A green and facile synthesis of an industrially important quaternary heterocyclic intermediates for baricitinib

Background Baricitinib, with a 2-(1-(ethylsulfonyl)azetidin-3-yl)acetonitrile moiety at N-2 position of the pyrazol skeleton, is an oral and selective reversible inhibitor of the JAK1 and JAK2 and displays potent anti-inflammatory activity. Several research-scale synthetic methods have been reported for the preparation of key quaternary heterocyclic intermediates of baricitinib. However, they were all associated with several drawbacks, such as the expensive materials, usage of pollutional reagents, and poor yields. Results In this manuscript, we established a green and cost-effective synthesis of 2-(1-(ethylsulfonyl)azetidin-3-ylidene)acetonitrile and tert-butyl 3-(cyanomethylene)azetidine-1-carboxylate for further scale-up production of baricitinib. This synthetic method employs commercially available and low-cost starting material benzylamine and an industry-oriented reaction of green oxidation reaction in microchannel reactor to yield important quaternary heterocyclic intermediates. Conclusion Generally, this procedure is reasonable, green and suitable for industrial production.


Background
Baricitinib, with a 2-(1-(ethylsulfonyl)azetidin-3-yl)acetonitrile moiety at the N-2 position of the pyrazol skeleton ( Fig. 1), is an oral and selective reversible inhibitor of the JAK1 and JAK2 and displays potent anti-inflammatory activity [1,2]. Besides, baricitinib has also been approved by the European Union in March 2017 and Japan in July 2017 for the treatment of moderate to severe rheumatoid arthritis for inhibiting the intracellular signaling of many inflammatory cytokines such as IL-6 and IL-23 [3][4][5] and for the patients with rheumatoid arthritis and poor response to the current standard treatment [2], respectively. For the above, the synthetic method of baricitinib has drew great attentions and been thoroughly investigated [1,2] in recent years.
However, the above synthetic methods have several defects. In Scheme 1, the yield of the first step is just only 43.4%, and the byproduct diphenylmethane in the second step is difficult to remove. Besides, in the third step, it will produce a large amount of mixed salt wastewater, which will bring great pressure to environmental protection and non-suitable for industrial production. In Schemes 2, 3, 4, the start materials are too expensive, which are also non-suitable for industrial production. Therefore, these drawbacks prompted us to consider some alternative approaches to synthesize the intermediates 2 and 3.
Herein, we presented our efforts for the development of a green and facile synthetic route with increased overall yield and suitable for industrial production, which were summarized in this manuscript.
In this green and facile synthetic route, we used benzylamine as the starting material instead of unstable reagent benzhydrylamine compared with the synthetic route in Scheme 1, as benzhydrylamine will be partly converted to dibenzophenone. Besides, the starting material benzylamine was much cheaper than benzhydrylamine, which was more suitable for industrial production. Moreover, in the second step, the by-product of deprotected toluene can be more easily removed by rectification process compared to the by-product diphenylmethane in the synthetic route in Scheme 1.
At first, traditional TEMPO reaction (sodium hypochlorite as an oxidant) in the third step was employed. Alkali with different concentrations were employed to reduce wastewater output and increase the yield. However, the by-product V-5-2 (tert-butyl 5-oxooxazolidine-3-carboxylate) was always yielded no matter how the reaction conditions were changed ( Table 1). The effects of different temperatures on the ratios of product and by-product were shown in Table 1, which suggested that − 10 °C was optimal temperature. Besides, we found that compound V-5 was converted to by-product V-5-2 by peroxidation and rearrangement reaction (Baeyer-Villiger oxidation rearrangement reaction). Peroxide H 2 O 2 was produced first as the following process (Fig. 2), which urged V-5 to by-product V-5-2 through Baeyer-Villiger oxidation rearrangement reaction.
Though lots of conditions screened, by-product V-5-2 was just controlled in 5% by traditional TEMPO reaction. To solve this problem, microchannel reactor was used with two methods instead of traditional TEMPO reaction, as it has the advantage of high heat efficiency and mass transfer property.
Method 1: TEMPO-H 2 O 2 system (Fig. 3), shortening residence time of product, inhibited the yield of by-product V-5-2, which reduced salt mixing wastewater and can be directly access to the sewage plant. In this step, the equivalents of V-5, TEMPO and H 2 O 2 was 1: 0.02: (2-10) and the best temperature was among 0-30 °C.

