Regio- and diastereoselective C-C coupling of α-olefins and styrenes to 3-hydroxy-2-oxindoles by Ru-catalyzed hydrohydroxyalkylation.

α-Olefins are abundant petrochemical feedstocks used in the manufacture of diverse chemical products.[1] Despite their commercial significance, intermolecular catalytic reductive C-C couplings of α-olefins to carbonyl compounds are unknown, withstanding the special case of hydroformylation.[2,3,4] Motivated in part by the prospect of addressing this deficiency, our laboratory embarked upon a systematic exploration of hydrogen mediated reductive couplings beyond hydroformylation.[5] These efforts led to a broad family of catalytic C-C couplings wherein two or more π-unsaturated reactants are hydrogenated to form a single, more complex product.[5] Such transformations typically proceed through pathways involving metallacycle formation and hydrogenolysis (Figure 1, top). Based on these studies, related C-C bond forming transfer hydrogenations were developed, in which hydrogen transfer from alcohols to π-unsaturated reactants via hydrometallation triggers generation of organometal-carbonyl pairs that combine to form products of addition (Figure 1, middle).[5] In a more recent advance, a third pathway involving alcohol mediated transfer hydrogenolysis of metallacycles was uncovered, as illustrated in couplings of α-hydroxy esters to isoprene or myrcene to form products of prenylation or geranylation, respectively, and related couplings dienes of 3-hydroxy-2-oxindoles (Figure 1, bottom).[6] The availability of this novel mechanistic pathway prompted a reinvestigation of the coupling of α-olefins. Here, we report that the ruthenium(0) catalyst generated from Ru3(CO)12 and tricyclohexylphosphine, PCy3, promotes direct C-C coupling of α-olefins and styrenes to 3-hydroxy-2-oxindoles to form branched products of hydrohydroxyalkylation as single diastereomers. To our knowledge, this work represents the first example of the metal catalyzed hydrohydroxyalkylation of unactivated olefins. Figure 1 Three distinct catalytic mechanisms enable C-C bond forming hydrogenation and transfer hydrogenation. In an initial experiment, N-benzyl-3-hydroxy-2-oxindole 1a was exposed to propylene 2b under conditions used to promote C-C coupling of substituted mandelic esters to dienes.[6] However, only trace quantities of the isopropyl-substituted oxindole 3b was observed. Nevertheless, the branch-regioselectivity suggested the predicted mechanism involving carbonylalkene oxidative coupling was operative but inefficient. As it was previously observed that carboxylic acids co-catalyze hydrogenolysis of oxa- and azametallacycles to enhance rate and conversion,[7] a range of carboxylic acids were screened under the aforementioned conditions. Remarkably, upon use of 1-adamantanecarboxylic acid (10 mol%) as co-catalyst, the isopropyl-substituted oxindole 3b was isolated in 90% yield as a single regioisomer (eqn. 1). (eqn. 1) These optimal conditions were applied to the C-C coupling of 1a to α-olefins 2a-2f (Table 1). The corresponding adducts 3a-3f were obtained in excellent yield with complete levels of branched regioselectivity. As illustrated in the formation of the 3-hydroxy-2-oxindoles 3c-3f, exceptional levels of diastereoselectivity are observed in couplings that generate stereogenic centers. The assignment of relative stereochemistry was corroborated by single crystal x-ray diffraction analysis of adduct 3c. In terms of functional group compatibility, the efficient coupling of olefins 2e and 2f, which incorporate alkoxy and silyl groups at the allylic position, is noteworthy. Finally, although Ru3(CO)12 derived catalysts are known to promote olefin isomerization,[8] the coupling of 1a to allylbenzene 2d occurs efficiently. This result is significant as β-methylstyrene and other 1,2-disubstituted olefins do not participate in C-C coupling under these conditions. Table 1 Ruthenium catalyzed hydrohydroxyalkylation of α-olefins 2a-2f and 3-hydroxy-2-oxindole 1a.a These conditions for ruthenium(0) catalyzed hydrohydroxyalkylation were applied to the coupling of 3-hydroxy-2-oxindole 1a to substituted styrenes 2g-2r (Table 2). With the exception of 