Charge Transfer Chromophores Derived from 3d-Row Transition Metal Complexes

A series of new charge transfer (CT) chromophores of “α-diimine-MII-catecholate” type (where M is 3d-row transition metals—Cu, Ni, Co) were derived from 4,4′-di-tert-butyl-2,2′-bipyridyl and 3,6-di-tert-butyl-o-benzoquinone (3,6-DTBQ) in accordance with three modified synthetic approaches, which provide high yields of products. A square-planar molecular structure is inherent for monomeric [CuII(3,6-Cat)(bipytBu)]∙THF (1) and NiII(3,6-Cat)(bipytBu) (2) chromophores, while dimeric complex [CoII(3,6-Cat)(bipytBu)]2∙toluene (3) units two substantially distorted heteroleptic D-MII-A (where D, M, A are donor, metal and acceptor, respectively) parts through a donation of oxygen atoms from catecholate dianions. Chromophores 1–3 undergo an effective photoinduced intramolecular charge transfer (λ = 500–715 nm, extinction coefficient up to 104 M−1·cm−1) with a concomitant generation of a less polar excited species, the energy of which is a finely sensitive towards solvent polarity, ensuring a pronounced negative solvatochromic effect. Special attention was paid to energetic characteristics for CT and interacting HOMO/LUMO orbitals that were explored by a synergy of UV-vis-NIR spectroscopy, cyclic voltammetry, and DFT study. The current work sheds light on the dependence of CT peculiarities on the nature of metal centers from various groups of the periodic law. Moreover, the “α-diimine-MII-catecholate” CT chromophores on the base of “late” transition elements with differences in d-level’s electronic structure were compared for the first time.


Introduction
Originating in ancient time as a household craft [1], dyes chemistry has received welldeserved recognition and the widest industrial application, working on the diverse needs of society and making our life unthinkable without the existence of colorants (textile, food, cosmetics, paint and varnish manufacturing, etc.). At present, a search for new effective colorants (dyes or chromophoric systems), corresponding to newly emerging challenges, remains one of the most demanded chemical fields, which involves a numerous range of different technologies to generate and design a huge set of chromophoric compounds and materials [2].
Summarizing, the further molecular design of CT colorants should be continued as a prospective, targeted action, directed to a synergy between fundamental research and a response to practical challenges. Involving the derivatives of 3d-row "late" transition elements (first of all, Cu and Co) other than 10 group metals promises to be interesting: considerable results can be achieved, taking into account the diverse structure of d-levels, and lability in the formation of a coordination environment. Following above trend, we present CT chromophores 1-3 of 3d-row "late" transition metals (Cu, Ni, Co) based on 4,4 -di-tert-butyl-2,2 -bipyridyl and 3,6-di-tert-butyl-o-benzoquinone in terms of used synthetic approaches, molecular structures' peculiarities, as well as CT characteristics, Cu II (3,6-Cat)(bipy tBu ), Ni II (3,6-Cat)(bipy tBu ) and [Co II (3,6-Cat)(bipy tBu )] 2 complexes, respectively. Interrelations between CT energy, structural features, and the nature of nodal metal center are explored in detail.
As an interesting fact, the ability to manage the CT chromophore's composition through a fine-tuning of the synthetic procedure (hydrous/anhydrous conditions), which allows generating monomer or dimeric species, was illustrated by the example of (3,5-ditert-butyl-catecholato)(2,2 -bipyridine)copper(II) derivatives [47]. Moreover, an oligomerization can be considered as a frequent enough phenomenon for homoleptic transition metal derivatives of 3,5-di-tert-butyl-o-benzoquinone type platform: a dimerization of bis-obenzosemiquinonato Cu II moieties [97], as well as a tetramerization of o-benzosemiquinonato Co II units [98] are fulfilled due to a dative bonding between oxygen atoms and metal ions. In contrast, a formation of monomeric molecular structures is a typical situation for coordination compounds derived from 3,6-di-tert-butyl-o-benzoquinone. A higher degree of the oxygen atom's steric hindrance prevents the existence of oligomeric species. In this regard, an obtaining of dimeric [Co II (3,6-Cat)(bipy tBu )] 2 complex (3), bearing two different bulky organic parts, is a noteworthy event. Summarizing, our strategy to prevent the oligomerization of D-M II -A parts by shielding of coordinating atoms from ligands due to an incorporation of substantially bulky tert-butyl groups turned out to be effective only in the Cu II derivative 1.
The synthesis of complexes 1-3 requires conditions excluding air moisture and oxygen. However, the resulting compounds are quite air stable and can be stored in open vessels in a crystalline form.

