3-Pyridylacetic-Based Lanthanide Complexes Exhibiting Magnetic Entropy Changes, Single-Molecule Magnet, and Fluorescence.

Four complexes from lanthanides, 3-pyridylacetate, and 1,10-phenanthroline, formulated as [Ln2(3-PAA)2(μ-Cl)2(phen)4](ClO4)2 [Ln = Gd(1), Dy(2), Eu(3), Tb(4), 3-PAA = 3-pyridylacetic acid, phen = 1,10-phenanthroline], were obtained. The four compounds were characterized by IR spectra, thermogravimetric analyses, powder X-ray diffraction, and single-crystal X-ray diffraction. Compounds 1-4 are isomorphous, and they have a dinuclear structure. Magnetic studies reveal that 1 shows the magnetocaloric effect with -ΔS m max = 19.03 J kg-1 K-1 at 2 K for ΔH = 5 T, and 2 displays a field-induced single-molecule magnet with U eff = 19.02 K. The photoluminescent spectra of 3 and 4 exhibit strong characteristic emission, which demonstrate that the ligand-to-EuIII/TbIII energy transfer is efficient.

Introduction

Lanthanide-based complexes have received considerable interest because of their remarkable properties stemming from unique 4f electrons of lanthanide ions.[1−4] On the one hand, Ln-based complexes hold a prominent position in magnetism.[5,6] For example, Gd(III) ion with a large spin ground state S, Dion = 0, and low-lying excited spin states is considered as a potential candidate for molecular refrigerant materials with a large magnetocaloric effect (MCE).[7,8] Such materials are sought because of their potential applications in cryogenic refrigeration.[9−11] Dy(III) ion with large magnetic moment and strong anisotropy is regarded as an ideal spin carrier for single-molecule magnets (SMMs),[12−14] which is the major breakthrough in the field of magnetic research in the recent two decades. Such magnets are favored because of their similar memory effects observed in magnetic nanoparticles and their potential application in high-density data storage quantum computing and spintronics of molecular dimensions.[15] To date, great progress has been made in Dy(III)-based SMMs, such as mononuclear and polynuclear DyIII systems.[16−20] For example, Tong and Mills’ groups reported two Dy(III)-SMMs showing hysteresis at a liquid nitrogen temperature.[21,22] The anisotropy of DyIII primarily originates from the unquenched orbital momentum of the ion, but it is also affected by the ligand field.[23] Therefore, the design and synthesis of SMMs remain a challenge, and ligands are important for the preparation of Ln-based SMM. In addition, previous achievements demonstrate that intramolecular magnetic exchange can decrease tunnel splitting (T), which can effectively suppress quantum tunneling of magnetization (QTM) and then greatly improve the effective barrier (Ueff).[24] Hence, the design and construction of the simplest polynuclear SMM, a Dy2 system, are important because such species can offer a good platform for a comprehensive understanding of magnetic exchange and the origin of magnetism in SMMs.

On the other hand, luminescence from lanthanides has aroused considerable attention because of their outstanding optical properties such as long-lived emissions, high quantum yields, narrow bandwidth, and large Stokes shifts, as well as their wide applications in lasers, sensing, white-light emission, luminescent thermometers, and color displays.[25−27] Among lanthanide ions, EuIII and TbIII are considered important optical centers because of their strong, visible, and easily detected emissions. However, the direct excitation of lanthanide ions is disfavored because of their spin- and parity-inhibited f–f transitions.[28,29] One effective method is to introduce suitable π-conjugated ligands, which can absorb light and transfer energy to LnIII ions (antenna effect), thereby improving lanthanide luminescence.[30,31] Hence, organic ligands play a key role in the construction of luminescent Ln complexes.

