Richard
Chlebík
a,
Csilla
Fekete
b,
Roman
Jambor
a,
Aleš
Růžička
a,
Zoltán
Benkő
*bc and
Libor
Dostál
*a
aDepartment of General and Inorganic Chemistry, University of Pardubice, Studentská 573, CZ 532 10 Pardubice, Czech Republic. E-mail: libor.dostal@upce.cz
bDepartment of Inorganic and Analytical Chemistry, Budapest University of Technology and Economics, Műegyetem rkp 3, H-1111 Budapest, Hungary. E-mail: benko.zoltan@vbk.bme.hu
cHUN-REN-BME Computation Driven Chemistry Research Group, Műegyetem rkp 3, H-1111 Budapest, Hungary
First published on 8th October 2024
A set of antimony(III) compounds containing a ligand (Ar) with a pendant guanidine function (where Ar = 2-[(Me2N)2CN]C6H4) was prepared and characterized. This includes triorgano-Ar3Sb, diorgano-Ar2SbCl and monoorgano-ArSbCl2 compounds and they were characterized by 1H and 13C NMR spectroscopy and by single-crystal X-ray diffraction analysis (sc-XRD). The coordination capability of Ar3Sb and Ar2SbCl was examined in the reactions with either cis-[PdCl2(CH3CN)2] or PtCl2 and complexes cis-[(κ2-Sb,N-Ar3Sb)MCl2] (M = Pd 1, Pt 2) and [(κ3-N,Sb,N-Ar2SbCl)MCl2] (M = Pd 3, Pt 4) were isolated, while their structures were determined by sc-XRD. Notably, the ligands Ar3Sb and Ar2SbCl exhibit different coordination modes – bidentate and tridentate, respectively – and the antimony exhibits three distinct bonding modes in complexes 1–4, which were also subjected to theoretical studies.
![]() | ||
Fig. 1 Examples of Z-type ligands based on the antimony σ-acceptor (A and B) and complexes from the current study (C). |
![]() | ||
Fig. 2 Molecular structure of starting antimony compounds showing 30% probability ellipsoids and the crystallographic numbering scheme. Hydrogen atoms are omitted for clarity. |
In the case of Ar3Sb, the central atom adopts the expected trigonal pyramidal geometry. All the Sb–N distances (in the range 2.983(2)–3.059(2) Å) are similar to each other and are far longer than expected for a Sb–N covalent bond (Σcov(Sb, N) = 2.11 Å),18 thereby ruling out any significant intramolecular N → Sb interaction. In contrast, there is an obvious difference between the Sb1–N1/2 distances in Ar2SbCl, i.e. 2.5440(17) vs. 3.0906(18) Å. These values are comparable to those of the related compound [2-(Me2N)C6H4]Sb(Mes)Cl (2.619(3) Å),19 Finally, the Sb1–N1 distance of 2.4382(15) Å is the shortest in this series being consistent with the highest Lewis acidity of the central atom, but slightly longer than in, e.g., [2-(Me2NCH2)C6H4]SbCl2 (2.407(5) Å)20 and [2-(DippNCH)C6H4]SbCl2 (2.416(2) Å).21 The enhanced Lewis acidity of the antimony atom also leads to the formation of intermolecular Sb(1)–Cl(1)⋯Sb(1a) contacts (3.225 Å; cf. Σcov(Sb, Cl) = 2.39 Å).18
Their coordination toward either cis-[PdCl2(CH3CN)2] or PtCl2 was examined. Unfortunately, no stable complexes could be isolated from the respective reactions with ArSbCl2. Nevertheless, Ar3Sb and Ar2SbCl provided the compounds cis-[(κ2-Sb,N-Ar3Sb)MCl2] (M = Pd 1, Pt 2) and [(κ3-N,Sb,N-Ar2SbCl)MCl2] (M = Pd 3, Pt 4, Scheme 1), respectively.
