Crystal Structures, Properties and Reactivity of Selected Macrocyclic and Chelate Complexes of Ni(II)

Abstract

In this dissertation we describe the structure, properties and decomposition reactions of a series of Ni(II) coordination complexes formed from reaction of the appropriate macrocyclic or chelating ligand with a simple nickel salt. The ligands used were 12aneS4 (1,4,7,10-tetrathiacyclododecane), 14aneS4 (1,4,8,11-tetrathiacyclotetradecane), cyclam (1,4,8,11-tetraazacyclotetradecane), dppe (1,2- (diphenylphosphino)ethane), and PP3 (tris-(2-(diphenylphosphino)ethyl)phosphine). The work falls into three broad sections related to, respectively: catalysis of the reversible decomposition of complex hydrides for hydrogen storage; the unusual structure of [Ni(H2O)6][Ni(cyclam)(SO4)2]; and the facile reversible hydration of Ni(12aneS4)(BF4)2. The ability of transition metal salts, particularly those of titanium, to catalyse the dehydrogenation/rehydrogenation of complex hydride hydrogen stores such as NaAlH4 is well known. Unfortunately the activity of these simple salts is not sufficient for commercial use in light vehicles, and further improvements to the catalysts are hampered both by the limited scope for adjusting the catalyst, and the lack of detailed knowledge of the active species. Instead of using such simple salts, we have investigated complexes of Ni(II) and 12aneS4, 14aneS4, cyclam, dppe and PP3. This approach was taken to serve two objectives: the first that should catalytic activity be found, we should have a good idea of the active species and thus be better able to improve upon it; and second, that such complexes often form molecular crystals which should be significantly easier to disperse in the hydrogen store by high-energy milling than the ionic crystals of the simple salts. The stability of the complexes was tested with a series of lithium and sodium borohydride and alanate compounds with progressively more aggressive reducing properties. The tetrathioether complexes were not resistant to reduction, even by dilute NaBH4 solution, forming black/brown tars or solids with poorly defined infrared spectra and showing no peaks in their XRD patterns. The cyclam complexes (Ni(cyclam)SO4, Ni(cyclam)(ClO4)2 or Ni(cyclam)(BF4)2), however, could be reacted with either NaBH4 or LiBH4 to form a nickel borohydride complex, Ni(cyclam)(BH4)2, of unusual stability. Both cis and trans isomers of this distorted octahedral complex were obtained, and their crystal structures investigated in some detail, including by powder neutron diffraction. The potential of Ni(cyclam)(BH4)2 for catalysis in complex hydride hydrogen stores was assessed by simultaneous thermogravimetric analysis and differential scanning calorimetry (TGA/DSC) and the gases evolved analysed in real-time by infrared spectroscopy and/or mass spectrometry. A difference in the decomposition profile of the two isomers (with the cis configuration decomposing at lower temperature than the trans) demonstrated the importance of the geometry of such complexes when iii considering them for such uses. However, the complexes did not show any catalytic behaviour toward the decomposition of either lithium or sodium borohydride when milled with them in a high- energy disc mill. Furthermore, both isomers of Ni(cyclam)(BH4)2 decomposed exothermically, ruling them out for use as catalysts in onboard reversible stores, for which an endothermic decomposition is required. The bidentate phosphine chelate dppe showed behaviour similar to that of the thioether macrocycles, i.e., it did not prevent reduction of Ni(II) even by dilute solutions of NaBH4. In contrast, the tetradentate PP3 ligand was able to prevent reduction even by the highly aggressive reducing agent LiAlH4, forming a penta-coordinated Ni(II) hydride complex. The formation of this hydride complex could also be performed in the solid state by high energy disc-milling Ni(PP3)(BF4)2 with one of the complex hydrides studied. Analysis of the decomposition of the PP3 complex by TGA/DSC, showed that as with Ni(cyclam)(BH4)2, they do not provide any catalytic behaviour when milled with complex hydrides, and that both the precursor (Ni(PP3)(BF4)2) and hydride complexes decompose exothermically, making them unsuitable as catalysts in onboard reversible stores. From this series of experiments, we conclude that the stability of the Ni(II) complexes with respect to reduction by complex hydrides varies according to the donor type in the following series O, S < N < P. Though catalysis of the decomposition (hydrogen evolving) reaction of borohydrides and alanates was not observed with any of the complexes, further work may yet allow for the development of such catalysts. In particular, the use of mixed-donor chelates or macrocycles and/or bridged bi-metallic complexes, with each metal centre donating or accepting a single electron in the two-electron oxidation/reduction associated with H2 + 2e− ⇌ 2H−, perhaps being necessary. In our view, complexes involving P donors show the most promise for future research due to the stability of such complexes even in highly reducing environments. However, other restrictions inherent to the technology may yet prevent the use of complex hydrides as hydrogen storage materials for use in cars, particularly the significant problems associated with heat management upon onboard re- fueling, and these are also discussed. The unusual structure of [Ni(H2O)6][Ni(cyclam)(SO4)2], synthesised as a by-product of the standard Ni(cyclam)SO4 complex, is peculiar in having all the stronger σ-donors (cyclam and SO42−) attached to the same nickel centre. Only two other complexes in the Cambridge Structural Database bear a superficial similarity to this complex, and upon closer inspection even these are actually quite different. A substantial network of strong hydrogen bonds links the cationic and anionic complexes of [Ni(H2O)6][Ni(cyclam)(SO4)2] and is likely the source of the stability of this highly novel structure. iv Ni(12aneS4)(BF4)2 was found to absorb and desorb water very easily, with hydration of a powdered sample occurring in minutes (and surface effects clearly visible in seconds) by simple exposure to atmospheric air, and dehydration within seconds by heating to about 100 °C, in minutes by reducing the pressure to a few mbar, or over several hours in a dry atmosphere at room temperature and pressure. The structures of the complexes were solved from powder synchrotron X-ray diffraction patterns and found to have a clear topotactic relationship with highly anisotropic expansion of lattice parameters. Upon hydration, the flexibility of the 12aneS4 ring allows it to fold back and make room for the two water molecules that attach directly to the nickel to form [Ni(12aneS4)(H2O)2](BF4)2. The thermodynamic and kinetic aspects of the solid state reaction were investigated by TGA/DSC. With a very slow temperature ramp (as low as 0.02 K min−1) and no purging gas, this technique showed two separate processes occurring for both hydration and dehydration, attributed to the addition or loss of one equivalent of water at each step. The enthalpy, entropy and activation energy for each step in static air and for the overall reaction with purging gas were estimated using both the Kissinger and Ozawa-Flynn-Wall approaches. The calculated activation energy was found to be highly dependent on whether the sample chamber was purged during the experiment with dry Ar gas or not, which in removing water from the crystallite surfaces, points to the important role the reversibility of the reaction plays in the overall kinetics. The anhydrous complex, expected to have a low-spin d8 electronic configuration, shows ‘anomalous’ magnetic susceptibility, most likely due to its slight deviation away from square planar geometry and therefore imperfect quenching of the orbital angular momentum. The octahedral hydrated complex, however, shows magnetic behaviour typical of a high-spin d8 configuration.