Molecular Structures of Complexes 1-3
It is acknowledged that the "α-diimine-M II -catecholate" type complexes have a squareplanar structure. Hence, the planar mutual arrangement of overlapping HOMO and LUMO, facilitates an effective intramolecular ligand-to-ligand charge transfer [30,32,41]. As a consequence, to date, there is a numerous set of LL'CT heteroleptic chromophores, each of which is distinguished by a planar molecular structure [99].
In turn, complexes 1 and 2 were no exception. The coordination polyhedra possess a minimal distortion (Figure 2), in accordance with values τ 4 (1, 2) = 0.05 [100,101]. Moreover, an ideal planarity (excluding tert-butyl groups) is featured for molecule of 1, since the molecule lies on a special position. As a result, zero angles are formed by ligand planes, as well as by O (1)  Notes: *-atoms from the second part of oligomer (Symmetry transformations used to generate equivalent atoms: -x + 1, -y, -z + 1); **-calculated by the authors of this work;
The "metal-heteroatom" lengths for Ni II complex 2 are almost identical to those presented for Ni II (3,6-Cat)(bipy) analog [32]. In turn, the same distances in Cu II (3,5-Cat)(bipy) [47] slightly exceed ones for Cu II derivative 1 (Table 3). A non-equidistance is observed for Co-O lengths owing to dimerization with a participation of oxygen atoms. As expected, the Co-O and Co-N bonds in 3 are elongated substantially compared to Co II (3,6-Cat)(DAD) ( Table 3). This fact is in a good agreement with a growth in the ionic radius of cobalt as the coordination number increases [106], as well as with a higher steric hindrance of metal center in complex 3.
The THF molecule in crystal 1 is disordered over two positions and is oriented by the oxygen atom to the hydrogen atoms of the bipyridine ligand. As a result, weak intermolecular hydrogen interaction O(THF)· · · H(16A) (2.653 Å) or O(THF)· · · H(19A) (2.600 Å) is realized.

1·THF 2 (Molecule A)/(Molecule B) 3·Toluene
Bond Å Bond Å Bond Å Table 3. Values of "metal-heteroatom" distances in some known monomeric "α-diimine-M IIcatecholate" type complexes. An ideal lamellar crystal structure without π-stacking is inherent for complex 1 ( Figure 3). The distances between molecular layers are equal to 4.559 Å. A similar situation with weakly interacting molecular layers was observed for a series of Pd II compounds on the base of perfluoroalkyl-bearing catechol and various substituted bipyridines [68]. In contrast, the stabilization of a crystal structure is characterized for less bulky Ni II (3,6-Cat)(bipy) reported recently [32] through intermolecular π-π contacts of 2,2 -bipyridines. Indeed, an existence of π-stacking can be assumed as a distinctive feature for many lamellar CT chromophores, containing conjugated organic platforms with a rigid carbon backbone [32,68,72,107,108]. Nevertheless, the presence of bulky tert-butyl groups in donor/acceptor ligands, as well as solvate THF molecules (one molecule per molecule of complex) in 1 makes any π-π interactions impossible. Molecules of complex 1 in odd layers are strictly under each other, and a similar pattern is observed for molecules of even layers. So, the complex 1 demonstrates a mutual molecular arrangement, minimizing a repulsion of tert-butyl moieties ( Figure S1). ture for many lamellar CT chromophores, containing conjugated organic platforms with a rigid carbon backbone [32,68,72,107,108]. Nevertheless, the presence of bulky tert-butyl groups in donor/acceptor ligands, as well as solvate THF molecules (one molecule per molecule of complex) in 1 makes any π-π interactions impossible. Molecules of complex 1 in odd layers are strictly under each other, and a similar pattern is observed for molecules of even layers. So, the complex 1 demonstrates a mutual molecular arrangement, minimizing a repulsion of tert-butyl moieties ( Figure S1). Despite the lamellar laying realized in 1, in general, the herringbone-like packing folded from T-motifs become the most preferable "building blocks" for a majority of CT chromophores of D-M II -A type with hindered organic moieties [32,67,73,87,89,90,95,109]. As a rule, molecules within such T-motif are stacked through the π-π interplays involving ligand orbitals, while there are no substantial π-π inter-stack contacts because of a steric shielding effect from bulky substituents in ligands (for instance, tert-butyls [95], iso-propylphenyls [73], perfluoroalkyls [68], etc). Complex 2 demonstrates the same pattern. The crystal structure contains isolated pairs of crystallographically independent molecules A and B (Figure 3). These T-motifs have a mutual arrangement close to perpendicularity, since molecular planes form angles at 72.31°/77.82°/83.37° (between molecules from different pairs-(A and A)/(A and B)/(B and B), respectively. Thus, the shortest contacts in dimeric pair are found: (1) between centroids of bipyridine ring Cdbipy and chelate cycle CdCh, which is adjacent to catecholate dianion (CdCh(A)···Cdbipy(B) 3.735 = Å, CdCh(B)···Cdbipy(A) 3.783 = Å); (2) between nickel atoms (Ni(1A)···Ni(1B) 3.635 Å), Figure S2).
The molecular packing of compound 3, continuing the tendencies for derivatives 1 and 2, is organized according to the principle of least repulsion between complex' molecules and solvate toluene molecules ( Figure 3). The plane of the C 6 -ring of the catecholate and the NC 5 -ring of the bipyridine ligand of the neighboring molecule are almost perpendicular to each other. The corresponding dihedral angle is 74.68 • . Thus, the possibility of realizing the π-π interaction is completely excluded. However, the shortest contact Cd Cat · · · C(23) = 3.695 Å (where Cd Cat is centroid of C 6 -ring, Figure S3) may indicate the presence of intermolecular C-H· · · π interaction in the crystal of 3.
It should be noted that an incorporation of sterically hindered functions into the ligands can be assumed as a rational strategy to increase the solubility of the complex. A tendency to π-π intermolecular stacking is weakening. For instance, a problem of solubility has been raised for Cu II CT chromophores [44]. As a result, all studied coordination compounds 1-3 possess good solubility in a wide range of solvents with various polarity.

