Synthesis of meta-Aminophenol Derivatives via Cu-Catalyzed [1,3]-Rearrangement—Oxa-Michael Addition Cascade Reactions

Cu-catalyzed reactions of N-alkoxy-2-methylanilines and alcohols in the presence of catalytic amounts of IPrCuBr and AgSbF6 afforded the corresponding meta-aminophenol derivatives in good to high yields. These reactions proceed via a [1,3]-rearrangement, in which the alkoxy group migrates from the nitrogen atom to the methyl-substituted ortho position, followed by an oxa-Michael reaction of the resulting ortho-quinol imine intermediate.


Scheme 1. Cu-catalyzed reaction of 1a and methanol 2a.
To reduce the number of regioisomers prior to the optimization of reaction conditions, we preliminary screened several starting materials 1. To our delight, the reaction of substrate 1b having a fluorine atom at the para position of the aniline ring at 70 °C generated two regioisomers, 3ba and 6b. Among the NHC ligands examined, IPr exhibited the best reactivity and product selectivity ( Table 1, entry 1), whereas the use of IMes [1,3-Bis(2,4,6-trimethylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene] and SIPr [1,3bis(2,6-diisopropylphenyl)imidazolidine] resulted in low catalytic activity (entries 2 and 3). The reactivity was significantly affected by counteranions; less coordinative counteranions such as hexafluoroantimonate were effective for the present transformations. On the other hand, when AgBF4 and AgNTf2 were used instead of AgSbF6, 3ba was formed in low chemical yields along with a considerable amount of recovered 1b (entries 4 and 5). The reaction in 1,2-dichloroethane (DCE) proceeded quickly to afford product 3ba in the best yield among the solvents examined (entries 6-9). Neither the chemical yield nor the product selectivity was improved when the temperature of the reaction of 1b and 2a was changed from 70 °C to 60 °C and 80 °C, respectively (entries 10 and 11). The reaction in the absence of either Cu or Ag did not yield desired product 3ba; N-methoxyaniline 1b was quantitatively recovered (entries 12 and 13). To reduce the number of regioisomers prior to the optimization of reaction conditions, we preliminary screened several starting materials 1. To our delight, the reaction of substrate 1b having a fluorine atom at the para position of the aniline ring at 70 • C generated two regioisomers, 3ba and 6b. Among the NHC ligands examined, IPr exhibited the best reactivity and product selectivity ( Table 1, entry 1), whereas the use of IMes [1,3-Bis(2,4,6-trimethylphenyl)-1,3-dihydro-2H-imidazol-2-ylidene] and SIPr [1,3-bis(2,6diisopropylphenyl)imidazolidine] resulted in low catalytic activity (entries 2 and 3). The reactivity was significantly affected by counteranions; less coordinative counteranions such as hexafluoroantimonate were effective for the present transformations. On the other hand, when AgBF 4 and AgNTf 2 were used instead of AgSbF 6 , 3ba was formed in low chemical yields along with a considerable amount of recovered 1b (entries 4 and 5). The reaction in 1,2-dichloroethane (DCE) proceeded quickly to afford product 3ba in the best yield among the solvents examined (entries 6-9). Neither the chemical yield nor the product selectivity was improved when the temperature of the reaction of 1b and 2a was changed from 70 • C to 60 • C and 80 • C, respectively (entries 10 and 11). The reaction in the absence of either Cu or Ag did not yield desired product 3ba; N-methoxyaniline 1b was quantitatively recovered (entries 12 and 13).
