Cu-Catalyzed Arylation of Bromo-Difluoro-Acetamides by Aryl Boronic Acids, Aryl Trialkoxysilanes and Dimethyl-Aryl-Sulfonium Salts: New Entries to Aromatic Amides

We describe a mechanism-guided discovery of a synthetic methodology that enables the preparation of aromatic amides from 2-bromo-2,2-difluoroacetamides utilizing a copper-catalyzed direct arylation. Readily available and structurally simple aryl precursors such as aryl boronic acids, aryl trialkoxysilanes and dimethyl-aryl-sulfonium salts were used as the source for the aryl substituents. The scope of the reactions was tested, and the reactions were insensitive to the electronic nature of the aryl groups, as both electron-rich and electron-deficient aryls were successfully introduced. A wide range of 2-bromo-2,2-difluoroacetamides as either aliphatic or aromatic secondary or tertiary amides were also reactive under the developed conditions. The described synthetic protocols displayed excellent efficiency and were successfully utilized for the expeditious preparation of diverse aromatic amides in good-to-excellent yields. The reactions were scaled up to gram quantities.


Introduction
The amide functional group is abundant in peptides and numerous natural products and is also ubiquitous in a vast range of biologically active compounds, marketed drugs, and a broad spectrum of agrochemicals [1][2][3][4][5][6][7]. The presence of the amide motif or its isosteres condition biological activity of many privileged scaffolds [7]. By recent estimates, almost a quarter of all marketed pharmaceuticals possesses an amide bond, making this functional group the most encountered in medicinal chemistry. Amides are prevalent in advanced materials [7,8], and many life science relevant substances; amides also play pivotal roles in supramolecular chemistry [9,10], molecular recognition [9][10][11], and catalysis [12,13]. The amide functional group can be tuned electronically and conformationally to gain desired structural, physical, and biological properties. The chemistry of amide group is vast, and by its virtue amides can be transformed into many other functional groups [14][15][16][17][18][19][20]. Due to the omnipresence and profound importance of the amide functionality, the development of principally new synthetic routes aiming at installation of the amide structural moiety is of current importance in both modern organic and medicinal chemistry. In this context, many new synthetic routes were elaborated [21][22][23][24]. Among those, it is important to mention such game-changing strategy as aminocarbonylations of aryl halides utilizing CO [25][26][27].
One conceptually underexplored strategy to prepare new amides was the installation of the amide structural unit by the attaching an appropriate substituent onto the prefunctionalized CO-N structural motif bearing a tuned leaving group on the amide carbon. Analysis of the literature revealed that this tactic has been realized using C-N synthons bearing Cl [28] and CHal 3 (Hal = Cl, Br, I) as a leaving group [29][30][31]. These strategies were predominately used for the construction of aromatic amides with different substituents on the nitrogen atom. Another method was developed that is based upon the transition-metalcatalyzed arylation of N-substituted formamides by different aryl-containing reagents, predominantly aryl halides [32][33][34].
Based on a mechanistic consideration, we considered that 2-bromo-2,2-difluoroacetamides 1 would be particularly attractive for the formation of aryl-amides by activation using transition-metal catalysis. Combining the halogens in this particular fashion on the trihaloacetamide enables us to harness the attractive features of copper catalysis and fluoride-mediated catalysis. We set out to explore 2-bromo-2,2-difluoroacetamides 1 in coupling reactions with aryl boronic acids 2 and (aryl)trialkoxysilanes 3 arylation agents as donors of aryl or heteroaryl substituents (Scheme 1a). We hypothesized (Scheme 1b) that using transition-metal-assisted catalysis, a 2-bromo-2,2-difluoroacetamide unit could undergo an oxidative addition on an appropriately tuned by ligands metal nuclei, forming an organometallic intermediate (structure 6) [35][36][37], followed by a rearrangement possibly via a CF 2 -carbene complex 7, which undergoes loss of difluorocarben and simultaneous exchange of Br versus F giving rise to an organometallic (intermediate 8) capable of undergoing reaction with aryl boronic acids or aryl trialkoxysilanes to deliver a new intermediate (9), which after the reductive elimination would result in the formation of a new C-C bond to yield the desired aryl amide (5). An alternative mechanistic pathway could be via copper-intermediate 11 (Scheme 1c), as a result of the reaction between a fluorinated transition-metal catalyst and an aryl boronic acid (or aryl trialkoxysilane). This species could react with a carbon-centered radical 10 to form the species 9, which then decomposes into the final amide product 5. The formation of the radical species 10 would be unusual from the mechanistic point of view. A similar mechanism has, however, been suggested on the instance of palladium-catalyzed carboxylate-assisted ethoxycarboxylation of aromatic acids by ethyl bromodifluoroacetate in a very recent study [38,39]. It is worth noting that the concept of F versus B(OR) 2 (or Si(OR) 3 ) exchange on the copper nuclei, which we are postulating here, was suggested by Giri and Brawn for the mechanism in their copper-catalyzed Suzuki-Miyaura C-C couplings. These protocols were operational not only for boronic esters, but also for a broad range of trialkoxysilanes [40][41][42]. Based on the assumption of a fluoride-bearing Ar-Cu-F intermediate being active (similar to structure 8), and in a view of the resent literature on copper-supported C-C coupling protocols, we envisioned the use of copper catalysts. We also envisioned the preparation of aromatic amides as a result of the C-C coupling between aryl boronic acids, aryl trialkoxysilanes, or sulphonium salts with 2-bromo-2,2-difluoroacetamides according to the general synthetic scenario depicted in the Scheme 1.