In addition, pyridylecarboxylic ligands are a class of attractive ligands bearing N and O coordination atoms, which have been extensively used to construct complexes showing structural variations and excellent properties because of their exceptional coordination ability and various coordinating modes.[32] However, the 3-pyridylacetic ligand (3-PAA) as an important and simple pyridylecarboxylic ligand has received less attention. Based on previous reports, no example of 3-pyridylacetate-based Ln complexes has been reported, and only several transition-metal complexes with 3-PAA have been reported.[33−36] As a part of our continuous studies on Ln complexes with magnetic and luminescent properties,[37,38] herein, a series of Ln-based complexes are obtained using 3-PAA and 1,10-phenanthroline (phen) as auxiliary ligands. This work is based on the following ideas: (1) Ln-based complexes fabricated by 3-PAA are unexplored, which provide substantial research opportunities; (2) phen as a kind of N,N′ bulky auxiliary ligands can block the high coordination sites of lanthanide centers, which can achieve polynuclear Ln-based SMMs with isolated structures; and (3) the 3-PAA and phen ligands are aromatic π-conjugated molecules, which can serve as chromophoric antenna ligands to sensitize lanthanide luminescence. In this work, four new Ln-based complexes, namely, [Ln2(3-PAA)4(μ-Cl)2(phen)4] [Ln = Gd(1), Dy(2), Eu(3), Tb(4), 3-PAA = 3-pyridylacetic, and phen = 1,10-phenanthroline], were prepared, and they show a dinuclear structure. 1 shows the MCE with −ΔSmmax = 19.03 J kg–1 K–1 at 2 K for ΔH = 5 T, whereas 2 exhibits SMM behavior. The photoluminescent properties of 3 and 4 have also been investigated. Notably, the four complexes represent the first lanthanide-based complexes constructed from 3-PAA.

Results and Discussion

Synthetic Aspects

Multiple factors such as the concentration and type of reactants, pH value, reaction temperature, reaction time, and type of solvent can affect the formation and crystallization of final products in a solvothermal system. We obtained a family of novel Ln-based complexes by solvothermal reactions of the corresponding Ln(ClO4)3·6H2O, 3-PAA·HCl, 1,10-phenanthroline, and Et3N in CH3CN. A series of systematic studies indicated that the pH and the temperature of the reaction play a key role for the resulting products. The initial pH values of the syntheses for 1–4 were in a range of 5.5–6.5. In addition, parallel experiments showed that the temperature of the reaction suitable products cannot be produced at 140, 160, and 180 °C (Scheme ).

Scheme 1

Synthetic Procedure of Complexes 1–4

Structural Description and Discussion

Single-crystal X-ray diffraction analyses revealed that complexes 1–4 were isostructural, and such complexes crystallized in the monoclinic space group P21/c. PXRD studies showed that solid samples of the as-synthesized complexes 1–4 were in agreement with single-crystal X-ray diffraction studies, indicating the phase purity of 1–4 (Figure S1 in the Supporting Information (SI)). Pertinent crystal data and structure refinement results for 1–4 are summarized in Table , and selected bond lengths are listed in Table S1 in SI. All of the compounds were composed of cationic dimers [Ln(PAA)(μ-Cl)(phen)2]2+, which were charge-compensated perchlorate in the accessible voids. Herein, as an example, the structure of complex 1 was discussed in detail. As shown in Figure a, the asymmetric unit of 1 contains one crystallographically independent GdIII ion, one PAA1– ligand, one Cl– anion, two phen molecules, and one ClO4– anion. The GdIII center was eight-coordinated in an O2Cl2N4 donor set, which was constructed by two carboxylate oxygen atoms from two different PAA1– ligands, two Cl atoms, and four nitrogen atoms from two different chelated phen molecules. The polyhedral geometry of the GdIII center was systematically analyzed using SHAPE 2.1 program.[39] The result showed that the coordination geometry of the GdIII ion can be regarded as a distorted square antiprism (D4 symmetry) with a continuous shape measurement (CShM) value of 1.381 (Figure b and Table S2 in the SI). The bond lengths of Gd–O were 2.349(2) and 2.352(2) Å; the bond length of Gd–Cl was 2.7759(7) Å, whereas the bond length of Gd–N ranged from 2.553(3) to 2.579(3) Å, which were close to those found in previously reported gadolinium–oxygen, chlorine, and nitrogen donor compounds.[40,41] As shown in Figure c, two neighboring Gd atoms were connected by two bridging Cl– ions and two bridging carboxyl groups of two PAA1– ligands to form a binuclear lanthanide cluster [Gd2(μ-Cl)2(μ-OCOPAA)2]2+, in which the Gd···Gd1A separation and Gd1–Cl1–Gd1A angle were 3.9059(3) Å and 89.747(21)°, respectively. In 1, the PAA1– ligand served as a bidentate ligand and adopted the μ2–η1:η1 coordination mode to bind GdIII ion by only using its carboxylate group, whereas its pyridine N atom was uncoordinated. This result was similar to other reported Ln complexes based on pyridylecarboxylic ligands. Then, these binuclear clusters were stacked together by face-to-face π–π stacking interactions from the phenyl rings of phen ligands, with a distance of 3.693(3) and 3.809(3) Å, generating a two-dimensional (2D) supramolecular layer along the ab plane (Figure ), and counteranions [ClO4]− were located in the accessible voids of these supramolecular layers along the ab plane (Figure S2 in the SI).