EPR Investigation for Complex 1
It is known that square-planar Ni II complexes are characterized by a low-spin state of the metal ion. As a result, compound 2 is diamagnetic, as evidenced by a well-resolved NMR spectrum. A related Cu II derivative 1 has the d 9 configuration of a metal ion and is paramagnetic. X-Band EPR spectra recorded for solid sample of 1 (Figure 4a) demonstrate anisotropic signals with an axial symmetry (g || = 2.18, g ⊥ = 2.08, A || ( 63,65 Cu) = 220 G), which is characteristic for square-planar Cu II compounds [110]. The finely resolved EPR spectrum for complex 1 in CH 2 Cl 2 solution at room temperature also exhibits a four-line pattern (Figure 4b the presence of intermolecular C-H···π interaction in the crystal of 3.
It should be noted that an incorporation of sterically hindered functions into the ligands can be assumed as a rational strategy to increase the solubility of the complex. A tendency to π-π intermolecular stacking is weakening. For instance, a problem of solubility has been raised for Cu II CT chromophores [44]. As a result, all studied coordination compounds 1-3 possess good solubility in a wide range of solvents with various polarity.

EPR Investigation for Complex 1
It is known that square-planar Ni II complexes are characterized by a low-spin state of the metal ion. As a result, compound 2 is diamagnetic, as evidenced by a well-resolved NMR spectrum. A related Cu II derivative 1 has the d 9 configuration of a metal ion and is paramagnetic. X-Band EPR spectra recorded for solid sample of 1 (Figure 4a) demonstrate anisotropic signals with an axial symmetry (g ║ = 2.18, g ┴ = 2.08, A ║ ( 63,65 Cu) = 220 G), which is characteristic for square-planar Cu II compounds [110]. The finely resolved EPR spectrum for complex 1 in CH2Cl2 solution at room temperature also exhibits a four-line pattern (Figure 4b  Dimeric compound 3 has two Co II paramagnetic ions. However, the cobalt derivative does not have any X-band EPR signal in solid or in solution at room temperature and being frozen down to 100K. This illustrates the preservation of the dimeric structure in the solution of 3. The formation of a low-spin Co II (SCo = 1/2) complex would be observed with dissociation. Such Co II derivatives show the corresponding signal in the EPR spectrum of frozen solutions [43].

Magnetochemical Study for Complex 3
An almost ideal square-pyramidal coordination environment of metal centers in 3 determined their high-spin state (SCo = 3/2), being a usual situation for transition elements in d 7 electronic configuration [30]. According to the temperature-variable magnetic susceptibility measurements, the high-temperature (T = 300 K) effective magnetic moment (μeff) value for complex 3 is equal to 6.39 μВ, which transcends insignificantly the theoretical spin only value of 6.30 μВ for two high-spin Co II ions (SCo = 3/2, quadruplet spin state) with g = 2.30. The magnitude of μeff changes slightly in the temperature range 300-100 K, whereas below 100 K the μeff(T) curve drops sharply down to 0.83 μВ at 2 K, Dimeric compound 3 has two Co II paramagnetic ions. However, the cobalt derivative does not have any X-band EPR signal in solid or in solution at room temperature and being frozen down to 100 K. This illustrates the preservation of the dimeric structure in the solution of 3. The formation of a low-spin Co II (S Co = 1/2) complex would be observed with dissociation. Such Co II derivatives show the corresponding signal in the EPR spectrum of frozen solutions [43].

