942
INDUSTRIAL A N D ENGINEERING CHEMISTRY
Vol. 17, No. 9
Analysis of Acetic Anhydride' By W. S. Calcott, F. L. English, and 0. C. Wilbur E. 1. DU
P O N T DE
NEMOURS & CO., INC.,WILMINGTON,DEL.
T
a
, H E usual method for determining the purity of acetic solution was filtered and cooled with ice. The hydrochloride a n h y d r i d e 4 e., hydrolysis of acetic anhydride with was filtered off and washed with about 500 cc. of 10 per cent an excess of standard alkali and titration of the excess hydrochloric acid. The crystals were dissolved in 4.5 liters alkali with standard acid-is unreliable because of the error of water, containing 200 cc. of 35 per cent hydrochloric acid, introduced through the assumption that acetic acid is the a t 60" to 65" C. This solution was filtered, cooled, and the only impurity in the anhydride. It also seems probable that free base precipitated by adding 30 per cent sodium hythere is a side reaction in the hydrolysis that gives lorn results droxide, with vigorous agitation, until the solution was only for the anhvdride. slightly acid to litmus. The dichloroThe metiods in which the acetylaaniline was filtered and washed until tion value is determined with aniline the wash water was neutral. The The Orton method for estimatare subject to an error due to the reacbase was dried a t 45" to 50" C. in ing small amounts of acetic antion of the aniline with the acetic acid vacuo. hydride in glacial acetic acid has present in the anhydride. Acetic Anhydride-The preparation been adapted to the determination The method of Walton and Withof pure acetic anhydride by fractionaof acetic anhydride. By this prorowz for acetic acid in acetic anhytion was found to be impractical. cedure the acetic anhydride is dedride is only readily applicable to samThe method of treating with metallic termined by acetylation with 2,4ples containing 5 per cent or less of sodium gave material of a high purity, dichloroaniline dissolved in glacial the acid and, as the anhydride is taken but the procedure is not suitable for acetic acid, and titrating the exas the difference, an error is introduced routine purposes, as a t 65" to 70" C. cess of dichloroaniline with sodue to impurities other than acetic acid t h e s o d i u m and anhydride react dium nitrite in the presence of in the anhydride. violently. the acetanilide formed. The method of Orton3 for estimaIt was found that magnesium filings This method has been applied to tion of small amounts of acetic anhyor turnings could be used in place of two samples of anhydride (Table dride in glacial acetic acid is adaptthe sodium. A sample of 1450 grams IV, Samples 2 and 4) which had able to the determination of acetic of 95 per cent anhydride was heated been treated to reduce the acetic anhydride. However, the procedure is a t 80" to 90" C. under a reflux conacid content to 0.3 per cent. The lengthy and involves several chlorodenser with 200 grams of coarse maganhydride contents were found to form extractions, which makes it unnesium filings for 5 days. The anhyaverage 99.0 and 98.8 per cent, redesirable for routine use. In addition, dride was protected from water vapor spectively, indicating the presence the 2,4-dichloroacetanilide is converted by a calcium chloride tube attached of about 1 per cent of an unknown to the corresponding c h l o r o a m i n e , to the top of the condenser. The impurity. which is unstable, decomposing rapidly anhydride was then refluxed slowly in direct sunlight, so that rapid handfor 17 hours, after which it was fractionated using a 30-em. (12line is necessarv to avoid errors due to -its decompoktion. The following modification of this inch) Hempel column. method has been developed which overcomes these difficulties. Experimental Data
Preparation of Materials I , d-Dichloroaniline4-Fourteen hundred grams of glacial acetic acid, 300 grams of acetanilide, and 390 grams of
freshly fused sodium acetate were weighed into a balloon flask fitted with a stopper carrying a reflux condenser and gas delivery tube. Chlorine was passed into the solution a t room temperature until the apparatus had gained about 150 grams; the flask was then placed in boiling water and the chlorination continued until the total gain in weight was 310 to 315 grams. The sodium chloride was filtered off from the hot solution and washed with a little hot glacial acetic acid. The hot filtrate was diluted with 3 liters of water and the suspension cooled to 10" to 15" C. The 2,4-dichloroacetanilide was filtered and washed thoroughly with distilled water . The crystals were then hydrolyzed by refluxing for 1 to 5 hours with 1 liter of water, 600 cc. of 35 per cent hydrochloric acid, and 25 cc. of 95 per cent alcohol. The hot 1 Presented before the Division of Dye Chemistry a t the 69th Meeting of the American Chemical Society, Baltimore, Md., April 6 t o 10, 1926. 2 J. A m . Chem. Soc., 46, 2689 (1923). 1. Chem. SOL. (London), 99, 1151 (1911). 4 I b i d , , 91, 1663 (1907).
