Chiral molecules, referring to three-dimensional molecules that are non-superimposable on their mirror images, have found significant applications in pharmaceuticals, agrochemicals, asymmetric synthesis and material science. Molecular chirality is generally classified into four categories of chiral elements: Central chirality, axial chirality, planar chirality and helical chirality (Fig. 1A). Calix[4]arenes are a type of privileged macrocyclic compounds, which have been extensively explored in supramolecular chemistry over recent decades, particularly for their applications in sensors, separations and catalysis [1–8]. The rigidly curved structure of calix[4]arenes engenders chirality through the asymmetric distribution of achiral substitutions within their frameworks, a concept coined as "inherent chirality" by Böhmer et al. in 1994 [9–11], which distinguishes it from the conventional four chiral elements. For instance, modifying of one phenyl ring of calix[4]arene with a meta-substitution [12], or altering two adjacent phenyl rings with different para-substitutions, would result in the unique inherent chirality (Fig. 1B) [13].

Given the significant potential of inherently chiral calix[4]arenes (ICCs) in chiral recognition, asymmetric catalysis and chiral material science [13–16], there is a pressing demand to develop efficient catalytic asymmetric methods for accessing these chiral compounds [17–21]. However, due to the unique scaffolds and relatively large size of calix[4]arenes, the early development of catalytic enantioselective methods met with limited success [[22], [23]]. In 2020, Tong, Wang and co-workers reported an elegant enantioselective Pd-catalyzed intramolecular C-N coupling method for the asymmetric synthesis of heterocalix[4]arenes [24–26]. However, due to the requirement of forming C-heteroatom bonds, this method is not applicable for conventional methylene-linked calix[4]arenes. On the other hand, enantioselective desymmetrization [27–29] has proven to be the most effective method for asymmetric synthesis of ICCs [30]. Notable advancements have been made in recent years, particularly through the directing group-enabled asymmetric meta-functionalization of the upper-rim, including approaches of C-H cyclization desymmetrization and C-H substitution desymmetrization (Fig. 1C). In 2022, both the Cai group [31] and the Tong group [32] independently reported the enantioselective synthesis of ICCs through a Pd-catalyzed intramolecular C-H arylation-enabled desymmetrization, albeit producing different cyclization products. Later, our group [33] and the Liu group [34] respectively presented the asymmetric synthesis of ICCs by desymmetrizing para-amino-substituted prochiral calix[4]arenes using sequential asymmetric Povarov reaction and aromatization. Recently, the Shi group [35] and the Niu group [36] detailed the asymmetric synthesis of ICCs using cobalt-catalyzed C-H annulation with alkynes in the presence of a directing group, leading to products exhibiting both inherent chirality and axial chirality. On the other hand, the enantioselective synthesis of ICCs through C-H substitution desymmetrization has received relatively less attention, as exemplified by recent demonstrations involving asymmetric electrophilic sulfenylation [37] and amination [38] of electron-rich arenes, along with directing group-enabled C-H acyloxylation [39]. Although numerous enantioselective desymmetrization methods have been developed to achieve high enantioselectivity, approaches that can facilitate the generation of ICCs with multiple modifications of phenyl rings and easy subsequent derivatization remain relatively limited.

Aryl halides play a crucial role in modern organic synthesis, which have been widely employed for synthesizing various functional molecules through cross-coupling reactions [40–42]. Given that electrophilic aromatic halogenation reaction represents the most classic and convenient way to produce aryl halides, the development of a catalytic enantioselective version holds significant importance, as the resulting chiral aryl halides would have broad applications in synthesizing chiral molecules through various cross-coupling reactions. Pioneered by Miller and co-workers, peptide-catalyzed enantioselective aromatic bromination of phenols has been extensively explored in the synthesis of axially chiral molecules through a dynamic kinetic resolution strategy, including biaryl [43–45], benzamides [46] and C-N atropoisomers [47–48]. Moreover, asymmetric aromatic bromination has been widely utilized for desymmetrizing prochiral molecules, resulting in the formation of carbon/heteroatom central chirality [49–52] and C-C/C-N axial chirality (Fig. 1D) [53–55]. However, while aryl bromide-containing ICCs hold promise as valuable precursors for straightforward derivatizations, the asymmetric synthesis of ICCs through aromatic C-H bromination reactions has yet to be revealed. In this work, we present the application of a chiral phosphoric acid (CPA) [56–62] catalyzed aromatic bromination reaction for the enantioselective synthesis of valuable aryl bromide-containing ICCs by desymmetrizing phenol-containing prochiral calix[4]arenes (Fig. 1E). Moreover, asymmetric di-bromination and even multi-bromination were also viable, which afford ICCs with three or even four phenyl rings precisely modified, a structure rarely attained by other known catalytic asymmetric methods.

