Building Up a Hexacopper(II)-Pyrazolate/Oxamate Magnetic Complex with Rare Ethane-1,2-Dioxide (–OCH₂CH₂O–) as a Bridge Between Copper(II) Units

Abstract

The synthesis, structural, and magnetic characterization of a novel neutral copper(II) hexanuclear complex [Cu₆(en)₄(OCH₂CH₂O)₂(pyox)₄]·3eg·en·12H₂O (1) was investigated [en = ethylenediamine, eg = ethylene glycol, and H₂pyox = 4-(1H-pyrazole-4-yl)phenylene-N-oxamic acid]. The crystal structure of 1, obtained by the single-crystal X-ray diffraction technique, revealed that the hexacopper(II) complex is built from two linear tricopper(II) complex subunits. Each subunit contains two [Cu(en)]²⁺ moieties connected to a [Cu(OCH₂CH₂O)] unit by two pyox²⁻ ligands acting as μ-κN:κN′ bridges, as well as a [OCH₂CH₂O]²⁻ ligand, which is ultimately found in the μ₃-κO,O′:κO:κO′ coordination form. The subunits are connected via the amide portion of the pyox²⁻ ligand, linked to copper atoms in the other subunit. They occupy the apical coordination positions, leading the trinuclear copper(II) segments to be almost perpendicular. The structural, chemical, and spectroscopic characterizations evidenced that ethylene glycol acted both as a solvent and a reactant upon deprotonation, forming the –OCH₂CH₂O– ligand due to the basic crystallization environment. DC magnetic studies revealed a strong antiferromagnetic interaction between the copper atoms within the trinuclear subunits, influenced by alkoxide and pyrazolate bridging ligands. Our findings offer new insights into the structural and magnetic properties of copper(II) complexes, enhancing the understanding of metal–ligand interactions in supramolecular chemistry.

1. Introduction

Polynuclear complexes have attracted significant attention in coordination chemistry, particularly their magnetic properties and potential applications in molecular materials with tunable magnetic behavior [1,2]. To achieve magnetic interactions between metal ions, orbital overlapping through the ligands is essential. In this context, bridging ligands with sp² or sp-hybridized atoms effectively facilitate spin density communication between metal centers [1,3,4]. Notably, oxamate and pyrazolate ligands have proven highly effective in bridging magnetic metal ions.

When oxamate acts as a bis-bidentate ligand (μ-κ²N,O:κ²O′,O″ coordination mode), it can achieve coupling constants exceeding −400 cm⁻¹ [5,6]. Pyrazolate, on the other hand, in its μ-κN:κN′ coordination form, exhibits coupling capabilities that are strongly influenced by the angle between the metal–nitrogen bonds, though it is not far behind oxamate in terms of performance [7]. Thus, incorporating both oxamate and pyrazolate moieties within the same ligand holds great promise for creating strongly coupled polymetallic systems.

On the other hand, synthesizing polynuclear complexes can be very challenging, and self-assembly processes can significantly facilitate this, especially when faced with poorly soluble products. Allowing the components to adopt the most stable conformation spontaneously has already led to the discovery of remarkable magnetic polynuclear compounds, such as the square-like [Cr^III₂(dmso)₈{μ-cis-Nb^VO₂(C₂O₄)₂}₂] [8]; the temperature influenced the formation of the 1D ladder-like coordination polymer [Mn₄(H₂mpba)₄(H₂O)₁₂]_n{[Mn₈Cu₈(mpba)₈(H₂O)₂₄]_n}·29.5nH₂O and 2D coordination polymer [Mn₄Cu₄(mpba)₄(H₂O)₉]_n·14nH₂O [9]; the discrete [Cu(bipy)(H₂mpba)]₂·2H₂O and 1D coordination polymer [Cu(bipy)(H₂mpba)]_n·ndmso systems, which can interchange based on solvent soaking [10]; the 2D coordination polymer with magnetization hysteresis [Co₂Cu₂(mpba)₂(H₂O)₆]_n⋅6nH₂O [11]; and so on (H₄mpba = N,N′-1,3-phenilenebis(oxamic acid), bipy = 2,2′-bipyridine). This synthetic strategy often reveals unusual phenomena, including ligand displacement [8], solvent coordination, ligand transformation [12], and, in the case of this work, the formation of the rare ligand ethane-1,2-dioxide from ethylene glycol (eg) solvent doubly deprotonated in the reaction–crystallization process.

The ethane-1,2-dioxide (OCH₂CH₂O) ligand is only found as a bridging ligand when coordinated to transition metal ions [13,14,15,16,17,18,19,20,21] and group 13–15 metal complexes from the periodic table [21,22,23,24,25,26,27,28,29,30,31]. It has been observed that using the diol itself leads to natural deprotonation, forming the polynuclear complex upon moderate heating (100–200 °C) [13,15,18,20,21,23,24,27,31] or via proton exchange in alkoxide complexes [14,16,24,28]. Manganese examples with ethane-1,2-dioxide are the only ones where the simple addition of a weak base is enough to deprotonate the ethylene glycol at room temperature [9,17]. While several polynuclear complexes containing ethane-1,2-dioxide have been reported, only one other crystal structure presenting it with copper(II) ions has been reported [20].

This contribution explores a novel hexanuclear copper(II) complex coordinated by a ligand combining the oxamate and pyrazolate groups—pyox²⁻ (H₂pyox = 4-(1H-pyrazole-4-yl)phenylene-N-oxamic acid), the ethane-1,2-dioxide ligand and ethylenediamine. To the best of our knowledge, this is a rare example from the literature that uses the ethane-1,2-dioxide ligand to connect copper atoms (Scheme 1). The synergistic interaction between these three distinct ligands allows for an in-depth investigation of the correlations between structure and magnetic behavior, with potential applications in developing new functional magnetic materials [32].

2. Experimental Method

All chemicals were used without further purification. The compound (n-Bu₄N)Hpyox·0.5H₂O was prepared as previously described in the literature [33,34]. [Cu(en)₂](NO₃)₂ (en = ethylenediamine) preparation followed the synthetic procedure described by Rafat et al. [35].

