Oxidations of Benzhydrazide and Phenylacetic Hydrazide by Hexachloroiridate(IV): Reaction Mechanism and Structure–Reactivity Relationship

Benz(o)hydrazide (BH) is the basic aryl hydrazide; aryl hydrazides have been pursued in the course of drug discovery. Oxidations of BH and phenylacetic hydrazide (PAH) by hexachloroiridate(IV) ([IrCl6]2−) were investigated by use of stopped-flow spectral, rapid spectral scan, RP-HPLC and NMR spectroscopic techniques. The oxidation reactions followed well-defined second-order kinetics and the observed second-order rate constant k′ versus pH profiles were established over a wide pH range. Product analysis revealed that BH and PAH were cleanly oxidized to benzoic acid and phenylacetic acid, respectively. A reaction mechanism was proposed, resembling those suggested previously for the oxidations of isoniazid (INH) and nicotinic hydrazide (NH) by [IrCl6]2−. Rate constants of the rate-determining steps were evaluated, confirming a huge reactivity span of the protolysis species observed previously. The enolate species of BH is extremely reactive towards reduction of [IrCl6]2−. The determined middle-ranged negative values of activation entropies together with rapid scan spectra manifest that an outer-sphere electron transfer is probably taking place in the rate-determining steps. The reactivity of neutral species of hydrazides is clearly not correlated to the corresponding pKa values of the hydrazides. On the other hand, a linear correlation, logkenolate = (0.16 ± 0.07)pKenol + (6.1 ± 0.8), is found for the aryl hydrazides studied so far. The big intercept and the small slope of this correlation may pave a way for a rational design of new antioxidants based on aryl hydrazides. The present work also provides the pKa values for BH and PAH at 25.0 °C and 1.0 M ionic strength which were not reported before.


Introduction
Hydrazides are widely employed for manufactures of polymers and glues and utilized as chemical preservers for plants in industry [1,2]. They are potent reagents for synthesis of various oxygen-, nitrogen-, and/or sulfur-containing heterocyclic rings in organic chemistry [1]. Hydrazide scaffold based clinical medicines involve isoniazid (INH), marplan, iproniazid, and indolylglyoxyl hydrazide [2] and INH has been a frontline anti-tubercular drug for a few decades [3][4][5]. Ascribed to the structural diversity and to the huge success of INH, hydrazides and their derivatives have been a base for new drug discoveries [2,[6][7][8][9][10].
Mechanistically, the anti-tubercular action of INH involves its activation by enzymes generating hydrazyl free radical(s) in the activation course [11][12][13][14]. The involvement of free radicals may not be surprising since numerous oxidation reactions are taking place in biologically and/or biomedically relevant processes which involve a single electron transfer. [IrCl 6 ] 2− is a well-known single electron oxidizing agent [15][16][17][18][19] and it has been utilized as a redox probe for acquiring chemical information of

Empirical Rate Law and Kinetic Data Collection
To find the reaction order in [BH] tot /[PAH] tot (the subscript tot represents the total concentrations), the effects of varying [BH] tot /[PAH] tot on the oxidation rates were investigated in each of an extended series of reaction media. However, the variation of [BH] tot /[PAH] tot in each medium was controlled to not induce any pH changes in that particular medium. Plots of k obsd versus [Hydrazide] tot are illustrated in Figure 2 in the case of BH and in Figure 3 for the reaction of PAH. No doubt, these plots are linear and passing through the origin, indicating that the oxidation reactions are also first order in [Hydrazide] tot . Hence an empirical rate law (expressed by Equation (2)) is established, where k represents the observed second-order rate constant and k obsd = k [Hydrazide] tot . −d[IrCl 6 2− ]/dt = k obsd [IrCl 6 2− ] = k [Hydrazide] tot [IrCl 6 2− ] The oxidation reaction of BH was investigated in a region of 0.11 ≤ pH ≤ 10.46; when pH > 10.5, the reaction became too fast to follow even by the stopped-flow technique. In the case of PAH, the reaction was investigated in a wider pH region (0.16 ≤ pH ≤ 11.78) since the oxidation reaction of PAH was slower than that of BH. Values of k were computed from the linear plots of k obsd versus [Hydrazide] tot at various pHs which were collected from a large amount of data, and are summarized in Table S1 in SM. More visually, the plots of logk versus pH are given in Figure 4 (data points).

