Photoinduced oxidation of [Mn(L)₃]²⁺ and [Mn₂O₂(L)₄]³⁺ (L = 2,2′-bipyridine and 4,4′-dimethyl-2,2′-bipyridine) with the [Ru(bpy)₃]²⁺/-aryl diazonium salt system

Abstract

The photophysical, photochemical and electrochemical studies of a mixture of bipyridinyl ruthenium ([Ru^II(bpy)₃]²⁺) and manganese ([Mn^II(L)₃]²⁺ or [Mn₂^III,IVO₂(L)₄]³⁺, L = bpy (2,2′-bipyridine) or dmbpy (4,4′-dimethyl-2,2′-bipyridine)) complexes have been investigated in CH₃CN. Electrochemical oxidations of [Mn^II(L)₃]²⁺ are irreversible and lead, by subsequent chemical reactions, to the corresponding di-μ-oxo complex [Mn₂^III,IVO₂(L)₄]³⁺ with a good yield. These latter complexes can be then reversibly oxidized into the stable [Mn₂^IV,IVO₂(L)₄]⁴⁺ species. The luminescence lifetime of the excited state of the photosensitizer [Ru^II(bpy)₃]²⁺* in the presence of variable concentration of manganese complexes has been determined. Using the Stern–Volmer equation, the quenching constant rate k_q have been estimated. It appears that the [Mn^II(L)₃]²⁺ mononuclear complexes quench only very weakly the excited state [Ru^II(bpy)₃]²⁺* since the magnitude of k_q determined for the bpy and dmbpy complexes is about 10⁷ M⁻¹ s⁻¹. In contrast, a strong decrease of the luminescence lifetime is observed by addition of an increasing concentration of [Mn₂^III,IVO₂(L)₄]³⁺. The k_q values obtained for the bpy and dmbpy complexes are, respectively, 2.3 × 10⁹ and 2.5 × 10⁹ M⁻¹ s⁻¹. The major quenching pathways of [Ru^II(bpy)₃]²⁺* by those binuclear complexes of manganese are presumably energy transfer processes. Finally, the possibility of photocatalytic oxidation of [Mn^II(L)₃]²⁺ and [Mn₂^III,IVO₂(L)₄]³⁺ has been evaluated by continuous irradiation in the presence of the photosensitizer [Ru^II(bpy)₃]²⁺ and an aryl diazonium salt, ArN₂⁺, playing the role of an irreversible electron acceptor. The photooxidation process transforming [Mn^II(L)₃]²⁺ into [Mn₂^III,IVO₂(L)₄]³⁺ by intermolecular electron transfers between photogenerated [Ru^III(bpy)₃]³⁺ and [Mn^II(L)₃]²⁺ occurs for the bpy and dmbpy complexes with a high efficiency. The subsequent photooxidation leading to [Mn₂^IV,IVO₂(L)₄]⁴⁺ is efficient only in the case of the dmbpy complex. The formation of those different species by electrochemical or photochemical ways has been demonstrated and quantified by coupled UV-visible absorption and EPR spectroscopy experiments. The efficiencies of the photoinduced oxidative processes have been correlated to the electrochemical data.

---

Introduction

In recent years, important efforts have been made to mimic the function of the donor-side of the photosystem II (PS II) using superstructured heterometallic complexes.^1–6 In the natural PS II system, under illumination, the photoactive P₆₈₀ chlorophylls center oxidizes water into dioxygen through a catalytic system containing an oxo cluster of four Mn and one Ca ion and a tyrosine residue; the detailed mechanism remaining not entirely elucidated, as is the case for the structure of the Mn cluster.^7–9 The main strategy used in literature to elaborate models of PS II involves the covalent coupling of a photoactive [Ru^II(bpy)₃]²⁺ (bpy = 2,2′-bipyridine) moiety, playing the role of the P₆₈₀ chlorophylls, to some mono or binuclear manganese complexes. In Scheme 1 are shown some relevant superstructured bimetallic complexes developed by Styring and co-workers.^{3–6,10–15} Laser flash photolysis experiments on the two first bimetallic complexes (Scheme 1A, B) have shown that an intramolecular electron transfer can occur from the Mn^II site to the photogenerated [Ru^III(bpy)₃]³⁺ species, using an external electron acceptor-like viologen in acetonitrile. However, it was found that the electron transfer (ET) rate constant is fairly slow (k_ET in the range 10⁵–2 × 10⁷ s⁻¹) and depends on the distance between the two metallic centers.⁶ These authors have also developed some [Ru^II(bpy)₃]²⁺ complexes covalently linked to a tyrosine residue.^16,17 In the presence of an external electron acceptor, an intramolecular electron transfer between the photogenerated [Ru^III(bpy)₃]³⁺ species and the tyrosine has been characterized by nanosecond laser flash photolysis experiments. The oxidized tyrosyl radical can then oxidize an oxo Mn₂^III,III complex added to the solution into a Mn₂^III,IV one.¹⁸ Finally, following these results, a superstructured heterobimetallic complexes has been proposed where the [Ru^II(bpy)₃]²⁺ center and a binuclear Mn₂^II,II complex are covalently linked by a tyrosine bridge (Scheme 1C);^3,13–15 the tyrosyl moiety playing the role of an electron relay between the photogenerated [Ru^III(bpy)₃]³⁺ species and the Mn₂^II,II complex. A relative faster electron transfer rate constant (k_ET > 10⁷ s⁻¹) has been determined, leading to the Mn₂^II,III species or to the Mn₂^III,IV analogue if experiments are carried out in the presence of water.¹⁴ All these studies have been conducted by flash photolysis, but the different oxidation states of the Mn center were not quantitatively generated by continuous photoirradiation of the solution.

Scheme 1 Structure of model complexes developed by Styring et al. [ref. 3, and refs. therein].

Paradoxically, photophysical and photochemical properties of a simple mixture of the [Ru^II(bpy)₃]²⁺ and [Mn^II(bpy)₃]²⁺ or [Mn₂^III,IVO₂(bpy)₄]³⁺ complexes have never been examined to our knowledge. In order to perform such photoinduced processes at the quantitative scale, one needs the use of sacrificial oxidant that is irreversibly reduced allowing the net conversion of Ru^II into Ru^III. Recently we demonstrated that the 4-bromophenyl diazonium salt (ArN₂⁺) played the role of efficient sacrificial oxidant towards a simple mixture of [Ru^II(bpy)₃]²⁺ and [Fe^II(bpy)₃]²⁺ complexes as well as some heterobinuclear polypyridinique complexes of Ru^II and Fe^II (Ru^II–Fe^II) in organic solvent-like CH₃CN.¹⁹ In these experimental conditions, the effective energy transfer from the ³MLCT excited state of the Ru^II(bpy)₃²⁺ center to the Fe^II(bpy)₃²⁺ unit is short-circuited allowing the photoinduced formation of Ru^III(bpy)₃³⁺ and the oxidation of the Fe^II(bpy)₃²⁺ subunit with an overall high efficiency. The oxidation of [Fe^II(bpy)₃]²⁺ occurs at a potential close to that of [Mn^II(bpy)₃]²⁺. However, in contrast to the Fe^II/Fe^III redox couple which is perfectly reversible in CH₃CN, the Mn^II/Mn^III one is irreversible, [Mn^III(bpy)₃]³⁺ reacting with the residual water to form the [Mn₂^III,IVO₂(bpy)₄]³⁺ complex.^20–22

Thus, we investigated the photophysical and photochemical properties of a mixture of [Ru^II(bpy)₃]²⁺ with [Mn^II(L)₃]²⁺ (L = bpy or dmbpy (4,4′-dimethyl-2,2′-bipyridine)) and with corresponding binuclear complexes [Mn₂^III,IVO₂(L)₄]³⁺ in CH₃CN in the presence of ArN₂⁺. In these experimental conditions, the successive photoinduced intermolecular electron transfers shown in Scheme 2 would be expected. The formation of the oxidized forms under continuous irradiation has been demonstrated by coupled UV-visible absorption and EPR spectroscopy experiments. Finally, we show that the efficiencies of the photoredox processes of the different systems can be correlated to the electrochemical data, especially for the second photooxidation.

