2776 GOSS AND INGOLD THE POSSIBLE ENHANCED CCCLXXXI1.-The Possible Enhanced Activity of Newly-formed Molecules. By FRANK ROBERT Goss and CHRISTOPHER KELK INGOLD. IN the course of investigations carried out during the past few years the authors have frequently been unable to confirm an apparently obvious reaction mechanism by preparing the supposed intermediate products and subjecting them to the conditions of the original experiment ; the substances either remained unaltered or behaved differently from expectation. Similar experiences have been recorded by others and the phenomenon appears sdciently widespread to give some support for the suggestion that intermedi-ates may be produced in a reactive condition in which they are capable of changes which cannot occur after the energy associated with the formation of the compound has been dissipated.This hypothesis of " nascent molecules " is of course not new; but it is one which should be accepted only after conclusive evidence has been furnished and whilst it is not claimed that the observations recorded in this paper constitute sufficient grounds for embracing ao far-reaching a hypothesis they appear to possess some sig-nificance from this point of view. Two mechanisms suggest themselves for the formation of the cyclobutane ester (IV) (Markovnikov and Krestovnikov Annalen, 1880 208 334) from a-chloropropionic ester (I) and sodium eth-oxide the elimination of hydrogen chloride is either inter-molecular, in which case 7-chloro-a-methylglutaric ester (11) is the inter-mediate product or intramolecular ethyl acrylate (In) being first formed : $!H2*CHCI*X CH,-CHCI-X X'CHmCH3 (11.) '4 (-i~(-p.x XCHXH (III.) ,H XCH-CH (IV.) X*mmCH (''1 1' \% CH,:CH.X (X=CO,Et) The chloro-ester (11) haa been prepared and subjected fo the It gave no action of sodium ethoxide under the same conditions ACTIVITY OF NEWLY-FORMED MOLECULES.2777 detectable quantity of the cyclobufane eater but on the other hand yielded the cyclopropane ester (V) together with the lactonic ester (VI) the unsaturated eater (VII) and the ethoxy-ester (VIII). H,C<CH(Co@t)'? (vI.) CH(CH,)- CO C02E t*CHMe-CH:CH*CO& t CO,Et*CEKMe*CBj.CE( OEt )*CO,Et (VII.) (VIII.) Acrylic ester (111) was treated in a similar manner. Again no cydobutane ester could be detected but only its unsaturated isomeride a-methyleneglutaric ester (IX) which waa the chief product apart from P-ethoxypropionic ester (X).(IS-) C02EtCH2*CH2*C( :CH2)*CO& t E t OC%42H2*C0&t (X.1 Although these experimenta provide no direct evidence of the intervention either of (11) or of (III) in the series of changes leading to the cyclobutane ester (IV) it may not be inappropriate to suggest fhat if the later change depends on the energy of formation of the intermediate probably it is acrylic ester that acts in this way; for not only are four-membered rings produced on the whole more easily by additive synthesis than by the closure of a chain but also the double linking in acrylic ester is more likely to derive temporary activity from it8 own formation than are the chlorine or methyl-hydrogen atoms of (11) from a reaction in which they are not directly concerned.In this connexion the suggestion may tentatively be advanced that the double bond in acrylic ester is possibly semi-polar when first produced (which it would be if hydrogen and chlorine were removed from chloropropionic ester, not as atoms but as ions) although it is undoubtedly non-polar in the ordinary substance (Sugden Reed and Wilkins this vol., p. 1525). It has been observed (J. 1922,121,1552) that a-campholenic acid (XI) when warmed with an alkaline suspension of silver oxide becomes partly converted into camphor (XIII) and the suggestion. was advanced that reduction to dihydrocampholenic acid (XII) first occurred. This reaction has again been investigated and both cis- and trans-dihydrocampholenic acids have been subjected to similar experimental conditions ; both however remained unaltered.C,HI,-CH2*C02H + C8Hl,DC]&oC02H + C H 1 4 q i 2 (XI.) (XII.) (XIII.) Other case8 have been observed in connexion with the opening of the cyckpropand ring. It has been shown that methoxy 2778 GOSS AND INQOLD TEEE POSSIBLE E " C E D cyclopropanes on treatment with demethylating agents yield first cyclopropanols which may undergo further conversion info open-chain ketones the extent of the latter reaction being limited by its reversibility and the consequent equilibrium which varies from case to case : In the case of the methoxy-ring acid (XIV) the corresponding keto-acid is of course acetosuccinic