Magnetic and Photo-Physical Properties of Lanthanide Dinuclear Complexes Involving the 4,5-Bis(2-Pyridyl-N-Oxidemethylthio)-4′,5′-Dicarboxylic Acid-Tetrathiafulvalene-, Dimethyl Ester Ligand

Abstract

The reaction between the 4,5-bis(2-pyridyl-N-oxidemethylthio)-4′,5′-dicarboxylic acid-tetrathiafulvalene-, dimethyl ester ligand (L) and the metallo-precursors Ln(hfac)₃·2H₂O leads to the formation of two dinuclear complexes of formula [Ln₂(hfac)₆(L)]·(CH₂Cl₂)·(C₆H₁₄)_0.5 (Ln^III = Dy^III (1) and Yb^III (2)). The X-ray structure reveals a quite regular square anti-prism symmetry for the coordination sphere of the lanthanide ion. UV-visible absorption properties have been experimentally measured and rationalized by TD-DFT calculations. The functionalization of the tetrathiafulvalene (TTF) core by two methyl ester moieties induces the appearance of an additional absorption band in the lowest-energy region of the spectrum. The latter has been identified as a HOMO (Highest Occupied Molecular Orbital)→LUMO (Lowest Unoccupied Molecular Orbital) Intra-Ligand Charge Transfer (ILCT) transition in which the HOMO and LUMO are centred on the TTF and methyl ester groups, respectively. Irradiation at 22,222 cm⁻¹ of this ILCT band induces an efficient sensitization of the Yb^III-centred emission that can be correlated to the magnetic properties.

1. Introduction

Since the discovery of the first mononuclear single molecule magnet (SMM) of lanthanide [1], f-elements have taken a preponderant place in the elaboration of coordination complexes in the field of molecular magnetism [2,3,4,5,6,7,8,9,10,11]. Lanthanide ions indeed have several advantages, such as strong single-ion anisotropy and large magnetic moment [12,13]. Recently, actinides have started to contribute to the development of new SMMs due to their strong magnetic anisotropy and larger exchange interactions coming from the more extended nature of the 5f orbitals [14,15,16,17,18,19,20]. On one hand, SMMs are molecular objects with potential applications in quantum computing, high-density memory data storage devices, and spintronics [21,22,23,24,25,26,27,28,29]. On the other hand, lanthanide ions are widely exploited for their specific luminescence which displays sharp f–f transitions ranging from the visible to the near infrared (NIR) region. Nevertheless, the very weak intensity of the f–f absorption bands (Laporte forbidden) [30] makes a direct excitation less straightforward, and consequently, lanthanide ions are usually coordinated to an organic ligand that strongly absorbs light in the UV-visible range. Thus, this organic ligand plays the role of organic chromophore for the sensitization of the lanthanide’s luminescence through an antenna effect [31,32]. In addition to the interest of lanthanide ions in coordination chemistry, a main challenge that chemists face is to use organic ligands that induce various behaviours such as redox activity, emission, magnetism, chirality, etc., and lead to multi-properties (functional) coordination complexes or SMMs [33,34,35]. In this context, a few years ago we started to combine lanthanide ions and tetrathiafulvalene (TTF)-based ligands which were used for their electron donating ability, redox activity and electronic conductivity [36,37,38,39,40,41,42,43,44]. We have already demonstrated that such ligands are suitable to obtain SMMs [45,46,47,48,49,50,51] and to sensitize lanthanide luminescence [52,53,54,55,56]. Previously, we associated the Ln(hfac)₃·2H₂O (hfac⁻ = 1,1,1,5,5,5-hexafluoroacetyacetonate) metallo-precursors with the (4,5-bis(2-pyridyl-N-oxidemethylthio)-4′,5′)-ethylenedithiotetrathiafulvene, -methyldithiotetrathiafulvene [57], or -4′,5′-ethylendioxotetrathiafulvalene ligands, leading to several dinuclear complexes [58].

In the present contribution we go one step forward, using the new 4,5-bis(2-pyridyl-N-oxidemethylthio)-4′,5′-dicarboxylic acid-tetrathiafulvalene-, dimethyl ester ligand (L). The coordination reactions of Ln(hfac)₃·2H₂O (Ln = Dy^III (1) and Yb^III (2)) with (L) are presented and the X-ray structure of [Dy₂(hfac)₆(L)]·(CH₂Cl₂)·(C₆H₁₄)_0.5 is described. The photo-physical and magnetic properties of 1 and 2 are presented and discussed.

2. Results and Discussion

2.1. Synthesis

The target ligand L (Scheme 1) was prepared by deprotection of the thiol function of the 4,5-dicarboxylic acid-4′,5′-bis(2-cyanoethyl)thio)-tetrathiafulvalene-, dimethyl ester [59] with sodium methoxide or cesium hydroxide and then grafting of the 2-methylpyridine-1-oxide coordinating groups.

Scheme 1. Chemical structure of the new 4,5-bis(2-pyridyl-N-oxidemethylthio)-4′,5′-dicarboxylic acid-tetrathiafulvalene-, dimethyl ester ligand, L.

