Can Serendipity Still Hold Any Surprises in the Coordination Chemistry of Mixed-Donor Macrocyclic ligands? The Case Study of Pyridine-Containing 12-Membered Macrocycles and Platinum Group Metal ions Pd^II, Pt^II, and Rh^III

Abstract

This study investigates the coordination chemistry of the tetradentate pyridine-containing 12-membered macrocycles L¹-L³ towards Platinum Group metal ions Pd^II, Pt^II, and Rh^III. The reactions between the chloride salts of these metal ions and the three ligands in MeCN/H₂O or MeOH/H₂O (1:1 v/v) are shown, and the isolated solid compounds are characterized, where possible, by mass spectroscopy and ¹H- and ¹³C-NMR spectroscopic measurements. Structural characterization of the 1:1 metal-to-ligand complexes [Pd(L¹)Cl]₂[Pd₂Cl₆], [Pt(L¹)Cl](BF₄), [Rh(L¹)Cl₂](PF₆), and [Rh(L³)Cl₂](BF₄)·MeCN shows the coordinated macrocyclic ligands adopting a folded conformation, and occupying four coordination sites of a distorted square-based pyramidal and octahedral coordination environment for the Pd^II/Pt^II, and Rh^III complexes, respectively. The remaining coordination site(s) are occupied by chlorido ligands. The reaction of L₃ with PtCl₂ in MeCN/H₂O gave by serendipity the complex [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O, in which two metal centers are bridged by an amidate ligand at a Pt1-Pt2 distance of 2.5798(3) Å and feature one square-planar and one octahedral coordination environment. Density Functional Theory (DFT) calculations, which utilize the broken symmetry approach (DFT-BS), indicate a singlet d⁸-d⁸ Pt^II-Pt^II ground-state nature for this compound, rather than the alleged d⁹-d⁷ Pt^I-Pt^III mixed-valence character reported for related dinuclear Pt-complexes.

1. Introduction

Macrocyclic chemistry is a very important and active area of chemical science with implications in a wide variety of applications, such as analytical chemistry, separation science, catalysis, and medicinal chemistry [1,2,3,4,5,6], and also in the development of fundamental aspects of supramolecular chemistry, such as molecular recognition, host-guest interactions, design of sensors, and smart artificial molecular devices [7,8,9,10].

Novel macrocyclic chemical structures—differing in molecular shape, architecture, flexibility, arrangement of structural groups, binding sites, and reactive functions—continue to be developed, with the aim of improving performances in the chemical functions of interest by achieving better control over the strength, selectivity, and dynamics of the binding processes of a variety of cationic, anionic, neutral, organic, and inorganic substrates.

However, the basic aspects of coordination chemistry of macrocyclic ligands towards different substrates, particularly metal ions, continue to be a fascinating area of research in the quest for systems capable of forcing the metal center to adopt unusual coordination geometries and/or oxidation states within stable complexes. For this purpose, the hard-soft nature of donor atoms and their spatial disposition, the cavity size, and flexibility of macrocyclic ligands are the most important parameters that define the coordination properties of these systems in relation to the stereo-electronic requirements of the metal ions of interest [11,12].

In this context, we have been engaged in the development of mixed N/O/S-donor macrocycles featuring rigid heterocyclic moieties, such as pyridine (py) [13,14,15,16,17,18,19,20,21], and 1,10-phenanthroline (phen) [22,23,24,25,26,27,28,29,30,31] as integral parts of the macrocyclic structure, which is completed by an aliphatic portion carrying different donor atoms. These systems proved to be highly efficient and selective ionophores in solid-phase extraction, selective transport, preparation of PVC-based ion-selective electrodes, and fluorimetric chemosensors for some transition and heavy metal ions. On the other hand, the conformational constraints on the aliphatic portion of these cyclic systems determined by the rigid heteroaromatic moieties, along with the fact that these heteroaryl frameworks carry one or more borderline N-donor atoms and are excellent π-acceptors groups, can be useful factors in expanding the scope of forcing unusual coordination behaviors, especially on d⁸ transition metal ions, such as Pd^II and Pt^II, having very strict stereo-electronic requirements [32,33].

We report herein the results of our investigation of the coordination chemistry of tetradentate macrocycles L¹-L³ (Figure 1) to platinum group metals, particularly Pd^II, Pt^II, and Rh^III. L¹-L³ contain one N- and two S-donor atoms as present in bis(thiomethyl)pyridine and complete their donor set with an aliphatic linker featuring the extra donor atom O(L¹), S(L²), and NH(L³) (Figure 1). Although these ligands, especially L³, have found interesting applications in the development of fluorescent chemosensors for heavy metal ions [13,14,15,16,17,18,19,20,21], their coordination chemistry is largely undeveloped.

In fact, in the case of L¹, only the X-ray crystal structure of the neutral complex [Cu(L¹)Cl₂] is known [34], while for L³, only the 1:1 complex cations [Cu(L³)]²⁺ and [Zn(L³)]²⁺ have been reported as nitrate salts [13]. In the case of L², the polymeric complexes [Ag(L²)]_n(CF₃CO₂)_n·nH₂O [35] and [Ag(L²)]_n(NO₃)_n [36] were structurally characterized along with the discrete 1:1 complex cations [Cu(L²)]²⁺ [35] and [Hg(L²)Cl]⁺ [37] (NO₃⁻ and HgCl₃⁻ counter-anions, respectively), and the binuclear compound [Ni₂(L²)₂Cl₂](BF₄)₂·1.5MeNO₂ [22].

The present study is strictly related to the previous ones performed on macrocyclic ligands similar to L¹ and L² but featuring the phen moiety instead of the py unit [22,32,33,38]. In those cases, the nature of the donor atom sets, the conformational constraints determined by the phen unit on the thioether linkers of the two pentadentate rings, and the locked [4+1] coordination sphere imposed on the Pd^II and Pt^II ions in their 1:1 complexes were responsible for the stabilization of the corresponding low-valent complexes of Pd^I and Pt^I [32,33]. The crystal structures of L¹ and L² are known [34,39]. In both structures, the aliphatic chain of the rings is tilted over the plane containing the pyridine unit—presumably because of the repulsion between the two sulfur atoms close to the aromatic ring [34]. The other donor atom, independently of its nature (oxygen in the case of L¹ and sulfur in the case of L²), adopts exodentate orientations with the lone pairs of electrons (LPs) pointing out of the ring cavity. Therefore, a conformational change is required for these ligands (and presumably also for L³) to coordinate a metal center with all four donor atoms. A similar situation is observed in the case of the analog of L¹ but featuring a phen moiety instead of the py unit, in the free ligand for which the crystal structure is known [33]. In this case, the explanation given for the conformational behavior observed for L¹ and L² cannot be applied, as the S-donors would be too far apart even in a completely planar conformation of the ligand. The tendency of the LPs on the S-donors to occupy exodentate positions pointing out of the ring cavity, with the effect of maximizing the number of gauche placements about the C-S bonds, seems more likely to be responsible for the tilted conformation also observed in the phen analogous of L¹. In both kinds of macrocyclic ligand, therefore, a conformational change in the aliphatic chains is required upon coordination to bring the lone pair(s) of all donors to adopt endodentate orientations suitable for metal coordination [13,22,32,33,34,35,36,38,39].

2. Results

2.1. Coordination Chemistry of L¹ towards Pd^II, Pt^II, Rh^III

The reaction of L¹ with one molar equivalent of PdCl₂ in refluxing MeOH/H₂O (1:1 v/v), followed by reduction of the volume of the reaction mixture under vacuum and slow evaporation in the air of the remaining solvent (water), afforded reddish prismatic crystals. Analytical data (Fast Atom Bombardment (FAB) Mass Spectrum, Figure S1 in the Supplementary Materials (SM), and elemental analysis in the Materials and Methods Section) indicate a Pd/L¹ molar ratio higher than 1:1 in the obtained compound.

The ¹³C-NMR spectrum of the complex recorded in CD₃CN solution at 25 °C shows only three peaks for the aromatic fragment of the macrocyclic ligand (δ_C = 122.8, 140.5, 164.1 ppm) and three for the aliphatic chain (δ_C = 45.0, 45.9, 65.4 ppm, Figure S2 in SM), thus suggesting that the complex exists in solution in only one form having a C_s symmetry with a symmetry plane passing through the N-donor atom of the ligand.

