Circularly Polarized Luminescence from New Heteroleptic Eu(III) and Tb(III) Complexes

The complexes [Eu(bpcd)(tta)], [Eu(bpcd)(Coum)], and [Tb(bpcd)(Coum)] [tta = 2-thenoyltrifluoroacetyl-acetonate, Coum = 3-acetyl-4-hydroxy-coumarin, and bpcd = N,N′-bis(2-pyridylmethyl)-trans-1,2-diaminocyclohexane-N,N′-diacetate] have been synthesized and characterized from photophysical and thermodynamic points of view. The optical and chiroptical properties of these complexes, such as the total luminescence, decay curves of the Ln(III) luminescence, electronic circular dichroism, and circularly polarized luminescence, have been investigated. Interestingly, the number of coordinated solvent (methanol) molecules is sensitive to the nature of the metal ion. This number, estimated by spectroscopy, is >1 for Eu(III)-based complexes and <1 for Tb(III)-based complexes. A possible explanation for this behavior is provided via the study of the minimum energy structure obtained by density functional theory (DFT) calculations on the model complexes of the diamagnetic Y(III) and La(III) counterparts [Y(bpcd)(tta)], [Y(bpcd)(Coum)], and [La(bpcd)(Coum)]. By time-dependent DFT calculations, estimation of donor–acceptor (D–A) distances and of the energy position of the S1 and T1 ligand excited states involved in the antenna effect was possible. These data are useful for rationalizing the different sensitization efficiencies (ηsens) of the antennae toward Eu(III) and Tb(III). The tta ligand is an optimal antenna for sensitizing Eu(III) luminescence, while the Coum ligand sensitizes better Tb(III) luminescence {ϕovl = 55%; ηsens ≥ 55% for the [Tb(bpcd)(Coum)] complex}. Finally, for the [Eu(bpcd)(tta)] complex, a sizable value of glum (0.26) and a good quantum yield (26%) were measured.


■ INTRODUCTION
Circularly polarized luminescence (CPL) is a chiroptical phenomenon that is attracting more interest in materials chemistry and physics thanks to the broad range of possible biological 1−5 and technological applications, 6−15 as in the case of the design of organic light-emitting diodes (OLEDs) emitting circularly polarized (CP) light. 16−18 Focusing our attention on these latter devices, we can find a logical application in displays in which the emitted polarized light is exploited to prevent reflection of ambient light and obtain high-contrast three-dimensional images and true black. 19 In addition, a CP screen reduces the perceived distortion found at some angles when the display is viewed through a linearly polarized (LP) filter, significantly improving the outdoor viewing of laptops or smartphones. In this scenario, as the emission of CP light by trivalent lanthanide ions is a quite efficient phenomenon, luminescent lanthanide complexes play a pivotal role. 18 The new quantity recently proposed by some of us 20 and known as CPL brightness (B CPL ) can be considered as useful tool for the design of efficient CPL phosphors for specific chiroptical applications, such as in CPL microscopy 21 and possibly in CPL security inks. 7,22 Moreover, a high B CPL is correlated with a higher signal available for a CPL measurement. 23,24 B CPL takes into account the absorption extinction coefficient and quantum yield along with the g lum factor and, for a selected lanthanide transition, is defined as B CPL = β i ε λ ϕ ovl |g lum |/2 = β i B|g lum |/2, where β i is the so-called branching ratio (β; 0 ≤ β ≤ 1), ε λ is the molar extinction coefficient, ϕ ovl is the overall quantum yield, 25 and g lum 26 is the dissymmetry factor associated with the considered transition β i = I i /∑ j I j , where I i is the integrated intensity of the considered transition and ∑ j I j is the summation of the integrated intensities over all of the transitions. In the case of lanthanide complexes, optimal B CPL values can be obtained in the presence of strong absorbing chromophores (high ε λ values), good ligand-tometal energy transfer (LMET or antenna effect), and luminescence efficiency from Ln (III) (both contributing to high ϕ ovl values) . Finally, the emitted light should be strongly polarized, with the difference in the intensity between the right-and left-handed components of the CP light being as large as possible. A promising strategy for obtaining a Ln(III)based luminescent complex exhibiting sizable B CPL values is to design heteroleptic complexes in which two different ligands are bound to the metal ion. One ligand takes care of efficiently sensitizing the Ln (III) luminescence, while the other one can trigger the required CPL activity, by virtue of the fact that it is chiral and nonracemic. As for the antenna that can sensitize Eu (III) and Tb (III) luminescence, the anion derived from 3acetyl-4-hydroxycoumarin (Coum) 27 has been selected for both ions, while 2-thenoyltrifluoroacetyl-acetonate (tta) has been selected as the ligand for only Eu (III) (Figure 1). It is well-known that the latter ligand efficiently sensitizes the luminescence of the Eu(III) ion by virtue of the small energy gap between the accepting levels of the europium ion and the triplet state of the ligand and the increased anisotropy around the metal ion. 28 As the chiral bpcd ligand can induce a significant chiroptical response stemming from Eu (III) and Tb (III) ions, 29 we decided to employ this ligand to produce CPL activity from our complexes. Therefore, in this work, we synthesized and spectroscopically characterized three different complexes (in both enantiomeric forms), namely, [Eu(bpcd)(tta)], [Eu-(bpcd)(Coum)], and [Tb(bpcd)(Coum)]; the structures are shown in Figure 2.
Also, an ultraviolet−visible (UV−vis) titration study and density functional theory (DFT) calculations were carried out to characterize the stability and structural features of the complexes present in solution. Also, by means of timedependent DFT (TD-DFT) calculations, the energies of the tta and Coum excited states involved in the energy transfer mechanism were determined, and a possible explanation for the different sensitization efficiencies of the different complexes has been provided.
ESI-MS. Electrospray ionization mass spectra (ESI-MS) were recorded with a Finnigan LXQ Linear Ion Trap (Thermo Scientific, San Jose, CA) operating in positive ion mode. The data were acquired under the control of Xcalibur software (Thermo Scientific). A methanol solution of the sample was properly diluted and infused into the ion source at a flow rate of 10 μL/min with the aid of a syringe pump. The typical source conditions were as follows: transfer line capillary at 275°C; ion spray voltage of 4.70 kV; and sheath, auxiliary, and sweep gas (N 2 ) flow rates of 10, 5, and 0 arbitrary units, respectively. Helium was used as the collision damping gas in the ion trap set at a pressure of 1 mTorr.
The Coum ligand precursor was synthesized by modifying the procedure previously reported in the literature. 27 In a 50 mL roundbottom flask equipped with a condenser under a nitrogen atmosphere, 4-hydroxy coumarine (839 mg, 5.18 mmol), dry and freshly distilled pyridine (10 mL), and 4-(dimethylamino)pyridine (DMAP, 32 mg, 0.26 mmol) were added. To this stirred suspension was added acetic anhydride (0.57 mL, 529 mg, 5.18 mmol), and the mixture was stirred at 50°C for 24 h. Pyridine was evaporated under reduced pressure; the residue was dissolved in ethyl acetate (AcOEt, 20 mL), and 1 M aqueous HCl was added until the pH reached 1. The water phase was extracted with AcOEt (3 × 15 mL); the combined organic layers were dried over anhydrous Na 2 SO 4 , and the solvent was removed under reduced pressure. The crude was subjected to flash column chromatography (SiO 2 ; 85:15 cyclohexane/AcOEt) to afford 3acetyl-4-hydroxy-2H-chromen-2-one in 64% yield (673 mg, 3.3 mmol). Spectroscopic data ( 1 H NMR and 13 C NMR in Figures S11 and S12) are in agreement with those reported in the literature. 