The oxyazo compounds are prepared by adding a solution of a diazonium salt to a cold slightly alkaline solution of a phenol. The diazo group takes up the para position with regard to the hydroxyl group, and if this be prevented it then goes into the ortho position. It never goes directly into the meta position.
The constitution of the oxyazo compounds has attracted much attention, some chemists holding that they are true azophenols of the type R·N·R·OH, while others look upon them as having a quinonoid structure, i.e. as being quinone hydrazones, type R·NH·N:R:O. The first to attack the purely chemical side were Th. Zincke (Ber., 1883,16, p. 2929; 1884, 17, p. 3026; 1887, 20, p. 3171) and R. Meldola (Jour. Chem. Soc., 1889, 55, pp. 114, 603). Th. Zincke found that the products obtained by coupling a diazonium salt with α-naphthol, and by condensing phenyl-hydrazine with α-naphthoquinone, were identical; whilst Meldola acetylated the azophenols, and split the acetyl products by reduction in acid solution, but obtained no satisfactory results. K. Auwers (Zeit. f. phys. Chem., 1896, 21, p. 355; Ber., 1900, 33, p. 1302) examined the question from the physico-chemical standpoint by determining the freezing-point depressions, the result being that the para-oxyazo compounds give abnormal depressions and the ortho-oxyazo compounds give normal depressions; Auwers then concluded that the para compounds are phenolic and the ortho compounds are quinone hydrazones or act as such.
A. Hantzsch (Ber., 1899, 32, pp. 590, 3089) considers that the oxyazo compounds are to be classed as pseudo-acids, possessing in the free condition the configuration of quinone hydrazones, their salts, however, being of the normal phenolic type. J. T. Hewitt (Jour. Chem. Soc., 1900, 77, pp. 99 et seq.) nitrated para-oxyazobenzene with dilute nitric acid and found that it gave a benzene azo-ortho-nitrophenol, whereas quinones are not attacked by dilute nitric acid. Hewitt has also attacked the problem by brominating the oxyazobenzenes, and has shown that when the hydrobromic acid produced in the reaction is allowed to remain in the system, a brombenzene-azo-phenol is formed, whilst if it be removed (by the addition of sodium acetate) bromination takes place in the phenolic nucleus; consequently the presence of the mineral acid gives the azo compound a pseudo-quinonoid character, which it does not possess if the mineral acid be removed from the sphere of the reaction.
Para-oxyazobenzene (benzene-azo-phenol), CHN:N(1)·CH·OH(4), is prepared by coupling diazotized aniline with phenol in alkaline solution. It is an orange-red crystalline compound which melts at 154° C. Ortho-oxyazobenzene, CHN:N(1)CH·OH(2), was obtained in small quantity by E. Bamberger (Ber., 1900, 33, p. 3189) simultaneously with the para compound, from which it may be separated by distillation in a current of steam, the ortho compound passing over with the steam. It crystallizes in orange-red needles which melt at 82.5-83° C. On reduction with zinc dust in dilute sal-ammoniac solution, it yields ortho-aminophenol and aniline. Meta-oxyazobenzene, CHN:N(1)CH·OH(3), was obtained in 1903 by P. Jacobson (Ber., 1903, 36, p. 4093) by condensing ortho-anisidine with diazo benzene, the resulting compound being then diazotized and reduced by alcohol to benzene-azo-meta-anisole, from which meta-oxyazobenzene was obtained by hydrolysis with aluminium chloride. It melts at 112-114° C. and is easily reduced to the corresponding hydrazo compound.