Catalytic electron drives host–guest recognition

Abstract

Electron injection is demonstrated to trigger electrocatalytic chain reactions capable of releasing a solvent molecule and forming a redox active guest molecule. One-electron reduction of a hydroxy anthrone derivative (AQH–CH₂CN) results in the formation of an anthraquinone radical anion (AQ˙⁻) and acetonitrile (CH₃CN). The resulting fragment of AQ˙⁻ exhibits high stability under mild reducing conditions, and it has enough reducing power to reduce the reactant of AQH–CH₂CN. Hence, subsequent electron transfer from AQ˙⁻ to AQH–CH₂CN yields the secondary AQ˙⁻ and CH₃CN, while the initial AQ˙⁻ is subsequently oxidized to AQ. Overall, the reactants of AQH–CH₂CN are completely converted into AQ and CH₃CN in sustainable electrocatalytic chain reactions. These electrocatalytic chain reactions are mild and sustainable, successfully achieving catalytic electron-triggered charge-transfer (CT) complex formation. Reactant AQH–CH₂CN is non-planar, making it unsuitable for CT interaction with an electron donor host compound (U_HAnt₂) bearing parallel anthracene tweezers. However, conversion of AQH–CH₂CN to planar electron acceptor AQ by the electrocatalytic chain reactions turns on CT interaction, generating a host CT complex with U_HAnt₂ (AQ ⊂ U_HAnt₂). Therefore, sustainable electrocatalytic chain reactions can control CT interactions using only a catalytic amount of electrons, ultimately affording a one-electron switch associated with catalytic electron-triggered turn-on molecular recognition.

---

Introduction

The addition of an electron can significantly change the chemical and electronic nature of molecules.^1,2 The extra electron (+e⁻) populated in the antibonding orbital weakens the chemical bond,³ leading to reductive fragmentations that form a redox-active fragment and a highly stable coproduct.^1,2 When the resulting active fragment has sufficient reducing power to reduce the reactant, electrocatalytic chain reactions can be triggered.^1,2 Synthetic studies on such electron upconversions have focused on challenging chemical transformations.^1,2,4–12 However, the advantages of electrons as a remote and accessible chemical input also show promise for the development of new dynamic host–guest systems,^13,14 whereby electrocatalytic chain reactions enable guest molecule release,^14–18 ultimately affording a one-electron switch using only a catalytic amount of electrons as the input dots.

Herein, we demonstrate electrocatalytic chain reactions capable of turning on charge-transfer (CT) complex formation, accompanied by solvent molecule release and redox-active anthraquinone formation, for the first time (Scheme 1).¹⁹ Quinone derivatives are important electron carriers widely distributed in nature,²⁰ and have recently been utilized as guest molecules in host–guest systems.^18,21–23 Electron injection to a hydroxy anthrone derivative (AQH–CH₂CN) results in the formation of an anthraquinone radical anion (AQ˙⁻) and a stable coproduct (acetonitrile, CH₃CN) [Scheme 1a]. Redox-active fragment AQ˙⁻ has sufficient reducing power to reduce reactant AQH–CH₂CN (Scheme 1b), which initiates the electrocatalytic chain reaction (Scheme 1a). This electron upconversion produces a much lower amount of energy (3.5 kcal mol⁻¹, 0.15 eV; Scheme 1b) compared with electron upconversions used for synthetic purposes (20–25 kcal mol⁻¹).² Even the redox-active fragment (AQ˙⁻) is stable under mild reducing conditions, and the stable coproduct (CH₃CN) is a solvent molecule that has no impact on the electrocatalytic chain reaction. Therefore, the present electrocatalytic chain reactions are mild and sustainable, providing new opportunities for catalytic electron-triggered host–guest systems (Scheme 1c). Reactant AQH–CH₂CN is non-planar, making it unsuitable for host–guest systems driven by CT interactions. However, the electrocatalytic chain reaction converts AQH–CH₂CN to AQ, which is a planar electron-acceptor suitable for CT interactions, making turn-on generation of a host CT complex possible. Therefore, the sustainable electrocatalytic chain reactions demonstrated here can control host–guest molecular recognition using only a catalytic amount of electrons. This paves the way for a “one-electron switch” associated with catalytic electron-triggered turn-on molecular recognition (Scheme 2).

