Reactivities of Hydroxycinnamic Acid Derivatives Involving Caffeic Acid toward Electrogenerated Superoxide in N,N-Dimethylformamide

Abstract

Reactivity of (2E)-3-(3,4-dihydroxyphenyl)prop-2-enoic acid (caffeic acid), classified as a hydroxycinnamic acid (HCA) derivative, toward electrogenerated superoxide radical anion (O₂^•−) was investigated through cyclic voltammetry, in situ electrolytic electron spin resonance spectrometry, and in situ electrolytic ultraviolet–visible spectrometry in N,N-dimethylformamide (DMF), aided by density functional theory (DFT) calculations. The quasi-reversible redox of dioxygen/O₂^•− is modified in the presence of caffeic acid, suggesting that O₂^•− is scavenged by caffeic acid through proton-coupled electron transfer. The reactivities of caffeic acid toward O₂^•− are mediated by the ortho-diphenol (catechol) moiety rather than by the acryloyl group, as experimentally confirmed in comparative analyses with other HCAs. The electrochemical and DFT results in DMF suggested that a concerted two-proton-coupled electron transfer mechanism proceeds via the catechol moiety. This mechanism embodies the superior kinetics of O₂^•− scavenging by caffeic acid.

1. Introduction

Caffeic acid ((2E)-3-(3,4-dihydroxyphenyl)prop-2-enoic acid, CafH₂(COOH)) is a natural phenolic compound classified as a hydroxycinnamic acid (HCA) derivative. All plants produce caffeic acid as an intermediate in the biosynthesis of lignin, a principal component of woody plant biomass [1]. CafH₂(COOH) delivers a surprising number of health benefits; for example, it mitigates carcinogenesis, neurodegeneration, and other age-related diseases, and induces pharmacological activities such as immunomodulatory and anti-inflammatory effects [2,3,4,5]. The carcinogenicities and anticancer activities of CafH₂(COOH) have been enthusiastically researched but with mixed results [2,6]. CafH₂(COOH) can target several diseases by scavenging the reactive oxygen species (ROS) generated in biochemical processes and pathologies. Therefore, ROS scavenging by HCAs involving CafH₂(COOH) as an antioxidant might play a therapeutic role. However, whether CafH₂(COOH) exerts a substantial effect on any human disease or delivers health benefits to humans remains unclear because solid evidence is lacking.

To confirm the medicinal effect of HCAs, especially of CafH₂(COOH) that possesses antioxidant activity, we must elucidate the chemical reaction mechanism, including the electron transfer (ET) reaction that scavenges ROS such as superoxide radical anions (O₂^•−, the precursor of the hydroperoxyl radical, HO₂^•) and the hydroxyl radical (HO^•). Clarifying the mechanism of ROS scavenging by HCAs is necessary for understanding the related medicinal effect, as the ROS generated around lesions and inflammatory organs are direct causes of several pathologies [4,6]. Previous studies in the field of electrochemistry have reported how HCA is oxidized by electrodes in aqueous or non-aqueous (aprotic) solvents [7,8,9], where the reactivity of HCAs toward the ROS was imitated by the heterogeneous electrochemical process. In these studies, antioxidant activity of HCA was assessed using cyclicvoltammetry, although the electrooxidation behavior varies depending on the presence or absence of protons. In the pioneering work by Hotta et al. [9], a well-developed two-electron reversible redox of CafH₂(COOH) in acidic-aqueous media was observed, demonstrating that protons from the aqueous solvent relate to the oxidation. Conversely, cyclic voltammograms (CVs) of CafH₂(COOH) in an aprotic acetonitrile solvent demonstrated an irreversible oxidation, as shown in the pioneering work by Masek et al. [8]. These electrochemical results imply that the chemical reaction mechanism involving proton transfer (PT) and ET between HCAs and the ROS cannot be observed in aqueous solvents, because solvent-derived protons interfere with the PT from hydroxyl groups of HCA. As is widely known, effective antioxidant activity usually requires the presence of ortho- or para-diphenolic hydroxyl groups (OH) for quinone–hydroquinone π-conjugation. CafH₂(COOH) belongs to the phenylpropanoids, possessing both an ortho-diphenol (catechol, CatH₂) moiety and an acryloyl group linked via a C6–C3 skeleton. Antioxidant studies of CafH₂(COOH) employing numerous assays and methodologies have been published, such as 1,1-diphenyl-2-picryl-hydrazyl free radical scavenging, 2-azino-bis(3-ethylbenzthiazoline-6-sulfonic acid) radical scavenging, total antioxidant activity by the ferric thiocyanate method, and O₂^•− scavenging with metal chelating activities [2,3]. Meanwhile, computational studies have provided mechanistic insights into the structure–antioxidant activity relationships of HCAs [10,11,12,13]. These studies evaluated the thermochemical properties of HCAs, such as their bond dissociation energies of OH and ionization potentials, thereby clarifying the energetics of the ROS-scavenging reaction. A pioneering work by Nsangou et al. [14] investigated the antioxidant activity of CafH₂(COOH) toward HO₂^• and HO^•. Being more reactive than O₂^•−, these radicals are substantial cellular toxins in the living body. From the structure–property relationships of ROS scavenging by CafH₂(COOH), the activity was mainly contributed by the CatH₂ moiety, which has a high degree of conjugation and can delocalize π-electrons with a resonance structure [11,13]. The characteristic resonance effect of CafH₂(COOH) is considered to thermodynamically stabilize its radical product. However, isolated HO₂^• and HO^• are highly reactive and their reactions are difficult to observe experimentally; consequently, conclusive evidence of the ROS-scavenging mechanism of CafH₂(COOH) is lacking in the literature.

