeCN was pumped through a packed bed containing 4Å molecular sieves (Reactor I) at 170 °C with a residence time of 8 min, and the system was maintained at 21 bar using a back-pressure regulator (BPR), affording better conversion than when THF was used as the solvent (entries 1–3). Because brown-colored impurities were evidently formed in THF and the conversion increased only marginally even when the residence time was extended to 8 min. The desired meso-cyclic anhydride 4 was obtained in 97% conversion and 93% isolated yield (entry 4). Further increase in residence time or reaction temperature was not beneficial for this transformation (entries 5–7). Increasing the concentration of 3 to 0.5 M resulted in reactor clogging (entry 8).

1. Optimization of Dehydration Reaction in a Flow Process .

2.

Entry Solvent t R1 (min) T 1 (°C) Conv. (%)
1 THF 5 170 76
2 THF 8 170 79
3 MeCN 5 170 90
4 MeCN 8 170 97 (93)
5 MeCN 10 170 94
6 MeCN 8 180 83
7 MeCN 8 200 75
8 MeCN 8 170 Clogging
Open in a new tab
a

Reactions were performed on 10 mmol of 3 using the flow setup depicted in the graphic.

b

Determined by HPLC analysis of the crude product.

c

0.11 M THF solution of 3.

d

0.11 M MeCN solution of 3 preheated at 76 °C.

e

0.5 M MeCN solution of 3 preheated at 76 °C.

f

Isolated yield after filtration and washing with cold MeCN.

With an efficient approach to the meso-cyclic anhydride 4 established, our attention next focused on the asymmetric construction of the crucial (3aS, 6aR)-lactone 2 under continuous flow conditions (Scheme ). In the conventional DSM-Tiger batch manufacturing process, the chiral auxiliary–controlled asymmetric alcoholysis of cyclic anhydride followed by selective ester reduction and lactonization constitutes the pivotal step that determines both the stereochemical outcome and the overall production cost of 1. In preliminary studies employing the commonly used chiral auxiliary (S)-1,1-diphenyl-1,2-propanediol ((S)-DPOL, 5a), we observed that its modest water solubility limited its recovery to below 90% after the reductive cleavage of hemiester 6, even after repeated extraction. This incomplete recovery of the costly chiral auxiliary posed a significant challenge to achieving an efficient and sustainable process. To overcome this limitation, we developed a telescoped pseudocatalytic continuous-flow approach under closed-loop control. Central to this design was a newly developed (S)-DPOL analogue exhibiting markedly reduced aqueous solubility, which enables efficient in-line recovery and reuse within the flow cycle.

1. Consecutive Flow Asymmetric Synthesis of 2 .

1

Open in a new tab

To improve auxiliary recovery and enable closed-loop recycling, we designed a series of (S)-1,1-di­(biphenyl)-1,2-propanediol analogues ((S)-DBPOLs, 5b–5g) by replacing the phenyl groups of (S)-DPOL (5a) with bulkier and more hydrophobic biphenyl substituents (Figure ). This structural modification was expected to substantially reduce water solubility through increased hydrophobic surface area and steric hindrance. Computational predictions using ChemBioDraw Ultra 14.0 supported this design rationale, giving Log P values of 6.20–9.68 for the (S)-DBPOLs, compared to 2.85 for 5a, consistent with markedly lower aqueous solubility. These novel (S)-DBPOL-type chiral auxiliaries were readily synthesized from commercially available (S)-ethyl lactate via Grignard addition in 75–88% yield and >99% ee (Supplementary Table S3).

1.

1

Open in a new tab

Structures of (S)-DBPOL-type chiral auxiliaries 5b–5g.

With a library of (S)-DBPOLs chiral auxiliaries in hand, we next examined the asymmetric alcoholysis of meso-cyclic anhydride 4 using auxiliaries 5b–5g under batch conditions. Screening was conducted with purified 4 and tributylamine (n-Bu3N) as base, affording the corresponding hemiesters in 56–88% yields and with >99:1 diastereoselectivity (Supplementary Table S4). Among these, (S)-DBPOL 5c proved to be the most effective auxiliary, and toluene was identified as the optimal solvent, giving the alcoholysis product 6c in 88% yield and >99:1 dr. Auxiliary 5c was also found to be readily recoverable after the subsequent reduction step due to its high solubility in toluene and low solubility in water. This optimized system was then adapted to a continuous-flow setup (Table ). In the initial flow experiment, two toluene solutions, one containing anhydride 4 and the other containing auxiliary 5c and n-Bu3N, were simultaneously pumped into a PTFE coil reactor (Reactor II) at 25 °C with a residence time of 6 min, affording the (4S,5R)-hemiester salt 6c in 60% yield and >99:1 dr (entry 1). Further optimization revealed that adjusting the molar ratio of 4:5c:n-Bu3N from 1:1:1 to 1:1.1:1.5 and extending the residence time from 6 to 10 min enabled complete conversion of 4, providing 6c in 98% yield and 99:1 dr (entries 2–4). The reaction proved insensitive to temperature elevation up to 45 °C (entries 5 and 6), whereas higher substrate concentrations enhanced the space-time yield but led to reactor clogging (entry 7).

