HAMER REDUCTION O F THE CARBOCXANINES. 211 XXXV. -Reduction of the Carbocyanines. By FRANCES &lARY HAJIER. THE constitution (I) of carbocyanine was deduced from general considerations supported by analytical evidence and established by the study of the oxidative breakdown of pinacyanole (Ifills and Hamer J. 1920 117 1550). The prediction of similar classes of dyes with the nuclei linked in the 4 4’- and 2 4’-positions, respectively has now been fulfilled (Mills and Braunholtz J., (1.1 (II.) 1923 123 2804; Mills and Odams J. 1924 125 1913) thus confirming the formula. Two syntheses have been adduced in its support. Konig’s ( B e y . 1922 55 3293) consists in condensation of quinaldine alkyl 212 HAMER REDUCTION OF THE CARBOCYANINES. halide with orthoformic ester in presence of acetic anhydride.It is not obvious why this is considered to throw more light on the constitution of ca,rbocyanine than does the earlier preparation by the action of formaldehyde on an alcoholic solution of quinaldine alkylhalide in presence of alkali (Mills and Hamer Zoc. cit.). The synthesis by means of orthoformic ester does however differ from the latter as from the preparation in which the formaldehyde is replaced by a trihalogenmethane (D.R.-P. 200,207) in that a 50% yield of analytically pure product is claimed. But when Konig’s procedure was repeated it was found that even twice the quoted amount of acetic anhydride was barely sufficient for solution of the given quantity of quinaldine ethiodide and that the con-siderable yield of crude product consisted chiefly of impurities.The method is said to be equally applicable to the more soluble methiodide but when this was used in order that the directions might be followed exactly the yield of pure carbocyanine was not more than 10%. In the second synthesis (Hamer J. 1923 123, 246) methylenediquinaldine dialkylhalide was prepared from quinaldine alkylhalide and formaldehyde according to the equation : The action of alkali in presence of quinoline alkylhalide elimin-ated halogen acid with apparently simultaneous removal of two hydrogen atoms since carbocyanine (I) was directly formed. This synthesis proves the existence of a three-carbon chain joining the quinoline nuclei of the carbocyanine molecule and is unequivocal except for the unexplained fact that the presence of quinoline alkylhalide is necessary.The converse transformation of carbocyanine into methylene-diquinaldine dialkylhalide has now been accomplished thus clearly demonstrating the relationship. The addition of hydriodic acid (b. p. 126”) to 1 1’-dirnethylcarbocyanine iodide gave yellow crystals too unstable to isolate for analysis but doubtless pos-sessing formula 11 which only requires addition of two hydrogen a toms to represent met h ylenediqu inaldine dime t hio dide . This reduction was brought about by heating with excess of hydriodic acid at a carefully regulated temperature and a 78% yield of methylenediquinaldine dimethiodide was the result. Simi-larly 1 1‘-diethylcarbocyanine iodide and 6 6’-dimethyl-1 1‘-diethylcarbocyanine iodide were reduced to methylenediquinaldine diethiodide and 6 6’-dimethylmethylenediquinaldine diethiodide, respectively.The temperature is an important factor in th HAMER REDUCTION O F THE CARBOCYANINES. 213 reduction since with a lowering of a few degrees unchanged carbo-cyanine remains whilst too high temperatures lead to decomposition of the dialkylhalide. p-Dimethylaminobenzylidenequinaldine ethiodide was similarly heated with hydriodic acid but analysis of the product proved that, besides reduction of the ethylenic linking the two methyl groups had been replaced by hydrogen atoms. That the compound was p-aminobenzylquinaldine ethiodide and not the hydriodide of p-dimethylaminobenzylidenequinaldine was established by the fact that it was neutral and unattacked by ammonia.F x P E R I M E N T A L. Preparation of 1 1'-Dimethylcarbocyanine Iodide by Use of Cirthoformic Ester (compare Konig loc. cit .).