Scheme 1 (a) One-electron reduction of AQH–CH₂CN-initiated CH₃CN release through electrocatalytic chain reactions. ORTEP view (50% probability) of AQH–CH₂CN. Hydrogen atoms are omitted for clarity. (b) Redox potentials of AQ and AQH–CH₂CN in the electrocatalytic chain reaction. (c) Catalytic electron turns-on complex formation between AQ and U_HAnt₂ through sustainable electrocatalytic chain reactions.

Scheme 2 (a)–(e) Concept of a one-electron switch: (a) electron injection into a host/guest(1) system yields a one-electron reduced species of guest(1) [guest(1)˙⁻]. (b) Guest(1)˙⁻ spontaneously converts to guest(2)˙⁻ associated with a release of a stable coproduct. (c) Electron transfer from guest(2)˙⁻ to guest(1) occurs to yield a neutral guest(2) and guest(1)˙⁻, where guest(2) forms a host–guest complex [guest(2) ⊂ host]. (d) and (e) Sustainable electrocatalytic chain reactions convert all guest(1) into guest(2) to generate the guest(2) ⊂ host complexes.

Results and discussion

The hydroxy anthrone derivative (AQH–CH₂CN)²⁴ was synthesized by solvolysis of anthraquinone (AQ) under strong basic conditions in acetonitrile (see the ESI†). The resulting AQH–CH₂CN was successfully crystallized to reveal a C–C bond (d_c–c = 1.56 Å) between the CH₂CN moiety and the carbon atom of the anthrone ring at the connected position (Scheme 1a). Although that C–C bond is slightly longer than a conventional C–C bond (≤1.54 Å), it is not considered as a significant high-energy bond. Density functional theory (DFT) calculations [DFT/B3LYP-6-31G+(d,p)] predict that the dissociated product (i.e., AQ + CH₃CN) is 13.0 kcal mol⁻¹ lower in energy than AQH–CH₂CN, while AQH–CH₂CN itself is stable in the long term, thus indicating a relatively large activation barrier for the conversion of AQH–CH₂CN to AQ and CH₃CN before electron injection.

Fig. 1 shows the cyclic voltammograms of AQ and AQH–CH₂CN in deaerated acetonitrile. In contrast to the reversible cyclic voltammogram of AQ (Fig. 1a), the cyclic voltammogram of AQH–CH₂CN shows an irreversible redox wave (Fig. 1b), suggesting that the one-electron reduction of AQH–CH₂CN was associated with an irreversible process (vide infra). The one-electron reduction potential of AQH–CH₂CN was then determined as E⁰_red = −1.10 V (vs. Ag/Ag⁺) using differential pulse voltammetry (Fig. 1c), which is +0.15 V higher than the one-electron reduction potential of AQ [E⁰_red = −1.25 V (vs. Ag/Ag⁺)]. Not surprisingly, AQH–CH₂CN containing the central sp³ carbon finds it rather difficult to accept an electron. Indeed, the optimized structure [DFT/UB3LYP-6-31G+(d,p)] of AQH–CH₂CN˙⁻ suggests that the electron density of the singly occupied molecular orbital (SOMO) is low at the central sp³ carbon (Fig. 1d), while the SOMO orbital of AQ˙⁻ is fully delocalized across the entire molecule (Fig. 1e). Consequently, the SOMO energy level of AQH–CH₂CN˙⁻ is 0.55 eV (12.7 kcal mol⁻¹) higher than that of AQ˙⁻. Conversely, a fully dissociated state (i.e., AQ˙⁻ + CH₃CN) is 1.39 eV (32.1 kcal mol⁻¹) lower than AQH–CH₂CN˙⁻ (Fig. 1f).^25,26 Therefore, the coupling of the dissociation of CH₃CN to the one-electron reduction of AQH–CH₂CN (inset of Fig. 1c) should result in a significant shift of its one-electron reduction potential to the positive direction.

