~ ~ 0 0 ~ AND LILLEY TRANSFORMATION OF 1 ~ ETC. 95 XVI.-Transformation of Mandelonitril e to M and el o -isonitrile. By CHARLES EDXUND WOOD and HAROLD SAMUEL LILLEY. MANDELONITRILE which has been. kept for some time becomes viscous and in some cases deposits a dark yellow solid owing to transformation of the normal nitrile to the iso-form; such material gives a poor yield (50%) of mandelic acid on hydrolysis. The rate of change of the nitrile to the iso-form is variable and depends on the purity of the potassium cyanide used in its prepar-ation. Crude commercial potassium cyanide gives products which remain clear for a long period (4-5 days) but 95yo potassium cyanide (Poulenc Freres) yields a nitrile in which the rate of change is much greater. The following results show the maximum observed rate of change with time in the nitrile and the corresponding reduc-tion in the yield of mandelic acid :-Time (days) ............0 0.25 0.5 0-75 1.0 2-0 3.0 4-0 yo Yield of acid ...... 94.6 91.0 86.5 82.0 78.1 65.0 56.4 50.3 The yields of acid obtained do not correspond with the corn-position of the mixture hydrolysed owing to a reversion of the is0 to the normal form under the conditions of hydrolysis. The results indicate that to give the best yield of acid the nitrile should be rapidly separated and hydrolysed (compare Pape Chem. Ztg. 1896, The change of the nitrile on standing is continuous no reaction product other than the yellow solid appears to be formed and a nearly theoretical yield of the isonitrile is ultimately obt~iiied.The reversibility of the reaction is indicated by the hydrolysis of the pure isonitrile which yields 38% of mandelic acid. A similar case is that of the cyclohexane analogue of mandelo-nitrile. The colourless oily hexahydrobenzaldehydecyanohydrin 20 90) 96 WOOD AND LILLEY TRANSFORMATION OF (Zelinsky and Gutt Ber. 1908 41 2677) becomes dark and viscous on standing and when hydrolysed gives as main product a dark brown tar and little hexahydromandelic acid. Investigation of the tar showed that the isonitrile was present in the product hydrolysed. E x P E R I M E N T A L . Fornzation of Mandeloisonitrile (a-liydroxy6enzylcarbylamine) from Mai2deZonitriZe.-Mandelonitrile on keeping deposits mandeloiso-nitrile as a yellow solid which after crystallisation from chloroform, melts a t 196" (corr.) (Found N = 10.7; 11.1 = 139 in chloroform, 136 in carbon tetrachloride 130 in camphor.Calc. N = 10.6y0; M = 133). The isonitrile is fairly soluble in alcohol ether or benzene more soluble in carbon tetrachloride and insoluble in water. It darkens on standing in air is not attacked by alkali, but is rapidly decomposed by dilute or concentrated mineral acids. With concentrated sulphuric acid it gives a bright green colour, whereas the normal nitrile gives a red one (Schiff Ber. 1899 32, 2701). The isonitrile structure was confirmed by reduction of the corn-pound to a secondary base oxidation to a carbimide and by examination of its hydrolytic products. This compound is distinct from the yellow solid m. p. 206", obtained by Minovici (Ber.1899,32,2206) by the action of hydrogen chloride on an ethereal solution of mandelonitrile for which the formula C,H,*CH(CN).C(oH)(CN).C,H was suggested. Even after repeated crystallisation our compound melted a t the temperature stated; moreover it differs from Minovici's compound in its solu-bility in alcohol and ether and in its inability to form a salt m. p. 222" with hydrochloric acid. Redaction.-The isonitrile (5 g.) was heated with a solution of 25 g. of " hydros " * in 100 C.C. of a mixture of equal parts of alcohol and water for 30 minutes on the water-bath the excess of hydros decomposed with warm hydrochloric acid and the liquid filtered into cone ent ra t ed aqueous ammonia when a- h y &ox y b e m y 1 -methylamine C,H5*CH(OH)*NH*CH, was precipitated.The crude product was purified by conversion to the ether-insoluble picrate decomposition of the salt with ammonia and crystallisation of the liberated base from petroleum (b. p. 60-80"). The low yield (50%) is probably due to sulphamation during reduc-tion. The base is a yellowish-grey solid m. p. 180" (decornp.), readily soluble in alcohol ether or mineral acids (Found N = 10.3. C,H,,ON requires N = 10-30/,). The nitrosoarnine prepared in the * Stannous chloride or ammonium eldoride and zinc dust may be employed MANDELONITRILE TO MANDELOZSONITRILE. 