Möbius Conformations of Hetero[12]annulenes

David Hall, Henry S. Rzepa^* and Nikola Sanderson

Department of Chemistry, Imperial College London, SW7 2AY

Summary: Conformations of the mono, di, tri and tetra hetero [12] annulenes which have C₂ or pseudo C₂ symmetry are explored using ab initio theory for their potential Möbius aromatic properties, as quantified using the nucleus independent chemical shift (NICS) values at the ring centroid. This is found to be greatest for the sulphur heterocycles, and a number of known examples of tri and tetrahetero systems show C₂ symmetric conformations for the crystal structures and modestly aromatic NICS values. The structures of several other annulenes show distorsions resulting from Möbius aromaticity, most promimently the tetrathiocine and tetraselenocine ring systems. We also propose forms of the elements nitrogen and sulfur which have some Möbius-aromatic character, the latter in particular being only 11 kcal mol^-1 higher than the most stable conformation.

Heilbronner¹ proposed in 1964 that a 180^o phase shift in the p_p-p_p overlap of a cyclic array of such orbitals occupied by 4n p electrons and delocalised over the entire ring could result in a so-called Möbius conformation having aromatic properties, rather than the Hückel antiaromaticity associated with planarity. A characteristic feature noted subsequently is the chirality of these systems, often associated with a formal or approximate C₂ axis of symmetry passing through the aromatic ring. In 1998, Schleyer and co-workers² reported that the lowest energy conformation of the known conjugated 4n p_p C₉H₉⁺ cation was of this type, and we followed this with the suggestion of a Möbius form with C₂ symmetry for the neutral [16] annulene³. However, this conformation did not exhibit any significant aromaticity, as quantified using the theoretical NICS (nucleus independent chemical shift) probe.⁴ More aromatic Möbius conformations for [12], [16] and [20] annulenes were subsequently published by Schleyer and co-workers⁵, and although their predicted energies were around 6 kcal mol^-1 higher than the non-aromatic forms corresponding to the X-Ray structure of e.g. [16] annulene, it was suggested that these conformations might nevertheless play a role in the chemistry of [4n]annulenes. We had earlier shown⁶ that Möbius minima can also be located for heteroatom derivatives of [8] annulenes, but these were predicted to be > 20 kcal mol^-1 higher than non-Möbius forms. To date therefore, no Möbius aromatic neutral annulene rings have been identified which are more stable than their non or anti-aromatic isomers, and hence capable of structural characterisation and study. We report here our findings for 12p rings, including those derived from [12] annulene by replacing between one to four double bonds with heteroatoms (1-6), along with several other heteroannulene ring systems.

Computational Procedure.

Molecular structures were normally optimised without symmetry constraints using Gaussian98⁷ at the closed shell B3LYP/6-31G(d) level. The aromaticity of these systems was quantified using a NICS⁴ probe placed at the ring centroid. Coordinates are available via the electronic supplemental data (ESI) at http://www.rsc.org/ESI/...

Results and Discussion

Crystal Structures

The hetero[12]annulene series reveals several entries for compounds of the general type 1-5. In contrast to a number of crystalline examples of 7-ring heteropine [8]annulenes, only a single example of the [12] analogue 1, X=S, has been reported,¹⁴ but with a methano-bridge constricting it to a plane of symmetry. Likewise one example of 2 is known; an oxastibecene (X=O, Y=SbPh), although no 3D structure is reported. More examples occur for the ring system3 including the benzo-substituted examples X=Y=O¹³ and X=Y=S¹³ which exhibit a formal C₂ symmetry axis rather than a plane. One example containing X=S, Y=N-acetyl¹⁵ can be identified as having a pseudo C₂ symmetry axis which passes through one S and the opposing C=C bond. These latter species can all be described as having conformations formed by replacing three C=C units in the recently identified Möbius form of [12] annulene⁵ with the corresponding heteroatom.

