Carboxyl-functionalized ionic liquids: synthesis, characterization and synergy with rare-earth ions†

Talita Jordanna de Souza Ramos ^a, Guilherme Henrique Berton ^b, Tania Maria Cassol ^b and Severino Alves Júnior *^a
aLaboratório de Terras Raras, Departamento de Química, Centro de Ciências Exatas e da Natureza (CCEN), Universidade Federal de Pernambuco, Recife – PE, 50740-560, Brazil. E-mail: salvesjr@ufpe.br
^bGrupo de Líquidos Iônicos e Metais, Departamento de Química e Biologia, Universidade Tecnológica Federal do Paraná, São Francisco Beltrão – PR, 85601-970, Brazil

First published on 18th May 2018

---

---

Introduction

Ionic liquids (ILs) are defined as molten organic salts. Most have melting points lower than 100 °C. It is estimated that there are as many as 10¹⁸ ILs available, including their binary and ternary mixtures. The sheer number of ILs provides enormous scope for industrial and scientific innovations.¹ Room temperature ionic liquids (RTILs) are ILs with melting points below room temperature. A RTIL typically consists of nitrogen or phosphorous-containing organic cations and large organic or inorganic anions.² Among the requirements for the study of ILs/RTILs are the development of new structures, a database with all of the properties of the structures already synthesized, a comparison between commercial solvents and ionic liquids with potential solvent applications, and less costly synthetic methodologies.³ Beyond the already well established applications as solvents or catalysts,⁴ ionic liquids have been the subject of reviews detailing their preparation, and their applications in polymer (polyILs) synthesis,⁵ battery/fuel cell electrolytes,⁶ biological activity and pharmaceuticals and medicine.⁷ Other applications include their use in high performance luminophores⁸ or as coordinating ligands. ILs have several advantages over typical organic systems, including high solubility, structural flexibility, and thermal/chemical stability. An area not yet well explored is the photophysical properties of ILs in luminescence, where lanthanides have been highlighted as high-performing luminophores due to their luminescence features, such as quantum efficiencies of up to 100%,⁹ narrow luminescence bands and long lifetimes.¹⁰

The introduction of rare-earth compounds into ionic liquids dates from 2006¹¹ and has aroused so much interest that there has been an emergence of a wide field of research focused specifically on soft materials.^12,13 This field of research was motivated by the potential of ILs to minimize vibration-induced deactivation processes to design a rigid metal-ion environment, free of high-energy vibrations and protecting the Ln³⁺ ion from solvent interaction. Soft materials favourably combine the properties of ionic liquids with the unique optical properties of rare-earth compounds, such as a sharp emission band, long decay time in the excited state, and a tunable emission from the UV to the infrared spectral region. Even with the highlighted photophysical features,¹⁰ luminophores formed with several types of oxo-ligands^14–16 coordinated to lanthanide ions do not show photostability. It is well known in the literature that carboxylate termini provide several coordination modes that favour the formation of three-dimensional structures.¹⁷ The primary goal of our work was to develop a new family of ionic liquids and apply them to the development of photostable structures.

All technologies under development are based on solid state electroluminescent materials and belong to the general area of solid-state lighting (SSL). The main technologies being developed in SSL are light-emitting diodes (LEDs), organic light-emitting diodes (OLEDs), and light-emitting electrochemical cells (LECs). Light-emitting electrochemical cells (LECs) have emerged as an alternative option.¹⁸ Generally, LECs have several advantages over OLEDs, such as a single-layer configuration, balanced carrier injection, low voltage operation and allowing the use of air-stable electrodes.¹⁹ The luminescent materials in LECs are either luminescent polymers together with ionic salts, or ionic species.¹⁸ The desired properties of materials for LECs are high electrical conductivity and visible light emission. Furthermore, higher electroluminescence efficiencies are expected due to the well-known phosphorescent nature of the metal complexes.²⁰ Owing to these advantages, LECs based on metal complexes of Ir³⁺, Cu³⁺ and Ru³⁺ have attracted much attention in recent years.^21–23 Even with the unprecedented photophysical properties of lanthanide–IL complexes, the synthesis of materials potentially applicable for the construction of light-emitting electrochemical cells has not been reported.

This report is a first example of four ionic liquids (ILs) formulated from anhydrides, imidazole-based cations with appended carboxylic acids as terminals. We examine the potential of RTILs as building blocks for coordination chemistry with Ln(III). We focus on the structure–property relationship for luminescent ionogels from experimental results and a computational study of the photoluminescence. Spectroscopic properties, such as intensity parameters Ω_λ (λ = 2, 4 and 6), energy transfer (WET) and back transfer (WBT) rates, radiative (A_rad) and nonradiative (A_nrad) decay rates, and the quantum efficiency (η), of Eu(III) compounds were computationally modelled using the electronic and spectroscopic semi-empirical approach²⁴ widely used in photoluminescence studies.^25–28

Experimental section

All reagents were purchased from commercial sources as described in the ESI.† The synthesis of ionic liquids and luminescent ionogels is given in detail in the ESI.†

Synthesis of ionic liquids

Monocationic carboxylate ionic liquids (MC1–4) were prepared starting with 1-methylimidazolium and 1-bromoethanol, as shown in Fig. 1. Translucent and colourless crystals of bromide 2-(1-methylimidazolium-1-yl)-1-ethanol were obtained with up to 90% yield. For the synthesis of ILs, 2-(1-ethylimidazoly)-1-ethanol bromide was added to acetonitrile and 4 mmol of the corresponding anhydride (succinic, phthalic, phenyl succinic or 3,3-dimethyl glutaric). The ionic liquids that were obtained and their yields were: 2-(3-methylimidazolyl-1-yl)-ethoxy-4-oxobutanoic acid bromide (MC1), 80%; ((2-(3-methylimidazolyl-1-yl)-ethoxy)-carboyl)-benzoic-phthalic acid bromide (MC2), 78%; 4-(2-(3-methylimidazolyl-1-yl)-methyl)-4-oxo-2-phenylbutanoic acid bromide (MC3), 75%; and 3,3-dimethyl-5-(2-(3-methylimidazolyl-1-yl)-ethoxy)-5-xopentanoic acid bromide (MC4), 98%. The proposed mechanisms for the reactions are presented in the ESI,† (Fig. S1).

