Magnetic Nature of the CrIII-LnIII Interactions in [CrIII2LnIII3] Clusters with Slow Magnetic Relaxation.

Two 3d-4f hetero-metal pentanuclear complexes with the formula {[CrIII2LnIII3L10(OH)6(H2O)2]Et3NH} [Ln=Tb (1), Dy (2); HL=pivalic acid, Et3N=triethylamine] have been produced. The metal core of each cluster is made up of a trigonal bipyramid with three LnIII ions (plane) and two CrIII ions (above and below) held together by six μ3-OH bridges. Also reported with this series is the diamagnetic CrIII-YIII analogue (3). Fortunately, we successfully prepared AlIII-LnIII analogues with the formula {[AlIII2LnIII3L10(OH)6(H2O)2]Et3NH⋅H2O} [Ln=Tb (4), Dy (5)], containing diamagnetic AlIII ions, which can be used to evaluate the CrIII-LnIII magnetic nature through a diamagnetic substitution method. Subsequently, static (dc) magnetic susceptibility studies reveal dominant ferromagnetic interactions between CrIII and LnIII ions. Dynamic (ac) magnetic susceptibility studies show frequency-dependent out-of-phase (χ'') signals for [CrIII2TbIII3] (1), [CrIII2DyIII3] (2), and [AlIII2DyIII3] (5), which are derived from the single-ion behavior of LnIII ions and/or the CrIII-LnIII ferromagnetic interactions.

Introduction

Molecule‐based magnetic materials have become an important area in modern coordination chemistry over the past two decades; therein, single‐molecule magnets (SMMs) exhibiting slow magnetic relaxation have attracted great concern, because of future prospects in information storage techniques.1, 2, 3, 4 Indeed, SMMs are a molecular approach to nanoscale particles, which can be magnetized in an external magnetic field and they retain magnetization when the field is turned off. Such behavior is derived from the combination of the large ground‐state spin (S) and a large magnetic anisotropy of the Ising (easy‐axis) type with a negative zero‐field splitting parameter, D. This leads to an energy barrier (U eff) to magnetic relaxation with exchange interactions between the magnetic centers.5

Since the first discovery of Mn12 SMM,6, 7, 8 coordination clusters containing MnIII ions are the richest family of SMMs, owing to its favorable Ising anisotropy;9, 10, 11, 12, 13, 14, 15, 16 however, SMMs with other 3d ions have been successfully isolated, such as FeIII,17, 18, 19, 20 CoII,21, 22, 23, 24, 25 and NiII,26, 27, 28, 29, 30, 31 all of which possess anisotropic characteristics. In 2003, a TbIII complex displayed SMM behavior,32 opening the upsurge in searching for lanthanide‐based SMMs; since then, plenty of lanthanide‐based SMMs have been reported, proving that lanthanide ions are good candidates to construct SMMs, owing to their large magnetic moments and anisotropy.33, 34, 35, 36, 37, 38, 39, 40, 41

Among the many lanthanide‐based complexes, 3d–4f coordination clusters have been considered one of the most interesting molecular systems for magnetic studies.42, 43, 44, 45, 46, 47, 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60 The strategy for constructing 3d–4f coordination clusters is to build metal cores with organic ligands, which act not only as bridging ligands to bridge 3d and 4f ions, but also as peripheral ligands to separate the discrete species from each other. However, such systems are quite difficult to analyze because: i) it is well‐known that lanthanide ions have contracted f‐orbitals and the magnetic exchange interactions between them are extremely weak, which has been reported for the majority of lanthanide‐based clusters; ii) the 3d–4f exchange interactions are always weak and difficult to determine, because of the complicated magnetic interactions in this system (d–d, d–f, and f–f interactions) and the intrinsic nature of 4f ions with an orbital contribution to their ground states.61

