{Gd^III₇} and {Gd^III₁₄} Clusters Based on Rhodamine 6G Ligand with Magnetocaloric Effect

1. Introduction

The exploration of new compounds in the field of coordination chemistry reveals many fascinating structures and properties. Among them, rare earth ions play an important role in the assembly of Complex polynuclear clusters [1,2,3]. Due to the large ion radius and the inherent ability to participate in various coordination environments, rare earth ions can promote the formation of polymetallic clusters [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20,21,22]. Multinuclear rare earth clusters with different structural types, such as {Ln₁₈} [23], {Ln₂₈} [24], {Ln₃₄} [25], {Ln₃₆} [26], {Ln₃₇} [27], {Ln₄₈} [28], {Ln₆₀} [29], {Ln₇₂} [30], {Ln₁₀₄} [31], {Ln₁₄₀} [32], etc., were synthesized, and the largest even has {Nd₂₈₈} structure [33]. These polynuclear clusters usually exhibit fascinating properties, such as single-molecule magnetism [8,12,14,15,19,30], luminescence [23,24], magneto-optical properties [34] and proton conductive properties [35].

The formation of lanthanide hydroxide clusters involves the hydrolysis of lanthanide ions in the presence of ligands. The hydrolysis process leads to the formation of small lanthanide hydroxide units, which then assemble to form larger clusters. The size and structure of the clusters can be controlled by adjusting the hydrolysis conditions and the choice of ligands. The hydrolysis-induced assembly mechanism of lanthanide hydroxide clusters is still not fully understood due to the elusive coordination configurations of lanthanide ions and the limited characterization methods available. However, recent studies have made progress in determining the intermediate species and the pathways of cluster formation [22]. The formation of high-nuclearity lanthanide clusters is believed to involve the assembly of low-nuclearity subunits, which are formed through initial hydrolysis. These small units are then connected to form the final cluster structure. Understanding the formation mechanism of lanthanide hydroxide clusters is important for the development of new functional materials and applications in various fields. In addition, the interactions between rare earth centers can affect the physical properties of the clusters. The ligand framework around these ions also plays an important role in determining the geometry, stability and functionality of the obtained clusters. By adjusting the ligand design, scientists can strategically guide the self-assembly process, thus forming the required multi-core architecture.

Some Gadolinium compounds have shown excellent magnetocaloric properties [36,37,38], such as Gd(HCOO)₃ [39], Gd(OH)₃ [40], Gd₂O(OH)₄(H₂O)₂ [40], GdPO₄ [41], Gd(OH)CO₃ [42], GdF₃ [43], Gd(OH)F₂ [44] and Ba₂Gd(BO₃)₂X (X = F, Cl) [45]. Because the magnetocaloric effect (MCE) in Gd is particularly strong, making it an ideal material for use in extremely low temperature magnetic refrigeration systems. Magnetic refrigeration is a technology that uses a magnetic field to cool objects. Magnetic refrigeration materials are playing an increasingly important role in the future of social development, particularly in the area of energy efficiency and sustainability. In addition, molecular clusters with diverse structures also exhibit enormous magnetic cooling potential and exhibit many physical properties, such as chirality, spin crossover, fluorescence, etc.

Rhodamine-derived ligands show ring-opened or ring-closed structures and can form rare earth [46,47,48,49] or transition metals [50,51] complexes. They are most mononuclear or low-nulearity complexes. We are interested in high-nuclearity clusters based on the rhodamine ligands using the hydrolysis approach. In this work, we used the rhodamine 6G ligand (Figure 1) to synthesize two different metal complexes: hexagonal heptanuclear {Gd^III₇} [Gd₇(L)₆(μ₂-CH₃O)₄(μ₃-CH₃O)₄(μ₃-OH)₄(NO₃)₂]NO₃·10CH₃CN·10CH₃OH·2H₂O (Complex 1) and tetradecanuclear {Gd^III₁₄} Gd₁₄(H_0.5L)₈(μ₆-O)(μ₄-O)₂(μ₃-OH)₁₆(NO₃)₁₆·9.5CH₃CN·2CH₃OH·11H₂O (complex 2) with unexpected three-layer double sandwich structure. These two complexes have excellent magnetic refrigeration performance, and their magnetic entropy changes are 17.4 J kg⁻¹ K⁻¹ at 3 K and 5 T, and 22.3 J kg⁻¹ K⁻¹ at 2 K and 5 T, respectively.

