Metal-organic frameworks (MOFs) represent a large and rapidly growing class of compounds which have remained a focus of attention of chemists around the world in recent decades [1,2,3]. The interest in MOFs is driven by the fact that they can be utilized in numerous application areas, such as catalysis [4,5,6], the deep separation of diverse organic substrates, including hydrocarbons, [7,8,9,10] the design of selective sensors [11,12,13,14,15], etc. It is very important that the selectivity of these processes is governed predominantly by the nature of the organic linker ligands connecting the metal centers, since those are mostly responsible for the system of non-covalent interactions with the substrates.

Usually, the main role is played by hydrogen bonds (this is expected due to the greater number of C-H fragments in most organic ligands). However, other non-covalent interactions can participate in the binding of substrates as well. Among them, there is the halogen bond (XB), a type of supramolecular contact which has been intensively investigated within the last decade [16,17,18,19,20,21,22]. In some the recent works, it was shown that the XB can play a decisive role in the appearance of certain properties of MOFs [13,23,24]. For this reason, we believe that this area has a great potential for further development.

Within our current research, we focused on the use of iodine-substituted aromatic polycarboxylic acids as linkers for MOFs. This class of compounds has several important advantages. Earlier, it was shown [25] that they can indeed form rather strong halogen bonds. On the other hand, they can be regarded as very suitable linkers—there are hundreds, if not thousands, of MOFs based on derivatives of terephthalic, isophthalic, biphenyldicarboxylic and other acids.

The results of our work with 2-iodoterephthalic acid (2-I-bdc) revealed that MOFs based thereupon feature sorption selectivity strikingly different from one demonstrated by MOFs based on non-substituted bdc [26]. In further experiments, we decided to focus on a more iodine-rich bdc derivative, 2,5-diiodoterephthalic acid (2,5-I-bdc), and on lanthanides as metal centers in order to examine the luminescent properties of the resulting MOFs. Here, we report five novel Ln³⁺-based MOFs—{[La₂(2,5-I-bdc)(DMF)₄}]·2DMF (1) and {[Ln₂(2,5-I-bdc)(DMF)₄}] (Ln = La (2), Nd (3), Sm (4) and Eu (5))—and discuss their structures and luminescent properties.

All reagents were obtained from commercial sources and used as purchased. H₂(2,5-I-bdc) was prepared according to the method described in the literature [27].

For the preparation of 1–5, we used the solvothermal approach, which is widely applied in MOF chemistry [31,32,33,34,35,36]. The crystals of 1 and 2 were isolated from the same sample; as follows from the PXRD data (see Supporting Information), it contains 1, 2 and some other unidentified phase. The data of the element analysis agree well with the composition corresponding to 1 and 2, so it can be assumed that the by-product has a similar composition. MOFs 3–5 were isolated as single phases.

It must be noted that all crystals in the 1–5 series were initially isolated in experiments where different N-linkers (such as 1,2-bis(4-pyridyl)ethane, 1,2-bis(4-pyridyl)ethylene and 4,4′-bipyridine) were added to the mixture. We aimed for the preparation of heteroleptic MOFs but this strategy did not work, clearly due to higher oxophilicity of Ln.

The main difference between 1 and isostructural series 2–5 is the absence/presence of solvate DMF molecules. In the crystal structure of 1, binuclear [La₂(C₆H₂I₂(COO)₂)₆(DMF)₄] building blocks (Figure 1; main bond distances are summarized in Table S2 of the supporting Information) combine into layers via bridging with dicarboxylate ligands. Such layers are oriented in the [110] crystal direction and stack together in an AAA… motif along the [001] crystal direction. Solvate DMF molecules fill the space between the layers. All [La₂(C₆H₂I₂(COO)₂)₆(DMF)₄] dimeric units are symmetrically equal. Carboxylate ligands of the dimer connect each building block with its four neighbors (Figure 2 and Figure 3).

In the crystal structure of 2, the binuclear geometry of [La₂(2,5-I-bdc)₆(DMF)₄] binuclear units is significantly different due to the absence of solvate DMF molecules. A comparison of the geometry of such dimers in the crystal structures of 1 and 2 is presented in Figure 4.

