Chains: Joint Magneto-Luminescence and ab Initio ... - ACS Publications

Jun 20, 2017 - Department of Chemistry, Brock University, 1812 Sir Isaac Brock Way, ... Institute of Materials, University of Aveiro, Aveiro 3810-193,...
1 downloads 0 Views 3MB Size
Download PDF
Article pubs.acs.org/IC

Dual-Property Supramolecular H‑Bonded 15-Crown‑5 Ln(III) Chains: Joint Magneto-Luminescence and ab Initio Studies Majeda Al Hareri,† Zineb Ras Ali,† Jeffery Regier,† Emma L. Gavey,† Luis D. Carlos,‡ Rute A. S. Ferreira,‡ and Melanie Pilkington*,† †

Department of Chemistry, Brock University, 1812 Sir Isaac Brock Way, St Catharines, Ontario L2S 3A1, Canada Physics Department and CICECOAveiro Institute of Materials, University of Aveiro, Aveiro 3810-193, Portugal

‡

S Supporting Information *

ABSTRACT: Two complexes comprising 9-coordinate capped square antiprismatic [Ln(NO3)3(OH2)2(MeOH)] units [Ln(III) = Dy 6; Tb 7] are reported in which the metal complexes are hydrogen-bonded to 15C5 (15-crown-5) macrocycles to form supramolecular chains, {[Ln(NO3)3(OH2)2(MeOH)]·(15C5)}n. Alternating current magnetic susceptibility measurements supported by ab initio studies show field-induced SMM (singlemolecule magnet) behavior for 6, but rapid relaxation of the magnetization for 7 because of the presence of dominant quantum tunneling processes as evidenced by the presence of a significant calculated tunnel splitting within the ground-state multiplet. Modeling the high-resolution emission spectra for 6 afforded energies of 37 ± 5 and 28 ± 5 cm−1 for the first-excited-state Stark sublevels of the two crystallographically independent Dy1 and Dy2 ions, in excellent agreement with the calculated values of 31 and 21 cm−1 for ΔE1 derived from ab initio studies.

■

INTRODUCTION Single-molecule magnets (SMMs) are coordination complexes composed of paramagnetic metal ions that are magnetized in the presence of an applied magnetic field and exhibit slow relaxation of the magnetization when this field is removed, below a certain blocking temperature TB.1 For transition metals, the energy barrier to relaxation Ueff in these systems is related to S, the total spin on the molecule, and D, the axial zero-field splitting parameter, such that Ueff = |D|S2 and Ueff = |D|(S2 − 1 /4) for integer and noninteger spins, respectively.2 Conversely, for lanthanides, the energy barrier arises from crystal-field splitting of the MJ states associated with the ground-state 2S+1ΓJ ground term. This energy barrier is therefore sensitive to the symmetry of the coordination environment and the nature of the ligands. In both lanthanide- and transition-metal-SMMs, an energy barrier serves to trap the magnetic moment in one direction, conferring magnetic hysteresis or a molecular memory on the system that can be potentially exploited for applications that include spin valves,3 transistors,4 quibits,5 and data storage.6 Magnetic relaxation in 3d-SMMs typically takes place either via a classical mechanism, where the spins relax back thermally over the energy barrier, or via quantum tunneling of magnetization (QTM), where electrons in degenerate microstates tunnel directly through the energy barrier.7 For the 4f-based systems, relaxation can occur via thermal population of excited mJ states whose anisotropy axes are not coincident with the ground state, or through more complex Orbach and/or Raman processes. Both lanthanideand transition-metal-SMMs exhibit superparamagnet-like prop© XXXX American Chemical Society

erties that are characterized by the presence of a frequencydependence of the out-of-phase component of the ac magnetic susceptibility, as well as a magnetic hysteresis.8 The first family of SMMs discovered were 3d-based polynuclear clusters, the most well-known of which is Mn12Ac, reported by Christou et al. in the early 1990s.9 Two decades later, the high-J ground state and large intrinsic anisotropy of Ln(III) ions are currently being exploited for the discovery of a second family of 4f-based SMMs, the first of which were [LnPc2]− complexes 1, comprising a Ln(III) center sandwiched between two phthalocyanine macrocycles, reported in 2003 by Ishikawa et al., Figure 1.10 For 4f-SMMs, their ground-state stability comes from the crystal-field splitting of the (2J + 1)mJ microstates within the

Figure 1. Molecular structures of the D4d sandwich complexes [Ln(Pc)2]− (1)10 and [Sm(15C5)2]3+ (2).16 Received: January 11, 2017

A

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry spin−orbit coupled 2S+1LJ ground term which can afford incredibly large magnetic anisotropy. Dy(III) is the most commonly employed f-block element since, as a Kramers ion, its ground state is always bistable, irrespective of the ligand field symmetry. In contrast, Tb(III), a non-Kramers ion, requires an appropriate axial ligand field to ensure a bistable ground state, and thus Tb(III)-SMMs are less common in the chemical literature.11 In these single-ion systems, magnetic relaxation typically occurs via an Orbach process involving excited ligand field states; however, in the presence of a transverse magnetic field, or in cases where uniaxiality cannot be achieved, QTM becomes dominant, severely compromising the SMM properties.12 In recent years, synthetic strategies to develop Ln-based SMMs with large energy barriers have included the following: (i) the introduction of very strong magnetic axiality by targeting the preparation of low-coordinate linear complexes12 and (ii) the development of complexes with high ligand field symmetries.13 Both approaches have served to enhance the intrinsic anisotropy of 4f-systems, providing a pathway to minimize QTM, affording SMMs with some of the largest energy barriers reported to date.10,14 Crown ethers first discovered by Pederson in the 1960s are a commercially available class of macrocyclic ligands in which both the cavity size and number of donor atoms can be structurally tuned to accommodate a broad range of metal cations.15 In addition to realizing the relationship between cavity diameter and cationic radius, Pederson established that crown ethers with smaller cavities such as 15C5 and 12C4 accommodate larger metal cations by forming sandwich-type topologies, e.g., 2, where the cation is displaced out of the plane of the macrocycle, reminiscent of the double-decker [LnPc2]− complex 1 shown in Figure 1.16 In addition to their applications in supramolecular host−guest chemistry, the photoluminescence properties of Ln(III) crown ether complexes have also been exploited for the development of luminescent probes.17 It occurred to us therefore that the emission spectra of Ln(III) crown ether complexes might also provide information regarding the Stark splitting of the ground-state 2S+1LJ multiplet of the 4f-ion into its component mJ microstates or sublevels.18 In theory, the energies of the aforementioned microstates can be gleaned spectroscopically via a Gaussian fit of the fine structure of the highest-energy emission band; however, in practice this is quite challenging, and only a handful of Ln(III) complexes with well-resolved emissive bands have been successfully fit to date.18−21 Given the abilities of the smaller crown ether ligands to coordinate metal cations in sandwich-type topologies, we proposed that the range of available crown ether macrocycles might afford Ln-SMMs, through which the effects of structural variation on magnetic properties could be studied in a systematic manner. We therefore set out to investigate their coordination chemistry with Ln(III) ions and to investigate the structural and magneto-optical properties of the resulting complexes. In our first study, we employed 15C5 and 12C4 macrocycles together with Dy(ClO4)3 for the preparation of two mononuclear complexes: the half-sandwich complex 3 and the pseudosandwich complex, 4 (Figure 2).20 Magneto-optical and theoretical studies reveal that, despite their comparable O9 coordination spheres, the studied complexes have remarkably different magnetic properties with only 4 displaying SMM behavior in zero field.20 Introduction of a benzo-substituent into the framework of the 15C5 macrocycle afforded a completely different structural topology in which

