Wingtip-Flexible N-Heterocyclic Carbenes: Unsymmetrical Connection between IMes and IPr

Abstract

IMes (IMes = 1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene) and IPr (IPr = 1,3-bis(2,6-diisopropylphenyl)imidazol-2-ylidene) represent by far the most frequently used N-heterocyclic carbene ligands in homogenous catalysis, however, despite numerous advantages, these ligands are limited by the lack of steric flexibility of catalytic pockets. We report a new class of unique unsymmetrical N-heterocyclic carbene ligands that are characterized by freely-rotatable N-aromatic wingtips in the imidazol-2-ylidene architecture. The combination of rotatable N–CH₂Ar bond with conformationally-fixed N–Ar linkage results in a highly modular ligand topology, entering the range of geometries inaccessible to IMes and IPr. These ligands are highly reactive in Cu(I)-catalyzed β-hydroboration, an archetypal borylcupration process that has had a transformative impact on the synthesis of boron-containing compounds. The most reactive Cu(I)–NHC in this class has been commercialized in collaboration with MilliporeSigma to enable broad access of the synthetic chemistry community. The ligands gradually cover %V_bur geometries ranging from 37.3% to 52.7%, with the latter representing the largest %V_bur described for an IPr analogue, while retaining full flexibility of N-wingtip. Considering the modular access to novel geometrical space in N-heterocyclic carbene catalysis, we anticipate that this concept will enable new opportunities in organic synthesis, drug discovery and stabilization of reactive metal centers.

Graphical Abstract

We report a new class of unsymmetrical N-heterocyclic carbene ligands that are characterized by freely-rotatable N-aromatic wingtips in the imidazol-2-ylidene architecture. The combination of a rotatable N–CH₂Ar bond with the conformationally-fixed N–Ar linkage results in a highly modular ligand topology, entering the range of geometries inaccessible to IMes and IPr.

Introduction

Since the seminal studies by Arduengo in the 1990s,¹ N-heterocyclic carbenes have had a major impact on chemical catalysis.^2,3 The capacity to stabilize metal centers by strong s-donation⁴ in combination with versatile steric environment⁵ has provided an unlimited access to novel catalytic methods, many of which have now been widely adopted by both the industry and academia. In this context, imidazol-2-ylidene-based IMes (1,3-bis(2,4,6-trimethylphenyl)imidazol-2-ylidene)^6a and IPr (1,3-bis(2,6-diisopropylphenyl)imidazol-2-ylidene)^6b represent by far the most frequently used N-heterocyclic carbene ligands in homogenous catalysis (Figure 1A).^2,3 These ligands are characterized by strong σ-donation (IMes: TEP = 2051, IPr: 2052, [Ir(NHC)(CO)₂Cl])⁴ and considerable sterics (IMes: %V_bur = 37.9%, IPr: 47.6%, [Cu(NHC)Cl]).⁵

Figure 1. Context of this work:

(a) symmetrical imidazol-2-ylidene-based N-heterocyclic carbene ligands. (b) Unsymmetrical, self-adjustable imidazol-2-ylidenes. (c) Cu-catalyzed borylcupration of π-systems.

The combination of these properties provides the most favorable reactivity for a variety of metal centers and catalytic manifolds.^3,7 However, the major limitation of these symmetrical imidazol-2-ylidenes is the lack of steric flexibility of catalytic pockets.⁵ The lack of flexibility of N-Ar wingtips of IMes and IPr prevents a facile access to new ligand geometries that are beneficial for stabilization of metal centers and catalysis.⁸

As part of studies on ligand development,⁹ we report a new class of unique unsymmetrical N-heterocyclic carbene ligands that are characterized by freely-rotatable N-aromatic wingtips in the imidazol-2-ylidene architecture (Figure 1B).

There are several noteworthy findings of our study: (1) The combination of a rotatable N–CH₂Ar bond with the conformationally-fixed N–Ar linkage results in a highly modular ligand topology, entering the range of geometries inaccessible to IMes and IPr. (2) These ligands are highly reactive in Cu(I)-catalyzed β-hydroboration of alkynes, an archetypal borylcupration process that has had a transformative impact on the synthesis of boron-containing compounds (Figure 1C).^10,11 (3) The synthetic utility is further demonstrated in the late-stage hydroboration of complex pharmaceuticals, enabling to introduce boron to drug molecules with excellent functional group tolerance.¹² (4) The most reactive Cu(I)–NHC complex^9d,13 in this class has been commercialized in collaboration with MilliporeSigma to enable broad access of the synthetic chemistry community (no. 930873, NHC·HCl salts: no. 930202, no. 930040).¹⁴ We anticipate that this concept will enable new opportunities in organic synthesis, drug discovery and stabilization of reactive metal centers.

