Asymmetric Gold(I)‐Catalyzed Tandem Hydroarylation–Nazarov Cyclization: Enantioselective Access to Cyclopentenones

Abstract

The asymmetric synthesis of cyclopentachromenones from gold‐catalyzed reaction of readily available skipped alkenynones is described. This cascade reaction involves an initial anti‐Michael hydroarylation of the ynone moiety to form a gold‐functionalized dialkenylketone intermediate, followed by a Nazarov cyclization that proceeds in an unprecedented enantioselective manner. Excellent enantiomeric ratios and chemical yields are obtained under mild reaction conditions.

A new enantioselective version of the Nazarov cyclization is reported. The reaction design uses readily available alkenynones to trigger a gold(I)‐catalyzed anti‐Michael hydroarylation of the ynone followed by Nazarov cyclization. A chiral gold complex is able to control the absolute stereochemistry of the process. Cyclopenta[c]chromenones, which combine the 2H‐chromene and cylopentanone cores, are synthesized with high yields and ee values.

Cyclopentenones are functional scaffolds for several transformations that allow access to bioactive target molecules. ^[1] Nazarov cyclization is among the most straightforward and commonly used methods for synthesizing this relevant core. ^[2] Nevertheless, its enantioselective version by means of asymmetric catalysis has demonstrated more challenges. ^[3] Notable advances have been achieved since 2003, when the first examples of the Nazarov cyclization under asymmetric catalysis were reported. ^[4] However, some of these pioneering contributions required sub‐ or stoichiometric amounts of the catalyst. In 2007, Rueping utilized chiral Brønsted acid organocatalysts to achieve the enantioselective synthesis of cyclopentenones (Scheme 1a). ^[5]

Scheme 1.

Previous work concerning asymmetric Nazarov cyclizations and our tandem gold‐catalyzed approach.

Since then, various approaches to asymmetric Nazarov reactions have been reported employing Lewis acid catalysts, ^[6] or chiral organocatalysts,[ ^3b , ⁷ , ⁸ ] or a combination of both metal Lewis acid and an organocatalyst. ^[9] These strategies are based on carbonyl activation by the chiral catalyst. Tius and Ateşin reported the only case in which chirality was achieved through an alternative approach through a Pd‐π‐allyl complex generated after coordination to one alkene moiety of a Pd⁰ complex bearing a phosphoramidite ligand, affording densely substituted cyclopentenones (Scheme 1a). ^[6d] Because of the spatial distance between the carbonyl group bonded to the chiral catalyst and the newly generated stereocenters, the design of the starting divinyl ketone plays a crucial role in achieving high enantioselectivities. For instance, polarized divinyl ketones ^[10] bearing an electron‐donating group at one of the α‐positions and an electron‐withdrawing group at the other α‐position were revealed as excellent substrates for asymmetric Nazarov cyclizations. Functionalities able to stabilize the cationic charge more efficiently over one of the α‐positions were also introduced, controlling the formation of the alkene and avoiding the formation of isomers. ^[9b] Notably, Rawal reported the enantioselective Nazarov cyclization of non‐polarized dienones using Cr‐salen complexes. ^[6c] In most cases, only divinyl ketones having substituents at both α‐positions were employed as suitable precursors for the asymmetric Nazarov cyclization. Only recently, List described the enantioselective synthesis of cyclopentenones using simple non‐activated α‐methyl‐substituted divinyl ketones and a confined Brønsted acid catalyst (Scheme 1b). ^[8]

The use of unconventional substrates in Nazarov cyclization has emerged considerably in recent years. ^[11] The development of gold catalysis has played a prominent role in designing Nazarov‐like reactions forming 5‐membered ring cycles. ^[12] Notably, Toste described the enantioselective synthesis of cyclopentenones from enantioenriched propargyl pivalates by Au^I‐catalyzed Rautenstrauch rearrangement. ^[13] Despite the challenges of asymmetric gold(I)‐catalysis, the same group designed a ligand‐controlled enantioselective dearomative Rautenstrauch rearrangement to obtain cyclopenta[b]indoles. ^[14]

Recently, we have described the synthesis of furanones from easily accessible skipped diynones ^[15] employing gold catalysis. ^[16] After initial Au‐catalyzed alkyne activation, the reaction proceeds by an unusual anti‐Michael hydration and subsequent endo‐cyclization (Scheme 1c). Inspired by these findings, we envisaged that dienones could be accessed by replacing one of the alkynes with an alkene group. An organogold dialkenyl ketone intermediate would be formed after alkyne activation followed by the addition of a suitable nucleophile. Considering that similar complexes have very low basicity compared with other C(sp²)−Au bonds, ^[17] we envisioned that carbonyl activation by the in situ generated proton could occur. If the dienone intermediate adopts s‐trans/s‐trans conformation faster than protodeauration of the alkenyl organogold bearing suitable chiral ligands, ^[18] an asymmetric Nazarov cyclization could be achieved, affording enantioenriched cyclopentenones (Scheme 1d).

