Trinuclear Ni(II) Complex with Rare Deprotonation of the Bridging Oxamate and Imino Nitroxide Radical as Blocking Ligand

1. Introduction

Polytopic organic ligands [1] of different topology are used for creating metallo-supramolecular assemblies with variable structural and physical properties. Bis-oxamate phenylene-based polydentate ligands offer a unique potential for a greater variety of coordination compounds, both homo- and heterometallic, in which spin-bearing metal ions are strongly magnetically coupled [2,3,4,5,6]. Despite a thirty-year history of studying such systems, no polynuclear complex in which N,N′-1,4-phenylenebis-oxamic acid (L) was triple-deprotonated (L³⁻) (Figures 1a and 1b) has yet been identified. In previous studies, the synthesis of compounds with half-deprotonated (L²⁻) [6,7,8,9,10] and fully deprotonated ligand (L⁴⁻) [11,12,13,14,15,16,17,18,19,20] (Figures 1c and 1e respectively) has been achieved. In addition, it has been shown that nickel(II) complexes in a distorted coordination octahedral environment can possess both optical and magnetic properties necessary for their use as quantum qubits [21,22].

The utilization of imino nitroxyl (IN) radicals as supplementary ligands confers several advantages. Firstly, it is a blocking function to prevent the formation of coordination polymers. Secondly, it has been established that, in contrast to nitronyl nitroxyl (NN) radicals, which exhibit strong antiferromagnetic coupling in Ni-NN pair [23,24,25,26], the Ni-IN interaction is ferromagnetic and strong [23,27,28,29,30,31,32,33,34,35,36]. The latter circumstance is of significant importance from the standpoint of enhancing the full spin value of the molecular cluster.

In this paper, we present the molecular, crystal structure, experimental and theoretical magnetic behavior of an exceptional neutral trinuclear complex [Ni(L³⁻)₂(IN)₃]∙5MeOH (1) having a cyclic triangular arrangement, where IN is an imino nitroxyl radical, 2-(6-methyl-2-pyridyl)-4,4,5,5,5-tetramethylimidazoline-1-oxyl (Figure 1f). Moreover, in this compound three Ni²⁺ ions are linked by the two bis-oxamate ligands playing rare tritopic function due to an unprecedented triple deprotonation of the related meta-phenylene-bis(oxamic acid).

2. Results and Discussion

2.1. Synthesis and Characterization

Two aqueous solutions, 1 and 2, were prepared simultaneously under agitation on magnetic stirrers at 80 °C. The utilized starting materials and their respective quantities are delineated in section 3.3.1. Solution 1 contained bis-oxamate ligand coordinated to the metal, while solution 2 contained the chelate complex [Ni(IN)(H₂O)₄]²⁺. Solution 2 was added dropwise to solution 1 without removing the heating. Following the heating and cooling stirring procedures, which were each 15 minutes in duration, the resultant red-orange precipitate was filtered through a porous glass filter. The precipitate was then washed with a small amount of water and dried on the filter until it became friable, while sucking in air. The solid was then dissolved in 5 mL of hot methanol. After filtering, the product solution was left undisturbed for several days. The dark red crystals that formed were used for structure determination. It is well established that crystals of the complex quickly lose solvate molecules when stored in air. For further characterization, we used either freshly prepared crystals (TG and IR) or powder brought to constant weight by heating at 80°C (elemental analysis).

In Supplementary Materials (SM) the results of the FTIR spectral and thermogravimetric investigations of [Ni₃(L³⁻)₂(IN)₃]∙5CH₃OH (1) are presented. The thermic study of 1 is illustrated in Figure S1, SM. After loss of solvated methanol molecules in the range of 20-100°C (theoretical mass loss is 10.46 %), the solid phase of three-nuclear cluster is thermally stable up to 210°C. After reaching a steady weight, dry complex [Ni(L³⁻)₂(IN)₃] was further heated to 360°C. At that final temperature, the mass loss was 55.15% of the initial mass. This result supports the assumption that the final product in this experiment was probably nickel(II) carbonate.

