Introduction

Despite advances in transition metal-catalyzed and photocatalyzed cross couplings, the site-selective functionalization of N-heterocyclic scaffolds remains a significant challenge in the diversification of this privileged architectural motif1,2,3,4,5,6. One of the most prominent intermediary scaffolds in alkaloid construction is the 2,3-disubstituted indole7,8,9. With the appropriate functional groups in place at the 2-, and 3-positions of the starting indole, this advanced intermediate can serve as a point of synthetic divergency that affords access to a fairly diverse and extensive number of indole alkaloid natural products of biological and therapeutic relevance (Fig. 1d)10,11,12,13,14,15,16,17. The ability to site-selectively incorporate specific coupling partners across a wide array of substrates has improved substantially with the advent of directed C–H activations, particularly as it pertains to the substitution of 3-carboxy indoles at the 2- and 4-positions18,19,20,21,22. In the context of target-directed syntheses, the effectiveness of this strategy is often antithetical to the desire for increased architectural diversity, and limited by the need to preinstall a reactive, Lewis basic directing group that requires either immediate derivatization or a non-trivial excision to avoid compatibility issues over subsequent transformations23,24,25,26. Recently, the in situ organocatalyzed derivatization of a non-directing functionality to one capable of satisfying the stereoelectronic requirements to direct metallation, typically in combination with a transition metal catalyst, has led to new avenues in directed C–H functionalizations (Fig. 1a)22,27,28,29,30. An elegant extension of this design concept was reported by Dong in which the use of norborandiene acted as a transient directing group by first ligation to the transitional metal catalyst followed by incorporation into the substrate, which upon C–C or C–X bond formation is then released through a carbometallation/β-hydride elimination sequence31,32.

Fig. 1: Strategies in C–H functionalization of (hetero)aromatic compounds.
figure 1

a Transient directing groups in transition metal-catalyzed C–H activations. b Pre-installed directing groups in assorted metallocarbene C–H insertions. c The ester ‘dance’ reaction developed by Yamaguchi. d Selected natural products bearing the 2,3-disubstituted indole scaffold; e This work: a transition metal-catalyzed sequential  indole carboxamide translocation/C3-functionalization with metallocarbenes.

Full size image

To complement these existing strategies, we sought to develop an approach wherein site selectivity is controlled through the use of a stable directing group capable of undergoing translocation, rather than outright extraction, resulting in a catalyst-controlled reorganization of the substrate. We surmised that metallation at the site of directing group attachment, followed by an intramolecular translocation of the directing group would allow access to a new site of functionalization (Fig. 1b). In this fashion, two competing pathways of vicinal C–H functionalization or 1,2-migration of the directing group and subsequent cross coupling in a net C–C activation process is controlled exclusively through the inherent reactivity of the catalyst chosen. The translocation of a directing group followed by ipso-functionalization represents a compelling strategy for the site-selective functionalization of N-heterocyclic motifs. While the metal-catalyzed 1,2-migration of aryl acyl groups constitutes an alternative approach to aryl functionalizations, combining an acyl translocation with an intermolecular C–C bond formation is relatively unexplored. Of the few examples that exist, PdII has proven most effective at promoting the translocation and cross-coupling33. For example, Saracoglu and coworkers reported a rather striking dependency for the 1,2-acetyl migration/C4-arylation of indoles bearing a free N–H with aryl iodides, while N-benzylated indole substrates suppressed acyl translocation34. Likewise, Li disclosed a Pd-catalyzed indole C4-arylation and a complementary C4-arylation/3,2-carbonyl migration cascade that was achieved by tuning the catalyst through the addition of a Brønsted acid35. Additionally, Miura reported a RhII-catalyzed ring-expansion of 2-triazolyl-1-indanones facilitated by a 1,2-acyl migration via an intermediate α-imino Rh-carbenoid36. In arguably the most impressive example to date of a directing group translocation/C–H functionalization cascade, Yamaguchi et al. reported an “ester dance” of aryl benzoates in the presence of Pd(OAc)2 and a bisphosphine ligand, which proved effective for the decarboxylative arylation and amination of heteroaromatics and benzene derivatives37,38,39.