To test our hypothesis, we designed a skipped alkenynone 1 a, ^[19] which would undergo an initial hydroarylation. ^[20] After some optimization, we found that XPhosAuCl, in combination with AgBF₄, afforded cyclopenta[c]chromenone 2 a in high yield (Scheme 2). ^[19] To account for its formation, we assumed that the initial hydroarylation takes place in an anti‐Michael manner ^[21] giving rise to an alkenyl gold intermediate like A, which could evolve via a Nazarov cyclization to the final cyclopentenone 2 a. At this stage, it is not clear if A is protodeaurated prior to the cyclization or not, which would be crucial for the success of an enantioselective version of this process by using chiral gold complexes as catalysts.

Scheme 2.

Proof of concept: tandem hydroarylation–Nazarov cyclization of skipped alkenynone 1 a.

Interestingly, this gold(I)‐catalyzed tandem transformation provides access to cyclopenta[c]chromenone derivatives in a straightforward manner from simple and readily available starting materials. These compounds combine the attractive 2H‐chromene core ^[22] with the cyclopentanone moiety. So, the scope of this process was tested with respect to the aryl moiety involved in the initial hydroarylation and the alkene substituents (Table S2). ^[19] Then, we decided to tackle if our initial proposal for developing a new strategy for enantioselective Nazarov cyclizations could be successful.

Using model alkenynone 1 a, we tested the reaction using various gold catalysts bearing chiral ligands. After an initial screening (see Table S3), ^[19] we selected three chiral BIPHEP‐like bisphosphines (Table 1). Our initial essays with L1–L3 in CH₂Cl₂ (rt, 3 h) using AgSbF₆ afforded cyclopentenone 2 a in high yields and good enantioselectivities, especially with L2 and L3 (entries 1–3). Curiously, the use of DCE as the solvent provides better ee values with L1 (entries 4–6). A subsequent counteranion screening with this ligand (entries 7–9) led to an excellent ee when employing AgOTf (entry 9). In conjunction with L2 and L3, this silver salt provides enantioselectivities similar to those achieved with AgSbF₆ (entries 10 and 11). Therefore, the combination of L1 with AgOTf was selected as optimal reaction conditions. An additional experiment lowering the reaction temperature did not improve the enantioselectivity (entry 12). Finally, with only 1 equiv. of AgOTf, similar yield and ee value were obtained (entry 13). ^[23]

Table 1.

Optimization of the reaction conditions for the enantioselective synthesis of 2 a.^[a]

Entry	L*	X	Solvent	Yield 2 a [b]	ee [c]
1	L1	SbF6	CH2Cl2	80	66
2	L2	SbF6	CH2Cl2	65	87
3	L3	SbF6	CH2Cl2	57	81
4	L1	SbF6	DCE	83	84
5	L2	SbF6	DCE	70	50
6	L3	SbF6	DCE	60	68
7	L1	BF4	DCE	71	76
8	L1	NTf2	DCE	77	76
9	L1	OTf	DCE	80	96
10	L2	OTf	CH2Cl2	38[d]	88
11	L3	OTf	CH2Cl2	37[e]	76
12[f]	L1	OTf	DCE	75	94
13[g]	L1	OTf	DCE	78	88

[a] Reactions were conducted with 1 a (0.2 mmol) and the catalyst in 2 mL of the corresponding solvent. [b] Isolated yield. [c] Determined by HPLC analysis. [d] Significant amounts of diketone 3 a, arising from hydration of the alkyne were obtained (see Supporting Information for details). [e] Uncomplete conversion. [f] Carried out at 0 °C. [g] Carried out with 3 mol % of AgOTf.

The optimized reaction conditions were then employed to evaluate the substrate scope for this new enantioselective synthesis of cyclopentenones. First, alkenynones 1 a–g bearing different aryl frameworks and a β‐styrenyl group were tested (Scheme 3). The reaction with substrates bearing aryl rings functionalized with alkyl, aryl, or alkoxy groups at para‐, meta‐ and ortho‐positions proceeds efficiently with high yields and enantioselectivities. The absolute configuration of compounds 2 was inferred from the X‐ray diffraction of 2 d, revealing a (S)‐configuration for the generated stereocenter. ^[24] In addition, the influence of the alkene geometry was studied by using (Z)‐1 a (as 6/1 mixture of Z/E isomers). Its reaction led to the opposite enantiomer but with lower enantioselectivity (68 % ee).

Scheme 3.

Scope of O‐aryl moiety for the enantioselective synthesis of cyclopentenones 2.