acid (XV) which if acid is used as demethylating agent passea info laewlic acid with loss of carbon dioxide.It is more remarkable however that when concentrated alkali is employed laevulic acid is still the sole product, although acetosuccinic esters yield mainly succinic and acetic acids, and only a small proportion of laevulic acid under these conditions. On the other hand if alkali of the same strength is used to open the ring in the cycbpropene acid (XV1)-a reaction which in view of the known tendency of glutaconic acids to add on water in the presence of alkalis giving p-hydroxy-acids (e.g. XVII) can scarcely be supposed to proceed ot,herwise than through acetosuccinic acid-only succinic and acetic acids are produced. (XIV.) (XVI.) WhiISt the most the mechanism formulated below has been establiahed probable one for the formation of Balbiano's acid (XXI) from camphoric acid (XVIII) by oxidation Pandya and Thorpe (J.1923 123 2858) synthesised the intermediate hydroxy-ring acid (XX) but could not convert it into Balbiano's acid. Although working with other ends in view they were struck by this and suggested that the real intermediary was the opposite stereo-isomeride to that which they had synthesised; it can however, readily be seen from models that the intervening hydroxy-acid should possess precisely the codguration of the acid which Pandya and Thorpe synthesised and found unreactive. Other similar instances involving the failure of a ring to open have been recorded from time (XVIII.) ax.) fo time-and some of these are now under investigation.HO-$JMe*CO,H HO*CH*C02H + ?Me2 (XIX.) ACTIVITY OF NEWLY-FORMED MOLECULES. 2779 E X P E R I M E N T A L . Action of Sodium E t M on Eayl a-ChlurFo@nate.-Ethyl a-chloropropionate prepared from pure dry lactic acid (Briihl, Ber. 1876 9 35) was treated with sodium ethoxide as deacribed by Markovnikov and Krestovnikov (loc. cit.) and by Haworth and Perkin (J. 1898 73 336). To 20 g. of the ester at SO" dry sodium ethoxide (10 g.) was added in small portions ; the temperature waa kally kept at 100" for 3 minutes and after cooling the product was poured into dilute acid and isolated by extraction with ether. As stated by Markovnikov and Krestovnikov it consisted mainly of ethyl a-ethoxypropionate b.p. 50°/4 mm. This was converted into ifs amide which melted a t 63" after crystallisation from petrol (b. p. 40-40") and was identified wit.h that described by Wurtz (Ann. chim. 1860 59 174) (Found C 50.9 ; H 10.0. Calc., C 51.2; H 9.5%). The remainder (about 10%) of the product, b. g. 120"/4 mm. consisted of the cycbbutane-1 3-dicarboxylic ester which was converted by hydrolysis into the corresponding acid m. p. 170° as described by Markovnikov and Krestovnikov. Action of Sodium Ethoxide on Ethyl u - C ~ r o - y - ~ t h ~ ~ l u ~ r ~ Ethyl a-chloro-y-methylglutarate (this vol. p. 393) waa prepared by treating the lactone of ethyl a-hydroxy-y-methylglutarate with phosphorus pentachloride and pouring into alcohol. The decom-poaition with sodium ethoxide (12 g.and 40 g. of the ester) was carried out as described above and the product was completely hydrolysed. The acids were digested for some days with an aqueous suspension of precipitated calcium carbonate and the filtered solution waa concentrated and allowed to crystallise; the calcium salt of the trans-cycbpropane acid (m. p. 168") then separated. The filtrate from the calcium salts was worked up for organic acids, which were converted through their silver salts into benzyl esters. The portion of these boiling below 240°/15 mm. yielded the original lactone on hydrolysis. By distillation at 1 mm. the leas volatile esters were separated into a small preliminary fraction from which on hydrolysis a-methylglutaconic acid together with a liquid acid (probably the ethoxy-acid described below) was obtained and a fraction (32 g.) b.p. 230-260" which on hydrolysis gave a liquid consisting as analysis indicated mainly of the ethoxy-acid. This waa purified by careful distillation of the ethyl ester b. p. 126-129"/12-14 mm. (Found C 58.1 ; H 9-0. C12HB05 requires C 58.5; H 8.9%); the liquid acid was recovered by hydrolysis (Found C 50.2; H 7-5. C,H,,O requires C 50.5; H 7.3%). Adion of Sodium E t h d e on Ethyl Acqlute.-Ethyl acrylate wit8 prepared from ethyl ap-dibromopropionate by R80hm's method VOL. CXXVII. 