CLARKE AND LAPWORTH : CYANOCARONE. 11I I. - Cy a? LO ccwonc.By REGINALD WILLIAM LANE CLARKE and ARTHUR LAPWORTH.ALTHOUGH carvone and its dihydro-derivative only differ inasmuchas the six-carbon ring in the former contains an et.hylenic linking,the products which their hydrobromides yield by loss of hydrogenbromide under the influence of alkalis are quite different instructure, as carone contains a three-carbon and a six-carbonring, whilst neither nucleus is present in eucarvone, which appearsto contain only one seven-membered ring (Wallach, Annulen, 1905,339, 94, e t sep.).The explanation of the formation of eucarvone which mostnaturally suggests itself is the one indicated by Wallach, namely,that in the first stage of the action of potassium hydroxide oncarvone hydrobromide, halogen hydride is removed in the samemanner as in the case of dihydrocarvone hydrobromide, but thatthe molecule of the resulting compound is less stable than that ofcarone, owing to the additional strain produced by the ethyleniclinking, and consequently the cyczopropane ring is a t once resolved,but between the two carbon atoms which previously formed partof the six-ring of the carvone hydrobromide ; hitherto, however, nodirect evidence confirming this idea has been forthcoming Had i12 CLARKE AND LAPWOKTH : CYANOCARONE.been possible to remove two hydrogen atoms from carone, or toseparate two groups from adjacent carbon atoms in a substitutionproduct of carone, in such a way as to produce an ethylenic linkinga t the position where this is found in the hypothetical intermediatecompound, it would have been possible to ascertain whether this atonce resulted in the formation of eucarvone, and thus to confirmthe view referred to, but no substituted derivatives of carone yetappear to have been obtained either by direct substitution or bypreparing them from substituted dihydrocarvones.The investigation of the action of alkalis on cyanodihydrocarvoneseemed likely to lead to the formation of such a substituted carvone,and since it has been shown by one of us that 8-cyano-derivativesof ketones are frequently convertible into a8-unsaturated ketonesby the action of alkalis in presence of ferrous hydroxide, a possiblemode of attacking this question was promised.After many unsuccessful attempts, pure cyanocarone was obtainedby a method similar to that used in preparing carone from dihydro-carvone, and under certain conditions it was found that cyano-dihydrocarvone hydrobromide might be converted nearly quantita-tively into the new substance, only traces of cyanodihydrocarvonebeing regenerated :A mixture of isodynamic forms is doubtless produced in thefirst instance, but in presence of the alkali, which acts asequilibrator, these, during the process of solidification, change withsuch e a e that only one isomeride remains, the equilibrium mixturebeing saturated with respect to one form.The product, when nearly pure, crystalliss in massive, transparentforms, and has the properties which it might be anticipated asubstance having the above constitution would po_ssess.By theaction of mineral acids, the cyclopropane complex is attacked inall cases before the cyano-group is affected, and with halogenhydrides the first product appears invariably to be the compoundof the acid with cyanodihydrocarvone, the reaction above repre-sented taking place in the reverse sense.The nitrile is saturated in character, and is attacked by coIdpermanganate solution only with great difficulty, but a t the water-bath temperature it is oxidised in alkaline solution, yielding caronicacid, the presence of the dimethylcyclopropane nucleus thus beingconfirmed. With acid permanganate, another acid, apparentlyisomeric with caronic acid, but not yet described, is producedCLARKE AND LAPWORTH : CYANOCARONE.13By the action of alkalis, however, the substance loses the elementsof hydrogen cyanide, and if hot dilute aqueous alkali is used inpresence of ferrous hydroxide to facilitate this reaction, the volatileproduct being allowed to pass away a t once in the steam, eucarvoneis obtained. The product which cyanocarone should normally yieldby the action of alkali is the ap-unsaturated ketone:CMeHCH*CH;CH _.