CARBON-NITROGEN BOND FORMATION VIA TANDEM CARBON-HYDROGEN (C-H) AND NITROGEN-HYDROGEN (N-H) BOND FUNCTIONALIZATION

The oxidative cross-coupling of carbon-hydrogen (C-H) and nitrogenhydrogen (N-H) bonds to form carbon-nitrogen (C-N) bonds is an important synthetic advance, as amine and amide functional groups are ubiquitous in biologically active molecules. This technique is orthogonal to conventional amination techniques, which rely on electrophilic nitration/reduction strategies or metal catalyzed coupling of prefunctionalized arenes. This dissertation’s main focus is on the development of oxidative methods for constructing N-arylamines and amides via tandem C-H/N-H bond activation and increasing synthetic efficiency for total synthesis of an inhibitor of botulinum neurotoxin via direct C–H functionalization. The first manuscript, “Metal-Free Intermolecular Oxidative C-N Bond Formation via Tandem C-H and N-H Bond Functionalization,” is focused on the development of a novel intermolecular oxidative amination reaction, a synthetic transformation that involves the simultaneous functionalization of both an N-H and a C-H bond. The process, which is mediated by an I(III) oxidant and contains no metal catalysts, provides a rapid and green method for synthesizing protected anilines from simple arenes and phthalimide. The second manuscript, “I(III)-Mediated Regioselective C-H Bond Amination of 2-Arylpyridine Derivatives,” is focused on the development of a novel, useful and economical process for the direct amination of 2-phenylpyridine derivatives. This process requires cheap and commercially available copper triflate and works for a variety of different 2-phenylpyridine derivatives. The third manuscript, “Increasing synthetic efficiency via direct C–H functionalization: formal synthesis of an inhibitor of botulinum neurotoxin,” is focused on designing an efficient scheme for the synthesis of one of the best known inhibitors of botulinum neurotoxin serotype A (BoNTA). The synthetic route involves two palladium catalyzed C–H functionalization reactions, formally activating three C–

vii

PREFACE
The following research has been presented in manuscript format according to guidelines of the Graduate School of the University of Rhode Island. The entire dissertation is divided into three manuscripts.
The formation of carbon-carbon (C-C) bonds via the oxidative cross-coupling of two carbon-hydrogen (C-H) bonds has recently become a field of intense interest and has resulted in the discovery of numerous novel synthetic methods. The analogous technology for the oxidative cross-coupling of C-H and nitrogen-hydrogen (N-H) bonds to form carbon-nitrogen (C-N) bonds would also be an important synthetic advance, as amine and amide functional groups are ubiquitous in biologically active molecules. However, a relatively small amount of work has focused on the development of oxidative methods for constructing N-arylamines and amides via tandem C-H/N-H bond activation.
The majority of literature in this field describes the insertion of nitrenoid intermediates into C-H bonds, mediated by transition metal catalysts. Cu and Pdcatalyzed reactions, as well as metal-free conditions, have also been recently explored, but these methods are limited to the intramolecular synthesis of carbazoles, oxindoles and diazapenones. The only reported examples of intermolecular oxidative amination involve azole-type heterocycles and proceed by attack of an amine nucleophile on the substrate's imine moiety. The ability to oxidatively couple phthalimide to unfunctionalized arenes is a useful method for synthesizing anilines that is orthogonal to conventional amination techniques, which rely on electrophilic nitration/reduction strategies or metal catalyzed coupling of pre-functionalized arenes. Phthalimide, in xii LIST OF TABLES   The formation of carbon-carbon (C-C) bonds via the oxidative cross-coupling of two carbon-hydrogen (C-H) bonds has recently become a field of intense interest and has resulted in the discovery of numerous novel synthetic methods. 1 The analogous technology for the oxidative cross-coupling of C-H and nitrogen-hydrogen (N-H) bonds to form carbon-nitrogen (C-N) bonds would also be an important synthetic advance, as amine and amide functional groups are ubiquitous in biologically active molecules ( Figure 1). However, a relatively small amount of work has focused on the development of oxidative methods for constructing N-arylamines and amides via tandem C-H/N-H activation. The majority of literature in this field describes the 3 insertion of nitrenoid intermediates into C-H bonds, mediated by transition metal catalysts. 2 Cu and Pd-catalyzed reactions, as well as metal-free conditions, have also been recently explored, but these methods are limited to the intramolecular synthesis of carbazoles, oxindoles and diazapenones. [3][4][5] The only reported examples of intermolecular oxidative amination involve azole-type heterocycles and proceed by Figure 1. Examples of high-value molecules containing anilines attack of an amine nucleophile on the substrate's imine moiety. 6,7 Herein, we disclose a novel intermolecular reaction that oxidatively constructs the C-N bond of phthalimide-protected anilines via the tandem activation of N-H and C-H bonds.

