Nature-Inspired Redox Active Organic Molecules: Design, Synthesis, and Characterization of Pyridine Derivatives

Abstract

In this article, we present experimental and theoretical studies of pyridine derivatives (pyDs) inspired by natural systems to investigate the electron transfer processes occurring in aqueous media and elaborate a theoretical model that adequately predicts the behavior of new derivatives. Our results might be relevant to scientific and technological applications, including energy storage, redox-active scaffolds for organic synthesis, photoredox catalysis, and new materials. The synthesis of eight pyDs is reported. To improve water solubility, six new compounds are hexafluorophosphate alkylammonium salts. The pyDs exhibit irreversible redox processes, with electron-donating substituents decreasing the cathodic peak potential while electron-withdrawing groups increase it; when both substituents are present, the latter effect prevails. A computational study was performed to investigate the electrochemical behavior of the synthesized compounds and design new electroactive pyDs. DFT calculations provided the predominant species’ redox potentials and acidity constants to elaborate Pourbaix diagrams for each compound. The synthesized molecules exhibit a two-electron-one-proton dismutation process in the water pH window. Beyond this range, stabilized radical species undergo one-electron exchange processes. We correlated experimental and calculated parameters, screening 22 additional derivatives to evaluate their electrochemical behavior, identifying potential candidates capable of performing a one-electron transfer process in the pH window of water, revealing new applications for pyDs.

1. Introduction

The applications and specific devices in daily life drive the search for materials displaying innovative properties [1]. The abundance and sustainability of organic molecules have played a crucial role in this endeavor. In addition, organic compounds can be “engineered” to modulate their chemical or physical properties. Among the different structural motifs investigated, heterocyclic compounds play a predominant role, particularly nitrogen-containing molecules, which are essential in many chemical reactions occurring in all organisms and have considerable pharmacological significance [2,3,4,5,6].

Pyridine derivatives have shown electrochemical properties with a variety of applications spanning from CO₂ reduction to diverse catalytic processes. For instance, in recent years, there has been a growing interest in studying molecular metabolism and redox biology to understand the mechanisms of cellular redox balance. In this context, nicotine adenine dinucleotides (NAD) have been called the cell’s “redox currency” [7]. Alkaloids, nitrogen-containing compounds, such as papaverine, strychnine, quinine, and nicotine, have been studied as inhibitors for metal corrosion [8,9]. The use of pyridine (py) and pyridinium ion (pyrH⁺) has been investigated to evaluate their catalytic role in the reduction of CO₂ to methanol or other high-energy chemical feedstocks [10,11,12,13]. Pyridines or 1,4-dihydropyridines (dhpys) substituted at 2,6-, 3,5-, and 1,4-positions, have been synthesized and widely used as hydride donors in organocatalysis, Lewis’s acid catalysis, hydrogen atom donors in radical reactions, and antioxidants [14,15,16]. The electrochemical behavior of N(1)-benzyl and N(1)-propyl-1,4-dihydronicotinamide was also investigated in aqueous solution at pH 7–13, showing a two-electron one-step oxidation process, whose half-wave potential and limiting current are pH-independent [17]. 4-Substituted dhpys have been considered analogs of NADH coenzymes, where the substituent attached to the N1 position of the dhpys derivatives considerably affects their electrochemical behavior [18]. In the search for efficient synthetic organic processes, photoredox catalysis is being widely investigated. Recently, the use of Hantzsch esters as visible-light photoredox catalysts for transition-metal-free coupling of aryl-halides and aryl-sulfinates [19], as well as their use as the emerging versatile class of reagents in photoredox-catalyzed organic synthesis [20], have been reported.

The introduction of pyridine groups into poly(p-phenylenebenzo bisoxazole) (PBO) improved the electronic and optical properties, increasing the oxidation potential from 1.65 to 1.82 V vs. SHE, compared to PBO. The different behavior was attributed to a strong protonation of the pyridine-containing polymer [21]. Similarly, the electronic and steric properties of novel molecular frameworks, such as aza-heteroarenes, can be adjusted by introducing substituents into the heteroaromatic ring. This enables the fine-tuning and expansion of their potential applications [22].

Studying a molecule’s electrochemical behavior for energy storage is an essential prelude to understanding its physicochemical properties. Thus, bipyridinium salts are currently under extensive investigation due to their potential application in redox flow batteries. The electrochemical properties of bipyridinium salts are affected by the number of quaternary nitrogen atoms, the position of the nitrogen atoms (2,2′, 3,3′, and 4,4′), and the presence of bridges between the pyridine rings [16,23]. Besides the above-mentioned investigations, the electrochemical behavior of other nitrogen containing heterocycles has been reported [24,25].

The quest for inexpensive electroactive organic molecules for use in different scientific areas is highly attractive. However, this endeavor requires synthesizing and screening the properties of many molecules. Fortunately, computational methods offer a reliable alternative to time-consuming experimental evaluations for obtaining a priori electrochemical knowledge of many redox-active molecules. For instance, the reduction potential of quinoxaline derivatives with a range of electron-donating and electron-withdrawing groups have been reported. Calculations indicate that adding electron-donating groups, particularly alkyl groups, can significantly lower the reduction potential, albeit with a concomitant decrease in oxidative stability [26]. Also, it has been demonstrated that Artificial Intelligence (AI) assists the design, synthesis, and measurement of functional organic molecules [27].

Several requirements should be considered to use these electroactive molecules for a technical application, such as water solubility, redox potential, and chemical and electrochemical stability. The solubility of organic molecules in water can be increased by functionalizing the moiety with the appropriate groups, particularly charged functionalities. The functional groups affecting solubility can be categorized into negatively charged groups (carboxyl, phosphate and sulfate), positively charged groups (pyridinium, imidazolium, and tetraalkylammonium), and non-charged groups (cyano, mercapto, hydroxyl, morpholino, and polyether). For negatively charged groups, the effect is highly dependent on the pH of the electrolyte [28].

The appealing use of organic molecules to develop new materials, particularly nitrogen-containing heterocyclic compounds, motivated our interest in studying the redox properties of pyridine derivatives (pyDs). Herein, we present experimental and theoretical studies of pyDs inspired by natural systems to investigate the electron transfer processes occurring in aqueous media. Additionally, a theoretical model that adequately predicts the behavior of new derivatives is disclosed. According to the electron transfer process, our findings might be relevant to diverse scientific and technological applications, such as those mentioned above.

Our investigation started with the chemical synthesis of eight pyridine derivatives (structures 1 to 7′ in Figure 1a), six of which are new compounds (1, 2, 3, 4, 5, and 6). For these compounds, electrochemical characterization in aqueous solution and a theoretical study of their Pourbaix diagrams (PDs) were carried out. Furthermore, a theoretical analysis of the PDs for selected pyridine derivatives was performed (Figure 1b, structures 7 to 28).

