Ruthenium(III)–aminopolycarboxylato complexes active for the reduction of the N–N bond of hydrazine and phenylhydrazine in aqueous acidic media

Abstract

Interactions of hydrazines N₂H₄X⁺ (X = H or Ph) with tri-, tetra- and penta-chelated ruthenium(III)–aminopolycarboxylic acid complexes giving the respective monomeric hydrazinium (Ru^III–N₂H₄X⁺) adducts have been investigated by potentiometry, spectrophotometry and voltammetry in aqueous acidic solution at 25 °C. The deprotonation and metal hydrolysis constants of the complexes and their N₂H₄X⁺ adducts in 0.1 M Na₂SO₄ solution were determined. At pH 2.8, the complexes exhibited a quasi-reversible one-electron reduction wave of Ru^III → Ru^II in sampled dc in the potential range between −0.16 and −0.37 V vs. SCE, while their hydrazinium adducts obtained in situ by adding an excess of N₂H₄X⁺ showed an additional two-electron reduction wave assigned to Ru^III–N₂H₄X⁺ → Ru^I–N₂H₄X⁺ in the potential range of −0.02 to −0.35 V vs. SCE. The species Ru^I–N₂H₄X⁺ on successive decomposition and hydrolysis give one mole of each of NH₃, NH₂X and ruthenium(III) species. Further, the Ru^III–N₂H₄X⁺ complexes have been used as electro-catalysts for the reduction of N₂H₄X⁺ to NH₃ and NH₂X at a mercury pool cathode in acidic solutions of pH 1.9 and 2.8. The quantity of ammonia produced in all cases is linear with time. The E_1/2 of Ru^III–N₂H₄X⁺ → Ru^I–N₂H₄X⁺ and the turnover number are correlated with the sigma basicity (ΣpK_a) of the aminopolycarboxylic acids and the results are discussed in terms of the hydrolytic tendency of the metal, the number of co-ordinating groups and the steric repulsion caused by the increase in size of the aminopolycarboxylic acid.

---

Introduction

Studies on the reactivity of hydrazine at the potential co-ordination site of an active metal centre in the absence and the presence of reducing agents have immense value, since the reduction of dinitrogen to ammonia catalysed by transition metals present in microorganisms ^1–7 or simple metal complexes or clusters has long been proposed to proceed via metal bound hydrazine. In the recent past intensive research has been carried out to evolve a simple mechanism for dinitrogen fixation. In this respect, numerous transition metals, cofactors and model compounds containing hydrazine were synthesized and characterized.^8–10 Structural analyses of a few of these complexes have also been made.^8,11 Meanwhile, reduction studies were also performed on hydrazine or its derivatives to ammonia and/or amine in both aqueous and non-aqueous solutions either in the absence or in the presence of external proton and electron sources or reducing agents.^12–14 Although, fairly good turnovers of ammonia have reportedly been obtained in organic solvents, most of these reactions did not meet much success in aqueous solutions. However, Schrauzer et al.¹⁵ reported the reduction of hydrazine to ammonia using K₂[Mo(O)(H₂O)(CN)₄] as a catalyst and NaBH₄ as reducing agent with a turnover rate of 4.2 in aqueous solutions. Later catalytic conversion of hydrazine into ammonia was obtained by [Mo₂Fe₆S₈L₉]³⁻ or [Fe₄S₄L₄]²⁻ (L = SPh or SCH₂CH₂OH) with a moderate turnover number.¹⁶ In another report,¹⁷ [WCp*Me₃(η²-NH₂NH₂)][O₃SCF₃] prepared from [WCp*Me₃(O₃SCF₃)] and N₂H₄ was shown to undergo reduction by Na–Hg to give NH₃. Recently, the co-ordinatively unsaturated diruthenium complex [Cp*Ru(μ-SPrⁱ)₂RuCp*] has been reported to catalyse the reduction of hydrazine more effectively.¹⁸ More sophisticated systems [MoFe₃S₄Cl₃(Cl₄C₆O₂)(MeCN)]²⁻ and [MoFe₃S₄Cl₃(citr)]³⁻ (citr = citrate) were also modelled and used to catalyse the reduction of N₂H₄ to NH₃ with good turnovers by CoCp₂ and 2,6-dimethylpyridine hydrochloride.^13,19 Moreover, the N₂H₄-bridged double cubane [{MoFe₃S₄Cl₃(Cl₄C₆O₂)}₂(μ-NH₂NH₂)] was isolated and its catalytic activity studied.²⁰ However, the parallel evolution of hydrogen gas which reduces the yield of NH₃ was reported as the major limiting step in almost all the above systems.

Recently, we have reported the electrochemical reduction of hydrazine and its phenyl derivative to NH₃ and/or NH₂X at high turnover rates and high coulombic efficiency ^21–24 without the liberation of hydrogen gas using terminally co-ordinated Ru^III–edta–N₂H₄X⁺ complexes as electro-catalysts in aqueous acidic solutions. In continuation, herein we describe the interaction of N₂H₄X⁺ (X = H or Ph) with tri-, tetra- and penta-chelated aminopolycarboxylates of ruthenium(III), viz. K₂[Ru(imda)Cl₃]·2H₂O 1, K₂[Ru(himda)Cl₂]·H₂O 2, K₂[Ru(nta)Cl₂]·H₂O 3, K[Ru(hedta)Cl] 4, K[Ru(Hedta)Cl]·2H₂O 5, K[Ru(Hpdta)Cl]·2H₂O 6, K[Ru(Hcdta)Cl]·H₂O 7 and K[Ru(H₂dtpa)Cl] 8 by potentiometry, spectrophotometry and voltammetry in aqueous solution. The function of the resulting hydrazinium complexes as catalysts for the reduction of hydrazine and its phenyl derivative electrochemically in acidic solution is elucidated. The effects of the sigma basicity (ΣpK_a) of the aminopolycarboxylic acid in complexes 1–8 on the half-wave potential (E_1/2) of the two-electron reduction wave of Ru^III–N₂H₅⁺ complexes and the turnover number of NH₃ produced in the electrolytic reduction of N₂H₄X⁺ have been investigated.

Experimental

Ruthenium trichloride (RuCl₃·xH₂O) from Arora Matthey India, iminodiacetic acid (H₂imda), N-(2-hydroxyethyl)iminodiacetic acid (H₂himda), nitrilotriacetic acid (H₃nta), N-carboxymethyl-N′-(2-hydroxyethyl)ethylenediiminodiacetic acid (H₃hedta), propylenedinitrilotetraacetic acid (H₄pdta), (±)-trans-cyclohexane-1,2-diyldinitrilotetraacetic acid (H₄cdta), and carboxymethyliminobis(ethylenenitrilo)tetraacetic acid (H₅dtpa) from Aldrich Chemical Company Inc. USA were purchased. Ethylenedinitrilotetraacetate disodium salt (Na₂H₂edta) from Polypharm India, hydrazine sulfate (N₂H₅HSO₄) from BDH Chemicals, India and phenylhydrazine hydrochloride (N₂H₄PhCl) from Allied Chemicals, USA. The zero grade Ar gas from IOLAR & Co. was used after purification with vanadium sulfate and alkaline pyrogallol solutions. All other reagents used were of AR grade. A mixture of 0.2 M CH₃COONa and 18 N H₂SO₄ solutions at a desired pH between 1 and 5 was used as supporting electrolyte in all voltammetric studies. Stock solutions of N₂H₅HSO₄ (0.1 M) were prepared and standardized with potassium iodate.²¹ A freshly prepared solution of 0.1 M N₂H₄PhCl was used to prepare the experimental solutions.

