Formation and degradation of strongly reducing cyanoarene-based radical anions towards efficient radical anion-mediated photoredox catalysis

Abstract

Cyanoarene-based photocatalysts (PCs) have attracted significant interest owing to their superior catalytic performance for radical anion mediated photoredox catalysis. However, the factors affecting the formation and degradation of cyanoarene-based PC radical anion (PC^•‒) are still insufficiently understood. Herein, we therefore investigate the formation and degradation of cyanoarene-based PC^•‒ under widely-used photoredox-mediated reaction conditions. By screening various cyanoarene-based PCs, we elucidate strategies to efficiently generate PC^•‒ with adequate excited-state reduction potentials (E_red^*) via supra-efficient generation of long-lived triplet excited states (T₁). To thoroughly investigate the behavior of PC^•‒ in actual photoredox-mediated reactions, a reductive dehalogenation is carried out as a model reaction and identified the dominant photodegradation pathways of the PC^•‒. Dehalogenation and photodegradation of PC^•‒ are coexistent depending on the rate of electron transfer (ET) to the substrate and the photodegradation strongly depends on the electronic and steric properties of the PCs. Based on the understanding of both the formation and photodegradation of PC^•‒, we demonstrate that the efficient generation of highly reducing PC^•‒ allows for the highly efficient photoredox catalyzed dehalogenation of aryl/alkyl halides at a PC loading as low as 0.001 mol% with a high oxygen tolerance. The present work provides new insights into the reactions of cyanoarene-based PC^•‒ in photoredox-mediated reactions.

Fig. 1. Schematic illustration of the current work.

Reaction scheme of the formation and photodegradation of cyanoarene-based photocatalyst radical anion (PC^•‒). Here, ISC, T₁, PET, and Sub denote intersystem crossing, triplet excited state, photoinduced electron transfer, and substituent, respectively.

Fig. 2. Characterization of 4DP-IPN and 4DP-IPN^•−.

a Reaction scheme of the formation of 4DP-IPN^•−. The calculated frontier molecular orbitals (MO) topologies of 4DP-IPN and 4DP-IPN^•− are shown. Their excited state redox potentials (E_red^*(PC) and E_ox^*(PC^•−)) were estimated from E_red* = E_0-0 + E_red⁰ and E_ox^*(PC^•‒) = −E_0-0(PC^•‒) + E_red⁰(PC); E_0-0(PC) and E_0-0(PC^•−) were evaluated by the onset of gated photoluminescence (PL) emission spectrum in CH₃CN at 65 K and the onset of UV-Vis absorption spectrum at room temperature (RT), respectively. b Jablonski diagram of 4DP-IPN. The rate constants of all photophysical processes were evaluated in the current work; here, IC, (R)ISC, D₀, and S₁/T₁/D_n denote internal conversion, (reverse)intersystem crossing, doublet ground state and singlet/triplet/doublet excited state, respectively. c UV-Vis absorption spectra of 4DP-IPN (black line) and 4DP-IPN^•− (orange line) in CH₃CN; here, a.u. denotes arbitrary units. UV-Vis absorption spectra of 4DP-IPN^•− were taken right after illumination by two 3 W 515 nm LEDs for 1 min at RT. TD-DFT results (oscillator strengths) are shown as stick spectra. d Time-dependent changes of the UV-Vis absorbance of 4DP-IPN^•‒ at 525 nm and Ir(dtbby)^•‒(ppy)₂PF₆ at 533 nm (compare Supplementary Fig. 9). PC^•‒ was generated under the illumination of two 3 W 515 nm LEDs for 3 min (for 4DP-IPN) or two 3 W 455 nm LEDs for 1 min (for Ir(dtbby)(ppy)₂PF₆) at RT. Changes in the UV-Vis absorption spectrum of freshly generated 4DP-IPN^•− were recorded every 2 min under dark conditions (inset). e Stern–Volmer plots for the PL decays quenching of 4DP-IPN and Ru(bpy)₃Cl₂ in CH₃CN by DIPEA at RT. f Results of kinetics simulation of the relative excited state population of selected PCs (5.0 × 10⁻³ M) over time under continuous 455 nm (or 390 nm for 10-phenylphenothiazine (PTH)) irradiation (see the SI for the full details of the kinetics simulation).