Inspired by this seminal work, we focused on the development of a unified strategy for indole functionalization from a readily accessible 3-acyl indole derivative wherein competitive C–H site selectivity is controlled exclusively by the choice of catalyst. Anticipating that electrophilicity of the transition metal to be a defining parameter to achieving the desired disparate reactivities, we chose diazo compounds as the cross-coupling partner wherein interception of an electrophilic metallocarbene would influence product distribution40,41,42,43. Herein, we report the catalyst-controlled site selective C–H functionalization of 3-carboxamide indoles 1 with diazo compounds 2 to provide C2-alkylated indole 3 in the presence of IrIII, while employing a RhI complex leads carboxamide translocation and concomitant C3-functionalization to give 2,3-disubstituted indole 4 (Fig. 1e). This study addresses limitations associated with architectural confinements to specific aryl motifs that hinders broader implementation of many modern directed C–H functionalization methods.

Results and discussion

We began our study in search of a robust, readily accessible functional group that embodied three important characteristics. These included (1) the capablity of directing a metal-catalyzed C–H activation event, (2) susceptibility to C–C bond insertion by a catalyst at the point of attachment to the core molecular framework, and (3) would remain within the ligation inner sphere of the metal enabling reincorporation at a site adjacent to the original point of attachement. Although a wide array of carbonyl derivatives have proven effective at directing C–H functionalizations, transition metal-catalyzed Csp2-C(acyl) insertions are arguably most well-developed employing aryl aldehyde and carboxylic acid substrates29,34,44,45,46. Aside from the inherent carbonyl electrophilicity of aldehydes and the Brønsted acidity of carboxylic acids, the propensity of these groups to dissociate as CO and CO2, respectively, complicates efforts to provide products derived from their translocated reinstallation. In contrast, few acyl derivatives are as thermally stable and resistant to nucleophilic or electrophilic additions as N-alkyl amides, and the electron donating capability of nitrogen should facilitate either a concerted 1,2-acyl shift or isocyanate formation resulting from C–C insertion followed by substrate recapitulation24. However, we were acutely aware that these factors could likewise lead to catalyst inhibition resulting from amide ligation. To circumvent this potential issue, we chose to examine the functionalization of tert-butyl amide indole 1a to preemptively diminish interference or catalyst sequestration by the Lewis basic nitrogen through steric hinderance42. Our initial foray with diazooxindole 2a sought to assess competitive indole C2–H activation versus a C–C bond activation and amide translocation of the amide directing group with concomitant C3 functionalization (Fig. 2). Treatment of 1a and 2a with [IrCp*Cl2]2 (2 mol%), AgNTf2 (8 mol%), and AgOAc (4 mol%) at 90 °C in Cl(CH2)2Cl provided exclusively indole 3a resulting from directed C2–H functionalization in 91% yield. While consistent with previous reports of C3-carbonyl derivatives directing 2-indole functionalization, the efficient use of diazo compounds as coupling partners is exceedingly rare in this circumstance. To the best of our knowledge, this represents the first example of constructing the indole-oxindole biaryl motif through an indole C2–H cross coupling with diazooxindole. In an effort to promote a 1,2-acyl migration/C3-functionalization pathway, we surmised that a sterically accessible metal center bearing electron withdrawing ligands would lead to stronger amide ligation, thereby weakening the C–C bond to the indole core. Thus, treatment of 1a and 2a with [Ir(COD)Cl]2 (2 mol%) in place of [IrCp*Cl2]2 while maintaining the relative stoichiometry of AgI salts, provided the translocation/C3-alkylated indole 4a, albeit in a mere 16% yield. Notably, the C2-alkylated indole 3a was not observed under these conditions, with recovered 1a and dimerized 2a accounting for the recovered mass balance, and the constitution of 4 was later confirmed by X-ray crystallography.

Fig. 2
figure 2

Initial experimental findings. Employing IrIII or IrI complexes led to disparate 2,3-disubstituted indole C–H functionalization products.

