166 PRATT AND ROBINSON A SYNTHESIS OF XXV1.-A Synthesis of Pyrylium Salts of Anthocyanidin Type. Part V. The Xynthesis of Cyanidin Chloride and of Delphinidin Chloride. By DAVID Do~a PRATT and ROBERT ROBINSON. IN Part I1 (J. 1923 123 750) it was shown that w-methoxy-acetoveratrone (I) can be obtained by the methylation of w-hydroxy-acetoveratrone by means of silver oxide and methyl iodide. The oily product yielded isomeric semicarbazones C1,H,,04N3 (a) m. p. 205" and ( b ) m. p. 168" whilst condensation with p-resorcylaldehyde in presence of hydrogen chloride gave rise to a pyrylium salt characterised as the ferrichloride C18H1,0,CI,Fe m. p. 156". The unsatisfactory character of the process is emphasised in the memoir quoted and as a preliminary to an attempt to synthesise cyanidin chloride it was clearly essential to work out an entirely novel method of prepara1;ion of o-methoxyacetoveratrone.We desired to proceed from veratric acid in order that the process might subsequently be applied to the trimethyl ether of gallic acid and so lead t o a synthesis of delphinidin salts. After unsuccessful trials in other directions e.g. in attempting the decomposition of diazoacetophenones by methyl alcohol the solution of the problem was found in an application of the synthesis of acetophenone from benzoyl chloride by way of ethyl benzoylacetate. The action of sodium on ethyl methoxyacetate leads to the sodio-derivative of ethyl ay-dimethoxyacetoacetate MeO*CH,-CO*CH(OMe)*CO,Et, which readily yields C-aroyl derivatives. Ethyl q-dimethoxy- a-benzoylacetoacetate was decomposed by boiling dilute sulphuric acid with formation of phenylglyoxal in poor yield but when it was subjected to the prolonged action of cold dilute potassium hydroxide and the mixture was subsequently boiled w-methoxy-acetophenone (Part 11 p.748) was obtained in accordance with the scheme : PYRYLIUN SALTS OF ANTHOCYASIDIN TYPE. PART v. 167 FO*CH,*OMe - + Ph*CO*CH,*OMe + MeO*CH,*CO,K Ph*CO*C( OMe)*CO,Et +3K0H + K,CO + EtOH. I n a similar manner w 4-dimethoxyacetophenone (Part 11, p. 750) was derived from anisic acid and the product of the inter-action of ethyl sodio-q-dimethoxyacetoacetate and veratroyl chloride in dry ethereal solution yielded w-methoxyacetoveratrone on careful hydrolysis with alkali. The analogous preparation of w 3 4 5-tetramethoxyacetophenone (11) offered no difficulty.Pure w-methoxyacetoveratrone is a crystalline substance which gives a single semicarba,zone m. p. 178" and this is identical with the previously described verstrylmethoxyacetaldehydesemicarb-azone-b the melting point of which was slowly raised to 178" by repeated crystallisation. Veratrylrnethoxyacetddehydesemicarb-azone-a m. p. 205" was the more sparingly solnble of the isomerides and its melting point is not raised on recrystallisation. This sub-stance is plainly the semicarbazone of a 3 4-trirnethoxyphenyl-acetaldehyde (111) and the formation of the latter in the methyl-ation of veratroylcarbinol is noteworthy. MeO/\CO*CH,*OMe MeO/\CO*CH,*OMe MeO/'\CH( OMe)*CI-f (3 MeO!, I MeO!,,! MeOl\) Me0 (1.) (11.) (III.) w-Methoxyacetoveratrone and w 3 4 5tetramethoxyaceto-phenone are much more soluble in cold than in hot water a pro-perty which doubtless indicates the formation and decomposition of hydrates.The former ketone on condensation with P-resorcyl-aldehyde in presence of hydrogen chloride yields trimethylfisetinidin chloride (IV) and the derived ferrichloride m. p. 156" is identical with the substance obtained as mentioned above. 2-Hydroxy-4 6-dimethoxybenzaldehyde and W-methoxyacetoveratrone con-dense under the agency of hydrogen chloride in ethereal solution with formation of cyanidin chloride pentamethyl ether (V) and this salt is demethylated by means of boiling hydriodic acid in presence of phenol with formation of cyanidin iodide.Cyanidin chloride (VI) was prepared from the iodide in the usual manner and the product exhibited in every detail the properties of cyanidin chloride of natuml origin as described by Willstatter and Everest (Annulen 1913 401 189) and Willstiitter and Nolan (ibid. 1915, 408 13). A direct comparison of the synthetical material with an authentic specimen disclosed no dif€erences existing between them. The production of cyanidin chloride by the reduction of quercetin in acid solution (Willstatter and Mallison Sitxungsber. Preu-ss 16s PR-4TT AND ROBINSON A SYNTHESIS OF Akad. Wiss. 1914 769) constitutes in one sense a synthesis of the substance and furthermore Willstatter and Kindler in 1919 made 8 communication to the Munchner Chemische Gesellschaft con-cerning experiments which they had conducted on the application of the Willstatter-Zechmeister pelargonidin synthesis to the case of cyanidin (‘‘ Wissenschaftliche E’orschungsberichte 111 Organische Chemie,” R.Pummerer 2nd Edition 1923 150). The details of the results obtained are not yet available.” c1 n 0 OMe c1 0 F-(VII.) The synthesis of delphinidin chloride (VII) followed the lines of that of cyanidin chloride delphinidin chloride hexamethyl ether being obtained from hydroxydimethoxybenzaldehyde and o 3 4 5-tetramethoxyacetophenone (11). As shown in the ex-perimental portion on page 174 there can be no doubt as to the identity of the synthetical product obtained on demethylation, with that of natural origin. E Y P E R I M E N T A L .Ethgl uy-Dimethoxyacetoccetate.-Ethyl chloroacetate (I22 9.) was gradually added with cooling and shaking to a solution of sodium methoxide (from 23 g. of sodium) in anhydrous methyl alcohol (200 c.c.). The mixture was refluxed until neutral most of the alcohol then removed a.nd the residue mixed with brine and ether. The dried ethereal solution on fractionation yielded 85 g. of ethyl methoxyacetste b. p. 1 3 1 O . j - Sodium (8 g.) cut in thin slices, * Willsttitter Zechmeister and Kindler have now published their work (Ber. 1924 57 [B] 1938). 