[MII(H2dapsc)]-[Cr(CN)6] (M = Mn, Co) Chain and Trimer Complexes: Synthesis, Crystal Structure, Non-Covalent Interactions and Magnetic Properties

Four new heterometallic complexes combining [MII(H2dapsc)]2+ cations with the chelating H2dapsc {2,6-diacetylpyridine-bis(semicarbazone)} Schiff base ligand and [Cr(CN)6]3− anion were synthesized: {[MII(H2dapsc)]CrIII(CN)6K(H2O)2.5(EtOH)0.5}n·1.2n(H2O), M = Mn (1) and Co (2), {[Mn(H2dapsc)]2Cr(CN)6(H2O)2}Cl·H2O (3) and {[Co(H2dapsc)]2Cr(CN)6(H2O)2}Cl·2EtOH·3H2O (4). In all the compounds, M(II) centers are seven-coordinated by N3O2 atoms of H2dapsc in the equatorial plane and N or O atoms of two apical –CN/water ligands. Crystals 1 and 2 are isostructural and contain infinite negatively charged chains of alternating [MII(H2dapsc)]2+ and [CrIII(CN)6]3− units linked by CN-bridges. Compounds 3 and 4 consist of centrosymmetric positively charged trimers in which two [MII(H2dapsc)]2+ cations are bound through one [CrIII(CN)6]3− anion. All structures are regulated by π-stacking of coplanar H2dapsc moieties as well as by an extensive net of hydrogen bonding. Adjacent chains in 1 and 2 interact also by coordination bonds via a pair of K+ ions. The compounds containing MnII (1, 3) and CoII (2, 4) show a significant difference in magnetic properties. The ac magnetic measurements revealed that complexes 1 and 3 behave as a spin glass and a field-induced single-molecule magnet, respectively, while 2 and 4 do not exhibit slow magnetic relaxation in zero and non-zero dc fields. The relationship between magnetic properties and non-covalent interactions in the structures 1–4 was traced.


Introduction
Molecular nanomagnets, the so-called single-molecular magnets (SMMs), single-chain magnets (SCMs) and single-ion magnets (SIMs) are attracting much attention due to their unique magnetic properties, such as superparamagnetism, relaxation, blocking and quantum tunneling magnetization [1][2][3], as well as the prospects for their application in information technology, spintronics and quantum calculations [4][5][6]. However, at present, the practical use of molecular nanomagnets is difficult, since they have low critical magnetization blocking temperatures (T B ), below which their unique magnetic properties appear. Although, in recent times, SIMs, mononuclear metallocene complexes of Dy, with blocking temperatures close to liquid nitrogen temperature have been synthesized [7,8], their application is unlikely since these compounds are extremely air-sensitive and have short magnetization relaxation times. For instance, for SIM ([(Cp iPr5 )Dy(Cp*)] + with T B = 80 K, long relaxation times (τ > 100 s) occur only below~30 K [8]. In 2021, Tang et al. reported an air-stable hexagonal bipyramidal Dy(III) SMM, [Dy(L N6 )(Ph 3 SiO) 2 ][PF 6 ], which displayed a record anisotropy barrier (U eff ) exceeding 1800 K (T B = 20 K) and the longest relaxation time approaching 2500 s at 2.0 K for all known air-stable SIMs [9]. However, it is difficult to [8]. In 2021, Tang et al. reported an air-stable hexagonal bipyramidal Dy(III) SMM, [Dy(L N6 )(Ph3SiO)2][PF6], which displayed a record anisotropy barrier (Ueff) exceeding 1800 K (TB = 20 K) and the longest relaxation time approaching 2500 s at 2.0 K for all known air-stable SIMs [9]. However, it is difficult to expect that sufficiently long magnetization relaxation times can be achieved with mononuclear complexes. Recently, a new concept for the design of advanced molecular nanomagnets has been proposed based on the use of low-spin (S = 1/2) pentagonal-bipyramidal complexes of 4d 3 and 5d 3 metals as building blocks for the synthesis of M(4d/5d)-M(3d) exchange-coupled pairs with Ising-type exchange interactions [10].
In this work, we investigated the reactions of complexes [Mn(H2dapsc) The presence of labile ligands in the axial positions of the PBP complexes makes them attractive building blocks for the construction of low-dimensional heteronuclear compounds by replacing these ligands with bridging cyanometallates or the PBP complexes with cyanide apical groups [17,23,24,37,[49][50][51][52][53][54][55][56]. Up to date, tri-, penta-, and decanuclear SMMs and coordination polymeric SCMs have been obtained along the way.