Fig. 1 (a) and (b) Cyclic voltammograms and (c) differential pulse voltammograms of (a) AQ (1.0 × 10⁻³ M) and (b) and (c) AQH–CH₂CN (1.0 × 10⁻³ M) in deaerated acetonitrile containing 0.1 M TBAP. Scan rate is 200 mV s⁻¹ (d) and (e) SOMO orbital for the optimized structures [DFT/UB3LYP-6-31G+(d,p)] of (d) AQH–CH₂CN˙⁻ and (e) AQ˙⁻. (f) Energy difference [DFT/UB3LYP-6-31G+(d,p)] between AQH–CH₂CN˙⁻ and its dissociated state (AQ˙⁻ + CH₃CN). ESR spectra of (g) AQH–CH₂CN (1.0 × 10⁻³ M) and (h) AQ (1.0 × 10⁻³ M) in the presence of Np˙⁻·Na⁺ (1.0 × 10⁻³ M) in deaerated acetonitrile. Inset: (c) schematic representation of the one-electron reduction of AQH–CH₂CN coupling with dissociation of CH₃CN.

Using electron spin resonance (ESR) spectroscopy, we then detected the product after electron injection to AQH–CH₂CN by addition of 1 equivalent of sodium naphthalenide (Np˙⁻·Na⁺).²⁷ The ESR spectrum for AQ˙⁻ was observed, which is identical to that obtained by the one-electron reduction of AQ (Fig. 1gvs.Fig. 1h), while no AQH–CH₂CN˙⁻ species was detected (Fig. 1g).

With these results, we then obtained the UV/Vis absorption spectra of AQH–CH₂CN with applied voltages of −0.8, −1.0, and −1.5 V (vs. Ag/Ag⁺) [Fig. 2a–c, respectively]. When the voltage of −0.8 V (vs. Ag/Ag⁺) was continuously applied to AQH–CH₂CN, no spectral change was observed (Fig. 2a). However, when the potential of −1.0 V (vs. Ag/Ag⁺) was applied to AQH–CH₂CN, UV/Vis absorption spectral changes were clearly observed with several isosbestic points (Fig. 2b). The resulting UV/Vis absorption spectrum after 290 s of applied voltage of −1.0 V (vs. Ag/Ag⁺) is identical to that of AQ (Fig. 2b red line vs.Fig. 2d), suggesting a quantitative conversion of AQH–CH₂CN to AQ (AQH–CH₂CN → AQ + CH₃CN), while no UV/Vis absorption bands due to AQ˙⁻ were observed (Fig. 2b). The one-electron oxidation potential of AQ˙⁻ [E⁰_ox = −1.25 V (vs. Ag/Ag⁺)] is lower than the applied potential [−1.0 V (vs. Ag/Ag⁺)], leading to rapid electron transfer to the working electrode and conversion of the temporarily generated AQ˙⁻ to AQ (Fig. 3a). In fact, the current flow was very low under these conditions (Fig. S3†).²⁸ No UV/Vis absorption spectral change took place for AQ under the −1.0 V (vs. Ag/Ag⁺) voltage (Fig. 2d), which is much higher than its one-electron reduction potential [E⁰_red = −1.25 V (vs. Ag/Ag⁺)]. Conversely, AQ was quantitatively converted to AQ˙⁻ under the more negative applied potential of −1.5 V (vs. Ag/Ag⁺) [Fig. 2e]. Then, the −1.5 V (vs. Ag/Ag⁺) potential was applied to AQH–CH₂CN, where AQH–CH₂CN also exhibited UV/Vis absorption spectral changes (Fig. 2c). The UV/Vis absorption spectrum of AQH–CH₂CN after 395 s of applied voltage of −1.5 V (vs. Ag/Ag⁺) is identical to that of AQ˙⁻ (Fig. 2c red line vs.Fig. 2e), suggesting that AQH–CH₂CN was directly converted to AQ˙⁻ (i.e., AQH–CH₂CN + e⁻ → AQ˙⁻ + CH₃CN) in this case. Here, the absorbance at 330 nm (due to AQ) increased upon applying the voltage of −1.5 V (vs. Ag/Ag⁺) to AQH–CH₂CN at t = 0–55 s, after which the absorbance decreased and reached saturation at t = 395 s (inset of Fig. 2c). The reactant of AQH–CH₂CN gave no appreciable absorption at 330 nm , and the molar absorption coefficient of AQ was larger than that of AQ˙⁻ at 330 nm (, inset of Fig. 2e). Hence, the initial increase and the subsequent decrease of absorbance observed at 330 nm (inset of Fig. 2c) indicated temporary generation of AQ and its subsequent conversion into AQ˙⁻ as explained below. Because the applied voltage of −1.5 V (vs. Ag/Ag⁺) is sufficiently lower than the one-electron oxidation potential of AQ˙⁻ [E⁰_ox = −1.25 V (vs. Ag/Ag⁺)], no further electron transfer to the electrode was expected (Fig. 3b(i)). The resulting AQ˙⁻ [E⁰_ox = −1.25 V (vs. Ag/Ag⁺)] has sufficient reducing power to reduce AQH–CH₂CN [E⁰_red = −1.10 V (vs. Ag/Ag⁺)], generating the secondary AQ˙⁻, while the original AQ˙⁻ was oxidized to AQ (Fig. 3b(ii)). Conversely, AQH–CH₂CN is easier to be reduced than AQ, and hence AQH–CH₂CN will more preferentially undergo one-electron reduction by the electrode than the subsequently formed AQ. In addition, chain electron transfer from AQ˙⁻ to AQH–CH₂CN also takes place to increase the concentration of AQ around the surface of the working electrode (Fig. 3b(i)–(iv)). When all the AQH–CH₂CN molecules were consumed, the AQ molecules generated by the electrocatalytic chain reactions started to be reduced by the electrode (Fig. 3b(v)).

