The Solvent Effect on Composition and Dimensionality of Mercury(II) Complexes with Picolinic Acid

Three new mercury(II) coordination compounds, {[HgCl(pic)]}n (1), [HgCl(pic)(picH)] (2), and [HgBr(pic)(picH)] (3) (picH = pyridine-2-carboxylic acid, picolinic acid) were prepared by reactions of the corresponding mercury(II) halides and picolinic acid in an aqueous (1) or alcohol–methanol or ethanol (2 and 3) solutions. Two different types of coordination compounds were obtained depending on the solvent used. The crystal structures were determined by the single-crystal X-ray structural analysis. Compound 1 is a one-dimensional (1-D) coordination polymer with mercury(II) ions bridged by chelating and bridging N,O,O′-picolinate ions. Each mercury(II) ion is four-coordinated with a bidentate picolinate ion, a carboxylate O atom from the symmetry-related picolinate ion and with a chloride ion; the resulting coordination environment can be described as a highly distorted tetrahedron. Compounds 2 and 3 are isostructural mononuclear coordination compounds, each mercury(II) ion being coordinated with the respective halide ion, N,O-bidentate picolinate ion, and N,O-bidentate picolinic acid in a highly distorted square-pyramidal coordination environment. Compounds 1–3 were characterized by IR spectroscopy, PXRD, and thermal methods (TGA/DSC) in the solid state and by 1H and 13C NMR spectroscopy in the DMSO solution.


Introduction
The transition metal coordination compounds have been of great interest over the past decades because of their interesting and diverse structural features and promising applications in different fields of industry. Such compounds can be used in various areas as materials that exhibit desired catalytic, ion-exchange, gas storage, pharmaceutical, luminescence, optical, electronic, or magnetic properties [1][2][3][4][5][6][7][8][9]. Electron-withdrawing metal ions, such as mercury(II), can change and tune the fluorescent properties of the organic ligands used in a design of coordination compounds [8][9][10][11][12][13]. Flexible coordination geometry of the mercury(II) ion, varying from tetrahedral to octahedral, enables the formation of different architectures with different properties [2,11,12,[14][15][16][17][18][19]. By changing additional parameters such as the nature of the ligands, type of the counter ions, selection of solvents, metalto-ligand ratios, temperature, pH etc., it is possible to impact the crystal structures and tune the desired properties of the coordination compounds [20][21][22][23][24]. Deeper understanding of the above-mentioned parameters is crucial for better prediction of the properties and structures of the solid-state materials. Approaches to crystal growth and design, together with the study and understanding of intermolecular interactions, represent one of the main challenges in the field of crystal engineering [25][26][27][28][29].
Pyridine-2-carboxylic acid (picolinic acid, picH) is a very popular and widely used chelating ligand that is able to coordinate to metal ions in an N,O-bidentate fashion via pyridine N and carboxylate O atoms. The coordination potential of picolinate ion (pic − ) can be used for designing different metal-organic architectures [30][31][32][33][34][35][36][37][38]. Indeed, picolinate complexes of almost every metal from the periodic system have been prepared and structurally characterized by now [39], including mercury(II). Monomeric [17] and polymeric [40][41][42] mercury(II) complexes with picolinate, but also with picolinate derivatives (e.g., methyl, ethyl, n-propyl and i-propyl esters [41,43] and ethyl ester of 6-methylpicolinate [41]), are known from the literature. The mercury(II) complex [HgI(pic)(picH)], containing iodide and both picolinate ion and picolinic acid molecule [17], has been prepared previously. As it became evident that slight changes in the preparation procedures (e.g., metal-to-ligand ratio, solvent choice, pH of the solution, temperature) of mercury(II) complexes with picolinate and its derivatives lead to complexes of various compositions and dimensionalities, the possibility of obtaining new picolinate mercury(II) complexes under defined experimental conditions needed to be examined further. A goal was to correlate the experimental conditions used during syntheses with the composition and dimensionality of the complexes.

Syntheses Aspects
The reactions of mercury(II) chloride and bromide with picolinic acid were studied in aqueous, methanol and ethanol solutions (using two different metal-to-ligand ratios, 1:1 and 1:2) to see how the solvent change and metal-to-ligand ratio would affect the obtained complexes (Scheme 1). The coordination polymer 1 was obtained in an aqueous solution by using a metal-to-ligand ratio of 1:1 or 1:2 and adjusting the pH of the reaction mixture until precipi- The coordination polymer 1 was obtained in an aqueous solution by using a metal-toligand ratio of 1:1 or 1:2 and adjusting the pH of the reaction mixture until precipitation of the compound was completed, while mononuclear complexes 2 and 3 were obtained in methanol or ethanol solutions by using a metal-to-ligand ratio of 1:1 or 1:2. Reactions of HgCl 2 with picH performed in water solution without increasing the pH value always led to the mixture of polymeric and mononuclear complexes. Regardless of the metal-toligand ratios (1:1 or 1:2), the mononuclear complex 3 was always obtained. Reactions were performed only in methanol and ethanol due to the very poor solubility of the mercury(II) bromide in water. The presence of water and a slight increase of the pH value favors the formation of the coordination polymer 1.

