Han-Qi
Zhou‡
a,
Xing-Wei
Gu‡
a,
Xiao-Hua
Zhou
a,
Li
Li
a,
Fei
Ye
a,
Guan-Wu
Yin
a,
Zheng
Xu
a and
Li-Wen
Xu
*ab
aKey Laboratory of Organosilicon Chemistry and Material Technology of Ministry of Education, Key Laboratory of Organosilicon Material Technology of Zhejiang Province, Hangzhou Normal University, No. 2318, Yuhangtang Road, Hangzhou 311121, P. R. China. E-mail: liwenxu@hznu.edu.cn
bState Key Laboratory for Oxo Synthesis and Selective Oxidation, Suzhou Research Institute and Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences, P. R. China
First published on 20th September 2021
Catalytic asymmetric variants for functional group transformations based on carbon–carbon bond activation still remain elusive. Herein we present an unprecedented palladium-catalyzed (3 + 2) spiro-annulation merging C(sp2)–C(sp2) σ bond activation and click desymmetrization to form synthetically versatile and value-added oxaspiro products. The operationally straightforward and enantioselective palladium-catalyzed atom-economic annulation process exploits a TADDOL-derived bulky P-ligand bearing a large cavity to control enantioselective spiro-annulation that converts cyclopropenones and cyclic 1,3-diketones into chiral oxaspiro cyclopentenone–lactone scaffolds with good diastereo- and enantio-selectivity. The click-like reaction is a successful methodology with a facile construction of two vicinal carbon quaternary stereocenters and can be used to deliver additional stereocenters during late-state functionalization for the synthesis of highly functionalized or more complex molecules.
Most of the related studies of enantioselective transformations based on the carbon–carbon bond cleavage of cyclopropenones are unsuccessful because of their low enantioselectivity (Fig. 2B), such as the organocatalytic synthesis of allenic esters (up to 83% ee),60,61 spirooxindoles from formal (3 + 2) cycloaddition with isatins (30–74% ee),62o-hydroxy aromatic aldimine-derived heterocyclic compounds,63 and other explorations with low enantioselectivity.64,65 Furthermore, the synthesis of chiral oxaspiranic compounds from cyclopropenones is especially challenging as previously reported synthetic procedures toward these species of spirocycles are based on Lewis acid66 and gold catalysis67 without suitable chiral ligands. Thus, much work is needed to develop highly stereoselective strategies with more flexible synthetic features, to expand the structural and functional pattern of heterocyclic compounds bearing at least a carbon-stereogenic center. Very recently, Li and co-workers reported a notable example that chiral nickel catalysis could complete an enantioselective (3 + 2) annulation of cyclopropenones and α,β-unsaturated ketones/imines with good ees.68 This work demonstrated for the only example that enantioselective transformations of cyclopropenones can also be achieved by a chiral ligand-controlled transition metal-catalyzed carbon–carbon bond activation process. Nevertheless, there is no precedent for the palladium-catalyzed enantioselective variant of the carbon–carbon bond functionalization of cyclopropenones involving a sequential C(sp2)–C(sp2) bond activation/(3 + 2) annulation process.
Considered the powerful potential of C–C bond activation and chiral ligands for the enantioselective construction of a single quaternary stereocenter,69 we propose a new strategy for the construction of more quaternary stereocenters that can use cyclopropenones as the active species to achieve the desymmetrization of bifunctional compounds (Fig. 2C). In this regard, one of the potential problems to control the diastereo- and enantio-selectivity lies in the precise recognition function of active intermediates of metal catalysts that are formed from the oxidative addition of the carbon–carbon bond of cyclopropenones to metal species, during the desymmetrization of the bifunctional groups before they undergo (3 + 2) annulation.
To address current limitations on the enantioselective and rarely reported transition-metal-catalyzed (3 + 2) spiro-annulation of cyclopropenones with ketones, we expect to design rigid and cyclic diketone substrates to realize a new strategy of constructing two vicinal carbon quaternary stereocenters based on C(sp2)–C(sp2) bond activation and desymmetrization that is similar to atom-economic (3 + 2) click additions.70 However, the utilization and re-organization of two small rings in transition-metal-catalyzed (3 + 2) spiro-annulation to create oxaspiro cyclopentenone–lactone scaffolds remain elusive.
Herein, we report a new strategy for the ring expansion and click cycloaddition of two small rings to access highly enantioselective transition-metal-catalyzed (3 + 2) spiro-annulation of cyclopropenones with cyclopentene-1,3-diones, which is achieved for the facile construction of complex spiro-molecules with two adjacent stereocenters. The reaction via palladium-promoted tandem carbon–carbon bond cleavage and (3 + 2) annulation featured with producing two quaternary carbon stereocenters by desymmetrization.
