The conjugate base of methyl 3-oxobutanoate as an antenna ligand in visible-emitting photoluminescent lanthanide complexes

Abstract

Coordination compounds with the formulae [Y(MAA)₃]_x (1^Y), [Ln(MAA)₃]_x (1^Ln), [Y(MAA)₃(phen)] (2^Y) and [Ln(MAA)₃(phen)] (2^Ln) (Ln = Sm, Eu, Tb, Dy, Yb; MAA = conjugate base of methyl 3-oxo-butanoate; phen = 1,10-phenantroline) were synthesized and characterized, and X-ray diffraction data were collected for [Tb(MAA)₃(phen)] (2^Tb) and [Yb(MAA)₃(phen)] (2^Yb). Most of the lanthanide derivatives showed appreciable luminescence in the solid state upon excitation with near-UV light. Strong emissions were observed in particular for complexes of lanthanide ions with high-energy resonance levels. Visible-emitting [Ln(MAA)₃(phen)] compounds were successfully used as dopants for the preparation of luminescent poly(methylmethacrylate) samples. Similar derivatives of the conjugate base of dimethyl malonate (DMM), with formulae [Y(DMM)₃(phen)] (3^Y) and [Ln(DMM)₃(phen)] (3^Ln) (Ln = Eu, Tb, Dy), were also synthesized and their photoluminescence behaviour was compared to that of the analogous methyl 3-oxo-butanoato-complexes.

---

Introduction

β-Diketonates are among the most used organic ligands in lanthanide coordination chemistry, and a huge number of neutral and charged complexes is known. The most usual substituents are alkyls, aryls, thiophene-based rings and perfluorinated chains, and the common coordination numbers are eight and nine. Even if the coordination sphere can be composed by only β-diketonates, often other ligands are present, usually mono- and polydentate oxygen and nitrogen donors. Besides their use as NMR shift reagents, MOCVD precursors and catalysts for organic reactions, these compounds have been studied in detail as luminescent materials. Several β-diketonate complexes of trivalent lanthanide ions are of current interest for the preparation of advanced materials for frontier technology applications, such as novel devices for light conversion and amplification and new biomedical sensors.¹

Quite surprisingly, lower attention has been devoted to other anionic bidentate O-donor ligands comparable to β-diketonates. As an example, we recently reported the synthesis and photophysical characterization of lanthanide derivatives with the conjugate bases of nitromalonaldehyde and bromomalonaldehyde, their use as dopants for luminescent plastic materials and the electrochemical tuning of photoluminescence of an europium complex of bromomalonalehyde.²

Other examples of ligands of potential interest for the preparation of new lanthanide complexes are the conjugate bases of methyl 3-oxo-butanoate (methyl acetoacetate, MAA) and dimethyl malonate (DMM). To the best of our knowledge no lanthanide compound with MAA is actually reported in the literature, even if some complexes of group 4 elements have been synthesized and characterized.³ For what concerns DMM and similar ligands, dinuclear dysprosium complexes of the type [Dy{ROC(O)CHC(O)OR}₃]₂ (R = Me, Et, ⁱPr, ^tBu, SiMe₃) and heteroleptic species having formulae [Dy{ROC(O)CHC(O)OR}₃(bpy)] (R = Et, ^tBu, SiMe₃; bpy = 2,2′-bipyridine) and [Dy{Me₃SiOC(O)CHC(O)OSiMe₃}₃(py)] (py = pyridine) have been prepared and used as MOCVD precursors, but no photophysical study has been carried out.⁴ In the framework of our interest towards new luminescent lanthanide complexes and polymeric materials containing lanthanide derivatives,⁵ the lack of experimental data regarding the use of methyl 3-oxo-butanoate as antenna-ligand towards lanthanide ions prompted us to start a research in this field. The results are summarized in this paper, together with comparative outcomes concerning new DMM-based lanthanide complexes.

Results and discussion

Synthesis and characterization of the complexes

Complexes having formulae [Y(MAA)₃]_x (1^Y) and [Ln(MAA)₃]_x (1^Ln) (MAA = conjugate base of methyl 3-oxo-butanoate; Ln = Sm, Eu, Tb, Dy, Yb) were obtained by reacting YCl₃ or LnCl₃ with three equivalents of K[MAA] in THF at room temperature, as depicted in Scheme 1. The elemental analyses agree with the proposed formulations. The IR spectra are all similar and show a quite intense signal involving the C=O moieties at wavenumbers comprised between 1625 and 1644 cm⁻¹. The ¹H NMR spectra of the MAA derivatives in (CD₃)₂SO or CDCl₃ show three signals attributable to the CH, C(O)CH₃ and OCH₃ groups of the coordinated ligands, which are therefore equivalent on the NMR timescale (see the ESI file, Fig. S1–S6†). These resonances are affected by paramagnetic shift and relaxation in lanthanide derivatives, in particular when the metal centre has high magnetic moment. As an example, the δ values for the diamagnetic derivative 1^Y in deuterated dimethylsulfoxide are 4.62, 3.44 and 1.69 ppm, while the chemical shift values for 1^Tb are 99.25, 78.19 and −32.30 ppm. No signal corresponding to coordinated THF is present and conductivity measurements have ruled out the formation of ionic compounds. Three MAA ligands are not able to saturate the coordination sphere of the yttrium and lanthanide ions, therefore the formation of oligomers is expected. On considering the dinuclear structure of the dialkyl malonate complexes [Dy{ROC(O)CHC(O)OR}₃]₂ already reported in the literature,⁴ where two [Dy{ROC(O)CHC(O)OR}₃] units are bridged by two oxygen atoms, one from an OR group and the other from a C=O moiety, we can tentatively suppose a comparable structure also for the new compounds 1^Y and 1^Ln.

Scheme 1 Synthesis of [Y(MAA)₃]_x (1^Y), [Ln(MAA)₃]_x, (1^Ln), [Y(MAA)₃(phen)] (2^Y), [Ln(MAA)₃(phen)] (2^Ln), [Y(DMM)₃(phen)] (3^Y) and [Ln(MAA)₃(phen)] (3^Ln).

Addition of 1,10-phenanthroline (phen) to THF solutions of 1^Y or 1^Ln at room temperature led to the formation of the complexes [Y(MAA)₃(phen)] 2^Y and [Ln(MAA)₃(phen)] 2^Ln (Ln = Sm, Eu, Tb, Dy, Yb), as depicted in Scheme 1. The elemental analyses confirm the proposed formulations and conductivity measurements indicate that these complexes are neutral. The IR stretching involving the C=O moieties falls between 1616 and 1622 cm⁻¹, at slightly lower wavenumbers with respect to the homoleptic precursors. All the ¹H NMR spectra, recorded using CDCl₃ or (CD₃)₂SO as solvents, show three signals for the CH, C(O)CH₃ and OCH₃ groups of MAA and four resonances attributable to phen (see the ESI file, Fig. S7–S12†). As for 1^Y and 1^Ln, no signal due to coordinated THF is observable. Paramagnetic shift and relaxation strongly influence the resonances of all the coordinated ligands. For example, the ¹H NMR signals of 2^Tb and 2^Dy at 298 K are spread over a range of about 140 ppm. The relative simplicity of the ¹H NMR spectra can be explained on admitting a fast fluxional behaviour of the complexes in solution, which makes the three MAA ligands equivalent on the NMR timescale. In the cases of 2^Tb and 2^Yb crystals suitable for X-ray diffraction were obtained from dichloromethane/diethylether solutions.

The asymmetric units of both 2^Tb and 2^Yb contain two molecules with similar geometrical parameters, only one of them is shown in Fig. 1. For each Ln(III), the coordination environment includes six O atoms from three methyl 3-oxobutanoato-O,O′ ligands and two N atoms from the 1,10-phenanthroline neutral ligand. Selected bond lengths and angles are set out in Table 1.

Fig. 1 Left: ORTEP⁶ view (20% probability level) of one of the molecules 2^Tb found in the asymmetric unit. Right: ORTEP⁶ view (30% displacement level) of one of the molecules found in the asymmetric unit of 2^Yb. For clarity, only one of the methyl 3-oxobutanoato ligands was labelled.

Table 1 Selected bond lengths [Å] and angles [°] for 2^Tb and 2^Yb

	2^Tb	2^Yb
Ln(1)–O(11)	2.323(4)	2.309(3)
Ln(1)–O(12)	2.330(4)	2.247(3)
Ln(1)–O(21)	2.298(5)	2.249(3)
Ln(1)–O(22)	2.406(5)	2.313(3)
Ln(1)–O(31)	2.315(4)	2.311(3)
Ln(1)–O(32)	2.348(4)	2.276(3)
Ln(1)–N(41)	2.633(5)	2.533(3)
Ln(1)–N(42)	2.605(5)	2.525(3)
Ln(2)–O(51)	2.305(6)	2.298(3)
Ln(2)–O(52)	2.398(5)	2.278(3)
Ln(2)–O(61)	2.316(4)	2.250(3)
Ln(2)–O(62)	2.345(5)	2.320(3)
Ln(2)–O(71)	2.314(4)	2.279(3)
Ln(2)–O(72)	2.352(5)	2.274(3)
Ln(2)–N(81)	2.592(6)	2.522(3)
Ln(2)–N(82)	2.633(7)	2.498(3)
O(11)–Ln(1)–O(12)	72.37(14)	74.73(10)
O(21)–Ln(1)–O(22)	72.37(16)	74.07(12)
O(31)–Ln(1)–O(32)	72.15(15)	73.18(11)
N(41)–Ln(1)–N(42)	61.90(16)	64.52(10)
O(51)–Ln(2)–O(52)	72.25(19)	73.18(11)
O(61)–Ln(2)–O(62)	72.11(17)	73.92(13)
O(71)–Ln(2)–O(72)	72.19(15)	74.11(15)
N(81)–Ln(2)–N(82)	61.81(18)	64.84(10)

---

The coordination polyhedron of the LnN₂O₆ was approximately assigned by using the SHAPE software⁷ and the calculation indicated that the LnN₂O₆ coordination polyhedron resulted to be different for each lanthanide. In the case of 2^Tb the polyhedron is best described as “triangular dodecahedron”, as seen in Fig. 2, with SHAPE values of 0.343 and 0.420 for the two molecules, while other possibilities⁸ as elongated trigonal prism are bigger than 2.5. The output of the SHAPE program are given in Table 2.