Results and Discussion

Our interest in expanding the range of geometries inaccessible to the common NHCs (Figure 1A–B) originated in electrophilic functionalization reactions using group 11 metals.⁹ These reactions convert π-systems derived from feedstock materials, such as olefins, alkynes, allenes, heterocycles, arenes, conjugated systems, to more complex materials used in the synthesis of complex molecules. We realized that especially Cu(I)–NHC borylcupration¹⁰ represents an extremely versatile approach for the synthesis of complex products by simultaneously introducing a bi-functional motif that can be engaged in the reactions with traditional electrophiles or merged with transition-metal-catalysis through transmetallation (Figure 1C).

In general, steric and electronic requirements for NHC ligands supporting group 11 metals at +1 oxidation state are distinct from the traditional cross-coupling chemistry utilizing Pd(0), where the requirement for (1) strong σ-donation,⁴ and (2) bulky wingtip substituents⁵ to facilitate oxidative addition and reductive elimination are the predominant directions in ligand development. In contrast, the use of strongly σ-donating NHC ligands that provide sufficiently strong metal-carbene bonds in the presence of limited steric hindrance has dominated ligand design for linear group 11 metal–NHC complexes. In this respect, IMes and IPr have been the most successful NHC ligands to date, wherein further modification of sterics and/or electronics resulted in a significant decrease of catalytic activity.¹³

The discovery of new, more active classes of NHC catalysts has been hindered by steric limitations of the symmetrical imidazol-2-ylidene scaffold. Considering NHC–metal catalysis, we hypothesized that unsymmetrical wingtip modification of the imidazol-2-ylidene scaffold could permit for (1) a rational modification of the buried volume and electronic properties by N-flexibility of the scaffold, which in turn could be further tuned by the backbone substitution. This approach could further allow for (2) a significantly improved synthesis of sterically-bulky NHC ligands, which is one of the major challenges in ligand synthesis, and (3) lead to exploration of new ligand space unavailable in in classical symmetrical imidazol-2-ylidene and related ligands. Our initial research focused on the synthesis and reactivity of Cu(I)–NHC complexes. Ligands and complexes selected for this study are presented in Figure 2. These ligands have been selected based on the combination of a fixed N–Ar wingtip in analogy to IPr and IMes (Ar = Dipp, Mes) with a flexible N–CH₂Ar wingtip (Ar = Ph, Mes) in the presence and absence of steric backbone substitution. The selection was based on route modularity and the potential for introduction of diverse peripheral substitution.

Figure 2.

Cu(I)–NHC complexes.

Ligand synthesis is presented in Scheme 1. The synthesis is based on the method for the synthesis of N-arylimidazoles. Although the literature procedure is low yielding and requires tedious chromatographic purification, at present, this is the method of choice for the synthesis of N-arylimidazoles and related NHC ligands.^15,16

Scheme 1.

Synthesis of Unsymmetrical NHC Precursors (see SI for detailed conditions).

We have developed two procedures for the synthesis of unsymmetrical ligands 2a–2e (Scheme 1A). First, a two-step method has been optimized to afford unsymmetrical imidazolium salts 2a–2e in a protocol involving the reaction of butane-2,3-dione (1 equiv) with the corresponding aniline (1.2 equiv) in the presence of NH₄OAc (1.2 equiv), paraformaldehyde (1 equiv) and acetic acid (5 equiv) in CHCl₃ at 60 °C. Alternatively, the reaction of glyoxal (1 equiv), aniline (1 equiv), paraformaldehyde (2 equiv) in the presence of NH₄Cl (2 equiv) and H₃PO₄ (5 equiv) in refluxing methanol was used for the synthesis of 4,5-unsubstituted N-Ar imidazoles. This step was followed by alkylation of N-Ar imidazole (1 equiv) with the corresponding halide (1 equiv) in THF at reflux. Although this sequence could be routinely performed on 10 mmol scale, the method was judged less suitable for large scale synthesis.