To analyze the 3d–4f exchange interactions, diamagnetic substitution is a successful approach to address the problem, which consists of comparing the magnetic behaviors of two isostructural systems (one with a 4f ion and a second paramagnetic center; the other with a 4f ion and a diamagnetic center) to garner insight into the magnetic nature between 3d and 4f ions. The comparison of two systems will eliminate the contribution of the spin–orbit coupling of 4f ions, thus revealing the magnetic nature of the 3d–4f interactions. However, to date, the use of this approach to obtain the magnetic nature of 3d–4f interactions is rather rare, owing to the complexity in obtaining the diamagnetic analogues, and then only a few 3d–4f coordination systems have been successfully applied to diamagnetic substitution to determine the nature of magnetic exchange interaction between 3d and 4f ions.62, 63, 64, 65, 66, 67 For instance, we use the diamagnetic substitution to obtain the magnetic nature between 3d and 4f ions of the hetero‐metal MnII–LnIII family.67

On the other hand, much work has concentrated on mixing 4f ions with anisotropic 3d ions to prepare the former 3d‐based SMMs; however, some isotropic ions can also be favorable in 3d–4f coordination clusters with enhanced SMM properties.68, 69, 70, 71, 72, 73, 74, 75 Since the [CrIII 4DyIII 4] SMM was reported,68 3d–4f SMMs containing CrIII ions have drawn great attention.69, 70, 71, 72, 73, 74, 75 More recently, a rare [CrIII 2DyIII 2] SMM with a high U eff of 77 K and large coercive magnetic fields was synthesized, which can significantly reduce the quantum tunneling of magnetization (QTM) due to the exchange interaction between CrIII and DyIII ions.74 Following work showed that the U eff was related to the magnitude of exchange between magnetic centers by comparing it to two families of CrIII‐containing SMMs.75 Interestingly, incorporating a diamagnetic metal ion provides a potential avenue to improve the U eff parameter in lanthanide‐based SMMs.76, 77, 78

Inspired by these aforementioned aspects, we chose the isotropic CrIII ion as a 3d metal ion to fabricate 3d–4f clusters, aiming to obtain novel SMMs. As a result, we obtained two hetero‐metal CrIII–LnIII clusters with a [CrIII 2O6LnIII 3] core [Ln=Tb (1) and Dy (2)], showing SMM‐like behavior. To understand the magnetic interactions between CrIII and LnIII ions, the diamagnetic YIII and AlIII ions were expected to substitute LnIII and CrIII ions, respectively, to obtain analogues. Fortunately, the corresponding CrIII–YIII (3) and AlIII–LnIII (4 and 5) analogues were successfully isolated and characterized. Static (dc) magnetic studies suggest strong ferromagnetic interactions between CrIII and LnIII ions within 1 and 2. Dynamic (ac) magnetic properties show that complexes 1, 2, and 5 display slow magnetic relaxation.

Results and Discussion

Syntheses

Hetero‐metal pentanuclear 3d–4f complexes with the general formula [CrIII 2LnIII 3L10(OH)6(H2O)2]Et3NH [where Ln=Tb (1), Dy (2)] have been produced by using a solution‐based method. The reaction of CrCl3⋅6 H2O, Ln(NO3)3⋅6 H2O, and HL in a 5:3:10 molar ratio with Et3N in mixed CH3CN and CH2Cl2 (1:1) gave a light purple solution, from which well‐shaped purple crystals were obtained after several days. Fortunately, using Y(NO3)3⋅6 H2O instead of Ln(NO3)3⋅6 H2O produced the CrIII–YIII complex (3); furthermore, when replacing CrCl3⋅6 H2O with AlCl3 as the original material, the corresponding AlIII–LnIII analogues with the formula {[AlIII 2LnIII 3L10(OH)6(H2O)2]Et3NH⋅H2O} [Ln=Tb (4), Dy (5)] were obtained successfully.

