2,2′-Azobisphenolate Ligand Bearing Electron-Donating Methoxy Substituents: Spin-Crossover Ligand Field for a Fe(III) Complex and Stabilization of Ligand-Centered Oxidation Species Due to Its Resonance Effect

Abstract

To investigate the electron-donating substitution effect on spin crossover (SCO) and electrochemical behaviors for homoleptic trivalent metal complexes with 2,2′-azobisphenolate (azp) ligands, the known 2,2′-azobisphenol with 5,5′-dimethyoxy substituents (H₂L^OMe) was synthesized and characterized by NMR spectroscopy and single-crystal X-ray diffraction (SCXRD). The homoleptic Fe^III and Al^III complexes 1-OMe and 2-OMe were prepared along with the homoleptic Fe^III complex with 5,5′-dimethyl-substituted azp ligands 1-Me. The temperature dependence of magnetic susceptibility and SCXRD revealed that 1-OMe exhibits incomplete gradual SCO with the rotational motion of the ligands. Cyclic voltammograms and their simulations revealed that both the Fe^III complex 1-OMe and the Al^III complex 2-OMe exhibit reversible one-electron oxidation waves, indicating that the introduction of the methoxy group enhances the stability of the oxidized species of the complexes. DFT calculations revealed that the oxidation waves for both the Fe^III and Al^III complexes are ligand-centered, while the reduction wave for the Fe^III complexes 1-OMe and 1-Me is metal-centered. In addition, the stability of the oxidized complex species arises from the resonance effect of the methoxy group.

1. Introduction

The phenomenon where molecular structure and/or electronic states switch in response to external stimuli such as temperature, pressure, or light has attracted interest across a wide range of molecular materials fields [1]. Among the molecular switching phenomena exhibited by metal complexes, several are well-known as spin crossover (SCO) [2,3,4,5,6,7], which involves a change in the spin state of a central metal ion between a low-spin (LS) and high-spin (HS) states; valence tautomerism (VT) [8,9,10], an electron transfer reaction between the central metal ion and redox-active ligands; and charge-transfer-induced spin transition (CTIST) [8,9,10], an electron transfer reaction between metal ions in multinuclear metal complexes. In recent years, attempts to control electrical conductivity [11,12,13,14,15], magnetism [16,17,18,19,20], optical properties [21,22,23], polarization [24,25,26], and mechanical properties [27,28,29] using these switching phenomena in metal complexes have also been actively pursued.

If a redox-active ligand gives a SCO complex, SCO or VT can be selectively prepared by either ligand substitution or metal ion exchange. It has been known that tetraoxolene-bridged dinuclear Co and Fe complexes exhibit VT [30] and SCO [31], respectively. Recently, Sato et al. reported SCO at low temperature followed by VT at higher temperature in a tetraoxolene-bridged hetero-dinuclear [FeCo] complex [32]. This suggests that competition between SCO and VT is expected to lead to the development of multi-stage switching or new types of switching phenomena.

2,2′-azobisphenol (H₂L^H) and its derivatives (Figure 1a) are known as intense coloring and fluorometric reagents for various metal ions. We hereafter refer to the deprotonated 2,2′-azobisphenolate dianion (L^R) as the azp ligand. Recently, we discovered that the homoleptic Fe^III complex TMA[Fe^III(L^H)₂] 1-H (TMA = tetramethylammonium) showed a gradual SCO transition (Figure 1b) [33]. After this discovery, we developed the homoleptic SCO Fe^III complexes with the π-extended azp ligand [34] and halogen-substituted azp ligands [35]. Furthermore, the neutral heteroleptic [Fe^III(L^H)(qsal)] complex (qsal = N-(8-quinolyl)salicylaldiminate) [36] and its derivatives [37,38] exhibited cooperative SCO transitions. Other groups have also reported the SCO transitions in the [Fe^III(L^H)₂]⁻ complexes with various cations [39,40,41]. Although the halogen substitution effect on the SCO behavior in the Fe^III complexes with the azp ligands has been reported very recently [35], there is no report on the electron-donating substituent effect on SCO behavior in the Fe^III complexes with the azp ligands.

In addition, recent investigations into the electrochemical properties of redox-innocent Al^III complexes derived from azp ligands with alkyl- and chloro-substituents revealed that the azp ligands are redox-active with substituent effects on their redox potentials [42]. Thus, we intended to investigate the possibility of VT by introducing electron-donating substituents into the SCO complex.

To investigate the electron-donating substituent effect on SCO behaviors and the redox properties in the homoleptic azp complexes, we focused on the methoxy-substituted azp ligand (L^OMe). Schetty and Steiner reported the synthesis of H₂L^OMe but described only the chemical shift of the methoxy substituent in the ¹H NMR spectrum and elemental analysis [43]. In this study, we synthesized the methoxy-substituted derivative H₂L^OMe and its homoleptic Fe^III and Al^III complexes, 1-OMe and 2-OMe, along with the homoleptic Fe^III complex with the methyl-substituted azp ligands, 1-Me. We fully characterized H₂L^OMe using the ¹H and ¹³C NMR spectra and single-crystal X-ray structure analysis. We investigated the crystal structures and magnetic properties of the Fe^III complexes 1-OMe and 1-Me. The Fe^III complex 1-OMe showed an incomplete gradual SCO transition, which is confirmed by the temperature variations in crystal structures. We also investigated the electrochemical properties of the Fe^III complexes 1-OMe and 1-Me, as well as the Al^III complex 2-OMe. The methoxy-substituted complexes 1-OMe and 2-OMe exhibited reversible oxidation waves, whereas the methyl-substituted complex 1-Me showed partially reversible oxidation waves. On the other hand, the Fe^III complexes 1-OMe and 1-Me exhibited a reversible reduction wave at similar potentials, whereas the Al^III complex 2-OMe showed an irreversible reduction wave at the substitution-dependent potentials. The density functional theory (DFT) calculations revealed that all the oxidation waves in the Fe^III and Al^III complexes are ascribed to the ligand center, while the reduction wave for the Fe^III complexes is ascribed to the metal center, and that for the Al^III complexes is ascribed to the ligand center. Furthermore, the resonance effect from the introduction of a methoxy group may contribute to the stability of the oxidized species of metal complexes.

