Influence of the Type of Macrocycle on the Stabilisation of the High Oxidation State of the Manganese Ion and Electrode Processes

Abstract

Dinuclear di-µ-oxo complexes of Mn³⁺ and Mn⁴⁺ ions, and mononuclear complexes of Mn³⁺ ions with tetraazamacrocycles ([12]aneN₄, [14]aneN₄, [15]aneN₄) and C-substituted derivative (Me₆[14]aneN₄) as well as mononuclear complexes of Mn²⁺ ions with N-substituted derivatives ((N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄, (N-Me)Me₂py [14]aneN₄) and with oxo₂[14]aneN₄ were studied. Based on spectroscopic (UV VIS and IR) and conductometric studies, the types of synthesised complexes (cis or trans isomers of mononuclear Mn³⁺ complexes, oxygen bridges and class II according to Robin and Day classification for dinuclear complexes) were determined. On the basis of voltammetric and spectroelectrochemical studies, trans-cis isomerisation at the level of Mn²⁺ ion complexes and cis-trans isomerisation at the level of Mn³⁺ ion complexes were demonstrated for complexes of ligands with free C positions. The N-substituted derivatives oxidise according to the EC mechanism, in which the follow-up reaction is a disproportionation reaction. The thermodynamic stabilisation of Mn³⁺ ions was determined by comparing the formal potentials (E_f⁰), the disproportionation constants (k₁) and the formation constants (β^III). The study showed the possibility of oxidation of mononuclear, pseudo-octahedral Mn³⁺ ion complexes to dinuclear complexes and the greatest stabilisation of Mn³⁺ ions, both in monomers and dimers of ligands with free N positions.

1. Introduction

One of the directions for the study of manganese ion complexes results from the thermodynamic and kinetic instability of Mn³⁺ ions, which undergo disproportionation reactions to Mn²⁺ and MnO₂ in aqueous media. Various manganese ion complexes have been studied in different aspects [1,2,3,4,5,6,7], among them azamacrocycles [8,9,10,11], which, relative to their non-cyclic counterparts, enhance the stabilisation of the thermodynamically unstable ion through a macrocyclic effect. This effect results from the superposition of entropic and enthalpic effects, associated with weaker solvation of the cyclic ligand and fewer conformational changes during complex formation.

Another direction of research of manganese ion complexes ensues from the role of this element in natural systems [12]. The structures in which it occurs in them and the way it performs certain functions are still under consideration. Therefore, complexes with geometries close to natural systems are still being studied. Their spectroscopic, structural, electrochemical and catalytic properties are being studied [13,14,15,16,17]. One of the most intensively studied natural systems is the water oxidation complex (OEC), a component of photosystem II (PS II) [18,19]. Studies of the s₀, s₁, s₂ states of the OEC centre indicate a close proximity between manganese atoms with mixed valence values, 16 linear EPR signal and Robin and Day class II type compound [12,20]. Various manganese core structures are considered, most commonly dimer of dimers in which two manganese ions connected by two oxygen bridges are linked to a second dimer via an oxygen bridge and two acetate bridges [21,22]. For this reason, dinuclear complexes with oxygen or acetate bridges are being investigated as structural and functional models, in addition to the most commonly considered tetranuclear complexes [21]. Among azamacrocycles, bi-, tetra- and tri-nuclear complexes of manganese ions have been analysed, e.g., with [9]anN₃, (N-Me)₃[9]anN₃, [12]aneN₄ and [14]aneN₄ [21,22].

Our choice of [14]aneN4 derivatives (Scheme 1) was due to the fact that the dinuclear complex of Mn³⁺ and Mn⁴⁺ ions with [14]aneN₄, described in the literature as a structural model of the s₂ state, was considered not only because of its typical features [23], but due to atypical characteristics such as dimer formation even from dilute solutions, cis-pseudo-octahedral coordination of ligands, H₂O as a source of oxygen bridges [24]. The selection of substituents at different ligand positions allowed us to test these unusual features in the complexes they formed and to evaluate the effect of the substituents on the stabilisation of thermodynamically unstable Mn³⁺ ions in aqueous solutions. Studies conducted for complexes with [12]aneN₄, [14]aneN₄ and [15]aneN₄ (Scheme 1) allowed us to analyse the effect of the size of the coordination cavity on the considered stabilisation process in water and to relate this effect to literature data for manganese complexes with [12]aneN₄ and [14]aneN₄ in acetonitrile [24,25]. We determined the degree of stabilisation by comparing the values of the formation constants and the formal potentials and disproportionation constants determined from electrochemical and spectroscopic analysis of the pH-dependent mechanisms of the electrode processes of the solutions.

2. Results and Discussion

2.1. Selected Aspects of the Synthesis of Complexes

The synthesis procedures used yielded different types of complexes, depending on the nature of the ligand and of the Mn²⁺ ion salt, as determined by spectroscopic studies.

All unsubstituted and C-substituted ligands formed mononuclear Mn³⁺ ion complexes in acidic media (Scheme 2). The very fact that Mn²⁺ ions can be oxidised without the use of an additional oxidant and that stable mononuclear complexes of Mn³⁺ ions and their stable solutions are obtained in neutral medium testify about the significant stabilisation of the high oxidation state by these ligands.

On the other hand, in an inert medium, all these ligands show a notable tendency to form dinuclear complexes of Mn³⁺ and Mn⁴⁺ ions in an air atmosphere, also without the use of an additional oxidant and regardless of the Mn²⁺ ion salt used (Scheme 2). The source of the oxygen bridges is water [24]. Syntheses carried out in aqueous-alcohol solvents of various ratios have shown that for these ligands, even a minimal amount of water is sufficient to form double-core complexes. Moreover, they are formed even in very dilute solutions (10⁻³ mol·dm⁻³). Typically, higher concentrations, at least 10⁻¹ mol·dm⁻³ [24], are required to obtain multicore metal ion complexes. In addition to these, mononuclear complexes of Mn³⁺ and MnO₂ ions may be present in the reaction mixture. Therefore, to obtain only dinuclear complexes, Mn(ClO₄)₂, containing a non-coordinating counter ion, was used, which, with a small number of water molecules and an unusual tendency to form dimers, blocks the formation of mononuclear, hexa-coordinated complexes. Complexes of Mn³⁺ and Mn⁴⁺ ions with [12]anN₄ are formed most rapidly and with the highest efficiency. Oxygen bridges hydrolyse very readily in acidic solutions, which, in such an environment, with Cl⁻ ions undergoing coordination, allowed mononuclear, hexa-coordinated Mn³⁺ complexes to be obtained. [12]anN₄ forms a pink complex with a cis-octahedral structure. For the other unsubstituted and C-substituted ligands, the pink-red Mn³⁺ ion complexes with cis-octahedral structures are transitional forms that transform into light green Mn³⁺ ion complexes with trans-octahedral structures. Thus, cis-trans isomerisation is observed at the level of Mn³⁺ ion complexes. The cis-[Mn^III([12]aneN₄)Cl₂]Cl complex does not isomerise, most likely due to the size of the coordination cavity of the macrocycle. It is the smallest among the investigated ligands and may be too small for the Mn³⁺ ion to enter the space defined by this cavity and form a trans-pseudo-octahedral complex in which the central atom would be in the plane of the four N atoms. Literature data show that even with a Co³⁺ ion, smaller than the Mn³⁺ ion, [12]anN₄ forms only cis-pseudo-octahedral complexes [26]. This result of the synthesis of a mononuclear complex explains the reason for the remarkable ease with which [12]anN₄ forms a dinuclear complex, in which the ligands are held by the manganese cation in cis-pseudo-octahedral coordination. The isomerisation process for the cis- [Mn^III([15]aneN₄)Cl₂]Cl complex occurs after a longer time, ~1 min, which may be due to the greater flexibility of the macrocycle ring caused by the larger size of the coordination cavity and, thus, may facilitate the adoption of a conformation suitable for cis coordination.