Crystal Structures of Complexes 1–5

These hetero‐metal complexes are isostructural and crystallize in tetragonal space group P42/nmc. The structure of complex 1 is described here as a representative, in which the CrIII and TbIII ions are held together by six μ 3‐OH bridges (O1, O2, and symmetry related) to display a trigonal dipyramidal configuration (Figure 1). The Tb1, Tb2, and Tb1A centers locate at the equatorial plane to form a nearly perfect trianglular plane with lengths of 3.960–3.974 Å and angles of approximately 60°, whereas Cr1 and Cr1A occupy the axial sites, and six μ 3‐OH bridge two TbIII ions to each CrIII ion, forming a [CrIII 2O6TbIII 3] core with a trigonal dipyramidal shape (Figure S1). The Tb⋅⋅⋅Cr distances are 3.478 (Cr1⋅⋅⋅Tb1) and 3.461 Å (Cr1⋅⋅⋅Tb2), and that of Cr⋅⋅⋅Cr is 5.217 Å. The Cr⋅⋅⋅Tb⋅⋅⋅Cr angles are 97.16° (Cr1⋅⋅⋅Tb1⋅⋅⋅Cr1A) and 97.86° (Cr1⋅⋅⋅Tb2⋅⋅⋅Cr1A), whereas the Cr⋅⋅⋅O⋅⋅⋅Tb and Tb⋅⋅⋅O⋅⋅⋅Tb angles are in the ranges of 103.13–104.15° and 109.32–110.11°, respectively (other Cr–Ln complexes are detailed in Tables S1 and S2). This core is further ligated by ten L− ligands and two terminal water molecules around the periphery. Among the ten L− ligands, six L− ligands adopt a syn–syn mode to bridge neighboring CrIII and TbIII ions, and two bidentately chelate one TbIII center, whereas the left two act as monodentate ligands to coordinate with one TbIII center.

Figure 1

a) Molecular structure of 1, representing the common topology observed in complexes 1–3; H atoms and solvent molecules are omitted for clarity. b) Top view of the core with the atom labels. c) Side view of the core, showing the trigonal bipyramidal configuration.

The asymmetric unit consists of a quarter of the molecule, and there is one crystallographically independent CrIII and two TbIII ions in the molecular structure. The CrIII center has close to the perfect octahedral geometry involving O6 donor sets, whereas the larger TbIII ions take up the nine and eight coordination sets, respectively, for Tb1 and Tb2, involving all O donor combinations. TbIII ions were systemically analyzed by using SHAPE 2.1 software,79 resulting in a capped square antiprism (D, with a value of 1.320) and a square antiprism (D 4, with a value of 0.943) for Tb1 and Tb2, respectively, as shown in Figure 2. The octahedral geometry around Cr1 is completed by six O atoms from three μ 3‐OH (O1, O2, O1C) and three carboxylate O atoms (O4, O5, O4C) of three L− ligands. The bond distances range from 1.977(6) to 1.983(5) Å, whereas the trans angles are close to linearity [175.77(27) and 176.256(181)°]. The nine coordination sets around the Tb1 ion contain four μ 3‐OH (O1, O2, O1B, O2A), two carboxylate O atoms (O3, O3B) of two carboxylate groups in syn–syn mode, one bidentate carboxylate group (O9, O10), and a terminal water molecule (O11). For Tb2, the difference from the Tb1 ion is that two monodentate carboxylate groups (O7, O7A) replace the bidentate carboxylate group and the coordinated water molecule is removed.

Figure 2

The coordination spheres around Tb1 and Tb2 in 1.

The AlIII 2LnIII 3 analogues (4 and 5) with the replacement of CrIII by AlIII display the similar structures (Figure S2). The AlIII centers are six‐coordinate with octahedral geometry, while the LnIII centers have two coordination environments: nine donor sets with capped square antiprism (Ln1) and eight ones for square‐antiprism (Ln2), as CrIII 2LnIII 3 series. Their relevant metrical parameters are shown in Table S1 and S2. The packing structure of complexes (1–5) shows 1D channels along c axis, and suggests that the counter ion (Et3NH) and lattice water molecules locate in the channels (Figure S3).

Thermogravimmetric and Powder X‐ray Diffraction Analyses

The thermogravimetric (TG) analyses of complexes 1–5 were measured under a N2 atmosphere with a heating rate of 10 °C min−1 (Figure S4). The weight losses for complexes 1–5 were equal to 5.7 % (170 °C), 5.5 % (170 °C), 5.5 % (170 °C), 6.3 % (150 °C), and 7.6 % (150 °C), which can be assigned to the loss of Et3N for complexes 1–3 and Et3N and H2O for complexes 4 and 5 (calcd 5.5, 5.5, 6.2, 6.6, 6.6 %). Further weight loss gives rise to the decomposition of complexes 1–5.

