Noncovalent interaction with a spirobipyridine ligand enables efficient iridium-catalyzed C-H activation

Noncovalent interactions are prevalent in Nature and have been used for molecular recognition, construction of complex architectures, and control of reactivity and selectivity across crystal engineering, supramolecular chemistry, organic synthesis, and catalysis. Among these, CH-π interactions are weak but potentially general for molecular recognition because C–H bonds and π electrons are ubiquitous in organic molecules. While CH-π interactions have been invoked to thermodynamically stabilize molecular systems and, in a few cases, proposed computationally to stabilize transition states, clear experimental evidence in catalysis remains scarce. Prior studies include Noyori’s enantioselective transfer hydrogenation of aryl ketones, where computation suggested a CH-π interaction between a ligand C–H and the aryl substrate governing stereocontrol; further computational reports rationalized enantio- and regioselectivity via CH-π interactions, including in transition-metal-catalyzed C–H functionalization. Building on efforts to use ligands for molecular recognition in C–H functionalization, the authors hypothesized that a ligand could directly recognize arene π electrons via CH-π interaction to stabilize the key transition state for undirected aromatic C–H cleavage. They report that a spirobipyridine (SpiroBpy) ligand enables efficient iridium-catalyzed undirected C–H borylation of arenes, particularly electron-rich substrates, by such an interaction, with computational studies and a ligand kinetic isotope effect (KIE) supporting the mechanistic proposal. Ir-catalyzed arene C–H borylation is a powerful route to aryl boronates, but borylation of electron-rich arenes under mild conditions with HBpin remains challenging using established bipyridine ligands. Installing a perpendicular fluorene motif onto a bipyridine core (SpiroBpy) was envisioned to promote attractive noncovalent CH-π interactions with arene substrates, accelerating C–H activation.

The paper situates its work within extensive literature on noncovalent interactions in catalysis, highlighting CH-π interactions as a subtle but potentially general design element. Notable prior work includes: (1) Noyori’s η6-ruthenium-catalyzed enantioselective transfer hydrogenation where computational analysis attributed enantioselectivity to a CH-π interaction between a ligand C–H and an aryl ketone substituent; (2) multiple computational studies proposing CH-π-stabilized transition states to explain observed enantio- and regioselectivity in various catalytic systems, including C–H functionalization; and (3) Musaev’s computationally proposed interaction where a ligand C–H engages the incoming arene’s π system during Pd-catalyzed directed C–H activation. In C–H borylation research, bidentate N-ligands such as dtbpy and tmphen have been widely used and optimized, but they often require higher temperatures or B2pin2 and exhibit reduced efficiency with HBpin for electron-rich arenes. Prior advances include alternative ligand families (e.g., phenanthrolines, bidentate boryl ligands, pincer complexes) and strategies invoking other noncovalent interactions (hydrogen bonding, ion-pairing) to influence site-selectivity. The present work uniquely targets undirected arene C–H activation by designing a ligand (SpiroBpy) to create a direct CH-π interaction between a ligand backbone C–H and the arene substrate, aiming to stabilize the C–H activation transition state and improve rates with HBpin.

