Short communication
Design and self-assembly of new [2 × 2] grids constructed by lanthanide ions and a Schiff base

https://doi.org/10.1016/j.inoche.2020.108067Get rights and content

Highlights

  • The system Ln(III)-Schiff base has been investigated in solution by potentiometric techniques.

  • Tetranuclear complexes containing Ln(III) and a Schiff base have been prepared in good yield.

  • The structure of the polynuclear complexes reveals the formation of neutral [2 × 2] grids.

Abstract

[2 × 2] grids of general formula [Ln4L4Cl4(DMF)4].2H2O (Ln = Sm (1), Eu (2), Tb (3); H2L = (E)-N’-((E)-(hydroxyimino)butan-2ylidene)salicylolhidrazide; DMF = dimethylformamide) were successfully constructed using a lanthanide ion as corners and L2-, a Schiff base, as linkers. The single crystal X-ray diffraction study reveals that compounds 1 and 3 are neutral tetranuclear complexes. They feature a [2 × 2] grid distorted skeleton. Coordination sphere of Ln ions are completed by one chloro ligand and a solvent molecule. The grids were obtained after a careful study of the chemical speciation of these systems performed by potentiometric titrations (25.0 °C in 40:60 DMF:water mixture, I = 0.50 mol L-1 NMe4Cl).

Introduction

The drive to shrink electronic devices to the nanoscale level has led to the design and evaluation of molecular-scale components, in particular those based on polynuclear complexes. They are endowed with sensing, switching, logic, and information storage functions [1], [2], [3]. With this in mind, grid-like metal ion arrays in which a short set of metal ions is held in a regular network of organic ligands in a quasi-perpendicular arrangement present particularly attractive features [4], [5]. They exhibit interesting redox, magnetic and spin-state transition properties [6], [7], [8]. In addition, the bidimensional arrangement of the grid allows deposition onto solid surfaces and formation of new materials with promising applications [9], [10].

The design of grids of specific nuclearity is a challenging task and this is why most of the reports found in literature rely solely on serendipity [4]. Different aspects of the ligands and metal ions should be considered. In the cases of square [n × n] or rectangular [n × m] grids, a specific molar ratio between metal ions and ligands must be fulfilled. A ligand with n coordination subunits is able to form a homoleptic [n × n] metallo-array, composed of 2n organic ligands and n2 metal ions, giving an overall [M(n2)L(2n)] stoichiometry. As a consequence, the amount of ligand should be carefully controlled to avoid the formation of 2:1 or 3:1 mononuclear complexes. The choice of the metal ion is also very important. Those possessing tetrahedral or octahedral coordination geometry are more adequate to provide perpendicular connections between ligands. For this reason, d metal ions have been the first choice to assemble grids. However, the key element is to use polytopic ligands having pockets with dimensions and donor arrangements that best match the coordination needs of the metal ions [4], [5].

Schiff bases are potential polytopic ligands to build [n × n] grids if prepared with suitable synthons. Many examples can be found in the literature including different metal ions, for example with Ag(I) [11], [12], [13], Cu(I) [14], Cu(II) [15], [16], [17], [18], [19], Zn(II) [17], [20], [21], Fe(II) [22], Cd(II) [23], Mn(II) [17], [23], Ni(II) [16], [24], Co(II) [17] and Mn(III) [25]. On the other hand, reports of grid-like compounds with lanthanide ions (Ln) are very scarce, in spite of a rich set of Ln(III)-Schiff base complexes [26]. Very recently, a Dy [2 × 2] grid built with the Schiff base (2-[(4-methoxyphenyl)imino]methyl]-8-hydroxy-quinoline, H2L) and an auxiliary ligand (2-thenoyltrifluoroacetone, TTA), [Dy4(TTA)4L4(H2O)2]⋅CH3OH, has been reported. This cluster exhibits a slow magnetic relaxation [8]. Inclusion of lanthanide ions suppose an added value owing to the possible applications that arise from their physical properties, which include molecular magnetism, magnetocaloric effect, luminescence and catalysis [27], [28], [29], [30].

In order to explore a rational design of this kind of compounds, we have been working with the Schiff base (E)-N’-((E)-(hydroxyimino)butan-2ylidene)salicylolhidrazide, H2L (Scheme 1). We used this ligand to prepare [Zn(HL)2]⋅DMF (DMF = dimethylformamide) [31]. In this work we expand the use of this Schiff base to build [2 × 2] grids with Ln ions with special care on the design of the synthetic conditions in order to avoid the obtention of other polynuclear complexes.

