Light Lanthanide Complexes with Crown Ether and Its Aza Derivative

Dec 12, 2016 - Light Lanthanide Complexes with Crown Ether and Its Aza Derivative Which Show Slow Magnetic Relaxation Behaviors. Hisami Wada†, Sayak...
3 downloads 10 Views 3MB Size
Download PDF
Article pubs.acs.org/IC

Light Lanthanide Complexes with Crown Ether and Its Aza Derivative Which Show Slow Magnetic Relaxation Behaviors Hisami Wada,† Sayaka Ooka,† Tomoo Yamamura,‡ and Takashi Kajiwara*,† †

Department of Chemistry, Faculty of Science, Nara Women’s University, Nara 630-8506, Japan Institute for Materials Research, Tohoku University, Aoba-ku, Sendai 980-8577, Japan

‡

S Supporting Information *

ABSTRACT: Two sets of isostructural Ln(III) mononuclear complexes, [Ln(NO3)3(18-crown-6)] (Ln = Ce (1), Pr (2), and Nd (3)) and [Ln(NO3)3(1,10-diaza-18-crown-6)] (Ln = Ce (4), Pr (5), and Nd (6)), were synthesized, and their slow magnetic relaxation behavior was investigated. Since Ln(III) ions are located in an axially stressed ligand field in both sets of complexes, they can exhibit single-molecule magnet (SMM) behavior owing to the oblate-type electronic distributions of the ground sublevels found in Ce(III), Pr(III), and Nd(III). Field-induced slow magnetic relaxation was observed for Ce(III) and Nd(III) complexes 1, 3, 4, and 6 under an applied bias dc field of 1000 Oe, whereas no slow relaxation was observed for Pr(III) complexes 2 and 5. The slow magnetic relaxation behavior of 1, 3, 4, and 6 was correlated with the even-numbered Jz sublevels of Ce(III) and Nd(III) ions, known as the Kramers system.

■

combination of LS states with (Lz, Sz) = (3, −1/2) and (2, 1/ 2), which leads to an oblate-shaped electronic distribution of this Jz sublevel due to orbitals with lz = 3 and 2. On the contrary, the Jz = 1/2 sublevel has a prolate-shaped electronic distribution, since this sublevel can be represented by the linear combination of (Lz, Sz) = (1, −1/2) and (0, +1/2) LS states. This simple assumption was confirmed by the theoretical calculations of Schmitt et al.10 and Long et al.11 If the oblatetype Ce(III) ion is placed in an axially stressed crystal field, the Jz sublevels possessing a larger electronic distribution along the z-axis are destabilized more than the oblate-shaped |Jz| = 5/2 sublevels, and hence, easy-axis magnetic anisotropy is realized. The situation is more complicated for the multi f-electron Pr(III) and Nd(III) ions; however, the synthetic strategy for the easy-axis anisotropy is very similar, since both the above ions are also oblate-type. To produce an appropriate anisotropic crystal field, we have focused on the coordination of 18-crown-6 to Ln(III) ions. This neutral ligand can occupy the equatorial positions around Ln(III) ions, and the vacant axial positions can accept the coordination by small anionic ligands. Previously, the syntheses and crystal structures of [Ln(NO3)3(18-crown-6)] (Ln = Nd)12 and [Ln(NO3)3(1,10diaza-18-crown-6)] (Ln = La)13 complexes have been reported. In this paper, we report the syntheses and magnetic properties of two isostructural complex families of [Ln(NO3)3(18-crown6)] (Ln = Ce (1), Pr (2), and Nd (3)) and [Ln(NO3)3(1,10diaza-18-crown-6)] (Ln = Ce (4), Pr (5), and Nd (6)), with

