An Oxygen Walk Approach for C3 Selective Hydroxylation of Pyridines

Abstract

Selective C3 functionalization of unbiased pyridines represents a significant challenge in organic synthesis. While seminal work in this area has enabled access to various C3-substituted pyridines via dearomatized intermediates, the direct introduction of a hydroxy group at this position is still challenging. In this context, we have developed a valence isomerization reaction triggered by photoexcitation of pyridine N-oxides to deliver synthetically challenging C3-hydroxy pyridine products.

1 Introduction

2 Recent Advances in Radical-Based Pyridine Functionalization via Pyridine N-Oxides

3 C3-Selective Hydroxylation of Pyridines through Oxygen Walking

4 Conclusion

 

MissingLinkText

Key words

pyridine N-oxides - hydroxylation - oxygen walk - photorearrangement - late-stage functionalization

1

Introduction

Pyridines are ubiquitous structural units in pharmaceuticals, agrochemicals, materials, and ligands for catalysts.[1] Substituents off the periphery of the pyridine nucleus provide a broad spectrum of physicochemical properties, adding to the versatility of pyridines. Along these lines, developing methodologies to directly functionalize the C–H bonds of unbiased pyridines represents an endeavor of particular value.[2]

Despite the significant progress in this field, decorating the C3 position of pyridines is still burdensome compared with the corresponding C2 and C4 positions, partially due to innate reactivity.[3] To date, robust means for C3 functionalization include transition-metal catalyzed reactions using designed catalyst systems[4] and transformations leveraging dearomatized intermediates having a reversed or enhanced reactivity compared with their nonoxidized precursors (e.g., C3 halogenation of pyridines through Zincke imine intermediates).[5] Despite these paradigm-shifting advancements, challenges linger in this field. For example, general and concise methods for installing a hydroxy group at the C3 position of unbiased pyridines remain elusive.

Tian Qin received his Ph.D. from Boston University in 2015 with John A. Porco, Jr. He completed his postdoctoral research with Phil S. Baran at the Scripps Research Institute, and started his independent group at UT Southwestern Medical Center in Dallas in 2018. Chen-Yan Cai received her Ph.D. under the guidance of Professor Hai-Chao Xu at Xiamen University in 2022, focusing on synthetic electrocatalysis. In October 2022, she began her postdoctoral appointment with Professor Tian Qin at UT Southwestern Medical Center in Dallas, where her research is focused on the development of novel transformations to access synthetically challenging targets.

In recent years, the renaissance in organic radical chemistry has spawned a tremendous growth in the chemistry community. Against this backdrop, pyridine N-oxides (serving as activated derivatives of pyridines) have been broadly used to provide C2 or C4 functionalized pyridine products, acting as reactive and readily available radical acceptors and/or precursors under mild conditions (Figure [1]).[6]

Figure 1 Synthetic applications of pyridine N-oxides in radical-based pyridine C–H functionalizations

In this article, we highlight our work on the C3 selective hydroxylation of pyridines through a photoactivated oxygen walk.[7] Additionally, we summarize some recent developments in radical-based pyridine C–H bond transformations through the intermediacy of pyridine N-oxides. This is by no means a comprehensive survey of this exciting field, as several excellent reviews have already been published.[6]

Scheme 1 Representative examples of pyridine C–H functionalization using pyridine N-oxides as radical acceptors

 

MissingLinkText

2

Recent Advances in Radical-Based Pyridine Functionalization via Pyridine N-Oxides

Pyridine N-oxides are readily available feedstock chemicals that display an ambiphilic character. They can therefore serve as electrophiles or nucleophiles in numerous ionic-type transformations.[6d] [8] Zwitterionic pyridine N-oxides are net neutral; the negatively charged oxygen attached to the ring nitrogen renders the aromatic nucleus more electron-deficient than pyridine but more electron-rich than their pyridinium counterparts, due to the existence of a π-backdonation of LP_O → π*_CN.[9] As such, zwitterionic pyridine N-oxides display a unique reactivity and regioselectivity in Minisci-type radical reactions (Scheme [1a]).[6e]