2. Materials and Methods

2.1. Synthesis

All experiments involving air- and/or moisture-sensitive compounds were conducted in an oven-dried round-bottom flask capped with a rubber septum under a positive pressure of nitrogen. Reaction temperatures refer to that of the bath. Concentration refers to removal of volatiles on a rotary evaporator below 35 °C. Analytical thin-layer chromatography (TLC) was performed using glass plates pre-coated with silica gel (0.25 mm, 60-Å pore size, 230–400 mesh, Merck, Rahway, NJ, USA) impregnated with a fluorescent indicator (254 nm). Materials on TLC plates were visualized under an ultraviolet lamp (254 nm) and/or by submersion of the plate in a solution of phosphomolybdic acid (5%) containing a trace of ceric sulfate in aqueous sulfuric acid (5% v/v) followed by charring on a hot plate. Column chromatography (CC) was performed with silica gel 60 (40–63 µm). All mixed solvent eluents are reported as v/v solutions. All reported compounds were analyzed by thin-layer chromatography (TLC): ¹H NMR, ¹³C NMR, IR, and HRMS.

All the reagents were purchased from Sigma-Aldrich (St. Louis, MO, USA) and were used as received unless other things are stated. Only anhydrous DMF was distilled on CaH₂. Acetone, methanol, ether, ethyl acetate, and dichloromethane were distilled prior to use. The solvents used for NMR spectra were purchased from Cambridge Isotope Laboratories, Inc., Tewksbury, MA, USA, and used as received.

Proton nuclear magnetic resonance (¹H NMR) spectra and carbon nuclear magnetic resonance (¹³C NMR) spectra were recorded on Agilent-Inova-300 (Santa Clara, CA, USA), Varian VNMRS-400 NMR spectrometers (Palo Alto, CA, USA), and Magritek-Spinsolve-80 (Malvern, PA, USA). Proton chemical shifts are expressed in parts per million (ppm, δ scale) and are referenced to tetramethylsilane (TMS: 0.0). Carbon chemical shifts are expressed in parts per million (ppm, δ scale) and are referenced to tetramethylsilane (TMS: 0.0). Data are represented as follows: chemical shift, multiplicity (s = singlet; d = doublet; t = triplet; q = quartet; m = multiplet; br = broad; app = apparent), and coupling constant (J) in Hertz (Hz), as required. IR spectra were recorded on a Perkin-Elmer Spectrum 400 FT-IR/FIR spectrometer with ATR (Springfield, IL, USA). Mass spectra assessments were carried out on a JEOL SMX-102a spectrometer (Peabody, MA, USA).

HE and HE-Ox: Diethyl 2,2′-(2,6-dimethyl-1,4-dihydropyridine-3,5-diyl)bis(2-oxoacetate), and diethyl 2,2′-(2,6-dimethylpyridine-3,5-diyl)bis(2-oxoacetate), respectively, were prepared following literature procedures [29,30].

pyD 1: A mixture of pyridine (3 mL) and TMAP-Br (100 mg, 0.38 mmol) was added to a round-bottom flask and the mixture was heated at 70 °C for 18 h. The precipitate that formed, corresponding to the bromide pyridinium salt, was filtered off under vacuum and washed with acetone; this compound has been reported by Gray et al. [31] The white solid was redissolved in methanol (5 mL), followed by the addition of ammonium hexafluorophosphate (160 mg, 0.89 mmol), and the solution was stirred at room temperature for 24 h. The precipitate was filtered under vacuum and washed with methanol to provide 143 mg (80% yield) of a white solid. ¹H NMR (500 MHz, DMSO-d₆): δ 9.1 (d, J = 5.5 Hz, 2H), 8.7 (tt, J = 7.9, 1.2 Hz, 1H), 8.2–8.2 (m, 2H), 4.6 (t, J = 7.5 Hz, 2H), 3.3 (s, 2H), 3.1 (s, 9H), 2.5–2.4 (m, 2H). ¹³C NMR (125 MHz, DMSO-d₆): δ 146.5, 145.5, 128.7, 62.3, 58.3, 53.0, 24.7. IR ν[cm⁻¹]: 3149.0(w) [ν(C-H, sp²)], 1642.7(m) [ν(C=C)], 1492.8(m) [δ(CH₃, CH₂)], 1187.8(m) [ν(C-N⁺)], 824.6(s) [ν(Asym P-F)], 556.0(s) [ν(Sym P-F)]. HRMS (DART⁺) (m/z) [M-PF₆]⁺ calculated for C₁₁H₂₀F₆N₂P: 325.1263, found: 325.1312. Spectral data are provided in Figures S1–S3.

pyD 2: Nicotinamide (300 mg, 2.46 mmol) and TMAP-Br (834 mg, 3.19 mmol) were dissolved in anhydrous DMF (6 mL) and the mixture was heated at 70 °C for 18 h. The precipitated dibromide salt was filtered off under vacuum and washed with acetone (the corresponding dichloride salt was reported by Craig et al. [32]). The white solid was redissolved in methanol (4 mL), followed by the addition of ammonium hexafluorophosphate (656.9 mg, 4 mmol), and the solution was stirred at room temperature for 24 h. The precipitate was filtered under vacuum and washed with methanol to afford 837 mg (85% yield) of a white solid. ¹H NMR (400 MHz, DMSO-d₆): δ 9.7 (s, 1H), 9.4 (d, J = 6.1 Hz, 1H), 9.0 (d, J = 8.2 Hz, 1H), 8.7 (s, 1H), 8.3 (dd, J = 8.0, 6.1 Hz, 1H), 8.2 (s, 1H), 4.8 (t, J = 7.5 Hz, 2H), 3.5–3.4 (m, 2H), 3.1 (s, 9H), 2.6 (q, J = 7.8 Hz, 2H). ¹³C NMR (100 MHz, D₂O): δ 165.7, 146.6, 144.7, 144.7, 134.3, 128.8, 62.4, 58.5, 53.2, 24.5. IR ν[cm⁻¹]: 3382.8(w) [ν(N-H)], 3147.3(m) [ν(C-H, sp²)], 1690.3(s) [ν(C=O)], 1615.9(m) [ν(C=C)], 1478.0(w) and 1400.3(m) [δ(CH₃, CH₂)], 818.0(s) [ν(Asym P-F)], 554.3(s) [ν(Sym P-F)]. HRMS (DART⁺) (m/z) [M-PF₆]⁺ calculated for C₁₂H₂₁F₆N₃OP: 368.1321, found: 368.1377. Spectral data are provided in Figures S4–S6.