A completely different synthesis result was obtained by applying this procedure to N-substituted and oxo₂[14]anN₄ derivatives. In the presence of these ligands, Mn²⁺ ions do not oxidise and, therefore, neither do mixed-valence dinuclear complexes. Mononuclear Mn²⁺ complexes are formed, which do not change their structure regardless of the pH of the solution (Scheme 2). These ligands do not form di-μ-oxo dinuclear complexes with Mn²⁺ ions, most likely due to the steric effect of the substituents and the limited flexibility of the macrocycle caused by the presence of a pyridine ring in the ligand structure, as well as the presence of tertiary and amide N atoms. These groups most likely hinder the macrocycle from adopting a conformation that allows the formation of the cis-pseudo-octahedral monomer structures necessary for the formation of a dimer of the type [Mn^IIIMn^IV(μ-O)₂L₂]³⁺, in which the N atoms form two equatorial and two axial bonds with the manganese ions each (L stands for ligand).

Based on the syntheses carried out in the presence of the investigated ligands, it can be seen that oxidation of the Mn²⁺ ion can occur both when access to the coordination cavity of the macrocycle is unrestricted and when the substituents do not hinder the adoption of a conformation that allows the formation of the cis-pseudo-octahedral structure of the complex.

2.2. Conductometric Investigations

Molar conductivity studies were conducted in ethanol (Et), acetonitrile (AN) and water. The choice of solvents with negligible coordinating properties and lower dielectric permeability constants than water was aimed at obtaining conditions that would enable the structure of the complexes in solutions to be preserved unchanged as far as possible, relative to this structure in the solid state. The results obtained for mononuclear solutions of the complexes in Et (36.2–42.7 S·cm²·mol⁻¹) and in AN (128.4–145.2 S·cm²·mol⁻¹), with the exception of the complex with oxo₂[14]aneN₄, are within the reference ranges for 1:1 type electrolytes (35–45 in Et, 120–160 S·cm²·mol⁻¹ in AN) [11,27]. In the case of Mn³⁺ ion complexes, these values correspond to hexa-coordinated complexes with [14]aneN₄, Me₆[14]aneN₄, [12]aneN₄ and [15]aneN₄, which contain Cl⁻ ions at two positions. In contrast, for Mn²⁺ ion complexes, the values indicate penta-coordinated structures, [Mn^II((N-Me)Me₂py [14]aneN₄)Cl]Cl, [Mn^II((N-Me)₂[14]aneN₄)Cl]Cl·and [Mn^II((N-Me)₄[14]aneN₄)Cl]Cl, in which only one position is occupied by Cl⁻ ions. For the complex of the Mn²⁺ ion with oxo₂[14]aneN₄, both in Et and AN, values of molar conductivities close to zero indicate a non-electrolyte and, thus, a hexa-coordinate complex, [Mn^II(oxo₂[14]aneN₄)Cl₂]. The molar conductivity values of the solutions of the dinuclear complexes (115.0–124.3 S·cm²·mol⁻¹ in Et, 368.4–414.5 in AN) are within the reference range for electrolytes of the 3:1 type (110–130 S·cm²·mol⁻¹ in Et, 340–420 S·cm²·mol⁻¹ in AN) [11,27] and, thus, correspond to complexes of Mn³⁺ and Mn⁴⁺ ions with two oxygen bridges. In contrast, conductivity studies conducted in water for single-core Mn³⁺ ion complexes with [14]aneN₄, Me₆[14]aneN₄, [12]aneN₄ and [15]aneN₄ showed values in the range of 418.2–431.2 S·cm²·mol⁻¹, falling within the reference range for electrolytes of the 3:1 type (408–435 S·cm²·mol⁻¹) [11], thus indicating an exchange of Cl⁻ ligands for water molecules, resulting from the labile nature of these complexes and, thus, complexes of the type [Mn^III(L)(H₂O)₂]Cl₃.

2.3. Spectroscopy

2.3.1. UV/VIS/NIR Spectroscopy of Mononuclear Complexes

On the UV/VIS/NIR spectra of the aqueous solutions of the ligands (with the exception of (N-Me)Me₂py [14]aneN₄) and the complexes of Mn²⁺ ions with (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄ and oxo₂[14]aneN₄ (Figure S1, yellow and blue lines, ESI), no bands are observed but only a slightly increased absorption in the UV range (mainly for the complexes). In contrast, the spectrum of (N-Me)Me₂py [14]aneN₄ (Figure S1, red line, ESI) and its complex with Mn²⁺ (Figure S1, black line, ESI) shows absorption bands that are the result of a bathochromic shift associated with the pyridine group of the macrocycle moiety. In the case of this ligand, the bands observed after complexation are shifted to a greater extent, which may indicate greater thermodynamic stability of this complex.

The spectrum of an aqueous solution of mononuclear Mn³⁺ complex with [12]aneN₄ (Figure 1, black and blue lines) is characterised by bands at 228 (14,580), 304 (4300) and a small band at 698 nm. The absorption bands of aqueous solutions of mononuclear Mn³⁺ complexes with [14]aneN₄, Me₆[14]aneN₄ and [15]aneN₄ (Table S1, ESI) correspond to those reported in the literature, for acetonitrile solutions of these complexes [28,29,30], indicating a deformed trans-pseudo-octahedral geometry of the ligand field in the coordination space of the manganese ion in the third oxidation state, i.e., an elongated tetragonal bipyramid. This is evidenced by bands at the borderline of the near ultraviolet and visible ranges, at 342–398 nm, and weak bands in the 700–712 nm range, corresponding to spin-allowed d-d transitions in the Mn³⁺ ion, in a high-spin configuration. The molar absorption coefficients corresponding to the maximum absorption of the d-d transitions (ε₂) are in the range ~2000–4000 dm³·mol⁻¹·cm⁻¹. Smaller than the others, the value of ε₂ (760 dm³·mol⁻¹·cm⁻¹) in the trans-[Mn^IIIMe₆[14]aneN₄)Cl₂]Cl complex may be related to the steric effect of the six methyl substituents of the ligand [29]. The absorption bands occurring in the UV range, at wavelengths of 271–312 nm, with the largest ε₁ values are indicative of LMCT transitions (Table S1, ESI).

