Kinetic Properties of Transformations Between Different Amorphous Ice Structures

Summary

We report on extensive kinetic experiments on different amorphous ice modifications of high density following their transformation at constant pressure and temperature into the low-density modification LDA. Monitoring the structural changes in situ by wide angle diffraction and small angle scattering experiments we establish the universal behavior of this transformation irrespective of the initial sample structure. The universality can be found in the formfactor changes as there is a strong peak in wide angle diffraction data shifting from high (2.2–2.4 Å^− 1) to low Q numbers (1.7 Å^− 1) whose width becomes transiently broadened reaching a maximum width close to the center of transformation. This broadening is generated by a pronounced heterogeneity of the sample, which can be directly monitored as a transient additional signal in the Q-range 0.04–0.6 Å^− 1. Local density variations are the origin of the additional signal and the samples can be understood as structural mixtures in the course of the transformation. Only very high-density amorphous structures and LDA prove to be homogeneous samples. The high–density amorphous modification HDA, a structure considered as a reference in literature, is unequivocally heterogeneous and its grade of heterogeneity can be reproduced with high precision. At lowest momentum numbers Q < 0. 04 Å^− 1 the formfactor follows the power–law Q ^− 4 of Porod-limit scattering giving evidence of a grainy consistency of the samples with an average grain size of the order of 10 μm. Furthermore, the universality of the transformation can be found in the kinetic response of the samples. When extracting kinetic information from diffraction data two apparent stages, a sluggish conversion crossing over into a sharp sigmoid-shaped step, are intrinsic to the entire transformation process. This stages can be found reproduced in the time response of the intensity of Porod-limit scattering in the small angle data. However, the transient signal at intermediate Q marks the transformation as a single continuing process, which could be understood in the most simple case as a nucleation and growth of a structure A in a matrix B. We show, however, that the driving forces, i.e. activation energies, of transformations of different structures are different, indicating that despite a close resemblance of formfactors the samples occupy different minima on a potential energy landscape. The bandwidth of activation energies estimated in our experiments is 33–65 kJ mol^− 1.