There are also several interesting heteroatom rich systems. Six examples of the 12p ring system 4 (X=S, a tetrathiocin) are known¹⁷ to exhibit chiral D₂ ring symmetry, although occurances of this ring with C_s symmetry are more common. The only known selenium analog 4 (X=Se, tetraselenocine) exhibits D₂ ring symmetry,¹⁸ as do all the 4,5,6,7 isomers of tetrathiocin (5, X=S) which are also not rare.¹⁹ The unusual hexa-aza[12] annulene 6¹⁶ has distinct C₂ like distorsions from planarity.

We conclude that a search of the Cambridge database does seem to reveal tantalising evidence of stable conformations of [12] heteroannulenes which could exibit to some degree Möbius characteristics. To understand how the geometrical preferences of these systems arise, we turned to ab initio modelling at the B3LYP/6-31G(d) theoretical level.

Monohetero[12]Annulenes 1

For all three heteroatoms an unsymmetric (C₁) conformation was the most stable, although the C₂ isomers were significantly more accessible in energy (4.0 kcal mol^-1 higher for X=S, 7.4 for X=O, 12.0 for X=N) than we had previously observed for the smaller ring [8] heteropines.⁶ The NICS(0) value for the C₂ Möbius forms continued the trend of increased aromaticity along the series 1-3, as do the C=C ring bond lengths which reveal a small degree of increasing delocalisation (1-3 X=S; 1.339, 1.343 and 1.346/1.355/1.363Å respectively).

2,6-Dihetero[12]Annulenes 2

For the homodihetero systems, two C₂ symmetric conformations were located, the slightly lower having an axis perpendicular to the approximate plane of the ring and the other passing directly through the two heteroatoms. The C_s symmetric conformation shows all cis alkenes and the plane again passing through the two heteroatoms. With X=O or NH, the latter conformation was the more stable, whereas with X=S, the C₂ form was. The NICS(0) of the latter in particular corresponds to a modest diatropic ring current corresponding to mild aromaticity.

1,4,7-Trihetero[12]Annulenes 3

The conformational surface of these 9-membered rings is relatively complex, and no attempt here has been made to identify all the possible conformations. We instead concentrate on comparing those with a proper or psuedo C₂ symmetry axis with conformations with a C_s plane of symmetry. Where searches have revealed conformations with no (C₁) symmetry these are also included (Table 2). Thus the triaza system 1, X=Y=NH reveals a number of significant conformations (Table 2), the lowest in energy of which has an approximate C₂ symmetry axis passing through one NH bond and the mid point of the opposite C-C bond, broken by the orientation of the two remaining NH bonds, respectively axial and equatorial with respect to the ring itself. Isomers with two axial or two equatorial NH bonds exhibit proper C₂ symmetry but are 2.4-3.0 kcal mol^-1 higher in energy. A C_s symmetric isomer with cis C=C double bonds is 11 kcal mol^-1 higher in energy; a form with two trans C=C double bonds being 41.6 higher. Despite the greater stability of the C₂ forms, the NICS(0) value at the ring centroid of the most stable conformation (+1.3 ppm) indicates neither aromatic or anti-aromatic character. No examples of this simple ring system have been hitherto reported, but experimental characterisation does appear desirable. The phosphorus analogue shows a subtle difference to nitrogen. The more stable form reveals a pseudo C₂ axis passing from a C=C mid point through an axis defined by the P atom and its lone pair of electrons, indicating perhaps a lower degree of conjugation of this lone pair to the ring. Again, this ring is not known experimentally.

The analogous oxygen heterocycle (X=Y=O) is conformationally simpler with the C₂ form 6.4 kcal mol^-1 lower than C_s, and this mode of course corresponds to the observed crystal structure of the tribenzo derivative.¹³ As with the triaza heterocycle, the NICS(0) value indicates no aromaticity. We have previously suggested³ that ring fluorination has the effect of enhancing Möbius characteristics, and in this instance the perfluoro derivative does have a more aromatic NICS(0) value, but at the expense of stability, since this conformer is now 3.0 kcal mol^-1 less stable than the corresponding C_s form. The trithia heterocycle (X=Y=S) is the most interesting of the series. Attempts to locate the all cis alkene C_s form all converged to a C₂ isomer; the only C_s symmetric form that could be located contained two trans alkenes and was 18.8 kcal mol^-1 higher in energy. As with oxygen, perfluorination tends to a more aromatic NICS(0) but less stable C₂ isomer. Here again a known tribenzo derivative is known to adopt the C₂ form.¹⁴

