Selective NH₃-to-N₂H₄ conversion electrocatalysed by ruthenium(II)-cymene complexes

Abstract

A series of ruthenium(II)-cymene complexes [(η⁶-p-cymene)Ru(pp)Cl] (1–4) and corresponding NH₃-ligated complexes [(η⁶-p-cymene)Ru(pp)(NH₃)]PF₆ ([1-NH₃]PF₆ to [4-NH₃]PF₆), where cymene = 4-isopropyltoluene and pp⁻ = pyridylpyrrole ligand, have been designed and synthesized. Structural modifications of pp⁻ ligands are accomplished through the use of an increasing number of electron-donating methyl groups on the pyrrole unit. Solid-state structural analysis shows that these complexes have a typical piano-stool structure. Electrochemical studies of these complexes illustrate that the introduction of a methyl group to the pp⁻ ligand can greatly decrease the oxidation potential of Ru^III/II from 0.49 V vs. Cp₂Fe^+/0 for [1-NH₃]PF₆ to 0.16 V vs. Cp₂Fe^+/0 for [4-NH₃]PF₆. Controlled potential coulometry experiments show that these complexes exhibit selective catalysis for the oxidation of NH₃ to N₂H₄ with a turnover number of up to 453.2 at E_app 0.8 V vs. Cp₂Fe^+/0 for the [4-NH₃]PF₆ complex. Kinetic and theoretical thermodynamic studies show that the pathway of bimolecular coupling of Ru^II-aminyl species and the pathway of ammonia nucleophilic attack of Ru^IV-imide (generated from the disproportionation of Ru^III-amide) are involved in N–N formation.

---

To meet the challenges of a large-scale energy crisis and environmental pollution, energy-rich H₂ of a green and sustainable nature has attracted much interest as an alternative energy source, although storage and distribution of liquid H₂ still suffer from harsh conditions along with a lack of infrastructure. Ammonia (NH₃) is a good candidate as a hydrogen energy carrier, and offers approximately 1.7 times the energy density of liquid H₂.^1–4 The NH₃ industry has not only seen applications worldwide but also has widespread facilities for storage, transport and handing. The classical catalytic cracking reaction for NH₃-to-H₂ conversion (eqn (1)) requires a precious metal catalyst and high temperature, which are of relatively high cost. Moreover, the generated N₂ product is directly discharged into the air, resulting in low atomic utilization and relatively poor economy in NH₃-to-H₂ conversion.

2NH₃(g) → N₂(g) + 3H₂(g) ΔG^θ = 7.9 kcal mol⁻¹ (1)

2NH₃(g) → N₂H₄(l) + H₂(g) ΔG^θ = 43.6 kcal mol⁻¹ (2)

N₂H₄(l) → N₂(g) + 2H₂(g) ΔG^θ = −35.7 kcal mol⁻¹ (3)

Selective electrocatalytic conversion of NH₃ into N₂H₄ and H₂ (NH₃-to-N₂H₄ conversion, eqn (2)) seems more appealing than NH₃-to-H₂ conversion due to the advantages of not only generating H₂, but also simultaneously producing high value-added N₂H₄ (the price of anhydrous hydrazine is about 58 000 USD t⁻¹). However, this route is a thermodynamically demanding process (ΔG^θ = 43.6 kcal mol⁻¹), and needs to overcome the competitive reaction of spontaneous dehydrogenation of N₂H₄ to N₂ (eqn (3), ΔG^θ = −35.7 kcal mol⁻¹). Hence, highly efficient and selective NH₃-to-N₂H₄ conversion is appealing, but remains a huge scientific challenge.⁵

Molecular catalysts can offer several advantages over their heterogeneous counterparts, such as controllable structure, convenient characterization, and well-defined active site nature, which allow for mechanistic studies to elucidate the factors controlling the catalytic activity and selectivity. Since the seminal work by Hamann and Smith III on the electro-oxidation of NH₃ to N₂ catalyzed by [(trpy)(bpy^NMe2)Ru(NH₃)](PF₆)₂ (trpy = 2,2′:6′,2″-terpyridine, bpy^NMe2 = 4,4′-bis(dimethylamino)-2,2′-bipyridine),⁶ significant progress has been made in the development of molecular catalysts for ammonia oxidation.^7–22 Nevertheless, the reported catalytic systems are mostly concerned with the oxidation of NH₃ to N₂, and there are few reports on the selective catalytic conversion of NH₃ into N₂H₄. In 2023, we demonstrated that [Ru(κ²-N,N′-dpp)(bpy)(dmso)(NH₃)]PF₆ (Hdpp = 2,5-di(pyridin-2-yl)-1H-pyrrole)^15a can catalyze the electrocatalytic conversion of NH₃ into N₂H₄ with unprecedentedly high selectivity (over 97.9%) and turnover frequency (238.9 h⁻¹). Although a similar ligated-N₂H₄ intermediate and a similar N–N formation pathway—such as either bimolecular coupling of a metal-imide^15,18,19,20a or ammonia nucleophilic attack of a metal-imide^6,10,17,22—are involved, this catalyst is in sharp contrast with conventional ones that usually generate N₂ as the major N–N coupling product with relatively low turnover frequencies. N₂H₄/N₂ selectivity is usually represented by a branch from a ligated-N₂H₄ intermediate,^14,15a which could directly release N₂H₄ or continue to be over-oxidized to release N₂. The release of N₂H₄ from a ligated-N₂H₄ intermediate to restart the catalytic cycle is one of the key issues for the selective oxidation of NH₃ to N₂H₄. Thus, rational design of the ancillary ligand backbone is still a desirable strategy for developing highly efficient and selective catalysts for NH₃-to-N₂H₄ conversion.