Table 1

X-ray Diffraction Crystallographic Data for 1–4

	Gd(1)	Dy(2)	Eu(3)	Tb(4)
formula	C62H44Cl4N10O12Gd2	C62H44Cl4N10O12Dy2	C62H44Cl4N10O12Eu2	C54H38Cl4N16O12Tb2
Fw	1577.37	1587.87	1566.79	1580.71
temp (K)	296(2)	296(2)	293(2)	296(2)
crystal system	monoclinic	monoclinic	monoclinic	monoclinic
space group	P21/c	P21/c	P21/c	P21/c
a (Å)	12.09360(10)	12.05930(10)	12.1286(12)	12.07650(10)
b (Å)	11.7499(2)	11.72510(10)	11.7919(11)	11.73330(10)
c (Å)	21.6432(3)	21.6822(2)	21.603(2)	21.6298(2)
β (deg)	92.6250(10)	92.7200(10)	92.461(3)	92.55
V (Å3)	3072.24(7)	3062.33(5)	3086.8(5)	3061.85(5)
Z	2	2	2	2
Dc (g cm–3)	1.705	1.722	1.686	1.715
μ (mm–1)	2.385	2.667	2.257	2.537
F (000)	1556	1564	1552	1560
reflns collected	26 260	17 604	18 856	20 243
independent reflns	5498	5327	5215	5626
Rint	0.0235	0.0255	0.0236	0.0237
theta range (deg)	1.88–25.50	1.90–25.01	3.23–25.00	1.88–25.50
params/restraints/data	400/23/5498	400/829/5327	406/18/5215	400/18/5626
R1 [I > 2σ(I)]a	0.0540	0.0424	0.0352	0.0407
wR2 (all data)b	0.2463	0.1558	0.1308	0.1475
GOF on F2	1.005	1.002	1.013	0.977
ρmax/ρmin (e Å–3)	0.7456/0.6682	0.7456/0.6592	1.000/0.784	0.7456/0.6372

R1 = ||Fo| – |Fc||/|Fo|.

w2 = [w(Fo2 – Fc2)2]/[w(Fo2)2]1/2.

Figure 1

(a) Ball and stick plot showing the asymmetric unit of 1. (b) Coordination polyhedron around the GdIII ion in 1. (c) Drawing of the dinuclear structure of 1. H atoms and perchlorate are omitted for clarity. Symmetry code A: 1 – x, 1 – y, −z.

Figure 2

Two-dimensional supramolecular layer constructed by π···π interactions (pink dashed lines) in 1 along the ab plane.

Thermal Behavior

As shown in Figure S3 in the SI, the thermogravimetric (TG) analysis (TGA) plots of complexes 1–4 showed that they had similar thermal behavior because they were isostructural, which is consistent with the reported isostructural complexes.[42] The thermogravimetric curve revealed that complexes 1–4 involved a loss of 34.50% (calcd 34.37%) for 1, 34.72% (calcd 34.15%) for 2, 34.55% (calcd 34.60%) for 3, and 34.02% (calcd 34.22%) for 4 from 25 to 379 °C, which can be attributed to the release of two counteranions [ClO4]− together with two Cl– and the removal of two coordinated PAA– ligand. And the subsequent pyrolysis process occurred from 379 to 800 °C due to the decomposition of the network. The result of the TG analysis basically agrees with that of the structure determination of 1–4.

