Lewis Acid-Catalyzed 2,3-Dihydrofuran Acetal Ring-Opening Benzannulations toward Functionalized 1-Hydroxycarbazoles

The development of a Lewis acid-catalyzed, intramolecular ring-opening benzannulation of 5-(indolyl)2,3-dihydrofuran acetals is described. The resulting 1-hydroxycarbazole-2-carboxylates are formed in up to 90% yield in 1 h. The dihydrofuran acetals are readily accessed from the reactions of enol ethers and α-diazo-β-indolyl-β-ketoesters. To highlight the method’s synthetic utility, a formal total synthesis of murrayafoline A, a bioactive carbazole-containing natural product, was undertaken.


Introduction
The carbazole scaffold and its derivatives represent a privileged class of nitrogencontaining heteroaromatic structures often found in bioactive natural products and pharmaceutical drugs ( Figure 1) [1][2][3][4][5][6][7]. For instance, Celiptium ® is a marketed drug utilized for metastatic breast cancer [8], while carvedilol is used treat high blood pressure and heart failure [9]. In another example, carprofen is an anti-flammatory drug prescribed in veterinary medicine [10]. Carbazoles are also frequently employed in the development of advanced materials due to their conjugated tricyclic structure, which provides a tunable π-extended system [11]. This feature has been exploited for optical and thermoelectronic applications such as fluorescent dyes for bioimaging and conductive polymers for transistors, light emitting diodes, biosensors, and photovoltaic devices [12][13][14].

Introduction
The carbazole scaffold and its derivatives represent a privileged class of nitrogencontaining heteroaromatic structures often found in bioactive natural products and pharmaceutical drugs ( Figure 1) [1][2][3][4][5][6][7]. For instance, Celiptium ® is a marketed drug utilized for metastatic breast cancer [8], while carvedilol is used treat high blood pressure and heart failure [9]. In another example, carprofen is an anti-flammatory drug prescribed in veterinary medicine [10]. Carbazoles are also frequently employed in the development of advanced materials due to their conjugated tricyclic structure, which provides a tunable πextended system [11]. This feature has been exploited for optical and thermoelectronic applications such as fluorescent dyes for bioimaging and conductive polymers for transistors, light emitting diodes, biosensors, and photovoltaic devices [12][13][14].  The synthesis of carbazoles generally follows two distinct approaches ( Figure 2): (a) an annulation reaction to generate the central pyrrole ring or (b) a benzannulation reaction in which a benzene ring is appended on the five-member ring of an indole [15][16][17][18][19]. While several syntheses have been reported , the abundant presence of the carbazole core in bioactive molecules and photoelectronic materials makes them attractive targets for the  [45,46] Substituted 1-hydroxycarbazoles are an important class of bioactive carbazoles that have been the targets of synthetic efforts over the last 15 years [47,48]. Many of these efforts begin with commercial 1-hydroxycarbazole, requiring numerous steps for regioselective substituent installation [49]. Direct regioselective methods to 1-hydroxycarbazoles are relatively scarce in the literature. Thus, a need still exists for the efficient syntheses of these structural motifs. Encouraged by the successful example of indole formation from a Our lab previously established a Lewis acid-catalyzed approach to intramolecular benzannulation using (hetero)aryl-substituted 2,3-dihydrofuran acetals as precursors ( Figure 3) [45,46]. Vicinal hydroxybenzoates are thus produced following dihydrofuran ring opening (via acetal hydrolysis), enolate isomerization, intramolecular π-attack (on the resulting oxonium II), and subsequent alcohol elimination. We demonstrated the versatility of this benzannulation approach using dihydrofurans substituted with arenes and oxygen-and sulfur-containing-heteroaromatics, resulting in the formation of naphthalenes, benzofurans, dibenzo[b,d]furans, phenanthrenes, and benzothiophenes in good to high yields ( Figure 3a) [45]. More recently, we demonstrated that a (2-pyrrolyl)-substituted 2,3-dihydrofuran acetal was readily converted to the corresponding 7-hydroxyindole-6carboxylate in 77% yield (Figure 3b) [46].
