Synthesis and Stabilty of Macrobicyclic Triphosphazides Substituted on its Bridgehead Carbon

José Berná, Mateo Alajarín, Carmen López-Leonardo

Departamento de Química Orgánica, Facultad de Química,

Universidad de Murcia, Campus de Espinardo, E-30100, MURCIA, SPAIN

 Introduction

Several synthetic strategies may be devised for the construction of macrobicyclic systems. The more direct one is tripod - tripod coupling, a molecular self-assembly process that requires the formation of three bonds in a single step.^[1] A major drawback of tripod - tripod coupling is the occurrence of extensive side reactions which minimize the yield of the expected bicyclic product. Only in limited cases^[2] have such processes been carried out in synthetically useful yields, provided that a fine tuning of reagents, reactions, and conditions could be achieved.

We have described^[3] the preparation and characterization of a number of macrobicyclic triphosphazides 1 by coupling of two tripodal subunits, aromatic triazides and 1,1,1-tris[(diphenylphosphino)methyl]ethane (triphos), via triple P-N bond formation be means of the Staudinger imination reaction.^[4]

Those C₃-symmetric cage compounds were shown to possess three intracyclic PN₃ units of Z configuration, propeller-like topology and the X (lp or O) and CH₃ groups on the bridgehead atoms pointing outward of the bicyclic cavity.

The stability in solution of these species was evaluated and found to be related to the following structural factors:

* the quaternization of the pivotal nitrogen atom, in the form of N-oxide, increased their stability;

* the presence of substituents flanking the N termini of the phosphazide functions decreased their stability (Scheme 1, e.g. if R², R⁴ or R⁶ are alkyl groups).

As a further step in our study we have now considered the coupling of tris(2-azidobenzyl)amine (2) and its N-oxide (3) with different 1,1,1-tris[(diphenylphosphino)methyl]methanes substituted at the pivotal carbon, under similar conditions. We were interested in the stability of the new compounds and the conformational changes that could be induced by the introduction of different pivotal groups at the bridgehead carbon.

 Results and discussion

The tripodal coupling of triazides^[3] 2 and 3 with a number of the triphosphanes^[5] 5 was carried out in diethyl ether solution at room temperature.^[6] The resulting macrobicyclic triphosphazides 6 precipitated from the reaction medium as yellow solids.

Scheme 2

Table. Synthesis of macrobicyclic triphosphazides 6.

The structure determination of compounds 6 was accomplished by means of their analytical and spectral data. These data were essentially coincident with those of the previously reported 1a and 1b.^[7] For this reason, tridimensional arrangements of the new compounds 6a, 6b, 6d and 6l here prepared were assumed to be similar to that of 1b, which has been unequivocally determined by X-ray analysis.

The entry 10 of the Table merits special consideration: the triphosphazide 6h derived from N-oxide 3 and the phosphane 5e has been obtained as a mixture of two rotamers in 1.7:1 ratio. The NMR spectra of this mixture show two sets of signals corresponding to two C₃-symmetric species (Figure 1). The methylene protons (CH₂N and CH₂P) and the methyl groups of the isopropyl unit are magnetically inequivalent in the ¹H  NMR spectra, as a consequence of their diastereotopic nature and accounting for the chirality of these propeller-shaped isomers.

Figure 1. ¹H NMR spectra (range: + 5 to - 1 ppm) of 6h (300 MHz, CDCl₃, 25 ºC). The signals of 6h’ are marked with a square figure (■) and the signals of 6h’’ with a circle figure (●).

We have proposed for these isomer two bicyclic structures which differs in the helical sense of their tripodal subunits, tribenzylamine (upper moiety) and neopentylic fragment (lower moiety) (Figure 2).

I                                                                                                    II

Figure 2. Proposed rotamers I and II for the isomers 6h viewed along the theefold axis. The lower tripodal moiety is in purple (except the two methyl carbons of the isopropyl group) .

The calculated^[8] energy difference between both isomers of 6h is 17.31 kcal·mol^-1 in favor of the structure I that has been assigned to 6h’, the major component of the mixture. Indeed, spectral data of the other N-oxides 6a, 6b, 6d, 6l and 1b are quite similar to those attributed to compound 6h. Then, the minor component of the mixture, 6h’’, would possess the structure II, in which the helical sense of the tribenzylamine fragment is contrary to that of the lower tripodal moiety.^[9]

            Unfortunately, the low stability (lability) of the triphosphazide 6h in solution limited the acquisition of enough solution NMR data to confirm the proposed structures.

