Liming Zhang‡
,
Chen Yang‡,
Xinhai Wang,
Taixin Yang,
Dandan Yang,
Yingchao Dou* and
Jun-Long Niu*
College of Chemistry, Pingyuan Laboratory, Zhengzhou University, Zhengzhou, Henan 450001, P. R. China. E-mail: yingchaodou@zzu.edu.cn; niujunlong@zzu.edu.cn
First published on 3rd September 2024
Inherently chiral calixarenes are known to exhibit versatile functions due to their delicate three-dimensional macrocyclic frameworks. However, the catalytic asymmetric synthesis of these compounds remains largely unexplored and poses a significant challenge. Herein, we report an unprecedented enantioselective electrochemical synthesis of inherently chiral calix[4]arenes. Our approach is based on a 3d metal cobalt-catalyzed asymmetric C–H acyloxylation of the prochiral macrocyclic frameworks. The easily accessible and modifiable chiral salicyloxazoline (Salox) was used as the ligand to efficiently regulate the enantioselectivity. This protocol proceeded smoothly under electrochemically mild conditions and was compatible with a wide range of carboxylic acids, including aryl carboxylic acids and tertiary, secondary, primary aliphatic carboxylic acids, yielding a variety of acyloxylated calix[4]arenes with good yields (up to 94% yield) and excellent enantioselectivities (95–99% ee). The synthetic practicability of this method was demonstrated by the scale-up reaction and the divergent derivatizations of the inherently chiral macrocyclic products.
Specifically, calix[4]arene is of enormous interest due to its moderate-sized cavity, stable conformation and easily accessible derivatizations.6 However, the problematic synthesis of enantiopure product severely limits application and further investigations. Therefore, the development of catalytic enantioselective synthesis methods for inherently chiral calix[4]arenes is highly attractive. While there have been sporadic reports of calix[4]arene synthesis achieved through lipase-catalyzed transesterification7 and palladium-catalyzed intramolecular coupling,8 it is worth noting that recently the Tong and Wang group described the enantioselective synthesis of ABCD-type heteracalix[4]aromatics9 via a palladium-catalyzed Buckwald–Hartwig reaction. Subsequently, two notable achievements (Fig. 1b) were reported almost simultaneously for the synthesis of inherently chiral calix[4]arenes.10,11 Both strategies involved noble metal-enabled enantioselective desymmetrization. Cai and coworkers10 achieved the enantioselective synthesis of calix[4]arenes with a fluorenone motif through a Pd-catalyzed intramolecular C–H arylation. Separately, the Tong group11 realized the Pd-catalyzed enantioselective synthesis of 9H-fluorene-embeded calix[4]arenes. A more complex cascade reaction sequence of oxidative addition, transannular 1,5-Pd migration and intramolecular C–H arylation was proposed to be involved. These syntheses mediated by precious 4d metal catalysts revealed a promising perspective for the synthesis of inherently chiral calixarenes. Despite these achievements, the development of more sustainable and practicable protocols is still challenging and desirable, particularly those enabled by economical and less toxic earth-abundant 3d metal catalysts with easily modified modular ligands.
The transition-metal-catalyzed C–H activation/acyloxylation has proved to be a direct and atom-efficient strategy for synthesizing oxygenated compounds,12–17 which are ubiquitous in natural products, pharmaceuticals and functional molecules. In recent years, electrochemically enabled asymmetric C–H functionalization catalyzed by earth-abundant 3d metal cobalt has emerged as an ideal and sustainable approach for generating various valuable enantiopure chiral compounds.18–28 Despite the relatively limited number of studies reported so far, its versatility and potential are undeniable. Significant breakthroughs have primarily focused on asymmetric constructions of chiral frameworks with central29–31 and axial chirality.32–39 However, there is a lack of literature on the catalytic asymmetric electrosynthesis of inherently chiral calixarenes.
To address this unprecedented challenge, we envisioned utilizing catalytic systems consisting of cobalt catalysts and diverse chiral ligands under electrochemical conditions. Specifically, we carefully evaluated the Salox ligands due to their potential for coordinating with cobalt catalyst, which could regulate the spatial orientation and enable enantioselective synthesis. Fortunately, the horizon of inherently chiral calix[4]arenes was successfully broadened via an electrochemically enabled cobalt-catalyzed Salox-mediated enantioselective acyloxylation, with the postulated octahedral catalytic model illustrated in Fig. 1c. Fruitful acyloxylated macrocycles were afforded with good yields and excellent enantioselectivities. The successful scale-up reaction and divergent derivatizations of the inherently chiral macrocyclic products demonstrated the synthetic practicability and versatility of this approach.