The effect of β‐substituent on alkenynones 1 was then studied, as summarized in Scheme 4. When using olefins with aromatic groups bearing halogen atoms at different positions, the corresponding cyclopentenones 2 h–m were obtained in high yields and excellent enantioselectivies. Similarly, electron‐rich aromatic rings were also well‐tolerated as R substituents providing access to cyclopentenones 2 n–r with good yields and high ee values. Alkenynones bearing a heterocycle, or an additional alkenyl group at the R position, were also suitable substrates leading to 2 s and 2 t, respectively, with high enantioselectivities. In addition, primary and secondary (c)‐alkyl groups were evaluated as β‐substituents of the alkenynone, and, gratifyingly, the desired cyclopentenones 2 u–x were obtained in high yields and enantioselectivities, even with a substrate possessing a simple methyl group (2 x). Moreover, a substrate bearing a functionalized alkyl group was also successfully realized (2 y). Finally, the cyclization of α,β‐disubstituted alkenynones 1 z and 1α was evaluated. In these cases, mixtures of diastereoisomers were obtained favoring the cis‐cyclopentenones, which were afforded with good enantioselectivities.

Scheme 4.

Scope of alkene moiety for the enantioselective synthesis of cyclopentenones 2.

As the initial hydroarylation reaction has also been described for other alkyne‐containing aromatics, related substrates were synthesized and evaluated (Scheme 5). Disappointingly, sulfonamide 4 a and sulfide 4 b did not evolve under the established conditions, even when employing the non‐chiral XPhosAuCl catalyst. Changing the −O−CH₂− tether by a −O−C(Me)₂− one, alkenynone 5, led to the expected cyclopentenone 6 in high yield and with good enantioselectivity. Moreover, with ketone 7, bearing a −CH₂−CH₂− tether, the tandem reaction proceeds efficiently leading to 8, although with almost no enantioselectivity.

Scheme 5.

Limitations and further scope of the enantioselective hydroarylation–Nazarov cyclization.

Interestingly, the scale‐up reaction using 1 g of 1 a can be conducted efficiently, providing 750 mg of 2 a with 96 % ee and with only a slight decrease in the yield (Scheme 6). In addition, a pure enantiomeric sample of 2 a could be obtained by recrystallization after chromatographic purification. In addition, a synthetic application of the product is shown in a further transformation involving oxidation to the coumarin derivative 9 a without erosion of the enantiomeric excess (Scheme 6).

Scheme 6.

Gram‐scale experiment and further derivatization. [a] ee value within parenthesis is obtained by recrystallization after column chromatography.

A control experiment with dialkenyl ketone 10 d was conducted to gain more mechanistic insights into the reaction. As shown in Scheme 7, when this plausible intermediate was treated with the gold catalyst under the established reaction conditions, no evolution occurred, thus suggesting that free chromene derivatives 10 are not involved as intermediates. Our mechanistic proposal implies the initial activation of the alkyne by the gold complex, 1‐[AuL*], that triggers an anti‐Michael hydroarylation reaction leading to cationic intermediate I. The removal of a proton, to recover the aromaticity, could be assisted by the carbonyl group favoring the formation of an intermediate such as II in which a competitive protodemetallation, leading to 10, could be slowed down, allowing subsequent Nazarov cyclization. In this way, intermediate III would be formed in which the stereogenic center has been created in an enantioselective manner due to the presence of the chiral complex attached to the molecule and the control over the absolute sense of conrotation. Similar to silicon‐directed Nazarov cyclizations, ^[9b] the gold complex in the intermediate III may be decisive in controlling the alkene formation and avoiding alternative isomers. Finally, demetallation and further tautomerization would lead to the cyclopenta[c]chromenones 2 (Scheme 7).

Scheme 7.

Control experiment and the postulated mechanism.

In conclusion, an unconventional approach to the asymmetric catalytic Nazarov cyclization has been described, probably due to the crucial formation of an alkenylgold intermediate. Accessible chiral bisphosphine‐gold complexes enable an enantioselective cascade reaction affording cyclopenta[c]chromenones with high enantiomeric excess, employing skipped alkenynones as substrates. The coordination of the gold catalyst to the alkyne prompted anti‐Michael hydroarylation followed by the Nazarov cyclization. The method is scalable and conveniently conducted at room temperature achieving cyclopentenones in good yields and excellent enantioselectivities. This approach could be useful in designing related enantioselective cascade reactions employing unconventional precursors.

Conflict of interest

The authors declare no conflict of interest.

Supporting information

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supporting Information

Supporting Information

M. Solas, S. Suárez-Pantiga, R. Sanz, Angew. Chem. Int. Ed. 2022, 61, e202207406; Angew. Chem. 2022, 134, e202207406.

Data Availability Statement

The data that support the findings of this study are available in the Supporting Information of this article.