2. Effect of Reaction Parameters on the Flow Alcoholysis of Anhydride 4 .

2.

Entry 4:5c:n-Bu3N t R2 (min) T 2 (°C) dr Yield (%)
1 1:1:1 6 25 >99:1 60
2 1:1:1 10 25 >99:1 77
3 1:1:1.3 10 25 >99:1 86
4 1:1.1:1.5 10 25 >99:1 98
5 1:1.1:1.5 10 35 >99:1 97
6 1:1.1:1.5 10 45 >99:1 97
7 1:1.1:1.5 10 25 blockage of reactor
Open in a new tab
a

Reactions were performed on 10 mmol of 4 using the flow setup depicted in the graphic.

b

Determined by chiral HPLC analysis of the crude product.

c

Determined by HPLC analysis of the crude product.

d

0.2 M suspension of 4 in toluene.

Building on these results, we carried out a telescoped continuous-flow sequence combining the alcoholysis and chemoselective reduction to produce alcohol 7 (Scheme ). The process was initiated by merging the outflow from Reactor II with a stock solution of LiBH4 (1.9 equiv) via a T-mixer. Efficient removal of hydrogen byproduct through an in-line tubular gas–liquid separator was essential to ensure stable residence-time control. The reduction proceeded to completion within 10 min in a PTFE coil reactor (Reactor III) at 60 °C under 8.0 bar back pressure. Following aqueous quenching, the resulting emulsified toluene stream was directed through a short glass column packed with Celite and SiO2 to scavenge boron-containing residues. The subsequent gravity-based liquid–liquid separation afforded a clear aqueous phase containing (4S,5R)-lithium hydroxymethyl carboxylate 7 and an organic phase containing the recovered auxiliary 5c and n-Bu3N. This telescoped two-step flow process furnished 7 in 94% yield and 99:1 dr, along with quantitative recovery of 5c and n-Bu3N.

2. Two-Step Synthesis of 7 in Continuous-Flow.

2

Open in a new tab

After validating the two-step alcoholysis/reduction sequence, the next challenge was to establish a continuous-flow pseudocatalytic cycle under closed-loop control for the streamlined synthesis of 2 by integrating the acid-lactonization into the alcoholysis/reduction steps. Three continuous reactions, in-line product separation, and real-time recycling of the auxiliary and n-Bu3N were incorporated into a single continuous-flow process, thereby eliminating time-consuming purification steps, minimizing yield losses in a fully continuous-flow manner (Scheme ). The aqueous stream of crude intermediate 7 from Reactor III, after quenching with water and in-line liquid/liquid separation, was mixed with 1.5 M aqueous HCl at a T-mixer, and the lactonization step was performed in PTFE Reactor IV at 80 °C with a residence time of 5 min under 10 bar back-pressure. After passing through a column packed with Celite and SiO2 to facilitate phase separation, the final toluene stream exiting the in-line extractor delivered the desired (3aS, 6aR)-lactone 2 in 89% overall yield and >98% ee over three steps with a total residence time of 25 min. The mixture of auxiliary 5c and n-Bu3N (1:1.36 molar ratio) was recovered in nearly quantitative yield and continuously recycled for reuse, without the need for further addition of either component (Table , Pass 1).

3. Continuous-Flow Asymmetric Synthesis of 2 Controlled by Chiral Auxiliary 5c with Closed-Loop Recovery.

3

Open in a new tab

3. Results of the Closed-Loop Continuous-Flow Synthesis of 2 via Alcoholysis/Reduction/Lactonization.

Pass Total time (min) Yield (%) ee (%) Recovery, (% Aux5c:n-Bu3N)
1 30 89.3 98.44 99 (1:1.36)
2 180 89.2 98.32 99 (1:1.36)
3 300 89.2 97.92 99 (1:1.37)
Open in a new tab
a

Product collected after the closed-loop flow system reached steady state.

b

Isolated yield of 2.

c

Determined by chiral HPLC analysis of 2.

d

Recovery yield of the 5c/n-Bu3N mixture.

e

The molar ratio of 5c to n-Bu3N was determined by NMR.

