Initial research for the purification was done on the solidified oil from the last batch. Recrystallization from water was attempted but no crystals formed, so 20 mL of 5% sulfuric acid was added and swirled to dissolve as much of the oil as possible. The aqueous solution was then used to extract the xylene mother liquor from the last reaction. The separated aqueous layer was filtered through a plug of cotton and strongly basified with strong sodium hydroxide solution, but an emulsion formed, so it was acidified again with sulfuric acid and then carefully brought to pH 7 with sodium hydroxide and the resulting white emulsion was extracted with 10 mL chloroform. The hazy aqueous layer was again filtered through a plug of cotton, which removed both solid and suspended solvent, and the filtrate was strongly basified with sodium hydroxide, producing an emulsion which changed to a suspension of white solid after a few seconds. After several hours the flask contained copious large, colorless, shiny flakes. These were gravity filtered, filling the funnel about half way. The solid was rinsed with about 25 mL of water to remove sodium hydroxide, but the solid seemed to shrivel away, down to about one quarter of the funnel! The aqueous solubility of tryptamine referenced on ChemFinder.com (34 g/L) had seemed dubious, but given that the chloroform leaves the tryptamine in the neutral water layer, the solubility of tryptamine in neutral water must be considerable. The filter paper was pressed between paper towels to remove all possible water and the air dried solid weighed 0.65 grams. The solid melted at 117-118.5°C (Merck 118°C) and had one tan spot (Rf ~ 0.1 - 0.2) on silica TLC, eluting with methanol containing ~50 mg of ammonium carbonate.
Another large scale reaction was run using 75 mL of turpentine, 7.14 grams of L-tryptophan, and 15 drops (0.25 grams) of spearmint oil in a 250 mL Erlenmeyer flask with a water cooled reflux condenser attached by a rubber stopper, boiling fast enough so that there was at least a drop/second coming out of the condenser. The mixture became clear in four hours and heating was turned off after another 30 minutes. There was a little yellow solid on the side of the flask above the liquid. After sitting overnight there was a clump of yellow crystals in the corner of the flask and solidified dark oil across the bottom. The flask was refrigerated for the day and the orangish mother liquor was decanted.
To the flask were added 150 mL of 5% distilled vinegar along with 5 mL of chloroform and the flask was briskly swirled until all solid was gone and there was only a little dark brown oil not dissolved in the yellow suspension. The hazy yellow liquid (pH 5-6) was filtered through a plug of cotton and the dark brown organic layer was extracted with another 10 mL of vinegar. To the combined filtrates were added 5 mL of chloroform and enough sodium bicarbonate (10.58 g) in portions so that further addition caused very little foaming. The flask was swirled thoroughly and the hazy yellow aqueous layer was filtered through a fresh plug of cotton. The filtrate was cooled in the freezer for 15 minutes, basified with 12 mL of 25% sodium hydroxide solution, and set back in the freezer for 30 minutes. The solid was dislodged from the sides with a metal scoop and the mixture was filtered through filter paper. The flask and crystals were rinsed with 100 mL of ice cold household ammonia in portions. The filter paper was pressed between paper towels until damp and set aside to dry. The light yellow crystals weigh 3.64 grams (65% yield).
The turpentine mother liquor from the last reaction, still containing spearmint oil and some tryptamine, was used directly to decarboxylate 7.23 grams of L-tryptophan. This time the reaction took seven hours, so apparently some of the catalyst was removed during the first reaction. This time both the turpentine and the solid product were extracted with vinegar as above, and brought through the same process, to give 5.21 grams (92% yield) of light yellow crystals. The combined yield for the last two reactions is 79%.
Note that the use of a water cooled condenser in this process is unnecessary - a foot or two of wide glass tubing would suffice.