Peng
Ji
a,
Xiang
Meng
a,
Jing
Chen
a,
Feng
Gao
a,
Hang
Xu
a and
Wei
Wang
*abc
aDepartment of Pharmacology and Toxicology, R. Ken Coit College of Pharmacy, University of Arizona, USA
bDepartment of Chemistry and Biochemistry, University of Arizona, USA
cUniversity of Arizona Cancer Centre, University of Arizona, 1703 E. Mabel Street, Tucson, AZ 85721-0207, USA. E-mail: wang@pharmacy.arizona.edu
First published on 28th February 2023
Reductive dearomatization has been a broadly explored means for rapid generation of sp3 complexity from simple planar arenes. Breaking the electron rich, stable aromatic systems requires strong reduction conditions. It has been notoriously challenging to dearomatize electron even richer heteroarenes. Herein we report an umpolung strategy enabling dearomatization of such structures under mild conditions. The reversal of the reactivity of these electron rich aromatics via photoredox mediated single electron transfer (SET) oxidation generates electrophilic radical cations, which can react with nucleophiles and break the aromatic structure to form a Birch type radical species. A crucial hydrogen atom transfer (HAT) is successfully engineered into the process to efficiently trap the dearomatic radical and minimize the formation of the overwhelmingly favorable, irreversible aromatization products. Particularly, a non-canonical dearomative ring-cleavage of thiophene/furan through selective C(sp2)–S bond breaking was first discovered. The preparing power of the protocol has been demonstrated for selective dearomatization and functionalization of various electron rich heteroarenes including thiophenes, furans, benzothiophenes and indoles. Furthermore, the process offers an unrivaled capacity for simultaneously introducing C–N/O/P bonds on these structures as exemplified by various “N”, “O” and “P” centered functional moieties with 96 examples.
Recently we have achieved the selective dearomative functionalization of benzene in polycyclic arenes;19 however, it is challenging to selectively dearomatize a single ring heteroarene, including thiophene/furan, which represents a long-standing issue.1 In order to address this issue, we envisioned that integrating photoredox catalysis into hydrogen atom transfer (HAT) could create a new strategy for direct dearomative functionalization of electron-rich heteroarenes (Scheme 1b and e). In light of the electron rich nature of aromatic structures, we conceived that single electron transfer (SET) oxidation of the system by an excited photocatalyst (PC*) could conveniently generate a radical cation (I, Scheme 1b and e).20 The reversed reactivity would make the nucleophilic addition more easy and generated an aromatic radical (II), a key species in Birch reduction, which then could undergo direct reduction by PC˙− to give an anion. Subsequent protonation would give the functionalized dearomative product. Alternatively, engineering a new hydrogenation process [e.g., hydrogen atom transfer (HAT)] could trap the radical (II) to directly furnish the functionalized dearomatization product. However, the intrinsic thermodynamic and kinetic favorable aromatization (e.g., formation of V)21–27 makes this unprecedented dearomatization particularly difficult. Below, we describe the successful development of a general selective method for the dearomative functionalization of diverse heteroarenes. Moreover, unexpectedly, the dearomative ring-cleavage functionalization of thiophene/furan through selective C(sp2)–S/O bond breaking was discovered for the first time. The HAT process is crucial for overcoming the overwhelmingly favourable, irreversible aromatization.