pyD 3: 3-(3-(methoxycarbonyl)-1-pyridin-1-yl)-N,N,N-trimthyl-propan-1-amonium hexafluorophosphate. Methylnicotinate (500 mg, 3.65 mmol) and TMAP-Br (1143 mg, 4.38 mmol) were dissolved in anhydrous DMF (6 mL) and the mixture was heated at 70 °C for 18 h. The mixture was cooled at ambient temperature and diluted with ether (6 mL). The precipitate thus formed was filtered off under vacuum and rinsed with ether. The solid was redissolved in methanol (4 mL) and ammonium hexafluorophosphate (1.18 mg, 7.3 mmol) was added and the solution was stirred at room temperature for 16 h. The precipitate was filtered under vacuum and rinsed with methanol to afford 1.25 g (65% yield) of a white solid. ¹H NMR (400 MHz, DMSO-d₆): δ 9.70 (s, 1H), 9.42 (d, 1H, J = 8.14), 9.03 (d, 1H, J = 6 Hz), 8.36 (d, 1H, J = 6 Hz), 8.33 (d, 1H, J = 6 Hz), 4.8 (app t, 2H, J = 7 Hz), 3.98 (s, 3H), 3.45 (m, 2H), 3.01 (s, 9H), 2.5 (m, 2H). ¹³C NMR: (76 MHz, DMSO-d₆): δ162.8, 148.8, 146.9, 146.1, 130.4, 129.06, 62.2, 58.6, 54.1, 53.0, 52.8, 24.9. IR ν[cm¹]: 3321.6(w) [ν(N-H)], 3095.1(w) [ν(C-H, sp²)], 3031.0(w) [ν(C-H, sp³)], 1735.6(m) [ν(C=O)], 1479.2(w) and 1437.3(w) [δ(CH₃, CH₂)], 1080.0(s) [ν(C-N⁺)], 829.1(s) [ν(Asym P-F)], 556.3(m) [ν(Sym P-F)]. Spectral data are provided in Figures S7–S9.

pyD 4: (S)-Nicotine (0.19 mL, 1.23 mmol) and TMAP-Br (417.6 mg, 1.60 mmol) were dissolved in anhydrous DMF (5 mL) and the mixture was heated at 70 °C for 18 h. Then, the DMF was concentrated in vacuum and the crude was washed with ethyl acetate. The product was obtained as a slightly red oil. Then, the oil was redissolved in methanol (3.5 mL) and ammonium hexafluorophosphate (411 mg, 2.52 mmol) was added. The solution was stirred at room temperature for 24 h. The precipitate was filtered under vacuum and washed with methanol to provide 490 mg (72% yield) of a beige solid. ¹H NMR (500 MHz, DMSO-d₆): δ 9.0 (s, 1H), 9.0 (d, J = 6.0 Hz, 1H), 8.6–8.6 (m, 1H), 8.2–8.2 (m, 1H), 4.6 (t, J = 7.7 Hz, 2H), 3.4 (d, J = 8.1 Hz, 2H), 3.3 (d, J = 10.1 Hz, 1H), 3.2 (t, J = 8.6 Hz, 1H), 3.1 (s, 9H), 2.4–2.4 (m, 3H), 2.3–2.3 (m, 1H), 2.2 (s, 3H), 1.9–1.8 (m, 2H), 1.7–1.6 (m, 1H). ¹³C NMR (125 MHz, DMSO-d₆): δ 145.5, 145.2, 144.2, 144.1, 128.5, 67.2, 62.3, 58.3, 56.8, 53.0, 35.3, 24.9, 23.2, 17.1. IR ν[cm⁻¹]: 3105.8(w) [ν(C-H, sp²)], 2975.9(w) [ν(C-H, sp³)], 2810.4(w) [ν(Sym cyclic-CH₂)], 1637.8(w) [ν(C=C)], 1492.9 and 1480.5(m) [δ(CH₃, CH₂)], 824.9(s) [ν(Asym P-F)], 555.3(s) [ν(Sym P-F)]. HRMS (DART⁺) (m/z) [M-PF₆]⁺ calculated for C₁₆H₂₉F₆N₃P: 408.1998, found: 408.2026. Spectral data are provided in Figures S10–S12.

pyD 5: (S)-Cotinine (300 mg, 1.7 mmol) and TMAP-Br (578 mg, 2.21 mmol) were dissolved in anhydrous DMF (6.1 mL) and the mixture was heated at 70 °C for 18 h. Then, the DMF was concentrated in vacuum and the crude was washed with ethyl acetate. The product was obtained as a slightly yellow oil. The oil was redissolved in methanol (3.5 mL) and ammonium hexafluorophosphate (583 mg, 3.5 mmol) was added. The solution was stirred at room temperature for 24 h. The precipitate was filtered under vacuum and washed with methanol to provide 752 mg (78% yield) of a white solid. ¹H NMR (400 MHz, DMSO-d₆): δ 9.0 (d, J = 4.9 Hz, 2H), 8.6–8.5 (m, 1H), 8.2–8.2 (m, 1H), 4.8 (dd, J = 8.4, 5.6 Hz, 1H), 4.6 (dt, J = 7.1, 3.6 Hz, 2H), 3.3–3.3 (m, 2H), 3.0 (s, 9H), 2.6 (s, 3H), 2.5–2.3 (m, 5H), 1.8–1.7 (m, 1H). ¹³C NMR (100 MHz, DMSO-d₆): δ 175.2, 144.9, 144.7, 143.9, 143.2, 128.9, 62.3, 60.9, 58.4, 52.9, 29.7, 28.5, 27.5, 24.8. IR ν[cm⁻¹]: 3105.8(w) and 3068.6(w) [ν(C-H, sp²)], 2979.7(w) [ν(C-H, sp³)], 1682.3(s) [ν(C=O)], 1482.2() [δ(CH₃, CH₂)], 821.6(s) [ν(Asym P-F)], 555.2(s) [ν(Sym P-F)]. HRMS (DART⁺) (m/z) [M-PF₆]⁺ calculated for C₁₆H₂₇F₆N₃OP: 422.1790, found: 422.1850. Spectral data are provided in Figures S13–S15.

pyD 6′ was prepared according to the procedure reported by Blank et al. [33] The product was purified by column chromatography (CH₂Cl₂/MeOH 5%) to afford 41.4 mg (65% yield) of desired product as a white solid. All the spectroscopic data match the reported values. ¹H NMR (500 MHz, DMSO-d₆): δ 8.3 (s, 1H), 2.7 (s, 6H). ¹³C NMR (125 MHz, DMSO-d₆): δ 169.4, 158.4, 139.8, 128.1, 24.6. Spectral data are provided in Figures S16 and S17.

