XXVIII.-Cis- and trans-Methylisopropylsuccinic acid. By WILLIAM HENRY BENTLEY, WILLIAM HENRY PERKIN, jun., and JOCELYN FIELD THORPE.

THE action of the ethereal salts of a-bromo-acids of the fatty series on the sodium compounds of ethylic malonate and its derivatives may take place in two different ways.

1. The action may proceed directly with the simple separation of sodium bromide, thus:

R.CH, CHBr COOC2H, + XCNa(COOC2Hs)2 =

RCH,CH•(COOC,H,)CX(COOC,H,), + NaBr,

2

forming an ethereal salt of a tribasic acid, which, on hydrolysis and subsequent elimination of CO2, yields a derivative of succinic acid, thus:

R.CH2 CH (COOH) CX(COOH), =

R.CH, CH(COOH).CHX.COOH + CO2.

2. The reaction is an indirect one. In this case an unsaturated ethereal salt is first produced by the removal of hydrogen bromide, R.CH CHBr COOC2H, R.CH:CH COOCH, + HBr,

and this unsaturated ethereal salt then condenses with the sodium derivative employed, as follows,

(COOC,H,),CXNa + RCH:CH-COOC,H =

(COỌC,H,),CX CHRCHNa•COOC,H,

yielding a sodium derivative of an ethereal salt from which the corresponding tribasic acid may be isolated by hydrolysis. This tribasic acid then readily decomposes on heating with formation of a derivative of glutaric acid,

(COOH)2CX CHR CH2 COOH = CO2 +

COOH CHX CHR.CH, COOH.

The direction in which the action proceeds depends generally on the conditions of the experiment; thus, when ethylic a-bromisobutyrate is digested with ethylic methyl malonate in alcoholic solution (Bischoff and Mintz, Ber., 1890, 23, 649), an ethereal salt of the formula (COOC2H5)2C(CH3)·CH2•CH(CH2)·COOC2H, is produced, from which, by hydrolysis and elimination of CO2, the two symmetrical dimethylglutaric acids are obtained; whereas if the experiment be performed in xylene solution at 200°, the action proceeds directly with separation of sodium bromide and formation of ethylic trimethylethanetricarboxylate, (COOC2H2)2C(CH3)·C(CH3)2•COOC2H,

from which trimethylsuccinic acid is readily produced (Bischoff, Ber., 1891, 24, 1078; Bredt and Helle, Inaugural Dissert., Bonn, 1893, 31; Auwers, Annalen, 1895, 285, 260 and 301).

During some experiments on the action of ethylic bromomethylisopropylacetate, (CH3)2CH·CBr(CH3) COOC2H,, on the sodium derivative of ethylic malonate, on which one of us has been engaged for some time, it was found to be exceedingly difficult to decide whether the substances obtained were derivatives of succinic, or of glutaric acid,* and, in order to throw some light on this point, a series of experiments on the action of the next lower homologue, namely, ethylic bromisopropylacetate (ethylic a-bromisovalerate) sodium derivative of ethylic methylmalonate were instituted, with the results described in this communication.

on the

If this decomposition proceed as indicated in equation 1, the end product would be methylisopropylsuccinic acid,

(CH3)2CH·CH(COOH)·CH(CH2)·COOH;

if, however, hydrogen bromide were eliminated, and subsequent condensation took place, the end product would be trimethylglutaric acid, COOH CH2C(CH3)2•CH(CH3)·COOH, and it would be easy to distinguish between these substances, since, in the former case, the acid would contain two asymmetric carbon atoms, and be capable, therefore, according to the Le Bel-van't Hoff theory, of existing in two CH, CH COOH distinct inactive modifications, namely, cis, CH, CH COOH' methylisopropylsuccinic acid; whereas the

trans,

C2H, CH.COOH
COOH CH CH,

and

above trimethylglutaric acid, containing only one asymmetric carbon atom, is capable, according to the same theory, of existing in one inactive modification only.

The action of ethylic a-bromisovalerate on the sodium derivative of ethylic methylmalonate was first carried out in boiling xylene solution in the usual manner, the purified ethereal salt obtained was hydrolysed, and the acid produced heated at 200° until all evolution of carbonic anhydride had ceased.

From the product, two well characterised acids melting at 174-175° and 124-125°, were isolated, which, from the study of their behaviour, were clearly shown to be stereoisomeric; these acids, therefore, are evidently the cis- and truns-methylisopropylsuccinic acids, which are formed by the direct action of ethylic a-bromisovalerate on the sodium derivative of ethylic methylmalonate, according to equation 1.

Cis-Methylisopropylsuccinic acid differs from the trans-acid in being

It is hoped that the results of these experiments, which are complete, will be ready for publication shortly.

readily volatile with steam when heated with a 50 per cent. solution of sulphuric acid. It melts at 125-126°, and, when treated with hydrochloric acid, is partially converted into the trans-modification melting at 174-175°. When heated with acetic anhydride, or when distilled, the cis-acid yields a liquid anhydride, and this, on treatment with water, is reconverted into the original acid.

The cis-anilic acid,

(CH3)2CH.CH COOH

CH,CHCONH CH

(?), melts at 160°, and is different from the anilic acid of the trans-acid, although both yield the same anil when heated at 200°.

Trans-methylisopropylsuccinic acid is much less soluble in water than the cis-acid; it melts at 174-175°, and when heated with hydrochloric acid at 180°, is partially converted into the cis-acid.

When distilled under reduced pressure, or heated with acetic anhydride, the trans-acid yields a solid anhydride melting at 46°, which by treatment with water is reconverted into the same acid. If, however, this anhydride be boiled in a reflux apparatus for some time under ordinary pressure, and then distilled, the distillate is found to consist of the cis-anhydride; the conversion of the trans- into the cisacid being complete under these circumstances. The trans-anhydride, on treatment with aniline, gives the anilic acid of the trans-acid, and this, when heated at 200°, yields the anil of the cis-acid.