Fig. 3. Formation of PC^•− of various cyanoarene-based PCs.

a Chemical structures of selected cyanoarene-based PCs and their calculated HOMO and LUMO energies. UV-Vis absorption spectra of selected PC (black line) and PC^•− (orange line). It should be noted that the PCs prepared here contain six completely new compounds (3DP-DMDP-IPN, 3DP-Cz-IPN, 3DP-DCDP-IPN, 4-p-MCDP-IPN, 4-o,p-DCDP-IPN, and 4-p,p-DCDP-IPN). All ground state reduction potentials of PCs (E_red⁰(PC)) were measured in the current work and their excited state reduction potentials (E_red^*(PC)) were estimated from E_red* = E_0-0 + E_red⁰; E_0-0(S₁) and E_0-0(T₁) were evaluated by the onset of PL emission and gated PL emission, respectively, in CH₃CN at 65 K (except for 4tCz-IPN in DMF). UV-Vis absorption spectra were taken from the degassed solutions of PCs (1.0 × 10⁻⁴ M) and DIPEA (0.5 M) in CH₃CN right after illumination of two 3 W 455 nm LEDs for 1 min at RT, b 3DP-DMDP-IPN, c 4tCz-IPN, d 4Cz-IPN, e 3DP-Cz-IPN, f 3DP-F-IPN, g 3DP-DCDP-IPN, h 4-p-MCDP-IPN, i 4-o,p-DCDP-IPN, and j 4-p,p-DCDP-IPN. All solutions were prepared in a glove box and fully degassed. TD-DFT calculation results (oscillator strengths) are shown as stick spectra.

Fig. 4. Photodegradation behavior of 4DP-IPN.

a Reactions were performed with 4DP-IPN (1.0 × 10⁻⁴ M) and DIPEA (0.5 M) in CH₃CN under the illumination of two 3 W 515 nm LEDs or two 3 W 455 nm LEDs at RT. PC degradations were monitored in situ by TLC with eluent conditions (CH₂Cl₂:hexanes, 7:3 v/v). The photodegraded products were isolated by column chromatography, and ¹H NMR spectra confirmed that a methyl (and hydrogen) substitution reaction occurred at the CN position of 4DP-IPN to yield 4DP-Me-BN (and 4DP-H-BN). b Proposed mechanistic pathway for the photodegradation behavior of 4DP-IPN in the presence of DIPEA and DFT calculations for the bond dissociation energies (ΔH) in DIPEA^•+; the values in parenthesis correspond to the calculated bond dissociation energies in DIPEA.

Fig. 5. Photodegradation behaviors of various cyanoarene-based PCs.

a Photodegradation behavior of 3DP-Cz-IPN and 3DP-DCDP-IPN. b Photodegradation behaviors of 4Cz-IPN and 4tCz-IPN and the proposed mechanism of photodegradation in the presence of DIPEA and DIPMA as a reducing agent. c Photodegradation behavior of PCs non-generating PC^•‒. d Photodegradation behaviors of PCs with labile groups. For the characterizations of the isolated products, see Supplementary Figs. 20–26 in the SI.

Fig. 6. Photodegradation of 4DP-IPN in actual dehalogenation reactions.

a Fate of catalyst during dehalogenation reactions. Reactions were performed with substrates (0.1 M), DIPEA (10.0 equiv.), and 4DP-IPN (1 mol%, 1.0 × 10⁻³ M) in CH₃CN (1 mL) under the illumination of two 3 W 455 nm LEDs for several hours at RT. Yields were determined by ¹H NMR using 1,3,5-trimethoxybenzene as an internal standard. PC degradation was monitored by in situ TLC with eluent conditions (EA:hexanes, 1:4 v/v). All redox potential values were obtained from the literature where the potential values were measured against the standard calomel electrode (SCE)^,,. b Catalytic performance of 4DP-Me-BN. Reactions were performed under the same conditions as in a. Early reaction kinetics of the dehalogenation reactions with 4DP-IPN (orange line) and 4DP-Me-BN (green line) were monitored on a 2 mL scale; the yields were determined by ¹H NMR using 1,3,5-trimethoxybenzene as an internal standard. c Proposed mechanistic pathways for the photodegradation of 4DP-IPN in the presence of aryl halides.