Powder X‐ray diffraction (PXRD) measurements of complexes 1–5 in crystalline samples were carried out and the results are shown in Figures S5 and S6. The experimental PXRD patterns are consistent with the simulated pattern (complexes 1 and 4, respectively). Minor inconsistencies between experimental and simulation data have been observed in the intensity and shape of peaks, which are attributed to the different orientation of crystals in crystalline samples. All crystallographic data are detailed in Table 1.

Table 1

Crystallographic data for complexes 1–5.

	1	2	3	4	5
Formula	Cr2Tb3 C56H116O28N	Cr2Dy3 C56H116O28N	Cr2Y3 C56H116O28N	Al2Tb3 C56H118O29N	Al2Dy3 C56H118O29N
M r [g mol−1]	1832.25	1842.99	1622.22	1800.23	1810.97
T [K]	153(2)	127.90(14)	143.05(10)	128.15(10)	134(2)
Crystal system	tetragonal
Space group	P42/nmc
a [Å]	18.5502(4)	18.2504(3)	18.4885(6)	18.5518(3)	18.5559(4)
b [Å]	18.5502(4)	18.2504(3)	18.4885(6)	18.5518(3)	18.5559(4)
c [Å]	29.6308(6)	29.7619(8)	29.6790(10)	29.2890(6)	29.3126(5)
α=β=γ [°]	90.00
V [Å3]	10196.3(5)	9913.0(4)	10145.0(7)	10080.3(4)	10093.0(4)
Z	4
ρ calcd [g cm‐3]	1.194	1.235	1.062	1.186	1.192
μ [mm−1]	2.315	2.502	1.960	2.156	2.272
F(000)	3704.0	3716.0	3392.0	3656.0	3668.0
2θ max [°]	50.02	50.02	50.02	50.02	50.01
Refl. collected/unique	32433/4782	19953/4672	23049/4720	20156/4691	20680/4695
R(int)	0.0500	0.0620	0.0835	0.0455	0.0529
GOF on F 2	1.074	1.063	1.058	1.047	1.043
R 1/wR 2 [I>2σ(I)]	0.0586/0.1438	0.0940/0.2299	0.0819/0.2356	0.0528/0.1270	0.0454/0.1085
R 1/wR 2 (all data)	0.0709/0.1555	0.1397/0.2729	0.1266/0.2678	0.0674/0.1377	0.0611/0.1185
Largest differencepeak/hole [eÅ‐3]	2.54/−1.07	2.12/−1.33	1.31/−0.62	1.15/−‐1.12	1.16/−0.71

Magnetic Properties

It is well‐known that lanthanide‐based complexes exhibit unique magnetic properties for their rather large and anisotropic magnetic moments. Over the past two decades, a large number of lanthanide‐based complexes in which the LnIII ion is exchange‐coupled with a second spin carrier, such as transition‐metal ion (M) and organic radical, have been obtained; however, except for isotropic GdIII, little is known about the nature and magnitude of the magnetic interactions within these complexes. This is because the thermal population of the MJ levels of the LnIII ions complicate the magnetic properties, making the analysis of the magnetic behaviors of M⋅⋅⋅LnIII coupling much more difficult. Herein, we successfully applied diamagnetic substitution to get insight into the magnetic nature of CrIII–LnIII interactions through the comparison of the magnetic behaviors of [CrIII 2LnIII 3] and [AlIII 2LnIII 3] complexes.

Static (dc) Magnetic Susceptibility Data

The temperature‐dependent magnetic susceptibilities of complexes 1–5 were collected in the field of 1000 Oe. Plots of χ m T or Δχ m T versus T are shown in Figures 3–6, and the magnetic data of complexes 1–5 are summarized in Table 2.

Figure 3

Plots of χ m T versus T for complexes 1 and 2 at 1000 Oe.

Table 2

Key magnetic data for complexes 1–5.