Synthesis of Ln(MCx)₃·yH₂O

Luminescent ionogels were synthesized from a simple metathesis reaction between the IL and lanthanide salt. Photographs of ionogels under daylight and UV are available in Fig. S2 (ESI†).

Characterization

Structural, thermal and photophysical characterizations of ionic liquids and luminescent ionogels are described in the ESI.†

Based on the theories developed by Malta,^29,30 we calculated the 4f–4f intensity parameters and emission quantum efficiency (η). The emission quantum efficiency was determined using eqn (1), where A_rad is the radiative decay rate, given by the sum of the spontaneous emission coefficients (A_Js) . The total radiative decay rate (A_total) is given by A_total = τ⁻¹, where τ is the lifetime for radiative decay.³¹ Finally, the non-radiative decay rate (A_nrad) is given by A_nrad = A_total − A_rad.

	(1)

Results and discussion

Synthesis and structural characterization

According to the methodology developed by our research group, four ionic liquids were obtained with high yields (75 to 98%). All ILs synthesized were translucent, colourless, and viscous at 25 °C and 1 atm (as seen in Fig. S2, ESI†).

The ¹H NMR spectra (Fig. S4, ESI†) show behaviour qualitatively similar to the ¹³C NMR spectra (Fig. S5, ESI†), and both are consistent with the COSY/HSQC characterization results (Fig. S6–S9, ESI†), suggesting the formation of the proposed structures of the ILs. The NMR spectra are shown in detail in the ESI† (Tables S1 and S2). FTIR spectroscopy interpretation was performed according to Table S3 (ESI†) that includes the intensity and attribution of each stretch. As expected, functionalized IL carboxyl groups can be identified by the absorption bands at 1560 and 1484 cm⁻¹, as shown in Fig. S10 (ESI†), which are assigned to the asymmetric stretching and symmetric stretching of the carboxylate group, respectively.³² FTIR spectra of MC1–4 showed bands at ∼1586 and 1560 cm⁻¹, which are assigned to the imidazole ring.³³ Vibrational bands in the far-IV spectra (<200 cm⁻¹) characterize the cation–anion interaction between the imidazolium cation and bromide ions.^34,35 Upon addition of Eu/Gd salts into the carboxyl-functionalized IL, the shift of absorption bands corresponding to the carboxylate moieties can be observed due to the coordination with Ln³⁺ ions (Fig. S11–S14, ESI†). The coordination of Ln³⁺ ions with ILs can be further confirmed by the luminescence data, which will be discussed later. A detailed discussion of the FTIR spectroscopy results for ligands and complexes is presented in the ESI.†

We describe a complete vibrational characterization via Raman spectroscopy, which is not commonly used for ionic liquids due to the fluorescence effect. Raman is more interesting than FTIR because the imidazolium-based ILs are highly ordered H-bonded polymeric liquids.³⁶ However, we obtained very sharp spectra, as shown in Fig. 2. We first focus on the region between 70–300 cm⁻¹, which is dominated by C–HA and/or N–HA (A = anion) interactions.³⁷ As can be clearly seen in Fig. 2, the maximum intensity of the measured spectrum occurs at ∼111 cm⁻¹ for MC3. Hydrogen bonds to the cation can be formed via N–H in the imidazolium cation and aromatic-H, explaining why the signal for MC2 is much less intense and shifted to 115 cm⁻¹ by steric hindrance. When the structure is incapable of strong hydrogen bonding via aromatic-H, the band is shifted to 104 cm⁻¹ (in MC1) and 100 cm⁻¹ (in MC4), and the intensity of the peak falls. The band shift to lower wavenumbers means a reduced interaction, suggesting that the strength of interaction between cation and anion is in an escalating order such that MC2 > MC3 > MC1 > MC4. Changes in the alkyl chain length cause changes in the interactive forces, which result in a decreased electrostatic attraction between the cation and anion.³⁸ Strong signals observed in the region at the high-frequency band (∼128 cm⁻¹) were enlarged due to a coupling with the peak associated with a wagging of the bonds in the N–C imidazolium ring,³⁹ and strong signals were also identified for all samples at ∼171 cm⁻¹. The Raman spectra of the ILs are quite similar to others in the 287–476 cm⁻¹ region, and at ∼1428 cm⁻¹; both are associated with CH₂(N)/CH₃(N)CN relative to the imidazolium structural stretch, as has been previously reported.⁴⁰ Low intensity peaks are identified in the 408 cm⁻¹ region associated with CH₂(N), CH₃ and CH₃(N)CH bonds.⁴¹ The O–C–O moiety has signals located at ∼588 cm⁻¹ (relating to symmetric bending) and 1103 to 1328 cm⁻¹ (relating to asymmetric bending).³⁰ Less significant peaks, and common to all samples, are observed at 788–888 cm⁻¹ arising from the ρ(CH₂) rocking.⁴² Pronounced peaks at 1028–1054 cm⁻¹, explained by the C–O symmetric strong peak,⁴⁰ and overlapped with ν(C–C), are observed in the MC1–MC4 structures.

An interesting behaviour is identified in the 1554–1744 cm⁻¹ region, related to CH₃(N)HCH symmetric bending with CH₂(N)/CH₃(N)CN.⁴⁰ For samples MC1 and MC4, these peaks disappear, while for samples MC2 and MC3, these peaks are closer and detached. This outcome can be explained by the similarities between the MC1/MC4 structures, which do not contain phenyl groups, and thus allow a greater flexibility in the organic chain. The relatively broad band ranging from 2800 to 2963 cm⁻¹, which is assigned to symmetrical and asymmetrical C–H stretching (δ(CH₃)) from the imidazolium cation (fingerprint bands), can serve as a useful probe to reflect structural change.⁴³ The subtle difference may be associated with the elongation of the MC4 organic chain compared to the MC1 structure, whereas in samples with a phenyl structure (MC2/MC3), this peak is superimposed on the enlargement of nearby peaks. The band identified at ∼2950 cm⁻¹ refers to ν_as(CH₃) and was identified in all samples. This peak is due to the presence of several CH₃ terminals in the structures.⁴⁴ Another signal common to all samples occurs at 3078–3089 cm⁻¹, attributed to ν(CH), and at ∼3096 cm⁻¹, assigned to ν((C–H)_ring.⁴⁴ Signals identified at 3159 cm⁻¹ (MC1), 3168 cm⁻¹ (MC4), and 3178 cm⁻¹ (MC2/MC3) are associated with the HCCH ring, and ν(C–H).³⁹Fig. 2 shows a strong band centred at 2959 cm⁻¹, together with a small peak seen as a shoulder located at approximately 3433 cm⁻¹. These peaks are associated with water molecules that interact preferentially with the anions, Br⁻⋯HOH⋯Br⁻.^34,45 We compared the results obtained with ionic liquids used as ligands, ionogels and lanthanide salts (Fig. S15–S22, ESI†). On the whole, in the luminescent materials we observed, all signals referring to the structures of the ionic liquids besides the emission of some electronic transitions to the trivalent ions, are already identified in the Raman lines.⁴⁶ The intensities of the characteristic peaks of the CO bond, strong in the IL, became weaker in the ionogels, accompanied by a shift to a higher frequency. This outcome might have resulted from coordination of the IL with the metal in this moiety.⁴⁷ Raman characterization of the Eu³⁺/Gd³⁺-ionogels is reported in detail in the ESI.† The assignments of several distinct Raman peaks of ionic liquids (MC1–4) and Gd³⁺/Eu³⁺ ionogels are listed in Tables S4 and S5 (ESI†), respectively.

