**Post Synthetic Defect Engineering of UiO-66 Metal–Organic Framework with An Iridium(III)-HEDTA Complex and Application in Water Oxidation Catalysis**

#### **Giordano Gatto, Alceo Macchioni \* , Roberto Bondi, Fabio Marmottini and Ferdinando Costantino \***

Department of Chemistry, Biology and Biotechnology, Università di Perugia and CIRCC, Via Elce di Sotto, 8, I-06123 Perugia, Italy; giordano.gatto@libero.it (G.G.); robertobondi11@gmail.com (R.B.); fabio.marmottini@unipg.it (F.M.)

**\*** Correspondence: alceo.macchioni@unipg.it (A.M.); ferdinando.costantino@unipg.it (F.C.); Tel.: +39-075-585-5579 (A.M.); +39-075-585-5563 (F.C.)

Received: 9 August 2019; Accepted: 5 October 2019; Published: 10 October 2019

**Abstract:** Clean production of renewable fuels is a great challenge of our scientific community. Iridium complexes have demonstrated a superior catalytic activity in the water oxidation (WO) reaction, which is a crucial step in water splitting process. Herein, we have used a defective zirconium metal–organic framework (MOF) with UiO-66 structure as support of a highly active Ir complex based on EDTA with the formula [Ir(HEDTA)Cl]Na. The defects are induced by the partial substitution of terephthalic acid with smaller formate groups. Anchoring of the complex occurs through a post-synthetic exchange of formate anions, coordinated at the zirconium clusters of the MOF, with the free carboxylate group of the [Ir(HEDTA)Cl]<sup>−</sup> complex. The modified material was tested as a heterogeneous catalyst for the WO reaction by using cerium ammonium nitrate (CAN) as the sacrificial agent. Although turnover frequency (TOF) and turnover number (TON) values are comparable to those of other iridium heterogenized catalysts, the MOF exhibits iridium leaching not limited at the first catalytic run, as usually observed, suggesting a lack of stability of the hybrid system under strong oxidative conditions.

**Keywords:** metal–organic framework; post-synthetic modification; iridium catalysis; water oxidation; water splitting

#### **1. Introduction**

Water oxidation (WO) to molecular oxygen is considered the ideal reaction to provide electrons and protons for the generation of renewable fuels [1–3]. In addition to being thermodynamically disfavored, WO is also an extremely complicated, multi-electron and multi-proton reaction from the kinetic point of view, asking for an efficient and robust catalyst [4]. WOCs (water oxidation catalysts) based on iridium are among the most efficient reported in the literature so far, having, however, in the little abundance and, consequently, high cost of iridium their "Achilles' heel" [5–7]. A possible strategy to alleviate this problem stems in the minimization of the amount of noble-metal exploited in the catalytic process, according to the noble-metal atom economy principle [5]. This can be accomplished by utilizing (i) extremely active molecular catalysts at very low concentration [8–15], (ii) layered heterogeneous catalyst in which almost all active sites are reachable by the substrate [16], and (iii) heterogenized hybrid materials derived from the anchoring of a well-defined molecular catalyst on a suitable support [17,18]. The latter strategy should guarantee a very high percentage of active

sites (potentially 100%), increased robustness of catalyst, mainly due to the inhibition of associative deactivation processes, and possible beneficial cooperation between the anchored catalyst and support. Many hybrid heterogenized catalysts have been reported in the literature [17,19–23] and, among them, those using MOF as support, pioneered by Lin and co-workers [24,25], have been particularly successful [26]. MOFs are a class of porous crystalline compounds constituted by the ordered connection of metal clusters and organic linkers, forming accessible pores and channels potentially useful for a plethora of applications, especially in catalysis and energy production [27–29], as support for metal nanoparticles [30] and for enhanced gas sorption properties [31]. Some of them are rapidly approaching the industrial world [32]. Zr-MOFs are particularly interesting for their chemical and thermal stability and low production cost. The archetype structure is that of UiO-66 which is constituted of hexanuclear clusters with the formula Zr6O4(OH)4(BDC)<sup>6</sup> (BDC = 1,4-benzenedicarboxylic acid) in a cubic framework with face-centered cubic (fcu) topology [33]. UiO-66, together with other Zr-MOFs with different topologies, such as NU-1000, was already employed as support for Ir-based catalytic complex, employing both the Zr-cluster site [34,35] and via post-synthetic modification of the ligands for the in situ formation of the complex [36]. UiO-66 possess the exotic feature to be defective when it is crystallized in the presence of a mono carboxylic modulator such as formic, acetic, or benzoic acid, which act as substituent of BDC linker attached to the Zr<sup>6</sup> cluster thus inducing missing linker or missing cluster defects into the structure. These defects can be considered as an opportunity to be employed for imparting targeted functionality to the MOFs by means the so called post-synthetic defect exchange (PSDE) of the monocarboxylic groups with other carboxylic linkers [37,38].

Herein, we report on the synthesis of formic acid (FA)-modulated UiO-66 with a high concentration of defects and its use a support for anchoring, by means of PSDE, an Ir(III) WOC complex based on EDTA (EDTA = *N*,*N*,*N*'*N*'-ethylenediaminotetraacetic acid). The complex, of formula [Ir(HEDTA)Cl]Na (see Figure 1) was already reported in literature as an efficient and durable homogeneous WOC under chemical oxidation with CAN as a sacrificial agent [39]. The molecular structure of the complex clearly shows that it possesses a free carboxylic group which could be employed as an anchoring functionality for its deposition onto a solid surface. Our approach here consists in a PSDE of the FA-UiO-66 MOF with the [Ir(HEDTA)Cl]<sup>−</sup> complex dissolved in water. The anchoring occurred through a topotactic exchange of the coordinated FA with the carboxylic group of the IrCl-EDTA complex. The hybrid material (IrEDTA@UiO-66) was characterized by means of surface area and porosity studies, inductively coupled plasma-optical emission spetrometry (ICP-OES) analysis, nuclear magnetic resonance (NMR) spectroscopy, and tested for WO reaction by using CAN as the sacrificial agent. The hybrid exhibited WO activity with TOF and TON values comparable to those of the best performing materials. However, a significant Ir leaching was observed not only during the first catalytic run, as usually observed, suggesting that strong oxidative conditions with Ce4<sup>+</sup> lead to a rapid decomposition of the hybrid material. −

**Figure 1.** Molecular structure of [Ir(HEDTA)Cl]Na complex.

#### **2. Results and Discussion**

#### *2.1. Synthesis and Characterization*

#### Synthesis of FA-Modulated UiO-66

FA-modulated UiO-66 was prepared according to the procedure reported by Taddei et al. [38]. The use of a large amount of formic acid as modulator (100 eq. with respect to Zr) induced the formation of a highly defective phase with respect to the defective free UiO structure, which can be obtained following other synthetic strategies present in literature [40]. FA acts as monocarboxylic modulator with the Zr clusters inducing two types of defects: missing linker defects (Figure 2b) and missing cluster defects (Figure 2c).

