90% KOH solid 265 g (4.25 mol, flakes) was loaded into a 1-L sized three-necked flask with a large egg-shaped stirbar. Water 180mL was added in one portion and the mixture was stirred without cooling. The flask was equipped with internal thermometer, an efficient reflux condenser and a side-arm septa. When the resulting hot KOH solution temperature dropped to about100C, neat nitromethane 90 g (80 mL, 1.475 mol) was gradually added by syringe through the side arm, with a vigorous stirring. A strongly exothermic reaction commenced, accompanied with ammonia gas evolution, foaming and precipitate formation. The temperature and gas evolution was controlled by the nitromethane addition rate so that the internal temperature remained between 120 and 140C. This addition took about 30 min. [Note 1] The mixture was stirred then for additional 20 min, the flask was then placed on a 130 C oil bath and stirred at this temperature for 1 hour with a gentle stream of Ar introduced into the flask through the side arm (the septa was replaced with a gas inlet adapter [Note 2]). The resulting light-colored slurry was cooled and stirred at RT for about 30 min. The flask was then placed into a refrigerator overnight. The precipitated bis-potassium salt of nitroacetic acid was collected by filtration on a large (350mL, medium porosity) glass Buchner funnel, the wet cake was compressed on the frit. The reaction flask and the filter cake were thoroughly rinsed with fridge-chilled methanol , 6x20mL. The solid was then dried by suction. [Note 3]
The obtained bis-potassium salt (84.4g of a cream-colored crystalline solid, 63% th) was gradually added as a solid into a freezer pre-cooled (-20C) stirred mixture of 85% H3PO4 400g with water 300mL and ethyl acetate 500 mL in a 2L Erlenmeyer flask. The mixture was then placed on a ice/water slush bath and stirred at 0C until all solids have dissolved (about 20 min). The phases were separated, the organic layer was washed with chilled sat NaCl 500 mL , the aqueous phases were re-extracted with additional chilled EtOAc 500 mL. The combined organic extracts were dried with MgSO4, filtered and evaporated from a 17-22C bath, to near dryness. [Note 4] The obtained crystalline residue was covered with ethyl acetate 10 mL, the slurry was diluted with chloroform 250mL and allowed to crystallise for 20 min. The precipitate was collected by filtration, washed with chloroform, dried by suction and then on highvac (30 min). Concentrating the supernatant and covering the residue with 10mL of 1:1 ethyl acetate + chloroform mixture and then diluting the slurry with chloroform 80mL provided a second crop of the pure product.
The combined yield was 44.85 g of a sugar-like white crystalline solid (Y=58% overall).
1H(CD3CN, 400MHz): 9.578 (very br s, 1H), 5.295 (s, 2H); 13C(CDCl3, 100MHz): 164.34, 77.51
Note 1: The nitromethane adition is promptly exothermic but the ammonia evolution and precipitate formation is slightly delayed. This can cause a foam-over accident if the nitromethane is added too rapidly below the optimal temperature. Above 120C the mixture is easier to stir and the foaming is suppressed but an efficient condenser is needed to keep nitromethane from being carried away by the evolving NH3 gas.
Basic impurities sensitise nitromethane - the used nitromethane should be transferred into a temporary storage flask and drawn from there so that the main bottle is not accidentally contaminated with alkali. If the addition funnel is employed, the funnel should be kept under positive pressure of Ar to keep NH3 from entering the funnel.
Note 2: Argon is passed through the flask sidearm to drive out the remaining ammonia. This is not essential but it seems to produce a lighter-colored reaction mix. Due to some evaporation the internal temperature of the mix holds at around 120C when the mix is stirred on a 130C bath.
Note 3: There are contradicting reports about the shock sensitivity of this bis-potassium salt. To stay on the safe side I tried to avoid grinding and pounding the material too vigorously. After the bis-K-salt washing with methanol, it became non-hygroscopic and the compressed cake could be fluffed up with a spatula on the Buchner funnel, to aid the drying by suction and to keep the cake from turning into a brick. Some residual MeOH is no problem in the next neutralisation step.
Note 4: The mono salt is extremely unstable. The acid is reasonably stable as a solid but it decarboxylates slowly at RT in solution, especially in the presence of water. The solutions were kept below +10C during the extraction workup. (The drying with MgSO4 and the evaporation can be done at ambient temperature, without a delay). The used chloroform was an ACS grade stabilised with amylene, free of ethanol and acidic impurities. Nitroacetic acid is stored in a fridge.