Of the water, 10 per cent escaped in twenty-four hours over sulphuric acid and only 0.4 per cent more in another like period, but a total of 14.4 per cent after ten days' uninterrupted exposure. The water thus lost is very slightly reabsorbed on exposure to air. From calculations by Prof. F. W. Clarke the substance may be regarded as a mixture of alunogen or the halotrichite group, with salts of the melanterite group, the empirical formula being nearly (Fe, Mn) (Zn, Mg)2Al2 (SO4), . 65 H2O. The outer white zone of the specimen contained only 39.62 per cent of water. The second sulphate examined was a beautiful sky-blue stalactite from the Anaconda mine at Butte, collected by Mr. G. W. Tower. Its composition was as follows: The stalactite was of some size, and was most readily soluble in cold water, the solution giving a strong acid reaction. The calculated ratio is strongly acid, showing either a mixture of highly acid salts or of normal salts with free acids. ON THE CHLORONITRIDES OF PHOSPHORUS AND THE METAPHOSPHIMIC ACIDS. By H. N. STOKES 1. ON TRI- AND TETRAPHOSPHONITRILIC CHLORIDES. "Chlorophosphuret of nitrogen" (Chlorphosphorstickstoff) was discovered in 1832 by Liebig," while attempting to prepare amides of phosphoric acid by acting on phosphorus pentachloride with gaseous ammonia and with ammonium chloride. Analysis led to the formula PN2Cls, and he observed that the compound could be distilled with steam or boiled with acids or alkalies without appreciable decomposition-properties unique in a phosphorus-chlorine compound. 4 5 Liebig did not pursue the subject much further, but at his suggestion Gladstone in 1849 continued the study of the phosphorus-nitrogen compounds and, incidentally, of chloro-phosphuret of nitrogen. In his papers on this body Gladstone detailed a method of preparation from phosphorus pentachloride and ammonium chloride, and pointed out some of its properties-among them its decomposition in aqueous ether or by alcoholic alkalies into hydrochloric acid and a nitrogenous compound, deutazophosphoric acid. Gladstone adopted Liebig's formula, PзNC15. The correctness of this was disputed on theoretical grounds by Laurent and Gerhardt, and the formula PNCI, suggested. At a later date Gladstone and Holmes revised the work of the former and adopted Laurent and Gerhardt's formula, at the same time showing, on the basis of the vapor density, that it must be tripled: P3N3C1. 8 The following work was begun in the laboratory of the School of Pharmacy, Northwestern University, Chicago, and continued in the laboratory of the United States Geological Survey. I wish to express my sincere thanks to Prof. J. H. Long, of the former institution, for his kindness in permitting me to work in his laboratory. 'Liebig-Wöhler, Briefwechsel, Vol. I, p. 63; Ann. Chem. (Liebig), Vol. XI, 1834, p. 146. The part of Wöhler in this investigation seems to have consisted merely in making the analyses for Liebig. 4Quart. Jour. Chem. Soc. London, Vol. II, 1850, p. 121. Ibid., Vol. III, pp. 135, 353; Ann. Chem. (Liebig), Vol. LXXVI, p. 74; Vol. LXXVII, p. 314. * Compt. rend. des Trav. de Chimie, 1851, p. 30; Ann. Chim. Phys. [3], Vol. XVIII, p. 204, note. Later, Wichelhaus' confirmed the results of Gladstone and Holmes, and suggested the structural formula NPC12 Cl2PN-NPC1 and A. W. Hofmann2 and Couldridge3 added some further observations. By distilling under diminished pressure the reaction product of phosphorus pentachloride and ammonia gas Besson obtained a substance of the composition PNC, which he describes as "un corps. solide donnent par sublimation des cristaux très réfringents, fusibles à 1060." The low melting point would indicate a body other than P3N3Cl (melting point 1140), but as the latter is the chief volatile product formed under ordinary circumstances it may be regarded as an open question whether Besson's compound is identical with Liebig's. Finally, White and Gilpin observed it as a secondary product of the action of phosphorus pentachloride on acid ammonium orthosulphobenzoate, doubtless originating in the action of the ammonia. None of these investigations have thrown much light on the constitution of the chlorophosphuret or on its relation to other phosphorus compounds. My attention being called to this substance while engaged in studying the amides of phosphoric acid, I have prepared it in large quantity with a view to further investigation. Although the experiments have not progressed far enough to establish any definite constitutional formula, they show that the chlorophosphuret is by no means as intractable as has been supposed, and that it is one of an homologous series of compounds having the general formula (PNC12), which are the chlorides of an homologous series of acids (PNO2H2), the metaphosphimic acids. From the reaction product of phosphorus pentachloride and ammonium chloride I have thus far isolated the body (PNC12)4, which almost equals (PNCl2); in stability, and which yields, on saponification, an acid (PNO2H2), also an extremely stable substance. A further product of the reaction is a stable, oily chloride (PNC12)x of high, but as yet unknown, molecular weight. I have also obtained from Liebig's chlorophosphuret the corresponding acid (PNO2H2)3, an intermediate chlorhydrine P ̧Ñ¿CO2H2, and a chloramide P3N3C1(NH2)2. The object of the present paper is mainly to describe the preparation and properties of the chlorides, the consideration of the acids being deferred to a later occasion. . Nomenclature.-In view of the large number of phosphorus-nitrogen compounds actually known or theoretically possible, it is desirable to have a more definite nomenclature than has thus far been used. I 'Ber. Deutsch. chem. Gesell., Berlin, Vol. III, 1870, p. 163. 