70 ADAM AND DYER SYNTBESIS OF XI.-Synthesis of Arachidic Acid and some Long-chain Compounds. By NEIL K. ADAM and JOSEPH W. W. DYER. THE melting points of methyl and ethyl arachidates and eicosyl alcohol recorded in the literature are several degrees higher than those proper to compounds of their length of chain in the respective series. The regular rise of melting point in the even series of long-chain compounds is more than a mere empirical relationship, since one of us showed (Proc. Roy. Xoc. 1922 [A] 101 528) that from this rise and other regularities a crystal structure could be predicted which has since been shown by X-ray methods to be correct (Muller and Shearer, J. 1923 123 2043 3152 3156). Having synthesised arachidic acid from stearic acid by means of malonic ester we find that the melting points of these derivatives are as follows.Ethyl arachidate 41.5" to 42.5" ; methyl arachidate, 46" to 47" ; eicosyl alcohol 65" to 65.5". Previously recorded values are; ethyl ester 49" (Schweizer Arch. Pharm. 1884 222 768), 50" (Gossmann Annalen 1854 89 10) ; methyl ester 54.5" (Haller, Compt. rend. 1907 144 598); 53" (Schweizer Zoc. cit.); alcohol, 71" (Haller Zoc. cit.). Our values are nearly the mean between the melting points of the compounds in the 18- and 22-carbon series; the older ones are in some cases higher than those in the 22-carbon series. The melting point of arachidic acid itself is 75" in agreement with previous results and with the position in the series. The synthesis from ethyl stearate gave a 41% overall yield of pure arachidic acid and appears to be much the easiest method of obtaining compounds of the Czo series pure.We record also a modification of the Bouveault and Blanc reduc-tion convenient for long-chain compounds ; and new preparations of eic 0s y 1 acetate eic 0s y lamin e hydrochloride eic 0s y Ic ar bamide , acetohexadecyl- and aceto-octadecyl-amides dioctadecylmalonic acid and dioctadecylacetic acid. E X P E R I M E N TAL. Reduction of Long-chain Esters with Sodium and Alcohol .-The higher ethyl esters are troublesome in the ordinary process owing to their small solubility in alcohol. Addition of a large bulk of light petroleum and substitution of wire or dust for the lumps of sodium give good yields of the alcohol. Octudecyl AZcohol.-Well-dried ethyl stearate (150 g.) dissolved in 1200 C.C.of light petroleum and 150 C.C. of absolute alcohol ARACHIDIC ACID AND SOME LONG-CHAM COMPOUNDS. 71 was added during $ hour to sodium wire (72 g.) covered with a little light petroleum to prevent oxidation at about 80"; heating was continued for 6 hours with occasional shaking. The solvents were distilled off the residual solid was treated with hot brine and after filtration the product was dried and extracted with ether. The ether was evaporated and the residue distilled a t 12 mm. pressure (yield 68.5%). For quantities less than 20 g. of ester the process was modified by using sodium pulverised under xylene and gave similar yields. Eicosyl alcohol from 4-82 g. of ethyl arachidate was obtained in 70% yield.It was not distilled but crystallised to constant m. p. (65-65.5') from 90% alcohol. Eicosyl acetate prepared by dissolving the alcohol in glacial acetic acid and passing hydrogen chloride was crystallised from aqueous acetone; m. p. 39*5-40.5" (Found C = 77.54; H = 13-04. Calc. C = 77-49; H = 13.0;jy0). Preparation of Iodides and of Derivatives of Jlnlonic Acid.