Excess and Hump Current in Esaki Diodes

Abstract

The origin of the excess and hump current in Esaki diodes is described in terms of discrete defect energy levels in the forbidden band. The hump current results from equi-energy transitions of electrons which originate in a defect level in the n side and tunnel to the valence band on the p side. Impurity conduction is required for appreciable hump currents. The position of the hump locates the defect level if the Fermi level is known for the n side. The excess current is explained in terms of energy dissipating transitions in which electrons start from a defect location within the junction, tunnel to a virtual state in a localized defect in the p region, and drop to the valence band by impact recombination. A crude derivation by the WKB method relates the logarithm of the excess current to the bias voltage where the proportionality factor contains the square root of the effective mass of the electron. From current-voltage curves for diodes fabricated from germanium, silicon, and gallium arsenide, the effective electron masses are estimated as 0.01, 0.1, and 0.05 m, respectively. Defect levels in germanium are estimated at 0.06 and 0.24 ev below the conduction band; levels in silicon are located at 0.04 and 0.42 ev; and in gallium arsenide 0.12 and 0.5 ev below the conduction band. The simultaneous existence of two defect levels can give rise to two distinct slopes of the logI vs V curve which have been observed in silicon and gallium arsenide diodes following irradiation with 2 Mev electrons.