Fig. 2 (a)–(e) UV/Vis absorption spectra of (a)–(c) AQH–CH₂CN (1.0 × 10⁻³ M) and (d) and (e) AQ (1.0 × 10⁻³ M) under applying potentials of (a) −0.8 V, (b) and (d) −1.0 V, (c) and (e) −1.5 V (vs. Ag/Ag⁺) in deaerated acetonitrile containing 0.1 M TBAP (1 mm cuvette). Insets: (c) and (e) corresponding time course at λ = 330 nm.

Fig. 3 (a) and (b) Schematic representation of the reactions that occurred around the surface of the working electrode under applying a potential of (a) −1.0 V or (b) −1.5 V (vs. Ag/Ag⁺) to AQH–CH₂CN, where diffusion of the molecules is omitted for clarity.

The electrocatalytic chain reaction was initiated by chemical reduction of AQH–CH₂CN (5.0 × 10⁻² M) using only 0.1 equivalent of Np˙⁻·Na⁺ (5.0 × 10⁻³ M) in deaerated THF (Fig. 4 and ESI Movie S1†). Before addition of Np˙⁻·Na⁺, the THF solution of AQH–CH₂CN was colorless and transparent; however, upon addition of Np˙⁻·Na⁺ (0.1 eq.), the solution color immediately turned red (Fig. 4b to c). The resulting red color corresponds to AQ˙⁻, whose red color was confirmed by electrochemical reduction of AQ (Fig. 4a). Then, the solution was suspended for 10 min after the Np˙⁻·Na⁺ addition, and the reaction was complete within 60 min under these conditions. The red color arising from AQ˙⁻ was maintained during the reaction (Fig. 4d and S4†). The resulting yellow precipitate was AQ that was not completely soluble in THF under the high concentration conditions employed ([C]₀ = 5.0 × 10⁻² M).

Fig. 4 (a) Photograph of the cuvette after electrochemical reduction of AQ. (b)–(d) Photographs of a deaerated THF solution containing AQH–CH₂CN (5.0 × 10⁻² M) (b) before addition, (c) upon addition, and (d) after addition of Np˙⁻·Na⁺ (5.0 × 10⁻³ M) at 0–60 min.

