PROBING OF BARRIER LAYERS WITH AN ELECTRON BEAM
barrier increases with the blocking potential. Also the application of a blocking potential leads to a very much increased gain (up to l0 times at low external resistance). At zero bias, the lowest value of Vm~ observed was l0 kV, and the highest 50kV, the gain at Vm~ being about 500. Values for the change of V~ with bias are from 10 to 20 kV for a change of bias from 0 to 1.5 volts or from 50 to 70 kV for 0 to 2 volts bias. Now it is known that these barrier layers are a few tenths of a micron thick and measurements of the capacity of the type of cells used confirm this. There is appreciable uncertainty concerning the range of electrons, but the depth for which the maximum dissipation of energy is at 50 kV can for selenium barely be less than 8 microns, and for 70 kV perhaps twice as much, the obvious explanation of the maximum being that it denotes the condition of highest energy loss in the barrier. A maximum of l0 kV on the other hand is not so much out of line with the results of capacity measurements.
Barring at present the possibility that the ideas of the thickness of the barrier layers require drastic revision, two explanations of the discrepancy offer themselves, viz. that the depth corresponding to the maximum represents a diffusion range rather than the thickness of the barrier or that the barrier commences only at an appreciable depth behind the surface. There are objections against either explanation.
We have also studied other aspects of the electron ~,oltaic effect, such as fatigue and recovery, the effect in the presence of visual light, and the influence of the external load but so far the results obtained have raised more questions than they have answered.