Fig. 2 Polyhedron (dodecahedron) around the Tb atom in 2^Tb.

Table 2 Output of the SHAPE software for 2^Tb and 2^Yba

Structure [ML8] (*)	OP	HPY	HBPY	CU	SAPR	TDD	JGBF	JETBPY	JBTPR	BTPR	JSD	TT	ETBPY
a Note that there are two molecules in the asymmetric unit in both compounds. (*) OP D8h octagon; HPY C7v heptagonal pyramid; HBPY D6h hexagonal bipyramid; CU Oh cube; SAPR D4d square antiprism; TDD D2d triangular dodecahedron; JGBF D2d Johnson gyrobifastigium J26; JETBPY D3h Johnson elongated triangular bipyramid J14; JBTPR C2v biaugmented trigonal prism J50; BTPR C2v biaugmented trigonal prism; JSD D2d Snub disphenoid J84; TT Td triakis tetrahedron; ETBPY D3h elongated trigonal bipyramid.
Tb1	33.409	24.288	15.162	8.110	3.116	0.343	15.224	29.231	3.308	2.744	3.144	8.831	25.273
Tb2	32.318	23.826	15.081	8.248	2.266	0.420	14.946	29.489	3.113	2.571	3.263	8.996	24.948
Yb1	29.333	22.793	16.868	10.327	0.391	2.107	16.038	27.550	2.385	1.984	4.765	11.124	23.641
Yb2	29.897	22.486	16.290	9.964	0.477	2.282	15.925	27.651	2.692	2.086	4.852	10.785	24.326

---

The Tb–O bond lengths vary in the range of 2.298(5) to 2.406(5) Å, similar (although in a slightly wide range) to that found in the acetylacetonate compound [Tb(acac)₃(tdzp)] where tdzp represents a substituted phenanthroline, between 2.320(2) and 2.385(2) Å,⁹ and the Tb–N bond distances are between 2.592(6) and 2.633(7) Å, slightly longer than the Tb–N distances (2.563 and 2.567 Å) found for other Tb(1,10-phen) complexes.¹⁰

Methyl 3-oxo-butanoato-O,O′ ligands are chelating the Tb(III) forming five membered rings almost planar¹¹ with root-mean-square deviation for these plane less than 0.0512 Å. Phenanthroline ligand is coordinated without any bending, and the Tb atoms are situated at less than 0.156(1) Å out of the phenanthroline best plane [that is, the angles between the phenanthroline best plane and the Tb–N bonds are between 2.7(2) and 2.9(2)°]. The O–Tb–O angles are in the range of 72.1(2) to 72.4(2)°, (five membered rings) whereas the N–Tb–N ones are 61.9(2) and 61.8(2)° (four membered rings) and these chelate rings are forming a paddle-wheel structure (Fig. 3) with dihedral angles rings of average 82.24°.

Fig. 3 Left: paddle-wheel structure for 2^Tb. Right: paddle-wheel structure for 2^Yb. Only chelate rings atoms were drawn.

Regarding with 2^Yb, once more the coordination environment of each Yb(III) is formed by six O atoms from three methyl 3-oxo-butanoato-O,O′ ligands and two N atoms from the 1,10-phenanthroline neutral ligand. However compounds 2^Tb and 2^Yb are not isomorphous. The SHAPE software⁷ provides for the coordination environment the data show in Table 2 which let us to conclude for this compound a “square antiprismatric polyhedron”. The two square planes consisting, for Yb1 of atoms N(41)/N(42)/O(11)/O(12) and O(31)/O(32)/O(21)/O(22), and for Yb2 of atoms N(81)/N(82)/O(71)/O(72) and O(51)/O(52)/O(61)/O(62) with a deviation of the root-mean-square of 0.0699, 0.0308, 0.0387 and 0.0770 Å, respectively. These square planes are rotated by an angle of almost 45° with respect to each other (see Fig. 4) and the distance between the centroids of the two square planes are about 2.54 Å. Yb(III) atoms are situated at about 1.34 Å from the N₂O₂ plane and at 1.18 Å from the O₄ plane.

Fig. 4 Square antiprism around Yb metal for 2^Yb. In yellow the square planes.

Methyl 3-oxo-butanoato-O,O′ ligands are chelating the Yb(III) forming five membered rings and they are strongly bent coordinated, in a different manner that occurs with the related Tb compound 2^Tb, with root-mean-square deviation for these plane of average 0.122 Å. If the metal atom is not considered in the calculation of the best plane, such atom deviated from the OCCO plane in average 0.56 Å, between 0.463(6) and 0.663(6) Å. Also the 1,10-phenathroline is coordinated in a bending mode, in such a way that the Yb atoms are situated at 0.467(1) and 0.536(1) Å from the phenanthroline best plane [that is, the angles between the phenanthroline best planes and the Yb–N bonds are between 9.7(2) and 11.7(2)°]. These deviations are clearly showed on the paddle-wheel structure shown in Fig. 3.

The Yb–O bond lengths are between 2.247(3) and 2.320(3) Å, as usual shorter than the Yb–N bond lengths, which range between 2.498(3) and 2.533(3) Å. These values are shorter to those in the terbium complex as expected due the well known lanthanide contraction. No important differences with other octacoordinated Yb(III) complexes are found.¹²

The replacement of K[MAA] with the potassium salt of dimethyl malonate, K[DMM], using the previously described experimental conditions allowed to isolate the complexes [Y(DMM)(phen)], 3^Y and [Ln(DMM)(phen)], 3^Ln (Ln = Eu, Tb, Dy), as depicted in Scheme 1. Elemental analyses and conductivity measurements support the proposed formulations. All the IR spectra are closely comparable and show a quite intense band between 1661 and 1667 cm⁻¹ attributable to the stretching of the carboxylic moieties. As for the previously described complexes 2^Y and 2^Ln, fluxional behaviour in solution led to quite simple ¹H NMR spectra, where only two resonances for DMM and four signals for phenanthroline are present, affected by meaningful paramagnetic shift in the case of 3^Eu (see Fig. 5). Unfortunately, paramagnetic relaxation did not allow to observe resolved resonances for the 3^Tb and 3^Dy derivatives. It is likely to suppose that also 3^Ln complexes are mononuclear, on considering the X-ray structures reported in the literature for the dysprosium derivatives [Dy{ROC(O)CHC(O)OR}₃(bpy)] (R = Et, ^tBu, SiMe₃; bpy = 2,2′-bipyridine).⁴

Fig. 5 ¹H NMR spectra of 3^Y and 3^Eu (CDCl₃, 298 K).

Preliminary electrochemical investigations on solutions of the europium derivatives suggested that the f-orbitals involved in the reduction of Eu³⁺ to Eu²⁺ are screened by the coordinated O-donor ligands, a behaviour already observed in the past for europium β-diketonates and nitromalonaldeyde complexes.^1a,2a A typical cyclic voltammogram is reported as an example in Fig. S13.†

Photoluminescence measurements on complexes and doped polymers

Photoluminescence studies were initially carried out on 1^Ln complexes. Compound 1^Eu is weakly luminescent in the red region upon excitation with UV light. The emission spectrum (see Fig. 6) shows that the most intense transition is the ⁵D₀ → ⁷F₂, centred at 611 nm (73.7%). The high ⁵D₀ → ⁷F₂/⁵D₀ → ⁷F₁ ratio, 10 : 1, suggests a low symmetry of the inner coordination sphere.¹³ The luminescence decay curve is mono-exponential and the measured lifetime (τ) is 0.255 ms, quite low if compared to the τ values of several homoleptic β-diketonate europium complexes.¹⁴ An intrinsic quantum yield (Q_i) around 18% has been estimated from the τ value and the relative intensity of the ⁵D₀ → ⁷F₁ transition.

Fig. 6 Emission spectra (solid samples, 298 K) of 1^Eu (red plot, λ_excitation = 319 nm), 1^Tb (green plot, λ_excitation = 325 nm) and 1^Dy (yellow plot, λ_excitation = 365 nm).

MAA showed to be an efficient antenna-ligand towards the Tb³⁺ ion. 1^Tb is an intense green-emitting compound under UV light. The typical ⁵D₄ → ⁷F_J (J = 6–0) transitions can be observed in the emission spectrum of 1^Tb and the most intense is the ⁵D₄ → ⁷F₅ one, centred at 544 nm (Fig. 6). The photoluminescence excitation spectrum indicates that the antenna-effect from the coordinated ligands is the dominant process of population of the resonance level of Tb³⁺ and occurs for λ_excitation below 380 nm, the maximum being reached around 305 nm (see Fig. S14†). The observed lifetime for the terbium emission in 1^Tb is 0.965 ms, corresponding to a Q_i value of about 20% by accepting 4.75 ms as the τ_rad value for Tb³⁺.¹⁵

The photoluminescence measurements were in turn extended to the samarium derivative 1^Sm, but no resolved emission spectrum was collected. Despite the fact that the transitions in the visible range of Sm³⁺ and Eu³⁺ fall at comparable wavelengths, the samarium compounds are more influenced by non-radiative decay because of the lower energy gap between the resonance level and the highest ground-state SO level (Sm³⁺: 7400 cm⁻¹; Eu³⁺: 12 300 cm⁻¹ (ref. 16)). This fact, combined with the hygroscopic character of 1^Sm, could explain the very poor photoluminescence exhibited by this compound.