5 This substance can also be obtained by the hydrolysis of methoxy-acetonitrile by Pinner’s method. Ethyl ethoxyacetate (Henry? Ber. 1871, 4 706) ; ethyl ay-diethoxyacetoacetate (Conrad Ber. 1878 11 58 ; Erlen-bach Annalen 1892 269 28).In order t o avoid the possibility of replace PYRYLIUM SALTS OF ANTHOCYAXIDIN TYPE. PART V. 169 was added in one portion to ethyl methoxyacetate (100 g.) ; the reaction which soon commenced was controlled by cooling in melting ice when necessary. The thick syrup which resulted was treated with 50% acetic acid (45 c.c.) and saturated brine and extracted with ether. The ethereal layer was separated washed with aqueous sodium cerbonate dried with anhydrous sodium sulphate and distilled; 40 g. of a colourless oil b. p. 130"/15 mm., were obtained (Found C = 50.4; H = 7.4. C,H,,O requires C = 50.5; €1 = 7.4%). This ester is moderately readily soluble in water and its alcoholic solution gives a violet coloration with ferric chloride the shade being bluer than that developed by ethyl acetoacetate under similar conditions.w 4-DimetAox~aceto~henone.-In the first place experiments on the hydrolysis of the condensation product from ethyl sodio-ay-dimethoxyacetoacetate and benzoyl chloride were made. After boiling with dilute sulphuric acid and distillation in a current of steam a yellow distillate was obtained and this by suitable treat-ment yielded phenylglyoxaldiphenylhydrazone m. p. 152". When, however the product was boiled with dilute aqueous sodium hydroxide o-methoxyacetophenone was obtained in small yield and identified as the semicarbazona m. p. 85". w 4-Dimethoxy-acetophenone may be obtained in the following manner Sodium (2.7 g.) was granulated under toluene washed with ether suspended in anhydrous ether (150 c.c.) and ethyl ay-dimethoxyacetoacetate (22.3 g.) gradually added.A sodio-derivative separated from the solution and when the sodium had disappeared anisoyl chloride (20 g.) was added in one portion. A gentle reaction occurred and after 1 hour the process was completed by heating on the steam-bath under reflux during 4 hours. Next day water was added the ethereal layer dried over sodium sulphate and after removal of the solvent 35 g. of a pale yellow oil remained. The whole wag agitated for 12 hours with 2.5% aqueous potassium hydroxide (500 c.c.) and the liquid then boiled for 3 hours cooled nearly saturated with potassium carbonate and extracted with ether ; 7.5 g. of the ketone m. p. 40° were obtained. On condensation with 6-aminopiperonal in alcoholic solution by means of potassium hydroxide o 4-dimethoxyacetophenone yields ment of the ethyl group of the esters by methyl and the production of mixtures Mr.N. L. Matthews has employed methyl chloroacetate and condensed this with sodium methoxide in dry methyl-alcoholic solution in the cold. Methyl chloroacetate (108 g.) yielded methyl methoxyacetate (70 g.) b. p. 129'1754 mm. Methyl methoxyacetate (100 g.) reacted with sodium (9 g.) to give a product from which methyl ay-dimethoxyxetoacetata b. p. 129'/17 mm. could be isolated (Found C = 47.5; H = 7.1. C,H,,O requires C = 47.7; R = 6.8%). The yield was 40-50% of that required by theory. G 170 PRATT AND ROBINSON A SYNTHESIS OF 3-methoxy-6 7-methylenedioxy-2-(4-methoxyphenyl)quinoline which crystallises from methyl alcohol in colourless laminze m.p. 152". o-Metlho~yucetoveratrone (I).-Ethyl ay-dimethoxyacetoacetate (36.2 g.) was gradually added to finely granulated sodium (4.4 g.) suspended in anhydrous ether (225 c.c.). Formation of the sodio-derivative appeared to be complete after an hour and a solution of distilled veratroyl chloride (38 g.) in ether (150 c.c.) was intro-duced causing a mild reaction. The mixture was gently heated on the steam-bath for 2 hours allowed t o remain over-night and heated during a further 2 hours. After washing with water drying, and evaporating the ether there remained 63 g. of a pale orange, viscous liquid which could not be crystallised. The whole product was hydrolysed by vigorous agitation €or 9 hours with cold 2.5% aqueous potassium hydroxide (800 c.c.) and finally by boiling the clear solution for 5 hours.The ketone was isolated by extraction with ether after the addition of much potassium carbona$e and purified by distillation; 22 g. of an oil b. p. 190"/15 mm. were obtained which solidified. The substance crystallises from light petroleum containing a little benzene in colourless prisms m. p. 62" (Found C = 63.1 ; H = 6.8. C,,H,,O requires C = 62-9; H = 607%). When a little of this substance is melted under water and the mixture cooled a clear solution is obtained from which the ketone may be precipitated as an oil either by heating or by the addition of an alkali carbonate or hydroxide. The substance is very readily soluble in organic solvents with the exception of light petroleum.On condensation with 6-aminopiperonal it yields 3 - methoxy - 6 7 - msthylenedioxy-2 -(3 4-dimethoxypheiayl)quinoline, which crystallises from ethyl alcohol in slender needles m. p. 155", and exhibits a violet fluorescence in alcoholic solution. Fisetinidin Chloride Trimethyl Ether (IV) . A slow stream of hydrogen chloride was passed through an ice-cold solution of p-resorcylaldehyde (2 g . ) and o-methoxyaceto-veratrone (3 g . ) in dry ether (40 c.c.) when the liquid became yellow and then red and crystalbation commenced after 20 minutes. The passage of the gas was discontinued after 30 minutes and after several hours the pyrylium salt which exhibited an intense green lustre was collected washed with ether and dried (3-4 g.).The ethereal mother-liquors were washed with dilute hydrochloric acid and a ferrichloride (0.2 g . ) was precipitated. This derivative crystallises from acetic acid in reddish-brown needles m. p. 156", and its properties are identical with those of the ferrichloride previously prepared from the product of methylation of veratroyl-carbinol (loc. cit. ). Fisetinidin chloride trimethyl ether dissolves i PYRYLIUM S,4LTS O F SNTHOCYANIDIX TYPE. PART V. 17 1 hot 70/6 hydrochloric acid to an orange-red solut'ion and separate.3 completely on cooling in reddish- brown