Synthesis of [Cu₆(en)₄(OCH₂CH₂O)₂(pyox)₄]·3eg·en·12H₂O (1). In the bottom of a small test tube (~5 mL capacity) was placed 0.5 mL of a solution containing 10.0 mg (32.5 μmol) of [Cu(en)₂](NO₃)₂ and 1.00 mg (4.14 μmol) of Cu(NO₃)₂·3H₂O dissolved in a mixture of ethylene glycol (eg) and methanol 1:1 v/v. Then, 2.0 mL of a mixture of methanol and ethylene glycol 4:1 v/v containing 1% in volume of Bu₄NOH (40% weight in water) was carefully placed over the bottom solution to form a separate layer. Finally, a third layer comprising a methanolic solution (0.250 mL) of (n-Bu₄N)Hpyox·0.5 H₂O (30.0 mg, 62.5 μmol) was placed on the top. The tube was sealed and the materials were left to diffuse at room temperature (20 °C) for two weeks. After this time, dark blue crystals were observed in the tube wall. The crystals were collected, washed with methanol, and left to dry in an open atmosphere for 48 h. This synthesis was replicated using several test tubes, diffusing simultaneously to obtain enough material for all studies. Yield: 71.0% (6.25 mg; 2.94 μmol). Elemental Analysis [Exp.(calcd.)] for C₆₄H₁₁₆N₂₀O₃₆Cu₆ (2123.01 g mol⁻¹): %C 37.28 (36.21), %H 4.63 (5.51), %N 14.17 (13.20), %Cu 17.89 (17.96). IR (cm⁻¹): 3250 and 3129 (ν N–H), 1632 (ν C=O), 1044 (ν C–O).

2.1. Chemical Characterizations

The thermogravimetric (TG) and differential thermal analysis (DTA) of 1 was carried out in a Shimadzu DTG-60H thermobalance using a ca. 5.0 mg sample in an alumina crucible, using synthetic air flow at 50.0 mL min⁻¹, from room temperature to 800 °C (Figure S1). The IR spectrum was obtained in a Perkin Elmer FTIR GX spectrometer in ATR mode, from 400 to 4000 cm⁻¹, at 4 cm⁻¹ spectral resolution (Figure S2). Elemental analysis was carried out in a CHN 2400 Perkin-Elmer analyzer (PerkinElmer, Waltham, MA, USA) for carbon, hydrogen, and nitrogen content analysis, and a VARIAN AA240FS spectrophotometer (VARIAN, Palo Alto, CA, USA) with flame atomization was used for copper content analysis.

2.2. X-Ray Diffraction

X-ray diffraction data collection on single crystals 1 was performed with a Rigaku Gemini diffractometer using Mo-Kα radiations (0.71073 Å). Measurements were performed at 120 K. Data integration and the reflections for all compounds were scaled with the CRYSALIS suite [36]. Final unit cell parameters were based on the fitting of all reflection positions. Analytical absorption corrections and the space group identification were performed using the CRYSALIS suite [36]. The structures of all compounds were solved by direct methods using the SUPERFLIP program [37]. For each compound, the positions of all atoms could be unambiguously assigned on consecutive difference Fourier maps. Refinements were performed using SHELXL-2019 [38] based on F² through the full-matrix least-squares routine. All non-hydrogen atoms and non-water molecules were refined with anisotropic atomic displacement parameters. The SQUEEZE technique [39] was applied to remove the scattered electronic density over the tubular voids occupied by disordered solvent molecules. The process removed 483 electrons over 1361 ų per unit cell. All hydrogen atoms were located in Fourier difference maps and included as fixed contributions according to the riding model [40]. The following restrain values were used in the structural refinements: C–H = 0.93 Å and U_iso(H) = 1.2 U_eq(C) for the aromatic carbon atoms, C–H = 0.97 Å and U_iso(H) = 1.2 U_eq(C) for methylene carbon atoms, N–H = 0.89 Å and U_iso(H) = 1.2 U_eq(N) for aromatic amides, and N–H = 0.91 Å and U_iso(H) = 1.2 U_eq(N) for coordinated amines. Bulk and single crystal identities were checked by comparing the calculated (from single crystals) and experimental X-ray powder diffraction patterns (PXRD). The PXRD of 1 was obtained in a PanAnalytical Empyrean diffractometer, using Cu-Kα radiation in θ-θ mode (0.02° s⁻¹) with crushed single crystals mounted in a gyratory zero-background silicon sample holder. Figures of the crystalline structure of 1 were prepared using Mercury^® [41] and Vesta [42] software. The crystal structure data were deposited under CCDC code 238974 and can be obtained free of charge from The Cambridge Crystallographic Data Centre via www.ccdc.cam.ac.uk/data_request/cif (accessed on 30 October 2024). Crystal data, data collection, temperature, and structure refinement details are summarized in

Table 1. Summary of the crystallographic parameters of 1.

Compound	1
Chemical Formula	C56H68N20O16Cu6
M/g mol−1	1658.54 *
λ/Å	0.71073
Crystal System	Triclinic
Space Group	$P\overline{1}$
a/Å	13.9356(4)
b/Å	15.5338(5)
c/Å	22.5074(7)
α/°	72.610(3)
β/°	76.845(3)
γ/°	67.435(3)
Volume/Å3	4258.2(3)
T/K	120(2)
Z	2
ρcalc/g cm−3	1.294
Crystal size/mm3	0.131 × 0.279 × 0.435
μ/mm−1	1.535
Reflections collected (Rint)	51,218 (0.0525)
Independent reflections	17,441
Reflections with I ≥ 2σ (I)	11,551
R (R total) a	0.0566 (0.1427)
wR (wR total) b	0.0849 (0.1568)
S c	1.048
ρmax, ρmin/e Å−3	+0.988; −0.520
CCDC number	2,389,741

^aR = Σ||F_o| − |F_c||/Σ|F_o|. ^b wR = [Σw(|F_o|² − |F_c|²)²/Σw|F_o|²]^1/2. ^c S = [Σw(|F_o|² − |F_c|²)²/(n_o − n_p)]^1/2, where $w\propto 1/\sigma$, n_o = observed parameters, and n_p = fitted parameters. * The SQUEEZE tool was applied to remove disordered solvent molecules inside the crystal structure pores.