We commenced our study by choosing the 4-phenol-containing calix[4]arene 1a as the model substrate, which can be readily prepared from unsubstituted calix[4]arene through sequential bromination, boronation and oxidation steps (Table 1). Utilization N-bromosuccinimide (NBS, 2a) as the electrophilic bromination reagent, and after extensive optimizations of reaction conditions, we were delighted to find that the reaction between 1a and NBS (1.05 equiv.) in the presence of CPA A1 (10 mol%) yielded the inherently chiral C-H brominated calix[4]arene 3a in 93% yield with 93% ee in DCM at −78 ℃ (entry 1). Our screening of the CPA catalysts suggested that both the 3,3'-substitutions and the chiral scaffolds of the CPA played a critical role in achieving high enantioselectivity in this reaction (entries 2-8). Notably, CPAs featuring 3,3'-bis-(2,4,6-(iPr)₃C₆H₂) groups exhibited notably superior enantioselectivity, with H8-BINOL-derived CPAs also surpassing the performance of the BINOL-derived CPAs. Examination of the solvents revealed that this reaction could proceed in various solvents; however, none of them could yield superior or comparable enantioselectivity (entries 9-12). Increasing the reaction temperature from −78 ℃ to −50 ℃ resulted in the formation of 3a with a decreased 85% ee (entry 13). Increasing the quantity of 2a from 1.05 equiv. to 1.2 equiv. resulted in slightly improved ee value but reduced yield due to the increased formation of the di-bromination product (entry 14).

With the optimal reaction conditions established, we set out to explore the scope of this method for the enantioselective synthesis of ICCs (Scheme 1). A series of C-ring modified prochiral phenol-containing calix[4]arenes were studied using this method. We were pleased to find various substitutions (3b-3d), including the functionalizable groups such as halides (3c) and boronic ester group (3d), were well tolerated by the method. Moreover, other heteroatom substitution on the C-ring were examined, such as the acyloxy (3e), sulfonyloxy (3f and 3g), alkoxy (3h) and amido (3i) groups, all of which exclusively yielded the desired phenol ring bromination products with high enantioselectivities, even though these electrondonating groups may potentially enhance the reactivity of the C-ring. Furthermore, a series of substituted phenyl rings at the para-position of the C-ring demonstrated compatibility with this enantioselective desymmetrization method, regardless of their substitution positions and electronic nature (3j-3n), including the sterically demanding ortho-substituted ones (3l-3m). The absolute configuration of these ICCs was assigned as (P) by analogy to the configuration of products 3i and 3j, whose structures were unambiguously determined by X-ray crystallography analysis. In addition, introducing 9-anthracenyl (3o), heteroaryl (3p) and acrylate (3q) substitution on the C-ring proved successful with this approach, yielding ICCs with good yield and high enantioselectivity. Modifications of the alkoxyl groups at the lower rim of calix[4]arenes were also studied, which indicated that various alkoxyl groups were feasible with the current method (3r-3s).

