Although benzene rings, which consist entirely of sp
2 carbon atoms, have traditionally been important in small-molecule drug discovery, recent findings have shown that a high fraction of sp
3 carbon atoms (Fsp
3, which is the ratio of the number of sp
3 carbons to the total number of carbons) is associated with the success of drug discovery [
1-
3]. The optimal Fsp
3 value is considered to be ≥0.42, and ap-proximately 84% of marketed drugs meet this criterion [
4-
7]. Moreover, data analysis has revealed that compounds with fewer than three aromatic rings are more likely to be developed as drug molecules. As a result, increasing saturation (referred to as “escape from flatland”) has gained attention in the medicinal chemistry community to meet the goal of improving clinical success in drug development by replacing flat benzenoid moieties with sp
3-carbon-rich three-dimensional bioisosteres [
2,
3,
8-
15]. Bicyclo[2.1.1]hexanes (BCHs) are among the most important bioisosteres of benzene rings because they not only have a high Fsp
3 but also mimic the relationships between ortho-substituents on the benzene rings (
Fig. 1a) [
16-
23]. Recently, catalytic synthesis of BCHs has been intensively studied. For example, in 2020 and 2023, Mykhailiuk
et al. reported pioneering work on photochemical intramolecular [2
π + 2
π] cycloadditions of 1,5-dienes to prepare multi-substutited BCHs [
24,
25]. Subsequently, methods for [2
σ + 2
π] cycloadditions of bicyclo[1.1.1]butanes (BCBs) to construct BCHs have emerged and have gradually come to dominate this field [
26-
40]. Specifically, the groups of Glorius, Bach, Brown and Lu developed a series of elegant, efficient methods for synthesizing BCH scaffolds by means of triplet-energy-transfer-enabled intermolecular [2
σ + 2
π] cycloadditions of BCBs and 2
π substrates, such as coumarins [
26], aza-renes [
27], cyclobutenones [
28], phenols [
29], styrenes [
30] and cyclic imines [
31]. Additionally, the groups of Procter, Li and Wang established methods for SmI
2-catalyzed [
32] and pyridine diboron–catalyzed [
33,
34] radical [2
σ + 2
π] cycloadditions. More recently, Glorius
et al. developed a new strategy for single-electron oxidative activation of BCBs
via photoredox catalysis to afford BCB radical cations, which can be captured with unactivated olefins to form BCHs [
35]. In addition, the groups of Studer, Deng, Feng, Chen and Zhou achieved Lewis acid–catalyzed ionic [2
σ + 2
π] cycloadditions of BCBs with ketenes [
36], indoles [
37,
38], 3-benzylideneindoline-2-thione derivatives [
39], silyl enol ethers [
40] and ynamides [
41].
BCHs can contain valuable chiral information that has the potential to be significant in drug design and development [
16,
42-
45]. However, the enantioselective synthesis of BCHs is still in its early stages (
Fig. 1b). In 2023, Bach reported the earliest success of asymmetric [2
σ + 2
π] cycloadditions, but stoichiometric chiral reagent was required to maintain high enantioselectivity [
46]. In 2024, the Jiang group reported the catalytic asymmetric radical [2
σ + 2
π] cycloadditions of BCBs with vinylazaarenes to obtain heteroaryl-substituted BCHs by utilizing a chiral photo-sensitive Brønsted acid catalyst [
47]. Lately, Yoon achieved a photochemical chromophore activation enabled enantioselective [2
σ + 2
π] cycloadditions of BCBs with
α,
β-unsaturated ketones [
48]. All these reports are based on photochemical radical mechanism and developing alternative pathways to prepare chiral BCHs with other substitution patterns are still high desirable.
Nitriles play a significant role in medicinal chemistry, with over 60 small molecule drugs currently on the market that contain the cyano functional group (
Fig. 1c) [
49]. The introduction of the cyano group can enhance a drug candidate’s pharmacodynamic (PD) and pharmacokinetic (PK) profiles, as well as improve water solubility [
50]. Additionally, it may facilitate beneficial noncovalent interactions [
50-
52] or covalent binding [
53] with the drug target. Giving the importance of cyano-arenes moiety in drug molecules, it would be attractive to develop cyano-BCHs as the potential bioisosteres for cyano-arenes. However, currently there are no methods available to access this structural interesting motif. Inspired by the fruitful donor-acceptor cyclopropane chemistry [
54-
64], we recently developed the precursor vinyl-carbonyl-bicyclo[1.1.0]butanes (VC-BCBs) that are capable of undergoing palladium(0)-catalyzed (3 + 2) cycloadditions with aldehydes (
Fig. 1d) [
65]. As part of our exploration of the potential utility of palladium-catalyzed cycloadditions of VC-BCBs, especially in terms of enantioselective variants, we herein report a palladium-catalyzed [2
σ + 2
π] cycloadditions of VC-BCBs with arylidenemalononitriles, providing a platform to synthesize enantioenriched cyano-BCHs (
Fig. 1e).