We have long been interested in the study of metal complexes based on the pyridylpyrrole (pp⁻) ligand, which is structurally analogous to bipyridine.²³ Notwithstanding the similar structure and coordination properties, the π-donation from the pyrrolyl group in the pp⁻ ligand increases the energy of the metal-based LUMO in the M–N₂H₄ intermediate, thus weakening the M–N₂H₄ bond and facilitating the release of N₂H₄.²⁴ In addition, the negative charge of the pyrrolyl unit not only lowers the overpotential of the metal complex but also reduces the overall positive charge of the reactive intermediate, which is a major cause of instability in the intermediates during the AO catalytic cycle.²³ Herein, we design an ancillary pp⁻ ligand using an increasing number of electron-donating methyl groups on the pyrrole unit to regulate the electronic structure of the metal complex. Corresponding ruthenium(II)-cymene complexes [(η⁶-p-cymene)Ru(pp)Cl] (where pp⁻ is HL1 = 2-(1H-pyrrol-2-yl)pyridine (1), HL2 = 2-(4-methyl-1H-pyrrol-2-yl)pyridine (2), HL3 = 2-(3,5-dimethyl-1H-pyrrol-2-yl)pyridine (3), and HL4 = 2-(3,4,5-trimethyl-1H-pyrrol-2-yl)pyridine (4)) and corresponding NH₃-ligated complexes [(η⁶-p-cymene)Ru(pp)(NH₃)]PF₆ ([1-NH₃]PF₆ to [4-NH₃]PF₆) are reported. Their selective catalysis for NH₃-to-N₂H₄ conversion and the catalytic mechanism are also presented.

As shown in Fig. 1a, complexes 1–4 are synthesized by treatment of dimeric precursor of [(η⁶-p-cymene)RuCl₂]₂ and a deprotonated pyridylpyrrole ligand in CH₂Cl₂ at 0 °C in moderate yield (∼50%). The corresponding NH₃-ligated complexes [(η⁶-p-cymene)Ru(pp)(NH₃)]PF₆ ([1-NH₃]PF₆ to [4-NH₃]PF₆) are synthesized in over 60% yield by treatment of 1–4 with one equiv. of AgPF₆ in CH₃CN followed by bubbling NH₃ gas. These complexes are stable in common organic solvents, and are fully characterized by ESI-MS, NMR and IR spectroscopy (Fig. S6–S29). Compared to 1–4, the ¹H NMR spectra of [1-NH₃]PF₆ to [4-NH₃]PF₆ show a newly added broad single peak at ∼2.0 ppm due to the ligated-NH₃. Similarly, the IR spectra of [1-NH₃]PF₆ to [4-NH₃]PF₆ show an additional band at ∼3340–3350 cm⁻¹ due to N–H stretching, also indicating that NH₃ binds to the Ru center. The solid-state structures of 1–4 and [2-NH₃]PF₆ are shown in Fig. 1b. The crystallographic data and selected bond distances and angles are listed in Tables S1–S6. These complexes exhibit a typical piano-stool structure with one pp⁻, one cymene and one Cl⁻ (or NH₃) ligand coordinating to the ruthenium center. The bond distance of Ru–N_pyrrole (2.042(10)–2.074(3) Å) is slightly shorter than that of Ru–N_pyridine (2.101(2)–2.113(3) Å), mainly due to the electrostatic interaction between the Ru atom and the anionic pyrrolide N atom. In [2-NH₃]PF₆, NH₃ coordinates to the Ru center with a Ru–N bond distance of 2.138(12) Å, which is similar to that of other NH₃-ligated Ru(II) complexes.

Fig. 1 (a) The synthetic routes of 1–4, and [1-NH₃]PF₆ to [4-NH₃]PF₆. (b) Solid-state structures of complexes 1–4 and [2-NH₃]⁺. The hydrogen atoms on cymene, the pp⁻ ligand and anionic PF₆⁻ in [2-NH₃]PF₆ are omitted for clarity.

The electrochemical behavior of the title complexes in CH₃CN is investigated by cyclic voltammetry (CV) and differential pulse voltammetry (DPV) using a AgCl/Ag electrode in a saturated KCl solution as a reference electrode. Unless otherwise specified, all potentials are converted into E_1/2 vs. Cp₂Fe^+/0 in CH₃CN by adding −0.43 V to the measured potential.

The electrochemical behavior of 1–4 is shown in Fig. 2a. The first oxidation wave is assigned to ruthenium center oxidation (Ru^II → Ru^III). The electro-donating nature of the pp⁻ ligand significantly influences the redox potential of these complexes, leading to a decreasing Ru^III/II reduction potential from 0.37 V for 1 to 0.04 V for 4 with an increase in the number of electron-donating methyl groups on the pyrrole unit. The redox potentials of the second oxidation wave at ∼1.2 V and third oxidation wave at ∼1.6 V (see the DPV curves in Fig. 2a) are independent of the methyl substituted pp⁻ ligand. These could clearly be assigned to Cl⁻ and cymene ligand oxidation events, respectively.^25,26 Fig. 2b shows that [1-NH₃]PF₆ to [4-NH₃]PF₆ exhibit two oxidation waves. The first one (0.49 V, 0.36 V, 0.23 V and 0.16 V for [1-NH₃]PF₆ to [4-NH₃]PF₆, respectively) is attributed to their metal center oxidation from Ru^II to Ru^III. Compared to 1–4, the Ru^III/II redox potentials of [1-NH₃]PF₆ to [3-NH₃]PF₆ are anodically shifted by ∼0.12 V, mainly due to the π-donating capability of the Cl⁻ ligand in the former. Upon expanding the voltage window, unlike the second oxidation peak corresponding to Ru^IV/III reported in our previous literature,¹⁵ these complexes exhibited no additional metal-centered oxidation waves except for the second oxidation peak assigned to cymene ligand oxidation at ∼1.6 V.²⁶ However, a weak new wave marked by a rhombus appears at 0.75–0.88 V when scanning in the cathodic direction (Fig. 2c). This suggests that a disproportionation of Ru^III species possibly occurs to give Ru^II and Ru^IV species, and then Ru^IV is reduced to Ru^III on the reverse scan. Taking [1-NH₃]PF₆ as an example, its Ru^III intermediate formed by 1e⁻ oxidation is proposed to undergo rapid disproportionation to Ru^II [1-NH₃]⁺ and a Ru^IV imido intermediate and simultaneously reach equilibrium. Thus, the wave at 0.88 V on the reverse scan can be assigned to the reduction of Ru^IV species to Ru^III species of [1-NH₃]⁺. To further confirm the 1e⁻ oxidation to form a Ru^III intermediate, CV was employed, with ferrocene added as the internal reference at an equimolar concentration to [1-NH₃]⁺. As shown in Fig. S30, the diffusion coefficient of complex [1-NH₃]⁺ (1.58 × 10⁻⁵ cm² s⁻¹) is very close to that of ferrocene (1.37 × 10⁻⁵ cm² s⁻¹), and the peak areas of [1-NH₃]⁺ and ferrocene are almost the same in their respective CV plots, preliminarily indicating that the first oxidation wave of [1-NH₃]⁺ involved only one electron transfer. The more critical evidence in the reversible charge transfer process is the slope analysis of E vs. lg[(I_l − I)/I] (where E is the potential and I_l is the limiting current).²⁷ As shown in Fig. S31, a linear relationship was observed between E and lg[(I_p − I)/I] (where I_p is the peak current, used in place of I_l due to the irreversible oxidation wave of [1-NH₃]⁺), with a slope m = 2.3RT/nF (where R is the ideal gas constant, T = 298.15 K, and F = 96 485C mol⁻¹). The fitted slope from the experimental data was 0.84, yielding an electron transfer number n ≈ 0.84. These results collectively confirm that only a single electron transfer occurs at the electrode surface, followed by a redox disproportionation chemical step.

