[www.rhodium.ws] [] [Chemistry Archive]

Synthesis of Methamphetamine
by Grignard Alkylation of Acetaldehyde Methylimine

Reactions of organomagnesium compounds with aliphatic azomethines
and with hexamethylenetetramine. I.

Vera Evdokimoff
Chemical Abstracts 42, 2586b (1948)
Gazz. Chim. Ital. 77, 31826 (1947)

HTML by Rhodium

Whereas the reaction RMgX + R'CH=NR" → R'CHRN(MgX)R" -H2O-> RR'NHR" has been studied by Busch1 with arylaliphatic and aromatic azomethines, there is no literature on the same reaction with aliphatic azomethines, perhaps because the latter tend to polymerize to products without azomethinic function. This difficulty can be overcome by making equimol. wts. of the primary amine and aldehyde react in Et2O, fixing with anhyd. Na2SO4 the water formed in the condensation, and pouring the soln. immediately into excess organomagnesium reagent.

Benzyl Chloride (24.4 g) in anhydrous Et2O, added slowly to 4.8 g Mg (the reaction can best be started by inoculating with a little Mg previously heated with iodine), and heated on a steam bath, yields a solution of benzylmagnesium chloride (I). Methylamine Hydrochloride (15 g) and 30 mL 40% aq. NaOH, distilled into ice-cold anhydrous Et2O containing Na2SO4, give a solution of MeNH2 (II), the concentration of which is determined by titration. An ice-cold mixture of 3.1g II in anhydrous Et2O, 4.4g freshly prepared acetaldehyde [by depolymerization of paraldehyde], and excess anhydrous Na2SO4, agitated, allowed to stand cold several min., the Et2O soln. added slowly to the ice-cold solution of I, refluxed 2h, the Et2O distilled, the residue heated 2h on a steam bath, Et2O added, ice and dilute H2SO4 added, the Et2O layer sepatated, and the aq. layer (III) extracted with Et2O, gives a combined Et2O solution (IV). The residual acidic aq. solution (V) contains bases which give a precipitate with picric acid (VI).

A riptographic study of this reaction (cf. Toffoli et al.2) indicated that VI ppts. a single compd. rather than a mixt., that under the exptl. conditions the solubility of the picrate is approx. 7 millimolar or 2.6%, that its solubility as an undissociated compound is very low, that the quantity of base formed is 0.04 M (160 µmols. per 250 mL), and that the reaction gives a theoretical yield of approx. 40%.

The main products of the reaction were sepd. by 2 methods. V, made alk. with NaOH, extd. with Et2O, the ext. dried by Na2SO4, filtered, evapd. to dryness, the oil residue salified by HCl, kept several days in a desiccator contg. H2SO4 and NaOH, and purified by acetone, yields PhCH2CHMe-NHMe.HCl (VII), mp 140C. A mixt. (VIII) of VII and d-rotatory VII (4:1) mp 140-5C; a mixt. of VII and VIII (1:1) mp 143-8C.

VI (0.2 that required for complete pptn.), added to V, filtered, excess VI added to the filtrate, and the ppt. dissolved in hot water, cooled very slowly, concd. below 45C in vacuo, and the part of the ppt. which mp 115-20C purified by water, yields the picrate, PhCH2CHMeNHMe·Ph(NO2)2OH, mp 126C. IV, concd. and fractionally distd., yields liquids below 200C, and bibenzil (IX), mp 52C, bp 284C.

A reproduction of this synthesis in modern times has been reported by Leno Latrunculus, but his authenticity is rather questionable, as certain details absolutely essential for the reaction to even start has been left out.


  1. Busch, Chem. Ber. 37, 2691 (1904); Chem. Ber. 38, 1761 (1905); Chem. Abs. 2, 2231
  2. Toffoli, Chem. Abs. 33, 697; Toffoli and Flamini, Chem. Abs. 40, 33604