Enantioconvergent Chan–Lam Coupling: Synthesis of Chiral Benzylic Amides via Cu-Catalyzed Deborylative Amidation

Abstract

The Chan–Lam coupling represents one of the most effective methods for constructing C–N bonds due to its mild reaction conditions and broad functional group compatibility. However, asymmetric versions of this transformation to forge C(sp³)–N bonds have remained elusive due to the need for a general mechanistic framework to engage alkylboron reagents. Herein, we demonstrate the first enantioconvergent Chan–Lam C(sp³)–N coupling using synthetically modular alkylboronic pinacol esters. The reported transformation imparts a high degree of enantioselectivity and tolerates a wide range of functional groups, heterocycles, and pharmaceutically relevant frameworks. The reaction mechanism was investigated through Hammett analysis, radical clock tests, and radical trapping experiments. Density functional theory (DFT) calculations support a radical-relay pathway: oxidative homolysis of the C–B bond generates prochiral alkyl radicals, which are functionalized by in situ generated Cu(II) intermediate through an inner-sphere mechanism.

Graphical Abstract

1. INTRODUCTION

The oxidative Chan–Lam amination of organoboron reagents represents an effective cross-coupling strategy for synthesizing active pharmaceutical ingredients.^1–9 However, despite the widespread use of the Chan–Lam coupling reactions for aryl amine synthesis, its application in the enantioselective construction of C(sp³)–N bonds remains elusive.^10–13 This challenge stems from the often-reliance on a super-stoichiometric amount of copper and the need for a general mechanistic framework to engage alkylboron reagents. Yet, such a transformation could unlock powerful synthetic disconnections by combining the generality of Chan–Lam coupling with the synthetic versatility of alkyl-boron reagents to construct complex alkyl frameworks (Figure 1A).^14–18 Herein, we disclose the first enantioconvergent Chan–Lam C(sp³)–N coupling reaction facilitated by ligand design and proceeding through a radical-mediated deborylative functionalization pathway.

Figure 1.

(A) Confronting the challenging yet synthetically advantageous asymmetric Chan–Lam C(sp³)–N coupling reaction. (B) Key disconnection for expedited access to chiral benzyl amides as essential pharmaceutical substructures. The reaction orthogonality of the deborylative approach is highlighted by multiple accessible C–H bonds. (C) Recent advances in Cu-catalyzed chiral benzylic amide synthesis by activating the most accessible C–H bond.^35–37 (D) Engaging alkyl boron reagents in enantioconvergent amidation by addressing two mechanistic challenges. Proposed Cu-catalytic cycle for enantioconvergent Chan–Lam amidation of alkylboronic pinacol esters through a radical relay mechanism.

We reasoned that the deborylative cross-coupling between alkylboronic pinacol esters and amides would provide an advantageous strategy for accessing chiral benzylic amides—a privileged class of pharmaceutical substructures (Figure 1B).^19–24 This protocol not only provides a streamlined and modular alternative to conventional multi-step procedures that rely on chiral auxiliaries^25–27 or resolution processes,^{23, 24} but also serves as an orthogonal strategy to emerging catalytic strategies,^28–33 such as those involving C–H functionalization recently reported by Zhou^34–36 and Kramer (Figure 1C).³⁷ The deborylative cross-coupling approach offers a particularly effective solution for substrates featuring multiple reactive positions in established methods, such as those highlighted in Figure 1B. We envisioned that this transformation could be achieved by addressing two key mechanistic challenges: (1) the homolytic cleavage of C–B bonds in alkylboronic pinacol esters and (2) the enantioselective functionalization of alkyl radicals via a Cu(II) amide intermediate (Figure 1D). The synthetic versatility of alkylboron reagents enabled by their empty boron p-orbital comes at the cost of a high oxidation potential.^38–40 We previously addressed this challenge through an in situ generated aminyl radical to facilitate C–B bond homolysis via an inner-sphere mechanism.^{40, 41} We hypothesized that highly reactive organic radicals formed through a radical relay pathway^42–44 (step a) could be utilized for a similar process (step b). The resulting prochiral alkyl radicals would react with an in situ generated Cu(II) amide intermediate, yielding the optically active product (step c-d). We herein report the successful implementation of this strategy for an enantioconvergent deborylative amidation protocol and the corresponding mechanistic studies to guide our future expansion of this asymmetric Chan-Lam-type cross-coupling platform.