Scheme 1. Chemical structure of the new 4,5-bis(2-pyridyl-N-oxidemethylthio)-4′,5′-dicarboxylic acid-tetrathiafulvalene-, dimethyl ester ligand, L.

L can be described as a multiply functionalized ligand with two pyridine-N-oxide moieties that are suitable for the coordination of lanthanide ions, leading generally to the formation of dinuclear complexes on one side and two ester groups on the other side, whose electron-withdrawing character will monitor the ligand’s electronic properties. These ester moieties could also be hydrolysed into carboxylate functions acting as additional coordinating sites [60].

2.2. Crystal Structure of [Dy₂(hfac)₆(L)]·(CH₂Cl₂)·(C₆H₁₄)_0.5 (1)

Compound 1 crystallizes in the Pcab (No.61) orthorhombic space group (Table 1 and Table 2). The asymmetric unit is composed of two Dy(hfac)₃ moieties, two L ligands and one dichloromethane molecule of crystallization. An ORTEP (Oak Ridge Thermal Ellipsoid Plot) view is depicted in Figure 1.

Table 1. X-ray crystallographic data for 1.

Compounds	[Dy2(hfac)6(L)2]·(CH2Cl2)·0.5C6H14 (1)
Formula	C78H51Cl2Dy2F36N4O24S12
M/g·mol−1	2892.84
Crystal system	Orthorhombic
Space group	Pcab (No.61)
Cell parameters	a = 17.3005(5) Å
	b = 34.2288(10) Å
	c = 37.6768(10) Å
Volume/Å3	22,311.3(11)
Z	8
T/K	150 (2)
2θ range/°	2.16 ≤ 2θ ≤ 54.96
ρcalc/g·cm−3	1.722
μ/mm−1	1.728
Number of reflections	103,297
Independent reflections	25,452
Rint	0.0815
Fo2 > 2σ(Fo)2	13,844
Number of variables	1396
R1, wR2	0.0805, 0.2175

Table 1. X-ray crystallographic data for 1.

Compounds	[Dy2(hfac)6(L)2]·(CH2Cl2)·0.5C6H14 (1)
Formula	C78H51Cl2Dy2F36N4O24S12
M/g·mol−1	2892.84
Crystal system	Orthorhombic
Space group	Pcab (No.61)
Cell parameters	a = 17.3005(5) Å
	b = 34.2288(10) Å
	c = 37.6768(10) Å
Volume/Å3	22,311.3(11)
Z	8
T/K	150 (2)
2θ range/°	2.16 ≤ 2θ ≤ 54.96
ρcalc/g·cm−3	1.722
μ/mm−1	1.728
Number of reflections	103,297
Independent reflections	25,452
Rint	0.0815
Fo2 > 2σ(Fo)2	13,844
Number of variables	1396
R1, wR2	0.0805, 0.2175

Table 2. Selected bond lengths (Å) for compound 1.

Compounds	1
Dy1–O9	2.263(8)
Dy1–O10	2.324(6)
Dy1–O11	2.376(7)
Dy1–O12	2.347(7)
Dy1–O13	2.359(7)
Dy1–O14	2.366(7)
Dy1–O15	2.381(7)
Dy1–O16	2.380(8)
Dy2–O17	2.344(7)
Dy2–O18	2.264(7)
Dy2–O19	2.359(7)
Dy2–O20	2.361(8)
Dy2–O21	2.352(8)
Dy2–O22	2.326(8)
Dy2–O23	2.382(7)
Dy2–O24	2.394(8)

Table 2. Selected bond lengths (Å) for compound 1.

Compounds	1
Dy1–O9	2.263(8)
Dy1–O10	2.324(6)
Dy1–O11	2.376(7)
Dy1–O12	2.347(7)
Dy1–O13	2.359(7)
Dy1–O14	2.366(7)
Dy1–O15	2.381(7)
Dy1–O16	2.380(8)
Dy2–O17	2.344(7)
Dy2–O18	2.264(7)
Dy2–O19	2.359(7)
Dy2–O20	2.361(8)
Dy2–O21	2.352(8)
Dy2–O22	2.326(8)
Dy2–O23	2.382(7)
Dy2–O24	2.394(8)

Figure 1. ORTEP view of 1 ([Dy₂(hfac)₆(L)₂]·(CH₂Cl₂)·0.5C₆H₁₄). Thermal ellipsoids are drawn at 30% probability. Hydrogen atoms and solvent molecules of crystallization are omitted for clarity.

Figure 1. ORTEP view of 1 ([Dy₂(hfac)₆(L)₂]·(CH₂Cl₂)·0.5C₆H₁₄). Thermal ellipsoids are drawn at 30% probability. Hydrogen atoms and solvent molecules of crystallization are omitted for clarity.