Compared with the batch process, this flow approach delivered a higher overall yield (89% over three steps vs 85% in batch), a substantial reduction in processing time to 25 min (batch reaction time: 4.5 h for alcoholysis, 4.5 h for reduction, 2 h for lactonization) and comparable enantioselectivity. The integrated reactor platform realized a pseudocatalytic closed-loop operation that combined three reactions, in-line separation, and auxiliary/base recycling, thereby enhancing operational efficiency and material utilization in the chiral auxiliary-mediated transformation. The robustness of this system was demonstrated by its stable operation: as shown in Table , the telescoped continuous-flow process consistently produced enantiopure 2 in 89.4% yield and 98.3% ee (Pass 2), with no detectable loss in yield or optical purity even after 5 h of continuous running (Pass 3).

Having consolidated the streamlined continuous-flow synthesis of lactone 2, we proceeded to develop a robust continuous-flow strategy for the key intermediate (3aS, 6aR)-thiolactone 8. Although the batch synthesis of 8 using inexpensive potassium ethylxanthate is known to afford high yields, , preliminary attempts to translate this protocol to flow conditions (160 °C, 10 bar back-pressure and 20 min of residence time) resulted in only 30% conversion of 2 (Supplementary Table S7, entry 1). A systematic evaluation of temperature, residence time, reagent ratio and back-pressure failed to improve the conversion of this thiolactonization reaction (Supplementary Table S7, entries 2–7). Thiolactonization of 2 with potassium thioacetate (KSAc) represents an alternative and well-established strategy for accessing 8 in industrial biotin production. However, prior experience from Lonza and our laboratories revealed that 40–45% of KSAc decomposed when the batch reaction was heated at 150–160 °C for about 1 h, affording 8 in low yield. Achieving complete thiolactonization within the shortest possible reaction time at elevated temperature is therefore essential, and the flow process offers a promising solution. Initial flow tests gave incomplete conversion when a 0.0912 M toluene solution of 2 (1.0 equiv) was treated with a 0.24 M solution of KSAc in dry DMF (1.2 equiv) at 150 °C for a residence time of 20 min and under 10 bar back pressure (Table , entry 1). Further optimization revealed that operating at 170 °C and 20 bar back-pressure with 1.3 equiv of KSAc afforded 99% conversion of 2 to 8 within a residence time of only 8 min (entries 2–11). Under these optimized conditions, thiolactonization product 8 was isolated in 92% yield and 99% purity after recrystallization from iso-PrOH.

4. Optimization of Thiolactonization under Continuous-Flow Conditions .

2.

Entry KSAc (equiv) T 5 (°C) t R5 (min) BPR (bar) Conv. (%)
1 1.2 150 20 10 72
2 1.2 160 20 10 84
3 1.2 170 20 10 92
4 1.2 170 15 10 90
5 1.2 170 10 10 87
6 1.2 170 10 12 91
7 1.2 170 10 15 94
8 1.2 170 10 18 97
9 1.2 170 10 20 98
10 1.3 170 10 20 99
11 1.3 170 8 20 99 (92)
Open in a new tab
a

All reactions were performed on 10 mmol of 2 using the flow setup depicted in the graphic.

b

Determined by HPLC analysis of the crude product.

c

Isolated yield.

The next stage involved developing a five-step continuous-flow protocol to install the biotin side chain (Scheme ). This sequence converts 8 into dibenzylbiotin 14 with the desired (3aS, 4S, 6aR)-stereochemistry. The first transformation is palladium-catalyzed Fukuyama cross-coupling between thiolactone 8 and organozinc 10, affording hydroxy ester 11. − The organozinc reagent 10 was prepared in situ from ethyl 5-bromovalerate 9 with active zinc powder in dry DMF under flow conditions (Reactor VI, 240 °C, 8 min residence time, 10 bar; Supplementary Table S8). The homogeneous palladium-catalyzed system provides a versatile platform for fine-tuning catalytic reactivity and markedly outperforms carbon-supported catalysts in this transformation. Five homogeneous palladium catalysts, PdCl2(PPh3)2, Pd2(dba)3, Pd­(PtBu3)2, Pd­(TFA)2, and Pd­(OAc)2 were evaluated using purified thiolactone 8 (Figure ). Initial screening with PdCl2(PPh3)2 (1.0 mol %) led to no conversion, and unreacted 8 was recovered almost quantitative after 15 min at 130 °C under flow conditions. Three other palladium catalysts, (Pd2(dba)3, Pd­(PtBu3)2, and Pd­(TFA)2), exhibithed partial conversion of 8 under identical conditions. Remarkably, Pd­(OAc)2 proved to be the most effictive and robust catalyst for this coupling, achiveving complete conversion of 8 to 11 within 15 min (Supplementary Table S9).