Full size image

Intrigued by these initial results, we first sought to improve the translocation/C3-functionalization pathway en route to cross-coupled product 4a by surveying a series of reaction parameters that included catalyst structure, AgI salts, solvent, and temperature (Fig. 3). The dimeric complex [Ir(COE)Cl]2 bearing relatively labile monodentate π-ligands resulted in a comparable yield of 4a, while [Ru(p-cymene)Cl2]2 provided only recovered indole 1a and the dimer of 2a (entries 1 and 2). The use of NiCl2•glyme led to a lower yield of 4a under slightly more mild thermal conditions, while [Rh(COD)Cl]2 proved comparable to its IrI counterpart (entries 2 and 4). However, lowering the reaction temperature to 50 °C and room temperature in the presence of [Rh(COD)Cl]2 led to substantially improved yields of 58% and 66% for 4a, respectively (entries 5 and 6). Performing the reaction in CH2Cl2 gave exclusively 4a in a satisfactory 73% yield, while non-chlorinated solvents PhMe and THF failed to provide any detectable indole functionalized products (entries 7 and 8). Interestingly, while the omission of AgNTf2 from the reaction mixture failed to provide indole 4a, the removal of AgOAc led to 53% yield of the anticipated adduct highlighting distinctly different roles for each AgI salt (entries 9 and 10). Curiously, conducting the reaction in the absence of [Rh(COD)Cl]2 and AgOAc while in the presence of AgNTf2 (12 mol%) led to ~5% of 4a, indicating that a mechanistic pathway involving Lewis acid activation of indole 1a may exist (entry 11). However, the formation of 4a was not observed when either AgNTf2 or both AgI salts were omitted from the reaction mixture in Cl(CH2)2Cl (entries 12 and 13). With an optimal set of reaction conditions for the formation of 4a, and insight into the mechanism for the formation of 4a in hand, we turned our attention toward evaluating the functional group tolerance and reaction efficiency for the formation of the IrIII-catalyzed, direct C2-functionalization and RhI-catalyzed amide translocation/C3-coupling.

Fig. 3: Identification of optimal reaction conditions for the formation of 4a[a].
figure 3

[a] Conditions: slow addition of diazooxindole 2a (0.2 mmol) via syringe pump in the indicated solvent (0.1 M) over 5 h to indole carboxamide 1a (0.1 mmol), MLn (2 mol%), AgNTf2 (8 mol%), and AgOAc (4 mol%) in the indicated solvent (0.1 M). See the Supplmentary Information for detailed experimental procedures and a complete list of experiments conducted. [c] Isolated yield of 4a. [d] Time = 48 h.

Full size image

With optimized conditions for the RhI-catalyzed translocation/C–H functionalization and IrIII-catalyzed indole C2–H activation in hand, we turned our attention toward assessing the architectural diversity and functional group tolerance of diazo compounds 2 in the cross coupling with N-methyl 3-carboxamide indole 1a (Fig. 4). Beginning with a series of diazooxindoles, we examined the impact of oxindole N-substitution on the formation of amide translocation/C3-functionalization adducts 4 and direct C2-functionalization cross-coupled products 3. Our decision to begin our study with this scaffold assembly was influenced by the prominence of indole and oxindole frameworks in biologically relevant natural products47. While diazooxindoles are formally classified as donor-acceptor diazo compounds, they tend to exhibit a reactivity profile unlike other metallocarbene precursors bearing a donor-acceptor 1,1-disubstitution pattern48,49. While the use of N-Me and N-Ac groups provided indole-oxindole adducts 4/3b and 4/3c in good to excellent yields, the corresponding N-Boc substituted diazooxindole provided 4d in 80% yield, but underwent deacylation under the IrIII-catalyzed C–H activation conditions to yield 3a in 92% (Fig. 4a). Electron donating methyl and methoxy substitution at the C5-position of diazooxindole did not hinder the formation of indole-oxindole cross coupled products 4/3e and 4/3f (Fig. 4b). In general, the presence of inductively withdrawing halogens at the C4-, C5-, and C7-positions of diazooxindole led to formation of adducts 4g-k in 43–64% yields under the RhI-catalyzed translocation/functionalization conditions, and comparable yields of indoles 3g-i were observed in the presence of IrIII/AgI (Fig. 4c). However, the direct C2-coupled products 3j and 3k were obtained in substantially better yields than their constitutional isomers 4j and 4k, while the presence of a CF3 group at the diazooxindole C6-position led to formation of both 4l and 3l in 72% and 90% yield, respectively. Although we were unable to identify side products resulting from a homo-diazo coupling at the site of halogen attachment, it is conceivable that the relatively low yields observed employing diazooxindoles derived from halogenated oxindole precursors may result from competitive oxidative addition pathways into the Csp2–X bond. In general, we observed diminished levels of conversion in those experiments where moderate to low yields of the products were obtained throughout our investigation, leading to ~95% yields based on recovered starting indole 1.

Fig. 4: Structural Variability of the Diazooxindole Component[a].
figure 4

[a] Conditions: A solution of 2 (0.2 mmol) in the indicated solvent (0.1 M) was added via syringe pump over 5 h to 1a (0.1 mmol), MLn (2 mol%), AgNTf2 (8 mol%), and AgOAc (4 mol%) in the indicated solvent (0.1 M) at the indicated temperature. See the Supplementary Information for detailed experimental procedures. [b] Structure validated via X-ray crystallography. [c] Diazo compound 2 recovered as a dimer.