Complex[MIII 2LnIII 3]	Ground multipletof LnIII	g J	Curie Constant forLnIII ion[a] [cm3  mol−1  K]	Predicted χ m T [a] [cm3  mol−1  K]	Measured χ m T [a] [cm3  mol−1  K]	Measured χ m T [b] [cm3  mol−1  K]
1 [CrIII 2TbIII 3]	7 S 6	3/2	11.82	39.21	35.98	69.66
2 [CrIII 2DyIII 3]	6 H 15/2	4/3	14.17	46.26	41.75	94.51
3 [CrIII 2YIII 3]	‐	g Cr=2	0	3.75	3.45	2.10
4 [AlIII 2TbIII 3]	7 S 6	3/2	11.82	35.46	32.63	18.68
5 [AlIII 2DyIII 3]	6H15/2	4/3	14.17	42.51	37.89	24.23

[a] At 300 K. [b] At 1.8 or 2.0 K.

As shown in Figure 3, complexes 1 and 2 have room‐temperature χ m T values of 35.98 and 41.75 cm3 mol−1 K, respectively, which are smaller than the expected values for two uncoupled CrIII ions and three respective LnIII ions (Table 2), maybe because of the spin‐orbit coupling of lanthanide ions. The curves of complexes 1 and 2 display similar features. As the temperature is decreased, the χ m T values are roughly constant from 300 to 50 K and then increase more and more rapidly, reaching a maximum value at 1.8 K. In fact, there is a slight gradual decrease between room temperature and 100 K, which may be attributed to the thermal depopulation of the Stark sublevels of the anisotropic LnIII ions. The continual increase of the χ m T values for 1 and 2 indicate that non‐negligible and significant magnetic exchange interactions are present between the DyIII/TbIII and the CrIII ions. Within the trigonal bipyramidal structure, the interaction between two CrIII ions, separated by more than 5 Å, is certainly very small (as following depicted) compared to the CrIII⋅⋅⋅LnIII and LnIII⋅⋅⋅LnIII interactions through the μ 3‐O bridges. Therefore, the CrIII⋅⋅⋅LnIII and/or LnIII⋅⋅⋅LnIII interactions within complexes 1 and 2 are ferromagnetic.

Complex 3 incorporates the diamagnetic YIII and paramagnetic CrIII ions into one system, and can give some insight into the nature of the CrIII⋅⋅⋅CrIII magnetic interaction. As shown in Figure 4, the room temperature χ m T value is 3.45 cm3 mol−1 K, which is consistent with the theoretical value of 3.75 cm3 mol−1 K of two uncoupled CrIII ions. As the temperature is decreased, the χ m T value remains constant until 50 K, and then decreases to reach 2.10 cm3 mol−1 K at the lowest temperature. The decrease at the lowest temperature might be caused by the antiferromagnetic coupling or zero‐field splitting on CrIII centers. CrIII 2YIII 3 can be viewed as a dimer of CrIII with a pair of exchange‐coupled spin‐only S=3/2 spins. The susceptibility is given in the exchange spin Hamiltonian written as H ex=‐2J S 1⋅S 2 and can be fitted to give J=−0.22(6) cm−1 and g=1.90(9) (Figure 4), showing the weak antiferromagnetic Cr–Cr interactions. Furthermore, the fitting of the Curie–Weiss law gives θ=−0.96 K and C=3.11 cm3 mol−1 K (Figure S7), and the negative θ confirms the antiferromagnetic interaction.

Figure 4

Temperature dependence of χ m T under a 0.1 T applied field for 3, and the red line shows the fitting.

The plots of χ m T versus T of the corresponding AlIII–LnIII series (complexes 4 and 5) are displayed in Figure 5, and the χ m T values of 32.63 and 37.89 cm3 mol−1 K at 300 K are smaller than the predicted values of three uncoupled LnIII ions (Table 2), maybe owing to the strong spin‐orbit coupling of lanthanide ions. The χ m T values of complexes 4 and 5 decrease slightly as the temperature decreases to 100 K, and then drop rapidly to reach minimum values at the lowest temperature (Table 2). For the AlIII–LnIII series, the AlIII ion is diamagnetic, thus the LnIII⋅⋅⋅LnIII interactions may play an important role in the magnetic behavior. Considering the magnetic results and the thermal depopulation of the LnIII ions, the LnIII⋅⋅⋅LnIII interactions within complexes 4 and 5 cannot be determined.

Figure 5

Plots of χ m T versus T for complexes 4 and 5 at 1000 Oe.