-Passing hydrogen iodide into the molten alcohols (see Krafft Ber., 1886,19,2984) gave quantitative yields of the iodides the gas being practically completely absorbed until the reaction was complete. The iodide washed in ethereal solution was heated with the theoretical amount of sodium and ethyl malonate in absolute alcohol. The mixture became neutral and gave ethyl mono-octadecyl-malonate after 2 hours' boiling and standing over-night.Hydrolysis with strong aqueous potash was nearly instantaneous. The calcium salt was prepared and extracted with alcohol and ether and the malonic acid liberated in 700; yield (calculated on the iodide used) by rubbing the purified calcium salt with hydrochloric acid in pre-sence of a little ether which promotes wetting. Ethyl dioctadecylmalonate could not be prepared in one operation, but was obtained by heating for 6 hours 12.2 g. of the crude ruono-octadecylmalonic ester with 0-682 g. of sodium 11.5 g. of octadecyl iodide and 30 C.C. of absolute alcohol a little dry ether being added to obtain a homogeneous liquid. After 48 hours' standing the mixture was neutral. The ester was hydrolysed by heating on the water-bath with alcoholic potash for 10 hours ; the dioctadecyl-malonic acid was purified by means of the calcium salt as before, and recrystallisation from glacial acetic acid.M. p. 93-5-94' [Found C = 76.88; H ==12.33. (C,8H,,),C(C0,H) requires Dioctadecylacetic Acid.-Heating dioctadecylmalonic acid at 160" left a residue m. p. 80-42"; this was raised to 81-82" by crystallisation from acetic acid [Found C = 80.63; H = 13.39. (C,,H,,),CH*CO,H requires C = 80.76; H = 13.560/,]. C = 76.89 ; H = 12.57%] 72 ADAM AND DYER SYNTHESIS OF AR,ACHIDIC ACID ETC. Guthzeif (Annalen 1881 206 362) describes dicetylmalonic acid (m. p. 86-87') and dicetylacetic acid (m. p. 69-70"); his work has been criticised by Ihafft since the cetyl alcohol used was not pure. If the crystal structure of these compounds is generally the same as that of the simple fatty acids with the long chains packed side by side and the molecules arranged in pairs of layers, the polar groups of one layer of the pair being next to the polar groups of the other layer we should expect the difference of two carbon atoms to make a difference of about 3" in the m.p.'s of the dialkylmalonic acids and of about 6" in those of the dialkylacetic acids which would indicate for the pure dihexadecylmalonic acid m. p. about 91', and for dihexadecylacetic acid m. p. about 75". Arachidic Acid and its Esters.-The arachidic acid obtained by heating the monooctadecylmalonic acid distilled a t 203-205" (uncorr.)/l mm. and crystallised in leaflets from light petroleum ; m.p. 75-75-5". Titration in alcoholic solution with soda gave the molecular weight 313 (theory 312.3) (Found C = 76-73; H = 12.96. Calc. C = 76-85; H = 12.90y0). Methyl arachidate was prepared by passing hydrogen chloride into a solution of the acid in pure methyl alcohol and recrystallising from alcohol. M. p. unchanged on recrystallisation 46-47' (Found C = 76.96; H = 12.98. Calc. C = 76.85; H = 12.90y0). Ethyl arachidate prepared similarly and crystallised to constant melting point melted at 41.542-6". Preparation of Long-chain Arnicles.-Fileti and Ponzio's method of adding a solution of the acid chloride in ether to aqueous ammonia (Gaxxetta 1593 23 391) gives much better results than methods in which the ether is omitted. These frequently give a product with a melting point 10 degrees too low which probably consists largely of ammonium salt.Using ether as solvent for the chloride, large quantities of the amide may be prepared the chloride being added very rapidly and the product is practically pure without crystallisation. The ether is almost entirely driven off by the heat of reaction and appears to act as a very efficient cooling agent. The aqueous ammonia may be a t room temperature. Arachidamide m. p. log" and the nitrile m. p. 49.5" were pre-pared by the usual methods from the acid. Eicosylamine hydrochloride was prepared by a modification of Krafft's method (Ber. 1889 22 812) for hexadecylamine. The nitrile (9-7 g.) in 150 C.C. of absolute alcohol was boiled under reflux, 15 g. of sodium being added during 1 hour.The mixture was poured warm into dilute hydrochloric acid heated with 250 C.C. of absolute alcohol filtered from any inorganic chlorides diluted to about 85% alcohol and