The synthesis of carbazoles generally follows two distinct approaches ( Figure 2): (a) an annulation reaction to generate the central pyrrole ring or (b) a benzannulation reaction in which a benzene ring is appended on the five-member ring of an indole [15][16][17][18][19]. While several syntheses have been reported , the abundant presence of the carbazole core in bioactive molecules and photoelectronic materials makes them attractive targets for the development of new synthetic methodologies, often targeting new substitution patterns, modularity and mild reaction conditions. Our lab previously established a Lewis acid-catalyzed approach to intramolecular benzannulation using (hetero)aryl-substituted 2,3-dihydrofuran acetals as precursors ( Figure 3) [45,46]. Vicinal hydroxybenzoates are thus produced following dihydrofuran ring opening (via acetal hydrolysis), enolate isomerization, intramolecular π-attack (on the resulting oxonium II), and subsequent alcohol elimination. We demonstrated the versatility of this benzannulation approach using dihydrofurans substituted with arenes and oxygen-and sulfur-containing-heteroaromatics, resulting in the formation of naphthalenes, benzofurans, dibenzo[b,d]furans, phenanthrenes, and benzothiophenes in good to high yields (Figure 3a) [45]. More recently, we demonstrated that a (2-pyrrolyl)-substituted 2,3-dihydrofuran acetal was readily converted to the corresponding 7-hydroxyindole-6-carboxylate in 77% yield (Figure 3b) [46].  [45,46] Substituted 1-hydroxycarbazoles are an important class of bioactive carbazoles that have been the targets of synthetic efforts over the last 15 years [47,48]. Many of these efforts begin with commercial 1-hydroxycarbazole, requiring numerous steps for regioselective substituent installation [49]. Direct regioselective methods to 1-hydroxycarbazoles are relatively scarce in the literature. Thus, a need still exists for the efficient syntheses of these structural motifs. Encouraged by the successful example of indole formation from a  Figure 3. Dihydrofuran acetals as building blocks for benzannulation to generate (a) arenes and oxygen-/sulfur-containg heteroaromatics and (b) indoles [45,46]. Substituted 1-hydroxycarbazoles are an important class of bioactive carbazoles that have been the targets of synthetic efforts over the last 15 years [47,48]. Many of these efforts begin with commercial 1-hydroxycarbazole, requiring numerous steps for regioselective substituent installation [49]. Direct regioselective methods to 1-hydroxycarbazoles are relatively scarce in the literature. Thus, a need still exists for the efficient syntheses of these structural motifs. Encouraged by the successful example of indole formation from a dihydrofuran acetal [46], we herein discuss our efforts to develop efficient Lewis acid-catalyzed intramolecular ring-opening benzannulations of the corresponding indolyl-substituted dihydrofuran acetals to form 1-hydroxy-9H-carbazole-2-carboxylates (Scheme 1). dihydrofuran acetal [46], we herein discuss our efforts to develop efficient Lewis acidcatalyzed intramolecular ring-opening benzannulations of the corresponding indolylsubstituted dihydrofuran acetals to form 1-hydroxy-9H-carbazole-2-carboxylates (Scheme 1). Scheme 1. Proposed intramolecular ring-opening benzannulation approach to 1-hydroxycarbazoles.

Synthesis of Dihydrofuran Acetals 3
Synthesis of dihydrofuran acetals 3 were accomplished through the Cu(hfacac)2-catalyzed decomposition of N-indolyl α-diazo-β-ketoesters 1 in the presence of enol ethers 2 (Scheme 2) [46]. In most cases, the dihydrofuran-forming reaction proceeded as expected with yields up to 76%. The outliers included the reactions with 1-aryl-substituted enol ethers 2f-2i, dihydropyran 2k, and 5-methyl-2,3-dihydrofuran 2l. Under the reaction conditions, no dihydrofuran products were detected with the 1aryl-susbtituted ethyl enol ethers 2f and 2g. Instead, the corresponding furans (4af and 4ag) were obtained. It is likely that the dihydrofuran serves as a short-lived intermediate which readily undergoes Cu-promoted EtOH elimination to provide the conjugated furan. In contrast, when a p-chloro group is placed on the phenyl ring, dihydrofuran 3ah is readily formed in 65% yield. Employing the corresponding 1-aryl silyl enol ether 2i in the reaction conditions, unexpectedly provided a 2.4:1 inseparable mixture of dihydrofuran 3ai (50%) along with carbazole 5ai (21%). After some minor attempts to design direct one-pot or tandem transformations, we determined that carbazole 5ai could only be isolated in up to 36% yield when diazo 1a and enol ether 2i were treated with 20 mol% Cu(hfacac) 2 in 1,2-dichloroethane at 70 • C for 24 h.