Preparations

K₂[RuCl₅(OH₂)].. The compound was prepared by the method reported elsewhere,²⁵ and used as the precursor to synthesize three- to five-co-ordinated ruthenium(III)–aminopolycarboxylic acid complexes 1–8 according to the following methods.

K₂[Ru(imda)Cl₃]·2H₂O 1.. To a hot solution of 10 mL K₂[RuCl₅(OH₂)] (0.375 g, 1 mmol) in 1 mM HClO₄, H₂imda (0.136 g, 1.02 mmol) dissolved in 10 mL of 1 mM HClO₄ was added slowly with constant stirring. The resultant mixture was refluxed (two hours) till the reddish brown colour changed to pale yellow. Later, the volume of the solution was reduced (5 mL) in vacuo on a Rotovapour. It was then precipitated with cold absolute alcohol, filtered off, washed with 9∶1 acetone–water till free of chloride and dried under vacuum. Calc. for C₄H₉Cl₃K₂NO₆Ru: C, 10.46; H, 1.92; N, 3.07. Found: C, 10.60; H, 2.01; N, 3.09%. IR (KBr): 1635 (COO⁻ asym); 1365 (COO⁻ sym); 1210 (CO); 840 (OCO def); 535 cm⁻¹ (M–N). UV-vis, λ/nm (ε/M⁻¹ cm⁻¹): 240 (2270); 285 (1400) and 365 (445).

K₂[Ru(himda)Cl₂]·H₂O 2.. The compound 2 was prepared from K₂[RuCl₅(OH₂)] (0.375 g, 1 mmol) and H₂himda (0.180 g, 1.02 mmol) by the procedure described for 1. Calc. for C₆H₁₀Cl₂K₂NO₆Ru: C, 16.04; H, 2.32; N, 3.09. Found: 16.29; H, 2.28; N, 3.17%. IR (KBr): 1640 (COO⁻ asym); 1365 (COO⁻ sym); 1235 (CO); 825 (OCO def); 535 cm⁻¹ (M–N). UV-vis, λ/nm (ε/M⁻¹ cm⁻¹): 245 (2595); 285 (2360) and 360 (640).

K₂[Ru(nta)Cl₂]·H₂O 3.. This compound was prepared from K₂[RuCl₅(OH₂)] (0.375 g, 1 mmol) and H₃nta (0.195 g, 1.02 mmol) in the same manner as described above for 1. Calc. for C₆H₈Cl₂K₂NO₇Ru: C, 15.85; H, 1.68; N, 3.13. Found: C, 15.79; H, 1.76; N, 3.07%. IR (KBr): 1635 (COO⁻ asym); 1358 (COO⁻ sym); 1230 (CO); 850 (OCO def); 525 cm⁻¹ (M–N). UV-vis, λ/nm (ε/M⁻¹ cm⁻¹): 285 (2420) and 360 (735).

K[Ru(H_mL)Cl]·xH₂O {m = 0, x = 0 for L = hedta 4; m = 1, x = 2 for edta 5; m = 1, x = 2 for pdta 6}.. These compounds were prepared from K₂[RuCl₅(OH₂)] and the respective aminopolycarboxylic acid according to the procedures reported earlier ^26–28 and subsequently characterized by physico-chemical methods.

K[Ru(Hcdta)Cl]·H₂O 7.. This complex was obtained by treating K₂[RuCl₅(OH₂)] (0.375 g, 1.0 mmol) in 10 mL 1 mM HClO₄ with Na₂H₂cdta (0.400 g, 1.02 mmol) in 20 mL 1 mM HClO₄ under a nitrogen atmosphere, initially for 30 min with stirring at room temperature followed by refluxing for about two hours. The greenish yellow solution was evaporated to small volume. The complex was precipitated with cold absolute ethanol, washed with 9∶1 acetone–water solution until free from chloride and dried under vacuum. Calc. for C₁₄H₂₁ClKN₂O₉Ru: C, 31.32; H, 4.19; N, 5.14. Found: C, 31.20; H, 4.30; N, 5.20%. IR (KBr): 3450 (OH); 1750 (COOH); 1655 (COO⁻ asym); 1375 (COO⁻ sym); 1220 (CO); 885, 685 (OCO def); 520 cm⁻¹ (M–N). UV-vis, λ/nm (ε/M⁻¹ cm⁻¹): 240 (2430); 285 (1640) and 370 (700).

K[Ru(H₂dtpa)Cl] 8.. This complex was isolated from K₂[RuCl₅(OH₂)] and H₅dtpa (0.403 g, 1.02 mmol) in 20 mL of HClO₄ by the procedure described in the case of 7. Calc. for C₁₆H₂₀ClKN₃O₁₂Ru: C, 29.82; H, 3.49; N, 7.34. Found: C, 29.7; H, 3.56; N, 7.43%. IR (KBr): 3560 (OH); 1750 (COOH); 1655 (COO⁻ asym); 1375 (COO⁻ sym); 1280 (CO); 885, 685 (OCO def); 525 cm⁻¹ (M–N). UV-vis, λ/nm (ε/M⁻¹ cm⁻¹): 235 (2235); 280 (1560) and 370 (375).

Instrumentation

Perkin-Elmer Series II-2400, CHNS/O Analyzer for C, H, N-data; Bio-Rad FT-40 spectrometer coupled to a SP-3200 computer for IR; Shimadzu UV-vis NIR scanning spectrophotometer UV-3101 PC for absorption spectra. An Adair Dutt digital pH meter (±0.01 pH) was used to record the pH of all experimental solutions other than those in potentiometry.

Electrochemical measurements were performed on EG&G Princeton Applied Research (PARC) instruments. A model PAR 174A Polarographic Analyzer and PAR 175 Universal Programmer coupled to a high precision Houston X-Y recorder were used to record sampled dc polarograms and cyclic voltammograms. A three electrode assembly, PAR 303 SMDE/HMDE comprising a dropping (DME, 3.85 mg s⁻¹)/hanging (HMDE, 0.021 cm²) mercury drop working, platinum wire auxiliary and SCE reference electrodes was employed.

The controlled potential coulometry was performed on a EG&G PAR model 173 Potentiostat/Galvanostat coupled to a model 179 digital coulometer provided with a three electrode cell assembly. The cell consisted of a mercury pool (4 cm convex diameter) working electrode in the main compartment along with a platinum mesh as counter electrode separated by a glass frit, SCE reference electrode and a glass disc agitator. An Orion 940 Ion Analyzer equipped with a model 9512 ammonia sensing membrane electrode (sensitive to 10⁻¹² M) was used for the estimation of ammonia in the electrolysed solutions.