4. Flow Synthesis of (+)-Dibenzylbiotin (14) via a Five-Step Telescoped Sequence.

4

Open in a new tab

2.

2

Open in a new tab

Results of palladium-catalyzed Fukuyama cross-coupling in flow.

Following the successful Fukuyama cross-coupling, the subsequent flow operations were designed to complete the side-chain construction through a series of telescoped transformations. The outlet from Reactor VII was combined with aqueous HCl (6.0 M), and the resulting mixture was introduced into PTFE Reactor VIII for dehydration at 95 °C with a 7.5 min residence time to obtain 12 (Supplementary Table S10). The crude dark-brown stream exiting Reactor VIII was passed through a PTFE column packed with activated carbon to scavenge colored impurities and undesired byproducts. The purified flow was subsequently directed through an in-line liquid–liquid gravity separator to remove the aqueous waste, affording the desired olefin 12 as a clear, nearly colorless toluene solution (Supplementary Figure S2). This toluene solution was then fed into Reactor IX containing a 7.5% Pd/C catalyst, where asymmetric hydrogenation proceeded smoothly within 2.0 min at 130 °C and 40 bar, affording complete conversion of 12 to 13 (Supplementary Table S11). The product stream was collected into a gas/liquid separator to release residual H2, and the resulting solution of ester 13 was subsequently combined with 5.0 M aqueous NaOH in a static mixer and pumped into PTFE Reactor X. Saponification was accomplished at 100 °C and 19 bar, achieving full conversion within 10 min of residence time (Supplementary Table S12). In-line acidification of the resulting mixture with 6.0 M aqueous HCl to pH = 2 was carried out at 0 °C in PTFE Reactor XI with a 4 min residence time. This five-step process efficiently converted 8 into (+)-N, N-dibenzylbiotin 14 in 79.7% overall isolated yield in a total residence time of 40 min.

The synthesis concluded with a telescoped three-step continuous-flow transformation of 14 to the final product 1 (Scheme ). It began with the cascade debenzylation/ring-opening of 14 using 47% aqueous HBr under flow conditions to generate the diamine·2HBr salt 15. This reaction was conducted in PEEK Reactor XII at 190 °C and 20 bar with a residence time of 20 min, affording 99% conversion after systematic optimization of temperature, pressure, and residence time (Supplementary Table S13). For long-term operation, the reactor was enclosed in a transparent explosion-proof shield as a safety precaution, while tantalum tubing is recommended for industrial scale-up due to its superior pressure tolerance, corrosion resistance, and durability, albeit at a higher material cost. Following gravity-based liquid–liquid separation and in-line vacuum evaporation for HBr recovery, the crude 15 was dissolved in 5% aqueous KOH owing to its limited solubility in water. The resulting stream was premixed with a toluene solution of triphosgene (BTC, 0.5 M) in PTFE coil reactor at room temperature, and subsequently directed into PTFE Reactor XIII for ring-closing at 70 °C and 20 bar (residence time 14 min, pH 10–11). The telescoped three-step sequence delivered (+)-biotin 1 in 71.0% isolated yield and 99% purity within a total residence time of 34 min.