After successfully demonstrating the excellent compatibility for catalytic enantioselective synthesis of ICCs through asymmetric mono-bromination, our focus shifted to asymmetric synthesis of ICCs through asymmetric di-bromination and multi-bromination (Scheme 2). Encouragingly, the reaction between the A-ring and C-ring bis-phenol-containing calix[4]arene 4a with NBS (2.1 equiv.) enabled by CPA A1 (20 mol%) yielded the di-brominated ICC 5a in 74% yield with 92% ee, featuring C-2 symmetry. Moreover, the prochiral bis-phenol-containing calix[4]arenes with B-ring and D-ring identical substitutions were explored with this method, resulting in the formation of ICCs where all four phenyl rings were modified (5b-5f). Furthermore, the calix[4]arenes with a single B-ring substitution (5g-5h) or B-ring and D-ring featuring different substitutions (5i-5j) proved to be viable substrates under these conditions, yielding the chiral calix[4]arenes with C-1 symmetry in good yield and high enantioselectivity. Notably, the tetra-phenol-containing calix[4]arene 4k was also investigated under the standard condition. However, due to its poor solubility, the reaction at -78 ℃ could not yield a major product. Interestingly, conducting this reaction at room temperature led to the isolation of a hexa-brominated ICC 5k in 39% yield with 90% ee.

To provide further insight into this method, a series of control experiments were performed. The reaction between 1a and NBS in the absence of CPA catalyst yield the brominated 3a in 85% yield at room temperature, with a yield of 47% achieved even when the reaction was performed at -78 ℃ (Fig. 2A). These results suggest a strong background reaction without chiral catalyst, thus highlighting the challenges to achieve high enantioselectivity control. Employing an O-Me protected calix[4]arene 6a as the substrate under the standard conditions failed to produce the expected bromination product 7a, which indicated that the free OH group is essential for both reactivity and enantioselectivity in this reaction, potentially engaging in hydrogen bonding with the CPA catalyst (Fig. 2B). Additionally, the use of hydrogen-bond-competing solvents, such as THF and MeOH, resulted in the formation of products with low enantioselectivities (see results in Table 1), further underscoring the critical role of hydrogen bonding interactions. Moreover, other electrophilic halogenation reagents were also explored using this method, and we were delighted to find that N-bromophthalimide could also yield ICC 3a with 93% ee (Fig. 2C). However, the use of NIS or NCS proved unsuccessful, resulting in either the formation of the product in nearly racemic form (3a') or a failure to yield the product. These results can be attributed to their inappropriate reactivities under these conditions: NCS exhibits insufficient reactivity, while NIS displays overly high reactivity, making it challenging to control stereoselectivity with the CPA catalyst. Given the observation of di-brominated byproducts for certain substrates, the potential involvement of kinetic resolution in the second bromination step was investigated (Fig. 2D). Treatment of racemic 3a with NBS (0.5 equiv.) under the standard conditions resulted in the achiral di-bromination calix[4]arene 8a and the recovered 3a in 47% yield with 68% ee (s = 8.0), with the same configuration as achieved through the desymmetrization method. This result indicated that the high enantiopurity of the obtained ICCs might arise from the synergistic effect of the enantioselective desymmetrization (the first bromination step) and kinetic resolution (the second di-bromination step) (Fig. 2E). In this process, the minor enantiomer (ent-3) formed during the initial desymmetrization step may be consumed more rapidly in the subsequent kinetic resolution step, ultimately enhancing the enantioselectivity of the products (3). Based on the results of the control experiments and previous studies [55], a plausible mechanism and the origin of enantioselectivity for inherent chirality were proposed (Fig. 2F). Through the dual-hydrogen bonding activation facilitated by the CPA catalyst on the phenol moiety of 1a and NBS, the ortho-selective electrophilic addition would lead to the formation of the dearomatized intermediate INT-A, which serves as both the rate-determining and enantio-determining step. Following that, the facile tautomerization step led to rearomatization, yielding the desired ICC 3a. Given the intrinsic cone-shaped structure of calix[4]arene, we presume that NBS would approach the phenol ring exclusively from the outer face. Accordingly, the addition of ortho-position a (route a) would lead to the formation of intermediate (S,P)-INT-A, whereas the addition using ortho-position b (route b) would yield intermediate (R,M)-INT-A. As demonstrated in the previous DFT calculation studies by Akiyama and co-workers [55], the (R)-CPA favors the formation of a brominative dearomatized intermediate with (S)-configuration. Consequently, the formation of the (S,P)-INT-A would be more favorable as (R)-CPA A1 was used in our reaction, which effectively explains the configuration of inherent chirality observed in these reactions.