Thus, we have developed a chromatography-free procedure that could be routinely performed on a decagram scale (Scheme 1B). The reaction of butane-2,3-dione (1 equiv) and aniline (1 equiv) in the presence of NH₄OAc (1.1 equiv), paraformaldehyde (1 equiv) in a mixture of CHCl₃/AcOH at reflux afforded the corresponding N-Ar imidazole, which after acid-base extraction, was directly subjected to N-alkylation with alkyl chloride (0.75 equiv) in THF at 70 °C. Recrystallization afforded the desired unsymmetrical N-Ar imidazole (2a) without chromatographic purification (100 mmol scale, 15.5 g, 36% yield). The sequence was extended to salts 2a–2d, providing chromatography-free, scalable access to the unsymmetrical precursors. 4,5-Dimethyl substitution of the imidazole ring (vs. 4,5-unsubstituted) is preferred due to higher stability of NHC salts and NHC–metal complexes. The method is advantageous over other protocols in that it permits for a modular, chromatography-free synthesis of unsymmetrical NHC salts by direct N-alkylation of N-Ar imidazoles and related precursors.

With access to imidazolium precursors 2a–2e, the synthesis of Cu(I)–imidazol-2-ylidenes was performed using CuCl (1.0 equiv) and t-BuONa (1.0 equiv) in CH₂Cl₂ at room temperature to afford the corresponding complexes in 66-75% yields (Scheme 2). Cu(I)–imidazol-2-ylidene complexes 3a–3e were found to be air- and moisture-stable.

Scheme 2.

Synthesis of Unsymmetrical NHC-Cu Complexes. Conditions: CuCl (1.0 equiv), NaOtBu (1.0 equiv), CH₂Cl₂, 23 °C, 16 h.

The complexes 3a–3d were fully characterized by X-ray crystallography (Figure 3). Complexes 3a–3d are monomeric. Interestingly, complex 3a crystallized in a N–CH₂Mes/Cu syn conformation (Figure 3A), complex 3b crystallized with two molecules in the unit cell; with one molecule in a N–CH₂Mes/Cu syn conformation and the other in a N–CH₂Mes/Cu anti conformation (Figure 3B), Complexes 3a–3d crystallized in a N–CH₂Ar/Cu anti conformation (Figure 3C–D). The complexes 3a–3d are linear (3a: C_(NHC)–Cu–Cl, 175.9°; C–Cu, 1.884 Å; 3b-syn: C_(NHC)–Cu–Cl, 172.6°; C–Cu, 1.882 Å; 3b-anti: C_(NHC)–Cu–Cl, 178.3°; C–Cu, 1.867 Å; 3c: C_(NHC)–Cu–Cl, 174.8°; C–Cu, 1.867 Å; 3d: C_(NHC)–Cu–Cl, 176.7°; C–Cu, 1.883 Å. These parameters can be compared with Cu(I)–imidazol-2-ylidenes, [Cu(IMes)Cl]: C_(NHC)–Cu–Cl, 180.0°; C–Cu, 1.956 Å; [Cu(IPr)Cl]: C_(NHC)–Cu–Cl, 176.6°; C–Cu, 1.881 Å.

Figure 3.

X-ray crystal structures of Cu(I) complexes 3a–3d. 50% ellipsoids. Hydrogen atoms have been omitted for clarity. CCDC 2270937 (3a); CCDC 2270938 (3b); CCDC 2270939 (3c); CCDC 1994536 (3d). Selected bond lengths [Å] and angles [°]: 3a: Cu–C1, 1.884(2); Cu–Cl, 2.1159(5); C1–N1, 1.354(2); C1–N2, 1.391(2); C16–N2, 1.444(2); N2–C3, 1.392(2); N1–C2, 1.354(2); N1–C6, 1.480(2); C1–Cu–Cl, 175.94(6); N1–C1–N2, 103.7(2); C16–N2–C1, 124.3(2); C6–N1–C1, 126.3(2); C7–C6–N1, 115.8(2). 3b-syn: Cu–C1, 1.882(4); Cu–Cl, 2.119(1); C1–N1, 1.368(6); C1–N2, 1.377(5); C16–N2, 1.450(6); N2–C3, 1.395(5); N1–C2, 1.377(5); N1–C6, 1.482(6); C1–Cu–Cl, 172.6(1); N1–C1–N2, 103.2(3); C16–N2–C1, 123.9(3); C6–N1–C1, 124.0(3); C7–C6–N1, 113.9(4). 3b-anti: Cu–C25, 1.867(4); Cu–Cl, 2.106(1); C25–N4, 1.348(5); C25–N3, 1.368(5); C30–N3, 1.436(5); N3–C26, 1.407(5); N4–C27, 1.409(5); N4–C39, 1.480(5); C25–Cu–Cl, 178.3(1); N3–C1–N4, 104.2(3); C30–N3–C25, 124.7(3); C39–N4–C25, 123.1(3); C40–C39–N4, 115.2(3). 3c: Cu–C1, 1.867(4); Cu–Cl, 2.097(2); C1–N1, 1.356(4); C1–N2, 1.361(4); C13–N2, 1.444(4); N2–C4, 1.398(5); N1–C2, 1.392(5); N1–C6, 1.462(4); C1–Cu–Cl, 174.8(1); N2–C1–N1, 103.4(3); C13–N2–C1, 122.9(3); C6–N1–C1, 122.8(3); C7–C6–N1, 113.2(3). 3d: Cu–C1, 1.886(3); Cu–Cl, 2.1085(9); C1–N1, 1.351(4); C1–N2, 1.359(3); C14–N2, 1.444(3); N2–C3, 1.382(4); N1–C2, 1.386(4); N1–C4, 1.473(3); C1–Cu–Cl, 176.46(9); N2–C1–N1, 111.4(2); C14–N2–C1, 123.6(2); C4–N1–C1, 123.9(2); C5–C4–N1, 112.2(2).