pyD 6: 2,6-Dimethylpyridine-3,5-dicarboxamide (200 mg, 1.03 mmol) and TMAP-Br (405 mg, 1.55 mmol) were dissolved in anhydrous DMF (2.6 mL) and the mixture was heated at 70 °C for 18 h. The precipitate formed was vacuum-filtered and washed with cold methanol. The white solid was redissolved in methanol/water (4 mL, 3:1 v/v) mixture and ammonium hexafluorophosphate (583 mg, 3.5 mmol) was added. The solution was stirred at room temperature for 24 h. The precipitate was filtered under vacuum and washed with methanol to provide a beige solid 330 mg (55% yield) of a beige solid. ¹H NMR (400 MHz, DMSO-d₆): δ 8.5 (s, 1H), 4.3 (t, J = 6.1 Hz, 2H), 3.4 (dd, J = 9.9, 6.4 Hz, 2H), 3.1 (s, 9H), 2.7 (s, 3H), 2.2–2.1 (m, 2H). ¹³C NMR (100 MHz, DMSO-d₆): δ 165.8, 161.9, 141.0, 123.1, 62.9, 52.8, 34.7, 25.2, 22.7. IR ν[cm⁻¹]: 3105.2(w) [ν(C-H, sp²)], 2977.1(w) [ν(C-H, sp³)], 1722.0(s) [ν(C=O)], 1283.1(m) [ν(C-N)], 1112.1(m) [ν(C-N⁺)], 827.2(s) [ν(Asym P-F)], 556.1(m) [ν(Sym P-F)]. Spectral data are provided in Figures S18–S20.

pyD 7′: N³,N⁵,2,6-tetramethylpyridine-3,5-dicarboxamide. The preparation of this compound is mentioned in the literature [34], however no details on its structural characterization are provided. A solution of diethyl 2,6-dimethylpyridine-3,5-dicarboxylate (200 mg, 0.79 mmol) in 2 mol L⁻¹ CH₃NH₂/MeOH (8 mL) was heated in a Schlenk storage tube at 70 °C overnight. The mixture was cooled to ambient temperature and concentrated in vacuum. The residue was taken up in ethyl acetate (20 mL), washed with water (2 × 5 mL), and sat. NaCl (5 mL) was added, and it was dried (Na₂SO₄). The solvent was removed, and the residue of the product was purified by column chromatography (silica gel, CH₂Cl₂/MeOH 5%) to afford 141 mg (35% yield) of a white solid. ¹H NMR (400 MHz, DMSO-d₆): δ 8.4 (s, 2H), 7.7 (s, 1H), 2.8 (d, J = 4.6 Hz, 6H), 2.5 (s, 6H). ¹³C NMR (100 MHz, DMSO-d₆) δ 168.0, 155.7, 134.5, 128.4, 26.0, 22.7. IR ν[cm⁻¹]: 3299.0(m) [ν(N-H)], 3077.3(w) [ν(C-H sp²)], 2960.8(w) [ν(C-H sp³)], 1618.9(s) [ν(C=O)], 1545.7(m) [ν(C=C)], 1296.5(m) [ν(C-N_amide)], 1156.3(m) [ν(C-N)]. HRMS (DART⁺) (m/z) [M + H]⁺ calculated for C₁₁H₁₆N₃O₂: 222.1243, found: 222.1250. Spectral data are provided in Figures S21–S23.

2.2. Electrochemistry

Cyclic voltamperometry tests were carried out with a potentiostat–galvanostat model PGSTAT 100N. All electrochemical experiments used a glassy carbon microelectrode (GCM) as the working electrode (WE), after polishing with diamond suspension (1 µm) and clear with deionized water. Platinum (Pt) wire was used as a counter electrode, and AgCl-Ag as a reference electrode (0.27 V vs. SHE). A scan rate of 0.1 V s⁻¹ was applied for cyclic voltamperograms (CVs). All CVs were referred to the SHE. The compounds (2 × 10⁻³ mol L⁻¹) were prepared in NaCl (1 mol L⁻¹, pH = 7). CVs were performed in positive and negative directions, starting from the zero current potential (E_i=0), after deoxygenation with N₂. Also, pyDs were studied in DMSO (99.7%, extra dry, over molecular sieves) using Bu₄NPF₆ (0.1 mol L⁻¹) as an electrolyte.

2.3. Computational Details

We performed redox potential and pK_a calculations using a workflow for the geometry optimization of species shown in Figure 1, with the following specifications. We generated the molecules by making substitutions with the different functional groups on the pyridine ring using SMILES strings. The first module in the workflow performs molecular mechanics (MM) calculations. The structures were converted from smiles to xyz coordinates using the Obabel 2.4.0. software [35], followed by a conformer search and an energy minimization with the MMFF94 force field [36,37,38,39,40]. The second module corresponds to the quantum mechanics (QM) module, which executes density functional theory (DFT) calculations through the Gaussian16 package [41].The QM module takes the MM-optimized coordinates as input for further geometry relaxation using the B3LYP functional [42], a 6-311+G(d,p) basis set [43,44], and an SMD solvation model in water [45]. A frequency analysis is specified to confirm a minimum and to compute the Gibbs free energy.

Redox Potential and pK_a Calculations

The estimation of redox potentials and pK_a values usually involves methodologies based on thermodynamic cycles [46,47,48]. In general, for a reduction equilibrium in solution (Equation (1)), the standard redox potential (E°_calc) is calculated through Equation (2), where ΔG_s^{0, redox} is the Gibbs free energy difference between the oxidized and reduced species (Equation (3)), n is the number of electrons, and F is the Faraday’s constant. An absolute value of 4.44 V for the standard hydrogen electrode (SHE) in water was considered as the reference potential value required for the calculations [49].

$$\mathrm{Ox}+e^{-} \leftrightarrows \mathrm{Red}; {\Delta G}_{\mathrm{s}}^{0,\mathrm{redox}},$$

(1)

$$E_{\mathrm{cal}}^0=-{\displaystyle \frac{{\Delta G}_{\mathrm{s}}^{0,\mathrm{redox}}}{\mathrm{nF}}}-\mathrm{SHE},$$

(2)

$${\Delta G}_{\mathrm{s}}^{0,\mathrm{redox}}=G_{\mathrm{s}}^0\left(\mathrm{Red}\right)-G_{\mathrm{s}}^0\left(\mathrm{Ox}\right),$$

(3)

The calculation of pK_a follows a similar procedure. For an acid-base equilibrium (Equation (4)), the standard Gibbs free energy change (ΔG_s^0,a) is related to the acid dissociation constant (K_a) in its negative logarithmic form (pK_a), as described by Equation (5), where R is the gas constant and T is the temperature. An approximation to calculate the ΔG_s^0,a requires the Gibbs free energy values for the species HA and A⁻ and the solvated proton in water (−269 kcal mol⁻¹) [50,51], as shown in Equation (6).

$$\mathrm{HA} \leftrightarrows {\mathrm{A}}^{-}+{\mathrm{H}}^{+} ; {\Delta G}_{\mathrm{s}}^{0,\mathrm{a}},$$

