It is now well established that a large class of important chemical reactions is controlled by the growth of obstructive films, but our knowledge of the mechanism of the growth of such films—especially at low temperatures—is still very imperfect. The present paper describes a detailed study of a particularly instructive case of film-growth, the action of iodine on silver; this reaction was chosen because silver iodide films—in contrast with oxide films—reach visible thickness rapidly at ordinary temperatures. The optical properties of these films have already been studied by Wernick, whilst much information regarding the velocity of film-growth has been obtained by Tammanm by Kohlschutter and Krahenbiihl, and by Hartung. Tammann used the interference colour as the means of arriving at the thickness of the film; this method has certain unique advantages, but the more recent work of Tammann and Bockow has indicated that—in the case of oxide-films at least—the particular form of the method employed gives values for the thickness differing widely from those obtained by the gravimetric method. Tammann expressed the relation between the thickness (
y
) and the time (
t
) by the parabolic equation
y
2
= 2
pt
, where
p
is a. constant. Kohlschütter and Krähenbiihl, and also Hartung, used microgravimetric methods to determine the amount of iodine taken up, and obtained curves connecting thickness and time ; these curves do not appear to follow the parabolic equation. All the experimenters mentioned used iodine vapour to attack the silver, but the concentration of iodine in the gas phase was not directly determined. In all cases the surfaces were prepared in air, and the possible effect of exposure to oxygen was not considered; yet in several of the reactions of copper and iron, pre-exposure to air or oxygen is known greatly to modify the result, owing to the fact that an oxide-film may become protective before reaching the thickness needed for interference colours. It may be mentioned that after exposure to vapour, the colour produced on metallic specimens is usually not uniform, but indicates a greater film-thickness near the edge. In the present research, it was decided to use a solution of iodine in an organic solvent; in this case, the concentration could be fixed or varied at will, and in general uniform colouration (indicating uniform thickness) could be obtained. Moreover, abrasion could be conducted, if required, below the surface of the solvent, instead of in air. Chloroform was .found to be a suitable liquid, since, whilst freely dissolving iodine, it had no appreciable solvent action on a silver iodide film.
Growth and Failure of Oxide Irregularities During Thermal Cycling: Interactions Between Stress, Geometry and Oxide Formation
