Stabilization of coordinatively unsaturated Ir₄clusters with bulky ligands: a comparative study of chemical and mechanical effects

Abstract

The synthesis and characterization of new cluster compounds represented by the series Ir(4)(CO)(12-x)L(x) (L = tert-butyl-calix[4]-arene(OPr)(3)(OCH(2)PPh(2)); x = 2 and 3) is reported using ESI mass spectrometry, NMR spectroscopy, IR spectroscopy and single-crystal X-ray diffraction. Thermally driven decarbonylation of the cluster compound series represented by x = 1-3 according to the formula above is followed via FTIR and NMR spectroscopies, and dynamic light scattering in toluene solution. The propensity of these clusters to decarbonylate in solution is shown to be directly correlated with number density of adsorbed calixarene phosphine ligands and controlled via Pauli repulsion between metal d and CO 5σ orbitals. The tendency for cluster aggregation unintuitively follows a trend that is exactly opposite to the cluster's propensity to decarbonylate. No cluster aggregation is observed for clusters consisting of x = 3, even after extensive decarbonylation via loss of all bridging CO ligands and coordinative unsaturation. Some of the CO lost during thermal treatment via decarbonylation can be rebound to the coordinatively unsaturated cluster consisting of x = 3. In contrast, the clusters consisting of x = 1 and x = 2 both aggregate into large nanoparticles when treated under identical conditions. Clusters in which the calixarene phosphine ligand is replaced with a sterically less demanding PPh(2)Me ligand 6 lead to significantly less coordinative unsaturation upon thermal treatment. Altogether, these data support a mechanical model of accessibility in coordinatively unsaturated metal clusters in solution, which hinges on having at least three sterically bulky organic ligands per Ir(4) core.

Structures, relative energies, and ligand dissociation energies of iridium carbonyl phosphine clusters

“…Thus, we suggest, if a transition state for CO loss is present, that such a transition state would have the lowest energy for a bri CO. This inference is consistent with the experimental result that bridging carbonyls dissociate more readily in thermal decarbonylation [42]. Because bri CO ligands and eq, ax CO ligands can be readily interconverted [40], substitution with a new ligand can take place at any basal position determined by its bulk and orbital direction; api substitutions are rare except through some selectively reactive decarbonylations that lower the activation energy, as they are not readily interconverted [43].…”

“…Thus, one should expect lower-energy transition states for the dissociation of bridging carbonyls. Because CO ligands can switch between the basal and bridging sites [40], basal substitution can occur more readily than apical substitution [42]. Similar to the trend found for Ir(PH 3 ) y (CO) z and Ir 2 (PH 3 ) y (CO) z , the LDEs of the PH 3 's in the tetra-iridium cluster gradually decrease with an increasing number of PH 3 's.…”

“…Such a transition state influence has been assumed for mononuclear complexes [98] and was used as a basis for inferring ligand dissociation rates for bi-and higher-nuclearity iridium clusters [93]. For Ir x (PH 3 ) y (CO) z clusters, this reasoning would suggest that decarbonylation will take place at an increasing rate as more phosphines are substituted, in agreement with what has been reported [42].…”

“…When hydrogen atom transfer and the rearrangement of the PH 3 from a bri to an eq position does not occur, the lowest CO LDE for a given site follows the order eq > ax > bri > api. Up to x = 3, the lowest energy CO LDEs are approximately the same as PH 3 ligands are substituted for CO. Experimentally, an increase in the numbers of phosphines leads to an increase in the decarbonylation rates [42], most likely a consequence of the use of much bulkier phosphines and potentially kinetic control of the decarbonylation. The addition of the phosphines makes the Ir atoms more electron rich, leading, it has been hypothesized, to a lower-energy transition state for thermal decarbonylation [93].…”

“…When L is a phosphine or phosphite, its bulk also strongly influences the degree of ligand substitution in the parent cluster Ir 4 (CO) 12 -that is, determining whether Ir 4 (CO) 8 L 4 or Ir 4 (CO) 9 L 3 is the final product [41] -but synthesis of Ir 4 (CO) 10 L 2 or Ir 4 (CO) 11 L directly from the parent through direct substitution is difficult [41]. Okrut et al [42] demonstrated that as the number of strong r-donor phosphine ligands increases, the thermomechanical stability against aggregation increases, whereas the thermochemical stability against decarbonylation varies inversely.…”