Two-Dimensional Nitronyl Nitroxide–Cu Networks Based on Multi-Dentate Nitronyl Nitroxides: Structures and Magnetic Properties

: Two multi-dentate nitronyl nitroxide radicals, namely, bisNITPhPy ([5-(4-pyridyl)-1,3-bis(1 (cid:48) -oxyl-3 (cid:48) -oxido-4 (cid:48) ,4 (cid:48) ,5 (cid:48) ,5 (cid:48) -tetramethyl-4,5-hydro-1 H -imidazol-2-yl)]benzene) and NIT-3Py-5-4Py (2-{3-[5-(4-pyridyl)]pyridyl}-4,4,5,5-tetramethylimidazoline-1-oxyl-3-oxide), were assembled with Cu II ions to obtain two-dimensional heterospin 2p–3d coordination polymers [Cu 7 (hfac) 14 (bisNITPhPy) 2 ] n ( 1 ) and [Cu 2 (hfac) 4 (NIT-3Py-5-4Py)] n ( 2 ) (hfac: hexaﬂuoroacetylacetonate). In both compounds, the bisNITPhPy and NIT-3Py-5-3Py radicals acted as pentadentate and tetradentate ligands, respectively, to connect with Cu II ions, generating a 2D layer structure. The analysis of the magnetic behavior indicated that strong antiferromagnetic coupling and ferromagnetic interaction ( J = 17.1 cm − 1 ) coexisted in 1 . For 2 , there were ferromagnetic couplings between the Cu II ion and NO group, as well as the Cu II ion and radical via the pyridine ring with J 1 = 32.8 and J 2 = 2.2 cm − 1 , respectively. for about 15 h to give cyan crystals. Yield 55%. Anal. calc. for C 37 H 23 Cu 2 F 24 N 4 O 10 (%): C 35.09, H 1.83, N 4.42; found. C 35.21, H 2.28, N 4.40; FT-IR (KBr, cm –1 ): 1650 (s), 1529 (m), 1465 (s), 1363 (m), 1251 (s), 1197 (s), 1133 (s), 799 (s), 662 (s), 578 (s), 526 (m).


Structural Investigation and Description
X-ray diffraction of both Cu II compounds 1 and 2 showed that they all crystallized in the triclinic space group Pī. The particulars of the crystallographic structural data of twodimensional polymers are included in Table 1. For compound 1, the asymmetric part contained seven Cu(hfac)2 moieties and two bisNITPhPy biradical ligands ( Figure 1). In 1, one NIT unit of each bisNITPhPy ligand bridged two different copper(II) ions (Cu4 and Cu5) through its two NO groups, leading to infinite linear chains. Meanwhile, 1D chains were connected to Cu(II) ions (Cu2 and Cu3) via another NIT moiety with the [Cu3-NIT-Cu2-NIT-Cu3] structural units to generate one two-dimensional layer. Meanwhile, the N atom of the pyridine ring of each bisNITPhPy ligand was connected to one Cu II ion (Cu1) in its equatorial position ( Figure 2).

