1356 WERNER AND FEARON: CXLVl I.-The Constitut iort of Chdamides. Part X I I L .The Constitution of Cyanic Acid and the Rorrzzation of Ureu from the Interaction of Ammonia and Cyanic Acid at Low Temperatures. . By EMIL ALPHONSE WERNER and WILLIAM ROBERT FEARON. THE theory which has been pub forward by one of us to explain the constitution of urea and the mechanism of its formation from ammonium cyanate (T. 1913 103 1013 2276; 1918 113 84) is based primarily on the conception that cyanic acid whether in the static condition or in solution is an equilibrium mixture, represented thus : Promoted by rise of t". (enol-form) I-lC).CN + HN:CO (keto-form). Promoted by fall of t". temperature. temperature. Stable only at low Stable at high Since cyanic acid is a6 all times very unstable the stability Permanent stability is only referred to above is only relative THE CONSTITUTION O F CARBAMIDES.PART XIII. 1357 attained by polymerisation with the production of cyanuric acid from the keto-form and of cyamelide from the enol-form of the bcid (Zoc. cit.). Now the ease with which aixmoniuni cyanate is changed to urea Es indicative of a close similarity in structure and hence the con-stitution of cyanic acid is s question of paramount importance in solving the problem of the relation which exists between the two isomerides. There are two reasons why previous investigators have given either little or no consideration to this important question namely, the apparent uncertainty which has been assumed to exist as regards the nature of cyanic acid and the general acceptance of the '' carbamide '' formula which has all along masked the true relationship between urea 2nd ammonium cyanate.As a matter of fact tho constitution of cyanic acid is clearly indicated by ( a ) its polymerisation and (6) its hydrolysis and behaviour towards ammonia at low temperatures. (a) Po?!ymerisation of Cymzic A cid. Seniar and TValsh (T. 1902 81 290) showed that the spontaneous polymerisation of cyanic acid yielded a product which contained in round numbers 70 per cent. of cyanuric acid and 30 per cent. of cyanielide. This was the result of a single experi-ment in which no particular effort was made to control the temperature a t which polymerisation took effect. If the theory of the polymerisation of cyanic acid which was propounded by one of us is sound (T.1913 103 1016),* and if cyanic acid is an equilibrium mixture as indicated above it follows that its composition a t any particular temperature will be revealed by the relative proportions of the two polymerides formed. This inference which was predicted when the theory was put forward, has now been verified by analyses of the products formed when liquid cyanic acid polymerised a t different temperatures. At zero the acid was stable for about four hours provided it was not agitated and it polymerised slowly. I n order to obtain a reasonable control of the temperature a t which polymerisation took place only very small quantities of the acid were dealt with in each experiment (see experimental part). The results were as follows: * For the sake of brevity the theory is not reproduced here ; in order to appreciate the significance of the result,s obtained it is necessary to consult the paper on the mechanism of the polymerisation of cyanic acid to which reference is made 1358 WERNER AND FEARON: Temperature of Weight polymerisation of (approximate).polymeride. 0" 0.53 gram. 5 0.42 ,, 10 0.105 ,, 20 0.24 ,, TABLE I. Percentage composition of C y anuri c polymeride. acid v-0.216 gram 59.25 40.75 0.174 , 58.58 41-42 0.045 , 57.27 42.73 0.136 , 42.92 57-08 found. Cyamelide. Cyanuric acid. I n several experiments where the acid polymerised when the containing vessel was plunged into water at 20* the temperature rose suddenly to about 70° and the proportions of cyanuric acid formed were from 70 to 80 per cent.Since the latter acid is almost the sole product found when polymerisation takes place a t high temperatures as for example when urea is heated above its melt-ing point ( 1 3 2 O ) (T. 1913 103 2276) it follows that under such conditions cyanic acid is liberated in the keto-form Liquid cyanic acid on the other hand must be an equilibrium mixture, the composition of which is a function of the temperature; thus, at O" it may be represented as approximately HO*CN HXCO 60 per cent. 40 per cent. (b) Hydrolysis of Cyanic A c i d a d its Reaction with Ammonia at Low Temperatures. Whenever urea is formed in a reaction