Analytical methods

All acid–base titrations were conducted in a jacketed glass-walled cell of about 150 mL capacity, connected to a Metrohm Swiss model 682 Auto Titroprocessor coupled with a 665 Dosimat and E-649 magnetic stirrer, through a combined glass-reference electrode. The cell also had inlet and outlet tubes for passing Ar gas through the solution. The electrode system was calibrated in terms of H⁺ concentration by means of a titration of HCl solution with NaOH ²⁹ in 0.1 M Na₂SO₄ at 25 ± 0.1 °C. The experimental data ranged over pH values between 1.6 and 11.2.

In a typical experiment, a weighed quantity of a complex 1–8, N₂H₅HSO₄ or N₂H₄PhCl (1 mM) was dissolved in 50 mL of 0.1 M Na₂SO₄ and titrated independently against carbonate free (0.1 M) NaOH solution in the region 2 < pH < 10 under Ar at 25 °C. Similar titrations were conducted in the presence of 1 equivalent of N₂H₅HSO₄/N₂H₄PhCl. The resulting data were used to calculate the acid dissociation (pK_a) constants corresponding to N₂H₄X⁺, 1–8 and adducts of both, and the metal hydrolysis (pK_OH) constants of the last two systems. The titration data at pH > 10 in the case of 1–8 and at pH > 9.0 in the case of N₂H₄X⁺ adducts were not considered in the present investigations. The acid dissociation constants of N₂H₄X⁺ and the unco-ordinated COOH/CH₂CH₂OH in complexes 1–8 and in their 1∶1 adducts of N₂H₄X⁺, where the dissociation of protons takes place in well defined buffer regions, were evaluated ²⁹ with the help of K_w (1.008 × 10⁻¹⁴ at 25 °C) and suitable mass and charge balance equations fitting to the generalized eqn. (1). Charges on the species are omitted for the sake of clarity. The stepwise first, second or third metal hydrolyses of 1–8 and their N₂H₄X⁺ adducts were considered according to reactions (2). The hydrolysis constants falling in the pH range 2 < pH < 10 were calculated with the help of suitable mass balance equations and K_w at 25 °C.

A graphical method ³⁰ was used to calculate the dissociation constants, in the case of simultaneous dissociation of two protons (either by the participation of COOH/CH₂CH₂OH and/or in the combination of dissociation of N₂H₄X⁺ proton and metal hydrolysis) described by the generalized eqn. (3) or combined eqns. (1) and (2), respectively.

The absorption spectral studies on complexes 1–8 (1 mM) in the absence and in the presence of various concentrations of N₂H₅⁺ or N₂H₄Ph⁺ (1–100 mM) were performed at pH 2.8 in a 10 mm quartz cuvette and the spectra were recorded in the wavelength range between 200 and 700 nm at 25 ± 0.1 °C.

The voltammetric responses were recorded with a weighed quantity of complex (1 mM) in the absence or presence of a known concentration (0.1–100 mM) of substrate (N₂H₅⁺/N₂H₄Ph⁺) in 10 mL electrolyte solution at the desired pH under Ar at 25 °C. The ivs.E plots were recorded between +0.2 and −0.8 V (vs. SCE) at the sweep rates 10 mV s⁻¹ in sampled dc and 20–500 mV s⁻¹ in cyclic voltammetry (CV). The effect of pH on the diffusion current (i_d) and half-wave potentials (E_1/2) was studied in the range 1–5. All the solutions were deaerated at least 15 min prior to measurements and the experiments were maintained at 25 ± 0.1 °C unless otherwise stated. The criteria for the electrochemical reversibility and diffusion current were established by following the reported techniques.^31,32

Constant potential electrolysis was performed in 25 mL of deaerated buffer solution having the desired pH (1.9/2.8) containing 1 mmol of complex (1–8) and 100 mmol of substrate at predetermined potential for at least 10 h under Ar at 25 ± 0.1 °C. The ammonia produced during the reaction was estimated every hour ²¹ while the aniline produced concomitantly with ammonia was tested qualitatively using vanadium(V) salt.³³

Results and discussion

Interaction of N₂H₄X⁺ with the complexes 1–8

Potentiometry.. Complexes 1–8 were independently titrated against standard sodium hydroxide in the absence and in the presence of one equivalent of N₂H₄X⁺. The representative titration data thus obtained with 1 are presented in Fig. 1. Complex 1 (Fig. 1(a)) showed three titratable protons dissociating in three independent buffer regions, two in acidic and one in basic pH by the metal hydrolysis. Further hydrolysis of 1 occurring beyond a > 3 resulted in a dark brown solution which may be due to its decomposition and hence was not considered here. Similarly, complexes 2 and 3 showed three buffer regions caused by metal hydrolyses at 2 < pH < 10 with two ill defined inflections at a = 1 and 2. On the other hand, 4–7 showed the stepwise liberation of two protons in two well defined buffer regions 0 < a < 1 and 1 < a < 2. Of the two, the first buffer region is assigned to the dissociation of an unco-ordinated CO₂H proton in the case of complexes 5–7 and may be the OH (CH₂CH₂OH) proton in the case of 4 while the second buffer region is considered for the metal hydrolysis.^33,34 On the contrary, the complex 8 showed one buffer region 0 < a < 2 in acidic pH followed by another buffer region 2 < a < 3 in basic pH. The first buffer region is assigned to the simultaneous liberation of two protons on the dtpa ligand, the second to metal hydrolysis.

Fig. 1 Potentiometric titration curve of (a) K₂[Ru(imda)Cl₃]·2H₂O, 1 mM; (b) K₂[Ru(imda)Cl₃]·2H₂O, 1 mM and N₂H₅⁺, 1 mM; (c) K₂[Ru(imda)Cl₃]·2H₂O, 1 mM and N₂H₄Ph⁺, 1 mM at 25 °C. I = 0.1 M (Na₂SO₄), a = mols of base added per mol of complex or N₂H₄X⁺.

The metal hydrolysis steps, three in the case of complex 1, the first two in the case of 2 and 3, the single one in the case of 4–6 and 8, and the first one in the case of 7, are accounted for by the rapid substitution of chloride ions by H₂O as verified kinetically in the case of 4–6, and the liberation of titratable protons following its dissociation.^26,35 The third metal hydrolysis step in the case of 2 and 3 could be explained if one of the loosely bound carboxylate groups of the tetradentate himda and nta is substituted by H₂O or the liberation of OH (CH₂CH₂OH) proton on tridentate himda in 2. The acid (pK_a) and metal hydrolysis (pK_OH) constants were calculated with the help of eqns. (1)–(3) and are given in Table 1. These showed that the complexes 1–3 are more hydrolysable because the chelation capability of imda (tri), himda (tri or tetra) and nta (tetra) ligands is low. All other complexes, 4–8, showed a single hydrolysis step which occurs almost at neutral pH. This implies that all these complexes have a potential site available to another ligand for co-ordination/substitution, though the number of co-ordinating sites in the aminopolycarboxylic acid ligand varies from 5 to 8. The present data also revealed that only five co-ordination sites on the hexacovalent ruthenium are used by all penta- or higher poly-dentate aminopolycarboxylic acids in the process of bond formation. The hydrolytic tendency of the metal increased in the order 8 < 4 < 5 < 6 < 7 < 2 < 3 < 1.

