transferring the crystals from the reaction flask to the Buchner funnel it is necessary to use a certain amount of water to dissolve the pasty chromium salts which are otherwise quite impossible to filter. The amount necessary varies greatly in different runs, according to the manner in which the chromium salts separate. The amount of this water is kept low in order to dissolve as little of the product as possible. Nevertheless, 10-15 g. of dichloroacetone are thus dissolved; this material, together with a little unchanged dichlorohydrin, may be recovered by a long procedure involving extraction with ether and sodium bisulfite. This is not profitable, however.

It is not necessary to wash the crystals in the centrifuge until they are white. A small amount of chromic salt will not interfere with the subsequent purification.

Commercial sodium dichromate is hygroscopic and contains varying amounts of water. The 375 g. required in these directions are equivalent to 319 g. of anhydrous material.

The total time required for the oxidation is twenty-four hours. It is convenient to start the reaction in the morning. In this way the last part of the reaction, which requires no attention, will be accomplished during the night. The regulation of the temperature is necessary, as the reaction proceeds very slowly below 20'0; on the other hand, the dichloroacetone itself is oxidized at a somewhat higher temperature than 25'0. 3. Other Methods of Preparation

The preparation of dichloroacetone by the following methods is described in the literature: the direct chlorination of acetone;[1] the oxidation of dichlorohydrin;[2] the action of silver chloride on diiodoacetone;[3] the action of dichloropropene (CH2Cl-CCl=CH2) and hypochlorous acid;[4] the action of hydrochloric acid on ethoxymonochloroacetoacetic ester;[5] and the hydrolytic cleavage of dichloroacetoacetic ester.[6]

[1] Jahresb. 1859, 345; 1871, 531; J. prakt. Chem. (2)4, 52 (1871); Ber. 7, 467 (1874); 8, 1330, 1438 (1875); 26, 598 (1893); 42, 3233 (1909); Ann. 279, 315 (1894)

[2] Ber. 6, 1210 (1873); 13, 1706 (1880); 42, 3233 (1909); Ann. 208, 355 (1881); 269, 46 (1892); Ann. chim. phys. (6) 9, 145 (1886); Bull. soc. chim. (2) 36, 19 (1881).

[3] Ann. 192, 93 (1878).

[4] Compt. rend. 94, 1428 (1882).

[5] Ann. 269, 18 (1892).

[6] Ber. 43, 3533 (1910).

V

p-DIMETHYLAMINOBENZALDEHYDE

(CH3)2NC6H5 + HNO2—> (CH3)2NC6H4NO + H2O (CH3)2NC6H4NO + 2HCHO
                        + 2C6H5N(CH3)2 —> (CH3)2NC6H4N
        = CHC6H4N(CH3)2 + 2H2) + (CH3)2NC6H4CHO
(CH3)2NC6H4N = CHC6H4N(CH3)2 + HCHO—>( CH3)2NC6H4N = CH2 + (CH3)2NC6H4CHO

Prepared by ROGER ADAMS and G. H. COLEMAN. Checked by H. T. CLARKE and W. W. HARTMAN.

1. Procedure

IN a 3-l. round-bottom flask fitted with a mechanical stirrer 150 g. of technical dimethylaniline are dissolved in 750 cc. of diluted hydrochloric acid (1 part concentrated acid to 1 part water). This solution is now cooled to 0'0 and a solution (previously cooled to 0'0) of 90 g. of technical sodium nitrite in 150 cc. of water is added through a separatory funnel. During the addition of the nitrite solution, mechanical stirring should be employed and the flask cooled well with ice and salt. The addition is made at such a rate (thirty to forty minutes for the entire addition) that the temperature does not rise above 5'0. The precipitate of nitroso dimethylaniline hydrochloride is filtered off with suction, then washed with about 300 cc. of diluted hydrochloric acid (1:1).

In a 2-l. beaker, 180 g. of technical dimethylaniline, 125 cc. of formaldehyde (technical 40 per cent), and 300 cc. of concentrated hydrochloric acid are mixed and heated for ten minutes on a steam bath. The mixture is now placed in a hood and the nitroso dimethylaniline added all at once, or as rapidly as possible. The beaker is then covered with a watch glass. A vigorous reaction soon occurs and is complete in about five minutes. The resulting solution is transferred to a 5-l. flask and diluted to 4 l.; stirring is started, and a 25 per cent solution of sodium hydroxide is added until the red color disappears (about 650 cc. are required). The yellow benzylidene compound separates, is filtered with suction and washed with water. The moist precipitate is transferred to a 4-l. glass jar, covered with 1000 cc. of 50 per cent acetic acid and 250 cc. of formaldehyde, and stirred until twenty minutes after the benzylidene compound has gone into solution. While the mixture is being stirred vigorously to prevent lumping of the precipitate, 400 cc. of water and 200 g. of cracked ice are added during the course of five minutes. The dimethylaminobenzaldehyde generally separates gradually in fifteen to twenty minutes, but in some cases does not. If the precipitate does not form, the solution is placed in a refrigerator for a few hours or overnight. The mixture is filtered with suction and washed at least ten times with 300 cc. of water. The precipitate is sucked as dry as possible for fifteen to thirty minutes.

The slightly moist aldehyde is distilled under diminished pressure from an oil bath, by means of a 1-l. Claisen flask. A small amount of water comes over first, then the thermometer rises rapidly to the boiling point of the aldehyde (180'0/22 mm.). In changing receivers between the water fraction and the aldehyde, care should be taken to keep the side-arm of the distilling flask warm; otherwise, on starting the distillation again, the aldehyde will solidify in the side-arm and cause trouble. It is advisable not to collect the very last portion of the distillate with the main portion, as the former is frequently quite red. This is best added to crude material from another run. The main distillate is dissolved in 100 cc. of alcohol in a 2-l. beaker, then 1000 cc. of water are gradually added with vigorous mechanical stirring to prevent lumping. The aldehyde separates, and is filtered with suction. The product, when dry, weighs 125-130 g. (56-59 per cent of the theoretical amount), and melts at 73'0.

The aldehyde prepared in this way is in the form of small granular crystals, which vary in different runs from a flesh color to a lemon yellow. For practically all purposes, this slightly colored product is entirely satisfactory and is essentially pure, as can be judged by the melting point. For reagent purposes it is desirable to remove the color completely, particularly since the product obtained as just described has a tendency to take on a reddish tinge on exposure to light. Further purification can be accomplished by dissolving the aldehyde (it dissolves slowly) in dilute hydrochloric acid (1 part of concentrated acid, sp. gr. 1.19, to 6 parts of water), 125 g. of aldehyde requiring 700 cc. of the acid. The solution is placed in a jar and diluted with half its volume of water, and dilute sodium hydroxide solution (15-20 per cent) is added slowly with mechanical stirring. At the beginning, the aldehyde comes down slightly colored. After about 10 to 30 g. are precipitated, however, the product appears white; this point can be readily seen. The first precipitate is filtered off and added to the next run of crude material, or fractionally precipitated again from hydrochloric acid. The rest of the aldehyde is now precipitated by means of more sodium hydroxide solution, and comes down almost white. At the very end of the neutralization, particularly if the original product was quite yellow, the last 4 to 5 g. of aldehyde should be precipitated separately, as they are inclined to be slightly colored. If too much alkali is added towards the end of the