There are several possible mechanisms of ROS (O₂^•−, HO^•, and HO₂^•) scavenging by HCAs: single ET, hydrogen atom transfer (HAT) involving proton-coupled electron transfer (PCET) [15,16,17,18,19], sequential proton-loss electron transfer [20], and superoxide-facilitated oxidation (SFO) [21,22,23]. Along the SFO mechanism, the initial PT from the substrate to O₂^•− generates HO₂^•, and a rapid dismutation into hydroperoxide (H₂O₂) and dioxygen (O₂) follows. Then, the O₂ formed in the dismutation process oxidizes the substrate anion [24]. On the other side, the other mechanism involves direct ET. In an aprotic solution, various phenolic antioxidants, polyphenols [18], 1,2- and 1,4-benzendiols (catechol [16] and hydroquinone [17]), and monophenols including aminophenols [19,25,26], scavenge electrogenerated O₂^•− through the PCET characterized by quinone–hydroquinone π-conjugation. Although O₂^•− is not so electrophilic but a Brønsted base, HO₂^• formed through the protonation of O₂^•− is a strong oxidant with a short lifetime. Therefore, the initial PT and subsequent oxidation (ET) are closely related to each other through the PCET between CafH₂(COOH)/anion and O₂^•−/HO₂^• that embodies the actual mechanism of O₂^•−/HO₂^• scavenging.

In this study, we investigate the homogeneous chemical reactivities between electrogenerated O₂^•− and HCAs involving CafH₂(COOH) using cyclicvoltammetry, in situ electrolytic electron spin resonance (ESR) spectral measurements, and in situ ultraviolet–visible (UV–vis) spectral measurements in N,N-dimethylformamide (DMF) solution. Avoiding the interference of solvent-derived protons, the ROS scavenging reaction by HCAs is inferred from the electrochemical measurements by applying density functional theory (DFT). In addition, this study provides mechanistic insights into the structural features of CafH₂(COOH), namely, the CatH₂ moiety and the acryloyl group for the PCET mechanism characterized by π-conjugated redox reaction. Accordingly, we present valuable information regarding O₂^•− scavenging by HCAs including CafH₂(COOH), which is assumed to provide substantial health benefits as a phytoalexin.

2. Materials and Methods

2.1. Chemicals

Highest-grade (2E)-3-(3,4-dihydroxyphenyl)prop-2-enoic acid (caffeic acid, CafH₂(COOH), >98.0%), (2E)-3-(4-hydroxy-3-methoxyphenyl)prop-2-enoic acid (ferulic acid, >99.0%), (E)-3-(3-hydroxy-4-methoxyphenyl)prop-2-enoic acid (isoferulic acid, >97.0%), ethyl (E)-3-(3,4-dihydroxyphenyl)prop-2-enoate (ethyl caffeate, Et-CafH₂, >95.0%), (E)-3-(4-hydroxyphenyl)-2-propenoic acid (p-coumaric acid, >98.0%), and sodium methoxide (CH₃ONa, 95.0%) were purchased from Sigma-Aldrich Inc. (Tokyo, Japan) and were used as received. DMF (spectrograde, 99.7%) was used as the solvent for electrochemical and electrolytic ESR/UV–vis spectral measurements, and ferrocene (Fc) used as a potential reference compound, were purchased from Nacalai Tesque Inc. (Kyoto, Japan) and used as received. Dinitrogen (N₂) gas (99.0%) and O₂ gas (99.0%) were obtained from Medical Sakai Co., Ltd. (Gifu, Japan). Tetrapropylammonium perchlorate (TPAP, >98.0%) used as a supporting electrolyte was purchased from Tokyo Chemical Industry Co., Ltd. (Tokyo, Japan) and prepared as described previously [27].

