Synthesis of indoles, indolines, and carbazoles via palladium-catalyzed C─H activation

Abstract

The construction of indole, indoline, and carbazole heterocycles has been of significant interest in the synthetic community over the last century due to their prevalence in natural products and other biologically active compounds. In particular for indoles, many conventional methods developed to date require highly pre-functionalized arene precursors, diminishing their attractiveness as “green” syntheses. Carbon–hydrogen bond activation, on the other hand, presents an elegant solution to this problem and can achieve the construction of indoles and their derivatives from comparatively simpler arene precursors. In this short review, we discuss various approaches for preparing indoles, indolines, and carbazoles via palladium-catalyzed C─H bond activation, highlighting their reaction mechanisms and synthetic applications.

Graphical abstract

Indoles and their close derivatives, indolines and carbazoles, have been of interest to synthetic chemists due to their prevalence in natural products and biologically active molecules. In particular, the activation of inert carbon–hydrogen (C─H) bonds presents an elegant means to prepare these heterocycles from simple precursors. This review will highlight the versatility of palladium-catalyzed C─H activation for the preparation of indoles, indolines, and carbazoles, discussing intramolecular cyclization and multi-component ring construction strategies.

1. Introduction

Indoles and their derivatives are among the most common heterocycles found in nature. As a result, a wide variety of compounds, including amino acids, natural products, and approved pharmaceutical drugs, contain such a structural motif (Fig. 1), many of which exhibit rich biological activity [1]. As a consequence, significant efforts have been made to prepare substituted indoles [2-4] and their close derivatives, indolines [5] and carbazoles [6]. Traditional approaches often involve harsh reaction conditions, complex substrates, or potentially toxic byproducts, diminishing their attractiveness as “green” syntheses. An emerging tool to mitigate these concerns is carbon–hydrogen (C─H) bond activation, whereupon an inert C─H bond is directly transformed into the desired carbon–carbon (C─C) or carbon–heteroatom (C─X) bond by a transition metal catalyst [7, 8]. In this regard, substrate complexity can be decreased, as pre-functionalization would be unnecessary, allowing for the use of more readily available precursors for heterocycle syntheses. In addition, harsh conditions or toxic byproducts can often be avoided.

Fig. 1.

Natural products and FDA-approved drug molecules bearing indoles and carbazoles.

There have been excellent reviews on the topic of indole synthesis via C─H bond activation reported until 2017 [9-11]; however, significant advances in the field have appeared in the literature since. In this short review, we will highlight different synthetic strategies that have been developed for the construction of indoles and their close derivatives, indolines and carbazoles, via palladium-catalyzed C─H bond activation. A particular focus is given to approaches developed after 2016. It is not our intent to comprehensively cover each transformation developed to date, but to emphasize the versatility of palladium catalysis for the construction of these heterocyclic rings from a variety of simple precursors. The content is divided based on whether the ring is formed through an intramolecular (unimolecular) or a multi-component (bi- or trimolecular) event. Besides reaction development, mechanistic details and synthetic applications of the transformations are also discussed.

2. Synthesis of indoles and their derivatives via palladium-catalyzed C─H bond activation

To date, many methods have been developed for indole synthesis, in addition to indole functionalization [12,13]. Among these synthetic approaches, the oldest and most well-known indole-forming reaction is likely the Fischer indole synthesis. Discovered in 1883, this transformation uses an aryl hydrazone, formed by condensation of an aryl hydrazine with a carbonyl, which is treated with acid at elevated temperatures in order to promote a [3,3]-sigmatropic rearrangement, eventually resulting in the formation of the indole ring with expulsion of gaseous ammonia (Scheme 1a) [14]. In a related work reported over a century later in 1998, Buchwald and co-workers showed that the aryl C─N bond can be forged via a palladium-catalyzed coupling reaction between an aryl bromide and a hydrazone, which can then be treated under standard Fischer conditions to form the indole ring (Scheme 1b) [15]. This advance allowed for a new disconnection to be made when considering how to prepare aryl hydrazones.

Scheme 1.

Fischer indole synthesis and Buchwald’s modification.

