General Synthesis of N-Alkylindoles from N,N-Dialkylanilines via [4+1] Annulative Double C–H Functionalization

Abstract

Strategies enabling the construction of indoles and novel polycyclic heterocycles from simple building blocks streamline syntheses in synthetic and medicinal chemistry. Herein, we report a C–H functionalization approach to N-alkylindoles proceeding via a double, site-selective C(sp³)–H/C(sp²)–H [4+1] annulation of readily accessed N,N-dialkylanilines. This protocol features a site-selective hydrogen atom transfer by a tuned N-tBu amidyl radical, and addition of a sulfonyl diazo coupling partner which promotes highly site-selective homolytic aromatic substitution of the (hetero)aromatic core. Mild decarboxylation of the annulation product enables the overall introduction of a carbyne equivalent into the N,N-dialkylaniline scaffold. Furthermore, the site-selectivity and mild conditions of the indolization facilitate direct access to N-alkyl indole scaffolds in late-stage functionalization settings.

Graphical Abstract

Indoles are among the most widely represented heterocycles in natural products and small molecule drugs.¹ In medicinal chemistry, structure activity relationship (SAR) studies of indole-containing lead compounds can require syntheses of matched molecular pairs wherein the indole is differentiated. Rapidly accessing these matched pairs accelerates the development of SAR and the identification of possible drug molecules.² An array of complementary synthetic approaches to N-alkylindoles involving either polar, radical, or organometallic reactivity have been developed to address this need.³ Despite the utility of these methods, there remains a need for general parallel indole syntheses which utilize readily available starting materials, display broad functional group compatibility, and encompass a range of substitution patterns.

The conversion of structurally diverse N,N-dialkylanilines—originating from abundant (hetero)aryl halides and secondary amines—to N-alkylindoles would facilitate rapid SAR exploration for medicinal chemistry. A representative example where such a transformation would prove valuable is illustrated in Figure 1a. During the development of LXR agonists, a diverse set of N,N-dialkylanilines were prepared and evaluated for potency.⁴ For comparison purposes, a multistep de novo synthesis of N-alkylindole analogs was then performed which led to structures with increased potency. The ability to directly transform the N,N-dialkylanilines to N-alkylindoles would significantly facilitate the generation of additional match pairs for further SAR studies.

Figure 1.

Transforming N,N-dialkylanilines to N-alkylindoles via site-selective, [4+1] annulative double C–H functionalization.

An appealing transformation to achieve this goal is a single-atom [4+1] annulation via double C–H functionalization (Figure 1b). Such an approach would use a (hetero)aromatic starting material to furnish diverse N-alkylindole frameworks. Very recently, initial efforts targeting this transformation have been reported.⁵ The requirement of substrate oxidation significantly limits the scope of these protocols however, leading to poor regiocontrol with non-symmetrical substrates and limited functional group tolerance unsuitable for late-stage applications. The regiocontrol of the two separate C–H functionalizations is a major challenge; both the hydrogen atom transfer (HAT) of the N-alkyl groups and the homolytic aromatic substitution (HAS) of the (hetero)aniline must be controlled. Furthermore, there are no examples introducing a carbyne equivalent⁶ in a single-atom annulation, allowing access to C3-unsubstituted indoles (Figure 1c).

We hypothesized that a radical-chain approach to the indolization involving a site-selective HAT using a bulky tBu amidyl radical could address these challenges.⁷ While previously not utilized in radical reactions, we envisioned sulfonyl diazo reagents as modular sources of one-carbon equivalents⁸ in the indolization which could also facilitate a regioselective HAS step. Herein, we report the successful development of such a transformation via site-selective, double C–H functionalization. This method extends to a broad range of (hetero)aromatic substrates and sulfonyl diazo partners and includes both the preparation of unique N-alkylindoles and the late-stage double C–H functionalization of complex molecules.

Our studies commenced with the indolization of a diverse set of N,N–dialkylanilines using O–alkenylhydroxamate 1 and several sulfonyl diazo coupling partners (Figure 2). The radical-chain C–H functionalization is conveniently initiated upon mild warming in dichloroethane (DCE). N,N-Dimethylanilines are excellent substrates, providing N-methyl indoles in good yields using several different sulfonyl diazo reagents (2–8). Notably, the direct indolization of these readily available substrates has not been previously reported.⁹ Annulation of an N,N-diethylaniline likewise provides indole 9, albeit in moderate yield.

Figure 2.

Indolizations of N,N-dialkylanilines. Reactions performed at either 45 °C or 60 °C, see Supporting Information for details. All yields are of isolated product. ^a1.2 equiv 1 and sulfonyl diazo partner used. ^b3 equiv 1 used.

We viewed the capability of performing site-selective C–H functionalizations of both the N-alkyl groups and the (hetero)aromatic core during the annulation as critical to the overall utility of the process. We therefore evaluated several N,N-dialkylanilines containing two inequivalent N-alkyl groups to assess the ability of our system to select between sterically and/or electronically differentiated C(sp³)–H sites in the HAT. These studies demonstrate high selectivity for functionalization of the N-methyl group over other N-alkyl groups (10–14). Notably, the site-selective annulation of N-benzyl-N-methylaniline (13) enables access to (NH) indoles via subsequent deprotection. The reaction is highly functional group compatible, as highlighted by the reaction delivering indole 14 in the presence of an unprotected alcohol (as well as results from Figure 3, vide infra).

Figure 3.