INTRODUCTION Single-molecule magnets (SMMs)1,2 are one of the most fascinating molecule-based nanomaterials. They are characterized by slow magnetization at low temperature, which originates from a large magnetic momentum as well as a large easy-axis magnetic anisotropy. Due to the unquenched orbital angular momentum, lanthanide(III) ions possess a larger magnetic momentum than those of the transition metal ions, which is correlated with the total angular momentum quantum number J. Here, the J values are defined as L − S for light Ln(III) ions and L + S for heavy Ln(III) ions, where S denotes the spin angular momentum and L denotes the orbital angular momentum. The J values expected for Ce3+ (L = 3, S = 1/2), Nd3+ (L = 6, S = 3/2), Tb3+ (L = 3, S = 3), and Dy3+ (L = 5, S = 5/2) are 5/2, 9/2, 6, and 15/2, respectively. For both the light3−5 and the heavy6−9 lanthanides, the advantageous contribution of orbital angular momentum leads to a strong magnetic anisotropy when the Ln(III) ion is placed in an anisotropic ligand field. Ce(III) has the simplest f1 configuration with a 2F5/2 ground state. When the Ce(III) ion is placed in an axially stressed ligand field, where the equatorial electronic repulsion is smaller than the axial one, the six-fold degenerate ground state is split into three sets of Kramers pairs, characterized by the projection of the J vector on the principle axis, |Jz| = 5/2, 3/2, and 1/2. Due to the different distribution of the orbital angular momentum Lz, each Jz state has a differently shaped electronic distribution, and the energy order of |Jz| sublevels can be roughly predicted by considering each electronic configuration. For example, the electronic configuration of the Jz = 5/2 sublevel is represented as a linear © XXXX American Chemical Society

Received: July 22, 2016

A

DOI: 10.1021/acs.inorgchem.6b01764 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry slow magnetic relaxation observed for the Ce(III) and Nd(III) complexes 1, 3, 4, and 6.

■

Table 1. Crystallographic Data for Complexes 1 and 4 formula FW crystal system space group a/Å b/Å c/Å α/deg β/deg γ/deg V/Å3 Z D/g cm−3 T/K wavelength/Å F(000) μ/mm−1 θ range/deg R1, wR2 (I > 2σ(I)) R1, wR2 (all data)

EXPERIMENTAL SECTION

All chemicals and reagents were of reagent grade and were used without further purification. All chemical reactions and sample preparations for physical measurements were performed in ambient atmosphere. Instrumentation. Variable temperature magnetic susceptibility measurements were performed on MPMS-5S and PPMS-9 magnetometers (Quantum Design). Diamagnetic corrections for each sample were applied using Pascal’s constants. Elemental analysis was carried out with the help of the Research and Analytical Center for Giant Molecules, Graduate School of Science, Tohoku University. Synthetic Procedures. Synthesis of [Ce(NO3)3(18-crown-6)] (1). Solutions of Ce(NO3)3·6H2O (0.5 mL, 0.2 M) and 18-crown-6 (0.5 mL, 0.2 M) in a mixture of MeCN and MeOH (3:1 v/v) were reacted at room temperature. Colorless crystals suitable for X-ray crystallography were obtained after several hours (45 mg, 76%). Elemental Anal. Calcd for 1: C, 24.41; H, 4.10; N, 7.12. Found: C, 24.62; H, 4.19; N, 7.22. Synthesis of [Pr(NO3)3(18-crown-6)] (2) and [Nd(NO3)3(18-crown6)] (3). Complexes 2 and 3 were obtained by the procedure described above, using Pr(NO3)3·6H2O and Nd(NO3)3·6H2O as starting materials for 2 (34 mg, 57%) and 3 (36 mg, 61%), respectively. Elemental Anal. Calcd for 2: C, 24.38; H, 4.09; N, 7.11. Found: C, 24.40; H, 4.22; N, 7.20. Elemental Anal. Calcd for 3: C, 24.24; H, 4.07; N, 7.07. Found: C, 24.31; H, 4.21; N, 7.14. Synthesis of [Ce(NO3)3(1,10-diaza-18-crown-6)] (4). Methanolic solutions of Ce(NO3)3·6H2O (0.5 mL, 0.1 M) and 1,10-diaza-18crown-6 (0.5 mL, 0.1 M) were reacted at room temperature and left to stand for some time. Soon, small needle-shaped microcrystals of [Ce(NO3)3(1,10-diaza-18-crown-6)]·MeOH were formed, with their composition and structure confirmed by X-ray analysis. The suspension was left to stand at room temperature for a week, during which the needle-shaped crystals disappeared, and colorless octagonal platelike crystals of 4 were formed (26 mg, 88%). Elemental Anal. Calcd for 4: C, 24.49; H, 4.45; N, 11.90. Found: C, 24.60; H, 4.54; N, 11.90. Synthesis of [Pr(NO3)3(1,10-diaza-18-crown-6)] (5) and [Nd(NO3)3(1,10-diaza-18-crown-6)] (6). Complexes 5 and 6 were obtained by the procedure described above, with Pr(NO3)3·6H2O and Nd(NO3)3·6H2O used as starting materials for 5 (16 mg, 55%) and 6 (21 mg, 70%), respectively. However, a longer time (1 or 2 months) was required to form the octagonal platelike crystals. Elemental Anal. Calcd for 5: C, 24.46; H, 4.45; N, 11.88. Found: C, 24.31; H, 4.55; N, 11.68. Elemental Anal. Calcd for 6: C, 24.32; H, 4.42; N, 11.82. Found: C, 24.08; H, 4.46; N, 11.55. Crystallography. X-ray data for 1−6 were collected at low temperature (153 K) on a Rigaku Varimax Saturn area detector diffractometer using confocal monochromated Mo Kα radiation. The intensity data were corrected for absorption using an empirical method included in the Crystal Clear software.14a The structures were solved by direct methods with SIR-97,14b and structure refinement was carried out using the full-matrix least-squares method on SHELXL97.14c Non-hydrogen atoms were anisotropically refined, and the hydrogen atoms were treated using the riding model. Tables 1 and S1 list the crystallographic data together with the R1 and wR2 values.

