Recent Advances in the C–H-Functionalization of the Distal Positions in Pyridines and Quinolines

Abstract

This review summarizes recent developments in the C–H-functionalization of the distal positions of pyridines, quinolines and related azaheterocycles. While the functionalization of the C2 position has been known for a long time and is facilitated by the proximity to N1, regioselective reactions in the distal positions are more difficult to achieve and have only emerged in the last decade. Recent advances in the transition metal-catalyzed distal C–H-functionalization of these synthetically-important azaheterocycles are discussed in detail, with the focus on the scope, site-selectivity and mechanistic aspects of the reactions.

Graphical Abstract

1. Introduction

C–H-functionalization has emerged as a powerful tool for the synthesis of heterocyclic compounds.¹ However, with numerous C–H-bonds simultaneously available for activation, site-selective functionalization of complex molecules remains a significant challenge.² In pyridines, quinolines, and other structurally-related N-heterocycles, the C2 position can be readily functionalized using a variety of transition metal-catalyzed reactions with Pd,^3,4 Cu,⁵ Ni,⁶ Rh,⁷ Ru,⁸ Fe,⁹ and Ag¹⁰ as catalysts, as well as by means of non-catalytic approaches.¹¹ The facility of the C2–H-functionalization is due to the favourable electronic factors, e.g. increased acidity of the C2–H bond and increased electrophilic character of the C2=N moiety; and due to the proximity of the nitrogen atom (or oxygen in the N-oxides) that can serve as a directing group.¹² Other positions in pyridines (C3, C4) and quinolines (C3–C8) have been less readily accessible, with far fewer reports of regioselective C–H-functionalization reactions. Recently, significant progress has been made in the area of the distal C–H-functionalization of azines. While some positions, e.g. C5, C6 and C7 in quinolines, have by and large remained inaccessible by means of catalytic techniques based on C–H-activation, other positions, e.g. C3, C4 and C8 have become a focal point of intense research. In the past several years a number of efficient synthetic techniques that exploit combinations of intrinsic (steric and electronic), proximal (directing groups) and extrinsic (influence of ligands and solvents) regioselectivity factors have been reported for these positions.

Pyridines, quinolines and related azines are important structural motifs of natural products,¹³ pharmaceuticals,¹⁴ advanced materials,¹⁵ catalysts¹⁶ and ligands.¹⁷ Catalytic C–H-functionalization in the distal positions of azines can significantly simplify access to these compounds and their structural analogues. Examples will be presented in such a way, as to allow the reader to view the progression of the area of regioselective C–H functionalization of azines through the course of time. This review focuses on the recent advances in the C–H-functionalization of the distal positions of pyridines and quinolines and related azines. Both directed, as well as sterically- and electronically-controlled catalytic processes are discussed from the mechanistic and synthetic perspectives. In those cases, where the synthetic scope of a particular method includes other classes of (hetero)aromatic compounds, these details are briefly discussed as well.

2. C3–H-Functionalization

In 2002, Ishiyama, Miyaura and co-workers disclosed two examples of borylation of heteroarenes catalyzed by [IrCl(cod)]₂/4,4’-di-tert-butyl-2,2’-bipyridine (dtbpy) in octane at 100 °C.¹⁸ The borylation of quinoline (1) and pyridine (3) in octane was sluggish at 80 °C, but proceeded smoothly at 100 °C. Excess quinoline (10 equiv.) was required to provide 3-borylated product 2 in 84 % yield with >99 % regioselectivity. This report of monoborylation stands out, as a number of other reported techniques provide the bis-borylation product as 3,7-bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)quinoline and 3,6-bis(4,4,5,5-tetramethyl-1,3,2-dioxaborolan-2-yl)quinoline.^19,20 Pyridine (3) (2 equiv.) gave a mixture of 3- and 4-regioisomers in a ratio of 67 : 33, respectively with an overall yield of 42 %. 2 examples are given of the boronate esters in 42 and 84 % yield (Scheme 1).

Scheme 1.

Ir-catalyzed C3-borylation of quinoline and pyridine.

In 2005, in the course of studies of the sterically-directed borylation of cyanoarenes, Smith and co-workers provided an example of borylation of the cyano-substituted pyridine 5.²¹ The ortho-selective borylation proceeded with [Ir(cod)(OMe)]₂/dtbpy as a catalyst in THF at 25 °C. Unlike directed lithiation methods that are regioselective due to electronic or directing properties of pendant functional groups, the Ir-catalyzed borylation is sterically-controlled. In the case of 5-bromo-2-cyanopyridine (5) the reaction yielded a 67:33 ratio of C3/C4 borylation products 6 and 7 in an overall 81% yield (Scheme 2). When the same reaction was attempted with 2-bromo-5-cyanopyridine (8), no borylated product was observed (Scheme 2).

Scheme 2.

Sterically controlled C3/C4-borylation.

In 2007, Hartwig presented a one-pot, two-step synthesis of 3- and 4-haloarenes and 3- and 4-halopyridines that utilized a C3-selective borylation catalyzed by [Ir(cod)(OMe)]₂/dtbpy.²² As this is a sterically-driven reaction, C3/C4 selectivity was very high with 2,6-dimethylpyridine and 3-methylpyridine (9). The borylation occurred meta to the bulky group (Scheme 3). Copper(II) chloride and bromide (3–3.5 equiv.) were then added to effect the halodeboronation.

Scheme 3.

C3-Halogenation of pyridines and the synthesis of altinicline precursor 12.

The method affords meta- and para-haloheteroarenes. Hartwig applied the Ir-catalyzed arene borylation to the synthesis of an intermediate en route to altinicline,²³ a selective nicotinic acetylcholine receptor agonist, previously prepared in the enantiomerically pure form in 5 steps in an overall yield of 32 %.²⁴ The improved synthesis begins with the Ir-catalyzed C3-borylation of the pyridine core that is followed by the halogenation with copper(II) bromide to give the drug candidate in four steps with an overall yield of 75% by intermediate 12 (Scheme 3).²² Advantageously, the meta-borylation/halodeboronation procedure allows for a streamlined and directing group-free synthesis of highly functionalized (hetero)arenes.

Although steric considerations are often invoked to explain C3-selectivity of C–H-borylation of azines, recent mechanistic and computational studies by Hartwig and Larsen suggest that the C2-pathway is disfavored due to the combination of the instability of the 2-borylazines under the reaction conditions, and the higher energy of the C2-pathway.²⁵

In 2010 Yu and co-workers reported a Pd-catalyzed C3/C4 arylation of a number of nicotinic and isonicotinic acid derivatives using an amide directing group.²⁶ No reaction was observed for the substrates that did not have the directing amide groups, e.g. isonicotinic acid. N-Phenyl isonicotinamide (13) gave an excellent overall yield, with low selectivity for monoarylated product 14. Conversely, N-3,5-dimethylphenyl amide increased the ratio of mono- and bisarylated products, but lowered the overall yield.²⁶ The divergent outcomes observed with the two directing groups can be exploited in the context of synthesis, e.g. the sterically hindered amide can be used for the monoarylation, whereas the less bulkier directing group can be used, when bis-arylation is desired.

Ortho substituents in the N-aryl amide moiety proved detrimental and led to substantially lower yields and poor selectivity for the monoarylation product. N-Phenyl isonicotinamide was chosen for the arylation of isonicotinic derivatives. Subsequent screening of phosphine ligands identified PCy₂t-Bu as the most effective ligand with cesium carbonate as a base at 130 °C. Various substituted N-phenyl 4-quinolinecarboxamides and N-phenyl isonicotinamides reacted with a variety of bromoarenes, giving rise to the C3-monoarylation products in up to 94% yields (Scheme 4). Isonicotinamide 13, in addition to monoarylation products, also afforded diarylation products in up to 51% yield. N-Phenyl nicotinamide was also successfully arylated, predominantly in C4 position, although C2-arylation products were also observed in up to 9% yield for several bromoarenes. It was noted that only one of the ortho-C–H bonds underwent the arylation, possibly due to the differences in the reactivities of the ortho-C–H bonds.²⁶

Scheme 4.

Pd-catalyzed C3/C4-selective C–H-arylation of isonicotinamides.

In 2010, Snieckus and co-workers disclosed the meta-borylation of N,N-diethyl amides of (hetero)aromatic acids, and N,N-diethyl (hetero)aryl O-carbamates.²⁷ The amide groups in these substrates were found to provide steric shielding for the C–H bonds in the adjacent positions under the conditions of the Ir-catalyzed borylation. Thus, regioselectivity of this method is complimentary to that of the directed ortho metalation (DoM).²⁸ Four pyridine substrates were subjected to the borylation (Scheme 5). It was found that presence of a trimethylsilyl group as a latent regiodirective moiety in the meta-position of N,N-diethyl picolinamide improved the yield of the borylation product from 47% for 18 to 71% for 19.

Scheme 5.

Sterically-controlled meta-borylation of N,N-diethyl pyridinecarboxamides and N,N-diethyl 3-pyridyl carbamate.

A high-yielding meta-borylation of N,N-diethyl 2-bromonicotinamide was also accomplished to obtain 21. The yields of the borylation products 18–21 were 47–84% (Scheme 5). As a proof of concept two heteroarenes (22 and 24) were subjected to borylation that was followed by a Suzuki-Miyaura cross-coupling reaction to give 23 and 25 in 90 and 71% yields (Scheme 6).²⁷

Scheme 6.

Suzuki-Miyaura cross-coupling products.²⁹

In 2010, Sarpong and co-workers reported the total synthesis of (+)-complanadine A using the Ir-catalyzed borylation reaction for the selective C3-borylation of the pyridine ring (Scheme 7).³⁰ The resulting boronate 27 was then subjected to a Suzuki-Miyaura cross-coupling to give the C2–C3’ bipyridine motif found in the natural product.

Scheme 7.

Ir-catalyzed C3-borylation in the total synthesis of (+)-complanadine A.

In 2011 Shi and co-workers described a C3-selective iridium-catalyzed nucleophilic addition of pyridines to aromatic aldehydes (Scheme 8).³¹ Much like the sterically-controlled meta-borylation reported by Hartwig and Ishiyama,^18,22,27 the site-selectivity is determined by the steric factors. Ir complexes were uniquely active, while Mn, Re, Ru, and Rh showed no catalytic activity.

Scheme 8.

Ir₄(CO)₁₂-catalyzed C3-addition of pyridines to aldehydes.

Ir₄(CO)₁₂ proved to be superior to other Ir sources. Bidentate aromatic nitrogenous ligands significantly improved the yield, with 1,10-phenanthroline (phen) emerging as a ligand of choice. The reaction was carried out in benzene in the presence of triethylsilane at 135 °C with a variety of substituted aromatic aldehydes. Halogen-substituted benzaldehydes, heteroaromatic aldehydes, and electron-deficient aromatic aldehydes gave good yields. Electron-rich aromatic aldehydes afforded addition products in slightly lower yields, while <10% product was formed from n-hexanal. Various meta-substituted pyridines were found suitable, as well as isoquinoline (C4-addition) and quinoline (C3-addition). A detailed study was carried out to elucidate the reaction mechanism (Schemes 9). No reaction occurred in the absence of trialkylsilanes, whereas in the absence of benzaldehyde C–H-silylation was observed in low yields. In an attempt to increase the yield of the C–H-silylation, norbornene and tert-butylethylene were added, however, these hydrogen acceptors did not increase the yield of silylated product 30.

Scheme 9.

Mechanistic studies of the Ir₄(CO)₁₂-catalyzed C3-addition of pyridines to aldehydes.

3-Phenyl-5-(triethylsilyl)pyridine (31) did not undergo the nucleophilic addition to benzaldehyde either directly or in the presence of the iridium catalyst. Upon subjecting 3-(triethylsilyl)pyridine (32) to the reaction under the optimized conditions the desired coupling product 33 was obtained, however, no desilylation or displacement of the silyl group by the aldehyde occurred, as evidenced by the absence of silyl ethers 34 and 35 in the mixture of products. These results indicated that a two-step mechanism involving C–H-silylation followed by the addition to the aldehyde is unlikely. No D/H scrambling was observed for 2-deuterated 3-phenylpyridne (D₁-29), suggesting that the migration of iridium from C2 to C3 is unlikely as well (Scheme 9). Additional experiments suggested that a silyl iridium complex could be an intermediate in the catalytic cycle (Scheme 10). Thus, the following mechanism was proposed: silyl iridium complex 36 oxidatively adds to the pyridine C3–H bond to produce intermediate 38. This step is then followed by an aldehyde C=O insertion into the Ir–Si bond to produce 39.

Scheme 10.

Proposed mechanism.

Subsequent reductive elimination yields product 40 and iridium hydride intermediate 41 that regenerates the catalyst upon reaction with triethylsilane. 25 examples were reported with yields up to 78%. A potential extension of this work, which has yet to be explored, is an asymmetric variant of this reaction, possibly using chiral chelating N-ligands.

In 2011, Yu and co-workers described a palladium-catalyzed C3-selective olefination of pyridine.³² It was envisioned that, a bidentate N,N-ligand can enhance the rate of the ligand exchange on Pd due to the trans-effect (Scheme 12).³³ While the equilibrium between π-complex 43, and σ-complex 44 is shifted to the right, fast ligand exchange can be beneficial for the turnover even if the concentration of the reactive π-complex 43 is low. It can then react with the available olefin to produce the desired product 42 (Scheme 11). The equilibrium between 43 and 44 was also envisioned to be influenced by the solvent, as well as the ligands and counterions.

Scheme 12.

Equilibrium between σ- and π-complexes of Pd with pyridine.

Scheme 11.

Pd-catalyzed C3-alkenylation of azines.