Scheme 2 Schematic presentation for the photooxidation mechanism of [Mn^II(L)₃]²⁺ by [Ru^II(bpy)₃]²⁺ in the presence of an external electron acceptor ArN₂⁺.

Experimental

Materials and general methods

Acetonitrile (CH₃CN, Rathburn, HPLC grade) and tetra-n-butylammonium perchlorate (Bu₄NClO₄, Fluka) were used as received and stored under an argon atmosphere in a dry glove box. X-band electron paramagnetic resonance (EPR) spectra were recorded with a Bruker ESP 300 E spectrometer at 100 K, with the following parameters: microwave 1 mW, modulation amplitude 0.197 G, time constant 327.68 ms, scan rate 1342 s, scan width 8 G, modulation frequency 100 kHz.

General synthesis

Synthesis of the diazonium salt. The 4-bromophenyl diazonium tetrafluoroborate BF₄⁻ (ArN₂⁺BF₄⁻) was synthesized as previously described.²³

Synthesis of the complexes. The [Mn(L)₃](PF₆)₂ and [Mn₂^III,IV(O)₂(L)₄](PF₆)₃ (L = bpy (2,2′-bipyridine) or dmbpy (4,4′-dimethyl-2,2′-bipyridine)) complexes were synthesized as previously described.^22,24,25 [Ru^II(bpy)₃](PF₆)₂ was synthesized by anion exchange starting from [Ru^II(bpy)₃](Cl)₂ (Aldrich).

Electrochemistry

All electrochemical measurements were run under an argon atmosphere in a dry glove box at room temperature. Cyclic voltammetry (CV) and controlled potential electrolysis experiments were performed using an EG&G model 173 potentiostat/galvanostat equipped with a PAR model universal programmer and a PAR model 179 digital coulometer. The standard three-electrodes electrochemical cell was used. Potentials were referred to an Ag/0.01 M AgNO₃ reference electrode in CH₃CN + 0.1 M Bu₄NClO₄. Potentials referred to that system can be converted to the ferrocene/ferrocenium couple by subtracting 87 mV, to SCE by adding 298 mV or to ENH reference electrode by adding 0.548 V. The working electrodes were platinum disks polished with 2 μm diamond paste (Mecaprex Presi) that were 5 mm in diameter for cyclic voltammetry (E_pa, anodic peak potential; E_pc, cathodic peak potential; E_1/2 = (E_pa + E_pc)/2; ΔE_p = E_pa − E_pc) and 2 mm in diameter for rotating disk electrode experiments (RDE). Exhaustive electrolyses were carried out on a platinum plate (5 cm²) or on a carbon felt electrode (RCV 2000, 65 mg cm⁻³). The auxiliary electrode was a Pt wire in CH₃CN + 0.1 M Bu₄NClO₄. For electrochemical experiments, electronic absorption spectra were recorded on a Hewlett-Packard 8452 A diode array spectrophotometer. Initial and electrolyzed solutions were transferred to a conventional 0.1 or 1 cm path length quartz cells in the glove box.

Spectroscopy

Absorption. UV-Visible spectra of the photolysed solutions were obtained using a Cary 100 absorption spectrophotometer on 0.1 or 1 cm path length quartz cells.

Luminescence. The steady-state emission spectra were recorded on a Photon Technology International (PTI) SE-900M spectrofluorimeter. All the samples for luminescence experiments were prepared in a dry glove box in deoxygenated CH₃CN + 0.1 M Bu₄NClO₄ and contained in a quartz cell. The samples were maintained in anaerobic conditions with a Teflon cap. For solutions containing a mixture of ruthenium and manganese complexes, the concentration of [Ru^II(bpy)₃]²⁺ is fixed to 0.013 mM in order to obtain a 0.2 absorbance at 454 nm, whereas the [Mn^II(L)₃]²⁺ or [Mn₂^III,IVO₂(L)₄]³⁺ complex concentrations are ranged between 0.05 and 5.4 or 1.3 mM, respectively.

The luminescence lifetime of [Ru^II(bpy)₃]²⁺* without or with a manganese complex was performed after irradiation at λ = 337 nm with a 4 ns pulsed N₂ laser (optilas VSL-337ND-S) and recorded at λ = 650 nm using filter and a photomultiplier tube (Hamamatsu E2380-01) coupled with an ultra-fast oscilloscope (LeCroy 9310 AM).

Irradiation. The irradiation experiments have been performed using a mercury lamp (Oriel 66901) (250 W) whose UV and IR radiation was cut off with filters. For these experiments, all the solutions were prepared in a dry glove box. The solutions were constituted by a mixture of [Ru(bpy)₃]²⁺ (0.05 mM), [Mn₂^III,IVO₂(L)₄]³⁺ (0.5 mM) or [Mn₂^II(L)₃]²⁺ (1 mM), and ArN₂⁺ (15 mM). In these conditions, the concentration of ArN₂⁺ can be considered as constant during the irradiation. Photooxidation experiments have been followed by UV-visible and EPR spectroscopies and by electrochemistry. For the latter experiments, the cyclic voltammograms have been recorded only in the potential range of the oxidation of the ruthenium and manganese complexes (between 0.5 and 1.4 V).

Results and discussion

Electrochemistry and spectroscopy

Since the photocatalytical oxidation experiments have been carried out in CH₃CN + 0.1 M Bu₄NClO₄ solutions containing a mixture of [Ru^II(bpy)₃]²⁺ and manganese complexes ([Mn^II(L)₃]²⁺ or [Mn₂^III,IVO₂(L)₄]³⁺) in the concentration ratio of about 1/10, it was important to known precisely the spectroscopic (UV-visible and EPR) and electrochemical properties of such solutions. We firstly reinvestigated in details the properties of each complex and their corresponding oxidized forms in this solvent and secondly those of a mixture of manganese and ruthenium complexes. Table 1 and Table 2 summarize, respectively, the electrochemical and UV-visible data of the complexes.

Table 1 Electrochemical data of [Ru^II(bpy)₃]²⁺, [Mn₂^III,IVO₂(L)₄]³⁺ and [Mn^II(L)₃]²⁺ (L = bpy and dmbpy) in deoxygenated CH₃CN + 0.1 M Bu₄NClO₄ at a scan rate of 100 mV s⁻¹

Complexes	Oxidation processes^a	Reduction processes^a
a Versus Ag/Ag^+ (0.01 M AgNO3 in CH3CN + 0.1 M Bu4NClO4). b Irreversible processes.
	E 1/2/V (ΔEp/mV)	E 1/2/V (ΔEp/mV)
	Ru^II ⇄ Ru^III	1st	2nd	3rd
[Ru^II(bpy)3]^2+	0.98 (80)	−1.64 (60)	−1.83 (60)	−2.08 (70)
	E 1/2/V (ΔEp/mV)	E pc/V
	Mn2^III,IVO2 ⇄ Mn2^IV,IVO2	Mn2^III,IVO2 ⇄ Mn2^III,IIIO2 → Mn^II
[Mn2^III,IVO2(bpy)4]^3+	1.02 (90)	0.04
[Mn2^III,IVO2(dmbpy)4]^3+	0.86 (80)	−0.10
	E pa/V^b	E 1/2/V (ΔEp/mV)
	Mn^II → Mn^III	1st	2nd	3rd
[Mn^II(bpy)3]^2+	0.84	−1.66 (60)	−1.83 (60)	−2.07 (80)
[Mn^II(dmbpy)3]^2+	—	−1.77 (80)	−1.95 (70)	−2.16 (80)