(4)

$${\mathrm{p}K}_{\mathrm{a}}={\displaystyle \frac{{\Delta G}_{\mathrm{s}}^{0,\mathrm{a}}}{2.303 \mathrm{RT}}},$$

(5)

$${\Delta G}_s^{0,\mathrm{a}}=G_s^0\left({\mathrm{A}}^{-}\right)+G_{\mathrm{s}}^0\left({\mathrm{H}}^{+}\right)-G_{\mathrm{s}}^0\left(\mathrm{HA}\right),$$

(6)

3. Results and Discussion

The pyridine ring is a π-deficient aromatic system due to the presence of an electronegative nitrogen atom. This characteristic is responsible for the typical chemical reactivity of this heterocyclic compound, i.e., pyridine reacts preferably with nucleophiles. Additionally, the lone pair of electrons on the nitrogen atom confers basic and nucleophilic characteristics to the molecule, forming quaternary ammonium salts readily. The possibility of modulating the electronic distribution on the heteroaromatic ring and its redox properties stimulated our interest in synthesizing nature-inspired pyridines to investigate their electrochemical behavior. For this purpose, we selected Hantzsch ester-derived pyridines and commercially available 3-substituted pyridines such as nicotine and its oxidized metabolite cotinine. Furthermore, computational studies were performed to study the Pourbaix diagram of the synthesized molecules and additional derivatives to find candidates which might have some application.

3.1. Chemical Synthesis

The structures shown in Figure 1 were proposed for chemical synthesis and structure relationship studies in silico. Compounds 1 to 7′ (Figure 1a) were synthesized and characterized to investigate their electrochemical behavior, while structures 7 to 28 (Figure 1b) are derivatives selected for computational studies.

PyDs 1 and 2 were obtained as the di-(hexafluorophosphate) salts; similar procedures have been reported to prepare the dibromide and dichloride salts, respectively [31,32]. PyD 6′ required for our synthetic and electrochemical studies was prepared according to literature procedures [33]. PyD 7′ was also synthesized and characterized; although the preparation of this compound is mentioned in the literature [34], no details on its structural characterization are provided.

The synthesis of the above-mentioned compounds was achieved either by N-alkylation of the corresponding commercially available pyridine ring (py, Scheme 1a), or by constructing the corresponding aromatic ring using literature procedures, followed by the previously described alkylation (Scheme 1b) [29,30].

For the alkylation reaction, a solution of the corresponding starting material of 1 to 5 in anhydrous DMF was treated with TMAP-Br to provide the corresponding pyridinium salts. The treatment of an alcoholic solution of the previously obtained bromide salts with an aqueous solution of NH₄PF₆ afforded the analogous hexafluorophosphate salts. All attempts to N-alkylate the oxidized Hantzsch ester (diethyl 2,6-dimethylpyridine-3,5-dicarboxylate) using TMAP-Br or methyl iodide were unsuccessful, as well as the N-alkylation of amide 7′. While this alkylation reaction might seem trivial, no reports in the literature describe this transformation, and the only reported method for its preparation involves the oxidation of diethyl 1,2,6-trimethyl-1,2,3,4-tetrahydropyridine-3,5-dicarboxylate [52]. Likewise, attempts to convert the oxidized Hantzsch ester into diethylamide derivative via an aminolysis reaction with a 70% aqueous solution of EtNH₂ failed. Two additional experiments were also performed, the first one included the hydrolysis of the ethyl ester, acid chloride formation, and subsequent reaction with a 70% solution of EtNH₂. In the second attempt, the carboxylic acid was reacted with a 70% aqueous solution of EtNH₂ in the presence of dicyclohexylcarbodiimide (DCC) and 1-hydroxybenzotriazole (HOBt). However, both experiments were unsuccessful and only trace amounts of desired product were observed in the ¹H NMR of the crude mixture, along with a mixture of several unidentified compounds.

Nevertheless, alkylation of amide 6′ (prepared by aminolysis of the oxidized Hantzsch ester, Scheme 1b) with TMAP-Br proceeded smoothly to provide a single product. Since all the attempts to obtain the diethylamide were unsuccessful, we prepared amide 7′ via aminolysis of the corresponding ethyl dicarboxylate (Scheme 2). However, treatment of 7′ with TMAP-Br to obtain derivative 7, following the general alkylation procedure, resulted in unreacted starting material. The above results show that changing the nature of the substituent on the heterocyclic moiety can dramatically affect the reactivity of the pyridine derivative.

3.2. Electrochemical Characterization

As demonstrated by the cyclic voltamperometry test, the reduction in pyD⁺ molecules involves the transfer of two electrons and one hydrogen ion at the C-4 position of the pyridine ring (Equation (7)) [53].

$${\mathrm{pyD}}^{+}+2e^{-}+{\mathrm{H}}^{+} \leftrightarrows \mathrm{HpyD},$$

(7)

The results obtained by cyclic voltamperometry show that the reduction processes of pyDs depend on the type of substituent and counterion [54]. Figure 2a shows the CV performed in a negative scan direction for pyD 1; the cathodic peak potential (E_cp) appears at −1.23 V vs. SHE, its corresponding oxidation signal at −0.03 V vs. SHE, and the peak-to-peak separation (ΔE_p) was 1.20 V. When the CV was recorded in a positive direction (Figure 2b), the absence of the oxidative signal at the start of the scan, from 0.10 to 1.00 V vs. SHE, confirms that the oxidative signal at −0.03 V vs. SHE is formed until the reduction in the compound takes place at −1.23 V vs. SHE.

To obtain more information about the electrochemical activity of pyD 1, inversion potential studies were performed (Figure 3). The results show that the intensity between the cathodic peak current (i_cp^red) and the anodic peak current (i_ap^ox) was disproportionate. According to this observation, pyD 1 presents an irreversible electrochemical profile.

Similar experiments were carried out for the other synthesized compounds under the same conditions as above. The studied derivatives displayed a low solubility (2.00 × 10⁻³ mol L⁻¹) at neutral medium.

Table 1. Electrochemical parameters, V vs. SHE, of pyridine derivatives 1 to 7′ in water.

pyDs	1	2	3	4	5	6′	6	7′
Ecp	−1.23	−1.13	−0.98	−1.27	−1.17	−1.18	−1.08	−1.26
Eap	−0.03	0.13	0.15	−0.02	0.08	0.65	0.58	0.68
ΔEp	1.20	1.26	1.13	1.25	1.25	1.83	1.66	1.94
E1/2	−0.63	−0.50	−0.42	−0.65	−0.55	−0.27	−0.25	−0.29

summarizes the experimental data, including half-wave potentials (E_1/2) and peak-to-peak separation (ΔE_p). The electrochemical evaluation of the synthesized compounds exhibits a similar behavior among them, showing a single reduction signal from −0.98 to −1.26 V vs. SHE under neutral conditions (Figure 3 and Figures S24–S30), which is characteristic of the pyridines and its derivatives [55,56,57,58]. The associated oxidative signals appear from −0.03 to 0.68 V vs. SHE.