In the trans-[Mn^III([15]aneN₄)Cl₂]Cl complex (Table S1, ESI), the absorption bands originating from the d-d transitions are shifted toward longer wavelengths, most likely due to the asymmetric coordination of the Mn³⁺ ion in the xy plane, by a ligand with less symmetry than [14]aneN₄. This translates into a higher energy d_xy orbital and a lower ${\mathrm{d}}_{{\mathrm{x}}^2-{\mathrm{y}}^2}$ orbital and, therefore, a lower energy difference between the B_2g and B_1g levels, resulting in an increased wavelength λ₂, responsible for the ⁵B_1g → ⁵B_2g and ⁵B_1g → ⁵E_g electron transitions in the trans pseudo-octahedral complex. The reduced symmetry in the Mn³⁺ coordination sphere in the trans-[Mn^III([15]aneN₄)Cl₂]Cl complex relative to the trans-[Mn^III([14]aneN₄)Cl₂]Cl complex is also reflected in the higher ε₂ value, which is also characteristic for complexes of other transition metal ions [31].

In contrast, the shift in the absorption bands on the electron spectrum of the pink solution of the cis-[Mn^III([12]aneN₄)Cl₂]Cl complex towards shorter wavelengths (304 nm), relative to those analysed above, suggests cis-pseudo-octahedral coordination of the central ion, which has been observed for other hexa-coordinated transition metal ion complexes that are trans and cis isomers [32]. Bands originating from d-d transitions in complexes with cis-pseudo-octahedral structures are also characterised by higher values of molar absorption coefficients [32]. The higher value of ε₂ of the cis-[Mn^III([12]aneN₄)Cl₂]Cl complex (4300 dm³·mol⁻¹·cm⁻¹) compared to the trans-[Mn^III([14]aneN₄)Cl₂]Cl complex (1980 dm³·mol⁻¹·cm⁻¹) may be related to the absence of a centre of symmetry in the cis-pseudo-octahedral structure.

2.3.2. UV/VIS/NIR Spectroscopy of Dinuclear Complexes

The VIS spectra of the aqueous solutions of dinuclear Mn³⁺ and Mn⁴⁺ complexes with Me₆[14]aneN₄ and [15]aneN₄ (Figure 2, Table S2, ESI) are very similar to acetonitrile solutions of dinuclear Mn³⁺ and Mn⁴⁺ complexes with [14]aneN₄ and [12]aneN₄ presented in the literature [24,25,33], indicating the presence of dimers in which central ions are connected by oxygen bridges. The multi-nuclear geometry of these complexes is indicated by broad bands observed in the low-energy NIR range, corresponding to transitions between metal atoms at different oxidation states. This is a characteristic feature of the Robin and Day class II compounds [12]. Bands at λ₃ wavelengths (Table S2, ESI) and their intensity indicate LMCT Mn(IV) transitions. Such transitions have been observed in mononuclear complexes [34]. The bands at λ₂, by analogy with the bands of complexes with Schiff bases [35], can be attributed to d-d transitions in Mn⁴⁺ ions. In contrast, the inflections at λ₁, by analogy with the spectra of di-μ-oxo Mn³⁺ and Mn⁴⁺ complexes with bipyridine [36] indicate d-d transitions in Mn³⁺, in dinuclear complexes with mixed valence. In the 300–400 nm range, bands corresponding to transitions characteristic of d-d transitions in the Mn³⁺ ion in a mononuclear complex are absent. Only a band in UV below 300 nm is visible, indicating LMCT transitions [24].

2.3.3. IR Spectroscopy of Complexes

On the IR spectra of macrocycles and their complexes, bands originating from vibrations of methyl, methylene, methanylidene groups, secondary amines, amides and pyridine groups are observed (Table S3, ESI).

Analysis of the literature shows that in the range of deformation vibrations of the amine (930–840) and methylene (830–790) groups, the complexes exhibit different numbers of bands, depending on the symmetry of the complex [37,38,39]. However, the nature of the spectrum in this range does not depend on the type of central ion or ligand. In general, complexes with trans symmetry are characterised by fewer bands in the range involving vibrations of both groups (930–790). These are usually one or two bands in the range of deformation vibrations of the amine groups and one corresponding to vibrations of the methylene groups. In contrast, complexes with cis symmetry correspond to four or five bands in the range of deformational vibrations of the amine groups and two are associated with vibrations of the methylene groups.

The IR spectrum of the single-core complex of the Mn³⁺ ion with Me₆[14]aneN₄ (Figure 3a) at 930–790 cm⁻¹ is characterised by two bands, thus indicating the trans-pseudo-octahedral symmetry of this complex. On the IR spectrum of the complex of the Mn³⁺ ion with [15]aneN₄ (Table S3, ESI), one band more is observed in the vibrational range of the amine groups, which may be due to the lower symmetry in the coordination sphere of this complex, resulting from the nature of the ligand.

In contrast, the IR spectrum of the Mn³⁺ ion complex with [12]aneN₄ (Figure 3b) in the range 840–930 cm⁻¹ is characterised by three bands and in the range 790–830 cm⁻¹ by two bands, which, by analogy with the described classification, allows the conclusion that the structure of the analysed complex is cis-pseudo-octahedral.

In the spectra of the Mn²⁺ ion complexes (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄, oxo₂[14]aneN₄ and (N-Me)Me₂py [14]aneN₄ in the range of deformation vibrations of the methylene and amine groups, the number of absorption bands compared to the spectra of the corresponding ligands is the same (Table S3, ESI).

The IR spectra of all dinuclear Mn³⁺ and Mn⁴⁺ ion complexes (Figure 4, Table S4, ESI) show the same characteristics, analogous to those described in the literature for the [Mn^IIIMn^IV(μ-O)₂([14]aneN₄)₂]³⁺ and [Mn^IIIMn^IV(μ-O)₂([12]aneN₄)₂]³⁺ complexes [25,33]. The main bands occur at ~680 cm⁻¹ and correspond to the vibrations of the [Mn^III(μ-O)₂Mn^IV]³⁺ moiety [36], indicating a dinuclear di-μ-oxo structure of the Mn³⁺ and Mn⁴⁺ ion complexes with the investigated ligand. The range characteristic of the deformation vibrations of the amine and methylene groups (790–930 cm⁻¹) shows an equal number of bands, two coming from the vibrations of the -CH₂- groups and three from the vibrations of the -NH- groups. This corresponds to the structures of complexes in which the ligands adopt cis-pseudo-octahedral coordination with respect to the [Mn^III(μ-O)₂Mn^IV]³⁺ moiety. The strong bands at 1120, 1108 and 626 cm⁻¹ (Figure 4) originate from vibrations of ClO₄⁻.

2.4. Electrochemical and Spectroelectrochemical Investigations

Spectroelectrochemical investigations were used during electrolysis runs with a controlled working electrode potential, allowing spectroscopic analysis throughout the experiments. During a given electrode process, at the potentials of the respective anodic or cathodic peaks, electron spectra were recorded at various times in the course of the process. In this way, a specific peak on the voltammogram was linked to a specific band on the electron spectrum.