prismatic needles exhibiting a green reflex (Found in material dried in a vacuum C = 56.1 : H = 3.3. This salt darkens a t 135" and decomposes a t 188-189". It is reddish-violet by transmitted light and makes a violet smcar on paper.1 ts alcoholic solutions are intensely reddish-violet and the orange solution in concentrated sulphuric acid exhibits a green fluorescence. Sparingly soluble in cold water it dissolves on warming to an orange-red solution which on dilution is rapidly decolorised with formation of the $-base. The violet colour-base is precipitated when sodium acetate is added to a moderately concentrated aqueous solution of the salt hut if a dilute solution is treated with sodium acetate the appearance of the violet colour is transient and the +-base is obtained. isoAmyl alcohol extracts the salt completely from an acid aqueous solution and the addition of sodium acetate to the red alcoholic extract gives a violet solution the colour of which fades.The oxonium salt is regenerated from solutions of the colour-base and $-base by the addition of hydrochloric acid. Cl,K1,O,Cl,BH,O requires C = 56-2 ; H = 5.5%). Cyaizidiii Chloride Pentamethyl Ether (V). I n order t o obtain good results in this preparation it is very necessary that the materials employed should be pure. Hydrogen chloride was passed for 1 hour through a solution of 2-hydroxy-4 5-dimethouybenzaldehyde (4-5 g.) and w-methosyacetoveratrone (5-2 g.) in ether (50 c.c.). The solution quickly became crimson and crystallisation of the product was induced by scratching after 30 minutes from the commencement. The tube containing the mixture and a dish containing solid potassium hydroxide were together covered by a bell-jar and after several hours the salt was collected washed with ether and dried (5.5 g.).The salt crystal-lises from alcohol containing a little hydrogen chloride in small, red necdles m. p. 152" which in mass exhibit an old-gold sheen, but give a chocolate-brown smear on paper (Found in material dried in a vacuum over phosphoric oxide and potassium hydroxide : C = 51.7 ; H = 5.9 ; C1 = 14.9. C,,,H,,B,Cl,HCl,2H20 requires C = 51.6; H = 5.6; C1 = 15.2%). The same dtchloride crystal-lised from a mixture of alcohol and concentrated aqueous hydro-chloric acid in slender needles m. p. 152" (Found C = 51.5; H = 5.5; C1 = 15.4%). From 776 aqueous hydrochloric acid the salt separates in long orange-brown needles which form a crystalline crust exhibiting a green reflex (Found in material dried in a vacuum C = 53-2 ; H = 6.0.C,,H,,O,C1,3H,O requires C = 53-7 ; H = 6.00/,). The substance is readily soluble in the simple c," 172 PRATT AND ROBINSON A SYNTHESIS OF alcohols to reddish-violet solutions the colour of which fades on dilution with water and can then be restored by the addition of a mineral acid. The ferrichloride crystallises from acetic acid in which it is very sparingly soluble in reddish-brown prismatic needles m. p. 196200' (decomp.) exhibiting a bronze lustre. The chloride is insoluble in chloroform but this derivative is moderately readily soluble and when the hot saturated solution is rapidly concentrated to one-quarter of its bulk the ferrichloride separates in microscopic needles exhibiting an intense green lustre.Cyanidin Chloride (VI). The foregoing pentamethyl ether (3.2 g.) and phenol (15 g.) were added to hydriodic acid (180 c.c.; d 1.7) and the mixture was boiled for 30 minutes in an atmosphere of carbon dioxide. Water (2 vols.) and much ether were added to the cooled solution and on standing glistening needles with a bright green reflex separated. The substance (1.8 g.) had properties identical with those of cyanidin iodide prepared by the demethylation of paeonidin chloride of natural origin (Willstatter and Nolan AnnaZen 1915 408 136). The iodide (3 g.) was thoroughly ground with aqueous sodium acetate the bluish-violet colour-base separated thoroughly washed, and heated to boiling with 3% aqueous hydrochloric acid (100 c.c.) in presence of a trace of silver and filtered 50 C.C.more 3% hydrochloric acid being used for washing. Concentrated hydro-chloric acid (50 c.c.) was added and on standing the dark red solution deposited the salt completely in masses of short reddish-brown prisms that had a green lustre (Found C = 52.5 ; H = 4.1. C1,HllO,C1,H,O requires C = 52.8; H = 343%). This product showed all the characteristic reactions of cyanidin chloride and a direct comparison with a specimen of the substance of natural origin disclosed no difference in properties of any kind. The behaviour on heating appearance of the crystals under the micro scope formation of colour-base and +-base solubilities in aqueous and alcoholic acid solutions colour of acid solutions absorption spectrum of the solution in ethyl alcohol and the reactions with ferric chloride sodium carbonate sodium acetate potassium acetate in alcoholic solution lead acetate and Fehling's solution were all examined.Willstatter and Everest (Zoc. cit.) state that cyanidin on oxidation with hydrogen peroxide yields a yellow crystalline product closely resembling a flavonol colouring matter. There is, however no clear evidence that quercetin can be obtained by the oxidation of cyanidin and furthermore it is remarkable that no chromone or chromonol derivative has yet been prepared by oxidising a benzopyrylium salt. We have made numerous experi PYRYLIUM SALTS OY ANTHOCYANIDIN TYPE. PART v. 173 ments with the object of achieving this transformation but any definite products isolated have been either carboxylic acids or cournarin derivatives.This investigation proceeds. w 3 4 5-Tetramethxyacetophenone (II).