A revision of the theory of THz detection by MOSFET in the light of the self-mixing model

CMOS technology has been extensively used for the realization of image sensors at Terahertz frequencies. The explanation of its strong efficiency was usually given invoking a mechanism described by the plasma wave detection theory. This model predicts that, when a high frequency potential is applied between gate and source electrodes of a MOSFET, oscillations of the 2D electron gas, located in the inversion layer, converts THz radiation into a DC voltage. Recently, we developed a new model of the self-mixing rectification process occurring in the depleted portion of a semiconductors crossed by a radiofrequency electric field. We studied both the new double barrier structure and the extensively used depleted region in MOS. In this paper, on the light of these new results, we review the theory of the THz detection in a MOS-FET structure. For a comparison with the former approach, we notice that the volume of interaction between free carriers and the RF electric field considered in this model is much higher that the volume considered in the plasma wave model. Technology Computer-Aided Design software simulations, using the Harmonic Balance analysis, will be adopted as evaluation tool. This consideration suggests that self-mixing effect may be more relevant in determining the rectification process. In the authors opinion, this approach substantially improves understanding of the THz rectification in semiconductors and in particular in MOS-FET structures.