Structural Investigation and Description
X-ray diffraction of both Cu II compounds 1 and 2 showed that they all crystallized in the triclinic space group Pī. The particulars of the crystallographic structural data of two-dimensional polymers are included in Table 1. For compound 1, the asymmetric part contained seven Cu(hfac) 2 moieties and two bisNITPhPy biradical ligands ( Figure 1). In 1, one NIT unit of each bisNITPhPy ligand bridged two different copper(II) ions (Cu4 and Cu5) through its two NO groups, leading to infinite linear chains. Meanwhile, 1D chains were connected to Cu(II) ions (Cu2 and Cu3) via another NIT moiety with the [Cu3-NIT-Cu2-NIT-Cu3] structural units to generate one two-dimensional layer. Meanwhile, the N atom of the pyridine ring of each bisNITPhPy ligand was connected to one Cu II ion (Cu1) in its equatorial position ( Figure 2).       Cu2, Cu4, and Cu5 adopted distorted octahedral geometries ( Figure S1). The equatorial positions were filled by two ß-diketonate coligands . These axial bond lengths were longer than equatorial distances, suggesting the Jahn-Teller effect was in action [42,43]. Five-coordinated Cu1 and Cu3 both had distorted pyramidal geometries, which were assessed using a shape program [44] ( Figure S1, Table S3). The coordination environment of Cu1 was built of a pyridyl N atom of the radical and four O atoms of the hfac − ligands, in which the O3 atom was located in the axial position (Cu1-O3: 2.166(3) Å). For Cu3, the equator plane was generated by several O atoms (O16, O5, O7, and O8) from a nitroxide unit of the bisNITPhPy ligand (Cu3-O16: 1.989(3) Å) and two hfac − ligands (Cu-O: 1.904(3)-1.963(3) Å), and the O6 atom of one hfac − ligand occupied the vertex (Cu3-O6: 2.146(3) Å). Within the two-dimensional sheet, the shortest distance of Cu· · · Cu was 6.246 Å ( Figure 2). Adjacent 2D planes stacked with the closest interlayer Cu· · · Cu distance of 9.364 Å.
For [Cu2(hfac)4(NIT-3Py-5-4Py)]n (2), asymmetric elements were made up of two Cu(hfac)2 units and one NIT-3Py-5-4Py radical ligand ( Figure 3). As shown in Figure 4, two NIT-3Py-5-4Py radicals linked Cu II ions by relying on nitroxide groups and nitrogen atoms of pyridines to produce an annular structure [Cu2(NIT-3Py-5-4Py)2]. Meanwhile, each dimer ring was linked to two Cu(hfac)2 units via two nitroxide units, yielding a 1D chain. Meanwhile, adjacent 1D chains coordinated with Cu II ions via N atoms of the pyridine rings to construct a network structure.   These Cu II ions all adopted an elongated octahedral geometry ( Figure S3). For Cu1, the apical positions were taken over by two oxygen atoms that arose from two different hfac − ligands (average Cu-O: 2.230(6) Å ), while the equatorial plane took shape via two These Cu II ions all adopted an elongated octahedral geometry ( Figure S3). For Cu1, the apical positions were taken over by two oxygen atoms that arose from two different hfac − ligands (average Cu-O: 2.230(6) Å), while the equatorial plane took shape via two nitrogen atoms from the two pyridine rings of two radicals (average Cu-N bond: 2.017(4) Å) and two oxygen atoms belonging to hfac − ligands (average Cu-O bond: 2.039 (5)  In a 2D layer, the nearest Cu· · · Cu separation was 5.911 Å (Figure 4). The shortest interlayer Cu· · · Cu interval between adjacent networks was 11.172 Å.