For tetrahydropyran 2k, the desired dihydrofuran 3ak was obtained as an inseparable mixture with unidentifiable material in a 56% 1 H NMR yield. The crude mixture was carried forward. 5-Methyl-2,3-dihydrofuran 2l underwent partial in situ alkene isomerization to form 2-methylene tetrahydrofuran 2m. This isomerization resulted in the formation of a 2.9:1 mixture of fused-bicyclic dihydrofuran 3al and spirocyclic dihydrofuran 3am in 50% total yield.

Acid Screening for Ring-Opening Benzannulation
Dihydrofuran 3aa was selected as the initial system to begin optimizing the ringopening benzannulation (Table 1). Based on our previous work, we began by screening Lewis and Brønsted acids at 10 mol% catalyst loading in toluene at 70 • C. We confirmed that no reaction occurs in the absence of an acid catalyst (entry 1). Al(OTf) 3 , the best performing Lewis acid in our previous study, provided carbazole 5aa in 79% yield (entry 2). Both Sc(OTf) 3 and Ga(OTf) 3 gave the carbazole with yields of 71% and 66%, respectively (entries 3 and 4). Divalent metals, Zn(OTf) 2 and Mg(OTf) 2 , offered little or no product (entries 5 and 6). In contrast, Yb(OTf) 3 generated carbazole 5aa in 90% yield (entry 7). In(OTf) 3 afforded 82% yield of carbazole (entry 8), while tetravalent Hf(OTf) 4 gave the product in 79% yield (entry 9). To test the influence of any potential TfOH formed in the reaction, we ran a series of control reactions. First, YbCl 3 and AlCl 3 were employed in the reaction. YbCl 3 gave low conversion with only 24% yield of 5aa (entry 10). AlCl 3 provided 5aa in 80% yield (entry 11). With TfOH, 5aa was formed in 66% yield (entry 12). Given that TfOH provided good reactivity as well, pTsOH•H 2 O was used as another comparative Brønsted acid. Carbazole 5aa was formed in 55% yield along with 19% yield of furan 4aa (entry 13). These control reactions demonstrate that while Lewis acid catalysis is definitely in play, we must acknowledge the presence of cooperative catalysis by TfOH. We decided to move forward with Yb(OTf) 3 as the Lewis acid catalyst of choice given the high product yield. Further changes to either the temperature, solvent, or concentration failed to provide better product outcomes. aryl-susbtituted ethyl enol ethers 2f and 2g. Instead, the corresponding furans (4af and 4ag) were obtained. It is likely that the dihydrofuran serves as a short-lived intermediate which readily undergoes Cu-promoted EtOH elimination to provide the conjugated furan. In contrast, when a p-chloro group is placed on the phenyl ring, dihydrofuran 3ah is readily formed in 65% yield. Employing the corresponding 1-aryl silyl enol ether 2i in the reaction conditions, unexpectedly provided a 2.4:1 inseparable mixture of dihydrofuran 3ai (50%) along with carbazole 5ai (21%). After some minor attempts to design direct onepot or tandem transformations, we determined that carbazole 5ai could only be isolated in up to 36% yield when diazo 1a and enol ether 2i were treated with 20 mol% Cu(hfacac)2 in 1,2-dichloroethane at 70 °C for 24 h.
For tetrahydropyran 2k, the desired dihydrofuran 3ak was obtained as an inseparable mixture with unidentifiable material in a 56% 1 H NMR yield. The crude mixture was carried forward. 5-Methyl-2,3-dihydrofuran 2l underwent partial in situ alkene isomerization to form 2-methylene tetrahydrofuran 2m. This isomerization resulted in the formation of a 2.9:1 mixture of fused-bicyclic dihydrofuran 3al and spirocyclic dihydrofuran 3am in 50% total yield.

