SEPARATION OF P-CROTONIC ACID FROM U-CROTONIC ACID, 345 XSXIX.-TlLe Sepcemtion of P-Crotonic Acid from a- Cyotonic Acid. By ROBERT SELBY ~IORRELL and ALBERT ERNEST BELLARS. THE separation of p-crotonic acid from a-crotonic acid is effected, according to Michael and Schulthess (J. pr. C'hern., 1892, [ii], 46, 245), by means of the different solubilities of the sodium salts in alcohol. One part of sodium a-crotonate dissolves in 340 parts of alcohol (98-99 per cent.) at 15", whilst at the same temperature, one part of sodium p-crotonate requires 12.7 parts of alcohol of the same strength; the more soluble p-crotonate is then decomposed by hydrochloric acid, and the free acid purified by distillation in vacuo. p-Crotonic acid, ob- tained by Michael's method or by the method described by Fittig and Kochs (AmnuZen, 1892,268, 15), is stated by J.Wislicenus (Chem. Centr., 1897, ii, 259) to contain a-crotonic acid, in spite of Michael's acid giving no precipitate in alcoholic precipitate on neutralisation346 MORRELL AND BELLARS: THE SEPARATION OF with alcoholic sodium hydroxide. The pure acid has been obtained (J. Wislicenus, Zoc. cit.) by transforming liquid p-crotonic acid, dissolved in alcohol, into its sodium salt, and by removing the sodium a-crotonate by concentrating the solution and precipitating the last traces of the salt by the addition of an equal volume of ether; the sodium p-crotonate left in solution is decomposed by hydrochloric acid, and the acid, which is purified by recrystallisation from pentane, crystallises in needles or prisms melting at; 15*45--15.5'; the yield must be small, and the method of preparation seems tedious.In a paper by one of us (Morrell and Hanson, Trans,, 1904, 85, 197), the optically active constituents of ap-dihydroxybutyric acid, prepared from a-crotonic acid, were described, and it was stated that the optically active constituents of a second up-dihydroxybutyric acid were under investigation. The oxidation of '' isocrotonic " acid, supplied by Kahlbaum, gave a n acid the brucine and quinine salts of which were virtually identi- cal with those of symmetrical ap-dihydroxybutyric acid, so that i t was necessary to devise a method of separation of the two isomeric acids, and to prepare salts of /3-crotonic acid, which are less soluble than the corresponding a-crotonates. The brucine salts of the two acids were found to have different solubilities, and the P-acid yielded the less soluble salt, but unfortunately these salts are very soluble in all ordinary solvents.The quinine salts, however, are sparingly soluble in water, and the salt of the P-acid is much less soluble than that of the a-acid; it is therefore quite easy to separate the two salts by crystallisation from water, and the quinine p-crotonate can be recrys- tallised from that solvent without undergoing any change in melting point. From the quinine salt, the barium p-crotonate has been pre- pared, and this has given the pure a,cid by treatment with dilute sulphuric acid in the presence of pure ether, The melting point agrees with that given by Wislicenus, namely, 15'.Determinations of the density, molecular weight, and molecular refraction have been made. The molecular weight in glacial acetic acid solution corresponds w i t h the formula C,H,O,, and there seems no justification for the double formula proposed by Wislicenus (Chem. Centr., 1897, ii, 259). Michael and Schulthess (Zoc. cit.) state that the purity of /I-crotonic acid can be shown by neutralising its alcoholic solution with 10 per cent. caustic soda solution; sodium p-crotonate is very soluble in alcohol, and, under the experimental conditions, no precipitate ought t o be formed on neutralisation. We have tested the p-crotonic acid prepared from the quinine salt, following Michael's directions, and have found that no insoluble sodium crotonate mas formed ; more- over, the determination of the solubility of sodium /3-crotonate gave values agreeing with bhose obtained by Michael.@-CROTONIC ACID FROM ~-CROTONIC ACID. 347 EXPERIMENTAL.Brucina a-Crotonate.-This salt, which was prepared by neutralising a boiling aqueous solution of a-crotonic acid with brucine, is exceed- ingly soluble in water, alcohol, or benzene; 5 grams dissolve in 7 C.C. of hot benzene, and the salt slowly crystallises out in hexagonal prisms, which soften a t 90' and melt at 125'. Quinine a-Crotor~ate.