Confronting Ice Vapor Growth Models with Measurements at Temperatures Below -25C

A newly designed thermal gradient diffusion chamber was used to grow columnar ice crystals near the tip of a fiberglass substrate at constant temperatures and supersaturations.. The substrate enters through the side of the chamber walls and terminates in the middle of the chamber. This design mitigates the vapor competition from surrounding crystals that may have influenced prior studies. Experiments were conducted at temperatures below -25C because few growth time-series measurements have been taken at these temperatures, and because the primary growth mode of single crystals at temperatures below -40C is not known. Crystals grown at temperatures below -45C were primarily columnar with deep hollowing. Crystal aspect ratios tended to increase with decreasing temperature, which is indicative of increasing kinetic limitations on the prism faces. All columnar crystals were substantially hollowed on the actively growing face (basal) face, with little growth and no hollowing on the strongly kinetics limited (prism) face. Most crystals approached a steady-state aspect ratio between 4 and 8 over about 1 hour of growth. Growth that approaches a constant aspect ratio is indicative of crystals grown by surface steps that nucleate near crystal corners and edges, providing evidence for the primary growth mechanism for single column crystals at low temperatures. Neither standard capacitance theory nor the theory of adaptive habit growth (Chen and Lamb, 1994) can reproduce the measured growth. Capacitance theory does not allow the aspect ratio to change, and if the theory is supplied with the observed aspect ratio the dimensional and mass growth rate becomes far too high. Chen and Lamb's (1994) theory produces continual aspect ratio increases causing the crystal dimensions to become too large. In contrast, the theory of Frank (1982) and Nelson and Baker (1996), that assumes steps nucleate near facet edges and corners, is capable of reproducing the measurements.