____ -\CMe2or the hypothctical intermediate product in the preparation ofeucarvone from carvone hydrobromide. It follows, therefore, thata t looo, even in presence of quite dilute aqueous alkali, thissubstance is unstable, and is at once converted into eucarvone. ' Asthe cyanocarone certainly contains the cy clopropane ring, whichis not stable to alkali, it seems certain that the presence of theethylenic linking in the six-carbon ring to which the cyclopropanenucleus is attached does in fact render the molecule unstable,and leads mainly to the opening of the three-carbon ring a t thepoint indicated by the dotted line.The matter is of further general interest, too, in contrasting themode in which the cyclopropane nucleus breaks down undervarying conditions.It would not appear reasonable t o suggestthat the complex *CMe,*CH: is less stable than :CH-CH:, as underthe influence of halogen hydride it is the former which is resolved.Nor can it be maintained that either is in such a position withreference t o the keto- or cyano-group as would render it more liableto attack on this account.It would rather appear that when thecarbon ring is saturated there is the less strain when the ring iscomposed of six atoms, bqt when there is an ethylenic linking inthe nucleus, at least in certain positions, then the reverse obtains.and the smaller ring is the less stable one.'420- CH/.EXPERIMENTAL.Formation of Cyanocarone,The hydrobromide of cyanodihydrocarvone, prepared as describedby Lapworth (Trans., 1906, 89, 1826), was rapidly crystallisedfrom alcohol, and treated in the following manner. The hydro-bromide (160 grams) was suspended in methyl alcohol (300 c.c.),cooled to Oo, and to the pasty mixture an ice-cold solution ofpotassium hydroxide (36 grams) in methyl alcohol (150 c.c.) wasadded gradually with frequent agitation. The resulting liqui14 CLARKE AND LAPWORTH : CYANOCAHONE.was kept for some hours, whiIe it gradually assumed a violet colour,potassium bromide being deposited.The whole was then saturatedwith carbon dioxide, and the precipitated potassium bromide andbicarbonate separated by filtration ; the methyl alcohol wasremoved by distillation, and the volatile material expelled with theaid of a rapid current of steam. I n some experiments the oilremaining in the distillation flask solidified on cooling, but it wasfound to be advantageous, as a rule, to shake the semi-solidmaterial for some time with an ice-cold solution of potassiumpermanganate, added gradually until the colour of the latter wasno longer discharged, the excess of permanganate and the pre-cipitated manganese dioxide being subsequently removed by sulphurdioxide. The crude cyanocarone, which solidified on again coolingthe liquid, was collected, and crystallised several times fromalcohol :0.2158 gave 0.589 CO, and 0.166 H,O.C=74*4; H=8*55.0.2206 ,, 0.606 CO, ,, 0.172 H,O. C=74.9; H=8*66.0.1264 ,, 9.0 C.C. N, (moist) at 19O and 757 mm. N=8*15.C,,H,,ON requires C = 74.6 ; H = 8-47 ; N = 7.92 per cent.Cya?zocuTome is very soluble in ethyl and methyl alcohols, ether,acetone, benzene, or ethyl bromide, fairly so in light petroleum,and almost insoluble in water. It crystallises with great readinessfrom its alcoholic solution in large, colourless, six-sided, transparentcrystals, which melt sharply at 54-55O.When strongly heated,cyanocarone boils and distils with some decomposition above 300O.0.201, made up to 25-05 C.C. with absolute alcohol, a t 18O gave,in a 2-dcm. tube, aD+4.79O, whence [a],+298O.0.2306, made up to 25.1 C.C. with absolute alcohol, at ZOO gave,in a 2-dcm. tube, a,? 5.45O, whence [aID + 297O.Cyanocarone is only very slowly attacked by a cold aqueoussolution of potassium permanganate, or by a solution of the samesalt in acetone at the boiling point of the solvent. It does notdecolorise a solution of bromine in glacial acetic acid in presence ofsodium acetate.The seniicarbazide, C11H1SN:NLH*CO*NH2, crystallises fromalcohol in thin, flat, rectangular plates, which melt ratherindefinitely, and decompose slightly a t 218-221O :0.1526 gave 31.6 C.C.N, (moist) at 15O and 758 mm.C,,H,,O", requires N = 23.93 per cent.Cyanocarone also yields an oxime, but this could not be obtainedN=24.18.in crystalline formCLARKE AND LAPWOHTH : CYANOCAEONE. 