Magnetic and Luminescence Properties of 8-Coordinated Pyridyl Adducts of Samarium(III) Complexes Containing 4,4,4-Triﬂuoro-1-(naphthalen-2-yl)-1,3-butanedionate

.T.) Abstract: A novel series of polypyridyl adducts, [Sm(ntfa) 3 (NN)] ( 2 – 4 ), with ntfa = 4,4,4-triﬂuoro-1-(naphthalen-2-yl)-1,3-butanedionate, NN = 2,2 (cid:48) -bipyridine (bipy), 4,4 (cid:48) -dimethyl-2,2 (cid:48) -bipyridine (4,4 (cid:48) -Me 2 bipy), and 5,5 (cid:48) -dimethyl-2,2 (cid:48) -bipyridine (5,5 (cid:48) -Me 2 bipy) were synthesized from the precursor complex [Sm(ntfa) 3 (MeOH) 2 ] ( 1 ) and the corresponding pyridyl ligands. Single X-ray crystallog-raphy showed that the complexes displayed 8-coordinated geometry. The solid pyridyl adducts 2 – 4 exhibited emission of luminescence in the NIR and visible regions with close quantum yields (QY = 0.20–0.25%). The magnetic data of 1 – 4 showed larger values than those expected for magnet-ically noncoupled Sm(III) complexes in the 6 H 5/2 ground state, with no saturation on the applied high magnetic ﬁeld static at a temperature of 2 K.


X-ray Crystal Structure Analysis
A Bruker-AXS APEX CCD single-crystal X-ray diffractometer (Bruker-AXS; Madison, WI, USA) with Mo-Kα radiation (λ = 0.71073 Å) and equipped with an Oxford Cryostream 700 cooling system was used for data collection at 100(2) K. Crystallographic data of the four title complexes are summarized in Table 1. SADABS and APEX computer programs [54,55] for absorption corrections and data processing were applied. The SHELX program package [56,57] was used for structure solution and refinement (direct methods, F 2 based full-matrix least-squares). The programs Mercury [58] and Platon [59] also were used. Partial disorder was observed in one -CF 3 group of 2. CCDC 1,964,549-CCDC 1,964,553 contained the crystallographic data in CIF format for 1-4, respectively.

Fluorescence Measurements
Solid-state fluorescence spectra of the title compounds 1-4 were recorded on a Horiba Jobin Yvon SPEX Nanolog fluorescence spectrophotometer (Fluorolog-3 v3.2, HORIBA Jovin Yvon, Cedex, France), which was equipped with a three-slit double-grating excitation and emission monochromator with dispersions of 2.1 nm/mm (1200 grooves/mm) at room temperature. The steady-state luminescence of the solid samples was excited by unpolarized light from a 450 W xenon CW lamp and detected at an angle of 22.5 • by a red-sensitive Hamamatsu R928 photomultiplier tube. Spectral corrections were made for both the emission spectral response (detector and grating) and the excitation source light-intensity variation (lamp and grating). NIR spectra (800-1400 nm) were recorded at an angle of 22.5 • using a solid InGaAs detector and liquid nitrogen as a coolant. The lifetime excited state (τ obs ) of the 4 G 5/2 emissions was measured on the same instrument in the phosphorescence mode using a 450 W xenon pulsed lamp (1.5 ns pulse), and the measured decays were analyzed using the Origin software package. The decay curves of the compounds 1-4 were fitted monoexponentially with Equation (1): The Pearson Chi 2 method was used to determine the fit quality. Absolute quantum yield (Φ TOT ) measurements were acquired in the G8 Quantum Integrating Sphere from GMP engaged with an interior reflective coating (Spectralon ® ). The Φ TOT was calculated using Equation (2): where L c is the calculated area of the outgoing amount of light after interaction with the sample, and L a is the calculated area without interaction with the sample (blank) at the λ exc and E c referees to the calculated area from the emission spectrum of the sample and E c (blank) from the corresponding spectrum of the blank.  (10 mL) was added to a methanol solution (15 mL) containing NaOH (3 mmol, 0.120 g) and Hntfa (1 mmol, 799 mg). This solution was stirred for 2 h at room temperature, then 10 mL of deionized water was added to the reaction mixture and stirred at 50 • C for another 6 h. The resulting solution was filtered through celite and then allowed to crystallize at room temperature. On the following day, the well-shaped shiny yellow crystals were filtered and dried in air (overall yield: 886 mg, 88%). Anal. Calcd. for C 44
The IR spectra of complexes 1-4 displayed general characteristic features, such as a strong vibrational band observed around 1608 ± 2 cm −1 that is typically assigned to the stretching frequency ν(C=O) of the coordinated carbonyl group of ntfa [47][48][49], as well as a series of medium intense bands over the frequency range of 1600-1506 cm −1 . In addition, complex 1 exhibited a weak broad band centered at 3410 cm −1 assigned to the ν(O-H) stretching frequency of the coordinated methanol ligands.

