/ \ / \ C3H5(OH)3 + C6H5NH2 + 4O(C6H5NO2)—> " " " + 4H2O \ / \n/

Prepared by H. T. CLARKE and ANNE W. DAVIS. Checked by ROGER ADAMS and A. W. SLOAN.

1. Procedure

IN a 5-l. round-bottom flask, fitted with an efficient reflux condenser of wide bore, are placed, in the following order, 80 g. of powdered crystalline ferrous sulfate, 865 g. of glycerol (c. p.), 218 g. of aniline, 170 g. of nitrobenzene, and 400 cc. of concentrated sulfuric acid (sp. gr. 1.84). The contents of the flask are well mixed and the mixture heated gently over a free flame. As soon as the liquid begins to boil, the flame is removed, since the heat evolved by the reaction is sufficient to keep the mixture boiling for one-half to one hour. If the reaction proceeds too violently at the beginning, the reflux condenser may be assisted by placing a wet towel over the upper part of the flask. When the boiling has ceased the heat is again applied and the mixture boiled for five hours. It is then allowed to cool to about 100'0 and transferred to a 12-l. flask; the 5-l. flask is rinsed out with a small quantity-of water. The 12-l. flask is then connected with the steam-distillation apparatus shown in Fig. 3, a 12-l. flask being used as a receiver; steam is passed in (without external heat) until 1500 cc. have distilled (ten to thirty minutes). This removes all the unchanged nitrobenzene (10-20 cc.). The current of steam is then interrupted, the receiver is changed, and 1500 g. of 40 per cent sodium hydroxide solution are added cautiously through the steam inlet. The heat of neutralization is sufficient to cause the liquids to boil and thus become thoroughly mixed. Steam is then passed in as rapidly as possible until all the quinoline has distilled. In this process, 6-8 l. of distillate are collected (two and a half to three and a half hours are required, unless a very efficient condensing apparatus is used, under which conditions the distillation may be complete in one-half to one and a half hours). The distillate is allowed to cool, and the crude quinoline separated. The aqueous layer of the distillate is again distilled with steam until all the quinoline has been volatilized and collected in about 3 l. of distillate.

These 3 l. of distillate are then mixed with the first yield of quinoline and 280 g. (150 cc.) of concentrated sulfuric acid are added. The solution is cooled to 0-5'0, and a saturated solution of sodium nitrite added until a distinct excess of nitrous acid is present (as shown either by starch-potassium iodide paper or by the odor). This generally requires 50 to 70 g. of sodium nitrite. The mixture is then warmed on a steam bath for an hour, or until active evolution of gas ceases, and is then distilled with steam until all the volatile material has been expelled (41. of distillate will result) The receiver is then changed and the mixture in the distillation flask is neutralized, as before, with 700 g. of 40 per cent sodium hydroxide solution. The quinoline is distilled exactly as described above, the aqueous portions of the distillate being distilled with steam until all the quinoline has been isolated. The crude product is then distilled under reduced pressure, and the fraction which boils at 110-114'0/14 mm. is collected. The foreruns are separated from any water which may be present, dried with a little solid alkali, and redistilled. The total yield is 255-275 g. (84-91 per cent of the theoretical amount based on the aniline taken).

2. Notes

Although these directions have been used many times with results exactly as described, in a few cases the yields have dropped to 60-65 per cent without any apparent reason. At present no explanation can be given for this.

In the Skraup synthesis of quinoline the principal difficulty has always been the violence with which the reaction generally takes place; it occasionally proceeds relatively smoothly, but in the majority of cases gets beyond control, with consequent loss of material through the condenser. By the addition of ferrous sulfate, which undoubtedly functions as an oxygen carrier, the reaction is extended over a longer period of time. It is thus possible to work with much larger quantities of material when ferrous sulfate is employed.

It is important that the materials should be added in the correct order; should the sulfuric acid be added before the ferrous sulfate, the reaction may start at once. It is also important to mix the materials well before applying heat; the aniline sulfate should have dissolved almost completely and the ferrous sulfate should be distributed throughout the solution. To avoid danger of overheating, it is well to apply the flame away from the center of the flask where any solids would be liable to congregate.

In the apparatus for steam distillation, the greater portion of the condensation is effected by the stream of water passing over the receiver. It is, therefore, necessary that the stream passing through the condenser should be sufficiently rapid to cause it to form a uniform film over the receiving flask. A 12-l. flask is even more efficient as a condenser than the 5-l. flask. It is important that the tube through which the vapors leave the distillation flask should be neither too short nor, especially, too narrow. Where the external diameter of the steam inlet tube is 5-8 mm., the internal diameter of this steam head should be not less than 28 mm. Were it less, the current of steam passing through it would be so rapid as to prevent small quantities of liquid from returning to the flask, and these would be driven over into the receiver.

Much time can be saved by the use of the steam distillation apparatus described, especially when large quantities have to be handled. The above directions avoid the use of extraction methods, which not only consume more time but may lead to appreciable losses of material. If the downward condenser is of iron, the apparatus is even more efficient and the time for the steam distillation is halved.

The percentage yields have been based on the amount of aniline taken. It would probably be more legitimate to base the calculation on the amounts of aniline taken and of nitrobenzene not recovered, since undoubtedly the latter is reduced to aniline during the course of the reaction. If this be done, the yield is found to be only 55 to 60 per cent of the calculated amount.

In a number of experiments, the glycerol used contained an appreciable amount of water. Under these conditions, the yield of product is much lower. "Dynamite" glycerol containing less than half a per cent of water is best employed; U. S. P. glycerol contains 5 per cent of water and usually gives lower yields.

3. Other Methods of Preparation

Quinoline has been produced by passing the vapor of allylaniline over red-hot lead oxide;[1a] by heating acrylideneaniline, or better, a mixture of aniline, glycerol and sulfuric acid;[2a] by heating aniline with glycerol and sulfuric acid, using nitrobenzene as an oxidizing agent;[1] by treating a mixture of glyoxal and o-toluidine with alkali;[2] by treating a solution of o-aminobenzaldehyde with acetaldehyde and alkali;[3] by heating methylacetanilide with zinc chloride;[4] by heating aminoazobenzene with glycerol and sulfuric acid;[5] by heating a mixture of aniline, glycerol and sulfuric acid with arsenic acid.[6]

[1a] Ber. 12, 453 (1879).

[2a] Ber. 13, 911 (1880); Monatsh. 1, 316 (1880).

[1] Monatsh. 2, 141 (1881); J. prakt. Chem. (2) 49, 549 (1894),

[2] Monatsh. 15, 277 (1894).

[3] Ber. 15, 2574 (1882); 16, 1833 (1883).

[4] Ber. 23, 1903 (1890).

[5] Ber. 24, 2623 (1891)

[6] Ber. 29, 704 (1896)

Of the above methods, the only ones which need be considered are those in which a mixture of aniline, glycerol and sulfuric acid is heated with an oxidizing agent. With the use of nitrobenzene, the reaction, according to the original method, takes place with extreme violence.

The method above described is the most satisfactory for the preparation of quinoline itself, but for the preparation of homologues of quinoline, the use of arsenic acid is preferable, since the yields are somewhat greater.

Since the work was carried out, a method has been published[7] in which aniline, glycerol and sulfuric acid are treated with ferric oxide. By this method Adams and Parks were unable to obtain yields comparable with those resulting from the above directions.

[7] Chem. News 121, 205 (1920).