---

Table 2 UV-visible data of [Ru(bpy)₃]^2+/3+, [Mn₂O₂(L)₄]^3+/4+ and [Mn(L)₃]^2+/3+ in deoxygenated CH₃CN + 0.1 M Bu₄NClO₄

Complexes	λ abs/nm (ε/M^−1 cm^−1)
[Ru^II(bpy)3]^2+	432 (sh. 13030) 452 (15000)
[Ru^III(bpy)3]^3+	416 (4700) 668 (460)
[Mn2^III,IVO2(bpy)4]^3+	406 (sh. 1950) 470 (sh. 1100) 525 (556) 555 (460) 684 (560)
[Mn2^III,IVO2(dmbpy)4]^4+	406 (sh. 2060) 470 (sh. 1250) 528 (550) 559 (450) 688 (560)
[Mn^II(dmbpy)3]^2+	249 (22900) 295 (32800) 304 (sh. 25000)
[Mn^II(bpy)3]^2+	243 (25300) 295 (31300) 307 (sh. 23000)

---

[Mn₂^III,IVO₂(L)₄]³⁺ (L = bpy and dmbpy). As previously described,^22,26,27 the electrochemical behaviour of those complexes in CH₃CN + 0.1 M Bu₄NClO₄ is characterized by a reversible one-electron oxidation process at E_1/2 = 1.02 or 0.86 V for bpy or dmbpy, respectively, leading to the Mn^IV–Mn^IV species (eqn. (1)):

[Mn₂^III,IVO₂(L)₄]³⁺ ⇄ [Mn₂^IV,IVO₂(L)₄]⁴⁺+ e⁻ (1)

An irreversible reduction process is also observed at E_pc = 0.04 for bpy or −0.10 V for dmbpy, corresponding to the formation of the Mn^III–Mn^III species (Table 1). These latter unstable species, in subsequent chemically coupled reactions, afford the mononuclear complexes [Mn(L)₃]²⁺, associated with MnO₂ and the release of L ligand²² (eqn. (2)):

[Mn₂^III,IVO₂(L)₄]³⁺ + e⁻ → [Mn^II(L)₃]²⁺ + (Mn^IVO₂)_x + L (2)

The dmbpy complex is easier to be oxidized and more difficult to be reduced than the bpy complex in accordance with the presence of the two methyl donor substitutions. Controlled potential oxidations at 1.10 V of [Mn₂^III,IVO₂(L)₄]³⁺ solutions consume one electron per molecule of complexes. The oxidation causes some decomposition only for the bpy derivatives. Indeed, as judged spectroscopically, the [Mn₂^IV,IVO₂(L)₄]⁴⁺ species were obtained with a 75% yield for L = bpy and 95% for L = dmbpy. The restitution with the same yield of [Mn₂^III,IVO₂(L)₄]³⁺ can be achieved by further exhaustive reductions at 0.60 V (eqn. (1)).

These oxidation processes are associated by significant changes in UV-visible absorption (Table 2). The visible absorption spectra of the initial green Mn^III–Mn^IV solutions show three bands at 525, 555 and 684 nm for L = bpy (528, 559 and 688 for L = dmbpy) (Fig. 1). Similar bands have been detected for other μ-oxo binuclear complexes²⁸ allowing us to assign them as Mn^IV d–d transitions for those at 525 and 555 nm and ligand to metal charge transfer (LMCT) from oxo ligands to Mn^IV for the band at 684 nm. In the fully oxidized brown Mn^IV–Mn^IV solutions, new shoulders appear at 565 and 638 nm for L = bpy (570 and 640 for L = dmbpy). In addition, the intensity of the absorbance between 400 and 500 nm increases strongly and is associated with the appearance of two clear shoulders at 420 and 490 nm (420 and 485 nm for L = dmbpy). These latter shoulders were assigned to the contribution of the Mn^IV d–d transitions and to the LMCT of oxo ligands to Mn^IV ion.²⁸

Fig. 1 Visible absorption spectra in CH₃CN + 0.1 M Bu₄NClO₄, of (a) 2 mM [Mn^II(dmbpy)₃]²⁺; (b) 0.9 mM [Mn₂^III,IVO₂(dmbpy)₄]³⁺ and (c) solution (b) after exhaustive oxidation at 1.10 V (formation of [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺), l = 1 cm. Insert, l = 1 mm.

The oxidation processes are also evidenced by EPR spectroscopy. Indeed, X-band EPR spectroscopy allows to easily distinguish the Mn^IV–Mn^IV and Mn^III–Mn^IV species since Mn^IV–Mn^IV is silent whereas the mixed-valence species presents the regular 16 line spectrum centred at g = 2.²⁹ Experimentally, exhaustive oxidations lead to the quasi-total disappearance of the initial 16 line signal of the starting complexes on the X-band EPR spectra.

[Mn₂^III,IVO₂(L)₄]³⁺ in the presence of [Ru^II(bpy)₃]²⁺. The electrochemical behaviour of the photosensitizer [Ru^II(bpy)₃]²⁺ in CH₃CN medium has been deeply studied. The cyclic voltammogram exhibits in the positive region one reversible metal-based oxidation process at E_1/2 = 0.98 V and, in the negative one, three successive ligand-based one-electron reversible reductions (Table 1). After addition of a [Ru^II(bpy)₃]²⁺ (0.1 mM) to a [Mn₂^III,IVO₂(L)₄]³⁺ (1.2 mM) CH₃CN solution, the resulting cyclic voltammograms and UV-visible spectra correspond to the superimposition of the electroactivity of both complexes with intensities proportional to their respective concentrations.

In the case of the dmbpy complex, the markedly lower oxidation potential of [Mn₂^III,IVO₂(dmbpy)₄]³⁺ (E_1/2 = 0.86 V) compared to [Ru^II(bpy)₃]²⁺ (E_1/2 = 0.98 V), ΔE_1/2 = 120 mV, leads to the clear observation of both of these successive oxidation systems in the cyclic voltammogram. This demonstrates that the oxidized Ru^III form could act as an efficient oxidant towards [Mn₂^III,IVO₂(dmbpy)₄]³⁺. An electrolysis performed at E = 1.10 V and stopped after one electron has been consumed allows the selective oxidation of the manganese complex into [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺ with a 95% yield. At this electrolysis potential, both [Mn₂^III,IVO₂(dmbpy)₄]³⁺ and [Ru^II(bpy)₃]²⁺ can be oxidized. The electrogenerated [Ru^III(bpy)₃]³⁺ oxidizes in turn the manganese complex. That is the reason why, when electrolysis is stopped after one electron has been consumed, only the manganese complex is oxidized. The pursuit of the electrolysis at this potential induced the quantitative oxidation of [Ru^II(bpy)₃]²⁺ into [Ru^III(bpy)₃]³⁺.

The [Mn₂^III,IVO₂(bpy)₄]³⁺ (E_1/2 = 1.02 V) complex is slightly more difficult to oxidize than [Ru^II(bpy)₃]²⁺. The very small difference of potential between Ru^II/Ru^III and Mn^IIIMn^IV/Mn^IVMn^IV redox systems (ΔE_1/2 = 40 mV) implies that these two oxidation processes cannot be clearly distinguished, the cyclic voltammogram of the mixture exhibiting only one single reversible system at E_1/2 = 1.00 V. Consequently, an electrolysis of the solution at 1.20 V yields to the concomitant formation of [Ru^III(bpy)₃]³⁺ and [Mn₂^IV,IVO₂(bpy)₄]⁴⁺ and does not allow the selective formation of [Ru^III(bpy)₃]³⁺. This result shows that [Ru^III(bpy)₃]³⁺ should not be a good candidate for an efficient oxidation of [Mn₂^III,IVO₂(bpy)₄]³⁺ in contrast to the case of [Mn₂^III,IVO₂(dmbpy)₄]³⁺ (see the ‘Continuous photolysis’ section).

