Direct Access to the Simplest N‑Substituted Pyrroles from Furans Enabled by Tailored Photocatalysts

Abstract

N-Methylpyrroles are important motifs in pharmaceuticals and biologically relevant compounds, yet their synthesis commonly relies on direct N-methylation of N–H pyrroles, a process often limited by the instability of pyrrole and the use of toxic alkylating reagents under harsh conditions. Herein, we report a photocatalytic strategy for the direct conversion of readily available furans to N-methylpyrroles. The reaction efficiency was markedly improved compared to the first-generation catalytic protocol through the development of tailored photocatalysts. Combining skeletal editing with continuous-flow photochemistry, this method provides efficient access to N-methylpyrroles under mild conditions while avoiding preformed N–H pyrrole substrates and conventional methylating agents. The utility of the protocol was further demonstrated through the late-stage diversification of structurally complex molecules.

Introduction

Furan and pyrrole are among the most fundamental five-membered aromatic heterocycles, and the two are formal isosteres that differ by only a single ring atom: an oxygen in furan versus a nitrogen in pyrrole. This seemingly minor distinction provides a profound divergence in how the two scaffolds engage biological targets. , The furan oxygen can serve only as a hydrogen-bond acceptor, whereas the pyrrole nitrogen bears an N–H that can act as a hydrogen-bond donor or be substituted to install additional functionality. This feature endows pyrrole with a dimension of target engagement, electronic tuning, and lipophilicity control, and it underlies the complementary roles the two scaffolds play across synthetic and medicinal chemistry. − For example, furan is among the most abundant oxygen heterocycles in natural products, yet it remains comparatively scarce among approved drugs. In contrast, pyrrole displays the inverse and more medicinally consequential profile: less conspicuous across the small-molecule natural-product inventory but a privileged scaffold that pervades modern pharmaceuticals, where it frequently functions as a key pharmacophore (Figure A). − Its pharmaceutical prominence motivates a closer look at how its nitrogen is functionalized.

1.

Prevalence of pyrroles in bioactive molecules and synthetic strategies for N-methylpyrroles.

N-Methylpyrrole, the simplest N-substituted pyrrole, occupies a privileged position that has drawn much attention in drug discovery. − N-Methylation is the most atom-economical and structurally conservative way , to cap the pyrrole nitrogen: It removes the N–H hydrogen-bond donor and markedly enhances oxidative and metabolic stability while leaving the aromatic core, sterics, and π-electronics essentially intact, thereby enabling fine modulation of polarity and electronic character with minimal structural perturbation. − , These attributes have made the N-methylpyrrole motif a cornerstone of both medicinal chemistry and the chemistry of complex natural products (Figure B). − It is embedded in clinically established nonsteroidal anti-inflammatory drugs of the pyrrole-acetic acid class, most notably the antiarthritic agent tolmetin, in which the N-methylpyrrole core is essential for activity, with the structurally related zomepirac. It also recurs throughout intricate, bioactive marine alkaloids, including the mauritamides, agelanin B, and citharoxazoles, many of them produced by organisms such as Moorea producens.

Despite this prevalence, the N-methylpyrrole motif is conventionally installed by S_N2-methylation of a preformed free N–H pyrrole, an approach that carries several intrinsic liabilities (Figure C). The electron-rich, sterically exposed pyrrole nitrogen is prone to overalkylation at the C2 position and oxidative degradation to form ill-defined “pyrrole black” materials that erode yield and purity; − the requisite free N–H pyrrole precursors are themselves unstable and troublesome to handle and store; and such reactions typically demand toxic alkylating reagents such as methyl iodide together with strong bases, raising operational, safety, and environmental concerns that become especially acute in the late-stage elaboration of complex substrates.

In this work, we address these long-standing limitations by assembling the N-methylpyrrole ring from the corresponding furan via an emerging structural reorganization strategy for heterocycle diversification. − We report a new photocatalytic platform in both batch and continuous-flow formats that constructs N-methylpyrroles directly from bench-stable furans, formally exchanging the furan oxygen to a N-methyl-bearing nitrogen (Figure D). Because the N-methylpyrrole ring is built from an abundant and readily available furan feedstock, this strategy − circumvents the two central liabilities of conventional routes: it does not traverse an unstable free N–H pyrrole intermediate, and it obviates the toxic alkylating reagents and strong bases that plague S_N2-methylation. By developing new sets of the next-generation photocatalysts, the transformation proceeds under mild photocatalytic conditions with broad functional-group tolerance.