With respect to the free macrocycle, the carbon atoms next to S-donors are deshielded [δ_C = 45.0 (36.6) and 45.9 (30.2) ppm for C7/C12 and C8/C11, respectively; see Figure 1 for the numbering scheme adopted, values in parentheses refer to the free macrocycle], whereas those next to the aliphatic O-donor are slightly shielded [δ_C = 65.4 (66.7) ppm for C9/C10; see Figure 1]. The ¹³C-NMR chemical shifts do not change on changing the temperature in the range allowed by the solvent CD₃CN. These data are consistent with a coordination sphere imposed in solution by L¹ at the Pd^II with possibly the O-donor atom weakly interacting with the metal center.

The ¹H-NMR spectrum of the complex recorded in CD₃CN solution at 25 °C (Figure S3) exhibits six distinct groups of aliphatic protons at 2.74–2.78 (multiplet), 3.32–3.34 (multiplet), 3.54–3.59 (multiplet), 4.07–4.10 (multiplet), 4.45 (doublet) and 4.88 (doublet) ppm each integrating for 2 protons, the assignments of which have been made on the basis of ¹H-¹³C-NMR Heteronuclear Single Quantum Correlation (HSQC) experiments in CD₃CN (see Experimental Section and Figure S4).

The doublets at 4.45 and 4.88 ppm are due to an AB spin system (J = 18.5 Hz) for each pair of protons on C7 and C12. This is confirmed by the observation in the HSQC that both doublets correlate with the same ¹³C resonance at 45.0 ppm. The same is found for the multiplets mentioned above, whose multiplicity is indicative of an AA’BB’ spin system. The four multiplets can be distinguished in two pairs, each correlating with a single ¹³C resonance in the HSQC spectrum. These observations indicate the presence of a symmetry plane bisecting the pyridine ring and passing through the metal ion in the complex structure, while the two geminal protons in each of the methylene groups are magnetically inequivalent. This is in agreement with inequivalent dispositions assumed by the protons on C7 and C12 (above and below the plane of the pyridine moiety) as a consequence of the ligand complexation.

An X-ray diffraction analysis was undertaken on the isolated reddish crystals to ascertain the ligation and stereochemistry of this complex. The crystal structure confirms the formation of the compound [Pd(L¹)Cl]₂[Pd₂Cl₆] containing [Pd(L¹)Cl]⁺ complex cations (Figure 2) balanced by [Pd₂Cl₆]^2– counter-anions and with the tetradentate macrocyclic ligand imposing a [3+1] coordination sphere at the Pd^II. The N-donor of the pyridine unit, Pd1-N1 2.013(3) Å, the two S-donors of the aliphatic linker, Pd1-S1 2.3062(10), Pd1-S2 2.2915(10) Å, and a Cl⁻ ligand, Pd1-Cl1 2.2984(11) Å are bound in a square-planar arrangement to the metal atom. The O-donor occupies an apical site at a distance of 2.654(3) Å from the metal center, which is much less than the sum (3.10 Å) of the relevant van der Waals radii [40]. A similar type of Pd···O interaction [2.779(4) Å] was observed in the complex cation [Pd([15]aneN₂OS₂)]²⁺ ([15]aneN₂OS₂ = 1-oxa-7,10-dithia-4,13-diazacyclopentadecane) [41], while much longer Pd···O interactions were found in the 1:1 complex of Pd^II with the pentadentate macrocyclic ligand similar to L¹, but having a phen unit instead of the py moiety [2.935(4) Å with the O-donor lying above the N₂PdS₂ coordination plane] [33], and in the half-sandwich complex [Pd([9]aneS₂O)Cl₂] ([9]aneS₃ = 1,4,7-trithiacyclononane), in which the macrocyclic ligand assumes a facial [2S + O] coordination mode at the metal center with the oxygen atom lying above the Cl₂PdS₂ coordination plane at a Pd···O distance of 2.968(3) Å [42]. The Pd-O vector is almost perpendicular to the Pd^II coordination plane as a consequence of the folded conformation adopted by L¹ in the complex cation [Pd(L¹)Cl]⁺ (Figure 2), which resembles an open book with the spine along the line connecting the S1-Pd1-S2 atoms and the N1-Pd1-O1 hinge angle of 89.5(1)°.

The charge neutrality of the complex is guaranteed by the dinuclear planar [Pd₂Cl₆]^2– anion featuring two Pd^II metal centers in a square-planar coordination sphere, which is determined by four coordinated chloride anions, two of which bridging the metal ions.

The [Pd₂Cl₆]^2– anions, which lie on a crystallographic inversion center, are sandwiched by two [Pd(L¹)Cl]⁺ cations and interact with them through CH···Pd and CH···Cl interactions of 2.67 and 2.71 Å, respectively (Figure S5 in SM). CH···Cl interactions among anions and complex cations ranging between 2.76 and 2.89 Å (Figure S6), together with intermolecular CH···Cl (2.73, 2.85 Å) and CH···O (2.80 Å) interactions among complex cations, contribute to determine the crystal packing in this compound (Figures S6 and S7).

Following the same synthetic procedures adopted for the synthesis of [Pd(L¹)Cl]₂[Pd₂Cl₆], we reacted L¹ with PtCl₂ in refluxing MeOH/H₂O (1:1 v/v). Yellow crystals were obtained after the addition of excess NH₄BF₄ to the reaction mixture, evaporation of MeOH under the vacuum, and subsequent crystallization in the air of the remaining aqueous solution by slow evaporation. The FAB mass spectrum of the compound (Figure S8) exhibits peaks with the correct isotopic distribution for [Pt(L¹)Cl]⁺ (m/z = 472). These data, together with elemental analysis, confirm the formulation [Pt(L₁)Cl](BF₄) for the isolated compound. ¹H- and ¹³C-NMR spectra (Figures S9 and S10, respectively) of the complex in CD₃CN show features very similar to those observed for [Pd(L¹)Cl]₂[Pd₂Cl₆], including the evidence of the AB spin system (J = 18.0 Hz) for the doublets at 4.62 and 4.85 ppm for each pair of protons on C7 and C12, respectively, and of the AA’BB’ spin system for the other methylene groups resonating at a lower frequency (assignments are made on the basis of ¹H-¹³C-NMR HSQC experiments in CD₃CN, Figure S11). This strongly suggests a very similar structure for the complexes formed with Pd^II and Pt^II. Furthermore, the carbon atoms next to S-donors are deshielded [δ_C = 46.7 (36.6) and 47.4 (30.2) ppm for C7/C12 and C8/C11, values in parentheses refer to the free macrocycle], whereas those next to the aliphatic O-donor are slightly upshifted [δ_C = 66.4 (66.7) ppm for C9/C10].

An X-ray diffraction analysis was undertaken on the obtained yellow crystals to ascertain the nature of this complex. The crystal structure confirms the formation of the complex cation [Pt(L¹)Cl]⁺ (Figure 3) balanced by a BF₄⁻ counter-anion.

The coordination environment at the metal center is very similar to that observed in the case of [Pd(L¹)Cl]⁺ with the macrocyclic ligand adopting the typically folded conformation and imposing a [NS₂+O] coordination sphere at the Pt^II metal ion, which reaches an overall square-based pyramidal geometry thanks to the coordination of a Cl⁻ ligand in the equatorial plane (Figure 3). The O-donor occupies the apical site of the square-pyramid at a distance of 2.752(4) Å from the metal center, which is slightly longer than the Pd···O distance observed in the complex cation [Pd(L¹)Cl]⁺. In the crystal packing, two units of complex cation interact via Pt···S and CH···Cl contacts of 3.625(2) Å and 2.88 Å, respectively, with the relevant equatorial coordination planes facing each other (Figure 4). Dimers of this kind interact head-to-tail via CH···Cl and CH···O H-bonds of 2.89 and 2.43 Å, respectively, to form zig-zag chains running along the [001] direction.

The zig-zag chains of [Pd(L¹)Cl]⁺ complex cations are joined in the crystal via CH···F H-bonds ranging from 2.38 to 2.48 Å involving BF₄⁻ counter anions (Figure S12).