Similar problems were expected for 1^Dy, even if the lower ionic radius¹⁷ makes this compound meaningfully less hygroscopic than 1^Sm. Quite surprisingly, an appreciable yellowish-white luminescence was observed for samples of 1^Dy under near-UV light, and the emission spectrum of 1^Dy (Fig. 6) shows resolved bands due to the transitions from ⁴F_9/2 to ⁶H_J (J = 15/2–11/2), the most intense being the hypersensitive ⁴F_9/2 → ⁶H_13/2 centred at 572 nm (72.9%). The measured lifetime of this emission, 26 μs at room temperature, is quite long if compared to the few data present in the literature. For example, a τ value of 7 μs has been reported for the compound [Dy(PM)₃(TPPO)₂] (PM = 1-phenyl-3-methyl-4-isobutyryl-5-pyrazolone; TPPO = triphenylphosphine oxide).¹⁸ The observed lifetime for the tris(benzoylacetonate) complex of dysprosium at 77 K is 12.3 μs.¹⁹ By admitting a τ_rad value for the Dy³⁺ ion around 400 μs,²⁰ the intrinsic quantum yield of 1^Dy is around 6–7%.

We tried to explain the good performances exhibited by the MAA complexes of emitters having high-energy resonance levels by performing DFT and TD-DFT calculations on the diamagnetic model system [Lu(MAA)₃] and, for comparative purposes, on [Lu(acac)₃] (acac = acetylacetonate). The Cartesian coordinates of the two optimized geometries are collected in the ESI.† The computed average Lu–O bond lengths, 2.195 Å for [Lu(acac)₃], 2.183 Å and 2.201 Å for [Lu(MAA)₃] respectively relative to the Lu-OC(CH₃) and Lu-OC(OCH₃) interactions, are closely comparable to experimental data reported in the literature for similar compounds. As an example, Inoue et al. reported an average Lu–O distance of 2.19 Å for the tris(2,2,6,6-tetramethylheptane-3,5-dionato)lutetium(III).²¹

MAA is formally an acetylacetonate derivative where one –CH₃ group is replaced by a –OCH₃ moiety. DFT outcomes predict a meaningful raise of both the first excited singlet (S) and triplet (T) states from [Lu(acac)₃] (S: 31 600 cm⁻¹; T: 23 900 cm⁻¹) to [Lu(MAA)₃] (S: 35 800 cm⁻¹; T: 26 900 cm⁻¹), attributable to the introduction of an electron-withdrawing group. The computed values support the idea of an easier energy transfer from MAA towards the resonance levels of Tb³⁺ (20 430 cm⁻¹) and Dy³⁺ (20 960 cm⁻¹)^1d with respect to acac. The long lifetimes recorded for 1^Tb and 1^Dy can be therefore attributed to the favourable excited-state energies of coordinated MAA, besides the lack of high-energy oscillators close to the lanthanide centres.

The addition of 1,10-phenanthroline to the coordination sphere on the lanthanide ions caused an increase of light harvesting in the UV range. Moreover, the observed lifetime for the 2^Eu derivatives is 0.828 ms, meaningfully higher than that of 1^Eu (see Fig. 7 for the luminescence decay curve of 2^Eu). The intrinsic quantum yield is 67% and makes 2^Eu a bright red-luminescent compound under excitation with near-UV light. The measured lifetime appears quite long if compared to the τ values of the phen-europium derivatives of dibenzoylmethane, dimethoxydibenzoylmethane, benzoylacetone and 4-methylbenzoyl-2-furanoylmethane, ranging from 0.34 to 0.47 ms.²² The same Q_i value was obtained in the past for the complex [Eu(NMA)₃(terpy)] (NMA = nitromalonaldehyde; terpy = 2,2′:6′,2′′-terpyridine).^2b The most intense band in the emission spectrum is the ⁵D₀ → ⁷F₂ one (62.1%) and, as for 1^Eu, the ⁵D₀ → ⁷F₂/⁵D₀ → ⁷F₁ ratio is around 10 : 1 (Fig. 8). It is likely to suppose that the ⁷F₁ level is composed by three sublevels, even if two of them overlap in Fig. 8, and that the ⁷F₂ is separated in five sublevels, the last one corresponding to the weak transition at 631 nm. This would agree with the very low symmetry of the first coordination sphere observed in the X-ray structures of the analogues 2^Tb and 2^Yb derivatives.^13c The photoluminescence excitation spectrum indicates antenna-effect over about all the UV range, as observable in Fig. 7. Comparable excitation spectra for wavelengths below 400 nm have been recorded for all the other 2^Ln derivatives (see Fig. S15–S18†). As a minor point, it is to be highlighted the high intensity of the hypersensitive ⁵D₂ ← ⁷F₀ transition at 464 nm in the PLE spectrum reported in Fig. 7, if compared to the ⁵L₆ ← ⁷F₀ band at 394 nm.

Fig. 7 Luminescence decay curve of 2^Eu (solid sample, 298 K, λ_excitation = 325 nm, λ_emission = 613 nm). Inset: excitation spectrum of 2^Eu (solid sample, 298 K, λ_emission = 610 nm).

Fig. 8 Emission spectra (solid samples, 298 K) of 2^Sm (orange plot, λ_excitation = 350 nm), 2^Eu (red plot, λ_excitation = 338 nm), 2^Tb (green plot, λ_excitation = 310 nm) and 2^Dy (yellow plot, λ_excitation = 365 nm).

Also in the case of terbium the coordination of 1,10-phenanthroline caused an increase of luminescence. The observed lifetime of 2^Tb is 1.212 ms, meaningfully higher than that of 1^Tb, and corresponds to an estimated Q_i of 26%. The bright green emission of this complex is related to the ⁵D₄ → ⁷F_J (J = 6–0) transitions, the most intense (67.6%) occurring for J = 5 and centred between 541 and 547 nm (Fig. 8).

Differently from the precursor 1^Sm, the compound 2^Sm is perceptibly photoluminescent. The bands in the 550–750 nm range, observable in Fig. 8, are associated to the ⁴G_5/2 → ⁶H_J (J = 5/2–11/2) transitions. The most intense emission is the hypersensitive one around 645 nm, corresponding to J = 9/2, as already reported for β-diketonate derivatives of samarium.²³ Having the ⁴G_5/2 → ⁶H_5/2 transition predominant magnetic dipole character, an elevated ⁴G_5/2 → ⁶H_9/2/⁴G_5/2 → ⁶H_5/2 intensity ratio indicates low symmetry and high polarizability of the first coordination sphere. In the case of 2^Sm this ratio is 7.7, higher than those of [Sm(Tp)₃] and [Sm(DPA)₃]³⁻ (Tp = tris(pyrazol-1-yl)-borate; DPA = dipicolinate), which are comprised between 4 and 5. Ratios greater than 10 have been however obtained for several tetrakis-β-diketonates.^5c,23a The τ value measured for 2^Sm, 31 μs, is a bit lower than that reported for [Sm(hfa)₃(phen)] (hfa = hexafluoroacetylacetonate).²⁴ On the basis of the τ_rad values present in the literature for Sm³⁺, around 3 ms,²⁵ the intrinsic quantum yield of 2^Sm is about 1%.

Even if the photoluminescence of 2^Dy appears enhanced by the increased absorption due to the coordination of 1,10-phenantholine, the measured τ value is 3 μs, meaningfully lower than that obtained for the precursor 1^Dy. As for the homoleptic compound, the yellow-whitish emission is attributable to the ⁴F_9/2 → ⁶H_J (J = 15/2–9/2) transitions, in particular the ⁴F_9/2 → ⁶H_15/2 (29%) at about 484 nm and the ⁴F_9/2 → ⁶H_13/2 (66.2%) centred at 576 nm, as observable in Fig. 8. This is the only case here reported of a reduction of lifetime caused by the coordination of phenanthroline ligand. It is worth noting that the triplet state of 1,10-phenanthroline is around 22 100 cm⁻¹ in the free molecule.²⁶ On considering the previously described TD-DFT results, coordinated 1,10-phenanthroline should be the ligand with the lowest triplet state energy in the coordination sphere of 2^Dy. The decrease of the observed lifetime from 1^Dy to 2^Dy could be therefore attributed to the closeness of the Dy³⁺ emitting level ⁴F_9/2 (20 960 cm⁻¹ (ref. 1d)) to the triplet state of the N-donor ligand.

After examining the photoluminescent features of the 1^Ln and 2^Ln complexes of visible-emitting lanthanide ions, attention has been devoted to the ytterbium derivatives 1^Yb and 2^Yb. Both the compounds showed the typical ⁷F_5/2 → ²F_7/2 transition at around 1000 nm upon excitation with UV light, as observable in Fig. 9. The emission spectrum is more resolved for compound 2^Yb, probably because of the lower hygroscopic character due to the saturation of the coordination sphere. In both the cases, however, the luminescence resulted too weak to allow the measurement of the lifetime.

Fig. 9 Emission spectra (solid samples, 298 K) of 1^Yb (light gray plot, λ_excitation = 326 nm) and 2^Yb (dark gray plot, λ_excitation = 351 nm).