In these examples, however, relatively harsh acidic conditions are required in order to furnish the final products. Other strategies have since been developed, such as the Mori-Ban (Scheme 2a) [16] and Larock (Scheme 2b) [17] indole syntheses, which employ milder palladium-catalyzed conditions to construct the heterocycles; however, these examples suffer from the necessity of more complex pre-functionalized arene starting materials. More recently, Jørgensen and coworkers have found 1,2-dihaloarenes to be effective substrates for Pd-catalyzed indole synthesis via an amination/cyclization cascade with allyl amines, although this work exhibits the same drawback, resulting in a somewhat limited aryl substrate selection (Scheme 2c) [18]. Additionally, this method was recently extended to nickel catalysis by Tian, allowing for the reaction to proceed from aryl dichloride starting materials [19]. Although all of these reactions efficiently deliver indole products, methods that can avoid using doubly functionalized arene precursors would further simply the syntheses.

Scheme 2.

Pd-catalyzed synthesis of indoles from pre-functionalized arenes.

Clearly, substrate complexity is one factor that should be considered when designing a synthesis, and it could influent the “greenness” of the overall process. Over the past few decades, many innovative approaches have emerged, which directly transform an aryl C─H bond into the desired C─C or C─X bond, reducing overall chemical byproducts and enabling a rapid increase in molecular complexity from simple precursors [7, 8]. Consequently, these C─H bond activation methods are advantageous when compared to traditional aryl halide cross-coupling strategies, which often require the installation of the halogen atom (or (pseudo)halogen unit) prior to the coupling reaction if the desired substrate is not commercially available (Scheme 3).

Scheme 3.

Traditional cross-coupling vs. C─H bond activation strategies.

Palladium catalysis, in particular, offers a wide variety of scaffolds and approaches for the synthesis of indoles and their close derivatives via both traditional cross-coupling and C─H bond activation strategies. The aminating reagents have also been designed to fulfill a dual role as the oxidant in many examples, mitigating the necessity for potentially harmful external oxidants. By applying C─H bond activation towards the synthesis of these heterocycles, an inert C─H bond can be directly transformed into a C─C or a carbon–nitrogen (C─N) bond. Specifically, two major classifications can be designated with regard to indole synthesis: (1) intramolecular and (2) multi-component construction of the five-membered nitrogen heterocycles, both of which can be further defined depending on the bonds that are formed (vide infra).

2.1. Intramolecular cyclization strategies for the synthesis of indoles and their derivatives

Intramolecular construction of indoles and their close derivatives via palladium-catalyzed C─H bond activation can be achieved through a few possible disconnections: either of the C─N bonds, or of the C─C bond (Scheme 4). The latter can offer complementary bond disconnection strategies compared to the Heck cyclization method (Scheme 2a), with the added benefit of allowing less functionalized arene precursors to deliver the heterocyclic products. A recent review has covered this area quite thoroughly [20], thus only a selection of significant examples are discussed here. While many of these transformations can be carried out under mild conditions with relatively benign oxidants, some transformations required the addition of stronger oxidants to promote the formation of Pd(IV) intermediates, which could more readily undergo C─N bond reductive elimination to form the desired products than their Pd(II) counterparts.

Scheme 4.

Strategies for the intramolecular construction of indoles, indolines, and carbazoles.

2.1.1. Synthesis of indoles, carbazoles, and indolines via intramolecular C─H bond amination

Carbazoles and indolines were the first indole derivatives synthesized via C─H bond amination. In 2005, Buchwald and co-workers first developed a palladium-catalyzed protocol for the intramolecular synthesis of carbazoles using N-acetyl 2-aminobiphenyl starting materials (Scheme 5a) [21, 22].

Scheme 5.

Intramolecular synthesis of carbazoles via C─H amination.

An electrophilic palladation pathway was initially proposed for the C─H bond activation (Scheme 5b); however, it was observed that substrates with electron-deficient substituents para to the C─N bond forming site performed better compared to those with electron-rich substituents, suggesting that some other C─H activation mechanism was operating (Scheme 6a). Moreover, the transformation was found to be inefficient if the temperature was reduced below 120 °C. From these results, two plausible alternative mechanisms were proposed: Heck- or Wacker-type cyclization pathways (Scheme 6b). Although these are the pathways proposed by the authors, it should be noted that a mechanism involving a concerted metalation-deprotonation (CMD) pathway with the acetate ligand cannot be excluded (Scheme 5b).

Scheme 6.

Substituent effects and revised mechanisms.