Next, we analyzed the geometry of complexes 3a–3d using the method by Cavallo to determine catalytic pockets (Figure 4). The % buried volume (%V_bur) of [Cu(NHC)Cl] complexes 3a–3d is 52.7% (3a), 49.8% (3b-syn), 37.3% (3b-anti), 40.2% (3c) and 38.9% (3d). These values can be compared with [Cu(IMes)Cl] of 37.9% and [Cu(IPr)Cl] of 47.6%.

Figure 4.

Topographical steric maps of unsymmetrical Cu(I)–imidazol-2-ylidenes 3a–3d (A–E), IMes (F) and IPr (G) showing % V_bur per quadrant.

The x-ray structures provide insight into the conformational geometry of unsymmetrical complexes 3a–3d. First, the results demonstrate that the ligands gradually cover the %V_bur geometries ranging from 37.3% to 52.7%. Second, the value of 52.7% determined for 3a represents the largest %V_bur described to date for an IPr analogue with Dipp N-wingtip. Finally, the results demonstrate ligand flexibility around the N-CH₂-Ar substituent depending on the wingtip and backbone substitution.

Additional insight is obtained from examination of the quadrant distribution in topographical steric maps of 3a–3d. There is a clear distinction of the steric impact between SW/NW vs. SE/NE quadrants in complexes 3a (SW: 45.1%, NW: 42.9%, NE: 60.4%, SE: 62.5%) and 3b-syn (SW: 39.8%, NW: 35.4%, NE: 57.5%, SE: 66.4%), which feature a syn N–CH₂Ar/Cu geometry. In contrast, complexes 3b-anti, 3c and 3d, which feature an anti N–CH₂Ar/Cu geometry, are characterized by a more symmetrical quadrant distribution 3b-anti (SW: 28.2%, NW: 44.0%, NE: 44.2%, SE: 32.9%), 3c (SW: 45.2%, NW: 53.5%, NE: 35.4%, SE: 26.5%) and 3d (SW: 51.8%, NW: 38.9%, NE: 36.9%, SE: 27.9%). The topographical maps in these latter cases are more characteristic of the classical imidazol-2-ylidenes, IMes (SW: 37.5%, NW: 38.3%, NE: 37.5%, SE: 38.3%) and IPr (SW: 55.5%, NW: 39.6%, NE: 39.6%, SE: 55.5%). Note that crystal packing affects geometry distribution in some cases (vide infra).

With access to unsymmetrical Cu(I)–NHC complexes 3a–3e, we next explored their reactivity in Cu(I)-catalyzed hydroboration. Cu(I)-catalyzed borylcupration has recently emerged as a powerful strategy for the synthesis of boron-containing compounds.¹³ The capacity of Cu(I)–NHCs to simultaneously install a boryl–copper species across a variety of π-systems has enabled the discovery of novel reactions. In this context, the design of new Cu(I)–NHC systems is a major direction in the development of Cu(I)-catalyzed borylcupration reactions, however, these studies have been focused on symmetrical imidazol-2-ylidenes. It has been postulated that this reaction framework necessitates NHC ligands that provide strong metal-carbene bond, while maintaining limited steric hindrance around the metal center. Our crystallographic studies (Figures 2–3) demonstrated a gradual modification of ligand topology with a of range of geometries traversing IMes and IPr. Accordingly, Cu(I)-catalyzed hydroboration, an archetypal borylcupration process, has emerged as an ideal testing platform for the reactivity of unsymmetrical Cu(I)–NHCs.