5. Flow Synthesis of (+)-Biotin (1) via a Three-Step Telescoped Sequence.

5

Open in a new tab

After successful optimization of each individual module, a fully integrated 13-step continuous-flow sequence was established for the asymmetric total synthesis of (+)-biotin 1 from meso-cyclic acid 3 (Scheme ). The sequence commenced with the dehydration of 3 in MeCN (0.11 M) over 4Å molecular sieves packed in Reactor I (170 °C, 21 bar, 8 min), affording meso-cyclic anhydride 4 after in-line evaporation and solvent exchange to toluene. The resulting toluene suspension of 4 (0.128 M) was combined with a toluene solution of auxiliary 5c (0.141 M) and n-Bu3N (0.192 M) and passed through Reactor II to effect alcoholysis at 25 °C with a residence time of 10 min. The output was merged with a toluene solution of LiBH4 (0.476 M, 1.9 equiv) via a T-mixer, followed by in-line gas–liquid separation to remove H2. The resulting stream was directed into Reactor III (60 °C, 8 bar, 10 min) to afford alcohol 7. A pseudocatalytic closed-loop flow cycle integrating alcoholysis, reduction, and acid-lactonization was then established, incorporating in-line purification, gravity-based liquid–liquid separation, and real-time recycling of 5c and n-Bu3N after in-line condensation and drying. The aqueous stream of 7 was combined with 1.5 M HCl and toluene in a cross-mixer and introduced into Reactor IV (80 °C, 10 bar, 5 min) to furnish Roche’s lactone 2. After in-line liquid–liquid gravity separation, the toluene phase containing 2 was passed through a Na2SO4 drying column, then merged with a DMF solution of KSAc (0.24 M, 1.3 equiv) and introduced into Hastelloy C-276 Reactor V (170 °C, 18 bar, 8 min) to yield thiolactone 8. After in-line liquid–liquid gravity separation to remove aqueous waste, the toluene phase containing 8 was directly mixed with Pd­(OAc)2 (1 mol %) in DMF and the organozinc reagent 10, which was prepared in situ by passing ethyl 5-bromovalerate 9 through a packed-bed Reactor VI (filled with Zn and SiO2; 240 °C, 10 bar, 8 min). The mixture underwent palladium-catalyzed Fukuyama cross-coupling in Reactor VII (130 °C, 18 bar, 15 min) to give hydroxy ester 11. The output was subsequently treated with 6.0 M HCl in Reactor VIII (95 °C, 18 bar, 7.5 min) for dehydration to 12, which was passed through an activated carbon column and an in-line liquid–liquid separator to remove aqueous zinc waste. The toluene solution of 12 was then hydrogenated with hydrogen in packed-bed Reactor IX (7.5% Pd/C and SiO2, 130 °C, 40 bar, 2 min) to give 13. After in-line gas–liquid separation, the stream was combined with 5.0 M NaOH in a static mixer and directed into Reactor X (100 °C, 19 bar, 10 min) for saponification, followed by in-line acidification with 6.0 M HCl in Reactor XI (0 °C, 4 min, pH = 2) to yield carboxylic acid 14. The aqueous solution of 14 was next subjected to debenzylation/ring-opening with 47% HBr in Reactor XII (190 °C, 20 bar, 20 min), producing the diamine·2HBr salt 15 after in-line separation and HBr recovery. The crude 15 was dissolved in 5% aq. KOH and fed into a PTFE coil reactor for premixing with a toluene solution of triphosgene (BTC, 0.5 M) at 20 °C for 3 min. The resulting stream was then combined with the stream of 40% aq. KOH to promote the ring-closing reaction in Reactor XIII (70 °C, 20 bar, 11 min) under pH 10–11. Following in-line gas–liquid separation and standard off-line operations (decoloration, concentration and crystallization), (+)-biotin 1 was obtained in 43% overall yield and 99% purity. The total residence time of the 13-step sequence was 121.5 min, and steady-state operation was reached within 317 min. The process was continuously operated for 4 h under steady-state conditions, producing 2.745 g of 1, corresponding to a production rate of 0.686 g h–1.

6. Integrated 13-Step Continuous-Flow Asymmetric Total Synthesis of (+)-Biotin (1).

6

Open in a new tab

3. Conclusion

In summary, we have developed a fully integrated and operationally robust platform that accomplishes the 13-step asymmetric total synthesis of (+)-biotin in continuous flow. The process delivers (+)-biotin in 43% overall yield and 99% purity within a total residence time of 121.5 min and maintains steady-state operation for over 4 h with a stable productivity of 0.686 g·h–1. By combining 13 reaction and separation modules, including dehydration, pseudocatalytic desymmetrization with closed-loop chiral auxiliary recycling, Pd-catalyzed Fukuyama coupling, and a telescoped debenzylation/ring-opening and ring-closing sequence, within a fully coupled modular flow architecture, the system minimizes manual handling, solvent switching, and waste generation while ensuring reproducibility and safety. At the heart of this achievement lies a pseudocatalytic closed-loop strategy that enables continuous chiral auxiliary turnover through the deliberate convergence of asymmetric synthesis and flow reactor design, offering a broadly applicable model for sustainable multistep asymmetric synthesis.

Supplementary Material

au5c01570_si_001.pdf (4.8MB, pdf)

Acknowledgments

Financial support from the Start-up Fund of Fudan University (No. JIH1615026Y) and Consulting Project on Science and Technology Strategy of the Chinese Academy of Engineering (2025-XZ-27) is acknowledged. We thank Shenzhen E-Zheng and Hangzhou Morotchem for the assistance with flow equipment.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.5c01570.

