Date of Award
Doctor of Philosophy (PhD)
Professor J. Shewchun
This thesis describes the theory of tunnel metal-insulator-semiconductor (MIS) and semiconductor-insulator semiconductor (SIS) solar cells. In both cases, the thickness of the insulating layer is < 40 Å and the current flow has been described by quantum mechanical tunneling process. The base-semiconductor in both cases in single-crystal and non-degenerate.
The introduction of a thin insulating layer [< 16 Å in the case of Al-SiOx-(p-type)Si solar cells] eliminates the pinning effect observed in Schottky barrier diodes, so that the minority carrier quasi-Fermi level is now pinned to the Fermi level in the metal to produce a higher open-circuit voltage. Recent experimental work on oxide-semiconductor/base-semiconductor solar cells indicates that the performance of such devices also could be dramatically controlled by the presence of a thin interfacial layer or insulator. We have developed a mathematical model to calculate the photovoltaic respose of the MIS diodes. The theory of SIS solar cells is an expansion of the MIS solar cell theory, where the metal is now allowed the property of a variable band gap which can be set at any value from zero to several electron volts. The theory is valid when the top semiconductor is an amorphous or heavily doped (degenerate) wide-gap oxide-semiconductor. The SIS model is capable of explaining why some oxide-semi-conductor base - semiconductor solar cells indicated good performance of such heterojunction devices while others show poor performance.
Our work shows that optimized MIS and SIS solar cells can have efficiencies comparable to p-n junction solar cells. The effects of various parameters on the device performance has been studied. The most important parameters are the thickness of the insulating layer and the metal work function.
The theory of tunnel MIS solar cells has been compared with the available experimental data and an excellent agreement is observed. The open-circuit voltage below a certain thickness of insulating layer (SiOx) shows slight deviation from the theory. This experimental observation has been accounted for by the presence of pin-holes associated with ultra-thin oxides and a composite model of Schottky barrier and MIS solar cells has been developed. Based on this model, we have calculated the complex dielectric constant of SiOx as a function of its thickness. Measurement of dark I-V characteristics as a function of temperature confirms the pseudo p-n junction behaviour of MIS solar cells and discards thermionic emission models.
The calculation of SIS solar cells has been compared with the experimental work. The measurement on the temperature effect, intensity effect, insulator thickness effect and the spectral response supports the SIS model. We have also discussed various main loss mechanisms associated with SIS solar cells. This includes the transmission-reflection loss from the top layer, recombination in the depletion later, open-circuit lowering and series and shunt resistance losses. At the end of the thesis, we have shown that simple methods based on the electrical measurements of oxide-semiconductor/base-semiconductor systems often lead to incorrect values for the electron affinity of oxide-semiconductors.
Singh, Rajendra, "Theory of Metal-Insulator-Semiconductor (MIS) and Semiconductor-Insulator-Semiconductor (SIS) solar cells" (1979). Open Access Dissertations and Theses. Paper 679.