The UV-visible spectra of the initial and oxidized solutions are the results of the superimposition of the absorption bands of the ruthenium and binuclear manganese complexes in their initial and oxidized forms.

The EPR spectra of the initial solutions present the typical 16 line spectra of the [Mn₂^III,IVO₂(L)₄]³⁺ species centred at g = 2 (Fig. 2A for L = dmbpy). In the fully oxidized solutions, the EPR spectra show now only a low intensity broad signal at g = 2.7 corresponding to [Ru^III(bpy)₃]³⁺ (Fig. 2B for L = dmbpy).¹⁸

Fig. 2 EPR spectra of a mixture of 1.2 mM [Mn₂^III,IVO₂(dmbpy)₄]³⁺ and 0.1 mM of [Ru^II(bpy)₃]²⁺ in CH₃CN + 0.1 M Bu₄NClO₄. (A) Initial solution, (B) after exhaustive electrolyses at 1.10 V: formation of [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺ and [Ru^III(bpy)₃]³⁺.

[Mn^II(L)₃]²⁺ (L = bpy and dmbpy). It has been previously demonstrated that the electrochemical oxidation of tris-bipyridine and tris-phenanthroline (substituted or not) manganese(II) complexes leads to the corresponding di-μ-oxo dimanganese(III,IV) complexes in CH₃CN + 0.1 M Bu₄NClO₄ (eqn. (3)).²¹ These latter complexes can be then reduced back to their initial Mn^II forms.The shape of the cyclic voltammograms of [Mn^II(L)₃]²⁺ solutions (1.2 mM) in the positive potential range, depends on the potential scan rate used (Fig. 3). At a low scan rate (20 mV s⁻¹, Fig. 3B), the electrochemical behaviour of [Mn^II(bpy)₃]²⁺ in CH₃CN + 0.1 M Bu₄NClO₄, is close to that observed in a buffered aqueous medium^24,25 with a broad oxidation peak at 0.84 V (shoulder) corresponding to the Mn^II/Mn^III redox couple. This Mn^III species reacts rapidly with residual water in the solvent via a disproportionation reaction to form the binuclear di-μ-oxo [Mn₂^III,IVO₂(bpy)₄]³⁺. The formation of this latter complex is displayed by the presence of its reversible oxidation system at E_1/2 = 1.02 V (E_pa = 1.07 V and E_pc = 0.98 V) into [Mn₂^IV,IVO₂(bpy)₄]⁴⁺, which follows the irreversible oxidation of [Mn^II(bpy)₃]²⁺, and on the reverse scan, by an irreversible reduction peak at E_pc = 0.04 V. It should be noted that in a previous study,²⁰ the Mn^II/Mn^III redox couple was erroneously assigned to that of the binuclear [Mn₂^III,IVO₂(bpy)₄]³⁺ (E_1/2 = 1.06 V). Otherwise, at higher scan rate, (100 mV s⁻¹, Fig. 3A) the expected reduction peak of the Mn^IV–Mn^IV species is masked by a new and more intense peak at E_pc = 0.92 V, corresponding probably to transient species such as a mono μ-oxo binuclear manganese complex.²² In addition, the intensity of the irreversible peak at 0.04 V becomes weaker.

Fig. 3 Cyclic voltammograms at a Pt electrode in CH₃CN + 0.1 M Bu₄NClO₄ of (A) 1.2 mM [Mn^II(bpy)₃]²⁺ with sweep rate ν = 100 mV s⁻¹; (B) ν = 20 mV s⁻¹; (C) after addition of 0.1 mM of [Ru^II(bpy)₃]²⁺, ν = 100 mV s⁻¹. (D) 1.2 mM [Mn^II(dmbpy)₃]²⁺, ν = 100 mV s⁻¹; (E) ν = 20 mV s⁻¹; (F) after addition of 0.1 mM of [Ru^II(bpy)₃]²⁺, ν = 100 mV s⁻¹.

An exhaustive potential-controlled oxidation at E = 0.84 V consumes 1.5 electrons per manganese atom and furnishes a green solution that exhibits the spectroscopic and electrochemical features of the [Mn₂^III,IVO₂(bpy)₄]³⁺ species (yield: 75%). In addition to the typical electroactivity of the binuclear complex, the cyclic voltammogram shows an additional irreversible peak at E_pc = −0.70 V typical of the reduction of the released protonated bipyridine ligand (eqn. (3)). This transformation is reversible as confirmed by a controlled-potential reduction at −0.1 V which consumes 3 electrons per molecule of binuclear and regenerates a solution of [Mn^II(bpy)₃]²⁺ with a 90% yield.

In the case of [Mn^II(dmbpy)₃]²⁺, the rate of formation of the corresponding di-μ-oxo binuclear complex is markedly slower than that of [Mn^II(bpy)₃]²⁺ as displayed on the cyclic voltammograms recorded at scan rate of 20 and 100 mV s⁻¹ (Fig. 3D, E). At 20 mV s⁻¹ the shape of the cyclic voltammogram is closed to that of the bpy complex recorded at 100 mV s⁻¹ with a shift towards the lower potential values of about 140 mV, due to the electronic donating effect of the methyl substitution. However, no clear shoulder related to the Mn^II/Mn^III oxidation process is observed. At 100 mV s⁻¹, the formation of the binuclear di-μ-oxo is less visible and the oxidation and reduction irreversible peaks (E_pa = 1.10 V and E_pc = 0.61 V, respectively) can be attributed to intermediate species produced during the electrochemical transformation of [Mn^II(dmbpy)₃]²⁺ into [Mn₂^III,IVO₂(dmbpy)₄]³⁺.²² This latter complex can be obtained with a 75% yield by an exhaustive oxidation performed at the foot of the anodic peak (E = 0.72 V) indicating that the oxidation of Mn^II to Mn^III occurs at this potential even if no clear shoulder is observed.

The UV-visible absorption spectra of the [Mn^II(L)₃]²⁺ complexes present absorption bands only in the near UV region corresponding to π → π* bands of the ligands (Table 2, Fig. 1), and exhaustive oxidations lead to the appearance of the typical visible absorption bands of the [Mn₂^III,IVO₂(L)₄]³⁺ complexes (Table 2).

The X-band EPR spectra of the initial solutions of [Mn^II(L)₃]²⁺ exhibit 6 line signals centred at g = 2 in accordance with a high-spin Mn^II (3d⁵) ion, the hyperfine splitting resulting from the interaction between the electronic (S = 5/2) and the nuclear spin (I = 5/2). After electrochemical oxidations into [Mn₂^III,IVO₂(L)₄]³⁺ complexes, the 6 line spectra are replaced by the typical 16 lines of these latter. However, these 16 line signal is slightly distorted by the presence of a weak 6 line signal of Mn(CH₃CN)₆²⁺ indicating a partial decoordination of manganese during the oxidation processes (see the next section).