2-methoxy styrene 2k, the coupling of electron neutral and electron rich styrene derivatives 2g-2m occurs efficiently with complete levels of branched regio- and diastereoselectivity. The assignment of relative stereochemistry was corroborated by single crystal x-ray diffraction analysis of adduct 3g. Formation of the 3-hydroxy-2-oxindole 3r, which incorporates a thiophene moiety, further illustrates functional group compatibility. For styrenes 2n-2q, increasing electron deficiency is accompanied by erosion in regioselectivity, which is attributed to a non-metal catalyzed background reaction involving classical conjugate addition. The veracity of this hypothesis is corroborated by the fact that N-benzyl-3-hydroxy-2-oxindole 1a spontaneously reacts with 2-vinylpyridine 2s in the absence of a metal catalyst to form the linear product exclusively (eqn. 2). (eqn. 2) Table 2 Ruthenium catalyzed hydrohydroxyalkylation of styrenes 2g-2r and 3-hydroxy-2-oxindole 1a.a The use of 1-adamantanecarboxylic acid as a co-catalyst suggests substrates that possess acidic functional groups should be tolerated in the C-C coupling. To probe this question, the 3-hydroxy-2-oxindole 1b, which lacks an N-protecting group, was exposed to styrene 2g under slightly modified conditions for hydrohydroxyalkylation. Due to the poor solubility of 1b in m-xylene, the reaction was conducted in chlorobenzene for 72 hours. The anticipated adduct 3g-NH was formed in 72% yield as a single regio- and diastereomer (eqn. 3). (eqn. 3) A plausible catalytic mechanism accounting for the effect of 1-adamantanecarboxylic acid begins with the combination of Ru3(CO)12 and tricyclohexylphosphine, PCy3, to form a discrete mono-metallic ruthenium(0) complex.[9] Oxidative coupling necessitates initial dehydrogenation of 3-hydroxy-2-oxindole 1a to form isatin 1c, which finds precedent in the Ru3(CO)12 catalyzed oxidation of alcohols employing olefins and alkynes as hydrogen acceptors,[10,11] as well as Ru3(CO)12 catalyzed aminations of secondary alcohols, which involves dehydrogenation of α-hydroxy esters.[12c] Indeed, for couplings of styrene 2g, small quantities of ethylbenzene are detected in the 1H NMR spectrum of crude reaction mixtures. Oxidative coupling of isatin 1c and styrene 2g to form oxametallacycle I finds precedent in the work of Chatani and Murai on Pauson-Khand type reactions of 1,2-diones,[13] and studies from our own laboratory on the prenylation of substituted mandelic esters.[6] We postulate that direct transfer hydrogenolytic cleavage of the oxaruthenacycle I, which would involve initial protonation of oxaruthenacycle I to form ruthenium alkoxide III, may be prohibitively slow, whereas protonolysis of oxaruthenacycle I by 1-adamantanecarboxylic acid to form ruthenium carboxylate II and subsequent exchange of the carboxylate to form ruthenium alkoxide III are relatively rapid. The ruthenium alkoxide III suffers β-hydride elimination to concomitantly generate ruthenium hydride IV and isatin 1c. Finally, C-H reductive elimination produces the product 3g and the initial ruthenium(0) complex to close the catalytic cycle. As depicted in the proposed stereochemical model, oxidative coupling should occur so as to avoid steric interactions between the olefin substituent and the arene moiety (Scheme 1). Scheme 1 Proposed catalytic mechanism illustrating the effect of the carboxylic acid co-catalyst and stereochemical model. In summary, we report the first examples of the metal catalyzed hydrohydroxyalkylation of unactivated olefins, as illustrated by their C-C coupling to N-benzyl-3-hydroxy-2-oxindole 1a. This process was enabled by the recent discovery of a novel reactivity pattern wherein the metallacyclic intermediates obtained upon carbonyl-olefin oxidative coupling suffer transfer hydrogenolysis mediated by a secondary alcohol reactant to liberate the product of hydrohydroxyalkylation and regenerate the carbonyl partner for oxidative coupling (Figure 1, bottom).[6] Future studies will focus on the development of related catalysts for the direct conversion of α-olefins and aliphatic primary and secondary alcohols to form higher alcohols in the absence of stoichiometric byproducts.