The X-ray structure reveals that two Dy(hfac)₃ moieties are bridged by two L ligands. The two water molecules of the starting metallo-precursor have been substituted by two pyridine-N-oxide groups. The Dy1 and Dy2 ions are surrounded by eight oxygen atoms that belong to three hfac⁻ ligands and the two pyridine-N-oxide coordinating acceptors. The average Dy–O distances are equal to 2.350(7) and 2.348(8) Å for Dy1 and Dy2, respectively. The arrangement of the ligands leads to a square antiprism (D_4d symmetry) as coordination polyhedron for both dysprosium ions (CShM_SAPR-8 = 0.385 for Dy1 and CShM_SAPR-8 = 0.396 for Dy2). The distortion is visualized by continuous shape measures performed with SHAPE 2.1 [61]. The Dy coordination polyhedra in 1 are the most symmetrical of all the series of similar dinuclear complexes that we have already obtained [57,58].

The central C=C bonds of the TTF core are equal to 1.342(13) and 1.340(16) Å, which attests the neutral form of L. The ester functions play an important role in the cohesion of the crystal packing. While the ester groups involving O5–O8 are almost parallel, those involving O1–O4 are almost perpendicular, in order to optimize the intermolecular S···O short contacts.

The combination of the shortest S···S contacts (3.668 Å (S3···S3) and 3.889 Å (S3···S6)) and shortest S···O contacts (3.281 Å (S8···O1), 3.364 Å (S1···O5), and 3.493 Å (S3···O7)) leads to the formation of tetramers of donors (Figure 2). The shortest intra- and inter-molecular Dy–Dy distances are equal to 10.052 Å (Dy1···Dy2), 11.039 Å (Dy2···Dy2), and 11.340 Å (Dy1···Dy1).

Figure 2. Crystal packing of 1 highlighting the formation of tetramers of L through short S···S and S···O contacts. Colour code: grey (C), red (O), orange (S), blue (D), green (F).

Figure 2. Crystal packing of 1 highlighting the formation of tetramers of L through short S···S and S···O contacts. Colour code: grey (C), red (O), orange (S), blue (D), green (F).

2.3. Electrochemical Properties

The redox properties of L and related complexes 1 and 2 ([Yb₂(hfac)₆(L)₂]·CH₂Cl₂·(C₆H₁₄)_0.5) are investigated by cyclic voltammetry (Figure S1) and the values of the oxidation potentials are listed in Table 3.

Table 3. Oxidation potentials (V vs. saturated calomel electrode, nBu₄NPF₆, 0.1 M in CH₂Cl₂ at 100 mV·s⁻¹) of the ligand L and complexes 1 and 2.

	E11/2/V		E21/2/V			E31/2/V
	OxE11/2	redE11/2	OxE21/2	redE21/2	OxE31/2		redE31/2
L	0.79	0.70	1.17	1.09	/		/
1	0.76	0.67	1.15	1.02	1.28		1.20
2	0.76	0.66	1.13	0.99	1.28		1.20

Table 3. Oxidation potentials (V vs. saturated calomel electrode, nBu₄NPF₆, 0.1 M in CH₂Cl₂ at 100 mV·s⁻¹) of the ligand L and complexes 1 and 2.

	E11/2/V		E21/2/V			E31/2/V
	OxE11/2	redE11/2	OxE21/2	redE21/2	OxE31/2		redE31/2
L	0.79	0.70	1.17	1.09	/		/
1	0.76	0.67	1.15	1.02	1.28		1.20
2	0.76	0.66	1.13	0.99	1.28		1.20

The cyclic voltammogram for L shows two mono-electronic oxidations at about 0.75 and 1.13 V, corresponding to the formation of a radical cation and a dication TTF fragment, respectively (Figure S1). These oxidation potentials are higher than those found for the functionalized TTF donor used in the synthesis of the other dinuclear complexes, due to the direct functionalization of the TTF core with two methyl ester groups [57,58]. The electrochemistry does not highlight significant effects on the oxidation potentials upon coordination of the electron attracting Ln(hfac)₃ fragments (Table 3). Nevertheless, an additional quasi-reversible oxidation wave is observed for the two complexes around 1.24 V. No clear explanation can be given for the origin of this additional redox activity at this point. The reversibility of the oxidation potentials is conserved and the electrochemical properties attest to the redox activity of L in the complexes.

2.4. Photo-Physical Properties

2.4.1. Absorption Properties

The UV-visible absorption properties of L and 2 have been studied in a CH₂Cl₂ solution and in a KBr pellet in solid-state (Figure 3a and Figure 4a, Figures S2 and S3). Rationalization by TD-DFT calculations was performed on L and the Y(III) analogue of 1 and 2 (Figure 3b and Figure 4b) following a computational strategy already used successfully on TTF-based systems [62]. The molecular orbital diagram is sketched in Figure 5.

Figure 3. (a) Experimental UV-visible absorption spectra of L in CH₂Cl₂ solution (C = 4 × 10⁻⁵ mol·L⁻¹) (open grey circles). Respective Gaussian decompositions (red, orange, and blue dashed lines correspond to ILCT, intra-donor and intra-acceptor transitions, respectively) and best fit (full black line) (R = 0.9994) (b) Theoretical absorption spectra of L (black line). The bars represent the mean contributions of the absorption spectra.

Figure 3. (a) Experimental UV-visible absorption spectra of L in CH₂Cl₂ solution (C = 4 × 10⁻⁵ mol·L⁻¹) (open grey circles). Respective Gaussian decompositions (red, orange, and blue dashed lines correspond to ILCT, intra-donor and intra-acceptor transitions, respectively) and best fit (full black line) (R = 0.9994) (b) Theoretical absorption spectra of L (black line). The bars represent the mean contributions of the absorption spectra.