Magnetic Properties
A direct current magnetic susceptibility study of [Cu 7 (hfac) 14 (bisNITPhPy) 2 ] n (1) was implemented over the 2-300 K range with an extrinsic dc field of 5000 Oe. As described in Figure 5, the χ M T value of [Cu 7 (hfac) 14 (bisNITPhPy) 2 ] n (1) at 300 K was 3.15 cm 3 Kmol -1 , which was lower than the desired value of 4.125 cm 3 Kmol -1 for uncoupled seven copper ions and four mono-radicals, but close to the theoretical values of 2.625 cm 3 Kmol -1 for uncoupled five copper ions and two mono-radicals (Cu II : C = 0.375 cm 3 Kmol −1 and S = 1/2; meanwhile, for the radical: S = 1/2), suggesting strong antiferromagnetic coupling. On cooling, the χ M T value increased slowly during the range of 300-35 K and then further increased abruptly to reach the maximum value of 8.36 cm 3 Kmol −1 at 2.0 K. The above phenomenon shows that strong antiferromagnetic coupling and ferromagnetic interaction coexisted in 1. Based on the structure of compound 1, the O16 atom of the NO unit lay on one side of the Cu3 equatorial plane to generate strong antiferromagnetic interaction and completely offset the opposite spin [45,46]. Furthermore, Cu II -rad magnetic coupling via the pyridyl and phenyl rings might be very weak [47]. Thus, the magnetic behavior of 1 was mainly derived from where the mono-radical bridged two Cu II chains involving Cu4 and Cu5 ions and uncoupled three copper ions (Cu1, Cu2, and Cu1a) (Scheme 2). The observed magnetic behavior of 1 should be analyzed using Equations (1)- (4). For the Cu II chain, an equation (Equation (2)) for the magnetic susceptibility was based on the Hamiltonian ̂= −̂•̂ [48]. The mean field (zJ′) denotes additional magnetic coupling. Based on the structure of compound 1, the O16 atom of the NO unit lay on one side of the Cu3 equatorial plane to generate strong antiferromagnetic interaction and completely offset the opposite spin [45,46]. Furthermore, Cu II -rad magnetic coupling via the pyridyl and phenyl rings might be very weak [47]. Thus, the magnetic behavior of 1 was mainly derived from where the mono-radical bridged two Cu II chains involving Cu4 and Cu5 ions and uncoupled three copper ions (Cu1, Cu2, and Cu1a) (Scheme 2). The observed magnetic behavior of 1 should be analyzed using Equations (1)- (4). For the Cu II chain, an equation (Equation (2)) for the magnetic susceptibility was based on the Hamiltonian H = −JŜ rad ·Ŝ Cu [48]. The mean field (zJ ) denotes additional magnetic coupling.
The isothermal M versus H curve for complex 1 was investigated at 2.0 K in 0-70 kOe ( Figure 5). The magnetization displayed a sharp rise below 15 kOe, suggesting the existence of ferromagnetic couplings, in agreement with the precipitous increase of the χMT value in the low-temperature range. Then, M increased slowly to 7.16 Nβ at 70 kOe, in accordance with the theoretical values (7.0 Nβ) found using the Brillouin functions for free seven S = 1/2 spins, which verified the strong Cu II -rad antiferromagnetic interaction in the equatorial plane.
The χMT-T curve for [Cu2(hfac)4(NIT-3Py-5-4Py)]n (2) is described in Figure 6. The room temperature value of χMT was 1.21 cm 3 Kmol −1 , which approached the desired value of 1.125 cm 3 Kmol −1 for two uncoupled Cu II ions and one radical. On cooling, the χMT value of compound 2 increased steadily to 4.66 cm 3 Kmol −1 at 2 K, showing that ferromagnetic coupling dominated the 2p-3d magnetic system. The rational evaluation of the measurement data produced J = 17.11 cm −1 , g = 2.08, zJ = −0.044 cm −1 , and R 2 = 0.9963. The positive J parameter signifies the ferromagnetic Cu-radical coupling, which was derived from the axial Cu II -ON bond to give rise to orbital orthogonality of the π* orbital of the NO group and the d x 2 −y 2 orbital of the Cu II ion [49,50]. The magnetic coupling is analogous to some reported Cu II -nitronyl nitroxide compounds [42,51].
The isothermal M versus H curve for complex 1 was investigated at 2.0 K in 0-70 kOe ( Figure 5). The magnetization displayed a sharp rise below 15 kOe, suggesting the existence of ferromagnetic couplings, in agreement with the precipitous increase of the χ M T value in the low-temperature range. Then, M increased slowly to 7.16 Nβ at 70 kOe, in accordance with the theoretical values (7.0 Nβ) found using the Brillouin functions for free seven S = 1/2 spins, which verified the strong Cu II -rad antiferromagnetic interaction in the equatorial plane.
The χ M T-T curve for [Cu 2 (hfac) 4 (NIT-3Py-5-4Py)] n (2) is described in Figure 6. The room temperature value of χ M T was 1.21 cm 3 Kmol −1 , which approached the desired value of 1.125 cm 3 Kmol −1 for two uncoupled Cu II ions and one radical. On cooling, the χ M T value of compound 2 increased steadily to 4.66 cm 3 Kmol −1 at 2 K, showing that ferromagnetic coupling dominated the 2p-3d magnetic system. ence of ferromagnetic couplings, in agreement with the precipitous increase of the χMT value in the low-temperature range. Then, M increased slowly to 7.16 Nβ at 70 kOe, in accordance with the theoretical values (7.0 Nβ) found using the Brillouin functions for free seven S = 1/2 spins, which verified the strong Cu II -rad antiferromagnetic interaction in the equatorial plane.
The χMT-T curve for [Cu2(hfac)4(NIT-3Py-5-4Py)]n (2) is described in Figure 6. The room temperature value of χMT was 1.21 cm 3 Kmol −1 , which approached the desired value of 1.125 cm 3 Kmol −1 for two uncoupled Cu II ions and one radical. On cooling, the χMT value of compound 2 increased steadily to 4.66 cm 3 Kmol −1 at 2 K, showing that ferromagnetic coupling dominated the 2p-3d magnetic system.  For the magnetic system of compound 2, a negligible magnetic Cu II -NIT exchange through the two pyridyl rings was anticipated. Based on the above analysis, two effective exchange pathways should be considered: (i) the directly axial coordinated NO-Cu exchange (J 1 ) and (ii) the Cu-NO coupling via pyridine heterocycle (J 2 ). Accordingly, from a magnetic point of view, this 2D complex could be considered as the radicals bridged 1D chains plus the uncoupled mononuclear copper(II) units. There is no available magnetic expression for such a 1D system. Thus, the MAGPACK procedure was employed to investigate the dc magnetic susceptibilities involving a closed cycle containing two [Cu-{Cu 2 NIT 2 }] units (Scheme 3) with two additional uncoupled Cu II ions with the Hamiltonian. For the magnetic system of compound 2, a negligible magnetic Cu II -NIT exchange through the two pyridyl rings was anticipated. Based on the above analysis, two effective exchange pathways should be considered: (i) the directly axial coordinated NO-Cu exchange (J1) and (ii) the Cu-NO coupling via pyridine heterocycle (J2). Accordingly, from a magnetic point of view, this 2D complex could be considered as the radicals bridged 1D chains plus the uncoupled mononuclear copper(II) units. There is no available magnetic expression for such a 1D system. Thus, the MAGPACK procedure was employed to investigate the dc magnetic susceptibilities involving a closed cycle containing two [Cu-{Cu2NIT2}] units (Scheme 3) with two additional uncoupled Cu II ions with the Hamiltonian. The best-fitting result gave J1 = 32.8 cm −1 , J2 = 2.2 cm −1 , and g = 2.16. The ferromagnetic interaction (J1) as a result of the NO group was axially bound to the Cu II ion [49,50]. Meanwhile, the ferromagnetic Cu II -rad coupling through one pyridine ring (J2) was explained in terms of a spin-polarization model (Scheme S1), which is comparable to other Cu-rad compounds [19,52].
The magnetization value was 3.24 Nβ at 2 K for 70 kOe, which is in agreement with the desired result of 3 Nβ. Beyond that, the experimental magnetization (M) was larger than the theoretical curve involving the Brillouin functions with one S = 1/2 and two S = 1/2 uncoupled spins, indicating the existence of a ferromagnetic property.