27 [Eu(bpcd)(tta)] (yield of 94%) was synthesized as follows ( Figure  3). At room temperature, KOH (60 mg, 1.07 mmol) was added to a methanol (8 mL) solution of the desired enantiomer (S,S or R,R) of H 2 bpcd (200 mg, 0.31 mmol, as trifluoroacetate salt). After 15 min, EuCl 3 ·6H 2 O (116 mg, 0.31 mmol) was added to form the first complex, [Eu(bpcd)]Cl. In another flask, KOH (21 mg, 0.37 mmol) was added to a methanol (4 mL) solution of Htta (2thenoyltrifluoroacetyl-acetone, 68 mg, 0.31 mmol). This mixture was slowly added to the previously prepared solution containing the first complex, and the final mixture was stirred for 1 h. Then, the solvent was removed under reduced pressure, and the desired product was obtained as a yellow powder upon extraction in dichloromethane (8 × 3 mL) followed by solvent removal in vacuo.    (Figure 4). At room temperature, EuCl 3 ·6H 2 O (113 mg, 0.31 mmol) or TbCl 3 ·6H 2 O (115.7 mg, 0.31 mmol) was added to a previously prepared methanol solution containing KOH (60 mg, 1.07 mmol) and H 2 bpcd (S,S or R,R) (200 mg, 0.31 mmol, as trifluoroacetate salt). In another flask, 3-acetyl-4-hydroxy-2H-chromen-2-one (63.3 mg, 0.31 mmol) was solubilized in methanol and added to a solution of NaOMe (16.7 mg, 0.31 mmol), in the same solvent (10 mL), with some drops of acetonitrile to improve its solubility. The deprotonated Coum ligand was slowly added to the solution containing the Ln(III) complex. The final mixture was stirred for 2 h. Then, the solvent was removed under reduced pressure, and the desired product was obtained as a yellow powder upon extraction in dichloromethane (8 × 3 mL) followed by solvent removal in vacuo.
Elemental Luminescence and Decay Kinetics. Room-temperature luminescence was measured with a Fluorolog 3 (Horiba-Jobin Yvon) spectrofluorometer, equipped with a Xe lamp, a double-excitation monochromator, a single-emission monochromator (mod. HR320), and a photomultiplier in photon counting mode for the detection of the emitted signal. All of the spectra were corrected for the spectral distortions of the setup. The spectra in solution were recorded on methanol (50 μM) solutions.
In decay kinetics measurements, a xenon microsecond flash lamp was used and the signal was recorded by means of a multichannel scaling method. True decay times were obtained using the convolution of the instrumental response function with an exponential function and the least-squares-sum-based fitting program (Spectra-Solve software package).
Circularly Polarized Luminescence. CPL spectra were recorded with the homemade spectrofluoropolarimeter described previously. 31 The spectra were recorded on methanol (0.4 mM) solutions in a 1 cm cell. Electronic Circular Dichroism (ECD). ECD spectra were recorded with a Jasco J1500 spectropolarimeter on 1 mM CH 3 OH solutions in a 0.01 cm cell.
Overall Quantum Yield Measurements. Overall quantum yields were measured by adopting the relative method. Fluorescein in 0.1 M NaOH (fluorescence quantum yield of 0.9) was used as the standard. Absorption spectra were recorded with a PerkinElmer Lambda 650 UV−vis spectrophotometer. Emission spectra were recorded with an Edinburgh FLS1000 fluorometer and corrected for excitation intensity and detector sensitivity. The samples were dissolved in methanol, while keeping their absorbance lower than 0.1. We obtained the same values of overall quantum yield for the two enantiomers.
Spectrophotometric Titrations. The UV−vis absorption spectra were recorded in the wavelength range of 220−500 nm on a Varian Cary 50 spectrophotometer using 1 cm optical path length quartz cells (Hellma Analytics) sealed with a Teflon stopper. The samples were prepared in dry methanol within a drybox with a N 2 atmosphere.