Melting points

The melting points of the MC1 and MC4 ILs are 21 °C and 20 °C, respectively. These structures have aliphatic chains between the ester and the carboxylic acid termini, whereas MC2 and MC3 ILs have the phenyl structure, which hinders structural packing, and their melting points are 19 °C. The lower structural packing observed for MC2 and MC3 is consistent with the presence of fewer intermolecular forces (such as H-bonds) and lower melting points.⁴⁸

IL miscibility with several solvents

Miscibility of the 1-methylimizodalium carboxylate ILs with protic, aprotic, polar and nonpolar solvents is reported in Table S6 (ESI†).

Thermal analysis

The mechanisms of thermal degradation of the ionic liquids and ionogels are discussed in detail in the ESI,† (Pages 28 and 29). It is noteworthy that the ionic liquids and the ionogels are stable up to 150 °C and 223 °C, respectively. TG profiles are reported in Fig. S24 for ionic liquids (MC1–4) and Fig. S25 (ESI†) for ionogels.

Photophysical characterization

Studies reported in the literature indicate that imidazolium-based ionic liquids have significant absorption at approximately 260 nm.⁴⁹Fig. 3 shows the UV-Vis spectra of the elaborated ILs, where the signals for absorption of the imidazolium ring occur at ∼200 nm (broad and intense band) and a weaker and broader signal occurs at ∼280 nm (π; n → π* transitions), indicating the formation of a new species via intermolecular interactions in solution.⁵⁰ To prove that the absorption is not due to any impurity, we adopted a strategy different from that of comparing absorption spectra from repeatedly purified IL samples (Fig. S25, ESI†). That the absorption is actually due to the imidazolium moiety is proved by starting with pure 1-methylimidazole used in the synthesis of ionic liquids. The UV-Vis spectra exhibited a band centred at ∼205 nm assigned to the absorption of carboxylic acid (n → σ* transition).⁵¹ This signal is superimposed with broad and intense bands associated with the imidazole cation. At 227 nm (for MC2) and 220 nm (for MC3), the signal for absorption of the benzene ring can be observed. The molar absorption coefficient of the compounds depends on the concentration, ranging from approximately 5 M⁻¹ cm⁻¹ up to 35 M⁻¹ cm⁻¹. This outcome indicates the formation of a new species via intermolecular interactions taking place in solution. This formation is common behaviour for ionic liquids and has been previously studied.⁵⁰

The UV-Vis spectra of the ionogels present similar behaviour to that observed in the ILs (Fig. S26, ESI†). The bands identified at ∼200 nm and ∼280 nm (π; n → π* transitions) refer to the imidazolium moiety and intermolecular interactions.⁵⁰ The complexes Eu(MC2)₃(H₂O)₂ and Eu(MC3)₃(H₂O)₂ show bands at ∼237 nm associated with absorption of the benzene ring. In comparison with the free ligand, these peaks were displaced by 10 nm (for the MC2 complex) and 17 nm (for the MC3 complex) to regions of lower energy. Fig. S26 (ESI†) shows the absorption spectra for ionogels with Eu³⁺. It's noteworthy that the ionogels are transparent in the visible range.

Steady-state photoluminescence spectra (Fig. 4) of the ionic liquids were obtained from pure samples and without the use of solvents. As observed in previous studies,⁵⁰ the excitation spectra of the ILs confirm the observation of the prominence of intermolecular interactions, with a main excitation maximum at ∼323 nm and a very small component at ∼300 nm, the contribution of which grows with increasing concentration. The excitation spectra reached a maximum approximately at 358 nm, 332 nm, and 339 nm, for MC1, MC2 and MC4 compounds, respectively. This excitation is related to the π → π* transition. The emission for a single excitation wavelength is found to consist of two components, as previously reported.⁴⁹ Ionic liquids, in particular those based on the imidazolium cation, should be more accurately considered as three dimensional networks of anions and cations linked together by weak interactions,³⁶ as recorded using UV-Vis/NMR measurements. The short wavelength emission (here, at ∼430 nm) is due to the monomeric form of the imidazolium ion, and the long wavelength emission (∼550 nm) is due to its well-known associated forms.³⁶ Emissions are identified as intraligand transitions and assigned as π* → π. The Stokes displacements were calculated to be: 79 nm for MC1, 104 nm for MC2, 96 nm for MC3, and 86 nm for MC4. We observed that the ionic liquids that have a benzene ring in their structure demonstrate a greater separation between the triplet levels, which causes a smaller overlap between the excitation–emission bands, and thus, a greater Stokes displacement.

The emission identified for the Gd³⁺ complexes corresponds to the emission of the ionic liquid because there is no transfer of energy from the IL to the Ln³⁺ ion. The Gd³⁺ emission (from the first excited state ⁶P_7/2) has a higher energy, typically 32200 cm⁻¹, corresponding to 310 nm.⁵² Emission and excitation spectra of the Gd³⁺–IL complexes are shown in Fig. S27–S30 (ESI†). In the comparison between these spectra and the photoluminescence results of the ionic liquids (MC1–4), it is noteworthy that there was a shift between the excitation and emission bands to lower wavelengths. The variation of the maximum emission (Δλ_Em) and excitation (Δλ_Ex) was calculated, for each complex, indicating the modification of the ligand before coordination with the gadolinium. Ligand triplet state energy was estimated from the luminescence spectra of Gd(MCx)₃(H₂O)_y complexes using the methods previously reported.²⁹ Triplet level values and spectrum modifications (Δλ_Em and Δλ_Ex) are set out in Table 1.