We next studied the action of ethylic a-bromisovalerate on the sodium derivative of ethylic methylmalonate in alcoholic solution, and in this case again, curiously enough, working up the product in the way described in the body of the paper, we were only able to isolate cis-methylisopropylsuccinic acid, the trans-modification which should have been formed being apparently converted into the cis-acid under the conditions of hydrolysis employed in this particular instance.

Lastly, in order that there might be no doubt as to the constitution of these acids, we have prepared them in the following way.

In the first place, the sodium derivative of ethylic malonate was digested with ethylic a-bromisovalerate, when a good yield of an ethereal salt was obtained, which has already been described by Roser (Annalen, 1883, 220, 277), and which is undoubtedly ethylic isopropylethanetricarboxylate, (CH,),CH·CH(COOC2H2)·CH(COOC2H2). In order to be certain of the constitution of this ethereal salt, it was hydrolysed, and the tribasic acid formed heated at 200°, whereby it was converted into isopropylsuccinic acid,

(CH3)2CH·CH(COOH)·CH, COOH,

a result which confirms Roser's experiments.

The ethylic isopropylethanetricarboxylate was now treated with sodium and methylic iodide in alcoholic solution,

(COOEt),CNa CH(COOEt)·CH(CH3)2 + CH‚I =

(COOEt),C(CH3) CH(COOEt) CH(CH3), + NaI. The ethereal salt thus formed was identical with that obtained in the previous experiments, since, on hydrolysis and subsequent elimination of carbon dioxide, a mixture of acids was obtained from which cis- and trans-methylisopropylsuccinic acid, melting at 174175° and 124-125° respectively, were readily isolated. These acids differed in no respect from those produced in the manner previously described.

The latter method gives by far the best yield of these acids; it is therefore placed first in this paper, and described in most detail.

In preparing this substance, 23 grams of sodium were dissolved in 250

grams of absolute alcohol and 160 grams of ethylic malonate alded, when the sodium derivative separated as a white, gelatinous precipitate. The flask containing the mixture was then connected with a reflux condenser, heated on the water bath, and 209 grams of ethylic a-bromisovalerate added in small portions at a time to the boiling solution. The action was not violent, although sodium bromide separated immediately on adding the bromisovalerate; the boiling was continued for three hours, after which the product was nentral. The alcohol was, as far as possible, distilled off, this being most quickly and completely effected by placing the flask in the boiling water bath, bumping being prevented by suspending a piece of string from the neck of the flask, so as to hang in the boiling liquid.

When the alcohol had ceased to come over, the residue in the flask was mixed with water, the oil separated, the aqueous liquor extracted four times with ether, the ethereal solution dried over calcium chloride, and the ether distilled off. The oily product was then fractionated under reduced pressure (37 mm.); a small portion boiled below 200°, but the chief portion came over between 180° and 182°. The weight of this fraction was 186 grams, or 66 per cent. of the theoretical yield of ethylic isopropylethanetricarboxylate.

0-1510 gave 0·3209 CO, and 0·1155 H2O. C = 57′96; H = 8·49. CHO requires C = 58′33; H = 8:33 per cent.

In order to prove that this ethereal salt has the constitution represented by the formula at the head of this section, 20 grams of the

pure substance were hydrolysed by boiling with an alcoholic solution of potash for two hours, evaporating the product three times with water, acidifying, and extracting several times with ether; after drying over calcium chloride and evaporating the ether, the oily acid was heated at 200° until the evolution of gas had ceased, and the product, which rapidly solidified and melted indefinitely at 90-110°, was recrystallised from hydrochloric acid until the melting point became constant at 116-117°.

0.2297 gave 0-4408 CO2 and 0·1555 H2O. C = 5233; H = 7·53. CH12O, requires C 5252; H = 7.52 per cent.

A careful examination of this acid proved conclusively that it is identical with the isopropylsuccinic acid obtained by Hlasiwetz and Grabowski (Annalen, 1868, 145, 207), from the fusion of camphoric acid with potash, and which Roser (Annalen, 1883, 220, 272) has shown to be identical with the acid obtained by the hydrolysis of ethylic isopropylacetosuccinate with potash,

C,H,C(C,H,O)(COOC,H,)-CH,COOCH + 3KOH =

C2H, CH(COOK)·CH2COOK + 2С2H ̧·OH + CH, COOK. Not only do the melting points of these acids coincide, but on comparing the anilic acid produced from the acid obtained by us with that prepared from a sample of isopropylsuccinic acid, which had been obtained by fusing camphoric acid with potash, they both melted at 145°, and were identical in all respects.

This anilic acid, which does not appear to have been previously prepared, is readily obtained by mixing isopropylsuccinic anhydride with aniline in benzene solution; it crystallises from a mixture of light petroleum and ethylic acetate in large, glistening plates melting at 145°. A nitrogen determination gave the following figures. 0.1991 gave 99 c.c. moist nitrogen at 12° and 760 mm. N = 6·00. C3H7NO3 requires N = 5.96 per cent.

17

On heating this anilic acid in a sulphuric acid bath at 200° for half an hour, it loses water, and is converted into the anil.

This compound crystallises from light petroleum (80-100°) in microscopic needles melting at 213°. The same substance was obtained both from the anilic acid prepared from the synthetically produced isopropylsuccinic acid, and from that prepared from the isopropylsuccinic acid from camphoric acid.

Ethylic Isopropylmethylethanetricarboxylate,

This was prepared as follows. 133 grams of sodium were dis. solved in 133 grams of absolute alcohol, and 166 grams of ethylic