In the vast majority of CrIIILnIII systems, the χ m T values decreased upon lowering the temperature, which suggests the spin‐orbit coupling of the LnIII ions and/or antiferromagnetic interactions between metal ions are dominant within these systems. In some CrIIILnIII cases, the χ m T products presented a small rise before a continual decrease, indicting the presence of weak ferromagnetic interactions, but no examples use diamagnetic substitution to evaluate the magnetic coupling between CrIII and LnIII ions.68, 69, 70, 71, 72, 80, 81, 82 Compared to these complexes, it is the first time diamagnetic substitution is used to determine the strong ferromagnetic interactions between CrIII and LnIII ions within complexes 1 and 2.

To obtain new insight into the nature of the M–LnIII (M=a transition‐metal ion or an organic radical) interactions, the method of diamagnetic ion substitution addressed by Kahn et al. proved to be effective.62 Here, the nature of the magnetic interactions between CrIII and LnIII within complexes 1 and 2 was investigated by comparing the magnetic susceptibilities of CrIII 2LnIII 3 with those of corresponding AlIII 2LnIII 3 and CrIII 2YIII 3 analogues involving the diamagnetic AlIII and YIII ions, respectively. Δχ m T is defined as Δχ m T=(χ m T) −(χ m T) ‐ (χ m T) and was obtained experimentally, which may eliminate the crystal‐field contribution of LnIII ions, and then the profile for Δχ m T could be characteristic of the CrIII⋅⋅⋅LnIII interactions within the complexes. The plots of Δχ m T versus T are displayed in Figure 6. The Δχ m T values for complexes with Tb and Dy increase slightly from 300 to about 100 K, and increase more and more rapidly as the temperature approaches zero. The profile of those curves clearly indicates strong ferromagnetic interactions between the CrIII and LnIII ions within the corresponding complexes.

Figure 6

Plots of Δχ m T defined as Δχ m T=(χ m T) ‐ (χ m T) −(χ m T) versus T.

The field dependence of magnetization was performed for complexes 1–5, and the corresponding field‐independent isothermal magnetization data are shown in Figures S8 and S9. The M versus H plots of complexes 1, 2, 4, and 5 show sharp increases with increasing H at low fields and low temperature, and then linear increases with larger fields. The magnetizations of complexes 1, 2, 4, and 5 are not saturated even at 2 K under 7 T (Figure S8), indicating the presence of magnetic anisotropy, as expected for LnIII‐based complexes. The plots of M versus H/T show non‐superposed curves (Figure S9), further confirming the highly anisotropic ground state and/or low‐lying excited states. The plot of M versus H at 2 K for complex 3 is shown in Figure S10, and the magnetization significant is not saturated under 7 T, reaching a value of 5.44 Nβ at 7 T.

Dynamic (ac) Magnetic Susceptibility Data

To probe the presence of slow magnetic relaxations in these systems, and thus the presence of SMM behaviors, ac magnetic susceptibilities were performed on these complexes at zero dc field. In the CrIII–LnIII series, complexes 1 and 2 exhibit frequency‐dependent out‐of‐phase (χ′′) signals (Figures 7 and 8), whereas complex 3 has no frequency‐dependent signals under the experimental conditions (Figure S11). Meanwhile, among the AlIII–LnIII series, complex 5 with DyIII exhibits frequency‐dependent χ′ and χ′′ signals (Figure 9), and complex 4 with TbIII does not display frequency‐dependent χ′ and χ′′ signals (Figure S12). These results indicate slow magnetic relaxations in complexes 1, 2, and 5. Complex 1, incorporating CrIII and TbIII ions, shows slow magnetic relaxation, whereas the similar complex 4 containing AlIII and TbIII ions does not. Thus, in this system, the CrIII–TbIII ferromagnetic interaction plays an important role in the slow magnetic relaxation for complex 1. Complex 2 based on CrIII and DyIII ions displays slow magnetic relaxation, and the result can be ascribed to the single‐ion behavior of DyIII and/or the CrIII–DyIII ferromagnetic interaction. Complex 5 based on AlIII and DyIII ions shows a similar behavior without the CrIII–DyIII ferromagnetic interaction, probably owing to the single‐ion behavior of the DyIII ion.

Figure 7

The ac susceptibility data for complex 1 at zero dc field.

