Field-Induced Dysprosium Single-Molecule Magnet Involving a Fused o-Semiquinone-Extended-Tetrathiafulvalene-o-Semiquinone Bridging Triad

: The reaction between the 2,2 (cid:48) -benzene-1,4-diylbis(6-hydroxy-4,7-di- tert -butyl-1,3-benzodithiol-2-ylium-5-olate biradical triad ( L ) and the metallo-precursor [Dy(hfac) 3 ] · 2H 2 O leads to the formation of a one-dimensional coordination polymer with the formula {[Dy(hfac) 3 ( L )] · 2C 6 H 14 } n ( 1 ). The X-ray structure reveals that the polymeric structure is formed by the bridging of the Dy(hfac) 3 units with the multi-redox triad L . Single-crystal X-ray diffraction and UltraViolet-visible absorption spectroscopy conﬁrm that the triad L in 1 is bound as a direduced, diprotonated form of o -quinone-extended tetrathiafulvalene- o -quinone (Q-exTTF-Q). Alternate Current (AC) measurements highlight a ﬁeld-induced single-molecule magnet (SMM) behavior with an energy barrier of 20 K, and thus 1 can be described as a one-dimensional assembly of mononuclear SMMs bridged by the L triad.


Introduction
Multifunctional materials are key systems for the communities of chemists and physicists studying exciting physical properties. One explored route to design multifunctional materials is the use of adequate organic ligands to contribute one property while metallic precursors carry additional properties such as magnetism and luminescence. To combine the two latter properties, lanthanide ions are the best candidates because of their intrinsic strong magnetic anisotropy and high magnetic moment combined with their specific luminescence [1][2][3][4]. Thus, a luminescent single-molecule magnet (SMM) [5][6][7], ferroelectric SMM [8], chiral SMM [9,10], and redox-active SMM [11,12] were obtained, associating chiral or redox-active ligands with lanthanide precursors. Following this strategy, one of the most widely used redox-active ligands is based on the tetrathiafulvalene (TTF) core. Even more interesting is the decoration of the TTF core with one or two p-quinone [13,14] and o-quinone molecules [15], because these can be easily oxidized or reduced in the coordination sphere of the metal, leading to the observation of a redox-isomerisation mechanism [16,17]. The resulting donor (D)-acceptor (A) dyads or A-D-A triads (where D refers to TTF and A refers to quinone) open the route for potential applications in molecular electronics and optoelectronics because of their low highest occupied molecular orbital to lowest unoccupied molecular orbital (HOMO-LUMO) gap [18]. A few Inorganics 2018, 6, 45 2 of 9 years ago, some of the current authors elaborated upon dinuclear complexes that displayed SMMs or luminescence using the 4,4 ,7,7 -tetra-tert-butyl-2,2 -bi-1,3-benzodithiole-5,5 ,6,6 -tetrone triad [19,20].