Acid Screening for Ring-Opening Benzannulation
Dihydrofuran 3aa was selected as the initial system to begin optimizing the ringopening benzannulation (Table 1). Based on our previous work, we began by screening Lewis and Brønsted acids at 10 mol% catalyst loading in toluene at 70 °C. We confirmed that no reaction occurs in the absence of an acid catalyst (entry 1). Al(OTf)3, the best performing Lewis acid in our previous study, provided carbazole 5aa in 79% yield (entry 2). Both Sc(OTf)3 and Ga(OTf)3 gave the carbazole with yields of 71% and 66%, respectively (entries 3 and 4). Divalent metals, Zn(OTf)2 and Mg(OTf)2, offered little or no product (entries 5 and 6). In contrast, Yb(OTf)3 generated carbazole 5aa in 90% yield (entry 7). In(OTf)3 afforded 82% yield of carbazole (entry 8), while tetravalent Hf(OTf)4 gave the product in 79% yield (entry 9). To test the influence of any potential TfOH formed in the reaction, we ran a series of control reactions. First, YbCl3 and AlCl3 were employed in the reaction. YbCl3 gave low conversion with only 24% yield of 5aa (entry 10). AlCl3 provided 5aa in 80% yield (entry 11). With TfOH, 5aa was formed in 66% yield (entry 12). Given that TfOH provided good reactivity as well, pTsOH•H2O was used as another comparative Brønsted acid. Carbazole 5aa was formed in 55% yield along with 19% yield of furan 4aa (entry 13). These control reactions demonstrate that while Lewis acid catalysis is definitely in play, we must acknowledge the presence of cooperative catalysis by TfOH. We decided to move forward with Yb(OTf)3 as the Lewis acid catalyst of choice given the high product yield. Further changes to either the temperature, solvent, or concentration failed to provide better product outcomes.

Benzannulation Substrate Scope
With dihydrofuran 3aa readily converting to carbazole 5aa with high yield, we next sought to explore the benzannulation reaction with the other synthesized substrates. We first examined 3-substituted dihydrofuran acetals 3ab-3ad (Scheme 3). Under the reactions, 2-ethoxy-3-methyl-substituted dihydrofuran 3ab (as a 3.4:1 trans:cis diastereomeric mixture) gave partial conversion to the corresponding carbazole-2-carboxylate 5ab in 58% yield (Scheme 3a). Interestingly, the recovered dihydrofuran (15%) was the cis-diastereomer. This outcome equates to a ~3.8:1 carbazole:cis-dihydrofuran ratio which closely parallels the initial dihydrofuran trans:cis ratio. Similar results were obtained with 3-ethyl substituted dihydrofuran 3ac (as 1.4:1 trans:cis mixture). Carbazole 5ac was formed in 42% yield along with 31% of recovered cis-dihydrofuran, a matching ~1.4:1 ratio (Scheme 3b). Seemingly contradictory, the trans-3-phenyl-substituted substrate 3ad afforded only trace (~5%) benzannulation product (Scheme 3c). Instead, epimerization of the acetal center occurred to generate cis-3ad in 64% yield, which is fairly unreactive under the reaction conditions. It is plausible that the phenyl substituent participates in the mechanism through anchimeric assistance resulting in stabilization of the dihydrofuran and reversible ringopening. Thus, we can infer that the trans-dihydrofuran isomers react faster than the corresponding cis-isomers under the reaction conditions. To overcome this reactivity issue, we first turned to revisiting some of the other Lewis acids that effectively formed the carbazole. Disappointingly, similar outcomes were obtained whether In(OTf)3, Sc(OTf)3, or Al(OTf)3 was used in place of Yb(OTf)3. At this point, we went back to our previous work for inspiration and decided to try Al(OTf)3 with a To overcome this reactivity issue, we first turned to revisiting some of the other Lewis acids that effectively formed the carbazole. Disappointingly, similar outcomes were obtained whether In(OTf) 3 , Sc(OTf) 3 , or Al(OTf) 3 was used in place of Yb(OTf) 3 . At this point, we went back to our previous work for inspiration and decided to try Al(OTf) 3 with a drop of H 2 O [45]. These conditions previously prevented furan formation from the corresponding dihydrofuran acetals. Satisfyingly, subjecting dihydrofurans 3ab-3ad to Al(OTf) 3 and 1 drop of H 2 O, provided the desired carbazoles 5ab, 5ac, and 5ad in 67%, 62%, and 65% yields, respectively (Scheme 4). Our rationale for the shift in reactivity with the added water is the likely formation of TfOH (in line with the catalyst screen) which facilitates generation of dihydrofuran hemiacetal intermediate V that undergoes ring-opening. To confirm our hypothesis, we treated 3ad with TfOH (10 mol%) in toluene at 70 • C and obtained