In order to improve the reactivity, and shift the equilibrium in favour of the desired π-coordination of pyridine to the Pd, a number of bidentate ligands were screened with 1,10-phenanthroline (phen) giving the best results. Pd(OAc)₂ proved to be the Pd source of choice. Under the optimized conditions, the reaction was carried out with silver carbonate as a base in DMF at 140 °C. The C3-alkenylation is applicable to substituted pyridines, as well as quinoline and pyrimidine.

Both electron-withdrawing and electron-donating groups in the pyridine, e.g. halogens, methoxy, methyl, and trifluoromethyl groups, are well-tolerated under the reaction conditions. While C2/C6 substituents retarded the addition of the olefin to the C3 position, higher yields were not always accompanied by higher site-selectivity. C4 substitution of the pyridyl unit also decreased the yields, presumably due to the steric hindrance around the C3 position.

The reaction is primarily selective for the C3 position, although some concomitant C2/C4-alkenylation was observed in up to 33% yield. Monosubstituted olefin coupling partners containing ester, amide, acetyl, and aryl groups gave acceptable yields, while internal olefins gave poor yields (<15%). The yields of the C3- and C4-alkenylpyridines were typically in the range of 27–73%. A kinetic isotope study was performed with D₅-pyridine (D₅-3) and pyridine (3). The resulting product showed a k_H/k_D value of 4.0. The authors postulate that the mechanism involves a Pd-mediated C–H bond cleavage, rather than a Lewis acid mediated Friedel-Crafts type reaction (Scheme 13).

Scheme 13.

Kinetic isotope effect study.

In 2011 Sames and co-workers disclosed a Pd-catalyzed regioselective arylation of electron-deficient pyridines (Scheme 14).³⁴ After optimization of the reaction conditions it was found that Pd(OAc)₂/n-BuAd₂P, Cs₂CO₃, Ag₂CO₃, and pivalic acid in toluene at 120 °C provided the highest selectivity and yield. Addition of Ag₂CO₃ was necessary to achieve high regioselectivity. Studies performed with sub- and superstoichiometric amounts of the silver salt found that 1 equiv. of Ag₂CO₃ was optimal for the yield and site-selectivity. With substoichiometric Ag₂CO₃ the regioselectivity eroded, while superstoichiometric amounts of Ag₂CO₃ did not improve the reaction performance. Cs₂CO₃ was also required as a base, with lower yields observed for K₂CO₃. Both bromo- and iodoarenes were suitable electrophiles under the cross-coupling conditions. No reaction was observed with chloroarenes suggesting that the chloro group can be further modified at a later stage of synthesis. Both electron-rich and electron-deficient haloarenes afforded coupling products in good yields. It was shown that the structure of the carboxylic acid additive should be further optimized depending on the electron-withdrawing group (EWG) used.

Scheme 14.

Pd-catalyzed C3/C4-arylation of electron poor-pyridines.

Bulkier, more soluble acids tend to provide higher selectivity and improved yields. In general, an electron-withdrawing group in the C4 position favours a smooth C3-addition, while an EWG in the C3 position directs the reaction in the C4 position. Most substrates had either a bulky group to prevent bis-addition in the C3 and C5 position, or employed haloarene as a limiting reagent. 4-Cyanopyridines were not compatible with silver carbonate, presumably, due to coordination of the cyano group with the silver salts.

This limitation was overcome by excluding Ag₂CO₃ from the reaction mixture. Similarly, fluorinated and chlorinated heteroarenes were arylated in good to moderate yields, with 2,2-dimethylhexanoic acid used in place of pivalic acid. While the mechanism of the reaction awaits further investigation, H/D exchange experiments hint to the role of Ag₂CO₃ as a Lewis base. 33 examples were reported with yields in the range of 15–88% (Scheme 14).

In 2011, Yu and co-workers published a study of non-directed C3-arylation³⁵ (Scheme 15) based on the Pd^II/phen catalytic system that they disclosed earlier that year. By adapting their previously reported C–H alkenylation conditions,³² formation of the C3-arylated pyridine was observed in an 11% yield. Further optimization identified Cs₂CO₃ as a base of choice that provided a 74% yield and an outstanding C3 selectivity (C3/C4/C2 ratio 25/1/1).

Scheme 15.

Non-directed Pd/phen-catalyzed C3–H-arylation of pyridines.

The optimal catalytic system was found to be formed with 15 mol% of 1,10-phenanthroline as a bidentate ligand and 5 mol% of Pd(OAc)₂. It was noted that a 3:1 ratio of ligand to palladium increased the yield by 19%. Excess pyridine substrate (up to 75 equiv.) was required to achieve high yields.

A decrease in the amount of pyridine from the optimal 75 equiv. to 6 equiv. resulted in significantly lower yields (a drop from 92% to 18%, respectively). Iodo- and bromoarenes were found to be suitable electrophiles, while aryl chlorides were unreactive. The reaction tolerated both electron-withdrawing and electron-donating substituents in the haloarene in the para, meta, and ortho positions, including a redox-active thiomethoxy group. The pyridine core can accommodate substituents in the ortho and meta positions. Quinoline was found to provide the arylated product in a 65% yield and a 3:0:1 (C3/C4/C2) selectivity. A primary kinetic isotope effect of 4.2 was obtained. This is higher than that for a typical Fujiwara-type electrophilic palladation (Scheme 16), pointing to a metalation/deprotonation reaction en route to C3-Pd species (Martinez pathway).³⁶ This suggests that an S_EAr mechanism involving a rate-limiting deprotonation cannot be ruled out (Scheme 16). A competition reaction between pyridine (3) and benzene (52) showed that pyridine is 3.26 times more reactive than benzene (Scheme 16).

Scheme 16.

KIE studies and the plausible mechanism of the Pd-catalyzed C3–H-arylation of pyridines.

The authors postulate that the coordination of the pyridine to the Pd center allows it to have an effectively higher molarity near Pd. In order to show the utility of the methodology a 4-step synthesis of the preclinical drug preclamol (63) from pyridine was carried out (Scheme 17).³⁷ Under the optimized conditions, a 70% yield of the C3-arylated pyridine 60 was obtained with a 19:1:1 C3/C2/C4 selectivity.

Scheme 17.

Synthesis of rac-preclamol. Reagents and conditions: (a) 1-bromopropane, CH₃CN, 110 °C; (b) PtO₂, MeOH, H₂ (60 psi), r.t.; (c) HBr in AcOH, reflux.

Propylation of the pyridine ring afforded pyridinium salt 61 that was further reduced to piperidine 62. Finally, demethylation gave rise to racemic preclamol (63) in 4 steps and in 69% overall yield.

In 2013, Chatani and Aihara reported the Ru-catalyzed direct arylation of C–H bonds in aromatic amides.³⁸ Only one example is pertinent to this review, that of arylation of N-(quinolin-8-yl)quinoline-4-carboxamide (64) in 86% yield (Scheme 18). The method called for use of [RuCl₂(p-cymene)]₂ or [Ru(OAc)₂(p-cymene)]₂ (5–10 mol%) as a catalyst and PPh₃ (40 mol%) as a ligand, in the presence of Na₂CO₃ (2 equiv.), and PhBr (1.2 equiv.) in PhMe at 140 °C. Aryl bromides and iodides proved to be suitable electrophilic coupling partners under the optimized reaction conditions.

Scheme 18.

Ru-catalyzed arylation of C4-substituted quinoline.

In 2013, Shi and co-workers reported the Rh-catalyzed olefination of azines assisted by an amide directing group (Scheme 19).³⁹ The study began with evaluation of the reaction conditions reported by Glorius for the Rh(III)-catalyzed oxidative Heck reaction of acetanilide with styrene.⁴⁰ While no product was obtained with N-(pyridin-2-yl)acetamide (66) and styrene, a reaction of 66 with ethyl acrylate (1.5 equiv.) afforded product 67 in a 6% yield under the same conditions.

Scheme 19.

Rh(III)-catalyzed olefination of azines.

Further optimization of the reaction conditions (DCE, Cu(OAc)₂, [RhCp*Cl₂]₂ (5 mol%) and AgSbF₆ (20 mol%) at 120 °C) led to a complete conversion of 66 to 67. Other additives such as Ag₂CO₃ or Zn(OAc)₂ were less effective. Acrylates were found to provide the corresponding olefinated products in good yields. Substituted 2-aminopyridines are compatible with a number of functional groups, e.g. halides, methoxy, ester, and trifluoromethyl, in the C4/C5/C6 positions.

The functionalization generally occurred in the C3 position, except for 2,6-N-(pyridin-2-yl)bisacetamide that provided a C3/C5 bis-substituted product 68 in a 98% yield with ethyl acrylate. A total of 38 examples were reported in up to 100% yield. The alkenylation products were then transformed to a number of analogues. For example, acidic cleavage of the pivalamide in 69 was followed by the diazotization and subsequent nucleophilic displacement to give the fluorinated pyridine 70. The methodology was then applied to the formal synthesis of pharmaceutically important compounds 71 and 72 (Scheme 20). Thus, a multigram-scale synthesis of 7-methoxy-3,4-dihydro-1,8-naphthyridin-2(1H)-one (73), – a precursor to the antibacterial agent 71, – can be achieved in three steps in 76% overall yield from pivalamide 74. The alkenylation of 74 was successfully carried out with 0.1 mol % catalyst resulting in a turnover number (TON) of 840. The intermediate 75 was then converted to 76 that produced dihydronaphthyridinone 73 upon exposure to HCl, thus completing the formal synthesis of the antibacterial agent 71.⁴¹

Scheme 20.

Application of the amide-directed dehydrogenative C3–H-alkenylation of pyridines to the synthesis of intermediates en route to medicinally important compounds 71 and 72.

The next synthetic target, – indole naphthyridinone 72, – has shown significant activity in animal studies of the reduction in bacteria fatty acids production through inhibition of bacterial enoyl-ACP reductase (FabI). This interruption halts bacterial biosynthesis and is a promising target for novel antibacterial agents. Thus, alkenylation product 77 was hydrolyzed under acidic conditions to give aminopyridine 78 in 80% yield. Ni-mediated conjugate reduction afforded saturated ester 79 that underwent a base-mediated cyclization to give lactam 80 that can be further converted to indole naphthyridinone 72.⁴²

In 2013, Miura, Hirano, and co-workers disclosed the 8-aminoquinoline-directed dehydrogenative cross-coupling of (hetero)arenes and 1,3-azoles.⁴³ The scope of the method includes arenes and heteroarenes that are not covered by this review, as well as an example of isonicotinamide derivative 82 that was obtained in 34% yield (Scheme 21).

Scheme 21.

Copper-mediated dehydrogenative cross-coupling of isonicotinamide 81 with benzoxazole.

To gain insights into the mechanism, experiments involving deuterium-labeled benzamide and benzoxazole were performed. When a mixture of benzoxazole (83) with the C2-deuteromer (D₁-83) was reacted with benzamide 85, under the standard conditions, a primary KIE value of 1.0 was obtained. On the other hand, a KIE value of 2.0 was observed for the ortho positions in benzamide 85. This KIE data may suggest that the rate-limiting step is the C–H bond cleavage in benzamide 85 (Scheme 22).⁴⁴

Scheme 22.

Primary kinetic isotope effect studies.

H/D exchange experiments were performed with D₅-benzamide (D₅-85) and D₁-benzoxazole (D₁-83) independently and together (Scheme 23). 57% H/D exchange in D₁-benzoxazole (D₁-83) in the absence of benzamide 85 was observed, while addition of 85 led to a 35% H incorporation (with 43% of the coupling product formed) in 10 min. Based on these results, the C–H activation of the benzoxazole was proposed to be reversible. When benzamide D₅-85 was subjected to the reaction under the standard conditions in the absence of benzoxazole 83 a 14% H incorporation was observed, while addition of benzoxazole 83 led to a 2% H incorporation (with 23% of the coupling product formed) in 10 min. These results suggest that the C–H-activation is irreversible. These two data sets show that benzoxazole should undergo C–H-activation first, followed by the C–H-activation of the benzamide (Scheme 24). The proposed mechanism begins with an acetate ligand-assisted cupration to generate azolylcopper intermediate 86. Subsequent ligand exchange forms complex 87.

Scheme 23.

H/D exchange experiments with benzamide D₅-85 and D₁-benzoaxazole (D₁-83).

Scheme 24.

Plausible mechanism of the Cu-mediated dehydrogenative cross-coupling.

A Cu(OAc)₂-promoted disproportionation induces the C–H cupration of the benzene ring (complex 89), followed by the reductive elimination of Cu(I) to give the desired product 84 (Scheme 24).⁴⁵

In 2013, Daugulis and co-workers disclosed the 8-aminoquinoline-directed, copper-catalyzed fluorination of (hetero)arene C–H bonds.⁴⁶ The optimal conditions for monofluorination involve CuI as a catalyst, N-methylmorpholine N-oxide (NMO) as an oxidant, and DMF as solvent, with AgF as the fluoride source at 50–125°C for 30–120 min.

Di-fluorination can be achieved if the amounts of the reagents are doubled at 75–105 °C. Two heterocycles pertinent to this review are reported in 62% and 61% yields for di- and mono-fluorinated products 90 and 91, respectively (Scheme 5).

Daugulis and Roane subsequently disclosed the copper-catalyzed etherification of arene C–H bonds (Scheme 26).⁴⁷ Previously reported amination conditions⁴⁸ were used first to give a 21% yield observed of the reaction between 8-aminoquinoline 3-trifluoromethylbenzamide with 4-tert-butylphenol. Upon changing the oxidant to NMO/O₂ a yield of 67% was obtained, with an 88% yield reported using (CuOH)₂CO₃ as a catalyst, K₂CO₃ as base, and air as oxidant.