Entry | Variation from the “standard conditionsa” | Yieldb (%) |
---|---|---|
a Standard conditions: a mixture of 2-methoxythiophene (1.5 equiv.), pyrazole (1.0 equiv.), Mes-Acr4 (2.5 mol%), PhSeH (0.2 equiv.), and 2,6-lutidine (0.2 equiv.) in DCM (0.05 M) under the irradiation of 40 W blue LEDs for 48 hours at room temperature. b Yield is determined by crude 1H NMR using 1,3,5-trimethoxybenzene as the internal reference. c Isolated yield. | ||
1 | None | 81 (77)c |
2 | Mes-Acr1 instead of Mes-Acr4 | 50 |
3 | Mes-Acr2 instead of Mes-Acr4 | 42 |
4 | Mes-Acr3 instead of Mes-Acr4 | 58 |
5 | TPT instead of Mes-Acr4 | Trace |
6 | DCA instead of Mes-Acr4 | Trace |
7 | 4CzIPN instead of Mes-Acr4 | Trace |
8 | PhSH instead of PhSeH | 56 |
9 | 4-NO2PhSH instead of PhSeH | 38 |
10 | DCM (0.025 M) | 82 |
11 | DCM (0.1 M) | 60 |
12 | DCM (0.2 M) | 30 |
13 | Without 2,6-lutidine | 68 |
14 | Without PhSeH | 23 |
15 | Without Mes-Acr4 | Trace |
16 | No light | Trace |
Having established the optimal reactions conditions, we explored the scope of the process with various heteroarenes including thiophenes, furans, benzothiophenes, and indoles (Scheme 2). The dearomative ring-cleavage functionalization of thiophenes using diverse pyrazoles was first investigated. Both electron donating and withdrawing groups are tolerated under the mild reaction conditions with moderate to good yield. They include iodide (4, 73%), bromide (5, 71%), chloride (6, 68%), fluoride (7, 54%), methyl (8, 78%; 16, 48%), tert-butyl (15, 70%), ester (9, 63%), boronic acid (10, 65%), methoxy (11, 61%), trifluoromethyl (12, 34%), phenyl (13, 58), ketone (14, 44%), disubstituted (19, 72%), trisubstituted (20, 67%; 21, 59%; 22, 55%), benzotriazole (17, 66%) and benzimidazole (18, 34%). Notably, boronic acid (10) can serve as a viable handle for cross coupling reactions in further synthetic elaboration. It appears that the Z/E selectivity is affected by the steric hindrance. A relatively high rr (>3:1) is observed with congested pyrazoles (15 and 19–21). In addition to monosubstituted thiophene, 2-methoxy, 3-methylthiophene (23, 25%) could also be dearomatized with reasonable yield. Notably exclusive E-configuration was formed in 23. Although low Z/E-selectivity was observed with thiophenes, 2-methoxyfuran (27–30) worked smoothly and delivered single Z-conformation.30 The simple 2-methoxythiophene could undergo the late-stage modification of complex molecules including pharmaceutically related heteroarenes (31, 71%; 32, 70%), ribose (33, 66%), and purine (34, 37%). As to polycyclic heteroarenes, benzothiophene (35) could selectively undergo dearomative non-cleavage hydroamination in the five-member ring at the 3-position, yielding 61% product with trans:cis = 8:1. Similarly, the simple indoles could selectively be dearomatized in the 2-position of the five member ring. Among three protecting groups (36–38), the sulfonamide delivered an excellent yield (94%). Diverse functional groups in the phenyl ring of indoles could be tolerated, such as methyl (39, 95%; 43, 86%; 46, 89%), halogens (40–42, 67–81%), ester (44, 92%), and aldehyde (48, 91%). Additionally, trans-products (49, 71%, 50, 91% and 51, 64%) were produced with 3-substituted indoles in high yields.
Scheme 2 Scope of heteroarenes. a Reaction conditions, unless specified, see Table 1, entry 1 and ESI.† Yields are isolated yields. bE/Z ratio was determined by crude 1H NMR. c Diastereoselectivity (dr) was determined by 1H NMR. |
Visible light-mediated photochemical dearomatization of indoles has been reported recently.31 These methods are limited to intramolecular and [4 + 2]/[2 + 2] cycloaddition processes.32–44 In contrast, intermolecular dearomative functionalization remains underexplored.45–47 Herein, we found that the intermolecular photoredox catalyzed dearomatization strategy shows a broader scope with a wide array of nucleophiles including amines, various azoles, carboxylates, alcohols, and phosphites. As shown in Scheme 3, a number of functional groups tethered to pyrazoles can be tolerated including ester (52, 94%), ketone (53, 82%), aldehyde (54, 67%), methyl (55, 91%), halogen (56–59, 91–95%), boronic acid (60, 77%), nitrile (61, 55%), trifluoromethyl (62, 63%), t-butyl (63, 88%), disubstituted (64, 95%; 65, 85%), and trisubstituted (66, 95%). Aside from pyrazoles, other common azoles, including imidazole (67, 95%), 1,2,3-triazole (68, 80%), 1,2,4-triazole (69, 83%), tetrazole (70, 93%), benzoimidazole (71, 84%), indazole (72, 79%), and benzotriazole (73, 96%), are also amenable in good to excellent yield. Amines such as benzylamine (74, 79%), 2-picolylamine (75, 67%), 2-aminomethyl)thiophene (76, 64%), and propargylamine (77, 51%) are also viable nucleophiles for direct functional dearomatization. Furthermore, remarkably, the less nucleophilic O-nucleophiles including carboxylic acids (78–84) and alcohols (85–87) could also work smoothly. These transformations are compatible with a number of functional groups such as cyclopropyl (80, 57%), alkyne (81, 79%), alkene (82, 83%), phenyl (83, 73%), azide (84, 73%), methanol (85, 89%; 87, 46%), and ethanol (86, 69%). Moreover, triethyl phosphite (88–91) was a validated nucleophilic agent, giving 2,3-dihydro-1H-indol-2-ylphosphonic acid in good yields (75–78%). The mild reaction conditions also enabled the late-stage modification of pharmaceutically relevant molecules such as aldenine (92, 63%), tryptamine (93, 81%), tyrosine (94, 90%), Oppolzer's camphorsultam (95, 74%), galactose (96, 89%), and 4-(1H-pyrazol-4-yl)-7H-pyrrolo[2,3-d]pyrimidine (97, 53%). In addition to intermolecular dearomatization, the intramolecular reaction could also work efficiently under the standard conditions, delivering the cyclized product (99, 83%, Scheme 3b). The reaction can be scaled up without loss of yield (Scheme 3c). These results clearly demonstrate the mildness, high regioselectivity, and broad generality of the newly developed synthetic manifold.