Earlier  6 ] and (Ph 4 P) 3 Cr(CN) 6 were studied. The interaction of 5 and 6 with K 3 [Cr(CN) 6 ] in ethanol/water led to the formation of the chain polymeric complexes 1 and 2, respectively. The use of (Ph 4 P) 3 Cr(CN) 3 6 ] moieties, two K + ions, one EtOH solvent and eight water molecules, all in general positions. Two water molecules (O8w, O9w) have 0.7 occupancy; one water (O4w) is disordered between two sites with an occupancy ratio of 0.8/0.2 in 1 and 0.6/0.4 in 2. An ORTEP drawing of 1 is shown in Figure 1, and the key bond distances and angles in 1 and 2 are listed in Table A1 of the Appendix A section.

Synthesis and Crystal Structure
The reactions of [Mn(H2dapsc) (2). The complexes 1 and 2 are isostructural and crystallize in the monoclinic P21 space group. The asymmetric unit includes two [M II (H2dapsc)], two [Cr(CN)6] moieties, two K + ions, one EtOH solvent and eight water molecules, all in general positions. Two water molecules (O8w, O9w) have 0.7 occupancy; one water (O4w) is disordered between two sites with an occupancy ratio of 0.8/0.2 in 1 and 0.6/0.4 in 2. An ORTEP drawing of 1 is shown in Figure 1, and the key bond distances and angles in 1 and 2 are listed in Table A1 of the Appendix A section.
N atoms from H2dapsc in the equatorial plane and two axial N atoms from CN-bridges. The M-O,N bond lengths in the [M II (H2dapsc)] moieties are shorter in the Co(II) compound 2 compared with the Mn(II) compound 1, the maximal difference ~0.1 Å being observed for M-N bonds (Table A1). Both independent [M II (H2dapsc)] fragments in 1 and 2 are almost flat: the dihedral angle between the two semicarbazone planes of the H2dapsc ligand defined by seven non-metallic atoms of two pentagonal cycles [for example, O1, C1, N3, N5, C4, C5, N7 and O2, C2, N4, N6, C10, C9, N7] is 1.48(4)° for Mn(1), 1.22(4)° for Mn(2), 3.05(5)° for Co(1) and 2.58(4)° for Co (2). The Cr-CCN distances in the anionic units are in the range of 2.06-2.08 Å. The ethanol molecule and five of the eight independent water molecules O(2w)-O(6w) also belong to the coordination sphere of K + ions; the corresponding K-O distances are 2.73-2.84 Å. Two K + cations linked to [M(H2dapsc)] from adjacent chains are at the K-K distance of 4.154(1) and 4.165(1) Å in 1 and 2, respectively, and bridged together through two oxygen atoms from one water and one ethanol molecule. The coordination bonds via the pairs of K + ions join adjacent infinite chains in the (1 0 -1) plane ( Figure 2). Additionally, the chains are fixed together by non-covalent interactions: by π-π stacking of the nearest H2dapsc ligands as well as by hydrogen bonds N−HH2dapsc···Owater, O−Hwater···Nanion in the (1 0 -1) plane and N−HH2dapsc...Nanion in the (0 1 0) plane (Figures 2 and S1 in the Supplementary Materials, hydrogen bond geometry is given in Tables S1 and S2).
{[Mn(H2dapsc)]2Cr(CN)6(H2O)2}Cl·H2O (3). Complex 3 crystallizes in the monoclinic P21/c space group. The asymmetric unit includes half of the formula unit with Cr(1) atom in the inversion center; a half-occupied Cl − anion and solvate water are mixed in the close sites. An ORTEP drawing of 3 is shown in Figure 4, and the key bond distances and angles are listed in Table A1. . Complex 3 crystallizes in the monoclinic P2 1 /c space group. The asymmetric unit includes half of the formula unit with Cr(1) atom in the inversion center; a half-occupied Cl − anion and solvate water are mixed in the close sites. An ORTEP drawing of 3 is shown in Figure 4, and the key bond distances and angles are listed in Table A1. The non-covalent intermolecular interactions are well represented in the structure 3. The trimers are connected with each other and with the anion/water site by hydrogen bonding (red dashed lines in Figure 5 and Table S3 for details). The shortest intermetallic distances are found in the ac plane.   Table S3 for hydrogen bond geometry).