Scheme 26.

Cu-catalyzed C–H-etherification of isonicotinamides.

The method can be used with a number of aliphatic alcohols in addition to substituted phenols. Mono- and bis-substitution with 4-tert-butylphenol at C3 and C5 positions in isonicotinamide 81 was reported (Scheme 26). The yields of the mono- and bis-addition products 92 and 93 were 51 and 47%, respectively.

In 2013, Su and co-workers disclosed the Rh(III)-catalyzed amide-directed cross coupling of pyridines with other heteroarenes (Scheme 27).⁴⁹ Initial screening began by evaluating the reaction between N-phenyl isonicotinamide (94) and 2-methylthiophene. [RhCp*Cl₂]₂ (2.5 mol%), AgSbF₆ (10 mol%), CsOPiv (1.5 equiv.) and Cu(OAc)₂ (2 equiv.) in 1,4-dioxane at 130 °C furnished product 95 in a 6% yield. Under the optimized conditions with K₂HPO₄ (1.5 equiv.) as a base and with 1.5 mol% of the catalyst, coupling product 95 was obtained in 70 % yield.

Scheme 27.

C3-selective heteroarylation of disubstituted pyridines.

The influence of the directing group of the amide was studied in an attempt to increase the yield. Sterically-hindered or electron-withdrawing N-polyfluorophenyl amides, N-4-methoxyphenyl amide, and alkyl amides afforded trace amounts of the cross-coupling product. Other aryl amides, e.g. N-4-fluorophenyl amide, N-4-methoxyphenyl amide, N-3,5-dimethylphenyl amide, furnished the target compounds in moderate yields. The role of the substituents in the directing group may be complex and remains to be clarified in the future.

A variety of thiophenes bearing both electron-donating and withdrawing substituents, including alkyl, aryl, halo, cyano, and ester groups, can be employed as coupling partners, as well as benzothiophene, benzofuran and 2-methylthiazole. Disubstituted pyridines were also used (Scheme 28).

Scheme 28.

Rh(III)-catalyzed dehydrogenative C2/C3-heteroarylation.

A variety of substituted N-phenyl pyridinecarboxamides, as well as N-phenylquinoline-3-carboxamide were successfully reacted with substituted thiophenes to give the C3-heteroarylation products (e.g. 100–104) in up to 73% yield and with exclusive C3-regioselectivity. A notable practical feature of this method is that the formation of the monoarylation products is not accompanied by concomitant diarylation. Based on the previous studies by Fagnou⁵⁰ and Ellman,⁵¹ the following mechanism was proposed: coordination of the Rh(III) species with N-phenylamide is followed by formation of arylrhodium(III) intermediate 106, which, upon reaction with 2-methylthiophene, gives rise to complex 107. Subsequent reductive elimination produces rhodium(I) species 108 and cross-coupling product 95. Finally, oxidation of rhodium(I) species 108 by copper (II) regenerates the rhodium(III) catalyst 105 (Scheme 29). ⁴⁹

Scheme 29.

Proposed catalytic cycle.

Later in 2014, Shi and co-workers reported the Rh(III)-catalyzed hydroarylation of alkynes via a directed C–H-functionalization of pyridines.⁵² Previous reports of the C2/C4/C6 alkenylation through the use of Ni(0) catalysts with and without Lewis acids were reported, but are not covered in this review.⁵³ For optimization, Shi and co-workers started with the reaction conditions reported for the oxidative olefination of N-(2-pyridyl)pivalamides (Scheme 19).³⁹ Under these conditions the coupling of N,N-diethyl 5-bromopicolinamide and diphenylacetylene afforded the product of type 111 in 74% yield. Upon addition of 4 equiv. of AcOH the yield increased to 92%. Stoichiometric Cu(OAc)₂ as an additive proved to be important, as its exclusion or use in catalytic amounts resulted in poor yields. Other copper salts (e.g. CuBr₂ or CuSO₄) were less effective. Mn(OAc)₂ and Co(OAc)₂ demonstrated comparable to Cu(OAc)₂ activity, suggesting that they act as Lewis acids.

The method tolerates various substituents in the C4, C5 and C6 positions including halides, methoxycarbonyl, p-methoxyphenyl and acetoxy groups. Electron-withdrawing substituents typically increased the yield, with the opposite result for electron-donating substituents. The alkyne scope was also briefly studied. Both symmetric and unsymmetric alkynes emerged as suitable reaction partners. While, di(p-methoxyphenyl)acetylene yielded alkenylation product 112 as a 1.7:1 mixture of Z and E isomers, most of the other alkynes tested, (e.g. 4-octyne, 1-phenyl-1-butyne, 1-phenyl-1-propyne, and p-diarylacetylenes bearing F, Cl, Me, t-Bu substituents) produced E isomers as major products. 26 examples were reported with overall yields in the range of 30–99% (Scheme 30).

Scheme 30.

Rh(III)-catalyzed alkenylation of picolinamide derivatives.

To probe the mechanism of the hydroarylation a number of experiments were performed. When 3-bromo-N,N-diethylpicolinamide was reacted under the optimized reaction conditions in the absence of the alkyne and with 4 equiv. of AcOD, for 20 min and for 4 h, 5% and 30% deuteration at C3 was observed, respectively. This result suggests a reversible Rh insertion into the C3 position, although a concomitant C6-deuteration (3–6%) was also noted. Primary KIE of 3.1 (k_H/k_D) indicates that the C–H bond cleavage could occur in the rate-limiting step. When 3-deuterated-N,N-diethyl 5-bromopicolinamide (80% D) was reacted with diphenylacetylene, no deuterium incorporation was observed in the product. This result suggests that oxidative addition of the ortho-C–H bond under reported conditions is unlikely. Alkyne activation by a cationic catalyst was eliminated due to the high E-stereoselectivity and the primary KIE data.

The following mechanism was proposed: initial Rh insertion in C3–H to form the five-membered rhodacycle 116 is then followed by the migratory insertion of the alkyne to form the seven-membered rhodacycle 117. Subsequent protodemetalation completes the catalytic cycle (Scheme 31).

Scheme 31.

Plausible mechanism for Rh(III)-catalyzed alkynylation.

Later in 2014, Shi and co-workers reported the oxidative C3–H-olefination of picolinamides with the amide functionality acting as a directing group.⁵⁴ Near-quantitative yields of the coupled products were achieved with N,N-dialkyl-substituted picolinamides. A variety of alkylamide groups (e.g. ethyl, isopropyl, cyclohexyl, benzyl, n-pentyl, and n-butyl) proved to be effective directing groups. In contrast, no reaction was observed with the N-phenylamide and Weinreb amide.

A number of activated alkenes were successfully employed as coupling partners, including acrylate esters, styrene and diethyl vinylphosphonate (Scheme 32). The method tolerates both electron-donating and electron-withdrawing groups in the pyridine ring. Quinoline- and pyrimidinecarboxamides were also found to be suitable substrates, and dual alkenylation can be achieved with a 2,6-pyridinediamide (Scheme 33). The catalyst loading can be further reduced to 0.5 mol%. 31 examples were provided with yields in the range of 20–100%.

Scheme 32.

Rh(III)-catalyzed alkenylation of picolinamides.

Scheme 33.

Products of C3-olefination of various azinecarboxamides.

Mechanistic studies were carried out to clarify the catalytic cycle. A competition reaction between N,N-diethyl 4-methylpicolinamide (130) and N,N-diethyl picolinamide (123) afforded a 4.8:1 mixture of 4-H and 4-Me products 131 and 132 pointing to the importance of the steric interactions. On the other hand, a competition experiment with 6-methoxypicolinamide (133) and N,N-diethyl 6-fluoropicolinamide (134) produced a 3.5:1 mixture of 6-MeO and 6-F products 135 and 136, indicating that the reaction favors more electron-rich substrates. Further, the H/D-exchange was not observed under the standard reaction conditions in the presence of MeOD.

The reaction has a primary KIE (k_H/k_D) of 2.5, indicating that C–H-cleavage could be the rate-determining step (Scheme 34). Based on the above mechanistic studies a catalytic pathway was proposed (Scheme 35). According to the mechanism, initial cyclometalation of the substrate by catalyst 139 is followed by the olefin insertion into the C–Rh bond of five-membered rhodacycle 140. Subsequent β-hydride elimination in the seven-membered intermediate 141 affords intermediate 142, that, upon protodemetalation and oxidation of Rh(I) by Cu(II), releases the coupling product 131 and catalyst 139.

Scheme 34.

Mechanistic studies for the Rh-catalyzed C3-olefination of picolinamides.

Scheme 35.

Proposed mechanism of Rh(III)-catalyzed C3-olefination.

In 2014, Hartwig and Cheng disclosed a Rh/bisphosphine system that efficiently catalyzed silylation of arene C–H bonds, however, a number of functional groups, e.g. heavy halogens and basic heterocyclic moieties, were not compatible with the catalytic system.⁵⁵ A year later the authors reported on a Ir-based catalytic system with broad functional group compatibility and tolerance for basic heterocycles.⁵⁶ The reactions were performed with HSiMe(OSiMe₃)₂ as an inexpensive and uniquely active silylating reagent.

Optimization studies conducted with [Ir(cod)(OMe)]₂ identified 2,4,7-trimethyl-1,10-phenanthroline (tmphen) as a ligand of choice that provided a high yield and high regioselectivity of the silylation. The reactions were carried out at 80–100 °C in THF with cyclohexene (1 equiv.) as a hydrogen acceptor. The Ir-catalyzed silylation procedure showed broad functional group tolerance (e.g. ester, ketone, bromide, iodide, nitrile, and sulfones). Several examples of the silylation of pyridines are reported with yields between 46 and 94%. (Scheme 36). 3-Methylpyridine was silylated in a 59% yield. Mirtazapine, an antidepressant, was also silylated in the meta-position of the pyridine ring in a 46% yield.

Scheme 36.

Ir-catalyzed C3–H-silylation.

Concomitant silylation of the benzene ring (14% yield) was also observed. In addition, 6-methoxyquinoline afforded C3-silyl derivative 145 that was further converted to 3-bromo-6-methoxyquinoline in a 64% yield.

In 2014, Song, Niu and co-workers reported the Cu-mediated alkoxylation of arenes using an N,O-bidentate directing system.⁵⁷ The authors sought to optimize the reaction of 2-benzamidopyridine 1-oxide (149) with ethanol.

This N,O-bidentate directing group was previously used by the same group in aryloxylation using stoichiometric amounts of copper.⁵⁸ Stoichiometric amounts of Cu(OAc)₂ in ethanol provided product 150 in a 21% yield. Inorganic bases were tested (e.g. KOt-Bu, NaOEt, KOH, Cs₂CO₃), with K₂CO₃ emerging as the base of choice. Under the optimized conditions with CuCl (1 equiv.) in a 1:1 mixture of ethanol and pyridine at 130 °C the reaction afforded 150 in an 84% yield. Catalytic amounts of CuCl provided the product in lower yields. The reaction tolerates aliphatic alcohols, and a number of substituents in the arene ring, including iodo, methyl, and chloro groups. In one example, isonicotinamide was reacted with hexafluoroisopropanol (HFIP) to give the C3/C5-ether in 75% combined yield with a 2.6:1 ratio of mono- and bis-etherification products (Scheme 37).

Scheme 37.

Cu-catalyzed C–H-etherification.

In 2014, Yu, Dai, and co-workers reported the use of an amide-tethered oxazoline directing group for the Cu(II)-mediated amination of arenes (Scheme 38).⁵⁹

Scheme 38.

Cu(II)-mediated amination.

Unlike Daugulis's⁶⁰ 8-quinolinamide directing group that forms a five-membered ring incorporating N–M–N, Yu's directing group forms a six-membered bidentate complex.⁶¹

During the optimization stage a reaction between N-(2-(4,5-dihydrooxazol-2-yl)phenyl)benzamide (155) and tosylamide in the presence of a stoichiometric amount of Cu(OAc)₂ and K₂CO₃ (2 equiv.) in DMSO at room temperature afforded the product in 8% yield. The yield was increased to 70% at 80 °C and with Na₂CO₃ as a base. Several isonicotinamide examples are reported with yields in the range of 31–70% (Scheme 38). After completion of the reaction, the amide-tethered oxazoline directing group can be removed with KOH in EtOH at 80 °C.

In 2014, Yu, Dai, and co-workers disclosed the Cu(II)-mediated ortho-C–H-alkynylation of (hetero)arenes with terminal alkynes using the same amide-tethered oxazoline directing group.⁶² The conditions⁵⁹ previously-reported for the amination (Scheme 38) yielded product 157 arising from a reaction of benzamide 155 with (4-methylphenyl)acetylene in 52% yield. Various bases were tested, with NaOAc (1 equiv.) providing coupling product 157 in 85% yield at 60 °C. Excess acetylene (typically 3 equiv.) is used, due to concurrent homocoupling, and stoichiometric amounts of Cu(OAc)₂ were used to achieve good yields. The method tolerates a variety of aliphatic alkynes, a conjugated enyne, and heteroarenes. 4-Pyridine- and quinolinecarboxamides were reported to give the alkynylation products 158–160 in moderate yields (Scheme 39).

Scheme 39.

Cu(II) catalyzed alkynylation with amide-tethered oxazoline directing group.