Figure 4. (a) Experimental UV-visible absorption spectra in CH₂Cl₂ solution of 2 (C = 4 × 10⁻⁵ mol·L⁻¹) (open grey circles). Respective Gaussian decompositions (green dashed line corresponds to intra-hfac transitions) and best fit (full black line) (R = 0.9998); (b) Theoretical absorption spectra of the Y^III analogue (black line). The bars represent the mean contributions of the absorption spectra.

Figure 4. (a) Experimental UV-visible absorption spectra in CH₂Cl₂ solution of 2 (C = 4 × 10⁻⁵ mol·L⁻¹) (open grey circles). Respective Gaussian decompositions (green dashed line corresponds to intra-hfac transitions) and best fit (full black line) (R = 0.9998); (b) Theoretical absorption spectra of the Y^III analogue (black line). The bars represent the mean contributions of the absorption spectra.

Figure 5. Molecular orbital diagram for L (left) and the Y^III analogue (right). The energy levels of the centred TTF, 2-methylpyridine-N-oxide, methyl ester, and hfac⁻ orbitals are represented in orange, blue, black, and green, respectively.

Figure 5. Molecular orbital diagram for L (left) and the Y^III analogue (right). The energy levels of the centred TTF, 2-methylpyridine-N-oxide, methyl ester, and hfac⁻ orbitals are represented in orange, blue, black, and green, respectively.

The experimental absorption curve of L has been decomposed into seven bands (Figure 3a and Figure S2, Table 4). The calculated UV-visible absorption spectrum for L reproduces the experimental curve well (Figure 3b). The low energy bands (red Gaussian decomposition) were attributed to π–π* HOMO→LUMO and HOMO→LUMO + 1 excitations. While the latter is identified as the classical TTF to Methyl-2-Py-N-oxide charge transfer (ILCT) (Table 4) [57,58], the lowest-energy excitation is identified as TTF to methyl ester charge transfer. Thus, the functionalization of the TTF core with methyl ester groups has successfully increased the intensity of the ILCT and induced the appearance of new ILCT at lower-energy (22,600 cm⁻¹, calculated at 18,810 cm⁻¹) than the TTF to Methyl-2-Py-N-oxide ILCT (27,600 cm⁻¹, calculated at 24,463 cm⁻¹). The remaining absorption bands were respectively identified as Intra-Acceptor (IA) (where the acceptor is the Py-N-oxide moiety), Intra-donor (ID), and ILCT excitations (Table 4).

Table 4. TD-DFT calculated excitations energies and main composition of low-lying electronic transitions for L and the Y(III) analogue of 1 and 2. In addition, charge transfer and the pure intramolecular transitions are reported. ID, IA, Ihfac, H, and L represent respectively the intramolecular TTF (Donor), 2-pyridine-N-oxide (Py) (Acceptor) or hfac⁻ ligand, HOMO and LUMO; therefore ILCT stands for Intra-Ligand Charge Transfer (from TTF to 2-pyridine-N-oxide (PyNO) or methyl ester (ester)). The theoretical values are evaluated at the PCM(CH₂Cl₂)-PBE0/SVP level of approximation. Bold style is used to highlight the main contributions. Exp.: experimental, Th.: theoretical, Osc.: oscillator strength.

	Energy Exp. (cm−1)	Energy Th. (cm−1)	Osc.	Type	Assignment	Transition
L	22,600	18,810	0.03	ILCT	πTTF→π*Ester	H→L (99%)
	27,600	24,463	0.03	ILCT	πTTF→π*PyNO	H→L + 1 (88%)
	30,600	32,414	0.06	IA	πPyNO→π*PyNO	H-1→L + 1 (74%)
		32,857	0.07			H-2→L + 1 (70%)
	32,200	33,652 34,426 34,640	0.18 0.22 0.11	ID + ILCT	πTTF→π*TTF + πTTF→π*PyNO	H→L + 7 (60%) H-3/-4→L + 1 (43/28%)
	34,000	36,605	0.11	ID + IA	πTTF→π*TTF +	H→L + 7/ + 9 (10/44%)
		36,788	0.14		πPyNO→π*PyNO	H-1→L + 2/ + 5/ + 6 (9/27/10%)
	38,200	39,635 40,437 40,773 41,331 41,698	0.08 0.05 0.07 0.06 0.04	ID + ILCT	πTTF→π*TTF + πTTF→π*PyNO	H-3/-4→L + 4 (38/40%) H-5→L + 1 (38%) H-3→L + 2 (25%) H-10→L (41%)
	44,400	/	/	IA	πPyNO→π*PyNO	/
Y	22,600	19,282	0.03	ILCT	πTTF→π*Ester	H/-1→L (76/22%)
		20,840	0.04			H/-1→L + 5 (22/72%)
	27,300	22,985	0.02	ILCT	πTTF→π*PyNO	H→L + 6/ + 8/ + 9 (20/19/21%)
		23,132	0.02			H-1→L + 6/ + 9 (17/33%)
	31,000	33,234 33,485	0.30 0.15	ID + ILCT	πTTF→π*TTF + πTTF→π*Ester	H/-1→L + 20 (9/30%) H→L + 21 (15%) H-4/-5→L (21/29%)
	33,200	36,440 36,455 36,618 36,639	0.27 0.10	Ihfac	πhfac→π*hfac	H→L + 10 (53%) + H-3→L (65%)
	34,500	34,432 34,886	0.38	ID	πTTF→π*TTF	H/-1→L + 21 (12/41%) H/-1→L + 22 (9/30%)
	36,600	35,662	0.12	ILCT	πTTF→π*Ester	H-4/-5→L + 5 (13/7%)
	40,000	/	/	ILCT	πTTF→π*PyNO	/
	44,700	/	/	IA	πL2→π*PyImPy	/