**Figure 2.** Structure of non-defective (**a**), missing linker defective (**b**), and missing cluster defective (**c**) UIO-66 phase. Formic acid is evidenced in yellow. Zirconium clusters are depicted in blue.

θ It is known that the materials obtained with FA as modulator most likely possess missing cluster defects [38]. Nitrogen adsorption and desorption analysis at 77 K was performed on FA-UiO-66 compound after activation at 120 ◦C overnight. The N<sup>2</sup> adsorption/desorption isotherm is reported in Figure 3a and the Brunauer-Emmett-Teller (BET) value is 1450 m<sup>2</sup> /g with a total micropore volume of 0.57 cm<sup>3</sup> /g. These values, quite higher than the normal surface area and micropore volume of a defect-free UiO-66 (about 1100 m<sup>2</sup> /g and 0.4 cm<sup>3</sup> /g), suggest the highly defective nature of the obtained material. X-ray powder diffraction (XRPD) pattern of FA\_UiO-66 (Figure 3b) shows the peaks at 7.3◦ and 8.5◦ of 2θ belonging to the (111) and (200) of the **fcu** UiO-66 phase and a good crystallinity degree. <sup>1</sup>H-NMR spectra on the hydrolyzed compound confirm the presence of a considerable amount of FA, as can be seen in Figure 3d. Integration of <sup>1</sup>H-NMR signals belonging to FA (8.3 ppm) and BDC (7.8 ppm) gives a FA/BDC ratio equal to 0.63. The obtained solution after the hydrolysis of the sample with NaOH was analyzed with ion chromatography resulting in the following BDC and FA contents in the starting solid: BDC = 2.78 mmol/g and FA = 1.72 mmol/g.

2

6െ ൌ 0.62

 ൌ

**Figure 3.** N<sup>2</sup> adsorption and desorption isotherm for FA\_UiO-66 (black line) and IrEDTA@UiO-66 (red line) (**a**). XRPD patterns of FA\_UiO-66 (black), Ir-EDTA@UiO-66 (red), and calculated pattern for UiO-66 (blue) (**b**). TGA curve for FA\_UiO-66 (black) and IrEDTA@UiO-66 (red) (**c**) and <sup>1</sup>H-NMR spectrum for hydrolyzed FA\_UiO-66 MOF (NaOD/D2O, 298 K) (**d**).

Given these results, the ratio FA/BDC = 0.63 is in very good agreement with the results of NMR experiments. Since FA is a monocarboxylic acid, the following equation can be used in order to determine the formula of the defective MOF:

$$\frac{FA}{BDC} = \frac{2x}{6-x} = 0.62$$

resulting in Zr6O4(OH)4(BDC)4.58(FA)2.74. Thermogravimetric analysis (Figure 3c) shows three different weight losses at 100 ◦C (7.5%), 330 ◦C (11%), and 540 ◦C (38%) due to the loss of water molecules and decomposition of the organic part of the MOF. If the plateau in the 550–1200 ◦C temperature range is assumed to be 6ZrO<sup>2</sup> (*Mw* = 123 g/mol), we can use this value as a reference (100%) for extrapolating the theoretical formula from the analysis. The normalized weight at 100 ◦C is therefore 213%. The experimental formula weight from TGA analysis at 100 ◦C is, therefore, 1572 g/mol. Given the mass of the defective, desolvated MOF of formula Zr6O4(OH)4(BDC)4.58(FA)2.74 = 1555 g/mol, this is in good agreement with the experimental data from TGA analysis. The PSDE process for anchoring the Ir-EDTA complex onto the cluster surface is shown in Figure S3. After soaking the evacuated MOF into a water solution containing the dissolved complex (0.02 M) and heating at 80 ◦C for 24 h, the partial exchange of FA with the free carboxylic group of the complex occurred. Three samples with different amounts of exchanged Ir-EDTA were prepared: The compound was exchanged with 0.096, 0.077, and 0.057 mmol of Ir-EDTA, respectively. ICP-OES analysis for the determination of Ir content gave the following results: IrEDTA@UiO-66(1) = 256 µmol/g; IrEDTA@UiO-66(2) = 226 µmol/g; and IrEDTA@UiO-66(3) =170 µmol/g. Figure 3a shows the nitrogen adsorption and desorption analysis

at 77 K performed in the same conditions on IrEDTA@UiO-66 containing 256 µmol/g (red curve). After the exchange the calculated BET value is reduced to 547 m<sup>2</sup> /g and the total micropore volume is reduced to 0.22 cm<sup>3</sup> /g suggesting that the complex is not simply linked to the particle surface but most likely occupies the micropores created by the defects. However, the Ir-EDTA complex is inserted in the micropores and the complex could obstruct a part of the micropore volume of the substrate. TGA analysis of IrEDTA@UiO-66 (Figure 3b, red curve) is similar to that of the pristine MOF although the decomposition of the material starts at lower temperature (around 350 ◦C) with respect to the unmodified MOF.

The XRPD patterns of the three samples are shown in Figure 4. Anchoring Ir-EDTA onto the cluster surface did not affect the structure of the MOF since the characteristic peaks remained unaltered.

**Figure 4.** XRPD pattern of IrEDTA@UiO66(1) black, (2) red and (3) blue.

Figure 5 shows the <sup>1</sup>H-NMR spectrum of the hydrolyzed IrEDTA@UiO-66 sample. The peak at 8.3 ppm attributed to FA exhibits a reduced intensity and the integration with that of BDC gave as result FA/BDC = 0.10. This value is about six times lower than the unmodified defective MOF (FA/BDC = 0.63) meaning that the most part of FA was successfully exchanged by Ir-EDTA complex. Peaks belonging to the Ir-EDTA complex are clearly visible at 3 and 2.2 ppm. With this new ratio, a suggested formula can be Zr6O4(OH)4(BDC)4.58(FA)0.6(Ir-EDTA)2.2.

**Figure 5.** <sup>1</sup>H-NMR spectrum of hydrolyzed Ir@UiO-66(1) sample [NaOD/D2O, 298 K; \* denote impurities present in the solvent, likely acetone (ca. 2 ppm) and Silicon Grease (slightly lower than 0 ppm)].