Raw Materials and Physical Investigation
Multi-dentate radicals bisNITPhPy [31] and NIT-3Py-5-4Py [53] were prepared according to the related literature. Elemental analyses of both 2D Cu complexes were implemented by means of Perkin-Elmer elemental analytical equipment. A Bruker TENOR 27 spectrograph was used to measure the Fourier infrared spectra. The magnetic data of both compounds were corrected using Quantum Design SQUID VS and diamagnetic corrections of both Cu compounds were carried out using Pascal's constants.
The magnetization value was 3.24 Nβ at 2 K for 70 kOe, which is in agreement with the desired result of 3 Nβ. Beyond that, the experimental magnetization (M) was larger than the theoretical curve involving the Brillouin functions with one S = 1/2 and two S = 1/2 uncoupled spins, indicating the existence of a ferromagnetic property.

Raw Materials and Physical Investigation
Multi-dentate radicals bisNITPhPy [31] and NIT-3Py-5-4Py [53] were prepared according to the related literature. Elemental analyses of both 2D Cu complexes were implemented by means of Perkin-Elmer elemental analytical equipment. A Bruker TENOR 27 spectrograph was used to measure the Fourier infrared spectra. The magnetic data of both compounds were corrected using Quantum Design SQUID VS and diamagnetic corrections of both Cu compounds were carried out using Pascal's constants.

Crystallographic Analysis
Structural data of both Cu II networks were gathered via a Rigaku Saturn diffractometer (Mo-Kα source). Utilizing SADABS, the empirical absorption correction of two Cu II compounds was carried out. The structures of both compounds were parsed via direct methods and reasonably refined via least-squares with SHELXL-2014 on F 2 [54,55]. All H atoms were attached to suitable positions. For 2, there was a large cavity of disordered solvent units, which was evaluated using the SQUEEZE [56] option. Anisotropic parameters were applied to all non-H atoms.

Conclusions
In summary, two novel rad-Cu heterospin 2D networks with different spin topologies were acquired by using multi-dentate nitronyl nitroxides. The magnetic behavior of compound 1 was mainly derived from mono-radical bridged Cu II 1D chains and [Cu-NIT-Cu-NIT-Cu] structural units. For compound 2, the ferromagnetic interaction dominated the magnetic system, originating from 1D loop chains and uncoupled Cu II ions. Our work not only provides intriguing 2D nitronyl nitroxide-Cu compounds but also promises a new strategy for constructing two-dimensional magnetic materials through polydentate nitronyl nitroxide. This work shows that multi-dentate nitronyl nitroxide can act as fine linkers for facilitating an improvement in dimensionality to obtain a high-dimensional heterospin complex with fascinating magnetic properties.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/magnetochemistry7050073/s1. Tables S1 and S2: Vital bond lengths and corresponding angles for two Cu complexes; Tables S3 and S4: Structural analysis for Cu II ions in two compounds; Figures S1 and S3: The coordination polyhedron of Cu(II) ions in both Cu II compounds; Figures S2 and S4: Molecular packing arrangement for both Cu II complexes; Scheme S1: Spin polarization model of the Cu II -rad interaction in compound 2; CCDC 2079603 and 2079604 involve crystallographic data of Cu compounds, where these data can be downloaded from www.ccdc.cam.ac.uk/data_request/cif (accessed on 23 April 2021).

Conflicts of Interest:
The authors declare no conflict of interest.