Deaminative Ring Contraction for the Modular Synthesis of Pyrido[n]helicenes

Click to copy article linkArticle link copied!

Helicenes are a fascinating class of polycyclic aromatics, aptly named after the chiral helicity (handedness) that results when ortho-condensed aromatic systems are extended beyond three consecutive rings. (1) Pyrido[n]helicenes, like 1 (n = 5) and 2 (n = 6), are of particular importance as their pyridyl moieties can be regioselectively functionalized and the pyridyl group can be leveraged for the scalable resolution of helical enantiomers (Figure 1A). (2,3) Due to its relative structural simplicity, room temperature configurational stability, and pH-dependent chiroptical properties, 1-aza[6]helicene 2a (3−5) has emerged as a promising chiral small molecule additive for the design of novel optical and electronic materials. (6−9) Additionally, 2a and its N-oxide derivative (not shown) are useful in asymmetric catalysis. (4,10) However, even the prototypical 1-aza[6]helicene, 2a, is not commercially available and can be challenging to prepare─requiring the multistep conversion of advanced building blocks and an intermediary resolution step (3) if a single enantiomer is desired (see 2a, (P)-isomer shown). (3−5)

Figure 1

High Resolution Image

Download MS PowerPoint Slide

While developed strategies have recognized the C and D rings as convergent assembly points (Figure 1A, shown in orange for 1 and 2), modular access to #-aza (e.g., # = 1-, 2-, 3-, 4-, 5-, or 6-) and #,#′-diaza variants (e.g., #,#′ = 1,14- or 1,16-) that could be useful in tuning materials and catalysis applications is underdeveloped. The general strategies for the synthesis of pyrido[n]helicenes (3−5,11−14) are limited because the incorporation of a single heteroatom (e.g., N) can complicate isomer outcome during alkene-forming events (i.e., intermediary E- and Z-alkene isomers) (4,11) and ring-forming events (i.e., two or more isomers possible). (12,13) Direct access to chiral pyrido[n]helicenes is further challenged (14,15) by the added Lewis basic site(s) that can compromise the catalytic efficiency and asymmetry of involved bond-forming steps. (5) Recently, Yang and co-workers developed a route to chiral 4-aza[4] and 4-aza[5]helicenes using an asymmetric multicomponent condensation reaction to build the pyridyl ring last; however, access to other N-atom positional isomers is limited. (15)

Motivated by the potential value of N-atom positional isomers, we envisioned the development of a process to convert simple-to-prepare prepare tertiary amines, 4 and 5, into 1-aza[5]helicenes (1a) and 1-aza[6]helicenes (2a), respectively (Figure 1B). Reductive cyclization (16,17) would regioselectively conjugate the amine-templated ring system (e.g., AB + DE (4), ABC + E (5), and ABC + EF (5)), and a subsequent deaminative contraction process would forge the C or D ring (shown in orange). The realization of these exemplars could motivate general and modular access to #-aza[5]helicenes (1) and #-aza[6]helicenes (2) from appropriate tertiary amine templates. Furthermore, the sequence has the potential to be telescoped, without the purification of intermediates, which, when coupled with a simple resolution protocol, would drastically simplify access to chiral pyridyl-containing helicenes.

In 1985, Maigrot and Mazaleyrat showed that the double alkylation of l-(−)-ephedrine 6 with (±)-7 formed separable chiral azepinium diastereomers that could be transformed into chiral 1,1′-binaphthyl derivatives (Figure 2). (18) Inspired by this, Závada and co-workers realized that the chiral auxiliary component could be removed via epoxide (8) formation (19) and, thus, advanced chiral azepinium, (S)-(+)-9, transformed into chiral [5]helicene (P)-(+)-10 via a [1,2]-Stevens rearrangement, dehydroamination cascade reaction (confer ‘Stevens intermediate’ 11). (20) This singular demonstration, where (±)-7 derives from the homocoupling of 1-bromo-2-methyl naphthalene in two steps, (18) further motivates the need for a more direct and asymmetric assembly strategy to prepare biaryl-linked dihydroazepines that can be advanced to pyrido[5] and [6]helicenes.

Figure 2

High Resolution Image

Download MS PowerPoint Slide

We began our studies by examining the scope of tertiary amines that would be transformed into a variety of representative (di)aza[5] (1) and (di)aza[6]helicenes (2) in two steps (Figure 1B). While the synthesis of some targeted (di)aza[5] and (di)aza[6]helicenes are reported, (3−5,11,13) the strategies for their assembly are disparate and of varied yield. The required tertiary amines (12) were prepared in moderate to high isolated yields by alkylating benzylic secondary amines with benzylic bromides (see the Supporting Information). (21,22) The ortho-bromo-bearing tertiary amines (12) are readily cyclized into biaryl-linked dihydroazepines (13) using Ni-promoted conditions (Figure 3A). (21−25) Alternatively, related tertiary amines can be prepared using reductive amination methods. (21−23) The C*-ring-forming cyclizations proceed with moderate to high efficiency to provide 13a–13i in 54–84% yields, except for 13d, obtained in 21% yield. We tentatively attribute this lower yield to the ability of 12d and its reduced intermediates to chelate the Ni catalyst nonproductively. (22,26) Using a similar sequence, the D*-ring-forming cyclizations proceed with moderate to high efficiency to provide 13j–13n in 33–84% yields.

