N-amination of indoles on pilot-plant scale via simultaneous and proportional metering of reagents

Organic Process Research and Development

Volumn 15, Issue 3, 2011, Pages 704-709

  Weiberth, Franz J ^a   Hanna, Reda G ^a   Lee, George E ^a   Polverine, Yvonne ^a   Klein, Joseph T ^a  

^a SANOFI   (France)

Author keywords

[No Author keywords available]

Indexed keywords

CORIOLIS; MASS FLOW METERING; PILOT PLANT SCALE; PUMPING SYSTEMS;

AMINES;

EID: 79957462670     PISSN: 10836160     EISSN: 1520586X     Source Type: Journal    
DOI: 10.1021/op100330h     Document Type: Article

References (27)

1. Somei, M.; Natsume, M. Tetrahedron Lett. 1974, 461
2. Somei, M.; Matsubara, M.; Kanda, Y.; Natsume, M. Chem. Pharm. Bull. 1978, 26 (8) 2522
3. Weiberth, F. J.; Lee, G. E.; Hanna, R. G.; Mueller-Lehar, J.; Dubberke, S.; Utz, R. U.S. Patent 7,112,682, September 26, 2006, (CAN 142:411224).
4. Indoles are not sufficiently nucleophilic to react directly with HOSA.
5. No desired product would be expected if solutions of the incompatible reagents HOSA and KO t Bu were mixed prior to the addition of indole.
6. Hynes, J., Jr.; Doubleday, W. W.; Dyckman, A. J.; Godfrey, J. D., Jr.; Grosso, J. A.; Kiau, S.; Leftheris, K. J. Org. Chem. 2004, 69, 1368
7. a of the N-H proton in indole is reported to be ∼17. See: Remers, W. A. Properties and Reactions of Indoles, Isoindoles, and Their Hydrogenated Derivatives. In Indoles, Part I; Houlihan, W. J., Ed.; The Chemistry of Heterocyclic Compounds; Wiley & Sons, Inc.: New York, 1972; p 14.
8. a of t -BuOH is commonly reported to be within the range of 17-19, depending on the specific literature source.
9. Conversions to 3a were typically 60-68 A% when only 1.0 equiv of HOSA was dispensed simultaneously with 2.0 equiv of KO t Bu.
10. Dipolar aprotic solvents were preferred because they solubilized both HOSA and KO t Bu. NMP was especially preferred because it was also suitable for downstream chemistry (telescoping).
11. The N -amination was a very fast reaction (e.g., 1 min stirring was sufficient to provide representative conversion assays after feeds were momentarily interrupted for in-process sampling). Consequently, this latter charging regimen has shown potential to be adapted into a flow reaction, although the advantage may be limited because reagent solutions will need to be prepared freshly in batch mode just prior to processing (see ref 16), and an amination vessel would still be needed to collect the product stream.
12. As determined by HPLC assay. The balance of material was mostly unreacted indole.
13. In this discussion, XP refers to Class 1, Division 1, Group C/D processing environments.
14. Major components: Bran & Luebbe duplex metering plunger pump model N-P32 with pulsation dampener SG-TW-25 and back-pressure valve 14010-2300; Micro Motion mass flow sensor D12H205SU (1/8-in tubes); Micro Motion transmitter model RFT9739E1SU; Micro Motion flow monitoring device model FMS1NA0EA.
15. All materials should be handled while minimizing exposure and while wearing proper protective equipment, most importantly when handling HOSA and KO t Bu, both of which can cause burns on skin and in the eyes and are extremely destructive to mucous membranes and the upper respiratory tract.
16. All materials, including NMP, were accurately preweighed.
17. HOSA was charged in 3 portions over about 15 min to minimize formation of clumps of the solid that were then difficult to dissolve.
18. Alternatively, a separately weighed charge of KO t Bu could be added directly into the amination reactor.
19. This failed batch was salvaged by isolating the product (extractive workup), then reprocessing the resulting mixture of 3c:1c by treatment with additional HOSA/NMP and KO t Bu/NMP in a simultaneous manner to give a 98 A% conversion to 3c.
20. In later campaigns, it was found that a precharge of 0.07-0.10 equiv of KO t Bu resulted in more robust performance.
21. The sparging tubes were fabricated from 3/8-in SS tubing by drilling a series of twenty 1/16-in. diameter holes in the submersed and plugged end of the tubing. The tubes were secured within the reactor through a 1-in dip pipe or by clamping to a baffle.
22. The amination was a fast reaction with no apparent accumulation of reagents (see ref 9). On a 133.5-mol scale (150-gal reactor) and while feeding the reagents over a 187-min duration, a reaction temperature of 25-30 °C was readily maintained by circulating 10 °C coolant through the jacket.
23. The efficiency of the amination decreased as the reaction neared completion. Additional charges of HOSA and KO t Bu only incrementally improved the overall conversion.
24. In some batches, up to 1% of the HOSA charge remained in the feed vessel (undissolved chunks). An in-line, polypropylene Y-strainer prevented any loose chunks of solid from reaching the pump.
25. Alternatively, indole can be dissolved in NMP in the amination vessel prior to the simultaneous feeds of the HOSA/NMP and KO t Bu/NMP solutions.
26. KO t Bu and its solutions should be protected from moisture.
27. The reagents were sometimes prepared in excess as a precaution, e.g., in case they were needed to further drive the amination towards completion.