We initiated our study by using VC-BCB
1a and arylidenemalononitrile
2a as model substrates to screen various chiral ligands in reactions in THF at 30 ℃ under argon for 12 h (
Table 1). Initially, we found that electron-rich phosphorus ligands showed poor enantioselectivity for the desired cycloaddition (for details, see Supporting information). Therefore, we shifted our focus to electron-deficient phosphoramidite ligands. Ligands
L1 and
L2 showed good activity, yielding desired product
3a in >90% yields, but the enantioselectivity was poor (entries 1 and 2). By switching to ligand
L3, we achieved moderate enantioselectivity (43%
ee), but the yield of
3a dropped dramatically (entry 3). In previous work, we [
66] and others [
67] had found that a bidentate bisphosphite ligand can accelerate the allylic substitution step. Therefore, we explored some typical BINOL-derived bisphosphite ligands (
L4–
L6;
Table 1, entries 4–6). We were pleased to find that using
L5 allowed us to obtain
3a in 73% yield with 61%
ee (entry 5). After carefully exploring various linkages for the bisphosphite, we found that
L7, a new ligand featuring a SPINOL linkage, greatly enhanced the enantioselectivity (entry 7). Furthermore, when we replaced the BINOL backbone of
L7 with H
8BINOL, resulting in ligand
L8, 3a was obtained in 62% yield with 80%
ee (entry 8). Subsequently, by varying the reaction temperature and solvent, entries 9–14), we established the optimal conditions. Specifically, by using
L8 as the ligand, methyl
tert-butyl ether as the solvent, and a reaction temperature of −40 ℃, we obtained
3a in 75% yield with 92%
ee. Moreover, we discovered that we could reduce the loadings of PdCp(
η3-C
3H
5) and
L8 to 2.5 mol% and 3 mol%, respectively, without affecting the yield of
3a (entry 15). However, extending the reaction time to 48 h was needed to achieve completion of the reaction. In the absence of
L8 or PdCp(
η3-C
3H
5), this reaction could not occur (entry 16).
Having established the optimal conditions, we next turned our attention to the substrate scope, beginning with the arylidenemalononitrile (
Scheme 1). The catalytic system proved to efficiently mediate reactions of malononitrile derivatives with either an electron-withdrawing or electron-donating substituent on the benzene ring, smoothly delivering chiral BCHs
3a–
3j in 63%–93% yields with 87%–96%
ee. Notably, substrates with a sterically hindered
ortho-methyl or 3,5-dimethyl-substituted benzene ring or a naphthyl motif also gave the target products (
3k–
3n) with satisfactory yields and enantioselectivities. In addition, substrates with a thiophene, benzothiophene, benzofuran, quinoline, or pyridine motif were tolerated, furnishing
3o–3u in 71%–95% yields with 77%–95%
ee. Alkylidenemalononitrile could also be used as a reaction partner in this cycloaddition; however, it produced poor enantioselectivity, as shown by the formation of
3v with 15%
ee. This unsatisfying outcome could be caused by the smaller steric hindrance of
2v. The structure and stereochemistry of the products were confirmed by single-crystal X-ray diffraction analysis of
3h.
With
2n as the reaction partner, we next explored the substrate scope with respect to the VC-BCB (
Scheme 2). Specifically, the carbonyl moiety of the VC-BCB was systematically varied. We found that benzoyl groups with various substituents, such as an electron-donating alkyl or methoxy group or electron-withdrawing halogen atoms or a trifluoromethyl group, were well tolerated, affording
3ab–3ak in 60%–95% yields with 89%–95%
ee. Moreover, naphthyl ketones (
3al, 3am) and heteroaryl ketones (
3an–3as) were suitable substrates, demonstrating the generality and adaptability of this method. An
n-butyl ketone–substituted substrate was compatible with the catalytic system as well, furnishing corresponding product
3at in 90% yield with 88%
ee. In addition to ketones, acyl pyrazole (
3au) proved to be competent substrates for this catalytic asymmetric cycloaddition. For carboxylate-derived VC-BCBs, the ligand
L8 yielded relatively low
ee values. Specifically, the phenyl and
tert-butyl carboxylate products
3av and
3aw were obtained with enantioselectivities of 69%
ee and 60%
ee, respectively, under the standard conditions. However, following further optimization, ligand
L9 was identified as a highly effective ligand for these challenging substrates, improving the enantioselectivities of
3av and
3aw to 81%
ee and 77%
ee, respectively. Additionally, the VC-BCB containing a (l)-menthol-moeity produced
3ay in 78% yield with >20:1
dr by using
L9 as the ligand. Regardless of whether
L8 or
L9 was used, no cycloaddition products were observed for VC-BCBs bearing an internal alkene (
3ax).
