THESTRUCTURE OF THE TRIMER OF PHOSPHONITRILE CHLORIDE
.lug., 19-13 ICUSrRrBL T I O N
F R O X TAS: CHEMICAI.
IJABDRATORY O F THE UNIVERSITY O F MICHIGANAND T O R Y O F HARVARD ITSIVERSITY]
THE &f.ALLINCKRODT
1551 LABORA-
The Structure of the Trimer of Phosphonitrile Chloride, PJN,CI, BY L. 0. BROCKW.\Y AND WILLARD M. BRIGHT' This electron diffraction study was undertaken in order to elucidate the structure of the trimer of phosphonitrile chloride, P3K3C16. A reaction between phosphorus pentachloride and ammonia producing the phosphonitrile chloride series of compounds was discovered over one hundred years ago.? Since that time several other methods of preparation have been reported, and nu merous investigations have been made of the physical and chemical properties of the resulting product^.^ The favored method of preparation involves the use of ammonium chloride as the ammonolytic reagent, and results in a series of polymers having the formula (PNC12),, where x may be 3, 4, 3, etc., reaching a hundred or more. The lower molecular weight compounds are believed to be cyclic while the higher ones, which may be rubber-like in character, have been shown to be long chains.' The trimer is the most abundant product of the reaction, forming nearly 50'7; of the mixture, although the amount depends on the precise method of preparation. The determination of the atomic arrangement and bond distances existing in this molecule was executed, and values are reported for the interatomic distances and bond angles. Our results are compared with those found by X-ray and crystallographic analysis of the trimer and tetramer.5 The structures proposed by the various investigators are discussed I
Experimental The phosphonitrile chloride was prepared using essentially the method of Schenck and Romer,6 which seemed to be the most convenient laboratory method. However, some slight changes were introduced into their procedure. T n e reaction proceeds according to the equation PCl:, I\:H,Cl ---f PNCl? f 5HC1 and is carried out in the presence of sym-tet rachloroet hatie. Four hundred and fifty grams of phosphorus penlachloride was added t o a reaction flask contaixiing 500 ml. of sym-tetrachloroethane (b. p . 144-145"j and 130 g .
+
~
. .
( I ) Present address: T h e Kendall Company, Boston, Mass (2) Liebig and Wohler, Arrn., 11, 139 (1834). (3) Audrieth, Steinman a n d Toy, C h o n . Rev., 32, 109 (1943): an excellent review of the earlier work on these compounds. (4) Meyer, Lotmar and Pankow, Helv. C h i n . Acto, 19,930 (1936). ( 3 ) (a) Jaeger a n d Bentema, Proc. Acad. Scz. Amsierdatn. 36, 766 (1932); (b) Renaud, A n n . chim., (11) 3, 445 (1935); ( c ) Ketelaar and Rec. Irau. chim., 68, 1081 (19RY). i F , ) S c h e n c k and R o m e r , &r 69. 1343 (1024')
devries.
(16% excess) of finely ground, dry ammonium chloride. The phosphorus pentachloride was not all dissolved at the beginning of the reaction, but this offered no difficulty, as the reaction product is extremely soluble. N o provision was made for stirring the reactants since the evolution of hydrogen chloride was found to give sufficient agitation. The mixture was heated under reflux for eighteen hours with the condenser connected to a n "Anhydrone" drying tube and thence to a hydrogen chloride absorbing system. At the end of this time the solution was allowed to cool, arid the excess ammonium chloride filtered off. The clear yellowish solution wai: then vacuum distilled at 20 mm. in order to remove the solvent. Both this distillation and the previous heating were carried out in a n all-glass system. After allowing the distillation residue to cool overnight, a yellowish-brown slush was obtained, which consisted of crystalline phosphonitrile chloride trimer and tetramer and more highly-polyinerized material in the form of an oil. The crystals were separated by filtering the mixture through a light wad of glass-wool, Approximately 115 g. of the crude crystal3 and 107 g. of oil were obtained in this manner. .sy~/2-Dichloroethanewas found to possess a satisfactory temperature coefficient of solubility, and the crude crystals were recrystallized three times from this solvent. A 35% yield of pure white material was obtained.: Pure trimer was separated from the mixed crystals by vacuum distillation of 50 g. of the product at 11 mm. pressure. An all-glass system made for the vacuuni distil'ation of solids was used, and 25 g. of material boiling a t 116-118' (11 rnm.) was collected. Upon redistillation 18 g. of the trimer was secured with the b. p . 116-117". The molecular weight was determined by measuring the depression of the freezing point in benzene and found
