Tetradentate NOO′O″ Schiff-Base Ligands as a Platform for the Synthesis of Heterometallic Cd^II-Fe^III and Cd^II-Cr^III Coordination Clusters ^†

Abstract

The chemistry of heterometallic metal complexes continues to attract the interest of molecular inorganic chemists mainly because of the properties that different metal ions can bring to compounds. Contrary to the plethora of 3d–4f- and 3d–3d′-metal complexes, complexes containing both 3d- and 4d-metal ions are much less studied. The choice of the bridging organic ligand is of paramount importance for the synthesis of such species. In the present work, we describe the use of the potentially tetradentate NOO′O″ Schiff bases N-(2-carboxyphenyl)salicylideneimine (saphHCOOH) and N-(4-chloro-carboxyphenyl)salicylideneimine (4ClsaphHCOOH) in Cd^II-M^III (M = Fe, Cr) chemistry. The complexes [Cd₂Fe₂(saphCOO)₄(NO₃)₂(H₂O)₂] (1), [Cd₂Cr₂(saphCOO)₄(NO₃)₂(H₂O)₂] (2), [Cd₂Fe₂(4ClsaphCOO)₄(NO₃)₂(H₂O)₂] (3) and [CdCr₂(4ClsaphCOO)₄(H₂O)₃(EtOH)] (4) have been structurally characterized, the quality of the structure of the latter being poor but, permitting the knowledge of the connectivity and the main structural features. Complexes 1–3 are isostructural, but not isomorphous, possessing a variety of lattice solvent molecules (EtOH, MeCN, CH₂Cl₂, H₂O). The metal topology can be described as two isosceles triangles sharing a common Cd^II…Cd^II edge. The two Cd^II atoms are doubly bridged by two μ-aqua groups. The M^III…Cd^II sides of the triangles are each asymmetrically bridged by one carboxylate oxygen atom of a 2.2111 saphCOO²⁻/4ClsaphCOO²⁻ ligand. The core of the molecules is {Cd₂M₂(μ-O_aqua)₂(μ-OR)₄}⁶⁺, where the OR oxygen atoms are the bridging carboxylate oxygens. The coordination spheres of the metal ions in the centrosymmetric molecules are [Cd(O_aqua)₂(O_carboxylato)₄(O_nitrato)₂] and [M(N_imino)₂(O_carboxylato)₂(O_phenolato)₂]. The biaugmented trigonal prism is the most appropriate for the description of the coordination geometry of the Cd^II atoms in 1 and 3, while the geometry of these metal ions in 2 is best described as distorted triangular dodecahedral. A combination of H-bonding and π–π stacking interactions give interesting supramolecular patterns in the three tetranuclear compounds. The three metal ions in 4 define an isosceles triangle with two almost equal Cd^II…Cr^III sides. The Cd^II center is linked to each Cr^III atom through one carboxylato oxygen of a 2.2111 4ClsaphCOO²⁻ ligand. The core of the molecule is {CdCr₂(μ-OR)₂}⁶⁺, where the OR oxygen atoms are the bridging carboxylato oxygens. A tridentate chelating 1.1101 4ClsaphCOO²⁻ ligand is bonded to each Cr^III. The coordination spheres are [Cd(O_aqua)₃(O_ethanol)(O_{bridging carboxylato})₂(O_{terminal carboxylate})₂] and [Cr(O_{bridging carboxylato})(O_{terminal carboxylato})(O_phenolato)₂(N_imino)₂]. Complexes 1–4 are the first heterometallic 3d–4d complexes based on saphHCOOH and 4ClsaphCOOH. The structures are critically compared with those of previous reported Zn^II-M^III (M = Fe, Cr) complexes. The IR and Raman spectra of the complexes are discussed in terms of the coordination modes of the ligands involved. UV/VIS spectra in CH₂Cl₂ are also reported, and the bands are assigned to the corresponding transitions. The δ and ΔE_Q ⁵⁷Fe-Mössbauer parameters of 1 and 3 at room temperature and 80 K suggest the presence of isolated high-spin Fe^III centers. Variable-temperature (1.8–310 K) and variable-field (0–50 kOe) magnetic studies for 1 and 2 indicate the absence of M^III…M^III exchange interactions, in agreement with the long distances (~8 Å) between the paramagnetic metal ions. The combined work demonstrates the ability of saphCOO²⁻ and 4ClsaphCOO²⁻ to give 3d–4d metal complexes.

1. Introduction

The field of “classical” (i.e., based on atoms or ions) mixed-metal materials continues to attract the intense interest of researchers working in condensed-matter physics, solid-state chemistry and theoretical aspects of science, mainly for their applications [1]. Such materials can be broadly divided into two categories. The majority include two (or more) different metal ions and monoatomic inorganic bridges (e.g., oxido, halido or sulfido). For example, one of the myriads of these materials is BiMnO₃; this shows a ferroelectric transition at T_FE = 800 K and a ferromagnetic transition at T_FM = 110 K, below which the two orders coexist, being a unique example in which both magnetization and electric polarization are reasonably large [2]. Compounds containing exclusively two (or more) different kinds of metal (or metalloid) atoms are defined as intermetallics [3]. A representative example is crystalline MnNi₂Ga, which has been shown to produce approximately 10% magnetic-field-induced strain, while the compositions FeGa₃, YNiGe₂, Sm₂NiGa₁₂ or SmNiSi₃ appear exotic to the average chemist.

The field of “molecular” (i.e., based on molecules) mixed-metal materials has emerged in the last 20 years or so, mainly due to its relevance in several areas including molecular nanomagnetism [4], catalysis [5], porous complexes [6], quantum technology [7] and the development of precursors for multifunctional materials [8]. Such “molecular” materials consist mainly of heterometallic coordination complexes (dinuclear or polynuclear), often deposited in surfaces (e.g., graphene and Au). In the complexes, the different metal ions are bridged by solely inorganic groups (e.g., CN⁻) [9], exclusively by organic ligands or a combination of them. This work concerns the second case. There are few strategies for the synthesis of molecular mixed-metal species [10,11,12]. One of them involves the use of polytopic organic ligands that contain compartments (or “pockets”) capable of binding different metal ions; such ligands introduce preprogrammed coordination information that is “stored” in the “pockets”. When the ligands react with different metal ions, the latter interpret this information according to their own “coordination algorithms”. The nuclearity of the products depends on the organic ligand chosen, the preferred coordination numbers and geometries of the metal centers, and the reaction conditions. This may in turn lead to a variety of magnetic exchange interactions, which depend on the number and nature of the bridging atoms or groups and the magnetic orbitals available.

Although the number of complexes containing two different transition-metal ions (d-d′) increases steadily [9,13], these remain less than the thousands of d-4f complexes that have been reported so far [10,11,12,14,15,16]. We recently embarked on a new program aiming at the synthesis and study of the properties of non-cyanido 3d-3d′, 3d-4d and 3d-5d complexes, starting from species containing one paramagnetic and one diamagnetic metal ion. There are significantly fewer 3d–4d mixed-metal complexes than 3d–3d′ and 3d–5d ones. Thus, there is great interest in the chemistry of such species. For some transition metal-ion pairs, e.g., Cu^II-Fe^III and Ni^II-Fe^III, the Hard and Soft Metal and Bases (HSAB) model is particularly useful. For other combinations, e.g., the Cd^II-Cr^III and Cd^II-Fe^III pairs of the present work, this model is expected to have limited success, and the reactions sometimes give products containing only one of them, or mixtures of homometallic compounds. For such pairs, a large number of “try and see” exercises are necessary for success, and the desired “self-assembly” process mainly relies on the choice of the organic ligand.

Polydentate O,N Schiff bases are candidates for both the designed and serendipitous assembly of 3d–3d′ and 3d–nd (n = 4, 5) complexes [17]. Advantages of these ligands are their logical modular synthesis and the control that can be achieved over the nature and position of donor atoms, denticity, combined chelating and bridging behavior, steric characteristics and electronic features. An important family of potentially tetradentate NOO′O″ Schiff-base ligands arises from the condensation of salicylaldehyde (or its derivatives) and 2-aminobenzoic (or its derivatives). The two members of this family that have been employed in this work are N-(2-carboxylphenyl)salicylideneimine, abbreviated as saphHCOOH [the IUPAC name is {((2-hydroxyphenyl)methylidene)amino}benzoic acid], and N-(4-chloro-carboxyphenyl)salicylideneimine, abbreviated as 4ClsaphHCOOH [the IUPAC name is 4-chloro{((2-hydroxyphenyl)methylidene)amino}benzoic acid]; the structural formulas of the two ligands are illustrated in Scheme 1. Single (–COOH) or double (–COOH and –OH) deprotonation produces mono- or dianionic ligands that can coordinate to two or more, same or different, metal ions, thus resulting in homo- or heterometallic dimers, polynuclear complexes (coordination clusters) or coordination polymers. Before starting our endeavors, only homometallic complexes, e.g., see refs. [18,19,20,21,22,23,24,25], had been synthesized and structurally characterized with saphHCOO⁻ and saphCOO²⁻. Concerning the coordination chemistry of 4ClsaphHCOOH and its anionic forms, this had been almost unexplored being confined to complex [Cu(4ClsaphCOO)(phen)] [26,27], where phen is the N,N’-chelating ligand 1,10-phenanthroline. Our simplistic “design” idea for the selection of these ligands was that a divalent, borderline (HSAB) nd (n = 3, 4, 5) metal (acid) would bind to the border-line O_hydroxyl, N_imine, O_carboxylate donor set (base), forming two six-membered, edge-sharing chelating rings, whereas the remaining hard carboxylate oxygen would be free to form a coordination bond with a trivalent hard acid 3d metal ion. Alternatively, the O_hydroxyl, N_imine pair would bind to the divalent metal, forming a six-membered chelating ring and the two carboxylate oxygens to the trivalent 3d metal in a chelating or/and bridging fashion.