Fig. 2 Electrochemical behavior of the title complexes in CH₃CN solution. (a) CV (solid lines) and DPV (dotted lines) plots of 1–4; (b) CV (solid lines) and DPV (dotted lines) plots of [1-NH₃]PF₆ to [4-NH₃]PF₆; and (c) DPV plots of [1-NH₃]PF₆ to [4-NH₃]PF₆ in the anodic direction (dotted lines) and cathodic direction (solid lines). Conditions: [Ru] = 1 mM, scan rate 0.1 V s⁻¹ for CV and 0.008 V s⁻¹ for DPV, 0.1 M Bu₄NPF₆ as the supporting electrolyte, glassy carbon as the working electrode, platinum wire as the counter electrode, AgCl/Ag in saturated KCl aqueous solution as the reference electrode.

The electrochemical behavior of [1-NH₃]PF₆ to [4-NH₃]PF₆ in the presence of 0.05 M NH₃ is shown in Fig. 3. Complexes [1-NH₃]PF₆ and [2-NH₃]PF₆ display a slightly increased catalytic current (i_cat) over Ru^III species with low onset potentials of 0.15 V and 0.10 V, respectively (inset figures of Fig. 3). Although Ru^IV species of the title complexes are not clearly observed in the CV studies, the catalytic current increases sharply at potentials over 0.6 V. For complexes [3-NH₃]PF₆ and [4-NH₃]PF₆, a catalytic current only appears at a high potential, and there is no catalytic current over Ru^III species. Anodic currents (i_p) of the Ru^III/II redox couple and the catalytic current of [1-NH₃]PF₆ to [4-NH₃]PF₆ in the presence of 0.05 M NH₃ increase linearly with the square root of the scan rate (Fig. S33), indicating diffusion-controlled behavior of these ruthenium catalysts under test conditions. The rate constant (k_cat) and maximum turnover frequency (TOF_max) of [1-NH₃]PF₆ to [4-NH₃]PF₆ for ammonia oxidation are estimated. The diffusion coefficients (D_Ru) and rate constants (k_cat) of [1-NH₃]PF₆ to [4-NH₃]PF₆ are 1.58 × 10⁻⁵ cm² s⁻¹, 1.71 × 10⁻⁵ cm² s⁻¹, 1.82 × 10⁻⁵ cm² s⁻¹, and 1.94 × 10⁻⁵ cm² s⁻¹; and 4.5 s⁻¹, 4.9 s⁻¹, 5.2 s⁻¹, and 5.4 s⁻¹, respectively, which are determined from the dependence of i_p and ν^1/2 based on the Randles–Sevcik relation (eqn (4)) and linear fitting of i_cat/i_p with ν^−1/2 based on eqn (5), respectively. The TOF_max values at a scan rate of 0.1 V s⁻¹ based on eqn (6) are estimated to be 4.92 × 10⁻² s⁻¹, 5.48 × 10⁻² s⁻¹, 5.81 × 10⁻² s⁻¹, and 6.29 × 10⁻² s⁻¹.

Fig. 3 CV plots of (a) [1-NH₃]PF₆, (b) [2-NH₃]PF₆, (c) [3-NH₃]PF₆ and (d) [4-NH₃]PF₆ solution with/without 0.05 M NH₃/NH₄PF₆ in CH₃CN. Conditions: [Ru] = 1 mM, scan rate 0.1 V s⁻¹, 0.1 M Bu₄NPF₆ as the supporting electrolyte, glassy carbon as the working electrode, platinum wire as the counter electrode, AgCl/Ag in saturated KCl aqueous solution as the reference electrode. Insets: CVs. in the voltage window of −0.25–1.2 V.

To confirm the electrocatalysis of [1-NH₃]PF₆ to [4-NH₃]PF₆ for NH₃-to-N₂H₄ conversion, controlled potential coulometry (CPC) experiments are conducted in a conventional sealed three-electrode cell with 2.0 M NH₃, 0.01 mM catalyst and 0.1 M [n-Bu₄N]PF₆ electrolyte in dried CH₃CN under an Ar atmosphere. Carbon cloth (1 cm²), a Pt plate (1 cm²) and Ag/AgCl in saturated KCl aqueous solution are used as the working electrode, counter electrode and reference electrode, respectively. Possible gas products (such as N₂, H₂) are determined by the GC method (Fig. S1). Possible products in the electrolyte (such as N₂H₄, NO₂⁻, NO₃⁻) are quantitatively analyzed via chemical methods (Fig. S2–S5).^28–30