Shortly after, Yu and co-workers employed a similar aryl C─N bond disconnection for the synthesis of indolines, with the major difference being the role of the oxidant (Scheme 7a) [23]. In this particular reaction, the oxidant was necessary in order to promote formation of a higher-valent palladium(IV) species, which could readily undergo the C─N reductive elimination; the oxidant was crucial for the transformation, as C─N bond formation was found to be quite challenging for the lower-valent Pd(II) species with these substrates (Scheme 7b). A wide range of oxidants was surveyed, finding an F⁺ reagent to be the best oxidant for the reaction, which could smoothly deliver the corresponding indoline products. This oxidant, in particular, minimized non-productive reductive elimination pathways (e.g., halogenation or acetoxylation side-products). The synthetic utility of the method was highlighted with a concise synthesis of 4-bromoindole, an important precursor in the synthesis of ergot alkaloids [24], which was more efficient than previously developed routes (Scheme 7c). The same group later improved upon this transformation with the more readily removable 2-pyridinesulfonyl protecting group, using PhI(OAc)₂ as the oxidant [25].

Scheme 7.

Synthesis of indolines via C(sp²)–H amination.

Concurrently, Glorius and co-workers developed a complementary C(sp³)–H amidation strategy for the synthesis of indolines [26]. A wide-ranging substrate scope was demonstrated, using the commercially available oxidant, AgOAc (Scheme 8a). In this transformation, an acetamide unit acts as an anionic directing group towards the palladium catalyst, facilitating the C(sp³)–H bond activation. After ruling out a potential acetoxylation/nucleophilic substitution pathway, it was proposed that the Pd(II) intermediate could either undergo reductive amination to furnish the indoline product with a Pd(0) species, or be oxidized to a Pd(IV) intermediate, which could readily undergo C─N bond formation to regenerate the active Pd(II) catalyst (Scheme 8b).

Scheme 8.

Synthesis of indolines via C(sp³)–H amidation.

Hartwig and co-workers developed the first redox-neutral intramolecular synthesis of indoles using oxime esters as both the amine source and oxidant in 2010 (Scheme 9a) [27]. The reaction was proposed to proceed through initial oxidative addition of Pd(0) into the oxime N─O bond, followed by tautomerization and subsequent C─H bond activation on the proximal phenyl ring. C─N reductive elimination could then smoothly deliver the desired indole product, with no need of an external oxidant (Scheme 9b). At the time of publication, Pd(0) was merely proposed to undergo oxidative addition into the N─O bond of the oxime; this proposed intermediate was isolated and its structure was unambiguously elucidated via X-ray crystallography in this work, confirming the proposed mechanism (Scheme 9c). This transformation exemplifies a “green” synthesis, as the oxime starting materials can readily be prepared from ketones, and the reaction itself generates a benign acetate by-product, without the need for any potentially toxic or harmful oxidants.

Scheme 9.

Redox-neutral intramolecular synthesis of indoles from oxime esters.

The analogous oxidative indole synthesis from amine substrates was later developed by Youn in 2014, using oxone as the terminal oxidant to prepare N-Ts indoles (Scheme 10a) [28]. In a closely-related synthesis of carbazoles [29] from the same group, it was proposed that the strong oxidant could generate a high-valent Pd(IV) intermediate, which could readily undergo reductive elimination to forge the C─N bond. In this regard, the oxidant fulfills a similar role as the oxidant in Yu’s indoline synthesis (Scheme 7) [23]. This template was further expanded by Stahl and co-workers in 2016, using molecular oxygen as the terminal oxidant, for the preparation of N-acetyl indoles (Scheme 10b) [30]. Mechanistically, this transformation was proposed to operate similarly to the intramolecular carbazole synthesis developed by Buchwald [21], which was proposed to proceed through a Pd(II)/Pd(0) catalytic cycle (Scheme 5b).

Scheme 10.

Intramolecular synthesis of indoles via oxidative C─H bond amination.

Other Pd-catalyzed C─H bond animation transformations have been developed for the synthesis of indoles from 2-nitrostyrene or β-nitrostyrene compounds [31-33]. However, the mechanisms of these transformations are not entirely understood, and likely do not involve a C─H activation step with the formation of a C–Pd bond; for example, a nitrene insertion pathway was initially proposed to be involved the C─N bond forming step [31]. As such, transformations of this type are not discussed.