Scheme 2 Representative examples of pyridine C–H functionalization by single-electron transfer of pyridine N-oxide species

For example, in 2015, Wu and co-workers reported an oxidative methylation of the C–H bonds of pyridine N-oxides (Scheme [1b]).[10] Their initial attempt to use pyridines as substrates was unsuccessful, whereas pyridine N-oxides worked smoothly in this coupling. In addition, incorporating pyridine N-oxides as a complementary activation mode to the classic Minisci-type transformation,[11] which typically requires a strong Brønsted acid (e.g., TFA, HCl) to activate the pyridines, is beneficial. In this vein, Postigo and co-workers developed a method for the fluoroalkylation of heteroarene N-oxide derivatives that employs rose bengal as an organic photocatalyst, affording products in excellent yields (Scheme [1c]).[12] Moreover, the reactivity of pyridine N-oxides can be amplified effectively by Lewis acids. In 2017, Baran and co-workers described a general iron-promoted Minisci-type alkene hydrofunctionalization (Scheme [1d]).[13] In that reaction, heterocyclic N-oxides combined with BF₃·Et₂O furnished products in typically higher yields and with improved regioselectivities. Noteworthily, the pyridine N-oxide products were deoxygenated in situ, obviating the need for a separate reduction step.

More-recent developments within this area have expanded the scope of pyridine N-oxides beyond acting just as radical acceptors. In 2018, Stephenson and co-workers reported a decarboxylative alkylation of heteroarenes under redox-neutral conditions.[14] A transient intermediate arising from the acylation of the pyridine N-oxide oxygen serves as both the coupling partner and radical source (Scheme [2a]).

In addition, synthetic applications employing pyridine N-oxyl radical cations generated through facile single-electron oxidation of pyridine N-oxides by photochemical or electrochemical processes have emerged recently.[6b] [c] Pyridine N-oxyl radical cations are highly reactive species that can react with alkynes, leading to alternative reactive vinyl radical intermediates that then undergo radical cascade processes to form C2 alkylated or acylated pyridines (Scheme [2b]).[15] Furthermore, pyridine N-oxides can also be used as versatile hydrogen-atom transfer (HAT) catalysts, taking advantage of their tunability in the site-selective transformation of nonactivated C(sp³)–H bonds (Scheme [2c]).[16] These oxygen-centered radicals can also react with easily accessible and commercially available alkyl boronates to provide C-centered radicals (Scheme [2d]).[17]

Scheme 3 Complementary options for pyridine C–H functionalization via diradical intermediates

As redox-active species, zwitterionic pyridine N-oxides have found rich applications in single-electron transfer (SET) processes. Nevertheless, whereas the photochemical properties of pyridine N-oxides have been studied since the nineteenth century, diradical intermediates formed through a formal intramolecular electron transfer upon direct photoexcitation have scarcely been discussed in the synthetic-chemistry literature.[18]

The singlet photochemistry of pyridine N-oxides has often been understood as involving an oxaziridine intermediate generated through an intramolecular radical combination that moves toward a range of photorearrangement products (Scheme [3b]).[18b] In 2019, the Harran group achieved greatly increased yields of ketopyrrolophane (to 25%) through flow-based photolysis. This process was performed on gram scales as a key step in an asymmetric synthesis of marineosin A (Scheme [3c]).[19] While in their triplet states, diradical species formed from pyridine N-oxides tend to engage in intermolecular processes. Oxygen transfer to solvent occurs through homolytic cleavage of the N–O bond.[20] More recently, Deng and co-workers used 4-nitropyridine N-oxide as a bifunctional catalyst to generate alkyl radicals from boronic acids without additional photocatalysts, overriding the competitive nonphotochemical process and thereby demonstrating their synthetic potential (Scheme [3d]).[21] Intrigued by the versatility of these intermediates, we undertook an exploration of the synthetic applications of diradicals of pyridine N-oxides.