To obtain general information for the CPC experiments, first, control experiments are employed. As shown in Table S7, only negligible N₂H₄ and N₂ are generated at low applied potential (such as E_app 0.39 V) for 24 h. However, 21.6 µmol of N₂H₄ and 0.5 µmol of N₂ are produced at E_app 0.8 V for 2 h, indicating that the bare electrode causes slow ammonia oxidation at relatively high potential. Subsequently, the complex [2-NH₃]PF₆ is chosen to investigate the relationship between catalytic performance over time and applied potential (Table 1 and Fig. S38, S39, Tables S8, S9). The amounts of N₂H₄ as an absolutely dominant anodic product and H₂ as a cathodic product increase with an increase of the applied potential from 0.2 V to 1.0 V (Fig. S38). Other anode products of NO₂⁻ and NO₃⁻ are not determined due to the absence of oxygen sources in the catalytic system. Holding the applied potential at 0.8 V, the generation of N₂H₄ and H₂ keeps increasing over time, however, the turnover frequency (TOF) and Faraday efficiency (FE) of N₂H₄ formation continuously decrease (Fig. S39 and Table S9). The loss of TOF and FE for long-term electrolysis might be caused by over-oxidation of the produced H₂ in the sealed CPC cell. A decrease in i_cat under a H₂ atmosphere indicating the current consumption of H₂ oxidation cannot be ignored (Fig. S40). The i_cat remains almost unchanged after 100 consecutive CV cycles or electrolysis for 2 h, indicating that these Ru catalysts under catalytic conditions have satisfactory stability (Fig. S41 and S42). A thoroughly rinsed electrode after catalysis shows no ruthenium deposition and no catalytic activity, indicating solution-based electrocatalysis (Fig. S43–4S5 and Table S7). Notably, N₂H₄ is prone to undergo either catalytic oxidation or disproportionation decomposition to yield N₂. In contrast, complexes [1-NH₃]PF₆–[4-NH₃]PF₆ exhibit good selectivity toward N₂H₄ formation. CV measurements of [1-NH₃]PF₆ in MeCN containing N₂H₄ confirm that they are inactive toward N₂H₄ oxidation (Fig. S54).

Table 1 The electrocatalytic performances of [1-NH₃]PF₆ to [4-NH₃]PF₆ in CH₃CN^a

Entry	Cat	[NH3] (M)	Eapp	Time (h)	TONH2/nH2(µmol)	TOFH2 (h^−1)	TONN2H4/nN2H4(µmol)	TOFN2H4 (h^−1)	TONN2/nN2(µmol)	TOFN2 (h^−1)	FEN2H4^b (%)	SN2H4^c (%)
a Conditions: carbon cloth (1 cm^2) as the working electrode, a platinum plate (1 cm^2) as the counter electrode, AgCl/Ag in saturated KCl aqueous solution as the reference electrode, [cat] = 0.01 mM. The background is subtracted to obtain the number of moles of the products H2, N2 and N2H4.b FEN2H4 = 2 nN2H4F/it × 100%.c SN2H4 = nN2H4/(nN2 + nN2H4) × 100%.
1	[1-NH3]PF6	2.0	0.39	24	17.2	0.7	16.5	0.7	Trace	—	86.3	100
					13.7		13.2
2	[2-NH3]PF6	2.0	0.26	24	11.5	0.5	11.0	0.5	Trace	—	89.7	100
					9.2		8.8
3	[3-NH3]PF6	2.0	0.13	24	Trace	—	Trace	—	Trace	—	—	—
4	[4-NH3]PF6	2.0	0.06	24	Trace	—	Trace	—	Trace	—	—	—
5	[1-NH3]PF6	2.0	0.8	2	377.3	188.7	356.2	178.1	5.3	2.7	88.2	98.6
					301.9		284.9		4.2
6	[2-NH3]PF6	2.0	0.8	2	403.5	201.8	393.9	197.0	2.2	1.1	90.8	99.4
					322.8		315.2		1.8
7	[3-NH3]PF6	2.0	0.8	2	430.1	215.1	418.2	209.1	4.3	2.2	93.5	99.0
					344.1		334.5		3.4
8	[4-NH3]PF6	2.0	0.8	2	461.6	230.8	453.2	230.8	2.4	1.2	91.9	99.5
					369.3		362.5		1.9
9	[4-NH3]PF6	0.05	0.8	2	156.6	78.3	151.0	75.5	1.5	0.8	87.6	99.0
					125.3		120.8		1.2
10	5 (ref. 15a)	2.0	1.0	24	5870	244.6	5735	238.9	43.9	1.8	36.6	97.8
					2348		2293.8		17.56
11	6 (ref. 15b)	2.0	1.0	2	801.5	400.8	721.5	360.8	13.7	6.9	99.2	98.1
					561.1		505.1		9.6
12	7 (ref. 15c)	2.0	1.0	2	844.2	422.1	786.8	393.4	19.8	9.9	99.3	93.1
					616.2		566.4		14.3

---

To clarify the catalytic performance of the catalysts at the initial stage of Ru^III species formation, we selected a potential 0.1 V lower than the Ru^III/II potential of the catalysts (0.39 V for [1-NH₃]PF₆, 0.26 V for [2-NH₃]PF₆, 0.13 V for [3-NH₃]PF₆, and 0.06 V for [4-NH₃]PF₆) as the applied potential in the CPC experiment. N₂H₄ and almost equiv. molar of H₂ are produced in the [1-NH₃]PF₆ and [2-NH₃]PF₆ catalyst systems, and the generation of N₂ could be negligible (entries 1 and 2). From comparison to the result of the control CPC experiments at E_app 0.39 V, this indicates that the catalytic activity originates from the catalyst but not the bare electrode. The catalytic conversion of NH₃ to N₂H₄ achieves ∼100% selectivity. This behavior also aligns with our earlier ruthenium catalysts containing pyridylpyrrole operated at low applied potentials via bimolecular N–N coupling of Ru^III-amide to form N₂H₄.^15,18,19,20a As expected in the CV studies, no ammonia oxidation products (N₂H₄ or N₂) are generated in the [3-NH₃]PF₆ and [4-NH₃]PF₆ catalyst systems, only over Ru^III species at low potential (entries 3 and 4). Next, we carried out CPC experiments at higher applied potential E_app of 0.80 V. Correspondingly, the catalytic efficiency of [1-NH₃]PF₆ to [4-NH₃]PF₆ significantly increased. As shown in Table 1, entries 5–8, N₂H₄ is still determined as the dominant anodic product with a yield range from 284.9 µmol to 362.5 µmol. The turnover frequency (TOF) of N₂H₄ formation reaches 178.1 h⁻¹ for [1-NH₃]PF₆, 197.0 h⁻¹ for [2-NH₃]PF₆, 209.1 h⁻¹ for [3-NH₃]PF₆ and 226.6 h⁻¹ for [1-NH₃]PF₆. The selectivity and Faraday efficiency based on N₂H₄ formation are over 98.6% and 88.2%, respectively. In this work, [4-NH₃]PF₆ also shows excellent selectivity and high catalytic activity for the oxidation of low-concentration NH₃ to generate N₂H₄. We carried out CPC experiments in a low-concentration NH₃ solution (0.05 M) at the same potential. After 2 h, catalytic amounts of H₂ (125.3 µmol), N₂H₄ (120.8 µmol) and N₂ (1.2 µmol) were generated. This result breaks through the limitation of conventional catalytic systems that rely on high ammonia concentrations to achieve high selectivity (entry 9).