A synthetic procedure similar to that adopted for the synthesis of [Pt(L¹)Cl](BF₄) was also employed for the preparation of the 1:1 complex of L¹ with Rh^III. Yellow crystals were obtained after the addition of excess NH₄PF₆ to the reaction mixture of RhCl₃·H₂O and L¹ in MeCN/H₂O (1:1 v/v), removal of the solvent under vacuum, and crystallization of the resulting solid from MeCN by slow diffusion of Et₂O vapors. The elemental analysis and the FAB mass spectrum (Figure S13) of the obtained crystals, which exhibits peaks with the correct isotopic distribution for [Rh(L¹)Cl₂]⁺ (m/z = 414), confirm the formulation [Rh(L¹)Cl₂](PF₆) for the isolated compound. Similar to the case of Pd^II and Pt^II, the ¹H-NMR spectrum reflects an AB spin system for each pair of protons on C7 and C12 (doublets at 5.01 and 5.25 ppm with J = 18.6 Hz, Figures S14–S16) and AA’BB’ spin system for the other methylene protons (four multiplets at 4.04–4.07, 3.57–3.62, 3.48–3.51 and 3.36–3.40 ppm). The two doublets showed correlation with a single ¹³C resonance at 46.0 ppm in the HSQC spectrum (Figure S16). The four multiplets can be divided into two pairs, with the two ¹H multiplets at 4.04–4.07 and 3.36–3.40 ppm showing correlation with a single ¹³C resonance at 74.1 ppm, while the other two ¹H multiplets at 3.57–3.62 and 3.48–3.51 ppm showed correlation with the same ¹³C resonance at 40.5 ppm, in the HSQC spectrum. These observations strongly indicate a coordination mode of the ligand analogous to that observed in the Pd^II and Pt^II complexes of L¹.

An X-ray diffraction analysis was undertaken on the obtained yellow crystals showing the presence of [Rh(L¹)Cl₂]⁺ complex cations counterbalanced by PF₆⁻ anions in the crystal structure. The complex cations feature a Rh^III ion in a distorted octahedral environment defined by the four donor atoms of a macrocycle L¹ and two chlorido ligands (Figure 5). The structure of the cation is conditioned by the meridional coordination of the 2,6-bis(thiomethyl)pyridine unit, as is observed in the value of the S-Rh-S angle, 170.23(3)°. The O-donor is located perpendicular to the pseudo-plane defined by the metal ion, the pyridine ring, and the two thioether sulfur atoms. The folded conformation, adopted by L¹ as in the cases of [M(L¹)Cl]⁺ complex cations (M = Pd, Pt), leaves the two coordination sites occupied by two Cl⁻ ligands in a relative cis orientation.

While the Rh-N, Rh-S, and Rh-Cl bond distances are similar to the corresponding bond lengths observed in the [M(L¹)Cl]⁺ complex cations (M = Pd, Pt), the Rh-O bond length [2.088(2) Å] is much shorter, in agreement with the stereoelectronic requirements of Pd^II, Pt^II, and Rh^III in their coordination chemistry. In the crystal packing, [Rh(L¹)Cl₂]⁺ units are joined head to tail via CH···Cl H-bonds of 2.66 Å to form zig-zag chains, which run along the [010] direction (Figure 6). As in the case of the compound [Pd(L¹)Cl](BF₄), H-bonded chains of [Rh(L¹)Cl₂]⁺ units are joined via CH···F bonds of lengths 2.33–2.53 Å involving PF₆⁻ counter-anions to afford a 3-dimensional network (Figure S17).

2.2. Coordination Chemistry of L² towards Pd^II, Pt^II, Rh^III

The reaction of L² with one molar equivalent of PdCl₂ or PtCl₂ in refluxing MeCN/H₂O (1:1 v/v) afforded orange and yellow microcrystalline powders, respectively (see Experimental Section). Unfortunately, we were not able to grow crystals suitable for X-ray diffraction analysis. However, mass spectra (Figures S18, S19) and elemental analyses suggest the formation of 1:1 metal-to-ligand complexes having the formulation [M(L²)Cl]Cl (M = Pd^II and Pt^II).

¹H-NMR spectra of the complex [Pd(L²)Cl]Cl (Figure S21) and [Pt(L²)Cl]Cl (Figure S23) recorded in D₂O and CD₃CN, respectively, show features that are very similar to those observed for the respective complexes with L¹, including the evidence of an AB spin system for each pair of protons on C7 and C12 (numbering scheme as in Figure 1) [doublets at 4.70 and 5.17 ppm (J = 18.8 and 18.4 Hz) for the Pd^II compound, 3.94 and 4.05 ppm (J = 12.8 Hz) for the Pt^II compound]. The carbon atoms next to the S-donors are all deshielded [δ_C = 46.9 (36.2), 48.0 (30.9), 33.2 (29.9) ppm for C7/C12, C8/C11, and C9/C10, respectively, in the case of the Pd^II complex (Figure S22, values in parentheses refer to the free macrocycle); 39.8, 39.2 and 34.3 ppm for the Pt^II complex (Figure S24)]. These results suggest a [3+1] coordination mode of L² at the Pd^II and Pt^II metal centers within a square-based coordination sphere, similar to those observed for L¹ in the case of the complex cations [Pd(L¹)Cl]⁺ and [Pt(L¹)Cl]⁺ (see above), respectively, with presumably stronger interactions of the metal ions with the apical S-donor. Interestingly, the free macrocycle L² is reported to prefer a “chair-like” conformation in which the central S-donor is oriented in the opposite direction with respect to the site perpendicular to the plane containing the remaining NS₂ donor set, and a conformational change is necessary to interact with the d_z² orbital of the coordinated transition metal ion [39].

No crystals could be grown for the brown solid isolated from the reaction of L² with one molar equivalent of RhCl₃·H₂O in MeCN/H₂O (1:1 v/v) followed by the addition of excess NH₄PF₆. Elemental analysis and the FAB mass spectrum (Figure S20) of the isolated compound showing peaks with the correct isotopic distribution for both [Rh(L²)Cl₂]⁺ (m/z = 430) and for [Rh(L²)Cl]⁺ (m/z = 396) suggest the presence in the solid crude product of a 1:1 complex having the formulation [Rh(L²)Cl₂](PF₆).

Surprisingly, the ¹H- and ¹³C-NMR spectra of the Rh^III complex (Figures S25 and S26, respectively) isolated with L² recorded in CD₃CN clearly show two distinct complexes in solution. Two series of homologous resonances can be seen, both for the aromatic and the methylene protons. Homologous resonances differ in position and relative intensity, but share the same fine structure, with very similar J couplings. By the relative intensity of 1.3, we can distinguish one major and one minor species, separated by a ΔG of 0.65 kJ mol⁻¹. As far as the methylene resonances are concerned, both species are characterized by two doublets around 5 ppm corresponding to an AB spin system and showing scalar correlation with the same ¹³C resonance in the HSQC spectrum (Figure S27). Both the species are characterized by four multiplets between 2.5 and 4.0 ppm corresponding to an AA’BB’ spin system and distinguished in couples by showing correlation with the same ¹³C resonance in the HSQC spectrum. Definitely, the two complexes appear to be very similar. Since L¹ formed the complex cation [Rh(L¹)Cl₂]⁺ with Rh^III, the major and minor species observed for L² could be tentatively assigned to the cis and the trans configurations of the two coordinated chloride ions. However, the latter is not compatible with the AB spin system observed for each pair of protons on C7 and C12, which are expected to be equivalent and to appear as a singlet in the ¹H-NMR spectrum. In order to clarify this point further and to characterize the two species in more detail, we acquired NOESY spectra on both the L¹ and L² complexes with Rh^III. Their analyses (see SI for discussion and Figure S28 for the molecular model compatible with NMR measurements) clearly point out that in the case of L² and Rh^III, the dichlorido complex is formed together with another species in which one chloride, at least, is substituted by some other ligands (presumably a solvent molecule, likely MeCN). All attempts to isolate the two complexes by chromatography were unsuccessful. Data available on the solid-state are not conclusive on the presence of the complex featuring only one coordinated chlorido ligand (see above). While the peak at m/z 430 can be unambiguously assigned to the species [Rh(L²)Cl₂]⁺, the peak at m/z = 396, which can be assigned to [Rh(L²)Cl]⁺, could either derive from [Rh(L²)Cl₂]⁺ by loss of one chlorido ligand or from the other complex by the loss of the coordinated solvent molecule.

2.3. Coordination Chemistry of L³ towards Pd^II, Pt^II, Rh^III

The reaction of L³ with one molar equivalent of PdCl₂ in refluxing MeCN/H₂O (1:1 v/v), followed by the addition of excess NH₄PF₆, the reduction of the solvent volume under vacuum, and the slow evaporation in the air of the remaining solvent (water), afforded a brown microcrystalline powder. The FAB mass spectrum of the compound exhibits a peak at m/z = 381 with the correct isotopic distribution expected for the cation [C₁₁H₁₆ClN₂PdS₂]⁺ (see Figure S29 in the SM). This, together with elemental analytical data, support the formation of a 1:1 complex having the formulation [Pd(L³)Cl](PF₆).