Results and Discussion
Reaction Discovery. We recently discovered that heating a solution of phthalimide (1) and PhI(OAc) 2 in benzene provides the phthalimide-protected aniline (2) in an 88% yield (Table 1, entry 4). We initially explored the potential of Cu(I)/(III)-catalysts to 4 perform this oxidative amination, but quickly realized that a metal catalyst was not necessary. 5 We are confident that trace metals are not the cause of the transformation, as reactions performed in new, acid washed flasks were similar in both yield and rate to those performed in old flasks. Furthermore, reagents from different commercial sources perform similarly.
While optimizing the process, we quickly discovered that the best conditions employed microwave heating and 2.5 equivalents of the I(III) oxidant, phenyliodine(III) diacetate (PIDA). Less than two equivalents of PIDA did not allow the reactions to proceed to completion (compare entries 2 and 3).
Importantly, we found that the arene substrate does not need to be in large excess (i.e. solvent) for the intermolecular oxidative amination reaction to occur. This is in stark contrast to many of the oxidative arylation reactions that have been previously studied. 8 As shown in Table 1, entry 9, acetonitrile was used as the solvent and the concentration of the arene substrate was lowered to a near-stoichiometric level (1.5 equiv). Under these conditions, the yield for the amination of benzene dropped from 88% to 51%, but we hypothesized that this was due to the volatility of the arene. This was verified by the reaction of the less volatile substrate, p-xylene, which provided an 80% yield (Scheme 1A).  [b] Oil bath. [c] Yield of isolated product after column chromatography. [d] GC yield calculated using dodecane as internal standard.

Oxidant
Other solvent and oxidant combinations proved to be less effective. In particular, 2,2,2-trifluoroethanol (TFE), a solvent that has been shown to be particularly compatible with I(III) mediated arene substitutions provided a minimal yield (entry 6).
The fluorinated derivative of PIDA, phenlyiodine(III) bis(trifluoroacetate) (PIFA), proved to be too harsh of an oxidant for these reactions. Any loading of the oxidant in excess of 1 equiv resulted in a complex mixture of products. Lowering the reaction temperature and altering the co-solvent, also failed to provide the desired anilne 2 (entries [13][14][15][16][17][18]. To determine whether the arene that participated in the oxidative amination originated from the benzene solvent or the oxidant, a crossover experiment was performed (Scheme 1B). When toluene was used as the solvent and PhI(OAc) 2 was the oxidant, the reaction did not produce any N-phenylphthalimide (2). Rather, an inseparable mixture of regiomers arising from the oxidative amination of the sp 2 -hybridized C-H bonds of toluene was observed (3). This confirmed our hypothesis that the source of the aryl group that forms the new C-N bond was the solvent, not the phenyl group that resides on the oxidant.