Results and Discussion

We recently reported a photocatalytic furan-to-pyrrole conversion wherein a furan and a primary amine are merged into an N-substituted pyrrole through a formal oxygen-to-nitrogen transmutation. Mechanistically, this transformation is initiated by photocatalytic single-electron oxidation of the electron-rich furan, and the resulting radical cation is intercepted by the amine; subsequent ring opening and recyclization with extrusion of the original oxygen then forge the new C–N bonds that deliver the pyrrole. At the outset, we hypothesized that this platform might be extended to N-methylpyrrole synthesis simply by employing methylamine as the nitrogen source. In practice, however, subjecting a furan 1a to the conditions from our previous work with methylamine hydrochloride as the methylamine source afforded none of the desired N-methylpyrrole 2a, while 1a was almost fully consumed (eq ). The high conversion of 1a without product formation indicated that the reported first-generation system indeed oxidizes the furan but fails to channel methylamine into the transmutation pathway, presumably due to the intrinsically weaker nucleophilicity. We therefore recognized that accessing N-methylpyrroles would require next-generation protocol specifically tailored to methylamine as the nitrogen source, which sets the stage for the development of the photocatalyst library.

1

During the course of extensive reaction optimization, we observed that changing the photocatalyst to PC2, an acridinium photocatalyst originally reported by the Glorius group and known for its high photothermal stability, afforded the desired N-methylpyrrole product 2a (Table ). Specifically, slight excess of methylammonium salt and NaOAc as a base in a 2:1 mixture of acetonitrile (MeCN) and water delivered the most effective reaction outcomes (entry 1). Solvents and their composition also played a crucial role in determining the reaction efficiency: when MeCN was used as the sole solvent, the yield decreased significantly, probably due to solubility issue of the ammonium salt in a pure organic solvent. The choice of base also exerted a profound impact on the reaction efficiency. Compared to the optimal NaOAc, the use of NaHCO₃ afforded a moderate yield of 72% (entry 2), while Na₂CO₃ led to a substantial decrease in yield (entry 3). Na₂HPO₄ also gave excellent reactivity (entry 4). We attribute this diminished reactivity to the stronger basicity of Na₂CO₃, which promotes the formation of free methylamine from the corresponding ammonium salt. Under the heated reaction conditions, the highly volatile free methylamine can readily escape from the reaction vessel, decreasing its effective concentration and thereby lowering the reaction efficiency. Finally, control experiments conducted in the absence of either photocatalyst (entry 5) or light (entry 6) yielded no product, confirming a photocatalytic reaction pathway consistent with our previous study.

1. Reaction Optimization .

^aReaction conditions: 1a (0.10 mmol), MeNH₃Cl (0.24 mmol), PC2 (8 μmol), and NaOAc (0.12 mmol) in MeCN:H₂O (2:1) (2.0 mL) under 456 nm irradiation at 60 °C for 24 h.

^bYields based on ¹H NMR analysis of the crude mixture with CH₂Br₂ or 1,3,5-trimethoxybenzene as an internal standard.

While reaction optimization with the 3-substituted furan 1a proved successful, applying the otherwise identical conditions to 2-phenylfuran 1b gave poor reactivity, delivering only 21% of the desired pyrrole 2b. This diminished efficiency was disappointing considering the importance of 2-substituted furans and pyrroles in medicinal chemistry and materials science. We therefore set out to further optimize the reaction, focusing on the structure–reactivity relationship of the photocatalyst (Figure ). We first surveyed a series of acridinium photocatalysts bearing different substitution patterns (Figure A). The parent Fukuzumi acridinium photocatalyst , (PC1) afforded only low yields (15% and 7%). Inspired by the catalyst diversification strategy of the Nicewicz group, we then examined a range of structurally modified acridinium salts, preparing a library of both known and new photocatalysts spanning various substitution patterns (PC3-PC7). Sterically protected acridinium photocatalysts have been reported to be less susceptible to nucleophilic/radical addition and demethylation, thereby improving catalyst robustness under photocatalytic conditions. From this set, we identified a promising new catalyst PC3 bearing an electron-deficient CF₃ group, whose structure was unambiguously established by single-crystal X-ray diffraction analysis. Notably, PC3 smoothly converted 2-phenylfuran 1b to pyrrole 2b in 65% yield in a 199:1 mixture of MeCN and H₂O. By contrast, installing an electron-donating group or a 2-bromo substituent on the catalyst (PC4 and PC5) gave poor performance for both 1a and 1b, while catalysts bearing bulkier aryl or extended π-systems (PC6 and PC7) retained high reactivity toward 1a but performed worse with 1b.

2.

Reactivity in batch (A) and flow (B) conditions, and application to previously reported reactions (C). ^aMeCN:H₂O = 199:1. ^b In MeCN, using MeNH₂(40% in MeOH, 0.48 mmol). (i.d., internal diameter).

