I've done a reaction twice now using the same conditions, and observed the desired product mass by UPLC-MS on both occasions, but by the end of the workup procedure (or column, in the second case) I only observed a single peak whose mass corresponded to the dehalogenated byproduct. Did SciFinder/Reaxys searches for similar reactions but couldn't come up with anything, so turning to you good people. It's pretty clear I'll need to explore other conditions to obtain the desired product, but I'm really baffled as to why I'm seeing dehalogenation.
The general procedure I'm following from a patent is dissolving the oxime in methanol/water (2:1), adding Zinc dust (2 eq.) and cooling to 0C and adding formic acid (same amount as water by volume) portionwise, continuing to stir at 0C for a few hours then allowing to warm to room temp. I generally need to add 1-2 additional eq. of Zinc to no longer observe starting material. Then dilute with EtOAc, extract a few times with water, and concentrate acidic aq. layers in vacuo to recover the product and use during the next step without further purification.
In the patent for the next step their yield is 30-40% for the bromo compound, while the same reaction using the des-bromo pyridine results in 80-90% yield, so I'm convinced that in their hands they experience the same dehalogenation, however, I'm baffled about the mechanism. Any pointers or empirical experience would be greatly appreciated.