of 15
The oxidation reaction of BH was investigated in a region of 0.11 ≤ pH ≤ 10.46; when pH > 10.5, the reaction became too fast to follow even by the stopped-flow technique. In the case of PAH, the reaction was investigated in a wider pH region (0.16 ≤ pH ≤ 11.78) since the oxidation reaction of PAH was slower than that of BH. Values of k′ were computed from the linear plots of kobsd versus [Hydrazide]tot at various pHs which were collected from a large amount of data, and are summarized in Table S1 in SM. More visually, the plots of logk′ versus pH are given in Figure 4 (data points).       Table S1 in Supplementary Materials). The solid curves were generated from the best fits of Equation (5) to the experimental data by a weighted nonlinear least-squares simulation.

Evaluation of the Reaction Stoichiometry
Spectrophotometric titration was proved to be a good method for determinations of reaction stoichiometries [14,28,29]; it was thus used in the present reaction systems (cf. the experimental section below). From the spectrophotometric titration data, plots of the measured absorbance at 488 nm as a function of [BH] [14,28]. The stoichiometric reaction in the present cases can be described by Equation (3), where n = 0 for BH and n = 1 for PAH [14].

Product Analysis
To confirm the oxidation products as inferred by Equation (3), RP-HPLC was employed for the analysis of the reaction of BH; Figure 5 displays the chromatogram acquired for a reaction mixture of 2.5 mM BH and 2.5 mM [IrCl 6 ] 2− in a phosphate buffer of pH 6.34 after a reaction time of about 10 min. In the figure, the peaks were assigned according to the retention times which were identical to those from authentic samples. Moreover, no late peaks in the chromatogram were eluted by an extended elution time. Hence, benzoic acid was identified as the oxidation product of BH.
the ratios in Table 1 point to a clean stoichiometry of ∆[IrCl6 2− ]:∆[Hydrazide]tot = 4:1. The same reaction stoichiometry was also derived previously for the oxidations of INH and hydrazines [14,28]. The stoichiometric reaction in the present cases can be described by Equation (3), where n = 0 for BH and n = 1 for PAH [14].

Product Analysis
To confirm the oxidation products as inferred by Equation (3), RP-HPLC was employed for the analysis of the reaction of BH; Figure 5 displays the chromatogram acquired for a reaction mixture of 2.5 mM BH and 2.5 mM [IrCl6] 2− in a phosphate buffer of pH 6.34 after a reaction time of about 10 min. In the figure, the peaks were assigned according to the retention times which were identical to those from authentic samples. Moreover, no late peaks in the chromatogram were eluted by an extended elution time. Hence, benzoic acid was identified as the oxidation product of BH. For the oxidation of PAH by Ir(IV), the 1 H-NMR spectra acquired are shown in Figure 6, together with the assignments of NMR signals; 3-(trimethylsilyl)propionic acid-d 4 sodium salt (TSP) was utilized as the reference of the NMR shifts in the spectra. The spectra indicate that when more than a stoichiometric amount of Ir(IV) was used in the reaction mixture, all the reactant PAH was cleanly oxidized to phenylacetic acid. It was observed that the excess of Ir(IV) could not oxidize phenylacetic acid that was produced from the oxidation reaction; this was not surprising since Ir(IV) did not oxidize the HAc-NaAc buffers. Phenylacetic acid was thus confirmed as the oxidation product of PAH, justifying Equation (3). stoichiometric amount of Ir(IV) was used in the reaction mixture, all the reactant PAH was cleanly oxidized to phenylacetic acid. It was observed that the excess of Ir(IV) could not oxidize phenylacetic acid that was produced from the oxidation reaction; this was not surprising since Ir(IV) did not oxidize the HAc-NaAc buffers. Phenylacetic acid was thus confirmed as the oxidation product of PAH, justifying Equation (3).