Cu(I)–NHC-catalyzed β-hydroboration of alkynes in air has been selected as a model system.^13d This reaction is the most effective catalyst system for β-hydroboration reported to date, allowing for reactivity at 0.04 mol% [Cu] loading in some cases. We evaluated the catalysts 3a–3e under standard conditions using 1-phenyl-1-propyne (4) (1 equiv) in the presence of B₂pin₂ (1.2 equiv) and NaOH (5 mol%) in THF/MeOH at 0.02 mol% [Cu–NHC] loading (Table 1). As shown, complexes 3a–3c featuring unsymmetrical wingtip substitution of the imidazole ring in the presence of 4,5-dimethyl backbone substitution resulted in excellent reactivity (entries 1-3). Interestingly, complexes 3d–3e were also reactive at 0.02 mol% loading, however, they were less effective than 3a–3c. Importantly, 3a–3c could promote the reaction with much higher efficiency than imidazol-2-ylidenes IMes and IPr (entries 6-7). Further differentiation between the reactivity of 3a–3c was established at 0.01 mol% loading, which demonstrated N-Dipp/N-CH₂Mes^Me (3a) as the most reactive catalyst in the series. It should be noted that NaOH was used instead of NaOtBu because the reaction was more efficient under these conditions (3a, NaOH: 98%, NaOtBu: 67%).

Table 1.

β-Hydroboration of Alkynes Catalyzed by Unsymmetrical Cu(I)–NHC^a

^aConditions: 1-Phenyl-1-propyne (1.0 equiv), B₂pin₂ (1.2 equiv), NaOH (5 mol %), Cu(I)–NHC (0.02-0.01 mol%), THF:MeOH (10/1 v/vol) (0.36 M), 23 °C, 16 h. See SI for details.

The Cu(I)–NHC-catalyzed β-hydroboration with the unsymmetrical [Cu(N-Dipp/N-CH₂Mes^Me)Cl] (3a) was then tested with a number of alkynes (Scheme 3). To demonstrate the efficiency of the process, 0.02 mol% [Cu(I)–NHC] loading was used. The loading was increased to a minimum loading that permitted full conversions with more challenging alkynes. As shown in Scheme 3, we found that unsymmetrical (3a) is a general and highly reactive catalyst for β-hydroboration of alkynes. As such, the reactions of differently substituted phenylacetylenes (Me, Et, Bu, Ph) afforded the borylation products in excellent yields with full β:α selectivity (5a–5d). Substitution of the aromatic ring was well-tolerated and included a number of highly sensitive functional groups, such as bromo (5e–5f), nitro (5g), cyano (5h), and dichloro (5i). Furthermore, heterocyclic alkynes, including coordinating thiophenes (5j) and pyridines (5k), could be readily employed, allowing for incorporation of the boron moiety.

Scheme 3.

β-Hydroboration of Alkynes Catalyzed by Unsymmetrical Cu(I)–NHCs. Conditions: alkyne (1.0 equiv), B₂pin₂ (1.2 equiv), NaOH (5 mol %), Cu(I)–NHC (3a) (0.02–1 mol%), THF/MeOH (10/1 v/vol) (0.36 M), 23 °C, 16 h.

Moreover, substrates bearing coordinating methoxy (5l) and electrophilic ester (5m) were viable, while maintaining full β:α selectivity. Finally, aliphatic conjugated alkynes also performed well, including amides (5n) and esters (5o), demonstrating the functional group tolerance and high activity of this system. Inspired by the reactivity of [Cu(N-Dipp/N-CH₂Mes^Me)Cl] (3a), we also tested its performance in β-carboboration (Scheme 4). Pleasingly, this catalyst promoted β-carboboration at 2 mol% loading, further attesting to its high reactivity.

Scheme 4.

β-Carboboration of Alkynes. Conditions: alkyne (1.0 equiv), B₂pin₂ (1.2 equiv), NaOtBu (1.2 equiv), Cu(I)–NHC (3a) (2 mol%), MeI (8 equiv), THF (0.4 M), 60 °C, 16 h.