Indeed, as already observed for the Pd^II complexes with L¹ and L² (see above), the ¹³C-NMR chemical shifts for the Pd^II complex with L³ in CD₃CN solution shows only three peaks for the aromatic region and three for the aliphatic chain (Figure S30), thus suggesting that the complex exists in solution in only one form having a C_s symmetry with a symmetry plane passing through the two N-donor atoms of the ligand. With respect to the free macrocycle, the carbon atoms next to S-donors are deshielded [δ_C = 45.8 (37.8) and 43.3 (31.7) ppm for C7/C12 and C8/C11, respectively; (values in parentheses refer to the free macrocycle)], whereas those next to the aliphatic N-donor are slightly deshielded [δ_C = 48.2 (47.0) ppm for C9/C10]. These features have also been observed in the ¹³C-NMR spectra of the 1:1 complexes of Pd^II and Pt^II with pentadentate macrocyclic ligands similar to L¹-L³, but having a phen unit instead of the py moiety [32,33], which showed a [4 + 1] coordination sphere at the metal centers with the central donor atom in the aliphatic linker occupying the apical site of a distorted square-based pyramid with a long-range interaction to the metal atom, and the ligand adopting a folded conformation. A similar coordination mode of L³ to the Pd^II center can be suggested for the complex cation [Pd(L³)Cl]⁺ in which the [4 + 1] pyramidal coordination sphere would be reached thanks to a chloride anion occupying one site in the basal coordination plane (see Pd^II and Pt^II complexes of L¹ and L² above).

This hypothesis is supported by the ¹H-NMR spectrum of the complex in D₂O (Figure S31) that exhibits four distinct groups of aliphatic protons at 3.03–3.06 (multiplet), 3.54–3.60 (multiplet), 4.61 (doublet), and 5.07 (doublet) ppm integrating for four, four, two, and two protons, respectively. The doublets at 4.61 and 5.07 ppm define an AB spin system for each pair of protons on C7 and C12 (Figure 1) (J = 18.6 and 18.0 Hz, respectively, assignment of the chemical shift is made for the analogy with the ¹HNMR shits observed for the Pd^II, Pt^II, and Rh^III complexes of L¹, whose assignment is made via ¹H-¹³C-heteronuclear correlation, HSQC, experiments, see below), which agrees with inequivalent dispositions assumed by these protons (above and below the plane of the pyridine moiety) as a consequence of metal complexation.

The reaction of L³ with one molar equivalent of PtCl₂ in refluxing MeCN/H₂O (1:1 v/v), followed by the addition of excess NH₄BF₄, the reduction of the solvent volume under vacuum, and the slow evaporation in the air of the remaining solvent (water), serendipitously afforded a few red-orange single crystals suitable for X-ray diffraction analysis. We repeatedly tried the complexation in MeCN/H₂O and also under different experimental conditions changing the solvent mixture by replacing MeCN with other solvents (MeNO₂/H₂O, MeOH/H₂O, THF/H₂O), but no reaction occurred, and the sole unreacted ligand was always recovered.

The crystal structure determination revealed the unusual complex [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O. The asymmetric unit consists of a binuclear [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ complex cation involving a bridging amidate ligand likely formed from the hydrolysis of acetonitrile solvent [43,44,45] (Figure 7), BF₄⁻ counter-anions and a co-crystallized water molecule. In the dimeric unit, one platinum atom is six-coordinated in a distorted octahedral geometry, being surrounded by the four donor atoms from the macrocyclic ligand L³ [Pt2-N1 2.006(3), Pt2-N2 2.237(3), Pt2-S1 2.2926(10), Pt2-S2 2.3067(10) Å] the O-donor atom from the amidate bridge [Pt2-O1 2.018(3) Å], and the other platinum atom [Pt1-Pt2 2.5798(3) Å]. Pt1 is four-coordinated in a square-planar geometry, due to the additional coordination of an acetonitrile molecule [Pt1-N3 1.971(4) Å], a chlorido ligand [Pt1-Cl1 2.3433(11) Å], and the N-donor atom from the amidate bridging ligand [Pt1-N4 1.981(4) Å] (Figure 7). The maximum deviation from the least-squares plane calculated through the atoms Pt1, Cl1, N3, N4, Pt2 is 0.04 Å for N4. The average coordination plane at the Pt1 atom, which also comprises the two N-donors from the macrocyclic ligand and the two donors from the amidate bridging ligand, is almost perpendicular to the plane containing the pyridine ring and the Pt2, S1, S2, and O1 donors with the interplanar angle being 89.09° (Figure 7). L³ adopts the folded conformation already observed in the complex cations [Cu(L³)]²⁺ and [Zn(L³)]²⁺ [13] resembling an open book with the spine along the line connecting the S1-Pt2-S2 atoms and the N1-Pt2-N2 hinge angle of 91.81(13)°. The aliphatic tertiary nitrogen is, therefore, located almost perpendicularly to the pseudo-plane defined by the metal ion, Pt2, the pyridine ring, the S-donors, and the amidate O-donor, in trans-position with respect to the other platinum atom, Pt1 (Figure 7). The four-coordinated platinum atom, Pt1, features the other platinum atom and the chlorido ligand in mutually trans-positions, giving an almost linear Cl-Pt-Pt-N arrangement in the binuclear cation, with the other two trans-positions being occupied by the coordinated acetonitrile molecule and the amidate N-donor. It is interesting to note that L³ binds metal atoms with almost equivalent M-N1 and M-N2 bond lengths [13]. In contrast, in the complex cation, the bond distance Pt2-N2 = 2.237(3) Å is longer than the Pt2-N1 = 2.006(3) Å due to the higher trans-influence of coordinating Pt1 compared to O1 donor atoms, thus suggesting a donor–acceptor nature for Pt-Pt bond, also confirmed by the short Pt1-Pt2 distance [2.5798(3) Å] consistent with a metal-metal bond.

The binuclear [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ complex cations interact through H-bonds involving the BF₄⁻ anions and the water molecule forming head-to-tail chains running along the crystallographic [001] direction (Figure 8). Symmetry-related chains pack through H-bonds involving the complex cations and the BF₄⁻ anions (Figure S32).

Platinum binuclear complexes similar to [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ are quite rare, the only other two known examples being the neutral two-electron mixed-valence complexes [Pt₂^I,III(tfepma)₂X₄] [X = Cl, Br; tfepma = bis(bis(trifluoroethoxy)-phosphino)methylamine] obtained by X₂ photo-elimination from homobimetallic meridionally coordinated Pt^III tri-halides bridged by two neutral tfepma ligands, [Pt₂^III,III(tfepma)₂X₆] [46,47].

To get a deeper insight into the structural features of diplatinum-based discrete complexes, we have performed a search in the Cambridge Structural Database [48] for all compounds of this type containing a Pt-Pt bond, sorted on the coordination number around each platinum atom.

The structurally characterized diplatinum systems were assigned to the suites of dⁿ-dⁿ and mixed-valence dⁿ-d^m complexes (n, n = 6, 7, 8, 9; n, m = 6, 8; 7, 8; and 7, 9), based on the reported formal oxidation states. Figure 9 shows the number of fragments found for the different suites versus the coordination number displayed by the platinum centers.

Indeed, an examination of the data reported in Figure 9 shows that reportedly mixed-valence dⁿ-d^m complexes are quite rare, with only five examples known in the literature, among the 505 items found [49,50,51,52,53] beside the complexes [Pt₂^I,III(tfepma)₂X₄] (X = Cl, Br) [46,47]. The category with the higher number of items (212) is that of complexes formally featuring two penta-coordinated Pt^II ions, which can be envisaged as two square-planar complexes interacting through long Pt-Pt bonds ranging from 2.53 to 3.41 Å, with a mean value as long as 2.94 Å and an overall distorted square-based pyramidal environment for both metal ions. (d⁸-d⁸, Pt5-Pt5 green column in Figure 9). Quite numerous (155) are also the binuclear complexes formally featuring two Pt^III ions sharing one of the six bonds in a distorted octahedral coordination for both metals, with Pt^III-Pt^III bond lengths in the range 2.39–3.08 and a mean value of 2.61 Å (d⁷-d⁷, Pt6-Pt6 blue column in Figure 9). For this kind of binuclear system, a significant number of structures (40 items) featuring the two metal ions in a different coordination environment (distorted octahedral/square-based pyramidal) are reported (d⁷-d⁷, Pt5-Pt6 orange column in Figure 9). These are often described as formal Pt^III-Pt^III dimers with significant Pt^IV and Pt^II influences for the octahedral and square pyramidal platinum center, respectively [52]. The Pt-Pt distances again occupy a quite narrow range of 2.50–2.85 Å with a mean value of 2.69 Å.

