Benzylic-type couplings provide an important asymmetric entry to functionalized, non-racemic helicenes

Abstract

Benzylic-type couplings are key reactions to make non-racemic helicenes. Their simplicity contributes to a short, efficient, and scalable asymmetric route to functionalized [7]helicenes. It widens the scope and uses of enantiopure helicenes, which are unequivocally needed for new chiroptical-electronic materials. A mechanistic proposal featuring an electrocyclization is based on experiments, X-ray crystallography and calculations.

---

Helicenes are known for over 100 years, but their importance is growing because of their attractive chiroptical, photophysical, physico-chemical, and electronic properties.¹ Those features are highly desirable for further developments in materials science and in other fields.² Today, a real challenge is the convenient preparations of enantiopure, tailor-made, non-racemic helicenes (including the longest ones) from the simplest scalable methods. One of the solutions to those problems is disclosed here in the non-racemic synthesis of a series of [7]helicenes, and a tentative mechanistic rationale of the benzylic couplings.

For nearly five decades, the preferred synthetic method to make racemic helicenes has been the oxidative photo-cyclodehydrogenation.³ A well-known drawback is the high dilution in solvents (100 mg L⁻¹; 10⁻³ M) for producing a modest amount of helicenes in yields averaging 50–60%. To circumvent those issues, some elegant, thermal methods have been reported.^1,2 However, only a few provide short, scalable and general entries to helicene chemistry, but more rarely from non-racemic, stereoselective routes.⁴

Racemic benzylic-type couplings have combined easy functionalization, high yields, cheap reagents and simple equipments for producing [5]helicenes,^5,6 which allowed further applications in nanoscience.⁷ All results focussed on racemic, configurationally labile, short helicenes.

This work is distinct from the previous ones because stereoselective benzylic couplings to enantiopure, long helicenes, were never reported by us or by others in helicene chemistry (Schemes 1 and 3). Additionally, functionalization of non-racemic (Scheme 3) and racemic (Scheme 4) heptahelicenes are reported, after an optimization of a biphenanthrol synthesis (Scheme 2). A direct access to either enantiomers of helicenes 1 (disubstituted) and 2 (monosubstituted), in only three steps from the known bis(triflate) (±), (−) or (+) 6, is a key feature.⁸ Those couplings also prompted us to propose a mechanistic rationale based on calculations (Scheme 5) and X-ray crystallography (Scheme 1).

Scheme 1 Asymmetric benzylic-type couplings in the synthesis of non-racemic [7]helicenes.

Scheme 2 Efficient synthesis of 3,3′-dihydroxy-4,4′-biphenanthryl 5.

Toward a non-racemic entry to [7]helicenes, we first efficiently produced (±)-biphenanthrol 5, as shown in Scheme 2. The oxidative coupling of 3-phenanthrol with a molar excess of a Cu(NO₃)₂/(±)1,2-diphenylethylamine complex under an inert atmosphere yielded (±)5 in a 77% yield (previously: 20%).⁹ For a better convenience, we optimized a catalytic method^8,10 in air with Cu(TMEDA)ClOH (15–20 mol%) to provide (±)5 in a 90% yield, (>3 g, previously 68% (ref. 9)). Racemic (±)5 was easily resolved to (−)5 (85% e.e.) and to (+)5 (92% e.e.) on a gram scale.^8,11 The stereochemistry of helicenes was thus set from the start from non-racemic (+)- or (−)-biphenanthrol 5 with a full retention of configuration.

Having 5 in hand, a triflation with pyridine in DCM, afforded bis(triflate) 6 (Scheme 3) in a racemic (98% yield) or in a non-racemic form ((−)-94% or (+)-96% yield). From (±)6, a Ni-catalyzed methylation produced (±)7 in a 86% yield (Scheme 3).¹² Similarly, non-racemic (+)7 (85% yield, 97% e.e.) or (−)7 (79% yield, 85% e.e.) were obtained with a retention of configuration. A radical tetrabromination of (±)7 to (±)3 (85–91% yield) required an excess of NBS. The satisfactory purity allowed to pursue the synthesis of (±)1. Under similar conditions, both enantiomers (−)3 (>95% e.e.) and (+)3 were also produced with a retention of configuration. A spontaneous ring closure of (±)3 provided (±)-9,10-dibromo-heptahelicene 1 (80–88% yield) or each enantiomers (M)(−)1 (>98% e.e.) and (P)(+)1 (>98% e.e.). Traces of 2 (<4%, HPLC) came from the cyclization of residual tribromide 4 (<4% in 3, HPLC).

Scheme 3 An efficient entry to non-racemic [7]helicenes via asymmetric benzylic-type couplings.