 

MissingLinkText

3

C3-Selective Hydroxylation of Pyridines through Oxygen Walking

Considering the growing importance of C3-oxygen-substituted pyridines in medicinal chemistry, we sought to develop a method for introducing a hydroxy group at this position.[22] Inspired by the underexplored rearrangement processes of pyridine N-oxides under UV irradiation, and mindful of their unpredictable reactivity, we envisaged that the reactive intermediate II (Scheme [3b]) might be harnessed to provide C3-hydroxylated pyridine products. The desired direction of oxygen walking might be driven by an irreversible, acid-promoted, epoxide opening.[23]

Our research commenced with pyridine N-oxide as the model substrate. To reach its excited singlet states through a π–π* transition efficiently,[18b] we selected 254 nm as the irradiation wavelength. Further optimizations revealed that both the solvent and acid additive are crucial for achieving high yields and good regioselectivity (Scheme [4], inset). A combination of (F₃C)₃COH and AcOH facilitated the most effective hydroxylation. The desired 3-hydroxypyridine (1) was isolated in 64% yield, along with the easily separable C2-hydroxylated byproduct in 18% yield.

Scheme 4 Selected examples of photochemical conversions of pyridine N-oxides into C3-hydroxy pyridines. ^a The yield and regioisomeric ratios were determined by ¹H NMR analysis with CH₂Br₂ as an internal standard. ^b AcOH instead of (F₃C)₃COH. ^c HFIP instead of (F₃C)₃COH.

This reaction was demonstrated with a broad selection of substrates containing various functional groups. Both electron-deficient and electron-rich groups were well tolerated (Scheme [4]). Compared with C4-monosubstituted pyridine N-oxides (2–7), C3-monosubstituted substrates showed diminished yields of the desired C5 products due to competitive C2 and C6 hydroxylations (8). For C2-monoalkyl-substituted pyridine N-oxides, three regioisomers of 9 were generated, but C5-pyridinols were obtained in higher yields. The C2-phenyl substituted substrate, however, delivered a reversed regioselectivity (10). Further exploration of di- and trisubstituted substrates with different substitution patterns also provided satisfactory results (11–16). Despite the moderate regioselectivity observed in several cases, each of the hydroxylated pyridines, which are usually challenging to access, could be delivered in one pot. Moreover, the methodology proved effective in the hydroxylation of the pyridine core of several medicinally relevant molecules, such as tropicamide (17), (S)-cotinine (18), pioglitazone (19), etofibrate (20), vismodegib (21), and abametapir (22) (Scheme [5]), which further promised its broader applications in drug discovery.

Scheme 5 Selected examples of late-stage diversification of medicinally relevant molecules. ^a HFIP instead of (F₃C)₃COH

Scheme 6 Synthetic derivatizations of C3-hydroxylated products

The synthetic utility of 3-hydroxypyridines was further demonstrated in a variety of downstream functionalizations (Scheme [6]). Through sequential triflation and palladium-catalyzed cross-couplings with organic nucleophiles, the hydroxy group could be used as a synthetic handle to furnish various C3-functionalized pyridines 24–29. On the other hand, with the presence of a trimethylsilyl substituent adjacent to the triflate group, a 3,4-pyridyne intermediate could be generated, which could be rapidly derivatized to form highly complex structures such as 30. Moreover, the benzylated pyridinium derivative of 23 was exploited in a [5+2] cycloaddition with vinyl sulfone to afford the azabicyclic [3.2.2] compound 31.