The good photoluminescent features of the 2^Ln complexes of europium, terbium and dysprosium prompted us to prepare and characterize the corresponding DMM derivatives 3^Ln. In all the cases compounds with appreciable luminescence in the visible range upon excitation with UV light were isolated. The emission spectra of 3^Tb and 3^Dy are roughly comparable to those previously described for the 2^Ln complexes of the same elements. In the case of the europium complex, even if the ⁵D₀ → ⁷F₂ transition remains the most intense one, the intensity ratio ⁵D₀ → ⁷F₂/⁵D₀ → ⁷F₁ of 3^Eu is meaningfully lower with respect to 2^Eu, being about 5.5 : 1. The change of chelating O-donor ligand also influences the observed luminescence lifetimes. The τ values measured for 3^Tb and 3^Dy at room temperature are respectively 1.268 ms and 6 μs, a bit longer than those recorded for the analogous 2^Ln compounds, this suggesting a slight increase of intrinsic quantum yield. On the other hand, the lifetime of the emission of 3^Eu is 0.540 ms, meaningfully shorter than that reported for 2^Eu. On considering the change of relative intensity of the ⁵D₀ → ⁷F₁ transition, the Q_i value estimated for 3^Eu is 25%. The measured lifetimes of 1^Ln, 2^Ln and 3^Ln compounds are summarized for clarity in Table 3.

Table 3 Observed lifetimes (μs) of complexes and doped plastic materials (298 K)

	1^Ln	2^Ln	2^Ln@PMMA	2^Ln@PVP	3^Ln	3^Ln@PMMA
a Freshly prepared sample.
Sm		31	36
Eu	255	828	337	686^a	540
Tb	965	1212	342		1268	28
Dy	26	3	5		6

---

The phen-complexes having luminescence in the visible range were studied as dopants for the preparation of photoluminescent poly(methylmethacrylate), PMMA. The samples were obtained by adding a known amount of the proper complex to a solution of PMMA in dichloromethane and subsequent slow removal of the solvent. The complex: PMMA ratios considered in this work range from 10 to 50 μmol g_PMMA⁻¹. The 2^Ln@PMMA samples (Ln = Sm, Eu, Tb, Dy) showed appreciable luminescence if irradiated with UV light. The relative intensities of the Ln³⁺ transitions in the emission spectra are roughly similar to those observed for the pure compounds and do not depend upon the concentration of the complex in the matrix, this suggesting that the embedding in PMMA does not meaningfully change the first coordination sphere of the complexes. The presence of PMMA chains has however a variable influence on the observed lifetimes (see Table 3). For the derivatives of the metal centres most affected by non-radiative decay, i.e. samarium and dysprosium, a small increase of the τ values is caused by the encapsulation in the polymer matrix (τ = 36 μs for 2^Sm@PMMA, 5 μs for 2^Dy@PMMA), probably because the complexes are protected from moisture. On the other hand, the measured lifetimes of 2^Eu@PMMA and 2^Tb@PMMA are meaningfully lower than those of the pure compounds, respectively 0.337 ms and 0.342 ms. This result can be tentatively explained on the basis of an incremented competition of luminescence from the excited states of the ligands. In fact, as clearly observable in Fig. 10 for Ln = Eu, a broad band in the violet-blue region appears in the emission spectrum of the doped polymer, which is absent in that of the pure complex. All the 2^Ln@PMMA doped polymers exhibited negligible variations of the lifetimes on changing the concentration of the complexes in the corresponding materials.

Fig. 10 Comparison of the emission spectra of solid 2^Eu (red line) and 2^Eu@PMMA (purple line) at 298 K. λ_excitation = 338 nm.

Also 3^Ln complexes were studied as dopants for the preparation of luminescent poly(methylmethacrylate), PMMA. Unfortunately, photophysical measurements on 3^Tb@PMMA suggest that the complex is decomposed by the polymer chains. Even if the emission spectrum shows the typical transitions of the Tb³⁺ ion, the measured lifetime is extremely low, around 28 μs, corresponding to an intrinsic quantum yield below 1%.

Besides PMMA, another polymer that was considered is polyvinylpyrrolidone PVP, which has recently shown to be a good matrix for the preparation of photoluminescent plastics doped with lanthanide β-diketonates.²⁷ Preliminary experiments using 2^Eu as luminescent species indicate that the coordination sphere is progressively altered by PVP. In fact, the photoluminescence of 2^Eu@PVP quenches almost completely in few days. Emission and excitation measurements on freshly prepared samples showed a relative increase of the ⁵D₀ → ⁷F₁ and ⁵D₀ → ⁷F₄ transitions with respect to pure 2^Eu and a reduction of about 50 nm of the excitation spectrum in the near-UV range.

Conclusions

In this paper we reported new coordination compounds of trivalent yttrium and lanthanide ions with the conjugate base of methyl 3-oxo-butanoate. This ligand resulted able to sensitize several luminescent lanthanide ions and good performances were observed in particular for homoleptic complexes of the high-energy emitters Tb³⁺ and Dy³⁺, a result probably attributable to the raise of the excited states with respect to common β-diketonates caused by the presence of an electron-withdrawing moiety. The introduction of 1,10-phenanthroline in the coordination sphere enhanced the luminescence of the samarium, europium and terbium complexes, but the τ value of [Dy(MAA)₃(phen)] resulted meaningfully lower than that of [Dy(MAA)₃]_n. Visibly luminescent plastic materials were obtained from the embedding of [Ln(MAA)₃(phen)] complexes in PMMA, but luminescence from the coordinated ligands became competitive for the derivatives of europium and terbium. Polyvinylpyrrolidone resulted not suited as matrix for the preparation of plastic materials doped with these compounds. Finally, some dimethyl malonato-complexes of the type [Ln(DMM)₃(phen)] were isolated and characterized. The photoluminescence features of the terbium and dysprosium complexes resemble those of the analogous methyl 3-oxo-butanoate derivatives, while the sensitization of Eu³⁺ is less efficient. Preliminary experiments indicated that these last compounds are strongly altered by encapsulation in PMMA.

Experimental section

Materials and methods

The reagents used for the preparation of the lanthanide complexes and the doped polymers were Aldrich products used without further purifications, with the exception of the anhydrous YCl₃ and LnCl₃ salts (Strem). The synthesis of the complexes was carried out under inert atmosphere (N₂, < 5 ppm O₂, < 1 ppm H₂O) in a MBraun glove-box. Solvents were purified using common techniques.²⁸ The potassium salt K[MAA] was prepared by slowly adding potassium tert-butoxide (1.040 g, 9.3 mmol) in anhydrous methanol (20 mL) to a solution of methyl 3-oxo-butanoate (1.0 mL, 9.3 mmol) in 30 mL of methanol, kept at 0 °C. At the end of the addition, the solvent was removed under reduced pressure and diethylether (30 mL) was added. The white solid obtained was filtered and dried. Yield 95%. The salt K[DMM] was prepared by adding 2.0 mL (17.5 mmol) of dimethyl malonate to 1.960 g (17.5 mmol) of potassium tert-butoxide in 100 mL of frozen, anhydrous THF. The reaction mixture was allowed to slowly reach room temperature, then the solvent was removed by evaporation under reduced pressure and diethylether (50 mL) was added. The solid which separated out was collected by filtration and dried. Yield 95%.

Conductivity measurements were carried out at 298 K on 10⁻³ mol dm⁻³ solutions using a radiometer Copenhagen CDM 83 instrument. Voltammetric experiments were performed at room temperature in an air-tight, three-electrode cell, which was located in a Faraday cage to avoid external noise. Glassy carbon working electrodes (1.5 mm radius) were employed. A platinum spiral was used as counter electrode, and aqueous Ag/AgCl saturated with KCl was used as reference electrode. All working electrodes were polished mechanically with graded alumina powder on a polishing microcloth. All voltammetric measurements were carried out using a CH Instruments 760b bipotentiostat. Elemental analyses were performed at the University of Padua using either a Fison EA1108 or a Thermo Scientific FLASH 2000 CHNS-O microanalyzer. NMR spectra were recorded at variable temperature on a Bruker Avance 300 or a Bruker AC 200 spectrometer using CDCl₃ or (CD₃)₂SO as solvents. ¹H chemical shifts are reported relative to tetramethylsilane. The solvent signals, quoted with respect to TMS (δ = 0 ppm), were used as internal references. IR spectra were recorded from 4000 to 400 cm⁻¹ using a Perkin Elmer Spectrum One spectrophotometer. Samples were dispersed in KBr or nujol. Photoluminescence emission (PL) and excitation (PLE) measurements were carried out on solid samples at room temperature by a Horiba Jobin Yvon Fluorolog-3 spectrofluorometer. A xenon arc lamp was used as continuous-spectrum source selecting the excitation wavelength by a double grating Czerny-Turner monochromator. A single grating monochromator coupled to a R928 Hamamatsu photomultiplier tube (PMT) was used as detection system for optical emission measurements. Excitation and emission spectra were corrected for the instrumental functions. Time-resolved analyses were performed in Multi Channel Scaling modality (MCS) by using a tuneable pulsed Nd:YAG laser system as excitation source. The luminescence lifetimes were derived from the luminescence decay curves, fitting the data with the least squares method by using mono-exponential equation for all the materials. The intrinsic quantum yield (Q_i) was estimated on the basis of eqn (1), where τ is the measured luminescence lifetime. τ_rad was obtained for the europium-based compounds from eqn (2), where η indicates the refractive index of the sample and the value of 1.5 is assumed in this work for comparative purposes. I(⁵D₀ → ⁷F_J)/I(⁵D₀ → ⁷F₁) is the ratio between the total integrated emission from the Eu(⁵D₀) level to the ⁷F_J manifold and the integrated intensity of the transition ⁵D₀ → ⁷F₁.²⁹ The τ_rad values for the samarium, terbium and dysprosium ions were taken from the literature (Sm³⁺: 3 ms;²⁵ Tb³⁺: 4.75 ms;¹⁵ Dy³⁺: 400 μs (ref. 19)).

Synthesis of [Y(MAA)₃]_x (1^Y) and [Ln(MAA)₃]_x (1^Ln) (Ln = Sm, Eu, Tb, Dy, Yb)

In a typical preparation, solid K[MAA] (0.463 g, 3.0 mmol) was slowly added at room temperature under inert atmosphere to 1.0 mmol of anhydrous YCl₃ or LnCl₃ in 20 mL of THF under vigorous stirring. The resulting reaction mixture was allowed to react at room temperature for 12 hours, then the solids were separated by centrifugation and the solvent was evaporated under reduced pressure. The residual white solid was triturated with diethylether (10 mL), filtered and dried under vacuum. Yield >70% in all the cases.