97 usual way is a yellow oil which solidifies on cooling. The picrate, prepared in ethereal solution is a dark yellow crystalline solid m. p. 238" (decomp.) (Found Picric acid = 62.1.C,H,,ON,C,H,O,N, requires 62.7%). Oxidation.-An ethereal or carbon tetrachloride solution of the isonitrile was heated with mercuric oxide (theor. quantity) on the water-bath for 1 hour. The dark red oil of disagreeable odour obtained on evaporating the filtered solution changed on standing, to a black amorphous mass probably a polymeride. The oily carbimide could not be purified for analysis but its structure was established by means of its reaction products. The oil was treated with gaseous ammonia in carbon tetrachloride solution n4en the red colour rapidly disappeared. The solution, on evaporation deposited almost colourless crystals of the carb-amide C,H,*CH(OH)*NH*CO*NH, m. p. 76" (Found N = 16.95. Theory requires N = 16.90/) readily soluble in water or alcohol, less soluble in benzene.The nitrate is fairly soluble in water but mercuric chloride precipitates from neutral aqueous solutions the slightly yellow almost insoluble mercurichloride, C,M,*CH (OH)*NH*CO*NH,,HgCl, (Found Eg = 45-58. The picrate prepared in ethereal solution was obtained as an orange crystalline solid easily soluble in ether or alcohol [Found : picric acid = 73.6. C6H,*C13( OH)*NH*CO*~\'H,,2C6H3O,N requires 7 3 *3 741. 1lydroZysis.-The isonitrile rapidly dissolved when treated on the water-bath with concentrated hydrochloric acid (5 parts). The small quantity of tarry matter produced was removed the filtrate distilled in steam and benzaldehyde extracted from the distillate, which then showed after neutralisation the characteristic reducing properties of formic acid.The benzaldehpde is produced by the decomposition of the primary base initially formed in the hydrolysis which may be represented as follows : OH*CHPh*NC + H*CO,H + OH-CHPh*P\rH + Ph*CHO+NH,. The amounts of formic acid and benzaldehyde found were not theoretical; in addition to the tarry matter deposited a certain amount of mandelic acid was extracted. By the hydrolysis of the pure isonitrile for 3 hours under the above conditions a 38% yield of mandelic acid was obtained. The mandelic acid was estimated by neutralisation of the filtered reaction product with ammonia followed by addition of zinc sulphate or cadmium chloride solution ; zinc or cadmium mandelate, Theory requires Hg = 45433%). VOL. CXXVII.98 SOPER THE HYDROLYSIS OF ACYLCHLOROAMINES IN WATER. precipitated on standing and shaking was filtered off dried and weighed. Small corrections (McKenzie J. 1899 75 969) were applied for the solubility of these salts. The use of the silver salt in the estimation is precluded owing to the presence of formic acid, and to a lesser extent benzoic acid. Hence the reaction normal nitrile -+ isonitrile is reversible proceeding under the conditions of hydrolysis from the is0 to the normal form (compare Guillemard, Compt. rend. 1907 144 141). Further when mixtures of the two forms are hydrolysed the yield of mandelic acid is greater than that corresponding with the initial composition of the mixture. Also on allowing the normal nitrile to stand the yield of acid on hydrolysis cannot fall below 38% although the change to the iso-form is complete.The observed tarry matter consists probably of compounds produced by the action of ammonia on the nitrile (SchifT Ber., 1899 32 2701). Hydrolysis of Hexahydromandeloisonitri1e.-The pure isonitrile has not been isolated but the normal form kept for 2 months and then hydrolysed gave large amounts of tarry matter and 20% of acid. The tarry product was polymerised hexahydrobenzaldehyde and its ammonia derivative for by treatment with dilute caustic soda solution followed by distillation in steam after acidification, the aldehyde was extracted from the distillate in quant'ity. Formic acid accompanied the viscous products. UNIVERSITY OF BIRMINGHAN, EDGBASTON. [Received October 2nd 1924.