We also explored several other variations of X or Y. Replacing Y=O by Y=CH^- forms a charged dihetero ring for which the C₂ form is no longer the more stable (by 3.1 kcal mol^-1), but for this system perfluorination makes the C₂ form more stable by 1.7 kcal mol^-1. The system X=O, Y=S has two C₂-like isomers differing by 2.0 kcal mol^-1 and both were much lower than the only C_s form locatable. With X=S, Y=CH^-, the stable conformation was again C₂, with ring perfluorination reducing the NICS(0) value significantly. The system X=S, Y=NR is of interest, since an example of this ring is known.¹⁵ With Y=NH, a pseudo C₂ structure corresponding to the crystal structure can be located; a C_s in this instance is computed to be 3.2 kcal mol^-1 more stable, but may well be the result of NH...S hydrogen bonding in the latter orientation. With X=S, Y=O, again C₂-like isomers differ by 1.6 kcal mol^-1, with the two trans C=C C_s form some 26.4 kcal mol^-1 higher.

In the trihetero ring system therefore, we see evidence of the preference for conformations with a C₂ rather than a C_s symmetry, and this preference does seem to be enhanced by sulfur in particular. The magnetic properties of these rings as quantified by the NICS(0) probe do not reveal any prominent aromaticity however, but as 4n (n=3) p systems, they are clearly not Hückel antiaromatics.

Other Heteroannulenes 4-7

Species 6 is a heteroatom rich [12] annulene with the unusual out of plane C₂-like distorsion as noted above. The calculations confirm this, and indicate also modest aromaticity (Table 2). This interpretation however must be moderated by the potential "trannulene" effect as described by Schleyer and co-workers²² which may also contribute to the diatropic ring current. The tetrathiocine ring 9, a=1, b=0 has a slightly aromatic NICS value, and this is more true of the alkene equivalent 7a, X=S, which also reveals a 5.7 kcal mol^-1 preference for D₂ symmetry. The molecular orbitals (Figure 1) for 7a, X=S are characteristic of systems with Möbius character³, revealing a catenane-like phase structure for the electronic Möbius strip. Ring fluorination in this case decreases the aromaticity and the stability of the Möbius form. The alternate isomers 7b, X=S are in absolute energies lower than 7a, X=S, and show similar mild aromaticity. The selenium heterocycle 7a, X=Se, R=H is predicted to be more twisted and with greater aromaticity than the sulfur analogue, an effect remarked upon previously in the context of spiroaromatic systems for the series N, P, As.²³. The known¹⁸ example (7a, X=Se, R=CF₃) has a measured C=C bond length of 1.385Å, which is approaching the fully aromatic delocalised value in e.g. benzene. The NICS value of -7.6 indicates this system is the most aromatic of the various discussed. The equivalent bond length in 7a, X=S is 1.33Å. The homologue tetrathia[20]annulene 9, a=2,b=1 reveals a strong D₂ twist in common with the series 4, but only a very slightly negative NICS value. This may in part be because the distance from the ring centroid to the nearest carbon atom is 2.64Å, which may be too far to induce any significant magnetic effect. The molecular orbitals of 9, a=2,b=1 have characteristically Möbius photogenic form!

One can remove all the carbon atoms entirely from the ring, resulting in S₈, for which a C₂ symmetric form 10 equivalent to a [16] annulene can be readily located, being only 11 kcal mol^-1 higher than the known stable form. This Möbius sulfur has modest NICS aromaticity, and again characteristic molecular orbitals (Figure 1). Continuing the removal of carbon atoms, the three benzene rings in 6 can be replaced by three N=N units, giving the N₁₂ allotrope 11. Unlike 6, it exhibits a true C₂ axis and modest aromaticity. Its energy relative to gaseous nitrogen (353 kcal mol^-1) suggests this is a purely theoretical species (we do not urge its synthesis!) but it does serve to illustrate the generality of the Möbius distorsion in the [12] and higher annulenes.