As a preliminary optimization of the benzylic couplings to (±)1, a few parameters were examined for the first time: (a) bases, (b) temperatures, (c) solvents, (d) addition rates. In short, tBuOK was one of the most practical and efficient bases. However, LiHMDS did not provide a significant coupling in the absence of HMPA, in spite of a higher basicity. Similarly, K₂CO₃ did not promote a cyclization to 1 in DMF. As for the temperature, incomplete reactions and slow rates were noticed at −35 °C with tBuOK in DMF. The best yields of 1 were thus obtained near 0–3 °C. A polar solvent was needed and DMF provided the best results with tBuOK. Less polar solvents such as THF, dioxane or CH₂Cl₂ were not successful with tBuOK, and a fast addition of tBuOK to the precursor 3 produced a high molar ratio 1/2.

As a convenience, crude 3 generated a molar ratio 1/2: 4.2 in a 80% yield. The tribromide 4 accounts for the formation of 2. If 3 is almost pure (HPLC: 96% of 1; 4% of 4), a 87% yield of 1 is then obtained. We next turned our attention to the reactivity of (±)1 in some metal-catalyzed couplings (Scheme 4). A double Sonogashira reaction in triethylamine, with PdCl₂(PPh₃)₂/CuI as catalysts, generated (±)8 in a 95% yield. Removal of TMS groups from (±)8 with K₂CO₃ in MeOH lead quantitatively to (±)9. As an innovative one-pot functionalization, a double cyanation with CuCN in moist N-methyl-pyrrolidinone at 180 °C directly produced the 9,10-imido-helicene (±)10 (75% yield).

Scheme 4 A new entry to functionalized [7]helicenes.

We wanted to understand the coupling mechanism from 11. A previous thought was a carbenoid coupling coming from α-eliminations of HBr at the benzylic positions. However, it was found unlikely with LiHMDS/HMPA at 0 °C from a [5]helicene derivative.⁵ For the first time, a hypothetical mechanism is outlined in Scheme 5. A long-range benzylic-type elimination from 11 is reasonable, based on numerous reports.¹³ It could then lead to trienes 12 and/or 13, followed by an electrocyclic rearrangement to 14. As a support of this proposal, another type of electrocyclization was suggested, but in a [5]helicene formation.¹⁴ Finally, an aromatization after a second elimination would generate 15. The first elimination might be a rate-determining step sensitive to a polar solvent (such as DMF), due to a stabilization of a polarized transition state. To get further insights, preliminary calculations with the GenMol® program were performed.¹⁵ In the [7]helicene series, the relative constraint energy (torsion, angle, steric, electrostatic energy) of each species under study was evaluated (Scheme 5). Compounds 11 (52.3 kcal mol⁻¹) and 15 (48.7 kcal mol⁻¹) were more energetic or comparable to 12 (47.5 kcal mol⁻¹) or to 13 (48.2 kcal mol⁻¹). As a consequence, 12 or 13 are thus energetically attainable intermediates. A similar biphenanthryl structure with a bridged double-bond supports the assumption of a thermal stability of 12 or 13.¹⁶ As for 11, a positive electrostatic potential at each C–H benzylic-type proton induces a strong benzylic repulsion, combined to a steric strain at both terminal benzene rings. The constraint energy of 14 is less than for 15 because of a staggered conformer at the 9,10-position, causing also less steric repulsion at the terminal rings. As for the π-orbitals overlap of 12 and 13, a distance of 3.2 Å and 3.3 Å, respectively, was found between the terminal methylene carbon atoms in the lowest-energy conformers. Even if a distorsion of 12 or 13 for shortening this distance costs in energy, a slight deviation from planarity over many atoms provides a minimal activation energy of the transition states.

Scheme 5 A mechanistic hypothesis based on GenMol® calculations.

Conclusions

For the first time, stereoselective benzylic-type couplings are shown as key reactions in some expeditive, non-racemic (or racemic¹⁷) routes to functionalized higher helicenes. All intermediates of the three-step sequence to helicenes proceed with a full retention of configuration, which is set from the start. Such efficient and scalable syntheses are rare in the literature for making enantiopure [7]helicenes. Based on X-ray crystallography, experiments, and GenMol® calculations, an electrocyclization mechanism is proposed. Carbenoid couplings are unlikely. This work significantly widens the production and uses of non-racemic helicenes, which are absolutely required for enantiopure chiroptical-electronic materials. It will unequivocally add a contribution to chiroscience, materials and nanoscience, chemical-biology, and supramolecular chemistry. Importantly, the simplicity of this method will encourage a spread use of it within the scientific community.