Conclusions
In conclusion, we provide a green and facile synthesis of an industrially important quaternary heterocyclic intermediate for baricitinib, which proceeds in six steps with multiple advantages. The most significant step of the route is the synthesis of intermediate tert-butyl 3-oxoazetidine-1-carboxylate (V-5), and there are many advantages of this method, such as inexpensive starting materials, less by-product, easily work up, and environmental protection. Moreover, the reaction reactant,  reaction time, temperature, and solvent of this step were preliminarily investigated. This environmental-friendly, cost-effective and facile process and the optimum conditions for the preparation of quaternary heterocyclic intermediates for baricitinib may form the basis of a future manufacturing route.

Experimental section
1 H NMR spectra was obtained on a Bruker AV-400 spectrometer (Bruker BioSpin, Fällanden, Switzerland) in the indicated solvent CDCl 3 . Chemical shifts were expressed in δ units (ppm), using TMS as an internal standard, and J values were reported in hertz (Hz). TLC was performed on Silica Gel GF254. Spots were visualized by irradiation with UV light (λ 254 nm). Flash column chromatography was carried out on columns packed with silica gel 60 (200-300 mesh). Solvents were of reagent grade and, if needed, were purified and dried by distillation. Starting materials, solvents, and the key reagents were purchased from commercial suppliers and were used as received without purification.

General procedure for the synthesis of tert-butyl 3-hydroxyazetidine-1-carboxylate (V-4)
To the mixture solution of 1-benzylazetidin-3-ol (V-3) (35.0 g, 214.4 mmol) in THF (350 mL) was added with 5% Pd/C (1.75 g). The reaction mixture was stirred at room temperature overnight under H 2 atmosphere for 20 h. Upon completion of the reaction, the reaction mixture was filtered by a suction filter and the filtrated was removed under vacuum and giving the desired crude compound tert-butyl 3-hydroxyazetidine-1-carboxylate (V-4). It was dissolved in n-heptane (105 mL) and stirred with 0-5 °C for 2 h under N 2 atmosphere, which was filtered again and the filter cake was dried to afford pure white solid V-4 (33.8 g, 91% yield). 1

General procedure for the synthesis of t tert-butyl 3-oxoazetidine-1-carboxylate (V-5) (traditional TEMPO reaction with oxidant NaClO)
To the solution of tert-butyl 3-hydroxyazetidine-1-carboxylate (V-4, 10.0 g, 57.7 mmol) in CH 2 Cl 2 (200 mL) 9.1% potassium bromide water solution (15.1 g) and TEMPO (0.18 g, 1.15 mmol) were slowly added under − 15 to 5 °C, which was added the mixture solution of KHCO 3 (104 g) and NaClO (86 g, 12% water solution) in water (389 mL) and stirred for half an hour. Upon completion of the reaction, the reaction mixture was quenched by 15% sodium thiosulfate aqueous solution (100 mL), extracted with ethyl acetate, washed with water, and then the solvent was removed under vacuum. The residue was dissolved in ethyl acetate again, which was added slowly 5 mL n-heptane and 0.1 g seed crystal under 10-15 °C with stirred for 20 min. And then another 5 mL n-heptane was added under − 5-0 °C and stirred for 20 min. The mixture was filtered and the filter cake was dried to afford desired compound V-5 with little by product V-5-2.

