2-Bromo-5-fluoro-3-nitropyridine

Scheme 80 is a preferred method for making compounds of Formula I and Ia where R2 is fluoro. This is exemplified specifically in the preparation of compound Example 216. The synthesis of 2-hydroxy-3-nitro-5-fluoropyridine 5-80 as shown was carried out generally via the methods of A. Marfat and R. P. Robinson U.S. Pat. No. 5,811,432 (column 25, example 5) and Nesnow and Heidleberger (J. Heterocyclic Chem. 1973, 10, pg 779) except that a number of procedural enhancements were incorporated as noted in the description of each step. 2-Hydroxy 5-fluoropyridine 4-80 is also commercially available. The formation of diazonium tetrafluoroborate salt 2-80 from 5-amino-2-methoxy pyridine 1-80 proceeded in essentially quantitative yield and was isolated via filtration. The Schiemann reaction provided poor yields of the desired 2-methoxy-fluoropyridine using the literature conditions due mainly to significant contamination with 3-fluoro 1-(N)-methyl pyridone and other byproducts. However, adoption of a procedure similar to that described in Sanchez, J. P.; Rogowski, J. W.; J Heterocycl Chem 1987, 24, 215 for a related compound provided very high yields of essentially clean but volatile 2-methoxy-5-fluoro pyridine 3-80 as a solution in toluene. In the interest of expediency, demethylation was achieved on large scale using aqueous HCl in pressure bottles at 140° C. for 1 hr. Prior to heating, the toluene solution was stirred with the aq HCl and then the toluene was removed by decantation. The literature method for carrying out this step using HBr at 100° C. was also successful on small scale and had the advantage of avoiding the use of pressure bottles. Nitration of 4-80 as described by Marfat provided lower than expected yields so the procedure was modified slightly, using guidance from A. G. Burton, P. J. Hallis, and A. R. Katritzky (Tetrahedron Letters 1971, 24, 2211-2212) on the control of the regiochemistry of nitration of pyridones via modulation of the acidity of the medium. The chemical yields of 2-hydroxy-3-nitro-5-fluoro pyridine 5-80 were significantly improved using the procedure described in the experimental section. Occasionally the product failed to precipitate during workup and then considerable efforts were necessary to isolate this highly water soluble compound from the aqueous layer. Using neat excess POBr3, compound 5-80 was converted to 2-bromo-3-nitro-5-fluoro pyridine 6 which could be used without further purification in the subsequent azaindole forming reaction. Addition of the pyridine 6 to excess vinyl magnesium bromide in THF at low temperature afforded the desired 4-fluoro-7-bromo-6-azaindol (precursor 5j) in yields of up to 35% following acidic work up and isolation via crystallization. A disadvantage of this method is the workup is difficult due to the large amounts of salts formed as co-products in the reaction and the low conversion to albeit clean product. The reaction is also exothermic and thus would require care on larger scales. Despite the moderate yields, as mentioned above the reaction proceeds cleanly and provides pure product precursor 5j without chromatography so it is anticipated that more detailed studies of this chemistry could result in yield enhancements. A selective copper/potassium carbonate mediated displacement of the 7-bromo group by commercially available 1,2,3-triazole provided an approximately 1:1 mixture of triazoles from which the desired 7-80 was isolated via chromatography in 25-35% yields. Copper-bronze rather than copper powder can also be used to carry out similar transformations. This reaction must not be allowed to overheat since concomitant displacement of the fluorine is possible and has been observed. Acylation occurred most efficiently under conditions that utilized excess acidic imidazolium chloro aluminate ionic liquid to provide highly activated glyoxylating reagent (K. S. Yeung et al. Tetrahedron Lett. 2002, 43, 5793). The acylation of 7-80 usually does not proceed to completion and typically results in about 75% conversion as measured by LC/MS. An advantage to these conditions is that the typical next step, ester hydrolysis, proceeded in situ to provide the desired acid 8-80 which was isolated directly by precipitation during workup. Coupling of the piperazine benzamide was found to be cleaner and produced higher yields of the compound of Example 216 using the depicted HATU based coupling than with other standard coupling reagents such as EDC or DEPBT.