This method was taken from
Patent US4242274
A 6 l flask containing 1375 ml toluene was charged with 50% solution of NaOH (from 440 g NaOH and 440 ml of water), 550 g diphenylacetonitrile, 550 g 1-(dimethylamino)-2-chloropropane hydrochloride, 440 ml water and 48 g TBAB. The mix was heated to reflux with powerful mechanical stirring (500-800 rpm) and was refluxed 2 hrs. The mix becomes homogeneous only then refluxed.
The mix was cooled and to it 2200 ml of water was added. The bottom strong alkaline aqueous layer was separated and discarded. To the organic layer was added 1100 ml of water and 350 ml concentrated HCl (pH of water phase should be 2-3). The phases were separated and to the organic layer additional water (550 ml) was added and 110 ml conc. HCl. The aqueous phase was separated and combined with aqueous layer from the first acid extraction. The organic layer was discarded. To the combined aqueous phase a 50% hot solution of NaOH was added until pH was increased about 13. The mixture should be hot (50-60°C) to promote separation which required about 30 min. The yellow oily layer was separated, dissolved in 800 ml IPA, heated to boiling, insoluble residue was filtered off and solution was cooled to 10°C and maintained at that temp about 1 hr so as to crystallyze the desired nitrile. The product was filtered and dried, affording 350 g (44%) white crystalls. Product is very pure.
The use of dibenzo-18-crown-6 as phase-transpher catalyst (
Post 481630 (missing)
(Aurelius: "PTC Synthesis of Methadone", Methods Discourse))gives an advantage in ratio of desired and undesired nitrile 72:28, but it is very expensive reagent at €69.50/50g (Aldrich).