Magnetochemical Study for Complex 3
An almost ideal square-pyramidal coordination environment of metal centers in 3 determined their high-spin state (S Co = 3/2), being a usual situation for transition elements in d 7 electronic configuration [30]. According to the temperature-variable magnetic susceptibility measurements, the high-temperature (T = 300 K) effective magnetic moment (µ eff ) value for complex 3 is equal to 6.39 µ B , which transcends insignificantly the theoretical spin only value of 6.30 µ B for two high-spin Co II ions (S Co = 3/2, quadruplet spin state) with g = 2.30. The magnitude of µ eff changes slightly in the temperature range 300-100 K, whereas below 100 K the µ eff (T) curve drops sharply down to 0.83 µ B at 2 K, that clearly indicates a presence of antiferromagnetic exchange interactions between spins of two Co II ions ( Figure 5). Analysis of the µ eff (T) dependence using dimer model (Spin Hamiltonian H = -2J·S 1 S 2 ) allows estimating exchange interaction parameter J. The best fit values of g-factor and J are 2.39 ± 0.01 and −4.76 ± 0.03 cm −1 , respectively. that clearly indicates a presence of antiferromagnetic exchange interactions between spins of two Co II ions ( Figure 5). Analysis of the μeff(T) dependence using dimer model (Spin Hamiltonian H = -2J·S1S2) allows estimating exchange interaction parameter J. The best fit values of g-factor and J are 2.39 ± 0.01 and −4.76 ± 0.03 cm −1 , respectively.

Electrochemical Study for Complexes 1-3
The electrochemical behavior of Cu II and Ni II monomeric derivatives 1 and 2 can be considered as similar, consisting in ligand-centered redox-stages ( Figure 6): reversible one-electron oxidative wave corresponds to a transition "catecholate dianion → semiquinone radical anion", and quasi-reversible one-electron reduction of coordinated bipyridine ligand is observed.

Electrochemical Study for Complexes 1-3
The electrochemical behavior of Cu II and Ni II monomeric derivatives 1 and 2 can be considered as similar, consisting in ligand-centered redox-stages ( Figure 6): reversible oneelectron oxidative wave corresponds to a transition "catecholate dianion → semiquinone radical anion", and quasi-reversible one-electron reduction of coordinated bipyridine ligand is observed. that clearly indicates a presence of antiferromagnetic exchange interactions between spins of two Co II ions ( Figure 5). Analysis of the μeff(T) dependence using dimer model (Spin Hamiltonian H = -2J·S1S2) allows estimating exchange interaction parameter J. The best fit values of g-factor and J are 2.39 ± 0.01 and −4.76 ± 0.03 cm −1 , respectively.

Electrochemical Study for Complexes 1-3
The electrochemical behavior of Cu II and Ni II monomeric derivatives 1 and 2 can be considered as similar, consisting in ligand-centered redox-stages ( Figure 6): reversible one-electron oxidative wave corresponds to a transition "catecholate dianion → semiquinone radical anion", and quasi-reversible one-electron reduction of coordinated bipyridine ligand is observed.  It should be noted that the values of electrochemical potentials for complex 2 and Ni II analog Ni II (3,6-Cat)(bipy) with unsubstituted 2,2 -bipyridine [32] are very close. However, the HOMO → LUMO gap is wider expectedly for 2 due to the presence of electron-donor tert-butyl fragments in acceptor diimine part (Table 4). On the contrary, dimeric Co II complex 3 displays another character of voltammogram ( Figure 6). Quasi-reversible one-electron redox process, which is detected with significant displacement towards reductive steps for 1 and 2, can be attributed to Co II → Co I reduction. Noticeably, the same metal-centered redox transformations were described in detail for Co II derivatives based on the polypyridyl multidentate ligand platforms [111,112]. Two consecutive one-electron reversible peaks are observed on the oxidative curve of the voltammogram. These processes can be attributed with equal probability to both the oxidation of catecholate ligands to o-semiquinone ones, and to the transition of Co II → Co III [112].