carbazole 5ad in 54% yield. sponding dihydrofuran acetals. Satisfyingly, subjecting dihydrofurans 3ab-3ad to Al(OTf)3 and 1 drop of H2O, provided the desired carbazoles 5ab, 5ac, and 5ad in 67%, 62%, and 65% yields, respectively (Scheme 4). Our rationale for the shift in reactivity with the added water is the likely formation of TfOH (in line with the catalyst screen) which facilitates generation of dihydrofuran hemiacetal intermediate V that undergoes ringopening. To confirm our hypothesis, we treated 3ad with TfOH (10 mol%) in toluene at 70 °C and obtained carbazole 5ad in 54% yield. With two sets of reaction conditions, we proceeded forward with the exploration of the substrate scope. Figure 4 summarizes the outcomes of the study, including those systems previously discussed (entries 1-4). The trisubstituted dihydrofuran 3ab (derived from ethyl vinyl ether) smoothly converted using Yb(OTf)3 to its 1-hydroxy carbazole-2carboxylate 5ab in 81% yield (entry 5). For 2-(4-chlorophenyl)-2-methoxy dihydrofuran 3ah, elimination to furan 4ah was the predominant outcome observed for both Yb(OTf)3 and Al(OTf)3/H2O. The formed carbazole 5ah (5% for Yb and 19% for Al) was inseparable from the furan product in both cases (entry 6). Next, we examined the fused bicyclic dihydrofurans 3aj and 3ak (entries 7 and 8). Dihydrofuran 3aj was synthesized as the cisdiastereomer. Under the Al(OTf)3/H2O conditions, only furan 4aj was obtained. After some minor optimization, we found that Al(OTf)3 without added water produced a 78% yield of lactono-carbazole 6aj which resulted from the intramolecular transesterification of 5aj (entry 7). Under the same conditions (Al(OTf)3 with no added water), tetrahydropyran-fused dihydrofuran acetal 3ak readily provided carbazole 5ak in 22% (entry 8). No lactonization was observed which is consistent with entropic and enthalpic considerations for seven-membered ring formation [50]. Based on the reactions with the bicyclic dihydrofurans, we rationalized two things relative to the tetrasubstituted monocyclic dihydrofurans 3ab-3ad: (1) Diastereoselectivity does not seem to be a factor in the reactivity; (2) It is likely that hemiacetal formation is unfavorable with the added water due to the stability of the bicyclic acetal framework. When the mixture of fused-bicyclic dihydrofuran 3al and spirocyclic dihydrofuran 3am was subjected to the same conditions (Al(OTf)3, no
With two sets of reaction conditions, we proceeded forward with the exploration of the substrate scope. Figure 4 summarizes the outcomes of the study, including those systems previously discussed (entries 1-4). The trisubstituted dihydrofuran 3ab (derived from ethyl vinyl ether) smoothly converted using Yb(OTf) 3 to its 1-hydroxy carbazole-2-carboxylate 5ab in 81% yield (entry 5). For 2-(4-chlorophenyl)-2-methoxy dihydrofuran 3ah, elimination to furan 4ah was the predominant outcome observed for both Yb(OTf) 3 and Al(OTf) 3 /H 2 O. The formed carbazole 5ah (5% for Yb and 19% for Al) was inseparable from the furan product in both cases (entry 6). Next, we examined the fused bicyclic dihydrofurans 3aj and 3ak (entries 7 and 8). Dihydrofuran 3aj was synthesized as the cis-diastereomer. Under the Al(OTf) 3 /H 2 O conditions, only furan 4aj was obtained. After some minor optimization, we found that Al(OTf) 3 without added water produced a 78% yield of lactono-carbazole 6aj which resulted from the intramolecular transesterification of 5aj (entry 7). Under the same conditions (Al(OTf) 3 with no added water), tetrahydropyran-fused dihydrofuran acetal 3ak readily provided carbazole 5ak in 22% (entry 8). No lactonization was observed which is consistent with entropic and enthalpic considerations for seven-membered ring formation [50]. Based on the reactions with the bicyclic dihydrofurans, we rationalized two things relative to the tetrasubstituted monocyclic dihydrofurans 3ab-3ad: (1) Diastereoselectivity does not seem to be a factor in the reactivity; (2) It is likely that hemiacetal formation is unfavorable with the added water due to the stability of the bicyclic acetal framework. When the mixture of fused-bicyclic dihydrofuran 3al and spirocyclic dihydrofuran 3am was subjected to the same conditions (Al(OTf) 3 , no water), lactono-carbazole 6al was obtained in 61% yield along with 19% yield of carbazole 5am (entry 9). water), lactono-carbazole 6al was obtained in 61% yield along with 19% yield of carbazole 5am (entry 9).  3 Inseparable mixture with furan 4ah. 4 No water was added. 5 No desired product formed using either procedure. Obtained furan 4ca instead.