A study was made of the method of hydrolyzing acetic anhydride with 0.1 N sodium hydroxide and titrating the excess alkali with 0.1 N hydrochloric acid. The water used showed no acidity or alkalinity. Pure sodium acetate was prepared by recrystallizing the C. P. material from water and drying 16 hours a t 50" C. in vacuo. The final material was assumed to be pure sodium acetate plus about 0.5 mol of water. Seven-tenths of a gram of this material (the amount present in an average titration) dissolved in 250 cc. of water, gave a barely perceptible pink with phenolphthalein. The color was completely discharged by 1drop of 0.1 N hydrochloric acid. When a similar solution was heated to boiling 0.08 cc. of 0.1 N hydrochloric acid was required. This amount is insignificant, since the end point cannot be detected closer than 1 drop. It was found that sodium acetate could not be boiled in dilute solution in the presence of 10 cc. of 0.1 iV hydrochloric acid without the loss of acetic acid proportional to the time of boiling and concentration of the acetate. At this point in the work samples of anhydride were prepared by fractionation, using a 30-cm. (12-inch) Hempel column. The data obtained from these distillations gave 140.2" C., 760 mm., as the boiling point for acetic anhydride.
September, 1925
I-VDUSTRIAL A N D ELVGINEERIYGCHEMISTRY
The fraction collected a t this point gave 94 to 98 per cent anhydride by hydrolyzing with 0.1 iV sodium hydroxide. Various methods of determining the acetic acid in the acetic anhydride v-ere next tried. In the first, an attempt was made to use the reaction between acetic acid and lead monoxide, which should give lead acetate and water, and the reaction of the water with calcium carbide to form acetylene. This procedure gave 13.1 per cent of acetic acid in a sample of C. P. acetic anhydride. h sample of magnesium oxide was ignited for 16 hours a t 1100" to 1200" C. and solvent naphtha was distilled from calcium oxide through a column containing calcium oxide. The foregoing experiment was repeated using these reagents. The volume of acetylene when the experiment was stopped represented 23 8 per cent of acetic acid, and gas was still being evolved rapidly. If a mixture of acetic acid and acetic anhydride is boiled with zinc dust the acid should react with the liberation of an equivalent amount of hydrogen. A IO-gram sample of twice-fractionated acetic anhydride, boiling range 139.8' to 140.8" C., was refluxed in a Knorr flask for 2 hours with 10 grams of zinc dust. The gas evolved calculated to 0.73 per cent acetic acid. The experiment was repeated using 50 grams of anhydride. During the first 2.5 hours the gas evolved calculated to 0.42 per cent of acetic acid. This reaction mixture was then boiled another hour and another 15 cc. of gas was collected, giving a total of 0.57 per cent of acetic acid. This reaction is either progressive, with gradual decomposition of the anhydride, or it is too slow to be of value as a routine procedure. Magnesium was substituted for zinc in the foregoing procedure and the following results were obtained, using a sample of acetic anhydride that had been distilled from zinc (boiling range 139.8' to 140.8" C.). T a b l e I-Acetic Acid in Acetic A n h y d r i d e as D e t e r m i n e d b y M e a s u r i n g H y d r o g e n Evolved on Refluxing w i t h M a g n e s i u m (25 grams acetic anhydride and 10 grams magnesium for each experiment) Expt. 1 Hours 2 3 4 Cc. total gas evolved 70.9 87.8 100.2 Per cent acetic acid 1.39 1.72 1.97 2 Hours 2 3 4 Cc total gas evolved 94.0 103 9 112.4 Per cent acetic acid 1.88 2.04 2.20 3 Hours 1 2 3 Cc. total gas evolved 49.5 76.5 92.0 Per cent acetic acid 0.57 1.50 1.81 4 Hours 1 2 3 63.4 75.6 Cc. total gas evolved 47.0 1.24 1.49 Per cent acetic acid 0.92
943
T a b l e 11-Extent of R e a c t i o n a t 25" C. b e t w e e n 2.0 G r a m s of 2,4D i c h l o r o a c e t a n i l i d e a n d 25 Cc. of G l a c i a l Acetic Acid (Cc. of 0 1 A' Na,S?Oi required) 7 Hours for Acetylation 1 17 23 68 0.60 1.40 2.20 3.8 0.60 2.10 5.5s 1.35
a
0.65 2.0 a 0.50 2.5 a 0 58 12 5 cc of glacial acetic acid.