To illustrate the synthetic utility of our method, we performed a gram-scale synthesis of chiral BCH
3b and found that the yield and
ee value were well-maintained upon scale up, making this method practical for large-scale production (
Scheme 3a). Subsequently, transformations of
3b were carried out. Aiming to overcome the limitation posed by the geminal cyano groups of the substrates, we set out to transform the cyano groups into other functionality (
Scheme 3b). One cyano group could be selectively hydrolyzed by using copper(Ⅱ) acetate as a catalyst for the generation of amide
4, which was obtained in 80% yield with the formation of a new quaternary stereocenter with a diastereomer ratio (
dr) of 8:1 [
68]. Decyanative coupling of
3b with chlorosilanes enabled by nickel/titanium dual catalysis [
69] allowed us to construct a tertiary C(sp
3)–Si bond to afford
5 in 19% yield with 90%
ee and >20:1
dr, together with decyanative protonation product
cis-6 (21% yield, >20:1
dr, 90%
ee). The structure and stereochemistry of
cis-
6 were confirmed by X-ray crystallography (CCDC 2380580). Interestingly, treatment of
5 with tetra-
n-butylammonium fluoride at −20 ℃ selectively afforded
trans-
6 with >20:1
dr. That is, stereodivergent synthesis of cyano-BCHs bearing vicinal stereocenters was achievable. Next, the vinyl group of
3b was converted to various other functional groups (
Scheme 3c). Specifically, oxidative cleavage of the double bond by RuCl
3/NaIO
4 and subsequent esterification furnished ester
7 in 90% yield. An iridium-catalyzed hydroboration/oxidation procedure transformed
3b to alcohol
8 in 58% yield. Additionally, the pyrazole moiety of
3au could be hydrolyzed with aqueous LiOH, then underwent esterification reaction to afford a 66% yield of redox-active ester
9 (NHPI =
N-hydroxyphthalimide) which could be converted to heteroaryl-substituted BCH
10 by means of a decarboxylative Minisci reaction (
Scheme 3d) [
70].
To get insight into the reaction mechanism, we carried out some control experiments (
Scheme 4a). First, the reaction of
1a and
2n was performed in the dark, which had little to no effect on the yield or enantiopurity of product
3n, a result that rules out a pathway involving catalysis by an excited-state palladium species. Moreover, addition of the radical scavenger TEMPO (2,2,6,6-tetramethylpiperidine 1-oxyl, 1.0 equiv.) did not notably affect the yield, suggesting that a radical pathway may not be involved. Furthermore, when methyl-substituted BCB ketone
1aa or des-methyl BCB ketone
1ab was subjected to reaction with
2n under the standard conditions, formation of the corresponding cycloaddition product was not observed. This result indicates that the vinyl group of the VC-BCB is crucial for the success of the cycloaddition with the arylidenemalononitrile. On the basis of these results, along with results of previous studies [
65,
71,
72] we propose that the reaction proceeds
via the mechanism shown in
Scheme 4b. Initially, a chiral palladium(0) complex coordinates with the vinyl group of
1a to form complex
A. The strong
π-donating ability of the palladium helps to cleave the strained C–C bond of the BCB, resulting in the formation of palladium-zwitterionic enolate
B. Subsequently,
B undergoes a stereoselective Michael addition to
2n to produce species
C, which then undergoes an intramolecular allyl substitution reaction to generate complex
D. Finally, disassociation of LPd from
D generates product
3n and releases the palladium(0) catalyst.
In summary, we have developed a palladium-catalyzed enantioselective [2
σ + 2
π] cycloadditions of VC-BCBs with arylidenemalononitriles. The method has a wide substrate scope and provides rapid access to cyano-BCHs with good enantioselectivity. Various transformations of the cyano-BCHs were accomplished to give diverse functionality for the products. Mechanistic experiments supported a pathway involving a zwitterionic intermediate generated by palladium-mediated ring opening of the VC-BCBs followed by stereoselective 1,2-addition and allylic substitution reactions. We believe this work not only demonstrates a platform for the preparation of chiral cyano-BCHs, but also provides a good complement to the rare asymmetric BCBs cycloaddition reactions [
46-
48,
73-
80].
PDF
Full
76
42
关闭全屏
No. 86 Xueyuan South Road, Haidian District, Beijing
010-62662699(Operation and Technical Support)
qkjq@cast.org.cn