  • Additional experimental details, materials, and methods, including NMR spectra for compounds and photographs of experimental setup (PDF)

#.

X. Z., L. G., and Y. Z. contributed equally to this work. L. W. and F. C. designed the project and wrote the original draft. J. W. and D. C. helped with data collection.

The authors declare no competing financial interest.

Published as part of JACS Au special issue “Continuous Flow Chemistry”.

References

  1. Kögl F., Tönnis B.. Over the Bios-Problem. Discovery of Crystallized Biotin from the Eigelb. 20. Mixture over flat Wachstumsstoffe Hopp Seyler Z. Physiol. Chem. 1936;242(1–2):43–73. doi: 10.1515/bchm2.1936.242.1-2.43. [DOI] [Google Scholar]
  2. Lis K.. Biotin Supplementation-The Cause of Hypersensitivity and Significant Interference in Allergy Diagnostics. Nutrients. 2025;17(15):2423. doi: 10.3390/nu17152423. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Eggersdorfer M., Laudert D., Letinois U., McClymont T., Medlock J., Netscher T., Bonrath W.. One Hundred Years of Vitamins-a Success Story of the Natural Sciences. Angew. Chem., Int. Ed. 2012;51(52):12960–12990. doi: 10.1002/anie.201205886. [DOI] [PubMed] [Google Scholar]
  4. Vigneaud V.. The Structure of Biotin. Science. 1942;96(2499):455–461. doi: 10.1126/science.96.2499.455. [DOI] [PubMed] [Google Scholar]
  5. Trotter J., Hamilton J. A.. The Absolute Configuration of Biotin. Biochemistry. 1966;5(2):713–714. doi: 10.1021/bi00866a044. [DOI] [PubMed] [Google Scholar]
  6. Waldrop G. L., Holden H. M., Maurice M. S.. The Enzymes of Biotin Dependent CO2 Metabolism: What Structures Reveal about Their Reaction Mechanisms. Protein Sci. 2012;21(11):1597–1619. doi: 10.1002/pro.2156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Karalia S., Meena V. K.. Biotin: DNA to Diet. J. Nutr. Biochem. 2025;146:110081. doi: 10.1016/j.jnutbio.2025.110081. [DOI] [PubMed] [Google Scholar]
  8. Johnson A. R., Hood R. L., Emery J. L.. Biotin and the Sudden Infant Death Syndrome. Nature. 1980;285(5761):159–160. doi: 10.1038/285159a0. [DOI] [PubMed] [Google Scholar]
  9. Decode the Future of Biotin, https://www.chemanalyst.com/industry-report/biotin-market-3087?utm_source=chatgpt.com (accessed 2025–11–04).
  10. Clercq P. J. D.. Biotin: A Timeless Challenge for Total Synthesis. Chem. Rev. 1997;97(6):1755–1792. doi: 10.1021/cr950073e. [DOI] [PubMed] [Google Scholar]
  11. Atodiresei I., Schiffers I., Bolm C.. Stereoselective Anhydride Openings. Chem. Rev. 2007;107(12):5683–5712. doi: 10.1021/cr068369f. [DOI] [PubMed] [Google Scholar]
  12. Seki M.. Biological Significance and Development of Practical Synthesis of Biotin. Med. Res. Rev. 2006;26(4):434–482. doi: 10.1002/med.20058. [DOI] [PubMed] [Google Scholar]
  13. Casutt M., Koppe T., Schwarz M.. Vitamins, 9. Biotin. Ullmann’s Encyclopedia of Industrial Chemistry. 2011;38:309–320. doi: 10.1002/14356007.o27_o12. [DOI] [Google Scholar]
  14. Deroose F. D., Clercq P. J. D.. Novel Enantioselective Syntheses of (+)-Biotin. J. Org. Chem. 1995;60(2):321–330. doi: 10.1021/jo00107a009. [DOI] [Google Scholar]
  15. Bonrath, W. ; Reinhard, K. ; Netscher, T. ; Roessler, F. ; Spindler, F. . Biotin-The Chiral Challenge. Chimia 2009, 63 (5), 265 10.2533/chimia.2009.265. [DOI] [Google Scholar]
  16. Wehrli, C. A Process for the Stereoselective Synthesis of Lactones. WO2004094367.
  17. Tonyonaka, Y. A. ; Takarazuka, H. S. ; Nishinomiya, H. A. ; Kawanishi, S. O. . Method for Synthesis of Optically Active Lactones. US3876656, 1975.