2.1.2. Synthesis of indoles and carbazoles via intramolecular C─H bond carbonation

As reported by Mori and Ban [16], a typical Heck cyclization strategy for indole synthesis begins with the formation of an arylpalladium(II) species from an aryl halide (Scheme 2a). Larock and co-workers hypothesized that this arylpalladium(II) species could instead be formed via a 1,4-palladium migration with an unactivated arene C─H bond, which could then undergo Heck cyclization with an allylic amine to afford indole products (Scheme 11a) [34, 35]. In this regard, multi-substituted indoles can be prepared in a single step from simple 3-iodoaniline precursors. Alternatively, subsequent C─H bond activation on a nearby phenyl ring can deliver multi-substituted carbazole products (Scheme 11b).

Scheme 11.

Synthesis of indoles and carbazoles via 1,4-palladium migration.

A deuterium labeling study was conducted in order to further elucidate the mechanism of the ortho-C─H bond migration. Using a deuterium labeled oxygen-tethered substrate, incorporation of 70% deuterium, or 85% deuterium with a large excess of D₂O, on the newly formed internal alkene was observed (Scheme 12a). Consequentially, it was proposed that the alkenylpalladium(II) intermediate likely oxidatively inserted into the ortho-C─H bond, resulting in the formation of a Pd(IV)-hydride species. This hydride intermediate was poised to undergo facile C─H reductive elimination with the alkenyl ligand to form the key arylpalladium(II) intermediate, which could go on to cyclize with a proximal olefin to form indole products or undergo another C─H activation to prepare carbazoles or dibenzofurans (Scheme 12b). Since the deuterium incorporation was not 100%, H/D exchange was proposed to occur with this Pd(IV)-hydride intermediate by way of reductive elimination of an equivalent of acid, which is evident from the increase in deuterium incorporation with an excess of a deuteric additive. Although, a mechanism involving concerted sigma bond metathesis, with reversible C─H activation slightly eroding the deuterium incorporation at the ortho position prior to the metathesis, cannot be excluded. This work highlights the utility of metal migrations to prepare intermediates in typical aryl halide-initiated cross-coupling reactions, allowing for a rapid increase in molecular complexity from simple substrates.

Scheme 12.

Probing the mechanism of the 1,4-palladium migration.

Glorius and co-workers later developed an intramolecular synthesis of indoles via construction of the C─C bond. Aryl enamine precursors were employed in a cross-dehydrogenative coupling (CDC), whereupon an annulation was achieved by the extrusion of two hydrogen atoms from the starting material (Scheme 13a) [36]. Moreover, the substrates could be easily prepared from commercially available anilines. Mechanistically, after electrophilic palladation of the enamine moiety, an electrophilic palladation pathway was initially proposed for the subsequent aryl C─H bond activation. However, electron-donating substituents para to the amine resulted in a slower reaction rate, suggesting against this proposed electrophilic palladation pathway (Scheme 13b). Additionally, a large primary kinetic isotope effect (KIE) of 4.6 suggested some other mechanism was operating (Scheme 13c). With these results in hand, the mechanism was revised to involve one of two potential pathways: σ-bond metathesis or base-assisted deprotonation of the arene C─H bond (Scheme 13d).

Scheme 13.

Cross-dehydrogenative coupling for the synthesis of indoles.

Clearly, indole synthesis is greatly simplified by the intramolecular C─H cyclization strategy, compared to conventional cross-coupling approaches with aryl halides. In these transformations, however, the substrates need to be carefully designed in order to facilitate the cyclization. Construction of the heterocyclic ring from two or three components, on the other hand, would allow for even simpler starting materials to afford the desired indole products.

2.2. Multi-component synthesis of indoles, indolines, and carbazoles

Multi-component construction of the pyrrole nucleus of indoles, along with their close derivatives, through C─H bond activation would simultaneously form two or three bonds; therefore, simpler building blocks have been employed as substrates. This section primarily focuses on various synthetic strategies that assemble the 5-membered nitrogen heterocycle through different bond disconnections (Scheme 14).

Scheme 14.

Strategies for the multi-component construction of indoles, indolines, and carbazoles.