[Mn^II(L)₃]²⁺ in the presence of [Ru^II(bpy)₃]²⁺. Addition of a [Ru^II(bpy)₃]²⁺ (0.1 mM) to [Mn^II(L)₃]²⁺ (1.2 mM) solutions induces some modifications of the shape of the cyclic voltammograms in the positive region (Fig. 3C, F). The presence of the ruthenium complex, even at a low concentration compared to the manganese one, increases the rate of formation of [Mn₂^III,IVO₂(L)₄]³⁺. This is displayed by the quasi-similarity of the shape of the CV of a solution containing the Mn and Ru complexes at 100 mV s⁻¹ (with the exception of the additional system of the ruthenium complex in the case of [Mn^II(dmbpy)₃]²⁺) with those at 20 mV s⁻¹ relative to the solution containing only the manganese complex. This means that the Ru complex acts as an electron relay in homogeneous phase increasing the oxidation kinetic of [Mn^II(L)₃]²⁺, the oxidation potential of these complexes being largely lower than that of [Ru^II(bpy)₃]²⁺. Electrolyses performed at E = 1.10 for L = bpy or dmbpy and stopped after 1.5 electrons per Mn^II complex have been consumed allow the selective formation of [Mn₂^III,IVO₂(L)₄]³⁺ with the same yield to that obtained in absence of ruthenium complex, whereas [Ru^II(bpy)₃]²⁺ remains unoxidized (Fig. 4). This result confirms that [Ru^III(bpy)₃]³⁺ can be considered as an efficient oxidant towards [Mn^II(L)₃]²⁺. At this stage, extended electrolyses at 1.10 V yield similar results to those obtained from solutions containing [Ru^II(bpy)₃]²⁺ and chemically prepared samples of [Mn₂^III,IVO₂(L)₄]³⁺. As expected, after these electrolyses, the UV-visible absorption spectra exhibit the superimposition of the absorbance of Mn^III–Mn^IV and Ru^II complexes, and then those of the Mn^IV–Mn^IV, with the ruthenium complex in the Ru^II or Ru^III oxidized state (see above).

Fig. 4 Cyclic voltammograms at a Pt electrode in CH₃CN + 0.1 M Bu₄NClO₄ obtained after oxidation at 1.10 V (1.5 electron consumed) of a mixture of 1.2 mM [Mn^II(L)₃]²⁺ and 0.1 mM of [Ru^II(bpy)₃]²⁺, (A) L = bpy, (B) L = dmbpy; ν = 100 mV s⁻¹.

Concerning the EPR spectra (Fig. 5, for L = dmbpy), after oxidation into [Mn₂^III,IVO₂(L)₄]³⁺, the 16 line signals, characteristic of these complexes, replaced the initial 6 lines of the starting Mn^II species. However, it should be noted that the signal is distorted by the presence of a 6 line signal centred at g = 2 corresponding to a new mononuclear Mn^II species. After oxidation into [Mn₂^IV,IVO₂(L)₄]⁴⁺, only this signal is present (Fig. 5C). This spectrum is identical to that of Mn(ClO₄)₂·6H₂O dissolved in CH₃CN, allowing one to assign it, without ambiguity, to Mn^II(CH₃CN)₆²⁺. Indeed, it has been demonstrated by EXAFS spectroscopy that in CH₃CN medium, Mn²⁺ is coordinated by six nitrogens of the solvent.³⁰ In the fully oxidized solutions, the additional signal centred at g = 2.7 of [Ru^III(bpy)₃]³⁺ is also observed (Fig. 5D).¹⁸

Fig. 5 Evolution of the EPR spectrum during electrolyses at 1.10 V of the mixture of 1.2 mM [Mn^II(dmbpy)₃]²⁺ and 0.1 mM of [Ru^II(bpy)₃]²⁺ in CH₃CN + 0.1 M Bu₄NClO₄. (A) Initial solution, (B) after formation of [Mn₂^III,IVO₂(dmbpy)₄]³⁺, (C) after formation of [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺, (D) after formation of [Ru^III(bpy)₃]³⁺.

Photophysics

Any molecule presenting an absorption in the visible region, even with a weak intensity, is likely to react by an electronic energy transfer (EET) process with the excited state of the photosensitizer, [Ru^II(bpy)₃]²⁺*. This is the case for the [Mn₂^III,IVO₂(L₄)]³⁺ complexes and at a markedly less extent, for the [Mn^II(L)₃]²⁺ ones (absorption in the near UV region; see Table 2). Since an energy transfer process between [Ru^II(bpy)₃]²⁺* and the different Mn complexes cannot be ruled out, and may limit the photooxidation rate constant of [Ru^II(bpy)₃]²⁺* by the irreversible electron acceptor used (ArN₂⁺) it was important to investigate the effect of the presence of the manganese complexes on the luminescence lifetime of [Ru^II(bpy)₃]²⁺* (τ) before studying the electron transfer (ET) processes between [Ru^III(bpy)₃]³⁺ and the manganese complexes.

Luminescence lifetime of [Ru^II(bpy)₃]²⁺* in the presence of the [Mn^II(L)₃]²⁺ complexes. Some previous works have already suggested that Mn^II salts do not quench the luminescence of [Ru^II(bpy)₃]²⁺-like species. Only the linkage of a Mn^II bipyridinyl subunit to a Ru^II one seems to have an influence on the luminescence lifetime presumably by an energy transfer process.^6,11,31 The possibility of an interaction between [Ru^II(bpy)₃]²⁺* and [Mn^II(L)₃]²⁺ in CH₃CN + 0.1 M Bu₄NClO₄ has been checked by adding a variable concentration (up to 5.4 mM) of these complexes to a [Ru^II(bpy)₃]²⁺ solution (about 0.013 mM). In all cases, the luminescence decay of [Ru^II(bpy)₃]²⁺* is monoexponential and decreases by increasing the manganese concentration. The quenching constant rate, k_q, has been estimated using the Stern–Volmer equation (eqn. (4)) in which τ₀ and τ are the luminescence lifetimes of [Ru^II(bpy)₃]²⁺*, without and with a variable concentration of quencher Q: [Mn^II(L)₃]²⁺ (Fig. 6(a)).

τ₀/τ = 1 + k_qτ₀[Q] (4)

It appears that the [Mn^II(L)₃]²⁺ complexes quench only weakly the excited state [Ru^II(bpy)₃]²⁺*, the magnitude of k_q determined for both [Mn^II(bpy)₃]²⁺ and [Mn^II(dmbpy)₃]²⁺ is of 10⁷ M⁻¹ s⁻¹. Quenching by a photoinduced redox process can be ruled out since the potential value of Ru^I/Ru^II* (E_1/2 = 0.46 V)³² is strongly lower to that of Mn^II/Mn^III (E_pa = 0.7–0.86 V). So the quenching can be due to some energy transfer between [Ru^II(bpy)₃]²⁺* and [Mn^II(L)₃]²⁺ as previously demonstrated for similar systems, although Mn^II transition are not observed in the visible region due to their spin forbidden character.¹¹

Fig. 6 Stern–Volmer plots in deoxygenated CH₃CN + 0.1 M Bu₄NClO₄ of [Ru^II(bpy)₃]²⁺* (0.013 mM) in the presence of variable concentration of quencher Q, Q = (a) [Mn^II(dmbpy)₃]²⁺, (b) [Mn₂^III,IVO₂(dmbpy)₄]³⁺ and (c) ArN₂⁺.