-Sodium (2.5 g.), granulated under toluene and washed with ether was suspended in ether (120 c.c.) and ethyl ccy-dimethoxyacetoacetate (20 8.) slowly added. When the formation of the sodio-derivative appeared to be complete an ethereal solution of trimethylgalloyl chloride (24.5 g.) was added in one portion and the mixture heated gently on the steam-bath for 8 hours. The product was isolated (34 9.) and shaken for 12 hours with 2.5% aqueous potassium hydroxide. The solution was boiled for 4 hours and the ketone isolated as in the case of w-methoxyacetoveratrone. The yield was 9 g. of a colourless oil b. p. 212"/15 mm. which solidified in contact with light petroleum.The substance crystallises from benzene-light petroleum in colourless needles m. p. 54" and closely resembles w-methoxyacetoveratrone (Found C = 60.0; H = 6.3. C,,H,,O, requires C = 60-0; H = 6.6%). The semicarbaxone crystalliser from aqueous alcohol in long colourless needles m. p. 158". 3-Metltoxy-6 ?-methylenedioxy-2-(3 4 5-trimethoxyphenyl)quinoE-ine crystallises from ethyl alcohol in colourless needles m. p. 159", and exhibits a violet fluorescence in alcoholic solution. Its salts, e.g. the sparingly soluble hydrochloride are yellow. The metho-svlphute is orange-yellow and it is evident from a comparison of this substance with the intensely coloured delphinidin chloride hexamethyl ether that the quinolinium and benzopyryhm nuclei function quite differently as chromophores.It may well be that the methoxyl groups are able to decentralise the cationic valency of the pyrylium nucleus and so produce changes of orbits of electrons owing to recurrent redistribution of the charge whilst on the other hand the powerfully basic quinolinium nucleus is able to ignore the claims of the weak auxochromes. We are thus led to anticipate that the aminophenylquinolinium salts will resemble tinctorially the related hydroxyphenylbenzopyrylium salts and it is hoped that an experimental test of this point will be made. Delphinidin Chloride Hexamethgl Ether (corresponding with VII). Hydrogen chloride was passed through an ice-cold solution of w 3 4 5-tetramethoxyacetophenone (4 g.) and 2-hydroxy-4 6-dimethoxybenzaldehyde (3 g.) in dry ether (50 c.c.) crystallisation of the product in crimson needles and prisms having a dark green reflex occurring after about 1 hour.Next day the salt (4 g . ) was collected and a further quantity (0.7 g.) was obtained from the ethereal mother-liquor. The substance crystallises from alcoho 174 A SYNTHESIS OF PYRYLIUM SALTS OF ANTHOCYANIDIN TYPE. containing a trace of hydrogen chloride in long red needles m. p. 163-164" (Found in material dried in a vacuum C = 58.5 ; H = 5.8. C21H,30,CI,0-5H20 requires C = 58.4; H = 5.6%). The substance exhibits a green lustre and is reddish-violet by transmitted light. It is readily soluble in the simple alcohols and moderately soluble in chloroform to a reddish-violet solution. The dark red colour of the solution in water rapidly fades and an almost colourless $-base is precipitated.Faintly acid solutions are also decolorised on great dilution but addition of acid restores the colour of the pyrylium salt. The ferrichloride crystalhes from glacial acetic acid in brownish-red needles which have a dark green appearance in mass. This derivative m. p. 169-170" is readily soluble in acetone chloroform or the simple alcohols. The sulphate crystallises from dilute sulphuric acid in slender red needles with green reflex m. p. 225" (decomp.). Delphinidin Chloride (VII). Delphinidin chloride hexamethyl ether (3-0 g.) was demethylated by means of a boiling mixture of hydriodic acid (180 c.c.; d 1.7) and phenol (15 g.) during 30 minutes a slow current of carbon dioxide being passed through the containing vessel.The cooled liquid deposited clusters of long pointed brown needles together with some squat prisms and these were collected washed with ether and dried (2.0 g.). The iodide was converted into chloride by treatment a t 60" with silver chloride in alcoholic solution in a silver-mirrored vessel. The filtered solution was mixed with an equal volume of concentrated hydrochloric acid and the amorphous precipitate so obtained allowed to remain when it slowly became resolved into a mass of microscopic needles. The substance was isolated and crystallised by solution in 5% hydrochloric acid and addition of 30% acid to the filtered solution so as to make the concentration of hydrochloric acid about 25%. The brown, crystalline mass was dried in a vacuum (Found C = 49.4; H = 4-0.C1,Hl10,Cl,1~5H20 requires C = 49.4; H = 3-8 ; H20 = 7.4y0). According to Willstatter and Weil (Annalen 1916 412, 178) delphinidin chloride crystallises from hydrochloric acid of 20% or greater concentrations with 1.5H20 which is completely lost in a vacuum. Our product did not behave in this way but the circumstance was an accident since when the material was powdered and again exposed to a vacuum it lost 7.6%. The anhydrous salt does not melt below 350". The hydrates containing 1H20 2H20 and 4H20 have been prepared by the methods devised by Willstatter and Weil in order to observe the crystalline forms, which were found to agree with the descriptions of these authors SYNTHESIS OF CERTAIN HIGHER ALIPHATIC COMPOUNDS.PART I. 175 The absorption spectrum of an alcoholic solution of the chloride, the formation of the +base the behaviour of aqueous and acid solutions on shaking with ether ethyl acetate and isoamyl alcohol, the reactions with ferric chloride in aqueous and alcoholic solutions and with sodium carbonate sodium bisulphite lead acetate and Fehling’s solution were all examined with results identical with those described in the case of delphinidin chloride of natural origin by Willstatter and Mieg (AnnaZen 1915 408 61) and Willstatter and Weil (Zoc. cit.). We were not able to make a direct comparison, but the properties of delphinidin chloride are so characteristic and have been recorded in such detail that we can entertain no doubt as to the identity of the synthetical product with that obtained by the hydrolysis of the pigments of the wild purple larkspur and the blue-black pansy.One of us (D. D. P.) desires to thank the Carnegie Trust for t3he Universities of Scotland for a Fellowship which has enabled him to take part in this investigation. THE UNIVERSITY MANCHESTER. [Received November 3 ~ d 1924. 