Working on the hypothesis that atmospheric ice-forming nuclei are largely of terrestrial origin, the nucleating ability of various types of soil particles and mineral dusts has been investigated. Of the thirty substances tested, twenty-one, mainly silicate minerals of the clay and mica groups, were found to produce ice crystals in supercooled clouds and also on supercooled soap films at temperatures of − 18° C, or above, and of these, ten were active above − 12° C. The most abundant of these is kaolinite with a threshold temperature of − 9° C. Ten natural substances, again mainly silicates, were found to become more efficient ice nuclei having once been involved in ice-crystal formation, i.e. they could be pre-activated or «trained». Thus, ice crystals grown on kaolinite nuclei, which are initially active at −9° C, when evaporated and warmed to near 0° C in a dry atmosphere, leave behind nuclei which are thereafter effective at − 4° C. Particles of montmorillonite, another important constituent of some clays, and which are initially inactive even at −25° C, may be pre-activated to serve as ice nuclei at temperatures as high as −10° C. It is suggested that although such particles can initially form ice crystals only at cirrus levels, when the ice crystals evaporate they will leave behind some «trained» nuclei which may later seed lower clouds at temperatures only a few degrees below 0° C. On this hypothesis, the fact that efficient nuclei are occasionally more abundant at higher levels would not necessarily imply that they originate from outer space. Indeed, in view of our tests on products of stony meteorites, produced both by grinding and vaporization, which show them to be ineffective at temperatures above − 17° C, it seems likely that atmospheric ice nuclei are produced mainly at the earth's surface, the clay minerals, particularly kaolinite, being a major source.
Although a good deal of work has been carried out in different laboratories on the ice-nucleating ability of a wide variety of inorganic compounds, there has been little agreement in the results. Careful tests carried out in our laboratory have revealed a number of reasons for this. Spurious results may be obtained because of the presence, in the air or the chemicals, of small traces of silver or free iodine, leading to the formation of silver iodide: if all such trace impurities are removed, many of the substances that have been claimed to provide efficient ice nuclei are found to be quite ineffective. It is dangerous to infer that all twinkling particles in a water cloud are ice crystals since particles of some seeding agents glitter even at positive temperatures. The threshold temperature of a nucleant will depend upon the criterion adopted for the onset of nucleation, i.e. upon the fraction of the total number of particles of seeding agent which are activated; this, in turn, will depend upon the fraction of particles which happen to possess suitable crystallographic faces for nucleation. Much may also depend upon the manner in which the test is performed. Since some nucleating materials produce ice crystals only after a delay of 30 seconds or more, they may appear to be ineffective if tested in the transient cloud of an expansion chamber but highly effective if allowed to remain in an ice-supersaturated atmosphere for a minute or more. Again, we have found that the efficiency of some nuclei is governed by the supersaturation as well as the temperature of the environment, and the supersaturation regimes in expansion, diffusion, and mixing-cloud chamber may be widely different. Highly soluble particles, although able to act as «sublimation» nuclei in atmospheres super-saturated relative to ice but sub-saturated relative to water, on entering a water cloud go quickly into solution and lose their nucleating ability.
Inorganic substances which definitely nucleate a supercooled water cloud in a mixing-cloud chamber at temperatures of −15° C and above are: AgI (−4° C), PbI2 (−6° C), CuS (−6° C), Ag2S (−8° C), Ag2O (−9° C), HgI2 (−8° C), V2O5 (−14° C), Cu2I2 (−15° C), the figures in brackets indicating the threshold temperatures at which about one particle in 104 becomes active as an ice nucleus. Cadmium iodide (−12° C), ammonium fluoride (−9° C) and iodine (−14° C) are examples of salts which will act as sublimation nuclei in an ice-supersaturated atmosphere and will nucleate a supercooled soap film, but which are ineffective in a water cloud because of their solubility.
Although the most efficient nucleating agents tend to be hexagonal in structure, there are some striking exceptions e.g. Ag2S, Ag2O, HgI2, but in most cases, we have been able to find a low-index crystal surface on which the ice lattice could grow with a misfit of only a few per cent.
In an attempt to investigate the nucleation mechanism in more detail, we have studied the growth of ice on single crystals of various nucleating agents. Perfect orientation of ice crystals has so far been observed on the basal faces of silver iodide, lead iodide, cupric sulphide, cadmium iodide, and freshly-cleaved mica, on the (001) plane of iodine, and on the (010) plane of mercuric iodide.