15Action of Alkali and Ferrous Hydroxide on Cyanocarone.On boiling cyanocarone with a 10 per cent. sodium hydroxidesolution, an oil with a peppermint-like odour is produced, and theaqueous solution gives the reactions of a cyanide. The removalof hydrogen cyanide appears to take place more readily in presenceof ferrous hydroxide, and for the investigation of this decom-position the following conditions were employed. Twenty gramsof cyanocarone, 12 grams of potassium hydroxide, 4 grams offerrous chloride, and 150 C.C.of water were gently heated in aflask attached to a condenser, and the water which distilled over,carrying with i t the odorous oil, was replaced by gradually addingwater to the flask. The process was continued until the aqueousdistillate no longer contained an appreciable quantity of oil. Thedistillate was then saturated with common salt, and the oilextracted with ether. On fractionation, 4.4 grams of liquid boilingbetween 205O and 208O, and 1.1 grams boiling between 208O and215O were obtained, a small amount of residue, which underwentdecomposition on further heating, remaining in the distilling flask.The oil thus obtained readily decolorised a solution of bromine,and gave a reddish-violet colour on boiling with methyl-alcoholicpot'ash. It yielded an easily crystallisable semicarbazide, which,on recrystallisation from alcohol, melted at 183-184O :0.1978 gave 35.5 C.C.N, (moist) at 19O and 751 mm.CllH,,ON, requires N = 20.29 per cent.On mixing this semicarbazide with eucarvone semicarbazide,prepared as described by Wallach and Lohr (Annalen, 1899, 305,237), the mixture melted at 183-184O; the product, after repeatedcrystallisation from methyl and ethyl aicohols, was opticallyinactive. The conversion of cyanocarone into eucarvone by theabove process is not quantitative, and a considerable amount of awhite substance crystallises out of the aqueous residue. This wasisolated by diluting the residual liquid in the flask with water,heating to boiling, and filtering, when, on cooling, the substanceseparated, and was purified by recrystallisation from water, andfinally from alcohol :N=20.38.0.2278 gave 0.5630 CO, and 0.1786 H,O.C = 67.43 ; H = 8.72.0.203 ,, 13.3 C.C. N, (moist) a t 20° and 756 mm. N=7*45.C,,H,,O,N requires C = 67.69 ; H = 8.72 ; N = 7'18 per cent.The substance has the properties of a saturated lactam oranhydramide, it is unaffected by a cold potassium permanganatesolution, or by boiling aqueous or alcoholic potassium hydroxid16 CLARKE AND LAPWORTH : CYANOCARONE.solutions, and is only slowly changed by fusion with potassiumhydroxide and a few drops of water. It crystallises from wateror alcohol in square plates or cubes, melting a t 210-212O.Action of Hydrogen Halides on Cyanocarone.When heated with concentrated hydrochloric acid on the water-bath, cyanocarone yielded an acidic substance, which appeared tobe a mixture of the stereoisomeric dihydrocarvonecarboxylic acids(Trans., 1906, 89, 1823) ; from this after repeated crystallisationfrom carbon tetrachloride and finally ethyl acetate, an unsaturatedacid melting a t 141-142O was obtained, which was identified by theniixed melting-point method as P-dihydrocarvonecarboxylic acid.With a cold saturated solution of hydrogen chloride, cyanocarone isfirst converted into a hydrogen chloride additive product identicalwith that obtained from cyanodihydrocarvone, the cyclopropane ringundergoing fission.This substance on further treatment withhydrochloric acid loses the elements of hydrogen chloride, and the*CN group is converted into the CO-NH, group, the amide of theunsaturated dihydrocarvonecarboxylic acid being formed.Twentygrams of cyanocarone were suspended in 100 C.C. of concentratedhydrochloric acid, and the mixture was saturated with hydrogenchloride in the cold. The cyanocarone dissolved, and after a shorttime a white, crystalline substance separated, which was purifiedby crystallisation from alcohol :0.3035, after being heated with fuming nitric acid and 0.325 ofsilver nitrate, required 4.5 C.C. of 0-ll2N-thiocyanate.C1= 16.5.C,,H,,ONCl requires C1= 16.6 per cent.The substance crystallised from alcohol in flattened, prismaticneedles, melting at 69O, and when mixed with cyanodihydrocarvonehydrochloride its melting point was unaltered.0.402, made up to 25 C.C.with absolute alcohol, a t 1 8 O , gave, ina 2-dcm. tube, a, + 0 ' 8 2 O , whence [a]= + 25*6O. Cyanodihydrocarvonehydrochloride has [aID + 25'3O a t 1 8 O (Trans., 1906, 89, 1826).When cyanocarone is dissolved in a saturated solution ofhydrogen bromide in glacial acetic acid, and kept for some time,a crystalline substance separates, of which a further amount canbe obtained by diluting the acetic acid solution with water; thiswas collected and crystallised from alcohol :0.2964, after being heated with fuming nitric acid and 0.2478of silver nitrate, required 2'68 C.C. of O.112N-thiocyanate.Br = 30.7.C,,H,,ONBr requires Br = 31.0 per centCLARKE AND LAPWORTH : CPANOCARONE.17The substance crystallised from alcohol ia flattened needles,melting a t 85O, and on mixing it with cyanodihydrocarvone hydro-bromide, its melting point was unaltered :0.3546, made up to 25.1 C.C. with absolute alcohol, a t 14O, gave,in a 2-dcm. tube, a, + OS72O, whence [a],+ 25'5O.0.3208 of cyalrodihydrocarvone hydrobromide, made up to 24.9C.C. with absolute alcohol, at 14O, gave, in a 2-dcm. tube, a,+ 0*665O,whence [aJD + 25.8O.Unsaturuted Amide.