General procedure for the synthesis of t tert-butyl 3-oxoazetidine-1-carboxylate (V-5) (the microchannel reactor with TEMPO-H 2 O 2 system)
Intermediate tert-butyl 3-hydroxyazetidine-1-carboxylate (V-4, 10.0 g, 57.7 mmol), TEMPO (0.18 g, 1.15 mmol) and CH 2 Cl 2 (120 mL) were added in premixed reactor A, which was derived to the micro-channel reactor with the speed of 6.5 g/min. Meanwhile, 30% H 2 O 2 solution was pumped into the micro-channel reactor at a speed of 4.5 g/min and the stay time was 30 s. Upon completion of the reaction, the mixture solution was pumped into oil-water separator for 20 min. The organic phase was washed by water (20 mL), concentrated under vacuum to give the residue, which was dissolved in 15 mL n-heptane under 30 °C. Then 0.1 g seed crystal was added under 10-15 °C and stirred for 20 min, which was stirred for another 20 min under − 5-0 °C. The mixture was filtered and the filter cake was dried to afford desired compound V-5 (9.1 g, 92.1% yield) without by-product V-5-2. HPLC: 99.07%.

General procedure for the synthesis of t tert-butyl 3-oxoazetidine-1-carboxylate (V-5) (the microchannel reactor with composite catalyst-O 2 system)
Intermediate tert-butyl 3-hydroxyazetidine-1-carboxylate (V-4, 5.0 g, 28.8 mmol), N-hydroxyphthalimide (0.94 g, 5.76 mmol) and CH 3 CN (50 mL) were added in premixed reactor A, which was derived to the micro-channel reactor with the speed of 1 mL/min. Meanwhile, the solution of cobalt acetate (0.14 g cobalt acetate in 25 mL acetic acid) was pumped into the micro-channel reactor at a speed of 4.5 g/min and the stay time was 90 s. Upon completion of the reaction, the mixture solution was pumped into treatment reactor for 55 min. The reaction solution was concentrated and 50 mL CH 2 Cl 2 was added, which was washed by 20 mL water and 20 mL salt solution, dried to afford white crude solid. The above solid was dissolved in 2.5 mL ethyl acetate under 15 °C, which was added slowly 5 mL n-heptane and 0.1 g seed crystal with stirred for 20 min under − 5-0 °C. The mixture was filtered and the filter cake was dried to afford desired compound V-5 (4.3 g, 87% yield) without by-product V-5-2. HPLC: 99%.

General procedure for the synthesis of 2-(1-(ethylsulfonyl) azetidin-3-ylidene)acetonitrile (V-7)
To the solution of tert-butyl 3-(cyanomethylene)azetidine-1-carboxylate (V-6, 36.0 g, 185 mmol) in CH 3 CN (252 mL) hydrochloric acid (252 mL, 3 mol/L) was added and stirred under room temperature for 16 h. After completion of the reaction, the mixture solution was concentrated under vacuum and dissolved in 144 mL CH 3 CN, which was stirred for 2 h under 30 °C. And then the solution was cooled to 5 °C and stirred for another 2 h. The mixture was filtered and the filter cake was dissolved in 432 mL CH 3 CN. Diisopropylethylamine (97.1 mL) and ethanesulfonyl chloride (26.3 mL) were added under 15 °C. The reaction mixture was stirred for 12 h under 20 °C. Upon completion of the reaction, the mixture solution was concentrated under vacuum, dissolved in 360 mL CH 2 Cl 2 , extracted by 180 mL 12.5% aqueous solution of NaCl, concentrated under vacuum again to afford crud compound 2-(1-(ethylsulfonyl)azetidin-3-ylidene)acetonitrile (V-7). The crud compound V-7 was dissolved in 36 mL ethyl acetate and warmed to 50 °C. N-Heptane (48 mL) was added and cooled to 30 °C. Then 0.2 g seed crystal was added and stirred for 20 min, another n-heptane (48 mL) was added, stirred for 50 min under − 5 to 0 °C. The mixture was filtered and the filter cake was dried to afford pure compound V-7 (30.5 g,