Full size image

Employing phenyl diazoacetate, arguably the most widely studied donor-acceptor diazo compound, provided the anticipated translocation/C–H functionalization product 4m in near quantitative yield under our RhI-catalyzed conditions (Fig. 4c)50. However, the expected C2-functionalized product 3m was not observed under our IrIII-catalyzed conditions, leading exclusively to products resulting from dimerization of the diazo compound. In contrast, acceptor-acceptor diazo substrates performed equally well in the presence of RhI/AgI and IrIII/AgI (Fig. 4e). Coupling of indole 1a with the diazo compound derived from dimethyl malonate afforded adducts 4n and 3n in 86% and 73% yield, respectively. Although the translocated adduct 4o and direct C2-functionalized product 3o were obtained in 99% yield from the benzyl ester-derived diazo compound, the corresponding tbutyl ester substrate failed to provide 3p under the IrIII-catalyzed coupling conditions. However, the translocation/C3-functionalization product 4p was obtained in 99% yield in the presence of RhI/AgI. Attempts to incorporate a methyl substituent at C3 employing either CH2N2 or TMSCHN2 resulted in complete recovery of the starting indole 1 that was most likely due to the rapid dimerizations of the diazo compound.

We next evaluated the architectural diversity within indole-3-carboxamide 1 beginning with an assessment of reaction efficiency related to a variety of synthetically versatile indole N-substituents (Fig. 5a). Indole substrates bearing N-benzyl, acyl, and sulfonyl groups provided the corresponding translocation/C3-functionalized products 4q-s in good to excellent yields in the presence of RhI/AgI. Even indole bearing a free N–H afforded adduct 4t in 43% yield. However, while the direct, C2-functionalized N-benzyl indole 3q was obtained in 64% yield, the IrIII/AgI-catalyzed C2-functionalization conditions proved incompatible with the electron withdrawing N-substitution and free N–H starting 3-carboxamide indoles. An indole substrate bearing a 4-methyl group gave the translocation product 4u in slightly diminished yield, presumably due to unfavorable steric encumbrance at C3, while the C2-functionalized adduct 3u was obtained in 64% yield (Fig. 5b). A similar trend between our RhI and IrIII-catalyzed conditions was observed with the 5-methoxy substituted indole products 4/3v. The inclusion of an electron withdrawing Br or Cl at the indole C5- and C6-positions resulted in the formation of adducts 4/3w and 4/3y, while a 4-chloro indole failed to give translocation product 4x and led to the direct C2-functionalized product in modest yield (Fig. 5c). The presence of a methyl ester at the indole C6-position gave 4z in good yield and 3z in a modest 47%. To assess the impact of an additional Lewis basic nitrogen, we evaluated C7-azaindole under both sets of conditions, but while the IrIII-catalyzed C2-functionalized product 3aa was obtained in 94%, our RhI/AgIconditions provided the corresponding translocation adduct 4aa in a mere 11% yield suggesting that the Lewis basicity of the pyridine motif hindered indole metallation. In general, recovered starting material constituted the mass balance for those substrates which provided moderate to low yields of products.

Fig. 5: Substrate scope: indole-3-carboxamide[a].
figure 5

[a] Conditions: A solution of 2a (0.2 mmol) in the indicated solvent (0.1 M) was added via syringe pump over 5 h to 1 (0.1 mmol), MLn (2 mol%), AgNTf2 (8 mol%), and AgOAc (4 mol%) in the indicated solvent (0.1 M) at the indicated temperature. See Supplementary Information for detailed reaction conditions.

Full size image

Our next objective was evaluating the efficiency with which various acyl derivatives at the indole 3-position would undergo translocation/C3-functionalization in the presence of [Rh(COD)Cl]2 and AgNTf2/AgOAc. Although the anticipated 2,3-disubstiuted indoles 4ab-af were obtained in good to excellent yields from the corresponding secondary and tertiary amides, the acyl morpholine proved significantly less reactive, affording indole 4ag in 12% yield and 98% based on recovered starting material (Fig. 6a)51,52. Not surprisingly, the 3-acyl indoles derived bearing an aldehyde, ester, ketone, and trifluoromethyl ketone failed to give the translocation products. An assessment of scalability was conducted employing 1 gram of N-benzyl-3-indole carboxamide 1b and diazo malonate 2b under our standard conditions, which provided the 2,3-substituted in dole 4ah in an isolated yield of 67% (Fig. 6b).