1

4

C12H24N3O15Ce 590.49 orthorhombic Pbca 12.1658(11) 15.5222(15) 21.704(2)

C12H26N5O13Ce 588.50 triclinic P1̅ 8.8581(10) 10.2562(13) 12.6506(16) 83.787(5) 76.197(5) 67.095(3) 1028.0(2) 2 1.901 153(2) 0.71075 590 2.291 2.56−30.00 0.025, 0.061 0.026, 0.062

4098.5(7) 8 1.914 153(2) 0.71075 2360 2.303 1.88−30.00 0.042, 0.087 0.048, 0.091

(featuring oblate-type lanthanides such as Ce, Nd, and Pr) by the reaction of equimolar amounts of lanthanide nitrates and 18-crown-6 ether in a mixture of CH3CN/CH3OH. In the case of [Ln(NO3)3(1,10-diaza-18-crown-6)], using CH3OH as a solvent provided the best yield and higher quality crystals. The reaction of lanthanide nitrates and 1,10-diaza-18-crown-6 in methanol first gave small needles of [Ln(NO3)3(1,10-diaza-18crown-6)]·MeOH. This complex was slowly converted into platelike crystals of solvent-free [Ln(NO3)3(1,10-diaza-18crown-6)] in the reaction solution. Since the solvated crystals lose methanol under ambient pressure, nonsolvated crystals were utilized for physical measurements. Figure 1 shows the crystal structures of 1 and 4. In both complexes, the macrocyclic ligand occupies equatorial positions and is slightly bent, with one and two chelating nitrate anions coordinating the Ce(III) ion above and below, respectively. The Ce(III) ion is in a dodeca-coordination and is sandwiched by three negatively charged nitrate anions. The dihedral angles of bent macrocyclic ligands were defined by two coordination planes (O15/O10/O11/O12 and O12/O13/O14/O15 for 1− 3 and N4/O10/O11/N5 and N5/O12/O13/N4 for 4−6) and were estimated as 37.94(8), 37.98(7), and 38.15(6)° for the former and 49.97(6), 49.85(6), and 49.75(6)° for the latter complexes. To compare the coordination structure around Ce(III) in 1 and 4 in detail, the relative positions of donor atoms from the equatorial plane (defined as the plane parallel to the least-squares plane of six donor atoms of the macrocyclic ligand) were estimated (Figure 2). The corresponding positions of the nitrate donor oxygen atoms in crown and aza-crown ether complexes are almost superimposable, although the donor atoms of the macrocyclic ligands showed different deviations from the equatorial plane. Hence, we can discuss the difference in magnetic properties between two families from the structural viewpoint of different macrocyclic ligands. The bond distances of anionic nitrate oxygens are slightly shorter (2.605(2)− 2.643(2) Å for 1, 2.6265(15)−2.6758(16) Å for 4) than those of the neutral macrocyclic ligand (2.619(2)−2.780(2) Å for 1, 2.6725(17)−2.8116(15) Å for 4), which enhances the ligand field anisotropy by applying stress above and below the Ln(III)