Description of the Crystal Structures
Partially labelled molecular plots of the title compounds 1-4 are presented in Figure 1, and coordination figures are depicted in Figure 2. In addition, the selected bond distances and bond angles of 1-4 are given in Table 2 The degree of distortion of the coordination polyhedra from their ideal polyhedron geometry were analyzed using the continuous shape measure theory with the SHAPE software [60,61]. Intermediate distortion from the ideal eight-vertex coordination polyhedra was observed for the LaO 8 Table 3 for the four compounds.    The packing plots of 1-4 are presented in Supplementary Figures S5-S8

Photoluminescent Properties
The UV-visible excitation spectra of solid samples of the complexes 1-4 represented in Supplementary Figure S9 show a slight redshift of the intense broad band over the 350-400 nm region, which may be taken as an indication of the complex formation. These bands located at 400, 394, 392, and 389 nm for 1-4, respectively, corresponded to π → π* electronic transition in the conjugated ligands coordinated to the central samarium(III) ion. Excitation of the solid sample at these wavelengths (λ ex ) led to the Sm 3+ -centered emission transitions in the visible and NIR regions.
The excitation and emission spectra of the compounds were recorded at room temperature in the visible and NIR regions and at the liquid nitrogen temperature (77 K) in the visible range; these spectra are shown in Figure 3 for the four complexes. Inspection of these spectra indicated an almost similar emission trend for all compounds. The bands in the visible region were assigned to 4 G 5/2 → 6 H 5/2 at 565 ± 1 nm, 4 G 5/2 → 6 H 7/2 at 608 ± 2 nm, and 4 G 5/2 → 6 H 9/2 at 649 ± 2 nm, while a less intense band at 714 ± 3 nm corresponded to the 4 G 5/2 → 6 H 11/2 transition. Additionally, the very-low-intensity band located at 535 nm could be discerned as the f-f transition from an upper emissive level to the ground state: 4 F 3/2 → 6 H 5/2 [12,62].