2. Results and Discussion

2.1. Synthesis and Characterization of H₂L^OMe

2.1.1. Synthesis of H₂L^OMe

The azp derivative H₂L^OMe was synthesized according to the method described in the literature [42]. The synthesis scheme is shown in Scheme 1. The azo coupling reaction of the diazonium cation obtained from 2,5-dimethoxyaniline 3 with 4-methoxyphenol gave azobenzene derivative 4 in good yield. The selective demethylation at the ortho position of the azo group in 4 occurred by the reaction with AlCl₃–pyridine, affording the target 5,5′-dimethoxy derivative H₂L^OMe in high yield.

2.1.2. NMR Spectra of 4 and Ligand H₂L^OMe

The ¹H NMR spectra for 4 and H₂L^OMe are shown in Figure 2a and Figure S1, and Figure 2b and Figure S3, respectively. The spectrum of H₂L^OMe in the aromatic region seems to be two anomalous double-doublet signals with very small coupling constants. On the other hand, the spectrum of 4 shows very complex signals. Assuming the number of methyl protons in one methoxy group is set to 3, the integrations of each region are 2, 2, and 2 protons, respectively, which are consistent with the number of aromatic protons in 4. These anomalous spectra can be observed when the frequency difference between two signals is close to their coupling constant. In the present case, these patterns are known as an ABX system [44]. By performing a least-squares fitting of the ABX system using the observed frequencies of each signal for 4 and H₂L^OMe, we can obtain reasonable chemical shifts, relative intensities, and coupling constants for 4 and H₂L^OMe (Figure 2c,d, and Tables S1 and S2).

The ¹³C NMR spectra for 4 and H₂L^OMe are shown in Figures S2 and S4, respectively. The spectrum of H₂L^OMe consists of one and six signals in the alkyl and aromatic regions, respectively, whereas that of 4 consists of three and twelve signals in the alkyl and aromatic regions, respectively. The numbers of signals for each compound are in good agreement with those expected by the chemical structure.

2.1.3. Crystal Structure of H₂L^OMe

In order to confirm the molecular structure of H₂L^OMe, single-crystal X-ray structure analysis was performed. The crystallographic data are listed in Table S3. The crystal structure of H₂L^OMe belongs to orthorhombic Pca2₁. The asymmetric unit of H₂L^OMe contains one molecule. Figure 3a shows the asymmetric unit at 290 K. The C-C bond lengths on the benzene rings are 1.370(6)–1.414(6) Å, and no remarkable bond alternation is observed. The N-N bond length is 1.278(2) Å, which is typical of the N=N double bond. The intramolecular distances of O1···N2 and O2···N1 are 2.600(4) and 2.585(5) Å, respectively, indicating the existence of strong hydrogen bonds. Since the dihedral angle between the two benzene rings is 1.1(2) °, the molecule has an almost planar structure.

The H₂L^OMe molecules are uniformly stacked in a face-to-face manner, forming one-dimensional columns along the b axis (Figure 3b). From the overlap of the two molecules, the azo group is in proximity to the phenyl ring (Figure S5). The intermolecular π-plane distances between the benzene rings in the one-dimensional column are 3.296(4) and 3.266(4) Å, and notable short intermolecular distances between C3 and C6, C6 and C13, and C13 and C10 are 3.323(5), 3.312(3), and 3.299(5) Å, respectively, which are shorter than the sum of van der Waals radii of carbon (1.70 Å) [45], indicative of the existence of strong π-stacking interactions. There are weak C-H···π interactions between one-dimensional columns along the a axis, whereas there is no short contact between one-dimensional columns along the c axis.

2.2. Synthesis and Characterization of M^III Complexes (M^III = Fe^III and Al^III) with Electron-Donating Ligands

2.2.1. Synthesis of M^III Complexes with H₂L^OMe and H₂L^Me (M^III = Fe^III and Al^III)

The homoleptic Fe^III and Al^III complexes were prepared by the reaction of corresponding metal salts with 2 equivalents of azp ligands L^OMe or L^Me, followed by the addition of corresponding onium bromides. Recrystallization in appropriate solvents gave the homoleptic Fe^III complexes 1-OMe and 1-Me as black platelets, and the Al^III complex 2-OMe as black rods, respectively. The single-crystal X-ray structure analysis described below revealed that 1-Me and 2-OMe were obtained as solvate crystals. The compositions of 1-OMe, 1-Me, and 2-OMe were also confirmed by elemental analysis. On the other hand, the thermogravimetry measurements of the air-dried samples 1-Me and 2-OMe suggested that desolvation occurred even at room temperature (Figure S6). In contrast to the ¹H NMR spectra for 4 and H₂L^OMe, 2-OMe gave well-defined signals attributed to the two trisubstituted benzenes, with the signals of the tetraphenylphosphonium ion in the aromatic region (Figure S7).

2.2.2. Magnetic Susceptibility of the Fe^III Complexes 1-OMe and 1-Me

The temperature variation in magnetic susceptibility for 1-OMe is shown in Figure 4. The χ_MT value for 1-OMe at 90 K is 0.45 cm³ K mol⁻¹, indicative of the LS state. On heating, the χ_MT value increases around 250 K. Then, the χ_MT values gradually increase and reach 3.12 cm³ K mol⁻¹ at 400 K. Since the spin-only χ_MT value for the HS Fe^III ion is 4.375 cm³ K mol⁻¹, 1-OMe exhibits an incomplete gradual SCO conversion. No thermal hysteresis is observed for 1-OMe. As the temperature dependence of the magnetic susceptibility of 1-OMe was fitted using the modified regular-solution model [46], the best-fit parameters were ΔH = 26.1 kJ mol⁻¹, ΔS = 70.0 J K⁻¹ mol⁻¹, and Γ = 0.6 kJ mol⁻¹. The fitted and measured curves are shown in Figure S8. Based on these thermodynamic parameters, the transition temperature T_1/2 is 372 K. The thermodynamic parameters closely resemble those of 1-H [33]. Note that there is almost no cooperative behavior.

The temperature variation in magnetic susceptibility for the air-dried sample of 1-Me is shown in Figure S9. The χ_MT value for 1-Me at 300 K is 2.62 cm³ K mol⁻¹, which is an intermediate value between the LS and HS states. This is inconsistent with the crystal structure described below, indicating that 1-Me at 296 K is in the HS state. The elemental analysis of the air-dried sample revealed that the solvent ratio was 0.5 molecule of acetone. Thus, desolvation could induce a change in the spin state. In the subsequent temperature scan, no reversible changes were observed. This suggests that the structure changes as the solvent is removed. Although single-crystal structure analysis is not possible, these results suggest that the desolvated sample of 1-Me may show possible SCO. Further experiments using a different recrystallization solvent or a different counter cation are required to confirm that the [Fe(L^Me)₂]⁻ anion exhibits SCO.

