Rapid Construction of Tetralin, Chromane, and Indane Motifs via Cyclative C−H/C−H Coupling: Four-Step Total Synthesis of (±)-Russujaponol F

Abstract

The development of practical C−H/C−H coupling reactions remains a challenging yet appealing synthetic venture because it circumvents the need to prefunctionalize both coupling partners for the generation of C−C bonds. Herein, we report a cyclative C(sp³)−H/C(sp²)−H coupling reaction of free aliphatic acids enabled by a cyclopentane-based mono-N-protected β-amino acid ligand. This reaction uses inexpensive sodium percarbonate (Na₂CO₃·1.5H₂O₂) as the sole oxidant, generating water as the only byproduct. A range of biologically important scaffolds, including tetralins, chromanes, and indanes, could be easily prepared by this protocol. Finally, the synthetic application of this methodology is demonstrated by the concise total synthesis of (±)-russujaponol F in a four-step sequence starting from readily available phenylacetic acid and pivalic acid through the sequential functionalizations of four C−H bonds.

Graphical abstract

Carbon-carbon (C−C) bond formation constitutes one of the most important classes of reactions in organic synthesis. Owing to its potential to shorten synthesis, the past two decades have witnessed rapid developments in using C−H activation strategies for the construction of C−C bonds.¹ While most coupling methods require prefunctionalized coupling partners (e.g. organoborons and organohalides), C−H/C−H coupling reactions offer a complementary strategy to construct a C−C bond directly from two simple C−H bonds (Scheme 1A).² Compared to traditional coupling methods, this green and atom-economical approach is highly attractive because water is potentially the sole stoichiometric byproduct of this process (Scheme 1A). To date, extensive studies have focused on the coupling of two relatively reactive C(sp²)−H bonds for biaryl synthesis,³ whereas only a few reactions have been reported for the construction of more challenging C(sp³)−C(sp²) bonds. Because these existing reaction protocols require exogenous directing groups (DGs) to promote cyclometallation, additional steps to install and remove the DG are necessary.^5,6 Additionally, reported methods pose practical limitations, such as the stoichiometric use of precious silver salts^4b,c,5,6b,c and harsh conditions^4b,c,5a,b,6 — with temperatures as high as 160 °C being reported. Moreover, current methods for C(sp³)−H/C(sp²)−H coupling initiated by C(sp³)−H activation are largely limited to more reactive heterocyclic C(sp²)−H bonds.^5a,b,6 Despite the great value that C−H/C−H coupling reactions might have for organic synthesis, the development of C(sp³)−H/C(sp²)−H coupling reactions that use both a practical oxidant and native substrates remains a significant challenge.

Scheme 1.

C−H Activation/C−C Bond-Forming Reactions

Recent advances in C−H functionalization have provided chemists with creative and strategic retrosynthetic disconnections that are otherwise difficult to achieve using traditional methods.⁷ However, for C−H functionalization strategies to truly improve the overall efficiency of synthesis, three criteria should be met: (1) the ability to use a wide range of simple starting materials to enable the synthesis of diverse natural product families; (2) the use of native functionalities as the DG; (3) the site-selectivity of C−H functionalization reactions should be precisely controllable. Given the ubiquitous nature of C−H bonds in organic molecules, synthetic sequences that incorporate multiple C−H functionalizations are particularly attractive for the efficient synthesis of natural products. However, approaches that meet these aforementioned criteria are challenging to execute and so uncommon in literature.^7a,8

To address these challenges, we herein report a cyclative C(sp³)−H/C(sp²)−H coupling reaction using a native free carboxylic acid as the DG (Scheme 1B). The use of a cyclopentane-based mono-N-protected β-amino acid ligand and a practical and inexpensive oxidant sodium percarbonate (Na₂CO₃·1.5H₂O₂) proved crucial to the success of this reaction. Tetralins, chromanes, or indanes, common frameworks in natural products could be readily prepared by this protocol (Figure 1). The synthetic application of this methodology is further demonstrated by a concise total synthesis of (±)-russujaponol F (the shortest and highest yielding to date) via multiple C−H functionalizations in four steps from readily available phenylacetic acid and pivalic acid (Scheme 1C), demonstrating the potential of C−H activation disconnections to enhance the ideality of synthesis⁹.