For compound 1, the 4 F 3/2 → 6 H 5/2 transition at room temperature was not detected, and this probably could be attributed to the low-intensity residual emission from the ligand in the 450-500 nm range. For 2-4, no emission from the ligand could be perceived, indicating a rather good energy transfer efficiency from the ligand's lowest triplet state to the resonant highest energy level of the Sm 3+ ion. Moreover, the presence of a broad band from the absorption of the ligands in the excitation spectra suggested a good antenna effect [63]. The most intense band observed for the 1-4 compounds, which corresponded to the 4 G 5/2 → 6 H 9/2 transition, was a hypersensitive band (electric dipole allowed transition, ∆J = 2) and due to its high intensity, it dominated in the final color that the compounds presented. Additionally, the 4 G 5/2 → 6 H 5/2 and 4 G 5/2 → 6 H 7/2 were magnetic dipole transitions (∆J = 0, 1 respectively), and the 4 G 5/2 → 6 H 9/2 / 4 G 5/2 → 6 H 5/2 ratio provided information about the polarizability of the Sm 3+ chemical environment. Ratio values of 6.25, 4.52, 4.62, and 4.17 were determined for compounds 1-4, respectively. The similar ratio values observed in compounds 2-4 indicated that their coordination environment was not altered when the NN-donor ligand was changed. The higher ratio value detected in 1 indicated a more polarized environment around the Sm 3+ ion, which obviously was attributed to the coordinated MeOH molecules [64]. The difference in the hypersensitive band intensity of 1-4 compounds was also perceived in the emission color, and could be seen by naked eye under a UV lamp and according to the CIE diagram represented in Figure 4.  The visible spectra of the compounds, which were also measured at the liquid nitrogen temperature (77 K), displayed the splitting of each band due to the crystal field perturbation, corresponding to the Stark sublevels of the Sm 3+ Kramer ion. The 6 H J energy levels of the four complexes should split to a maximum of J + 1/2 Stark according to symmetries lower than cubic for J half-integer values. Higher asymmetry around the Sm 3+ ion is related to an enhancement of the emission intensity [2,21]. Thus, based on the results obtained in the SHAPE measurements, compounds 1-4 presented symmetries lower than cubic (O h , O, T d , T h and T). Thus, it was expected that each band corresponding to different 6 H J energy levels should split into three for 6 H 5/2, four for 6 H 7/2 , five for 6 H 9/2 , and six for 6 H 11/2 . However, if the symmetry around the Sm 3+ ion was cubic, then the splitting due to the crystal field would be two, three, three, and four for each energy level, respectively [65,66]. The magnetic-dipole-allowed transition 4 G 5/2 → 6 H 7/2 was taken as a reference. For compounds, deconvolution of the 4 G 5/2 → 6 H 7/2 band was conducted, and the best fittings for the four compounds were performed when four Gaussian functions were used (Supplementary Figure S10), corroborating that the symmetry of the coordination polyhedron was lower than cubic. Moreover, the low-intensity transition 4 F 3/2 → 6 H 5/2 seen at 535 nm at room temperature measurements was not discerned at 77 K. The 4 F 3/2 level, which was close in energy to the emitting 4 G 5/2 level, was thermally populated at room temperature [52]. Furthermore, the 4 G 5/2 → 6 H 9/2 / 4 G 5/2 → 6 H 5/2 intensity ratio was measured for the 77 K spectra. The obtained values were 6.29 for 1, 4.55 for 2, 4.66 for 3, and 4.24 for 4. These values were very similar to those obtained from the room temperature spectra, suggesting that there was no change in the coordination Sm 3+ environment when the temperature changed.
The emission spectra of 1-4 enabled us to analyze the transition from the lowest emitting energy level to the ground state ( 4 G 5/2 → 6 H J ) for each compound, and as can be seen in Figure 3, this emission presented five different bands that corresponded to the five expected levels for J = 5/2-13/2 centered at 565, 606, 647, 711, and 904 nm for 1; at 564, 606, 649, 712, and 905 nm for 2; at 565, 611, 651, 717, and 906 nm for 3; and at 566, 610, 649, 714, and 903 nm for 4. These transitions provided information about the crystal-field energy of the samarium ions in the complexes, and allowed us to estimate the energy between the ground ( 6 H 5/2 ) and first-excited ( 6 H 7/2 ) J states. The energy differences between these two transitions were about 1197, 1228, 1335, and 1275 cm −1 for the 1-4 complexes, respectively. The energy separation between the 6 H 5/2 ground state of the samarium(III) and the firstexcited state 6 H 7/2 was evaluated as 1000 cm −1 [67], which was in good accordance with the calculated values for 1-4. From the energy separation between the first state 6 H 7/2 and the 6 H 5/2 ground state calculated previously from the emission spectra, we could evaluate the spin-orbit coupling parameter, λ, through the expression E(J) = λJ(J + 1)/2 [67]. These calculations led to λ values of 342, 351, 381, and 364 cm −1 for 1-4, respectively.