~ ~ 0 0 ~ AND LILLEY TRANSFORMATION OF 1 ~ ETC. 95 XVI.-Transformation of Mandelonitril e to M and el o -isonitrile. By CHARLES EDXUND WOOD and HAROLD SAMUEL LILLEY. MANDELONITRILE which has been. kept for some time becomes viscous and in some cases deposits a dark yellow solid owing to transformation of the normal nitrile to the iso-form; such material gives a poor yield (50%) of mandelic acid on hydrolysis. The rate of change of the nitrile to the iso-form is variable and depends on the purity of the potassium cyanide used in its prepar-ation. Crude commercial potassium cyanide gives products which remain clear for a long period (4-5 days) but 95yo potassium cyanide (Poulenc Freres) yields a nitrile in which the rate of change is much greater.The following results show the maximum observed rate of change with time in the nitrile and the corresponding reduc-tion in the yield of mandelic acid :-Time (days) ............ 0 0.25 0.5 0-75 1.0 2-0 3.0 4-0 yo Yield of acid ...... 94.6 91.0 86.5 82.0 78.1 65.0 56.4 50.3 The yields of acid obtained do not correspond with the corn-position of the mixture hydrolysed owing to a reversion of the is0 to the normal form under the conditions of hydrolysis. The results indicate that to give the best yield of acid the nitrile should be rapidly separated and hydrolysed (compare Pape Chem. Ztg. 1896, The change of the nitrile on standing is continuous no reaction product other than the yellow solid appears to be formed and a nearly theoretical yield of the isonitrile is ultimately obt~iiied.The reversibility of the reaction is indicated by the hydrolysis of the pure isonitrile which yields 38% of mandelic acid. A similar case is that of the cyclohexane analogue of mandelo-nitrile. The colourless oily hexahydrobenzaldehydecyanohydrin 20 90) 96 WOOD AND LILLEY TRANSFORMATION OF (Zelinsky and Gutt Ber. 1908 41 2677) becomes dark and viscous on standing and when hydrolysed gives as main product a dark brown tar and little hexahydromandelic acid. Investigation of the tar showed that the isonitrile was present in the product hydrolysed. E x P E R I M E N T A L . Fornzation of Mandeloisonitrile (a-liydroxy6enzylcarbylamine) from Mai2deZonitriZe.-Mandelonitrile on keeping deposits mandeloiso-nitrile as a yellow solid which after crystallisation from chloroform, melts a t 196" (corr.) (Found N = 10.7; 11.1 = 139 in chloroform, 136 in carbon tetrachloride 130 in camphor.Calc. N = 10.6y0; M = 133). The isonitrile is fairly soluble in alcohol ether or benzene more soluble in carbon tetrachloride and insoluble in water. It darkens on standing in air is not attacked by alkali, but is rapidly decomposed by dilute or concentrated mineral acids. With concentrated sulphuric acid it gives a bright green colour, whereas the normal nitrile gives a red one (Schiff Ber. 1899 32, 2701). The isonitrile structure was confirmed by reduction of the corn-pound to a secondary base oxidation to a carbimide and by examination of its hydrolytic products. This compound is distinct from the yellow solid m.p. 206", obtained by Minovici (Ber. 1899,32,2206) by the action of hydrogen chloride on an ethereal solution of mandelonitrile for which the formula C,H,*CH(CN).C(oH)(CN).C,H was suggested. Even after repeated crystallisation our compound melted a t the temperature stated; moreover it differs from Minovici's compound in its solu-bility in alcohol and ether and in its inability to form a salt m. p. 222" with hydrochloric acid. Redaction.-The isonitrile (5 g.) was heated with a solution of 25 g. of " hydros " * in 100 C.C. of a mixture of equal parts of alcohol and water for 30 minutes on the water-bath the excess of hydros decomposed with warm hydrochloric acid and the liquid filtered into cone ent ra t ed aqueous ammonia when a- h y &ox y b e m y 1 -methylamine C,H5*CH(OH)*NH*CH, was precipitated.The crude product was purified by conversion to the ether-insoluble picrate decomposition of the salt with ammonia and crystallisation of the liberated base from petroleum (b. p. 60-80"). The low yield (50%) is probably due to sulphamation during reduc-tion. The base is a yellowish-grey solid m. p. 180" (decornp.), readily soluble in alcohol ether or mineral acids (Found N = 10.3. C,H,,ON requires N = 10-30/,). The nitrosoarnine prepared in the * Stannous chloride or ammonium eldoride and zinc dust may be employed MANDELONITRILE TO MANDELOZSONITRILE. 