Luminescence lifetime of [Ru^II(bpy)₃]²⁺* in the presence of the [Mn₂^III,IVO₂(L)₄]³⁺ complexes. In contrast, τ strongly decreases by addition of an increasing concentration of [Mn₂^III,IVO₂(L)₄]³⁺ complexes from 0 to 1.3 mM, the emission decay remaining a single exponential decay (Fig. 6(b)). The k_q values obtained according to eqn. (4) are equal to 2.3 × 10⁹ and 2.5 × 10⁹ s⁻¹ M⁻¹ for the bpy and dmbpy complexes, respectively. The three following processes (eqns. (5–7)) could be considered as possible quenching pathways for [Ru^II(bpy)₃]²⁺*:

[Ru^II(bpy)₃]²⁺* + [Mn₂^III,IVO₂(L)₄]³⁺ → [Ru^I(bpy)₃]⁺ + [Mn₂^IV,IVO₂(L)₄]⁴⁺ (5)

[Ru^II(bpy)₃]²⁺* + [Mn₂^III,IVO₂(L)₄]³⁺ → [Ru^III(bpy)₃]³⁺ + [Mn₂^III,IIIO₂(L)₄]²⁺ (→Mn^II) (6)

[Ru^II(bpy)₃]²⁺* + [Mn₂^III,IVO₂(L)₄]³⁺ → [Ru^II(bpy)₃]²⁺ + [Mn₂^III,IVO₂(L)₄]³⁺* (7)

Eqns. (5) and (6) involve a redox process (reduction and oxidation of [Ru^II(bpy)₃]²⁺*, respectively), while in eqn. (7), the quenching of [Ru^II(bpy)₃]²⁺* is due to an energy transfer process. These three quenching processes may proceed in parallel, the measured quenching rate constants k_q resulting from contributions of all of them. Nevertheless, it is possible to deduce the major quenching mechanism by a simple examination of the different redox potentials of the complexes.

The reductive process following eqn. (5) has to be ruled out since in the most favourable case, this process would be endergonic of 400 mV (E_1/2 Mn^IV–Mn^IV/Mn^III–Mn^IV = 0.86 V for L = dmbpy and E_1/2 Ru^II*/Ru^I = 0.46 V). For the oxidative one, (eqn. (6)), although it is presumably strongly exoergonic, its contribution to the quenching process is rather difficult to evaluate. Indeed the [Mn₂^III,IVO₂(L)₄]³⁺ complexes are irreversibly reduced and the potential measured (E_pc = 0.04 and −0.10 V for bpy and dmbpy, respectively) does not represent the thermodynamic E_1/2 value of the couple [Mn₂^III,IVO₂(L)₄]³⁺/[Mn₂^III,IIIO₂(L)₄]²⁺. Moreover, after a continuous visible irradiation (3 h), no net changes of the absorption spectra of the solutions are detected. Meanwhile, this observation is not a conclusive proof that the quenching following eqn. (6) does not occur. Indeed, as observed from the electrochemical study, these complexes are irreversibly reduced in mononuclear [Mn^II(L)₃]²⁺ species. These latter complexes are then easily reoxidized via the transient photogenerated [Ru^III(bpy)₃]³⁺ one, restoring the initial [Mn₂^III,IVO₂(L)₄]³⁺ complexes, explaining the absence of spectroscopic change during continuous irradiation experiments.

So for thermodynamic and kinetic point of view, the energy transfer (eqn. (7)) represents presumably the main contribution to the quenching process of [Ru^II(bpy)₃]²⁺*. This process is likely owing to the large overlap of emission band centred at 600 nm of [Ru^II(bpy)₃]²⁺* with the visible absorption bands of the [Mn₂^III,IVO₂(L)₄]³⁺ complexes (Table 2).

The origin of the energy transfer process is rather difficult to determined since it might be due to a coulombic interaction (Förster) or a short range electron exchange (Dexter).^33,34 In our case, due to the important likeness of the UV-visible spectra of the binuclear bpy and dmbpy, the values of k_q are similar for both complexes studied and smaller than the diffusion rate constant of the molecules. That could suggest that the interaction is due to a short-range electron exchange, since this process involves the interchange of electrons between [Ru^II(bpy)₃]²⁺* and the [Mn₂^III,IVO₂(L)₄]³⁺ complexes and thus these complexes must diffuse in order to permit the overlap of the orbitals of the two components.³⁵ Moreover, the energy transfer processes allowed by the coulombic interaction are those in which there is no change in spin multiplicity in either component. In the case of [Ru^II(bpy)₃]²⁺*-like complexes, the MLCT excited states are a mixture of three states in rapid Boltzmann equilibrium but largely triplet in character (more than 90%).³⁶ The ground state of [Ru^II(bpy)₃]²⁺ is, on contrary, a pure singlet one which makes the Förster interaction unlikely. Finally we must notified that the Dexter mechanism, which is not favourable if we consider the strong electronic repulsion between the complexes, has already been suspected in the case of similar complexes of Ru^II* and Fe^II.^37,38 After the quenching process, the populated excited state of the [Mn₂^III,IVO₂(L)₄]³⁺ complexes quickly relax to the ground state via a radiationless deactivation pathway.

Continuous photolysis

Photogeneration of [Ru^III(bpy)₃]³⁺. As previously reported,^19,23 addition of an irreversible electron acceptor-like 4-bromophenyl diazonium tetrafluoroborate (ArN₂⁺BF₄⁻) to a solution of [Ru^II(bpy)₃]²⁺ in CH₃CN + 0.1 M Bu₄NClO₄ allows a very efficient production of [Ru^III(bpy)₃]³⁺ under continuous irradiation (quantum yield 0.34). This permanent build up of [Ru^III(bpy)₃]³⁺ arises because of the following electron transfer quenching reaction (eqns. (8) and (9)), the back-electron transfer reaction between [Ru^III(bpy)₃]³⁺ and ArN₂˙ is largely avoided by the fast irreversible evolution of the ArN₂˙ radical into ArH and N₂.

[Ru^II(bpy)₃]²⁺* + ArN₂⁺ → [Ru^III(bpy)₃]³⁺ + ArN₂˙ (8)

The quenching rate constant of [Ru^II(bpy)₃]²⁺* by the electron transfer process following eqn. (8), k_ET, measured using the Stern–Volmer equation is estimated to 3.4 ± (0.2) 10⁹ s⁻¹ M⁻¹ (Fig. 6(c)). Although k_ET value appears to be slightly higher to those of k_EET for the [Ru^II(bpy)₃]²⁺* and [Mn₂^III,IVO₂(L)₄]³⁺ systems (respectively, equal to 2.3 × 10⁹ and 2.5 × 10⁹ s⁻¹ M⁻¹ for bpy and dmbpy), these values are in the same order of magnitude, indicating that the deactivation route of [Ru^II(bpy)₃]²⁺* may follow two competitive pathways. Meanwhile, energy transfer could be advantageously short-circuited by adding a large excess of ArN₂⁺ (as previously obtained for similar binuclear Ru^II–Fe^II complexes¹⁹). In this case the main deactivation process of [Ru^II(bpy)₃]²⁺* involves then an electron transfer to ArN₂⁺ and lead to the [Ru^III(bpy)₃]³⁺ species.

In addition, it should be notified that no interactions have been identified between the mono or binuclear manganese complexes and the diazonium salt.

Photolysis of [Mn₂^III,IVO₂(L)₄]³⁺: generation of [Mn₂^IV,IVO₂(L)₄]⁴⁺. A CH₃CN solution containing the [Ru^II(bpy)₃]²⁺ and [Mn₂^III,IV(O)₂(L)₄]³⁺ complexes in the presence of an excess of ArN₂⁺ has been irradiated with a Mercury lamp (250 W). Fig. 7 illustrates the spectral changes during the photolysis for L = dmbpy. The initial spectra of the solutions result from the overlap of those of the two complexes, and exhibit the intense MLCT band at 432 nm of [Ru^II(bpy)₃]²⁺ slightly distorted due to the absorption of the [Mn₂^III,IVO₂(L)₄]³⁺ complexes in the 400–500 nm region, and the three visible bands of these latter (Table 2). Upon visible irradiation, the evolution of the absorption spectra is quite different according to the bpy and dmbpy complexes.

