GIFT OF The University of Chicago Founded by JOHN D. ROCKEFELLER The C r Saccharinic Acids. I. The Resolu tion of J/-2,3-Dioxybutyric Acid into the Optically-Active Components. The Derivatives of These Acids A DISSERTATION SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE SCHOOL OF SCIENCE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF CHEMISTRY By GEORGE ELMER MILLER EASTON, PA.: PRESS OF THE ESCHENBACH PRINTING CO. 1920 The University of Chicago Founded by JOHN D. ROCKEFELLER The C 4 -Saccharinic Acids. I. The Resolu tion of J/-2,3~Dioxybutyric Acid into the Optically- Active Components. The Derivatives of These Acids A DISSERTATION SUBMITTED TO THE FACULTY OF THE OGDEN GRADUATE SCHOOL OF SCIENCE IN CANDIDACY FOR THE DEGREE OF DOCTOR OF PHILOSOPHY DEPARTMENT OF CHEMISTRY By GEORGE ELMER MILLER EASTON, PA.: PRESS OF THE ESCHENBACH PRINTING CO. 1920 THE C 4 -SACCHARINIC ACIDS. I. THE RESOLUTION OF dl- 2,3-DIOXYBUTYRIC ACID INTO THE OPTICALLY-ACTIVE COMPONENTS. THE DERIVATIVES OF THESE ACIDS. 1 The saccharinic acids are acids which would result from the oxidation of the H C = O group of an aldo-monosaccharide of the formula C n H 2w O n to the O = C OH groug at the expense of one, of the I I I H C OH groups which would be reduced to the H C H I I group. These acids have the same molecular formula as the correspond- ing aldo-monosaccharides and may thus be looked upon as the results of an internal oxidation-reduction reaction. The saccharinic acids formed from some of the pentoses and hexoses have been the direct subject of study 2 by Xef and his students at various times, and have also entered into consideration in connection with all the sugar-oxidation experiments conducted in this laboratory in recent years. These oxidations were carried out in alkaline solution, under which conditions the production of saccharinic acids was a possibility. In a recent paper from this" labora- tory 3 there was reported, in the oxidation of maltose in alkaline solution, the production of an acid, the phenylhydrazid of which gave a perfect analysis for that of a 4-carbon-atom saccharinic acid. The properties of the free acid furthermore were those which would be expected of one of these acids. Its configuration could not be reported because of lack of 1 Reprinted from an article in the Journal of the American Chemical Society for November, 1920, by J. W. E. Glattfeld and G. E. Miller. 2 Nef, Ann., 376, 1-120 (1910). 3 /. Am. Chem. Soc., 40/973 (1918). 458682 data as to the properties and constants of the 4-carbon-atom saccharinic acids. Nef has referred 1 to the handicap which this lack of data has imposed on the work with sugars in alkaline solution. In order to supply some of this deficiency we have attempted the preparation of the saccharinic acids in a systematic way and have started with the C 4 -saccharinic acids to throw light on the first question of interest in this connection, namely, which of the possible optically-active Ci-saccharinic acids is formed in the oxidation of maltose in alkaline solution. The present paper is a report on the preparation of the first pair of saccharinic acids, the 2,3-dioxy- butyric acids. Work on the preparation of the other 8 active acids is now in progress. The dl-2, 3 -dioxy -butyric acid has been prepared by Hanriot. 2 The synthesis was repeated by Nef 3 who studied the acid in somewhat more detail and reported a few of its properties as well as those of the phenyl- hydrazid and brucine salt. The constitution of the d/-acid was proved by Nef 3 by the method of synthesis and by oxidation to malic acid. We prepared our acid in accordance with the procedure of these investigators. The analysis of the vacuum-dried barium salt showed the substance to be pure. The acid was finally obtained by precipitating the barium from an aqueous solution of the salt quantitatively by means of sulfuric acid. The solution of the acid was subjected to complete distillation in vacuo and the acid was then obtained as a slightly yellow oil. It was freed from inorganic salt by repeated solution in ethyl acetate, and was then used in the resolution described below. The phenylhydrazid was made according to Nef's directions. 3 After many recrystallizations it was found to melt at 100-101 instead of at 99 as reported by Nef. In order to prove that the acid was actually d/-2,3-dioxy-butyric acid it was prepared also from vinylacetic acid CH 2 : CHCH 2 COOH by oxidation with permanganate. The vinylacetic acid was prepared accord- ing to the directions of Hauber 4 and oxidized according to the directions of Penschuk. 5 The acid so prepared gave a phenylhydrazid which melted at 100-101. A mixed melting-point determination of this phenyl- hydrazid with the one from the dl-acid above, showed them to be identical. Fichter and Sonneborn 6 also prepared the 2,3-dioxy acid from vinyl- acetic acid by the above method and state that the acid is identical with that made by Hanriot from the a-chlorohydrine. As they base their 1 Nef, Ann., 376, 9 (1910). 