2. Results and Discussion

2.1. Synthesis

Rhodamine 6G-type ligands have attracted significant attention in the realm of fluorescent sensor. Previous studies have primarily focused on the synthesis of mononuclear rare earth or transition metal complexes [46,47,48,49,50,51,52]. However, these ligands have yet to be fully unearthed. We speculate that the reaction between Ln^III and the rhodamine 6G ligands might lead to high-nuclearity lanthanide complexes under high pH value. On the basis of the hydrolysis strategy, we have successfully synthesized two new Gd^III clusters 1 and 2 (Figure 1). Additionally, our research has unveiled the influential factors that impact the synthesis process, including the ratio of central ions to ligands, solvent, alkaline, and reaction temperature.

Figure 1. Transformation of ring-opened and ring-closed form of Rhodamine 6G-type ligands and its reactions.

Figure 1. Transformation of ring-opened and ring-closed form of Rhodamine 6G-type ligands and its reactions.

The synthetic procedure of the two clusters is similar with the exception of the molar ratio of Gd:L, i.e. an excess amount of gadolinium nitrate was used in the synthesis of {Gd₁₄}. Both Complexes were prepared by the reaction of the ligand HL with gadolinium nitrate in a mixed solution of methanol and acetonitrile. A quantity of triethylamine was used to induce the ring close of HL and hydrolysis. The resulting mixture is left undisturbed at room temperature for one week, facilitating formation of yellow plate-like single crystals for {Gd₇}. The reactant mixture was heated at 60°C in an oven for three days, resulting in the formation of yellow, cubic-shaped samples of {Gd₁₄}. These two Complexes tend to lose solvents in the air. When the crystals are taken out of the solution, they turn from yellow to red at room temperature, and they lose the crystallinity. The powder XRD pattern of Gd₇ illustrates that the main strong peaks show disagreement with that simulated, indicating the desolvation of the crystals (Figure S1). The peaks at low 2θ angles of PXRD data for Gd₁₄ are approximately consistent with that simulated.

2.2. Structure

A yellow single crystal of complex 1 was selected for single crystal X-ray diffraction at 100 K. Complex 1 crystallizes in the monoclinic system with the P2₁/n space group. The volume of the unit cell is very large with a monoclinic system, and therefore the diffraction data are not so good. The command MASK was used during the structural refinement, and 452 electrons were masked per formula unit which account for the missing NO₃^- anion, 10 acetonitrile, 10 methanol and 2 H₂O molecules with the total electrons of 451. The crystallographic data and selected bond distances and angles are given in Tables S1 and S2. The asymmetric unit cell contains three crystallographically independent {Gd^III₇} moieties. Because they have similar molecular structures, only one of them is described in detail as a representative. The coordination number of each Gd^III ion in {Gd₇} (Figure 2) is between 7 and 9. Their coordination patterns are shown in Figures S2 and S3. Using SHAPE software for calculation, their coordination patterns were obtained as shown in Tables S3–S5. The central Gd1 ions are coordinated by eight oxygen atoms and have a square antiprism structure D_4d. Among the eight oxygen atoms, four are μ₃-OH^- and another four are μ₃-CH₃O^-. The remaining six Gd^III ions are evenly distributed around the central Gd^III ion, forming a saddle-shaped structure, which is relatively rare in rare earth complexes [53] (Figure S4). One ring-closed ligands L^- is coordinated to one peripheral Gd^III ion, and adjacent Gd^III ions are bridged by CH₃O^-/μ₃-OH^- and/or phenoxy oxygen atoms as shown in Figure 3.