2.2.3. Single-Crystal Structure Analysis

The single-crystal X-ray structure analysis was performed for 1-OMe, 1-Me, and 2-OMe. Since 1-OMe exhibited the SCO transition, temperature variations in crystal structure for 1-OMe were carried out. All the crystallographic data are listed in Table S4.

The crystal structures of 1-OMe at all temperatures belong to isostructural monoclinic P2₁/c, indicating no structural phase transition accompanying the SCO transition. The asymmetric unit of 1-OMe contains one TBA cation and one [Fe^III(L^OMe)₂]⁻ anion. Figure 5a–c show the asymmetric unit of 1-OMe at 90 K. The two ligands in the [Fe^III(L^OMe)₂]⁻ anion exhibit an orientation disorder. We hereafter refer to the major-oriented ligands with C1 to C14 and C15 to C28 atoms in 1-OMe at 90 K as ligands L1 and L2, respectively. The ligands with the opposite minor-oriented ligands and their atoms are labeled with a prime symbol, such as L1′ and C1′. For the TBA cation, the butyl-group orientation is ordered at 90 K (Figure 5c), whereas positional disorder of the butyl groups is observed at 296 and 400 K (Figure 5d). The coordination bond lengths, distortion parameters of the coordination octahedron, and occupancies of L1, L2, and major-positioned cation are listed in

Table 1. Coordination bond lengths and distortion parameters Σ and Θ of the complex anion with major-oriented ligands, and occupancies of major-oriented cation and ligands.

Complex	1-OMe			1-Me	2-OMe	1-H
Temp. (K)	90	296	400	296	90	90	293
M1-O1 (Å)	1.941(2)	1.982(5)	1.978(5)	1.997(4)	1.942(3)	1.9233(15)	1.975(2)
M1-O2 (Å)	1.842(2)	1.829(5)	1.889(5)	1.945(4)	1.807(5)	1.8593(16)	1.947(2)
M1-N1 (Å)	1.920(2)	1.943(4)	2.083(5)	2.168(3)	2.005(3)	1.9523(16)	2.166(2)
M1-O5 (Å)	1.9312(16)	1.954(4)	1.967(4)	1.997(4) a	1.896(2)	1.9217(13)	1.9836(19)
M1-O6 (Å)	1.8586(16)	1.854(3)	1.898(4)	1.945(4) a	1.845(2)	1.8898(13)	1.942(2)
M1-N3 (Å)	1.9153(17)	1.941(4)	2.087(4)	2.168(3) a	2.005(2)	1.9220(14)	2.1537(19)
Σ (°) b	31.4(4)	39.5(8)	65.1(9)	86.2(6)	34.4(5)	39.4(3)	89.6(3)
Θ (°) c	42.4(4)	40.3(9)	104.8(10)	170.5(7)	74.4(6)	53.2(3)	169.6(4)
L1 occupancy	74.6(3)%	68.8(5)%	64.4(6)%	82.5(5)%	72.8(5)%	86.7%	86.7%
L2 occupancy	91.9(3)%	83.2(5)%	81.9(6)%	82.5(5)% a	100%	100%	100%
Cation occupancy	100%	85.5(3)%	58.0(3)%	50.4(3)%	100%	100%	100%
Spin-state	LS	LS	HS	HS	—	LS	HS
Reference	This work	This work	This work	This work	This work	[33]	[33]

^a The values in the table are the same due to a 2-fold rotation relation to the L1 ligand. ^b The sum of absolute deviations of 12 bite angles from 90°. ^c The sum of absolute deviations of 24 angles of 8 surface triangles of the coordination octahedron from 60°.

. The coordination bond lengths and distortion parameters of 1-OMe at 90 and 296 K are very similar to those of 1-H in the LS state [33]. On the other hand, most coordination bond lengths at 296 K are slightly longer than those at 90 K, which is consistent with the observation of a slight increase in the χ_MT value as described in the magnetic susceptibility section. The coordination bond lengths of 1-OMe at 400 K are longer than those at 90 and 293 K, while they are shorter than those of 1-H in the HS state [33]. The distortion parameters at 400 K are also smaller than those in the HS state. These are in good agreement with the occurrence of an incomplete SCO transition at 400 K. Note that the occupancies of both ligands L1 and L2 in 1-OMe are definitely changed from 90 to 373 K. These observations are reminiscent of a SCO-induced pedal-like rotational motion of the azp ligand [35,38].

The crystal structure of 1-Me belongs to monoclinic C2/c. The asymmetric unit of 1-Me contains half of a TBA cation, half of the [Fe^III(L^Me)₂]⁻ anion, and two halves of an acetone molecule. The Fe1 in the complex anion and N3 atoms in the TBA cation are on the two-fold axis. The alkyl chains in the TBA cation show an orientational disorder, whose occupancy is almost 50% (Figure S10). The carbonyl group of one of the acetone molecules is also on the two-fold axis, whereas the other acetone molecule is close to the two-fold axis. Thus, occupancy of the latter acetone molecule is 50%. The coordination bond lengths and distortion parameters of 1-Me are very similar to those of the [Fe^III(L^H)₂]⁻ anion in the HS state [33], suggesting the crystal of 1-Me was in the HS state. Since acetone molecules are present in a one-dimensional channel along the a axis, this is the reason why desolvation occurs even at room temperature.

The crystal structure of 2-OMe belongs to monoclinic P2₁/n. The asymmetric unit of 2-OMe contains one TPP cation, one [Al^III(L^OMe)₂]⁻ anion, one diethyl ether molecule, and one methanol molecule. The ligand L1 exhibits an orientational disorder, whereas the orientation of the ligand L2 is ordered (Figure S11). The coordination bond lengths are quite similar to those in the Al^III complexes with the azp derivatives reported in the literature [42], and are shorter than those of the Fe^III complex due to the smaller effective ion radii of the Al ion [47].

2.2.4. Cyclic Voltammetry

To investigate the redox properties of 1-OMe, 1-Me, and 2-OMe, the cyclic voltammograms were recorded at a scan rate of 100 mV s⁻¹ in a 0.1 M Bu₄NPF₆ dichloromethane solution. The voltammograms for all the complexes are shown in Figure 6 and Figure S12. The redox potentials are summarized in

Table 2. Anodic and cathodic peak potentials ^a (E_pa and E_pc in V vs. Fc/Fc⁺) and the midpoint potentials (E_mid) for the cyclic voltammograms obtained for the complexes 1 and 2.