2.2.1. Bimolecular synthesis of indoles from anilines

Jiao and co-workers have developed a modified indole synthesis based on the cross-dehydrogenative coupling (CDC) strategy [36], which involves a Michael addition to prepare the aryl enamines in situ, followed by the CDC reaction [37]. In this regard, simple, commercially available anilines and internal alkynes can be directly transformed to indoles (Scheme 15a). Moreover, molecular oxygen is used as the sole oxidant in the reaction, avoiding the use of potentially toxic reagents in the reaction. While the authors proposed a tandem Michael addition/electrophilic palladation/CDC pathway, they could not rule out an initial aminopaladation of the alkyne to directly form the alkenylpalladium(II) species (Scheme 15b). Beyond this first C─N bond formation, the transformation also differs mechanistically from Glorius’ report (Scheme 13a) [36]; an electrophilic palladation pathway was proposed for the second C─H activation because of an observed intramolecular competition KIE of 1.2 (Scheme 15c). Finally, the synthetic utility of the transformation was highlighted by conducting a formal synthesis of a high-affinity 5-HT₃ receptor antagonist (Scheme 15d) [38].

Scheme 15.

Synthesis of indoles from anilines and alkynes.

This template was later expanded to better accommodate unsymmetrical alkynes by Zhang, Cao, and co-workers for the synthesis of 2-perfluoroalkyl indoles [39]. Additionally, Yoshikai [40] and Xiao [41] demonstrated that alkynes could be replaced with ketones for the one-pot synthesis of indoles from anilines.

2.2.2. Bi- and tri-molecular synthesis of indoles, indolines, and carbazoles via construction of both C─N bonds

A diaziridinone reagent, di-tert-butyldiaziridinone, has been employed extensively in the multi-component synthesis of N-tBu indoles, although this reagent was first used in the synthesis of indolines. This reagent is also proposed to serve a dual role as an oxidant towards key pallada(II)cycle intermediates, promoting the formation of high-valent Pd(IV) intermediates, which can readily form the desired C─N bonds (vide infra). The resulting N-tBu-protected products can be readily deprotected by treatment with strong acid, making them attractive substrates for nitrogen heterocycle synthesis (Scheme 16).

Scheme 16.

Synthesis of indoles and indolines with a diaziridinone reagent

Inspired by their previous studies of the reactivity of di-tert-butyldiaziridinone with terminal olefins [42], Shi and co-workers first used this diaziridinone reagent for the synthesis of spirocyclic indoline products from α-methylstyrenes (Scheme 17) [43]. This transformation was proposed to involve two catalytic processes – the formation of an allylic urea intermediate and subsequent transformation of this intermediate into the spirocyclic indoline product.

Scheme 17.

Synthesis of spirocyclic indolines via arene C─H bond activation.

The reaction mechanism was proposed to begin with oxidative addition of Pd(0) into the N–N bond of the diaziridinone, which could form an allylic Pd(II) species upon reaction with the α-methylstyrene. This allylic species was proposed to form an allylic amine upon reductive elimination, which could coordinate with another Pd(II)-urea unit. Aminopalladation of the amine with the olefin could form the 5-membered cyclic urea motif and an alkylpalladium(II) intermediate, which was poised to undergo a C─H bond activation at the arene ortho position to form a 5-membered pallada(II)cycle. This palladacycle was proposed to undergo yet another round of oxidative addition into the N–N bond of the diaziridinone to form a higher-valent Pd(IV)-urea species, which could deliver the product through potentially two distinct pathways: pathway (a) could involve C─N reductive elimination to form potentially two 8-membered palladacycles, which then underwent β-nitrogen elimination and C─N reductive elimination to produce the same indoline product (Scheme 18a); pathway (b) could involve formation of a Pd(IV)-nitrene species, which could furnish the indoline product via a sequence involving nitrene insertion and C─N reductive elimination (Scheme 18b).

Scheme 18.

Proposed mechanisms for the Pd-catalyzed spirocyclic indoline synthesis.

In an effort to confirm their proposed reaction mechanism, the intermediacy of the proposed allylic urea intermediate was probed by subjecting it to the standard reaction conditions, obtaining the corresponding indoline product in 35% yield (Scheme 19a), and an aryl pallada(II)cycle was found to be a competent catalyst for the transformation, even producing its corresponding indoline product when used in stoichiometric quantities (Scheme 19b). Together, these results support the proposed reaction mechanism.

Scheme 19.

Mechanistic probes for the spirocyclic indoline formation.

On the other hand, a strategy that can catalytically prepare the palladacycle without the spirocyclic moiety would further increase the synthetic potential of the diaziridinone reagent. Gratifyingly, the same group later modified their strategy to selectively install an N-tBu amine moiety at the ortho position of aryl iodides, which furnished indolines upon reaction with norbornenes (Scheme 20a) or a tethered alkene (Scheme 20b) [44]. Moreover, the product obtained from 2,5-norbornadiene allows for a retro Diels-Alder reaction to occur at an elevated temperature, affording the aromatic indole product (Scheme 20c).

