Influences of light illumination on (Cs,O) activated GaAs photocathode

Some experiments were made on the effects of light illumination on the (Cs,O) activated GaAs(110) photocathode in previous work. It was observed that if the GaAs photocathode activated without illumination were illuminated by light of proper intensities, its photosensitivity would increase to a new maximum in several minutes and that too intensive illumination could decrease the photosensitivity on the contrary. We also made a stability and recovery study of GaAs photocathodes under varies intensities illumination, which was sealed in image intensifier and placed in UHV chamber respectively, with the help of the self-developed spectral response testing instruments. We found that when the photocathode sealed in the image intensifier illuminated by light of weak intensities, the photosensitivity first increased to a new maximum and then decreased. After removing the light for several hours, the photosensitivity can be recovered itself. Contrast to the weak illumination case, the photosensitivity of photocathode under intensive illumination would decrease monotonously and could not be recovered. By another experiment on photocathode that had been activated successfully in the UHV chamber, we found some new phenomena. Monotonous degradation is produced not only by intensive illumination, but also by weak illumination. Furthermore, photosensitivity after illuminating was unrecoverable. One of the most important phenomena observed by us was that the intensity of pressure in UHV system varied with the intense of illumination. The more intense the illumination was, the more intense would be the pressure in vacuum chamber in the several minutes of beginning. In this paper, the phenomena above-mentioned are analyzed and discussed. It is suggested that excess minority carrier induced by incident photons would damage the surface composition and structure of (Cs,O) activated GaAs photocathode. According to our research, an effective solution to increase photoemission stability is put forward. In our discussion the dipole model is strongly supported.