Fig. 6: Versatility of the RhI-catalyzed translocation/C–H functionalization of 3-carboxamide indoles.
figure 6

a Acyl groups prone to translocation. b Gram-scale accessibility of the cross-coupling adducts.

Full size image

In an effort to interrogate the mechanism by which this unusual carboxamide translocation/C–H functionalization process occurs under our RhI/AgI-catalyzed conditions, we focused on the fate of the C–H upon conversion of 3-carboxamide indole 1 to the 2,3-disubstituted indole product 4, and process by which the acyl 1,2-migration event occurs. To assess the extent to which the C2–H in indole 1a is transferred to the benzylic position in 4m, we conducted the RhI/AgI-catalyzed translocation/C–H functionalization on C2-D indole d-1a (94% D-incorporation) with diazo ester 2c and observed the formation of 4m in 58% yield with 20% D-incorporation in the crude reaction mixture, which was reduced to 10% following purification by flash chromatography (Fig. 7a). Speculating that the low D-incorporation was due in part to the relative acidity of the newly formed bisbenzylic, α-ester proton, we subjected the isolated indole 4m (10% D-incorporation) to 10 equiv of D2O in THF at rt for 24 h and observed no deuterium enrichment. However, treatment of indole 4ac, derived from the corresponding diazo malonate, with D2O resulted in 66% D-incorporation, presumably due to the enhanced acidity of the α-proton. We then conducted a series of crossover experiments featuring N-methyl indole 1a bearing a tbutyl amide at C3 and N-benzyl indole 1c with a 3-ipropyl amide (Fig. 7b). Exposure of a 1:1 mixture of 1a and 1c to excess diazo malonate 2b under our standard conditions provided 2,3-disubstituted indoles 4n and 4ae in 78% and 64% yield, respectively, with no detectable formation of indoles 4ai and 4aj resulting from crossover of the 3-carboxamide groups. This result suggests amide migration is occurring through either a concerted 1,2-acyl shift or if a stepwise process involving ionization to a Rh-bound isocyanate is operable, then reincorporation of the isocyanate at the indole C2-position is faster than dissociation. Attempts to promote the acyl migration under our RhI-catalyzed conditions while leading to H-incorporation at the indole C3-position by replacement of the diazo component with a proton source such as MeOH, p-TsOH, or TFA led to complete recovery of the starting indole (≥95%). Furthermore, a preliminary evaluation of rendering the Rh-catalyzed translocation/C3-functionalization of indole 1a and diazo ester 2b enantioselective employing phosphoramidite ligand L provided indole 4m in a promising 45% ee (Fig. 7c)53,54.

Fig. 7: Mechanistic studies, enantioselective variant, proposed mechanism, and further derivatizations of translocated product.
figure 7

a Deuterium labeling experiment and deuterium exchange of migrated product 4m and 4ac. b Cross-over experiment with two amides that have demonstrated individual reactivity under standard [Rh] condition. c Application of chiral phosphroamidate ligand in migration reaction. d Proposed mechanistic cycle of rhodium catalyzed carboxamide translocation reaction. e Derivatization of translocated product 4ah towards strategic intermediate in syntheses of indole alkaloids.

Full size image

Based on our composite results, with consideration of established Rh-catalyzed indole C–H functionalization methods, a possible mechanism for the formation of 2,3-disubstituted indoles 4m begins with the nucleophilic addition of indole 1a to a cationic RhI complex resulting in metalated indole 5 (Fig. 7d). Amide translocation via a 1,2-acyl migration leads to carbenoid 6 that upon rearomatization gives intermediate 7. Decomposition of diazo 2c gives metallocarbene 8 followed by migratory insertion to construct the crucial C–C bond. Our deuterium labeling studies highlighting the low level of deuterium transfer from the C2 position would indicate a protodemetallation step to furnish the translocated/C–H functionlized adduct 4m and regenerate RhI–X, which upon exposure to a AgI salt provides the active cationic RhI complex.