Crystal Structures
The asymmetric unit of {[HgCl(pic)]} n (1) consists of a mercury(II) ion, a chloride ion and a picolinate ion. The mercury(II) ion is four-coordinated with the chloride Cl1 ion, pyridine N1, and carboxylate O1 and O2 i atoms (symmetry code (i) x, -y + 3/2, z + 1/2), resulting in a coordination environment that can be described as a highly distorted tetrahedron (τ 4 value [44] of 0.67; Figure 1a and Table S1 [45] (0.08 for 2 and 0.02 for 3) suggest that the mercury(II) coordination environment in both 2 and 3 can be best described as a square pyramid. The mercury(II) ion is significantly displaced from the center of the basal plane toward the halide ion in both 2 and 3, with a maximum out-of-plane deviation of 1.081(2) Å (in 2) and 1.079(6) Å (in 3).
The highly distorted tetrahedral coordination environment in 1 can be also seen in a wide range of bond angles around the mercury(II) ion (72.6(2) • -147.0(1) • ; Table S1 in the Supplementary Materials). The square-pyramidal coordination environments around the mercury(II) ions in 2 and 3 are also highly distorted, as the angles for the trans bond lengths in 1-3 are comparable to those seen in the related mercury(II) complexes containing picolinate ions [17,[40][41][42] and the mentioned picolinate derivatives [41,43].
A fairly similar 1-D mercury(II) coordination polymer with picolinate, {[Hg(pic) 2 ]} n , is known from the literature [40,41]. Although picolinate ions in {[Hg(pic) 2 ]} n act in the same way (chelating and bridging ligands) as in 1, there are no halide ions in {[Hg(pic) 2 ]} n , and two picolinate ions bridge two neighboring mercury(II) ions (as opposed to one bridging picolinate in 1), leading to an octahedral coordination of mercury(II) ions in {[Hg(pic) 2 ]} n [40,41]. The complexes 2 and 3 are very similar to the known [HgI(pic)(picH)], but they are not isostructural, as the unit cell parameters and the space group of [HgI(pic)(picH)] differs [17]. However, while mercury(II) ions in 2 and 3 exhibit a distorted square-pyramidal coordination, mercury(II) ion in [HgI(pic)(picH)] shows a distorted trigonal-bipyramidal coordination [17].     PXRD was used to confirm the phase purity and bulk composition of compounds 1-3 ( Figure 4). The powder diffraction traces (bulk sample) of all three compounds are consistent with the traces calculated from the single crystal diffraction data, indicating pure phases of compounds 1-3. The almost identical powder patterns of isostructural compounds 2 and 3 are consistent with their crystal structures.

Spectroscopic studies (FT-IR, NMR)
The recorded IR spectra of coordination polymer 1 and mononuclear comp and 3 ( Figure S1 in the Supplementary Materials) are in a good agreement w crystal structures. Very broad bands in the spectra of 2 and 3 centered around 34 can be attributed to the hydrogen-bonded (O-H⋅⋅⋅O) carboxylic groups [47,48]. Th bands in the IR spectrum of free picH centered around 2600 and 2100 cm −1 are co with the O-H⋯N intermolecular hydrogen bonding. The disappearing above-mentioned bands from the IR spectra of the metal complexes suggests a co tion of the N atom to the mercury(II) ion [48,49]. The characteristic ν(C=O) band carboxylic group at 1718 cm −1 in the IR spectra of 2 and 3 are connected with the e of neutral picolinic acid molecules [47,48]. On the other hand, the absence of th around 3400 and 1700 cm −1 in the IR spectrum of 1 indicates a complete deproton the carboxylic group [41,[48][49][50][51]. The IR spectra of the metal complexes show str sorptions of the antisymmetric νas(COO − ) and symmetric νs(COO − ) stretching vi of the carboxylate ligands in the 1650−1590 cm −1 and 1380−1350 cm −1 regions. The 13 C NMR data for picH and compounds 1-3 are given in Table 1. The n tion of the atoms is consistent with the IUPAC nomenclature (Scheme 2). The co tion effects due to the mercury(II)-induced electronic redistribution throughout ordinated ligands are well-known from the literature [18][19]41]. The NMR dat show small but significant complexation effects. The greatest changes upon comp with mercury(II) are observed for the coordination polymer 1. In the 13 C NMR sp all complexes, the greatest change (deshielding effect, 4.27-1.38 ppm) upon comp with mercury(II) was observed for the chemical shift of the C-4 atoms. The shielding effects (2.89-1.66 ppm) were observed for the carboxylate carbon atom only two bonds away from the coordinated mercury(II) ion. The shielding effec sistent with a reduction of the π-electron density of the C=O bond due to the ylic/carboxylate coordination to mercury. Small but significant complexation eff served in the NMR spectra of prepared complexes are in agreement with the exis mononuclear (2 and 3) and polymeric (1) species in the DMSO solutions.