The CH₃CN generated during the electrocatalytic chain reactions was monitored by NMR spectroscopy (Fig. 5). Upon introduction of 0.1 equivalent of Np˙⁻·Na⁺ (2.0 × 10⁻³ M) into AQH–CH₂CN (2.0 × 10⁻² M) in deaerated THF-d₈, ¹H NMR signals corresponding to AQH–CH₂CN (blue triangles) gradually decreased and a concomitant increase of signals arising from the formation of AQ (red circles) and CH₃CN (green triangles) was observed (Fig. 5a and S5†). The corresponding time course curves (Fig. 5b) showed that the initial 20 mM of AQH–CH₂CN (blue triangles) was almost completely consumed at t = 138 min, where the formation of AQ (red circles) and the associated CH₃CN released (green triangles) reached saturation.²⁹

Fig. 5 (a) ¹H NMR spectra of AQH–CH₂CN (2.0 × 10⁻² M) upon addition of Np˙⁻·Na⁺ (2.0 × 10⁻³ M) in THF-d₈. Signal denoted by an asterisk is from the solvent. The intensity of the spectra at 1.5–2.0 ppm is scaled vertically (1/2) for clarity. (b) Corresponding time course.

In light of these results, we can summarize the electrocatalytic chain reaction (Scheme 1) as follows: at the initial step, electron injection to AQH–CH₂CN produces AQ˙⁻, with the process driven by bond breaking associated with proton migration. The initially formed AQ˙⁻ can subsequently reduce the AQH–CH₂CN reactant to generate secondary AQ˙⁻ and CH₃CN, while the initial AQ˙⁻ is oxidized to AQ (Scheme 1a). Then, the secondary-formed AQ˙⁻ is also capable of reducing the reactant of AQH–CH₂CN to generate a next AQ˙⁻ associated with the concomitant release of CH₃CN (Scheme 1a). Thus, conversion of AQH–CH₂CN to AQ and CH₃CN is the electrocatalytic chain reaction process mediated by AQ˙⁻ (Scheme 1a). The electrocatalytic chain reactions are terminated when all the AQH–CH₂CN is consumed, while the concentration of AQ˙⁻ remains unchanged during the chain reaction. If the present electrocatalytic chain reaction mechanism is valid, the rates of the concentration changes (|dC/dt|) of AQ, CH₃CN, and AQH–CH₂CN would obey pseudo-first-order kinetics because the observed rate (|dC/dt|) can be represented as k_et[AQ˙⁻][C], where the [AQ˙⁻] term is constant during the chain reaction. This was confirmed by time course curves obtained from the above ¹H NMR kinetics experiments, where each time course obeyed first-order kinetics with very similar pseudo-first-order rate constants (k_obs = k_et[AQ˙⁻] = 3.3–3.9 × 10⁻⁴ s⁻¹) [Fig. 5b]. The observed reaction rate also increased with an increase in the initial concentration of the added Np˙⁻·Na⁺ (i.e., [Np˙⁻·Na⁺]₀ = [AQ˙⁻]) [Fig. S6†]. Therefore, we can extract the electron-transfer rate constant (k_et = 1.6–1.9 × 10⁻¹ M⁻¹ s⁻¹) by using the initial concentration of Np˙⁻·Na⁺ (2.0 × 10⁻³ M). The determined k_et value (k_et = 1.6–1.9 × 10⁻¹ M⁻¹ s⁻¹) is reasonable for slightly exergonic electron transfer with a large reorganization energy arising from the bond breaking/proton migration.³⁰ Moreover, the conversion of AQH–CH₂CN to AQ and CH₃CN can be completed only with 0.025 equivalent of Np˙⁻·Na⁺ used as the initiator (Fig. S7†), suggesting that the turnover number of the electrocatalytic chain reactions is at least greater than 40.

