Validity of the Padovani–Stratton formulas for analysis of reverse current–voltage characteristics of 4H–SiC Schottky barrier diodes

Abstract

In this present study, we test the accuracy of the Padovani-Stratton models in terms of the percent error in the barrier height extracted by this model from the reverse characteristics I-V simulated by the Tsu-Esaki model over a range −1000 V at room temperature. The thermionic field emission model without image force barrier lowering is much less accurate for low doping concentration (<10 15 cm −3 ), in particular at low reverse bias voltages (<−200 V). While the field emission model is less accurate for high doping concentrations (>10 17 cm −3 ) in particular for high reverse bias voltages. For medium doping concentration, thermionic field emission or field emission models provide an acceptable accurate for medium and high reverse bias voltages (>−100 V), however, it is less accurate at lower reverse bias voltages (<−100 V) and the best accuracy is occurred at about 10 16 cm −3 . When the Padovani-Stratton models is combined with image force barrier lowering model, the percent error in extracted barrier height takes large values; hence, it is inaccurate on entire range bias.

Electrical properties of inhomogeneous tungsten carbide Schottky barrier on 4H-SiC

“…However, there is no consensus in literature on the model describing the reverse characteristics of 4H-SiC Schottky diodes. In fact, while some papers use the classical TFE model [34,42], whose validity depends on doping concentration and bias range [43], other works include the image force lowering effect in the TFE model to fit the experimental data [37,44].…”

A Compact Model for SiC Schottky Barrier Diodes Based on the Fundamental Current Mechanisms

“…A free electron model is nearly always used in order to describe the metal's electronic structure [2]- [9], [22]- [24]. Many works also use a Wentzel-Kramers-Brillouin (WKB) approximation to simplify the tunneling probability, although the exact shape of the barrier that is modeled varies [2], [3], [5], [7]- [9], [22]- [24].…”

“…These equations, however, often perform poorly for SiC for two reasons; one, the devices are often operated in the intermediate regions between the limiting cases, and two, these models do not account for the image force barrier lowering. The second point is crucially important for SiC [7]. Since the breakdown field is so much larger, the electric field at the metal-SiC surface can be much greater than for Si Schottky diodes.…”

“…To prevent this, we defined upper and lower bounds to E P [see (7)]. If E P reaches one of the bounds, rather than continue to increase or decrease, it instead becomes a fixed value.…”

Conduction mechanisms of the reverse leakage current of β-Ga2O3 Schottky barrier diodes

“…In our recent study [36], we tested the accuracy of the Padovani-Stratton model (Eqs. (6) and (11)) in terms of the percent error in the barrier height over a range -1000 V at room temperature with and without the image force barrier lowering.…”