Fig. 7 Evolution of the visible absorption spectra of a CH₃CN + 0.1 M Bu₄NClO₄ solution containing a mixture of [Mn₂^III,IVO₂(dmbpy)₄]³⁺ (0.7 mM), [Ru^II(bpy)₃]²⁺ (0.05 mM) and ArN₂⁺ (15 mM) under visible irradiation. (A): (a) initial solution, (b) after 140 s, (c) 270 s, (d) 480 s, (e) 680 s, (f) 900 s, (g) 1160 s, (h) 1650 s. (B): (h) 1650 s, (i) 2300 s, (j) 3500 s; l = 1 cm.

In the case of the dmbpy complex, two successive steps are observed on the absorption spectra during irradiation (Fig. 7). In the first step (Fig. 7A), the visible absorption bands at 528, 559 and 688 nm of [Mn₂^III,IVO₂(dmbpy)₄]³⁺ progressively decrease in favour of the emergence of those at 570 and 640 nm, characteristics of the [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺ species. In addition, the intensity of the band at 432 nm increases due to the more intense absorption in this area (between 400 and 500 nm) of the Mn^IV–Mn^IV species than that of the Mn^III–Mn^IV one (Table 2). All these spectral changes are in accordance with the photooxidation of the [Mn₂^III,IVO₂(dmbpy)₄]³⁺ into [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺ by the photogenerated Ru^III species. The maximum yield of formation of [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺ (80%; estimated by spectroscopic titration) is reached after 1650 s of irradiation (Fig. 7, spectrum (h)). This yield is only slightly lower to that obtained by an electrochemical oxidation of [Mn₂^III,IVO₂(dmbpy)₄]³⁺ with or without [Ru^II(bpy)₃]²⁺, demonstrating the efficiency of the photooxidation process. A more prolonged irradiation (up to 3500 s) produces quantitatively [Ru^III(bpy)₃]³⁺. This second step is illustrated by the decrease of the absorbance at 432 nm (Fig. 7B).

These two successive photoredox processes are also confirmed by following irradiation by EPR spectroscopy, the EPR spectra being almost superimposable on those obtained by electrochemical oxidation (Fig. 2). Initially, since [Mn₂^IV,IVO₂(dmbpy)₄]⁴⁺ is EPR silent, the light exposure leads to the progressive fading out of the 16 line EPR spectrum of [Mn₂^III,IVO₂(dmbpy)₄]³⁺ until it almost disappears. Then, in a second step, oxidation of [Ru^II(bpy)₃]²⁺ is verified by the observation of a new weak signal at g = 2.7 typical of the Ru^III form of the photosensitizer. These results are consistent with the fact that [Ru^III(bpy)₃]³⁺ is an efficient oxidant toward [Mn₂^III,IVO₂(dmbpy)₄]³⁺, as expected by comparison of the E_1/2 values of the redox couples Ru^III/Ru^II and Mn^IV–Mn^IV/ Mn^III–Mn^IV (Table 1).

In contrast with [Mn₂^III,IVO₂(bpy)₄]³⁺, the photooxidation process is poorly efficient. A continuous irradiation of a solution containing this complex, [Ru^II(bpy)₃]²⁺ and ArN₂⁺ allows the formation of only 20% of [Mn₂^IV,IVO₂(bpy)₄]⁴⁺ and at the same time of [Ru^III(bpy)₃]³⁺. The yield of the oxidized manganese complex has been determined by electrochemistry by comparison of the height of the Mn^III–Mn^IV/Mn^IV–Mn^IV waves before and after irradiation at a rotating disk electrode.

EPR analysis during irradiation unambiguously confirms the poor efficiency of the photoredox process. Indeed, contrary to the case of the dmbpy derivative, a continuous irradiation leads to a large degradation of the manganese complex into a diamagnetic, unidentified species; this is attested by the progressive decrease of the typical 16 line spectrum of [Mn₂^III,IVO₂(bpy)₄]³⁺ associated with the appearance of a weak 6 line signal due to the formation of a small amount of Mn(CH₃CN)₆²⁺. These results are in good accordance with the electrochemical data, since the E_1/2 value of the Ru^III/Ru^II couple is 40 mV lower than that of the [Mn₂^III,IVO₂(bpy)₄]³⁺/[Mn₂^IV,IVO₂(bpy)₄]⁴⁺. The comparison of the efficiency of the photocatalytic process starting from [Mn₂^III,IVO₂(L)₄]³⁺ demonstrated that the formation of the [Mn₂^IV,IVO₂(L)₄]⁴⁺ oxidized species is thermally governed and very sensitive to the substitution of the bipyridyl ligands.

Photolysis of [Mn^II(L)₃]²⁺: generation of [Mn₂^III,IVO₂(L)₄]³⁺. For this photochemical oxidation, the initial solutions contain a mixture of [Mn(L)₃]²⁺ (1 mM), [Ru^II(bpy)₃]²⁺ (0.05 mM) and an excess of ArN₂⁺ (15 mM). Between 380 and 900 nm the spectra of the solutions are very similar to that of [Ru^II(bpy)₃]²⁺, since neither [Mn(L₃)]²⁺ nor ArN₂⁺ have absorption bands in this area. Upon visible irradiation of the solutions, in both case, the typical visible absorption bands of the [Mn₂^III,IVO₂(L)₄]³⁺ complexes appear progressively (525, 555 and 684 nm for L = bpy and 528, 559 and 688 nm for L = dmbpy) (Fig. 8). These complexes are obtained with a fairly good yield (60–65%) estimated by UV-visible titration at around 685 nm.

Fig. 8 Evolution of the visible absorption spectra of a CH₃CN + 0.1 M Bu₄NClO₄ solution containing a mixture of [Mn^II(dmbpy)₃]²⁺ (1 mM), [Ru^II(bpy)₃]²⁺ (0.05 mM) and ArN₂⁺ (15 mM) under visible irradiation. (A): (a) after 60, (b) 180, (c) 360, (d) 540, (e) 900, (f) 1260, (g) 1620 s; l = 1 cm.

The photochemical induced oxidations of [Mn^II(L)₃]²⁺ by [Ru^III(bpy)₃]³⁺ is efficient for L = bpy and dmbpy, as predicted by the less anodic potential values of Mn^II/Mn^III compared to that of Ru^II/Ru^III (Table 1). EPR spectroscopy attests also the formation of [Mn₂^III,IVO₂(L)₄]³⁺ by the progressive conversion of the initial 6 line signal of the Mn^II complexes into the typical 16 line spectra of the Mn^III–Mn^IV ones.

A prolonged irradiation of the solutions yield similar results to those observed during irradiation of solutions containing [Ru^II(bpy)₃]²⁺, [Mn₂^III,IVO₂(L)₄]³⁺ and ArN₂⁺. As expected, the Mn^IV–Mn^IV species is photogenerated with a good yield only in the case of the dmbpy complex.

Conclusion

This study demonstrates that di-μ-oxo complexes [Mn₂^III,IVO₂(L)₄]³⁺; L = 2,2′-bipyridine and 4,4′-dimethyl-2,2′-bipyridine can be generated with high efficiency by photoinduced oxidation in CH₃CN of the corresponding mononuclear [Mn^II(L)₃]²⁺ using [Ru^II(bpy)₃]²⁺ as photosensitizer in the presence of an aryl diazonium salt acting as an irreversible electron acceptor. A further photoinduced oxidation with the same system into the corresponding [Mn₂^IV,IVO₂(L)₄]⁴⁺ species can be achieved for the complex by having the bipyridine ligands substituted by an electron donating group (L = dmbpy) in agreement with the thermodynamic features of the reaction determined from electrochemical data. It is noteworthy that the efficiency of this photoinduced oxidation is also quite high (only 15% lower than that obtained by electrochemical oxidation) in spite of the strong quenching, presumably by an electronic energy transfer mechanism, of the ³MLCT excited state of [Ru^II(bpy)₃]²⁺ by [Mn₂^III,IVO₂(L)₄]³⁺.