In consideration of the excellent reactivity of [Cu(N-Dipp/N-CH₂Mes^Me)Cl] (3a), we tested its performance in β-hydroboration of complex substrates (Scheme 5). Although Cu-catalyzed β-hydroboration is a common process in organic synthesis, the vast majority of reactions involve simple substrates. Therefore, it was of significant interest to determine if [Cu(N-Dipp/N-CH₂Mes^Me)Cl] (3a) is capable of tolerating sensitive functional groups in complex pharmaceuticals. In the event, we found that this catalyst shows remarkable functional group tolerance, enabling hydroboration of complex alkynes derived from probenecid (antigout), ciprofibrate (hypolipidemic), camphorosultam (chiral auxiliary), quinine (antimalarial), febuxostat (antihyperuricemic), indomethacin (anti-inflammatory), oxaprozin (anti-inflammatory), zaltoprofen (COX-2 inhibitor), cholesterol (steroid) and erlotinib (anticancer) (7a–7j). These reactions demonstrate excellent functional group tolerance towards sulfonamides, esters, aliphatic halides, activated esters, coordinating heterocycles, such as quinolines and piperidines, benzylic electrophiles, nitriles, thiazoles, indoles, activated amides, oxazoles, ketones, diaryl sulfides, quinazolines and anilines.

Scheme 5.

Late-Stage Hydroboration of Complex Pharmaceuticals Catalyzed by Unsymmetrical Cu(I)–NHCs. Conditions: alkyne (1.0 equiv), B₂pin₂ (1.2 equiv), Cu(I)–NHC (3a) (1 mol%), NaOH (5 mol%), THF/MeOH (10/1 v/vol) (0.36 M), 23 °C, 16 h. Yields and β:α selectivity determined by ¹HNMR of crude reaction mixtures. Isolated yields and β:α ratio are shown in parenthesis.

In a broader sense, this catalytic approach should offer an attractive entryway to the modification of complex pharmaceuticals to enable installation of boron moiety. The presented examples represent some of the most complex applications of β-hydroboration reported to date.

Encouraged by the broad substrate scope in the β-hydroboration of alkynes, we investigated the reactivity of [Cu(N-Dipp/N-CH₂Mes^Me)Cl] (3a) in hydroboration of alkenes (Scheme 6). As shown, this catalyst was highly effective in promoting hydroboration of a range of electronically and sterically-differentiated aryl-substituted alkenes, including electronically-neutral (9a), electron-rich (9b–9c), halogen-substituted (9d–9f) as well as sterically hindered (9g–9h), conjugated (9i–9j), disubstituted (9k) and cyclic (9l). These reactions proceeded with full β-regioselectivity to afford valuable alkyl boronates.

Scheme 6.

Hydroboration of Alkenes. Conditions: alkene (1.0 equiv), B₂pin₂ (1.2 equiv), NaOtBu (10 mol%), Cu(I)–NHC (3a) (1.0 mol%), MeOH (2 equiv), THF (0.25 M), 25 °C, 16 h. ^aCu(I)–NHC (7.5 mol%).

Furthermore, although the goal of this study was to develop achiral ligands, we have established that this class of unsymmetrical Cu–NHCs is capable of promoting β-borylcupration in an asymmetric fashion (Scheme 7). The results in borylcupration of alkenes bode well for the broad application of this class of unsymmetrical NHCs in catalysis.

Scheme 7.

Asymmetric β-Borylcupration of Alkenes. Conditions: alkene (1.0 equiv), B₂pin₂ (1.2 equiv), NaOtBu (10 mol%), Cu(I)–NHC (3f) (10 mol%), MeOH (2 equiv), THF (0.4 M), −78 °C to RT, 24 h, then H₂O₂ (5 equiv), NaOH (5 equiv), THF (0.2 M), rt, 30 min.

In order to eliminate impact from steric packing in the x-ray crystallographic analysis of complexes 3a–3e, the percent buried volume (%V_bur) was calculated from the optimized structures at the B3LYP 6-311++g(d,p) level (Figure 5). X-ray structures of 3a–3d were used as a starting geometry. A noteworthy feature is that complexes 3a–3b optimized to two geometries, N–CH₂Mes/Cu syn conformation and N–CH₂Mes/Cu anti conformation (3a-syn/3a-anti: Figure 5A–B; 3b-syn/3b-anti: Figure 5C–D), while complexes 3c–3e optimized to a single N–CH₂Ar/Cu anti geometry (Figure 5E–G). Steric maps of the standard imidazol-2-ylidenes, [Cu(IMes)Cu] and [Cu(IPr)Cl] are shown for comparison (Figure H–I).