In all cases, ΔE_p suggests an irreversible electrochemical process. The presence of electron donor/acceptor groups slightly modifies the reduction peak potential (E_cp) for pyD derivatives. The difference in E_cp would be related to the nature of the synthesized compound’s charge and the substituent’s energy density. The reduction potentials become more negative, depending on the nature of the functional group, observing the following trend: 3, 6, 2, 5, 6′, 1, 7′, and 4 (Table 1). Moreover, in aqueous solutions, the presence of protons facilitates the reduction in the pyDs. These results demonstrate that the substituent and the molecular structure influence the electrochemical properties.

Compound 1 (Figure 1a) was the simplest molecule studied without substituents and is used for comparison purposes. When the pyridine is mono-substituted at the C-3 position with an electron-withdrawing group, the reduction potential increases due to a lower electron density on the pyridine ring. For instance, E_cp V vs. SHE changes from −1.23 to −0.98 (3, R₃ = CO₂Me), to −1.13 (2, R₃ = CONH₂), and to −1.17 (5, R₃ = pyrrolidone). In contrast, if R₃ is an electron-donating group, such as a pyrrolidine in compound 4, the reduction reaction is slightly more challenging due to the increased electron density. Therefore, E_cp for compound 4 appears at a lower potential value, −1.27 V vs. SHE, compared to pyD 1.

When pyD is tetra-substituted with R₃ = R₅ = CONH₂ and R₂ = R₆ = CH₃ (pyD 6), the effect of the electron-withdrawing group dominates over the electron-donating inductive effect of the methyl groups. As a result, the potential shifts to more positive values, with an E_cp = −1.08 V vs. SHE, Figure S29. To evaluate the effect of the alkyl chain attached to the nitrogen atom of the pyridine ring on E_cp, we compared pyD 6′ (Figure S28) with pyD 6. In this comparison, pyD 6′ showed an E_cp = −1.18 V vs. SHE, whereas the potential for pyD 6 was E_cp = −1.08 V vs. SHE. The shift to more positive values is due to a positive charge on the pyridine nitrogen atom. The possibility of resonance structures (Scheme 3) delocalizes the positive charge on a larger surface, facilitating the reduction process. In the case of pyD 6′ and 7′, the E_cp value is more negative for pyD 7′ −1.26 V vs. SHE (Figure S30). Although both compounds contain a carbonyl moiety at C-3 and C-5, for pyD 7′ R₃ and R₅ are secondary amides; additionally, this compound has two methyl groups at C-2 and C-6 whose inductive effect might be responsible for lowering the reduction potential value.

Due to the amphoteric properties of water, several organic compounds often undergo acid/base reactions when they are oxidized or reduced at the electrode surface. In order to modulate these effects, electrochemical measurements were conducted in the aprotic polar solvent DMSO (99.7%, extra dry, stored over molecular sieves).

The voltamperograms obtained for the compounds in DMSO show a reduction process and its corresponding oxidation process in all cases (Figures S31–S37). The values for each compound are presented in

Table 2. Electrochemical parameters, V vs. SHE, of pyDs in DMSO.

pyDs	1	2	3	4	5	6′	6	7′
Ecp	−0.81	−0.74	−0.63	−0.96	−0.86	−1.71	−1.54	−2.14
Eap	0.41	0.44	0.46	0.20	0.31	0.27	0.48	−0.23
ΔEp	1.22	1.18	1.09	1.16	1.17	1.98	2.02	1.91
E1/2	−0.63	−0.50	−0.42	−0.65	−0.55	−0.27	−0.25	−0.29

. For compounds 1, 2, 3, 4, and 5, the reduction potential in DMSO has a ~0.4 V anodic shift with respect to those observed in NaCl (aq). These effects are related to the electron pair donation ability of the solvent playing an important role in the redox potential of an ionic compound, which can be estimated by the parameter known as the donor number, DN [59]. Therefore, less energy is required to carry out the reduction in ionic species in a solvent with high electron donor capacity, resulting in a shift toward anodic potential. For instance, in DMSO (DN = 124.7 kJ mol⁻¹), the cathodic process occurs before than the observed in aqueous solution (DN = 75.312 kJ mol⁻¹).

Additionally, important differences are observed in the E_cp values (Table 2) for the first five compounds with respect to pyDs 6′, 6, and 7′ in DMSO (Figures S31–S37). To understand the difference in redox behavior, a series of experiments consisting in the addition of known amounts of KOH/D₂O (0.04 mol L⁻¹) to a DMSO solution of pyDs 1, 2, and 7′, followed by acquisition of the corresponding ¹H NMR spectra. The results allowed us to conclude that a nucleophilic attack by the HO⁻ ion occurs at the free positions of the pyD ring (C-2, C-4, C-5, and C-6) in compounds 1 and 2, leading to the formation of a new electroactive species (see Figures S38 and S39). The hydroxylated products undergo an easier oxidation process; however, since compound 7′ has substituents at C-2, C-3, C-5, and C-6 positions, the nucleophilic attack does not occur. In the electrochemical process, the reduced species are more reactive toward the nucleophilic attack. As a consequence, a lower reduction potential can be observed for pyDs 1, 2, 3, 4, and 5 compared to 6, 6′, and 7′.

3.3. Computational Studies

3.3.1. Proton-Coupled Electron Transfer Reactions

For the pyDs, the main redox equilibria in solution were established through a pH predominance zone diagram (PZD), using the pKa values of the oxidized and reduced species involved in Equation (1). Scheme 4 shows the processes for the species considered, where we identify electron transfer reactions (E⁰₁, E⁰₂, and E⁰₃), proton-coupled electron transfer (PCET) reactions (E⁰_1A and E⁰_2B), and proton transfer reactions (pK_a^A and pK_a^B).

For these equilibria, the redox potential and pKa were calculated at a B3LYP/6-311+G(d,p) level of theory in water, following the methodology described in computational details. The results obtained for derivatives 1 to 7′ are presented in

Table 3. Calculated standard redox potentials, E⁰/V vs. SHE, and pK_a values for pyridine derivatives 1–7′ in water.

