Ice electrification

Abstract

When water drops are cooled to -1 C, nucleated with tiny ice crystals and then allowed to freeze at -15 C, a thin shell of ice is formed. Further freezing of the water causes internal stresses to be set up and the water escapes through the shell to form spicules. If the pressure becomes too great then some of the ice shell or of the spicule may be ejected, and the residue is found to have a charge of the order of 10(^-3) e.s.u. A diffusion chamber was designed and constructed. The base was cooled directly being an aluminium heat sink surrounded by solid carbon dioxide. The steady temperature gradient thus set up was measured with horizontal thermo-couples. A 1mm diameter water drop was suspended in the centre of the chamber on the end of a fine insulating fibre which could be easily raised and lowered. The charge on the drop was measured by raising it into the centre of a Faraday cage which was connected to a sensitive ballistic galvanometer through a Vibrating Reed Electrometer (V.R.E.). The limit of sensitivity was 0.05 x 10(^-3) e.s.u. It was found that the sign of the charge on the residue was dependant upon the type of break which had occurred. When more water than ice was given off then the residue had a predominantly negative charge. When no break occurred, no charge was detected. Of the 633 drops studied, 118 broke in some way. The average charge on the residue was +0.34 x 10(^-3) e.s.u. and ranged from -17.10 to +25.30 x 10(^-3) e.s.u. It was found impossible to account for either the magnitude or the sign of the charge in terms of the Latham and Mason temperature gradient theory but both could be accounted for if the Workman and Reynolds effect was invoked. It is suggested that in a thundercloud the effect of the turbulence present is to increase the number of breaks which occur while the spicule is still partially liquid, thereby increasing the number of negative residues and giving the cloud the correct polarity, when minute water drops freeze on graupel and break.