---

We compare the diagrams of the singlet and triplet levels of the ILs and of Eu³⁺ in Fig. 5. All the systems showed that the triplet level of the ionic liquid is between 23000 and 30000 cm⁻¹. This fact is of immense relevance in the elaboration of high performance luminescent systems, due to the formation of high energy phonons, which deactivate the non-radiative ions.⁴⁶ Thus, the ligands reported in this work present a remarkable potential in synergistic coordination with the Eu³⁺ ion. This synergy is discussed below.

Excitation spectra were obtained by monitoring the ⁵D₀ → ⁷F₂ line (619 nm for Eu(MC1)₃(H₂O)₃; 614 nm for Eu(MC2)₃(H₂O)₂; 616 nm for Eu(MC3)₃(H₂O)₂; and 617 nm for Eu(MC4)₃(H₂O)₃). The excitation spectrum consists of a series of sharp absorption lines, positioned in the 320–580 nm region, which can be attributed to transitions within the 4fⁿ configurations of Eu³⁺ ions.¹⁰ The most intense transition of the excitation spectrum indicates the wavelength at which the direct excitation in the ion should give the greatest contribution to the photoluminescence of the compound. This wavelength is equal to 394 nm for Eu(MC1)₃(H₂O)₃, Eu(MC2)₃(H₂O)₂ and Eu(MC₄)₃(H₂O)₃, and 395 nm for Eu(MC3)₃(H₂O)₂. Fig. 6, and Fig. S31–S34 (ESI†) show the corresponding emission spectra (red line) of the complexes obtained from direct ion excitation in Eu³⁺ (transition ⁷F₀ → ⁵L₆). In these spectra, five emission bands can be seen, centred at 580, 595, 612, 657 and 702 nm, distinguished in all complexes, and assigned to the ⁵D₀ → ⁷F_J (J = 0–4) transitions in Eu³⁺.⁵³

The emission spectrum (Fig. S35, ESI†) obtained from the excitation in the band of each ligand (358 nm for Eu(MC1)₃(H₂O)₃; 332 nm for Eu(MC2)₃(H₂O)₂; 325 nm for Eu(MC3)₃(H₂O)₂; and 340 nm for Eu(MC1)₃(H₂O)₃) does not show the ligand phosphorescence band between 392–450 nm. Excitation of the complexes in the ligand part (MC1–4) gives the same spectral characteristics as the ion-centred Eu³⁺ emission. Emission spectra of Eu(MC1)₃(H₂O)₃ show transition 0–4 with relevant intensity, which can be explained by the greater polarizability of the MC1 ligand compared to the other ILs (MC2, MC3 and MC4).⁵⁴ The presence of donor groups or acceptors in the ligand structures modifies the electron density around the ion and interferes with the position of the ⁵D₀ → ⁵F₀ transition.⁴⁹ This transition was observed at 579 nm for Eu(MC1)₃(H₂O)₃, Eu(MC3)₃(H₂O)₂ and Eu(MC4)₃(H₂O)₃, and at 577 nm for Eu(MC2)₃(H₂O)₂. This outcome suggests that MC2 presents a lower polarizability when compared to the other ligands. The lower intensity of the 0–4 transition of Eu(MC2)₃(H₂O)₂ corroborates this fact. Less polarizable environments, in general, contribute to decreasing the covalent character of the metal–ligand bonds in the complexes (nephelauxetic effect), which implies a lower displacement for the 0–0 transition, and a lower intensity for the 0–4 transition.⁴⁸ The presence of the ⁵D₀ → ⁵F₀ transition indicates that the Eu³⁺ ion is located in a symmetry site of the type C_s, C_n, or C_nv. This fact, together with the mono-exponential profile of each mono-exponential system of the ⁵D₀ excited state radiative decay curve of the Eu³⁺ ion (Fig. S36 and S37, ESI†), indicates that Eu³⁺ ions predominantly exhibit coordination environments with only one symmetry centre.⁵⁵ Emission spectra of the complexes (Fig. S31–S34, ESI†) show the multiplicities of the transitions, indicating that the point group for the environment of symmetry of the europium ion is around C_2v.⁵⁵ This outcome corroborates the hypothesis, represented by the ⁵D₀ → ⁵F₀ transition, that the chemical environment around Eu³⁺ must be of low symmetry. The fluorescence lifetimes were analysed throughout the spectral range of the europium ion (excitation at ⁷F₀ → ⁵L₆ and emission at ⁵D₀ → ⁷F₂), as shown in Fig. S36 (ESI†). We also examined the decay profiles measuring the temporal behaviour of the fluorescence by exciting the organic part of the samples and monitoring the fluorescence emission at ⁵D₀ → ⁷F₂ (Fig. S37, ESI†). All fluorescence decay profiles could be best fitted to a mono-exponential decay function, as shown in Fig. S36 and S37 (ESI†). The fluorescence lifetimes reported in this work are four times greater than those of luminescent gels, elaborated with europium and carboxylate ligands, recognized as high emission performance gels reported in the literature.⁵⁶

We calculated the parameters of the experimental intensities: Ω₂ and Ω₄, Einstein spontaneous emission coefficient (A_rad), the non-spontaneous emission coefficient (A_nrad), quantum emission efficiency (η), and finally, R_0–2/0–1 and R_0–4/0–1, following well-established models in the literature for the determination of the energy transfer of ligand–lanthanide,²⁹ the position and nature of excited states,⁵⁷ and the intensity parameters.⁵⁸Table 2 shows these results and shows the values of emission lifetimes from ⁵D₀ in Eu³⁺.

