Y(BH₄)₃—an old–new ternary hydrogen store aka learning from a multitude of failures

Fourteen different synthetic approaches towards pure solvent-free Y(BH₄)₃ have been tested, thirteen of which have failed. Attempted reactions of YCl₃ or Y(OC₄H₉)₃ with LiBH₄ in THF, those of YCl₃ with (C₄H₉)₄N⁺ BH₄⁻, as well as between YH_x3 and R₄NBH₃ (R = CH₃, C₂H₅) in the presence or absence of a solvent (n-hexane or CH₂Cl₂) did not lead to the expected product. The mechanochemical solid/solid reactions (MBH₄ + 3 YX₃→ Y(BH₄)₃ + 3 LiCl, where M = Li, Na; X = F, Cl) have succeeded only for the LiBH₄ and YCl₃ reagents, but the separation of the crystalline reaction products (Y(BH₄)₃ in its Pa phase and LiCl) by dissolution or flotation in various solvents has not been successful. The thermal decomposition process of Y(BH₄)₃ in a mixture with LiCl has been investigated with thermogravimetric (TGA) and calorimetric analysis (DSC) combined with spectroscopic evolved gas analysis (EGA). Three major endothermic steps could be distinguished in the DSC profile at ca. 232, 282, 475 °C (heating rate 10 K min⁻¹) corresponding to a phase transition and two steps of thermal decomposition. Solid decomposition products are amorphous except for the new cubic polymorph of Y(BH₄)₃ overlooked in previous work. The high-temperature phase forms at the onset of thermal decomposition and it may be prepared by heating of the low-temperature phase up to a narrow temperature range (194–210 °C) followed by rapid quenching. Y(BH₄)₃ constitutes a novel highly efficient hydrogen storage material (theor. 9.0 wt% H) but, unfortunately, the evolved H₂ is contaminated by toxic boron hydrides and products of their pyrolysis.