Table 4. TD-DFT calculated excitations energies and main composition of low-lying electronic transitions for L and the Y(III) analogue of 1 and 2. In addition, charge transfer and the pure intramolecular transitions are reported. ID, IA, Ihfac, H, and L represent respectively the intramolecular TTF (Donor), 2-pyridine-N-oxide (Py) (Acceptor) or hfac⁻ ligand, HOMO and LUMO; therefore ILCT stands for Intra-Ligand Charge Transfer (from TTF to 2-pyridine-N-oxide (PyNO) or methyl ester (ester)). The theoretical values are evaluated at the PCM(CH₂Cl₂)-PBE0/SVP level of approximation. Bold style is used to highlight the main contributions. Exp.: experimental, Th.: theoretical, Osc.: oscillator strength.

	Energy Exp. (cm−1)	Energy Th. (cm−1)	Osc.	Type	Assignment	Transition
L	22,600	18,810	0.03	ILCT	πTTF→π*Ester	H→L (99%)
	27,600	24,463	0.03	ILCT	πTTF→π*PyNO	H→L + 1 (88%)
	30,600	32,414	0.06	IA	πPyNO→π*PyNO	H-1→L + 1 (74%)
		32,857	0.07			H-2→L + 1 (70%)
	32,200	33,652 34,426 34,640	0.18 0.22 0.11	ID + ILCT	πTTF→π*TTF + πTTF→π*PyNO	H→L + 7 (60%) H-3/-4→L + 1 (43/28%)
	34,000	36,605	0.11	ID + IA	πTTF→π*TTF +	H→L + 7/ + 9 (10/44%)
		36,788	0.14		πPyNO→π*PyNO	H-1→L + 2/ + 5/ + 6 (9/27/10%)
	38,200	39,635 40,437 40,773 41,331 41,698	0.08 0.05 0.07 0.06 0.04	ID + ILCT	πTTF→π*TTF + πTTF→π*PyNO	H-3/-4→L + 4 (38/40%) H-5→L + 1 (38%) H-3→L + 2 (25%) H-10→L (41%)
	44,400	/	/	IA	πPyNO→π*PyNO	/
Y	22,600	19,282	0.03	ILCT	πTTF→π*Ester	H/-1→L (76/22%)
		20,840	0.04			H/-1→L + 5 (22/72%)
	27,300	22,985	0.02	ILCT	πTTF→π*PyNO	H→L + 6/ + 8/ + 9 (20/19/21%)
		23,132	0.02			H-1→L + 6/ + 9 (17/33%)
	31,000	33,234 33,485	0.30 0.15	ID + ILCT	πTTF→π*TTF + πTTF→π*Ester	H/-1→L + 20 (9/30%) H→L + 21 (15%) H-4/-5→L (21/29%)
	33,200	36,440 36,455 36,618 36,639	0.27 0.10	Ihfac	πhfac→π*hfac	H→L + 10 (53%) + H-3→L (65%)
	34,500	34,432 34,886	0.38	ID	πTTF→π*TTF	H/-1→L + 21 (12/41%) H/-1→L + 22 (9/30%)
	36,600	35,662	0.12	ILCT	πTTF→π*Ester	H-4/-5→L + 5 (13/7%)
	40,000	/	/	ILCT	πTTF→π*PyNO	/
	44,700	/	/	IA	πL2→π*PyImPy	/

For 2, the absorption spectrum has been decomposed into eight bands for the solution and solid-state measurements.

The absorption spectra for the free ligand and the dinuclear complex are similar. Only an additional intense absorption band is observed around 33,000 cm⁻¹ that corresponds to π–π* intra-hfac⁻ excitations [62]. Complexation induces a very weak red shift of the ILCT transitions involving the Py-N-oxide acceptor (300 cm⁻¹, calculated red shift 1400 cm⁻¹), due to the Lewis acid behaviour of the Ln(hfac)₃ moieties enforcing the electron withdrawing character of the 2-pyridine-N-oxide fragments. This effect is weak because of the poor electronic communication through the methylthio arms. Nevertheless, the red-shift of these absorption bands in coordination complexes compared to those in L is a first indication of the stability of the dinuclear complexes in such solvents.