We have recently reported the use of saphHCOOH and 4ClsaphHCOOH in Zn^II/M^III (M = Fe, Co) chemistry, and isolated and studied the hexanuclear clusters [Zn₄Fe₂(saphCOO)₆(NO₃)₂(EtOH)₂], [Zn₄Cr₂(saphCOO)₆(NO₃)₂(H₂O)₂] and [Zn₄Fe₂(4ClsaphCOO)₆(NO₃)₂(EtOH)₂] [28]. Here, we report analogous synthetic studies in Cd^II/M^III reaction systems. Since the chemistries of Cd^II and Zn^II have similarities but also differences, we were interested to see if and how the group 12 divalent metal replacement affects the chemical and structural identity of the products. Cd^II-Fe^III compounds attract the interest of the inorganic scientific community because (i) Cd^II is an ideal NMR-active substitute for Zn^II in model Zn^II-Fe^III compounds, allowing studies of the heterodinuclear {Zn^IIFe^III} unit of purple acid phosphatase and human calcineurin, enzymes that facilitate the hydrolysis of phosphodiesters into phosphomonoesters, and construction of efficient catalysts [29]; (ii) the Cd^II/Fe^III combination has led to coordination clusters and polymers with interesting structures and metal topologies [30,31,32]; (iii) Cd^II-Fe^III clusters with suitable ligands exhibit band gap energy values (~2.5 eV) in the visible region, and have been used for the development of effective Schottky devices and as photocatalysts in aqueous media (e.g., for methylene blue dye degradation [33]); (iv) trinuclear Fe^III₂Cd^II complexes display high-temperature spin-crossover behaviors [34]; and (v) {Fe^III₇Cd^II} wheels are of interest for potential applications in quantum computing [35]. The interest in Cd^II-Cr^III chemistry arises from (i) the development of synthetic routes for {Cr^III_xCd^II} rings via templates (x = 7, 8) [36,37]; (ii) the investigation of metal distribution and disorder in the crystal structures of these wheel molecules [38]; (iii) the synthesis, structures and dynamic properties of hybrid organic–inorganic rotaxanes [39]; (iv) the development of the Primary Molecular Building Block (PMBB) approach in crystal engineering [40]; and (v) the discovery of new architectures based on Cr^III-containing Anderson-type polyoxoanions and Cd^II fragments [41].

2. Results and Discussion

2.1. Synthetic Comments

As mentioned in the Introduction, our general goal was to isolate Cd^II-Fe^III and Cd^II-Cr^III compounds with the anionic forms of saphHCOOH and 4ClsaphHCOOH as primary bridging ligands. A variety of synthetic parameters were studied, including the source of the two metal ions, the solvent, the temperature, the pressure (solvothermal techniques), the presence or absence of external base (to help deprotonation of the ligands), the time of the reaction (Cr^III is a kinetically non-labile species) and the crystallization techniques before arriving at the optimized procedures described in Section 3 (vide infra). For the isolation of single crystals of the products, the best choice of metal-containing starting materials were Cd(O₂CMe)₂·2H₂O, Fe(NO₃)₃·9H₂O and Cr(NO₃)₃·9H₂O. High temperatures were necessary for the preparation of the Cr^III compounds to overcome the well-known inertness of this 3d³ metal ion. Assuming that the structurally characterized complexes are the only products in solution, the formation of the compounds are summarized in Equations (1)–(3) [LH₂ = saphHCOOH, 4ClsaphHCOOH].

$$\begin{gathered} 2\mathrm{Cd}{({\mathrm{O}}_2\mathrm{CMe})}_2·2{\mathrm{H}}_2\mathrm{O}+2\mathrm{Fe}{({\mathrm{NO}}_3)}_3·9{\mathrm{H}}_2\mathrm{O}+4{\mathrm{LH}}_2\stackrel{EtOH/C{H}_{2}C{l}_2,20°\mathrm{C}}{\to } \\ [\mathrm{C}{\mathrm{d}}_2\mathrm{F}{\mathrm{e}}_2{(\mathrm{L})}_4{(\mathrm{N}{\mathrm{O}}_3)}_2{({\mathrm{H}}_2\mathrm{O})}_2]+4\mathrm{M}\mathrm{e}\mathrm{C}{\mathrm{O}}_2\mathrm{H}+4{{\mathrm{H}}_3\mathrm{O}}^{+}+4{\mathrm{N}{\mathrm{O}}_3}^{-}+16{\mathrm{H}}_2\mathrm{O} \end{gathered}$$

(1)

                     1, 3

$$\begin{gathered} 2\mathrm{C}\mathrm{d}{\left({\mathrm{O}}_2\mathrm{C}\mathrm{M}\mathrm{e}\right)}_2·2{\mathrm{H}}_2\mathrm{O}+2\mathrm{C}\mathrm{r}{\left(\mathrm{N}{\mathrm{O}}_3\right)}_3·9{\mathrm{H}}_2\mathrm{O}+4\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{h}\mathrm{H}\mathrm{C}\mathrm{O}\mathrm{O}\mathrm{H} \stackrel{EtOH/MeCN,T}{\to } \\ [\mathrm{C}{\mathrm{d}}_2\mathrm{C}{\mathrm{r}}_2{(\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{h}\mathrm{C}\mathrm{O}\mathrm{O})}_4{(\mathrm{N}{\mathrm{O}}_3)}_2{({\mathrm{H}}_2\mathrm{O})}_2] + 4 \mathrm{M}\mathrm{e}\mathrm{C}{\mathrm{O}}_2\mathrm{H} + 4 {{\mathrm{H}}_3\mathrm{O}}^{+} + 4 {\mathrm{N}{\mathrm{O}}_3}^{-} +16 {\mathrm{H}}_2\mathrm{O} \end{gathered}$$

(2)

                     2

$$\begin{gathered} \mathrm{C}\mathrm{d}{\left({\mathrm{O}}_2\mathrm{C}\mathrm{M}\mathrm{e}\right)}_2·2{\mathrm{H}}_2\mathrm{O}+2\mathrm{C}\mathrm{r}{\left(\mathrm{N}{\mathrm{O}}_3\right)}_3·9{\mathrm{H}}_2\mathrm{O}+4\mathrm{C}\mathrm{l}\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{h}\mathrm{H}\mathrm{C}\mathrm{O}\mathrm{O}\mathrm{H}+\mathrm{E}\mathrm{t}\mathrm{O}\mathrm{H} \stackrel{EtOH/MeCN,T}{\to } \\ [\mathrm{C}\mathrm{d}\mathrm{C}{\mathrm{r}}_2{(4\mathrm{C}\mathrm{l}\mathrm{s}\mathrm{a}\mathrm{p}\mathrm{h}\mathrm{C}\mathrm{O}\mathrm{O})}_4{({\mathrm{H}}_2\mathrm{O})}_3(\mathrm{E}\mathrm{t}\mathrm{O}\mathrm{H})]+2\mathrm{M}\mathrm{e}\mathrm{C}{\mathrm{O}}_2\mathrm{H}+6{{\mathrm{H}}_3\mathrm{O}}^{+}+6{\mathrm{N}{\mathrm{O}}_3}^{-}+11{\mathrm{H}}_2\mathrm{O} \end{gathered}$$

(3)

                     4

Some synthetic points deserve discussion: (i) heterometallic compounds containing the singly deprotonated ligands saphHCOO⁻ and 4ClsaphHCOO⁻ could not be obtained. (ii) Small changes in the reactants’ molar ratios do not affect the identity of the products, as evidenced by IR spectroscopy and microanalytical samples (for few samples only). (iii) Somewhat to our surprise, the Cd(O₂CMe)₂·2H₂O/Cr(NO₃)₃·9H₂O/4ClsaphHCOOH reaction system does not give tetranuclear clusters, but instead the trinuclear complex 4. Efforts were made to prepare a {Cd^II₂Cr^III₂} cluster employing higher Cd^II:Cr^III reaction ratios, but in vain; the product was again 4, suffice to say that this complex is the thermodynamically stable product under the conditions employed. (iv) Looking at the stoichiometric Equations (1)–(3), we observe that one of the byproducts is dilute HNO₃ (strong acid), which could decompose the products by attacking the doubly deprotonated ligands. A logical explanation for the stability of the tetranuclear complexes is that the released H₃O⁺ ions are neutralized by the MeCO₂⁻ (base) ions that are in excess in the reaction systems. The stoichiometric MeCO₂⁻:saphHCOOH/4ClsaphHCOOH molar ration indicated by (1) and (2) is 1:1 (i.e., 4:4), whereas the experimental one (vide infra) is 2:1 (i.e., 8:4); the excess of acetates act as additional effective proton acceptors giving weak MeCO₂H acid, which cannot decompose the products. This viewpoint is supported by the fact that complexes 1–3 could not be isolated by using an 1:1.5 Cd(O₂CMe)₂·2H₂O:LH₂ (i.e., MeCO₂⁻:LH₂ = 4:6) ratio. If this is correct, an alternative, more realistic description of the actual process can be represented by Equation (4), where S = CH₂Cl₂ for 1 and 3, S = MeCN for 2, LH₂ = saphHCOOH for 1 and 2, and LH₂ = 4ClsaphHCOOH for 3. This means that Fe^III and Cr^III are in significant “deficiency” (more than that resulting from the experiments), justifying the moderate yields of the reactions. Analogous remarks could be valid for 4 explaining its stability, despite the appearance of HNO₃ as byproduct in Equation (3).

$$\begin{gathered} 4\mathrm{C}\mathrm{d}{\left({\mathrm{O}}_2\mathrm{C}\mathrm{M}\mathrm{e}\right)}_2·2{\mathrm{H}}_2\mathrm{O}+2\mathrm{M}{\left(\mathrm{N}{\mathrm{O}}_3\right)}_3·9{\mathrm{H}}_2\mathrm{O}+4\mathrm{L}{\mathrm{H}}_2 \stackrel{EtOH/S}{\to } \\ [\mathrm{C}{\mathrm{d}}_2{\mathrm{M}}_2{(\mathrm{L})}_4{(\mathrm{N}{\mathrm{O}}_3)}_2{({\mathrm{H}}_2\mathrm{O})}_2]+8\mathrm{M}\mathrm{e}\mathrm{C}{\mathrm{O}}_2\mathrm{H}+2\mathrm{“}\mathrm{C}\mathrm{d}{(\mathrm{N}{\mathrm{O}}_3)}_2\mathrm{”}+24{\mathrm{H}}_2\mathrm{O} \end{gathered}$$

(4)

                     1–3

The purity of the complexes was confirmed by microanalyses and ⁵⁷Fe-Mössbauer spectra for the iron(III)-containing compounds (vide infra), which clearly show one high-spin Fe^III site.

2.2. Description of Structures

The structures of the four complexes were solved by single-crystal, X-ray crystallography. The crystals of the trinuclear {CdCr₂} cluster were poorly diffracting and twin, and thus the quality of the structure is not good (for details, see Section 3). The molecular formula of this compound is [CdCr₂(4ClsapCOO)₄(H₂O)₃(EtOH)]·xEtOH (4·xEtOH). Various structural plots are shown in Scheme 2, Scheme 3, Schemes S1 and S2 and Figure 1, Figure 2, Figure 3, Figure 4, Figure 5, Figure 6 and Figure 7 and Figures S1–S14. Crystallographic data are presented in

Table 1. Crystallographic data for complexes 1·6EtOH·2CH₂Cl₂·2H₂O, 2·2EtOH·2MeCN and 3·4.5CH₂Cl₂·4H₂O.