2.2. Cyclic Voltammetry and In Situ Electrolytic ESR/UV–Vis Spectrum Measurements

Cyclic voltammetry was performed in a standard three-electrode system: working electrode, a 1.0-mm-diameter glassy carbon (GC); counter electrode, a coiled platinum (Pt, 99.99%); reference electrode, a silver/silver nitrate (Ag/AgNO₃) (containing an acetonitrile solution of 0.1 mol dm⁻³ tetrabutylammonium perchlorate and 0.01 mol dm⁻³ AgNO₃). All measurements were performed at 25 °C using an ECstat-301 electrochemical analyzer (EC-frontier Co., Ltd., Kyoto, Japan) and electrochemical software (Supplementary Materials, Table S1). Before the experiments, the working electrode was polished with alumina paste on a polishing wheel, rinsed with deionized water and acetone, and dried. Calibration of the reference electrode was referenced to the ferrocenim ion/ferrocene couple (Fc⁺/Fc), and all reported potentials are referenced to the potential of the Fc⁺/Fc couple. A JES-FA200 X-band spectrometer (JEOL Ltd., Tokyo, Japan) and an OCEAN HDX spectrometer (OptoSirius Co., Ltd., Tokyo, Japan) were used for ESR and UV–vis spectral measurements, respectively. The controlled-potential electrolysis was performed at room temperature in two types of cells: an in situ electrolytic ESR cell with a 0.5-mm-diameter straight Pt wire sealed in a glass capillary as the working electrode (5.0-mm-length), and an optically transparent thin-layer electrochemical (OTTLE) cell (path length: 1.0 mm) with a Pt-mesh working electrode (10.0 mm × 20.0 mm) sandwiched between two pieces of quartz glass (Supplementary Materials, Figure S1). The working electrodes (planar glassy carbon, Pt wire, and Pt mesh) were chosen because they do not cause side reactions such as electrode-absorption with solutions or substrates. Samples were prepared in a glove box completely filled with N₂ gas to prevent contamination by moisture. The DMF solutions were saturated with O₂ by air-bubbling the gas for ca. 2–3 min, and the gas was passed over the solutions during the measurements to maintain a constant concentration of O₂ at 4.8 × 10⁻³ mol dm⁻³.

2.3. Theoretical Calculations

All calculations were performed at the DFT level with the Becke three-parameter Lee–Yang–Parr (B3LYP) hybrid functional implemented in the Gaussian 16 Program package [28]. This functional was chosen because it obtains good geometries of the reactants, transition states (TS), and products in PCET reactions between phenolic compounds and free radicals [29]. The highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) were computed from geometry optimization with frontier orbital theory. In the calculations for geometry optimization, vibrational frequency calculation, intrinsic reaction coordinate (IRC) calculation, and population analysis of each compound, we employed the standard split-valence triple ζ basis sets augmented by the polarization 3df,2p and diffusion orbitals 6-311+G(3df,2p). The polarized continuum model (PCM) under the default settings of Gaussian 16 was used to estimate the solvent contribution of DMF to the standard Gibbs free energies, which is widely employed in thermodynamic characteristic studies of solvation. The zero-point energies and thermal correction, together with the entropy, were used to convert the internal energies to standard Gibbs energies at 298.15 K. The natural bond orbital (NBO) technique was used for electron and spin calculations in the population analysis [30].

3. Results and Discussion

3.1. Cyclic Voltammetry of O₂/O₂^•− in the Presence of HCAs

Reactivities of HCAs shown in Scheme 1 toward electrogenerated O₂^•− were demonstrated in the CVs of 4.8 × 10⁻³ mol dm⁻³ of O₂ in the presence of HCAs in DMF (Figure 1). In aprotic solvents such as DMF, the relationship between ET and PT is well demonstrated in electrochemical and electrolytic-spectral measurements, avoiding interference of solvent-derived protons. In the CVs, one-electron reduction of O₂ (Equation (1)) is demonstrated in a quasi-reversible redox reaction with the initial cathodic scan generating O₂^•− and following reoxidation to O₂ in the returned anodic scan (1c/1a, bold lines in Figure 1). The generated O₂^•− is lowly reactive toward aprotic DMF. The reversible CVs investigated here became irreversible in the presence of phenolic HCAs (Figure 1a–e) at various concentrations (0 to 5.0 × 10⁻³ mol dm⁻³). As the CVs of bubbled N₂ showed no peak over the potential range, the loss of reversibility in the CVs of O₂/O₂^•− was attributed to an acid–base reaction; specifically, the O₂^•− acts as a Brønsted base forming HO₂^• along the initial PT (Equation (2)).