#### *2.2. Water Oxidation Catalytic Activity of IrEDTA@UiO-66*

Herein, the catalytic activity of IrEDTA@UiO-66 hybrid materials toward water oxidation to molecular oxygen (Equation (1)) is described. First, a blank experiment by using only the FA\_UiO-66 without Ir was performed by adding a 25 mM solution of CAN to 3 mg of MOF (see Figures S1 and S2). No oxygen evolution was observed confirming the inactivity of MOF toward water oxidation. Catalytic tests with Ir containing MOF were carried out by using Ce4<sup>+</sup> (added as CAN) as a sacrificial oxidant, dispersing the proper amount of catalyst in acidic water (pH 1, 0.1 M HNO3) at 25 ◦C.

$$4\text{Ce}^{4+} + 2\text{H}\_2\text{O} \rightarrow 4\text{Ce}^{3+} + 4\text{H}^+ + \text{O}\_2 \tag{1}$$

The evolved gas, according to Equation (1), was quantified by differential manometry (See Materials and Methods). In a first series of experiments, a consecutive triple addition (100, 150, and 500 µL) of a 1.25 M solution of CAN to 4.9 mL of a 51.5 µM IrEDTA@UiO-66 suspension was executed (Table 1, entries 1–3; Figure 6). IrEDTA@UiO-66 was found to be a competent catalyst for water oxidation and exhibited a TOF of ca. 5 min−<sup>1</sup> and TON values included between 62 and 308 with yields = 30%–50%. A second series of measurements was performed with the aim of evaluating possible leaching of the molecular catalyst from the MOF support. Particularly, a catalytic run was executed by using 73.12 µM IrEDTA@UiO-66 and 75 mM CAN (Table 1, entry 4). At the end of O<sup>2</sup> evolution IrEDTA@UiO-66 was recovered by filtration and the supernatant solution tested by the addition of another aliquot of 75 mM CAN (Table 1, entry 5). Moreover, the recovered solid was tested under the same conditions (Table 1, entry 6). At the end of the reaction the solid catalyst was again recovered by filtration and the second supernatant tested (Table 1, entry 7). The measured TOF (4 min−<sup>1</sup> ) and TON (108, yield = 42%) values of the starting IrEDTA@UiO-66 are nicely consistent with those observed in the first series of experiments. Furthermore, the recovered solid IrEDTA@UiO-66 exhibits similar TOF (6 min−<sup>1</sup> ) and TON (180, yield = 67%) values. Nevertheless, the two supernatants are active, with even higher TOF (10 and 13 min−<sup>1</sup> ) but comparable TON (363, yield = 44% and 1013, yield = 46%) values, evidencing some leaching of iridium in solution. ICP-OES measurements indicate that 30.98% and 30.75% of iridium leached out from IrEDTA@UiO-66 after the first and second catalytic run, respectively. In order to check the stability of the MOF before catalysis we evaluated the Ir leaching by dispersing IrEDTA@UiO-66 in a 0.1 M HNO<sup>3</sup> solution for 2 h, without the addition of CAN. The measured Ir leaching was about 35%, which is similar to that observed in the first catalytic run. It means that the grafted complex is scarcely stable upon acidic conditions. The catalytic activity of IrEDTA@UiO-66 compares well with those of the molecular precursor [34] and hybrid material IrEDTA@TiO<sup>2</sup> [20], tested under similar conditions, in terms of TOF (Table 1, entries 8–10 and 12). The TON values are clearly lower than those observed for the molecular precursor, which provide 100% yield, and somewhat smaller also than those of IrEDTA@TiO<sup>2</sup> (Table 1, entries 8–10 and 12). Nevertheless, the main criticality of IrEDTA@UiO-66 seems to be the leaching of iridium, occurring also after the second catalytic run, contrary to what observed for IrEDTA@TiO<sup>2</sup> (Table 1, entries 11 and 13) and other heterogenized iridium catalysts reported before [17,20]. Several explanations might be provided for such a phenomenon. It can be hypothesized some Ce4<sup>+</sup> might undergo an exchange with the Zr4<sup>+</sup> ions of MOF, becoming not available anymore for driving the oxidative splitting of water. Alternatively, it might be hypothesized that the oxidative potential of iridium inIrEDTA@UiO-66 is slightly higher than in the molecular precursor and hybrid material IrEDTA@TiO2, thus asking for a higher Ce4+/Ce3<sup>+</sup> ratio in order to reach the appropriate "Nernstian" potential for WO [41,42]. Both the explanations are consistent with the observation that the addition of a second aliquot of CAN restores the catalytic activity.


**Table 1.** Summary of the water oxidation (WO) catalytic data for Equation (1). "Sur" indicates supernatant. −

**Figure 6.** [O<sup>2</sup> ] (**bottom**) and d[O<sup>2</sup> ]/dt (**up**) versus time trends for a WO triple cerium ammonium nitrate (CAN) addition experiment (Table 1, entries 1–3).

A catalytic run with a large amount of IrEDTA@UiO-66 (50 mg, 2.61 mM; CAN = 75 mM) was performed in order to recover and analyze IrEDTA@UiO-66 post-catalysis. The <sup>1</sup>H-NMR spectrum of the recovered solid digested in NaOD is significantly different than that before catalysis (Figure 7). In particular, the typical resonances of the –CH<sup>2</sup> protons of EDTA in the 2.0–3.2 ppm range are not visible anymore in the post-catalysis sample, suggesting a complete degradation of the ligand framework [43]. XRPD pattern of the MOF after three catalytic runs (Figure S4) shows no crystallinity loss. The FA/BDC ratio (Figure S5) post catalysis is 0.20 suggesting that the framework remained most likely unaltered and the degradation involved a small fraction of BDC together with the Ir-EDTA complex. However, because the recovered solid is still active in WO, it might be hypothesized that after EDTA degradation some iridium remains attached at the MOF structure, possibly through the formation of Zr–O–Ir oxo bridges, as observed in heterogenized WOCs prepared by anchoring an Ir-Kläui molecular precursor onto BiVO<sup>4</sup> nanopyramids [17].

**Figure 7.** <sup>1</sup>H-NMR spectra (NaOD/D2O, 298K) before (**bottom**) and after (**up**) a catalytic run, showing the disappearance of the aliphatic resonance of the EDTA ligand at 2.2–3.2 ppm.

#### **3. Materials and Methods**

#### *3.1. Synthetic Procedures*

All reagents were used as received without further purification: ZrCl4, cerium ammonium nitrate (CAN), formic acid (FA), terephthalic acid (BDC) and *N*,*N*-dimethylformamide (DMF) was purchased from Sigma Aldrich (St. Louis, MO, USA). [Ir(HEDTA)Cl]Na was prepared according to Reference [34].

#### 3.1.1. Synthesis of FA-UiO-66

θ

ZrCl<sup>4</sup> (0.60 g, 2.5 mmol) was dissolved in DMF (40 mL). Then, water (0.135 mL, 7.5 mmol), FA (9.4 mL, 250 mmol), and BDC (0.435 g, 2.5 mmol) were added to the solution. The mixture was sonicated until complete dissolution and divided in four vials (10 mL each) and heated in an oven at 120 ◦C for 16 h. After the reaction, the solid was recovered for centrifugation and washed with DMF (one time after 2 h soaking), water (2 h soaking), and acetone (one time after 10 min soaking). At the end, the solid was dried in an oven at 80 ◦C for 2 h.