Complex	First Cathodic Couple			First Anodic Couple			Second Anodic Couple			Third Anodic Couple
	Epa	Epc	Emid b	Epa	Epc	Emid b	Epa	Epc	Emid b	Epa	Epc	Emid b
1-OMe	−1.36	−1.50	−1.43	0.05	−0.07	−0.01	0.34	0.22	0.28	0.83	0.71	0.77
2-OMe	−1.70 c	−1.93 d	−1.82	0.05	−0.05	0.00	0.37	0.27	0.32	0.83	0.72	0.78
1-Me	−1.38	−1.53	−1.45	0.21	0.08 c	0.14	0.56	0.45 c	0.50	—	—	—
2-Me e	−1.73 c	−1.96	−1.85	0.24	0.11	0.18	0.60	0.50	0.55	—	—	—
2-Cl e	−1.50	−1.62	−1.56	0.57	0.40 c	0.49	0.90 d	—	—	—	—	—

^a The scan rate was 100 mV s⁻¹. ^b E_mid = (E_pa + E_pc)/2. ^c Broad peaks. ^d Shoulder peak. ^e Reference [42].

.

For the methoxy-substituted complexes 1-OMe and 2-OMe, each showed two well-defined reversible waves with oxidation and re-reduction peaks, as well as a slightly ill-defined reversible wave (Figure 6a). As is also seen in the figure, the methyl-substituted complex 1-Me showed two couples of oxidation and re-reduction waves similar to those of the Al^III complex with methyl substituents 2-Me [42]. However, their re-reduction peaks were somewhat smaller than the corresponding oxidation peaks. This indicates that the oxidation forms of the methyl-substituted complexes are relatively unstable. Although additional oxidation waves in 1-OMe and 2-OMe were observed in the further anodic scan, the re-reduction waves coupled to the oxidation waves for 1-OMe and 2-OMe became small after the anodic scan exceeded the third oxidation wave (Figure S12). The oxidation potentials of the first to third oxidation waves in 1-OMe and 2-OMe are almost the same values, indicating that the oxidations are ligand-centered and have no significant effect on the central metal ions. A similar observation applies to the oxidation potentials of 1-Me and 2-Me. Therefore, the Fe ion remains trivalent during these oxidation processes. The ligand-centered oxidations of the azp complexes are also supported by a distinct substituent effect on the oxidation potentials between the methoxy and methyl substituents.

To determine the number of electrons responsible for the redox processes, the digital simulation analysis of cyclic voltammograms [48,49] was performed for two well-defined redox waves of 1-OMe and 2-OMe. In this simulation, it has been postulated that the redox mechanism for the voltammetric behaviors of interest can be explained in terms of a two-step electron transfer mechanism:

$$\mathrm{R}-{n_{1}e}^{-}\rightleftarrows {\mathrm{O}}_1$$

$${\mathrm{O}}_1-{n_{2}e}^{-}\rightleftarrows {\mathrm{O}}_2$$

where R is the original reduced form of the complex, O₁ and O₂ are the electrochemically produced one- and two-step oxidized species, respectively, and n₁ and n₂ are the number of electrons for each oxidation step. In this study, we have used conventional, i.e., Butler–Volmer-type kinetic equations for describing the forward rate constant ($k_{\mathrm{f}1}$ or $k_{\mathrm{f}2}$) and backward rate constant ($k_{\mathrm{b}1}$ or $k_{\mathrm{b}2}$) for the respective redox reactions:

$$k_{\mathrm{f}1}=k_1^{°}\exp \left[{\displaystyle \frac{\left(1-{\alpha }_1\right)n_{1}F}{RT}}(E-E_1^{{\circ }^{\prime }})\right], k_{\mathrm{b}1}=k_1^{°}\exp \left[-{\displaystyle \frac{{\alpha }_1{n}_{1}F}{RT}}(E-E_1^{{\circ }^{\prime }})\right]$$

$$k_{\mathrm{f}2}=k_2^{°}\exp \left[{\displaystyle \frac{\left(1-{\alpha }_2\right)n_{2}F}{RT}}(E-E_2^{{\circ }^{\prime }})\right], k_{\mathrm{b}2}=k_2^{°}\exp \left[-{\displaystyle \frac{{\alpha }_2{n}_{2}F}{RT}}(E-E_2^{{\circ }^{\prime }})\right]$$

where $k_1^{\circ }$ or $k_2^{\circ }$ is the standard rate constant, ${\alpha }_1$ or ${\alpha }_2$ is the transfer coefficient, $E_1^{\circ \prime }$ or $E_2^{\circ \prime }$ is the formal potential, and R, T, and F have their usual meanings. The observed and simulated voltammograms are shown in Figure S13, and the simulated parameters are listed in Table S5. It has thus been revealed that the two successive waves observed for 1-OMe and 2-OMe can be thoroughly elucidated in terms of a simple reversible electron transfer by one-electron steps (n₁ = n₂ = 1). Then, the two oxidation species, [M^III(L^OMe)₂]⁰ and [M^III(L^OMe)₂]⁺, have been shown to be sufficiently stable. It is known that the introduction of bulky substituents improves the stability of oxidation species by kinetic control. On the other hand, since the methoxy group is not bulky, we will investigate why the oxidation species are stabilized with theoretical calculations below.

In view of exploring the reductive behaviors of the Fe^III complexes 1-OMe and 1-Me, and the Al^III complex 2-OMe, the potential scan was started in the cathodic direction as shown in Figure 6b. The Fe^III complexes 1-OMe and 1-Me showed one well-defined reversible wave with reduction and re-oxidation peaks. The first reduction waves for 1-OMe and 1-Me appeared at a similar potential around −1.5 V, and thus, there is no significant substituent effect on the first reduction potentials for the Fe^III complexes. Regarding 2-OMe, the voltammogram was ill-defined, probably because of the instability of the reduced species of the complex. When the reduction peak potential of the Al^III complex 2-OMe is compared with those of 2-Me and 2-Cl [42], a distinct substitution effect is observed, which may exhibit a ligand-centered reduction. These observations suggest that the first reduction wave in the Fe^III complexes 1-OMe and 1-Me is metal-centered.