As for complex 2, a yellow cube single crystal was selected for single crystal X-ray diffraction at 100 K. A solvent mask was used and 183.4 electrons were found in a volume of 3448.9 ų in 9 voids per unit cell, which is consistent with the presence of 2 CH₃OH, 2 H₂O and 1.5 CH₃CN per formula unit with 178 electrons. The crystalographic data and selected bond distances and angles are given in Tables S1 and S6. Complex 2 crystallizes in the tetragonal system with space group of P4/n and has a D_4h symmetry. The asymmetric unit contains 1/4 of the tetradecanuclear molecule and there are five different kinds of Gd^III ions (Gd1-Gd5, Figure 4b). They have three different coordination modes (Table S7), and their coordination patterns are shown in the Figures S5 and S6. The tetradecanuclear cluster core is neutral, and has a highly symmetrical three-layer double sandwich structure (Figure 4b). In the structure, four nine-coordinated Gd^III ions form a square plane layer, and a nine-coordinated Gd^III ion is located between the layers. The distance between the two layers is 5.792 Å. The center of the middle layer is six-coordinated μ₆-O^2-, while the outer layers on both sides are μ₄-O^2-. The sandwiched Gd^III ions and the square-shaped layer are connected by μ₃-OH^- ions. The Gd ions are linked together through hydrophilic hydroxo bridges, forming a [Gd₁₄(μ₆-O)(μ₄-O)₂(μ₃-OH)₁₆] core. This core contains one octahedral [Gd₆(μ₆-O)(μ₃-OH)₈] unit that shares two apexes with two [Gd₅(μ₄-O)(μ₃-OH)₄] square pyramid moieties. The cluster core is surrounded by eight ring-closed L^- ligands. Additionally, the Gd^III of middle layer are coordinated with two nitrate ions, and Gd^III of outer layers on both sides are coordinated with a nitrate ion and a L^- ligand. Square plane layers are bridged by phenolic oxygen on the ligand as shown in the Figure 3b.

Although there are many reports on tetranuclear clusters of planar quadrilateral [54,55] and nine-nuclear molecules of double-layer sandwich type [56,57], the molecular structure of rare earth in the form of a three-layer double sandwich is rare. Similar molecules have been reported before, as shown in the Figures S7–S10. The tetradecanuclear hydroxo–lanthanide acetylacetonato complexes formulated as Ln₁₄(μ₄-OH)₂(μ₃-OH)₁₆(μ-η²-acac)₈(η²-acac)₁₆ (Ln = Tb and Eu, acac^- = acetylacetonato) [58] and chiral tetradecanuclear hydroxo-lanthanide clusters Ln₁₄(μ₄-OH)₂(μ₃-OH)₁₆(μ-η²-acac)₈(η²-acac)₁₆·6H₂O (Ln = Dy and Tb) [59] have been reported. The ligands used in these two works are both based on acetylacetonato, but the present tetradecanuclear {Gd^III₁₄} are completely different ligands, i.e. ring-closed rhodamine L^-. The use of ortho-nitrophenolate afforded the tetradecanuclear H₁₈[Ln₁₄(μ-η²-o-O₂N-C₆H₄O)₈(η²-o-O₂N-C₆H₄O)₁₆(μ₄-O)₂(μ₃-O)₁₆] (Ln = Dy and Tm; o-O₂N-C₆H₄O^- = o-nitrophenolate) [60]. Despite the above similar Ln₁₄ complexes, the μ₆-O^2- in 2 is unique among them.

It is worth mentioning that similar hexadecanuclear molecules [Eu^III₁₆(tfac)₂₀(CH₃OH)₈(μ₃-OH)₂₄(μ₆-O)₂] have also been reported based on trifluoroacetylacetone (tfac^-) [22]. In addition to the tetradecanuclear {Eu₁₄}, there is another two Eu ions on opposite sides. The study provides insights into the formation, evolution, and assembly of lanthanide hydroxide clusters. The formation of {Gd₇} and {Gd₁₄} in this work further verifies that the hydrolysis under high pH values is an effective way of constructing high-nuclearity Ln^III species.