166 PRATT AND ROBINSON A SYNTHESIS OF XXV1.-A Synthesis of Pyrylium Salts of Anthocyanidin Type. Part V. The Xynthesis of Cyanidin Chloride and of Delphinidin Chloride. By DAVID Do~a PRATT and ROBERT ROBINSON. IN Part I1 (J. 1923 123 750) it was shown that w-methoxy-acetoveratrone (I) can be obtained by the methylation of w-hydroxy-acetoveratrone by means of silver oxide and methyl iodide. The oily product yielded isomeric semicarbazones C1,H,,04N3 (a) m.p. 205" and ( b ) m. p. 168" whilst condensation with p-resorcylaldehyde in presence of hydrogen chloride gave rise to a pyrylium salt characterised as the ferrichloride C18H1,0,CI,Fe m. p. 156". The unsatisfactory character of the process is emphasised in the memoir quoted and as a preliminary to an attempt to synthesise cyanidin chloride it was clearly essential to work out an entirely novel method of prepara1;ion of o-methoxyacetoveratrone. We desired to proceed from veratric acid in order that the process might subsequently be applied to the trimethyl ether of gallic acid and so lead t o a synthesis of delphinidin salts. After unsuccessful trials in other directions e.g. in attempting the decomposition of diazoacetophenones by methyl alcohol the solution of the problem was found in an application of the synthesis of acetophenone from benzoyl chloride by way of ethyl benzoylacetate.The action of sodium on ethyl methoxyacetate leads to the sodio-derivative of ethyl ay-dimethoxyacetoacetate MeO*CH,-CO*CH(OMe)*CO,Et, which readily yields C-aroyl derivatives. Ethyl q-dimethoxy- a-benzoylacetoacetate was decomposed by boiling dilute sulphuric acid with formation of phenylglyoxal in poor yield but when it was subjected to the prolonged action of cold dilute potassium hydroxide and the mixture was subsequently boiled w-methoxy-acetophenone (Part 11 p. 748) was obtained in accordance with the scheme : PYRYLIUN SALTS OF ANTHOCYASIDIN TYPE. PART v. 167 FO*CH,*OMe - + Ph*CO*CH,*OMe + MeO*CH,*CO,K Ph*CO*C( OMe)*CO,Et +3K0H + K,CO + EtOH.I n a similar manner w 4-dimethoxyacetophenone (Part 11, p. 750) was derived from anisic acid and the product of the inter-action of ethyl sodio-q-dimethoxyacetoacetate and veratroyl chloride in dry ethereal solution yielded w-methoxyacetoveratrone on careful hydrolysis with alkali. The analogous preparation of w 3 4 5-tetramethoxyacetophenone (11) offered no difficulty. Pure w-methoxyacetoveratrone is a crystalline substance which gives a single semicarba,zone m. p. 178" and this is identical with the previously described verstrylmethoxyacetaldehydesemicarb-azone-b the melting point of which was slowly raised to 178" by repeated crystallisation. Veratrylrnethoxyacetddehydesemicarb-azone-a m.p. 205" was the more sparingly solnble of the isomerides and its melting point is not raised on recrystallisation. This sub-stance is plainly the semicarbazone of a 3 4-trirnethoxyphenyl-acetaldehyde (111) and the formation of the latter in the methyl-ation of veratroylcarbinol is noteworthy. MeO/\CO*CH,*OMe MeO/\CO*CH,*OMe MeO/'\CH( OMe)*CI-f (3 MeO!, I MeO!,,! MeOl\) Me0 (1.) (11.) (III.) w-Methoxyacetoveratrone and w 3 4 5tetramethoxyaceto-phenone are much more soluble in cold than in hot water a pro-perty which doubtless indicates the formation and decomposition of hydrates. The former ketone on condensation with P-resorcyl-aldehyde in presence of hydrogen chloride yields trimethylfisetinidin chloride (IV) and the derived ferrichloride m.p. 156" is identical with the substance obtained as mentioned above. 2-Hydroxy-4 6-dimethoxybenzaldehyde and W-methoxyacetoveratrone con-dense under the agency of hydrogen chloride in ethereal solution with formation of cyanidin chloride pentamethyl ether (V) and this salt is demethylated by means of boiling hydriodic acid in presence of phenol with formation of cyanidin iodide. Cyanidin chloride (VI) was prepared from the iodide in the usual manner and the product exhibited in every detail the properties of cyanidin chloride of natuml origin as described by Willstatter and Everest (Annulen 1913 401 189) and Willstiitter and Nolan (ibid. 1915, 408 13). A direct comparison of the synthetical material with an authentic specimen disclosed no dif€erences existing between them.The production of cyanidin chloride by the reduction of quercetin in acid solution (Willstatter and Mallison Sitxungsber. Preu-ss 16s PR-4TT AND ROBINSON A SYNTHESIS OF Akad. Wiss. 1914 769) constitutes in one sense a synthesis of the substance and furthermore Willstatter and Kindler in 1919 made 8 communication to the Munchner Chemische Gesellschaft con-cerning experiments which they had conducted on the application of the Willstatter-Zechmeister pelargonidin synthesis to the case of cyanidin (‘‘ Wissenschaftliche E’orschungsberichte 111 Organische Chemie,” R. Pummerer 2nd Edition 1923 150). The details of the results obtained are not yet available.” c1 n 0 OMe c1 0 F-(VII.) The synthesis of delphinidin chloride (VII) followed the lines of that of cyanidin chloride delphinidin chloride hexamethyl ether being obtained from hydroxydimethoxybenzaldehyde and o 3 4 5-tetramethoxyacetophenone (11).As shown in the ex-perimental portion on page 174 there can be no doubt as to the identity of the synthetical product obtained on demethylation, with that of natural origin. E Y P E R I M E N T A L . Ethgl uy-Dimethoxyacetoccetate.-Ethyl chloroacetate (I22 9.) was gradually added with cooling and shaking to a solution of sodium methoxide (from 23 g. of sodium) in anhydrous methyl alcohol (200 c.c.). The mixture was refluxed until neutral most of the alcohol then removed a.nd the residue mixed with brine and ether. The dried ethereal solution on fractionation yielded 85 g. of ethyl methoxyacetste b.p. 1 3 1 O . j - Sodium (8 g.) cut in thin slices, * Willsttitter Zechmeister and Kindler have now published their work (Ber. 1924 57 [B] 1938). 