398 COULDRIDGE SOME INTERAUTIOXS OF XXX.- Some Interactions of Nitrogen Chlorophosphide. By WARD CODLDRIDGE B.A. CHLOROPHOSPHIDE of nitrogen was carefully and at length examined by Gladstone a t the request of its discoverer Liebig ; and about the same time Gerhardt examined the interaction in which it is formed. Since that date the only work which has been done with it excepting Wichelhaus' confirmatory determination of its high vapour-density is that of Hofmann who stxidied the interaction of aniline and chloro-phosphide of nitrogen. My object in resuming the investigation of chlorophosphide of nitrogen was i f possible to elucidate its forma-tion and to examine in more detail its reaction with amines. As my work proceeded other issues suggested themselves. I endeavoured to displace the chlorine-atomg by cyanogen-groups ; and as the effort NITROGEN CHLOROPHOSPHIDE.399 which had been made to remove by the action of sodium and potassium. the chlorine and to isolate the radicle of phosphorus and nitrogen had not given very definite results I tried the action of zinc ethide expecting by this means to displace the chlorine by ethyl, and thus to form compounds of which the reduction products would have been interesting. But to work with this substance one has to meet the difficulty of its preparation and as Gladstone found the yield is but small. My method of preparation was a slight modification of that given by Gladstone (Jour. Chem. Xoc. 3 135). Instead of attaching receivers directly to the flask in which the ammonium chloride and the phosphorus pentachloride are heated by a Bunsen burner a straight condenser was first fixed to the flask in order to return to the sphere of action any unaltered vaporised pentachloride and to condense the chlorophosphide of nitrogen which would otherwise have been carried over by the evolved hydrogen chloride and would, in part have been lost.The chlorophosphide of nitrogen was purified by driving over with steam. The objection to this method of purifica-tion is that the steam decomposes some of the product; but i n its favour is the fact that it is less trouble than t o extract with anhydyous ether. The yield was somewhat variable; but in no case was it large. The maximum yields I obtained amounted to 10 grams of chlorophosphide of nitrogen from 100 grams of phosphorus penta-chloride and 200 grams of ammonium chloride; whereas the theoretical yield calculated from the equation-3PC1 + 3NH3 = P,N,Cltj + 9HC1, would be 41 grams.Gladstone states that his yield was uniformly about 6 per cent. of the pentachloride used. The smallness of the yield is most probably explained by the following experiments. I found that when dry ammonia gas is passed through a dry tube over melted chlorophospliide of nitrogen a reac-tion takes place which results in the transformation of the greater part of the latter substance into a greyish-white infusible powder, and in the volatilisation of the smaller portion on to the cooler part of the tube where it remains unaltered; this same greyish-white infusible substance is also obtained when chlorophosphide of nitrogen and ammonium chloride are heated in a sealed tube at a temperature of 150" for an hour or more; at the end of which time mere traces only of chlorophosphide of nitrogen remain.The fol-lowing equation represents the reaction :-P,N,CI,j + 3NH3 = P,N?(NH)3 + 6HC1. P hospharn 400 COULDRIDGE SOME ISTERXCTIOSS O F The greyish-white infusible substance is readily characterised as phospham. This easy transformat ion of chlorophosphide of nitrogen into phospham renders the view distinct that the molecule of phospham is not PN(NH) but is P3N3(NH), just as chlorophosphide of nitrogen is not PNC1 but is P,N,CI,. The production in the first instance of chlorophosphide of nitrogen receives a simple explanation in terrns of the following equation :-C1 H Cl H 3C1,PCl + 3HN = P,N,Cl + 9HC1.Gerhardt (Ann. Clzim. Phys. 18,205) who investigated the action of phosphorus pentachloride on ammonium chloride concluded that the fortnrttion of phosphorus nitrogen chloride,* occurring as it does in small quantities was accidental. But the above-mentioned action of ammonia would lead to another conclusion. And indeed not only does ammonia react with phosphorus nitrogen chloride but also amines generally appear to remove the chlorine. Hofmann (Bey. 17, 1909) found that by dissolving phosphorus nitrogen chloride in aniline, and warming the solution on a water-bath the whole solidified ; and from the solid mass by extracting the aniline hydrochloride and the unaltered aniline by hydrochloric acid and water he obtained a white, solid substance which after crystallisation from glacial acetic acid, gave on analysis numbers corresponding to the constitution I repeated this experiment using somewhat larger quantities of substances.There was formed the white substance which Hofmann described