For the future purposes of synthesis, the effects of changing the N-methyl substituent to a N-benzyl group was explored. The corresponding carbazole 5ba was obtained in 80% yield using Yb(OTf)3 for dihydrofuran 3ba (entry 10), whereas Al(OTf)3/H2O was used to generate 67% yield of carbazole 5bb from dihydrofuran 3bb (entry 11).  3 Inseparable mixture with furan 4ah. 4 No water was added. 5 No desired product formed using either procedure. Obtained furan 4ca instead.
For the future purposes of synthesis, the effects of changing the N-methyl substituent to a N-benzyl group was explored. The corresponding carbazole 5ba was obtained in 80% yield using Yb(OTf) 3 for dihydrofuran 3ba (entry 10), whereas Al(OTf) 3 /H 2 O was used to generate 67% yield of carbazole 5bb from dihydrofuran 3bb (entry 11).
Lastly, we attempted the benzannulation of the 3-indolyl dihydrofuran 3ca (entry 12). In all cases, regardless of what procedure or Lewis acid employed, only furan 4ca was obtained.

Formal Synthesis of Murrayafoline A
Using carbazole 5bb as a precursor, we undertook the formal synthesis of murrayafoline A (8, Scheme 5). Murrayafoline A is a monomeric carbazolic alkaloid that has been isolated from kilograms of the dried powdered roots of several species of the genus Murraya, Glycosmis, and Clausena in up to 3% yield [51][52][53]. Murrayafoline A was shown to exhibit a number of interesting biological activities including strong fungicidal activity (12.5 µg dose against C. cucumerinum) [52], cancer growth inhibitory activity (HT-1080 cells) [53], and Wnt/β-catenin signaling antagonist activity [54]. It has been the target of numerous several syntheses due to its role as a precursor to access a variety of congeners and non-natural derivatives. From carbazole 5bb, we could readily access N-benzyl murrayafoline A 9 in 66% yield over two steps following Krapcho decarbalkoxylation and O-methylation. Carbazole 9 has been shown by Mal and coworkers [47] to readily convert to the target molecule in in one step. Lastly, we attempted the benzannulation of the 3-indolyl dihydrofuran 3ca (entry 12). In all cases, regardless of what procedure or Lewis acid employed, only furan 4ca was obtained.

Formal Synthesis of Murrayafoline A
Using carbazole 5bb as a precursor, we undertook the formal synthesis of murrayafoline A (8, Scheme 5). Murrayafoline A is a monomeric carbazolic alkaloid that has been isolated from kilograms of the dried powdered roots of several species of the genus Murraya, Glycosmis, and Clausena in up to 3% yield [51][52][53]. Murrayafoline A was shown to exhibit a number of interesting biological activities including strong fungicidal activity (12.5 µg dose against C. cucumerinum) [52], cancer growth inhibitory activity (HT-1080 cells) [53], and Wnt/β-catenin signaling antagonist activity [54]. It has been the target of numerous several syntheses due to its role as a precursor to access a variety of congeners and non-natural derivatives. From carbazole 5bb, we could readily access N-benzyl murrayafoline A 9 in 66% yield over two steps following Krapcho decarbalkoxylation and Omethylation. Carbazole 9 has been shown by Mal and coworkers [47] to readily convert to the target molecule in in one step.