5 This substance can also be obtained by the hydrolysis of methoxy-acetonitrile by Pinner’s method. Ethyl ethoxyacetate (Henry? Ber. 1871, 4 706) ; ethyl ay-diethoxyacetoacetate (Conrad Ber. 1878 11 58 ; Erlen-bach Annalen 1892 269 28). In order t o avoid the possibility of replace PYRYLIUM SALTS OF ANTHOCYAXIDIN TYPE. PART V. 169 was added in one portion to ethyl methoxyacetate (100 g.) ; the reaction which soon commenced was controlled by cooling in melting ice when necessary. The thick syrup which resulted was treated with 50% acetic acid (45 c.c.) and saturated brine and extracted with ether.The ethereal layer was separated washed with aqueous sodium cerbonate dried with anhydrous sodium sulphate and distilled; 40 g. of a colourless oil b. p. 130"/15 mm., were obtained (Found C = 50.4; H = 7.4. C,H,,O requires C = 50.5; €1 = 7.4%). This ester is moderately readily soluble in water and its alcoholic solution gives a violet coloration with ferric chloride the shade being bluer than that developed by ethyl acetoacetate under similar conditions. w 4-DimetAox~aceto~henone.-In the first place experiments on the hydrolysis of the condensation product from ethyl sodio-ay-dimethoxyacetoacetate and benzoyl chloride were made. After boiling with dilute sulphuric acid and distillation in a current of steam a yellow distillate was obtained and this by suitable treat-ment yielded phenylglyoxaldiphenylhydrazone m.p. 152". When, however the product was boiled with dilute aqueous sodium hydroxide o-methoxyacetophenone was obtained in small yield and identified as the semicarbazona m. p. 85". w 4-Dimethoxy-acetophenone may be obtained in the following manner Sodium (2.7 g.) was granulated under toluene washed with ether suspended in anhydrous ether (150 c.c.) and ethyl ay-dimethoxyacetoacetate (22.3 g.) gradually added. A sodio-derivative separated from the solution and when the sodium had disappeared anisoyl chloride (20 g.) was added in one portion. A gentle reaction occurred and after 1 hour the process was completed by heating on the steam-bath under reflux during 4 hours. Next day water was added the ethereal layer dried over sodium sulphate and after removal of the solvent 35 g.of a pale yellow oil remained. The whole wag agitated for 12 hours with 2.5% aqueous potassium hydroxide (500 c.c.) and the liquid then boiled for 3 hours cooled nearly saturated with potassium carbonate and extracted with ether ; 7.5 g. of the ketone m. p. 40° were obtained. On condensation with 6-aminopiperonal in alcoholic solution by means of potassium hydroxide o 4-dimethoxyacetophenone yields ment of the ethyl group of the esters by methyl and the production of mixtures Mr. N. L. Matthews has employed methyl chloroacetate and condensed this with sodium methoxide in dry methyl-alcoholic solution in the cold. Methyl chloroacetate (108 g.) yielded methyl methoxyacetate (70 g.) b.p. 129'1754 mm. Methyl methoxyacetate (100 g.) reacted with sodium (9 g.) to give a product from which methyl ay-dimethoxyxetoacetata b. p. 129'/17 mm. could be isolated (Found C = 47.5; H = 7.1. C,H,,O requires C = 47.7; R = 6.8%). The yield was 40-50% of that required by theory. G 170 PRATT AND ROBINSON A SYNTHESIS OF 3-methoxy-6 7-methylenedioxy-2-(4-methoxyphenyl)quinoline which crystallises from methyl alcohol in colourless laminze m. p. 152". o-Metlho~yucetoveratrone (I).-Ethyl ay-dimethoxyacetoacetate (36.2 g.) was gradually added to finely granulated sodium (4.4 g.) suspended in anhydrous ether (225 c.c.). Formation of the sodio-derivative appeared to be complete after an hour and a solution of distilled veratroyl chloride (38 g.) in ether (150 c.c.) was intro-duced causing a mild reaction.The mixture was gently heated on the steam-bath for 2 hours allowed t o remain over-night and heated during a further 2 hours. After washing with water drying, and evaporating the ether there remained 63 g. of a pale orange, viscous liquid which could not be crystallised. The whole product was hydrolysed by vigorous agitation €or 9 hours with cold 2.5% aqueous potassium hydroxide (800 c.c.) and finally by boiling the clear solution for 5 hours. The ketone was isolated by extraction with ether after the addition of much potassium carbona$e and purified by distillation; 22 g. of an oil b. p. 190"/15 mm. were obtained which solidified. The substance crystallises from light petroleum containing a little benzene in colourless prisms m.p. 62" (Found C = 63.1 ; H = 6.8. C,,H,,O requires C = 62-9; H = 607%). When a little of this substance is melted under water and the mixture cooled a clear solution is obtained from which the ketone may be precipitated as an oil either by heating or by the addition of an alkali carbonate or hydroxide. The substance is very readily soluble in organic solvents with the exception of light petroleum. On condensation with 6-aminopiperonal it yields 3 - methoxy - 6 7 - msthylenedioxy-2 -(3 4-dimethoxypheiayl)quinoline, which crystallises from ethyl alcohol in slender needles m. p. 155", and exhibits a violet fluorescence in alcoholic solution. Fisetinidin Chloride Trimethyl Ether (IV) . A slow stream of hydrogen chloride was passed through an ice-cold solution of p-resorcylaldehyde (2 g .) and o-methoxyaceto-veratrone (3 g . ) in dry ether (40 c.c.) when the liquid became yellow and then red and crystalbation commenced after 20 minutes. The passage of the gas was discontinued after 30 minutes and after several hours the pyrylium salt which exhibited an intense green lustre was collected washed with ether and dried (3-4 g.). The ethereal mother-liquors were washed with dilute hydrochloric acid and a ferrichloride (0.2 g . ) was precipitated. This derivative crystallises from acetic acid in reddish-brown needles m. p. 156", and its properties are identical with those of the ferrichloride previously prepared from the product of methylation of veratroyl-carbinol (loc. cit. ). Fisetinidin chloride trimethyl ether dissolves i PYRYLIUM S,4LTS O F SNTHOCYANIDIX TYPE.PART V. 17 1 hot 70/6 hydrochloric acid to an orange-red solut'ion and separate.3 completely on cooling in reddish- brown prismatic needles exhibiting a green reflex (Found in material dried in a vacuum C = 56.1 : H = 3.3. This salt darkens a t 135" and decomposes a t 188-189". It is reddish-violet by transmitted light and makes a violet smcar on paper. 