Evgeny D.
Boltkov
*a,
Mikhail E.
Buzoverov
b,
Elmira Kh.
Lermontova
b,
Yury A.
Belousov
ad,
Dmitry A.
Steshenko
a,
Victoria E.
Gontcharenko
cd and
Tatyana Yu.
Glazunova
a
aM.V. Lomonosov Moscow State University, Department of Chemistry, Moscow 119991, Russia. E-mail: boltkov@inorg.chem.msu.ru
bN.S. Kurnakov General and Inorganic Chemistry Institute, Russian Academy of Sciences, Leninskii Pr. 31, Moscow 119991, Russia
cFaculty of Chemistry, National Research University Higher School of Economics, 20 Miasnitskaya Str., Moscow 101000, Russia
dP. N. Lebedev Physical Institute of the Russian Academy of Sciences, 53 Leninskiy 1. Prospect, Moscow 119991, Russia
First published on 4th August 2025
This work presents a unique example of the self-assembly of {Ln6F8} clusters composed of Ln3+ cations and F− and CF3COO− anions. The crystal structures of new fluorotrifluoroacetate compounds with the general formula A2[Ln6(μ3-F)8(tfa)12L6], where A = Na and Ln = Pr(I), Eu (IIa), Tb(III), Dy(IV) and Tm(V); A = Na and H3O+ and Ln = Eu (IIb); or A = K and Ln = Yb(VI); tfa = CF3COO− (trifluoroacetate anion); L = Htfa (CF3COOH, trifluoroacetic acid) and H2O, determined by the single-crystal XRD method revealed that they were constructed from the complex anion [Ln6(μ3-F)8(tfa)12L6]2− and alkali metal (or alkali metal and hydroxonium) cations and included solvate molecules in some cases. The core of the anion presented a rhombododecahedral unit of six lanthanide atoms linked by bridging fluoride and trifluoroacetate anions. The nature of the counterion led to the formation of compounds with different dimensions: from 1D for potassium cations and 2D for sodium cations to 3D for sodium and hydroxonium cations. The {Ln6F8} clusters were obtained for a wide range of lanthanides, with Pr and Yb being the end members; thus, the existence of heterometallic compounds was presumed. At a temperature of 100 °C, the new compounds started to release solvent molecules, whereas the [Ln6(μ3-F)8(tfa)12] core was stable up to 270–300 °C according to the TG-MS data. Substances IIb, III and IV revealed metal-centered luminescence and could be used for the development of new optoelectronic materials.
Our research group has been intensively interested in fluorocarboxylates of various transition metals.4,5 The addition of fluoride ions to the system allows the assembly of small clusters stabilized by carboxylate anions. In this case, only the stoichiometry of the reagents determines the nuclearity and structure of the cluster as opposed to oxo- or hydroxo-clusters, for which hydrolysis processes play an important role, and as a consequence, the acidity of the reaction mixture and the amount of water also require control.
A literature review showed that most characterized lanthanide fluorocarboxylate compounds contain the {Ln6(μ3-F)8} core.6–10 The first work studying such compounds was published in 2020.9 A compound [Tb6(μ3-F)8(piv)10(Hpiv)4DMF]·xDMF·yH2O (Hpiv = pivalic acid, DMF = dimethylformamide) was characterized by SCXRD, and researchers were able to describe its structure. The molecular fragment [Tb6(μ3-F)8(piv)10(Hpiv)4DMF] was based on an octahedron of six terbium atoms. Eight fluorine atoms were placed over each {Ln3} face of the octahedron, and ten pivolate anions connected the terbium atom pairs. DMF and Hpiv molecules completed the coordination environment of the terbium atoms. The presence of F− ions was confirmed by XPS and MALDI-TOF methods. A study on the photophysical properties of [Tb6(μ3-F)8(piv)10(Hpiv)4DMF]·xDMF·yH2O revealed its extremely high quantum yield (99.6%) and luminescent lifetime (2.03 ms). These characteristics were better than those for similar hydroxoclusters and nanoscale TbF3. Analyzing the data, the authors conclude that the central {Tb6(μ3-F)8} core acts as an energy cage that minimizes the energy loss caused both by energy migration and multiphonon relaxation. Thus, the presented work described lanthanide fluorocarboxylates as promising substances that could be used to produce luminescent materials.