IR spectrum of 1 confirms (Figure S2, SM) the absence of perchlorate ions from starting nickel salt, indicating the neutrality of the cluster. The vibrations due to the participation of solvated alcohol molecules in the hydrogen bonding system create an intense and broad band in the 3750-2700 cm⁻¹ region (ν_max=3400 cm⁻¹), covering also the absorption bands from >N−H (shoulders at 3268 and 3102 cm⁻¹) and C−H (2965, 2925 and 2857 cm⁻¹) from the methyl groups of IN. An intense asymmetric absorption band in the 1800–1500 cm⁻¹ region with a maximum at 1590 cm⁻¹ covers few vibrations: shoulder at 1717 cm⁻¹ of (C=O), band at ~1656 cm⁻¹ of (O−C=O)_asym and the C=C and C=N stretching vibrations. The two peaks at 1448 cm^-1 and 1335 cm⁻¹ apparently correspond to ν(C−N_Py) and C=O vibrations respectively. The other absorptions can be assigned to: a symmetric mode of O−C=O at 1472 cm⁻¹, a shoulder at 1386 cm⁻¹ to ν(N−O), at 1072 and 700 cm⁻¹ to δC−H in-plane and δC−H out-of-plane respectively. The other less intense absorption bands in 1200−400 cm⁻¹ area predominantly belong to IN ligand.

2.2. Crystal and Molecular Structure

The crystal structure of 1 contains trinuclear molecular clusters of [Ni₃(L³⁻)₂(IN)₃] and solvate methanol molecules. It has been established that each Ni²⁺ cation coordinates an imino nitroxide (IN) paramagnetic ligand in a chelate manner. {Ni(IN)}²⁺ fragments are interlinked with two m-phenylene-bis(oxamate) anions, L³⁻, to form trinuclear complex, Figure 2.

All the metallic centers are situated within a distorted octahedral coordination environment. Ni1 one is linked with two nitrogens atoms of the radical ligand (N11, N12) and four oxygen atoms of two L³⁻ anions (O41, O43, and O53, O53). Both Ni2 and Ni3 cations possess an identical coordination environment, comprising three N atoms of an IN ligand (N21, N22 for Ni2, and N31, N32 for Ni3) and an L³⁻ anion (N51 for Ni2 and N41 for Ni3). In addition, the coordination environment incorporates three O atoms of two L³⁻ anions (O44, O46, O51 for Ni2, and O54, O56, O42 for Ni3), which are linked in a fac manner. Ni–N and Ni–O bond lengths and valence angles are listed in the Table S2 (Supplementary Material, SM). As demonstrated in Table S3, the coordination environment of each Ni ion is a distorted octahedron. The matching ratios are very close for Ni1 and Ni2, and the ratio for Ni3 slightly exceeds them. The arrangement of the triangle molecules along the c axis gives rise to the formation of a supramolecular chain. In the latter, each complex is connected with adjacent molecules by means of hydrogen bounding between −N–H (N42 and N52) and O=C− (O45 and O55) groups of two L³⁻ ligands (Figure 3) with the N…O distances of 2.795(5) and 2.829(5) Å. Interactions between neighboring chains are established via CH··∙O contacts between the pyridyl −CH group (C103) of the radical and the coordinated oxygen atom (O46) of the L³⁻ anion. These results in the formation of a wave-shaped supramolecular layer that is oriented parallel to the bc plane (Figure S3, SM) with the N…O distance of 3.182(8) Å. The layers exhibit an alternating pattern along crystallographic axis a. The structure contains 28% solvate accessible void volume, which was estimated with PLATON [37]. This volume is occupied by guest methanol molecules. The latter form a system of hydrogen bonds, which involves CH₃OH···HOCH₃ and CH₃OH···O(L³⁻) contacts, the O…O distances being in range 2.48(3)–2.851(6) Å.

2.3. Magnetic Studies

2.3.1. Cryomagnetic Measurements

The temperature dependence of susceptibility for a polycrystalline sample of 1 is presented in Figure 4 as a χT vs T graphic. The room temperature χT value (4.93 emu K mol⁻¹) is closed to the value corresponding to the sum uncorrelated spins of constituent paramagnetic centers. As the temperature decreases to 2 K, the χT value steadily decreases, indicating the predominance of antiferromagnetic interactions in the solid phase of the compound. This is also shown by the field dependence data of the magnetization (Figure 5), as the value of M(H) at H = 7 T is only 3.26 µ_B.

