Oxidative Dearomatization of Pyridines

Abstract

Dearomatization of pyridines is a well-established synthetic approach to access piperidines. Although remarkably powerful, existing dearomatization processes have been limited to hydrogenation or addition of carbon-based nucleophiles to activated pyridiniums. Here, we show that arenophile-mediated dearomatizations can be applied to pyridines to directly introduce heteroatom functionalities without prior substrate activation. The arenophile platform in combination with olefin oxidation chemistry provides access to dihydropyridine cis-diols and epoxides. These previously elusive compounds are now readily accessible and can be used for the downstream preparation of diversly functionalized piperidines.

Graphical Abstract

The abundance of pyridine and piperidine motifs in FDA-approved drugs and natural products has motivated extensive research in the area of synthetic methodology.¹ Pyridine dearomatization, one of the most enabling strategies to access functionalized piperidines and related precursors, traditionally proceeds through the hydrogenation of pyridines² (Figure 1a) or the addition of nucleophiles to activated pyridinium salts (Figure 1b).³ Significant advances in catalytic hydrogenation have provided improvements in chemo-, regio-, and stereoselectivity,⁴ and the emergence of frustrated Lewis pairs offers a complementary approach to traditional metal-catalyzed processes.⁵ Similarly, numerous contributions to the dearomatization of activated pyridines has advanced the state-of-the-art. Noteworthy strides have been made towards more sustainable transient activation with metals and various electrophilic groups,⁶ as compared to traditional stoichiometric activators. Furthermore, recent advances have enabled new modes of introducing unconventional functionality,⁷ including asymmetric processes.⁸ Although these methods rapidly build complexity, most are limited to the introduction of hydrogen atoms or formation of C─C bonds.

Figure 1.

Dearomative chemistry of pyridines, including this work.

Chemoenzymatic approaches have potential to remedy this diversification gap, as previously demonstrated with arene dihydroxylation chemistry.⁹ However, the translation of arene dioxygenases to pyridines proved to be unsuccessful due to the inherent instability of dihydropyridine cis-diols, which spontaneously aromatize towards hydroxylated pyridines (Figure 1C).¹⁰ The absence of protocols to install heteroatom functionalities hinders access to decorated piperidines.

Dearomative cycloadditions offer complementary reactivity and open novel avenues of diversification.¹¹ Several groups, including our own, have recently reported a series of dearomative strategies based on cycloadditions between non-activated arenes and arenophiles that provide access to functionalized cyclohexadienes.¹² Noting that this strategy could deliver elusive piperidine scaffolds functionalized with heteroatoms, we desired to extend this work to pyridines. Importantly, dearomative cycloadditions with pyridines are exceedingly rare¹³ and the first practical [4+2] cycloaddition was only recently demonstrated in intramolecular settings.¹⁴ Herein, we show that pyridines are viable substrates for arenophile-mediated para-cycloaddition and the resulting cycloadducts are amenable to in-situ oxidations. Specifically, N-methyl-1,2,4-triazoline-3,5-dione (MTAD, 1) in combination with olefin oxidations delivered dihydropyridine cis-diols and epoxides (Figure 1D). These unique products were used to prepare a series of heteroatom-substituted piperidines that would be challenging to prepare through other means.

