In recent experimental investigations, it was found that secondary emission ratios as high as 10,000 to 1 could be attained utilizing field dependent secondary emission from magnesium oxide. Early tests showed the mechanism causing the high gains to be fundamentally different from the more standard secondary emission phenomenon.

The hypothesis was made that the mechanism of field dependent secondary emission was a process similar to that of the "Townsend avalanche" occurring in gas discharges. As the surface of the dielectric film was bombarded with primary electrons, the high resistivity of the material in combination with the secondary emission current caused the surface to charge to the potential of the collector grid, producing a high field within the dielectric. Electrons released within the material could then gain enough energy to liberate additional electrons, and an avalanche type process resulted.

Experiments were conducted to test this hypothesis and each proved to be consistent with the above theory. The main content of these experiments can be summarized in the following statements:

(1) The high yields appeared to be independent of the base material. This indicated that the surface or volume effects were most important, implying that a Fowler type field emission from the base metal was not a significant factor.

(2) In studying the secondary current as a function of field, the gas discharge equations were found to be correct. These equations predicted a straight line plot of the $\ln \ln \delta \mathrm{vs} \frac{1}{E}$, and in addition, gave a close estimate of the magnitude of the secondary emission ratio.

(3) By means of retarding potential measurements, the energies and mean free paths of the emitted secondary electrons were determined. These data were in good agreement with the results in item (2).

(4) The rise time for surface charging was determined by using square wave variations of bombarding currents, and was found to be consistent with the original hypothesis.

• Received 3 April 1952

DOI:https://doi.org/10.1103/PhysRev.88.492