Table 1 Acid dissociation (pK_a) and metal hydrolysis (pK_OH) constants of complexes 1–8 at 25 °C, I = 0.1 M (Na₂SO₄)

	pKa		pKOH
Complex	CO2H	CH2CH2OH	First	Second	Third
Values given in parentheses were evaluated by a graphical method.
1	—	—	2.42 ± 0.02	5.78 ± 0.03	8.86 ± 0.03
2	—	—	2.86 ± 0.02	5.72 ± 0.02	8.53 ± 0.03
3	—	—	2.47 ± 0.01	4.80 ± 0.03	9.62 ± 0.03
4	—	4.82 ± 0.03	7.96 ± 0.03	—	—
5	2.34 ± 0.02	—	7.64 ± 0.02	—	—
6	2.38 ± 0.02	—	7.60 ± 0.02	—	—
7	2.55 ± 0.03	—	6.20 ± 0.03	—	—
8	(2.58, 3.40)	—	8.28 ± 0.02	—	—

---

The titration data of N₂H₅HSO₄ showed two buffer regions separated by an inflection at a = 1. The buffer region 0 < a < 1 is considered for the neutralization of the strong acidic proton due to the HSO₄⁻ group, since the N₂H₆⁺ (K_a 11) ¹¹ does not form inder the present experimental conditions. The buffer region corresponding to this proton was ignored in all other studies presented below and hence not included in Fig. 1(b), but considered in the evaluation of relevant constants. The second buffer region 1 < a < 2 is assigned to the neutralization of proton in the equilibrium N₂H₅⁺ ⇌ N₂H₄ + H⁺. Similarly, the titration data of N₂H₄PhCl contained a single buffer region 0 < a < 1, representing the equilibrium N₂H₄Ph⁺ ⇌ N₂H₃Ph + H⁺.

The titration curves of complexes 1–8 in the presence of N₂H₅⁺ possessed buffer regions, four in the case of 1 (Fig. 1(b)) and three in the case of 2–8. The last buffer region (fourth in the case of 1, and third in the case of other complexes), where Ru^III–N₂H₅⁺ species were decomposed to give ruthenium(II) complexes, as found by sampled dc at pH > 9, was not considered in the present study. The third buffer region in the case of 1 and the second in the case of 2–8, which were assigned earlier to metal hydrolysis in the absence of N₂H₅⁺, are assigned to the equilibrium N₂H₅⁺ ⇌ N₂H₄ + H⁺ in the vicinity of the metal co-ordination sphere. The first and second buffer regions in the case of 1 and the first in the case of 2 and 3 are assigned to the metal hydrolysis, while the first in the case of 4–7 and the first and second buffer regions in the case of 8 are assumed for the deprotonation of the CO₂H/CH₂CH₂OH group on the aminopolycarboxylic acid as found above in the absence of N₂H₅⁺.

The potentiometric data obtained in the presence of N₂H₄PhCl showed two buffer regions: a = 0–1, 1–3 in the case of complex 1 (Fig. 1(c)), a = 0–1, 1–2 in the case of 2, 3 and 5–7 and a = 0–2, 2–3 in the case of 8, while one buffer region a = 0–2 in the case of 4. It was noticed that the hydrolysis steps, third in the case of 1, second in the case of 2 and 3, and first in the case of 5–8 which were observed in the absence of N₂H₄Ph⁺, were not seen. Thus, the deprotonation of N₂H₄Ph⁺ was assumed in the buffer region a = 1–3 in the presence of 1 along with the second metal hydrolysis, 1–2 in the presence of 2, 3, 5–7, 0–2 in the case of 4 together with the deprotonation of the OH (CH₂CH₂OH) proton and 2–3 in the case of 8.

The acid dissociation constants and metal hydrolysis constants corresponding to free N₂H₄X⁺ and its adducts with complexes 1–8 were calculated with the help of eqns. (1)–(3) and presented in Table 2. The values of K_a (CO₂H/CH₂CH₂OH) and K_OH are closely comparable to the corresponding values seen in Table 1. Moreover, the values obtained for N₂H₄X⁺ ⇌ N₂H₃X + H⁺ in the presence of 1–8 are in close agreement with those obtained in their absence. However, the K_a value is reduced to only 0.38–1.0 logarithmic units in the case of N₂H₅⁺ and negligibly affected in the case of N₂H₄Ph⁺ after their interaction with the ruthenium complex. On the basis of these observations together with the disappearance of metal hydrolysis steps (third in the case of 1, second in the of 2 and 3, and first in the other complexes as seen in Table 1) in the presence of N₂H₄X⁺, the co-ordination of N₂H₄X⁺ to ruthenium occurs at the vacant site where the poorly dissociable water was attached. Further, the small changes in pK_a values of the N₂H₄X⁺ ligand after its interaction with ruthenium suggests that its co-ordination takes place through the weakly basic nitrogen (NH₂ of N₂H₅⁺ or NHPh of N₂H₄Ph⁺) leaving the other nitrogen in the protonated form, which is further confirmed by the voltammetric data presented at pH < 3.0 in the latter part of this paper.

Table 2 Acid dissociation (pK_a) and metal hydrolysis (pK_OH) constants of N₂H₄X⁺ and their adducts with complexes 1–8 at 25 °C, I = 0.1 M (Na₂SO₄)

Adduct		pKa		pKOH
N2H4X^+	Complex	N2H4X^+	CO2H/CH2CH2OH	First	Second
Values given in parentheses were evaluated by a graphical method.
N2H5^+	—	8.05 ± 0.01	—	—	—
	1	7.67 ± 0.02	—	2.43 ± 0.02	5.82 ± 0.03
	2	7.42 ± 0.03	—	2.82 ± 0.02
	3	7.36 ± 0.02		2.45 ± 0.02	—
	4	7.65 ± 0.02	4.83 ± 0.02	—	—
	5	7.52 ± 0.02	2.40 ± 0.02	—	—
	6	7.28 ± 0.03	2.42 ± 0.02	—	—
	7	7.15 ± 0.02	2.58 ± 0.02	—	—
	8	7.30 ± 0.03	(2.60, 3.43)	—	—
N2H4Ph^+	—	5.27 ± 0.02	—	—	—
	1	(5.24)	—	2.45 ± 0.02	(5.84)
	2	5.25 ± 0.02	—	2.89 ± 0.03	—
	3	5.34 ± 0.02	—	2.50 ± 0.03	—
	4	(5.32)	(4.62)	—	—
	5	5.20 ± 0.02	2.35 ± 0.02	—	—
	6	5.22 ± 0.02	2.41 ± 0.02	—	—
	7	5.31 ± 0.02	2.60 ± 0.02	—	—
	8	5.30 ± 0.02	(2.64, 3.49)	—	—