Using the procedure outlined, a sample of commercial 95 per cent acetic anhydride was analyzed, giving the results, corrected for the blank, shown in Table 111. T a b l e 111-Analyses Time for acetylation Hours 1
of C o m m e r c i a l 95 Per c e n t Acetic A n h y d r i d e Acetic anhydride Per cent REMARKS 94.8 High owing t o incomplete ei.aporation of chloroform
IT
93.7 93.0 53.5 93.3 93.i 61.0
17 23 23 68 68
92.2 52.5 93.4 93.0 93.6
1 1 1 1 1
Low light owing on chloroamine t o action of sun-
A sample of pure 2,4-dichloroacetanilide was converted t o the chloroamine, as in the foregoing procedure, in order to test the accuracy of the latter part of the procedure. The results obtained were 99.2, 99.2, 99.7, and 99.7 per cent, which indicates that the conversion and titration are reliable. Homever, this method requires considerable time and very careful manipulation, JT-hich makes it undesirable for a routine method. T a b l e I\'-Analysis Sample
Acetic acid Per cent
1
3.2
of P u r e Acetic A n h y d r i d e Acetic Total anhydride accounted for Per cent Per cent 55.7 96.3 55.0
Average, . . 9.5. 7 2
0.28
3 4
99.3
0.50
98.2
9s.i
0.30
98.9 98.7 98.9 98.8 . 98.8
99.1
Average..
Average
The same criticism applies to this procedure as to that using zinc. A study of OrtonJs3method for acetic anhydride in acetic acid showed that it could be applied, with slight modifications. to the determination of acetic anhydride. In brief, this procedure as used in this investigation consists in the quantitative formation a t 25' C. of 2,4-dichloroacetanilide from acetic anhydride in the presence of an excess of 2,Pdichloroaniline dissolved in glacial acetic acid. The anilide and part of the aniline were extracted from the diluted acetic acid with chloroform. The excess 2,4-dichloroaniline was extracted with 10 per cent hydrochloric acid and the anilide was converted to the chloroamine. Orton's method for remoring the chloroform was found to be too slow. Eraporation under reduced pressure a t 25' C. was satisfactory. The chloroamine was finally treated with water, potassium iodide, and acetic acid; the iodine liberated was titrated with 0.1 -1-sodium thiosulfate solution. The extent of the reaction of 2,4-dichloroaniline with acetic acid was determined by following this procedure. The results are given in Table 11.
T a b l e V-Analysis
1
Acetic acid Per cent 2.5
2
4.1
Sample
98.9
99.2 98.7 59.4 98.7 59.0
,
.
of C o m m e r c i a l A c e t i c A n h y d r i d e Acetic Total anhydride accounted for Per cent Per cent Average.
Aierdge
..
96.5 95,s 96.1
98.6
.
95,o 55.1 95.6 55 2
99 3
A modification of Orton's procedure was developed in which the chloroamine conversion was eliminated. After the acetylation the 2,4-dichloroacetanilide was separated from the excess 2,4-dichloroaniline by the usual extraction and washing with 10 per cent hydrochloric acid. The 2,4-dichloroacetanilide was then hydrolyzed by refluxing with 19 per cent hydrochloric acid and the resulting 2,kdichloroaniline was titrated with 0.1 11- sodium nitrite. It seemed feasible to use a known amount of 2,4-dichloroaniline and determine the excess by titrating with 0.1 i V sodium nitrite. It was known that such a procedure is possible in the case of other amines
944
I N D U S T R I A L A N D ENGINEERING CHEMISTRY
Vol. 17, No. 9
and their anilides. The titration of 2,4-dichloroaniline with 0.1 N sodium nitrite in the presence of hydrochloric acid a t 25' C. was found to proceed rapidly, giving a purity of 99.6 per cent for material prepared as described above. The presence of the 2,4-dichloroacetanilide had no effect upon the titration of the 2,4-dichloroaniline.
pleted within 1 hour of the time of dilution of the acetylation mass. ( b ) Weigh carefully 0.9 to 1.0 gram of the dichloroaniline, and proceed as above with the exception that no anhydride is added. Titrate with 0.1 N sodium nitrite as above ( L ) . (c) Determine the amount (0)of 0.1 N sodium nitrite required to give a test for nitrous acid in the same volume of water, hydrochloric and acetic acids used in the above determinations.