The disparity between non-radiative rates can be explained by observing the vibrational spectra in the infrared region of these complexes. The FTIR spectra of Eu(MC1)₃(H₂O)₃ and Eu(MC4)₃(H₂O)₃ complexes show O–H (water-group) absorption bands at 3384 and 3393 cm⁻¹, and the Eu(MC2)₃(H₂O)₂ and Eu(MC3)₃(H₂O)₂ complexes show lower intensity signals at 3476 and 3424 cm⁻¹, respectively. The bands for the MC2/MC3 complexes are in a region of lower energy with respect to the bands for the MC1/MC4 complexes. The energy difference between ⁵D₀ and ⁷F₆ (for Eu³⁺ ion) is ∼12300 cm⁻¹. Therefore, there is a better resonance condition involving three phonons (3 × 3393 cm⁻¹) in the case of the Eu(MC1)₃(H₂O)₃/Eu(MC4)₃(H₂O)₃ complexes, causing a considerable increase in the rate of non-radiative decay. As can be seen in Table 2, MC1/MC4 complexes show lower efficiency. TGA results for Eu(MC1)₃(H₂O)₃/Eu(MC4)₃(H₂O)₃ showed 3 coordination water molecules, whereas for Eu(MC2)₃(H₂O)₂ and Eu(MC3)₃(H₂O)₂, only 2 water molecules were observed, which corroborates the vibrational spectra (FTIR/Raman) and justifies the higher quantum efficiency presented by these complexes. An analysis of Table 2 can be understood from the separation of the structures of the ligand. For ligands with alkyl chains (Eu(MC1)₃(H₂O)₃ and Eu(MC4)₃(H₂O)₃), we observe a high nonradiative decay rate in relation to the radiative decay rate, the lowest quantum efficiency, and the shortest lifetime. These observations can be associated with the high incidence of non-radiative processes, as vibrational coupling of O–H oscillators from water molecules coordinated with Eu³⁺. Eu(MC2)₃(H₂O)₂ and Eu(MC3)₃(H₂O)₂ had low nonradiative decay rates (A_nrad) in relation to high radiative decay rates (A_rad), higher quantum efficiency, and a lifetime beyond 1 ms. The A_rad values and lifetimes are the highest reported compared to luminescent ionogels with europium ions reported in the literature.^59–61 The values reported for the photophysical properties of Eu(MC2)₃(H₂O)₂ and Eu(MC3)₃(H₂O)₂ can be explained by the position of the triplet level in the energy diagram of the ionic liquids being resonant with the emitting level of Eu³⁺ (Fig. 5). We observed that Eu(MC2)₃(H₂O)₂ showed greater efficiency compared to complexes with ligand structures similar to those proposed in this work.⁶² We compared the luminescence properties of different europium complexes to evaluate the effect of the alkyl chain length and flexibility. The ligand structures of MC1 and MC4 did not present steric impediments, besides having a greater structural flexibility, which implies a greater absorption of water and a closer approximation between neighbouring ions. The self-quenching process between neighbouring europium complexes could cause the nonradiative dissipation of excited states.¹³ The effective isolation and confinement of europium complexes inside more rigid structures, such as Eu(MC2)₃(H₂O)₂ and Eu(MC3)₃(H₂O)₂, were crucial to increasing luminescence efficiency, which explains the higher values of quantum efficiency and lifetime associated with these systems. Due to the nature of the chemical environment in terms of the intensity of ⁵D₀ → ⁷F₂ and ⁵D₀ → ⁷F₁ transitions, their intensity ratio (R_0–2/0–1) could be used to assess the asymmetry of the Eu³⁺ coordination environment.⁶³ A relatively higher ratio usually denotes a relatively higher degree of asymmetry and better luminescent monochromaticity.¹³ The disparity in the values found for R_0–2/0–1 and R_0–4/0–1 represents the structural differences caused by the ionic liquids used as ligands. The high values of R_0–2/0–1 can be explained by the ionic nature of the ligands that involves a load compensation through the coordination network. The fact that no load transfer band in the spectrum can be seen supports this explanation. The high values of R_0–2/0–1 suggest environments with low symmetry, as registered by the emission spectrum.

---

The Judd–Ofelt parameter Ω₂ reported for Eu(MC2)₃(H₂O)₂ is similar to the values found in the literature for organic ligands with benzenes and carboxylates.^63,64 The disparity of the results reported in the literature compared to the other structures described in this work (Eu(MC1)₃(H₂O)₃, Eu(MC3)₃(H₂O)₂ and Eu(MC4)₃(H₂O)₃) indicates the strong differences in the chemical environments for Eu³⁺ ions between these compounds. These differences can be explained by the benzene structure in the main chain (MC2) and benzenehexa carboxylic acid.⁶³ In MC3, the benzene ring is only a branch of the main chain. In Eu(MC1)₃(H₂O)₃ and Eu(MC4)₃(H₂O)₃, which show great differences between Ω₂ values, there is no benzene ring. The parameter Ω₄ reflects the stiffness of the chemical environment around the lanthanide ion. The lowest Ω₄ values were exhibited by the coordination network Eu(MC3)₃(H₂O)₂, reflecting the network effect on the structural rigidity of the material. Similar ligand structures, Eu(MC1)₃(H₂O)₃ and Eu(MC4)₃(H₂O)₃, show congruent results that indicate the large size of the structural difference caused by the absence of the benzene ring. Preliminary measures of absolute emission quantum yields are reported in Table S8 (ESI†). We calculated the sensitization efficiency for the ligand based on previously reported methods.⁶⁵ The high quantum yield observed for these complexes seems to be due to the calculated high yield energy transfer from these ligand states to the quasi-resonant Eu³⁺ energy levels.

The photostability of the Eu³⁺-complexes was evaluated by monitoring the emission/excitation spectra intermittently under continuous exposure to UV irradiation. Under UV exposure (UVA at 365 nm and UVB at 320 nm), the luminescence intensities of the intra configurational Eu³⁺ transitions suffer insignificant changes, as observed in Fig. S38 and S39 (ESI†). Using the intensity of the ⁵D₀–⁷F₂ transition, we produced a graph that provides evidence of the photostability of the systems against UV radiation (Fig. 7).

We studied the photostability by comparing the measurements performed on the same sample over three years (2016, 2017 and 2018). The emission and excitation spectra are reported in Fig. S41 (ESI†) for all the coordination systems. The Eu complexes showed great chemical stability and photostability to produce exactly the same spectral emissions/excitations even three years after synthesis of the ionogels.