---

Absorption spectra.. The electronic absorption spectra of complexes 1–8 were recorded under identical conditions in the absence and presence of N₂H₄X⁺ salts at different compositions at pH 2.8 to verify the interaction of the latter (N₂H₄X⁺) through the weakly basic nitrogen with ruthenium. In the absence of N₂H₄X⁺ the ligand charge transfer between 240 and 250 nm and the ligand to metal charge transfer bands in the regions 280–290 and 360–370 nm were observed for all complexes. The absorption band at 360 nm, characteristic of a Ru–OH₂ bond,³⁶ was reduced in intensity on introduction of N₂H₅⁺ due to the replacement of bound water molecule by the latter. In the case of addition of N₂H₄Ph⁺ two new absorption bands at 400–450 and 540 nm appeared rapidly. The band at 540 nm did not occur in the case of complexes 7 and 8. Concomitantly, the yellow colour of these complexes instantly changed to reddish brown accounting for the formation of N₂H₄Ph⁺ adducts in solution. Further, plots of change in absorbance (ΔA) vs. the ratio [N₂H₄X⁺]∶[complex] confirmed the formation of 1∶1 N₂H₄X⁺ adducts as suggested above with 1–8. Fig. 2 shows the typical spectral changes observed with 4 in the absence and in the presence of N₂H₄X⁺ (1 mM) at pH 2.8. The spectral features observed with all complexes 1–8 are summarized in Table 3.

Table 3 Absorption spectral data of complexes 1–8 in the absence and in the presence of N₂H₄X⁺ at pH 2.8

	λ max/nm (εmax/M^−1 cm^−1)
Complex	Absence	Presence of N2H5^+	Presence of N2H4Ph^+
1	240(2270); 285(1400); 365(445)	240(2240); 285(1340)	420(1070); 540(350)
2	245(2595); 285(2360); 360(640)	245(2370); 285(2320)	425(1450); 540(350)
3	245(2330); 285(2200); 360(735)	245(2290); 285(2120)	420(1360); 540(350)
4	240(2270); 285(1400); 360(445)	240(2240); 285(1375)	400(1400); 540(260)
5	245(2600); 285(2370); 360(640)	245(2570); 285(2360)	420(2250); 540(480)
6	245(2565); 285(2420); 360(630)	245(2560); 285(2380)	400(1350); 540(130)
7	240(2430); 285(1630); 360(700)	240(2390); 285(1580)	430(1230)
8	235(2240); 280(1560); 360(375)	235(2230); 280(1540)	450(1270)

---

Fig. 2 Absorption spectra of (a) K[Ru(hedta)Cl], 1 mM; (b) K[Ru(hedta)Cl], 1 mM and N₂H₅⁺, 1 mM; (c) K[Ru(hedta)Cl], 1 mM and N₂H₄Ph⁺, 1 mM. In all cases the pH was 2.8.

Voltammetry.. Sampled dc polarograms of complexes 1–8 were recorded in the absence and in the presence of various equivalents of N₂H₄X⁺. In the absence, each complex exhibited a well defined one-electron cathodic wave in the potential range between −0.150 and −0.400 V (E_1/4 − E_3/4 55–65 mV; I (diffusion current constant) 1.2–1.5) vs. SCE assignable to the Ru^III → Ru^II process. The height of this wave proportionately enhanced with increase in complex concentration, while the plots E_de (potential at the electrode) vs. log [i/(i_d − i)] were linear with a slope of about 70 mV, indicating the electrode reactions are one-electron, diffusion controlled and quasi-reversible processes. The wave shifted to a measurable extent (60 mV per pH unit) in the case of 1–3, less in the case of other complexes with the change in pH between 1.0 and 5.0. The polarogram of 1 representing this one-electron wave is shown in Fig. 3(a) while the half-wave potential (E_1/2) data of all the complexes 1–8 are listed in Table 4. It decreased in the order 8 > 5 > 4 > 6 > 7 > 3 > 2 > 1 which is in good agreement with the order of sigma basicity of the co-ordinated aminopolycarboxylic acid.³⁷ However, it is noted that the E_1/2 values for 1–3 are relatively more negative than those of other complexes. This may be attributed to the strong hydrolytic tendency of these complexes when dissolved in aqueous solution. Accordingly, the hydrolytic tendency of these complexes may be given as 3 < 2 < 1. On the other hand, the lower E_1/2 value for complex 8 than those of 4–7 may be accounted for by its destabilization due to steric repulsion because of the large size of dtpa, despite its large sigma basicity.

Table 4 Half wave potentials of the Ru^III → Ru^II wave and those of complexes 1–8 observed in the presence of 100 equivalents of N₂H₄X⁺ at pH 2.8

	−E1/2/V vs. SCE
Complex	Absence	Presence of N2H5^+	Presence of N2H4Ph^+
1	0.372	0.346, 0.475	0.360, 0.482
2	0.364	0.328, 0.450	0.340, 0.462
3	0.354	0.300, 0.440	0.322, 0.456
4	0.248	0.030, 0.256	0.165, 0.248, 0.335
5	0.238	0.025, 0.230	0.150, 0.225, 0.346
6	0.251	0.034, 0.250	0.156, 0.235, 0.340
7	0.286	0.045, 0.230	0.168, 0.260, 0.386
8	0.165	0.028, 0.250	0.154, 0.250, 0.395

---

Fig. 3 Sampled dc polarogram of (a) K₂[Ru(imda)Cl₃]·2H₂O, 1 mM; (b) K₂[Ru(imda)Cl₃]·2H₂O, 1 mM and N₂H₅⁺, 100 mM; (c) K₂[Ru(imda)Cl₃]·2H₂O, 1 mM and N₂H₄Ph⁺, 100 mM. The pH in each case was maintained at 2.8.

