Endothermic features on heating of glasses show that the second glass to liquid transition of water was phenomenologically-mistaken

When glasses formed by cooling at different rates are heated at the same rate, they show different shapes of the glass to liquid transition endotherm in their C-p-T plot and different onset temperatures of the C-p-endotherm. (This temperature is taken as the glass to liquid transition temperature. To prevent confusion we denote it as T-g -> l.) According to the phenomenology of glass to liquid transition, (i) a glass formed by slower cooling shows on heating a higher T-g -> l than a glass formed by rapid cooling and (ii) the C-p-endotherm's shape changes and its enthalpy recovery overshoot becomes higher. But a recent study arbitrarily asserted the opposite, namely, that a glass formed by slower cooling of a liquid has a lower T-g -> l and this feature is "a typical signature of a glass transition". By using that "signature" the study concluded that an ill-defined state of solid water has a "second T-g" (T-g -> l in the terminology here) at 116K. The assertion caused us to perform a calorimetric study of the glass to liquid transition phenomenon in eight materials by using the same cooling/heating protocols that had led to the assertion of the signature. Our study confirms the already known glass phenomenology. Therefore, "a typical signature for a glass transition" is false, and the second T-g (or T-g -> l ) at 116 K based on the study of an ill-defined solid water (formed by annealing the pressure-collapsed state of ice) was phenomenologically-mistaken. In a different context, we quantify the dependence of T-g -> l on the cooling rate and compare the effect of slow cooling against the effects of annealing (ageing) on the properties of the glassy state. The study has adverse consequences for some models of the structure of water. (C) 2016 Elsevier B.V. All rights reserved.