In 2014, Yu, Dai, and co-workers reported the Cu(II)-catalyzed ortho-C–H-trifluoromethylation of arenes using the Ruppert-Prakash reagent (CF₃TMS).⁶³ Using their previously reported amide-tethered oxazoline directing group,⁵⁹ the authors set out to optimize the trifluoromethylation reaction. Treatment of benzamide 155 with 30 mol% of Cu(OAc)₂, 5 equiv. of TMSCF₃, 1 equiv. of Ag₂CO₃, and 4 equiv. of KF in DMSO at 100 °C gave the coupling product 161 in 37% yield. NaF and CsF provided no increase in yield, however increasing the amounts of Cu(OAc)₂ and Ag₂CO₃ to 1 and 1.5 equiv., respectively, raised the yield to 80% of mono/bis-trifluoromethylation products. Reducing the reaction time from 5 h to 30 min increased the ratio of mono- and bis-trifluoromethylation products to 3.8:1. The method was applied to a number of substituted heteroarenes. 4-Pyridine- and quinolinecarboxamides were reported to give the alkynylation products 162–164 in 40–62% (Scheme 40).

Scheme 40.

Cu(II)-promoted trifluoromethylation.

3-Fluoroisonicotinamide and 4-quinolinecarboxamide afforded the mono-trifluoromethylation product exclusively, while the unsubstituted isonicotinamide gave rise to a ratio of 27:25 mono-/bis-functionalization products. To probe the mechanism, intra- and intermolecular kinetic isotope effects were measured. An intermolecular competition experiment gave a KIE of 3.2, while an intramolecular competition experiment produced a KIE of 3.5 (Scheme 41).

Scheme 41.

Cu(II)-promoted trifluoromethylation.

No discernible effect was observed with the addition of a radical quencher (TEMPO), ruling out a possible radical pathway.

The reaction is believed to proceed through organocopper intermediate 165 that is oxidized to copper(III) species 166. Subsequent transmetalation with TMSCF₃ affords organocopper(III) intermediate 167 that undergoes reductive elimination to yield the desired product 161 and Cu(I).

In 2014, Yu, Dai, and co-workers reported the Cu(II)-catalyzed coupling of aromatic C–H bonds with arylboronates using the same oxazoline directing group.⁶⁴ One pyridine, N- 2-(4,5-dihydrooxazol-2-yl)phenyl isonicotinamide (153), was arylated in the C3 position (meta to the pyridine nitrogen) with PhBpin (168), Cu(OAc)₂, Ag₂O, Na₂CO₃, and KOAc in DMSO at 70 °C, giving rise to product 169 in 26% yield (Scheme 42).

Scheme 42.

Cu-catalyzed arylation of isonicotinamide derivative 153.

A possible mechanism may involve initial copper acetate-mediated C–H-activation en route to 170, followed by disproportionation to yield Cu(III) intermediate 171. Transmetalation with 168 affords intermediate 172 that undergoes reductive elimination to give Cu(I) complex 173. Subsequent protonolysis yields CuOAc and product 169. Finally, oxidation of the Cu(I) species leads to regeneration of the catalyst (Scheme 43).

Scheme 43.

Cu-catalyzed arylation with amide-oxa directing group.

In 2014, Gooßen and co-workers reported the Cu(II)-catalyzed ortho-nitration of arenes and heteroarenes with 8-amidoquinoline as a directing group.⁶⁵ A number of directing groups were tested, with Daugulis's⁶⁶ 8-aminoquinoline bidentate directing group providing the highest yield and selectivity. The reaction conditions included Cu(NO₃)₂ and AgNO₃ in propylene carbonate with N-methylmorpholine N-oxide (NMO) as the oxidant at 50 °C to yield 30% of 2-nitrobenzamide 174. Addition of phosphine ligands improved the yield, with triphenylphosphine emerging as a ligand of choice. The preformed Cu(PPh₃)₂NO₃⁶⁷ provided product 174 in 78% yield. Two N-heterocyclic C3-nitration products 175 and 176 were obtained in 76 and 52% yields (Scheme 44). The kinetic isotope effect of 2.7 was observed, confirming that the C–H bond activation could the rate-determining step. The reaction was also performed with radical scavengers (p-benzoquinone or TEMPO), with an insignificant decrease in yields, suggesting that a radical pathway is unlikely.⁶⁸

Scheme 44.

Cu(II)-mediated nitration.

The following mechanism was proposed, based on the aforementioned experimental data: initial cyclometalation by copper(II) to give organocopper species 177 is followed by oxidation to Cu(III) intermediate 178 and transmetalation to nitrate 179; lastly, reductive elimination furnishes product 175 (Scheme 45).⁶⁹

Scheme 45.

Plausible mechanism of directed C3–H-nitration.

In 2014, Shi and co-workers reported the Cu(II)-mediated C–H-hydroxylation of arenes and heteroarenes with a 2’-(pyridine-2-yl)isopropylamine (PIP) directing group (Scheme 46).⁷⁰ The initial experiment was carried out with N-(2-(pyridin-2-yl-propan-2-yl)benzamide (180), Cu(OAc)₂ (1 equiv.), Ag₂CO₃ (2 equiv.), and NaHCO₃ (1 equiv.) in DMF at 100 °C to give the hydroxylation product 181 in 51% yield. Among the additives tested (e.g. NaHCO₃, KOAc, PhCO₂Na) tetrabutylammonium iodide (TBAI, 2 equiv.) provided the highest yield (90%) within 1 h. The reaction tolerates both arenes and heteroarenes bearing a variety of functional groups (e.g. fluoro, chloro, trifluoromethyl, methoxy, and methyl). Substituted isonicotinamides of type 182 were successfully converted to the hydroxylated products 183 in 53–65% yields (Scheme 46). When 2-chloro- or 2-fluoroisonicotinamides were employed, a mixture of C3- and C5-oxidation products 184 and 185 was obtained, with 3-hydroxy amide 183 being the major product. On the other hand, N-PIP 2-methoxyisonicotinamide afforded 5-hydroxy amide 186 as a sole product.

Scheme 46.

Cu(II)-mediated hydroxylation.

Interestingly when the N-PIP picolinamide was subjected to the oxidation with the aim of obtaining a C3-oxidation product, no reaction occurred. In addition, C2-oxidation was not observed under the reaction conditions. The authors carried out a number of experiments to gain insight into the mechanism. Addition of radical scavengers (1,4-dinitrobenzene, TEMPO, or 1,1-diphenylethylene) had no influence on the yield, eliminating a radical pathway. A KIE of 5.3 was observed, indicating that the C–H bond cleavage is the rate-determining step.

No deuterium incorporation was observed under standard conditions in the presence of D₂O or AcOD. This result eliminates a reversible C–H activation step. The following mechanism was, therefore, put forward: complexation of Cu^II produces chelate 187 that gives rise to organocopper(II) species 188 via a concerted metalation deprotonation pathway. Subsequent oxidation with CuX₂ delivers Cu(III) intermediate 189 that undergoes reductive elimination to give acetate 190. Rapid hydrolysis of the acetate produces ortho-hydroxy amide 181. Alternatively, a disproportionative C–H-activation⁷¹ may yield Cu(III) intermediate 189 directly from chelate 187.

In 2015, Shi and co-workers utilized the PIP directing group in the copper- and silver-mediated ortho-ethynylation of (hetero)aryl C–H bonds with triisopropylsilylacetylene.⁷² Initial experiments with benzamide 180 in the presence of Na₂CO₃ (2 equiv.) and Ag₂CO₃ (1 equiv.) with 50 mol% Cu(OAc)₂ gave alkynylation product 191 in 55% yield. Adding triethylamine (0.8 equiv.) and increasing the amount of Ag₂CO₃ to 1.5 equiv. with 1 equiv. of Na₂CO₃ improved the yield to 80% with only 30 mol% of Cu(OAc)₂. The reaction tolerates a number of functional groups (e.g. chloro, fluoro, methyl, phenyl, and nitro). Isonicotinamides 182 underwent C3-alkynylation (Scheme 47) producing monoacetylenes 192 along with substantial amounts of the C3,C5-dialkynylation products in 50–82% combined yields. The ratio of mono- and dialkynylation products varied from 4 : 1 for isonicotinamide 193 to 0.7 : 1 for 2-chloroisonicotinamide 194.

Scheme 47.

Cu-catalyzed directed alkynylation of isonicotinamides.

In a series of competition experiments electron-deficient substrates were found to be more reactive than electron-rich substrates, ruling out a simple electrophilic aromatic substitution (S_EAr). The reaction proceeded smoothly in the presence of radical scavengers, suggesting that a radical mechanism is unlikely. This data, together with the high values of the intermolecular KIE (k_H/k_D = 3.6) and the intramolecular KIE (k_H/k_D = 7.5), led to the following mechanistic proposal: initial C–H-activation by concerted metalation deprotonation (CMD) is followed by oxidation and transmetalation to generate Cu(III). Lastly, reductive elimination furnishes the desired product.

In 2015, Shi and co-workers reported the Cu(II)-catalyzed methoxylation of (hetero)arenes⁷³ using their bidentate (pyridin-2-yl)isopropylamine (PIP-amine) directing group (Scheme 49). The reaction proceeds with 5 mol% of Cu₂(OH)₂CO₃, with KOCN as base in MeOH at 120 °C, and with air as an oxidant. Other copper salts (e.g. Cu(OAc)₂, CuBr₂ and Cu(OTf)₂) were less efficient even at higher catalyst loadings. Methoxylation of PIP-isonicotinamide 195 afforded a 4:1 mixture of C3-monomethoxylation product 196 and the C3/C5-bismethoxylation product in a 95% yield. The catalytic system tolerates methyl, chloro, and methoxy groups in C2. The 2-methoxy amide 197 afforded a mixture of regioisomers 198 and 199. N-PIP-nicotinamides afforded the corresponding C4-methoxylated products.

Scheme 49.

PIP-amide-directed methoxylation.

3 C4–H-Functionalization

In 1997, Grigg and Savic published a method of C3/C4-directed Ru-catalyzed addition of heteroarenes to alkenes (Scheme 50).⁷⁴ They surmised that an acetyl group could direct the substitution at either C3 or C4 faster than at the C2 position.

Scheme 50.

Ru-catalyzed addition of pyridines to alkenes.

Reactions were carried out with RuH₂CO(PPh₃)₃ (5 mol%) as a catalyst and with a 1:3–4 ratio of heteroarene to alkene. The initial experiment with 3-acetylpyridine (200) and 1-octene gave rise to the desired product 201 in an 11% yield. No C2- or bis-C3/C5-alkenylation products were observed. The low yield was attributed to the competitive double bond isomerization.

This side-reaction was avoided by using alkenes that lack C–H bonds in the allylic positions, e.g. 3,3-dimethylbutane that yielded 50% of the desired product. A number of silylalkenes were also used. With trimethyl(vinyl)silane, addition to both the C4 and C2 was observed. 3-Benzoylpyridine 202 was also reacted with triethoxy(vinyl)silane to give a C4-addition product 203 in 48% yield.

In 2000, Smith and co-workers reported a sterically controlled C4-borylation of 2,6-dimethylpyridine (204) with pinacolborane by dual C–H-activation.⁷⁵ The method was developed for meta-borylation (meta to the two methyl groups) of disubstituted arenes. The reaction with 2,6-dimethylpyridine produced boronate 205 in 41% yield (Scheme 51).

Scheme 51.

Ir-catalyzed C4-selective borylation of 2,6-dimethylpyridine.

Another example was reported in 2002 with 2,6-dichloropyridine (206).⁷⁶ The catalyst used was (Ind)Ir(cod) with bis-1,2-bis(diphenylphosphino)ethane (dppe) as a ligand. The reaction afforded boronate 207 in 69 % yield at 100 °C in cyclohexane under solvent-free conditions (Scheme 52). The methodology subsequently was further expanded to a one-pot borylation/oxidation protocol.⁷⁷

Scheme 52.

Ir-catalyzed C4-selective borylation of 2,6-dimethylpyridine.

Using the previously reported borylation conditions⁷⁶ 206 was successfully borylated at C4 position. Subsequent oxidation by Oxone in acetone afforded the corresponding 4-hydroxypyridine 208 in 64% yield (Scheme 53).

Scheme 53.

Sequential Ir-catalyzed C4–H-borylation and oxidation.

In 2004, Nishida and Tagata disclosed an Ir-catalyzed C4-selective-borylation of 2,6-dichloropyridine with two different ligands (209 and 210, Scheme 54) and two boronate esters.⁷⁸

Scheme 54.

Ligands used for selective borylation.

The isopropyl groups in the ligands 209 and 210 were important for obtaining high yields. Pinacolborane gave higher yields than bis(pinacolato)diboron by ~20%. The yields for C4-borylation of 2,6-dichloropyridine were in the range of 71–93% (Scheme 55).

Scheme 55.

C4-selective-borylation of pyridines.

In 2009, Marder, Steel, and co-workers developed a one-pot borylation/Suzuki-Miyaura cross-coupling procedure based on the group's prior work in this area.^73,19

The key to the success of the one-pot procedure was the discovery of a solvent (MTBE) that was suitable for both the borylation and the cross-coupling reaction. The compatibility of the Ir/dtbpy with MTBE was explained by the steric hindrance that prevents coordination of MTBE to the Ir catalyst. This result allowed for replacement of hexane that is not suitable for the Suzuki-Miyaura reaction. Using this one-pot method, 2,6-dichloropyridine (206) was transformed to methyl 4’-(2,6-dichloropyridin-4-yl)benzoate (211) in a 94% yield (Scheme 56).

Scheme 56.

One-pot reaction with 2,6-dichloropyridine and methyl 4-iodobenzoate.

Marder, Steel, and co-workers reported the microwave-assisted borylation of a number of heteroarenes, as a continuation of their sterically directed borylation.⁷⁹ The use of microwave irradiation allowed for the reduction in reaction time from 2 h to as little as 3 min (Scheme 57).

Scheme 57.

Microwave-assisted Ir-catalyzed C4–H-borylation.