than those using metalorganic reagents, thus enabling the creation larger amide structural diversities.

Results and Discussion
We selected three model reactions and performed a set of trial experiments to identify the trends and generalities depicted in Scheme 2 and Tables 1-3. After testing numerous reaction parameters, among which are catalysts, ligands, solvents, and bases, we noticed that some of the copper salts in combination with nitrogen-containing ligands (not indicated in the optimization Tables), in particular solvents, facilitate the expected C-C-coupling reaction and thus the formation of the desired aromatic amide. Furthermore, we succeeded in establishing the optimal reaction conditions for synthetic protocols (a) and (b), which were identical and consisted in the use of CuBr2 (0.1 equiv.), KF (2 equiv.), MgCl2 (1 equiv.) with hexafluoropropanol as the solvent, where all reactions were conducted in ACE pressure tubes at 70 °C for 8 h. One crucial aspect appeared to be the addition of calix [4]arene derivatives, which most probably act as ligands for the coper salt. The best efficiency was observed for the corresponding calix [4]arene L1. The magnesium salt, due to the high affinity of Mg 2+ towards electron rich fluoride ion (hardness of Mg 2+ in terms of the Pearson Hard-Soft acid-base theory), is most probably involved in the activation of one of the C-Hal bonds, like the corresponding C-F bond, by the coordination onto fluorine (where the fluoride ion in turn is a hard base, as per the Pearson Hard-Soft acid-base theory) and formation of the Mg-haloalkane complex [44,45]. The optimized reaction conditions allowed the efficient preparation of the model amide compound 5a in 87% and 90%, respectively (Tables 1 and 2). This success encouraged further exploration of the scope and limitation of these two new protocols. We set out to test the scope and limitations of these coupling reactions by selecting twenty-two 2-bromo-2,2-difluoroacetamides 1 and reacting those with a range of aryl boronic acids 2 (twenty-three different substrates) and aryl trialkoxysilanes 3 (seventeen substrates). In a result of this study, we successfully prepared thirty-one amide derivativities 5 in good-to-excellent yields. We first considered the use of 2-bromo-2,2-difluoroacetamides as a source of the -CO-NR 2 synthon. The only literature example known to date where ethoxycarboxylation of aromatic acids occurs using ethyl bromodifluoroacetate was described recently by Zhao et al. [38]. Similar access was proposed by Shi and co-workers in an alkoxycarbonylation of benzamides utilizing chloroformates [28]. Trifluoroacetyl amides have been used for the construction of aromatic and aliphatic amides via C(O)-CF 3 bond cleavage utilizing the reaction with Grignard reagents [43]. The routes proposed by us utilize commercially or readily available reagents aryl donors and are visibly more atom economic and efficient than those using metalorganic reagents, thus enabling the creation larger amide structural diversities.