The initial PT and consequent reduction of HO₂^• (Equation (3)) gave rise to bielectronic CVs involving its cathodic current (2c) in the presence of ferulic acid, isoferulic acid, and p-coumaric acid (Figure 1b,c,e, respectively). This bielectronic CV did not appear in the presence of CafH₂(COOH) and Et-CafH₂ (Figure 1a,d, respectively) and was replaced by a prepeak (3c), demonstrating scavenging of HO₂^• by ET from the deprotonated anions (CafH(COOH)⁻ and Et-CafH⁻) (Equation (4)) forming substrate radicals (CafH(COOH)^• and Et-CafH^•) and hydroperoxyl anion (HO₂⁻). Consequently, a monoelectronic CV appeared.

O₂ + e− ↔ O₂^•− (E° = −1.284 V vs. Fc⁺/Fc)

(1)

O₂^•− + CafH₂ → HO₂^• + CafH(COOH)⁻ (the initial PT)

(2)

HO₂^• + e⁻ → HO₂⁻ (E° = −0.4 to −0.2 V vs. Fc⁺/Fc)

(3)

HO₂^• + CafH(COOH)⁻ → HO₂⁻ + CafH(COOH)^• (ET)

(4)

HO₂⁻ + CafH(COOH)^• → H₂O₂ + Caf(COOH)^•− (the second PT)

(5)

Based on the CVs, we rationalized that O₂^•− formation associated with PT from HCA leads to the irreversible bielectronic reduction of O₂ to H₂O₂ driven by the exergonic reduction of HO₂^•/HO₂⁻. Additionally, we divided the CVs of O₂/O₂^•− in the presence of HCAs (Figure 1a–e) into two typical curves: type A, an irreversible two-electron process observed in electro–chemical–electro reactions (Equations (1)–(3)); and type B, an irreversible one-electron process observed in electro–chemical–chemical reactions (Equations (1), (2), (4) and (5)) leading to the scavenging of O₂^•−. Figure 2 summarizes Equations (1)–(5) showing the electrochemical mechanisms of O₂/O₂^•− in the presence of (a) CafH₂(COOH) and (b) ferulic acid.

In this scenario, the CVs (Figure 1) recorded in the presence of ferulic acid, isoferulic acid, and p-coumaric acid were type A, demonstrating the absence of O₂^•− scavenging and the appearance of a cathodic-current of HO₂^•. Conversely, the CVs in the presence of CafH₂(COOH) and Et-CafH₂ were type B, demonstrating the scavenging of O₂^•−/HO₂^•. The prepeak in the CVs of type B is mainly ascribed to radical anions (Caf(COOH)^•− and Et-Caf^•−), which are formed through the PCET from the CatH₂ moiety of the substrate. The slightly different cathodic prepeaks in Figure 1a,d can be explained by the reduction of HO₂^• via PT from the acryloyl group of CafH₂(COOH). These CV results suggest a PCET reaction between O₂^•− and its CatH₂ moiety, which is independent of PT from the acryloyl group.

3.2. In Situ Electrolytic ESR/UV–Vis Spectral Analyses of O₂/O₂^•− in the Presence of HCAs

To confirm the O₂^•− scavenging by HCAs, the solutions of the CV experiment were analyzed by electrolytic ESR using an in situ cell (scanning time = 4 min) and by UV–vis spectral measurements using an OTTLE cell (Supplementary Materials, Figure S1). These spectra were acquired under a potential applied at −1.3 V, which corresponds to the electrogeneration of O₂^•− (Equation (1)). ESR spectra were obtained only in the presence of CafH₂(COOH) and Et-CafH₂, suggesting that O₂^•− was scavenged in these cases (Figure 3a). An almost identical ESR spectrum was obtained in the presence of Et-CafH₂ (data not shown). In these ESR spectra, the fine coupling constants for hydrogen (a_H/mT) were assigned to hydrogens of CatH₂ (H^a, H^b, H^c: 0.23, 0.21, 0.19 mT) in the substrate radical anions (Caf(COOH)^•− and Et-Caf^•−) formed through the PCET. This result demonstrates that radical spins are barely distributed on the carboxyl group of CafH₂(COOH) and the ethyl ester group of Et-CafH₂. The in situ ESR system, which is sensitive to hyperfine couplings, showed no clear splitting derived from acryloyl hydrogens (H^d, H^e: 0.01 mT).