The non-covalent intermolecular interactions are well represented in the structure 3. The trimers are connected with each other and with the anion/water site by hydrogen bonding (red dashed lines in Figure 5 and Table S3 for Table S3 for hydrogen bond geometry).
The non-covalent intermolecular interactions are well represented in the structure 3. The trimers are connected with each other and with the anion/water site by hydrogen bonding (red dashed lines in Figure 5 and Table S3 for details). The shortest intermetallic distances are found in the ac plane. The adjacent [Mn(H2dapsc)] units with the smallest Mn-Mn distance, 7.2853(5) Å, interact through a pair of strong N-H…O bonds formed between -NH2 function and oxygen atoms of H2dapsc ligands. The nearby Mn-Cr dis-  Table S3 for hydrogen bond geometry).
. Complex 4 crystallizes in the triclinic P1 space group. The asymmetric unit includes half of the formula unit with the Cr(1) atom in the inversion center; the half-occupied Cl − and water are mixed in the close sites. An ORTEP drawing of 4 is shown in Figure 6, and the key bond distances and angles are listed in Table A1.
Structure 4 contains centrosymmetric positively charged trimers like structure 3, although these compounds are not isostructural. The Co-Cr-Co trimer in 4 is more compact than the Mn-Cr-Mn trimer in 3 due to the smaller ionic radius of Co(II) (0.72 Å) in comparison with Mn(II) (0.82 Å) and the stronger bending of the Co chain (the Co-N-C angle is 151.3(4) • ). The Co-Cr distance in the trimer is 5.1281(9) Å. The dihedral angle between the two semicarbazone planes in the H 2 dapsc ligand is 1.2(1) • . teraction is 7.3022(5) Å. The π-π stacking in pairs of Mn(H2dapsc) moieties from adjacent trimers is also observed (Mn-Mn is 7.5018(5) Å, Figure 5).
{[Co(H2dapsc)]2Cr(CN)6(H2O)2}Cl·2EtOH·3H2O (4). Complex 4 crystallizes in the triclinic P1 � space group. The asymmetric unit includes half of the formula unit with the Cr(1) atom in the inversion center; the half-occupied Cl − and water are mixed in the close sites. An ORTEP drawing of 4 is shown in Figure 6, and the key bond distances and angles are listed in Table A1.  The shortest intermetallic distances are observed for the non-covalently interacting trimers in the ab plane (Figure 7). A pair of strong O-HH2O…OH2dapsc hydrogen bonds (red dashed lines in Figure 7 and Table S4 for details) are observed between adjacent trimers along the [1 −1 0] direction whose terminal water molecules are close to each other. The Co-Co distance in this contact, 5.214(1) Å, is very short and only slightly larger than the    Table S4 for hydrogen bond geometry).

Static (dc) Magnetic Properties
Variable-temperature magnetic measurements were performed on polycrystalline samples of complexes 1-4 in the range of 1.8-300 K under a dc field of 1000 Oe. The plots of temperature dependencies of the magnetic susceptibility (χM) and effective magnetic moment (μeff) for the isostructural chain complexes 1 and 2 are depicted in Figure 8a Table S4 for hydrogen bond geometry).

Static (dc) Magnetic Properties
Variable-temperature magnetic measurements were performed on polycrystalline samples of complexes 1-4 in the range of 1.8-300 K under a dc field of 1000 Oe. The plots of temperature dependencies of the magnetic susceptibility (χ M ) and effective magnetic moment (µ eff ) for the isostructural chain complexes 1 and 2 are depicted in Figure 8a,b. Variable-temperature magnetic measurements were performed on polycrystalline samples of complexes 1-4 in the range of 1.8-300 K under a dc field of 1000 Oe. The plots of temperature dependencies of the magnetic susceptibility (χM) and effective magnetic moment (μeff) for the isostructural chain complexes 1 and 2 are depicted in Figure 8a The µ eff value at 300 K for 1 is close to the expected value (7.06 µ B ) for non-interacting Mn(II) and Cr(III) ions with S = 5/2 (g = 2.0) and 3/2 (g = 2.0), respectively. For 1, at cooling, the µ eff value gradually decreases down to 50 K, and then, it decreases rapidly to the minimum value at ∼20 K. Upon further lowering the temperature, the µ eff increases abruptly to reach a maximum at 3 K, followed by a slight drop to 1.8 K. The latter feature is probably attributed to the presence of antiferromagnetic (AF) coupling between the chains through the hydrogen bonding network, π−π stacking and the interchain coordination bonds via solvent molecules and K + ions. The general behavior of the µ eff with temperature suggests AF exchange interactions between Mn(II) and Cr(III) spin carriers within the chains, resulting in a ferrimagnetic spin arrangement along the chains. The Curie-Weiss fit of the magnetic susceptibility data above 75 K gave a Weiss constant of −75.0 K ( Figure S2, Table S5). The large negative Weiss constant confirms the dominance of AF interactions in the chains of complex 1. The rapid growth of µ eff below 20 K indicates a ferrimagnetic ordering in the chains of 1.