Parameter	1·6EtOH·2CH2Cl2·2H2O	2·2EtOH·2MeCN	3·4.5CH2Cl2·4H2O
Formula	C70H84N6O28Cl4Cd2Fe2	C64H58N8O22Cd2Cr2	C60.5H53N6O24 Cl13 Cd2Fe2
Formula weight	1935.73	1619.98	2045.44
Crystal system	triclinic	monoclinic	triclinic
Space group	P$\overline{1}$	P21/n	P$\overline{1}$
Radiation (wavelength/Å)	Mo Kα (0.71073)	Mo Kα (0.71073)	Mo Kα (0.71073)
T/°C	−83	−83	−113
a/Å	11.6608(10)	12.6281(8)	13.3233(7)
b/Å	13.9884(11)	14.1793(9)	13.8231(7)
c/Å	14.3258(12)	18.3556(12)	13.8893(7)
α/°	67.611(2)	90.00	116.159(1)
β/°	73.211(2)	97.033(2)	96.288(1)
γ/°	70.602(2)	90.00	104.868(1)
V/Å3	2002.4(3)	3262.0(4)	2144.8(2)
Z	1	2	1
Dcalcd/g cm−3	1.605	1.649	1.584
μ/mm−1	1.10	1.05	1.30
2θmax/°	54	54	54
Reflections collected	45,107	46,117	38,911
Reflections unique (Rint)	8717(0.045)	7105(0.083)	9322(0.0744)
Reflections with I > 2σ(I)	7777	5887	6887
No. of parameters	591	520	412
R1a [I > 2σ(I)]	0.0643	0.0476	0.0528
wR2b [I > 2σ(I)]	0.1741	0.1204	0.1444
GOF(F2)	1.07	1.05	1.03
Δρmax/Δρmin (e Å−3)	1.60/−1.46	1.22/−0.82	1.44/−0.61
CCDC	2,374,217	2,374,218	2,374,219

^a R₁ = Σ(|F_o|-|F_c|)/Σ(|F_c|). ^b wR₂ = { Σ[w(F_o²-F_c²)]/Σ[w(F_o²)²]}^1/2, with w = 1/σ²(F_o²) + (aP)² + (bP)], where P = (F_o² + 2F_c²)/3 [a = 0.751 and b = 10.6120 for complex 1·6EtOH·2CH₂Cl₂·2H₂O, a = 0.0707 and b = 2.5008 for complex 2·2EtOH·2MeCN, a = 0.0887 and b = 0.0 for complex 3·4.5CH₂Cl₂·4H₂O].

, while numerical data for interatomic distances, bond angles, coordination polyhedra and H-bonding interactions are listed in Table 2, Table 3 and Table 4 and Tables S2–S4, and the caption of Figure S9.

The molecular structures of the three tetranuclear complexes are similar, and thus only the structure of 1·6EtOH·2CH₂Cl₂·2H₂O will be described in detail. This cluster crystallizes in the triclinic space group P$\overline{1}$, and the asymmetric unit of the cell contains half of the tetranuclear molecule, and three EtOH, one CH₂Cl₂ and one H₂O molecules. The structure consists of [Cd₂Fe₂(saphCOO)₄(NO₃)₂(H₂O)₂], EtOH, CH₂Cl₂ and H₂O molecules in a 1:6:2:2 ratio: the latter three will not be described here, but later, in the context of its supramolecular features. The cluster molecule possesses a crystallographically imposed inversion center, which is located at the midpoint of the Cd1…Cd1′ (or Fe1…Fe1′) vector. The metal topology can be described as two isosceles triangles sharing a common edge (Cd1…Cd1′). The two Cd^II atoms are doubly bridged by μ-aqua ligands. The Cd1O1WCd1′O1W′ rhombus is perfectly planar (due to symmetry). Its sides are 2.308(4) and 2.362(4) Å, its angles are 76.8(1) and 103.2(1)°, while its long diagonal is 3.659(2) Å. The Fe^III…Cd^II sides of the triangles are each asymmetrically bridged by one carboxylato oxygen atom (O2, O5 and symmetry equivalents) of a 2.2111 (Harris notation [42]) saphCOO²⁻ ligand, Scheme 2. Thus, the core of the molecule is {Cd₂Fe₂(μ-O_aqua)₂(μ-OR)₄}⁶⁺, where the OR oxygen atoms are the bridging carboxylato oxygens of the saphCOO²⁻ ligands (Scheme 3). A slightly asymmetric, bidentate chelating nitrato group and two terminal carboxylate O atoms complete 8-coordination at each Cd^II atom. The coordination sphere of each 6-coordinate Fe^III atom is completed by two imino N atoms and two terminally ligated phenolato O atoms. The coordination spheres of the two crystallographically independent metal ions are thus [Cd(O_aqua)₂(O_carboxylato)₄(O_nitrato)₂] and [Fe(N_imino)₂(O_carboxylato)₂(O_phenolato)₂].

Table 2. Selected interatomic distances (Å) and bond angles (^o) for complex 1·6EtOH·2CH₂Cl₂·2H₂O ^a.

Interatomic Distances	(Å)	Bond Angles	(°)
Cd1…Cd1′	3.659(2)	O1W–Cd1–O1W′	76.8(1)
Fe1…Fe1′	8.122(2)	O1′–Cd1–O4	82.3(1)
Cd1…Fe1	4.371(2)	O1′–Cd1–O1W	133.5(1)
Cd1…Fe1′	4.535(2)	O1′–Cd1–O1W′	91.2(1)
Cd1–O1′	2.285(4)	O4–Cd1–O5	52.8(1)
Cd1–O4	2.302(4)	O4–Cd1–O9	112.7(2)
Cd1–O1W	2.308(4)	O5–Cd1–O1W′	77.1(1)
Cd1–O1W′	2.362(4)	O5–Cd1–O7	75.5(1)
Cd1–O5	2.593(3)	O5–Cd1–O9	127.8(1)
Cd1–O7	2.449(5)	O7–Cd1–O9	52.7(2)
Cd1–O9	2.379(4)	O2–Fe1–N2	166.2(2)
Cd1–O2′	2.731(4)	O3–Fe1–O6	163.1(2)
Fe1–O2	2.002(3)	O5–Fe1–N1	166.0(2)
Fe1–O3	1.952(4)	O7–N3–O9	116.6(5)
Fe1–O5	2.010(3)	O7–N3–O8	123.0(6)
Fe1–O6	1.950(4)	O8–N3–O9	120.4(6)
Fe1–N1	2.131(4)	Cd1–O1W–Cd1′	103.2(1)
Fe1–N2	2.139(4)

^a Symmetry code: (′) −x + 2, −y + 2, −z.

Table 2. Selected interatomic distances (Å) and bond angles (^o) for complex 1·6EtOH·2CH₂Cl₂·2H₂O ^a.

Interatomic Distances	(Å)	Bond Angles	(°)
Cd1…Cd1′	3.659(2)	O1W–Cd1–O1W′	76.8(1)
Fe1…Fe1′	8.122(2)	O1′–Cd1–O4	82.3(1)
Cd1…Fe1	4.371(2)	O1′–Cd1–O1W	133.5(1)
Cd1…Fe1′	4.535(2)	O1′–Cd1–O1W′	91.2(1)
Cd1–O1′	2.285(4)	O4–Cd1–O5	52.8(1)
Cd1–O4	2.302(4)	O4–Cd1–O9	112.7(2)
Cd1–O1W	2.308(4)	O5–Cd1–O1W′	77.1(1)
Cd1–O1W′	2.362(4)	O5–Cd1–O7	75.5(1)
Cd1–O5	2.593(3)	O5–Cd1–O9	127.8(1)
Cd1–O7	2.449(5)	O7–Cd1–O9	52.7(2)
Cd1–O9	2.379(4)	O2–Fe1–N2	166.2(2)
Cd1–O2′	2.731(4)	O3–Fe1–O6	163.1(2)
Fe1–O2	2.002(3)	O5–Fe1–N1	166.0(2)
Fe1–O3	1.952(4)	O7–N3–O9	116.6(5)
Fe1–O5	2.010(3)	O7–N3–O8	123.0(6)
Fe1–O6	1.950(4)	O8–N3–O9	120.4(6)
Fe1–N1	2.131(4)	Cd1–O1W–Cd1′	103.2(1)
Fe1–N2	2.139(4)

^a Symmetry code: (′) −x + 2, −y + 2, −z.

Scheme 2. The coordination modes (using Harris notation) of the saphCOO²⁻ and 4ClsaphCOO²⁻ ligands in complexes 1–4. M = Fe, Cr and X = H, Cl. The coordination bonds are drawn with bold lines.

Scheme 2. The coordination modes (using Harris notation) of the saphCOO²⁻ and 4ClsaphCOO²⁻ ligands in complexes 1–4. M = Fe, Cr and X = H, Cl. The coordination bonds are drawn with bold lines.

Scheme 3. The labeled core of 1·6EtOH·2CH₂Cl₂·2H₂O, emphasizing the metal topology, which consists of two isosceles triangles sharing the common Cd1…Cd1′ edge. The lines between the metal ions do not represent real bonds. The coordination bonds are drawn with bold lines. The primes (′) are used for atoms generated by the symmetry operation −x + 2, −y + 2, −z.

Scheme 3. The labeled core of 1·6EtOH·2CH₂Cl₂·2H₂O, emphasizing the metal topology, which consists of two isosceles triangles sharing the common Cd1…Cd1′ edge. The lines between the metal ions do not represent real bonds. The coordination bonds are drawn with bold lines. The primes (′) are used for atoms generated by the symmetry operation −x + 2, −y + 2, −z.