2. RESULTS AND DISCUSSION

2.1. Reaction Development and Optimization

We began our investigation by examining the deborylative coupling between alkylboronic pinacol ester 1a and benzamide (B_zNH₂). We explored chiral bixoxazoline (BOX) ligands, various copper sources, chemical oxidants, and reaction solvents (see Supporting Information, Table S1 for full optimization details). We initially identified that a combination of CuCl, sodium tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (NaBAr^F₄), and ligand L1 afforded product 2a with high yield and good stereocontrol (ligand optimization, Table 1). Modification at the meso-position resulted in heightened enantioselectivity (L1–L3). To guide our ligand design, we obtained the solid-state structure of ionic complex [(L3)Cu(MeCN)](PF₆) (P1). We identified two potential strategies to enhance the reaction selectivity: (1) since the meso-substituents are oriented toward the metal center, extended substituents such as adamantyl groups could enforce the ligand-substrate interaction;⁴⁵ and (2) distal bulky alkyl substituents could facilitate non-covalent interactions with incoming radicals and further improve the reaction stereoselectivity.^46–48 After an extensive ligand design campaign (L4–L6, Table S2), we discovered that the use of L6, which features 3,5-di-tert-butylphenyl and 4-adamantylbenzyl groups, led to excellent enantioselectivity and yield. Computational analysis suggests that meso-substitution by 4-adamantylbenzyl groups sterically blocks two diagonally opposing quadrants, likely contributing to the high level of selectivity (Scheme S21; see SI for details).

Table 1.

Optimization of the Enantioconvergent Chan–Lam Amidation of Alkylboronic Pinacol Ester 1a^a

^aReaction conditions unless otherwise noted: 2-(1-([1,1’-biphenyl]-4-yl)ethyl)-4,4,5,5-tetramethyl-1,3,2-dioxaborolane (1a, 0.10 mmol, 1.0 equiv), BzNH₂ (0.20 mmol, 2.0 equiv), 4-picoline (0.02 mmol, 20 mol%), specified Cu and ligand mixture, dicumyl peroxide (0.25 mmol, 2.5 equiv), and chlorobenzene (0.25 ml, 0.4 M); reaction yields were determined by ¹H NMR spectroscopy of the crude product mixture using 1,1,2,2-tetrachloroethane as an internal standard (see SI for details). The enantiomeric ratio (er) of product 2a was determined by chiral high-performance liquid chromatography (HPLC) analysis.

^bSolid-state molecular structure of P1 with thermal ellipsoids at 50% probability level. Color scheme: Cu, gold; N, blue; O, red; C, gray; H, white.

Compared to other copper sources tested, a combination of CuCl and NaBAr^F₄ yielded the most optimal result (entries 2–3, Table 1).^{36, 49–51} Among the oxidants screened, the feedstock chemical dicumyl peroxide provided the highest yield compared to other oxidants commonly employed in radical relay transformations (entries 4–6). Using chlorobenzene as the reaction solvent further improved the yield (entries 7–8). The presence of a Lewis basic additive proved beneficial,^{41, 52} with optimal results achieved using a sub-stoichiometric amount of 4-picoline (20 mol%, entries 9–12), which we propose promotes the NaBAr^F₄-mediated halide abstraction from CuCl and facilitates the generation of a cationic catalyst.^{36, 49–51} The reaction is the most efficient at 60 °C (entries 13–14).

2.2. Substrate Scope of the Enantioconvergent Deborylative Amidation Protocol

With the optimal reaction conditions established for the formation of 2a, we next explored the functional group tolerance and the scope of compatible alkylboronic pinacol esters (Table 2) and amides (Table 3). The reaction demonstrated broad substrate compatibility, proceeding efficiently in the presence of both electron-donating and -with-drawing groups on either coupling partner. A variety of pharmaceutically relevant heterocycles were well-tolerated, including pyridine (3t, 3ah), piperidine (3ae), quinoline (3e), azetidine (3i, 3ac), pyrrole (3s), pyrazole (3k), thiazole (3b), oxazole (3al), tetrazole (3c), benzoxazole (3m), oxetane (3o), tetrahydropyran (3ad), dibenzofuran (3r), thiophene (3w), and dioxolane (3g). The absolute configuration of the major enantiomer was determined by a combination of optical rotation measurements³⁶ and the solid-state structures of products 3b and 3w.

Table 2.

Scope of Alkylboronic Pinacol Esters in Enantioconvergent Chan–Lam Amidation^a

^aYields were obtained using alkylboronic pinacol ester (1, 0.50 mmol) unless otherwise noted. The enantiomeric ratio (er) was determined by chiral HPLC analysis.

^bSolid-state molecular structure for 3b with thermal ellipsoids at 50% probability level. Color scheme: S, yellow; N, blue; C, gray; O, red; H, white.

Table 3.

Amide Scope in Enantioconvergent Chan–Lam Amidation^a

^aYields were obtained using alkylboronic pinacol ester (1, 0.50 mmol) unless otherwise noted. The enantiomeric ratio (er) was determined by chiral HPLC analysis.

^bReaction carried out on a 5.00 mmol scale of the pinacol boronic ester substrate.

^cSolid-state molecular structure for 3u with thermal ellipsoids at 50% probability level. Color scheme: S, yellow; N, blue; C, gray; H, white.