Fig. 7. Characterization of 4DP-Me-BN.

a Calculated energies and topologies of the frontier MO of 4DP-Me-BN. b PL decay of 4DP-IPN (1.0 × 10⁻⁵ M; orange line) and 4DP-Me-BN (1.0 × 10⁻⁵ M; green line) in CH₃CN at RT. c Steady-state PL emission spectra of 4DP-IPN (1.0 × 10⁻⁵ M; orange line) and 4DP-Me-BN (1.0 × 10⁻⁵ M; green line) in CH₃CN at RT. d UV-Vis absorption spectra of 4DP-IPN (1.0 × 10⁻⁵ M; orange line) and 4DP-Me-BN (1.0 × 10⁻⁵ M; green line) in CH₃CN at RT. TD-DFT results (oscillator strengths) are shown as stick spectra. e Stern–Volmer plots for the PL quenching of 4DP-Me-BN (1.0 × 10⁻⁵ M) in CH₃CN by DIPEA at RT. Stern–Volmer plots were obtained from PL decays of 4DP-Me-BN, excitation at λ_ex = 377 nm, and detection at λ_det = 470 nm.

Fig. 8. Oxygen tolerance in the dehalogenation reactions.

a Proposed mechanism of the photocatalyzed reductive dehalogenation of aryl halides mediated by ²PC^•‒. Here, (B)ET and XAT denote (back)electron transfer and halogen atom transfer, respectively. b Effect of oxygen for the reductive dehalogenation of 4-bromobenzonitrile with 4DP-IPN as a PC in the presence of Ar (green-filled squares), air (blue-filled squares), and O₂ (orange-filled squares). Reactions were performed with 4-bromobenzonitrile (0.1 M), DIPEA (10.0 equiv.), and 4DP-IPN (5–0.001 mol%) in CH₃CN under the illumination of two 3 W 455 nm LEDs for 8 h. c Experimental (filled squares) and simulated (empty squares) reaction kinetics of photoredox-mediated reductive dehalogenation of 4-bromobenzonitrile with 4DP-IPN (0.03 mol%) under Ar and air atmospheres in a closed glass vial; simulated (empty circles) reaction kinetics indicate that the BET process (k_BET = 1 × 10¹⁰ M⁻¹ s⁻¹) and a slower ET process (k_PET = 4.9 × 10⁶ M⁻¹ s⁻¹) are involved. Kinetics simulations were performed based on the proposed mechanism described in Supplementary Fig. 15. The rate constants for all processes were evaluated from experiments or calculations (see Supplementary Fig. 15 and Supplementary Table 7 for the full details of the kinetics simulation).

Fig. 9. Results of reductive dehalogenation of various aryl/alkyl halides in the presence of 4DP-IPN as a PC.

Reactions were performed with substrates (0.1 M), DIPEA (10.0 equiv.), and 4DP-IPN (0.005 mol%) in CH₃CN (1 mL) under the illumination of two 3 W 455 nm LEDs for 6 to 48 h at RT. The gram-scaled reaction was performed with 4-bromobenzonitrile (3.0 g, 16.48 mmol), DIPEA (10.0 equiv.), and 4DP-IPN (0.01 mol%) in CH₃CN (30 mL) with irradiation by four 3 W 455 nm LEDs under ambient conditions without any degassing. Reactions were performed with ^a0.01 mol%, ^b0.05 mol%, ^c0.5 mol%, and ^d5 mol% 4DP-IPN, and ^eillumination by four 3 W 455 nm LEDs. Injection of a total of ^f1.5 mol% and ^g3 mol% 4DP-IPN divided over three additions every 16 h during the course of the reaction. Yields were determined by GC-FID or ^†1H NMR using 1,3,5-trimethoxybenzene as an internal standard. All redox potential values were obtained from the literature, where the potential values were measured against the SCE^–.