As the present electrocatalytic chain reactions were mild and sustainable, we investigated their application to CT interaction turn-on systems (vide infra). Electron donor host compound U_HAnt₂, first synthesized by Lehn et al., was used for this purpose.³¹ U_HAnt₂ bears anthracene tweezers suitable for insertion of a planar electron-acceptor molecule to form a stable D–A–D-type CT complex.³¹ The possibility of CT complex formation between U_HAnt₂ and AQ was predicted by DFT modeling [DFT/CAM-B3LYP-6-31G(d)], which suggested that D–A–D-type CT complex formation (AQ ⊂ U_HAnt₂) was reasonable (Fig. 6a and S8†). Furthermore, the ¹H NMR spectrum of U_HAnt₂ showed that anthracene aromatic protons (H⁷, H⁸, and H⁹) were shifted upfield upon addition of AQ (Fig. 6b and S9–S13†). This observation was consistent with the stacked D–A–D CT model comprising anthracene rings shielded by the central anthraquinone plane (Fig. 6a). The electrocatalytic conversion of AQH–CH₂CN to AQ was then examined by applying a potential of −1.0 V (vs. Ag/Ag⁺) to AQH–CH₂CN in the presence of U_HAnt₂. The absorption band resulting from AQ formation was successfully observed, along with the appearance of a broad absorption band at around 410 nm (Fig. 7a). Successive addition of AQ to U_HAnt₂ resulted in appearance of the same broad absorption band in a longer wavelength region (Fig. S14†). Conversely, both U_HAnt₂ alone and AQ alone showed no absorption band in this longer wavelength region (Fig. S15 and S16†). Time-dependent (TD) DFT of AQ ⊂ U_HAnt₂ indicated that the newly observed broad absorption band was attributed to the CT transition from the anthracene tweezers to the inserted central AQ molecule (Fig. 7b). Further assignment of the CT band was performed with quinones having different one-electron reduction potentials (see details in Fig. S17–S20†). During the electrocatalytic conversion, U_HAnt₂ showed no degradation (Fig. 7a). Besides, DFT and electrochemical studies (Fig. S8, S21 and S22†) revealed that the mediator AQ˙⁻ (reduced form) has no interaction with U_HAnt₂, enabling the sustainable electrocatalytic chain reaction even in the presence of U_HAnt₂. In contrast, a mixture of U_HAnt₂ and AQH–CH₂CN provided no CT absorption band (Fig. S23†), suggesting no effective CT interaction between them. Reactant AQH–CH₂CN is a non-planar molecule containing several sp³ carbon atoms, making it unsuitable for CT interaction with the anthracene tweezers of U_HAnt₂. Furthermore, the CT interactions could also be turned on by electrocatalytic conversion of AQH–CH₂CN to AQ initiated by chemical reduction using Np˙⁻·Na⁺ (Fig. S24†). Therefore, the CT interactions could be turned on by electrocatalytic conversion of AQH–CH₂CN to AQ for CT complex (AQ ⊂ U_HAnt₂) formation (Scheme 1c; inset of Fig. 7a).³²

Fig. 6 (a) Optimized structures [DFT/CAM-B3LYP-6-31G(d)] of U_HAnt₂ and AQ ⊂ U_HAnt₂. (b) Stacked ¹H NMR spectra (for H³′, H⁹, and H^a) of U_HAnt₂ (3.5 × 10⁻³ M) in the presence of AQ (0–7.0 × 10⁻³ M) in CDCl₃.

Fig. 7 (a) UV/Vis absorption spectra of AQH–CH₂CN (8.0 × 10⁻⁴ M) in the presence of U_HAnt₂ (8.0 × 10⁻⁴ M) under applying a potential of −1.0 V (vs. Ag/Ag⁺) (0–395 s) in deaerated THF containing 0.1 M TBAP (1 mm cuvette). (b) Calculated UV/Vis absorption spectrum [TD-DFT/CAM-B3LYP-6-31G(d), IEFPCM:THF] of the optimized structure of AQ ⊂ U_HAnt₂ [DFT/CAM-B3LYP-6-31G(d), IEFPCM:THF]. Insets: (a) schematic representation of CT complex formation turned on by the catalytic electron. (b) Summary of excited states 1 and 2.

Conclusions

In conclusion, we successfully demonstrated the electron injection-triggered solvent (CH₃CN) molecular release associated with formation of a redox active quinone derivative (AQ) through sustainable electrocatalytic chain reactions. The presented electrocatalytic chain reactions enable guest molecular release and should be mild enough to apply to a wide range of host–guest systems. This study reports a first example of successful catalytic electron-triggered CT complex formation. These findings provide new opportunities for creating a “one-electron switch” (Scheme 2) in association with catalytic electrons as the input dots to trigger specific molecular recognition.