When 0.1–5.0 equivalents of N₂H₅⁺ were introduced into the complex solution the Ru^III → Ru^II wave was enhanced while its plateau extended to anodic potentials in cases of complexes 1–8. As a result, the E_1/2 value apparently shifted to anodic potentials. A plot of the enhanced current increased linearly as the number of equivalents of added N₂H₅⁺ approached unity and remained constant thereafter indicating the instant formation of monomeric N₂H₅⁺ complexes in acidic solutions, as suggested above. The plateau of the enhanced wave gradually split into two as [N₂H₅⁺] reached twenty times more than that of the complex and the wave finally appeared as two distinguishable waves (w₁ and w₂) at all other compositions of N₂H₅⁺. The polarographic responses of 1 in the presence of 100 equivalents of N₂H₅⁺ are shown in Fig. 3(b). In all cases, the wave w₁ (E_1/4 − E_3/4 32 mV; I 1.9) shifted anodically with the increase in [N₂H₅⁺] while the wave w₂ (E_1/4 − E_3/4 65 mV; I 1.0) remained steady. On the other hand, the diffusion currents of w₁ and w₂ were directly proportional to [complex]. The plots E_devs. log [i/(i_d − i)] for w₁ and w₂ were linear with slopes of 33 and 68 mV, respectively. The diffusion current (i_d) of the composite wave or of the wave w₁ or w₂ was maximum in the pH range 2.5–3.5 and somewhat reduced when the ionic strength was raised beyond 0.5 M. Besides, the wave w₁ shifted negligibly with increase in pH between 1 and 4. The E_1/2 data (Table 4) revealed that the wave w₁ is 0.02–0.05 V less negative in the case of 1–3, while it is 0.13–0.22 V less negative in the case of 4–8 compared to that of the corresponding one-electron Ru^III → Ru^II wave observed before the addition of N₂H₅⁺. Similarly, the wave w₂ is about 0.09–0.10 V more negative in the case of 1–3 and 8 and similar in the case of 4–7 to that of the initial one-electron Ru^III → Ru^II wave. The wave w₁ in the case of 1–3 is more negative than that of the complexes 4–7. The E_1/2 of w₁ is correlated with ΣpK_a of the primary ligand,³⁷ aminopolycarboxylic acid. The large negative E_1/2 for complexes 1–3 is responsible for their hydrolytic tendency, while the negligible dependence of E_1/2 in complexes 4–8 shows that there is no considerable change in the redox properties of the metal with the increase in sigma basicity of the aminopolycarboxylic acid. Probably, this could be due to identical binding of penta-, hexa- or hepta-dentate aminopolycarboxylic acids to ruthenium through the same number of co-ordinating groups.

The polarographic responses of complex 1 in the presence of 0.1–100 equivalents of N₂H₄Ph⁺ were probed. The response in 100 equivalents of N₂H₄Ph⁺ is depicted in Fig. 3(c). It displayed two closely separated waves (w₁ and w₂) at potentials close to that of the Ru^III → Ru^II wave with a doubled diffusion current. The responses in the presence of other concentrations (0.1–100 equivalent of N₂H₄Ph⁺) overlapped those in Fig. 3(c) except for a negligible negative shift in E_1/2 and a minor change in the overall diffusion current. Complexes 2 and 3 also exhibited two cathodic waves replacing the Ru^III → Ru^II wave while the other complexes (4–8) showed an additional low intensity wave poorly resolved from w₁ and w₂. The E_1/2 data pertaining to these waves are summarized in Table 4. The measured data (E_1/4 − E_3/4 34 mV and I 2.0 for w₁; 65 mV and 0.9 for w₂) indicated w₁ to be a two-electron process while w₂ and the additional wave are one-electron processes. This was further substantiated by plots of E_devs. log [i/(i_d − i)] which were linear having slopes of 32–35 mV for w₁ and 67–70 mV for both w₂ and the additional wave. The overall diffusion current at −0.5 V was proportional to [1] while the enhanced current (Δi_d) was a maximum in the pH range 2.0 to 3.5. The data in Table 4 revealed that, for a given complex, the E_1/2 of both w₁ and w₂ are more negative with N₂H₄Ph⁺ than with N₂H₅⁺, indicating that the former hydrazine binds to ruthenium more strongly than the latter can through a sigma bond. This may due to the effect of phenyl ring substitution on the co-ordinating nitrogen in the former hydrazine. The E_1/2 values of w₁ obtained with complexes 1–8 in the presence of N₂H₄Ph⁺ are correlated with ΣpK_a of the aminopolycarboxylic acid ³⁷ attached to ruthenium. Again, the large negative values in the case of 1–3 and negligible dependence of E_1/2 in the case of 4–8 are accounted for by the reasons given before.

The CV responses of complexes 1–8 in the absence and in the presence of N₂H₄X⁺ were run subsequently for the above solutions. The results in the absence and in the presence of 100 equivalents of N₂H₄X⁺ are discussed below. All the complexes, in the absence of N₂H₄X⁺, yielded a pair of cathodic and anodic peaks corresponding to the quasi-reversible Ru^III–Ru^II couple. Fig. 4(a) shows a typical cyclic voltammogram of 4, which peaked at −0.320 V in the forward scan and at −0.252 V in the reverse scan. The peak separation which varied between 70 and 65 mV at 0.1 V s⁻¹ increased with increase in scan speed, while the peak width E_p/2 − E_p tended to 62–65 mV which is close to the theoretical value for a one-electron process. On the other hand, the peak currents increased linearly with increase in the square root of the scan speed, while the peak to peak ratio (i_pa∶i_pc) was less than unity.

Fig. 4 Cyclic voltammograms of (a) K[Ru(Hcdta)Cl]·H₂O, 1 mM; (b) K[Ru(Hcdta)Cl]·H₂O, 1 mM and N₂H₅⁺, 100 mM; (c) K[Ru(Hcdta)Cl]·H₂O, 1 mM and N₂H₄Ph⁺, 100 mM. The pH in each case was maintained at 2.8.

All the complexes, in the presence of 100 equivalents of N₂H₅⁺, exhibited two well resolved cathodic peaks (p₁ and p₂ corresponding to the polarographic waves w₁ and w₂, respectively) and a single anodic peak (p₃, counterpart of p₂). The anodic counterpart of p₁ did not appear under the present experimental conditions. Representative CV responses of complex 4 are shown in Fig. 4(b). The complexes 1–3 exhibited two well resolved cathodic peaks (p₁ and p₂) in the presence of 100 equivalents of N₂H₄Ph⁺. Contrarily, they showed a single anodic peak (p₄) as counterpart of p₁ and no peak for p₂. Complexes 4–8 behaved similarly in the presence of 100 equivalents of N₂H₄Ph⁺. Additionally, they showed a well defined cathodic peak and its counterpart (Fig. 4(c)) corresponding to the additional wave observed in sampled dc between w₁ and w₂. The peak analyses data for p₁ and p₂, confirmed them to be a two- and one-electron reduction process, respectively. Peak p₁ shifted anodically with increase in [N₂H₄X⁺] but cathodically with increase in scan speed, but plots of i_pcvs.ν^1/2 were linear for both p₁ and p₂.