To validate a putative mechanism featuring an intramolecular oxygen-atom migration, we conducted ¹⁸O labeling studies and crossover experiments. The complete preservation of the isotopic label and the absence of detectable crossover products provided compelling evidence for a unimolecular process (Schemes 7a and 7b). The C5-hydroxylated product 33 was detected from 34, supporting a possible mechanism involving a 1,3-oxazepine intermediate (Scheme [7c]), as generally reported in the recent literature.[24]

Scheme 7 Mechanistic studies

Scheme 8 Proposed mechanism

Based on the mechanistic studies and literature precedents,[18b] [19b] we proposed a mechanism for this transformation (Scheme [8]). After irradiation, oxaziridine II is produced. Subsequent UV-light-promoted N–O bond homolysis generates the diradical intermediate II-int, which is transformed into the dearomatized and highly strained epoxide intermediate III. Meanwhile, in a manner analogous to a norcaradiene-type rearrangement,[25] a reversible photoinduced 6π electrocyclic ring expansion of III provides the 1,3-oxazepine IIIa. Finally, irreversible acid-promoted ring opening of the epoxy moiety and rearomatization gives the C3-hydroxylated pyridine product.

 

MissingLinkText

4

Conclusion

Whereas pyridine N-oxides have extensive synthetic applications in radical-mediated transformations, their diradical species, accessible by direct photoirradiation, remain underexplored in pyridine C–H bond functionalization. Here, we present an oxygen walking strategy for the efficient derivatization of pyridine N-oxides to give previously synthetically challenging C3-hydroxy pyridines. This operationally simple and scalable protocol features a broad substrate scope. Subsequent successful applications in late-stage functionalization of biologically relevant complex molecules and building-block derivatizations together highlight the practicality and robustness of this transformation. During the editorial process of this manuscript, an elegant approach related to this topic was published. Bhattacharya, D.; Studer, A. Angew. Chem. Int. Ed. 2025, e202423512.

 

MissingLinkText

 

MissingLinkText

Conflict of Interest

The authors declare no conflict of interest.

• References


For selected reviews and examples, see:
• 1a Bhutani P, Joshi G, Raja N, Bachhav N, Rajanna PK, Bhutani H, Paul AT, Kumar R. J. Med. Chem. 2021; 64: 2339
• 1b Kallitsis JK, Geormezi M, Neophytides SG. Polym. Int. 2009; 58: 1226
• 1c Zafar MN, Atif AH, Nazar MF, Sumrra SH, Gul-E- Saba, Paracha R. Russ. J. Coord. Chem. 2016; 42: 1

For selected reviews and examples, see:
• 2a Josephitis CM, Nguyen HM. H, McNally A. Chem. Rev. 2023; 123: 7655
• 2b Failla M, Lombardo GW, Orlando P, Fiorito D, Bombonato E, Ronchi P, Passarella D, Fasano V. Eur. J. Org. Chem. 2023; 26: e202300074
• 3 Cao H, Cheng Q, Studer A. Angew. Chem. Int. Ed. 2023; 62: e202302941

For selected reviews and examples, see:
• 4a Wright JS, Scott PJ. H, Steel PG. Angew. Chem. Int. Ed. 2021; 60: 2796
• 4b Larsen MA, Hartwig JF. J. Am. Chem. Soc. 2014; 136: 4287
• 4c Yang L, Uemura N, Nakao Y. J. Am. Chem. Soc. 2019; 141: 7972

For selected examples, see:
• 5a Boyle BT, Levy JN, de Lescure L, Paton RS, McNally A. Science 2022; 378: 773
• 5b Cao H, Cheng Q, Studer A. Science 2022; 378: 779
• 5c Muta R, Torigoe T, Kuninobu Y. Org. Lett. 2022; 24: 8218
• 5d Tian J.-J, Li R.-R, Tian G.-X, Wang X.-C. Angew. Chem. Int. Ed. 2023; 62: e202307697