Scheme 20.

Synthesis of indolines via a three-component coupling.

Interestingly, a bidentate ligand (such as BINAP) failed to deliver any of the desired product, possibly because it prevents coordination of the diaziridinone to the key pallada(II)cycle intermediate. Mechanistically, the transformation was proposed to begin with oxidative addition of Pd(0) into the aryl iodide, followed by migratory insertion of the alkene into the aryl–Pd bond to give an alkylpalladium(II) species, which could undergo a C─H activation with the ortho-C─H bond to form the key pallada(II)cycle with two carbon ligands. As shown in their previous study [43], Shi proposed that this electron-rich palladacycle could then react with the diaziridinone via an oxidative addition pathway to furnish a Pd(IV) intermediate, eventually resulting in the formation of the desired indoline products through the previously proposed pathways (Scheme 21a). Gratifyingly, the potential side-reaction of the diaziridinone with Pd(0) was avoided, likely due to the high reactivity of aryl iodides towards Pd(0). Alternative reactivity was also observed when using a phenyl-substituted alkene. This led to C─H activation taking place on this phenyl substituent, instead of the starting aromatic ring, presumably to minimize the overall strain of the pallada(II)cycle intermediate. Consequentially, this less strained palladacycle furnishes a spirocyclic indoline product (Scheme 21b), the structure of which was unambiguously confirmed by X-ray crystallography of a related derivative.

Scheme 21.

Proposed mechanism for the synthesis of indolines from aryl iodides, norbornene, and di-tert-butyldiaziridinone and spirocyclic indoline formation.

The diaziridinone reagent was also shown to be effective for the preparation of carbazoles and an indole by Zhang and co-workers (Scheme 22a) [45]. In this work, C(sp²)–H bond activation occurred on a proximal arene/olefin after oxidative addition of Pd(0) into the aryl iodide, resulting in the formation of a 5-membered pallada(II)cycle, which could further react with the diaziridinone as in the previous examples (Scheme 22b). The intermediacy of this proposed palladacycle was proved in a stoichiometric study, albeit in a reduced yield, presumably because of the bipyridine ligand being a less efficient ligand than triphenylphosphine for the amination process (Scheme 22c). In particular, this transformation highlights the potential of the diaziridinone reagent to react with many pallada(II)cycles bearing two carbon ligands, regardless of the hybridization of those carbon ligands (e.g. aryl, alkyl, alkenyl).

Scheme 22.

Palladium-catalyzed synthesis of carbazoles or indoles from 2-iodobiphenyls or 2-iodostyrenes and di-tert-butyldiaziridinone.

The same group later showed that the diaziridinone reagent can react with pallada(II)cycles generated from C(sp³)–H activation with a proximal alkyl group (Scheme 23a) [46]. Notably, 3,3-dimethylindoline structures are attractive scaffolds in the drug discovery process [47] and can easily be accessed by using this method. A study of the stoichiometric palladacycle showed that the phosphine ligand was necessary to promote the amination process, as the desired indoline product could not be obtained when 1,5-cyclooctadiene (COD) was used as the sole ligand (Scheme 23b). Additionally, a kinetic isotope effect of 7 was found in an intramolecular competition experiment (Scheme 23c) and a KIE of 2.1 was found in a parallel study, indicating that the C─H activation was likely involved in the rate-determining step of the transformation.

Scheme 23.

Synthesis of indolines via C(sp³)–H bond activation.

The same group also directly prepared indoles via a three-component reaction among aryl iodides, alkynes, and the diaziridinone (Scheme 24a) [48]. This transformation was proposed to be mechanistically similar to Shi’s prior work (Scheme 20) [44], with the alkyne fulfilling the role of the alkene. The versatility of this transformation was highlighted in the substrate scope study – a wide variety of electron-rich and -poor functional groups can be tolerated, along with several internal alkynes. Zhang later expanded this method to include tethered alkynes for the preparation of 3,4-fused tricyclic indoles (Scheme 24b) [49], which was also reported by Jiang, Yu, and co-workers shortly after [50]. Zhang further illustrated the utility of their intramolecular cyclization strategy by developing a short synthesis of the FDA-approved cancer drug, Rucaparib (Scheme 24c).