1 ts alcoholic solutions are intensely reddish-violet and the orange solution in concentrated sulphuric acid exhibits a green fluorescence. Sparingly soluble in cold water it dissolves on warming to an orange-red solution which on dilution is rapidly decolorised with formation of the $-base. The violet colour-base is precipitated when sodium acetate is added to a moderately concentrated aqueous solution of the salt hut if a dilute solution is treated with sodium acetate the appearance of the violet colour is transient and the +-base is obtained.isoAmyl alcohol extracts the salt completely from an acid aqueous solution and the addition of sodium acetate to the red alcoholic extract gives a violet solution the colour of which fades. The oxonium salt is regenerated from solutions of the colour-base and $-base by the addition of hydrochloric acid. Cl,K1,O,Cl,BH,O requires C = 56-2 ; H = 5.5%). Cyaizidiii Chloride Pentamethyl Ether (V). I n order t o obtain good results in this preparation it is very necessary that the materials employed should be pure. Hydrogen chloride was passed for 1 hour through a solution of 2-hydroxy-4 5-dimethouybenzaldehyde (4-5 g.) and w-methosyacetoveratrone (5-2 g.) in ether (50 c.c.).The solution quickly became crimson and crystallisation of the product was induced by scratching after 30 minutes from the commencement. The tube containing the mixture and a dish containing solid potassium hydroxide were together covered by a bell-jar and after several hours the salt was collected washed with ether and dried (5.5 g.). The salt crystal-lises from alcohol containing a little hydrogen chloride in small, red necdles m. p. 152" which in mass exhibit an old-gold sheen, but give a chocolate-brown smear on paper (Found in material dried in a vacuum over phosphoric oxide and potassium hydroxide : C = 51.7 ; H = 5.9 ; C1 = 14.9. C,,,H,,B,Cl,HCl,2H20 requires C = 51.6; H = 5.6; C1 = 15.2%).The same dtchloride crystal-lised from a mixture of alcohol and concentrated aqueous hydro-chloric acid in slender needles m. p. 152" (Found C = 51.5; H = 5.5; C1 = 15.4%). From 776 aqueous hydrochloric acid the salt separates in long orange-brown needles which form a crystalline crust exhibiting a green reflex (Found in material dried in a vacuum C = 53-2 ; H = 6.0. C,,H,,O,C1,3H,O requires C = 53-7 ; H = 6.00/,). The substance is readily soluble in the simple c," 172 PRATT AND ROBINSON A SYNTHESIS OF alcohols to reddish-violet solutions the colour of which fades on dilution with water and can then be restored by the addition of a mineral acid. The ferrichloride crystallises from acetic acid in which it is very sparingly soluble in reddish-brown prismatic needles m.p. 196200' (decomp.) exhibiting a bronze lustre. The chloride is insoluble in chloroform but this derivative is moderately readily soluble and when the hot saturated solution is rapidly concentrated to one-quarter of its bulk the ferrichloride separates in microscopic needles exhibiting an intense green lustre. Cyanidin Chloride (VI). The foregoing pentamethyl ether (3.2 g.) and phenol (15 g.) were added to hydriodic acid (180 c.c.; d 1.7) and the mixture was boiled for 30 minutes in an atmosphere of carbon dioxide. Water (2 vols.) and much ether were added to the cooled solution and on standing glistening needles with a bright green reflex separated. The substance (1.8 g.) had properties identical with those of cyanidin iodide prepared by the demethylation of paeonidin chloride of natural origin (Willstatter and Nolan AnnaZen 1915 408 136).The iodide (3 g.) was thoroughly ground with aqueous sodium acetate the bluish-violet colour-base separated thoroughly washed, and heated to boiling with 3% aqueous hydrochloric acid (100 c.c.) in presence of a trace of silver and filtered 50 C.C. more 3% hydrochloric acid being used for washing. Concentrated hydro-chloric acid (50 c.c.) was added and on standing the dark red solution deposited the salt completely in masses of short reddish-brown prisms that had a green lustre (Found C = 52.5 ; H = 4.1. C1,HllO,C1,H,O requires C = 52.8; H = 343%). This product showed all the characteristic reactions of cyanidin chloride and a direct comparison with a specimen of the substance of natural origin disclosed no difference in properties of any kind.The behaviour on heating appearance of the crystals under the micro scope formation of colour-base and +-base solubilities in aqueous and alcoholic acid solutions colour of acid solutions absorption spectrum of the solution in ethyl alcohol and the reactions with ferric chloride sodium carbonate sodium acetate potassium acetate in alcoholic solution lead acetate and Fehling's solution were all examined. Willstatter and Everest (Zoc. cit.) state that cyanidin on oxidation with hydrogen peroxide yields a yellow crystalline product closely resembling a flavonol colouring matter. There is, however no clear evidence that quercetin can be obtained by the oxidation of cyanidin and furthermore it is remarkable that no chromone or chromonol derivative has yet been prepared by oxidising a benzopyrylium salt.We have made numerous experi PYRYLIUM SALTS OY ANTHOCYANIDIN TYPE. PART v. 173 ments with the object of achieving this transformation but any definite products isolated have been either carboxylic acids or cournarin derivatives. This investigation proceeds. w 3 4 5-Tetramethxyacetophenone (II).