Scheme 1. Oxidative amination of substituted arenes
Cleavage of the phthalimide protecting group allowed for the determination of the regiomeric ratios by comparison to the commercially available toluidine isomers.
Interestingly, the major product of the reaction was o-toluidine (4). Amination of the sp 3 -hybridized C-H bonds was not observed.
The regioselectivity of these amination reactions appears to be slightly directed by electronic factors. For example, p-methylanisole was 50% more likely to be aminated at the position ortho to the larger, but more electron donating, methoxy group (15).
Likewise, both the ortho and para-positions of toluene were 67% more likely to be aminated than the meta-position. The only exception to this rule was the amination of p-tert-butylanisole (16), where the size of the tert-butyl group prevented orthoamination.
Other amine sources, in addition to phthalimide, were also investigated. Succinimide also provided very good yields of oxidative amination products (20), while pyrrolidin-2-one performed the oxidative amination reaction, but resulted in a poor yield (16%, 21). N-Tosylamide and pyrrolidine failed to oxidatively aminate benzene. These data led us to conclude that the requirements for the amine coupling partner were two-fold: 1) the N-H bond must be relatively acidic (phthalimide pKa = 8.3) and 2) the amine coupling partner must be secondary, preferably a cyclic imide. In keeping with these requirements, other imides with acidic N-H bonds performed the oxidative amination reaction, albeit in lower yields (22)(23)(24). Surprisingly, acyclic imides, such as diacetamide, N-acetylanilines and hetercycles such as indole and benzimidazole produced only trace amounts of products.

Competition and Kinetic Studies.
Competition studies were performed to explore the mechanism of the novel process (Table 3). In all cases, the amination of electron-rich arenes was favored. In particular, the competition between p-xylene and pdifluorobenzene shows a dramatic preference for the amination of the electron rich pxylene substrate (entry 2).
The kinetic isotope effect (KIE) of the reaction was assessed using a competition experiment between equimolar amounts of benzene and benzene-d 6 . The near-unity KIE of 1.03 implies that C-H bond breaking is not involved in the rate-determining step of the reaction. This is also a stark contrast with the literature describing oxidative arylation processes for forming C-C bonds, where C-H bond breaking is often rate limiting, as indicated by large KIEs. 9 The KIE of the reaction involving equimolar amounts of phthalimide and phthalimide-d was observed to be 0.98, which indicates that the cleavage of the N-H bond is also not rate-limiting.

Ar 1 -H Ar 2 -H
[a]    The preference for imides and electron-rich arenes is consistent with two possible mechanisms (Scheme 2). The first involves the in situ formation of a PhI(OAc)(NR 2 ) species (25 and 26), which is highly electrophilic, functioning essentially as an R 2 N + equivalent. 10 Electrophilic aromatic substitution then forms the desired C-N bond (2) 13 (Scheme 2A). 11 Alternatively, a single electron transfer may occur, forming an ion pair with the arene substrate and the I(III) reagent. The radical cation 27 could then undergo a nucleophilic attack by phthalimide (or its anion), giving rise to the desired product (3, Scheme 2B). Kita has previously described such a mechanism for oxidative reactions between arenes and soft nucleophiles such as b-diketones and TMS-N 3 . 12 N-centered radicals have also been proposed in I(III) mediated C-N forming reactions, 13 but they are not likely to be involved in the reactions shown in Table 1. The weaker C-H bonds of the methyl group of toluene were not aminated, as one would expect for reactions involving N-centered radicals.