The CV and CPC experiments illustrate that complexes [1-NH₃]PF₆ and [2-NH₃]PF₆ for Ru^III species and [1-NH₃]PF₆ to [4-NH₃]PF₆ for Ru^IV species readily undergo ammonia oxidation to generate N₂H₄. To explore the detailed mechanism of these complexes, we carried out theoretical calculations on the activation energy of each step in the catalytic procedure for the [1-NH₃]PF₆ and [4-NH₃]PF₆ catalysts. A summary of the proposed mechanism and alternative pathway is shown in Fig. 4.

Fig. 4 Calculated pathways of ammonia oxidation catalysed by [1-NH₃]PF₆ and [4-NH₃]PF₆, where the free energy changes (ΔG) of individual steps for [1-NH₃]PF₆ and [4-NH₃]PF₆ are presented in green and orange, respectively. (a) At low potential, hydrazine formation via bimolecular coupling of Ru^III amide; (b) at high potential, hydrazine generation through ammonia nucleophilic attack of Ru^IV imide via redox disproportionation. [Ru^II]⁺ = [(η⁶-p-cymene)Ru(pp)]⁺. ΔG is calculated relative to Cp₂Fe^+/0, with values given in kcal mol⁻¹.

As shown in Fig. 4a, complexes [1-NH₃]⁺ and [4-NH₃]⁺ are first oxidized to Ru^III-amide ²[1-I]⁺ and ²[4-I]⁺ (the left superscript shows the spin state) via a 1e⁻/H⁺ transfer process with ΔG of 8.0 and 6.3 kcal mol⁻¹, respectively. Spin density distribution calculations (Fig. S55) reveal that the spin densities of ²[1-I]⁺ and ²[4-I]⁺ are primarily located on the Ru center (0.50 and 0.47 e⁻) and N atom of NH₂ (0.50 and 0.50 e⁻), indicating that the Ru^III-amide species has Ru^II-aminyl character via inter-conversion of Ru^III–NH₂⁻ ⇋ Ru^II–NH_2͘͘˙. The Ru^II-aminyl species is an active intermediate and can readily generate N₂H₄ via a bimolecular Ru^II–NH₂⁻ coupling reaction.^15,18,19,20a For the [1-NH₃]PF₆ catalyst, theoretical calculations elucidate the coupling process between two ²[1-I]⁺ molecules, which proceeds through transition states ¹[1-III]²⁺ and ¹[1-IV]²⁺ to ultimately form the thermodynamically stable hydrazine-bridged bimetallic complex ¹[1-V]²⁺. The catalytic cycle is completed by a mildly endothermic hydrazine dissociation process, with an overall energy barrier of 12.3 kcal mol⁻¹.

This bimolecular coupling pathway is supported by the CPC experiment of [1-NH₃]PF₆ at low applied potential (0.39 V), where only N₂H₄ as an ammonia oxidation product is observed. In addition, its catalytic current over Ru^III species is independent of the increase of [NH₃] (10–50 mM), also indicating that ammonia seems not to be involved in N₂H₄ formation (Fig. S35) only over Ru^III species. A similar bimolecular coupling pathway to form N₂H₄ over Ru^III species is observed in our reported ruthenium catalysts,¹⁵ and has recently been confirmed in the [(trpy)(bpy^NMe2)Ru(NH₃)](PF₆)₂ catalyst system.¹⁴ Interestingly, in the bimolecular coupling pathway, the energy barrier for the conversion of ¹[4-III]²⁺ to ¹[4-V]²⁺ (17.3 kcal mol⁻¹) is only 5 kcal mol⁻¹ higher than that for ¹[1-III]²⁺ to ¹[1-V]²⁺. However, CV and CPC studies on the structurally analogous complex [4-NH₃]PF₆ demonstrate that its Ru^III species (²[4-I]⁺) cannot trigger ammonia oxidation. This suggests that the difference in energy barrier does not play a decisive role in determining whether N–N bond formation occurs. From a kinetic perspective, an increased number of methyl groups (three methyl groups in [4-NH₃]PF₆) enhances steric hindrance, which disfavors the bimolecular coupling pathway and appears to be the dominant controlling factor.