Independently, Morsbach et al.6,10 serendipitously obtained complexes [(NH4)0.75(H3O)0.25]2[Yb6(μ3-F)8(tfa)12(H2O)4]·4H2O,6 (NH4)2[Eu6(μ3-F)8(tfa)12(Htfa)6] and (NH4)2[Eu6(μ3-F)8(pfa)12(Hpfa)6]·8Hpfa10 (Hpfa = pentafluoropropionic acid) during attempts to synthesize the simple perfluorocarboxylates of Eu2+ and Yb2+ in liquid ammonia. The europium hexanuclear side-products exhibited longer luminescence decay times than related μ3-OH-capped compounds. Notably, the synthetic mixtures did not contain any fluoride sources, but the {Ln6(μ3-F)8} cores were assembled through slow oxidation processes. The possible cause of cluster formation may be activation of C–F bonds by binding of a fluoride atom to a lanthanide ion at an intermediate stage, as suggested by Vizuet et al.7 In that work, the question of the composition of the UiO-66 type lanthanide MOF (containing {Ln6(μ3-OH)8} cores) formed in the presence of a fluorine modulator was addressed for the first time. The reexamination was supported by SCXRD, XPS, IR, SEM/EDS and NMR experimental data and showed that the presence of a fluoride atom in the μ3 positions is more consistent with this data. Thus, the number of known compounds based on {Ln6(μ3-F)8} clusters could be significantly higher due to the inclusion of metal organic frameworks synthesized in the presence of 2-fluorobenzoic acid (2-fba) or 2,6-difluorobenzoic acid (2,6-fba) as modulators. Additionally, when the authors7 examined the use of an extra amount of modulator under the same conditions, leading to the formation of HoF3, the results also confirmed the C–F bond activation. To sum up, in terms of their photoluminescent properties, compounds containing the {Ln6(μ3-F)8} unit are more promising than their μ3-OH analogues, and their synthesis is relatively straightforward. Given that these compounds have not yet been studied in depth, there is significant motivation for work in this area.
Herein, we report the synthesis under mild conditions of a series of new isotype fluorotrifluoroacetate hexanuclear complexes of Pr(III), Eu(III), Tb(III), Dy(III), Tm(III) and Yb(III), with the general anion fragment [Ln6(μ3-F8)(tfa)12L6]2−, containing a unique {Ln6(μ3-F)8} core. Trifluoroacetate anions were chosen due to the absence of X–H bonds capable of acting as vibrational quenchers. The optimized conditions allowed large crystals of seven compounds containing alkali metal (Na and K) counterions to be obtained: Na2[Pr6(μ3-F8)(tfa)12(Htfa)2(H2O)4](Htfa)0.5 (I), Na2[Eu6(μ3-F8)(tfa)12(Htfa)0.88(H2O)5.12] (IIa), (H7O3)0.5Na1.5[Eu6(μ3-F8)(tfa)12(Htfa)0.78(H2O)5.22](Htfa)0.72(H2O)1.5 (IIb), Na2[Tb6(μ3-F8)(tfa)12(Htfa)0.88(H2O)5.12] (III), Na2[Dy6(μ3-F8)(tfa)12(Htfa)2(H2O)4](Htfa)0.5 (IV), Na2[Tm6(μ3-F8)(tfa)12(Htfa)2(H2O)4](Htfa)0.5 (V), and K2[Yb6(μ3-F8)(tfa)12(H2O)5](Htfa)2(H2O) (VI). The compounds were studied by SCXRD, PXRD, IR, and DSC-TG analyses. Their crystal structure is discussed in comparison with the above-mentioned analogues. The luminescent properties of compounds IIb, III and IV were also studied.
The synthesis was performed using the same procedure for all compounds. The hexahydrate of the lanthanide carbonate (Ln = Pr, Eu, Tb, Dy, Tm or Yb) was dissolved in 90% trifluoroacetic acid at room temperature, then a mixed solution of alkali metal fluoride and hydrofluoric acid was added. The molar ratio of the ions M+ (M = Na and K), Ln3+ and F− in the reagent was 1:
3
:
4 (Table S1).