The magnetic properties of complex 2 differ significantly from those of 1. The µ eff value at room temperature is higher than the spin-only value (5.47 µ B ) for one Co(II) of S = 3/2 (g = 2.0) and one Cr(III) of S = 3/2 (g = 2.0) due to the orbital contributions of Co(II) ions, Figure 8b. As the temperature decreases, the effective magnetic moment does not decrease, as in the case of 1, but, on the contrary, increases to a sharp maximum around 6.0 K, indicating an intrachain ferromagnetic coupling between the Co(II) and Cr(III) ions, which is expected for pseudolinear Co(II)-NC-Cr(III) bridges [49,51,57,58]. The Weiss constant is positive and amounts to 7.0 K ( Figure S2, Table S5), which confirms the presence of ferromagnetic interactions in the chains of complex 2. X-ray diffraction analysis of 1 and 2 showed that the Co-Cr chains are more compact; the Co-Cr distances are~0.07 Å shorter than the Mn-Cr distances. Below 6 K, the value of the effective magnetic moment for 2 shows a sharp and deep drop to 1.8 K, which is probably due to the presence of intrinsic magnetic anisotropy of the Co(II) centers [59] and/or antiferromagnetic ordering associated with coupling between chains through non-covalent interactions. The χ M vs. T behavior also exhibits a maximum (Figure 8b), suggesting antiferromagnetic ordering. The interchain AF interactions in structure 2 are much stronger than in 1. The shortest interchain distance between Co(1) and Co(2) centers linked through π-π-stacking of H 2 dapsc ligands is 0.15 Å shorter than the Mn(1)-Mn(2) distance in 1. The study of the field-dependent magnetization at 2 K indicates a metamagnetic-like transition from an AF state to a state of spontaneous magnetization under the field H c = 1.5 kOe, Figure 9a. The dM/dH(H) dependence exhibits distinct peaks in the fields 1.5 kOe and −1/5 kOe, Figure 9b. The metamagnetic nature of the transition is also evidenced by the character of the magnetization curves M(T) and the curves χ mol(T) recorded at different magnetic fields ( Figure S3) as well as the data of ac magnetic susceptibility. The field dependence of the real part (χ') of ac susceptibility shows a maximum of χ' at a field of 1.5 kOe ( Figure S4). The application of an external constant magnetic field leads to the suppression of interchain AF coupling and causes a metamagnetic transition. Such transitions have been observed in other heterospin as well as homospin compounds with interchain AF ordering [52,[60][61][62][63][64][65][66][67]. Unlike 1 and 2, the trinuclear complexes 3 and 4 are not isostructural. Their magnetic properties differ markedly and, in some respects, are close to those of the corresponding chain complexes. The temperature dependencies of the magnetic susceptibility (χM) and effective magnetic moment (μeff) for the Mn2Cr trimer 3 are depicted in Figure 8c. The μeff value at 300 K is close to the expected value (9.2 μB) for two non-interacting Mn(II) and one Cr(III) ions. Upon cooling from room temperature, the μeff value decreases to approximately 16 K, as in the case of complex 1, indicating antiferromagnetic interaction between the metallic centers, as was observed in the similar trimer [68]. On further cooling, the μeff increases to 1.8 K. Unlike 1, this increase is not accompanied by a subsequent decrease in μeff associated with AF interactions between the chains in structure 1. Theoretical analysis of the antiferromagnetic interactions between the Mn II (H2dapsc) units belonging to the neighboring chains, which contact through π-π stacking of planar H2dapsc ligands and a system of hydrogen bonds in the complex {[Mn(H2dapsc)][Fe(CN)6][K(H2O)3.5]}n·1.5nH2O, showed that contacts through π-π stacking play a decisive role in these superexchange AF interactions [52]. In the structure 1, along the a-axis, there are infinite chains of ...-Mn1-Mn2-Mn1-Mn2-... through π-π stacking of the H2dapsc ligands with distances between metals of 8.2341(6) and 8.2357(6) Å (Figure 3), whereas in 3, such contacts are present only in pairs ( Figure 5). Probably for this reason, interchain AF interactions in 3 are very weak compared to those in 1.