.

2.3. Magnetic Measurements

The magnetic susceptibility measurements were carried out on single crystals 1 in the temperature range of 3.0–265 K with a Quantum Design SQUID magnetometer and using an applied magnetic field of 0.1 T. The sample was inserted in the SQUID at 265 K to quickly freeze it to minimize crystallization solvent evaporation during the air removal process. Upon reaching 3.0 K, the temperature was fixated, and the magnetization was evaluated by varying the magnetic field in the 0.1–7.0 T range. Diamagnetic corrections of the constituent atoms were estimated from Pascal’s constants [43]. Experimental susceptibilities were also corrected for the magnetization of the sample holder (a plastic bag).

3. Results

3.1. Synthesis and Product Formation Comments

Compound 1 was obtained in a basic medium. Neutral or acidic environments did not lead to the formation of 1. The medium was tested by basifying it with a strong base (n-Bu₄NOH) or a weak base (Et₃N), but only the strong one led to the formation of 1. The strong base is necessary to deprotonate the pyrazole group—which could be deprotonated with weaker bases—and to deprotonate the coordinated ethylene glycol and promote the formation of the ethane-1,2-dioxide ligand. It is the second time that ethylene glycol has been observed in a doubly deprotonated form, acting as a bridging ligand for three copper(II) ions. It is the first time it has led to the formation of a discrete polynuclear compound. Its deprotonation is highly supported due to the lack of vibrations in the IR spectrum characteristic of nitrate anions, which, if present, would indicate that the hexanuclear complex is cationic due to the coordination of ethylene glycol in its neutral form. Its deprotonation is also supported by the fact that a crystallization ethylene glycol molecule (before the SQUEEZE tool was applied) is guiding its O–H group towards the coordinated oxygen atoms, indicating a lack of hydrogen atoms in the coordinated ethane-1,2-dioxide, acting as a hydrogen bond acceptor.

The thermal analysis (Figure S1) was used to characterize the solvent content in the voids. It revealed an initial mass loss of 10.1% between room temperature and 130 °C, as an endothermic event. It is associated with the evaporation of crystallization solvent molecules. Between 130 °C and 400 °C, four exothermic events were observed related to organic matter oxidation. After that, up to 800 °C, a plateau is observed at 22.5% of the residual mass, indicating complete oxidation of the organic matter, with CuO most likely as the residue. To investigate the solvent loss, we must consider that the methanol and ethylene glycol used in the synthesis of 1 were used without a drying process, meaning that they had a considerable water content, which must be occupying the voids. Additionally, after letting the sample dry in the open atmosphere and at room temperature, the methanol molecules inside the voids can evaporate and be replaced by the moisture in the air. The first weight loss corresponds to 12 water molecules per hexanuclear complex (calculated = 10.2%). The organic matter oxidation masks the release of crystallization ethylene glycol and ethylenediamine molecules, but the weight loss of 68.1% is associated with four pyox ligands, five ethylene glycol molecules (two in the form of ethane-1,2-dioxide ligand and three of crystallization), and five ethylenediamines (four as ligands and one of crystallization), which, when combined, correspond to 70.3% of 1′s mass. The residual mass corresponds to six moles of CuO per complex, as expected (calculated = 22.5%).

To support the characterization of the void content, the SQUEEZE tool of PLATON was used to remove all electron density associated with the crystallization solvent molecules. Per unit cell, a total of 483 electrons were removed, spread across 1360 ų. Since Z = 2 for 1, a total of 242 electrons are associated with the crystallization molecules per hexanuclear unit. Combining the 12 water molecules, the three ethylene glycol molecules, and one ethylenediamine group (C₈H₅₀N₂O₁₈) resulted in 220 electrons. Considering the atomic volume (approximately. 14.0 ų for non-hydrogen atoms, i.e., atomic radii of about 1.5 Å, and 5.6 ų for hydrogen atoms, with atomic radii of about 1.1 Å), these molecules would occupy a volume 667 ų, almost half of the volume occupied by the electron density eliminated using the SQUEEZE tool. The combination of the electron count and squeezed volume is very close to the proposed void content, in agreement with the observed thermal analysis, and also aligns with the results of both elemental and thermogravimetric analysis.

3.2. Crystal Structure Description

Compound 1 is a neutral copper(II) hexanuclear complex crystallized in the centrosymmetric space group $P\overline{1}$, as seen in Figure 1a. It crystallizes with several solvent molecules that were omitted due to their high disorder, occupying a large tubular void. This polynuclear complex is composed of six metal atoms, four pyox²⁻ ligands, four ethylenediamine ligands, and two doubly deprotonated ethylene glycol ligands. It consists of two subunits, each containing half of the complex atoms, resulting in two trimetallic subunits, as shown in Figure 1b,c. The three copper atoms in each subunit are in a linear arrangement, linked by pyrazolate groups in a μ-κN:κN′ coordination fashion. The central metal ion is coordinated to ethane-1,2-dioxide, while the terminal copper atoms are each coordinated to one ethylenediamine molecule, as well as one oxygen atom of the ethane-1,2-dioxide ligand, which is found in the μ₃-κ²OO′:κO:κO′ double-bridging mode. Thus, the metal atoms are bridged by pyrazolate and ethane-1,2-dioxide ligands, forming two five-membered rings fused on the central copper atom. The metal–ligand distances corroborate the presence of ethane-1,2-dioxide, as they are in the range of 1.918–1.964 Å, shorter than those of the ethylenediamine and pyrazolate ligands, whose bond lengths range from 1.981 to 2.018 Å (see

Table 2. Selected bond lengths and angles for metal atoms in 1.