#### 3.1.2. Synthesis of IrEDTA@UiO-66 via PSDE

FA\_UiO66 (60 mg) was suspended in 5 mL of a 0.02 M water solution of a [Ir(HEDTA)Cl]Na (0.02M) for 24 h at 80 ◦C. After completion of the reaction, the solid was centrifuged and washed with DMF (one time, two-hour soaking), water (two times, two-hour soaking), and acetone (two times, two-hour soaking). The solid was dried in an oven at 80 ◦C for two hours. Two other syntheses with

α

different Ir contents were carried out: 30 mg of UiO-66 in 0.01M Ir-EDTA solution (5 mL) and 40 g in 0.015 Ir-EDTA solution (5 mL).

#### *3.2. Analytical and Instrumental Procedures*

*Powder X-Ray Di*ff*raction (PXRD).* PXRD patterns were collected in reflection geometry in the 4–40◦ 2θ range, with a 40 s per step counting time and with a step size of 0.016◦ on a PANalytical X'PERT PRO diffractometer (Malvern Panalytical Ltd., Malvern, UK), PW3050 goniometer, (Malvern Panalytical Ltd., Malvern, UK) equipped with an X'Celerator detector (Malvern Panalytical Ltd., Malvern, UK) by using the Cu-Kα radiation. The long fine focus (LFF) ceramic tube operated at 40 kV and 40 mA.

*Thermogravimetric analysis (TGA).* TGA was performed using a Netzsch STA490C thermoanalyzer (NETZSCH Group, Selb, Germany) under a 20 mL min−<sup>1</sup> air flux with a heating rate of 10 ◦C min−<sup>1</sup> .

*Nitrogen adsorption and desorption isotherms.* N<sup>2</sup> adsorption/desorption isotherms were performed using a Micromeritics ASAP 2010 analyzer (Micromeritics, Norcross, GA, USA). Prior of the analysis, the samples were degassed overnight under vacuum at 120 ◦C. BET analysis and t-plot analysis of the adsorption data were used to calculate specific surface area and micropore volume respectively. The Harksin and Jura equation was used as reference for the statistical thickness calculation.

*Ion-Chromatography Analysis.* Ion chromatography was carried out using a Dionex 500 (Dionex Corp., Sunnyvale, CA, USA) apparatus with a CD20 suppressed conductivity module. Sample analysis was performed as follow: About 30 mg of sample was dispersed in 40 mL of NaOH 0.0125 M and refluxed for 2 h. After reflux, the solution was diluted to 100 mL by water. The resulting solution was analyzed by ion chromatography using a Dionex AS11 column and eluted with a flux of 1.5 mL/min with NaOH 6 mM in the case of BDC analysis or NaOH 0.1 mM in the case of FA analysis.

*ICP-OES Analysis*. The ICP-OES analysis was carried out using a Varian 700-ES series (Agilent Technologies, Santa Clara, CA, USA) with a standard (2,5,7, and 10 mg/L respectively) of Iridium solution.

*WO catalytic experiments*. Catalytic experiments were performed using two homemade jacketed glass reactors coupled to a Testo 521-1 manometer. In a typical catalytic run, IrEDTA@UiO-66 suspended in a 0.1 M HNO<sup>3</sup> solution was loaded into the first reaction vessel (working cell), whereas an equal amount of neat water was loaded into the second one (reference cell). Both reactors were sealed with a rubber septum, connected to the manometer, kept at a constant temperature of 25 ◦C, and placed under stirring for 20 min. Acquisition was started. When a steady baseline was achieved, an equal volume of a solution of CAN and neat water were injected into the working cell and reference cell, respectively, to reach a final volume of 5 mL in each reactor. The concentration of the stock solution of CAN was adjusted, depending on the final concentration desired, in order to have a maximum injection volume of 500 µL. The total gas evolved was estimated by measuring the differential pressure between the working and reference cell.

*Fitting methodology and kinetic data analyses*. All trends of [O2] evolution versus time were fitted by a composite mathematical function developed by Peters and Baskin (PB) for distinguishing sigmoidal and bilinear growth profiles of plant roots [38]. The derivative of the PB fits provided reaction rate (v = d[O2]/dt) trends as function of time. Reaction rate over catalyst concentration led to TOF (= v/[Ir]), which was plotted versus the factor conversion X (= 4[O2]/[CAN]0) [9].

#### **4. Conclusions**

In this paper a catalytic active Ir complex based on EDTA was successfully anchored onto a defective Zr-MOF with UiO-66 structure. The post-synthetic modification of defective MOF for designing a new heterogenous catalyst was here validated for the first time demonstrating that substitution of small formate anions linked to zirconium clusters with a larger carboxylate-bearing complex is possible. The material was employed for water oxidation reaction using Ce4<sup>+</sup> as the sacrificial agent. The catalyst showed a good catalytic activity, which is comparable to that observed for already reported iridium-supported Zr-MOF [25] and slightly lower than Ir-EDTA@TiO2 heterogenized

catalysts [20]. However, Ir leaching occurs not only during the first catalytic run, as usually observed, but also for the successive ones. Moreover, leaching of Ir was also observed simply dispersing the solid in the nitric acid solution, without CAN addition. This fact suggests that the material is not stable under the acidic and strong oxidative conditions due to the high redox potential of Ce4+. Furthermore, the WO reaction yield is somewhat lower than that observed for other heterogenized iridium WOCs, indicating a possible exchange of the zirconium atom of MOF with cerium of CAN or a higher "Nernstian" potential. Despite those drawbacks, the results reported in this paper suggest that anchoring a molecular WOC onto a defective MOF is a viable strategy to assemble a hybrid material to be integrated into a device for the generation of renewable fuels. Future developments of this work will be devoted to the stability improvement of the system by performing photo- or electro-catalysis which avoid the use of Ce4<sup>+</sup> and strong acidic conditions.

**Supplementary Materials:** The following are available online at http://www.mdpi.com/2304-6740/7/10/123/s1, Figure S1: Manometric oxygen evolution of IrEDTA@UiO-66 and UiO-66. Figure S2: Differential manometric oxygen evolution of IrEDTA@UiO-66 and UiO-66. Figure S3: PSDE of FA with IrEDTA complex onto the structure of FA\_UiO-66. Figure S4: XRPD patterns of IrEDTA@UiO-66(3) before and after three catalytic runs. Figure S5: <sup>1</sup>H-NMR spectrum of IrEDTA@UiO-66 after 3 catalytic runs.

**Author Contributions:** Conceptualization, F.C. and A.M.; methodology, F.C., F.M., A.M.; formal analysis, G.G., R.B. and F.M.; data curation, G.G., F.M. and F.C.; writing—original draft preparation, F.C. and A.M.; writing—review and editing, F.C. and A.M.

**Funding:** This research received no external funding.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


© 2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (http://creativecommons.org/licenses/by/4.0/).