The luminescence quantum yield (QYs) in the solid state were determined using an integrating sphere. Only the visible emission range was determined. For 1, the QY could not be calculated due to the low emission intensity that the sample presented, evidencing the effect of high-energy oscillators in the effectiveness of radiative deactivation. For 2, 3, and 4, the close QY values were evaluated as 0.23, 0.25, and 0.20%, respectively.
Excited-state lifetimes (τ obs ) of the 4 G 5/2 emitting state in the 1-4 complexes were collected at the maximum emission band in the visible range ( 4 G 5/2 → 6 H 9/2 , 649 ± 2 nm), where the decay curves were fitted monoexponentially as (I(t) = I 0 · exp − t τ obs ) ( Figure 5), according to single emitting species for all Sm 3+ complexes. The measured τ obs values were in the 30-74 µs range. The lowest value of τ obs was obtained for the precursor compound. For compounds 2-4, τ obs increased when the NN-donor ligand was used, in the order: 5,5 -Me 2 bipy < bipy < 4,4 -Me 2 bipy. The same tendency was followed for the measured QYs. Compound 3 showed more luminescence intensity, suggesting an enhancement of the electron density around the Sm 3+ ion due to the electron-donating methyl groups in the 4,4 positions of the bipyridine ligand [69]. The photoluminescent data for QYs and τ obs are compiled in Table 4.   20 65 In general, Sm 3+ coordination compounds did not produce relatively high QYs because the energy gap between the lowest emitting sublevel 5 G 5/2 and the lower energy level 6 F 11/2 was 7500 cm −1 . This value was low compared to the energy difference from Tb 3+ or Eu 3+ ions, which were 12,500 cm −1 {∆E ( 5 D 0 → 7 F 6 )} and 14,800 cm −1 {∆E ( 5 D 4 → 7 F 6 )}, respectively. The small energy gap in Sm 3+ ion favored nonradiative relaxation processes that lowered the emission efficiency and luminescence lifetimes. However, samarium(III) coordination compounds are very interesting due to their capability of emission over a wide range within the electromagnetic spectrum, thus covering both the visible and NIR regions [70][71][72][73][74]. The results obtained here were similar to those for other previously published samarium(III)-β-diketonate systems [2,63,65,70]

Magnetic Properties DC Magnetic Susceptibility Studies
Powder samples of complexes 1-4 were measured under applied magnetic fields of 0.3 T (300-2 K). The data are plotted as χ M T products versus T in Figure 6. Magnetization dependence of the applied field at 2 K for compounds were also recorded and are shown in Figure 7. The magnetic measurements of 1-4 revealed that the χ M T values at 300 K were 0.35, 0.48, 0.35, and 0.29 cm 3 ·mol −1 ·K, respectively, which were larger than the theoretical value for a free Sm(III) ion (0.09 cm 3 ·mol −1 ·K) in the 6 H 5/2 ground state (g J = 2/7) [75], but in accordance with the room temperature χ M T values of previously reported Sm(III) complexes [76][77][78]. Upon cooling the samples, the χ M T values decreased practically linearly, reaching values of 0.04, 0.03, 0.04, and 0.03 cm 3 ·mol −1 ·K for compounds 1-4, respectively. These relatively small low temperature χ M T values found in the compounds could be attributed to the significant crystal field (CF) splitting under an anisotropic coordination environment [78].  Field dependence of the magnetization on the magnetic static applied field at T = 2 K for complexes 1-4 ( Figure 7) revealed no saturation in high fields, with similar values of 0.12, 0.09, 0.12, and 0.11 Nµ B at 5 T for 1-4, respectively. Taking into consideration the 4f 5 ground configuration of the Sm(III) ion value, the saturated magnetization should have been 5/7 Nµ B (Msat = g J ·J·Nµ B ; g J = 2/7, J = 5/2).