Conclusions

The prototypic molecule from which the concept of aromaticity and anti-aromaticity is largely derived is the monocyclic annulene, and although the suggestion that a Möbius aromatic form might be possible was made almost forty years ago, experimental characterisation of any such annulenes has hitherto been largely unaccomplished or unrecognised. Properties such as the molecular geometry, NMR characteristics and reactivity have thus far only been inferred by theoretical calculation. Following the recent suggestion that low energy Möbius conformations of medium sized 4np annulenes may be accessible, we here reinforce this by drawing attention to various hetero-[12]annulenes, particularly those containing sulfur, whose Möbius characteristics in stable (C₂ symmetric) conformations may not have hitherto been recognised. We feel it appropriate to conclude that the prospects for the synthesis of new stable forms of Möbius aromatic molecules must be good.

References

1. E. Heilbronner, Tetrahedron Lett., 1964, 29, 1923.
2. M. Mauksch, V. Gogonea, H. Jiao and P. von R. Schleyer Angew. Chemie Int. Edition, 1998, 37, 2395.
3. S. Martin-Santamaria and H S. Rzepa, ChemComm., 2000, 1089; idem, J. Chem. Soc., Perkin Transactions 2, 2000, 2372-2377
4. P. von R. Schleyer, C. Maerker, A. Dransfeld, H. Jiao, and N. J. R. van Eikema Hommes, J. Am. Chem. Soc., 1996, 118, 6317; H. Jiao and P. von R. Schleyer, J. Phys. Org. Chem., 1998, 11, 655.
5. C. Castro, C. M. Isborn, W. L. Karney, M. Mauksch and P. von R. Schleyer, Organic Letters, 2002, 4, 3431-3434.
6. W. L. Karney, C. J. Kastrup, S. P. Oldfield and H. S. Rzepa, J. Chem. Soc., Perkin Trans 2, 2002, 388-392.
7. Gaussian 98 (Revision A.11), M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, V. G. Zakrzewski, J. A. Montgomery, Jr., R. E. Stratmann, J. C. Burant, S. Dapprich, J. M. Millam, A. D. Daniels, K. N. Kudin, M. C. Strain, O. Farkas, J. Tomasi, V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C. Adamo, S. Clifford, J. Ochterski, G. A. Petersson, P. Y. Ayala, Q. Cui, K. Morokuma, D. K. Malick, A. D. Rabuck, K. Raghavachari, J. B. Foresman, J. Cioslowski, J. V. Ortiz, A. G. Baboul, B. B. Stefanov, G. Liu, A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R. L. Martin, D. J. Fox, T. Keith, M. A. Al-Laham, C. Y. Peng, A. Nanayakkara, C. Gonzalez, M. Challacombe, P. M. W. Gill, B. G. Johnson, W. Chen, M. W. Wong, J. L. Andres, M. Head-Gordon, E. S. Replogle and J. A. Pople, Gaussian, Inc., Pittsburgh PA, 1998.
8. F. H. Allen and O. Kennard, Chemical Design Automation News, 1993, 8, pp 1 & 31-37.
9. A. J. Matzger and K. P. C. Vollhardt, Tetrahedron Letters, 1998, 39, 6791-6794;
10. TDHBZD: H. Irngartinger Chem. Ber., 1973, 106, 761.
11. [16] D. Solooki, J. D. Bradshaw, C. A. Tessier, W. J. Youngs, R. F. See, M. Churchill and J. D. Ferrara, J. Organomet. Chem., 1994, 470, 231; RICVEE: [20] R. Boese, A. J. Matzger and K. P. C. Vollhardt, J. Am. Chem. Soc.,, 1997, 117, 2052; HEGGOP: [24] L. Guo, J. D. Bradshaw, C. A. Tessier and W. J. Youngs, Chem.Commun., 1994, 243