The solutions obtained were placed in a desiccator containing phosphorus pentoxide at room temperature, leading to the formation of bulk crystals with linear lengths up to 0.3 cm, which were suitable for SCXRD. The reaction mixture for the europium sample gave two types of crystals with different shapes (IIa and IIb) that could be separated manually. After two weeks, the formed crystals were removed from the mother liquor and dried in an argon flow. Compounds I–VI were stable in air for long times.
PXRD data were collected using a Huber G670 Guinier Camera (CuKα1 radiation, Ge monochromator, λ = 1.5406 Å). Le Bail refinement was performed using Jana2020 software.11
Single crystals of compounds I–VI were subjected to X-ray single-crystal measurements with a Bruker diffractometer (AXS D8 VENTURE PHOTON III C14 IuS 3.0 and SMART-APEX-II) using graphite monochromatized MoKα radiation under a stream of cooled nitrogen. Data reduction was performed using the SAINT program.12 Absorption correction based on measurements of equivalent reflections was applied.13 Structures were solved by direct methods and refined by F2 by full-matrix least-squares in anisotropic approximation for non-hydrogen atoms using the SHELXTL-Plus software.14 For all structures, hydrogen atoms were placed in the calculated positions. In all the structures, the –CF3 groups are rotationally disordered and were refined with restrained C–F and F⋯F distances (SADI). In the structures of monoclinic compounds IIa and III, there is statistical disordering of the axial ligands. The occupancies of the H2O (PART 1) and Htfa (PART 2) molecules were refined using a free variable (oc(H2O) + oc(Htfa) = 1) and then fixed. The Htfa molecule was refined isotropically with restriction of the C–F, C–C and F⋯F distances (DFIX commands). The ADP of these fluorine atoms were unified using the EADP command. The same disordering was observed in IIb, but one more Htfa solvate molecule was observed in the H2O PART connected to water via a hydrogen bond. All the restrictions applied to the Htfa molecules were identical to those described above, and the disordered molecules were refined isotropically. Moreover, there were voids in this net-like structure, filled by H3O+ cations and H2O molecules. Because of the poor sensitivity of the SCXRD method to hydrogen atoms and their complex disordering, only oxygen atoms (five different PART positions) in the voids were added to the model. Potassium complex VI has another type of disordering. One of the bridging trifluoroacetate anions either occupies the standard position between two Yb atoms (major PART) or connects the Yb and K atoms (minor PART). Both PARTs were supplemented by water molecules on the atom that did not bind with trifluoroacetate anions (K and Yb atom in the major and minor PART, respectively). The occupancies of the PARTs were refined using a free variable and then fixed. The minor PART was refined isotropically. The Htfa molecule was refined isotropically with restriction of the C–F (DFIX commands), C–C and F⋯F distances (SADI commands). The ADP of these fluorine atoms were unified using the EADP command.
A summary of the crystallographic data and structure determination parameters is provided in Table S2. Crystal data have been deposited in the Crystallographic Data Centre as supplementary publications under the CCDC numbers 2391137–2391142 and 2427723.
Infrared spectra were recorded with an FTIR Spectrum One (PerkinElmer). The samples were pressed into KBr pellets and measured. The spectral resolution was 0.5 cm−1, from 4000 cm−1 to 400 cm−1.
The TG-DSC thermal analysis of samples combined with a simultaneous QMS evolved gas analysis was performed in a dynamic stream of argon (30 mL min−1) with an STA 409PC/PG (NETZSCH) at 25 °C to 500 °C with a 5 K min−1 heating rate.
Photoluminescence and excitation spectra in the visible region for all the complexes were measured in the solid state using a Horiba Jobin–Yvon Fluorolog QM-75-22-C spectrofluorimeter fitted with a 75 W xenon arc lamp (ArcTune). A Hamamatsu R13456 cooled photomultiplier tube that was sensitive in the UV-Vis-NIR region (200–950 nm) was used as the detector. Photoluminescent decays were recorded for all the complexes in the solid state using the same device, except the excitation source was changed to a pulsed xenon lamp with a 50 μs pulse duration and 100 Hz repetition rate.