2.4.2. Emission Properties

No significant emission for the free ligand was detected, as already observed in previous TTF-based ligands in which the TTF core is not conjugated with the coordinated moieties [57,58]. Emission properties of 2 were measured in solid state at 77 K (Figure 6). The characteristic luminescence profile of Yb^III, corresponding to the ²F_5/2→²F_7/2 transition, is observed upon irradiation at 22,222 cm⁻¹ (450 nm). This low-energy irradiation is possible due to the ILCT involving the methyl ester groups. No residual emission centred on the ligand L was observed in the visible range after coordination. Based on the previous published TTF-based complexes of Yb^III [57,63,64], it is expected that in the antenna-effect sensitization process, such energy transfer takes place between the singlet CT excited state of the chromophore to the ²F_5/2 states of the Yb^III ion.

Four main emission maxima are clearly identified at the following energies: 9718, 9843, 9980, and 10,204 cm⁻¹. The total splitting is determined to be equal to 486 cm⁻¹. The values of this splitting for an Yb^III ion in a distorted and regular D₃ symmetry are 455 and 372 cm⁻¹, respectively [65], while a splitting of 528 cm⁻¹ is found for an Yb complex in a lower symmetry, increasing up to 880 cm⁻¹ for organometallic derivatives [66]. The value of 486 cm⁻¹ thus seems to correspond to a quite low symmetry. Nevertheless, this value is lower than the one found in a dinuclear analogue (508 cm⁻¹) in which the distortion of the square anti-prism symmetry of the Yb^III coordination sphere is more pronounced (CShM_SAPR-8 = 0.484).

Figure 6. Luminescence spectrum of 2 in the Near Infrared (NIR) range for λ_ex = 22,222 cm⁻¹ (450 nm) in solid-state at 77 K.

Figure 6. Luminescence spectrum of 2 in the Near Infrared (NIR) range for λ_ex = 22,222 cm⁻¹ (450 nm) in solid-state at 77 K.

2.5. Magnetic Properties

The temperature dependence of χ_MT, with χ_M being the molar magnetic susceptibility and T being the temperature in Kelvin, in powdered samples of compounds 1 and 2 are represented in Figure 7. The room temperature values are equal to 27.6 and 4.69 cm³ K·mol⁻¹ for 1 and 2, respectively. These values are in good agreement with the two expected values for the two multiplet ground states: ⁶H_15/2 (14.17 × 2 cm³ K·mol⁻¹) and ²F_7/2 (2.57 × 2 cm³ K·mol⁻¹) for two Dy^III and two Yb^III ions, respectively [67]. χ_MT decreases on cooling due to the crystal field splitting of the multiplet ground states down to 20.2 cm³·K·mol⁻¹ for 1, and 1.03 cm³·K·mol⁻¹ for 2. At 2 K, the magnetization saturates around 10 Nβ for 1 and 2.8 Nβ for 2. Neither of the two complexes show frequency dependence of ac susceptibility.

Figure 7. (a) Temperature dependence of χ_MT for 1. Inset: field variation of the magnetization measured at 2 K; (b) Temperature dependence of χ_MT for 2. Inset: field variation of the magnetization measured at 2 K. The red line corresponds to the best-fitted curve (see text). χ_M, molar magnetic susceptibility; T, temperature in Kelvin.

Figure 7. (a) Temperature dependence of χ_MT for 1. Inset: field variation of the magnetization measured at 2 K; (b) Temperature dependence of χ_MT for 2. Inset: field variation of the magnetization measured at 2 K. The red line corresponds to the best-fitted curve (see text). χ_M, molar magnetic susceptibility; T, temperature in Kelvin.

2.6. Correlation Magnetism-Luminescence

In the last five years, we have proven that magnetism and luminescence are intimately correlated [50,54,56,57,64]. In the case of compound 2, the excitation ²F_5/2→²F_7/2 can be viewed as a picture of the crystal field splitting of the ground state multiplet, while temperature dependence of the magnetic response can be viewed as a thermal picture of the same ground state multiplet splitting. In the first approximation, the coordination polyhedron of Yb^III in 2 has a D_4d symmetry (CShM_SAPR-8~0.4), with an environment close to an antiprismatic square. In this approximation, and in the formalism of Stevens, the ground state splitting can be accounted for by the following Hamiltonian for one metallic centre:

$$\hat{H}=B_2^0\widehat{O_2^0}+B_4^0\widehat{O_4^0}+B_6^0\widehat{O_6^0}$$

Where $\widehat{{\text{O}}_{\text{k}}^{\text{q}}}$ are the operators equivalents expressed as polynomials of the total angular momentum ($\widehat{J^2}$, $\widehat{J_Z}$, $\widehat{J_{+}}$, and $\widehat{J_{-}}$), and $B_k^q$ are connected to the crystal field parameters [68,69]. The perturbation due to the application of an external magnetic field can be easily estimated and the magnetization calculated. In a D_4d environment, only the terms with q = 0 and k = 2, 4, and 6 are non zero [70]. The temperature variation of χ_MT in 2 is fitted and the best curve is represented in Figure 7b (Yb1 and Yb2 are treated identically since they have a very similar environment). To ensure that the obtained set of parameters is correct, the M vs. H curve is also calculated at 2 K (inset of Figure 7b). In a D_4d symmetry M_J is a good quantum number, so the eigenstates are pure M_J states. In this frame, the ground state corresponds to M_J = ±5/2, which is taken at the energy origin. The first excited state (M_J = ±3/2) is located at 215 cm⁻¹, the second at 468 cm⁻¹, and the last at 569 cm⁻¹ above the ground state. The gap between the first two levels almost perfectly matches the gap between the two emission lines at 10,204 and 9980 cm⁻¹. The last two emission lines (9843 and 9780 cm⁻¹) are not so well reproduced, which might be due to the relative insensitivity of the magnetism to the high energy levels. Indeed, the last two levels are almost not thermally populated, even at room temperature.

