Heat capacity and glass transition in P2O5–H2O solutions: support for Mishima's conjecture on solvent water at low temperature

Abstract

The P(2)O(5)-water system has the widest range of continuously glass-forming compositions known for any glassformer + water binary system. Despite the great range of structures explored by the glasses and liquids in this system, the glass transition temperature (T(g)) itself varies in a simple monotonic fashion. However the values of T(g) reported in the literature show wide disagreement, linked to the different methods of measurement employed. In this work we use differential scanning calorimetry (DSC) to obtain both T(g) itself and the jump in heat capacity that occurs as the metastable equilibrium of the supercooled liquid relieves the non-ergodic glassy state. Our study covers the molar ratio range of H(2)O/P(2)O(5) from 1.5 to 14 (corresponding to the mass fraction of P(2)O(5) between 0.36 and 0.84), which includes the compositions corresponding to pyrophosphoric acid (H(4)P(2)O(7)) and orthophosphoric acid (H(3)PO(4)). The theoretical model of Couchman and Karasz predicts very well the glass transition temperatures of the P(2)O(5)-H(2)O system over the whole composition range if the relatively large heat capacity change associated with water in aqueous solutions at the glass transition temperature is adopted, instead of the vanishingly small value observed for vapor deposited or hyperquenched pure water. Therefore, solvent water in this ambient pressure P(2)O(5)-H(2)O system behaves like a different liquid, more closely resembling a high-density liquid (HDL) polyamorph, as suggested by Mishima for electrolytes at high pressures.

Experimental evidence for excess entropy discontinuities in glass-forming solutions

“…(5) into 94 and more recently Corti et al 95 found a value of 35 J K −1 mol −1 for c l 1 − c g 1 from binary aqueous solutions containing hydrazine (N 2 H 4 ), hydrogen peroxide (H 2 O 2 ), and phosphorus pentoxide (P 2 O 5 ), respectively. These discrepancies are explained in terms of a "fragile-to-strong" liquid transition of very dilute solutions which is experimentally inaccessible due to homogeneous ice freezing.…”

“…These discrepancies are explained in terms of a "fragile-to-strong" liquid transition of very dilute solutions which is experimentally inaccessible due to homogeneous ice freezing. 82,95 This transition and hence the c l 1 − c g 1 which should be used in Eq. (6) depend on the solute.…”

An extended topological model for binary phosphate glasses

“…The T g of P 2 O 5 has been measured to be between 590 K 33 and 665 K 34 by (10) and experimental data. Data for x > 0.5 were determined by DSC 93 and for x = 0.5 by dilatometer. 32 The largest discrepancy between the experimentally determined and the modeled T g values is 4 K. The dotted line is for a topological model taking the DBO as a network forming atom with one constraint for the P-DBO bond and two additional α constraints for each Q n .…”

Glass Transition Temperature of Saccharide Aqueous Solutions Estimated with the Free Volume/Percolation Model

“…55,56 We suggest here that something similar occurs in polyol aqueous solutions, and hence the volumetric properties of water are different from that of bulk water (as represented by the linear temperature dependence shown in Figure 2). …”