Characterization of 1^Y. C₁₅H₂₁O₉Y: calcd C 41.49, H 4.87; found C 41.30, H 5.00. IR (KBr): ν_CO = 1630 cm⁻¹. ¹H NMR ((CD₃)₂SO, 298 K, δ): 4.62 (s, 1H, CH); 3.44 (s, 3H, OCH₃); 1.69 (s, 3H, C(O)CH₃). ¹H NMR (CDCl₃, 315 K, δ): 4.71 (s, 1H, CH); 3.56 (s, 3H, OCH₃); 1.81 (s,3H, C(O)CH₃).

Characterization of 1^Sm. C₁₅H₂₁O₉Sm: calcd C 36.35, H 4.27; found 36.19, H 4.37. IR (nujol): ν_CO = 1626 cm⁻¹. ¹H NMR ((CD₃)₂SO, 315 K, δ): 5.44 (s, slightly br, 1H, CH); 3.16 (s, slightly br, 3H, OCH₃); 2.26 (s, 3H, C(O)CH₃).

Characterization of 1^Eu. C₁₅H₂₁O₉Eu: calcd C 36.23, H 4.26; found 36.18, H 4.40. IR (KBr): ν_CO = 1644 cm⁻¹. ¹H NMR ((CD₃)₂SO, 295 K, δ): 6.07 (s, slightly br, 3H, OCH₃); −0.82 (s, br, 1H, CH); −4.23 (s, slightly br, 3H, C(O)CH₃). ¹H NMR ((CD₃)₂SO, 324 K, δ): 6.79 (s, slightly br, 3H, OCH₃); −0.6 (s, very br, CH); −3.16 (s, br, 3H, C(O)CH₃). PL (solid sample, 298 K, λ_excitation = 319 nm, nm): weak, 578 (⁵D₀ → ⁷F₀, 0.7%); 591 (⁵D₀ → ⁷F₁, 7.2%); 611 (⁵D₀ → ⁷F₂, 73.7%); 652 (⁵D₀ → ⁷F₃, 3.8%); 690, 701 (⁵D₀ → ⁷F₄, 14.6%). PLE (solid sample, 298 K, λ_emission = 611 nm, nm): max 319 (ligand excitation); 393, 464 (Eu³⁺ excitation). Emission decay time (solid sample, λ_excitation = 325 nm, λ_emission = 612 nm): τ = 0.255 ms. Q_i = 18%.

Characterization of 1^Tb. C₁₅H₂₁O₉Tb: calcd C 35.73, H 4.20; found 35.59, H 4.38. IR (KBr): ν_CO = 1644 cm⁻¹. ¹H NMR ((CD₃)₂SO, 298 K, δ): 99.25 (s, br, 1H, CH); 78.19 (s, br, 3H, CH₃); −32.30 (s, br, 3H, CH₃). ¹H NMR ((CD₃)₂SO, 310 K, δ): 89.65 (s, br, 1H, CH); 70.12 (s, br, 3H, CH₃); −30.59 (s, br, 3H, CH₃). PL (solid sample, 298 K, λ_excitation = 305 nm, nm): medium, 487, 494 (⁵D₄ → ⁷F₆, 21.2%); 541, 546 (⁵D₄ → ⁷F₅, 65.3%); 582, 591 (⁵D₄ → ⁷F₄, 7.4%); 617, 622 (⁵D₄ → ⁷F₃, 4.7%); 635–695 (⁵D₄ → ⁷F₂, ⁵D₄ → ⁷F₁, ⁵D₄ → ⁷F₀, 1.4%). PLE (solid sample, 298 K, λ_emission = 546 nm, nm): max 305 (ligand excitation). Emission decay time (solid sample, λ_excitation = 325 nm, λ_emission = 545 nm): τ = 0.965 ms. Q_i = 20%.

Characterization of 1^Dy. C₁₅H₂₁O₉Dy: calcd C 35.48, H 4.17; found 35.37, H 4.30. IR (nujol): ν_CO = 1628 cm⁻¹. ¹H NMR ((CD₃)₂SO, 319 K, δ): 54.1 (s, very br, 1H, CH); 38.1 (s, very br, 3H, CH₃); −55.0 (s, very br, 3H, CH₃). PL (solid sample, 298 K, λ_excitation = 365 nm, nm): strong, 478 (⁴F_9/2 → ⁶H_15/2, 19.9%); 572 (⁴F_9/2 → ⁶H_13/2, 72.9%); 662 (⁴F_9/2 → ⁶H_11/2, 7.2%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 585 nm): τ = 26 μs. Q_i = 7%.

Characterization of 1^Yb. C₁₅H₂₁O₉Yb: calcd C 34.76, H 4.08; found 34.61, H 4.12. IR (KBr): ν_CO = 1625 cm⁻¹. ¹H NMR ((CD₃)₂SO, 319 K, δ): 14.18, 8.27, −15.59 (3s, br, MAA). PL (solid sample, 298 K, λ_excitation = 326 nm, nm): weak, 950–1060, max 978 (²F_5/2 → ²F_7/2). PLE (solid sample, 298 K, λ_emission = 1000 nm, nm): max 327 (ligand excitation).

Synthesis of [Y(MAA)₃(phen)] (2^Y) and [Ln(MAA)₃(phen)] (2^Ln) (Ln = Sm, Eu, Tb, Dy, Yb)

Solid K[MAA] (0.463 g, 3.0 mmol) was slowly added at room temperature under inert atmosphere to 1.0 mmol of anhydrous YCl₃ or LnCl₃ in 20 mL of THF under vigorous stirring. The resulting reaction mixture was left under stirring at room temperature for 6 hours, then a solution of anhydrous 1,10-phenantholine (0.180 g, 1.0 mmol) in 5 mL of THF was added drop by drop. The reaction mixture was allowed to react overnight at room temperature, then the solids were removed by centrifugation, the solvent was evaporated under reduced pressure and diethylether (5 mL) was added. The solid which separated out was filtered, washed with 5 mL of cold diethylether and dried under vacuum. Yield >75% in all the cases.

Characterization of 2^Y. C₂₇H₂₉N₂O₉Sm: calcd C 52.78, H 4.76, N 4.56; found 52.59, H 4.80, N 4.61. IR (KBr): ν_CO = 1622 cm⁻¹. ¹H NMR (CDCl₃, 298 K, δ): 9.50 (d, 2H, ³J_HH = 4.5 Hz, phen-H₂/H₉); 8.36 (d, 2H, ³J_HH = 8.3 Hz, phen-H₄/H₇); 7.85 (s, 2H, phen-H₅/H₆); 7.76 (dd, 2H, ³J_HH = 4.5 Hz, ³J_HH = 8.3 Hz, phen-H₃/H₈); 4.69 (s, 3H, CH); 3.39 (s, 9H, OCH₃); 1.79 (s, 9H, C(O)CH₃).

Characterization of 2^Sm. C₂₇H₂₉N₂O₉Sm: calcd C 47.98, H 4.32, N 4.14; found 47.90, H 4.48, N 4.24. IR (KBr): ν_CO = 1616 cm⁻¹. ¹H NMR (CDCl₃, 298 K, δ): 9.20 (s, br, 3H, CH); 8.28 (d, slightly br, 2H, ³J_HH = 7.8 Hz, phen-H₄/H₇); 7.82 (s, 2H, phen-H₅/H₆); 7.66 (m, slightly br, 2H, phen-H₃/H₈); 7.55 (s, br, 2H, phen-H₂/H₉); 3.30 (s, slightly br, 9H, OCH₃); 2.35 (s, br, 9H, C(O)CH₃). PL (solid sample, 298 K, λ_excitation = 345 nm, nm): strong, 558, 564 (⁴G_5/2 → ⁶H_5/2, 7.2%); 593, 599, 604, 608 (⁴G_5/2 → ⁶H_7/2, 26.9%); 639, 646, 650 (⁴G_5/2 → ⁶H_9/2, 55.1%); 703, 711 (⁴G_5/2 → ⁶H_11/2, 10.8%). PLE (solid sample, 298 K, λ_emission = 646 nm, nm): max 310–357 (ligand excitation); 403, 418 (Sm³⁺ excitation). Emission decay time (solid sample, λ_excitation = 320 nm, λ_emission = 646 nm): τ = 31 μs. Q_i = 1%.