In our experiments, solutions of tta or Coum (initial concentrations of 5.0 × 10 −5 mol L −1 ) were titrated with a [Eu(bpcd)]Cl solution (1.0 × 10 −3 mol L −1 ) with a final ligand:complex molar ratio ∼3.0. The cell was stirred for ∼10 min before each spectrum was recorded. The formation constants of the [Eu(bpcd)(tta)] and [Eu(bpcd)-(Coum)] complexes were obtained by multiwavelength analysis of the absorption spectra using HypSpec. 32 DFT Calculations. All molecular structures of the complexes were obtained by means of DFT calculations performed in Gaussian 16 (version A.03). 33 In previous works, 30,34 the paramagnetic Eu(III) ion was replaced by Y(III), which was a suitable substitute. This choice is also supported for the isostructural complexes found with analogous hexadentate ligands EDTA and CDTA. In the crystal structures with the latter two ligands, 35−37 Y(III) and Eu (III) ions are ninecoordinated with EDTA (ligand and three water molecules bound to the metal) and eight-coordinated with CDTA (two bound water molecules). Because the ionic radius of Y(III) is smaller than those of Eu (III) and Tb (III), the calculations for the same complexes with the larger La(III) ion were also carried out. To identify the solvent molecules bound to the complexes, the geometries of [Ln(bpcd)-(Coum)]·4H 2 O (L = Y or La), in which water molecules replaced methanol, were also considered.
The functional B3LYP 38,39 was used with the 6-31+G(d) basis set for all ligand atoms and MWB28 pseudopotential and valence  Inorganic Chemistry pubs.acs.org/IC Article electron basis set for the metal ions. 40,41 Geometry optimizations were carried out at the DFT level with a polarizable continuum model (PCM) to simulate solvation. 42 As in a previous work, 43 the excited state (T 1 and S 1 ) energies were obtained by employing the time-dependent DFT approach (TD-DFT) on the [Y(bpcd)L] complexes using the same level of theory as in the geometry optimizations, as it was shown that the B3LYP functional accurately predicts the UV−vis spectra of coumarin derivatives. 44 Figure S1.
As it is well-known, the absorption band around 350 nm, in the case of [Eu(bpcd)(tta)], can be attributed to the diketonate-centered singlet−singlet π−π* transition of the tta ligand 46 while the absorption band peaking around 270 nm is assigned to electronic transitions involving both pyridine ring (i.e., π−π* and n−π*) transitions of the bpcd 2− ligand. 30 The lower-energy ECD bands indicate that the chiral ligand can dictate a preferred sense of twist of the diketonate, as demonstrated by a dichroic signal around 350 nm, where the absorption of tta takes place.
The UV−vis electronic absorption spectra of [Eu(bpcd)-(Coum)] and [Tb(bpcd)(Coum)] complexes are practically identical, demonstrating the ligand-centered nature of the electronic transitions, which is independent of the type of metal ion. The absorption peaks centered around 230, 300, and 320 nm, as previously demonstrated, 27 are related to Coum electronic transitions, while the shoulder at 270 nm is due to the absorption of the bpcd 2− ligand. Although both ECD spectra show an overall similar trend, the dichroic signals show some differences in intensity and shape especially in the range of 280−350 nm. Possibly, this could be due to the different size of the lanthanide cation (Eu 3+ vs Tb 3+ ), which in turn impacts the position of the Coum moiety relative to the bpcd ligand. Geometrical variations within a series of analogue lanthanide complexes are possible, especially across the socalled Gd break. For comparison, the ECD of the [Gd(bpcd)-(Coum)] analogue was also measured. It displayed a trend similar to that shown by the Eu complex but with a somehow stronger band at 270 nm ( Figure S1). Again, the CD signals in the range of 280−350 nm, which are associated with the achiral coumarin, indicate a stereochemically defined arrangement of the whole complex, including the coumarin.