Cu1				Cu2				Cu3
Cu1–L/Å		L–Cu1–L/°		Cu2–L/Å		L–Cu2–L/°		Cu3–L/Å		L–Cu3–L/°
Cu1–N2	1.981(3)	N2–Cu1–N13	94.94(14)	Cu2–N3	1.952(3)	N3–Cu2–N5	100.11(13)	Cu3–N6	1.964(3)	N6–Cu3–N15	173.16(16)
Cu1–N13	2.029(4)	N2–Cu1–N14	176.47(14)	Cu2–N5	1.961(3)	N3–Cu2–O9	88.8(3)	Cu3–N15	1.992(4)	N6–Cu3–N16	93.04(15)
Cu1–N14	1.993(3)	N2–Cu1–O12	92.24(10)	Cu2–O9	2.71(1)	N3–Cu2–O13	87.13(12)	Cu3–N16	2.019(4)	N6–Cu3–O9	78.0(7)
Cu1–O12	2.527(4)	N2–Cu1–O13	85.56(12)	Cu2–O13	1.926(3)	N3–Cu2–O14	169.07(13)	Cu3–O9	2.64(2)	N6–Cu3–O14	87.95(13)
Cu1–O13	1.918(3)	N13–Cu1–N14	83.87(15)	Cu2–O14	1.921(3)	N5–Cu2–O9	86.2(3)	Cu3–O14	1.940(3)	N15–Cu3–N16	84.25(16)
		N13–Cu1–O12	84.61(10)			N5–Cu2–O13	163.58(13)			N15–Cu3–O9	96.3(7)
		N13–Cu1–O13	176.10(13)			N5–Cu2–O14	88.53(13)			N15–Cu3–O14	94.58(14)
		N14–Cu1–O12	85.35(10)			O9–Cu2–O13	108.9(3)			N16–Cu3–O9	100.8(7)
		N14–Cu1–O13	95.86(13)			O9–Cu2–O14	85.1(3)			N16–Cu3–O14	178.08(14)
		O12–Cu1–O13	99.24(10)			O13–Cu2–O14	86.30(12)			O9–Cu3–O14	77.8(7)
Cu4				Cu5				Cu6
Cu4–L/Å		L–Cu4–L/°		Cu5–L/Å		L–Cu5–L/°		Cu6–L/Å		L–Cu6–L/°
Cu4–N8	1.983(3)	N8–Cu4–N17	95.70(15)	Cu5–N9	1.952(3)	N9–Cu5–N12	99.68(14)	Cu6–N11	1.963(4)	N11–Cu6–N19	163.27(16)
Cu4–N17	2.016(4)	N8–Cu4–N18	173.99(16)	Cu5–N12	1.956(3)	N9–Cu5–O6	97.18(10)	Cu6–N19	2.018(4)	N11–Cu6–N20	93.10(15)
Cu4–N18	2.021(3)	N8–Cu4–O3	87.02(10)	Cu6–O6	2.934(4)	N9–Cu5–O15	86.11(13)	Cu6–N20	2.018(4)	N11–Cu6–O6	85.25(10)
Cu4–O3	2.442(3)	N8–Cu4–O15	85.74(13)	Cu5–O15	1.930(3)	N9–Cu5–O16	168.82(14)	Cu6–O6	2.846(4)	N11–Cu6–O16	87.19(13)
Cu4–O15	1.921(3)	N17–Cu4–N18	83.63(15)	Cu5–O16	1.951(3)	N12–Cu5–O6	84.66(10)	Cu6–O16	1.964(3)	N19–Cu6–N20	84.26(16)
		N17–Cu4–O3	95.48(19)			N12–Cu5–O15	168.58(14)			N19–Cu6–O6	79.74(19)
		N17–Cu4–O15	171.64(15)			N12–Cu5–O16	88.78(13)			N19–Cu6–O16	96.89(14)
		N18–Cu4–O3	87.11(19)			O6–Cu5–O15	104.49(10)			N20–Cu6–O16	174.90(16)
		N18–Cu4–O15	95.79(13)			O6–Cu5–O16	76.22(10)			N20–Cu6–O6	106.82(19)
		O3–Cu4–O15	92.81(10)			O15–Cu5–O16	86.86(12)			O6–Cu6–O16	78.29(10)

), and also shorter than the other observed examples of ethylene glycol coordinated to divalent copper ions [44,45,46].

The sum of the bridging ligands culminates in the copper atoms being very close to each other inside each trinuclear subunit. The intramolecular metal–metal distances are 3.3317(6) Å for Cu1···Cu2, 3.2572(7) Å for Cu2···Cu3 and 6.3156(7) Å for Cu1···Cu3, with an angle of 146.88(2)° for Cu1···Cu2···Cu3. For the second trinuclear subunit, the distances are equal to 3.2787(7) Å for Cu4···Cu5, 3.2264(9) Å for Cu5···Cu6, and 6.2565(8) Å Cu4···Cu6, with an angle of 148.22(2)° for Cu4···Cu5···Cu6. The angle between the coordination planes of the connected metal atoms is equal to 30.9(4)° for Cu1N2N13N14 and Cu2N3N5O13O14 mean planes, 35.8(4)° for the angle between the latter and the Cu3N6N15N16O14 mean plane, 37.8(5)° for Cu4N8N17N18O15 and Cu5N9N11O15O16 mean planes, and finally 47.5(4)° between the latter and the Cu6N12N19N20O16 mean plane. The torsion angle around the pyrazolate N–N bond is equal to −4.1(1.4)° for Cu1–N2–N3–Cu2, −5.5(1.2)° for Cu2–N5–N6–Cu3, 0.3(1.4)° for Cu4–N8–N9–Cu5, and −0.43(1.3)° for Cu5–N11–N12–Cu6. It is also important to note the angle concerning the alkoxide bridging ligand of the ethane-1,2-dioxide ligand: 120.23(14)° for Cu1–O13–Cu2, 115.20(14)° for Cu2–O14–Cu3, 116.76(14)° for Cu4–O15–Cu5, and 110.96(14)° for Cu5–O16–Cu6.

