Patent Watch

August 20, 2012

Here are two ways to make an isoquinoline intermediate. S. A. Springfield and co-inventors describe a process for preparing 1,8-dichloro-4-methoxyisoquinoline (8) and other 1-chloro-4-methoxyisoquinolines. Product 8 is an intermediate in the synthesis of 1, a compound that is active against the hepatitis C virus.

The inventors report two methods for making 8, but only one is covered by the patent claims. This route, outlined in Figure 1, starts by treating acid 2 with (COCl)2 and a catalytic amount of DMF, followed by adding NH4OH, to produce amide 3. The amide is recovered as a solution in 2-methyltetrahydrofuran (MeTHF); after the solvent is switched to toluene, it reacts with Me2NC(OMe)2 to give amidine 4. The amidine is not isolated, but it is treated with strong base to produce isoquinolone 5.

Compound 5 is recovered in 89% yield and 98% purity; it is then oxidized with a hypervalent iodine compound. The example uses PhI(OAc)2 in the presence of methanesulfonic acid (MsOH), although other reagents are mentioned and covered in the claims. The reaction initially gives dimethoxy compound 6, which is not isolated, but is heated to drive off MeOH and form isoquinolone 7.

Product 7 is isolated in yields of 75–85% and >97% purity. In the last step, 7 is chlorinated with POCl3 to produce target compound 8, recovered in 77–87% yield of with >98% purity.

The alternative route to 8 via 7 is shown in Figure 2. The patent’s examples contain very detailed steps, but only the main reagents are shown in the figure. The procedure starts with the reaction of benzoic acid ester 9 and vinyl ether 10 in the presence of a palladium–phosphine catalyst and EtN-i-Pr2 to produce ester 11. In the same pot, 11 is hydrolyzed with acid to give ketone 12, which is recovered in toluene solution in 87.3% yield. The inventors stress that the concentration of this solution must be between 95 and 108 mg/mL. If it is any higher, ketone 12 decomposes to give a cyclized acetophenone that can amount to as much as ≈12% of the product.

The solution of 12 is treated with bromine to produce α-bromo ketone 13, isolated in 60% yield as an orange solid. Product 13 undergoes a biphasic reaction with NaN(CHO)2 to produce dialdehyde 14, which is converted to isoquinolone 15 in the same pot by treating it with MeOH and then HOAc. Compound 15 is recovered as a brown solid in 78% yield; it is then methylated with MeOH–MsOH to obtain 7. The reaction takes ≈48 h; product 7 is isolated in 90% yield after workup.

Compound 7 is chlorinated to 8 with POCl3 as in the first synthesis, but the subsequent workup uses K2HPO4 rather than K3PO4. The patent includes numerous notes about the workup that describe how the various stages should be performed. Although the synthesis of 8 by this route is not covered in the claims, its examples are all carried out on a kilogram scale, whereas those for the first route are on a 10–20-g scale. The detailed description suggests that the second route may have greater commercial significance than the first.

The examples are unusual because they list the precise quantities of reagents and solvents used and give more experimental details than one would expect to find even in journal articles. The processes provide efficient ways to prepare the desired compound without isolating and purifying many intermediates. (Bristol-Myers-Squibb [Princeton, NJ]. US Patent 8,207,341, June 26, 2012; Keith Turner)

What do you think of Patent Watch? Let us know.