Dinuclear complexes of Mn³⁺ and Mn⁴⁺ ions in inert medium oxidise and reduce in two steps in an irreversible and independent manner (one redox system from the other) (Figure S2, ESI), analogous to our previously described [Mn^IIIMn^IV(μ-O)₂(iso [14]aneN₄)₂](ClO₄)₃ complex [40], thus indicating the presence of two redox systems, [Mn^IIIMn^IV(μ-O)₂L₂]³⁺/[Mn^IIIMn^III(μ-O)₂L₂]²⁺ and [Mn^IVMn^IV(μ-O)₂L₂]⁴⁺/[Mn^IIIMn^IV(μ-O)₂L₂]³⁺. The oxidation of Mn³⁺ and Mn⁴⁺ ion dimers to Mn⁴⁺ ion dimers is confirmed by the electron spectra obtained after electrolysis with a controlled working electrode potential, carried out at the oxidation potential of the second redox system (e.g., for [Mn^IIIMn^IV(μ-O)₂([15]aneN₄)₂](ClO₄)₃ at 1 V, see Figure S2). The spectra do not show the absorbance enhancements in the VIS NIR range, characteristic of dinuclear complexes with mixed valence, corresponding to metal–metal transitions (see in Figure S3, ESI, at 800 nm, redline), nor the inflections at ~540 nm, corresponding to d-d transitions in Mn³⁺ ions. There are, however, bands associated with d-d transitions in Mn⁴⁺ ions (at ~550 nm) and with LMCT O-Mn(IV) transitions (at~650 nm). In contrast, the reduction of Mn³⁺ and Mn⁴⁺, during electrolysis at the reduction potential of the first redox system (e.g., for [Mn^IIIMn^IV(μ-O)₂([15]aneN₄)₂](ClO₄)₃ at −0.02 V, see Figure S2), to Mn³⁺ dimers is evidenced by the absence of bands in the spectra obtained after this process (Figure S3, ESI, blue line), as demonstrated in the literature for [Mn^IIIMn^III(μ-O)₂([14]aneN₄)₂]²⁺ in AN [24].

The mononuclear Mn³⁺ ion complexes of [12]aneN₄, [14]aneN₄, Me₆[14]aneN₄ and [15]aneN₄ in acidic medium are reduced to Mn²⁺ ion complexes with the same symmetry, as indicated by only one signal on the voltammetric curves. For the cis-[Mn^III([12]aneN₄)Cl₂]Cl complex, one redox system in the 0.2–0.4 V potential limit (Figure S4, ESI) corresponds to the reduction and oxidation of the cis structures. However, for trans-pseudo-octahedral Mn³⁺ ion complexes, one redox system in the −0.25 to 0 V potential limit (Figure S5, ESI) indicates redox processes occurring in trans structures. Considering the fact that the reduction and oxidation potentials of the cis and trans forms differ, the presence of only one redox system on the voltammetric curve confirms the presence of only one structure of a given complex. The absence of other peaks on the voltammograms during the recording of successive cycles excludes possible isomerisation during the electrode processes recorded in acidic solutions. Very weakly formed oxidation peaks in the case of complexes with trans structures (Figure S5, ESI) may be indicative of the instability of trans-pseudo-octahedral complexes of Mn²⁺ ions, what taking into account the difficulty of obtaining such complexes under current-free conditions, may be justified.

In contrast, mononuclear Mn³⁺ ion complexes with [12]aneN₄, [14]aneN₄, Me₆[14]aneN₄, and [15]aneN₄ in neutral medium, during recording of successive voltammetric curves, are oxidised to dinuclear Mn³⁺ and Mn⁴⁺ ion complexes, as evidenced by the obtained voltammetric curves (Figure 5, black line), analogous to those for the dimers of Mn³⁺ and Mn⁴⁺ ions (Figure S2, ESI) and electronic spectra. The voltammetric curves and electronic spectra were obtained based on spectroelectrochemical measurements.

In the case of cis-[Mn^III([12]aneN₄)Cl₂]Cl, the voltammetric curves in the initial two cycles show only a single, weakly marked redox system (Figure 5, red line), analogous to the curve recorded in acidic medium (Figure S4, ESI), indicating reduction and oxidation of the complex to a complex of the same structure. In subsequent cycles, the formation of a two-step process is already observed (Figure 5, blue, yellow and black lines). The electron spectrum during electrolysis with a controlled working electrode potential, carried out at the oxidation potential of the first redox system (0.4 V) shows the presence of dimers of Mn³⁺ and Mn⁴⁺ ions (Figure 1, yellow line), with a simultaneous decrease in monomer concentration (Figure 1, red line).

The observed process of transition from the cis configuration of Mn³⁺ ion monomer to the di-μ-oxo dimer of Mn³⁺ and Mn⁴⁺ ions must, therefore, be preceded by the exchange of axial Cl⁻ ions for water molecules, deprotonation and dimerization. The former already occurs under current-free conditions, as shown by molar conductivity studies of an aqueous monomer solution, indicating a 3:1 type electrolyte. Dimerization, on the other hand, most likely occurs at the level of mononuclear Mn³⁺ ion complexes, as suggested in the literature for the trans-[Mn^III([14]aneN₄)Cl₂]⁺ complex [24]. We excluded the possibility of dimerization at the level of a mixture of mononuclear Mn³⁺ complexes and Mn⁴⁺ complexes, due to the absence of a peak from the mononuclear Mn⁴⁺ ion complex. Only peaks corresponding to the following redox systems are observed on the voltammetric curve: cis-[Mn^III([12]aneN₄)Cl₂]⁺/cis-[Mn^II([12]aneN₄)Cl₂], [Mn^IIIMn^IV(μ-O)₂([12]aneN₄)₂]³⁺/[Mn^IIIMn^III(μ-O)₂([12]aneN₄)₂]²⁺ and [Mn^IVMn^IV(μ-O)₂([12]aneN₄)₂]⁴⁺/[Mn^IIIMn^IV(μ-O)₂([12]aneN₄)₂]³⁺.

The mode of oxidation of Mn³⁺ complexes with [14]aneN₄, Me₆[14]aneN₄ and [15]aneN₄,determined on the basis of spectroelectrochemical studies is analogous to the trans-[Mn^III(iso[14]aneN₄)Cl₂]Cl complex we described earlier [40]. In the voltammetric curves, in the first cycle there is one redox system (Figure S6, red line), corresponding to the reduction in the complex with the trans structure to a complex with the same structure, in the second cycle two redox systems are already observed (Figure S6, blue line). From about the tenth cycle, a two-step electrode process can be seen, corresponding to the oxidation and reduction in dinuclear complexes of Mn³⁺ and Mn⁴⁺ ions (Figure S6, black line), whose presence is confirmed by the electron spectra of the solutions after the oxidation process, carried out at the potentials of the first redox system. In the case of the investigated complexes, the process of transition from monomers of Mn³⁺ ions with trans configurations to di-μ-oxo dimers of Mn³⁺ and Mn⁴⁺ ions, holding the ligands in cis-pseudo-octahedral coordination, must still be preceded by an additional isomerisation process. In the first cycle, recorded from 0 V, the presence of only one oxidation signal indicates the absence of any other structure of the mononuclear complex of the Mn³⁺ ion and, thus, trans–cis isomerisation at the level of the Mn³⁺ ion complexes does not occur. At the level of the complexes of these ions, an isomerisation process in the opposite direction takes place, as shown by the synthesis of the complexes. Only after the reduction process in the first cycle, another peak is observed on the oxidation curve of the second cycle at a potential of ~0.5 V (Figure S6, blue line), which may indicate, by analogy with the curve for cis-[Mn^III([12]aneN₄)Cl₂]Cl (Figure 5, red line), the cis-pseudo-octahedral structure of the Mn³⁺ complex, which reduces at a potential of ~0.4 V. The isomerisation process, thus, occurs at the level of Mn²⁺ ion complexes. Literature studies have shown such a process in DMSO for trans-[Mn^III([14]aneN₄)Cl₂]⁺ [37]. In contrast, the process of formation of dinuclear complexes of Mn³⁺ and Mn⁴⁺ ions under current conditions from mononuclear complexes with trans-pseudo-octahedral configurations took longer than the formation of [Mn^IIIMn^IV(μ-O)₂([12]aneN₄)₂], for which clear signals were observed already in the sixth cycle. This is reasonable given the cis-pseudo-octahedral symmetry of the mononuclear Mn³⁺ complex with [12]aneN₄ and, thus, one less step in the pathway of formation of the mononuclear complex. Easy dimer formation is also observed under current-free conditions, as demonstrated by the synthesis and low stability of the solutions of cis-[Mn^III([12]aneN₄)Cl₂]Cl complex. The [Mn^IIIMn^IV(μ-O)₂Me₆[14]anN₄)₂]³⁺ complex took the longest time to form, which may be related to the steric effect of the six methyl groups in one ligand, making it difficult to adopt the cis-pseudo-octahedral conformation.