Figure 5.

Topographical steric maps of unsymmetrical Cu(I)–imidazol-2-ylidenes 3a–3d (A–F), IMes (G) and IPr (H) showing % V_bur per quadrant at B3LYP 6-311++g(d,p) level. Note that IPr is symmetrical.

The data demonstrate a gradual change of the geometry of the catalytic pocket in the series of 3a–3e. The %V_bur of [Cu(NHC)Cl] (3a-syn) and (3b-syn) is 49.5% (SW: 47.8%, NW: 46.5%, NE: 55.3%, SE: 48.3%) and 45.0% (SW: 38.2%, NW: 39.1%, NE: 47.4%, SE: 55.2%). This can be compared with the corresponding anti isomers, (3a-anti) and (3b-anti), of 39.9% (SW: 46.5%, NW: 44.7%, NE: 40.0%, SE: 28.2%) and 35.7% (SW: 37.9%, NW: 36.7%, NE: 40.0%, SE: 28.1%). The %V_bur of [Cu(NHC)Cl] further decreases in the order of (3c) > (3d) > (3e) of 38.6% (SW: 46.1%, NW: 45.0%, NE: 36.1%, SE: 27.3%), 36.7% (SW: 43.2%, NW: 42.4%, NE: 33.2%, SE: 28.0%), 33.6% (SW: 36.5%, NW: 36.4%, NE: 33.3%, SE: 28.0%), respectively. These values can be compared with the %V_bur of 36.4% (SW: 36.4%, NW: 36.4%, NE: 36.4%, SE: 36.4%) and 42.6% (SW: 42.6%, NW: 42.6%, NE: 42.6%, SE: 42.6%) for symmetrical [Cu(IMes)Cl] and [Cu(IPr)Cl].

Overall, the optimized geometry of the catalytic pockets in 3a–3e clearly reveals flexibility of the N-CH₂Ar wingtip. The N-CH₂Ar substituent is rotated away from the metal center in an anti conformation (3a: C_(car)–N–C_(CH2)–C_(Ar) dihedral angle of 131.3°) or rotated towards the metal center in a syn eclipsed conformation (3a: C_(car)–N–C_(CH2)–C dihedral angle of 40.6°). The less sterically-demanding anti conformation of 3a (%V_bur of 39.9%) is between that of IMes (36.4%) and IPr (42.6%), while the syn conformation of 3a is more sterically-demanding than IPr (49.5%). The %V_bur of complexes 3b–3e cover additional geometries spanning a remarkable range from 33.6% to 45.0% (Chart 1). The unique feature of unsymmetrical imidazol-2-ylidene ligands 3a–3e is flexibility of the N-wingtip that can accommodate to the steric environment.

Chart 1.

Graphical representation of the steric space accessible by unsymmetrical Cu(I)–NHC imidazol-2-ylidenes.

Rotational studies were performed to interrogate the flexibility of N-aromatic wingtip in unsymmetrical imidazol-2-ylidenes 3a–3e. Detailed rotational profiles of the parent carbenes 3a–3e were obtained by a systematic rotation along the C_(Me)–N–CH₂–C_(Ar) dihedral angle. The rotation was performed in both directions. Figure 6 shows rotational profiles of unsymmetrical imidazol-2-ylidenes 3a–3e.

Figure 6.

Rotational profiles of unsymmetrical imidazol-2-ylidenes (ΔE, kcal/mol, vs. C_(Me)–N–CH₂–C_(Ar) [°]).