To showcase the utilities of this method, a variety of derivatizations of the ICC products were performed (Scheme 3). Electrophilic aromatic substitution of 3a with diazo reagent 9 afforded the ortho-amination product 10 in 57% yield (Scheme 3A). Interestingly, treatment of 3a with NIS did not yield the expected iodinated product; instead, the oxidative quinone-containing calix[4]arene 11 was obtained. Moreover, Suzuki coupling of 3a with phenylboronic acid in the presence of a palladium catalyst led to the cross-coupling product 12. Notably, treatment of 3a with a Pd catalyst without adding other coupling component resulted in the formation of the transannular arene–arene coupling product 13. Moreover, the trifluoromethanesulfonylation of 3a produced 14, featuring two cross-coupling handles for further derivatizations (Scheme 3B). Treatment of 14 with n-BuLi generated the active benzyne intermediate, which underwent a Diels-Alder reaction with isobenzofuran derivatives 15 to give the corresponding cycloaddition product 16 [63]. Subsequent deoxygenative aromatization [64] of 16 led to the formation of the anthracene-containing ICCs 17 in 86% yield with 99% ee. The derivatizations of the 1,3-bis-phenyl ring modified ICC products were also investigated (Scheme 3C). Starting with the C-ring benzoyloxy-substituted chiral calix[4]arene 3f, the removal of the O-Bz group under basic conditions afforded diphenol 18, which then underwent the bis-alkylative macrocyclization to give the upper rim-bridged inherently chiral calix[4]arene 19, potentially creating a rigidified chiral cavity [65].

While the brominative ICC products did not exhibit notable photophysical properties, we were thrilled to find that the anthracene-containing ICC derivative 17 demonstrates potent photophysical and chiroptical properties, which have been extensively studied (Fig. 3). The UV-vis absorption spectrum of 17 revealed a maximum absorption band at 273 nm (Fig. 3A). The fluorescence spectrum of ICC 17 displayed its strongest emission band at 465 nm, with the compound exhibiting a light-blue fluorescence in solution upon excitation (Fig. 3B). To investigate the chiroptical properties of this compound, both enantiomers of 17 were synthesized through the same route from the pair of enantiomers of 3a, which were derived by employing the CPA A1 catalyst with opposite configurations. The circular dichroism (CD) spectrum of (P)/(M)-17 displayed a clear mirror image relationship, exhibiting remarkable Cotton effects (Fig. 3C). Moreover, the circularly polarized luminescence (CPL) analysis of (P)/(M)-17 also revealed a mirror image spectrum, indicating its CPL activity (Fig. 3D). The luminescence dissymmetry factor |g_lum| value of this compound was measured to be 1.72×10⁻³ (at 465 nm), suggesting its promising applications in chiroptical materials.

In conclusion, we have developed an efficient method for asymmetric synthesis of inherently chiral calix[4]arenes through an organocatalyzed enantioselective desymmetrization protocol. Leveraging the chiral phosphoric acid-catalyzed asymmetric ortho-bromination of phenols, the symmetry of a series of prochiral calix[4]arenes were successfully broken, which led to the formation of various 1,3-bis-phenyl ring-modified ICCs with high enantioselectivities. Moreover, this method is applicable for bis-phenol-containing prochiral calix[4]arene substrates, resulting in ICCs with three or even four phenyl rings precisely decorated. The versatile reactivity of the introduced bromide handle and the intrinsic phenolic OH group in products allowed for a variety of derivatizations, producing a diverse range of ICC derivatives. Notably, one ICC derivative containing anthracene moiety exhibited promising photophysical and chiroptical properties, highlighting the potential of this method.