Very few ethers of phenyltropanes are actually known to have been reported. NS2330 and NS2359 are exemplary, although it must be stressed that these compounds have α,β stereochemistry. Since further development of NS2214 appears to have been halted, the importance of a metabolically stable compound that is not just readily metabolized is evident.
For some reason that is not entirely clear, the authors of the present document decided that they wanted to make all 8 stereoisomers for the phenyltropane homolog of paroxetine.
To solve the problem of the unexpected aza-bicyclo[3.2.2]nonane rearrangement product, the original synthesis had to be modified as follows;WIN 35428 was N-demethylated and then the NH amine was reacted with a suitable protecting group so that is no longer nucleophilic. In their case they used a tosyl. (Satendra Singh, 2000) Page 952
^Ogier, L.; Turpin, F.; Baldwin, R. M.; Riché, F.; Law, H.; Innis, R. B.; Tamagnan, G. (2002). "Rearrangement of a mesylate tropane intermediate in nucleophilic substitution reactions. Synthesis of aza-bicyclo3.2.1octane and aza-bicyclo3.2.2nonane ethers, imides, and amines". The Journal of Organic Chemistry67 (11): 3637–3642. doi:10.1021/jo010973x. PMID12027674.edit
^Runyon, S.; Burgess, J.; Abraham, P.; Keverline-Frantz, K.; Flippen-Anderson, J.; Deschamps, J.; Lewin, A.; Navarro, H.; Boja, J.; Kuhar, M. J.; Carroll, F. I. (2005). "Synthesis, structural identification, and ligand binding of tropane ring analogs of paroxetine and an unexpected aza-bicyclo3.2.2nonane rearrangement product". Bioorganic & Medicinal Chemistry13 (7): 2439–2449. doi:10.1016/j.bmc.2005.01.046. PMID15755646.edit