The protecting group on the nitrogen atom significantly affected the product selectivity ( Table 2, entries 1-4); desired meta-anisidines 3 were obtained with better product selectivity with carbamate-type protecting groups (entries 2-4) than with the amide-type protecting groups (Table 1 entry 6, and Table 2, entry 1). The use of one equivalent of methanol 2a was effective, whereas the use of a large excess (five equivalents) of methanol 2a diminished the catalytic activity (entry 2 versus entry 6). It should be noted that the reaction using half an equivalent of methanol at 80 • C afforded 3da in the 71% yield, suggesting that the methoxy group eliminated from substrate 1d participated in the reaction as a nucleophile (entry 8). Finally, the chemical yield was improved by increasing the scale from 0.2 mmol to 0.5 mmol (entry 9, Appendix A).  The protecting group on the nitrogen atom significantly affected the product selectivity ( Table 2, entries 1-4); desired meta-anisidines 3 were obtained with better product selectivity with carbamate-type protecting groups (entries 2-4) than with the amide-type protecting groups (Table 1 entry 6, and Table 2, entry 1). The use of one equivalent of methanol 2a was effective, whereas the use of a large excess (five equivalents) of methanol 2a diminished the catalytic activity (entry 2 versus entry 6). It should be noted that the reaction using half an equivalent of methanol at 80 °C afforded 3da in the 71% yield, suggesting that the methoxy group eliminated from substrate 1d participated in the reaction as a nucleophile (entry 8). Finally, the chemical yield was improved by increasing the scale from 0.2 mmol to 0.5 mmol (entry 9, Appendix A).   PhCl  70  70  28  <1  2  IMesCuCl AgSbF6  PhCl  70  21  3  76  3  SIPrCuCl AgSbF6  PhCl  70  13  5  47  4  IPrCuBr  AgBF4  PhCl  70  43  13  37  5  IPrCuBr  AgNTf2  PhCl  70  <10  <15  b  6  IPrCuBr  AgSbF6  DCE  70 c  76  22  <2  7  IPrCuBr  AgSbF6  toluene  70  63  18  11  8 IPrCuBr AgSbF6  The protecting group on the nitrogen atom significantly affected the product selectivity ( Table 2, entries 1-4); desired meta-anisidines 3 were obtained with better product selectivity with carbamate-type protecting groups (entries 2-4) than with the amide-type protecting groups (Table 1 entry 6, and Table 2, entry 1). The use of one equivalent of methanol 2a was effective, whereas the use of a large excess (five equivalents) of methanol 2a diminished the catalytic activity (entry 2 versus entry 6). It should be noted that the reaction using half an equivalent of methanol at 80 °C afforded 3da in the 71% yield, suggesting that the methoxy group eliminated from substrate 1d participated in the reaction as a nucleophile (entry 8). Finally, the chemical yield was improved by increasing the scale from 0.2 mmol to 0.5 mmol (entry 9, Appendix A). The optimized conditions ( Table 2, entry 9) were applied to the reactions of various N-methoxyanilines 1, as summarized in Table 3. The reactions of substrates 1g and 1h, having a methyl and a phenyl group, respectively, at the para position, proceeded at 70 • C, affording corresponding meta-anisidines 3ga and 3ha in good yields (entries 1 and 2). Substrate 1j, having a bromo group, was converted into desired product 3ja when the loading amount of the Cu catalyst was increased (20 mol%, entry 7). The chemical Molecules 2023, 28, 4251 6 of 11 yield of 3ia, which has a chloro group at the para position, was slightly improved when the reaction was performed at 90 • C using chlorobenzene instead of DCE as the solvent (entries [3][4][5]. An iodo group (1k) was tolerated under the present reaction conditions, affording multi-substituted meta-aminophenol derivative 3ka in an acceptable yield (entry 8). 3-Anisidine 3la, which has an alkynyl group at the para position of the nitrogen atom, could also be generated (entry 9). In contrast, substrate 1m having a methoxycarbonyl group at the para position, exclusively afforded domino-rearrangement byproduct 5m (entry 10). The reaction of N-ethoxyaniline 1n and ethanol 2b afforded corresponding 3-ethoxyaniline 3nb in a good yield (Scheme 2). Table 3. Cu-catalyzed reactions of N-methoxyanilines 1g-m and methanol 2a a . scale; 17 h. Isolated yield. At 80 °C. The reaction was conducted at a 0.5 mmol scale.