Scheme 2. Two possible mechanisms for the oxidative amination
Proposed Mechanism. Based these data, a mechanism involving electrophilic aromatic substitution (EAS) seems unlikely, as the ratio of products obtained from the oxidative amination of toluene (3) are not indicative of the expected EAS regioselectivity (Scheme 2A). While the ortho and para-aminated products were somewhat favored in the reaction of arenes containing electron donating groups, reactions operating via S E Ar mechanisms tend to form little, if any, meta-substituted products. Alternatively, PhI(OAc) 2 could oxidize the electron-rich arene substrate to a radical cation (27), and nucleophilic attack on such a radical cation should be relatively non-regioselective (Scheme 2B). The consequent radical intermediate (28) could then be oxidized to form a Wheland-type arenium ion. These two individual oxidation steps may indicate why two equivalents of PhI(OAc) 2 are required to achieve complete conversion. The fates of the reduced iodine intermediates are less clear, but the two other by-products that are observed in all of these reactions are iodobenzene and phenyl acetate. Consequently, we hypothesize that a single electron transfer mechanism is operating in these oxidative amination reactions. This hypothesis is supported by the observation that the reaction was partially inhibited by BHT and completely inhibited by TEMPO, common radical inhibitors. Our hypothesis is supported by the prior work of Kita, who has extensively shown that arene radical cation intermediates such as 27 are formed by the action of I(III) oxidants and can be directly observed by EPR spectroscopy. 12 It should be noted that Cho and Chang have recently proposed an electrophilic mechanism for the same reaction described herein. 7b This hypothesis was supported by the observation of 25 by mass spectrometry. In comparing our data with those of Kita and both Cho and Chang, it seems likely that 15 several I(III) species are simultaneously present in solution. However, we reason that the observed regioselectivites for aminations of monosubstituted arenes that were observed by both us and Cho and Chang are best explained by the intermediacy of a radical cation intermediate (27).

Scheme 3.
Resonance structures of the aryl radical cation explain the observed regioselectivities.
Furthermore, the unique regioselectivities that were observed in the oxidative amination of toluene (o:m:p = 10:6:5) corroborate the intermediacy of an aromatic radical cation (Scheme 3). The yields of ortho and para-aminated products, which arise from the nucleophilic attack on resonance forms 27b and 27c, are statistically equivalent. This seems plausible as both contain a tertiary radical and a secondary carbocation. Additionally, reactions with p-methylanisole and p-tert-butylanisole (28) not only produced the aforementioned aminated products (14)(15)(16), but substitution products, arising from S N Ar-type attack on the tertiary cation intermediate (29b) followed by ejection of the methoxide leaving group (30), were also observed (Scheme 4).

Conclusion
In conclusion, the ability to oxidatively couple phthalimide to unfunctionalized arenes is a useful method for synthesizing anilines that is orthogonal to conventional amination techniques, which rely on electrophilic nitration/reduction strategies or metal catalyzed coupling of pre-functionalized arenes. Phthalimide, in particular, is an ideal starting point for the development of the aforementioned oxidative amination technology. It is commercially available, inexpensive, easy to handle and, once coupled, it can be readily converted to a primary amine, which can be further derivatized. Additionally, N-arylphthalimides like those shown in Table 1 have recently been shown to have anti-cancer activity. 14 Future work in our laboratory will be dedicated to the mechanistic study and application of this unique method for constructing C-N bonds.

Instrumentation
Reactions were carried out in a CEM Discover microwave. GC/MS analysis was carried out on an Agilent Technologies 6890 GC system fixed with a 5973 mass selective detector. NMR spectrum were acquired using a Bruker Avance 300MHz spectrometer.

Synthesis of 2-phenylisoindoline-1, 3-Dione (2)
A magnetically stirred solution of phthalimide (0.10 g, 0.68mmol), iodobenzene diacetate (0.55 g, 1.7mmol) in 4 mL of benzene was microwave heated at 145 0 C for 3 h. The excess solvent from the mixture is removed at reduced pressure and the crude product was purified by column chromatography to give pure 2 (0.133 g, 88 %

Synthesis of 17, 18, 19
A magnetically stirred solution of phthalimide (0.10 g, 0.68mmol), iodobenzene diacetate (0.55 g, 1.7mmol) in 4 mL of m-xylene was microwave heated at 145 0 C for 3 h. The excess solvent from the mixture is removed at reduced pressure and crude product was purified by column chromatography to give a mixture of 17, 18

2) Order in Substrate
The order in the substrate (1)

Competition experiment
A magnetically stirred solution of phthalimide (0.10 g, 0.68mmol), iodobenzene diacetate (0.55 g, 1.7 mmol) with equimolar amounts of Ar-H (2.0 mL) and Ar'-H (2.0 mL,) was microwave heated at 145 0 C for 3 h. The reaction was then cooled, and an aliquot was removed and analyzed by GC/MS.