Entry | Variation from “standard conditions” | Yieldb (%) | ee (%) | dr |
---|---|---|---|---|
a Unless otherwise noted, the standard reaction conditions were as follows: 1a (0.2 mmol), 2a (0.4 mmol), and in solvent (0.4 mL). For entries 2–9, the solvent is toluene. b The yield and dr value were determined by crude 1H NMR analysis using dibromomethane as an internal standard. c The enantiomeric excess of 3a was determined using chiral UPLC. d Under the reported reaction conditions with Ni catalysis,15 no product of the annulation reaction (nr) was detected in this case. | ||||
1 | None | 87 | 92 | >19:1 |
2 | With L1 instead of L15 | <5 | — | — |
3 | With L2 instead of L15 | 30 | 22 | >19:1 |
4 | With L3 instead of L15 | <5 | — | — |
5 | With L4 instead of L15 | 16 | 57 | 12:1 |
6 | With Pd2(dba)3 as Pd source | 61 | 90 | >19:1 |
7 | With Pd(dba)2 as Pd source | 59 | 89 | >19:1 |
8 | With Pd(MeCN)2Cl2 as the Pd source | nr | — | — |
9 | With PdBr2 as the Pd source | nr | — | — |
10 | With THF as solvent | 32 | 92 | 13:1 |
11 | With DMF as solvent | 69 | 94 | >19:1 |
12 | With NMP as solvent | 67 | 67 | >19:1 |
13 | With DMAc as solvent | 74 | 92 | >19:1 |
14 | With toluene at 0 °C | 44 | 93 | >19:1 |
15 | With toluene at 40 °C | 21 | 78 | 16:1 |
16 | With toluene at 80 °C | 34 | 70 | 16:1 |
17 | With Ni(cod)2 under Li's conditionsd,15 | nr | — | — |
Having optimized the reaction conditions, we then explored the substrate scope of the catalytic asymmetric (3 + 2) spiro-annulation reaction. Notably, there is no other reaction process that could be applied for the enantioselective construction of such oxaspiro cyclopentenone–lactone scaffolds. As such, catalytic access to structurally diverse products bearing different substituents on the starting materials, cyclopentene-1,3-diones or cyclopropenones, is valuable. The results in Fig. 3 show that highly efficient (3 + 2) spiro-annulation takes place for cyclopentene-1,3-diones, where the aromatic rings of benzyl groups are substituted with methyl or halide groups at the para, ortho, or meta position, with the corresponding oxaspiro molecules 3 being produced with high enantio- and diastereo-selectivity (90–94% ee, up to >19:1 dr). The yields were also high for the aryl cyclopentene-1,3-diones bearing ortho-substituted groups, such as 3b, 3e, 3i and 3n with 81–99%. Notably, cyclopentene-1,3-dione substrate 1p containing a transition-metal-sensitive nitro group was really suitable for this reaction because of its good enantioselectivity (92% ee) and moderate yield (56%), where it is generally believed that the nitro group is not conducive to the stereocontrol of transition metal complexes and likely problematic in catalytic cycles. It is also possible to use thiophene-substituted substrate 1r to finish the (3 + 2) spiro-annulation with excellent diastereo- and enantio-selectivity (92% ee and >19:1 dr). This provides access to a valuable S– heterocyclic oxaspiro molecule. In addition, alkynyl and other substrates bearing electron-withdrawing groups, such as CO2Me, CN, and CF3, were found to undergo efficient and highly diastereo- and enantio-selective cycloaddition with cyclopropenones (Fig. 3). These products bearing such functional groups provide additional access to the synthesis of more complex molecules. Interestingly, when the methyl group on the cyclopentenone is replaced with an ethyl group, the steric hindrance effect exhibits negative function, and accordingly the yields of 3v–3x are very low under the same optimized reaction conditions. Therefore, the reaction temperature needs to be increased to 50 °C to obtain an improved yield, and correspondingly, its stereoselectivity is also significantly reduced to a moderate level. As expected, substituted cyclopropenones gave the desired oxaspiro products 3y–3aa with good diastereo- and enantio-selectivity in moderate to good yields. Encouraged by these reaction results, we next sought to extend the catalytic asymmetric (3 + 2) spiro-annulation to unusual cyclopentene-1,3-diones or cyclopropenones. Using spiro-type cyclopentene-1,3-dione 1y as an acceptor allowed access to the enantioselective construction of double spiro cyclopentenone–lactone scaffold bearing two vicinal carbon quaternary stereocenters with 83% ee and 94% yield (eqn (1) of Fig. 4), whereas moderate diastereoselective control presumably occurred due to the unfavorable steric repulsion between rigid spiro substrate 1y and nucleophilic palladium-2a species.