Temperature behavior of μeff for the Co2Cr trimer (4) (Figure 8d) is close to that for the chain complex (-Cr-CN-Co-NC-)n (2), Figure 8b. Upon cooling, the μeff value increases, reaching a maximum near 7.8 K, which indicates a ferromagnetic exchange between the metal centers in the trimer, and then decreases. The Weiss constant is positive and equal to 27.0 K ( Figure S2, Table S5). It is possible that the strong ferromagnetic interaction in 4 Unlike 1 and 2, the trinuclear complexes 3 and 4 are not isostructural. Their magnetic properties differ markedly and, in some respects, are close to those of the corresponding chain complexes. The temperature dependencies of the magnetic susceptibility (χ M ) and effective magnetic moment (µ eff ) for the Mn 2 Cr trimer 3 are depicted in Figure 8c. The µ eff value at 300 K is close to the expected value (9.2 µ B ) for two non-interacting Mn(II) and one Cr(III) ions. Upon cooling from room temperature, the µ eff value decreases to approximately 16 K, as in the case of complex 1, indicating antiferromagnetic interaction between the metallic centers, as was observed in the similar trimer [68]. On further cooling, the µ eff increases to 1.8 K. Unlike 1, this increase is not accompanied by a subsequent decrease in µ eff associated with AF interactions between the chains in structure 1. Theoretical analysis of the antiferromagnetic interactions between the Mn II (H 2 dapsc) units belonging to the neighboring chains, which contact through π-π stacking of planar H 2 dapsc ligands and a system of hydrogen bonds in the complex {[Mn(H 2 dapsc)][Fe(CN) 6 ][K(H 2 O) 3.5 ]} n ·1.5nH 2 O, showed that contacts through π-π stacking play a decisive role in these superexchange AF interactions [52]. In the structure 1, along the a-axis, there are infinite chains of ...-Mn1-Mn2-Mn1-Mn2-... through π-π stacking of the H 2 dapsc ligands with distances between metals of 8.2341(6) and 8.2357(6) Å (Figure 3), whereas in 3, such contacts are present only in pairs ( Figure 5). Probably for this reason, interchain AF interactions in 3 are very weak compared to those in 1.
Temperature behavior of µ eff for the Co 2 Cr trimer (4) (Figure 8d) is close to that for the chain complex (-Cr-CN-Co-NC-) n (2), Figure 8b. Upon cooling, the µ eff value increases, reaching a maximum near 7.8 K, which indicates a ferromagnetic exchange between the metal centers in the trimer, and then decreases. The Weiss constant is positive and equal to 27.0 K ( Figure S2, Table S5). It is possible that the strong ferromagnetic interaction in 4 is due to the fact that trimers in its structure are joined into infinite chains [-Co(H 2 dapsc)-(2H 2 O)-Co(H 2 dapsc)-Cr(CN) 6 -] n through the pairs of strong intertrimer hydrogen bonds O-H H2O . . . O H2daps (Figure 7 and Table S4). The metal-metal separations along the formed chain are rather uniform: the Co-Cr distance inside the trimer is 5.128 Å, whereas the Co-Co distance between the trimers is only slightly longer, 5.214 Å. The drop of µ eff to 1.8 K is not as sharp and deep as in the case of complex 2 (Figure 8b,d). This is probably due to the weaker interchain AF interactions in the Co-Co pairs coupled via π-π stacking of the H 2 dapsc ligands in complex 4 (Figure 7) compared to such interactions in infinite Co-Co chains of π-stacked complexes in 2 (similar to the Mn-Mn chains in 1 shown in Figure 3).