Figure 3

High Resolution Image

Download MS PowerPoint Slide

With biaryl-linked dihydroazepine intermediates (13) in hand, we show that these stable intermediates can undergo mild, pyridyl-tolerant deaminative contractions to give (di)aza[5] (1) and (di)aza[6]helicenes (2) (Figure 3B). Under optimized conditions using trimethyl phosphate, (21,22,27,28) the C-ring-forming contractions proceed with moderate to high efficiency, providing 1a–1i in 50–89% yields. Optimization with 13a to form 1a found 18-crown-6 equivalency and anhydrous THF (<20 ppm of H₂O) to be critical factors for the high conversion and reproducible isolation of 1a (CSD 2402615, for substrates with observable side products, reducing the reaction exposure time can improve the isolated yields; see the Supporting Information). Similarly, the D-ring-forming contractions proceed with moderate efficiency, providing 1d, 1k–1l, and 2a–2b, in 31–67% yields. Having optimized the cyclization and deaminative contraction conditions for #-aza[n]helicenes, we investigated the scalability of the developed methods and pursued the synthesis of novel C-ring-alkylated 1-aza[n]helicenes. We targeted 1-aza[6]helicene 2a to demonstrate synthesis on scale due to its configurational stability (3) and established use in optical devices. (6−9) The gram-scale synthesis of 2a proceeds smoothly (see the Supporting Information for minor procedural deviations). Starting with 4 g of 12m (7.7 mmol), its reductive cyclization produces 13m in 72% isolated yield (1.99 g, 5.5 mmol). Next, we converted 13m (1.94 g, 5.4 mmol) into more than 1.5 g of (±)-1-aza[6]helicene 2a─an 85% isolated yield (1.52 g, 4.6 mmol). We attribute the improved yields to the larger scale, which may benefit from less adventitious water.

With the successful assembly of pyridyl-containing [5] (1) and [6]helicenes (2) (Figure 3), we considered if the tertiary amine substrate could bear an α-alkyl substituent that could then be translated into a C7/C8-alkylated 1-aza[5]helicene variant (Figure 4A). We envision that C-ring-alkylated 1-aza[5]helicenes derivatives can be designed or functionalized (e.g., radical halogenation) to improve their properties, enable covalent conjugation handles, and motivate novel applications in materials science. By design, the alkyl group can be incorporated at either the C7- or C8-position of the target (confer 1o). To evaluate this, we prepared C8-α-methyl tertiary amine 12o (see the Supporting Information), which undergoes reductive cyclization to yield α-methyl biaryl-linked azepine 13o in 53% isolated yield as an inconsequential mixture of diastereomers (d.r. 1:1, determined by integration of the ¹H NMR spectrum). Subjecting the diastereomers (13o) to the standard deaminative contraction conditions gives 1-aza-8-methyl[5]helicene 1o (25%) alongside vinyl-bearing naphthyl derivative, 14 (see the Supporting Information), which likely arises via a competitive Hofmann elimination pathway. (29) Efforts to optimize the desired [1,2]-Stevens, dehydroamination pathway toward 1o and with other C7/C8-alkylated variants are ongoing.

Figure 4

High Resolution Image

Download MS PowerPoint Slide

Next, we set out to show that a designed aza[6]helicene can be efficiently prepared on scale by two iterations of a cyclize–contract sequence (Figure 4B). Advancing our previously developed two-step method, where biaryl-linked dihydroazepines were isolated, we streamlined the synthesis of the key 9-methyl-benzo[h]quinoline 15 building block from tertiary amine 16, itself available on multigram scale in two steps from commercial reagents. (22) The purification of biaryl-linked dihydroazepines using column chromatography on preparative scales can be challenging due to their highly polar tertiary amine moieties that tend to streak on SiO₂ and coelute with triphenylphosphine oxide. To mitigate this, the cyclize–contract events (steps 3a and 3b) can be performed in series, without the purification of intermediates, to form 15 in 67% isolated yield from 16. This overall efficiency compares favorably to that of the two-step process. Next, 15 is brominated at C10 (step 4) (30) and then again at the C9 methyl (step 5), and the resultant product reacts with secondary amine 17 (step 6) to afford tertiary amine 12m in 44% yield over three steps. (4,22) Encouragingly, the same cyclize–contract events (steps 7a and 7b) proceed to form the target 1-aza[6]helicene 2a in 31% yield on a 1.9 mmol scale. Here, the overall efficiency is near that of the two-step process (step 7a: 67% of isolated 13m; step 7b: 67% of 2a from 13m; calculated 45% over two steps). Efforts to further improve this streamlined process, where the B and D rings are modular assembly points, are ongoing. Finally, (±)-2a can be readily resolved by leveraging the differential solubility of its dibenzoyl-l-tartrate salts to access (+)-(P)-2a (shown) and (−)-(M)-2a on scale. (3) A dimeric complex (18) suitable for X-ray diffraction forms when (+)-(P)-2a (calculated 92% ee) is treated with 0.5 equiv of silver(I) trifluoromethanesulfonate in THF (Figure 4B, (+)-(P)-2a, CSD 2402616). Overall, the telescoped process is modular, provides access to both enantiomers, and is anticipated to be useful for the design of other room temperature configurationally stable #-aza[6]helicenes, like 2a.

Pyridyl-containing helicenes are interesting molecules on account of their unique chiroptical properties and the ease by which their enantiomers can be resolved, but further exploration of their applications has been thwarted by their limited and challenging syntheses. We have developed a reductive cyclization–deaminative contraction strategy that enables modular access to various pyridyl-containing helicenes (15 demonstrated examples), including access to enantioenriched 1-aza[6]helicenes (2a), showing that the methods can be quickly adapted to synthesize desired analogues, including N-atom positional isomers (1a–1l) and (di)aza[6]helicene variants, 2a and 2b, on preparative scales (Figure 4C). (31) Previous efforts show that 1-aza[4]helicene 19 and [4] and [5]carbohelicenes (e.g., 20 and 21) are also accessible, connecting general access to pyrido- and carbohelicenes. (22) Efforts to modify the prepared parent helicenes and C-ring-alkylated variant (1o) to enable further design granularity are ongoing.