aqueous solutions of sodium benzoate.

[1] Ann. 152, 131 (1869).

[2] Gmelin's Handbuch der Organ. Chem. 3, 40.

[3] Ber. 20, 649 (1887). Cf. also J. Chem. Soc. 75, 1155 (1899).

[4] J. Am Chem. Soc. 42, 2059 (1920).

The Claisen method (c) furnishes the most convenient and practical procedure for the preparation of this ester. The materials are cheap, the experimental procedure simple, and the product obtained is free from objectionable traces of benzyl chloride. Unfortunately the method has been found to be extremely erratic in regard to yield (10-95 per cent), as well as in regard to purity of the product (87-97 per cent ester).[1] As a result of the present study,[2] causes for variations are fully accounted for and the procedure has been converted into a satisfactory method of preparation.

[1] C. A. 14, 3500 (1920).

[2] J. Am. Pharm. Assoc. 11, 599 (1922).

III

BENZYL CYANIDE

C6H5CH2Cl + NaCN—> C6H5CH2CN + NaCl

Prepared by ROGER ADAMS and A. F. THAL Checked by O. KAMM and A. O. MATTHEWS.

1. Procedure

IN a 5-l. round-bottom flask, fitted with a stopper holding a reflux condenser and separatory funnel, are placed 500 g. of powdered sodium cyanide (96-98 per cent pure) and 450 cc. of water. The mixture is warmed on a water bath in order to dissolve most of the sodium cyanide, and then 1 kg. of benzyl chloride (b. p. 170-180'0) mixed with 1 kg. of alcohol is run in through the separatory funnel in the course of one-half to three-quarters of an hour. The mixture is then heated with a reflux condenser on the steam bath for four hours, cooled and filtered with suction to remove most of the sodium chloride. It is well to wash the filtered salt with a small portion of alcohol in order to remove any benzyl cyanide which may have been mechanically held. The flask is now fitted with a condenser, and as much alcohol as possible is distilled off on the steam bath. The residual liquid is cooled, filtered if necessary, and the layer of benzyl cyanide separated. This crude benzyl cyanide is now placed in a Claisen distilling flask and distilled in vacuo, the water and alcohol coming over first, and finally the cyanide. It is advantageous to use a fractionating column or, better still, a Claisen flask with a modified side-arm[1] (Vol. I, p. 40, Fig. 3) which gives the same effect as a fractionating column. The material is collected from 135-140'0/38 mm. (115-120'0/10 mm.). The yield is 740-830 g. (80-90 per cent of the theoretical amount).

[1] J. Am. Chem. Soc. 39, 2718 (1917). 2. Notes

The quality of the benzyl chloride markedly affects the yield of pure benzyl cyanide. If a poor technical grade is used, the yields will not be more than 60-75 per cent of the theoretical, whereas consistent results of about 85 per cent or more were always obtained when a product was used that boiled over 10'0. The technical benzyl chloride at hand yielded on distillation about 8 per cent of high-boiling material; a technical grade from another source was of unusual purity and boiled over a 2'0 range for the most part.

It is advisable to distil off the last portion of alcohol and water in vacuo and also to distil the benzyl cyanide in vacuo, since under ordinary pressures a white solid invariably separates during the distillation.

One method of purifying the benzyl cyanide is to steam distil it after the alcohol has been first distilled from the reaction mixture. At ordinary pressures, this steam distillation is very slow and, with an ordinary condenser, requires eighteen to twenty hours in order to remove all of the volatile product from a run of 500 g. of benzyl chloride. The distillate separates into two layers; the benzyl cyanide layer is removed and distilled. The product obtained in this way is very pure and contains no tarry material, and, after the excess of benzyl chloride has been removed, boils practically constant. This steam distillation is hardly advisable in the laboratory.

The benzyl cyanide, prepared according to the procedure as outlined, is collected over a 5'0 range. It varies in appearance from a colorless to a straw-colored liquid and often develops appreciable color upon standing. For a product of special purity, it should be redistilled under diminished pressure and collected over a 1-2'0 range. For most purposes, such as the preparation of phenylacetic acid or ester, the fraction boiling 135-140'0/38 mm. is perfectly satisfactory. 3. Other Methods of Preparation

Benzyl cyanide occurs naturally in certain oils.[1] The only feasible method of preparing it that has been described in the literature is the one in which alcoholic potassium cyanide and benzyl chloride[2] are employed. The cheaper sodium cyanide is just as satisfactory as the potassium cyanide and therefore is the best material to use. Gomberg has recently prepared benzyl cyanide from benzyl chloride and an aqueous solution of sodium cyanide.[3]

[1] Ber. 7, 519, 1293 (1874); 32, 2337 (1899)

[2] Ann. 96, 247 (1855); Ber. 3, 198 (1870); 14, 1645 (1881); 19, 1950 (1886).

[3] J. Am. Chem. Soc. 42, 2059 (1920).

IV

a, g-DICHLOROACETONE

CH2ClCHOHCH2Cl + O(Na2Cr2O7 + H2SO4)—> CH2ClCOCH2Cl + H2O

Prepared by J. B. CONANT and O. R. QUAYLE. Checked by A. W. DOX, L. YODER, and O. KAMM.

1. Procedure

IN a 2-l. flask are placed 375 g. of commercial sodium dichromate, 225 cc. of water, and 300 g. of dichlorohydrin (b. p. 68-75'0/14 mm.). The flask is set in a water bath and equipped with a thermometer and mechanical stirrer. The contents are vigorously stirred, and 450 g. of sulfuric acid, diluted with 115 g. of water, are introduced during the course of seven to eight hours. It is convenient to add the acid at ten-minute intervals. The temperature is kept between 20'0 and 25'0 during the entire reaction; this is accomplished by adding a little ice to the water bath from time to time. The stirring is continued for sixteen to seventeen hours after all the acid has been added; as there is very little heat evolved during this part of the reaction, the water bath may be allowed to come to room temperature.

Sufficient water is now added to the mixture to dissolve the pasty chromium salts (300-800 cc.). The mass of crystals is then rapidly filtered on a Buchner funnel and sucked as dry as possible. The crystals are then transferred to a small laboratory centrifuge and centrifuged for several minutes. The crystals are washed in the centrifuge with about 15-25 cc. of ice water, then with 10-15 cc. of cold petroleum ether, and finally centrifuged till as dry as possible. The crude dichloroacetone is dried in a vacuum desiccator over sulfuric acid overnight It weighs about 220 g.

The crude product is best purified by distillation from a 250-cc. distilling flask fitted with an air condenser. A very small fraction (10-15 g.) of low-boiling material is obtained, and the dichloroacetone (170-175'0) is then collected. It solidifies in the receiver to a white crystalline mass which weighs 200-220 g. (65-70 per cent of the theoretical amount). A few grams more may be obtained by chilling the low-boiling fraction and filtering off the water.

2. Notes

Great caution should be exercised in working with dichloroacetone, as it is extremely lachrymatory and blisters the skin.

In