Characterization of 2^Eu. C₂₇H₂₉N₂O₉Eu: calcd C 47.87, H 4.31, N 4.13; found 47.69, H 4.50, N 4.04. IR (KBr): ν_CO = 1616 cm⁻¹. ¹H NMR (CDCl₃, 298 K, δ): 10.88 (d, 2H, ³J_HH = 7.8 Hz, phen-H₄/H₇); 10.62 (s, 2H, phen-H₅/H₆); 9.21 (s, slightly br, 2H, phen-H₂/H₉); 8.58 (dd, slightly br, 2H, ³J_HH = 4.0 Hz, ³J_HH = 7.8 Hz, phen-H₃/H₈); 5.85 (s, 9H, OCH₃); 0.09 (s, 3H, CH); -3.42 (s, 9H, C(O)CH₃). ¹H NMR (CDCl₃, 315 K, δ): 10.56 (d, slightly br, 2H, phen-H₄/H₇); 10.26 (s, slightly br, 2H, phen-H₅/H₆); 9.20 (s, br, 2H, phen-H₂/H₉); 8.40 (m, slightly br, 2H, phen-H₃/H₈); 5.80 (s, 9H, OCH₃); 0.82 (s, 3H, CH); −3.04 (s, 9H, C(O)CH₃). ¹H NMR (CDCl₃, 324 K, δ): 10.34 (s, br, 2H, phen-H₄/H₇); 10.08 (s, br, 2H, phen-H₅/H₆); 9.18 (s, br, 2H, phen-H₂/H₉); 8.31 (s, br, 2H, phen-H₃/H₈); 5.74 (s, 9H, OCH₃); 1.18 (s, 3H, CH); −2.82 (s, 9H, C(O)CH₃). PL (solid sample, 298 K, λ_excitation = 338 nm, nm): very strong, 578 (⁵D₀ → ⁷F₀, 0.6%); 590, 594 (⁵D₀ → ⁷F₁, 6.1%); 610, 613, 619, 624, 631 (⁵D₀ → ⁷F₂, 62.1%); 649, 654 (⁵D₀ → ⁷F₃, 3.5%); 685, 694, 701, 705 (⁵D₀ → ⁷F₄, 27.7%). PLE (solid sample, 298 K, λ_emission = 610 nm, nm): max 307–346 (ligand excitation); 394, 464 (Eu³⁺ excitation). Emission decay time (solid sample, λ_excitation = 325 nm, λ_emission = 613 nm): τ = 0.828 ms. Q_i = 67%.

Characterization of 2^Tb. C₂₇H₂₉N₂O₉Tb: calcd C 47.38, H 4.27, N 4.09; found 47.21, H 4.43, N 3.96. IR (KBr): ν_CO = 1616 cm⁻¹. ¹H NMR ((CD₃)₂SO, 298 K, δ): 109.61 (s, br, 2H, phen); 99.03 (s, br, 2H, phen); 77.64 (s, br, 3H, CH); 62.97 (s, br, 9H, CH₃); 7.84 (s, slightly br, 2H, phen); −21.08 (s, slightly br, 2H, phen); −31.65 (s, 9H, CH₃). PL (solid sample, 298 K, λ_excitation = 310 nm, nm): very strong, 486, 490, 496 (⁵D₄ → ⁷F₆, 19.2%); 541, 547 (⁵D₄ → ⁷F₅, 67.6%); 581, 590 (⁵D₄ → ⁷F₄, 6.7%); 617, 621 (⁵D₄ → ⁷F₃, 4.2%); 635–695 (⁵D₄ → ⁷F₂, ⁵D₄ → ⁷F₁, ⁵D₄ → ⁷F₀, 2.3%). PLE (solid sample, 298 K, λ_emission = 548 nm, nm): max 307–351 (ligand excitation). Emission decay time (solid sample, λ_excitation = 325 nm, λ_emission = 545 nm): τ = 1.212 ms. Q_i = 26%.

Characterization of 2^Dy. C₂₇H₂₉N₂O₉Dy: calcd C 47.13, H 4.25, N 4.07; found 46.97, H 4.45, N 3.95. IR (KBr): ν_CO = 1619 cm⁻¹. ¹H NMR (CDCl₃, 298 K, δ): 86.2 (s, very br, 3H, CH); 9.03 (s, slightly br, 2H, phen); 8.33 (s, slightly br, 2H, phen); 7.88 (s, slightly br, 2H, phen); 7.63 (s, slightly br, 2H, phen); −50.04 (s, br, 9H, CH₃); −55.47 (s, br, 9H, CH₃). PL (solid sample, 298 K, λ_excitation = 365 nm, nm): strong, 481, 486 (⁴F_9/2 → ⁶H_15/2, 29.0%); 576 (⁴F_9/2 → ⁶H_13/2, 66.2%); 663 (⁴F_9/2 → ⁶H_11/2, 3.4%); 751 (⁴F_9/2 → ⁶H_9/2, 1.4). PLE (solid sample, 298 K, λ_emission = 576 nm, nm): max 307–355 (ligand excitation). Emission decay time (solid sample, λ_excitation = 373 nm, λ_emission = 576 nm): τ = 3 μs. Q_i = 1%.

Characterization of 2^Yb. C₂₇H₂₉N₂O₉Yb: calcd C 46.42, H 4.18, N 4.01; found 46.29, H 4.30, N 3.90. IR (KBr): ν_CO = 1622 cm⁻¹. ¹H NMR ((CD₃)₂SO, 315 K, δ): 20.30 (s, br, 2H, phen); 19.08 (s, br, 2H, phen); 14.36 (s, br, 3H, CH); 13.05 (s, br, 2H, phen); 9.00 (s, br, 2H, phen); 8.33 (s, br, 9H, CH₃); −16.52 (s, br, 9H, CH₃). PL (solid sample, 298 K, λ_excitation = 351 nm, nm): medium, 978, 1010, 1040 (²F_5/2 → ²F_7/2). PLE (solid sample, 298 K, λ_emission = 1000 nm, nm): max 351 (ligand excitation).

Synthesis of [Y(DMM)₃(phen)] (3^Y) and [Ln(DMM)₃(phen)] (3^Ln) (Ln = Eu, Tb, Dy)

The experimental procedure for the preparation of 3^Y and 3^Ln is the same above described for 2^Y and 2^Ln, replacing K[MAA] with K[DMM] (0.511 g, 3.0 mmol). Yield >75% in all the cases.

Characterization of 3^Y. C₂₇H₂₉N₂O₁₂Y: calcd C 48.95, H 4.41, N 4.23; found 48.79, H 4.52, N 4.15. IR (nujol): ν_CO = 1667 cm⁻¹. ¹H NMR (CDCl₃, 298 K, δ): 9.52 (dd, 2H, ³J_HH = 4.8 Hz, ⁴J_HH = 1.7 Hz, phen-H₂/H₉); 8.41 (dd, 2H, ³J_HH = 8.2 Hz, ⁴J_HH = 1.7 Hz, phen-H₄/H₇); 7.89 (s, 2H, phen-H₅/H₆); 4.19 (s, 3H, CH); 3.47 (s, 9H, OCH₃).

Characterization of 3^Eu. C₂₇H₂₉N₂O₁₂Eu: calcd C 44.70, H 4.03, N 3.86; found 44.57, H 4.20, N 3.91. IR (nujol): ν_CO = 1661 cm⁻¹. ¹H NMR (CDCl₃, 298 K, δ): 12.87 (d, slightly br, 2H, ³J_HH = 3.5 Hz, phen-H₂/H₉); 10.00 (d, 2H, ³J_HH = 8.1 Hz, phen-H₄/H₇); 9.28 (s, 2H, phen-H₅/H₆); 8.78 (dd, 2H, ³J_HH = 3.5 Hz, ³J_HH = 8.1 Hz, phen-H₃/H₈); 3.56 (s, 18H, OCH₃); 0.06 (s, 3H, CH). ¹H NMR (CDCl₃, 298 K, δ): 12.15 (s, slightly br, 2H, phen-H₂/H₉); 9.79 (d, slightly br, 2H, ³J_HH = 8.1 Hz, phen-H₄/H₇); 9.06 (s, 2H, phen-H₅/H₆); 8.55 (m, 2H, phen-H₃/H₈); 3.54 (s, 18H, OCH₃); 0.66 (s, 3H, CH). PL (solid sample, 298 K, λ_excitation = 365 nm, nm): medium, 579 (⁵D₀ → ⁷F₀, 1.1%); 588, 591, 596 (⁵D₀ → ⁷F₁, 10.8%); 610, 615, 619 (⁵D₀ → ⁷F₂, 59.2%); 651 (⁵D₀ → ⁷F₃, 3.8%); 700, 707 (⁵D₀ → ⁷F₄, 25.1%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 616 nm): τ = 0.540 ms. Q_i = 25%.

Characterization of 3^Tb. C₂₇H₂₉N₂O₁₂Tb: calcd C 44.27, H 3.99, N 3.82; found 44.35, H 4.20, N 3.90. IR (nujol): ν_CO = 1664 cm⁻¹. No NMR data available because of paramagnetic relaxation. PL (solid sample, 298 K, λ_excitation = 365 nm, nm): very strong, 489 (⁵D₄ → ⁷F₆, 14.7%); 546 (⁵D₄ → ⁷F₅, 62.0%); 582, 588 (⁵D₄ → ⁷F₄, 9.9%); 620 (⁵D₄ → ⁷F₃, 7.7%); 635–695 (⁵D₄ → ⁷F₂, ⁵D₄ → ⁷F₁, ⁵D₄ → ⁷F₀, 5.7%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 557 nm): τ = 1.268 ms. Q_i = 27%.

Characterization of 3^Dy. C₂₇H₂₉N₂O₁₂Dy: calcd C 44.06, H 3.97, N 3.81; found 43.90, H 4.13, N 3.69. IR (nujol): ν_CO = 1664 cm⁻¹. No NMR data available because of paramagnetic relaxation. PL (solid sample, 298 K, λ_excitation = 365 nm, nm): strong, 480, 487 (⁴F_9/2 → ⁶H_15/2, 22.7%); 575 (⁴F_9/2 → ⁶H_13/2, 70.8%); 662 (⁴F_9/2 → ⁶H_11/2, 6.5%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 588 nm): τ = 6 μs. Q_i = 1–2%.

Preparation of doped polymers

2^Ln@PMMA (Ln = Sm, Eu, Tb, Dy) and 3^Tb@PMMA samples were prepared by adding under N₂ atmosphere a known amount of a freshly prepared 10⁻⁴ M solution of the proper complex in CH₂Cl₂ to a solution of 1.00 g of poly(methylmethacrylate) (PMMA, atactic, average M_w = 86 000 g mol⁻¹) in 50 mL of dichloromethane, to obtain complex: PMMA ratios ranging from 10 to 50 μmol g_PMMA⁻¹. The solutions were then concentrated to ca. 5 mL by evaporation under reduced pressure and transferred in a plastic polyethylene holder having 20 mm diameter. The residual solvent was allowed to evaporate overnight at room temperature and then the samples were kept under 10⁻³ torr pressure for 6 hours. Plastic glasses having about 2–3 mm thickness were obtained. A similar procedure was used for the preparation of 2^Eu@PVP (PVP = polyvinylpyrrolidone, average M_w = 10 000 g mol⁻¹), using anhydrous methanol instead of dichloromethane as solvent.