The optimized conditions (Table 2, entry 9) were applied to the reactions of various N-methoxyanilines 1, as summarized in Table 3. The reactions of substrates 1g and 1h, having a methyl and a phenyl group, respectively, at the para position, proceeded at 70 °C, affording corresponding meta-anisidines 3ga and 3ha in good yields (entries 1 and 2). Substrate 1j, having a bromo group, was converted into desired product 3ja when the loading amount of the Cu catalyst was increased (20 mol%, entry 7). The chemical yield of 3ia, which has a chloro group at the para position, was slightly improved when the reaction was performed at 90 °C using chlorobenzene instead of DCE as the solvent (entries [3][4][5]. An iodo group (1k) was tolerated under the present reaction conditions, affording multi-substituted meta-aminophenol derivative 3ka in an acceptable yield (entry 8). 3-Anisidine 3la, which has an alkynyl group at the para position of the nitrogen atom, could also be generated (entry 9). In contrast, substrate 1m having a methoxycarbonyl group at the para position, exclusively afforded domino-rearrangement byproduct 5m (entry 10). The reaction of N-ethoxyaniline 1n and ethanol 2b afforded corresponding 3ethoxyaniline 3nb in a good yield (Scheme 2).  The reaction of N-methoxyaniline 1d and one equivalent of 2-phenethyl alcohol 2c afforded 3-phenylethoxyaniline 3dc in a 51% yield, along with 3da, which was produced through the oxa-Michael addition of methanol derived from 1d, in a 14% yield (Table 4, entry 1). The use of two equivalents of 2c did not improve the product selectivity (entry 2). The reaction of 1d and allyl alcohol 2d also afforded 3-allyloxyaniline 3dd as the major product (entry 3). These results suggest that substrates 1 react more preferentially with external alcohols 2c and 2d than with methanol 2a derived from 1, although the product selectivity depends on the structure of 2. When tert-butanol (2e) and phenol (2f) were employed as an alcohol nucleophile, the reactions gave a mixture of unidentified products (entries 4 and 5). Table 4. Cu-catalyzed reactions of N-methoxyanilines 1d and alcohols 2. The reaction of N-methoxyaniline 1d and one equivalent of 2-phenethyl alcohol 2c afforded 3-phenylethoxyaniline 3dc in a 51% yield, along with 3da, which was produced through the oxa-Michael addition of methanol derived from 1d, in a 14% yield (Table 4, entry 1). The use of two equivalents of 2c did not improve the product selectivity (entry 2). The reaction of 1d and allyl alcohol 2d also afforded 3-allyloxyaniline 3dd as the major product (entry 3). These results suggest that substrates 1 react more preferentially with external alcohols 2c and 2d than with methanol 2a derived from 1, although the product selectivity depends on the structure of 2. When tert-butanol (2e) and phenol (2f) were employed as an alcohol nucleophile, the reactions gave a mixture of unidentified products (entries 4 and 5). Table 4. Cu-catalyzed reactions of N-methoxyanilines 1d and alcohols 2.
The reaction of N-methoxyaniline 1d and one equivalent of 2-phenethyl alcohol 2c afforded 3-phenylethoxyaniline 3dc in a 51% yield, along with 3da, which was produced through the oxa-Michael addition of methanol derived from 1d, in a 14% yield (Table 4, entry 1). The use of two equivalents of 2c did not improve the product selectivity (entry 2). The reaction of 1d and allyl alcohol 2d also afforded 3-allyloxyaniline 3dd as the major product (entry 3). These results suggest that substrates 1 react more preferentially with external alcohols 2c and 2d than with methanol 2a derived from 1, although the product selectivity depends on the structure of 2. When tert-butanol (2e) and phenol (2f) were employed as an alcohol nucleophile, the reactions gave a mixture of unidentified products (entries 4 and 5). A proposed mechanism for the Cu-catalyzed reactions of N-methoxyanilines 1 and alcohol nucleophiles 2 is illustrated in Figure 5a. The cationic copper catalyst coordinates to 1 to form chelate complex 8, and this is followed by an oxidative addition of the N-O bond to the Cu(I) catalyst to form Cu(III) complex 9 [29]. Because of the contribution of canonical form 9′, a C-O bond is formed at the methyl-substituted ortho position, generating ortho-quinol imine intermediates 10 that coordinate to the cationic Cu catalyst. Electrophilically activated ortho-quinol imine intermediates 10 undergo nucleophilic addition of 2 to form Cu enamide species 11. Finally, proton transfer and elimination of methanol 2a give products 3 along with the regenerated cationic Cu catalyst. Byproducts 5 are formed via a Cu-catalyzed [1,2]-rearrangement of ortho-quinol imine intermediates 12 [25]. The product selectivity of the present reaction system was greatly influenced by the para substituent; a fluorine atom (1d) was effective in suppressing the formation of the domino [1,3]/[1,2]-rearrangement product (5,  A proposed mechanism for the Cu-catalyzed