Notably, translating the optimized batch protocol to continuous-flow , conditions enabled room-temperature operation and dramatically accelerated the transformation, shortening the reaction to a residence time (τ) of less than 3 h, nearly an order of magnitude faster than the previously reported batch process with other amines (Figure B). Moreover, the product was isolated in an excellent 85% yield, demonstrating that the switch to a flow format preserves high catalytic efficiency without any significant loss in reactivity. Reducing the catalyst loading to 4 mol % led to a significant decrease in reactivity, affording 2a in only 18% yield.

To assess the broader applicability of the newly developed photocatalysts, we next evaluated PC2 and PC3 using representative primary amine substrates from our previous study (Figure C). Gratifyingly, both photocatalysts outperformed PC1, with PC3 affording the highest reactivity. For representative low-yielding substrates, the yields increased from 35% and 42% with PC1 to 81% and 75% with PC3, respectively. These results indicate that the catalyst design developed in this work is broadly applicable beyond methylamine-based transformations.

To better understand the superior performance of PC3, we carried out experimental mechanistic studies. We first recorded its photophysical properties (Figure A). The UV–vis absorption spectrum of PC3 displays a broad band extending into the visible region (350–480 nm) and centered at approximately 377 nm. Bright and broad emission centered at 504 and 523 nm was observed, and time-resolved photoluminescence analysis, performed with an excitation wavelength (λ_ex) of 405 nm and monitored at the emission maximum (λ_em) of 523 nm, revealed a clean monoexponential decay corresponding to an excited-state lifetime (τ_PL) of 23.2 ns. This confirms that PC3 possesses an adequate excited-state window to engage with the surrounding species.

3.

Mechanistic and photophysical studies. (A) Electronic absorption and emission spectra of PC3 in acetonitrile at room temperature. Inset: Photoluminescence decay of PC3 (τ_PL = 23.2 ns). (B) Stern–Volmer quenching plots of PC2 and PC3 with 1b and MeNH₃Cl:NaOAc (2:1). Numbers above and below each fitted line denote the Stern–Volmer constant (K _SV) and the corresponding R ² value, respectively.

To directly compare the photoinitiation processes, we next conducted Stern–Volmer quenching studies with PC2 and PC3 (Figure B). Interestingly, the Stern–Volmer quenching constant for PC3 was markedly larger than that for PC2 when 2-phenylfuran 1b was used as the quencher (K _SV = 139 vs 50 M^–1), indicating that PC3 interacts preferentially with the substrate. We reasoned that the hydrophobic 2-(trifluoromethyl)­phenyl moiety lowers the polarity of the catalyst, thereby promoting favorable donor–acceptor π–π interactions with furan 1b. Conversely, PC3 was quenched less efficiently than PC2 by the polar methylammonium salt (K _SV = 14 vs 29 M^–1). Although ionic acridinium photocatalysts are known to engage in electrostatic or ion-pairing interactions that facilitate electron transfer, ,, our new system instead benefits from suppressing such associations, thereby avoiding premature amine oxidation. Accordingly, we attribute the superior performance of PC3 under the less polar conditions to its favorable interaction with the nonpolar furan substrate over the polar methylammonium salt, which suppresses unproductive amine oxidation while promoting productive coupling.

With the optimized conditions in hand, we evaluated the substrate scope under continuous-flow conditions (Table ). We first examined the effect of the electronic properties of the aryl substituent at the 3-position of the furan ring. The model substrate 1a delivered the corresponding pyrrole 2a in 85% yield under the flow protocol, and a fluoro-substituted derivative (1c) was also well tolerated. Under slightly elevated reaction temperatures, the electron-rich methyl-substituted substrate 1d afforded the corresponding N-methylpyrrole 2d in 81% yield. Notably, strongly electron-withdrawing substituents such as cyano- and trifluoromethyl groups were likewise compatible, providing the desired products 2e and 2f in high yields. In addition, a variety of N-heteroaryl substituents could be incorporated successfully, furnishing products 2g and 2h. Notably, they could be converted selectively without any detectable methylation of the pyridine nitrogen, highlighting the excellent chemoselectivity of the present strategy. A naphthyl-substituted substrate was also amenable to the transformation, although mild heating was required to obtain the corresponding product 2i in moderate yield.

2. Substrate Scope .

^aReaction conditions: 1 (0.20 mmol), MeNH₃Cl (0.48 mmol), PC2 (16 μmol), and NaOAc (0.24 mmol) in MeCN:H₂O (2:1) (4.0 mL) under 456 nm irradiation; flow reactor (i.d. 0.75 mm, length 1.8 m, reactor volume 0.795 mL, τ 120 min, flow rate 6.63 μL/min) equipped with a 40 psi back-pressure regulator. Room temperature (laboratory temperature: 21–25 °C); reactor temperature during irradiation: 30–35 °C (average 33 °C), measured by an infrared thermal camera. Isolated yields after flash column chromatography.

