Azoxybenzene, from nitrobenzene

Azoxybenzene from Nitrobenzene by Electrolysis.— Nitrobenzene can be conveniently converted into azoxybenzene by electrolytic reduction. The apparatus required is shown in Fig. 77. 

It can be seen from this figure that the steady state production of nitrosobenzene is preceded by an induction period, in which aniline is the main product. Further, small amounts of azobenzene and azoxybenzene are formed throughout the reaction. The existence of an autoredox reaction implies that a selectivity of 100% from nitrobenzene to nitrosobenzene is impossible. After the induction period the selectivity of nitrobenzene to nitrosobenzene becomes above 90% of the reduction products. The extent of conversion of nitrobenzene is also time dependent. In the steady state situation about 20% of the nitrobenzene is converted, after an initial conversion of 65%. 

When reduced by electrolysis, nitrobenzene and its homologues yield the same products as may be obtained by the various chemical methods of reduction. Aniline, azobenzene, azoxybenzene, hydrazobenzene, and -amino-phenol, as well as phenylhydroxylamine, can thus be obtained from nitrobenzene, and most, if not all, of these products could be prepared satisfactorily on an industrial scale by electrolysis, by adjusting the manner of working so that economy of energy is combined with maximum yields. Many of these products demand a comparatively high price, so that low power cost is not so important in this class of manufacture as high percentage yields. 

Zinc and Weak Alkali The mechanism of reduction with zinc and strongly alkaline solutions leads to the formation of azoxybenzene as the first stable product in the reduction of nitrobenzene. Bamberger has shown that, when faintly alkaline systems are employed, the principal initial stable product of reduction is N-phenylhydroxylamine. The systems —zinc + calcium chloride and zinc + ammonium chloride—are not neutral but quite alkaline to phenolphthalein. In the preparation of N-phenylhy-droxylamine from nitrobenzene by means of zinc dust and aqueous calcium chloride, the reduction liquid has a pH Between 10.5 and 11.7, which corresponds to the alkalinity of 0.01 N alkali. When NH4CI is substituted for CaCb in the reducing system, the pH is between 8 and 9. 

Equip a 500 ml. three-necked flask with an efficient stirrer (e.g., a Hershberg stirrer. Fig. II, 7, 8) and a reflux condenser stopper the third neck. Place a solution of 30 g. of sodium hydroxide in 100 ml. of water, and also 20-5 g. (17-1 ml.) of pure nitrobenzene in the flask, immerse it in a water bath maintained at 55-60°, and add 21 g. of anhydrous dextrose in small portions, with continuous stirring, during 1 hour. Then heat on a boiUng water bath for 2 hours. Pour the hot mixture into a 1 litre round-bottomed flask and steam distil (Fig. II, 40, 1) to remove aniline and nitrobenzene. When the distillate is clear (i.e., after about 1 htre has been collected), pour the residue into a beaker cooled in an ice bath. The azoxybenzene soon sohdifies. Filter with suction, grind the lumps of azoxybenzene in a mortar, wash with water, and dry upon filter paper or upon a porous plate. The yield of material, m.p. 35-35-5°, is 13 g. Recrystallise from 7 ml. of rectified spirit or of methyl alcohol the m.p. is raised to 36°. ... 

Hurley and Testa (Ref 17) exposed nitrobenzene in isopropyl alcohol, degassed and in air, to a mercury lamp at 3660A Products in the absence of air were acetone and phenyl-hydroxylamine (PHA). In air PHA was oxidized to nitro sob enzene which couples with PHA to form azoxybenzene. They hypothesized that the triplet molecule abstracted H-atoms from the solvent no effect was noted with ben zene as solvent. They also worked with nitrobenzene in isopropyl alcohol-water mixts containing HC1 with a mercury lamp at 3660A (Ref 18), and found that the quantum yields depended on pH and isopropyl alcohol content, but were independent of oxygen with acid present. Their conclusion was that the quantum yield consisted of two parts, H abstraction by the triplet, and protonation of the triplet... 

The photoreduction of nitrobenzene using p5o ex filtered light from a medium pressure mercury arc was studied in petroleum, toluene, ether, 2-propanol, tert-butyl alcohol, diethylamine, triethylamine, aqueous solutions of 2-propanol and diethylamine and also in aqueous t-butylalcohol containing sodium boro-hydiide 3 >. Varying amounts of aniline, azo- and azoxybenzene were obtained. In the presence of a fourty-fold excess of benzophenone, a six-fold increase in the rate of aniline formation in ethereal solution was observed, and aniline formation was completely suppressed by addition of biacetyl or octafluomaphthalene Since unreacted nitrobenzene could be recovered in these experiments, it is demonstrated that the triplet state of nitrobenzene was quenched. 

From Figure 5 it can be clearly seen that nitrosobenzene totally inhibited nitrobenzene hydrogenation. The rapid adsorption and formation of azoxybenzene indicated that nitrosobenzene was more strongly adsorbed than nitrobenzene and that, given its high surface concentration, the principal surface reaction was coupling to form azoxybenzene with loss of water as shown in the reaction sequence ... 

In the reduction of nitrobenzene in a 2% aqueous sodium-hydroxide solution, according to previous publications, azoxy-benzene is formed at platinum and nickel electrodes, azobenzeno at lead, tin, and zinc cathodes, and aniline at copper cathodes especially in the presence of copper powder. It was found that, in an unchangeable experimental arrangement, a cathodo potential of 1.8 volts, as measured in connection with the deci-normal electrode, could be carried out with all the chosen cathodes and additions. At this constant potential, by using different metals and adding various metallic hydroxides, the whole reduction was carried out and the nature and quantity of the reduction products determined in each case. It turned out that the emphasized differences in the results disappeared and that, with an equal potential of all cathodes, similar yields of azoxybenzene and aniline and traces of azobenzene resulted. The cathodes were of platinum, copper, copper and copper powder, tin, platinum with addition of stannous hydroxide zinc, platinum with addition of zinc hydroxide, lead, platinum with addition of lead hydroxide, and nickel. The yields of azoxybenzene varied from 41-65% of aniline 23-53%. 

An investigation of nitrosobenzene in alkaline solution is complicated by the reaction of nitrosobenzene with hydroxyl ions [163,164]. There is a relatively fast, reversible addition of hydroxyl ion to the nitroso group. Besides that, there is in aqueous solutions at pH > 13 in the absence of dioxygen a relatively slow reaction giving mainly a mixture of azoxybenzene and 4-hydroxynitrosobenzene (p-quinone monoxime) in a reaction suggested as involving a hydride ion transfer from the initially formed 4-hydroxycyclohexa-diene oxime anion to nitrosobenzene hydride ion transfer to nitrosobenzene probably also occurs in alcoholic solutions when an alkoxide is present. In the presence of dioxygen, some nitrobenzene is formed. 

Place in an eight-inch test tube arranged for reflux 1 g of nitrobenzene, 6 ml of methanol, and 1.4 g of solid sodium hydroxide. Heat for thirty minutes or until the odor of nitrobenzene is very faint. Add the reaction mixture to 30 ml of water, and cool. The oily azoxybenzene which separates out soon solidifies. Filter the crystals. It may be lecrystallized from methanol. 

---