Spectroscopic Studies (FT-IR, NMR)
The recorded IR spectra of coordination polymer 1 and mononuclear compounds 2 and 3 ( Figure S1 in the Supplementary Materials) are in a good agreement with their crystal structures. Very broad bands in the spectra of 2 and 3 centered around 3400 cm −1 can be attributed to the hydrogen-bonded (O-H···O) carboxylic groups [47,48]. The broad bands in the IR spectrum of free picH centered around 2600 and 2100 cm −1 are connected with the O-H· · · N intermolecular hydrogen bonding. The disappearing of the above-mentioned bands from the IR spectra of the metal complexes suggests a coordination of the N atom to the mercury(II) ion [48,49]. The characteristic ν(C=O) bands of the carboxylic group at 1718 cm −1 in the IR spectra of 2 and 3 are connected with the existence of neutral picolinic acid molecules [47,48]. On the other hand, the absence of the bands around 3400 and 1700 cm −1 in the IR spectrum of 1 indicates a complete deprotonation of the carboxylic group [41,[48][49][50][51]. The IR spectra of the metal complexes show strong absorptions of the antisymmetric ν as (COO − ) and symmetric ν s (COO − ) stretching vibrations of the carboxylate ligands in the 1650−1590 cm −1 and 1380−1350 cm −1 regions. The 13 C NMR data for picH and compounds 1-3 are given in Table 1. The numeration of the atoms is consistent with the IUPAC nomenclature (Scheme 2). The complexation effects due to the mercury(II)-induced electronic redistribution throughout the coordinated ligands are well-known from the literature [18,19,41]. The NMR data of 1-3 show small but significant complexation effects. The greatest changes upon complexation with mercury(II) are observed for the coordination polymer 1. In the 13 C NMR spectra of all complexes, the greatest change (deshielding effect, 4.27-1.38 ppm) upon complexation with mercury(II) was observed for the chemical shift of the C-4 atoms. The largest shielding effects (2.89-1.66 ppm) were observed for the carboxylate carbon atoms, being only two bonds away from the coordinated mercury(II) ion. The shielding effect is consistent with a reduction of the π-electron density of the C=O bond due to the carboxylic/carboxylate coordination to mercury. Small but significant complexation effects observed in the NMR spectra of prepared complexes are in agreement with the existence of mononuclear (2 and 3) and polymeric (1) species in the DMSO solutions.

Thermal analysis (TGA/DSC)
The thermal behavior of compounds 1-3 has been investigated by simultaneous TGA/DSC analysis from room temperature up to 450 °C in flowing nitrogen ( Figure S2 in the Supplementary Materials).

Materials and physical measurements
All commercially available chemicals were of reagent grade and were used as received without further purification. CHN elemental analyses were carried out with a Perkin-Elmer 2400 Series II CHNS analyzer in Analytical Services Laboratories of the Ruđer Bošković Institute, Zagreb, Croatia. The mercury content in the complexes was determined by complexometric titration with sodium diethyldithiocarbamate of the solution obtained after decomposition of the compounds in aqua regia [54]. The pH values were measured by a Mettler Toledo MP220 Basic pH/mV/°C Meter.
IR spectra were obtained from KBr pellets in the range 4000-500 cm −1 on a Perkin-Elmer Spectrum Two FT-IR spectrometer. The one-dimensional homonuclear 1 H and 13 C NMR spectra were recorded with a Bruker AV 600 spectrometer, operating at 600.133 MHz for the 1 H nucleus and 150.917 MHz for the 13 C nucleus. The samples were dissolved in DMSO-d6 in 5 mm NMR tubes. Chemical shifts, in ppm, are referred to TMS as internal standard.
Thermogravimetric analysis was performed using a simultaneous TGA-DSC analyzer Mettler-Toledo TGA/DSC 3+. The samples of compounds 1-3 were placed in alumina pans (70 μL) and heated in flowing nitrogen (50 mL min −1 ) from room temperature up to 450 °C at a rate of 10 °C min −1 . Data collection and analysis were Scheme 2. Numeration of the atoms in picH.