At the onset of our investigations, we tested a range of pyridine substrates to probe for feasibility and establish guiding reactivity principles. Although pyridine did not engage in the desired para-cycloaddition, further screening revealed that a substitution at the 2-position was necessary to promote [4+2] photocycloaddition with 1. Using 2-phenylpyridine (2a) as a substrate for optimization, we explored in situ dihydroxylation of the corresponding cycloadduct (see Table inset, Figure 2a; for additional details see Supporting Information). We began with our previously reported conditions¹⁵ where compatibility with cycloaddition and in situ dihydroxylation necessitated use of acetone as the solvent. Several modifications to the established procedure for benzene-derived substrates were necessary for efficient dihydroxylation of the pyridine-derived adduct, including use of cosolvent quantities of water. Although p-TsNH₂ was found to be an ineffective additive, citric acid proved to be crucial.¹⁶ No product was observed in the absence of citric acid, and inclusion of two equivalents provided the best outcome. Concentration, temperature, and the loading of osmium tetroxide also showed significant effects on yield. A 3-fold excess of 2a relative to 1 proved optimal, with attempts to decrease the equivalents of 2a having a detrimental impact on the reaction time and yield. Our optimized conditions provided product 3a in 58% isolated yield as a single constitutional and diastereoisomer. Our two-step, one-pot protocol was applied to diverse substituted 2-arylpyridines to assess the scope of arene compatibility. A range of diol products containing o-chloro (3b), m-nitro (3c), p-nitrile (3d), and p-pivaloyloxy (3e) handles was obtained. In addition to arenes, 2-alkyl and 2-methoxy substituted pyridines also proved to be viable photocycloaddition partners as demonstrated by products 3f–3j. Furthermore, this protocol was shown to tolerate ester functionalities (3g and 3j) as well as 2,6- and 2,4-disubstitution patterns (3i and 3j). Finally, the robustness and scalability of this dearomative oxidation was tested by conducting this protocol on a larger scale, and we observed only a slight decrease in the yield of product 3h. In all cases, the dihydroxylation product was obtained as a single constitutional isomer and diastereoisomer.

Figure 2.

Arenophile-mediated dearomative dihydroxylation of pyridines. (a) Optimization and scope. (b) Cycloreversion. (c) Derivatization of products. See Supporting Information for additional details.

With several dihydroxylated bicyclic products in hand (Figure 2A, inset), we focused on cycloreversion of the arenophile motif to reveal dihydropyridine cis-diols (Figure 2b). Initial efforts revealed the exceptional instability of dihydropyridine cis-diols, which spontaneously underwent elimination and aromatization as previously observed in the chemoenzymatic dihydroxylation of pyridines. Based on our experience handling sensitive dihydrodiols¹⁵ and arene oxides,¹⁷ we protected the dihydroxylation products as acetonides (3a→4a) and applied mild cycloreversion conditions (4a→5a). Partial urazole hydrolysis followed by nickel peroxide-mediated cycloreversion¹⁸ provided the desired dihydropyridine cis-diol acetonide 5a. This protocol was applied to other representative diol substrates, including 5h and 5k. In the case of 6-methyl-2-methoxypyridine (2k), the dihydroxylation product was unstable to purification and was directly converted to acetonide 4k, which ultimately afforded 5k. Notably, the mild conditions of this arenophile-mediated dearomative strategy provided each dihydropyridine cis-diol without undesired aromatization byproducts.

Although dihydropyridine cis-diol acetonides are not stable to standard work-up procedures, purification, or long-term storage, they could be used immediately in various functionalization reactions (Figure 2c). For example, several dihydropyridine cis-diols were directly reduced with sodium borohydride to deliver dihydroxylated piperidines 6a, 6b, and 6f. Ionic reduction with trifluoroacetic acid and triethylsilane provided monohydroxylated imine 7.¹⁹ We also examined the reactivity of dihydropyridine cis-diol 5a with Grignard reagent in the presence of TMSOTf and observed an interesting double adduct 8,²⁰ likely proceeding through S_N2′ and imine addition. Furthermore, we explored the cycloaddition of dihydropyridine cis-diol acetonides with nitroso compounds to introduce additional heteroatom functionalities. Substrate 5a reacted with nitrosobenzene in a highly diastereo- and regioselective manner,²¹ and the corresponding cycloadduct underwent N─O bond cleavage with CuCl²² or hydrogenolysis²³ to deliver functionalized piperidine derivatives 9 and 10. Seeking an alternative to cycloreversion, we developed a protocol for reductive cleavage of the urazole motif.^12a To this end, we demonstrated that treatment of 4a or 4h with lithium aluminum hydride followed by Raney-Ni yielded aminohydroxylated piperidine derivatives 11 and 12.