Crystal Structure Description of {[Dy(hfac)3(L)]·2C6H14}n (1)
Compound 1 crystallizes in the P21/c (N°14) monoclinic space group ( Table 1). The asymmetric unit is composed of one Dy(hfac)3 moiety, one L triad and two n-hexane molecules of crystallization. An Oak Ridge thermal ellipsoid plot (ORTEP) view of 1 is depicted in Figure S1.   Table 1). The asymmetric unit is composed of one Dy(hfac) 3 moiety, one L triad and two n-hexane molecules of crystallization. An Oak Ridge thermal ellipsoid plot (ORTEP) view of 1 is depicted in Figure S1.
The X-ray structure of 1 reveals the formation of a coordination polymer ( Figure 1). The coordination between the Dy III ion and the bridging ligand takes place through the non-protonated oxygen atoms (O1 and O4). The two protonated oxygen atoms (O2 and O3) remain uncoordinated. The bridging ligand in 1 is almost planar, and its structure is very similar to that of the free ligand in its protonated semiquinone form [22]. In fact, the torsion angle between the planes defined by the benzo-1,3-dithio system and the p-phenylene ring takes the average value of 6.32 (11) • , which can be compared with the value of 8.70(10) • found for the free L ligand. The direduced, diprotonated form of the ligand L in 1 was also confirmed by the distribution of C-C distances in the dioxolene rings as well as the C-OH (1.347(4) Å) and C-O· (1.307(4) Å) bond lengths. The C-OH and C-O· distances in 1 were slightly longer and shorter, respectively, than in the free biradical triad because of the electronic effect of the coordination of the trivalent metal. Finally, the semiquinone form was also confirmed by absorption spectroscopy (see Section 2.3). The arrangement of the ligands around the Dy III center leads to an O 8 surrounding with a coordination polyhedron symmetry intermediate between the C 2v biaugmented trigonal prism (CShM BTPR-8 = 1.140), the D 4d square antiprism (CShM SAPR-8 = 1.191), and the D 2d triangular dodecahedron (CShM TDD-8 = 1.238). The numbers correspond to the deviation from the ideal symmetries determined with the SHAPE 2.1 program [24]. The Dy1-O hfac distances (2.382(3) Å) were longer than those of Dy1-O SQ (2.267(3) Å) ( Table 2).  The X-ray structure of 1 reveals the formation of a coordination polymer ( Figure 1). The coordination between the Dy III ion and the bridging ligand takes place through the non-protonated oxygen atoms (O1 and O4). The two protonated oxygen atoms (O2 and O3) remain uncoordinated. The bridging ligand in 1 is almost planar, and its structure is very similar to that of the free ligand in its protonated semiquinone form [22]. In fact, the torsion angle between the planes defined by the benzo-1,3-dithio system and the p-phenylene ring takes the average value of 6.32(11)°, which can be compared with the value of 8.70(10)° found for the free L ligand. The direduced, diprotonated form of the ligand L in 1 was also confirmed by the distribution of C-C distances in the dioxolene rings as well as the C-OH (1.347(4) Å) and C-O· (1.307(4) Å) bond lengths. The C-OH and C-O· distances in 1 were slightly longer and shorter, respectively, than in the free biradical triad because of the electronic effect of the coordination of the trivalent metal. Finally, the semiquinone form was also confirmed by absorption spectroscopy (see Section 2.3). The arrangement of the ligands around the Dy III center leads to an O8 surrounding with a coordination polyhedron symmetry intermediate between the C2v biaugmented trigonal prism (CShMBTPR-8 = 1.140), the D4d square antiprism (CShMSAPR-  The crystal packing of 1 is shown in Figure 2. No significant intermolecular interactions, such as S···S, S···O, or π···π interactions, as observed in the X-ray structure of the free ligand L [22], were identified. The L triad and Dy(hfac) 3 metallo-precursor formed two subnetworks aligned along the c-axis (Figure 2). The shortest intra-and inter-molecular Dy···Dy distances were determined to be equal to 21.079 and 11.079 Å, respectively. Table 2. Selected bond lengths (Å) for compound 1.
The crystal packing of 1 is shown in Figure 2. No significant intermolecular interactions, such as S···S, S···O, or π···π interactions, as observed in the X-ray structure of the free ligand L [22], were identified. The L triad and Dy(hfac)3 metallo-precursor formed two subnetworks aligned along the caxis (Figure 2). The shortest intra-and inter-molecular Dy···Dy distances were determined to be equal to 21.079 and 11.079 Å, respectively.

Infrared Spectroscopy
The Infrared (IR) spectra for both the free L ligand and 1 were recorded in the KBr solid state. The protonated semiquinone form of L was confirmed by C-O· vibrations at 1549 and 1502 cm −1 [22]. For 1, these semiquinone carbonyl vibrations were identified in the frequency range of 1485-1565 cm −1 . The red shift is explained by the coordination of the Dy(hfac)3 moiety, and this was in good agreement with the slight changes in bond lengths observed in the X-ray structure. The additional strong band centered at 1655 cm −1 was attributed to the C=O carbonyl groups of the hfac − anions. As previously observed [22], the characteristic signals of hydroxyl groups were observed for neither L nor 1.

Absorption Spectroscopy
The absorption spectrum of 1 was recorded in CH2Cl2 solution at room temperature ( Figure 3). It displayed a broad absorption band in the near infrared (NIR) region (890 nm), which should correspond to HOMO to LUMO excitations [22]. Such a wavelength is characteristic of the semiquinone form of the triad in 1, whereas in the case of the quinone form, it should be centered at 750 nm. [22] The less-intense absorption band centered at 540 nm was attributed to the α+β-SOMOs to LUMO+1 excitations for the free L [22]. Clear changes in the electronic spectrum of 1 were observed compared to that of the free triad. Thus, more excitations were observed in both the NIR and visible regions likely as a result of the desymmetrization of the triad after complexation. Nevertheless, the most important difference between the electronic spectra of free L and 1 was the intensity of the