Results and Discussion
We selected three model reactions and performed a set of trial experiments to identify the trends and generalities depicted in Scheme 2 and Tables 1-3. After testing numerous reaction parameters, among which are catalysts, ligands, solvents, and bases, we noticed that some of the copper salts in combination with nitrogen-containing ligands (not indicated in the optimization Tables), in particular solvents, facilitate the expected C-C-coupling reaction and thus the formation of the desired aromatic amide. Furthermore, we succeeded in establishing the optimal reaction conditions for synthetic protocols (a) and (b), which were identical and consisted in the use of CuBr 2 (0.1 equiv.), KF (2 equiv.), MgCl 2 (1 equiv.) with hexafluoropropanol as the solvent, where all reactions were conducted in ACE pressure tubes at 70 • C for 8 h. One crucial aspect appeared to be the addition of calix [4]arene derivatives, which most probably act as ligands for the coper salt. The best efficiency was observed for the corresponding calix [4]arene L1. The magnesium salt, due to the high affinity of Mg 2+ towards electron rich fluoride ion (hardness of Mg 2+ in terms of the Pearson Hard-Soft acid-base theory), is most probably involved in the activation of one of the C-Hal bonds, like the corresponding C-F bond, by the coordination onto fluorine (where the fluoride ion in turn is a hard base, as per the Pearson Hard-Soft acid-base theory) and formation of the Mg-haloalkane complex [44,45]. The optimized reaction conditions allowed the efficient preparation of the model amide compound 5a in 87% and 90%, respectively (Tables 1 and 2). This success encouraged further exploration of the scope and limitation of these two new protocols. We set out to test the scope and limitations of these coupling reactions by selecting twenty-two 2-bromo-2,2-difluoroacetamides 1 and reacting those with a range of aryl boronic acids 2 (twenty-three different substrates) and aryl trialkoxysilanes 3 (seventeen substrates). In a result of this study, we successfully prepared thirty-one amide derivativities 5 in good-to-excellent yields. Focusing first on the reactions utilizing aryl boronic acids and aryl trialkoxysilanes, these synthetic protocols were tolerant to numerous functional groups placed on both coupling partners. In particular, both methodologies allowed the coupling of aryl substrates bearing a vast range of electron-withdrawing and electron-donating substituents placed in ortho-, meta-, and para-positions, respectively; among those are alkyl groups, alkoxy groups, Ph, halogens including fluorine, as well as CF3, CF3O, and CF3S groups. Substrates bearing 1-naphthyl, 1-thiophenyl, and 3-pyridyl moieties also showed excellent efficiency with some discrepancy for the formation of the thionyl derivative 5n (Scheme 3). Interestingly, both protocols were operational for aryl substrates bearing diverse ortho substituents (Me, F, Cl, Br, CF3, CF3O). Of note, highly fluorinated boronic acids and aryl trialkoxysilanes were prone to enter those protocols readily delivering the corresponding amides 5g, 5o, 5q. Regarding the reactivity of 2-bromo-2,2-difluoroacetamide counterparts 1, we did not observe any influence on the reaction efficiency of the substituents placed on the amide nitrogen-both alkyl and aryl groups as well as mixed derivatives exerted excellent tolerability within the developed protocols (Scheme 3). These reactions were not affected by changing a substitution pattern on the 2-bromo-2,2difluoroacetamides: Species with alkyl as well as aryl substituents on the amide motif were equally effective within both synthetic protocols (Scheme 2). To further demonstrate the synthetic utility of these methodologies, the gram-scale reactions were successfully performed using 10 mmol of the 2-bromo-2,2-difluoroacetamides, which yielded the expected products in high yields.  Focusing first on the reactions utilizing aryl boronic acids and aryl trialkoxysilanes, these synthetic protocols were tolerant to numerous functional groups placed on both coupling partners. In particular, both methodologies allowed the coupling of aryl substrates bearing a vast range of electron-withdrawing and electron-donating substituents placed in ortho-, meta-, and para-positions, respectively; among those are alkyl groups, alkoxy groups, Ph, halogens including fluorine, as well as CF 3 , CF 3 O, and CF 3 S groups. Substrates bearing 1-naphthyl, 1-thiophenyl, and 3-pyridyl moieties also showed excellent efficiency with some discrepancy for the formation of the thionyl derivative 5n (Scheme 3). Interestingly, both protocols were