UV-vis-NIR Spectroscopy for Complexes 1-3
As it was highlighted above, a comparison of chromophores from a standpoint of CT energy is a quite rare final goal for corresponding investigations. H.-C. Chang and co-workers have shown [67] that changes in absorptivity among analogous CT dyes on the base of 10 group metals (Ni, Pd, Pt) cannot be considered as "substantial": the wavelength shift of LL'CT band did not transcend at~20 nm (in the same solvent) between derivatives of metals from adjacent periods. On the other side, to the best of our knowledge, there are no targeted investigations establishing how the nature (d-level structure, hence a coordination sphere volume and preferable coordination environment) of metals from other groups affects CT energy. However, this task may represent certain perspectives within CT chromophores' development and design. Thus, in this paper, we will try to work on this issue.
Regarding the Ni II complex 2, the energetic parameters of LL'CT (according to CV data, see above) are identical to those for the related Ni II chromophore, containing 3,5di-tert-butyl-catecholate ligand instead of 3,6-substituted counterpart (Table 5). An incorporation of tert-butyl functions to 2,2 -bipyridine platforms in 2 expectedly initiates a slight hypsochromic shift (within 20-35 nm in dependence on solvent) of the LL'CT bands (Table 5) in comparison with recently published (3,6-di-tert-butyl-catecholato)(2,2bipyridine)nickel(II) [32]). As can be seen from the current results (see Table 5), an equivalent effect towards LL'CT (according to CV data, see above) energy can be caused by a replacement of the metal center from Ni II to Cu II ion with the unchanged ligands and a similar molecular structure (monomer, square-planar coordination core): the LL'CT bands' shifts to shortwave region are equal to 26-40 nm in the case of Cu II analog 1 (Table 5). Another important fact that can be seen when comparing the photophysical characteristics of isostructural complexes 1 and 2 is that replacing the Ni II cation with a Cu II almost halves the extinction coefficient.