Marder and co-workers also reported a sterically-controlled Ir-catalyzed C4-borylation of 2-phenylpyridine (212) that produced a mixture of C3- and C4-pyridineboronates. The boronates were coupled with 1-iodonapthalene and a Pd catalyst to yield the C3/C4 products 213 in 61% yield over two steps (Scheme 58).⁸⁰

Scheme 58.

Ir-catalyzed borylation/Suzuki-Miyaura coupling of 2-phenylpyridine.

In 2013, Krska, Maleczka, Smith and co-workers reported a Ir-catalyzed C–H borylation of aminopyridines 214 guided by a traceless directing group (Scheme 59).⁸¹ HBpin first engages the amino group to give aminoboronate 215 that undergoes a selective meta- or para-borylation. The resulting aminoboronate 216 is then cleaved hydrolytically to give amine 217, thus acting as a traceless directing group.

Scheme 59.

Intermediates in the Bpin-directed borylation.

The reaction was conducted with 1.2 equiv. of HBpin, 0.5 equiv. of B₂pin₂, 1.5 mol% of [Ir(cod)(OMe)]₂ and 3 mol% of 3,4,7,8-tetramethyl-1,10-phenanthroline (tmphen) in THF at 80 °C. All aminopyridines had at least one C2-substituent, e.g. a methoxy, chloro, trifluoromethyl, or methyl group. C4-Borylation occurred both with 2- and 3-aminopyridines. With 2-aminopyridines the reaction was not directed by the NHBpin group, but instead took place at the least sterically hindered C4 position. Several substituted 4-pyridineboronates (218–221) were produced in good to excellent yields (Scheme 60). In order to better understand the mechanism, ¹H and ¹¹B NMR studies were performed.

Scheme 60.

Traceless borylation of aminopyridines.

The spectroscopic data indicated that ArNHBpin species was formed. This intermediate did not react further with HBpin to ArN(Bpin)₂ even at elevated temperatures. Previous computational work on the borylation of N-Boc-protected anilines showed that ortho-borylation was favored due to the hydrogen bonding between the NH moiety of the aniline and an oxygen atom of the Bpin reagent in the transition state.⁸² No C- or N-borylation was observed, when HBpin was omitted, supporting the proposed mechanism.

In 2006, Fagnou and co-workers published a general method for a cross coupling reaction of aryl halides with perfluoroaromatics, e.g. 2-chlorotoluene with 2,3,5,6-tetrafluoropyridine (222).⁸³ The cross-coupling was carried out with Pd(OAc)₂, S-Phos, K₂CO₃ as a base, in isopropyl acetate at 80 °C. The reaction afforded product 223 in 97% yield (Scheme 61).

Scheme 61.

C4-Arylation of tetrafluoropyridine.

Later in the year the same group published a follow-up study, where 222 was coupled with 4-bromotoluene to give the biaryl product 223 in 86% yield.⁸⁴ Pd(OAc)₂ was used as a catalyst with Pt-Bu₂Me as a ligand.

In 2008, Daugulis and Do reported the copper catalyzed arylation of fluoropyridines 222 and 224.⁸⁵ The reactions were carried out with catalytic amounts of CuI and phen, with K₃PO₄ as a base in DMF at 130–140 °C. It is believed that the reaction proceeds first through deprotonation of the fluoropyridines, followed by a Li/Cu transmetalation. The intermediate organocopper species then reacts with the aryl iodide. This reaction requires 3 equiv. of the heteroarene. Lithium tert-butoxide was used as a stronger base in the case of the less acidic 3-fluoropyridine (224). The coupling products 223 and 225 were formed in 91 and 40% yields, respectively (Scheme 62).

Scheme 62.

Cu-catalyzed arylation of polyfluoropyridines.

In 2009, Su and co-workers reported the Pd-catalyzed cross-coupling of arylboronic acids with electron-deficient polyfluoro(hetero)arenes.⁸⁶ The reaction was carried out under basic conditions to effect the cleavage of the C–H bond in the polyfluoro(hetero)arene. However, these conditions also accelerated the transmetalation of the arylboronic acid, which in turn promoted the undesired homocoupling of the phenylboronic acid to biphenyl.

In order to avoid the homocoupling, a combination of a weak base and a weak acid, – p-MeC₆H₄CO₂K and t-BuCO₂H, – was employed. Under these conditions, with Ag₂CO₃ as an oxidant and Pd(OAc)₂ as a catalyst, the coupling of 2,3,5,6-terafluoropyridine (222, 3 equiv.) and phenylboronic acid proceeded in 71% yield (226, Scheme 63).

Scheme 63.

Pd-catalyzed arylation of 2,3,5,6-tetrafluoropyridine (222).

In 2010, Zhang and co-workers reported the Pd-catalyzed C–H-olefination of perfluoroarenes.⁸⁷ The authors optimized reaction conditions included Pd(OAc)₂ (10–20 mol%), Ag₂CO₃ (2 equiv.) in DMF/DMSO at 120 °C. Internal and branched olefins were successfully cross-coupled with both electron-poor and electron-rich (hetero)arenes. For example, 2,3,5,6-tetrafluoropyridine (222) was reacted with tert-butyl acrylate and styrene to give products 227 and 228 in 69 and 72% yields (Scheme 64).

Scheme 64.

Pd-catalyzed olefination of electron-deficient heteroarenes.

Subsequently, Zhang and co-workers reported the Pd-catalyzed cross-coupling of polyfluoroarenes with aromatic heterocycles.⁸⁸ A 72% yield of thiophene 229 was obtained in the reaction between pentafluorobenzene (230) and 2-acetylthiophene with Pd(OAc)₂ as a catalyst and Ag₂CO₃ as an oxidant in DMF/DMSO at 120 °C. The yield dropped substantially upon removal of DMSO or switching to other polar solvents (e.g. DMF, NMP, 1,4-dioxane). The authors surmised that DMSO acts as a ligand to the Pd, preventing the formation of Pd black. A 1:1 mixture of Ag₂CO₃ and AgOAc afforded a higher yield. Use of an excess of Ag₂CO₃ (1.5 equiv.) with AcOH (1 equiv.) afforded 229 in 77%. For the reaction involving pyridine 222 the thiophene was used as the limiting reagent. Pyridine 222 was reacted with three 2-substituted thiophenes to give the coupling products 231–233 in high yields (Scheme 65).

Scheme 65.

Cross coupling of 2,3,5,6-tetrafluoropyridine (222) with substituted thiophenes.

In 2010, Su and co-workers reported the Cu-catalyzed alkynylation of electron-deficient polyfluoroarenes.⁸⁹ A coupling reaction between pentafluorobenzene (230) and phenylacetylene in the presence of CuCl₂ (30 mol%), LiOt-Bu (3 equiv.), DDQ (30 mol%), and phen (30 mol%) in the atmosphere of O₂ (1 atm) proceeded in 72% yield. A five-fold excess of the electron-deficient arene was used. 2,3,5,6-Tetrachloropyridine (234) was successfully coupled with phenylacetylene to yield acetylene 235 in 41% yield (Scheme 66).

Scheme 66.

Alkynylation of 2,3,5,6-tetrachloropyridine (234).

In 2010, Su and Wei reported the Pd-catalyzed dehydrogenative C–H-cross-coupling of electron-deficient polyfluoroarenes with simple arenes.⁹⁰ The majority of the substrates studied were substituted polyfluoroarenes that were reacting with benzene or substituted benzenes. In one case, 2,3,5,6-tetrafluoropyridine (222) underwent a cross-coupling reaction with benzene by way of dual C–H-activation.

Under the optimized reaction conditions with 45 equiv. of benzene, Pd(OAc)₂ (20 mol%) as a catalyst, Cu(OAc)₂ (2 equiv.) as an oxidant, Na₂CO₃ (0.75 equiv.) as a base, and pivalic acid (1.5 equiv.) as an acidic additive in dimethylacetamide at 110 °C, 2,3,5,6-tetrafluoro-4-phenylpyridine (226) was produced in 70% yield (Scheme 67). In order to determine, which of the two C–H bond-cleavage processes occurred first, the authors measured the KIE for both coupling partners. A KIE of 1.3 was observed for the C4–H bond in 236. This result showed that the rate-limiting step did not involve the C–H cleavage in the fluorobenzene substrate. On the other hand, KIEs of 6.5 and 4.8 were obtained for benzene and 1,2-dichlorobenzene. These high KIEs suggested that the C–H cleavage in the arene component was the rate-determining step. This assessment was supported by lower yields, when electron-poor arenes were employed as coupling partners with polyfluoroarenes.

Scheme 67.

Pd-catalyzed dehydrogenative C–H-cross-coupling reaction of 2,3,5,6-tetrafluoropyridine (222) and the reaction kinetic data.

In 2011, Daugulis and Do reported the cross dimerization of electron-rich and electron-deficient arenes.⁹¹ The reaction is believed to be a two-step process. In the first, fast step, one of the coupling partners undergoes iodination.

In the second step, the regioselective copper-catalyzed arylation of a C–H bond ensues, as reported by Daugulis earlier.⁴⁷ The iodination proceeds with iodine, in the presence of the copper iodide/phenanthroline catalyst.

The reaction was performed with K₃PO₄ or K₃PO₄/LiOt-Bu as a base in either 1,4-dioxane or 1,2-dichlorobenzene, with pyridine as an additive (Scheme 68). A number of 4-(hetero)arylpyridines were prepared using this method (for example, 244–251 in Scheme 69). The notable features of the method are low ratios of coupling partners (1.5–3 equiv.), low rates of homocoupling side product formation and excellent functional group tolerance.

Scheme 68.

Dehydrogenative cross-coupling reaction of electron-rich arenes and electron-deficient pyridines.

Scheme 69.

Cu-catalyzed cross-coupling of heteroarenes.

In 2011, Ohmura, Suginome, and Oshima reported the palladium-catalyzed C4/N-silaboration of 2- and 3-disubstituted pyridines 252 to dihydropyridines 253 under mild conditions (Scheme 70).⁹² The initial experiment that was carried out with pyridine (10 equiv.), Me₂PhSi-Bpin (1 equiv.), in D₆-benzene with (η³-C₃H₅)PdCl(PPh₃) (2 mol%) at 50 °C yielded no product. Upon replacement of PPh₃ with PCyPh₃ the silaboration product 254 was obtained in 36% yield, while a reaction with PCy₃ as a ligand furnished 254 in 79% yield after 96 h. Changing the palladium source to Pd(dba)₂ with PCy₃ as a ligand gave product 254 in 85% yield. The reaction was then carried out with several 2- and 3-disubstituted pyridines. A number of functional groups, e.g. methyl, phenyl methoxy, and methyl ester, were found to be compatible with the catalytic system. Quinoline was also subjected to the silaboration reaction, yielding the 1,4-adduct 255 as the major product in 81% yield (by NMR).

Scheme 70.

Pd-catalyzed silaboration of 2,3-substituted azines.

Surprisingly, no reaction was observed with 2-methylpyridine. All reactions were carried out with either Me₂PhSi–Bpin or Me₂ClSi–Bpin. The silaboration adducts 253 can be isolated, if stable, or easily converted to 4-silylpyridines 256 upon heating with benzaldehyde in a single-flask fashion. For example, silylpyridines 257 and 258 were prepared using this method in 86 and 62% yields, respectively.

The following mechanism was proposed for the silaboration reaction (Scheme 71).⁹³ Oxidative addition of the silylboronic ester 259 to Pd(0) and coordination to the pyridine produce complex 260. The regioselective insertion of pyridine into the Pd–B bond with the introduction of the boryl group onto the nitrogen atom produces the π-allylpalladium complex 261. Lastly, reductive elimination gives rise to dihydropyridines 262 and 263 and regenerates the Pd(0) catalyst.⁹² Involvement of Pd(0) is supported by the experiments with Pd(dba)₂ as a catalyst that gave the product in yields similar to those obtained under the standard conditions.

Scheme 71.

Proposed mechanism of the Pd-catalyzed silylboration.

In 2012, Daugulis and co-workers reported the copper catalyzed bis-trifluoromethylsulfenylation of N-(8-quinolinyl)nicotinamide (264) to give disulfide 265 in 43% yield (Scheme 72).⁹⁴ The bulk of the paper is devoted to the sulfenylation of benzamide derivatives catalyzed by copper(II) acetate. The optimal conditions include Cu(OAc)₂ (50 mol%), and bis(trifluoromethyl) disulfide (1.8–2.5 equiv.) at 90–110 °C in DMSO.

Scheme 72.

Cu-catalyzed trifluoromethylsulfenylation of nicotinamide 264.

A control experiment with 99.999% Cu(OAc)₂ verified that the reaction is catalyzed by copper. No reaction occurred in the absence of copper.

In 2013, Daugulis and co-workers reported the directed amination of arene C–H bonds by copper(II).⁴⁸ Initial studies focused on the reaction between 8-aminoquinoline p-methoxybenzamide (266) and morpholine with Cu(OAc)₂ as a catalyst in N-methylpyrrolidinone at 110 °C. Under these conditions coupling product 267 was obtained in 39% yield. Use of N-methylmorpholine N-oxide (NMO) as an oxidant and Ag₂CO₃ (0.13 equiv.) with 10 mol% Cu(OAc)₂ afforded 267 in 87% yield. K₂CO₃ and O₂ gave lower yields. Arenes featuring fluoro, methyl, methyl ester, methoxy, trifluoromethyl, and tert-butyl groups were suitable substrates for the reaction conditions.

The C4-amination of isonicotinamide 268 furnished the amination product 269 in 56% yield (Scheme 73). A mechanism was proposed that involves activation of the aryl C–H bond by Cu(II), followed by oxidation to a Cu(III) species, coordination of the morpholine, and subsequent reductive elimination that gives rise to the Cu(I) species.⁹⁵

Scheme 73.