  18. Yang, L. Y. ; Shuang, Z. Y. ; Jing, Z. Z. ; Cha, P. Z. ; Peng, W. C. . A Method for Recovering Dextroamine. CN109456206, 2018.
  19. Cole K. P., Groh J. M., Johnson M. D., Burcham C. L., Campbell B. M., Diseroad W. D., Heller M. R., Howell J. R., Kallman N. J., Koenig T. M.. et al. Kilogram-Scale Prexasertib Monolactate Monohydrate Synthesis under Continuous-flow CGMP Conditions. Science. 2017;356(6343):1144–1150. doi: 10.1126/science.aan0745. [DOI] [PubMed] [Google Scholar]
  20. Liao J. Y., Zhang S. L., Wang Z. S., Song X., Zhang D. L., Kumar R., Jin J., Ren P., You H. Z., Chen F. E.. Transition-Metal Catalyzed Asymmetric Reactions under Continuous Flow from 2015 to Early 2020. Green Synth Catal. 2020;1(2):121–133. doi: 10.1016/j.gresc.2020.08.001. [DOI] [Google Scholar]
  21. Zhu K. J., Jiang M. F., Ye B. J., Zhang G. T., Li W. J., Tang P., Huang Z. D., Chen F. E.. A Unified Strategy to Prostaglandins: Chemoenzymatic Total Synthesis of Cloprostenol, Bimatoprost, PGF­(2alpha), Fluprostenol, and Travoprost Guided by Biocatalytic Retrosynthesis. Chem. Sci. 2021;12(30):10362–10370. doi: 10.1039/D1SC03237B. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Saito Y., Nishizawa K., Laroche B., Ishitani H., Kobayashi S.. Continuous-Flow Synthesis of (R)-Tamsulosin Utilizing Sequential Heterogeneous Catalysis. Angew. Chem., Int. Ed. 2022;61(13):202115643. doi: 10.1002/anie.202115643. [DOI] [PubMed] [Google Scholar]
  23. Jiang M. F., Liu M. J., Huang H. S., Chen F. E.. Fully Continuous Flow Synthesis of 5-(Aminomethyl)-2-methylpyrimidin-4-amine: A Key Intermediate of Vitamin B. Org. Process Res. Dev. 2021;25(10):2331–2337. doi: 10.1021/acs.oprd.1c00253. [DOI] [Google Scholar]
  24. Pastre J. C., Browne D. L., Ley S. V.. Flow Chemistry Syntheses of Natural Products. Chem. Soc. Rev. 2013;42(23):8849–8869. doi: 10.1039/c3cs60246j. [DOI] [PubMed] [Google Scholar]
  25. Britton J., Raston C. L.. Multi-step Continuous-flow Synthesis. Chem. Soc. Rev. 2017;46(5):1250–1271. doi: 10.1039/C6CS00830E. [DOI] [PubMed] [Google Scholar]
  26. Snead D. R., Jamison T. F.. A Three-minute Synthesis and Purification of Ibuprofen: Pushing the Limits of Continuous-flow Processing. Angew. Chem., Int. Ed. 2015;54(3):983–987. doi: 10.1002/anie.201409093. [DOI] [PubMed] [Google Scholar]
  27. Adamo A., Beingessner R. L., Behnam M., Chen J., Jamison T. F., Jensen K. F., Monbaliu J. C., Myerson A. S., Revalor E. M., Snead D. R.. et al. On-demand Continuous-flow Production of Pharmaceuticals in A Compact, Reconfigurable System. Science. 2016;352(6281):61–67. doi: 10.1126/science.aaf1337. [DOI] [PubMed] [Google Scholar]
  28. Ziegler R. E., Desai B. K., Jee J. A., Gupton B. F., Roper T. D., Jamison T. F.. 7-Step Flow Synthesis of the HIV Integrase Inhibitor Dolutegravir. Angew. Chem., Int. Ed. 2018;57(24):7181–7185. doi: 10.1002/anie.201802256. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Russell M. G., Jamison T. F.. Seven-Step Continuous Flow Synthesis of Linezolid Without Intermediate Purification. Angew. Chem., Int. Ed. 2019;58(23):7678–7681. doi: 10.1002/anie.201901814. [DOI] [PubMed] [Google Scholar]
  30. Anderson N. G.. Using Continuous Processes to Increase Production. Org. Process Res. Dev. 2012;16(5):852–869. doi: 10.1021/op200347k. [DOI] [Google Scholar]
