34 CHATTAWAY AND RAY: 11.- The Decmyiosition of Tartaric Acid by Heut. By FREDERICK DANIEL CHATTAWAY and FRANCIS EARL RAY. ALTHOUGH the decomposition which tartaric acid undergoes when heated has been studied many times during the past hundred years there is no general agreement as t o the substances produced or as to their relative amounts nor has the course of the decomposition been made clear. Since many of the products described by previous observers are obviously the result of secondary reactions the decomposition has been carried out a t the lowest possible temperature and under diminished pressure. As under these conditions the primary products are removed immediately they are formed any side-reac-tions might be expected t o be avoided or reduced to a minimum.When heated in this way no substances of complicated structure are formed but the tartaric acid decomposes in two stages; in the first water alcne is liberated ; in the second carbon monoxide and carbon dioxide are produced together with formic acetic and pyruvic acids. The reactions which occur during the heating may be explained as follows: The tartaric acid a t first loses one molecule of water and there is produced a solid colourless substance of the nature of a lactide having half the acidity of the parent acid and reconvertible into it by hydrolysis. CO,H*CH (0 11 ) CH (OH) *C'O,H This stage may be formulated thus: -+ CO,H* CH( OH) *yH*yO CO,H*CH (OH) CH( OH) C0,H 2H20+ o i) I I CO-CH*CH (OH)*CO,H This loss of water must be followed a t a somewhat higher tem-perature by an intramolecular rearrangement of the nature of the Beckmann transformation hydrogen and hydroxyl changing places, thus : OH 13 H OH I 1 CO,H*CH-C-CO I I CO,H*CH*C-CO -+ I I 0 0 I I 0 0 I I I I COG-CH* C0,H CO* C-CH* CO,H 1 1 I t H OH OH THE DECOMPOSITlON OF TARTARIC ACID BY HEAT.35 The compound momentarily formed in this manner then breaks down in two ways producing in one decomposition carbon monoxide, carbon dioxide and acetic acid t h u s : Co2H'CH,*C(oH)*Co CO,H.(3H + CO + (lo I 0 b GO + CO + CH,*CO,H 1 I CO-C( OH) *CH,*CO,H and in the other decomposition carbon dioxide and pyruvic acid a ti ansf erence of hydroxyl again taking place thus : Cc),H*CH2*C( 0 E3)*CO I I 0 0 -+ I I UO--U(UH)*CH,*CO,H CH,* C(0H) *CO I.I 2c0 + 0 I I 0 c'o-- C(OH)*CH, The formic acid which is prodnced in sinall amount in the decom-position is probably formed by an intramolecular rearrangement of the undehpdrated acid followed by a rupture of the molecule similar to that which occurs when a tartrate is fused with sodium hydr-oxide the oxalic acid which is one of the primary products of the decomposition subsequently breaking down into carbon dioxi'de and formic acid. C02H C0,H CO,H I I I CH, CH*OH 1 CO,H E X P E R I M E N T A L . The heating was carried out in a bath of fusible metal the bulb of the flask being completely immersed. Reduction of pressure was effected by a water-pump or by a Sprengel pump when the gases were to be examined.Heating was not begun until the pressure had been reduced as far as possible. During the heating a pressure of 10-20 mm was maintained the pressure varying somewhat with the stage of the decomposition and the rapidity of the accompanying evolution of gas. The liquid products were condensed in one or more strongly cooled c 36 CHnTTAWAY AND RAY: receivers whilst the escaping gases after having been passed through cotton-wool were collected over mercury. Tartaric acid when heated in this way melts when the tempera-ture of the bath reaches about 150° and appears to boil vigorously, a clear colourless liquid condensing in the receivers. I f the pressure is kept at 19-15 inm. and the temperature of the bath not allowed to rise above 1 6 5 O the distillate is neutral but it becomes slightly acid if this temperature is exceeded.As the decomposition proceeds, the temperature of the bath being raised a few degrees above 165O, the contents of the flask gradually become more viscid a slight evolution of gas occurs and care is needed to prevent the mass distended by this gas and by water-vapour from frothing over. After some time tjhe frothing ceases liquid no longer condenses, and action a t this temperature appears t o come to an end. T'he contents of the flask if cooled a t this stage have a white spongy appearance somewhat resembling dried bread. When the temperature of the bath is further raised to about 180° action recommences the mass darkens slightly gas is very freely evolved and a clear yellow liquid of a pungent odour distils over.This decomposition continues until the whole mass has dis-appeared a very slight carbonaceous residue only being left behind in the flask. When a low pressure is maintained and the heating is slowly and carefully conducted this small amount of residue is unweighabl; and negligible. The liquid obtained in the first stage of the decomposition is water 1 gram-molecule approximately being given off by each gram-mclecule of tartaric acid. 'T'he colourless spongy solid then remaining is undoubtedly a compound of t.he nature of a lactide. It is brittle absorbs moisture from the air and therefore becomes sticky on exposure. It does not dissolve appreciably in benzene chloroform or ether. It dissolves very sparingly in cold water more freely in hot.When quickly dissolved in cold water it is neutralised by little more than half the amoant of alkali required to neutralise the parent acid. Such a neutralised solution becomes again acid on keeping or heating until finally the amount of alkali required for neutralisa-tion equals that which would have been required to neutralise the tartaric acid from which it was derived. On thus neutralising a large amount with potassium hydroxide and crystallising the salt in eight successive crops each proved to be ordinary neutral potass-ium tartrate,. On boiling it with ethyl alcohol and distilling under reduced pressure ethyl tartrate was obtained. The gas obtained in the second stage of the dewmposition con THE T)ECOMPOSITION OF’ TARTARIC ACID BY REAT. 37 sisted of carbon nionoxide and carbon dioxide roughly in the pro-portion of one volume of monoxide t o two volumes of dioxide.If the first stage of the heating is not carried out very slowly and the teniperature of the bath is allowed t o rise rapidly t o effect ccmplete decomposition more quickly the percentage of carbon dioxide in the evolved gas is considerably increased owing to the greater amount of tartaric acid decomposing without the initial formation of the lactide. The liquid distilling in the second stage of the decomposition is a mixture of formic acetic and pyruvic acids no other substances being produced. By fractionation under diminished pressure the distillate was resolved into its three constituents and each was separately purified and identified.The amount of each acid yielded by a known weight of tartaric acid was estimated the formic acid by Franzen and Greve’s method (J. pr. Chem. 1909 [ii] 80 383) the acetic acid by titration and the pyruvic acid by conversion into its p-bromophenylhydrazone. The average of several concordant esti-mations gave the following as the products obtainable by theadistil-lation of 1 gram-molecule of tartaric acid (150 grams) under diminished pressure and a t a temperature not much exceeding 180° water 18.8; formic acid 2.4; acetic acid 49; pyruvic acid, 14.1; carbon monoxide 21.3; carbon dioxide 43 grams a total less by 1.4 grams than the weight of acid used. The substances termed by Fremy (Annalen 1839 29 142) tar-tralic and tartrelic acids represent stages in the formation of the acid lactide which together with water is the sole product of the first stage of the decomposition. The formation of tarry matters and the subsequent charring which occur when tartaric acid is heated under the ordinary pressure are due to the decomposition of the pyruvic acid first formed and in no respect differ from the changes which take place when this acid is similarly heated. THE QUEEN’S COLLEGE OXFORD. [Received. November 26th 1920.