Na2[Pr6(μ3-F8)(tfa)12(Htfa)2(H2O)4](Htfa)0.5 (I)
Na2[Eu6(μ3-F8)(tfa)12(Htfa)0.88(H2O)5.12] (IIa)
(H7O3)0.5Na1.5[Eu6(μ3-F8)(tfa)12(Htfa)0.78(H2O)5.22](Htfa)0.72(H2O)1.5 (IIb)
Na2[Tb6(μ3-F8)(tfa)12(Htfa)0.88(H2O)5.12] (III)
Na2[Dy6(μ3-F8)(tfa)12(Htfa)2(H2O)4](Htfa)0.5 (IV)
Na2[Tm6(μ3-F8)(tfa)12(Htfa)2(H2O)4](Htfa)0.5 (V)
K2[Yb6(μ3-F8)(tfa)12(H2O)5](Htfa)2(H2O) (VI)
The synthetic method used was simple and allowed to obtain large crystals of these compounds (Fig. 1). Both big crystal sizes and stability at ambient atmosphere made it possible to study the samples by IR-spectroscopy, TG-MS and luminescence measurements using parts of one single crystal. Nevertheless, phase purity of the obtained solid products was explored by the PXRD method (Fig. S1–S7, and Table S3), which was needed to characterize the synthetic procedure.
Crystalline compounds formed in solutions during the concentration process were not produced as single-phase samples in some cases. Phase analysis of the europium system shows the presence of two compounds with slightly different compositions and structures (the crystals could be separated manually based on their morphology). It was also found that the solid substance formed in the synthesis of I contained small amounts of impurities. We repeated the synthesis of Pr compound I several times and found that the resulting solids contained the same impurity. However, the single notable peak did not allow us to conduct a phase analysis to establish its nature; thus, some of these systems require further studies.
IR spectra of I–VI show the presence of absorption bands corresponding to vibrations of atom groups in trifluoroacetate anions, trifluoroacetic acid and water (Fig. S8, and Table S4). Band assignments were made based on data from published articles15–17 in which the IR spectra of trifluoroacetates were also discussed. No bands corresponding to v(Ln–O) oscillations were observed in the studied region; these probably absorb slightly more to the right than 400 cm−1. However, a shift of the absorption band corresponding to the planar vibration of the CCOO-group (that is directly related to the energy of the Ln–O bonds) was observed. The reduction of the lanthanide ionic radius in the series of compounds I–VI leads to increasing Ln–O bond energy and, correspondingly, to a shift in the δoop(CCO2) band to the higher wavenumbers.
Despite the diversity of the lanthanide carboxylates and their tendency to form oligo- and poly-nuclear coordination compounds, to our knowledge, there are only five reports of compounds featuring a {Ln6F8} core.6,7,9,10,18 A pivalate complex [Tb6(μ3-F)8(piv)10(Hpiv)4DMF]·xDMF·yH2O was the first reported compound of this type.9 Trifluoroacetate and pentafluoropropionate analogies with the same fragment [Ln6(μ3-F8)(tfa)12L6]2− were subsequently described.10 Meanwhile, the results reported in the present work indicate the existence of a family of compounds structurally related to fluorocarboxylatolanthanate for the wide row of lanthanides (at least from Pr to Yb).
The presence of the second phase IIb ((H7O3)0.5Na1.5[Eu6(μ3-F)8(tfa)12(Htfa)0.78(H2O)5.22](Htfa)0.72(H2O)1.5) in the system Na–Eu–tfa–F additionally indicates the stability of the fluorotrifluoroacetic fragment [Ln6(μ3-F)8(tfa)12L6]. We associate the formation of IIb with some non-stoichiometry of the europium carbonate hydrate, leading to a lack of sodium ions in the reaction mixture. In this case, the compound contains hydroxonium cations as compensation.
![]() | ||
Fig. 2 Structure of the anionic fragment [Ln6(μ3-F)8(tfa)12]. In the last picture, fluorine atoms of the CF3-groups are omitted for clarity. |
Axial positions (the ninth donor atom positions for Ln3+ ions) are occupied by molecules of water or trifluoroacetic acid. In the structures of compounds IIa, IIb, III, and VI, these molecules are statistically disordered. The hydrogen atoms of the trifluoroacetic acids in the axial positions form intramolecular hydrogen bonds with oxygen atoms of the bridging trifluoroacetate anions (Fig. 3).