The extension of this work is currently under way using a unimolecular polymetallic complex combining the properties of [Ru^II(bpy)₃]²⁺ and [Mn^II(L)₃]²⁺moieties, a better efficiency in terms of kinetic of the electron transfer between the ground state [Ru^III(bpy)₃]³⁺ and [Mn^II(L)₃]²⁺ through an intramolecular process should be expected.

Acknowledgements

We thank A. Jeunet from Laboratoire d’Etude Dynamiques et Structurales de la Sélectivité, UMR CNRS 5616, Université Joseph Fourier Grenoble France, Institut de Chimie Moléculaire de Grenoble, FR CNRS 2607, for EPR experimental assistance.

References

1. D. Burdinski, K. Wieghardt and S. Steenken, J. Am. Chem. Soc., 1999, 121, 10781.
2. D. Burdinski, E. Bothe and K. Wieghardt, Inorg. Chem., 2000, 39, 105.
3. L. Sun, M. Burkitt, M. Tamm, M. K. Raymond, M. Abrahamsson, D. le Gourriérec, Y. Frapart, A. Magnuson, P. H. Kenéz, P. Brandt, A. Tran, L. Hammarström, S. Styring and B. Akermark, J. Am. Chem. Soc., 1999, 121, 6834.
4. L. Sun, L. Hammarström, B. Akermark and S. Styring, Chem. Soc. Rev., 2001, 30, 36.
5. L. Hammarström, L. Sun, B. Akermark and S. Styring, Spectrochim. Acta, 2001, 37A, 2145.
6. M. L. A. Abrahamsson, H. B. Baudin, A. Tran, C. Philouze, K. E. Berg, M. K. Raymond-Johansson, L. Sun, B. Akermark, S. Styring and L. Hammarström, Inorg. Chem., 2002, 41, 1534.
7. A. Zouni, H.-T. Witt, J. Kern, P. Fromme, N. Krauss, W. Saenger and P. Orth, Nature, 2001, 409, 739.
8. N. Kamiya and J.-R. Shen, Proc. Natl. Acad. Sci. USA, 2003, 100, 98.
9. K. N. Ferreira, T. M. Iverson, K. Maghlaoui, J. Barber and S. Iwata, Science, 2004, 303, 1831.
10. L. Sun, L. Hammarström, T. Norrby, H. Berglund, R. Davydov, M. Andersson, A. Börje, P. Korall, C. Philouze, M. Almgren, S. Styring and B. Akermark, Chem. Commun., 1997, 607.
11. H. Berglund-Baudin, L. Sun, R. Davidov, M. Sundahl, S. Styring, B. Akermark, M. Almgren and L. Hammarström, J. Phys. Chem. A, 1998, 102, 2512.
12. L. Sun, H. Berglund, R. Davydov, T. Norrby, L. Hammarström, P. Korall, A. Börje, C. Philouze, K. Berg, A. Tran, M. Andersson, G. Stenhagen, J. Martensson, M. Almgren, S. Styring and B. Akermark, J. Am. Chem. Soc., 1997, 119, 6996.
13. L. Sun, M. K. Raymond, A. Magnuson, D. LeGourriérec, M. Tamm, M. Abrahamsson, P. H. Kenéz, J. Martensson, G. Stenhagen, L. Hammarström, S. Styring and B. Akermark, J. Inorg. Biochem., 2000, 78, 15.
14. P. Huang, A. Magnuson, R. Lomoth, M. Abrahamsson, M. Tamm, L. Sun, B. van Rotterdam, J. Park, L. Hammarström, B. Akermark and S. Styring, J. Inorg. Biochem., 2002, 91, 159.
15. A. Johansson, M. Abrahamsson, A. Magnuson, P. Huang, J. Martensson, S. Styring, L. Hammarström, L. Sun and B. Akermark, Inorg. Chem., 2003, 42, 7502.
16. A. Magnuson, H. Berglund, P. Korall, L. Hammarström, B. Akermark, S. Styring and L. Sun, J. Am. Chem. Soc., 1997, 119, 10720.
17. M. Sjödin, S. Styring, B. Akermark, L. Sun and L. Hammarström, J. Am. Chem. Soc., 2000, 122, 3932.
18. A. Magnuson, Y. Frapart, M. Abrahamsson, O. Horner, B. Akermark, L. Sun, J.-J. Girerd, L. Hammarström and S. Styring, J. Am. Chem. Soc., 1999, 121, 89.
19. F. Lafolet, J. Chauvin, M.-N. Collomb, A. Deronzier, H. Laguitton-Pasquier, J.-C. Leprêtre, J.-C. Vial and B. Brasme, Phys. Chem. Chem. Phys., 2003, 5, 2520.
20. S. A. Richer, P. K. S. Tsang and D. T. Sawyer, Inorg. Chem., 1989, 28, 2471.
21. M. M. Morrison and D. T. Sawyer, Inorg. Chem., 1978, 17, 333.
22. M. N. Collomb Dunand-Sautier and A. Deronzier, J. Electroanal. Chem., 1997, 428, 65.
23. H. Cano-Yelo and A. Deronzier, J. Chem. Soc., Faraday Trans. 1, 1984, 80, 3011.
24. M.-N. Collomb Dunand Sauthier, A. Deronzier and X. Pradon, J. Am. Chem. Soc., 1997, 119, 3173.
25. M.-N. Collomb Dunand Sauthier, A. Deronzier, A. Piron, X. Pradon and S. Ménage, J. Am. Chem. Soc., 1998, 120, 5373.
26. S. R. Cooper and M. Calvin, J. Am. Chem. Soc., 1977, 99, 6623.
27. M. M. Morrison and D. T. Sawyer, J. Am. Chem. Soc., 1977, 99, 257.
28. D. R Gamelin, M. L. Kirk, T. L. Stemmler, S. Pal, W. H. Armstrong, J. E. Penner Hahn and E. I. Solomon, J. Am. Chem. Soc., 1994, 116, 2392.
29. C. Hureau, G. Blondin, M. Cesario and S. Un, J. Am. Chem. Soc., 2003, 125, 11637.
30. Y. Inada, T. Sugata, K. Ozutsumi and S. Funahashi, Inorg. Chem., 1998, 37, 1886.
31. K. E. Berg, A. Tran, M. K. Raymond, M. Abrahamsson, J. Wolny, S. Redon, M. Andersson, L. Sun, L. Hammarström, H. Toftlund and B. Akermark, Eur. J. Inorg. Chem., 2001, 1019.
32. (a) C. P. Anderson, D. J. Salmon, T. J. Meyer and R. C. Young, J. Am. Chem. Soc., 1977, 99, 1980 . The potentials given in this reference vs. SCE have been converted into Ag/Ag⁺ (0.01 M AgNO₃ in CH₃CN + 0.1 M Bu₄NClO₄) by subtracting 298 mV following (b); (b) V. V. Pavlishchuk and A. W. Addison, Inorg. Chim. Acta, 2000, 298, 97.
33. T. H. Förster, Discuss. Faraday Soc., 1959, 27, 7.
34. D. L. Dexter, J. Chem. Phys., 1953, 21, 838.
35. J. A. Barltrop and J. D. Coyle, in Excited States in Organic Chemistry, John Wiley & Sons Ltd, New York, 1975, ch. 4.
36. E. M. Kober and T. J. Meyer, Inorg. Chem., 1982, 21, 3967.
37. C. Crentz, M. Chou, T. L. Netzel, M. Okumura and N. Sutin, J. Am. Chem. Soc., 1980, 102, 1309.
38. R. H. Schmehl, R. A. Auerbach, W. F. Wacholtz, C. M. Elliott, R. A. Freitag and J. W. Merkert, Inorg. Chem., 1986, 25, 2440.

---