Following the 1e⁻/H⁺ transfer process, the subsequent second 1e⁻/H⁺ oxidation to form Ru^IV-imide species ¹[1-II]⁺ and ¹[4-II]⁺ is less endergonic with ΔG of 5.3 and 1.5 kcal mol⁻¹, respectively. Compared to the first oxidation step, the significantly reduced ΔG values indicate thermodynamically more favorable formation of Ru^IV-imide species. Analysis of the combined free energy changes reveals that the redox disproportionation of Ru^III-amide to generate Ru^II-ammine and Ru^IV-imide exhibits ΔG values of −2.7 and −4.8 kcal mol⁻¹, respectively, demonstrating a highly spontaneous thermodynamic process. Therefore, the redox disproportionation pathway to form Ru^IV-imide species is thermodynamically preferred over direct oxidation of Ru^III-amide, explaining the absence of an observable Ru^IV/III redox couple under CV test conditions. Notably, the redox disproportionation of ²[4-I]⁺ seems to be more favorable than that of ²[1-I]⁺, which is less exergonic. Subsequently, the Ru^IV-imide species ¹[1-II]⁺ and ¹[4-II]⁺ initiate ammonia oxidation via nucleophilic attack pathways. First, ¹[1-II]⁺ and ¹[4-II]⁺ readily interact with NH₃ to generate ¹[1-VI]⁺ and ¹[4-VI]⁺ intermediates (ΔG = −4.5 and −4.3 kcal mol⁻¹, respectively) due to the formation of a hydrogen bond between the H atom of the imide and the N atom of the approaching NH₃. Subsequently, terminal N₂H₄-ligated ¹[1-IX]⁺ and ¹[4-IX]⁺ are formed through transition states ¹[1-VII]⁺ and ¹[4-VII]⁺ (ΔG^‡ = 5.4 and 6.0 kcal mol⁻¹, respectively) and intermediates ¹[1-Vⁱⁱⁱ]⁺ and ¹[4-Vⁱⁱⁱ]⁺. The formation of ¹[1-IX]⁺ and ¹[4-IX]⁺ N₂H₄-ligated intermediates through the reaction of ammonia and Ru^IV-imides ¹[1-II]⁺ and ¹[4-II]⁺ is highly exergonic by 46.4 and 44.1 kcal mol⁻¹, respectively. Finally, the catalytic cycle is restarted by endergonic evolution of N₂H₄ through N₂H₄-by-NH₃ substitution of ¹[1-IX]⁺ and ¹[4-IX]⁺ (ΔG = 10.2 and 9.8 kcal mol⁻¹, respectively). Unlike other ruthenium molecular catalysts only generating N₂ as the N–N coupling product, the π-donor capability of the pyrrolyl group of the ancillary pp⁻ ligand in the title complexes helps to release N₂H₄ through N₂H₄-by-NH₃ substitution of the N₂H₄-ligated intermediate, thus hindering N₂H₄ overoxidation to generate N₂.²⁴

The kinetic studies of the title complexes also support an ammonia nucleophilic attack route over Ru^IV-imide species. As shown in Fig. S35 and S36, the catalytic current i_cat (at E = 1.2 V) linearly increases with the increase of [NH₃] (0.010–0.050 M) and [cat] (0.2–1.0 mM), clearly indicating that there is a single-site molecular catalytic pathway. Notably, unlike i_cat at 0.05 M NH₃ showing a linear relationship with ν^1/2, the i_cat at 1.0 M NH₃ almost does not change with increasing scan rate, indicating that the i_cat is no longer determined by the bulk diffusion of catalyst or NH₃ but by the rate of regeneration of active Ru^IV-imide species at the electrode.¹¹ This seems to confirm that when ammonia is present in high concentrations, redox disproportionation to generate Ru^IV-imide could be the slow step, and the nucleophilic coupling pathway could prevail, which is consistent with the recent results of mechanism studies of the [(trpy)(bpy^NMe2)Ru(NH₃)](PF₆)₂ catalyst system.¹⁴

In summary, a series of ruthenium(II)-cymene NH₃-ligated complexes are synthesized and fully characterized. By regulating the electronic structure of the ancillary ligand, the oxidation potential of the ruthenium center is gradually reduced from 0.49 V for [1-NH₃]PF₆ to 0.16 V for [4-NH₃]PF₆. Unlike the structurally analogous half-sandwich ferric catalyst [Cp*Fe(1,2-Ph₂PC₆H₄NH)(NH₃)]⁺ containing a phosphinoamido ligand to only generate N₂H₄ stoichiometrically,¹⁷ [1-NH₃]PF₆ to [4-NH₃]PF₆ exhibit good performance for the selective electrocatalytic conversion of NH₃ to N₂H₄ with at least 98.6% selectivity and 86.3% Faraday efficiency. The mechanism studies illustrate that the Ru^III-amide intermediate has radical Ru^II-aminyl character via the inter-conversion Ru^III–NH₂ ⇋ Ru^II–NH_2͘͘˙. Bimolecular coupling of the Ru^II-aminyl species readily generates a N₂H₄-bridged biruthenium intermediate in [1-NH₃]PF₆ and [2-NH₃]PF₆ catalyst systems, but not in [3-NH₃]PF₆ and [4-NH₃]PF₆ catalyst systems. Another pathway of N₂H₄ formation in [1-NH₃]PF₆ and [4-NH₃]PF₆, the nucleophilic attack of Ru^IV-imide species by ammonia, is more feasible due to lower energy barriers of 5.4 kcal mol⁻¹ and 6.0 kcal mol⁻¹, compared to the bimolecular coupling pathway with energy barriers of 12.3 and 17.3 kcal mol⁻¹, respectively. This single-site molecular catalytic pathway is supported by a linear relationship between the catalytic current and concentration of catalyst and ammonia.

Author contributions

Xiao-Yi Yi as the corresponding author contributed to project design and paper revision. Xi Zhang mainly contributed to synthesis and electrocatalysis studies of the titled complexes, writing the preliminary draft. Shan Zhao and Guo Chen contributed to the assist with synthesis. Contribution of Liru Cao and Jian Lin lied in assisting with project design. Zhi-Yan Liu contributed to supplement data during the paper revision process. Piao He contributed to the DFT calculations.

Conflicts of interest

The authors declare no competing financial interest.

Data availability

The raw data including synthesis, characterization, and photophysical and catalytic properties of these complexes are available from the corresponding author, upon reasonable request.

CCDC 1 (2320736), 2 (2324232), 3 (2321021), 4 (2321506) and [2-NH₃]PF₆ (2427138) contain the supplementary crystallographic data for this paper.

The authors confirm that the data supporting the findings of this study are available within the article and/or its supplementary information (SI). Supplementary information: general methods for synthesis and characterization, crystallographic refinement, cyclic voltammetry experiments and electrolysis experiments, and DFT calculations. See DOI: https://doi.org/10.1039/d5sc08826g.