Figure 2. Half-sandwich complex of 12C4 with Dy(III) (3),20 pseudosandwich complex of 15C5 with Dy(III) (4),20 and the encapsulated [Dy(OH2)8]3+ cation (5).21

three uncoordinated benzo-15C5 macrocycles encapsulate [Dy(OH2)8]3+ cation 5 with close-to-ideal D4d geometry that exhibits both SMM and photoluminescence properties (Figure 2).21 We report herein detailed magneto-structural, optical, and theoretical studies of two new 4f-15C5 complexes with 1-D structural topologies.

■

EXPERIMENTAL SECTION

Synthesis. All reagents were supplied by Sigma-Aldrich and used without further purification. Preparation of {[Dy(NO3)3(H2O)2(CH3OH)]·(15C5)}n·0.5H2O (6). A sample of 15-crown-5 (655 mg, 1.04 mmol) was added in one portion to a solution of Dy(NO)3·5H2O (655 mg, 1.04 mmol) in MeOH/ MeCN (1:3, 6 mL). The solution was heated to 55 °C and vigorously stirred for 3 h. After being cooled to room temperature, the solution was filtered through a plug of cotton wool and left to slowly evaporate at room temperature. After 3 weeks, single crystals of 6 were isolated as colorless plates (87 mg, 13%). IR (cm−1): 3215 (br), 2921 (w), 2883 (w), 1655 (w), 1453 (m), 1295 (s), 1268 (s), 1088 (s), 943 (m), 840 (w), 814 (w), 741 (m), 656 (w). UV−vis (MeOH, nm): λmax = 270 (ε = 1080 L mol−1 cm−1). FAB-MS m/z: 598 [M − 2H2O]+. Anal. Calcd for C11H29O17.5N3Dy: C, 20.46; H, 4.30; N, 6.51%. Found: C, 20.19; H, 4.30; N, 6.51%. Preparation of {[Tb(NO3)3(H2O)2(CH3OH)]·(15C5)}n·2H2O (7). Tb(NO3)3·5H2O (469 mg, 1.08 mmol) and 15-crown-5 (214 μL, 1.08 mmol) were vigorously stirred in CH3OH/CH3CN (1:3, 6 mL) at 55 °C for 3 h. After being cooled to room temperature, the reaction mixture was filtered through a plug of cotton wool, and the solvent was evaporated to yield a viscous yellow solid. The solid was redissolved in a 1:3 mixture of MeOH/CH3CN and left to crystallize via slow evaporation to yield, after 3 weeks, light brown single crystals of 7 suitable for X-ray diffraction (318 mg, 44%). IR (cm−1): 3209 (br), 2921 (w), 2883 (w), 1656 (w), 1458 (m), 1295 (s), 1267 (s), 1090 (s), 1022 (m), 943 (s), 840 (w), 813 (w), 741 (m), 649 (w). UV−vis (MeOH, nm): λmax = 208 (ε = 7050 L mol−1 cm−1). FAB-MS m/z: 508 [M − 2H 2 O − CH 3 OH − NO 3 − ] + . Anal. Calcd for C11H28N3O17Tb: C, 19.48; H, 4.90; N, 6.19%. Found: C, 19.47; H, 4.55; N, 6.27%. B