reactions of N-methoxyanilines 1 and alcohol nucleophiles 2 is illustrated in Figure 5a. The cationic copper catalyst coordinates to 1 to form chelate complex 8, and this is followed by an oxidative addition of the N-O bond to the Cu(I) catalyst to form Cu(III) complex 9 [29]. Because of the contribution of canonical form 9 , a C-O bond is formed at the methyl-substituted ortho position, generating orthoquinol imine intermediates 10 that coordinate to the cationic Cu catalyst. Electrophilically activated ortho-quinol imine intermediates 10 undergo nucleophilic addition of 2 to form Cu enamide species 11. Finally, proton transfer and elimination of methanol 2a give products 3 along with the regenerated cationic Cu catalyst. Byproducts 5 are formed via a Cucatalyzed [1,2]-rearrangement of ortho-quinol imine intermediates 12 [25]. The product selectivity of the present reaction system was greatly influenced by the para substituent; a fluorine atom (1d) was effective in suppressing the formation of the domino [1,3]/[1,2]rearrangement product (5, Table 2, entry 9). This is possibly because the mesomeric electron-donating effect of the fluorine atom increases the electron density of the carbon next to the quaternary carbon in ortho-quinol imine intermediates 10d/10d , decelerating the [1,2]-rearrangement of the methyl group ( Figure 5(bi)). In contrast, when an ester group was present, the selective formation of the domino rearrangement product (5m) occurred even in the presence of methanol (Table 3, entry 10), presumably because of the undesired [1,2]-rearrangement reaction facilitated by the electron-withdrawing effect ( Figure 5(bii)). In addition, because substrate 1m reacted with stronger carbon nucleophiles, such as 1,3,5-trimethoxybenzene, via [1,3]-rearrangement/Michael addition [26], the low nucleophilicity of alcohols 2 also causes the selective formation of byproduct 5m. Our preliminary DFT calculations for the [1,2]-rearrangement of para-substituted ortho-quinol imine intermediates 10 were in good agreement with these experimental results (Figure 5c). The reaction of N-methoxyaniline 1d and alcohol 2c afforded 3dc (51%) in a higher yield than 3da (14%) derived from the reaction of 1d and generated methanol 2a (Table 4, entry 1). Because the product selectivity is dependent on the structure of external alcohols 2c and 2d, the elimination of methanol 2a from enamide intermediates 11 may be mediated by external alcohol 2, affecting the concentration of methanol 2a. Further investigations to understand the reactivity of key intermediates 10 are under way in our laboratory. results (Figure 5c). The reaction of N-methoxyaniline 1d and alcohol 2c afforded 3dc (51%) in a higher yield than 3da (14%) derived from the reaction of 1d and generated methanol 2a (Table 4, entry 1). Because the product selectivity is dependent on the structure of external alcohols 2c and 2d, the elimination of methanol 2a from enamide intermediates 11 may be mediated by external alcohol 2, affecting the concentration of methanol 2a. Further investigations to understand the reactivity of key intermediates 10 are under way in our laboratory.

Materials and Methods
The general procedure for the Cu-catalyzed reactions of N-methoxyaniline 1d and methanol 2a is as follows: DCE (1.0 mL) and methanol 2a (20.3 µL, 0.5 mmol) were added to a mixture of IPrCuBr (26.6 mg, 0.05 mmol), AgSbF6 (24.0 mg, 0.05 mmol), and 1d (144.8 mg, 0.5 mmol) under an argon atmosphere, and the mixture was stirred at 70 °C for 48 h. After complete consumption of 1d, as monitored via TLC, the mixture was passed through a pad of silica gel with ethyl acetate (50 mL). The solvents were removed in vacuo, and the

Materials and Methods
The general procedure for the Cu-catalyzed reactions of N-methoxyaniline 1d and methanol 2a is as follows: DCE (1.0 mL) and methanol 2a (20.3 µL, 0.5 mmol) were added to a mixture of IPrCuBr (26.6 mg, 0.05 mmol), AgSbF 6 (24.0 mg, 0.05 mmol), and 1d (144.8 mg, 0.5 mmol) under an argon atmosphere, and the mixture was stirred at 70 • C for 48 h. After complete consumption of 1d, as monitored via TLC, the mixture was passed through a pad of silica gel with ethyl acetate (50 mL). The solvents were removed in vacuo, and the crude product was purified via silica gel column chromatography using hexane/ethyl acetate (20/1) as an eluent to afford 3da (123.3 mg, 0.426 mmol, 85% yield) in an analytically pure form (Supplementary Materials).

Conclusions
In conclusion, we have developed a new approach to prepare meta-aminophenol derivatives via Cu-catalyzed cascade reactions involving [1,3]-rearrangement, oxa-Michael addition, and aromatization in an efficient manner. Because a variety of functional groups are tolerated in this transformation, the present method is potentially useful for synthesizing a new class of meta-aminophenol derivatives.

Data Availability Statement:
The data presented in this study are available in this article.