5 2780 GOSS AND INGOLD THE POSSIBLE ENHANCED (Ber. 1901 34 573). The reaction was carried out exactly as described in the case of ethyl a-chloropropionate except that for 20 Q. of ester 15 g. of sodium ethoxide were employed.The products were isolated in an identical manner and again consisted to the extent of nine-tenths of a liquid distilling at 50"/4 mm. and one-tenth of a liquid distilling at 120'14 mm. The former waa shown to be ethyl @-ethoxypropionate (compare Purdie and Marshall, J. 1891 59 475) by conversion into the corresponding amide deacribed by Kilpi (2. physikd. Chm. 1912 80 184) which was crystallised from water; m. p. 50" (Found C 50.6; H 10.1. Calc. for C,H,,O,N C 51.2; H 9.5%). The high-boiling fraction consisted of ethyl a-methyleneglutarate and yielded on hydrolpis the corresponding acid m. p. 130" (Rohm Ber. 1901 34 427). a- Campholenic Acid and a-Cumphlenumide.-dl-Campholeno-nitrile b. p. 100"/5 mm. was prepared by the method of Tiemann (Ber.1895 28 2167) from d-camphoroxime and obtained in 75% yield. It was converted into a mixture of the acid and amide by boiling 3 g. with a solution of 5 g. of potassium hydroxide in 25 C.C. of ethyl alcohol. Addition of water precipitated the amide, m. p. 122" and more was obtained by extracting the filtrate. The acid extracted after acidification of the mother-liquor distilled at 137"/1 mm. a-Campblanic Acid.-The cis-acid was prepared by the reduc-tion of a-campholenic acid with hydrogen and platinum black (Lipp Ber. 1922 55 1883). The truns-acid was obtained by reducing a-campholenamide in the same way (Eoc. cit.) and hydrolysing the a-campholanamide h t formed by boiling for 50 hours with 3 g. of potassium hydroxide in 15 C.C. of alcohol. Action of Silver Oxide on a-Camphlenic Acid and a-Campholanic Acid.-a-Campholenic acid and cis- and tram-a-campholanic acids were added in separate experimenfs to a suspension of 2-5 g.of silver oxide in a solution of 0.5 g. of calcium oxide in 25 C.C. of water. In the experiment with a-campholenic acid the silver oxide was rapidly reduced and the amount of camphor deposited in the condenser reached a maximum after 4 hours' heating. The saturated acids however gave no trace of camphor under similar conditions even after several days and the silver oxide remained apparently unaltered. Action of Hydrochloric Acid on 3-Met7wxy-3-methykyclopropane-1 2-dicarboxylic Acid.-The acid prepared as described by Goss, Ingold and Thorpe (J. 1923 123 3358; this vol. p. 468) was boiled for 2 hours with ten times its weight of 20% hydrochloric acid.From the product a quantitative yield of lawulic mid was It was distilled at 140"/1 mm. The acid was distilled a t 141"/1 mm ACTIVITY OF NEWLY-FOBXED MOLECULES. 2781 obtained which was identified by direct comparison and through its semimbazone and phenylhydmzone the m. p’s of which were not depressed by admixture with genuine specimens of the respec-tive substances. The lac tone of 3- hydroxy -3-methylc~propane-1 2-dimboxylic acid was unaffected by this treatment. Action of Potassium Hydmri.de on 3-Mei%xy-3-dykyclo-~zrcrpane-1 2 - d i . c u ~ l i c Acid-The acid ww boiled f p 30 minuta with 64% aqueous potassium hydroxide. From the cooled solution ether extracted lsevulic acid which was identified as described above.The lactone of 3-hydroxy-3-methylcyclopropane- 1 2-dicarboxylic acid wa8 unafiected by the same treatment. Action of Hydrobromic Scid on the Metbxy-acid Fomndon of 3- Methyl- A2-cyclopropene- 1 2-diCarboxyZic Acid.-The methoxy-acid was boiled for 1 hour with 20 paris by weight of concentrated hydrobromic acid and the resulting solution was diluted and extracted with ether methylcyclopropenedicarboxy~c acid being obtained. Action of Potassium Hydroxide on 3-M&yl-A2-cycloprope~-1 2-dicarboxylic Acid.-(1) The ring-acid was boiled for 2 minub with 55% aqueous potassium hydroxide and the solution was cooled strongly acidified and extracted with ether. Unchanged material was recovered. (2) The experiment was repeated the time being increased to 30 minutes. The product consisted of a mixture of the cy&-propene acid and succinic acid which were separated by fractional cryst&Uisation and identified by direct comparison. (3) The ring-acid wit8 boiled for 2 minutes with 64% potassium hydroxide. On working up for acids a mixture of the cycbpropene acid and succinic acid WM obtained. (4) The above experiment wit8 repeated the time being increased to 30 minutes. The solid product consisted solely of succinic acid. Acetic acid also was produced but this wit8 detected by qualitative tests only and was not purified. We desire to thank the Chemical Society for a grant which has defrayed part of the expense of this investigation. THE UNIVERSITY LEEDS. [Received May 2let 1925.