2.3. Magnetic Measurements

The temperature dependence of the magnetic susceptibility of Complexes 1 and 2 are measured under a 1000 Oe magnetic field in the range of 2–300 K (Figure 5). At room temperature, the χ_MT value of 54.1 cm³ K mol^-1 for {Gd₇} and 109.0 cm³ K mol^-1 for {Gd₁₄} is close to the theoretical value of 55.09 cm³ K mol^-1 for heptanuclear and 110.18 cm³ K mol^-1 for tetradecanuclear uncoupled Gd^III (S = 7/2, g = 2, C = 7.87 cm³ K mol^-1 per Gd), respectively. For 1, upon lowering the temperature, the χ_MT value slightly decreases to 49.76 cm³ K mol^-1 at 20 K and then rapidly falls to 30.22 cm³ K mol^-1 at 2 K. 2 exhibits a similar behavior: when lowering the temperature, the χ_MT value slightly decreases to 95.4 cm³ K mol^-1 at 30 K and then rapidly falls to 33.0 cm³ K mol^-1 at 2 K. These changes indicate the presence of dominant antiferromagnetic interactions between the Gd^III ions in the clusters. The data can be perfectly fitted to the Curie-Weiss law, giving C = 54.44 cm³ K mol^-1 and θ = -1.712 K for {Gd₇} and C = 110.50 cm³ K mol^-1 and θ = -4.629 K for {Gd₁₄}. Larger absolute θ value in {Gd₁₄} suggests that the antiferromagnetic interaction is stronger than that for {Gd₇} (Figure S11).

The field dependence of the magnetizations (M) for Complexes 1 and 2 was measured in the temperature range of 2 - 10 K (Figure S12). It can be seen that the magnetization has not reached to saturation at 5 T and 2 K. At 2 K, the experimental maximum magnetization value of 47.03 Nβ for 1 and 89.35 Nβ for 2 is lower than the theoretical saturation value of Gd^III (49 Nβ for 1 and 98 Nβ for 2, respectively), which may be owing to the antiferromagnetic coupling and higher magnetic field is needed to suppress the magnetic coupling effect. For complexes 1 and 2, the experimental M-H curves at 2 K lie below the calculated Brillouin curve for non-interacting S_Gd spins (Figure S12), also suggesting the presence of intermetallic antiferromagnetic coupling. The difference between the experimental and the calculated ones for Gd₁₄ is obviously larger than that for Gd₇, indicating that the former shows stronger antiferromagnetic coupling than that of the latter. The presence of μ₆-O^2--bridged Gd₆O moiety may be responsible for this.

The half-filled 4f electronic configuration in a Gd^III ion makes it magnetically isotropic. This makes gadolinium a valuable material in various applications, especially magnetic refrigeration. Thus, the magnetocaloric effect (MCE) of Complexes 1 and 2 was studied using the Maxwell equation:

$$∆S_m{\left(T\right)}_{∆H}=\int {\left[\frac{\partial M(T,H)}{\partial T}\right]}_{H}dH$$

At 3 K and ∆H = 5 T, the value of ∆S_m is 17.44 J kg⁻¹ K⁻¹ for 1(Figure 6a), which is slightly lower than the expected value of 14.56R (25.37 J kg⁻¹ K⁻¹) calculated for 7 uncorrelated Gd^III using the equation ∆S_m = nRln(2S+1) (R ≈ 8.314 J mol^-1 K^-1). The value of ∆S_m for 2 is 22.30 J kg⁻¹ K⁻¹ at 2 K and ∆H = 5 T (Figure 6b), which is close to the expected value of 29.11R (28.72 J kg⁻¹ K⁻¹) calculated for 14 uncorrelated Gd^III. To improve the magnetic refrigeration effect of gadolinium clusters [37], several approaches can be considered. Firstly, the experimental conditions can be optimized. For instance, the temperature and magnetic field can be carefully controlled to ensure the most efficient operation of the gadolinium clusters. The thermal conductivity of the environment and the pressure during the refrigeration cycle can also be adjusted to minimize energy loss. Secondly, the chemical composition of the clusters can be varied. Gadolinium can be alloyed with other metals to create compounds with different magnetic properties. Thirdly, the size of the clusters can be optimized. The optimal size will depend on the specific setup and application, but generally, smaller clusters have a higher surface-to-volume ratio, which leads to more efficient heat exchange and therefore a stronger refrigeration effect. However, too small clusters may also suffer from higher energy barriers between spin states, which can decrease the magnetic entropy change. Overall, a combination of these strategies can be used to improve the magnetic refrigeration effect of gadolinium clusters for various applications, such as cryogenic cooling of scientific instruments, temperature control in electronics, and energy-efficient refrigeration in households and industries.