Dynamic (ac) Magnetic Properties
In order to examine possible SCM or SMM properties of the complexes 1, 2 and 3, 4, respectively, the in-phase χ' and out-of-phase χ" components of ac magnetic susceptibility were measured in the frequency range of 0.1 Hz-1000 Hz at temperatures 1.8 K-5.0 K in a zero and in non-zero magnetic field. Chain (2) and trimeric (4) complexes, in which the Co(II) and Cr(III) metal centers are linked by cyanide bridges, did not show a frequency dependence of χ" in zero and non-zero static magnetic fields, thus precluding the single chain or single molecule magnetic behavior for 2 and 4, respectively ( Figures S5-S7). This result agrees with the statement made in [23] that seven-coordinated Co(II) PBP complexes with planar anisotropy are not suitable as building blocks for the creation of exchangecoupled polynuclear ensembles demonstrating slow magnetization relaxation, in contrast, for example, to PBP complexes of Fe(II) and Ni(II) with Ising anisotropy.
Unlike 2 and 4, complexes 1 and 3 which contain Mn(II) instead of Co(II) exhibit frequency-dependent signals χ' and χ". The plots of temperature and frequency dependencies of χ' and χ" for 1 in a zero dc field are shown in Figure 10 and Figure S8. Below 4.7 K, an out-of-phase signal reveals a frequency-dependent maximum that indicates a slow relaxation of magnetization. However, the relative change of the temperature of the χ" maximum depending on the frequency of the oscillating field, expressed by the so-called Mydosh parameter φ = (∆Tp/Tp)/∆(logf) [69], turned out to be 0.023, which is in the range of values characteristic of spin glass (0.004 > φ < 0.08) and not superparamagnets for which the Mydosh parameter is usually an order of magnitude larger (>0.1) [70,71]. The relaxation time (τ) was fitted to the Arrhenius equation, τ = τ 0 exp(U eff /k B T), where τ = 1/2πν, to allow the estimation of a pre-exponential factor, τ 0 = 6.5·10 −55 s, and the effective barrier to the relaxation of magnetization, U eff = 512 K ( Figure S9). These values are outside the normal range for typical SCMs (τ 0 usually >10 −13 s) and indicate that 1 shows typical spin-glass behavior. Similar behavior was found in other 1D chain complexes [72][73][74][75][76]. Antiferromagnetic interactions between different structural units can lead to randomness (disorder, defects) and frustration, which are responsible for the spin-glass system [73,74,76,77]. In 1, the antiferromagnetic interaction between the chains, transferred through π-π stacking and hydrogen bonds, disrupts the ferrimagnetic ordering in the chains and leads to the disorder of the electron spin system, which causes the spin-glass behavior of complex 1. Moreover, the loss of the coordination solvent (EtOH), which is involved in the formation of hydrogen bonds in 1, may contribute to the formation of the spin-glass state, as was observed in the works [73,77].
In the case of trimeric complex 3, the studies of ac susceptibility showed frequencydependent signals of χ" when applying a dc field of 1500 Oe (Figure 11), indicating a slow relaxation of the magnetization. The pronounced χ" maxima appear one after another in the frequency range above 100 Hz and shift towards higher frequencies with increasing temperature, which is typical for SMMs. The Mydosh parameter is 0.9. The temperature dependence of the χ" peak frequency follows an Arrhenius law with an effective energy barrier U eff = 4.1 K and a pre-exponential factor τ 0 being 4.2·10 −6 s, suggesting a thermally activated relaxation ( Figure 12). Thus, complex 3 is a field-induced SMM.
