Direct Measurement of Mechanical Properties of Vitreous Ice by Cryo-FIB

Low-density amorphous ice, also known as “vitreous” ice, is the most abundant phase of water in the universe, although it is not naturally found on this planet. The most common scientific/commercial application for vitreous ice is the preparation of specimens for cryo-electron microscopy (cryo-EM) in their native state by plunge-freezing or high-pressure freezing. Simulation/modeling of vitreous ice has produced controversial and sometimes contradictory results (add two or more references), so an experimental approach is warranted. Understanding the physics of vitreous ice is significant not only to model extraplanetary phenomena such as cryovolcanism but also to improve the performance of biological applications such as cryo-EM. During data collection by cryo-EM, the beam-induced motion significantly degrades image quality [1]. It has been theorized that this motion is largely caused by the relief of mechanical stress generated during the plunge freezing process [2]. As a first step towards understanding the mechanics of amorphous ice, including the mechanical behavior on cryo-EM grids, we developed a new approach to perform nanoindentation tests on vitreous and crystalline ice specimens. We conducted nanoindentation tests on water forms to determine the Young’s modulus of vitreous (amorphous) and crystalline ice. After plunge-freezing water sheets on TEM grids (Fig. 1A), we used cryogenic focused ion beam micromachining (cryo-FIB) to fabricate ice cantilevers in situ (Fig. 1). Then, we employed the tungsten needle of the Aquilos 2 cryo-FIB instrument’s native nanomanipulator as an indenter. Each needle is calibrated by SEM observation of the the strain curve when indenting against a pre-calibrated AFM cantilever; the Young’s modulus of the cantilever and the strains of the cantilever and needle are used to infer the full stress/strain curve of the needle. To produce crystalline ice, the same plunge-frozen water sheets can be warmed inside the SEM chamber to a defined annealing temperature such as 155K by varying the flow rate of the cryogen. We found that vitreous ice is significantly less stiff than crystalline ice (Fig. 2). Moreover, whereas crystalline ice exhibits a temperature-dependent shift in modulus (with values increasing as the temperature is lowered), vitreous ice does not exhibit a significant dependence of elastic modulus on temperature between 78K and 137K. This difference in Young's modulus between the two forms of ice is likely due to the distinct structural arrangements of the molecules in each phase. Crystalline ice possesses a well-defined, ordered lattice structure that imparts greater stiffness to the material. In contrast, vitreous ice lacks long-range order and exhibits a more disordered, amorphous structure, leading to lower stiffness. These findings illustrate that the mechanical properties of ice are strongly influenced by its phase and (for some phases) by its temperature, with the Young's modulus providing a useful metric for characterizing such behavior. Further studies building on this nanoindentation technique aim to relate the mechanical properties of amorphous ice to vitrification-induced stress and beam-induced motion. a) Diagrammatic illustration of the in situ fabrication of an ice cantilever by plunge-freezing a water sheet and ablating material to form a plank by cryo-FIB. b) Scanning electron micrograph of a representative beam-type cantilever micromachined out of a water sheet. Temperature-Dependent Young's Modulus of Crystalline and Vitreous Ice. Young’s modulus was computed at two temperatures for both crystalline (blue) and vitreous (green) phases of ice.