In the context of AC magnetic measurements conducted at temperatures below 25 K, no significant correlation was identified between the magnetic susceptibility and the field oscillation frequency, irrespective of the presence or absence of an external magnetic field with a strength of B = 0.1 T.

2.3. Magnetic Behavior Modeling and Theoretical Calculations

2.3.1. Theoretical Model and Exhaustive Parameter Set Required for Magnetic Data Simulation

Theoretical modeling of the magnetic behavior of the investigated complex, considering two different types of magnetic coupling (Scheme 1), was performed using the expression for the spin-Hamiltonian (SH) outlined below

$$\begin{array}{cc}\hat{H}=& -{2J}_1{\mathrm{Ŝ}}_{M1}·{\mathrm{Ŝ}}_{R1}-2J_2{\mathrm{Ŝ}}_{M2}·{\mathrm{Ŝ}}_{R2}-2J_3{\mathrm{Ŝ}}_{M3}·{\mathrm{Ŝ}}_{R3}-\hfill \\ & -2J_{11}{\mathrm{Ŝ}}_{M1}·{\mathrm{Ŝ}}_{M2}-2J_{12}{\mathrm{Ŝ}}_{M1}·{\mathrm{Ŝ}}_{M2}-2J_{13}{\mathrm{Ŝ}}_{M2}·{\mathrm{Ŝ}}_{M3}+\hfill \\ & +{\mathrm{Ŝ}}_{M1}{\mathit{D}}_1{\mathrm{Ŝ}}_{M1}+{\mathrm{Ŝ}}_{M2}{\mathit{D}}_2{\mathrm{Ŝ}}_{M2}+{\mathrm{Ŝ}}_{M3}{\mathit{D}}_3{\mathrm{Ŝ}}_{M3}+\hfill \\ & +{g_{Ni}\beta (\mathrm{Ŝ}}_{M1}+{\mathrm{Ŝ}}_{M2}+{\mathrm{Ŝ}}_{M3})·B_0{{+g}_R\beta (\mathrm{Ŝ}}_{R1}+{\mathrm{Ŝ}}_{R2}+{\mathrm{Ŝ}}_{R3})·B_0\hfill \end{array}$$

Scheme 1. Topological structure of magnetic exchange interactions between spin carriers in complex 1: M_n = Ni₁, Ni₂ and Ni₃, R_n = IN radical coordinated to the corresponding Ni-center.

Scheme 1. Topological structure of magnetic exchange interactions between spin carriers in complex 1: M_n = Ni₁, Ni₂ and Ni₃, R_n = IN radical coordinated to the corresponding Ni-center.

The initial line of the expression delineates the exchange interaction between nickel ions (M_n) and the IN-radicals (R), while the subsequent line accounts for the M-M exchange interaction. The third and fourth lines represent the zero-field splitting (ZFS) for Ni²⁺ -centers having S = 1 and the Zeeman interaction with the external magnetic field B₀, respectively. In order to circumvent the over-parameterization problem, the isotropic g-factor for nickel (g_M) was utilized, and the value of g_R was assumed to be 2.

The results of DFT calculation of exchange integrals by broken symmetry (BS) method using different functionals are summarized in Table 1. The latter shows that the nickel-radical magnetic couplings (J₁, J₂, J₃) are roughly the same within a method and ferromagnetic (FM), whereas the metal-metal interaction in M₁−M₂ and M₁−M₃ pairs are has the opposite sign. Hybrid functionals (B3LYP and TPSSh) give about one and a half times larger coupling values than the more recent range-separated DF variants. A separate calculation using the B3LYP method also showed that the exchange integral J₁₂ between spins Ni₂ and Ni₃ is small and does not exceed 0.05 cm⁻¹. Therefore, it was ignored and assumed equal to zero for SH.