2.2.5. Density Functional Theory Calculations

To give an insight into the origin of the first oxidation and reduction waves, the density functional theory (DFT) calculations were carried out for the [Fe^III(L^R)₂]⁻ anions in the LS and HS states and the [Al^III(L^R)₂]⁻ anions. All the molecules in CH₂Cl₂ could be fully optimized at the B3LYP/6-311+G(d,p) level of theory using the polarizable continuum model (PCM) [50]. The energy level diagram near the frontier molecular orbitals is shown in Figure 7, and their energies are listed in

Table 3. Energy level (eV) near the frontier orbitals for the [Fe^III(L)₂]⁻ and [Al^III(L)₂]⁻ anions.

Anion	[FeIII(LOMe)2]−				[AlIII(LOMe)2]−	[FeIII(LMe)2]−				[AlIII(LMe)2]− a
Spin-State	LS		HS		—	LS		HS		—
	α	β	α	β		α	β	α	β
LUMO + 1	−2.117	−2.105	−2.179	−2.382	−2.178	−2.050	−2.039	−2.116	−2.395	−2.132
LUMO	−2.161	−2.315	−2.303	−2.805	−2.388	−2.128	−2.262	−2.231	−2.770	−2.324
HOMO	−4.796	−4.777	−4.811	−4.795	−4.757	−4.985	−4.959	−4.997	−4.979	−4.946
HOMO − 1	−4.830	−4.810	−4.852	−4.841	−4.762	−4.992	−4.967	−5.016	−4.998	−4.953

^a Reference [42].

.

DFT calculations revealed that for the [Fe^III(L^R)₂]⁻ anions, the highest occupied molecular orbital (HOMO) was the β spin orbital, and the lowest unoccupied molecular orbital (LUMO) was also the β spin orbital. These will henceforth be referred to as β-HOMO and β-LUMO.

The β-HOMO energy levels of the [Fe^III(L^R)₂]⁻ anions are similar to the HOMO energy levels of the [Al^III(L^R)₂]⁻ anions with identical ligands. The difference in the β-HOMO energy level between the methoxy and methyl substituents is considered a substituent effect. Indeed, the first oxidation potentials in both Fe^III and Al^III complexes follow this trend.

The β-LUMO energy levels for the [Fe^III(L^R)₂]⁻ anions vary significantly depending on the spin states: the β-LUMO energy levels of the [Fe^III(L^R)₂]⁻ anions in the LS state are close to the LUMO energy levels of the [Al^III(L^R)₂]⁻ anions with identical ligands. On the other hand, the β-LUMO energy levels of the [Fe^III(L^R)₂]⁻ anions in the HS state are significantly lower than those in the LS state. Since the Fe^III complexes 1-OMe and 1-Me are more easily reduced than the Al^III complexes 2-OMe and 2-Me [42] with identical ligands, the [Fe^III(L^R)₂]⁻ anions in solution should be in the HS state. To confirm the assumption, the magnetic susceptibility of the CDCl₃ solution of the Fe^III complexes was determined using the Evans method [51]. The χ_MT products of 1-OMe and 1-Me are 4.28 and 4.03 cm³ K mol⁻¹, respectively. These values support each Fe^III complex solution in the HS state, because the spin-only χ_MT value for the Fe^III complex in the HS state is 4.38 cm³ K mol⁻¹.

The MO surfaces near the frontier orbitals of the [Fe^III(L^OMe)₂]⁻ and [Fe^III(L^Me)₂]⁻ anions in the HS state are shown in Figure 8. The MO coefficients of β-HOMO for the Fe^III complexes are relatively large on the ligands. Therefore, the first oxidation wave involves oxidation at the ligand center. On the other hand, the MO coefficients of β-LUMO for the Fe^III complexes are relatively large on the Fe^III ion, whereas those of LUMO for the Al^III complexes are nearly absent on the Al^III ion [42]. Therefore, the first reduction wave in the Fe^III complexes originates from the metal center.

To investigate a reason to stabilize the one-electron oxidation species of the methoxy complexes, DFT calculations were carried out for the neutral one-electron oxidized [Al^III(L^OMe)₂] and [Al^III(L^Me)₂] molecules in CH₂Cl₂. The spin density surfaces of the one-electron oxidized neutral [Al^III(L^OMe)₂] molecule along with the neutral [Al^III(L^Me)₂] molecule [42] are shown in Figure 9. The difference in spin density between the [Al^III(L^OMe)₂] and [Al^III(L^Me)₂] molecules is a large spin density distribution on the oxygen atom in the methoxy substituent. Accompanying this change, the spin density on the entirety of the phenolate moiety has relatively decreased. On the other hand, the spin density of the [Al^III(L^Me)₂] molecule is distributed mainly on the hydrogen atom of the methyl substituent, suggesting that the [Al^III(L^Me)₂] molecule may readily induce intermolecular radical reactions. Therefore, stabilization of the radical species of the methoxy-substituted complexes may originate from the resonance effect of the methoxy substituent.

3. Discussion

We summarize the results obtained so far. The Fe^III complexes containing the azp ligands with the methoxy and methyl substituents exhibit SCO in the solid state. These results indicate that the introduction of electron-donating groups into azp ligands can also afford ligand fields that can induce SCO. On the other hand, the Fe^III complexes in solution are in the HS state at room temperature.

The cyclic voltammogram and its digital simulation for the Fe^III complex with the methoxy-substituted azp ligands reveal that it undergoes a reversible one-electron reduction and a reversible two-step one-electron oxidation. The reversible one-electron reduction in the Fe^III complexes with the methoxy- and methyl-substituted azp ligands shows no significant substituent effects, in contrast to the distinct substitution effect on the Al^III complexes. Thus, the one-electron reduction in the Fe^III complexes is metal-centered.

Note that the DFT calculations of the Fe^III complexes reveal that the differences in spin state strongly influence the one-electron reduction behavior. When considering the electrochemical redox reactions of the [Fe^III(L^R)₂]⁻ anions in solution at room temperature, it is necessary to discuss the ligand field in each redox species. In general, Fe^II complexes exhibit a weaker ligand field than Fe^III complexes. In other words, if a Fe^III complex with ligands is in the HS state, it is reasonable to assume that the Fe^II complex with the identical ligands is also in the HS state. This situation applies to the reduction process from the [Fe^III(L^R)₂]⁻ anions in the HS state to the [Fe^II(L^R)₂]²⁻ dianions, suggesting that the [Fe^II(L^R)₂]²⁻ dianions are in the HS state.