## *Article* **Photoluminescent Coordination Polymers Based on Group 12 Metals and 1***H***-Indazole-6-Carboxylic Acid**

**Antonio A. García-Valdivia <sup>1</sup> , Estitxu Echenique-Errandonea <sup>2</sup> , Gloria B. Ramírez-Rodríguez <sup>1</sup> , José M. Delgado-López <sup>1</sup> , Belén Fernández <sup>3</sup> , Sara Rojas <sup>1</sup> , Javier Cepeda 2,\* and Antonio Rodríguez-Diéguez 1,\***


**Abstract:** Two new coordination polymers (CPs) based on Zn(II) and Cd(II) and 1*H-*indazole-6 carboxylic acid (H2L) of general formulae [Zn(L)(H2O)]<sup>n</sup> (**1**) and [Cd<sup>2</sup> (HL)<sup>4</sup> ]n (**2**) have been synthesized and fully characterized by elemental analyses, Fourier transformed infrared spectroscopy and single crystal X-ray diffraction. The results indicate that compound **1** possesses double chains in its structure whereas **2** exhibits a 3D network. The intermolecular interactions, including hydrogen bonds, C–H···π and π···π stacking interactions, stabilize both crystal structures. Photoluminescence (PL) properties have shown that compounds **1** and **2** present similar emission spectra compared to the free-ligand. The emission spectra are also studied from the theoretical point of view by means of time-dependent density-functional theory (TD-DFT) calculations to confirm that ligand-centred *π-π\** electronic transitions govern emission of compound **1** and **2**. Finally, the PL properties are also studied in aqueous solution to explore the stability and emission capacity of the compounds.

**Keywords:** group 12 metals; 1*H*-indazole-6-carboxylic acid; coordination polymer; photoluminescence properties

#### **1. Introduction**

The study of coordination polymers (CPs) and metal-organic frameworks (MOFs) is at the forefront of modern inorganic chemistry due to their broad range of potential applications, spanning from magnetism and luminescence, through catalysis and sensing, to gas separation and storage, and biomedicine [1,2]. Through an adequate selection of their building blocks (metal ions and organic ligands), CPs and MOFs can be designed to enhance a particular property [3–5]. It is well known that nitrogen-containing heterocycles are molecules commonly employed as ligands owing to not only their good coordination ability, but also pharmacological relevance, given that they are important scaffolds widely present in numerous commercially available drugs [6]. The most famous are diazepam, isoniazid, chlorpromazine, metronidazole, barbituric acid, captopril, chloroquinine, azidothymidine and anti-pyrine. As a result of their diverse biological activities, nitrogen heterocyclic compounds have always been attractive targets to develop new active compounds. This is the case for 1*H-*indazole-6-carboxylic acid (H2L), a common moiety in the pharmaceutical industry [7]. Polysubstituted indazole-containing compounds furnished with different functional groups usually present significant pharmacological activities and serve as structural motifs in drug molecules (i.e., niraparib-anticancer drug, pazopanibapproved by the FDA for renal cell carcinoma, bendazac and benzydamine-antiinflamatory

**Citation:** García-Valdivia, A.A.; Echenique-Errandonea, E.; Ramírez-Rodríguez, G.B.; Delgado-López, J.M.; Fernández, B.; Rojas, S.; Cepeda, J.; Rodríguez-Diéguez, A. Photoluminescent Coordination Polymers Based on Group 12 Metals and 1*H*-Indazole-6-Carboxylic Acid. *Inorganics* **2021**, *9*, 20. https:// doi.org/10.3390/inorganics9030020

Academic Editor: Andrea Rossin

Received: 16 January 2021 Accepted: 17 March 2021 Published: 22 March 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

drugs) [8]. From the structural point of view, indazole is an aromatic heterocyclic molecule with a benzene ring fused to a pyrazole ring [9]. It shows three tautomeric forms (Scheme 1) being tautomer **A** favoured over **B** and **C** due to its higher degree of aromaticity [10]. ‐

‐ ‐

‐

‐ ‐

‐

‐

‐

‐

‐

‐

‐

‐

‐

‐ ‐

**Scheme 1.** Indazole tautomerism (**A**, **B**, and **C** from left to right).

‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ H2L is presented here as an ideal candidate to form CPs or MOFs as it possesses multiple coordination possibilities, not only derived from its carboxylate group, but also from its pyrazole function. Thus, a great variety of coordination modes are possible, according to similar ligands containing carboxylate and pyrazole chemical functions in crystallized complexes (Scheme 2). Until now, only one complex based on this ligand has been reported so far [11]. In that work, Kruger et al. described in detail four substituted indazole derivatives containing pyridine or carboxylic functionalities upon coordination with Cu(II) ions in solution and solid state. In the complex, 1*H*-indazole-6-carboxylate acts as a bridging ligand showing a tridentate coordination mode: the carboxylate group coordinates to two Cu(II) atoms in a *syn,syn* mode to establish a dimeric paddle-wheel shaped entity, whereas the non-protonated nitrogen atom of the pyrazole ring links to a third Cu(II) atom in a monodentate way (see the highlighted modes in Scheme 2). Aside from this work mainly focused on the description of a new compound, it should be pointed out that some Co(II)-based complexes with indazole derivatives have shown a capacity to bind to DNA [12]. ‐ ‐ ‐ ‐ ‐ ‐

‐

‐ ‐

*‐* ‐ ‐ *‐* ‐ ‐ **Scheme 2.** Possible coordination modes of *1H-*indazole-6-caboxylate ligand. Note that only those two modes highlighted in black have been described in bibliography whereas the rest correspond to potential binding modes.

‐

‐ ‐

‐

‐

‐

On the other hand, H2L may also present interesting photoluminescence (PL) properties due to its aromatic nature and the presence of carboxylic groups, with potentially strong light absorption [13]. When these indazole-carboxylate ligands are coordinated to metal centres in the crystal structure of a CP, PL tends to be enhanced by means of the well-known crystal-induced luminescence effect [14]. Among others, metal ions from group 12 are particularly appropriate for their use in PL as they present a closed-shell electronic configuration in which d-d transitions cannot occur [15,16]. In fact, many CPs and MOFs formed by these metal ions have been reported during the last decade [17,18], some of which present not only strong and bright fluorescent emissions, but also long-lived phosphorescence that may be traced by the naked eye [19–21]. Moreover, the presence of these ions may also promote ligand-to-metal charge transfer (LMCT) as metal ions possess empty orbitals that can be populated in the excited state, and therefore the PL emission may be modulated with regard to the ligand-centred (LC) emissions [22,23]. Irrespective of the luminescence mechanism occurring in these systems, the interest for group 12-based compounds has increased given their potential application as not only lighting devices, but also as luminescence-based molecular detectors [24], thermometers [25] and anti-counterfeiting inks, among others [26]. Particularly for indazole derivatives playing as ligands, many Zn-/Cd-indazole complexes have already proved efficient luminescent CPs under UV irradiation [27]. ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐ ‐