Figure 3b shows the in situ electrolytic UV–vis spectra of the CV solution containing CafH₂(COOH) (1.0 × 10⁻³ mol dm⁻³). These spectra were measured in the absence and presence of O₂ (in the former case, O₂ was removed by purging with N₂). The spectrum of CafH₂(COOH) alone presented characteristic absorption bands centered at 218, 278, and 323 nm. This spectrum was unchanged under an applied potential of 1.0 to −2.0 V vs. Fc⁺/Fc without O₂ (data not shown), demonstrating that CafH₂(COOH) was not electrolyzed without deprotonation. Additionally, the spectrum did not alter in the presence of 2.0 × 10⁻³ mol dm⁻³ CH₃ONa without applying potentials (red line in Figure 3b). As CafH₂(COOH) is deprotonated by the Brønsted base CH₃ONa, this spectrum was attributed to a mixture of CafH₂(COOH), CafH(COOH)⁻, and CafH₂(COO)⁻. Under O₂ (4.8 × 10⁻³ mol dm⁻³), the spectrum diminished at applied cathodic potentials above −1.0 V. From these spectral changes, the homogeneous reaction between CafH₂(COOH) and the electrogenerated O₂^•− was inferred to involve the initial PT and the subsequent ET forming Caf(COOH)^•−. This radical product was detectable in the in situ electrolytic ESR system, conversely undetectable in the UV–vis spectra, showing a decomposition of the radical product (Caf(COOH)^•−). Analogously to the CV results, the spectra revealed that the initial PT (Equation (2)) and the following reactions involving ET between HO₂^• and CafH(COOH)⁻ (Equation (4)) rapidly underwent base-catalyzed oxidation. The CV and spectral measurements imply that O₂^•− was successfully scavenged by CafH₂(COOH) through the PCET over the time scale of the CV measurements.

3.3. DFT Optimization of the Stable Structure of CafH₂(COOH) and Its Deprotonated Anion

To clarify the mechanism of the PCET between CafH₂(COOH) and O₂^•− in DMF, we performed DFT calculations using the B3LYP hybrid functional employing the PCM method. First, the stable structures of CafH₂(COOH) and the conformers of its deprotonated anions (CafH₂(COO)⁻, CafH(COOH)⁻) along the initial PT route were obtained from energy scanning of the dihedral angle around the acryloyl group. Figure 4 shows the optimized structures along with their calculated standard Gibbs free energy changes (ΔG°/kJ mol⁻¹, 298.15 K) along the PT route. The distributed charges on the OH protons of CafH₂(COOH) were obtained in an NBO analysis (Supplementary Materials, Table S2).

Comparing the charge distributions on the three OH protons of CafH₂(COOH) (0.502 on 3OH, 0.507 on 4OH, and 0.502 on 1OH of the acryloyl group), one observes that the 4OH proton was more reactive in the acid–base reaction of O₂^•− than the other OH protons. The ΔG°s of the initial PT also show the higher reaction feasibility of CafH(COOH)⁻ (−8.95 kJ mol⁻¹) than of CafH₂(COO)⁻ (5.81 kJ mol⁻¹). According to these calculation results, PT is initiated at the OH of the catechol moiety, although the 3OH and 4OH are indistinguishable along the dihedral rotation. The electrochemical results indicate the occurrence of PT from the three OHs of CafH₂(COOH), but only 3OH and 4OH of the CatH₂ moiety are involved in the O₂^•− scavenging reaction.