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Physical Methods. Infrared spectra were measured on a Bruker Alpha FT-IR spectrometer. UV−vis data were collected on a Beckman Coulter DU 720 General-Purpose UV−vis spectrophotometer. Mass spectrometry data were measured on a Carlo Erba/Kratos GC/MS acquisition system and processed using a SPARC workstation. Elemental analysis data were collected by Atlantic Microlab. X-ray Crystallography. Crystals of 6 and 7 were grown via the slow evaporation of a 3:1 mixture of MeOH/H2O after 3 weeks. Suitable single crystals were mounted on a cryoloop with paratone oil and measured on a Bruker APEX-II CCD X-ray diffractometer equipped with an Oxford Cryoflex device. X-ray data were collected at 150(2) K with Mo Kα radiation (λ = 0.710 73 Å) using the Bruker software.22 SAINT was used for cell refinement and data-reduction. A multiscan absorption correction was applied to the data using the SADABS program.23 The structures of 6 and 7 were solved by direct methods (SHELXS-97)24 and refined using SHELXL-2014 in the Bruker SHELXTL suite.24 Hydrogen atoms were added at calculated positions and refined with a riding model. For complex 6, disordered solvent was removed using the SQUEEZE procedure in PLATON.25 Bond lengths and angles for 6 and 7 are summarized on page S3 in the Supporting Information. Complexes 6 and 7 have been assigned the following CCDC deposition numbers: CCDC 1519141−1519142, respectively. Photoluminescence Spectroscopy. Photoluminescence data for 6 and 7 were measured at 300 and 14 K, respectively, on a modular double-grating excitation spectrofluorimeter equipped with a TRIAX 320 emission monochromator (Fluorolog-3, Horiba Scientific) coupled to an R928 Hamamatsu photomultiplier, using a front-face acquisition mode. A 450 W Xe arc lamp was used as the excitation source. The excitation spectra were corrected for the spectral distribution of the lamp intensity with a photodiode reference detector, and the emission spectra were corrected for the detection and optical spectral response of the spectrofluorimeter. The emission decay curves for 7 at 300 K were measured as previously described for the luminescence spectra using a pulsed Xe−Hg lamp (6 ms pulse at half width and 20−30 × 10−6 s tail). For compound 7, the decays were acquired on a Fluorolog TCSPC spectrofluorimeter (Horiba Scientific) coupled to a TBX-04 photomultiplier tube module (950 V), with 200 ns time-to-amplitude converter and 70 ns delay; the excitation source was a Horiba Jobin Yvon pulsed diode (NanoLED390, peak at 390 nm, 1.2 ns pulse duration, 1 MHz repetition rate, and 150 ns synchronization delay). The absolute emission quantum yields were measured at 300 K using a Hamamatsu (C9920-02) quantum yield measurement system equipped with a 150 W xenon lamp coupled to a monochromator for wavelength discrimination, an integrating sphere as the sample chamber, and a multichannel analyzer for signal detection. Data was collected for each sample in triplicate, and the average value is reported to an accuracy of within 10%. Magnetic Susceptibility. Variable-temperature magnetic measurements were carried out on freshly prepared crystalline samples of 6 and 7. The samples were encapsulated in a diamagnetic gelatin sample holder. Pascal’s constants were used to estimate the diamagnetic corrections, which were subtracted from the experimental susceptibilities to afford the molar paramagnetic susceptibility (χM). Direct current susceptibility data were collected on a Quantum Design SQUID MPMS magnetometer in an applied field of 0.1 T, from 5 to 300 K. Alternating current measurements were made using a Quantum Design PPMS magnetometer. There was one set of ac measurements made at 2 K with frequencies in the range 10−1000 Hz with static fields from 0 to 5000 Oe with a 5 Oe oscillating field. A second set of ac measurements used 13 frequencies between 50 and 10 000 Hz below 10 K and 7 frequencies in the range 10−15 K with static applied fields between 0 and 0.2 T and a 3.5 Oe oscillating field. Theoretical Calculations. Ab initio calculations of the CASSCF + RASSI/SINGLE_ANISO type were performed on 6 and 7 using the MOLCAS quantum chemistry package.26 The calculations were performed using coordinates determined from single-crystal X-ray diffraction without any further geometry optimization. For both complexes, the structural model that best describes the coordination sphere of the Ln(III) ions comprises a [Ln(III)(NO3)3(OH2)-

(MeOH)] core plus the two H-bonded uncoordinated 15C5 macrocycles (model 1, page S14 in the SI). For 6, because of the inversion symmetry, one Dy(III) center was calculated while the other magnetic center was replaced with a diamagnetic Lu(III) ion. The complete active space approach was used where the active space was chosen to include the nine electrons of the seven 4f orbitals. Relativistic basis sets from the ANO-RCC library were used exclusively for all calculations. For a determination of whether or not the choice of basis sets influences the calculations, two types of basis sets (long and short) were used as summarized on page S15 in the SI. Strong, spin− orbit coupling was included in the CASSI procedure which uses the spin-free eigenstates of the CASSCF procedure as the elements in the state interaction determinant. In the CASSCF procedure, 21 roots were used for the sextets, and 224 roots were used for the quartets. The doublet configurations were omitted because of limited computer resources. In the state interaction procedure, strong spin−orbit coupling was introduced into the calculation in the RASSI module where 21 sextet and 128 quartet spin-free states were mixed. The doublet states were omitted because of limited computer resources. Local magnetic properties were calculated in the SINGLE_ANISO module which uses the resulting spin−orbit multiplets from the RASSI procedure. The g-tensors of each of the Kramers doublets (KDs) were calculated using the S = 1/2 pseudospin formalism. The results of these calculations are also summarized in the SI.

■

RESULTS AND DISCUSSION Synthesis and Structural Studies. Following our previous studies,20,21 we set out to further investigate the coordination chemistry of 15C5 macrocycles with f-block ions, by varying the counterion of the metal salt. In this context, the reaction of 1 equiv of 15C5 and Ln(NO3)3 in a 1:3 mixture of methanol and acetonitrile afforded single crystals of H-bonded chains of formula {[Ln(NO3)3(H2O)2(CH3OH)]·(15C5)}n, where Ln(III) = Dy 6 and Tb 7. Both complexes crystallize in the monoclinic space group P21/c, with similar unit cells, but are not isostructural (page S2 in the SI). The asymmetric unit of 6 comprises two independent, 9-cooordinate [Dy(NO3)3(H2O)(CH3OH)] units that are H-bonded to the oxygen atoms of two crystallographically independent 15C5 macrocycles affording a 1-D chain topology (Figure 3). For the Dy1 ion,

Figure 3. Molecular structure of the asymmetric unit of {[Dy(NO3)3(H2O)2(CH3OH)]·(15C5)}n (6) with an appropriate atomic numbering scheme. H bonds are shown as blue ---. Only selected H atoms are shown.

intermolecular H-bonding interactions between the O10 coordinated water molecule and O25 and O29 of a 15C5 macrocycle are 1.855 and 1.838 Å, and those between the O11 coordinated water molecule and O32 and O34 of a second 15C5 macrocycle are 1.858 and 1.845 Å. For the second Dy2 ion, the water O22 is H-bonded to O25 and O28 (1.903 and 1.797 Å) of one crown, and O23 is H-bonded to O30 and O33 of the second crown (1.830 and 1.833 Å). In addition, the H C