The oxamate portion of each subunit plays a role in connecting the subunits. The amide oxygen atoms coordinate with the copper atoms of the other subunit. In each subunit, one amide group is monodentate, while the other bridges the central and a terminal copper atom. Consequently, all copper atoms are found in five coordination spheres, as shown in Figure 2a,b. The outer copper atoms (Cu1, Cu3, Cu4, and Cu6) are in a CuN₃O₂ environment, while the central ones (Cu2 and Cu5) are in a CuN₂O₃ coordination sphere. The trigonality parameter τ varies from 0.01 to 0.19, indicating a coordination geometry closer to square-based pyramidal (τ = 0 represents perfect square-based geometry, and τ = 1 represents trigonal bipyramidal geometry) [47]. The major distortion in the five-coordinate sphere for the copper atoms arises from the apical amide oxygen atom, which is shared by the Cu2, Cu3, Cu5, and Cu6 ligands. This sharing results in several geometric restrictions that cause both bond length elongation and bond angle deviation from the ideal (see Table 2). Due to this significant distortion in the coordination geometry and to corroborate the results obtained from the trigonality parameter analysis, the SHAPE software [48] was employed, using the metal and ligand atomic positions extracted from the crystal structure. Although deviations from perfect coordination geometries were confirmed for all copper atoms, they are closest to square-based geometry (See Table S2 in Supplementary Information). Finally, it is important to highlight that in the hexanuclear complex, the trimetallic units are entirely separated from each other. The closest intramolecular metal–metal distance is 9.5055(3) Å for Cu2···Cu5.

Although the carboxylate group has a higher electron density than the carbonyl portion of the amide, its inability to coordinate with copper atoms is attributed to molecular recognition involving intramolecular hydrogen bonds. All carboxylate groups are found to interact with the N–H groups of ethylenediamine (see Figure 2). The combination of intramolecular hydrogen bonds and the coordination of the metal by the amide carbonyl stabilizes the complex more than if the carboxylate were only coordinated with the copper, as the apical Cu–O bond lengths are quite large and might not provide sufficient stabilization if the carboxylate were solely coordinated.

From a supramolecular perspective, crystal packing is dictated by the hydrogen bonds between the hexanuclear units (see all hydrogen bond geometry in Table S1 in the Supporting Information). The N–H groups of ethylenediamine and oxamate amide portion act as hydrogen bond donors, recognizing carboxylate groups around them. The hexanuclear complexes interact via amide-to-carboxylate bonds in dimeric forms along with the crystallographic a-axis This means that while one oxamate amide recognizes the neighboring oxamate carboxylate portion [N1–H1···O7ⁱ, i = x + 1, y, z], the latter interacts back, with its amide engaging with the carboxylate of the first oxamate [N7ⁱ–H7ⁱ···O1]. This interaction is defined by Etter et al. [49] as $R_2^2\left(10\right)$, indicating that it forms a ring with two hydrogen bond donors (both amide N–H) and two acceptors (one carboxylate of each oxamate) composed of ten atoms (two ···H–N–C–C–O··· units). This interaction forms a supramolecular 1D structure (See Figure 3a). In addition, also along the crystallographic a-axis direction, ethylenediamine groups are connecting to the neighboring supramolecular chain’s carboxylate, and once again, it interacts back, forming a motif denominated as $R_4^2\left(8\right)$. This means that two ethylenediamine NH₂ groups interact with two terminal oxygen atoms in a ring configuration, with a total of eight interacting atoms [N19–H19A···O15 and N19–H19B···O15^iv, iv = −x + 1, −y + 2, −z + 1]. Thus, the combination of both interactions leads to a supramolecular 1D double chain along the a-axis, as shown in Figure 3a. These double chains interact along the crystallographic b- and c-axis via ethylenediamine–carboxylate discrete hydrogen bonds [N16—H16B···O10ⁱⁱⁱ, iii = −x, −y + 2, −z; N20–H20–O11^v, v = −x + 1, −y + 2, −z] to reach a three-dimensional supramolecular structure. The sum of all interactions forms a net, leaving a large zigzag tubular void, which hosts disordered crystallization solvent molecules removed from the crystal structure using the SQUEEZE technique, as depicted in Figure 3b.

3.3. Magnetic Studies

The direct current (DC) magnetic susceptibility measurement shows that at 265 K, the ${\chi }_{M}T$ is equal to 1.44 cm³ mol⁻¹ K and is in a decreasing regime as the sample is cooled down (Figure 4). The ${\chi }_{M}T$ expected value for six copper(II) atoms, with respect to the spin-only approximation, is 2.48 cm³ mol⁻¹ K (S = 1/2 and g = 2.1). The lower experimental value is associated with an active antiferromagnetic regime at high temperatures, indicating a somewhat strong coupling between the copper atoms. Upon cooling, a plateau is reached at 50 K, and the ${\chi }_{M}T$ is equal to 0.84 cm³ mol⁻¹ K, which is close to the expected value for two isolated spin doublets (0.83 cm³ mol⁻¹ K). The observed ${\chi }_{M}T$ and the magnetic behavior can be related to an antiferromagnetic coupling between the copper(II) atoms in the trinuclear subunits and a very weak to negligible magnetic interaction between the subunits, although they are connected via pyox ligands. Below 15 K, a second antiferromagnetic regime is observed, evidenced by a decrease in the ${\chi }_{M}T$ value. This probably occurs due to dipolar interactions throughout the space of the neighboring hexanuclear complexes. This coupling mechanism is evidenced by the low temperature at which the second antiferromagnetic regime occurs and the slight decrease observed at the lowest temperature (3.0 K) that accounts for a ${\chi }_{M}T$ equal to 0.78 cm³ mol⁻¹ K, which is also a signature of a very weak antiferromagnetic coupling.