Procedure
Calculations
(a) Weigh into a 300-cc. Erlenmeyer flask 2.000 grams of
2,4-dichloroaniline and add 25 cc. of glacial acetic acid. Weigh carefully from 0.6 t o 0.7 gram of acetic anhydride (100 per cent) in a small glass-stoppered weighing bottle. Remove the stopper from the bottle and lower the bottle into the Erlenmeyer, stopper the flask, sind rotate it so as to wash the anhydride out of the bottle. Allow the flask to stand at 25' C. for 1 hour. Transfer the solution to a 400-cc. beaker, using I50 cc. of water containing 25 cc. of 35 per cent hydrochloric acid a t 15' to 20' C. Adjust the temperature of the solution to 20'-25' C. and titrate the excess dichloroaniline with 0.1 N sodium nitrite, using a 1-minute end point with potassium iodide starch papers (M). The dichloroacetanilide present does not interfere, provided the temperature is kept below 26' C. and the titration is com-
(L-0)X wt. of dichloroaniline used in ( a )
= c c . of NaNOz Wt. of dichloroaniline used in ( b ) equivalent to the total dichloroaniline used in (a) = P - ( M - )' - Per cent of acetic anhydride Wt. of sample of acetic anhydride -
Results
Proceeding according to this method, samples of pure and of commercial acetic anhydride were analyzed, with the results shown in the Tables IV and v. The acetic acid was determined by the method of Walton and Withrow.2
Acetone b y Distillation of Wood with Lime' By A. W. Schorger C. F. BURGESS LABORATORIES. MADISON, Wis.
T H E utilization of wood waste is still largely an unsolved problem. The wood distillation industry, hitherto one of the largest consumers of small-dimension refuse, is threatened in its very existence by the synthesis of acetic acid from acetylene, methanol from carbon monoxide and hydrogen, and the production of acetone in the butyl fermentation of starch. Within recent years considerable work has been done on the distillation of wood in the presence of various chemicals in order to increase the yield of valuable products. The most striking statement is that of Basset,* who claimed a yield of 26 per cent of acetone and mixed ketones by distilling wood with 2.5 parts of lime. Hawley,s using up to 60 per cent of lime, did not obtain an increase in acetone, the yield varying from 0.01 to 0.05 per cent. Fremy4in 1835 showed that acetone was obtained by distilling sugar, gum, and starch with 8 parts of lime. No yields of acetone were given and, contrary to the statement of Basset, Fremy does not mention having used wood. Several experiments were made to determine the yield of acetone from wood and modified celluloses. From 25 to 75 grams of material passing a 40-mesh sieve were thoroughly mixed with 4 parts of hydrated lime and distilled from an iron retort having a capacity of 1.3 liters. Preliminary trials with starch showed no increase in acetone by increasing the lime to 8 parts. The retort was gradually brought to a temperature of 500" C. over a distillation period of 7 to 8 hours. The distillate obtained was redistilled three times, twice after having been made slightly alkaline with sodium hydroxide, and finally in the presence of a small amount of sulfuric acid. About 70 per cent of distillate was collected each time. The final distillate was passed through a wet filter paper to remove any oil carried over, made up to a definite volume, and titrated for acetone by Messenger's method. The results are as follows:
Acetone Per cent MATERIAL by weight White pine 1.76 Aspen 2.19 Yellow birch (a) 2.17 Sellow birch ( b ) 2.05 Hydrolyzed white pine 1 . 2 8
Aoetonc Per cent 8r . weizht 1 33
a
20
0 SI 2 34
The hydrocellulose was prepared by exposing cotton to the fumes of concentrated hydrochloric acid for 50 hours and washing thoroughly. The oxycellulose was prepared with nitric acid by the method of Nastjukoff.6 The hydrolyzed wood was the residue from the semicommercial manufacture of ethyl alcohol. It will be noted that the yield of acetone is of the order of 2 per cent; it is probable that these results are somewhat too high owing to the presence of unsaturated compounds. That acetone is formed was shown by the preparation of dibenzylidene acetone (m. p. 114' C.). Basset does not give his method of analysis, but his results are explainable on the supposition that he used the titration method on a distillate containing a large amount of unsaturated compounds. A distillate from sugar maple that was merely redistilled and filtered to remove oil showed 6 per cent of acetone. Owing t o the formation of mesityl oxide by boiling acetone with calcium oxide,6 it was possible that acetone had been lost by condensation to mesityl oxide and phorone. These higher ketones are decomposed by boiling dilute acids,' also by alkalies.* T o determine this point two distillations of 50 grams of white pine with lime were made in the regular way. Sufficient dilute sulfuric acid and distilled water were added to the distillate to give a volume of 200 cc. having an acid concentration of 0.25 per cent. After refluxing 7 hours the solution was distilled, then redistilled twice after having been made slightly alkaline. The yields of acetone were 1.37 and 1.25 per cent, respectively.
1
8
2
8
Received May 23, 1925. Chem. Met. Eng., 20, 190 (1919). 8 THISJOURNAL, 14, 43 (1922). 4 Ann. chim. phys., [a] 69, 5 (1835); A n n . , 16,277 (1836).
MATERIAL Western larch hark Hydrocell u1 ose Oxycell iilose Starch
7 8
Bcr., 94, 3589 (1901). HoEman, J . A m . Chem. Soc.. 01,722 (1909) Claissen, A n n . , 180, 19 (1876). Harries, Ber., Sa, 1328 (1899).