barrier to the relaxation of magnetization, Ueff = 512 K ( Figure S9). These values are outside the normal range for typical SCMs (τ0 usually >10 −13 s) and indicate that 1 shows typical spin-glass behavior. Similar behavior was found in other 1D chain complexes [72][73][74][75][76]. Antiferromagnetic interactions between different structural units can lead to randomness (disorder, defects) and frustration, which are responsible for the spin-glass system [73,74,76,77]. In 1, the antiferromagnetic interaction between the chains, transferred through π-π stacking and hydrogen bonds, disrupts the ferrimagnetic ordering in the chains and leads to the disorder of the electron spin system, which causes the spin-glass behavior of complex 1. Moreover, the loss of the coordination solvent (EtOH), which is involved in the formation of hydrogen bonds in 1, may contribute to the formation of the spin-glass state, as was observed in the works [73,77] In the case of trimeric complex 3, the studies of ac susceptibility showed frequency-dependent signals of χ″ when applying a dc field of 1500 Oe (Figure 11), indicating a slow relaxation of the magnetization. The pronounced χ″ maxima appear one after another in the frequency range above 100 Hz and shift towards higher frequencies with increasing temperature, which is typical for SMMs. The Mydosh parameter is 0.9. The temperature dependence of the χ″ peak frequency follows an Arrhenius law with an effective energy barrier Ueff = 4.1 K and a pre-exponential factor τ0 being 4.2·10 −6 s, suggesting a thermally activated relaxation ( Figure 12). Thus, complex 3 is a field-induced SMM.  Figure 12. The dependence of lnτ vs. 1/T for 3 in 1500 Oe dc field and its approximation using the Arrhenius formula (solid line). The points were obtained from the frequency dependencies χ″(ν) at different temperatures ( Figure 11, main text). Approximation parameters: activation energy Ueff = 4.1 K, a pre-exponential factor τ0 = 4.2·10 −6 s.

Materials and Methods
All chemicals and solvents were reagent grade and used without further purification.
The The C, H, N, O elemental analyses were carried out with a Vario Micro Cube analyzing device. The infrared spectra were measured on solid samples using a VERTEX 70v (Bruker) spectrometer in the range of 4000-500 cm −1 . The thermogravimetric analysis  Figure 12. The dependence of lnτ vs. 1/T for 3 in 1500 Oe dc field and its approximation using the Arrhenius formula (solid line). The points were obtained from the frequency dependencies χ″(ν) at different temperatures ( Figure 11, main text). Approximation parameters: activation energy Ueff = 4.1 K, a pre-exponential factor τ0 = 4.2·10 −6 s.

Materials and Methods
All chemicals and solvents were reagent grade and used without further purification.
The The C, H, N, O elemental analyses were carried out with a Vario Micro Cube analyzing device. The infrared spectra were measured on solid samples using a VERTEX 70v Figure 12. The dependence of lnτ vs. 1/T for 3 in 1500 Oe dc field and its approximation using the Arrhenius formula (solid line). The points were obtained from the frequency dependencies χ"(ν) at different temperatures ( Figure 11, main text). Approximation parameters: activation energy U eff = 4.1 K, a pre-exponential factor τ 0 = 4.2·10 −6 s.

Materials and Methods
All chemicals and solvents were reagent grade and used without further purification. The starting compounds [Mn(H 2 6 ] and PPh 4 Cl. The C, H, N, O elemental analyses were carried out with a Vario Micro Cube analyzing device. The infrared spectra were measured on solid samples using a VERTEX 70v (Bruker) spectrometer in the range of 4000-500 cm −1 . The thermogravimetric analysis (TGA) was performed in an argon atmosphere with a heating rate of 5.0 • C min −1 using a NETZSCH STA 409 C Luxx thermal analyzer. X-ray powder diffraction spectra were recorded using a Siemens D500 powder diffractometer with a linear detector at room temperature (CuKα1radiation, λ = 1.5406 Å, step = 0.02 • , single crystal sample holder). The dc and ac magnetic susceptibility of powder samples of complexes 1-4 were measured using a Quantum Design MPMS-5 SQUID magnetometer. The experimental data were corrected for the sample holder and for the diamagnetic contribution calculated from Pascal constants.  Figure S13).
The thermal analysis of complex 2 after drying in vacuum showed that a mass loss begins at~50 • C with an endothermic peak at 112.7 • C and reaches 9.5% at 150 • C, that corresponds to the loss of 3.5 molecules H 2 O ( Figure S12). In the mass spectrum of the gas phase, the peaks are observed at m/z = 18 and m/z = 17 from H 2 O molecules. The peaks from ethanol molecules are absent. The second weight-loss step appears above 200 • C with the release of fragments of the decaying complex.
For the single crystal X-ray diffraction analysis, crystals of 1, 2 were kept in contact with mother liquid to prevent an ethanol loss. After drying, the complexes preserve crystallinity. X-ray powder diffraction pattern of the dried sample of 2 shows a coincidence with the simulated pattern calculated from the single crystal data ( Figure S14).