![]() | ||
Fig. 3 Fragment of the anion complex in the crystal structure of IV. The intramolecular hydrogen bonds are shown in dashed lines, and the fluorine atoms of the CF3-groups are omitted for clarity. |
Axial ligands are absent at some lanthanide atoms in compound VI, which is based on ytterbium atoms with a smaller radius than the other lanthanides in compounds reported in this work. Consequently, a coordination number of 8 for the Yb ion was observed. The observation was also noted in recently published works6 on ytterbium fluorotrifluoroacetate complexes with a rhombododecahedral core. Another reason for the lower coordination number could be steric hindrance, as observed for similar pivalate complexes.9
Coordination polyhedra of the Ln atoms were determined using the continuous shape measures (CShM) approach implemented in SHAPE2.1 software19 (Table S6). According to the calculated data, the best description of the polyhedra for lanthanides with CN = 9 is a capped square antiprism (C4v), whereas for the ytterbium atoms in VI (CN = 8) without axial ligands, the most accurate description of a polyhedron is a square antiprism (D4d).
Reduction of the lanthanide ionic radii along the Pr–Yb row (structures I–VI) causes the shortening of the Ln–Ln, Ln–F, and Ln–O lengths and is consistent with theoretically calculated data20 (Table 1).
Pr(I) | Eu (IIa) | Eu (IIb) | Tb(III) | Dy(IV) | Tm(V) | Yb(VI) | |
---|---|---|---|---|---|---|---|
d(Ln⋯Ln) | 4.0738(5) | 3.9859(7) | 3.9804(9) | 3.9457(8) | 3.9232(6) | 3.8617(5) | 3.8537(4) |
d(Ln–F) | 2.428(3) | 2.378(4) | 2.376(4) | 2.352(6) | 2.342(3) | 2.309(3) | 2.304(2) |
d(Ln–O tfa ) | 2.474(4) | 2.405(6) | 2.400(5) | 2.375(8) | 2.359(4) | 2.318(3) | 2.307(5) |
d(Ln–F)theor. | 2.44 | 2.37 | 2.37 | 2.34 | 2.33 | 2.29 | 2.28 (CN 8) |
d(Ln–O)theor. | 2.49 | 2.42 | 2.42 | 2.39 | 2.38 | 2.34 | 2.33 (CN 8) |
Alkali metal cations (sodium or potassium) play the role of counterions for the described anion complexes. In structures I–VI, the cations are located between two neighboring anions and have six oxygen atoms in their environment (Fig. 4). For the larger potassium cation, the environment also includes one additional oxygen atom from a trifluoroacetic acid molecule.
![]() | ||
Fig. 4 Environment of the alkali metal ions: (a) sodium in compound IV and (b) potassium in compound VI. Fluorine atoms of the CF3-groups are omitted for clarity. |
It is worth noting that, in structures I–V, a few fluorine atoms of the CF3 groups are close to the sodium ions. As we highlight above, the reduction of the lanthanide radius leads to the shrinkage of the {Ln6(μ3-F)8} core. At the same time, the Ln–O distances increase (O atoms are from the trifluoroacetate anion connected to Na ions). This, in turn, makes sodium atoms more available to contact the fluorine atoms of the CF3 groups, so the Na⋯F distances become shorter in the series of compounds I–V (Fig. 5). Calculations (Fig. S17, Tables S8–S13) by the overlapping spheres method21 indicate that the coordination number of sodium ions in I–V is almost constant (excluding Na1 atom in structure of I) and is 10. At the same time, the proportion of the specific surface area of the Voronoi–Dirichlet polyhedra per six oxygen atoms in the sodium atom surrounding decreases for heavier lanthanides, which confirms the above trend (Fig. S17).
![]() | ||
Fig. 5 Binding sodium atom between two {Ln6F8} fragments in IIb (a) and III (b) structures (short Na–F contacts are shown in dashed lines). |
The sodium ions are linked to the anion complexes in the same way as realized in the fluorotrifluoroacetate of 3d-metals.22 Fluorotrifluoroacetate families containing 4f- and 3d-elements are similar, with the difference in their crystal structures determined by atom radii, resulting in differences in preferred coordination numbers. Based on the structure data for recently published fluorotrifluoroacetate complexes with 3d-metals,4,23 it was concluded that varying the cations in the lanthanide family of complexes also allows us to obtain substances with various packing of the anion complex and different structure dimensions.