Acknowledgements

This work was supported by the National Natural Science Foundation of China (22471291) and the Foundation of Central South University Research Programme of Advanced Interdisciplinary (2023QYJC019) and Key Program of Hunan Provincial Natural Science Foundation of China (2025JJ30006).

References

1. (a) H. Tan, Z. Sang, Y. Tian, W. Peng, X. Liu and J. Liang, Ammonia as a Green Carbon-Free Fuel: A Pathway to the Sustainable Energy Economy, ACS Energy Lett., 2024, 9, 5120–5136; (b) D. T. Tran, T. H. Nguyen, H. Jeong, P. K. L. Tran, D. Malhotra, K. U. Jeong, N. H. Kim and J. H. Lee, Recent Engineering Advances in Nanocatalysts for NH₃-to-H₂ Conversion Technologies, Nano Energy, 2022, 94, 106929.
2. S. A. Lee, M. G. Lee and H. W. Jang, Catalysts for Electrochemical Ammonia Oxidation: Trend, Challenge, and Promise, Sci. China Mater., 2022, 65, 3334–3352.
3. L. Jiang and X. Fu, An Ammonia-Hydrogen Energy Roadmap for Carbon Neutrality: Opportunity and Challenges in China, Engineering, 2021, 7, 1688–1691.
4. N. M. Adli, H. Zhang, S. Mukherjee and G. Wu, Review—Ammonia Oxidation Electrocatalysis for Hydrogen Generation and Fuel Cells, J. Electrochem. Soc., 2018, 165, J3130–J3147.
5. (a) J. Li, F. Zhang, H. Xiong, Y. Cai and B. Zhang, Molecular Catalysts for Electrocatalytic Ammonia Oxidation, Sci. China Chem., 2024, 67, 3976–3993; (b) D. N. Stephens and M. T. Mock, Molecular Complexes for Catalytic Ammonia Oxidation to Dinitrogen and the Cleavage of N-H Bonds, Eur. J. Inorg. Chem., 2024, e202400039; (c) H. Y. Liu, H. M. C. Lant, C. C. Cody, J. Jelusic, R. H. Crabtree and G. W. Brudvig, Electrochemical Ammonia Oxidation with Molecular Catalysts, ACS Catal., 2023, 13, 4675–4682.
6. F. Habibzadeh, S. L. Miller, T. W. Hamann and M. R. Smith, Homogeneous Electrocatalytic Oxidation of Ammonia to N₂ under Mild Conditions, Proc. Natl. Acad. Sci. U. S. A., 2019, 116, 2849–2853.
7. K. Nakajima, H. Toda, K. Sakata and Y. Nishibayashi, Ruthenium-catalysed oxidative conversion of ammonia into dinitrogen, Nat. Chem., 2019, 11, 702–709.
8. P. L. Dunn, S. I. Johnson, W. Kaminsky and R. M. Bullock, Diversion of Catalytic C-N Bond Formation to Catalytic Oxidation of NH₃ through Modification of the Hydrogen Atom Abstractor, J. Am. Chem. Soc., 2020, 142, 3361–3365.
9. J. Holub, N. Vereshchuk, F. J. Sanchez-Baygual, M. Gil-Sepulcre, J. Benet-Buchholz and A. Llobet, Synthesis, Structure, and Ammonia Oxidation Catalytic Activity of Ru-NH₃ Complexes Containing Multidentate Polypyridyl Ligands, Inorg. Chem., 2021, 60, 13929–13940.
10. M. J. Trenerry, C. M. Wallen, T. R. Brown, S. V. Park and J. F. Berry, Spontaneous N₂ formation by a diruthenium complex enables electrocatalytic and aerobic oxidation of ammonia, Nat. Chem., 2021, 13, 1221–1227.
11. S. I. Jacob, A. Chakraborty, A. Chamas, R. Bock, L. Sepunaru and G. Ménard, Rapid Aqueous Ammonia Oxidation to N₂ Using a Molecular Ru Electrocatalyst, ACS Energy Lett., 2023, 8, 3760–3766.
12. A. M. Beiler, A. Denisiuk, J. Holub, F. J. Sánchez-Baygua, M. Gil-Sepulcre, M. Z. Ertem, D. Moonshiram, A. Piccioni and A. Llobet, Heterogeneous Electrochemical Ammonia Oxidation with a Ru-bda Oligomer Anchored on Graphitic Electrodes via CH-π Interactions, ACS Energy Lett., 2023, 8, 172–178.
13. S. Feng, J. Chen, R. Wang, H. Li, J. Xie, Z. Guo, T.-C. Lau and Y. Liu, Dual Pathways in Catalytic Ammonia Oxidation by a Ruthenium Complex Bearing a Tetradentate Bipyridine–Bipyrazole Ligand: Isolation of a Diruthenium Intermediate with a µ-Hexazene Derivative, J. Am. Chem. Soc., 2024, 146, 21490–21495.
14. B. Yuan, G. L. Tripodi, M. T. G. M. Derks, Y. Pereverzew, S. Zhou and J. Roithová, Mapping the Catalytic Cycle of Ru-Catalyzed Ammonia Oxidation, Angew. Chem., Int. Ed., 2025, 64, e202501617.
15. (a) G. Chen, P. He, C. Liu, X.-F. Mo, J.-J. Wei, Z.-W. Chen, T. Cheng, L.-Z. Fu and X.-Y. Yi, Direct synthesis of hydrazine by efficient electrochemical ruthenium-catalysed ammonia oxidation, Nat. Catal., 2023, 6, 949–958; (b) C. Zhou, X. Zhang, S. Zhao, S.-D. Zhong, X.-L. Ding, S.-P. Yang, F. Pan, P. He and X.-Y. Yi, Catalytic Selective Conversion of Ammonia into Hydrazine by a RuII(trpy) Complex Bearing a Pyridylpyrrole Ligand, ACS Catal., 2025, 15, 3535–3545; (c) S. Zhao, X. Zhang, G. Chen, T. Cheng, X.-L. Ding, S.-D. Zhong, S.-P. Yang, P. He and X.-Y. Yi, Selective Electrocatalytic Oxidation of Ammonia by Ru-dpp Complexes Containing Aromatic Nitrogen Donor as Axial Ligand, Inorg. Chem., 2024, 63, 23150–23157.