Discrete dimers formally featuring a Pt^I-Pt^I bond can be found in complexes featuring distorted square-planar/square-planar, square-planar/square-based pyramidal, and square-based pyramidal/square-based pyramidal coordination environments and ligands able to stabilize low oxidation states, such as phosphine derivatives, carbon monoxide, cyanides, hydrides and carbanions, and comprise metal-metal distances in the quite narrow range 2.53–2.76 Å, with a mean value of 2.62 Å (d⁹-d⁹ columns in Figure 9, 67 items). Only two binuclear complexes are known showing hepta-coordinated platinum Pt^III-Pt^III or Pt^IV-Pt^IV metal ions (Pt7-Pt7 purple columns in Figure 9) [54,55].

The complex [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O belongs to the very unusual category of discrete Pt-dimers, featuring one square-planar and one octahedral platinum center connected by a metal-metal bond (d⁹-d⁷, Pt4-Pt6 red column in Figure 9) and formally considered as mixed-valence Pt^I-Pt^III systems. It is interesting to note that the only two examples known of binuclear complexes formally sharing a Pt^I-Pt^III bond, namely, [Pt₂^I,III(tfepma)₂X₄] [X = Cl, Br; tfepma = bis(bis(trifluoroethoxy)-phosphino)methylamine], contain the same trifluoroethyl-imidophosphito ligand bridging the metal centers counterbalanced by halides that complete the platinum coordination spheres [46,47]. Our compound would be the first example supported by a macrocyclic ligand that does not bridge the two metal centers. In these complexes the Pt-Pt distance is quite short [2.6187(7) and 2.6270(9) Å for X = Cl and Br, respectively], and the coordination environment is distorted octahedral for the reportedly Pt^III center and square-planar for the Pt^I one. This structural feature seems to be peculiar to formally defined d⁹-d⁷ Pt₂^I,III binuclear complexes, In fact, d⁷-d⁷ binuclear Pt₂^III,III complexes, also characterized by short metal-metal distances, see above, generally feature both metal centers either in distorted octahedral environments or octahedral/square-based pyramidal coordination spheres. In contrast, binuclear d⁸-d⁸ Pt₂^II,II complexes are characterized by both metal centers in a distorted square-based pyramidal environment.

Following a synthetic procedure analogous to that adopted for the preparation of [Pd(L³)Cl](PF₆), we were able to isolate a compound corresponding to the formulation [Rh(L³)Cl₂](BF₄)·MeCN from the reaction of L³ with RhCl₃·H₂O in MeCN/H₂O (1:1 v/v) (see Materials and Methods section and Figure S33 for FAB Mass Spectrum). ¹³C- and ¹H-NMR spectra (Figures S34 and S35, respectively) presented features similar to those observed in the corresponding NMR spectra of [Rh(L¹)Cl₂](PF₆), suggesting that the complex of Rh^III with L³ has structural features similar to those found for [Rh(L¹)Cl₂](PF₆). An X-ray diffraction analysis was undertaken on the crystals obtained by slow diffusion of Et₂O vapors into a MeCN solution of the crude product.

Indeed, the pseudo-octahedral coordination environment around the Rh^III metal center in the complex cation [Rh(L³)Cl₂]⁺ (Figure 10) resembles that observed for [Rh(L¹)Cl₂]⁺ (Figure 5), with small variations on the structural parameters (bond distances and angles). In both complex cations [Rh(L)Cl₂]⁺ (L = L¹, L³), the ligands adopt a similar folded conformation upon coordination, leaving the other two coordination sites in a relative cis orientation occupied by two Cl⁻ ligands.

In the crystal, [Rh(L³)Cl₂]⁺ complex units are joined in chains running along the [100] direction via NH2···Cl2’ H-bonds (Figure 11a). Chains of complex cation interact with each other via soft-soft Cl···S weak interactions to form undulated sheets in the (101) plane (see Figure 11b).

Sheets of this kind stack along the [010] direction and are connected through an intricate network of C-H···F hydrogen bonds involving the BF_4⁻ counter-anions. The MeCN molecules sit in between the layers and are anchored to the complex cation via an H-bond: N1S···H9^v 2.565 Å, N1S···C9^v 3.435(11) Å, N1S···H9^v-C9^v 146° (v = −x, −y, 1 − z).

DFT Calculations on the Complex Cation [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺

Theoretical calculations carried out at the Density Functional Theory (DFT) [56] level represent an invaluable tool in understanding the electronic structure of metal complexes. They have been extensively used by chemists to investigate structure/property relationships in a large variety of compounds, in fields as varied as material science and bioinorganic chemistry [57,58,59,60], including metal···metal interactions [61,62,63,64]. Within the variety of DFT methods available, the broken symmetry approach (DFT-BS) proposed by Noodleman [65,66,67], widely used in the field of molecular magnetism for the calculation of exchange coupling constants in multinuclear complexes [68], allows unrestricted calculations for low-spin open-shell molecular systems, in which the α and β electrons are allowed to be localized on different atomic centers. This approach was applied to the complex cation [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ and the previously reported related system [Pt₂(tfepma)₂Cl₄] [46,47] in an attempt to elucidate the oxidation state of the platinum ions in these species. For the latter, a mixed-valence d⁹-d⁷ Pt^I-Pt^III nature was proposed for the binuclear system. Based on the excellent results previously obtained on several coordination compounds containing group 10 metal ions [69,70], the mPW1PW functional [71] was adopted, in combination with the full-electron split valence basis sets (BSs) def2-SVP [72,73] for all atomic species but Pt, for which the LANL08(f) basis set [74], including pseudopotentials, was adopted to keep into account relativistic effects of the core electrons [75].

The metric parameters of the complex cation [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ and [Pt₂(tfepma)₂Cl₄] were optimized, starting from crystal structure data, in their closed-shell singlet state (2S + 1 = 1; Figure S36a) and triplet ground-state (2S + 1 = 3, two unpaired electrons; Figure S36b). The complexes were then modeled at the DFT-BS level (Figure S36c), starting from different configurations with different combinations of charges (Q = +1, +2, +3) and spin multiplicities (2S + 1 = 1, 2, 3) assigned to the two Pt ions.

These combinations correspond to all the possible configurations for both d⁸-d⁸ Pt^II-Pt^II [singlet configuration, I in Figure S36c, and triplet configuration, with the unpaired electrons either on the octahedrally (Pt_O) or square-planar (Pt_SP) coordinated Pt ions, II and III in Figure S36c, respectively] and mixed-valence d⁹-d⁷ Pt^I-Pt^III systems (singlet and triplet configurations, with either Pt_O or Pt_SP carrying the Q = +1 charge; IV–VII in Figure S36c).

This scheme includes the mixed-valence configuration previously reported for [Pt₂(tfepma)₂Cl₄], where the charge Q = +3 was assigned to the Pt_O center, and the charge Q = +1 to the Pt_SP one (VI–VII in Figure S36c). When the optimization of the two model compounds was performed starting from the electron density guess of ground-state configurations I–VII, all calculations converged to two geometries only, corresponding to the closed-shell singlet and triplet ground-states.

According to these results, it appears that the spin densities on the two metal centers in this type of binuclear complexes cannot be separately modeled, probably because of the close proximity of the Pt ions, therefore both should be better described as binuclear d⁸-d⁸ Pt^II-Pt^II complexes. An examination of the optimized geometries in the singlet and triplet ground-states shows that for both complexes the total electronic energy of the geometry in the singlet state is lower than that in the triplet state (by 111.1 and 133.7 kJ mol⁻¹ for [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ and [Pt₂(tfepma)₂Cl₄], respectively). Accordingly, a better agreement between the optimized geometry and the structural data was found for both complexes for the singlet ground-state. In the case of [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺, the bond lengths and angles of the optimized geometry in the singlet state differ from the experimental ones by less than 0.05 Å and 6°, respectively, with the sole exception of the Pt2-N2 bond distance, which is overestimated by 0.111 Å (Tables S1–S3, Figure S37). On the other hand, in the optimized geometry in the triplet state, a significant elongation (about 0.3 Å) of the Pt-S bonds within the coordination sphere of the octahedral Pt ion is observed, along with a divergence from the square-planar coordination geometry for Pt1, with a Cl1-Pt1-Pt2-N2 dihedral angle of 89.51° (Figure S37). In the case of [Pt₂(tfepma)₂Cl₄], an even further deviation from the experimental geometry was observed for the optimized geometry in the triplet state, featuring both Pt ions pentacoordinated in a trigonal bipyramidal geometry, while a very good agreement was found between the experimental structure and that optimized in the singlet ground-state (Tables S4–S6, Figure S38).