The Cd1…Cd1′ distance [3.659(2) Å] is short because of the presence of two monoatomic bridges between these metal ions. The Cd–O distances are in the wide range 2.302(4)–2.731(4) Å. These bond lengths are unremarkable [43], except the 2.731(4) Å one, which is long and can be considered as a bonding interaction (and not a real bond). The Cd1–O5 and Cd1–O2′ distances for the bridging carboxylato O atoms are larger [2.593(3) and 2.731(4) Å] than the distances exhibited by the terminal carboxylato oxygens [Cd1-O4 = 2.302(4) and Cd1–O1′ = 2.285(4) Å] to the same metal ion. The increase in bond distance upon bridging relative to terminal coordination is common in complexes containing carboxylato ligands with one bridging O atom. Theoretical and experimental studies [44,45,46] have revealed that the syn lone pairs of the carboxylate functional group are more basic than the anti lone pairs. Due to this trend, we might expect that the Cd1–O5 and Cd1–O2′ (and symmetry equivalents) distances would be shorter than the Cd1–O4 and Cd1–O1′ (and symmetry equivalents) bond lengths; however, the opposite trend is observed in 1·6EtOH·2CH₂Cl₂·2H₂O. This result, which is typical for other Cd^II carboxylate complexes [43,47], suggests that the Cd-O bond lengths involving η¹:η²:μ (2.21) carboxylato groups are mainly dependent on geometric rather than electronic properties of the carboxylato group [43]. The Fe1/Fe1′–O/N bond lengths are typical for high-spin iron(III) atoms in an octahedral coordination environment [28]. The Fe1/Fe1′ geometry is distorted octahedral, with the trans angles being in the narrow 163.1(2)–166.2(2)° range. Of the accessible eight-coordinate polyhedra of metal centers, the biaugmented trigonal prism J50 is the most appropriate for the description of the donor atoms around Cd1/Cd1′ (Figure 2, Table S1) [48]. The rather large distortion (the CShM value is 4.40, very close to that of the normal biaugmented trigonal prism) is primarily a consequence of the small bite angle imposed by the chelating nitrato group.

Figure 2. The biaugmented trigonal prismatic (J50) coordination geometry of Cd1/Cd1′ in the structure of 1·6EtOH·2CH₂Cl₂·2H₂O, see also Table S1. The plotted polyhedron represents the ideal, best-fit one using the program SHAPE. Primed atoms are generated by the symmetry operation −x + 2, −y + 2, −z.

Figure 2. The biaugmented trigonal prismatic (J50) coordination geometry of Cd1/Cd1′ in the structure of 1·6EtOH·2CH₂Cl₂·2H₂O, see also Table S1. The plotted polyhedron represents the ideal, best-fit one using the program SHAPE. Primed atoms are generated by the symmetry operation −x + 2, −y + 2, −z.

Compound 3·4.5CH₂Cl₂·4H₂O (Table 1, Table 3 and Table S2; Figure 3 and Figures S3–S5) is isostructural with 1·6EtOH·2CH₂Cl₂·2H₂O, but not isomorphous. As is apparent, the centrosymmetric molecules 1 and 3 are very similar (for convenience, an almost analogous atom labeling scheme has been adopted, with the exception of the nitrato oxygens). Again, the coordination spheres are [Cd1O₈] and [Fe1N₂O₄], and the core is {Cd₂Fe₂(μ-O_aqua)₂(μ-OR)₄}⁶⁺. Thus, it seems that the Cl atom of the iminobenzoato ring has no structural effect.

Figure 3. Partially labeled plot of the structure of the molecule [Cd₂Fe₂(4ClsaphCOO)₄(NO₃)₂(H₂O)₂] that is present in 3·4.5CH₂Cl₂·4H₂O. Thermal ellipsoids are shown at the 50% probability level. A plot without thermal ellipsoids is presented in Figure S3. Symmetry code: (‴) −x + 1, −y + 2, −z.

Figure 3. Partially labeled plot of the structure of the molecule [Cd₂Fe₂(4ClsaphCOO)₄(NO₃)₂(H₂O)₂] that is present in 3·4.5CH₂Cl₂·4H₂O. Thermal ellipsoids are shown at the 50% probability level. A plot without thermal ellipsoids is presented in Figure S3. Symmetry code: (‴) −x + 1, −y + 2, −z.

Table 3. Selected interatomic distances (Å) and bond angles (°) for complex 3·4.5CH₂Cl₂·4H₂O ^a.

Interatomic Distances	(Å)	Bond Angles	(°)
Cd1…Cd1‴	3.703(2)	O1W–Cd1–O1W‴	75.9(1)
Fe1…Fe1‴	8.007(1)	O1‴–Cd1–O4	83.9(1)
Cd1…Fe1	4.460(1)	O1‴–Cd1–O1W	96.0(1)
Cd1…Fe1‴	4.361(2)	O4–Cd1–O1W	93.3(1)
Cd1–O1‴	2.333(3)	O4–Cd1–O7	82.8(1)
Cd1–O4	2.302(2)	O4–Cd1–O8	122.6(1)
Cd1–O1W	2.345(3)	O7–Cd1–O8	52.0(1)
Cd1–O1W‴	2.352(3)	O7–Cd1–O1W	148.7.(1)
Cd1–O2‴	2.589(2)	O7–Cd1–O1W‴	85.1(1)
Cd1–O5	2.668(3)	O8—Cd1–O1W	143.9(1)
Cd1–O7	2.396(4)	O2–Fe1–N2	169.0(1)
Cd1–O8	2.424(3)	O3–Fe1–O6	163.1(2)
Fe1–O2	1.997(2)	O5–Fe1–N1	169.6(1)
Fe1–O3	1.957(2)	O7–N3–O8	117.0(4)
Fe1–O5	2.001(2)	O7–N3–O9	120.6(5)
Fe1–N1	2.117(3)	O8–N3–O9	122.2(5)
Fe1–N2	2.124(3)	Cd1–O1W–Cd1‴	104.1(1)

^a Symmetry code: (‴) −x + 1, −y + 2, −z.

Table 3. Selected interatomic distances (Å) and bond angles (°) for complex 3·4.5CH₂Cl₂·4H₂O ^a.

Interatomic Distances	(Å)	Bond Angles	(°)
Cd1…Cd1‴	3.703(2)	O1W–Cd1–O1W‴	75.9(1)
Fe1…Fe1‴	8.007(1)	O1‴–Cd1–O4	83.9(1)
Cd1…Fe1	4.460(1)	O1‴–Cd1–O1W	96.0(1)
Cd1…Fe1‴	4.361(2)	O4–Cd1–O1W	93.3(1)
Cd1–O1‴	2.333(3)	O4–Cd1–O7	82.8(1)
Cd1–O4	2.302(2)	O4–Cd1–O8	122.6(1)
Cd1–O1W	2.345(3)	O7–Cd1–O8	52.0(1)
Cd1–O1W‴	2.352(3)	O7–Cd1–O1W	148.7.(1)
Cd1–O2‴	2.589(2)	O7–Cd1–O1W‴	85.1(1)
Cd1–O5	2.668(3)	O8—Cd1–O1W	143.9(1)
Cd1–O7	2.396(4)	O2–Fe1–N2	169.0(1)
Cd1–O8	2.424(3)	O3–Fe1–O6	163.1(2)
Fe1–O2	1.997(2)	O5–Fe1–N1	169.6(1)
Fe1–O3	1.957(2)	O7–N3–O8	117.0(4)
Fe1–O5	2.001(2)	O7–N3–O9	120.6(5)
Fe1–N1	2.117(3)	O8–N3–O9	122.2(5)
Fe1–N2	2.124(3)	Cd1–O1W–Cd1‴	104.1(1)

^a Symmetry code: (‴) −x + 1, −y + 2, −z.

Compound 2·2EtOH·2MeCN crystallizes in the monoclinic space group P2₁/n, and the asymmetric unit of the cell contains half of the tetranuclear molecule, one EtOH molecule and one MeCN molecule. Thus, the cluster consists of centrosymmetric [Cd₂Cr₂(saphCOO)₄(NO₃)₂(H₂O)₂], EtOH and MeCN molecules in a 1:2:2 ratio. The molecule 2 has a rather similar structure (Figure 4, Figures S6 and S8; Scheme S1; Table 4 and Table S3) to that of 1, the main difference being the replacement of the Fe^III atoms in the latter by the Cr^III atoms in the former. A minor difference exists in the coordination geometry of the Cd^II atoms. This is the best described as biaugmented trigonal prismatic in 1 (Figure 2, Table S1), and as triangular dodecahedral (Figure S7, Table S3) in 2. Since the geometries are distorted and the CShM values are close, the polyhedra in the two complexes can be considered as intermediate between the biaugmented trigonal prism and the triangular dodecahedron. The Cr–N/O bond distances are typical [28] for octahedral chromium (III) complexes.

Figure 4. Partially labeled plot of the structure of the molecule [Cd₂Cr₂(saphCOO)₄(NO₃)₂(H₂O)₂] that is present in 2·2EtOH·2MeCN. Thermal ellipsoids are shown at the 50% probability level. A plot without ellipsoids is presented in Figure S6. Symmetry code: (″) −x + 1, −y, −z.

Figure 4. Partially labeled plot of the structure of the molecule [Cd₂Cr₂(saphCOO)₄(NO₃)₂(H₂O)₂] that is present in 2·2EtOH·2MeCN. Thermal ellipsoids are shown at the 50% probability level. A plot without ellipsoids is presented in Figure S6. Symmetry code: (″) −x + 1, −y, −z.

Table 4. Selected interatomic distances (Å) and bond angles (°) for complex 2·2EtOH·2MeCN ^a.

Interatomic Distances	(Å)	Bond Angles	(°)
Cd1…Cd1″	3.946(2)	O1W–Cd1–O1W″	72.7(1)
Cr1…Cr1″	7.321(1)	O1″–Cd1–O4	79.8(1)
Cd1…Cr1	4.080(1)	O1″–Cd1–O1W	141.6(1)
Cd1…Cr1″	4.235(2)	O2″–Cd1–O5	148.2(1)
Cd1–O1″	2.270(3)	O4–Cd1–O5	54.0(1)
Cd1–O4	2.417(2)	O4–Cd1–O8	110.5(1)
Cd1–O1W	2.343(3)	O5–Cd1–O7	83.3(1)
Cd1–O1W″	2.554(3)	O5–Cd1–O8	134.5.(1)
Cd1–O5	2.379(2)	O7–Cd1–O8	51.7(1)
Cd1–O7	2.459(3)	O8–Cd1–O1W	91.8(1)
Cd1–O8	2.377(4)	O3–Cr1–O6	178.0(1)
Cd1–O2″	2.514(2)	O2–Cr1–N2	173.3(1)
Cr1–O2	1.979(2)	O5–Cr1–N1	176.8(1)
Cr1–O3	1.946(2)	O3–Cr1–O5	88.6(2)
Cr1–O5	2.001(2)	O7–N3–O8	115.6(4)
Cr1–O6	1.931(2)	O7–N3–O9	121.7(4)
Cr1–N1	2.007(3)	O8–N3–O9	122.7(4)
Cr1–N2	2.005(5)	Cd1–O1W–Cd1″	102.3(1)

^a Symmetry code: (″) −x + 1, −y, −z.