Entry | Catalysts | Additives | 3aa (%) | ee (%) |
---|---|---|---|---|
a Undivided cell, GF anode (15 mm × 10 mm × 6 mm submerged), platinum plated cathode (10 mm × 10 mm × 0.1 mm submerged), constant current of 2 mA; isolated yields are indicated; ee values were determined by chiral HPLC analysis.b 1a (0.1 mmol), 2a (0.2 mmol), Co(OAc)2·4H2O (10 mol%), ligand (20 mol%), NaOH (2 equiv.), 40 °C, 6 h, air, HFIP (5 mL). | ||||
1 | Co(OAc)2·4H2O | NaOH | 89 | >99 |
2 | Co(OAc)2 | NaOH | 85 | >99 |
3 | Co(OOCC6H5)2 | NaOH | 55 | 96 |
4 | Co(acac)2 | NaOH | N.R. | — |
5 | CoSO4·H2O | NaOH | Trace | — |
6 | Co(OAc)2·4H2O | AcOH | Trace | — |
7 | Co(OAc)2·4H2O | DBU | 85 | 96 |
8 | Co(OAc)2·4H2O | DMAP | 27 | 89 |
9 | Co(OAc)2·4H2O | Pyridine | 11 | 89 |
10 | Co(OAc)2·4H2O | Na2CO3 | 69 | 94 |
11 | Co(OAc)2·4H2O | tBuONa | 85 | >99 |
Fortunately, ligand L1 yielded 52% of product 3aa with a 94% ee value. It was hypothesized that the presumed π–π interaction (Fig. 1c) between substrates and phenyl group on the oxazoline motif could be crucial. To verify this hypothesis, the ligands L2–L4 with alkyl substitutes on the oxazoline motif were used, but they led to trace amounts of the product (Table 1 and Table S1†). Consequently, the electronic effect was examined using para-substituted ligands L5–L6, which resulted in higher yields and enantioselectivities with an electron-donating group (MeO). The steric hindrance effect was studied using the ligands L7–L9 with ortho-substituted electron-donating groups (Me, tBu, MeO), indicating that the bulky tBu substituent was ideal. The ligand L10, bearing a tBu substituent at the para- and ortho-positions of the phenol motif, was then tested, yielding 89% product with a >99% ee value. This is consistent with previous studies36–38 and the literature.34,35 No better results were observed after evaluating various cobalt salts (Table 1, entries 1–5) with different counterions such as Co(OOCC6H5)2, Co(acac)2 and CoSO4·H2O. The desired transformation was inhibited in the presence of acidic additives (AcOH, Table 1, entry 6). Then other organic (DBU, DMAP, pyridine) and inorganic (Na2CO3, tBuONa) basic additives were tested (Table 1, entries 7–11), but they led to lower yields and ee values compared to the case of NaOH. The influence of the solvent was evaluated by using several protic and aprotic solvents (Table S2†), but no improvement was observed. After further evaluations of other parameters, including current intensity, temperature, dosages of catalysts and ligands, reaction time and electrode materials (Tables S3-8 and S8-1†), the optimal conditions were found to be those in entry 1 of Table 1, yielding the most satisfactory results.