These data suggest that both complexes are better described as featuring closed-shell singlet Pt₂ systems. Accordingly, the Kohn-Sham (KS) frontier molecular orbitals (MOs) are distributed between the two Pt ions in both complexes (Figure 12). Moreover, a natural population analysis (NPA) shows comparable natural charges on Pt_O (Q = 0.047 and −0.862 |e|) and Pt_SP (Q = 0.243 and −0.367 |e| for [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ and [Pt₂(tfepma)₂Cl₄], respectively). Consistently, very similar natural electron configurations were found in both complexes for the two Pt atoms, the electron populations differing by less than 0.5 |e|.

Finally, Wiberg (0.472 and 0.507) and Mayer bond indices (0.375 and 0.502 for [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ and [Pt₂(tfepma)₂Cl₄], respectively) for the Pt-Pt bond, suggest for both complexes a non-negligible covalent character.

3. Materials and Methods

All melting points are uncorrected. Elemental analyses were obtained using a Fison EA CHNS-O instrument operating at 1000 °C. FAB mass spectra were measured at the EPSRC National Mass Spectrometry Service at Swansea (UK). ¹H and ¹³C NMR experiments were conducted at 25 °C with a Varian VXR400 spectrometer operating at 400 MHz for ¹H and 100.62 MHz for ¹³C or with a Bruker Avance III HD spectrometer operating at 600 MHz for ¹H and 150.9 MHz for ¹³C, using TMS as an internal standard. Data are reported as chemical shifts (multiplicity, coupling constants where applicable, number of hydrogen atoms, and assignment where possible). Abbreviations are: s (singlet), d (doublet), t (triplet), m (multiplet). Coupling constants (J) are quoted in Hertz (Hz) to the nearest 0.1 Hz. Absorption spectra were recorded with Varian Model Cary 5 UV-Vis-NIR spectrophotometer. Solvents and starting materials were purchased from commercial sources and used as received. Solvents used for the synthesis of the ligands were dried following conventional methods. The ligands 5-oxa-2,8-dithia[9](2,6)-pyridinophane (L¹) and 2,5,8-trithia[9](2,6)-pyridinophane (L²), were synthesized according to reported procedures using as starting materials 2,6-dichloromethylpyridine and the dithiols O(HSCH₂CH₂)₂ and S(HSCH₂CH₂)₂ for L¹ and L², respectively [22,33]. 2,8-Dithia-5-aza-2,6-pyridinophane (L³) was prepared according to the reported procedure starting from 2,6-dithiomethyl-pyridine and N-(tert-butoxycarbonyl)bis-(2-chloroethylamine) [13].

3.1. General Procedure for the Synthesis of the Pd^II, Pt^II, Rh^III Complexes of L¹-L³

The appropriate metal chloride was added to a solution of L, ¹L² or L³ in 1:1 molar ratio. No excess of ligand or starting metal salt was considered, to avoid the formation of coordination compounds with stoichiometries other than 1:1, which is very likely when the macrocyclic ligand is not able to satisfy the stereo-electronic requirements of the metal ion. The reactions were all conducted in MeCN/H₂O (20 mL, 1:1 v/v) solvent mixture, except in two cases where the MeOH/H₂O (20 mL, 1:1 v/v) solvent mixture was used for solubility reasons, as specified below. The reaction mixture was refluxed under N₂ for 5 h in all cases. When a pure solid product was not obtainable/isolable from the reaction mixture of the ligands with the chloride of the metal under investigation, a counter-anion metathesis reaction was performed to replace coordinating chlorido ligands with non-coordinating BF₄⁻ or PF₆⁻ anions and facilitate the crystallization or formation of solid products. This was necessary in the preparation of [Pt(L¹)Cl](BF₄), [Rh(L¹)Cl₂](PF₆), [Rh(L²)Cl₂](PF₆), [Pd(L³)Cl](PF₆), [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O, [Rh(L³)Cl₂](BF₄)·MeCN for which a ten-fold molar excess NH₄BF₄ or NH₄PF₆ was added at room temperature after refluxing of the reaction mixture. This is a well-established synthetic procedure in the field of macrocyclic ligand chemistry [76,77].

[Pd(L¹)Cl]₂[Pd₂Cl₆]. To a solution of L¹ (0.051 g, 0.211 mmol) in MeOH/H₂O (20 mL, 1:1 v/v) was added PdCl₂ (0.037 g, 0.226 mmol). Reddish prismatic crystals (0.048 g, yield 19%) were obtained by reduction under the vacuum of the volume of the reaction mixture and subsequent crystallization in the air from the remaining aqueous solution by slow evaporation. Mp: 220 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₅Cl₄NOPd₂S₂): C, 22.4 (22.2); H, 2.6 (2.5) N, 2.5 (2.4); S, 10.6 (10.8). ¹H-NMR (600 MHz, CD₃CN): δ_H 2.74–2.78 (m, 2H, H_9a/H_10a or H_9b/H_10b), 3.32–3.34 (m, 2H, H_8a/H_11a or H_8b/H_11b), 3.54–3.59 (m, 2H, H_8b/H_11b or H_8a/H_11a), 4.07–4.10 (m, 2H, H_9b/H_10b or H_9a/H_10a), 4.45 (d, J = 18.5 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 4.88 (d, J = 18.5 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.53 (d, J = 7.9 Hz, 2H), 7.97 (t, J = 7.9 Hz, 1H). ¹³C-NMR (150.9 MHz, CD₃CN): δ_C 45.0 (ArCH₂S), 45.9 (SCH₂CH₂O), 65.4 (SCH₂CH₂O), 122.8, 140.5, 164.1 (aromatic carbons). UV-Vis spectrum (H₂O): λ (ε) 211 (2390), 261 (2190), 388 nm (190 dm³ mol⁻¹cm⁻¹). MS (FAB): m/z 419 ([C₁₁H₁₅Cl₂NOPdS2]⁺).

[Pt(L¹)Cl](BF₄). To a solution of L¹ (0.035 g, 0.145 mmol) in MeOH/H₂O (20 mL, 1:1 v/v) was added PtCl₂ (0.038 g, 0.145 mmol). Yellow crystals (0.040 g, yield 49%) were obtained after the addition of NH₄BF₄ to the mixture, reduction under the vacuum of the volume of the reaction mixture and subsequent crystallization in the air from the remaining aqueous solution. Mp: 200 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₅BClF₄NOPtS₂): C, 23.7 (23.6); H, 2.3 (2.7) N, 2.9 (2.5); S, 11.3 (11.5). ¹H-NMR (600 MHz, CD₃CN): δ_H 2.58–2.62 (m, 2H, H_9a/H_10a or H_9b/H_10b), 3.48–3.61 (m, 4H, H_8a/H_11a or H_8b/H_11b), 4.09–4.11 (m, 2H, H_9b/H_10b or H_9a/H_10a), 4.62 (d, J = 18.0 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 4.85 (d, J = 18.0 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.61 (d, J = 6.0 Hz, 2H), 8.08 (t, J = 6.0 Hz, 1H). ¹³C-NMR (150.9 MHz, CD₃CN): δ_C 46.7 (ArCH₂S), 47.4 (SCH₂CH₂O), 66.4 (SCH₂CH₂O), 123.3, 140.5, 163.4 (aromatic carbons). UV-Vis spectrum (H₂O): λ (ε) 265 (2360), 360 nm (56 dm³ mol⁻¹ cm⁻¹). MS (FAB): m/z 472 ([C₁₁H₁₅ClNOPtS₂]⁺).

[Rh(L¹)Cl₂](PF₆). To a solution of L¹ (0.040 g, 0.166 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was added RhCl₃·H₂O (0.038 g, 0.166 mmol). Yellow crystals (0.038 g, yield 41%) were obtained after the addition of NH₄PF₆ to the reaction mixture, evaporation under the vacuum of the solvent and subsequent crystallization of the solid obtained, by diffusion of Et₂O vapor into an MeCN solution. Mp: 220 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₅Cl₂F₆NOPRhS₂): C, 24.0 (23.6); H, 2.9 (2.7); N, 2.9 (2.5); S, 11.3 (11.4)%. ¹H-NMR (600 MHz, CD₃CN): δ_H 3.36–3.40 (m, 2H, H_9a/H_10a or H_9b/H_10b), 3.48–3.51 (m, 2H, H_8a/H_11a or H_8b/H_11b), 3.57–3.62 (m, 2H, H_8b/H_11b or H_8a/H_11a), 4.04–4.07 (m, 2H, H_9b/H_10b or H_9a/H_10a), 5.01 (d, J = 18.6 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 5.25 (d, J = 18.6 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.77 (d, J = 8.0 Hz, 2H), 8.06 (t, J = 8.0 Hz, 1H). ¹³C-NMR (150.9 MHz, CD₃CN): δ_C 40.5 (ArCH₂S), 46.0 (SCH₂CH₂O), 74.1 (SCH₂CH₂O), 124.2, 140.0, 162.2 (aromatic carbons). UV-Vis spectrum (MeCN): λ (ε) 322 (727), 371 (770), 424sh nm (544 dm³ mol⁻¹ cm⁻¹). MS (FAB): m/z 414 ([C₁₁H₁₅Cl₂NORhS₂]⁺).

