Rocket science in the making

Confirmation of the synthesis of the elusive pentazole molecule, of interest for its possible role in future rocket propellants, has come almost a decade after it was made, writes Richard Butler

Azole molecules are enormously important throughout organic and biological chemistry and materials science. Pyrrole is the parent molecule of this series of aromatic planar pentagonal fivemembered rings with nitrogen and carbon atoms. However, the highest member – a planar pentagon of five N-atoms (HN₅), pentazole – has long been sought experimentally and has proved an attractive target for many theoretical chemists.¹

Many questions arise about this enigmatic molecule. Can it exist? Is it the final member of the azole series with no carbon? Or should it be classified in the realm of inorganic chemistry?

Notwithstanding this intrinsic interest, high nitrogen molecules – with three or more linked nitrogen atoms – also have potential as high energy sources and they attract particular interest for practical uses as fuels or fuel additives. The higher oxides dinitrogen trioxide N₂O₃, dinitrogen tetroxide N₂O₄ and dinitrogen pentoxide N₂O₅, are fairly unstable and explosive, a consequence of the chemical stability of their constituent nitrogen bonds and the high oxidation state of the nitrogen. N₂O₄ is one of the most important oxidisers of rocket fuels, used to oxidise hydrazine (N₂H₄) in the Titan rocket and in the recent NASA Messenger probe to Mercury.

As a probable source of pentazole anion, N₅-, the synthesis of HN₅ was of huge importance as researchers speculated that if the anion were to be combined with its cation N₅ ⁺, the result would be a stable nitrogen fuel that could deliver a substantially better propulsion power than the same quantity of hydrazine.

Quest for the ring
The possibility of the molecule HN₅ has long been considered by organic chemists and in 1915 the first speculative – but incorrect – claim of a derivative of such a ring, R-N₅, where R = CH₂- (C=NH)-NHNH₂, was made. In the mid-1950s, by the combined work of Rolf Huisgen and Ivar Ugi in Munich and Klaus Clusius and Hans Hürzeler in Zurich, the existence of the ring was proved.²

It was established as an intermediate in the reactions of aryl diazonium salts, Ar-N₂ ⁺.X–, with azide ion (N₃ ^-), at 0^oC, by ¹⁵N isotopic labeling experiments. A number of aryl derivatives were subsequently isolated and a low temperature X-ray crystal structure of one was obtained.3 These aryl pentazole derivatives are unstable and evolve N₂ gas, changing to aryl azides even at low temperatures in the range 0 to –10^oC. Since this early work there has been considerable interest in HN₅ and a number of attempts were made to remove the aryl group from Ar-N₅, the first by Ugi in 1961.⁴

However, searches for HN₅ became more vigorous after the discovery of N₅ ⁺, an acyclic chain of five N-atoms as a cation in inorganic salts, was announced by Karl Christe et al in 1999.⁵ The possibility of obtaining a material that combines N₅ ⁺ with the then unknown anion N₅ ^– became a realistic target. By 2002, two groups^6,7 had independently detected N₅ – in the gas phase by bombarding aryl pentazole molecules with high energy particles in high energy mass spectrometric degradation experiments.

My own group’s interest in pentazole chemistry, at the National University of Ireland in Galway, was sparked much earlier, in the mid-1990s, following a request from the late Charles Rees, Hoffmann professor of organic chemistry at Imperial College London, UK, for us to contribute a chapter on pentazole chemistry for a book on Comprehensive Heterocyclic Chemistry II.³ During work for this chapter we identified significant challenges in the field which we subsequently went on to address, but none was more demanding than HN₅. In subsequent years Rees took a keen interest in the work that developed from this chapter and we often discussed it during our friendship. In 1996 my research group reported a variable temperature proton NMR kinetic study of the mechanism of degradation of aryl pentazoles to N₂(g) and aryl azides. In 1998 we reported the mechanism of the formation of the aryl pentazoles as intermediates in the normal reactions of aryl diazonium salts with azide anion by using proton and ¹⁵N NMR to directly monitor the reaction at –80 to –85^oC. The aryl pentazoles arose from cyclisations in fleeting aryl pentazene intermediates, acyclic chains of five N-atoms, Ar-N=N-N₃ (Scheme 1). Also the protonation sites and very low basicities of aryl pentazoles were established from low temperature ¹⁵N NMR work.