In the next section, the structural features of I–VI, which have different axial ligand sets and packing, will be considered.
![]() | ||
Fig. 6 Crystal structure of the anion complex [Dy6(μ3-F8)(tfa)12(H2O)4(Htfa)2]2− in IV (a) and its bond with solvate Htfa molecule (b). Fluorine atoms of the CF3-groups are omitted for clarity. |
![]() | ||
Fig. 7 Solvate trifluoroacetic acid molecule between two neighboring layers (shown in different colors). Fluorine atoms of the CF3-groups are omitted for clarity. |
The triclinic compound IV contains two crystallographically independent sodium atoms (Na1 and Na2). Each anion complex is connected with four sodium atoms; thus, layers in the structure of IV are formed (Fig. 8). The layers are arranged parallel to the {bc} plane. Molecules of solvate acid are observed between the layers but do not connect them.
There are three crystallographically independent lanthanide atoms in the structure of IIa. These three lanthanide atoms, along with their symmetry equivalents, form one hexanuclear anion complex [Eu6(μ3-F8)(tfa)12(H2O)5.12(Htfa)0.88]2−. An axial position on Eu1 atom is occupied by an oxygen atom from statistically disordered molecules of trifluoroacetic acid, O13, or water, O13A (with occupancy 0.44 and 0.56, respectively, according to SCXRD refinement data), as shown in Fig. 9.
The water molecules’ oxygen atoms O22 and O21 complete the coordination environment of the Eu2 and Eu3 atoms, respectively. The O22 atom is also linked to the sodium atom Na1 and plays a bridging role.
In contrast to the triclinic phases described above, IIa has a single independent sodium atom. However, the same kinds of cation–anionic layers are observed, which are parallel to the {bc} plane, as in the triclinic compounds (Fig. 10).
The single independent sodium atom in the structure of IIb is linked to the anion complexes. Furthermore, anions formed by Eu4 atoms have six neighboring {Ln6(μ3-F)8} fragments, and those formed by Eu1–Eu3 atoms have only two neighboring fragments (Fig. 12). Thus, IIb exhibits a net-like organization.
The sodium cations in IIb cannot compensate for the anion charges completely, so the electroneutrality principle requires the presence of another counterion. The only such counterion may be the hydroxonium cation located in the cavities of the cation–anionic framework (Fig. 13). Apparently, there are two independent H3O+ ions connected to disordered water molecules (O27–O30). The localization of the hydrogen atoms in these H2O molecules looks impossible based on the available data.
Each anion complex is surrounded by four potassium atoms. The environment of the cation contains four oxygen atoms of bidentate trifluoroacetate anions, the disordered oxygen atoms O11B and O22A (see above), and the carbonyl oxygen atom O14 of the trifluoroacetic acid. The hydroxy group of this acid (O13A–H13A) forms a hydrogen bond with partially occupied atoms O11B and O22A (d(O⋯O) is 2.767(9) and 2.558(7) Å, respectively). The anion complexes are linked via potassium atoms into infinite chains placed along the crystallographic axis b (Fig. 15). Distances between neighboring chains (or layers in cases compounds I–V) correspond to van der Waals interactions between fluorine atoms of trifluoroacetate groups belonging to adjacent chains (layers).
Ln–F and Ln–O distances are almost the same for each lanthanide atom within each structural function. A full set of this distance data is given in Table S5; Table 2 contains the average distances for each lanthanide atom type.