To clarify and verify the results of DFT calculations, ab initio CASSCF/NEVPT2 computations with partial diamagnetic substitution of paramagnetic centers in 1 were also performed. For this purpose, the Ni²⁺-centers (M₂ and M₃) were replaced by Zn²⁺ ions in the trinuclear molecule. In addition, the radicals (1 and 2) were also replaced by their reduced analogs 4,4,5,5-tetramethyl-2-(6-methylpyridin-2-yl)-4,5-dihydro-1H-imidazol-1-olhaving a hydroxyl amine moiety, >N-OH, instead of the nitroxyl group >N-O. In this way, a cluster with only one exchange-bonded pair was obtained: Ni²⁺ (S = 1) and imino nitroxyl radical (S = 1/2), respectively (Figure 6). The ground state of such a spin system is a quartet, and the energy gap between the ground state and, nearest to it, the spin doublet can be calculated by the formula: J₁ = (E(D) − E(Q))/3. The results of the calculation of J₁ obtained in this way are summarized in Table 2.

A similar diamagnetic substitution approach (see Figure 7) was implemented for the CASSCF/NEVPT2 computation of the J₁₂ exchange integral between Ni²⁺ ions (M₁ and M₂). The constant J₁₂ in this case is found by the formula: J₁₂ = (E(T) − E(S))/2 (see Table 3).

As evidenced by a comparison the data from the Table 2 and Table 3 with those of Table 1, there is a substantial discrepancy between the exchange integral values obtained from DFT calculations and CASSCF/NEVPT2 data, with a range from 4 to 7 times.

To estimate the other SH parameters, the g-tensor and D-tensor of the paramagnetic center M₁ (Ni²⁺, S = 1), were calculated using the diamagnetic substitution shown in Figure 5. The results of the computations are summarized in Table 4. Considering the obtained anisotropy of the Ni g-tensor, which is less than 2%, only the isotropic part of the g-tensor is presented in Table 4, in which the parameters corresponding to the standard form of the D tensor along the principal axes are also given.

According to the DFT calculations, it was hypothesized that all exchange integral values of Ni−IN are equivalent (J = J₁ = J₂ = J₃), as well as J(Ni₁−Ni₂) = J(Ni₁−Ni₃) during the simulation of experimental data on χT(T). This approach was adopted to circumvent the issue of over-parameterization. Furthermore, the parameters of the g- and D-tensors for all Ni-centers were considered equal. The results of the experimental data fitting performed in the Phi program are displayed in Figure 3.

The best fitting parameters are as follows: g_Ni = 2.3; D_Ni = 2.0 cm⁻¹; J = J₁ = J₂ = J₃ = 36.2 cm⁻¹; J₁₂ = J₁₃ = −18.9 cm⁻¹. These parameters were also used to fit the experimental magnetization data, M(H), registered at a temperature of 2K (see Figure 4).

The comparison of the optimal values for the exchange integrals, g_Ni and D_Ni, obtained from the χT(T) simulation with the results of quantum-chemical calculations (Table 2, Table 3 and Table 4), shows that the best value for g_Ni = 2.29 is provided by cas(8,5) calculation. DFT computation, especially cam-B3LYP, most accurately reproduces D_Ni. The CAS(13,9)/NEVPT2 method gives a very close value for Ni-IN exchange integral, J, but ab initio approach can’t reproduce the J₁₂ (Ni₁-Ni₂) exchange integral. The CAM-B3LYP calculation is the most accurate in this case.

The estimation of the magnetic exchange interactions between neighboring hydrogen-bonded triangular molecules of the complex 1 (see Figure 5) obtained from the DFT computation yielded a negative J_inter, with its absolute value ranging from 0.1 to 0.2 cm⁻¹.

3. Materials and Methods

3.1. Instrumental and Physical Measurements

Elemental (C,H,N) analysis was performed on a Euro-Vector 3000 analyzer (Eurovector, Redavalle, Italy). FTIR spectra were registered with a NICOLET spectrophotometer (Thermo Electron Scientific Instruments LLC, Madison, WI, USA) in the 4000–400 cm⁻¹ range. Thermal stability of 1 was studied by means of Thermo Microbalance TG 209 F1 Iris (Netzsch, Selb, Germany). The cryomagnetic investigation of the compound was performed using a Quantum Design MPMS 5XL SQUID magnetometer (Quantum Design, Inc., San Diego, CA, USA) in the temperature range of 1.8–300 K and under a magnetic field of up to 7 T. The diamagnetic contribution for the magnetic susceptibility of the complex 1 was corrected using Pascal’s constants for the constituent atoms and organic ligand bonds [38].