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry atoms of the two coordinated methanol ligands are also Hbonded to an oxygen atom of a neighboring 15C5 macrocycle in the chain such that O12−H12···O26 = 2.356 Å and O24− H24···O31 = 2.190 Å. The Dy−O bond distances are in the range 2.329(4)−2.480(4) Å for Dy1 and 2.343(4)−2.453(4) Å for Dy2. Shape analyses27 of the two Dy(III) centers reveal that they are closest to a 9-coordinate, capped square antiprismatic geometry with continuous shape measures (CSMs) of 1.81 and 1.96 for Dy1 and Dy2, respectively (page S5 in the SI). Detailed analysis of the crystal packing reveals that the chains pack along the c-axis of the unit cell, stabilized by interchain Hbonding interactions from a nitrato oxygen atom of one chain to the CH proton from an uncoordinated 15C5 macrocycle of a neighboring chain such that O20···H7−C7 = 2.506 Å (Figure 4). Intrachain Dy···Dy distances are 8.809 and 8.806 Å, and the

reveals that, like the Dy(III) ions in 6, it adopts a 9-coordinate geometry that is also closest to a capped square antiprism with a CSM of 2.29 (page S6 in the SI).27 Analysis of the crystal packing reveals that the molecules pack as H-bonded chains that are aligned coparallel along the b-axis of the unit cell, with interchain H-bonding interactions between nearest neighbors giving rise to 2-D sheets (Figure 6a). Interestingly,

Figure 4. Crystal packing of 6. H atoms are removed for clarity. View down the b-axis of the unit cell. H-bonding interactions are shown as blue ---.

Figure 6. (a) Crystal packing of 7 showing the 1-D chains that are connected via interchain H bonds to afford 2-D sheets in the bc-plane. (b) View down the b-axis of the unit cell showing the large channels between the 2-D layers. The Tb···Tb distance between the sheets is shown as a blue ---. H atoms are omitted for clarity.

shortest interchain distances are 8.854 Å, which are slightly shorter than the Dy···Dy distances of 8.880(8) and 8.875(5) Å, reported for the half-sandwich and pseudosandwich complexes 3 and 4, respectively.20 The Tb(III) complex crystallizes with one [Tb(NO3)3(OH2)(MeOH)] unit H-bonded via its two ligated water molecules to 15C5 macrocycles such that O11−H11A···O15 = 1.76 and O11−H11B···O15 = 1.77 Å, and O10−H10A···O16 = 1.87 and O10−H10B···O14 = 1.71 Å (Figure 5). The Tb−O bond distances are in the range 2.318(9)− 2.5261(12) Å. Continuous shape analysis of the Tb(III) ion

perpendicular to these sheets there are very large channels that accommodate disordered solvent molecules which were removed from the crystallographic model, but serve to magnetically isolate the Tb(III) ions in the chains along the a-direction (Figure 6b). Magnetic Studies. Direct current susceptibility measurements were carried out on freshly prepared crystalline samples of 6 and 7 in an applied static field of 0.1 T between 5 and 300 K (page S7 in the SI). Above 15 K, both complexes display Curie−Weiss behavior with Curie constants of 13.86 and 12.02 cm3 mol−1 K, in excellent agreement with the theoretical values of 14.17 and 11.82 cm3 mol−1 K for noninteracting Dy(III) (6H15/2, S = 5/2, g = 4/3) and Tb(III) (7F6, S = 3, g = 3/2) ions, respectively. For both complexes, the decrease in χT below 50 K is likely due to the thermal depopulation of excited Stark sublevels within the ground multiplet of the respective Ln(III) ion and/or the presence of weak antiferromagnetic interactions in the crystal lattice. Alternating current susceptibility measurements were carried out in an oscillating field of 3.5 Oe in zero and nonzero static fields, between 2 and 15 K, using frequencies between 50 and 10 000 Hz. Additional field-dependent measurements at 2 K were recorded in fields up to 5000 Oe with frequencies ranging from 30 to 1000 Hz (see page S11 in the SI). The Dy(III) complex displays no frequency-dependent signals in zero field, most likely because of fast quantum tunnelling of magnetization

Figure 5. Molecular structure of {[Tb(NO3)3(H2O)2(CH3OH)]· (15C5)}n (7) with an appropriate atomic numbering scheme. Interchain H-bonding interactions are shown as blue ---. The crystallographically unique atoms in the asymmetric unit are shown in color. D

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (QTM). However, when a series of static fields between 300 and 2000 Oe were applied to suppress QTM, the out-of-phase χ″ components of the susceptibility show a clear frequencydependence consistent with SMM behavior. Plots of χ″ versus temperature in an applied field of 2000 Oe (Figure 7) reveal that, at higher ac frequencies, well-resolved

For an investigation into these relaxation pathways, further Cole−Cole plots were examined at different temperatures (Figure 9, top). In the low-temperature region below 5 K,

Figure 7. Out-of-phase χ″ vs temperature in an applied field of 2000 Oe with selected frequencies in the range 50−10 000 Hz ( are a guide for the eye).

maxima in χ″ can be observed below 5 K with partial resolution of the maxima into two components observed at the upper limit of our hardware setup (10 000 Hz) indicating the presence of two relaxation pathways. When these data are replotted in the form of χ″ versus frequency at different temperatures (Figure 8a), we again see the diagnostic frequency-dependence of the out-of-phase susceptibility expected for SMM behavior with well-resolved maxima emerging below 6.4 K. Upon further cooling below 4 K (Figure 8b), a significant change in the temperature-dependence of χ″ is observed in which the maximum in χ″ around 6000 Hz decreases in intensity, and a low-frequency “tail” begins to emerge although no lowfrequency maximum could be observed down to 2 K. The evolution of the temperature-dependence reflects a change in the dominant relaxation processes occurring in this system upon cooling. Alternating current measurements in other static fields are available on page S10 in the SI.