2 Hanriot, Ann. chim. phys. [5] 17, 62 (1879). 8 Nef, Ann., 376, 35 (1910). 4 Hauber, Ber., 36, 2897 (1903). 6 Penschuk, Ann., 283, 109 (1894). 8 Fichter and Sonneborn, Ber., 35, 942 (1902). statement only on the analysis of the barium salt for barium oxide, which would obviously be the same for the barium salts of all dioxy-butyric acids, they cannot be said to have proved the identity of the acids from the two sources. It was, therefore, considered necessary to make the proof of the identity absolute by preparing the phenylhydrazids and proving them identical before proceeding with further studies of the acid. The Resolution of the dl-2,3 Dioxy-butyric Acid into the Active Com- ponents. Brucine, cinchonine, quinine and strychnine were tried and brucine found to be best for the resolution. The brucine salt was made in the usual way 1 by adding a slight excess of the alkaloid to an aqueous solu- tion of the acid and heating the mixture on the water-bath. The use of an electrically-driven mechanical stirrer greatly hastened the solution of the alkaloid. Eighty-three g. of the dl-acid and 305 g. of brucine were heated in about 2 liters of water on the boiling water-bath until the solution reacted alkaline to litmus. Three g. of brucine did not go into solution. After extraction with benzol and subjection of the extracted solution to complete vacuum distillation, 337 g. of crude brucine salt was obtained. This was treated with 750 cc. of boiling absolute alcohol. Fifteen g. of floccu- lent material remained undissolved. This material was separated by hot filtration. The clear filtrate was allowed to stand overnight and deposited crystals which weighed 58.4 g. The mother liquor was subjected to com- plete vacuum .distillation at 60. The residue, weighing 257.5 g., was dissolved in 300 cc. of boiling absolute alcohol. This solution deposited crystals which weighed 13.55 g- The mother liquor from the second crop was subjected to complete vacuum distillation at 70. The residue weighed 204 g. This was treated with 140 cc. of boiling alcohol. This solution deposited a third crop of crystals which weighed 54.8 g. A fourth crop of crystals which weighed 17 g. was obtained in the same way by treating the residue, which weighed 141 g., with 72 cc. of hot absolute alcohol. The mother liquor from the fourth crop was sub- jected to complete vacuum distillation. The residue weighed 130 g. No more crystals could be obtained. The rotations of these 4 crops of brucine salts were taken in exactly 4% aqueous solution. The density of the solution was considered to be i.on. The following results were obtained. [i20 a in one <*JD- dcm. tube. I 29.42 I.I9 II 27.95 1.12 III 27.95 I -12 IV 26 .95 i . OQ 1 J. Am. Chem. Soc., 40, 976 (1918). Footnote. The rotation of the successive crops of brucine salts convinced us that a partial separation of the dl-acid into its optical components had been effected. Our next effort was, therefore, to obtain the less soluble brucine salt in its purest form. Another quantity of crude brucine salt was pre- pared and recrystallized as above. Again the specific rotation of the first crop was 29.42. The successive crops had approximately the same rotations as the corresponding ones in the first experiment. The salt with the specific rotation of 29.42 was now dissolved in the smallest possible quantity of hot absolute alcohol. The solution deposited a crop of crystals whose specific rotation was taken. This crop was again dissolved in the smallest possible quantity of absolute alcohol and allowed to stand overnight during which time it deposited a crop of crystals, which was separated by filtration and dried to constant weight in vacua over sulfuric acid. The rotation of this crop was taken. This process was repeated 7 times. The rotations of the successive crops were in- variably between 29.18 and 29.42. It was, therefore, concluded that the pure brucine salt of one of the optical isomers had the specific rotation of approximately 29.42. The Free Acid from the Brucine Salt, [Q:]D O , 29.42. One hundred and thirty-seven g. of the brucine salt, [a] 29.42, was dissolved in about 4 liters of hot water and treated with a hot solution of 118 g. of crystallized barium hydroxide in the usual way 1 to remove brueine from its salts. After the removal of brucine by filtration and extraction of the filtrate with benzol, the barium was removed with sulfuric acid and the filtrate from the barium sulfate was subjected to complete vacuum dis- tillation. The residue was taken up in absolute alcohol and the solution was filtered to remove any inorganic salts present. The alcohol was com- pletely removed by distillation in vacuo finally at 100, and the acid was left as a clear yellow and very mobile oil. The rotation of a portion of this acid was then taken and another portion was titrated. Other portions were converted into the barium and calcium salts and the phenylhydrazid. Rotation. The specific rotation of the acid in approximately 4% solu- tion was found to be 8.29, i. e., 2.32 g. acid dissolved in 45.78 g. water gave a in a one dcm. tube 0.40. The density of the solution was assumed to be i.oo and the temperature was approximately 20. Titration. The titra.tion showed that the acid was present largely as lactone. The amounts of free acid and lactone were determined 2 and the following results obtained. 