-Tx'hen cyanocarone is allowed to remainwith saturated aqueous hydrogen chloride for some hours, thehydrochloride a t first formed slowly dissolves. When the liquidno longer gave the reactions of a nitrile, it was diluted with twiceits volume of water, and rendered alkaline with strong ammonia.After cooling, the separated solid was collected and crystallisedseveral times from water:0.2055 gave 0.5120 CO, and 0.1648 H,O.C=68*0; H=8.91,0.2022 ,, 12-8 C.C. N, (moist) a t 1 8 O and 751 mm. hT=7*22.C,,H,,O,N requires C = 67.7 ; H = 8-72 ; N = 7-18 per cent.0.2335, made up to 25 C.C. with absolute alcohol, a t 18*5O, gave,in a 2-dcm. tube, a,+ 1*33O, whence [alD + 71.2O.The amide is readily soluble in hot water or benzene, very solublein alcohol, acetone, chloroform, or ethyl acetate, and sparingly soin cold water o r light petroleum. It crystallises from alcohol orwater in small, flattened, white needles, melting a t 130O.It evolves ammonia when boiled with 10 per cent. aqueoussodium hydroxide, reduces permanganate solution immediately inthe cold, and decolorises a solution of bromine in acetic acid inpresence of sodium acetate.When heated on the water-bath with concentrated hydrochloricacid, the amide was converted into an acidic substance, which, ondilution, was precipitated as an oil; this was collected, and finallyobtained as a solid, which was recrystallised several times fromethyl acetate.It melted at 141-142O, and when mixed withP-dihydrocarvonecarboxylic acid its melting point was unaltered.The amide was therefore in all probability an isomeride of thedihydrocarvonecarboxylic amide previously described (Trans., 1906,89, 958).Oxidation of Cyunocarone.An aqueous solution of potassium permanganate oxidises cyano-carone fairly rapidly when heated with it on the water-bath.Asolution of 140 grams of potassium permanganate in 3500 C.C. ofwater was added gradually to a mixture of 20 grams of cyano-carone and 200 C.C. of water. The liquid was filtered from theVOL. XCVII. 18 CLARKE AND LAPWOnTH : CYAKOCAROKE.precipitated manganese dioxide, evaporated to small bulk, saturatedwith salt, and extracted with ether twice t o remove any unchangedcyanocarone or other neutral material. The liquid was thenacidified with hydrochloric acid, and extracted twelve times withether. The extracted material was freed from volatile material inthe usual manner, and was finally obtained as a dark semi-solidmass. This was purified by triturating it with chloroform, andcrystallising from a relatively small quantity of chloroform, usinga Soxhlet extractor, as it is sparingly soluble in this solvent.Afterit subsequent crystallisation from water, it was dried a t looo:C = 52.98 ; H = 6.28. 0.2065 gave 0.4012 CO, and 0-1168 H20.C7HI0O4 requires C = 53.16 ; H = 6.33 per cent.0.1075 required 13.6 c c. of N/l0-sodium hydroxide a t - 5 O forneutralisation, using phenolphthalein as indicator, twhence theequivalent = 79.6. A dibasic acid, C,H,,O,, requires equivalent = 79.The acid crystallised from water in small, white masses, meltingat 173-174O. The anhydride, prepared from the acid by meansof acetyl chloride, was crystallised from light petroleum, an_d wasfound to melt a t 55O.The properties of the acid prove it t o be identical with the cis-caronic acid prepared by Baeyer and Ipatieff from carone (Ber.,1896, 29, 2796), and synthesised by Perkin and Tliorpe (Trans.,1899, 75, 48).The first experiment on the oxidation of cyanocarone was carriedout with an aqueous solution of potassium perrnanganate containingrather more sulphuric acid than that required to combine with thepotassium hydroxide which is formed during the oxidation. I nthis instance, an acid similar in solubility to caronic acid wasisolated, which melted, however, a t 204O. On titration withsodium hydroxi-de solution, 0.104 required 6.15 C.C. of 0.1075N-alkali for neutralisation, using phenolphthaleiii as indicator,whence equivalent = 157.3. On adding a further 7 C.C. of the alkaliand heating for one hour on the water-bath, the excess of alkalirequired 0.9 C.C. f 0~1003n’-hydrochloric acid for neutralisation,whence the equivalent calculated from the total amount of alkalineutralised = 79.The data indicate that the substance is the lactone of a saturatedhydroxydicarboxylic acid, and probably isomeric with terebic acid.GOLDSMITHS’ COLLEGE, NEW CROSS. ‘r HE UNIVERSITY,M ANCHESTER