The computational method²⁴ employed in this work has already been well established in the literature and its main objective has been to provide data for a detailed study of the photoluminescence of for ionogels,²⁵ crystals,²⁶ composities²⁷ and general complexes.²⁸ This methodology presented consistent results for each of the elaborated complexes. All information pertaining to each structure and the geometry optimization method are reported in the ESI,† (Table S15). In the determination of the coordination sphere of the complexes, we found very reliable results for which we have the lowest percentage errors that have been reported for this computational procedure.^26,62 The structures of the ionogels were optimized using the Sparkle/PM7 model. The numbers of water molecules are in accordance with those predicted by thermogravimetric analysis. Fig. 8 shows a representative coordination sphere for Eu(MC2)₃(H₂O)₂. Figures of coordination geometries for Eu(MC1)₃(H₂O)₃, Eu(MC3)₃(H₂O)₂, and Eu(MC4)₃(H₂O)₃ complexes are available in the ESI,† (Fig. S38–S41). As shown in Fig. 8, the ligand coordination comes from its carboxylate group in a chelating coordination, as predicted by FTIR analysis. The coordination polyhedron is organized in a distorted symmetry as a bicapped trigonal prism, which indicates that the arrangement is in point group symmetry C_2v.⁶⁶ This symmetry requires sp³d³f hybrids that are common coordination polyhedra for a coordination number of eight or nine.⁶⁶ The proposed symmetry is consistent with the estimate from the emission spectra of all compounds. The distances between the metal centre and the oxygen atoms of the ionic liquid and coordinated water are shown in Tables S9–S11 in the ESI.†

Spectroscopic properties as intensity parameters Ω_λ (λ = 2, 4 and 6), energy transfer (WET) and back transfer (WBT) rates, radiative (A_rad) and nonradiative (A_nrad) decay rates, the quantum efficiency (η) and quantum yield (q) of europium compounds were calculated using semi-empirical models specialized for the study of photoluminescence of europium ions (LUMPAC).¹⁴ All results are exhibited in Table 3 and are in agreement with those obtained experimentally, evidencing the accuracy of the computational methodology. The outstanding photophysical properties inspired us to investigate the mechanisms of population transfer involved in the systems. Intramolecular energy transfer (ET) and back transfer (BT) were calculated, considering that the Eu³⁺ levels arose from the metal ion at an intermediate coupling.²⁴ Theoretical data are reported for all complexes in Tables S12–S15 (ESI†). Energy transfer for Eu(MC1)₃(H₂O)₃, Eu(MC2)₃(H₂O)₂ and Eu(MC4)₃(H₂O)₃ complexes follows a multipolar mechanism. For Eu(MC3)₃(H₂O)₂, the profile suggests that the energy transfer occurs by an exchange mechanism.^67,68 The population diagrams of the complexes showed the synergy between the carboxylate ionic liquids and the Eu³⁺ ions (Fig. S50–S54, ESI†).

---

We note that Eu³⁺/Gd³⁺ ionogels presented a similar coordination mode (as evidenced by FTIR and Raman spectra) and thermal stability until 223 °C (Fig. S23, ESI†) associated with high viscosity at room temperature. The interesting features of the soft materials include outstanding ionic conductivity due to a high content of ILs, easy coating on surfaces, and excellent luminescence properties (e.g., long lifetime, high colour purity, highlighted quantum efficiency, and photostability). These features might render them extremely valuable for various optical applications. The ionogels can also be entirely solution-processed, which is largely favourable to the development of low-cost and large-area lighting, or yet applicable for the composition of flexible displays in light emitting devices (LEDs), light-emitting electrochemical cells, and luminescent solar concentrators.

Conclusions

In summary, we reported, designed and synthesized an imidazolium based-ionic liquid with appended carboxylic acids as termini, which was directly derived from anhydrides. The ionic liquids produced show the following: high yields (up to 98%) and high purity (proven by NMR, elementary analysis and melting point); thermal stability up to 100 °C, associated with melting points below room temperature, chemical stability in water and common organic solvents, as demonstrated by miscibility tests; and featured luminescence with intensities on the order of 10⁶ photons s⁻¹. With these aforementioned characteristics, carboxyl-functionalized ILs showed triplet levels in the best possible region (23000–30000 cm⁻¹) for synergistic coordination with lanthanide ions.

Our study demonstrates a complete analysis of the structure–property relationship for each complex (Eu(MC1)₃(H₂O)₃, Eu(MC2)₃(H₂O)₂, Eu(MC3)₃(H₂O)₂ and Eu(MC4)₃(H₂O)₃). The coordination polyhedron is in bicapped trigonal prism symmetry, which is consistent with emission spectrum estimates from all the compounds. The supramolecular ionogels had high viscosity at room temperature, transparency, and thermal stability until 223 °C, and strong photoluminescence (τ > 1 ms, η = 48% and φ = 27%). The extraordinary photophysical properties (presented in a detailed experimental and theoretical study) thus make these materials extremely valuable for optical applications such as light-emitting electrochemical cells. Our future research will focus on the fabrication of emitting devices based on Eu(MC1)₃(H₂O)₃, Eu(MC2)₃(H₂O)₂, Eu(MC3)₃(H₂O)₂ and Eu(MC4)₃(H₂O)₃ as thin films.

Conflicts of interest

The authors declare no competing financial interests.

Acknowledgements

This work was supported by FACEPE: APQ-0675-1.06/14. We gratefully acknowledge LAC-UFPE, Analytical Central (dQF-UFPE) and CETENE for carrying out several analyses. We thank CAPES/CNPQ for financial support APQ-0549-1.06/17 INCT.