cooled. The solution deposited 8-5 g THE ADSORPTION OX CATALYTICALLY POISOXOUS JIETdLS KTC. of pure eicosylamine hydrochloride IL g. more being obtained 011 concentratling the mother-liquors. For analysis the chlorine was precipitated with alcoholic silver nitrate (Found Cl =T= 10-66. C,,H,,N,HCl requires C1 = 10.64y0). Eicosylcarbamide was easily prepared by evaporating the preceding compound to dryness n.ith excess of potassium cyariate and recrystal-k i n g from alcohol. M. p. 111.5" (corr.) (Pound C = 73.98; H = 13.13; C,oH,l*rU'M-C@-XK requires C =L 74.00; Moiiomolecular films of all thc compounds of the C, series were examined according to thc methods described in previous papers (Proc.Roy. Xoc. 1922 [ A ] 184 452 516) and were found t o have the properties expected for their i-especdive series and thi C, chiill, within experimental error ; a fact which confirms their identity. Aceto-octadecyZamicle.-Octadecylami:ie hydrochloride (1.4 g.), prepared in the same manner as eicosyIaniine hydrochloride i ~ a ~ distilled with quicklime the distillate warmcd with acetic mhyclride a few minutes and the product crystallised from acetic acid nit11 the aid of charcoal; m. p. unchanged by further crysdlisation, 79.5-80" (yield 66%). The action of acetyl chloride on the aririne gave a very poor yield of the desired substance (Found C = 77-23 ; $3 = 13.4 ; N = 4.73.Cl,H3i-NH*CO*CH requires &' == 75439 ; Acetohexadec$amicle was similarly prepared and crystallised from acetic acid and acetone to constant melting point (Found : C = 76.46; H = 13.35. C,,H,,-NH*CQ*CH requires C = 76-69; H = 13.22%). I n the monomolecular films the acetamides showed a characteristic behaviour described in another paper, which indicated that the c'ompunds were members of the same liomologous series and differed by two CM groups. N = S.25. H = 13.04; N = S*25",). H = 13.2; N = 4.50%). THE SORBY RESEARCH LABOR~~TORY, UNIVERSITY OF SHEWIELD. [Receiced June l l f h 1924. 70 ADAM AND DYER SYNTBESIS OF XI.-Synthesis of Arachidic Acid and some Long-chain Compounds. By NEIL K. ADAM and JOSEPH W. W. DYER.THE melting points of methyl and ethyl arachidates and eicosyl alcohol recorded in the literature are several degrees higher than those proper to compounds of their length of chain in the respective series. The regular rise of melting point in the even series of long-chain compounds is more than a mere empirical relationship, since one of us showed (Proc. Roy. Xoc. 1922 [A] 101 528) that from this rise and other regularities a crystal structure could be predicted which has since been shown by X-ray methods to be correct (Muller and Shearer, J. 1923 123 2043 3152 3156). Having synthesised arachidic acid from stearic acid by means of malonic ester we find that the melting points of these derivatives are as follows. Ethyl arachidate 41.5" to 42.5" ; methyl arachidate, 46" to 47" ; eicosyl alcohol 65" to 65.5".Previously recorded values are; ethyl ester 49" (Schweizer Arch. Pharm. 1884 222 768), 50" (Gossmann Annalen 1854 89 10) ; methyl ester 54.5" (Haller, Compt. rend. 1907 144 598); 53" (Schweizer Zoc. cit.); alcohol, 71" (Haller Zoc. cit.). Our values are nearly the mean between the melting points of the compounds in the 18- and 22-carbon series; the older ones are in some cases higher than those in the 22-carbon series. The melting point of arachidic acid itself is 75" in agreement with previous results and with the position in the series. The synthesis from ethyl stearate gave a 41% overall yield of pure arachidic acid and appears to be much the easiest method of obtaining compounds of the Czo series pure.We record also a modification of the Bouveault and Blanc reduc-tion convenient for long-chain compounds ; and new preparations of eic 0s y 1 acetate eic 0s y lamin e hydrochloride eic 0s y Ic ar bamide , acetohexadecyl- and aceto-octadecyl-amides dioctadecylmalonic acid and dioctadecylacetic acid. E X P E R I M E N TAL. Reduction of Long-chain Esters with Sodium and Alcohol .