Magnetic Properties

In recent years, molecular cryomagnetic coolants have received considerable interest not only because they have potential applications in cryogenic refrigeration with high energy efficiency and environmental friendliness but also because they have a stoichiometric composition, modifiability, and monodispersity.[43−45] Previous studies showed that those spin centers bearing large spin, single-ion isotropy, and low-lying excited spin states such as Fe3+, Mn2+, and Gd3+ had great potential application in constructing high-performance materials.[46,47] In particular, compared with strong magnetic couplings existing in Fe3+ and Mn2+ compounds, weak magnetic interactions in Gd3+ compounds stemmed from the shielding of its 4f orbital improved magnetic refrigeration materials with large MCE. Thus, here, the magnetic properties and MCE of complex 1 were investigated. The plots of χM and χMT of 1 in a constant field of 1000 Oe in the temperature range of 2.0–300 K are illustrated in Figure a. χM displayed a gradual increase from 0.052 cm3 mol–1 at 300 K to 0.54 cm3 mol–1 at 28 K and then exponentially reached the maximum value of 5.87 cm3 mol–1 at 2 K. Correspondingly, χMT of 15.61 cm3 K mol–1 at 300 K for 1 was in good agreement with the theoretical value (15.75 cm3 K mol–1) for two uncoupled GdIII (S = 7/2, L = 5, 8H7/2, g = 2) ions.[48] As the temperature decreased, χT increased gently and remained nearly constant and then sharply declined to the minimum value of 11.73 cm3 K mol–1 at 2 K, which was related to the total zero-field splitting or the antiferromagnetic exchange interactions between two GdIII cations. The 1/χ–T plot of 1 followed the Curie–Weiss law with a negative Weiss constant θ = −1.14 K in the temperature range of 2–300 K (Figure S4 in the SI), which further confirmed antiferromagnetic couplings among adjacent GdIII ions. We studied the magnetic entropy changes −ΔSm to investigate the MCE of the Gd2 complex. Therefore, the magnetization measurements of 1 were measured in the range of 0–5 T at 2 K (Figure S5 in the SI), and the M–H curves increased gradually with the increase of applied field. Moreover, M of 1 was 14.31 Nβ at 2 K and 5 T, which was consistent with the expected value of 14.00 Nβ for two isolated GdIII ions (g = 2.00). The magnetic entropy change (ΔSmmax), that is, a key parameter in evaluating the MCE, can be calculated on the basis of the Maxwell equation ΔSmmax = ∫[∂M(T,H)/∂T] dH.[49,50] Based on the experimental magnetization data, ΔSm values of 1 at variable temperatures and applied magnetic fields were obtained (Figure S6 in the SI). The largest −ΔSm value was 19.03 J kg–1 K–1 at 2 K under the condition ΔH = 5 T (Figure b), which was slightly smaller than the theoretical value for two noninteracted GdIII ions based on the equation −ΔSmmax = nR ln(2SGd + 1)/MW, in which n represents the number of GdIII per mole, R is the gas constant, and S is the spin state. This phenomenon can be assigned to the presence of weak antiferromagnetic couplings among the GdIII ions and a small metal/ligand mass ratio.[51] The magnetocaloric effect (MCE) of 1 (19.03 J K–1 kg–1 at 2 K and 5 T) is compared to the reported complex [Gd4(acac)4(μ2-OH)(L)6]·xCH3CN·yC2H5OH (Hacac = acetylacetone, HL = 5-(4-ethylbenzylidene)-8-hydroxylquinoline, 21.2 J kg–1 K–1 at 2 K and 5 T).[52] Moreover, its MCE of 12.5 J K–1 kg–1 at 2 K and 2 T is slightly lower than that of commercial Gd3Ga5O12 (14.6 J kg–1 K–1 at 2 K and 2 T)[53] but can be comparable to that of the reported Gd5BSi2O13 (12.2 J kg–1 K–1 at 2 K and 2 T).[54]

Figure 3

(a) Temperature dependence of the magnetic susceptibility for complex 1 between 2 and 300 K. (b) −ΔSm for complex 1 calculated using the magnetization data at variable fields and temperatures.