16. (a) M. D. Zott, P. Garrido-Barros and J. C. Peters, Electrocatalytic Ammonia Oxidation Mediated by a Polypyridyl Iron Catalyst, ACS Catal., 2019, 9, 10101–10108; (b) M. D. Zott and J. C. Peters, Enhanced Ammonia Oxidation Catalysis by a Low-Spin Iron Complex Featuring cis Coordination Sites, J. Am. Chem. Soc., 2021, 143, 7612–7616; (c) M. D. Zott and J. C. Peters, Improving Molecular Iron Ammonia Oxidation Electrocatalysts via Substituent Effects that Modulate Standard Potential and Stability, ACS Catal., 2023, 13, 14052–14057.
17. Y. Li, J.-Y. Chen, Q. Miao, X. Yu, L. Feng, R.-Z. Liao, S. Ye, C.-H. Tung and W. Wang, A Parent Iron Amido Complex in Catalysis of Ammonia Oxidation, J. Am. Chem. Soc., 2022, 144, 4365–4275.
18. L. Liu, S. I. Johnson, A. M. Appel and R. M. Bullock, Oxidation of Ammonia Catalyzed by a Molecular Iron Complex: Translating Chemical Catalysis to Mediated Electrocatalysis, Angew. Chem., Int. Ed., 2024, 63, e202402635.
19. M. E. Ahmed, R. J. Staples, T. R. Cundari and T. H. Warren, Electrocatalytic Ammonia Oxidation by Pyridyl-Substituted Ferrocenes, J. Am. Chem. Soc., 2025, 147, 6514–6522.
20. (a) M. E. Ahmed, M. Raghibi Boroujeni, P. Ghosh, C. Greene, S. Kundu, J. A. Bertke and T. H. Warren, Electrocatalytic Ammonia Oxidation by a Low-Coordinate Copper Complex, J. Am. Chem. Soc., 2022, 144, 21136–21145; (b) H. Y. Liu, H. M. C. Lant, J. L. Troiano, G. Hu, B. Q. Mercado, R. H. Crabtree and G. W. Brudvig, Electrocatalytic, Homogeneous Ammonia Oxidation in Water to Nitrate and Nitrite with a Copper Complex, J. Am. Chem. Soc., 2022, 144, 8449–8453.
21. (a) D. N. Stephens, R. K. Szilagyi, P. N. Roehling, N. Arulsamy and M. T. Mock, Catalytic Ammonia Oxidation to Dinitrogen by a Nickel Complex, Angew. Chem., Int. Ed., 2023, 62, e202213462; (b) N. X. Gu, P. H. Oyala and J. C. Peters, Hydrazine Formation via Coupling of a Nickel(III)-NH₂ radical, Angew. Chem., Int. Ed., 2021, 60, 4009–4013.
22. H. Toda, K. Kuroki, R. Kanega, S. Kuriyama, K. Nakajima, Y. Himeda, K. Sakata and Y. Nishibayashi, Manganese-Catalyzed Ammonia Oxidation into Dinitrogen under Chemical or Electrochemical Conditions, ChemPlusChem, 2021, 86, 1511–1516.
23. J. N. McPherson, B. Das and S. B. Colbran, Tridentate pyridine–pyrrolide chelate ligands: An under-appreciated ligand set with an immensely promising coordination chemistry, Coord. Chem. Rev., 2018, 375, 285–332.
24. G. Chen, X. L. Ding, P. He, T. Cheng, Y. Chen, J. Lin, X. Zhang, S. Zhao, N. Qiao and X.-Y. Yi, Understanding the factors governing the ammonia oxidation reaction by mononuclear ruthenium complex, Chem. Sci., 2025, 16, 7573–7578.
25. P. Štěpnička, J. Ludvík, J. Canivet and G. Süss-Fink, Relating catalytic activity and electrochemical properties: The Case of Arene-ruthenium Phenanthroline Complexes Catalytically Active in Transfer Hydrogenation, Inorg. Chim. Acta, 2006, 359, 2369–2374.
26. J. P. da Silva, O. M. Fuganti, G. Kramer, G. Facchin, L. E. N. Aquino, J. Ellena, D. F. Back, A. C. S. Gondim, E. H. S. Sousa, L. G. F. Lopes, S. Machado, I. D. L. Guimarães, K. Wohnrath and M. P. de Araujo, Electrochemical, Mechanistic, and DFT Studies of Amine Derived Diphosphines Containing Ru(ii)-cymene Complexes with Potent in vitro Cytotoxic Activity Against HeLa and Triple-negative Breast Cancer Cells MDA-MB-231, Dalton Trans., 2020, 49, 16498–16514.
27. A. J. Bard, L. R. Faulkner and H. S. White, Electrochemical Methods: Fundamentals and Applications, 2022, p. 248.
28. G. W. Watt and J. D. Chrisp, Spectrophotometric Method for Determination of Hydrazine, Anal. Chem., 1952, 24, 2006–2008.
29. H. Y. Liu, H. M. C. Lant, J. L. Troiano, G. Hu, B. Q. Mercado, R. H. Crabtree and G. W. Brudvig, Electrocatalytic, Homogeneous Ammonia Oxidation in Water to Nitrate and Nitrite with a Copper Complex, J. Am. Soc. Chem., 2022, 144, 8449–8453.
30. Q. Chen, J. Liang, L. Yue, Y. Luo, Q. Liu, N. Li, A. A. Alshehri, T. Li, H. Guo and X. Sun, CoO Nanoparticle Decorated N-doped Carbon Nanotubes: a High-efficiency Catalyst for Nitrate Reduction to Ammonia, Chem. Commun., 2022, 58, 5901–5904.
31. (a) CCDC 2320736: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2hwxgs; (b) CCDC 2324232: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2j0k7c; (c) CCDC 2321021: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2hx6n9; (d) CCDC 2321506: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2hxq9g; (e) CCDC 2427138: Experimental Crystal Structure Determination, 2026,  DOI:10.5517/ccdc.csd.cc2mgmsj.

---