-The higher ethyl esters are troublesome in the ordinary process owing to their small solubility in alcohol. Addition of a large bulk of light petroleum and substitution of wire or dust for the lumps of sodium give good yields of the alcohol. Octudecyl AZcohol.-Well-dried ethyl stearate (150 g.) dissolved in 1200 C.C. of light petroleum and 150 C.C. of absolute alcohol ARACHIDIC ACID AND SOME LONG-CHAM COMPOUNDS.71 was added during $ hour to sodium wire (72 g.) covered with a little light petroleum to prevent oxidation at about 80"; heating was continued for 6 hours with occasional shaking. The solvents were distilled off the residual solid was treated with hot brine and after filtration the product was dried and extracted with ether. The ether was evaporated and the residue distilled a t 12 mm. pressure (yield 68.5%). For quantities less than 20 g. of ester the process was modified by using sodium pulverised under xylene and gave similar yields. Eicosyl alcohol from 4-82 g. of ethyl arachidate was obtained in 70% yield. It was not distilled but crystallised to constant m. p. (65-65.5') from 90% alcohol. Eicosyl acetate prepared by dissolving the alcohol in glacial acetic acid and passing hydrogen chloride was crystallised from aqueous acetone; m.p. 39*5-40.5" (Found C = 77.54; H = 13-04. Calc. C = 77-49; H = 13.0;jy0). Preparation of Iodides and of Derivatives of Jlnlonic Acid.-Passing hydrogen iodide into the molten alcohols (see Krafft Ber., 1886,19,2984) gave quantitative yields of the iodides the gas being practically completely absorbed until the reaction was complete. The iodide washed in ethereal solution was heated with the theoretical amount of sodium and ethyl malonate in absolute alcohol. The mixture became neutral and gave ethyl mono-octadecyl-malonate after 2 hours' boiling and standing over-night. Hydrolysis with strong aqueous potash was nearly instantaneous. The calcium salt was prepared and extracted with alcohol and ether and the malonic acid liberated in 700; yield (calculated on the iodide used) by rubbing the purified calcium salt with hydrochloric acid in pre-sence of a little ether which promotes wetting.Ethyl dioctadecylmalonate could not be prepared in one operation, but was obtained by heating for 6 hours 12.2 g. of the crude ruono-octadecylmalonic ester with 0-682 g. of sodium 11.5 g. of octadecyl iodide and 30 C.C. of absolute alcohol a little dry ether being added to obtain a homogeneous liquid. After 48 hours' standing the mixture was neutral. The ester was hydrolysed by heating on the water-bath with alcoholic potash for 10 hours ; the dioctadecyl-malonic acid was purified by means of the calcium salt as before, and recrystallisation from glacial acetic acid.M. p. 93-5-94' [Found C = 76.88; H ==12.33. (C,8H,,),C(C0,H) requires Dioctadecylacetic Acid.-Heating dioctadecylmalonic acid at 160" left a residue m. p. 80-42"; this was raised to 81-82" by crystallisation from acetic acid [Found C = 80.63; H = 13.39. (C,,H,,),CH*CO,H requires C = 80.76; H = 13.560/,]. C = 76.89 ; H = 12.57%] 72 ADAM AND DYER SYNTHESIS OF AR,ACHIDIC ACID ETC. Guthzeif (Annalen 1881 206 362) describes dicetylmalonic acid (m. p. 86-87') and dicetylacetic acid (m. p. 69-70"); his work has been criticised by Ihafft since the cetyl alcohol used was not pure. If the crystal structure of these compounds is generally the same as that of the simple fatty acids with the long chains packed side by side and the molecules arranged in pairs of layers, the polar groups of one layer of the pair being next to the polar groups of the other layer we should expect the difference of two carbon atoms to make a difference of about 3" in the m.p.'s of the dialkylmalonic acids and of about 6" in those of the dialkylacetic acids which would indicate for the pure dihexadecylmalonic acid m. p. about 91', and for dihexadecylacetic acid m. p. about 75". Arachidic Acid and its Esters.