Figure 8

The ac susceptibility data for complex 2 at zero dc field.

Figure 9

The ac susceptibility data for complex 5 at zero dc field.

However, the expected maximum of χ′′ lies outside the temperature of 1.8 K, mainly owing to fast QTM. Thus, the energy barrier U eff and relaxation time τ 0 cannot be obtained through the conventional Arrhenius method. Recently, Bartolomé et al. employed another method to obtain the U eff and τ 0, assuming that there is only one relaxation process of the Debye type with one energy barrier and one time constant.83 Then, with this assumption for complexes 1, 2, and 5, by plotting ln(χ′′/χ′) versus 1/T at the different frequencies, a linear plot according to the equation ln(χ′′/χ′)=ln(2πντ 0) + E a/k B T serves for the estimation of the energy barrier and the characteristic time (Figures S13 and S14). For complex 1, these estimates are U eff≈E a/k B=17 K and τ 0=7× 10−9 s. For complex 2, the values are U eff≈E a/k B=10 K and τ 0=1.3× 10−9 s. For complex 5, we cannot obtain reasonable parameters. Therefore, we performed them in a static magnetic field to probe the effect on relaxation time.84 As shown in Figure S15, the broad peaks are observed at low frequencies and the relaxation time remain roughly constant, which can be found in many complexes.71, 85 Cole–Cole plots are shown in Figure S16, and show that there is more than one relaxation processes in complex 5. We chose a moderate magnetic field of 2000 Oe to obtain the ac magnetic susceptibility for complex 5 (Figure S17). Although the expected maxima of χ′′ can be observed in out‐of‐phase magnetic susceptibility, the peaks remain constant as the frequencies increase probably owing to QTM.

Conclusions

We have described the syntheses, structures, and magnetism of five new hetero‐metallic MIII 2LnIII 3 complexes (1–5). All complexes possess similar structures, showing a trigonal bipyramidal configuration. Complexes 1 and 2 are associated with 3d CrIII, incorporating TbIII and DyIII, respectively. Complex 3 combines the diamagnetic YIII ion and CrIII in order to elucidate the CrIII–CrIII magnetic interaction. Complexes 4 and 5 incorporate the diamagnetic AlIII ion in the place of the CrIII ion as diamagnetic “blanks” for complexes 1 and 2, which were used for the evaluation of the CrIII–LnIII magnetic interactions in complexes 1 and 2. Magnetic measurements reveal that the CrIII–LnIII magnetic interactions are strongly ferromagnetic, and the CrIII–CrIII magnetic interaction is weak antiferromagnetic. The ac magnetic measurements show that [CrIII 2TbIII 3] (1), [CrIII 2DyIII 3] (2), and [AlIII 2DyIII 3] (5) display slow magnetic relaxations, behaving as SMMs. These results enrich the 3d–4f molecule‐based magnetic materials and illustrate the 3d–4f magnetic nature by incorporating the main‐group metal ion by diamagnetic substitution.

Experimental Section

Materials and Physical Measurements

All reactions were carried out under aerobic conditions with commercially available chemicals and reagents, which were used as received, without further purification. Elemental analyses for C, H, and N were performed by using an Elementar Vario‐EL CHNS elemental analyzer and carried out at Sun Yat‐sen University. TG analyses were performed by a NETZSCH TG 209F1 Iris analyzer in Sun Yat‐sen University. PXRD data were collected on a D/Max‐RA diffractometer (DX‐2600, Dan‐Dong China) with Cu Kα radiation (λ=1.548 Å) operating at 40 kV and 100 Ma. The magnetic properties were measured on a Quantum Design MPMS‐XL7 and a PPMS‐9 ACMS magnetometer. Diamagnetic correction was made with Pascal's constants for all constituent atoms.