This work⁸ gave us considerable experience of the low temperature techniques and the experimental difficulties involved in working with these unstable and sensitive high nitrogen compounds. Unsuccessful direct attempts to remove the aryl group from aryl pentazoles led us eventually to seek a general N-dearylation reaction for lower azoles and we focused on cerium(IV) ammonium nitrate, Ce(NH₄)₂(NO₃)₆ (CAN). We found that this reagent could remove a p-anisyl group, p-MeO-C₆H₄ –, from the N atom of an azole molecule. This work effectively rendered the p-anisyl group as a useful azole NH protecting group since the p-anisyl substituent can be readily restored as well as removed.⁸

Under the appropriate conditions, the p-anisyl group comes off as a p-benzoquinone molecule (2) (Scheme 2) and the remaining azole compound could be readily isolated. We tested this reaction on a wide range of different N-(p-anisyl)-azoles, at temperatures down to –40^oC and developed an understanding of the scope and limitations of the process. With proof of principle established together with my PhD students Mark Hanniffy and John Stephens, we applied this reaction to N-p-anisylpentazole at –40^oC by 2001 with results similar to the other azoles, particularly pyrazoles and tetrazoles, and later 1,2,3-triazoles.

These results gave a strong indication that HN₅ had been generated in the solutions at –40^oC, and this was indeed the case. It survived for a short time, probably about a minute,⁹ before degrading to nitrogen gas and azide anion, N₃ –. Strategic synthetic placement of initially one – and later two and three – ¹⁵N isotopic labeling atoms in the starting pentazole (1) showed that the labeled isotopic N atoms had been scrambled in the symmetrical tautomeric HN₅ and N₅ – species in the solutions and the labels appeared at all positions in the azide anion fragment (3).

The positions of the labeled ¹⁵N atoms were followed through the reactions by low temperature ¹⁵N NMR spectra. When more than one ¹⁵N isotope atom was present, the splitting patterns between the adjacent ¹⁵N atoms in the NMR spectra of the fragments confirmed that the observed signals were indeed from the azide fragments of HN₅/N₅ –.

Proof of synthesis
This paper on our synthesis of HN₅ appeared in 2003. But despite our own convictions, there were some chemists who questioned whether what we had actually produced was in fact NO₃ –. Earlier this year, the controversy has finally been laid to rest with the publication of a paper by Rod Bartlett and coworkers at the University of Florida, US,¹⁰ after comparisons of our experimental splitting constants, J values, were compared with predicted values. These were the first calculated splitting constants between two bonded ¹⁵N atoms and were determined by the Bartlett-Perera group using coupled-cluster theory. The agreement between the experimental and theoretical values was excellent.

There is no doubt that if a new allotrope of nitrogen, N₅ ⁺.N₅ ^– – overall N₁₀ – can be obtained it will be a remarkable substance. The species N₅ ⁺. N₅ – contains both an oxidiser and a reducer and the transfer of one electron should trigger the release of its energy to give five molecules of nitrogen gas with no pollution other than production of air, minus the oxygen. For its use as a rocket propellant taming it would probably be a challenge, but stabilisers for other propellants are well established. The technical challenge of producing significant quantities of N₅ – would undoubtedly prove difficult and daunting. On this 40th anniversary of the first Moon landing, however, even the possibility of a new powerful nonpolluting rocket propellant based on a new form of the element nitrogen alone seems particularly apt.

Acknowledgement: Dedicated to the memory of Charles W. Rees FRS.

Richard Butler is emeritus professor at the school of chemistry, National University of Ireland, Galway.

References

1. Bartlett, R.J., Chem. Ind., 2000, 4, 140.
2. Huisgen, R. and Ugi, I., Angew.Chem., 1956, 68, 705; Clusius, K. and Hürzeler, H., Helv.Chim.Acta, 1954, 37, 798.
3. Butler, R. N. in Comprehensive Heterocyclic Chemistry II, 1996, Eds. Katritzky, A. R., Rees, C. W. and Scriven, E. F., Elsevier Science, New York, vol. 4, (Ed., Storr, R. C.), 897-904.
4. Ugi, I., Angew.Chem., 1961, 73, 172.
5. Christe, K. O. et al, Angew.Chem.Int.Ed., 1999, 38, 897.
6. Ostmark, H. et al, Chem.Phys.Lett., 2003, 379, 539.
7. Vij, A et al, Angew.Chem.Int.ed., 2002, 41 , 3051.
8. Butler, R. N. et al, J. Org. Chem,, 2008, 73, 1354; Butler, R. N., Stephens, J. C. and Burke, L. A., Chem. Comm., 2003, 1016.
9. da Silva, G. and Bozzelli, J. W., J. Org. Chem., 2008, 73, 1343.
10. Perera, S. A., Gregusová, A. and Bartlett, R. J. J.Phys. Chem., 2009, 113, 3197.