Mechanism of electrode reactions

The complexes 1–8 were reduced in solution in the absence and in the presence of one equivalent of N₂H₄X⁺ at pH 1.9 and 2.8 by holding the potential at a mercury pool cathode at 0.1 V away from the E_1/2 of Ru^III → Ru^II wave as given in Table 4. In the absence, all the complexes consumed one faraday of charge and produced the corresponding ruthenium(II) complexes. In the presence of N₂H₄X⁺ all the complexes consumed three faradays. The electrolysed solutions contained an ammonia concentration two times that of [N₂H₅⁺] but equal to that of [N₂H₄Ph⁺] taken initially. On the basis of these data and the data in Figs. 3 and 4, it is proposed that the Ru^III–N₂H₄X⁺ complexes take two electrons from the electrode at w₁ and produce Ru^I–N₂H₄X⁺ species (reaction (4), charges on the species are omitted for clarity). These ruthenium(I) species are less stable and hence rapidly decompose as shown in reaction (5) to Ru^III–NH₂X and one mole of NH₃ by abstracting one equivalent of H⁺ from the medium. The species Ru^III–NH₂X are reduced by one electron in a subsequent step at w₂ (reactions (6) and (10)) to give Ru^II–NH₂X which on aquation produces the corresponding Ru^II–H₂O species and NH₂X (reactions (8) and (12)), or hydrolyse to Ru^III–OH₂ producing an equivalent amount of NH₂X (reactions (7) and (11)). The Ru^III–OH₂ species thus produced may be reduced along with Ru^III–NH₂X at w₂ in the case of N₂H₅⁺ (reaction (9)) or at potentials where a new reduction step appeared between w₁ and w₂ in the case of N₂H₄Ph⁺ (reaction (13)). The Ru^II–NH₂Ph complexes produced at w₁ are not as stable as the corresponding Ru^III–NH₃ due to the presence of the Ph group and hence these species and their reduced forms Ru^II–NH₂Ph quickly hydrolyse to the corresponding aqua species and NH₂Ph (reactions (11) and (12)). This may be the reason for the irreversible nature of wave w₂ and reversible nature of the poorly resolved new wave between w₁ and w₂.

Ru^III–N₂H₄X⁺ + 2e⁻ ⇌ Ru^I–N₂H₄X⁺ (4)

Ru^I–N₂H₄X⁺ + H⁺ → Ru^IIINH₂X + NH₃ (5)

When X = H

Ru^III–NH₃ + e⁻ ⇌ Ru^II–NH₃ (6)

Ru^III–NH₃ + H₂O → Ru^III–H₂O + NH₃ (7)

Ru^II–NH₃ + H₂O → Ru^II–H₂O + NH₃ (8)

Ru^III–H₂O + e⁻ ⇌ Ru^II–H₂O (9)

When X = Ph

Ru^III–NH₂Ph + e⁻ ⇌ Ru^II–NH₂Ph (10)

Ru^III–NH₂Ph + H₂O → Ru^III–H₂O + NH₂Ph  (11)

Ru^II–NH₂Ph + H₂O → Ru^II–H₂O + NH₂Ph  (12)

Ru^III–H₂O + e⁻ ⇌ Ru^II–H₂O (13)

Electrolytic reduction of N₂H₄X⁺ in the presence of complexes 1–8

The hydrazines N₂H₄X⁺ were electrolytically reduced at pH 1.9 and 2.8 under Ar by constant potential coulometry by holding the potential at 50 mV away from the corresponding E_1/2 of w₁ given in Table 5 in order to measure the catalytic ability of the ruthenium as a function of the sigma basicity of the aminopolycarboxylic acid on the reduction of the N–N bond in N₂H₄X⁺. In all cases the constant reduction of N₂H₄X⁺ to NH₃ and NH₂X occurred, eqn. (14), justifying the wave w₁ as a multi-electron process.

Table 5 Turnover number, moles of NH₃ produced per mol of complex per hour, during the reduction of N₂H₄X⁺

	Turnover number
	N2H5^+			N2H4Ph^+
Complex	Potential/V	pH 1.9	pH 2.8	Potential/V	pH 1.9	pH 2.8
1	−0.350	1.44	2.54	−0.400	0.94	1.10
2	−0.200	1.76	4.12	−0.300	1.02	1.76
3	−0.350	2.98	6.06	−0.400	1.78	2.98
4	−0.200	3.75	7.18	−0.250	1.62	2.95
5	−0.050	9.50	18.40	−0.200	2.85	5.98
6	−0.100	7.96	14.84	−0.220	2.68	5.32
7	−0.175	4.46	8.98	−0.300	2.17	3.84
8	−0.100	1.34	2.95	−0.230	0.92	1.27

---

At both pH studied, the number of mols of NH₃ produced per mol of complex was linear, in all cases, as seen in Fig. 5 for N₂H₅⁺ reduction. The turnover number and mols of NH₃ formed per mol of complex per hour are given in Table 5. For a given substrate, the turnover number at pH 2.8 is nearly double that at pH 1.9. The observed turnover number of NH₃ obtained with N₂H₄Ph⁺ is less than half of that achieved with N₂H₅⁺. However, the coulombic efficiency in all these cases was almost 100%. The solution at the end of electrolysis showed the same voltammetric features as described above and produced NH₃ at the same turnover rate when an additional amount of N₂H₄X⁺ was added to the electrolysed solution and the electrolysis continued after adjusting the pH to the initial value. Thus, the catalytic efficiency of the metal was shown to be intact during the reduction cycle.

Fig. 5 Plots of mols of NH₃ produced per mol of complex (1–8) during the reduction of N₂H₅⁺ at pH 2.8 vs. electrolysis time.

The turnover rate is correlated in Fig. 6 with the sigma basicity of the polyaminopolycarboxylic acid. The data revealed that the turnover rate increases with the increase in sigma basicity of the aminopolycarboxylic acid to nearly ΣpK_a = 22 and then decreases thereafter. With the given hydrazine N₂H₅⁺/N₂H₄Ph⁺, it increases in the order 1 < 2 < 3 < 4 < 5 ≈ 6 as the sigma basicity of the aminopolycarboxylic acid increases in the order 2 < 1 < 3 < 4 < 5 < 6. The lower turnover rate for complex 1 than that of 2 is explained by its high hydrolytic tendency. The decrease in the turnover rate with 5–8 with the increase in the sigma basicity may be related to the structural constraints developed within the aminopolycarboxylic acid due to increase in size, affecting the overall catalytic ability of the ruthenium.

Fig. 6 Plot of turnover number of NH₃ produced by the reduction of N₂H₄X⁺vs. ΣpK_a of the aminopolycarboxylic acid: (a) N₂H₅⁺, (b) N₂H₄Ph⁺.

Conclusion

Ruthenium(III)–aminopolycarboxylic acid complexes of tri- (imda, himda), tetra- (himda, nta), penta- (hedtra), hexa- (edta, pdta and cdta) and octa- (dtpa) dentate ligands were prepared and characterized. Interaction of these complexes with hydrazines N₂H₄X⁺ (X = H or Ph) was investigated by pH-metry, voltammetry and spectrophotometry. The results confirmed that these complexes rapidly form monomeric adducts with N₂H₄X⁺ by its substitution at the less hydrolysable metal co-ordination site in acidic media. These adducts readily accept two electrons to form the corresponding Ru^I–N₂H₄X⁺ which on decomposition give one mol of ammonia and an electroactive species which in a subsequent chemical step gives one mol of NH₂X. Employing these hydrazinium adducts, N₂H₄X⁺ were reduced electrolytically. Ammonia and/or amine was obtained in catalytic amounts with nearly 100% coulombic efficiency. The turnover rate of ammonia produced per mol of complex per hour was evaluated and correlated with the sigma basicity (ΣpK_a) of the aminopolycarboxylic acid. The data revealed that the turnover rate increased with increase in sigma basicity in the order imda < himda < nta < hedtra < edta ≈ pdta. Further increase in the sigma basicity of the aminopolycarboxylic acid reduced the catalytic ability of the ruthenium which is attributed to the development of complexity caused by the concomitant increase in size of the aminopolycarboxylic acid. The most probable chemical and charge transfer steps were suggested for the reactions at the electrode.