^bAt 80 °C.

^cIn MeCN, using MeNH₂ (40% in MeOH, 0.48 mmol) and PC3 instead of MeNH₃Cl/NaOAc and PC2, respectively.

We next turned our attention to 2-aryl-substituted furans. Although reactivity was diminished relative to the 3-substituted series, moderate to good levels could be restored through fine-tuning of the reaction conditions. In particular, the model substrate 1b afforded the desired N-methylpyrrole 2b in up to 57% yield with PC3 as the photocatalyst. Halogen- and trifluoromethyl-substituted substrates (1j and 1k) were likewise tolerated, as was a meta-substituted aryl substrate (1l), which furnished the corresponding product 2l in synthetically useful yield. 2-Heteroaryl-substituted furans (1m–1o) also proved to be effective substrates, providing the corresponding N-methylpyrroles 2m–2o in modest yields. The successful conversion of nucleophilic pyridyl- and other N-heterocycle-containing substrates further underscores the broad substrate compatibility of the present protocol. Notably, PC3 delivered approximately twice the yield of PC2 in the conversion of 2m, reinforcing the advantage of the optimized photocatalyst system for challenging substrates. Difunctionalized furans bearing substituents at both the 3- and 4-positions (1p and 1q) were also competent, affording the corresponding N-methylpyrroles 2p and 2q in 51% and 65% yield, respectively. Likewise, a 2,4-disubstituted furan (1r) was successfully converted to the corresponding N-methylpyrrole (2r), further demonstrating the tolerance of the transformation toward multiply substituted furans. We next explored the reactivity of alkyl-substituted furans. A 3-benzyl-substituted furan (1s) underwent the transformation smoothly to afford the desired 2s in 44% yield without detectable benzylic oxidation. Alkyl substitution at the 2-position was also tolerated. A thioester-substituted furan (1t) could be transformed to desired product 2t in up to 30% yield.

To demonstrate the practical utility and scalability of this platform, we conducted a continuous-flow scale-up synthesis. Remarkably, 383 mg of the desired product was isolated after 8 h of continuous operation under our system, corresponding to a productivity of more than 1 g per day (Figure A), establishing the suitability of the method for rapid, preparative-scale synthesis. Encouraged by these results, we explored the applicability of the protocol to the late-stage modification of structurally complex molecules (Figure B). Owing to their limited solubility under continuous-flow conditions, these substrates were evaluated in batch. Notably, the O-to-N transmutation was successfully applied to limonin, a highly intricate furan-containing natural product renowned for its anti-inflammatory properties, affording the corresponding N-methylpyrrole derivative 2u in 81% yield. Likewise, diosbulbin B, a structurally complex furanoid natural product known for its potent antitumor activity, proved compatible with the transformation, delivering the desired product 2v in 70% yield. Together, these results highlight the exceptional functional-group tolerance and late-stage synthetic utility of the present protocol, even within highly complex molecular architectures.

4.

Practical applications of photocatalytic N-methylpyrrole synthesis.

Conclusion

In conclusion, we have developed a robust and efficient photocatalytic platform for the direct synthesis and late-stage modification of N-methylpyrroles. A key achievement in this study was overcoming the intrinsic limitations of conventional N-methylation by gaining mechanistic insights into the reaction components. Specifically, we demonstrated that attenuating the undesirable electrostatic or ion-pairing interactions between the photocatalyst and the highly hydrophilic methylammonium salt is critical for achieving efficient and selective furan oxidation. The synthetic utility of the method was further demonstrated through the late-stage modification of structurally complex natural products under batch conditions, while the continuous-flow protocol enabled rapid and scalable synthesis under mild reaction conditions. Together, these advancements provide a sustainable, atom-economical, and operationally simple alternative to conventional N-methylation methods that rely on toxic alkylating reagents.

The data supporting this article have been included as part of the SI.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.6c01035.

• General procedures for the synthesis of PC, 1, and 2; ¹H, ¹³C, and ¹⁹F NMR spectrum of PC,1, and 2 in CDCl₃ at 25 °C; Stern–Volmer quenching data of PC1 and PC2 by varying concentrations of 1b in MeCN; crystal data and structure refinement details for PC3 (PDF)

∥.

H.B. and H.H. contributed equally to this work.

The authors declare no competing financial interest.

Published as part of JACS Au special issue “Molecular Editing in Natural Products and Drug Molecules”.