-Sodium (2.5 g.), granulated under toluene and washed with ether was suspended in ether (120 c.c.) and ethyl ccy-dimethoxyacetoacetate (20 8.) slowly added. When the formation of the sodio-derivative appeared to be complete an ethereal solution of trimethylgalloyl chloride (24.5 g.) was added in one portion and the mixture heated gently on the steam-bath for 8 hours.The product was isolated (34 9.) and shaken for 12 hours with 2.5% aqueous potassium hydroxide. The solution was boiled for 4 hours and the ketone isolated as in the case of w-methoxyacetoveratrone. The yield was 9 g. of a colourless oil b. p. 212"/15 mm. which solidified in contact with light petroleum. The substance crystallises from benzene-light petroleum in colourless needles m. p. 54" and closely resembles w-methoxyacetoveratrone (Found C = 60.0; H = 6.3. C,,H,,O, requires C = 60-0; H = 6.6%). The semicarbaxone crystalliser from aqueous alcohol in long colourless needles m. p. 158". 3-Metltoxy-6 ?-methylenedioxy-2-(3 4 5-trimethoxyphenyl)quinoE-ine crystallises from ethyl alcohol in colourless needles m.p. 159", and exhibits a violet fluorescence in alcoholic solution. Its salts, e.g. the sparingly soluble hydrochloride are yellow. The metho-svlphute is orange-yellow and it is evident from a comparison of this substance with the intensely coloured delphinidin chloride hexamethyl ether that the quinolinium and benzopyryhm nuclei function quite differently as chromophores. It may well be that the methoxyl groups are able to decentralise the cationic valency of the pyrylium nucleus and so produce changes of orbits of electrons owing to recurrent redistribution of the charge whilst on the other hand the powerfully basic quinolinium nucleus is able to ignore the claims of the weak auxochromes. We are thus led to anticipate that the aminophenylquinolinium salts will resemble tinctorially the related hydroxyphenylbenzopyrylium salts and it is hoped that an experimental test of this point will be made.Delphinidin Chloride Hexamethgl Ether (corresponding with VII). Hydrogen chloride was passed through an ice-cold solution of w 3 4 5-tetramethoxyacetophenone (4 g.) and 2-hydroxy-4 6-dimethoxybenzaldehyde (3 g.) in dry ether (50 c.c.) crystallisation of the product in crimson needles and prisms having a dark green reflex occurring after about 1 hour. Next day the salt (4 g . ) was collected and a further quantity (0.7 g.) was obtained from the ethereal mother-liquor. The substance crystallises from alcoho 174 A SYNTHESIS OF PYRYLIUM SALTS OF ANTHOCYANIDIN TYPE. containing a trace of hydrogen chloride in long red needles m.p. 163-164" (Found in material dried in a vacuum C = 58.5 ; H = 5.8. C21H,30,CI,0-5H20 requires C = 58.4; H = 5.6%). The substance exhibits a green lustre and is reddish-violet by transmitted light. It is readily soluble in the simple alcohols and moderately soluble in chloroform to a reddish-violet solution. The dark red colour of the solution in water rapidly fades and an almost colourless $-base is precipitated. Faintly acid solutions are also decolorised on great dilution but addition of acid restores the colour of the pyrylium salt. The ferrichloride crystalhes from glacial acetic acid in brownish-red needles which have a dark green appearance in mass. This derivative m. p. 169-170" is readily soluble in acetone chloroform or the simple alcohols.The sulphate crystallises from dilute sulphuric acid in slender red needles with green reflex m. p. 225" (decomp.). Delphinidin Chloride (VII). Delphinidin chloride hexamethyl ether (3-0 g.) was demethylated by means of a boiling mixture of hydriodic acid (180 c.c.; d 1.7) and phenol (15 g.) during 30 minutes a slow current of carbon dioxide being passed through the containing vessel. The cooled liquid deposited clusters of long pointed brown needles together with some squat prisms and these were collected washed with ether and dried (2.0 g.). The iodide was converted into chloride by treatment a t 60" with silver chloride in alcoholic solution in a silver-mirrored vessel. The filtered solution was mixed with an equal volume of concentrated hydrochloric acid and the amorphous precipitate so obtained allowed to remain when it slowly became resolved into a mass of microscopic needles.The substance was isolated and crystallised by solution in 5% hydrochloric acid and addition of 30% acid to the filtered solution so as to make the concentration of hydrochloric acid about 25%. The brown, crystalline mass was dried in a vacuum (Found C = 49.4; H = 4-0. C1,Hl10,Cl,1~5H20 requires C = 49.4; H = 3-8 ; H20 = 7.4y0). According to Willstatter and Weil (Annalen 1916 412, 178) delphinidin chloride crystallises from hydrochloric acid of 20% or greater concentrations with 1.5H20 which is completely lost in a vacuum. Our product did not behave in this way but the circumstance was an accident since when the material was powdered and again exposed to a vacuum it lost 7.6%.The anhydrous salt does not melt below 350". The hydrates containing 1H20 2H20 and 4H20 have been prepared by the methods devised by Willstatter and Weil in order to observe the crystalline forms, which were found to agree with the descriptions of these authors SYNTHESIS OF CERTAIN HIGHER ALIPHATIC COMPOUNDS. PART I. 175 The absorption spectrum of an alcoholic solution of the chloride, the formation of the +base the behaviour of aqueous and acid solutions on shaking with ether ethyl acetate and isoamyl alcohol, the reactions with ferric chloride in aqueous and alcoholic solutions and with sodium carbonate sodium bisulphite lead acetate and Fehling’s solution were all examined with results identical with those described in the case of delphinidin chloride of natural origin by Willstatter and Mieg (AnnaZen 1915 408 61) and Willstatter and Weil (Zoc. cit.). We were not able to make a direct comparison, but the properties of delphinidin chloride are so characteristic and have been recorded in such detail that we can entertain no doubt as to the identity of the synthetical product with that obtained by the hydrolysis of the pigments of the wild purple larkspur and the blue-black pansy. One of us (D. D. P.) desires to thank the Carnegie Trust for t3he Universities of Scotland for a Fellowship which has enabled him to take part in this investigation. THE UNIVERSITY MANCHESTER. [Received November 3 ~ d 1924.