In contrast, mononuclear complexes of Mn²⁺ ions with (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄, (N-Me)Me₂py [14]aneN₄ and oxo₂[14]aneN₄ under both current and current-free conditions, do not oxidise to dinuclear complexes. In the voltammetric curves of these complexes (Figure 6a), irrespective of the pH of the solution, only one redox system is observed at potentials of 0.8–1.1 V. When recording successive cycles (at the same electrode polarisation rate v), no other anodic peaks appear, thus ruling out possible isomerisation in the −0.5–1.4 V potential range. The cathodic peaks are wider than the anodic peaks and, as the rate of electrode polarisation increases, their increasingly complex nature becomes apparent. The currents of the anodic peaks are higher than the currents of the corresponding cathodic peaks and these differences decrease as the scanning rate increases. This character of the curves indicates the EC mechanism of the electrode process, in which the subsequent reaction is a disproportionation reaction of the complexed Mn³⁺ ions [41] and, thus, a weaker stabilisation of these ions by N-substituted macrocycles and oxo₂[14]aneN₄.

The values of the disproportionation reaction rate constants (k₁) were determined from voltammetric curves recorded at different electrode polarisation rates, from the slope of the k₁C⁰τ = f(τ) relation (Figure 6b), based on the values of the kinetic parameters and τ, defined as log = log(k₁c⁰τ) + 0.047(ατ-4) and τ = (E_λ − E⁰_f)/v, where E_λ denotes the potential at which the electrode polarisation changes, according to the procedure presented in [42]. The obtained results (

Table 1. Values of disproportionation constants (k₁) and stability constants of Mn²⁺ and Mn³⁺ complexes.

Ligand	k1	logβ(II)	logβ(III)
[14]anN4	-	11.06 ± 0.06	-
Me6[14]anN4	-	11.12 ± 0.08	-
[12]anN4	-	11.74 ± 0.06	-
[15]anN4	-	10.81 ± 0.07	-
(N-Me)2[14]anN4	3.19 ± 0.14	11.41 ± 0.06	22.68 ± 0.12
(N-Me)4[14]anN4	2.87 ± 0.16	11.35 ± 0.06	22.14 ± 0.10
(oxo)2[14]anN4	4.52 ± 0.12	10.26 ± 0.07	18.27 ± 0.12
(N-Me)Me2py [14]anN4	1.01 ± 0.14	11.98 ± 0.08	24.06 ± 0.11

) indicate that, in this group of complexes, the Mn³⁺ ion is stabilised to the greatest extent by (N-Me)Me₂py [14]aneN₄. On the other hand, to the smallest extent by oxo₂[14]aneN₄, most likely due to the reduced electron density of the two amide N atoms, affecting a worse complexation effect and thus an easier disproportionation reaction. All these ligands, however, stabilise Mn³⁺ ions better than the phenanthroline derivatives we studied earlier [42]—lower k₁ values for the macrocyclic complexes.

The formal potentials (E⁰_f) of the analysed redox systems were determined at low electrode polarisation rates, using the polarisation curve method, due to the non-reversibility of the electrode processes [43]. All the values of the formal potentials of the redox systems [Mn^IIILCl₂]⁺/[Mn^IILCl₂] and [Mn^IIIMn^IV(μ-O)₂L₂]³⁺/[Mn^IIIMn^III(μ-O)₂L₂]²⁺ are lower than the formal potential of the Mn³⁺/Mn²⁺ system and indicate stabilisation of the central ion according to macrocyclic effect, to a degree dependent on the type of the ligand (Table S5, ESI). The Mn³⁺ ion is best stabilised by mononuclear Mn³⁺ complexes with a trans-pseudo-octahedral structure. Formal potential values of the redox trans-[Mn^IIILCl₂]⁺/trans-[Mn^IILCl₂] systems are the lowest, ranging from −0.04 to −0.16 V. Among the dinuclear complexes, the formal potentials of the first steps of the electrode processes are higher, varying between 0.01 and 0.07 V. The lowest value of the formal potential is exhibited by the [Mn^IIIMn^IV(μ-O)₂([12]aneN₄)₂](ClO₄)₃ complex, which is justified, taking into account the fact that it is this complex that is also most easily obtained under current-free conditions. For the complexes [14]aneN₄ and its derivative, the direction of changes in the formal potential values of the Mn^IIIMn^IV(μ-O)₂L₂]³⁺/[Mn^IIIMn^III(μ-O)₂L₂]²⁺ system is opposite to those for the trans-[Mn^IIILCl₂]⁺/trans-[Mn^IILCl₂] one. In the trans-[Mn^IIILCl₂]⁺ complex, electron-donor substituents reduce the stabilisation of the high oxidation state (E⁰_f of the trans-[Mn^IIIMe₆[14]aneN₄Cl₂]⁺/trans-[Mn^IIMe₆[14]aneN₄Cl₂] system is higher than E⁰_f of the trans-[Mn^III[14]aneN₄Cl₂]⁺/trans-[Mn^II[14]aneN₄Cl₂] system, Table S5, ESI), whereas in the [Mn^IIIMn^IV(μ-O)₂L₂]³⁺ complex they increase the stabilisation (the E⁰_f values for the dimers of these ligands are inversely related). The reason for the higher stabilisation of the dimers of smaller ions, Mn⁴⁺ and Mn³⁺, by two ligands containing electrodonor substituents linked by a manganese core into cis-pseudo-octahedral structures may be the lower flexibility of such a structure. The steric effects of the substituents in the case of the manganese core stiffening the structure of the complex may hinder the change to one that would allow the coordination of the larger central ions (two Mn³⁺ ions). In contrast, for the trans-[Mn^III([15]anN₄)Cl₂]⁺/trans-[Mn^II([15]anN₄)Cl₂] and [Mn^IIIMn^IV(μ-O)₂([15]anN₄)₂]³⁺/[Mn^IIIMn^III(μ-O)₂([15]anN₄)₂]²⁺ systems, the highest E⁰_f values (−0.04 V and 0.07 V, respectively) may be related to the larger size of the coordination cavity of this ligand, facilitating the coordination of the larger ion. The even higher value of the formal potential of the cis-[Mn^III([12]anN₄)Cl₂]⁺/cis-[Mn^II([12]anN₄)Cl₂] system (E⁰_f = 0.34 V), indicating weaker stabilisation of the Mn³⁺ ion by this ligand, is due to the different symmetry of this complex.