Acknowledgements

ANR BioNanoPlasmon, MESR France, CNRS, Aix-Marseille Université and CINAM. Dr M. Giorgi (for XRD), Mr E. Pizarro-Cartier and Mr X. Robillard (Univ. Sherbrooke).

Notes and references

1. M. Gingras, Chem. Soc. Rev., 2013, 42, 968; M. Gingras, G. Félix and R. Peresutti, Chem. Soc. Rev., 2013, 42, 1007; Y. Shen and C.-F. Chen, Chem. Rev., 2012, 112, 1463; I. G. Stará and I. Starý, Sci. Synth., 2010, 45b, 885; S. K. Collins and M. P. Vachon, Org. Biomol. Chem., 2006, 4, 2518.
2. M. Gingras, Chem. Soc. Rev., 2013, 42, 1051.
3. M. Schölz, M. Mühlstaed and F. Dietz, Tetrahedron Lett., 1967, 8, 665; M. Flammand-Barbieux, J. Nasielski and R. H. Martin, Tetrahedron Lett., 1967, 743.
4. See ref. 1. It mainly involves Diels–Alder, [2π + 2π + 2π] cycloisomerizations, radical, metal-catalyzed C–O, C–N, C–S, C–C couplings.
5. M. Gingras and F. Dubois, Tetrahedron Lett., 1999, 40, 1309; F. Dubois and M. Gingras, Tetrahedron Lett., 1998, 39, 5039.
6. S. Goretta, C. Tasciotti, S. Mathieu, M. Smet, W. Maes, Y. M. Chabre, W. Dehaen, R. Giasson, J.-M. Raimundo, C. R. Henry, C. Barth and M. Gingras, Org. Lett., 2009, 11, 3846.
7. C. Barth, M. Gingras, A. S. Foster, A. Gulans, G. Félix, T. Hynninen, R. Peresutti and C. R. Henry, Adv. Mater., 2012, 24, 3228; B. Hoff, M. Gingras, R. Peresutti, C. R. Henry, A. S. Foster and C. Barth, J. Phys. Chem. C, 2014, 118 DOI:10.1021/jp501738c.
8. K. Nakano, Y. Hidehira, K. Takahashi, T. Hiyama and K. Nozaki, Angew. Chem., Int. Ed., 2005, 44, 7136.
9. B. Feringa and H. Wynberg, Tetrahedron Lett., 1977, 4447; K. Yamamoto, K. Noda and Y. Okamoto, J. Chem. Soc., Chem. Commun., 1985, 1065.
10. Cu(TMEDA)ClOH and 2-naphthols: M. Nakajima, I. Miyoshi, K. Kanayama, S.-i. Hashimoto, M. Noji and K. Koga, J. Org. Chem., 1999, 64, 2264; M. Noji, M. Nakajima and K. Koga, Tetrahedron Lett., 1994, 35, 7983.
11. The configurations of (−)12 or (+)12 were assigned by comparison to optical rotation of (S)-(+)12 and (R)-(−)12 in ref. 9. % e.e. for 8, 9, 12 and 14 were determined by chiral HPLC (Chiralcel OD-H, n-heptane/i-PrOH), UV-Vis and CD chiral detector.
12. S. Sengupta, M. Leite, D. S. Raslan, C. Quesnelle and V. Snieckus, J. Org. Chem., 1992, 57, 4066.
13. For long-range eliminations involving benzylic groups: M. Gingras, J.-M. Raimundo and Y. M. Chabre, Angew. Chem., Int. Ed., 2007, 46, 1010 and references therein. Classic examples of eliminations from 1,2- or 1,4-bis(bromomethyl) benzene have lead to quinodimethanes.
14. R. Pathak, K. Vandayar, W. A. L. van Otterlo, J. P. Michael, M. A. Fernandes and C. B. de Koning, Org. Biomol. Chem., 2004, 2, 3504.
15. G. Pèpe, B. Serres, D. Laporte, G. Del Ré and C. Minichino, J. Theor. Biol., 1985, 115, 571; G. Pèpe, GenMol, Version 2008, CNRS UMR 7325 CINaM.
16. B. Feringa and H. Wynberg, J. Am. Chem. Soc., 1977, 99, 602.
17. For a few brominated, racemic heptahelicenes: J. Zadny, P. Velisek, M. Jakubec, J. Sykora, V. Cirkva and J. Storch, Tetrahedron, 2013, 69, 6213; C. Collet and M. Gingras, Synlett, 2005, 2337.

---

Footnote

† CCDC reference number 951885. For crystallographic data in CIF or other electronic format see DOI: 10.1039/c4ra04398g

---