Indolizations furnishing tricyclic molecules and late-stage [4+1] annulation. See Fig. 2 for reaction scheme. Reactions performed at either 45 °C or 60 °C with 1.1–3.5 equiv 1 and 1.1–2 equiv sulfonyl diazo partner, see Supporting Information for details. All yields are of isolated product. ^a3.5 equiv 1 used. ^b3 equiv 1 used.

We also evaluated the site-selectivity of the HAS step on non-symmetrical (hetero)aromatic substrates (15–18). The annulation of 3-bromo-N,N-dimethylaniline delivers the 6-bromo indole with excellent regioselectivity (15), while that of 3-methoxy-N,N-dimethylaniline favors the 6-methoxy indoles (16, 5.9:1 rr and 17, 6.3:1 rr) with minor 4-methoxy regioisomers. Annulation of a substituted pyridine is also efficient, providing azaindole 18 with high site selectivity on both the HAT and HAS functionalization of the pyridine core.

We next sought to apply the double C–H functionalization to the synthesis of a wide variety of [1,2]-annulated indoles, which form the core of diverse natural products and drug compounds (Figure 3).¹⁰ Reactions involving piperidine (19–23), morpholine (24–28), and piperazine (29–33) heterocycles afford indoles in good yield using a range of sulfonyl diazo partners, including ((1-diazo-2,2,2-trifluoroethyl)sulfonyl)benzene which directly provides valuable 3-trifluoromethyl indoles. Indolization involving the five-membered ring of 1-(4-bromophenyl)pyrrolidine yields fused indole 34, albeit in moderate yield.

Site-selective annulation delivering tricyclic indoles from non-symmetrical substrates is also efficient with respect to both the HAT and HAS steps. For example, the reaction of 2-methyl-1-phenylpiperidine occurs selectively at the unsubstituted methylene site to deliver 35, although a substrate with 3-amido substitution affords modest levels of regiocontrol (36). Transformations involving substrates with meta substituents with respect to the aniline nitrogen atom uniformly deliver tricyclic products with regioselective annulation occurring para to the ring substituent, regardless of whether the substituent is electron-donating or electron-withdrawing (37–40, 42–43). Substrates with disubstitution on the aromatic ring also react with high selectivity (44–47).

N,N-Dialkyl(hetero)anilines are a common structural motif in biologically active small molecules—out of the 89 small molecule drugs with sales over $1B in 2022, 15 contain this structure—positioning the indolization as a unique transformation for late-stage functionalization.¹¹ For example, complex drug compounds ulipristal acetate and mifepristone are both transformed into indole derivatives using our approach (48–49). As a final test involving the derivatization of a complex, functional group rich heterocycle, we applied the annulation to the late-stage double C–H functionalization of antineoplastic agent ribociclib. In the event, the indolization proceeds with high regioselectivity on the central pyridine ring to deliver pyrrolopyridine 50 in good yield.

To further enhance the generality of the [4+1] process, we targeted a single-atom annulation involving a carbyne equivalent via introduction of a removable ester group in the indolization (Figure 4). The modularity of the sulfonyl diazo partner is a useful aspect of our approach, and we selected the tBu ester as ideal for mild deprotection following the annulation. This two-step approach to C3-unsubstituted indoles involves reaction with tBu 2-diazo-2-(phenylsulfonyl)acetate, followed by mild TFA-mediated removal of the tBu ester in HFIP. Notably, this transformation is efficient with both cyclic and acyclic N-alkyl groups regardless of the electron-richness of the aromatic substrate.

Figure 4.

Single-atom annulations involving the introduction of a carbyne equivalent. All yields are of isolated product over two steps.

A mechanistic outline for the indolization is depicted in Scheme 1. We propose that site-selective HAT by the hindered tBu amidyl radical is followed by a rare example of the addition of a carbon-centered radical to a diazo trap.¹² Following stepwise loss of N₂, the electron-poor carbon-centered radical (see Supporting Information for further details) adds to the (hetero)aromatic ring. The high regioselectivity of this addition is likely due to the steric demand of the two electron-withdrawing substituents on the carbon-centered radical. At this stage, rearomatization and loss of phenylsulfonyl radical delivers the product and addition to reagent 1 propagates the chain process. While the precise mechanistic details of the rearomatization and loss of the phenylsulfonyl radical remain to be determined, we have found that reactions using less than two equivalents of both reagent 1 and the sulfonyl diazo trap lead to decreased yield (see Supporting Information), and it is possible that either an additional HAT or oxidation step could be facilitated by either reagent. We also considered alternative pathways via carbenoid intermediates, but calculations suggest these involve prohibitively high energies (see Supporting Information for details).

Scheme 1.

Proposed radical chain mechanism for the [4+1] annulation.

In conclusion, we have developed a [4+1] indolization of N,N-dialkylanilines proceeding via site-selective, double C–H functionalization. This method directly accesses a diverse array of N-alkylindoles and polycyclic heterocycles from simple starting materials that would otherwise present a synthetic challenge. The notable site-selectivities of both the C(sp³)–H and C(sp²)–H functionalizations are a result of HAT via a tuned amidyl radical and the use of a sulfonyl diazo reagent as a modular coupling partner. A formal single-atom annulation via the two-step introduction of a carbyne equivalent was also developed via double C–H functionalization and decarboxylation. The mild conditions are amenable to common functionality and applications in the late stage functionalization of complex molecules. We anticipate that this indolization will serve as a valuable complement to alternative established methods to this synthetically and medicinally valuable heterocycle.