Significantly higher formal potential values are observed for [Mn^IIILCl]²⁺/[Mn^IILCl]⁺, where L = (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN_4, (N-Me)Me₂py [14]aneN₄ and [Mn^III(oxo₂[14]aneN₄)Cl₂]⁺/[Mn^II(oxo₂[14]aneN₄)Cl₂] (0.8–0.89 V) redox systems, thus indicating a worse stabilisation of the Mn³⁺ ions. The [Mn^II((N-Me)Me₂py [14]anN₄)Cl]Cl complex is the easiest to oxidise (E⁰_f = 0.8 V), while [Mn^II(oxo₂[14]anN₄)Cl₂] is the most difficult (E⁰_f = 0.89 V).

2.5. Potentiometric Investigations

Equilibrium studies of the complexation of Mn²⁺ and Mn³⁺ ions were conducted to determine the thermodynamic stability of the complexes. The formation constants of manganese complexes were determined from potentiometric studies. The formation constants of Mn²⁺ complexes (β^II) were determined based on pH measurements (Figure 7). Mn³⁺ complex formation constants (β^III) were determined from EMF measurements of cells containing the Mn³⁺/Mn²⁺ redox system in the presence of a protonated ligand, at pH = 2 and ionic strength constant μ = 0.1(KCl). Mn³⁺ ions in the cell under study were produced following the addition of the MnO₄⁻ oxidant, to a solution containing, among other species, Mn²⁺ ions. In the medium of pH = 2, in the presence of a complexing agent and a tenfold excess of Mn²⁺ ions, the reaction occurred rapidly and quantitatively and no disproportionation reaction of Mn³⁺ ions was observed. Despite the low pH, complexation of Mn³⁺ ions occurred, as evidenced by the pink solutions obtained during this procedure and the β^II values obtained (Table 1) higher than the protonation constants. The values of the protonation constants (K) of the analysed macrocycles, necessary for the determination of β are available in the literature, but they were also determined assuming that the same conditions of the experiment would reduce the error of the calculated β values.

Values of the protonation and stability constants of Mn²⁺ complexes were calculated using Hyperquad 2003 software [44]. Automating weighting scheme were used for automatic refining of the constants.

The logK values obtained (Table S6) indicate a reduction in the basicity of the macrocycles as a result of the introduction of electrodonor and electroacceptor substituents into the [14]aneN₄ structure, which may be related to the steric effect of the substituents and the mesomeric effect of the pyridine and amide groups of the ligands.

The formation constants of the Mn²⁺ ion complexes (β^II) with the studied macrocycles (Table 1) are relatively small and not significantly different, characteristic of an ion not stabilised by the ligand field. The presence of complexed Mn²⁺ ions in solutions, in the pH range studied, is indicated by the titration curve (Figure 7), the presence of such forms in the Mn²⁺/Mn²⁺L partition curves (Figure S7) and the obtained β^II values (Table 1).

Among the unsubstituted ligands, the complex with [12]aneN₄ shows the highest value, which is most likely due to the cis-pseudo-octahedral structure, for which the size of the central ion is less important than in the case of trans-pseudo-octahedral structures. The values of the formation constants of the complexes with [14]aneN₄ and Me₆[14]anN₄ are similar. This may indicate a balanced result of generally oppositely acting inductive and steric effects related to the presence of methyl substituents in C-substituted [14]aneN₄.

Among the N-substituted and oxo₂[14]aneN₄ ligands, it is the latter that forms the least stable complexes. The reason for this is that the electron density of the two amide N atoms is lowered so much that they do not protonate at all in acidic solution, as shown by protonation equilibrium studies. Also, a relatively low logβ^(II) value is exhibited by the complex of the Mn²⁺ ion with (N-Me)₄[14]anN₄, which may be related to its spatial structural. Literature studies show that the (N-Me)₄[14]anN₄ ligand has an undulating structure due to the sp³ hybridisation of the C and N atoms. The Mn²⁺ ion in complex with this ligand is located above the macrocycle plane, on the same side as the methyl substituents [45]. This spatial orientation may hinder the access of the central ion to the N atoms, which in effect results in a lower stability of the complex of the Mn²⁺ ion with (N-Me)₄[14]anN₄. In contrast, the lower steric effect of the substituents in the N-substituted, dimethyl derivative of [14]anN₄ influences the increase in logβ^(II) value. In this group of ligands, the complex of the Mn²⁺ ion with (N-Me)Me₂py [14]aneN₄ shows the highest logβ^(II) value. Such a result may indicate that the mesomeric effect of the pyridine group, reducing the electron density of the ligand atom, is outweighed by the inductive effect of the methyl substituents, increasing the electron density. The two of methyl substituents in the C positions are unlikely to hinder coordination. An additional factor increasing the stability of this complex may be the weaker solvation of the ligand with less hydrophilicity, which facilitates complexation.

The determination of the values of the formation constants (β^III) of Mn³⁺ ions complexes, which are always accompanied by Mn²⁺ ions in excess in solutions, based on EMF measurements, required the development of a suitable computational programme [46]. The programme was written in Delphi programming environment and is based on the PKAS [47], MINIGLASS and BEST [48] algorithms published in the literature. The input data for the programme, in addition to the concentrations and volumes of solutions, were a curve describing the dependence of the formal potential on the concentration of the oxidised form of the system, with a constant concentration of the reduced form of the redox system and constant concentrations of the ligand and hydrogen ion, as well as the ligand protonation constants and constants for the formation of its complexes with the reduced form of the system, determined in independent experiments. The next step is to select a model of the complexation process that the oxidised form of the metal ion undergoes. The model includes the maximum number of ligands that can bind to the central ion (up to 6), the maximum number of hydrogen ions that can be bound in the protonated complexes (0–3) and the maximum number of hydroxyl ions (0–3) included in the complexes. For each measurement point, a system of mass balance equations [46]. is solved on the basis of the proposed model. The equilibrium concentrations [Mn³⁺], [Mn²⁺] and [L] are solved, which allows to calculate the potential that the studied redox system should have for an assumed set of formation constants. The curve thus obtained is compared with the experimental curve (calculates the sum of squares of deviations) and corrections are made to the initial set of constants. These operations are repeated until convergence is achieved (the sum of squares of deviations of the calculated curve from the experimental curve becomes smaller than the assumed accuracy). The performance of the programme was tested on the Fe(III)/Fe(II) system, for which many values of complex formation constants are available in the literature. This system was tested in the presence of picolinic acid, oxalic acid, citric acid and nitrilotriacetic acid. The results obtained [46], coinciding with those in the literature, indicate that the measurement procedure and the calculation programme work correctly.

The values of the complex formation constants of Mn³⁺ (β^III) ions with (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN_4, oxo₂[14]aneN₄ and (N-Me)Me₂py [14]aneN₄ are higher than the corresponding values of logβ^(II), which, in addition to the smaller radius of the central ion, is influenced by the crystal field stabilisation energy for the d⁴ ion in the weak field of the ligands. The direction of change in the stability of Mn³⁺ ion complexes for this group of ligands is the same as the direction of change in the stability of Mn²⁺ ion complexes.