operational for aryl substrates bearing diverse ortho substituents (Me, F, Cl, Br, CF 3 , CF 3 O). Of note, highly fluorinated boronic acids and aryl trialkoxysilanes were prone to enter those protocols readily delivering the corresponding amides 5g, 5o, 5q. Regarding the reactivity of 2-bromo-2,2-difluoroacetamide counterparts 1, we did not observe any influence on the reaction efficiency of the substituents placed on the amide nitrogen-both alkyl and aryl groups as well as mixed derivatives exerted excellent tolerability within the developed protocols (Scheme 3). These reactions were not affected by changing a substitution pattern on the 2-bromo-2,2-difluoroacetamides: Species with alkyl as well as aryl substituents on the amide motif were equally effective within both synthetic protocols (Scheme 2). To further demonstrate the synthetic utility of these methodologies, the gram-scale reactions were successfully performed using 10 mmol of the 2-bromo-2,2-difluoroacetamides, which yielded the expected products in high yields.    To the general scope and limitations, it is also important to note: (1) Within both described syntactic protocols we tried numerous other N-substituted and N-unsubstituted derivatives of 2-bromo-2,2-difluoroacetic acid, for instance: 2-bromo-2,2-difluoroethanethioamides, 2-bromo-2,2-difluoroacetimidamides, 2-bromo-2,2-difluoroacetohydrazonamides; all these substrates were not prone to enter the developed arylation protocols; (2) Aryl pinacol borates as well as aryl trifluoroborates in the form of potassium salts act as arylation agents in the frames of both synthetic protocols (2 and 4  As the final accord of this work, we turned our attention to aryl sulphonium salts 4. These are donors of aryl groups and are often considered as equivalents of aryl halides, possessing low reduction potentials [46][47][48]. We assumed that those species might have capacity to enter the title synthetic protocol (Scheme 2c). These compounds did not react well under previously optimized reaction conditions, where the model compound 5a was obtained in 47% yield ( Table 3, Entry 1). Thus, we embarked once more on the search for new operational reaction conditions for the model reaction. It is worthwhile to note that in the case of this reaction, we had to increase the amount of copper salt to 0.3 equiv. and add 0.2 equiv. of [Ru(p-cymene)Cl 2 ] 2 , which was superior to other TM co-catalysts (Table 3). Finally, by employing CuBr 2 (0.3 equiv.), [Ru(p-cymene)Cl 2 ] 2 (0.2 equiv.), KF (2 equiv.), MgCl 2 (1 equiv.) and 0.25 equiv. of calix [5]arene derivative (L2), in hexafluoropropanol, the model amide 5a was prepared in 84% yield. Further study of the scope resulted in the preparation of ten amides in total (Scheme 3c).
To gain the insight to the reaction mechanism, we performed several control experiments: (a) Reactions without addition of calixarenes; (b) reactions without CuBr 2 and MgCl 2 ; (c) reactions in the dark and (d) reactions with 2 equiv. and 3 equiv. of TEMPO, which led to the modest decrease of the yield of title model amid compound. All these experiments are depicted in the Tables 1-3.

Materials and Methods
Commercially available starting materials, reagents, catalysts, anhydrous, and degassed solvents were used without further purification. Flash column chromatography was performed with Merck Silica gel 60 (230-400 mesh). The solvents for column chromatography were distilled before the use. Thin layer chromatography was carried out using Merck TLC Silica gel 60 F254 and visualized by short-wavelength ultraviolet light or by treatment with potassium permanganate (KMnO 4 ) stain. 1 H, 13 C, and 19 F-NMR spectra were recorded on a Bruker 250 and 500 MHz at 20 • C. All 1 H-NMR spectra are reported in parts per million (ppm) downfield of TMS and were measured relative to the signals for CHCl3 (7.26 ppm) and DMSO (2.50 ppm). All 13 C{ 1 H}-NMR spectra were reported in ppm relative to residual CHCl 3 (77.00 ppm) or DMSO (39.70 ppm) and were obtained with 1 H decoupling. Coupling constants, J, are reported in Hertz (Hz). Gas chromatographic analyses was performed on Gas Chromatograph Mass Spectrometer GCMS-QP2010 Ultra instrument.
The optimal reaction conditions were identified by microscale high- †hroughput experimentation screening. Parallel synthesis was accomplished in an MBraun glovebox operating with a constant Ar-purge (oxygen and water <5 ppm). Screening reactions were carried out in 10 mL vials using suitable heating blocks. Liquid chemicals were dosed using gas tight micro syringes. Isolation of obtained compounds was achieved by column chromatography on Silica gel.