Figure 9. (Top) Modeled out-of-phase χ″ vs in-phase χ′ in a 2000 Oe applied field with kinks in the Cole−Cole semicircles clearly visible below 5 K ( represent the fit to the Cole−Cole equations). (Bottom) Fit of the data to an Arrhenius model.

several of the χ′ versus χ″ semicircles are visibly kinked, indicative of two partially merged arcs, each corresponding to a distinct relaxation domain. This behavior is reminiscent of the previously studied 15C5 sandwich-type complexes and was therefore modeled using two combined modified Debye functions (page S13 in the SI).20,21 In this respect, the Arrhenius plot for the relaxation comprises two regions; at high temperature, the Cole−Cole data were best fitted as a single relaxation process, whereas at low temperature a two-

Figure 8. Out-of-phase χ″ vs frequency plot for 6 in a 2000 Oe applied field: (a) from 9.8 to 4.3 K and (b) from 4.3 to 2.2 K ( are a guide for the eye). E

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry component Debye term was used. The temperature-dependence of the relaxation can be written in terms of different relaxation mechanisms: τ −1 = τ0−1 exp( −U /kBT ) + ABm T + CT n + D

Here, the first term refers to the thermally activated Orbachtype relaxation with the latter terms reflecting direct, Raman, and quantum tunnelling processes, respectively, where m = 2 and n = 5 for Kramers ions such as Dy(III).28 Above 5 K, the data were fitted to a single thermally activated process [Ueff = 26 K (18 cm−1), τ0 = 4.10 × 10−7 s] consistent with an Orbach mechanism. In the low-temperature region, the separation of the relaxation processes reflects the persistence of an Orbach process [Ueff = 41 K (29 cm−1), τ0 = 1.35 × 10−8 s] coupled with a temperature-independent term consistent with a quantum tunnelling mechanism (τ0 = 2.30 × 10−5 s) arising from tunnelling between ±mJ levels of the ground state. We emphasize here that the poor deconvolution of the two relaxation processes gives rise to considerable uncertainty in the fitting of the Cole−Cole data. Nevertheless, the features in the χ″ versus T data (Figure 7) and the marked temperaturedependence of the χ″ versus frequency data (Figure 8) are all consistent with two relaxation processes in the low-temperature regime. Furthermore, the α parameters of 0.28−0.45 for domain A and 0.39−0.44 for domain B suggest that multiple relaxation processes are at play. In sharp contrast, ac susceptibility studies on the corresponding Tb(III) analogue 7 in both zero and applied static dc fields showed no frequency-dependence, and thus, we conclude that this complex displays no SMM behavior (page S11 in the SI). Ab Initio Studies. For a further elucidation of the dynamic magnetic properties of the two complexes, ab initio calculations were carried out with MOLCAS 8.26 For 6, the relative energies and the directionality of the main anisotropy axes of the KD in the 6H15/2 ground state were evaluated. The main magnetic axes of the ground- and first-excited-state KDs are summarized in Table 1 and projected onto the symmetry axis of the molecule in Figure 10. The calculations reveal that both the ground- and

Figure 10. (a, b) Plots of KD1± and KD2± for Dy1 and Dy2 in 6 showing the average matrix elements of the transition moments between states (numbers next to arrows) for model 1, long basis set. Red arrows are used for thermally activated transitions, blue arrows for ground and excited QTM, and green arrows for Orbach spin−lattice relaxation. (c, d) Crystal structures of Dy1 and Dy2 coordination spheres showing the H-bound crown ethers and the main magnetic axes of the ground (green) and first-excited (purple) KD.

first-excited-state Kramers doublets (KD1± and KD2±, respectively) contain a significant contribution from the transverse gxy parameters, deviating significantly from Ising with gz values of 16.360 and 13.183, respectively, for Dy1 and 17.474 and 12.391 for Dy2. For both crystallographically unique Dy(III) ions, the ground wave functions consist of mainly mJ = ±15/2, whereas the first excited state includes a large contribution from all mJ states. Examination of the magnitude of the matrix elements between the KD1± and KD2± spin states (0.60 for Dy1 and 0.44 for Dy2) reveals that QTM involving the ground-state doublets of both ions is accessible in a zero dc field. In addition, the large matrix element connecting the ground and first excited state of opposite spin for both ions also supports the presence of a second, thermally assisted relaxation process (Figure 10). In the presence of a small applied dc field, the QTM involving the ground state of the two Dy(III) ions is likely suppressed. In this case, the calculations support the presence of a dominant thermally activated Orbach relaxation process which is consistent with the experimental data. The calculated energy gap between the ground (1±) and the first (2±) excited state in 6 is 21 cm−1 for Dy1 and 31 cm−1 for Dy2. Furthermore, the lack of well-isolated first excited states and the large deviation from Ising character for both ions also support the presence of rapid QTM, consistent with the absence of zero-field SMM behavior. Theoretical studies were also carried out on the Tb(III) analogue 7 to shed more light on its magnetic properties. The g-tensors and the relative energies and angles of the principle anisotropy axes of the ground and first excited states are summarized in Table 2. As expected for the non-Kramers Tb(III) ion, all of the pseudodoublets in the model are of the pure Ising-type. The ground-state pseudodoublet possesses a gz value of 17.40, approaching that expected for a pure mJ = ±6 state where gz =

Table 1. Energies of the Eight Kramers Doublets (KDs) in the 6H15/2 Multiplet for the Dy1 and Dy2 Ions in 6 KD

E (cm−1)

gx

1 2 3 4 5 6 7 8

0.000 21.442 58.452 82.734 116.830 142.992 171.438 222.416

1.093 3.551 4.548 1.365 5.488 5.980 4.266 1.312

1 2 3 4 5 6 7 8

0.000 30.676 56.850 85.727 122.264 156.264 185.508 257.141

0.878 2.597 0.194 3.606 0.440 10.616 1.482 0.213

gy Dy1 2.504 3.551 4.548 1.365 5.488 5.980 4.266 1.312 Dy2 1.741 2.878 3.203 4.799 2.627 6.424 4.222 0.626

gz

angle (deg)

16.360 13.183 11.609 17.167 0.405 11.816 11.673 17.594

0.000 129.637 57.182 42.533 90.396 37.416 83.543 86.211

17.494 12.391 9.779 6.655 11.391 0.702 12.610 18.555

0.000 91.763 78.655 67.594 40.468 106.811 96.588 109.126 F

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry 5

L9,10, and 5D2−4 levels, and a broad band (240−290 nm) assigned to the contribution of the spin-allowed (low-spin, LS) and spin-forbidden (high-spin, HS) interconfigurational Tb(III) fd transitions.30 Similar to what is found for the emission spectra, apart from a decrease of the fwhm for the intra-4f lines, the excitation spectrum observed at 14 K resembles that measured at 300 K (not shown). The lifetime values of the excited levels for complexes 6 (4F9/2) and 7 (5D4) were evaluated at 300 K. Whereas, for the former complex, the decay lies below the detection limits of the experimental setup (10−9 s), the emission decay curve for complex 7 reveals a single exponential behavior, yielding a lifetime value of (0.742 ± 0.002) × 10−6 s. The observation of a single exponential decay is in good agreement with the presence of a single average local environment found for 7. The emission features of both complexes were further quantified by the measurement of the absolute emission quantum yield. Whereas for complex 6 very low values were found ( 0.98) for a judgment of the fit quality.