Entry Ar-H Ar ' -H PhthN-Ar PhthN-Ar
The ratios of the products were determined by GC/MS.

Phthalimide Deprotection
N-Hydrazine hydrate (0.15 g, 4.5 mmol) and 3 mL of water was added to a solution of  4 Recently, quests to discover highly efficient, economical, and environmentally benign methods for C-N bond formation have led to advances in copper mediated and copper catalyzed C-H amination. 5 Yu and coworkers reported a C-N bond formation process using stoichiometric copper acetate (Eq 1). 2b Similarly, Nicholas and associates reported a copper catalyzed C-N bond formation process; however this required the incorporation of additives and extremely harsh conditions (Eq 2). 6 Likewise, Li reported a copper catalyzed direct amidation of 2phenylpyridine (1) with acetanilide through a cross dehydrogenative coupling (CDC) process. 7 (1) Recently, we and others have reported metal free intermolecular oxidative C-N bond formation processes. Though these reactions provide high yield of amination products, the regioselevtivity appears to be directed by electronic factors (Eq 3). 2d, 2e, 8 In addition , Hartwig, reported a palladium catalyzed amination reaction to sterically control the regioselectivity. 9

81
(3) Our initial approach to aminate 2-phenylpyridine substrates using our reported I(III) mediated metal free oxidative amination process were not successful. 2d Recently, Muniz has reported a variety of novel hypervalent iodine (III) reagent for metal free intermolecular allyic amination and di-aminiation of alkenes (Eq 4). 10 (4) Furthermore, diaryl iodonium salts have been used by the Sanford group to acetoxylate C-H bonds in presence of catalytic palladium; 11 and by the Gaunt group to arylate C-H bonds in the presence of catalytic copper (Eq 5). 12 (5) 82 As a result, we synthesized a phenyliodine (III)bis[phthalimidate] containing iodine-nitrogen (I-N) bond (Eq 6). 13 (6) This new Iodane (2) was subjected to both metal-free and catalytic palladium/copper conditions with 2-arylpyridine substrates. We hypothesized that 1) Iodane could be used to generate electrophilic amine moieties which could react nonselectively with arenes and 2) regioselectivity can be achieved by using selective metalation of specific C aryl -H bonds using a transition metal catalyst. Our results proved fruitful, and lead us to conclude that a metal was necessary to facilitate C-N bond formation. Herein, we report a novel method to selectively aminate 2-phenyl pyridine substrate using the Iodane and copper triflate (Eq 7).

