Growth and analysis of gallium phosphide on silicon for very high efficiency solar cells
University of Delaware
Photovoltaics research and development is marked by a continual effort to reach higher efficiencies. The highest efficiencies are achievable through the use of multijunction solar cells - devices designed to utilize the solar spectrum to its fullest potential. State-of-the-art multijunction devices are comprised of III-V materials both for the substrates and the junction layers. While these materials provide the high efficiency sought, they are also very expensive. On the other hand, other high efficiency devices utilize silicon (Si) - a material that is highly abundant and inexpensive with a well-developed technological base. The goal of this body of work is the integration of Si and III-V materials in a multijunction system for the realization of high efficiency and increased affordability. This work is comprised of two portions. The first section is the experimental implementation of the growth of gallium phosphide (GaP) on Si through liquid phase epitaxy (LPE) based on previous foundational work. This approach provides unique challenges for the further growth of GaP or other III-V devices on a Si substrate. This body of work addresses the effects of substrate orientation, growth time, temperature, rate, and area, and principles of supersaturation. The second section is a theoretical analysis of two potential dual-junction devices: a GaP solar cell on a Si solar cell and a GaAsP solar cell on a Si solar cell. The results of this work show promise for the future of these devices. The experimental results of this study demonstrate marked improvements in GaP buffer layer quality for subsequent layer growths. Growth procedure optimization steps have led to a reduction of Si concentration from 20% to 3-10%. Additionally, theoretical modeling from a first principles approach shows relative efficiency gain of 20-71% and absolute efficiency gains of 3-13% for devices adding either a GaP junction or a GaAsP junction above a representative Si device. Both the experimental and the theoretical analysis shows that, while there is still work to be done to realize the goal of a high efficiency multijunction device utilizing a Si solar cell as the substrate, there is significant potential for these structures.