The direct-current (dc) magnetic susceptibility of 2 was also investigated in the temperature range of 1.8–300 K under 1000 Oe (Figure a). Upon reducing the temperature, the χM value of 2 slowly increased from 0.098 emu mol–1 at 300 K to 0.327 emu mol–1 at 100 K and then sharply reached the maximum value of 8.32 emu mol–1 at 2 K. The χMT value of 2 at 300 K was 28.08 cm3 K mol–1, which was in good agreement with the expected value of 28.34 cm3 K mol–1 for two isolated DyIII ions (S = 5/2, L = 5, 6H15/2, g = 4/3). Upon cooling, a gradual increase in χMT was observed for 2, reaching the maximum value of 35.75 cm3 K mol–1 at 36 K, and then χMT decreased rapidly below this temperature, reaching the maximum value of 16.55 cm3 K mol–1 at 2 K. This phenomenon is the result of the combination of the depopulation of M levels of the Dy(III) ion and the ferromagnetic interaction between the Dy(III) ions.[55−58] The field dependence of the magnetization of 2 was examined at 2–5 K (Figure b). The magnetization of 16.8 Nβ at 50 kOe and 2 K was smaller than the theoretical saturation value of 20 Nβ (10 Nβ for each DyIII ion). Moreover, M vs H/T curves displayed a nonsuperimposable nature. These phenomena indicated the presence of strong magnetic anisotropy associated with 2.[59]

Figure 4

(a) Temperature dependence of the magnetic susceptibility for complex 2 between 2 and 300 K. (b) M vs H/T curves for complex 2.

The frequency dependence of ac susceptibilities was tested to study the magnetic properties of 2. As shown in Figure S7 in SI, no out-of-phase signal (χ″) appeared for the high frequency of 997 Hz at Hac = 2.50 Oe and Hdc = 0 Oe in the temperature range of 1.8–15 K. When a dc field of 2000 Oe was applied, QTM was evidently suppressed. As shown in Figure , both in-phase (χ′) and out-of-phase (χ″) curves showed a clear frequency dependence, which suggested the SMM behavior of 2.[60] In addition, the Cole–Cole plots of 2 between 2 and 4 K displayed one characteristic magnetic relaxation (Figures b and S8), and the breadth of the distribution of relaxation was analyzed using a generalized Debye model (Table S3 in the SI). These α values (0.186–0.258) were not small, indicating that 2 had a relatively wide distribution of relaxation time. Moreover, the Debye model based on the relationship ln(χ″/χ′) = ln(ωτ0) + Ea/KBT was used to fit the frequency-dependent ac susceptibility data in the range of 650–1599 Hz (Figure a), and the energy barrier of 19.02 K and τ0 value of 1.12 × 10–6 s for 2 were obtained, which were consistent with the expected numbers (τ0 = 10–6–10–11 s) for SMMs. The Ueff/k value of 2 was consistent with that of carboxylic-based DyIII-SMMs [Dy2(L)3(H2O)3]·DMF (Ueff = 24.57K; H2L = 3-(3,5-dicarboxylphenoxy)pyridine).[61]

Figure 5

Plots of the temperature-dependent (a) in-phase (χ′) and (b) out-of-phase (χ″) ac susceptibilities of complex 2 under a 2000 Oe dc field.

Figure 6

(a) Magnetization relaxation time ln τ vs T–1 plots for 2 under a dc field of 2000 Oe (the red solid line is fitted with the Arrhenius law). (b) Cole–Cole plots of 2 measured at 2, 3, and 4 K with a 2000 Oe dc field (the solid line represents the least-squares fitting using CC-FIT software).

Photoluminescent Properties

Ln-based complexes displayed unique and excellent photoluminescent properties, which were caused by the internal electron transitions in the 4f shell of Ln ions.[62,63] Thus, the luminescent properties of Ln-based complexes have attracted considerable attention because of their wide applications in sensors, light-emitting diodes, optical switches, displays, and functional probes in biological systems.[64,65] Herein, the photoluminescent properties of complexes 3 and 4 have been studied in solid state at room temperature.