Ar3Sb behaves as an Sb,N-chelate in 1 and 2 containing tetracoordinated Pd and Pt centers in a square planar coordination environment (Fig. 3). To our knowledge, these complexes represent rare structurally characterized examples containing five-membered C2SbNM metallacycles (M = Pd or Pt) as only a few related six-membered metallacycles have been reported: i.e. cis-[κ2-Sb,N-(2-(Me2NCH2)C6H4SbMes2)PdCl2],22cis-[κ2-Sb,N-(2,6-(Me2NCH2)2C6H3SbPh2)PtCl2],23cis-[κ2-Sb,N-(2-(Me2NCH2)C6H3)3SbPtCl2]24 and cis-[κ2-Sb,N-(2-(Me2NCH2)Fc)3SbPtCl2].25 The Sb1–Pd/Pt bond lengths of 2.4573(7)/2.4661(7) Å in 1/2 are a little shorter than those found in the abovementioned six-membered cycles (cf. Sb–Pd 2.4831(5) and Sb–Pt 2.4935(5)–2.5162(4) Å) and the same applies to the N1–Pd/Pt bonds of 2.082(3)/2.072(5) Å (cf. literature data 2.096(5)–2.133(4) Å).22–25
![]() | ||
Fig. 3 Molecular structures of 1 and 2 showing 30% probability ellipsoids and the crystallographic numbering scheme. Hydrogen atoms are omitted for clarity. |
The M–Cl bond located trans to the antimony is elongated to 2.4004(12) vs. 2.2988(13) Å in 1 and 2.3863(14) vs. 2.3024(15) Å in 2. In line with the solid-state structures, the 1H and 13C NMR spectra of 1 and 2 revealed two sets of signals for the aryl ligands in a 1:
2 integral ratio, indicating that only one of the guanidines forms a chelate with the metal. The methyl groups within this set of signals are magnetically non-equivalent (δ(13C) = 40.4/41.7 and 40.1/42.0 ppm, for 1 and 2, respectively) suggesting hindered rotation around the phenyl–N bond due to the closure of this metallacycle. The formation of the chelate is also reflected in a pronounced down-field shift of one of the guanidine carbons N3C 159.4/168.9 and 159.9/169.6 ppm in 1 and 2, respectively, and this is also consistent with the value found in the non-coordinated parent Ar2SbCl (158.9 ppm). Although both compounds were shown to be unstable at elevated temperatures by high-temperature 1H NMR experiments (Fig. S17 and S18†), the 1H, 1H-EXSY experiment (with a mixing time of 1 s) in the case of palladium complex 1 clearly showed chemical exchange between all accessible guanidine donors in solution (Fig. S10†), proving its dynamic behavior. In contrast, similar experiments with 2 (with a mixing time of up to 3 s) revealed chemical exchange only in the region of the methyl groups of the coordinated guanidine function, while solely NOESY cross-peaks were obtained in the aromatic region of the spectra (Fig. S14–S16†). This suggests that the metallacycle in complex 2 is more rigid in solution.
In marked contrast to the conventional coordination mode in 1 and 2, Ar2SbCl in 3 and 4 behaves as an N,Sb,N-pincer ligand (Scheme 1 and Fig. 4) pointing to a Z-type coordination of the antimony atom (vide infra). The central metals are coordinated by two chlorine and two nitrogen atoms in a cis fashion. The M–N (2.055(3) and 2.060(3) Å for 3; 2.049(5) and 2.049(7) Å for 4) and M–Cl bond lengths (2.3004(10) and 2.3138(10) Å for 3; 2.3085(18) and 2.319(2) Å for 4) are in the expected range for single bonds (cf. Σcov(Pd/Pt, N) = 1.91/1.94 Å; Σcov(Pd/Pt, Cl) = 2.19/2.22 Å).18 Importantly, the coordination number of the metals is increased to five by the coordination of the antimony atom, giving rise to a distorted square-pyramidal geometry for the Pt and Pd atoms. The atoms-in-molecules (AIM) study (see the ESI†) also located five bond critical points around the Pd/Pt centers (ωB97X-D/def2-TZVP level). This observation contrasts with the coordination non-innocent behaviour of the related systems containing pendant P-donors, where the coordination is accompanied by the formation of a new Sb–Cl bond, thereby preserving the square planar environment at the metal centers (Fig. 1B).14b,15b,i The bond lengths Pd–Sb 2.8829(5) Å in 3 and Pt–Sb 2.8163(6) in 4 are elongated compared to those in 1/2 and with the closest counterparts from Fig. 1B (R = Cl), cf. Sb–Pd 2.4230(3)14b and Sb–Pt 2.4407(5) Å.15b Nevertheless both values are still close to the respective Σcov(Pd/Pt, Sb) = 2.60/2.63 Å18 indicating significant bonding interactions (vide infra). The coordination sphere of the antimony atom in 3 and 4 is completed by a chlorine atom (Sb–Cl 2.4605(12)/2.515(2) Å for 3/4) that is coordinated trans to the metal (Cl–Sb–M 171.35(3)/170.91(5)° for 3/4) and is best described as having a see-saw geometry bearing a stereochemically active lone pair. Although complex 4 after isolation remains insoluble, the solubility of 3 was at an acceptable level that allowed recording of 1H and 13C NMR spectra, which revealed two sets of signals for both ligands in a mutual 1
:
1 integral ratio. The chemical shifts of both N3C guanidine carbons at 165.9/169.9 are consistent with coordination toward the palladium atom, when compared with the values found in 1 and 2 (vide supra).
