(46c) Process Developmet for Fluorination of a Steroid with Dast and Deoxofluor

Selective presence of fluorine in some organic molecules is known to alter their physiological properties¹. In the past few decades there has been increasing interest in the development of pharmaceutical products that contain fluorine in their structure². The conversion of C-O to C-F bond represents a viable method to produce selectively fluorinated organic compounds. These fluorination reactions, also sometimes termed as deoxofluorination, need fluorinating reagents that are often very hazardous³. Microreaction technology holds a big potential to benefit this field, mainly by offering improved safety features realized due to smaller reaction volumes and better heat transfer in microreactors compared to conventional reactors. Furthermore with some fluorinating reagents, scale-up in batch systems is constrained by the recommend limit for the maximum reactor volume. In such cases continuous processing also offers the advantage of higher throughput with a single reactor with much smaller hold-up than the batch reactor.

Figure 1. Deoxofluorination of a steroid.    

In this work, we have utilized these advantages in the development of a process for the deoxofluorination of a C=O bond in a steroid (Figure 1). After testing some commercially available fluorinating agents, Diethylaminosulfurtrifluoride (DAST)⁴ and Bis(2-methoxyethyl)aminosulfurtrifluoride (Deoxofluor)³ were found suitable for this reaction. These fluorinating agents are thermally unstable and have high thermal potential. DAST is known to decompose violently at temperatures above 90°C and it is not recommended for production on an industrial scale³. Preliminary reaction studies showed similar results with both fluorinating reagents. Calorimetry was performed with pure reagents and reaction mixtures to identify the boundaries for safe operating conditions. Formal kinetics of the reaction was investigated in a 50 mL batch reactor with Deoxofluor. It was observed that higher temperatures favour the decomposition of the fluorinating reagent more than the deoxofluorination reaction. Since the reaction and decomposition both are exothermic, a precise temperature control is needed to maintain optimum reaction conditions. A simulation based experimental design approach was implemented to find the optimum conditions of temperature, reagent amount and residence time for the flow reaction. Figure 2 shows the process flowsheet for this reaction. The reactants were pumped through a micromixer into a residence time module. Depending on the operating temperature and other specific requirements, a metal or fluoropolymer tube with submillimeter to a few millimetres diameter and a specific length was used as residence time module for the reaction. Inline monitoring of the product stream was performed by IR spectroscopy using a specially fabricated miniaturized ATR probe. Figure 3 shows some results of batch and flow reactions. The throughput of the laboratory process with DAST was increased fifteen times by applying a combined scale-up and numbering-up concept. Kilograms of the fluorinated steroid intermediate were produced by this process for further processing for clinical tests.

Figure 3. Steroid conversion in batch and flow reactor with 2 molar equivalents of Deoxofluor in the feed. Toluene and Tetralin were used as solvent.

References

1.       A. Strunecka, J. Patocka, P. Connet. Fluorine in Medicine. J. App. Biomed, 2004, 2, 141-150

2.       H.J. Böhm, D. Banner, S. Bendels, M. Kansy, B. Kuhn, K. Müller, U. Obst-Sander, M. Stahl. Fluorine in Medicinal Chemistry. ChemBioChem, 2004, 5, 637-643

3.       U.S. Patent 6,242,645. 2001

4.        U.S. Patent 3,976,691. 1976