Published online by Cambridge University Press: 24 October 2008
The fact that the rate of production of atomic hydrogen at a tungsten surface at a given temperature is proportional to the square root of the hydrogen pressure means either that the important process is the evaporation of atoms from an adsorbed film which over the whole range of experimental conditions is sparsely occupied, or that the production of atoms is in the main due to a process in which a hydrogen molecule strikes a bare tungsten atom in the surface, one atom being adsorbed and the other evaporating and the surface being almost completely covered over the whole range of experimental conditions. Either process leads to a temperature variation in the rate of atom production in agreement with experiment. A definite decision between the two processes cannot yet be made.
* Bryce, , Proc. Camb. Phil. Soc. 32 (1936), 648.CrossRefGoogle Scholar
† Roberts, , Proc. Camb. Phil. Soc. 32 (1936), 154Google Scholar. This paper will be referred to as Roberts I.
‡ See also Fowler, , Proc. Camb. Phil. Soc. 31 (1935), 262.CrossRefGoogle Scholar
§ In the present paper we shall only be applying equation (1) over a narrow range of conditions. In one application the surface is always very sparsely covered and in another it is almost completely covered, so that this restriction does not introduce any appreciable error. It may be noted however that the actual value of A will be different in the two applications.
* An attempt has been made to estimate the value of θ from equation (6) on p. 155 of Roberts I. A lower limit to the value of β, the condensation coefficient for molecules, was estimated at room temperature from the rapid formation of the adsorbed film (cf. Roberts I, p. 156); it was assumed that β does not change appreciably with temperature (see later in this paper). The value of η, the rate of evaporation of molecules from a complete film, was estimated at 700°K., the temperature at which evaporation becomes appreciable (cf. Roberts I, p. 156 and Proc. Roy. Soc. A, 152 (1935), 460Google Scholar—this paper will be referred to as Roberts II). The variation of η with temperature was assumed to be given by an exponential function of Φ/RT, where Φ is the heat of desorption of molecules, which has been measured in Roberts II. These calculations indicated that, even at the highest temperature and lowest pressure used, θ > 0·6 which suggests that process (iii) (b) is the important one. They will not be given here in detail because it is felt that the result cannot in itself be regarded as conclusive until a proposed direct experimental test has been carried out.
† Roberts I, equation (5).
* Roberts II, p. 459.
† Jones, Lennard and Devonshire, , Proc. Roy. Soc. A, 156 (1936), 26Google Scholar, have given an explicit formula for the condensation coefficient for van der Waals' fields which shows that the coefficient varies very slowly with temperature.
‡ Giauque, , J. Amer. Chem. Soc. 52 (1930), 4816.CrossRefGoogle Scholar
* Roberts I, equation (7).