Cu-catalyzed amination of isonicotinamide 268.

In 2013, Yu and co-workers reported the Pd(II)-catalyzed iodination of (hetero)arenes with elemental iodine (Scheme 74).⁹⁶

Scheme 74.

Pd(II)-catalyzed iodination.

It was found that CsOAc/NaHCO₃ facilitated iodine abstraction for the catalytic cycle. A mixture of DMF/t-AmylOH solubilized the PdI₂ sufficiently well with catalyst loading as low as 2 mol%. The reactions were run with 4Å molecular sieves at 80 °C. Ortho-substitution to the amide directing group is observed in all cases. Using this method, nicotinamide 270 was converted to 4-iodopyridine 271 in 54% yield (Scheme 74).

K₂S₂O₈ (20 mol %) was used in place of NaHCO₃, to prevent dimerization, and DMSO as the solvent at 65–80°C. The ortho-selectivity of the iodination was explained by a strong trans-effect and the steric effects that favor structure 272 over bispyridine complex 273 (Scheme 75).

Scheme 75.

Assembly of the reactive precursor.

In 2013, Cui, Wu, and co-workers reported a method that allows for a regioselective synthesis of various 2-arylsulfonylquinoline N-oxides.⁹⁷ The reaction was carried out with arylsulfonyl chloride (4 equiv.), CuI (10 mol%), K₂CO₃ (2 equiv.) in 1,2-dichloroethane at 100 °C. Most of the examples are provided for 2-arylsulfonylquinoline N-oxides, with one example for isoquinoline N-oxide (274) that produced sulfone 275 in 88% yield (Scheme 76). The unusual C4-substitution pattern that was observed with isoquinoline N-oxide remains to be explained.

Scheme 76.

Cu-catalyzed C4-phenylsulfonylation of isoquinoline N-oxide.

In 2008, Hiyama, Nakao, and co-workers reported the alkenylation and alkylation of fluoroarenes using Ni(cod)₂ as a catalyst.⁹⁸ The electron-deficient pyridines 222 and 239 underwent alkenylation with 4-octyne in C4 position to give 4-vinylpyridines 276 and 277 in 85 and 99% yields, respectively (Scheme 77).

Scheme 77.

Ni-catalyzed alkenylation of electron-deficient pyridines.

In 2010, Ong and co-workers reported a Ni-catalyzed conversion of substituted pyridines 16 to 4-vinylpyridines 278 (Scheme 78).⁹⁹ Concomitant formation of 3-vinylpyridines 279, typically as minor products, was also observed. In the optimization screening, a reaction of pyridine with 4-octyne in the presence of a complex of AlMe₃with amino-NHC ligand 280 (20 mol%), and Ni(cod)₂ (10 mol%) at 80 °C afforded the desired product 281 in 83% yield.

Scheme 78.

Ni-catalyzed C4-alkenylation of pyridines.

Other Lewis acids such as ZnMe₂ or BH₃ proved detrimental to the reaction (7% and 5%, respectively). Good yields were reported for monosubstituted pyridines containing phenyl, methyl, and methoxy groups. Quinoline and 6-methyl quinoline were also studied. 3-Methylpyridine gave rise to product 282 along with the 3-isomer in a 5:2 C4/C3 ratio. Quinoline afforded product 283 with a 10 : 1 C4/C3 ratio.

In order to understand factors influencing the regioselectivity, Ni(cod)₂ was treated with stoichiometric amounts of ligand 280 and pyridine (Scheme 79). After a few hours all of ligand 280 had been consumed and complex 284 formed. A crystal structure unambiguously confirmed a three-coordinate nickel(0) center bound to two 243 units and one pyridine unit. The Lewis acid, AlMe₃ is coordinated to the nitrogen of the pyridine ring, providing a plausible explanation of the C4/C3-selectivity.⁵³

Scheme 79.

Complexation of Ni⁰ with NHC and pyridine.

Competition experiments yielded a small kinetic isotope effect of 1.25. This result shows that the C–H bond cleavage is not the rate-limiting step. The proposed mechanism is shown in Scheme 80. Formation of 285 is followed by the oxidative addition of the alkyne to produce Ni hydride 286. Subsequent alkyne insertion into the Ni–H bond gives rise to intermediate 287. Finally, reductive elimination yields product 278 and the catalytically-active Ni(0) complex.

Scheme 80.

Plausible mechanism of the Ni-catalyzed C4-alkenylation of pyridines.

In 2010, Hiyama, Nakao, and co-workers reported the C4-selective Ni-catalyzed alkylation of pyridines (Scheme 81).¹⁰⁰ After developing a C2-selective alkylation of pyridine using a nickel-based catalytic system with electron-rich phosphine ligands,¹⁰¹ the authors used a Lewis acid to coordinate to the N1-nitrogen, allowing for the steric and electronic factors to dictate the selectivity away from C2. For their optimization study the research team used pyridine, 1-tridecene, Ni(cod)₂, 1,3-2,6-diisoprpylphenyl)imidazole-2-ylide (IPr, 288) and AlMe₃ in toluene at 130 °C. This reaction yielded 4-tridecylpyridine (289) in 70% yield. No product was observed upon replacement of the NHC with various phosphine ligands. A number of Lewis acids were screened, with AlMe₃ and MAD (290) providing the highest yields and C4 selectivity. On the other hand, diorganozinc reagents were inefficient Lewis acids, with no product obtained. 30 mol% of the Lewis acid was optimal with higher loadings providing no benefit, and lower loadings providing lowered yields. IMes (291) and IPr (288) gave the optimal yields of alkylpyridines 292 and 293. High selectivity of the linear C4-alkylated product 292 was observed, with the branched product 293 as a minor contaminant with aliphatic 1-alkenes. With styrene, however, the branched addition product 294 was obtained in 95% yield.

Scheme 81.

Ni-catalyzed C4-alkylation of pyridines.

The reaction proceeds with a number of substrates in 12–95% yields. Aliphatic 1-alkenes having a phenylsilyl- or pivaloyl-protected hydroxyl group and a terminal or internal double bond as well as vinylsilanes were suitable substrates for the alkylation reaction.

One reaction was also carried out with 4-octyne under standard conditions. The reaction produced a Z-alkenylated pyridine 281 in 53% yield, along with the corresponding C3-isomer in 15% yield. In order to better understand the mechanism, a reaction with D₅-pyridine and 1-tridecene (1.5 equiv.) was performed under the standard conditions. While some loss of deuterium at C2 and C3 was observed after 9 h, no D/H exchange was detected after 3 min (value in parentheses) (Scheme 82).

Scheme 82.

H/D exchange in the Ni-catalyzed C4-alkylation of pyridines and plausible mechanism.

While there is an H/D exchange at the C2 and C3 positions, lack of a large amount of these alkylated products implies that the coordination of Ni to these sites is reversible, in contrast to the irreversible reaction at the C4 position (Scheme 82). This result also implies a catalytic cycle initiated by the oxidative addition of the C4–H bond of the pyridine in the Ni/pyridine-MAD complex 295 The resulting intermediate 296 engages the alkene to give complex 297. Subsequent migratory insertion of the alkene into the Ni–H bond produces alkyl nickel species 298. Finally, reductive elimination from 298 gives rise to alkylpyridine 299 and, after complexation with pyridine and MAD, the Ni(0) complex 295.

4 C5/C6/C7–H-Functionalization

In 1995, Huang and co-workers reported the perfluoroalkylation of pyridine, quinoline, and isoquinoline.¹⁰² Based on the previous research by the group,¹⁰³ sodium perfluoroalkanesulfinates were reacted with heteroarenes and Mn(OAc)₃ in MeCN/AcOH/Ac₂O at 80–85 °C. The reaction generates perfluoroalkyl radicals by oxidation with Mn(OAc)₃. The radicals add at C2 in substituted pyridines with some C3/C4 addition as well. Quinoline (1), 4-methylquinoline (300), and isoquinoline (301) were also reacted under the standard conditions, with all three giving mixtures of C6-isomers 302–304 and C8-isomers 305–307 in 45–61% yields and with a 1.3:1 average C6/C8 selectivity (Scheme 83).

Scheme 83.

Radical perfluoroalkylation of heteroarenes.

In 2010, Yamakawa and co-workers reported the trifluoromethylation of arenes and heteroarenes with trifluoroiodomethane and Fe(II).¹⁰⁴ Previous work performed by the same group described the radical trifluoromethylation of a number of nucleobases.¹⁰⁵ The method uses CF₃I, FeSO₄ or Cp₂Fe, H₂O₂, DMSO as a radical promoter, and H₂SO₄.

A number of arenes, heterocycles, and pyridine derivatives were converted to the corresponding trifluoromethylated derivatives by this procedure. When quinoline was reacted under these conditions, the C5-trifluromethylated product 308 was obtained in 10% yield. With 8-aminoquinoline (309), a mixture of C5-mono-, C7-mono-, and C5/C7-bistrifluoromethylation products 310–312 were obtained in a 2:1:5.8 ratio with an overall 79% yield. The reaction is postulated to proceed via an initial formation of a hydroxyl radical through the reduction of H₂O₂ by Fe(II) that is followed by the interception of the OH radical by DMSO to form a methyl radical. The methyl radical reacts with CF₃I to give the trifluoromethyl radical.¹⁰⁶

In 2013, Ertem, Stahl, and co-workers reported the radical chlorination of 8-amidoquinolines under acidic conditions.¹⁰⁷ The work focused on the reactivity of the N-(quinolin-8-yl)benzamide 85. Previous work on the Cu(II)-catalyzed aerobic C–H-oxidation had been studied on electron-rich arenes, but not on substituted azines. The optimization study began with conditions similar to the oxidative chlorination of electron-rich arenes: 1 atm O₂, 2 equiv. LiCl in AcOH at 100 °C with CuCl or CuCl₂. Surprisingly, instead of the chlorination of the aryl ring of the benzamide they observed formation of the C5-chlorination product 313 in 34% and 81% yields with CuCl₂ and CuCl, respectively. Upon addition of 20 mol% LiOAc with CuCl₂ the yield was increased to 88% with exclusive C5-chlorination. Lithium acetate was proposed to act as a Brønsted base that promotes substrate binding to the Cu center.¹⁰⁸ Mechanistic studies were performed to better understand chlorination of the quinoline ring. No deuterium incorporation was observed in CD₃CO₂D in place of AcOH, indicating that the C–H bond cleavage is irreversible. Based on the KIE studies a single electron transfer (SET) C–H-functionalization pathway was proposed for the quinoline chlorination. A calculated mechanism for the Cu(II) mediated chlorination is shown in Scheme 85. Cu(II) complex 314 undergoes an SET oxidation of the quinoline system to give radical cation 315.

Scheme 85.

Calculated mechanism for the copper-mediated C5-chlorination.

A chlorine atom is then transferred from CuCl₂ to C5 position with the C5–H of the azine activated by a chloride and acetic acid via a transition state 316 to give the cationic species 317. Further deprotonation proceeding via transition state 318 furnishes the C5-chlorinated quinoline complex 319.

In 2014, Zeng and Cong described a chelation-induced C5-allylation of 8-amidoquinolines 320 (Scheme 86).¹⁰⁹ The reaction was catalyzed by iron(III) chloride. Preliminary mechanistic analysis pointed to the activation of the substrates 320 via chelation of the Fe catalyst, and the activation of the allylic alcohol by Fe(III). The products 322 were obtained in good to excellent yields and with an excellent C5-regioselectivity.

Scheme 86.

Fe(III)-catalyzed C5-allylation of 8-amidoquinolines.

In 2013, Dong and Dong reported the Pd and norbornene catalyzed C–H amination using aryl halides (Scheme 87).¹¹⁰ The authors utilized a reductive Catellani reaction pathway to install an amino group ortho to the position previously occupied by the halogen. Numerous (hetero)arene examples are provided, including three quinolines. Under the optimized conditions 5-iodoquinoline (322) was reacted with N-benzoyloxymorpholine (323) to give aminoquinoline 324 in 88% yield. Substituted 6-aminoquinolines 325 and 326 were also produced in 50 and 99% yields, respectively.

Scheme 87.

Pd- and norbornene-catalyzed ortho-amination.

5 C8–H-Functionalization

In 2011, Chang and co-workers reported the Rh(NHC)-catalyzed regioselective synthesis of 8-arylquinolines 327 from quinolines 328.¹¹¹ Prior work by the group¹¹² led them to postulate the possibility of a Rh-NHC catalytic system for direct regioselective C8-arylation of quinoline.

Quinoline, 4-bromotoluene, and Rh₂(OAc)₄ with and without IMes ligand 291 were reacted at 95 °C. C8-arylation product 329 was obtained in 3% yield in the absence of the ligand 291. With the ligand added, product 329 was obtained in 58% yield with >99:1 C8/C2 regioselectivity. With other bulkier NHC ligands, e.g. IAd (330) or IPr (288), the selectivity decreased to 2:1 and 10:1, respectively. Phosphine ligands proved ineffective, providing low yields and site-selectivity. For example, the yield of 329 was only 5% and the C8:C2 ratio was 5:1 with PCy₃. The method tolerates a number of para-substituents in the aryl bromides (e.g. Me, MeO, F, CF₃). ortho-Substituted aryl bromides were not reactive, while aryl chlorides were as active as aryl bromides. Quinolines bearing a number of functional groups (e.g. methyl, benzyl, phenyl, methoxymethyl ethers, alkyl, aryl) in C3, C4, and C6 positions provided C8-arylation products 327 in 64–93% yields (Scheme 88). Quinolines 331 and 332 containing a tertiary amino group and an acetal were also prepared in good yields.

Scheme 88.