Author contributions

Y. O. and S. I. performed the synthesis and characterization of materials, and also contributed the titration experiments. D. O. contributed the analysis of the experimental data. J. Y. designed the study, analysed the experimental data and wrote the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

This work was partly supported by JSPS KAKENHI, Grant Numbers: JP19H02693, JP19K22207, and JP21H05404 in Scientific Research on Innovative Areas “Dynamic Exciton”, a Grant-in-Aid for the ASAHI Glass Foundation, and the Iketani Foundation. We thank Iain Mackie, PhD, and Simon Partridge, PhD, from Edanz (https://jp.edanz.com/ac) for editing a draft of this manuscript.

Notes and references

1. (a) A. Studer and D. P. Curran, Nat. Chem., 2014, 6, 765; (b) A. Studer and D. P. Curran, Angew. Chem., Int. Ed., 2016, 55, 58.
2. M. A. Syroeshkin, F. Kuriakose, E. A. Saverina, V. A. Timofeeva, M. P. Egorov and I. V. Alabugin, Angew. Chem., Int. Ed., 2019, 58, 5532.
3. C. Costentin, M. Robert and J.-M. Saveant, Chem. Phys., 2006, 324, 40.
4. M. A. Syroeshkin, I. B. Krylov, A. M. Hughes, I. V. Alabugin, D. V. Nasybullina, M. Y. Sharipov, V. P. Gultyai and A. O. Terent'ev, J. Phys. Org. Chem., 2017, 30, e3744.
5. X. Tang and A. Studer, Angew. Chem., Int. Ed., 2018, 57, 814.
6. M. Vacher, I. F. Galvan, B.-W. Ding, S. Schramm, R. Berraud-Pache, P. Naumov, N. Ferre, Y.-J. Liu, I. Navizet, D. Roca-Sanjuan, W. J. Baader and R. Lindh, Chem. Rev., 2018, 118, 6927.
7. D. Davis, V. P. Vysotskiy, Y. Sajeev and L. S. Cederbaum, Angew. Chem., Int. Ed., 2012, 51, 8003.
8. (a) A. Studer and D. P. Curran, Angew. Chem., Int. Ed., 2011, 50, 5018; (b) S. Zhou, G. M. Anderson, B. Mondal, E. Doni, V. Ironmonger, M. Kranz, T. Tuttle and J. A. Murphy, Chem. Sci., 2014, 5, 476; (c) J. P. Barham, G. Coulthard, R. G. Kane, N. Delgado, M. P. John and J. A. Murphy, Angew. Chem., Int. Ed., 2016, 55, 4492.
9. J. Kim, D. J. Darley, W. Buckel and A. J. Pierik, Nature, 2008, 452, 239.
10. M. Majek and A. Jacobi von Wangelin, Angew. Chem., Int. Ed., 2015, 54, 2270.
11. J. Du, K. L. Skubi, D. M. Schultz and T. P. Yoon, Science, 2014, 344, 392.
12. Similar situations can be found in oxidant upconversion, where hole catalysis has been utilized as an alternative input to induce the isomerization of photochromic compounds, see: (a) T. Nakashima, Y. Kajiki, S. Fukumoto, M. Taguchi, S. Nagao, S. Hirota and T. Kawai, J. Am. Chem. Soc., 2012, 134, 19877; (b) S. Lee, Y. You, K. Ohkubo, S. Fukuzumi and W. Nam, Chem. Sci., 2014, 5, 1463; (c) A. Goulet-Hanssens, C. Rietze, E. Titov, L. Abdullahu, L. Grubert, P. Saalfrank and S. Hecht, Chem, 2018, 4, 1740.
13. J. Yuasa and S. Fukuzumi, J. Am. Chem. Soc., 2007, 129, 12912.
14. (a) V. Croue, S. Goeb, G. Szaloki, M. Allain and M. Salle, Angew. Chem., Int. Ed., 2016, 55, 1746; (b) G. Szaloki, V. Croue, V. Carre, F. Aubriet, O. Aleveque, E. Levillain, M. Allain, J. Arago, E. Orti, S. Goeb and M. Salle, Angew. Chem., Int. Ed., 2017, 56, 16272.