12. JAJTIX: J. D. Tovar, N. Jux, T. Jarrosson, S. I. Khan and Y. Rubin J. Org. Chem., 1997, 62, 3432
13. HMTHBZ: K. von Deuten and G. Klar Z. Naturforsch.,Teil B, 1981, 36, 1526.
14. K. von Deuten, J. Kopf and G. Klar Cryst. Struct. Commun., 1979, 8, 569.
15. NIJKIA: K. Kyongtae and M. Nyoung, J. Heterocyclic Chem., 1997, 34, 1-12.
16. HACKBEB: H. Hilpert, L. Hoesch and A. S. Dreiding, Helvetica Chimica Acta, 1985, 68 325-33; cited in the Cambridge database as A. S. Dreiding, J. H. Bieri, R. Prewo, A. Linden, H. Hilpert and L. Hoesch, Private Communication, 1993.
17. BIBNIJ: T. Chivers, M. Parvez, I. Vargas-Baca and G. Schatte, Can. J. Chem., 1998, 76, 1093; HONKUQ T. Shimizu, H. Murakami, Y. Kobayashi, K.Iwata and N.Kamigata, J. Org. Chem., 1998, 63, 8192; HONLAX: T. Shimizu, H. Murakami, Y. Kobayashi, K.Iwata and N.Kamigata, J.Org.Chem., 1998, 63, 8192; POSHOU: A. I. Kotov, S. V. Konovalikhin, R. V. Pisarev, G. V. Shilov, O. A. Dyachenko and E. B. Yagubskii, Mendeleev Commun., 1994, 180; TUBCIC: T. Shimizu, K. Iwata and N. Kamigata, Angew.Chem.,Int.Ed.Engl.,1996,35, 2357; ZINPOB: cited in the Cambridge database as E. Fanghanel, R. Herrmann, J. Bierwisch, H. Hartung, U. Baumeister, G. Maier and H. P. Reisenauer, Private Communication, 1996.
18. HETTAB: K. B. Borisenko, M. Broschag, I. Hargittai, T. M. Klapotke, D.Schroder, A.Schulz, H. Schwarz, I. C. Tornieporth-Oetting and P. S. White, J. Chem. Soc., Dalton Trans., 1994, 2705.
19. HOLPED: M. C. Aragoni, M. Arca, F. Demartin, F. A. Devillanova, A.Garau, F. Isaia, F. Lelj, V. Lippolis and G.Verani J. Am. Chem. Soc., 1999, 121, 7098; JUSSAR: F. Bigoli, P. Deplano, M. L. Mercuri, M. A. Pellinghelli, E .F. Trogu and A.Vacca, Semin. Virol., 1991, 62, 53; SOJTIU: F.Bigoli, M. A. Pellinghelli, P.Deplano and E.F.Trogu, Inorg. Chim. Acta, 1991, 182, 33; VEBCEK D. Atzei, P. Deplano, E. F. Trigu, F. Bigoli, M. A. Pellinghelli, A.Sabatini and A.Vacca, Can. J. Chem., 1989, 67, 1416; VIDKOI: F. Bigoli, M. A. Pellinghelli ,P.Deplano, E.F.Trogu, Inorg. Chim. Acta, 1990, 170, 245; ZARDOL: F. Bigoli, P. Deplano, M. L. Mercuri, M. A. Pellinghelli, A. Sabatini, E. F. Trogu and A.Vacca, Can. J. Chem., 1995, 73, 380.
20. COVCEV20: A. Muller, M. Romer, H. Bogge, E. K. Rickemeyer and M. Zimmermann Z. Anorg. Allg. Chem., 1986, 534, 69; CILPUI: G. Marbach and J. Strahle Angew. Chem., Int. Ed. Engl., 1984, 23, 246.
21. C. J. Kastrup, S. Oldfield and H. S. Rzepa, ChemComm., 2002, 6452-643.
22. A. A. Fokin, H. Jiao and P. von Schleyer, J. Am. Chem. Soc. 1998, 120, 9364-9365; A. D. Darwish, I. V. Kuvytchko, X.-W. Wei, O. V. Boltalina, I. V. Gol'dt, J. M. Street and R. Taylor, J. Chem. Soc., Perkin Transactions 2, 2002, 1118-1121.
23. D. Hall and H. S. Rzepa, Org. Biomol. Chem., 2002, in press; H. S. Rzepa and K. Taylor, J. Chem. Soc., Perkin Trans 2, 2002, 1499-1501.