A nanostructural investigation of mechanochemically synthesised hydrogen storage materials

Abstract

Aluminium, aluminium hydride (alane), and magnesium hydride nanoparticles have been mechanochemically synthesised in order to study their hydrogen sorption properties in contrast to the bulk. Nanoparticle formation was facilitated by the addition of a salt phase to ball milled chemical reagents that matched the reaction byproduct phase. The presence of a salt buffer during ball milling prevents agglomeration and thus restricts particle growth.Aluminium nanoparticles were mechanochemically synthesised with particle sizes from 40 – 55 nm. The LiCl salt by-product phase was removed by washing with a nitromethane/AlCl[subscript]3 solution resulting in 55 nm Al particles (single crystals) that did not display any crystalline oxide phases. High pressure hydrogen absorption experiments were undertaken up to 2 kbar at temperatures from 77 – 473 K to examine if there were any major thermodynamic changes to the Al. No hydrogen absorption could be detected proving that either smaller Al is required to form AlH3 under these conditions or higher pressures are needed. Ni-coated and Ti-doped Al nanoparticles were also synthesised in order to verify if catalytic metals could enhance hydriding kinetics and allow hydrogenation to occur at lower pressures. However the doped samples did not display any hydrogen absorption up to 108 bar.Alane nanoparticles were synthesised using both room temperature and cryogenic mechanochemical synthesis with particle sizes < 100 nm. The evolution of alane production was investigated as a function of milling time under a variety of milling conditions. Cryogenic milling was verified to form higher yields of AlH[subscript]3 than room temperature milling and four different alane phases (α, α', β, γ) were identified by XRD structural investigations. The LiCl reaction by-product phase was removed by washing with a nitromethane/AlCl[subscript]3 solution, which adversely reacted with the AlH[subscript]3 nanoparticles. The hydrogen desorption kinetics in washed samples were hindered, and the maximum H[subscript]2 wt.% was halved although no crystalline oxide or hydroxide phases were found using XRD. Unwashed mechanochemically synthesised AlH[subscript]3 was found to desorb at room temperature over months and significantly at 50ºC in a 24 hr period. Quantitative Rietveld results coupled with hydrogen desorption measurements suggested the presence of an amorphous AlH[subscript]3 phase in the mechanochemically synthesised samples.The mechanochemical synthesis of MgH[subscript]2 was undertaken with varying LiCl buffer quantities. Increasing the buffer resulted in MgH[subscript]2 crystallite sizes down to 6.7 nm, measured by XRD, whilst TEM investigations showed that increasing the buffer resulted in smaller, more highly dispersed MgH[subscript]2 nanoparticles. The size of these MgH[subscript]2 particles approached theoretical predictions for thermodynamic changes, where the MgH[subscript]2 is only physically bound by the LiCl. Hydrogen equilibrium pressure measurements were used to determine the decomposition enthalpy and entropy for MgH[subscript]2 nanoparticles that were mechanochemically synthesised. A reduction in both the decomposition enthalpy (ΔH decrease of 2.84 kJ/mol H[subscript]2) and entropy (ΔS decrease of 3.8 J/mol H[subscript]2/K) was found for ~7 nm MgH[subscript]2 nanoparticles in relation to bulk MgH[subscript]2. The consequence of this thermodynamic destabilization is a drop in the 1 bar hydrogen equilibrium pressure of ~6°C. The temperature drop is not as large as theoretical predictions due to the decrease in reaction entropy which partially counteracts the effect from the decrease in reaction enthalpy.