Scheme 24.

Three-component synthesis of indoles from aryl iodides.

In a departure from using the diaziridinone reagent for C─H bond amination, Luan and co-workers recently showed that hydroxylamines can be effective reagents for the preparation of fused tricyclic N-Ts indoles (Scheme 25a) [51]. Related to the work of Zhang (Scheme 24) [49], along with recent advances in ortho-amination in the palladium/norbornene cooperative catalysis (vide infra) [52], hydroxylamines were chosen as the coupling partner due to their ease of preparation, derivatization, and precedence in ortho-amination reactions. By carefully tuning the electronics of the benzoyloxy leaving group, the desired 3,4-fused tricyclic indoles could be obtained in yields as high as 97% (Scheme 25b). In general, more electron-deficient leaving groups performed better under the reaction conditions, which indicates that better leaving group ability may be a key component in this reaction. Notably, various ring sizes and positions can be tolerated in this transformation, albeit typically resulting in lower overall yields for larger fused rings.

Scheme 25.

Synthesis of fused tricyclic indoles using hydroxylamine electrophiles.

An oxidative addition pathway involving the hydroxylamine reagent was ruled out, since the more acidic N─H bond would likely be deprotonated under the reaction’s basic conditions. Instead, two pathways involving a concerted 1,2-migration or formation of a Pd(IV) nitrene species were proposed for the formation of the aryl C─N bond (Scheme 26).

Scheme 26.

Proposed mechanisms for C─N bond formation.

2.2.3. Indole and indoline synthesis via the palladium/norbornene cooperative catalysis

The palladium/norbornene (Pd/NBE) cooperative catalysis, also known as Catellani-type reactions, has been established as an effective tool for the vicinal difunctionalization of an aryl halide and its ortho-C─H bond [52-57]. Consequentially, it has used to prepare indoles and their derivatives, among other heterocycles. In this dual catalyst system, norbornene acts as a catalytic directing group, enabling C─H activation to occur at the arene ortho position. This results in the formation of a key intermediate known as the aryl-norbornyl palladacycle, typically referred to as the ANP intermediate, which can then react with an electrophile to functionalize the arene ortho position. Subsequently, the norbornylpalladium species can undergo β-carbon elimination to regenerate the norbornene co-catalyst; finally, the resulting arylpalladium species can participate in a typical cross-coupling reaction to close the catalytic cycle (Scheme 27).

Scheme 27.

Mechanism of a typical Pd/NBE-catalyzed aryl iodide difunctionalization.

The first example of an indole synthesis using this dual catalyst system was reported by Lautens in 2010, where 2H-azirines were used to prepare indoles in a single step from aryl iodides (Scheme 28a) [58]. It should be noted that, unlike the previously discussed multi-component heterocycle construction methods, no C─H bond amination takes place in this reaction, but rather a C─H carbonation. Mechanistically, this reaction was proposed to occur by oxidative addition of the ANP intermediate into the azirine C─N single bond, followed by C─C reductive elimination at the arene ortho position. After norbornene extrusion, the ipso position was aminated via C─N reductive elimination to first form a 3H-indole, which isomerized to the aromatic 1H-indole (Scheme 28b). Notably, this method can produce unprotected indoles in a single step from aryl iodides.

Scheme 28.

Synthesis of indoles from aryl iodides and 2H-azirines.

One drawback of using 2H-azirines, though, was their high reactivity under the reaction conditions – it was critical for them to be added slowly and under dilute conditions, as fast addition or a more concentrated reaction promoted the formation of undesired dihydroimidazoles. These dihydroimidazole side-products were proposed to be formed from a palladium-catalyzed formal [3+2] cycloaddition of the 3H-indole with another 2H-azirine, supporting the importance of diluted conditions to favor the aromatization (Scheme 29).

Scheme 29.

Conversion of the indole products into dihydroimidazoles.