‐

‐

‐

‐

‐

‐ ‐

‐ ‐ ‐

Considering all the above, in this work we present the synthesis, structural characterisation and PL properties of two new coordination polymers based on group 12 metals and 1*H-*indazole-6-carboxylic acid of general formula [Zn(L)(H2O)]<sup>n</sup> (**1**) and [Cd2(HL)4]<sup>n</sup> (**2**). Their emission characteristics have been studied both from the theoretical and experimental points of view, involving the measurements in the solid state as well as in aqueous medium. *‐* ‐ ‐

#### **2. Results and Discussion**

The reaction of 1*H-*indazole-6-carboxylic acid gave rise to two compounds based on group 12 metals which exhibit a different structural dimensionality. In particular, the solvothermal reaction of the 1*H-*indazole-6-carboxylic acid ligand with zinc acetate salt (Zn(CH3COO)2) using a 1:2 molar ratio in a *N*,*N*-dimethylformamide/water (DMF/H2O) mixture afforded a 1D CP, namely **1** (see Experimental Section for further details). Similarly, the use of cadmium acetate salt (Cd(CH3COO)2) salt in the synthesis, successfully led to a 3D MOF, namely **2**. This fact can be explained by the larger ion size of Cd(II), which may admit higher coordination numbers, involving the participation of additional ligands and increasing the metal-to-ligand connectivity. *‐* ‐ ‐ *‐* ‐ ‐ ‐ ‐ ‐

#### *2.1. Description of the Structures*

#### 2.1.1. Structural Description of [Zn(L)(H2O)]<sup>n</sup> (**1**)

Compound **1** crystallizes in the *P*21/*n* space group and consists of a double chain structure in which Zn(II) ions are bridged by nitrogen atoms of L2<sup>−</sup> in a bidentate way, giving rise to a stable and in plane Zn2N<sup>4</sup> dimeric core as a six membered ring (Figure 1). −

**Figure 1.** Representation of the 1D polymeric chain in which Zn2N<sup>4</sup> planar six membered ring is observed (zinc, nitrogen, oxygen, and carbon are represented in green, blue, red, and grey, respectively; hydrogen atoms are omitted for clarity).

The Zn(II) ion is also coordinated to a carboxylate moiety of the indazole derivative ligand in a monodentate way, which extends the dimeric entity into infinite 1D chains running along the crystallographic [100] direction. The coordination sphere of Zn is completed by the coordination of a water molecule (see Table S1 in the ESI for further information about bond lengths and angles). The ZnN2O<sup>2</sup> coordination sphere can be described as a tetrahedron, although Zn ions show a geometry close to an axially vacant trigonal bipyramid according to continuous-shape-measures (CShMs) using SHAPE software (Tables S2 and S3, in the ESI) [28].

The packing of the double chains is ruled by intermolecular interactions, among which hydrogen bonding interactions established between coordination water molecules and carboxylate oxygen atoms are to be highlighted (Figure 2). In particular, the coordinated water molecule is involved in hydrogen bonding interactions in which non-coordinated carboxylate oxygen atoms belonging to adjacent chains act as receptors. Additionally, the angle formed among neighbouring chains allows for the formation of C–H···π interactions between aromatic rings, reinforcing the stability of the supramolecular crystal building (see Figure S5 in the ESI). ‐ ‐ ∙∙∙π

‐

‐

‐

‐ **Figure 2.** Perspective view of the chains of [Zn(L)(H2O)]<sup>n</sup> packed in the framework (hydrogen atoms have been omitted for clarity).

#### 2.1.2. Structural Description of [Cd2(HL)4]<sup>n</sup> (**2**)

‐ ‐ ‐ ‐ *‐* ‐ ‐ ‐ ‐ ‐ Compound **2** crystallizes in the triclinic *P*-1 space group. The asymmetric unit contains two non-equivalent Cd(II) atoms and four ligand molecules. Each Cd(II) ion is connected to two monodentate indazole nitrogen atoms and four oxygen atoms of the carboxylate group of the ligand. Cd1 and Cd2 ions are doubly linked by ancillary *syn-anti* carboxylate moieties of 1-*H*-indazole-6-carboxylate ligands (namely A, C and D). However, the carboxylate group of B ligand presents a different coordination pattern, in which O1B connects in a monodentate way to both Cd1 and Cd2 atoms giving rise to alternating five and six membered rings (Figure 3, see also the view along *b* axis in Figure S6 in the ESI), whereas O2B atom remains unconnected to any metal centre.

‐ ‐ **Figure 3.** View of the coordination of Cd(II) ions to HL in compound **2** (cadmium, nitrogen, oxygen, and carbon are represented in dark-yellow, blue, red, and grey, respectively; hydrogen atoms have been omitted for clarity).

‐ ∙∙∙ ‐ ∙∙∙ ‐ ‐ CShMs indicate that different ligand coordination modes affect the connectivity of the metal centres, which leads to the formation of a distinct crystal structure. When comparing **1** and **2** compounds, the coordination spheres of Cd1 and Cd2 are described as octahedra according to SHAPE measurements (Tables S2 and S3 in the SI). M···N2 distances are slightly shorter than in compound **1** (in the 1.969 and 2.293–2.316 Å ranges, respectively), similarly to the M···O1carboxylate bond distances (between 1.935 and 2.320 Å, see Table S1 in the ESI). As a result, a 3D framework is obtained in the case of **2** by the further linkage of the carboxylate groups to Cd(II) atoms along a metal-carboxylate rod (Figure 4). Considering the connectivity of the metal ions and HL ligands, this framework may be described as a **5,6T24** topological network with the (3<sup>2</sup> .42 .52 .63 .7)2(3<sup>2</sup> .44 .54 .62 .73 ) point symbol, as previously observed in the MOF of [Al2(OH)2(H2O)2(C10O8H2)] or MIL-118A [29].

‐ **Figure 4.** Perspective of 3D structure of compound **2** (hydrogen atoms have been omitted for clarity).

− −

−

‐ ‐ ‐ Ʋ

‐ ‐

‐

‐

‐ *π π*

‐

‐ ‐

‐

−

*‐* ‐ ‐

−

−

*‐* ‐ ‐

Ʋ

−

Ʋ

−

−

− −

−

To end up with the structural description, it is worth mentioning that compound **2** presents some remarkable supramolecular interactions that reinforce its packing. Unlike with hydrogen bonding and C–H··· π interactions governing the crystal structure of **1**, **2** contains π···π stacking interactions. In particular, the aromatic rings of HL promote strong face-to-face contacts among the whole structure (see Figure S5 in the ESI).