Derivative	${\mathbf{p}\mathit{K}}_{\mathbf{a}}^{\mathbf{A}}$	${\mathbf{p}\mathit{K}}_{\mathbf{a}}^{\mathbf{B}}$	${\mathit{E}}_1^0$	${\mathit{E}}_{1\mathbf{A}}^0$	${\mathit{E}}_2^0$	${\mathit{E}}_{2\mathbf{B}}^0$	${\mathit{E}}_3^0$
1	3.35	49.23	−1.42	−1.22	−2.57	0.35	0.15
2	0.12	47.02	−1.15	−1.14	−2.30	0.48	0.48
3	−0.53	46.24	−1.14	−1.17	−2.23	0.50	0.53
4	4.69	49.45	−1.47	−1.19	−2.58	0.35	0.07
5	3.73	45.69	−1.43	−1.21	−2.40	0.30	0.08
6′	−0.24	46.09	−1.25	−1.26	−2.31	0.42	0.43
6	−0.49	41.49	−1.23	−1.26	−2.00	0.46	0.49
7′	0.90	46.36	−1.30	−1.25	−2.36	0.38	0.33

, where we can note that the values are consistent through all the derivatives. Notably, the pK_a^B has an extremely high value, exceeding the pH water scale, which can be attributed to the high basicity of the pyD⁻.

Based on the pK_a^A and pK_a^B values, we can plot the PZD of the pyDs (see example in Figure S40 for pyD 1) and establish the predominant equilibria (Equations (8)–(12)), where we identified the following general behavior. At pH < pK_a^A, the two-electron transfer process is given by Equations (8) and (9), where the redox potential of the first reduction (E⁰_1A) is lower than the second reduction (E⁰₃), leading to a dismutation process of the HpyD^•+ radical species. A similar behavior is observed for the region where pK_a^A < pH < pK_a^B (Equations (10) and (11)) since the redox potential E⁰₁ is lower than E⁰_2B, now producing the dismutation of the pyD^• species. Finally, the region where pH > pK_a^B has a redox potential E⁰₁, higher than E⁰₂ (the equilibria involve Equations (10) and (12)), indicating that a two-step electron transfer occurs, resulting in the stabilization of the pyD^• radical species.

$${\mathrm{pyD}}^{+}+{\mathrm{H}}^{+}+e^{-} \leftrightarrows {\mathrm{HpyD}}^{•+} ; E_{1\mathrm{A}}^0,$$

(8)

$${\mathrm{HpyD}}^{•+}+e^{-} \leftrightarrows \mathrm{HpyD} ; E_3^0,$$

(9)

$${\mathrm{pyD}}^{+}+e^{-} \leftrightarrows {\mathrm{pyD}}^{•} ; E_1^0,$$

(10)

$${\mathrm{pyD}}^{•}+{\mathrm{H}}^{+}+e^{-} \leftrightarrows \mathrm{HpyD} ; E_{2\mathrm{B}}^0,$$

(11)

$${\mathrm{pyD}}^{•}+e^{-} \leftrightarrows {\mathrm{pyD}}^{-} ; E_2^0,$$

(12)

The previous analysis allowed us to establish the general behavior on the pyDs based on the standard redox potential. A further calculation of the pH-dependent redox potential (conditional potential) to construct the Pourbaix diagram could confirm this trend. In the following section, we present the theoretical study of the Pourbaix diagrams for the synthesized (1–7′) and proposed (7–28) pyridine derivatives shown in Figure 1. Although the PDs are applicable only within the stability limits of water (considering pH and redox potential), an extended theoretical PD can help to identify how to direct the processes occurring on a molecule towards a desired region.

3.3.2. Pourbaix Diagrams

To compute the Pourbaix diagram (PD), the conditional potential (E′) was calculated for each pH zone through the Nernst equation (Equation (13)), where E⁰ is the standard reduction potential, Q is the reaction quotient, and n is the number of electrons transferred. The pH dependence is obtained by considering equal concentrations of oxidized and reduced species.

$$E^{\prime }=E^0-{\displaystyle \frac{\mathrm{R}\mathrm{T}}{\mathrm{n}\mathrm{F}}}\mathrm{l}\mathrm{n}Q=E^0-0.059 \mathrm{p}\mathrm{H},$$

(13)

Four main zones were established from the Pourbaix diagram for pyD 1, shown in Figure 4a. The first one, at pH < 3.35, is characterized by a dismutation of the radical species HpyD^•+ (obtained by subtracting Equation (8) from Equation (9)). The dismutation occurs since the conditional potential of the first reduction (dotted green line) is lower than the second (dotted blue line). Consequently, a single-proton two-electron transfer reaction takes place (red line). A similar reaction takes place in the second zone, at 3.35 < pH < 29.91, where the pyD^• species undergoes a dismutation process (subtracting Equation (10) from Equation (11)). The third zone is defined at 29.91 < pH < 49.23, where a crossing between the first and second reduction lines is observed (pHcross = 29.91) and the reduction process occurs in two steps: a single-electron transfer (Equation (10)) and a proton-coupled electron transfer reaction (Equation (11)). This zone is characterized by stabilizing the radical species (pyD^•). The last zone, at pH > 49.23, is pH-independent and a two-step electron transfer takes place (Equations (10) and (12)).

The PDs for pyridine derivatives 2 to 7′ were calculated following the same strategy used for derivative 1. Figure 4b illustrates the global process for these derivatives, compared to pyD 1 (black line), showing similar behavior among them. The crossing of pH from the dismutation zone to the radical stability window is in the range of 27−31 pH units. Below this value, the reaction is a single-proton two-electron transfer, which spans the water pH window (0–14) and close to the redox potential lower boundary of the water stability window (gray shaded area). Beyond pHcross occurs the stabilization of the radical species, producing a two-step electron transfer process.

We correlate the experimental and calculated parameters to select the parameters that allow estimating peak potential values from a theoretical parameter or from the peak potentials to predict the undergoing electrochemical transfer processes associated with a molecule. Among the parameters evaluated are the cathodic and anodic peak potentials in water and DMSO (E_cp and E_ap) and different calculated parameters from the PDs, such as the conditional potential at pH = 0 (E′_pH=0), the crossing pH between the dismutation zone and the radical stability window (pHcross), and the pKa^A and pKa^B.

In general, the E_cp in water (Figure 5) shows higher correlation with the computed parameters than in DMSO (Figure S41). However, the E_ap in water and DMSO have low correlation (Figure S42); this poor correlation could be related to the origin of the oxidation signal and the species generated at the surface electrode. Notably, the peak values for compounds 1, 2, 3, 4, and 5 are different from those values for compounds 6, 6′, and 7′, because the former values exhibited coupled chemical hydroxylation reactions via a nucleophilic attack. Experimentally, the peak-to-peak separation (ΔE_p) is an indicator of the system reversibility. The electrochemically characterized compounds (1–7′) have a high ΔE_p value that can be considered irreversible (see above discussion of Figure 3). For a reversible system, a low peak-to-peak separation is desired (associated with the oxidation and reduction processes). Since only a good correlation was observed for the E_cp, it would be hard to predict reversibility for these systems from the calculated parameters. However, we can identify other relevant parameters such as the dismutation and radical stability zones.