3. Experimental Section

3.1. Synthesis

General Procedures and Materials: The precursors Ln(hfac)₃·2H₂O (Ln = Dy^III and Yb^III, hfac⁻ = 1,1,1,5,5,5-hexafluoroacetylacetonate anion) [71] and the 4,5-dicarboxylic acid-4′,5′-bis(2-cyanoethyl)thio)-tetrathiafulvalene-, dimethyl ester [59] were synthesised following previously reported methods. All other reagents were purchased from Aldrich Co., Ltd. (Saint-Quentin Fallavier, France) and used without further purification.

3.2. Synthesis of 4,5-Bis(2-Pyridyl-N-Oxidemethylthio)-4′,5′-Dicarboxylic Acid-Tetrathiafulvalene-, Dimethyl Ester (L)

Method A. A solution of 0.5 M EtONa/EtOH (14 mL) was added to a suspension of 4,5-dicarboxylic acid-4′,5′-bis(2-cyanoethyl)thio)-tetrathiafulvalene-, dimethyl ester [72] (0.676 g, 1.38 mmol) in anhydrous degassed EtOH (41 mL) under argon. After being stirred at room temperature for 4 h, the mixture was reacted with a solution of 2-(chloromethyl)pyridine-1-oxide [73] (0.593 g, 4.13 mmol) in anhydrous degassed EtOH (27 mL), and then the mixture was stirred for 16 h. H₂O (41 mL) was added to quench the reaction, and the mixture was poured into CH₂Cl₂ (345 mL), washed with saturated NaHCO₃ (3 × 55 mL) and water (210 mL). The organic extract was concentrated in vacuum to give a red-brown oil, which was purified by chromatography on alumina gel, initially with CH₂Cl₂ (to remove the unreacted alkyl halide), and then with CH₂Cl₂/MeOH (10:1) to give the pure desired ligand. Yield: 437 mg (53%). Anal. Calcd (%) for C₂₂H₁₈N₂O₆S₆: C 44.15, H 3.01, N 4.68; found: C 44.03, H 3.05, N 4.74. ¹H-NMR (CDCl₃): 8.24–8.20 (m, 4H), 7.31–7.17 (m, 4H), 3.80 (s, 6H), 3.34 (s, 4H). I.R. (KBr): 2967 (w), 1738 (s), 1724 (s), 1584 (m), 1428 (m), 1415 (m), 1253 (s), 1092 (m), 1020 (m), 886 (w), and 769 (w) cm⁻¹.

Method B. CsOH·H₂O (335 mg, 2.00 mmol) dissolved in a minimum of MeOH was slowly added to 4,5-dicarboxylic acid-4′,5′-bis(2-cyanoethyl)thio)-tetrathiafulvalene-, dimethyl ester (0.490 g, 1.00 mmol) in 10 mL of DMF. After being stirred at room temperature for 30 min, the mixture was reacted with a solution of 2-(chloromethyl)pyridine-1-oxide [73] (0.430 g, 2.99 mmol) in anhydrous degassed MeOH (25 mL), and then the mixture was stirred overnight. Then the organic solvents were removed under vacuum and the residue extracted with CH₂Cl₂ and washed with H₂O. The organic phase was extracted and dried with MgSO₄. Pure ligand was obtained by chromatography on alumina gel, initially with CH₂Cl₂ (to remove the unreacted alkyl halide) and then with CH₂Cl₂/MeOH (10:1). Yield: 486 mg (59%).

3.3. Synthesis of Complexes 1 and 2

[Dy₂(hfac)₆(L)₂]·CH₂Cl₂·(C₆H₁₄)_0.5 (1). 32.8 mg of Dy(hfac)₃·2H₂O (0.04 mmol) were dissolved in 10 mL of CH₂Cl₂ and then added to a solution of 10 mL of CH₂Cl₂ containing 24.0 mg of L (0.04 mmol). After 15 min of stirring, 25 mL of n-hexane were layered at room temperature in the dark. Slow diffusion leads to dark red single crystals, which are suitable for X-ray studies. Yield: 100 mg (86%). Anal. Calcd (%) for C₇₅H₄₄Cl₂Dy₂F₃₆N₄O₂₄S₁₂: C 31.58, H 1.54, N 1.97; found: C 31.69, H 1.62 N, 2.01. I.R. (KBr): 2956 (w), 2928 (w), 2870 (w), 1734 (m), 1653 (s), 1556 (m), 1531 (m), 1508 (m), 1447 (w), 1256 (s), 1211 (s), 1144 (s), 1101 (w), 798 (w), 769 (w), 661 (m), and 587 (w) cm⁻¹.