Rotational profiles of 3a (Dipp/CH₂Mes^Me) and 3b (Mes/CH₂Mes^Me) identified two energy minima at ca. 60° C_(Me)–N–CH₂–C_(Ar) angle in an anti C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 127.5°) and at ca. 140° in a syn C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 46.5°). The energy maxima are located at ca.100° C_(Me)–N–CH₂–C_(Ar) dihedral angle (2.7 kcal/mol) in a staggered C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 68.0°) and at ca. 0° C_(Me)–N–CH₂–C_(Ar) dihedral angle (3.2 kcal/mol) in an anti C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 180.0°). In contrast, rotational profiles of 3d (Dipp/CH₂Mes) and 3e (Mes/CH₂Mes) reveal the energy minimum at ca. 40° C–N–CH₂–C_(Ar) angle in an anti C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 143.0°), while the energy maximum (2.4 kcal/mol) is located at ca. 180° C–N–CH₂–C_(Ar) dihedral angle in a syn C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 0.0°). The rotational profile of 3c (Dipp/CH₂Ph^Me) reveals the energy minimum at ca. 70° C_(Me)–N–CH₂–C_(Ar) angle in a syn C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 105.0°), while the energy maxima (6.5 kcal/mol) are located at ca. 0° C_(Me)–N–CH₂–C_(Ar) dihedral angle in a syn C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 180.0°) and ca. 180° C_(Me)–N–CH₂–C_(Ar) dihedral angle in an anti C_(carbene)–N–CH₂–C_(Ar) conformation (ca. 0.0°). The rotational profiles provide further evidence for the flexible rotation of N-wingtips in 3a–3e. The ligand geometry depends on the interplay of N-CH₂Ar substitution and backbone substitution. The optimal combination of C^(Me) and CH₂Mes permits for a full flexibility of the N-wingtip with low rotational barriers between syn and anti conformers. The avoidance of steric interactions in these systems is likely to provide a range of modular steric environments around the metal centers.

To gain insight into the electronic properties of the unsymmetrical class of imidazol-2-ylidenes, HOMO and LUMO energy levels were determined at the B3LYP 6-311++g(d,p) level (Figure 7 and SI). It is now well-established that HOMO and LUMO provide the most accurate estimation of nucleophilicity and electrophilicity of NHC ligands.

Figure 7.

(A) HOMO and LUMO energy levels (eV). (B) HOMO and LUMO+1 (eV) of 3a calculated at B3LYP 6-311++g(d,p). See SI.

Interestingly, the HOMO of 3a (syn conformation: −5.68 eV; anti: −5.75 eV) is much higher than of the classic IPr (−6.01 ev) and in the same range as N-alkyl NHCs (e.g. ItBu, −5.67 eV). This can be ascribed to the combination of N-alkyl and C^(Me) backbone substitution, resulting in a strongly σ-donating carbene ligand. The HOMO of 3b (syn: −5.62 eV; ant: −5.68 eV), 3c (−5.83 eV), 3d (−5.93 eV) and 3e (−5.86) indicate that all of these ligands are stronger s-donors than IPr. The π-accepting orbitals (LUMO+1 and LUMO+2 due to required symmetry) of 3a (syn: −0.35 eV; anti: −0.36 eV) can be compared with IPr (−0.48 eV) and with N-alkyl NHCs (e.g. ItBu, 0.36 eV), while the values for 3b–3e range from −0.38 eV to −0.04 eV (see SI).

Overall, these results indicate that the unsymmetrical imidazol-2-ylidenes are strongly σ-nucleophilic and sterically-flexible ligands with sterics spanning the range of IMes/IPr geometries,^[17] and electronics matching those of N-alkyl NHCs. This unique combination should enable further tuning of the ligand design on the interface of N-aromatic and N-aliphatic carbene ligands, while retaining flexibility of the ligand environment.

Conclusion

In conclusion, we have reported a unique class of unsymmetrical N-heterocyclic carbene ligands that are characterized by freely-rotatable N-aromatic wingtips in the imidazol-2-ylidene architecture. One of the long-standing challenges in NHC ligand engineering has been the lack of steric flexibility of catalytic pockets of the most common symmetrical imidazol-2-ylidenes. We have shown that the combination of a rotatable N–CH₂Ar bond with the conformationally-fixed N–Ar linkage results in a highly modular ligand topology. Most importantly, these ligands now enter the range of geometries inaccessible to IMes and IPr. We have provided a practical access from readily accessible materials in a modular synthetic pathway. These ligands are highly reactive in Cu(I)-catalyzed β-hydroboration of alkynes, the most representative of borylcupration methods with profound importance in organic synthesis. We have further demonstrated the synthetic utility in the late-stage hydroboration of complex pharmaceuticals. Full structural characterization of steric and electronic properties has been presented. The broad access to this ligand class has been established in collaboration with MilliporeSigma to enable applications by the synthetic chemistry community. We anticipate that the facile access to new modular ligand geometries will enable significant opportunities and broader interrogation of N-heterocyclic carbenes in organic synthesis and catalysis.