Acknowledgements

The authors are grateful to the Department of Science and Technology (No. SP/S1/F-12/93) New Delhi for supporting this work. R. P. is grateful to the Council of Scientific and Industrial Research (CSIR), New Delhi for awarding a Senior Research Fellowship (31/28(22)/97 EMR-I and 31/28(29)/99 EMR-I).

References

1. H. Dalton and L. E. Mortenson, Bacteriol. Rev., 1972, 21, 29.
2. J. Chatt, J. R. Dilworth and R. L. Richards, Chem. Rev., 1978, 78, 589; P. Pelikan and R. Boca, Coord. Chem. Rev., 1984, 55, 55.
3. R. A. Handerson, G. J. Leigh and C. J. Pickett, Adv. Inorg. Chem. Radiochem., 1983, 27, 198.
4. B. K. Burgess and D. J. Lowe, Chem. Rev., 1996, 96, 2983; J. B. Howard and D. C. Rees, Chem. Rev., 1996, 96, 2965.
5. D. Sellmann and J. Sutter, Acc. Chem. Res., 1997, 30, 460.
6. B. E. Smith, M. C. Durrant, S. A. Fairhurst, C. A. Gormal, K. L. C. Gornburg, R. A. Henderson, S. K. Ibrahim, T. Le. Gall and C. J. Pickett, Coord. Chem. Rev., 1999, 185–186, 669.
7. R. N. F. Thorneley, R. R. Eady and D. J. Lowe, Nature (London), 1978, 272, 557.
8. B. T. Heaton, C. Jacob and P. Page, Coord. Chem. Rev., 1996, 154, 193 and references therein..
9. P. B. Hitchock, D. L. Hughes, M. J. Maguire, K. Marjani and R. L. Richards, J. Chem. Soc., Dalton Trans., 1997, 4747.
10. J. V. Barkely, B. T. Heaton, C. Jacob, R. Mageswaran and J. T. Sampanthar, J. Chem. Soc., Dalton Trans., 1998, 697.
11. T. Furuhashi, M. Kawano, Y. Koide, R. Somazawa and K. Matsumoto, Inorg. Chem., 1999, 38, 109.
12. R. R. Schrock, T. E. Glassman, M. G. Vale and M. Kol, J. Am. Chem. Soc., 1993, 115, 1760.
13. S. M. Malinak, K. D. Demadis and D. Coucouvanis, J. Am. Chem. Soc., 1995, 117, 3126.
14. T. A. George, D. N. Kurk and J. Redepenning, Polyhedron, 1996, 15, 2377.
15. G. N. Schrauzer, F. R. Robinson, E. L. Moorehead and J. M. Vickery, J. Am. Chem. Soc., 1976, 98, 2815.
16. Y. Hozumi, Y. Imasaka and T. Tanaka, Chem. Lett., 1983, 897.
17. S. Cai and R. R. Schrock, Inorg. Chem., 1991, 30, 4106.
18. S. Kuwata, Y. Mizobe and M. Hidai, Inorg. Chem., 1994, 33, 3619.
19. K. D. Demadis and D. Coucouvanis, Inorg. Chem., 1995, 34, 3658; K. D. Demadis, S. M. Malinak and D. Coucouvanis, Inorg. Chem., 1996, 35, 4038; S. M. Malinak, A. M. Simeonov, P. E. Mosier, C. E. Mckenna and D. Coucouvanis, J. Am. Chem. Soc., 1997, 119, 1662.
20. R. E. Palermo, R. E. Singh, J. K. Bashkin and R. H. Holm, J. Am. Chem. Soc., 1984, 106, 2600.
21. G. Ramachandraiah, J. Am. Chem. Soc., 1994, 116, 6733.
22. R. Prakash, B. Tyagi, D. Chatterjee and G. Ramachandraiah, Polyhedron, 1997, 12, 1235.
23. R. Prakash and G. Ramachandraiah, Stud. Surf. Sci. Catal., 1998, 113, 519; D. Chatterjee, Coord. Chem. Rev., 1998, 168, 273.
24. R. Prakash and G. Ramachandraiah , J. Mol. Catal. A, Chem., in the press..
25. E. R. Mercer and R. R. Buckley, Inorg. Chem., 1965, 4, 1962.
26. T. Matsubara and C. Creutz, Inorg. Chem., 1979, 18, 1956; J. Am. Chem. Soc., 1978, 100, 6255..
27. A. A. Diamantis and J. V. Dubrawski, Inorg. Chem., 1981, 20, 1142; M. M. Taqui Khan, A. Kumar and Z. Shirin, J. Chem. Res. (S), 1986, 5, 1003.
28. M. M. Taqui Khan, M. A. Moiz and A. Hussain, J. Coord. Chem., 1991, 23, 245; M. M. Taqui Khan and R. M. Naik, J. Mol. Catal., 1989, 54, 139.
29. M. M. Taqui Khan, A. Hussain, G. Ramachandraiah and M. A. Moiz, Inorg. Chem., 1986, 25, 3023; M. M. Taqui Khan and A. Hussain, Indian J. Chem., Sect. A, 1979, 19, 50; CRC Handbook of Chemistry and Physics, eds. R. C. Weast and M. J. Astle, CRC Press, Boca Raton, FL, 1979, p. D-168..
30. A. E. Martell and M. Calvin , in Chemistry of Metal Chelate Compounds, Prentice-Hall, New York, 1952..
31. L. Meites , in Polarographic Techniques, Interscience, New York, 2nd edn., 1957, pp. 95–301..
32. A. J. Bard and L. R. Faulkner , in Electrochemical Methods Fundamentals and Applications, Wiley, New York, 1980..
33. V. N. Alexeyev , in Qualitative Chemical Semimicro Analysis, ed. P. K. Agasyan, Mir Publishers, Moscow, 1975, p. 475..
34. J. F. Endicott and H. Taube, Inorg. Chem., 1965, 4, 437; K. Shimzu, T. Matsubara and G. P. Sato, Bull. Chem. Soc. Jpn., 1974, 47, 1651.
35. P. D. Maria Mattioli and A. A. Luiz Oliveria, Polyhedron, 1987, 6, 603.
36. H. Ogino, T. Watanabe and N. Tanaka, Inorg. Chem., 1975, 14, 2093; H. Ogino, M. Shimura and N. Tanaka, Inorg. Chem., 1979, 18, 2497.
37. A. E. Martell and R. M. Smith , in Critical Stability Constants, Plenum, New York, 1982, vols. 1–5..

---