Table 4. Selected interatomic distances (Å) and bond angles (°) for complex 2·2EtOH·2MeCN ^a.

Interatomic Distances	(Å)	Bond Angles	(°)
Cd1…Cd1″	3.946(2)	O1W–Cd1–O1W″	72.7(1)
Cr1…Cr1″	7.321(1)	O1″–Cd1–O4	79.8(1)
Cd1…Cr1	4.080(1)	O1″–Cd1–O1W	141.6(1)
Cd1…Cr1″	4.235(2)	O2″–Cd1–O5	148.2(1)
Cd1–O1″	2.270(3)	O4–Cd1–O5	54.0(1)
Cd1–O4	2.417(2)	O4–Cd1–O8	110.5(1)
Cd1–O1W	2.343(3)	O5–Cd1–O7	83.3(1)
Cd1–O1W″	2.554(3)	O5–Cd1–O8	134.5.(1)
Cd1–O5	2.379(2)	O7–Cd1–O8	51.7(1)
Cd1–O7	2.459(3)	O8–Cd1–O1W	91.8(1)
Cd1–O8	2.377(4)	O3–Cr1–O6	178.0(1)
Cd1–O2″	2.514(2)	O2–Cr1–N2	173.3(1)
Cr1–O2	1.979(2)	O5–Cr1–N1	176.8(1)
Cr1–O3	1.946(2)	O3–Cr1–O5	88.6(2)
Cr1–O5	2.001(2)	O7–N3–O8	115.6(4)
Cr1–O6	1.931(2)	O7–N3–O9	121.7(4)
Cr1–N1	2.007(3)	O8–N3–O9	122.7(4)
Cr1–N2	2.005(5)	Cd1–O1W–Cd1″	102.3(1)

^a Symmetry code: (″) −x + 1, −y, −z.

As mentioned earlier, the X-ray data set for 4·xEtOH was poor. However, we were able to see its gross molecular structure (Figure 5 and Figures S9–S11; Scheme S2). The compound crystallizes in the triclinic space group P$\overline{1}$, and the asymmetric unit of the cell contains one [CdCr₂(4ClsaphCOO)₄(H₂O)₃(EtOH)] molecule and lattice EtOH molecules. The metal topology is triangular, and the three metal ions define an isosceles triangle, the Cd1…Cr1, Cd1…Cr2 and Cr1…Cr2 distances being 4.34, 4.25 and 6.85 Å, respectively. The Cd1 center is linked to each Cr^III atom through one bridging carboxylato oxygen (O5 and O8 for Cr1 and Cr2, respectively) of a 2.2111 4ClsaphCOO²⁻ ligand (Scheme 2, left; M = Cr, and X = Cl). Three aqua groups, one EtOH molecule and the two terminally ligated carboxylato oxygen atoms (O4, O7) from two 2.2111 ligands complete eight-coordination at cadmium(II). Thus, the coordination sphere of Cd1 is [Cd1(O_aqua)₃(O_ethanol)(O_{bridging carboxylato})₂(O_{terminal carboxylato})₂]. A terminal 1.1101 4ClsaphCOO²⁻ ligand (Scheme 2, right) and the N_imino, O_phenolato pair of an 2.2111 ligand complete a coordination number of six at each chromium(III). Thus, the coordination sphere of each Cr^III atom is [Cr(O_{bridging carboxylato})(O_{terminal carboxylato})(O_phenolato)₂(N_imino)₂]. The core of the molecule is {CdCr₂(μ-OR)₂}⁶⁺, where the OR oxygen atoms are the bridging carboxylato oxygens of the 2.2111 4ClsaphCOO²⁻ ligands.

Figure 5. Partially labeled plot of the molecule [CdCr₂(4ClsaphCOO)₄(H₂O)₃(EtOH)] that is present in the structure of 4·xEtOH. Thermal ellipsoids are shown at the 50% probability level. Selected interatomic distances and bond angles are listed in the caption of Figure S9, in which the plot is without thermal ellipsoids. Note: This plot has resulted from a poorly diffracting single-crystal.

Figure 5. Partially labeled plot of the molecule [CdCr₂(4ClsaphCOO)₄(H₂O)₃(EtOH)] that is present in the structure of 4·xEtOH. Thermal ellipsoids are shown at the 50% probability level. Selected interatomic distances and bond angles are listed in the caption of Figure S9, in which the plot is without thermal ellipsoids. Note: This plot has resulted from a poorly diffracting single-crystal.

We now discuss the intermolecular interactions in the tetranuclear {Cd₂M₂} (M = Fe, Cr) clusters. Plots are shown in Figure 6, Figure 7 and Figures S12–S14, while numerical data are listed in Table S4.

Figure 6. (a) π–π interactions contributing in the formation of layers in the crystal structure of compound 1·6EtOH·2CH₂Cl₂·2H₂O. Dashed yellow and cyan lines indicate the interactions between the phenyl pairs C15C17…C21/C23C24…C28 and C2C3…C7/C9C10…C14, respectively (see also Figure 1). The C6–H(C6)…Cl2, C20–H(C20)…Cl1 and O1W–H1(O1W)…O10, O10–H(O10)…O3 H-bonding interactions involving the CH₂Cl₂ and EtOH lattice solvents are indicated with violet and orange colors, respectively. The C35–H_A(C35)…O8 and C35–H_B(C35)…O8 H bonds involving CH₂Cl₂ are also indicated with violet color. (b) The O1W–H2(O1W)…O2W, O2W–H1(O2W)…O6, O2W–H2(O2W)…O11, O11–H(O11)…O12, O12–H1(O12)…·O1, C33–H_A(C33)…O4 and C34–H_C(C34)…O9 chain of H bonds connecting molecules that belong to different layers.

Figure 6. (a) π–π interactions contributing in the formation of layers in the crystal structure of compound 1·6EtOH·2CH₂Cl₂·2H₂O. Dashed yellow and cyan lines indicate the interactions between the phenyl pairs C15C17…C21/C23C24…C28 and C2C3…C7/C9C10…C14, respectively (see also Figure 1). The C6–H(C6)…Cl2, C20–H(C20)…Cl1 and O1W–H1(O1W)…O10, O10–H(O10)…O3 H-bonding interactions involving the CH₂Cl₂ and EtOH lattice solvents are indicated with violet and orange colors, respectively. The C35–H_A(C35)…O8 and C35–H_B(C35)…O8 H bonds involving CH₂Cl₂ are also indicated with violet color. (b) The O1W–H2(O1W)…O2W, O2W–H1(O2W)…O6, O2W–H2(O2W)…O11, O11–H(O11)…O12, O12–H1(O12)…·O1, C33–H_A(C33)…O4 and C34–H_C(C34)…O9 chain of H bonds connecting molecules that belong to different layers.

The main characteristic of the packing in the three complexes is the formation of layers through π–π interactions. Figure 6a,b illustrates the intermolecular intra- and interlayer interactions, respectively, for complex 1·6EtOH·2CH₂Cl₂·2H₂O. Figure S12 shows the layer formed by the molecules through π–π interactions parallel to the (100) plane for this compound. The centroid…centroid distance between phenyl ring planes defined by atoms C16C17…C21 and C23C24…C28 (−x + 2, −y + 2, −z + 1) is 4.528(1) Å, and the angle between them is 34.3(2)°; the respective parameters for the phenyl rings defined by atoms C2C3…C7 and C9C10…C14 (−x + 2, −y + 3, −z) are 4.353(1) Å and 25.4(2)°. CH₂Cl₂ and EtOH solvent molecules are H-bonded to the molecules that form the layers through C6–H(C6)…Cl2, C20–H(C20)…Cl1, O1W–H1(O1W)…O10 and O10–H(O10)…O3 interactions (Figure 6a, Table S4). The CH₂Cl₂ molecules interact further with the non-coordinated nitrato oxygens O8/O8′ through their methylene H atoms, i.e., C35–H_A(C35)…O8 and C35–H_B(C35)…O8 (Figure 6a, Table S4). Neighboring layers, as the one shown in Figure S12, are stacked along the a axis through a chain of O1W–H2(O1W)…O2W, O2W-H1(O2W)…O6, O2W–H2(O2W)…O11 and O11–H(O11)…O12 H bonds, where O2W is the oxygen atom of the lattice H₂O molecule, and O11 and O12 are the oxygens of the lattice EtOH molecules. The O12-containing EtOH molecule is H-bonded to a cluster molecule of the neighboring layer through the O12–H(O12)…O1, C33–H_A(C33)…O4 and C34-H_C(C34)…O9 H bonds (Figure 6b; Table S4).

Figure 7. Intramolecular and intralayer H bonds observed in the crystal structure of complex 2·2EtOH·2MeCN. The intramolecular H bond O1W–H1(O1W)…O6 is indicated with dashed light green lines. The O10–H(O10)…O3 and O1W–H2(O1W)…O10 H bonds are indicated with dashed orange lines, while the dashed violet lines represent the C20–H(C20)…N4 and C22–H(C22)…N4 H bonds.

Figure 7. Intramolecular and intralayer H bonds observed in the crystal structure of complex 2·2EtOH·2MeCN. The intramolecular H bond O1W–H1(O1W)…O6 is indicated with dashed light green lines. The O10–H(O10)…O3 and O1W–H2(O1W)…O10 H bonds are indicated with dashed orange lines, while the dashed violet lines represent the C20–H(C20)…N4 and C22–H(C22)…N4 H bonds.

In compound 2·2EtOH·2MeCN, the lattice EtOH molecules participate in H bonds through the O10–H(O10)…O3 and O1W–H2(O1W)…O10 interactions (O10 is the EtOH oxygen atom). The purely intramolecular O1W-H1(O1W)…O6 H bond also gives extra stabilization in the tetranuclear cluster molecule. The MeCN molecules are also H-bonded to the cluster molecules through the C20–H(C20)…N4 and C22–H(C22)…N4 interactions. C20 and C22 are aromatic carbon atoms of saphCOO²⁻. All the abovementioned characteristics are illustrated in Figure 7. A layer parallel to the (−101) plane through π–π interactions is shown in Figure S13. The centroid–centroid distance for the planes of the phenyl rings defined by atoms C2C3…C7 and C16C17…C21 (−x + 3/2, y − 1/2, −z + 1/2) is 4.291(1) Å and the angle between them is 28.6(1)°; the respective parameters for the phenyl rings defined by atoms C9C10…C14 and C23C24…C28 (−x + 3/2, y − 1/2, −z + 1/2) are 4.245(1) Å and 15.0(1)°.