Catalyst system and optimization: Reactions employed [Ir(OMe)(cod)]2 (2 mol%) with ligands (4 mol%) in THF (0.1 M) at 50 °C for 16 h using HBpin as the borylating agent. Benchmark substrate 1,3-dimethoxybenzene (1a, 0.10 mmol) was used to compare ligands including dtbpy, tmphen, a methylene-bridged bipyridine (L1), a diphenyl-bridged variant (L2), and a family of R-substituted SpiroBpy ligands. Reaction progress was monitored (GC or 1H NMR with internal standards) and optimized against temperature, solvent, and HBpin loading. Representative alternative conditions included 30 °C, cyclohexane solvent, and near-stoichiometric HBpin (120 mol%). Kinetic comparisons between SpiroBpy and tmphen were conducted for electron-rich substrates (1a, 1m) under optimized conditions to assess initial rate differences. Substrate scope: Electron-rich arenes (multi-alkoxy, amino, alkyl, silyl) were evaluated to minimize regioisomeric complications by selecting meta-/ortho-disubstituted and polysubstituted substrates. Standard conditions were typically 200 mol% HBpin, 2 mol% [Ir(OMe)(cod)]2, 4 mol% ligand, THF, 50 °C, 16 h. For more challenging anilines and certain substrates, reactions were run at 70 °C and/or with 300 mol% HBpin and adjusted concentration (0.2 M) per figure notes. Scale-up demonstrations: Gram-scale borylations were performed for pharmaceutically relevant molecules: lidocaine (3a), a phenylalanine derivative (3b), and an indole derivative (3c) relevant to tambromycin synthesis, using SpiroBpy under slightly modified conditions (e.g., 70 °C, increased HBpin) as indicated. Computational studies: DFT calculations compared C–H activation barriers for LIrBpin3 complexes bearing L1, SpiroBpy, and tmphen with 1,3-bis(dimethylamino)benzene (5) as model substrate. Energies were computed at M06/SDD:6-311+G(d,p)THF(SMD)//B3LYP-D3/SDD:6-31+G(d,p) (298.15 K). A distortion/interaction (activation strain) analysis was performed to decompose activation barriers. Noncovalent interaction characterizations employed IGMH and NCI plot analyses, alongside NBO analysis to quantify donor–acceptor interactions between the arene π orbitals and the σ* orbital of the ligand backbone C–H. Ligand kinetic isotope effect (KIE): An octadeuterated ligand (SpiroBpy-d8) was synthesized to probe CH-π involvement. Parallel reactions of 1,3-bis(dimethylamino)benzene (5) with HBpin using SpiroBpy vs SpiroBpy-d8 were run at three temperatures (two runs each) to measure ligand KIE. A differential Eyring analysis provided enthalpic and entropic components to rationalize temperature dependence of the inverse KIE. General procedure: In an oven-dried J-Young Schlenk tube under N2, [Ir(OMe)(cod)]2 (2 mol%) and SpiroBpy (4 mol%) were combined with arene substrate (0.10 mmol), dry THF (1.0 mL), and HBpin (200 mol%), then stirred at 50 °C for 16 h (or 70 °C for certain anilines). Workup involved dilution with EtOAc, GC or 1H NMR yield determination using internal standards, solvent removal, and purification by silica gel chromatography or GPC. A representative example for substrate 1m at 70 °C furnished 2m in 78% NMR yield (71% isolated).

• SpiroBpy substantially accelerates Ir-catalyzed undirected arene C–H borylation with HBpin compared to state-of-the-art ligands (dtbpy, tmphen), particularly for electron-rich arenes.
• Optimization with 1,3-dimethoxybenzene (1a): dtbpy gave 20% 2a (78% 1a remaining), tmphen 50% (44% 1a), whereas SpiroBpy gave 82% (78% isolated; 13% 1a remaining). Modified SpiroBpy ligands gave lower yields (e.g., Bpin-SpiroBpy 48%; Ph-SpiroBpy 61%; tBu-SpiroBpy 80%). L1 and L2 were less effective (26% and 65%, respectively). Lower temperature (30 °C) decreased yield; near-stoichiometric HBpin slightly reduced yield; cyclohexane solvent was viable (79%).
• Kinetic monitoring showed faster initial rates with SpiroBpy than tmphen for electron-rich substrates (1a, 1m) under optimized conditions.
• Broad substrate scope: SpiroBpy delivered high yields for challenging electron-rich arenes. Examples (SpiroBpy vs dtbpy/tmphen, GC/NMR yields; isolated in parentheses): anisoles 2a >80% (e.g., 2a 82% (78%)); alkylbenzenes 2g–j 72–96% (e.g., 2j 96% (87%)); silyl-protected phenol 2e 80% (76%); meta-terphenyl 2f quantitative; aniline derivatives 2m–p high yields (e.g., 2m 78% (71%)) whereas dtbpy/tmphen often ≤19%. For more reactive halogenated/EDG substrates (2r–x), SpiroBpy maintained higher or comparable yields, generally surpassing dtbpy and tmphen.
• Gram-scale demonstrations: Lidocaine (3a) borylated to 4a in 62%; a phenylalanine derivative (3b) to 4b in 96%; indole derivative (3c) to 4c in 95% (key intermediate toward tambromycin).
• Computational support: Calculated barriers for C–H cleavage (TSBC) with model 1,3-(Me2N)2benzene: ΔG‡ = 32.8 kcal/mol (L1), 30.3 (SpiroBpy), 32.1 (tmphen). Activation strain analysis attributes lower barrier mainly to more favorable interaction energy with SpiroBpy (ΔEint ≈ −56.0 kcal/mol vs −53.2 for L1/tmphen). IGMH and NCI analyses reveal attractive interactions between the SpiroBpy backbone C–H and the arene π system in TS and Ir(V) intermediate. NBO shows donor (arene π) → acceptor (σ* of ligand C–H) interactions, stronger for more electron-rich substrates, aligning with experimental rate enhancements for diaminobenzenes.
• Mechanistic probe via ligand KIE: An inverse ligand KIE observed using SpiroBpy-d8 (faster reactions than with SpiroBpy-H) across temperatures supports involvement of a ligand C–H/π interaction in the turnover-limiting transition state. Differential Eyring analysis suggests the protiated ligand pathway is enthalpically favored, while the deuterated is entropically favored; stronger inverse KIE at higher temperature is consistent with prior reports and may relate to differences in CH vs CD bond length and polarizability affecting CH-π interaction strength.