  31. Gutmann B., Cantillo D., Kappe C. O.. Continuous-flow Technology-A Tool for the Safe Manufacturing of Active Pharmaceutical Ingredients. Angew. Chem., Int. Ed. 2015;54(23):6688–6728. doi: 10.1002/anie.201409318. [DOI] [PubMed] [Google Scholar]
  32. Plutschac M. B., Pieber B., Gilmore K., Seeberger P. H.. The Hitchhiker’s Guide to Flow Chemistry Parallel. Chem. Rev. 2017;117(18):11796–11893. doi: 10.1021/acs.chemrev.7b00183. [DOI] [PubMed] [Google Scholar]
  33. Li W. J., Jiang M. F., Chen W. C., Chen Y., Yang Z., Tang P., Chen F. E.. Total Synthesis of (−)-Canadine, (−)-Rotundine, (−)-Sinactine, and (−)-Xylopinine Using a Last-Step Enantioselective Ir-Catalyzed Hydrogenation. J. Org. Chem. 2021;86(12):8143–8153. doi: 10.1021/acs.joc.1c00602. [DOI] [PubMed] [Google Scholar]
  34. Xia Y. Q., Jiang M. F., Liu M. J., Zhang Y., Qu H. G., Xiong T., Huang H. S., Cheng D., Chen F. E.. Catalytic Syn-Selective Nitroaldol Approach to Amphenicol Antibiotics: Evolution of a Unified Asymmetric Synthesis of (−)-Chloramphenicol, (−)-Azidamphenicol, (+)-Thiamphenicol, and (+)-Florfenicol. J. Org. Chem. 2021;86(17):11557–11570. doi: 10.1021/acs.joc.1c01124. [DOI] [PubMed] [Google Scholar]
  35. Li W. J., Jiang M. F., Liu M. J., Ling X., Xia Y. Q., Wan L., Chen F. E.. Development of a Fully Continuous-Flow Approach Towards Asymmetric Total Synthesis of Tetrahydroprotoberberine Natural Alkaloids. Chem. – Eur. J. 2022;28(33):e202200700. doi: 10.1002/chem.202200700. [DOI] [PubMed] [Google Scholar]
  36. Rebire F., Kagan H. B.. An Efficient Route to Chiral t-Butyl Sulfoxides. Tetrahedron Lett. 1989;30(28):3659–3662. doi: 10.1016/S0040-4039(01)80475-5. [DOI] [Google Scholar]
  37. Wang B., Liu F. Q.. Synthesis and Properties of A Stimulus-responsive Block Polymer. RSC Adv. 2020;10(48):28541–28549. doi: 10.1039/D0RA05343K. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Chen F. E., Chen X. X., Dai H. F., Kuang Y. Y., Xie B., Zhao J. F.. Synthetic Studies on d-Biotin, Part 8:[1] An Efficient Chemoenzymatic Approach to the Asymmetric Total Synthesis of d-Biotin via a Polymer-Supported PLE-Mediated Desymmetrization of meso-Symmetic Dicarboxylic Esters. Adv. Syn. Catal. 2005;347(4):549–554. doi: 10.1002/adsc.200404311. [DOI] [Google Scholar]
  39. Xiong F., Li J., Li G., Song B., Zhou Y. X., Chen F. E., Li D.. Preparation of (+)-Biotin: Process Development and Scale-Up. Heterocycles. 2016;92(3):544–551. doi: 10.3987/COM-15-13379. [DOI] [Google Scholar]
  40. Tokuyama H., Yokoshima S., Yamashita T., Fukuyama T.. A Novel Ketone Synthesis by A Palladium-catalyzed Reaction of Thiol Esters and Organozinc Reagents. Tetrahedron Lett. 1998;39(20):3189–3192. doi: 10.1016/S0040-4039(98)00456-0. [DOI] [Google Scholar]
  41. Shimizu T., Seki M.. Facile Synthesis of (C)-Biotin via Fukuyama Coupling Reaction. Tetrahedron Lett. 2000;41(26):5099–5101. doi: 10.1016/S0040-4039(00)00781-4. [DOI] [Google Scholar]
  42. Takahashi Y., Seki M.. Finkelstein Reaction in Non-polar Organic Solvents: A Streamlined Synthesis of Organic Iodides. Org. Process Res. Dev. 2021;25(8):1974–1978. doi: 10.1021/acs.oprd.1c00226. [DOI] [Google Scholar]
  43. Mori Y., Seki M.. A Practical Synthesis of Multifunctional Ketones through the Fukuyama Coupling Reaction. Adv. Syn. Catal. 2007;349(11–12):2027–2038. doi: 10.1002/adsc.200600610. [DOI] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

au5c01570_si_001.pdf (4.8MB, pdf)

Articles from JACS Au are provided here courtesy of American Chemical Society

RESOURCES