Palladium-Catalyzed Room Temperature Acylative Cross-Coupling of Activated Amides with Trialkylboranes

A highly efficient acylative cross-coupling of trialkylboranes with activated amides has been effected at room temperature to give the corresponding alkyl ketones in good to excellent yields by using 1,3-bis(2,6-diisopropyl)phenylimidazolylidene and 3-chloropyridine co-supported palladium chloride, the PEPPSI catalyst, in the presence of K2CO3 in methyl tert-butyl ether. The scope and limitations of the protocol were investigated, showing good tolerance of acyl, cyano, and ester functional groups in the amide counterpart while halo group competed via the classical Suzuki coupling. The trialkylboranes generated in situ by hydroboration of olefins with BH3 or 9-BBN performed similarly to those separately prepared, making this protocol more practical.


Introduction
Amides are unique and ubiquitous substructures in natural and artificial organic functional molecules, because the strong resonance between the carbonyl and amino groups leads to a highly inert and significantly planar linkage. However, the synthetic utility of amides as an acyl source had remained underdeveloped until the seminal publications in 2015 independently from Garg [1], Szostak [2], and ourselves [3], taking advantage of palladium or nickel catalysis for cleavage of electronically or sterically activated amide C-N bond, and formation of new carbon-carbon or carbon-oxygen bonds. In the past three years, many efforts have been made to expand the scope of amide counterparts, developing a variety of activated amides suitable for the acylative cross-coupling, e.g., N-acyl imides [4][5][6], N-Boc and N-Ts/Ms amides [7][8][9][10][11][12][13][14][15], N-acylsaccharins [16][17][18][19][20], and amides of heteroaromatic compounds [21,22]. Comparably, the carbon-centered nucleophile counterparts are still rather undeveloped, in particular, with respect to alkyl ones, although alkyl ketones have been widely found in biologically important molecules and synthetic building blocks for fine chemicals. In fact, besides the closely related esters [23][24][25], there are only two reports on the acylative cross-coupling of amides with alkyl reagents effected by using nickel catalysts. Garg and co-workers reported the first alkylation of amide derivatives by nickel-catalyzed acylative coupling with alkyl zinc reagents in 2016 [26]. Early this year, Rueping and co-workers described the other nickel-catalyzed cross-coupling of B-alkyl-9-BBN with amides [27]. After extension of the nucleophile counterparts from arylboronic acids to cost-effective diarylborinic acids and tetraarylboronates in palladium-catalyzed acylative cross-coupling of amides [28,29], we are interested in the reactivity of trialkylboranes, which could be readily prepared from alkyl Grignard reagents or olefins via hydroboration. Herein we report a highly efficient palladium-catalyzed acylative cross-coupling of activated amides with trialkylboranes at The catalyst system Pd(PCy3)2Cl2/PCy3, which we previously established for acylative cross-coupling of amides with aryl boron compounds, was investigated, at first, in the presence of K2CO3 as base in THF. Unfortunately, palladium black developed immediately upon heating and no reaction was detected after 6h at 60 °C, indicating the incompatibility of the Pd(PCy3)2Cl2/PCy3 system with the high reducing ability of trialkylboranes. We then turned to the sterically demanding N-heterocyclic carbene (NHC) supported palladium catalysts pioneered by Nolan group [31], for example, 1,3-bis(2,6-diisopropylphenyl)imidazolium chloride (IPr-HCl)/palladium chloride (IPr-HCl/PdCl2) [32], IPr supported 2-((dimethylamino)methyl)phenyl palladium chloride (Palladacycle(IPr)) [33], and IPr and 3-chloropyridine co-supported PEPPSI catalyst (IPr)PdCl2(3-ClPy) [30]. Although the IPr-HCl/PdCl2 system also rapidly decomposed to palladium black and lost activity, the palladacycle/IPr and PEPPSI catalysts provided the desired cross-coupling product 3a in 27% and 61% yields, respectively, from 1a with 1.0 equiv 2a in the presence of 2 equiv K2CO3. To test if dialkylboranes could be used in the acyl alkylation with amides, we also carried out the model reaction by using diethylborinate (Et2B(OMe), 1M in THF) under otherwise identical conditions (Scheme 2). Surprisingly, the C-O cross-coupling product, methyl 4-methoxybenzoate, was obtained in 74% yield, while the C-C coupling product 3a was isolated in low (~5%) yield, implying a much slower transmetalation of alkyl group from boron to palladium than