UV-vis-NIR Spectroscopy for Complexes 1-3
As it was highlighted above, a comparison of chromophores from a standpoint of CT energy is a quite rare final goal for corresponding investigations. H.-C. Chang and co-workers have shown [67] that changes in absorptivity among analogous CT dyes on the base of 10 group metals (Ni, Pd, Pt) cannot be considered as "substantial": the wavelength shift of LL'CT band did not transcend at ~20 nm (in the same solvent) between derivatives of metals from adjacent periods. On the other side, to the best of our knowledge, there are no targeted investigations establishing how the nature (d-level structure, hence a coordination sphere volume and preferable coordination environment) of metals from other groups affects CT energy. However, this task may represent certain perspectives within CT chromophores' development and design. Thus, in this paper, we will try to work on this issue.
The key feature of absorption spectra for complexes 1-3 is the presence of high-intense broadened CT bands extending in the visible/NIR regions (Figure 7), and a pronounced negative solvatochromic effect (Table 5). It is necessary to note that compounds 1, 2, and abovementioned metal complexes reported earlier [32,67] can be classified as so-called "visible CT dyes (i.e., have the maximums of CT band in visible region)" based on λ max values (without considering the band width). The generation of effective NIR chromophores is extremely relevant and dictated by different challenges, thereby one of the prior strategies within molecular design of CT dyes should be considered a targeted decrease of CT energy until to NIR absorptivity. One way is a modification of the donor catecholate moiety against acceptor diimine part. This is a quite efficient approach: (1) a slight turn in electronic properties of 3,6-di-tertbutyl-catecholate ligand in the course of an annelation with electron-donor glycol fragment shifted the LL'CT band at 95 nm, producing NIR dye (λ max = 810 nm, in toluene solution) of Ni II (Table 5) [32]; (2) a cardinal design (A. F. Heyduk with co-workers) [42]) of catecholates' composition and structure (3,5-di-tert-butyl-o-benzoquinone, tetrachloro-1,2-quinone and 9,10-phenanthrenequinone were used) among a wide set of "α-diimine-Ni II -catecholate" species resulted a record-breaking red LL'CT band's displacement at about~430-550 nm up to deep NIR region (the largest value λ max = 1370 nm, in THF solution).
Here, we report another possible strategy of molecular design on the way to NIR CT chromophores: a turn from 10 group metal in Ni II complex 2 (convenient standard in LL'CT chemistry due to a preference of a square-planar coordination polyhedron) to the 3d-row transition element of different nature with more labile coordination sphere. Dimeric Co II derivative 3 demonstrate UV-vis-NIR spectra with high-intense (within studied complexes 1-3) (M + L)L'CT (according to CV data, see above) absorption bands which are shifted from the visible to the NIR region (Table 5). It should be remarked especially that the dimeric structure of chromophore 3 is maintained regardless of the polarity and coordination ability of used solvents (CH 3 CN, DMF, CH 2 Cl 2 , THF) that is proved by a synergy of UV-vis-NIR and EPR-spectroscopy data: (1) the general character of UV-vis-NIR spectra was independent towards concentration effect, as well as solvent nature-high-intense (M + L)L'CT bands are presented ( Figure 7); (2) no X-band EPR signal for complex 3 was observed in CH 2 Cl 2 solution. Obviously, dimer destruction should generate "αdiimine-Co II -catecholate" species, and an implementation of a square-planar or tetrahedral molecular geometry is equiprobable for similar four-coordinated cobalt derivatives, as highlighted in Section 1. The tetrahedral mutual arrangement between HOMO and LUMO eliminates CT and, therefore, an observation of corresponding CT band in UV-vis-NIR spectra is impossible. In turn, square-planar coordination environment provides a low spin state of divalent cobalt ion (S Co = 1/2) [30,31], conditioning the paramagnetism of such "α-diimine-Co II -catecholate" monomer. Thus, the existence of a dimeric structure for 3 in solution is beyond doubt.
The pronounced negative solvatochromism is established for chromophores 1-3 as expressed by a fine linear dependence E T N (λ) (Figure 8) between CT bands and the normalized empirical parameter E T N by Dimroth and Reichardt [113]. The E T N parameter is obtained empirically on the base of the solvent-dependent CT band's shift for the standard N-phenoxypyridinium betaine dye, and, hence, describes the non-specific solvent polarity. CT bands' displacement for chromophores 1-3 is explained by fine-sensitive energy of non-polar excited states towards the solvent polarity. So, a turn from CH 3 CN to THF or toluene facilitates CT, causing substantial red shifts (Table 5) due to CT energy's decrease. The found parameters of solvatochromic displacement are comparable with those for known "α-diimine-Ni II -catecholate" analogs (Table 5), being one of the largest reported values [32,42,67]. meric Co II derivative 3 demonstrate UV-vis-NIR spectra with high-intense (within studied complexes 1-3) (M + L)L'CT (according to CV data, see above) absorption bands which are shifted from the visible to the NIR region (Table 5). It should be remarked especially that the dimeric structure of chromophore 3 is maintained regardless of the polarity and coordination ability of used solvents (CH3CN, DMF, CH2Cl2, THF) that is proved by a synergy of UV-vis-NIR and EPR-spectroscopy data: (1) the general character of UV-vis-NIR spectra was independent towards concentration effect, as well as solvent nature-high-intense (M + L)L'CT bands are presented ( Figure 7); (2) no X-band EPR signal for complex 3 was observed in CH2Cl2 solution. Obviously, dimer destruction should generate "α-diimine-Co II -catecholate" species, and an implementation of a square-planar or tetrahedral molecular geometry is equiprobable for similar four-coordinated cobalt derivatives, as highlighted in Section 1. The tetrahedral mutual arrangement between HOMO and LUMO eliminates CT and, therefore, an observation of corresponding CT band in UV-vis-NIR spectra is impossible. In turn, square-planar coordination environment provides a low spin state of divalent cobalt ion (SCo = 1/2) [30,31], conditioning the paramagnetism of such "α-diimine-Co II -catecholate" monomer. Thus, the existence of a dimeric structure for 3 in solution is beyond doubt.
The pronounced negative solvatochromism is established for chromophores 1-3 as expressed by a fine linear dependence ET N (λ) (Figure 8) between CT bands and the normalized empirical parameter ET N by Dimroth and Reichardt [113]. The ET N parameter is obtained empirically on the base of the solvent-dependent CT band's shift for the standard N-phenoxypyridinium betaine dye, and, hence, describes the non-specific solvent polarity. CT bands' displacement for chromophores 1-3 is explained by fine-sensitive energy of non-polar excited states towards the solvent polarity. So, a turn from CH3CN to THF or toluene facilitates CT, causing substantial red shifts (Table 5) due to CT energy's decrease. The found parameters of solvatochromic displacement are comparable with those for known "α-diimine-Ni II -catecholate" analogs (Table 5), being one of the largest reported values [32,42,67].