-The arachidic acid obtained by heating the monooctadecylmalonic acid distilled a t 203-205" (uncorr.)/l mm. and crystallised in leaflets from light petroleum ; m. p. 75-75-5". Titration in alcoholic solution with soda gave the molecular weight 313 (theory 312.3) (Found C = 76-73; H = 12.96.Calc. C = 76-85; H = 12.90y0). Methyl arachidate was prepared by passing hydrogen chloride into a solution of the acid in pure methyl alcohol and recrystallising from alcohol. M. p. unchanged on recrystallisation 46-47' (Found C = 76.96; H = 12.98. Calc. C = 76.85; H = 12.90y0). Ethyl arachidate prepared similarly and crystallised to constant melting point melted at 41.542-6". Preparation of Long-chain Arnicles.-Fileti and Ponzio's method of adding a solution of the acid chloride in ether to aqueous ammonia (Gaxxetta 1593 23 391) gives much better results than methods in which the ether is omitted. These frequently give a product with a melting point 10 degrees too low which probably consists largely of ammonium salt. Using ether as solvent for the chloride, large quantities of the amide may be prepared the chloride being added very rapidly and the product is practically pure without crystallisation.The ether is almost entirely driven off by the heat of reaction and appears to act as a very efficient cooling agent. The aqueous ammonia may be a t room temperature. Arachidamide m. p. log" and the nitrile m. p. 49.5" were pre-pared by the usual methods from the acid. Eicosylamine hydrochloride was prepared by a modification of Krafft's method (Ber. 1889 22 812) for hexadecylamine. The nitrile (9-7 g.) in 150 C.C. of absolute alcohol was boiled under reflux, 15 g. of sodium being added during 1 hour. The mixture was poured warm into dilute hydrochloric acid heated with 250 C.C. of absolute alcohol filtered from any inorganic chlorides diluted to about 85% alcohol and cooled.The solution deposited 8-5 g THE ADSORPTION OX CATALYTICALLY POISOXOUS JIETdLS KTC. of pure eicosylamine hydrochloride IL g. more being obtained 011 concentratling the mother-liquors. For analysis the chlorine was precipitated with alcoholic silver nitrate (Found Cl =T= 10-66. C,,H,,N,HCl requires C1 = 10.64y0). Eicosylcarbamide was easily prepared by evaporating the preceding compound to dryness n.ith excess of potassium cyariate and recrystal-k i n g from alcohol. M. p. 111.5" (corr.) (Pound C = 73.98; H = 13.13; C,oH,l*rU'M-C@-XK requires C =L 74.00; Moiiomolecular films of all thc compounds of the C, series were examined according to thc methods described in previous papers (Proc. Roy.Xoc. 1922 [ A ] 184 452 516) and were found t o have the properties expected for their i-especdive series and thi C, chiill, within experimental error ; a fact which confirms their identity. Aceto-octadecyZamicle.-Octadecylami:ie hydrochloride (1.4 g.), prepared in the same manner as eicosyIaniine hydrochloride i ~ a ~ distilled with quicklime the distillate warmcd with acetic mhyclride a few minutes and the product crystallised from acetic acid nit11 the aid of charcoal; m. p. unchanged by further crysdlisation, 79.5-80" (yield 66%). The action of acetyl chloride on the aririne gave a very poor yield of the desired substance (Found C = 77-23 ; $3 = 13.4 ; N = 4.73. Cl,H3i-NH*CO*CH requires &' == 75439 ; Acetohexadec$amicle was similarly prepared and crystallised from acetic acid and acetone to constant melting point (Found : C = 76.46; H = 13.35. C,,H,,-NH*CQ*CH requires C = 76-69; H = 13.22%). I n the monomolecular films the acetamides showed a characteristic behaviour described in another paper, which indicated that the c'ompunds were members of the same liomologous series and differed by two CM groups. N = S.25. H = 13.04; N = S*25",). H = 13.2; N = 4.50%). THE SORBY RESEARCH LABOR~~TORY, UNIVERSITY OF SHEWIELD. [Receiced June l l f h 1924.