In complex 3·4.5CH₂Cl₂·4H₂O, due to the application of the squeeze procedure (vide infra), only the formation of layers through π–π interactions is discussed. A layer parallel to (100) through π–π interactions is shown in Figure S14. In this case, only the phenyl rings without the Cl atom overlap. The centroid-centroid distance for the planes of the centrosymmetrically related phenyl rings defined by atoms C9C10…C14 and C9C10…C14 (−x + 1, −y + 2, −z + 1) is 4.578(1) Å; the respective parameter for the planes of the centrosymmetrically related phenyl rings defined by atoms C23C24…C28 (−x + 1, −y + 1, −z) is 4.754(1) Å.

The centroid–centroid distances in the structures are of the offset type and fall within the range of the usually observed corresponding distances of π–π interactions [49,50].

Complexes 1–4 are members of a very small group of structurally characterized heterometallic complexes based on anionic forms of saphHCOOH and 4ClsaphHCOOH. The previous members were {Zn^II₄Fe^III₂} and {Zn^II₄Cr^III₂} clusters, which also contain the doubly deprotonated ligands [28]. {Cd^II₂Cr^III₂} and {Cd^IICr^III₂} nuclearities have never been observed in discrete coordination clusters. The nuclearity {Cd^II₂Fe^III₂} has only been found in the complex [Cd₂Fe₂OI₄L₂], where L is the dianion 6,6′-(1E,1E′)(ethane-1,2diylbis(azanelyidene))bis(methaneylylidene))bis(2-ethoxyphenol) [33]; however, the metal topology, consisting of two dinuclear Cd^IIFe^III moieties connected by a μ-oxido group, is different compared to that in 1 and 3. Discrete {Cd^IICr⁰₂} [51] and {Cd^II₂Fe^II₂} [52], i.e., with the 3d metal in zero or +II oxidation states, are known in organometallic chemistry.

2.3. Conventional Spectroscopic Characterization in Brief

IR, Raman and UV/VIS spectra of the complexes were obtained on well-dried, analytically pure (Section 3) samples. Selected spectra are shown in Figure 8 and Figures S15–S22. The weak-to-medium intensity IR band in the 3570–3470 cm⁻¹ region is assigned to the ν(OH) vibration of the coordinated and lattice (complexes 1, 3) or only coordinated (2, 4) H₂O molecules [28]; the broadness of the band is indicative of H bonding. The spectra of 3 and 4 show a second well-resolved band at ~3265 cm⁻¹, which may be due to the first overtone mode of the band at ca. 1610 cm⁻¹, with the possible contribution of ν(OH)_{coord. EtOH} in 4. This mode is hardly seen in the Raman spectra. The IR spectra exhibit a strong band at 1608–1593 cm⁻¹ which can be safely assigned to the ν(C=N) vibration of the imino group, most probably overlapping with an aromatic stretch [28]. This spectral mode appears at 1611, 1581, 1612 and 1574 cm⁻¹ in the Raman spectra of 1, 2, 3 and 4, respectively; it is obvious that the 3d-metal ion affects this mode, shifting the corresponding peak to lower wavenumbers in the Cr^III complexes. The IR bands at 1483–1445 and 1317–1290 cm⁻¹ are assigned [53] to the ν₁(A₁)[ν(Ν=O)] and ν₅(Β₂)[ν_as(NO₂)] modes, respectively, of the nitrato group, assuming C_2ν symmetry; their separation (165 cm⁻¹ for 1, 155 cm⁻¹ for 2 and 166 cm⁻¹ for 3) is large, indicating a bidentate nitrato group [53] (as confirmed by crystallography). These bands are absent in the IR spectrum of 4, which does not contain nitrates. The nitrato peaks appear at 1468 and 1318 cm⁻¹ for 1, at 1440 and 1274 cm⁻¹ for 2 and at 1486, 1317 cm⁻¹ for 3 in their Raman spectra. The location of the carboxylato stretching bands [ν_as(CO₂), ν_s(CO₂)] in the expected 1600–1300 cm⁻¹ region is difficult due to the appearance of the aromatic and nitrato stetching bands. We tentatively assign [53] the IR bands at 1540–1521 and 1384–1381 cm⁻¹ in the spectra of 1–3 to the ν_as(CO₂) and ν_s(CO₂) vibrations, respectively. Their large difference (156–140 cm⁻¹) is indicative of the bidentate bridging coordination of the carboxylate groups of saphCOO²⁻ and 4ClsaphCOO²⁻. The ν_as(CO₂) and ν_s(CO₂) Raman peaks appear at 1550–1530 and 1374–1356 cm⁻¹, respectively; again, the large separations indicates bidentate bridging ligation [53]. The simultaneous presence of two types of carboxylato ligation (monodentate, bidentate bridging) in 4 renders assignments risky.

The complexes are soluble in the non-donor solvent CH₂Cl₂. This solubility, combined with the negligible value of molar conductivity (Λ_M = 2–6 S cm² mol⁻¹) in this solvent, suggest that the compounds do not decompose in solution. Thus, the ideal solvent to record UV/VIS spectra was CH₂Cl₂. The spectra of the {Cd^II₂Fe^III₂} complexes 1 and 3 are similar. The bands in the 300–220 nm region can be assigned to transitions of the aromatic rings, with the longest wavelength maximum possibly also having a –CH=N– character [54]. The bands at ca. 360–390, 440 and 520 nm are due to ligand-to-metal charge transfer transitions (LMCT, where M = Fe^III) [55]; the transitions are from the π orbitals of the organic ligands saphCOO²⁻ and 4ClsaphCOO²⁻ to the t_2g orbitals of Fe^III. Since iron(III) is an oxidizing metal ion, these LMCT bands obscure the very weak d–d transitions. The very low-intensity band at ca. 590 nm could be due to the ⁶A_1g → ⁴T_2g transition of the high-spin iron(III) in an octahedral N,O-environment [28,55]. The spectra of 2 and 4 are rather similar, partly due to the common {Cr^IIIO₄N₂} chromophore in the two complexes. Chromium(III) is neither a good oxidizing nor a good reducing agent; therefore, the charge-transfer bands do not generally obscure the three spin-allowed d–d transitions in an octahedral field. The d–d bands are located at ~290, ~445 and ~520 nm and assigned [55] to the ⁴A_2g → ⁴T_1g(P), ⁴A_2g → ⁴T_1g(F) and ⁴A_2g → ⁴T_2g transitions, respectively, in an octahedral 3d³ crystal field. The possibility that the band at ~290 nm has a mixed –CH=N–/d–d character cannot be ruled out. A band at ~590 nm is clearly visible in the spectra of both complexes and is assigned to the ⁴A_2g → ²T_2g, ²E_g spin-forbidden transition [55]. In the spectrum of 4, a well-defined band is seen at 320 nm, which can be tentatively assigned to an LMCT transition. The longest wavelength spin-allowed d–d transition at ~520 nm leads directly to the e_g-t_2g gap (10 Dq), which is ~19,200 cm⁻¹. This value is typical for octahedral Cr^III complexes with O/N ligation [28,55].

2.4. Mössbauer Spectra

Zero-field ⁵⁷Fe-Mössbauer spectra of well-dried (analytically pure) samples of 1 and 3 were recorded at 295 or 300 and 80 K (Figure 9 and Figure S23). Both samples exhibit similar behavior, a consequence of their similar structures. At 295 K, the spectrum of 3 (Figure 9) consists of a broad asymmetric doublet. The doublet can be simulated with an isomer shift (δ) value of 0.40(2) mm s⁻¹ and a quadrupole splitting value (ΔE_Q) of 0.94(3) mm s⁻¹. These values are consistent with a high-spin iron(III) in an N/O-octahedral environment [28,56,57]. At 80 K, the doublet becomes broader with δ = 0.48(2) mm s⁻¹ and ΔE_Q = 0.92(3) mm s⁻¹. The increase in δ is attributed to the second-order Doppler effect [57]. The broadness of the spectra is due to relaxation effects, as it is often observed in isolated high-spin Fe^III atoms in the solid state. Such a behavior was also observed in the spectra of the recently studied {Zn^II₄Fe^III₂} clusters with the saphCOO²⁻ and 4ClsaphCOO²⁻ Schiff-base ligands [28], in which the iron(III) sites are well separated. As the temperature decreases, the spin–lattice relaxation ratio decreases. This decrease in the relaxation rate leads to the conversion of the Mössbauer spectrum from a doublet to sextet(s); this conversion is incomplete at 80 K in the present case.

The ⁵⁷Fe-Mössbauer spectra of 1 (Figure S23) are broader and less resolved than those of 3. In any case, we may assume that the spectra consist of an asymmetric doublet (as in 3), but an accurate and independent determination of the parameters is less reliable. Nevertheless, a fit can be obtained by assuming an asymmetric doublet with parameters similar to those of 3. This assumption is justified based on the similarity of the environment around Fe^III centers as revealed by crystallography. A satisfactory fit is obtained by assuming broader lines for the spectra of 1. As in the case of 3, we attribute the Mössbauer behavior of complex 1 to spin relaxation effects.

In summary, in both clusters, Mössbauer spectroscopy indicates that the Fe^III ions are well isolated. This is in agreement with the crystal structures, which give intramolecular Fe^III…Fe^III distances of 8.122(2) and 8.007(1) Å for 1 and 3, respectively. Also, the closest intermolecular Fe^III…Fe^III distances are large [8.476(1) Å for 1 and 8.195(1) Å for 3]. This conclusion is further supported by the magnetic susceptibility data (vide infra).

2.5. Magnetochemistry

Direct current (dc) magnetic susceptibility data (χ) on dried analytically pure samples of clusters 1 and 2 were collected in the 1.8–310 K range in an applied field of 1 kOe. Data are presented in Figure 10, Figure 11 and Figures S24–S27. We present the results using the emu convention.

Magnetization (M) data for 1 at 1.8 K (Figure S24) approach saturation at 50 kOe, with a value of ~51,700 emu/mol, in good agreement with the expected value of ~56,700 emu per mole for two independent S = 5/2 Fe^III atoms and g ~ 2.00. The χ value increases monotonically with decreasing absolute temperature (T) and shows no sign of a maximum down to 1.8 K (Figure S25). A Curie–Weiss fit to the data (Figure 10) between 10 and 310 K yields a Curie constant of 9.30 emu K/mol Oe, in agreement with the expected value for two high-spin Fe^III centers, and a Weiss constant of zero within experimental error [θ = 0.007(14) K]. The slight deviations in χΤ seen at the lowest temperatures (Figure 10) might be indicative of a very weak Fe^III…Fe^III interaction, but are most likely due to a trace impurity given the zero value of θ exhibited by the high-temperature data.