Infrared Spectroscopy
The Infrared (IR) spectra for both the free L ligand and 1 were recorded in the KBr solid state. The protonated semiquinone form of L was confirmed by C-O· vibrations at 1549 and 1502 cm −1 [22]. For 1, these semiquinone carbonyl vibrations were identified in the frequency range of 1485-1565 cm −1 . The red shift is explained by the coordination of the Dy(hfac) 3 moiety, and this was in good agreement with the slight changes in bond lengths observed in the X-ray structure. The additional strong band centered at 1655 cm −1 was attributed to the C=O carbonyl groups of the hfac − anions. As previously observed [22], the characteristic signals of hydroxyl groups were observed for neither L nor 1.

Absorption Spectroscopy
The absorption spectrum of 1 was recorded in CH 2 Cl 2 solution at room temperature ( Figure 3). It displayed a broad absorption band in the near infrared (NIR) region (890 nm), which should correspond to HOMO to LUMO excitations [22]. Such a wavelength is characteristic of the semiquinone form of the triad in 1, whereas in the case of the quinone form, it should be centered at 750 nm [22]. The less-intense absorption band centered at 540 nm was attributed to the α+β-SOMOs to LUMO+1 excitations for the free L [22]. Clear changes in the electronic spectrum of 1 were observed compared to that of the free triad. Thus, more excitations were observed in both the NIR and visible regions likely as a result of the desymmetrization of the triad after complexation. Nevertheless, the most important difference between the electronic spectra of free L and 1 was the intensity of the absorption band centered near 300 nm. This absorption band originated from the π-π* excitations localized on the hfac − anions [25]. absorption band centered near 300 nm. This absorption band originated from the π-π* excitations localized on the hfac − anions [25].

Magnetic Properties
The temperature dependence of the χMT product for 1 is depicted in Figure 4a. The room temperature χMT value was 13.95 cm 3 ·K·mol −1 , which was in good agreement with the expected value of 14.17 cm 3 ·K·mol −1 for a magnetically isolated Dy III ion ( 6 H15/2 with g = 4/3) [26]. It is worth noting that the magnetic contribution from the two semiquinone radicals was not observed because, even at room temperature, they were strongly antiferromagnetically coupled (1092 cm −1 ). Upon cooling, the χMT(T) curve monotonically decreased until 2 K. At this temperature, the χMT product reached the value of 11.84 cm 3 ·K·mol −1 . The reason for such behavior is the thermal depopulation of the ligand field levels (MJ combination) in the ground state multiplet. In the inset of Figure 4a is depicted the field dependence of the magnetization. At 50 kOe, the magnetization reached a value of 5.30 Nβ, in agreement with the expected value of 5 Nβ for an Ising ground state.
An out-of-phase signal of the magnetic susceptibility (χM") was observed at a high frequency (1000 Hz) at 2 K in a zero-Direct Current (DC) magnetic field (Figures 4b and S2). Nevertheless, no maxima were visible in the experimental time window available with our magnetometer because of an efficient zero-field quantum tunneling of the magnetization (QTM), which allows a fast relaxation as usually observed for mononuclear SMMs of Dy III [27,28]. In order to suppress the QTM, a DC magnetic field was applied, and a scan field of the magnetic susceptibility at 2 K was performed (Figure 4b). The application of a small DC field shifts the maxima of the χM" to lower frequencies. Such a shift is a sign of the suppression of the zero-field QTM [29,30], dipolar interaction [31], and/or hyperfine interaction [32,33], as well as of the breaking of the transverse magnetic anisotropy [34].