Rh(NHC)-catalyzed C8–H-arylation of quinolines.

In order to better understand the mechanism primary kinetic isotope effect was measured. A KIE of 2.76 was observed for the C8-arylation of quinoline. Chang and co-workers attribute this result to a base-assisted concerted proton abstraction and metalation in the C8-position of quinoline. Both bimetallic 333 and monomeric 334 are plausible intermediates, and more studies are needed to clarify the mechanism (Scheme 88).¹¹³

In 2014, Marder, Sawamura and co-workers reported the C8-selective borylation of quinolines catalyzed by a silica-supported Ir/phosphine system (Scheme 89).¹¹⁴ 8-Borylquinolines 335 had previously been prepared by multistep sequences via borylation of 8-haloquinolines.¹¹⁵ As discussed earlier in this review, prior to this work borylation of pyridines and quinolines under Ir or Rh catalysis occurred preferentially at the C3 or C4 positions with bidentate N-containing ligands (e.g. dtbpy). The authors began their study by reacting quinoline and B₂pin₂ in MTBE at 60 °C in the presence of an immobilized Ir-catalyst precursor prepared from silica-SMAP ligand 336 and [Ir(cod)(OMe)]₂. The C8-borylated quinoline 337 was formed in an 81% yield, with 13% of the 1,2,3,4-tetrahydroquinoline (338) as a side product. HBpin was also tested as a boronic ester source, however, product 337 was obtained in only 17% yield. In addition, side product 338 was also formed in 30% yield. A similar immobilized Rh-catalyst afforded no C8-borylation product. Other phosphine ligands or silica supported ligands yielded only trace amounts of products (Scheme 90). The method is applicable to a number of mono- and di-substituted quinolines bearing alkyl, chloro, methoxy, and aryl groups. For example, the 2-substituted 8-borylquinolines 339 and 340 were readily prepared in 91 and 74% yields. It was noted that with 2-methylquinoline and Ir-dtbpy catalyst, in place of 336, a mixture of C4/C6/C7 borylated products was obtained in 16%, 27%, and 20% yields, respectively.

Scheme 89.

Ir-catalyzed C8–H-borylation of quinolines.

Scheme 90.

Phosphines providing trace C8–H-borylation product.

A number of 8-borylquinolines 335 were too unstable to be isolated, hence the crude product was subjected to the oxidation with NaBO₃, followed by treatment with Boc₂O (Boc = tert-butoxycarbonyl). This sequence furnished four 8-tert-butoxycarboxyquinolines 341 in 60–71% yields (Scheme 91). To further demonstrate the utility of this method, the corticotropin-releasing factor₁ (CRF₁) receptor antagonist 342 was synthesized from quinoline 343 using the C8-borylation that was followed by a Suzuki-Miyaura reaction of the intermediate boronate 344 in the late stage of the synthesis (Scheme 92).¹¹⁶

Scheme 91.

8-tert-butoxycarboxy-substituted quinolines prepared using the Ir/336-catalyzed C8-borylation of quinolines.

Scheme 92.

Synthesis of the CRF₁ receptor antagonist 342 using a late-stage C8–H borylation strategy.

In 2014, Shibata and Matsuo reported the Rh-catalyzed regioselective C8-alkenylation of quinoline N-oxides 345 (Scheme 93) that produced 8-vinyl derivatives 346.¹¹⁷ This method uses the N-oxide moiety as a directing group that is presumed to facilitate formation of a five-membered rhodacycle. Initial experiment with quinoline N-oxide (347), diphenylacetylene, [Rh(cod)₂BF₄] (10 mol%), BINAP (10 mol%), in chlorobenzene at 135 °C afforded product 348 in 34% yield.

Scheme 93.

Rh-catalyzed C8-selective alkenylation of quinoline N-oxide.

It was found that shorter reaction times were key, as prolonged reaction times led to extensive decomposition. By changing the Rh-counterion to triflate and the ligand to 3,5-xylyl-BINAP, and shortening the reaction time to 1 h, 348 was obtained in >95% yield.

Heating at 135 °C was necessary to achieve a high yield in chlorobenzene. On the contrary, in cyclopentyl methyl ether (CPME) the reaction gave comparable results already at 110 °C. A number of para-diarylacetylenes afforded products 346 in 41–89% yields. The best E/Z ratio of the addition product was >20:1 with the trifluoromethyl- (in chlorobenzene at 135 °C), and chloro-substituted diarylacetylenes (in CPME at 110 °C).

A 1/1 E/Z ratio was obtained with bis(4-methoxyphenyl)acetylene. C2-, C3-, C4-, C6- and C2/C6-substituted quinolines bearing methyl, methoxy, chloro, and nitro groups afforded corresponding products 346 (e.g. 348–351) in 50–89% yields. E/Z ratio was highest (>20:1) for the products 349 and 350 that arouse from 3- and 4-methylquinolines, while quinoline 351 was formed with the 5 : 1 E/Z ratio.

In order to shed light on the mechanism of the reaction, quinoline N-oxide (347) was reacted in the presence of D₂O and in the absence of the alkyne under the standard reaction conditions. 69% of H/D exchange was observed in C2- and C8-positions. With pyridine N-oxide (352) 50–60% of C2-deuteration was observed under the same conditions (Scheme 94). These results suggested that the C–H-activation by Rh occurs in both positions and produces intermediates 353 and 354 (Scheme 95). However, alkyne insertion only occurs in C8 position due to the stability of the five-membered metalacycle 353. Subsequent reductive elimination from intermediate 355 affords N-oxide 348.

Scheme 94.

Rh-catalyzed C2- and C8-H/D exchange in azine N-oxides.

Scheme 95.

Plausible mechanism of the C8-selective alkenylation.

In 2014, Chang and co-workers reported the C8-selective Rh-catalyzed iodination and amidation of quinoline N-oxides.¹¹⁸ After initial screening of the reaction conditions it was found that a Rh(III) catalytic system enabled the iodination of substituted quinoline N-oxides 345.

Bromination and chlorination reactions were not synthetically useful providing corresponding C8-halogenated products in less than 20% yields. The optimized reaction conditions include [RhCp*Cl₂]₂ (4 mol%), N-iodosuccinimide (NIS, 1.5 equiv.), AgNTf₂ (16 mol%) in DCE at 50 °C. Products 356 were formed in 32–93% yields (Scheme 96). A number of functional groups were compatible with the catalytic system.

Scheme 96.

Rh-catalyzed C8–H-iodination.

For example, nitro- and formyl-substituted 8-iodoquinoline N-oxides 357 and 358 were produced in 77 and 93% yields. 8-Iodoquinoline N-oxides are valuable synthetic intermediates that can be used for further functionalization of the C8-position.¹¹⁹ Attention was further turned to the catalytic synthesis of 8-amidoquinoline N-oxides 359 (Scheme 97). Using conditions reported in their previous work¹¹¹ with p-toluenesulfonyl azide, [IrCp*Cl₂]₂/AgNTf₂ in DCE at 80 °C 8-tosylamide 360 was obtained in 24% yield. While addition of sodium acetate did not improve the catalytic performance, addition of acetic acid (30 mol%) at a lower temperature (50 °C) resulted in an improved yield (92%).

Scheme 97.

Ir-catalyzed C8–H-amidation of quinoline N-oxides.

No C2-amidation was observed under the reaction conditions. Other acids were tested as additives (e.g. trifluoroacetic acid, camphorsulfonic acid, pivalic acid and benzoic acid), with acetic acid emerging as an additive of choice, due to the ease of handling and the ease of removal for purification. It was also noted that rhodium-, ruthenium-, and palladium-based catalytic systems, were ineffective at 50 °C.

Various functional groups were tolerated under the reaction conditions, e.g. nitro, esters, acetal, alkoxy, siloxy, carbamate, and aldehyde in the quinoline core (361-366). Alkanesulfonyl, benzenesulfonyl, naphthalene, and acyl azides were suitable, giving rise to products 367–369, whereas aryl and alkyl azides did not react under these conditions (Scheme 97). The reaction proceeds with a small kinetic isotope effect (k_H/k_D = 1.23 with acetic acid, k_H/k_D = 1.14 without acetic acid) (Scheme 98). Due to the small observed KIE the C–H bond cleavage is the rate-limiting step.¹²⁰ Crystalline C8-Ir complex 370 was isolated and was subjected to the reaction conditions as a catalyst. The reaction gave rise to the desired product in a 62% yield, suggesting that complex 370 could be an intermediate in the catalytic cycle, although it could also be an off-cycle intermediate. Based on the aforementioned information the following mechanism was proposed (Scheme 99). The silver salt converts the dimeric Ir to a cationic species that reacts with the N-oxide substrate to give the five-membered iridacycle 370.

Scheme 98.

Study of the primary kinetic isotope effect.

Scheme 99.

Proposed mechanism for the Ir-catalyzed C8–H-amidation.

Subsequent coordination of azide 371 produces intermediate 372. Insertion of the amido group into the C–Ir bond affords complex 373 that subsequently undergoes protodemetalation, as shown in structure 374, to give complex 375. Acid-assisted release of the cationic Ir catalyst completes the catalytic cycle. The synthetic utility of the method was demonstrated by the synthesis of zinquin ethyl ester 376. The ester is a derivative of an important fluorescent sensor for Zn(II) in cells.¹²¹ The starting material 377 was synthesized from 6-hydroxy-2-methylquinoline (378) in two steps. The C8–H-amidation furnished sulfonamide 379 in a 76% yield and with high C8-selectivity. Subsequent reduction with zinc produced zinquin ethyl ester 376 in a 54% overall yield (Scheme 100).

Scheme 100.

Synthesis of zinquin ethyl ester.

Later that year Chang and co-workers reported new examples of C8–H-functionalization catalyzed by Rh(III) at room temperature with hypervalent iodoalkynes and diazoesters (Scheme 101).¹²² The reaction optimization was performed for the reaction of 6-methylquinoline N-oxide with methyl diazomalonate catalyzed by [RhCp*Cl₂]₂ and AgSbF₆ at various temperatures in DCE. Addition of 20 mol% pivalic acid at 25 °C provided N-oxide 380 in 94% yield. Solvents other than DCE retarded the reaction.

Scheme 101.

Rh-catalyzed C8–H-alkylation of quinoline N-oxides.

Changing the catalyst to [IrCp*Cl₂]₂ or [Ru(p-cymene)Cl₂]₂ severely reduced the yield (<5%). For the alkylation of quinoline N-oxides with diazo compounds it was found that alkyl, nitro, halogen, carbamate, siloxy (381), acetoxy, acetal (382), aldehyde, keto, and ester groups in the quinoline N-oxides were well-tolerated. The diazoesters bearing phenylsulfonyl and benzoyl groups, as well as diethyl diazomalonate were suitable for the reaction, while phosphoryl-, cyano- and nitro-substituted diazoesters did not provide the desired products. 8-Alkylquinoline N-oxides 383 were typically formed in 40–96% yields.

In addition to C8–H-alkylation, Chang and co-workers developed alkynylation with I-alkynyl-substituted hypervalent iodine reagents previously reported by Loh, Li and Glorius groups independently: 1-[(triisopropylsilyl)ethynyl]-1,2-benziodoxol-3(1H)-one (TIPS-EBX, 384) (Scheme 102).¹²³ A variety of 8-alkynylquinoline N-oxides 385 were prepared in 62–95% yields.

Scheme 102.

Rh-catalyzed C8–H-alkynylation of azine N-oxides.

The optimal catalyst for this reaction was found to be [RhCp*(MeCN)₃](SbF₆)₂. Addition of 4Å molecular sieves improved the yield dramatically. Quinoline N-oxides bearing halogen, alkyl, as well as ester, carbamate, siloxy, acetoxy, acetal, aldehyde and ketone groups proved to be viable substrates (see, for example, products 386–390 in Scheme 102). To gain insights into the mechanism, isotope-labeling experiments were carried out for both reactions under the standard conditions (Scheme 103). For the alkylation a k_H/k_D value of 5.0 was obtained, while a somewhat smaller but significant KIE (k_H/k_D = 2.7) was observed for the alkynylation. It was, therefore, proposed that the C–H-bond cleavage was involved in the rate-limiting step. Based on the KIE data, the following mechanism was proposed: initial formation of the five-membered rhodacycle 393 is followed by coordination of the diazo species to give intermediate 394 that, upon elimination of N₂, produces the metal carbenoid species 395 (pathway A, Scheme 104). Subsequent migratory insertion gives rise to a 6-memebered rhodacycle 396 that undergoes protodemetalation to give the desired product 397 and the catalyst. In addition, a pathway that does not involve a discrete carbenoid intermediate en route to the 6-membered rhodacycle cannot be ruled out (pathway B, Scheme 104).¹²²

Scheme 103.

Kinetic isotope effect experiments.

Scheme 104.

Proposed mechanism of the C8–H-alkylation of quinoline N-oxides.

In 2014, Li and co-workers reported the Rh(III)-catalyzed C8–H-alkylation of quinoline N-oxides followed by O-atom transfer, resulting in ketone products 398 (Scheme 105).¹²⁴ The initial experiment with quinoline N-oxide (347), diphenylacetylene and [RhCp*Cl₂]₂ in DCE at 110 °C afforded no product. Addition of AgSbF₆, to cleave the Rh-dimer, led to formation of product 399 in 32% yield. A number of additives were tested including acetic and pivalic acids, as well as Zn(OTf)₂. Under the optimal conditions in 1,4-dioxane, with AcOH (2 equiv.), and Zn(OTf)₂ (20 mol%) as additives product 399 was formed in 90% yield. Several symmetric para-substituted diarylacetylenes were tested, and the corresponding products containing halogen (400, 401), alkyl (402, 403), trifluoromethyl (404), methoxy (405), and other groups were isolated in good yields. No reduction of yields was observed with C2-substituted N-oxides (e.g. products 406 and 407).