To highlight the synthetic utility of the RhI/AgI-catalyzed translocation/C–H functionalization strategy toward the construction of high value 2,3-disubstituted indoles, we targeted 2-carboxaldehyde indole 11 by advancing indole 4ah as an expedient gateway toward complex indole alkaloids (Fig. 7e). Although a bearing a relatively modest collection of functional groups, aldehyde 11 has served as a crucial intermediate in several indole alkaloid total syntheses, including Macmillan’s inspirational syntheses of strychnine, aspidospermidine, kopsinine and 3 other indole alkaloids55. Additionally, conventional approaches toward this key synthetic building block typically require either construction of indole scaffold from non-indole containing substrates, or beginning frequently with halogenated indole core56,57,58. A Krapcho decarboxylation of the substituted malonate side chain proceeded in 95% yield, which was followed by dehydration of the tbutyl amide to afford nitrile 10 in 99%59. Reduction of the ester in the presence of DIBAL-H proceeded with concomitant partial reduction of the nitrile group to furnished aldehyde 11 in excellent yield over a relatively straightforward three step sequence60.

In conclusion, we developed complementary RhI/AgI- and IrIII/AgI-catalyzed approaches toward 1,2-disubstituted indoles proceeding through either a 1,2-acyl translocation/C–H functionalization or directed C2–H functionalization, respectively, that enables the site selective incorporation of diazo substrates into 3-carboxamide indoles. Highly functionalized N-heterocycles are pervasive architectural features across numerous facets of academic and industrial chemical research. Thus, new methods for the selective diversification of a central lynchpin intermediate will broadly impact synthetic efforts in the target directed synthesis of N-heterocycles. However, that significance increases if the method achieves these aims while using readily available feedstock precursors with a high degree of chemoselectivity in an operationally simple and cost-effective fashion. Using the biologically active indole alkaloids as motivating templates for design, this study demonstrates how a directing group translocation/C–H functionalization event is controlled exclusively through the choice of catalyst system, rather than complementary preinstallation of directing functionality. Both methods demonstrated excellent functional group compatibility across a diverse assortment of indole arene and N-substitution, and various classes of substituted diazo compounds. A series of experiments designed to probe key mechanistic details revealed insight into the indole metalation and acyl translocation events. Likewise, the synthetic utility of the products obtained was demonstrated by the assembly of a functionalized indole that serves as a lynchpin building block en route to an array of indole alkaloid natural products. Future studies directed at providing a detailed mechanistic understanding of the mechanistic divergency between the RhI/AgI and IrIII/AgI systems, offering asymmetric modifications, and exploring new, structurally valuable coupling partners are ongoing and will be reported in due course.

Methods

General procedures for the Ir-catalyzed C2–H functionalization of indole-3-carboxamides 1

An oven dried 10 mL round bottom flask was charged with 1 (0.1 mmol, 1.0 equiv.), [IrCp*Cl2]2 (1.6 mg, 0.002 mmol, 2 mol%), AgNTf2 (3.1 mg, 0.008 mmol, 8 mol%), and AgOAc (0.6 mg, 0.004 mmol, 4 mol%). A reflux condenser was attached, and the completed apparatus purged and backfilled with N2 (3x). Freshly distilled 1,2-dichloroethane (1.0 mL, 0.1 M) was added, and the resulting solution was heated to 90 °C. A solution of 2 (0.2 mmol, 2.0 equiv) in 1,2-dichloroethane (2.0 mL, 0.1 M) was added slowly over 5 h employing a syringe pump, and stirring continued for 24 h. The mixture was allowed to cool to room temperature by removal of the heating bath, then filtered through a short plug of silica gel eluting with 1:1 ethyl acetate/hexane (10 mL) and concentrated under reduced pressure. The crude mixture was purified by flash column chromatography eluting with hexane/EtOAc to afford 3.

General procedures for the Rh-catalyzed C-3 translocation/functionalization of 1

An oven dried 10 mL round bottom flask was charged with 1 (0.1 mmol, 1.0 equiv.), [Rh(COD)Cl]2 (1.0 mg, 0.002 mmol, 2 mol%), AgNTf2 (3.1 mg, 0.008 mmol, 8 mol%), and AgOAc (0.6 mg, 0.004 mmol, 4 mol%). A reflux condenser was attached, and the completed apparatus purged and backfilled with N2 (3x). The mixture was constituted in CH2Cl2 (1.0 mL, 0.1 M), a solution of 2 (0.2 mmol, 2.0 equiv) in CH2Cl2 (2.0 mL, 0.1 M) was added over 5 h employing a syringe pump, and the mixture stirred at room temperature for an additional 48 h. The reaction was filtered through a short plug of silica gel eluting with 1:1 ethyl acetate/hexane (10 mL) and concentrated under reduced pressure. The crude mixture was purified by flash column chromatography eluting with hexane/EtOAc to afford 4.