Mechanistic Analysis
For BH and PAH in aqueous solution in the present work, three protolysis species (I-III shown in Figure 7) are involved across the wide pH range used in present work [14]. The elucidated kinetic characters for the present reaction systems (such as well-defined second-order kinetics, rapid scan spectra, the reaction stoichiometry and the oxidation products) echo those revealed in the INH-Ir(IV) reaction system [14]. Moreover, even the shape of logk′ versus pH profiles in Figure 4 is also similar to that obtained for the INH-Ir(IV) reaction. By analog, a reaction mechanism portrayed in Figure 7 is suggested for the present reaction systems in which the reactions denoted by k1-k3 are the ratedetermining steps. Two types of hydrazyl free radicals (species IV and V) were inferred to be generated in the rate-determining steps [11][12][13][14]26,27], and were followed by three consecutive and fast reactions, leading to formation of benzoic acid/phenylacetic acid [14].

Mechanistic Analysis
For BH and PAH in aqueous solution in the present work, three protolysis species (I-III shown in Figure 7) are involved across the wide pH range used in present work [14]. The elucidated kinetic characters for the present reaction systems (such as well-defined second-order kinetics, rapid scan spectra, the reaction stoichiometry and the oxidation products) echo those revealed in the INH-Ir(IV) reaction system [14]. Moreover, even the shape of logk versus pH profiles in Figure 4 is also similar to that obtained for the INH-Ir(IV) reaction. By analog, a reaction mechanism portrayed in Figure 7 is suggested for the present reaction systems in which the reactions denoted by k 1 -k 3 are the rate-determining steps. Two types of hydrazyl free radicals (species IV and V) were inferred to be generated in the rate-determining steps [11][12][13][14]26,27], and were followed by three consecutive and fast reactions, leading to formation of benzoic acid/phenylacetic acid [14].
Rate expression in Equation (4) was attained according to the reaction mechanism in Figure 7, where a H represents the proton activity which corresponds exactly to the pH measurements.
Equation (4) conforms to the empirical Equation (2), rendering:  Rate expression in Equation (4) was attained according to the reaction mechanism in Figure 7, where aH represents the proton activity which corresponds exactly to the pH measurements.

pKa Values and Rate Constants of the Rate-Determining Steps
Determination of pKa values for hydrazines from the well-defined kinetic data offered a good approach for the Ir(IV)-hydrazine reaction systems [28]; this was based on the measured kinetic data in the pH ranges covering the pKa values of hydrazines. For the present reaction systems, the protolysis constants Ka1 and Ka2 of BH and PAH at 25.0 °C and μ = 1.0 M have not been reported in the literature. The pH range studied for the PAH reaction was from 0.16 pH 11.78, probably covering both pKa1 and pKa2 of PAH, and consequently enabling us to derive these pKa values from our kinetic data. Equation (5) was then utilized to simulate the k′-pH dependence data by use of a weighted nonlinear least-squares method; in the simulation, k1, k2, k3, Ka1 and Ka2 were all treated as tunable parameters. The simulation provided with a good fit shown in the bottom part of Figure 4, conferring simultaneously the values for these parameters (listed in Table 2). The value of pKa2 = 11.7 ± 0.2 obtained from the simulation is indeed within the pH region studied kinetically. Table 2. Rate constants for the rate-determining steps and protolysis constants in Figure 7

pK a Values and Rate Constants of the Rate-Determining Steps
Determination of pK a values for hydrazines from the well-defined kinetic data offered a good approach for the Ir(IV)-hydrazine reaction systems [28]; this was based on the measured kinetic data in the pH ranges covering the pK a values of hydrazines. For the present reaction systems, the protolysis constants K a1 and K a2 of BH and PAH at 25.0 • C and µ = 1.0 M have not been reported in the literature. The pH range studied for the PAH reaction was from 0.16 pH 11.78, probably covering both pK a1 and pK a2 of PAH, and consequently enabling us to derive these pK a values from our kinetic data. Equation (5) was then utilized to simulate the k -pH dependence data by use of a weighted nonlinear least-squares method; in the simulation, k 1 , k 2 , k 3 , K a1 and K a2 were all treated as tunable parameters. The simulation provided with a good fit shown in the bottom part of Figure 4, conferring simultaneously the values for these parameters (listed in Table 2). The value of pK a2 = 11.7 ± 0.2 obtained from the simulation is indeed within the pH region studied kinetically. Table 2. Rate constants for the rate-determining steps and protolysis constants in Figure 7 for the reactions of BH and PAH determined at 25.0 • C and ionic strength µ = 1.0 M.