Fig. 2 Scope of the substrates (details are provided in the ESI†). |
Rationally, in terms of evaluating the steric hindrance influence of the acids on the reactivity and enantioselectivity, a series of secondary and primary aliphatic acids was subjected to the reaction (Fig. 2B and C), which were less bulky compared with the previously studied tertiary aliphatic carboxylic acids. The desired products were furnished with excellent enantioselective control and a slight decrease of yields (3ai–3aq, 31–78% yield, 96–>99% ee). Undoubtedly, the reaction of mono- and bisaryl substituted secondary carboxylic acids delivered the products in good yields (3ai–3ak, 64–73% yield). To our delight, excellent ee values were also observed (>99% ee). The dialkyl and cyclobutyl substituted secondary acids also reacted well to give the desired products (3al–3an, 54–78% yield, 99–>99% ee). Meanwhile, the absolute configuration of 3al was determined by X-ray diffraction analysis (see details in the ESI, Table S10, CCDC 2339942†). Notably, compound 3ak was obtained efficiently (64% yield, >99% de) from the chiral carboxylic acid 2k, illustrating that the chirality of substrates could be retained during the transformation. Satisfactory yields and excellent ee values were also obtained when we screened the primary aliphatic acids (3ao–3ap, 52–62% yield, 97–99% ee). Furthermore, several simple primary acids were also tested. No desired product was observed when acetic acid was used, and only trace product could be obtained in the case of propionic acid. The use of butanoic acid afforded 3aq in 31% yield and 96% ee.40
Subsequently, the compatibility of various aryl carboxylic acids with diverse functionalities was examined carefully (Fig. 2D, 3ar–3az). We were pleased to find that the para-substituted substrates with electron-donating (Me) and electron-withdrawing (CF3, Cl, CO2Me) groups could afford the corresponding acyloxylated calix[4]arenes (3ar–3av) in good yields (52–72%) and excellent enantioselectivities (95–99% ee). For meta- and ortho-substituted substrates (2w, 2x), the transformation took place smoothly to provide 3aw and 3ax in satisfactory yields (57%, 68%) and excellent ee values (>99%, 96%). Compound 3ay was also obtained starting from 3,5-dimethoxy substituted substrate. To our delight, 2-naphthoic acid was also compatible with this reaction, affording compound 3az in 62% yield and 99% ee. Furthermore, the influence of the modifications at the “lower rim” of the macrocycle was also evaluated (Fig. 2E). Pleasingly, the replacement of nPr with Et or nBu had no effect on the reaction efficiency, leading to calix[4]arenes 3ba and 3ca in high yields (78–94%) and excellent ee values (98%). The successful synthesis of intricate acyloxylated inherently chiral calix[4]arenes demonstrates the broad generality of this protocol.
To gain insights into the reaction, mechanism investigations were also conducted. The non-linear effect experiments (Fig. 3c) revealed a linear relationship between the ee values of the chiral ligand and the corresponding acyloxylated calix[4]arenes. Dynamic electrode potential analysis (Fig. 3d) demonstrated that a stable potential (1.33 V) could be maintained under the standard reaction conditions. Furthermore, cyclic voltammetry experiments (Fig. 3e) were performed and revealed that the oxidative potential of 1a in HFIP (Fig. 3e, no. 2) is consistant with the aforementioned results, being lower than those of the other four cases (no. 3–6): L10; L10 with NaOH; L10 with Co salt; and L10 with Co salt and NaOH. This observation could explain the absence of ligand decomposition in this current process. The electrolysis was irreversible, as the value of ipa/ipc was far less than 1, which is consistent with the absence of a reductive wave. The base NaOH could facilitate the oxidative process of Co(II) to Co(III), as well as the oxidation of the ligand, as evidenced by the decreased oxidation peaks observed with the addition of base. Notably, in the case (no. 7) of 1a, Co salt, L10 and NaOH, the oxidative potential was slightly lower than that of 1a alone. This indicates that the substrates could be activated by interactions between 1a and the catalytic species.
Based on the previous cobalt-catalyzed C–H activation of arylamides and the aforementioned explorations, we propose a plausible mechanism for this cobalt-catalyzed asymmetric electrochemical synthesis, as shown in Fig. 3f. Initially, the Co(II) salt coordinates with L10 and PivOH. The resulting complex undergoes in situ anodic oxidation to furnish the octahedral Co(III) species, denoted as Int-1. The ligand exchange with substrate 1a then forms Int-2. Subsequently, the pivotal stereocontrolling C–H activation and anodic oxidation of Int-2 produce the high-valent Co(IV), denoted as Int-3. The inherently chiral calix[4]arenes 3aa could be obtained following the reductive elimination, which also regenerates the Co(II) complex. It is assumed that the introduction of the directing group pyridine N-oxide and the phenyl substituent on the oxazoline motif are crucial and may be responsible for the success of this transformation. The presumed π–π interaction between them could dictate their spatial relationship. The enantioselective differentiation of the identical reactive sites on the macrocyclic scaffold is regulated under the electrochemical reaction conditions, as shown in Fig. 3f. The formation of Int-2′ and its corresponding Int-3′ is proposed to be unfavorable due to the steric hindrance between the Salox ligand and macrocycle.
The datasets supporting this article have been uploaded as part of the ESI.†
Footnotes |
† Electronic supplementary information (ESI) available: General information, experimental section, X-ray crystallographic data, compound characterization, and NMR spectra (PDF). CCDC 2339942 (3al). For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4gc02877e |
‡ These authors contributed equally to this work. |
This journal is © The Royal Society of Chemistry 2024 |