The study tests the hypothesis that a ligand-designed CH-π interaction can directly recognize arene π electrons to stabilize the transition state of undirected C–H activation, thereby accelerating Ir-catalyzed borylation. Experimentally, SpiroBpy consistently delivers higher initial rates and higher overall yields for electron-rich arenes compared to dtbpy and tmphen under otherwise identical conditions, including challenging anilines and alkylbenzenes with HBpin. Computational analyses corroborate a reduced activation barrier for C–H cleavage with SpiroBpy, with activation strain analysis indicating enhanced stabilizing interactions at the transition state. Visualization (IGMH/NCI) and orbital analysis (NBO) identify an attractive interaction between the SpiroBpy backbone C–H and the arene π system that persists into the Ir(V) intermediate. The inverse ligand KIE provides experimental evidence that a ligand C–H bond participates in a noncovalent interaction in the turnover-limiting transition state. The rate enhancement is more pronounced for electron-rich arene substrates, consistent with stronger donor (π) interactions. Collectively, these findings support that appropriately oriented, three-dimensional ligand frameworks can harness weak CH-π interactions to modulate kinetics of undirected C–H activation. This approach is inherently general for arenes since it relies on π electron density rather than specific directing groups or heteroatoms, contrasting strategies based on hydrogen bonding, Lewis acid-base, or ion pairing that require particular functional groups.

This work introduces a spirobipyridine (SpiroBpy) ligand that exploits a ligand–substrate CH-π interaction to accelerate iridium-catalyzed, undirected C–H borylation of arenes using HBpin, enabling high yields with otherwise challenging electron-rich substrates under relatively mild conditions. Key contributions include: (i) experimental demonstration of enhanced reactivity and broad scope, including gram-scale applications; (ii) computational evidence that SpiroBpy lowers the C–H activation barrier via increased stabilizing interaction energy at the transition state, visualized and quantified by IGMH/NCI/NBO analyses; and (iii) a ligand kinetic isotope effect indicating direct involvement of a ligand C–H in the turnover-limiting transition state via CH-π interaction. The authors propose that this generalizable recognition of arene π electrons can guide future ligand designs for diverse C–H functionalizations. Future directions include expanding spirobipyridine derivatives, exploring other catalytic transformations leveraging CH-π acceleration, and developing ligand KIE as a broader mechanistic probe of catalyst–substrate noncovalent interactions.

While computations and KIE data support the CH-π interaction hypothesis, the authors note that this attractive interaction alone may not fully account for all transition-state stabilization, and detailed mechanistic understanding remains under investigation. The interpretation of the inverse ligand KIE and its temperature dependence is preliminary. Reaction conditions typically require 50–70 °C and 2–3 equivalents of HBpin, and although the method excels with electron-rich arenes, comprehensive evaluation across less electron-rich or highly deactivated arenes is not detailed here.