The first step was the study of the systems Ln(III)-ligand in solution. La, Sm, Eu, Tb and Yb were selected as metal ions to cover a wide range of metal radii. Protonation equilibrium constants of the ligand (Table S1) were previously reported at 25.0 °C in 40:60 DMF:water mixture, I = 0.50 mol L-1 NMe4Cl [31]. Hydrolysis constants of the Ln(III) ions were also reported under similar conditions [32]. Both sets of equilibrium constants were taken into account for the determination of the stability constants of the species Ln(III)-L under such conditions. The results are presented in Table S1. Only species 1:1 and 1:2 (metal to ligand) are detected in solution, with similar stability constant values for the five ions.

With these results, it is possible to build species distribution diagrams for the five systems. Fig. 1 shows the case of Eu, while Fig. S1-S4 show similar diagrams for La, Sm, Tb and Yb. Species distribution diagrams show that complexes are predominant at basic pH values, i.e., its formation would require the use of an external base during synthesis. 1:1 species ([Eu(HL)]2+, [EuL]+, [EuL(OH)], [EuL(OH)2]- are clearly predominant, even under ligand excess.

The above results, in particular the low relative presence of 2:1 species, together with the adequate chemical structure of L2-, stimulated us to prepare Ln grids with this Schiff base. The structure of the ligand contains two different pockets. One contains O,O or O,N atom donor set, and includes the phenol residue. The second pocket would be bidentate (N,N or N,O) or tridentate (N,N,O) and includes the oxime functional group (Scheme 1). Spatial and donor atoms arrangements could match Ln ions in [2 × 2] grids [4]. According to previous studies by J.M. Lehn et al. [33], a mixture of ligand and metal ion in 1:1 molar ratio is adequate to obtain these grids; this strategy was used in this work (experimental details are presented in SI). We successfully prepared three new complexes [Ln4L4Cl4(DMF)4]⋅2H2O (Ln = Sm (1), Eu (2), Tb (3)) by direct reaction of equimolar amounts of LnCl3⋅6H2O, H2L and ethylenediamine to deprotonate the ligand. This is a one-step synthesis that suggests a self-assembling process. The predominance of 1:1 species in solution, together with the low solubility of the neutral species in the mixed-solvent system, are probably the driving forces for this process.

It was possible to obtain single crystals of 1 and 3, which crystallize in the tetragonal space group P4¯21c. They are isostructural, so, we will only discuss the structure of 1. The molecular structure consists of neutral units containing four Sm(III) ions, four deprotonated L2- ligands, four chloro ligands, and four coordinated DMF molecules with 4¯ symmetry. The asymmetric unit is formed by one fourth of the molecule, containing one unit of each component as shown in Fig. 2. The structure is completed with half a crystallization water molecule per asymmetric unit. The four metal ions are symmetry equivalent. They are surrounded by two nitrogen (N2, N3) and four oxygen atoms (O2, O1i, O2i, O1ii) from three L2- anions, O4 from a DMF molecule and a chloro ligand (i: -x + 1, -y + 1, z; ii: y, -x + 1, -z + 1). SmO5N2Cl polyhedra can be described as triangular dodecahedron (TDD-8) (Fig. S5) [34], [35].

From the ligand point of view, the Schiff base is doubly deprotonated. Previous structures contain this ligand either as HL- or L2- [31], [36]. Both forms can also be found in the same compound, like in the polynuclear [Dy6(HL)3(L)3(OH)2(CO3)3(H2O)5(DMF)4][(HL)2Dy(HL)2]⋅6.5DMF⋅3.5H2O [36]. It should be noted that ref. [36] uses H3L as the label for the neutral ligand, in spite of having only two ionizable protons in solution. We have uniformized the nomenclature throughout this work using H2L as the neutral form of the Schiff base.

Protons from phenol (bound to O1) and acylhydrazone group (bound to the proximal nitrogen atom N1) are lost in 1. These two deprotonated acid sites can be found in the Dy complex, but in different independent units of the ligand. When hydrogen bound to N1 is lost, tridentate units through N2, N3 and O2 are observed. On the other hand, when O1 hydrogen is lost, monodentate coordination is verified [36]. Despite these similarities, the conformation of the ligand in 1 is very different from previously observed ones. In the structures of the neutral ligand [37] and other reported complexes [31], [36], O1 and N1 lie at the same side of the molecule, opposite to O2. The ligand is basically planar with slight deviations in the oxime end. This is not the case in 1, where O1 and O2 are found on the same side (as shown in Scheme 1), due to a rotation of 155° of the phenol residue leaving both negatively charged atoms O1 and N1 at opposite sides of the molecule. In this conformation, the phenol residue is not coplanar with the rest of the molecule as extracted from the rotation angle mentioned above.