-Quinaldine methiodide (6 g.) was boiled with acetic anhydride (60 c.c. b. p. 135-140"), and during 5 minutes orthoformic ester (2.1 g. b. p. 142-146") was added. The mixture was concentrated to half volume and the solid removed when cold. The greater part of the impurity was extracted by boiling methyl alcohol (15 c.c.) and the undissolved crude carbocyanine was recrystallised from methyl alcohol (yield 67; instead of 50). The method recommended in a later paper (Ber. 1924 57 685) gave a 10% yield. Reduction of 1 1 '-Dimethylcarbocyaizine Iodide.-By heating the carbocyanine (1.5 g . ) with hydriodic acid (10 c.c.b. p. 126") in a sealed tube a t 182-185" for 6 hours black crystals of periodide were produced; addition of water gave a yellow precipitate. The total solid was suspended in boiling 1.5% hydrochloric acid and sulphur dioxide passed in until a clear solution was obtained when potassium iodide was added (1.5 g.). The product (yield 780/) mas recrystallised from dilute hydrochloric acid with charcoal treatment and addition of potassium iodide. The crystals were ground with pyridine (5 c.c.) to remove any acid impurity (see below) washed with acetone recrystallised from absolute alcohol, and dried in the steam-oven (Found C = 47.27; H = 4-19: I = 43.41. Calc. for C2,H2,N,T2 C = 47.42; H = 4.16; I = 43.617;) ; m. p. 20'7" (decornp.) alone or mixed with methylene-diquinaldine dimethiodide.The compound had the properties of this substance and its crystallisation could be started by inoculation with it. By reduction of the carbocyanine a t 203-209" the iodide obtained after two recrystallisations from absolute alcohol was strongly acid ; it appeared to be free from methylenediquinaldine dimethiodide since it was completely and easily soluble in cold pyridine in which the dimethiodide is practically insoluble 214 HAMER REDUCTION OF THE CARBOCYANINES. Reduction of 1 1'-Diethylcarbocyanine Halide.-The iodide or bromide (1 g.) was heated with hydriodic acid (7 c.c.) at 178-181" and on reduction of the resultant periodide the crystalline product (yield 76%) was almost pure (Found I = 41.27y0).It was treated with pyridine and recrystallised from absolute alcohol (yield 54%). This reduction product methylenediquinaldine diethiodide and their mixture melted simultaneously a t 205" (decomp.). For analysis it was dried in the steam-oven (Found : C = 48.80; H = 4.73; I = 41.62. Calc. for C25H28N212 C = 49.18; H = 4.63; I = 41-61y0). Reduction of 6 6'-Dimethyl-1 1 '-diethylcarbocyanine Iodide.-Hydriodic acid (5 c.c.) and the carbocyanine (0.5 g.) [fine needles of the almost colourless hydriodide formed] were heated a t 178-185" for 5 hours and the solid was treated with sulphur dioxide as in other cases (yield 83%). For analysis the iodide was dried over sulphuric acid and soda-lime (Found I = 38-97. Calc. for C,,H,,N21,,H,0 I = 38.68%). It melted at 219" alone or mixed with 6 6'-dimethylmethylenediquinaldine diethiodide.The m. p. was unaltered by recrystallisation. Reduction of p-Dimethylaminobenxylidenequinaldine Ethiodide.-This was prepared by the method of Konig and Treichel ( J . pr. Chenx. 1921 [ii] 102 63). By increasing the time of heating from 1 hour to 24 hours the yield of recrystallised product was raised from 23 to 76% (Found I = 29.33; calc. 29.51%); m. p. 259-263" (decomp.). p-Dimethylaminobenzylidenequinaldine ethiodide ( 1 -5 g.) and hydriodic acid (10 c.c.) were heated for 3 hours at 178-181" water was added and the solid in boiling dilute hydrochloric acid treated with sulphur dioxide. The clear solution gave no precipitate with potassium iodide (1.5 g.) but ammonia produced an orange ta,r which hardened and was powdered filtered and dried in a vacuum desiccator.From absolute alcohol it separated as clear red crystals (yield 59%). A thrice recrystallised specimen was dried in a vacuum desiccator and was shown by analysis to be p-aminobenzyl-quinaldine ethiodide (Found C = 56.64 ; H = 5.26 ; N = 7.08 ; I = 31.30. CI9H,,N,I requires C = 56.41 ; H = 5-24 ; N = 6-93 ; I = 31.41y0). It is moderately soluble in alcohol or water ; m. p. 200-201" with incipient softening a couple of degrees lower. UNIVERSITY CREMICAL LABORATORY CAMBRIDGE. DAVY FARADAY RESEARCH LABORATORY ROYAL INSTITUTION. [Received November 22nd 1924. HAMER REDUCTION O F THE CARBOCXANINES. 