Accession Codes

CCDC 1519141−1519142 contain the supplementary crystallographic data for this paper. These data can be obtained free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by emailing [email protected], or by contacting The Cambridge Crystallographic Data Centre, 12 Union Road, Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

components using Gaussian functions, whose energy was constrained to the peak position by careful analysis of the spectrum taking into account the experimental error (±5 cm−1); the fwhm and the relative intensity were free to vary. The two high-energy components are hot bands (marked with an asterisk in Figure 12) arising from transitions from the first 4 F9/2 Stark sublevel.18b From the best fit of the emission spectra, two average energy barriers, ΔE1 = 28 ± 5 cm−1 and ΔE2 = 37 ± 5 cm−1, were estimated, which are in excellent agreement with the calculated values for ΔE1 from the ab initio studies.

■

AUTHOR INFORMATION

Corresponding Author

*E-mail: [email protected]. ORCID

Luis D. Carlos: 0000-0003-4747-6535 Melanie Pilkington: 0000-0002-9274-8512 Notes

The authors declare no competing financial interest.

■

■

CONCLUSIONS We report the synthesis and structural characterization of two H-bonded 15C5 Ln(III) chains. Although both complexes crystallize in similar unit cells, they are not isostructural, with complex 6 containing two crystallographically independent Dy(III) ions and the Tb(III) analogue 7 comprising just a single unique 4f-ion. Joint magneto-luminescence studies reveal that 6 is a dual-property emissive SMM, whereas the presence of a fast quantum tunneling mechanism in 7 thwarts the observation of slow relaxation of magnetization. Magnetic susceptibility studies of 6 reveal the presence of two effective energy barriers (29 and 0 cm−1) below 5 K and a third Ueff of 18 cm−1 between 5 and 7 K. At 14 K, the well-resolved solidstate emission spectrum of 6 permitted a Gaussian fit of the fine structure of the highest-energy emission band affording an energy separation between the ground- and first-excited-state KD of 28 ± 5 and 37 ± 5 cm−1 for Dy1 and Dy2, respectively. Ab initio studies shed additional light on the magnetization dynamics of the two complexes; the energy gap ΔE1 between the ground- and the first-excited-state KD in 6 was calculated to be 21 cm−1 for Dy1 and 31 cm−1 for Dy2, in excellent agreement with the luminescence data. Furthermore, in the presence of a small applied dc field to suppress quantum tunnelling processes, theoretical studies support the presence of a dominant Orbach relaxation process, consistent with the experimental ac susceptibility data. For the Tb(III) derivative 7, the calculations reveal a significant tunnel splitting (Δtun = 0.30 cm−1) within the ground multiplet suggesting that the mJ states

ACKNOWLEDGMENTS This work was supported by NSERC (DG, M.P.; USRA, M.A.H.), CRC (Tier II, M.P.), Brock University, CFI (New Opportunities, M.P.), OIT (matching funds, M.P.), and the Lybian North American Scholarship program (Z.R.A.). This work is partially developed in the scope of the projects CICECOAveiro Institute of Materials (UID/CTM/50011/ 2013) financed by national funds through the Fundaçaõ para a Ciência e a Tecnologia/Ministério da Educaçaõ e Ciência (FCT/MEC) and when applicable cofinanced by FEDER under the PT2020 Partnership Agreement.

■

REFERENCES

(1) Benelli, C.; Gatteschi, D. Introduction to Molecular Magnetism; Wiley−VCH Verlag GmbH: Weinheim, 2015. (2) For example, see: (a) Gatteschi, D.; Sessoli, R. Quantum Tunneling of Magnetization and Related Phenomena in Molecular Materials. Angew. Chem., Int. Ed. 2003, 42, 268−297. (b) Aromi, G.; Brechin, E. K. Synthesis of 3D Metallic Single Molecule Magnets. Struct. Bonding (Berlin) 2006, 122, 1−69. (3) Urdampilleta, M.; Klyatskaya, S.; Cleuziou, J.-P.; Ruben, M.; Wernsdorfer, W. Supramolecular Spin Valves. Nat. Mater. 2011, 10, 502−506. (4) (a) Vincent, R.; Klyatskaya, S.; Ruben, M.; Wernsdorfer, W.; Balestro, F. Electronic Read-out of a Single Nuclear Spin Using a Molecular Spin Transistor. Nature 2012, 488, 357−360. (b) Thiele, S.; Balestro, F.; Ballou, R.; Klyatskaya, S.; Ruben, M.; Wernsdorfer, W. Electrically Driven Nuclear Spin Resonance in Single-Molecule Magnets. Science 2014, 344, 1135−1138.