Hydrazide
k m /pK am Values 3.37 ± 0.09 pK a2 12.6 ± 0.1 3.24 ± 0.08 pK a2 11.7 ± 0.2 The pK a1 value of BH is expected to be between 3 and 4 [30], which is in the pH region studied kinetically in this work while that of pK a2 is anticipated to be >12 at 25.0 • C and µ = 1.0 M [30], being beyond the pH region of the kinetic data collection. We thus determined the pK a2 value of BH spectrophotometrically [31]. The UV-vis spectra recorded for 0.10 mM BH in the buffer solutions of pH 6.00 and 12.68 are given in the top part of Figure 8, where the spectra originate predominantly Equation (6) was then employed to simulate the data [31], using a nonlinear squares method, whereε 2 andε 3 represent the molar absorptivities of species II and III, respectively. The simulation resulted in a good fit, generating the values ofε 2 = (1.00 ± 0.02) × 10 3 M −1 cm −1 , ε 3 = (7.3 ± 0.4) × 10 3 M −1 cm −1 , and pK a2 = 12.6 ± 0.1 at 25.0 • C and µ = 1.0 M.
Equation (5) was then used to simulate the k -pH dependence data for the BH reaction; in the simulation, k 1 , k 2 , k 3 , and K a1 were treated as adjustable parameters and the value of K a2 obtained above was used as a direct input. The simulation provided an excellent fit shown in the top part of Figure 4 whereas the acquired values of k 1 , k 2 , k 3 , and K a1 are listed in Table 2.

Probing the Activation Process
After evaluation of the protolysis constants and the rate constants of the rate-determining steps, we were able to create species of BH/PAH versus pH distribution diagrams (the top parts of Figures  S7 and S8 in SM) and the reactivity of BH/PAH species versus pH distribution diagrams (bottom parts of Figures S7 and S8 in SM) [14]. Figures S7 and S8 demonstrate that species II of BH/PAH (cf. Figure 7) contributes predominantly in both distributions between pH 5 and 6, which corresponds to the plateau regions in the logk′ versus pH plots. In this small region, Equation (5) can be simplified to k′ ≈ 4k2 (or k2 ≈ k′/4). The oxidation reactions were thus investigated at several temperatures in this

Probing the Activation Process
After evaluation of the protolysis constants and the rate constants of the rate-determining steps, we were able to create species of BH/PAH versus pH distribution diagrams (the top parts of Figures S7  and S8 in SM) and the reactivity of BH/PAH species versus pH distribution diagrams (bottom parts of Figures S7 and S8 in SM) [14]. Figures S7 and S8 demonstrate that species II of BH/PAH (cf. Figure 7) contributes predominantly in both distributions between pH 5 and 6, which corresponds to the plateau regions in the logk versus pH plots. In this small region, Equation (5) can be simplified to k ≈ 4k 2 (or k 2 ≈ k /4). The oxidation reactions were thus investigated at several temperatures in this region; the results are summarized in Figures S9 and S10 in SM (the top parts) and in Table 3. The Eyring plots for k 2 are displayed in the bottom parts of Figures S9 and S10 for the reactions of BH and PAH. Activation parameters were calculated from these plots and are also listed in Table 3. Activation entropies ∆S 2 ‡ = −72 J·K −1 ·mol −1 for BH and ∆S 2 ‡ = −66 J·K −1 ·mol −1 for PAH are very close to each other and are of the middle-ranged negative values. These values, reflecting the activation processes between [IrCl 6 ] 2− and the neutral forms of BH and PAH, are consistent favorably with the nature of the second-order kinetics, where a compact structure of the transition state is expected. When the salient features of the rapid scan spectra are put together with the middle-ranged negative values of activation entropies, an outer-sphere electron transfer likely took place in the rate-determining step denoted by k 2 [14,28]. The same mode of electron transfer is expected to occur for the reactions expressed by k 1 and k 3 although it was not possible to determine the activation parameters for these reactions.