3. Materials and Methods

3.1. Ligands

1,4,8,11-tetrazacyclotetradecane ([14]aneN₄) and its pyridine, methyl and oxo derivatives ((N-Me)Me₂py [14]aneN₄, Me₆[14]aneN₄, (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄, oxo₂[14]aneN₄), 1,4,7,10-tetraazacyclododecane ([12]aneN₄) and 1,4,8,12-tetraazacyclopentadecane ([15]aneN₄)) were from Sigma Aldrich and were used as received.

3.2. Complexes

3.2.1. Mononuclear Complexes

Mononuclear manganese ion complexes were synthesised based on the previously described procedures [30,40,45].

The solution of MnCl₂·4H₂O (10⁻³ mol) in ethanol (10 cm³) was added dropwise to the solution of ligands (10⁻³ mol) in ethanol (10 cm³)—green-brown solids are formed. (In the case of (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN_4, (N-Me)Me₂py [14]aneN₄ and oxo₂[14]aneN₄−white solids are formed) The reaction mixtures were left with stirring for 1.5 h at room temperature. Addition of concentrated hydrochloric acid (0.1 cm³) gave a red-pink solids (In the case of (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄, (N-Me)Me₂py [14]aneN₄ and (oxo)₂[14]aneN₄ white solids remained unchanged). After ~30 s red-pink solids turn into a clear green precipitate (In the case of [12]anN₄ the precipitate remained pink). The white solids of (N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN₄, (N-Me)Me₂py [14]aneN₄ and oxo₂[14]aneN₄ remained unchanged regardless of time. All precipitates were filtered off, washed with ethanol, recrystallized from water and dried under vacuum at room temperature. As a result of this procedure, the following complexes were obtained:

• trans-[Mn^III([14]aneN₄)Cl₂]Cl·2H₂O; Yield: 54%. Anal. Calcd. for C₁₀H₂₄N₄Cl₃Mn·2H₂O: C, 30.20; H, 7.04; N, 14.09. Found C, 30.39; H, 7.00; N, 13.92;
• trans-[Mn^IIIMe₆[14]aneN₄)Cl₂]Cl·2H₂O; Yield: 44%. Anal. Calcd. for C₁₆H₃₆N₄Cl₃Mn·2H₂O: C, 39.89; H, 8.30; N, 11.63. Found C, 39.97; H, 8.51; N, 11.68;
• cis-[Mn^III([12]aneN₄)Cl₂]Cl·3H₂O; Yield: 53%. Anal. Calcd. for C₈H₂₀N₄Cl₃Mn·3H₂O: C, 24.79; H, 6.71; N, 14.46. Found C, 23.65; H, 6.75; N, 14.51;
• trans-[Mn^III([15]aneN₄)Cl₂]Cl·3H₂O; Yield: 51%. Anal. Calcd. for C₁₁H₂₆N₄Cl₃Mn·3H₂O: C, 30.75; H, 7.45; N, 13.04. Found C, 31.18; H, 7.57; N, 13.08;
• [Mn^II((N-Me)₂[14]aneN₄)Cl]Cl·3H₂O; Yield: 52%. Anal. Calcd. for C₁₂H₂₈N₄Cl₂Mn·3H₂O: C, 35.30; H, 8.33; N, 13.73. Found C, 35.15; H, 7.99; N, 13.65;
• [Mn^II((N-Me)₄[14]aneN₄)Cl]Cl·5H₂O; Yield: 50%. Anal. Calcd. for C₁₄H₃₂N₄Cl₂Mn·5H₂O: C, 35.60; H, 8.89; N, 11.87. Found C, 35.98; H, 8.63; N, 11.91;
• [Mn^II((N-Me)Me₂py[14]aneN₄)Cl]Cl·H₂O; Yield: 52%. Anal. Calcd. for C₁₆H₂₉N₄Cl₂Mn·H₂O: C, 45.62; H, 7.36; N, 13.30. Found C, 45.74; H, 7.22; N, 13.37;
• [Mn^IIoxo₂[14]aneN₄)Cl₂]·5H₂O; Yield: 52%. Anal. Calcd. for C₁₀H₂₂O₂N₄Cl₂Mn·5H₂O: C, 26.92; H, 7.17; N, 12.56. Found C, 26.52; H, 7.02; N, 12.60.

3.2.2. Dinuclear Complexes of Mn(III) and Mn(IV)

Dinuclear Mn(III) and Mn(IV) ions complexes were synthesized according to the previously described procedures [37,40].

The solution of Mn(ClO₄)₂·6H₂O (10⁻³ mol) in ethanol (10 cm³) was added dropwise to the solution of ligands ([14]aneN₄, Me₆[14]aneN₄, [12]aneN₄ and [15]aneN₄) (10⁻³ mol) in ethanol (10 cm³). The reaction mixture was then stirred for 1.5 h at room temperature. A small quantity of green-brown solid was formed. The reaction mixture was left in contact with the atmosphere for several days. A dark green crystalline product formed, which was collected by vacuum filtration, washed with ethanol and dried under vacuum at room temperature. As a result of this procedure, the following complexes were obtained:

• [Mn^IIIMn^IV(μ-O)₂([14]aneN₄)₂](ClO₄)₃·2H₂O; Yield: 54%. Anal. Calcd. for C₂₀H₄₈O₁₄N₈Cl₃Mn₂·2H₂O: C, 27.39; H, 5.93; N, 12.78. Found C, 27.33; H, 5.82; N, 12.71;
• [Mn^IIIMn^IV(μ-O)₂Me₆[14]aneN₄)₂](ClO₄)₃·H₂O; Yield: 45%. Anal. Calcd. for C₃₂H₇₂O₁₄N₈Cl₃Mn₂·2H₂O: C, 37.42; H, 7.21; N, 10.91. Found C, 37.86; H, 7.05; N, 10.80;
• [Mn^IIIMn^IV(μ-O)₂([12]aneN₄)₂](ClO₄)₃·4H₂O; Yield: 81%. Anal. Calcd. for C₁₆H₄₀O₁₄N₈Cl₃Mn₂·4H₂O: C, 22.69; H, 5.67; N, 13.24. Found C, 23.91; H, 5.58; N, 13.20;
• [Mn^IIIMn^IV(μ-O)₂([15]aneN₄)₂](ClO₄)₃·2H₂O; Yield: 51%. Anal. Calcd. for C₂₂H₅₂O₁₄N₈Cl₃Mn₂·2H₂O: C, 29.20; H, 6.19; N, 12.39. Found C, 28.88; H, 6.02; N, 12.36.

Ligands: N-Me)₂[14]aneN₄, (N-Me)₄[14]aneN_4, (N-Me)Me₂py [14]aneN₄ and oxo₂[14]aneN₄ do not form dinuclear complexes.

The chemical composition of the complexes was confirmed by colorimetric analysis, conductivity studies and complexometric titration with EDTA solutions of ascorbic acid and Eriochrome Black T indicator. Colorimetric analysis was based on the measurement of the absorbance of MnO₄⁻ ions at 526 nm, obtained from the oxidation reaction of manganese ion complexes at lower oxidation states by IO₄⁻ ions.