The magnetization curve as a function of the applied field (M vs. H) at 3.0 K revealed incomplete saturation at 7.0 T (M_sat = 1.87 Nβ). This corroborates the presence of two spin doublets due to antiferromagnetic coupling between the metal atoms in each trinuclear subunit (S_total = 1/2), which is expected to exhibit a saturation of the magnetization of 2.1 Nβ. However, as observed, at 3.0 K, in the ${\chi }_{M}T$ vs. T curve, the active antiferromagnetic dipolar interactions may cause the magnetization to not reach the expected saturation value.

To investigate the magnitude of the magnetic coupling, the susceptibility data were fitted considering the coupling between the copper atoms in the subunits with the central copper atom coupled to the outer ones, as indicated in Equation (1). J′ and J″ are the magnetic coupling constants for each central atom and the outer ones in the different subunits. The terms j′ and j″ are the magnetic coupling between the outer metal atoms in each trinuclear subunit

$$\begin{array}{l}\mathit{H}=-J^{\prime }\left({\mathit{S}}_{Cu1}·{\mathit{S}}_{Cu2}+{\mathit{S}}_{Cu2}·{\mathit{S}}_{Cu3}\right)+J^{\prime }({\mathit{S}}_{Cu1}·{\mathit{S}}_{Cu3})\\ \hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}\hspace{1em}-J″\left({\mathit{S}}_{Cu4}·{\mathit{S}}_{Cu5}+{\mathit{S}}_{Cu5}·{\mathit{S}}_{Cu6}\right)+j″({\mathit{S}}_{Cu4}·{\mathit{S}}_{Cu6})\end{array}$$

(1)

It is important to remark that the magnetic coupling between the outer copper atoms can be neglected, since they are not directly bonded. One can also assume that the subunits can be considered magnetically equivalent because the coordination parameters around the metal atoms are very similar, although they are not symmetry related. It will simplify the Hamiltonian for the presented molecule, as described in Equation (2). Therefore, the fitted coupling constant J will be the average of the coupling constants J′ and J″ for each trinuclear subunit.

$$\mathit{H}=-2J({\mathit{S}}_{Cu1}·{\mathit{S}}_{Cu2}+{\mathit{S}}_{Cu2}·{\mathit{S}}_{Cu3})$$

(2)

Following the Hamiltonian, the susceptibility must adapt to Equation (3) [50]

$$\begin{array}{c}{\chi }_{M}T(T)={\displaystyle \frac{N{g}^2{\beta }^2}{2k(T-\theta )}}{\displaystyle \frac{A}{B}}\\ A=1+e^{-{\displaystyle \frac{J}{kT}}}+10e^{{\displaystyle \frac{J}{2kT}}}\\ B=1+e^{-{\displaystyle \frac{J}{kT}}}+2e^{{\displaystyle \frac{J}{2kT}}}\end{array}$$

(3)

$N$, $g$, $\beta$, and $k$ are the Avogadro number, the average Landé factor for all copper(II) atoms, the Bohr magneton, and the Boltzmann constant, respectively. $\theta$ is the deviation from the Curie law indicating dipolar interactions, which only occur at very low temperatures. An excellent fitting between the experimental and calculated data was obtained with $J$ = −106.3 cm⁻¹, $g$ = 2.12, and $\theta$ = −0.23 K with $R$ = 2.9 × 10⁻⁵. The ${\chi }_{M}T$ curve fitting was also performed with PHI software (version 3.1.6) [51], which provided very close results, obtaining $J$ = −116.5(3) cm⁻¹, $g$ = 2.12(1), and $\theta$ = −0.08(1) K with $R$ = 1.4 × 10⁻³ ($R$ is the agreement factor defined as $R=\sum {\left(P_{obs}-P_{calc}\right)}^2/\sum {\left(P_{obs}\right)}^2$, with $P$ being the physical parameter). This convergence in both analytical and numerical methods is indicative that the trinuclear subunits are almost magnetically isolated and that the provided simplified Hamiltonian in Equation (2) describes the magnetic behavior of this copper(II) hexanuclear compound very well. A comparison between the curve adjustments is shown in Figure 4. The obtained Landé factor is consistent with that of copper(II) ions, and the found coupling constants fall within the range typically found for complexes containing three metal ions in a linear arrangement, as can be concluded from the content of Table 3.

It is important to note that when only oxamate bridges are present in the trinuclear units, blocking ligands are important to orient the overlapping magnetic orbitals. In the presence of meridional orienting blocking ligands (Table 3), the bidentate oxamate can only bite the metal in apical–basal positions, causing a large angle between the basal planes and thus weakening antiferromagnetic coupling, as observed in [{Cu(bpca)}₂Cu(opba)(H₂O)]·H₂O [bpca = bis(2-pyridylcarbonyl)amidate; opba = N,N′-phenylene-1,2-bis(oxamate)] [50]; and [{Cubpca)}₂{Cu(dmopba)(H₂O)}]₂·4H₂O [dmopba = N,N′-4,5-dimethylphenylene-1,2-bis(oxamate)] [52]. When bidentate-blocking ligands are employed, the oxamate usually occupies two basal positions, leading to a smaller angle between the basal planes and drastically increasing magnetic orbital overlapping. For instance, in [{Cu(Me₄en)(H₂O)}₂Cu(pba)(ClO₄)] [53] and [{Cu(Me₄en)(H₂O)}₂Cu(pba)(H₂O)](PF₆)₂ (Me₄en = N,N,N′N′-tetramethylethylenediamine; pba = N,N′-propylbis(oxamate)] [54], the angles between the basal planes are lower than the previous examples, exhibiting antiferromagnetic coupling that is ten times more intense than observed for [{Cubpca)}₂{Cu(dmopba)(H₂O)}]₂·4H₂O (J values of −334.4, −342.1, and −31.96 cm⁻¹, respectively). In addition, for [{Cu(Me₄en)(H₂O)}₂Cu(pba)(ClO₄)] and [{Cu(Me₄en)(H₂O)}₂Cu(pba)(H₂O)](PF₆)₂, it is observed that the one with the lower angle between the basal planes presented a slightly stronger antiferromagnetic coupling constant due to more optimal orbital overlapping. Ferromagnetic couplings in the oxamate derivatives bridging examples arise from the bridging of metal ions with orthogonal magnetic orbitals [55,56] or different bridging forms other than µ-κ²N,O:κ²O′,O″ [52,57,58].