Figure 1.

Biologically significant natural products containing tetralin, chromane, or indane frameworks.

Aliphatic carboxylic acids are ubiquitous and synthetically versatile motifs and are often inexpensive reagents in organic chemistry; as such, they are privileged substrates for C−H activation reactions.¹⁰ Following our recent disclosure of the β-C(sp³)−H lactonization¹⁰ⁱ and acyloxylation^10j of free carboxylic acids using tert-butyl hydrogen peroxide (TBHP) as the sole oxidant, we initiated our investigation of cyclative C(sp³)−H/C(sp²)−H coupling reactions by selecting TBHP as the bystanding oxidant and aliphatic acid 1a as a model substrate. Under the optimal conditions of the aforementioned β-acyloxylation reaction^10j, we were delighted to observe a 50% ¹H NMR yield of the desired product 2a without forming competing reductive elimination products, such as the β-lactone or β-hydroxy acid (see Supporting Information, Table S1). Further investigation of the bystanding oxidants and bases revealed that the combination of Na₂CO₃·1.5H₂O₂ and LiOAc could further improve the yield to 57% (see Supporting Information, Table S1−S2). The yields using LiOAc are generally better than those using NaOAc as the metal additives under the optimized conditions (see Supporting Information, Table S4). The use of sodium percarbonate, one of the cheapest and most easily handled oxidants,¹¹ potentially renders this protocol practical and scalable. In light of recent advances in ligand-accelerated Pd(II)-catalyzed C−H activation,¹² we next searched for ligands that could substantially improve the reactivity of the catalyst. Guided by mono-N-protected amino acid (MPAA) ligand-enabled C(sp³)−H activation reactions of free carboxylic acids^10c,d,g,i,j, we tested a series of commercially available MPAA ligands (L1−L4): β-amino acid ligand L4 showed superior reactivity over α-amino acid ligands L1−L3 (57% vs. 19−45%), as was also observed in other C(sp³)−H functionalization reactions of free acids via Pd(II)/Pd(IV) catalytic cycles^10d,i,j. Through systematic modifications to the backbone of the β-amino acid ligand (L5−L10), we found that cis-cyclopentane-based ligand (±)-L9 gave the optimal reactivity (78% isolated yield). The superior reactivity of (±)-L9 might be attributed to the more rigid conformation enforced by the cyclopentane linkage. Control experiments showed that the yields were low in the absence of the ligand or in the presence of the γ-amino acid ligand (L11) (23% or 20%, respectively), indicating the importance of six-membered chelation by the ligand for reactivity.