I | IIa | IIb | III | IV | V | VI | |
---|---|---|---|---|---|---|---|
LnHtfa–F | 2.414(3) | 2.361(4) | 2.383(4) | 2.334(6) | 2.321(3) | 2.276(3) | 2.247(2) |
LnH2ONa/K–F | 2.416(3) | 2.370(4) | 2.368(4) | 2.337(6) | 2.335(3) | 2.307(3) | 2.326(2) |
LnH2O–F | 2.454(3) | 2.402(4) | 2.378(4) | 2.384(6) | 2.370(3) | 2.343(3) | 2.340(2) |
LnHtfa–Otfa | 2.488(4) | 2.414(6) | 2.390(5) | 2.386(8) | 2.363(4) | 2.320(3) | 2.291(5) |
LnH2ONa/K–Otfa | 2.476(4) | 2.408(6) | 2.409(5) | 2.379(8) | 2.364(3) | 2.321(3) | 2.307(3) |
LnH2O–Otfa | 2.459(4) | 2.394(6) | 2.402(5) | 2.359(8) | 2.351(4) | 2.314(3) | 2.322(7) |
LnHtfa–OHtfa | 2.750(4) | 2.879(19) | 2.662(19) | 3.02(3) | 2.974(5) | 3.252(5) | — |
LnH2ONa/K–OH2O | 2.671(4) | 2.623(11) | 2.633(6) | 2.70(2) | 2.609(4) | 2.613(3) | |
LnH2O–OH2O | 2.573(4) | 2.524(6) | 2.575(13) | 2.515(8) | 2.508(4) | 2.488(4) |
According to the data, Ln–F distances increase in the series LnHtfa–F, LnH2ONa–F, and LnH2O–F. A slight increase of the “LnHtfa”–Otfa distance (Otfa is an oxygen atom of the bridging trifluoroacetate anion) is caused by participation of this oxygen atom in intramolecular hydrogen bonds. In contrast to Ln–F distances, the bond length of Ln–O for the axial oxygen atom reduces in the series LnHtfa–F, LnH2ONa–F, and LnH2O–F.
The currently available structural data for the [Ln6(μ3-F8)(RCOO)12] fragment are summarized in Table 3. It lists the average Ln–Ln, Ln–F and Ln–O distances for the compounds obtained in this work, as well as for previously reported compounds.
As shown in Table 3, a lanthanide coordination number of nine is observed in most of the compounds. However, there are some examples in which the coordination is reduced to eight due to the small lanthanide radius (Yb compounds) and bulky sterically hindered carboxylate anions (piv or bdc). The presence of an available axial site in most cases can be used to place antenna ligands that can enhance the luminescence of such compounds.
The {Ln6(μ3-F)8} rhombododecahedral sizes vary over a wide range. This may indicate that such the units with can apparently be formed for the most lanthanides. This fact makes it possible to obtain heterometallic compounds based on this core.
Gas products of decomposition were determined to be in the m/z range 16–70 using a mass analyzer. As expected, the ion currents for particles that are characteristic for trifluoroacetic acid24 and water had considerable intensity (44 (CO2), 69 (CF3), 16 (O), 17 (OH), and 18 (H2O)). In the first stage, species with m/z 16, 17, 18 are observed for all compounds, which correspond to the removal of solvate water molecules. For compound IV, the removal of one solvate trifluoroacetic acid molecule was also observed. Simultaneously, weak peaks appear in the ion current curves at m/z 44 and 69. The assumed elimination of H2O and Htfa molecules was confirmed by weight loss calculations.
No distinct first stage of decomposition was observed for compound IIb, owing to the large amount of solvate water molecules linked together in the complex hydrogen bonding system. Its dehydration takes place smoothly in a temperature range of 100–200 °C.
The main mass losses are observed in the second stage of the decomposition over a narrow temperature interval. The heat release effect observed during this stage, according to the DSC curve, may indicate a mass removal effect. Such phenomena lead to slight differences between the calculated and experimental mass loss. An absence of pyrohydrolysis may be confirmed by the absence of peaks for the oxide/oxofluoride phases in the PXRD patterns of the decomposition products.
In contrast to the luminescent complexes of lanthanides with aromatic25 and heteroaromatic carboxylic acids1 and acylpyrazolones,26 the excitation spectra of complexes IIb, III and IV (Fig. S25) do not contain π–π bands of the ligand but contain pronounced electronic transitions of lanthanides. This is certainly because the intrinsic absorption of trifluoroacetate ligands is in the hard UV range (λ < 250 nm). Nevertheless, compounds IIb and III under UV irradiation exhibit intense red and green luminescence, respectively, visible to the naked eye. The yellow-green luminescence of the IV complex is much less bright.
The luminescence spectrum of Eu contains bands related to europium transitions 5D0 → 7Fj, j = 0–4.27 Interestingly, transitions from higher 5D1 and 5D2 levels do not appear, although they can often be observed for such rigid high-symmetry systems.27,28 The hypersensitive transition 5D0–7F2 is the most intense.