4 Compt. rend., Vol. CXIV, 1892, pp. 1264, 1480. 2Ibid. Vol. XVII, 1884, p. 1910. 5 Dissertation, Baltimore, 1891, p. 13. Dissertation, Baltimore, 1892, pp. 7, 16. Jour. Chem. Soc. London, Vol. LIII, 1888, p. 398. он have proposed' to reserve the name amido-orthophosphoric acid for the ortho acid in which one hydroxyl is replaced by the amido group, and to call the isomeric form P(NH)(OH), orthophosphimic acid. By analogy an acid P(NH). might be named metaphosphimic acid. Corresponding to the polymeric metaphosphoric acids we may imagine an homologous series of metaphosphimic acids, but for each of these several forms would be possible, according as the polymerization is effected by means of oxygen or of nitrogen atoms, assuming them to be otherwise built on the same general type. For example, an acid of the formala P2N2O,H,, with cyclic structure, might have any of the following constitutional formulas, omitting mixed forms: (a) and (b) represent the amide of a dimetaphosphoric acid with its desmotropic form; (c) and (d) represent the two forms of an acid isomeric with the former, but not directly derivable from a dimetaphosphoric acid. In the former class, phosphorus is united by means of oxygen; in the latter, by nitrogen. (b) and (c) are polymers of (PO) (NH) (OH) while (d) is a polymer of an acid N = P(OH)2, which we may call phosphonitrilic acid. Acids of the type (c) I call x-metaphosphimic acids, and those of the type (d), x-phosphonitrilic acids.2 In the chlorides P2N,Cl, and P,N,Cl, phosphorus must be united by means of nitrogen3 (in the absence of oxygen); hence also in all probability in the acids derived from them. The direct replacement of chlorine by hydroxyl would result in phosphonitrilic acids, but as the type (c) is equally plausible, in the absence of definite experimental grounds, I call them, provisionally, metaphosphimic acids. For the chlorides themselves, instead of the old term chlorophosphuret of nitrogen, which is no longer characteristic, the more definite names tri- and tetraphosphonitrilic chloride may be used. EXPERIMENTAL PART. Preparation of the chlorides.-The theoretical yield of triphosphonitrilic chloride is 55.5 per cent of the phosphorus pentachloride; in reality, it falls far below this, no matter what process may be employed, and at best is but a by-product. Gladstone' states the yield as J Am. Chem. Jour., Vol. XVI, p. 124, note. 2 Iz mono-, di-, tri-, etc. 3 Unless we assume P to be joined to P, and N to N, a view which is not supported by their general behavior. Quart. Jour. Chem. Soc. London, Vol. III, p. 135; Ann. Chem. (Liebig), Vol. LXXVI, p. 76. about 6 per cent of the pentachloride (11 per cent of the theoretical), when 1 molecular weight is distilled with 8 molecular weights ammonium chloride. Couldridge' obtained by a similar method a maximum yield of 10 per cent of the pentachloride (18 per cent of the theoretical). As a large excess of ammonium chloride is used, much "phospham" is formed. A modification by Gladstone and Holmes," consisting in acting on the pentachloride with mercuric chloramide, gave no better results. I made numerous experiments in sealed tubes at 200°-260°, using sal ammoniac and pentachloride in theoretical proportions. In this case there is no formation of "phospham," but the yield of the desired body is not increased-in fact, seems to be even less than by distilling with an excess of ammonium chloride from an open retort. What results is a mixture of chloronitrides, largely crystalline, but of various degrees of solubility and stability toward water, of which but a very small portion is volatile with steam. A study of this mixture would, in all probability, lead to the discovery of other members of the series (PNC12) x. I finally adopted a slight modification of Gladstone's original method: A mixture (which need not be very intimate) of 1 part pentachloride and 2 parts dry ammonium chloride is rapidly heated in a tubulated retort (one-third filled) fitted to a receiver containing water, which is connected with one or two Woulff's bottles, with water, to condense the small portions carried over with the escaping hydrochloric acid. The water in the receiver should be gently agitated occasionally in order to break up the crust forming on the surface. When the decomposition is about half finished the heating is interrupted and the hard cake of phospham and sal ammoniac turned over with a rod inserted through the tubulure; this is necessary, as it is so poor a conductor that the bottom of the retort melts before the upper portions of the cake are affected. Heating must not be continued too long, because the subliming ammonium chloride acts on the chlorides condensed in the neck of the retort. After cooling, the substances condensed in the neck are removed by scraping and injecting hot water through the tubulure. The yield varies considerably, according to the quantity of mixture taken; the less the amount, the greater the yield. I have obtained as high as 11.5 per cent. Practically it is not desirable to push this too far, and I have obtained the most satisfactory results by distilling 200 grams of mixture at a time. The proportion of ammonium chloride should not be less than 2 parts; otherwise much pentachloride sublimes unchanged. The distillate, after washing with water, is by no means. pure triphosphonitrilic chloride; it is a mixture of chlorides, of which about one-half is quite stable toward cold water, but decomposed on distilling with steam. During this operation nearly all the triphosphonitrilic chloride, P3N3Cl, is deposited as a hard crust in the con Ibid. [2], Vol. II, p. 227. 1 Jour. Chem. Soc. London, Vol. LIII, p. 399. |