H

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry (5) Leuenberger, M. N.; Loss, D. Quantum Computing in Molecular Magnets. Nature 2001, 410, 789−793. (b) Ardavan, A.; Blundell, S. J. J. Mater. Chem. 2009, 19, 1754−1760. (6) Gatteschi, D.; Sessoli, R.; Villain, J. Molecular Nanomagnets; Oxford University Press: Oxford, U.K., 2006. (7) (a) Wernsdorfer, W.; Sessoli, R. Quantum Phase Interference and Parity Effects in Magnetic Molecular Clusters. Science 1999, 284, 133− 135. (b) Wernsdorfer, W.; Aliaga-Alcalde, N.; Hendrickson, D. N.; Christou, G. Exchange-Biased Quantum Runnelling in a Supramolecular Dimer of Single-Molecule Magnets. Nature 2002, 416, 406− 409. (c) Hill, S.; Edwards, R. S.; Aliaga-Alcalde, N.; Christou, G. Quantum Coherence in an Exchange-Coupled Dimer of SingleMolecule Magnets. Science 2003, 302, 1015−1018. (8) Athanasopoulou, A. A.; Abbasi, P.; Alexandropoulus, D.; Hayward, J. J.; Beavers, C. M.; Teat, S. J.; Wernsdorfer, W.; Escuer, A.; Pilkington, M.; Stamatatos, Th. C. A {NiII18} Single-Molecule Magnet with a ‘Flying Saucer’ Topology and a Large Energy Barrier of 381 K. Angew. Chem., Int. Ed. 2016, submitted. (9) Sessoli, R.; Tsai, H. L.; Schake, A. R.; Wang, S.; Vincent, J. B.; Folting, K.; Gatteschi, D.; Christou, G.; Hendrickson, D. N. High-Spin Molecules: [Mn12O12(O2CR)16(H2O)4]. J. Am. Chem. Soc. 1993, 115, 1804−1816. (10) Ishikawa, N.; Sugita, M.; Ishikawa, T.; Koshihara, S.; Kaizu, Y. Lanthanide Double-Decker Complexes Functioning as Magnets at the Single-Molecular Level. J. Am. Chem. Soc. 2003, 125, 8694−8695. (11) Feltham, H. L. C.; Brooker, S. Review of Purely 4f and MixedMetal nd-4f Single-Molecule Magnets Containing Only One Lanthanide Ion. Coord. Chem. Rev. 2014, 276, 1−33. (12) Chilton, N. F. Design Criteria for High-Temperature SingleMolecule Magnets. Inorg. Chem. 2015, 54, 2097−2099. (13) For examples, see: (a) Woodruff, D. N.; Winpenny, R. E. P.; Layfield, R. A. Lanthanide Single Molecule Magnets. Chem. Rev. 2013, 113, 5110−5148. (b) Gavey, E. L.; Beldjoudi, Y.; Rawson, J. M.; Stamatatos, Th. C.; Pilkington, M. Slow relaxation in the first penta-aza Dy(III) macrocyclic complex. Chem. Commun. 2014, 50, 3741−3743. (c) Liu, J.-L.; Chen, Y.-C.; Zheng, Y.-Z.; Lin, W.-Q.; Ungur, L.; Wernsdorfer, W.; Chibotaru, L. F.; Tong, M.-L. Switching the Anisotropy Barrier of a Single-ion Magnet by Symmetry Change from Quasi-D5h to Quasi-Oh. Chem. Sci. 2013, 4, 3310−2216. (d) Meihaus, K. R.; Long, J. R. Magnetic Blocking at 10 K and a Dipolar-Mediated Avalanche in Salts of the bis(η8-cyclooctatetraenide) Complex [Er(COT)2]−. J. Am. Chem. Soc. 2013, 135, 17952−17957. (14) (a) Ungur, L.; Le Roy, J. J.; Korobkov; Murugesu, M.; Chibotaru, L. F. Fine-tuning the Local Symmetry to Attain Record Blocking Temperature and Magnetic Remanence in a Single-Ion Magnet. Angew. Chem., Int. Ed. 2014, 53, 4413−4417. (b) Liu, J.; Chen, Y.-C.; Liu, J.-L.; Vieru, V.; Ungur, L.; Jia, J.-H.; Chibotaru, L. F.; Lan, Y.; Wernsdorfer, W.; Gao, S.; Chen, S.-M.; Tong, M. L. A Stable Pentagonal Bipyramidal Dy(III) Single-ion Magnet with a Record Magnetization Reversal Barrier Over 1000K. J. Am. Chem. Soc. 2016, 138, 5441−5450. (15) (a) Pedersen, C. J. The Discovery of Crown Ethers (Nobel Leture). Angew. Chem., Int. Ed. Engl. 1988, 27, 1021. (b) Pedersen, C. J. Cyclic Polyethers and their Complexes with Metal Salts. J. Am. Chem. Soc. 1967, 89, 7017−7036. (16) Felinto, M. C. F. C.; Tomiyama, C. S.; Brito, H. F.; Teotonio, E. E. S.; Malta, O. L. Synthesis and Luminescent Properties of Supramolecules of β-Diketonate of Eu(III) and Crown Ethers as Ligands. J. Solid State Chem. 2003, 171, 189−194. (17) Belian, M. F.; de Sá, G. F.; Alves, S., Jr.; Galembeck, A. J. Systematic Study of Luminescent Properties of New Lanthanide Complexes Using Crown Ethers as Ligand. J. Lumin. 2011, 131, 856− 860. (18) (a) Cucinotta, G.; Perfetti, M.; Luzon, J.; Etienne, M.; Car, P. E.; Caneschi, A.; Calvez, G.; Bernot, K.; Sessoli, R. Magnetic Anisotropy in a Dysprosium/DOTA Single-Molecule Magnet: Beyond Simple Magneto-Structural Correlations. Angew. Chem., Int. Ed. 2012, 51, 1606−1610. (b) Long, J.; Vallat, R.; Ferreira, R. A. S.; Carlos, L. D.; Almeida Paz, F. A.; Guari, Y.; Larionova, J. A Bifunctional Luminescent