Scheme 105.

Cascade C8–H-alkynylation/O-transfer reaction of quinoline N-oxides.

On the other hand, 7-chloroquinoline N-oxide afforded the corresponding product 408 in a lower yield (38%), indicating that steric hindrance close to the addition position is detrimental to the reaction. Symmetric para- and meta-substituted diarylacetylenes proved to be suitable substrates. An electronically-biased alkyne, (p-methoxyphenyl)(p-trifluoromethylphenyl)-acetylene afforded a mixture of isomers 406 and 407 with a 1.5 : 1 406/407 ratio (Scheme 105).

Acetylenes bearing a phenyl and an alkyl group also proved to be suitable coupling partners, giving rise to aromatic ketones, e.g. 409 and 410. In contrast, dialkyl-substituted acetylenes and 1-phenyl-2-trimethylsilylacetylene did not give the expected products. Extensive mechanistic studies were performed in order to elucidate the mechanism of the C–H-activation, and O-atom transfer process. The reaction was run under standard conditions with di(p-tolyl)acetylene in the presence of 4 equiv. of D₄-acetic acid. Low deuterium incorporation (5%) at C8 was observed (at 20% conversion), suggesting an irreversible C8–H activation. A competition reaction between quinoline N-oxide (347 and its D₇-isotopomer (D₇-347) with di(p-tolyl)acetylene under standard conditions at 20% conversion afforded a KIE value of 3.6–4.0, indicating that the C–H-activation may be involved in the rate-determining step.

In order to gain insights into the O-atom transfer (OAT) process, quinoline N-oxide (347) was reacted with diphenylacetylene under standard conditions with 6 equiv. of H₂¹⁸O. No ¹⁸O incorporation was observed in the isolated product, indicating that water is not involved, and that the OAT is an intramolecular process. Furthermore, the reaction was carried out in the presence of 3-bromoquinoline (411). Bromide 411 is expected to compete with the quinoline, if the oxygen transfer occurs prior to C8–H-activation. Formation of a product arising from 3-bromoquinoline was not observed, suggesting that a direct O-atom transfer to alkynes that produces quinoline and a rhodium α-oxocarbene species is unlikely. The possibility of an unassisted insertion of a rhodium(III)/α-oxocarbene intermediate into the C8–H bond of quinoline was also excluded. It was, therefore, proposed that the C–H-activation occurs prior to the O-atom transfer. When 8-(1,2-diphenylvinyl)quinoline N-oxide (412) was subjected to the reaction, no O-atom transfer was observed. This result suggests that an intermediate, such as olefin 412, is not plausible. Rhodium(III) η³-benzyl complex 413 was isolated, that produced ketone 399 when treated with 347 (Scheme 107). Metalacycle 414 was detected by HRMS during in the reaction mixture. These results suggest that 413 is a resting state of the catalyst in the absence of AcOH, and that the release of the catalyst, likely through participation of quinoline N-oxide may be the rate-determining step.

Scheme 107.

Isolation of rhodium(III) η³-benzyl complex 413 and further mechanistic studies.

Two possible mechanisms were proposed (Scheme 108), – path A and path B. In path A, the cyclometalation followed by alkyne coordination and the migratory insertion produces the seven membered rhodacycle 415. Subsequent reductive elimination generates a rhodium(I) species and intermediate 416. Oxidative addition of 416 to the rhodium species generates an O- or C-bound enolate (417 or 418). Tautomerization produces rhodium(III) η³-benzyl complex 413 that liberates the product upon reaction with quinoline N-oxide. In path B formation of the five-membered rhodacycle 419 is followed by the alkyne insertion to give the seven membered rhodacycle 420. Cleavage of the N–O bond affords a metal α-oxocarbene species 421 that undergoes migratory insertion to yield rhodium(III) η³-benzyl complex 418, which then follows path A.

Scheme 108.

Two possible mechanistic pathways.

In 2015 Lan, Li, and co-workers reported a computation study of this Rh-catalyzed reaction,¹²⁵ using the M11L DFT method.¹²⁶ While four different pathways were ultimately evaluated, the path A in Scheme 108 was found to be the most energetically favorable one.

Later that year Chang and co-workers reported a similar Rh(III)-catalyzed C8-alkylation/O-atom transfer cascade for quinoline N-oxides (Scheme 109).¹²⁷ Under the optimal conditions for Chang's method, an excess of quinoline N-oxide is reacted with the alkyne in the presence of [RhCp*Cl₂]₂ (2.5 mol%) and Cu(OAc)₂ (5 mol%) in DMF at 110 °C. Kinetic isotope studies produced a KIE of 2.1, – almost half of that reported by Li.

Scheme 109.

Kinetic isotope effect studies and the scope of the Rh-catalyzed C8-alkylation/O-atom transfer.

Further, 8-(1,2-diphenylvinyl)quinoline N-oxide (412) was synthesized and subjected to the reaction. Less than 5% yield of the OAT product was formed in this case. The ¹⁸O-isotopomer of 3-methylquinoline N-oxide (¹⁸O-422) was also reacted with diphenylacetylene. Complete incorporation of the ¹⁸O in the ketone moiety of the product 423 was observed. The products 398 were typically isolated in 17–93% yields.

In 2015, Larionov and co-workers reported the Pd-catalyzed site-selective C8-arylation of quinoline N-oxides (Scheme 110).¹²⁸ In a departure from the previously reported Rh- and Ir-based methodology, the authors focused their efforts on catalysis with palladium. Initial screening with quinoline N-oxide, 4-iodobenzotrifluoride, Pd(OAc)₂ and acetic acid (10 equiv.) at 120 °C gave product 424 in 8% yield. With the addition of AgOAc (3 equiv.) the yield improved to 39%. Use of Ag₃PO₄ (0.5 equiv.) and AcOH/H₂O (30/5.5 equiv.) raised the yield to 95% and the C8/C2 selectivity to 23:1. Microwave conditions were also developed that allowed for sub-hour reaction times. The reactions were generally complete within 50 min at 180 °C under the microwave irradiation, while under thermal conditions at 120 °C the reaction required ~16 h.

Scheme 110.

Pd-catalyzed C8-arylation of quinoline N-oxides.

A one-pot synthesis of 8-(4-bromophenyl)quinoline (426) can be accomplished under the microwave irradiation within 3 h. In this case, the N-oxidation of quinoline with hydrogen peroxide in acetic acid to give N-oxide 347 that was subjected to the C8–H-coupling with 4-bromoiodobenzene. Reduction of the N-oxide with hypophosphorous acid produced 426 in 67% overall yield. A number of 8-arylquinoline N-oxides 425 bearing diverse substituents (e.g. aryl, nitro, halogen, ester, methoxy, and alkyl groups) at C2/C3/C4/C5/C6 in the quinoline core were synthesized in 54–93% for the thermal conditions, and in 69–94% for the microwave-assisted reactions (Scheme 110). The synthesis of 8-phenylquinoline N-oxide (427) was carried out on a 2 gram scale with 82% yield.

A mechanistic study was undertaken to clarify the origin of the site-selectivity. The H/D exchange occurred in the C8 position, but not in C2 position of quinoline N-oxide under standard conditions in deuterated solvents. A primary KIE of 2.0 was observed for the C8-arylation. These results, together with the data on the substituent effects obtained from the Hammett studies suggest that the C8-cyclometalation may be the rate-determining step. With this information in hand, a computational study was performed to determine the factors that influence the site-selectivity of the cyclopalladation of quinoline N-oxides. It was found that acetic acid plays a crucial role as a ligand that makes Pd more electrophilic and diverts the reaction in the more electron-rich C8 position.

On the other hand, Fagnou's Pd/phosphine systems react predominantly in the C2 position due to the higher acidity and strength of the C2–H bond.¹² (Scheme 111).

Scheme 111.

Computational study of the Pd-catalyzed C8-arylation of quinoline N-oxides.

C8–H-Arylation of quinoline N-oxides 345 was also recently accomplished by Chang and co-workers, with arenediazonium salts as the electrophilic coupling partners (Scheme 112).¹²⁹ The reaction was catalyzed by [IrCp*Cl₂]₂ (5 mol%)/AgNTf₂ (20 mol%) at 45 °C in trifluoroethanol. Eight 8-arylquinoline N-oxides 425 were prepared in 49–89% yields.

Scheme 112.

Pd-catalyzed C8-arylation of quinoline N-oxides.

In addition to the C8–H-arylation of N-oxides 345, efficient ortho-C–H-arylation of benzamides and C3–H-arylation of acrylamides were also described.

In 2015, Larionov and co-workers added another example of a Pd-catalyzed C8–H-functionalization, – the C–H homocoupling of quinoline N-oxides (Scheme 113).¹³⁰ During optimization of their Pd-catalyzed C8-site-selective arylation,¹²⁸ a formation of a side product was observed in some instances. Upon removal of the iodoarene from the reaction mixture, the yield of the side product increased to almost 50%.

Scheme 113.

Pd-catalyzed oxidative C8–H-homocoupling of quinoline N-oxides and proposed mechanism.

Single crystal X-ray crystallographic analysis of the side product confirmed that it was 8,8’-biquinolyl N,N′-dioxide (428) that was formed by way of oxidative C8–H-homocoupling of quinoline N-oxide. The dimerization products were obtained in 42–83% yields in the presence of Pd(OAc)₂ (10 mol%), AgOAc (4 equiv.), AcOH (5 equiv.) and H₂O (1.5 equiv.) at 120 °C. In line with earlier observations, the addition of water significantly increased the yield. It was suggested that water assists the breakdown of trimetric Pd(OAc)₂.¹³¹

In order to gain insights into the mechanism of the reaction, a Hammett substituents effect and kinetic isotope effect studies were undertaken. A ρ value of −1.28 was obtained that was significantly smaller then for the H/D exchange and for the C8–H-arylation. No primary KIE (k_H/k_D = 1) was observed for the homocoupling, showing that formation of palladacycles 429 and 430 was not a turnover-limiting step, in contrast to the Pd-catalyzed C8–H-arylation. It was, therefore, proposed that the reaction involved higher oxidation state Pd intermediates 431 (Scheme 113).

In 2015, Li, Wan and Yu reported the Rh(III)-catalyzed regioselective C–H-selenation of (hetero)arenes (Scheme 114).¹³² The reaction was carried out with [RhCp*Cl₂]₂ as a catalyst in the presence of AgSbF₆ (1.5 equiv.), NaOAc (1.2 equiv.), and phenylselenyl chloride (1.2 equiv.) in THF at 80 °C. Low yields of 8-selenylquinoline N-oxides 432 were observed with other silver salts or other acetate salts. Products 433–437 were isolated in 45–67% yields. The reaction is believed to proceed through the initial formation of rhodacycle 438 that engages phenylselenyl chloride to give the cationic species 439. In the next step, silver salt-assisted electrophilic selenylation of the C–Rh bond affords complex 440. Subsequent reaction of complex 440 with substrate 441 affords selenation product 442 and rhodacycle 438, thus completing the catalytic cycle. A Rh(III)/Rh(V) pathway, where the Se–Cl bond oxidatively adds to a Rh(III) species may also be operative.¹³³

Scheme 114.

Rh(III)-catalyzed selenation.

6 Summary and Outlook

The reactions discussed in this review attest to the significant progress that has been achieved in the area of distal C–H-functionalization of pyridines and quinolines in the past decade. A number of site-selective transformations have been developed for C3 and C4 positions, as well as C8 position in quinolines. On the other hand, C5, C6 and C7 positions in quinolines have generally only been functionalized by using a directing group in an adjacent position. Hence, methods of non-directed C–H-functionalization in these positions have yet to be established. In addition, diversification of functional groups that can be installed by means of C–H-functionalization is an important area of the future systematic mechanism-driven studies. Site-selective C–X (e.g. F and Cl), C–O, C–N and C–C bond-forming reactions (e.g. trifluoromethylation and alkylation) will also likely become the focus of the future work. New catalytic systems based on other metals, – in particular the more abundant Mn, Co, Ni, Fe, and Cu, – should also become a focal point for future research. Together, these developments are expected to significantly simplify access to the important class of azaheterocycles.

Figure 1.

Examples of important distally-substituted pyridines and quinolines.

Scheme 25.

8-Quinolylamide-directed fluorination.

Scheme 48.

Kinetic isotope effect studies.

Scheme 84.

Radical trifluoromethylation of azines.

Scheme 106.

Studies performed to elucidate the mechanism.

Scheme 115.

Plausible mechanism for the Rh(III)-catalyzed selenation.

Biography

Oleg Larionov graduated with a Master's degree in Chemistry from the Higher Chemical College of the Russian Academy of Sciences. After the doctoral studies under the supervision of Prof. Armin de Meijere at Georg-August-University in Göttingen, Germany, he moved to Mülheim an der Ruhr, Germany to work with Prof. Alois Fürstner at Max-Planck Institut für Kohlenforschung. In 2007 he joined the group of Prof. E. J. Corey at Harvard University to work on the synthesis of caryophylloid natural products. Since 2010 he has been an Assistant Professor at the University of Texas at San Antonio. His research is focused on the development of catalytic methods of synthesis of N- and N,O-heterocycles, and the total synthesis of bioactive heterocyclic natural products.

David Stephens graduated with a B.S. degree in Chemistry from St. Edward's University in Austin, TX. Since 2011 he has been conducting his doctoral studies under the guidance of Prof. Oleg Larionov at the University of Texas at San Antonio. His current work is focused on the catalytic methods of synthesis of N- and N,O-heterocycles and their applications in medicinal chemistry.