[Pd(L²)Cl]Cl. To a solution of L² (0.058 g, 0.226 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was added PdCl₂ (0.040 g, 0.226 mmol). An orange microcrystalline solid (0.053 g, yield 54%) was obtained by removal under reduced pressure of solvent from the reaction mixture and subsequent crystallization of the crude product obtained by diffusion of Et₂O vapor into a MeCN solution. Mp: 210 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₅Cl₂NPdS₃): C, 30.3 (30.4); H, 3.3 (3.5) N, 3.3 (3.2); S, 22.6 (22.1)%. ¹H-NMR (400 MHz, D₂O): δ_H 2.13–2.20 (m, 4H, H_9a/H_10a), 3.56–3.78 (m, 4H, H_8a/H_11a), 4.70 (d, J = 18.8 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 5.17 (d, J = 18.4 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.69 (d, J = 8.0 Hz, 2H), 8.09 (m,1H). ¹³C-NMR (100.62 MHz, (CD₃)₂CO): δ_C 33.2 (SCH₂CH₂S), 46.9 (ArCH₂S), 48.0 (SCH₂CH₂S), 123.1, 141.1, 164.6 (aromatic carbons). UV-Vis spectrum (H₂O): λ (ε) 265 (7000), 317 (200), 398 nm (80 dm³ mol⁻¹cm⁻¹). MS (FAB): m/z 433 ([C₁₁H₁₅Cl₂NPdS₃]⁺).

[Pt(L²)Cl]Cl. To a solution of L² (0.048 g, 0.186 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was added PtCl₂ (0.050 g, 0.186 mmol). A yellow solid (0.070 g, yield 78%) was obtained by evaporation under the vacuum of the reaction mixture and subsequent crystallization of the crude product obtained by diffusion of Et₂O vapor into a MeCN solution. Mp: 215 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₅Cl₂NPtS₃): C, 24.7 (25.2); H, 3.0 (2.9) N, 3.2 (2.7); S, 18.7 (18.4)%. ¹H-NMR (600 MHz, CD₃CN): δ_H 2.56–2.62 (m, 2H, H_9a/H_10a or H_9b/H_10b), 2.67–2.72 (m, 2H, H_8a/H_11a or H_8b/H_11b), 2.93–2.96 (m, 2H, H_8b/H_11b or H_8a/H_11a), 3.28–3.32 (m, 2H, H_9b/H_10b or H_9a/H_10a), 3.94 (d, J = 12.8 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 4.05 (d, J = 12.8 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.57 (d, J = 8.0 Hz, 2H), 7.83 (t, J = 8.0 Hz, 1H). ¹³C-NMR (150.9 MHz, CD₃CN): δ_C 34.3 (SCH₂CH₂S), 39.2 (ArCH₂S), 39.8 (SCH₂CH₂S), 124.5, 138.6, 158.4 (aromatic carbons). UV-Vis spectrum (H₂O): λ (ε) 265 (7500), 317 nm (120 dm³ mol⁻¹cm⁻¹). MS (FAB): m/z 488 ([C₁₁H₁₅ClNPtS₃]⁺).

[Rh(L²)Cl₂](PF₆). To a solution of L² (0.040 g, 0.155 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was added RhCl₃·H₂O (0.035 g, 0.155 mmol). A brown solid (0.034 g) was obtained after the addition of NH₄PF₆ to the mixture, evaporation of solvent from the reaction mixture under vacuum and subsequent crystallization in the air from the aqueous solution. Mp: 230 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₅PCl₂F₆NRhS₃): C, 22.4 (22.9); H, 2.3 (2.6); N, 3.0 (2.4); S, 16.1 (16.7)%. ¹H-NMR (600 MHz, CD₃CN): δ_H 2.59–2.65 (m, 2H), 2.73–2.79 (m, 2H), 3.31–3.36 (m, 2H), 3.46–3.51 (m, 2H), 3.57–3.63 (m, 2H), 3.86–3.94 (m, 4H), 4.03–4.06 (m, 2H), 4.81 (d, J = 18.2 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 4.87 (d, J = 18.3 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 5.14–5.18 (m, 4H), 7.69 (d, J = 7.9 Hz, 2H), 7.74 (d, J = 7.9 Hz, 2H), 8.02 (t, J = 7.9 Hz, 1H), 8.11 (t, J = 7.9 Hz, 1H) ¹³C-NMR (150.9 MHz, CD₃CN): δ 36.2/36.6, 44.5/45.1, 45.4/46.8, 123.8/124.6, 139.9/141.2, 161.4/162.3. UV-Vis spectrum (MeCN): λ (ε) 325 (730), 365 (765), 421sh nm (540 dm³ mol⁻¹cm⁻¹). MS (FAB): m/z 430 ([C₁₁H₁₅Cl₂NRhS₃]⁺), 396 (C₁₁H₁₅ClNRhS₃]⁺). One of the two sets of signals, in both the ¹H- and ¹³C-NMR spectra belongs to the species [Rh(L²)Cl₂](PF₆), the other set belongs to a species having only one coordinated chlorido ligand (see discussion above). However, it is not possible to uniquely identify which set of signals corresponds to which complex.

[Pd(L³)Cl](PF₆). A mixture of L³ (0.020 g, 0.083 mmol) and PdCl₂ (0.015 g, 0.083 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was refluxed for 2 h. A brown solid (0.020 g, yield 46%) was obtained after the addition of excess NH₄PF₆ to the reaction mixture, reduction of the volume of the reaction mixture under vacuum, and subsequent crystallization in the air from the resulting aqueous solution by slow evaporation. Mp: 220 °C with decomposition. Elem. Anal. found (calc. for C₁₁H₁₆ClF₆N₂PPdS₂): C, 25.4 (25.1); H, 2.8 (3.1); N, 5.7 (5.3); S, 12.6 (12.2)%. ¹H-NMR (400 MHz, D₂O): δ_H 3.03–3.06 (m, 4H, H_9a/H_10a), 3.54–3.60 (m, 4H, H_8a/H_11a), 4.61 (d, J = 18.6 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 5.07 (d, J = 18.0 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.65 (d, J = 8.1 Hz, 2H), 8.04 (m,1H). ¹³C-NMR (100.62 MHz, CD₃CN): δ_C 43.3 (SCH₂CH₂N), 45.8 (ArCH₂S), 48.2 (SCH₂CH₂N), 122.1, 139.5, 163.6 (aromatic carbons). UV-Vis spectrum (MeCN): λ (ε) 274 (12480), 374 nm (3880 dm³ mol⁻¹cm⁻¹). MS (FAB): m/z 381 ([C₁₁H₁₆ClN₂PdS₂]⁺).

[Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O. To a solution of L³ (0.020 g, 0.083 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was added PtCl₂ (0.022 g, 0.083 mmol), and the reaction mixture was refluxed for 2 h. Very few orange crystals were obtained after the addition of NH₄BF₄ to the mixture, reduction of the volume of the reaction mixture under vacuum, and subsequent crystallization in the air from the resulting aqueous solution by slow evaporation. Mp: 230 °C with decomposition. Elem. Anal. found (calc. for C₁₅H₂₅B₂ClF₈N₄O₂Pt₂S₂): C, 18.4 (18.8); H, 2.8 (2.6) N, 5.7 (5.9); S, 6.5 (6.7)%.