3. Materials and Methods

3.1. Synthesis and Preparations

All of the reagents we used were commercially available and used without further purification. The ligand HL we used was synthesized by the literature method [46,47,48,49].

3.1.1. Synthesis of [Gd₇(L)₆(μ₂-CH₃O)₄(μ₃-CH₃O)₄(μ₃-OH)₄(NO₃)₂]NO₃·10CH₃CN·10CH₃OH·2H₂O (1).

The ligand HL (0.2 mmol, 106 mg) was suspended in the mixed solution of MeOH (10 mL) and MeCN (10 mL), to which Gd(NO₃)₃·6H₂O (0.25 mmol, 108 mg) was added affording a red solution. The solution was heated and stirred for a few minutes, then triethylamine (300 µL) was add to give a yellow solution. The mixture was filtered, and the filtrate was placed undisturbed at room temperature. Yellow crystals suitable for X-ray diffraction analysis were collected after about 7 days. Yield: about 20%. Desolvated samples were used for elemental analysis and physical measurements. Elemental analysis calcd (%) for C₂₀₆H₂₁₄N₂₇O₃₉Gd₇ (4793.23): C, 51.62; H, 4.50; N, 7.89. Found: C, 51.79; H, 4.91; N, 7.92.

3.1.2. Synthesis of Gd₁₄(H_0.5L)₈(μ₆-O)(μ₄-O)₂(μ₃-OH)₁₆(NO₃)₁₆·9.5CH₃CN·2CH₃OH·11H₂O (2).

The ligand HL (0.2 mmol, 106 mg) was suspended in the mixed solution of MeOH (10 mL) and MeCN (10 mL). Gd(NO₃)₃·6H₂O (0.35 mmol, 150 mg) was added to afford a red solution. After heating and stirring for a few minutes, triethylamine (150 µL) was add, giving rise to a yellow solution. The solution was filtered and the filtrate was heated in an oven at 60ºC. Yellow crystals suitable for X-ray diffraction analysis were obtained after about 2 days and collected. Yield: about 30%. Elemental analysis calcd (%) for C₂₈₅H_330.5N_57.5O₁₀₄Gd₁₄ (8427.05): C, 39.62; H, 3.95; N, 9.56. Found: C, 39.27; H, 4.10; N, 9.25.

3.2. Physical measurements

Elemental analyses (C, H, and N) were performed on an Elementar Vario Cario Erballo analyzer. Powder X-ray diffraction (PXRD) measurements were recorded on a Bruker D8 ADVANCE X-ray diffractometer using CuKα radiation (λ = 1.54184 Å) at room temperature from 5° to 50° with sweeping speed of 10°/min. Single-crystal X-ray data were collected on a Rigaku SuperNova, Dual, Cu at zero, AtlasS2. The structure was solved by program SHELXT and refined by a full matrix least-squares method based on F2 using SHELXL-2014/7 method. Hydrogen atoms were added geometrically and refined using a riding model. Temperature- and field-dependent magnetic susceptibility measurements were carried out on a Quantum Design MPMS-XL5 SQUID magnetometer.

4. Conclusions

In conclusion, novel high-nuclearity clusters have been obtained via the hydrolysis strategy: saddle-shaped {Gd₇} and three-layer double sandwich {Gd₁₄}. Both complexes exhibit good magnetocaloric properties with the magnetic entropy changes of 17.4 J kg⁻¹ K⁻¹ for Gd₇ at 3 K and 5 T and 22.3 J kg⁻¹ K⁻¹ for Gd₁₄ at 2 K and 5 T, respectively. Further work on new clusters {Ln^III₇} and {Ln^III₁₄} (Ln = Dy, Tb, Eu and Ho) are in progress in our laboratory, which maybe behave as SMMs or fluorescent materials. The combination of rare earth ions and new ligands paves the way for the discovery of new polynuclear clusters with unprecedented structures and functionalities.