We next investigated the reactivity of pyridine-arenophile cycloadducts under epoxidizing conditions to provide pyridine oxides (2→13→14, Figure 3a). These oxidized heterocycles have been largely underexplored;²⁴ however, they are known to undergo valence bond isomerization like their better-known arene oxide counterparts.²⁵ Accordingly, pyridine oxides readily interconvert with their 1,4-oxazapine tautomers (15). Mn-based epoxidation²⁶ showed promise in initial studies. Notably, a significant increase in catalyst loading relative to the initial report was required for optimal yields. This could be attributed to the challenging nature of the epoxidation and/or to competitive binding of the excess pyridine substrate that remains after the cycloaddition. The cycloadduct epoxides were subjected to the described cycloreversion to provide pyridine oxides 14. This two-step protocol successfully yielded more than a dozen pyridine oxides. For example, 2-phenylpyridine and various substituted 2-arylpyridines bearing chlorine, nitro, or cyano groups delivered oxides 14a–14d. Additionally, 2-bipyridines, such as 2,2′- and 2,3′-bipyridine proved to be viable substrates, delivering 2-bipyridine oxides 14l and 14m. In all cases, the cycloaddition and subsequent epoxidation occurred chemoselectively at the 2-substituted pyridine. This dearomative epoxidation can also tolerate alkyl-substituted pyridines (14f, 14n, and 14o), although benzylic C–H amination with photoexcited MTAD was a notable side reaction²⁷ in the case of 2-isopropylpyridine (2o). Finally, 2-methoxypyridine and several substituted derivates performed well and provided the corresponding oxides 14h, 14p–14r. Our epoxidation protocol delivered products as single constitutional and anti-diastereoisomers (see Figure 3a for X-ray of 13a), and N-oxidation was not observed.

Figure 3.

Arenophile-mediated dearomative epoxidation of pyridines. (a) Reaction scope. (b) Cycloreversion and derivatization. See Supporting Information for additional details.

The described epoxidation strategy streamlines the preparation of pyridine oxides and represents a significant advance from previous 5- and 6-step methods using electrocyclization or fragmentation of pyridines and 2-pyridinones.²⁴ Our protocol also offers an improved scope as previous methods were limited to few substrates. The only reactivity of pyridine oxides previously reported is their rearrangement to 3-hydroxypyridines, which parallels arene oxides through a NIH shift-like pathway.^24,28 We planned to leverage the equilibrium between pyridine oxides and 1,4-oxazepines to further functionalize both tautomers (Figure 3b). First, 6-phenylpyridine 3,4-oxide (14a), prepared from bicycle 13a and used without purification, was exposed to nitrosobenzene to deliver bicyclic hydroxylamine derivative 16. Subsequent reduction of the N─O bond and aminal afforded aminohydroxylated piperidine 17. Additionally, 5-methoxy-7-phenyl-1,4-oxazepine (15r) was successfully captured by hydrogenolysis to provide partially reduced dihydro-1,4-oxazepine derivative 18.

In summary, we have developed a distinct dearomatization of pyridines based on chemoselective olefin oxidations. This strategy provides direct entry to dihydropyridine cis-diols and pyridine oxides through one-pot photochemical para-cycloaddition of pyridines and subsequent dihydroxylation or epoxidation. Our protocol was demonstrated on a variety of substituted pyridines and delivered products that are amenable to diverse downstream functionalizations. This dearomative method does not require substrate preactivation, avoids decomposition of sensitive products, and enables the installation of heteroatom functionalities into piperidine scaffolds. Given the scarcity of similar dearomative approaches in chemistry and biocatalysis, we anticipate this method will be an important contribution to the preparation of high-value heterocycles and further development of arenophile-mediated approaches towards heteroarenes.