^dReaction carried out using L3.

^eReactions carried out using 2.0 equiv of 1a and 1.0 equiv of the corresponding amides.

Benzylic boronic pinacol esters with ortho- (3m), meta- (3g, 3r), and para- (3b, 3j, 3k, 3t) substituents served as suitable substrates. Reactions with aryl (3w, 3ag-3al) and 1° (3z), 2° (3y, 3aa, 3ad, 3am), or 3° (3ab, 3ac, 3ae, 3an) alkyl amides delivered high yields and excellent enantioselectivity. The amidation of substrates containing electron-deficient aza-heterocycles (3e, 3k, 3t, 3ah) proceeded smoothly without any detectable side products from radical addition. Deborylative amidation yielded exclusively product 3l with no side reactions detected, whereas attempts to access 3l through established protocol resulted in a mixture of secondary and primary benzylic amidation products.³⁷ High yields and selectivity were obtained with substrates featuring tertiary (3b, 3t), weak benzylic (3c, 3e, 3f, 3n, 3s, 3z, 3ai), and allylic (3d, 3j) sites that are likely to be competitive in C–H functionalization protocols.^{36, 37} The amidation of substrates containing alkyl (3h) and activated benzyl (3ai) halides proceeded without detectable halogen abstraction. These results highlight the orthogonality of this amidation protocol to existing asymmetric radical cross-coupling reactions.^{32, 36, 37, 53} While ligand L6 delivers high yields and enantioselectivity in the amidation of alkylboronic pinacol esters bearing methyl groups (1a, 1k–m, 1r), substrates with bulkier alkyl substituents (1b–j, 1n–q, 1s–v) gave low yields, likely due to excessive steric repulsion between the ligand and substrate. To address this, we employed the less sterically demanding ligand L3, which afforded consistently high yields and enantioselectivities for these more hindered substrates.

The transformation also exhibits high selectivity between different types of alkylboronic pinacol esters. For substrates containing additional aryl (3f) or primary alkyl boronic esters (3p), only the benzylic functionalization product was obtained. The broad applicability of this protocol was further demonstrated through the derivatization of pharmaceuticals, including Vitamin A derivative Adapalene (3aj), nonsteroidal anti-inflammatory Oxaprozin (3al), antiepileptic Valpromide (3am), and lipid-lowering agent Gemfibrozil (3an). To showcase the synthetic utility of the enantioconvergent amidation, the synthesis of 3a was scaled up to 5.0 mmol, resulting in comparable yield and enantioselectivity to the 0.5 mmol scale reactions.

After successfully implementing the new amidation strategy across a wide range of substrates, we sought to investigate the reaction mechanism and lay the groundwork for expanding the asymmetric Chan–Lam coupling platform to incorporate additional coupling partners. By integrating experimental and theoretical approaches, our objectives were to (1) analyze the stereoablative homolytic cleavage of C–B bonds and (2) explore the potential pathway for Cu-mediated single-electron C–N bond formation.⁵⁴

2.3. Mechanistic Investigation into the Homolytic Activation of Alkylboronic Pinacol Esters

We began our mechanistic investigation by probing the stereoablative C–B bond homolysis. Amidation of radical clock substrate 4 led to the formation of both the direct amidation product 5a and the ring-opened product 5b, indicating the intermediacy of an alkyl radical (Figure 2A). Radical trapping experiments were carried out using TEMPO or CBrCl₃ (Figure 2B). In the presence of TEMPO, amidation product 3a was observed in addition to the corresponding TEMPO-coupled product (6, 19%).⁵⁵ The addition of CBrCl₃ completely inhibited 3a formation, and 24% of (1-bromoethyl)benzene (7) was detected. Results from these two radical trapping experiments further corroborate the involvement of an alkyl radical intermediate. Subjecting enantiomerically enriched (R)- or (S)-1a to the amidation conditions resulted in the generation of product 3a with a similar enantiomeric ratio favoring the same major enantiomer, indicating the loss of stereochemical information prior to C–N bond formation (Figure 2C).

Figure 2.

Mechanistic probes for the asymmetric Chan–Lam amidation (Reactions were carried out on a 0.10 mmol scale. Yields were determined by ¹H NMR spectroscopy of the crude product mixture using 1,1,2,2-tetrachloroethane as an internal standard.). (A) The proposed alkyl radical formation is supported by radical ring opening of the cyclopropyl substituent in substrate 4. (B) The alkyl radical intermediates can be intercepted using radical traps. (C) Stereochemical information is lost after the generation of an alkyl radical.