Each ligand is connected with three different Sm ions (Fig. 3) using both ligation pockets. N2 and N3 are coordinated together with O2 to the same Sm(III) ion in a tridentate fashion, with bond distances of 2.544(5), 2.632(6) and 2.399(4) Å, respectively. The other pocket, with O1 and O2, provides two μ2-O bridges between Sm ions. Each oxygen atom is bis(monodentate) with average Sm-O distance of 2.3988(4) Å. N1 and O3 do not participate in the coordination.

Analyses of the previous structures which contain the Schiff base used in this work shows that in the solid-state deprotonation of N1 is more frequent than that of O1. Deprotonation of O3 in the oxyme group is rarely observed. The observed electron density in the X-ray crystal structure and observed bond distances and hydrogen-bond scheme in 1 confirms the location of H-atoms. On one side, there is no residual electron density close to N1 that could suggest the presence of an H atom, while residual electron density peaks appear close to O3 if an H atom is not included. Moreover, the position of this H atom in the oxime group clearly suggests the formation of a O3-H3B⋅⋅⋅Cl1 intra-complex hydrogen bond further stabilizing the protonation of O3. Besides, C7-O2 bond distance is longer than average for a C = O amide carbonyl pointing to the delocalization of the negative charge over N1.

Ideal [2 × 2] grids contain a fully orthogonal or planar M4 metal ion array. However, most M4 arrays usually adopt a butterfly form. This depends on the coordination geometry of the central atom and the planarity of the connecting ligand [33], [38], [39]. In the case of 1, the four non-bonding Sm-Sm distances are 3.895 Å, and internal Sm-Sm-Sm angles are 87.05° (Fig. S6). The skeleton is far from a perfect square and can be better described as a butterfly shape with a puckering angle of 36.415°. The distortion is not originated in the Schiff base which remains almost planar, but in the presence of the double μ2-O bridge between Sm ions. The bridge forces one Sm ion out of the ligand’s plane as can be seen in Fig. S7.

The use of an asymmetric ditopic ligand such as L gives rise to the possible formation of four types of [2 × 2] isomeric grids. They differ in the way the four ligands cross the corner of the grid. They can be in some cases detected in solution, but it is common that only one of them is obtained in the solid state [33]. In the case of compound 1, only the optically inactive isomer is crystallized (Fig. S8).

Section snippets

Author statement

Gabriela Mendoza-Sarmiento – Conducting the experimental research and investigation process, specifically performing the experiments in solution and synthesis. Preparation, creation and presentation of the published work.

Fernando Igoa – Conduced the experimental research and investigation process, specifically performing the experiments in solution. Preparation, creation and presentation of the published work.

Leopoldo Suescun – Conduced the experimental research and investigation process,

Declaration of Competing Interest

The authors declared that there is no conflict of interest.

Acknowledgements

We are grateful for the financial support from the Uruguayan organizations CSIC (Comisión Sectorial de Investigación Científica), PEDECIBA (Programa para el Desarrollo de las Ciencias Básicas) and ANII (Agencia Nacional de Investigación e Innovación). G. M.-S. is grateful to ANII for a postdoctoral scolarship (PD-NAC-2016-1-133445).

References (39)

  • J.-Q. Liu et al.

    Coord. Chem. Rev.

    (2020)
  • A. Schmidt et al.

    Coord. Chem. Rev.

    (2014)
  • Y.-L. Hou et al.

    Inorg. Chem. Commun.

    (2020)
  • Y.-Z. Li et al.

    Coord. Chem. Rev.

    (2019)
  • J. Duan et al.

    Coord. Chem. Rev.

    (2019)
  • S.O. Malinkin et al.

    Polyhedron

    (2012)
  • H. Wang et al.

    Inorg. Chim. Acta

    (2009)
  • X.-J. Hong et al.

    Inorg. Chem. Commun.

    (2014)
  • M.T. Kaczmarek et al.

    Coord. Chem. Rev.

    (2018)
  • S.-J. Liu et al.

    Coord. Chem. Rev.

    (2019)
  • G.D. Klungness et al.

    Polyhedron

    (2000)
  • S. Erbas-Cakmak et al.

    Chem. Soc. Rev.

    (2018)
  • L.N. Dawe et al.

    Chem. Soc. Rev.

    (2009)
  • M. Ruben et al.

    Angew. Chemie. Int. Ed.

    (2004)
  • M. Ruben et al.

    Chem. Eur. J.

    (2003)
  • L.H. Uppadine et al.

    Chem. Eur. J.

    (2005)
  • J.R. Price et al.

    Dalton Trans.

    (2006)
  • J.R. Price et al.

    Inorg. Chem.

    (2008)
  • H.-C. Wu et al.

    Inorg. Chem.

    (2006)
  • Cited by (0)

    View full text