[Rh(L³)Cl₂](BF₄)·MeCN. To a solution of L³ (0.020 g, 0.083 mmol) in MeCN/H₂O (20 mL, 1:1 v/v) was added RhCl₃·H₂O (0.019 g, 0.083 mmol). Yellow crystals (0.020 g, yield 44%) were obtained after the addition of NH₄BF₄ to the mixture, reduction of the reaction mixture under vacuum and subsequent crystallization of the solid obtained by diffusion of Et₂O vapors into a MeCN solution. Mp: 210 °C with decomposition. Elem. Anal. found (calc. for C₁₃H₁₉BCl₂F₄N₃S₂Rh): C, 28.7 (28.8); H, 3.2 (3.5); N, 7.6 (7.7); S, 12.2 (11.8)%. ¹H-NMR (400 MHz, CD₃CN): δ 2.39–2.48 (m, 2H, H_9a/H_10a or H_9b/H_10b), 3.01–3.09 (m, 2H, H_8a/H_11a or H_8b/H_11b), 3.33–3.42 (m, 2H, H_8b/H_11b or H_8a/H_11a), 3.55–3.62 (m, 2H, H_9b/H_10b or H_9a/H_10a), 4.93 (d, J = 18.4 Hz, 2H, H_7a/H_12a or H_7b/H_12b), 5.21 (d, J = 18.4 Hz, 2H, H_7b/H_12b or H_7a/H_12a), 7.71 (d, J = 8.4 Hz, 2H), 8.04 (t, J = 8.0 Hz, 1H). ¹³C-NMR (100.62 MHz, CD₃CN): δ 42.6 (SCH₂CH₂N), 46.6 8 (ArCH₂S), 52.6 (SCH₂CH₂N), 124.4, 140.2, 162.7 (aromatic carbons). UV-Vis spectrum (MeCN): λ (ε) 274 (3110), 380 nm (933 dm³ mol⁻¹cm⁻¹). MS (FAB): m/z 413 ([C₁₁H₁₆Cl₂N₂RhS₂]⁺).

3.2. Theoretical Calculations

Theoretical calculations were performed on the complex cation [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]²⁺ and on [Pt₂(tfepma)₂Cl₄] (tfepma = ((CF₃CH₂O)₂P)₂NCH₃) [46,47] at the density functional theory (DFT) [56] level with the Gaussian 16 (Rev. B.01) suite of programs [78], on a IBM x3755 server with four 12-core processors and 64 Gb of RAM (OS: SUSE Linux Enterprise Server 11 SP3). The mPW1PW functional [71] was adopted, in combination with the full-electron split valence basis sets (BSs) def2-SVP [72,73] for light atoms and the LANL08(f) BS [74], including f polarization functions for the outer electron shell and Relativistic Effective Core Potentials (RECPs) [75], for the Pt atomic species. All basis sets and RECPs were obtained from Basis Set Exchange and Basis Set EMSL Library [79].

The geometries of all compounds were optimized starting from crystal structure data in their triplet ground-state (2S + 1 = 3, two unpaired electrons), closed-shell singlet state (2S + 1 = 1) after verification of the wavefunction stability (stable = opt), or by means of a broken-symmetry (DFT-BS) approach. The procedure recently developed for bis(1,2-dithiolene) metal complexes was followed [70]. In particular, the BS electron density guess was obtained through a fragmented approach (guess = fragment = n, the fragments being the two Pt ions and the various ligands) starting from the geometry optimized at the largest spin multiplicity, by attributing different combinations of charges (Q = +1, +2, +3) and corresponding spin multiplicities (2S + 1 = 1, 2, 3) to the Pt ions, eventually optimizing (opt) the geometry of the complexes for the different combinations and verifying (and in case re-optimizing) the stability of the wavefunctions [70]. Fine numerical integration grids (Integral = ultrafine keyword) were used, and the nature of the minima of each optimized structure was verified by harmonic frequency calculations (freq = raman keyword). A natural population analysis was carried out at the optimized geometries using the natural bonding orbital (NBO) partitioning scheme [80]. The programs GaussView 6.0.16 [81], Molden 6.6 [82], and Chemissian 4.53 [83] were used to investigate the optimized structures and the shapes of Kohn–Sham molecular orbitals.

3.3. X-ray Crystallography

A summary of the crystal data and refinement details for the compounds discussed in this paper is given in Tables S7 and S8 (SM). Diffraction data for [Pd(L¹)Cl]₂[Pd₂Cl₆] and [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O were collected at 293(2) K as ω scans on an APEX II CCD Diffractometer. Diffraction data for [Rh(L¹)Cl₂](PF₆) were collected at 294(2) K as ω scans on an ENRAF NONIUS CAD4 Diffractometer. For [Pt(L¹)Cl](BF₄) and [Rh(L³)Cl₂](BF₄)·MeCN diffraction data were collected at 150(2) K as ω scans on, respectively, a Bruker SMART1000 CCD Area Detector Diffractometer and a Bruker SMART-APEX CCD Area Detector Diffractometer equipped with an Oxford Cryosystem open-flow cryostat. Data were collected using graphite-monochromated MoKα radiation (λ = 0.71073 Å). Absorption corrections were treated by semiempirical corrections based on multiple scans, as specified in the CIF files.

All the structures were solved by direct methods using SIR92 [84] ([Pd(L¹)Cl]₂[Pd₂Cl₆], [Rh(L¹)Cl₂](PF₆), [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O, [Rh(L³)Cl₂](BF₄)·MeCN) and SHELXS-97 [85] ([Pt(L¹)Cl](BF₄)) and completed by iterative cycles of full-matrix least-squares refinement and ΔF syntheses using the software package SHELXL [85]. For [Rh(L³)Cl₂](BF₄)·MeCN diffraction was poor in certain directions; twinning was modeled as non-merohedral by 180° rotation about the [100] axis with a twin fraction of 0.0463(1). In all cases, non-H atoms were refined with anisotropic displacement parameters, while H atoms were introduced at calculated positions and refined using a riding model. In [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O and [Rh(L³)Cl₂](BF₄)·MeCN the N-H hydrogen atom on the amidate ligand and the solvent methyl H-atoms, respectively, were found in a difference Fourier map and thereafter refined using a riding model.

4. Conclusions

In this paper, the coordination chemistry of the mixed-donor tetradentate macrocycles L¹-L³ featuring a pyridine moiety towards platinum group metal ions Pd^II, Pt^II, and Rh^III has been investigated. In all isolated 1:1 metal-to-ligand complexes, the ligands adopt a folded conformation and impose a [3 + 1] coordination mode at the Pd^II and Pt^II metal centers within a distorted square-based coordination sphere. In the case of Rh^III complexes, the tetradentate ligands occupy four of the six positions of a distorted octahedral geometry with the other two coordination sites in a relative cis orientation occupied by two Cl⁻ ligands. A rare example of a discrete Pt₂ dimer was isolated by serendipity from the reaction of L³ and PtCl₂ in refluxing MeCN/H₂O (1:1 v/v), and structurally characterized. This complex, based on data from the literature, could have been formally defined as a d⁹-d⁷ Pt₂^I,III mixed-valence binuclear complex featuring a Pt-Pt bond linking a square-planar and an octahedral platinum centers. DFT calculations, following the broken symmetry approach (DFT-BS), identify a singlet ground-state nature (d⁸-d⁸ Pt^II-Pt^II) both for the isolated compound, as well as for the only other example of a discrete binuclear Pt₂ complex of the same type reported in the literature, despite the different coordination environments of the two metal centers typical for a d⁸ Pt^II center (square-planar) and a d⁷ Pt^III center (octahedral). Notwithstanding the theoretical limits inherent to a non-multireference DFT-BS approach, the case of [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]⁺ suggests that a more in-depth re-evaluation may be needed for the electronic configurations assigning mixed oxidation states to Pt ions in dinuclear complexes where two directly interacting Pt ions show different coordination geometries and numbers.

The obtained results, especially in the case of L³, can be of help in understanding the sensing properties toward metal ions of fluorescent chemosensors featuring this macrocycle as receptor units [13,14,15,16,17,18,19,20,21]. However, as far as the question posed in the title is concerned, based on the results obtained, we can conclude that well-established fields of coordination chemistry, such as that of macrocyclic ligands and Pt^II, can still hold some unexpected outcomes. The serendipitous and unexpected isolation of complex [Pt(L³)(η-1,3-MeCONH)PtCl(MeCN)](BF₄)₂·H₂O stands as a proof of principle for the unexplored synthetic possibilities still available in the coordination chemistry of well-known classes of macrocyclic ligands and Platinum Group metals. We have shown that the dimeric and unique complex cation [Pt(L³)(μ-1,3-MeCONH)PtCl(MeCN)]⁺ can exist, despite the fact we were not able to reproduce it or explain its formation. This is still interesting and could open new perspectives in the coordination compounds of Pt^II. Furthermore, we strongly believe that it is essential to perform fundamental research even when all the available information suggests that only predictable, trivial results will be obtained.