In a round flask with a large egg-shaped stirbar, a solution of diisopropyl amine 3.1 g (4.2 mL; 30 mmol) in anh. THF (approx 140mL) under Ar was cooled to 0C, 2.5M BuLi solution in hexane 11 mL (27.5 mmol) was added over 5 min and the mixture was stirred on ice for additional 5 min and then cooled to -78C. To this LDA solution, an ice-cooled solution of 2,5-dibromopyridine 5.925 g (25 mmol) in anh THF (approx 50 mL) was slowly added via thin canula (gauge 18) along the flask wall over 20 min with vigorous stirring (at near-maximum speed). The flask and the canula were washed with additional anh THF (2x5mL). [Note 1] After the complete addition the mixture was stirred at -78C for extra 20 min. Anhydrous DMF 2.4 mL was then added drop-by drop over 6 min period, the mixture was then stirred for additional 30 min at -78C. The reaction was quenched by dropwise water addition (10mL), the cooling bath was replaced with ambient water bath and when the mixture warmed above 0 C additional water 90mL was added. The mixture was stirred at RT for 30 min during which time the color turned from purplish to yellow-brown. The reaction mixture was combined with hexane 100mL and sat NaCl 100mL in a separatory funnel and separated. The organic phase was washed twice with sat. NaHCO3, 2x150mL. The aqueous phases were re-extracted with ether 100mL. The combined organic extracts were dried (MgSO4) and evaporated. The residue was re-evaporated from benzene 60mL and dried on highvac. The obtained semi-solid residue was dissolved in refluxing cyclohexane 60mL, the solution was decanted and the remaining dark gummy insoluble residue was re-extracted with some additional cyclohexane (2x10mL) at reflux. The combined cyclohexane solutions were allowed to cool down to RT, after 1 hour the cloudy solution was filtered through a medium-porosity Buchner funnel (to remove a small amount of insoluble sticky impurities; the filtration funnel was washed with additional cyclohexane) and the filtrates were evaporated.
The obtained solid residue was re-crystallized from cyclohexane 25mL (reflux to RT, overnight). The supernatants were decanted and the crystallized product was suspended in a mixture cyclohexane-hexane 1:1 (approx 10mL), filtered, washed with some additional 1:1 cyhex-hex mixture and dried on highvac to yield 3.398g of pure product. The supernatants were combined with the washings and placed in refrigerator (+2C) overnight. The supernatants were decanted from the crystals, the obtained second fraction was suspended in 1:1 cyhex-hexane mixture, filtered, covered with a small volume (4 mL) of cyclohexane on a Buchner funnel and crushed and stirred with spatula (to remove some sticky oil adhering to the crystals), then filtered and washed again with some 1:1 cyhex-hex mixture and then dried - to provide additional 633mg of pure product.
Combined Y = 4.031g (61% th) of a light orange-tan crystalline solid
1H(d6-DMSO, 400MHz): 10.080(s, 1H), 8.814(s, 1H), 7.899(s, 1H); 13C(d6-DMSO, 100MHz): 189.70, 154.09, 141.43, 140.71, 127.43, 120.68
Note 1: 5-halo-4-pyridyl lithium species are very sensitive, they decompose readily above -78C and they also tend to equilibrate to isomeric 3-pyridyl lithiums. It is important to avoid overheating when adding the pyridine to the LDA solution. Efficient stirring aids the heat transfer. The halopyridine substrate solution should be added slowly, pre-cooled. As 2,5-dibromopyridine is poorly soluble in THF at low temperatures, it was pre-cooled only to 0C. (Some dibromopyridine precipitation occurs on the flask wall during the canula addition but the material is washed down into the reaction mixture during the canula wash with additional THF.)
Credit: Lithiation of 2,5-dibromopyridine in the 4 position is not described in the literature but the procedure is based on Schlosser work. I am grateful to my colleague, Par, whose advice and pyridines I took.