We envisioned two potential S_H2 pathways for the homolytic cleavage of C–B bonds in alkylboronic pinacol esters: (1) the direct participation of cumyloxy radical 8 or (2) the formation of an amidyl radical 9 through hydrogen atom transfer (HAT) with the amide substrate (Scheme S10).^56–58 To assess the feasibility of these pathways, we analyzed the reaction between 1a and benzamide using density functional theory (DFT). The results showed that single-electron activation of 1a using either cumyloxy radical 8 or amidyl radical 9 is highly exergonic (ΔG = −44.2 kcal/mol with 8 and ΔG = −34.7 kcal/mol with 9), with activation barriers of 14.5 kcal/mol (TS0) and 15.2 kcal/mol (TS3), respectively. However, the generation of amidyl radical 9 is endergonic (ΔG = 8.7 kcal/mol) and subjected to an activation barrier of 22.0 kcal/mol (TS2). Overall, these results indicate that the direct involvement of cumyloxy radical 8 in generating 1-phenylethyl radical 10 is kinetically preferred (see SI for details).^{41, 52}

2.4. Cu-Mediated Pathways for C–N Bond Formation

Computational studies were carried out to investigate the C–N bonding-forming mechanism and to elucidate the origins of enantioconvergence. Dispersion-corrected DFT calculations were performed at the PW6B95-D3(BJ)/def2-TZVP-SMD(PhCl)//PW6B95-D3(BJ)/def2-SVP-SMD(PhCl) level of theory. We modeled the coupling between alkylboronic pinacol ester 1a and benzamide 2a with truncated but functional ligand L2 (Figure 3A). Computational results⁵⁹ indicate that the formation of divalent copper intermediate 11 is exothermic and proceeds through oxidation with dicumyl peroxide, followed by transmetalation with benzamide 2a (ΔG = −2.5 kcal/mol, Scheme S13). The association of alkyl radical 10 and divalent copper intermediate 11 is slightly endergonic, forming the formal Cu(III) intermediates 12(R) (ΔG = 2.8 kcal/mol) and 12(S) (ΔG = 1.6 kcal/mol). The subsequent C–N bond formation proceeds via reductive elimination transition state structures (TSs) TS4(R) and TS4(S) with retention of configurations at the benzylic positions. A ΔΔG^‡ of 1.9 kcal/mol in favor of the S-enantiomer was observed, in line with experimental observations. Close inspection of the reductive elimination TSs (Figure 3B) ascribed the preference toward S-enantiomer to a stronger N_amideH–π interaction in TS4(S) (2.58 Å vs 2.74 Å in TS4(R)).^{60, 61} In addition to ligand-substrate non-covalent interactions, polarity matching between alkyl radicals and their coupling partners can significantly influence the reaction efficiency.^62–64 To examine the polarity requirements in the C–N bond-forming process, we conducted Hammett analyses using para-substituted benzyl boronic pinacol esters and benzamides. A binary competition approach^65–67 was employed, using parent substrates 1a and 2a as references, respectively (see SI for details). Correlation between the Hammett parameters (σ_p)⁶⁸ of substituted alkylboronic pinacol esters and reaction rates revealed a positive slope (Figure 4A). This observation can be ascribed to (1) a favorable association between the cumyloxy radical 8 and more electrophilic benzyl boronic esters prior to C–B bond cleavage and (2) more electrophilic benzyl radicals increasing the efficiency of C–N coupling. Correspondingly, electron-deficient amides result in slower transformations compared to benzamide likely due to a polarity mismatch (Figure 4B). However, we also observed slower rates with electron-rich amides, leading to a change in the slope of the Hammett correlation. A concave down Hammett plot indicates a change in the rate-limiting step.⁶⁹ We propose that the lower acidity of more electron-rich amides results in slower transmetalation. Building on the mechanistic framework, efforts are underway to develop asymmetric Chan–Lam-type cross-coupling protocols to accommodate a broad range of nucleophiles.

Figure 3.

Computational investigation into the C–N bond forming mechanism. (A) Potential energy surface of amidation via an inner-sphere pathway. (B) Reductive elimination transition state structures reveal favorable non-covalent interactions in TS4(S). (Red: N_amideH–π distance; green: N_amide–C_benzyl distance; blue: C_benzylH–π distance)

Figure 4.

Elucidating the polarity requirements of the Chan–Lam-type amidation through competition Hammett analysis.

3. CONCLUSION

In summary, we present the first asymmetric Chan-Lam C(sp³)–N coupling protocol. This method is compatible with a wide range of functional groups and heterocycles. The amidation product can be obtained on a gram scale with high enantioselectivity. Experimental and computational studies support a radical-relay pathway through an inner-sphere C–N coupling mechanism, where the stereoselectivity is ascribed to ligand-substrate non-covalent interactions.

Supplementary Material

The Supporting Information is available free of charge via the Internet at http://pubs.acs.org.

Experimental procedures and characterization data for all new compounds, including NMR spectra, HPLC traces, computational details, and Cartesian coordinates of all computed structures (PDF)