In the two-step one-electron oxidation process, the ligand substitution effect and DFT calculations indicate that the two-step one-electron oxidation occurs at the HOMO of ligands. Thus, it is presumed that the effect on the ligand field is due solely to the charge of a ligand. However, the spin states of the oxidized Fe^III ion remain unclear and require future investigation. Note that the oxidized species of the methoxy-substituted Fe^III and Al^III complexes are more stable due to the resonance effect of the methoxy group compared to the Al^III complexes with the other azp ligand derivatives [42].

Based on the above considerations, the electrochemical redox process in this study can be represented by the following electrochemical reaction equilibria:

$${\left[{\mathrm{Fe}}_{\mathrm{HS}}^{\mathrm{II}}{\left({\mathrm{L}}^{2-}\right)}_2\right]}^{2-}\begin{array}{c}-e^{-}\\ \rightleftarrows \\ +e^{-}\end{array}{\left[{\mathrm{Fe}}_{\mathrm{HS}}^{\mathrm{III}}{\left({\mathrm{L}}^{2-}\right)}_2\right]}^{-}\begin{array}{c}-e^{-}\\ \rightleftarrows \\ +e^{-}\end{array}{\left[{\mathrm{Fe}}^{\mathrm{III}}\left({\mathrm{L}}^{2-}\right)\left({\mathrm{L}}^{-·}\right)\right]}^0\begin{array}{c}-e^{-}\\ \rightleftarrows \\ +e^{-}\end{array}{\left[{\mathrm{Fe}}^{\mathrm{III}}{\left({\mathrm{L}}^{-·}\right)}_2\right]}^{+}$$

$${\left[{\mathrm{Al}}^{\mathrm{III}}\left({\mathrm{L}}^{3-·}\right)\left({\mathrm{L}}^{2-}\right)\right]}^{2-}\begin{array}{c}-e^{-}\\ \rightleftarrows \\ +e^{-}\end{array}{\left[{\mathrm{Al}}^{\mathrm{III}}{\left({\mathrm{L}}^{2-}\right)}_2\right]}^{-}\begin{array}{c}-e^{-}\\ \rightleftarrows \\ +e^{-}\end{array}{\left[{\mathrm{Al}}^{\mathrm{III}}\left({\mathrm{L}}^{2-}\right)\left({\mathrm{L}}^{-·}\right)\right]}^0\begin{array}{c}-e^{-}\\ \rightleftarrows \\ +e^{-}\end{array}{\left[{\mathrm{Al}}^{\mathrm{III}}{\left({\mathrm{L}}^{-·}\right)}_2\right]}^{+}$$

where L^3−·, L²⁻, and L^−· are the formal charges and spin states of ligands for tri-anion radical, dianion, and anion radical, respectively. Note that the possibility arises that the neutral trivalent metal complex, ${\left[{\mathrm{M}}^{\mathrm{III}}\left({\mathrm{L}}^{2-}\right)\left({\mathrm{L}}^{-·}\right)\right]}^0$, may exhibit a mixed-valence (MV) state due to two ligands with different valences. The MV state is responsible for various chemical and physical properties such as conductivity, magnetism, dielectrics, optics, catalysis, and so on; therefore, the substitution of metal ions and/or ligands is expected to lead to diverse functionalities among SCO, VT, and MV.

4. Materials and Methods

All the chemicals were purchased and used without further purification. 1,2-bis(5-methyl-2-hydroxyphenyl)diazene (H₂L^Me) was prepared according to the literature [42]. ¹H and ¹³C NMR spectra were recorded on a Bruker Avance 400 spectrometer. Elemental analyses were performed on a Yanaco CHN corder MT-5 elemental analyzer. Thermogravimetry was performed using a Rigaku TG8120 simultaneous thermal analyzer.

4.1. Synthesis of Ligand

4.1.1. 1-(2-Hydroxy-5-methoxyphenyl)-2-(2,5-dimethoxyphenyl)diazene (4)

A solution of 2,5-dimethoxyaniline 3 (5.00 g, 32.6 mmol) in 25 mL of water, 25 mL of methanol, and 8.0 mL of conc. HCl was stirred and cooled to 0 °C. A solution of NaNO₂ (2.48 g, 36.0 mmol) was added dropwise to the solution in 7.0 mL of water. The mixture became a dark red solution and was kept below 5 °C. After confirming the existence of HNO₂ using a KI-starch paper, the solution was quickly transferred to a solution of 4-methoxyphenol (4.05 g, 32.6 mmol) in 25 mL of an aqueous solution of NaOH (2.13 g, 53.2 mmol) under N₂ atmosphere. The reaction mixture was stirred for 2 h below 5 °C and then warmed to room temperature. To this suspension, we added 2 mL of 35% HCl and then extracted with ethyl acetate five times. The solution was dried over MgSO₄ and evaporated. Silica gel chromatography using CH₂Cl₂ as an eluent gave 4 as a brown powder (7.90 g, 84%).

¹H NMR (400 MHz, CDCl₃) δ 13.30 (s, 1H), 7.43 (d, J = 3.34 Hz, H_X), 7.43 (d, J = 3.12 Hz, H_X′), 7.05 (dd, J = 9.01, 3.12 Hz, H_A′), 7.02 (d, J = 9.01 Hz, H_B′), 6.97 (dd, J = 9.0, 3.34 Hz, H_A), 6.97 (d, J = 9.0 Hz, H_B), 3.97 (s, 3H), 3.87 (s, 3H), and 3.85 (s, 3H) ppm. ¹³C NMR (100 MHz, CDCl₃) δ 154.14, 152.64, 151.05, 148.01, 138.90, 137.64, 121.41, 119.93, 119.18, 114.35, 113.62, 99.41, 56.51, 55.92, and 55.87 ppm. Anal. Calcd. for C₁₅H₁₆N₂O₄: C, 62.49; H, 5.59; N, 9.72%. Found: C, 62.30; H, 5.58; N, 9.74%.