and together with it a viscid dark-coloured oil which was separated by digestion with cold alcohol after treating the solid mass with hydrochloric acid and water. This oily substance is extremely soluble in alcohol ether and benzene and appears to be unaltered by boiling with a large quantity of water for 10 hours. All attempts to crjstallise it were futile ; and after standing for two months it shows no signs of crystallisation. This resinous matter was formed in every case from the action of phosphorus nitrogen chloride on the amines which I used.Hofmann considered that if this compound P,N,( NH.C,H,), were heated with hydrochloric acid it might lose 3 mols. of aniline and be converted into the corresponding phospham-derivative P3N3( NCsH,),. But 1 found that when heated in a sealed tube with strong hydro-chloric acid it remained unaltered a t a temperature of 150" and f Tbis name seemR preferable to " chlorophosphide of nitrogen. NITROGEN CHLOROPHOSPHIDE. 401 when the temperature was I-aised to 250" it was completely deconi-posed into phosphoric acid ammonium chloride and aniline. Thus-P3N,(NH*C,H,) + 12H20 + 9HC1 = 3NHAC1 + 6C6H5*NH2,HCl + 3H3POa. I next used instead of aniline orthotoluidine. On mixing together tlie phosphorus nitrogen chloride and the orthotoluidine the mixture gets warm; and on heating on the water-bath it finally solidifies.Instead of a t first treating the solid mass with hydrochloric acid I used ether to remove the excess of orthotoluidine fearing lest the hydrochloric acid might be the cause of the production of the resinous substance. After filtering off the ethereal extract a white solid re-mained but its ready solubility in water proved it to be orthotoluidine hydrochloride. Theethereal extract gave a residue after thr removal of the free orthotoluidine of a viscid oil similar to that obtained when aniline was used. Thus a t loo" no product of the type P,N,R was formed. But when the orthotoluidine and the phosphorus nitrogen chloride were heated a t temperatures of 150" or 250' such a substance was produced together with the resin; so that on extracting the solid mass with ether and with hydrochloric acid and water a white substance remained which could be crgstallised in slender needles from hot alcohol.I t s melting point is 241-242'. The formula P3N3(NH*C6H,.CH3)6 requires 12.17 per cent. of nitrogen ; 1 found 12.16. When phenylhydrazine was mod there was at a temperature of loo" formed in addition to the viscid oil a crystallisable substance which I extracted as in the previous instance. It was crysrallised from alcohol. Its melting point was 200". The constitution, P,N3(NH*NH.C,H,)6 requires 11.97 per cent. of phosphoiws and 27.03 per cent. of nitrogen ; my analysis gave 12-15 per cent.of phos-phorus and of nitrogen 27.27. The reaction with piperidine and phosphorus nitrogen chloride is very rapid. So much heat is evolved that the piperidine begins to boil. The reaction is completed by warming on a water-bath. The products are here again a viscid oil in much smaller quantity than in the other instances and a crystallisable componnd. The latter can with difficiilty be crystallised from alcohol. On heating it decom-posed before me1 ting. The formula P3N,(NC5H,,) would require-Pound. 14.3 Phosphorus . . . . . . . 14-08 per cent. Nitrogen. . . . . . . . . . 19.89 , 19.8 Diphenylamine and diamylamine do not at ordinary temperatures react with phosphorus nitrogen chloride 402 STUART AC'I'IOX OF Attempts were made t o prepare cyanogen-derivatives of phosphorus nitrogen chloride by dissolving it in alcohol and heating the solution with silver cyanide a t a 100" for several hours.But the phosphorus nitrogen chloride was entirely decomposed and prussic acid was set free. On evaporating the alcohol a thick oil with a strongly acid reaction was obtained. It readily gave off ammonia when warmed with baryta-water and otherwise reacted as a mixture of acid am-monium salts. It showed Gladstone's test for azophosphoric acid ; for on adding a drop of ferric chloride to the ammoniacal solution of the acid the liquid became brown and no trace of precipitate was formed. But on applying this test to a solution of ammonium phos-phate I obta.ined exactly the same result. Previous efforts which had been made to remove the chlorine and t o isolate the radicle had not yielded very definite results. I found that sodium or sodium amalgam acts very slowly on a benzene solu-tion of phosphoric nitrogen chloride ; so I mixed it with dry zinc-dust exhausted the tube and heated the mixture. Much gas was evolved smelling of cyanogen. The phosphorus nitrogen chloride was entirely decomposed and the phosphorus remained behind in combination with the zinc. I next tried the action of zinc ethide on the phosphorus nitrogen chloride ; but at ordinary temperatures the substances do not react. The University Laboratory, Ca 112 bridge