UV−Vis Titrations. The absorption spectra of tta/Coum with an increasing concentration of [Eu(bpcd)]Cl were recorded in anhydrous methanol between 220 and 400 nm. The spectra of both systems present a maximum of absorption at 266 nm related to the pyridine group of the [Eu(bpcd)] + complex. 30 As for tta (Figure 5a), the initial absorption maximum at 340 nm undergoes a marked red-shift upon formation of the [Eu(bpcd)(tta)] complex in solution. On the contrary, for Coum (Figure 5b), the intensity of the initial absorption at 310 nm increases with a slight red-shift and a shoulder at 320 nm. The data were analyzed by simultaneous least-squares fitting of the absorbance data in the wavelength range associated with the formation of the new species (300− 380 and 290−340 nm for tta and Coum, respectively). As an example of the goodness of fit, calculated and experimental absorbances at selected wavelengths are reported in Figure S2. The fitting procedure confirms the formation of quite stable 1:1 species (Table 1). Interestingly, we show that the complex formed with Coum is more stable (∼1.5 log units) than that with tta. In previous works, the formation of the [EuL j ] 3−j species was studied in water for tta (log K 1 = 4.65, log β 2 = 9.67, and log β 3 = 12.0 at 25°C in water) 47 and mixed solvents for Coum (50% water/dioxane; log K 1 = 3.92 and log β 2 = 6.89 at 35°C). 48 However, our results (Table 1) are not comparable with the literature data for the 1:1 species. The reaction reported in those works indeed occurs between the   Figure S3. The excitation spectrum is very similar to the absorption spectrum, and the presence of two excitation peaks at 270 and 350 nm is related to an efficient ligand-to-metal energy transfer involving both ligands. The emission spectrum is dominated by the presence of the hypersensitive 5 D 0 → 7 F 2 transition of Eu (III), which is usually observed when the emitting Eu(III) is located in a site, whose point symmetry lacks the inversion center. The high value (9.54) of the asymmetry ratio is compatible with a highly distorted geometric environment around the metal ion. The luminescence decay curve of the 5 D 0 excited state of Eu (III) in methanol can be properly fitted by a singleexponential function (Figure 6, right). The calculated observed lifetime [0.68(1) ms] is reported in Table 2.
To estimate the number of methanol molecules in the inner coordination sphere of the Eu(III) ion (m), the observed lifetime was also measured in CD 3 OD (1.18 ms). The value obtained here for m is 1.3(5) and has been determined by using the equation m = 2.1(1/τ MeOH − 1/τ CDd 3 OD ) ( Table 2). 50 This number has a significant impact on the luminescence efficiency of the emitting lanthanide ion. In fact, the OH group of methanol possesses high-energy vibrations (3300−3400 cm −1 ), which are particularly effective in the nonradiative

Inorganic Chemistry
pubs.acs.org/IC Article quenching of the lanthanide emitting level, by means of the multiphonon relaxation process (MPR). 51 A higher m value would lead to stronger quenching of the Ln(III) luminescence. By using the equation reported by Werts, 52 which is known to be applicable to only the emission spectra of Eu(III), we calculated the radiative lifetime (τ rad ). We also determined the intrinsic quantum yield (ϕ int = τ obs /τ rad ; i.e., the ratio of emitting/absorbed photons upon direct excitation into a luminescent level of the lanthanide ion) and the overall quantum yield upon excitation of the tta ligand (ϕ ovl ; i.e., the ratio of emitting/absorbed photons upon excitation of the ligand) by using the secondary methods described in the literature 53 and η sens , which is the overall energy transfer efficiency (η sens = ϕ ovl /ϕ int ). It is worth noting that the obtained value of η sens is ∼100%, which underlines the high efficiency of the antenna effect of tta molecules in sensitizing the Eu(III) luminescence in the [Eu(bpcd)(tta)] complex. The CPL spectra of [Eu(bpcd)(tta)] ( Figure 6) show intense bands with g factors on the order of 10 −1 and 10 −2 for the 5 D 0 → 7 F 1 and 7 F 2 transitions, respectively (see Table 3 and Figure S4). Notably, three and two bands are clearly resolved for the 5 D 0 → 7 F 1 and 7 F 2 transitions, respectively, corresponding to the M J splitting due to the crystal field. [

Eu(bpcd)(Coum)] and [Tb(bpcd)(Coum)].
