In a 3-I. round-bottom flask are placed 410 g. of crude benzoin and 1 l. of 72% nitric acid (sp. gr. 1.42; 2.5 cc. of nitric acid per gram of benzoin). The mixture is heated on a water bath for about one hour with occasional shaking (note 1), until the evolution of oxides of nitrogen ceases (note 2). The reaction mixture is then poured with stirring into four to five times its volume of cold water; the benzil precipitates as small yellow lumps. The crude product is filtered and washed several times with water to remove the nitric acid. The yield after drying is 390 to 395 g. (96—97 per cent theory). Absolutely pure benzil may be obtained by crystallization from an alcohol solution. 390 g. of the crude benzil may be dissolved in about 1 l. of boiling alcohol: and since pure benzil is quite insoluble in cold alcohol, there is no difficulty in obtaining 370 g. of the pure product from the 390 g. used.
Notes
1. A number of experiments upon the oxidation of benzoin to benzil with nitric acid were made in which the mixture was stirred continuously. An inferior grade of product and smaller yields (seldom over 75 per cent) were obtained. The probable reason for this is to be found in the fact that benzoin is readily soluble in nitric acid while benzil is not, and the latter therefore floats on the top as soon as it has formed. If a stirrer is employed, it tends to bring the precipitated benzil into intimate contact with the nitric acid, causing some of it to become further oxidized or nitrated, and thus reducing the yield and quality of the product.
2. E. Fischer recommends the following test to determine when all the benzoin has been oxidized. A drop of the oil is treated with water to cause it to solidify. The crystals thus produced are dissolved in alcohol, diluted with water, a little Fehling’s solution is added and the mixture is heated to 60—70°. If benzoin is present, cuprous oxide is precipitated. In general, however, the test is unnecessary and the cessation of the evolution of oxides of nitrogen is a satisfactory end-point.
© Collected Ruslan Anatolievich Kiper, burewestnik@mail.ru