15. (a) D. Ogata and J. Yuasa, Angew. Chem., Int. Ed., 2019, 58, 18424; (b) Y. Imai and J. Yuasa, Chem. Sci., 2019, 10, 4236; (c) K. Hamashima and J. Yuasa, Angew. Chem., Int. Ed., 2022, 61, e202113914.
16. M. Han, R. Michel, B. He, Y.-S. Chen, D. Stalke, M. John and G. H. Clever, Angew. Chem., Int. Ed., 2013, 52, 1319.
17. S. Hiraoka, K. Harano, M. Shiro and M. Shionoya, Angew. Chem., Int. Ed., 2005, 44, 2727.
18. L. S. Lisboa, J. A. Findlay, L. J. Wright, C. G. Hartinger and J. D. Crowley, Angew. Chem., Int. Ed., 2020, 59, 11101.
19. After we submitted this manuscript, Stoddart et al. reported an important and excellent example, which also underlined the importance of an electron as a catalyst (trigger) of molecular recognition, see: Y. Jiao, Y. Qiu, L. Zhang, W.-G. Liu, H. Mao, H. Chen, Y. Feng, K. Cai, D. Shen, B. Song, X.-Y. Chen, X. Li, X. Zhao, R. M. Young, C. L. Stern, M. R. Wasielewski, R. D. Astumian, W. A. Goddard III and J. F. Stoddart, Nature, 2022, 603, 265.
20. C. C. Moser, J. M. Keske, K. Warncke, R. S. Farid and P. L. Dutton, Nature, 1992, 355, 796.
21. S. Iseki, K. Nonomura, S. Kishida, D. Ogata and J. Yuasa, J. Am. Chem. Soc., 2020, 142, 15842.
22. (a) D. P. August, G. S. Nichol and P. J. Lusby, Angew. Chem., Int. Ed., 2016, 55, 15022; (b) V. Martí-Centelles, A. L. Lawrence and P. J. Lusby, J. Am. Chem. Soc., 2018, 140, 2862.
23. S. Sugiura and H. Maeda, Chem. Commun., 2021, 57, 6983.
24. S. M. Lukonina, S. V. Mikhailova, I. K. Korobeinicheva, v. A. Loskutov and E. P. Fokin, Russ. Chem. Bull., 1983, 32, 2379.
25. The free energy change of radical-anionic version of the fragmentation should be much smaller in the real system (see details in Fig. S1†).
26. Although the detailed mechanism of the fragmentation of acetonitrile in AQH–CH₂CN˙⁻ could not be established, DFT study indicates that the fragmentation is likely to be triggered by the intramolecular proton transfer (see details in Fig. S2†).
27. Y. Inui, S. Miyazaki, K. Ohkubo, S. Fukuzumi and T. Kojima, Angew. Chem., Int. Ed., 2012, 51, 4623.
28. Accurate determination of the coulombic efficiency could be difficult in this system, because the current flow was extremely low under the potential of −1.0 V (vs. Ag/Ag⁺) (Fig. S3†).
29. The remaining (ungenerated) 2 mM of AQ should still exist as AQ˙⁻ formed by the initial electron transfer from Np˙⁻·Na⁺ to AQH–CH₂CN. Conversely, undetected remaining CH₃CN may result from volatilization.
30. J. Yuasa, S. Yamada and S. Fukuzumi, J. Am. Chem. Soc., 2008, 130, 5808.
31. A. Petitjean, R. G. Khoury, N. Kyritsakas and J.-M. Lehn, J. Am. Chem. Soc., 2004, 126, 6637.
32. The titration experiment indicated that the binding constant (K_CT) between U_HAnt₂ and AQ is 140 M⁻¹ (Fig. S25†).

---

Footnote

† Electronic supplementary information (ESI) available. CCDC 2096046. For ESI and crystallographic data in CIF or other electronic format see https://doi.org/10.1039/d2sc01342h

---