The luminescence excitation and emission spectra of a methanol solution (50 μM) of the complexes are shown in Figure S3. Both excitation spectra are very similar to the corresponding absorption spectra. The presence of the excitation bands at 270 and 320 nm indicates, also for these Coum-based complexes, the presence of an efficient ligand-to-metal energy transfer involving both ligands. As in the case of the [Eu(bpcd)(tta)] complex, the emission spectrum of [Eu-(bpcd)(Coum)] is dominated by the presence of the hypersensitive 5 D 0 → 7 F 2 transition of Eu (III) (Figure S3), although the geometric environment of the metal ion is less distorted than in the case of the tta-based complex, as suggested by the lower value of the asymmetry ratio (7.13). In the case of the [Tb(bpcd)(Coum)] complex upon excitation in the ligand absorption bands, the typical Tb(III) emission stemming from the f−f transition is detected ( Figure S3).
The luminescence decay curves of the excited states of Eu (III) and Tb(III) ( 5 D 0 and 5 D 4 levels, respectively) are fitted by a single-exponential function (Figure 7), and the calculated observed lifetimes (reported in Table 2) are 0.80(1) and 1.42(1) ms for the Eu (III) and Tb(III) complexes, respectively. As discussed above, the number of methanol molecules bound to the metal ion (m) can be obtained by measuring the observed lifetimes also in deuterated methanol. This number is slightly greater than 1 in the case of the [Eu(bpcd)(Coum)] complex and slightly less than 1 for [Tb(bpcd)(Coum)] ( Table 2). A possible explanation for this behavior is given in DFT Calculations. A similar trend was observed by Arauzo et al. 54 in heteroleptic Eu and Tb complexes containing the Coum and phenanthroline ligands. Interestingly, we can also conclude that the Coum ligand is not a good sensitizer for Eu(III) luminescence (as η sens is only ∼21%) but can better sensitize Tb(III) luminescence [ϕ ovl = 55%; η sens ≥ 55 (Table 2)]. This statement is in agreement with previous works on Tb (III) and Eu(III) tris chelates of Coum, 27 where the values of ϕ ovl are 29% and 12%, respectively, and coumarin dipicolinate Eu(III) complexes (ϕ ovl < 2%). 55 When heteroleptic complexes of the Coum ligand contain phenanthroline or bathophenathroline ligands, the overall quantum yields increase for both Eu(III) (40− 45%) and Tb(III) (58−76%). 54 This feature is related to the involvement of the phenanthroline-based ligands in the sensitization of Ln(III) luminescence. As expected, both CPL spectra of [Eu(bpcd)(Coum)] and [Tb(bpcd)(Coum)] (Figure 7) show mirror image spectra. Unlike the pattern of bands shown by [Eu(bpcd)(tta)] for the 5 D 0 → 7 F 1 transition, the coumarin complex has a single monosignate band (Figure 7). [Tb(bpcd)(Coum)] shows transitions associated with 5 D 4 → 7 F 6,5,4,3 , with the most prominent feature being the manifold 5 D 4 → 7 F 5 transition around 547 nm. Despite the relatively low g lum factors ( Figure  S6), because of its high quantum yield, [Tb(bpcd)(Coum)] shows high values of B CPL , comparable to those obtained for [Eu(bpcd)(tta)] (Table 3).    (III) and Tb(III) luminescent complexes. 57 DFT Calculations. DFT calculations were carried out to obtain structural information about the [Ln(bpcd)L] complexes by studying the diamagnetic analogues of the Eu and Tb complexes. As previously described, 30,58 several isomeric forms can be present, depending on the arrangement of the pyridine and acetate groups in the bpcd ligand, namely, trans-OO, trans-NN, and cis-OO,NN. These isomers can host either the Coum (Figure 8a−c) or tta (Figure 8d−f) ligand, which replaces the coordinated solvent molecules. The energy of the isomers does not differ significantly (ΔE < 0.3 and 1.0 kcal mol −1 for the adduct with Coum and tta, respectively), so that a mixture should be present in solution.