211 XXXV. -Reduction of the Carbocyanines. By FRANCES &lARY HAJIER. THE constitution (I) of carbocyanine was deduced from general considerations supported by analytical evidence and established by the study of the oxidative breakdown of pinacyanole (Ifills and Hamer J.1920 117 1550). The prediction of similar classes of dyes with the nuclei linked in the 4 4’- and 2 4’-positions, respectively has now been fulfilled (Mills and Braunholtz J., (1.1 (II.) 1923 123 2804; Mills and Odams J. 1924 125 1913) thus confirming the formula. Two syntheses have been adduced in its support. Konig’s ( B e y . 1922 55 3293) consists in condensation of quinaldine alkyl 212 HAMER REDUCTION OF THE CARBOCYANINES. halide with orthoformic ester in presence of acetic anhydride. It is not obvious why this is considered to throw more light on the constitution of ca,rbocyanine than does the earlier preparation by the action of formaldehyde on an alcoholic solution of quinaldine alkylhalide in presence of alkali (Mills and Hamer Zoc.cit.). The synthesis by means of orthoformic ester does however differ from the latter as from the preparation in which the formaldehyde is replaced by a trihalogenmethane (D.R.-P. 200,207) in that a 50% yield of analytically pure product is claimed. But when Konig’s procedure was repeated it was found that even twice the quoted amount of acetic anhydride was barely sufficient for solution of the given quantity of quinaldine ethiodide and that the con-siderable yield of crude product consisted chiefly of impurities. The method is said to be equally applicable to the more soluble methiodide but when this was used in order that the directions might be followed exactly the yield of pure carbocyanine was not more than 10%.In the second synthesis (Hamer J. 1923 123, 246) methylenediquinaldine dialkylhalide was prepared from quinaldine alkylhalide and formaldehyde according to the equation : The action of alkali in presence of quinoline alkylhalide elimin-ated halogen acid with apparently simultaneous removal of two hydrogen atoms since carbocyanine (I) was directly formed. This synthesis proves the existence of a three-carbon chain joining the quinoline nuclei of the carbocyanine molecule and is unequivocal except for the unexplained fact that the presence of quinoline alkylhalide is necessary. The converse transformation of carbocyanine into methylene-diquinaldine dialkylhalide has now been accomplished thus clearly demonstrating the relationship.The addition of hydriodic acid (b. p. 126”) to 1 1’-dirnethylcarbocyanine iodide gave yellow crystals too unstable to isolate for analysis but doubtless pos-sessing formula 11 which only requires addition of two hydrogen a toms to represent met h ylenediqu inaldine dime t hio dide . This reduction was brought about by heating with excess of hydriodic acid at a carefully regulated temperature and a 78% yield of methylenediquinaldine dimethiodide was the result. Simi-larly 1 1‘-diethylcarbocyanine iodide and 6 6’-dimethyl-1 1‘-diethylcarbocyanine iodide were reduced to methylenediquinaldine diethiodide and 6 6’-dimethylmethylenediquinaldine diethiodide, respectively.The temperature is an important factor in th HAMER REDUCTION O F THE CARBOCYANINES. 213 reduction since with a lowering of a few degrees unchanged carbo-cyanine remains whilst too high temperatures lead to decomposition of the dialkylhalide. p-Dimethylaminobenzylidenequinaldine ethiodide was similarly heated with hydriodic acid but analysis of the product proved that, besides reduction of the ethylenic linking the two methyl groups had been replaced by hydrogen atoms. That the compound was p-aminobenzylquinaldine ethiodide and not the hydriodide of p-dimethylaminobenzylidenequinaldine was established by the fact that it was neutral and unattacked by ammonia. F x P E R I M E N T A L. Preparation of 1 1'-Dimethylcarbocyanine Iodide by Use of Cirthoformic Ester (compare Konig loc.cit .).