that of alkoxyl group in dialkylborinates. To increase the yield of the desired C-C coupling product in the PEPPSI-catalyzed acyl alkylation of amides, the reaction conditions were optimized with respect to substrate ratio, solvent, base, and reaction temperature, etc. Bases proved to be crucial since no reaction took place in the absence of bases or in the presence of organic base, e.g., NEt3 and pyridine (Table 1, entries 2, 8 and Scheme 1. Screening of palladium catalyst for acyl alkylation of amides with alkylboranes. The catalyst system Pd(PCy 3 ) 2 Cl 2 /PCy 3 , which we previously established for acylative cross-coupling of amides with aryl boron compounds, was investigated, at first, in the presence of K 2 CO 3 as base in THF. Unfortunately, palladium black developed immediately upon heating and no reaction was detected after 6h at 60 • C, indicating the incompatibility of the Pd(PCy 3 ) 2 Cl 2 /PCy 3 system with the high reducing ability of trialkylboranes. We then turned to the sterically demanding N-heterocyclic carbene (NHC) supported palladium catalysts pioneered by Nolan group [31], for example, 1,3-bis(2,6-diisopropylphenyl)imidazolium chloride (IPr-HCl)/palladium chloride (IPr-HCl/PdCl 2 ) [32], IPr supported 2-((dimethylamino)methyl)phenyl palladium chloride (Palladacycle(IPr)) [33], and IPr and 3-chloropyridine co-supported PEPPSI catalyst (IPr)PdCl 2 (3-ClPy) [30]. Although the IPr-HCl/PdCl 2 system also rapidly decomposed to palladium black and lost activity, the palladacycle/IPr and PEPPSI catalysts provided the desired cross-coupling product 3a in 27% and 61% yields, respectively, from 1a with 1.0 equiv 2a in the presence of 2 equiv K 2 CO 3 . To test if dialkylboranes could be used in the acyl alkylation with amides, we also carried out the model reaction by using diethylborinate (Et 2 B(OMe), 1M in THF) under otherwise identical conditions (Scheme 2). Surprisingly, the C-O cross-coupling product, methyl 4-methoxybenzoate, was obtained in 74% yield, while the C-C coupling product 3a was isolated in low (~5%) yield, implying a much slower transmetalation of alkyl group from boron to palladium than that of alkoxyl group in dialkylborinates.

Results and Discussion
Reaction of 4-methoxy-N-methyl-N-tosylbenzamide (1a) with the commercially available triethylborane (2a) (BEt3, 1M in THF) was chosen as the model to establish an optimal catalyst system for the acyl alkylation of activated amides (Scheme 1). Scheme 1. Screening of palladium catalyst for acyl alkylation of amides with alkylboranes.
The catalyst system Pd(PCy3)2Cl2/PCy3, which we previously established for acylative cross-coupling of amides with aryl boron compounds, was investigated, at first, in the presence of K2CO3 as base in THF. Unfortunately, palladium black developed immediately upon heating and no reaction was detected after 6h at 60 °C, indicating the incompatibility of the Pd(PCy3)2Cl2/PCy3 system with the high reducing ability of trialkylboranes. We then turned to the sterically demanding N-heterocyclic carbene (NHC) supported palladium catalysts pioneered by Nolan group [31], for example, 1,3-bis(2,6-diisopropylphenyl)imidazolium chloride (IPr-HCl)/palladium chloride (IPr-HCl/PdCl2) [32], IPr supported 2-((dimethylamino)methyl)phenyl palladium chloride (Palladacycle(IPr)) [33], and IPr and 3-chloropyridine co-supported PEPPSI catalyst (IPr)PdCl2(3-ClPy) [30]. Although the IPr-HCl/PdCl2 system also rapidly decomposed to palladium black and lost activity, the palladacycle/IPr and PEPPSI catalysts provided the desired cross-coupling product 3a in 27% and 61% yields, respectively, from 1a with 1.0 equiv 2a in the presence of 2 equiv K2CO3. To test if dialkylboranes could be used in the acyl alkylation with amides, we also carried out the model reaction by using diethylborinate (Et2B(OMe), 1M in THF) under otherwise identical conditions (Scheme 2). Surprisingly, the C-O cross-coupling product, methyl 4-methoxybenzoate, was obtained in 74% yield, while the C-C coupling product 3a was isolated in low (~5%) yield, implying a much slower transmetalation of alkyl group from boron to palladium than that of alkoxyl group in dialkylborinates. To increase the yield of the desired C-C coupling product in the PEPPSI-catalyzed acyl alkylation of amides, the reaction conditions were optimized with respect to substrate ratio, solvent, base, and reaction temperature, etc. Bases proved to be crucial since no reaction took place in the absence of bases or in the presence of organic base, e.g., NEt3 and pyridine ( To increase the yield of the desired C-C coupling