Results and Discussion
Our initial investigation began by heating a solution of phthalimide, 2phenylpyridine (1), and iodobenzene diacetate (PIDA) in acetonitrile at elevated temperatures in a microwave. However, we determined that PIDA was not able to assist the amination of pyridine derivative. As a result, the incorporation of catalytic palladium, along with PIDA or iodobenzene bis(trifluoroacetate) (PIFA), was also explored. While progress was made and amination was observed, this methodology still did not facilitate amination with acceptable yields (See SI, Table 4).
Thus, we concluded that either a different catalyst or different hypervalent iodine reagents, which are serving as our oxidant, were required. These theories lead to the construction of the bis-phthalimide iodane (2), to replace our original nitrogen source, phthalimide. A control experiment was conducted where this iodane, along with with 2-phenyl pyridine (1), was heated without any catalyst or oxidant. This experiment did not yield the desired aminated product (3). The next step was to use either an oxidant or metal catalyst to achieve the required amination. Initial experiments had shown that oxidants like PIDA and PIFA had no effect on the amination reaction.
[c] GC yield calculated using calibration curve method..
As a result, we elected to screen metal additives and discovered that copper (II) triflate showed higher amounts of amination products when compared to the previously successful experiment with palladium acetate. It is worth noting that freshly prepared iodane exhibited higher yields in comparison to salts left exposed to atmospheric conditions. We theorize this is a result of atmospheric moisture hydrolyzing the iodane (2). Low equivalents of copper triflate, reduced reaction time, and decreasing the temperature also showed a reduction in reaction yield. (See SI, Table 4).
Encouraged by our initial screening with various amounts of copper triflate, we further explored the potential of Cu(I)/(II) catalysts to determine if higher conversions of arene substrates could be obtained. Dichloroethane was determined to be the preferred solvent, and copper acetate and copper chloride provided the aminated product along with acetylated and chlorinated byproducts of 2-arylpyridine substrate (4) ( Table 1, entries 6,7). This was observed by GC/MS. When determining the stoichiometric necessity for the iodane an interesting pattern was observed. Since the salt acts as both our nitrogen source as well as oxidant, it has to be used in excess as reactions ran with equimolar quantities showed lower yields (See SI, Table 1).
Optimized conditions required heating the iodane (2), 2-arylpyridine substrate (4) and 1 eq. (relative to the 4) of copper triflate in dicholoethane at 80 0 C for 48 hrs on a hotplate. Amination was exclusively observed at the ortho position, relative to the pyridine substituent to yield (5) and was confirmed via COSY NMR. Table 2 shows a variety of 2-phenyl pyridine substrates that could be aminated using our novel reaction conditions. Substrates with electron withdrawing functional groups (9-11) exhibited diminished yields in amination when compared to substrates with electron rich groups. First, substrates containing different phenyl-substituted groups were screened. 2-toylpyridine (5) showed the highest yield as compared to the methoxy derivative (7).

Substrate scope
A chlorinated derivative (10) showed a substantially lower yield as compared to its competitive halogen counterpart, fluorine (9). The aminated product was also observed when the aldehyde derivate (11) was subjected to our reaction conditions; albeit at reduced yields. This is a promising result as aldehydes can be easily converted into a carboxylic acid and can be used to obtain the carboxy aminated product. Substitution on the pyridine side of the substrate also provided the required products in moderate to high yields (12)(13)(14). Additionally, the process also works with a bis-saccharin iodane, to yield the expected aminated product (15).

Competition and kinetic studies
To elucidate the mechanism of this novel reaction, competition studies were conducted using a mixture of equimolar amounts of two different pyridine substrates in the lead reaction. In most of the cases, amination of electron rich pyridine substrates was more favored over electron poor pyridine substrates. The kinetic isotope effect was also studied using an intramolecular competition between a C-H bond and C-D bond in 2-phenyl pyridine. An observed KIE of 1.15 shows that C-H bond cleavage is not involved in the rate determining step of the reaction.

Proposed mechanism
A number of insights were obtained from the experiments carried out for determination of the reaction mechanism. The deuterated substrate ( ( ring with electron donating group will be much faster than that from an aryl ring with electron withdrawing group. Thus SET step has to be the rate limiting step in this reaction.

Conclusion
In conclusion, we have developed a novel, useful and economical process for the direct amination of 2-phenylpyridine derivatives. This process requires cheap and commercially available copper triflate and works for a variety of different 2-phenylpyridine derivatives. Additionally, the process also works with a bis-sachhrain iodane and thus future endeavors aim to synthesize such novel iodanes with various other amine sources in order to further extend this process. Future work in our laboratory will focus on studying the mechanistic insights and application of this method to heterocyclic intermediates.
Flash chromatography was performed on Silicycle silica gel (60Å, 40-63 µm). All reagents were stored under an inert atmosphere before use.

Instrumentation
Microwave reactions were carried out in a CEM Discover microwave. Flash chromatography was performed using CombiFlash®Rf 200. GC/MS analysis was carried out on an Agilent Technologies 6890 GC system fixed with a 5973 mass selective detector. NMR spectrum were acquired using a Bruker Avance 300MHz spectrometer.