-Quinaldine methiodide (6 g.) was boiled with acetic anhydride (60 c.c. b. p. 135-140"), and during 5 minutes orthoformic ester (2.1 g. b. p. 142-146") was added. The mixture was concentrated to half volume and the solid removed when cold. The greater part of the impurity was extracted by boiling methyl alcohol (15 c.c.) and the undissolved crude carbocyanine was recrystallised from methyl alcohol (yield 67; instead of 50). The method recommended in a later paper (Ber. 1924 57 685) gave a 10% yield. Reduction of 1 1 '-Dimethylcarbocyaizine Iodide.-By heating the carbocyanine (1.5 g . ) with hydriodic acid (10 c.c. b. p. 126") in a sealed tube a t 182-185" for 6 hours black crystals of periodide were produced; addition of water gave a yellow precipitate.The total solid was suspended in boiling 1.5% hydrochloric acid and sulphur dioxide passed in until a clear solution was obtained when potassium iodide was added (1.5 g.). The product (yield 780/) mas recrystallised from dilute hydrochloric acid with charcoal treatment and addition of potassium iodide. The crystals were ground with pyridine (5 c.c.) to remove any acid impurity (see below) washed with acetone recrystallised from absolute alcohol, and dried in the steam-oven (Found C = 47.27; H = 4-19: I = 43.41. Calc. for C2,H2,N,T2 C = 47.42; H = 4.16; I = 43.617;) ; m. p. 20'7" (decornp.) alone or mixed with methylene-diquinaldine dimethiodide. The compound had the properties of this substance and its crystallisation could be started by inoculation with it.By reduction of the carbocyanine a t 203-209" the iodide obtained after two recrystallisations from absolute alcohol was strongly acid ; it appeared to be free from methylenediquinaldine dimethiodide since it was completely and easily soluble in cold pyridine in which the dimethiodide is practically insoluble 214 HAMER REDUCTION OF THE CARBOCYANINES. Reduction of 1 1'-Diethylcarbocyanine Halide.-The iodide or bromide (1 g.) was heated with hydriodic acid (7 c.c.) at 178-181" and on reduction of the resultant periodide the crystalline product (yield 76%) was almost pure (Found I = 41.27y0). It was treated with pyridine and recrystallised from absolute alcohol (yield 54%). This reduction product methylenediquinaldine diethiodide and their mixture melted simultaneously a t 205" (decomp.).For analysis it was dried in the steam-oven (Found : C = 48.80; H = 4.73; I = 41.62. Calc. for C25H28N212 C = 49.18; H = 4.63; I = 41-61y0). Reduction of 6 6'-Dimethyl-1 1 '-diethylcarbocyanine Iodide.-Hydriodic acid (5 c.c.) and the carbocyanine (0.5 g.) [fine needles of the almost colourless hydriodide formed] were heated a t 178-185" for 5 hours and the solid was treated with sulphur dioxide as in other cases (yield 83%). For analysis the iodide was dried over sulphuric acid and soda-lime (Found I = 38-97. Calc. for C,,H,,N21,,H,0 I = 38.68%). It melted at 219" alone or mixed with 6 6'-dimethylmethylenediquinaldine diethiodide. The m. p. was unaltered by recrystallisation.Reduction of p-Dimethylaminobenxylidenequinaldine Ethiodide.-This was prepared by the method of Konig and Treichel ( J . pr. Chenx. 1921 [ii] 102 63). By increasing the time of heating from 1 hour to 24 hours the yield of recrystallised product was raised from 23 to 76% (Found I = 29.33; calc. 29.51%); m. p. 259-263" (decomp.). p-Dimethylaminobenzylidenequinaldine ethiodide ( 1 -5 g.) and hydriodic acid (10 c.c.) were heated for 3 hours at 178-181" water was added and the solid in boiling dilute hydrochloric acid treated with sulphur dioxide. The clear solution gave no precipitate with potassium iodide (1.5 g.) but ammonia produced an orange ta,r which hardened and was powdered filtered and dried in a vacuum desiccator. From absolute alcohol it separated as clear red crystals (yield 59%). A thrice recrystallised specimen was dried in a vacuum desiccator and was shown by analysis to be p-aminobenzyl-quinaldine ethiodide (Found C = 56.64 ; H = 5.26 ; N = 7.08 ; I = 31.30. CI9H,,N,I requires C = 56.41 ; H = 5-24 ; N = 6-93 ; I = 31.41y0). It is moderately soluble in alcohol or water ; m. p. 200-201" with incipient softening a couple of degrees lower. UNIVERSITY CREMICAL LABORATORY CAMBRIDGE. DAVY FARADAY RESEARCH LABORATORY ROYAL INSTITUTION. [Received November 22nd 1924.