A study of the structure of abraded metal surfaces

Abstract

Kramer and other workers have found that freshly abraded metal surfaces produced counts when placed beneath open Geiger-Müller tubes. It has also been shown that such surfaces give photo-electrons at wavelengths longer than the photo-electric threshold of the metal. The latter effect was investigated in the present work. The emission at different ranges of wavelength was studied for a number of freshly abraded metals. All the metals investigated give emission in the range 3000 to 3700 Å, while aluminium, magnesium and zinc gave emission also in the visible range. A strong emission peak was found at 4700 Å (2.64 eV) and a smaller peak at about 5200 Å. A broad band at 6000 to 7000 Å was also found to be present. The emission current increased as the square of the negative potential applied to the specimen, so that the ‘activity’ of a specimen, α , could be defined as counts min -1 V -2 . The emission current also increased with the intensity of illumination. The decay of the emission peak at 4700 Å from abraded aluminium was studied in detail. When the specimen was kept in an inert atmosphere (argon), and in darkness, the decay followed the equation for a first-order reaction with a rate constant k 1 = 1.5 x 10 -5 s -1 . Continuous illumination and flow of emission current increased the rate of decay. Unfiltered high-intensity illumination caused temporary fatigue of the emissive properties. When specimens were kept in atmospheres containing oxygen, decay was more rapid. The decay in the presence of oxygen was found to follow the equation α t = α 0 /(l + k 2 t ) 2 , where α t denotes the activity, α 0 the initial activity, t the time and k 2 a rate constant which increased proportionally with the concentration of oxygen in the gas. The distribution of the activity in depth was investigated, and it was found that when an outermost ‘dead’ layer had been removed by etching, a layer of high activity was encountered, which extended to 5 to 10 µ below the surface. The suggestion by other workers that the emissive properties derived from the metal and depended on exothermal processes was found untenable. Emission in the visible range, observed with metals which give ‘excess-metal’ oxides, is due to lattice imperfections in the oxide. The peak at 4700 Å is thought to be due to excited F' -centres (oxygen ion vacancies occupied by two electrons) which, when returning to the ground state, cause emission from shallow centres near the surface. The nature of the structures responsible for the other peaks and bands is yet uncertain.