| dc.description.abstract | 3–12. Conclusions – The present work embodies the first systematic investigations of the reactions of octachlorocyclotetraphosphazene with two primary aliphatic amines, viz., ethylamine and t-butylamine. In both cases, in addition to compounds with various degrees of replacement of chlorine atoms, isomeric products have been isolated and characterized. Examples include N?P?Cl?(NHEt)? (mp 116° and 124°), N?P?Cl?(NHBu?)? (mp 158°) and N?P?Cl?(NHBu?)? (mp 127° and 171°).
Some earlier observations and data in the literature have been clarified. For instance, the bis-t-butylamino derivatives reported by two groups of workers are proved to be not the same compound but two distinct bis?isomers. Structures have been proposed for most of the compounds prepared on the basis of infrared, ¹H and ³¹P spectroscopic data.
One important feature of the present work is that it is not necessary to employ anhydrous ethylamine for the preparation of the partially substituted compounds and the octakis?compound. The use of aqueous amine is quite satisfactory and in many cases gives better yields than obtained by using anhydrous ethylamine. The reactions of ethylamine with the tetrameric chloride (I) give only nongeminal products and although the octakis?compound can be prepared easily, no authentic penta?, hexa? or hepta?substituted compounds could be isolated.
The reactions of t-butylamine with the tetrameric chloride, N?P?Cl? (I), follow a nongeminal pathway up to the tris?stage of chlorine replacement. It is reported that the reactions of t-butylamine with the trimeric chloride, N?P?Cl?, give exclusively geminal products; formation of nongeminal compounds in the case of the tetrameric system is attributed to the greater reactivity of the tetrameric chloride (I). The complete replacement of chlorine atoms by t-butylamino groups in the tetrameric chloride requires milder conditions than the analogous reactions of the trimeric chloride. The present method developed for the preparation of octakis?t-butylamino compound (XXV) is a distinct improvement over the earlier method reported.
In both the reactions of ethylamine and t-butylamine with the tetrameric chloride (I), the yields of the expected phosphazene derivatives, N?P?Cl???(NHR')? (R' = Et or Bu?), are only modest. A considerable amount of resinous material is formed, particularly when the tetrameric chloride to amine stoichiometry exceeds 1:8. The insolubility of the resinous material in the usual organic solvents precluded the detailed study necessary to characterize it adequately. However, infrared and NMR evidence suggests that the resin contains cross?linked tetrameric units. A possible reason for the formation of the resin is discussed.
There is pronounced solvent effect in the reactions of the tetrameric chloride even at the bis?stage of chlorine replacement. In the reactions of the trimeric chloride with secondary amines, the solvent effects become detectable at the tris?stage of chlorine replacement. In the case of the tetrameric chloride, the solvent not only alters the relative proportion of one compound over another but sometimes gives an entirely different set of compounds. The role of solvents in the reactions of halogenocyclophosphazenes with nitrogenous bases is poorly understood as yet. Further experimental work is necessary on selected systems before a satisfactory theory can be evolved.
The most significant highlight of the present work has been the isolation of bicyclic phosphazene compounds possessing a P–N(Et)–P bridge in the reactions of the tetrameric chloride (I) with ethylamine or of 2?trans?6?bisethylamino isomer, N?P?Cl?(NHEt)?, with dimethylamine in chloroform. This is the first time the formation of these bicyclic compounds has been observed in the aminolysis reactions of chlorocyclophosphazenes as a result of intramolecular trans?annular nucleophilic substitution. A tentative mechanism has been advanced to explain these observations. The IR and NMR data for these compounds are discussed in detail. The discovery of the reaction leading to the formation of bicyclic compounds opens up a new area of investigation in cyclophosphazene chemistry. The generality of the reactions with other primary amine systems, the factors that influence the reaction (solvent, presence of tertiary base in the reaction medium, presence of other substituents in the ring, etc.) and the mechanistic details remain to be explored.
Another important aspect of the present work is the isolation and characterization of a number of octakisaminocyclotetraphosphazene hydrochloride adducts which have not been reported so far in the literature. They include N?P?(NHBu?)?·HCl, N?P?Cl?(NHEt)?·HCl (from two different bis?isomers), and N?P?(NMe?)?(NHEt)?·2HCl. The infrared and ¹H NMR spectroscopic data suggest protonation at a ring nitrogen atom; a variable?temperature ¹H NMR study would be expected to throw light on any exchange processes occurring in these systems. Also, a detailed X?ray crystallographic study of the hydrochloride adducts and the corresponding free bases would conclusively prove the site of protonation and the associated changes in the skeletal bonding.
The products obtained from the reactions of monoethylamino?, N?P?Cl?(NHEt), and mono?t?butylamino?, N?P?Cl?(NHBu?), compounds with two equivalents of t-butylamine and ethylamine respectively show that the attacking amine appears to be more important than the amino substituent in determining the course of the replacement reactions of the tetrameric chloride (I). A detailed investigation of several mixed amine systems is necessary before definitive conclusions can be drawn about the dominant role of the nucleophile in determining the nature of products formed. Kinetic studies would be particularly valuable in understanding the mechanism(s) of these reactions.
The preparation of mixed dimethylamino derivatives of many cyclotriphosphazene derivatives has previously simplified the structural assignment of the precursors. This technique is not very informative when applied to the ethylamino? and t-butylamino derivatives of the tetrameric chloride. In many cases, owing to the similarity of chemical shifts and the presence of strong "virtual coupling" the expected number of dimethylamino and methoxy environments are not resolved in the proton NMR spectra of these derivatives. On the other hand, ³¹P NMR spectroscopic data appear to be more useful for the structural assignments of cyclotetraphosphazene derivatives. | |
