A 2L round flask equipped with a large egg-shaped stirbar was flushed with a stream of Ar. Solid LiBr 0.96g (11mmol) was added, followed by anhydrous THF 0.5L and diisopropyl amine 17mL (120 mmol). The obtained solution was cooled to 0C, BuLi 2.5M solution 44mL (110 mmol) was added by syringe over 10min, the mixture was stirred for additional 5 min and the ice bath was then replaced with a dry ice/acetone bath. Meanwhile, 2-bromo-5-cyanopyridine 19.838g (108.4 mmol) was dissolved under Ar in anh THF 0.25L and the solution was cooled to 0C. Using a wide canula, the ice-cooled solution of the cyanopyridine was transferred into the LDA solution at -78C, along the flask wall and with vigorous stirring, over a 10 min period (the flask and the cannula was rinsed with additional anh THF, 2x15mL). After complete addition the reaction mixture was gently swirled by hand – to wash down any crystals of the cyanopyridine on the flask wall above the reaction mixture. The mixture was then stirred for additional 15 min at -78C. Iodine 30.5g (120 mmol) solution in anh. THF 220mL pre-cooled on ice bath was added rapidly (4 min, along the flask wall), the reaction mixture was stirred at -78C for additional 30 min, the cooling bath was then removed and the flask was allowed to warm up close to the ambient temperature (over approx 1 hour interval). Without any quench, the reaction mixture was concentrated on rotovap from a 25C bath. A solution of sodium metabisulfite 35g in water 250mL was carefully combined with saturated bicarbonate solution 550mL (CO2 effervescence!) and this combined solution was then added to the oily evaporation residue. The resulting slurry was stirred vigorously for 30 min, the precipitated product was then collected by filtration, washed with copious quantity of water and dried by suction. The obtained crude product was suspended in methanol 1L and the mixture was refluxed for 30 min. Activated charcoal 4g was added (carefully) and the mix was refluxed for additional 10 min, then filtered while warm. The charcoal was washed with some additional refluxing methanol, 100mL. The combined filtrates were placed in a fridge at 0C overnight. The crystallised product (tan prisms, 19.434g, 99% pure by HPLC) was collected by filtration, washed with chilled methanol (100mL) and dried by suction and on highvac. The supernatants were evaporated and the residue was re-crystallised from methanol 0.3L (re-dissolved at reflux, then placed in a freezer at -20C overnight), to yield the second, dark brown-colored crop 5.920g (95% pure by HPLC). The combined yield was 25.354g (75.5% th).
1H(d6-DMSO, 400MHz): 8.752(d, 0.3Hz, 1H); 8.495(d, 0.3Hz, 1H)
Note: One should use 1:1 LDA/substrate ratio and keep the lithiation time short. Exceeding the LDA stoechiometry encourages formation of non-polar impurities that are rather difficult to remove. Also, the starting cyanopyridine has tendency to fall out of the THF solution at low temperature – so it is important to wash down any crystals stuck on the flask walls above the reaction mix.
Credit: I am grateful to my colleague Par who found that this substrate lithiates with LDA into the 4-position.