With the optimal ligand and reaction conditions in hand, we evaluated the scope of the cyclative C(sp³)−H/C(sp²)−H coupling reaction (Table 2). A wide range of tertiary aliphatic acids bearing a single α-methyl group (1a−1e and 1h) or an α-gem-dimethyl group (1f and 1g) were all compatible, affording the tetralin products in moderate to good yields (52−78%). The reaction could also be conducted on a 2.0 mmol scale, delivering 2a in 69% yield. The attempted desymmetrization of the α-gem-dimethyl group of 1f using enantioenriched L9 resulted in racemic product. Less reactive free carboxylic acids containing α-hydrogens (1i−1l) also reacted in synthetically useful yields (35−65%). Among these, a variety of functionalities on the aryl rings such as methyl (2b), methoxy (2j and 2k), fluoro (2c, 2g, and 2l), and chloro (2d) as well as naphthyl (2e) were tolerated, with the halogen moiety (2d) serving as a useful synthetic handle for subsequent derivatization. This protocol could also be successfully extended to the synthesis of biologically important chromane products. β-Phenoxy carboxylic acids containing an α-gem-dimethyl group (1m−1r) or α-hydrogens (1s, from Roche ester) were all reactive substrates. While a range of electron-donating (methoxy, tert-butyl, cyclohexyl, and benzyl) (2s and 2n−2p) groups on the aryl ring were well tolerated to afford the desired products in good yields (70−85%), aliphatic acids containing electron-withdrawing (bromo and trifluoromethyl) groups (2q and 2r) showed comparatively low reactivity (31% and 23%), likely due to the sluggish nature of C(sp²)−H activations of electron-deficient arenes. Although intermolecular KIE experiments of electron-rich 1m and 1m-d₅ (k_H/k_D = 1.1) suggest that C(sp²)−H activation is not the rate-determining step (see Supporting Information, KIE experiments section), the possibility of C(sp²)−H activation with electron-deficient substrates as the rate-determining step cannot be ruled out. It is noteworthy that high regioselectivity was observed for the aliphatic acids containing meta-methoxy groups (1j and 1s), while the substrate bearing a tert-butyl group (1n) afforded a mixture of regioisomers (2n/2n’ = 3/1). Considering the previously observed high para-selectivity of Pd(IV)-mediated C−H coupling reactions of anisoles^3g,5c,22, it is likely that the alkyl-Pd(II) intermediate is oxidized to Pd(IV) prior to C−H activation. Under the present conditions, the carboxylic acid 1t containing either NBoc or NTs groups failed to deliver tetrahydroisoquinoline (THIQ) product 2t. This cyclative C−H/C−H coupling reaction was also amenable to the syntheses of indane scaffolds (2u−2w). Notably, an [F⁺] oxidant^3g,13 (1-fluoro-2,4,6-trimethylpyridinium tetrafluoroborate) showed superior reactivity for tertiary aliphatic acids containing an α-gem-dimethyl group (2v and 2w) (see Supporting Information, Table S5).

Table 2.

Substrate Scope of the Cyclative C(sp³)−H/C(sp²)−H Coupling Reaction^a,b

^aConditions A: 1 (0.1 mmol), Pd(OAc)₂ (10 mol%), (±)-L9 (10 mol%), LiOAc (1.0 equiv), Na₂CO₃·1.5H₂O₂ (2.0 equiv), HFIP (1.0 mL), 60 °C, 12 h.

^bIsolated yields.

^cConditions B: 1 (0.1 mmol), Pd(CH₃CN)₄(BF₄)₂ (10 mol%), Ag₂CO₃ (1.0 equiv), 1-fluoro-2,4,6-trimethylpyridinium tetrafluoroborate (2.0 equiv), HFIP (1.0 mL), 90 °C, 12 h. ^dOn 2.0 mmol scale.

Illudalane sesquiterpenes comprise a large family of natural products, which typically feature an indane core (for which various oxidation states are possible) bearing a challenging all-carbon quaternary center (Scheme 2A).¹⁴ Owing to their promising biological activities, tremendous efforts have been devoted to the total syntheses of these targets.^15,16 Given the power of this methodology for the construction of indane scaffolds, we embarked on the total synthesis of (±)-russujaponol F via multiple C−H functionalizations (Scheme 2B). Baudoin’s group reported the first total synthesis of russujaponol F in racemic and enantioselective forms based on a C(sp³)−H arylation strategy in 13 steps (26% yield) and 15 steps (12% yield) respectively.¹⁵ Beginning with phenylacetic acid 3 that is commercially available or synthesized through ortho-C−H methylation¹⁷, we were able to prepare aryl iodide 4 by esterification and subsequent mono-iodination¹⁸ of 3 using I₂ and Selectfluor in 79% yield. Investigation of the C−H arylation of pivalic acid indicated that, with ligand L12^10f,19, the mono-arylated product 5 could be obtained in 62% yield, along with 12% of the cyclative C−H/C−H coupling product 6 (see Supporting Information, Table S6). The formation of 6 under these conditions might be attributed to a second arylation of 5 with additional aryl iodide serving as the bystanding oxidant.²⁰ The cyclative C−H/C−H coupling was then performed under the standard conditions using an [F⁺] oxidant to give the desired product 6 in 41% yield. Finally, global reduction of 6 using LAH cleanly delivered (±)-russujaponol F in 96% yield, completing the total synthesis in four steps and 28% overall yield: the shortest and highest yielding total synthesis of russujaponol F to date.