1 /. Am. Chem. Soc., 40, 981 (1918), footnote 2. 2 The method of determining the free acid and lactone in a mixture of the two is as follows. The weighed sample is dissolved as rapidly as possible in cold water and immediately titrated to pink with o.i N sodium hydroxide solution, using phenol- phthalein as indicator. A quantity of o. i N sodium hydroxide solution in excess of Weight O.lA'NaOH 0.1 .V NaOH Total Calc. for C< of sample. for free acid. for lactone. NaOH used. saccharinic acid. a G. Cc. Cc. Cc. Cc. I 0.4765 5-72 39-75 45-47 45,70 II 0.4843 5.83 4076 4 6 -59 4647 In calculating the theoretical cc. account must, of course, be taken of the amounts of free aoid and lactone present in the samples as indicated by the titration figures. These titrations show that about 14.4% of the sample in water solution exists as free acid and the rest as lactone. This agrees with Nef's observa- tions. 1 The Barium Salt. A mixture of one g. of acid. 3 g. of barium carbonate and 75 cc. of water was heated to make this salt. The salt was gummy and could not be made to crystallize except by triturating with absolute alcohol. There was obtained 0.4 g. of the vacuum-dried salt ; [a]& + 1 .48 , i. e., 0.4 g. salt in 9.6 g. water gave a +0.03 in a half -decimeter tube. The Calcium Salt. A mixture of 1.2 g. of acid, 3 g. of calcium carbonate and 75 cc. of water was heated to make this salt. It became crystalline only when triturated with absolute alcohol. There was obtained 0.4 g. of vacuum-dried salt; rotation, [a\ +2.47, i. e., 0.4 g. of salt in 9.6 g. water gav^e a +0.05 in a half -decimeter tube. The Phenylhydrazid. A mixture of 1.75 g. of acid, 2 cc. of phenyl- hydrazine and 2 cc. of ethyl acetate gave, in the usual way 3.15 g. of the crude phenylhydrazid. This was recrystallized from 18 cc. of ethyl acetate and gave one g. of phenylhydrazid with the melting point of 102- 103. This was recrystallized from 3.7 cc. of ethyl acetate and gave 0.75 g. of crystals with a melting point of 102-103. The specific rotation of this compound was found to be +1.71, i. e., 0.54 g. in 12.06 g. of water gave a +0.07 in a one dcm. tube. The Free Acid from the Non-crystallizable Brucine Salt. The mother liquor from the 4 crops of crystalline brucine salts was subjected to com- plete vacuum distillation at 60 and gave a residue which weighed 130 g. as mentioned above. The brucine was set free in the usual way and T 5- 2 5 g- f a light brown ether-soluble oil was obtained. The specific rotation of this acid was determined as +7.18, i. e., 1.65 g. of acid in 39.6 g. of water gave a +0.287 in a one dcm. tube. As the specific rotation of the optical isomer of the acid from the brucine salt rotating 29.42 should be +8.29, the acid now under consideration was evi- the calculated amount is then added and the mixture heated on the boiling water-bath for 15 minutes. It is then cooled and a measured quantity of o.i N hydrochloric acid is added to acid reaction and the mixture boiled to expel carbon dioxide. It is again cooled and the excess hydrochloric acid is determined with o. i N sodium hydroxide. solution. The quantity of sodium hydroxide solution added at first in the cold de- termines the amount of free acid present and the remainder added determines the -quantity of lactone. 1 Ann., 376, 35 (1910). dently still contaminated with some of the racemic acid. It was, there- fore, converted into the barium salt. The 15.25 g. of acid yielded 14.8 g. of the vacuum-dried barium salt. This salt had a specific rotation of 148, i. e., one g. of salt in 24 g. of water gave a 0.03. This indicates that the salt was pure, as this rotation is equal and opposite to that of the barium salt from the crystalline brucine salt. The gum was set free from this salt and purified in the usual way. The weight of free acid was 6 g. It was a light yellow mobile oil. The rotation of a portion of this acid was taken, another portion was titrated and another portion was converted into the phenylhydrazid with the following results. The Rotation. The specific rotation of the acid was found to be +8.00, i. e., 0.83 g. of acid dissolved in 19.92 g. of water gave a +0.32 in a one dcm. tube. The Titration. A titration of another portion of the acid gave the following results. Only one titration was made because of the small amount of acid available. Weight O.lTVNaOH O.lTVNaOH Calc. for C 4 of sample. for free acid. for lactone. Total. saccharinic acid. G. Cc. Cc. Cc. Cc. 0.4014 6.l8 31.93 38.11 38.18 The Phenylhydrazid. This compound was obtained only as a gum, very soluble in ethyl acetate. Determination of the Configuration of the Two Optical Isomers. The foregoing data prove that the