#### *2.2. Fourier Transformed Infrared (FTIR) Spectroscopy*

The analysis of the FTIR spectra of **1** and **2** confirms the coordination of zinc(II) and cadmium(II) ions to the *N*-containing carboxylate ligand (Figure S1, see the ESI). FTIR spectra of both compounds confirmed a shift in the wavelengths in comparison to the pure linker, suggesting the formation of interactions between the linker and the metals. The main vibrations of 1*H*-indazole-6-caboxylic acid associated with the he ƲC=N stretching − − − Ʋ − − − stretching vibration at 1633 cm−<sup>1</sup> , and the asymmetric and symmetric vibrations of the carboxylate groups at 1683 and 1423 cm−<sup>1</sup> are shifted when compared with the spectra of **1** and **2**. The bands found at 1537 and 1589 cm−<sup>1</sup> for complex **1** and **2**, respectively, are attributed to the he ƲC=N stretching stretching vibration of the indazole ring [30]. Moreover, the strong absorption peak observed at 1558 and 1402 cm−<sup>1</sup> for **1**, and 1541 and 1411 cm−<sup>1</sup> for **2**, respectively, revealed the asymmetric and symmetric vibrations of the carboxylic groups [31]. Finally, the strong broad band in the range of 3317–3086 cm−<sup>1</sup> was assigned to the O–H stretching vibration of the coordinated water molecule in complex **1** [32].

#### *2.3. Luminescence Properties*

*π π π π* As previously mentioned, complexes consisting of metal ions with d<sup>10</sup> electronic configuration are known to yield strong PL emissions. The completely filled d-orbitals disable ligand field d-d transitions, eliminating fluorescence quenching and allowing the occurrence of PL [14]. Thus, the development of d10-based compounds is interesting for photochemical, electroluminescence and sensing applications [17,33]. The extended aromaticity of the 1*H-*indazole-6-carboxylate ligand coordinated to Zn(II) and Cd(II) atoms suggests the existence of emissive properties of **1** and **2**. The emission of these compounds are found to be similar to ligand emission, which may stem from the ligand-centred *π–π\** electronic transitions, as shown in Figure 5. Consequently, it can be suggested that the highly conjugated 1*H-*indazole-6-caboxylate ligand is the main part contributing to the emission [34]. An intense broad band at 350–450 nm dominates the emission spectra of all compounds upon 325 nm excitation (in view of the maxima found in the excitation spectra), among which the maxima at 362 and 388 nm, 363 and 381 nm, and 363 and 391 nm can be distinguished for the ligand, and compounds **1** and **2**, respectively; which imbues all compounds with blue emission. The similar emission band of **2** and the free ligand must be attributed to the fact that **2** possesses the protonated form of the ligand (HL−) whereas it is completely deprotonated (L2−) in **1**. In a comparative scale, the ligand spectrum shows two well-defined maxima (not that easily identified for the compounds) and relatively higher intensity (Figure S2, SI). It is worth noticing that the observed luminescence resembles to that shown by other previously reported CPs containing other isomers of indazolecarboxylates [35,36].

−

‐

‐

−

‐

‐

‐

‐ ‐ ‐ λ ‐ ‐ **Figure 5.** Room temperature Time-dependent density-functional theory (TD-DFT) computed (dashed lines) and experimental (solid line) photoluminescence emission under λex = 325 nm of compounds **1** (blue-up) and **2** (green-down). The insets show the most representative molecular orbitals involved in the electronic transitions.

‐ ‐ ‐ ‐ π In order to get a deeper insight into the emission mechanism, TD-DFT calculations were performed on suitable models of compounds **1** and **2**. The calculated spectra reproduce fairly well the experimental ones, indicating that the process is driven by singlet transitions occurring between the molecular orbitals shown in Figure 5. In both cases, the electron density in HOMO orbitals, HOMO-2 and HOMO-4 for compound **1** and **2**, respectively, extend over the bonds over the whole ligand molecule (signifying a π orbital) whereas LUMO orbitals, LUMO+2 and LUMO+1 for compound **1** and **2**, respectively, feature a π\* character. Therefore, it can be stated that the transitions involved in the photoluminescence of compound **1** and **2** are mainly of π\*←π nature induced by ligand centred emission.

Inspired by the potential biomedical properties of the ligand on the basis of its similar structure to other indazole derivatives [7,8], we studied the stability and fluorescence performance of these compounds in aqueous media in order to explore their performance as luminescent probes. First, the stability of both compounds was confirmed by recording UV-Vis absorption spectra on aqueous solutions of both compounds immediately after their solution and also after 24 h (see Figure S7 in the ESI). These spectra show the presence of three main absorption bands (sited at ca. 215, 265 and in the 340 nm for **1** and 220, 265 and 305 nm for **2**), corresponding to intraligand and/or ligand-to-metal charge transfers occurring in the complexes. It is worth noticing that these bands are in good agreement with the experimental and TD-DFT computed excitation bands, finding only slight shifts that may be attributed to the different media in which the spectra are recorded (water for UV-Vis and solid state for excitation spectra). Moreover, these solutions were also employed to measure the PL emission spectra of both compounds. As observed in Figure S4, the emission spectra acquired in this medium do not significantly differ from those measured in solid state but for a drop in the intensity of the signal, which is an expected behaviour given that the capacity of this solvent to quench the PL is a largely reported effect [37]. All these results suggest that these new compounds could show potential activity as luminescent probes in some particular biological media (i.e., as biosensor), a fact that *a priori* excludes most of biological tissues owing to their low transparency (high light absorption capacity) to the blue emission (λem = 350–450 nm) shown by these compounds.

#### **3. Materials and Methods**

#### *3.1. Materials and Physical Measurements*

All the reagents were purchased commercially and used without any previous purification. Elemental analysis (C, H and N) were carried out at the Centro de Instrumentación Científica (University of Granada) on a Fisons-Carlo Erba analyzer model EA 1108 (Thermo Scientific, Waltham, MA, USA). FTIR spectra (400–4000 cm−<sup>1</sup> ) were recorded on a Nicolet FT-IR 6700 spectrometer (Thermo Scientific, Madrid, Spain) in KBr pellets.

#### *3.2. Synthesis of [Zn(L)(H2O)]<sup>n</sup> (***1***)*

0.010 g (0.006 mmol) of 1*H-*indazole-6-carboxylic acid (H2L) was dissolved in 0.5 mL of DMF. Then, 0.5 mL of water was added to the ligand solution. In a separate vial, 0.0134 g (0.03 mmol) of Zn(CH3COO)<sup>2</sup> was dissolved in 0.5 mL of water. Similarly, once metal salt was dissolved, 0.5 mL DMF were added to the solution. Metal solution was added dropwise to the ligand solution, and the resulting colourless mixture was placed in a closed glass vessel and heated in an oven at 100 ◦C for 24 h. X-ray quality crystals of **1** were obtained during heating process under autogenous pressure and washed with water. Yield: 64% based on Zn. Anal Calcd. for C8H4N2O3Zn: C, 39.46; H, 2.48; N, 11.50. Found: C, 39.39; H, 2,41; N, 11.59. In addition to the elemental analyses, the purity of all the samples was checked by FT-IR spectra.