Magnetic Properties
The temperature dependence of the χ M T product for 1 is depicted in Figure 4a. The room temperature χ M T value was 13.95 cm 3 ·K·mol −1 , which was in good agreement with the expected value of 14.17 cm 3 ·K·mol −1 for a magnetically isolated Dy III ion ( 6 H 15/2 with g = 4/3) [26]. It is worth noting that the magnetic contribution from the two semiquinone radicals was not observed because, even at room temperature, they were strongly antiferromagnetically coupled (1092 cm −1 ). Upon cooling, the χ M T(T) curve monotonically decreased until 2 K. At this temperature, the χ M T product reached the value of 11.84 cm 3 ·K·mol −1 . The reason for such behavior is the thermal depopulation of the ligand field levels (M J combination) in the ground state multiplet. In the inset of Figure 4a is depicted the field dependence of the magnetization. At 50 kOe, the magnetization reached a value of 5.30 Nβ, in agreement with the expected value of 5 Nβ for an Ising ground state.
An out-of-phase signal of the magnetic susceptibility (χ M ") was observed at a high frequency (1000 Hz) at 2 K in a zero-Direct Current (DC) magnetic field (Figure 4b and Figure S2). Nevertheless, no maxima were visible in the experimental time window available with our magnetometer because of an efficient zero-field quantum tunneling of the magnetization (QTM), which allows a fast relaxation as usually observed for mononuclear SMMs of Dy III [27,28]. In order to suppress the QTM, a DC magnetic field was applied, and a scan field of the magnetic susceptibility at 2 K was performed (Figure 4b). The application of a small DC field shifts the maxima of the χ M " to lower frequencies. Such a shift is a sign of the suppression of the zero-field QTM [29,30], dipolar interaction [31], and/or hyperfine interaction [32,33], as well as of the breaking of the transverse magnetic anisotropy [34].

Crystallography
Single crystals of 1 were mounted on a APEXIII D8 VENTURE Bruker-AXS diffractometer (Bruker, Billerica, MA, USA) for data collection (Mo Kα radiation source, λ = 0.71073 Å), from the Centre de Diffractométrie (CDIFX), Université de Rennes 1, France. Structures were solved with a direct method using the SHELXT program [42] and were refined with a full matrix least-squares method on F 2 using the SHELXL-14/7 program [43]. Crystallographic data are summarized in Table 1. Complete crystal structure results as a CIF file including bond lengths, angles, and atomic coordinates are given in the Supplementary Materials (CCDC number 1825935).

Physical Measurements
The elementary analyses of the compounds were performed at the Centre Régional de Mesures Physiques de l'Ouest, Rennes. Absorption spectra were recorded on a Varian Cary 5000 UV-Visible-NIR spectrometer (Varian Inc. (Agilent Technologies), Palo Alto, CA, USA) equipped with an integration sphere. The DC magnetic susceptibility measurements were performed on solid polycrystalline samples with a Quantum Design MPMS-XL SQUID magnetometer (Quantum Design Inc., CA, USA) for temperatures between 2 and 300 K under an applied magnetic field of 200 Oe, between 2 and 20 K for 2 kOe, and between 20 and 80 K for 10 kOe and above. These measurements were all corrected for the diamagnetic contribution as calculated with Pascal's constants.

Conclusions and Outlook
A one-dimensional coordination polymer with the formula {[Dy(hfac) 3 (L)]·2C 6 H 14 } n (1) was synthesized. Its X-ray structure revealed that the Dy(hfac) 3 unit is bridged by the L biradical triad through the monodentate coordination of C-O· only. The bond-length distribution analysis, IR, and UV-vis absorption spectroscopy data confirm that the Q-exTTF-Q triad in 1 exists in a direduced, diprotonated state. The last technique highlighted the two expected HOMO-LUMO intra-ligand charge-transfer and α+β-SOMOs to LUMO+1 excitations for the semiquinone form. The AC measurements indicated field-induced SMM behavior with an energy barrier of 13.3 cm −1 . The coordination polymer 1 can be described as a one-dimensional assembly of mononuclear SMMs bridged by the L triad. The triad L is a multi-redox active ligand for which the two semiquinone moieties can be either reduced into a catechol or oxidized into an o-quinone form; consequently such redox activity is expected to strongly influence the magnetic properties of the polymer. Such a study is under progress in our laboratory.
Supplementary Materials: The following are available online at http://www.mdpi.com/2304-6740/6/2/45/s1: CIF and CIF-checked file; Figure S1: ORTEP View of the asymmetric unit of 1; Figure S2: Frequency dependence of the in-phase and out-of-phase signal of the magnetic susceptibility measured under a zero DC applied magnetic field; Figure S3: Frequency dependence of the in-phase signal of the magnetic susceptibility measured under a DC applied magnetic field of 1200 Oe; Figure S4: Normalized Cole-Cole plots for 1 at several temperatures between 2 and 8 K.
Author Contributions: S.N., V.K. and F.P. made the synthesis, J.F.G. and O.C. performed the magnetic and electronic spectra measurements and interpreted them. All authors participated to the writing process of the manuscript. All authors read and approved the final version of the manuscript.