![]() | ||
Fig. 4 Molecular structure of 3 and 4 showing 30% probability ellipsoids and the crystallographic numbering scheme. Hydrogen atoms are omitted for clarity. |
To explore the bonding situations in the experimentally obtained complexes, quantum chemical calculations were performed using various DFT functionals (ωB97X-D, BP86, BP86-D3, B3LYP-D3, and M06-2X) in combination with the def2-TZVP basis set and the more accurate LNO-CCSD(T)/aug-cc-pVTZ level. Based on comparing the computed geometrical parameters with those obtained from sc-XRD experiments, the ωB97X-D (and BP86-D3) functional shows the best performance and, in the following, we only discuss the findings obtained at the ωB97X-D/def2-TZVP level. To scrutinize the nature of possible metal–antimony interactions that may stabilize these complexes, we performed NBO (Natural Bonding Orbital) calculations. In the case of complexes 1 and 2, two important stabilizing donor–acceptor interactions were found: one from the N atom and another from the Sb center into the σ*(M–Cl) antibonding orbital (Fig. 5A). In contrast, for the Pd–Sb interaction in complex 3, the NBO analysis identifies two different types of interactions that are similar to those in a previous Au complex reported by the Gabbaï group:11a (i) a donation from the d-orbital at the palladium center into the σ* antibonding orbital of the Sb–Cl bond, lp(Pd) → σ*(Sb–Cl) (Fig. 5B), with a stabilization energy of only E(2) = 10.6 kcal mol−1 (according to the second order perturbation theory in the Fock matrix); and (ii) additional interactions arising from a donation of the s-type Sb lone pair toward the σ*(Pd–Cl) orbitals with a summarized stabilization energy of 8.1 kcal mol−1 (Fig. 5C). Although complex 4 shows similar characteristics, (i) the donation from the d-type Pt “lone pair” into the σ*(Sb–Cl) orbital exhibits a significantly larger E(2) of 25.7 kcal mol−1 (compared to 3), while (ii) the two lp(Sb) → σ*(Pt–Cl) donations have a lower cumulative stabilization energy of 7.5 kcal mol−1. The deletion procedures on the Fock matrix elements (for the interactions between the Sb–metal and Sb–Cl) in both 3 and 4 also delivered similar results for the E(2) values.
![]() | ||
Fig. 5 Various M–Sb bonding interactions obtained in 1–4 (A–C) and the definition of borane affinity used in the study (D). |
These donor–acceptor interactions also affect the geometrical parameters. On one hand, the backdonation from the M center into the σ*(Sb–Cl) antibonding orbital is manifested in the slight elongation of the Sb–Cl bond in both complexes (2.461/2.515 Å, for 3/4) compared to the free ligand Ar2SbCl (2.442 Å). On the other hand, we focused on gaining evidence for the presence of the stabilizing interaction from the Sb lone pair toward the M center. Therefore, the lone pair of Sb was blocked by a BH3 molecule in complexes 3 and 4 to hamper the lp(Sb) → σ*(M–Cl) donation. As a result, the Sb–M bond distance lengthens significantly upon the BH3 addition from 2.88 to 3.17 Å for complex 3, and a smaller but still notable change from 2.81 to 3.04 Å is observed for complex 4. These Sb–M bond elongations are consistent with the lack of stabilizing interaction from the Sb center toward the metal in the borane adducts.
Finally, we aimed to identify the reasons why the Ar3Sb and Ar2SbCl ligands prefer different kinds of coordination modes. A simple and straightforward explanation seems to be the higher Lewis acidity of the Ar2SbCl ligand compared to that of the Ar3Sb counterpart. However, the backdonation in the Pd complex 3 offers stabilization similar to that of the Sb → Pd donation, and this does not justify the formation of a pentacoordinate Pd center. To clarify this situation, we studied the borane affinities (BA, Fig. 5D) of the ligands at both Sb and N centers defined as the complexation energy of the donor center (LB, Lewis base) utilizing BH3 as the Lewis acid (which cannot offer backdonation). The BA of the Sb center in the Ar3Sb ligand is much lower (−32.4 kcal mol−1) than that in Ar2SbCl (−14.4 kcal mol−1), highlighting that the Sb in the Ar2SbCl ligand is a weaker donor than that in Ar3Sb. Interestingly, the BA of the nitrogen in a guanidine arm (−24.4 kcal mol−1) has a value intermediate between the BAs of the two types of Sb centers. Comparing the BAs of the various centers in these ligands reveals that the best stabilization using the triaryl ligand Ar3Sb can be achieved via bidentate coordination utilizing the Sb and N donors.
In contrast to the Ar3Sb ligand, the Sb center of the Ar2SbCl ligand cannot compete with another N donor due to its highest (least negative) BA among the studied ones, and thus coordination is established by two imino-N centers rather than by an Sb and N donor. Altogether, the reason for the coordination fashion observed for complexes 3 and 4 is the lower donor strength of the Sb center compared to that of an imino-group rather than its higher Lewis acidity.
Footnote |
† Electronic supplementary information (ESI) available: All experimental data including NMR spectra and theoretical studies. CCDC 2313661 (Ar2SbCl); 2313662 (Ar2SbCl); 2313664 (Ar3Sb); 2313665 (1); 2313666 (2); 2313660 (3); 2313659 (4). For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4dt02787f |
This journal is © The Royal Society of Chemistry 2024 |