On the other hand, we envisaged that the unsuccessful coupling of tris(2-azidobenzyl)amine and its N-oxide and triphosphanes 5d (R = PhCH₂) and 5f [R = (CH₃)₃C], bearing bulky substituents at the pivotal carbon atom, may be a consequence either:

a) of a decrease in the population of the optimal reactive conformer for the coupling process.

or/and

b) of the fast decomposition of the resultant macrobicyclic triphosphazide. This unstability would be originated by the steric repulsion between the pivotal group and the three proximal pseudoaxial phenyl groups. These interactions could distort the stabilizing planar conformation of the phosphazide groups. Moreover, this situation might cause the dissociation of the phosphazide groups into their primitive phosphane and azide constituents.

However, in some cases this unstability can be reduced if the bridgehead nitrogen atom is quaternizated in its N-oxide form. This fact allowed the isolation of the byciclic compounds 6d, 6h and 6l with a pivotal ethyl, isopropyl or phenyl group, respectively.

 Conclusions

The size of the group at the pivotal carbon of the triphosphane is essential for determining the preparation and stability of the bicyclic triphosphazide. In this sense, only the triphosphanes with an hydrogen or methyl group at the pivotal carbon allow the preparation of triphosphazides, independently of the X group (lone pair or oxygen) attached to bridgehead nitrogen. To obtain triphosphazides with a pivotal ethyl, isopropyl or phenyl group is necessary that the bridgehead nitrogen was quaternizated as its N-oxide form. Finally, the present methodology is not suitable for the synthesis of triphosphazides with a pivotal benzyl or tert-butyl group.

 Acknowledgements

This work was supported by MCYT and FEDER (Proyect BQU2001-0010) and Fundación Séneca-CARM (Proyect PI-1/00749/FS/01). José Berná also thanks the MEC for a fellowship.

 References

[1] B. Dietrich, M. W. Hosseini, J. M. Lehn, R. B. Sessions, Helv. Chim. Acta 1985, 68, 289.

[2] F. Vögtle, R. G. Lichtenthaler, Angew. Chem. Int. Ed. Engl. 1972, 11, 535; R. G. Lichtenthaler, F. Vögtle, Chem. Ber. 1973, 106, 1319; F. Vögtle, P. Neumann, J. Chem. Soc. Chem. Commun. 1970, 1464; P. L. Anelli, F. Montanari, S. Quici, J. Chem. Soc. Chem. Commun. 1983, 194; M. Kanishi, J. I. Kunizaki, J. Inanaga, M. Yamaguchi, Bull. Chem. Soc. Jpn. 1981, 54, 3828; F. Vögtle, G. Hohner, Angew. Chem. Int. Ed. Engl. 1975, 14, 497; C. Bolm, K. B. Sharpless, Tetrahedron Lett. 1988, 29, 5101; M. Nakazaki, K. Yamamoto, T. Toya, J. Org. Chem. 1980, 45, 2553; P. Osvath, A. M. Sargeson, J. Chem. Soc. Chem. Commun. 1993, 41.

[3] M. Alajarín, A. López-Lázaro, A. Vidal, J. Berná, Chem. Eur. J. 1998, 4, 2558.

[4] H. Staudinger, J. Meyer, Helv. Chim. Acta 1919, 2, 635.

[5] The starting triphosphanes were prepared following previously described methodologies in J. Berná. Thesis Dissertation. University of Murcia, 2003 and references therein.

[6] General procedure for the preparation of the triphosphazides 6: Two solutions of the corresponding tris(2-azidobenzyl)amine (1.5 mmol) in diethyl ether or dichloromethane (10 mL) and the corresponding tris[(diphenylphosphino)methyl]methane (1.5 mmol) in diethyl ether (10 mL) were simultaneously added to a round-bottom flask containing diethyl ether (15 mL) under nitrogen at room temperature over a period of 30 min with stirring. The resulting mixture was then stirred for 2 h. The precipitated pale yellow solid was filtered, washed with diethyl ether (3 x 10 mL), and dried under vacuum.

[7]  M. Alajarín, P. Molina, A. López-Lázaro, C. Foces-Foces, Angew. Chem. Int. Ed. Engl. 1997, 36, 67.

[8] Molecular Mechanics calculations were carried out by using the MM+ force field included in the Hyperchem^® v. 6.01 program.

[9] Macrobicyclic systems with tripodal moieties having opposite helical sense: I. Bkouche-Waksman, J. Guilhem, C. Pascard, B. Alpha, R. Deschenaux, J. M. Lehn, Helv. Chim. Acta 1991, 74, 1163; I. M. Atkinson, D. C. R. Hockless, L. F. Lindoy, O. A. Matthews, G. V. Meehan, B. W. Skelton, A. H. White, Aust. J. Chem. 2000, 53, 799; c) X. Hu, Y. Tang, P. Gantzel, K. Meyer, Organometallics 2003, 22, 612.