Scheme 3 Scope of nucleophiles. Reaction conditions, unless specified, see Table 1, entry 1 and ESI.† Yields are isolated yields. |
In the photoredox dearomative functionalization reaction, one of the important steps is the selective SET oxidation of heteroarenes. Although visible light mediated SET oxidation of arenes including indoles in organic transformations has been reported,21–27,31,48–56 this study represents the first example of oxidation of thiophenes, furans and benzothiophenes. The excited state PC+* Mes-Acr4* possesses strong oxidizing power (Ered* = 2.15 V)57 and can oxidize these substrates (2-methoxythiophene: Eox = 1.44 V; 2-methoxyfuran: Eox = 1.28 V, and N-SO2Ph indole: Eox = 1.90 V vs. SCE, Fig. S1–S3†) in principle (Scheme 4a using 2-methoxythiophene as an illustration example). We found that although the indole had high oxidation potential, the reaction occurred with it favorably in the mixture of indole and 2-methoxythiophene (Scheme 4b). A similar trend was observed with 2-methoxyfuran. Furans and thiophenes have higher resonance stabilization than indoles.29
Additional energy is needed to cleave the heteroarene ring bond, which was usually achieved via transitional-metal-catalyzed C–S insertion.58 These factors may pose a challenge for dearomative ring-opening functionalization of simple thiophenes and furans with photoredox catalysis. The electrophilic radical cation IIa reacts with pyrazole followed by subsequent deprotonation to form the radical IIIa (Scheme 4a), which readily undergoes the ring-opening β-scission and generate radical Va. The radical Va can abstract a hydrogen atom from PhSeH or undergoes the SET process, providing a non-classic dearomative ring-cleavage product. On the other hand, the radical IIIa can undergo thermodynamically favorable rearomatization20–27 to give functionalized aromatic product IVa. However, the application of the photoredox C–H functionalization of these heteroarenes has not been reported. In our effort, we further challenged the chemistry by redirecting to an overwhelmingly unfavorable dearomatized product. We found that the incorporation of a new HAT process can significantly inhibit the aromatized process and facilitate the dearomative product (Table 1, entries 1 and 14). The HAT agents are crucial for the process. Without it, only a small amount of product was formed (entry 14 in Table 1). In addition, the choice of HAT agents is also important. Selenol with a higher H-atom transfer rate (K20 = 1.3 × 109 M−1 s−1vs. PhSH: K20 = 9.0 × 107 M−1 s−1)59–61 gives the desired product in a better yield (entries 8 and 14 in Table 1). Furthermore, control experiments further validated the vital role of the HAT agent. In the absence of PhSeH, the standard conditions failed to deliver the dearomatized indole product (Scheme 4c). The lack of the HAT agent also significantly compromises the reaction efficacy in the reaction of 2-methoxyfuran. Finally, the deuteration experiment consolidates the HAT process engaged in the reaction (Scheme 4d). Taken together, these studies showed that the HAT is an important factor contributing to the success of the new dearomatization process.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2234192. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d3sc00060e |
This journal is © The Royal Society of Chemistry 2023 |