■

RESULTS AND DISCUSSION Syntheses and Crystal Structures. Crown ether based lanthanide complexes [Ln(NO3)3(15-crown-5)] (Ln = La, Pr, and Eu) and [Ln(NO3)3(18-crown-6)] (Ln = La, Pr, and Nd) were initially synthesized by Bünzli et al.,12,15 Benetollo et al.,12,16 and Backer-Dirks et al.17 for the investigation of their crystal structures and luminescent properties. We have focused on the Ln(III) coordination structures of this family and synthesized high-quality crystalline light lanthanide complexes B

DOI: 10.1021/acs.inorgchem.6b01764 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 1. ORTEP drawings of 1 (left) and 4 (right) at 50% probability. Hydrogen atoms are omitted for clarity.

Figure 3. Temperature dependence of the products of molar susceptibility and temperature, χMT, measured under a 1000 Oe dc field. (Right) plots for the whole temperature range up to 300 K and (left) a magnified view below 16 K, where slow magnetic relaxation is discussed.

Figure 2. (Left) Coordinate structure in 1 and (Right) a schematic relative location of donor atoms around the Ce(III) ion in 1 and 4. In the schematic drawing, Ce(III) is located at the origin, and the abscissa axis represents the equatorial plane. The locations of oxygen donor atoms in 1 are represented as open circles, whereas the oxygen (red) and nitrogen (blue) donor atoms in 4 are represented as closed circles. Dotted circles denote coordination radii of 2.6 and 2.8 Å, respectively.

For the Pr(III) complexes, the χMT values continuously decreased as the sample was cooled, approaching the diamagnetic state at the low temperature limit. For Ce(III) and Nd(III) complexes, the temperature dependence of χMT was rather small, possibly reaching finite values at the low temperature limit. A strong temperature dependence observed for Pr(III) complexes may be attributed to the thermal depopulation of Jz sublevels, which exhibit a small splitting under the anisotropic ligand field that causes thermal depopulation phenomena. In the case of Ce(III) and Nd(III) complexes, the separation between the ground and excited Jz sublevels is large due to an appropriate anisotropic ligand field. As a result, the ground sublevels of Ce(III) in 1 and 4 and Nd(III) in 3 and 6 with large |Jz| are sufficiently separated from the other excited sublevels and are mainly occupied at low temperatures (below 10 K). To confirm the presence of slow magnetic relaxation properties, the dynamic susceptibility data were initially measured under zero dc field condition for all complexes (Figure S2), which did not exhibit any out-of-phase components. The Curie-like behavior of Ce(III) and Nd(III) complexes were confirmed by the χM′T values, which were almost independent of the temperature in the range studied. On the contrary, the χM′T values of 2 and 5 showed a strong temperature dependence, decreasing linearly with decreasing temperature and approaching 0 emu mol−1 K. These observations were consistent with those of dc susceptibility and temperature discussed above.

ions. The shortest Ce···Ce distances were estimated to be 7.6968(6) Å for 1 (symmetry code: 1 − x, −y, 1 − z) and 7.2964(8) Å for 2 (symmetry code: 1 − x, 1 − y, 1 − z), respectively. The packing diagrams of the molecules in the unit cell are given as Figure S1. Magnetic Properties. dc susceptibility data were recorded at variable temperature for all complexes under a dc field of 1000 Oe. Figure 3 shows the temperature dependence of χMT products. The Curie constants were estimated as 0.80 emu K mol−1 for Ce(III) (J = 5/2, g = 6/7), 1.60 emu K mol−1 for Pr(III) (J = 4, g = 4/5), and 1.64 emu K mol−1 for Nd(III) (J = 9/2, g = 8/11). The observed χMT values at 300 K were 0.76 and 0.68 emu K mol−1 for Ce(III) complexes 1 and 4, 1.52 and 1.48 emu K mol−1 for Pr(III) complexes 2 and 5, and 1.41 and 1.60 emu K mol−1 for Nd(III) complexes 3 and 6, respectively. All values were close to, but slightly smaller than, the expected values, which may reflect the presence of magnetic anisotropy. Ce(III) complexes exhibited only a slight temperature dependence in the whole temperature range, whereas Pr(III) and Nd(III) complexes showed a stronger temperature dependence. However, below 16 K (where the ac susceptibility is discussed), Pr(III) complexes showed a different temperature dependence from those of Ce(III) and Nd(III) complexes (Figure 3, left). C

DOI: 10.1021/acs.inorgchem.6b01764 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 4. Frequency dependence of χM′T (closed circles) and χM″T (open circles) of 1, 3, 4, and 6 measured under an oscillating field of 3 Oe and a dc field of 1000 Oe. Solid curves represent the fits to a generalized Debye model, while the dotted curves are visual guides.