-This salt crystallises from a hot aqueous solu- tion in aggregates of slender needles, which melt a t 136' without decomposition ; a solubility determination a t 17' showed that 0.5903 gram was dissolved in 13-7 grams of solution, hence the solubility of the salt in water at 17' is 2.4." isoCotonic " Acid.-The acid was obtained from Kahlbaum, who stated that it was prepared by heating P-hydroxybutyric acid, and freed from a-crotonic acid, as far as possible, by distillation in vucuo and by freezing out the last traces of the a-isomeride at - 20'. Several analyses of the acid were made, but the percentages of carbon were invariably found to be too lorn (compare Michael and Schulthess, loc. cit.). 0.1845 gave 0.3732 CO, and 0.1145 H,O. C=55.16 ; H=6.89. 0,1950 ,, 0.3920 CO, ,, 0.1243 H,O. C = 54.82 ; H = 7.08. C,H,02 requires C = 55.8 ; H = 6.97 per cent. On distilling 100 grams of b L isocrotonic" acid, about 40 grams boiled at 87' under 15-30 mm. pressure; the residue in the distilling flask became solid at the ordinary temperature, and the distillate was no purer than the original acid.0.1773 gave 0.3515 CO, and 0.1190 H,O. The analysis showed that it was not possible to purify the acid by further distillation. The liquid p-crotonic acid passed slowly into a-crotonic acid and solidified completely. Brucine P-Crotonate.-This salt is very soluble in water, alcohol, or benzene ; from the last solvent, it crystallises out in prismatic needles, which melt between 98' and 100' without decomposition. Quinine P-Crotonate.-The salt was prepared from the crude p-cro- tonic acid, and also from the acid wbich had been freshly distilled; the first crop of crystals melted a t 147', and on recrystallisation from water the fraction which separated first melbed at 156-1558', the melting point being unchanged on repeated crystallisation from this solvent.The mother liquors, from the crystallisation of the fraction melting at 147', gave crystals which melted at 134-136'. It would seem as if the two quinine salts crystallised together in about C=54.06 j H=7*45.348 MORRELL AND BELLARS: THE SEPARATION OF equal quantities, and, on recrystallisation, separated into a less soluble form melting at 156-158' and a more soluble form melting at 134-136'. p-Crotonic acid, which had been distilled in vacuo at 87', gave a quinine salt which melted at 147". Seventy grams of the quinine salt (m. p. l47'), obtained from 30 grams of redistilled p-crotonic acid, gave, on recrystallisa.tion, 60 grams of salt melting a t 156'.The mother liquors were found to contain the quinine a-crotonate (m. p. 132'). There is no evidence that quinine p-crotonate changes into the a-salt on crystallisation from water. The quinine p-crotonate crystallises in prismatic needles ; an air-dried specimen of the crystals, on heating a t 105', does not lose in weight. A determina- tion of the solubility of the salt gave the following numbers: 13.5930 grams of a saturated aqueous solution at 17'contained 0.1410 gram of the salt, hence its solubility in water a t 17' is 1.04. 0,3045 gave 18 C.C. moist nitrogen a t 16' and 757 mm. 0.1858 ,, 0.4805 CO, and 0.1212 H,O. C = 70.5 ; H = 7.24. N = 6.97. C2,H2,02N2,C,H,0, requires C = 70.24 ; H = 7.31 ; N = 6.83 per cent. Bwiunz p-Crotonate.-The quinine p-crotonate (m.p. 157") was dissolved in hot water and decomposed by a slight excess of baryta water. The quinine was filtered off and a current of carbon dioxide passed through the filtrate in order to remove excess of baryta. The clear liquid was concentrated t o n small bulk in vacuo at 50' and poured into alcohol ; the barium salt was precipitated in large, lustrous, rhomboid plates, which contained a molecule of water of crystallisation. 