where cyanic acid and ammonia are concerned it has always been assumed that it must originate from the transf orination of ammonium cyanate produced in the first instance.Since the difference between the two isomerides is nothing more than that of the products of the union of ammonia with the enol- and keto-forms respectively of cyanic acid it is obvious that the above assumption is superfluous. A quantitative study of the hydrolysis of cyanic acid has sup-plied convincing evidence that both isomerides are simultaneously formed when ammonia reacts with the acid a t low temperatures. Not less than six consecutive changes are involved during the pro-gress of this reaction and these are conveniently divided into two groups as follows: Primary changes : (1) (HOCN t BNCO) + H20 = CO + NN,. (2) HOCN + NH = NH4*OCN. (3) HN:CO 4- NH = HN:C <xH3 THE CONSTITUTION OP CARBAMIDES.PART XIII. 1359 Secondary changes : (4) NH3 + CO + H,O = NH4HC03. (5) Formation of biuret from interaction of urea and HNCO. (6) Production of cyamelide. The main question was to prove the validity of reaction (3). This was successfully accomplished when an approximately N / 4-solution of cyanic acid was allowed to hydrolyse a t Oo. Under such conditions the secondary changes were almost completely suppressed up to the point a t which the primary changes were completed. A solution of cyanic acid was prepared by the addi-tion of the theoretical amount (62.5 c.c.) of N-nitric acid to 5-05 grams of pure potassium cyanate dissolved in 187.5 C.C. of water. The solution (250 c.c.) was prepared a t Oo and maintained a t this point during all the analyses.The following results were obtained with 25 c.c. taken at intervals of fifteen minutes (Expt. 11). TABLE Ir. Time in minutes. 15 30 45 60 75 90 1230 -Cyanic acid present (free and as NH,OCN). Gram. 0.230* 0. i C6 0.165 0.129 0.109 0.101 0.094 0-061 Urea formed (theoretical). Gram. 0.024 0.090 0.094 0.118 Urea found. Gram. 0.018 0.038 0.063 0.081 0.087 0.089 not estimated. -* 0.267 Gram is the theoretical amount from the weight of potassium cyanate taken. The deficiency found a t the outset was due to (a) lose by volatilisakion-the solution had a very pungent odour ; ( b ) time elapsed during addition of nitric wid and impossibility of avoiding t o a small extent the change HOCN + H20 + HNO =NH,NO + CO,.The theoretical values for urea formed were calculated from the amounts of cyanic acid which had disappeared after each interval on the basis 2HNCO -+ CON,H,. It will be seen that under the conditions stated the hydrolysis of cyanic acid was comparatively slow; thus about ninety minutes were required before all free acid had disappeared. At this stage the three primary reactions had been completed hence there was no object in estimating urea formed beyond this point since it could only then arise from the slow transformation of ammonium cyanate as a result of its hydrolytic dissociation thus: NH,OCN = NH3 + HOCN HNCO 1360 WERNER AND FEARON: This is strikingly shown by the last result where in the interval between 90 and 1230 minutes only 0.033 gram of cyanic acid (as ammonium cyanate) had been removed and when this is considered in connexion with the fact that in 75 minutes from the commence-ment of the change 0'313 gram of urea was formed there can be no doubt that the latter can only have been produced as the direct result of reaction (3).Now after 75 minutes when the 'disturb-ing effects of reaction (4) were just noticeable,* half of the solution (125 c.c.) had been used that is 1.15 grams of cyanic acid capable of yielding 0.8 gram of either urea or ammonium cyanate had taken part in the completion of reactions (l) (2) and (3). Hence (0-8-0.313) 0.487 gram of ammonium cyanate was formed which shows that cyanic acid in aqueous solution at Oo had reacted with ammonia as a mixture of HOCN=60-9 per cent.and of HNCO= 39.1 per cent. a result almost identical with that arrived a t from the study of the polymerisation of the anhydrous acid a t the same temperature. As regards the secondary changes the production of biuret con-firms reaction (3) since it must be a sequence of it (compare Werner T. Zoc. cit.; P. 1914 30 262) and whilst its format,ion does not affect the above result as regards the proportion of cyanic acid which reacted as HN:CO its presence was mainly responsible for the low values found for urea formed in the early stages of the reaction. I n agreement with theory the formation of biuret was