#### *3.3. Synthesis of [Cd2(HL)4]<sup>n</sup> (***2***)*

The same synthetic procedure was carried out to obtain complex **2**, by replacing Zn(CH3COO)<sup>2</sup> by 0.01651 g (0.03 mmol) of Cd(CH3COO)2. X-ray quality crystals were obtained and washed with water. Yield: 54% based on Cd. Anal Calcd. for C32H20Cd2N8O8: C, 44.21; H, 2.32; N, 12.89. Found: C, 44.16; H, 2.29; N, 12.91. In addition to the elemental analyses, the purity of all the samples was checked by FT-IR spectra.

#### *3.4. Crystallographic Refinement and Structure Solution*

Single crystals of suitable dimensions were used for data collection. For compound **1** and **2**, diffraction intensities were recorded on a Bruker X8 APEX II and Bruker D8 Venture with a Photon detector (Bruker, Madrid, Spain) equipped with graphite monochromated MoKα radiation (λ = 0.71073 Å). The data reduction was performed with the APEX2 software [38] and corrected for absorption using SADABS [39]. In all cases, the structures were solved by direct methods and refined by full-matrix least-squares with SHELXL-2018 [40]. The main refinement parameters are listed in Table 1. Details of selected bond

lengths and angles are given in Table S2 in the ESI. CCDC reference numbers for the structures are 1,948,382 and 1,948,383 for Cd and Zn coordination polymers, respectively.


**Table 1.** Crystallographic data and structure refinement details of compounds **1** and **2**.

<sup>a</sup> *R<sup>1</sup>* = S||Fo| − |Fc||/S|Fo|. <sup>b</sup> Values in parentheses for reflections with *I* > 2s(*I*). <sup>c</sup> *wR<sup>2</sup>* = {S[*w(F<sup>o</sup> <sup>2</sup>* − *F<sup>c</sup> 2 ) 2* ] / S[*w(F<sup>o</sup> 2 ) 2* ]} 1 <sup>2</sup> ; where w = 1/[σ 2 (F<sup>0</sup> 2 ) + (aP)<sup>2</sup> + bP] and P = (max(F<sup>0</sup> 2 ,0) + 2Fc<sup>2</sup> )/3 with a = 0.0.0319 (**1**), 0.0380 (**2**) and b = 6.9969 (**1**).

#### *3.5. Photophysical Measurements*

UV-Vis absorption spectra were recorded on UV-2600 UV/vis Shimadzu spectrophotometer using polycrystalline samples of compounds **1** and **2**. PL measurements were carried out on crystalline samples at room temperature using a Varian Cary-Eclipse fluorescence spectrofluorometer equipped with a Xe discharge lamp (peak power equivalent to 75 kW), Czerny–Turner monochromators, and an R-928 photomultiplier tube. For the fluorescence measurements, the photomultiplier detector voltage was fixed at 600 V, and the excitation and emission slits were set at 5 and 2.5 nm, respectively. Phosphorescence spectra were recorded with a total decay time of 20 ms, delay time of 0.2 ms and gate time of 5.0 ms. The photomultiplier detector voltage was set at 800 V, and both excitation and emission slits were open to 10 nm.

#### **4. Conclusions**

The reaction between 1*H*-indazole-6-carboxylic acid ligand and Zn(II) or Cd(II) leads the formation of two new coordination polymers with different dimensionalities. Compound **1** possesses a double chain structure, whereas compound **2** exhibits a 3D structure. Emissive properties of both complexes have been studied demonstrating that their photoluminescent emission is driven by the ligand centred π\*←π transition. The similar luminescent properties between compound **2** and the linker may be consequence of the partially protonated HL<sup>−</sup> ligand present in **2**. This work is pioneer in studying and comparing the luminescent properties of 1*H-*indazole-6-carboxylic acid (H2L) and its complexes, which represent a common moiety in pharmaceutical industry. In this regard, novel materials based on this ligand are being developed in our laboratory using lanthanide ions to enhance their luminescent properties.

**Supplementary Materials:** The following are available online at https://www.mdpi.com/2304-674 0/9/3/20/s1, Figure S1: Infrared spectra of the ligand and compounds **1** and **2**, Figure S2: Emission spectra of the ligand and compounds **1** and **2** under λex = 325 nm, Figure S3: Excitation spectra of compounds monitored at the emission maxima: (a) λem = 381 nm for **1** and (b) λem = 391 nm for **2**, **Figure S4**: Comparative view of the absorption spectra of compounds (a) **1** and (b) **2** in solid state and aqueous solution, Figure S5: The most representative intermolecular interactions and packing modes for complexes **1** (up) and **2** (down). H bonds, π···π and C-H···π interactions are shown with dashed blue, green and orange lines, respectively, Figure S6: View along a (left), b (middle) and c (right) axis of complex **1** (up) and **2** (down), Figure S7: UV-Vis spectra of compounds (a) **1** and (b) **2** in aqueous solutions acquired at times (0 h and after 24 h), Table S1: Selected bond lengths (Å) and angles (◦ ) for complexes **1** and **2**, Table S2: Continuous Shape Measurements for the ZnN2O<sup>2</sup> coordination environment, Table S3: Continuous Shape Measurements for the CdN2O<sup>4</sup> coordination environment.

**Author Contributions:** Conceptualization, A.R.-D.; methodology, A.A.G.-V. and E.E.-E.; software, J.C.; validation, J.M.D.-L. and S.R.; formal analysis, B.F. and S.R.; investigation, A.A.G.-V. and G.B.R.- R.; resources, B.F. and A.R.-D.; data curation, G.B.R.-R. and J.C.; writing—original draft preparation, E.E.-E.; writing—review and editing, J.C. and A.R.-D.; visualization, B.F.; supervision, A.R.-D.; project administration, J.C. and A.R.-D.; funding acquisition J.C. and A.R.-D. All authors have read and agreed to the published version of the manuscript.

**Funding:** Financial support was given by Junta de Andalucía (Spain) (FQM-394), University of the Basque Country (GIU 17/13), Gobierno Vasco/Eusko Jaurlaritza (IT1005-16), and the Spanish Ministry of Science, Innovation and Universities (MCIU/AEI/FEDER, UE) (PGC2018-102052-A-C22, PGC2018-102052-B-C21). J.M.D.L. and G.B.R.R. acknowledge the FEDER/MCIU/AEI for their Ramón y Cajal (RYC-2016-21042) and Juan de la Cierva (JdC-2017) fellowships, respectively. S.R. acknowledges the Juan de la Cierva Incorporación Fellowship (grant agreement n<sup>o</sup> . IJC2019-038894-I). E.E.-E. is grateful to the Government of the Basque Country for the predoctoral fellowship. The authors thank for technical and human support provided by SGIker of UPV/EHU and European funding (ERDF and ESF).

**Institutional Review Board Statement:** Not applicable.

**Informed Consent Statement:** Not applicable.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**


*Article*