On the other hand, both alkoxide and pyrazolate ligands lead to antiferromagnetic coupling between copper(II) ions. For the pyrazolate group, it is observed that the torsion angle around the N–N bond is a determinant for orbital overlapping. A smaller torsion angle means that the basal planes of the linked metal atoms are closer to coplanarity, and the spin density can be delocalized over the plane of the pyrazolate aromatic ring. To exemplify, both [Cu₃(L1)₂(NO₃)₂(H₂O)₂]·3H₂O and [Cu₃(L1)₂Cl₂(CH₃OH)₂]·2H₂O [H₂L1 = 6,6-((1E,10E)-hydrazine-1,2-diylidenebis(methanylylidene))bis(2-methoxyphenol)] [59] have almost the same alkoxide bridging angles, but the small variation of the torsion angle from 28.86° to 30.54° causes the antiferromagnetic coupling to weaken from −113.1 to −98.3 cm⁻¹. For alkoxide ligands, the direct bridging angle is crucial to magnetic overlapping. The bridging angle closest to 120° indicates better fitting to sp² hybridization for the oxygen atom; thus, better magnetic overlapping can arise from the ligand. As seen in Table 3, the strongly antiferromagnetically coupled atoms have a Cu–O–Cu angle close to 120°, suppressing even the downside of having a somewhat greater torsion angle along the pyrazolate N–N bond. However, it can be concluded that this bridging angle has the drawback that even small variations can drastically impact the magnetic coupling between the bridged metal atoms.

Finally, the strength of copper(II) magnetic coupling depends not solely on the nature of the ligands. Copper(II) ions can have ${\mathrm{d}}_{{\mathrm{z}}^2}$ and ${\mathrm{d}}_{{\mathrm{x}}^2-{\mathrm{y}}^2}$ as their magnetic orbitals, and thus small variations in ion coordination geometry, as well as the angle and/or bond length between the bridging atoms/ligands, cause variations in the crystal field and consequently in the metal’s magnetic orbitals. Since each orbital resides in a different atomic axis, swapping them will significantly affect the quality of orbital overlap, finally resulting in the wide range of magnetic couplings constants observed in the literature.

Table 3. Magneto-structural properties of some oxamate-bridged or pyrazolate/oxo-bridged linear arranged trinuclear copper(II) compounds.

Compound	Cu–NN–Cu a/°	Cu–O–Cu b/°	Mean Planes Angle c/°	J/cm−1	g	θ or zJ′/K	Ref.
1	−4.1 (1.4)/−5.5 (1.2)/ 0.3 (1.4)/−0.43 (1.3)	120.23(14)/115.20(14)/ 116.76(14)/110.96(14)	30.9(4)/35.8(4)/37.8(5)/47.5(4)	−106.3 d	2.12 d	−0.23 d	This work
				−116.5 d	2.12 d	−0.08 d
[{Cu(bpca)}2Cu(opba)(H2O)]·H2O	--- e	--- e	78.45/84.62	−65.8	2.08	---	[50]
[{Cu(bpca)}2{Cu(dmopba)(H2O)}]2·4H2O	--- e	--- e	78.09/81.67	−31.96/+1.34 f	2.12	---	[52]
[{Cu(Me4en)(H2O)}2Cu(pba)(ClO4)]	--- e	--- e	11.29/18.50	−334.4	2.08	−0.36	[53]
[{Cu(Me4en)(H2O)}2Cu(pba)(H2O)](PF6)2	--- e	--- e	7.94/10.75	−342.1	2.14	−0.61	[54]
[Cu3(L1)2(BF4)(H2O)2]BF4	40.5(2)/33.0(1)	116.23(7)/118.40(7)	17.04	−170	2.16	---	[59]
[Cu3(L1)2(NO3)2(H2O)2]·3H2O	28.86	116.95	5.01	−113.1	2.08	---
[Cu3(L1)2Cl2(CH3OH)2]·2H2O	30.54	116.89	4.64	−98.3	2.07	---
[Cu3(L1)2(ClO4)2(H2O)2]	40.19/28.92	117.67/117.91	16.20	−249.7	2.14	---
{[Cu3(L2)2(H2O)2](ClO4)4}·4H2O	17.2(8)/13.7(8)	120.5(3)/116.6(3)	8.19/18.36	−179	2.0/2.2	---	[60]
[Cu3(HYDRAV)2Cl2]	35.9	115	0.96/11.7	−7.21	2.03	---	[61]
{[Cu3(L3)(L4)2(NO3)2(H2O)2]NO3·1.5H2O}n	−6.9(8)/8.0(8)	119.3(3)/124.5(3)	22.9/33.6	−149/−175	2.20	−0.7	[62]

^a Torsion angle along N–N bond. ^b Oxo bridge angle. ^c For metal atoms with NC = 5 (square-based pyramid geometry), basal planes were used. For those with NC = 6 (octahedral geometry), planes perpendicular to the elongated bonds’ direction were considered. ^d J, g, and zJ/θ were calculated using Equation (2) (top line) and using PHI software (bottom line). ^e Not present for oxamate-bridged metal atoms. ^f Ferromagnetic coupling due to syn-anti μ³-carboxylate from oxamate group. Ligand abbreviations: bpca = bis(2-pyridylcarbonyl)amidate; opba = N,N′-phenylene-1,2-bis(oxamate); dmopba = N,N′-4,5-dimethylphnylene-1,2-bis(oxamate); Me₄en = N,N,N′,N′-tetramethylethylenediamine; pba = N,N′-propilbis(oxamate); H₂L1 = 6,6-((1E,10E)-hydrazine-1,2-diylidenebis(methanylylidene))bis(2-methoxyphenol); HL2 = [(5-amino-1H-1,2,4-triazol-3-yl)amino](1H-imidazol-5-yl)methanolate. HYDRAV2 = Schiff base made with 1-(hydrazineylidenemethyl)naphthalen2-ol, o-vanillin and (2-hydroxy-3-methoxybenzaldehyde); H₂L3 = 4-[bis(pyridin-2-yl-methanol)]amino1,2,4-triazole; L4 = 4-amino-4H-1,2,4-triazole.