3.4. Change in HOMO-LUMO Energies during PCET between CafH₂(COOH) and O₂^•−

Figure 5 shows changes in HOMO-LUMO (the singly occupied molecular orbital, SOMO) energies (Hartree/a.u.) during the PCET between CafH₂(COOH) and O₂^•− in DMF, supplemented by a frontier molecular-orbital analysis. After the initial PT, the reactant species CafH₂(COOH), O₂^•−, CafH(COOH)⁻, (CafH₂(COO)⁻), and HO₂^• are involved in the solution. The SOMO energy of HO₂^• (−0.3142) is much lower than the HOMO energies of CafH₂(COOH) (−0.2278) and its anions (CafH(COOH)⁻, −0.1782; CafH₂(COO)⁻, −0.2127), indicating that the electron acceptor is HO₂^• rather than O₂^•−. Figure 1a demonstrates that HO₂^• was scavenged, and thus, the electron donor is CafH(COOH)⁻/CafH₂(COO)⁻ where the downhill energy relationship during the ET is shown (bold red lines in Figure 5). These HOMO-LUMO changes occur during the PT between CafH₂(COOH) and O₂^•− forming CafH(COOH)⁻/CafH₂(COO)⁻ and HO₂^•, implying that the initial reaction is PT. Next, the HOMO-LUMO relationship between the products is reversed after the subsequent ET (red dotted lines in Figure 5). However, since H₂O₂ formed after the second PT has a lower HOMO (−0.2754) than HO₂⁻ (−0.1648), the reverse ET cannot proceed, so the ET direction is dominantly determined by the subsequent PT. Judging from the HOMO-LUMO relationship for O₂^•− scavenging by CafH₂(COOH), the net PCET involves two PTs and one ET.

3.5. Change in Free Energies during PCET between CafH₂(COOH) and O₂^•−

Vibrational frequency calculations were performed to obtain changes in standard Gibbs free energy (ΔG°/kJ mol⁻¹, 298.15 K) along the PCET in DMF. Figure 6 shows the equilibrium schemes and ΔG° of the six diabatic electronic states in the PCET involving two PTs and one ET between CafH₂(COOH) and O₂^•−, and CafH₂(COO)⁻ formed after the deprotonation of acryloyl OH and O₂^•−. The main drivers of these sequential processes are the ΔG°s of the individual reactions, the acid–base interaction and the redox potentials of the components. Since ET1 (a, 406.8; b, 346.7) is strongly endergonic, PT1 (a, −8.9; b, 16.2) dominantly occurs forming CafH(COOH)⁻/CafH₂(COO)⁻ and HO₂^•. In the following pathway (the lower panels in Figure 6a,b), both PT3 (a, 302.9; b, 388.9) and ET2 (136.0, 6.5) are uphill endergonic, so the sequential PCET is unlikely to proceed. Efficient O₂^•− scavenging then requires a concerted PCET or HAT reaction involving one ET and one PT, corresponding to movement along the diagonal of the lower panels (red straight arrows). Notably, PT4 in Figure 6b is exergonic (−63.6), meaning that two PTs from the CatH₂ moiety must be coupled to one ET for successful O₂^•− scavenging. That is, PT from the acryloyl OH does not associate with ET. Then, another plausible pathway is a concerted ET and two PTs in one step after preforming the HB complexes between O₂^•− and the CatH₂ moiety of CafH₂(COOH)/CafH₂ (COO)⁻ without generating high energy intermediates (red curved arrows). We refer to this pathway as the concerted two-proton-coupled electron transfer (2PCET) mechanism [16,29].

For a comparative study, the ΔG°s of the PCET of the other compounds were also calculated (

Table 1. ΔG°s (kJ mol⁻¹, 298.15 K) of PCET between O₂^•− and HCAs (CafH₂(COOH), ferulic acid, isoferulic acid, Et-CafH₂, and p-coumaric acid) in DMF, calculated using DFT at the (U)B3LYP/PCM/6-311+G(3df,2p) level.

Compounds	PT1	PT2	PT3	PT4	ET1	ET2	ET3	Concerted 1	Total 2
CafH2(COOH)	−8.9	−279.7	302.9	−177.3	406.8	136.0	−344.2	−41.3	−50.2
Ferulic acid	6.2	−476.9	313.1	−79.6	532.2	49.0	−343.6	−30.5	−24.3
Isoferulic acid	4.5	−471.2	298.2	−77.1	517.0	41.2	−334.1	−35.9	−31.4
Et-CafH2	−5.4	−365.8	371.3	−85.9	402.9	42.5	−414.6	−43.3	−48.7
p-Coumaric acid	7.7	−353.7	301.9	−79.4	420.1	58.6	−322.8	−20.8	−13.1

¹ Concerted values are the summed ΔG°s of ET2 and PT4 (ET3 and PT3). ² Total values are the summed ΔG°s of two PTs and one ET.