Notes and references

1. K. Dong, X. Liu, H. Dong, X. Zhang and S. Zhang, Chem. Rev., 2017, 117, 6636–6695 CrossRef PubMed .
2. P. Wasserscheid and W. Keim, Angew. Chem., 2000, 39, 3772–3789 CrossRef .
3. T. Murphy, R. Atkin and G. G. Warr, Curr. Opin. Colloid Interface Sci., 2015, 20, 282–292 CrossRef .
4. Y. Qiao, W. Ma, N. Theyssen, C. Chen and Z. Hou, Chem. Rev., 2017, 117, 6881–6928 CrossRef PubMed .
5. M. Suckow, M. Roy, K. Sahre, L. Häußler, N. Singha, B. Voit and F. Böhme, Polymer, 2017, 111, 123–129 CrossRef .
6. M. Watanabe, M. L. Thomas, S. Zhang, K. Ueno, T. Yasuda and K. Dokko, Chem. Rev., 2017, 117, 7190–7239 CrossRef PubMed .
7. K. S. Egorova, E. G. Gordeev and V. P. Ananikov, Chem. Rev., 2017, 117, 7132–7189 CrossRef PubMed .
8. S. Hisamitsu, N. Yanai and N. Kimizuka, Angew. Chem., Int. Ed., 2015, 54, 11550–11554 CrossRef PubMed .
9. M. Bredol, U. Kynast and C. Ronda, Adv. Mater., 1991, 3, 361–367 CrossRef .
10. K. Binnemans, in Handbook on the Physics and Chemistry of Rare Earths, ed. J.-C. Bünzli and V. Pecharsky, Elsevier Amsterdam, Amsterdam, 1st edn, 2005, pp. 1–480 Search PubMed .
11. P. Nockemann, B. Thijs, N. Postelmans, K. Van Hecke and L. Van Meervelt, J. Am. Chem. Soc., 2006, 128, 13658–13659 CrossRef PubMed .
12. D. Wang, H. Wang and H. Li, ACS Appl. Mater. Interfaces, 2013, 5, 6268–6275 Search PubMed .
13. S. Yi, J. Wang, Z. Feng and X. Chen, J. Phys. Chem. B, 2017, 121, 9302–9310 CrossRef PubMed .
14. Y. Ding, Y. Wang, H. Li, Z. Duan, H. Zhang and Y. Zheng, J. Mater. Chem., 2011, 21, 14755 RSC .
15. E. B. Gibelli, J. Kai, E. E. S. Teotonio, O. L. Malta, M. C. F. C. Felinto and H. F. Brito, J. Photochem. Photobiol., A, 2013, 251, 154–159 CrossRef .
16. P. Nockemann, E. Beurer, K. Driesen, R. Van Deun, K. Van Hecke, L. Van Meervelt and K. Binnemans, Chem. Commun., 2005, 4354 RSC .
17. G.-L. Law, K.-L. Wong, K.-K. Lau, S. Lap, P. A. Tanner, F. Kuo and W.-T. Wong, J. Mater. Chem., 2010, 20, 4074 RSC .
18. R. D. Costa, E. Ortí, H. J. Bolink, F. Monti, G. Accorsi and N. Armaroli, Angew. Chem., Int. Ed., 2012, 51, 8178–8211 CrossRef PubMed .
19. F. Dumur, in Luminescence in Electrochemistry, ed. F. Miomandre and P. Audebert, Springer, Cham, 2017, pp. 327–361 Search PubMed .
20. R. Sun, C. T. Liao and H. C. Su, Org. Electron., 2014, 15, 2885–2892 CrossRef .
21. S. Gaillard, J.-L. Renaud and M. Elie, in Light-Emitting Electrochemical Cells, ed. R. D. Costa, 2017, pp. 287–327 Search PubMed .
22. E. Matteucci, A. Baschieri, A. Mazzanti, L. Sambri, J. Ávila, A. Pertegás, H. J. Bolink, F. Monti, E. Leoni and N. Armaroli, Inorg. Chem., 2017, 56, 10584–10595 CrossRef PubMed .
23. C. D. Ertl, C. Momblona, A. Pertegás, J. M. Junquera-Hernández, M.-G. La-Placa, A. Prescimone, E. Ortí, C. E. Housecroft, E. C. Constable and H. J. Bolink, J. Am. Chem. Soc., 2017, 139, 3237–3248 CrossRef PubMed .
24. J. D. L. Dutra, T. D. Bispo and R. O. Freire, J. Comput. Chem., 2014, 35, 772–775 CrossRef PubMed .
25. T. Jordanna, D. S. Ramos, S. Viana, L. Schaidhauer, T. Cassol and S. A. Junior, J. Mater. Chem. C, 2015, 3, 10934–10942 RSC .
26. C. V. Rodrigues, L. L. Luz, J. D. L. Dutra, S. A. Junior, O. L. Malta, C. C. Gatto, H. C. Streit, R. O. Freire, C. Wickleder and M. O. Rodrigues, Phys. Chem. Chem. Phys., 2014, 16, 14858–14866 RSC .
27. T. C. de Oliveira, H. P. Santos, M. G. Lahoud, D. F. Franco, R. O. Freire, J. D. L. Dutra, A. Cuin, J. F. de Lima and L. F. Marques, J. Lumin., 2017, 181, 196–210 CrossRef .
28. F. C. M. Lippy, F. Marques, H. P. Santos, K. A. D’Oliveira, N. P. Botezine, M. C. R. Freitas, R. O. Freire, J. D. L. Dutra, J. S. Martins, C. Legnani and W. G. Quirino, Inorg. Chim. Acta, 2017, 458, 28–38 CrossRef .
29. O. L. Malta, H. F. Brito, J. F. S. Menezes, F. R. G. Silva, S. Alves, F. S. Farias and A. V. M. de Andrade, J. Lumin., 1997, 75, 255–268 CrossRef .
30. O. L. Malta, J. Non-Cryst. Solids, 2008, 354, 4770–4776 CrossRef .
31. A. Florez, Y. Messaddeq, O. L. Malta and M. A. Aegerter, J. Alloys Compd., 1995, 227, 135–140 CrossRef .
32. M. Q. Ao, G. Y. Xu, Y. Y. Zhu and Y. Bai, J. Colloid Interface Sci., 2008, 326, 490–495 CrossRef PubMed .
33. Y.-S. Ding, M. Zha, J. Zhang and S.-S. Wang, Colloids Surf., A, 2007, 298, 201–205 CrossRef .
34. Z. Liu, S. Z. El Abedin and F. Endres, Phys. Chem. Chem. Phys., 2015, 17, 15945–15952 RSC .
35. K. Fumino, A. Wulf and R. Ludwig, Angew. Chem., Int. Ed., 2008, 47, 3830–3834 CrossRef PubMed .
36. J. Dupont, R. F. de Souza and P. A. Z. Suarez, Chem. Rev., 2002, 102, 3667–3692 CrossRef PubMed .