[Yb₂(hfac)₆(L)₂]·CH₂Cl₂·(C₆H₁₄)_0.5 (2). This compound was obtained using the same protocol as for 1, starting from 33.2 mg of Yb(hfac)₃·2H₂O (0.04 mmol) instead of Dy(hfac)₃·2H₂O. Yield: 82 mg (71%). Anal. Calcd (%) for C₇₅H₄₄Cl₂Yb₂F₃₆N₄O₂₄S₁₂: C 31.47, H 1.54, N 1.96; found: C 31.57, H 1.60, N, 2.03. I.R. (KBr): 2954 (w), 2928 (w), 2871 (w), 1734 (m), 1657 (s), 1558 (m), 1533 (m), 1513 (m), 1448 (w), 1264 (s), 1216 (s), 1151 (s), 799 (w), 769 (w), 662 (m), and 588 (w) cm⁻¹.

3.4. Crystallography

A single crystal of 1 was mounted on a APEXII Bruker-AXS diffractometer for data collection (MoK_α radiation source, λ = 0.71073 Å), at the Centre de Diffractométrie (CDIFX), Université de Rennes 1, France. The structure was solved with a direct method using the SIR-97 program and refined with a full matrix least-squares method on F² using the SHELXL-97 program [74,75]. Crystallographic data are summarized in Table 1. Complete crystal structure results as a CIF file including bond lengths, angles, and atomic coordinates are deposited as Supplementary files.

3.5. Physical Measurements

The elementary analyses of the compounds were performed at the Centre Régional de Mesures Physiques de l’Ouest, Rennes. Cyclic voltammetry was carried out in CH₂Cl₂ solution, containing 0.1 M N(C₄H₉)₄PF₆ as supporting electrolyte. Voltammograms were recorded at 100 mV·s⁻¹ with a platinum disk electrode. The potentials were measured versus a saturated calomel electrode (SCE). Absorption spectra were recorded on a Varian Cary 5000 UV-Visible-NIR spectrometer equipped with an integration sphere. The luminescence spectra were measured using a Horiba-Jobin Yvon Fluorolog-3^® spectrofluorimeter, equipped with a three slit double grating excitation and emission monochromator with dispersions of 2.1 nm/mm (1200 grooves/mm). The steady-state luminescence was excited by unpolarized light from a 450 W xenon CW lamp. Quartz tubes containing the samples were immersed in liquid nitrogen and near infra-red spectra were recorded at right angle using a liquid nitrogen cooled, solid indium/gallium/arsenic detector (850–1600 nm). Spectra were reference corrected for both the excitation source light intensity variation (lamp and grating) and the emission spectral response (detector and grating). The dc magnetic susceptibility measurements were performed on a solid polycrystalline sample with a Quantum Design MPMS-XL SQUID magnetometer between 2 and 300 K in applied magnetic field of 0.2 T for temperatures of 2–20 K and 1 T for temperatures of 20–300 K. These measurements were all corrected for the diamagnetic contribution calculated with Pascal’s constants.

3.6. Computational Details

DFT geometry optimizations and TD-DFT excitation energy calculations of the ligand L and Y^III analogue of the dinuclear complexes were carried out with the Gaussian 09 (revision A.02) package [76] employing the PBE0 hybrid functional [77,78]. The “Stuttgart/Dresden” basis sets and effective core potentials were used to describe the yttrium atom [79], whereas all other atoms were described with the SVP basis sets [80]. The first 50 monoelectronic excitations were calculated for ligand L while the first 100 monoelectronic excitations were calculated for the Y^III analogue of the dinuclear complexes. In all steps, a modelling of bulk solvent effects (solvent = dichloromethane) was included through the Polarisable Continuum Model (PCM) [81], using a linear-response non-equilibrium approach for the TD-DFT step [82,83]. Molecular orbitals were sketched using the Gabedit graphical interface [84].

4. Conclusions

Two dinuclear complexes of formula [Ln₂(hfac)₆(L)₂]·CH₂Cl₂·(C₆H₁₄)_0.5 (Ln = Dy (1) and Yb (2)) have been synthesised. The analysis of the polyhedron symmetry around the metallic centres reveals that the dinuclear complexes of the present work have the most symmetric coordination sphere (square anti-prism symmetry) compared to the previously published dinuclear compounds. Nevertheless, they do not display any out-of phase signals of their magnetic susceptibility, demonstrating that the symmetry of the coordination polyhedron is not the crucial parameter for displaying slow magnetic relaxation, but it might be the electronic distribution of the first neighbouring atoms. The functionalization of the TTF core with two methyl ester groups leads to a low-energy ILCT band, which is identified as the HOMO→LUMO excitation. The LUMO is localized on the methyl ester moieties. Irradiation of this absorption band leads to the sensitization of the Yb^III-centred luminescence. Static magnetic properties of the Yb^III analogue and its luminescence have been correlated. More work is in progress in the group in order to assemble this kind of dinuclear complexes through chemical modifications of the methyl ester groups, for example using carboxylic functions.