DFT Calculations
Quantum chemical calculations by the Density functional theory (DFT) method at the B3LYP/6-311++g(2d,2p) level of theory were performed to study the electronic structure of complexes under investigation. The experimental (X-ray) geometries were used as the starting points and optimized. Compound 1 was calculated in an open-shell approximation with S = 1/2 ground state. EPR spectroscopy data confirms doublet spin state for Cu II derivative 1. Complexes 2 and 3 were calculated in a closed-shell approximation with S = 0 ground state. The diamagnetic nature of Ni II complex 2 was evidenced by wellresolved NMR spectrum with no line broadening. The singlet ground state for dimeric Co II derivative 3 is clear from the solid state magnetochemical measurements. The selected frontier orbitals are presented in Figures 9 and 10. proximation with S = 1/2 ground state. EPR spectroscopy data confirms doublet spin state for Cu II derivative 1. Complexes 2 and 3 were calculated in a closed-shell approximation with S = 0 ground state. The diamagnetic nature of Ni II complex 2 was evidenced by well-resolved NMR spectrum with no line broadening. The singlet ground state for dimeric Co II derivative 3 is clear from the solid state magnetochemical measurements. The selected frontier orbitals are presented in Figures 9 and 10. The LUMO orbital (−2.51 eV) (Figure 9, (4)) is located mainly on the acceptor diimine part of the Ni II complex 2. This well-defined HOMO (−4.38 eV) ( Figure 9, (2)) occupies catecholate ligand with small nickel dxy and bipyridine system contribution. Orbitals energy and the nature of their distribution is quite obvious for such type of complexes [32]. The energy of the HOMO orbital increases significantly (0.41 eV) when tert-butyl substituents are introduced into the 2,2′-bipyridine [32]. These changes lead to an increase in the HOMO-LUMO energy gap (1.87 eV) and a corresponding hypsochromic shift observed in the electronic spectrum of compound 2 compared to the previously published derivative containing unsubstituted bipyridine. The energy (1.82 eV) of longwave absorption band (680 nm in toluene) is in a good agreement with calculated HOMO-LUMO energy gap. The nearest orbitals with a prevailing contribution of metal AO are HOMO-2 The LUMO orbital (−2.51 eV) (Figure 9, (4)) is located mainly on the acceptor diimine part of the Ni II complex 2. This well-defined HOMO (−4.38 eV) ( Figure 9, (2)) occupies catecholate ligand with small nickel d xy and bipyridine system contribution. Orbitals energy and the nature of their distribution is quite obvious for such type of complexes [32]. The energy of the HOMO orbital increases significantly (0.41 eV) when tert-butyl substituents are introduced into the 2,2 -bipyridine [32]. These changes lead to an increase in the HOMO-LUMO energy gap (1.87 eV) and a corresponding hypsochromic shift observed in the electronic spectrum of compound 2 compared to the previously published derivative containing unsubstituted bipyridine. The energy (1.82 eV) of longwave absorption band (680 nm in toluene) is in a good agreement with calculated HOMO-LUMO energy gap. The nearest orbitals with a prevailing contribution of metal AO are HOMO-2 (d z2 , −6.19 eV) and LUMO+3 (d x2-y2 , −1.14 eV). Thus, complex 2 can be considered as a LL'CT donor-acceptor chromophore which is controlled by the ligand framework.
The situation with the formation of frontier orbitals for the related Cu II complex 1 turns out to be very similar. HOMO (−4.28 eV) and α-LUMO (−2.58 eV) are located on the donor catecholate and acceptor diimine ligands respectively (Figure 9, (1) and (3)). The appearance of an additional electron at the d-level of the metal leads to a general decrease in the energy of metal atomic orbitals. Thus, the highest for squire-planar configuration semioccupied d x2-y2 orbital is α-HOMO-2 (−5.93 eV) for complex 1. This orbital is responsible for the doublet spin state of the compound (Figure 9, (5)). A decrease in the energy of the dorbitals in the complex 1 (compared to the Ni II derivative 2) leads to a decrease in the contribution of the d xy orbitals in the HOMO and LUMO. It reduces the efficiency of charge transfer between the donor and acceptor fragments of the molecule and the molar extinction coefficient for the Cu II compound 1 is about two times lower than the Ni II analog 2.

Materials and Methods
The comprehensive general information towards used materials and methods is collected in Supporting Information.

Syntheses and Characterization of Complexes 1-3
To prepare complexes 1-3, combined conditions were implemented: the second stages of the two-step synthetic procedure in the case of compounds 2 and 3, as well as the synthesis of 1, were performed under Schlenk line, while an argon atmosphere glovebox was used to carry out the second stages for 2 and 3.

Complex 1•THF, [Cu II (3,6-Cat)(bipy tBu )]•THF.
Equimolar amounts of 3,6-di-tert-butyl-o-benzoquinone (3,6-DTBQ) (0.5 g, 2.27 mmol) and 4,4′-di-tert-butyl-2,2′-bipyridyl (0.61 g, 2.27 mmol) reacted with an excess of sliced copper foil in dry THF (20 mL) in evacuated ampoule. The resulting solution, which turned from bright-green to ink-blue color, was decanted from copper and partially evaporated, after that dark fine-crystalline product formed. Precipitate was collected by filtration and washed with THF under reduced pressure. Single crystals of [Cu II (3,6-Cat)(bipy tBu )]•THF with C36H52CuN2O3 composition, which are suitable for X-ray diffraction experiment, were obtained from THF through a slow evaporation of solution under residual pressure. The composition and energy of the frontier orbitals undergo a significant change in the Co II complex 3 ( Figure 10). HOMO (−3.82 eV) and LUMO (−2.25 eV) have a prevailing metal character which is confirmed by the data of electrochemical studies. A noticeable contribution of the diimine fragment is observed for LUMO. The nearby HOMO-1 (−4.19 eV) and HOMO-2 (−4.44 eV), as well as LUMO+1 (−2.00 eV) and LUMO+2 (1.97 eV), are occupied by catecholate and bipyridine ligands respectively. These orbitals have an appreciable contribution of cobalt d-orbitals. Thus, the cobalt ion in the resulting donor-acceptor complex provides effective charge transfer due to the active involvement of metallic d-orbitals into the frontier orbitals formed by redox-active organic fragments. It should be noted that the decrease in the molar extinction coefficient in low-polar solvents for Co II derivative 3 is not observed, despite the formation of a five-coordinate environment of the metal atom. This configuration of the complex can provide charge transfer between the donor and the acceptor [114]. However, it is significantly inferior to structures with a square-planar geometry.