Under inert atmosphere (glovebox operating with a constant Ar-purge), to an 18 mL ACE pressure tube equipped with a stir bar, consequently, an appropriate 2-bromo-2,2difluoroacetamide (1.0 mmol, 1.0 equiv.), KF (116 mg, 2.0 mmol, 2.0 equiv.), MgCl 2 (95 mg, 1.0 mmol, 1.0 equiv.), appropriate aryl boronic acid (1.3 mmol, 1.3 equiv.), the L1 (0.2 mmol, 0.2 equiv.), and finally CuBr 2 (0.1 mmol, 0.1 equiv.) were placed; then the hexafluoropropanol (0.12 mmol/mL) was added and the reaction vessel was properly capped by Teflon stopper. Finally, the reaction vessel was removed from the glovebox and subjected to heating under vigorous stirring for 8 h. The progress of the reaction was controlled by TLC. After completion, the reaction mixture was evaporated until it reached dryness using a rotary evaporator, the content of the flask was generously treated with distilled water, filtered, and finally properly dried in vacuum. The resulting crude was directly subjected to gradient flash chromatography on silica gel using a mixture of hexane/ethyl acetate as eluent to isolate the desired amide derivative.
The gram scale synthesis was performed on 10 mmol of the starting 2-bromo-2,2difluoroacetamide.
Under inert atmosphere (glovebox operating with a constant Ar-purge), to an 18 mL ACE pressure tube equipped with a stir bar, an appropriate 2-bromo-2,2-difluoroacetamide (1.0 mmol, 1.0 equiv.), KF (116 mg, 2.0 mmol, 2.0 equiv.), MgCl 2 (95 mg, 1.0 mmol, 1.0 equiv.), appropriate trialkoxysilane (1.4 mmol, 1.4 equiv.), the L1 (0.2 mmol, 0.2 equiv.), and finally CuBr 2 (0.1 mmol, 0.1 equiv.) was consequently placed; then the hexafluoropropanol (0.12 mmol/mL) was added and the reaction vessel was properly capped by Teflon stopper. Finally, the reaction vessel was removed from the glovebox and subjected to heating under vigorous stirring for 8 h. The progress of the reaction was controlled by TLC. After completion, the reaction mixture was evaporated until it reached dryness using a rotary evaporator, the content of the flask was generously treated with distilled water, filtered, and finally properly dried in vacuum. The resulting crude was directly subjected to gradient flash chromatography on silica gel using a mixture of hexane/ethyl acetate as eluent to isolate the desired amide derivative.
The gram scale synthesis was performed on 10 mmol of the starting 2-bromo-2,2difluoroacetamide.
Under inert atmosphere (glovebox operating with a constant Ar-purge), to an 18 mL ACE pressure equipped with a stir bar, an appropriate 2-bromo-2,2-difluoroacetamide (1.0 mmol, 1.0 equiv.), KF (116 mg, 2.0 mmol, 2.0 equiv.), MgCl 2 (95 mg, 1.0 mmol, 1.0 equiv.), aryl sulphonium salt (1.6 mmol, 1.6 equiv.), the L2 (0.25 mmol, 0.25 equiv.), [Ru(p-cymene)Cl 2 ] 2 (0.2 mmol, 0.2 equiv.), and finally CuBr 2 (0.3 mmol, 0.3 equiv.) was consequently placed; then the hexafluoropropanol (0.12 mmol/mL) was added and the reaction vessel was properly capped by Teflon stopper. Finally, the reaction vessel was removed from the glovebox and subjected to heating under vigorous stirring for 11 h. The progress of the reaction was controlled by TLC. After completion, the reaction mixture was evaporated until it reached dryness using rotary evaporator, the content of the flask was generously treated with distilled water, filtered, and finally properly dried in vacuum. The resulting crude was directly subjected to gradient flash chromatography on silica gel using a mixture of hexane/ethyl acetate as eluent to isolate the desired amide derivative. The gram scale synthesis was performed on 10 mmol of the starting 2-bromo-2,2-difluoroacetamide.
Flash column chromatography was performed using a mixture of hexane/ethyl acetate 4:1 as an eluent to provide the corresponding amide product.
White solid, mp 116-117 • C. Flash column chromatography was performed using a mixture of hexane/ethyl acetate 8:1 as an eluent to provide the corresponding amide product.
Flash column chromatography was performed using a mixture of hexane/ethyl acetate 7:1 as an eluent to provide the corresponding amide product.
Flash column chromatography was performed using a mixture of hexane/ethyl acetate 8:1 as an eluent to provide the corresponding amide product.