Magnetization data for 2 at 1.8 K approach saturation at 50 kOe (Figure S26) with a value of ~32,000 emu/mol, in good agreement with the expected value of ~34,000 emu/mol for six unpaired electrons, i.e., two Cr^III atoms, and a g value of ~2.00. As observed for 1, the χ value increases monotonically with decreasing T and shows no sign of a maximum (Figure S27). The data in the 10–310 K range were fit to the Curie–Weiss law (Figure 11), yielding a Curie constant of 3.754 emu K/mol Oe, in good agreement with that expected for two S = 3/2 centers and a Weiss constant θ = −0.33(4) K; the data indicate a lack of significant Cr^III…Cr^III interactions in the sample. The slight downturn in the χΤ product at the lowest temperatures (Figure 11) may be indicative of very weak intra- or/and intermolecular antiferromagnetic Cr^III…Cr^III magnetic exchange interactions, or single-ion anisotropy, but lower temperature data (<1.8 K) would be required to attempt to distinguish between them. Attempts to fit the χT vs. T data to a single-ion anisotropy model for S = 3/2 with a Curie-Weiss correction for potential interactions (the solid line in Figure 11) yielded a Curie constant of 3.74(2) emu K/mol Oe, D = 1.0(1) K and θ = −0.15(1) K, in very good agreement with the Curie–Weiss fit to the 1/χ vs. T data. Given the magnitude of both D and θ, it is inadvisable to attempt to attribute the minor deviation at the lowest temperatures to either single-ion anisotropy or antiferromagnetic exchange.

The negligible (if any) magnetic coupling for the two complexes studied was expected considering the long (>8 Å, vide supra) intra- and intermolecular distances between the spin centers. However, there have been cases in which appreciable coupling has been observed for complexes with Cr^III…Cr^III distances of ca. 7 Å [58].

We did not consider it necessary to study the magnetic properties of compounds 3 and 4 due to the facts that (a) the distance between the Fe^III or Cr^III atoms is long (~8 Å), like in compounds 1 and 2; (b) the practically isolated character of the Fe^III centers in complex 3 has been proven by its ⁵⁷Fe-Mössbauer spectrum at room temperatures and 80 K; and (c) the chloro groups in one of the aromatic rings of the 4ClsaphCOO²⁻ ligands (which would in theory affect the magnetic exchange) are far from the donor atoms of the paramagnetic centers.

3. Experimental Section

3.1. Materials and Instrumentation

All experiments were performed under aerobic conditions using reagents and solvents (Alfa Aesar, Karlsruhe, Germany; Sigma-Aldrich, Tanfrichen, Germany) as received. Deionized water was received from the in-house facility. The free ligands saphHCOOH [18,20] and 4ClsaphHCOOH [26,27] were synthesized as previously reported; their purity was checked by ¹H NMR spectra in d₆-DMSO on a Bruker Avance DPX spectrometer (Bruker AVANCE, Billerica, MA, USA) at a resonance frequency of 400.13 MHz. Microanalyses (C, H, N) were conducted by the University of Patras Instrumental Analysis Service. Conductivity measurements in CH₂Cl₂ were carried out at room temperature with a Metrohm-Herisau E-527 bridge and a cell of standard design (Metrohm AG, Herisau, Switzerland); the concentration of the solutions was ~10⁻³ M. FT-IR spectra (4000–400 cm⁻¹) were recorded using a Perkin-Elmer 16PC spectrometer (Waltham, MA, USA); the samples were in the form of KBr pellets. For the Raman measurements, a T64000 Horiba-Jobin Yvon micro-Raman setup was used. The excitation wavelength was 514.5 nm, emitted from a DPSS laser (Cobolt Fandango TMISO Laser, Norfolk, UK). The laser power on the samples was 1 mW. Backscattered radiation was collected from a single configuration of the monochromator after passing through an appropriate edge filter (LP02-633RU-25, Laser200 Ltd., Huntingdon, Cambridgeshire, UK). The calibration of the instrument was achieved via the standard Raman peak position of Si at 520.5 cm⁻¹. The spectral resolution was 5 cm⁻¹. UV/VIS solution spectra were recorded using a Specord 50 Plus spectrophotometer (Analytik Jena, Germany). The concentrations of the solutions were ~0.03 mM. ⁵⁷Fe-Mössbauer spectra from powdered, analytically pure samples of the {Cd^II₂Fe^III₂} complexes were recorded using a constant-acceleration conventional spectrometer with a source of ⁵⁷Co (Rh matrix). Spectra in the 300–80 K range were obtained using a Janis cryostat (Cryogenic Technology, Woburn, MA, USA). Isomer shift (δ) values are reported relative to Fe foil at 293 K. For the magnetic measurements, powdered samples were mounted in gelatin capsules in clear plastic straws. The data were obtained using a Quantum Design MPMS-XL-SQUID magnetometer (San Diego, CA, USA). Magnetization data were collected as function of field at 1.8 K. Several data points were recollected as the field returned to zero to check for hysteresis effects; none were observed. Compound 1 showed a small increase in magnetization as the field decreased, which is attributed to a degree of sample reorientation in the 50 kOe field; the effect vanished when the field returned to zero. No remnant magnetization was observed. Data were then collected in a 1 kOe field as a function of temperature from 1.8 to 310 K. Susceptibility values were corrected for the gelatin capsule and straw (measured independently) and the diamagnetism of the constituent atoms as estimated from Pascal’s constants [59].

3.2. Preparation of the Complexes

[Cd₂Fe₂(saphCOO)₄(NO₃)₂(H₂O)₂]·6EtOH·2CH₂Cl₂·2H₂O (1·6EtOH·2CH₂Cl₂·2H₂O). Solid Cd(O₂CMe)₂·2H₂O (0.053 g, 0.20 mmol) was added to a yellow solution of saphHCOOH (0.048 g, 0.20 mmol) in EtOH/CH₂Cl₂ (14 mL, 1:1 v/v). To the resulting yellow suspension was added solid Fe(NO₃)₃·9H₂O (0.040 g, 0.10 mmol). A dark red solution was obtained, which was stirred for 1 h and stored in a closed flask. X-ray quality red crystals of the product were precipitated within 1 week. The crystals were collected by filtration, washed with Et₂O (2 × 2 mL) and dried in a vacuum desiccator over silica gel. The yield was ~60% (based on the Fe^III available). The sample was analyzed satisfactorily as lattice CH₂Cl₂- and EtOH-free, i.e., as 1·2H₂O. Anal. Calcd. (%) for C₅₆H₄₄Cd₂Fe₂N₆O₂₂: C, 45.61; H, 2.98; N, 5.64. Found (%): C, 45.77; H, 2.89; N, 5.43. IR (KBr, cm⁻¹): 3567mb, 1598s, 1557sh, 1541s, 1465m, 1446m, 1384m, 1360w, 1340m, 1300m, 1229w, 1187m, 1149m, 1124w, 1081sh, 1046w, 1027w, 981w, 922m, 875m, 847m, 825w, 795m, 759s, 709m, 688w, 646w, 604m, 563w, 521m, 457m, 416m. Selected Raman peaks (cm⁻¹): 1611s, 1581m, 1550w, 1439s, 1374w, 1339m, 1318m, 1272w, 1185w, 1149w, 603s, 564s. Λ_Μ (CH₂Cl₂, ~10⁻³ M, 25 °C) = 3 S cm² mol⁻¹. UV/VIS (CH₂Cl₂, nm): 235sb, 275sb, 358mb, 442mb, 520wb, 592wb. ⁵⁷Fe-Mössbauer (mm s⁻¹): δ = ~0.4, ΔE_Q = ~1.3 at room temperature (the parameters are not so reliable, see text in 2.4).

[Cd₂Cr₂(saphCOO)₄(NO₃)₂(H₂O)₂]·2EtOH·2MeCN (2·2EtOH·2MeCN). Solid Cd(O₂CMe)₂·2H₂O (0.053 g, 0.20 mmol) was added to a yellow solution of saphHCOOH (0.048 g, 0.20 mmol) in EtOH/MeCN (8 mL, 1:1 v/v). To the resulting yellow suspension was added solid Cr(NO₃)₃·9H₂O (0.040 g, 0.10 mmol). An olive-green solution was obtained which was refluxed under stirring for 2 h and stored in a closed vial. X-ray-quality olive-green crystals of the product were precipitated within 1 week. The crystals were collected on a filter paper, washed with Et₂O (2 × 2 mL) and dried in a vacuum desiccator, first over silica gel and later over anhydrous CaCl₂. The yield was 54% (based on the Cr^III available). The sample was analyzed satisfactorily as lattice solvent-free, i.e., as 2. Anal. Calcd. (%) for C₅₆H₄₀Cd₂Cr₂N₆O₂₀: C, 46.52; H, 2.79; N, 5.81. Found (%): C, 46.80; H, 2.73; N, 5.66. IR (KBr, cm⁻¹): 3470mb, 1593s, 1540m, 1445m, 1383s, 1360sh, 1291w, 1221w, 1190w, 1150m, 1125w, 1099w, 1061w, 1029w, 981w, 929w, 884w, 859w, 796w, 759m, 711w, 671w, 649w, 623w, 534w, 508w, 468sh, 440m, 418m. Selected Raman peaks (cm⁻¹): 1611sh, 1581s, 1530m, 1440s, 1374m, 1358m, 1274w, 1249w, 1189m, 1153m, 1127m, 1026w, 927m, 799m, 607m, 569w, 353m. Λ_Μ (CH₂Cl₂, ~10⁻³ M, 25 °C) = 2 S cm² mol⁻¹. UV/VIS (CH₂Cl₂, nm): 240sb, 288mb, 435wb, 521wb, 588wb.