97 usual way is a yellow oil which solidifies on cooling. The picrate, prepared in ethereal solution is a dark yellow crystalline solid m. p. 238" (decomp.) (Found Picric acid = 62.1.C,H,,ON,C,H,O,N, requires 62.7%). Oxidation.-An ethereal or carbon tetrachloride solution of the isonitrile was heated with mercuric oxide (theor. quantity) on the water-bath for 1 hour. The dark red oil of disagreeable odour obtained on evaporating the filtered solution changed on standing, to a black amorphous mass probably a polymeride. The oily carbimide could not be purified for analysis but its structure was established by means of its reaction products. The oil was treated with gaseous ammonia in carbon tetrachloride solution n4en the red colour rapidly disappeared. The solution, on evaporation deposited almost colourless crystals of the carb-amide C,H,*CH(OH)*NH*CO*NH, m. p. 76" (Found N = 16.95. Theory requires N = 16.90/) readily soluble in water or alcohol, less soluble in benzene.The nitrate is fairly soluble in water but mercuric chloride precipitates from neutral aqueous solutions the slightly yellow almost insoluble mercurichloride, C,M,*CH (OH)*NH*CO*NH,,HgCl, (Found Eg = 45-58. The picrate prepared in ethereal solution was obtained as an orange crystalline solid easily soluble in ether or alcohol [Found : picric acid = 73.6. C6H,*C13( OH)*NH*CO*~\'H,,2C6H3O,N requires 7 3 *3 741. 1lydroZysis.-The isonitrile rapidly dissolved when treated on the water-bath with concentrated hydrochloric acid (5 parts). The small quantity of tarry matter produced was removed the filtrate distilled in steam and benzaldehyde extracted from the distillate, which then showed after neutralisation the characteristic reducing properties of formic acid.The benzaldehpde is produced by the decomposition of the primary base initially formed in the hydrolysis which may be represented as follows : OH*CHPh*NC + H*CO,H + OH-CHPh*P\rH + Ph*CHO+NH,. The amounts of formic acid and benzaldehyde found were not theoretical; in addition to the tarry matter deposited a certain amount of mandelic acid was extracted. By the hydrolysis of the pure isonitrile for 3 hours under the above conditions a 38% yield of mandelic acid was obtained. The mandelic acid was estimated by neutralisation of the filtered reaction product with ammonia followed by addition of zinc sulphate or cadmium chloride solution ; zinc or cadmium mandelate, Theory requires Hg = 45433%).VOL. CXXVII. 98 SOPER THE HYDROLYSIS OF ACYLCHLOROAMINES IN WATER. precipitated on standing and shaking was filtered off dried and weighed. Small corrections (McKenzie J. 1899 75 969) were applied for the solubility of these salts. The use of the silver salt in the estimation is precluded owing to the presence of formic acid, and to a lesser extent benzoic acid. Hence the reaction normal nitrile -+ isonitrile is reversible proceeding under the conditions of hydrolysis from the is0 to the normal form (compare Guillemard, Compt. rend. 1907 144 141). Further when mixtures of the two forms are hydrolysed the yield of mandelic acid is greater than that corresponding with the initial composition of the mixture. Also on allowing the normal nitrile to stand the yield of acid on hydrolysis cannot fall below 38% although the change to the iso-form is complete. The observed tarry matter consists probably of compounds produced by the action of ammonia on the nitrile (SchifT Ber., 1899 32 2701). Hydrolysis of Hexahydromandeloisonitri1e.-The pure isonitrile has not been isolated but the normal form kept for 2 months and then hydrolysed gave large amounts of tarry matter and 20% of acid. The tarry product was polymerised hexahydrobenzaldehyde and its ammonia derivative for by treatment with dilute caustic soda solution followed by distillation in steam after acidification, the aldehyde was extracted from the distillate in quant'ity. Formic acid accompanied the viscous products. UNIVERSITY OF BIRMINGHAN, EDGBASTON. [Received October 2nd 1924.