The preparation of homogentisic acid and of 2:5-dihydroxyphenylethylamine.

This acid was synthesized soon after its discovery in alcaptonuric urine by Baumann & Fraenkel (1894) from 2:5-dimethoxybenzaldehyde by successive conversion to the alcohol, chloride and nitrile, followed by hydrolysis. The overall yield was very poor, due mainly to side reactions occurring in the conversion of the chloride to the nitrile. Schoepf & Winterhalder (1940), who studied systematically the formation of substituted phenylacetic acids by various methods, have found that a large excess of cyanide has to be used in such reactions if a good yield of the nitrile is to be obtained. We hoped that the chloride might be conveniently prepared by chloromethylation of 1:4-dimethoxybenzene; however, even under the most varied conditions, only the dichloromethyl compound could be obtained. Since, as will be shown later, homogentisic acid can be prepared from 2:5-dimethoxybenzaldehyde in good overall yield by another method, we did not explore the method ofBaumann & Fraenkel (1894) any further. Various other methods for the synthesis of homogentisic acid have been described, but the yields reported were in most cases very poor. Hahn & Stenner (1929) prepared 2:5-dibenzoyloxy-1-allylbenzene from hydroquinone, and claimed to have oxidized the allyl compound directly with very dilute ozone to benzoylhomogentisic acid in good yield. In our hands the method was not found to be very satisfactory. The benzoyl groups are very labile, and the material had, as Hahn & Stenner themselves found, to be rebenzoylated during the synthesis. Owing to the low concentration of ozone to be employed, the oxidation has to be carried out over many hours or even days, if moderately large amounts of material are used (cf. Schoepf & Winterhalder, 1940). Moreover, apart from homogentisic acid, other substances are produced in the oxidation. Hill & Short (1937) obtained good yields in the oxidation of 3-allyl-o-tolyJ methyl ether with aqueous KMnO4 containing the required amount of acetic acid. We have applied their method to the oxidation 1:4-dimethoxy-2-allylbenzene. This compound is readily obtained by Claisen rearrangement from 1-methoxy-4allyloxybenzene (Mauthner, 1921) followed by methylation. However, oxidation of this compound with KMnO4 gave, under the many conditions which were tried, a mixture of substances, and the yield of the desired 2:5dimethoxyphenylacetic acid was generally poor. Apart from 2:5-dimethoxyphenylbenzoic acid, a large amount of a neutral compound was obtained which was identified as 3-(2':5'-dimethoxyphenyl)-propane-1:2-diol. This glycol reacted with periodic acid to give formaldehyde and 2:5dimethoxyphenylacetaldehyde which was not isolated, but oxidized directly to the corresponding acid. The diol was also prepared, albeit in poor yield, by oxidizing the allyl compound with performic acid (Swern, Billen & Scanlan, 1946; English & Gregory, 1947). Oxidation of o-hydroxyphenylacetic acid with persulphate gave homogentisic acid, but again the yield was poor. Homogentisic acid was finally prepared in good yield by oxidation of 2:5-dimethoxyphenylpyruvic acid with H202 followed by hydrolysis. 2:5-Dimethoxybenzaldehyde was condensed with hippuric acid (Gulland & Virden, 1928; Neuberger, 1948) and the azlactone ring opened with alkali. Gulland & Virden (1928) then hydrolyzed the benzamidoacrylic acid and separated the keto and benzoic acids with SO2. It was found more convenient to oxidize the mixture of the two acids directly, and to separate the benzoic and 2:5-dimethoxyphenylacetic acids by esterification and fractional distillation (see Snyder, Buck & Ide, 1943). Short hydrolysis with HBr yielded homogentisic acid in good overall yield. Two other methods have been published which might be equally suitable for the preparation of this acid. Schwenk & Bloch (1942) have prepared 2:5-dimethoxyphenylacetic acid by a modification of the Willgerodt reaction from the corresponding acetophenone and recently McElvain & Engelhardt (1944) have obtained the lactone of homogentisic acid byhydrolysis of 5-hydroxy2-ethoxycoumarone which had been made by condensation of p-quinone with ketene acetal.