The reorientation of dicarboxylic ligands (Figure 5) leads to a change in the topology of the coordination polymer from 2D to 3D. In the crystal structure of 2, each dimer is connected with six neighboring ones (Figure 6). The crystal packing of 2 (Figure 7) demonstrates the presence of 1D channels in the [001] crystal direction. These channels are decorated by I-atoms potentially accessible for non-covalent interactions with different substrates.

Luminescence of Ln³⁺ carboxylate complexes has been widely studied in recent years [37,38,39,40,41,42], so we decided to study these features for the 1/2 mixture as well as the pure 3, 4 and 5 phases. The excitation and emission spectra of the complexes (λ_ex = 340 nm) are shown in Figure 8 (see also Figure 9 and Figure 10 for details). The emission spectra of the solid containing 1 and 2 exhibits only the band at 445 nm. Because La³⁺ is a non-luminescent ion, these spectra show the optical properties of the ligand luminescence [43]. The measured lifetime of the luminescence is τ = 5 ns. The blue emission has the colorimetric coordinates (0.150, 0.128).

The photoluminescence spectra of the (3) complex have pronounced peaks corresponding to ⁴F^3/2 → ⁴I_9/2 (880 nm) and ⁴F^3/2 → ⁴I_11/2 (1060 nm) transitions in Nd³⁺. The most intensive emission band observed at 1060 nm corresponds to the ⁴F^3/2 → ⁴I_11/2 transition. The excitation spectra obtained by monitoring the luminescence at λ = 1060 nm showed some narrow excitation bands (λ = 500–850 nm region) and a large band (λ < 400 nm), which were attributed to the absorption of the Nd³⁺ central ion and ligand moieties, respectively. The luminescence spectrum of compound (4) in the visible region exhibited the characteristic emission bands for Sm³⁺ (⁴G_5/2 → ⁶H_5/2, ⁴G_5/2 → ⁶H_7/2, ⁴G_5/2 → ⁶H_9/2 and ⁴G_5/2 → ⁶H_11/2). The most intense emission line corresponded to the hypertensive ⁴G_5/2 → ⁶H_7/2 (601 nm) transition and a ligand-centered emission at 445 nm induces a purple light emission with the colorimetric coordinates (0.238, 0.175). The luminescent lifetime for complex (4) is 50 μs and the quantum yield is < 1%. The excitation spectrum of (4) has a broad band in the region of 290–370 nm, which is ascribed to the π–π* electronic transition of the ligand, and narrow bands in the region of 400–500 nm, which are attributed to the intraconfigurational f−f transitions of the Sm³⁺ ion.

For compound (5), the following transitions are observed: ⁵D₀ → ⁷F₁, ⁵D₀ → ⁷F₂, ⁵D₀ → ⁷F₃, and ⁵D₀ → ⁷F₄. The intensity of the emission of the forced electric dipole transition of Eu^{3+ 5}D₀ → ⁷F₂ (615 nm) of compound (5) dominated the spectrum. The rest of the emission bands were small compared with the latter. It has been previously reported that these characteristics are indicative of low-symmetry Eu³⁺ coordination compounds [44]. Additionally, the emission peak centered at 615 nm is the strongest one, which results in a red emission with the colorimetric coordinates (0.627, 0.342). The luminescent lifetime for complex (5) is 1.1 ms and the quantum yield is 4.5%. The excitation spectrum of (3) contains a broad band ranging from 290 to 360 nm, which is ascribed to the π–π* electronic transition of the ligand. The narrow bands in the region of 360–600 nm are attributed to the intraconfigurational f−f transitions of the Eu³⁺ ion.

TGA data for 4 and 5 are given in SI (Figure S5 and Figure S6, respectively). Those are almost identical: there is a clear stage of DMF elimination (≈16% of total mass) occurring within the T range up to ≈120 °C.

To conclude, we demonstrated that 2,5-diiodoterephthalate is an efficient linker for the design of lanthanide-based MOFs. However, in the absence of auxiliary O-containing linkers, additional coordination sites are occupied by DMF molecules so that they can form either 2D or 3D polymers. Most likely, this series can be expanded to other Ln³⁺ -based MOFs, potentially resulting in complexes with greater emission properties. Corresponding experiments are underway.