Glass transition and positional ordering of hydrogen in bulk and nanocrystalline palladium

Abstract

Palladium hydride ${\mathrm{PdH}}_{x}\phantom{\rule{0.28em}{0ex}}(0<x<1)$, which may be the most well known metal hydride, still has some unexplained phenomena. In order to elucidate the origin of the ``50 K anomaly,'' we have measured the heat capacity of the $\ensuremath{\beta}$ phase of ${\mathrm{PdH}}_{x}\phantom{\rule{0.28em}{0ex}}(x\phantom{\rule{0.28em}{0ex}}=\phantom{\rule{0.28em}{0ex}}0.638,\phantom{\rule{0.28em}{0ex}}0.725,\phantom{\rule{0.28em}{0ex}}0.782,\phantom{\rule{0.28em}{0ex}}0.829)$ and ${\mathrm{PdD}}_{0.641}$ by means of a custom-built adiabatic calorimeter. A steplike anomaly in heat capacity appeared at a temperature depending on $x$ between 50 and 75 K. An exothermic effect followed by an endothermic one, implying an enthalpy relaxation to the equilibrium state, took place around the anomaly. This anomaly is attributed to not a phase transition as has been believed so far, but a glass transition corresponding to the freezing of the jump motion of hydrogen atoms in a short-range ordered state. If the hydrogen motion is not frozen, the completely ordered state could be realized at a hypothetical transition which appears below the glass transition temperature. Interestingly, the glass transition disappeared in ${\mathrm{PdH}}_{x}\phantom{\rule{0.28em}{0ex}}(x\phantom{\rule{0.28em}{0ex}}=\phantom{\rule{0.28em}{0ex}}0.553,\phantom{\rule{0.28em}{0ex}}0.629)$ and ${\mathrm{PdD}}_{0.372}$ particles with a diameter of ca. 8 nm. This could be due to a quantum tunneling effect which is caused by surface and/or distortion effect of nanoparticles.