From Figure 5, a direct correlation is observed between E_cp and E′_pH=0 (Figure 5a); that is, higher peak potentials are obtained at higher E′_pH=0. However, the trend is inverse for pHcross (Figure 5b) or a pKa value (Figure 5c,d), which results in lower peak potentials at higher calculated values. The correlation between E_cp and pHcross has the higher R² = 0.788, when considering only the N-alkylated pyridine derivatives (1–6). We noted that pHcross is a parameter that depends on the calculated standard redox potential and the pKa^A and pKa^B values. Modifying any of these parameters would result in a different conditional potential and pHcross. Therefore, we can establish that the correlation for E_cp is low when a single parameter is evaluated (E′_pH=0, pKa^A, or pKa^B) but increases by using pHcross since it depends intrinsically on several parameters.

The correlation plot of E_cp and pHcross in Figure 5b also shows the influence of the electron-withdrawing or electron-donating functional groups on the redox potential and pH values compared to pyD 1. The electron-withdrawing groups at C-3 position of the pyridine ring, such as pyD 2 (R₃ = CONH₂), 3 (R₃ = COOMe), and 5 (R₃ = pyrrolidone) shift the potential toward higher values, while the pHcross toward lower values. The opposite effect is observed for an electron-donating group such as pyD 4 (R₃ = pyrrolidine), displacing the E_cp to lower values and pHcross to higher values. Compound 6 has both types of functional groups, where the effect of the electron-withdrawing substituent at C-3 and C-5 positions (R₃ = R₅ = CONH₂) prevails over the effect of the electron-donating group at C-2 and C-6 positions (R₂ = R₆ = CH₃), resulting in an increase in the redox potential and a decrease in the pHcross. Nevertheless, this effect is not stronger than the observed for compounds 2 and 3. Finally, for derivatives 6′ and 7′, the main structural difference is the lack of the alkyl chain on the pyridine nitrogen atom, which results in a lower redox potential (as can be seen between compounds 6 and 6′), making the reduction process more difficult to conduct, as previously established.

From the previous analysis, we can establish the following remarks: in the PD, pHcross allows the identification of the radical stability window. The shifting of this parameter toward the water pH window would lead to the stabilization of the radical species in solution. Furthermore, from the correlation between E_cp and pHcross, the reduction potential peak can be used to estimate the pH for the radical stabilization and vice versa. Electron-withdrawing groups on the pyridine ring shift the peak potential and pHcross toward the potential and pH water stability window, favoring the stabilization of the radical species.

Given the trends established, different pyridine derivatives were proposed (compounds 7 to 28 in Figure 1b), and their PDs calculated (Figure S43) to modify the electron transfer processes occurring toward the water stability window. Most derivatives exhibit a similar behavior compared to derivative 1, where the radical stability window is at high pH values (pHcross > 25). However, compounds with electron-withdrawing groups such as 25 (R₂ = R₆ = CF₃ and R₃ = R₅ = COOEt), 26 (R₂ = R₆ = CF₃ and R₃ = R₅ = CONH₂), and 27 (R₂ = R₆ = CF₃ and R₃ = R₅ = CONHMe) have low pHcross values of 4.77, 4.54, and 7.01, respectively. These derivatives also show a shift toward positive redox potentials due to a higher electron-withdrawing effect on the pyridine ring, which makes them potential candidates for a single-electron transfer process in the water stability window.

The general trend is represented in Figure 6 for pyDs 2, 6, and 26 compared to 1. When the electron-withdrawing effect is weak on the pyridine ring such as in 2 or 6, the process observed in the water stability window corresponds to a 1H⁺, 2e⁻ transfer process. This process may change by attaching strong electron-withdrawing groups as CF₃ in pyD 26, which increases the redox potential and shifts the pHcross toward the water stability window, leading to the stabilization of the radical species and a two-step one-electron transfer process. According to the linear fitting performed previously, compound 26 would have an E_cp of about 0.50 V vs. SHE.

4. Conclusions

Eight pyridine derivatives with various substituents attached to the ring were synthesized. To increase the water solubility, six of these compounds were alkylated at the pyridine ring using TMAP-Br, followed by an ion exchange with ammonium hexafluorophosphate (NH₄PF₆), obtaining new compounds. All efforts to N-alkylate diethyl 2,6-dimethylpyridine-3,5-dicarboxylate (HE-Ox), and 7′ with TMAP-Br or methyl iodide were unsuccessful. These observations, combined with the scarcity of experimental procedures for N-alkylation of 2,6-dimethyl-3,5-dicarbonyl pyridine derivatives, underscore the importance of the nature of substituents attached to the pyridine ring on their physicochemical properties.

Electrochemical studies of synthesized pyDs indicate an irreversible reduction along with its corresponding oxidation processes. The substituents on the pyridine ring influence the cathodic and anodic peak potentials. In water solution, electron-donating groups at C-3 position shift the reduction potential toward more negative values (E_cp = −1.27 V vs. SHE, compound 4), whereas electron-withdrawing groups displace it to less negative values (E_cp = −0.98 V vs. SHE, compound 3). When both groups coexist in the same structure (pyDs 6′, 6, and 7′), the effect of the electron-withdrawing groups predominates, resulting in a less negative redox potential. On the other hand, the alkyl chain attached to N-1 of the pyridine ring facilitates the reduction process due to a delocalization of the positive charge by resonance.

The stability of the reduced species was determined using the E_ap value and the difference in peak potentials (ΔE_p). This difference increases with the number of substituents on the pyridine ring. Among the derivatives studied, the smallest value was obtained for compound 3 (ΔE_p = 1.13 V vs. SHE). Thus, the reversibility of the redox process could be improved by selecting suitable functional groups and substitution positions, leaving the possibility for exploring derivatives that have small ΔE_p values, closer to the expected value of 30 mV, for a use in a two-electron or 60 mV for a one-electron exchange process.

The standard reduction potentials and acidity constants were calculated for the synthesized derivatives (1–7′). The predominant equilibria were identified from these values, and the theoretical Pourbaix diagram was constructed for each derivative. We established different correlations between experimental (E_cp and E_ap) and theoretical parameters of the PD (E′_pH=0, pH_cross, pK_a^A, and pK_a^B), where a high correlation between E_cp and pH_cross was obtained (R² = 0.788), noting that the latter is a parameter that depends on the calculated standard redox potentials and pK_a values. We confirmed the increase of the redox potential caused by the electron-withdrawing groups on the pyridine ring. From this analysis, derivatives with different substitutions were proposed to obtain their PD according to the methodology established. We found three derivatives that modify the processes observed in the pH water stability window. In particular, pyDs 25 (R₂ = R₆ = CF₃ and R₃ = R₅ = COOEt), 26 (R₂ = R₆ = CF₃ and R₃ = R₅ = CONH₂), and 27 (R₂ = R₆ = CF₃ and R₃ = R₅ = CONHMe) have pH_cross values of 4.77, 4.54, and 7.01, respectively. Therefore, these compounds would exhibit both two- and one-electron transfer processes in the water stability window.