Synthesis of phenyliodine(III) bis(saccharin)Iodane (2a)
A mixture of (0.27 g, 0.633 mmol) phenyliodine(III) bis(trifluoroacetate) and sodium saccharin (0.26 g, 1.267 mmol in 100 mL of acetonitrile is stirred at 40 0 C in an oil bath for 12 h. The white precipitate is collected, washed with acetonitrile and Initial Optimization Studies Table 4. Initial Optimization Studies

Competition experiment
To a solution of equimolar amounts of Py-Ar 1 -H (1eq.) and Py-Ar 2 -H (1eq.) in 1,2dichloroethane ( 4 mL) was added the iodane 2 (2.5 eq) and Cu(OTf) 2 ( 1 eq.). The reaction was mixture was stirred for the 48 h at 80 o C in an oil bath. The reaction was then cooled, and an aliquot was removed and analyzed by GC/MS.

Calibration Curve
The pure product 5 was isolated by column chromatography. Known Concentrations in ppm for (1-100% yield) were prepared in DCE. The plot of various concentrations against their area under the curve from GC-MS spectrum generates the calibration curve.

Introduction
Botulinum neurotoxin serotype A (BoNTA) is a protein produced by a sporeforming bacteria called Clostridium botulinum. It is the world's most poisonous protein having a median lethal dose of approximately 1 ng/Kg. 1 The toxin is associated with numerous food-borne illnesses and is a potential bioweapon.
Consequently, the synthesis of small molecule inhibitors of BoNTA is of high importance. 2,3 Previous work has shown that devleopment of BoNTA inhibitors is particularly challenging because the enzyme-substrate interface is unusually large.

Results and Discussion
In general, two synthetic strategies can be envisioned for the synthesis of these aryl Contrastingly, Itahara developed a method for oxidatively coupling heteroarenes such as indoles, pyrroles and furans with benzene in the early 1980's. [8][9][10] These original reactions required superstoichiometric amounts of palladium acetate.
We and others have recently developed catalytic methods for directly arylating heterocycles involving either one or two C-H functionalizations. [11][12][13][14][15][16] These methods offer the benefit of greater step economy by using readily available hydrocarbon starting materials. 17,18 Herein, we present a novel method for synthesizing inhibitor 1 using two key palladium catalyzed C-H activation steps.
We have recently shown that electron-rich aromatic heterocycles and benzenederived arenes can be oxidatively cross-coupled by palladium catalysis. For example, both we and the Fagnou group have shown that N-acetylindoles can be regioselectively coupled with benzene at either their 2-or 3-position depending on the nature of the oxidant chosen for the reaction. [19][20][21][22][23] We have extended this methodology to include N-alkylindoles, a considerably more challenging class of compounds due to their tendency to decompose in the oxidative reaction conditions. In particular, we have found that N-alkylindoles bearing electronwithdrawing groups provide high levels of regioselectivity, favoring the 2-arylproduct. 16 Consequently, 1 appeared to be an ideal target for the application of our oxidative coupling technology.