The splitting of the least intense 5D0 → 7F0 transition is well described by a single Gaussian curve, peaking at 17254 cm−1 (579.6 nm) (Fig. S26). This transition cannot split in any symmetry field, making it a marker for nonequivalent europium cations in the structure.28 The lack of splitting indicates a high symmetry in the coordination environment of Eu1–Eu4 atoms. Analysis using the SHAPE 2.1 program19,29 suggests that this environment is best represented as a one-capped square antiprism (C4v) or a Muffin (Cs), with the former being more accurate. For these symmetries, the 5D0 → 7F1 transition should split into 2 or 3 components,28,30 but the observed spectrum shows four components, three of which are dominant and one weak. This suggests that the “extra” components of the 5D0 → 7F1 transition are linked to nonequivalent europium cations, especially given the small energy differences between them (Fig. S27). The hypersensitive 5D0 → 7F2 transition is the most intense, with four components identified (Fig. S28), consistent with predictions from Judd–Ofelt theory for C4v symmetry.28
Integration of the spectrum components (Table S13) allows us to calculate the radiative τrad of europium ions, which is 4370 μs. Luminescence decay curves of the Eu complex at 77 K and at room temperature are described by the monoexponential law, . The values of τobs at 300 K and 77 K are (981.3 ± 0.4) μs (Fig. S29) and (1002.0 ± 0.4) μs (Fig. S30), respectively. Thus, the temperature dependence of τobs is small, which correlates well with the rigid crystal structure of IIb and the absence of vibrational quenching on OH, CH, or NH bonds, which is common for most europium27 complexes. The internal quantum yield,
, is 22.5 and 22.9% at 300 K and 77 K, respectively.
The luminescence spectra of III and IV (Fig. 16), as well as in the case of the europium complex considered above, do not contain any bands, except for the transitions of these ions from the most stable excited terms 5D4 (Tb) and 7F9/2 (Dy) to the sublevels of the main multiplet 7Fj, j = 6–0 (Tb) and 5Dj, j = 15/2–9/2 (Dy). The most intense transition of terbium ions (j = 5) has a maximum at 545 nm, which corresponds to bright green luminescence. It is interesting that transitions with j = 2,1 and 0, which usually do not appear in organic complexes, but are often found in solid-state phosphors,27 appear as low-intensity bands in the red region of the spectrum. The most intense Dy transition (j = 13/2) has a maximum at 575 nm, providing yellow-green luminescence of the Dy complex. The observed lifetimes of terbium and dysprosium complexes are (2563.4 ± 0.8) μs (Fig. S31) and (34.4 ± 0.4) μs (Fig. S32), respectively. In both the cases, one can speak of very high values of τobs, exceeding most of the carboxylate complexes, which is associated with the suppression of vibrational quenching in the rigid fluoroacetate framework.1
![]() | ||
Fig. 16 High-resolution emission spectra of IIb, III, and IV recorded at 77 K (λex = 395, 385 and 393 nm, respectively). |
Investigations on the luminescence properties of IIb, III and IV revealed the high lifetime of the excited state, along with weak dependence of the photophysical characteristics on temperature, which may be explained by the rigid nature of the central fluorinated fragment containing the lanthanide ions. The six close crystallographic metal positions in the core open great opportunities for the syntheses of bimetallic systems. For example this can allow to reach decreasing of the concentration quenching and to obtain systems with sensibilisation or up-conversion effects.
The family of compounds is a good base for creating new luminescent materials due to its thermal stability and wide range of rare earth elements which can form this structures.
Supplementary information (synthesis details, X-ray diffraction data, IR spectra, selected crystallographic data, decomposition details and luminescence spectra) is available. See DOI: https://doi.org/10.1039/d5dt01105a
CCDC 2391137–2391142 and 2427723 contain the supplementary crystallographic data for this paper.31a–g
This research was carried out with partial support from MSU Shared Research Equipment Center “Technologies for obtaining new nanostructured materials and their complex study”, National Project “Science” and MSU Program of Development.
This research was partially performed using the equipment of the JRC PMR IGIC RAS.
This journal is © The Royal Society of Chemistry 2025 |