Scheme 2.

Total Synthesis of (±)-Russujaponol F^a

^aConditions: (a) SOCl₂, EtOH, reflux, overnight; I₂ (0.5 equiv), Selectfluor (0.5 equiv), CH₃CN, 60 °C, 3 h. (b) Pd(OAc)₂ (10 mol%), L12 (10 mol%), pivalic acid (3.0 equiv), CsOAc (1.0 equiv), Ag₂CO₃ (2.0 equiv), HFIP, 80 °C, 12 h. (c) Pd(CH₃CN)₄(BF₄)₂ (10 mol%), Ag₂CO₃ (1.0 equiv), 1-fluoro-2,4,6-trimethylpyridinium tetrafluoroborate (2.0 equiv), HFIP, 90 °C, 12 h. (d) LAH (3.0 equiv), THF, 0 °C to rt, overnight.

On the basis of literature precedents³⁻⁵ and our recent work on the C−H activation of free acids^10i,j, we propose that this cyclative C(sp³)−H/C(sp²)−H coupling reaction proceeds via a Pd(II)/Pd(IV) catalytic cycle as outlined in Scheme 3. First, coordination of Pd(OAc)₂ to an MPAA ligand generates the active LPd(II) species. After coordination of the acid substrate 1 to Pd, both the countercation Na⁺ or Li⁺ and the MPAA ligand accelerate the cyclopalladation of the β-C(sp³)−H bond to form int-I. Next, oxidative addition of the hydrogen peroxide occurs to produce int-II, a process established in previous studies on the oxidation of Pd(II) to Pd(IV) by benzoyl peroxide^21a, tert-butyl peroxyacetate^21b, or hydrogen peroxide^21c,d. In the previously reported β-lactonization¹⁰ⁱ or acetoxylation^10j reactions, selective reductive elimination yields β-lactone or β-acetoxylated carboxylic acid. In this case, a reactive phenyl group on the side chain of the substrate undergoes a second C(sp²)−H activation of int-II to deliver int-III via a seven or six-membered palladacycle, enabled by the facile dissociation of the weakly coordinating free acid.²² However, it is also possible that the Pd(II) intermediate int-I performs the second C−H activation prior to the oxidative addition of hydrogen peroxide that generates int-III. Finally, reductive elimination of int-III generates the cyclative C−H/C−H coupling product 2 and regenerates the LPd(II) species.

Scheme 3.

Proposed Mechanism for Cyclative C(sp³)−H/C(sp²)−H Coupling Reaction

In summary, we have realized a Pd(II)-catalyzed cyclative C(sp³)−H/C(sp²)−H coupling reaction enabled by a cyclopentane-based mono-N-protected β-amino acid ligand. The use of inexpensive sodium percarbonate as the sole oxidant and native free carboxylic acids as the directing group renders this reaction highly practical and potentially amenable to large-scale manufacturing. A range of biologically significant scaffolds, including tetralins, chromanes, and indanes, could be readily prepared by this protocol. The synthetic application of this methodology was demonstrated by a concise total synthesis of (±)-russujaponol F via multiple C−H functionalizations in four steps from readily available phenylacetic acid and pivalic acid.

Table 1.

Ligand Investigation for the Cyclative C(sp³)−H/C(sp²)−H Coupling Reaction^a,b

^aConditions: 1a (0.1 mmol), Pd(OAc)₂ (10 mol%), ligand (L) (10 mol%), LiOAc (1.0 equiv), Na₂CO₃·1.5H₂O₂ (2.0 equiv), HFIP (1.0 mL), 60 °C, 12 h.

^bThe yields were determined by ¹H NMR analysis of the crude product using CH₂Br₂ as the internal standard.

^cIsolated yield.