Under a dc bias field of 1000 Oe, 2 and 5 retained their temperature dependent features (Figure S3). On the contrary, Ce(III) and Nd(III) complexes exhibited out-of-phase signals being observed at temperatures of up to 7 K (1), 3.8 K (3), 6.5 K (4), and 6 K (6) at an ac frequency of up to 10 000 Hz (Figure 4). This observation was indicative of slow magnetic relaxation of these four complexes, which obeyed Debye relaxation. The frequency dependences of both the in-phase and out-of-phase signals were analyzed using the generalized Debye equations to estimate four parameters such as isothermal and adiabatic susceptibilities, χT and χS, and relaxation time and its distribution, τ and α (the four parameters are summarized in

Tables S2−S5); the corresponding Cole−Cole plots18 showed a semicircular shape (Figure 5). The estimated α values were sufficiently small, indicating that slow magnetic relaxation occurs via a single process, and hence, the Ce(III) and Nd(III) complexes are considered to be field-induced SMMs. The values of χTT were almost constant in these temperature ranges, and the isothermal susceptibility χT obeyed the Curie law, which is consistent with the dc and zero-field ac susceptibility data. To investigate the slow magnetic relaxation in detail, the bias field dependence of the slow relaxation in 1 was first revealed at a field range of 0−5000 Oe for a temperature range of 3−6 K D

DOI: 10.1021/acs.inorgchem.6b01764 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 5. Cole−Cole plots of 1, 3, 4, and 6 measured under an oscillating field of 3 Oe and a dc field of 1000 Oe. Solid curves represent the fits to a generalized Debye model, while the dotted curves are visual guides.

Figure 6. (Left) Bias field dependence of χM′T (top, closed circles) and χM″T (bottom, open circles) of 1 measured at 3.0 K. Solid curves represent theoretical calculations on the basis of the generalized Debye model. (Right) Bias field dependence of the relaxation rate τ−1 of 1 measured at the temperature range of 3.0−6.0 K.

with the decrease of χST. During this process, the peaks of the out-of-phase signals were slightly shifted into a slower frequency region, indicating that the magnetization relaxation under these conditions occurs via quantum tunneling relaxation. This slow relaxation frequency was maintained up to 1500 Oe, becoming slightly faster at fields higher than 2000 Oe. The enhancement of magnetization flipping at higher dc fields indicates that relaxation occurs via a direct process. All data were well reproduced using generalized Debye equations

(Figures 6 and S4). Under zero-field condition, the relaxation via quantum tunneling is predominant, which is suppressed by the application of a dc bias field. When the dc field was applied, the amplitude of susceptibility was slightly increased, accompanied by an appearance of frequency-dependent outof-phase signals (Figure 6, left), which is enhanced upon the increase of the dc field. In the low-field region, it increased when a field of up to 300 Oe was applied and then reached a constant value above 500 Oe. This occurred concomitantly E

DOI: 10.1021/acs.inorgchem.6b01764 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry

Figure 7. Arrhenius plots for complexes 1, 3, 4, and 6 measured under a 1000 Oe bias field. Solid lines and dotted curves represent theoretical fitting taking into account the TA-QTM process (black), the TA-QTM and the Raman processes with n = 5 (blue), and the TA-QTM and the Raman processes with n = 9 (red).

Oe bias field were used to eliminate the effect of the dc field (Figure 7). A slightly bent Arrhenius plot was obtained over the entire temperature range being indicative of the presence of several relaxation processes such as Raman and/or TA-QTM processes. The data were first analyzed using the linear Arrhenius equation (τ = τ0 exp(ΔE/kBT)) for the data above 4 K, which gave the values of ΔE/kB = 31.4(4) K and τ0 = 1.71(13) × 10−7 s. The whole data was then analyzed using eq 2, which considers both Raman and TA-QTM processes.

with small α values (