Characterization of 2^Sm@PMMA. PL (solid sample, 298 K, λ_excitation = 356 nm, nm): strong, 564 (⁴G_5/2 → ⁶H_5/2, 5.5%); 599, 606, 609 (⁴G_5/2 → ⁶H_7/2, 25.4%); 646, 653 (⁴G_5/2 → ⁶H_9/2, 63.2%); 709 (⁴G_5/2 → ⁶H_11/2, 5.9%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 646 nm): τ = 36 μs. Q_i = 1%.

Characterization of 2^Eu@PMMA. PL (solid sample, 298 K, λ_excitation = 338 nm, nm): very strong, 578 (⁵D₀ → ⁷F₀, 1.6%); 588, 592, 595 (⁵D₀ → ⁷F₁, 8.6%); 611, 616, 623 (⁵D₀ → ⁷F₂, 71.4%); 651 (⁵D₀ → ⁷F₃, 3.5%); 691, 703 (⁵D₀ → ⁷F₄, 14.9%). PLE (solid sample, 298 K, λ_emission = 610 nm, nm): max 335 (ligand excitation); 464 (Eu³⁺ excitation). Emission decay time (solid sample, λ_excitation = 325 nm, λ_emission = 613 nm): τ = 0.337 ms. Q_i = 19%.

Characterization of 2^Tb@PMMA. PL (solid sample, 298 K, λ_excitation = 310 nm, nm): very strong, 489, 496 (⁵D₄ → ⁷F₆, 24.3%); 545 (⁵D₄ → ⁷F₅, 64.6%); 583, 589 (⁵D₄ → ⁷F₄, 6.5%); 620 (⁵D₄ → ⁷F₃, 3.2%); 635–695 (⁵D₄ → ⁷F₂, ⁵D₄ → ⁷F₁, ⁵D₄ → ⁷F₀, 1.4%). PLE (solid sample, 298 K, λ_emission = 545 nm, nm): max 308–345 (ligand excitation). Emission decay time (solid sample, λ_excitation = 325 nm, λ_emission = 544 nm): τ = 0.342 ms. Q_i = 7%.

Characterization of 2^Dy@PMMA. PL (solid sample, 298 K, λ_excitation = 365 nm, nm): strong, 481, 484 (⁴F_9/2 → ⁶H_15/2, 17.9%); 575 (⁴F_9/2 → ⁶H_13/2, 75.4%); 662 (⁴F_9/2 → ⁶H_11/2, 6.7%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 573 nm): τ = 5 μs. Q_i = 1%.

Characterization of 2^Eu@PVP. PL (freshly prepared solid sample, 298 K, λ_excitation = 294 nm, nm): strong, 578 (⁵D₀ → ⁷F₀, 0.5%); 591 (⁵D₀ → ⁷F₁, 10.2%); 614, 618 (⁵D₀ → ⁷F₂, 53.0%); 650 (⁵D₀ → ⁷F₃, 1.8%); 687, 694, 697 (⁵D₀ → ⁷F₄, 34.5%). PLE (freshly prepared solid sample, 298 K, λ_emission = 616 nm, nm): max 273, 294 (ligand excitation). Emission decay time (freshly prepared solid sample, λ_excitation = 325 nm, λ_emission = 613 nm): τ = 0.686 ms. Q_i = 33%.

Characterization of 3^Tb@PMMA. PL (solid sample, 298 K, λ_excitation = 365 nm, nm): medium, 488 (⁵D₄ → ⁷F₆, 18.4%); 544 (⁵D₄ → ⁷F₅, 54.5%); 583, 588 (⁵D₄ → ⁷F₄, 14.5); 621 (⁵D₄ → ⁷F₃, 8.5%); 635–695 (⁵D₄ → ⁷F₂, ⁵D₄ → ⁷F₁, ⁵D₄ → ⁷F₀, 4.1%). Emission decay time (solid sample, λ_excitation = 350 nm, λ_emission = 544 nm): τ = 0.028 ms. Q_i < 1%.

Theoretical calculations

The computational geometry optimization of the lutetium model complexes Lu(MAA)₃ and Lu(acac)₃ was carried out without symmetry constrains using the hybrid DFT B3PW91 functional³⁰ in combination with the ECP-based basis set SDD, where the 4f shell of the lanthanide centre is included in a relativistic pseudopotential.³¹ The relative energies of the excited states were obtained from single-point time-dependent DFT calculations on the equilibrium geometries, using the previously described DFT functional and basis set.³² All the calculations were performed on a Intel-based x86-64 workstation and the software used was Gaussian '09.³³

Crystal structure determination of 2^Tb and 2^Yb

Crystallographic data for 2^Tband 2^Yb were collected at room temperature using a Bruker Smart 6000 CCD detector and Cu-Kα radiation (λ = 1.54178 Å) generated by a Incoatec microfocus source equipped with Incoatec Quazar MX optics. The software APEX2 (ref. 34) was used for collecting frames of data, indexing reflections, and the determination of lattice parameters, SAINT³⁴ for integration of intensity of reflections, and SADABS³⁴ for scaling and empirical absorption correction. The crystallographic treatment was performed with the Oscail program.³⁵ The structures were solved by charge flipping in arbitrary dimensions [Superflip program]³⁶ and refined by a full-matrix least-squares based on F².³⁷ Non-hydrogen atoms were refined with anisotropic displacement parameters. Hydrogen atoms were included in idealized positions and refined with isotropic displacement parameters. For 2^Yb, the Squeeze program³⁸ was used to correct the reflection data for the diffuse scattering due to disordered molecule present in the unit cell. And also for latter 2^Yb some positional disorder was found, but it was not modelled. In fact, all the methyl 3-oxo-butanoato-O,O′ ligand labelled as O71, O72, O73, C71, C72… is disordered over two positions, but the model including both positions does not improve the agreement factors, so only one position was refined, although BUMP and SIMU restrains were applied. CCDC 1448054 (2^Tb) and 1448055 (2^Yb) contain the ESI† crystallographic data for this paper. Details of crystal data and structural refinement are given in Table 4.

Table 4 Crystal data and structure refinement

Identification code	2^Tb	2^Yb
Empirical formula	C27H29N2O9Tb	C27H29N2O9Yb
Formula weight	684.44	698.56
Temperature	296(2) K	296(2) K
Wavelength	1.54178 Å	1.54178 Å
Crystal system	Triclinic	Triclinic
Space group	P1̄	P1̄
Unit cell dimensions	a = 9.3684(2) Å	a = 13.1594(6) Å
	b = 16.6936(3) Å	b = 16.6828(8) Å
	c = 18.2672(4) Å	c = 16.7793(8) Å
	α = 83.1135(10)°	α = 88.5817(16)°
	β = 88.7925(10)°	β = 72.2913(14)°
	γ = 87.0863(8)°	γ = 66.8382(14)°
Volume	2832.25(10) Å^3	3207.3(3) Å^3
Z	4	4
Density (calculated)	1.605 Mg m^−3	1.447 Mg m^−3
Absorption coefficient	12.743 mm^−1	5.803 mm^−1
F(000)	1368	1388
Crystal size	0.154 × 0.121 × 0.108 mm	0.171 × 0.144 × 0.087 mm
Θ range for data collection	2.437 to 68.172°	2.781 to 68.480°
Index ranges	−11 ≤ h ≤ 11	−15 ≤ h ≤ 15
	−20 ≤ k ≤ 20	−18 ≤ k ≤ 19
	−21 ≤ l ≤ 21	−20 ≤ l ≤ 20
Reflections collected	57 616	91 364
Independent reflections	10 084 [R(int) = 0.0612]	11 448 [R(int) = 0.0443]
Reflections observed (>2σ)	7399	10 211
Data completeness	0.973	0.970
Absorption correction	Semi-empirical from equivalents
Max. and min. transmission	0.7531 and 0.3489	0.7531 and 0.4872
Refinement method	Full-matrix least-squares on F^2
Data/restraints/parameters	10 084/0/714	11 448/42/714
Goodness-of-fit on F^2	1.063	1.036
Final R indices [I > 2σ(I)]	R1 = 0.0496	R1 = 0.0350
	wR2 = 0.1252	wR2 = 0.0998
R indices (all data)	R1 = 0.0698	R1 = 0.0385
	wR2 = 0.1436	wR2 = 0.1043
Largest diff. peak and hole	1.174 and −1.413 e Å^−3	1.307 and −0.644 e Å^−3

---

Acknowledgements

Ca' Foscari University of Venice is gratefully acknowledged for financial support (Progetti di Ateneo 2014). We also thank to the CACTI (University of Vigo) for X-ray data collection.