As shown in Figure a, upon the excitation at 349 nm (Figure S9 in the SI), the emission spectra of 3 in the visible region exhibited five emission peaks at 5D0 → 7F0 (579 nm), 5D0 → 7F1 (593 nm), 5D0 → 7F2 (614 nm), 5D0 → 7F5 (653 nm), and 5D0 → 7F4 (702 nm) with the characteristic transitions of the EuIII ion.[66−68] The literature shows that the 5D0 → 7F2 transitions of the Eu(III) cation were electric dipole transitions being hypersensitive to their local environments, whereas its 5D0 → 7F1 transitions were magnetic dipole transitions being insensitive to the local environment.[69−71] Thus, the intensity ratio of I (5D0 → 7F2) and I (5D0 → 7F1) was often used as a probe to determine the site symmetry of a Eu(III) ion. In 3, the value of I (5D0 → 7F2):I (5D0 → 7F1) was 1.6, indicating that the Eu(III) ion was located on the low-symmetry ligand field. This result was in good agreement with the above-mentioned single-crystal X-ray results, in which the Eu(III) ion of 3 had low-symmetry coordination configuration with a distorted square antiprism. Fluorescence lifetime was measured by monitoring the excitation at 349 nm and emission at 614 nm to study the fluorescence behavior of 3 (Figure b). The decay curve of 3 can be fitted well by a monoexponential function I = I0 exp(−t/τ),[72,73] leading to the luminescence lifetime τ of 1.49 ms. This lifetime value was consistent with those reported of the EuIII complex based on pyridinecarboxylic ligands.[74]

Figure 7

(a) Soild-state emission spectra for 3 at room temperature; the inset is the Commission internationale de l’éclairage (CIE) plot of 3. (b) Decay curves of the EuIII (5D0 → 7F2) complex 3.

As for 4, the characteristic emission spectra in the visible region were obtained upon excitation at 349 nm (Figure S10 in the SI), and they revealed four groups of characteristic single bands at 489, 545, 589, and 622 nm (Figure a), which can be assigned to 5D4 → 7F6, 5D4 → 7F5, 5D4 → 7F4, and 5D4 → 7F3 electron transitions of the TbIII ion, respectively. Moreover, the fluorescence lifetime of 4 in the visible region was examined using the strongest emission (545 nm) and excitation (349 nm) with the decay curve (Figure b). The decay behavior followed the double-exponential function I = A1 exp(−t/τ1) + A2 exp(−t/τ2), leading to the lifetime value of approximately 0.20 ms, which was in comparison with those reported for TbIII complexes.[75]

Figure 8

(a) Soild-state emission spectra for 4 at room temperature; the inset is the CIE plot of 4. (b) Decay curves of the TbIII (5D4 → 7F5) complex 4.

Furthermore, the chromaticity coordinates for 3 and 4 based on the visible fluorescence spectrum are shown in the CIE 1931 diagram, which can reflect the specific emission color of the complexes. As shown in the insets of Figures a and 8a, the CIE chromaticity coordinates (X, Y) for 3 and 4 were (0.6512, 0.3433) and (0.2871, 0.5361), corresponding to the intense red and green emission of EuIII and TbIII ions, respectively. Therefore, the systematic study on the luminescence properties of these EuIII and TbIII species showed that they had strong emissions and long lifetime, indicating that the photoenergy transfer from the PAA1– and phen linker excited state to the excited state of EuIII/TbIII was efficient.

Conclusions

Four lanthanide (GdIII, DyIII, EuIII, and TbIII) coordination molecules featuring dinuclear structures were prepared using 3-pyridylacetic acid ligand and the simplest N,N′ bulky (1,10-phenanthroline) as the auxiliary ligand. In particular, the four complexes were the first lanthanide-based complexes constructed from 3-PAA. The magnet studies showed that the GdIII derivative possessed MCE, and the DyIII derivative displayed a typical SMM behavior. Moreover, the EuIII and TbIII derivatives displayed strong characteristic emission and long lifetime, indicating that the ligands were good luminescent sensitizers to EuIII and TbIII ions. Thus, the four complexes might be a good candidates in the molecular luminescent/magnet fields.