Syntheses of Complexes 1–5

{[Cr2Tb3L10(OH)6(H2O)2]Et3NH} (1)

Et3N (0.6 mL) was added to a stirred slurry of HL (0.102 g, 1.0 mmol) and CrCl3⋅6 H2O (0.2 mmol, 0.053 g) in 15 mL CH3CN. After stirring for 1 h at room temperature, CH2Cl2 (15 mL) was added and stirred for another 1 h before Tb(NO3)3⋅6 H2O (0.3 mmol, 0.135 g) was added. The resulting mixture was heated to reflux for 30 min to give a light purple mixture. The mixture was cooled to room temperature and filtered, and the filtrate was allowed to stand at room temperature. After about 2 weeks, well‐shaped purple crystals were obtained, which were subsequently washed with CH3CN and CH2Cl2 (10 mL) and dried in air. Yield: 70 mg (34.4 %, based on Tb). Elemental analysis: calcd for Cr2Tb3C56H116O28: C 36.71, H 6.38, N 0.76; found: C 36.89, H 6.80, N 0.48.

{[Cr2Dy3L10(OH)6(H2O)2]Et3NH} (2)

The same procedure as complex 1 was used, but with Dy(NO3)3⋅6 H2O (0.3 mmol, 0.137 g) in place of Tb(NO3)3⋅6 H2O. Yield: 42 mg (22.8 %, based on Dy). Elemental analysis: calcd for Cr2Dy3C56H116O28: C36.49, H 6.34, N 0.76; found: C 36.62, H 6.74, N 0.44.

{[Cr2Y3L10(OH)6(H2O)2]Et3NH} (3)

The same procedure as complex 1 was used, but with Y(NO3)3⋅6 H2O (0.3 mmol, 0.115 g) in place of Tb(NO3)3⋅6 H2O. Yield: 46 mg (24.5 %, based on Y). Elemental analysis: calcd for Cr2Y3C56H116O28: C 41.46, H 7.21, N 0.86; found: C 41.21, H 7.57, N 0.53.

{[Al2Tb3L10(OH)6(H2O)2]Et3NH⋅H2O} (4)

Et3N (0.6 mL) was added to a stirred slurry of HL (0.102 g, 1.0 mmol) and AlCl3 (0.2 mmol, 0.027 g) in 15 mL CH3CN. After stirring for 1 h at room temperature, CH2Cl2 (15 mL) was added and stirred for another 1 h before Tb(NO3)3⋅6 H2O (0.3 mmol, 0.135 g) was added. The resulting mixture was heated to reflux for 30 min to give a light purple mixture. The mixture was cooled to room temperature and filtered, and the filtrate was allowed to stand at room temperature. After about 2 weeks, well‐shaped purple crystals were obtained, which were subsequently washed with CH3CN and CH2Cl2 (10 mL) and dried in air. Yield: 65 mg (33.7 %, based on Tb). Elemental analysis: calcd for Al2Tb3C56H118O29: C 37.36, H 6.61, N 0.78; found: C 37.88 H 6.96, N 0.51.

{[Al2Dy3L10(OH)6(H2O)2]Et3NH⋅H2O} (5)

The same procedure as complex 4 was used, but with Dy(NO3)3⋅6 H2O (0.3 mmol, 0.137 g) in place of Tb(NO3)3⋅6 H2O. Yield: 77 mg (40.8 %, based on Dy). Elemental analysis: calcd for Al2Dy3C56H118O29: C 37.14, H 6.58, N 0.77; found: C 37.51 H 6.88, N 0.46.

Single‐Crystal X‐ray Diffraction

The crystal data and cell parameters for 1–5 are given in Table 1. Crystallographic data were collected with a SuperNova, Single source at offset (1, 2, 4 and 5) and Xcalibur (3), Eos diffractometers using graphite monochromated Mo/Kα radiation (λ=0.71073 Å). The data integration and empirical absorption corrections were carried out by SAINT programs. The structure was solved by the SHELXTL 2014 program suite by full‐matrix least‐square methods on all F 2 data.86 All the non‐hydrogen atoms were refined anisotropically on F 2 by full‐matrix least‐squares techniques. All hydrogen atoms, except for those of disordered atoms and lattice water molecules, were generated geometrically and refined isotropically using the riding model. The highly disordered solvent molecules in complexes 1–5 were treated by the “SQUEEZE” method as implemented in PLATON,87 and the results were appended to the CIF files.

The supplementary crystallographic data for this paper were deposited with the Cambridge Crystallographic Data Centre (CCDC) as entry CCDC 1540747 (1), 1540748 (2), 1540749 (3), 1540750 (4), and 1540751 (5).88

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.