References

1. (a) K. Binnemans, Rare Earths Beta Diketonates, in Handbook on the Physics and Chemistry of Rare Earths, ed. K. A. Gschneidner Jr., J.-C. G. Bünzli and V. K. Pecharsky, Elsevier, Amsterdam, 2005, vol. 35, ch. 225, pp. 107–272; (b) S. Cotton, Scandium, Yttrium, and the Lanthanides, in Comprehensive Coordination Chemistry II, ed. G. F. R. Parkin, Elsevier, Amsterdam, 2005, vol. 3; (c) J.-C. G. Bünzli, S. Comby, A.-S. Chauvin and C. D. B. Vandevyver, J. Rare Earths, 2007, 25, 257; (d) K. Binnemans, Chem. Rev., 2009, 109, 4283; (e) J.-C. G. Bünzli and S. V. Eliseeva, J. Rare Earths, 2010, 28, 824; (f) S. V. Eliseeva and J.-C. G. Bünzli, Chem. Soc. Rev., 2010, 39, 189; (g) Y. Ma and Y. Wang, Coord. Chem. Rev., 2010, 254, 972; (h) B. Francis, D. B. Ambili Raj and M. L. P. Reddy, Dalton Trans., 2010, 39, 8084; (i) S. Liu, P. He, H. Wang, J. Shi and M. Gong, J. Lumin., 2010, 130, 855; (j) E. Di Piazza, L. Norel, K. Costuas, A. Bourdolle, O. Maury and S. Rigaut, J. Am. Chem. Soc., 2011, 133, 6174; (k) C. Tan and Q. Wang, Inorg. Chem. Commun., 2011, 14, 515; (l) V. I. Verlan, M. S. Iovu, I. Culeac, Y. H. Nistor, C. I. Turta and V. E. Zubareva, J. Non-Cryst. Solids, 2013, 360, 21; (m) O. Moudam, B. C. Rowan, M. Alamiry, P. Richardson, B. S. Richards, A. C. Jones and N. Robertson, Chem. Commun., 2009, 6649.
2. (a) M. Bortoluzzi, D. Battistel, S. Roppa, S. Daniele, A. Perosa and F. Enrichi, Dalton Trans., 2014, 43, 9303; (b) M. Bortoluzzi, E. Bianchin, S. Roppa, V. Bertolasi and F. Enrichi, Dalton Trans., 2014, 43, 10120.
3. (a) E. Erasmus, J. Conradie, A. Muller and J. C. Swarts, Inorg. Chim. Acta, 2007, 360, 2277; (b) L. A. Tomachynski, V. Y. Chernii, Y. Yu. Kolotilova, A. N. Chernega, J. A. K. Howard and S. V. Volkov, Inorg. Chim. Acta, 2007, 360, 1493; (c) R. Jain, A. K. Rai and R. C. Mehrotra, Polyhedron, 1986, 5, 1017.
4. A. P. Milanov, R. W. Seidel, D. Barreca, A. Gasparotto, M. Winter, J. Feydt, S. Irsen, H.-W. Becker and A. Devi, Dalton Trans., 2011, 40, 62.
5. (a) G. Gheno, M. Bortoluzzi, R. Ganzerla and F. Enrichi, J. Lumin., 2014, 145, 963; (b) M. Bortoluzzi, G. Paolucci, D. Fregona, L. Dalla via and F. Enrichi, J. Coord. Chem., 2012, 65, 3903; (c) M. Bortoluzzi, G. Paolucci, M. Gatto, S. Roppa, F. Enrichi, S. Ciorba and B. S. Richards, J. Lumin., 2012, 132, 2378; (d) M. Bortoluzzi, G. Paolucci, S. Polizzi, L. Bellotto, F. Enrichi, S. Ciorba and B. S. Richards, Inorg. Chem. Commun., 2011, 14, 1762; (e) M. Bortoluzzi, D. Battistel, G. Albertin, S. Daniele, F. Enrichi and R. Rumonato, Chem. Pap., 2016, 70, 43.
6. L. J. Farrugia, J. Appl. Crystallogr., 1997, 30, 565.
7. M. Llunell, D. Casanova, J. Cirera, P. Alemany and S. Álvarez, SHAPE, v. 2.1, Universitat de Barcelona and The Hebrew University of Jerusalem, Barcelona and Jerusalem, 2013.
8. D. Casanova, M. Llunell, P. Alemany and S. Alvarez, Chem.–Eur. J., 2005, 11, 1479.
9. H. Gallardo, G. Conte, A. J. Bortoluzzi, I. H. Bechtold, A. Pereira, W. G. Quirino, C. Legnani and M. Cremona, Inorg. Chim. Acta, 2011, 365, 152.
10. L. Jia, Y. C. Hui, Z. Li, H. L. Sun and Z. Wang, CrystEngComm, 2014, 16, 6483.
11. O. V. Konnik, V. F. Shul'gin, S. V. Abchairova, S. B. Meshkova, E. B. Rusanov and V. V. Minin, Russ. J. Coord. Chem., 2014, 40, 28.
12. P. Martín-Ramos, P. S. P. da Silva, V. Lavín, I. R. Martín, F. Lahoz, P. Chamorro-Posada, M. R. Silva and J. Martín-Gil, Dalton Trans., 2013, 42, 13516.
13. (a) K. Binnemans, Chem. Rev., 2007, 107, 2592; (b) J.-C. G. Bünzli and S. V. Eliseeva, Basics of Lanthanide Photophysics, Springer-Verlag, Berlin, 2010; (c) K. Binnemans, Coord. Chem. Rev., 2015, 295, 1.
14. (a) Q.-P. Li and B. Yan, Dalton Trans., 2012, 41, 8567; (b) S. M. Bruno, R. A. S. Ferreira, F. A. Almeida Paz, L. D. Carlos, M. Pillinger, P. Ribeiro-Claro and I. S. Gonçalves, Inorg. Chem., 2009, 48, 4882.
15. J. L. Vázquez-Ibar, A. B. Weinglass and H. R. Kaback, Proc. Natl. Acad. Sci. U. S. A., 2002, 99, 3487.
16. J.-C. G. Bünzli, Chem. Rev., 2010, 110, 2729.
17. (a) R. D. Shannon and C. T. Prewitt, Acta Crystallogr., Sect. B: Struct. Crystallogr. Cryst. Chem., 1969, 25, 925; (b) R. D. Shannon, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1976, 32, 751.
18. Z.-F. Li, L. Zhou, J.-B. Yu, H.-J. Zhang, R.-P. Deng, Z.-P. Peng and Z.-Y. Guo, J. Phys. Chem. C, 2007, 111, 2295.
19. J. J. Freeman and G. A. Crosby, J. Phys. Chem., 1963, 67, 2717.
20. Z. Li and J. Yu, J. Lumin., 2013, 143, 169.
21. S. Onuma, H. Inoue and S. Shibata, Bull. Chem. Soc. Jpn., 1976, 49, 644.
22. (a) J.-C. G. Bünzli, E. Moret, V. Foiret, K. J. Schenk, W. Mingzhao and J. Linpei, J. Alloys Compd., 1994, 207–208, 107; (b) S. Alves Jr, F. V. De Almeida, G. F. de Sá and C. de Mello Donegá, J. Lumin., 1997, 72–74, 478.
23. (a) K. Lunstroot, P. Nockermann, K. Van Hecke, L. Van Meervelt, C. Görrler Walrand, K. Binnemans and K. Driese, Inorg. Chem., 2009, 48, 3018; (b) H. F. Brito, O. L. Malta, M. C. F. C. Felinto, E. E. S. Teotonio, J. F. S. Menezes, C. F. B. Silva, C. S. Tomiyama and C. A. A. Carvalho, J. Alloys Compd., 2002, 344, 293.
24. Y. Hasegawa, S.-i. Tsuruoka, T. Yoshida, H. Kawai and T. Kawai, J. Phys. Chem. A, 2008, 112, 803.
25. (a) X. Y. Chen, M. P. Jensen and G. K. Liu, J. Phys. Chem. B, 2005, 109, 13991; (b) Z. Li, J. Yu, L. Zhou, H. Zhang and R. Deng, Inorg. Chem. Commun., 2008, 11, 1284.
26. M. Montalti, A. Credi, L. Prodi and M. T. Gandolfi, Handbook of Photochemistry, CRC Press Taylor & Francis Group, 3rd edn, 2006, ch. 3.
27. K. Buczko and M. Karbowiak, J. Lumin., 2013, 143, 241.
28. D. D. Perrin and W. L. F. Armarego, Purification of Laboratory Chemicals, Pergamon Press, Oxford, 3rd edn, 1988.
29. J.-C. G. Bünzli, A.-S. Chauvin, H. K. Kim, E. Deiters and S. V. Eliseeva, Coord. Chem. Rev., 2010, 254, 2623.
30. (a) P. Geerlings, F. De Proft and W. Langenaeker, Chem. Rev., 2003, 103, 1793; (b) T. Ziegler, Chem. Rev., 1991, 91, 651; (c) A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
31. (a) A. Bergner, M. Dolg, W. Kuechle, H. Stoll and H. Preuss, Mol. Phys., 1993, 80, 1431; (b) P. Schwerdtfeger, M. Dolg, W. H. E. Schwarz, G. A. Bowmaker and P. D. W. Boyd, J. Chem. Phys., 1989, 91, 1762.
32. F. Jensen, Introduction to Computational Chemistry, Wiley, 2nd edn, 2007.
33. M. J. Frisch, et al., Gaussian 09, Revision C.01, Gaussian, Inc., Wallingford, CT, 2010.
34. APEX2, SMART, SAINT, Bruker AXS Inc., Madison, Wisconsin, USA, 2007.
35. P. McArdle, J. Appl. Crystallogr., 1995, 28, 65.
36. L. Palatinus and G. Chapuis, J. Appl. Crystallogr., 2007, 40, 786.
37. G. M. Sheldrick, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 2008, 64, 112.
38. P. V. D. Sluis and A. L. Spek, Acta Crystallogr., Sect. A: Cryst. Phys., Diffr., Theor. Gen. Crystallogr., 1990, 46, 194.

---

Footnote

† Electronic supplementary information (ESI) available: ¹H NMR spectra of the 1^Y, 1^Ln, 2^Y and 2^Ln derivatives (Fig. S1–S12); plot of a typical voltammogram of an europium MAA derivative (Fig. S13); selected excitation spectra (Fig. S14–S18). Cartesian coordinates of the DFT-optimized structures of Lu(acac)₃ and Lu(MAA)₃. CCDC 1448054 and 1448055. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c6ra01741j

---