Experimental Section

Materials and Instruments

3-Pyridylacetic acid, 1,10-phenanthroline, and Ln(ClO4)3·6H2O were used as purchased without further purification. Fourier transform infrared (FT-IR) spectra were recorded using a PerkinElmer Spectrum One Spectrometer in the range of 4000–400 cm–1 using KBr pellets as bases. Element analyses of C, H, and N were conducted using an Elementar Vario EL III microanalyzer. Powder X-ray diffraction (PXRD) patterns at room temperature were collected on a Rigaku Miniflex II diffractometer using Mo Kα radiation (λ = 1.540598 Å). Simulated PXRD patterns were obtained from Mercury version 1.4 software (http://www.ccdc.cam.ac.uk/products/mercury). TGA measurements have been performed on polycrystalline samples under air atmosphere with a heating rate of 10 °C min–1 in the temperature range of 25–800 °C. Magnetic susceptibilities were performed on a Quantum Design PPMS model 6000 magnetometer. Photoluminescence analyses were conducted using an Edinburgh FL S920 fluorescence spectrometer.

Preparations of [Ln2(3-PAA)2(μ-Cl)2(phen)4](ClO4)2 [Ln = Gd(1), Dy(2), Eu(3), Tb(4)]

A mixture of Ln(ClO4)3·6H2O (0.5 mmol), 3-PAA·HCl (0.5 mmol), phen (0.5 mmol), Et3N (0.20 mL), and CH3CN (10 mL) with a starting pH of 6 was sealed into a 25 mL Teflon-lined stainless steel container under autogenous pressure, kept at 170 °C for 2 days, and then cooled to room temperature at a rate of 6.4 °C h–1. Sheet single crystals were filtered off from the solution, then washed with CH3CN, and dried in air. Yield: 25% (based on Gd) for 1. Yield: 22% (based on Dy) for 2. Yield: 27% (based on Eu) for 3. Yield: 28% (based on Tb) for 4. IR (KBr, cm–1) for Gd1: 623m, 725s, 851s, 933w, 1090vs, 1300w, 1394s, 1423s, 1515m, 1571vs, 1623s, 3075w, 3405w, 3849w; for Dy2: 621m, 725s, 848s, 933w, 1089vs, 1299w, 1394s, 1423s, 1515m, 1568vs, 1623s, 3080w, 3431w, 3849w; for Eu3: 623m, 725s, 849s, 932w, 1088vs, 1294w, 1391s, 1423s, 1513m, 1569s, 1618s, 3057w, 3359w, 3849w. for Tb4: 623m, 725s, 852s, 936w, 1088vs, 1298w, 1384s, 1423s, 1513m, 1575s, 1613s, 3071w, 3361w, 3849w. Anal. calcd for 1: C, 47.21; H, 2.81; N, 8.88. Found: C, 46.8; H, 2.92; N, 9.09. Anal. calcd for 2: C, 46.90; H, 2.79; N, 8.82. Found: C, 46.12; H, 2.89; N, 9.07. Anal. calcd for 3: C, 47.53; H, 2.83; N, 8.94. Found: C, 47.01; H, 2.93; N, 9.13. Anal. calcd for 4: C, 47.11; H, 2.80; N, 8.86. Found: C, 46.43; H, 2.92; N, 9.08. In the IR spectra of 3-pyridylacetic and 1,10-phenanthroline, and complexes 1–4, the absorption bands at 1400–1600 cm–1 represent the skeletal vibrations of the pyridyl rings, and the broad band at ca. 3300–3450 cm–1 suggests that the O–H stretching of the carboxylic group.

Single-Crystal XRD Structure Determination

Single crystals were mounted on a Bruker SMART APEX CCD diffractometer using graphite-monochromated Mo Kα radiation (=0.71073 Å) at 293 K for complexes 1–4. The structures were solved by direct methods using the Siemens SHELXTL version 5 package[76] and refined by full-matrix least-squares techniques. All nonhydrogen atoms were refined anisotropically. Nonhydrogen atoms were located by difference Fourier maps and subjected to anisotropic refinement. No higher space groups for 1–4 were found using the Platon software from the IUCr website (http://www.iucr.org/). The crystallographic data of 1–4 in CIF format were deposited in the Cambridge Crystallographic Data Center (CCDC nos. 2101132, 2101133, 2101134, and 2101135).