Materials and Methods
The comprehensive general information towards used materials and methods is collected in Supporting Information.

Syntheses and Characterization of Complexes 1-3
To prepare complexes 1-3, combined conditions were implemented: the second stages of the two-step synthetic procedure in the case of compounds 2 and 3, as well as the synthesis of 1, were performed under Schlenk line, while an argon atmosphere glovebox was used to carry out the second stages for 2 and 3.
The synthesis in situ of CatNa 2 was carried out through the direct interaction of 3,6-DTBQ with an excess of metallic Na in THF. Reagents were stirred for an hour in evacuated ampoule, while the bright-green initial color of the reaction mixture turned to pale. The corresponding technique is well-known and described earlier [115].
The polycrystalline samples of 3 for an elemental analysis, IR-, EPR-, UV-vis-NIR spectroscopy, cyclic voltammetry, and magnetic susceptibility measurements were prepared from CH 2 Cl 2 .

Conclusions
A novel strategy within a molecular design of CT chromophores aimed at the CT energy management was implemented by the example of new "α-diimine-M II -catecholate" CT dyes 1-3 (where M is 3d-row transition metals-Cu, Ni, Co): the varying of the metal center's nature (electronic structure of d-level) was performed. As regards the preparation of CT dyes, three synthetic routes were applied and optimized, providing a high yield of complexes with different composition and molecular structure. So, the chosen approaches allow obtaining the square-planar monomeric LL'CT species (Cu II and Ni II derivatives 1 and 2, respectively), as well as dimeric Co II (M + L)L'CT chromophore 3 containing two "α-diimine-M II -catecholate" moieties united through the donor ability of oxygen atoms from catecholate dianions.
Corresponding changes in CT energy were investigated in detail by the synergy of experimental and theoretic tools, namely UV-vis-NIR spectroscopy, cyclic voltammetry, and DFT study. In particular, a considerable red CT band's displacement at approximately 130-150 nm was achieved by the turn from Cu II to Co II center with the same ligand platforms (neutral 4,4 -di-tert-butyl-2,2 -bipyridyl and 3,6-di-tert-butyl-catecholate dianion). Moreover, a fine-tunable interrelation between CT energy and solvent polarity was established, expressed in a pronounced negative solvatochromic effect with substantial blue CT band's shifts. Total values are equal to 153, 137, and 58 nm for complexes Cu II (1), Ni II (2), and Co II (3), respectively (from toluene to CH 3 CN for 1, 2; from THF to CH 3 CN for 3).
Supplementary Materials: The following are available online at https://www.mdpi.com/article/ 10.3390/molecules27238175/s1, Figure S1: The fragment of crystal packing for complex 1-the molecules from odd and even layers are represented as bright and faded pictures, respectively. Structures are given with 30% thermal probability ellipsoids. Hydrogen atoms, tert-butyl groups and THF molecules are omitted for clarity; Figure S2: The shortest intra-stack contacts Cd Cat(A) · · · Cd bipy(B) , Cd Cat(B) · · · Cd bipy(A) , and Ni(1A)· · · Ni(1B) for complex 2. Structures are given with 30% thermal probability ellipsoids. Hydrogen atoms and tert-butyl groups are omitted for clarity; Figure S3. The shortest contact Cd Cat · · · C(23) = 3.695 Å for complex 3. Structures are given with 30% thermal probability ellipsoids. Hydrogen atoms and tert-butyl groups are omitted for clarity; Table S1: X-ray diffraction data collection and structure refinement for complexes 1-3. CIF and CheckCIF files for 1 and 2. Crystallographic data for the structural analysis has been deposited with the Cambridge Crystallographic Data Centre, CCDC 2213958 (1·THF), 2213959 (2) and 2213960 (3·toluene). Copies of the above information may be received free of charge at ccdc.cam.ac.uk/getstructures from the Cambridge Crystallographic Data Centre. References [113,[116][117][118][119][120][121] are cited in the supplementary materials.