Experimental
A. General. All reactions were performed under protection of N2 in flame-dried glassware unless water was applied as solvent. Chromatographic purification was performed as flash chromatography with Silicycle SiliaFlash P60 silica gel (40-63 µm) or preparative thin-layer chromatography (prep-TLC) using silica gel F254 (1000 µm) plates and solvents indicated as eluent with 0.1-0.5 bar pressure. For quantitative flash chromatography, technical grades solvents were utilized. Analytical thin-layer chromatography (TLC) was performed on Silicycle SiliaPlate TLC silica gel F254 (250 µm) TLC glass plates. Visualization was accomplished with UV light. Infrared (IR) spectra were obtained via thin film IR on a salt plate using a Nicolet 6700 Fourier-transform infrared spectrophotometer. The IR bands are characterized as broad (br), weak (w), medium (m), and strong (s). Proton and carbon nuclear magnetic resonance spectra (Supplementary Materials: 1 H NMR, 13 C NMR and 19 F NMR) were recorded on a Bruker 400 MHz spectrometer or on a Bruker 500 MHz spectrometer or on a Bruker 700 MHz spectrometer with solvent resonances as the internal standard ( 1 H NMR: CDCl3 at 7.26 ppm; 13 C NMR: CDCl3 at 77.0 ppm). 1 H NMR data are reported as follows: chemical shift (ppm), multiplicity (s = singlet, d = doublet, dd = doublet of doublets, dt = doublet of triplets, ddd = doublet of doublet of doublets, t = triplet, q = quartet, p = pentet, m = multiplet, br = broad), coupling constants (Hz), and integration. Mass spectra were obtained through EI on a Micromass AutoSpec machine or through ESI on a Thermo Orbitrap XL. The accurate mass analyses run in EI mode were at a mass resolution of 10,000 and were calibrated using PFK (perfluorokerosene) as an internal standard. The accurate mass analyses run in EI mode were at a mass resolution of 30,000 using the calibration mixture supplied by Thermo. Crystal structures for carbazoles 5ad (Deposition Number 2213553) and 5ai (Deposition Number 2213554) were deposited in the Cambridge Structural Database.
B. Synthesis of Indolyl-α-diazo-β-ketoesters 1. General Procedure: Using the following modified literature procedure [55]: To a dry flask charged with a stir bar and the corresponding indole carboxylic acid (1.0 equiv.), dry CH2Cl2 was added to make a 0.5 M solution, followed by addition of catalytic DMF (a few drops). The solution was then

Experimental
A. General. All reactions were performed under protection of N 2 in flame-dried glassware unless water was applied as solvent. Chromatographic purification was performed as flash chromatography with Silicycle SiliaFlash P60 silica gel (40-63 µm) or preparative thin-layer chromatography (prep-TLC) using silica gel F254 (1000 µm) plates and solvents indicated as eluent with 0.1-0.5 bar pressure. For quantitative flash chromatography, technical grades solvents were utilized. Analytical thin-layer chromatography (TLC) was performed on Silicycle SiliaPlate TLC silica gel F254 (250 µm) TLC glass plates. Visualization was accomplished with UV light. Infrared (IR) spectra were obtained via thin film IR on a salt plate using a Nicolet 6700 Fourier-transform infrared spectrophotometer. The IR bands are characterized as broad (br), weak (w), medium (m), and strong (s). Proton and carbon nuclear magnetic resonance spectra (Supplementary Materials: 1 H NMR, 13 C NMR and 19 F NMR) were recorded on a Bruker 400 MHz spectrometer or on a Bruker 500 MHz spectrometer or on a Bruker 700 MHz spectrometer with solvent resonances as the internal standard ( 1 H NMR: CDCl 3 at 7.26 ppm; 13 C NMR: CDCl 3 at 77.0 ppm). 1 H NMR data are reported as follows: chemical shift (ppm), multiplicity (s = singlet, d = doublet, dd = doublet of doublets, dt = doublet of triplets, ddd = doublet of doublet of doublets, t = triplet, q = quartet, p = pentet, m = multiplet, br = broad), coupling constants (Hz), and integration. Mass spectra were obtained through EI on a Micromass AutoSpec machine or through ESI on a Thermo Orbitrap XL. The accurate mass analyses run in EI mode were at a mass resolution of 10,000 and were calibrated using PFK (perfluorokerosene) as an internal standard. The accurate mass analyses run in EI mode were at a mass resolution of 30,000 using the calibration mixture supplied by Thermo. Crystal structures for carbazoles 5ad (Deposition Number 2213553) and 5ai (Deposition Number 2213554) were deposited in the Cambridge Structural Database.
B. Synthesis of Indolyl-α-diazo-β-ketoesters 1. General Procedure: Using the following modified literature procedure [55]: To a dry flask charged with a stir bar and the corresponding indole carboxylic acid (1.0 equiv.), dry CH 2 Cl 2 was added to make a 0.5 M solution, followed by addition of catalytic DMF (a few drops). The solution was then cooled to 0 • C and oxalyl chloride (1.2 equiv.) was slowly added over 1 min. After 15 min, the reaction was allowed to warm to room temperature with continued stirring. After 3 h at room temperature, the reaction was concentrated under reduced pressure and the acid chloride residue was dissolved in dry THF to make a 1 M solution (keeping it under inert atmosphere). The solution was added slowly to the prepared enolate at −78 • C.