largely confined to this period. The production of a trace of cyamelide in these esperiments proves that cyanic acid was liberated in the enol-form since the generation of this polymeride takes effect from the change HOCN -+ HNCO.When a solution of ammonia in pure ether a t -80 was gradually added to a similar 'solution of cyanic acid, the crystalline product which immediately separated was found to be a mixture of ammonium cyanate and urea in the proportion of 6 to 2-6 respectively. Note on the XanthhydroZ Test for Urea. The use of xanthhydrol for the detection and estimation of urea depends on the formation of a very sparingly soluble condensation product. Posse (Compt. rend. 1907 145 813; 1913 156 1938) recommends pure acetic acid as the solvent in applying the test. This introduces certain limitations and where the detection of a * For this reason 75 minutes must be taken as the limit up to which the change had proceeded undisturbed in accordance with the primary reactions THE CONSTITUTION OF CARBAMIDES.PART XIII. 1361 trace of urea in a relatively large volume of water is desired the process is tedious. As a matter of fact xanthhydrol is not a test for “free” urea; thus in alcoholic solution no reaction was effected even after heating a t 100. in a sealed tube for several clays. After the addi-tion of one drop of concentrated hydrochloric acid the condensa-tion product was precipitated and its formation was completed within ten minutes. The solubility of dixanthylurea in pure alcohol a t 1 5 O was equal t o 0.009 gram in 100 C.C. A salt of urea must be formed in order to bring about the necessary configuration of the urea molecule before it can react with xanthhydrol and in accordance with this view the test can be applied as follows.A saturated aqueous solution of xanthhydrol (containing 0.13 gram in 1000 C.C. at 15.) is readily prepared by adding the reagent, previously dissolved in 2 C.C. of alcohol to a litre of boiling water. The cold solution (filtered if necessary) when added in consider-able excess (not less than 6 vols. to 1) to an aqueous solution con-taining urea to which a few drops of hydrochloric acid have been added will reveal 1 part in 10,000 within fifteen seconds whilst 1 part of urea in 800,000 can be detected in about ten minutes. The solution of xanthhydrol loses its sensitiveness after about a week on account of gradual oxidation to xanthone.EX P E R I M E N TAL. I Polymerisation of Cyanic Acid.-The results given in table I were obtained as follows cyariic acid was prepared from pure, dry cyanuric acid which was heated electrically in a hard glass tube so arranged that the heating could be continued right up to the neck of the receiver. The construction of the latter was such that liquid cyanic acid collected in the narrow space between an inner and an outer vessel each of which was kept a t Oo. A relatively large surface of the acid was thus maintained a t a constant temperature and by careful avoidance of agitation which was found to be a great promoter of the change polymerisation was allowed to proceed as slowly as possible. The ice in the outer vessel only was displaced by water a t the temperature a t which it was desired to bring about polymerisation; in this way only was it possible to control within reasonable limits the temperature a t which the change took effect.A weighed quantity of the polymeride was extracted with hot water; the cyanuric acid present was estimated by titration with N / 10-so.dium hydroxide using phenolphthalein as indicator. 3 D 1362 FARMER AND INGOLD: Whilst it was necessary to prepare small portions of cyanic acid for each experiment the values given represent the mean of many more experiments than are recorded. 11. Formation of Urea at Oo.-Cyanic acid was estimated by precipitation with an excess of silver nitrate and the silver cyanate dissolved in dilute nitric acid was titrated by Volhard's method. The filtrate freed from the excess of silver was rendered just alkaline by addition of pure lime and after the removal of all traces of ammonia urea was estimated in the residue by ( a ) the '' hypobromite " method ( b ) decomposition by urease and (e) precipitation with xanthhydrol. The presence of biuret was proved by the copper test. Part of the expense of this researoh was defrayed by the Mackinnon Research Studentship of the Royal Society granted to one of us (W.R.F.). UNIVERSITY CHEMICAL LABORATORY, TRINITY COLLEGE DUBLIN. [Received September 22nd 1920.