37. T. Peppel, C. Roth, K. Fumino, D. Paschek, M. Köckerling and R. Ludwig, Angew. Chem., Int. Ed., 2011, 50, 6661–6665 CrossRef PubMed .
38. Y. Yoshimura, M. Shigemi, M. Takaku, M. Yamamura, T. Takekiyo, H. Abe, N. Hamaya, D. Wakabayashi, K. Nishida, N. Funamori, T. Sato and T. Kikegawa, J. Phys. Chem. B, 2015, 119, 8146–8153 CrossRef PubMed .
39. K. Malek, A. Puc, G. Schroeder, V. I. Rybachenko and L. M. Proniewicz, Chem. Phys., 2006, 327, 439–451 CrossRef .
40. K. Noack, P. S. Schulz, N. Paape, J. Kiefer, P. Wasserscheid and A. Leipertz, Phys. Chem. Chem. Phys., 2010, 12, 14153 RSC .
41. T. Moumene, E. H. Belarbi, B. Haddad, D. Villemin, O. Abbas, B. Khelifa and S. Bresson, J. Mol. Struct., 2015, 1083, 179–186 CrossRef .
42. G. Socrates, Infrared and Raman Characteristic Group Frequencies: Tables and Charts, 3rd edn, 2004 Search PubMed .
43. A. M. Moschovi, S. Ntais, V. Dracopoulos and V. Nikolakis, Vib. Spectrosc., 2012, 63, 350–359 CrossRef .
44. M. El Hadri, A. Achahbar, J. El Khamkhami, B. Khelifa, V. Faivre, T. T. Cong, F. Bougrioua and S. Bresson, Vib. Spectrosc., 2013, 64, 78–88 CrossRef .
45. L. Cammarata, S. G. Kazarian, P. A. Salter and T. Welton, Phys. Chem. Chem. Phys., 2001, 3, 5192–5200 RSC .
46. G. M. Begun, G. W. Beall, L. A. Boatner and W. J. Gregor, J. Raman Spectrosc., 1981, 11, 273–278 CrossRef .
47. HORIBA.
48. S. Trohalaki and R. Pachter, QSAR Comb. Sci., 2005, 24, 485–490 Search PubMed .
49. A. Paul, P. K. Mandal and A. Samanta, Chem. Phys. Lett., 2005, 402, 375–379 CrossRef .
50. E. Guillet, D. Imbert, R. Scopelliti and J.-C. G. Bünzli, Chem. Mater., 2004, 16, 4063–4070 CrossRef .
51. M. Younas, Organic Spectroscopy and Chromatography, Ilmi Kitab Khana, Lahore, Pakistan, 2015 Search PubMed .
52. S. V. Bunzli and J.-C. G. Eliseeva, in Lanthanide Luminescence, ed. P. Hanninen, Springer, Berlin Heidelberg, 2011, pp. 1–45 Search PubMed .
53. K. C. Görller-Walrand, in Handbook on the Physics and Chemistry of Rare Earths, ed. J. A. Marinsky, L. E. Glendenin and K. A. Gschneidner, Elsevier B. V., 1998, pp. 101–264 Search PubMed .
54. R. A. Sá Ferreira, S. S. Nobre, C. M. Granadeiro, H. I. S. Nogueira, L. D. Carlos and O. L. Malta, J. Lumin., 2006, 121, 561–567 CrossRef .
55. K. Binnemans, Coord. Chem. Rev., 2015, 295, 1–45 CrossRef .
56. J. Y. R. Silva, L. L. Da Luz, F. G. M. Mauricio, I. B. Vasconcelos Alves, J. N. D. S. Ferro, E. Barreto, I. T. Weber, W. M. De Azevedo and S. A. Júnior, ACS Appl. Mater. Interfaces, 2017, 9, 16458–16465 Search PubMed .
57. H. J. Batista, A. V. M. de Andrade, R. L. Longo, A. M. Simas, G. F. de Sá, N. K. Ito and L. C. Thompson, Inorg. Chem., 1998, 37, 3542–3547 CrossRef PubMed .
58. O. L. Malta, S. J. L. Ribeiro, M. Faucher and P. Porcher, J. Phys. Chem. Solids, 1991, 52, 587–593 CrossRef .
59. K. Lunstroot, K. Driesen, P. Nockemann, K. Van Hecke, L. Van Meervelt, C. Görller-Walrand, K. Binnemans, S. Bellayer, L. Viau, J. Le Bideau and A. Vioux, Dalton Trans., 2009, 298–306 RSC .
60. H. Li, D. Li, Y. Wang and Q. Ru, Chem. – Asian J., 2011, 6, 1443–1449 CrossRef PubMed .
61. J. R. Diniz, J. R. Correa, D. D. A. Moreira, R. S. Fontenele, A. L. de Oliveira, P. V. Abdelnur, J. D. L. Dutra, R. O. Freire, M. O. Rodrigues and B. A. D. Neto, Inorg. Chem., 2013, 52, 10199–10205 CrossRef PubMed .
62. A. S. Borges, J. D. L. Dutra, R. O. Freire, R. T. Moura, J. G. Da Silva, O. L. Malta, M. H. Araujo and H. F. Brito, Inorg. Chem., 2012, 51, 12867–12878 CrossRef PubMed .
63. M. Rodrigues, N. Júnior, C. de Simone, A. Araújo, A. Brito-Silva, F. Paz, M. de Mesquita, S. Junior, R. Freire, S. a. Júnior and R. O. Freire, J. Phys. Chem. B, 2008, 112, 4204–4212 CrossRef PubMed .
64. L. L. da Luz, B. F. Lucena Viana, G. C. O. da Silva, C. C. Gatto, A. M. Fontes, M. Malta, I. T. Weber, M. O. Rodrigues and S. A. Júnior, CrystEngComm, 2014, 16, 6914–6918 RSC .
65. J.-C. G. Bünzli and C. Piguet, Chem. Soc. Rev., 2005, 34, 1048 RSC .
66. R. B. King, J. Chem. Educ., 1996, 73, 993 CrossRef .
67. O. L. Malta and F. R. G. Silva, Spectrochim. Acta, Part A, 1998, 54, 1593–1599 CrossRef .
68. M. E. de Mesquita, F. R. G. Silva, R. Q. Albuquerque, R. O. Freire, E. C. da Conceição, J. E. C. da Silva, N. B. C. Júnior and G. F. de Sá, J. Alloys Compd., 2004, 366, 124–131 CrossRef .

---

Footnote

† Electronic supplementary information (ESI) available: Details of nuclear magnetic resonance, FTIR/Raman spectroscopy, elemental/thermophysical analysis, UV-Vis and photoluminescence spectrophotometry. See DOI: 10.1039/c8tc00658j

This journal is © The Royal Society of Chemistry 2018