). Thermodynamically, the total ΔG° of the net PCET obtained by summing the ΔG°s of the two PTs and one ET is the energetic driving force. However, in case the PCET occurs along a pathway involving an infeasible single PT/ET, the total ΔG° cannot embody the energetic driving force. Along the plausible pathways, the concerted PCET (ET2–PT4/PT3–ET3) after the initial PT1, and 2PCET, was exergonic for all compounds. Therefore, the total ΔG° embodies the exergonic driving force similar to the case of CafH₂(COOH) (concerted, −41.3; total, −50.2). These ΔG°s cannot explain the higher reactivities of CafH₂(COOH) and Et-CafH₂ than the other compounds toward electrogenerated O₂^•−. In the electrochemical results (Figure 1 and Figure 2), ferulic acid, isoferulic acid, and p-coumaric acid did not show O₂^•−/HO₂^• scavenging ability, indicating that the CatH₂ moiety of CafH₂(COOH) and Et-CafH₂ confers the superior kinetics of the O₂^•− scavenging on CV time scales.

3.6. Potential-Energy Surfaces of the PCET between CafH₂(COOH) and O₂^•−

To gain more insight into the PCET mechanism of O₂^•− scavenging by CafH₂(COOH) in DMF, we investigated the potential-energy surfaces with DFT and NBO analyses. The reaction is assumed to involve three elementary steps: (i) formation of the prereactive HB complex (PRC) from the free reactants (FRs), (ii) PCET reaction to the product complex (PC) via a TS, and (iii) dissociation of the PC yielding free products (FPs). First, we performed a geometry optimization of the stable HB complexes (PRC, intermediate complex, and PC) along the PCET reaction (Figure 7a). Then, optimized structures of the plausible PRC (CafH₂(COOH)–O₂^•−) and PC (Caf(COOH)^•−–H₂O₂) formed from the FRs and the FPs via two HBs (step i) were obtained. After optimization, the ΔG° was reduced by 39.8 kJ mol⁻¹ (the ΔG° of the PRC was set to zero in Figure 7a). Then, an energy profile (ΔG°, kJ mol⁻¹) along the IRC for the 2PCET, which forms the PC (step ii). The IRC shows that 2PCET occurs between the CatH₂ moiety of CafH₂(COOH) and O₂^•− in a one-kinetic process via a TS with low activation energy (E_a) at 53.8 kJ mol⁻¹ without generating any intermediates such as HO₂^•, HO₂⁻, CafH(COOH)⁻, and CafH(COOH)^•.

The dependence of OH bond distances (OH¹, black line; OH², red line) on the number of electrons in the π-orbital of the planar molecule of CafH₂(COOH) (blue circles) along the IRC are shown in Figure 7b. Spin density distributions localized on the radicals before and after the TS are also demonstrated, where the radical localized on O₂^•− in the initial PRC is transferred to Caf(COOH)^•− in the resulting PC. The changes in spin on the electron-donor side (CafH₂(COOH)) correlated well with the changes in the π-electrons of CafH₂(COOH). Furthermore, the changes in structures coupled with π-electron transfer occur simultaneously with sequential lengthening of the two OH bond distances (OH¹/OH²) of the CatH₂ moiety. First, one phenolic proton (H¹) is attracted by O₂^•−. At the TS, this attraction results in nearly complete deprotonation (OH¹) with one-half of the π-electrons transferred from CafH₂(COOH) to O₂^•−. The second PT (H²) accelerates the ET forward from the TS, eventually forming the PC.

For a comparative study, the potential-energy surfaces of the PCET with an IRC and TS between O₂^•− and the other substrates—CafH₂(COO)⁻, Et-CafH₂, and CatH₂—were also investigated (Supplementary Materials, Figure S2). The 2PCET mechanisms were similarly mediated by the CatH₂ moiety and generated no intermediates.

Table 2. ΔG° and E_a values (kJ mol⁻¹, 298.15 K) of the 2PCET between O₂^•− and the substrates (CafH₂(COOH), CafH₂(COO)⁻, Et-CafH₂, and CatH₂) in DMF, calculated using DFT at the (U)B3LYP/PCM/6-311+G(3df,2p) level.