Comparison of the Rate Constants
For the oxidations of aryl hydrazides by [IrCl 6 ] 2− , the most surprising observation was that the reactivities of the protolysis species of hydrazides vary by about nine orders of magnitude [14]. This huge reactivity difference is also observed in the oxidation reaction of BH by [IrCl 6 ] 2− , being as k 1 :k 2 :k 3 = 1:1.3 × 10 4 :3.2 × 10 9 . BH is very close to INH in structure and the reactivities of its neutral and enolate forms are about the same as those of INH. INH is a frontline anti-tubercular drug but BH has essentially no anti-tubercular activity. Thus, the vital role played by the pyridine nitrogen in INH is not related directly to their reactivity in the reduction of a single electron oxidant. For the PAH-[IrCl 6 ] 2− reaction, the ratio of k 2 :k 3 = 1:7.6 × 10 3 becomes smaller but is still large. Thus, it can be concluded that the enolate forms of aryl hydrazides are exceptionally reactive towards reduction of [IrCl 6 ] 2− . This exceptionally high reactivity makes it possible that aryl hydrazides are potentially good candidates for antioxidants [32]. This also accounts for the good chemical preserving properties of hydrazides [2]. Another surprising observation in this work is that a methylene group reduces the reactivity of the enolate form of PAH about 120 times from that of BH. Table 4 summarizes the main results acquired so far for the oxidations of hydrazides by [IrCl 6 ] 2− including a very recent one (2-furoic hydrazide (FH), cf. Figure S1 in SM) studied by the Shi group [33]. The pK a values for the deprotonation of R-CONHNH 3 + locate in a small region, only varying from 3.04 to 3.67 whereas the reactivity of the neutral forms of hydrazides changes from 157 to 1120

Structure-Reactivity Relationship
Clearly, there is no correlation between the reactivity and pK a values (Table 4), which supports the conclusion drawn in the oxidations of hydrazine and substituted hydrazines by [IrCl 6 ] 2− [28]. On the other hand, a linear correlation is found between the reactivity of the enolate forms of the aryl hydrazides (INH, NH, FH, and BH) and the pK enol values as shown in Figure 9; the correlation is expressed by: logk enolate = (0.16 ± 0.07)pK enol + (6.1 ± 0.8). The big intercept and the small slope of the correlation indicate that the intrinsic reactivities of the enolate forms of aryl hydrazides are very high but are not very sensitive to the enolate basicities. These characters may pave a way for a rational design of new antioxidants based on aryl hydrazides. PAH no longer being an aryl hydrazide drops the correlation, suggesting that other aliphatic hydrazides may fall off the correlation as well. Hence, more data are needed to test whether aliphatic hydrazides together with PAH will follow another correlation.

Buffers and Reaction Media
The following buffering pairs of AcOH/NaOAc, NaH 2 PO 4 /Na 2 HPO 4 , NaHCO 3 /Na 2 CO 3 , and Na 2 HPO 4 /Na 3 PO 4 (all about 0.2 M) were combined to cover the pH range from 3.15 to 12.68; all the buffers which contained 2 mM PDCA [19] were adjusted to an ionic strength (µ) of 1.0 M by use of

Stoichiometric Investigation
The reactions were investigated in two reaction media: a phosphate buffer of pH 6.