Ligands and complexes solutions were prepared immediately before measurements.

3.3. Other Reagents

MnClO₄·6H₂O and EDTA solution were from Fluka, MnCl₂·4H₂O was from Sigma-Aldrich. KIO₄, Eriochrome Black T, ethanol, acetonitrile, KCl, KMnO₄ and NaOH were purchased from POCH Gliwice, Poland. All of them were used without further purification.

3.4. Instruments and Procedures

UV-VIS spectra were obtained on a PU 8630 Philips spectrophotometer (Philips, Eindhoven, The Netherlands).

Infrared spectra were performed on a Specord M80 (Carl Zeiss, Jena, Germany) spectrometer using KBr pellets.

Conductance measurements were recorded on N-5772 TELECO conductancemeter (TELECO, Wrocław, Poland) at 298 K in ethanolic, acetonitrile and aqueous solutions of the complexes.

Spectroelectrochemistry measurements were performer using spectroelectrochemical instrument SPELEC DROPSENS (Metrohm DropSens, Oviedo, Spain) with reflection probe UV-VIS.

Cyclic voltammetry ware carried using AUTOLAB PGSTAT 10 (Eco Chemie BV, Utrecht, The Netherlands) in a three-electrode system. A glassy carbon disc electrode (GCE) was used as a working electrode. The SCE (MINERAL, Łomianki-Sadowa, Poland) connected to the bulk of the solution by a Luggin capillary was used as a reference electrode. A platinum plate was used as a counter electrode. Measurements were recorded in the presence of an aqueous solution of KCl (0.1 mol·dm⁻³) or HCl (0.1 mol·dm⁻³), as the supporting electrolyte. Solutions were deoxidised with argon before measurements.

Electrolyses with controlled potential were carried out at values corresponding to the respective anodic and cathodic peaks potential, using a spectroelectrochemical instrument, which allowed for recording spectra at any time during electrolysis.

Potentiometric measurements were carried out using microburette controlled by two-channel computer system (Cerko, Gdynia, Poland). For pH measurements (to determine K and β^II) ligand (10⁻³ mol·dm⁻³), HCl (4 × 10⁻³ mol·dm⁻³, and/or MnCl₂ (10⁻³ mol·dm⁻³) and a standard NaOH solution (4.5 × 10⁻² mol·dm⁻³) were used. The exact concentration of the NaOH titrant and the CO₃²⁻ content were determined using the Gran method. The titrated NaOH solutions contained ~1.4% CO₃²⁻. A series of three titrations were performed for each ligand. The duration of one titration step was 40 s, with a permissible pH change of 0.018 pH units. In contrast, a series of five titrations was performed for each ligand in the presence of Mn²⁺, due to the longer equilibrium setting time. The duration of one titration step was 100 s, with a permissible pH change of 0.06 pH units. In both cases, 100 doses of titrant were added to each sample, to a titration fraction of ~4.5. Measurements covered a pH range of ~3 to ~11. pH measurements were carried out using a combined glass electrode with built-in Ag/AgCl as reference electrode (MINERAL, Łomianki-Sadowa, Poland), calibrated each day on three buffers, pH 4.01; 7.00 and 10.00. The slope of the calibration curve was usually within 57.6–57.9 range. For EMF measurements (to determine β^III) ligand (10⁻³ mol·dm⁻³), HCl (10⁻² mol·dm⁻³, MnCl₂ (10⁻² mol·dm⁻³) and a standard KMnO₄ solution (c_1/5 = 10⁻¹ mol·dm⁻³) as oxidising agent, were used. Mn³⁺ ions in the tested cell containing KCl, MnCl₂, ligand and HCl were produced following the addition of the oxidising agent, MnO₄⁻. After each added portion of oxidant, once equilibrium was set, the potential of the tested redox system was measured using a combined electrode (Au, Ag/AgCl, MINERAL, Łomianki-Sadowa, Poland). The permissible EMF variation was set at 0.5 mV (20 readings per packet, packet interval 20 s). A series of five titrations were performed for each ligand in the presence of Mn²⁺. Each included _~40 points. The low number of measurement points is due to difficulties in designing the Mn(II) oxidation experiment resulting from the tendency of Mn(III) to disproportionate. With more concentrated solutions of Mn(II) and ligand, this reaction could not be completely prevented, while with less concentrated oxidant solutions, the oxidation of Mn(II) to Mn(III) did not occur. The Au, Ag/AgCl electrode was tested prior to measurements against a solution containing equimolar amounts of K₄[Fe(CN)₆] (10⁻³ mol·dm⁻³) and K₃[Fe(CN)₆] (10⁻³ mol·dm⁻³). The measure of the suitability of the electrode was the potential corresponding to that of the reference electrode. Triple-distilled water was used to prepare the solutions. All solutions were deoxygenated with argon during the experiments to prevent possible oxidation (by O₂) of Mn²⁺ ions in the presence of ligands and precipitation of MnO₂. The low pH of the medium prevented the disproportionation reaction of the complexed Mn³⁺ ions. The formation constants of mononuclear complexes of Mn³⁺ ions with unsubstituted and C-substituted ligands were not determined, due to the high tendency to form dimers of Mn³⁺ and Mn⁴⁺ ions.

4. Conclusions

In the presence of unsubstituted and C-substituted tetraazamacrocycles, in neutral mediaMn²⁺ form dinuclear Mn³⁺ and Mn⁴⁺ complexes; in acidic media mononuclear Mn³⁺ complexes. The electron spectra of the mononuclear cis- and trans structures are characterised by the same type of bands shifted with respect to each other. IR spectra differ in the number of bands within the ranges characteristic for deformation vibrations of the amine and methylene groups. The electron and IR spectra of the dinuclear complexes show the same features, independent of the ligand type. Mononuclear Mn³⁺ complexes oxidise very readily to dimers, a process preceded by trans–cis isomerisation at the level of the Mn²⁺ complexes. These ligands significantly stabilise the high oxidation state of the manganese ion by the fact that Mn³⁺ ions in complexes with these ligands do not disproportionate and the very low formal potentials (E_f⁰) of the trans-[Mn^IIILCl₂]⁺/trans-[Mn^IILCl₂] and [Mn^IIIMn^IV(μ-O)₂L₂]³⁺/[Mn^IIIMn^III(μ-O)₂L₂]²⁺ redox systems.

Among the N-substituted tetraazamacrocycles and oxo₂[14]aneN₄ ligands. Under current conditions, these complexes oxidise to mononuclear Mn³⁺ ion complexes according to the EC mechanism and undergo disproportionation. These ligands stabilise the high oxidation state of the manganese ion to a much lesser extent, as indicated by the much higher E_f⁰ value and the fact of disproportionation. Among these complexes, Mn³⁺ ions are best stabilised by (N-Me)Me₂py [14]aneN₄, most likely as a result of the predominantly inductive over mesomeric effect and the weaker solvation of the ligand with lower hydrophilicity, as indicated by the lowest values of the disproportionation constant, k₁ and E_f⁰, and the highest value of the stability constant (β^III). In contrast, Mn³⁺ ions are least stabilised by oxo₂[14]aneN₄, due to the reduced electron density of the two amide N atoms. The relation of k₁ and E_f⁰ as well as (β^III) is reversed.