product in the PEPPSI-catalyzed acyl alkylation of amides, the reaction conditions were optimized with respect to substrate ratio, solvent, base, and reaction temperature, etc. Bases proved to be crucial since no reaction took place in the absence of bases or in the presence of organic base, e.g., NEt 3 and pyridine (Table 1,   The stronger inorganic bases, Cs 2 CO 3 (25%) and K 3 PO 4 (33%), gave the lower yields of 3a than that of K 2 CO 3 , albeit the reaction proceeded faster while the weaker ones, e.g., Na 2 CO 3 , NaOAc, and NaHCO 3 , resulted in recovering most of amide substrate 1a (Table 1, entries 3-7). The yield of 3a could be increased to 80% with 1.5 equiv BEt 3 (Table 1, entry 10). Methyl tert-butyl ether (MTBE) appeared to be the choice of solvent, giving 3a in 90% yield (Table 1, entry 12). The reaction occurred even at room temperature and an excellent yield (98%) for 3a was obtained in 24 h (Table 1, entry 14). Given the advantages of room temperature organic synthesis [34], we reinvestigated the model reaction in THF or MTBE with K 2 CO 3 or K 3 PO 4 as the base (Table 1, entries [15][16][17]. The results confirmed the best performance of the combination of MTBE with K 2 CO 3 . Table 1. PEPPSI-catalyzed acyl alkylation reaction of amide 1a with triethylborane. a Molecules 2018, 23, x 3 of 10 9). The stronger inorganic bases, Cs2CO3 (25%) and K3PO4 (33%), gave the lower yields of 3a than that of K2CO3, albeit the reaction proceeded faster while the weaker ones, e.g., Na2CO3, NaOAc, and NaHCO3, resulted in recovering most of amide substrate 1a (Table 1, entries 3-7). The yield of 3a could be increased to 80% with 1.5 equiv BEt3 (Table 1, entry 10). Methyl tert-butyl ether (MTBE) appeared to be the choice of solvent, giving 3a in 90% yield (Table 1, entry 12). The reaction occurred even at room temperature and an excellent yield (98%) for 3a was obtained in 24 h (Table 1, entry 14). Given the advantages of room temperature organic synthesis [34], we reinvestigated the model reaction in THF or MTBE with K2CO3 or K3PO4 as the base (Table 1, entries [15][16][17]. The results confirmed the best performance of the combination of MTBE with K2CO3. Scope and limitations of the PEPPSI-catalyzed room temperature acylative cross-coupling of activated amides with trialkylboranes were briefly explored ( Table 2). Influence of the amide structure was investigated at first under the optimized reaction conditions. Similar to tosyl-activated amide 1a, mesyl (Ms, 1b) analog also reacted efficiently to give 3a in 98% yield, while tert-butyloxycarbonyl (Boc) activated amide (1c) showed comparably lower reactivity (75%) ( Table 2, entries 1 and 2). Reaction of benzamides bearing an electron-withdrawing group, acyl (1e), cyano (1f), or ester (1g), at the para-position of the benzene ring gave the corresponding ketones (3b, 3c, or 3d) in 92%, 70%, or 91% yields, respectively, indicating good functional group compatibility of the PEPPSI-catalyzed acyl alkylation ( Table 2, entries 4-6). When 4-chloro-N-methyl-N-tosylbenzamide (1h) was used as the substrate, the desired acylative cross-coupling product 1-(4-chlorophenyl)propanone was not obtained. Instead, double alkylation product 1-(4-ethylphenyl)propanone (3e) was isolated in 47% yield, which could be increased to 88% yield with 2.5 equiv BEt3 (Table 2, entry 7), similar to our previous investigation on the palladium-catalyzed acylative Suzuki coupling of arylboronic acids [3]. A small ortho-substituent at Scope and limitations of the PEPPSI-catalyzed room temperature acylative cross-coupling of activated amides with trialkylboranes were briefly explored ( Table 2). Influence of the amide structure was investigated at first under the optimized reaction conditions. Similar to tosyl-activated amide 1a, mesyl (Ms, 1b) analog also reacted efficiently to give 3a in 98% yield, while tert-butyloxycarbonyl (Boc) activated amide (1c) showed comparably lower reactivity (75%) ( Table 2, entries 1 and 2). Reaction of benzamides bearing an electron-withdrawing group, acyl (1e), cyano (1f), or ester (1g), at the para-position of the benzene ring gave the corresponding ketones (3b, 3c, or 3d) in 92%, 70%, or 91% yields, respectively, indicating good functional group compatibility of the PEPPSI-catalyzed acyl alkylation (Table 2, entries 4-6). When 4-chloro-N-methyl-N-tosylbenzamide (1h) was used as the substrate, the desired acylative cross-coupling product 1-(4-chlorophenyl)propanone was not obtained. Instead, double alkylation product 1-(4-ethylphenyl)propanone (3e) was isolated in 47% yield, which could