0.9925 (air-dried) lost, at l l O o , 0.0562 H,O. 0.2255 (dried at 110') gave 0.1723 BaSO,. (C,H,02),Ba,H20 requires H,O = 5.53. Ba = 44-63 per cent. H,O=5*64. Ba=44*85. (C,H,O,),Ba requires p-Crotolzic Acid. The barium p-crotonate was covered with pure ether and a little water added; the calculated quantity of 10 per cent. sulphuric acid was added drop by drop, the mixture being thoroughly shaken; the barium sulphate mas removed by filtration and the clear ethereal solution of p-crotonic acid was concentrated in, VGCCZCO a t the ordinary temperature.On cooling, the liquid acid solidifies immediately in long needles melting a t 15'. Wislicenus gives the melting point of the acid as 15.5'; the melting point with the thermometer in the solid is 14-154P-CROTONIC ACID FROM a-CROTONIC ACID. 349 0,1856 gave 0,3765 GO, and 0.1161 H,O. C = 55.32 ; H = 6.94. 0.1895 ,, 0,3845 CO, ,, 0.1180 H,O. C=55.33; H=6*91. C4H,02 requires C = 55.8 ; H = 6.97 per cent. A molecular weight determination in glacial acetic acid gave the 0,3903 in 16.7390 glacial acetic acid gave At - 1.062'. M. W. = 85.6. The density of the acid was found to be Dg = 1.0342 ; Wislicenus (Eoc.cit.) obtained Di5 = 1.0313. Mr. Gold, of St. John's College, was kind enough to determine the refractive index of the acid for sodium light : p? = 1.4483, molecular refraction = 37.27. The molecular refraction for a-crotonic acid is 36.94 (Kannonikoff, J. p, Chenz., 1885, [ii], 31, 347). In order to test the purity of the p-crotonic acid obtained from the quinine salt, tbe following experiments were performed : (a) 0.5 gram of the acid was dissolved in 7.5 C.C. of absolute alcohol and neutralised by a 10 per cent. absolute alcoholic solution of sodium hydroxide. The solution remained quite clear, which showed t h a t the acid was free from a-crotonic acid. ( b ) Two grams of the acid mere dissolved in 29.5 C.C.of absolute alcohol and the solution carefully neutralised by the alcoholic sodium hydroxide solution. An exceedingly slight precipitate formed, which was found to contain sodium carbonate. After filtration, the solution was evaporated to dryness, and the solid residue was extracted with 50 C.C. of absolute alcohol; the solid dissolved completely. From the alcoholic solution, the sodium p-crotonate was obtained, analysed, and its solubility in the solvent was determined. 0.255, dried ztt looo, gave 0.1657 Na2S0,. Na=21. C,H,O,Na requires Na = 21.3 per cent. 4,759 grams of a saturated 99 per cent. alcoholic solution of sodium p-crotonate at 15" contained 0.3395 gram of the salt, therefore 1 part of the salt dissolves in 14.4 parts of 99 per cent. alcohol.Michael and Schulthess found that one part of the sodium ,f3-crotoaate dissolves in 16.7 parts of 99.5 per cent. alcohol a t ll", and in 12.7 parts of 98-99 per cent. alcohol a t 15O. following numbers : C,H,O, requires M. W. =86 per cent. TiLe Oxidation of Liquid p-Crotonic Acid. " isoCrotonic " acid, obtained from Kahl bauu, was oxidised by barium permanganate according t o the directions given by Fittig and Kochs (Annalen, 1892, 268, 16). The up-dihydroxybutyric acid was obtained as a syrup, although in one preparation some crystals appeared350 DIXON : CERTAIN ORGANIC PHOSPHORUS COMPOUNDS. in the form of long prisms with domed shaped ends, identical in ap- pearance with those described by Fittig and Kochs (Zoc. cit.) as characteristic of up-dihydroxybutyric acid prepared from a-crotonic acid. The yield of the syrup was the same as in the case of the oxidation of a-crotonic acid. The brucine salt of this up-dihydroxy- butyric acid is identical in melting point, specific rotatory power, and solubility with the up-dihydroxybutyric acid described in a former paper. Fractional crystallisation of the brucine salt gave a pro- duct melting a t 285O, which amounted to more than 50 per cent. of the total weight of the salt taken. The “i~ocrotonic” acid con- tained so much of the a-crotonic acid that it was found impossible to confirm Pittig and Koch’s results. Investigation of the quinine salt gave the same results as in the case of the brucine salt. The propor- tion of quinine cis-up-dihydroxybutyrate was a t least 90 per cent. of the weight of the unrecrystallised quinine salt. The silver salt of this ap-dihydroxybutyric acid was prepared, and its solubility compared with that of the silver cis-up-dihydroxybutyrate. The solubilities a t 17’ were practically the same, being 3.2 and 3.1 ; the solubilities of the two up-dihydroxybutyrates at 15” and 13’ are 1.45 and 5.1 re- spectively (Fittig and Kochs, Zoc. cit.).