Fig. 4 Exploration of the catalytic asymmetric (3 + 2) spiro-annulation with other unusual substrates. |
Then, unsymmetrical cyclopropenone 2f bearing phenyl and i-propyl groups was evaluated to access the desired product 3ac (eqn (2) of Fig. 4), demonstrating the viability of (3 + 2) spiro-annulation with controllable chemo-, diastereo-, and enantioselectivity by palladium catalysis. In addition, an alkyl oxaspiro cyclopentenone–lactone scaffold was also accessed by using propyl cyclopropenone 2g and cyclopentene-1,3-dione 1a as coupling partners, and the desired product 3ad could be achieved with excellent diastereoselectivity and moderate yield and enantioselectivity due to the lack of aromatic interaction in comparison to that of phenyl cyclopropenone. To further demonstrate the substrate scope and applicability of the palladium-catalyzed (3 + 2) annulation reaction, we next investigated the use of 5-(tert-butyldimethylsilyloxy)naphthalene-1,4-dione 4 as an acceptor in this cycloaddition. Although the structure of 4 is largely different from that of cyclopentene-1,3-diones, the desired product 5 was obtained with excellent enantioselectivity (92% ee) and promising regioselectivity (4:1 rr) as well as good yield (86%), and we reasoned that this reaction would provide useful information on the reaction mechanism of palladium-catalyzed (3 + 2) spiro-annulation.
On the basis of experimental results and 31P NMR analysis (see Fig. S1–S3 of the ESI†), a plausible catalytic cycle for the palladium-catalyzed (3 + 2) spiro-annulation of cyclopropenones with cyclic 1,3-diketones is depicted in Fig. 5. First, the Pd0–L complex is coordinated with the three-membered cyclopropenone to obtain the complex A, and the intermediate B is obtained by oxidative addition with carbon–carbon bond activation. And then the coordination of Pd species B with the carbon–carbon double bond of cyclopentene-1,3-dione is affected by the steric repulsion of the substituted group, giving a favorable coordinated model as complex C that the Pd center located on the methyl side because of less steric hindrance, and subsequent migratory insertion of a Pd–C bond into the same-directional ketone group to give intermediate E. In this regard, the control experiment with 2-benzyl-2- methylcyclopentane-1,3-dione or other enones as a substrate instead of 1a in this reaction resulted in no reaction (Fig. S2†), which revealed the importance of both cyclic 1,3-diketone and carbon–carbon double bonds of 1a in the catalyst–substrate interaction. And finally, the product 3a is obtained by reductive elimination and the release of Pd0–L continues to participate in the next catalytic cycle. And the experimental results determined that the recognition of one of the carbonyl groups is a key step in this desymmetrization and (3 + 2) annulation reaction, indicating the crucial role of the TADDOL-derived bulky P-ligand bearing a large cavity in the alkene-directed migratory insertion of the Pd–C intermediate to the carbon–carbon double bond. The formation of Pd/substrate species D is necessary for the high level of enantioselective induction during the stereospecific migratory insertion and subsequent formation of an oxaspiro skeleton.
To evaluate whether this (3 + 2) spiro-annulation and its natural product-like complex molecules could be applied to late-stage functionalization, we performed a gram-scale experiment to provide the starting material for the downstream transformations (Fig. 6). Next, we focused on our attention on the further functionalization of the cyclopentenone moiety. The strong base (NaOMe)-promoted transesterification reaction of 3a with methanol resulted in sequential oxa-Michael addition of alcohol to the intramolecular cyclopentenone moiety, which gave unexpected product 6 bearing an epoxide moiety with good enantioselectivity. And the treatment of 3a with t-BuOOH in the presence of DBU gave the corresponding product 7 with almost quantitative yield and good enantioselectivity (99% yield and 90% ee, Fig. 6). In addition, an enantioselective conjugate addition reaction of azide was also performed because of its importance in synthetic chemistry. The desired product 8, having an additional azide moiety on the oxaspiro cyclopentenone–lactone scaffold, was obtained with good enantioselectivity. These representative transformations supported the powerful potential of the oxaspiro cyclopentenone–lactone scaffold in synthetic chemistry.
Footnotes |
† Electronic supplementary information available. CCDC 2024499. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/d1sc04558j |
‡ These authors contributed equally. |
This journal is © The Royal Society of Chemistry 2021 |