Crystal Structure Determination
X-ray single crystal diffraction data were collected at low temperatures on an Oxford Diffraction Gemini-R CCD diffractometer equipped with an Oxford cryostream cooler [λ(MoK α ) = 0.71073 Å, graphite monochromator, ω-scans]. Single crystals of the complexes 1-4 were taken from the mother liquid using a nylon loop with oil and immediately transferred into the cold nitrogen stream of the diffractometer. Data reduction with empirical absorption correction of experimental intensities (Scale3AbsPack program) was made with the CrysAlisPro software [79].
The structures were solved by a direct method and refined by a full-matrix least squares method using SHELX-2016 program [80]. All non-hydrogen atoms were refined anisotropically. The positions of H-atoms were refined in a riding model with isotropic displacement parameters depending on the U eq of the connected atom. The hydrogen atoms in water molecules were found from the difference Fourier map. All N-H and O-H bond distances were refined, with additional geometrical restraints (SADI/DFIX) applied in some cases. Acentric structures 1 and 2 were refined as two-component inversion twins, with refined twin fractions of 0.50(1) and 0.48(1), respectively. Main crystal and experimental data for 1-4 are listed in Table 1.   6 ] bridging units. The pentadentate H 2 dapsc ligands in all four compounds keep the protons in both hydrazine -NH moieties and retain a neutral state. Structural analysis revealed the important role of intermolecular non-covalent interactions such as hydrogen bonding and π . . . π stacking in the stabilization of crystal packing. In the chain compounds 1 and 2, neighbor chains are linked additionally by coordination bonds via pairs of potassium cations and oxygen atoms from water/ethanol molecules. Pairs of compounds 1, 3 and 2, 4 containing Mn(II) and Co(II), respectively, differ significantly in their dc and ac magnetic properties. In contrast to 1, in compound 2, a field-induced metamagnetic transition with a critical field of 1500 Oe was found. Complexes 1 and 3 exhibit a frequency-dependent ac magnetic susceptibility due to the spin-glass behavior in the chain complex 1 and the field-induced single molecule magnetism in the trimeric complex 3. At the same time, 2 and 4 did not show a frequency dependence of χ" signals of ac susceptibility in the zero and non-zero dc field, thus precluding the single chain or single molecule magnetic behavior for 2 and 4, respectively. This result implies that seven-coordinated Co(II) PBP complexes with planar anisotropy are not suitable as building units for the design of advanced molecular nanomagnets based on the M(4d/5d)-M(3d) exchange-coupled pairs with Ising-type exchange interactions [10,81]. At the same time, Mn(II) PBP complexes with high spin (S = 5/2) can be used for these purposes.

Conclusions
Supplementary Materials: The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/molecules27238518/s1, Cif and CheckCif files; Figure S1: Interchain hydrogen bonds in the (0 1 0) plane of the structure 1; Figure S2: Temperature dependencies of inverted magnetic susceptibility for 1-4; Figure S3: Temperature dependencies of the magnetization (a) and dc magnetic susceptibility (b) for 2 at different magnetic fields; Figure S4: Magnetic dc field dependencies of real (χ') and imaginary (χ") parts of ac magnetic susceptibility at 2.0 K for complex 2; Figure S5: Frequency dependencies of ac magnetic susceptibility at 2.0, 2.5 and 3.0 K for complex 2 in a zero dc field; Figure S6: Frequency dependencies of ac magnetic susceptibility at 2.0-3.5 K temperature range for complex 2 under 1.5 kOe dc field; Figure S7: Frequency dependencies of ac magnetic susceptibility at 1.8-5.0 K temperature range for complex 4 under 1.5 kOe dc field; Figure  S8: Frequency dependencies of χ' and χ" at different temperatures for complex 1 in a zero dc field; Figure S9: The dependence of ln(τ) vs. T −1 for 1 in a zero dc field; Figure S10: TG-DSC curves and mass spectra for complex 1 after drying in vacuum; Figure S11: IR (a) and Raman (b) spectra of the complex 1; Figure S12: TG-DSC curves and mass spectra for complex 2 after drying in vacuum; Figure S13: IR spectrum of the complex 2; Figure S14: Powder diffraction pattern for a dried sample of 2; Figure S15: IR spectrum of the complex 3; Figure S16: IR spectrum of the complex 4; Tables S1, S2, S3, S4: Hydrogen bond geometry in 1, 2, 3, 4, respectively; Table S5: Weiss temperatures in the complexes 1-4.