4.1.2. 1,2-bis(2-Hydroxy-5-methoxyphenyl)diazene (H₂L^OMe)

To a stirred solution of 4 (7.90 g, 27.4 mmol) in 160 mL of chloroform under N₂ atmosphere, we added portionwise a finely powdered AlCl₃ (14.7 g, 110 mmol). The resulting dark red solution was warmed to 30 °C and then 57 mL of pyridine was added. The solution became purple. The mixture was heated to reflux for 2 h and then cooled to room temperature. After reducing the volume of the solution to about a quarter of its original volume using an evaporator, it was acidified by adding a mixture of 50 mL of 36% HCl and 75 g of ice. The mixture was heated to reflux for 2 h. The resulting precipitate was filtered and dried in vacuo. Recrystallization from ethyl acetate gave H₂L^OMe as reddish-purple needles (6.21 g, 83%).

¹H NMR (400 MHz, CDCl₃) δ 11.93 (s, 2H), 7.18 (d, J = 3.48 Hz, 2H), 6.99 (dd, J = 9.00, 3.48 Hz, 2H), 6.99 (d, J = 9.00 Hz, 2H), and 3.86 (s, 6H) ppm. ¹³C NMR (100 MHz, CDCl₃) δ 153.13, 147.31, 134.72, 121.52, 119.29, 112.99, and 56.01 ppm. Anal. Calcd. for C₁₄H₁₄N₂O₄: C, 61.31; H, 5.15; N, 10.21%. Found: C, 61.26; H, 5.13; N, 10.21%.

4.2. Synthesis of Trivalent Metal Complexes

4.2.1. TBA[Fe^III(L^OMe)₂] (1-OMe)

To a suspension of H₂L^OMe (200 mg, 0.730 mmol) in 12 mL of methanol under N₂ atmosphere, we added a 28% methanol solution of sodium methoxide (0.9 mL, 4.41 mmol) dropwise. The suspension was heated to reflux and stirred for 30 min, to become a dark red solution. To this solution was added a solution of FeCl₃ (59 mg, 0.365 mmol) in 5 mL of methanol. Then the solution was heated to reflux for 1.5 h. To the solution, we added tetrabutylammonium bromide (351 mg, 1.09 mmol) in 18 mL of methanol. The solution was heated again and stirred for 30 min, then cooled to room temperature. After evaporating the solution to a few mL, water was added. The precipitate was filtered and washed with water and dried in vacuo. Recrystallization from acetonitrile–diethyl ether gave 1-OMe as black platelets (276 mg, 89%).

Anal. Calcd. for C₄₄H₆₀FeN₅O₈ (1-OMe): C, 62.70; H, 7.18; N, 8.31%. Found: C, 62.57; H, 7.21; N, 8.45%.

4.2.2. TBA[Fe^III(L^Me)₂]·2acetone (1-Me·2acetone)

To a suspension of H₂L^Me (500 mg, 2.06 mmol) in 13 mL of methanol under N₂ atmosphere, we added a 28% methanol solution of sodium methoxide (0.9 mL, 4.56 mmol) dropwise. The suspension was heated to reflux and stirred for 30 min, to become a dark red solution. To this solution was added a solution of FeCl₃ (167 mg, 1.03 mmol) in 5 mL of methanol. Then the solution was heated to reflux for 1.5 h. To the solution, we added tetrabutylammonium bromide (832 mg, 2.58 mmol) in 10 mL of methanol. The solution was heated again and stirred for 30 min, then cooled to room temperature. After evaporating the solution to a few mL, water was added. The precipitate was filtered and washed with water and dried in vacuo. Recrystallization from acetone–hexane gave 1-Me as black platelets (612 mg, 66%).

Anal. Calcd. for C₅₀H₇₂FeN₅O₆ (1-Me·2acetone): C, 67.10; H, 8.11; N, 7.83%. Found: C, 67.04; H, 8.12; N, 7.98%.

The formula of the air-dried sample for magnetic susceptibility measurement was also confirmed by elemental analysis.

Anal. Calcd. for C_45.5H₆₃FeN₅O_4.5 (1-Me·0.5acetone): C, 67.65; H, 7.86; N, 8.67%. Found: C, 67.28; H, 7.76; N, 8.92%.

4.2.3. TPP[Al^III(L^OMe)₂]·Et₂O·CH₃OH (2-OMe·Et₂O·CH₃OH)

To a suspension of H₂L^OMe (200 mg, 0.730 mmol) in 10 mL of methanol under N₂ atmosphere, we added a 28% methanol solution of sodium methoxide (0.9 mL, 4.41 mmol) dropwise. The suspension was heated to reflux and stirred for 30 min, to become a dark red solution. To this solution was added a solution of Al(NO₃)₃·9H₂O (137 mg, 0.365 mmol) in 5 mL of methanol. Then the solution was heated to reflux for 2 h. To the solution, we added tetraphenylphosphonium bromide (457 mg, 1.09 mmol) in 3.0 mL of methanol. The solution was heated again and stirred for 1 h, then cooled to room temperature. After evaporating the solution to a few mL, water was added. The precipitate was filtered and washed with water and dried in vacuo. Recrystallization from methanol–diethyl ether gave 2-OMe as black rod-shaped crystals (302 mg, 81%). The elemental analysis was performed using a completely desolvated sample.

¹H NMR (400 MHz, CDCl₃) δ 7.79–7.75 (m, 4H), 7.66–7.61 (m, 8H), 7.48–7.43 (m, 8H), 7.23 (d, J = 3.2 Hz, 2H), 7.04 (d, J = 3.3 Hz, 2H), 6.57 (dd, J = 9.0, 3.1 Hz, 2H), 6.55 (dd, J = 9.1, 3.3 Hz, 2H), 6.35 (d, J = 9.0 Hz, 2H), 6.27 (d, J = 9.2 Hz, 2H), 3.75 (s, 6H), and 3.74 (s, 6H) ppm. Anal. Calcd. for C₅₂H₄₄AlN₄O₈P (2-OMe): C, 68.57; H, 4.87; N, 6.15%. Found: C, 68.40; H, 4.81; N, 6.17%.

4.3. NMR Spectrum Fitting

Theoretical expressions for the energies (Hz) and relative intensities of the 14 allowed transitions in the ABX-spin system are described in the book [44]. The sum of the squares of the differences between the observed signal energies and the corresponding calculated energies is minimized using Excel’s Solver. Since the signals from the protons H_X and H_X′ in 4 overlap with each other, all the energies (ν_A′, ν_B′, ν_X′) and coupling constants (J_A′B′, J_A′X′, J_B′X′) are fitted with 8 observed signal energies derived from the protons H_A′ and H_B′ in 4. On the other hand, the energy differences between the protons H_A and H_B for H₂L^OMe and 4 are too small and become zero after minimization. Thus, their energies ν_A and ν_B are fixed as the coefficient of determination would be as high as possible. The energy ν_X and coupling constants (J_AB, J_AX, J_BX) are fitted with 6 observed signal energies in H₂L^OMe. Since the available signals derived from the protons H_A and H_B in 4 are limited, the energy (ν_X) and coupling constant (J_AX) are fitted with 4 observed signal energies derived from the protons H_A and H_B in 4.