Product Analysis by RP-HPLC and NMR Spectra
For identification of the oxidation product of BH, reaction mixtures of BH with [IrCl 6 ] 2− were analyzed by reversed-phase high performance liquid chromatography (RP-HPLC) using a Shimadzu LC-20 AD HPLC system equipped with a UV detector (Shimadzu Corporation, Kyoto, Japan). A C18 column of Shimadzu (250 × 4.6 mm, 5 µm in particle size) and an injection loop of 20 µL were used for sample separations and injections. Moreover, the injection loop was always fully filled with samples. After optimizations of mobile phase in an isocratic elution mode, a solvent mixture of H 2 O:MeOH = 4:1 (v/v) was chosen as the mobile phase. The UV detector was set at 261 nm and the flow rate was at 1.0 mL/min. Under the optimized conditions, a reaction mixture containing 2.5 mM BH and 2.5 mM [IrCl 6 ] 2− in a phosphate buffer of pH 6.34 after a reaction time of 10 min was subjected to analysis.
In the case of PAH, 1 H-NMR spectroscopy (AVANCE NEO 400 MHz NMR spectrometer, Bruker, Switzerland) was utilized to analyze the oxidation product. Two samples were prepared for NMR experiments: (a) 1 mM PAH in D 2 O which contained 0.02% TSP and (b) a reaction mixture of 1 mM PAH and 5 mM [IrCl 6 ] 2− after a reaction time of about 5 hrs.

Kinetic Measurements
A stock solution of 1.0 mM [IrCl 6 ] 2− was prepared and used daily by dissolving the desired amount of Na 2 IrCl 6 ·6H 2 O in a solution mixture containing 0.99 M NaClO 4 and 0.01 M HCl. Stock solutions of BH/PAH were prepared by adding the required amount of BH/PAH into a reaction medium of specific pH and then flushed for 5 min with nitrogen of high purity. For kinetic measurements, solutions of [IrCl 6 ] 2− and BH/PAH were prepared by dilution of the stock solutions with the same medium and then flushed for about 5 min with the nitrogen. Reactions were initiated by mixing equal volumes of the [IrCl 6 ] 2− and BH/PAH solutions directly on an SX-20 stopped-flow spectrometer (Applied Photophysics Ltd., Leatherhead, UK); the temperature was also controlled to ±0.1 • C using another thermostat of Lauda Alpha RA8. Moreover, the reaction solutions were only used for a couple of hours. The reactions were investigated under pseudo first-order conditions with [Hydrazide] tot ≥ 10·[IrCl 6 2− ].

Conclusions
The oxidation reactions of BH and PAH by [IrCl 6 ] 2− have been characterized in a wide pH range by use of stopped-flow spectral, rapid spectral scan, RP-HPLC, and NMR spectroscopic techniques. The nature of well-defined second-order kinetics of the reactions warrants the evaluation of the rate constants of rate-determining steps and the protolysis constants for both BH and PAH. This work clearly confirmed the earlier findings made by the Shi group [14] that the enolate forms of aryl hydrazides have exceptionally high reactivities towards reduction of Ir(IV). Moreover, for the aryl hydrazides, a linear correction of logk enolate = (0.16 ± 0.07)pK enol + (6.1 ± 0.8) is unraveled for the first time. Collectively, this work together with previous investigations [14,33] may pave a way for a rational design of new antioxidants based on aryl hydrazides. Additionally, the present work also offers the pK a values for BH and PAH at 25.0 • C and µ = 1.0 M.

Supplementary Materials:
The following are available online, Table S1: Observed second-order rate constants k' for oxidations of BH and PAH by [IrCl 6 ] 2− as a function of pH at 25.0 • C and 1.0 M ionic strength. Figure S1: Structures of hydrazides including aryl hydrazides BH, INH, NH, and FH. Figure S2: Kinetic traces acquired from the data points in Figure 1. The solid curves were obtained from the best fits of the experimental data to Equation (1). Figure Table 2. (Bottom): Reactivity versus pH distribution diagram for the BH species in the reduction of [IrCl 6 ] 2− ; the above pK a values and k 1 = 0.046, k 2 = 597, and k 3 = 1.47 × 10 8 M −1 s −1 in Table 1 were utilized in the calculations. Species I-III of BH are described in Figure 7. Figure S8: (Top): PAH species versus pH distribution diagram at 25.0 • C and µ = 1.0 M, which was calculated by use of pK a1 = 3.24 and pK a2 = 11.7 in Table 2. (Bottom): Reactivity versus pH distribution diagram for the PAH species in the reduction of [IrCl 6 ] 2− ; the above pK a values and k 1 = 0, k 2 = 157, and k 3 = 1.19 × 10 6 M −1 s −1 in Table 1 were utilized in the calculations. Species I-III of PAH are described in Figure 7. Figure