In our previous work, 30 the number of water molecules coordinated to the metal ion in the structures of [Y(bpcd)-H 2 O n ] + (n = 2−5) complexes was always 2. The additional waters remained dissociated and formed hydrogen bonds with the ligand or the coordinated ones. Nevertheless, a similar calculation for the [La(bpcd)H 2 O n ] + counterpart showed that more than two water molecules could interact directly with the metal ion.
The calculated number of coordinated methanol molecules in this study is 1.3 and 1.1 for Eu (III) and 0.6 for Tb (III). To confirm the solvent coordination, we considered the [Y-(bpcd)(Coum)]·4H 2 O complexes, in which water was employed in place of methanol to reduce the number of degrees of freedom of the system. The first result is that water cannot interact with the metal in all [Y(bpcd)(Coum)]·4H 2 O isomers (Figure 9a−c). Second, the differences in energy [ΔE = E(isomer) − E(a)] are 1.2 and 4.6 kcal mol −1 for structures b and c, respectively, in agreement with our previous observation that the cis-OO,NN isomer was the less stable one.
In the obtained minimum energy structures, the shortest metal−O water distances fall in the range of 4.2−4.5 Å. Even if it has been proposed that closely diffusing solvent molecules (like those shown in Figure 9a−c) can contribute to decrease the luminescence decay lifetime, the number of methanol molecules calculated from the experimental measurements is definitely greater than 1 for Eu (III).
In  (III) and Y(III) have ionic radii, longer and shorter than that of Eu(III), b respectively, it is likely that the small differences allow the coordination of an additional solvent molecule in solution. Interestingly, the experimental m ( Table  2) clearly decreases from Eu (1.3 or 1.1) to Tb (0.6), in agreement with the hypothesis that small changes in ionic radius here determine a different number of coordinated solvent molecules in the complex. Therefore, in this case, Y(III) models are more representative of the [Tb(bpcd)-(Coum)] coordination in solution.
The energies of the S 1 and T 1 excited states for both tta and Coum ligands were determined by TD-DFT calculations for the Y-and La-based complexes ( Table 4). The HOMO and LUMO molecular orbitals, mainly localized on these two ligands, are mostly involved in the S 0 → S 1 and S 0 → T 1 transitions ( Figures S7−S10), and the two most intense peaks at lower energies [around 315 nm for Coum-based complexes and around 350 nm for tta-based complexes (Figures 5 and 6)] are related to these S 0 → S 1 transitions. As previously demonstrated by some of us, at shorter wavelengths (at 270 nm), an additional electronic transition involving the molecular orbitals localized on the bpcd ligand can be exploited to sensitize Eu (III) 43 and Tb(III) 30 luminescence (not discussed here).