[Cd₂Fe₂(4ClsaphCOO)₄(NO₃)₂(H₂O)₂]·4.5CH₂Cl₂·4H₂O (3·4.5CH₂Cl₂·4H₂O). Solid Cd(O₂CMe)₂·2H₂O (0.053 g, 0.20 mmol) was added to a yellow solution of 4ClsaphHCOOH (0.055 g, 0.20 mmol) in EtOH/CH₂Cl₂ (10 mL, 1:1 v/v). To the resulting yellow suspension was added solid Fe(NO₃)₃·9H₂O (0.040 g, 0.10 mmol). A dark red solution was obtained, which was stirred for 1 h at room temperature and then layered with Et₂O (3 mL). Red single crystals of the product were grown in a period of 1 week. The crystals were collected on a filter paper, washed with Et₂O (2 × 2 mL) and dried in a vacuum desiccator over anhydrous CaCl₂. The yield was 42% (based on the Fe^III available). The sample was analyzed satisfactorily as lattice CH₂Cl₂-free, i.e., as 3·4H₂O. Anal. Calcd. (%) for C₅₆H₄₄Cd₂Fe₂N₆O₂₄Cl₄: C, 40.44; H, 2.19; N, 5.05. Found (%): C, 40.49; H, 2.30; N, 4.93. IR (KBr, cm⁻¹): 3485wb, 3263m, 1606s, 1582s, 1521s, 1503sh, 1484sh, 1415m, 1381s, 1317w, 1280w, 1230m, 1169sh, 1152m, 1107m, 1078w, 1030s, 907s, 860m, 818w, 788s, 759w, 708sh, 654sh, 656s, 597sh, 580w, 508w, 457w, 423m. Selected Raman peaks (cm⁻¹): 1612s, 1579m, 1534m, 1511w, 1486w, 1438m, 1373w, 1339w, 1317w, 1149w, 604s, 569m. Λ_Μ (CH₂Cl₂, ~10⁻³ M, 25 °C) = 5 S cm² mol⁻¹. UV/VIS (CH₂Cl₂, nm): 230sb, 295sb, 390mb, 450wb, 521wb, 588wb. ⁵⁷Fe-Mössbauer (mm s⁻¹): δ = 0.40(2), ΔE_Q = 0.94(3) at room temperature and δ = 0.48(2), ΔE_Q = 0.92(3) at 80 K.

[CdCr₂(4ClsaphCOO)₄(H₂O)₃(EtOH)]·xEtOH (4·xEtOH). Solid Cd(O₂CMe)₂·2H₂O (0.053 g, 0.20 mmol) was added to a yellow solution of 4ClsaphHCOOH (0.055 g, 0.20 mmol) in EtOH/MeCN (10 mL, 1:1 v/v). To the resulting yellow suspension was added solid Cr(NO₃)₃·9H₂O (0.040 g, 0.10 mmol). A new, olive-green slurry was obtained, and the reaction mixture was refluxed for 2 h. The solid material was dissolved during the reflux, and the resultant solution was layered with Et₂O (4 mL). Poor-quality single crystals (vide infra) of the product were precipitated within 8–9 d. The crystals were collected on filter paper, washed with Et₂O (2 × 2 mL) and dried in air for 2–3 d. The yield was 58% (based on the Cr^III available). The sample was analyzed as lattice EtOH-free, i.e., as 4. Anal. Calcd. (%) for C₅₈H₄₄CdCr₂N₄O₁₆Cl₄: C, 48.88; H, 3.12; N, 3.93. Found (%): C, 49.43; H, 3.05; N, 3.79. IR (KBr, cm⁻¹): 3565wb, 3263w, 1608m, 1583w, 1521m, 1383s, 1230w, 1152w, 1106w, 1029m, 905m, 859w, 815w, 787m, 667m, 578w, 500w, 415w. Selected Raman peaks (cm⁻¹): 1574m, 1528m, 1438s, 1360m, 615m, 576m, 457m. Λ_Μ (CH₂Cl₂, ~10⁻³ M, 25 °C) = 6 S cm² mol⁻¹. UV/VIS (CH₂Cl₂, nm): 230sb, 285sb, 320mb, 390mb, 440mb, 522wb, 585wb.

3.3. Single-Crystal X-ray Crystallography

Red crystals of complexes 1·6EtOH·2CH₂Cl₂·2H₂O (0.07 × 0.24 × 0.43 mm) and 3·4.5CH₂Cl₂·4H₂O (0.22 × 0.25 × 0.33 mm), and olive-green crystals of 2·2EtOH·2MeCN (0.12 × 0.15 × 0.44 mm) and 4·xEtOH (data were collected for several crystals) were taken directly from the mother liquor and immediately cooled to −83 °C, −83 °C, −113 °C and −83 °C, respectively. Diffraction data were collected on a Rigaku R-AXIS SPIDER Image Plate Diffractometer (Rigaku Americas Corporation, The Woodlands, TX, USA) using graphite-monochromated Mo Kα radiation. Data collection (ω-scans) and processing (cell refinement, data reduction and empirical absorption correction) were performed using the CrystalClear program package [60]. The structures were solved by direct methods using SHELXS ver. 2013/1 [61] and refined by full-matrix least-squares techniques on F² using SHELXL ver. 2014/6 [62]. Important crystallographic and refinement details are listed in Table 1. All non-H atoms were refined anisotropically. The H atoms were either located by difference maps and refined isotropically or were introduced at calculated positions and refined as riding on their corresponding bonded atoms. Plots of the structures were drawn using the Diamond 3 program package [63].

For compound 3, the SQUEEZE procedure [64] was applied, and estimated numbers of 4 H₂O and 4.5 CH₂Cl₂ molecules per formula unit were derived based on the accessible solvent voids volume or electron counts within the voids volume. For 4, several crystals were examined, but all diffracted poorly and were twinned. This compound crystallizes in the triclinic system (space group P$\overline{1}$) with unit cell dimensions: a = 13.882(1) Å, b = 14.831(1) Å, c = 19.432(2) Å, α = 97.085(4)°, β = 110.009(5)°, γ = 109.232(3)° and V = 3421.3(5) ų. The refinement was convergent to the R₁/wR₂ values of 0.1995/0.3719 (for all 11519 data), using the non-merohedral twin law of a 2-fold axis parallel to the [4, 1, −1] crystallographic axis expressed by the matrix [(0.59, 0.43, −0.41), (0.00, −1.00, 0.000), (−1.59, −0.43, −0.59)] and a BASF parameter equal to 0.437(6). No.CIF was deposited for this structure. The molecular formula of the cluster is [CdCr₂(4ClsaphCOO)₄(H₂O)₃(EtOH)]·xEtOH. The connectivity model and important characteristics of the molecular structure are clear, and are briefly discussed in this paper.

The X-ray crystallographic data for the three complexes with high-quality structures were deposited with the Cambridge Crystallographic Data Center, Nos 2374217 (1·6EtOH·2CH₂Cl₂·2H₂O), 2374218 (2·2EtOH·2MeCN) and 2374219 (3·4.5CH₂Cl₂·4H₂O).

4. Concluding Comments and Prognosis for the Future

It is difficult to conclude on a research project that is still at its infancy. The chemical message of this work is that the saphHCOOH and 4ClsaphHCOOH ligands can support the preparation of heterometallic 3d–4d complexes; complexes 1–4 are the first heterometallic species of these ligands with this metal combination. The O_phenolate, N_imine, O_carboxylate donor-atom system of the doubly deprotonated ligands is ideal for tridentate chelating behavior comprising two six-membered rings, while leaving the coordinated phenolato oxygen, and the “free” carboxylate oxygen, or both the coordinated and “free” carboxylate oxygens available for bridging other metal ions (one in the present case). A priori prediction for the preference of homometallic or heterometallic compounds is rather impossible. Thermodynamic factors seem to be responsible for the self-assembly process that governs the nature of the product. Based on HSAB principles, we expected that Cd^II would be capable of the formation of the two chelating rings, whereas Fe^III and Cr^III would occupy bridging positions of the ligands. However, the opposite was observed here, questioning the utility of this model for the synthesis of products containing two different metal ions with varying HSAB characteristics. From the inorganic chemistry viewpoint, the complexes have interesting molecular and supramolecular structures, unusual cores, and enrich the rather small families of Cd^II-Fe^III and Cd^II-Cr^III compounds; moreover, {Cd^II₂Cr^III₂} and {Cd^IICr^III₂} nuclearities have been observed for the first time in discrete coordination clusters. It is also worth mentioning that the nuclearity of the Cd^II-Cr^III complexes seems to depend on the ligand used, i.e., {Cd^II₂Cr^III₂} with saphCOO²⁻ and {Cd^IICr^III₂} for 4ClsaphCOO²⁻. From the magnetochemistry viewpoint, the long M^III… M^III distances are responsible for the almost magnetic isolation of the M^III ions and the absence of significant exchange interactions.

As mentioned in the “Introduction”, the only previously characterized heterometallic clusters based on the two ligands are {Zn^II₄M^III₂} compounds (M = Fe, Cr) [28]. The replacement of Zn^II by Cd^II has a dramatic influence on the product identity. The Zn^II-M^III clusters are hexanuclear, whereas the Cd^II-Cr^III ones are tetranuclear and trinuclear. The coordination modes of the ligands are also completely different. In the hexanuclear clusters, the ligands adopt the 3.2111 (with exclusive bridging of three Zn^II atoms) and 2.1111 (where the chelating part of the bridging ligand is again bonded to Fe^III or Cr^III as in 1–4) modes; in contrast, the bridging behavior of the ligands in the tetranuclear clusters is of the 2.2111 type (Scheme 2). The observed differences are primarily due to the ability of Cd^II to attain coordination numbers higher than six (8 in 1–4), whereas the maximum coordination number of Zn^II is generally six (four and five in the {Zn^II₄M^III₂} compounds [28]).

We believe that the research reported herein is not exhausted of new results; recent results in our groups indicate that we have scratched only the surface of the heterometallic chemistry based on saphCOO²⁻ and 4ClsaphCOO²⁻. Current efforts are directed towards the following: (1) the replacement of Fe^III in 1 and 3 with the diamagnetic Ga^III and In^III or 4f-metal ions (e.g., Eu^III and Tb^III) in order to isolate clusters with interesting photoluminescence properties based on the organic ligands or lanthanide(III) centers, respectively; (2) the replacement of Zn^II and Cd^II with paramagnetic divalent 3d-metal ions, e.g., Mn^II, Co^II, Ni^II and Cu^II, with the goal to study the magnetic properties of clusters containing solely paramagnetic centers; (3) the incorporation of electron-withdrawing (except Cl used in this work) or electron-releasing, non-donor groups in one or both the aromatic rings of saphHCOOH, to investigate the difference of the substitutes on the electronic features and structural characteristics of the products; and (4) the replacement of the H atom on the imine carbon with Me or Ph groups, and study of their heterometallic chemistry.