Thermal Analysis (TGA/DSC)
The thermal behavior of compounds 1-3 has been investigated by simultaneous TGA/DSC analysis from room temperature up to 450 • C in flowing nitrogen ( Figure S2 in the Supplementary Materials).

Materials and Physical Measurements
All commercially available chemicals were of reagent grade and were used as received without further purification. CHN elemental analyses were carried out with a Perkin-Elmer 2400 Series II CHNS analyzer in Analytical Services Laboratories of the Ruder Bošković Institute, Zagreb, Croatia. The mercury content in the complexes was determined by complexometric titration with sodium diethyldithiocarbamate of the solution obtained after decomposition of the compounds in aqua regia [54]. The pH values were measured by a Mettler Toledo MP220 Basic pH/mV/ • C Meter.
IR spectra were obtained from KBr pellets in the range 4000-500 cm −1 on a Perkin-Elmer Spectrum Two FT-IR spectrometer. The one-dimensional homonuclear 1 H and 13 C NMR spectra were recorded with a Bruker AV 600 spectrometer, operating at 600.133 MHz for the 1 H nucleus and 150.917 MHz for the 13 C nucleus. The samples were dissolved in DMSO-d 6 in 5 mm NMR tubes. Chemical shifts, in ppm, are referred to TMS as internal standard.
Thermogravimetric analysis was performed using a simultaneous TGA-DSC analyzer Mettler-Toledo TGA/DSC 3+. The samples of compounds 1-3 were placed in alumina pans (70 µL) and heated in flowing nitrogen (50 mL min −1 ) from room temperature up to 450 • C at a rate of 10 • C min −1 . Data collection and analysis were performed using the program package STARe Software 15.01 MettlerToledo GmbH, 2015.
Powder X-ray diffraction experiments (PXRD) were measured on a Malvern Panalytical Aeris XRD diffractometer with CuKα (1.5406 Å) radiation, Ni filter, and solid-state PIXcel3D-Medipix3 detector. Samples were prepared as a thin layer on a silicon zero-background plate. Data were collected in the 2θ range from 5 • to 50 • with a step size of 0.02173 • , scan rate 10 s/ • , 1 4 inch divergence slit, and 13 mm beam mask. for single-crystal X-ray structure determination, were formed in a few hours. The mixture of 1 and 2 was isolated by allowing the mother liquor to stand for a few days.

Synthesis of [HgCl(pic)(picH)] (2)
A solution of picH (0.46 g, 3.74 mmol) in 20 mL MeOH was added dropwise to a solution of HgCl 2 (0.50 g, 1.84 mmol) in 10 mL MeOH. Colorless crystals suitable for single-crystal X-ray structure determination were formed in a few days. The colorless crystals were filtered off, washed with MeOH, and dried in air. Yield: 48% (based on HgCl 2 ). Anal. Calc. for HgClC 12

X-ray Crystallographic Analysis
Suitable single crystals of 1-3 were selected and mounted in Paratone-N oil onto cryoloops. Data collection was carried out on an XtaLAB Synergy-S Dualflex diffractometer with PhotonJet (Mo) microfocus X-ray source and HyPix-6000HE hybrid photon counting (HPC) X-ray area detector, using graphite monochromated MoKα (λ = 0.71073 Å) radiation at low temperature (170(2) K) and by applying the CrysAlisPro Software system [55]. Data reduction and cell refinement were performed by the CrysAlisPro Software system [55]. The structures were solved by SHELXT [56] and refined by SHELXL-2018/3 [57]. The refinement procedure was done by the full-matrix least-squares methods based on F 2 values against all reflections. The figures were made with MERCURY (Version 2020.2.0) [58]. The crystallographic data for 1-3 are summarized in Table 2.

Conclusions
The neighboring mercury(II) ions are bridged by an N,O,O -picolinate ion into a polymeric chain of 1, while both complexes 2 and 3 are mononuclear, containing the respective halide ion and N,O-bidentate picolinate ion and picolinic acid molecule. It is possible to control both the composition and dimensionality of the mercury(II) complexes formed in the system containing mercury(II), picolinic acid, halide ions, and a solvent. This control was achieved by varying the solvent type and metal-to-ligand ratio in this system. The IR, 1 H, and 13 C NMR data correlate well with the solid-state structures of 1, 2, and 3. According to the NMR data, there is no decomposition of 1, 2, and 3 in the respective DMSO solutions and 1 remains polymeric, while 2 and 3 remain mononuclear in the solutions. The observed differences in the chemical shifts were ascribed to complexation effects and were reliable enough to distinguish between polymeric (1) and monomeric (2 and 3) compounds.