Scheme 2 Oxidative arylation of N-alkyl protected indole
However, selective oxidative arylation of the SEM-protected indole 15, 24 using catalytic palladium acetate (PdOAc 2 ) and silver acetate (AgOAc) as an oxidant, afforded 16 in a good yield with a 9:1 regioselectivity favoring arylation at the indole's 2-position, and these regiomers could be separated via flash chromatography. The optimal conditions contained a slight excess of AgOAc (3 equiv) compared to pivalic acid (PivOH, 2.5 equiv), which likely prevents the acidpromoted oxidative decomposition of the indole substrate and products (Scheme 3). 19 Alkaline hydrolysis of 16 afforded the free acid 17 in a good yield. In situ conversion of 17 to the acid chloride, followed by Friedel-Crafts acylation, as described by Pang,4 was attempted, but the desired ketone was not observed. Later, the acid chloride was reacted with the anion of methyl-3-thiopheneacetate to afford the desired product 18 in a modest 38% isolated yield. The isomer which is coupled at the thiophene's 2-position (proximal to the ester), was isolated as a minor product (6% yield). We attribute the selectivity favoring the desired product to steric factors (Scheme 3).
In an effort to expand our methodology to form biaryl C-C bonds by activating two C-H bonds, we initially attempted to oxidatively couple benzene to thiophene 18 using our earlier optimization studies as a guide, 19,22 but the starting material failed to convert. This may be the result of the ability of the thiophene to deactivate the palladium catalyst via coordination of its sulfur atom. However, using Mori's conditions, the C-H bond of the thiophene 18 was selectively arylated using both palladium and silver catalysts and iodobenzene as the arene source. 25 The SEM protecting group of compound 19 was cleaved by treatment with tetrabutylammonium fluoride (TBAF) in DMF to afford 20 and subsequent Nalkylation was carried out with N-(4-bromobutyl)-phthalimide in the presence of potassium tert-butoxide to afford the N-alkylindole 21. Finally, the BoNTA inhibitor 1 was synthesized following the procedure of Pang by treating 21 with excess hydroxylamine, which simultaneously converts the methyl ester and phthalimide to hydroxamic acid and amine, respectively (Scheme 3). 4

Conclusion
In conclusion, we have developed a novel synthetic route for the synthesis of an inhibitor of botulinum neurotoxin serotype A (BoNTA). The 4.6% overall yield is nearly identical to that which was reported by Pang, but the step economy of the overall synthesis has been increased due to the incorporation of reactions which do not rely on prefunctionalized starting materials; rather the biaryl C-C bonds were directly formed from the C-H bonds of simple arenes and heteroarenes. It should also be noted that the C-H functionalizing reactions are two of the highest yielding steps in this synthesis. While our initial attempts to form biaryl C-C bonds with thiophene-type substrates such as 18 were not successful, efforts in our laboratory are currently focused on expanding our oxidative coupling methods to encompass these and other classes of heteroarenes for application to the synthesis of more high-value targets.

Synthesis of methyl-1-(4-(1,3-dioxoisoindolin-2-yl)butyl)-1H-indole-6-carboxylate (13)
To a solution of potassium tert-butoxide (0.143g, 1.28mmol) in 3mL of DMF at 0 o C was added a solution of methylindole-6-carboxylate (0.1g, 0.57mmol) in 1mL of DMF and the reaction mixture is stirred for 15min. After stirring for 15mins add a solution of N-(4-bromobutyl)-phthalimide (0.264g, 0.93mmol) in 1mL DMF and stirring is continued for 6h. The mixture was diluted with 40mL of water and extracted with three 40mL portions of ethyl acetate. The combined organic extracts were washed with two portions of 40mL of water and dried with anhydrous magnesium sulfate.

Synthesis of methyl-5-(2-phenyl-1-[2-(trimethylsilyl)-ethoxymethyl]-indole-6carbonyl)thiophen-3-yl)acetate (18)
A magnetically stirred solution of 2-phenyl-1-[2-(trimethylsilyl)-ethoxymethyl]indole-6-carboxylicacid (0.560g, 1.52mmol) and oxalyl chloride (0.967g, 7.61mmol) in 10mL of dry dichloromethane was stirred for 2h at room temperature. The solvent was removed at reduced pressure, reaction mixture was re-diluted in 10mL of dry THF and cooled at -78 0 C. The cooled reaction mixture is transferred using a canula to a magnetically stirred solution of methyl-3-thiopheneacetate (0.261g, 1.67mmol) and nbutyl lithium (0.312g, 5.01mmol) at -78 0 C.After stirring the reaction mixture for 2h, the solvent was removed under reduced pressure and the mixture was diluted with 40mL of water and extracted with three 60mL portions of ethyl acetate. The combined organic extracts were washed with two portions of 40mL of water and dried with anhydrous magnesium sulfate. After filtration, the solvent is removed at reduced pressure and crude product was purified by column chromatography to give pure 18 0.292g (38%).