Reactants 1	FR	TS (Ea)	PC	FP
CafH2(COOH) (+O2•−)	39.8	53.8	−19.9	−10.4
CafH2(COO)− (+O2•−)	30.1	50.0	−27.6	−10.8
Et-CafH2 (+O2•−)	38.7	53.2	−20.3	−10.0
CatH2 (+O2•−)	71.6	52.5	−20.9	45.3

¹ ΔG°s (kJ mol⁻¹) of PRC were set as a zero point.

lists the ΔG° of FR, PC, and FP, and E_a (the geometries of the TS are given in Tables S3–S5 of the Supplementary Materials). The E_as of CafH₂(COOH) (53.8 kJ mol⁻¹), CafH₂(COOH)⁻ (50.0 kJ mol⁻¹), Et-CafH₂ (53.2 kJ mol⁻¹), and CatH₂ (52.5 kJ mol⁻¹) were very similar and as low as the hydrogen-bonding energy. In Table 2, the relationship in which the difference in the E_a values is proportional to the difference in their enthalpy of reaction shown as the ΔG°s of the PC suggests that the 2PCET mechanism of the same framework (between CatH₂ moiety and O₂^•−) follows the Bell–Evans–Polanyi principle [31], justifying the conclusion that “CafH₂(COOH) scavenges O₂^•− through the 2PCET” and the methods adopted in this study. The ΔG°s of dissociating the PC to the FPs were −10.4 kJ mol⁻¹ for CafH₂(COOH), −10.8 kJ mol⁻¹ for CafH₂(COO)⁻, and −10.0 kJ mol⁻¹ for Et-CafH₂, 55–56 kJ mol⁻¹ lower than that for CatH₂ (45.3 kJ mol⁻¹). Thus, the superior O₂^•−-scavenging ability of CafH₂(COOH) is mainly attributable to the ΔG°s of dissociating the PC (step iii) that promote the net PCET reaction, rather than to the kinetics of 2PCET via the TS (step ii).

Collectively, DFT results clarified that scavenging of O₂^•− by CafH₂(COOH) in DMF is governed by 2PCET involving two PTs and one ET in the formed PRC via two HBs, which corresponds to moving along the red curved arrows in Figure 6a. Figure 8 shows the net mechanism of O₂^•− scavenging by CafH₂(COOH) in DMF. In the 2PCET mechanism, ET occurs between the π orbitals of oxygen orthogonal to the molecular frameworks of the donor and acceptor, and PT occurs between the σ orbitals of oxygen along the HBs [16]. The numbers of spin distributed on the oxygen species (O₂^•−, H₂O₂), the CatH₂ moiety, and the acryloyl group, along step iii, the dissociation of the PC (9.5 kJ mol⁻¹), are also given in Figure 8. Notably, the spins are distributed on the acryloyl group of the PC (0.079) and on the dissociated Caf(COOH)^•− (0.110) that expands the π-conjugated plane, demonstrating that the radical product (PC and Caf(COOH)^•−) is more stabilized by the acryloyl group than the orthoquinone radical generated from CatH₂. These results are in good correlation with the ΔG°s of dissociating PC to FP (Table 2), suggesting that the acryloyl group of CafH₂(COOH), CafH₂(COO)⁻, and Et-CafH₂, thermodynamically promotes step iii, the dissociation of PC. These results imply that the acryloyl group play a promoting role in the net 2PCET mechanism (Figure 8) between O₂^•− and the catechol moiety via the dissociation (step iii) for efficient O₂^•− scavenging abilities of CafH₂(COOH).

4. Conclusions

In the present study, we investigated the reactivities of HCAs involving CafH₂(COOH) toward the electrogenerated O₂^•− in DMF. In the CV and spectral measurements, it was confirmed that O₂^•− was successfully scavenged by CafH₂(COOH) and Et-CafH₂ through PCET mediated by the CatH₂ moiety over the time scale of the CV measurements. Then, DFT calculations clarified that the concerted 2PCET mechanism involving two PTs and one ET via the CatH₂ moiety (not involving PT from the acryloyl group) embodies the superior kinetics of O₂^•− scavenging by CafH₂(COOH) and Et-CafH₂. The net mechanism involves the initial formation of PRC followed by 2PCET and dissociation of the PC into FPs. The acryloyl group contributes to the thermodynamic stability of the product radical, promoting the dissociation of the PC. Efficient O₂^•− scavenging abilities of CafH₂(COOH) and Et-CafH₂ are derived from both the CatH₂ moiety and the acryloyl group.

Although the presented results are specific to chemical reactions in aprotic DMF solvent, the PCET theory is adaptable to biological processes in biotic structures such as lipid bilayers. Therefore, we hope that the findings will reveal the mechanistic actions of O₂^•− scavenging by HCAs and the health benefits of CafH₂(COOH), thus securing its pharmacological use as a phytoalexin.