Liang and co-workers later showed that the saturated derivatives of 2H-azirines, aziridines, could efficiently undergo the difunctionalization reaction to deliver indoline products (Scheme 30a) [59]. This transformation was proposed to proceed through a similar reaction mechanism as Lautens’ indole synthesis (Scheme 28b) [58]. The versatility of the transformation was demonstrated by forming a collection of 3-substituted and 2,3-disubstituted indoline products, although the selectivity between 2- and 3-substituted indoline isomers for the former was difficult to control if an aromatic ring was bound to the 2-position of the aziridine. Conversely, 2-alkyl aziridines dramatically favored formation of the 3-alkyl indoline products. Additionally, the use of an enantiopure chiral aziridine resulted in full transfer of chirality to the indoline products (Scheme 30b). Moreover, the inversion of configuration found with the 3-phenylindoline product isomer suggested that an S_N2-type oxidative addition with the ANP intermediate was operating, which is in agreement with a previously reported ortho-alkylation stereochemical investigation by Lautens [60].

Scheme 30.

Synthesis of indolines from aryl iodides and aziridines.

Pd/NBE cooperative catalysis has also been used to achieve indole synthesis though C─H bond amination. Liang and co-workers recently developed a three-component indole synthesis from aryl iodides, internal alkynes, and an N,N-dimethylhydroxylamine electrophile (Scheme 31a) [61]. In principle, this transformation could conceivably occur in the absence of norbornene, as in Zhang’s three-component indole synthesis [48]; however, no formation of the desired indole product was observed in the absence of the norbornene co-catalyst. Since it was not clear when the demethylation step occurred, density functional theory (DFT) calculations were employed to show that the skeleton of the indole was formed prior to the C─N bond cleavage, which is thermodynamically driven by aromatization of the indole heterocycle. The alternative pathway, termed “concerted metallization” by the authors, where the demethylation occurs prior to the cyclization, was found to have an incredibly high activation barrier of 75.4 kcal/mol. Conversely, demethylation of a quaternary indolium species after the cyclization was found to have a much lower activation barrier of 27.9 kcal/mol (Schemes 31b and c). While this method is effective for the preparation of linear N-alkyl indoles, namely N-methyl, ethyl, or propyl indoles, N-benzyl or -aryl indoles were found to be challenging products to access using this transformation.

Scheme 31.

Three-component synthesis of indoles via the Pd/NBE cooperative catalysis

3. Conclusions and outlook

In summary, recent strategies of constructing indoles and their close derivatives via Pd-catalyzed C─H bond activation are discussed in this short review article. Compared to conventional approaches, these palladium-catalyzed methods represent “greener” choices, as highly functionalized products could be accessed from simpler precursors under milder conditions. It is particularly attractive to directly use aryl halides or anilines as substrates, which greatly streamlined the synthesis of these heterocycles. Among the reported examples, the directed orthopalladation has been a powerful strategy for preparing indoles either intramolecularly or through multi-component couplings. In addition, the 1,4-palladium migration provides unique opportunities to transfer a reactive site to its adjacent position. Moreover, the palladium/norbornene cooperative catalysis offers a versatile platform to construct not only indoles and indolines, but also other heterocycles, due to the wide-range of ipso-terminating reagents and ortho-coupling partners that have been developed to date [52].

As an outlook, the future development of indole, indoline, and carbazole syntheses would benefit from the use of more readily available reagents, lower catalyst loading, broader substrate scope, and milder reaction conditions. We anticipate that new C─H bond activation strategies will continue being developed for preparing indoles and other biologically important heterocycles, and eventually find their utilities in preparing pharmaceuticals. It is our hope that direct and regioselective double/vicinal C─H bond functionalization on the arene could be realized one day for multi-component indole synthesis, which would further reduce the complexity of the precursor molecules.

Biographies

Alexander Rago obtained his BS in Chemistry from the University of Illinois at Urbana-Champaign, where he conducted research on the hydroamination of alkenes and the use of chloroform as a carbon monoxide precursor in cross-coupling reactions under Prof. Kami L. Hull. Since 2016, he has been pursuing his PhD in Organic Chemistry under the direction of Prof. Guangbin Dong at the University of Chicago, focusing on the development of new transformations of haloarene ortho-C─H bonds via palladium/norbornene cooperative catalysis.

Guangbin Dong received his BS degree from Peking University and completed his PhD degree in chemistry at Stanford University with Professor Barry M. Trost, where he was a Larry Yung Stanford Graduate fellow. In 2009, he began his postdoctoral research with Professor Robert H. Grubbs at California Institute of Technology, as a Camille and Henry Dreyfus Environmental Chemistry Fellow. In 2011, he joined the department of chemistry and biochemistry at the University of Texas at Austin as an assistant professor and a CPRIT Scholar. Since 2016, he has been a Professor of Chemistry at the University of Chicago.