be increased to 88% yield with 2.5 equiv BEt 3 (Table 2, entry 7), similar to our previous investigation on the palladium-catalyzed acylative Suzuki coupling of arylboronic acids [3]. A small ortho-substituent at the benzene ring appeared to slightly hamper the coupling. In fact, 1-(o-tolyl)propan-1-one (3f) was isolated in 72% yield from N-methyl-N-tosyl-2-methylbenzamide (1i) ( Table 2, entry 8). Alkyl amides (1j and 1k) reacted similarly to give dialkyl ketones 3g and 3h in 95% and 98% yields, respectively (Table 2, entries 9 and 10). the benzene ring appeared to slightly hamper the coupling. In fact, 1-(o-tolyl)propan-1-one (3f) was isolated in 72% yield from N-methyl-N-tosyl-2-methylbenzamide (1i) (  (1j and 1k) reacted similarly to give dialkyl ketones 3g and 3h in 95% and 98% yields, respectively (Table 2, entries 9 and 10). Tri(n-butyl)borane (2b) reacted with 1a similarly to triethylborane (2a) while no reaction of tricyclohexylborane was observed, implying the failure of transmetalation of secondary-alkyl group from boron to palladium. Hydroboration of alkenes represents an alternative route to primary alkyl boranes. However, the long-chain primary alkyl boranes prepared by hydroboration of alkenes are generally contaminated by the presence of branched-isomers due to the non-regiospecific addition of B-H to C-C double bonds [35]. Due to the inertness of the secondary-alkyl group in alkylboranes, reaction of 1a with tri(octyl)boranes prepared by hydroboration of 1-octene proceeded comparably to that of tri(n-octyl)borane (2c) from Grignard reagent [36], giving 3j in 91% and 99% yields, a Reaction was run at 0.5 mol scale with respect to amides with 1.5 equiv trialkylboranes; b Isolated yield; c With 2.5 equiv BEt 3 ; d B(n-C 8 H 17 ) 3 and branched isomers in situ generated from hydroboration of octene with 1.0 equiv BH 3 in THF at room temperature was used.

General Procedure for the PEPPSI-Catalyzed Cross-Coupling of N-Methyl-N-Tosylamides with Trialkylboranes
A Schlenk tube (20 mL) charged with amide (0.5 mmol), PEPPSI (0.025 mmol, 5 mol%), and K2CO3 (2 equiv) was degassed and then refilled with nitrogen, three times. Then, solvent MTBE (6.0 mL) was added via syringe followed by 0.75 mL (1.5 equiv) borane solution (1 M in THF), which was commercially available or was in situ prepared by hydroboration of olefin. The resulted mixture was stirred for 24 h at room temperature under N2 atmosphere. The reaction was quenched with iced water (5 mL) and extracted with MTBE (2 × 5 mL). The combined MTBE extracts were dried over anhydrous Na2SO4. After filtration, solvents were removed by rotavapor to afford the crude product,

General Procedure for the PEPPSI-Catalyzed Cross-Coupling of N-Methyl-N-Tosylamides with Trialkylboranes
A Schlenk tube (20 mL) charged with amide (0.5 mmol), PEPPSI (0.025 mmol, 5 mol%), and K 2 CO 3 (2 equiv) was degassed and then refilled with nitrogen, three times. Then, solvent MTBE (6.0 mL) was added via syringe followed by 0.75 mL (1.5 equiv) borane solution (1 M in THF), which was commercially available or was in situ prepared by hydroboration of olefin. The resulted mixture was stirred for 24 h at room temperature under N 2 atmosphere. The reaction was quenched with iced water (5 mL) and extracted with MTBE (2 × 5 mL). The combined MTBE extracts were dried over anhydrous Na 2 SO 4 . After filtration, solvents were removed by rotavapor to afford the crude product, which was purified by flash column chromatography on silica gel using petroleum ether/EtOAc as the eluents.

Conclusions
In summary, we have developed a highly efficient acylative cross-coupling of trialkylboranes with activated amides by using 1,3-bis(2,6-diisopropyl)phenylimidazolylidene and 3-chloropyridine co-supported palladium chloride, the PEPPSI catalyst, under mild conditions. Bases appeared to play a key role in the reaction, among which K 2 CO 3 performed best in MTBE at room temperature. The reaction proceeded to give alkyl ketones in good to excellent yields, tolerating a variety of functional groups in the amide counterpart. Unlike the high-order arylboron compounds, in which all the aryl groups react effectively, only one of the three primary alkyl groups in trialkylboranes could be used as alkyl source for the acyl alkylation. The trialkylboranes generated in situ by hydroboration of olefins with BH 3 or 9-BBN performed comparably to those separately prepared. This protocol complements ketone synthesis via palladium-catalyzed acylative cross-coupling of amides, providing a feasible access to both monoalkyl and dialkyl ketones.