For selected reviews and examples, see:
• 6a Proctor RS. J, Phipps RJ. Angew. Chem. Int. Ed. 2019; 58: 13666
• 6b Satheesh V, Deng Y. J. Org. Chem. 2024; 89: 11864
• 6c Ang HT, Miao Y, Ravelli D, Wu J. Nat. Synth. 2024; 3: 568
• 6d Habib I, Singha K, Hossain M. ChemistrySelect 2023; 8: e202204099
• 6e Oudeyer S, Levacher V, Beucher H, Brière J.-F. Molecules 2023; 28: 1071
• 7 Cai C.-Y, Chen S.-J, Merchant RR, Kanda Y, Qin T. J. Am. Chem. Soc. 2024; 146: 24257

For selected reviews and examples, see:
• 8a Kutasevich AV, Perevalov VP, Mityanov VS. Eur. J. Org. Chem. 2021; 2021: 357
• 8b Petrosyan A, Hauptmann R, Pospech J. Eur. J. Org. Chem. 2018; 2018: 5237
• 8c Wang Y, Zhang L. Synthesis 2015; 47: 289
• 9 Kobus M, Friedrich T, Zorn E, Burmeister N, Maison W. J. Med. Chem. 2024; 67: 5168
• 10 Li G, Yang S, Lv B, Han Q, Ma X, Sun K, Wang Z, Zhao F, Lv Y, Wu H. Org. Biomol. Chem. 2015; 13: 11184
• 11 Duncton MA. J. Med. Chem. Commun. 2011; 2: 1135
• 12 Lantaño B, Barata-Vallejo S, Postigo A. Org. Biomol. Chem. 2018; 16: 6718
• 13 Lo JC, Kim D, Pan C.-M, Edwards JT, Yabe Y, Gui J, Qin T, Gutiérrez S, Giacoboni J, Smith MW, Holland PL, Baran PS. J. Am. Chem. Soc. 2017; 139: 2484
• 14 Sun AC, McClain EJ, Beatty JW, Stephenson CR. J. Org. Lett. 2018; 20: 3487
• 15 Xu J.-H, Wu W.-B, Wu J. Org. Lett. 2019; 21: 5321
• 16 Schlegel M, Qian S, Nicewicz DA. ACS Catal. 2022; 12: 10499
• 17 Yue F, Li M, Yang K, Song H, Liu Y, Wang Q. Chem. Sci. 2024; 15: 14241

For selected reviews and examples, see:
• 18a Albini A, Alpegiani M. Chem. Rev. 1984; 84: 43
• 18b Poole JS. In Heterocyclic N-Oxides . Larionov OV. Springer; Cham: 2017: 111
• 19a Feng Z, Allred TK, Hurlow EE, Harran PG. J. Am. Chem. Soc. 2019; 141: 2274
• 19b Hurlow EE, Lin JB, Dweck MJ, Nuryyeva S, Feng Z, Allred TK, Houk KN, Harran PG. J. Am. Chem. Soc. 2020; 142: 20717

For selected reviews and examples, see:
• 20a Streith J, Danner B, Sigwalt C. Chem. Commun. (London) 1967; 979b
• 20b Basistyi VS, Frederich JH. Org. Lett. 2022; 24: 1907
• 21 Ascenzi-Pettenuzzo C, Nganda J, Deng Y. ChemCatChem 2023; 15: e202300953

For selected reviews and examples, see:
• 22a Liang J, Han Q, Tan Y, Ding H, Li J. Front. Mol. Biosci. 2019; 6: 4
• 22b Cramer J, Sager CP, Ernst B. J. Med. Chem. 2019; 62: 8915
• 23 Thirumalaikumar M. Org. Prep. Proced. Int. 2022; 54: 1
• 24 Woo J, Christian AH, Burgess SA, Jiang Y, Mansoor UF, Levin MD. Science 2022; 376: 527
• 25 Berson JA, Willcott III MR. J. Am. Chem. Soc. 1965; 87: 2751

Publication History

Received: 02 January 2025

Accepted after revision: 30 January 2025

Article published online:
14 March 2025