Single-ion Magnet: Towards Correlation Between Luminescence Studies and Magnetic Slow Relaxation Processes. Chem. Commun. 2012, 48, 9974−9976. (19) For examples, see: (a) Long, J.; Rouquette, J.; Thibaud, J.- M.; Ferreira, R. A. S.; Carlos, L. D.; Donnadieu, B.; Vieru, V.; Chibotaru, L. F.; Konczewicz, L.; Haines, J.; Guari, Y.; Larionova, J. A HighTemperature Molecular Ferroelectric Zn/Dy Complex Exhibiting Single-Ion-Magnet Behavior and Lanthanide Luminescence. Angew. Chem. 2015, 127, 2264−2268. (b) Costes, J. P.; Titos-Padilla, S.; Oyarzabal, I.; Gupta, T.; Duhayon, C.; Rajaraman, G.; Colacio, E. Effect of Ligand Substitution around the DyIII on the SMM Properties of Dual-Luminescent Zn−Dy and Zn−Dy−Zn Complexes with Large Anisotropy Energy Barriers: A Combined Theoretical and Experimental Magnetostructural Study. Inorg. Chem. 2016, 55, 4428−4440. (20) Gavey, E. L.; Al Hareri, M.; Regier, J.; Carlos, L. D.; Ferreira, R. A. S.; Razavi, F. A.; Rawson, J. M.; Pilkington, M. Placing a Crown on DyIII − A Dual Property LnIII Crown Ether Complex Displaying Optical Properties and SMM Behaviour. J. Mater. Chem. C 2015, 3, 7738−7747. (21) Al. Hareri, M.; Gavey, E. L.; Regier, J.; Ras Ali, Z.; Carlos, L. D.; Ferreira, R. A. S.; Pilkington, M. Encapsulation of a [Dy(OH2)8]3+ cation: magneto-optical and theoretical studies of a caged, emissive SMM. Chem. Commun. 2016, 52, 11335−11338. (22) Sheldrick, G. M. SADABS: Program for Empirical Absorption Correction; University of Gottingen: Germany, 1996. (23) SADABS; Bruker Analytical X-ray Systems: Madison, WI, U.S.A., 2001. (24) Sheldrick, G. M. Crystal Structure Refinement with SHELXL. Acta Crystallogr., Sect. C: Struct. Chem. 2015, C71, 3−8. (25) Spek, A. L. PLATON SQUEEZE: A Tool for the Calculation of the Disordered Solvent Contribution to the Calculated Structure Factors. Acta Crystallogr., Sect. C: Struct. Chem. 2015, C71, 9−18. (26) Aquilante, F.; De Vico, L.; Ferré, N.; Ghigo, G.; Malmqvist, P.A.; Neogrády, P.; Pedersen, T. B.; Pitoňaḱ , M.; Reiher, M.; Roos, B. O.; Serrano-Andrés, L.; Urban, M.; Veryazov, V.; Lindh, R. MOLCAS 7: The Next Generation. J. Comput. Chem. 2010, 31, 224−247. (27) Zabrodsky, H.; Peleg, S.; Avnir, D. Continuous symmetry measures. J. Am. Chem. Soc. 1992, 114, 7843−7851. (28) Hazra, S.; Titis, J.; Valigura, D.; Boca, R.; Mohanta, S. Bisphenoxido and bis-acetato bridged heteronuclear {CoIIIDyIII} single molecule magnets with two slow relaxation branches. Dalton Trans. 2016, 45, 7510−7520. (29) Ojea, M J. H.; Milway, V. A.; Velmurugan, G.; Thomas, L. H.; Coles, S. J.; Wilson, C.; Wernsdorfer, W.; Rajaraman, G.; Murrie, M. Enhancement of TbIII−CuII Single-Molecule Magnet Performance through Structural Modification. Chem. - Eur. J. 2016, 22, 12839− 12848. (30) van Pieterson, L.; Reid, M. F.; Burdick, G. W.; Meijerink, A. 4fn→4f n−15d Transitions of the Heavy Lanthanides: Experiment and Theory. Phys. Rev. B: Condens. Matter Mater. Phys. 2002, 65, 045114− 045127. (31) Belian, M. B.; de Sa, G. F.; Alves, S., Jr.; Galembeck, A. Systematic Study of Luminescent Properties of New Lanthanide Complexes Using Crown Ethers as Ligands. J. Lumin. 2011, 131, 856− 860. (32) (a) Yamashita, K.; Miyazaki, R.; Kataoka, Y.; Nakanishi, T.; Hasegawa, Y.; Nakano, M.; Yamamura, T.; Kajiwara, T. A Luminescent Single-Molecule Magnet: Observation of Magnetic Anisotropy Using Emission as a Probe. Dalton Trans. 2013, 42, 1987−1990. (b) Pointillart, F.; Le Guennic, B.; Golhen, S.; Cador, O.; Maury, O.; Ouahab. A Redox-Active Luminescent Ytterbium Based Single Molecule Magnet. Chem. Commun. 2013, 49, 615−617. (c) Shintoyo, S.; Murakami, K.; Fujinami, T.; Matsumoto, N.; Mochida, N.; Ishida, T.; Sunatsuki, Y.; Watanabe, M.; Tsuchimoto, M.; Mrozinski, J.; Coletti, C.; Re, N. Crystal Field Splitting of the Ground State of Terbium(III) and Dysprosium(III) Complexes with a Triimidazolyl Tripod Ligand and an Acetate Determined by Magnetic Analysis and Luminescence. Inorg. Chem. 2014, 53, 10359−10369. (d) Ren, M.; Bao, S.-S.; Ferreira, R. A. S.; Zheng, L.-M.; Carlos, L. D. A Layered I

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX

Article

Inorganic Chemistry Erbium Phosphonate in Pseudo-D5h Symmetry Exhibiting FieldTunable Magnetic Relaxation and Optical Correlation. Chem. Commun. 2014, 50, 7621−7624. (e) Rechkemmer, Y.; Fischer, J. E.; Marx, R.; Dörfel, M.; Neugebauer, P.; Horvath, S.; Gysler, M.; BrockNannestad, T.; Frey, W.; Reid, M. F.; van Slageren, J. Comprehensive Spectroscopic Determination of the Crystal Field Splitting in an Erbium Single-Ion Magnet. J. Am. Chem. Soc. 2015, 137, 13114− 13120. (f) Gregson, M.; Chilton, N. F.; Ariciu, A.-M.; Tuna, F.; Crowe, I. F.; Lewis, W.; Blake, A. J.; Collison, D.; McInnes, E. J. L.; Winpenny, R. E. P.; Liddle, S. A Monometallic Lanthanide bis(methanediide) Single Molecule Magnet with a Large Energy Barrier and Complex Spin Relaxation Behaviour. Chem. Sci. 2016, 7, 155−165. (g) Bi, Y.; Chen, C.; Zhao, Y.-F.; Zhang, Y.-Q.; Jiang, S.-D.; Wang, B.-W.; Han, J.B.; Sun, J.-L.; Bian, Z.-Q.; Wang, Z.-M.; Gao, S. Thermostability and Photoluminescence of Dy(III) Single-Molecule Magnets Under a Magnetic Field. Chem. Sci. 2016, 7, 5020−5031. (33) Mamontova, E.; Long, J.; Ferreira, R. A. S.; Botas, A. M. P.; Luneau, D.; Guari, Y.; Carlos, L. D.; Larionova, J. MagnetoLuminescence Correlation in the Textbook Dysprosium(III) Nitrate Single-Ion Magnet. Magnetochemistry 2016, 2 (4), 41−51.

J

DOI: 10.1021/acs.inorgchem.7b00089 Inorg. Chem. XXXX, XXX, XXX−XXX