4.4. Magnetic Susceptibility Measurements

Variable-temperature direct-current magnetic susceptibilities of polycrystalline samples were measured on a Quantum Design MPMS-XL magnetometer under a field of 0.5 T in the temperature range of 2 or 5 to 400 K. The magnetic susceptibilities were corrected for diamagnetic contributions estimated by Pascal constants [52].

Magnetic susceptibilities of a ca. 0.47 mM chloroform-d solution of 1-OMe and 1-Me at room temperature were determined by the Evans method [51] using a Bruker Avance 400 spectrometer. The solution samples were prepared using a Wilmad^® 517-COMPLETE coaxial tube.

4.5. Crystal Structure Determinations

A crystal was mounted on a MiTeGen MicroMounts^TM. All data except H₂L^OMe were collected on a Bruker APEX II CCD area detector with monochromated Mo-Kα radiation generated by a Bruker Turbo X-ray Source coupled with Helios multilayer optics. A Nihon Thermal Engineering nitrogen gas flow temperature controller was used for the temperature variable measurements. All data collections except H₂L^OMe were performed using the APEX2 crystallographic software package (Bruker AXS). The APEX3 crystallographic software package (v2019.1-0, Bruker AXS) was used to determine the unit cell parameters. Data were integrated by using SAINT. Numerical absorption correction was applied by using SADABS. Diffraction data of H₂L^OMe were collected on a Rigaku XtaLAB Synergy X-ray diffractometer using multi-layer mirror monochromated Cu-Kα radiation generated by a PhotonJet microfocus X-ray source and a HyPix detector. The CrysAlis^Pro software (v1.171.43.144a, Rigaku Oxford Diffraction) was used to control the diffractometer and to integrate and correct the data. All the structures were solved by direct methods and refined by full-matrix least-squares methods based on F² using the SHELXTL program (v2019/3). All non-hydrogen atoms were refined anisotropically. Hydrogen atoms were generated by calculation and refined using the riding model. CCDC 2575300–2575305 contains the supplementary crystallographic data for this paper. These data can be obtained free of charge via https://www.ccdc.cam.ac.uk/structures/? (accessed on 23 July 2026) (or from the CCDC, 12 Union Road, Cambridge CB2 1EZ, UK; Fax: +44 1223 336033; E-mail: deposit@ccdc.cam.ac.uk).

4.6. Cyclic Voltammetry

Cyclic voltammograms were recorded using an Ivium Technologies CompactStat.h standard electrochemical interface. A conventional three-electrode electrolytic cell was used. Platinum wires were used as the working and counter electrodes. A BAS RE-7 nonaqueous reference electrode (Ag/Ag⁺) was used as the reference electrode. A few mg of the complexes were dissolved in a 0.1 M Bu₄NPF₆ dichloromethane solution. Nitrogen gas was bubbled through the solution before the measurement. Ferrocene (Fc) was used as an internal reference to correct the potential axis by using the midpoint potential for the Fc/Fc⁺ redox couple. The calculation of cyclic voltammograms was performed by the normal explicit finite difference digital simulation technique [48,49].

4.7. Density Functional Theory (DFT) Calculations

The geometry optimizations for the Fe^III and Al^III complex anions were performed at the B3LYP level [53,54] using the Gaussian 16 program package [55]. The 6-311+G(d,p) basis sets for H, C, O, N [56,57], and Al [58,59,60] atoms were used. The Wachters–Hay [61,62] basis sets were used for the Fe atom. The atomic coordinates for the [Fe^III(L^OMe)₂]⁻ anion in the LS and HS states were taken from the major component of the disordered ligands in the crystal structure of 1-OMe at 90 and 400 K, respectively. The atomic coordinates for the [Fe^III(L^Me)₂]⁻ anion in the HS state were taken from the major component of the disordered ligands in the crystal structure of 1-Me at 296 K, whereas those for the [Fe^III(L^Me)₂]⁻ anion in the LS state were obtained by methyl substitution of the optimized coordinates for the [Fe^III(L^H)₂]⁻ anion from the major component of the disordered ligands in the crystal structure of 1-H at 90 K [33]. The atomic coordinates for the [Al^III(L^OMe)₂]⁻ anion were taken from the major component of the disordered ligands in the crystal structure of 2-OMe. The atomic coordinates for the neutral one-electron oxidized Al^III complexes were taken from the optimized [Al^III(L^OMe)₂]⁻ anion under vacuum. After the geometry optimization under vacuum, the geometry optimizations in the presence of CH₂Cl₂ were performed using the polarizable continuum model (PCM) [50]. No imaginary frequencies were found for any of the calculated optimized structures. Cartesian coordinates of all the complex anions calculated at the B3LYP level of theory are summarized in Tables S6–S11.

5. Conclusions

We synthesized and characterized the azp derivative bearing electron-donating methoxy groups, H₂L^OMe, and elucidated the structures, magnetic properties, and electrochemical properties of its homoleptic Fe^III and Al^III complexes, 1-OMe and 2-OMe, along with the methyl derivative 1-Me. The Fe^III complex composed of the methoxy-substituted azp ligands, 1-OMe, exhibited SCO, revealing that the azp ligand provides a ligand field that induces SCO regardless of any substituents introduced. Note that cyclic voltammograms for both the Fe^III complex 1-OMe and the Al^III complex 2-OMe exhibit reversible oxidation waves, indicating that the introduction of the methoxy group enhances the stability of the oxidized species of the complexes. This is the first observation of stable oxidized complex species from the azp complexes. DFT calculations revealed that the oxidation waves for both the Fe^III and Al^III complexes, 1-OMe and 2-OMe, are ligand-centered, while the reduction wave for the Fe^III complexes 1-OMe and 1-Me is metal-centered. In addition, the stability of the oxidized complex species originates from the resonance effect of the methoxy group. The stabilization of the oxidized complex species by the methoxy substitution opens the potential of the chemistry of metal complexes with azp ligands to exhibit not only SCO and VT phenomena but also a new functionality derived from MV. Further progress in research along this line is expected.