The electrification of ice and supercooled water drops

Abstract

The electrical structure of a thunderstorm consists essentially of a positive dipole, with values of the segregated charge estimated at 20 coulombs. Between the main charge centres there exists mixed charge, of both signs, of the order of 1000 to 1500 coulombs. As this charge is generated in regions where the temperature is well below 0 C the phenomena have been associated with the charging of ice. In recent years several workers have found that if a temperature gradient is set up in ice, the colder part acquires a positive charge and the warmer a negative charge. This has been explained by assuming that protons migrate from the warmer to the colder part of the ice, and Mason and Latham (1961 a) have put forward a quantitative theory, which is in very good agreement with their experimental results. In the present work the charging due to the momentary contact of vapour grown ice crystals at different temperatures was investigated. Ice crystals were grown in a diffusion chamber on a fine insulating fibre and a fine horizontal wire supported by a cam driven rod. Contact could be made between ice crystals grown at different levels in the diffusion chamber and the fibre then raised into a bronze Faraday cylinder, which was connected to a Vibron electrometer adapted for measuring single charges. Contact times of the order of 1/100th second were employed, this being the time for maximum charge separation predicted by Mason and Latham’s theory. However, although temperature differences of up to 10 C were involved no charging was observed. When water droplets are nucleated just below 0 C a strong ice shell is formed and further freezing produces a pressure inside the drop which finally causes the drop to burst. Mason and Maybank (1960) measured the charging which accompanies the bursting and attributed it to the migration of protons along the radial temperature gradient set up in the ice shell; the charge separation being caused by an excess of either the outer of inner surface of the shell carried away when the drop bursts. Mason and Latham (1961) suggest that this process may account for the charging of a hall pellet growing by the accretion of supercooled droplets. An attempt to measure the charging of a single supercooled droplet, brought into contact with an ice surface, was unsuccessful, but the charging of nucleated supercooled droplets was observed. One millimetre diameter droplets suspended from the insulating fibre were cooled to just below 0 C and then nucleated by dropping a small piece of solid carbon dioxide into the diffusion chamber. The droplets were raised into the Faraday cylinder after freezing was completed and no charging was observed unless fragmentation of the drop had taken place. The results obtained suggested that the charging was too large to he accounted for by the temperature gradient theory, and also that the sign of the charge remaining on the drop was determined by the amount of water present when the drop bursts.