Table 3. Magneto-structural properties of some oxamate-bridged or pyrazolate/oxo-bridged linear arranged trinuclear copper(II) compounds.

Compound	Cu–NN–Cu a/°	Cu–O–Cu b/°	Mean Planes Angle c/°	J/cm−1	g	θ or zJ′/K	Ref.
1	−4.1 (1.4)/−5.5 (1.2)/ 0.3 (1.4)/−0.43 (1.3)	120.23(14)/115.20(14)/ 116.76(14)/110.96(14)	30.9(4)/35.8(4)/37.8(5)/47.5(4)	−106.3 d	2.12 d	−0.23 d	This work
				−116.5 d	2.12 d	−0.08 d
[{Cu(bpca)}2Cu(opba)(H2O)]·H2O	--- e	--- e	78.45/84.62	−65.8	2.08	---	[50]
[{Cu(bpca)}2{Cu(dmopba)(H2O)}]2·4H2O	--- e	--- e	78.09/81.67	−31.96/+1.34 f	2.12	---	[52]
[{Cu(Me4en)(H2O)}2Cu(pba)(ClO4)]	--- e	--- e	11.29/18.50	−334.4	2.08	−0.36	[53]
[{Cu(Me4en)(H2O)}2Cu(pba)(H2O)](PF6)2	--- e	--- e	7.94/10.75	−342.1	2.14	−0.61	[54]
[Cu3(L1)2(BF4)(H2O)2]BF4	40.5(2)/33.0(1)	116.23(7)/118.40(7)	17.04	−170	2.16	---	[59]
[Cu3(L1)2(NO3)2(H2O)2]·3H2O	28.86	116.95	5.01	−113.1	2.08	---
[Cu3(L1)2Cl2(CH3OH)2]·2H2O	30.54	116.89	4.64	−98.3	2.07	---
[Cu3(L1)2(ClO4)2(H2O)2]	40.19/28.92	117.67/117.91	16.20	−249.7	2.14	---
{[Cu3(L2)2(H2O)2](ClO4)4}·4H2O	17.2(8)/13.7(8)	120.5(3)/116.6(3)	8.19/18.36	−179	2.0/2.2	---	[60]
[Cu3(HYDRAV)2Cl2]	35.9	115	0.96/11.7	−7.21	2.03	---	[61]
{[Cu3(L3)(L4)2(NO3)2(H2O)2]NO3·1.5H2O}n	−6.9(8)/8.0(8)	119.3(3)/124.5(3)	22.9/33.6	−149/−175	2.20	−0.7	[62]

^a Torsion angle along N–N bond. ^b Oxo bridge angle. ^c For metal atoms with NC = 5 (square-based pyramid geometry), basal planes were used. For those with NC = 6 (octahedral geometry), planes perpendicular to the elongated bonds’ direction were considered. ^d J, g, and zJ/θ were calculated using Equation (2) (top line) and using PHI software (bottom line). ^e Not present for oxamate-bridged metal atoms. ^f Ferromagnetic coupling due to syn-anti μ³-carboxylate from oxamate group. Ligand abbreviations: bpca = bis(2-pyridylcarbonyl)amidate; opba = N,N′-phenylene-1,2-bis(oxamate); dmopba = N,N′-4,5-dimethylphnylene-1,2-bis(oxamate); Me₄en = N,N,N′,N′-tetramethylethylenediamine; pba = N,N′-propilbis(oxamate); H₂L1 = 6,6-((1E,10E)-hydrazine-1,2-diylidenebis(methanylylidene))bis(2-methoxyphenol); HL2 = [(5-amino-1H-1,2,4-triazol-3-yl)amino](1H-imidazol-5-yl)methanolate. HYDRAV2 = Schiff base made with 1-(hydrazineylidenemethyl)naphthalen2-ol, o-vanillin and (2-hydroxy-3-methoxybenzaldehyde); H₂L3 = 4-[bis(pyridin-2-yl-methanol)]amino1,2,4-triazole; L4 = 4-amino-4H-1,2,4-triazole.

4. Conclusions

In this study, we report the synthesis, crystal structure, and magnetic properties of a novel hexacopper(II) complex, in which ethylene glycol serves both as a solvent for its crystallization and as a ligand in its doubly deprotonated form, ethane-1,2-dioxide. The crystal structure reveals that the hexanuclear complex consists of two linear trinuclear copper(II) units, with the central copper atom partially coordinated by ethane-1,2-dioxide, while each terminal copper atom coordinates one ethylenediamine molecule. Both pyrazolate (from the pyox ligand) and alkoxide groups (from ethane-1,2-dioxide) serve as bridging ligands between the external and central copper ions. The two trinuclear units are connected by the coordination of a carbonyl oxygen atom (from the oxamate portion of the pyox ligand) to the copper centers of the opposing trinuclear unit. However, the distance between them is relatively large and involves a pathway consisting of non-resonant sp² atoms. DC magnetic measurements indicate that the copper ions exhibit strong antiferromagnetic interactions at high temperatures, indicating an active antiferromagnetic regime. The plateau observed at 50 K corresponds to two isolated spin doublets, suggesting that the trinuclear units are not magnetically connected. This is consistent with the assumption that they are only linked to the extremities of the pyox ligand. The magnetic data were fitted using analytical and numerical methods, providing the best fits of the exchange interaction parameters of $J$ = −106.3 cm⁻¹ and $g$ = 2.12 and $J$ = −116.5 cm⁻¹ and $g$ = 2.12, respectively. The found values align with previously reported values for pyrazolate- and alkoxide-bridged copper(II) complexes.