Assuming the participation of the excited triplet state in the antenna process, the sensitization of Tb(III) luminescence by tta can be ruled out. In fact, the lowest excited state of Tb(III) ( 5 D 4 ) is located around 20500 cm −1 ; that being above the energy of the triplet state of the ligand (19618 cm −1 averaged on the possible isomers) could be responsible for a back energy transfer mechanism (from the metal to the ligand) that makes the antenna process ineffective. On the contrary, the energy position of the T 1 state of tta seems to be optimal to transfer energy to the 5 D 0 emitting level of Eu(III) (located at 17300 cm −1 ), in agreement with the calculated η sens of ∼100% ( Table  2). As for the energy transfer mechanism in the case of the Coum ligand, both the emitting 5 D 4 level of Tb(III) located at 20500 cm −159 and the 5 D 2 level of Eu(III) (around 21500 cm −1 ) can be considered suitable acceptor states of the T 1 donor state of the Coum molecule [around 25600/26000 cm −1 ( Table 4)]. However, in our complexes, the sensitization efficiency of Eu(III) luminescence by the Coum ligand is lower than that of Tb (III) {η sens values around 20% and ≥55% for the [Eu(bpcd)(Coum)] and [Tb(bpcd)(Coum)] complexes, respectively}. Although a detailed study of the dynamics of the energy transfer process would be necessary to clearly understand the reasons underlying this behavior, a possible explanation for the lower sensitization efficiency in the case of the Eu(III) complex can be found in the D−A (donor− acceptor) distance (R L ) and the S 1 −T 1 energy gap (fifth and fourth columns, respectively, in Table 4). In fact, it is wellknown that the R L distance strongly affects the probability of energy transfer from D to A: the shorter the R L , the higher the Inorganic Chemistry pubs.acs.org/IC Article probability. 60 As it can be evinced from the inspection of Table  4, with an increase in the ionic radius of the lanthanide ion (passing from Y to La), the D−A distance increases. Therefore, we expect a sensitization efficiency that improves as the lanthanide ion becomes smaller. The decrease (|Δr| = 0.03 Å) in the ionic radius (and in turn of the R L distance) by passing from the Eu (III) to Tb(III) ion can justify the higher sensitization efficiency observed in the case of the Tb(III)based complex. Analogously, it seems that an increase in the S 1 −T 1 energy gap occurs with an increase in the size of the metal ion [from Y to La (Table 4) and presumably from Tb to Eu]. The increase in this energy gap is often related to a decrease of the intersystem crossing and sensitization efficiencies. Finally, as usual for lanthanide-based coordination compounds, in the energy transfer process we can assume the exchange mechanism as dominant, being the most sensitive to the D−A distance. 60 Therefore, its involvement could account for the significant decrease in sensitization efficiency, even if only a very small increase in the ionic radius (and in turn the R L distance) occurs by passing from the Tb (III) to Eu(III) ion.

■ CONCLUSIONS
Upon excitation of the antenna ligands, a moderate to good overall quantum yield is measured for the [Eu(bpcd)(tta)] (26%) and [Tb(bpcd)(Coum)] (55%) complexes. This is mainly due to the very good sensitization efficiency of the Eu (III) and Tb(III) luminescence by tta and Coum antenna ligands, respectively. A shorter D−A distance and a better intersystem crossing efficiency, in the case of [Tb(bpcd)-(Coum)], could account for the observed higher sensitization efficiency of the Coum ligand toward Tb(III) (≥55%), with respect to Eu(III) (21%). In addition, the access to the first coordination sphere of the metal ion by the methanol molecules, which decrease the luminescence efficiency, is more limited in the case of the Tb(III) complex (m = 0.6) than in the case of the Eu(III) complex (m = 1.1). This is probably due to an ionic size effect, Eu(III) being slightly larger than Tb(III) (Δr = 0.03 Å; r is the ionic radius in water). As suggested by the D−A distances (R L ), the Coum ligand is more tightly bound to the metal ion than tta, and this reflects the fact that the stability of [Eu(bpcd)(Coum)] (log Κ = 5.59) is higher than that of [Eu(bpcd)(tta)] (log Κ = 4.12). The ECD experiments suggested that the chiral bpcd ligand can dictate a preferred sense of twist of both tta and Coum ligands and the CPL activity of the magnetic dipole transition around 586 nm of Eu (III)  ■ ADDITIONAL NOTES a Because (i) ϕ int , ϕ ovl , and η sens can assume values between 0 and 1, (ii) ϕ ovl must be less than or equal to ϕ int , and (iii) ϕ ovl is 55% for the [Tb(bpcd)(Coum)] complex, we can conclude that η sens must be ≥55%.