Single-crystal XRD experimental details are presented in Table S2 (, SM). Crystallographic data were deposited with the Cambridge Crystallographic Data Centre (deposit number CCDC 2451012).

3.2. Theoretical Calculations

Quantum chemical study was fulfilled using crystallographic geometry. The calculations of magnetic exchange J integrals, D tensors, and g tensors was performed by both broken-symmetry DFT (BS-DFT) [39] and ab initio CAS/NEVPT2 methods with def2-QZVPP basis set for Ni and def2-TZVP for other atoms using the Orca-6.0 software package [40,41]. Fitting of the experimental χT(T) and M(H) dependences to obtain optimal parameters of employed spin–Hamiltonian was carried out using the PHI 3.16 package [42].

3.3. Preparations

Solvents of the reagent grade (EKOS-1, Moscow, Russia) were distilled prior to use. The complex was synthesized under ambient conditions. The radical 2-(6-methyl-2-pyridyl)-4,4,5,5,5-tetramethylimidazoline-1-oxyl (IN) was synthesized according to a literature procedure [23]. Tritopic derivative, Na₃L, was prepared in situ by hydrolysis of ethyl-2- [3- [(2-ethoxy-2-oxo-acetyl)amino]anilino]-2-oxo-acetate [12].

3.3.1. Synthesis of [Ni₃(L³⁻)₂(IN)₃]∙5CH₃OH

The solution 1: Under stirring and heating (80 C), a solution of IN (38 mg, 0.165 mmol) in 1.5 mL H₂O was added dropwise to a solution of Ni(ClO₄)₂·6H₂O (60 mg, 0.165 mmol) in 2.5 mL H₂O.

The solution 2. The hydrolysis of ethyl-2- [3- [(2-ethoxy-2-oxo-acetyl)amino]anilino]-2-oxoacetate [43,44], the esterified form of 2- [3-(oxaloamino)anilino]-2-oxoacetic acid was performed in 2 mL of distilled water? Which was added to the mixture of ethyl-2- [3- [(2-ethoxy-2-oxo-acetyl)amino]anilino]-2-oxoacetate (33 mg, 0.11 mmol) and NaOH (18 mg, 0.45 mmol). The mixture was then stirred at a magnetic stirrer at a temperature of 80°C for 15 minutes.

An elemental analysis of complex 1 was performed on a desolvated sample that was heated to 110°C until constant weight was reached: Anal Calcd. for Ni₃C₅₉H₆₄N₁₃O₁₅ (mass. %): C, 51.74; H, 4.71; N, 13.30; found: C, 51.68; H, 4.6; N, 13.25. IR spectrum (KBr, ν, cm⁻¹): 3400, 3268, 3102, 2965, 2925, 2857, 1717, 1656, 1590, 1472, 1448, 1386, 1335, 1072, 700.

4. Conclusions

Among the complexes of d-metals with phenylene-based bis-oxamate ligands, the compound studied by us is distinguished by its atypical tri-nuclear triangular molecular structure, in which three metal centers are connected by only two anionic bridges, representing a triply deprotonated form of 2- [3-(oxaloamino)anilino]-2-oxoacetic acid, complexes with such a form of the latter having not been previously known.

Despite the presence of six possible magnetic couplings in the trinuclear cluster 1, its magnetic behavior was well reproduced using the three-J model and magnetic anisotropy D under the assumption that three different Ni-IN interactions are equal to each other, and considering only two equivalent Ni-Ni interactions, with the third one equating to zero. This restriction was implemented to circumvent the issue of over parameterization. The magnetic plots simulation results indicate the presence of two opposite in nature types of magnetic interactions within the triangular core. DFT and detailed CASSCF/NEVPT2 computations were also performed, thereby providing support for the experimental modeling of the magnetic data.+